Fatty acid productivity

Abstract

The present disclosure relates to an engineered microbe capable of improved productivity of fatty acid or fatty acid derivative. An NAD.sup.+-dependent 3-oxoacyl-ACP reductase or NAD.sup.+-dependent 3-oxoacyl-CoA reductase replaces or supplements the native NADP.sup.+-dependent 3-oxoacyl-ACP reductase so as to utilize the higher availability of NAD.sup.+ rather than NADP.sup.+ in the cell. Higher production, yield and titer of fatty acids are therefore obtained. Such microbes can be combined with other mutations to further improve yield of fatty acids or fatty acid derivatives.

Claims

1. An engineered FASII bacteria with improved productivity of a fatty acid or fatty acid derivative, said FASII bacteria comprising type II fatty acid synthesis enzymes (FASII) and a) an overexpressed NAD+-dependent 3-oxoacyl-ACP reductase or an overexpressed NAD.sup.+-dependent 3-oxoacyl-CoA reductase that replaces or supplements a native NADP.sup.+-dependent 3-oxoacyl-ACP reductase or 3-oxoacyl-CoA reductase (respectively); and b) one or more overexpressed acyl-ACP thioesterases (TE) wherein said NAD+-dependent 3-oxoacyl-ACP reductase has the amino acid sequence of SEQ ID NO. 2.

2. The engineered microbe of claim 1, said FASII bacteria further comprising reduced activity of one or more enzymes selected from beta-oxidation cycle enzymes, acetate synthesis enzymes, lactate synthesis enzymes, formate synthesis enzymes, ethanol synthesis enzymes, glycolytic enzymes or tricarboxylic acid (TCA) cycle enzymes.

3. The engineered FASII bacteria of claim 1, wherein said NAD+-dependent 3-oxoacyl-ACP reductase or NAD.sup.+-dependent 3-oxoacyl-CoA reductase is from Mycobacterium tuberculosis.

4. The engineered FASII bacteria of claim 1, said engineered FASII bacteria further comprising at least one downregulated or disrupted gene selected from one or more of fadD, fabG, sth and pntAB.

5. The engineered FASII bacteria of claim 4, wherein the downregulated or disrupted gene is achieved by using anti-sense RNA (asRNA).

6. The engineered FASII bacteria of claim 1, wherein i) at least one protein from the tricarboxylic acid cycle is reduced, or ii) at least one protein from glycolysis is reduced, or both i) and ii) are reduced.

7. A method of making fatty acids, comprising growing engineered FASII bacteria of claim 1 in a nutrient broth for a time sufficient to make fatty acids, and isolating said fatty acids.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1A is a simplified diagram of central aerobic metabolic pathway and the fatty acid synthesis pathway of E. coli, including the newly added NAD.sup.+-dependent 3-oxoacyl-[acyl-carrier-protein] (ACP) reductase (underlined).

(2) FIG. 1B is a simplified diagram of typical fatty acid synthesis pathway of E. coli.

(3) FIG. 1C provides each of the enzymatic reactions used in fatty acid synthesis of a FASII microbe, the two bolded enzymes using NADPH and being replaced or supplemented herein by an NADH-dependent enzyme.

(4) FIG. 2 is the schematic diagram of the plasmid construct pXZ18G1 that carries an acyl-ACP thioesterase from Ricinus communis (accession no.: XM002515518) and a NAD.sup.+-dependent 3-oxoacyl-ACP reductase from Mycobacterium tuberculosis (accession no.: Rv0242c). Abbreviations: RBS, ribosomal binding site; rcTE, acyl-ACP thioesterase from R. communis; MtG1, an NAD.sup.+-dependent 3-oxoacyl-ACP reductase (FabG4) from M. tuberculosis; lacI, regulator gene of trc promoter system; AmpR, ampicillin resistant gene; T1 terminator and T2 terminator, transcriptional terminator of rrnB; pBR322 origin, origin of replication.

(5) FIG. 3A is a schematic diagram of the plasmid construct pXZ18G2 that carries an acyl-ACP thioesterase from Ricinus communis (accession no.: XM002515518) and a NAD+-dependent 3-oxoacyl-ACP reductase from Mycobacterium tuberculosis (accession no.: Rv0242c). Abbreviations: RBS, ribosomal binding site; rcTE, acyl-ACP thioesterase from R. communis; MtG2, a NAD.sup.+-dependent 3-oxoacyl-ACP reductase (FabG4) from M. tuberculosis with an omission of the first 16 amino acids; lacI, regulator gene of trc promoter system; AmpR, ampicillin resistant gene; T1 terminator and T2 terminator, transcriptional terminator of rrnB; pBR322 origin, origin of replication.

(6) FIG. 3B is a schematic diagram of the plasmid construct pXZ18DG2 which is derived from the plasmid pXZ18G2 by adding an additional copy of mtG2. Similar to pXZ18G2, the plasmid also carries an acyl-ACP thioesterase from Ricinus communis (accession number: XM002515518) (Zhang et al., 2011) and a NAD+ dependent 3-oxoacyl-ACP reductase from Mycobacterium tuberculosis (accession number: Rv0242c). Abbreviations: RBS, ribosomal binding site; rcTE, acyl-ACP thioesterase from R. communis; mtG2, a NAD+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis with an omission of the first 16 amino acids; lacI, regulator gene of trc promoter system; AmpR, ampicillin resistant gene; T1 terminator and T2 terminator, transcriptional terminator of rrnB; pBR322 origin, origin of replication.

(7) FIG. 4: SEQ ID NO: 1 (XM002515518) rice TE.

(8) FIG. 5: SEQ ID NO:2 (Rv0242c) Mycobacterium FabG4.

(9) FIG. 6: SEQ ID NO:3 (truncated Rv0242c) Mycobacterium truncated FabG4.

(10) FIG. 7: Simplified overview of the metabolites and representative pathways in microalgal lipid biosynthesis.

(11) FIG. 8A-B: Effect of higher expression of mt FabG2 on fatty acid production. Plasmid pWL1Tg2 carries an acyl-ACP thioesterase and truncated fabG from Mycobacterium tuberculosis under a constitutive promoter system. Plasmid pXZ18DG2 carries an acyl-ACP thioesterase and two copies of truncated fabGs from Mycobacterium tuberculosis under an inducible trc promoter system.

(12) FIG. 9: Effect of co-overexpression of mt FabG2a and ec FabZ on fatty acid production. Plasmid pWL1TG2 carries an acyl-ACP thioesterase and truncated fabG from Mycobacterium tuberculosis under a constitutive promoter system. Plasmid pWL1TG2Z carries an acyl-ACP thioesterase, truncated fabG from Mycobacterium tuberculosis and fabZ gene from E. coli under an inducible trc promoter system.

(13) FIG. 10: Effect of antisense ec fabG on fatty acid production. Plasmid pWL1TG2 carries an acyl-ACP thioesterase and truncated fabG from Mycobacterium tuberculosis under a constitutive promoter system. Plasmids pWL1TG2AS, pWL1TG2AS2 and pWL1TG2AS3 are derived from plasmid pWL1TG2 with additional anti-sense RNAs of ec fabG of different length. The number 1 to 3 denotes the length of the anti-sense of 450, 300 and 150 bp, respectively.

(14) FIG. 11: Effect of antisense ec fabG on fatty acid production. Plasmid pWL1TG2 carries an acyl-ACP thioesterase and truncated fabG from Mycobacterium tuberculosis under a constitutive promoter system; plasmids pHWTAS1, pHWTAS2, pHWTAS3 is pBAD33 based which carries an acyl-ACP thioesterase under a constitutive promoter and different length of anti-sense RNAs of ec fabG with a loop design as a stabilizer under an inducible trc promoter. The number 1 to 3 denotes the length of the anti-sense of 450, 300 and 150 bp, respectively.

(15) FIG. 12A-C: Constructs for generating anti-sense RNA against FabG without loop design. FIG. 12A: AS1; FIG. 12B: AS2; FIG. 12C: AS3.

(16) FIG. 13A-C: Constructs for generating anti-sense RNA against FabG with loop design. FIG. 13A: AS1-looped; FIG. 13B: AS2-looped; FIG. 13C: AS3-looped.

DETAILED DESCRIPTION

(17) This disclosure provides the inventive concept of replacing or supplementing the NADP+-dependent enzyme in a type II fatty acid synthesis pathway with an NAD+-dependent enzyme so as to take advantage of the higher concentration of NADH/NAD.sup.+ in cells. In E. coli, as in many species, the 3-oxoacyl-ACP reductase is NADPH-dependent, and is thus rate limiting, and adding an NADH-dependant reductase alleviates this bottleneck, allowing more fats to be made.

(18) To demonstrate the concept, we used previously constructed host strain E. coli strain, ML103 (MG1655, fadD) for fatty acid production with a deleted long-chain fatty acyl coenzyme A synthase gene fadD. The fadD is not an essential component of the invention, although it does improve fatty acid accumulation. FadD is the first step in the fatty acid beta-oxidation pathway. It activates the fatty acid to acyl-CoA before going into the beta-oxidation cycle, thus its deletion helps to conserve the fatty acids that are made. However, any enzyme in the beta-oxidation pathway can provide similar effects if reduced or knocked out.

(19) The base strain also contained a thioesterase (TE) gene from Ricinus communis, which functions to release free fats from the ACP thus allow increased levels of free fatty acids to accumulate.

(20) TE expression of some kind is needed to allow free fatty acid production. The host's native TE (such as TesA and TesB) is capable of providing some activity, but we have shown that overexpression of either endogenous or exogenous TE significantly improves free fatty acids levels. Furthermore, by tailoring the TE gene used, we are able to influence the length of the free fatty acids produced.

(21) A plasmid carrying an acyl-ACP thioesterase from Ricinus communis (Acc no.: XM002515518) (Zhang et al., 2011) and a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from Mycobacterium tuberculosis (Acc. no.: Rv0242c) is used as an example. Two versions, one with the full length reductase (named G1) and the other one with an omission of the first 16 amino acids (named G2) of the 3-oxoacyl-ACP reductase were tested, because the first 16 amino acids were found to hinder the solubility of the recombinant protein (Dutta et al, 2011). The schematics of the plasmid constructs pXZ18G1 and pXZ18G2 are shown in FIG. 2 and FIG. 3, respectively.

(22) By replacing or supplementing NADP.sup.+-dependent 3-oxoacyl-ACP reductase with NAD.sup.+-dependent 3-oxoacyl-ACP reductase, the higher availability of NAD.sup.+ in the organism will facilitate synthesis of fatty acids and result in higher yield thereof, especially long-chain fatty acid of 14 or more carbons. This significantly increases the efficiency of fatty acid production and reduces the cost thereof.

(23) LB medium supplemented with approximately 15 g/L glycerol as a carbon source and 100 mg/L ampicillin for selection were used for culturing cells. Isopropyl--D-thiogalactopyranoside (IPTG) was added to the medium to a final concentration of 200 M, thus inducing the expression of acyl-ACP thioesterase and NAD.sup.+ dependent 3-oxoacyl-ACP reductase.

(24) A single colony of strain (ML103-18 (control), ML103-18G1 (NADH-dependent enzyme) or ML103-18G2 (truncated NADH-dependent enzyme) was inoculated into 5 ml of Luria-Bertani (LB) and incubated in an orbital shaker operated at 250 rpm at 30 C. overnight. The pre-culture was inoculated into a flask containing 50 mL of the culture medium with 1% (v/v) inoculum. The culture medium contained 50 ml LB and about 15 g/L of glycerol.

(25) Shake flask experiments were performed at 30 C. with shaking at 250 rpm. Samples were taken at four specific time points (0, 24, 48 and 72h) to quantify the fatty acids produced and glycerol consumed. All experiments were carried out in triplicates. The results are shown in Tables 1 and 2.

(26) TABLE-US-00009 TABLE 1 Percentage improvement of fatty acid production and yield using a NAD+ dependent 3-oxoacyl-ACP reductase at 24, 48 and 72 hours. Values shown are averages of triplicates Concentrations of free fatty acid Yield FFA % yield % Relevant (g/L) (g fatty acid/ improve- improve- Strain genotype Time C14 C16 C16:1 C18 Total g glycerol) ment ment ML103-18 fadD, rcTE.sup.+ 24 h 0.651 0.622 0.368 0.063 1.704 0.169 ML103-18G1 fadD, rcTE.sup.+ 0.947 0.327 0.763 0.040 2.077 0.211 21.91 24.82 G1.sup.+ ML103-18G2 fadD, rcTE.sup.+ 1.071 0.337 0.719 0.030 2.157 0.209 26.57 23.58 G2.sup.+ ML103-18 fadD, rcTE.sup.+ 48 h 0.850 0.733 0.383 0.063 2.030 0.146 ML103-18G1 fadD, rcTE.sup.+ 1.346 0.399 0.946 0.043 2.733 0.194 34.68 32.41 G1.sup.+ ML103-18G2 fadD, rcTE.sup.+ 1.323 0.376 0.796 0.031 2.526 0.203 24.46 38.57 G2.sup.+ ML103-18 fadD, rcTE.sup.+ 72 h 0.957 0.792 0.464 0.064 2.276 0.142 ML103-18G1 fadD, rcTE.sup.+ 1.548 0.449 1.089 0.045 3.130 0.195 37.51 37.68 G1.sup.+ ML103-18G2 fadD, rcTE.sup.+ 1.609 0.433 0.946 0.032 3.020 0.194 32.67 36.81 G2.sup.+ rcTE.sup.+: overexpression of acyl-ACP thioesterase from Ricinus communis; G1.sup.+: overexpression of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis G2.sup.+: overexpression of a 16 aa truncant of the NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis

(27) TABLE-US-00010 TABLE 2 Fatty acid distribution comparison (derived from Table 1) Fatty acid distribution Relevant (% of total free fatty acid) % increase % increase Strain genotype Time C14 C16 C16:1 C18 C14 in % C14 ML103-18 fadD, rcTE.sup.+ 24 h 38.218 36.520 21.577 3.685 ML103-18G1 fadD, rcTE.sup.+ 45.581 15.725 36.751 1.943 45.390 19.265 G1.sup.+ ML103-18G2 fadD, rcTE.sup.+ 49.640 15.646 33.340 1.374 64.399 29.886 G2.sup.+ ML103-18 fadD, rcTE.sup.+ 48 h 41.875 36.137 18.894 3.094 ML103-18G1 fadD, rcTE.sup.+ 49.227 14.600 34.605 1.568 58.321 17.556 G1.sup.+ ML103-18G2 fadD, rcTE.sup.+ 52.394 14.876 31.508 1.223 55.717 25.118 G2.sup.+ ML103-18 fadD, rcTE.sup.+ 72 h 42.022 34.814 20.366 2.798 ML103-18G1 fadD, rcTE.sup.+ 49.451 14.341 34.780 1.428 61.814 17.678 G1.sup.+ ML103-18G2 fadD, rcTE.sup.+ 53.273 14.336 31.337 1.054 68.191 26.772 G2.sup.+ rcTE.sup.+: overexpression of acyl-ACP thioesterase from Ricinus communis; G1.sup.+: overexpression of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis G2.sup.+: overexpression of a 16 aa truncant of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis

(28) In summary, as shown in Table 1, both NAD.sup.+ dependent 3-oxoacyl-ACP reductase carrying strains, ML103-18G1 and ML103-18G2, produced more fatty acids than that of the control strain ML103-18 (Table 1). At 72 hours, the ML103-18G1 and ML103-18G1 strains accumulated more than 34% and 32% of free fatty acids than that of the control strain ML103-18, respectively (Table 1). In addition, both NAD.sup.+ dependent 3-oxoacyl-ACP reductase-carrying strains, ML103-18G1 and ML103-18G2, gave higher yields than that of the control strain ML103-18, more than 35% at 72 hours (Table 1).

(29) Both NAD.sup.+ dependent 3-oxoacyl-ACP reductase-carrying strains, ML103-18G1 and ML103-18G2, also showed changes in the free fatty acid distribution as compared to that of the control strain ML103-18 (Table 2). The ML103-18G1 and ML103-18G2 strains accumulated more than 61% and 68% of C14 free fatty acids than that of the control strain ML103-18, respectively (Table 2), but this is due to the substrate specificity of the added TE gene. Of course, the exit points can be modified by tailoring the exit point for the fatty acid elongation cycle (see e.g., WO2013096665). Thus, by changing the added TE gene, one can influence the fatty acid length.

(30) The ability of the NAD.sup.+ dependent 3-oxoacyl-ACP reductase to improve free fatty acid production in two pyridine nucleotide transhydrogenase mutant strains was also examined. The pyridine nucleotide transhydrogenases normally function to reoxidize NADPH, according to the following:

(31) EC Number: 1.6.1.2/1.6.1.3

(32) NAD.sup.++NADPH<=>NADP.sup.++NADH

(33) Thus, deleting these would prevent the conversion of NADH to NADPH.

(34) A strain WLK09 with the cytoplasmic transhydrogenase (sth) deactivated and the other strain WLK310 with the membrane bound transhydrogenase (pntB) deactivated were used. Both strains also have the beta-oxidation pathway blocked by deleting the fadD gene to prevent the degradation of fatty acids produced in vivo.

(35) LB medium supplemented with approximately 15 g/L glycerol as a carbon source and 100 mg/L ampicillin for selection were used for culturing cells. IPTG was added to a final concentration of 200 M, thus inducing the expression of the added genes.

(36) A single colony of each strain (ML309-18, ML309-18G1 & ML309-18G2 or ML310-18, ML310-18G1 or ML310-18G2) was inoculated into 5 ml of Luria-Bertani (LB) and treated as above. The results are shown in Tables 3 and 4 below.

(37) TABLE-US-00011 TABLE 3 Effect of NAD+ dependent 3-oxoacyl-ACP reductase in a transhydrogenase mutant host Concentrations of free fatty acid Yield FFA % yield % Relevant (g/L) (g fatty acid/ improve- improve- Strain genotype Time C14 C16 C16:1 C18 Total g glycerol) ment ment WLK309-18 fadD, sth, 24 h 0.150 0.157 0.197 0.023 0.527 0.145 rcTE.sup.+ WLK309-18G2 fadD, sth, 0.375 0.157 0.327 0.011 0.871 0.148 65.24 1.87 rcTE.sup.+ WLK310-18 fadD, pntAB, 0.204 0.297 0.539 0.123 1.163 0.123 rcTE.sup.+ G2.sup.+ WLK310-18G2 fadD, pntAB, 0.568 0.272 0.786 0.050 1.676 0.160 44.09 30.12 rcTE.sup.+ G2.sup.+ WLK309-18 fadD, sth, 48 h 0.707 0.550 0.581 0.069 1.907 0.162 rcTE.sup.+ WLK309-18G2 fadD, sth, 1.279 0.356 0.674 0.018 2.326 0.173 21.98 7.10 rcTE.sup.+ WLK310-18 fadD, pntAB, 0.211 0.317 0.552 0.149 1.230 0.080 rcTE.sup.+ G2.sup.+ WLK310-18G2 fadD, pntAB, 0.582 0.279 0.793 0.067 1.720 0.111 39.84 39.55 rcTE.sup.+ G2.sup.+ WLK309-18 fadD, sth, 72 h 0.916 0.621 0.619 0.080 2.237 0.158 rcTE.sup.+ WLK309-18G2 fadD, sth, 1.592 0.421 0.790 0.023 2.825 0.180 26.33 13.83 rcTE.sup.+ WLK310-18 fadD, pntAB, 0.233 0.360 0.623 0.168 1.384 0.089 rcTE.sup.+ G2.sup.+ WLK310-18G2 fadD, pntAB, 0.618 0.288 0.844 0.074 1.824 0.118 31.76 32.72 rcTE.sup.+ G2.sup.+ rcTE.sup.+: overexpression of acyl-ACP thioesterase from Ricinus communis; G1.sup.+: overexpression of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis G2.sup.+: overexpression of a derivative of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis pntAB = deactivation of the membrane bounded transhydrogenase sth = deactivation of the soluble transhydrogenase

(38) TABLE-US-00012 TABLE 4 Fatty Acid distribution Comparison (derived from Table 3) Fatty acid distribution Relevant (% of total free fatty acid) % increase % increase Strain genotype Time C14 C16 C16:1 C18 C14 in % C14 WLK309-18 fadD, sth, 24 h 0.284 0.298 0.374 0.044 rcTE.sup.+ WLK309-18G2 fadD, sth, 0.430 0.180 0.376 0.013 150.71 51.72 rcTE.sup.+ G1.sup.+ WLK310-18 fadD, pntAB, 0.176 0.255 0.463 0.106 rcTE.sup.+ G2.sup.+ WLK310-18G2 fadD, pntAB, 0.339 0.162 0.469 0.030 177.82 92.82 rcTE.sup.+ G2.sup.+ WLK309-18 fadD, sth, 48 h 0.371 0.288 0.305 0.036 rcTE.sup.+ WLK309-18G2 fadD, sth, 0.550 0.153 0.290 0.008 80.85 48.26 rcTE.sup.+ G1.sup.+ WLK310-18 fadD, pntAB, 0.172 0.258 0.449 0.121 rcTE.sup.+ G2.sup.+ WLK310-18G2 fadD, pntAB, 0.338 0.162 0.461 0.039 175.57 97.06 rcTE.sup.+ G2.sup.+ WLK309-18 fadD, sth, 72 h 0.409 0.278 0.277 0.036 rcTE.sup.+ WLK309-18G2 fadD, sth, 0.563 0.149 0.279 0.008 73.81 37.59 rcTE.sup.+ G1.sup.+ WLK310-18 fadD, pntAB, 0.169 0.260 0.450 0.121 rcTE.sup.+ G2.sup.+ WLK310-18G2 fadD, pntAB, 0.339 0.158 0.463 0.040 164.81 100.97 rcTE.sup.+ G2.sup.+ rcTE.sup.+: overexpression of acyl-ACP thioesterase from Ricinus communis; G1.sup.+: overexpression of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis G2.sup.+: overexpression of a derivative of a NAD.sup.+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis pntAB = deactivation of the membrane bounded transhydrogenase sth = deactivation of the soluble transhydrogenase

(39) Again, both NAD.sup.+ dependent 3-oxoacyl-ACP reductase carrying strains, WLK309-18G2 and WLK310-18G2, produced significantly more fatty acids and with higher yields than that of the corresponding control strains WLK309-18 and WLK310-18 (Table 3). At 72h, both NAD.sup.+ dependent 3-oxoacyl-ACP reductase carrying strains, WLK309-18G2 and WLK310-18G2, produced 13% and 30% more fatty acids with 26% and 30% higher yield than that of the corresponding control strains WLK309-18 and WLK310-18, respectively (Table 3). In addition, both NAD.sup.+ dependent 3-oxoacyl-ACP reductase carrying strains, WLK309-18G2 and WLK310-18G2, produced more C14 fatty acids, 73% and 164%, than that of the corresponding control strains, respectively at 72 h (Table 4).

(40) Therefore, these results demonstrate that NADPH is a limiting factor, and it can be alleviated by the introduction of a NAD-dependent 3-oxoacyl-ACP reductase. In addition, the introduction of a NAD-dependent 3-oxoacyl-ACP reductase changes the composition of the fatty acids produced, yielding more C14 chain length fatty acid.

(41) The native NADPH-dependant 3-oxoacyl-ACP reductase was believed to be an essential gene. Therefore, we first tried to reduce its expression with antisense, so that some amount of gene/enzyme activity would remain.

(42) We made expression plasmids encoding 150, 300 and 450 bp antisense against the gene of NADPH-dependant 3-oxoacyl-ACP reductase under the control of an IPTG inducible promoter (lacZ). We measured fatty acid levels at 24 and 48 hrs. See Table 5 and 6.

(43) Surprisingly, those bacteria with reduced native NADPH-dependant 3-oxoacyl-ACP reductase and added NADH-dependant 3-oxoacyl-ACP reductase made more fatty acids that those with wild type levels of expression of NADPH-dependant 3-oxoacyl-ACP reductase and added NADH-dependant 3-oxoacyl-ACP reductase.

(44) TABLE-US-00013 TABLE 5 Effect of overexpression NAD+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis and inhibition expressing of host FabG using normal design of anti-sense. Yield FFA % Yield % Relevant IPTG Concentration of fatty acid (g fatty acid/ improve- improve- Strains genotype (mM) Time C14 C16:1 C16 C18 Total g glucose) ment ment ML103 fadD, rcTE.sup.+ 1 24 h 1.548 0.541 0.935 0.118 3.142 0.218 (pWL1TG2) G2.sup.+ Control ML103 fadD, rcTE.sup.+ 0 1.786 0.490 0.898 0.099 3.273 0.221 4.16 1.14 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 0 0.831 0.770 0.442 0.134 2.178 0.158 30.70 27.56 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 0 0.925 0.837 0.472 0.146 2.379 0.181 24.28 16.92 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ ML103 fadD, rcTE.sup.+ 1 1.712 0.471 0.866 0.103 3.152 0.238 0.32 8.86 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 1 1.077 0.981 0.476 0.150 2.684 0.193 14.58 11.38 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 1 1.031 0.870 0.439 0.134 2.474 0.171 21.25 21.83 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ ML103 fadD, rcTE.sup.+ 1 48 h 1.640 0.580 1.005 0.127 3.352 0.217 (pWL1TG2) G2.sup.+ Control ML103 fadD, rcTE.sup.+ 0 2.122 0.569 1.065 0.116 3.872 0.257 15.52 18.81 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 0 1.218 0.994 0.796 0.182 3.190 0.217 4.84 0.00 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 0 1.215 1.117 0.595 0.186 3.113 0.207 7.12 4.31 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ ML103 fadD, rcTE.sup.+ 1 2.167 0.566 1.053 0.122 3.907 0.270 16.55 24.43 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 1 1.305 1.209 0.598 0.196 3.308 0.232 1.33 6.94 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 1 1.328 1.144 0.557 0.176 3.204 0.210 4.41 2.89 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ rcTE+: overexpression of acyl-ACP thioesterase from Ricinus communis; G2+: overexpression of a derivative of a NAD+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis AS1.sup.+~AS3.sup.+: anti-sense fabG targeting the E. coli fabG structural gene with different lengths of 400, 250 and 150, respectively

(45) TABLE-US-00014 TABLE 6 Fatty Acid distribution Comparison (derived from Table 5) Fatty acid distribution Relevant IPTG (% of total free fatty acid) % increase % increase Strains genotype (mM) Time C14 C16:1 C16 C18 C14 in % C14 ML103 fadD, rcTE.sup.+ 1 24 h 49.25 17.22 29.76 3.77 (pWL1TG2) G2.sup.+ Control ML103 fadD, rcTE.sup.+ 0 54.58 14.97 27.42 3.03 15.43 10.81 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 0 38.18 35.37 20.28 6.17 46.28 22.47 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 0 38.87 35.18 19.82 6.13 40.24 21.08 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ ML103 fadD, rcTE.sup.+ 1 54.32 14.93 27.49 3.26 10.65 10.30 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 1 40.12 36.54 17.75 5.59 30.41 18.53 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 1 41.65 35.17 17.75 5.43 33.40 15.42 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ ML103 fadD, rcTE.sup.+ 1 48 h 48.93 17.31 29.97 3.79 (pWL1TG2) G2.sup.+ Control ML103 fadD, rcTE.sup.+ 0 54.80 14.71 27.49 3.00 29.39 12.01 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 0 38.19 31.17 24.95 5.69 25.73 21.95 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 0 39.03 35.86 19.12 5.99 25.91 20.23 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ ML103 fadD, rcTE.sup.+ 1 55.46 14.48 26.94 3.12 32.10 13.34 (pWL1TG2AS1) G2.sup.+ AS1.sup.+ ML103 fadD, rcTE.sup.+ 1 39.44 36.55 18.07 5.94 20.46 19.39 (pWL1TG2AS2) G2.sup.+ AS2.sup.+ ML103 fadD, rcTE.sup.+ 1 41.44 35.71 17.37 5.48 19.03 15.30 (pWL1TG2AS3) G2.sup.+ AS3.sup.+ rcTE+: overexpression of acyl-ACP thioesterase from Ricinus communis; G2+: overexpression of a derivative of a NAD+ dependent 3-oxoacyl-ACP reductase from M. tuberculosis AS1.sup.+~AS3.sup.+: anti-sense fabG targeting the E. coli fabG structural gene with different lengths of 400, 250 and 150, respectively

(46) We also knocked out the native NADPH-dependant 3-oxoacyl-ACP reductase and added NADH-dependant 3-oxoacyl-ACP reductase, but these cells did less well that those cells with wild type NADPH-dependant 3-oxoacyl-ACP reductase and added NADH-dependent 3-oxoacyl-ACP reductase (data not shown) possibly because the expression level of the added gene was not high enough. However, we expect that further fine-tuning of the NADH-dependent 3-oxoacyl-ACP reductase will improve the fatty acid production, and experiments are in progress to show this.

(47) We expect that combining the introduction of NADH-dependent 3-oxoacyl-ACP reductase with overexpressed UDH+ and/or PntAB+ will improve the fatty acid production because it will allow efficient usage of both NADH-dependent 3-oxoacyl-ACP reductase and the native NADPH-dependent 3-oxoacyl-ACP reductase, and experiments are in progress to show this.

(48) Our lab has made many engineered bacteria that produced increased amount of fatty acids, and many of those modifications are compatible herewith. Direct strategies that have been tested and proven effective can be classified into two broad categories: i) overexpression of enzymes catalyzing key steps in the fatty acid synthesis pathway, including endogenous or heterologous thioesterases (TE), acetyl-CoA carboxylase (ACC), and acyl-CoA ligases (ACL); and ii) deletion of enzymes involved in the -oxidation pathway that degrades fatty acids, such as acyl-CoA dehydrogenase (FadE), acyl-CoA synthetase (FadD), and a long-chain fatty acid outer membrane transporter (FadL). In one of the latest studies, efforts along this direction led to a titer of 5.1 g/L extracellular fatty acids and a yield of 4.1% (g per g glucose supplied) in a fed-batch culture with online product extraction.

(49) The research efforts described above focused on local pathways directly related to fatty acids. However, modifications in distant pathways, such as glycolysis or TCA cycle, can also improve fatty acid synthesis through redistribution of metabolite precursors towards fatty acid production.

(50) For example, the level of malonyl-CoA, a precursor for fatty acids was improved 15-fold through the deletion of ackA-pta and adhE, together with the overexpression of acetyl-CoA synthetase (Acs).

(51) For another example, there are two other lactate dehydrogenases in E. coli encoded by ldhA and lldD. Knocking one or both out would block the formation of lactate from pyruvate and direct more carbon towards fatty acid biosynthesis.

(52) Second, there is another acetate-producing pathway catalyzed by poxB encoded pyruvate oxidase. Even though the amount of acetate was quite low after pta was deleted in the above study, further knockout of poxB would lead to complete elimination of this by-product.

(53) Finally, it has been suggested that derepression of the glyoxylate bypass by iclR deletion alone cannot draw isocitrate from the TCA cycle to the glyoxylate bypass because enzyme IcdA has a stronger affinity to isocitrate than enzymes AceA and AceB. Hence, to fully activate the glyoxylate bypass, icdA may need to be knocked out in addition to iclR.

(54) These genetic combinations with the invention described herein will be explored in our future study, and the work is expected to proceed quickly as many base strains and/or expression plasmids are already available.

(55) High fat producing microbes can also be combined with genes that would allow the microbes to use less energy intensive food sources than glucose. For example, glycerol is a by-product of biodiesel production and is a very inexpensive food-source, and microbes can be altered to allow growth on glycerol. See Murarka (2008). As another example, cellulosic food-sources are also readily available, and microbes have been engineered to secrete cellulose degrading enzymes and thus are able to grow or e.g., switchgrass. Bokinsky (2011). Ultimately, the engineered microbes described herein may be combined with this additional type of engineering as the microbes are adapted for large scale production of fats or their derivatives.

(56) We predict that the inventive concept can be applied to other organisms having Type II fatty acid synthesis systems to achieve similar improvement of fatty acid production, as long as suitable NAD.sup.+-dependent 3-oxoacyl-ACP reductase, native or engineered or exogenous, is available to replace or augment the native NADP.sup.+-dependent 3-oxoacyl-ACP reductase.

(57) As shown herein, there are thousands of such enzyme sequences that can be used when placed into a suitable expression vector for the chosen host species. If expression levels are low, the codon usage can be optimized for the species in question, as optimized codon charts are available for many species. Further, the genes are fairly small, and complete synthesis of an optimized codon ORF would be fairly quick and inexpensive.

(58) We expect that the higher availability of NAD.sup.+ than NADP.sup.+ in such organisms will make the concept equally beneficial in these FASII organisms. Examples of FASII organisms include most bacteria, algae and plants, including but not limited to Escherichia, Bacillus, Lactobacillus, Staphylococcus, Salmonella, Haemophilus, Lemnoideae, Chlamydomonas, Chlorella, Nannochloropsis. Yeast mitochondria have FASII genes, as well. Future experiments may test one of the microalgae or other bacteria, and we expect that improved production will be found on replacing or supplementing NADP-based enzymes with NADH-based enzymes.

(59) The above experiments can be repeated in Bacillus subtilis. The same genes can be used, especially since Bacillus has no significant codon bias. A protease-deficient strain like WB800N is preferably used for greater stability of heterologous protein. The E. coli-B. subtilis shuttle vector pMTLBS72 exhibiting full structural stability can be used to move the genes easily to a more suitable vector for Bacillus. Alternatively, two vectors pHT01 and pHT43 allow high-level expression of recombinant proteins within the cytoplasm. As yet another alternative, plasmids using the theta-mode of replication such as those derived from the natural plasmids pAM1 and pBS72 can be used. Several other suitable expression systems are available. Since the FAS genes are ubiquitous, the invention is predicted to function in bacillus.

(60) The inventors further tested the effect of co-overexpressing mtFabG (FabG gene obtained from Mycobacterium tuberculosis) and FabZ in order to improve fatty acid productivity. This is prompted by the observation of the following experiment with higher overexpression of mt FabG2 (3-ketoacyl-ACP reductase obtained from Mycobacterium tuberculosis). FabZ is 3R-hydroxymyristoyl ACP dehydratase, and as shown in FIG. 1B, it is involved in fatty acid synthesis.

(61) An experiment was designed to test if the expression of mt FabG is the limiting factor by cloning two copies of the mt fabG genes in the same plasmid (the resulting plasmid is called pXZ18DG2, as shown in FIG. 3B). Fermentation experiments were performed with ML103 as the host carrying either the plasmid pXZ18G2 or pXZ18DG2 in glucose or glycerol supplemented LB medium. The results are shown in FIG. 8A-B. The values shown are the average of triplicate runs. Unlike previous experimental observations that overexpression of a single copy of mt fabG gene increases fatty acid production, the double mt FabG did not show any improvement when comparing the strain ML103(pXZ18G2) with ML103(pXZ18DG2) in both glucose and glycerol. The decrease in fatty acid titer is more significant in glycerol than that of glucose. This reduction of fatty acid production in strains carrying pXZ18DG2 suggests that too much -ketoacyl reductase activity or NADH supply would lead to discoordination within the fatty acid elongation cycle. We hypothesize that increased -ketoacyl reductase activity resulted in an accumulation of 3-hydroxyacyl-ACP and thus might cause feedback inhibition. It has been reported that acyl-ACP intermediates might act as feedback inhibitors for fatty acid production.

(62) The inventors further tested the hypothesis of recovering the coordination among the reactions within the fatty acid elongation cycle by co-overexpressing mt FabG and FabZ in order to improve fatty acid productivity. The results shown in FIG. 9 support the hypothesis that FabG might not be the only limiting factor. The fatty acid production by the mt FabG and FabZ double-overexpression strain increased by about 20% when compared to the strain with mt FabG overexpression alone. Co-overexpression of FabZ, downstream of mt FabG, alleviates the imbalance caused by the overexpression of mt FabG alone, leading to an improved performance.

(63) The inventors also examined the effect of down regulation of the native E. coli NADPH-dependent FabG (ec FabG) on fatty acid production. Since NADH is more readily available in E. coli and that the NADH-dependent FabG (mt FabG) should be more efficient than the native ec FabG, we speculate that fatty acid production can be improved by increasing the relative ratio of the newly introduced mt FabG to that of the native ec FabG. We chose to use the anti-sense RNA techniques to decrease the expression of ec fabG since fabG is an essential gene.

(64) The fermentation data showed that only the strain carry the plasmid with the longest anti-sense RNA, pWL1TG2AS1, produced similar amount of fatty acid as control strain at 24 h, the other two strains carrying plasmids pWL1TG2AS2 or pWL1TG2AS3 were significantly lower (FIG. 10). These two strains however caught up at 48 h, but the final total fatty acid concentrations were similar to that of the control. At 48 h, the strain carrying pWL1TG2AS1 produced about 3.9 g/L fatty acid, which is 16.5% higher than that of the control. Bacterial RNAs are typically short lived, and some of the reports suggested that half-life is an important factor in anti-sense RNA efficiency. It is therefore possible that the shorter sequences are more likely to be degraded quicker. As a result, only the longest anti-sense RNA showed the positive improvement effect. The results also indicated that a system with mt FabG overexpression and native FabG down regulation provides the best performance.

(65) In order to better compare the effect of down-regulating the native ec fabG, a paired-termini (PT) design was used to stabilize the anti-sense RNA. All three newly paired termini anti-sense RNA constructs (pHWTAS1, pHWTAS2, pHWTAS3) shared the same length and sequence as the earlier design, except being stabilized by the addition of a hairpin structure. A control plasmid pHWTASC was also constructed using a dummy sequence to replace the anti-sense portion.

(66) FIG. 11 shows the effect of anti-sense RNAs expression, all three strains carrying the new anti-sense constructs showed similar improved fatty acid production over the control strain. The final titers are 3.59, 3.72 and 3.76 g/L compared to the control of 2.74 g/L, which represent improvements of 31.1%, 35.2% and 36.8%, respectively. The results further indicate that overexpression of a NAD-dependent mt FabG and down-regulating the native NADPH-dependent ecFabG is a practical approach to improve fatty acid production in E. coli.

(67) Although work is still needed to scale up microalgae production for use in making biofuels, they are especially attractive as a source of fuel from an environmental standpoint because they consume carbon dioxide and can be grown on marginal land, using waste or salt water.

(68) Indeed, Ann Ruffings group from Sandia National Laboratories has already engineered two strains of cyanobacteria to produce free fatty acid, and is working with a third. The cyanobacteria were chosen because fuel from engineered cyanobacteria is excreted outside the cell, in contrast to eukaryotic algae, in which fuel production occurs inside the cell. This greatly simplifies scale up, as the cyanobacteria continue to grow, while fats are skimmed from the top of the culture media.

(69) In addition, Radakovitz has overexpressed two genes encoding acyl-ACP thioesterase (TE) of plant origin in P. tricornutum to produce medium-chain fatty acids in the oil fraction. These results provide adequate foundation for applying this invention to microalgae, such that the NADPH dependent 3-oxoacyl-ACP reductase is supplemented or replaced with an NADH-dependent enzyme.

(70) Further, significant advances in microalgal genomics have been achieved during the last decade. Expressed sequence tag (EST) databases have been established; nuclear, mitochondrial, and chloroplast genomes from several microalgae have been sequenced; and several more are being sequenced. Historically, the green alga Chlamydomonas reinhardtii has been the focus of most molecular and genetic physiological research. Therefore, most of the tools for the expression of transgenes and gene knockdown have been developed for and are specific for this species. However, tools are now also being rapidly developed for diatoms and other algae that are of greater interest for industrial applications.

(71) Additionally, successful genetic transformation has been reported for the green (Chlorophyta), red (Rhodophyta), and brown (Phaeophyta) algae; diatoms; euglenids; and dinoflagellates, although the efficiency of transformation seems to be strongly species dependent, and the method of transformation has to be carefully selected and optimized for each microalga, and the stability of expression improved through proper codon usage, the use of strong endogenous promoters, and inclusion of species-specific 5, 3, and intron sequences.

(72) The following references are incorporated by reference in their entirety for all purposes.

(73) All accession numbers referenced herein, and the sequences and data therein, are incorporated by reference herein in their entireties for all purposes. Accession numbers can be accessed, e.g., at GenBank, EMBL, Brenda, UniProt and the like. Bergler H, et a., Protein EnvM is the NADH-dependent enoyl-ACP reductase (FabI) of Escherichia coli, J Biol Chem. 269(8):5493-6 (1994). Bokinsky G. et al., Synthesis of three advanced biofuels from ionic liquid-pretreated switchgrass using engineered Escherichia coli, PNAS, 108(50): 19949-19954 (2011). Caughey I, Kekwick R G, The characteristics of some components of the fatty acid synthetase system in the plastids from the mesocarp of avocado (Persea americana) fruit, Eur J Biochem. 123(3):553-61 (1982). Dutta D, et al., Crystal structure of FabG4 from Mycobacterium tuberculosis reveals the importance of C-terminal residues in ketoreductase activity, J Struct Biol. 174(1):147-55 (2011). Fuhrer T, Sauer U. Different biochemical mechanisms ensure network-wide balancing of reducing equivalents in microbial metabolism, J Bacteriol. 191(7):2112-21 (2009). Jing et al., Phylogenetic and experimental characterization of an acyl-ACP thioesterase family reveals significant diversity in enzymatic specificity and activity, BMC Biochemistry 2011, 12:44. Martnez I, et al., Replacing Escherichia coli NAD-dependent glyceraldehyde 3-phosphate dehydrogenase (GAPDH) with a NADP-dependent enzyme from Clostridium acetobutylicum facilitates NADPH dependent pathways, Metab Eng. 10(6):352-9 (2008). Murarka A., et al., Fermentative Utilization of Glycerol by Escherichia coli and Its Implications for the Production of Fuels and Chemicals, Appl Environ Microbiol. February 2008; 74(4): 1124-1135. Radakovits, R., Eduafo, P. M. and Posewitz, M. C. Genetic engineering of fatty acid chain length in Phaeodactylum tricornutum. Metab. Engin., 13, 89-95 (2011). Ruffing, A. M. & Jones, H. D. T. Physiological effects of free fatty acid production in genetically engineered Synechococcus elongatus PCC 7942. Biotechnology and Bioengineering, 2012; 109 (9): 2190. Sanchez A M, et al., Effect of overexpression of a soluble pyridine nucleotide transhydrogenase (UdhA) on the production of poly(3-hydroxybutyrate) in Escherichia coli, Biotechnol Prog. 22(2):420-5 (2006). Wang Y, at al, Improvement of NADPH bioavailability in Escherichia coli by replacing NAD+-dependent glyceraldehyde-3-phosphate dehydrogenase GapA with NADP+-dependent GapB from Bacillus subtilis and addition of NAD kinase, J Ind Microbiol Biotechnol. 2013. PMID: 24048943. Zhang X, Li M, Agrawal A, San K Y, Efficient free fatty acid production in Escherichia coli using plant acyl-ACP thioesterases, Metab Eng. 13(6):713-22 (2011). WO2013096665 Long chain organic acid bioproduction. WO2011064183 Novel fatty acid elongase and uses thereof WO2010142522 Novel fatty acid elongation components and uses thereof WO2012052468 Novel fatty acid desaturases, elongases, elongation components and uses thereof US20140093921, WO2011116279 Bacteria and method for synthesizing fatty acids WO2013059218 Bacteria and method for synthesizing fatty acids US20140212935 Short chain fatty acids from bacteria US20140193867 Microbial odd chain fatty acids