The US-based agricultural products company, Cargill, initiated research on biological 3-HP production with the purpose of commercial production (Cargill, 2012). As a result of their studies, they proposed and patented seven metabolic pathways for 3-HP production from glucose (Gokarn et al., 2001; Liao et al., 2005, 2007; Marx et al., 2007). These pathways are mediated by either pyruvate or phosphoenolpyruvate (PEP), both of which are common intermediates of sugar metabolism (Figure 14.1).
The maximum theoretical yield (mol/mol) of 3-HP from glucose, based on pathway stoichiometry and/or comparison of energy content between substrate (glucose) and product (3-HP), is 2.0 (Dugar and Stephanopoulos, 2011). To be practically feasible, the pathway leading to 3-HP production should be redox-balanced and generate net ATP for use in cell growth, maintenance, product excretion, and so on. All biosynthetic pathways shown in Figure 14.1 are redox-balanced; however, many do not generate ATP and some are even ATP-consuming.
The ATP yield and thermodynamic feasibilities (predicted on the basis of group contribution method) of these pathways are shown in Table 14.2. The ATP yield for a pathway can vary depending upon the enzyme(s) involved in the pathway. For exam- ple, the carboxylation of PEP into oxaloacetate produces one ATP if mediated by PEP carboxykinase; however, the same reaction does not generate ATP when catalyzed by PEP carboxylase. Henry et al. (2010) compared and evaluated Cargill’s seven path- ways along with several new pathways constructed by BNICE (Biochemical Network Integrated Computational Explorer). Based on several criteria including pathway length, thermodynamic feasibility, maximum achievable yield of 3-HP from glucose, and maximum achievable activity (intracellular concentration), they identified four pathways which were promising, one from Cargill and three from BNICE (Henry et al., 2010; Figure 14.2). In addition, they claimed that, among these four pathways, the one developed by Cargill which occurred viaα- andβ-alanines (pyruvate→α- alanine→β-alanine→3-oxopropanoate→3-HP) was most promising. In this pathway, the energy consuming step was bypassed by the use of alanine-2,3-aminomutase, an enzyme that catalyzes the transfer of an amino group from theα- toβ-carbon (Jiang et al., 2009). This allowed the highest possible yield (2.0 mol 3-HP from 1.0 mol glucose) (Willke and Vorlop, 2004; Henry et al., 2010). This pathway was tested by Liao et al. (2007b) in a recombinantEscherichia colioverexpressing a vitamin B12-dependent lysine 2,3-aminomutase, an isozyme of alanine-2,3-aminomutase.
However, detailed information pertaining to titer, yield, and productivity of 3-HP have not been revealed.
Three other pathways suggested to be promising by BNICE (Figure 14.2) also deserve some attention. The pathway via lactate (pyruvate→lactate→3-HP) is attrac- tive since it involves the fewest number of reactions. However, the conversion of lactate into 3-HP is thermodynamically unfavorable and can only proceed in the for- ward direction when the lactate concentration is very high and the 3-HP is very low (Herrmann et al., 2005; Jiang et al., 2009; Henry et al., 2010). This condition can be
FIGURE14.1Biochemicalpathwaysfor3-HPproduction.Glu,glutamate;α-KG,α-ketoglutarate.(Jiangetal.,2009;Henryetal., 386
TABLE 14.2 The ATP Yield and Thermodynamic Feasibility for Various Metabolic Pathways Starting with Glucose for the Production of 3-HP
Net ATP yield Thermodynamic
Metabolic pathway (mol/mol 3-HP) feasibility
Pyruvate→lactate→lactoyl-CoA→acryloyl- CoA→3-HP-CoA→3-HP
1/0 Unfavorable
Pyruvate→acetyl-CoA→malonyl-CoA→3-
oxopropanoate→3-HP 0 Favorable
Pyruvate/PEP→OAA→aspartate→β-
alanine→3-oxopropanoate→3-HP 1/0 Favorable
Pyruvate/PEP→OAA→aspartate→β- alanine→β-alanyl-CoA→acryloyl-CoA→3- HP-CoA→3-HP
0/-1 Unfavorable
Pyruvate→α-alanine→β-alanine→3- oxopropanoate→3-HP
1 Favorable
Pyruvate→α-alanine→β-alanine→β-alanyl-
CoA→acryloyl-CoA→3-HP-CoA→3-HP 1/0 Unfavorable
Source: Adapted from Jiang et al. (2009) and Henry et al. (2010).
met when lactate is accumulated at high concentrations and 3-HP is quickly excreted as soon as it is produced. The other two pathways suggested by BNICE occur viaα- andβ-alanines (pyruvate→α-alanine→β-alanine→propenoate→3-HP). These path- ways are similar to the one proposed by Cargill in that all intermediary compounds are the same except for 3-oxopropanoate, which is replaced by propenoate. The ami- nating and deaminating reactions in these two pathways are also somewhat different.
FIGURE 14.2 Novel pathways for 3-HP biosynthesis designed by Biochemical Network Integrated Computational Explorer (BNICE) (Henry et al., 2010).
Specifically, glutamate andα-ketoglutarate along with transaminases are involved in Cargill’s pathway, while ammonia along with aminase and deaminase are involved in the BNICE pathway (Henry et al., 2010). However, the pathways suggested by BNICE have not been tested experimentally and their practical applications have not been demonstrated. Beside these three pathways, one more pathway constructed by BNICE is worthy of discussion. The pathway consists of three steps only and 3-HP production in this pathway takes place via oxaloacetate and 3-oxopropanoate (pyruvate→oxaloacetate→3-oxopropanoate→3-HP). This pathway is thermodynam- ically feasible, but requires the consumption of one ATP; thus, the yield is as low as 1.22 mol 3-HP/mol glucose (Henry et al., 2010).
Interestingly, the only pathway that has been seriously studied experimentally for the production of 3-HP from glucose is the one that occurs via malonyl-CoA. This pathway was not thought to be promising by Henry et al. (Figure 14.1). However, Rathnasingh et al. (2012) developed a recombinantE. colistrain overexpressingmcr, pntAB, and accADBCencoding malonyl-CoA reductase, nicotinamide nucleotide transhydrogenase, and acetyl-CoA carboxylase and biotinilase, respectively. Themcr gene was derived fromC. aurantiacus, while thepntABandaccADBCgenes were fromE. coliK-12. In this pathway, acetyl-CoA is carboxylated into malonyl-CoA (by acetyl-CoA carboxylase and biotinilase), which is then reduced into 3-HP by malonyl-CoA reductase using two molecules of NADPH as electron donors. The conversion of NADH into NADPH was also facilitated by nicotinamide nucleotide transhydrogenase (pntAB). The recombinant strain yielded 2.16 mM 3-HP from glucose in 24 hours. It should be noted that malonyl-CoA is a well-known intermediate for fatty acid synthesis. To improve 3-HP production via the malonyl-CoA reductase pathway, Lynch et al. (2011) attempted to increase carbon flux from malonyl-CoA toward 3-HP synthesis by reducing the expression level of enoyl-ACP reductase (fabI) (key enzyme of fatty biosynthesis). They could produce 5.8 g/L 3-HP in shake flask cultures and 20.7 g/L 3-HP in bioreactor cultures at 38.5 hours. Further modification of the recombinantE. coli by deleting competing pathways leading to by-products formation (lactate, ethanol, acetate, and methylglyoxal) resulted in a drastic improvement in 3-HP production to 49 g/L in 69 hours, which is the maximum titer reported to date using glucose as the substrate (Lynch et al., 2011).