Missense Mutations Allow a Sequence-Blind Mutant of SpoIIIE to Successfully Translocate Chromosomes during Sporulation

Abstract

SpoIIIE directionally pumps DNA across membranes during Bacillus subtilis sporulation and vegetative growth. The sequence-reading domain (γ domain) is required for directional DNA transport, and its deletion severely impairs sporulation. We selected suppressors of the spoIIIEΔγ sporulation defect. Unexpectedly, many suppressors were intragenic missense mutants, and some restore sporulation to near-wild-type levels. The mutant proteins are likely not more abundant, faster at translocating DNA, or sequence-sensitive, and rescue does not involve the SpoIIIE homolog SftA. Some mutants behave differently when co-expressed with spoIIIEΔγ, consistent with the idea that some, but not all, variants may form mixed oligomers. In full-length spoIIIE, these mutations do not affect sporulation, and yet the corresponding residues are rarely found in other SpoIIIE/FtsK family members. The suppressors do not rescue chromosome translocation defects during vegetative growth, indicating that the role of the γ domain cannot be fully replaced by these mutations. We present two models consistent with our findings: that the suppressors commit to transport in one arbitrarily-determined direction or delay spore development. It is surprising that missense mutations somehow rescue loss of an entire domain with a complex function, and this raises new questions about the mechanism by which SpoIIIE pumps DNA and the roles SpoIIIE plays in vivo.

Fig 1. Intragenic spoIIIEΔγ suppressor mutations alter residues in the linker and motor domains.

A. Positions of intragenic suppressor mutations are indicated on a schematic of the spoIIIEΔγ linear sequence. The γ domain is not shown, but is C-terminal to the β subdomain in full-length SpoIIIE. Numbers in parentheses indicate the number of times each mutation was isolated. Underlined mutations were identified only after the P492 codon was mutated from ccg to cct to lessen the chances of obtaining P492Q. In this study, SpoIIIE codons are numbered for the protein beginning with the sequence MSVAKKKRKS. This presumes an earlier translation start and thus the codons are numbered +2 relative to some annotations of SpoIIIE [50]. (B-E). Positions of intragenic suppressor mutations are shown in a 3D-model of the SpoIIIE motor domain, obtained by threading the SpoIIIE sequence onto a FtsK crystal structure [25, 44]. B. Two of the six subunits of a SpoIIIE hexamer are shown. The Walker A and Walker B sites are shown in red and ADP is shown in brown. P319 (pink), A343 (orange), E347 (yellow), P492 (green), H493 (cyan), D586 (blue), and T617 (purple) are shown as space-filled residues. C. A343 (orange) and E347 (yellow) lie on the same face of a helix in the α domain. D. P319 (pink), P492 (green) and H493 (cyan) are near each other. E. D586 (blue) and T617 (purple) are near the Walker B motif (red) in the β domain.

Fig 2. Missense mutations in spoIIIEΔγ rescue sporulation and chromosome transport in vivo.

A. Suppressor mutations rescue spore formation. Intragenic mutations identified by suppressor selection were remade in a spoIIIEΔγ allele by site-directed mutagenesis and expressed from the spoIIIE promoter at an ectopic locus (ycgO) in a ΔspoIIIE strain. Strains were induced to sporulate for 24–36 h in DSM medium, unsporulated cells were eliminated by heat-kill, and the number of spores was measured by plating for cfu. The number of heat-resistant spores per ml is indicated for strains harboring full-length (“f.l.”) spoIIIE (bKM776), spoIIIEΔγ (BOSE2042), and 11 spoIIIEΔγ mutants: P260L (BOSE2286), S264I (BOSE2540), E310K (BOSE2411), E312A (BOSE2121), Y316D (BOSE2284), P319S (BOSE2321), A343V (BOSE2288), E347G (BOSE2323), P492Q (BOSE2120), H493Y (BOSE2538), and T617A (BOSE2123). Each number is the average of at least 3 replicates. Error bars indicate one standard deviation in each direction. B. Suppressor mutations rescue chromosome transport in vivo. Sporulation was induced by resuspension and DNA transport was evaluated using a previously-established fluorescent microscopy assay [12]. yfp and cfp genes are expressed from a forespore-specific promoter (PspoIIQ). yfp is integrated near the origin (yycR), and its expression indicates that asymmetric septation is complete. cfp is integrated near the terminus (pelB), and its expression indicates that the terminus has been transported into the forespore. Percent of termini in forespores is the percent of YFP+ cells that are also CFP+. Data are shown for full-length (“f.l.”) spoIIIE (bBB128), spoIIIEΔγ (bBB412), and 5 spoIIIEΔγ mutants: P260L (BOSE2331), E312A (BOSE2201), A343V (BOSE2332), P492Q (BOSE2200), and T617A (BOSE2202). Each data point represents the average of ≥ 3 replicates, with ≥ 500 forespores scored for each.

Fig 3. SftA is not required for rescue of spore formation by spoIIIEΔγ intragenic suppressor mutations.

Strains were induced to sporulate for 24–36 h in DSM medium, unsporulated cells were eliminated by heat-kill, and the number of spores was measured by plating for cfu. Sporulation levels are similar for sftA+ and ΔsftA strains bearing spoIIIEΔγ mutant alleles: P260L (BOSE2286; BOSE2492), E312A (BOSE2121; BOSE2486), Y316D (BOSE2284; BOSE2490), P319S (BOSE2321; BOSE2496), A343V (BOSE2288; BOSE2494), E347G (BOSE2323; BOSE2498), P492Q (BOSE2120; BOSE2935), and T617A (BOSE2123; BOSE2488). Sporulation efficiencies for strains harboring full-length (“f.l.”) spoIIIE (bKM776) and spoIIIEΔγ (BOSE2042) are plotted for comparison. The average of at least 3 replicates is plotted. Error bars indicate one standard deviation in each direction.

Fig 4. Intragenic suppressor mutations do not alter protein levels or ATPase activity.

A. SpoIIIE levels are similar in strains expressing spoIIIEΔγ or a suppressor allele. Samples were harvested 2.5 h after cells were induced to sporulate by resuspension. Protein levels were evaluated using Western blots with antibodies against SpoIIIE. The upper arrow indicates the position of full-length SpoIIIE, and the lower arrow indicates the position of SpoIIIEΔγ on the blot. Asterisks indicate SpoIIIEΔγ variants whose migration was altered. All strains bear ΔspoIIIE::neo. As indicated, samples were from strains with full-length (“f.l.”) spoIIIE (1:10 dilution; bKM776), no ectopic spoIIIE (bDR1066), spoIIIEΔγ (BOSE2042), or a spoIIIEΔγ mutant: P260L (BOSE2286), S264I (BOSE2540), E310K (BOSE2411), E312A (BOSE2121), Y316D (BOSE2284), P319S (BOSE2321), A343V (BOSE2288), E347G (BOSE2323), P492Q (BOSE2120), H493Y (BOSE2538), T617A (BOSE2123). One representative set from at least three replicates is shown here. All three replicates are shown in S2B Fig. ATPase activity of intragenic suppressor mutants is similar to that of SpoIIIEΔγ. ATPase activity of soluble variants of SpoIIIE and SpoIIIEΔγ was measured using an NADH+-coupled assay, as described previously [11, 38, 41]. Results obtained in this study for these two proteins are similar to those obtained previously [38]. Three different intragenic suppressor mutations were introduced to the SpoIIIEΔγ-encoding plasmid by site-directed mutagenesis, and the resulting soluble versions of each mutant were expressed in and purified from E. coli. Each suppressor mutant exhibited the same ATPase activity as that of SpoIIIEΔγ. Averages of at least three replicates are plotted. Error bars indicate plus one standard deviation.

Fig 5. Spore formation in various genetic backgrounds.

Strains were induced to sporulate for 24–36 h in DSM medium, unsporulated cells were eliminated by heat-kill, and the number of spores was measured by plating for cfu. The average of at least 3 replicates is plotted. Error bars indicate one standard deviation in each direction. A. Intragenic mutants are partially or fully dominant to spoIIIEΔγ. Strains expressing both spoIIIEΔγ and an intragenic spoIIIEΔγ suppressor mutant exhibit sporulation efficiencies that are intermediate between those of strains with only the spoIIIEΔγ or the mutant allele, or that are similar to that of the mutant allele. All displayed strains express the indicated spoIIIE allele from the spoIIIE promoter at the ectopic locus ycgO. White bars indicate strains that also express spoIIIEΔγ from the spoIIIE promoter at the ectopic locus yhdGH. Asterisks mark pairs of sporulation efficiencies that were significantly different from each other by t-tests (p<0.05). Sporulation efficiencies are shown for spoIIIEΔγ (BOSE2042; BOSE2301) and 10 spoIIIEΔγ mutants: P260L (BOSE2286; BOSE2311), S264I (BOSE2540; BOSE3089), E312A (BOSE2121; BOSE2305), Y316D (BOSE2284; BOSE2309), P319S (BOSE2321; BOSE3083), A343V (BOSE2288; BOSE2313), E347G (BOSE2323; BOSE3085), P492Q (BOSE2120; BOSE2303), H493Y (BOSE2538; BOSE3087), and T617A (BOSE2123; BOSE2307). B. Missense mutations that suppress the spoIIIEΔγ phenotype do not affect the function of full-length SpoIIIE. Site-directed mutagenesis was used to introduce each indicated mutation into a spoIIIE allele that was then expressed under its native promoter at ycgO in ΔspoIIIE strains. Each mutant sporulated as well as cells with wild-type spoIIIE (bKM776). Seven mutants were tested: P260L (BOSE2298), E312A (BOSE2290), Y316D (BOSE2296), E347G (BOSE2325), P492Q (BOSE2294), D586N (BOSE3091), and T617A (BOSE2292). All eight sporulation efficiencies were not significantly different from each other by single factor ANOVA (p = 0.503).

Fig 6. Mutations that rescue the sporulation defect of spoIIIEΔγ strains do not rescue a vegetative translocation defect.

Growth of strains in various concentrations of the replication-stress-inducing antibiotic novobiocin (nov) was evaluated. Cells were grown to mid-exponential phase in LB, and then diluted to OD₆₀₀ 0.0125 in LB containing 960 ng ml^-1 nov (A), 480 ng ml^-1 (B), or no nov (C). OD600 measurements are plotted versus hours after dilution. All strains harbor ΔspoIIIE::neo. As indicated, samples were from strains with no ectopic spoIIIE (bDR1066), spoIIIE (bKM776), spoIIIEΔγ (BOSE2042), or a spoIIIEΔγ mutant: P260L (BOSE2286), E312A (BOSE2121), Y316D (BOSE2284), A343V (BOSE2288), E347G (BOSE2323), P492Q (BOSE2120), T617A (BOSE2123). Representative data from one of at least two replicates are plotted.