Isobutanol production by combined in vivo and in vitro metabolic engineering

Abstract

The production of the biofuel, isobutanol, in E. coli faces limitations due to alcohol toxicity, product inhibition, product recovery, and long-term industrial feasibility. Here we demonstrate an approach of combining both in vivo with in vitro metabolic engineering to produce isobutanol. The in vivo production of α-ketoisovalerate (KIV) was conducted through CRISPR mediated integration of the KIV pathway in bicistronic design (BCD) in E. coli and inhibition of competitive valine pathway using CRISPRi technology. The subsequent in vitro conversion to isobutanol was carried out with engineered enzymes for 2-ketoacid decarboxylase (KIVD) and alcohol dehydrogenase (ADH). For the in vivo production of KIV and subsequent in vitro production of isobutanol, this two-step serial approach resulted in yields of 56% and 93%, productivities of 0.62 and 0.074 g L^-1 h^-1, and titers of 5.6 and 1.78 g L^-1, respectively. Thus, this combined biosynthetic system can be used as a modular approach for producing important metabolites, like isobutanol, without the limitations associated with in vivo production using a consolidated bioprocess.

Keywords: Bicistronic; CRISPR; CRISPRi; Fed-batch; Isobutanol; α-Ketoisovalerate.

Fig. 1

Schematic illustration of isobutanol pathway (A) Pathway design for in vivo production of KIV in E. coli and in vitro production of Isobutanol in a tube. ALsS, acetolactate synthase from Bacillus subtilis; IlvC, 2-hydroxy-3-ketol-acid reductoisomerase from E. coli; IlvD, dihydroxy-acid hydratase from E. coli; KIVD, 2-keto acid decarboxylase; ADH, Alcohol dehydrogenase. (B) KIV pathway in bicistronic design (BCD) with its two Shine-Dalgarno motifs (SD1 and SD2) is shown. +1 is mRNA and 5′UTR is leader sequence. SD2 sites adhere to the BCD (Mutalik et al., 2013).

Fig. 2

CRISPR Cas9 mediated integration of KIV pathway into E. coli genome (A) General outline of integration of KIV pathway and curing of plasmids (B) Screening of transformed colonies by PCR to verify upstream junction (5′) and (C) downstream junction 3’. L represents 10 kb ladder; expected sizes for correct integration are represented by arrows. PCR products are separated on 0.8% agarose gel. ‘C’ represents control (without integration) Pathway is showing 80% integration efficiency (D) PCR amplification using primers (MB67/MB68) flanking outside of the integrated region to confirm whole pathway integration 5′-3’ (lane 3).

Fig. 3

dCas9 mediated deregulation of ilvE gene (A) Schematic diagram of ilvE promoter from EcoCyc E. coli database showing spacer sequences within promoter IlvE (p) and gene region IlvE (g) designed for dCas9 targeting. (B) Colony PCR for screening of transformed colonies for successful integration of spacer regions into dCas9. L, represents 10 kb ladder; expected sizes for correct integration are represented by arrows. PCR products are separated on 1.5% agarose gel. ‘C’ represents control. (C) Effect of dCas9-mediated downregulation at the ORF using dcas9IlvE (g) primer (MG02) and promoter region using dcas9IlvE (p) primer (MG03) on the growth of E. coli having KIV pathway integrated. Data is replicate of three values. Error bars represent standard deviation.

Fig. 4

2-Ketoisovalerate and other metabolites production (A) Graph showing KIV production in shake flasks at 30 °C and 37 °C before and after downregulation of ilvE gene. (B) Glucose consumption pattern of different strains at different temperatures in shake flasks. (C) Comparative metabolite profile of different metabolites in the culture supernatant of control and modified E. coli strains at two different temperatures (37 °C and 30 °C) as detected by HPLC. (D) KIV and acetate production pattern in glucose fed batch reactor at different growth stages of MG03. Values are means of three replications. Bar values (mean ± S.D) sharing same alphabet(s) do not differ significantly by LSD at P ≤ 0.05.

Fig. 5

NADH production in E. coli. (A) under slow growth rate and glucose limited condition sufficient electron transport chain exists to keep the NADH concentration low. (B) Under high growth rate electron transport chain surface limitation unable to metabolize all the NADH created, leading to NADH imbalance and excretion to outside the cell (Szenk et al., 2017) (C) NADH concentration in outer culture media in control and modified strains. (D) Standard curve showing NADH concentration with respect to fluorescence units. Values are means of three replicates. Error bars represents the standard deviation.

Fig. 6

Copy Number Effect on KIV Production: (A) OD of cultures grown at 37 °C with and without the inducible KIV pathway plasmid. (B) KIV concentration for cultures grown at 37 °C with and without the inducible KIV pathway plasmid.