Identification of Bacillus subtilis SipW as a Bifunctional Signal Peptidase That Controls Surface-Adhered Biofilm Formation

Abstract

Biofilms of microbial cells encased in an exopolymeric matrix can form on solid surfaces, but how bacteria sense a solid surface and upregulate biofilm genes is largely unknown. We investigated the role of the Bacillus subtilis signal peptidase, SipW, which has a unique role in forming biofilms on a solid surface and is not required at an air-liquid interface. Surprisingly, we found that the signal peptidase activity of SipW was not required for solid-surface biofilms. Furthermore, a SipW mutant protein was constructed that lacks the ability to form a solid-surface biofilm but still retains signal peptidase activity. Through genetic and gene expression tests, the non-signal peptidase role of SipW was found to activate biofilm matrix genes specifically when cells were on a solid surface. These data provide the first evidence that a signal peptidase is bifunctional and that SipW has a regulatory role in addition to its role as a signal peptidase.

INTRODUCTION

Biofilms are a nonmotile community of microbial cells that are encased in an exopolymeric matrix. Biofilms are proposed to be the predominant lifestyle of microbes in nature and to contribute to medically significant infections and biofouling in industrial settings (9, 17, 24, 27). Biofilms can form on solid surfaces, at air-liquid interfaces (i.e., pellicle biofilms), or in liquid as cellular clusters (i.e., flocs) (20, 38). Solid-surface biofilms are of particular interest; they form on teeth, on urinary catheters and cardiac implants, on food industry equipment and drinking water distribution systems, and on plant roots (1, 15, 19, 21, 30, 36, 44). Surface-adhered biofilms are initiated when individual planktonic cells adhere to a surface. After adherence, the cells grow or move to form small clusters called microcolonies (26, 42, 46, 56). An exopolymeric structure, such as exopolysaccharide, is produced, allowing the cells of the microcolonies to adhere to each other (31, 39, 40, 61). The microcolonies continue to grow and produce the exopolymeric structure to make the matrix that surrounds the mature biofilms (8, 22, 39, 45). How bacteria sense that they are on a surface and then upregulate exopolymeric matrix genes is largely unknown.

Bacillus subtilis forms solid-surface biofilms and pellicle biofilms. Biofilm formation is controlled by a complicated regulatory network whose main function is to antagonize the SinR transcription repressor and thus derepress the operons involved in matrix production (11–14, 32). One of these matrix operons is the epsA-to-epsO (epsA-O) operon that encodes exopolysaccharide production enzymes (6, 7, 32, 41). The other operon is tapA-sipW-tasA that encodes the major protein component of the biofilm matrix, TasA, which forms amyloid fibers (5, 7, 47, 49). Of the other two proteins encoded by the operon, TapA is required for the assembly of and attachment of TasA fibers to the cell wall, and the type I signal peptidase SipW is required for processing and transport of TasA and TapA to the extracellular matrix (35, 48, 53, 55). In addition, SipW has a second role in forming biofilms beyond processing TasA and TapA. Mutants lacking TasA and TapA form surface-adhered biofilms that have slightly different structures than a wild-type strain, but mutants lacking SipW are only able to adhere to surfaces as single cells (29). Here, we address what this second function of SipW may be and whether it is to process an as-yet-unknown protein or is independent of its signal peptidase activity.

SipW is a member of the endoplasmic reticulum (ER) subfamily of type I signal peptidases (54, 55). Type I signal peptidases recognize N-terminal signal sequences of proteins that are being transported through the secretory machinery; they then cleave the protein just C terminal to the signal sequence so that it may be released from the membrane (43, 57). The ER-type subfamily of signal peptidases are membrane bound by an amino-terminal transmembrane anchor and share a highly conserved catalytic domain, which contains the active-site serine (43, 55, 58). Beyond these conserved features, the carboxy-terminal region of ER-type signal peptidases has a variable amino acid sequence and putative structure (18, 55, 58). The C-terminal region of SipW is proposed to contain an extracellular loop, a transmembrane domain, and a short cytoplasmic tail (55). It is not known what functions these carboxy-terminal regions have for these ER-type signal peptides.

In this study, we characterized the contribution of SipW to surface-adhered biofilm formation. We found that SipW is a bifunctional signal peptidase. Disruption of its active-site serine did not affect the ability of B. subtilis to form solid-surface biofilms, although it did prevent SipW from processing TasA. In contrast, deletion of the carboxy-terminal 20 amino acids of SipW was sufficient to render B. subtilis unable to form a solid-surface biofilm, while it was still able to process TasA. The novel non-signal peptidase role of SipW appears to be to upregulate the expression of the genes that are needed for the biosynthesis of the biofilm matrix after cells have adhered to a solid surface. These findings represent the first demonstration that a signal peptidase has a function that is independent of its signal peptidase activity.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The B. subtilis strains used in this study are listed in Table 1. Details of plasmid and strain construction and of growth conditions are provided in Experimental Procedures in the supplemental material.

Table 1.

Bacillus subtilis strains used in this study

Strain	Relevant genotypea	Source or reference
BAL218(JH642)	ttrpC2 pheA1	1
BAL722	ΔtapA::pBL105(cat)	2
BAL888	ΔsipW:: ΔtasA::spc	BAL982→BAL983
BAL982	ΔtasA::spc	2
BAL983	ΔsipW::	2
BAL988	ΔtapA:: ΔtasA::spc	2
Ssw2(BAL1049)	ΔsipW:: sinR2	This study
Ssw1(BAL1050)	ΔsipW:: sinR1	This study
Ssw3(BAL1051)	ΔsipW:: sinR3	This study
BAL1061	ΔsipW:: amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL983
BAL1062	ΔtapA:: ΔtasA::spc amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL988
BAL1343	ΔsipW:: ΔsinR::cat	IS432b→BAL983
BAL1459	ΔsipW::thrC::Φ(Pspachy-sipW)erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3423
BAL1460	ΔsipW::thrC::Φ[Pspachy-sipW(S47A)]-erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3424
BAL1461	ΔsipW::thrC::Φ[Pspachy-sipW(H87A)]-erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3425
BAL1462	ΔsipW::thrC::Φ[Pspachy-sipW(D106A)]-erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3426
BAL1924	ΔsipW::thrC::Φ[Pspachy-sipWΔ(120-190)]-erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3455
BAL1925	ΔsipW::thrC::Φ[Pspachy-sipWΔ(150-190)]-erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3456
BAL1926	ΔsipW::thrC::Φ[Pspachy-sipWΔ(170-190)]-erm amyE::Φ(Pspachy-gfpmut2)cat	BAL3165→BAL3457
BAL1927	ΔsinR::cat ΔtapA:: ΔtasA::spc amyE::Φ(Pspachy-gfpmut2)cat::erm	IS432 and BAL3171→BAL988
BAL2339	amyE::Φ(PtapA-gfp)cat	pBL585→BAL218
BAL2342	ΔsipW::amyE::Φ(PtapA-gfp)cat	BAL983→BAL2339
BAL2430	ΔsipW:: ΔsinR::cat ΔepsG::spc	pBL736→BAL1343
BAL2507	ΔsipW::thrC::erm	pBL112→BAL3193
BAL2739	ΔepsG::spc ΔsinR::cat	IS432→BAL3222
BAL3150	amyE::Φ(Peps-gfp)-cat	pBL741→BAL218
BAL3151	ΔsipW::amyE::Φ(Peps-gfp)-cat	BAL3150→BAL983
BAL3161	amyE::Φ(PtapA-lacZ)-cat	pBL589→BAL218
BAL3162	ΔsipW::amyE::Φ(PtapA-lacZ)-cat	pBL589→BAL983
BAL3165	amyE::Φ(Pspachy-gfpmut2)cat	pBL165→BAL218
BAL3171	amyE::Φ(Pspachy-gfpmut2)-cat::erm	pCm::Erb→BAL3165
BAL3172	thrC::Φ(Peps-lacZ)-erm
BAL3173	ΔsipW::thrC::Φ(Peps-lacZ)-erm	BAL3172→BAL983
BAL3193	ΔsipW::	AGS215b→BAL218
BAL3200	ΔsinR::cat ΔsipW:: amyE::Φ(Pspachy-gfpmut2)-cat::erm	BAL3171→BAL1343
BAL3222	ΔepsG::spc	pBL736→BAL218
BAL3404	ΔsinR::cat ΔsipW::thrC::sinR-erm	pBL749→BAL1343
BAL3405	ΔsipW:: sinR1 thrC::sinR-erm	pBL749→Ssw1
BAL3406	ΔsipW:: sinR2 thrC::sinR-erm	pBL749→Ssw2
BAL3407	ΔsipW:: sinR3 thrC::sinR-erm	pBL749→Ssw3
BAL3423	ΔsipW::thrC::Φ(Pspachy-sipW)erm	pBL760→BAL3193
BAL3424	ΔsipW::thrC::Φ[Pspachy-sipW(S47A)]-erm	pBL761→BAL3193
BAL3425	ΔsipW::thrC::Φ[Pspachy-sipW(H87A)]-erm	pBL762→BAL3193
BAL3426	ΔsipW::thrC::Φ[Pspachy-sipW(D106A)]-erm	pBL763→BAL3193
BAL3455	ΔsipW::thrC::Φ[Pspachy-sipWΔ(120-190)]-erm	pBL767→BAL3193
BAL3456	ΔsipW::thrC::Φ[Pspachy-sipWΔ(150-190)]-erm	pBL768→BAL3193
BAL3457	ΔsipW::thrC::Φ[Pspachy-sipWΔ(170-190)]-erm	pBL769→BAL3193
BAL3459	ΔtasA::spc epsA::pBL584Φ(Pspachy-epsA)-cat	pBL584→BAL982
BAL3460	ΔsipW:: ΔtasA::spc epsA::pBL584Φ(Pspachy-epsA)-cat	pBL584 and BAL2340→BAL982
BAL3465	ΔsipW:: ΔtasA::spc epsA::pBL584Φ(Pspachy-epsA)-cat amyE::Φ(Pspachy-gfpmut2)-cat::erm	BAL3171→BAL3460
BAL3466	ΔtasA::spc amyE::Φ(Pspachy-gfpmut2)-cat	BAL3165→BAL982

^aOnly relevant genotypes are given. The trpC2 and pheA1 mutations are present in all strains. The and markers differ with regard to the direction of transcription of the neo gene. The ΔsipW:: and ΔtapA:: alleles result in their downstream genes, tasA and sipW, respectively, being constitutively expressed from the neo promoter.

^bReferences are as follows: IS432, reference 3; pCm::Er, reference 4; and AGS215, reference 5.

Isolation of sipW suppressor mutant strains.

Three sipW suppressor mutant strains, named Ssw1, Ssw2, and Ssw3 for suppressor of sipW, were isolated. Strain BAL983 (ΔsipW::) was incubated in a microtiter plate for a total of 72 h. Cells that were adherent to the surface of the wells were recovered and tested for their ability to form biofilms after 24 h of incubation in a standard microtiter plate assay. See Experimental Procedures in the supplemental material for details. Three independently isolated strains that possessed the ability to form biofilms after 24 h of incubation were retained.

Mapping the location of the suppressor mutation.

The Ssw1 strain was chosen for mapping the location of the suppressor mutation. To map the location of the mutation, a method similar to that in Stanley et al. (51) was used, in which different strains, each with an antibiotic resistance marker associated with a particular locus on the chromosome and the ΔsipW mutation, were crossed with the Ssw1 mutant strain. The antibiotic-resistant transformants were screened by microtiter plate assay for a reversal to the ΔsipW mutant phenotype, which would indicate that a region near the location of the antibiotic resistance marker in the Ssw mutant contained the suppressor mutation. As the mapping indicated that the suppressor mutation mapped near the sinR locus, a PCR product containing the sinR coding region was amplified from the chromosomes of Ssw1, Ssw2, and Ssw3 using primers BL659 (5′-CAC AGT GGA ACG GCT TGA-3′) and BL300 (5′-CAG TGC AGC TTA CAG TTG A-3′). The PCR product was cloned into pCR2.1-TOPO (Invitrogen) and sequenced using the universal primer M13 forward (Davis Sequencing).

Microtiter plate assay.

The microtiter plate assay measures the amounts of cells adhering to the surfaces of microtiter plate wells by staining adhered cells with crystal violet. These assays were performed as previously described (28, 29, 51, 52) and are detailed in Experimental Procedures in the supplemental material.

CLSM.

Confocal laser scanning microscopy (CLSM) was utilized to examine the structures of biofilms formed by various strains. CLSM analysis was carried out as previously described (28, 29, 51) and is detailed in Experimental Procedures in the supplemental material.

Flat-bottom plate assay.

To microscopically analyze the progression of surface-adhered biofilm formation, cells containing a green fluorescent protein (GFP) transcriptional fusion were incubated in the wells of a tissue culture dish and then analyzed by epifluorescence microscopy. This is a modified version of the protocol used by Caiazza and O'Toole (10) and is detailed in Experimental Procedures in the supplemental material.

Growth of pellicle biofilms.

Bacterial cells were grown in beakers or the well of a tissue culture dish under stationary conditions. Mature pellicles were imaged after 48 h, and early cells transitioning to pellicles were imaged after 12 h. See Experimental Procedures in the supplemental material for details.

Western blot analysis.

The levels of TasA protein that were cell associated and released into the extracellular medium were analyzed as a measure of pre-TasA and mature TasA, respectively. Cells were grown under biofilm formation conditions, and then lysed cell and culture supernatant fractions were subjected to Western blot analysis with anti-TasA antibodies using standard protocols (4). See Experimental Procedures in the supplemental material for details.

RESULTS

SipW has a surface-specific function in biofilm formation.

SipW appears to have two roles in biofilm formation under some conditions; one is the processing of TasA and TapA, and the second is independent of TasA and TapA. The simplest model is that SipW processes an as-yet-unidentified protein. We and others have demonstrated that derivatives of the laboratory strain JH642 that lack SipW have a far greater defect in biofilm formation than strains lacking TasA and TapA when assayed for adherence to a microtiter plate or a glass slide (5, 29). In contrast, derivatives of wild-type B. subtilis strain 3610 that lack SipW, TasA, or TapA showed a similar defect in forming air-liquid interface biofilms (i.e., pellicles) (5, 7). One possible explanation for these different results is that this second function of SipW is specifically required for surface-adhered biofilm formation.

To test whether the second SipW function was surface specific, we used isogenic strains to compare the ability of SipW and TasA TapA mutant strains to form biofilms on two solid surfaces, one that was positively charged (i.e., poly-l-lysine-coated glass) and one that was noncharged (i.e., polystyrene), and for the ability to form biofilms at an air-liquid interface. We assessed biofilm formation on the surface of a poly-l-lysine-coated glass slide using confocal laser scanning microscopy (CLSM) (Fig. 1A). The ΔtapA ΔtasA mutant had a small defect in biofilm formation, with less depth and a flatter surface than the wild-type strain. However, the ΔsipW mutant attached to the surface only as single cells. Similar results were also observed on a polystyrene surface (Fig. 1B). The ΔtapA ΔtasA mutant cells were able to aggregate in large microcolonies; however, very few cells of the ΔsipW mutant strain were able to adhere. In contrast to these data, the ΔsipW and ΔtapA ΔtasA mutant strains formed indistinguishable, fragile pellicles (Fig. 1C). In addition, the cells of the ΔsipW and ΔtapA ΔtasA mutant strains did not show the typical cell-chain bundling that is characteristic of cells that form pellicle biofilms (Fig. 1D). Taken together, these data suggest that the second function of SipW is not required for biofilms that form at an air-liquid interface but is specifically required for surface-adhered biofilm formation.

Fig 1.

SipW has a surface-specific function in biofilm formation. (A) Biofilms grown on poly-l-lysine-coated glass for 24 h were viewed using confocal laser scanning microscopy (CLSM) and are at ×630 magnification. For each panel, the center image is a single section in the xy plane, the lower image is the xz plane, and the right image is the yz plane. Red lines indicate the locations of the xz, yz, and xy planes depicted. The scale bar is 50 μm. (B) Biofilms grown on polystyrene for 10 h were viewed using phase-contrast microscopy at ×200 magnification. (C) Biofilms were grown as pellicles in beakers and imaged after 48 h (no magnification). (D) Biofilms were grown as pellicles, and cells within the pellicles were imaged after 10 h (×400 magnification). All images are representative of at least 3 independent experiments. The wild-type strain was BAL3165, the ΔsipW strain was either BAL1061 (CLSM image) or BAL983, and the ΔtapA ΔtasA strain was either BAL1062 (CLSM image) or BAL988.

The signal peptidase activity of SipW is not required for its surface-specific function in biofilm formation.

A simple model for the second function of SipW is that it processes an as-yet-unidentified protein. To test this model, we asked whether the second, surface-specific function of SipW was dependent on its signal peptidase activity. The SipW signal peptidase activity requires a catalytic triad of Ser-47, His-87, and Asp-106 (55). Strains were created that expressed only a mutated allele of sipW, with one of its three catalytic residues replaced with an alanine (Fig. 2A). Surprisingly, each of these strains formed surface-adhered biofilms to the same extent as the isogenic wild-type strain (Fig. 2B; also see Table S1 in the supplemental material). Consistent with these mutant SipW proteins lacking signal peptidase activity, a strain expressing the SipW(S47A) allele did not form pellicles (Fig. 3), which depends on the ability of SipW to process TasA (5, 14). These data indicate that the surface-specific function of SipW is independent of its signal peptidase activity, suggesting that SipW is a bifunctional signal peptidase.

Fig 2.

Signal peptidase-defective mutants of SipW form surface-adhered biofilms, but truncated SipW mutants do not. (A) A linear representation of SipW and the relative locations of substitutions and truncations. The boxes labeled TM indicate the locations of putative transmembrane helices, and the box labeled SP is the location of the signal peptidase domain. Small arrows point to the locations of amino acid substitutions. Large arrows point to the endpoints of truncation mutants. (B) Biofilms formed by the mutants after 24 h of incubation were viewed using CLSM. The strains all express GFP and are as follows: wild-type (BAL3165), ΔsipW (BAL1061), ΔsipW P_spachy-sipW (BAL1459), sipW(S47A) (BAL1460), sipW(H87A) (BAL1461), sipW(D106A) (BAL1462), sipWΔ(120-190) (BAL1924), sipWΔ(150-190) (BAL1925), and sipWΔ(170-190) (BAL1926). Images are representative of at least three independent experiments. For each panel, the center image is a single section in the xy plane, the lower image is the xz plane, and the right image is the yz plane. Red lines indicate the particular locations of the xz, yz, and xy planes depicted, and the scale bar is 50 μm.

Fig 3.

The signal peptidase activity of SipW is required for pellicle formation. Biofilms were grown as pellicles in 25 ml of biofilm growth medium in 250-ml beakers and imaged after 48 h (no magnification). Images are representative of at least 3 independent experiments. Shown is a strain expressing wild-type SipW (ΔsipW P_spachy-sipW; BAL3423), a strain lacking sipW (ΔsipW; BAL3193), and a strain expressing the SipW(S47A) allele [ΔsipW P_spachy-sipW(S47A); BAL3424].

Identification of a region of SipW that is specifically required for surface-adhered biofilm formation.

To further verify that SipW is a bifunctional signal peptidase, we isolated mutant versions of sipW that retain signal peptidase activity but lack the surface-specific function of SipW. SipW is a member of the ER family of signal peptidases, which have a variable C-terminal region and a conserved N-terminal region with the signal peptidase active site (55, 58). We constructed three sipW mutant alleles that encode truncated SipW proteins lacking various amounts of the C-terminal region (Fig. 2A). Based on the predicted structure of SipW, the SipWΔ(120-190) protein would contain only the N-terminal membrane anchor and the signal peptidase domain, the SipWΔ(150-190) protein would contain that plus the remainder of the extracytoplasmic region, and the SipWΔ(170-190) protein would contain all but the 20-amino-acid C terminus, which is predicted to be cytoplasmic.

Strains expressing the truncated sipW were first assessed for their ability to form surface-adhered biofilms. Each of these mutant strains, with the sipWΔ(120-190), sipWΔ(150-190), or sipWΔ(170-190) mutation, showed a defect in biofilm formation similar to that of a ΔsipW mutant (Fig. 2B; also see Table S1 in the supplemental material). These data indicate that strains containing the truncated alleles of sipW are unable to form surface-adhered biofilms.

We next asked whether any of these truncated alleles of sipW encoded proteins that retained signal peptidase activity. As the ability to form pellicle biofilms is solely dependent on the signal peptidase activity of SipW, we asked whether any of the strains with the truncated alleles of sipW were able to form pellicles (Fig. 4A). Two strains, those containing the sipWΔ(120-190) or sipWΔ(150-190) allele, showed the same defect in pellicle formation as a ΔsipW mutant strain. However, the strain expressing the sipWΔ(170-190) allele formed a pellicle similar to that of the wild-type strain that covered the surface area and did not break apart or fall to the bottom of the beaker.

Fig 4.

The sipWΔ(170-190) mutant processes TasA. (A) Assessment of TasA processing through pellicle biofilm formation. Pellicle biofilms were grown in 25 ml of biofilm growth medium in 250-ml beakers and imaged after 48 h (no magnification). In the ΔsipW strain (BAL3193) background, the following alleles were expressed from the P_spachy promoter: wild-type sipW (ΔsipW P_spachy-sipW; BAL3423), sipWΔ(120-190) (BAL3455), sipWΔ(150-190) (BAL3456), and sipWΔ(170-190) (BAL3457). The pellicles formed by the mutants were compared with those produced by a wild-type (WT) strain (BAL218) and a ΔsipW (BAL3193) strain. Images are representative of at least 3 independent experiments. (B and C) TasA processing was also assessed through Western blot analysis using an anti-TasA antibody. (B) Samples are B. subtilis cell lysates. Note that the gel was slanted and there is pre-TasA present in ΔsipW and sipW(S47A) samples. (C) Samples are B. subtilis culture supernatants. The following strains were used: ΔsipW (BAL983), wild-type (BAL218), sipwΔ(170-190) (BAL3457), and sipW(S47A) (BAL3424).

That a strain expressing this sipWΔ(170-190) allele formed a pellicle suggested that this SipW protein retains signal peptidase activity. To further verify this, we assayed the sipWΔ(170-190) strain for the ability to process TasA. Cellular accumulation of unprocessed TasA (i.e., pre-TasA) was observed in strains that lack SipW or express a catalytically inactive version of SipW, SipW(S47A) (Fig. 4B). In strains expressing either wild-type SipW or SipWΔ(170-190), mature TasA accumulated extracellularly due to processing and release to the extracellular milieu (Fig. 4C). These data indicate that the sipWΔ(170-190) allele retains signal peptidase activity and appears to be defective specifically for the surface adherence function of SipW. That it is possible to genetically separate the signal peptidase activity and surface-specific function of SipW supports the model of SipW as a bifunctional protein.

Isolation of sipW suppressor mutants.

To understand the surface-specific function of SipW, we isolated spontaneous suppressor mutants that restored surface-adhered biofilm formation to the ΔsipW mutant. To isolate these suppressors, we took advantage of the observation that a ΔsipW mutant strain would gain the ability to form a surface-adhered biofilm if the cells were propagated in a microtiter plate for more than 72 h. To propagate these biofilms, every 24 h, all non-surface-adhered cells were removed, including any pellicle biofilm cells, and fresh medium was added. This method has the possibility of enriching for surface-adhered cells in a population. Consistent with what was previously observed (29), after 24 h in this microtiter plate assay, the wild-type strain had formed approximately 8-fold larger amounts of surface-adhered biofilm than the ΔsipW mutant strain (Fig. 5A). After 48 h of incubation, the wild-type strain still formed larger amounts of surface-adhered biofilm than the ΔsipW mutant strain. However, after 72 h of incubation, only a 1.5-fold difference in the amounts of biofilm was observed between the wild-type and ΔsipW mutant strains.

Fig 5.

Isolation of ΔsipW suppressor mutant strains that form biofilms and have point mutations in sinR. (A) The amount of biofilm formed by the ΔsipW mutant strain increased after 60 h of incubation. The amounts of biofilm formed by the wild-type (WT; BAL218) and ΔsipW mutant (BAL983) strains were measured over 72 h using a microtiter plate assay. Shown are the means and standard errors of the means of the A₅₇₀ of the solubilized crystal violet from the wells of the microtiter plate from one representative assay. (B) ΔsipW mutant cells that adhered to the microtiter plate wells after 72 h were able to form biofilm at a level similar to that of a wild-type strain after only 24 h of incubation. Shown are the results for the wild-type strain (WT; JH642), the ΔsipW mutant (BAL983), and a representative sipW suppressor mutant strain, Ssw1 (BAL1050). The average A₅₇₀ value obtained for each strain was normalized to the A₅₇₀ obtained for the wild-type strain. The error bars represent the standard errors of the means. (C) The sipW suppressor mutant strains Ssw1, Ssw2, and Ssw3 contained a mutation in sinR. The sequence of the sinR coding region is shown along with the corresponding amino acid sequence. The numbers represent either the amino acid (upper line) or nucleotide (lower line) position relative to the start codon. The gray boxed region indicates the location of the A inserted in a run of As in the Ssw strains. The alteration of the amino acid sequence due to the presence of the point mutation is indicated underneath the wild-type sinR sequence. The asterisks represent stop codons.

The ability of the ΔsipW mutant strain to form surface-adhered biofilms after 72 h of incubation could be caused either by a delay in the ability to form biofilms or by the acquisition of a suppressor mutation. To distinguish between these two possibilities, the ΔsipW mutant cells that were adhered to the surface of microtiter plate wells after 72 h of incubation were clonally purified. These clones were tested for their ability to form biofilms in the microtiter plate assay after 24 h of incubation. All of the colonies tested formed large amounts of biofilms, more comparable to the amounts observed for the wild-type strain than the amounts observed for the parental ΔsipW strain (Fig. 5B and data not shown). These data suggest that the ΔsipW mutant had acquired a suppressor mutation that allowed biofilm formation to occur.

The sipW suppressor mutants contain a frameshift mutation in sinR.

We sought to map the locus that led to enhancement of biofilm formation in the ΔsipW suppressor mutants. To achieve this, we characterized one ΔsipW suppressor mutant strain, named Ssw1, for suppressor of sipW mutant 1. Ssw1 was transformed with chromosomal DNA from different strains, which each had an antibiotic resistance cassette at a different location on the chromosome, as well as a deletion of sipW. Transformants were then selected for acquisition of the antibiotic resistance cassette, and at least 16 individual colonies from the transformants were tested for their ability to form biofilms in the microtiter plate assay after 24 h of incubation (data not shown). The percentage of transformants having a biofilm-negative phenotype indicated the linkage of the antibiotic resistance cassette to the suppressor mutation. Using this approach, we observed approximately 85% linkage of the suppressor mutation in the Ssw1 mutant strain to either the ΔtasA::neo or the ΔtapA::neo allele. Note that the ΔtasA::neo and ΔtapA::neo alleles themselves do not result in a defect in biofilm formation as measured by the microtiter plate assay (7, 29). These data indicate that the mutation suppressing the biofilm defect of a ΔsipW mutation was located in or near the tapA-sipW-tasA operon.

The sinR gene is located immediately adjacent to the tapA operon on the chromosome. As mutations in sinR had previously been shown to be able to suppress the defect in pellicle formation of some B. subtilis mutants (32), we hypothesized that the Ssw strains may contain a mutation in sinR. The sinR coding region was PCR amplified and sequenced from 3 independently isolated Ssw strains (i.e., isolated from independent 72-h microtiter plate enrichments). Each Ssw strain contained a single adenine insertion into a homopolymeric track of 6 adenine residues at the 3-prime end of sinR that would result in a frameshift and a truncated SinR protein being produced (Fig. 5C). This point mutation was similar to that described in Kearns et al. (32), with the exception that an insertion rather than a deletion of an adenine residue was observed.

To confirm that the increase in biofilm formation seen in the Ssw1 mutant was due to the frameshift mutation leading to a loss of SinR function, we introduced a sinR deletion mutation into a ΔsipW mutant strain. To avoid the complication of the second function of SipW in processing TasA, a ΔtasA mutation was introduced into both strains. The ΔsinR ΔsipW ΔtasA mutant strain formed biofilms to the same extent as the ΔsinR ΔtasA strain as determined by microtiter plate assay and CLSM (Fig. 6; also see Fig. S1 in the supplemental material). These data suggest that loss of sinR function is able to suppress the ΔsipW biofilm defect. The introduction of a wild-type copy of sinR at the heterologous thrC locus to the Ssw1 mutant or the ΔsinR ΔsipW mutant restored biofilm formation to the level of a ΔsipW mutant (see Table S2 in the supplemental material). Altogether, these data indicate that loss of sinR suppresses the surface-adhered biofilm defect caused by the ΔsipW mutation.

Fig 6.

Deletion of sinR and derepression of eps bypass the requirement of SipW for surface-adhered biofilm formation. Biofilms were grown with 20 μM IPTG to induce P_spachy-eps expression, and morphology was assessed by CLSM after 48 h of incubation. The ΔtasA (BAL3466) (A), ΔsipW ΔtasA (BAL1061) (B), ΔsinR ΔtasA (BAL1927) (C), ΔsinR ΔsipW ΔtasA (BAL3200) (D), P_spachy-eps ΔtasA (BAL3464) (E), and P_spachy-eps ΔtasA ΔsipW (BAL3465) (F) strains were compared. Images are representative of at least three independent experiments and are at ×630 magnification. For each panel, the large, center image is a single section in the xy plane, the lower image is the xz plane, and the right image is the yz plane. Red lines indicate the particular locations of the xz, yz, and xy planes depicted.

Suppression of the sipW biofilm defect by a sinR mutation requires the eps operon.

We next considered how loss of SinR function might suppress the biofilm defect of a ΔsipW mutant. Since SinR is a transcriptional repressor of the EPS biosynthesis genes (epsA-O) (32) and the tapA-sipW-tasA operon (5) and TasA and TapA are not essential for surface-adhered biofilm formation in the assays used here, we hypothesized that ΔsinR suppression may be achieved through derepression of the eps operon. Consistent with this, deleting a gene required for EPS production, epsG, prevented surface-adhered biofilm formation even in the ΔsipW suppressor strains (see Fig. S2 in the supplemental material; also data not shown).

If derepression of eps expression through inactivation of sinR suppressed the biofilm defect of the ΔsipW mutant, then the expression of eps from a heterologous promoter should result in the same suppression phenotype. To test this, we constructed a strain in which the isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible promoter, P_spachy, replaced the epsA promoter at the endogenous locus (P_spachy-eps). When grown in the presence of 20 μM IPTG, the P_spachy-eps strain formed biofilms similar to those of the wild-type strain (data not shown). We then tested the biofilm phenotype of P_spachy-eps and P_spachy-eps ΔsipW mutant strains, which also contained a ΔtasA mutation to allow us to specifically examine suppression of the surface-specific function of SipW. The P_spachy-eps ΔsipW strain formed biofilms that were similar to those of the P_spachy-eps strain as determined by CLSM and microtiter plate assay (Fig. 6; also see Fig. S1 in the supplemental material). The biofilm formed by the P_spachy-eps ΔsipW mutant was also similar to the surface-adhered biofilms formed by the wild-type strain and the ΔsinR ΔsipW strain (Fig. 6). These data indicate that heterologous expression of eps can suppress the ΔsipW biofilm defect. This further indicates that the ability of ΔsinR to suppress the surface-specific function of SipW is a consequence of derepression of the eps genes.

SipW affects the expression of the biofilm matrix components.

The above-described data indicate that derepression of eps (ΔsinR or P_spachy-eps) is epistatic to the surface-specific biofilm function of SipW. Since SinR is a transcriptional repressor of the eps and tapA operons, these results suggest that SipW may be upstream from and required for the expression of these biofilm matrix operons during surface-adhered biofilm formation. To test this hypothesis, we analyzed eps and tapA expression in wild-type and ΔsipW mutant strains under surface-adhered biofilm formation conditions. To assess the expression of the eps and tapA operons, we used P_eps-gfp and P_tapA-gfp reporter fusions, where the eps promoter (P_eps) or the tapA promoter (P_tapA) was fused to the gene for green fluorescent protein (GFP). Strains containing these fusions were inoculated into the wells of a 24-well plate with a poly-l-lysine-coated glass coverslip placed in the bottoms of the wells. The plates were incubated at an angle such that the air-liquid interface crossed the bottom of the well. At various time points, the coverslips were removed to image the surface-adhered cells by phase-contrast microscopy and eps and tapA expression by epifluorescence microscopy (Fig. 7). At the earliest time point, 2.5 h, only a monolayer of single cells was observed and eps and tapA were not expressed. By 4.5 h of incubation, small clusters of cells were observed for the wild-type strain, and individual cells within these clusters were observed to express eps and tapA. The expression of eps and tapA continued to increase in the wild-type strain at later time points as the cell cluster increased in size. In contrast, the ΔsipW mutant strain remained adhered as a monolayer of cells, and significant levels of eps and tapA expression were not detected at any time point. Therefore, we conclude that expression from the eps and tapA promoters is activated in cells that have formed cell-cell interactions to create microcolonies and that sipW is required for this expression. These data suggested that SipW is required for upregulation of the expression of biofilm matrix components. Consistent with this regulatory role for SipW only being required for surface-adhered biofilm formation, the expression of eps and tapA was not observed to be affected by a ΔsipW mutation during planktonic growth or during pellicle formation (see Fig. S3 in the supplemental material; also data not shown). These data together indicate that SipW is required for biofilm matrix gene expression specifically during surface-adhered biofilm formation.

Fig 7.

SipW affects tapA and eps expression. Cells were grown in biofilm growth medium in the presence of poly-l-lysine-coated glass slides and were viewed after 2.5, 4.5, and 7 h of incubation by phase-contrast (Phase) and epifluorescence (GFP) microscopy. (A, B) Expression of tapA was monitored using a P_tapA-gfp reporter fusion and was compared in wild-type (BAL2339) and ΔsipW (BAL2342) backgrounds. (C, D) Expression of eps was monitored using a P_eps-gfp reporter fusion and was compared in wild-type (BAL3150) and ΔsipW (BAL3151) backgrounds. Images are representative of at least 4 independent experiments and are at ×400 magnification. Any differences observed in numbers of cells adhered simply reflect day-to-day variability in this assay.

DISCUSSION

Signal peptidases are ubiquitous enzymes critical for the proper processing of proteins being transported across a membrane. Here, we present data that indicate that the SipW signal peptidase of B. subtilis is bifunctional, with a second function that is unrelated to the ability of SipW to act as a signal peptidase (Fig. 8). This second function appears to be regulatory, as SipW was found to be required for the expression of the biofilm matrix tapA and eps operons when cells were forming a biofilm on a solid surface. Consistent with SipW being a bifunctional enzyme, the signal peptidase and regulatory function of SipW were found to be genetically separable. To the best of our knowledge, this is the first demonstration of a signal peptidase being bifunctional.

Fig 8.

Model of the two separable functions of SipW. The putative structural model for SipW is shown, with the gray bars indicating the transmembrane domains. The dashed black line divides SipW into the regions that code for the signal peptidase (SP) and regulatory functions. The precise demarcation between these two functions is not known. The signal peptidase activity requires the signal peptidase domain (black circle) and is shown processing (white arrow) TasA (red). The regulatory function requires the C-terminal cytoplasmic tail of SipW at a minimum. Through an unknown mechanism, the C-terminal tail antagonizes the activity of the transcriptional repressor SinR, which leads to induction of the biofilm matrix genes. “In” and “Out” refer to the cytoplasmic and extracytoplasmic areas of the cell.

The bifunctionality of SipW.

We had previously shown that SipW has a function in biofilm formation that is independent of its requirement to process TasA and TapA (29). The simple model for this other function would be that SipW processes an unknown protein that contributes to biofilm formation, but this other function of SipW did not require signal peptidase activity. Furthermore, we isolated a SipW mutant allele, SipWΔ(170-190), that lacks the ability to support surface-adhered biofilm formation but retains the ability to act as a signal peptidase. This is, to our knowledge, the first evidence that a signal peptidase can be bifunctional with a function that is independent of its protease activity.

The signal peptidase-independent, regulatory function of SipW requires its C-terminal tail. The C-terminal 20 amino acids are predicted to be cytoplasmic, and a simple prediction is that this cytoplasmic tail interacts with a protein to stimulate biofilm formation. However, it is conceivable that a greater portion of SipW beyond the C-terminal 20 residues may be necessary for the regulatory function. The other two mutant alleles that removed larger portions of SipW (amino acids 120 to 190 and 150 to 190) produced nonfunctional proteins. Thus, it is possible that some portion of SipW's C-terminal transmembrane or periplasmic domain is also involved in the regulatory function of SipW.

The regulatory function of SipW.

The non-signal peptidase function of SipW in biofilm formation appears to be to induce the expression of the biofilm matrix tapA and eps operons. Removing the normal regulatory control from the eps operon, either through loss of SinR or through replacement of the eps promoter, resulted in bypass of the requirement of SipW. Furthermore, the expression of P_tapA-gfp and P_epsA-gfp fusions was dependent on SipW during surface-adhered biofilm formation. These data indicate that the non-signal peptidase function of SipW is to stimulate the expression of biofilm matrix gene expression during biofilm formation on a solid surface.

How SipW affects the expression of biofilm matrix genes is unknown. That the regulatory function of SipW can be suppressed by the loss of SinR suggests that SipW functions to antagonize SinR activity, leading to derepression of the biofilm matrix operons. SinR activity can be antagonized directly via the SinI, SlrA, and SlrR proteins and indirectly via YwcC or the Spo0A phosphorelay network (2, 11–13, 23, 34, 37). SipW could work through any of these proteins to affect SinR or through an as-yet-unidentified protein, and we are actively investigating the mechanism by which SipW controls matrix gene expression.

The regulatory function of SipW is specific to surface-adhered biofilms.

Several lines of evidence support the surface specificity of the regulatory function of SipW. First, deletion of sipW does not result in an obvious TasA-/TapA-independent phenotype during air-liquid interface biofilms (aka pellicles) as it does in surface-adhered biofilms. Second, substitution of the signal peptidase active-site residues severely disrupted pellicle formation while having no significant effect on surface-adhered biofilms. Third, deletion of the C-terminal tail of SipW abolished surface-adhered biofilm formation and the regulatory function of SipW but supported normal pellicle formation. These data collectively suggest that the regulatory function of SipW is specifically required for biofilms to form on a solid surface. Consistent with this, we did not observe a role for SipW in stimulating matrix gene expression in liquid cultures. It is not unexpected that some of the components that control the expression of the biofilm matrix would differ between surface-liquid and air-liquid interface biofilms. As such, these components would most likely have a role in perceiving or transducing the signals that tell a cell to form a biofilm in these different environments, and SipW is such a candidate component.

SipW, endoplasmic reticulum signal peptidases, and other bifunctional proteins.

SipW belongs to the endoplasmic reticulum (ER) family of signal peptidases. Members of this family posses different C termini, and a function for these varied C termini has not been described (55). The research presented here with SipW is the first to demonstrate a function for the C-terminal region of ER signal peptidases. That the function associated with the C-terminal region of SipW was regulatory and not dependent on its signal peptidase activity suggests that any of a range of possible functions could be associated with the other ER signal peptidases and that hypotheses for them do not need to be limited to the substrate specificity category.

SipW joins a growing list of bifunctional enzymes whose enzymatic activities are distinct from their regulatory activities. Interestingly, one of these other bifunctional proteins, EpsE, also affects biofilm formation by B. subtilis. EpsE coordinates the production of the matrix with inhibition of flagellar motility (3, 25). However, SipW more clearly groups with a subgroup of bifunctional proteins that have an enzymatic activity and a regulatory function that controls gene expression (16). Perhaps the greatest parallels can be drawn between the GmaR protein of Listeria monocytogenes and SipW. GmaR is a glycosylase of flagellin and a regulator that stimulates the transcription of the gene for flagellin (50). Similarly, SipW processes the matrix protein TasA and stimulates transcription of the operon encoding TasA. GmaR acts by antagonizing the activity of the transcriptional repressor through direct interaction (50). SipW also appears to antagonize the activity of a transcriptional repressor, SinR. We do not know if this is due to a direct interaction, and unlike GmaR, which is a cytoplasmic protein, SipW is a membrane protein. Could SipW act more like the membrane protein DcuB in modulating the activity of a two-component regulatory system? DcuB is an antiporter for fumarate and succinate and a regulator of genes for fumarate respiration, and transport activity is not required for DcuB to regulate gene expression (33). DcuB does not appear to regulate gene expression directly but instead modulates the activity of the DcuSR two-component regulatory system (33). An expanded two-component regulatory system, the Spo0A phosphorelay, stimulates biofilm formation by B. subtilis (6, 11, 23, 28, 59, 60). Thus, studying how SipW controls gene expression will expand our knowledge of how bifunctional proteins can act to control gene expression.

This study has identified the first bifunctional signal peptidase having genetically separable and independent signal peptidase and non-signal peptidase functions. This localizes in one protein the functions required for the production of the biofilm matrix proteins and the expression of the genes for those same matrix proteins. The role of SipW in controlling matrix gene expression only occurs during biofilm formation on a solid surface, and gaining a better understanding of the mechanism by which SipW stimulates matrix gene expression may yield greater insight into the mechanisms that bacteria utilize to sense and respond to cell-surface attachment. This will also augment our understanding of the different functions that signal peptidases may perform.