Thermoadaptation-directed enzyme evolution in an error-prone thermophile derived from Geobacillus kaustophilus HTA426

Abstract

Thermostability is an important property of enzymes utilized for practical applications because it allows long-term storage and use as catalysts. In this study, we constructed an error-prone strain of the thermophile Geobacillus kaustophilus HTA426 and investigated thermoadaptation-directed enzyme evolution using the strain. A mutation frequency assay using the antibiotics rifampin and streptomycin revealed that G. kaustophilus had substantially higher mutability than Escherichia coli and Bacillus subtilis. The predominant mutations in G. kaustophilus were A · T→G · C and C · G→T · A transitions, implying that the high mutability of G. kaustophilus was attributable in part to high-temperature-associated DNA damage during growth. Among the genes that may be involved in DNA repair in G. kaustophilus, deletions of the mutSL, mutY, ung, and mfd genes markedly enhanced mutability. These genes were subsequently deleted to construct an error-prone thermophile that showed much higher (700- to 9,000-fold) mutability than the parent strain. The error-prone strain was auxotrophic for uracil owing to the fact that the strain was deficient in the intrinsic pyrF gene. Although the strain harboring Bacillus subtilis pyrF was also essentially auxotrophic, cells became prototrophic after 2 days of culture under uracil starvation, generating B. subtilis PyrF variants with an enhanced half-denaturation temperature of >10°C. These data suggest that this error-prone strain is a promising host for thermoadaptation-directed evolution to generate thermostable variants from thermolabile enzymes.

FIG 1

pGKE70 structure. pGKE70 was constructed to integrate genes into the trpE locus and to force gene expression under the control of the P_gk704 promoter. Amp^R, ampicillin resistance gene; pUC, pUC replicon; oriT, conjugative-transfer origin; TK101, thermostable kanamycin nucleotidyltransferase gene; P_gk704, the P_gk704 promoter functional in G. kaustophilus (29); and trpE, an internal and defective gene for anthranilate synthase component I of strain HTA426. Restriction enzyme sites unique to multiple cloning sites are also indicated.

FIG 2

Mutability of G. kaustophilus, E. coli, and B. subtilis. Generation frequencies of Rif^r or Str^r cells (per 10⁹ viable cells) from G. kaustophilus MK242^p70, E. coli DH5α, and B. subtilis 168 were analyzed using Rif and Str at low (solid bars) and high (open bars) concentrations. Analysis was performed using four independent culture experiments (n = 4), and the data are presented as means and standard errors (SE).

FIG 3

In-frame deletions in DNA repair genes in G. kaustophilus MK242. Correct deletions to generate the ΔmutSL (A), ΔmutM (B), ΔmutY (C), ΔmutT (D), Δung (E), and Δmfd (F) mutants were verified by Southern blotting. The open boxes represent DNA probes. The dotted lines indicate the deletion regions. The numbers indicate DNA fragment lengths (kb).

FIG 4

Effects of deletion and complementation of mutSL, mutM, mutY, mutT, ung, and mfd genes on G. kaustophilus mutability. The generation frequencies of Rif^r or Str^r cells were evaluated using Rif and Str at 10 mg liter⁻¹ (solid bars) and 50 mg liter⁻¹ (open bars). Analyses were performed in three to five independent experiments for each condition (n = 3 to 5). The data are presented as means and SE of fold changes relative to the mean generation frequency of the MK242^p70 strain.

FIG 5

Comparison of mutability (A) and growth curves (B) between the G. kaustophilus strains MK242^p70 and MK480^p70. (A) Generation frequencies of Rif^r or Str^r cells from MK480^p70 were evaluated using Rif and Str at 10 mg liter⁻¹ (solid bars) and 50 mg liter⁻¹ (open bars). The data are presented as means and SE (n = 5) of fold changes relative to the mean generation frequency of the MK242^p70 strain. (B) The strains MK242^p70 and MK480^p70 were cultured at 60°C in LB medium with monitoring of the OD₆₀₀.

FIG 6

Generation of thermostable variants BSpyrF_e1 and BSpyrF_e2 from BSpyrF_wt. (A) Phenotype transition from prototrophy to auxotrophy for uracil in G. kaustophilus MK480^BSpyrF. G. kaustophilus MK480^p70, MK480^BSpyrF, and MK242^BSpyrF were incubated in MM at 60°C (left) and 65°C (middle). MK480^BSpyrF and MK242^BSpyrF cells that had been grown at 60°C were incubated in MM at 65°C (right). (B) Mutations in BSpyrF_e1 and BSpyrF_e2 genes. BSpyrF_e1 carried C497T, and BSpyrF_e2 carried C169A and C497T mutations. The C169A and C497T mutations are responsible for L57I and A166V amino acid substitutions, respectively.

FIG 7

Thermostability assay of BSpyrF_wt, BSpyrF_e1, and BSpyrF_e2 proteins. (A) SDS-PAGE analysis of purified recombinant BSpyrF_wt (lane 1), BSpyrF_e1(lane 2), and BSpyrF_e2 (lane 3) proteins with molecular markers (lane M; kDa). (B) Thermal denaturation curves of BSpyrF_wt (solid circles), BSpyrF_e1 (open circles), and BSpyrF_e2 (solid squares) proteins. The protein solution was incubated for 1 h at the indicated temperatures and centrifuged. Proteins retained in the supernatant were analyzed by the Bradford method. The data are presented as means and SE (n = 4 to 8). The amount of protein remaining after 30°C incubation was taken to be 100%. (C) Time course analysis of BSpyrF_wt (solid circles), BSpyrF_e1 (open circles), and BSpyrF_e2 (solid squares) aggregations. The protein solution was incubated at 30°C (left) and 37°C (right) for the indicated periods. After centrifugation, proteins retained in the supernatant were analyzed by the Bradford method. The data are presented as means (n = 4; SE are less than 3% [not shown]).

FIG 8

Three-dimensional structures surrounding the L57 (A) and A166 (B) residues of the BSpyrF^wt protein.