TYPE II FATTY ACID SYNTHESIS ENZYMES IN REVERSE ß-OXIDATION

PRIOR RELATED APPLICATIONS

This application claims priority to 61/932,057, filed Jan. 27, 2014, and incorporated by reference herein in its entirety for all purposes.

FEDERALLY SPONSORED RESEARCH STATEMENT

This invention was made with government support under Grant Nos: CBET-1134541, CBET-1067565, and EEC-0813570, awarded by the NSF. The government has certain rights in the invention.

FIELD OF THE DISCLOSURE

The disclosure generally relates to the use of microorganisms to make feedstock chemicals, e.g., fatty acids and derivatives thereof, by driving the beta oxidation cycle in reverse.

BACKGROUND OF THE DISCLOSURE

Much effort has been devoted in recent years to the production of biofuels, such as ethanol, butanol, higher-chain alcohols, and hydrocarbons via microbial fermentation of sugars and other biomass constituents.

To date, the fatty acid biosynthesis pathway has been widely used as the means to generate higher-chain (C≧4) acyl-CoA thioesters required for the synthesis of the aforementioned products. However, the operation of this pathway is not efficient because it consumes ATP in the synthesis of malonyl-ACP, which is the donor of two-carbon units for chain elongation. As a consequence, the ATP yield associated with the production of hydrocarbon through the fatty acid synthesis pathway is very low. This, in turn, greatly limits cell growth and hydrocarbon and other product production.

We have implemented an entirely novel approach, driving beta oxidation in reverse to make fatty acids instead of degrading them (see US20130316413, WO2013036812, each incorporated by reference in its entirety for all purposes). Unlike the fatty acid biosynthesis pathway, the reversal of the β-oxidation cycle operates with coenzyme-A (CoA) thioester intermediates and uses acetyl-CoA directly for acyl-chain elongation (rather than first requiring ATP-dependent activation to malonyl-CoA).

A engineered microorganism having a reverse beta oxidation cycle that produces alcohols, carboxylic acids, and hydrocarbons, and derivatives thereof, generally includes i) expression of the β-oxidation cycle in the absence of fatty acids and presence of a non-fatty acid carbon source, ii) functional operation of the β-oxidation cycle in the reverse or biosynthetic direction (e.g. making fats rather than degrading them), iii) overexpression of one or more termination enzymes that convert reverse beta oxidation cycle intermediates to a desired alcohol, carboxylic acid, or hydrocarbon, thus exiting or terminating the cycle for that intermediate. Further, any of the alcohols, carboxylic acids, and hydrocarbon products can be further modified to make other products, such as aldehydes, and the like, in secondary termination pathways.

The utility of this technology relates to the efficient synthesis of hydrocarbons, etc. using an engineered reversal of the β-oxidation cycle, which in turn will establish a new paradigm for the production of advanced biofuels. The ubiquitous nature of β-oxidation enzymes should enable the combinatorial synthesis of non-native products in industrial organisms with a minimum number of foreign genes, an approach that would ensure the efficient functioning of the engineered pathways. By enabling the production of products through a functional reversal of the β-oxidation cycle, this technology will contribute to the creation of fundamentally new approaches that could enable efficient production of second-generation biofuels.

We take the above research forward in this disclosure, adding further diversification of enzymes to be used as part of this pathway.

SUMMARY OF THE DISCLOSURE

This disclosure demonstrates that enzymes from the type II (a discrete set of enzymes) fatty acid synthesis (“FAS”) pathway can be used in combination with thiolases to operate a functional reversal of the β-oxidation cycle. Specifically, a combination of thiolases with one or more of 3-oxoacyl-[acyl-carrier-protein] reductase (FabG, others), 3-hydroxyacyl-[acp] dehydratase (FabA, FabZ, others), and enoyl-[acyl-carrier-protein] reductase (FabI, FabK, FabL, FabV, others) yields a functional reversal of the β-oxidation cycle. If only one or two enzymes are used, the remaining enzymes will be traditional beta oxidation enzymes. Once this cycle is coupled with the appropriate priming and termination pathways, the production of carboxylic acids, alcohols, hydrocarbons, amines and their α-, β-, and ω-functionalized derivatives from renewable carbon sources can be achieved.

A thiolase that condenses acetyl-CoA with an acyl-CoA (or functionalized acyl-CoA) of varying chain length. One or more enzymes from the FAS pathway (see Table below) are then used for the dehydrogenation (FabG), dehydration (FabA, FabZ), and reduction (FabI, FabK, FabL, FabV) steps resulting in an acyl-CoA that is 2 carbons longer than the starting unit. The termination enzyme(s) pulls intermediates from the cycle, and makes one or more final product(s).

RXN

FAS ENZYME

BOX ENZYME

catalyzes the reduction of a β-

3-oxoacyl-[acyl-carrier-protein]

3-hydroxyacyl-CoA

ketoacyl-CoA to a (3R)-β-

reductase

dehydrogenase

hydroxyacyl-CoA

catalyzes the dehydration of a

hydroxyacyl-[acyl-carrier-

enoyl-CoA hydratase or

(3R)-β-hydroxyacyl-CoA to a

protein] dehydratase

3-hydroxyacyl-CoA dehydratase

transenoyl-CoA

catalyzes the reduction of a

enoyl-[acyl-carrier-protein]

acyl-CoA dehydrogenase or

transenoyl-CoA to an acyl-CoA

reductase

trans-enoyl-CoA reductase

that is two carbons longer than

the acyl coA primer or starting

unit

The key difference between this and the FAS pathway is that we are circumventing the energy intensive step of the FAS pathway (decarboxylative condensation) with a non-decarboxylative condensation through the aforementioned thiolases. This and the fact that the enzymes will be working with CoA intermediates as opposed to the ACP intermediates used in the FA synthesis pathway are the key distinctions in this remaining a “reverse beta-oxidation pathway.”

Gene or Abbreviation

Definition

Δ

Refers to reduced activity wherein reduced activity is at least an 75% reduction of wild type

activity, and preferably, 80, 85, 90, 95 or 100% reduction. 100% reduction in activity may also

be called knockout or null mutant herein.

abaT

Gene encoding 4-aminobutyrate transaminase from Mus musculus

ACH2

Gene encoding thioesterase from Arabidopsis thaliana

ackA

Gene encoding acetate kinase, required for synthesis of acetate from acetyl-CoA.

acot8

Gene encoding thioesterase from Mus musculus

acr1

Gene encoding a fatty aldehyde-forming acyl-CoA reductases from Acinetobacter calcoaceticus

acrM

Gene encoding a fatty aldehyde-forming acyl-CoA reductases from Acinetobacter sp. strain M-1

adhE

Gene encoding aldehyde/alcohol dehydrogenase, required for synthesis of ethanol from acetyl-

CoA

adhE2

Gene encoding aldehyde/alcohol dehydrogenase from Clostridium acetobutylicum

ald

Gene encoding an aldehyde-forming coenzyme-A thioester reductase from Clostridium

beijerinckii

arcA

Encodes the cytosolic transcription factor of E. coli's ArcAB two-component system, a global

regulator of gene expression under microaerobic and anaerobic conditions. ArcAB regulates the

expression of a large number of operons involved in respiratory and fermentative metabolism

(including repression of fad regulon).

arcB

Encodes the membrane associated sensor kinase and phosphatase of E. coli's ArcAB two-

component system, a global regulator of gene expression under microaerobic and anaerobic

conditions. ArcAB regulates the expression of a large number of operons involved in respiratory

and fermentative metabolism (including repression of fad regulon).

At3g11980

Gene encoding an alcohol-forming coenzyme-A thioester reductase from Arabidopsis thaliana

At3g44560

Gene encoding an alcohol-forming coenzyme-A thioester reductase from Arabidopsis thaliana

At3g56700

Gene encoding an alcohol-forming coenzyme-A thioester reductase from Arabidopsis thaliana

At5g22500

Gene encoding an alcohol-forming coenzyme-A thioester reductase from Arabidopsis thaliana

atoB

Gene encoding an acetyl-CoA acetyltransferase

atoC

Encodes the cytosolic transcription factor of E. coli's AtoSC two-component system, which

induces the ato operon (atoDAEB) for metabolism of short chain fatty acids in response to the

presence of acetoacetate.

atoC(c)

atoC mutant that induces constitutive expression of the ato operon (atoDAEB) in the absence of

acetoacetate.

atoD

Gene encoding acetyl-CoA:acetoacetyl-CoA transferase

atoS

Encodes the membrane associated sensor kinase of E. coli's AtoSC two-component system,

which induces the ato operon (atoDAEB) for metabolism of short chain fatty acids in response to

the presence of acetoacetate.

betA

Gene encoding choline dehydrogenase; used as a surrogate of alcohol dehydrogenase in the

synthesis of n-alcohols

bktB

Gene encoding beta-ketothiolase from Ralstonia eutropha

buk

Gene encoding carboxylate kinase from Clostridium acetobutylicum or Enterococcus faecalis

cat1

Gene encoding Succinyl-CoA:coenzyme A transferase from Clostridium kluyveri

cat2

Gene encoding 4-hydroxybutyrate CoA transferase from Clostridium kluyveri

catF

Gene encoding beta-ketoadipyl-CoA thiolase from Pseudomonas sp. Strain B13

CER4

Gene encoding an alcohol-forming coenzyme-A thioester reductase from Arabidopsis thaliana

cddC

Gene encoding alcohol dehydrogenase/oxidase from Rhodococcus ruber SC1

cddD

Gene encoding aldehyde dehydrogenase from Rhodococcus ruber SC1

chnD

Gene encoding alcohol dehydrogenase/oxidase from Acinetobacter sp. SE19

chnE

Gene encoding aldehyde dehydrogenase from Acinetobacter sp. SE19

crp

Encodes transcriptional dual regulator CRP, which upon binding to its allosteric effector cyclic

AMP (cAMP) regulate the expression of about 200 genes (most of them involved in the

catabolism of carbon sources, including the fad regulon).

crp*

crp mutant encoding a cAMP-independent CRP (i.e. CRP*, which does not require cAMP to

regulate gene expression and hence prevents catabolite repression of fad regulon in the

presence of glucose)

ctfB

Gene encoding butyrate-acetoacetate CoA-transferase from Clostridium acetobutylicum

CYP

Gene encoding a carboxylic acid omega hydroxylase family enzyme, such as those from

Arabidopsis (CYP94B1, CYP94C1, and CYP86A subfamily), V. sativa (CYP94A1, CYP94A2),

Nicotiana tabacum (CYP94A5), Ps. hybrida (CYP92B1, CYP703A1), Zea mays (CYP78A1),

C. tropicalis (CYP52A1, CYP52A2), rat (CYP4A1), or human (CYP4A11, CYP4B1, and CYP4F2)

egTER

Gene encoding an NAD(P)H-dependent transenoyl-CoA reductase from T. gracilis

eutE

Gene encoding predicted aldehyde dehydrogenase with high sequence similarity to adhE

eutG

Gene encoding predicted alcohol dehydrogenase

fabA

Gene encoding 3-hydroxyacyl-[acp] dehydratase

fabG

Gene encoding 3-oxoacyl-[acyl-carrier-protein]/p-ketoacyl-[ACP] reductase

fabI

Gene encoding enoyl-[acyl-carrier-protein] reductase

fabK

Gene encoding enoyl-[acyl-carrier-protein] reductase from E. faecalis

fabL

Gene encoding enoyl-[acyl-carrier-protein] reductase from B. subtilis (

fabV

Gene encoding enoyl-[acyl-carrier-protein] reductase from V. cholerea

fabZ

Gene encoding 3-hydroxyacyl-[acp] dehydratase

FACoAR (Maqu2507)

Gene encoding an alcohol-forming coenzyme-A thioester reductase from Marinobacter

aquaeolei VT8

fadA

Gene encoding 3-ketoacyl-CoA thiolase (thiolase I), component of fatty acid oxidation complex

fadB

Gene encoding hydroxyacyl-CoA dehydrogenase, aka fused 3-hydroxybutyryl-CoA epimerase

and delta(3)-cis-delta(2)-trans-enoyl-CoA isomerase and enoyl-CoA hydratase, part of fatty acid

oxidation complex

fadBA

Both fadB and fadA

fadD

Gene encoding acyl-CoA synthetase (long-chain-fatty-acid--CoA ligase), part of fatty acyl-CoA

synthetase complex

fadE

Gene encoding acyl-CoA dehydrogenase, a medium-long-chain fatty acyl-CoA dehydrogenase

fadI

Gene encoding 3-ketoacyl-CoA thiolase, part of fatty acid oxidation complex

fadJ

Gene encoding hydroxyacyl-CoA dehydrogenase, aka fused enoyl-CoA hydratase and

epimerase and isomerase

fadK

Gene encoding short chain acyl-CoA synthetase

fadL

Gene encoding long-chain fatty acid outer membrane transporter

fadM

Gene encoding long-chain acyl-CoA thioesterase

fadR

Gene encoding a dual regulator of fatty acid metabolism, which exerts negative control over the

fad regulon and positive control over expression of unsaturated fatty acid biosynthesis genes

fadR*

fadR mutant that allows expression of the fad regulon in the absence of fatty acids

Fnr

Gene encoding transcriptional dual regulator, regulates genes involved in the transition from

aerobic to anaerobic growth

frdA

Gene encoding fumarate reductase, required for synthesis of succinate from fumarate

fucO

Gene encoding L-1,2-propanediol oxidoreductase

gabT

Gene encoding 4-aminobutyrate transaminase

ldhA

Gene encoding lactate dehydrogenase

mhpF

Gene encoding acetaldehyde dehydrogenase

MIT9313_pmt1231

Gene encoding an aldehyde decarbonylase from or Prochlorococcus marinus MIT9313

mgsA

Gene encoding methylglyoxal synthase; key enzyme in the synthesis of lactate through the

methylglyoxal bypass

MXAN_0191

Gene encoding fatty acid alpha-hydroxylase from Myxococcus xanthus

oleA

Gene encoding the enzyme that catalyzes non-decarboxylative Claisen condensation of CoA-

thioesters in Xanthomonas campestris

oleB

Gene encoding a member of the α/β-hydrolase superfamily in Xanthomonas campestris

oleC

Gene encoding a member of the AMPdependent ligase/synthase superfamily or acetyl-CoA

synthetase-like superfamily in Xanthomonas campestris

oleD

Gene encoding a member of the short-chain dehydrogenase/reductase superfamily in

Xanthomonas campestris

paaJ

Gene encoding β-ketoadipyl-CoA thiolase catalyzing two beta-oxidation steps in phenylacetate

catabolism

pcaF

Gene encoding acetyl-CoA acetyltransferase from Rhodococcus opacus or beta-ketoadipyl-CoA

thiolase from Streptomyces sp. or Pseudomonas putida

PCC73102

Gene encoding an aldehyde decarbonylase from Nostoc punctiforme PCC73102

npun_R1711

PCC7942_orf1593

Gene encoding an aldehyde decarbonylase from Synechococcus elongatus PCC7942

pduL

Gene encoding phosphotransacylase from Salmonella enterica

pduW

Gene encoding carboxylate kinase from Salmonella enterica

phaA

Gene encoding beta-ketothiolase from Ralstonia eutropha

pmAD

Gene encoding an aldehyde decarbonylase from Prochlorococcus marinus MIT9313

poxB

Gene encoding pyruvate oxidase, which catalyzes the oxidative decarboxylation of pyruvate to

form acetate, reduced ubiquinone (ubiquinol), and CO₂

pta

Gene encoding phosphotransacetylase, required for synthesis of acetate from acetyl-CoA

ptb

Gene encoding phosphotransacylase from Clostridium acetobutylicum or Enterococcus faecalis

PTE1

Gene encoding thioesterase from S. cerevisiae

STIAU_3334

Gene encoding fatty acid alpha-hydroxylase from Stigmatella aurantiaca

tdTER

Gene encoding an NAD(P)H-dependent transenoyl-CoA reductase from T. denticola

tesA

Gene encoding multifunctional acyl-CoA thioesterase I and protease I and lysophospholipase L1

tesB

Gene encoding thioesterase II from E. coli or A. Borkumensis

thlA

Gene encoding acetoacetyl-CoA thiolase from Clostridium acetobutylicum

thlB

Gene encoding acetoacetyl-CoA thiolase from Clostridium acetobutylicum

ucpA

predicted oxidoreductase, sulfate metabolism protein (used as surrogate of aldehyde-forming

acyl Coenzyme A Reductase)

yahK

Gene encoding alcohol dehydrogenase/oxidase

ybbO

predicted oxidoreductase with NAD(P)-binding Rossmann-fold domain (used as surrogate of

aldehyde-forming acyl Coenzyme A Reductase)

ybdH

Gene encoding predicted oxidoreductase

ybgC

Gene encoding esterase/thioesterase

yciA

Gene encoding acyl-CoA thioesterase

ydiF

Gene encoding predicted acetyl-CoA:acetoacetyl-CoA transferase

ydiI

Gene encoding esterase/thioesterase

ydiL

Gene encoding fused predicted acetyl-CoA:acetoacetyl-CoA transferase: α subunit/β subunit

ydiO

Genes encoding predicted acyl-CoA dehydrogenase

ydiQ

Gene encoding putative subunit of YdiQ-YdiR flavoprotein

ydiR

Gene encoding putative subunit of YdiQ-YdiR flavoprotein

ydiS

Gene encoding putative flavoprotein

ydiT

Gene encoding putative ferredoxin

yiaY

Gene encoding predicted Fe-containing alcohol dehydrogenase

yjgB

Gene encoding alcohol dehydrogenase/oxidase

yqeF

Gene encoding predicted acetyl-CoA acetyltransferases

yqhD

Gene encoding NADP-dependent aldehyde/alcohol dehydrogenase

+

Refers to an overexpressed activity, meaning at least 150% wild type activity, and preferably

200, 500, 1000% or more.

See also Table A for various termination enzymes

As used herein a “reverse beta oxidation” or BOX-R cycle is a pathway that results in the synthesis of fatty acids by adding 2 carbon units to a primer molecule in each turn of the cycle. BOX-R uses acetyl-CoA as the extender unit, added on by a thiolase that uses a non-decarboxylating mechanism. Further, as noted above, any of the BOX-R intermediates can be pulled out of the cycle and further modified in the many ways shown herein.

By contrast, the fatty acid biosynthesis pathway uses keto-acyl-ACP synthases, which employ a decarboxylating mechanism, and malonyl-ACP as the extender or donar unit.

As used herein, a “primer” is a starting molecule for the BOX-R cycle to add two carbon donar units to. The initial primer is either typically acetyl-CoA or propionyl-CoA, but as the chain grows by adding donar units in each cycle, the primer will accordingly increase in size. In some cases, the bacteria can also be provided with larger primers, e.g, C4 primers, etc. added to the media or obtained from other cell pathways. Further, non-traditional primers can be used wherever atypical products are desired (i.e., hydroxylated primers, carboxylated primers, etc. . . . ). As used herein, the “donar” of the 2 carbon units is acetyl-CoA.

As used herein “type II fatty acid synthesis enzymes” refer to those enzymes that function independently, e.g., are discrete, monofunctional enzymes, used in fatty acid synthesis. Type II enzymes are found in archaea and bacteria. Type I systems, in contrast, utilise a single large, multifunctional polypeptide.

As used herein, a “thiolase” is an enzyme that catalyzes the condensation of a acyl-CoA or other primer with a 2-carbon donor acetyl-CoA to produce a β-ketoacyl-CoA in a non-decarboxylative condensation reaction.

Many examples of thiolase enzymes are provided herein and the following table provides several examples:

TABLE A

Example Thiolase Enzymes (EC Number 2.3.1.—)

Protein

Source organism

Accession

and gene name

Numbers

E. coli atoB

NP_416728.1

E. coli yqeF

NP_417321.2

E. coli fadA

YP_026272.1

E. coli fadI

NP_416844.1

Ralstonia eutropha bktB

AAC38322.1

Pseudomonas sp. Strain B13 catF

AAL02407.1

E coli paaJ

NP_415915.1

Pseudomonas putida pcaF

AAA85138.1

Rhodococcus opacus pcaF

YP_002778248.1

Streptomyces sp. pcaF

AAD22035.1

Ralstonia eutropha phaA

AEI80291.1

Clostridium acetobutylicum thlA

AAC26023.1

Clostridium acetobutylicum thlB

AAC26026.1

As used herein a “3-oxoacyl-[acyl-carrier-protein] reductase” or “3-oxoacyl-[ACP] reductase” is an enzyme that catalyzes the reduction of a β-ketoacyl-CoA to a (3R)-β-hydroxyacyl-CoA:

As used herein, a “3-hydroxyacyl-[ACP] dehydratase” is an enzyme that catalyzes the dehydration of a (3R)-β-hydroxyacyl-CoA to a transenoyl-CoA:

As used herein, an “enoyl-[ACP] reductase” that catalyzes the reduction of a transenoyl-CoA to an acyl-CoA:

Many examples of FAS enzymes catalyzing these reactions are provided herein and the following table provides several examples

TABLE B

Example FAS Enzymes for use in BOX-R (See

above for reaction illustrations)

Protein

Source organism

Accession

Reaction

and gene name

Numbers

β-ketoacyl-CoA →

E. coli fabG

NP_415611.1

(3R)-β-hydroxyacyl-CoA

(3R)-β-hydroxyacyl-CoA →

E. coli fabA

NP_415474.1

transenoyl-CoA

E. coli fabZ

NP_414722.1

transenoyl-CoA →acyl-CoA

E. coli fabI

NP_415804.1

Enterococcus faecalis

NP_816503.1

fabK

Bacillus subtilis fabL

KFK80655.1

Vibrio cholerae fabV

ABX38717.1

As used herein “termination pathway” refers to one or more enzymes (or genes encoding same) that will pull reaction intermediates out the BOX-R cycle and produce the desired end product.

By “primary termination pathway” what is meant is a intermediate from the BOX-R is pulled out of the BOX-R by one (which can have more than one activity) or more termination enzymes and results in i) carboxylic acids, ii) primary alcohols, iii) hydrocarbons, or iv) primary amines, from CoA intermediates as described in FIG. 1.

By “secondary termination pathway” what is meant is that the intermediate pulled out of the BOX-R by a primary termination pathway enzyme is further modified by one or more enzymes.

Many examples of termination pathways are provided herein and the following table provides several examples:

TABLE C

Termination Pathways

Protein

EC

Enzyme

Source organism

Accession

Reaction

Illustration

Numbers

names

and gene name

Numbers

Acyl-CoA4 → Carboxylic acid

3.1.2.-

Thioesterase

E. coli tesA E. coli tesB E. coli yciA E. coli fadM E. coli ydil E. coli ybgC

NP_415027.1 NP_414986.1 NP_415769.1 NP_414977.1 NP_416201.1 NP_415264.1

Alcanivorax

YP_692749.1

borkumensis tesB2 Fibrobacter succinogenes Fs2108 Prevotella ruminicola Pr655

  YP_005822012.1     YP_003574018.1  

Prevotella

YP_003574982.1

ruminicola Pr1687

2.8.3.8

Acyl-

E. coli atoD

NP_416725.1

CoA:acetyl-

Clostridium kluyveri

AAA92344.1

CoA

cat2

transferase

Clostridium

NP_149326.1,

acetobutylicum

NP_149327.1

ctfAB

E. coli ydiF

NP_416209.1

2.3.1.-;

Phospho-

Clostridium

NP_349676.1

2.7.2.1;

transacylase +

acetobutylicum ptb

2.7.2.15

Carboxylate

Enterococcus

AAD55374.1

kinase

faecalis ptb

Salmonella enterica

AAD39011.1

pduL

Clostridium

AAK81015.1

acetobutylicum buk

Enterococcus

AAD55375.1

faecalis buk

Salmonella enterica

AAD39021.1

pduW

Acyl-CoA → Alcohol

1.2.1.84

Alcohol- forming CoA reductase

Clostridium acetobutylicum adhE2 Arabidopsis thaliana At3g11980 Arabidopsis

YP_009076789.1     AEE75132.1   AEE77915.1

thaliana At3g44560

Arabidopsis thaliana At3g56700 Arabidopsis thaliana At5g22500

AEE79553.1   AED93034.1  

Arabidopsis

AEE86278.1

thaliana CER4

Marinobacter

YP_959486.1

aguaeolei VT8

maqu_2220

Marinobacter

YP_959769.1

aguaeolei VT8

maqu_2507

Acyl-CoA → Aldehyde

1.2.1.10

Aldehyde forming CoA reductase

Acinetobacter calcoaceticus acr1 Acinetobacter sp Strain M-1 acrM Clostridium beijerinckii ald

AAC45217.1   BAB85476.1   AAT66436.1  

E. coli eut E

NP_416950.1

Salmonella enterica eutE E. coli mhpF

AAA80209.1   NP_414885.1

Aldehyde → Alcohol

1.1.1.-

Alcohol dehydrogenase

E. coli betA E. coli dkgA E. coli eutG E. coli fucO E. coli ucpA E. coli yahK

NP_414845.1 NP_417485.4 NP_416948.4 NP_417279.2 NP_416921.4 NP_414859.1

E. coli ybbO

NP_415026.1

E. coli ybdH E. coli yiaY E. coli yjgB

NP_415132.1 YP_026233.1 NP_418690.4

Aldehyde → Alkane

4.1.99.5

Aldehyde decarbonylase

Synechococcus elongatus PCC7942 orf1593 Nostoc punctiforme PCC73102 npun_R1711

Q54764.1     B2J1M1.1    

Prochlorococcus

Q7V6D4.1

marinus MIT9313 pmt1231

Aldehyde → Amine

2.6.1.-

Transaminase

Arabidopsis thaliana At3g22200 Alcaligenes

NP_001189947.1     AAP92672.1

denitrificans AptA

Bordetella bronchiseptica BB0869 Bordetella

WP_015041039.1     WP_010927683.1

parapertussis

BPP0784

Brucella melitensis

EEW88370.1

BAWG_0478

Burkholderia

AFI65333.1

pseudomallei

BP102613_I0669

Chromobacterium

AAQ59697.1

violaceum CV2025

Oceanicola

WP_007254984.1

granulosus

OG2516_07293

Paracoccus

ABL72050.1

denitrificans

PD1222

Pden_3984

Pseudogulbenkiania

WP_008952788.1

ferrooxidans

ω-TA

Pseudomonas

P28269.1

putida ω-TA

Ralstonia

YP_002258353.1

solanacearum

ω-TA

Rhizobium meliloti

NP_386510.1

SMc01534

Vibrio fluvialis

AEA39183.1

ω-TA

Mus musculus

AAH58521.1

abaT

E. coli gabT

YP_490877.1

Carboxylic Acid → ω-hydroxyacid

1.14-

Carboxylic acid omega hydroxylase

Pseudomonas putida alkBGT   Marinobacter aguaeolei CYP153A Mycobacterium

YP_009076004.1, Q9WWW4.1, Q9L4M8.1 ABM17701.1     YP_001851443.1

marinum

CYP153A16 Polaromonas sp. CYP153A Nicotiana tabacum CYP94A5

  YP_548418.1   AAL54887.1

Vicia sativa

AAD10204.1

CYP94A1

Vicia sativa

AAG33645.1

CYP94A2

Arabidopsis

BAB08810.1

thaliana CYP94B1

Arabidopsis

CAC67445.1

thaliana CYP86A8

Candida tropicalis

AAA63568.1,

CYP52A1

AAA34354.1,

AAA34334.1

Candida tropicalis

AAA34353.2,

CYP52A2

CAA35593.1

Homo sapiens

AAQ56847.1

CYP4A11

ω- hydroxyacid → ω-oxo- acid

1.1.1.-

Alcohol oxidase/ alcohol dehydrogenase

Rhodococcus ruber SC1 cddC Acinetobacter sp. SE19 chnD E. coli yahK E. coli yjgB

AAL14237.1   AAG10028.1   NP_414859.1 NP_418690.4

ω-oxo- acid → dicarboxylic acid

1.2.1.-

Aldehyde dehydrogenase

Rhodococcus ruber SC1 cddD Acinetobacter sp SE19 chn

AAL14238.1     AAG10022.1

Carboxylic Acid → α- hydroxyacid

1.14.-

Carboxylic acid alpha hydroxylase

Myxococcus xanthus MXAN_0191 Stigmatella aurantiaca STIAU_3334

YP_628473.1     YP_003957653.1

As used herein, references to “cells,” “bacteria,” “microbes,” “microorganisms” or “strains” and all such similar designations include progeny thereof. It is also understood that all progeny may not be precisely identical in DNA content, due to deliberate or inadvertent mutations that have been added to the parent. Mutant progeny that have the same function or biological activity as screened for in the originally transformed cell are included. Where distinct designations are intended, it will be clear from the context.

As used herein, “NAD(P)H” means either cofactor could be used, or both, depending on the enzyme selected and availability of the cofactors I the cell.

The terms “operably associated” or “operably linked,” as used herein, refer to functionally coupled nucleic acid sequences.

As used herein “recombinant” is relating to, derived from, or containing genetically engineered material. In other words, the genome (including extrachromosomal elements) was intentionally manipulated by the hand of man in some way.

“Reduced activity” or “inactivation” is defined herein to be at least a 75% reduction in protein activity, as compared with an appropriate control species. Preferably, at least 80, 85, 90, 95% reduction in activity is attained, and in the most preferred embodiment, the activity is eliminated (100%, aka a “knock-out” or “null” mutants). Proteins can be inactivated with inhibitors, by mutation, or by suppression of expression or translation, and the like. Use of a frame shift mutation, early stop codon, point mutations of critical residues, or deletions or insertions, and the like, can completely inactivate (100%) gene product by completely preventing transcription and/or translation of active protein. Reduced activity genes can also be indicated by a minus supercript, e.g. Adh⁻.

“Overexpression” or “overexpressed” is defined herein to be at least 150% of protein activity as compared with an appropriate control species, and preferably 200, 500, 1000%) or more. Overexpression can be achieved by mutating the protein to produce a more active form or a form that is resistant to inhibition, by removing inhibitors, or adding activators, and the like. Overexpression can also be achieved by removing repressors, adding multiple copies of the gene to the cell, or upregulating the endogenous gene, and the like. Increased activity genes can also be indicated by a positive supercript, e.g. PYC⁺.

The term “endogenous” or “native” means that a gene originated from the species in question, without regard to subspecies or strain, although that gene may be naturally or intentionally mutated, or placed under the control of a promoter that results in overexpression or controlled expression of said gene. Thus, genes from Clostridia would not be endogenous to Escherichia, but a plasmid expressing a gene from E. coli would be considered to be endogenous to any genus of Escherichia, even though it may now be overexpressed. By contrast, an “exogenous” gene is from a different species.

The use of the word “a” or “an” when used in conjunction with the term “comprising” in the claims or the specification means one or more than one, unless the context dictates otherwise. The use of the term “or” in the claims is used to mean “and/or” unless explicitly indicated to refer to alternatives only or if the alternatives are mutually exclusive.

The term “about” means the stated value plus or minus the margin of error of measurement or plus or minus 10% if no method of measurement is indicated.

The terms “comprise”, “have”, “include” and “contain” (and their variants) are open-ended linking verbs and allow the addition of other elements when used in a claim. The phrase “consisting of” is closed, and excludes all additional elements. The phrase “consisting essentially of” excludes additional material elements, but allows the inclusions of non-material elements that do not substantially change the nature of the invention, such as instructions for use, buffers, and the like.

The following viations are used herein:

ABBREVIATION

TERM

BOX-R

Beta oxidation pathway in reverse.

FA

Fatty acid

FAD

Flavin adenine dinucleotide

FAS

Fatty acid synthesis

ACP

Acyl carrier protein

BOX

Beta oxidation

ENR

enoyl-acyl carrier protein reductase

PCT

propionate CoA transferase

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A. Modified BOX-R cycle with type II fatty acid biosynthesis enzymes. AtoB, BktB, YqeF, FadA: example of overexpressed thiolases that catalyze the non-decarboxylative condensation of an acyl-CoA primer with 2-carbon donor acetyl-CoA to produce a β-ketoacyl-CoA; FabG: example of overexpressed 3-oxoacyl-[acyl-carrier-protein] reductase that catalyzes the reduction of a β-ketoacyl-CoA to a (3R)-β-hydroxyacyl-CoA; FabA, FabZ: example of overexpressed 3-hydroxyacyl-[acyl-carrier-protein] dehydratases that catalyzes the dehydration of a (3R)-β-hydroxyacyl-CoA to a trans-enoyl-CoA; FabI, FabK, FabL, FabV: example of overexpressed enoyl-[acyl-carrier-protein] reductases that catalyzes the reduction of a trans-enoyl-CoA to an acyl-CoA; Thioesterase: example of overexpressed termination pathway.

FIG. 1B. Primary termination pathways. Pathways that act on the CoA thioester group/carbon, resulting in the synthesis of i) carboxylic acids, ii) primary alcohols, iii) hydrocarbons, and iv) n primary amines, along with their β-hydroxy, β-keto, and α,β-unsaturated derivatives are illustrated.

FIG. 1C. Secondary termination pathways continuing from the primary pathways shown in FIG. 1B. Pathways for the production of omega-hydroxylated carboxylic acids (va), dicarboxylic acids (vii), omega-hydroxylated primary amines (ix), and omega carboxylic acid primary amines (viiib) along with their β-hydroxy, β-keto, and α,β-unsaturated derivatives from the carboxylic acids (i) and primary amines (iv) generated from BOX-R with primary termination pathways are illustrated.

FIG. 1D. Secondary termination pathways. Pathways for the production of omega-hydroxylated primary alcohols (vi), omega carboxylic acid primary alcohols (vb), and omega amino primary alcohols (viiia) along with their β-hydroxy, β-keto, and α,β-unsaturated derivatives from the primary alcohols (ii) generated from BOX-R with primary termination pathways are illustrated.

FIG. 1E. Secondary termination pathways. Pathways for the production of alpha-hydroxylated carboxylic acids (x), alpha-hydroxylated primary alcohols (xii), and alpha-hydroxylated primary amines (xi) along with their β-hydroxy, β-keto, and α,β-unsaturated derivatives from the carboxylic acids (i), primary alcohols (ii), and primary amines (iv), generated from BOX-R with primary termination pathways are illustrated.

FIG. 2. Impact of deletion and over-expression of genes encoding enzymes potentially responsible for the reduction of crotonyl-CoA on cell growth and product synthesis by strain JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA. Chromosomal gene deletions (Δ“gene”) or plasmid based (vector pETDuet) gene overexpression are indicated. Strain JC01 is a fermentation-deficient derivate of E. coli MG1655 containing the following gene deletions: ΔldhA, ΔpoxB, Δpta, ΔadhE, and ΔfrdA.

FIG. 3. Characterization of crotonyl-CoA reduction to butyryl-CoA by FabI. FIG. 3A. Mass spectrum analysis of FabI reduction of crotonyl-CoA to butyryl-CoA. (i) Controls of crotonyl-CoA (836.3 m/z—gray) and butyryl-CoA (838.3 m/z—black). (ii) Reaction mixture without FabI. (iii) Reaction mixture with FabI. FIG. 3B. pH activity profile of FabI with crotonyl-CoA and NADH relative to pH 7.5 (1.0). FIG. 3C. Inhibition profile of crotonyl-CoA reduction activity for FabI with the indicated CoA thioesters at either 20 or 200 μM relative to activity in absence of inhibitor (100%).

FIG. 4A-B. Synthesis of longer chain carboxylic acids via multiple turns of a BOX-R using FabI as the enoyl-CoA reductase. FIG. 4A. Both short-chain (AtoB) and long-chain (FadA) thiolases were activated in strain JC01(DE3) atoB^CT5fadBA^CT5 to facilitate operation of several turns of the BOX-R. The impact of deletion of fadD and yciA on product synthesis is also shown. Gene deletion represented by Δ“gene”. All strains included either pETDuet or pETDuet-fabI (indicated by fabI⁺) vector. FIG. 4B. Production of odd chain carboxylic acids is facilitated in strain JC01(DE3) atoB^CT5fadBA^CT5 ΔfadD ΔyciA pETDuet-fabI when mePCT is co-expressed from pCDFDuet and the medium is supplemented with 15 mM propionic acid.

FIG. 5. Cell growth and product synthesis for strain JC01(DE3) and its derivatives demonstrating the FabI-mediated production of carboxylic acids proceeds through a functional reversal of the β-oxidation cycle. Modifications to JC01(DE3) are indicated by either a gene deletion (Δ“gene”) or overexpression (chromosome or vector pETDuet).

FIG. 6. Operation of a functional reversal of the BOX-R using different enoyl-ACP reductases from the type II fatty acid biosynthesis pathway. FIG. 6A. Butyrate production upon overexpression of enoyl-ACP reductases from Enterococcus faecalis (efFabK), Vibrio cholerae (vcFabV), and Bacillus subtilis (bsFabL) in strain JC01(DE3) atoB^CT5fadB^CT5 ΔfadA; FIG. 6B. Longer chain carboxylic acids in JC01(DE3) atoB^CT5fadBA^CT5 ΔfadD ΔyciA upon the overexpression of fabI, effabK or vcfabV. Gene overexpression from pETDuet indicted by listed gene.

FIG. 7. BOX-R with the use of enzymes from the type II FAS pathway. Carboxylic acid products from a one-turn reversal are only observed with the overexpression of FAS enzymes (FabG, FabZ, FabI) in combination with thiolase (AtoB) overexpression. fadB and fadJ deleted in the case of thiolase expression to ensure product formation is not proceeding through β-oxidation enzymes. Data shown for strain JC01(DE3) with indicated overexpression (+) and deletions (Δ). fabG, fabZ, and fabI overexpressed from pETDuet vector while atoB expression through cumate controlled chromosomal construct at atoB locus.

FIG. 8. Multiple cycle BOX-R with the use of enzymes from the type II FAS pathway for the production of longer acid carboxylic acids. The use of longer chain specific thiolase BktB from Ralstonia eutropha in place of AtoB enables the multiple cycle turns leading to carboxylic acid synthesis with endogenous thioesterase termination. Data shown for strain JC01(DE3) ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ) with either atoB or btkB overexpression under cumate control at the atoB chromosomal locus. The vectors (pET or pCDF) have 2 promotors that genes can be cloned behind. P1 and P2 indicate first or second promoter use.

FIG. 9. Primary alcohol production through BOX-R with the use of enzymes from the type II FAS pathway with aldehyde-forming acyl-CoA reductase and alcohol dehydrogenase termination. Product chain length controlled through use of thiolase enzymes (AtoB or BktB) with varying chain length specificity. Data shown for strain JC01(DE3) ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ) with either atoB or btkB overexpression under cumate control at the atoB chromosomal locus and Clostridium beijerinckii aldehyde-forming acyl-CoA reductase ALD (cbjALD) and E. coli alcohol dehydrogenase FucO (fucO) overexpressed from pCDFDuet vector (pCDF-P1-cbjALD-P2-fucO).

FIG. 10. Odd chain carboxylic acid production with propionate priming through BOX-R with the use of enzymes from the type II FAS pathway. Data shown for strain JC01(DE3) bktB^CT5 ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ) in the presence of 20 mM propionate with (Activation) or without (No Activation) vector for expression M. elsdenii propionyl-CoA transferase (mePCT) for the activation of propionate to propionyl-CoA. mePCT overexpressed from pCDFDuet vector (pCDFDuet-mePCT). Note: pCDPDuet may be abbreviated pCDF throughout.

FIG. 11. Primary odd-chain alcohol production with propionate priming through a BOX-R with type II FAS enzymes and alcohol-forming acyl-CoA reductase termination. Data shown for strain JC01(DE3) bktB^CT5 ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ) in the presence of 20 mM propionate with pCDF-P1-mePCT-P2-maqu2507 vector for the overexpression of M. elsdenii propionyl-CoA transferase (mePCT) for the activation of propionate to propionyl-CoA and M. aquaeolei VT8 alcohol-forming acyl-CoA reductase Maqu2507 (maqu2507) for alcohol forming termination pathway.

FIG. 12. Synthesis of ω-functionalized products through BOX-R with type II FAS enzymes and ω-oxidation termination. Endogenous thioesterase termination combined with ω-oxidation pathways enables the synthesis of ω-hydroxyacids of varying chain length. ω-hydroxyacids products shown for strain JC01(DE3) bktB^CT5 ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ) expressing P. putida alkane monooxygenase (AlkBGT) in pCDFDuet vector (pCDF-P1-alkBGT).

FIG. 13. Genetic constructs for the synthesis of ω-functionalized products with succinate priming through BOX-R with enzymes from type II FAS. FIG. 13A. Chromosomal expression construct for cumate-controlled overexpression of P. putida thiolase PcaF at the E. coli atoB locus. FIG. 13B. pCDFDuet vector for the overexpression of C. kluyveri succinyl-CoA:coenzyme A transferase Cat1 for the activation of succinate to succinyl-CoA (pCDF-P1-cat1).

DETAILED DESCRIPTION

The technology herein is based on developing an alternative strategy to the efficient production of α-, β-, and ω-functionalized carboxylic acids, alcohols, hydrocarbons, and amines that focuses on the use of type II fatty acid biosynthesis pathway genes/enzymes in E. coli and S. cerevisiae (as examples) to assemble a functional reversal of the β-oxidation cycle by combining these type II fatty acid biosynthesis enzymes with the non-decarboxylating condensation reaction catalyzed by thiolases. However, these pathways are ubiquitous, and any type II FAS enzyme from any species can be used.

Technologies developed prior to this are based on the native version of the FAS pathway, which uses a decarboxylative condensation step catalyzed by keto-acyl-ACP/CoA synthases. However, the operation of this pathway is less efficient because it consumes ATP in the synthesis of malonyl-ACP, which is the donor of two-carbon units for chain elongation/decarboxylating condensation reaction. As a consequence, the ATP yield associated with the production of products through the FAS pathway is very low. This, in turn, greatly limits cell growth and production of products.

In more detail, the recombinant engineering to make the BOX-R is:

1) Express the Enzymes Required for a Functional β-Oxidation Cycle Reversal Using Enzymes of the Type II Fatty Acid Biosynthesis Pathway.

Previously, expression of the β-oxidation cycle was accomplished through an approach in which, first 1) mutations fadR and atoC(c) enable expression of the genes encoding beta oxidation enzymes in the absence of fatty acids; 2) an arcA knockout (ΔarcA) enabled the expression of genes encoding beta oxidation cycle enzymes/proteins under anaerobic/microaerobic conditions (microaerobic/anaerobic conditions are used in the production of fuels and chemicals but lead to repression of beta oxidation genes by ArcA); and 3) replacement of native cyclic AMP receptor protein (crp) with a cAMP-independent mutant (crp*) enabled the expression of genes encoding beta oxidation cycle enzymes/proteins in the presence of a catabolite-repressing carbon source such as glucose (glucose is the most widely used carbon source in fermentation processes and represses the beta oxidation genes).

However, the cycle can also be made to run in reverse by individually overexpressing the rate limiting enzymes as opposed to this regulatory approach. Furthermore, the active enzymes can be purified and combined in vitro and made to run the BOX-R pathway in a test tube or flask, or resting cells could be used as a bioreactor for same.

The overall idea in this disclosure, however is to replace some or all genes (other than the thiolase) with type II FAS enzymes (i.e. 3-oxoacyl-[acyl-carrier-protein] (FabG, others), 3-hydroxyacyl-[acyl-carrier-protein] dehydratase (FabA, FabZ, others), and enoyl-[acyl-carrier-protein] reductase (FabI, FabK, FabL, FabV)).

Thus, for a functional reversal of the BOX-R using type II FAS enzymes, 1) non-decarboxylative thiolase(s) are expressed in combination with 2) 3-oxoacyl-[acyl-carrier-protein]/p-ketoacyl-[ACP] reductase (FabG, others), 3) 3-hydroxyacyl-[acp] dehydratase (FabA, FabZ, others), and 4) enoyl-[acyl-carrierprotein] reductase/enoyl-ACP reductase (FabI, FabK, FabL, FabV) to enable the generation of a diverse set of CoA thioester intermediates (FIG. 1). This approach ensures that the key requirements of a beta-oxidation reversal, non-decarboxylative condensation and the use of CoA thioester intermediates, are preserved even with the use of type II FAS enzymes. See e.g., FIG. 1A.

2) Driving the Beta Oxidation Cycle in the Reverse/Biosynthetic Direction (as Opposed to its Natural Catabolic/Degradative Direction).

In addition to functionally expressing the β-oxidation cycle with non-decarboxylative thiolase(s) and type II fatty acid biosynthesis enzymes, we propose the following modifications to improve yields on the reverse operation of this pathway: 1) the use of microaerobic/anaerobic conditions prevents/minimizes the metabolism of acetyl-CoA through the tricarboxylic acids (TCA) cycle and makes acetyl-CoA available to drive the beta oxidation cycle in the reverse/biosynthetic direction; 2) pta (or ackA or both), poxB, adhE, yqhD, and eutE knockouts block/reduce the synthesis of acetate (Δpta or ΔackA and poxB) and ethanol (ΔadhE, ΔyqhD, and ΔeutE) from acetyl-CoA and therefore make acetyl-CoA available to drive the beta oxidation cycle in the reverse/biosynthetic direction; 3) ldhA, mgsA, and frdA knockouts block/reduce the synthesis of lactate (ΔldhA and ΔmgsA) and succinate (ΔfrdA) from pyruvate and phosphoenolpyruvate, respectively, making more phosphoenolpyruvate and pyruvate available for the synthesis acetyl-CoA and therefore making acetyl-CoA available to drive the beta oxidation cycle in the reverse/biosynthetic direction.

3) Conversion of CoA Thioester Intermediates to the Desired End Products.

Generally speaking, there are several termination enzymes that will pull reaction intermediates out the reverse β-oxidation cycle and produce the desired end product (FIG. 1B-E), and a nonexclusive list is provided in Table A, below.

One or more of these termination enzymes can be overexpressed, as needed depending on the desired end product. The termination enzymes can be native or non-native as desired for particular products, but it may be preferred that the reverse beta oxidation cycle use native genes.

4. Regulation of Product Chain Length.

The chain length of thioester intermediates determines the length of end products, and can be controlled by using appropriate termination enzymes with the desired chain-length specificity. Additionally, chain elongation can be inhibited or promoted by reducing or increasing the activity of thiolases with the desired chain-length specificity. These two methods can be used together or independently. For example:

1) knockout of fadA, fadI, and paaJ to avoid chain elongation beyond 1-to-2 turns of the cycle (generates 4- & 6-carbon intermediates and products, or 5- & 7-carbon intermediates and products, depending on the use of acetyl-CoA or propionyl-CoA as primer/starter molecule) and overexpression of the short-chain thiolases yqeF or atoB and short chains alcohol dehydrogenases such as fucO or yqhD for alcohol production for example.

2) overexpression of thiolase(s) of appropriate substrate length specificity (such as E. coli atoB, E. coli yqeF, E. coli fadA, E. coli fadI, Ralstonia eutropha bktB, Pseudomonas sp. Strain B13 catF, E coli paaJ, Pseudomonas putida pcaF Rhodococcus opacus pcaF, Streptomyces sp. pcaF, Ralstonia eutropha phaA, Clostridium acetobutylicum thlA, or Clostridium acetobutylicum thlB) to promote chain elongation and overexpression of specific chain length termination pathways such as long-chain thioesterases (such as E. coli tesA, tesB, fadM, ybgC or yciA, among others) or long chain alcohol dehydrogenases (such as ucpA, ybbO, yiaY, betA, ybdH or eutG) for carboxylic acid or alcohol production, respectively. The term “appropriate” is used herein to refer to an enzyme with the required specificity toward a given intermediate (i.e. acyl-CoA, enoyl-CoA, hydroxyacyl-CoA, and ketoacyl-CoA) of a specific chain length.

Clomburg et al (2012) focused on bottom-up/synthetic approach to reconstruct a beta-oxidation pathway in reverse (“BOX-R”) and hence address short-coming of an early system-level approach that focused on the use of global regulators. Clomburg et al (2012) was successful on identifying and exploiting native enzymes for 3 of the four steps of the BOX-R: thiolase (e.g. AtoB, FadA), 3-hydroxyacyl-CoA dehydrogenase (FadB) and enoyl-CoA dehydratase (fadB). While both of our previous reports (Dellomonaco et al. 2011 and Clomburg et al. 2012) indicate that there is a native E. coli enzyme that catalyzes the last step of the BOX-R (trans-enoyl-CoAs to acyl-CoAs), the identity of this enzyme remained elusive.

In this study, we identified fabI-encoded enoyl-CoA reductase as the activity responsible for this conversion and demonstrated that enoyl-ACP reductases of different families can support the efficient operation of an engineered functional reversal of the β-oxidation cycle.

Since acetyl-coA dehydrogenases (ACDH) utilize a bound FAD cofactor to reduce the 2,3 enoyl-CoA bonds, they often require additional enzymes such as flavoproteins and ferrodoxins in order to function, although the initial reducing equivalents may be sourced to NAD(P)H. Because these enzymes utilize FAD molecules, the chemical reduction/dehydrogenation can be considered reversible, though that might not be the practical result in vivo and they are generally considered to be slow enzymes. In contrast to ACDHs, trans-enoyl-CoA reductase (TER) enzymes that directly utilize NAD(P)H molecules to reduce enoyl bonds are considered irreversible reactions as exampled by their inability to oxidize butyryl-CoA substrates (Tucci & Martin, 2007) and do not require electron transfer proteins to function.

This effectively irreversible activity for reduction of crotonyl-CoA by TER-like enzymes has previously been identified as a beneficial property in improving butanol titers (Atsumi et al., 2008; Bond-Watts, Bellerose, & Chang, 2011b; Shen et al., 2011). In addition, modeling studies have shown that E. coli engineered for butyrate production by the reversal of β-oxidation cycle would benefit from crotonyl-CoA reduction by enzymes that utilize NADH, given the stoichiometric constraints imposed by the use of ferrodoxins (Cintolesi et al., 2014).

Strains, Plasmids, and Methods

Genomic DNAs from E. faecalis V583 (ATCC 700802), B. subtilis (Ehrenberg) Cohn (ATCC 23857), and V. cholerea N16961 (ATCC 39315), as well as, the M. elsdenii Rogosa strain were acquired from ATCC (Manassas, Va.). E. coli genomic DNA from MG1655 (Kang et al., 2004) and M. elsdenii Rogosa were purified using the Wizard Genomic DNA Purification Kit (PROMEGA™, Madison, Wis.). E. gracilis TER was amplified from a plasmid harboring a codon-optimized egTER synthesized by GENSCRIPT™ (Piscataway, N.J.).

All restriction enzymes were purchased from NEW ENGLAND BIOLABS™ (Ipswich, Mass.). Plates were prepared using LB medium containing 1.5% agar, and antibiotics were included at the following concentrations when appropriate: ampicillin (100 μg/mL), kanamycin (50 μg/mL), spectinomycin (50 μg/mL), chloramphenicol (12.5 μg/mL chromosomal/34 μg/mL plasmids) and zeocin (25-50 μg/mL).

Gene overexpression was achieved by cloning the desired gene(s) into either pETDuet-1 (pET) or pCDFDuet-1 (pCDF) (NOVAGEN™, Darmstadt, Germany) digested with NcoI and EcoRI restriction enzymes utilizing In-Fusion PCR cloning technology (CLONTECH LABORATORIES™, Mountain View, Calif.). These vectors have 2 promotors that genes can be cloned behind, referred to herein as P1 and P2.

Cloning inserts were created via PCR of ORFs of interest from their respective genomic or codon-optomized DNA with Phusion DNA polymerase (THERMO SCIENTIFIC™, Waltham, Mass.). The resulting In-Fusion products were used to transform E. coli Stellar cells (CLONTECH LABORATORIES™) and PCR identified clones were confirmed by DNA sequencing.

Strain JC01 (MG1655 ΔldhA ΔpoxB Δpta ΔadhE ΔfrdA) (Clomburg, Vick, Blankschien, Rodriguez-moya, & Gonzalez, 2012), a derivative of wild-type K12 E. coli strain MG1655 (Kang et al., 2004), was used as the host for all genetic modifications. JC01 (DE3) was constructed from JC01 using a λDE3 Lysogenization Kit (NOVAGEN™) to allow the expression of genes under the T7lac promoter.

Gene knockouts were introduced in JC01 (DE3) and its derivatives by P1 phage transduction (Miller, 1972; Shams Yazdani & Gonzalez, 2008). Single gene knockout mutants from the National BioResource Project (NIG, Japan) (Baba et al., 2006) were used as donors of specific mutations. All mutations were confirmed by polymerase chain reaction and the disruption of multiple genes in a common host was achieved as previously described (Shams Yazdani & Gonzalez, 2008).

Strain JC01 (DE3) was further modified to allow for cumate inducible control (Choi et al., 2010) from the genome. To enable this, first the construction of pUCBB-p^CT5-ntH6-eGFP based on previous BioBrick™ vector designs (Vick et al., 2011).

E. coli atoB and fadB genes were PCR amplified and digested with BglII and NotI and ligated by T4 ligase (Invitrogen, Carlsbad, Calif.) into pUCBB-P^CT5-ntH6-eGFP that was previously digested with BglII and NotI to produce pUCBB-P^CT5-atoB and pUCBB-P^CT5-fadB. The resulting ligation products were used to transform E. coli DH5α (INVITROGEN™, Carlsbad, Calif.) and positive clones identified by PCR were confirmed by DNA sequencing.

To integrate the cumate-controlled atoB and fadB constructs onto the chromosome, first the cumate repressor (cymR), promoter/operator regions (P^CT5), and respective ORFs were PCR amplified, as well as, kanamycin or chloramphenicol drug constructs (respectively via pKD4 and pKD3 (Datsenko & Wanner, 2000). These respective products were linked together via overlap extension PCR to create a final chromosomal targeting construct. Integration of the cumate-controlled constructs was achieved via standard recombineering protocols using strain HME45 and selecting on respective LB drug plates (Thomason et al., 2007).

The gene fadA was separately deleted via recombineering in the HME45 derivative harboring the cumate-controlled fadBA contract by replacement of the fadA ORF with a zeocin-resistance marker amplified from pKDzeo (Magner et al., 2007). All constructs were verified via PCR and sequencing.

All chemicals were obtained from FISHER SCIENTIFIC™ (Pittsburgh, Pa.) and SIGMA-ALDRICH™ (St. Louis, Mo.). Fermentations were conducted as previously described using the identical medium formulation (Clomburg et al., 2012) with the exception of 5.0 μM Isopropyl β-D-1-thiogalactopyranoside (Vector-based expression) and 0.1 mM cumate (chromosomal-based expression) included for induction when appropriate.

Measurement of cell growth, quantification of glycerol and metabolic products by high-performance liquid chromatography (HPLC), quantification of fatty acids (C4-C6) by HPLC, and quantification of fatty acids (C7-C14,C16,C18) by Gas Chromatography—Flame Ionization Detection (GC-FID) were performed as previously described (Clomburg et al., 2012).

For quantification purposes of odd-chain fatty acids, samples were run with and without the C13 fatty acid internal standard, verifying that C13 was not produced. Propionic (C_3:0), Valeric (C_5:0), Enanthic (C_7:0), Pelargonic (C_9:0) and Undecyclic (C_11:0) acid (SIGMA-ALDRICH CO.) standards were used to calibrate HPLC and GC-FID analysis.

For measurement of enzymatic activities from fermentation samples, cells were disrupted as previously reported (Dellomonaco, Clomburg, Miller, & Gonzalez, 2011). The HIS Tagged FabI protein was expressed from pCA24N-FabI from the ASKA collection (Kitagawa et al., 2005) in BL21(DE3) cells (NEW ENGLAND BIOLABS™) induced at OD₆₀₀≈0.4 with 0.1 mM IPTG and shaken O/N at Room Temperature. FabI protein was harvested and purified as previously reported using Talon Metal Affinity Resin (CLONTECH LABORATORIES™) (Clomburg et al., 2012).

Thiolase activity (Wiesenborn, Rudolph, & Papoutsakis, 1988) and β-hydroxybutyryl-CoA dehydrogenase activity (Bond-Watts, Bellerose, & Chang, 2011a) were monitored accordingly as previously described (Clomburg et al., 2012). Trans-enoyl-CoA reductase activity for egTER, FabI, bsFabL, efFabK and vcFabV were monitored by loss of NADH (Bond-Watts et al., 2011a; Clomburg et al., 2012). Acyl-CoA dehydrogenase activity was monitored by measuring the reduction of Methyl-Thiazolyl Blue (MTT) coupled to the oxidation of butyryl-CoA (O'Brien & Frerman, 1977) in cellular extracts disrupted as above with the addition of 5 μM flavin adenine dinucleotide (FAD).

Reactions conditions were as follows: 25 mM Tris pH 7.5, 240 μM MTT, 1.7 mM phanazine methosulfate (PMS), 15 mM sodium cyanide, and 250 μM butyryl-CoA in a final volume of 200 μL at 30° C. All substrates and chemicals for enzyme assays were obtained from FISHER SCIENTIFIC™ and SIGMA-ALDRICH™.

High performance liquid chromatography-mass spectrometry (HPLC-MS) was performed on a BRUKER™ MicroTOP ESI (Fremont, Calif.) with an AGILENT™ 1200 HPLC System (Santa Clara, Calif.). Enzymatic reactions were quenched with 10% of IN HCl and diluted 10 fold with dH₂O, 2 μL of which was used for injection. HPLC-MS assays were performed using a SHIMADZU™ Shim-pack XR-ODS II column (2.0 mm I,d.×75 mm) (Tokyo, Japan) at a flow rate of 0.15 mL/min with a maximum pressure of 380 bar at room temperature. Liquid chromatography was performed with 40 mM ammonium acetate (Buffer A) and methanol (Buffer B). The time course was as follows: 0 min-2.5% methanol, linear gradient to 70% methanol at 7 min, hold at 70% methanol to 12 min, linear gradient to 2.5% methanol at 12.1 min, hold at 2.5% methanol to 24 min.

Chromosomal Expression

Previously, an engineered reversal of BOX-R was constructed in E. coli by utilizing strain JC01, a fermentation-deficient derivative of E. coli K12 strain MG1655 (MG1655 ΔldhA ΔpoxB Δpta ΔadhE ΔfrdA), with the vector level expression of native thiolase (AtoB), 3-hydroxyacyl-CoA dehydrogenase (FadB), and enoyl-CoA hydratase (FadB), as well as the foreign trans-enoyl-CoA reductase from Euglena gracilis (egTER) (Clomburg et al., 2012).

An interesting result of this work was the ability to produce small amounts of butyrate in the absence of an overexpressed acyl-CoA dehydrogenase (ACDH) or trans-enoyl-CoA reductase (TER). Using strain JC01 with the pTH.atoB.fadB vector, the two main products from the reversal of the β-oxidation cycle were 3-hydroxybutyrate (3-HB) (2.5 g/L) and butyrate (0.2 g/L). When E. gracilis trans-enoyl-CoA reductase (egTER) is overexpressed via pTH.atoB.fadB.egTER, a 17-fold increase in butyrate production is observed (3.4 g/L) while virtually eliminating production of the 3-hydroxybutyrate (Clomburg et al., 2012).

The product profile can then be altered to longer-chain carboxylic acids (up to C12) with the inclusion of a longer chain specific thiolase (FadA). This work clearly established thiolases (AtoB, FadA), 3-hydroxyacyl-Coa dehydrogenase (FadB), and enoyl-CoA dehydratase (FadB) as key native enzymes necessary to facilitate the functional reversal of β-oxidation (FIG. 1). However, the identity of native ACHD or TER-like enzyme(s) reported to be able to catalyze the conversion of enoyl-CoA to acyl-CoAs (Clomburg et al., 2012; Dellomonaco et al., 2011) was elusive.

In order to establish a clean, vector-free platform to identify the unknown native ACDH or TER-like enzymes that catalyze the final step of the β-oxidation reversal, we constructed a strain with tunable chromosomal expression of AtoB and FadB by engineering their native chromosomal loci. Using an in-house developed expression system based on a cumate-inducible promoter (Choi et al., 2010) adapted into a BioBrick vector system (Vick et al., 2011), atoB and fadB genomic level expression was altered to be controlled by cumate while the fadA gene was knocked out.

For atoB and fadB, the native promoter was replaced by both the T5 promoter controlled by the CymR operator (CT5) as well as an upstream constitutive expression system for CymR. By removing FadA expression, butyrate production can be utilized as proxy for flux through a single turn of the reversal of β-oxidation cycle, thus facilitating the identification of native enzyme(s) responsible for the reduction of crotonyl-CoA.

Prior to the aforementioned modifications, strain JC01 was modified to include the DE3 cassette to allow the use of Novagen's Duet vector system for the expression of hypothesized ACDHs and TER-like enzymes. The resulting strain will be referred to throughout the manuscript as JC01(DE3) atoB^CT5fadB^CT5 ΔfadA.

Enzyme activities and fermentation profiles of JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA were compared to those of the vector-based expression system in strain JC01(DE3) [pETDuet-atoB pCDFDuet-fadB]. The chromosomal expression strain produced more butyrate (0.63±0.01 g/L) and 3-hydroxybutyrate (4.33±0.04 g/L) than the vector based system (butyrate=0.42±0.03 g/L, 3-hydroxybutyrate=3.23±0.07 g/L) (data not shown). The chromosomal strain had similar 3-hydroxybutyryl-CoA dehydrogenase activity (2.2±0.7 μmol/mg/min) to that of pCDFDuet-fadB (1.7±0.7 μmol/mg/min) while thiolase activity was greater from pETDuet-atoB (25.0±3.2 μmol/mg/min) in comparison to atoB expression from the chromosome (3.8±0.3 μmol/mg/min). The higher product yield from JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA indicates that adequate amounts of thiolase and 3-hydroxybutryl-CoA dehydrogenase activities are produced by chromosomal expression.

Identification of Native Enzyme(s)

Multiple enzymes in E. coli have the potential to reduce a 2,3-trans enoyl-CoA (e.g. crotonyl-CoA) to the corresponding acyl-CoA (e.g. butyryl-CoA). E. coli has two reported ACDHs as part of β-oxidation pathways. FadE is the reported aerobic ACDH (Campbell & Jr, 2002; O'Brien & Frerman, 1977) while YdiO is a member of a butyryl-CoA dehydrogenase (BCD) electron transfer protein (Etf) complex (BCD/EtfBA) necessary for growth on fatty acids under anaerobic conditions (Campbell 2003). Two additional ACDH enzymes are present that could potentially reduce crotonyl-CoA as a promiscuous substrate; the crotonobetaine reductase complex CaiAB (Elssner 1999) and AidB, the isovaleryl-CoA dehydrogenase (Landini 1994; Rohankhedkar 2006), although AidB is reported to not work on butyryl-CoA. In an attempt to identify the native enzyme reducing crotonyl-CoA, the genes encoding these ACDH enzymes were knocked out and their effect on butyrate production in JC01(DE3) atoB^ct5 fadB^ct5 ΔfadA was assessed. Interestingly, the individual deletion of these enzymes had little impact on either butyrate or 3-HB production levels (FIG. 2).

The genes encoding the aforementioned ACDHs were deleted and overexpressed in strain JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA and their effect on butyrate production assessed. The deletions had little impact on either butyrate or 3-hydroxybutyrate production levels and did not impair growth (FIG. 2). The only ACDH over-expression that affected butyrate yield was FadE, which resulted in a 44% increase in comparison to a control with the pETDuet vector alone (FIG. 2). Similarly, only the overexpression of FadE resulted in a change in measurable dehydrogenase activity on butyryl-CoA with a specific activity of 0.019±0.0017 μmol/mg/min, which is an order of magnitude greater than any other tested (Table 3). Since deletion or overexpression of ACDHs did not result in significant changes to butyrate levels when combined with AtoB and FadB overexpression, alternative candidates for enzymes catalyzing the conversion of crotonyl-CoA to butyryl-CoA were explored.

Given the advantageous nature of TER enzymes and the fact that none of the investigated ACDH enzymes significantly altered butyrate production, another potential native enzyme that can catalyze the NAD(P)H-dependent conversion of crotonyl-CoA to butyryl-CoA is FabI, an essential enzyme from the fatty acid biosynthesis pathway (Heath & Rock, 1995). Although the fatty acid biosynthesis pathway utilizes Acyl Carrier Proteins (ACP) as substrates, FabI has been demonstrated to reduce crotonyl-CoA to butyryl-CoA (Helmut Bergler, 1996; Weeks & Wakil, 1968).

Despite the potential for the native enzyme, when crude extracts from JC01(DE3) atoB^ct5 fadB^ct5 ΔfadA were assayed for NAD(P)H dependent reduction of crotonyl-CoA no detectable activity was observed (Table 1). However, this is not necessarily an unexpected result, as FabI only reduces crotonyl-CoA as a promiscuous substrate with a K_M (2.7 mM) (Helmet Bergler 1994) that is roughly 30-fold higher than the concentration of crotonyl-CoA (80 μM) used in the standard assay of trans-enoyl-CoA reductases (80 μM crotonyl-CoA) (Bond-Watts 2011a).

To better assess the potential involvement of FabI, the trans-enoyl-CoA reductase assay was performed using 1 mM crotonyl-CoA, a concentration that is similar to the reported K_M of 2.7 mM. Utilizing this substrate concentration, NADH-dependent crotonyl-CoA reduction activities of 0.100±0.001 μmol/mg/min and 0.080±0.007 μmol/mg/min were observed in crude cell extracts from fermentation samples of JC01(DE3) atoB^ct5 fadB^ct5 ΔfadA harvested at 8 hrs and 48 hrs, respectively (Table 1). These results revealed the potential involvement of FabI in butyrate production from a one-turn reversal of the β-oxidation cycle.

To further investigate the potential role of FabI, vectors containing fabI or egTER were constructed and transformed into JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA to assess butyrate production from a one-turn BOX-R. In FIG. 2 FabI over-expression resulted in significant increase in butyrate production compared to that with only AtoB and FadB expression. Even more surprising, the overexpression of FabI resulted in butyrate production at levels comparable to egTER overexpression (3.66 g/L and 3.64 g/L, respectively), while also showing similar reduction in 3-hydroxybutyrate levels (to 0.9 and 0.5 g/L respectively) (FIG. 2). When comparing the trans-enoyl-CoA reductase activities of these fermentations samples (using the standard 80 μM crotonyl-CoA), fabI-overexpressing cultures showed an activity which is six times lower than the activity observed upon egTER-overexpression (Table 2). Despite these lower levels of activity, the comparable levels of butyrate production with either fabI or egTER overexpression demonstrates the potential of an enoyl-ACP reductase from the type II fatty acid biosynthesis pathway to function with CoA intermediates in the context of a reversal of the β-oxidation cycle.

In order to provide further evidence that promiscuous crotonyl-CoA reduction by FabI is involved in native BOX cycle, fabI deletion is required. Given the essential nature of FabI (this is the only ENR in E. coli), construction of a ΔfabI strain requires complementation with another trans-enoyl-ACP activity. Moreover, to verify the hypothesis that FabI indeed functions as the enoyl-CoA reductase of the BOX-R, the complementing enoyl-ACP reductase should not promiscuously reduce crotonyl-CoA as well. Since no enoyl-ACP reductase has been reported with these explicit characteristics, we investigated the following representative members of bacterial ENR families for their lack of crotonyl-CoA reductase activity: Bacillus subtilis FabL (bsFabL) (Heath 2000), Enterococcus faecalis FabK (efFabK) (Bi 2014, Zhu 2013), and Vibrio cholerae FabV (vcFabV) (Massengo-Tiasse 2008).

Genes encoding each of the aforementioned ENRs were cloned into pETDuet and the resulting vectors transformed into JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA, and the resulting strains assayed for crotonyl-CoA reductase activity using 80 μM crotonyl-CoA. While detectable activity was measured for cells with efFabK or vcFabV, no activity was detected for the cells overexpressing bsFabL (Table 1). Even when assayed with 1 mM crotonyl-CoA, the levels of crotonyl-CoA reductase activity shown by cell extracts of JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA ΔfabI [pETDuet-bsfabL] are similar to the background levels measured for JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA [pETDuet], making bsfabL an ideal candidate to complement a fabI deletion without also providing the ability to promiscuously reduce crotonyl-CoA. As such, JC01(DE3) atoB^CT5fadB^CT5 ΔfadA ΔfabI strains were constructed in the presence of either pETDuet fabI or pETDuet-bsfabL. These vectors complemented the fabI deletion equally well, as evidenced by the isolation of viable transductions as well as the similar growth of both strains during fermentations (FIG. 2). However, significant differences in butyrate production (FIG. 2) and enoyl-CoA reductase activity (Table 1) were observed between the two strains. JC01(DE3) atoB^CT5fadB^CT5 ΔfadA ΔfabI [pETDuet fabI] still produced large amounts of butyrate and exhibited enoyl-CoA reductase activity, characteristics not observed with JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA ΔfabI [pETDuet-bsfabL]. These results strongly support the hypothesis that FabI is the unknown enzyme responsible for enoyl-CoA reduction in the experiments discussed above and as such could also have played a role in product synthesis through the previously engineered BOX-R cycle in E. coli (Clomburg 2012, Delmonaco 2011).

When assayed with NADPH, FabI is able to reduce enoyl-ACPs at an optimum of pH 6.5 on chain lengths of C4 to C10, while not able to reduce enoyl-CoAs. With NADH, FabI has a broader pH range for enoyl-ACP reduction that is optimal at pH 7.5 and is active over a chain length from C4 to C16. NADH is also necessary for FabI to reduce CoA substrates, though the reported catalytic efficiency with crotonyl-CoA (k_cat/K_M=8.52×10¹ M⁻¹ sec⁻¹) is several orders of magnitude lower than that reported with crotonyl-ACP (k_cat/K_M=1.91×10⁶M⁻¹ sec⁻¹) at pH 7.5.

The low catalytic efficiency reported for FabI with crotonyl-CoA contrasts with our observation that this enzyme supports butyrate synthesis fluxes as high as those supported by egTER, an enzyme with a very high catalytic efficiency with crotonyl-CoA (k_cat/K_M=1.2±0.1 M⁻¹ sec⁻¹) (Clomburg et al., 2012). A comparison of crotonyl-CoA reduction activity observed in our cultures to the expected activity based on reported kinetic parameters (Helmet Bergler et al., 1994) revealed an interesting picture. After 8 hours of cultivation, strain JC01(DE3) atoB^ct5 fadB^ct5 ΔfadA exhibited an NADH-dependent crotonyl-CoA reduction activity of 0.100±0.001 μmol/mg/min, which corresponds to an observed rate of 4.88×10⁻⁸ M s⁻¹. The total protein concentration of the same culture was 2.94 mg/mL, from which the concentration of FabI is estimated to be between 6.83×10⁻⁸ M and 6.83×10⁻¹²M (Ishihama et al., 2008). Considering this concentration of FabI in the culture, the reported kcat (0.23 s⁻¹) and Km (2.7 mM) (Helmet Bergler et al., 1994), and a 1 mM crotonyl-CoA concentration, the expected rate of crotonyl-CoA reduction by FabI was calculated to be between 7.86×10⁻¹³M s⁻¹ and 7.86×10⁻⁹M s⁻¹. These calculations indicate that the expected rate calculated from reported kinetic parameters is only a small fraction of the rate observed in our cultures, which warrants re-evaluation of the kinetic parameters of FabI using crotonyl-CoA as the substrate.

N-terminal His-tagged FabI was purified from an ASKA collection vector (Kitagawa et al., 2005) expressed in the strain BL21(DE3). The purified enzyme had a K_M (4.6±0.6 mM) quite similar to the reported K_M of 2.7 mM (Helmet Bergler et al., 1994), but our measurement of k_cat was 5.3±0.6 s⁻¹ in comparison to 0.23 s⁻¹ that was previously reported, resulting in a 13-fold improvement of catalytic efficiency (k_cat/K_M) from 8.52×10¹ M⁻¹ sec⁻¹ to 1.15×10³. This change in k_cat increases the expected rate of crotonyl-CoA reduction by FabI to a value between 1.02×10⁻⁷M s⁻¹ and 7.86×10⁻¹¹ M s⁻¹, which is in agreement with the rate observed in our cultures. To verify that FabI is utilizing NADH to convert crotonyl-CoA to butyryl-CoA, the reaction was monitored via HPLC-MS, which confirmed that FabI converts crotonyl-CoA almost completely to butyryl-CoA (FIG. 3a).

Factors such as the presence of Mg²⁺ during our purification as well as Trisma-HCl vs. phosphate based buffer systems had marginal effect on activity levels (data not shown), and the most likely cause for this discrepancy is that previous studies used a maximum of approximately 1.25 mM crotonyl-CoA, which is less than their reported K_M indicating that there was an inaccurate estimation of V_max.

As an essential gene for FAS, FabI is inhibited when large amounts of palmitic acid (C16) is present, and has been demonstrated to have a K_i of 5.4 μM towards palmitoyl-ACP and 20 μM towards palmitoyl-CoA, but this inhibition is quickly lost for shorter fatty acids as the K_i for decaonyl-CoA is approx 900 μM (Helmut Bergler 1996).

To understand the potential of utilizing FabI for BOX-R, we examined the effect of key CoA thioester intermediates of the BOX-R as well as a broad range of acyl-CoAs on FabI activity (FIG. 3B). Palmitoyl-CoA demonstrated close to 50% inhibition at 20 μM, while myristoyl-CoA (C14) had the highest percent inhibition at 200 μM. The shorter acyl-CoAs thioesters, acetoacetyl-CoA and 3-hydroxy-butyryl-CoA demonstrated minimal inhibitory effects.

The pH profile for FabI was also examined for comparison to the reported profile for crotonyl-ACP (Weeks & Wakil, 1968) (FIG. 3C). Using crotonyl-CoA, FabI has a broad range of activities from pH 5.5 up to 9.0. This range is not dissimilar to the reported pH range with crotonyl-ACP, although the optimal pH switched to pH 6.5 and the enzyme remained active at higher pHs.

FABI Supports Synthesis of C>4 Carboxylic Acids

Employing AtoB as the thiolase limits the operation of the BOX-R pathway to one cycle and product synthesis to C4 molecules. FadA, which has longer chain length specificity, produces longer chain products. When fabI is overexpressed from pETDuet in JC01(DE3) atoB^CT5 fadBA^CT5, the production of extracellular longer chain length carboxylic acids is observed (FIG. 4A). The most abundant carboxylic acid produced is C6 (0.53±0.03 g/L), while C8, C10 and C12 carboxylic acids are produced at 57.8±4.7, 56.8±4.4 and 11.5±0.3 mg/L, respectively (FIG. 4A).

The deletion of fadD has been utilized in numerous studies to facilitate the accumulation of longer chain carboxylic acids by preventing the uptake of free fatty acids from the media (Lennen 2012). Under these conditions, the deletion of fadD in JC01(DE3) atoB^CT5fadBA^CT5 with pETDuet-fabI resulted in the extracellular production of 19.4±0.1 mg/L tetradecanoic acid (C14) and a slight increase in dodecanoic acid (C12), with minimal impact on the levels of C4-C10 carboxylic acids produced (FIG. 4A). In order to alter the product profile away from C4 and C6 carboxylic acids, yciA, encoding a thioesterase previously shown to have a marked effect on short chain product profiles (Clomburg 2012), was also deleted. This resulted in significant increases to C8 (97.4±1.9 mg/L) and C10 (102.5±8.0 mg/L) carboxylic acids while greatly reducing the amount of butyrate (0.47±0.01 g/L) (FIG. 4A). Surprisingly, the JC01 (DE3) atoB^CT5 fadBA^CT5 ΔfadD ΔyciA variant also resulted in a loss of the long-chain carboxylic acids, indicating that YciA may play a role as a longer chain acyl-CoA thioesterase as well as the shorter chains, while other thioesterases are potentially present to remain active on the medium chain acyl-CoAs (e.g. TesA, TesB, FadM).

Production of Odd Chain Carboxylic Acids

In order to investigate the potential for FabI to be used in the production odd-chain carboxylic acids, strain JC01(DE3) atoB^CT5 fadBA^CT5 ΔfadD ΔyciA was further complemented with the propionate CoA transferase (PCT) from M. elsdenii (Taguchi et al., 2008), which has been used to successfully produce a variety of C5 fatty acid molecules (Tseng & Prather, 2012). mePCT was cloned into the pCDFDuet vector to facilitate co-expression with pETDuet-fabI. The two vectors were co-transformed into JC01 (DE3) atoB^CT5 fadBA^CT5 ΔfadD ΔyciA to determine odd chain carboxylic acid production in the presence of 15 mM propionate. This strain produced odd chain carboxylic acids ranging from 187±4 mg/L of valerate (C5) to a maximum chain length of C11 at 30.9±0.6 mg/L (FIG. 4B).

FABI Supports Product Synthesis

One concern when utilizing FabI as a trans-enoyl-CoA reductase is that product synthesis (e.g. butyrate and other products) could actually be proceeding through the type II FAS pathway, instead of the BOX-R. Fortunately, the condensation reactions responsible for carbon chain elongation in these two pathways are very different, providing an opportunity to clearly distinguish them. The BOX-R, which employ acetyl-CoA as the extender unit and a non-decarboxylating condensation mechanism (Binstock 1981, Jenkins 1987). The fatty acid biosynthesis pathway uses keto-acyl-ACP synthases, which employ a decarboxylating condensation mechanism and malonyl-ACP as the extender unit (White 2005).

Overexpression of FabI in strain JC01(DE3) resulted predominantly in the production of acetate, a fermentation profile essentially undistinguishable from that of JC01(DE3) carrying the empty pETDuet vector (FIG. 5). In fact, butyrate production with FabI expression was only observed when a thiolase (AtoB or FadA) was used in conjunction with FadB (FIG. 5). The requirement of a non-decarboxylating condensation enzyme (e.g. thiolases AtoB or FadA), as well as the β-oxidation enzyme FadB (which only functions with CoA substrates) to observe the synthesis of butyrate indicates that FabI is facilitating a reversal of the β-oxidation cycle despite being an enzyme of the type II fatty acid biosynthesis pathway.

Other Enoyl-ACP Reductases

The identification of FabI as an enzyme that enables efficient operation of the β-oxidation reversal prompted us to investigate whether other ENRs from the type II fatty acid biosynthesis pathway can play a similar role. While bsFabL was found to complement a fabI deletion without promiscuous activity on CoA substrates, measurement of crotonyl-CoA reduction activity with strains expressing effabK and vcfabV indicated their potential to support a β-oxidation reversal (Table 1). When overexpressed in JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA, both efFabK and vcFabV resulted in a marked increase in butyrate production, similar to that fabI overexpression (FIG. 6A). As expected based on the results from ΔfabI complementation and NAD(P)H-dependent crotonyl-CoA reduction measurements (Table 1), bsFabL had a profile quite similar to the empty pETDuet vector in that only small amounts of butyrate were produced (0.33±0.02 g/L) (FIG. 6A). Observed differences in butyrate production are not due to an altered growth phenotype as all the strains grew to similar densities. In addition to supporting butyrate production from a one-turn β-oxidation reversal, when pETDuet-effabK and pETDuet-vcfabV were overexpressed in JC01(DE3) atoB^CT5 fadBA^CT5 ΔfadD ΔyciA, longer chain carboxylic acids were produced at levels similar to that with pETDuet-fabI (FIG. 6B).

Other Type II Fatty Acid Synthesis Enzymes

In addition to the use of FabI or other enoyl-[acyl-carrier-protein] reductases from the type II fatty acid biosynthesis pathway, the 3-oxoacyl-[acyl-carrier-protein]/β-ketoacyl-[ACP] reductase (FabG, others) and 3-hydroxyacyl-[ACP] dehydratase (FabA, FabZ, others) enzymes of the type II FAS pathway can be utilized during BOX-R in place of BOX enzyme(s) (FadB). In order to establish the potential of these enzymes for product formation when expressed with non-decarboxylative thiolase(s), enoyl-[acyl-carrier-protein] reductases, and appropriate termination pathways, product formation was first investigated from a one-turn BOX-R with the overexpression of various combinations of FabG, FabZ, and FabI.

For this purpose, pETDuet vectors containing FabG (pET-P1-P2-fabG), FabG and FabZ (pET-P1-P2-fabG-fabZ), and all of FabG, FabZ, and FabI (pET-P1-fabI-P2-fabG-fabZ) were utilized for the overexpression of these enzymes in both JC01(DE3) and a JC01(DE3) variant containing the cumate-controlled atoB expression construct. This latter strain also included deletions to fadB and fabJ to ensure any product formation was not a result of endogenous β-oxidation enzymes (JC01(DE3) atoB^CT5 ΔfadB ΔfadJ). As seen in FIG. 7, the overexpression of FabG, FabZ, and FabI only resulted in β-oxidation reversal product formation (i.e. 3-hydroxybutyrate and/or butyrate) when these enzyme(s) were overexpressed in combination with atoB. Furthermore, the concentration of butyrate, representing the final product of a full one-turn BOX-R with thioesterase termination, increases as FabG is overexpressed in combination with FabZ and further when all 3 FAS enzymes are overexpressed (FabG, FabZ, and FabI). These results provide strong evidence to the fact that a β-oxidation reversal is taking place through a non-decarboxylative condensation mechanism and CoA intermediates even with the use of enzymes from the type II fatty acid biosynthesis pathway.

The ability for BOX-R with the use of type II FAS pathway to support longer chain product synthesis was investigated through the replacement of AtoB with the longer chain specific thiolase BktB from Ralstonia eutropha (Kim 2014). With this thiolase was utilized with FabG, FabZ, and FabI expression in the JC01(DE3) derivative lacking key β-oxidation enzymes (JC01(DE3) btkB^CT5 ΔfadB ΔfadJ) the synthesis of extracellular longer chain carboxylic acids up to ten carbons in length was observed (FIG. 8), demonstrating the ability for FAS enzymes to support longer chain product formation when utilized with a non-decarboxylative thiolase of longer chain specificity.

Other Primary Termination Pathways

The establishment of the ability for three E. coli type II FAS pathway enzymes to support one- and multiple-turn BOX-R cycles when overexpressed with a non-decarboxylative thiolase utilized endogenous termination pathways for the direct conversion of pathway intermediates to carboxylic acids. However, further product diversification can be achieved through the use of additional primary termination pathways for the synthesis of varying product families from β-oxidation reversal intermediates.

To demonstrate this potential, the overexpression of a primary termination pathway for alcohol production was investigated in strains shown to produce one-turn and multiple-turn BOX-R products (i.e. JC01(DE3) atoB^CT5 ΔfadB ΔfadJ or JC01(DE3) btkB^CT5 ΔfadB ΔfadJ containing pET-P1-fabI-P2-fabG-fabZ). For this purpose, the combination of the Clostridium beijerinckii aldehyde-forming acyl-CoA reductase ALD (cbjALD) (Yan 1990) and E. coli alcohol dehydrogenase FucO (Dellomonaco 2011) were cloned into pCDFDuet to provide the ability to co-express all FAS enzymes along with this alcohol producing termination pathway.

As seen in FIG. 9, the overexpression of this termination pathway resulted in primary alcohol production whose chain length was also dependent on the thiolase utilized. With AtoB as the thiolase, the production of butanol was observed from a one-turn β-oxidation reversal, while the use of BktB enabled the synthesis of hexanol and octanol as well as butanol production. In addition to demonstrating the ability for the selection of the primary termination pathway to dictate product synthesis, the use of an aldehyde-forming acyl-CoA reductase provides further evidence to the operation of BOX-R with CoA intermediates even with the use of type II FAS enzymes, as this enzyme directly utilizes CoA substrates as opposed to ACP intermediates.

Production of Odd Chain Products

The ability for type II FAS enzymes to support odd-chain product synthesis through a BOX-R was investigated through the use of propionyl-CoA as the initial primer. This was accomplished through the overexpression of the propionate CoA transferase (PCT) from M. elsdenii (Taguchi et al., 2008) for the activation of propionate to propionyl-CoA. When this enzyme was overexpressed from a pCDF vector in strain JC01(DE3) btkB^CT5 ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ) in the presence of propionate, the synthesis of C5, C7, and C9 carboxylic acids was observed through endogenous termination pathways (FIG. 10). Furthermore, when propionate activation was combined with the overexpression of the alcohol-forming acyl-CoA reductase Maqu2507 from Marinobacter aquaeolei VT8 (Willis 2011) in JC01(DE3) btkB^CT5 ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ), the production of odd-chain primary alcohols was observed (FIG. 11).

Thus, in addition to demonstrating the ability for odd-chain product synthesis through a full BOX-R with enzymes from the type II FAS pathway, the production of alcohols with Maqu2507 represents the use of an additional primary termination pathway. As opposed to alcohol production with the combination of an aldehyde-forming acyl-CoA reductase and an alcohol dehydrogenase, the production of primary alcohols in this case makes use of a bi-functional enzyme for the 2-step reduction of an acyl-CoA to an alcohol.

Secondary Termination Pathways

Another route to additional product diversification is the use of secondary termination pathways to add additional functional groups to the above mentioned products of primary termination pathways (FIG. 1). An example of this is the use of a carboxylic acid omega hydroxylase to introduce omega functionality in the form of an alcohol group to carboxylic acids produced from a β-oxidation reversal with the use of type II fatty acid biosynthesis enzymes and primary termination pathways. While several potential enzymes can be utilized for this purpose, the alkane monooxygenase AlkBGT from Pseudomonas putida was been linked to the ability to omega hydroxylate carboxylic acids (Kusunose 1964; McKenna & Coon 1970) with more recent studies demonstrating the ability for AlkBGT to ω-hydroxylate medium chain length (C5-C12) fatty acid methyl esters (Julsing 2012; Schrewe 2011; Schrewe 2014).

In order to demonstrate the potential for the use of secondary termination to provide increased product functionality and diversity, the alkBGT genes from P. putida were cloned into pCDF to allow for co-expression with all other enzymes required to generate carboxylic acids from a BOX-R with type II FAS enzymes. When AlkBGT was overexpressed in strain JC01(DE3) btkB^CT5 ΔfadB ΔfadJ (pET-P1-fabI-P2-fabG-fabZ), the synthesis of 6-hydroxyhexanoic, 8-hydroxyoctanoic, and 10-hydrodecanoic acids was observed from the omega oxidation of carboxylic acids generated from primary termination pathways (FIG. 12). The production of these compounds demonstrates the potential of secondary termination pathways to generate a diverse set of products dependent on the combination of primary and secondary termination pathways used.

For example, the synthesis of dicarboxylic acids can be achieved through the further oxidation of omega-hydroxyacids, through the use of an alcohol and aldehyde dehydrogenase (Cheng 2000), while omega amino acids such as 6-aminocaproic acid can be synthesized from omega-hydroxyacids through the overexpression of an alcohol dehydrogenase and a transaminase (Schrewe 2013). These product classes represent just a small sample of the potential products that can be generated through a β-oxidation reversal with the use of type II fatty acid biosynthesis enzymes (FIG. 1). As such, the combinatorial selection of various primary and secondary termination pathways can be further exploited to synthesize a wide variety of industrially important fuel and chemical compounds.

Product Diversity Through Functionalized Primers

An additional route to generating further product diversity is the introduction of functional groups at the initial priming step of the BOX-R. For example, the use of a priming molecule such as succinyl-CoA or glycolyl-CoA for initial condensation with acetyl-CoA can introduce ω-carboxyl or ω-hydroxyl groups, respectively, with subsequent β-oxidation cycle turns and primary termination leading to various product classes (FIG. 1). This approach relies on the ability a thiolase to condense a given functionalized acyl-CoA molecule with acetyl-CoA as well as the enzymes involved in the cycle steps (reduction, dehydration, and reduction) to accept substrates with these functional groups.

Several thiolases have been identified with the potential to condense functionalized acyl-CoA molecules such as succinyl-CoA and glycolyl-CoA with acetyl-CoA, including Ralstonia eutropha bktB (Martin 2013), E coli paaJ (Ismail 2010), and Pseudomonas putida pcaF (Harwood 1994) among others. Genetic constructs for the overexpression of these enzymes have been assembled and are currently being investigated along with enzymes for the activation of potential acid primer molecules to their CoA derivatives (such as mePCT and Clostridium kluyveri Cat1) to determine their potential for this route to product synthesis (FIG. 13). These constructs will enable further investigation into the ability of type II FAS enzymes to accept substrates with varying functionalities in this context. In should be noted that type II FAS enzymes have been implicated in the biotin synthetic pathway in which FabG, FabZ, and FabI are involved in the reduction, dehydration, and reduction of various chain length omega methyl ester ACP intermediates (Lin 2010), which holds great potential for the utilization of these enzymes in a BOX-R with functionalized intermediates. This type of priming, combined with various primary and secondary termination pathways provides an additional route to potential expand the product diversity that can be achieved through the BOX-R and type II FAS pathways.

Enzyme activity (μmol/mg protein/min)^a

NADH-Dependent

Trans-enoyl-CoA reductase

Acyl-CoA

Crotonyl-CoA concentration^b

Strain

Vector

dehydrogenase

80 μM

1 mM

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet (8 hours)

0.0034 ± .0001

n.d.^c

 0.10 ± .001

pETDuet (48 hours)

0.0020 ± .0002

n.d.^c

0.080 ± .007

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet

 0.0016 ± .00002

—

—

ΔfadE

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet

0.0018 ± .0001

—

—

ΔydiO

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet

0.0011 ± .0001

—

—

Δcai4

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet

 0.0026 ± .00003

—

—

ΔaidB

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet-fadE

0.019 ± .002

—

—

pETDuet-ydiO

0.0031 ± .0003

—

—

pETDuet-caiA

0.0014 ± .0002

—

—

pETDuet-egTER

—

3.1 ± 0.4

—

pETDuet-fabI

—

 0.5 ± 0.05

 3.4 ± .27

pETDuet-effabK

—

0.05 ± .010

0.14 ± .01

pETDuet-bsfabL

—

n.d.^c

0.019 ± .006

pETDuet-vcfabV

—

0.053 ± .010 

 4.3 ± 1.6

JC01(DE3) atoB^CT5 fadB^CT5 ΔfadA

pETDuet-fabI

0.0010 ± .0032

0.14 ± .006

 1.8 ± .10

ΔfabI

pETDuet-bsfabL

0.0020 ± .0002

n.d. 

n.d.

Each reference below is incorporated by reference its entirety for all purposes.

• Atsumi, S., Cann, A. F., Connor, M. R., Shen, C. R., Smith, K. M., Brynildsen, M. P., . . . Liao, J. C. (2008). Metabolic engineering of Escherichia coli for 1-butanol production. Metabolic engineering, 10(6), 305-11.
• Baba, T., Ara, T., Hasegawa, M., Takai, Y., Okumura, Y., Baba, M., . . . Mori, H. (2006). Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Molecular systems biology, 2, 2006.
• Bergler, H., et al., (1994) Protein EnvM Is the NADH-dependent Enoyl-ACP Reductase (FabI) of Escherichia coli. The Journal of biological chemistry, 269(8), 5493-5496.
• Bergler, H. et al., (1996) which catalyzes a key regulatory step in fatty acid biosynthesis, accepts NADH and NADPH as cofactors and is inhibited by palmitoyl-CoA, 694, 689-694.
• Bi H. K., et al., (2014) Inefficient Translation Renders the Enterococcus faecalis fabK Enoyl-Acyl Carrier Protein Reductase Phenotypically Cryptic. J. Bacteriol. 196:170-179.
• Binstock J. F. & Schulz H. (1981) Fatty acid oxidation complex from Escherichia coli. Methods Enzymol. 71 Pt C:403-411.
• Bond-Watts, B. B., et al., (2011a) Enzyme mechanism as a kinetic control element for designing synthetic biofuel pathways. Nature chemical biology, 7(4), 222-7. doi:10.1038/nchembio.537
• Bond-Watts, B. B., et al., (2011b) Enzyme mechanism as a kinetic control element for designing synthetic biofuel pathways. Nature chemical biology, 7(4), 222-7.
• Campbell, J. W. & Jr, J. E. C. (2002) The Enigmatic Escherichia coli fadE Gene Is yafH, 184(13), 3759-3764.
• Campbell, J. W., et al., (2003) A new Escherichia coli metabolic competency: growth on fatty acids by a novel anaerobic beta-oxidation pathway. Molecular microbiology, 47(3), 793-805.
• Cheng, Q., T et al., (2000) Genetic analysis of a gene cluster for cyclohexanol oxidation in Acinetobacter sp strain SE19 by in vitro transposition. J. Bacteriol. 182, 4744-4751.
• Choi, Y. J., et al., (2010) Novel, versatile, and tightly regulated expression system for Escherichia coli strains. Applied and environmental microbiology, 76(15), 5058-66.
• Clomburg J. M., et al., (2012) A Synthetic Biology Approach to Engineer a Functional Reversal of the β-Oxidation Cycle. ACS Synth. Biol. 1:541-554.
• Datsenko, K. A. & Wanner, B. L. (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proceedings of the National Academy of Sciences of the United States of America, 97(12), 6640-5.
• Dellomonaco, C., et al., (2011) Engineered reversal of the β-oxidation cycle for the synthesis of fuels and chemicals. Nature, 476(7360), 355-359.
• Elssner, T., et al., (2000) Isolation, identification, and synthesis of gamma-butyrobetainyl-CoA and crotonobetainyl-CoA, compounds involved in carnitine metabolism of E. coli. Biochemistry, 39(35), 10761-9.
• Harwood C. S., et al., (1994) Identification of the pcaRKF gene cluster from Pseudomonas putida: involvement in chemotaxis, biodegradation, and transport of 4-hydroxybenzoate. J. Bacteriol. 176, 6479-6488.
• Heath, R. J. & Rock, C. O. (1995) Enoyl-Acyl Carrier Protein Reductase (fabI) Plays a Determinant Role in Completing Cycles of Fatty Acid Elongation in Escherichia coli. Journal of Biological Chemistry, 270(44), 26538-26542.
• Heath R. J., et al., (2000) The enoyl-acyl-carrier-protein reductases FabI and FabL from Bacillus subtilis. J. Biol. Chem. 275:40128-40133.
• Ishihama, Y., et al., (2008) Protein abundance profiling of the Escherichia coli cytosol. BMC genomics, 9, 102.
• Ismail W., et al., (2003) Functional genomics by NMR spectroscopy. Phenylacetate catabolism in Escherichia coli. Eur. J. Biochem. 270, 3047-3054.
• Jenkins L. S. & Nunn W. D. (1987) Genetic and molecular characterization of the genes involved in short-chain fatty-acid degradation in Escherichia coli: The ato system. J. Bacteriol. 169:42-52.
• Julsing, M. K., et al., (2012) Outer Membrane Protein AlkL Boosts Biocatalytic Oxyfunctionalization of Hydrophobic Substrates in Escherichia coli. Appl. Environ. Microbiol. 78, 5724-5733.
• Kang, Y., et al., (2004) Systematic mutagenesis of the Escherichia coli genome. Journal of bacteriology, 186, 4921-4930.
• Kim, E. J., et al., (2014) Crystal structure and biochemical characterization of β-keto thiolase B from polyhydroxyalkanoate-producing bacterium Ralstonia eutropha H16. Biochem. Biophys. Res. Commun 444, 365-369.
• Kitagawa, M., et al., (2005) Complete set of ORF clones of Escherichia coli ASKA library (a complete set of E. coli K-12 ORF archive): unique resources for biological research. DNA research: an international journal for rapid publication of reports on genes and genomes, 12(5), 291-9.
• Kusunose, M., et al., (1964) Enzymatic ω-Oxidation of Fatty Acids: I. Products of Octanoate, Decanoate, and Laurate Oxidation. J. Biol. Chem. 239, 1374-1380.
• Landini, P., et al., (1994) Structure and transcriptional regulation of the Escherichia coli adaptive response gene aidB. J. Bacteriol., 176(21), 6583-9.
• Lennen R. M. & Pfleger B, F. (2012) Engineering Escherichia coli to synthesize free fatty acids. Trends Biotechnol. 30:659-667.
• Lin S., et al., (2010) Biotin synthesis begins by hijacking the fatty acid synthetic pathway. Nat. Chem. Biol. 6, 682-688.
• Magner, D. B., et al., (2007) RecQ promotes toxic recombination in cells lacking recombination intermediate-removal proteins. Molecular cell, 26(2), 273-86.
• Martin C. H., et al., (2013) A platform pathway for production of 3-hydroxyacids provides a biosynthetic route to 3-hydroxy-γ-butyrolactone. Nat. Commun 4, 1414.
• Massengo-Tiasse R. P. & Cronan J. E. (2008) Vibrio cholerae FabV defines a new class of enoyl-acyl carrier protein reductase. J. Biol. Chem. 283:1308-1316. Miller, J. H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press.
• McKenna, E. J. & Coon, M. J. (1970) Enzymatic ω-Oxidation: IV. Purification and Properties of the ω-Hydroxylase of Pseudomonas oleovorans. J. Biol. Chem. 245, 3882-3889.
• O'Brien, W. J. & Frerman, F. E. (1977) Evidence for a complex of three beta-oxidation enzymes in Escherichia coli: induction and localization. Journal of bacteriology, 132(2), 532-40.
• Preusser, A., et al., (1999) Crotonobetaine reductase from Escherichia coli consists of two proteins. Biochimica et biophysica acta, 1431(1), 166-78.
• Rohankhedkar, M. S., et al., (2006) The AidB Component of the Escherichia coli Adaptive Response to Alkylating Agents Is a Flavin-Containing, DNA-Binding Protein, 188(1), 223-230.
• Shams Yazdani, S. & Gonzalez, R. (2008) Engineering Escherichia coli for the efficient conversion of glycerol to ethanol and co-products. Metabolic engineering, 10(6), 340-51.
• Shen, C. R., et al., (2011) Driving forces enable high-titer anaerobic 1-butanol synthesis in Escherichia coli. Applied and environmental microbiology, 77(9), 2905-15.
• Schrewe, M., et al., (2011) Kinetic Analysis of Terminal and Unactivated C—H Bond Oxyfunctionalization in Fatty Acid Methyl Esters by Monooxygenase-Based Whole-Cell Biocatalysis. Adv. Synth. Catal. 353, 3485-3495.
• Schrewe M., at al., (2013) Direct Terminal Alkylamino-Functionalization via Multistep Biocatalysis in One Recombinant Whole-Cell Catalyst. Adv. Synth. Catal. 355, 1693-1697.
• Schrewe, M., et al., (2014) Reaction and Catalyst Engineering to Exploit Kinetically Controlled Whole-Cell Multistep Biocatalysis for Terminal FAME Oxyfunctionalization. Biotechnol. Bioeng. 111, 1820-1830.
• Taguchi, S., et al., (2008) A microbial factory for lactate-based polyesters using a lactate-polymerizing enzyme. PNAS(USA), 105(45), 17323-7.
• Thomason, L., et al., (2007) Recombineering: genetic engineering in bacteria using homologous recombination. Current protocols in molecular biology/edited by Frederick M. Ausubel . . . [et al.], Chapter 1, Unit 1.16.
• Tseng, H.-C., & Prather, K. L. J. (2012). Controlled biosynthesis of odd-chain fuels and chemicals via engineered modular metabolic pathways. Proceedings of the National Academy of Sciences of the United States of America, 109(44), 17925-30. doi:10.1073/pnas.1209002109
• Tucci, S. & Martin, W. (2007) A novel prokaryotic trans-2-enoyl-CoA reductase from the spirochete Treponema denticola. FEBS letters, 581(8), 1561-6.
• Vick, J. E., et al., (2011) Optimized compatible set of BioBrick™ vectors for metabolic pathway engineering. Applied microbiology and biotechnology, 1275-1286.
• Weeks, G. & Wakil, S. J. (1968) GENERAL PROPERTIES OF THE REDUCTASES FROM ESCHERICHIA on the Mechanism of Fatty Acid.
• Willis R. M., et al., (2011) Characterization of a Fatty Acyl-CoA Reductase from Marinobacter aquaeolei VT8: A Bacterial Enzyme Catalyzing the Reduction of Fatty Acyl-CoA to Fatty Alcohol. Biochemistry 50:10550-8.
• White S. W., et al., (2005) The structural biology of type II fatty acid biosynthesis, p. 791-831, Annual Review of Biochemistry, vol. 74. Annual Reviews, Palo Alto.
• Wiesenborn, D. P., et al., (1988) Thiolase from Clostridium acetobutylicum ATCC 824 and Its Role in the Synthesis of Acids and Solvents. App. Environ. Microbiol., 54(11), 2717-22.
• Yan R. T. & Chen J. S. (1990 (Coenzyme A-acylating aldehyde dehydrogenase from Clostridium beijerinckii NRRL B592. Appl. Environ. Microbiol. 56, 2591-2599.
• Zhu L., et al. (2013) The Two Functional Enoyl-Acyl Carrier Protein Reductases of Enterococcus faecalis Do Not Mediate Triclosan Resistance. MBio 4:10.