METHODS FOR DECOUPLING YIELD AND PRODUCTIVITY OF A NON- CATABOLIC COMPOUND PRODUCED BY A HOST CELL

BACKGROUND OF THE INVENTION

[0001] Yield and productivity are two major cost drivers for any biomanufacturing process. For the biomanufacturing of non-catabolic compounds like isoprenoids, productivity and yield are frequently a function of other cellular rates such as the rate of sugar and/or oxygen consumption. Design of fermentation processes that achieve the optimal combination of productivity and yield for the lowest cost production requires close characterization of this relationship. For the production of isoprenoids, under standard fermentation conditions, yield is consistently inversely correlated with the cell specific rates of oxygen and sugar uptake and by extension productivity, i.e. the faster oxygen and sugar is taken up by isoprenoid producing cells, the lower the isoprenoid yield. This relationship has been termed“rate-yield coupling.” This inverse coupling of yield and productivity is problematic because yield and productivity are two key cost drivers in isoprenoid production. Because of rate-yield coupling, any attempt to increase productivity by increasing the rate of oxygen and/or sugar transfer results in a concomitant decrease in yield, thereby negating the cost benefit of increased productivity. The elimination of the inverse correlation between yield and productivity would be beneficial because it would be possible to simultaneously achieve high yield and high productivity and thereby maximize the efficiency (maximize isoprenoid product produced per cost of fermentation) of the isoprenoid production.

SUMMARY OF THE INVENTION

[0002] The invention relates generally to methods of uncoupling yield and productivity during the production of a non-catabolic compound during fermentation of a host cell that produces the non-catabolic compound.

[0003] In one aspect the invention provides method of decoupling yield and productivity of a non-catabolic compound produced in a host cell capable of making the non-catabolic compound involving the step of reducing ATP utilization during fermentation.

[0004] In an embodiment the ATP utilization is reduced by addition of one or more ATP depleting agents. In another embodiment the one or more ATP depleting agents is a weak organic acid. In certain embodiments the weak organic acid is selected from sorbic acid, acetic acid, benzoic acid, and propionic acid. In a preferred embodiment the weak organic acid is benzoic acid.

[0005] In other embodiments of the invention the ATP utilization is reduced by over expression of one or more ATP dissipation enzymes. In an embodiment the one or more ATP dissipation enzymes are selected from Saccharomyces cerevisiae SSB1 and ATP-diphosphohydrolase. In another embodiment the ATP utilization is reduced by over expression of one or more ATP uncoupling enzymes. In particular embodiments the one or more ATP uncoupling enzymes are selected from NADH oxidase (NOX) and alternative oxidase (AOX).

[0006] In further embodiments of the invention the ATP levels are reduced by expression of a futile cycle in the host cell. In certain embodiments the futile cycle is selected from

simultaneous over expression of phosphofmctokinase and fructose- 1,6-bisphosphatase and simultaneous over expression of phosphoenolpyruvate carboxykinase and pyruvate carboxylase.

[0007] In an embodiment of the method of the invention the non-catabolic compound is selected from the group consisting of an amino acid, a fatty acid, an isoprenoid, and a polyketide. In certain embodiments the non-catabolic compound is an isoprenoid. In particular embodiments the isoprenoid is selected from the group consisting of a hemiterpene, monoterpene, diterpene, triterpene, tetraterpene, sesquiterpene, and polyterpene. In other embodiments the isoprenoid is selected from the group consisting of abietadiene, amorphadiene, carene, a-famesene, b- famesene, famesol, geraniol, geranylgeraniol, isoprene, linalool, limonene, myrcene, nerolidol, ocimene, patchoulol, b-pinene, sabinene, g-terpinene, terpinolene, and valencene. In a preferred embodiment the isoprenoid is b-farnesene.

[0008] In an embodiment of the method of the invention the host cell is selected from a bacterial cell, a plant cell, and a yeast cell. In certain embodiments the host cell is a yeast cell. In a preferred embodiment the yeast cell is a Saccharomyces cerevisiae.

[0009] In another aspect of the invention the invention provides a method of decoupling yield and productivity of a non-catabolic compound produced in a host cell capable of making the isoprenoid compound involving the step of reducing carbon flux through the citric acid cycle (TCA) in the host cell.

[0010] In an embodiment of the method of the invention carbon flux through the TCA cycle is reduced by inhibition of one or more TCA enzymes. In another embodiment the one or more TCA enzymes are downregulated. In a further embodiment the TCA enzymes are selected from citrate synthase, aconitate hydratase, NAD-dependent isocitrate dehydrogenase, 2-ketoglutarate dehydrogenase, succinyl-CoA ligase, succinate dehydrogenase, fumarate hydralase, peroxisomal malate dehydrogenase, and pyruvate carboxylase. In a preferred embodiment the TCA enzymes are pyruvate carboxylase and citrate synthase.

BRIEF DESCRIPTION OF THE DRAWINGS

[0011] FIG.l is a graph of yield of farnesene at different cell specific oxygen uptake values (q02).

[0012] FIG.2 is a graph of yield of farnesene at different cell specific sugar uptake values (qS).

[0013] FIG.3 is a graph of yield of farnesene at different productivity values.

FIG.4 is a graph of cell specific sugar uptake values (qS) at different oxygen uptake values (q02).

[0014] FIG.5 is a graph plotting the ratio of alpha keto glutarate to isocitrate at different oxygen uptake rates.

[0015] FIG.6 is a diagram of a modification on a genome scale model that enabled the modeling of the impact of ATP production on yield of a product, by eliminating hydrolysis or futile cycles as potential sinks for ATP. (NGAM is“Non-growth associated maintenance.”)

[0016] FIG.7 is a graph plotting the data from a simulation of farnesene yield as a function of qS, at different assumed fluxes through the TCA cycle. Actual experimental data are plotted over the simulation values as black circles.

[0017] FIG.8 is a set of graphs showing the effect of different benzoic acid concentrations (X axis = mM benzoic acid) on specific sugar uptake (qS) (Y axis top panel); growth rate (Y axis second panel from top); specific farnesene productivity (qP) ( Y axis third panel from top); and calculated product yield (qP/qS) (Y axis bottom panel).

[0018] FIG.9 is a graph showing farnesene yield over different specific oxygen uptake rates (q02) for three strains: Y21901 (control); Y22021 (control); and Y31655 (Y22021 over expressing NOX). [0019] FIG.10 is a graph showing famesene yield over different specific oxygen uptake rates (q02) for four strains: Y21601 (control); Y27662 (PYC1 downregulation); Y29438 (PYC1 and CITl downregulation); and Y39666 (CITl downregulation).

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] As used herein, the term“productivity” or“qp” with respect to the biomanufacturing of isoprenoid products refers to the moles of isoprenoid product produced per gram of dry cell weight per hour.

[0021] As used herein, the term“yield” or“Yp _/ s” with respect to the biomanufacturing of isoprenoid products refers to the rate of isoprenoid product formation over the rate of sugar consumption.

[0022] As used herein, the term“qs” or“sugar consumption” refers to the rate of sugar consumed during a fermentation presented as moles sugar consumed per gram dry cell weight per hour.

[0023] As used herein, the term“qo2” or“oxygen consumption” refers to the rate of oxygen consumed during a fermentation presented as moles O2 consumed per gram dry cell weight per hour.

[0024] As used herein, the term“rate-yield coupling” refers to the inverse correlation between yield and productivity during the fermentation based production of an isoprenoid.

[0025] As used herein, the term“weak organic acid” or“(WOA)” refers to organic acids having a pKa of 4.0 or greater. Non-limiting illustrative examples of WOAs include sorbic acid, acetic acid, benzoic acid, and propionic acid.

[0026] As used herein, the term“futile cycle” refers to at least two metabolic cycles or pathways that when run concurrently in opposite directions have no effect other than the dissipation of energy in the form of hydrolysis of ATP.

[0027] As used herein, the term“ATP dissipation reaction” refers to a biochemical reaction that hydrolyzes ATP without utilizing the energy for any physiological process.

[0028] As used herein, the term“ATP uncoupling reaction” refers to a biochemical reaction that uncouples NADH oxidation or proton transport from ATP generation.

[0029] As used herein, the term“heterologous” refers to what is not normally found in nature. The term“heterologous nucleotide sequence” refers to a nucleotide sequence not normally found in a given cell in nature. As such, a heterologous nucleotide sequence may be: (a) foreign to its host cell (i.e., is“exogenous” to the cell); (b) naturally found in the host cell (i.e.,“endogenous”) but present at an unnatural quantity in the cell (i.e., greater or lesser quantity than naturally found in the host cell); or (c) be naturally found in the host cell but positioned outside of its natural locus.

[0030] As used herein, to“functionally disrupt” or a“functional disruption” of a target gene, e.g., one or more genes of the TCA pathway, means that the target gene is altered in such a way as to decrease in the host cell the activity of the protein encoded by the target gene. In some embodiments, the activity of the protein encoded by the target gene is eliminated in the host cell. In other embodiments, the activity of the protein encoded by the target gene is decreased in the host cell. Functional disruption of the target gene may be achieved by deleting all or a part of the gene so that gene expression is eliminated or reduced, or so that the activity of the gene product is eliminated or reduced. Functional disruption of the target gene may also be achieved by mutating a regulatory element of the gene, e.g., the promoter of the gene so that expression is eliminated or reduced, or by mutating the coding sequence of the gene so that the activity of the gene product is eliminated or reduced. In some embodiments, functional disruption of the target gene results in the removal of the complete open reading frame of the target gene.

[0031] As used herein, the term“parent cell” refers to a cell that has an identical genetic background as a host cell disclosed herein except that it does not comprise a particular heterologous nucleotide sequence, and that serves as the starting point for introducing said heterologous nucleotide sequence leading to the generation of a host cell disclosed herein.

[0032] As used herein, the term“biosynthetic enzyme” refers to an enzyme that functions in a biosynthetic pathway leading to the production of a naturally occurring molecule.

Genetically Modified Microbes Producing Isoprenoids

Host Cells

[0033] Host cells useful in the practice of the present invention include archae, prokaryotic, or eukaryotic cells.

[0034] Suitable prokaryotic hosts include but are not limited to any of a variety of gram-positive, gram-negative, or gram-variable bacteria. Examples include but are not limited to cells belonging to the genera: Agrobacterium, Alicyclobacillus, Anabaena, Anacystis, Arthrobacter, Azobacter, Bacillus, Brevibacterium, Chromatium, Clostridium, Corynebacterium, Enterobacter, Erwinia, Escherichia, Lactobacillus, Lactococcus, Mesorhizobium, Methylobacterium, Microbacterium, Phormidium, Pseudomonas, Rhodobacter, Rhodopseudomonas, Rhodospirillum, Rhodococcus, Salmonella, Scenedesmun, Serratia, Shigella, Staphlococcus, Strepromyces, Synnecoccus, and Zymomonas. Examples of prokaryotic strains include but are not limited to: Bacillus subtilis, Bacillus amyloliquefacines, Brevibacterium ammoniagenes, Brevibacterium immariophilum, Clostridium beigerinckii, Enterobacter sakazakii, Escherichia coli, Lactococcus lactis,

Mesorhizobium loti, Pseudomonas aeruginosa, Pseudomonas mevalonii, Pseudomonas pudica, Rhodobacter capsulatus, Rhodobacter sphaeroides, Rhodospirillum rubrum, Salmonella enterica, Salmonella typhi, Salmonella typhimurium, Shigella dysenteriae, Shigella flexneri, Shigella sonnei, and Staphylococcus aureus. In a particular embodiment, the host cell is an Escherichia coli cell.

[0035] Suitable archae hosts include but are not limited to cells belonging to the genera:

Aeropyrum, Archaeglobus, Halobacterium, Methanococcus, Methanobacterium, Pyrococcus, Sulfolobus, and Thermoplasma. Examples of archae strains include but are not limited to:

Archaeo globus fulgidus, Halobacterium sp., Methanococcus jannaschii, Methanobacterium thermoautotrophicum, Thermoplasma acidophilum, Thermoplasma volcanium, Pyrococcus horikoshii, Pyrococcus abyssi, and Aeropyrum pernix.

[0036] Suitable eukaryotic hosts include but are not limited to fungal cells, algal cells, insect cells, and plant cells. In some embodiments, yeasts useful in the present methods include yeasts that have been deposited with microorganism depositories ( e.g . IFO, ATCC, etc.) and belong to the genera Aciculoconidium, Ambrosiozyma, Arthroascus, Arxiozyma, Ashbya, Babjevia, Bensingtonia, Botryoascus, Botryozyma, Brettanomyces, Bullera, Bulleromyces, Candida, Citeromyces, Clavispora, Cryptococcus, Cystofilobasidium, Debaryomyces, Dekkara,

Dipodascopsis, Dipodascus, Eeniella, Endomycopsella, Eremascus, Eremothecium,

Erythrobasidium, Fellomyces, Filobasidium, Galactomyces, Geotrichum, Guilliermondella, Hanseniaspora, Hansenula, Hasegawaea, Holtermannia, Hormoascus, Hyphopichia,

Issatchenkia, Kloeckera, Kloeckeraspora, Kluyveromyces, Kondoa, Kuraishia, Kurtzmanomyces, Leucosporidium, Lipomyces, Lodderomyces, Malassezia, Metschnikowia, Mrakia, Myxozyma, Nadsonia, Nakazawaea, Nematospora, Ogataea, Oosporidium, Pachysolen, Phachytichospora, Phaffia, Pichia, Rhodosporidium, Rhodotorula, Saccharomyces, Saccharomycodes,

Saccharomycopsis, Saitoella, Sakaguchia, Saturnospora, Schizoblastosporion, Schizosaccharomyces, Schwanniomyces, Sporidiobolus, Sporobolomyces, Sporopachydermia, Stephanoascus, Sterigmatomyces, Sterigmatosporidium, Symbiotaphrina, Sympodiomyces, Sympodiomycopsis, Torulaspora, Trichosporiella, Trichosporon, Trigonopsis, Tsuchiyaea, Udeniomyces, Waltomyces, Wickerhamia, Wickerhamiella, Williopsis, Yamadazyma, Yarrowia, Zygoascus, Zygosaccharomyces, Zygowilliopsis, and Zygozyma, among others.

[0037] In some embodiments, the host microbe is Saccharomyces cerevisiae, Pichia pastoris, Schizosaccharomyces pombe, Dekkera bruxellensis, Kluyveromyces lactis (previously called Saccharomyces lactis), Kluveromyces marxianus, Arxula adeninivorans , or Hansenula polymorpha (now known as Pichia angusta). In some embodiments, the host microbe is a strain of the genus Candida, such as Candida lipolytica, Candida guilliermondii, Candida krusei, Candida pseudotropicalis, or Candida utilis.

[0038] In a particular embodiment, the host microbe is Saccharomyces cerevisiae. In some embodiments, the host is a strain of Saccharomyces cerevisiae selected from the group consisting of Baker’s yeast, CBS 7959, CBS 7960, CBS 7961, CBS 7962, CBS 7963, CBS 7964, IZ-1904, TA, BG-1, CR-1, SA-1, M-26, Y-904, PE-2, PE-5, VR-1, BR-1, BR-2, ME-2, VR-2, MA-3, MA-4, CAT-1, CB-1, NR-1, BT-1, and AL-1. In some embodiments, the host microbe is a strain of Saccharomyces cerevisiae selected from the group consisting of PE-2, CAT-1, VR-1, BG-1, CR-1, and SA-1. In a particular embodiment, the strain of Saccharomyces cerevisiae is PE-2. In another particular embodiment, the strain of Saccharomyces cerevisiae is CAT-1. In another particular embodiment, the strain of Saccharomyces cerevisiae is BG-1.

[0039] In some embodiments, the host microbe is a microbe that is suitable for industrial fermentation, e.g., bioethanol fermentation. In particular embodiments, the microbe is conditioned to subsist under high solvent concentration, high temperature, expanded substrate utilization, nutrient limitation, osmotic stress due to sugar and salts, acidity, sulfite and bacterial contamination, or combinations thereof, which are recognized stress conditions of the industrial fermentation environment.

NADH-using HMGRs

[0040] In another aspect, provided herein is a genetically modified host cell capable of producing an isoprenoid, the cell comprising one or more heterologous nucleotide sequences encoding acetylaldehyde dehydrogenase acetylating (ADA, EC 1.2.1.10) and one or more heterologous nucleotide sequences encoding one or more enzymes of a biosynthetic pathway for the isoprenoid, wherein the one or more enzymes of the biosynthetic pathway comprise an NADH-using enzyme. Without being bound by theory, it is believed that the increased intracellular pool of NADH generated by ADA in the conversion of acetaldehyde to acetyl-CoA is utilized by the NADH-using biosynthetic enzyme, thus helping to restore intracellular redox balance while increasing the yield of the acetyl-CoA derived product.

[0041] In some embodiments, the NADH-using enzyme is an enzyme that is non-native to the biosynthetic pathway. For example, the NADH-using enzyme can replace an NADPH-using enzyme that is native to the biosynthetic pathway. In other embodiments, the NADH-using enzyme is co-expressed with a NADPH-using enzyme that is native to the biosynthetic pathway. In some embodiments, the genetically modified host cell comprises HMGR(s) that can only utilize NADH as a cofactor.

[0042] In some embodiments, the genetically modified host cell is capable of producing an isoprenoid, and the cell comprises one or more heterologous nucleotide sequences encoding one or more enzymes of a mevalonate (MEV) pathway for making isopentenyl pyrophosphate, wherein the one or more enzymes comprise a NADH-using HMG-CoA reductase (HMGR).

HMG-CoA reductases catalyze the reductive deacylation of (S)-HMG-CoA to (R)-mevalonate, and are composed of two classes, class I and class II HMGrs. Class I includes the enzymes from eukaryotes and most archaea, and class II includes the HMG-CoA reductases of certain prokaryotes and archaea. In addition to the divergence in the sequences, the enzymes of the two classes also differ with regard to their cofactor specificity. Unlike the class I enzymes, which utilize NADPH exclusively, the class II HMG-CoA reductases vary in the ability to discriminate between NADPH and NADH. See, e.g., Hedl et al., Journal of Bacteriology 186 (7): 1927-1932 (2004). Co-factor specificities for select class II HMGRs are provided below.

Table 1. Co-factor specificities for select class II HMGRs

[0043] Useful HMGRs for the compositions and methods provided herein include HMGRs that are capable of utilizing NADH as a cofactor, e.g., HMGR from P. mevalonii, A. fulgidus or S. aureus. In particular embodiments, the HMGR is capable of only utilizing NADH as a cofactor, e.g., HMGR from P. mevalonii, S. pomeroyi or D. acidovorans.

[0044] In some embodiments, the NADH-using HMGR is from Pseudomonas mevalonii. The sequence of the wild type mvaA gene of Pseudomonas mevalonii, which encodes HMGR (E.C. 1.1.1.88), has been previously described. See Beach and Rodwell, J. Bacteriol. 171:2994-3001 (1989). Representative mvaA nucleotide sequences of Pseudomonas mevalonii include Genbank accession number M24015, and SEQ ID NO: 3as provided herein. Representative HMGR protein sequences of Pseudomonas mevalonii include Genbank accession number AAA2583, and SEQ ID NO: 4 as provided herein.

[0045] In some embodiments, the NADH-using HMGR is from Silicibacter pomeroyi. A representative HMGR nucleotide sequence of Silicibacter pomeroyi includes SEQ ID NO: 5 as provided herein. Representative HMGR protein sequences of Silicibacter pomeroyi include Genbank accession number YP_164994 and SEQ ID NO: 6 as provided herein.

[0046] In some embodiments, the NADH-using HMGR is from Delftia acidovorans . A representative HMGR nucleotide sequence of Delftia acidovorans includes SEQ ID NO: 7 as provided herein. Representative HMGR protein sequences of Delftia acidovorans include Genbank accession number YP_001561318 and SEQ ID NO: 8 as provided herein.

[0047] NADH-using HMGRs also useful in the compositions and methods provided herein include those molecules which are said to be "derivatives" of any of the NADH-using HMGRs described herein, e.g., from P. mevalonii, S. pomeroyi and D. acidovorans . Such a“derivative” has the following characteristics: (1) it shares substantial homology with any of the NADH-using HMGRs described herein; and (2) is capable of catalyzing the reductive deacylation of (S)- HMG-CoA to (R)-mevalonate using NADH as a cofactor. A derivative of an NADH-using HMGR is said to share“substantial homology” with NADH-using HMGR if the amino acid sequences of the derivative is at least 80%, and more preferably at least 90%, and most preferably at least 95%, the same as that of NADH-using HMGR.

[0048] In some embodiments, the NADH-using HMGR is selective for NADH over NADPH as a cofactor. In some embodiments, the NADH-using HMGR is selective for NADH over NADPH as a cofactor at a A ^' ,,, ^' ' ¹⁴¹³¹¹ : _m ^NADPH ratio of at least 1:2, 1:3, 1:4, 1:5 or greater than 1:5. In some embodiments, the NADH-using HMGR is engineered to be selective for NADH over NAPDH, for example, through site-directed mutagenesis of the cofactor-binding pocket.

Methods for engineering NADH-selectivity are described in Watanabe et al., Microbiology 153:3044-3054 (2007), and methods for determining the cofactor specificity of HMGRs are described in Kim et al., Protein Sci. 9: 1226-1234 (2000), the contents of which are hereby incorporated by reference in their entireties.

[0049] In some embodiments, the NADH-using HMGR is derived from a host species that natively comprises a mevalonate degradative pathway, for example, a host species that catabolizes mevalonate as its sole carbon source. Within these embodiments, the NADH-using HMGR, which normally catalyzes the oxidative acylation of internalized (R)-mevalonate to (S)- HMG-CoA within its native host cell, is utilized to catalyze the reverse reaction, that is, the reductive deacylation of (S)-HMG-CoA to (R)-mevalonate, in a genetically modified host cell comprising a mevalonate biosynthetic pathway. Prokaryotes capable of growth on mevalonate as their sole carbon source have been described by: Anderson et al., J. Bacteriol, 171(12):6468- 6472 (1989); Beach et al., J. Bacteriol. 171:2994-3001 (1989); Bensch et al., J. Biol. Chem. 245:3755-3762; Fimongnari et al., Biochemistry 4:2086-2090 (1965); Siddiqi et al., Biochem. Biophys. Res. Commun. 8: 110-113 (1962); Siddiqi et al., J. Bacteriol. 93:207-214 (1967); and Takatsuji et al., Biochem. Biophys. Res. Commun.110: 187-193 (1983), the contents of which are hereby incorporated by reference in their entireties.

Methods of Making Genetically Modified Cells

[0050] The methods provided herein include methods for producing a host cell that is genetically engineered to comprise an ADA and/or an NADH-using biosynthetic enzyme. Expression of an ADA and/or an NADH-using biosynthetic enzyme in a host cell can be accomplished by introducing into the host cells a nucleic acid comprising a nucleotide sequence encoding the ADA and/or NADH-using biosynthetic enzyme under the control of regulatory elements that permit expression in the host cell. In some embodiments, the nucleic acid is an

extrachromosomal plasmid. In other embodiments, the nucleic acid is a chromosomal integration vector that can integrate the nucleotide sequence into the chromosome of the host cell. [0051] Nucleic acids encoding these proteins can be introduced into the host cell by any method known to one of skill in the art without limitation (see, for example, Hinnen et al. (1978) Proc. Natl. Acad. Sci. USA 75:1292-3; Cregg et al. (1985) Mol. Cell. Biol. 5:3376-3385; Goeddel et al. eds, 1990, Methods in Enzymology, vol. 185, Academic Press, Inc., CA; Krieger, 1990, Gene Transfer and Expression— A Laboratory Manual, Stockton Press, NY; Sambrook et ah, 1989, Molecular Cloning— A Laboratory Manual, Cold Spring Harbor Laboratory, NY ; and Ausubel et ah, eds., Current Edition, Current Protocols in Molecular Biology, Greene Publishing

Associates and Wiley Interscience, NY). Exemplary techniques include but are not limited to spheroplasting, electroporation, PEG 1000 mediated transformation, and lithium acetate or lithium chloride mediated transformation.

[0052] The copy number of an enzyme in a host cell may be altered by modifying the

transcription of the gene that encodes the enzyme. This can be achieved for example by modifying the copy number of the nucleotide sequence encoding the enzyme (e.g., by using a higher or lower copy number expression vector comprising the nucleotide sequence, or by introducing additional copies of the nucleotide sequence into the genome of the host cell or by deleting or disrupting the nucleotide sequence in the genome of the host cell), by changing the order of coding sequences on a polycistronic mRNA of an operon or breaking up an operon into individual genes each with its own control elements, or by increasing the strength of the promoter or operator to which the nucleotide sequence is operably linked. Alternatively or in addition, the copy number of an enzyme in a host cell may be altered by modifying the level of translation of an mRNA that encodes the enzyme. This can be achieved for example by modifying the stability of the mRNA, modifying the sequence of the ribosome binding site, modifying the distance or sequence between the ribosome binding site and the start codon of the enzyme coding sequence, modifying the entire intercistronic region located“upstream of’ or adjacent to the 5’ side of the start codon of the enzyme coding region, stabilizing the 3’-end of the mRNA transcript using hairpins and specialized sequences, modifying the codon usage of enzyme, altering expression of rare codon tRNAs used in the biosynthesis of the enzyme, and/or increasing the stability of the enzyme, as, for example, via mutation of its coding sequence.

[0053] The activity of an enzyme in a host cell can be altered in a number of ways, including, but not limited to, expressing a modified form of the enzyme that exhibits increased or decreased solubility in the host cell, expressing an altered form of the enzyme that lacks a domain through which the activity of the enzyme is inhibited, expressing a modified form of the enzyme that has a higher or lower Kcat or a lower or higher Km for the substrate, or expressing an altered form of the enzyme that is more or less affected by feed-back or feed-forward regulation by another molecule in the pathway.

[0054] In some embodiments, a nucleic acid used to genetically modify a host cell comprises one or more selectable markers useful for the selection of transformed host cells and for placing selective pressure on the host cell to maintain the foreign DNA.

[0055] In some embodiments, the selectable marker is an antibiotic resistance marker.

Illustrative examples of antibiotic resistance markers include but are not limited to the BLA, NAT1, PAT, AUR1-C, PDR4, SMR1, CAT , mouse dhfr, HPH, DSDA, KAN ^R , and SH BLE gene products. The BLA gene product from E. coli confers resistance to beta-lactam antibiotics (e.g. , narrow- spectrum cephalosporins, cephamycins, and carbapenems (ertapenem), cefamandole, and cefoperazone) and to all the anti-gram-negative-bacterium penicillins except temocillin; the NAT1 gene product from S. noursei confers resistance to nourseothricin; the PAT gene product from S. viridochromogenes Tu94 confers resistance to bialophos; the A UR1-C gene product from Saccharomyces cerevisiae confers resistance to Auerobasidin A (AbA); the PDR4 gene product confers resistance to cerulenin; the SMR1 gene product confers resistance to sulfometuron methyl; the CAT gene product from Tn9 transposon confers resistance to chloramphenicol; the mouse dhfr gene product confers resistance to methotrexate; the HPH gene product of Klebsiella pneumonia confers resistance to Hygromycin B; the DSDA gene product of E. coli allows cells to grow on plates with D-serine as the sole nitrogen source; the KAN ^R gene of the Tn903 transposon confers resistance to G418; and the SH BLE gene product from Streptoalloteichus Hindus tanus confers resistance to Zeocin (bleomycin). In some embodiments, the antibiotic resistance marker is deleted after the genetically modified host cell disclosed herein is isolated.

[0056] In some embodiments, the selectable marker rescues an auxotrophy (e.g., a nutritional auxotrophy) in the genetically modified microorganism. In such embodiments, a parent microorganism comprises a functional disruption in one or more gene products that function in an amino acid or nucleotide biosynthetic pathway and that when non-functional renders a parent cell incapable of growing in media without supplementation with one or more nutrients. Such gene products include but are not limited to the HIS3, LEU2, LYS1, LYS2, MET15, TRP1, ADE2, and URA3 gene products in yeast. The auxotrophic phenotype can then be rescued by transforming the parent cell with an expression vector or chromosomal integration construct encoding a functional copy of the disrupted gene product, and the genetically modified host cell generated can be selected for based on the loss of the auxotrophic phenotype of the parent cell. Utilization of the URA3, TRP1, and LYS2 genes as selectable markers has a marked advantage because both positive and negative selections are possible. Positive selection is carried out by auxotrophic complementation of the URA3, TRP1, and LYS2 mutations, whereas negative selection is based on specific inhibitors, i.e., 5-fluoro-orotic acid (FOA), 5-fluoroanthranilic acid, and aminoadipic acid (aAA), respectively, that prevent growth of the prototrophic strains but allows growth of the URA3, TRP1, and LYS2 mutants, respectively. In other embodiments, the selectable marker rescues other non-lethal deficiencies or phenotypes that can be identified by a known selection method.

TCA cycle enzymes

[0057] The tricarboxylic acid cycle (TCA), also known as the citric acid cycle (CAC) or Krebs cycle is a series of biochemical reactions that are used to generate energy in the form of ATP through the oxidation of acetyl-CoA derived from carbohydrates, fats, and proteins. The TCA also provides precursors for the production of certain amino acids as well as the reducing agent NADH. In an embodiment of the invention, the rate and yield of isoprenoid production are uncoupled by down regulating the activity of one or more of the enzymes that participate in the TCA cycle either directly or indirectly (for example, by providing carbon to the TCA cycle).

[0058] In some embodiments, the TCA enzyme is citrate synthase also known as citrate condensing enzyme, CoA-acetylating citrate oxaloacetate-lyase, citric-condensing enzyme, citrogenase, oxaloacetate transacetase, CITl, CIT3, (comprising the amino acid sequence NP 015325.1 or NP 014398.1), EC 2.3.3.1, EC 2.3.3.8, and EC 2.3.3.3. Citrate synthase catalyzes the conversion of oxaloacetic acid, acetyl-CoA, and water to citrate and Coenzyme A.

[0059] In some embodiments, the TCA enzyme is aconitate hydratase also known as cis- aconitase, aconitase, ACOl, (comprising the amino acid sequence NP 013407.1), and EC 4.2.1.3. Aconitate hydratase catalyzes the conversion of citrate to isocitrate through a cis- aconitate intermediate.

[0060] In some embodiments, the TCA enzyme is NAD-dependent isocitrate dehydrogenase also known as IDH2, IDH1, EC 1.1.1.42, EC 1.1.1.41, EC 1.1.1.286, and (comprising the amino acid sequence of NP 014779.1 or NP 014361.1). NAD-dependent isocitrate dehydrogenase catalyzes the NAD dependent conversion of isocitrate to 2-oxoglutarate, carbon dioxide, and NADH.

[0061] In some embodiments, the TCA enzyme is 2-ketoglutarate dehydrogenase complex also known as dihydrolipoamide dehydrogenase, KGD2, KGD1, LPD1, EC 1.2.4.2, EC 2.3.1.61, and (comprising the amino acid sequence of NP 010432.3, NP 012141.1, or NP 116635.1). 2- Ketoglutarate dehydrogenase complex catalyzes the NAD dependent conversion of 2- oxoglutarate and Coenzyme A to succinyl-CoA, carbon dioxide, and NADH.

[0062] In some embodiments, the TCA enzyme is succinyl-CoA ligase also known as LSC2, LSC1, EC 6.2.1.4, EC 6.2.1.5, EC 2.8.3.18, and (comprising the amino acid sequence of NP 011670.3 or NP 014785.3). Succinyl-CoA ligase catalyzes the conversion of succinyl-CoA, ADP, and phosphate to succinate, ATP, and Coenzyme A.

[0063] In some embodiments, the TCA enzyme is minor succinate dehydrogenase also known as SDH1, SDH2, SDH3, SDH4, succinate dehydrogenase, EC 1.3.5.4, EC 1.3.5.1, and (comprising the amino acid sequence of NP 012774.1, NP 013059.1, or NP 012781.1). Succinate

dehydrogenase catalyzes the conversion of succinate and ubiquinone to fumarate and ubiquinol.

[0064] In some embodiments, the TCA enzyme is fumarate hydralase also known as FUM1, EC 4.2.1.2, and (comprising the amino acid sequence of NP 015061.1). Fumarate hydralase catalyzes the conversion of fumarate and water to malate.

[0065] In some embodiments, the TCA enzyme is peroxisomal malate dehydrogenase also known as MDH3, mitochondrial malate dehydrogenase, MDH1, cytosolic malate

dehydrogenase, MDH2, EC 1.1.1.37, EC 1.1.5.4, and (comprising the amino acid sequence of NP 010205.1, NP 014515.2, or NP 012838.1). Peroxisomal malate dehydrogenase catalyzes the conversion of malate and NAD to oxaloacetic acid and NADH.

[0066] In some embodiments, the TCA enzyme is pyruvate carboxylase also known as PYC1, PYC2, EC 4.1.1.32, EC. 4.1.1.49, and (comprising the amino acid sequence of NP 011453.1 or NP 09777.1). Pyruvate carboxylase catalyzes the conversion of pyruvate, bicarbonate, and ATP to phosphate, oxaloacetic acid, and ADP.

Futile Cycles

[0067] In an aspect of the invention, yield and productivity of isoprenoids can be uncoupled in host cells producing the isoprenoids by the introduction of one or more futile cycles into the host cell. Futile cycles comprise at least two metabolic cycles or pathways that when run concurrently in opposite directions have no effect other than the dissipation of energy in the form of hydrolysis of ATP. Accordingly, the introduction of one or more futile cycles into the host cell reduces the ATP levels of the cell and thereby uncouple the yield and productivity of isoprenoid production.

[0068] In an embodiment, the futile cycle comprises the over expression of phosphofmctokinase and fructose- 1,6-bisphosphatase. Phosphofmctokinase catalyzes the conversion of D-fmctose-6- phosphate to D-fructose- 1,6-biphosphate. In contrast, fructose- 1,6-biophatase (EC 3.1.3.11) catalyzes the hydrolysis of D-fructose- 1,6-biphosphate to D-fmctose-6-phosphate in a reaction that consumes one molecule of ATP. Accordingly, simultaneous expression of both enzymes results in the dissipation of ATP. See for example, US Patent Application Publications Nos: US20150322461 and US20120088290 both of which are incorporated herein in their entireties.

[0069] In another embodiment, the futile cycle comprises the simultaneous over expression of phosphoenolpyruvate carboxykinase and pyruvate carboxylase. Phosphoenolpyruvate carboxykinase catalyzes the conversion of oxaloacetate into phosphoenolpyruvate whereas pyruvate carboxylase catalyzes the inverse reaction. Each enzyme hydrolyzes one ATP molecule per reaction. However, the net cycle also generates one ATP molecule. Accordingly, each cycle reaction dissipates one net ATP molecule. See for example, US Patent Application Publications Nos: US20150322461 and US20120088290.

ATP Dissipation Enzymes

[0070] In an aspect of the invention, yield and productivity of isoprenoids can be uncoupled in host cells producing the isoprenoids by the expression of an enzyme that dissipates ATP without producing any other physiologic effect.

[0071] In one embodiment, the over expression of the Saccharomyces cerevisae SSB1 gene or fragment thereof in a host cell producing an isoprenoid results in the uncoupling of the yield and productivity of the isoprenoid. The SSB 1 gene encodes a chaperone protein that hydrolyses an ATP molecule as it binds nascent unfolded proteins. Accordingly, over expression of SSB1 or an enzymatically active fragment thereof results in the dissipation of ATP without producing any other physiologic effect. See for example, US Patent Application Publications Nos:

US20150322461 and US20120088290.

[0072] In another embodiment, the over expression of ATP-diphosphohydrolase or fragment thereof in a host cell producing an isoprenoid results in the uncoupling of the yield and productivity of the isoprenoid. ATP-diphosphohydrolases are enzymes that catalyze the hydrolysis of both the b- and g- phosphates of ADP and ATP. Accordingly, over expression of ATP-diphosphohydrolase or an enzymatically active fragment thereof results in the dissipation of ATP without producing any other physiologic effect. See for example, US Patent Application Publications Nos: US20150322461 and US20120088290.

[0073] In another embodiment, the over expression of NADH oxidase; EC 1.6.3.4 (NOX) or a functional fragment thereof in a host cell producing an isoprenoid results in the uncoupling of the yield and productivity of the isoprenoid. NOX reduces NADH to NAD+ by directly transferring hydrogen to O2 without generating ATP. NOX lowers intracellular ATP concentrations by bypassing the native electron transport chain which would otherwise generate ATP upon oxidation of NADH to NAD+. Accordingly, over expression of NOX or a functional fragment thereof results in the dissipation of ATP without producing any other physiologic effect.

[0074] In another embodiment, the over expression of alternative oxidase (AOX) or a functional fragment thereof in a host cell producing an isoprenoid results in the uncoupling of the yield and productivity of the isoprenoid. Electron flow from ubiquinol to AOX resulting in the reduction of 02 to H20, is not coupled to proton transport and therefore reduces the motive force used by ATP synthase to produce ATP. Accordingly, over expression of AOX or a functional fragment thereof results in the dissipation of ATP without producing any other physiologic effect.

ATP Depleting Agents

[0075] In some embodiments, ATP levels within the host cell are lowered by addition of one or more ATP depleting agents. ATP depleting agents are compounds or molecules that are capable of lowering the ATP levels within the host cell when the host cell is cultured in media containing the ATP depleting agent. In some embodiments, the ATP depleting agent is one which uncouples electron transport from ATP generation. In preferred embodiments, the ATP depleting agent is a weak organic acid. Non-limiting illustrative examples of weak organic acids are acetic acid, propionic acid, sorbic acid, and benzoic acid. The host cells can be cultured in media that contains an amount (concentration) of weak organic acid sufficient to lower ATP levels and thereby uncouple yield and productivity of the non-catabolic compound. In some embodiments the amount of weak organic acids is 0.25 mM or more. In particular embodiments, the host cell culture media has at least 0.25 mM, 0.3 mM, 0.35 mM, 0.40 mM, 0.45 mM, 0.5 mM, 0.55 mM, 0.6 mM, 0.65 mM, 0.70 mM, 0.75 mM, 0.80 mM, 0.85 mM, 0.90 mM, 0.95 mM, 1.0 mM, 2.0 mM, 3.0 mM, 4.0 mM, 5.0 mM, 6.0 mM, 7.0 mM, 8.0 mM, 9.0 mM, or 10.0 mM.

MEV Pathway

[0076] In some embodiments, the host cell further comprises one or more heterologous enzymes that function in a biosynthetic pathway for the production of a cytosolic isoprenoid. The production of the elevated level of the cytosolic isoprenoid can be affected through targeted genetic engineering of the host cell. A number of enzymes are known to function in the production of cytosolic isoprenoids or in the utilization of cytosolic acetyl-CoA and its precursors, and any one of these enzymes can be manipulated to change the level of a cytosolic isoprenoid in a host cell.

[0077] In some embodiments, the host cell comprises one or more heterologous enzyme of the MEV pathway. In some embodiments, the host cell comprises a heterologous mevalonate kinase. In other embodiments, the host cell comprises a heterologous HMG-CoA reductase. In some embodiments, the host cell comprises a heterologous IPP isomerase. In some

embodiments, the host cell comprises a heterologous polyprenyl synthase. In some

embodiments, the host cell comprises a heterologous FPP synthase. In some embodiments, the host cell comprises a heterologous terpene synthase. In some embodiments, the host cell comprises a heterologous famesene synthase.

[0078] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can condense two molecules of acetyl-coenzyme A to form

acetoacetyl-CoA, e.g., an acetyl-CoA thiolase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (NC_000913 REGION:

2324131.2325315; Escherichia coli ), (D49362; Paracoccus denitrificans ), and (L20428;

Saccharomyces cerevisiae).

[0079] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can condense acetoacetyl-CoA with another molecule of acetyl-CoA to form 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA), e.g., a HMG-CoA synthase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to:

(NC_001145. complement 19061.20536; Saccharomyces cerevisiae ), (X96617; Saccharomyces cerevisiae ), (X83882 Arabidopsis thaliana ), (AB037907; Kitasatospora griseola ), (BT007302; Homo sapiens), and (NC_002758, Locus tag SAV2546, GenelD 1122571; Staphylococcus aureus).

[0080] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can convert HMG-CoA into mevalonate, e.g., a HMG-CoA reductase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (NM_206548; Drosophila melanogaster), (NC_002758, Locus tag SAV2545, GenelD 1122570; Staphylococcus aureus), (NM_204485; Gallus gallus), (AB015627; Streptomyces sp. KO 3988), (AF542543; Nicotiana attenuata), (AB037907; Kitasatospora griseola), (AX128213, providing the sequence encoding a truncated HMGR; Saccharomyces cerevisiae), and

(NC_001145: complement (115734.118898; Saccharomyces cerevisiae).

[0081] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can convert mevalonate into mevalonate 5-phosphate, e.g., a mevalonate kinase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (L77688; Arabidopsis thaliana), and (X55875; Saccharomyces cerevisiae).

[0082] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can convert mevalonate 5-phosphate into mevalonate 5-pyrophosphate, e.g., a phosphomevalonate kinase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (AF429385; Hevea brasiliensis), (NM_006556; Homo sapiens), and (NC_001145. complement 712315.713670; Saccharomyces cerevisiae).

[0083] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can convert mevalonate 5-pyrophosphate into IPP, e.g., a mevalonate pyrophosphate decarboxylase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (X97557; Saccharomyces cerevisiae), (AF290095;

Enterococcus faecium), and (U49260; Homo sapiens).

[0084] In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding more than one enzyme of the MEV pathway. In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding two enzymes of the MEV pathway. In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding an enzyme that can convert HMG-CoA into mevalonate and an enzyme that can convert mevalonate into mevalonate 5-phosphate. In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding three enzymes of the MEV pathway. In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding four enzymes of the MEV pathway. In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding five enzymes of the MEV pathway. In some embodiments, the host cell comprises one or more heterologous nucleotide sequences encoding six enzymes of the MEV pathway.

[0085] In some embodiments, the host cell produces a Cs isoprenoid. These compounds are derived from one isoprene unit and are also called hemiterpenes. An illustrative example of a hemiterpene is isoprene. In other embodiments, the isoprenoid is a Cio isoprenoid. These compounds are derived from two isoprene units and are also called monoterpenes. Illustrative examples of monoterpenes are limonene, citranellol, geraniol, menthol, perillyl alcohol, linalool, thujone, and myrcene. In other embodiments, the isoprenoid is a C15 isoprenoid. These compounds are derived from three isoprene units and are also called sesquiterpenes. Illustrative examples of sesquiterpenes are periplanone B, gingkolide B, amorphadiene, artemisinin, artemisinic acid, valencene, nootkatone, epi-cedrol, epi-aristolochene, famesol, gossypol, sanonin, periplanone, forskolin, and patchoulol (which is also known as patchouli alcohol). In other embodiments, the isoprenoid is a C20 isoprenoid. These compounds are derived from four isoprene units and also called diterpenes. Illustrative examples of diterpenes are casbene, eleutherobin, paclitaxel, prostratin, pseudopterosin, and taxadiene. In yet other examples, the isoprenoid is a C20+ isoprenoid. These compounds are derived from more than four isoprene units and include: triterpenes (C30 isoprenoid compounds derived from 6 isoprene units) such as arbrusideE, bruceantin, testosterone, progesterone, cortisone, digitoxin, and squalene;

tetraterpenes (C40 isoprenoid compounds derived from 8 isoprenoids) such as b-carotene; and polyterpenes (C40+ isoprenoid compounds derived from more than 8 isoprene units) such as polyisoprene. In some embodiments, the isoprenoid is selected from the group consisting of abietadiene, amorphadiene, carene, a-farnesene, b-famesene, famesol, geraniol, geranylgeraniol, isoprene, linalool, limonene, myrcene, nerolidol, ocimene, patchoulol, b-pinene, sabinene, g- terpinene, terpinolene and valencene. Isoprenoid compounds also include, but are not limited to, carotenoids (such as lycopene, a- and b-carotene, a- and b-cryptoxanthin, bixin, zeaxanthin, astaxanthin, and lutein), steroid compounds, and compounds that are composed of isoprenoids modified by other chemical groups, such as mixed terpene-alkaloids, and coenzyme Q-10. [0086] In some embodiments, the host cell further comprises a heterologous nucleotide sequence encoding an enzyme that can convert IPP generated via the MEV pathway into DMAPP, e.g., an IPP isomerase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (NC_000913, 3031087.3031635; Escherichia coli), and (AF082326;

Haematococcus pluvialis).

[0087] In some embodiments, the host cell further comprises a heterologous nucleotide sequence encoding a polyprenyl synthase that can condense IPP and/or DMAPP molecules to form polyprenyl compounds containing more than five carbons.

[0088] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can condense one molecule of IPP with one molecule of DMAPP to form one molecule of geranyl pyrophosphate (“GPP”), e.g., a GPP synthase. Illustrative examples of nucleotide sequences encoding such an enzyme include, but are not limited to: (AF513111; Abies grandis), (AF513112; Abies grandis), (AF513113 ; Abies grandis),

(AY534686; Antirrhinum majus), (AY534687; Antirrhinum majus), (Y17376; Arabidopsis thaliana), (AE016877, Focus API 1092; Bacillus cereus ATCC 14579), (AJ243739; Citrus sinensis), (AY534745; Clarkia breweri), (AY953508; Ips pini), (DQ286930; Lycopersicon esculentum ), (AF 182828; Mentha x piperita), (AF 182827; Mentha x piperita), (MPI249453; Mentha x piperita), (PZE431697, Focus CAD24425; Paracoccus zeaxanthinifaciens),

(AY866498; Picrorhiza kurrooa), (AY351862; Vitis vinifera), and (AF203881, Focus

AAF 12843; Zymomonas mobilis).

[0089] In some embodiments, the host cell comprises a heterologous nucleotide sequence encoding an enzyme that can condense two molecules of IPP with one molecule of DMAPP, or add a molecule of IPP to a molecule of GPP, to form a molecule of famesyl pyrophosphate (“FPP”), e.g., a FPP synthase. Illustrative examples of nucleotide sequences that encode such an enzyme include, but are not limited to: (ATU80605; Arabidopsis thaliana), (ATHFPS2R;

Arabidopsis thaliana), (AAU36376; Artemisia annua), (AF461050; Bos taurus), (D00694;

Escherichia coli K-12), (AE009951, Focus AAF95523; Fusobacterium nucleatum subsp.

nucleatum ATCC 25586), (GFFPPSGEN; Gibber ella fujikuroi), (CP000009, Focus AAW60034; Gluconobacter oxydans 621H), (AF019892; Helianthus annuus), (HUMFAPS; Homo sapiens), (KFPFPSQCR; Kluyveromyces lactis), (FAU15777; Lupinus albus), (FAU20771; Lupinus albus), (AF309508; Mus musculus), (NCFPPSGEN; Neurospora crassa), (PAFPS1; Parthenium argentatum), (PAFPS2; Parthenium argentatum), (RATFAPS; Rattus norvegicus), (YSCFPP; Saccharomyces cerevisiae ), (D89104; Schizosaccharomyces pombe), (CP000003, Locus AAT87386; Streptococcus pyogenes), (CP000017, Locus AAZ51849; Streptococcus pyogenes), (NC_008022, Locus YP_598856; Streptococcus pyogenes MGAS 10270), (NC_008023, Locus YP_600845; Streptococcus pyogenes MGAS2096), (NC_008024, Locus YP_602832;

Streptococcus pyogenes MGAS 10750), (MZEFPS; Zea mays), (AE000657, Locus AAC06913; Aquifex aeolicus VF5), (NM_202836; Arabidopsis thaliana), (D84432, Locus BAA12575; Bacillus subtilis), (U12678, Locus AAC28894; Bradyrhizobium japonicum USDA 110), (BACFDPS; Geobacillus stearothermophilus), (NC_002940, Locus NP_873754; Haemophilus ducreyi 35000HP), (L42023, Locus AAC23087; Haemophilus influenzae Rd KW20), (J05262; Homo sapiens), (YP_395294; Lactobacillus sakei subsp. sakei 23K), (NC_005823, Locus YP_000273; Leptospira interrogans serovar Copenhageni str. Fiocruz Ll-130), (AB003187; Micrococcus luteus), (NC_002946, Locus YP_208768; Neisseria gonorrhoeae FA 1090), (U00090, Locus AAB91752; Rhizobium sp. NGR234), (J05091; Saccharomyces cerevisae), (CP000031, Locus AAV93568; Silicibacter pomeroyi DSS-3), (AE008481, Locus AAK99890; Streptococcus pneumoniae R6), and (NC_004556, Locus NP 779706; Xylella fastidiosa

Temeculal).

[0090] In some embodiments, the host cell further comprises a heterologous nucleotide sequence encoding an enzyme that can combine IPP and DMAPP or IPP and FPP to form geranylgeranyl pyrophosphate (“GGPP”). Illustrative examples of nucleotide sequences that encode such an enzyme include, but are not limited to: (ATHGERPYRS; Arabidopsis thaliana), (BT005328; Arabidopsis thaliana ), (NM_119845; Arabidopsis thaliana), (NZ_AAJM01000380, Locus ZP_00743052; Bacillus thuringiensis serovar israelensis, ATCC 35646 sql563), (CRGGPPS; Catharanthus roseus), (NZ_AABF02000074, Locus ZP_00144509; Fusobacterium nucleatum subsp. vincentii, ATCC 49256), (GFGGPPSGN; Gibber ella fujikuroi), (AY371321; Ginkgo biloba), (AB055496; Hevea brasiliensis), (AB017971; Homo sapiens), (MCI276129; Mucor circinelloides f lusitanicus), (AB016044; Mus musculus), (AABX01000298, Locus NCU01427; Neurospora crassa), (NCU20940; Neurospora crassa), (NZ_AAKL01000008, Locus

ZP_00943566; Ralstonia solanacearum UW551), (AB 118238; Rattus norvegicus), (SCU31632; Saccharomyces cerevisiae), (AB016095; Synechococcus elongates), (SAGGPS; Sinapis alba), (SSOGDS; Sulfolobus acidocaldarius), (NC_007759, Locus YP_461832; Syntrophus aciditrophicus SB), (NC_006840, Locus YP_204095; Vibrio fischeri ESI 14), (NM_112315; Arabidopsis thaliana ), (ERWCRTE; Pantoea agglomerans), (D90087, Locus BAA14124;

Pantoea ananatis ), (X52291, Locus CAA36538; Rhodobacter capsulatus), (AF195122, Locus AAF24294; Rhodobacter sphaeroides), and (NC_004350, Locus NP_721015; Streptococcus mutans UA159).

[0091] In some embodiments, the host cell further comprises a heterologous nucleotide sequence encoding an enzyme that can modify a polyprenyl to form a hemiterpene, a monoterpene, a sesquiterpene, a diterpene, a triterpene, a tetraterpene, a polyterpene, a steroid compound, a carotenoid, or a modified isoprenoid compound.

[0092] In some embodiments, the heterologous nucleotide encodes a carene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to: (AF461460, REGION 43.1926; Picea abies ) and (AF527416, REGION: 78.1871; Salvia stenophylla).

[0093] In some embodiments, the heterologous nucleotide encodes a geraniol synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to: (AJ457070; Cinnamomum tenuipilum), (AY362553; Ocimum basilicum), (DQ234300; Perilla frutescens strain 1864), (DQ234299; Perilla citriodora strain 1861), (DQ234298; Perilla citriodora strain 4935), and (DQ088667; Perilla citriodora).

[0094] In some embodiments, the heterologous nucleotide encodes a linalool synthase.

Illustrative examples of a suitable nucleotide sequence include, but are not limited to:

(AF497485; Arabidopsis thaliana), (AC002294, Locus AAB71482; Arabidopsis thaliana), (AY059757; Arabidopsis thaliana), (NM_104793; Arabidopsis thaliana), (AF154124; Artemisia annua), (AF067603; Clarkia breweri), (AF067602; Clarkia concinna), (AF067601; Clarkia breweri), (U58314; Clarkia breweri), (AY840091; Lycopersicon esculentum), (DQ263741; Lavandula angustifolia), (AY083653; Mentha citrate), (AY693647; Ocimum basilicum), (XM_463918; Oryza sativa), (AP004078, Locus BAD07605; Oryza sativa), (XM_463918,

Locus XP_463918; Oryza sativa), (AY917193; Perilla citriodora), (AF271259; Perilla frutescens), (AY473623; Picea abies), (DQ195274; Picea sitchensis), and (AF444798; Perilla frutescens var. crispa cultivar No. 79).

[0095] In some embodiments, the heterologous nucleotide encodes a limonene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to: (+)- limonene synthases (AF514287, REGION: 47.1867; Citrus limon) and (AY055214, REGION: 48.1889; Agastache rugosa) and (-)-limonene synthases (DQ195275, REGION: 1.1905; Picea sitchensis), (AF006193, REGION: 73.1986; Abies grandis), and (MHC4SLSP, REGION:

29.1828; Mentha spicata).

[0096] In some embodiments, the heterologous nucleotide encodes a myrcene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to: (U87908; Abies grandis ), (AY 195609; Antirrhinum majus), (AY 195608; Antirrhinum majus),

(NM_127982; Arabidopsis thaliana TPS 10), (NM_113485; Arabidopsis thaliana ATTPS-CIN), (NM_113483; Arabidopsis thaliana ATTPS-CIN), (AF271259; Perillafrutescens), (AY473626; Picea abies ), (AF369919; Picea abies ), and (AJ304839; Quercus ilex).

[0097] In some embodiments, the heterologous nucleotide encodes an ocimene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to:

(AY195607 ; Antirrhinum majus), (AY195609; Antirrhinum majus), (AY195608; Antirrhinum majus), (AK221024; Arabidopsis thaliana), (NM_113485; Arabidopsis thaliana ATTPS-CIN), (NM_113483; Arabidopsis thaliana ATTPS-CIN), (NM_117775; Arabidopsis thaliana

ATTPS03), (NM_001036574; Arabidopsis thaliana ATTPS03), (NM_127982; Arabidopsis thaliana TPS 10), (AB 110642; Citrus unshiu CitMTSL4), and (AY575970; Lotus corniculatus var. japonicus).

[0098] In some embodiments, the heterologous nucleotide encodes an a-pinene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to: (+) a- pinene synthase (AF543530, REGION: 1.1887; Pinus taeda), (-)a-pinene synthase (AF543527, REGION: 32.1921; Pinus taeda), and (+)/(-) a-pinene synthase (AGU87909, REGION: 6111892; Abies grandis).

[0099] In some embodiments, the heterologous nucleotide encodes a b-pinene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to: (-) b-pinene synthases (AF276072, REGION: 1.1749; Artemisia annua) and (AF514288, REGION: 26.1834; Citrus limon).

[0100] In some embodiments, the heterologous nucleotide encodes a sabinene synthase. An illustrative example of a suitable nucleotide sequence includes but is not limited to AF051901, REGION: 26.1798 from Salvia ojficinalis. [0101] In some embodiments, the heterologous nucleotide encodes a g-terpinene synthase.

Illustrative examples of suitable nucleotide sequences include: (AF514286, REGION: 30.1832 from Citrus limon) and (AB 110640, REGION 1.1803 from Citrus unshiu).

[0102] In some embodiments, the heterologous nucleotide encodes a terpinolene synthase.

Illustrative examples of a suitable nucleotide sequence include but is not limited to: (AY693650 from Oscimum basilicum) and (AY906866, REGION: 10.1887 from Pseudotsuga menziesii).

[0103] In some embodiments, the heterologous nucleotide encodes an amorphadiene synthase. An illustrative example of a suitable nucleotide sequence is SEQ ID NO. 37 of U.S. Patent Publication No. 2004/0005678.

[0104] In some embodiments, the heterologous nucleotide encodes a a-farnesene synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to DQ309034 from Pyrus communis cultivar d'Anjou (pear; gene name AFS 1) and AY182241 from Malus domestica (apple; gene AFS1). Pechouus et al., Planta 219(l):84-94 (2004).

[0105] In some embodiments, the heterologous nucleotide encodes a b-farnesene synthase.

Illustrative examples of suitable nucleotide sequences include but is not limited to GenBank accession number AF024615 from Mentha x piperita (peppermint; gene Tspal 1), and AY835398 from Artemisia annua. Picaud et al., Phytochemistry 66(9): 961-967 (2005).

[0106] In some embodiments, the heterologous nucleotide encodes a farnesol synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to GenBank accession number AF529266 from Zea mays and YDR481C from Saccharomyces cerevisiae (gene Pho8). Song, L., Applied Biochemistry and Biotechnology 128:149-158 (2006).

[0107] In some embodiments, the heterologous nucleotide encodes a nerolidol synthase. An illustrative example of a suitable nucleotide sequence includes, but is not limited to AF529266 from Zea mays (maize; gene tpsl).

[0108] In some embodiments, the heterologous nucleotide encodes a patchouliol synthase.

Illustrative examples of suitable nucleotide sequences include, but are not limited to AY508730 REGION: 1.1659 from Pogostemon cablin.

[0109] In some embodiments, the heterologous nucleotide encodes a nootkatone synthase.

Illustrative examples of a suitable nucleotide sequence includes, but is not limited to AF441124 REGION: 1.1647 from Citrus sinensis and AY917195 REGION: 1.1653 from Perilla frutescens . [0110] In some embodiments, the heterologous nucleotide encodes an abietadiene synthase. Illustrative examples of suitable nucleotide sequences include, but are not limited to: (U50768; Abies grandis ) and (AY473621; Picea abies).

Methods of Producing Isoprenoids

[0111] In some embodiments where the genetically modified host cell comprises a heterologous nucleotide sequence encoding an NADH-using HMGR, the genetically modified host cell produces an increased amount of the isoprenoid compound compared to a host cell not comprising a heterologous nucleotide sequence encoding an NADH-using HMGR, but is otherwise genetically identical. In some embodiments, the increased amount is at least 10%, as measured, for example, in grams per liter of cell culture, milligrams per gram of dry cell weight, on a per unit volume of cell culture basis, on a per unit dry cell weight basis, on a per unit volume of cell culture per unit time basis, or on a per unit dry cell weight per unit time basis.

[0112] In some embodiments where the genetically modified host cell comprises both a heterologous nucleotide sequence encoding an ADA and a heterologous nucleotide sequence encoding an NADH-using HMGR, the genetically modified host cell produces an increased amount of the isoprenoid compound compared to: (i) a host cell not comprising a heterologous nucleotide sequence encoding an ADA, but is otherwise genetically identical; (ii) a host cell not comprising a heterologous nucleotide sequence encoding an NADH-using HMGR, but is otherwise genetically identical; or (iii) a host cell not comprising a heterologous nucleotide sequence encoding an ADA or a heterologous sequence encoding an NADH-using HMGR, but is otherwise genetically identical. In some embodiments, the increased amount is at least 10%, as measured, for example, in grams per liter of cell culture, milligrams per gram of dry cell weight, on a per unit volume of cell culture basis, on a per unit dry cell weight basis, on a per unit volume of cell culture per unit time basis, or on a per unit dry cell weight per unit time basis.

[0113] The methods generally involve growing a host cell under suitable conditions in a suitable medium comprising a carbon source. Suitable conditions and suitable media for growing microorganisms are well known in the art. In some embodiments, the carbon source is a monosaccharide (simple sugar), a disaccharide, a polysaccharide, a non-fermentable carbon source, or one or more combinations thereof. Non-limiting examples of suitable

monosaccharides include glucose, galactose, mannose, fructose, ribose, and combinations thereof. Non-limiting examples of suitable disaccharides include sucrose, lactose, maltose, trehalose, cellobiose, and combinations thereof. Non-limiting examples of suitable polysaccharides include starch, glycogen, cellulose, chitin, and combinations thereof. Non limiting examples of suitable non-fermentable carbon sources include acetate and glycerol. In some embodiments, the suitable medium is supplemented with one or more additional agents, such as, for example, an inducing compound (e.g., when one or more nucleotide sequences encoding a gene product are under the control of an inducible promoter), a repressing compound (e.g., when one or more nucleotide sequences encoding a gene product are under the control of a repressible promoter), or a selection agent (e.g., an antibiotic to select for microorganisms comprising the genetic modifications).

EXAMPLES

Example 1: Yield is a Function of Cell-Specific Rates

[0114] Single colonies were inoculated in 15 ml of 2% sucrose, 1% maltose, 2g/L lysine LGM with 50 mM succinate pH 5.0 in a 125 ml flask, then grown at 28 °C, with shaking at 200 r.p.m. to an OD600 between 4 to 9, with residual glucose between 3 to 6 g 1-1.50% glycerol was added to culture to a concentration of 20%, then 1 ml vials of cell suspension were stored at - 80 °C. 1- 2 vials of cells were thawed and grown in media with 3 g 1-lyeast extract, 7 g 1-1NH4H2P04, 1 g 1-1 KH2P04, 0.5 g 1-1 MgS04 7H20, 50 mM succinate pH 5.0, 4% sucrose, 2% maltose, 2g/L lysine and a trace metal and vitamin solution for 24 h, then sub-cultured to OD600 reading of 0.1 in the same media for 24 h. 25 ml of culture was used to inoculate a 0.5-litre fermenter (Sartorius, Germany) with 225 ml fermentation media containing 15 g 1-1 NH4H2P04, 20 g 1-1 total reducing sugar (TRS) from cane syrup (Florida Crystals, West Palm Beach), and a trace metal and vitamin solution. The fermenter temperature was cycled between 30 - 34 °C and pH was maintained at 5.0 with addition of NH40H. In an initial batch phase, the fermenter was aerated at 0.5 volume per volume per minute (VVM) and agitation ramped to maintain 30% dissolved oxygen. After the initial sugar was consumed, the rise in dissolved oxygen triggered feeding of Florida cane syrup (~ 800 g glucose equivalents, (also known as total reducing sugars (TRS)) per litre) at 10 g TRS per litre per hour in pulses of 10 g TRS per litre doses. Between pulses, the feed rate was lowered to 1-5 g TRS per litre per hour. The high feed rate resumes when the dissolved oxygen spikes, indicating the exhaustion of residual carbon; the high feed rate ends after a set amount of sugar is added. As cell density increased, dissolved oxygen was allowed to reach 0%, and the pulse dose was increased to 50 g TRS per litre. Oxygen transfer rate was maintained at particular rates by adjusting agitation as volume increased. Hereon, feedrate was adjusted dynamically to meet demand using an algorithm (the feedrate algorithm) that alternates between a high feedrate and low feedrate. During the low feedrate, cells consume sugar and any overflow metabolites accumulated during the high feedrate. A rise in dissolved oxygen then triggers the high feedrate to resume. The length of time spent in the low feedrate reflects the extent to which cells were over- or under-fed in the prior high feedrate pulse; this information is monitored and used to tune the high feedrate up or down, keeping the low feedrate within a defined range. Over time, feed rate matches sugar demand from cells. The feedrate algorithm ensures minimal net accumulation of fermentation products other than famesene, biomass, and C02. The process continued for 8-13 days. The fermentation tank undergoes fill and draw. Accumulated broth was removed daily and assayed for biomass and famesene concentration. A concentrated solution of NH4H2P04, trace metals and vitamins was added periodically to maintain steady state concentrations.

[0115] Oxygen delivery is typically between 100 and 120 mmol/L/hr, also referred as oxygen transfer rate or OTR. Different oxygen transfer rates (20, 110, 180, 225) during fermentation runs were achieved by combination of variable agitation rate, air flow and feed rate. Once the peak biomass levels reached (grams dry cell weight or gDCW) in the production phase, cells experience microaerobic conditions and the dissolved oxygen (d02) is nearly zero; in other words, the oxygen uptake rate (OUR) is then equal to the oxygen transfer rate (OTR).

[0116] Fermentation yield: Famesene yield (Ysp) is calculated as weight/weight amount of famesene produced divided by the amount of total reducing sugars (TRS, or glucose equivalents) added to a fermentor. Grams of famesene produced divided by grams of total reduced sugar added, expressed as a percentage, also referred simply as‘yield’. gDW or gDCW is referred as grams of dry weight, a measure of cellular biomass or amount of cells in the fermentor.

[0117] Specific oxygen uptake rate (q02) is the specific rate of oxygen consumption by the biomass in the fermentor expressed as mmol/02/gDCW/h. Also known as specific oxygen utilization rate (sOUR).

[0118] Specific sugar uptake rate (qS) is the specific rate of sugar consumption by biomass in the fermenter expressed as mmol/TRS/gDCW/h. [0119] Famesene was quantitated as previously described (Sandoval, C. M. et al. (2014) Metab Eng vol. 25, pp. 215-226).

[0120] Yield and productivity are two major cost drivers for any biomanufacturing process. For the biomanufacturing of non-catabolic products, the cell- specific rate of production (qp, in units of mol product / gram dry cell weight / hr) and Yield (Yp _/ s , or rate of product formation / rate of sugar consumption) are frequently a function of other cellular rates such as specific growth rate (1/hour) or (hour ¹ ), qs (mol sugar consumed / gram dry cell weight / hr) or qo2 (mol 02 consumed / gram dry weight /hr). Design of fermentation processes that achieve the optimal combination of yield and productivity for lowest cost production requires close characterization of this relationship.

[0121] For the production of isoprenoids, we have observed that yield (Yp _/ s) is consistently anticorrelated with cell-specific rates of oxygen and sugar uptake rates, and by extension, volumetric productivity. In other words, the faster oxygen and sugar is taken up by our cells, the lower the isoprenoid yield. We refer to this relationship as“rate-yield coupling.” FIG.l through FIG.3 show this phenomenon for a single famesene producing strain grown under different oxygen transfer rates (the rate at which oxygen is delivered to the bioreactor and taken up by cells). Yield goes down as cell specific sugar uptake, oxygen uptake, and productivity increase. FIG. 4 shows that using the feedrate algorithm, sugar uptake rate is directly proportional to oxygen uptake rate.

[0122] The anticorrelation between yield and productivity shown in FIG. 3 is problematic for industrial production of non-catabolic compounds - such as the isoprenoid famesene - because productivity and yield are the two key cost drivers for the generation of non-catabolic compounds. When rate-yield coupling exists, any attempt to increase volumetric productivity by increasing the rate of oxygen transfer and/or sugar transfer to cells results in a concomitant decrease in yield, negating the cost benefit of increased productivity. If rate-yield coupling could be eliminated, then it would be possible to simultaneously achieve high yield and high productivity.

Example 2: Computational Modeling and Metabolomics Measurements Suggest ATP may Drive Coupling Between Rate and Yield [0123] To evaluate the possible mechanistic cause of rate yield coupling, we analyzed the absolute concentrations of central metabolites from glycolysis, the TCA cycle, the pentose phosphate pathway, and the isoprenoid pathway in fermentation samples from a single strain run at either low OTR (30 mmol C /L/hr) or high OTR (180 mmol C /L/hr). Sampling from bioreactors was performed using a rapid sampling protocol wherein a sample is immediately quenched upon removal from the tank, in order to capture the metabolic state. While the majority of central metabolites were measured at similar absolute concentrations in the low and high OTR conditions, two metabolites (isocitrate and alpha ketoglutarate) stood out as having very different concentrations (See FIG. 5). Interestingly, these two metabolites represent consecutive steps in the TCA cycle: isocitrate can be converted to alpha ketoglutarate through the action of the enzyme isocitrate dehydrogenase. We observed that isocitrate concentrations decrease and alpha ketoglutarate concentrations increased in the high OTR condition relative to the low OTR condition, suggesting that the flux rate through this step relative to others increases with the high OTR condition.

[0124] Previously it was established that excess ATP produced in catabolic pathways can be detrimental to rates and yields of a bioprocess. In terms of rates, this is because elevated ATP concentrations can be inhibitory to glycolysis. However, excess ATP can also decrease the yield of a bioprocess, since the excess ATP can drive formation of biomass, which then acts as a carbon sink that reduces product yield. An approximate stoichiometry for this is 1.5 mol excess ATP is removed or cleared from the system for every mol of Biomass formed.

[0125] To determine whether higher flux through the TCA cycle at high OTRs could account for the lower farnesene yields, we developed a novel computational modeling framework (see FIG. 6) for evaluating the impact of ATP formation, based on a comprehensive genome-scale metabolic model. Whereas the typical genome scale model enables excess ATP to be

hydrolyzed or“wasted” through futile reactions that may not happen in vivo , our new model was built such that a small amount of ATP to be apportioned to non-growth associated maintenance (or NGAM, assumed to be 0.4 mmol gDW-1 h- 1.), while the remainder is strictly coupled to biomass formation, as has been commonly empirically observed. No futile cycles were permitted, and we assumed 1.5 mol ATP needed per mol Biomass formed.

[0126] Using this model, we examined the potential impact of the tricarboxylic acid cycle (TCA) flux on isoprenoid in silico. Our modeling demonstrated that an increase in TCA cycle flux from 20 to 180 OTR could indeed account for the rate-yield coupling effect observed. In other words, the rate-yield coupling effect could be explained by assuming that ATP produced from the TCA cycle at higher rates was“sunk” into biomass at a ratio of 1.5 mol ATP per mol biomass (assuming a conservative estimates of -5.667 mol ATP / mol PYR that enters the TCA cycle.

Properties of relevant simulations plotted above to come to this conclusion:

• Simulations were excluded that were not within the 95% confidence interval (or Cl) computed via 13C analysis for the proportion qS going through PTA at 180 OTR (*we assume it doesn’t' change wildly with OTR).

• TCA cycle activity up to 25% of incoming qS. Our best estimate for TCA cycle flux as a proportion of incoming qS at 180 OTR is - 12% (around where qS - 1.4).

[0127] Materials and methods for Example 2.

[0128] Rapid sampling and Absolute quantitation of metabolite concentrations.

[0129] Definitions: 13C IDMS means Carbon- 13 isotope dilution mass spectrometry; MSTFA means /V-Tri methyl si lyl-/V-mcthyl trifluoroacetamide; and MRM mode means Multiple reaction monitoring mode.

[0130] Absolute intracellular concentrations of isocitrate and a-ketoglutarate were obtained in accordance with the procedures originally described by Canelas and Wahl. Briefly, tank broth is sampled into -80°C methanol using a custom-built rapid sampling device, vortexed and weighted. The biomass is poured into a fast filtering apparatus and washed with 100% -80°C methanol. The filtrate is added to a 50 mL centrifuge tube containing 30 mL of 75% v/v ethanol and 200pL of‘13C IDMS’ internal standard extract. The mixture is boiled at 95°C for 3 minutes then placed back on dry ice. The extract tubes are then evaporated to dryness in a CentriVap and resuspended in 600 pL of water. The water is filtered and dried again via lyophilization prior to ‘MSTFA’ derivatization and analysis.

[0131] Absolute intracellular quantification of isocitrate and a-ketoglutarate was achieved using an Agilent 7000 triple quadrupole GC/MS in‘MRM mode’ comparing signal ratio from sample extract to authentic standards (Sigma). Absolute concentrations were normalized to dry cell weight (measured at time of extraction).

[0132] Modeling:

[0133] A genome-scale model representing yeast metabolism was produced following standard procedures. All reactions from the publically available reconstruction ΪT0977 were incorporated into the starting model. In addition, we added a“second generation” Famesene pathway including 6 reactions as follows:

^■ Alternate HMG-CoA reductase (or NADH HMGR)

o 's_3_hydroxy_3_methylglutaryl_coa_c + 2.0 nadh_c <=> 2.0 nad_c + r_mevalonate_c + coenzyme_a_c',

^■ Acetaldehyde dehydrogenase, acetylating (or ADA)

o 'nad_c + coenzyme_a_c + acetaldehyde_c— > acetyl_coa_c + nadh_c',

^■ Phosphoketolase acting on F6P (or PK-f6p)

o 'phosphate_c + beta_d_fructofuranose_6_phosphate_c— > h2o_c +

acetyl_phosphate_c + d_erythrose_4_phosphate_c',

^■ Phosphoketolase acting on X5P (or PK-x5p)

o 'phosphate_c + d_xylulose_5_phosphate_c— > h2o_c + acetyl_phosphate_c + d_glyceraldehyde_3_phosphate_c'

^■ Phosphotransacetylase (or PTA)

o 'coenzyme_a_c + acetyl_phosphate_c <=> phosphate_c + acetyl_coa_c',

^■ Farnesene Synthase (or FS)

o '_2_trans6_trans_famesyl_diphosphate_c— > diphosphate_c + beta_farnesene_c',

[0134] The model was verified to produce Farnesene with a maximum theoretical yield of -29.5% (g Farnesene / g Sugar). During this process we deactivated the following reactions to prevent uncontrolled free cycling among NAD/NADH/NADP/NADPH:

^■ nadp_specific_glutamate_dehydrogenase_ 1

^■ methylenetetrahydrofolate_dehydrogenase_nad_

[0135] A reaction named“atp_drain_flux_for_constant_maintanence_requirements ” which simply represents the hydrolysis of ATP to ADP was added and constrained to a constant value of 0.4 mmol gDW-1 h-1.

[0136] Default environmental conditions were set by applying a custom function written in Python to work with the model objects of the cobrapy module version 0.3.2. The growth media was set to "glucose_aerobic_minimal" and functionally this allowed for glucose uptake at the rate of 1 mmol gDW-1 h-1 and unlimited o2, nh3, phosphate, sulfate, and h2o uptake.

[0137] Next a set of reactions referred to as the CORE farnesene biosynthetic pathway was defined. These reactions included: 'acetyl_coa_acetyltransferase'; 'glucokinase_glkl'; 'inorganic_pyrophosphatase'; 'dimethylallyltranstransferase';

'atp_drain_flux_for_constant_maintanence_requirements ; geranyltranstransferase' ;

'phosphomevalonate_kinase'; 'hydroxymethylglutaryl_coa_synthase';

'isopentenyl_diphosphate_delta_isomerase'; 'diphosphomevalonate_decarboxylase';

'galactosejxansporter'; 'farnesene_synthase'; 'exchange_of_betafarnesene_c';

'exchange_of_phosphate_e' ; 'exchange_of_alphadgluco se_e' ; 'exchange_of_h2o_e' ;

'gluco se_6_pho sphate_isomerase ; pho sphofructokinase_ G ; 'fructo se_bispho sphate_aldolase' ; 'triosephosphate_isomerase' ; 'enolase_ 1 ' ; 'transport_of_h2o_extracellular' ;

'phosphoglycerate_kinase' ; 'phosphoglycerate_mutase_ 1_ 1 ' ; 'pyruvate_kinase_ 1 ' ;

'pyruvate_decarboxylase_isozyme_l'; 'acetaldehyde_dehydrogenase_acetylating_';

'_3_hydroxy_3_methylglutaryl_coenzyme_a_reductase_ 1 ' ;

'_3_hydroxy_3_methylglutaryl_coenzyme_a_reductase_l_NADH'; 'mevalonate_kinase';

'exchange_of_co2_e'; 'phosphoketolase_f6p'; 'phosphoketolase_x5p'; 'phosphotransacetylase'; 'transaldolase'; 'transketolase_l'; 'glucose_6_phosphate_l_dehydrogenase';

'probable_6_phosphogluconolactonase_ 1 ' ;

'_6_phosphogluconate_dehydrogenase_decarboxy lating_ 1 ' ; 'transketolase_ 1_ 1 ' ;

'ribose_5_phosphate_isomerase'; 'ribulose_phosphate_3_epimerase';

'transport_of_carbon_dioxide_extracellular' ; and

'glyceraldehyde_3_phosphate_dehydrogenase_l'.

[0138] The following reaction, which represents the Alternate HMG-CoA reductase (or NADH HMGR) was constrained to have 0 flux. This is because evidence suggests the majority of flux is carried by the native Sc. HMGR that uses NADPH vs. NADH:

'_3_hydroxy_3_methylglutaryl_coenzyme_a_reductase_ 1 _NADH'

[0139] No flux was permitted to occur outside of these core reactions.

[0140] As flux is not permitted to occur outside the core reactions, any excess NADH would make the simulation unfeasible. So a reaction was added to allow extra NADH produced to convert to ATP (at an assumed stoichiometry of 1:1):

[0141] We also permitted controlled free cycling among NAD/NADH/NADP/NADPH with the udhA reaction [0142] NADPH + NAD <=> NADP + NADH as this is a heterologous enzyme present in our top famesene producers. This means excess NADPH can be converted to NADH and then to ATP using the reactions previously described.

[0143] We also introduced a reaction simulating loss of Pyruvate to the TCA cycle.

'PYR_leak_to_TCA’ :

PYR + 5.6667 ADP + 5.6667 Pi = 3 C02 + 5.6667 ATP + 5.6667 H20

[0144] Finally, we added yet another reaction that hydrolyzes ATP called the

“CUSTOM_NGAM” reaction:

[0145] Once this constrained model was fully constructed it, we adjoined it to ANOTHER copy of the same yeast genome-scale metabolic model, but in this copy all reactions are unconstrained with one critical exception. Importantly, flux through the CUSTOM_NGAM reaction in the first model (the only possible drain of ATP, and NADH converted to ATP for that matter) was strictly coupled to the biomass formation reaction in the second (generally unconstrained) model. The default biomass reaction (in second model) was set to "biomass_1060_bk>mass". Flux through this reaction is reported as mu, or growth rate, with units 1/h. See FIG. 6 for a visual depiction of how the models are setup and interact. Both models SHARE the allotted glucose uptake maximum of 1 mmol gDW-1 h-1, so if there is excess energy produced by the first model it will come at the cost of having to send some sugar to the second model to be“sunk” into growth (biomass formation).

[0146] The coupling constraint was as follows:

Growth in (unconstrained) model > CUSTOM_NGAM*

( 1./ 1000. ) * ( 1./mol_atp_per_c_mol_biomas s) * 12.0107 * 2

Where :

^■ Growth in (unconstrained) model has units: h-1

^■ CUS T OM_N GAM has units: mmol ATP gDW-1 h-1

^■ The“(1./1000.)” term converts from mmol to mol, so after application of this term we have units: mol ATP gDW-1 h-1.

mol_atp_per_c_mol_biomass = 1.4647 (unitless), so after application of this term we have units: cmol biomass gDW-1 h-1. ^■ 12.0107 is the atomic mass of Carbon (12.0107 g / 1 cmol), so after application of this term we have units: g Carbon gDW-1 h-1.

^■ The number“2” is included as the final term, since approximately 2 gDCW can be

composed from 1 g Carbon, assuming the carbon content of a cell is about 50% of the dry weight (PMID 10482783). After application of this term, we have gDW gDW-1 h-1, and the gDW cancels leaving h-1 (matching the units on the left side of the equation for mu or growth rate).

To generate the final results, Famesene yield in the first (constrained) model was optimized over the following parameter values for qS:

qS (20 to 180 OTR in equal intervals): [ 0.5 , 0.92, 1.34, 1.76, 2.18, 2.6 ]

[0147] At each qS, we simulate all possible flux splits to the TCA cycle (from 0 flux into the TCA cycle up to the maximum) (see FIG. 7). To generate additional variation, we also simulate the unknown phosphotransacetylase (2 ^nd generation famesene pathway) flux from 0 to the maximum at each fixed qS and TCA cycle flux. As shown in FIG. 7, the experimental data demonstrates that the flux through the TCA cycle (as a fraction of qS) must increase.

Example 3. Futile ATP Burning Increases Product Yield Relative to Biomass

[0148] We hypothesized that excess ATP may affect rate-yield coupling, if biomass is a preferred sink for excess ATP. ATP expenditure for cell maintenance is constant regardless of specific rate of ATP production. Therefore, at low cell-specific rates, proportionately less ATP is available for biomass or non-catabolic compound production, whereas at high cell-specific rates, proportionately more ATP is available to spend on biomass or non-catabolic compound. If the most efficient way of expending excess ATP is for the cell to sink it into biomass, less carbon would be available for non-catabolic compound production and yield would decrease at high cell-specific rates. Conversely, at low cell-specific rates, there is less excess ATP to force biomass formation and proportionately more carbons can be shunted into non-catabolic compound production. This hypothesis predicts that reducing ATP levels to eliminate ATP excess that would otherwise be sunk into biomass would allow a more favorable partition of carbon into non-catabolic compounds like famesene. [0149] Benzoic acid can be used to deplete ATP in the cell by forcing the cells to use ATP to pump out excess protons that are moved into the cytoplasm by the benzoyl cation. A farnesene- producing strain was treated with different concentrations of benzoic acid and the effect on specific sugar uptake rate (qS), specific farnesene production rate (qP), and specific growth rate (mu) was measured. The yield (qP/qS) was calculated from measured specific sugar uptake rate and specific farnesene production rate.

[0150] The effect of benzoic acid is shown in FIG. 8. Interestingly, while specific growth rates decreased linearly with increasing concentrations of benzoic acid, specific productivity did not change and even increased a little at intermediate concentrations. Increased benzoic acid concentrations (and associated increased ATP-wasting) was associated with more favorable partitioning of carbons into farnesene as reflected in the increased calculated yields (qP/qS).

This data was intriguing and caused us to further directly measure rate-yield coupling in a strain that had been genetically modified to reduce ATP production.

Methods for Example 3.

[0151] Single colonies were grown on CSM agar plates, then picked into sterile 96-well microtiter plates (1.1 mL working volume Axygen) containing 360ul of defined liquid growth medium (LGM; as referenced in Example 1, Westfall et al, 2012) with 50mM succinate (pH 5.0) and 2% sucrose + 1% maltose + 2g/L lysine grown for 72 h at 28 C. 14uL was sub- cultured in 360uL of fresh defined LGM with 50mM succinate (pH 5.0) with the indicated amounts of benzoic acid, then grown for 72 h at 33.5 C. Sucrose is spiked to 8% final concentration at day 3, then after 6 hours of incubation the culture (at early log phase) is diluted 26-fold into the production plate containing 8% sucrose and different concentrations of benzoic acid, thus avoiding a lag phase. Taking measurements of farnesene, biomass, and total residual sugar during log phase growth (at two times, T1 and T2) allows for the determination of specific productivity, growth rate, and specific sugar uptake rate. Farnesene is measured using full- well extraction with isopropanol, and quantified by UV absorbance at 220 nm with reference to a standard curve. Biomass was measured by assaying the fluorescent signal of intracellular tryptophan with an excitation wavelength of 290 nm and detection at 350 nm (UVOD). The relationship between this tryptophan signal and actual biomass is strain-dependent, and is determined empirically for each strain assayed. We do this by measuring both UVOD and biomass absorbance (OD) just prior to the start of production plates; this gives a per-well OD/UVOD conversion factor that is then used to convert UVOD signal at the end of production runs back to biomass. Another conversion we must make in determining biomass is from optical density (OD) to dry cell weight; this is also determined empirically. To eliminate any contribution of farnesene emulsion to the OD signal, cultures were diluted in a solution of 20% (v/v) PEG 20, 20% (v/v) ethanol, 2% (v/v) Triton X-114. Growth rates were determined by applying a linear regression to LN (OD) vs time. Total reducing sugars was measured by using an enzymatic determination of sucrose, fructose and glucose with an output of NADH absorption read at 340nM, as described in various commercial kits, as such those sold by Sigma Aldrich.

[0152] Benzoic acid, acetic acid, sorbic acid, lactic acid, and propionic acid can all used to decrease biomass yields/cause ATP-wasting (when added at different levels). Many other carboxylic acids should be capable of reducing biomass yield (or causing ATP-wasting). The degree of ATP-wasting is generally related to the pKa of the acid and the octanol-water partitioning coefficient (logP) which both influence the permeability of the molecule across the membrane. At low extracellular pH, weak acids should occur predominantly in the un dissociated form, which has relatively high membrane permeability. After entry into the cell via passive diffusion, the higher pH of the cytosol causes dissociation of the acid, thus acidifying the cell and triggering the ATP-dependent efflux of protons. Consequently, weak acids can cause, at the very least, a transient reduction of intracellular ATP levels. At high concentrations, ATP exhaustion, acidification of the cytoplasm and dissipation of the proton-motive force may occur. This‘weak-acid uncoupling’ mechanism is customarily cited as the major mechanism underlying weak organic acid toxicity. Examples of weak acids that can be used to deplete cellular ATP levels are shown in Table 1 below.

Table 1. Examples of weak organic acids that can be used to deplete cellular ATP levels. The concentrations required to reduce the biomass yield to 50% of the reference condition (YRC50) and the predicted concentration of un-dissociated acid at pH 5.0 are indicated along with the most commonly cited pKa and partition coefficient.

Example 4. Genetic Modifications that Lower ATP Levels Reduce Rate- Yield Coupling in Tanks

[0153] A famesene -producing strain was modified by engineering overexpression of NOX (NADH oxidase) under the TDH3 promoter. The oxidation of NADH by NOX prevents NADH from being used as an electron donor for ATP synthesis by ATP synthase in the mitochondria. Thus, overexpression of NOX decreases intracellular ATP levels. This strain, Y31655, together with control strains (Y21901, Y22021) without the NOX engineering, were separately run in tanks set at different OTRs. The specific OUR (q02) is a function of the OTR divided by the total biomass present in the tank. The rate-yield coupling effect was measured by plotting the product yield against the specific oxygen uptake rate from each tank condition. In FIG. 9, the relationship between yield and specific rate is shown in two different shades of gray for the control strains, and in black for the strain overexpressing NOX. The rate-yield coupling slope was significantly reduced (by half) in the NOX-overexpressing strain relative to the controls. This experiment shows that ATP levels affects rate-yield coupling; reduced intracellular ATP levels is associated with reduced coupling between yield and specific rate. This is consistent with our hypothesis that rate-yield coupling is driven by excess ATP that is sunk into biomass, drawing carbons away from farnesene production.

Materials and methods

[0154] See Example 1 for details on bioreactor conditions.

[0155] The coding sequence of NADH Oxidase (defined as the nucleotide sequences spanning the start codon to the stop codon of the NADH Oxidase gene) from Lactococcus lactis was fused to the native S. cerevisiae TDH3 promoter at its 5’ end, and to the terminator of the native S. cerevisiae TDH3 gene at its 3’ end. The TDH3 promoter was defined to be the nucleotide sequence -830 bp immediately upstream of the TDH3 start codon. The TDH3 terminator was defined to be the nucleotide sequence ~300bp immediately downstream of the stop codon of the TDH3 gene. The TDH3 promoter-NOX-TDH3 terminator construct was integrated into the native GAS4 locus using flanking homology sequences, approximately 500 bp upstream and 500 bp downstream of the GAS4 gene, in accordance with standard yeast molecular genetic techniques.

Example 5. Reducing Flux to TCA Cycle Reduces Rate- Yield Coupling

[0156] To determine the effect on rate-yield coupling of reduced flux to the TCA cycle, which is a major source of electrons for ATP synthesis in the mitochondria, strains were made in which PYC1 or CITl were down-regulated, either singly or in combination. Pycl converts cytoplasmic pyruvate into oxaloacetate, which can be transported into the mitochondria, entering the TCA cycle. Citl is the rate-limiting enzyme of the TCA cycle. Down-regulation of both of these enzymes should significantly reduce carbon flux into the TCA cycle and reduce ATP production by ATP synthase in the mitochondria. We engineered down-regulation of CITl or PYC1 by replacing their native promoters with synthetic promoters that are active in the presence of maltose (such as during seed build conditions) but are inactive in the absence of maltose (such as during production conditions). We measured rate-yield coupling in the following farnesene- producing strains Y27662 (engineered to down-regulate PYC1), Y39666 (engineered to down- regulate CITl), Y29438 (engineered to down-regulate both CITl and PYC1) and non-engineered control Y21601. Strikingly, the strain in which both PYC1 and CITl were down-regulated (Y29438), rate appears almost completely decoupled from yield. This observation proves that rate can be decoupled from yield, and identifies a solution to construct strains capable of maintaining both high yield and high specific rates that is compatible with manufacturing at scale. This solution is the culmination of observations from modeling and experimental evidence (all detailed above) that pointed to excess ATP as a major driver of the fate of carbons diverted into biosynthetic reactions.

Materials and methods: Same as for NOX experiment.

[0157] PYC1 and/or CITl down-regulation was achieved by replacing the native promoters of each gene with synthetic promoters that turned off in the absence of maltose, using standard yeast molecular genetic techniques for replacing or inserting DNA sequences in the yeast genome using the host’s native homologous recombination machinery.

[0158] All publications, patents and patent applications cited in this specification are herein incorporated by reference as if each individual publication or patent application were specifically and individually indicated to be incorporated by reference. Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity of understanding, it will be readily apparent to those of ordinary skill in the art in light of the teachings of this invention that certain changes and modifications may be made thereto without departing from the spirit or scope of the appended claims.