Supporting Information Combinatorial Pathway Engineering for Optimized Production of the Anti-malarial FR900098 Todd S. Freestone1 and Huimin Zhao1,2,3,4* 1 Department of Chemical and Biomolecular Engineering and Institute for Genomic Biology, 2 School of Molecular and Cellular Biology, 3Department of Chemistry, 4Departments of Biochemistry and Bioengineering, Center for Biophysics and Computational Biology, University of Illinois at Urbana-Champaign, Urbana, IL 61801 1 Supplementary methods Finding optimal overhangs for Golden Gate assembly. A Perl script was written that used UNAfold (Markham and Zuker 2008) to calculate the Gibbs free energy of all 256 possible 4base pair overhangs against themselves. Those having a negative free energy against themselves were screened out as well as those that had a free energy greater than -1 kcal/mol (an arbitrarily chosen cutoff) when hybridized with their reverse complement, leaving 154 overhangs. The Gibbs free energy for binding between each of the remaining overhangs was calculated against all other overhangs and their reverse complements. A set of four overhangs were found, ACAC, CCCT, TAGC, and TTCG, that gave strong binding to their own reverse complements but not to themselves, the other overhangs, or the reverse complements of the other overhangs (Supplementary Table 5). Knockout of E. coli BL21(DE3) genes. Genes aceE, acs, adhE, atpF, atpH, frdC, ldhA, pflB, poxB, ppc, pta, and pykF were knocked out using P1 phage recombineering. E. coli BW25113 (Datsenko and Wanner 2000) knockout strains from the Keio collection (Baba et al. 2006) were used as a source of the mutations, and genomes were compared between this strain and E. coli BL21(DE3) to ensure that other modifications would be unlikely. To 1 mL of LBCG (LB broth with 2.5 mM CaCl2 and 0.1% glucose), 5 µL of P1 phage was added. 50 µL of the P1/LBCG solution was then added to 0.1 mL of LB-grown donor cells in a disposable culture tube and incubated at room temperature for 10 minutes for phage adsorption. To this solution 2.5 mL LBCG was added and gently mixed. 2.5 mL of molten LB top agar, cooled to 55 °C, was added and poured onto a thick (40 mL) LBCG agar plate and placed in a 37 °C incubator. After 4 hours, plates were checked for lysis. If lysis had begun, phages were harvested by adding 5 mL of ice cold Tris-Phage buffer (10 mM Tris-HCl, 100 mM NaCl, 10 mM MgCl2, 7.2 pH) to each plate, and plates were left overnight at 4 °C. Buffer was then collected, and a few drops of chloroform were added, mixed gently, and centrifuged for 10 minutes. Buffer containing phage was then transferred to a new tube for another round of decontamination by addition of chloroform. For P1 transduction, a BL21(DE3) colony was grown for 4 hours at 37 °C in 5 mL LB, and 1 mL of culture was centrifuged for 10 minutes and resuspended in 1 mL MC (10 mM MgCl2, 4 mM CaCl2) and stood for 5 minutes and then vortexed. 5 µL phage lysate was added to 0.1 mL LB to which 0.1 mL of resuspended cells were added and mixed gently. Cell/phage mixture stood for 20 minutes, then 1 mL LB with 10 mM ethylene glycol tetraacetic acid (EGTA) was added and vortexed. Cells were collected by centrifugation, the supernatant was discarded, and the cells were resuspended in another 1 mL of LB+EGTA. Cells were placed on a 37 °C shaker for 2 hours and then spread on LB agar plates with 25 µg/mL kanamycin and incubated at 37 °C for up to 2 days. Colonies were then streaked out on LB + 50 µg/mL kanamycin plates, and colony PCR was performed to see if the correct deletions were present. Primers used to check for the deletions are named with the ending “KO-seq-for” and “KO-seqrev”. Knockout strain ackA/pta/adhE was previously constructed (Zha et al. 2009). 2 Enriched library simulations. Simulations were performed using MATLAB (Mathworks, Natick, MA). For each simulation, a library of 1800 pathways was made where each of nine genes was assigned a promoter strength, pi, that matched the strengths of our promoter library of 0.03, 0.15, 0.45, and 1. The output for each pathway was then modeled using the equations outlined below, where the theoretical maximum output is 1. The total product output (PT) of each pathway is the cellular output (PC) multiplied by a cell mass term (m): ๐๐ = ๐๐๐ถ (Equation 1) Cellular output is determined by the rate-limiting step of the simulated pathway: ๐๐ถ = โก min (๐๐ ๐๐ ) ๐=1…8 (Equation 2) where k represents the turnover rate for enzyme i. Random values between 1 and 10 were assigned for each k (values 10, 2, 3.5, 1, 3, 1.5, 6, and 4 were assigned to enzymes 1 through 8). At least one pathway enzyme was assigned a turnover rate of 1 which with the rate-limiting output would ensure the maximum output did not exceed 1. The cell mass (also with a maximal possible value of 1) was dependent on the metabolic load caused by the promoter strengths (L), toxicity of an intermediate (tI), and toxicity of the product (tP): ๐ = (1 − ๐ฟ − ๐ก๐ผ )๐ก๐ (Equation 3) The metabolic load was calculated by summing the promoter strengths and dividing it by a normalization factor (n1, described in more detail below) to ensure that the growth rate will be positive: 9 1 ๐ฟ = ∑ ๐๐ ๐1 (Equation 4) ๐=1 The toxic intermediate was only present if a rate-limiting step upstream of the intermediate (steps 1 through 4) was greater than enzymatic step 5. Another normalization factor was used, again to ensure a positive growth rate: 0,โกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโก min (๐๐ ๐๐ ) ≤ ๐5 ๐5 ๐=1…4 ๐ก๐ผ = { 1 [ min (๐ ๐ )−๐5 ๐5 โก],โกโกโกโกโกโกโกโก min (๐๐ ๐๐ ) > ๐5 ๐5 ๐=1…4 ๐2 ๐=1…4 ๐ ๐ 3 (Equation 5) Product toxicity was determined by the cellular output as well as the expression of the immunity enzyme (gene 9): ๐ก๐ = { 1 − (๐๐ถ − ๐9 ),โกโกโกโกโกโกโกโก๐9 < ๐๐ถ 1,โกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโกโก๐9 ≥ ๐๐ถ (Equation 6) According to equation 3, it is possible for the cell mass to be negative if L + tI > 1; however, this is dependent on the turnover values (k) that the enzymes are assigned. For our mock pathway, normalization factors of n1 = 12 and n2 =10 were used, giving a maximal value of (L + tI)max = 0.761. These values were chosen to give the metabolic load greater weight (Lmax = 0.75) in determining cell mass as compared to the toxic intermediate (tI,max = 0.091), which is known to have low toxicity in the FR900098 pathway. Although these values to an extent are arbitrary, it should be remembered that the primary purpose of the model is to show the searching capabilities of the enriched library screening strategy and not to elucidate the actual metabolic characteristics of the FR900098 pathway. Supplementary Table I. Primers used in this work. No. Primer Sequence 1 NdeI-FrbA-for tgacgccatatgcgcgacctgttacgggac 2 FrbA-G1374T-for gtcccgtcctgggttaagaccagtctggcgccgggctcgc 3 FrbA-G1374T-rev gcgagcccggcgccagactggtcttaacccaggacgggac 4 FrbA-C1725G-for gacccggacggcgccgaggtgttcctcgcggacctgtggc 5 FrbA-C1725G-rev gccacaggtccgcgaggaacacctcggcgccgtccgggtc 6 HindIII-FrbA-rev tgacctaagctttcaggcctcggcgtcgagcag 7 NdeI-FrbC-for 8 FrbC-C395G-for 9 10 FrbC-C395G-rev HindIII-FrbCC1077G-rev tgacgccatatgcgcaacgacttagtgctcg caaggaccggggcgcgttcgtgtcgatcagcgccgaggac atc gatgtcctcggcgctgatcgacacgaacgcgccccggtcc ttg tgacctaagctttcaggcggcggccttgttgtagatctcg acgagctgctcgtgcgacacggcgccgccgatcctgatc 11 HindIII-FrbC-rev tgacctaagctttcaggcggcggccttgttgtag 12 NdeI-FrbD-for 13 FrbD-G384A-rev Use 14 FrbD-G468A-for tgacgccatatgaccaagcgaaccatgttac cttggccgccgcgagcttggcggcgaagccctcgatcgtt tcctggtcctgggccccgtc cttcgccgccaagctcgcggcggccaagaaggcccagcag accgacgacttcgtggtcgtggcgcggatcgaaaccttca tcgccg 15 HindIII-FrbD-rev NdeI-FrbE-A33Tfor tgacctaagctttcagcgctggagctcgaagacc tgacgccatatgcgtaaacacacagtcactctgatcgcgg gtgacggcagcggcccg HindIII-FrbE-rev NdeI-FrbG-C45Tfor tgacctaagctttcatccgatgtcccgcaggcgctcg tgacgccatatgacgcactacgcgaccgtcatctgcggcg gtggccccgcgggtgtctccgcggtggtg 16 17 18 4 FrbA mutations FrbC mutations FrbD mutations FrbE mutation FrbG mutation 19 HindIII-FrbG-rev tgacctaagctttcaacgcccatcctccgtacgatc 20 NdeI-FrbH-for 21 FrbH-G1773T-rev 22 HindIII-FrbH-rev 23 1A-low-for 24 1A-med-for 25 1A-hi-for 26 1A-WT-for 27 1B-low-for 28 1B-med-for 29 1B-hi-for 30 1B-WT-for 31 1C-low-for 32 1C-med-for 33 1C-hi-for 34 1C-WT-for 35 2A-low-for 36 2A-med-for 37 2A-hi-for 38 2A-WT-for 39 2B-low-for 40 2B-med-for 41 2B-hi-for 42 2B-WT-for 43 2C-low-for 44 2C-med-for 45 2C-hi-for 46 2C-WT-for tgacgccatatgaacgagaaccggaccttcgccac gccccgcgggccgccgacgatctcggcgaccgcggcgacc accacctcgttctccgcgcggctgcggctggagacacgca ggtaggcgtc tgacctaagctttcagccgtcggcacggcccgccccctcg gtcgtcgcggcgccccgcgggccgccgacgatc cggaagactaacacgtctcaacacgaaattaatacgactc actaccaccgaattgtgagc cggaagactaacacgtctcaacacgaaattaatacgactc actaggtgagaattgtgagc cggaagactaacacgtctcaacacgaaattaatacgactc actaaggcggaattgtgagc cgcctaggaagactaacacgtctcaacacgaaattaatac gactcactataggggaattg cggaagactaccctcgaaattaatacgactcactaccacc gaattgtgagc cggaagactaccctcgaaattaatacgactcactaggtga gaattgtgagc cggaagactaccctcgaaattaatacgactcactaaggcg gaattgtgagc cgcctaggaagactaccctcgaaattaatacgactcacta taggggaattg cggaagactatagcgaaattaatacgactcactaccaccg aattgtgagc cggaagactatagcgaaattaatacgactcactaggtgag aattgtgagc cggaagactatagcgaaattaatacgactcactaaggcgg aattgtgagc cgcctaggaagactatagcgaaattaatacgactcactat aggggaattg cgggtctcaacacgtctcaccctcgaaattaatacgactc actaccaccgaattgtgagc cgggtctcaacacgtctcaccctcgaaattaatacgactc actaggtgagaattgtgagc cgggtctcaacacgtctcaccctcgaaattaatacgactc actaaggcggaattgtgagc cgcctaggtctcaacacgtctcaccctcgaaattaatacg actcactataggggaattg cgggtctcaccctgcgaaattaatacgactcactaccacc gaattgtgagc cgggtctcaccctgcgaaattaatacgactcactaggtga gaattgtgagc cgggtctcaccctgcgaaattaatacgactcactaaggcg gaattgtgagc cgcctaggtctcaccctgcgaaattaatacgactcactat aggggaattg cgggtctcatagcgaaattaatacgactcactaccaccga attgtgagc cgggtctcatagcgaaattaatacgactcactaggtgaga attgtgagc cgggtctcatagcgaaattaatacgactcactaaggcgga attgtgagc cgcctaggtctcatagcgaaattaatacgactcactatag gggaattg 5 FrbG mutation FrbH mutation Pathway assembly 47 3A-low-for 48 3A-med-for 49 3A-hi-for 50 3A-WT-for 51 1A-rev 52 1B-rev 53 1C-rev 54 2A-rev 55 2B-rev 56 2C-rev 57 3C-rev 58 BB-I-for 59 BB-I-rev 60 BB-II/III-for cgggtctcaacacgtctcatagcgaaattaatacgactca ctaccaccgaattgtgagc cgggtctcaacacgtctcatagcgaaattaatacgactca ctaggtgagaattgtgagc cgggtctcaacacgtctcatagcgaaattaatacgactca ctaaggcggaattgtgagc cgcctaggtctcaacacgtctcatagcgaaattaatacga ctcactataggggaattg cttgaggaagactaagggctcctttcagcaaaaaacccct caag cttgaggaagactagctactcctttcagcaaaaaacccct caag cttgaggaagactacgaacgtctcaagggctcctttcagc aaaaaacccctcaag cttgagggtctcaagggctcctttcagcaaaaaacccctc aag cttgagggtctcagctactcctttcagcaaaaaacccctc aag cttgagggtctcacgaacgtctcagctactcctttcagca aaaaacccctcaag cttgagggtctcacgaacgtctcacgaactcctttcagca aaaaacccctcaag tgactcgaagactattcgcaataaaccggtaaaccagcaa tagac cttgaggaagactagtgtcgaccgatgcccttgagagcct tcaac tgactcggtctcattcgcaataaaccggtaaaccagcaat agac cttgagggtctcagtgtcgaccgatgcccttgagagcctt caac tgactcggtctcacccttatgcgacacacagagacgtcag ccgctacagggcgcgtccc tgactcggtctcacaagcgaatgagacgaccggtggaaag cgggcagtgag 61 BB-II/III-rev 62 GGBB-lacZa-for 63 GGBB-lacZa-rev 64 GGBB-pET-1-for tgactcggtctcacttgccagcgccctagcgcccgc 65 GGBB-pET-1-rev 66 GGBB-pET-2-for tgactcggtctcatagccgcggtatcattgcagcactgg tgactcggtctcagctaccacgctcaccggctccagattt atcag 67 tgactcggtctcacgaaaactcacgttaagggattttg ctcggtctcattcgttccactgagcgtcagaccccgtaga aacgtaacggcaaaagcacc 71 GGBB-pET-2-rev GGBB-pACYC-1for GGBB-pACYC-1rev GGBB-pACYC-2for GGBB-pACYC-2rev 72 T7-WT-for-check cccgcgaaattaatacgactcactata 73 T7-hi-for-check cccgcgaaattaatacgactcactaggtga 74 T7-med-for-check cccgcgaaattaatacgactcactaaggc 75 T7-low-for-check cccgcgaaattaatacgactcactaccacc 68 69 70 Pathway assembly pFRGG1 backbone pFRGG2/3 backbone pFRGG-BB assembly tgactcggtctcagtgtcactggtgaaaagaaaaacc tgactcggtctcaacacgggcaacagctgattgcc tgactcggtctcaagggagagcgtcgagatcccgg 6 PCR assay for promoter assignments 76 FrbA-prom-seq-rev gttcgtactcacggcgcatg 77 FrbB-prom-seq-rev acgaactcggaaacccggtc 78 FrbC-prom-seq-rev tggcactgcacgtcgatgtc 79 FrbD-prom-seq-rev gagttgcgcttcgggaagac 80 FrbE-prom-seq-rev tgatcggcttcagcgagaagg 81 FrbF-prom-seq-rev gagaacgtcggcatgaccag 82 FrbG-prom-seq-rev agtagaagatcggcggcagc 83 FrbH-prom-seq-rev tcggaggttgttggtcgtcg 84 DxrB-prom-seq-rev 85 T7-mut-3N 86 T7-mut-6N ccttgttcgccaggatgagc cgtagcatgccgaaattaatacgactcactaNaNgNgaat tgtgagcggataacaattc cgtagcatgccgaaattaatacgactcactNNNNNNgaat tgtgagcggataacaattc 87 XhoI-GFP-rev 88 T7-low-for 89 T7-med-for 90 T7-hi-for cacgtactcgagtcatttgtatagttcatccatg cgtagcatgccgaaattaatacgactcactaccaccgaat tgtgagcggataacaattc cgtagcatgccgaaattaatacgactcactaggtgagaat tgtgagcggataacaattc cgtagcatgccgaaattaatacgactcactaaggcggaat tgtgagcggataacaattc 91 aceE KO seq for cgcatcgccatctggcctttatcg 92 aceE KO seq rev caactttgtcgcccactttgaccagg 93 acs KO seq for cccgctcccttatgggagaagg 94 acs KO seq rev caacagcatgcataactgcatgttcctc 95 adhE KO seq for gcatgagcagaaagcgtcaggc 96 adhE KO seq rev cgccacctggaagtgacgc 97 atpF KO seq for cgactcggcgagcgtttctgg 98 atpF KO seq rev tcccgatgatcgctgtaggtctgg 99 atpH KO seq for cccggcagggagatcatttcacc 100 atpH KO seq rev gcagattctggacgaagcgaaagctg 101 frdC KO seq for ccaatgaagctctgcgcgaacg 102 frdC KO seq rev ccacggtaagaaggagcgtatggc 103 ldhA KO seq for tgctgtagctgttctggcgtaacagc 104 ldhA KO seq rev ccgagcgtcatcagcagcg 105 pflB KO seq for attgcggtgtttctccagatgtggcc 106 pflB KO seq rev tattgtaatccgcgacttcgcatccccg 107 poxB KO seq for gcctgagtgccggtaggcag 108 poxB KO seq rev cgtaccgtgatgacctgcggc 109 ppc KO seq for ccaacccagggctttccagc 110 ppc KO seq rev cgcatcttatccgacctacacctttgg 111 pta KO seq for cctggctgcacgtttcggc 112 pta KO seq rev ggaactacccaggtggcaaggc 113 pykF KO seq for cagcgtataatgcgcgccaattgac 114 pykF KO seq rev gttcgctcaaagaagcatcgaacgc 7 PCR assay for promoter assignments T7 promoter library generation Reconstruct chosen T7 promoter mutants To check for correct gene knockouts Supplementary Table II. Mutations made in FR900098 biosynthetic genes for Golden Gate compatibility. Gene Mutation 1 RE site Mutation 2 RE site frbA frbC frbD frbE frbG frbH 1374G>T 396C>G 384G>A 33A>T 45C>T 1775G>T BbsI Esp3I Esp3I Esp3I Esp3I BsaI 1735C>G 1077C>G 468G>A BbsI Esp3I BsaI Supplementary Table III. PCR and OE-PCR for the introduction of mutations in FR900098 biosynthetic genes. Gene frbA Template pET26b-frbA pET26b-frbA pET26b-frbA A6 pET26b-frbC pET26b-frbC C3 pET26b-frbC pET26b-frbC D3 pET26b-frbE Forward primer/fragment NdeI.FrbA for FrbA G1374T for FrbA C1725G for A1 A2 A4 NdeI.FrbA for NdeI.FrbC for FrbC C395G for C1 NdeI.FrbC for NdeI.FrbD for FrbD G468A for D1 NdeI.FrbD for frbE Fragment A1 A2 A3 A4 A5 A6 A7 C1 C2 C3 C4 D1 D2 D3 D4 E1 frbG G1 pET26b-frbG frbH H1 H2 pET26b-frbH H1 NdeI.FrbG C45T for NdeI.FrbH for NdeI.FrbH for frbC frbD NdeI.FrbE A33T for 8 Reverse primer/ fragment FrbA G1374T rev FrbA C1725G rev HindIII.FrbA rev A2 A3 A5 HindIII.FrbA rev FrbC C395G rev HindIII.FrbC C1077G rev C2 HindIII.FrbC rev FrbD G384A rev HindIII.FrbD rev D2 HindIII.FrbD rev HindIII.FrbE rev Method PCR PCR PCR OE-PCR OE-PCR OE-PCR PCR PCR PCR OE-PCR PCR PCR PCR OE-PCR PCR PCR HindIII.FrbG rev PCR FrbH G1773T rev HindIII.FrbH rev PCR PCR Supplementary Table IV. combinatorial pathway. Fragment Template frbA-lo frbA-med frbA-hi frbA-WT frbB-lo frbB-med frbB-hi frbB-WT frbC-lo frbC-med frbC-hi frbC-WT frbD-lo frbD-med frbD-hi frbD-WT frbE-lo frbE-med frbE-hi frbE-WT pFRGG-A pFRGG-A pFRGG-A pFRGG-A pET26b-frbB pET26b-frbB pET26b-frbB pET26b-frbB pFRGG-C pFRGG-C pFRGG-C pFRGG-C pFRGG-D pFRGG-D pFRGG-D pFRGG-D pFRGG-E pFRGG-E pFRGG-E pFRGG-E PCR reactions for fragments used to construct the FR900098 Forward primer 1A low for 1A med for 1A hi for 1A WT for 1B low for 1B med for 1B hi for 1B WT for 2A low for 2A med for 2A hi for 2A WT for 2B low for 2B med for 2B hi for 2B WT for 2C low for 2C med for 2C hi for 2C WT for Reverse primer 1A rev 1A rev 1A rev 1A rev 1B rev 1B rev 1B rev 1B rev 2A rev 2A rev 2A rev 2A rev 2B rev 2B rev 2B rev 2B rev 2C rev 2C rev 2C rev 2C rev Fragment Template frbF-lo frbF-med frbF-hi frbF-WT frbG-lo frbG-med frbG-hi frbG-WT frbH-lo frbH-med frbH-hi frbH-WT dxrB-lo dxrB-med dxrB-hi dxrB-WT BB-I BB-II/III pET26b-frbF pET26b-frbF pET26b-frbF pET26b-frbF pFRGG-G pFRGG-G pFRGG-G pFRGG-G pFRGG-H pFRGG-H pFRGG-H pFRGG-H pET26b-dxrB pET26b-dxrB pET26b-dxrB pET26b-dxrB pACYCDuet-1 pACYCDuet-1 Forward primer 3A low for 3A med for 3A hi for 3A WT for 1C low for 1C med for 1C hi for 1C WT for 2B low for 2B med for 2B hi for 2B WT for 2C low for 2C med for 2C hi for 2C WT for BB-I for BB-II/III for Supplementary Table V. PCR reactions for the construction of pFRGG-BB. Fragment GGBB lacZα GGBB pET 1 GGBB pET 2 GGBB pACYC 1 GGBB pACYC 2 Template lacZα pETDuet-1 pETDuet-1 pACYCDuet-1 pACYCDuet-2 Forward primer GGBB lacZa for GGBB pET 1 for GGBB pET 2 for GGBB pACYC 1 for GGBB pACYC 2 for 9 Reverse primer GGBB lacZa rev GGBB pET 1 rev GGBB pET 2 rev GGBB pACYC 1 rev GGBB pACYC 2 rev Reverse primer 2A rev 2A rev 2A rev 2A rev 1C rev 1C rev 1C rev 1C rev 2B rev 2B rev 2B rev 2B rev 3C rev 3C rev 3C rev 3C rev BB-I rev BB-II/III rev Supplementary Table VI. Gibbs free energy (kcal/mol) between the four chosen overhangs (ACAC, CCCT, TAGC, TTCG). Dark green denotes strong binding. Maximum calculation for free energy is 5 kcal/mol, which has been listed for combinations with no binding. Overhangs Forward sequences ACAC CCCT TAGC TTCG Reverse complements GTGT AGGG GCTA CGAA ACAC 5 5 5 5 -2.32 5 5 5 CCCT 5 5 0.61 5 5 -4.05 5 5 TAGC 5 0.61 0.003 1.859 5 5 -2.762 5 TTCG 5 5 5 5 -2.511 5 1.859 1.569 10 Supplementary Figure 1. Number of screens necessary for 1x coverage of a combinatorial pathway library depending on library size compared and screening rate. Numbers in blue circles represent pathways studied in the references below the table. The star represents this study. 1. 2. 3. 4. Wu, J. J., Du, G. C., Zhou, J. W., and Chen, J. 2013. Metabolic engineering of Escherichia coli for (2S)-pinocembrin production from glucose by a modular metabolic strategy. Metab Eng 16: 48-55. Ajikumar, P. K., Xiao, W. H., Tyo, K. E. J., Wang, Y., Simeon, F., Leonard, E., Mucha, O., Phon, T. H., Pfeifer, B., and Stephanopoulos, G. 2010. Isoprenoid pathway optimization for taxol precursor overproduction in Escherichia coli. Science 330:70-74. Du, J., Yuan, Y. B., Si, T., Lian, J. Z., and Zhao, H. M. 2012. Customized optimization of metabolic pathways by combinatorial transcriptional engineering. Nucleic Acids Res 40, e142. Latimer, L. N., Lee, M. E., Medina-Cleghorn, D., Kohnz, R. A., Nomura, D. K., and Dueber, J. E. 2014. Employing a combinatorial expression approach to characterize xylose utilization in Saccharomyces cerevisiae. Metab Eng 25:20-29. 11 Supplementary Figure 2. Promoter strengths of a chosen library. Originals promoters screened in at least quadruplicates, and reconstructed promoters in duplicates except for the High promoter, which only had one sample. Fluorescence (au) 10000 1000 100 Original Reconstructed 10 1 Low Weak High Promoter 12 WT Supplementary Figure 3. (A) Feeding assay of various amounts of supernatant from all WT T7 promoter strain fed to 120 µL culture of the phosphonate uptake strain WM6242. Each point shows average of 48 culture wells (half of 96-well plate), with error bars showing the minimum and maximum OD 600 measured for each supernatant volume fed. Although the assay is not quantitative, it can show whether the concentration is above a certain threshold. (B-D) OD600 measured of WM6242 cultures fed supernatants from strains of each round of screening. Samples are ordered from strongest to weakest growth inhibition, with hits carried over for direct LC-MS quantification highlighted in red. Z-factors for the random library (B), first enriched library (C), and second enriched library (D) are 0.253, 0.383, and -0.250, and were calculated by using the hits as controls and non-hits as background (Zhang et al. 1999). The negative Z-factor for the second enriched library is due to all hits being along the upper half of the dynamic range, giving them a high average and large standard deviation. 1.2 A. 1 OD 600 0.8 0.6 0.4 0.2 0 0.01 0.1 1 10 Production supernatant fed (µl) B. OD 600 Random library growth inhibition of WM6242 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Non-hits Hits 0 500 1000 1500 Sample 13 2000 2500 C. First enriched library growth inhibition of WM6242 1.4 1.2 OD 600 1 0.8 0.6 Non-hits 0.4 Hits 0.2 0 0 500 1000 1500 2000 2500 Sample D. Second enriched library growth inhibition of WM6242 1.4 1.2 OD 600 1 0.8 0.6 Non-hits 0.4 Hits 0.2 0 0 500 1000 1500 Sample 14 2000 2500 Supplementary Figure 4. Production of top strains from each round that were tested as 5 mL cultures. Error bars show the standard deviation of duplicates. Below the chart, a table of the promoter assignments is given of each round’s top pathway (boldfaced) as well an additional three strains from the 2nd enriched library. Relative FR900098 production 6 5 4 All WT T7 promoters 3 Random library 1st enriched library 2 2nd enriched library 1 0 4pT 4pTR 4pTRA 4pTRB 4pTRN 4pTRQ FrbD Med Low Low Low Low Low FrbC High High High High High High FrbA Med High High FrbB Low Low Med Med High Med Low Low Low 15 FrbE Low Low Low Low Low FrbH Low Low Low Low Low FrbG High High High High High FrbF Med High High High High DxrB High High Low High Low Low Low High High High Supplementary Figure 5. Gene knockout effects on FR900098 output in strain 4pTR. FR900098 (mg/L) 120 100 80 60 40 20 0 Knockout strains References Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H. 2006. Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2. Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. P Natl Acad Sci USA 97:6640-6645. Markham NR, Zuker M. 2008. UNAFold: software for nucleic acid folding and hybridization. Methods Mol Biol 453:3-31. Zha WJ, Rubin-Pitel SB, Shao ZY, Zhao HM. 2009. Improving cellular malonyl-CoA level in Escherichia coli via metabolic engineering. Metab Eng 11:192-198. Zhang JH, Chung TDY, Oldenburg KR. 1999. A simple statistical parameter for use in evaluation and validation of high throughput screening assays. J Biomol Screen 4:67-73. 16