A thiostrepton resistance gene and its mutants serve as selectable markers in Geobacillus kaustophilus HTA426

Abstract

Effective utilization of microbes often requires complex genetic modification using multiple antibiotic resistance markers. Because a few markers have been used in Geobacillus spp., the present study was designed to identify a new marker for these thermophiles. We explored antibiotic resistance genes functional in Geobacillus kaustophilus HTA426 and identified a thiostrepton resistance gene (tsr) effective at 50 °C. The tsr gene was further used to generate the mutant tsr^H258Y functional at 55 °C. Higher functional temperature of the mutant was attributable to the increase in thermostability of the gene product because recombinant protein produced from tsr^H258Y was more thermostable than that from tsr. In fact, the tsr^H258Y gene served as a selectable marker for plasmid transformation of G. kaustophilus. This new marker could facilitate complex genetic modification of G. kaustophilus and potentially other Geobacillus spp.

Graphical abstract

A thiostrepton resistance gene (tsr) and its mutant were identified as new selectable markers for Geobacillus kaustophilus HTA426.

Key words:

• Geobacillus kaustophilus
• thiostrepton resistance
• thermoadaptation-directed evolution
• error-prone thermophile
• selectable marker

The members of the genus Geobacillus include aerobic or facultatively anaerobic, gram-positive, and Bacillus-related thermophiles.¹⁾ These organisms have been isolated from a wide range of environments²⁾ and usually exhibit optimum growth at temperatures ranging from 55 to 60 °C. Because of their thermophilic properties, the members of this genus can be used in microbial bioprocesses performed at high temperature. High-temperature bioprocesses have considerable advantages over ordinary-temperature bioprocesses, as exemplified by the growth inhibition of animal pathogens, low energy expenditure for agitation and cooling, and easy removal of volatile products.³⁾ In addition, several Geobacillus spp. have exhibited remarkable properties useful in microbial bioprocesses. For instance, G. thermoglucosidasius exhibits ethanol tolerance and has been studied for high-temperature ethanol and isobutanol production.^4–7) Another strain can produce acetoin and 2,3-butanediol.⁸⁾ Some others have the ability to utilize cellulose⁹⁾ or hemicelluloses,¹⁰⁾ to accumulate arsenate¹¹⁾ or toxic metal ions¹²⁾ and to degrade hydrocarbons,¹³⁾ long-chain alkanes,^14,15) herbicides,²⁾ or polyvinyl alcohol.¹⁶⁾ These properties make Geobacillus spp. attractive for use in high-temperature bioprocesses intended to bioproduction, biodegradation, biomineralization, and bioremediation.

In general, effective utilization of microbes requires genetic modification such as gene inactivation and forced gene expression to enhance the bioactivity and/or alter the metabolic flux of the microbes. Genetic modification is commonly performed using antibiotic resistance genes as selectable markers. In addition, the combined use of multiple markers enables more complex microbial modifications, including the stable maintenance of multiple plasmids and inactivation of multiple genes. However, fewer antibiotic resistance genes are used in thermophiles relative to mesophiles. In Geobacillus spp., only kanamycin, chloramphenicol, and tetracycline resistance genes have been demonstrated to serve as selectable markers.^5,17–23) Therefore, there is room to expand new markers for more effective utilization of Geobacillus spp. through complex genetic modifications.

Geobacillus kaustophilus HTA426 is an aerobic strain that had been isolated from deepest sea mud of the Mariana Trench.^24,25) We have studied this strain as a model for biotechnological applications of Geobacillus spp., because basal genetic tools have been established for this strain^18,26–28) and a whole genome sequence is also available.²⁹⁾ In this study, we explored antibiotic resistance genes functional in G. kaustophilus HTA426 to expand selectable markers for Geobacillus spp. Here, we report a new marker, tsr, which is a thiostrepton resistance gene from Streptomyces azureus.³⁰⁾ Thiostrepton inhibits protein translation by firmly binding to the complex formed by 23S rRNA and ribosomal protein L11 in bacterial ribosomes.³¹⁾ The tsr gene encodes an RNA methyltransferase that prevents thiostrepton from binding to ribosomes by 23S rRNA methylation.³²⁾ We also generated tsr mutants and characterized their availability as selectable markers.

Materials and methods

Bacterial strains, media, and plasmids

Table 1 lists G. kaustophilus strains used. An error-prone strain, MK480, was used for thermoadaptation-directed evolution of tsr gene. Strain MK242 was used for gene characterization because it is genetically stable. These strains were grown at 60 °C in Luria–Bertani (LB) medium. However, G. kaustophilus [pGKE75 derivative], where square brackets denote the plasmid-carrier state, was grown in LB medium supplemented with 5 mg/L of kanamycin (LK5). Escherichia coli DH5α (Takara Bio, Otsu, Shiga, Japan) and E. coli BL21-CodonPlus(DE3)-RIPL (Agilent Technologies, Santa Clara, USA) were used for DNA manipulation and recombinant protein production, respectively. E. coli strains were grown at 37 °C in LB medium. Ampicillin (50 mg/L), kanamycin (50 mg/L), and chloramphenicol (13 mg/L) were added when necessary. The plasmid pGKE75 was constructed previously.³³⁾ pUC19 and pET-16b were purchased from Takara Bio and Merck KGaA (Darmstadt, Germany), respectively.

Table 1. G. kaustophilus strains used in this study.

Strain	Relevant description	Reference
MK242	Derivative of the wide-type strain HTA426; ΔpyrF ΔpyrR ΔhsdM1S1R1 Δ(mcrB1-mcrB2-hsdM2S2R2-mrr) GK0707::Pgk704-bgaB	35
MK480	Derivative of the strain MK242; ΔpyrF ΔpyrR ΔhsdM1S1R1 Δ(mcrB1-mcrB2-hsdM2S2R2-mrr) GK0707::Pgk704-bgaB ΔmutSL ΔmutY Δung Δmfd	35

Notes: G. kaustophilus MK242 and MK480 lack genes related to pyrimidine biosynthesis (pyrF and pyrR) and DNA restriction–modification (hsdM₁S₁R₁, mcrB₁, mcrB₂, hsdM₂S₂R₂, and mrr). The MK480 strain further lacks genes related to DNA repair (mutSL, mutY, ung, and mfd). Both strains contained bgaB gene under the control of P_gk704 promoter at the GK0707 chromosomal locus (GK0707::P_gk704-bgaB).

Construction of pGKE75-tsr

The tsr gene was amplified from plasmid pIJ4123³⁴⁾ using primers 5′-GCCGCATGCCTGAGTTGGACACCATC-3′ (SphI site underlined; start codon italicized) and 5′-GCCGGATCCTTATCGGTTGGCCGCGAGATTC-3′ (BamHI site underlined; stop codon italicized). The amplified fragment was trimmed with SphI and BamHI and subcloned between the SphI and BamHI sites of pGKE75 to generate pGKE75-tsr (Fig. 1).

Fig. 1. pGKE75-tsr structure.

Notes: pGKE75-tsr is an E. coli–Geobacillus shuttle plasmid used to force tsr expression in G. kaustophilus. Abbreviations: P_gk704, a promoter functional in G. kaustophilus²⁸⁾; tsr, a thiostrepton resistance gene from S. azureus³⁰⁾; bla, an ampicillin resistance gene; TK101, a thermostable kanamycin nucleotidyltransferase gene¹⁹⁾; pUC, pUC replicon functional in E. coli; pBST1, pBST1 replicon functional in Geobacillus spp⁵⁾.; and oriT, conjugative transfer origin from IncP-1 plasmid RK2.

Plasmid transformation of G. kaustophilus

The derivatives of pGKE75 were introduced into G. kaustophilus by conjugative DNA transfer from E. coli.²⁷⁾ E. coli [pGKE75-tsr derivative] and E. coli [pUB307] were used as the DNA donor and conjugation helper, respectively. G. kaustophilus MK242 was used as the DNA recipient. These strains were cultured until the optical density at 600 nm reached 0.3. Cultures (E. coli, 10 mL; G. kaustophilus, 100 mL) were then mixed and concentrated by centrifugation. Cell mixture was spotted on LB plates and incubated for 4 h at 37 °C to achieve conjugation. When kanamycin was used for transformant selection, cells were recovered from LB plates and incubated at 60 °C on LK5 plates. When thiostrepton was used, cells were recovered following further preincubation at 52 °C for 4 h and subsequently incubated on LB plates supplemented with thiostrepton (1 mg/L) at 52 °C to isolate transformants.

Thiostrepton resistance assay

G. kaustophilus MK242 [pGKE75-tsr derivative] was precultured at 50 °C and then incubated at 50, 55, and 60 °C on LK5 plates with and without thiostrepton (0.5 mg/L). The colonies were counted to calculate the thiostrepton resistance efficiency, which was defined as the ratio of thiostrepton-resistant cells (grown on LK5 with thiostrepton) to the total number of cells (grown on LK5 without thiostrepton).

Generation of the tsr^H258Y mutant using thermoadaptation-directed evolution

G. kaustophilus MK480 [pGKE75-tsr] was incubated on LK5 plates containing thiostrepton (0.1 mg/L) at 45 °C for 2 days. The cells were collected and further incubated on LK5 plates containing thiostrepton (0.5 mg/L) at 50 °C for 1 day. Similar successive cultivations were repeated on LK5 plates with thiostrepton (0.5 mg/L) at 55 °C for 3 days, 55 °C for 1 day, and 60 °C for 1 day. In addition, apparent colonies were regrown on LK5 plates with thiostrepton (0.5 mg/L) at 60 °C. pGKE75-tsr was isolated from the clones that showed evident thiostrepton resistance, and the tsr sequence was analyzed.

Site-directed mutagenesis

The tsr gene was excised from pGKE75-tsr with SphI and BamHI and subcloned between the SphI and BamHI sites of pUC19. The resulting plasmid was used as a template to amplify tsr partial fragments (361 bp) with the primers 5′-GGGGATCATCCTGGTAGACAGTGACATCACC-3′ and 5′-CCTGTCGATCCTCTCxxxCAGCGCGATTCCGAG-3′ (xxx: CGC for tsr^H258A; GCA for tsr^H258C; GTC for tsr^H258D; TTC for tsr^H258E; AAA for tsr^H258F; GCC for tsr^H258G; GAT for tsr^H258I; TTT for tsr^H258K; CAA for tsr^H258L; CAT for tsr^H258M; GTT for tsr^H258N; CGG for tsr^H258P; TTG for tsr^H258Q; GCG for tsr^H258R; CGA for tsr^H258S; CGT for tsr^H258T; GAC for tsr^H258V; CCA for tsr^H258W). Using the fragments as mutagenic primers, pUC19 harboring tsr was replicated in vitro to generate plasmids carrying the tsr mutants. The template plasmid was digested with DpnI treatment. The resulting DNA was subsequently propagated in E. coli. The tsr mutants were excised from the obtained plasmids and subcloned between the SphI and BamHI sites of pGKE75 to generate pGKE75-tsr derivatives, including pGKE75-tsr^H258W and pGKE75-tsr^H258F.

Preparation of recombinant proteins

The tsr, tsr^H258Y, tsr^H258W, and tsr^H258F genes were amplified from appropriate pGKE75-tsr derivatives using the primers 5′-GCCCATATGCCTGAGTTGGACACCATC-3′ (NdeI site underlined; start codon italicized) and 5′-GCCGGATCCTTATCGGTTGGCCGCGAGATTC-3′ (BamHI site underlined; stop codon italicized) and subcloned between the NdeI and BamHI sites of pET-16b. The resulting plasmid was introduced into E. coli BL21-CodonPlus(DE3)-RIPL. The transformant was cultured at 20 °C for 20 h in LB medium containing 1% lactose. Cells were harvested by centrifugation (5000 × g for 5 min) and homogenized in buffer (20 mM sodium phosphate, pH 7.4, 0.5 M NaCl) by ultrasonication. The lysate was clarified by centrifugation (18,000 × g for 10 min at 4 °C) and then applied to a Talon column (metal ion, Co²⁺; bed volume, 0.2 mL; Takara Bio) equilibrated with buffer. The column was washed with buffers containing 20 mM imidazole (2 mL) and then 50 mM imidazole (0.5 mL). Subsequently, recombinant proteins were eluted with buffer containing 200 mM imidazole. Purified proteins were dialyzed against 10 mM sodium phosphate (pH 7.0) containing 0.5 M NaCl. The purity of proteins was analyzed using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Protein concentration was determined by the Bradford method using bovine serum albumin as a standard.

Thermostability assay of Tsr proteins

A protein solution (20 µL; 0.4 g/L protein) was incubated at 40–55 °C for 1 h in a gradient temperature heater (Thermal Cycler Dice Gradient; Takara Bio). Aggregated proteins were removed by centrifugation (18,000 × g for 10 min at 4 °C). Soluble proteins retained in the supernatant were analyzed by SDS-PAGE and the Bradford method. The temperature at which the residual protein = 50% (T_1/2 value) was calculated by the linear regression of thermal aggregation curves (Tsr, 44.3–48.3 °C; Tsr^H258Y, 48.3–51.1 °C; Tsr^H258W, 48.3–51.1 °C; Tsr^H258F, 44.3–48.3 °C).

Gel filtration column chromatography

Gel filtration was performed using a Superdex 75 10/300 GL column (GE Healthcare Japan, Tokyo, Japan) and an ÄKTA pure 25 system (GE Healthcare Japan) by isocratic elution in 20 mM sodium phosphate (pH 7.4) containing 0.5 M NaCl at a flow rate of 0.5 mL/min. Conalbumin (75 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), chymotrypsinogen (25 kDa), and ribonuclease A (14 kDa) were used as molecular weight standards.

Thermostability assay of thiostrepton

LB plates supplemented with thiostrepton (0.5 mg/L) were preincubated at 50, 55, and 60 °C for 24, 48, and 72 h. These plates were then used for incubation of G. kaustophilus MK242 (10⁹ cells) at 50 °C to assess effective thiostrepton remained.

Results

The tsr gene confers thiostrepton resistance on G. kaustophilus

Thiostrepton, ampicillin, and spectinomycin resistance genes had been retained in our laboratory; therefore, we tentatively examined the utility of these genes as selectable markers for G. kaustophilus. Spectinomycin (100 mg/L) was ineffective against G. kaustophilus. Ampicillin was effective but the ampicillin resistance gene (bla) from pUC19 could not confer resistance to ampicillin (0.1 mg/L) on G. kaustophilus even at 45 °C. However, although G. kaustophilus was intrinsically sensitive to thiostrepton at 45–70 °C (IC₅₀ = 10–50 µg/L), G. kaustophilus MK242 [pGKE75-tsr] could grow in LK5 media supplemented with thiostrepton (0.5 mg/L) at 50 °C following preculture at 50 °C (Fig. 2(A)). Preculture was essential for thiostrepton resistance presumably because of sufficient rRNA methylation prior to thiostrepton exposure. It was likely that Tsr protein undergoes thermal denaturation at 55 °C, because MK242 [pGKE75-tsr] could not grow at 55 °C (Fig. 2(B)) and cells that were precultured at 55 °C were sensitive to thiostrepton even at 50 °C (data not shown).

Fig. 2. Thiostrepton resistance conferred by tsr and tsr^H258Y genes.

Notes: G. kaustophilus MK242 harboring pGKE75 (Empty), pGKE75-tsr (tsr), and pGKE75-tsr^H258Y (tsr^H258Y) were precultured in liquid LK5 at 50 °C and subsequently cultured using LK5 media at 50 °C (A) and 55 °C (B) for 36 h. The media contained thiostrepton at the indicated concentration (mg/L).

Thermostability of thiostrepton

LB plates supplemented with thiostrepton completely inhibited G. kaustophilus growth even after preincubation at 50–55 °C for 48 h (Table 2). However, a few cells grew on the plates preincubated at 55 °C for 72 h. When the plate was preincubated at 60 °C, thiostrepton was obviously inactivated by 72-h incubation. These results suggest that thiostrepton (0.5 mg/L) is effective at temperatures below 55 °C for 48 h but may be inactivated by longer or higher temperature incubations.

Table 2. The frequency of cells growing on LB plates that contained thiostrepton but were preincubated at high temperatures.

Preincubation time	Preincubation temperature
	50 °C	55 °C	60 °C
24 h	<1 × 10^−9	<1 × 10^−9	<1 × 10^−9
48 h	<1 × 10^−9	<1 × 10^−9	3.5 × 10^−9 ± 3.1 × 10^−9
72 h	<1 × 10^−9	7.2 × 10^−9 ± 5.9 × 10^−9	1.9 × 10^−6 ± 1.0 × 10^−6

Notes: G. kaustophilus MK242 (10⁹ cells) was grown on LB plates with thiostrepton (0.5 mg/L) following preincubation of the plate under the indicated conditions. The colonies were counted to calculate the frequency of cells that showed apparent thiostrepton resistance per total cells. The data are presented as means ± standard deviations (n = 4).

Plasmid transformation of G. kaustophilus using tsr as a selectable marker

To determine whether tsr actually serves as a selectable marker in G. kaustophilus, we introduced pGKE75-tsr into strain MK242 and selected transformants using thiostrepton selection. Plasmid introduction was performed using conjugative DNA transfer from E. coli, because it was an only practical method for plasmid introduction into G. kaustophilus. In this method, plasmid was transferred from E. coli [pGKE75-tsr] to G. kaustophilus with the mediation of E. coli [pUB307] during incubation of their cell mixture at 37 °C. The mixture was then preincubated at 52 °C for intracellular Tsr production and 23S rRNA methylation. Subsequently, an aliquot of cells was incubated at 52 °C on LB plates supplemented with thiostrepton. This process successfully provided 78 colonies after 3-day incubation, although an equal aliquot provided 188 colonies on LK5 plates within 24 h. Repeated analysis indicated that the colonies obtained using thiostrepton accounted for approximately 42% of those obtained using kanamycin (Table 3). Several transformants obtained using thiostrepton were sensitive to kanamycin. These false positives can be attributed to thermal inactivation of thiostrepton by 3-day incubation (Table 2). When preincubation and incubation were performed at temperatures above 53 °C, few transformants grew probably because of Tsr denaturation and subsequent insufficient 23S rRNA methylation.

Table 3. Plasmid transformation of G. kaustophilus with pGKE75-tsr derivatives using thiostrepton.

Plasmid	Incubation time^a	Selection efficiency (%)^b			False positive^c
		52 °C	53 °C	54 °C	52 °C	53 °C	54 °C
pGKE75-tsr	>3 days	42 ± 3	2 ± 2	1 ± 1	7/80	0/11	6/6
pGKE75-tsr^H258Y	48 h	66 ± 8	64 ± 8	50 ± 9	0/80	0/80	0/64
pGKE75-tsr^H258W	60 h	52 ± 6	45 ± 7	11 ± 6	3/80	2/80	1/29

^a Incubation time required until transformants form obvious colonies.

^b Transformant cells were incubated at the indicated temperatures on LB plates containing thiostrepton or kanamycin. Selection efficiency was defined as the ratio of transformant numbers obtained using thiostrepton to those obtained using kanamycin. Data are presented as means ± standard errors (n = 4).

^c Transformants obtained using thiostrepton were checked for kanamycin resistance to determine false positives. A maximum 20 clones were analyzed per experiments (n = 4); when fewer than 20 transformants were present, all clones were analyzed. Data are presented as the number of kanamycin-sensitive clones per the total number of analyzed clones.

Generation of the tsr^H258Y mutant by thermoadaptation-directed evolution

We previously constructed an error-prone thermophile, G. kaustophilus MK480, and demonstrated that this strain can generate mutant genes encoding more thermostable variants than the parent enzymes during periods of cell growth at high temperatures.^33,35) The results suggested the possibility that this approach, thermoadaptation-directed evolution, might generate tsr mutants that confer substantial thiostrepton resistance on G. kaustophilus at temperatures above 50 °C. Thus, we examined thermoadaptation-directed tsr evolution by serially culturing G. kaustophilus MK480 [pGKE75-tsr] under thiostrepton selection pressures. This approach generated clones that were resistant to thiostrepton at 60 °C. pGKE75-tsr was isolated from eight clones and analyzed for the tsr sequence. All eight plasmids carried the tsr^H258Y mutant that has C772T mutation and encodes H258Y protein variant (Tsr^H258Y).

Thiostrepton resistance conferred by tsr mutants

In addition to tsr^H258Y, we generated additional tsr mutants (tsr^H258A, tsr^H258C, tsr^H258D, tsr^H258E, tsr^H258F, tsr^H258G, tsr^H258I, tsr^H258K, tsr^H258L, tsr^H258M, tsr^H258N, tsr^H258P, tsr^H258Q, tsr^H258R, tsr^H258S, tsr^H258T, tsr^H258V, and tsr^H258W) by in vitro mutagenesis approach. These saturation mutants regarding the 258th codon were subcloned in pGKE75 and introduced into G. kaustophilus MK242 to analyze the thiostrepton resistance conferred by the genes. As expected, pGKE75-tsr^H258Y conferred the thiostrepton resistance at 55 °C (Fig. 2(B)). Quantitative analysis showed that approximately 68% of MK242 [pGKE75-tsr^H258Y] cells were resistant to thiostrepton (Fig. 3). pGKE75-tsr^H258W also conferred substantial resistance at 55 °C. The tsr^H258F, tsr^H258A, tsr^H258G, tsr^H258I, and tsr^H258R mutants were functional at 50 °C but not 55 °C. Other mutants were not functional even at 50 °C.

Fig. 3. Comparison of thiostrepton resistance conferred by tsr mutants.

Notes: G. kaustophilus MK242 [pGKE75-tsr derivative] was preincubated at 50 °C and subsequently incubated on LK5 plates with and without thiostrepton (0.5 mg/L) at 50 °C (solid bars) or 55 °C (hollow bars). Thiostrepton resistance efficiency was defined as the ratio of thiostrepton-resistant cells per total cells. Data are presented as means ± standard errors (n = 4). MK242 [pGKE75-tsr^H258C, -tsr^H258D, -tsr^H258E, -tsr^H258K, -tsr^H258L, -tsr^H258M, -tsr^H258N, -tsr^H258P, -tsr^H258Q, -tsr^H258S, -tsr^H258T, or -tsr^H258V] was sensitive to thiostrepton at 50 °C and 55 °C (efficiency, <1%). The efficiency was <1% at 60 °C for all strains examined.

Plasmid transformation of G. kaustophilus using tsr^H258Y and tsr^H258W as selectable markers

When G. kaustophilus was transformed with pGKE75-tsr using thiostrepton selection, it was necessary for 3 days to identify transformant colonies (see above). However, pGKE75-tsr^H258Y provided transformants within 48 h even when preincubation and incubation were performed at 53 °C or 54 °C (Table 3). Moreover, the selection was achieved with higher efficiency relative to tsr and produced no false positives. pGKE75-tsr^H258W also provided transformants, but these contained false positives and the selection efficiency was lower than that of pGKE75-tsr^H258Y (Table 3). Thus, tsr^H258Y served as the most efficient marker among tsr and its mutants examined.

Thermostability of Tsr, Tsr^H258Y, Tsr^H258W, and Tsr^H258F proteins

Tsr, Tsr^H258Y, Tsr^H258W, and Tsr^H258F proteins were heterologously produced as soluble proteins fused with an N-terminal octahistidine-tag and were purified using immobilized metal ion affinity chromatography (Fig. 4(A)). Gel filtration analysis of the purified proteins revealed single protein peaks corresponding to 64–68 kDa. These observations indicated that Tsr forms a homodimer, as suggested by X-ray crystallographic analysis,³⁶⁾ and that the homodimeric structure is not disrupted by H258Y, H258W, and H258F replacements. Because it is difficult to quantify Tsr activity in vitro, we analyzed thermostability of recombinant proteins using protein aggregation assay (Fig. 4(B) and (C)). Tsr protein was completely aggregated by incubation at approximately 48 °C. A similar profile was observed for Tsr^H258F. However, complete aggregation of Tsr^H258Y and Tsr^H258W was observed at temperatures above 51 °C. The T_1/2 values of Tsr, Tsr^H258Y, Tsr^H258W, and Tsr^H258F were 45.3, 49.5, 49.3, and 45.7 °C, respectively. These data indicated that Tsr^H258Y and Tsr^H258W are more thermostable than Tsr and Tsr^H258F.

Fig. 4. Thermostability of Tsr, Tsr^H258Y, Tsr^H258W, and Tsr^H258F proteins.

Notes: (A) SDS-PAGE analysis of molecular markers (left lane) and Tsr protein purified (right lane). Purified Tsr^H258Y, Tsr^H258W, and Tsr^H258F proteins also exhibited comparable purity. (B) A protein solution was incubated at the indicated temperatures for 1 h, followed by centrifugation. Proteins retained in the supernatant were analyzed by SDS-PAGE. (C) Protein solutions of Tsr (solid circles), Tsr^H258Y (hollow circles), Tsr^H258W (solid squares), and Tsr^H258F (hollow squares) were incubated for 1 h and centrifuged. Protein concentration in the supernatant was determined using the Bradford method. The amount of protein remaining after incubation at 40 °C was considered to be 100%.

Discussion

While the tsr gene has been used as a selectable marker exclusively in Streptomyces spp.,³⁷⁾ we found that tsr is functional at 50 °C in G. kaustophilus HTA426 even though the performance as a selectable marker was unsatisfactory. To improve the performance, we examined thermoadaptation-directed evolution of tsr and identified a more efficient marker, tsr^H258Y, which conferred thiostrepton resistance at 55 °C. The enhanced functional temperature is attributable to the enhanced thermostability of the gene product, as the recombinant Tsr^H258Y protein was more thermostable than Tsr. The T_1/2 enhancement by the H258Y replacement was only 4.2 °C, but tsr^H258Y efficiently selected transformants within 48 h at temperatures higher than 52 °C. The thiostrepton concentration required for transformant selection was only 1 mg/L. Moreover, the selection produced no false positives. This is probably due to short periods required for transformant selection because thiostrepton was substantially effective at 50–55 °C within 48 h although not over the time. tsr^H258Y marker may be unsuitable for stable plasmid maintenance at temperatures above 55 °C but is available for complex genome modification, such as inactivation and integration of targeted genes, along with other proven selectable markers. Thus, tsr^H258Y and thiostrepton constitute a practical marker system for G. kaustophilus.

Thiostrepton resistance assay indicated that MK242 [pGKE75-tsr^H258Y] cells resistant to thiostrepton were less than those resistant to kanamycin (Fig. 3). In G. kaustophilus transformation, transformants selected by tsr^H258Y with thiostrepton did not outreach transformants selected by the kanamycin resistance gene with kanamycin (Table 3). Preincubation was also required to select transformants using thiostrepton. These observations can be explained by inefficient expression of tsr^H258Y due to codon bias, because tsr^H258Y gene has a high GC content (61.4%) in contrast to a moderate content in G. kaustophilus (52.1%).²⁹⁾ This idea is supported by negligible production of Tsr^H258Y protein in E. coli BL21(DE3), which has no extra copies of rare tRNA genes unlike E. coli BL21-CodonPlus(DE3)-RIPL. Therefore, more efficient tsr^H258Y marker may be generated by codon optimization of the gene for G. kausotphilus.

In the course of tsr^H258Y generation, G. kaustophilus MK480 [pGKE75-tsr] cells had adapted to thiostrepton at 60 °C, whereas MK242 cells that had been freshly transformed with pGKE75-tsr^H258Y were sensitive to thiostrepton at 60 °C. This difference may arise from a factor to enhance thiostrepton resistance during thiostrepton-exposed cultivations, such as ribosomal gene mutation to prevent thiostrepton from binding to ribosomes,³⁸⁾ activation of drug efflux pumps to eliminate thiostrepton to the extracellular space, and induction of protein chaperones (or chaperone-like factors) to thermostabilize the Tsr protein in vivo. Because of such adaptation, our attempts to generate other variants from tsr^H258Y have not yielded mutations, even though MK480 [pGKE75-tsr^H258Y] cells were found to adapt to thiostrepton at 70 °C. Therefore, we conducted saturation mutagenesis regarding the H258 codon to generate a tsr mutant that confers thiostrepton resistance more efficiently than does tsr^H258Y. Unfortunately, this approach did not provide a mutant that functions at 60 °C but determined that tsr^H258W functions at 55 °C as with tsr^H258Y. The enhanced functional temperature is also attributable to the enhanced thermostability of the gene product, as suggested by thermostability assay of Tsr^H258W protein (Fig. 4).

In a homodimeric structure of the Tsr protein,³⁶⁾ two pairs of K204 and E259 residues interact with tetragonal ionic networks at the contact surface between the two subunits (Fig. 5). The two H258 residues are located on the same side of the tetragonal ionic network, in which the residue of each subunit interacts to reciprocally with the E259 residue of the other subunit. Thus, the H258 residues appear to optimize the ionic networks formed between K204 and E259 residues. This observation implies that replacement of histidine at residue 258 to tyrosine or tryptophan, which is aromatic and bulky amino acid, may optimize the ionic networks and thereby render the structure more robust. However, Tsr^H258F protein showed the thermostability comparable to Tsr protein despite an aromatic side chain at residue 258 as with Tsr^H258Y and Tsr^H258W. Moreover, there was no clear association between the side chain structures of residue 258 in the Tsr variants and their functionality at 55 °C. Therefore, it is likely that the H258 replacements affected the catalytic activity and/or protein thermostability through complex molecular mechanisms.

Fig. 5. Three-dimensional structure of Tsr protein around the H258 residues.

Notes: In a homodimeric Tsr structure (PDB ID: 3GYQ), the K204 and E259 residues of one subunit (K204^A and E259^A) interact with those of the other subunit (K204^B and E259^B) through tetragonal ionic networks. In addition, the H258 residue of one subunit (H258^A) interacts with E259^B residue, while H258^B interacts reciprocally with E259^A.

In summary, we generated tsr^H258Y as a selectable marker available for G. kaustophilus. This new marker extends the genetic toolbox of G. kaustophilus and potentially other Geobacillus spp., providing new opportunities for the effective utilization of these thermophiles through complex genetic modifications.

Author contributions

H.S. designed the study. K.W., J.K., M.F., and H.S. performed the experiments. K.W., K.D, T.O., and H.S. analyzed the data. K.W. and H.S. wrote the manuscript. K.D. and T.O. contributed to the critical revision of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by the Programme for Promotion of Basic and Applied Researches for Innovations in Bio-oriented Industry, Japan, under Grant: The Science and Technology Research Promotion Program for Agriculture, Forestry, Fisheries and Food Industry, Japan [grant number 26033A]; JSPS KAKENHI [grant number 25450105]; The Institute for Fermentation, Osaka, Japan.