Clostridium XII Engineering Improved Cellulosomes Michael E. Himmel National Renewable Energy Laboratory BioEnergy Science Center - 2012 Focus area 2: Biomass Deconstruction and Conversion Conversion of Biomass to Fuels Recalcitrance and multi-scale complexity • transport phenomena - tissue/cellular scale • microfibril/matrix interaction - switchgrass Meters cellular/macromolecular • cellulose morphology - molecular scale CLSM Stereo Nanometers TEM cell debris AFM milled biomass vascular bundle secondary cell wall xylem cellulose microfibrils SEM Acid (hot water) pretreatment • Delaminates cell walls/increases porosity after hot water pretreatment • Solubilizes hemicellulose • CAFI3 switchgrass samples (Purdue) before middle lamella plant cell debris cell wall cell wall 1 µm 1 µm cell lumen Innovation for Our Energy Future Alkaline pretreatment • Erodes wall surfaces • Solubilizes lignins • CAFI3 switchgrass after lime pretreatment samples (Texas A&M) before exposed cellulose microfibrils cell wall surface 5 µm Innovation for Our Energy Future enzymes free confined localized secretion complexed large complexed isolated free small complexed cell tethered matrix confined diffuse localized concentrated digestion T. reesei enzymes C. cellulolyticum bacteria α-Cel7a::15 nm Au digested scalloped surface 200 nm cell wall 200 nm cell wall Free vs. complexed enzymes Trichoderma reesei C. cellulolyticum Our strategy is information based Molecular Structure (experimental parameters) X-ray crystallography Structure diversity (genomics) Homology modeling Physical Biochemistry Numerical Models (experimental parameters) (subsets to entire system) Molecular dynamics QM/MM Multi-scale modeling Code development Force fields Supercomputers Protein purification Physical chemical analyses MS and spectro. analyses Special and HTP activity testing Mechanistic Model (kinetic and thermodynamic) F* Hydrolysis Hydrolysis {…} Hydrolysis + Processivity Surface binding Recognition Processivity Initial processivity/decrystallization by cellobiose Processivity RC We take a reductionist approach Example: T. reesei Cel7A Linker peptide •Define function and functionality •Spring action •Interactions with substrate and water Binding Domain •Adsorption, binding energy •Mobility on cellulose surface •Interaction with broken strands 800,000 atoms Cellulose Substrate •Define most likely form of cell wall cellulose •How does pretreatment change it? •Are other isomorphs better substrates? Catalytic Domain •Free energy of motion of cellodextrin in tunnel •Exiting of cellobiose •QM/MM of reaction and structural changes Example: Improving Cel7A through enhanced understanding • Our approach to enhanced cellulose conversion: use experiments and modeling as complementary tools Carbohydratebinding module Linker Catalytic Domain Cellulose 1 0 CBM1 translates along cellulose, pausing every 1 nm Four residues form strategic hydrogen bonds: Y5, Q7, N29, Y32 1 nm 1 nm Homology at these sites is conserved across many cellulases and species: Beckham et al., JPCB 2010 The C. thermocellum Cellulosome 1 primary scaffoldin 4 anchoring scaffoldins 91 enzymes Fontes et. al. (2010) Illustration of Enzymatic Mechanisms Bryon Donohoe & Mike Resch, NREL What advantage from highly articulated GHs ? Mike Crowley, NREL C. thermocellum cellulosome GH GH GH GH 1 2 3 4 5 GH GH GH GH 6 7 8 9 CBM3 1 primary scaffoldin 4 different anchoring scaffoldins 72 various dockerins GH proteins with GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 SdbA CBM3 At least 92 potential places for cell-wall-bound enzymes? GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 Orf2p CBM3 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 GH GH GH GH GH GH GH GH GH 1 2 3 4 5 6 7 8 9 CBM3 CBM3 CBM3 CBM3 CipA 1 OlpC (Cthe_0452) cell 2 3 OlpB 4 GH CBM3 CBM3 CBM3 5 6 7 Cthe_0736 OlpA ?? Cthe_0735 ? Purified cellulosome performance • Cellulosomes perform better on the substrates they were grown on. • Cellulosomes grown on cellobiose perform poorly on Avicel and PTSG. Understanding cellulosomes: the critical enzymes GH48 and GH9 (CbhA) Family 48 cellulases are essential components of CBP organisms • Family 48 cellulases are essential components in several biomass-degrading bacteria. • Deletion of CelS reduces the activity of C. thermocellum by more than 40%. • Product inhibition is a major problem. • Understanding and improving these cellulases will lead to better microbes. Four new structures of GH family 48 from NREL • We have solved the structures of C. bescii, B. pumillus, H. chejuensis and T. fusca GH48 enzymes in addition to the two already known unique structures • B. pumillus GH48 stands out from the others enzymes due to its enlarged loops near the active site tunnel *in collaboration with D. Wilson C. bescii CelA GH48 B. pumilus GH48 H. Chejuensis GH48 T. fusca GH48 Comparison of family 48 cellulases CelA, CelS, CelF (Blue) Cel48(Red) Cel48 Tm ~ 45°C CelF Tm ~ 55°C CelS Tm ~ 65°C CelA Tm ~ 85°C Computational scheme to characterize product expulsion Initial Final Chen and Brady, Cornell University Reaction coordinate Understanding T. thermocellum CbhA Dockerin PDB CBM4 NREL 2009 Ig-GH9 PDB Vlad Lunin & Markus Alahuta, NREL Fn31- Fn32 NREL 2009 CBM3b Bayer et al 2009 Clostridium thermocellum CbhA X-ray structures SP • CBM 4 IG GH 9 FN 3 CBM3 b DO C The structures of three new modules of CbhA have been solved A family 4 carbohydrate binding module (CBM4) and two fibronectin(III)-like modules. • CBM4 binds to cellobiose, where the aromatic side chains of tyrosine 110 and tryptophan 68 constitute the main interactions with one glucose unit of cellobiose. • Tryptophan 118 is a unique feature of CbhA CBM4 and other clostridial CBM4s. • FN 3 The CBM4 binding pocket with bound cellobiose Our structural and computational studies indicate a possible role in binding for Trp118 Treatment of dilute acid pretreated corn stover with Fn(III)-like domains showed no significant improvement in digestion relative to Spezyme CP alone. The role of the fibronectin domains in CbhA might not be related to digestion. Molecular dynamics simulation snapshot of CbhA CBM4 with cellohexaose Domain swapping leads to an enhanced cellulase Domain-swapping doubles activity of CbhA! Wild type CbhA New cellulase Wild-type CbhA 24 “Coated” C. cellulolyticum morphology Enzymatic domain Dockerin domain CBM3 Cohesin domain CipC (Scaffoldin) Using CHARMM (MD) to begin to visualize these systems Mike Crowley, NREL Acknowledgements • • • • • • • • • • • • • • Steve Decker Roman Brunecky Shi-You Ding Bryon Donohoe John Baker Yannick Bomble Qi Xu Peter Ciesielski Deanne Sammond Mike Resch John Yarbrough Michael Crowley Marcus Alahuhta Vladimir Lunin • • • • • • • • • • • Ed Bayer (Weizmann) David Wilson (Cornell) Maxim Kostylev (Cornell) Adam Guss (ORNL) Bob Hettich (ORNL) Rich Giannone (ORNL) Lee Lynd (Dartmouth) Dan Olsen (Dartmouth) Mo Chen (Cornell) John Brady (Cornell) Igor Zhulin (UT-Knoxville)