12.3 METABOLIC ENGINEERING STRATEGIES FOR DIRECTED PRODUCTION OF ISOBUTANOL
12.3.1 Isobutanol Production with Escherichia coli
The initial step toward the biotechnological production of isobutanol was done by Atsumi et al. (2008) withE. coli. Implementation ofkivdfromL. lactis andadh2 from S. cerevisiae, and additional overexpression of the E. coli l-valine biosyn- thetic pathway genesilvIHCD, encoding AHAS, acetohydroxyacid isomeroreductase (AHAIR), and dihydroxyacid dehydratase (DHAD) resulted in an improved yield due to an increased drain-off of pyruvate toward 2-ketoisovalerate (Figure 12.3). To avoid by-product formation and to further increase pyruvate availability, the genesadhE, ldhA,frdAB,pta, andfnrencoding the bi-functional AdhE, d-lactate dehydrogenase (LDH), fumarate reductase (FRD), phosphotransacetylase (PTA), and the fumarate and nitrate reductase regulator FNR, respectively, were stepwise deleted (Figure 12.3).
The resulting strain produced isobutanol with a yield of about 0.21 g of isobutanol per gram of glucose and was further improved by overexpression ofalsSencoding AHAS fromB. subtilisinstead ofilvIH. The latter improvement in combination with the deletion of thepflB gene, encoding pyruvate formate lyase (PFL), resulted in E. coliJCL260/pSA55/pSA69 and led to the production of about 300 mM (22 g/L) with a yield of about 0.86 mol/mol (0.35 g/g) in minimal medium with yeast extract and glucose under microaerobic conditions (Atsumi et al., 2008; Table 12.3).
Applying anaerobic conditions, the maximal theoretical yield can be calcu- lated to be 0.41 g isobutanol per gram of glucose (1 mole/mole; Li et al., 2010).
TABLE12.2OverviewoftheBiochemicalPropertiesofRelevantEnzymeswith2-KetoacidDecarboxylaseorAlcoholDehydr Activity Enzyme(organism)SubstrateCofactorKM(mM)kcat/skcat/KMReference KDC(L.lactis)2-KetoisovalerateTPP2.2±0.938.3±9.817Zhangetal. AlsS(B.subtilis)2-KetoisovalerateTPP300±358.9±1.20.03Atsumietal. YqhD(E.coli)IsobutyraldehydeNADPH1.8±1.31±0.10.7Atsumietal. Adh2(S.cerevisiae)IsobutyraldehydeNADH385±310.9±0.12.2×10−3Atsumietal. AdhA(L.lactis)IsobutyraldehydeNADH9.1±2.96.6±0.20.8Atsumietal. AdhA(C.glutamicum)IsobutyraldehydeNADH–––Smithetal. Adh,alcoholdehydrogenase;AlsS,acetohydroxyacidsynthase;KDC,2-ketoaciddecarboxylase;YqhD,broad-rangealcoholdehydrogenase;TPP,thi pyrophosphate.
332
TABLE12.3RelevantYieldsandTitersofIsobutanol,2-Methyl-1-Butanol,3-Methyl-1-Butanol,1-ButanolDerivedofthe PathwaysofBranched-ChainAminoAcids StrainProcessProductYP/S(mole/mole), (g/g)Titer(g/L)Reference E.coliJCL260/pSA55/pSA69FlaskIsobutanol0.86,0.3522Atsumietal. E.coli1993(pGVferm6)FlaskIsobutanol1.03,0.4213.4Bastianetal. E.coliCRS22Flask2-Methyl-1-butanol0.35,0.171.3CannandLiao E.coliAL2Flask3-Methyl-1-butanol0.22,0.119.5Connoretal. E.coliCRS-BuOH23Flask1-Butanol∼0.09,0.040.8ShenandLiao
333
PEP
Pyruvate
Acetyl-CoA PK
PDHC PEPCx
PEPCk
Oxaloacetate
Malate Fumarate
Citrate
MQO Mdh
Isocitrate
Succinate FUM
Acetyl-P AK
Acetate PQO
ODHC Ketoglutarate
Succinyl-CoA Glyceraldehyde-3P
Glucose
Glucose-6P Fructose-6P
1,3-Bisphosphoglycerate GAPDH
6P-gluconate PGDH
Ribulose-5P
ICD Glyoxylate
MS
ICL
Lactate
NAD+ NADH+H+
+
NAD+ NADH+H+
NADH+H+
NAD+
NAD+ NADH+H+
NADPH+H+ NAPD+
GPDH
NAPDH+H+ NADP+
SfcA
NADH+H+
NAD+
MalE
NAPDH+H+
NADP+
Ethanol PPS
SDH
NADH+H+
NAD+
NAPDH+H+ NADP+
PntAB
AHAS
AcetolactateAHAIR DHAD 2-Ketoiso- TA L-Valine Isobutyraldehyde
Isobutanol Adh
valerate KDC
FIGURE 12.3 The enzymes of the central metabolism with the biosynthetic pathway of l- valine inE. coliand the syntethic pathway from 2-ketoisovalerate to isobutanol. Adh, alcohol dehydrogenase; AdhE, alcohol dehydrogenase E; AHAIR, acetohydroxyacid isomeroreduc- tase; AHAS, acetohydroxyacid synthase; AK, acetate kinase; DHAD, dihydroxyacid dehy- dratase; FNR, fumarate and nitrate reductase regulator; FUM, fumarase; FRD, fumarate reductase; GAPDH, glyceraldehyde-3P dehydrogenase; GPDH, G6P dehydrogenase; ICD, isocitrate dehydrogenase; ICL, isocitrate lyase; KDC, 2-ketoacid decarboxylase fromL. lactis;
LDH, d-lactate dehydrogenase; MalE, NADPH-dependent malic enzyme; Mdh, malate dehy- drogenase; MQO, malate:quinone oxidoreductase; MS, malate synthase; ODHC, oxoglutarate dehydrogenase complex; PDHC, pyruvate dehydrogenase complex; PEP, phosphoenolpyru- vate; PEPCk, PEP carboxykinase; PEPCx, PEP carboxylase; PFL, pyruvate formate lyase;
PGDH, 6P-gluconate dehydrogenase; PK, pyruvate kinase; PntAB, membrane-bound tran- shydrogenase; PPS, PEP synthetase; PTA, phosphotransacetylase; PQO, pyruvate:quinone oxidoreductase; SDH, succinate dehydrogenase; SfcA, NADH-dependent malic enzyme; TA, transaminase B.
However, maintaining a balanced redox state is crucial for an efficient production process under oxygen limitation. AHAIR enzymes are predominantly NADPH depen- dent, whereas different types of Adhs accept either NADH+H+ or NADPH+H+. Thus, the formation of one mole isobutanol from pyruvate requires one mole NADH+H+ and one mole NADPH+H+ or 2 moles NADPH+H+, respectively.
Regarding that most bacteria generate 2 moles NADH+H+ per mole of glucose in the course of glycolysis, the conversion of NADH+H+to NADPH+H+or the con- struction of a completely NADH-dependent isobutanol-forming pathway is essential for efficient production (Bastian et al., 2011). Both strategies were compared in an approach to optimize cofactor utilization for isobutanol production withE. coli (Bastian et al., 2011). Protein engineering of the NADPH-dependent AHAIR identi- fied a mutated variant (IlvCA71S,R76D,S78D,Q110V) that showed a strong preference for NADH+H+over NADPH+H+. Furthermore, the catalytic efficiency and the affinity toward isobutyraldehyde of the NADH-dependent AdhA fromL. lactiswas signif- icantly improved by random mutagenesis and recombination of useful mutations.
Overexpression of the engineered genes for AHAIR and AdhA as well as for AHAS (fromB. subtilis), DHAD, and KDC (fromL. lactis) inE. coliwith inactivated LdhA, AdhE, FRD, PFL, AHAIR, and PTA resulted inE. coli1993 (pGVferm6), producing isobutanol with the maximal theoretical yield of 1.03 mole isobutanol per mole of glucose (0.42 g/g) under anaerobic conditions (Table 12.3). Also overexpression of the membrane-bound transhydrogenase PntAB in the same strain background with a completely NADPH+H+-dependent pathway, overexpressing nativeilvC(encoding AHAIR), andyqhd(encoding Adh), resulted in the maximal theoretical yield, how- ever, with reduced productivity compared toE. coli1993 (pGVferm6) (Bastian et al., 2011).
All these results demonstrate thatE. coliis a very promising host for biotechno- logical production of isobutanol. However, industrial relevant processes also have to result in high final titers, which do not match to the known cytotoxicity of higher alcohols on microorganisms (Knoshaug and Zhang, 2009; Ezeji et al., 2010). As outlined further in section 12.4, the maximal tolerance ofE. coliagainst isobutanol is lower compared to that of some Gram-positive bacteria, such asC. glutamicumor B. subtilis(Brynildsen and Liao, 2009; Smith et al., 2010; Li et al., 2011) or to that ofS. cerevisiae. Consequently, several metabolic engineering approaches focused on these alternative hosts.