The ABE producing strains can hydrolyze starch to glucose or other hexose by amylases. Glucose was firstly converted to pyruvate through the Embden-Meyerhoff pathway (EMP, or glycolysis). Pyruvate was then cleaved to acetyl-CoA by pyruvate ferredoxin oxidoreduc — tase. Acetyl-CoA is the common precursor of all the fermentation intermediate and end products. The enzyme activity and the coding genes have been widely assayed and descri­bed in butanol-producing strains (Durre et al., 1995; Gheshlaghi et al., 2009).

The ABE fermentation process can be divided into two successive and distinct phase as acidogenesis phase and solvetogenesis phase. The acidogenesis phase is accompanied with cell exponential growth and pH drop, accumulation of acetate and butyrate. Solventogenesis phase begins with endospore forming and the cells entering stationary state. The products of acidogenesis phase include acetate and butyrate. Acetate forms from Acetyl-CoA, which is catalyzed by two enzymes, phosphotransacetylase (PTA, or phosphate acetyltransferase, endoced by pta gene) and acetate kinase (AK, encoded by ak gene). The butyrate synthesis is a little complicated with more steps. At first, two molecular of acetyl-CoA is catalyzed by thiolase (thl, or acetyl-CoA acetyltransferase, encoded by thl gene) and transforms into one molecular C4 unit acetoacetyl-CoA, which is another important node and precursor of buty­rate, acetone, and butanol synthesis. The acetoacetyl-CoA is subjected to three enzymes in turn and another C4 unit butyryl-CoA is the intermediate product. The three enzymes are hydroxybutyryl-CoA dehydrogenase (encoded by hbd gene) (Youngleson et al., 1995), croto — nase (CRT, or hydroxybutyryl-CoA dehydrolase, encoded by crt gene), and butyryl-CoA de­hydrogenase (BCD, encoded by bcd gene). Accordingly, three encoded genes coexist in the BCS operon with additional two genes coding for the a and p subunit of electron transfer protein (Bennett and Rudolph, 1995). Butyryl-CoA was then catalyzed by phosphotransbu- tylase (PTB, or phosphate butyltransferase, encoded by ptb gene) and butyrate kinase (BK, encoded by bk gene) to form butyrate during acidogenesis phase.

As the organic acid accumulation, pH drop to the lowest point during the fermentation. This leads to the switch of acidogenesis phase to solventogenesis phase. Acetate and butyrate are reassimilated and participate in the solvent formation. Under the catalyzing of CoA transfer­ase (CoAT, two unit encoded by ctfa and ctffi), acetate and butyrate was transformed into acetyl-CoA and butyryl-CoA respectively again. The alcohols formation share the same key enzymes, NAD(P)H dependent aldehyde/alcohol dehydrogenases (encoded by adhl and adh2 gene) (Chen, 1995). In addition, Butanol owns its unique butanol dehydrogenase (en­coded by bdh gene) (Welch et al., 1989). The formation of acetone from acetoacetyl-CoA is a two-step reaction. Acetoacetyl-CoA is catalyzed to acetoacetate by CoA transferase. Acetone is produced after a molecular CO2 released from acetoacetate by decarboxylase (AADC, en­coded by aadc gene) (Janati-Idrissi et al., 1988; Cary et al., 1993). Both acid reassimilation and acetone formation utilize CoA transferase, however, the butyrate uptake was not concomi­tant with the production of acetone (Desai et al., 1999). The metabolic pathway accompanied by electron transfer and reduction force forming. The main ABE fermentation pathway was illustrated in Fig.3.

Solventogenic genes aad, ctfA, ctfB and adc constitute the sol operon (Durre et al., 1995). In some conditions, butanol producing strains lose the ability to produce solvents after repeat­ed subculturing, called as degenerated (DGN) strain. In C. acetobutylicum ATCC 824, the plasmid pSOL1 carrying the sol operon was found missing during degenerating process (Cornillot et al., 1997). For C. saccharoperbutyl acetonicum strain N1-4, the sol genes main­tained in degenerated DGN3-4 strain, while the sol operon was hardly induced during sol- ventogenesis. Extract from the culture supernatants of wild-type N1-4 is enough to induce the transcription of the sol operon in DGN3-4 (Kosaka et al., 2007). It suggested that the de­generation maybe caused by the incompetence of the induction mechanism of the sol oper — on. The transcription of sol operon may be under the control of the quorum-sensing mechanism in C. saccharoperbutyl acetonicum.

Though the metabolic pathway is clear, the underlying regulation mechanism is poorly un­derstood, such as the phase switch of fermentation, the relationship between solventogene — sis and sporulation. Answering these questions is critical to improve the efficiency of butanol producing fundamentally. Proteomics and transcriptomics can provide more un­known details, which will be helpful for solving these problems (Sivagnanam et al., 2011; Sivagnanam et al., 2012).