Synthesis of Heterologous Mevalonic Acid Pathway Enzymes in Clostridium ljungdahlii for the Conversion of Fructose and of Syngas to Mevalonate and Isoprene

ABSTRACT There is a growing interest in the use of microbial fermentation for the generation of high-demand, high-purity chemicals using cheap feedstocks in an environmentally friendly manner. One example explored here is the production of isoprene (C5H8), a hemiterpene, which is primarily polymerized to polyisoprene in synthetic rubber in tires but which can also be converted to C10 and C15 biofuels. The strictly anaerobic, acetogenic bacterium Clostridium ljungdahlii, used in all of the work described here, is capable of glycolysis using the Embden-Meyerhof-Parnas pathway and of carbon fixation using the Wood-Ljungdahl pathway. Clostridium-Escherichia coli shuttle plasmids, each bearing either 2 or 3 different heterologous genes of the eukaryotic mevalonic acid (MVA) pathway or eukaryotic isopentenyl pyrophosphate isomerase (Idi) and isoprene synthase (IspS), were constructed and electroporated into C. ljungdahlii. These plasmids, one or two of which were introduced into the host cells, enabled the synthesis of mevalonate and of isoprene from fructose and from syngas (H2, CO2, and CO) and the conversion of mevalonate to isoprene. All of the heterologous enzymes of the MVA pathway, as well as Idi and IspS, were shown to be synthesized at high levels in C. ljungdahlii, as demonstrated by Western blotting, and were enzymatically active, as demonstrated by in vivo product synthesis. The quantities of mevalonate and isoprene produced here are far below what would be required of a commercial production strain. However, proposals are made that could enable a substantial increase in the mass yield of product formation. IMPORTANCE This study demonstrates the ability to synthesize a heterologous metabolic pathway in C. ljungdahlii, an organism capable of metabolizing either simple sugars or syngas or both together (mixotrophy). Syngas, an inexpensive source of carbon and reducing equivalents, is produced as a major component of some industrial waste gas, and it can be generated by gasification of cellulosic biowaste and of municipal solid waste. Its conversion to useful products therefore offers potential cost and environmental benefits. The ability of C. ljungdahlii to grow mixotrophically also enables the recapture, should there be sufficient reducing equivalents available, of the CO2 released upon glycolysis, potentially increasing the mass yield of product formation. Isoprene is the simplest of the terpenoids, and so the demonstration of its production is a first step toward the synthesis of higher-value products of the terpenoid pathway.


Strain Construction and Western Blotting
Construction of plasmid pJF100 (Table 1) (Table 1) A vector (pJF200) for the expression of IspS (isoprene synthase) and Idi (isopentenyl diphosphate isomerase) using the Acetobacterium woodii promoter, Awo1181gi, with a pCB102

Construction of pJF200
Gram positive replicon, was derived from the plasmids pMTL83151 (Supplementary Figure S3) and pMCS337 (also called pDW253; SEQ ID NO:23(1); plasmid map Supplementary Figure   S5). Further details on the construction of plasmid pJF200 are provided in Beck et al. (1).
C. ljungdahlii was transformed by electroporation with pJF200 as previously described for pJF100. The transformants obtained, carrying pJF200, showed no evidence, however, of IspS or Idi expression by Western blotting (not shown). Plasmid isolated from these strains, used to 9 back transform E. coli TV3007, did, however, express IspS and Idi (by Western blot, Supplementary Figure S6) indicating that in the proper host the pJF200 plasmid was functional for expression. We concluded that while active in E. coli TV3007, the A. woodii promoter Awo1181gi was not active in C. ljungdahlii under the growth conditions used. Consequently, this promoter was replaced with the C. sporogenes Pfdx promoter that was used for MvaE and MvaS expression from plasmid pJF100. (Table 1) Genes ispS and idi were placed under the control of the Pfdx promoter by replacing the mvaE and mvaS in vector pJF100 with these coding regions. A 3.5-kb DNA vector fragment which contains the Pfdx promoter was PCR amplified using pJF100 as template and primers (see  Transformants of E. coli TV3007 were prepared as described earlier (1) with plasmid isolated from two C. ljungdahlii pJF200 Fdii transformants named nos.1LD and 2LD. IspS and Idi were synthesized using the Awo1181 promoter in E. coli TV3007 and were used as markers for IspS and Idi. These transformants were grown up as described earlier (see also Beck et al.

Construction of plasmid pJF100 Fdii
(1)) and harvested and solubilized with LDS sample buffer as before.
Gel electrophoresis and Western blotting were performed as described in the Materials and Methods section (see Supplementary Figure S6). His codons at the 5' and 3' ends of the ispS gene in the pJF100 Fdii plasmid is given below.

Adaptation of C. ljungdahlii to growth on syngas
Plasmid pJF100 Fdii producing hexaHistidine tagged Isps (Table 1) Plasmid pJF100 Fdii was used as template for the independent addition of N-and C- 14 The electrophoresis gel system was run as described earlier (also see Materials and Methods) and the gels were transferred, incubated with antibody and then with fluorescenttagged goat anti-rabbit antibody as before (see Supplementary Figure S7).
Placing all of these genes in one plasmid produced no C. ljungdahlii transformants following electroporartion, possibly due to the size of the plasmid (9844 bp). Consequently, we divided the genes between two plasmids, one containing mvk, pmk and mvd (pJF102) and the other containing idi and ispS (pJF101). Electrocompetent C. ljungdahlii cells were electroporated with plasmid pJF102 and cultured as described previously (also see Materials and Methods section) except that antibiotic selection was with 1 mg spectinomycin/mL rather than 5 µg thiamphenicol/mL. Confirmation of successful transformation involved plasmid isolation and back transformation into E. coli Top10 followed by plasmid reisolation and sequencing as described earlier.