Delivery of Biomolecules into Mammalian Cells Using Anthrax Toxin ARCHNES by MASSACHUSETTS INSTITUTE Amy Ellen Rabideau OF TECHNOLOGY B.S. Chemistry and Biology Syracuse University, 2010 NOV 0 9 2015 LIBRARIES Submitted to the Department of Chemistry in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy at the Massachusetts Institute of Technology September 2015 C 2015 Massachusetts Institute of Technology All rights reserved Signature of Author: Signature redacted 1) Certified by: Department of Chemistry ugust 28th, 2015 Signature redacted Bra ley L. Pentelute Pfizer-Laubach Career Development Professor of Chemistry Thesis Supervisor Signature redacted Accepted by: Robert W. Field Haslam and Dewey Professor of Chemistry Chairman, Departmental Committee for Graduate Students 2 This doctoral thesis has been examined by a committee of the Department of Chemistry as follows: Signature redacted Alexander M. Klibanov Novartis Professor of Chemistry and Bioengineering Thesis Committee Chair Signature redacted Bradley L. Pentelute Pfizer-Laubach Career Development Professor of Chemistry Thesis Supervisor Signature redacted John M. Essigmann William R. and Betsy P. Leitch Professor of Chemistry and Biological Engineering Signature redacted Barbara Imperiali Class of 1922 Professor of Chemistry and Biology 3 4 Delivery of Biomolecules into Mammalian Cells Using Anthrax Toxin by Amy Ellen Rabideau Submitted to the Department of Chemistry on August 2 8th, 2015 in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Abstract The intracellular delivery of biomolecules into mammalian cells is a major challenge due to the plasma membrane, which acts as a barrier between the extracellular environment and intracellular components. Recently, a non-toxic delivery platform derived from anthrax lethal toxin has been developed to overcome this challenge for the delivery of biomolecules into the cytosol of mammalian cells. The PA/LFN delivery platform has been used to deliver over 30 known biomolecules of diverse sequences, structures, and functionalities. Collectively, these translocation studies have helped to elucidate the translocation mechanism and to probe intracellular biological processes. In this thesis, the PA/LFN delivery platform was used to analyze the delivery of assorted biomolecules through the PA pore. A facile, modular ligation strategy using sortase A was developed for the conjugation of biomolecules to LFN. The biomolecules for this analysis included antibody mimic proteins with defined sizes and secondary structures, mirror image peptides and proteins, polypeptides containing non-canonical amino acids or small molecule drugs, and cyclic peptides. Our translocation analyses have led to guidelines for translocation as well as insight into design parameters for the efficient delivery of new cargos. The PA/LFN delivery platform has also been used to translocate bioactive cargos for the disruption of intracellular protein-protein interactions (PPI). The translocation efficiency and bioactivity of a tandem monobody to Bcr-Abl, an affibody to hRaf- 1, and a mirror image peptide to MDM2 were analyzed. Efficient translocation and disruption of the intended PPI in each case indicated that the delivery platform could be used to deliver bioactive cargos into cells for therapeutic utility. As an application of this technology, the PA/LFN delivery platform was employed to analyze the intracellular stability of mixed chirality proteins. One major factor that governs a protein's stability is the N-end rule, which states that the N-terminal residue of a protein impacts its intracellular stability through the ubiquitin (Ub)/proteasome system. Utilizing the PA/LFN delivery platform, the stability of proteins containing one N-terminal D-amino acid was analyzed. In contrast to N-terminal L-amino acids, each N-terminal D-amino acid abrogates protein degradation by the N-end rule pathway. Thesis Supervisor: Bradley L. Pentelute Pfizer-Laubach Career Development Professor of Chemistry 5 6 Acknowledgements Throughout my upbringing my dad emphasized the philosophy, "Learning is the journey!" I do not think I entirely understood the true meaning behind what he has been saying all these years until graduate school, where some of the most important lessons are not taught in a classroom. My graduate career has been filled with new experiences, troubleshooting, incredible colleagues, and a lot of really amazing science. I am grateful for everyone who has supported and guided me throughout this phase of my learning journey. I am indebted to my advisor, Prof. Brad Pentelute for taking me on as his first graduate student and exposing me the complexities of biochemistry research. I will never forget walking into our empty lab in 56-546 on my first day and thinking, "What did I get myself into?" Fortunately, in true Pentelute lab style (fast), after only a few months of ordering and organizing, we were up and running-expressing proteins, synthesizing peptides, growing cells, and writing grants. Since the beginning, Brad has encouraged me to think big and explore new ideas with confidence. He helped me realize that simple experiments are often the most successful and the more controls, the better. Under Brad's mentorship, I have developed the intuition to think critically through challenging questions. Thank you to Prof. Alex Klibanov, my thesis chair, for his advice and helpful conversations over the past five years. I also thank Prof. John Essigmann for his encouragement and support during my transition into the Pentelute lab. Thank you to Prof. Barbara Imperiali for advising me during my first MIT experience in Summer 2009 and for her constant support throughout my graduate career. Completing a Ph.D. would be difficult without the help of incredible lab mates. I am so grateful to have developed the translocation project with Dr. Xiaoli Liao during the first two years in the Pentelute lab. Together we expressed the lab's first proteins, worked through troublesome ligations, and got the project off the ground. I learned so much from Xiaoli-team translocation's progress would not have been possible without her help. I also thankful for all the former and current Pentelute lab members for everything they taught me and the fun timesAlex L., Alex M., Alex S., Alex V., Anthony, Chi, Colin, Dale, Dan C., Dan D., Daphne, Ethan, Faycal, Gizem, Guillaume, Hansol, Jingjing, Jun, Justin, Kyle, Mark, Michael, Mike, Peng, Richard, Rocco, Surin, Tatiana, Ted, Tessa, Tuang, Yekui, Yuta, Zak, and Zi-Ning. I am extremely proud of and impressed by the lab's success to date; I am excited to see what the future holds! Graduate school would have been completely different had it not been for my colleagues and classmates. These are the people who shared in the victories, helped troubleshoot the struggles, and listened to the problems: Ali, Katya, Jingnan, Jon, Nootaree, Vinita, Whitney, and especially my biological classmates, Austin, Haritha, Kanchana, and Megan. A special thank you to Austin, a talented scientist and devoted friend, who always had time to help me think through tough problems, to eat lunch at Cosi, or to take a run around the Charles. His enthusiasm for science was infectious and I am forever grateful for all his support. My Yorktown and Syracuse friends have been extremely encouraging and always willing to talk about non-science things over the past five years as I navigated the ups and downs of graduate school: Anna, Dana, Eileen, Emily, Heather, Kaitlyn, and Rachel. My family has been my rock and greatest support system throughout my graduate years. No matter what time of day it is, someone is always available to listen, laugh, or advise. My best friend and sister, Erin, for always knowing exactly how to calm me down and make me smile 7 and for buying me my first yoga mat. My mom, for always having time to talk and listen to me and for teaching me perseverance and dependability. My dad, for the pep talks and wise words before all my big presentations and meetings and for teaching me to always think about what comes after what comes next. Indeed, learning is the journey. 8 Table of Contents Abstract 5 Acknowledgements 7 Table of Contents 9 List of Figures 14 List of Tables 18 Chapter 1: Delivery of Non-Native Cargo into Mammalian Cells Using Anthrax Lethal Toxin 1.1. 18 Introduction 19 1.2. Anthrax Lethal Toxin 23 1.3. Fusion and Conjugation Strategies 24 1.4. Methods to Study Translocation 28 1.5. Delivery of Non-Native Cargos 31 1.5.1. Natural Proteins 31 1.5.2. Engineered Protein Variants 32 1.5.3. Stabilized Protein Conjugates 34 1.5.4. Mirror Image Polypeptides 35 1.5.5. Non-Natural Peptides 36 1.5.6. Cyclic Peptides 37 1.5.7. Small Molecule Drugs 37 1.5.8. Modified Delivery System 38 1.5.9. Retargeting PA 39 1.6. Summary and Outlook 42 1.7. References 43 Chapter 2: Delivery of Antibody Mimics into Mammalian Cells via Anthrax Toxin Protective Antigen 48 2.1. Introduction 49 2.2. Results 53 2.2.1. Delivery of antibody mimic proteins 53 9 2.2.2. Translocation requires a functional PA and endocytosis 54 2.2.3. Western blot analysis of cytosolic fraction 57 2.2.4. Comparison to TAT-mediated delivery 58 2.2.5. Delivery of a tandem monobody for SH2 domain of Bcr-Abl 61 2.2.6. Delivery of an affibody for Raf-1 67 2.3. Discussion 69 2.4. Experimental 74 2.4.1. Materials 74 2.4.2. One-pot sortagging reaction using Staphylococcus aureus SrtA* 74 2.4.3. Protein synthesis inhibition assay 75 2.4.4. Uptake of Lvs or TAT-HA-I to -4 in CHO-KI cells 75 2.4.5. Cytosolic protein extraction and whole cell lysate preparation 76 2.4.6. Western blot 76 2.4.7. Co-immunoprecipitation of Lv5 with Abl kinase 76 2.4.8. TUNEL assay with Lv5 in K562 cells 77 2.4.9. Transfection of HEK 293T with pcDNA3-ABRaf 77 2.4.10. Delivery of Lv6 into HEK 293T cells 78 2.4.11. Construction of plasmids for recombinant proteins and transfection 78 2.4.12. Protein expression and purification 79 2.4.13. Synthesis of TAT peptide and native chemical ligation (NCL) of TAT-DTA 80 2.4.14. Synthesis of TAT-HA-LPSTGG peptide and sortagging to G5 -proteins 80 2.5. Acknowledgements 81 2.6. Appendix 82 2.6.1. LC-MS Traces 2.7. References Chapter 3: Delivery of Mirror Image Polypeptides into Cells 10 101 106 110 3.1. Introduction 111 3.2. Results 113 3.2.1. Delivery of mirror image polypeptides 113 3.2.2. Translocation requires a functional PA and endocytosis 116 3.2.3. Translocation of a D-peptide for MDM2 118 3.2.4. Translocation of mirror image proteins 122 3.3. Discussion 125 3.4. Experimental 126 3.4.1. Materials 126 3.4.2. Solid phase peptide synthesis (Boc) 127 3.4.3. Solid phase peptide synthesis (Fmoc) 127 3.4.4. Analytical LC-MS 128 3.4.5. Preparative, semi-preparative, and analytical RP-HPLC 128 3.4.6. Synthesis of D-affibody 129 3.4.7. Synthesis of D-affibody-alkyne 131 3.4.8. Synthesis of D-GB1 131 3.4.9. Circular dichroism (CD) spectroscopy of folded proteins 133 3.4.10. Synthesis of biotinylated p53/MDM2 Inhibitor D-peptide 133 3.4.11. Construction of plasmids for recombinant proteins 134 3.4.12. Protein expression and purification 134 3.4.13. One-pot sortagging reaction using Staphylococcus aureus SrtA* 135 3.4.14. Translocation of SDvs and protein inhibition assay in CHO-Ki cells 136 3.4.15. Cytosolic protein extraction and whole cell lysate preparation for western blot 137 3.4.16. Western Blot of Svl, 2, 4, and 5 translocated into CHO-KI cells 137 3.4.17. Trypsin digestion of L- and D-Affibody 138 3.4.18. Pull down of Sv4-biotin in U-87 MG cells 138 3.4.19. Translocation of Sv4 and Sv4-biotin in U-87 MG or K562 cells 138 3.4.20. Binding interaction between SUMO-2 5- o 9MDM2 and 4 or Sv4 139 3.5. Acknowledgements 140 3.6. Appendix 141 3.6.1. LC-MS Traces 3.7. References 157 160 Chapter 4: Translocation of Non-Canonical Polypeptides into Cells Using Protective Antigen 163 11 4.1. Introduction 164 4.2. Results 168 4.2.1. Translocation of non-canonical polypeptide cargos with backbone or side chain modifications 168 4.2.2. Translocation of cyclic peptides 171 4.2.3. Translocation of complex small molecules 173 4.2.4. Translocation of intact cargo 176 4.3. Discussion 179 4.4. Experimental 181 4.4.1. Materials 181 4.4.2. 'H Nuclear magnetic resonance (' H NMR) 182 4.4.3. Synthesis of docetaxel-maleimide 182 4.4.4. Synthesis of doxorubicin-maleimide 183 4.4.5. Solid phase peptide synthesis (Boc) 184 4.4.6. Solid phase peptide synthesis (Fmoc) 184 4.4.7. Protein expression and purification 185 4.4.8. Sortase-mediated ligation 186 4.4.9. LC-MS analysis 186 4.4.10. Preparative, semi-preparative, and analytical RP-HPLC 187 4.4.11. Conjugation of docetaxel, doxorubicin, and MMAF to G 5-LRRLRAC 187 4.4.12. Cyclization of linear peptide using native chemical ligation 188 4.4.13. Protein synthesis inhibition assay 189 4.4.14. Translocation of LDn1-1 I and cytosolic and total cell lysate extraction 190 4.4.15. Western blot of extracted material 190 4.5. Acknowledgements 191 4.6. Appendix 192 4.6.1. LC-MS Traces 4.7. References 200 204 Chapter 5: A D-Amino Acid at the N-terminus of a Protein Abrogates its Degradation by the N-End Rule Pathway 12 206 5.1. Introduction 207 5.2. Results 209 5.2.1. Sortase A Attaches One D-Amino Acid onto the N-Terminus of LFN-DTA 209 5.2.2. One N-Terminal D-Amino Acid Stabilizes LFN-DTA to Proteasomal Degradation 209 5.2.3. Proteasomal Stabilization is Not an Artifact of the Sortag 212 5.2.4. LFN-DTA with One N-terminal D-Amino Acid is Stable In Vitro 214 5.2.5. LFN-DTA with One N-terminal D-Amino Acid is Not Ubiquitinated 214 5.2.6. N-terminal Stabilization is Not Protein-Specific 216 5.2.7. RRSPc 2 is a Ras/Rap 1-Specific Endoprotease 219 5.2.8. EGFR-Targeted Delivery of RRSPc 2 Interrupts the MAPK Pathway 219 5.3. Discussion 222 5.4. Experimental 225 5.4.1. Materials 225 5.4.2. Protein Expression and Purification 225 5.4.3. Fmoc Solid Phase Peptide Synthesis 226 5.4.4. Sortase A-Mediated Ligation of X-LFN-DTAut Constructs 227 5.4.5. Protein Synthesis Inhibition Assay with X-LFN-DTAmut Constructs 228 5.4.6. Translocation and Western Blot Analysis with X-LFN-DTAmut Constructs 228 5.4.7. Native Chemical Ligation of Native X-LFN-DTAmut 229 5.4.8. In Vitro Stability of X-LFN-DTAmut Constructs 230 5.4.9. Streptavidin Pulldown of Ubiquitinated Constructs 231 5.4.10. Stabilization of X-DTAmut or X-DARPin after Translocation 231 5.4.11. EGFR-Targeted Translocation of RRSPc 2 233 5.5. Acknowledgements 234 5.6. Appendix 235 5.6.1. LC-MS Traces 5.7. References 250 265 13 List of Figures Figure 1.1.1. Anthrax lethal toxin is comprised of two discreet components. 21 Figure 1.1.2. Translocation of biomolecules using PA/LFN delivery platform. 22 Figure 1.3.1. Modular chemistry to modify LFN or LFN-DTA at N- or C-terminus. 27 Figure 1.4.1. Representative assay results from common translocation assays. 30 Figure 1.5.1. Translocation efficiency has been analyzed for various peptides and proteins. 40 Figure 2.1.1. Delivery of antibody mimics into the cytosol by the PA/LFN system. 52 Figure 2.2.1. Control experiments validating the translocation mechanism of LDv1 and Lvl. 56 Figure 2.2.2. TAT peptide mediated translocation of DTA and antibody mimics. 60 Figure 2.2.3. Delivery of Lv5 and binding of Lv5 to Abl kinase in K562 cells. 63 Figure 2.2.4. Monitoring of apoptosis of K562 cells treated with Lv5 and PA by TUNEL assay. 66 Figure 2.2.5. Perturbation of the MAPK signaling pathway by PA mediated delivery of an affibody (Lv6) that targets Raf. 68 Figure 2.6.1. One-pot sortagging reaction. 82 Figure 2.6.2. Coomassie stained SDS-page gel of LDvs and Lvs obtained after sortagging and purification. 83 Figure 2.6.3. Thermal stability of 1 OFN3 (4) and HA4 (4mut) was monitored by circular dichroism (CD) spectroscopy. 84 Figure 2.6.4. Western blot analysis of delivered Lvs. 85 Figure 2.6.5. Coomassie stained SDS-PAGE gel of TAT-HA-I to -4 and -4mut (1 pg). 86 Figure 2.6.6. Cellular uptake of TAT-HA-I to -4 and -4mut. CHO-KI cells. 87 Figure 2.6.7. The level of protein synthesis inhibition in K562 cells. 88 Figure 2.6.8. SPR curves for SUMO-SH2 and varying concentrations of Lv5 or HA4-7c 12. 89 Figure 2.6.9. Linear relationship between signal intensity of each band and the amount of protein loaded. 90 Figure 2.6.10. Phosphorylation analysis of Bcr-Abl. 91 Figure 2.6.11. Apoptosis measurement of K562 cells. 92 Figure 2.6.12. MTS cell viability assay. 93 Figure 2.6.13. Protein sequences. 95 14 Figure 3.2.1. Delivery of D-cargo sortagged onto LFN and LFN-DTA. 114 Figure 3.2.2. Translocation of mirror peptides using PA/LFN- 117 Figure 3.2.3. Translocation of a D-binder to MDM2. 121 Figure 3.2.4. Synthesis and translocation of mirror image proteins. 124 Figure 3.6.1. Immunoblot of media from CHO-KI cells treated with Svl-3. 146 Figure 3.6.2. Linear relationship between Sv4 band intensity and the amount of protein loaded. 147 Figure 3.6.3. Calibration curve for the interaction between immobilized biotin- -29 p53 and SUMO-25- 09MDM2. 148 Figure 3.6.4. Binding interaction between SUMO- 2 5- ' 09MDM2 and 4 or Sv4. 149 Figure 3.6.5. Quantification of MDM2, p53, and p21 protein levels for U-87 MG cells. 150 Figure 3.6.6. LC-MS characterization of D-affibody synthesis. 151 Figure 3.6.7. LC-MS characterization of D-GB 1 synthesis. 152 Figure 3.6.8. CD of L- and D-affibody and L- and D-GB 1. 153 Figure 3.6.9. Trypsin digestion of L-affibody, D-affibody, Sv5-L, and Sv5-alkyne over time. 155 Figure 3.6.10. LC-MS characterization of D-affibody-alkyne and D-affibody-biotin. 156 Figure 4.1.1. Delivery of non-canonical polypeptide cargo into the cytosol. 167 Figure 4.2.1. Translocation of non-canonical peptides. 170 Figure 4.2.2. Translocation of cyclic peptides. 172 Figure 4.2.3. Translocation of small molecules. 175 Figure 4.2.4. Translocation of C-terminally biotinylated cargo. 178 Figure 4.6.1. Cyclization of L-linear peptide using native chemical ligation. 196 Figure 4.6.2. Synthesis of doxorubicin-maleimide and docetaxel-maleimide. 197 Figure 4.6.3. Western blot of total extraction of LDn1-8. 198 Figure 4.6.4. Western blot of total extraction of LDn9- 11. 199 Figure 5.2.1. Intracellular stability was monitored for X-LFN-DTA constructs delivered through protective antigen pore. 210 Figure 5.2.2. One N-terminal D-amino acid on LFN-DTA enhances protein stability. 213 Figure 5.2.3. One N-terminal D-amino acid prevents ubiquitination of LFN-DTA. 215 Figure 5.2.4. N-terminal D-amino acid stabilization is not limited to LFN. 218 15 Figure 5.2.5. Precision delivery of stabilized RRSPc 2 through EGFR into pancreatic cancer cells interrupts the MAPK pathway. 221 Figure 5.6.1. Translocation of X-LFN-DTAmUt constructs. 235 Figure 5.6.2. Western blot analysis of X-LFN-DTAmut constructs in CHO-KI cells. 237 Figure 5.6.3. Western blot analysis of X-LFN-DTAmut constructs delivered into HEK-293T or HeLa cells. 238 Figure 5.6.4. Western blot analysis of sortagged and native X-LFN-DTAmut constructs. 239 Figure 5.6.5. In vitro degradation of X-K(bio)-LFN-DTAmut. 240 Figure 5.6.6. Translocation of X-K(bio)-LFN-DTAmut into CHO-Kl cells. 241 Figure 5.6.7. LFN-C* X-cargo-C is the oxidation product of LFN-C* and X-G 5-cargo-C- Ellman's. 242 Figure 5.6.8. Rate of reduction for 1-X. 243 Figure 5.6.9. Delivery of 1-X. 244 Figure 5.6.10. Translocation of LFN-DTA-RRSPc 2 . 245 Figure 5.6.11. SDS-PAGE analysis of in vitro cleavage of KRas by RRSPC 2 conjugates. 246 Figure 5.6.12. LC-MS analysis of in vitro cleavage of KRas by RRSPc 2 conjugates. 247 Figure 5.6.13. Targeted delivery of LFN-DTA through EGFR. 248 Figure 5.6.14. Quantification of pErkl/2 levels in 3-X-treated AsPC-1 cells. 249 List of Tables Table 1.5.1. More than 30 different non-native cargos have been delivered into the cytosol of cells. Table 2.6.1. PCR primers. 41 96 Table 2.6.2. Observed molecular masses of expressed protein constructs when analyzed by LC- MS. 97 Table 2.6.3. Isolated yields of sortagging ligations from SrtA* reaction. 98 Table 2.6.4. EC5 0 values of 30-minute protein synthesis inhibition assay. 99 Table 2.6.5. List of variants. 100 Table 3.6.1. Peptides used in this investigation. 141 16 Table 3.6.2. Observed molecular masses of expressed protein constructs when analyzed by LCMS. 142 Table 3.6.3. List of variants. 143 Table 3.6.4. Isolated yields of sortagging ligations from SrtA* reaction. 144 Table 3.6.5. EC50 values of 30-minute protein synthesis inhibition assay. 145 Table 4.6.1. Peptides used in this investigation. 192 Table 4.6.2. List of variants. 193 Table 4.6.3. EC 50 values of 30-minute protein synthesis inhibition assay. 195 Table 5.6.1. EC50 values for X-LFN-DTAut constructs translocated in CHO-Ki cells. 236 17 Chapter 1: Delivery of Non-Native Cargo into Mammalian Cells Using Anthrax Lethal Toxin 18 1.1. Introduction Pathogenic bacteria often express protein toxins capable of delivering cytotoxic payloads into the cytosol of cells.' The cytotoxic payloads are often referred to as effector proteins and have diverse functionalities in the host cell-protease activity, modification of intracellular substrates, or interruption in cell signaling pathways-for the benefit of the bacterium. One example is anthrax lethal toxin from the gram-positive bacterium, Bacillus anthracis,which has been extensively studied for the past forty years. Thorough biophysical and biochemical analyses have led to an increased understanding of each component as a discreet protein and together as a macromolecular nanomachine. 2 Anthrax lethal toxin has evolved to deliver lethal factor (LF), a cytotoxic protein payload, into the cytosol of mammalian cells (Figure 1.1.1).3 To accomplish this transport, anthrax lethal toxin utilizes a second component called protective antigen (PA), which oligomerizes on the cell surface to form the PA pre-pore. Following endocytosis the PA pre-pore forms a ~-12 A channel in the endosomal membrane and acts as a conduit for delivery of LF into the Cytosol.4-6 Once in the cytosol, LF is a Zn protease that cleaves mitogen activated protein kinase kinases (MAPKK) and causes cell death.7 While native anthrax lethal toxin expressed by B. anthraciscontinues to be a bioterrorism threat, in the last two decades the toxin has been modified to serve as a delivery platform for the transport of biomolecules. The PA/LFN delivery platform consists of PA and the non-toxic, Nterminal PA-binding domain of LF known as LFN (Figure 1.1.2). The delivery of cargos other than LF (i.e. non-native) such as enzymes, polypeptides comprised of non-natural amino acids, or small molecule drugs has been explored. Fusions of non-native cargos composed of canonical amino acids to LFN have been achieved through recombinant expression. Non-recombinant 19 methods of ligation like native chemical ligation (NCL) or enzyme-mediated ligation have been utilized to study the translocation of cargos containing non-natural functionalities. This review provides an in-depth analysis of the PA/LFN delivery platform for the translocation of non-native cargos into the cytosol of cells. Most notably, the PA pore has been used to deliver more than 30 proteins and polypeptides containing natural and non-natural amino acids as well as small molecule drugs. Collectively, these analyses provide insight into the promiscuity of the PA pore and demonstrate its potential for the delivery of bioactive cargos to disrupt intracellular protein-protein interactions or to study biological processes. Furthermore, these studies support the current model of protein translocation through the PA pore and provide insight into design parameters for delivering new cargos efficiently. 20 a. PA6 LF PA 8 Anthrax receptor Hnosm cytosol 4 -{MAPKK b. catalytic domain C. d- LFN PA2 2PAI 3 PA-binding domain (LFN) 4 e. 90* PA pore Phe clamp Figure 1.1.1. Anthrax lethal toxin is comprised of two discreet components. a. Cytosolic delivery via anthrax lethal toxin is achieved by anthrax receptor recognition by PA83 then activation to form PA6 3 by a cell-surface furin family protease (1). Seven or eight PA 63 molecules self-assemble to form the PA pre-pore (2) then LF binds (3) and the entire complex is endocytosed into the endosome (4). Acidification triggers pore formation and translocation of LF into the cytosol (5). b. Crystal structure of LF reveals two domains (pdb: 1J7N)-PA binding domain (green; LFN) and catalytic domain (gray) c. PA83 is composed of 4 domains (blue domain I initiates pre-pore formation, pink domain 2 and orange domain 3 are involved in forming the pore itself, and purple domain 4 is the receptor-binding domain) (pdb: 1ACC) d. LFN binds to two adjacent PA subunits (pdb: 3KWV) e. PA pore structure reveals the restrictive Phe clamp (within domain 2) at the center of the pore (pdb: 3J9C). 21 cargo PA 83 Anthrax receptor PA 63 cytosol 2LFN H+ endosome 4 Figure 1.1.2. Translocation of biomolecules using PA/LFN delivery platform. The PA/LFN delivery platform was developed such that the catalytic domain of LF could be replaced with various cargo molecules to determine their translocation efficiency through PA pore or to analyze their biological function in the cytosol of cells. 22 1.2. Anthrax Lethal Toxin Anthrax lethal toxin is a two-component system in which lethal factor (LF; 90 kDa) is transported into the cytosol of a host cell through protective antigen (PA83 ; 83 kDa) pore.2 After nearly forty years of mechanistic and structural analyses, a model for protein translocation via anthrax lethal toxin has emerged (Figure 1.1. la). In the presence of a divalent metal ion, PA83 is recognized by and binds to either of two cell surface receptors, tumor endothelial marker 8 or capillary morphogenesis protein 2 (TEM8 or CMG2) with nanomolar or picomolar affinity, respectively. 8-' 0 The 1000-fold disparity in binding affinity has been attributed to non-conserved residues of CMG2 that interact with domain 2 of PA." While the exact function of each receptor ' remains unknown, both TEM8 and CMG8 have been shown to regulate angiogenic processes.1 2 13 Furthermore, the anthrax receptors are expressed on most human cells at approximately 2,000 - 50,000 receptors per cell.' 4 Once PA83 (Figure 1.1.1b) is receptor-bound, a furin family protease proteolytically activates the protein by cleaving the N-terminal 20 kDa portion, leaving PA6 3 to oligomerize into the PA pre-pore heptamer or octamer.' 15-1" The PA pre-pore is capable of binding up to three or ' four molecules of LF (Figure 1.1.1 c) between two adjacent PA 6 3 subunits with 1-2 nM affinity.' 8 19 The entire complex is endocytosed and encapsulated in an endosome. Acidification of the endosome (pH -5.5) results in a conformational rearrangement of the PA6 3 subunits, which leads to PA pore formation (~12 A diameter) in the endosomal membrane. 2 0, 21 The pH gradient generated between the two compartments leads to translocation of protonated LF into the cell cytosol (pH ~7.0). Translocation through PA pore is considered to be a charge state-dependent Brownian ratchet motion.22' 23 23 Biochemical and biophysical studies have demonstrated that the structural components of anthrax lethal toxin relate to protein translocation. Mutagenesis analyses, kinetic studies, computational models, and a recent crystal structure have revealed how the N-terminal, PA 24-26 4 LF; LFN) interacts with two adjacent subunits of the PA pre-pore.' 9 ' binding domain of LF (1-1 Specifically, there are key electrostatic and hydrophobic interactions between the first a- helix and P-sheet of LFN and a deep amphipathic cleft on the surface of PA (alpha clamp) that bind LF such that the N-terminal region of the protein is partially unfolded and poised for translocation from N- to C-terminus (Figure 1.1.1d).19, 27 A recent 2.9 A resolution cryogenic electron microscopy structure was solved for the PA pore that supports the current model for protein translocation (Figure 1.1.1e).2 The structure confirmed mutagenesis studies that predicted a narrow ring of solvent-exposed Phe427 residues in the lumen of the channel. 2 9 ,3 0 The ring of Phe residues, called the Phe clamp, is the most restrictive part of the pore and interacts with hydrophobic stretches. The structure also revealed negatively charged residues surrounding the Phe clamp, which are hypothesized to deprotonate the translocated protein and guide unidirectional translocation. 1.3. Fusion and Conjugation Strategies Prior to the elucidation of the LF and PA structures, researchers utilized protein fusions to explore the translocation mechanism and to analyze the biological function of bioactive payloads inside the cytosol. The earliest work relied on recombinant expression to create fusions with the catalytic domains of select protein toxins; however, recombinant expression is not adequate for probing the delivery of cargos containing non-natural functionalities such as amino acids with inverted chirality, non-canonical side chain residues, or modified backbone structures. To work with cargos containing non-natural amino acids, semisynthetic and enzymatic 24 techniques like native chemical ligation (NCL) and enzyme-mediated ligation using sortase A (SrtA) have been utilized. For protein fusions comprised of the 20 canonical amino acids, recombinant expression is often the simplest and most high yielding technique (Figure 1.3.la). However, proteins containing natural amino acids are susceptible to proteolysis and proteasomal degradation in the cytosol. Advantages of incorporating non-natural functionalities into protein fusions include stabilization to intracellular degradation, use of affinity handles, and perturbed binding affinities to target molecules. Pentelute, et al. developed a semisynthetic approach using NCL to ligate non-natural peptides onto the N-terminus of LFN in order to explore the specificity of PA with regard to translocation initiation (Figure 1.3.1 b). 3 ' While NCL facilitates the site-specific ligation of peptides, oftentimes it is performed under denaturing conditions to achieve optimal substrate concentrations. As a result, ligated products must be purified by RP-HPLC, followed by refolding. While LFN has been found to refold well, some cargos may not refold properly, resulting in inactivity.3 1 Enzyme-mediated ligation using SrtA has allowed for the specific attachment of nonnatural biomolecules under non-denaturing conditions. The catalytic domain of SrtA from Staphylococcus aureus has been demonstrated to be useful for in vitro ligations. As a cysteine protease and transpeptidase, SrtA ligates substrates containing an N-terminal pentaglycine tag onto molecules containing the LPXTG motif (Figure 1.3. 1c). Chen, et al. recently evolved SrtA enzymes to have -50-140-fold increase in LPETG substrate coupling activities such that ligation reactions can be carried out in less than one hour at physiological conditions with high yields.34 SrtA-mediated ligation has been used to attach antibody mimic proteins and peptides containing 25 non-natural functionalities on the C-terminus of LFN and LFN-DTA as well as mixed chirality peptides to the N-terminus of LFNRecombinant expression, NCL, and enzyme-mediated ligation each install a natural, amide linkage between LFN and the cargo of interest. While this is critical for initial analysis of translocation efficiency, there are a variety of other bioconjugation techniques that can be employed to fuse cargos onto LFN- Common examples include the genetic code expansion technique, disulfide linkage, maleimide conjugation, or click chemistry.3 5 Furthermore, combinations of techniques can be used to produce the desired construct. Recently, Ling, et al. demonstrated that SrtA can accommodate peptide thioester substrates, which can be subsequently used for NCL to form protein conjugates separated by non-natural sequences such as LFN-DTA in which the two proteins were joined together by a D-peptide linker. 3 6 26 a. Recombinant Expression E. coil Expression Expression E vector E Oil Expression vectorW Expression Eci vector W - > b. Native Chemical Ligation HS S'R H X L H 2N + 0 H2 N 0 I 0 + H2N CO 2 H HS - AS'R + U SHH NN CO 'H 0 S, R + * ; H2N f c. Enzyme-Mediated Ligation ORtA -LPSTGG -LPSTG 5 - Ca2 SrtA -LPSTGG + Gso -LPSTG +5-GSrtA +aG24% tLPSTG 5 - + -LPSTGG SrtA *LPSTGG + G5- LPSTG A Figure 1.3.1. Modular chemistry to modify LFN or LFN-DTA at N- or C-terminus. a. Recombinant expression in bacteria (e.g. E. coli) serves as a straightforward method to obtain fusion proteins containing canonical amino acids b. A semisynthetic platform using native chemical ligation (NCL) allows for the ligation of N-terminal cysteine biomolecules with Cterminal thioester biomolecules activated using mercaptophenylacetic acid (MPAA) c. Enzymemediated ligation using sortase A (SrtA) has been developed for the facile ligation of any cargo to LFN or LFN-DTA under non-denaturing conditions. 27 1.4. Methods to Study Translocation The delivery of protein and peptide cargos with non-natural functionalities has provided insight into the specificity of the PA pore, efficiency of the PA/LFN delivery platform, and a deeper understanding of the mechanism for protein translocation. Translocation efficiency of cargos at the C-terminus of LFN (or LFN-DTA) has been analyzed using three common laboratory assays: planar lipid bilayer,37 protein synthesis inhibition or cytotoxicity analysis based on enzymatic activity, 38, 39 and western blot analysis of the delivered material 40 . Each approach has provided new insight into key features of translocation. As an in vitro assay, planar lipid bilayer has been used to measure the change in ion current after LFN constructs have been added to a chamber containing PA pore embedded in an artificial membrane. Electrophysiological measurements from the bilayer experiments have revealed important features of translocation initiation and the mechanism of delivery through the PA pore. Cell-based, enzymatic assays have been developed using reporter protein cargos, providing a sensitive measure of translocation into eukaryotic cells. One specific enzymatic assay relies on the activity of the A chain of diphtheria toxin (DTA), which inactivates ' elongation factor 2 (EF-2) through ADP ribosylation and inhibits cytosolic protein synthesis.4 Also known as the protein synthesis inhibition assay, this assay utilizes the activity of DTA to monitor the delivery of assorted cargos fused to LFN-DTA by 3H-Leu incorporation (Figure 1.4.1 a). The Pseudomonas exotoxin A (PE) catalytic domain, which inhibits protein synthesis by a similar mechanism as DTA, has also been used as a reporter protein.4 1 Furthermore, cytotoxicity assays using tetrazolium salts like MTT have been used to measure the translocation efficiency of toxic enzymatic domains. Western blot analysis has been utilized as a third method to measure translocation efficiency. After translocation, the cytosolic fraction of cells is lysed 28 using digitonin, a non-ionic detergent used to permeabilize the cytosolic membrane, then analyzed by western blot (Figure 1.4.1b). 42 Immunostaining for LF, DTA, or biotin (if present) provides a semi-quantitative measure of translocated material. Further development of more sensitive assays is underway in order to precisely quantify the amount of material that has been translocated into the cytosol. 29 a. c 1.41.21.0. 0.8 -a-LF N-DTA -.- LF N-DTA, No PA o- 0.6- 8 -45 0.4- u- 0.2 0.0- -14 -13 -12 -1 -1 b -9 -8d -7-Y Log [Protein Concentration (M)] b. PA only PA + LFN anti-LF anti-Erkl/2 anti-Rab Figure 1.4.1. Representative assay results from common translocation assays. a. Protein synthesis inhibition assay measures the activity of DTA fused to LFN after translocation. After incubating cells with LFN-DTA, 3H-Leu is added to monitor the extent of protein synthesis inhibition caused by DTA. EC50 values of conjugates can be compared with the LFN-DTA control to determine translocation efficiency. b. Western blot analysis of the cytosolic fraction measures the amount of material delivered into the cytosol. Cytosolic extraction is achieved using a lysis buffer containing digitonin. 30 1.5. Delivery of Non-Native Cargos 1.5.1. Natural Proteins Early experiments of protein translocation through PA pore monitored the delivery of protein fusions as a method to characterize the structural requirements of initiating and sustaining translocation. These studies utilized recombinant expression to generate protein fusions of LFN with the A chain of diphtheria toxin (DTA), 3 8 , 19,4 Pseudomonas exotoxin A (PE), 44 A chain of Shiga toxin (STA), dihydrofolate reductase (DHFR),43 and P lactamase (Figure 1.5.1a). 4 5 All together these studies demonstrated that regions near the N-terminus of LFN are important for initiating translocation into eukaryotic cells. 27 The proteins however, must be unfolded or in an extended conformation for efficient translocation, which was verified by the impeded translocation of LFN-DTA containing an artificial disulfide (Figure 1.5.1b) and LFN$ DHFR bound to methotrexate.4 3 The efficient translocation of DTA, STA, PE, DHFR, and lactamase provided the first indication that the PA pore can accommodate non-native protein cargos. After translocation, the measured enzymatic activity for each protein cargo indicated that these proteins were folded correctly in the cytosol and recognized their intracellular substrates. Since the development of the PA/LFN delivery platform, various protein cargos have been delivered into cells to understand their biological function. Examples include a cytotoxic T lymphocyte epitope from Listeria monocytogenes listeriolysin 0 (LLO), 46' 47 Legionella pneumophila flaggellin protein,48 actin cross-linking domain (ACD) of RTX from Vibrio cholerae,4 9 Rho inactivation domain (RID) from Vibrio cholerae50 and a Ras/Rapl -specific endopeptidase (RRSP) from Vibrio vulnificus. Delivery of a cytotoxic T lymphocyte LLO epitope was analyzed for its immune response activity in mice as a potential method for developing vaccines against pathogens. 6 The inflammasome plays a major role in the innate 31 immune response. In order to study the biochemical and physiological consequences of inflammasome stimulation, von Moltke, et al. delivered a Legionella pneumophila flagellin protein to stimulate the inflammasome and monitored eicosanoid release. 48 Satchell and coworkers have utilized the PA/LFN delivery platform to study the functions of various effector domains from multifunctional autoprocessing repeats in toxin (RTX) domains (MARTX) expressed by pathogenic bacteria. The bioactivity of the actin cross-linking domain (ACD) from Vibrio cholerae was analyzed after translocation into HEp-2 cells. Cordero, et al. demonstrated that ACD directly catalyzes the covalent cross-linking of actin providing insight into the mechanism of cell death caused by Vibrio cholerae.49 Sheahan, et al. analyzed the inactivation of Rho GTPases by the Rho Inactivation Domain (RID) from Vibrio cholera after delivery into HEp-2 cells using the PA/LFN delivery platform.50 The bioactivity of a toxic domain within MARTX from Vibrio vulnificus whose function was previously unknown was characterized using the PA/LFN delivery platform. Antic, et al. delivered the domain into HeLa cells and demonstrated that the Ras/Rapl -specific endopeptidase (RRSP) effector domain cleaves the Switch I region of Ras and Rapi proteins and thus interferes with downstream signaling in the MAPK pathway.5 1 ,5 2 1.5.2. Engineered Protein Variants Antibodies serve as powerful tools for medical diagnostics and the treatment of disease. The use of antibodies, however, is limited to outside the cell due to their inability to cross the plasma membrane into the cytosol and the presence of disulfide crosslinks. Recently, singledomain, cysteine-free scaffold proteins have been developed as antibody mimics. The scaffolds include monobody from the tenth type III domain of human fibronectin (1OFN3), 5 3 ,5 4 affibody from immunoglobulin binding protein A,55' 56 designed ankyrin repeats protein (DARPin), 32 and the B1 domain of protein G (GBI)." Researchers have recently analyzed the translocation efficiency of these scaffolds, which can be evolved to bind extracellular receptors or intracellular 59 targets with high affinity. ' 60 Liao, et al. recently analyzed the delivery and intracellular activity of select affibody and monobody antibody mimics using the PA/LFN delivery platform. 59 Specifically, the researchers analyzed the delivery and associated bioactivity of a tandem monobody designed by Koide and 61 62 co-workers to bind the Src homology 2 domain (SH2) of Bcr-Abl (HA4-7cl2; KId 12 nM) , and an affibody designed to bind Raf (ABRaf; Kd 100 nM) developed by Nygren and co- workers. 63 The HA4-7c12 tandem monobody conjugate was translocated into chronic myeloid leukemia cells (K562) and intracellular binding to Bcr-Abl was confirmed by co- immunoprecipitation. The inhibition of Bcr-Abl kinase activity and induction of apoptosis was observed by TUNEL staining, which detects DNA fragmentation by labeling the termini. The ABRaf affibody conjugate was translocated in human embryonic kidney 293T (HEK 293T) cells and the interruption of the mitogen activated protein kinase (MAPK) pathway was monitored. ABRaf was found to significantly reduce the phosphorylation levels of Erkl/2 after activation with epidermal growth factor (EGF). Taken together, the HA4-7c12 and ABRaf delivery data indicate that antibody mimics can be efficiently delivered into the cell cytosol through the PA/LFN delivery platform, refold after translocation, and perturb intracellular PPIs. The GB1 and DARPin antibody mimics have been shown to translocate efficiently into the cytosol of cells. 59 These scaffolds have yet to be delivered into cells using PA pore to perturb PPIs; however, the delivery of DARPin constructs with different thermostabilities has been analyzed. Pl0ckthun and co-workers recently demonstrated that very stable DARPin constructs 33 (i.e. >90'C melting temperatures) cannot translocate efficiently through the PA pore. 60 The researchers addressed the thermostability of DARPin by engineering constructs with reduced stability, amenable for delivery via PA pore. A similar observation was made for 1OFN3 (unpublished results) in which wild-type 1OFN3 (Tm ~88C) translocated less efficiently than HA4, a mutant 1OFN3 construct with Tm~-75'C. Together, these observations indicated that the PA pore requires destabilization of the cargo for efficient translocation. 1.5.3. Stabilized Protein Conjugates The ubiquitin (Ub)/proteasome system is responsible for the majority of protein degradation in the cell. The N-end rule described by Varshavsky and co-workers states that the N-terminal amino acid of a protein impacts the protein's intracellular stability with regard to proteasomal degradation. Since LFN follows the N-end rule, strategies have been developed to increase cytotoxic activity and decrease immunogenicity of the translocated protein conjugates. 64 Leppla and co-workers demonstrated that reductive methylation to dimethylate the epsilon amino group of lysine residues improves cytoxicity of LFN-PE conjugates by stabilizing the proteins to intracellular degradation.65 LFN-PE conjugates including those prone to degradation (i.e. contain N-terminal His residue) were reductively methylated at all 36 lysine residues using borane dimethyl amine and formaldehyde. After translocation into several eukaryotic cell lines, cytotoxicity and western blot assays revealed each methylated conjugate was stabilized to degradation. A disadvantage of this approach is non-specific methylation, which means that for substrates requiring lysine for catalysis or structural integrity, reductive methylation cannot be used for stabilization. 34 Stabilization of proteins using one N-terminal D-amino acid was recently demonstrated using the PA/LFN system (Chapter 5). Incorporation of N-terminal D-amino acids was achieved through SrtA ligation or NCL for a native N-terminus. Proteasomal degradation of LFN-DTA constructs containing L- or D-amino acids was screened using the protein synthesis inhibition assay as well as western blot analysis. In both assays, while constructs containing N-terminal Lamino acids followed the N-end rule, constructs with N-terminal D-amino acids were stabilized to degradation. We demonstrated that a protein containing one N-terminal D-amino acid is not ubiquitinated. In order to prove that this phenomenon was not protein-specific, a hindered disulfide cleavable linker was incorporated and similar observations of protein stabilization were made for DTA, DARPin, and RRSPC 2 . This work updated the N-end rule to include D-amino acids as stabilizing amino acids. 1.5.4. Mirror Image Polypeptides The biological impact of polypeptides containing amino acids with non-natural side chains, backbone composition, mirror image chirality and many others remains relatively unexplored.66 Biomolecules composed of mirror image amino acids are of particular interest for their unique biological stability and reduced immunogenicity. The translocation efficiency of mixed chirality fusions to LFN through PA pore was recently analyzed.6 7 Protein synthesis inhibition assays indicated that mirror image polypeptides and proteins translocate as efficiently into the cell cytosol as their L-counterparts. Western blot analysis indicated that an unstructured L-peptide cargo on the C-terminus of LFN resulted in rapid degradation of the protein conjugate in the cell cytosol; however, a conjugate containing a D-peptide cargo of the same sequence was found to be stabilized to degradation. Capping the C-terminus with two D-amino acids (D-cap) 35 stabilized the L-peptide cargo. Evidently, the incorporation of a short D-cap at the C-terminus of an unstructured polypeptide can provide stabilization to intracellular degradation. Delivery of stable, bioactive cargos into the cell cytosol is an attractive application for translocation. Using the PA/LFN delivery platform, a D-peptide MDM2 antagonist (TAWYANF*EKLLR, where F* is p-CF 3-D-Phe)68 was recently delivered into the glioblastoma U-87 MG cell line, which overexpresses MDM2.67 A biotinylated form of the LFN conjugate (Kd 12.3 4.3 nM) towards MDM2 was found to interact with MDM2 after delivery into U-87 MG cells. Moreover, after delivery into U-87 MG cells, the conjugate was found to disrupt the p53/MDM2 pathway, as evidenced by upregulation of MDM2, p53, and p21 protein levels. For the first time, the PA/LFN delivery platform was utilized to deliver a bioactive D-peptide into the cytosol of a eukaryotic cell, where it subsequently bound the target protein and disrupted a critical PPI. 1.5.5. Non-Natural Peptides The translocation of proteins containing non-natural moieties is of particular interest for their bioactivity, stability, or affinity properties. Conjugates containing peptide cargo with p- alanine, N-methyl alanine, propargylglycine, or 2,4,5-trifluorophenylalanine modifications were found to translocate efficiently into CHO-Ki cells by western blot and protein synthesis inhibition assays.69 Thus, subtle non-natural modifications to the peptide cargo do not affect translocation efficiency through PA pore. The same conjugates with biotinylated C-termini were also found to translocate efficiently, as indicated by co-staining of DTA and streptavidin on the western blot. These results demonstrated efficient translocation of the biotin moiety and provided further support for the delivery of intact conjugates into the cytosol. There are several other 36 examples of translocating biotinylated cargo, which collectively confirm the efficient translocation of biotinylated protein conjugates and the accessibility of biotin as an affinity 60 handle. , 67, 70 1.5.6. Cyclic Peptides The cyclization of peptides is a useful approach to develop proteolytically stable therapeutics with large surface areas for protein binding. 69 The delivery of L and D-forms of a cyclic peptide comprised of 11 amino acids was recently explored using the PA/LFN delivery platform. According to the protein synthesis inhibition assay and western blot analysis, the cyclic peptide conjugates were unable to translocate through the PA pore. Rabideau, et al. hypothesized that the cyclic peptides' constrained conformation and inability to unfold contributed to their inefficient translocation (Figure 1.5.1b). Further investigation is required to fully understand the potential of cyclic peptides as cargo of the PA/LFN delivery platform. 1.5.7. Small Molecule Drugs To further probe translocation through PA pore, the delivery of small molecule drugs has been explored. Small molecule drugs have diverse properties, functionalities, and threedimensional structures. The translocation of three common chemotherapeutics, doxorubicin, docetaxel, and monomethyl auristatin F (MMAF) was analyzed by Rabideau, et al. 69 Analysis by protein synthesis inhibition assay as well as western blot of the cytosolic fraction indicated that small molecule drugs can translocate through the PA pore, but with limitations. Of the three molecules tested, docetaxel was unable to translocate into the cytosol through PA pore. We hypothesized that constrained or rigid molecules such as docetaxel cannot unfold or adopt a conformation amenable to translocation through the -12 A pore (Figure 1.5. 1b). 37 1.5.8. Modified Delivery System The PA/LFN delivery platform has been modified in some cases for the delivery of fluorescent cargos such as small molecule probes or proteins. Using the genetic code expansion technique, Zheng et al. installed an alkynyl-pyrrolysine residue within LF at position K581 then used click chemistry to site-selectively label the protein with AlexaFluor545. 7 1 The researchers monitored endocytic trafficking of the labeled protein in BHK fibroblast cells. Zornetta, et al. demonstrated the PA-mediated delivery of GFP into the cytosol fused to LF.7 The delivery of mCherry, however, proved inefficient. The researchers hypothesized that the difference in translocation was due to a higher resistance of mCherry to unfolding. The delivery of fluorescent proteins requires further investigation with respect to the cargo protein's melting temperature and the fate of the chromophore during translocation. The PA pore is cation-selective, thus favoring the passage of protonated or neutral species. While the N-terminal 28 residues (18 are charged) are critical for initiating translocation, the incorporation of cysteic acid (pKa -1.9) in this region halts translocation (Figure 1.5.1b). 73 Nterminal fusions of polycationic stretches (Lys, Arg, His) have been investigated for the delivery of DTA using PA only. Blanke, et al. and Sharma, et al. showed that DTA fused to polycationic residues rather than LFN can be delivered into cells through PA pore. 74' 75 Wright, et al. recently utilized PA-mediated delivery to investigate the delivery of peptide nucleic acids (PNA).76 Reporter cells containing luciferase transgenes with mutant splice sites were treated with PA and antisense PNA(Lys) 8 oligomers, which bind to the mutant splice sites. The researchers demonstrated that the PNA oligomers were delivered into cells, corrected the splice defect, and induced luciferase expression. 38 1.5.9. Retargeting PA Recently, the PA/LFN delivery platform has been modified for the translocation of material into specific cells such as those that overexpress specific receptors. Collier and coworkers retargeted PA to recognize non-native receptors such as HER2 and EGFR by mutating two key residues responsible for binding the anthrax receptors and adding a new targeting domain to the C-terminus of PA. 77~79 Retargeting PA to cells that overexpress specific receptors is a method to increase the amount of material delivered into the cell. Investigations are underway to study the delivery efficacy and bioactivity of delivered non-native cargos using retargeted PA proteins. 39 a. Efficient translocating cargos LFN L LFNAffibody LFN D-Affibody DARPin DTALFL LI FJ LN LFN O LFN a LFN D-GB1 GB1 PE ® 1OFN3 (HA4) LFN-QOjK JONH, (-CONH, S k f r p d s n v r N G ONH2 ~ o01y HN-NH H N LFN Q = L-amino acid - = D-amino acid HO b. Inefficient translocating cargos LF N-Gi(S3(D(PCONH, LFN. S F S LFN 7q, V GR IK fl-c. E~~S5 0K 0 0 0 ~YONO ., K *0& NK I H yz , DTA with N58C+S146C disulfide OH Figure 1.5.1. Translocation efficiency has been analyzed for various peptides and proteins. a. Representative protein and peptide cargos that translocated efficiently through the PA pore include DTA, PE, affibody (L and D forms), GB1 (L and D forms), DARPin, HA4, AKFRPDSNVRG peptide (L and D forms), biotinylated AKFRPDSNVRG peptide, and a doxorubicin-peptide conjugate b. Representative protein and peptide cargos that did not translocate efficiently (gray box) through the PA pore include DTA containing N58C+S146S disulfide, AKFRPDSNVRG cyclic peptide, docetaxel-peptide conjugate, and peptide containing three cysteic acid residues. 40 Table 1.5.1. More than 30 different non-native cargos have been delivered into the cytosol of cells. Cargos range from proteins comprised of natural or non-natural amino acids, peptides with different functionalities, and small molecules. The cargos that could not be efficiently translocated by the PA/LFN delivery platform are highlighted in gray. Cell Type Fusion and Pseudomonas exotoxin A (PE) CHO Recombinant Diphtheria toxin, A chain (DTA) CHO-KI, RAW264.7, MC3T3, RBL-1, VERO, L6 CHO-KI Mouse, P815 (H-2d) Mouse spleen, CHO, HeLa CHO-KI, L6 HEp-2 Recombinant or enzymatic or NCL Recombinant Recombinant Recombinant, chemical Recombinant NCL Recombinant AlexaFluor545 (conjugated to K581 of LF) Affibody B 1domain of protein G (GB 1) Tenth human fibronectin type three domain (IOFN3) Designed ankyrin repeats protein (DARPin) Mouse HEp-2 HN6 BHK CHO-KI HeLa, HCT 116, MDAMB-231, HEK 293T RAW264.7, CHO-KI, HeLa, HN6 J774A.1, BHK-21 CHO-KI, HEK 293T CHO-KI CHO-KI CHO-KI HA4-7c12 tandem monobody AKFRPDSNVRG peptide akfrpdsnvrG (all D) peptide AKFRPDSNvrG (Dcap) peptide tawyanf*ekllr (all D; P is p-CF 3-D-Phe) peptide Affibody (all D) GBl (all D) Biotinylated affibody (all D) [D-AlaKFRPDSNVRG peptide [N-Me-Ala]KFRPDSNVRG peptide [Prop-Gly]KFRPDSNVRG peptide AK[F 3-Phe]RPDSNVRG peptide AK(Cys)FRPDSNVRG peptide LRRLRAC(Doxorubicin) peptide-drug conjugate K562 CHO-KI CHO-KI CHO-KI U-87 MG CHO-KI CHO-KI CHO-KI CHO-KI CHO-KI CHO-KI CHO-Ki CHO-KI CHO-KI A chain of Shiga toxin (STA) Listeriolysin 0 epitope (LLO) 0 lactamase Dihydrofolate reductase (DHFR) Actin cross-linking domain (ACD) of RTX from Vibrio cholerae Legionella pneumophila flagellin protein Rho Inactivation Domain (RID) Reductively methylated PE Green fluorescent protein (GFP) (at C-terminus of LF) KEKEKNKDENKRKDEER (ligated to N-terminus of LFN) Ras/Rap I-specific endopeptidase (RRSP) Cytolethal distending toxin B (CdtB) Citation conjugation strategies Non-Native Cargo Wn-,47 Recombinant Recombinant 46,47 Recombinant Recombinant Recombinant 49 Recombinant, NHS bioconjugation Site specific click Enzymatic Enzymatic Enzymatic Enzymatic or Recombinant Enzymatic Enzymatic Enzymatic Enzymatic Enzymatic Enzymatic Enzymatic Enzymatic, click Enzymatic Enzymatic Enzymatic Enzymatic Enzymatic Enzymatic 8 _ _ 59 __ 59 67 69 69 69 69 41 1.6. Summary and Outlook The PA/LFN delivery platform permits the facile delivery of biomolecules with diverse structures and functionalities into the cytosol of eukaryotic cells (Figure 1.5.1 and Table 1.5.1). The PA pore is relatively promiscuous for the delivery of non-native cargo on the C-terminus of LFN. Analyses by several groups have demonstrated that once translocation is initiated by LFN, there are a few guidelines that cargos must follow in order to gain efficient cytosolic entry through the PA pore. First, the cargo must be able to adopt an unfolded or extended conformation in the endosome. Second, non-natural moieties such as non-natural backbone structures and side chain modifications like mirror image or modified amino acids do not disrupt the translocation process. Third, cargos containing moieties with low pKa values that cannot be protonated in the endosome may inhibit translocation. Taken together, these design principles can be employed for the delivery of previously unexplored cargos such as oligonucleotides or post-translationally modified proteins. There are several questions that remain unanswered regarding the PA/LFN delivery platform. The amount of material delivered into the cytosol varies based on the cell type (i.e. number of anthrax receptors expressed), incubation time (i.e. rate of endocytosis and receptor recycling), concentration of PA and LFN constructs, and translocation efficiency of the LFN construct. While western blot analysis provides a semi-quantitative analysis of the amount of material delivered into the cytosol, a more accurate and sensitive method will provide researchers with a better glimpse of translocation efficiency and a more precise measure of the amount of material delivered into the cytosol. Furthermore, exploration into the in vivo effects of the PA/LFN delivery platform will provide insight into the platform's efficacy and possible immunogenicity in multicellular organisms. 42 1.7. References 1. Falnes, P.O. & Sandvig, K. Penetration of protein toxins into cells. Current Opinion in Cell Biology 12, 407-413 (2000). Young, J.A.T. & Collier, R.J. Anthrax toxin: Receptor binding, internalization, pore formation, and translocation. Annu. Rev. Biochem. 76, 243-265 (2007). Pannifer, A.D. et al. Crystal structure of the anthrax lethal factor. Nature 414, 229-233 (2001). Miller, C.J., Elliott, J.L. & Collier, R.J. Anthrax protective antigen: Prepore-to-pore conversion. Biochemistry 38, 10432-10441 (1999). Milne, J.C., Furlong, D., Hanna, P.C., Wall, J.S. & Collier, R.J. Anthrax protective antigen forms oligomers during intoxication of mammalian-cells. J Biol. Chem. 269, 20607-20612 (1994). Lacy, D.B., Wigelsworth, D.J., Melnyk, R.A., Harrison, S.C. & Collier, R.J. Structure of heptameric protective antigen bound to an anthrax toxin receptor: A role for receptor in pH-dependent pore formation. Proceedings of the National Academy of Sciences of the United States ofAmerica 101, 13147-13151 (2004). Duesbery, N.S. et al. Proteolytic inactivation of MAP-kinase-kinase by anthrax lethal factor. Science 280, 734-737 (1998). Scobie, H.M., Rainey, G.J.A., Bradley, K.A. & Young, J.A.T. Human capillary morphogenesis protein 2 functions as an anthrax toxin receptor. Proc. Nati. Acad Sci. U S. A. 100, 5170-5174 (2003). Bradley, K.A., Mogridge, J., Mourez, M., Collier, R.J. & Young, J.A.T. Identification of the cellular receptor for anthrax toxin. Nature 414, 225-229 (2001). Lacy, D.B., Wigelsworth, D.J., Scobie, H.M., Young, J.A.T. & Collier, R.J. Crystal structure of the von Willebrand factor A domain of human capillary morphogenesis protein 2: An anthrax toxin receptor. Proceedingsof the NationalAcademy ofSciences of the United States ofAmerica 101, 6367-6372 (2004). Scobie, H.M. et al. Anthrax toxin receptor 2-dependent lethal toxin killing in vivo. Plos Pathogens2, 949-955 (2006). Nanda, A. et al. TEM8 interacts with the cleaved C5 domain of collagen alpha 3(VI). Cancer Research 64, 817-820 (2004). Reeves, C.V., Dufraine, J., Young, J.A.T. & Kitajewski, J. Anthrax toxin receptor 2 is expressed in murine and tumor vasculature and functions in endothelial proliferation and morphogenesis. Oncogene 29, 789-801 (2010). Abi-Habib, R.J. et al. BRAF status and mitogen-activated protein/extracellular signalregulated kinase kinase 1/2 activity indicate sensitivity of melanoma cells to anthrax lethal toxin. Mol. Cancer Ther. 4, 1303-1310 (2005). Kintzer, A.F. et al. The protective antigen component of anthrax toxin forms functional octameric complexes. J. Mol. Biol. 392, 614-629 (2009). Klimpel, K.R., Molloy, S.S., Thomas, G. & Leppla, S.H. Anthrax toxin protective antigen is activated by a cell-surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad Sci. U. S. A. 89, 10277-10281 (1992). Petosa, C., Collier, R.J., Klimpel, K.R., Leppla, S.H. & Liddington, R.C. Crystal structure of the anthrax toxin protective antigen. Nature 385, 833-838 (1997). 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 43 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 44 Mogridge, J., Cunningham, K. & Collier, R.J. Stoichiometry of anthrax toxin complexes. Biochemistry 41, 1079-1082 (2002). Feld, G.K. et al. Structural basis for the unfolding of anthrax lethal factor by protective antigen oligomers. Nature Structural& Molecular Biology 17, 1383-U245 (2010). Nassi, S., Collier, R.J. & Finkelstein, A. PA(63) channel of anthrax toxin: An extended beta-barrel. Biochemistry 41, 1445-1450 (2002). Krantz, B.A., Trivedi, A.D., Cunningham, K., Christensen, K.A. & Collier, R.J. Acidinduced unfolding of the amino-terminal domains of the lethal and edema factors of anthrax toxin. J. Mol. Biol. 344, 739-756 (2004). Krantz, B.A., Finkelstein, A. & Collier, R.J. Protein translocation through the anthrax toxin transmembrane pore is driven by a proton gradient. J. Mol. Biol. 355, 968-979 (2006). Blaustein, R.O. & Finkelstein, A. Diffusion limitation in the block by symmetric tetraalkylammonium ions of anthrax toxin channels in planar phospholipid bilayer membranes. J. Gen. Physiol. 96, 943-957 (1990). Arora, N. & Leppla, S.H. Residues 1-254 of anthrax toxin lethal factor are sufficient to cause cellular uptake of fused polypeptides. J. Biol. Chem. 268, 3334-3341 (1993). Elliott, J.L., Mogridge, J. & Collier, R.J. A quantitative study of the interactions of Bacillus anthracis edema factor and lethal factor with activated protective antigen. Biochemistry 39, 6706-6713 (2000). Lacy, D.B. et al. A model of anthrax toxin lethal factor bound to protective antigen. Proceedings of the National Academy of Sciences of the United States of America 102, 16409-16414 (2005). Zhang, S., Finkelstein, A. & Collier, R.J. Evidence that translocation of anthrax toxin's lethal factor is initiated by entry of its N terminus into the protective antigen channel. Proceedings of the National Academy of Sciences of the United States of America 101, 16756-16761 (2004). Jiang, J., Pentelute, B.L., Collier, R.J. & Zhou, Z.H. Atomic structure of anthrax protective antigen pore elucidates toxin translocation. Nature 521, 545 - 549 (2015). Krantz, B.A. et al. A phenylalanine clamp catalyzes protein translocation through the anthrax toxin pore. Science 309, 777-781 (2005). Sun, J., Lang, A.E., Aktories, K. & Collier, R.J. Phenylalanine-427 of anthrax protective antigen functions in both pore formation and protein translocation. Proc. Natl. Acad. Sci. U. S. A. 105, 4346-4351 (2008). Pentelute, B.L., Barker, A.P., Janowiak, B.E., Kent, S.B.H. & Collier, R.J. A Semisynthesis Platform for Investigating Structure-Function Relationships in the NTerminal Domain of the Anthrax Lethal Factor. Acs Chemical Biology 5, 3 59-364 (2010). Dawson, P.E., Muir, T.W., Clark-Lewis, I. & Kent, S.B. Synthesis of proteins by native chemical ligation. Science 266, 776-9 (1994). Popp, M.W., Antos, J.M., Grotenbreg, G.M., Spooner, E. & Ploegh, H.L. Sortagging: a versatile method for protein labeling. Nat. Chem. Biol. 3, 707-708 (2007). Chen, I., Dorr, B.M. & Liu, D.R. A general strategy for the evolution of bond-forming enzymes using yeast display. Proc. Natl. Acad Sci. U S. A. 108, 11399-11404 (2011). Sletten, E.M. & Bertozzi, C.R. Bioorthogonal Chemistry: Fishing for Selectivity in a Sea of Functionality. Angewandte Chemie-InternationalEdition 48, 6974-6998 (2009). 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Ling, J.J.J., Policarpo, R.L., Rabideau, A.E., Liao, X.L. & Pentelute, B.L. Protein Thioester Synthesis Enabled by Sortase. J. Am. Chem. Soc. 134, 10749-10752 (2012). Blaustein, R.O., Koehler, T.M., Collier, R.J. & Finkelstein, A. Anthrax toxin: channelforming activity of protective antigen in planar phospholipid bilayers. Proc Natl Acad Sci USA 86, 2209-13 (1989). Arora, N. & Leppla, S.H. Fusions of anthrax toxin lethal factor with shiga toxin and diphtheria toxin enzymatic domains are toxic to mammalian cells. Infect. Immun. 62, 4955-4961 (1994). Milne, J.C., Blanke, S.R., Hanna, P.C. & Collier, R.J. Protective antigen-binding domain of anthrax lethal factor mediates translocation of a heterologous protein fused to its amino- or carboxy-terminus. Mol. Microbiol. 15, 661-666 (1995). Dmochewitz, L. et al. Role of CypA and Hsp90 in membrane translocation mediated by anthrax protective antigen. CellularMicrobiology 13, 359-373 (2011). Wilson, B.A. & Collier, R.J. Diphtheria-toxin and Pseudomonas aeruginosa exotoxin A active-site structure and enzymatic mechanism. Curr. Top. Microbiol. Immunol. 175, 2741(1992). Adam, S.A., Marr, R.S. & Gerace, L. Nuclear-protein import in permeabilized mammalian-cells requires soluble cytoplasmic factors. J. Cell Biol. 111, 807-816 (1990). Wesche, J., Elliott, J.L., Falnes, P.O., Olsnes, S. & Collier, R.J. Characterization of membrane translocation by anthrax protective antigen. Biochemistry 37, 15737-15746 (1998). Arora, N., Klimpel, K.R., Singh, Y. & Leppla, S.H. Fusions of anthrax toxin lethal factor to the ADP-ribosylation domain of Pseudomonasexotoxin A are potent cytotoxins which are translocated to the cytosol of mammalian cells. J. Biol. Chem. 267, 15542-15548 (1992). Hu, H.J. & Leppla, S.H. Anthrax Toxin Uptake by Primary Immune Cells as Determined with a Lethal Factor-beta-Lactamase Fusion Protein. Plos One 4 (2009). Ballard, J.D., Collier, R.J. & Starnbach, M.N. Anthrax toxin-mediated delivery of a cytotoxic T-cell epitope in vivo. Proc. Natl. Acad Sci. U S. A. 93, 12531-12534 (1996). Ballard, J.D., Collier, R.J. & Starnbach, M.N. Anthrax toxin as a molecular tool for stimulation of cytotoxic T lymphocytes: Disulfide-linked epitopes, multiple injections, and role of CD4+ cells. Infection and Immunity 66, 4696-4699 (1998). von Moltke, J. et al. Rapid induction of inflammatory lipid mediators by the inflammasome in vivo. Nature 490, 107-U 126 (2012). Cordero, C.L., Kudryashov, D.S., Reisler, E. & Satchell, K.J.F. The actin cross-linking domain of the Vibrio cholerae RTX toxin directly catalyzes the covalent cross-linking of actin. J. Biol. Chem. 281, 32366-32374 (2006). Sheahan, K.-L. & Satchell, K.J.F. Inactivation of small Rho GTPases by the multifunctional RTX toxin from Vibrio cholerae. Cellular Microbiology 9, 1324-1335 (2007). Antic, I., Biancucci, M. & Satchell, K.J.F. Cytotoxicity of the Vibrio vulnificus MARTX toxin Effector DUF5 is linked to the C2A Subdomain. Proteins-StructureFunction and Bioinformatics82, 2643-2656 (2014). - 36. 45 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 46 Antic, I., Biancucci, M., Zhu, Y., Gius, D.R. & Satchell, K.J.F. Site-specific processing of Ras and RapI Switch I by a MARTX toxin effector domain. Nature communications 6, 7396-7396 (2015). Koide, A., Bailey, C.W., Huang, X.L. & Koide, S. The fibronectin type III domain as a scaffold for novel binding proteins. J. Mol. Biol. 284, 1141-1151 (1998). Koide, S., Koide, A. & Lipovsek, D. in Methods Enzymol. (eds. Wittrup, K.D. & Verdine, G.L.) 135-156 (Elsevier Academic Press Inc, San Diego, 2012). Lofblom, J. et al. Affibody molecules: Engineered proteins for therapeutic, diagnostic and biotechnological applications. FEBS Letters 584, 2670-2680 (2010). Nygren, P.A. Alternative binding proteins: Affibody binding proteins developed from a small three-helix bundle scaffold. Febs Journal275, 2668-2676 (2008). Parizek, P. et al. Designed Ankyrin Repeat Proteins (DARPins) as Novel IsoformSpecific Intracellular Inhibitors of c-Jun N-Terminal Kinases. ACS Chem. Biol. 7, 13561366 (2012). Mandal, K. et al. Chemical synthesis and X-ray structure of a heterochiral {D-protein antagonist plus vascular endothelial growth factor} protein complex by racemic crystallography. Proceedingsof the National Academy of Sciences of the United States of America 109, 14779-14784 (2012). Liao, X., Rabideau, A.E. & Pentelute, B.L. Delivery of Antibody Mimics into Mammalian Cells via Anthrax Toxin Protective Antigen. Chembiochem 15, 2458-2466 (2014). Verdurmen, W.P.R., Luginbuehl, M., Honegger, A. & Plueckthun, A. Efficient cellspecific uptake of binding proteins into the cytoplasm through engineered modular transport systems. Journal of Controlled Release 200, 13-22 (2015). Grebien, F. et al. Targeting the SH2-Kinase Interface in Bcr-Abl Inhibits Leukemogenesis. Cell 147, 306-319 (2011). Wojcik, J. et al. A potent and highly specific FN3 monobody inhibitor of the Ab SH2 domain. Nat. Struct. Mol. Biol. 17, 519-U173 (2010). Grimm, S., Salahshour, S. & Nygren, P.A. Monitored whole gene in vitro evolution of an anti-hRaf- 1 affibody molecule towards increased binding affinity. New Biotechnology 29, 534-542 (2012). Gupta, P.K., Moayeri, M., Crown, D., Fattah, R.J. & Leppla, S.H. Role of N-Terminal Amino Acids in the Potency of Anthrax Lethal Factor. Plos One 3 (2008). Bachran, C. et al. Reductive Methylation and Mutation of an Anthrax Toxin Fusion Protein Modulates its Stability and Cytotoxicity. Scientific Reports 4 (2014). Milton, R.C.D., Milton, S.C.F. & Kent, S.B.H. Total Chemical Synthesis of a D-Enzyme - the Enantiomers of Hiv- 1 Protease Show Demonstration of Reciprocal Chiral SubstrateSpecificity. Science 256, 1445-1448 (1992). Rabideau, A.E., Liao, X. & Pentelute, B.L. Delivery of mirror image polypeptides into cells. Chemical Science 6, 648-653 (2015). Zhan, C. et al. An Ultrahigh Affinity D-Peptide Antagonist Of MDM2. J. Med Chem. 55, 6237-6241 (2012). Rabideau, A.E., Liao, X., Akqay, G. & Pentelute, B.L. Translocation of Non-Canonical Polypeptides into Cells Using Protective Antigen. 5 (2015). 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. Bachran, C. et al. Cytolethal distending toxin B as a cell-killing component of tumortargeted anthrax toxin fusion proteins. Cell Death & Disease 5 (2014). Zheng, S., Zhang, G., Li, J. & Chen, P.R. Monitoring Endocytic Trafficking of Anthrax Lethal Factor by Precise and Quantitative Protein Labeling. Angew. Chem. Int. Ed 53, 16 (2014). Zometta, I. et al. Imaging the cell entry of the anthrax oedema and lethal toxins with fluorescent protein chimeras. Cellular Microbiology 12, 1435-1445 (2010). Pentelute, B.L., Sharma, 0. & Collier, R.J. Chemical dissection of protein translocation through the anthrax toxin pore. Angew. Chem. Int. Ed. Engl. 50, 2294-2296 (2011). Blanke, S.R., Milne, J.C., Benson, E.L. & Collier, R.J. Fused polycationic peptide mediates delivery of diphtheria toxin A chain to the cytosol in the presence of anthrax protective antigen. Proceedings of the NationalAcademy of Sciences of the United States ofAmerica 93, 8437-8442 (1996). Sharma, 0. & Collier, R.J. Polylysine-Mediated Trans location of the Diphtheria Toxin Catalytic Domain through the Anthrax Protective Antigen Pore. Biochemistry 53, 69346940 (2014). Wright, D.G., Zhang, Y. & Murphy, J.R. Effective delivery of antisense peptide nucleic acid oligomers into cells by anthrax protective antigen. Biochemical and Biophysical Research Communications376, 200-205 (2008). McCluskey, A.J. & Collier, R.J. Receptor-directed chimeric toxins created by sortasemediated protein fusion. Mol. Cancer Ther. (2013). McCluskey, A.J., Olive, A.J., Starnbach, M.N. & Collier, R.J. Targeting HER2-positive cancer cells with receptor-redirected anthrax protective antigen. Mol. Oncol. 7, 440-51 (2013). Mechaly, A., McCluskey, A.J. & Collier, R.J. Changing the Receptor Specificity of Anthrax Toxin. Mbio 3 (2012). Hobson, J.P., Liu, S.H., Rono, B., Leppla, S.H. & Bugge, T.H. Imaging specific cellsurface proteolytic activity in single living cells. Nature Methods 3, 259-261 (2006). Bachran, C. et al. Anthrax toxin-mediated delivery of the Pseudomonas exotoxin A enzymatic domain to the cytosol of tumor cells via cleavable ubiquitin fusions. Mbio 4, e00201-13 (2013). 47 Chapter 2: Delivery of Antibody Mimics into Mammalian Cells via Anthrax Toxin Protective Antigen Portions of the work presented in this chapter were published in the following manuscript and is reproduced with permission from John Wiley & Sons: Liao, X.,* Rabideau, A.E.* & Pentelute, B.L. Delivery of Antibody Mimics into Mammalian Cells via Anthrax Toxin Protective Antigen. ChemBioChem 15, 2458-2466 (2014). *: co-first author 48 2.1. Introduction Antibodies have proven to be powerful tools in science and therapeutics by facilitating the elucidation of disease mechanisms and generating novel and effective therapeutics. However, the use of antibodies has been limited to outside of the cell because of two major factors: antibodies containing disulfide bonds may be unstable in the reducing environment of cytosol, and antibodies are unable to cross the cell plasma membrane to reach the cytosol. There are numerous intracellular targets and protein-protein interactions with large flat contact areas that are considered difficult to perturb by small molecules. We believe it would be of immediate interest to use antibody mimics to target the intracellular protein-protein interactions provided that they can be transported into the cytosol using a straightforward delivery platform. In recent years, certain scaffold proteins that are robust, single-domain, and cysteine-free have emerged as antibody mimics. These antibody mimics include monobodies derived from the tenth type III domain of human fibronectin (1OFN3), affibodies derived from the immunoglobulin binding protein A, DARPins based on ankyrin repeat modules, and B 1 domain of protein G (GB 1).1~6 These antibody mimics have not only been engineered to bind extracellular receptors such as EGFR, HER2 receptor, VEGFR and integrin, but also various intracellular targets including caspases, Raf, Erk, c-Jun N-Terminal Kinases (JNK), Abl-SH2, and c-Jun.7'1 4 Advances in directed evolution and molecular display technologies such as phage display, yeast display, and ribosome display make it possible to routinely generate a wide variety of high affinity binders for specific protein targets.15-17 Some research efforts are now focused on applying these antibody mimics inside the cell (intrabodies) to target cytosolic proteins.' 8 To achieve this goal, strategies are needed to allow for 49 the facile and reliable delivery of these bioactive antibody mimics into the cytosol of various cell types. Delivery methods based on lipid-derived compounds,' 9 polymeric nanoparticles,2 0 , 21 inorganic nanocarriers, 2 2 supercharged proteins,2 3' 24 and most commonly, cell-penetrating peptides (CPPs) including the transactivator of transcription (TAT) of HIV-1, oligoarginine, and penetratin peptide derived from the Drosophilaantennapedia protein,25-28 have been extensively developed to deliver proteins of interest to the cytosol of mammalian cells. In most cases, high concentrations of these agents are required to achieve even modest effects often due to inefficient cargo escape from the endosome to the cytosol. Nature has evolved a variety of mechanisms to transport proteins across lipid membranes into the cytosol of mammalian cells.2 9 One protein transport nanomachine from bacteria is protective antigen (PA; 83 kDa), a component of anthrax toxin. PA is a receptor-binding, poreforming transporter, which has evolved to convey the enzymatic moieties of the toxin (lethal factor and edema factor) from the external milieu to the cytosol of mammalian cells. PA binds to ) host cell receptors and is cleaved by a furin-family protease to yield a 63 kDa species (PA 63 (Figure 2.1.1 a; step 1)30 that self-assembles to form ring-shaped heptamers 3 1 and octamers.3 2 These oligomers then form complexes with the cargo proteins (K 1I nM) and are endocytosed to the endosome (Figure 2.1.la; steps 2-4), where the acidification triggers conformational change of the PA 6 3 oligomers to form a transmembrane pore that unfolds and translocates the bound cargo proteins to the cytosol (Figure 2.1.1a; step 5).33 The N-terminal domain of lethal factor (LFN; ~30 kDa) is sufficient to bind PA oligomers and initiate translocation. 34 Prior studies have shown fusions of cargo to the C-terminus of LFN can be transported to the cytosol via PA; however, most efforts have focused on the delivery of peptides for vaccine development,3 5 or enzymes including P-lactamase, 36' 37 or enzymatic domains from diphtheria toxin (DTA), Shiga 50 toxin, Pseudomonas exotoxin A (PEIII), and RTX toxin (ACD).38-41 More recently, the PA/LFN system was shown to deliver Legionellapneumophilaflagellin into macrophages. 42 Despite these prior efforts, no study has investigated the ability of PA to translocate the above-mentioned antibody mimics for perturbation of intracellular protein-protein interactions, and the possibility of using the PA/LFN system and its active endosomal escape mechanism as another delivery tool to expand antibody therapeutics inside the cell. Herewith we use the transpeptidase sortase (SrtA) 43 to conjugate several commonly used antibody mimics to the C-terminus of LFN and find that PA can mediate their transport into the cytosol of several different cell lines. We confirm the refolding and binding of a tandem monobody to the protein target Bcr-Abl inside cell by co-immunoprecipitation. We observe inhibition of Abl kinase activity and subsequent cell death caused by the PA-delivered monobody. In a separate case, we show that the PA system delivers an hRaf-1 binder based on the affibody scaffold that disrupts MAPK signaling pathway. 51 b) PA 3 PA PA 63 LDv: LF N 2 Lv: W LPSTG LPSTG 5 * a) H+ c) Antibody mimics d) 1.2 U) M*44C-term ~LFN-DTA 1.0 OD -A-LDv2 1: Affibody -w- LDv3 LDv4 0.8 'c-err 4: 1OFN3 -+-LDv4 0. mr-m 4C:eHAm 0.2 . 2: GB1 LL 4rrl HA4 0.2 AA -14 -13 -12 -11 -10 -9 -8 -7 Log [Protein Concentralon (M)] -6 -5 3: DARPin Figure 2.1.1. Delivery of antibody mimics into the cytosol by the PA/LFN system. a) Mechanism of entry into cells. The star represents the antibody mimic to be delivered. b) Antibody mimics 1-4 and 4mut: affibody (PDB: 1Q2N), GB1 (PDB: 1PGB), DARPin (adapted from PDB: 3ZU7), 1OFN3 (PDB: lTTF), and HA4 (PDB: 3K2M). c) Variants of antibody mimics attached onto the C-terminus of LFN (Lv) or LFN-DTA (LDv). d) Protein synthesis inhibition assay using CHO-Ki cells treated with varying concentrations of LDvl-4 and 4 mut in the presence of 20 nM PA for 30 minutes. The cells were then washed 3 times with PBS and incubated with 1 pCi mL-1 3H-Leucine for 1 h. Subsequently, the cells were washed three times with PBS and scintillation fluid was directly added to each well and 3H incorporation into the cellular proteome was measured to determine the level of protein synthesis inhibition by LDvs (n = 3). The radioactive counts were normalized to that of cells treated with only PA (set to 1). 52 2.2. Results 2.2.1. Delivery of antibody mimic proteins In this study, our antibody mimics consisted of scaffolds widely used to generate highly specific and potent binders: an affibody, protein GB1, DARPin, and monobody (Figure 2.1. 1b). These scaffolds are disulfide-free, avoiding possible interference with passage through the PA translocase and potential stability problems in the reducing environment of the cytosol. Our chemoenzymatic bioconjugation route based on an evolved SrtA (SrtA*) is shown in Figure 2.6.1. SrtA* catalyzed the formation of covalent conjugates between LFN containing C- terminal LPXTG recognition motif and antibody mimics containing N-terminal oligoglycine, designated Lv (Figure 2.1.1c). We also prepared a series of conjugates between LFN-DTA and each antibody mimic, (designated LDv, Figure 2.1.1c) to measure their PA-mediated translocation into the cytosol. In anthrax toxin translocation studies, the A chain of diphtheria toxin (DTA), which catalyzes the ADP-ribosylation of EF-2 and inhibits protein synthesis, has been frequently used as a straightforward measure and a gold standard assay of PA-mediated translocation into the cytosol. Therefore, the LFN-DTA variants (LDv) allow us to compare our findings to existing reports that used the same assay.38'40'46'47 For each antibody mimic, after confirming the translocation of the LDv conjugate, we then carried out additional studies using Lvs, which lack the toxic DTA protein, thus avoiding interference with further characterization of antibody mimic function and delivery into the cytosol. To begin our study, each purified LFN-DTA variant (LDvl-4 and 4mut, Figure 2.6.2) was added to CHO-Ki cells in the presence of 20 nM PA. After 30 minutes of treatment, the cells 53 were washed and incubated with medium supplemented with 3H-Leu. The efficiency of antibody mimic translocation was measured by the incorporation of 3H-Leu in the cellular proteome. This indicates the level of protein synthesis inhibition and is determined by the amount of LFN-DTAcontaining variants that reach the cytosol. As shown in Figure 2.1.1d, despite their structural differences, all four variants, LDvl-3 and 4 mut translocated efficiently into cell cytosol at levels comparable to the positive control, LFN-DTA. Only LDv4 containing the highly stable 1OFN3 (Tm ~-88'C) translocated ~ 0-fold less efficiently, while the variant, LDv4mut containing a mutant form of 10FN3 (HA4) with reduced stability (Tm -75'C) (Figure 2.6.3) translocated as efficiently as LFN-DTA, indicating that the cargo thermal stability can affect translocation efficiency. To confirm that the full-length protein was required for translocation to the cytosol, we performed control experiments by treating cells with DTA plus PA, or DTA together with LFN-affibody (Lvi) plus PA, and observed no protein synthesis inhibition in either case (Figure 2.1.1 d). 2.2.2. Translocation requires a functional PA and endocytosis To further investigate if the LFN-DTA-antibody mimics (LDv) translocated into the cytosol by the same mechanism as LFN-DTA, we carried out a series of control experiments with LDv1 containing affibody as the cargo (Figure 2.2.la). We investigated the role of LFN in LDvl during the translocation process using an excess amount of LFN (1 gM) to outcompete LDvI from binding to PA. We found that 1 pM of LFN significantly abolished translocation of LDvl, confirming the importance of LFN-mediated binding to PA (step 3, Figure 2.1. 1a). We then studied the role of endocytosis in LDvI translocation by incubating cells with PA and LFN-DTA or LDvl at 4'C instead of 37'C to arrest endocytosis. The translocation of both LFN-DTA and LDvl were abolished under this condition, indicating that endocytosis is a necessary step in the 54 translocation process (step 4, Figure 2.1.1 a). To test the role of endosome acidification and active translocation (step 5, Figure 2.1.1 a), we performed two additional control experiments. First we treated the cells with PA and LFN-DTA or LDvl in the presence of 200 nM Bafilomycin Al, a specific inhibitor of the vacuolar H+-ATPase that blocks endosome acidification. We found that inhibition of endosome acidification abolished translocation for both LFN-DTA and LDvl, confirming the importance of the pH gradient for translocation. An additional control experiment was performed with a PA mutant (PA[F427H]) that binds and delivers the cargo to endosomes but arrests translocation through PA pore.48 ' 4 We found that PA[F427H] completely arrested translocation of LDvl, indicating that the functional PA was required for LDv1 to reach the cytosol. These controls indicate that the translocation of LDvI follows the same mechanism as LFN-DTA. 55 a), .2L 1.0- 0.8 0.LFN-DTA 0.4- o -e - LDv1 -LDv1 + LFN -4 - LFN-DTA 40C -4 -LDv1 40C r -LFN-DTA + BA -I Lvl.RA LDv1 + PA[F427H] ,u -15 b) -14 -13 -12 -11 -10 -9 -8 Log [Protein Concentration (M)] Digitonin extracted - + + - Total cell lysate - - + -6 -. -- + - + PA[F427H] PA -7 + + - + BA 4r g LF 1 LFN Lvi Lv1 Lv1 LFN Lv1 Lv1 - 0.0 -- + 0.2 BA 40C Lv1 Lv1 anti-LF i anti-Erki /2 Oo -w6em6 .mw %.o..* -. 4 anti-Rab5 Figure 2.2.1. Control experiments validating the translocation mechanism of LDv1 and Lv1. a) Protein synthesis inhibition assay of CHO-KI cells treated with varying concentrations of variants in the presence of 20 nM PA for 30 minutes. The treatment conditions of control experiments were modified as follows: BA = addition of 200 nM Bafilomycin Al, 4*C = incubation at 4'C instead of 37*C, or PA[F427H] = mutant PA instead of PA. b) CHO-Ki cells were treated with 250 nM Lvi in the presence of 40 nM PA or PA[F427H] for 12 hours. The treatment conditions of control experiments were modified as described above. The total lysate and digitonin extracted cytosolic proteins were prepared separately (~300,000 cells per lane). 56 2.2.3. Western blot analysis of cytosolic fraction After confirming the key steps responsible for the translocation of antibody mimics, we investigated the delivery of Lvs into the cytosol by digitonin extraction and western blot. Digitonin is a weak, nonionic detergent and at low concentrations selectively permeabilizes the plasma membrane, thereby releasing cytosolic components from cells while the nuclear envelope and other major membrane organelles remain intact.5 0 CHO-Ki cells were incubated with Lvl and PA overnight to allow for multiple rounds of receptor-mediated endocytosis and translocation, washed, and trypsin digested to remove cell surface receptors and bound proteins. In order to obtain the cytosolic fraction, cells were incubated with a phosphate buffer containing 50 pg mUl digitonin and 250 mM sucrose. Immunoblot analysis with antibodies against Rab5, an early endosome marker, and Erkl/2, a cytosolic marker protein, was carried out to confirm only the cytosolic fraction was present. Minor amounts of Rab5 protein were detected in the digitonin-extracted cytosolic fraction, indicating little contamination by early endosomes (Figure 2.2.1b). We observed a significant amount of LvI in the cytosolic extraction, similar to that in the total cell lysate obtained by lysis buffer containing 1% NP-40. Again, we performed control experiments to further validate translocation of Lvl into cytosol. No Lvl was observed even in the total cell lysate when the cells were incubated at 4'C in which endocytosis is arrested. The absence of detectable Lvl at 4'C also validated that the wash and trypsin digestion protocol was sufficient to eliminate potential contamination from surface bound proteins. When the cells were treated with Bafilomycin Al, a significant amount of Lvl was observed in the total cell lysate while the band was not detectable in the cytosolic fraction, indicating that Lvi was endocytosed but trapped in the endosome. In addition, the absence of Lvl in the cytosolic fractions for the Bafilomycin Al treated cells served as further evidence that our digitonin extraction protocol 57 was sufficient to separate the cytosolic proteins from the rest of the cell lysate, including endosomes. Finally, when the cells were treated with LvI and PA[F427H] instead of Lvl and PA, some amount of Lvl was observed in the total cell lysate but no Lvi was detected in the cytosolic fraction (Figure 2.2.1b). Under this condition, the endocytosed LvI was trapped in the endosome due the non-functional pore formed by the mutant PA, and subsequently sorted to the lysosome for degradation, leaving only a small amount of LvI detectable in the total cell lysate, but not detectable in the cytosolic fraction. The western blot results further corroborated the protein synthesis inhibition results based on LFN-DTA (vide supra, Figure 2.1.1 d) and confirmed the translocation mechanism of the antibody mimic cargos. The PA-mediated translocation of other analogues (Lv2, 3, 4, 4 mt) was also studied by digitonin extraction and western blot (Figure 2.6.4). The detection of bands at different molecular weight by anti-LF antibody confirmed the cytosolic presence of these antibody mimic conjugates. 2.2.4. Comparison to TAT-mediated delivery To compare the efficiency of PA/LFN system with the CPP system, we first used the protein synthesis inhibition assay. We prepared DTA with a TAT peptide covalently attached at the N-terminus. This construct showed at least 1000 times lower efficiency in protein synthesis inhibition than LFN-DTA and PA (Figure 2.2.2a), which is consistent with a previous report 5 and indicates the PA/LFN system is more efficient in delivering DTA into cytosol. Next, we compared the efficiency of the PA system and CPP system in delivering antibody mimics. We conjugated the oligoglycine-containing antibody mimics 1-4 and 4mut to a TAT peptide containing an HA tag and LPSTGG motif to make the TAT-HA-antibody mimic constructs (Figure 2.6.5).52 We studied their translocation again by digitonin extraction and western blot analysis using the anti-HA antibody. We used a concentration of 2.5 pM for the TAT-HA58 antibody mimic constructs, 10-fold more than the amount of Lv used, to treat the cells for 4 hours in serum free media and observed minor amounts of material in total cell lysate (Figure 2.6.6a). Further, using the trypsin digestion and digitonin extraction protocol described above, we detected none of the TAT-HA-antibody mimic proteins by anti-HA immunoblot in the cytosolic fractions, while a significant amount of Lvi with HA tag (Lvl-HA) was detected on the same blot for cells treated with 250 nM Lvl-HA and 40 nM PA for 16 hours (Figure 2.2.2b). To test if a different incubation time would improve translocation by a TAT peptide, we treated cells with 2.5 jM TAT-HA-antibody mimics for 16 hours, but again observed no material in the total cell lysate (Figure 2.6.6b). These results demonstrate the higher efficiency of the PA/LFN system in delivering antibody mimics over the TAT peptide. Additionally, the amount of LvlHA detected by anti-HA immunoblot was similar to that of LvI detected by anti-LF immunoblot. The presence of a C-terminal HA tag further validated the presence of full-length Lvi -HA in the cytosol. 59 a) 1.2D' 1.0 90.8- 0.40.2. -M-LFN-DTA + PA -+-TAT-DTA 0.-15 b) -14 -13 -12 -11 -10 -9 -8 Log [Protein Concentration (M)] -6 -7 Digitonin extracted cc) 31 9' V V-1* -t7 J~-I kDa 49 38 anti-HA 28 17 14 6 anti-Erki /2 anti-Rab5 Figure 2.2.2. TAT peptide mediated translocation of DTA and antibody mimics. a) Protein synthesis inhibition assay of CHO-Ki cells treated with varying concentrations of TAT-DTA for 30 minutes, in comparison to LFN-DTA plus 20 nM PA. The assay conditions were described in Figure 2.1.1. b) Western blot analysis of cytosolic fractions extracted by digitonin from CHOKi treated with 2.5 pM TAT-HA-i to -3 and -4mut (1-3 and 4mut see Figure 2.1.1b) for 4 h (4'C = incubation at 4*C instead of 37*C), in comparison to treatment with 250 nM Lvl-HA and 40 nM PA for 16 h. The cytosolic fraction was extracted using digitonin (~280,000 cells per lane). The corresponding immunoblots of total lysates from cells treated by TAT-HA-i to -3 and - 4 mut for 4 hours and 16 hours are shown in Figure 2.6.6. 60 2.2.5. Delivery of a tandem monobody for SH2 domain of Bcr-Abl Our next question was to ask whether the delivered antibody mimics can refold and bind to their targets inside the cytosol. Based on the protein synthesis inhibition assay, we confirmed that DTA could correctly refold in the cytosolic environment after translocation. For the antibody mimics, we chose to study the tandem 1OFN3 monobody (HA4-7c12), which binds with nanomolar affinity to the Src homology 2 (SH2) domain of the oncoprotein Bcr-Abl. First, we sortagged this tandem monobody to LFN-DTA (LDv5) and used the protein synthesis inhibition assay as a first measure to study the variant's translocation efficiency in chronic myeloid leukemia (CML) K562 cells. The protein synthesis inhibition assay showed that PA translocated LDv5 as efficiently as the LFN-DTA control (Figure 2.6.7). After confirming that the PA system efficiently translocated the tandem monobody HA4-7c12, we sortagged it to LFN (Lv5, Figure 2.2.3a), which lacks the cytotoxic DTA protein, and studied if Lv5 could bind to Bcr-Abl in K562 cells after translocation into the cytosol. To ensure that the binding affinity of the monobody was not affected by the presence of LFN, which apart from initiating translocation provided an important epitope for detection in the cytosol by western blot, we first measured the binding affinity of Lv5 toward the Abl SH2 domain. We obtained a Kd value of 12 nM, similar to that of the monobody by itself (6 nM, Figure 2.6.8). Next, we investigated if the delivered Lv5 could refold and bind to Abl kinase inside the cell. K562 cells were treated with Lv5 and PA, and subjected to immunoprecipitation with an anti-Abl antibody linked to agarose beads. Proteins eluted from the beads were subjected to immunoblot analysis with an anti-LF antibody. As shown in Figure 2.2.3b, we observed a pull-down band corresponding to Lv5 when cells were treated with Lv5 and PA, confirming that at least some of the monobody could properly fold after reaching the cytosol and was functional to bind to intracellular Abl kinase. We also carried 61 out control experiments where PA[F427H] was used instead of PA to arrest translocation from the endosome to cytosol. No band was observed under this condition, confirming cytosolic access of Lv5 was critical for the binding. Finally, the cells were treated with the binding mutant Lv5mut (HA4: Y87A; 7c62: Y62E/F87K) instead of Lv5. Similar amounts of Lv5 and Lv5mut were delivered to the cell; however, the pull-down band was absent in the Lv 5 mut condition (Figure 2.2.3b), indicating the binding interaction only occurred when the LFN variant contained the functional binder. These results demonstrated that PA-delivered tandem monobody could refold and bind to its target inside the cell. 62 a) W LPSTG 7c12 HA4 - PA[F427H] Lv5 Lv5mut - + + - Input IP: anti-Abi + + PA - - + - - + - - + Lv5mut + + - - - PA[F427H] Lv5 anti-Abi anti-Abi anti-LF anti-LF anti-GAPDH **mto + - - PA + b) + Lv5 MWMw Figure 2.2.3. Delivery of Lv5 and binding of Lv5 to Ab kinase in K562 cells. a) Construct of Lv5 containing HA4 (PDB: 3K2M), GS-rich linker, and 7c12 (PDB: 3T04). b) Western blot analysis (left) of total cell lysate from K562 cells treated for 24 hours with 50 nM Lv5 or Lv5mut in the presence of 20 nM PA or PA[F427H]. The lysate was then subjected to coimmunoprecipitation (right) by anti-Abl agarose beads. Cells treated with Lv5mut/PA or Lv5/PA[F427H] served as negative controls for the co-IP. 63 Our next goal was to investigate the possibility of using the delivered tandem monobody to perturb protein function and related signaling pathways in cancer cells. Overexpression of HA4-7c 12 in K562 cells was reported to strongly inhibit kinase activity and induce apoptosis by disrupting a critical intramolecular SH2-kinase domain-domain interaction. Despite the high affinity and specificity of the monobody in targeting the allosteric module in Bcr-Abl and its utility in fighting the Bcr-Abl drug resistance, Hantschel et aL indicated the intracellular delivery was the biggest hurdle to push its practical application. 53 In order to test if PA-mediated delivery could overcome this challenge, we used PA to translocate Lv5 into K562 cells. Based on the linear relationship between the amount of protein loaded and the signal intensity of each band detected by anti-LF antibody (Figure 2.6.9), we estimated a total of I ng Lv5 delivered into ~100,000 cells (Figure 2.6.10), that was ~10 fg or 110,000 molecules of Lv5 in each cell, giving a cytosolic concentration of -80 nM. Although this concentration was above the Kd towards Abl SH2 domain, we would not expect strong inhibition of Bcr-Abl kinase due to the high concentration of Bcr-Abl inside K562 cells. As detected by western blot, monobody binding resulted in modest reduction of activation loop (Tyr412) phosphorylation of Bcr-Abl (Figure 2.6.10). We then investigated the effect of this activity inhibition in inducing apoptosis of K562 cells. Using TUNEL staining that detects DNA fragmentation by labeling the terminal ends, we observed apoptosis after K562 cells were treated with Lv5 and PA (Figure 2.2.4a). This effect was not observed when PA[F427H] or Lv5mut were used. The amount of apoptotic cells caused by delivered Lv5 was 20% of that caused by the small molecule imatinib (I pM, Figure 2.6.11), which inhibits kinase activity by binding close to the ATP binding site of Bcr-Abl. The different inhibition mechanism and the much higher concentration of imatinib may explain the difference in the amount of apoptotic cells as compared to the Lv5 and PA treatment. In addition, the extent 64 of apoptosis caused by delivered Lv5 was also lower than that induced by overexpression of HA4-7c12 as shown by Grebien et al. The high protein level of the monobody achieved by overexpression and selection for only transfection positive cells overcame the challenge of requiring high local concentration to interfere with the intramolecular interactions between the SH2 domain and kinase domain in Bcr-Abl. With the modest inhibition of kinase activity and induction of apoptosis by PA mediated delivery of Lv5, it serves as a proof-of-concept example showing delivered monobodies can perturb the activity of an oncoprotein. 65 a)0._ 8.54 PA P+m597 PA + Lv5mut PA + Lv5 0 IIi- 102 0. lo, 0 PA[F427H] P911 063PA[F427H] PA[F427H + Lv5 + Lv5mut 104- C) Lj_ 10 3 0, 0 1& 10 0i 0S 2 103 101 101 PefCP-A b) 0 ftrCP-A 102 1&~ PerCP-A 10. 0 24- I 20 : 8 16 L .J J 12 D8 4- T A- PA + - + - + - - + Lv5 Lv5mut - 0- PA[F427H] Figure 2.2.4. Monitoring of apoptosis of K562 cells treated with Lv5 and PA by TUNEL assay. a) K562 cells treated with the indicated analogs (500 nM Lv5 or Lv5mut in the presence of 80 nM PA or PA[F427H]) for 3 days and analyzed for apoptosis. 1 pM imatinib served as a positive control. Representative dot plots from flow cytometry showed terminal deoxynucleotidyl transferase (TdT) catalyzed BrdUTP incorporation into the DNA strand breaks of apoptotic cells, which was detected by Alexa Fluor 488-labeled anti-BrdU antibody (FITC). PerCP indicates the DNA fraction. b) Quantification of TUNEL-positive cells, where the intensities were normalized to K562 cells treated with imatinib (set to 100%) and non-treated cells (set to 0%). The dot-plots of imatinib-treated and non-treated cells are shown in Figure 2.6.11. Each data point represents an average of three independent experiments. 66 2.2.6. Delivery of an affibody for Raf-1 In a separate case, we investigated PA-mediated delivery of another binder based on affibody. The affibody (ABRaf) was evolved to bind to human Raf- 1 (hRaf-1, Kd = 100 nM), 9 a protein kinase of central importance in the MAPK/ERK proliferation pathway. Although ABRaf was shown to inhibit the Ras/Raf interaction in vitro, it has not been tested for direct inhibition of signaling through the MAPK pathway in cells. We transfected human embryonic kidney 293 T (HEK293T) cells with ABRaf, and observed on average 37% reduction in phosphorylation levels of MAPK (p-Erkl/2, downstream of Ras/Raf pathway) upon EGF activation (Figure 2.2.5a), validating the cellular function of this binder in blocking MAPK pathway. We then conjugated ABRaf to LFN (Lv6), and found that PA effectively delivered Lv6 to an estimated concentration of 240 nM (~5 fg or 79,000 molecules per cell) in HEK293T cells (Figure 2.2.5b). We measured the phosphorylation level of MAPK (p-Erkl/2) upon EGF activation in cells treated with Lv6 and PA or PA[F427H]. As shown in Figure 2.2.5c, the cells treated with Lv6 and PA showed -25% reduction of p-Erkl/2 when comparing to EGF-activated untreated cells (P < 0.01, Figure 2.2.5c). PA[F427H] was used as a negative control and showed similar level of p-Erkl/2 as the EGF-activated untreated cells (P = 0.27, Figure 2.2.5c). This result provides another example of using PA to translocate functional binders to disrupt protein-protein interactions and related signaling pathway in cells. 67 + PA[F427H] EGF anti-pErki /2 + - + + + - + anti-pErk1 /2 W anti-Erki /2 - - Lv6 PA - + + + + a )pcDNA3-ABRaf EGF anti-Erki/2 Lv6 PA PA[F427H] anti-LF rn P= 0.27 1.2. LPSTG + + 4 ng w - - IP< 0.01 1.0. 0.8. - o+ 0.6- + b) anti-Erki /2 U anti-Rab5 c 0.2. n.n . I W Figure 2.2.5. Perturbation of the MAPK signaling pathway by PA mediated delivery of an affibody (Lv6) that targets Raf. a) HEK293T cells were transfected with pcDNA3-ABRaf for 24 hours, starved overnight, treated with 5 ng mL~ 1 EGF for 7 minutes, lysed with buffer containing 1% NP-40, and subjected to anti-pErkl/2 immunoblotting analysis. The membrane was stripped and re-blotted with anti-Erk1/2 antibody to serve as loading control. b) Western blot analysis of cytosolic fractions extracted by digitonin (as described above) from HEK293T cells treated with 500 nM Lv6 and 80 nM PA or PA[F427H] in serum-free medium for 12 hours (-400,000 cells per lane). c) HEK293T cells were treated with 500 nM Lv6 and 80 nM PA or PA[F427H] in serum-free medium for 12 hours, treated with 5 ng mL- 1 EGF for 7 minutes, lysed and analyzed as described in (a). Bar graph represents the quantification of level of phosphorylation of Erkl/2 (pErkl/2) (n = 3), where each bar is corresponding to the lane in the western blot (top). The pErk1/2 band intensities were normalized to that of Erkl/2, and compared to cells treated with EGF (set to 1). Each data point represents an average of three experiments. P values were calculated from Student's t-Test. 68 2.3. Discussion In this paper, we demonstrate the ability of PA/LFN to deliver antibody mimics to the cytosol of cells and also the possibility of using the delivered binders to perturb critical proteinprotein interactions in cancer cells. Based on prior work with PA/LFN mediated delivery, the use of this system to perturb protein-protein interactions was not precedented. Efforts have mainly focused on the delivery of enzymes that can exert a strong biological effect at very low concentrations, which is demonstrated by the LFN-DTA activity in the cell assays presented here. However, for an antibody mimic binder to exert inhibitory effects on protein-protein interactions inside cells, the concentration of the antibody mimic binder needs to reach a sufficient level that is above the K4 as well as close to the concentration of competing endogenous binding proteins. We aimed to answer the question of whether PA could not only efficiently deliver the antibody mimics, but also transport enough cargo to perturb protein-protein interactions. We first systematically investigated the translocation mediated by PA of four different antibody mimics: all a-helical (affibody, DARPin), all P-sheet (monobody), or a-helical and p- sheet proteins (GB1). We did not assume these scaffolds would translocate efficiently because prior investigations have indicated that certain C-terminal modifications of LFN or LFN-DTA can abrogate translocation through PA. In particular, imaging studies have indicated that fluorescent proteins fused to the C-terminus of LF significantly attenuate the efficiency of translocation. Additional studies have shown DTA with an artificial disulfide or DHFR complexed with methotrexate can arrest translocation when attached to the C-terminus of LFN. 56 Using the DTAbased protein synthesis inhibition assay, we found all four antibody mimics translocated as efficiently as the positive control LFN-DTA. We are unaware of any system that permits the facile delivery of antibody mimics ranging in structural diversity as shown here. 69 Our translocation studies began with the protein synthesis inhibition assay because it is commonly used to probe anthrax toxin entry into cell cytosol thereby allowing us to compare our findings to other reports. However, in order to use antibody mimics to perturb protein-protein interactions, we needed to eliminate the interference of DTA. Therefore, we also investigated the translocation of LFN-antibody mimics (Lvs) in the presence of PA. To facilitate the analysis of Lv delivery, we made extensive use of a reliable western blot approach to monitor full-length cargos in the cytosolic fraction. The cytosolic fraction was obtained using digitonin that selectively permeabilizes the plasma membrane. For all experiments, we confirmed successful extraction of cytosolic proteins from the rest of the cellular components by staining for the presence of Erkl/2 and absence of Rab5. The successful extraction of cytosolic proteins was further validated by the absence of Lvl in the digitonin-extracted fractions from cells treated with Bafilomycin Al or PA[F427H]. Additionally, we observed no Lvl in the total cell lysate under 4'C treatment condition, validating the effectiveness of trypsin digestion in removing surface bound Lv, leaving only intracellular proteins for detection. We detected the conjugates of antibody mimics to LFN by the anti-LF antibody, and confirmed the presence of the antibody mimics at the C-terminus of LFN by the correct molecular weight. Further evidence that fulllength material translocated into the cytosol was the detection of Lvi -HA by the C-terminal HA tag, which gave a similar amount of material detected by the anti-LF antibody. We further used this reliable digitonin extraction and western blot method to gauge the amount of material delivered into the cell cytosol. Based on the linear relationship between the amount of loaded protein and the signal intensity from the anti-LF immunoblot, we were able to estimate the amount of material delivered by comparing the immunoblot signal to a known amount of protein loaded on the same blot. Since the antibody mimics do not interfere with 70 translocation through PA pore, the amount delivered to the cytosol is dependent on the number of PA receptors on cells, which are present on most human cells. Such receptors have been reported to range in copy number from 2,000 to 50,000 and are boosted in certain cancer cells.5 7 Theoretically, one round of translocation for cells harboring 50,000 receptors would give ~20,000 molecules in one cell assuming every seven receptors deliver three copies of LFN-CargO variant. Therefore we estimate, from the use of western blot quantification, that after multiple rounds of translocation we achieve mid-nanomolar concentrations of cargo in the cytosol when accounting for the cell volume. With knowledge of the amount of material delivered to the cytosol, we next studied whether a functional antibody mimic binder can properly refold after cytosolic delivery since translocation through PA pore requires protein unfolding. Based on the co-immunoprecipitation of the monobody with the anti-Abl antibody, we confirmed that the monobody correctly refolds in the cytosolic environment after translocation. The efficiency of refolding and the portion of functional protein is presumably dependent on the cytosolic stability or degradation rates of these variants. For the binders based on monobody and affibody, we observed the functional inhibition of Bcr-Abl kinase and disruption of the MAPK pathway, respectively. Both results again provided evidence that the antibody mimics properly fold and function after translocation into the cytosol. Due to the fact that the amount of tandem monobody delivered was not significantly higher than the Kd for the Bcr-Abl target, the extent of apoptosis caused by Lv5 was not as robust as that achieved by overexpression of the tandem monobody, where its cytosolic concentrations reach a much higher level. The mild biological effect caused by Lv5 is probably due to the high concentration of AbI kinase and also the difficulty to interfere with the intramolecular domain-domain interaction within Abi kinase. In contrast, even though the 71 amount of affibody binder for hRaf- 1 delivered was not much higher than the Kd values for its target, the inhibition of the MAPK pathway by Lv6 was close to that achieved by overexpression of the affibody binder. This effect could be attributed to the lower target concentration and the higher efficiency to disrupt Ras/Raf intermolecular interaction, as compared to the Bcr-Abl target. The efforts to deliver more material or to increase the potency of the delivered cargo for its target are additional challenges in biomolecular delivery. The high adaptability and promiscuity of the PA transporter enabled delivery without the need for protein engineering or screening, unlike delivery methods such as CPPs where a range of sequences or linkers need to be screened for efficient delivery of each cargo. In the case of CPPs, the stability and identity of the peptide transduction sequence may lead to significant differences in delivery efficiency, and oftentimes strong adherence to the cell surface and endosomal membranes blocks cargo escape into the cytosol. For example, it was reported that CPP fusions to DTA were not able to impart delivery into the cell,5 1 even at concentrations a thousand times more than that required for PA-mediated translocation; we confirmed this result. Additionally, we investigated the efficiency of the TAT peptide in the delivery of the antibody mimics. We found that even after treatment of the same cell lines with ten-fold more material, the TAT peptide was not able to deliver any of the four antibody mimics. These results demonstrate the significant improvement in delivery efficiency by PA over the TAT peptide. 58 Additionally, due to the highly inefficient endosomal escape for delivery into the cytosol, ' 59 most delivery methods use high concentrations of these components that can give rise to background cellular toxicity. We studied if the delivery components PA or LFN by themselves or in combination are toxic to cells. Under the conditions reported here, we did not see any toxicity 72 associated with PA/LFN, based on a TUNEL apoptosis assay (Figure 2.2.4) or MTS assay (Figure 2.6.12). In summary, for the first time, we report PA-mediated delivery of antibody mimics into the cell cytosol and successful disruption of critical protein-protein interactions inside cells. As one example, we found that an SH2-binding tandem monobody can be delivered and function as an inhibitor of the oncoprotein Bcr-Abl inside cancer cells. We also demonstrated the inhibition of the Ras/Raf interaction by an affibody, thereby blocking the MAPK pathway that plays a central role in the control of cell proliferation, survival, and growth. The delivery of antibody mimics to the cytosol of cells as indicated by DTA activity and western blot, together with the functional tandem monobody binder to the oncoprotein Bcr-Abl and affibody binder to hRaf-1, supports our belief that the PA-mediated protein delivery system designed by nature will significantly expand the biomolecular delivery toolbox. Our future efforts are to increase the amount of material delivered or to deliver more potent antibody mimics. This delivery platform provides new possibilities to apply modern intrabody technology to disrupt processes in cells. 73 2.4. Experimental 2.4.1. Materials All reagents were purchased from Sigma-Aldrich and Life Technologies, or as otherwise indicated. The following primary and secondary antibodies were used goat anti-LF (bD-17, Santa Cruz Biotechnology), rabbit anti-HA (Sigma), rabbit anti-Abi and agarose beads (K-12, Santa Cruz Biotechnology), rabbit anti-pAbl (pTyr412, Cell Signaling), rabbit anti-Erkl/2, rabbit antiphospho-Erkl/2 (Thr202/Tyr2O4, Cell Signaling), rabbit anti-Rab5 (Cell Signaling), goat antirabbit IRdye 800CW (LI-COR Biosciences), donkey anti-goat IRdye 680LT (LI-COR Biosciences). 2.4.2. One-pot sortagging reaction using Staphylococcus aureus SrtA* Enzyme-mediated ligation using SrtA* was utilized to ligate the proteins to His 6 -SUMO- LFN-DTA-LPSTGG-His 5 or His 6 -SUMO-LFN-LPSTGG-His 6 as previously described (Figure 2.6.1).45 For all ligations, 50 jtM LFN-DTA-LPSTGG-His 5 or LFN-LPSTGG-HiS 6 , 5 iM SrtA*, and 100 to 500 pM G 5-protein were incubated with Ni-NTA beads in sortase buffer (10 mM CaCl 2 , 50 mM Tris-HCl, 150 mM NaCl, pH 7.5) for 30 min at RT nutating. Immediately after 30 min, the Ni-NTA beads, which bound any unreacted starting material, SrtA*-His 6 , His 6 -SUMO, and GG-His 6 , were spun down at 4 'C to minimize the formation of the hydrolyzed side product. The supernatant containing the sortagged product was collected. The beads were washed three times. The supernatant and three washes were followed by an extra step of gel-filtration to separate excess oligoglycine protein reactants. SDS-PAGE gel and LC-MS were used to confirm the purity of the products (Figure 2.6.2, Table 2.6.2, and LC-MS traces). The isolated yields of sortagging ligations were listed in Table 2.6.3. A list of all the protein variants is supplied in Table 2.6.5. 74 2.4.3. Protein synthesis inhibition assay CHO-K1 cells were maintained in F-12K media supplemented with 10% (v/v) fetal bovine serum at 37 'C and 5% CO 2. The cells were plated in 96-well plate at a density of approximately 30,000 per well 16 hours prior to the assay. The LDvs were prepared in ten-fold serial dilutions in 50 ptL, and added 50 tL of F-12K media containing 20 nM protective antigen . (PA8 3). The 100 p.L samples were added to CHO-Ki cells for 30 minutes at 37 'C and 5% CO 2 The cells were washed three times with PBS and 100 [LL leucine-free F-12K medium supplemented with 1 pCi mL)' 3H-leucine (Perkin Elmer, MA) was added for 1 h at 37 'C 5% CO 2 . The cells were washed three times with PBS and suspended in 150 tL of scintillation fluid. 3H-Leu incorporation into cellular proteins was measured to determine the inhibition of protein synthesis by LFN-DTA. The scintillation counts from cells treated with only PA were used as control value for normalization. Each experiment was done in triplicate. The data were fitted using Origin8 using Sigmoidal Boltzmann Fit using equation: Y= A2+ A, -A. where xO represents the logEC5o values (Table 2.6.4). 2.4.4. Uptake of Lvs or TAT-HA-1 to -4 in CHO-K1 cells CHO-Ki cells were plated in 12-well plates 16 hours prior to the treatment. For Lvs, cells were treated with 250 nM Lv in the presence of 40 nM PA in complete medium overnight at 37 'C and 5% CO 2 . For TAT-HA-antibody mimic constructs, cells treated with 2.5 pM TATHA-i to -4 and - 4mut in serum free medium for 4 h at 37 'C and 5% CO 2 . Translocation controls included PA[F427H] instead of PA, addition of 200 nM Bafilomycin Al, or incubation at 4 'C instead of 37 'C. 75 2.4.5. Cytosolic protein extraction and whole cell lysate preparation After uptake of the antibody mimics, cells were washed with PBS, detached and digested with 0.25% Trypsin-EDTA for 5 minutes at 37 'C to remove surface bound proteins, then washed twice with PBS. The cell pellets were then subjected to cytosolic extraction or whole cell lysate preparation. For cytosolic protein extraction, 1 x 106 cells were resuspended in 100 uL of 50 pg mL 1 digitonin in 75 mM NaCl, 1 mM NaH 2 PO 4, 8 mM Na 2HPO 4, 250 mM sucrose supplemented with Roche protease inhibitor cocktail for 10 min on ice, and centrifuged for 5 minutes at 13,000 rpm. For whole cell lysate, cells were lysed in IP lysis buffer (25 mM Tris, 150 mM NaCl, 1% v/v NP-40, pH 7.5) supplemented with Roche protease inhibitor cocktail on ice for 30 minutes and centrifuged for 10 minutes at 13,000 rpm. Both supernatants were collected for blotting and analysis. 2.4.6. Western blot Transfer was done with TE 70 Semi-Dry Transfer Unit (GE) onto nitrocellulose membrane (Whatman) using buffer containing 48 mM Tris, 39 mM glycine, 0.0375% SDS, 20% methanol. The membrane was blocked at room temperature for 2 hours with LI-COR blocking buffer and was then incubated with goat anti-LF antibody or rabbit anti-HA in LI-COR blocking buffer overnight at 4 'C. The membranes were washed with TBST (50 mM Tris, 150 mM NaCl, 0.1% Tween-20) and blotted with secondary antibody conjugated with IRdye from LI-COR and imaged with LI-COR Odyssey Infrared Imaging System. The western blot images were analyzed and quantified using the Image Studio Lite program (LI-COR). 2.4.7. Co-immunoprecipitation of Lv5 with Abl kinase K562 cells were treated with 40 nM PA or PA[F427H] and 50 nM LFN-antibody mimic for 24 hours. Approximately I x106 cells were trypsinized, washed with PBS, frozen at -80 'C, 76 and then lysed in 500 gL IP lysis buffer supplemented with Roche protease inhibitor cocktail on ice for 30 min. After centrifugation at 13,000 rpm for 15 min, 450 pL lysates were incubated with 12.5 pg anti-Abi agarose beads for 4 hours. The resulting immune complexes were washed three times with lysis buffer and once with lysis buffer without NP-40. The bound proteins were eluted with 0.2% SDS and 0.1% Tween 20 and subjected to SDS-PAGE separation. Standard immnuoblotting was performed using anti-Abl or anti-LF antibody. 2.4.8. TUNEL assay with Lv5 in K562 cells K562 cells were plated in a 24-well plate at a density of 150,000 cells mUl and treated with 500 nM LFN-antibody mimic and 60 nM PA or PA[F427H] in serum-free medium for 1 day. FBS (10%) was added to the medium and the cells were treated for additional two days. Cells treated with 1 pM imatinib served as the positive control. The cells were fixed with 3.2% paraformaldehyde, treated with in 90% methanol on ice or stored at -20 'C, followed by TUNEL staining according to the manufacturer's instructions (Life Technologies). 2.4.9. Transfection of HEK 293T with pcDNA3-ABRaf HEK 293T cells were plated in 24 well plates overnight to reach ~90% confluency. The cells were then transfected with the plasmid pcDNA3-ABRaf-GFP using Lipofectamine 2000 (Life Technologies). The medium was changed after 5 hours of transfection. The total transfection time was 24 hours. For the pErkl/2 analysis, the cells were starved for 12 hours before they were treated with 5 ng mL-' EGF for 7 minutes, washed with cold PBS, and lysed in plate with IP lysis buffer containing 1% NP-40, Roche protease inhibitor cocktail, and Roche phosSTOP. The lysates were subjected to SDS-PAGE separation, transferred onto nitrocellulose membrane, and blocked with 5% BSA in the presence of Na 3VO 4 and NaF. The immunoblotting with anti-pErkl/2 antibody (Cell Signaling) was performed in 3% BSA in the presence of 77 Na 3VO 4 and NaF overnight. After stripping the membrane with Restore Plus Western Blot Stripping Buffer (Thermo Scientific), the membrane was blocked again and immunoblotted with anti-Erkl/2 antibody (Cell Signaling). 2.4.10. Delivery of Lv6 into HEK 293T cells HEK 293T cells were plated in 24 well plates overnight to reach ~80% confluency. The cells were then treated with 80 nM PA and 500 nM Lv6 in serum-free medium for 12 hours. For detection of cytosolic Lv6, the cells were detached with trypsin, washed with PBS, and resuspended in 100 ptg mLU digitonin for 10 minutes as described above. For detection of pErkl/2, the cells were treated with 5 ng mL- EGF for 7 minutes, washed with cold PBS, lysed in plate with IP lysis buffer containing 1% NP-40, Roche protease inhibitor cocktail, and Roche phosSTOP. The immunoblotting with anti-pErkl/2 and anti-Erkl/2 were as described above. 2.4.11. Construction of plasmids for recombinant proteins and transfection The gene for Abl-SH2 was purchased from Addgene (pDONR223-ABL1, 23939).0 The genes for affibody," ABRaf, DARPin," and 7cl212 were purchased from DNA2.0 (Menlo Park, CA). The gene for 1OFN3 was kindly provided by K. Dane Wittrup (MIT, Department of Biological Engineering). The 1OFN3 mutant, HA463 was generated by mutagenesis. The amino acid sequences are shown in Figure 2.6.13. The constructs, pET SUMO-G 5-affibody, pET SUMO-G 5 -lOFN3, and pET SUMO-SH2, were prepared using the Champion pET SUMO protein expression system (Life Technologies). AccuPrime Taq DNA polymerase (Life Technologies) was used to PCR amplify the DNA using the primers listed in Table 2.6.1. pET SUMO-G 5 -HA4 with two restriction sites after the stop codon (BamHI and XhoI) was generated by mutagenesis based on pET SUMO-G 5 -1OFN3. 7c12 was PCR amplified, digested with BamHI and XhoI, and ligated in pET SUMO-G 5 -HA4 vector with a 20-amino acid linker 78 GGSG(GGSGG) 3G between HA4 and 7c12. For pET SUMO-LFN-DTA (C186S)-LPSTGG-His 5 and pET SUMO-LFN-LPSTGG-His 6, a G2S mutation was introduced to minimize undesired loss of Methionine. The plasmid for transfection of eGFP-ABRaf was generated by a two-step ligation. eGFP was inserted into pcDNA3 using NotI and BamHI restriction sites and ABRaf was then inserted using BamHI and XhoI restriction sites, with a GSGGGGSGGGGG between eGFP and ABRaf. 2.4.12. Protein expression and purification His6-SUMO-Gs-affibody, His 6-SUMO-G 5-GB 1, His 6-SUMO-G 5-DARPin, His 6 -SUMOG 5-HA4, His 6 -SUMO-G 5 -HA4-7c 12, His6-SUMO-G 5 -HA4-7c 1 2mut and His6 -SUMO-SH2 were expressed in F. coli BL21 (DE3) cells in 1 L LB culture and approximately 5 g of cell pellet was obtained for each. HiS 6-SUMO-LFN-DTA (C186S)-LPSTGG-His 5 , His 6 -SUMO-LFN-LPSTGGHis 6 , His 6-SUMO-LFN-DTA (C186S), SrtA*-His 6 , WT anthrax protective antigen (PA), and PA[F427H] were expressed in . coli BL21 (DE3) cells at New England Regional Center of Excellence/Biodefense and Emerging Infectious Diseases (NERCE). For each purification, approximately 40 g of cell pellet was lysed by sonication in 100 ml of 50 mM Tris-HCl, 150 mM NaCl, pH 7.5 buffer containing 200 mg lysozyme, 4 mg Roche DNAase I, and 2 tablet of Roche protease inhibitor cocktail. The suspension was centrifuged at 17,000 rpm for 40 minutes. The supernatant was loaded onto three 5 ml HisTrap FF crude Ni-NTA column (GE Healthcare, UK) and washed with 100 mL of 20 mM Tris-HCl, 150 mM NaCl, at pH 8.5 and 100 mL of 40 mM imidazole in 20 mM Tris-HCl, 500 mM NaCl, at pH 8.5. The protein was eluted from the column with buffer containing 500 mM imidazole in 20 mM Tris-HCl, 500 mM NaCl, pH 8.5, and buffer exchanged into 20 mM Tris-HCI, 150 mM NaCl, pH 8.5 using a HiPrep 26/10 79 Desalting column (GE Healthcare). WT PA and PA[F427H] was overexpressed in the periplasm of E. coli BL21 (DE3) cells and purified by anion exchange chromatography." 2.4.13. Synthesis of TAT peptide and native chemical ligation (NCL) of TAT-DTA The TAT peptide (YGRKKRRQRRRLLG) 2 was synthesized by Fmoc solid phase peptide synthesis on hydrazide resin (calculated mass: 1856.1 Da; observed monoisotopic mass: 1856.1 0.1 ). The crude peptides were purified by preparative RP-HPLC on Agilent Zorbax 300SB C18 column (9.4 x 250 mm, 5 jim). The hydrazide C-terminus was converted to an MPAA thioester according to the previously published method by Fang, et al.65 TAT-MPAA was ligated to 'C-DTA by NCL using 1.4 mM TAT-MPAA and 270 ptM 1 C-DTA in 100 mM phosphate buffer pH 7.0, 150 mM NaCl, 20 mM TCEP in 4 h. The reaction was quenched with 200 mM sodium 2-mercaptoethane sulfonate and desalted to 20 mM Tris buffer pH 7.5, 150 mM NaCl. To alkylate the cysteine, 50 mM bromoacetamide was added to react for 10 minutes. The reaction was again quenched with 200 mM sodium 2-mercaptoethane sulfonate and buffer exchanged into 20 mM Tris buffer pH 7.5, 150 mM NaCl to obtain TAT-DTA protein. 2.4.14. Synthesis of TAT-HA-LPSTGG peptide and sortagging to G 5 -proteins The TAT-HA-LPSTGG peptide (AGYGRKKRRQRRRGYPYDVPDYAGLPSTGG) was synthesized by Fmoc solid phase peptide synthesis on aminomethyl resin (calculated mass: 3397.8 Da; observed average mass: 3397.8 0.1). The crude peptides were purified by - preparative RP-HPLC on Agilent Zorbax 300SB C18 column (9.4 x 250 mm, 5 pm). The G5 proteins (1 - 4 and 4mut) were conjugated to TAT-HA-LPSTGG using 10 [tM SrtA* in sortase buffer (10 mM CaCl 2, 50 mM Tris-HCl, 150 mM NaCl, pH 7.5) for 1 h at RT. For each reaction, - the TAT-HA-LPSTGG peptide (230 - 600 jM) was used in two-fold excess compared to the G5 protein (115 - 300 pM). After 1 h, all of ligation reactions were diluted approximately 20-fold 80 with 20 mM Tris-HCl pH 8.5. The samples were purified by anion exchange chromatography (GE HiTrap Q HP) at 0-30% 1 M NaCl in 20 mM Tris-HCl pH 8.5 over 100 mL. The fractions containing the ligated product were concentrated using a 3 kDa or 10 kDa concentrator. For the TAT-HA-I (affibody) ligation, the product eluted with TAT-HA-LPSTGG peptide in the flow through, which was subsequently concentrated and buffer exchanged three times to remove excess peptide. Purity of proteins was analyzed by LC-MS and SDS-PAGE gel (Figure 2.6.5). 2.5. Acknowledgements This research was generously sponsored by MIT start-up funds, MIT Reed Fund, and Damon Runyon Cancer Research Foundation award for B.L.P, and National Science Foundation Graduate Research Fellowship for A.E.R. We would like to thank R. J. Collier (Harvard) for his encouragement, support, and some of the laboratory equipment. We are indebted to the NERCE facility (grant: U54 A1057159) for expressing the toxin proteins. Provisional patent applications covering material reported in this manuscript were filed by MIT-TLO. We thank Jingjing Ling, Mike Lu, Dan Chinnapen, and Daphne van Scheppingen for their helpful discussions. We also thank Dane Wittrup for providing the fibronectin clone. 81 2.6. Appendix H6-SUMO- LPSTGG-H 5 D@i) 30 min, RT H-SUMO + LPSTGG-H5 G 30 min, RT NI-NTA SrtA* 40 _( His tagged reagents LPSTG5- Figure 2.6.1. One-pot sortagging reaction. N-terminal small ubiquitin-like modifier (SUMO) was first cleaved from the protein substrate, His6-SUMO-LFN-DTA-LPSTGG-His , using 1 g 5 SUMO protease per mg of protein substrate at RT for 30 minutes to expose the native Nterminus of LFN-DTA-LPSTGG-His 5 , then the sortagging reaction was carried out as described. 82 i nv 1 1nval 1 Dv4 Lnvs LnvR Lv1 Lv2 Lv3 Lv4,Ut Lv5 Lv6 Figure 2.6.2. Coomassie stained SDS-page gel of LDvs and Lvs obtained after sortagging and purification. 83 -1000-2000 1OFN3 -4-HA4 -30002 -400 -5000-6000 z -7000-8000 20 30 40 50 60 70 80 Temperature (*C) 90 100 Figure 2.6.3. Thermal stability of I0FN3 (4) and HA4 (4mut) was monitored by circular dichroism (CD) spectroscopy. CD spectra were recorded on an Aviv model 202 instrument at 25 'C. A 1 mm path length cell was used. The proteins were prepared by dissolving 0.06 mg in 50 mM sodium sulfate and 5 mM Tris-HCl at pH 8.5. The molar ellipticity (0 in deg cm 2 dmol-') was calculated by [N]x = Oobs x 1/(10 lcn), where Oobs = observed ellipticity at k, 1 = path length (cm), c = concentration of protein (M), n = # of amino acids. To monitor the thermal stability, the ellipticity at 216 nm was measured from 25 'C to 95 'C. 84 3 4 4t + + + 2 + Lv 2 (4 ng) PA anti-LF anti-Erki/2 anti-Rab5 Figure 2.6.4. Western blot analysis of delivered Lvs. CHO-Ki cells were treated with 250 nM Lv2-4 and 4 mut in the presence of 40 nM PA for 12 hours. After treatment, cells were detached with trypsin and the cytosolic proteins were extracted with digitonin buffer and analyzed by immunoblotting. Lane 1 contains 4 ng of Lv2. 85 TAT-HA 1 2 3 4 4mut kDa 62 49 38 14 - 28 6 Figure 2.6.5. Coomassie stained SDS-PAGE gel of TAT-HA-I to -4 and -4mut (1 pg). 86 Total cell lysate a) 1 TAT-HA4 3 2 4,U1 b) 2 Lv1 -HA 4*C + PA Ud Total cell lysate 4 ng Lvl-HA TAT-HA- - 1 2 3 4,t 4*C 2 4 ng 2 38 anti-HA28 17 14 anti-HA (increase intensity) 17j 14 anti-HA I ng TAT-HAanti-Erk1/2 %004 anti-Rab5 anti-Erkl/2 anti-Rab5 Figure 2.6.6. Cellular uptake of TAT-HA-I to -4 and -4mut. CHO-KI cells were treated with 2.5 pM TAT-HA-1 to -3 and -4mut in serum-free medium for 4 hours (a) or 16 hours (b) then lifted with trypsin and washed with PBS. The total lysates were analyzed by immunoblotting. Inset showed even with increased anti-HA intensity there was little amount of TAT-HA-2 and -3 detectable (a). 87 LDv5 1.2- LPSTG HA4 7c12 1.00-8- 0.8LF.-DTA -4LDv5 5 0.2 U- 0.0 -16 -14 -12 -10 -8 Log [Protein Concentration (M)I -6 Figure 2.6.7. The level of protein synthesis inhibition in K562 cells treated with varying concentrations of LFN-DTA or LDv5 in the presence of 20 nM PA for 4 hours. 88 a b Lv5 2500- -100 2000- 180C nM -50 nM -25 nM - 12.5 nM 1400 1500- HA4-7c12 -100 nM -5o nM - 1000- . CE 1000- 600- 500- 2000- 0- -100 0 100 Time (s) 200 300 40 -100 110 200 300 400 Time (s' Figure 2.6.8. SPR curves for SUMO-SH2 and varying concentrations of Lv5 (a) or HA4-7cl2 (b). SPR measurements were taken in 10 mM HEPES, pH 7.4, 150 mM NaCl, 0.005% P20, and 3 mM EDTA with a BIAcore3000 instrument. The SUMO fusion of Abl SH2 domain was immobilized to an ethyl(dimethylaminopropyl) carbodiimide/N-hydroxysuccinimide activated CM5 sensor chip. Purified HA4-7c 12 or Lv5 was then flowed over the sensor chip at a rate of 20 pl min-' and the association and dissociation kinetics were monitored. Surface was regenerated by 10 mM glycine pH 2.0 between different runs. Data was fit to a 1:1 Langmuir binding model using multiple kinetic traces with the BlAevaluation software. The Kd was calculated as 12 nM for Lv5 (a) and 6 nM for HA4-7c 12 (b). 89 Lv5 (ng) 0.1 0.25 1.0 2.5 5.0 7.5 anti-LF 5000000. Lv5 40000003000000 8 200000010000000 0 1 2 3 4 5 Lv (ng) 6 7 8 Figure 2.6.9. Linear relationship between signal intensity of each band and the amount of protein loaded (Lv5: y = 378000x - 137000 R2 = 0.989). The western blot images were analyzed in Image Studio Lite program (LI-COR Biosciences). The signal intensity of each band was compared to that of 0.1-7.5 ng of pure protein loaded on the same blot. The amount of protein in each lane was calculated based on the linear relationship between the signal intensity and the amount of protein. 90 b a 4 ng Lv5 PA PA[F427H] Lv5 + 1.2 Lv5mut - + 0 U5 - *Lv5mut - 0.8 anti-pAbI (pY4l 2) anti-LF 0.4 anti-GAPDH Z + PA + PA[F427H] Figure 2.6.10. Phosphorylation analysis of Bcr-Abl. a) Western blot analysis of Lv5 and BcrAbl pY421. b) The Bcr-Abl pY412 intensity was normalized to GAPDH and was set to 1 in the non-treated condition. K562 cells were treated with 500 nM Lv5 and 60 nM PA or PA[F427H] in serum-free medium for 1 day and FBS (10% v/v) was added to treat the cells for an additional day. Cells were lysed and immunoblotted with antibodies recognizing Abl phosphorylated on Tyr412 or LF. The amount of Lv5 on the blot was quantified as 1 ng (a). The estimation of the molecules per cell and concentration of Lv5 in the cell was based on cell number of ~ 100,000 cells and cell volume of- 2 pL. 91 I- 0652 Cell 45 2.. Imatinib 101054 104 10 3 103 - - 104 - 103 10 10 2 - IL 102 03 0,2 3 0 104 102 PerCP-A 105 0 102 103 PerCP-A 104 105 Figure 2.6.11. Apoptosis measurement of K562 cells that were not treated or treated with 1 RM imatinib for 3 days. PerCP indicates the DNA fraction and FITC reports the TUNEL positive fraction by detecting the AlexaFluor 488 dye-labeled anti-BrdU antibody. 92 120- 120- 100. 100. -8080 60. 40 -11-LFN-DTA only -4-LFN-DTA + 20 nM PA 2 > 60 2 40- -0-Lv1 +20 nM PA -4-Lv only orgy -PA 2 20 20- 0 1E-4 1E-3 0.01 0.1 Protein Concentration (nM) 10 100.1000 Protein Concenration (nM) Figure 2.6.12. MTS cell viability assay. CHO-Ki cells were plated in a 96-well plate at 2,500 cells per well and treated with LFN-DTA or Lvi with or without PA, or with PA at the indicated concentration for 48 hours. CellTiter 96@ Aqueous One Solution Cell Proliferation Assay (Promega, 10 ptL) was added to each well and incubated for 1 hour according the manufacturer's instructions. The absorbance at 490 nm was normalized to 1 nM LFN-DTA + PA (set as 0) and untreated cells (set as 100). 93 Affibody Val 'AspAsnLysPheAsnLysGuGnGln 0 AsnAlaPheTyrGluleLeuHisLeuPro 2O Asn LeuAsnGluGluGlnArgAsnAlaPhe Ser 4 1AlaAsnLeuLeuAlaGluAlaLysLys 30 lleGlnSerLeuLysAspAspProSerGln 40 50LeuAsnAspAlaGlnAlaProLys GB1 Met]ThrTyrLysLeulleLeuAsnGlyLys 0 ThrLeuLysGIyGluThrThrThrGluAla 20 44 Val AspAlaAlaThrAlaGluLysValPhe 30 LysGlnTyrAlaAsnAspAsnGlyValAsp Gly 4'GluTrpThrTyrAspAspAlaThrLys 50 ThrPheThrValThrGlu DAPRin Ser'AspLeuGlyLysLysLeuLeuGluAlal0 AlaArgAlaGlyGlnAspAspGluValArg 21 Ile LeuMetAlaAsnGlyAlaAspValAsn 30 His 4LeuAlaAlaPheLeuGlyHisLeuGlu 50IleVatGIuValLeuLeuLysTyrGlyAla 60 AlaTyrAspAspAsnGlyValThrProLeu Asp 61ValAsnAlaAlaAspSerTrpGyThr 70 ThrProLeuHisLeuAlaAlaThrTrpGy 20 40 80 His 81LeuGlulteVaIGluValLeuLeuLys oHisGlyAlaAspValAsnAlaGtnAspLysloo Phe 10 1GlyLysThrAlaPheAspIIeSerIle' ""AspAsnGlyAsnGluAspLeuAlaGlulle Leu1 GnLysLeuAsn 1OFN3 Val SerAspValProArgAspLeuGluVallValAlaAlaThrProThrSerLeuLeulle 20 Ser TrpAspAlaProAlaValThrValArg 0 TyrTyrArgleThrTyrGlyGluThrGly 40 Gly 4 1AsnSerProValGlnGluPheThrVal 50 ProGlySerLysSerThrAlaThrIleSer6 0 Gly 6 LeuLysProGlyValAspTyrThrlle 0 ThrValTyrAlaValThrGlyArgGlyAsp 94 20 Ser8 1ProAlaSerSerLysProlleSerl le 9 AsnTyrArgThrGlulleAspLysProSer 00 Gin' 01' HA4 Val' SerSerValProThrLysLeuGluVal '0 ValAlaAlaThrProThrSerLeuLeule 20 Ser2 1TrpAspAlaProMetSerSerSerSer 30ValTyrTyrTyrArgleThrTyrGlyGlu 40 Thr 41GlyGlyAsnSerProValGlnGluPhe 50 ThrValProTyrSerSerSerThrAlaThr60 Ile 6 1SerGlyLeuSerProGlyValAspTyr 7 0 ThrIleThrValTyrAlaTrpGlyGluAsp 80 Ser 8 AlaGlyTyrMetPheMetTyrSerPro 9 lteSerIleAsnTyrArgThr 7c12 Gly' GlySerGlySerSerValSerSerVal 0 ProThrAsnLeuGluValValAspAlaThr 20 Pro 2 1ThrSerLeuLys1leSerTrpAspAla 3 0TyrTyrSerSerTrpGlnAsnValLysTyr40 Tyr 4 t ArgIleThrTyrGlyGluThrGlyGly 5 "AspSerProValGlnGluPheThrValProO Gly 61TyrTyrSerThrAlaThrIeSerG ly 7 LeuSerProGlyValAspTyrThrIleThr 80 Val 8 'TyrAlaTyrAspThrPhePheProGly 90 TyrGluProAsnSerProlleSerIleAsn 00 Tyr'0 1ArgThr ABRaf Val' AspAsnLysPheAsnLysGluValAsn 0 LeuValAlaAspG lu1leTrpLeuLeuPro 20 Asn2 ' LeuAsnAsnGlnGlnValTrpAlaPhe 30 IleThrSerLeuLysAspAspProSerG In40 Ser4' AlaAsnLeuLeuAlaGluAlaLysLys 50 LeuAsnAspAlaGlnGluProLys Figure 2.6.13. Protein sequences 95 Table 2.6.1. PCR primers 1 OFN3 fwd 5'-GGCGGTGGCGGTGGCGTTTCTGATGTTCCGAGGGAC-3' 1 OFN3 rev 5'-TAACTGGGATGGTTTGTCAATTTCTGTTC-3' Affibody fwd 5'-GGCGGTGGCGGTGGC GTCGACAATAAGTTCAATAAAGAACAACAG-3' Affibody rev 5'-TTACTTCGGCGCTTGTGCG-3' GB 1 fwd 5'-GGCGGTGGCGGTGGCATGACGTACAAACTGATCCT GAACGG-3' GB1 rev 5' ttaTTCGGTTACCGTGAAGGTTTTGG-3' DARPin fwd 5'-GGCGGTGGCGGTGGATCCGATCTGGGCAAGAAAC-3' DARPin rev 5'- ttaGTTCAGTTTCTGCAGAATCTCC-3' SH2 fwd 5'-TCCCTGGAGAAACACTCCTGGTACC-3' SH2 rev 5'-TTAGTTGCGCTTTGGGGCTG-3' 7c 12 fwd 5'-TTTTTGGATCCGGCGGCGGTTCTGGTGGCGGTGGTAGCG-3' 7c12 Rv 5'- TTTTTCTCGAGTTAGGTGCGATAATTGATGCTAATCGG-3' 96 Table 2.6.2. Observed molecular masses of expressed protein constructs when analyzed by LCMS Calculated MW Observed MW (Da) Protein (Da; average) SrtA*-His 6 19214.6 0.4 19214.5 LFN-DTA 52047.0 0.4 52046.2 66700.0 0.4 66700.1 45289.8 0.4 45289.8 WT PA 83754.1 0.4 83751.6 PA[F427H] 83742.4 0.4 83738.3 Affibody 6925.8 0.2 6925.6 GB1 6481.4 0.2 6481.0 DARPin 13701.5 0.2 13701.3 10FN3 11022.4 0.2 11022.2 HA4 10977.1 0.2 10977.1 HA4-7c12 23244.1 0.4 23244.2 HA4-7c12mut 23099.1 0.4 23099.1 SUMO-SH2 24626.6 0.4 24626.4 ABRaf 6864.9 His 6-SUMO-LFN-DTA-LPSTGG-His His 6 -SUMO-LFN-LPSTGG-His C-DTA 6 5 20840.60 DTA 20897.5 TAT-DTA 22722.84 0.2 0.4 0.4 0.4 6864.6 20840.1 20898.2 22723.4 97 Table 2.6.3. Isolated yields of sortagging ligations from SrtA* reaction. Protein Isolated Yield (%) LDv1 21 LDv2 33 LDv3 34 LDv4 44 LDv4mut 28 LDv5 22 Lvi 42 Lv2 47 Lv3 38 Lv4 44 Lv4mut 35 Lv5 37 Lv5mut 38 Lv6 42 98 Table 2.6.4. EC50 values of 30-minute protein synthesis inhibition assay (Figure 2.1.1d). The errors represent fitting errors from Sigmoidal Boltzmann Fit. Protein EC50 (pM) LFN-DTA 41 8 LDvl 33 11 LDv2 56 29 LDv3 62 9 LDv4 532 LDv4mut 38 209 4 99 Table 2.6.5. List of variants Abbreviation LDvIPT Variant Ctr C-term LDv2 (affibody) C-tr LPSTG __ _ _ LDv3 (GB1) PS*G" (DARPin) LDv4 _____ ____(FN _____ 41V~t PSTG, C-term LDv4mut F33) t OD-( TALPSTG Cer (HA4) LDv5 4___"" Lvi PSTG C term Lv2 (HA4-7c12) (affibody) C LPSTG(B (DARPin) Lv4 PSTG, ~LPSTG> C term (IOFN3) LV4mut G Cterm (HA4) Lv5 _____ (HA4-7cl2) Lv5mut LV~ m t ~ -term (HA4(Y87A)-7c12(Y62E/F87K)) Lv6 W LPSTG .1 ""(ABRaf) 100 2.6.1. LC-MS Traces LDvl obs.59353.1 ca 593522 "'4-T)-PSTG5-- . C-term 62500 57600 2 1 6 5 4 3 8 7 9 0.4 C 11 10 Tlme 12 (mini 13 16 15 14 19 16 17 21 20 LDv2 obs.58907.7 t OA4 ca. C LP"TG GW( 58907.7 -lerm 57500 60000 2 1 3 4 6 5 7 8 9 10 f3 12 11 rime [min) 15 14 16 17 19 18 20 21 LDv3 obs.66128.8 66127.9 0.4 ca. S S PSTG ;~:W0 L7vsw 1 Time 11 10 9 8 7 6 5 4 3 2 lmn LDv4 0.4 obs.63449 7 ca. 63448.8 DA -WtAC-term PSTG,, 62560 1 2 3 4 5 6 7 8 9 10 12 13 11 ime [mini 14 15 16 17 S000 18 19 20 21 101 LDv4mut obs.64112.2 0.4 64111.5 ca. MD-CDT ALPSTG -- C-term ca. 7567059 72500 77500 1 2 3 4 5 6 7 8 9 10 11 ime (fln] Lvl obs.3780b 5 ca 37805 4 PSTG C-term 35000 1 2 3 0.4 4 5 6 7 8 9 10 11 Time 13 12 14 15 16 17 40000 18 19 20 21 [minj Lv2 obs.37361.6 ca. 37360.8 C-term PSTG,""0 ] 30040000 1 2 3 4 5 6 Time [min) 102 7 8 9 10 11 *OA. Lv3 obs 44581.4 20.4 ca. 44581 1 PSTG, 4:0e.4 sA 4 A4 5 4 3 2 1 9 8 7 6 Time [min] 11 10 Lv4 obs.41902.5 * 0.4 ca. 41902.0 C-term LPSTG 45000 40000 1 3 2 4 5 6 8 7 10 9 12 11 13 14 15 17 16 18 19 21 20 Time [min Lv4mut 0.4 obs.418571 ca. 41856.9 LPSTG Cterm 4000 1 3 2 4 5 6 8 7 10 9 11 12 Time [min] 13 14 15 17 16 18 45000 19 21 20 Lv5 obs.54124.9 t. 0.4 ca 541240 52589 96089 1 2 3 4 5 6 7 8 9 11 10 Time [min] Lv 5 mut obs ca, 97,9 * A4 3978,9 62SW0 48000 1 2 3 4 5 6 7 9 1 11 ime (minj 103 Lv6 obs.37745.0 37744.4 ca. LPSTG 1 0.4 * JA pO 2 3 4 5 6 Time [minj 7 8 9 11 10 Lvl-HA obs23 39.1 T ca. 388W95 W , PSTG,-$ C-term 3750040000 TTH i 1 2 3 4 5 r 6 7 Time [min) 8 9 10 11 TAT-HA-1 obs.10192.5 Ctermca. LPST~'- 10192.2 9000 1 2 3 4 5 6 7 Time [min) 8 9 10000 1low 10 11 obs.9747.9 02 TAT-HA-2 -termca. LPSTG 1 104 2 3 4 5 6 ime [min] 9747.6 7 8 9 10 11 *0,2 0.4 TAT-HA-3 obs.16965.0 ca. 16967 9 0.2 LPSTG- 16"0 17000 lebo0 1 6 5 4 3 2 1 7 Time 9 10 1l (min) TAT-HA-4 obs. 14289.0 i0 2 ca -term LWPSTG, 1L3000 3 1 14289.2 4 5 6 Time (min1 7 8 9 14000 15000 10 11 TAT-HA-4mut obs 13950.7 0.2 ca. 13950.4 L - , PSTG "C-term 13060 14000 15000 2 3 4 5 6 7 9 9 10 11 Time [min) 105 2.7. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 106 Caravella, J. & Lugovskoy, A. Design of next-generation protein therapeutics. Curr. Opin. Chem. Biol. 14, 520-528 (2010). Gilbreth, R.N. & Koide, S. Structural insights for engineering binding proteins based on non-antibody scaffolds. Curr. Opin. Struct. Biol. 22, 413-420 (2012). Koide, S., Koide, A. & Lipovsek, D. in Methods Enzymol. (eds. Wittrup, K.D. & Verdine, G.L.) 135-156 (Elsevier Academic Press Inc, San Diego, 2012). Tamaskovic, R., Simon, M., Stefan, N., Schwill, M. & Pluckthun, A. in Methods Enzymol. (eds. Wittrup, K.D. & Verdine, G.L.) 101-134 (Elsevier Academic Press Inc, San Diego, 2012). Lofblom, J. et al. Affibody molecules: Engineered proteins for therapeutic, diagnostic and biotechnological applications. FEBS Letters 584, 2670-2680 (2010). Mandal, K. et al. Chemical synthesis and X-ray structure of a heterochiral {D-protein antagonist plus vascular endothelial growth factor} protein complex by racemic crystallography. Proc. Natl. Acad. Sci. U S. A. 109, 14779-14784 (2012). Schweizer, A. et al. Inhibition of caspase-2 by a designed ankyrin repeat protein: Specificity, structure, and inhibition mechanism. Structure 15, 625-636 (2007). Schroeder, T. et al. Specific Inhibition of Caspase-3 by a Competitive DARPin: Molecular Mimicry between Native and Designed Inhibitors. Structure 21, 277-289 (2013). Grimm, S., Salahshour, S. & Nygren, P.A. Monitored whole gene in vitro evolution of an anti-hRaf-1 affibody molecule towards increased binding affinity. New Biotechnology 29, 534-542 (2012). Parizek, P. et al. Designed Ankyrin Repeat Proteins (DARPins) as Novel IsoformSpecific Intracellular Inhibitors of c-Jun N-Terminal Kinases. ACS Chem. Biol. 7, 13561366 (2012). Kummer, L. et al. Structural and functional analysis of phosphorylation-specific binders of the kinase ERK from designed ankyrin repeat protein libraries. Proc. Natl. Acad. Sci. U S. A. 109, E2248-E2257 (2012). Grebien, F. et al. Targeting the SH2-Kinase Interface in Bcr-Abl Inhibits Leukemogenesis. Cell 147, 306-319 (2011). Wojcik, J. et al. A potent and highly specific FN3 monobody inhibitor of the Abl SH2 domain. Nat. Struct. Mol. Biol. 17, 519-U173 (2010). Lundberg, E., Brismar, H. & Graslund, T. Selection and characterization of Affibody (R) ligands to the transcription factor c-Jun. Biotechnol. Appl. Biochem. 52, 17-27 (2009). in Methods Enzymol. (eds. Wittrup, K.D. & Verdine, G.L.) 1-344 (2012). in Methods Enzymol. (eds. Wittrup, K.D. & Verdine, G.L.) 1-348 (2012). Moore, S.J. et al. Engineered knottin peptide enables noninvasive optical imaging of intracranial medulloblastoma. Proc. Natl. Acad. Sci. U S. A. 110, 14598-14603 (2013). Marschall, A.L.J., Frenzel, A., Schirrmann, T., Schungel, M. & Dubel, S. Targeting antibodies to the cytoplasm. mAbs 3, 3-16 (2011). Zelphati, 0. et al. Intracellular delivery of proteins with a new lipid-mediated delivery system. Journalof Biological Chemistry 276, 35103-35110 (2001). 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Hasadsri, L., Kreuter, J., Hattori, H., Iwasaki, T. & George, J.M. Functional Protein Delivery into Neurons Using Polymeric Nanoparticles. Journalof Biological Chemistry 284, 6972-6981 (2009). Kamaly, N., Xiao, Z.Y., Valencia, P.M., Radovic-Moreno, A.F. & Farokhzad, O.C. Targeted polymeric therapeutic nanoparticles: design, development and clinical translation. Chem. Soc. Rev. 41, 2971-3010 (2012). Gu, Z., Biswas, A., Zhao, M.X. & Tang, Y. Tailoring nanocarriers for intracellular protein delivery. Chem. Soc. Rev. 40, 3638-3655 (2011). Cronican, J.J. et al. Potent Delivery of Functional Proteins into Mammalian Cells in Vitro and in Vivo Using a Supercharged Protein. Acs Chemical Biology 5, 747-752 (2010). Thompson, D.B., Cronican, J.J. & Liu, D.R. in Methods Enzymol.: Protein Engineering for Therapeutics, Vol 203, Pt B (eds. Wittrup, K.D. & Verdine, G.L.) 293-319 (Elsevier Academic Press Inc, San Diego, 2012). Caron, N.J. et al. Intracellular delivery of a Tat-eGFP fusion protein into muscle cells. Molecular Therapy 3, 310-318 (2001). Foerg, C. & Merkle, H.P. On the biomedical promise of cell penetrating peptides: Limits versus prospects. J Pharm. Sci. 97, 144-162 (2008). Cai, S.R. et al. The kinetics and tissue distribution of protein transduction in mice. EuropeanJournalofPharmaceuticalSciences 27, 311-319 (2006). Heitz, F., Morris, M.C. & Divita, G. Twenty years of cell-penetrating peptides: from molecular mechanisms to therapeutics. British Journal of Pharmacology 157, 195-206 (2009). Alouf, J.E. & Popoff, M.R. (eds.) The Comprehensive Sourcebook of Bacterial Protein Toxins (Academic Press, 2006). Klimpel, K.R., Molloy, S.S., Thomas, G. & Leppla, S.H. Anthrax toxin protective antigen is activated by a cell-surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad Sci. U S. A. 89, 10277-10281 (1992). Milne, J.C., Furlong, D., Hanna, P.C., Wall, J.S. & Collier, R.J. Anthrax protective antigen forms oligomers during intoxication of mammalian-cells. J. Biol. Chem. 269, 20607-20612 (1994). Kintzer, A.F. et al. The protective antigen component of anthrax toxin forms functional octameric complexes. J Mol. Biol. 392, 614-629 (2009). Krantz, B.A., Trivedi, A.D., Cunningham, K., Christensen, K.A. & Collier, R.J. Acidinduced unfolding of the amino-terminal domains of the lethal and edema factors of anthrax toxin. J. Mol. Biol. 344, 739-756 (2004). Pannifer, A.D. et al. Crystal structure of the anthrax lethal factor. Nature 414, 229-233 (2001). Ballard, J.D., Collier, R.J. & Starnbach, M.N. Anthrax toxin-mediated delivery of a cytotoxic T-cell epitope in vivo. Proc. Natl. Acad Sci. U. S. A. 93, 12531-12534 (1996). Hobson, J.P., Liu, S.H., Rono, B., Leppla, S.H. & Bugge, T.H. Imaging specific cellsurface proteolytic activity in single living cells. Nat. Methods 3, 259-261 (2006). Bachran, C. et al. Anthrax toxin-mediated delivery of the Pseudomonas exotoxin A enzymatic domain to the cytosol of tumor cells via cleavable ubiquitin fusions. Mbio 4, e00201-13 (2013). 107 38. Arora, N. & Leppla, S.H. Fusions of anthrax toxin lethal factor with shiga toxin and diphtheria toxin enzymatic domains are toxic to mammalian cells. Infect. Immun. 62, 4955-4961 (1994). Arora, N. & Leppla, S.H. Residues 1-254 of anthrax toxin lethal factor are sufficient to cause cellular uptake of fused polypeptides. J Biol. Chem. 268, 3334-3341 (1993). Milne, J.C., Blanke, S.R., Hanna, P.C. & Collier, R.J. Protective antigen-binding domain of anthrax lethal factor mediates translocation of a heterologous protein fused to its amino- or carboxy-terminus. Mol. Microbiol. 15, 661-666 (1995). Cordero, C.L., Kudryashov, D.S., Reisler, E. & Satchell, K.J.F. The actin cross-linking domain of the Vibrio cholerae RTX toxin directly catalyzes the covalent cross-linking of actin. J. Biol. Chem. 281, 32366-32374 (2006). von Moltke, J. et al. Rapid induction of inflammatory lipid mediators by the inflammasome in vivo. Nature 490, 107-U126 (2012). Popp, M.W., Antos, J.M., Grotenbreg, G.M., Spooner, E. & Ploegh, H.L. Sortagging: a versatile method for protein labeling. Nat. Chem. Biol. 3, 707-708 (2007). Chen, I., Dorr, B.M. & Liu, D.R. A general strategy for the evolution of bond-forming enzymes using yeast display. Proc. Natl. Acad Sci. U S. A. 108, 11399-11404 (2011). Ling, J.J.J., Policarpo, R.L., Rabideau, A.E., Liao, X.L. & Pentelute, B.L. Protein Thioester Synthesis Enabled by Sortase. J Am. Chem. Soc. 134, 10749-10752 (2012). Collier, R.J. & Kandel, J. Structure and activity of diphtheria toxin. 1. Thiol-dependent dissociation of fraction of toxin into enzymically active and inactive fragments. J. Biol. Chem. 246, 1496 (1971). Wilson, B.A. & Collier, R.J. Diphtheria-toxin and Pseudomonas aeruginosa exotoxin A active-site structure and enzymatic mechanism. Curr. Top. Microbiol. Immunol. 175, 2741(1992). Sun, J., Lang, A.E., Aktories, K. & Collier, R.J. Phenylalanine-427 of anthrax protective antigen functions in both pore formation and protein translocation. Proc. Natl. Acad Sci. U. S. A. 105, 4346-4351 (2008). 39. 40. 41. 42. 43. 44. 45. 46. - 47. 48. 49 50. 51. 52. 53. 54. 55. 108 . z A. et al . A penylalanine clamp catalyzes protein translocation through the anthrax toxin pore. Science 309, 777-781 (2005). Adam, S.A., Marr, R.S. & Gerace, L. Nuclear-protein import in permeabilized mammalian-cells requires soluble cytoplasmic factors. J. Cell Biol. 111, 807-816 (1990). Falnes, P.O., Wesche, J. & Olsnes, S. Ability of the Tat basic domain and VP22 to mediate cell binding, but not membrane translocation of the diphtheria toxin A-fragment. Biochemistry 40, 4349-4358 (2001). Nagahara, H. et al. Transduction of full-length TAT fusion proteins into mammalian cells: TAT-p27(Kipl) induces cell migration. Nature Medicine 4, 1449-1452 (1998). Hantschel, 0., Grebien, F. & Superti-Furga, G. Targeting allosteric regulatory modules in oncoproteins: "Drugging the Undruggable". Oncotarget2, 828-829 (2011). Zornetta, 1. et al. Imaging the cell entry of the anthrax oedema and lethal toxins with fluorescent protein chimeras. CellularMicrobiology 12, 1435-1445 (2010). Zheng, S., Zhang, G., Li, J. & Chen, P.R. Monitoring Endocytic Trafficking of Anthrax Lethal Factor by Precise and Quantitative Protein Labeling. Angew. Chem. Int. Ed. 53, 16 (2014). 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. Wesche, J., Elliott, J.L., Falnes, P.O., Olsnes, S. & Collier, R.J. Characterization of membrane translocation by anthrax protective antigen. Biochemistry 37, 15737-15746 (1998). Abi-Habib, R.J. et al. BRAF status and mitogen-activated protein/extracellular signalregulated kinase kinase 1/2 activity indicate sensitivity of melanoma cells to anthrax lethal toxin. Mol. Cancer Ther. 4, 1303-1310 (2005). Wadia, J.S., Stan, R.V. & Dowdy, S.F. Transducible TAT-HA fusogenic peptide enhances escape of TAT-fusion proteins after lipid raft macropinocytosis. Nat. Med 10, 310-315 (2004). Pirie, C.M., Hackel, B.J., Rosenblum, M.G. & Wittrup, K.D. Convergent Potency of Internalized Gelonin Immunotoxins across Varied Cell Lines, Antigens, and Targeting Moieties. J. Biol. Chem. 286, 4165-4172 (2011). Johannessen, C.M. et al. COT drives resistance to RAF inhibition through MAP kinase pathway reactivation. Nature 468, 968-U370 (2010). Engfeldt, T., Renberg, B., Brumer, H., Nygren, P.A. & Karlstrom, A.E. Chemical synthesis of triple-labelled three-helix bundle binding proteins for specific fluorescent detection of unlabelled protein. Chembiochem 6, 1043-1050 (2005). Parizek, P. et al. Designed Ankyrin Repeat Proteins (DARPins) as Novel IsoformSpecific Intracellular Inhibitors of c-Jun N-Terminal Kinases. Acs Chemical Biology 7, 1356-1366 (2012). Wojcik, J. et al. A potent and highly specific FN3 monobody inhibitor of the Abl SH2 domain. Nature Structural & Molecular Biology 17, 519-U173 (2010). Miller, C.J., Elliott, J.L. & Collier, R.J. Anthrax protective antigen: Prepore-to-pore conversion. Biochemistry 38, 10432-10441 (1999). Fang, G.-M. et al. Protein Chemical Synthesis by Ligation of Peptide Hydrazides. Angewandte Chemie-InternationalEdition 50, 7645-7649 (2011). 109 Chapter 3: Delivery of Mirror Image Polypeptides into Cells The work presented in this chapter was published in the following manuscript and is open access from the Royal Society of Chemistry: Rabideau, A.E.,* Liao, X.* & Pentelute, B.L. Delivery of mirror image polypeptides into cells. Chemical Science 6, 648-653 (2015). *: co-first author 110 3.1. Introduction The chirality of biomolecules is critical for their structural integrity and function; 1 proteins produced by the ribosome are primarily composed of L-amino acids and achiral glycine. Polypeptides that contain mirror image or D-amino acids therefore provide a few key advantages over their natural counterparts. Mirror image polypeptides are not recognized by natural proteases and are thus less susceptible to degradation and have been found to cause a low immunogenic response. 2 Solid phase peptide synthesis from D-amino acids (and in some cases native chemical ligation) is used to synthesize mirror image polypeptides.3 ,4 Mirror image phage display was developed to generate functional D-polypeptide binders for L-target proteins. 5 The evolution of bioactive mirror image binders toward intracellular targets has been largely unexplored in part due to the challenge of delivery into the cell cytosol. There are only a few bioactive mirror image peptides that have been delivered inside cells to perturb a biological function, including a D-peptide evolved from mirror image phage display to bind to MDM2 protein6, 7 or a retro-inverso 8 peptide that activates the p53 protein. In addition, there are no examples of mirror image proteins delivered into the cell cytosol. Despite extensive efforts in developing intracellular delivery techniques based on cell penetrating peptides (CPPs), 9' 10 cationic lipids,1 ' nanoparticles, 2 or liposomes,"3 limited success has been achieved for D- polypeptides.1 4 In the case of D-peptide binders towards MDM2, the fusions to CPPs were nonspecifically cytotoxic in a p53-independent manner, and the most active binders could not be packaged into liposomes for delivery. ' 7 A system capable of efficient delivery of Dpolypeptides into the cell cytosol is of immediate interest since it allows us to not only investigate the fundamental properties of these modem agents in the cell, but also to use them for perturbation of intracellular processes. 111 Nature has evolved machineries to transport various biomolecules into cells.1 5 One example is Bacillus anthracis, which produces anthrax toxin. Anthrax toxin is a three-component system containing protective antigen (PA), a receptor-binding, pore-forming protein, lethal factor (LF)1 6 and edema factor (EF)' 7 as the enzymatic moieties. PA binds to receptors on host cells and is cleaved by furin-family proteases (Figure 3.2.1 a, step 1).18 The resulting fragment PA6 3 self-assembles into the ring-shaped heptameric1 9 and octameric 2 0 pre-pore (Figure 3.2.1 a, step 2), forming complexes with LF and EF with high affinity (Figure 3.2.1 a, step 3). The complexes are then endocytosed (Figure 3.2.1 a, step 4) and acidification triggers conformational rearrangement of the pre-pore to form an ion-conductive P-barrel transmembrane pore. 2 ' The pore then translocates LF and EF into the cytosol to act on their selective target. LF binds stereospecifically to PA pre-pore through the 263-residue N-terminal domain (LFN), which initiates the unfolding and translocation of the protein in an N- to C-terminal direction through the narrow P-barrel channel (Figure 3.2.1a, step 5).22 Prior work has shown this system can transport cargo into cells but never for mirror image cargo. Here we combined synthetic chemistry with enzyme-mediated bioconjugation to attach mirror image polypeptides as cargo to the C-terminus of LFN- We found that the PA pore can translocate mirror image polypeptide cargos as efficiently as the all L constructs. For the first time, we demonstrated the delivery of intact mirror image protein cargo into the cytosol of eukaryotic cells. We further demonstrated that a mirror image peptide binder delivered by PA was able to disrupt the intracellular p53/MDM2 interaction in cancer cells. 112 3.2. Results 3.2.1. Delivery of mirror image polypeptides We used an evolved sortase A (SrtA*) 2 3 for the facile attachment of different D-cargo with an N-terminal oligoglycine motif to the C-terminus of LFN or LFN-DTA containing the LPXTG recognition site to give sortagged variants (Sv) or sortagged DTA variants (SDv) (Figure 3.2.1b). LFN-DTA contains DTA, which is the A-chain of diphtheria toxin and is extensively used as a reporter of translocation based on its enzymatic activity to ADP ribosylate elongation factor-2 and block protein synthesis once in the cytosol.2 4 25 After confirming the translocation of the LFN-DTA variants (SDv), we investigated delivery into the cytosolic compartment via western blot using LFN variants that lack DTA (Sv). 113 a) D-cargo PA63 PA83 * 2 13 LFN __3 4 [D-cargo b) -LPSTGG or G 5-D-ro SV: or -> LPSTGG SrtA -LPSTG,-Ei SDv: LPSTG,-c Figure 3.2.1. Delivery of D-cargo sortagged onto LFN and LFN-DTA. a) Translocation of Dcargo conjugated to LFN is mediated by protective antigen (PA) from anthrax toxin. b) G5 -Dcargo are conjugated to LFN and LFN-DTA using SrtA to yield sortagged variants (Sv) and sortagged DTA variants (SDv). 114 - To begin our studies, we synthesized (Table 3.6.1) and sortagged two peptides, (G5 AKFRPDSNVRG) one containing all L-amino acids and the other with all D-amino acids, to LFN-DTA (Table 3.6.2) to give SDv1 and SDv2 (Table 3.6.3 and Table 3.6.4), respectively (Figure 3.2.2a). The SDv1 and SDv2 conjugates were added to CHO-Ki cells in the presence of 20 nM PA for 30 minutes, which was sufficient time to allow entry of DTA cargo enzyme into cytosol. After treatment, the cells were washed and incubated with medium supplemented with 3 H-Leu to determine the efficiency of cargo delivery by measuring protein synthesis (Table 3.6.5). As shown in Figure 3.2.2b, both protein conjugates translocated efficiently, relative to the positive control (LFN-DTA), indicating that the PA/LFN system is adaptable for the intracellular delivery of mirror image peptides. Once unfolding and translocation is initiated by LFN the cargo can 'piggy-back' through PA pore regardless of the stereochemistry. We further analyzed the cytosolic delivery of the LFN variants Svl and Sv2 by western blot. CHO-Ki cells were incubated with Svl and Sv2 in the presence of PA overnight to allow for multiple rounds of receptor mediated endocytosis and translocation. Cells were washed, trypsin digested to remove cell surface receptors and bound proteins, and then extracted with buffer containing 50 pg mL-' digitonin. Digitonin is a mild detergent used to permeabilize plasma membrane and extract only the cytosolic fraction.2 6 Immunoblotting with antibodies against Rab5 (an early endosome marker) and Erkl/2 (a cytosolic protein) is used to validate the extraction of cytosolic proteins. Immunoblot analysis using anti-LF antibody revealed that both conjugates translocated into cells and Sv2 reached a similar steady state concentration to LFN; however for SvI, we observed a significantly less intense band than Sv2 (Figure 3.2.2c). Since the delivery efficiencies were the same for all three constructs from the protein synthesis inhibition assay, the steady state levels in the cell could reflect the difference in extracellular or 115 intracellular stability. We observed similar stability in serum for Svl, Sv2, and LFN (Figure 3.6.1), we therefore hypothesize that the peptides used here acted as unstructured cargo on the Cterminus of LFN, and facilitated degradation. In this case, the L-peptide promoted degradation while the D-peptide did not. We investigated this observation with a variant containing the Lpeptide capped with two D-amino acids at the C-terminus (Sv3). We observed a similar concentration of Sv3 in the cytosol as wild-type LFN, indicating the C-terminal D-amino acids played an essential role in the variant's intracellular stability. These observations suggest that a D-peptide cap can possibly promote stabilization inside the cell. 3.2.2. Translocation requires a functional PA and endocytosis In order to confirm the translocation mechanism of Svi and Sv2, we performed three control experiments and monitored their translocation by western blot (Figure 3.2.2d). We first incubated the cells at 4 'C instead of 37 'C, which inhibited translocation of both Svl and Sv2, indicating that endocytosis is necessary for translocation. Next, we found that treatment with 200 nM Bafilomycin Al, a vacuolar H+-ATPase inhibitor, prevented translocation, indicating the importance of an acidic endosome. Finally, we used a PA mutant, PA[F427H], that binds LFN and is endocytosed but arrests translocation. We observed no cargo in the cytosol when cells were incubated with PA[F427H] instead of PA, indicating that Svl and Sv2 must pass through the translocase for entry into the cytosol. These results confirmed that the translocation of Svl and Sv2 followed the same mechanism as LFN- 116 a) 1 QQ 2 a k Q QQ f I r I pId I s In Q b) S C r r I NH G CONH 2 [~~ = D-amino acid =L-amino acid 1.4 _1.2. 1.0 0.8 a. -a-LF N-DT A I -*-SDv-1 0.6 -+-SDv-2 -*-LF N-DTA, No PA -&-SDv-1, No PA 0.41. 0.2' 0.0 -6 4 -6 -14 -1'3 -1'2 -1'1 -0 Log [Protein Concentration (M)] C) 3 -6 G 4 ng + + 2 - LFN + 2 + 1 + PA Sv- NH 2 3 100i000 anti-LF wAli WWON anti-Erkl/2 " Ow Owf PA PA[F427H] Sv- + + - - 4 ng - - + + 2 1 2 1 2 + - + - + - 1 2 1 4*C 4*C Ba - d) + anti-Rab5 2 Ba anti-LF 4* anti-Erki /2 ."n - *am anti-Rab5 Figure 3.2.2. Translocation of mirror peptides using PA/LFN. a) Peptide cargo 1 and 2 contain either L or D-amino acids, respectively. b) Translocation efficiency of SDv1 and SDv2 were analysed using the protein synthesis inhibition assay in CHO-Ki cells treated with varying concentrations of each variant in the presence of 20 nM PA for 30 minutes and protein synthesis was measured using 3H-Leu. c) Peptide cargo 3 and western blot for CHO-Ki treated with 250 nM wild-type LFN or Svl-3 in the presence of 40 nM PA overnight. d) CHO-Kl cells were treated with 250 nM Svl or Sv2 in the presence of 40 nM PA overnight. The treatment conditions for the control experiments included mutant PA (PA[F427H]), the addition of 200 nM Bafilomycin Al (Ba), and 4*C instead of 37'C incubation (4*C). 117 3.2.3. Translocation of a D-peptide for MDM2 We next explored PA-mediated delivery of a bioactive D-peptide to disrupt an intracellular protein-protein interaction. We chose a D-peptide (TAWYANF*EKLLR, where F* is p-CF 3 -D-Phe) that has a Kd of 0.45 nM towards MDM2. 7 Prior work with variants of this peptide proved challenging as fusions to CPPs were toxic and the most active binders could not be packaged into liposomes for delivery. To study if PA-mediated delivery of the D-peptide binder into the cytosol of human glioblastoma (U-87 MG) cells, we sortagged the peptide and its biotinylated form onto LFN-DTA to give SDv4 and SDv4-biotin, respectively (Figure 3.2.3a). Both variants translocated equally well based on the protein synthesis inhibition assay (Figure 3.2.3b). We then studied the cellular function of the D-peptide binder by sortagging the peptide and its biotinylated form onto LFN to give Sv4 and Sv4-biotin, avoiding the interference of DTA toxicity to the cell. Both constructs were delivered and detected by western blot in U-87 MG cells. Based on the linear relationship between the amount of protein loaded and the band intensity detected by anti-LF antibody, we estimated a total of 0.9 ng Sv4 delivered into 76,000 cells, giving 250,000 molecules per cell or 110 nM of cytosolic concentration (Figure 3.2.3d and Figure 3.6.2). We investigated whether the attachment of LFN would affect the interaction between the peptide and MDM2. We expressed -' 0 9MDM2 as a SUMO fusion (SUMO- 2 5- 09MDM2) and 25 used this construct to measure the binding affinity for cargo 4 and its LFN conjugate Sv4 using a bilayer interferometry system (Figure 3.6.3 and Figure 3.6.4). Both cargo 4 and Sv4 effectively competed with immobilized biotin- 15 -2 9p53 for SUMO- 2 51- 0 9MDM2 binding, yielding Kd values of 1.0 0.7 nM and 12.3 4.3 nM, respectively. Despite the reduction in binding affinity as compared to the cargo 4 by itself, Sv4 still had low nanomolar affinity for MDM2. 118 We analyzed whether the delivered peptide could interact with MDM2 in the cell cytosol. Streptavidin agarose beads were used to capture Sv4-biotin from the lysate of cells treated with Sv4-biotin and PA. We then eluted the bound proteins from the beads and analyzed by immunoblotting with anti-LF and anti-MDM2 antibody. As shown in Figure 3.2.3c, we observed bands corresponding to Sv4-biotin and the bound MDM2 from the lysate of cells treated with Sv4-biotin and PA, whereas no MDM2 was detected in the control experiments when PA[F427H] was used instead of PA, or a MDM2 non-binding control (Sv5-biotin, Figure 3.6.5) was used instead of Sv4-biotin. These results show the specific binding between delivered Sv4biotin and intracellular MDM2. To confirm that the binding event occurred inside the cell and not post-lysis, we added similar amount of Sv4-biotin directly into the U-87 MG cell lysate, and no MDM2 was detected in the pull-down fraction. The absence of MDM2 pulldown in this control experiment confirmed that the binding between delivered peptide and MDM2 occurred in the cytosol. Next, we studied the biological effect of the binding between Sv4 and MDM2 inside U87 MG cells. Binding an antagonist in the p53-binding domain of MDM2 disrupts this interaction and results in activation and expression of MDM2, p53, and p21 .27 Therefore, we quantified the relative change in protein levels of p53, MDM2, and p21 in U-87 MG cells under different treatment conditions, and used the small molecule MDM2 antagonist nutlin-3 as a positive control. According to the western blot, we observed increased levels of MDM2, p53, and p21 for U-87 MG cells treated with Sv4 and PA, similar to the cells treated with 1 jM nutlin-3, in comparison to the PA only or PA[F427H] conditions (Figure 3.2.3d). In order to confirm that this upregulation is dependent on the disruption of p53/MDM2 interaction, we performed the same analysis on the K562 leukemia cell line, which lacks p53.28 Despite the 119 delivery of Sv4 into K562 cells, we observed no perturbation of MDM2 and p21 protein levels. These findings strongly support Sv4 regulated protein levels by specifically inhibiting the p53/MDM2 interaction. Taken together, our results indicate a D-peptide can be delivered to the cytosol of U-87 MG cancer cells and reach a concentration sufficient to bind the target protein and disrupt a critical protein-protein interaction. 120 4 -4iFIyEEJLv1nT blotin l(G2' Pull down: streptavidin + + PA PACAF47U |-ONH, hT2 s)3-f Frat I wyI a I n y I kIl I C) t. j 4-biotin (doped in) 4-biotin 4-biotin Sv 5-biotin anti-LF -CONH anti-MDM2 , SDv4 PA + d) PA[F427H] 5 ng Sv 4 1 sM 1.0- -0.8- anti-LF h. 0.6 anti-MDM2 anti-p53 * 0.4u- 4 nutlin-3 4 U-87 MG -0000 6660 K562 ( a - 1.2. + -- LF N-DTA SDv4-biotin 1.4- + b) anti-p21 0.2- anti-GAPDH 0.0. -15 -14 -13 -12 -11 -10 -9 -8 -7 Log [Protein concentration (M)l - +- = D-amino acid - - - a) K562 (p53-null) -6 anti-LF M anti-MDM2 anti-p53 4 anti-p21 anti-GAPDH ""W-OM Figure 3.2.3. Translocation of a D-binder to MDM2. a) Structure of a D-peptide binder to MDM2 (4) and its biotinylated form (4-biotin). b) Translocation efficiency of SDv4 and SDv4biotin were analysed using the protein synthesis inhibition assay in CHO-KI cells treated with varying concentrations of each variant in the presence of 20 nM PA for 2 hours. c) U-87 MG cells were treated with 150 nM Sv4-biotin in the presence of 20 nM PA for 24 h. The cell lysates were then treated with streptavidin beads to co-precipitate Sv4-biotin and MDM2. Sv5-biotin with PA and Sv4-biotin with PA[F427H] served as negative controls. An additional negative control was run where 5 nM Sv4-biotin was doped into cell lysate then treated to the same pull down procedure. d) Western blot analysis of U-87 MG cells treated with 150 nM Sv4 in the presence of 20 nM PA or PA[F427H] or 1 gM nutlin-3 for 24 h. As a negative control, K562 (p53-null) cells were analysed under the same conditions. 121 3.2.4. Translocation of mirror image proteins After demonstrating that mirror image peptides can be efficiently delivered via the PA pore, we next investigated the possibility of translocating mirror image proteins-a feat never accomplished before. We chose to study two mirror image proteins, D-affibody containing three a-helices 29 and D-protein G B1 containing a-helices and a P-sheet.30 We chemically synthesized these proteins from three fragments (Figure 3.6.6 and Figure 3.6.8) using in situ neutralization Boc SPPS. 3 1After cleavage and purification of each fragment, the peptide segments were ligated together using native chemical ligation, according to the synthetic strategies in Figure 3.2.4a and Figure 3.2.4b. 4 The cysteine residues of D-affibody were alkylated to give pseudo-glutamine within the full-length protein, while the cysteine residues of D-GB1 were desulfurized 32 to give the native sequence. The mirror image conformation of the folded proteins was evidenced by circular dichroism (CD) spectra (Figure 3.6.8). An in vitro tryptic digest of L- and D-affibody was performed to measure the stability of the proteins over time (Figure 3.6.9). Based on LC-MS analysis, we observed complete degradation of the L-affibody and no degradation of the Daffibody. This demonstrated that mirror image polypeptides are less susceptible to proteolytic degradation. When conjugated to LFN, we found that LFN-D-affibody (Sv5) was more stable than the all L-protein LFN-L-affibody (Sv5-L) towards trypsin digestion, indicating a unique property of the hybrid protein. To study translocation efficiency, the D-proteins were sortagged onto LFN-DTA to give SDv5 and SDv6, respectively. As evidenced by the DTA-mediated inhibition of protein synthesis, both mirror image variants translocated as efficiently as LFN-DTA (Figure 3.2.4c). In order to confirm that D-affibody was delivered to the cytosol as an intact protein, we studied the translocation of LFN conjugates of D-affibody (Sv5) and installed an alkyne functional group at 122 the C-terminus for functionalization with azide-biotin to yield Sv5-biotin (Figure 3.6.10). After translocation into CHO-KI cells, the digitonin-extracted cytosolic fraction was immunoblotted with anti-LF antibody and showed Sv5-biotin delivery into the cytosol (Figure 3.2.4d). Delivery of Sv5-biotin was further confirmed by staining with streptavidin conjugated to an IR680 dye, which revealed a similar amount to that detected by anti-LF antibody (Figure 3.2.4d). For the first time we show the delivery of intact mirror image proteins into the cytosol by PA, thereby facilitating the investigation of their biological properties inside the cell. 123 GGGGGVDNKF'NKEQQNAFYE20ILHLPNLNEE * a) QRNAF IQSLK 4*DDPS.QSANLL'OAEAKKLNDAQ60APK NH - COSR Thz- 4 -COSR CyO-4-63 1-CONH, 1) NCL 2) -ThZ 1) NCL 2) Akylation NH- ( 33-44 -Cys'- Cys 45-63 -CONH, Fold b) * D-affibody 5 GGGGGMTYKL'ILNGKTLKGE20TTTEAVDAAT30 AEKVFKQYAN40DNGVDGEWTY5*DDATKTFTVT6*E NH -2 COSR Thz-29-38C-COSR Cys- 39-81 -CONH, 1) NCL 2) -ThZ 1) NCL 2) Desollugization NH 1-61 -CONH, Fold 6 D-GB1 1.4 C) 1.2 1.0 --- LF -DTA -- SDv5 S0.8 0,6- 0- -*--SDv6 -*-SDv5-L -A-SDV6-L 0.4 U- 0.20. + PA Sv5-biotin 4 ng + d) Sv5-biotin -5 + -14 -13 -12 -11 -10 9 -8 -6 Log [Probin Concentration (M)] anti-LF -LPSTG" Strep-680 anti-Erkl/2 anti-Rab5 Figure 3.2.4. Synthesis and translocation of mirror image proteins. a) Synthetic scheme for D-affibody (5) and LC-MS analysis of D-affibody (TIC with deconvoluted mass inset). b) Synthetic scheme for D-GB1 (6) and LC-MS analysis of D-GB1 (TIC with deconvoluted mass inset). c) Translocation efficiency of SDv5 and SDv6 were analyzed using the protein synthesis assay in CHO-Ki cells treated with varying concentrations of each variant in the presence of 20 nM PA for 30 min and compared to the efficiencies of L-affibody (SDv5-L) and L-GB 1 (SDv6L) were also analyzed. d) CHO-KI cells were treated with 100 nM Sv5-biotin in the presence of 40 nM PA for 24 hours and analyzed after digitonin extraction with anti-LF antibody and streptavidin-IRDye 680. 124 3.3. Discussion In this work, we showed that once translocation is initiated by LFN, the interaction between the PA pore and cargo attached at the C-terminus of LFN is independent of the stereochemistry. It was not clear that the PA/LFN system would allow for the delivery since most protein-protein interactions are stereospecific and disruption of chirality commonly results in loss of function. The proper function of PA/LFN is dependent on stereospecific interactions that are required for binding and initiation of translocation; however, we demonstrated that the interaction between the PA translocase and cargo is not necessarily stereospecific. In this study, we used two assays: a protein synthesis inhibition assay based on DTA activity and western blot analysis of cytosolic proteins. Collectively, these assays showed that PA can efficiently translocate mirror cargos into the cell cytosol. We found that translocation of mirror cargo is not limited to the length or fold of the polypeptide. We demonstrated the translocation of two mirror image proteins: affibody and GBI. We further confirmed delivery of an intact mirror image protein by staining the translocated biotinylated D-affibody with anti-LF antibody and fluorescently labeled streptavidin. However, we currently have no tools to assess their folding or activity inside the cytosol. Based on the protein synthesis inhibition data, the foreign protein DTA correctly refold in the cytosolic environment. Our future efforts to generate bioactive mirror image proteins will give us the capability to assay their intracellular function and provide insight into refolding after translocation into the cell. We further demonstrated that PA/LFN system efficiently delivers functional D-peptides for the disruption of intracellular protein-protein interaction in cancer cells. We focused on a mirror image peptide that was previously reported a challenge to deliver into cells. 6, 7 Our own 125 efforts to load this peptide into liposomes for delivery failed. In contrast, for the first time we found that the PA/LFN efficiently delivered this D-peptide. We found it bound to its target after reaching the cytosol as confirmed by capture assays and western blot. Furthermore, the D-binder perturbed the p53/MDM2 and activated this pathway. The delivery problem has been the major obstacle to place mirror image polypeptides in the cytosol of cells. Using SrtA-mediated ligation and the PA/LFN delivery platform, we overcame this challenge. 3.4. Experimental 3.4.1. Materials 2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU), N- a-Boc, and N-a-Fmoc protected L- and D-amino acids were purchased from Chem-Impex International, IL, Peptide Methylbenzhydrylamine Institute, (MBHA) Japan, resin was and Midwest obtained Bio-Tech, from Anaspec, Inc., CA. IN. 4N,N- dimethylformamide (DMF), dichloromethane (DCM), methanol (MeOH), diethyl ether, HPLCgrade acetonitrile (MeCN), and guanidine hydrochloride (guanidine-HCl) were from VWR, PA. Trifluoroacetic acid (TFA) was purchased from NuGenTec, CA and Halocarbon, NJ. 2,2'Azobis[2-(2-imidazolin-2-yl)propane]dihydrochloride (VA-044) was purchased from Wako Pure Chemical Industries, Ltd., Japan. All other reagents were purchased from Sigma-Aldrich, MO and Life Technologies, CA. The following primary and secondary antibodies were used goat anti-LF (bD-17, Santa Cruz Biotechnology, TX), rabbit anti-MDM2 (N-20, Santa Cruz Biotechnology, TX), rabbit antip21 (C-21, Santa Cruz Biotechnology, TX), mouse anti-p53 (DO-1, Santa Cruz Biotechnology, TX), rabbit anti-GAPDH (Sigma Aldrich, MO), goat anti-mouse IRdye 680RD (LI-COR 126 Biosciences, NE), goat anti-rabbit IRdye 800CW (LI-COR Biosciences, NE), and donkey antigoat IRdye 680LT (LI-COR Biosciences, NE). 3.4.2. Solid phase peptide synthesis (Boc) Select peptides were synthesized on a 0.2 mmol scale on MBHA resin using in situ neutralization Boc chemistry protocols.3 1 Peptide thioesters were prepared using a S-trityl mercaptopropionic acid (MPA) strategy.33 Side-chain protection for L- and D-amino acids included: Arg(Tos), Asn(Xan), Asp(OcHex), Cys(4-MeBzl), Glu(OcHex), His(Bom), Lys(2ClZ), Lys(Alloc), Ser(Bzl), Thr(Bzl), Trp(CHO), and Tyr(2-BrZ). After completion of stepwise SPPS, peptides containing Trp were subjected to 10% piperidine and 5% H 2 0 (v/v) in DMF for 2 h at RT to remove the formyl protecting group prior to final deprotection of Boc group. The allyloxycarbonyl (Alloc) protecting group was removed using 4.85 mmol phenylsilane and 39.5 pmol tetrakis(triphenylphosphine)palladium(0) in DCM for 20 min at with DCM DMF. RTE 32 34 The resinn was a washed ahdwt C then hnDF The peptides were simultaneously cleaved from the resin and side-chains were deprotected by treatment with 10% (v/v) p-thiocresol and 10% (v/v) p-cresol in anhydrous HF for 1 h at 0 'C. Peptides were triturated with cold diethyl ether, dissolved in 50:50 (v/v) H 2 0: MeCN containing 0.1% TFA (v/v) and lyophilized. The solvent compositions used in the experiments will be referred to as A: 0.1 % TFA in H 20 (v/v), B: 0.1 % TFA in MeCN (v/v). 3.4.3. Solid phase peptide synthesis (Fmoc) Select peptides were synthesized on a 0.1 mmol scale on aminomethyl resin with a Rink amide linker using fast flow Fmoc synthesis. 35' 36 Side-chain protection for the D-amino acids (Chem-Impex International) included: Arg(Pbf), Asn(Trt), Glu(OtBu), Lys(Boc), Lys(Alloc), 127 Ser(tBu), Thr(tBu), Trp(Boc), and Tyr(tBu). After synthesis, peptides were cleaved from resin with 94% TFA containing 2.5% EDT, 2.5% H2 0, and 1% TIPS (v/v) for 2 h at RT. After cleavage, TFA was dried under N 2 (g) and triturated three times with cold ether then dissolved in 50:50 A:B and lyophilized. Purification of the crude peptides was achieved using RP-HPLC. 3.4.4. Analytical LC-MS All peptides, proteins and semi-synthetic products were analyzed on an Agilent 6520 Accurate-Mass quadrupole time-of-flight (Q-TOF) liquid chromatography-mass spectrometry (LC-MS) system. Solvent A' 0.1% formic acid in H2 0 and solvent B': 0.1% formic acid in MeCN was used. LC-MS used Agilent Zorbax 300SB C 3 column (2.1 x 150 mm, 5 ptm) using a linear gradient of 5-65% B' over 15 min at a flow rate of 0.4 mL/min. The observed mass was generated by averaging the major peak in the total ion current (TIC). The charge-state series of the species were deconvoluted using Agilent MassHunter Bioconfirm using maximum entropy setting. 3.4.5. Preparative, semi-preparative, and analytical RP-HPLC The crude peptides were dissolved in 99:1 or 95:5 A:B. If the peptide was insoluble under these solvent conditions, 6M guanidine HCl was added to the solution. Peptides were purified by preparative RP-HPLC on Agilent Zorbax SB C 18 column (21.2 x 250 mm, 7 pim) at a flow rate of 10 mL/min at 1-41% B, 5-45% B, or 10-50% B over 80 min. For semi-preparative RP-HPLC, a Agilent Zorbax 300SB C 18 column (9.4 x 250 mm, 5 ptm) was used at a flow rate of 5 mL/min over the same gradients. HPLC fractions were spotted with MALDI matrix alpha-cyano-4hydroxycinnamic acid (a-CHCA) in 50:50 A:B and checked for the correct molecular masses. The analytical RP-HPLC Agilent C 18 Zorbax SB column (2.1 x 150 mm, 5 pm) was used to 128 confirm the purity of fractions at a flow rate of 0.5 mL/min over a linear gradient of 1-51% B over 12 min. Analytical HPLC UV absorbance traces were measured at 214 nm. 3.4.6. Synthesis of D-affibody The synthesis of D-affibody containing G 5 at the N-terminus (63-mer) was performed by native chemical ligation (NCL) of three peptide segments (Figure 3.2.4a). The peptide segments (and the corresponding masses) used in this synthesis were as follows (X: -S-CH 2-CH 2-CO-, Z: thiazolidine): [Gly -Arg 32 ]-"thioester: GGGGGVDNKFNKEQQNAFYEILHLPNLNEEQRXA (calc. average: 3773.8 Da, obs. 3773.8 0.1) [Thz 3 3-Ser 4 4] thioester: ZAFIQSLKDDPSXA (calc. monoisotopic: 1493.7, obs: 1493.7 [Cys 45 -Lys 63]-CONH 2: 0.1) CSANLLAEAKKLNDAQAPK (calc. monoisotopic: 1983.1 Da, obs: 1983.1 0.1) Purification of peptide segments All three segments were dissolved in 99:1 A:B and purified by preparative RP-HPLC. Fractions containing the purified fragments were combined and lyophilized: [Glyl-Arg32 ]-athioester 43.1 mg (10.2 ptmol), [Thz 3 -Ser 44]-"thioester 118.7 mg (68.9 pmol) and [Cys 45 --Lys 6 3]-CONH 2 152.1 mg (62.3 pLmol). NCL of three peptide segments Peptide fragments [Thz 33 Ser 44]-"thioester (11.0 mg, 6.4 gmol) and [Cys 4 5-Lys 6 3]-CONH2 (14.0 mg, 5.8 pmol) were dissolved to a concentration of 4 129 mM in NCL buffer (6 M guanidine-HC, 20 mM TCEP-HCl, and 40 mM MPAA in 0.2 M sodium phosphate buffer) (Figure 3.6.6A) at pH 6.95 at RT for 7 h. MeONH 2 HCl was then added to the crude reaction mixture at a final concentration of 0.2 M and pH 4.0 at RT overnight to give [Cys3 3 Ser4 4 ]-[Cys 4 5-Lys athioester 63 ]-CONH2 32 1(Figure 3.6.6B). In the same pot, [Gly -Arg] (26.6 mg, 6.3 pmol) was added (Figure 3.6.6C) and incubated at pH 7.0 at RT for 7 h to give [Glyl-Arg 3 2]-[Cys -Ser ]-[Cys 4 5-Lys 63]-CONH 2 (Figure 3.6.6D). The product was purified by semi-preparative RP-HPLC and fractions containing the pure product were combined and lyophilized to give 11.7 mg, 1.5 pmol (26% yield). Alkylation [Gly'-Arg 2]-[Cys- Ser4 4 ]-[Cys4-Lys6 3 r2ir-, 33 44-iC TCONH2The two cysteine positions 45_YS3H of [Gly'-Arg ]-[C-ys -Ser ]-[Cys -Lys l,2 ]-CONH 2 were alkylated in 6 M guanidine HCl, 20 mM TCEP HCl, and 50 mM 2-bromoacetamide in 0.2 M sodium phosphate buffer. The reaction was incubated at pH 7.1 for 30 minutes at RT and quenched with MESNa (100 mM final concentration) to give [Glyl-Lys 63]-CONH 2 . The product was purified by semi-preparative RPHPLC and fractions containing pure product were combined and lyophilized to give 9 mg, 1.1 pmol (76% yield) (Figure 3.6.6E). Folding of [Gly'-Lys 63 ]-CONH 2 Full-length [Gly' -Lys 63]-CONH 2 was dissolved in 6 M guanidine-HCl, 20 mM Tris-HCl, and 150 mM NaCl, pH 8.5, and was diluted from 6 M to 2 M guanidine-HCl using the same buffer without guanidine-HCl. The peptide solution was desalted into 20 mM Tris-HCl, 150 mM NaCl, pH 8.5 using a HiTrap Desalting column. The folded protein was concentrated using a 3 kDa concentrator to give 5.5 mg, 0.79 pmol (69% yield) (Figure 3.6.6F). 130 3.4.7. Synthesis of D-affibody-alkyne D-affibody-alkyne was synthesized using the same strategy but with an additional propargyl-glycine incorporated at the C-terminus of the protein. D-affibody-biotin was obtained by Cu(I)-catalyzed azide-alkynyl click reaction to label the D-affibody-alkyne with biotin-azide. For the labeling reaction, 50 piL D-affibody-alkyne (332 pig, 46.8 nmol) was mixed with final concentration of 100 mM Tris pH 8.5, 1 mM biotin-azide (70 nmol), 2 mM CuSO 4 and 200 mM ascorbic acid in a total 70 ptL reaction. LC-MS analysis showed 80% yield after 4 hours reaction at room temperature. The reaction mixture was dissolved in 6 M guanidine HCl, 20 mM TrisHCl, and 150 mM NaCl, pH 8.5, and was serially diluted from 6 M to 3 M then to 1 M guanidine-HCl using the same buffer without guanidine-HCl. The solution was desalted into 20 mM Tris-HCl, 150 mM NaCl, pH 7.5 using a HiTrap Desalting column. The folded protein was concentrated using a 3 kDa concentrator to give 202 ptg, 28.9 nmol (61% yield). 3.4.8. Synthesis of D-GB1 The synthesis of D-GB1 containing G5 at the N-terminus (61-mer) was performed by NCL of three peptide segments (Figure 3.2.4b). The peptide segments (and the corresponding masses) used in this synthesis were as follows (X: -S-CH 2-CH 2 -CO-, Z: thiazolidine): [Glyl-Ala 28]thioester: GGGGGMTYKLILNGKTLKGETTTEAVDAXA (calc. monoisotopic: 2940.4 Da, obs: 2940.5 0.1 Da) [Thz 29-Tyr 3 s]-thioester: ZTAEKVFKQYXA (calc. monoisotopic: 1387.6 Da, obs: 1387.6 0.1 Da) 131 [Cys 39-Glu 6 1]-CONH 2 : CNDNGVDGEWTYDDATKTFTVTE (calc. monoisotopic: 2579.0 Da, obs: 2579.1 0.1 Da) Purification ofpeptide segments All segments were dissolved in 6 M guanidine-HCl in 95:5 A:B and purified by preparative RP-HPLC. Peptide segments [Glyl-Ala28 ]-"thioester and [Cys 3 9-Glu 6 1 ] were further purified by semi-preparative RP-HPLC. Fractions containing the purified segments were combined and lyophilized: [Gly'-Ala 28]-"thioester 34.3 mg (11.65 gmol), [Thz 29-Tyr 38 ] -thioester 57.3 mg (41.3 pimol), and [Cys 3 9-Glu 6 1] 40.9 mg (15.8 pmol). NCL of three peptide segments Peptide segments [Thz2 9-Tyr 3 8]-"thioester (13.2 mg, 9.5 ptmol) and [Cys 3 9-Glu 6 1] (20.5 mg, 7.9 pmol) were dissolved to a concentration of 4 mM in NCL buffer (Figure 3.6.7A). The reaction was incubated at pH 7.0 at RT for 3 h. MeONH 2 'HCl was then added to the crude reaction mixture at a final concentration of 0.2 M and pH 4.0 at RT overnight to give [Cys29-Tyr ]-[Cys 39 -Glu 6 1 ]-CONH 2 (Figure 3.6.7B). The product was purified by semi-preparative RP-HPLC to give 23.5 mg, 6.21 pmol (78.2 % yield). Peptide segments [Cys 29 -Tyr]-[Cys 9 -Glu"J-CONH 2 (12.7 mg, 3.4 gmol) and [Glyl--Ala 2 ]8thioester (9.0 mg, 3.1 pmol) were dissolved in NCL buffer (Figure 3.6.7C) and incubated at pH 7.0 for 4.5 h to give [Gly'-Ala 28 ]-[Cys 2 9 -Tyr 3 8]-[Cys 39 -Glu 6' ]-CONH 2 (Figure 3.6.7D). The product was purified by semi-preparative RP-HPLC to give 6.0 mg, 0.92 pmol (38.5% yield). Desulfurization of [Gly1 -Ala_]-[Cys Tyr ]-[Cys39 -Glu6 1]-CONH2 Desulfurization was performed according to the protocol by Murakami, M., et al.3 2, 3 4 Full-length peptide [GlylAla2 8 ]-[Cys2 9-Tyr3 8 ]-[Cys 39 -Gu6 1 ]-CONH 2 (6.0 mg) was dissolved to a concentration of 0.25 gM in 6 M guanidine HCl, 0.45 M TCEP-HCl, 0.3 M MESNa and 5 p.M VA-044 in 0.2 M sodium phosphate buffer at pH 7.0. The reactions were incubated at RT for 8 h and pH was 132 adjusted back to 7.0 every hour. Upon completion of the desulfurization reactions, the products were purified by semi-preparative RP-HPLC. Fractions containing the desulfurized [GlylGlu 6 l]-CONH 2 were combined and lyophilized to give 2.8 mg, 0.43 tmol (47.1% yield) (Figure 3.6.7E). Folding of [Gly'-Glu 6 1]-CONH 2 The desulfurized peptide [Glyl-Glu 6 I]-CONH 2 was dissolved in 6 M guanidine-HCl, 20 mM TCEP-HCl, 20 mM Tris, and 150 mM NaCl, pH 8.5, and was serially diluted from 6 M to 3 M to 1 M guanidine-HCl using the same buffer without guanidine-HCl. The peptide solution was desalted into 20 mM Tris-HCl, 150 mM NaCl, pH 8.5 using a HiTrap Desalting column (GE Healthcare, UK). The folded protein was concentrated using a 3 kDa concentrator to give 1.3 mg, 0.20 tmol (46.4% yield) D-GB1 (Figure 3.6.7F). 3.4.9. Circular dichroism (CD) spectroscopy of folded proteins Circular dichroism spectra of D-affibody and D-GB1 were recorded on an Aviv model 202 instrument at 25 'C. A 1 mm path length cell was used. The proteins were prepared by dissolving 0.06 mg in 50 mM sodium sulfate and 5 mM Tris-HCl at pH 8.5. The molar ellipticity (0 in deg cm 2 dmol-) was calculated by [O]x = Oobs x 1/(10 lcn), where Oobs = observed ellipticity at X, 1 = pathlength (cm), c = concentration of peptide (M), n = # of amino acids. The mirror image form of the L proteins showed the same optical rotation, but with the opposite sign thus confirming the mirror image fold of the D-proteins (Figure 3.6.8). 3.4.10. Synthesis of biotinylated p53/MDM2 Inhibitor D-peptide Biotinylated p53/MDM2 inhibitor D-peptide was synthesized using Fmoc SPPS. The Nterminal glycine residue was protected with Boc. After the peptide was synthesized, the Alloc (allyloxycarbonyl) protecting group was removed as described above. Biotin (0.5 mmol) 133 dissolved in 0.2 M HBTU (0.95 eq) was activated with 1.4 eq DIEA then coupled for 25 min at RT. The peptide was cleaved and purified and described above. 3.4.11. Construction of plasmids for recombinant proteins The gene for 25 -109MDM2 was purchased from Addgene (pGEX-4T MDM2 WT, 16237).37 The pET SUMO-25- 9MDM2 was prepared using the Champion pET SUMO protein expression system (Invitrogen, CA). AccuPrime Taq DNA polymerase (Invitrogen, CA) was used to PCR amplify the DNA using the following primers: 5'-GAGACCCTGGTTAGACCAAAGC-3' (forward) 5'-TTATACTACCAAGTTCCTGTAGATCATGG-3' (reverse) The PCR product was purified by the QlAquick PCR purification kit (Qiagen, Netherlands), then cloned into pET SUMO by an overnight ligation at 15 'C with 3 ng PCR product, 50 ng nET SIMO vector, and 1 pl T4 DNA ligase. 2 p of the ligation pI-odu' was transformed into One Shot Machl-Ti competent cells and plated on 30 ptg/mL kanamycin plates and incubated overnight at 37 'C. Colonies were grown in LB media containing 30 jig/mL kanamycin. The plasmid DNA was isolated using the Qiaprep spin miniprep kit (Qiagen, Netherlands). 3.4.12. Protein expression and purification - SUMO- 25 - 109 MDM2 was expressed in Rosetta (DE3) pLysS cells in IL LB culture. His 6 SUMO-LFN-DTA (C186S)-LPSTGG-His 5 , His 6-SUMO-LFN-LPSTGG-His 6 , His6 -SUMO-LFNDTA (C186S), SrtA*-His 6 , WT anthrax protective antigen (PA), and PA[F427H] were expressed 134 in E coli BL21 (DE3) cells at New England Regional Center of Excellence/Biodefense and Emerging Infectious Diseases (NERCE). Approximately 40 g of cell pellet was resuspended in 100 ml of 50 mM Tris-HCl, 150 mM NaCl, pH 7.5 buffer containing 200 mg lysozyme, 4 mg Roche DNAase I, and 2 tablet of Roche protease inhibitor cocktail then sonicated for three times for 20 seconds. After sonication, the suspension was centrifuged at 17,000 rpm for 40 minutes. The lysate was loaded onto three 5 ml HisTrap FF crude Ni-NTA column (GE Healthcare, UK) and washed with 100 mL of 20 mM Tris-HCl pH 8.5, 150 mM NaCl, at pH 8.5 and 100 mL of 40 mM imidazole in 20 mM Tris-HCl pH 8.5, 500 mM NaCl. The protein was eluted from the column using 500 mM imidazole in 20 mM Tris-HCl pH 8.5, 500 mM NaCl. The eluted protein was buffer exchanged into 20 mM TrisHCL pH 8.5, 150 mM NaCl using a HiPrep 26/10 Desalting column (GE Healthcare, UK). WT PA and PA[F427H] was overexpressed in the periplasm of . coli BL21 (DE3) cells and purified by anion exchange chromatography. 38 3.4.13. One-pot sortagging reaction using Staphylococcus aureus SrtA* Enzyme-mediated ligation using Sortase A was utilized to ligate the peptide and protein cargo to LFN-DTA-LPSTGG (for SDvs) or LFN-LPSTGG (for Svs), depending on the assay. We used Staphylococcus aureus SrtA5 9-206 that was evolved (P94S/D160N/KI96T) by Chen et al. (SrtA*).3 9 4 1 N-terminal small ubiquitin-like modifier (SUMO) was first cleaved from the protein substrate, His 6-SUMO-LFN-DTA-LPSTGG-His 5 (or His 6-SUMO-LFN-LPSTGG-His 6 ), using 1 ptg SUMO protease per mg of protein substrate at RT for 30 minutes to give the native Nterminus of LFN-DTA-LPSTGG-His 5 (or LFN-LPSTGG-His 6 ). 135 The ligation reactions were performed on Ni-NTA agarose beads using SrtA*. Ni-NTA beads were washed three times with SrtA* buffer (10 mM CaCl2 , 50 mM Tris-HCl, 150 mM NaCl, pH 7.5). In one pot, 50 pM LFN-DTA-LPSTGG-His 5 (or LFN-LPSTGG-His 6 ), 5 pM SrtA*, and 100 to 500 pM G5 -cargo were incubated with Ni-NTA beads in SrtA* buffer for 30 min at RT nutating. After 30 min, the Ni-NTA beads containing any unreacted LFN-DTALPSTGG-His 5 , SrtA*-His6 , His6 -SUMO, and GG-His 6 were spun down at 4 'C to minimize the formation of the hydrolyzed LFN-DTA-LPST (or LFN-LPST) side product. The supernatant containing the sortagged LFN-DTA variant (SDv) or (sortagged LFN; Sv) was collected. The beads were washed three times with 1 mL of 20 mM Tris-HCI, 150 mM NaCl, pH 7.5. Three buffer exchanges were performed into 20 mM Tris-HCI, 150 mM NaCl, pH 7.5 to remove the excess G5 -peptide. Size exclusion chromatography into 20 mM Tris-HCl, 150 mM NaCl, pH 7.5 was used to remove the excess G 5-protein. The purity of the ligated product was analyzed by LCMS. 3.4.14. Translocation of SDvs and protein inhibition assay in CHO-Ki cells The CHO-Ki cells were maintained in F-12K media supplemented with 10% (v/v) fetal bovine serum at 37 'C and 5% CO 2. The cells were plated at 3.0 x 104 per well in a 96-well plate one day prior to the assay. The SDvs were prepared in ten-fold serial dilutions in 50 pL, and added 50 jtL of F-12K media containing 20 nM PA. The 100 tL samples were added to CHOKI cells and incubated for 30 minutes at 37 'C and 5% CO 2 . (SDv4 and SDv4-biotin was incubated for 2 hours.) After incubation with SDvs, the cells were washed three times with PBS and then 100 ptL leucine-free F-12K medium supplemented with 1 tCi/mL 3 H-leucine (Perkin Elmer, MA) was added and incubated for 1 h at 37 'C 5% CO 2 . The cells were washed three times with PBS and suspended in 150 ptL of scintillation fluid. 3 H-Leu incorporation into cellular 136 proteins was measured to determine the inhibition of protein synthesis by LFN-DTA. The scintillation counts from cells treated with only PA were used as control value for normalization. Each experiment was done in triplicate. The data were plotted using Origin8 using Sigmoidal Boltzmann Fit using equation y = A2 + A1-A2 X-02 and xO represents the logEC5o values. 1+e x- 3.4.15. Cytosolic protein extraction and whole cell lysate preparation for western blot CHO-KI cells were plated in a 12-well plate. The cells were treated with 250 nM Svl, 2, 4, and 5 in the presence of 40 nM PA overnight at 37 'C and 5% CO 2 . After treatment, the cells were washed with PBS and lifted with 0.25% trypsin-EDTA for 5 minutes to remove surface bound proteins then washed twice with PBS. For cytosolic protein extraction, 1 x 106 cells were resuspended in 100 ptL of 50 pg mL-' digitonin in 75 mM NaCl, 1 mM NaH 2PO 4 , 8 mM Na2HPO 4, 250 mM sucrose supplemented with Roche protease inhibitor cocktail for 10 min on ice, and centrifuged for 5 minutes at 13,000 rpm. For whole cell lysate, cells were lysed in IP lysis buffer (25 mM Tris, 150 mM NaCl, 1% v/v NP-40, pH 7.5) supplemented with Roche protease inhibitor cocktail on ice for 30 minutes and centrifuged for 10 minutes at 13,000 rpm. Both supernatants were collected for blotting and analysis. 3.4.16. Western Blot of Sv, 2, 4, and 5 translocated into CHO-Ki cells Lysates were run on an SDS-PAGE gel then transferred onto a nitrocellulose membrane soaked in 48 mM Tris, 39 mM glycine, 0.0375% SDS, 20% methanol and using a TE 70 SemiDry Transfer Unit (GE). After transferring, the membrane was blocked at room temperature for 2 hours with LI-COR blocking buffer. The membrane was then incubated with anti-LF or anti- 137 Erkl/2, and anti-Rab5 in TBST overnight at 4 'C. The membrane was washed and incubated with the appropriate secondary antibodies in TBST for 1 h then imaged by the LI-COR Odyssey infrared imaging system. 3.4.17. Trypsin digestion of L- and D-Affibody Digestion of 3 pg L- and D-affibody with 3 ng trypsin at 37 'C in 100 gL 20 mM tris, 150 mM NaCl, pH 7.5 buffer were monitored up to 23 hours. The digestion was monitored by LC-MS. 3.4.18. Pull down of Sv4-biotin in U-87 MG cells U-87 MG cells were grown in T-75 flasks at a density of 5x10 6 per flask. Cells were treated with 100 nM Sv3-biotin in the presence of 20 nM PA or PA[F427H] for 24 hours at 37 'C and 5% CO 2. After treatment, the cells were lifted with 0.25% trypsin-EDTA, washed three times, and lysed with 500 pL IP buffer for 10 min on ice. As a negative control, untreated cells. Lysates were incubated with 20 tL (42 ptg) streptavidin-agarose beads (Sigma Aldrich) for 4 hours at 4 'C. The beads were washed three times with lvsis huffer and nnce with lIs uIIffer without NP-40. The bound proteins were eluted in IX Laemmli sample buffer (Bio-Rad), boiled for 10 min, and then analyzed by SDS-PAGE gel. Standard immunoblotting was performed using anti-MDM2 antibody and streptavidin-680. 3.4.19. Translocation of Sv4 and Sv4-biotin in U-87 MG or K562 cells U-87 MG cells were grown in T-25 flasks at a density of lx106 per flask. K562 cells were treated in 24 well dishes at a density of 8.0 x 10 5 per well. Cells were treated with 150 nM of Svl, Sv4, or Sv4-biotin in the presence of 20 nM PA or 20 nM PA[F427H] for 24 hours at 37 'C and 5% CO 2 . As a positive control, cells were treated with 1 pM nutlin-3 for 24 h. After 138 treatment, cells were lifted by trypsin and lysed with IP lysis buffer. The isolated lysates were subjected SDS-PAGE separation and transferred to a nitrocellulose blot as previously described. The membrane was immunoblotted with anti-LF, anti-MDM2, anti-p21, anti-p53, and antiGAPDH overnight in TBST. The membrane was washed and incubated with the appropriate secondary antibody in TBST for 1 h and then imaged. The western blot images were analyzed and quantified using the Image Studio program (LI-COR). Each band was normalized to the GAPDH loading control. The band intensity of each protein was set to one in the PA only control. Percent increase in amount of MDM2, p53, and p21 present in each lane compared to the PA only control was plotted in Figure 3.6.5. 3.4.20. Binding interaction between SUMO-2s-' 0 9 MDM2 and 4 or Sv4 A competition binding assay was performed using a bilayer interferometry system (Octet RED96, ForteBio, CA) to determine the binding affinity of 4 (peptide) and Sv4 to SUMO- 2 5 109MDM2. Super streptavidin (SSA) sensors were soaked in IX kinetics buffer (IX PBS containing 0.02% Tween-20, 0.1% BSA, pH 7.4) for 10 minutes at 30 'C. For immobilization, the wild-type 15- 2 9p53 peptide was synthesized containing a biotin on the N-terminus (biotin- 15 29p53). The biotin- 1529p53 peptide (1 RM in IX kinetics buffer) was immobilized on the SSA sensor surface for 5 minutes at 30 'C and 1000 rpm. The surface was washed for 5 minutes with IX kinetics buffer to establish a baseline. Two-fold dilutions of SUMO- 2 51- 09 MDM2 in IX kinetics buffer were analyzed for binding over 5 minutes at 30 'C and 1000 rpm followed by dissociation with IX kinetics buffer. A calibration curve was generated using GraphPad Prism 6 software using non-linear regression analysis (one-site, specific binding fit). The Kd was found to be 578 170 nM. 139 For the competition assay, various concentrations of 4 (peptide) or Sv4 were incubated with 50 nM SUMO- 2 5-'09MDM2 at room temperature for 30 minutes. Meanwhile, SSA sensors were soaked in IX kinetics buffer for 10 minutes at 30 'C. The biotin- 15-29 p53 peptide (1 gM) was immobilized on the SSA sensor surface then washed to establish a baseline, as previously described for the calibration curve. After immobilization, the association and dissociation of SUMO-25- OMDM2 pre-incubated with 4 or Sv4 samples were analyzed at 30 'C and 1000 rpm and using IX kinetics buffer. Based on the binding (nm) values, the concentration of unbound MDM2 was interpolated for each sample using the calibration curve. Non-linear regression analysis was performed using GraphPad Prism 6 software to determine the Kd value based on the equation: Kd following equation to generate fitted curves: y = = [peptide][MDM2]/[complex].' We used the (b-x-K)+(b+X+Kd) 2 2 _4(bX); where y is [MDM2] in nM, x is [inhibitor] in nM, Kd is the dissociation constant, and b is ymax. 3.5. Acknowledgements This work was funded by MIT start-up funds, MIT Reed Fund, Damon Runyon Cancer Research Foundation Innovation Award, National Science Foundation (NSF) CAREER Award (CHE-1351807) for B.L.P, and NSF Graduate Research Fellowship for A.E.R. We thank R. J. Collier (Harvard) for his continued support, the NERCE facility (grant: U54 AI057159) for expression of toxin proteins, and the MIT biophysical instrument facility (grant: NSF-00703 1). We also thank Prof. Barbara Imperiali, Prof. Douglas Lauffenburger, Deborah Pheasant, Jingjing Ling, Mike Lu, and Daphne van Scheppingen for their comments. 140 3.6. Appendix Table 3.6.1.Peptides used in this investigation Observed (Da) Calculated (Da; monoisotopic) 1529.8 0.1 1529.8 2: (D)-G5 -akfrpdsnvrg-CONH 2 1529.8 0.1 1529.8 3: (D-cap) G 5-AKFRPDSNvrg-CONH 2 1529.8 0.1 1529.8 4: (D) G 5-tawyanf(CF 3)ekllr 1865.0 0.1 1864.9 4-biotin: (D) G 5-k(biotin)-(GGs) 3-tawyanf(CF 3)ekllr 2820.0 0.1 2820.5 2031.2 0.1 2031.2 Sequence 1: (L)-G 5-AKFRPDSNVRG-CONH 2 biotin- 1s-29p53: (L) biotin-SQETFSDLWZLLPEN-CONH 2 141 Table 3.6.2. Observed molecular masses of expressed protein constructs when analyzed by LCMS Calculated MW Protein Observed MW (Da) (Da; average) SrtA*-His 6 19214.6 0.4 19214.5 LFN-DTA 52047.0 0.4 52046.2 66700.0 0.4 66700.1 45289.8 0.4 45289.8 WT PA 83754.1 0.4 83751.6 PA[F427H] 83742.4 0.4 83738.3 SUMO- 2 5-'0 9MDM2 23297.3 0.4 23296.6 L-affibody 6925.8 0.2 6925.6 L-GBl 6481.4 0.2 6481.0 His6 -SUMO-LFN-DTA-LPSTGG-His His 6 -SUMO-LFN-LPSTGG-His 142 6 5 Table 3.6.3. List of variants Sequence Variant L PSTG GF OO @@N. N S~V1 Sv -@LPSTG @ -LPSTGA S v2 Sv ~-PSTG SDv4 ( Fp A LPSTGF-jIIL ~~~~~LPSTG-EEIELih Sv4 P dAFA G)1DCNEG~ C)K)(DO, (DT KP ® XOJ O Q®J~Ih3 PSTG 5 v SO KF ,- IZIEI- k SDv4-biotin Sv4-biotin Sv5-alkyne LSG WLPSTG , Sv5 IIp.I LSG SDv5 N LPSTGA-,, A G Sv5-biotin (~'$ LPTAb-' 0 SDv6 W LSG 143 Table 3.6.4. Isolated yields of sortagging ligations from SrtA* reaction Protein Isolated Yield (%) SDvl 61 Svl 41 SDv2 66 Sv3 53 Sv4 76 Sv4-biotin 77 SDv5 14 Sv5 51 Sv5-alkyne 38 SDv6 29 144 Table 3.6.5. EC 50 values of 30-minute protein synthesis inhibition assay. The errors represent fitting errors from Sigmoidal Boltzmann Fit. (* represents 2 h assay) Protein EC 50 (pM) LFN-DTA 21 SDv1 37+5 SDv2 24 3 SDv5 20 4 Sv5-alkyne 51 15 SDv6 33 6 3 LFN-DTA* 2.0 0.6 SDv4* 3.2 0.7 SDv4-biotin* 7.1 2.5 145 4 ng Sv- antiLF 2 LFN 1 *"map 4WOMW *Jjj* 2 3 ONWO Figure 3.6.1. Immunoblot of media from CHO-Ki cells treated with Svl-3 (Figure 3.2.2c). Serum-containing F12K with 250 nM wild-type LFN or Svl-3 and 40 nM PA was removed after overnight treatment with CHO-Ki cells. The stability of the LFN variants in media was analyzed by immunostaining with anti-LF. 146 Sv4 (ng) 0.25 0.1 1.0 2.5 3 4 5.0 7.5 10.0 5 6 7 anti-LF 5000000- * Sv4 4000000- 3000000- o 2000000- 1000000- - 0 0 1 2 8 Lv (ng) Figure 3.6.2. Linear relationship between Sv4 band intensity and the amount of protein loaded 0.1 (ng): y = 566000x + 107000, R2 = 0.993). The signal intensity of each band was compared to - 10 ng of pure Sv4 protein loaded on the gel and immunoblotted with anti-LF. 147 a) b) 0.7 0.6 - ,0.5, E - 0.3 - - 0.2- 400 nM 400 nM 300 nM 200 nM 100 nM 50 nM 25 nM 0.80.69 0.4- 0.1 . -0 0 0.2- 1 0 100 200 300 400 Time (s) 500 0.0600 700 0 100 200 300 Concentration (nM) 4 400 5 500 Figure 3.6.3. Calibration curve for the interaction between immobilized biotin-' 5-29 p53 and SUMO-25- 9MDM2 determined using Octet RED96 bilayer interferometry. a) Association and dissociation curves of various concentrations of SUMO- 2 5-'09MDM2 with biotin-' 5 -29p53 immobilized to super streptavidin sensors. b) A calibration curve was generated using GraphPad Prism 6 software using non-linear regression analysis (one-site, specific binding fit). The Kd was found to be 578 170 nM. 148 60 - A 4 Sv4 1-140 cij 02 L.. 0* 0.1 100 10 1 (nM) Concentration Inhibitor 1000 Figure 3.6.4. Binding interaction between SUMO-21- '9MDM2 and 4 or Sv4 based on a competition binding assay at 30 'C. Non-linear regression analysis was used to generate fitted curves and determine the Kd values based on the equation: Kd = [peptide][MDM2]/[complex]. Non-linear regression analysis was performed using GraphPad Prism 6 software to determine the Kd value based on the equation: Kd = [peptide][MDM2]/[complex]. We used the following 2 -4(bx); where y is [MDM2] in nM, x equation to generate fitted curves: y = (b-x-Kd)+ (b+x+Kd) 2 is [inhibitor] in nM, Kd is the dissociation constant, and b is ymax. The Kd values calculated for 4 and Sv4 were 1.0 0.7 nM and 12.3 4.3 nM, respectively. 149 1 4 4 anti-LF - + 1 + - + - + + - 5 ng 4 - PA PA[F427H] 1 [tM Sv nutlin + a) +- 4-biotin 4-biotin -. OWM* A"WW*OW anti-MDM2 anti-p53 -ms anti-p21 A-N- -si anti-GAPDH b) 300250 - * Sv4 + PA * Sv4 + PA[F427H] * Sv4-biotin + PA a) a) C, 200 - CO, 150- 4-0 a) 50 - CL 100- 0 I - -50 I MDM2 p53 M9 Sv4-biotin + PA[F427H] Sv + PA Sv + PA[F427H] * Nutlin-3 p21 Figure 3.6.5. Quantification of MDM2, p53, and p21 protein levels for U-87 MG cells treated with Svl, Sv4, or Sv4-biotin in the presence of PA or PA[F427H] based on western blot data from Figure 3.2.3d. All levels were normalized to the anti-GAPDH loading control and the PA only condition. 150 ACys- 45.83 H 6 l-{X*NH 3-44 Thz- 95 -COSF 8 766.4 BCY5- P-5. E ys4-O- f -M+5H], +M+ M+Hr mar 114 9.1 Bys 33-4 cs-4-61c5H]. M. + nvZ [M+SHr [+H' 37.3 EM9H .4 4 H - H r 11681 823.2 (M+3Hr 1097.2 800 700 m0 9])0 10001/0 C 860 1100 0 m/z F __ ___ CYs- Q3.44 -C-4 1 83 7JHCO ca 7002.1 NH,'0R 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 1 2 3 4 5 Time (min] 6 7 8 9 12 10 11 Time [nmin] 13 14 15 16 17 18 19 Figure 3.6.6. LC-MS characterization of D-affibody synthesis 151 20 21 A Th|zf-iOOSA D NH se .[ -qn 39- CONH [+8J' 1091.5 9359 Cys- 39-1 ONH Cys- 29-38 -Cys{ (M8i5H13. 1309.8 JCONH, 7 B D-CONH 1 Cvs- 29-38 -Cys{ E (M.7H [M+7HJ' NH, 1 +HHr - Co fM+qr 1060.9 926.7 1260.6 12968 Lmo C R NH 1 2 3 4 5 6 7 8 9 10 C 11 Time o rNH3 12 13 14 [nn] . bo 90001100 1S ob.6460.4 10.1 ca. 6480.0 F 15 16 17 16 19 20 21 1 2 3 4 Figure 3.6.7. LC-MS characterization of D-GB 1 synthesis 152 00 5 6 7 8 9 10 11 12 imo [min 13 14 15 18 17 16 19 20 21 30000-I- * * L-affibody D-affibody * 20000-0 o 10000ai) 0 U 0- C> 0 , -10000-6 * E -20000- U -z ? -30000210 200 240 230 220 wavelength (nm) 200001500010000- 250 * 0 * 260 L-GB1 D-GB1 p. 5000CD 0) CD -0 00. -5000- U -10000- I -15000L> -20000-25000- 0 S -30000 200 210 240 230 220 wavelength (nm) 250 260 Figure 3.6.8. CD of L- and D-affibody and L- and D-GB 1. 153 A t=O L-affibody t=3h t=18h 1 2 3 4 5 6 7 Time [min] 8 9 10 11 400 600 WO0 0 1 00 12 400 1600 1800 m/z B D-affibody t=0 t=3h Lt=18h L1 154 2 3 4 5 6 7 Time [min] 8 9 10 1 400 600 800 1000 1200 r~z 1400 1600 1800 C 37805.5 t=O 37805.5 t=9h t=23h 2 1 7 Time [min] 3 8 10 11e 1500 20000 25000 40000 30000 3500 Deconvaluted mass [Dal 45000 37978.0 t=O 37978.0 t=9h 37978.3 t=23h so A.o 1 2 3 4 5 71 6 Time [min] 1 15000 000 25000 Ioo iAI ohsoo 40000 3A000 3000 Deconvoluted mass [Dal 415000 Figure 3.6.9. Trypsin digestion of L-affibody (A) and D-affibody (B) at t=0, 3, and 18 h and Sv5-L (C) and Sv5-alkyne (D) at t=O, 9, and 23 h. For A-D, the left panel corresponds to the total ion current (TIC). For A and B, the right panel corresponds to the mass spectra of the major peak area and for C and D, the right panel corresponds to the deconvoluted mass of the major peak area. 155 A B Os. 7098.20O C& 7M008 O.74101 ca. 7713.8 S.M. S.M 000 1 2 3 4 5 6 7 8 9 10 I1 Trn. 12 ln-m 13 14 ;5 ;6 17 18 19 &XW 20 000 21 1 2 3 4 5 6 Time 7 8 9 [min Figure 3.6.10. LC-MS characterization of D-affibody-alkyne (a) and D-affibody-biotin (b) 156 10 800 3.6.1. LC-MS Traces SDvl obs.53958.0 0.4 ca. 53957.3 40 PSTG,5Mo 52500 2 1 4 3 7 6 5 8 9 17 15 16 11 12 13 14 Time [min] 10 55000 21 18 19 20 SvO obs. 32410.7 04 c&. 324 11.4 32200 32400 32600 5 4 3 2 1 11 10 9 8 7 6 Time [min] SDv2 obs.53957.9 ca. 53957.3 iI I PSTGnd - I I I 0.4 G 52500 55000 2 1 4 3 5 7 6 8 9 10 11 12 13 14 15 16 17 18 19 20 21 time (min) Sv2 obs.32411 2 t 0 4 ca 324 114 a @& ON | ::E PSTG,-5 35000 3000 2 1 3 4 5 6 7 8 9 10 11 Time [min} S15 cb. 32410. 1 .4D 4 caob. 32410. 30000" 1 2 3 4 5 6 7 1 9 10 35000 ;1 Time [min] 157 SDv4 obs54290.4 ca. 54291.6 LPSTG, ft 1 2 Sv4 3 SG5 J . k y 5 4 [III 3 3k 0.4 -H 8 6 7 Time [min) 9 10 11 IIT II~obs.327444 t 0 4 ca. 32744.8 200 1 2 3 7 8 Time [mn] 9 1 1 6400 1 11 SDv4-biotin obs.55249 1 t04 c, . 55249.Z 1 2 41 3 6 7 9 8 10 11 Time [min] Sv4-biotin obs 33702 4 *0 ca 33702 8 c~ W W (G2 D), LPSTG, ' I 2 Sv50 4 3 1 1 Time 8 7 [minj 0 34000 ;1 9 36000 11 SDv5 obs.59430.2 t OA4 ca. 59429.2 C-term 4* (t>TA)LPSTG, 62s5 s5s1" 1 2 3 4 5 6 7 8 9 10 11 12 Time [min) Sv5 158 13 14 15 16 17 18 19 20 21 4 abs 37582.7 0.4 ca. 37882.4 C-term LPSTG,"- 35000 2 1 4 3 6 5 7 8 9 10 11 12 Time (mini 13 15 14 16 17 40000 19 18 20 21 Sv5-alkyne ca 37977.4 LPSTGJ 40000 35000 4 3 2 1 6 5 9 8 7 10 11 12 13 14 15 16 17 18 19 20 21 Time (min) Sv5-biotin 38594.5 t 0A Obs ca. 38593.2 LPSTG,-H 35000 40 3 2 1 7 6 Tie imin} 5 4 11 10 9 8 SDv6 obW,58907.5t:0.4 ca. 58907 7 LPSTG, C-term 1 2 3 4 5 6 7 8 9 10 12 13 11 ime [mini 14 15 16 17 IS 19 20 21 159 3.7. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 160 Milton, R.C.D., Milton, S.C.F. & Kent, S.B.H. Total Chemical Synthesis of a D-Enzyme - the Enantiomers of Hiv- 1 Protease Show Demonstration of Reciprocal Chiral SubstrateSpecificity. Science 256, 1445-1448 (1992). Dintzis, H.M., Symer, D.E., Dintzis, R.Z., Zawadzke, L.E. & Berg, J.M. A comparison of the immunogenicity of a pair of enantiomeric proteins. Proteins 16, 306-8 (1993). Pentelute, B.L., Gates, Z.P., Dashnau, J.L., Vanderkooi, J.M. & Kent, S.B.H. Mirror image forms of snow flea antifreeze protein prepared by total chemical synthesis have identical antifreeze activities. Journalof the American Chemical Society 130, 9702-9707 (2008). Dawson, P.E., Muir, T.W., Clark-Lewis, I. & Kent, S.B. Synthesis of proteins by native chemical ligation. Science 266, 776-9 (1994). Schumacher, T.N.M. et al. Identification of D-peptide ligands through mirror-image phage display. Science 271, 1854-1857 (1996). Liu, M. et al. D-peptide inhibitors of the p53-MDM2 interaction for targeted molecular therapy of malignant neoplasms. Proc. Natl. Acad Sci. U S. A. 107, 14321-14326 (2010). Zhan, C. et al. An Ultrahigh Affinity D-Peptide Antagonist Of MDM2. J. Med Chem. 55, 6237-6241 (2012). Li, C. et al. Functional consequences of retro-inverso isomerization of a miniature protein inhibitor of the p53-MDM2 interaction. Bioorganic & Medicinal Chemistry 21, 40454050 (2013). Caron, N.J. et al. Intracellular delivery of a Tat-eGFP fusion protein into muscle cells. Molecular Therapy 3, 310-318 (2001). Heitz, F., Morris, M.C. & Divita, G. Twenty years of cell-penetrating peptides: from molecular mechanisms to therapeutics. British Journal of Pharmacology 157, 195-206 (2009). Zelphati, 0. et al. Intracellular delivery of proteins with a new lipid-mediated delivery system. Journalof Biological Chemistry 276, 35103-35110 (2001). Gu, Z., Biswas, A., Zhao, M.X. & Tang, Y. Tailoring nanocarriers for intracellular protein delivery. Chem. Soc. Rev. 40, 3638-3655 (2011). Torchilin, V.P., Rammohan, R., Weissig, V. & Levchenko, T.S. TAT peptide on the surface of liposomes affords their efficient intracellular delivery even at low temperature and in the presence of metabolic inhibitors. Proceedings of the National Academy of Sciences of the United States ofAmerica 98, 8786-8791 (2001). Foerg, C. & Merkle, H.P. On the biomedical promise of cell penetrating peptides: Limits versus prospects. J. Pharm. Sci. 97, 144-162 (2008). Alouf, J.E. & Popoff, M.R. (eds.) The Comprehensive Sourcebook of Bacterial Protein Toxins (Academic Press, 2006). Pannifer, A.D. et al. Crystal structure of the anthrax lethal factor. Nature 414, 229-233 (2001). Leppla, S.H. Anthrax toxin edema factor: a bacterial adenylate cyclase that increases cyclic AMP concentrations of eukaryotic cells. Proc. Natl. Acad Sci. U S. A. 79, 31623166 (1982). 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Klimpel, K.R., Molloy, S.S., Thomas, G. & Leppla, S.H. Anthrax toxin protective antigen is activated by a cell-surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad. Sci. U S. A. 89, 10277-10281 (1992). Milne, J.C., Furlong, D., Hanna, P.C., Wall, J.S. & Collier, R.J. Anthrax protective antigen forms oligomers during intoxication of mammalian-cells. J. Biol. Chem. 269, 20607-20612 (1994). Kintzer, A.F. et al. The protective antigen component of anthrax toxin forms functional octameric complexes. J. Mol. Biol. 392, 614-629 (2009). Krantz, B.A., Trivedi, A.D., Cunningham, K., Christensen, K.A. & Collier, R.J. Acidinduced unfolding of the amino-terminal domains of the lethal and edema factors of anthrax toxin. J. Mol. Biol. 344, 739-756 (2004). Arora, N. & Leppla, S.H. Residues 1-254 of anthrax toxin lethal factor are sufficient to cause cellular uptake of fused polypeptides. J. Biol. Chem. 268, 3334-3341 (1993). Chen, I., Dorr, B.M. & Liu, D.R. A general strategy for the evolution of bond-forming enzymes using yeast display. Proc. Natl. Acad Sci. U S. A. 108, 11399-11404 (2011). Collier, R.J. & Kandel, J. Structure and activity of diphtheria toxin. 1. Thiol-dependent dissociation of fraction of toxin into enzymically active and inactive fragments. J Biol. Chem. 246, 1496 (1971). Wilson, B.A. & Collier, R.J. Diphtheria-toxin and Pseudomonas aeruginosa exotoxin A active-site structure and enzymatic mechanism. Curr. Top. Microbiol. Immunol. 175, 2741(1992). Adam, S.A., Marr, R.S. & Gerace, L. Nuclear-protein import in permeabilized mammalian-cells requires soluble cytoplasmic factors. J Cell Biol. 111, 807-816 (1990). Eldeiry, W.S. et al. Wafl, a Potential Mediator of P53 Tumor Suppression. Cell 75, 817825 (1993). Law, J.C., Ritke, M.K., Yalowich, J.C., Leder, G.H. & Ferrell, R.E. Mutational inactivation of the p53 gene in the human erythroid leukemic K562 cell line. Leuk Res 17, 1045-50 (1993). Lofblom, J. et al. Affibody molecules: Engineered proteins for therapeutic, diagnostic and biotechnological applications. FEBS Letters 584, 2670-2680 (2010). Mandal, K. et al. Chemical synthesis and X-ray structure of a heterochiral {D-protein antagonist plus vascular endothelial growth factor} protein complex by racemic crystallography. Proc. Natl. Acad Sci. U S. A. 109, 14779-14784 (2012). Schnolzer, M., Alewood, P., Jones, A., Alewood, D. & Kent, S.B.H. In situ neutralization in Boc-chemistry solid phase peptide synthesis. Rapid, high yield assembly of difficult sequences. InternationalJournalof Peptide and Protein Research 40, 180-193 (1992). Wan, Q. & Danishefsky, S.J. Free-radical-based, specific desulfurization of cysteine: A powerful advance in the synthesis of polypeptides and glycopolypeptides. Angew. Chem. Int. Ed. 46, 9248-9252 (2007). Hackeng, T.M., Griffin, J.H. & Dawson, P.E. Protein synthesis by native chemical ligation: Expanded scope by using straightforward methodology. Proceedings of the NationalAcademy of Sciences of the United States ofAmerica 96, 10068-10073 (1999). Murakami, M., Okamoto, R., Izumi, M. & Kajihara, Y. Chemical Synthesis of an Erythropoietin Glycoform Containing a Complex-type Disialyloligosaccharide. Angew. Chem. Int. Ed. 51, 3567-3572 (2012). - 18. 161 35. 36. 37. 38. 39. 40. 41. 162 Simon, M.D. et al. Rapid flow-based peptide synthesis. ChemBioChem 15, 713-720 (2014). Mong, S.K., Vinogradov, A.A., Simon, M.D. & Pentelute, B.L. Rapid Total Synthesis of DARPin pE59 and Bamase. Chembiochem 15, 721-733 (2014). Zhou, B.H.P. et al. HER-2/neu induces p53 ubiquitination via Akt-mediated MDM2 phosphorylation. Nature Cell Biology 3, 973-982 (2001). Miller, C.J., Elliott, J.L. & Collier, R.J. Anthrax protective antigen: Prepore-to-pore conversion. Biochemistry 38, 10432-10441 (1999). Mazmanian, S.K., Liu, G., Hung, T.T. & Schneewind, 0. Staphylococcus aureus sortase, an enzyme that anchors surface proteins to the cell wall. Science 285, 760-763 (1999). Chen, I., Dorr, B.M. & Liu, D.R. A general strategy for the evolution of bond-forming enzymes using yeast display. Proceedings of the National Academy of Sciences of the United States ofAmerica 108, 11399-11404 (2011). Popp, M.W., Antos, J.M., Grotenbreg, G.M., Spooner, E. & Ploegh, H.L. Sortagging: a versatile method for protein labeling. Nat. Chem. Biol. 3, 707-708 (2007). Chapter 4: Translocation of Non-Canonical Polypeptides into Cells Using Protective Antigen The work presented in this chapter was published in the following manuscript and is open access from Nature Publishing Group: Rabideau, A.E.*, Liao, X.,* Akqay, G. & Pentelute, B.L. Translocation of Non-Canonical Polypeptides into Cells Using Protective Antigen. Scientific Reports 2015, 5:11944. *: co-first author 163 4.1. Introduction Nature has evolved several types of translocation systems for delivering toxic proteins into host cells.' Certain pathogenic bacteria such as Vibrio cholerae, Corynebacterium diphtheria, and Bacillis anthracis infect host cells with cholera, diphtheria, and anthrax toxin, respectively. 2 These protein toxins access the cytosol by two main pathways: transport through a translocase from the host cell endosome or retrograde endocytosis from the endosome to the Golgi and endoplasmic reticulum. 2 Anthrax toxin is one example that utilizes a translocase to deliver protein toxins from the endosome into the cell cytosol. 3 Anthrax lethal toxin is comprised of two proteins, protective antigen (PA) and lethal factor (LF). 3 The pore-forming protein, PA83 binds to an anthrax toxin receptor on the host cell surface and is then proteolytically activated by furin protease to give PA 6 3 ,4 triggering oligomerization of PA 63 into a heptameric5 or octameric6 pre-pore. LF binds to the pre-pore through the 263-residue N-terminal domain (LFN) with nanomolar affinity. 7 Up to three8 or four 6 molecules of LF can bind to the heptameric or optameric pre-pore, respectively. The entire complex is then endocytosed and the low pH triggers a conformational rearrangement of the prepore to form a pore that spans through the endosomal membrane, 9 initiating translocation of LF into the cytosol.10 Biophysical and biochemical experiments of LF translocation through the cationselective PA pore have elucidated the translocation mechanism into the cell cytosol."' 12 The working model points toward a number of key factors that govern translocation through PA and subsequent escape from the endosome: 164 e PA pre-pore undergoes conformational rearrangement to form a pore in the endosomal membrane 5 ' * 1 Charge state-dependent Brownian ratchet drives translocation through the PA pore" " * Protein cargo partially unfold in acidic endosome prior to translocation14' e 2 Cargo passes through the PA conduit of approximately 12 A diameter , 16 Taking advantage of these drivers of translocation, researchers have investigated the delivery of various protein cargos fused to the C-terminus to LFN through PA pore. Enzymes or enzymatic domains such as P-lactamase, 18 Pseudomonas exotoxin A, Diphtheriatoxin A chain, 19,20 actin cross-linking domain from the Vibrio cholerae RTX toxin, 21 Shiga toxin, 2 and Legionella pneumophila flagellin protein2 3 have been delivered into cells using the PA/LFN system. These studies showed that the PA pore is efficient in transporting naturally occurring or recombinantly expressed proteins. However, the question of whether the PA pore can accommodate and transport molecules beyond natural polypeptides has yet to be investigated. Only one example has shown the effects on translocation when D-amino acids and cysteic acid were incorporated at the N-terminus of LFN- 24 Some non-natural molecules, including small molecule drugs and polypeptides with backbone or side chain modifications, have attracted research interest because of their therapeutic potential. Investigating the translocation of these non-canonical polypeptides would not only aid in their cytosolic delivery, but also provide insight into the mechanism and promiscuity of the PA translocase. In the present study, we investigated the translocation of various non-canonical polypeptide moieties through the PA translocase (Figure 4.1.1a). We designed our experiments to seek answers to three questions: 1, Do non-natural amino acids affect translocation through 165 PA pore? 2, Can PA pore still tolerate cargos containing non-natural peptide backbones? 3, Can a constrained polypeptide or rigid natural product pass through the 12 A pore? To carry out these investigations, we leveraged the use of sortase A (SrtA)-mediated ligation. The non-canonical polypeptide cargos were synthesized with an N-terminal oligoglycine motif and attached to the C-terminus of LFN-DTA using the LPXTG tag and SrtA2 5 to give LDn constructs (Figure 4.1.1b). DTA (A chain of diphtheria toxin) serves as a reporter of translocation since it inhibits protein synthesis upon entry into the cytosol. 5 , 22, 26 From our investigations, we found that peptides containing minor modifications to the backbone or amino acid side chains, peptidomimetic monomethyl auristatin F (MMAF), and the small molecule drug, doxorubicin translocate efficiently. In contrast, cyclic peptides and the small molecule drug, docetaxel caused disruption to the translocation process. Our findings support the hypothesis that the chemical composition of the cargo does not inhibit the Brownian ratchet provided the cargo is able to adopt a conformational state in which it can pass through the narrow pore. 166 cytosol endosome non-canonical polypeptide cargo LFN PA pore bLPSTGG SrtA G 5-cargo cargo LDn: LPSTG 5 - Figure 4.1.1. Delivery of non-canonical polypeptide cargo into the cytosol. a) Non-canonical polypeptide cargo (green star) ligated to the C-terminus of LFN (pdb: LJ7N) to translocate through PA pore (pdb: 3J9C) b) Sortase A-mediated ligation of LFN-DTA-LPSTGG and G5 cargo to form constructs are comprised of LFN-DTA and non-canonical polypeptide cargo (LDn). 167 4.2. Results 4.2.1. Translocation of non-canonical polypeptide cargos with backbone or side chain modifications We first investigated the translocation of peptides containing either backbone or side chain modifications. We chose P-alanine and N-methyl-alanine for backbone modifications, and propargyiglycine and fluoro-phenylalanine for side chain modifications. These modifications were incorporated into a model peptide containing all natural amino acids, G 5 -AKFRPDSNVRG (Table 4.6.1). The peptides were sortagged onto LFN-DTA to give LDn1-5 (Figure 4.2.1a and Table 4.6.2). We first used a protein synthesis inhibition assay to analyze translocation efficiency -a LAcnstrut. Fru,22,26is assay, ten-fold serial dilutions of each purified conjugate were added to CHO-Ki in the presence of 20 nM PA and incubated for 30 minutes. Cells were washed three times with PBS then incubated with medium supplemented with 3H-Leu for 1 h. After incubation, medium was removed, cells were washed, and 3H-Leu incorporation was measured by a scintillation counter to determine the translocation efficiency of each cargo. According to Figure 4.2.1b, each non-canonical cargo (LDn2-5) translocated as efficiently as the controls (Table 4.6.3), LFN-DTA and LDn1, indicating that these peptides, which contain backbone and side chain modifications can be appended to the C-terminus of LFN-DTA without major interference to the translocation process. We further analyzed the cytosolic presence of the non-canonical peptide conjugates using digitonin extraction and western blot. CHO-KI cells were treated with 100 nM LDn1-5 in the presence of 20 nM PA for 12 hours to allow for multiple rounds of endocytosis and translocation. After trypsin digestion of surface bound materials and washing with PBS, the 168 cytosolic proteins were extracted using a lysis buffer containing 50 pg mL- digitonin, a mild detergent used to permeabilize only the plasma membrane.2 7 The proteins extracted by digitonin were analyzed by western blot immunostained with anti-Erkl/2 (a cytosolic marker) and antiRab5 (an early endosome marker) to confirm the presence of only cytosolic proteins. Using immunostaining by anti-DTA antibody, we observed bands corresponding to LDn2, 4, and 5 with intensities similar to that of LDnl (Figure 4.2.1c). We also analyzed the total cell lysate using a buffer containing 1% NP-40 and observed comparable band intensities detected by the anti-DTA antibody (Figure 4.6.3). This data further confirmed the results from the protein synthesis inhibition assay, indicating that these non-natural modifications did not interfere with the translocation through PA pore. In addition, we performed control experiments with LDn2 by incubating the cells at 4'C instead of 37'C, incubating with an inhibitor of endosomal acidification (Bafilomycin Al), or using a mutant PA (PA[F427H]) that binds LFN but arrests translocation. In all three cases, we observed no band corresponding to LDn2, indicating the translocation process followed the same mechanism as LFN-DTA, which requires active endocytosis, an acidic endosome, and a functional PA (Figure 4.2.1c). 169 a 1 -000000O00-CONH. F I.2 b 5 0 F H 1.4S1.21.0 .8 08 0.61 a- +LF N-DTA LDn1 LDn2 + LDn3 LDn4 8 0-44+ E ' 0.4LLn +LDn5 -DTA, No PA + iZ 0.2 0.0 -14 -13 -12 -11 -10 -9 -8 - -6 -5 - - - - - - - + PA[F427H] PA 2ng LDn 1 + - + 1 + 2 + 4 + + 2 + 2 - M Log [Protein Concentration 4*C Ba 5 2 anti-DTA anti-Erkl/2 anti-Rab5 Figure 4.2.1. Translocation of non-canonical peptides. a) Peptides 1-5 containing noncanonical side-chain modifications were ligated to LFN-DTA using SrtA to form LDnl-5 b) Protein synthesis inhibition assay in CHO-KI. Ten-fold serial dilutions of LDn1-5 and LFN-DTA were added to CHO-Ki in the presence of 20 nM PA and incubated for 30 minutes at 37 'C and 5 % CO 2 then washed with PBS. Treated cells were chased with 3H-Leu in Leu-free medium for 1 hour then washed and read using a scintillation counter. c) CHO-Ki cells were treated with 100 nM LDns in the presence of 20 nM PA for 12 hours. The cells were lifted and the cell surface was trypsin digested for 5 minutes then subsequently washed with PBS. The cell cytosol was extracted using 50 pg ml' digitonin for 10 min on ice. The cytosol extraction was analyzed by western blot. The following controls were also analyzed: incubation at 40 C (instead of 37 *C), incubation with Bafilomycin Al, and use of mutant PA, PA[F427H]. Cropped blots are used in western blot data (c). 170 4.2.2. Translocation of cyclic peptides We expanded our library of non-canonical peptides under investigation to include cyclic peptides. In comparison to linear peptides, cyclic peptides have increased structural rigidity and proteolytic stability, making them attractive therapeutic candidates.28 We cyclized L and D forms of the model peptide (1) using native chemical ligation (Figure 4.2.2a). As demonstrated in Figure 4.6.1, we installed a Cys residue on the c-amino group of the Lys residue of the peptide thioester in both L (7) and D (8) forms. The L model peptide containing an alkylated Cys residue at the Lys position served as a linear control (6) for the translocation of the cyclic peptides. Each peptide was sortagged onto LFN-DTA then analyzed using both the protein synthesis inhibition and western blot assays. We found that there was a substantial reduction in translocation efficiency for both LDn7 and LDn8, as compared to LFN-DTA and LDn6 (Figure 4.2.2b and Table 4.6.3). Furthermore, we treated CHO-KI with 100 nM LDn6-8 in the presence of 20 nM PA overnight then analyzed cytosolic delivery in digitonin extracted materials. The western blot immunostained with anti-DTA showed no detectable band for LDn7 and LDn8, while the linear control peptides LDnl and LDn6 showed similar band intensity (Figure 4.2.2c). A small amount of LDn7 and LDn8 was detected in the total cell lysate (Figure 4.6.3), indicating some cargos were trapped in endosomes. The western blot result was consistent with the protein synthesis inhibition assay results, showing that translocation was substantially reduced for the cyclic peptide conjugates, LDn7 and LDn8, compared to the linear peptide controls. 171 a 6 NH2 0 A K F R R P D S N VOO-CONH2 D F 7 S 9N 8 R b 2.0 1.81.6 1.4 -- 1.2 -'- LDn8 LF N-DTA -- LDn6 -A- LDn7 .C 1.0 0.80.6 0.4' 0.2 0.0-0.2-14 -13 -12 -11 -10 -9 -8 -7 -6 -5 C PA 4 ng LDn 1 anti-DTA + - + 1 + 6 + 7 + Log [Protein Concentration (M)] 8 anti-Erkl/2 anti-Rab5 Figure 4.2.2. Translocation of cyclic peptides. a) Linear control peptide and L and D cyclic peptides to form LDn6-8, respectively b) CHO-Ki were treated with ten-fold serial dilutions of LDn6-8 and LFN-DTA in the presence of 20 nM PA for 30 minutes. The same experimental conditions were used as previously described. c) CHO-Ki were treated with 100 nM LD6-8 in the presence of 20 nM PA for 12 hours then subjected to cytosolic extraction for western blot, as previously described. Cropped blots are used in western blot data (c). 172 4.2.3. Translocation of complex small molecules We next investigated the ability of PA to translocate complex, cytotoxic small molecules that are frequently used in antibody-drug conjugates: doxorubicin, docetaxel, and monomethyl auristatin F (MMAF). 29 Direct translocation of these small molecules into the cytosol would be of great interest for the delivery of impermeable chemotherapeutics or other small molecule - drugs. These small molecules were conjugated through the Cys residue of the peptide, G5 LRRLRAC, using a maleimide functional group (Figure 4.2.3a and Figure 4.6.2). The peptidesmall molecule products were then sortagged onto LFN-DTA (LDn9- 11, respectively) for investigation of translocation through PA pore. Again, we used the protein synthesis inhibition assay and found that LDn9 and LDn1 1 translocated as efficiently as the LFN-DTA control, while translocation was completely arrested for LDn1O (Figure 4.2.3b and Table 4.6.1). We then treated the cells with 100 nM LDn9-11 in the presence of 20 nM PA overnight and analyzed cytosolic delivery by immunostaining with anti-DTA antibody. We observed similar band intensities for LDn9 and LDnl1 as compared to LFN-DTA, while no band was detectable for LDn1O (Figure 4.2.3c). We also analyzed the total cell lysate and observed comparable band intensities with the anti-DTA antibody (Figure 4.6.4). The western blot results were consistent with protein synthesis inhibition results demonstrating that the PA pore was efficient in accommodating cytotoxic molecules whose structures were either similar to (i.e. MMAF) or distinct from (i.e. doxorubicin) polypeptides. We hypothesized that the three-dimensional size of the molecule was a major factor contributing to translocation efficiency. Therefore, we estimated the longest linear distance for doxorubicin and docetaxel based on published crystal structures. Despite different possible conformational arrangements, we found that the longest linear distance for doxorubicin in the crystal structure to be about 10.3 A, while that of docetaxel was about 173 13.5 A,3' which is close to the estimated PA pore size (12 A).' 6 Taken together, these data suggested that PA pore was promiscuous enough to translocate cytotoxic small molecules, but with a size limitation due to the inherent size of the pore. 174 -Q®®Q®®~ONH SH 9 10 ON 0 0 0 H H 0 OH 0 0 N0 Hf4YYH N 1114N ~0,. 0" 0 b N 0 1.41.2" 0 0.8 0.6 PA[F427H] PA LDn anti-DTA -v LDn10 ~-"Ln"I I -6 14 -13 -12 -11 -10 -9 -9 Log [Protein Concentration (M)] + 4ng 11 LF,-DTA 4* 9 + 10 11 + 9 BA + C L~9 - 0.01 -15 A + u. 0.2 -*-LF N-DTA + -0. 4 9 ,0-.* anti-Erki/2 anti-Rab5 Figure 4.2.3. Translocation of small molecules. a) Small molecules doxorubicin, docetaxel, and monomethyl auristatin F (MMAF) were conjugated at the cysteine residue of G 5-LRRLRAC then ligated to LFN-DTA to form LDn9- 11. b) CHO-KI cells were treated with ten-fold serial dilutions of LDn9-11 and LFN-DTA in the presence of 20 nM PA for 30 minutes. The same experimental conditions were used as previously described. c) CHO-Ki cells were treated with 100 nM LDn9-11 in the presence of 20 nM PA for 12 hours then digitonin extracted and run on western blot, as previously described. Cropped blots are used in western blot data (c). 175 4.2.4. Translocation of intact cargo Our studies indicate that a variety of non-canonical polypeptide cargos can translocate through PA pore into the cell cytosol. To verify that the protein conjugates containing noncanonical polypeptide cargos passed through the pore and into the cytosol without any truncations, we analyzed the translocation of intact cargos. For this analysis, we installed a biotin at the C-terminus of each non-canonical polypeptide cargo that translocated efficiently (Figure 4.2.4a and Figure 4.2.4b). The C-terminal biotin provided a small tag for western blot detection of intact cargos, which were sortagged onto LFN-DTA as previously described (LDnl-bio to LDn6-bio, LDn9-bio, and LDnll-bio). Translocation of the biotinylated conjugates was achieved by treating CHO-KI cells with 100 nM LDnI-bio to LDn6-bio, LDn9-bio, and LDn I1bio in the presence of 20 nM PA for 12 hours. The cells were subjected to the same harvest and digitonin lysis conditions as previously described. Prior to western blot analysis, the SDS-PAGE gel was run for approximately twice as long as the previous experiments in order to differentiate the shifts in molecular weights of the conjugates containing intact cargo. The western blot in Figure 4.2.4c was stained with anti-DTA (subsequently stained with goat anti-mouse IRdye 800CW secondary antibody) and streptavidin IRdye 680LT. The co-staining of anti-DTA and streptavidin for all the biotinylated conjugates (LDnl -bio to LDn6-bio, LDn9-bio, and LDn 1Ibio) confirmed the presence of the intact protein conjugates in the cytosol of cells. Since doxorubicin and MMAF were conjugated through non-amide linkages (i.e. maleimide) we relied on shifts in molecular weight in the western blot to confirm that the conjugates remained intact after translocation. As demonstrated in Figure 4.2.4c, the shifts in molecular weight for the biotinylated conjugates corresponded with the 1 ng loading controls (LFN-DTA, LDnl-bio, 176 LDn 1 -bio) as well as the non-biotinylated translocated materials (LDn2, LDn9, and LDn 11) indicating that the translocated material was indeed intact. 177 aN NH G R K ONH H 0 3-bio 0= 0 - 2-bio 0 HO'- 0 F NHN0= = 5-bio 0 NH2 ~ ~ C K,.., O 0 0K PA LDn 1ng LD 1ng ing 1-bio 11-bio 2 + + 9 11 + + + + + + + + C 0 H 0 + 6-bio F + 4-blo F 1-blo 2-bio 3-bNo 4-blo 5-bio 6-bio 9-bio 11 -bio anti-DTA streptavidin-680 anti-Erk1 /2 "' ""* " ""w " "' O anti-Rab5 Figure 4.2.4. Translocation of C-terminally biotinylated cargo. a) Non-canonical peptides 1-6 were modified to contain biotin on a C-terminal lysine residue (1 -bio to 6-bio) then ligated to LFN-DTA to form LDnl-bio to 6-bio b) Peptides containing the small molecules, doxorubicin and monomethyl auristatin F (MMAF) were modified to contain biotin on a C-terminal lysine residue (9-bio and 11 -bio, respectively) then ligated to LFN-DTA to form LDn9-bio and LDn 11bio c) CHO-KI cells were treated with 100 nM LDn2, LDn9, LDnl 1, LDnl-bio to LDn6-bio, LDn9-bio, and LDn 11 -bio in the presence of 20 nM PA for 12 hours then digitonin extracted and run on western blot, as previously described. Cropped blots are used in western blot data (c). 178 4.3. Discussion In this paper, we investigated whether PA pore could efficiently transport non-canonical polypeptides into the cell cytosol. We used two separate assays to evaluate translocation efficiency. The protein synthesis inhibition assay based on DTA activity is widely used in the anthrax toxin field to probe translocation, enabling direct comparison to the previous reports. We also used cytosolic extraction by digitonin and western blot analysis to further investigate delivery of the cargos in the cytosol. Based on these two assays, we found that several variants did not perturb translocation through PA pore. The peptide cargos contain backbone modifications, or side chain modifications that can be used to increase proteolytic stability (e.g. P-alanine, N-methyl-alanine), binding affinity (e.g. fluoro-phenylalanine), or serve as chemical handles (e.g. alkyne or biotin groups) for follow-up analysis. We also demonstrated the translocation of small molecule drugs such as doxorubicin and MMAF through PA pore. All together these examples indicate that the PA pore is relatively promiscuous in terms of the substrates that can be translocated. Once unfolding and translocation is initiated by the interaction of LFN with the pore in the acidic endosome, the pore is able to accommodate the trailing segment, with high tolerance for chemical modifications. The digitonin extraction and western blot analysis not only confirmed the cytosolic delivery of these materials through PA pore, but also supported the reported translocation mechanism. The requirement of endocytosis, endosome acidification, and active PA pore was indicated by the control experiments where no bands were detected using 4 'C, Bafilomycin Al, or PA[F427H]. 179 Model cyclic peptides and complex small molecules such as docetaxel perturbed translocation through PA pore, as indicated by the substantial reduction or abolishment of protein synthesis inhibition. We hypothesize that the three-dimensional size of the molecules was responsible for arresting translocation. For the cyclic peptides, although the structures are unknown, the potential rigidity and large size could perturb translocation. Furthermore, translocation efficiency is independent of stereochemistry, as evidenced by the efficient translocation of linear mirror peptides3 2 and inhibited translocation of the D cyclic peptide cargo. Although there was some detectable protein synthesis inhibition activity with the cyclic peptides, we observed no material by western blot of the cytosolic fractions. We note that the protein synthesis inhibition assay is a more sensitive assay since its readout is correlated with the activity of the DTA enzyme, which turns over at very low concentrations, while the western blot approach does not allow us to detect very low levels of protein. Nevertheless, both detection methods demonstrated that these large and constrained cargos had difficulty passing through PA pore. These results provide additional evidence to support the translocation mechanism, where cargos need to be in an extended confirmation to pass through the pore; large and rigid molecules cannot efficiently pass through the pore. This result also provides design principles for different cargos of interest. Moreover, we analyzed the translocation of intact polypeptide cargo. The non-canonical polypeptide cargos that were found to translocate efficiently were biotinylated on the C-terminus then translocated into CHO-Kl cells. After translocation and digitonin extraction, the western blot was analyzed for the presence of DTA as well as biotin using streptavidin labeled with an JRdye. Co-staining of DTA and streptavidin for each conjugate as well as band shifts corresponding to the correct molecular weights indicated that the protein conjugates containing 180 non-canonical polypeptide cargos remained intact after translocation through PA pore into the cytosol. Nature evolved to utilize 20 amino acids to make up proteins. Recently, there has been great interest to incorporate non-canonical amino acids into polypeptides in order to enhance biological properties or to study biological processes. For our study, we asked the basic question of whether these non-natural amino acids could be efficiently translocated into the cell cytosol through the anthrax toxin PA pore. Indeed, we found the PA translocase still functions despite the presence of non-canonical cargo. 4.4. Experimental 4.4.1. Materials All chemicals were reagent grade and used as supplied except where noted. 2-(1HBenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU), N-Fluorenyl-9- methoxycarbonyl (Fmoc) and di-tert-butyl-dicarbonate (Boc) protected amino acids were purchased from CreoSalus or Chem-Impex International. 4-Methylbenzhydrylamine (MBHA) resin was obtained from Anaspec, CA. NN-dimethylformamide (DMF), dichloromethane (DCM), methanol (MeOH), diethyl ether, HPLC-grade acetonitrile (MeCN) and guanidine hydrochloride (guanidine-HCl) were from VWR. Trifluoroacetic acid (TFA) was purchased from NuGenTec and Sigma-Aldrich. 3-maleimidopropionic acid was purchased from Toronto Research Chemicals, Toronto, Ontario. Doxorubicin hydrochloride was purchased from AvaChem Scientific, docetaxel was purchased from A ChemTek Inc., and maleimidomonomethyl auristatin F (MMAF) was purchased from Concortis. All other reagents were purchased from Sigma-Aldrich and Life Technologies. 181 All moisture sensitive reactions were performed under argon atmosphere in oven-dried glassware and in anhydrous solvents. Reactions were monitored by thin-layer chromatography (TLC) carried on silica gel 60F 25 4 plates (EMD Chemical Inc.) and were visualized under an UV lamp or by charring with phosphomolybdic acid or ninhydrin stain. Flash column chromatography was performed on 60A silica gel (230-400 mesh) purchased from Whatman Inc. The following primary and secondary antibodies were used goat anti-LF (bD- 17, Santa Cruz Biotechnology), rabbit anti-DTA (ab8308 abcam), anti-Erkl/2 (137F5 cell signaling), antiRab5 (C8B1 cell signaling), goat anti-mouse IRdye 680RD (LI-COR Biosciences), goat antimouse IRdye 800CW (LI-COR Biosciences), goat anti-rabbit IRdye 800CW (LI-COR Bis e) dny -g IDy Lc T IC1R1 Biosciences), and streptavidin iRdye 680LT (LI-COR Biosciences). 4.4.2. 'H Nuclear magnetic resonance ('H NMR) H NMR spectra were obtained with a Bruker Advance 1II 400 MHz instrument and referenced to tetramethylsilane (TMS) at 0.00 ppm. Chemical shifts are recorded as parts per million (ppm) in 8 scale and coupling constants J, are in hertz (Hz). Multiplicities are indicated as "s" (singlet), "bs" (broad singlet), "d" (doublet), "t" (triplet), "dd" (doublet of doublets), "ddd" (doublet of doublet of doublets), or "m" (multiplet). 4.4.3. Synthesis of docetaxel-maleimide Docetaxel (100 mg, 0.124 mmol) and maleimidopropionic acid (25 mg, 0.149 mmol) were taken in anhydrous DCM (1.5 mL), followed by addition of Mukaiyama's reagent; 2chloro-l-methylpyridinium iodide (57 mg, 0.225 mmol) and excess triethylamine (0.2 mL) at 182 00 C. The reaction mixture was slowly warmed up to room temperature and stirred 16 hours, at which TLC analysis (5% v/v methanol in dichloromethane) indicated consumption of starting materials and formation of a major product. The reaction was quenched by addition of ethanol and additional stirring for 10 min, followed by concentration to dryness. The crude material was subjected to silica flash chromatography to give the thiol reactive docetaxel derivative, docetaxel-maleimide, 62 mg (52.2 % yield). The identity of the product was confirmed by high resolution LC-MS and 'H-NMR. m/z.'HMR (500 MHz, CDCl 3): 8 Calculated mass: 958.4 Da; observed [M+H]+: 959.4 8.16 (d, J=7.55 Hz, C25, C29-H, 2H), 8.02 (s, 2H), 7.60 (t, J=7.30, 111), 7.50 (t, J=7.6, 2H), 7.48-7.30 (m, 8H+CDCl 3), 6.70 (s, maleimide, CH, 2H) 6.58 (d, J=9.15 Hz, 2H) 6.15 (m, 2H), 5.75 (d, J= 9.80 Hz, lH), 5.70 (d, J=6.95 Hz, C2-CH, lH), 5.50 (bs, C3'-CH, 1H), 5.32 (bs, 1H), 5.26 (s, C1O-CH, lH), 4.88 (dd, J=8.20 Hz, C5-CH, IH) 4.32 (d, J= 8.55 Hz, C20-CHb, 1H), 4.27 (in, C7-CH, 1H), 4.2 (d, J=4.75 Hz, C20-CHa 1H), 3.95 (d, J=6.30 Hz, 1H), 3.85 (m, 1H), 3.70 (m, lH), 3.50 (m, 7H), 2.55 (m, 2H), 2.48 (s, C22-CH 3, 3H), 1.98 (s, 2H), 1.90 (d, J=11.95 C14-CH2 ,2H) 1.80 (s, C18-CH 3,3H), 1.72 (s, C19-CH 3,3H), 1.35 (9Hs of -tBu), 1.22 (s, C16-CH3 , 3H), 1.10 (s, C17-CH 3 , 3H) 4.4.4. Synthesis of doxorubicin-maleimide Doxorubicin (50 mg, 0.086 mmol) and N-succinimidyl ester of maleimidopropionic acid (45.78 mg, 0.172 mmol) were taken in DMF (1.6 mL) and reacted for 1 hour in the presence of N,N-diisopropylethylamine DIEA (50 iL), at which TLC analysis (20% v/v methanol in dichloromethane) indicated completion of the reaction and formation of a major product. The reaction mixture was quenched by diluting with DCM, followed by repetitive aqueous extractions to remove DMF and unreacted doxorubicin. The combined organic phase was dried over magnesium sulfate (MgSO 4 ), inorganic salts were filtered off and concentrated in vacuo to 183 dryness. The crude material was purified by silica flash chromatography (20% v/v methanol in dichloromethane) to give the thiol reactive doxorubicin derivative, doxorubicin-maleimide, 47.7 mg (80% yield). The identity of the product was confirmed by MALDI and 'H-NMR. Calculated mass: 694.2; Observed [M+Na]+ 717.0 m/z and [M+K]+ 732.9 m/z. 'HMR (500 MHz, CH 30D): 8 8.25 (d, J=7.65 Hz, IH), 8.00 (s, 1H), 7.85 (t, J=8.05 Hz, IH), 7.61 (d, J=8.65 Hz, I H), 6.78 (s, maleimide, CH, 2H), 5.42 (d, J=3.65 Hz, 1H), 5.21 (bs, lH), 4.73 (d, J=3.5 Hz, 2H), 4.56 (s, COCH 2 -OH, 2H), 4.25 (dd, J=6.75, 10.15 Hz, 2H), 4.08 (m, 2H), 4.05 (s, 3H, OCH3 ), 3.72 (t, J= 6.95 Hz, 2H)) 3.60 (s, lH), 3.44 (s, 2H), 3.20 (s, 2H), 3.12 (d, 2H), 2.90 (s, 2H), 2.45 (t, J=6.90 Hz, 2H) 2.40 (d, J=14.80 Hz, 2H) 2.20 (dd, J=4.80, 9.57 Hz, 2H), 1.90 (ddd, 1H), 1.68 (dd, ) J=4.35 and 12.95 Hz, 2H), 1.30 (s, 4H), 1.28 (d, J=6.60 Hz, 3H, CH3 4.4.5. Solid phase peptide synthesis (Boc) Select peptides were synthesized using in situ neutralization boc chemistry. Peptides were synthesized on 0.2 mmol scale on MBHA resin and the following side chain protection was used for L- and D-amino acids: Arg(Tos), Asn(Xan), Asp(OcHex), Lys(2-ClZ), Lys(Alloc), and Ser(Bzl). For the cyclic peptides, peptide thioesters were prepared using the S-trityl mercaptopropionic acid (MPA) strategy. After peptide synthesis, the peptides were cleaved from the resin and side chains were deprotected using 10% (v/v) p-thiocresol and 10% (v/v) p-cresol in anhydrous HF for 1 h at 0 'C. Peptides were then triturated with cold diethyl ether, dissolved in 50:50 A:B (A: water + 0.1% TFA and B: acetonitrile + 0.1% TFA), and then lyophilized. 4.4.6. Solid phase peptide synthesis (Fmoc) Select peptides were synthesized using fast flow Fmoc synthesis on a 0.1 mmol scale on aminomethyl resin with the Rink amide linker.3 3 Side-chain protection for the amino acids 184 included: Arg(Pbf), Asn(Trt), Glu(OtBu), Lys(Boc), Lys(Alloc), Ser(tBu), Thr(tBu), Trp(Boc), and Tyr(tBu). After synthesis, peptides were cleaved from the resin with 94% TFA containing 2.5% EDT, 2.5% H 20, and 1% TIPS (v/v) for 7 min at 60 'C. After cleavage, TFA was dried under N 2 (g), triturated with cold diethyl ether, dissolved in 50:50 A:B, and then lyophilized. The allyloxycarbonyl (Alloc) protecting group was removed using 4.85 mmol phenylsilane and 39.5 pimol tetrakis(triphenylphosphine)palladium(0) in DCM for 20 min at RT. The resin was washed with DCM then DMF. 4.4.7. Protein expression and purification His6 -SUMO-LFN-DTA(C 1 86S)-LPSTGG-His 5 , His 6 -SUMO-LFN-DTA(C186S), SrtA*- His 6 , wild-type protective antigen (PA), and PA[F427H] were expressed in E coli BL21 (DE3) cells at New England Regional Center of Excellence/Biodefense and Emerging Infectious Diseases (NERCE). Each His6-tagged protein was purified using Ni-NTA columns. Each cell pellet (approximately 40 g) was resuspended in 100 ml of 50 mM Tris-HCI, 150 mM NaCl, pH 7.5 buffer containing 200 mg lysozyme, 4 mg Roche DNAase I, and 2 tablets of Roche protease inhibitor cocktail. The suspension was sonicated on ice three times for 20 seconds. After sonication, the suspension was centrifuged at 17,000 rpm for 40 minutes. The lysate was loaded onto three 5 ml GE HisTrap FF crude Ni-NTA column pre-equilibrated with binding buffer (20 mM Tris-HCl pH 8.5, 150 mM NaCl, at pH 8.5). After loading the lysate, the columns were washed with 100 mL binding buffer then 100 mL 40 mM imidazole in 20 mM Tris-HCl pH 8.5, 500 mM NaCl. The protein was eluted using 500 mM imidazole in 20 mM Tris-HCl pH 8.5, 500 mM NaCl. The eluted protein was buffer exchanged to remove the imidazole using a HiPrep 26/10 Desalting column (GE Healthcare, UK). Wild-type PA and PA[F427H] were 185 overexpressed in the periplasm of E. coli BL21 (DE3) cells and purified by anion exchange chromatography. 4.4.8. Sortase-mediated ligation Sortase A was used to ligate peptides containing the N-terminal oligoglycine motif to LFN-DTA-LPSTGG (LDn). Staphylococcus aureus 59-21ISrtA (P94S/D160N/K196T; SrtA*) evolved by Chen, et al. was used for our SrtA-mediated ligations. In order to obtain a native Nterminus, the small ubiquitin-like modified (SUMO) was cleaved off LFN-DTA-LPSTGG. SUMO cleavage was achieved using I pg SUMO protease per mg of protein substrate at RT for 1 hour followed by gel or LC-MS analysis to confirm complete cleavage. We performed the SrtA-mediated ligations in the presence of Ni-NTA beads in order to bind all His 6-tagged reagents and release the His6 -free product in the supernatant. Ni-NTA beads were equilibrated with SrtA buffer (10 mM CaCl 2 , 50 mM Tris-HCl, 150 mM NaCl, pH 7.5). In one pot, 50 [LM LFN-DTA-LPSTGG-Hiss, 5 pM SrtA*, and 300 pM Gs-peptide were incubated with Ni-NTA beads in SrtA buffer for 30 min at RT while rotating. After incubation, the beads were spun down at 4 'C and the supernatant was collected. The beads were washed twice with SrtA buffer and twice with 10 mM imidazole in 20 mM Tris-HCl pH 7.5, 150 mM NaCI (to remove any nonspecifically bound LFN). The supernatant and all washes were combined, concentrated, and buffer exchanged three times into 20 mM Tris-HCl, 150 mM NaCl, pH 7.5 to remove the excess G 5-peptide. The purity of the ligated product (LDn) was analyzed by LC-MS. Concentrations of the ligated products containing non-natural functionalities were determined using Bradford assay. 4.4.9. LC-MS analysis 186 The purity of all peptides and proteins were analyzed by high resolution LC-MS (Agilent 6520 Accurate-Mass quadrupole time-of-flight liquid chromatography mass spectrometry system). The samples were analyzed on an Agilent Zorbax 300SB C 3 column (2.1 x 150 mm, 5 + pm) using 0.8 mL min~' and a linear gradient of 1-61% or 5-65% B' over 9 min (A': water 0.1% formic acid; B': acetonitrile + 0.1% formic acid). The observed mass was reported by averaging the major peak in the total ion current (TIC). For each protein, the charge state series was deconvoluted (maximum entropy setting) using Agilent MassHunter Bioconfirm. 4.4.10. Preparative, semi-preparative, and analytical RP-HPLC Crude peptides were purified by reverse phase-HPLC. The peptides were dissolved in 99:1 or 95:5 A:B and 6 M guanidine hydrochloride was added depending on the solubility of the peptide. For preparative RP-HPLC purification, we used an Agilent Zorbax SB C8 column (21.2 x 250 mm, 7 tm) at a flow rate of 10 mL min' at 1-41% or 5-45% B over 80 min. For semipreparative RP-HPLC purification, we used an Agilent Zorbax SB C 18 column (9.4 x 250 mm, 5 ttm) at a flow rate of 5 mL/min over the same gradient. UV absorbance was monitored at 214 nm. Purity of the fractions was analyzed by MALDI or LC-MS. HPLC fractions from preparative or semi-preparative HPLC were spotted with MALDI matrix alpha-cyano-4-hydroxycinnamic acid (CHCA) in 50% A: 50% B and checked for the correct molecular masses. The analytical RPHPLC Agilent C 18 Zorbax SB column (2.1 x 150 mm, 5 pim) was used to confirm the purity of fractions at a flow rate of 0.5 mL/min over a linear gradient of 1-51% B over 12 min. Analytical HPLC UV absorbance traces were measured at 214 nm. 4.4.11. Conjugation of docetaxel, doxorubicin, and MMAF to G 5-LRRLRAC 187 Doxorubicin-maleimide (16 mg) or docetaxel-maleimide (30 mg) was added to a 16.5 mM solution of peptide G 5-LRRLRAC in DMF to a final concentration of 16.5 mM (1:1). The reaction mixtures were allowed to stir at 36'C for 5 hours, followed by additional 10 hours at room temperature, at which time LC-MS analysis indicated consumption of the starting peptide and formation of G5-LRRLRAC(doxorubicin) and G5-LRRLRAC(docetaxel) peptide conjugates. - The products were purified by semi-preparative RP-HPLC to give 35 mg (82% yield) of G5 LRRLRAC(doxorubicin) and 50 mg (75% yield) of G5-LRRLRAC(docetaxel). MMAF was also ligated to G 5-LRRLRAC using 2 mg maleimide-MMAF (20 mM) in DMSO and 20 mM G 5-LRRLRAC (2.5 mg). Reaction was incubated at room temperature for 2 - h. LC-MS indicated consumption of the peptide starting material and formation of G5 LRRLRAC(MMAF) product. The product was purified by semi-preparative RP-HPLC to give 4.8 mg (86% yield). strategies LRRLRAC(doxorubicin)K(biotin) and were utilized for the synthesis G5-LRRLRAC(MMAF)K(biotin) in of G5 which G5 - coupling - Similar LRRLRACK(biotin) was used as a starting material. 4.4.12. Cyclization of linear peptide using native chemical ligation Cyclization of the L-linear precursor peptide was performed by conversion of the thiazolidine residue into a Cys residue and subsequent native chemical ligation (NCL), according to Figure 4.6.1. To convert the thiazolidine residue into a Cys residue, 30 mg peptide (2 mM) was dissolved in 0.2 M sodium phosphate buffer containing 6 M guanidine-HCl, 20 mM TCEP-HC, and 0.2 M methoxyamine hydrochloride (MeONH 2 HCl). The reaction was incubated at pH 4.0 at RT for 6 h. In the same pot, 4 NCL was performed to cyclize the peptide 188 by adding 10 mM 4-mercaptophenylacetic acid (MPAA) and incubating at pH 7.0 for I h at RT.35 The cyclized product was purified by semi-preparative RP-HPLC. Pure fractions that were confirmed by MALDI and analytical RP-HPLC were combined and lyophilized to give 7 mg, 4.3 [tmol (26.1% yield). The Cys residue in the cyclized peptide was alkylated using 50 mM bromoacetamide and 20 mM TCEP-HCl in 0.2 M sodium phosphate buffer at pH 7.0 for 15 min at RT. Once complete, the reactions were quenched with 100 mM 2-mercaptoethanesulfonate (MESNa) and purified by semi-preparative RP-HPLC. Fractions containing the pure, alkylated L-cyclic peptide were combined and lyophilized to give 2.3 mg, 1.4 pimol (31.7% yield). Cyclization of the D peptide was performed using the same method as described for the L cyclic peptide. The pure D cyclic peptide yield was 6.0 mg, 3.7 jimol (19.1% yield) and the alkylated D cyclic peptide yield was 2.2 mg, 1.3 pimol (35.4% yield). 4.4.13. Protein synthesis inhibition assay CHO-K1 cells were maintained in F-12K medium containing 10% (v/v) fetal bovine serum, 100 U mL' penicillin and 100 ptg mUL streptomycin at 37 'C and 5% CO 2 . CHO-KI (3.0 X 104 per well) were seeded into a 96-well plate and incubated overnight. The following day, the cells were treated with ten-fold serial dilutions of each construct (LDn 1-11) in the presence of 20 nM PA for 30 minutes at 37 'C and 5% CO2 . After treatment, the cells were washed three times with PBS and then treated with 100 jiL leucine free medium containing I pCi ml]' 3H-Leucine (PerkinElmer) for 1 hour at 37 'C and 5% CO2 . After treatment with 3H-Leu, the cells were washed three times with PBS and then suspended in 150 gL of scintillation fluid. 3H-Leu incorporation was measured using a scintillation counter to determine the inhibition of protein synthesis by LFN-DTA in the cytosol. Cells treated with PA only were used for normalization. 189 Each experiment was done in triplicate. The data were plotted using Origin8 using Sigmoidal Boltzmann Fit using equation y = A2 + 1+e 0 and xo represents the log(EC5 o) values. 4.4.14. Translocation of LDn1-11 and cytosolic and total cell lysate extraction CHO-KI (2.0 x 105 per well) were seeded in a 12-well plate and incubated overnight. The cells were treated with 100 nM LDn1-11, LDnI-bio to LDn6-bio, LDn9-bio, and LDnl l-bio in the presence of 20 nM PA for 12 hours at 37 'C and 5% C02. After treatment, the cells were lifted and the cell surface was digested with 0.25% trypsin-EDTA for 5 minutes. The cells were then washed twice with PBS. For cytosolic protein extraction, 0.5 x 105 cells were resuspended in 50 pL of 50 pg mL-1 digitonin in 75 mM NaCl, 1 mM NaH 2PO 4, 8 mM Na2 HPO 4. 250 mM sucrose supplemented with Roche protease inhibitor cocktail for 10 min on ice, and centrifuged for 5 minutes at 13,000 rpm. For total protein extraction, cells were lysed in total cell lysis buffer (25 mM Tris, 150 mM NaCl, I% v/v NP-40, pH 7.5) supplemented with Roche protease inhibitor cocktail on ice for 30 minutes and centrifuged for 10 minutes at 13,000 rpm. Both supernatants were collected for western blot. 4.4.15. Western blot of extracted material Extracted material was run on an SDS-PAGE gel (Life Technologies) for 35 minutes at 165 V and then transferred onto a nitrocellulose membrane soaked in 48 mM Tris, 39 mM glycine, 0.0375% SDS, 20% methanol and using a TE 70 Semi-Dry Transfer Unit (GE) for 1 hour at 17 V. The membrane was blocked at room temperature for 2 hours using LI-COR Odyssey blocking buffer (PBS). The membrane was immunostained overnight with anti-DTA, anti-LF, anti-Erkl/2, or anti-Rab5 in TBST at 4 'C. After incubation, the membrane was washed 190 with TBST and incubated with the appropriate secondary antibody in TBST for 1 hour at room temperature then washed with TBST. The membrane was imaged by LI-COR Odyssey infrared imaging system. 4.5. Acknowledgements This work was funded by MIT start-up funds, MIT Reed Fund, Damon Runyon Cancer Research Foundation Innovation Award, National Science Foundation (NSF) CAREER Award (CHE-1351807) for B.L.P, and NSF Graduate Research Fellowship for A.E.R. We thank Prof. R. J. Collier (Harvard) for his continued support, the NERCE facility (grant: U54 A1057159) for expression of toxin proteins, and Prof. D. Lauffenburger for access to the LI-COR Odyssey Infrared Imager. We also thank Jingjing Ling and Mike Lu their support and comments. 191 4.6. Appendix Table 4.6.1. Peptides used in this investigation (*: alkylation) Sequence Observed (Da) Calculated (Da; mono.) I G 5-AKFRPDSNVRG-CONH 2 1529.8 0.1 1529.8 2 G 5-(P-Ala)KFRPDSNVRG-CONH2 1529.8 0.1 1529.8 3 G 5-(N-Me-Ala)KFRPDSNVRG-CONH 2 1543.8 0.1 1543.8 4 G 5-(propargyl-Gly)KFRPDSNVRG-CONH 2 1553.8 0.1 1553.8 5 G 5-AK(PheF 3)RPDSNVRG-CONH 2 1583.8 0.1 1583.7 6 G 5-AK(C*)FRPDSNVRG-CONH 1689.8 0.1 1689.8 7 G 5-AK(C*)FRPDSNVRG (cyclic) 1672.8 + 0.1 1672.8 8 (D)-G 5-AK(C*)FRPDSNVRG (cyclic) 1672.8 0.1 1672.8 9 G 5-LRRLRAC(doxorubicin)-CONH 1864.9 0.1 1864.9 10 G 5-LRRLRAC(docetaxel)-CON H 2 2129.0 0.1 2129.0 1I G 5-LRRLRAC(MMAF)-CONH 2095.2 0.1 2095.2 1-bio G 5-AKFRPDSNVRGK(biotin)-CONH-, 1884.0 0.1 1884.2 2-bio G 5-(p-Ala)KFRPDSNVRGK(biotin)-CONH 2 1884.0 0.1 1884.2 3-bio G 5-(N-Me-Ala)KFRPDSNVRGK(biotin)-CON H 2 1898.0 0.1 1898.0 1908.0 0.1 1908.0 5-bio G 5-AK(PheF 3)RPDSNVRGK(biotin)-CON H2 1938.0 0.1 1938.0 6-bio G 5-AK(C*)FRPDSNVRGK(biotin)-CONH 1987.0 0.1 1987.0 2219.1 0.1 2219.0 2449.4 + 0.1 2449.4 2 2 2 4-bio G 5-(propargyl-Gly)KFRPDSNVRGK(biotin)-CONH 2 9-bio G 5-LRRLRAC(doxorubicin)K(biotin)-CONH 11-bio G 5-LRRLRAC(MMAF)K(biotin)-CONH, 192 2 2 Table 4.6.2. List of variants Variant LDnI Sequence PSTG( Jn(O JD LDn2 LDn3 LDn4 Ok:PSG.' Go LDn5 LDn6NHa LDn7 R P SG LDn8 d r UA PSTG, faaI. F.I LDn9 LDnIO >(o o L-InIIA gn !A TAPSTGXK13 LDnI Io @WC:T::)4PSTG-- 0,~XJ o LDn3-bio PffGL-AySXCOXIYSTG, 193 LDn4-bio X PSTG .. LDn5-bio TAPSTG aJ LDn6-bio LDn9-bio CL& -C T A)-PSTGJX~1Df ,2 LDnlI1-bio )-PSTGJ&kyZ)Do( D 194 r R Table 4.6.3. EC 50 values of 30-minute protein synthesis inhibition assay. The errors represent fitting errors from Sigmoidal Boltzmann Fit. #represents data from a separate assay. n/a indicates that the data could not be properly fit with the Sigmoidal Boltmann Fit. Protein EC50 (pM) Protein EC5o (pM) LFN-DTA 21 3 LFN-DTA# 74+ 12 LDnI 37 5 LDn9" 128 LDn2 31 10 LDn10' LDn3 31 5 LDnI1' LDn4 48 15 LDn5 24 3 LDn6 36 5 LDn7 n/a LDn8 n/a 18 n/a 428 90 195 0 0H - HN S G, A NH 2 SH K- F~7f RXPYDs8 NYVYR G A 6 MGnd-HC 20 mM TCEP-HC 0.2 M MeONH 2-HCI 0.2 M NaHPO4 pH 4.0, RT 6 h R 10 mM MPAA pH 7.0, RT, 1h D R S s, F F GGGGGAN Alkylation 0 G- V RG0.2 > G H2 N NCL 1. 20 mM TCEP HCI 50 mM bromoacetamide M NaHPO, pH 7.0, RT, 15 min 2. 100 mM MESNa GN 0 Figure 4.6.1. Cyclization of L-linear peptide using native chemical ligation 196 D F N V G (H R HMkyaasONH 0- *.H a) OH 0 0 0 OH 0 OH 0 0 OHOH + O 0 OH 0 doxorubicin 0 0 D 0 0 F /0 lhr, RT, 80% doxorubicin-maleimide NH 2 OH KrNHOH 1tNH b) ) 0 0HO O. ,NH , OH k~k 0 H .. Ho COHi 0 docetaxel 0 0 OH 0 O + H OH OH HO 0 Mukaiyama's reagent H OHO .400 E Anhydrous DCM 16hr, O*C, 72% 0 0 N 0 docetaxel-maleimide Figure 4.6.2. Synthesis of doxorubicin-maleimide (a) and docetaxel-maleimide (b). 197 2 ng 1 + - LFN-DTA + + + + 1 2 3 5 + 6 + 7 + PA LDn + "Ic 8 anti-DTA anti-Erki /2 anti-Rab5 Figure 4.6.3. Western blot of total extraction of LDn 1-8. Cropped blots are used in western blot data. 198 + LF..-DTA 9 + 10 + 4ng 9 + PA LDn 11 anti-DTA anti-Erk1/2 anti-Rab5 Figure 4.6.4. Western blot of total extraction of LDn9-l 1. Cropped blots are used in western blot data. 199 obs.53958.O 0 4 ca. 53957.3 LPSTG, -( ) C., C, ) -I-op )(-!J D,(-R G),,,xi, 55000 52500 1 2 3 4 5 6 7 8 9 10 11 12 13 Time [min] 14 15 16 18 17 19 20 21 LDn2 obs.53957.8 t 0.4 ca. 53957.3 PSTG 5~ - F 52500 55000 I I I I I 1 2 3 4 5 I I I 7 6 8 9 10 11 12 13 Time [min] 14 15 16 17 18 19 20 21 LDn3 ca. 53971 3 JM-DCA:LPSTG> OO®OO()CDD@- &cNI 52500 55000 1 2 3 4 5 6 7 8 9 10 11 12 13 Time [min] 14 15 16 18 17 19 20 21 LDn4 ca. 53981.3 LPSTG, , III 4.6.1. LC-MS Traces LDnl 52500 55000 1 2 3 4 5 6 7 8 9 10 11 12 13 I 14 15 16 17 18 19 20 21 Time [min] LDn5 ca. 54011.3 -cN LPSTG I . 1 200 2 52500 55000 I 3 I 4 I I I 5 6 7 I 8 I 9 I I I I 10 11 12 13 Time [min] I I I I I I I I 14 15 16 17 18 19 20 21 LDn6 PSTG 1 @VIs -1 obs.54118.3 Ca 54118.4 0 NH2 5 O S- 1 1 0.4 N ,oNH (D- @@ l----lGO 52500 55000 04 401. IL I I I I I I I I I I I I I I I I I I 19I 11 12 13 Time [min] 10 9 8 7 6 5 4 3 2 1 14 15 16 21 20 18 17 LDn7 R obs.54101.2 * 0.4 D F ca. 54100.6 s N aw(fLPSTG, NHV R G S2500 55000 I I IIIII IIIIIII II 1 2 4 3 6 5 8 7 14 11 12 13 Time [min] 10 9 15 17 16 19 18 21 20 .Dn8 LEd obs.54101.0 * ca. 54100.6 LPSTGF 50000 2 1 4 3 6 5 8 7 14 12 13 11 Time [min] 10 9 15 17 16 0.4 55000 19 18 21 20 LDn9 obs. 54293.7 0.4 "ooOH o o 0 ca. 54291 5 NH 005000 5 oNH, LPSTG ~Th 1 2 4 3 5 I I 6 7 Time [min] 8 9 11 10 LDn1O 0 NH obs.54558.3 ca. 54555.6 0 0.4 OH0 C0 N 5600055000 H P 6 Time [min] 9 10 11 201 LDn11 obs.54524.0 * 0.4 ca. 54521.8 LPSTG,-D& G (D 50000 55000 I 1 2 3 IL 4 5 6 7 Time [min] 8 I 9I 10 1 11 1 LDnl-bio obs.54310.8 ca. 54312.4 LPSTG, - A (i ( Qi--(-))( -0(C 0 WN -) - 1 2 3 4 5 ('a , V )( Xi 6 Time [min] ONH &4-0 00 5-60 0 K 7 9 8 10 11 LDn2-bio ca1 54000 1 2 3 4 5 6 7 Time [min] 8 9 0.4 54312.4 56000 10 11 LDn3-bio * obs.54328.2 ca. 54325.8 0.4 PSTG>)~ 52500 55000 1 2 1 2 3 7 '6 3 Time [min] 8 9 10 11 LDn4-bio obs.54338.7 0.4 ca. 54335.8 52500 55000 1 202 2 3 4 5 6 7 Time [min] 8 9 10 11 0.4 LDn5-bio ca. 54365.7 DA PSTG5- o ()C.Q's (-)- 52500 5 4 3 2 1 6 -ime [mb] 8 7 9 55000 10 11 LDn6-bio obs54475.5 0.4 ca. 54471.9 1H2 PSTGOA ( NH ON N 52500 55000 1 2 3 4 5 T6 [min mn 7 9ime 8 10 9 11 LDn9-bio Wi)5 1 2 3 obs.54650.3 * 0.4 ca. 54647.1 LPTG 4 5 6 Time [min] 7 8 10 9 11 LDn1I-bio * obs.54880.5 0.4 ca. 54878.6 LPSTG5 J~K~XCX.~ 520 1 2 3 4 5 6 Time [min[ 7 8 9 5057500 10 11 203 4.7. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 204 Alouf, J.E. & Popoff, M.R. (eds.) The Comprehensive Sourcebook of Bacterial Protein Toxins (Academic Press, 2006). Sandvig, K. & van Deurs, B. Delivery into cells: lessons learned from plant and bacterial toxins. Gene Ther. 12, 865-872 (2005). Young, J.A.T. & Collier, R.J. Anthrax toxin: Receptor binding, internalization, pore formation, and translocation. Annu. Rev. Biochem. 76, 243-265 (2007). Klimpel, K.R., Molloy., S.S., Thomas, G. & Leppla, S.H. Anthrax toxin protective antigen is activated by a cell-surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad Sci. U S. A. 89, 10277-10281 (1992). Milne, J.C., Furlong, D., Hanna, P.C., Wall, J.S. & Collier, R.J. Anthrax protective antigen forms oligomers during intoxication of mammalian-cells. J. Biol. Chem. 269, 20607-20612 (1994). Kintzer, A.F. et al. The protective antigen component of anthrax toxin forms functional octameric complexes. J. Mol. Biol. 392, 614-629 (2009). Pannifer, A.D. et al. Crystal structure of the anthrax lethal factor. Nature 414, 229-233 (2001). Mogridge, J., Cunningham, K. & Collier, R.J. Stoichiometry of anthrax toxin complexes. Biochemistry 41, 1079-1082 (2002). Krantz, B.A., Trivedi, A.D., Cunningham, K., Christensen, K.A. & Collier, R.J. Acidinduced unfolding of the amino-terminal domains of the lethal and edema factors of anthrax toxin. J. Mol. Biol. 344, 739-756 (2004). Arora, N. & Leppla, S.H. Residues 1-254 of anthrax toxin lethal factor are sufficient to cause cellular uptake of fused polypeptides. J Biol. Chem. 268, 3334-3341 (1993). Krantz, B.A., Finkelstein, A. & Collier, R.J. Protein translocation through the anthrax toxin transmembrane pore is driven by a proton gradient. J. Mol. Biol. 355, 968-979 (2006). Krantz, B.A. et al. A phenylalanine clamp catalyzes protein translocation through the anthrax toxin pore. Science 309, 777-781 (2005). Nassi, S., Collier, R.J. & Finkelstein, A. PA(63) channel of anthrax toxin: An extended beta-barrel.Biochemistry 41, 1445-1450 (2002). Thoren, K.L., Worden, E.J., Yassif, J.M. & Krantz, B.A. Lethal factor unfolding is the most force-dependent step of anthrax toxin translocation. Proceedings of the National Academy of Sciences of the United States ofAmerica 106, 21555-21560 (2009). Feld, G.K. et al. Structural basis for the unfolding of anthrax lethal factor by protective antigen oligomers. Nature Structural & Molecular Biology 17, 1383-U245 (2010). Blaustein, R.O. & Finkelstein, A. Diffusion limitation in the block by symmetric tetraalkylammonium ions of anthrax toxin channels in planar phospholipid bilayer membranes. J. Gen. Physiol. 96, 943-957 (1990). Bachran, C. et al. Anthrax toxin-mediated delivery of the Pseudomonas exotoxin A enzymatic domain to the cytosol of tumor cells via cleavable ubiquitin fusions. Mbio 4, e00201-13 (2013). Hobson, J.P., Liu, S.H., Rono, B., Leppla, S.H. & Bugge, T.H. Imaging specific cellsurface proteolytic activity in single living cells. Nat. Methods 3, 259-261 (2006). 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Arora, N., Klimpel, K.R., Singh, Y. & Leppla, S.H. Fusions of anthrax toxin lethal factor to the ADP-ribosylation domain of Pseudomonas exotoxin A are potent cytotoxins which are translocated to the cytosol of mammalian cells. J. Biol. Chem. 267, 15542-15548 (1992). Milne, J.C., Blanke, S.R., Hanna, P.C. & Collier, R.J. Protective antigen-binding domain of anthrax lethal factor mediates translocation of a heterologous protein fused to its amino- or carboxy-terminus. Mol. Microbiol. 15, 661-666 (1995). Cordero, C.L., Kudryashov, D.S., Reisler, E. & Satchell, K.J.F. The actin cross-linking domain of the Vibrio cholerae RTX toxin directly catalyzes the covalent cross-linking of actin. J. Biol. Chem. 281, 32366-32374 (2006). Arora, N. & Leppla, S.H. Fusions of anthrax toxin lethal factor with shiga toxin and diphtheria toxin enzymatic domains are toxic to mammalian cells. Infect. Immun. 62, 4955-4961 (1994). von Moltke, J. et al. Rapid induction of inflammatory lipid mediators by the inflammasome in vivo. Nature 490, 107-U126 (2012). Pentelute, B.L., Sharma, 0. & Collier, R.J. Chemical dissection of protein translocation through the anthrax toxin pore. Angew. Chem. Int. Ed. Engl. 50, 2294-2296 (2011). Popp, M.W., Antos, J.M., Grotenbreg, G.M., Spooner, E. & Ploegh, H.L. Sortagging: a versatile method for protein labeling. Nat. Chem. Biol. 3, 707-708 (2007). Collier, R.J. & Kandel, J. Structure and activity of diphtheria toxin. 1. Thiol-dependent dissociation of fraction of toxin into enzymically active and inactive fragments. J. Biol. Chem. 246, 1496 (1971). Adam, S.A., Marr, R.S. & Gerace, L. Nuclear-protein import in permeabilized mammalian-cells requires soluble cytoplasmic factors. J. Cell Biol. 111, 807-816 (1990). Li, P. & Roller, P.P. Cyclization strategies in peptide derived drug design. Curr. Top. Med Chem. 2, 325-41 (2002). Chari, R.V.J., Miller, M.L. & Widdison, W.C. Antibody-drug conjugates: An emerging concept in cancer therapy. Angew. Chem. Int. Ed. Engl. 53, 3796-3827 (2014). Eicher, T. et al. Transport of drugs by the multidrug transporter AcrB involves an access and a deep binding pocket that are separated by a switch-loop. Proc. Natl. Acad Sci. U. S. A. 109, 5687-5692 (2012). Nogales, E., Wolf, S.G. & Downing, K.H. Structure of the alpha beta tubulin dimer by electron crystallography. Nature 391, 199-203 (1998). Rabideau, A.E., Liao, X. & Pentelute, B.L. Delivery of mirror image polypeptides into cells. Chemical Science 6, 648-653 (2015). Simon, M.D. et al. Rapid flow-based peptide synthesis. ChemBioChem 15, 713-720 (2014). Bang, D. & Kent, S.B.H. A one-pot total synthesis of crambin. Angew. Chem. Int. Ed. 43, 2534-2538 (2004). Johnson, E.C.B. & Kent, S.B.H. Insights into the mechanism and catalysis of the native chemical ligation reaction. J. Am. Chem. Soc. 128, 6640-6646 (2006). 205 Chapter 5: A D-Amino Acid at the N-terminus of a Protein Abrogates its Degradation by the N-End Rule Pathway Portions of the work presented in this chapter have been submitted for publication: Rabideau, A.E. & Pentelute, B.L. "A D-Amino Acid at the N-terminus of a Protein Abrogates its Degradation by the N-End Rule Pathway." 206 5.1. Introduction The chirality of biomolecules in nature is critical for substrate recognition, protein binding, and product formation. While E3 ubiquitin (Ub) ligases of the Ub/proteasome system have promiscuous substrate binding sites, the chirality of protein substrates has never been investigated. 1-3 Since the homeostasis of a cell's protein and amino acid concentrations is regulated by the proteasome, perturbation of proteasomal activity through substrate modification can drastically affect the intracellular equilibrium. Here, we investigated the effect on proteasomal degradation after the incorporation of one mirror image (D)-amino acid on the Nterminus of otherwise all L-proteins. Varshavsky and co-workers have characterized the N-end rule as it relates to the intracellular stability of proteins.4-6 According to the N-end rule, the identity of the N-terminal amino acid mediates the selective degradation of specific proteins through the Ub/proteasome system. The N-terminal degradation signals are termed N-degrons, which can range from stabilizing to destabilizing residues. The primary destabilizing residues are R, K, H, L, F, Y, W, and I, while D, E, N, Q, and C are also destabilizing after modifications. N-degrons are recognized by N-recognins, or E3 Ub ligases, which interact with E2 Ub ligases to polyubiquitinate proteins for proteasomal degradation. 6 To date, the N-end rule has only been defined for L-amino acids.7- 9 The main challenge with probing the N-terminus with non-natural functionalities is cytosolic delivery of the proteins. To overcome this, we utilized a platform derived from nature that enables the delivery of different types of proteins into the cytosol of cells. 207 Anthrax lethal toxin from Bacillus anthracis utilizes the protective antigen (PA) pore to deliver lethal factor (LF) into the cytosol.' 0 Protein translocation by anthrax lethal toxin has been extensively characterized. In short, to obtain entry into the cell, PA binds to an anthrax receptor on the cell surface and forms the PA pre-pore.' 1-16 LF binds to the PA pre-pore then the entire complex is endocytosed and endosomal acidification triggers a conformational rearrangement of the PA pre-pore to form a pore.1 7- 18 LF translocates through the pore via a charge state- dependent Brownian ratchet into the cytosol.12, 19 The N-terminal domain of LF (LFN) is sufficient to bind to the PA pre-pore, but does not cause any intracellular toxicity. 2' The PA/LFN delivery system has been engineered to deliver various peptide, protein, and small molecule cargos into the cell cytosol.2 1 2 3 The majority of eukaryotic proteins from ribosomal translation are composed of L-amino acids and achiral glycine. In order to study the intracellular stability of proteins containing Damino acids, we used Sortase A (SrtA) from Staphylococcus aureus to ligate one D-amino acid onto the N-terminus of natural proteins for the delivery though PA pore. 24 Furthermore, we incorporated a cleavable linker that releases the cargo protein from LFN after translocation into the cytosol further allowing us to characterize N-terminal D-amino acids on proteins other than LFN, including A-chain of diphtheria toxin (DTA),2 5 (DARPin), 27, 28 2 26 designed ankryin repeats protein and a Ras/Rap 1-specific endopeptidase (RRSPc 2 ). 29, 30 2 We opted to use a hindered disulfide cleavable linker that was small in structure such that it could translocate ' through the PA pore efficiently and readily cleaved in the reducing environment of the cytosol. 3 Furthermore, we applied this methodology to the targeted delivery of stabilized RRSPc 2 into pancreatic cancer cells (AsPC-1) using the epidermal growth factor receptor (EGFR) to interrupt the mitogen activated protein kinase (MAPK) pathway. 208 5.2. Results 5.2.1. Sortase A Attaches One D-Amino Acid onto the N-Terminus of LFN-DTA The X-LFN-DTAmut constructs were produced through enzyme-mediated ligation of XALPSTGG onto the N-terminus of the LFN-DTAmut. The N-terminal amino acid (X) represents a natural L-amino acid (LX) or its mirror image D-amino acid (DX), while the remaining residues had all L conformation (Figure 5.2.la). Each XALPSTGG peptide was ligated to G5 -LFNDTAmut using Staphylococcus aureus sortase A to yield XALPSTG-LFN-DTAmut constructs (XLFN-DTAmut; Figure 5.2. 1b). A one-pot ligation scheme was used for each reaction. 5.2.2. One N-Terminal D-Amino Acid Stabilizes LFN-DTA to Proteasomal Degradation We used the A chain of diphtheria toxin (DTA) as a first measure of proteasomal degradation inside cells. We specifically used a mutant form of DTA (E148S; DTAmut) that is reported to be 300-fold less active than wild-type DTA, allowing us to capture the translocation efficiency of each X-LFN-DTAmut construct over a wider dynamic range than wild-type DTA. DTA inhibits protein synthesis by ADP ribosylating elongation factor-2.2 5 , 26 To a first approximation, less protein synthesis inhibition (i.e. fewer DTA molecules) indicates decreased protein stability or hampered translocation. In order to analyze protein stability, we used two assays: 1, protein synthesis inhibition due to DTA; and 2, western blot analysis of the cytosolic fraction. Utilizing these two established assays we probed the intracellular stability of constructs containing N-terminal D-amino acids after translocation through PA pore (Figure 5.2.1c). 209 a. 0 4 .H2 N H2N R R L-amino acid D-amino acid LX DX b. SrtA X-ALPSTGG+G X-ALPSTG PA receptor proteasome endosome ? cytosol Figure 5.2.1. Intracellular stability was monitored for X-LFN-DTA constructs delivered through protective antigen pore. a. LX- (left) and DX- (right) amino acids ligated to the Nterminus of LFN-DTA (LFN: green; DTA: orange). b. XALPSTGG peptides, where X represents either L or D amino acids, are ligated onto G5 -LFN-DTA using Sortase A (SrtA) to form X-LFNDTA constructs. c. Translocation of X-LFN-DTA constructs is achieved using protective antigen (PA) of anthrax toxin. 210 For the protein synthesis inhibition assay, Chinese hamster ovary (CHO-Ki) cells were treated with ten-fold serial dilutions of each construct for 6 hours followed by incubation with 3 H-Leu in leucine-free medium for 1 hour. The fraction of protein synthesis was counted by a scintillation counter (Figure 5.6.1 and Table 5.6.1). According to Figure 5.2.2a, regardless of chirality, stabilizing amino acids such as A and V on the N-terminus of LFN-DTAmut had relatively similar EC5o values as the wild-type control (LFN-DTAmIt), which contains LA at the N-terminus. Furthermore, destabilizing residues like LW on the N-terminus of LFN-DTAm,,t had significantly higher EC 5o values (i.e. less DTA activity) than the wild-type control while the DW- LFN-DTAmut construct had similar activity as the control, which suggests that D-amino acids act as stabilizing residues. As a second measure of protein stability endowed by N-terminal D-amino acids, the XLFN-DTAmut constructs were delivered into CHO-KI cells, lysed using digitonin lysis buffer, and analyzed by western blot. CHO-Kl cells were treated with X-LFN-DTAmut constructs in the presence of PA for 6 hours. As a control, select conjugates (LV-, DV_. LA-, DA-, LW, and DW- LFN-DTAm 1 Ut) were incubated in the presence of lactacystin, a proteasome inhibitor. After 6 hours, the cells were lysed using a digitonin lysis buffer then analyzed by western blot. The western blot was immunostained with LF, Erkl/2, and Rab5 antibodies. Based on the findings in Figure 5.2.2b, the western blot results corroborated the protein synthesis inhibition data. The samples treated with lactacystin all showed strong anti-LF bands, indicating that the proteasome played a key role in degrading the LX-LFN-DTAmut constructs but had no observable effect on the DX-LFN-DTAmut constructs. Furthermore, these data indicated that each construct translocated efficiently into the cells, regardless of N-terminal amino acid. A similar analysis was made for 211 the remaining constructs. These data support the N-end rule for N-terminal L-amino acids, while all N-terminal D-amino acids stabilized X-LFN-DTAmt to degradation. To verify the mechanism of translocation and endosome escape, we incubated the X-LFNDTAmut constructs with a mutant PA (PA[F427H]), 4 a vacuolar H+-ATPase inhibitor (Bafilomycin Al), or at 4 'C with CHO-Ki for 6 hours. In all cases, no material was found to translocate into the cytosol by western blot (Figure 5.6.2). These controls indicated that delivery of the X-LFN-DTAmut constructs into the cytosol is dependent on functional endocytic machinery and PA. Moreover, we found that our observations were not cell-specific. After translocation, we observed protein stabilization in human embryonic kidney cells (HEK-293T) and human cervical cancer cells (HeLa), similar to the stabilization obscrvcd in CO-KI cells (Figure 5.6.3). 5.2.3. Proteasomal Stabilization is Not an Artifact of the Sortag SrtA ligation adds a short linkage (i.e. LPSTG5 ) between the N-terminal amino acid (X) and the start of the protein. We used native chemical ligation (NCL)35 to prepare constructs with native sequences for comparison with sortagged proteins. For our analysis. native LF1-T)TA.a was synthesized containing L-alanine at the N-terminus and compared to NCL synthesized constructs containing N-terminal DA, 'W, or DW.3 6 Each native construct was translocated in CHO-Ki cells and their protein stability was compared to the sortagged conjugates using western blot. Based on western blot of the translocated material in Figure 5.6.4, both native and sortagged constructs containing LA, DA, and DW had similar protein stability, while LW in both cases was highly degraded. These observations indicated that stabilization of LFN-DTAmnt through the incorporation of one N-terminal D amino acid is not an artifact of the sortag. 212 a. 10.0- -r DX-LF N-DTAt E LX-LFN-DTA, 1.0-- C 0. E LFN-DTAt - ' . E DYDWDMD FDHj Dp LP DKDL DRDE DTDV LV DALM LI DD LDNLT DS LA LL SLD LN LR LE LH LF LK LYLW X-LF N-DTAmut b. 1 ng X-LF N-DT \ LDO 20 [M lactacystin LV DV LA (A LW DW LD DD LE DE LF DF LM DM LN DN LP Dp LV DV LA DA LW DW LDm LH DH LI DI LK DK LL DL LR DR LS DS LT DT LD, anti-LF anti-Erk1/2 anti-Rab5 1 ng X-LFN-DTAm LD anti-LF anti-Erk1/2 anti-Rab5 1 ng X-LFN-DTAmt LDm LY DY anti-LF anti-Erk1/2 anti-Rab5 Figure 5.2.2. One N-terminal D-amino acid on LFN-DTA enhances protein stability. a. Translocation X-LFN-DTA constructs was analyzed by protein synthesis inhibition assay in CHO-Ki cells after 6 hours (n = 3). EC5o values from the protein synthesis inhibition assay were graphed for all LX-LFN-DTA or DX-LFN-DTA constructs. EC50 values (and error bars) were determined using a Boltzmann distribution fit. b. LV, DV-, L A-, DA-, LW, and DW-LFN-DTAmut were translocated into CHO-Ki cells in the presence of 20 nM PA for 6 hours then extracted using digitonin lysis buffer and analyzed by western blot. As a groteasomal inhibitor, 20 1 M lactacystin was used. Translocation of all LX-LFN-DTA or X-LFN-DTA constructs was analyzed by western blot. 213 5.2.4. LFN-DTA with One N-terminal D-Amino Acid is Stable In Vitro Protein translocation through the PA pore requires docking of LFN, specifically the Nterminal portion on the surface of the PA pre-pore." To decouple protein translocation and protein degradation, we analyzed the in vitro rates of degradation of the X-LFN-DTAmut constructs in rabbit reticulocyte lysate (RRL). Pure X-LFN-DTAmut proteins were incubated in the presence of 70% RRL at 37 'C. Samples at various time points were pulled and analyzed by western blot using LFN and P-actin antibodies (Figure 5.2.3a). As indicated in Figure 5.2.3b, only LW-LFN-DTAmut experienced significant protein degradation after 120 minutes. These data further support the in vivo protein synthesis inhibition and western blot analyses, which collectively suggest that N-terminal D-amino acids stabilize LFN-DTAmUt to protein degradation. 5.2.5. LFN-DTA with One N-terminal D-Amino Acid is Not Ubiquitinated Polyubiquitination of proteins by El, E2, and E3 Ub ligases is a critical step before proteasomal degradation.3 7 In order to identify the mode in which proteins with N-terminal Damino acids are stabilized, we used a pull-down assay. Protein constructs containing biotin (XK(bio)-LFN-DTAmut, where K(bio) represents biotin and X represents LV, LW, DW, LR, or DR) were synthesized (Figure 5.6.5 and Figure 5.6.6). Each construct was incubated with 70% RRL for 10 minutes to allow for polyubiquitination followed by pull-down using streptavidin beads and western blot analysis (Figure 5.2.3c). Streptavidin and anti-ubiquitin staining indicated that only the constructs containing N-terminal LW and 'R were ubiquitinated, while the negative control LV- as well as both DW- and DR-K(bo)-LFN-DTAmut constructs contained no detectable ubiquitination. These results indicate that destabilizing amino acids like LW and LR recognized by the Ub/proteasome system and are readily degraded, while are not ubiquitinated nor degraded by the proteasome. (Figure 5.2.3d). 214 DW and DR are a. X-LF-DTA, anti-LF 60 %pW W DV X-LF N-DTAMA time (min) 0 60 10 anti-LF *-W" % - # 0 60 10 -W 120 0 10 60 10 DA 120 0 10 DW 60 120 0 $woo W .- 120 oW WO M o %=0 anti-p-actin % anti-f-actin 120 :: : : I! :: t- . b . 1004 A : -- LW LA LV 10 0 % time (min) OW 10 u- 60 120 tow %WW W W 1 --. LV-LF N-DTA -+-'A-LFN-DTA,. ,k'W-LFN-DTAm~ - ,- DV-LF i-DTAm N mul --- DA-LFk-DTA, +-DW-LF -DTA 0 20 40 60 80 100 Time (minues) 120 Pull down: Streptavidin ladder RRL RRL X-K(bio)-LFN-DTAmuT (kDa) only only LV LW OW LR DR C. 188 Streptavidin 49 98 anti-Ubiquitin 62 Figure 5.2.3. One N-terminal D-amino acid prevents ubiquitination of LFN-DTA. a. The stability of LV-, DV-, LA- , DA-, LW-, and DW-LFN-DTAmut (4 ng) was monitored in 70% RRL over time at 37 'C and then analyzed by western blot. b. The concentration of X-LFN-DTAmut (%) was plotted against time, based on the western blot in Figure 5.2.3a. c. X-K(bio)-LFNDTAmut constructs (1 [tM; X represents 'V, LW, DW, LR, and DR) were incubated in 70% RRL for 10 minutes at 37 'C then pulled down using streptavidin beads for 1 hour. Elution samples were analyzed by western blot (streptavidin and anti-ubiquitin staining). 215 5.2.6. N-terminal Stabilization is Not Protein-Specific In order to study the stabilization of proteins other than LFN, we incorporated a cleavable linker to separate the attached cargo from LFN once the entire construct translocated into the cytosol but was also stable outside of the cell. The cleavable linker allowed for the intracellular stabilization of different types of cargo to be explored using the PA/LFN delivery platform. We used a hindered disulfide cleavable linker (Figure 5.6.7), which contains steric hindrance around a disulfide bond thus increasing its stability toward reduction.3 1 While hindered disulfide bonds have a wide range of rates of reduction, the penicillamine-cysteine bond was chosen since it is more stable than unhindered disulfide bonds, but can be readily reduced at physiological conditions over the time scale of our experiments (Figure 5.6.8). Enabled by the hindered disulfide cleavable linker, we analyzed the stability of X-DTAm,,t and X-DARPin protein cargo after translocation into the cell cytosol. For our analyses, 1-X (Figure 5.2.4a) and 2-X (Figure 5.6.9a) conjugates were synthesized containing X-DTAmut and X-DARPin linked to LFN through a hindered disulfide, respectively, where X represents LV, DV, LA, DA, LW, and DW. The protein stability of X-DTAmut was analyzed using the protein synthesis inhibition assay after translocation in CHO-KI cells for 6 hours. According to the results in Figure 5.2.4b, the activity of X-DTAmut was stabilized through the addition of one Nterminal D-amino acid. These data were confirmed through the use of western blot analysis. CHO-Ki cells were treated and lysed using the same conditions previously described. According to the anti-LF immunostaining in Figure 5.2.4c, there was no detectable full-length material after the 6 hour incubation. These results indicate sufficient cleavage of the hindered disulfide material in the reducing environment of the cytosol after translocation. The bands corresponding to cleaved LFN further demonstrate the reduction. The presence of bands corresponding to DTA 216 in Figure 5.2.4c corroborated the protein synthesis inhibition assay data; the cleaved X-DTAmut proteins in which X represented a D-amino acid were stabilized to degradation upon cleavage from LFN. A similar analysis was made for the X-DARPin protein cargo, in which we demonstrated the stabilization of biotinylated DARPin using one N-terminal D-amino acid (Figure 5.6.9b). Regardless of the side-chain functionality, both DTAmut and DARPin proteins were stabilized using one N-terminal D-amino acid. 217 a. LA S0 LPSTGLRLA-N X-ALPSTG C. )kNH2 media 1 ng LFNDTA 1-X - LD G5 VL 'V OV LA DA LWDW LV anti-LF OH H0 anti-DTA 1-X anti-Erk1/2 W40U5RW9 anti-Rab5 b. 'A AIWLPSTGLRLkN.ANH2 biotin X-AGAKLPSTG . 2-X NOH d. X-K(bio)-DARPin 2-X 'V 1 ng LFN D V media LV DV - LV DV LA DA LW DW LV DV Q*su"usmma&#6 4 anti-LF Streptavidin anti-Erki /2 - ~. - _- -~ anti-Rab5 Figure 5.2.4. N-terminal D-amino acid stabilization is not limited to LFN. a. Molecular composition of X-DTAmut conjugated to LFN through a hindered disulfide (1-X), where X represents G5 , LV DV LA, DA, LW, or DW. b. Molecular composition of X-DARPin conjugated to LFN through a hindered disulfide (2-X), where X represents LV, DV, LA, DA, LW, or DW. c. CHO-Ki cells were treated with 100 nM 1-X conjugates in the presence of 20 nM PA for 6 hours then extracted using digitonin lysis buffer and analyzed by western blot. The absence of full-length material suggests that each construct was appropriately reduced in the cytosol. Furthermore, LFN (LA as the native N-terminus) and X-DTAmut bands indicated cleavage and stabilization of the X-DTAmut cargo with one N-terminal D-amino acid. The post-incubation media was analyzed by Western blot to indicate the stability of the hindered disulfide over the time of the experiment. d. CHO-KI cells were treated with 100 nM 2-X conjugates in the presence of 20 nM PA for 6 hours then extracted using digitonin lysis buffer and analyzed by western blot using anti-LF and streptavidin staining. LFN and X-DARPin bands indicated cleavage and stabilization of the X-DARPin cargo with one N-terminal D-amino acid. 218 5.2.7. RRSPC 2 is a Ras/Rapl-Specific Endoprotease Ras is a GTPase responsible for activating signal transduction pathways. Interrupting the Ras GTPase activity interferes with the signal transduction of the PI3K/Akt and MAPK/Erk pathways, which deregulates cell growth and survival processes.38 Recently, a Ras/Rap 1-specific endopeptidase (RRSP) has been characterized from the Vibrio vulnificus MARTX toxin.2 9' 30 Since RRSPC 2 can recognize mutant forms of Ras, it is an attractive candidate for treating cancer cells expressing mutant Ras proteins. 39 The N-terminus of RRSPC 2 is a destabilizing residue (i.e. Phe3669); therefore, delivery would lead to rapid degradation in the cytosol. Thus, we applied our methodology of protein stabilization to stabilize the catalytic domain of RRSP (RRSPc 2) for its targeted delivery into pancreatic cancer cells for interruption of the MAPK pathway. LFN conjugates containing X-RRSPc 2 were synthesized (3-X), where X represents LF or DF (Figure 5.2.5a). The Ras protease activity of each conjugate was analyzed in vitro (Figure , 5.6.10 and Figure 5.6.11) using 25-fold excess KRas substrate. In all cases containing RRSPC 2 the KRas substrate was completely cleaved after 10 minutes. These results indicated that Nterminal modifications on RRSPC 2 did not affect the Ras protease activity. Furthermore, the translocation efficiency of RRSPC 2 through PA pore was analyzed by the protein synthesis inhibition assay. According to the results, RRSPC 2 translocated as efficiently as the wild-type LFN-DTA control (Figure 5.6.12). 5.2.8. EGFR-Targeted Delivery of RRSPc 2 Interrupts the MAPK Pathway Modified PA proteins have recently been developed to retarget PA for the recognition of non-native receptors. 40-42 Delivery of LFN can be targeted to specific cells through a mutant PA (N682A/D683A; mPA) and an added targeting domain, such as EGF to target EGFR. Since 219 EGFR is overexpressed in certain types of cancer cells, including pancreatic cancer, it is an ideal target for specific delivery of bioactive cargos. We investigated the delivery of stabilized RRSPC 2 (3-DF) using EGFR-targeted PA (mPA-EGF), which recognizes EGFR. We analyzed the effect of RRSPC 2 on the MAPK pathway in the AsPC-1 pancreatic cancer cell line, which overexpresses EGFR (Figure 5.6.13) and contains mutant KRas (G12D). AsPC-l cells were treated with the 3-X conjugates containing X-RRSPc 2 in the presence of mPA-EGF, a mutant mPA-EGF (mPA[F427H]-EGF) that inhibits translocation, or I pM EGF to outcompete mPA-EGF for EGFR. In addition, constructs that lack the cleavable linker ('A-LFN-RRSPc 2 and LF-LFN-RRSPC2) were used as controls. After treatment, the cells were lysed with digitonin lysis buffer or RIPA buffer to analyze translocation or bioactivity, respectively. According to Figure 5.2.5b, after cleavage, RRSPC 2 was significantly more stable than 'F-RRSPC 2 DF- in the cytosol. Furthermore, 3- F-treated cells contained cleaved KRas and had approximately 6-fold reduction in pErkl/2 protein levels in the total cell lysate, similar to the positive control, LA-LFN-RRSPC 2 (no disulfide linker), with respect to the mPA-EGF only control (Figure 5.6.14). The 3-LF-treated cells showed no obvious Ras protease activity, similar to the negative control LF-LFN-RRSPC 2 (no disulfide linker). These results indicated that one N-terminal DF stabilized RRSPC 2 to degradation and perturbed the MAPK signaling pathway in AsPC-1 pancreatic cancer cells. 220 a. LRLA-N LPSTG 5 'A - NH, biotin X-AG 5 KLPSTG5 RRSP NOH 3-X 1 [tM EGF - - - - - - - LF DF LF DF LF DF + + - - + + - - - - + + + - - mPA[F427H]-EGF - - + + mPA-EGF - LFN DF LA - - LF LA + 3-X - - 2 .5ng --- - X-LFN-RRSPC - DF-K(bio)-RRSPC 2 - b. Streptavidin 2> L CO 0 anti-Erki /2 anti-Rab5 anti-KRas C KRas KRasc U- Aii- a ftw a 0 anti-pErki /2 0 low" 00 of Is to 01 Nfa,&-" is us - anti-Erk1 /2 00000 Figure 5.2.5. Precision delivery of stabilized RRSPC 2 through EGFR into pancreatic cancer cells interrupts the MAPK pathway. a. Molecular composition of X-RRSPC 2 conjugated to LFN through a hindered disulfide (3-X), where X represents 'F or DF. b. AsPC-1 cells were treated with 100 nM 3-DF or 3-LF in the presence of 10 nM mPA-EGF for 9 hours. Controls included mPA-EGF only, -DF only, X-LFN-RRSPC2 (X represents 'A or LF) plus mPA-EGF, and wild-type LFN plus mPA-EGF. In addition, incubation with 1 tM EGF was used to outcompete mPA-EGF or 10 nM mPA[F427H]-EGF to arrest translocation from the endosome. The cytosolic fraction was extracted using digitonin lysis buffer and total cell lysate was C-terminal extracted using RIPA buffer. Both lysates were analyzed by western blot (KRasc: fragment of KRas). 221 5.3. Discussion For the first time, we have demonstrated the stabilization of proteins to proteasomal degradation using one D-amino acid at the N-terminus. This phenomenon was evident for LFNDTAmut, DTAmut, DARPin, and RRSPC 2 proteins delivered into the cytosol using the PA/LFN delivery system. Our findings suggest that stabilization using D-amino acids at the N-terminus of proteins is not protein-specific. Taken together, we updated the N-end rule to include D-amino acids as stabilizing residues. In the ubiquitin/proteasome system, an E3 ubiquitin ligase forms a complex with an E2 ubiquitin ligase, in order to conjugate ubiquitin onto the fated protein. 43 ' 44 In particular, the UBR box domain within E3 ubiquitin ligases is responsible for recognition of the N-terminus of protein substrates. Analysis of crystal structures of the UBR box domains from the Ubrl and Ubr2 E3 ubiquitin ligases revealed critical hydrogen bonds between the first two residues of the substrates. 45' 46 We hypothesized that inverted stereochemistry at the alpha carbon of N-terminal residue would interrupt the hydrogen bonding network, prevent substrate binding in the UBR box, and inhibit ubiquitination. Through a pull-down assay, we demonstrated that LFN-DTAmUt constructs containing one N-terminal D-amino acid were not ubiquitinated, while the constructs containing an N-terminal L-amino acid were highly ubiquitinated. Thus, proteins with Nterminal D-amino acids cannot be ubiquitinated for proteasomal degradation. The PA/LFN delivery system has been used to deliver a variety of cargos into the cytosol of cells. Many examples of delivery incorporate the protein, peptide, or small molecule cargo on the C-terminus of LFN, leaving cargo attachment to the N-terminus of LFN relatively unexplored. 2 1 , 22, 222 47 Through these studies, we have found that the PA pore can tolerate short peptide modifications at the N-terminus of LFN. Furthermore, we analyzed the intracellular stability of three different proteins, DTAmut, DARPin, and RRSPC 2 , by installing a hindered disulfide cleavable linker between LFN and the protein cargo. The hindered disulfide was shown to be stable outside of the cell, yet readily reduced in the cell cytosol, freeing the cargo for interactions with intracellular substrates. Previous explorations of translocation have involved the delivery of proteins from N- to C-termini. For the first time, we demonstrated that the PA pore is capable of translocating protein cargo from C- to N-termini, providing further support to the payload promiscuity of PA. Using the PA/LFN delivery platform, we aimed to target stable, bioactive protein cargo to specific cancer cells. Since several pancreatic cancer cell types overexpress EGFR and more than 80% have mutant KRas, the AsPC-l pancreatic cancer cell line was an excellent model for applying our recent findings. We targeted the delivery of the stabilized Ras/RapI endopeptidase, RRSPc 2 to AsPC-l cells through EGFR using a mPA-EGF fusion protein. The EGFR-directed delivery of stabilized DF-RRSPC 2 caused cleavage of KRas, which resulted in downstream interruption of the MAPK pathway. We believe this methodology can be expanded to stabilize disordered proteins prone to degradation as well as to drastically extend the intracellular halflives of therapeutic proteins. D-amino acid incorporation into polypeptides occurs in nature, albeit infrequently when compared to L-amino acid incorporation. Select organisms including bacteria and some eukaryotes utilize racemases to convert L- to D-amino acids or nonribosomal protein synthetases to site specifically insert a D-amino acid within a growing peptide chain 48, . The precise roles of naturally occurring D-amino acid containing polypeptides remains an active area of 223 investigation; however, early studies have shown that polypeptides containing D-amino acids generally occur in higher abundance and are often more active compared to their L-counterparts 48. Furthermore, naturally occurring cyclic peptides which would also evade the N-end rule have been demonstrated to have enhanced stability in cells ways to circumvent the N-end rule. 224 50. Perhaps nature has evolved unexpected 5.4. Experimental 5.4.1. Materials Peptides were synthesized using Fmoc-protected L- and D-amino acids, N,NN',N'Tetramethyl-O-(1 H-benzotriazol- 1 -yl)uronium (HBTU), hexafluorophosphate and 1- [Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate (HATU) purchased from Creosalus and Chemlmpex. Dimethylformamide, piperidine, diisopropylethylamine, trifluoroacetic acid, and triisopropylsilane were purchased from VWR or Sigma Aldrich. All cloning was accomplished using the QuikChange Lightning kit (Agilent) or HiFi DNA Taq Polymerase (LifeTechnologies) and pET SUMO Champion kit (LifeTechnologies). All proteins were expressed in BL21(DE3) from LifeTechnologies. All media for tissue culture was from LifeTechnologies and fetal bovine serum was from Sigma Aldrich. For western blots, nitrocellulose membranes (GE), filters (BioRad), and PBS blocking buffer (LI-COR) were used. We used the following primary and secondary antibodies: LF (Santa Cruz), DTA (abcam), Erkl/2 (Cell Signaling), Rab5 (Cell Signaling), p-actin (Sigma Aldrich), ubiquitin (Santa Cruz), KRas (Cell Signaling), pErk1/2 (Cell Signaling), donkey anti-goat IRdye800 (LI-COR), donkey anti-goat IRdye680 (LI-COR), goat anti-mouse IRdye800 (LICOR), goat anti-mouse IRdye680 (LI-COR), goat anti-rabbit IRdye680, goat anti-rabbit IRdye800 (LI-COR), and streptavidin IRdye680 (LI-COR). Unless specified otherwise, all other reagents were purchased from VWR, Sigma Aldrich, or LifeTechnologies. 5.4.2. Protein Expression and Purification G 5-LFN-DTAmUt, LFN-LPSTGG-H 6 , G 5-DTAnut-C, 17C-LFN-DTAmut, LPSTGG-H5 , G 5-RRSPc 2, and Gs-RRSPc 2 -C were expressed in BL21(DE3) LFN-DTA- . coli cells. The cells were grown at 37 'C to an OD600 0.6-1.0. The proteins were induced with 0.4 mM IPTG 225 overnight at 30 'C. After induction, the cells were pelleted and resuspended in 20 mM Tris pH 7.5 and 150 mM NaCl with protease inhibitor cocktail (Roche), DNaseI (Roche), and lysozyme. The cells were sonicated three times for 20 seconds on ice then spun down for 30 minutes at 35,000 g and 4 IC. The lysate was purified over a HisTrap FF NiNTA column (GE Healthcare) pre-equilibrated with 20 mM Tris pH 8.5 and 150 mM NaCl. Each protein was loaded onto the column then washed with the equilibration buffer and then with 20 mM Tris pH 8.5, 500 mM NaCl, and 40 mM imidazole. Finally, each protein was eluted in buffer containing 20 mM Tris pH 8.5, 500 mM NaCl, and 500 mM imidazole. The protein elutions were desalted using a HiLoad 26/10 desalt column (GE Healthcare) into 20 mM Tris pH 7.5 and 150 mM NaCl. After desalting the proteins, the native N-termini were obtained by cleaving the SUMO protein fusion. For SUMO cleavage, I pg SUMO protease was added per mg protein for 1 hour at room temperature. The cleaved sample was run over a second NiNTA column in order to obtain the pure protein in the flow through and buffer wash. Wild-type PA, PA[F427H], PA-EGF, and PA[F427H]-EGF were expressed in the periplasm of E. coli BL21 (DE3) cells and purified by anion exchange chromatography followed by size exclusion chromatography. 5.4.3. Fmoc Solid Phase Peptide Synthesis All peptides were synthesized using Fmoc fast flow solid phase peptide synthesis (SPPS). All peptides were synthesized on a 0.1 mmol scale on aminomethyl resin with the Rink amide linker. Side-chain protection for the amino acids included: Arg(Pbf), Asn(Trt), Glu(OtBu), Lys(Boc), Lys(Alloc), Ser(tBu), Thr(tBu), Trp(Boc), and Tyr(tBu). For each coupling, 1 mmol (10 eq) amino acid was dissolved in 0.4 M HBTU or HATU in DMF and diisopropylethylamine 226 was used as the activating reagent. Peptides were synthesized according to the protocol in Simon, et al.5 using 3 minute cycles at 60 'C of 40 sec amino acid coupling then 1 min DMF wash then 20 sec 20% piperidine in DMF (v/v) deprotection and finally 1 min DMF wash. After synthesis, peptides were cleaved from the resin with 94% TFA containing 2.5% EDT, 2.5% H 20, and 1% TIPS (v/v) for 7 min at 60 'C. After cleavage, TFA was dried under N2(g>, triturated three times with cold diethyl ether, dissolved in 50:50 A:B, and then lyophilized. For biotinylated peptides, the allyloxycarbonyl (Alloc) protecting group was removed using 4.85 mmol phenylsilane and 39.5 pimol tetrakis(triphenylphosphine)palladium(0) in DCM for 20 min at RT. The resin was washed with DCM then DMF. Biotin (1 mmol) was coupled using 0.2 M HBTU in DMF for 20 min at RT. 5.4.4. Sortase A-Mediated Ligation of X-LFN-DTAmut Constructs The X-LFN-DTAmut constructs were synthesized using the one-pot SrtA-mediated ligation strategy with an evolved SrtA (SrtA*), as reported in Liao, et al. 2 1, 24 GS-LFN-DTAmut was ligated onto each XALPSTGG peptide using the following conditions: 50 jtM GS-LFN-DTAmut, 1 mM XALPSTGG, and 5 ptM SrtA*. These conditions were optimized to maximize the amount of product formed with respect to G5-LFN-DTAmut. The sortagging reactions were incubated at room temperature for 25 minutes then NiNTA was added to each reaction mixture and rotated for an additional 5 minutes to remove the SrtA* from the reaction. At the completion of the reaction, the samples were spin filtered at 4 'C and then buffer exchanged three times into 20 mM Tris pH 7.5 and 150 mM NaCl to remove the excess peptide. LC-MS was used to analyze the purity of each X-LFN-DTAut construct. 227 5.4.5. Protein Synthesis Inhibition Assay with X-LFN-DTAmut Constructs Chinese hamster ovary (CHO-Ki) cells (ATCC) were grown in F-12K medium . containing 10% (v/v) fetal bovine serum and IX penicillin-streptomycin at 37 'C and 5% CO2 For the protein synthesis inhibition assay, 20,000 CHO-KI cells were plated per well in 96-well plates 16 hours before the assay. Each X-LFN-DTAmut construct was diluted ten-fold and in triplicate then PA was added to each well for a final concentration of 20 nM. The plates were incubated for 6 hours at 37 'C and 5% CO 2 . After incubation, the medium was removed and the cells were washed three times with PBS. Leucine-free F-12K containing 3H-leucine (1 ptCi mL-, PerkinElmer) was added to each well and incubated for 1 hour at 37 'C and 5% CO 2. The radioactivity was removed and the cells were washed three times with PBS and suspended in scintillation fluid. Scintillation counting was used to measure the amount of 3H-Leu present, which is indicative of DTA activity (i.e. fraction of protein synthesis). For each sample, the scintillation counts were normalized to a PA only control. The data were fitted with a sigmoidal Boltzmann fit using OriginLab software. 5.4.6. Translocation and Western Blot Analysis with X-LFN-DTAm.t Constructs For western blot analysis, 200,000 CHO-Ki cells were plated per well in 12-well plates 16 hours prior to treatment. Cells were treated with 100 nM X-LFN-DTAmut construct in the presence of 20 nM PA in serum-containing F-12K for 6 hours at 37 'C and 5% CO2 . In select experiments, lactacystin was used to inhibit the proteasome. For this treatment, cells were preincubated with 20 pM lactacystin for 1 hour at 37 0C and 5% CO 2 and then subsequently treated with the X-LFN-DTAmut constructs in the presence of PA. After translocation, the medium was removed and 0.25% trypsin-EDTA was added to each well for 5 minutes at 37 'C and 5% CO2 to remove and non-specifically bound material from the cell surface as well as lift the cells from the 228 plate. The cells were washed twice with PBS at 500 g for 2 minutes at room temperature. In order to obtain the cytosolic fraction, cells were lysed according to the conditions previously reported. In brief, the cells were lysed using 50 pg mL' digitonin in buffer containing 75 mM NaCl, 1 mM NaH2PO 4, 8 mM Na2HPO 4, 250 mM sucrose and protease inhibitor cocktail (Roche) for 10 minutes on ice then spun down at 4 'C for 10 minutes. The extracted lysates were analyzed by western blot. Nitrocellulose membrane and filters were soaked in buffer containing 48 mM Tris-HC, 39 mM glycine, 0.0375% SDS (v/v), and 20% methanol (v/v). Proteins from the gel were transferred to the membrane at 17 V for 1 hour using a TE 70 Semi-Dry Transfer Unit (GE Healthcare). After transfer, the membrane was blocked for two hours at room temperature with blocking buffer (LI-COR) and then incubated with the appropriate primary antibody (LF, Erkl/2, or Rab5) in TBST (50 mM Tris-HCI, 150 mM NaCl, 0.1% Tween 20 (v/v)) overnight at 4 'C. The membranes were washed three times with TBST then stained with a secondary antibody and imaged using an Odyssey Infrared Imaging System (LI-COR). The efficiency of lysis was analyzed by anti-Erkl/2 (cytosolic protein) and anti-Rab5 (early endosome) immunostaining. 5.4.7. Native Chemical Ligation of Native X-LFN-DTAmut The following hydrazide peptides were synthesized in preparation for native chemical ligation of native LFN-DTAmut: XGGHGDVGMHVKEKEK, where X represents LA, DW. DA, LW, or Additionally, 17C-LFN-DTAmut was expressed and purified from E coli using the expression and purification protocol described above. For native chemical ligation, each peptide hydrazide (LA DA, LW, and DW-NHNH 2) was converted to a mercaptophenylacetic acid (MPAA) thioester using the following protocol.36 Each peptide hydrazide (5 mM; 1 mg) was dissolved in 0.2 M 229 phosphate buffer pH 3.0 then 1:10 dilution 0.2 M sodium nitrite (in 0.2 M phosphate buffer pH 3.0) was added to a final concentration of 20 mM sodium nitrite. The reaction was incubated for 10 minutes at -10 'C while stirring to form the peptide azide. After 10 minutes, the reaction was warmed to room temperature and the pH was adjusted to pH 7.0. A 1:1 dilution of 0.1 M MPAA (in 0.2 M phosphate buffer pH 7.0) was added to the reaction for a final concentration of 50 mM MPAA for 20 minutes at room temperature. Formation of the MPAA thioester peptides was confirmed by LC-MS. In one pot, 1 eq 17C-LFN-DTAmut (1 mg, 79.5 pM) was added to 28 eq of MPAA thioester (1 mg, 2.0 mM) and the pH was adjusted to pH 7.0 and incubated for 10 hours at room temperature to allow the NCL to go to completion. The reactions were monitored by LC-MS. Upon completion of the reactions, 50 mM DTT was added to the samples for 20 minutes at room temperature to reduce any disulfides. The samples were buffer exchanged twice into PBS pH 7.0 to remove excess starting materials. The ligated proteins (LA, DA, LW, and 'W-LFN-DTAmUt) were alkylated at the cysteine residue to form pseudoglutamine at position 17, which is asparagine in native LF. The cysteine residue was alkylated using 20 mM 2-bromoacetamide (in 0.2 M phosphate buffer pH 7.0) and pH was adjusted to pH 7.0 then incubated for 15 min then quenched with 50 mM sodium 2-mercaptoethane sulfonate (MESNa in 0.2 M phosphate buffer pH 7.0). The alkylated ligated proteins were desalted into 20 mM Tris pH 7.5, 150 mM NaCL using a HiTrap desalt column (GE Healthcare) then concentrated and analyzed by LC-MS. 5.4.8. In Vitro Stability of X-LFN-DTAm.t Constructs The in vitro stability of X-LFN-DTAmt constructs (where X represents LW, 230 or DW) LV DV LA, DA, was analyzed in rabbit reticulocyte lysate (RRL). Each X-LFN-DTAmut construct (4 ng) was incubated in a 70% RRL solution for up to 120 minutes. Time points were pulled at 0, 10, 60, and 120 minutes. At each time point, 2 jiL of each sample was added to 20 pL IX loading dye and flash frozen. Time points were analyzed by western blot, which was immunostained with LF and P-actin antibodies. The bands were quantified using LI-COR Image Studio software. The rate of degradation graph was plotted according to normalized values. 5.4.9. Streptavidin Pulldown of Ubiquitinated Constructs In order to analyze the ubiquitination of the biotinylated constructs, 1 pM X-K(bio)-LFNDTAmut in 70% RRL (20 pL total volume) at 37 'C for 10 minutes. The samples were then incubated with 20 piL Dynabeads MyOne Streptavidin C1 beads (LifeTechnologies; washed twice with 200 uL 50 mM HEPES pH 7.1, 200 mM KCl, 10% glycerol, 0.02% NP-40) for 1 hour at room temperature. After incubation, the beads were washed twice with the same HEPES buffer then eluted in 20 pl 2X loading dye for 10 minutes at 95 'C. Samples were analyzed by western blot, which was stained with streptavidin and ubiquitin antibody. 5.4.10. Stabilization of X-DTAmut or X-DARPin after Translocation The hindered disulfide conjugates (1-X, 2-X, and 3-X) were synthesized through Cterminal penicillamine (C*) on LFN (LFN-C*) and a C-terminal cysteine (C) on X-DTAmut, XDARPin, or X-RRSPc 2. A three-step ligation strategy was optimized for the synthesis of the hindered disulfide conjugates: 1, sortagging to form LFN-C*; 2, sortagging to form X-DTAmut-CEllman's, X-DARPin-C-Ellman's, or X-RRSPc 2-C-Ellman's; and 3, oxidation to form 1-X, 2-X or 3-X conjugates, where X is any amino acid on the N-terminus of DTAmut, DARPin, or RRSPc 2 . According to the ligation scheme in Figure 5.6.7, the peptide G3LRLAC* was synthesized and purified then sortagged onto LFN-LPSTGG under standard sortagging conditions 231 to yield LFN-C*. Protein cargos like DTAmut, DARPin, and RRSPc 2 were expressed with a C- terminal cysteine (G 5-cargo-C) for oxidation with C*. Each G 5-cargo-C protein was expressed and purified then activated with Ellman's reagent to form G 5-cargo-C-Ellman's. In order to obtain the desired N-terminus, XALPSTGG peptides, where X represents and DW, LV, DV, 'A, D LW, were sortagged onto G 5-cargo-C-Ellman's using optimized conditions, giving a product yield between 90-95%. Finally, LFN-C* and X-G 5-cargo-C-Ellman's were combined to form the hindered disulfide then purified to give the desired conjugates, 1-X, 2-X, or 3-X. The hindered disulfide comprised of cysteine and penicillamine was found to be 2-fold more stable to reduction than a cysteine-cysteine disulfide (Figure 5.6.8). As a first measure of X-DTAmut 's protein stability, we used the protein synthesis inhibition assay with 1-X conjugates. CHO-Ki cells were treated with 10-fold dilutions of 1-X conjugates for 6 hours in the presence of 20 nM PA for 6 hours at 37 'C and 5% CO 2. After 6 hours, the cells were washed three times with PBS then treated with leucine-free F-12K medium containing 3H-leucine (1 pCi mL-, PerkinElmer) for 1 hour at 37 'C and 5% CO 2. The cells were washed three times then resuspended in scintillation fluid and 3H radioactivity was counted. For each sample, the scintillation counts were normalized to a PA only control. Both the protein stability of X-DTAmut and X-DARPin was determined using western blot analysis. CHO-KI cells were treated with 100 nM 1-X or 2-X (where X represents 'A, DA, LV, DV, LW, or DW) in the presence of 20 nM PA for 6 hours at 37 'C and 5% CO 2. After 6 hours, the medium was removed and 0.25% trypsin-EDTA was added to each well for 5 minutes at 37 'C and 5% CO 2 . The cells were washed twice with PBS at 500 g for 2 minutes at room temperature. The cytosolic fraction was extracted using the digitonin lysis conditions and 232 analyzed by western blot as previously described. The membrane was stained with LF, DTA or streptavidin, Erkl/2, and Rab5 antibodies then stained with the appropriate secondary antibodies prior to imaging. 5.4.11. EGFR-Targeted Translocation of RRSPC 2 The bioactivity of stabilized RRSPC 2 was analyzed after translocation through EGFR into AsPC-I pancreatic cancer cells (ATCC). AsPC-1 cells were treated with 100 nM 3-X (where X represents LF or DF) for 9 hours in the presence of 10 nM mPA-EGF in serum-free RPMI medium at 37 'C and 5% CO2 . Additionally, samples were treated with 100-fold excess EGF (1 ptM; LifeTechnologies) to outcompete mPA-EGF binding or 10 nM PA[F427H]-EGF to block translocation. The following samples were used as controls: 100 nM 3DF only, 100 nM LA-LFNRRSPC 2 (N-terminal Ala), 100 nM LF-LFN-RRSPC 2 (N-terminal Phe), 100 nM wild-type LFN, and 10 nM mPA-EGF only. After treatment, the medium was removed and 0.25% trypsin-EDTA was added to each well for 5 minutes at 37 'C and 5% CO 2 . The cells were washed twice with PBS at 500 g for 2 minutes at room temperature. The cytosolic fraction was extracted using the digitonin lysis conditions and analyzed by western blot as previously described. The membrane was stained with streptavidin, anti-Erkl/2, and anti-Rab5. The total cell lysate was extracted using RIPA buffer (50 mM Tris HCl pH 8.0, 150 mM sodium chloride, 0.1% SDS, 0.5% sodium deoxycholate, 1% NP-40) plus protease inhibitor cocktail and PhosStop phosphatase inhibitor (Roche) for 30 minutes on ice then spun at 4'C for 10 minutes. The total cell lysate was analyzed by western blot, which was immunostained using pErkl/2, KRas, and Erkl/2 antibodies then stained with the appropriate secondary antibodies prior to imaging. The bands were quantified using LI-COR Image Studio software and normalized to Erkl/2. The PA only sample was set to 1. 233 5.5. Acknowledgements This research was generously supported by MIT start-up funds, the MIT Reed Fund, NSF CAREER Award (CHE-1351807), and a Damon Runyon Cancer Research Foundation award for B.L.P, and a National Science Foundation Graduate Research Fellowship for A.E.R. We would like to thank Prof. R. John Collier (Harvard) for his continued support. We thank the NERCE facility (grant: U54 A1057159) for expressing some toxin proteins and Prof. Douglas Lauffenburger (MIT) for the use of the Odyssey Infrared Imaging System. We also thank Prof. Alexander Varshavsky (CalTech), Dr. Jang-Hyun Oh (CalTech), Dr. Xiaoli Liao (MIT), Dr. Faycal Touti (MIT), and Mr. Ethan Evans (MIT) for helpful conversations. 234 5.6. Appendix b. a. 1.6- (n - U) -C 1.4- 1.2- LFN-DTA., LD-LFN-DTA -A -D-LF F N-DT A.. .2 LF-LFN-DTA C0 1.2- c C 0 1.0- -4- DF-LFN-DTA"'" -+-LL-LFN Uk. -4-L-LFN 0 C: >1 - 0.6 0.4- LF N-DT -- LR-LFN-DTA A DR -LFN-DTA 0.8- u-DT 4. 0.6- -!-LLFN-DTA 0 - - DM-LF -DTA -*-'P-LF N-DTAmt -P -LFN t 0.8- 1.0- -- T-LF N-DTA d ' V-LF NJ-TA. DV-LFN-DTA - CC- 0 0.4. 0.2. 0.20.0. - 0.0 -15 -14 -13 -12 -11 -10 -9 Log[Concentration(M)] -15 -6 -7 -8 -14 -13 -12 -11 -10 -9 Log[Concentration(M)] -8 -6 -7 d. C. 1.4- -- LF-DTA,, t S -A- 1.6- U) 1.2- -C 1.0- -V- LW-LF N-DTk -4- DW-LF N-DT A., 1.4LFN-DTA - -H-LFTN At - , 1.2 ADH-LFN-DTA C _ C LE-LFN-DTA 4- DE-LFN-DTA - --0 - 9C 0.6-. LI-LF N-D 0 T C 0L , .0 ~-0.8 D4j-LFN-DTA O LY-LFN-DTA 0.8- - 0 - Y-LFN-DTA,, LA-LFN -0-DA-LF 0.6- N-DTA,, LN-LFN -- DTA,. N-LFN-DT A , U) -LFN-DTA, 0.4- 0.4- I 0.2- 0.2- 0.0- - 0.0 -15 -14 -13 -6 -7 -8 -12 -11 -10 -9 Log[Concentration(M)] -15 -14 -13 -12 -11 -10 -9 Log[Concentration(M)] -8 -7 -6 e. , 1.21.0U) - C- 0.8 0 0.6 Ca U) 0.4- --- LF N-DTA ,, -- KLFN-DTA. -A-K-LF N-DTA , - 0C 0.2- 0.0 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 Log[Concentration(M)] Figure 5.6.1. Translocation of X-LFN-DTAmut constructs were analyzed by the protein synthesis inhibition assay in CHO-KI cells treated for 6 hours in the presence of 20 nM PA. 235 Table 5.6.1. EC50 values for X-LFN-DTAmut constructs translocated in CHO-Ki cells from Figure 5.6.1. X-LFN-DTA, X represents: Dy DW 0.077 0.078 DM DF DH Di Dp Lp DK D DL DR DE DT DV LV DA 0.084 0.085 0.089 0.092 0.094 0.094 0.099 0.10 0.11 0.11 0.11 0.11 0.12 LM 0.13 i 0.13 0.13 DD DN LT DS LA LL LS LD LN LR LE LH LF K Ly LW 236 EC5 0 (nM) 0.074 0.15 0.16 0.17 0.21 0.21 0.56 0.96 1.6 1.6 1.9 2.4 2.5 3.7 3.8 5.5 DV + + Bafilomycin Al anti-LF DA LW DW + + alMZ, 4 anti-Erk/2 LA - LV + DW - LW + DA + LA + + + + DV + + LV + + + LD, LD, + 1 ng X-LFN-DTAmt PA w OW*O* two* anfi-Rab5 1 ng LA DA LW + + + + + 4*C 400 400 4*C 4*C DW LV DV LA + + + DA LW DW + DV + + PA[F427H] LV + PA LDt LDmt + X-LFN-DTA, 4*C anti-LF 10OW IPPW fpow & - APIWO anti-Erk1/2 anti-Rab5 + Figure 5.6.2. Western blot analysis of X-LFN-DTAmut constructs in CHO-Ki cells in the presence of 200 nM Bafilomycin A1, 20 nM PA[F427H], or incubated at 4'C over 6 hours. 237 a. HEK-293T 20 [tM lactacystin 1 ng X-LFN-DTAt LD, LV DV LA DA LW DW LV DV DA LA 1ng X-LFN-DTAt LD, DW = It = b. HeLa LW 20 [M lactacystin LV DV LA DA LW DW LV DV LA DA LW DW I Figure 5.6.3. Western blot analysis of X-LFN-DTAmUt constructs delivered into HEK-293T (a) or HeLa (b) cells in the presence of 20 nM PA over 6 hours. 238 NH, X-ALPSTG 1 ng X-LF N-DTAmUt LDmut LD LA DA X/\\17C INJ 5 -- LW DW LA DA LW DW anti-LF anti-Erki /2 tsa emp Mne4mi M emstee tnt aitwo anti-Rab5 Figure 5.6.4. Sortagged (left) and native (right) X-LFN-DTAmut constructs (where X represents LA, DA, LW, and DW) were translocated in CHO-Ki cells in the presence of 20 nM PA for 6 hours then analyzed by western blot. 239 t =O0min X-Kbio-LFN-DTAmat Streptavidin anti-p-actin LV LW DW LR .% t =10 min DR LV LW DW LR t = 60 min DR LV LW DW DR two 6".. 6" 6 No W b" 0. *0% y~~ Figure 5.6.5. In vitro degradation of X-K(bio)-LFN-DTAmut, where X represents and DR, in rabbit reticulocyte lysate over time. 240 LR WooM44 LV, LW, DW, LR, 20 [M lactacystin 1 ng PA LFNX-Kbio-LFN-DTAmut DTAmUt only LV LW DW LR DR LV LW DW LR DR anti-LF Streptavidin anti-Erk1/2 anti-Rab5 Figure 5.6.6. LV-, LDW-, and LDR-K(bio)-LFN-DTAmut were translocated into CHO-Ki cells in the presence of 20 nM PA for 6 hours then analyzed by western blot. As a proteasomal inhibitor, 20 [M lactacystin was used. 241 G,,LRLA- TANH2 HS-T- LPSTGG I- H - 0 LPSTG LRLA-NyNH oxidation HSI - 0 LPSTG LRLA- iNANH2 2 S G5 os c,CO 5 ) X-ALPSTGG G X-ALPSTG, X-ALPSTG, -- OH LFN-C* X-cargo-C Figure 5.6.7. LFN-C*_X-cargo-C is the oxidation product of LFN-C* and X-G -cargo-C5 Ellman's. 242 a. LF N-C G 5-DTAmu-$ Time (min) 0 5 LF N-C* G5 -DTAmut-C 15 30 60 120 240 0 5 15 30 60 120 240 420 ft"O * %-wow -FN-C(*) G5-DTAmut-C (53 kDa) - ,~~~. 4"**-4 _FN-C(*) (32 kDa) 35 -DTA-C (21 kDa) b. 0.50.0- -C * LFN-C G -DTA * LFN-C* G5 -DTAmut- -0.5- <.E -1.0- ) I-- U. - S-2.0- LL C: -2.5-3.0-3.5-20 0 20 40 60 80 100 120 140 160 180 200 220 240 260 Time (minutes) Figure 5.6.8. Rate of reduction for 1-X. a. Coomassie-stained SDS Page gel of 1 JAM 1-X (disulfide) or 1-X (hindered disulfide) incubated with 1 mM DTT in 20 mM Tris pH 7.5 and 150 mM NaCl incubated at 37 'C. b. The first order reaction rates were determined by plotting ln([1X]) over time. The rate of reduction (kb,) for 1-X (disulfide) was 0.0225 min-' (R2 = 0.972) and 1-X (hindered disulfide) was 0.0116 min-1 (R2 = 0.995). 243 a. - SGLRLA-N",.NH LA -LPST 5 NH s X-ALPST G5 N f 4OH H 0 1_x b. 1.6- 1.4A) Uf) 1.2- UI) 1.0- 0.8- LFNDTAmu 1-G5 -0 CT3 a> 0.6- A ILV -y 1 -"V -4-- LA .. q- 0.40.2- ---- 0.0-15 -1 -14 1 -DA 1 LW -DW -13 I I -12 -11 -10 -9 -8 -7 -6 Log[Concentraion(M)] Figure 5.6.9. Delivery of 1-X. a. Molecular composition of 1-X. b. CHO-Ki cells were treated with ten-fold dilutions of 1-X in the presence of 20 nM PA then analyzed by protein synthesis inhibition assay. The results indicated that X-DTAmut was stabilized by one N-terminal D-amino acid after cleavage from LFN- 244 - 1.0- LF -DTA - LFN-DTA-R RSP C2 C/) 0.8H Cl) C 0.6+-P 0 0.40 L- 0.2- LL 0.0 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 Log (Concentration[M]) Figure 5.6.10. Translocation of LFN-DTA-RRSPc2 was analyzed by the protein synthesis inhibition assay in CHO-KI cells treated for 30 minutes in the presence of 20 nM PA. 245 t=0 LFN- 3- LF WT RRSPLFN DF t=10min LFN- - WT RRSP LFN LF DF LFN LF WT RRSPLFN DF kDa 98 62 3-X (79 kDa) LFN-RRSPC 2(77 kDa) 49 38 WT LFN (30 kDa) 28 KRas (20 kDa) KRas, (17 kDa) 14 6 3 KRasN (3.5 kDa) Figure 5.6.11. In vitro cleavage of 10 pM KRas by 0.4 pM RRSPc 2 conjugates in 20 mM Tris pH 7.5 and 150 mM NaCl at 37 *C. Time points were taken and quenched with 2x non-reducing loading dye after 0 and 10 minutes then analyzed by coomassie-stained SDS-PAGE gel. 246 a. la la 20608.5 Da KRas only KRas on1v 18000 20000 22000 I: 2 lb 3a 1b 881.7 m/ 17103.4 Da 79594.8 Da 1175.3 3a 2 KRas + 3'F 16000 18000 20000 800 1000 1200 1400 2 C 881.7 m/z lb 75000 80000 85000 3b lb 17103.4 Da 79595.3 Da 1175.3 m/z 800 1000 1200 1400 3c 2 KRas + 3-DF 2 d. 881.7 mz 16000 1800 20000 lb 80000 7500 3c 85000 77464.7 Da lb 17103.5 Da 1175.3 3c 2 KRas + 'A-LFN-RRSPC2 L_ 800 1000 1200 14D0 16000 18000 20000 la e. 75000 3d 80000 85000 300631.6 Da 20608.8 Da 1a 3d KRas + LFN 1 2 4 5 6 Time 7 Imin] 8 9 10 11 18000 20000 22000 30000 3200 Figure 5.6.12. In vitro cleavage of 10 pM KRas by 0.4 pM RRSPC 2 conjugates in 20 mM Tris pH 7.5 and 150 mM NaCl at 37 'C after 10 minutes. Time points were taken (from the same reactions as in Figure 5.6.11) and quenched with 1:1 water:acetonitrile with 0.1% TFA after 10 minutes then analyzed by LC-MS. Peak la is uncleaved KRas (calc 20608.2 Da), peak lb is Cterminal portion of cleaved KRas (calc 17103.1 Da), peak 2 is N-terminal portion of cleaved KRas (calc 3523.1 Da), peak 3a is LFNC* LF-RRSPC 2 C (calc 79582.5 Da), peak 3b is LFNC* _DF-RRSPC 2C (calc 79582.5 Da), peak 3c is LFN-RRSPc2 (calc 77459.7 Da), and peak 3d is wild-type LFN (calc 30631.2 Da). 247 -- 1.8 - LF N-DTA + PA SLFN-DTA + PA-EGF ~LFN-DTA + PA +1 pM EGF - LFN-DTA + PA-EGF + 1 sM EGF .In 1.6 1.4C 0.I " 0.8C .2) 0.60. 0.2 -15 -14 -13 -12 -11 -10 -9 -8 Log[Concentration(M)] -7 -6 Figure 5.6.13. Targeted delivery of LFN-DTA through EGFR was analyzed by the protein synthesis inhibition assay in AsPC-1 treated for 30 minutes in the presence of 10 nM PA or PAEGF. For competition controls, 1 pM EGF was added to 100 nM LFN-DTA plus 10 nM PA or PA-EGF. 248 LA - LF LA - - - 3-X - - DF - - - LF DF LF DF LF DF LFN - - - mPA-EGF + - - + + + + + + + - - mPA[F427H]-EGF - - - - - - - - - - + - - - - + X-LFN-RRSPC2 - - + 1 [M EGF anti-pErkl/2 S1.2. a) 1 .0 30.8 c" .2~0.6- - -J. 4 set to 1. 249 5.6.1. LC-MS Traces NH2 * obs.52052.4 0.4 ca. 52051.2 LA \\17C 500.0 52500 1 2 3 4 5 6 7 Time [min] 8 10 9 11 LA-LFN-DTA NCL obs.52053.2 Ca. 52051.2 DA /\^17C 50000 1 2 3 4 5 1 3 4 5 DA-LFN-DTA 6 7 6 7 Time [min] 8 9 0.4 52500 10 11 NCL NHW obs.52168.5 * 0.4 ca. 52166.3 LW /\/\17C 3 50000 52500 1 2 3 5 4 6 Time (min] 7 8 10 9 11 LW-LFN-DTA NCL NH < obs.52168.8* ca. 52166.3 DW^^YX1 7C 50000 1 2 DW-LFN-DTA 250 3 NCL 4 5 6 Time [min] 7 8 9 52500 10 11 0.4 * obs.52829.9 LA-ALPSTG5 0.4 ca. 52830.0 52000 54000 2 1 5 4 3 6 Time [mini 7 8 9 10 11 LA-LFN-DTA * obs.52830.1 0.4 cCa. 52830.0 A-ALPSTG5 52000 54000 2 1 3 4 5 6 Time [min] 7 8 9 10 11 DA-LFN-DTA obs.52874.6 ca. 52874.0 LD-ALPSTG 0.4 52000 5400 1 2 3 4 5 6 Time [min] 7 8 9 10 11 LD-LFN-DTA obs.52874.2 * 0.4 ca. Dr" L/-ALPSTG 52874.0 52000 54"0 1 2 3 4 5 6 Time [min] 7 8 9 10 11 DD-LFN-DTA 251 obs.52888.1 ca. 52888.0 LE-ALPSTG 5 0.4 52000 54000 I 1 2 3 --~ T 4 5 8 7 8 ime [min] 9 10 11 LE-LFN-DTA DE-ALPSTG obs.52888.1 * G ca. 52888.0 52000 1 2 3 4 5 6 8 7 lime [min] 9 0.4 54000 10 11 DE-LFN-DTA obs.52906.8 0.4 ca. 52906.1 LFALPSTG 52000 54000 1 2 3 4 5 6 ime [min] 8 i 9 0 11 LF-LFN-DTA obs.52906.6 ca. 52906.1 DFALPSTG 5 52000 54000 1 2 3 4 5 6 ime [mini DF-LFN-DTA 252 7 8 9 10 11 0.4 obs.52896.0 *t 0.4 ca. 52896.1 LH-ALPSTG 52000 8 8 7 6 Time [min] 5 4 3 2 1 9I 54000 1 11 1 10 LH-LFN-DTA DH-ALPSTG 5 obs.52895.9 ca. 52896.1 .- 52000 6 Time [mini 5 4 3 2 1 7 54000 11 10 9 8 0.4 DH-LFN-DTA obs.52872.3 0.4 ca. 52872.1 L 1 -ALPSTG 52000 54000 1 2 3 4 5 7 6 Time [mini 8 9 10 11 LI-LFN-DTA obs.52871.9 0.4 ca. 52872.1 DI -ALPSTG 52000 54000 1 1 I 2 I 3 I 4 I 5 7 6 Time [mini 8 9 10 11i DI-LFN-DTA 253 obs.52887.1 LK-ALPSTG 0.4 ca. 52887.1 .. 50000 55000 1 2 3 4 8 Time [min] 9 10 11 LK-LFN-DTA * obs.52887.2 DK-ALPSTG 5 52000 1 T8 2 0.4 ca. 52887.1 3 4 9 i ime [min) 54000 10 11 DK-LFN-DTA obs.52872.7 ca. 52872.1 LL -ALPSTG 0.4 52000 54000 1 2 3 4 5 6 Time [min] 7 8 9 10 11 LL-LFN-DTA obs.52872.8 0.4 ca. 52872.1 DL-ALPSTG I 1 2 DL-LFN-DTA 254 3 4 5 6 Time [min) 8 5200054000 9 10 11 0.4 obs.52890.1 ca. 528 90.4 I LM-ALPSTG, 52000 54W0 1 5 4 3 2 6 ime [min] 8 7 9 10 11 LM-LFN-DTA * obs.52890.1 ca. 52890.3 DM-ALPSTG. 52000 2 4 3 5 8 Time [min] 9 0.4 54000 10 11 DM-LFN-DTA LN-ALPSTG 5 ca. 52873.0 Imam 54000 _]i200 2 1 3 0.4 obs.52873.4 4 5 6 Time [min] 7 8 9 10 11 LN-LFN-DTA * obs.52873.3 0.4 52873.0 D N-ALPSTG 5ca. 52000 54"0 1 2 3 6 lime [mini 7 i 9 10 11 DN-LFN-DTA 255 obs.52856.4 ca. 52856.0 LP-ALPSTG 5 0.4 52000 54000 1 2 3 4 5 6 Time [minj 8 7 9 10 11 LP-LFN-DTA obs.52856.3 ca. 52856.0 DRALPSTG 0.4 52000 54000 ---- 1 2 v 4 3 5 T Time [min) 7 8 9 10 11 DP-LFN-DTA obs.52915.7 ca. 52915.1 0.4 . LR-ALPSTG 5 52000 54000 3 4 5 6 Time [mini 7 8 9 10 11 LR-LFN-DTA 5 obs.52915.6 * 0.4 Ca. 52915.1 - DR-ALPSTG 52000 54000 1 2 DR-LFN-DTA 256 3 4 5 6 Time [min 7 8 9 10 11 * obs.52845.5 0.4 Ca. 52844.9 LS-ALPSTG5 52000 I I I 1 I 1 2 3 4 5 I i 6 7 Time [min] 8 9 40 10 11 LS-LFN-DTA obs.52845.0 D *0.4 ca. 52844.9 S-ALPSTG 52000 54000 3 7 lm2 Time [min] 4e 8 9 1 1 DS-LFN-DTA obs.52860.0 0.4 ca. 52860.0 LT-ALPSTG5 52000 54000 1 2 3 4 5 6 7 8 9 10 11 Time [min] LT-LFN-DTA obs.52860.1 0.4 DT-ALPSTG ca. 52860.0 5 52000 54000 7 Time [min] 8 9 10 11 DT-LFN-DTA 257 obs.52858.3 ca. 52858.0 LV-ALPSTG 0.4 52000 54000 1 2 3 4 5 6 7 8 9 10 11 Time [mini LV-LFN-DTA obs.52858.2 ca. 52858.0 DV-ALPSTG 0.4 52000 54000 1 2 3 4 5 6 Time [mini 7 8 9 10 11 DV-LFN-DTA obs.52944.7 ca. 52945.1 LW -ALPSTG 5 0.4 152000 64000 1 2 3 4 5 6 Time [min] 7 8 9 10 11 LW-LFN-DTA obs.52944.8 ca. 52945.1 * DW-ALPSTG 52000 54000 1 2 3 4 6i Time DW-LFN-DTA 258 [min[ 9 10 11 0.4 * obs.52922.2 0.4 LY-ALPSTG . ca. 52922.1 52000 54000 1 2 3 4 5 6 Time [mini 7 8 9 10 11 LY-LFN-DTA obs.52922.3 ca. 52922.1 DYALPSTG 0.4 52000 54000 51 5 6 Time [mini 7 8 9 10 11 DY-LFN-DTA obs.53500.9 * 0.4 a. 53499.0 biotin LV -AG KLPSTG 50000 58000 4 6 Time [min] 7 8 9 10 11 LV-K(bio)-LFN-DTA obs.53587.6 ca. 53584.6 biotin LWAGKPSTG .. 50000 2 3 4 5 6 Time [mini 7 8 9 0.4 500 10 11 LW- K(bio)-LFN-DTA 259 * obs.53587.6 biotin 0.4 DWAGKLPSTG ca. 53584.6 50000 55000 1 2 3 4 5 1 2 3 4 5 6 6 7 Time [min] 8 9 10 11 DW-K(bio)-LFN-DTA biotin obs.53558.3 LR -AGKLPSTG 0.4 ca. 53554.6 5 50000 55000 1 2 3 4 5 6 7 8 Time [mini 9 10 11 LR-K(bio)-LFN-DTA biotin obs.53557.6 ca. 53554.6 DR-AG5LPsTG 0.4 50000 55000 1 2 3 4 5 6 Time [mini 7 8 9 10 11 D RK(bio)LFDTA LPSTGLRLA-Nfl I. NH, obs.53089.1 ca. 53087.5 XOH G GWO 52500 55000 1 2 3 LFN-C*_G 5 -DTA-C 260 4 5 6 Time [mini 7 8 9 10 11 0.4 LPSTGLRA NEW obs.53629.1 53628.2 'ca. 0.4 .NX .OH A-ALPSTG 52500 55000 3 2 6 5 4 Time [mini 9 8 7 11 10 LFN- *_LA-DTA-C M LPSTGLRLA-N NH, DA-ALPSTG, obs.53629.3 ca. 53628.2 0.4 OH 52500 55000 6 5 4 3 2 1 Time [mini 7 8 10 9 11 LFNC* DA-DTA-C 0 NH LPSTG,LRL-A-N obs.53657.1 0.4 s ca. 1 'V-ALPSTG, 53656.2 OH w 52500 55DOO - - -L 2 1 6 5 4 3 Time [mini 7 I8 9 10 11 LFN-C _LV-DTA-C NV-ALPSTG 1 obs.53656.4 H MLSGRA-sJ .OH ca. H *0.4 53656.2 52500 55000 1 2 3 5 6 7 8 i 1o i1 Time [mini LFN-C*_DV-DTA-C 261 0 OEN=LPSTGLRLAgNH OH .N W-ALPSTG, obs.53744.6 0.4 ca. 55743.3 52500 55000 1 2 4 3 5 6 Time [min] 7 8 I', LFN-C* LW-DTA-C 0 - LPSTGLLA- N NH obs.53744.5 ca. 55743.3 sA 0.4 lW-ALPSTGWN 52500 55000 ' 2 '3 ' ' ' 4 5 6 7 ' I 1 ime [min] 8 9 10 11 LFN-C* DW-DTA-C -LPSTGLRLA-N obs.46637.4 * 0.4 ca. 46634.2 biotin A-AGLPSTG os 45000 47500 1 2 3 4 5 6 Time [mini 7 8 9 10 11 LFN-C* L A-DARPin-C 4LPSTGLRLA64 6.4 4 ca. biotin 46634.2 osf -A-AG,KLPSTG 46000 47500 1 LFNC* 262 2 3 4 DA-DARPin-C 5 6 Time [min] 7 89 10 11 LVAG 5 obs.46665.4 KLILPSTGLRLA- os, - LV-AG,KLPSTG 0.4 ca. 46662.3 4500047500 ~~1 ~~ 3 2 5 4 6 Time [min] 10 9 8 7 11 LFN-C* LV-DARPin-C LPSTGLALA- obs.46664.7 NH ca. bioinT 0.4 46662.3 X -V-AGKLPSTG, 4.7500 2 1 3 7 6 5 4 Time [min] 8 9 10 11 LFNC _ DV-DARPin-C -LPSTGLRLAbiotin LW-AGKLPSTG,G obs.46751.9 ca. 46749.4 0.4 ,&- 5 45000 47500 2 1 3 4 5 6 Time [mini 7 8 9 10 11 LFN-C* LW-DARPin-C MLPSTGLRLA obs.46751.5 <O, 0.4 ca. 46749.4 biotin OW-AG KLPSTGo, 45000 47500 1 1 2 3 4 5 2 3 4 5 6 6 Time [min] 7 8 9 7 8 9 10 11 LFNC* DW-DARPin-C 263 obs.77456.1 LPSTG, RRSP ca. 77450.2 75000 1 2 3 4 5 0.4 6 Time [min[ 8 7 9 80000 10 11 LA-LFN-RRSPC2 obs.77534.7 0.4 'FU LPSTG, RRSP ca. 77526.3 75000 80000 1 2 3 4 5 6 7 Time [mini 8 9 10 11 F-LFN-RRSPC2 LPSTGLRLA obs.79591.2 ca. 79582.5 biotin LF-AGKPSTG 1 LFN-C* 2 3 RRSPJ 4 5 os L_ 0 6 7 Time [min] 75000 8 9 80000 10 11 F-RRSPc 2 -C obs.79590.4 T ca. 79582. 5 biinITDF-AGKLPSTGjR sf pOH 75000 ....... oj Time [min] LFN-C* DF-RRSPC2-C 264 0.4 8 9 80000 10 11 0.4 5.7. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Tasaki, T. et al. A family of mammalian E3 ubiquitin ligases that contain the UBR box motif and recognize N-degrons. Molecular and CellularBiology 25, 7120-7136 (2005). Xia, Z. et al. Substrate-binding sites of UBRI, the ubiquitin ligase of the N-end rule pathway. Journalof Biological Chemistry 283, 24011-24028 (2008). Sriram, S. et al. Development and Characterization of Monomeric N-End Rule Inhibitors through In Vitro Model Substrates. Journal of Medicinal Chemistry 56, 2540-2546 (2013). Bachmair, A., Finley, D. & Varshavsky, A. In vivo Half-Life of a Protein Is a Function of Its Amino-Terminal Residue. Science 234, 179-186 (1986). Gonda, D.K. et al. Universality and structure of the N-end rule. JBiol Chem 264, 1670012 (1989). Varshavsky, A. The N-end rule pathway and regulation by proteolysis. Protein Science 20, 1298-1345 (2011). Falnes, P.O. & Olsnes, S. Modulation of the intracellular stability and toxicity of diphtheria toxin through degradation by the N-end rule pathway. Embo Journal 17, 615625 (1998). Gupta, P.K., Moayeri, M., Crown, D., Fattah, R.J. & Leppla, S.H. Role of N-Terminal Amino Acids in the Potency of Anthrax Lethal Factor. Plos One 3 (2008). Bachran, C. et al. Reductive Methylation and Mutation of an Anthrax Toxin Fusion Protein Modulates its Stability and Cytotoxicity. Scientific Reports 4 (2014). Young, J.A.T. & Collier, R.J. Anthrax toxin: Receptor binding, internalization, pore formation, and translocation. Annu. Rev. Biochem. 76, 243-265 (2007). Feld, G.K. et al. Structural basis for the unfolding of anthrax lethal factor by protective antigen oligomers. Nature Structural& MolecularBiology 17, 1383-U245 (2010). Krantz, B.A., Finkelstein, A. & Collier, R.J. Protein translocation through the anthrax toxin transmembrane pore is driven by a proton gradient. J. Mol. Biol. 355, 968-979 (2006). Scobie, H.M., Rainey, G.J.A., Bradley, K.A. & Young, J.A.T. Human capillary morphogenesis protein 2 functions as an anthrax toxin receptor. Proc. Natl. Acad Sci. U. S. A. 100, 5170-5174 (2003). Nassi, S., Collier, R.J. & Finkelstein, A. PA(63) channel of anthrax toxin: An extended beta-barrel. Biochemistry 41, 1445-1450 (2002). Milne, J.C., Furlong, D., Hanna, P.C., Wall, J.S. & Collier, R.J. Anthrax protective antigen forms oligomers during intoxication of mammalian-cells. J. Biol. Chem. 269, 20607-20612 (1994). Klimpel, K.R., Molloy, S.S., Thomas, G. & Leppla, S.H. Anthrax toxin protective antigen is activated by a cell-surface protease with the sequence specificity and catalytic properties of furin. Proc. Natl. Acad Sci. U. S. A. 89, 10277-10281 (1992). Krantz, B.A., Trivedi, A.D., Cunningham, K., Christensen, K.A. & Collier, R.J. Acidinduced unfolding of the amino-terminal domains of the lethal and edema factors of anthrax toxin. J. Mol. Biol. 344, 739-756 (2004). Miller, C.J., Elliott, J.L. & Collier, R.J. Anthrax protective antigen: Prepore-to-pore conversion. Biochemistry 38, 10432-10441 (1999). 265 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Jiang, J., Pentelute, B.L., Collier, R.J. & Zhou, Z.H. Atomic structure of anthrax protective antigen pore elucidates toxin translocation. Nature 521, 545 - 549 (2015). Arora, N. & Leppla, S.H. Residues 1-254 of anthrax toxin lethal factor are sufficient to cause cellular uptake of fused polypeptides. J Biol. Chem. 268, 3334-3341 (1993). Liao, X., Rabideau, A.E. & Pentelute, B.L. Delivery of Antibody Mimics into Mammalian Cells via Anthrax Toxin Protective Antigen. Chembiochem 15, 2458-2466 (2014). Rabideau, A.E., Liao, X. & Pentelute, B.L. Delivery of mirror image polypeptides into cells. Chemical Science 6, 648-653 (2015). Milne, J.C., Blanke, S.R., Hanna, P.C. & Collier, R.J. Protective antigen-binding domain of anthrax lethal factor mediates translocation of a heterologous protein fused to its amino- or carboxy-terminus. Mol. Microbiol. 15, 661-666 (1995). Chen, I., Dorr, B.M. & Liu, D.R. A general strategy for the evolution of bond-forming enzymes using yeast display. Proc. Natl. Acad. Sci. U. S. A. 108, 11399-11404 (2011). Collier, R.J. & Kandel, J. Structure and activity of diphtheria toxin. 1. Thiol-dependent dissociation of fraction of toxin into enzymically active and inactive fragments. J. Biol. Chem. 246, 1496 (1971). Wilson, B.A. & Collier, R.J. Diphtheria-toxin and Pseudomonas aeruginosa exotoxin A active-site structure and enzymatic mechanism. Curr. Top. Microbiol. Immunol. 175, 2741 (1992). Parizek, P. et al. Designed Ankyrin Repeat Proteins (DARPins) as Novel IsoformSpecific Intracellular Inhibitors of c-Jun N-Terminal Kinases. ACS Chem. Biol. 7, 13561366 (2012). Tamaskovic, R., Simon, M., Stefan, N., Schwill, M. & Pluckthun, A. in Methods Enzymol. (eds. Wittrup, K.D. & Verdine, G.L.) 101-134 (Elsevier Academic Press Inc, San Diego, 2012). Antic, I., Biancucci, M. & Satchell, K.J.F. Cytotoxicity of the Vibrio vulnificus MARTX toxin Effector DUF5 is linked to the C2A Subdomain. Proteins-StructureFunction and - 19. viuinfurmatics 0"(., Z-U'+.6-Z.)06U(2'4). 30. 31. 32. 33. 34. 35. 266 Antic, I., Biancucci, M., Zhu, Y., Gius, D.R. & Satchell, K.J.F. Site-specific processing of Ras and Rap 1 Switch I by a MARTX toxin effector domain. Nature communications 6, 7396-7396 (2015). Kellogg, B.A. et al. Disulfide-Linked Antibody-Maytansinoid Conjugates: Optimization of In Vivo Activity by Varying the Steric Hindrance at Carbon Atoms Adjacent to the Disulfide Linkage. Bioconjugate Chemistry 22, 717-727 (2011). Popp, M.W., Antos, J.M., Grotenbreg, G.M., Spooner, E. & Ploegh, H.L. Sortagging: a versatile method for protein labeling. Nat. Chem. Biol. 3, 707-708 (2007). Wilson, B.A., Reich, K.A., Weinstein, B.R. & Collier, R.J. Active-site mutations of diphtheria toxin: effects of replacing glutamic acid-148 with aspartic acid, glutamine, or serine. Biochemistry 29, 8643-51 (1990). Sun, J., Lang, A.E., Aktories, K. & Collier, R.J. Phenylalanine-427 of anthrax protective antigen functions in both pore formation and protein translocation. Proc. Natl. Acad Sci. U. S. A. 105, 4346-4351 (2008). Dawson, P.E., Muir, T.W., Clark-Lewis, 1. & Kent, S.B. Synthesis of proteins by native chemical ligation. Science 266, 776-9 (1994). 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Pentelute, B.L., Barker, A.P., Janowiak, B.E., Kent, S.B.H. & Collier, R.J. A Semisynthesis Platform for Investigating Structure-Function Relationships in the NTerminal Domain of the Anthrax Lethal Factor. A cs Chemical Biology 5, 359-364 (2010). Bachmair, A. & Varshavsky, A. The degradation signal in a short-lived protein. Cell 56, 1019-32 (1989). Rodriguez-Viciana, P., Sabatier, C. & McCormick, F. Signaling specificity by Ras family GTPases is determined by the full spectrum of effectors they regulate. Molecular and Cellular Biology 24, 4943-4954 (2004). Cox, A.D., Fesik, S.W., Kimmelman, A.C., Luo, J. & Der, C.J. Drugging the undruggable RAS: Mission Possible? Nature Reviews Drug Discovery 13, 828-851 (2014). Mechaly, A., McCluskey, A.J. & Collier, R.J. Changing the Receptor Specificity of Anthrax Toxin. Mbio 3 (2012). McCluskey, A.J. & Collier, R.J. Receptor-directed chimeric toxins created by sortasemediated protein fusion. Mol. Cancer Ther. (2013). McCluskey, A.J., Olive, A.J., Starnbach, M.N. & Collier, R.J. Targeting HER2-positive cancer cells with receptor-redirected anthrax protective antigen. Mol. Oncol. 7, 440-51 (2013). Hershko, A. & Ciechanover, A. The ubiquitin system for protein degradation. Annu Rev Biochem 61, 761-807 (1992). Bartel, B., Wunning, I. & Varshavsky, A. The recognition component of the N-end rule pathway. EMBO J9, 3179-89 (1990). Choi, W.S. et al. Structural basis for the recognition of N-end rule substrates by the UBR box of ubiquitin ligases. Nature Structural& Molecular Biology 17, 1175-+ (2010). Matta-Camacho, E., Kozlov, G., Li, F.F. & Gehring, K. Structural basis of substrate recognition and specificity in the N-end rule pathway. Nature Structural & Molecular Biology 17, 1182-+ (2010). Pentelute, B.L., Sharma, 0. & Collier, R.J. Chemical dissection of protein translocation through the anthrax toxin pore. Angew. Chem. Int. Ed Engl. 50, 2294-2296 (2011). Bai, L., Sheeley, S. & Sweedler, J.V. Analysis of Endogenous D-Amino Acid-Containing Peptides in Metazoa. Bioanal. Rev. 1, 7-24 (2009). Ollivaux, C., Soyez, D. & Toullec, J.-Y. Biogenesis of D-amino acid containing peptides/proteins: where, when and how? Journalof Peptide Science 20, 595-612 (2014). Cascales, L. & Craik, D.J. Naturally occurring circular proteins: distribution, biosynthesis and evolution. Organic& Biomolecular Chemistry 8, 5035-5047 (2010). Simon, M.D. et al. Rapid flow-based peptide synthesis. ChemBioChem 15, 713-720 (2014). 267