Functional analysis of the methylerythritol phosphate pathway terminal enzymes IspG and IspH from Zymomonas mobilis

Abstract

Isoprenoids are a diverse family of compounds that are synthesized from two isomeric compounds, isopentenyl diphosphate and dimethylallyl diphosphate. In most bacteria, isoprenoids are produced from the essential methylerythritol phosphate (MEP) pathway. The terminal enzymes of the MEP pathway IspG and IspH are [4Fe-4S] cluster proteins, and in Zymomonas mobilis, the substrates of IspG and IspH accumulate in cells in response to O₂, suggesting possible lability of their [4Fe-4S] clusters. Here, we show using complementation assays in Escherichia coli that even under anaerobic conditions, Z. mobilis IspG and IspH are not as functional as their E. coli counterparts, requiring higher levels of expression to rescue viability. A deficit of the sulfur utilization factor (SUF) Fe-S cluster biogenesis pathway did not explain the reduced function of Z. mobilis IspG and IspH since no improvement in viability was observed in E. coli expressing the Z. mobilis SUF pathway or having increased expression of the E. coli SUF pathway. Complementation of single and double mutants with various combinations of Z. mobilis and E. coli IspG and IspH indicated that optimal growth required the pairing of IspG and IspH from the same species. Furthermore, Z. mobilis IspH conferred an O₂-sensitive growth defect to E. coli that could be partially rescued by co-expression of Z. mobilis IspG. In vitro analysis showed O₂ sensitivity of the [4Fe-4S] cluster of both Z. mobilis IspG and IspH. Altogether, our data indicate an important role of the cognate protein IspG in Z. mobilis IspH function under both aerobic and anaerobic conditions.

Importance: Isoprenoids are one of the largest classes of natural products, exhibiting diversity in structure and function. They also include compounds that are essential for cellular life across the biological world. In bacteria, isoprenoids are derived from two precursors, isopentenyl diphosphate and dimethylallyl diphosphate, synthesized primarily by the methylerythritol phosphate pathway. The aerotolerant Z. mobilis has the potential for methylerythritol phosphate pathway engineering by diverting some of the glucose that is typically efficiently converted into ethanol to produce isoprenoid precursors to make bioproducts and biofuels. Our data revealed the surprising finding that Z. mobilis IspG and IspH need to be co-optimized to improve flux via the methyl erythritol phosphate pathway in part to evade the oxygen sensitivity of IspH.

Keywords: DMADP; Fe-S cluster stability; HMBDP; IDP; IspG; IspH; MEP pathway optimization; MEcDP; alpha proteobacteria; isoprenoid synthesis; oxygen-sensitive Fe-S clusters.

Fig 1

E. coliΔispGΔispH MVA⁺ (PK14487) with plasmid variants containing ispG and ispH from either Z. mobilis (ispH_ZM ispG_ZM) or E. coli (ispH_EC ispG_EC) were grown in LB with mevalonate, arabinose, and spectinomycin. After 16 hours of overnight growth, bacteria were washed with LB to remove mevalonate and were normalized to OD₆₀₀ of 1. The bacteria were then serially diluted in LB, and 5 μL was used from each dilution tube to spot on solid TYE media, with or without mevalonate (1.0 mM), containing the indicated concentrations of IPTG (25 μM, 50 μM, 100 μM, or 200 μM) to induce ispG and ispH co-expression. The agar plates were then incubated at 37°C overnight under either (A) anaerobic or (B) aerobic conditions.

Fig 2

E. coliΔispG MVA⁺ (PK14465) with plasmid variants containing ispG from Z. mobilis (ispG_ZM) or E. coli (ispG_EC) were grown in LB with mevalonate, arabinose, and spectinomycin. Viability of cells was assayed as described in Fig. 1.

Fig 3

E. coliΔispH MVA⁺ (PK14464) with plasmid variants containing ispH from Z. mobilis (ispH_ZM) or E. coli (ispH_EC) or ispG and ispH from Z. mobilis (ispH_ZM ispG_ZM) were grown in LB with mevalonate, arabinose, and spectinomycin. Viability of cells was assayed as described in Fig. 1.

Fig 4

E. coliΔispGΔispH MVA⁺ (PK14487) with plasmid variants containing ispG from E. coli and ispH from Z. mobilis (ispH_ZM ispG_EC) or ispG from Z. mobilis and ispH from E. coli (ispH_EC ispG_ZM) were grown in LB with mevalonate, arabinose, and spectinomycin. Viability of cells was assayed as described in Fig. 1.

Fig 5

Loss of the [4Fe-4S] cluster from IspG and IspH following exposure to air. UV-visible absorbance changes of anaerobically isolated Z. mobilis Strep-tag-II-IspG (60 μM) (A) and Z. mobilis Strep-tag-II-IspH (60 μM) (B) following exposure to air. (C) Plot of the change in absorbance at 410 nm after exposure of E. coli Strep-tag II IspH to air. (D) Plot of the change in absorbance at 410 nm from the data in panel A. (E) Plot of the change in absorbance at 410 nm from the data in panel B. Relative absorbance at 410 nm is the normalized absorbance value at the initial time point.

Fig 6

E. coliΔispH MVA⁺ (PK14464) with plasmid variants containing ispH from Z. mobilis (ispH_ZM) with point mutations E127A or C13A were grown in LB with mevalonate, arabinose, and spectinomycin. Viability of cells was assayed as described in Fig. 1 at 100 μM IPTG.

Fig 7

E. coliΔispH MVA⁺ or E. coliΔispH P_fnr suf_ZM MVA⁺ with plasmid variants containing ispH from Z. mobilis (ispH_ZM) or E. coli (ispH_EC) or ispG and ispH from Z. mobilis (ispH_ZM ispG_ZM) were grown in LB with mevalonate, arabinose, and spectinomycin. Viability of cells was assayed as described in Fig. 1.

Fig 8

E. coliΔispH MVA⁺ or E. coliΔispH P_fnr suf_EC MVA⁺ with plasmid variants containing ispH from Z. mobilis (ispH_ZM) or E. coli (ispH_EC) or ispG & ispH from Z. mobilis (ispH_ZM ispG_ZM) were grown in LB with mevalonate, arabinose, and spectinomycin. Viability of cells was assayed as described in Fig. 1

Fig 9

Proposed model describing Z. mobilis and E. coli IspH and IspG function under anaerobic (top panel) and aerobic (bottom panel) conditions. The arrows represent the proposed flux through the terminal steps of the MEP pathway; thick lines represent high flux, whereas thin lines represent low flux and assumes nonoptimal interactions with redox proteins (flavodoxin or ferredoxin) indicated by X. The shade of green indicates the relative amount of DMADP and IDP produced, with dark green being the most and light green being the least.