All experiments were performed in accordance with the “Regulations on biosafety of Pathogenic Microorganism Laboratory” (2004) No. 424 set by the State Council of the People’s Republic of China and approved by Biosafety Committee of Northwest A&F University.

The bacterial strains used in this study were derived from the B. suis vaccine strain S2 (Di et al., 2016). The strains were cultured in tryptic soy broth (TSB) with shaking or on tryptic soy agar (TSA) at 37°C. To determine the concentrations of viable B. suis S2 strains, single bacterial colonies from freshly streaked TSA plates were inoculated into 50 ml TSB and grown at 37°C for 2 days with gentle shaking. Subsequently, the cultures were diluted 100-fold with TSB and incubated at 37°C to exponential phase. The cultures were harvested and resuspended in sterile phosphate-buffered saline (PBS), then serially diluted in sterile PBS and plated onto TSA plates to confirm the concentrations of viable bacteria. Antibiotics, when required, were added at the following concentrations: kanamycin, 50 μg/ml; ampicillin, 100 μg/ml. All work involving live B. suis S2 strains was performed in a biosafety level 3 (BSL-3) facility.

To obtain the brbI deletion mutant strain, a resistance gene replacement procedure was carried out. Briefly, the upstream homologous arm of brbI was amplified using the UP-F and UP-R primers, while the downstream homologous arm of brbI was amplified using the DW-F and DW-R primers. A 1375-bp kanamycin-resistance cassette (KanR) fragment was amplified from the pEGFP-C1 plasmid using the KanR-F and KanR-R primers. The KanR fragment and the two homologous arms were then inserted into the pUC18 vector to generate the recombinant pUC18-brbI-KanR plasmid which was transfected into electrocompetent B. suis S2 cells by electroporation (McQuiston et al., 1995). Transformants were screened in the presence of 50 μg/ml kanamycin. The mutant strain was further confirmed via polymerase chain reaction (PCR) using primers BrBI-TF and BrBI-TR and via quantitative real time polymerase chain reaction (qRT-PCR) using primers BrBI-qF and BrBI-qR.

To obtain the complemented strain, an expression plasmid containing a FLAG-tag (pBBARpc) was constructed (Supplementary Material). The intact brbI open reading frame (ORF) was amplified using the BrBI-F/BrBI-R primer pair, and the PCR product was cloned into pBBARpc to generate the recombinant pBBARpc-brbI plasmid. Electrocompetent cells of the mutant strain were prepared and electroporated according to standard procedures. Subsequently, transformants were screened in the presence of 100 μg/ml ampicillin. The complemented strain was further confirmed via western blot (WB) using an anti-FLAG antibody (abs830005, Absin, Shanghai, China) and via qRT-PCR using the primers, BrBI-qF and BrBI-qR. The sequences of all the primers used in this study are listed in Table 1.

The three B. suis S2 strains were inoculated at equal densities (1 × 10⁷ colony-forming units [CFUs]) into 10 ml of TSB and incubated at 37°C with shaking. The cultures were collected at specific times (0, 8, 16, 24, 32, 40, 48, 56, 64, 72, and 80 h), and the optical density (OD) at 600 nm was measured in a 96-well plate using a microplate reader (Bio-Rad, CA, United States). The three B. suis S2 strains were streaked on antibiotic-free TSA plates and incubated at 37°C. After 3 days, the colony sizes of the three B. suis S2 strains were measured.

The bacterial aggregation assay was performed as previously reported, with some modifications (Liu et al., 2017). Briefly, the three B. suis S2 strains were cultured to the exponential phase as described above, then the cultures were left standing on test-tube racks at room temperature for 24 h. To quantify the bacterial aggregation, the OD_{600 nm} of 100 μl of the culture obtained before and after standing was measured in a 96-well microplate using a microplate reader (Bio-Rad). Percent bacterial aggregation was calculated using the following formula:

The three B. suis S2 strains were cultured to the exponential phase as described above, then inocula of the three strains were fixed in 2.5% glutaraldehyde solution for 3 h at room temperature. After washing three times with PBS, the bacterial samples were dehydrated in an ascending ethanol gradient (30, 50, 70, 80, 90, 95, and 100% for 15 min each). After critical point drying and gold coating, bacterial cells were observed and imaged using a scanning electron microscope (Regulus8100, Hitachi, Tokyo, Japan).

The three B. suis S2 strains were prepared as described above and adjusted to the same OD_{600 nm} with PBS. The concentrations of viable bacteria from the three B. suis S2 strains that had the same OD_{600 nm} were measured by plating onto TSA plates as described above. Propidium iodide (PI) staining was also performed. 2 × 10⁷ CFUs of each bacterial suspension was diluted into 100 μl with PBS and pipetted into separate wells of a 96-well, flat-bottom microplate, then incubated with isovolumetric PI stain solution (20 μM) at room temperature in the dark for 15 min. The fluorescence intensity was then measured (excitation wavelength: 535 nm; emission wavelength: 615 nm) using a fluorescence microplate reader (Spark 10M, TECAN, Männedorf, Switzerland).

The three B. suis S2 strains were cultured and counted as described above. For the sodium dodecyl sulfate (SDS) sensitivity assay, 10 μl of 10-fold diluted samples from the three B. suis S2 strains (10⁴–10⁶ CFUs/ml) were added to a TSA plate containing 0.00325% SDS, then the plate was incubated for 3 days at 37°C.

For the antibiotic-resistance assay, bacterial samples were seeded into a 96-well microplate (2 × 10⁴ CFUs/well) and treated with serially diluted antibiotics (polymyxin B: final concentrations of 1200, 600, 300, 150, 75, 37.5, 18.75, 9.325, and 0 μg/ml; lincomycin: final concentrations of 500, 375, 250, 125, 62.5, 31.25, 15.6, 7.8, 3.9, 1.95, and 0 μg/ml) at 37°C for 1h. The bacterial samples were then diluted 10-fold and plated onto TSA plates, then the plates were incubated for 3 days at 37 °C. The survival rate of each strain was calculated by dividing the antibiotic-treated bacterial CFUs by the untreated bacterial CFUs.

For the environmental stress tolerance assay, samples from the three B. suis S2 strains were inoculated into TSB containing H₂O₂ (1.25 and 2.5 mM) for 1h and HCl (pH 3.5 and 4.5) for 30 min at 37°C. After treatment, a sample of each strain was diluted serially 10-fold and plated onto TSA plates. The plates were then incubated at 37°C for 3 days and the survival rate of each strain was calculated by dividing the treated bacterial CFUs by the untreated bacterial CFUs.

The B. suis S2 strains were grown in TSB to the exponential phase, then collected via centrifugation at 4°C. Total RNA was then isolated using TRIzol reagent (Invitrogen, Inc., Carlsbad, CA, United States). The resulting RNA samples were reverse-transcribed into cDNA using HiScript III RT SuperMix (Vazyme, Nanjing, China) per the manufacturer’s recommended protocols. qRT-PCR was then performed using the ChamQ SYBR qPCR Master Mix (Vazyme, Nanjing, China) and Bio-Rad CFX96 Real-Time PCR Systems (Bio-Rad, United States). The 2^–ΔΔCt method was used to analyze the relative transcription levels. The results for each target mRNA were normalized to 16S rRNA transcript levels and averaged.

The amino acid sequences were aligned using Clustal Omega Webservers⁵ (Madeira et al., 2019). The protein transmembrane helices were predicted using TMHMM Server v. 2.0⁶. The BrBI three-dimensional (3D) structure was predicted using the I-TASSER Webserver⁷ (Roy et al., 2010; Yang and Zhang, 2015).

All experiments were independently repeated at least three times, and the data are expressed as the mean ± standard deviation (SD) of representative experiments performed in triplicate. All data were analyzed with SPSS (SPSS, Inc., Cary, NC, United States). One-way analysis of variance (ANOVA) with Tukey’s post-test was used to analyze differences between groups. A p-value < 0.05 was considered statistically significant.

As human BI-1 is the first and best characterized eukaryotic BI-1, and E. coli YccA and B. subtilis YetJ are relatively well-characterized prokaryotic homologs of eukaryotic BI-1, we compared BrBI with these three BI-1 family proteins using bioinformatic methods. BrBI shares 21.09% identity and 47.64% similarity to human BI-1, 21.12% identity and 59.36% similarity to E. coli YccA, and 21.43% identity and 47.01% similarity to B. subtilis YetJ in amino acid sequences (Figure 1A). Additionally, BrBI contains seven predictive transmembrane helices with a 100% probability, which is concordant with human BI-1, E. coli YccA, and B. subtilis YetJ (Figure 1B). As human BI-1, E. coli YccA, and B. subtilis YetJ have been identified as transmembrane proteins, we deduced that BrBI is also a transmembrane protein, and BI-1 family proteins may be structurally conserved. We further predicted the 3D structural model of BrBI. The predicted 3D structure of BrBI contained seven transmembrane segments (Figure 1C) and is homologous to B. subtilis YetJ according to the Protein Data Bank library, further revealing that BrBI is a transmembrane protein homologous to other BI-1 family proteins.

To explore the functions of BrBI, we generated a B. suis S2 brbI deletion mutant strain (ΔbrbI) via homologous recombination, replacing the intact brbI ORF with the kanamycin-resistance gene used as a selection marker (Figure 2A). To further confirm successful generation of the ΔbrbI strain, we designed a primer pair BrBI-TF/TR residing within the brbI ORF (Figure 2A). We randomly selected four kanamycin-resistant transformants for PCR verification and identified a transformant lacking brbI (Figure 2B). This brbI mutant strain was further confirmed by qRT-PCR (Supplementary Figure S2A). Additionally, qRT-PCR indicated that the deletion of brbI did not affect the downstream gene expression (Supplementary Figure S2B). The complemented strain (ΔbrbI::brbI) was generated via exogenous expression of the intact brbI ORF in the ΔbrbI strain using an expression vector harboring a FLAG-tag. ΔbrbI::brbI strain generation was confirmed via WB using a mouse anti-FLAG antibody (Figure 2C) and qRT-PCR (Supplementary Figure S2A).

Because brbI deletion suppressed B. suis S2 growth, and because of the conserved cytoprotective capacity of eukaryotic BI-1 proteins and the bacterial cytoprotective potential of B. subtilis YetJ and E. coli YccA, we hypothesized that brbI deletion might lead to increased cell death in B. suis S2. To test this, we investigated how deleting brbI would affect the viability of B. suis S2. When adjusted to the same OD_{600 nm} (representing the same bacterial cells), the ΔbrbI strain showed significantly reduced viable bacterial concentrations (represented as CFUs/ml) compared with those of the B. suis S2 and ΔbrbI::brbI strains (reduced ∼4.8- and 3.3-fold, respectively, data not shown), indicating that brbI deletion suppressed the survival of B. suis S2 (Figure 3C). We performed PI staining to further confirm this. PI is a red fluorescent nucleic acid stain that cannot penetrate intact biological membranes; therefore, the fluorescence intensity of bacteria stained with PI can indicate the cell membrane integrity status. If PI penetrates the membrane, the cell is usually assumed to be dead. When bacterial samples containing 2 × 10⁷ CFUs were stained with PI, the ΔbrbI strain displayed considerably stronger fluorescence intensity than did either the B. suis S2 strain or the complemented strain (increased ∼3.7- and 2.5-fold, respectively, data not shown), indicating a greater proportion of dead bacteria in the ΔbrbI strain samples (Figure 3D). These results revealed that BrBI is crucial for B. suis S2 viability, suggesting that BrBI is a bacterially cytoprotective protein.

In addition to growth suppression and viability defects, another noticeable phenotype was observed in the ΔbrbI strain. After standing at room temperature for 24 h, the B. suis S2 strain culture harvested at the exponential phase remained uniform, whereas conspicuous bacterial aggregation was observed in the ΔbrbI strain (Figure 3E). Moreover, brbI complementation rescued the bacterial aggregation phenotype. We speculated that brbI deletion may have affected the membrane properties of B. suis S2. We then performed an SDS-sensitivity assay to determine the membrane properties of the ΔbrbI strain. As expected, brbI deletion significantly increased the sensitivity of B. suis S2 to SDS (Figure 3F), suggesting that BrBI may be important for the membrane properties in B. suis S2.

Scanning electron microscopy analysis was conducted to assess the effect of brbI deletion on the B. suis S2 morphology (Figure 4). The B. suis S2 strain displayed a typical coccobacillus morphology. Conversely, the cell morphology of the ΔbrbI strain was aberrant: some of the mutant strain showed Y-shaped or branched morphology, while some were intumesced mid-cell, indicating that brbI deletion impaired B. suis S2 cell division. Complementation of the brbI relieved the morphological irregularities. Together, BrBI may involve in the process of cell division in B. suis S2, which may contribute to the defective growth and viability in the ΔbrbI strain.

In bacteria, the cell envelope is the first layer to sense environmental stresses. Given the altered membrane properties of the ΔbrbI strain, we assessed whether brbI deletion would affect the sensitivity of B. suis S2 to environmental stresses. When exposed to an acidic environment, the survival rate was significantly lower in the ΔbrbI strain than in the B. suis S2 and ΔbrbI::brbI strains (Figure 5A), indicating that BrBI is crucial for B. suis S2 to tolerate acidic stress. When stimulated with H₂O₂, cell viability was significantly impaired in the ΔbrbI strain (Figure 5B), implying that BrBI also promotes B. suis S2 tolerance to oxidative stress.

Regarding antibiotic resistance, the bacterial cell envelope is both a barrier to antibiotic penetration and the target of several antibiotics. Hence, we evaluated the impact of brbI deletion on the resistance of B. suis S2 to polymyxin B and lincomycin. Polymyxin B is a peptide antibiotic that strongly interacts with bacterial membranes and is frequently used to assess the membrane characteristics of Brucella. Exposure to polymyxin B concentrations of up to 75 μg/ml significantly decreased the survival rate of the ΔbrbI strain compared with that of the complemented and B. suis S2 strains, indicating increased sensitivity of the ΔbrbI strain to polymyxin B (Figure 5C). Lincomycin is a lincosamide antibiotic that inhibits bacterial protein synthesis by interacting with the ribosomal 50S subunits. When pre-treated with different lincomycin concentrations for 30 min at 37°C, the B. suis S2 and ΔbrbI::brbI strains displayed similar resistances to lincomycin; however, resistance of the ΔbrbI strain to lincomycin decreased significantly at 62.5 μg/ml of lincomycin (Figure 5D). Further, brbI deletion significantly reduced the MICs of polymyxin B and lincomycin against B. suis S2 (Supplementary Figure S3). Thus, BrBI played a crucial role in modulating the sensitivity of B. suis S2 to antibiotics.

Given the prominent phenotypes induced by brbI deletion, we compared the proteomes of the three B. suis S2 strains. In total, 2270 proteins were identified in the three B. suis S2 strains, and 363 significantly affected proteins were differentially expressed in the ΔbrbI strain (Figure 6A). Among these differentially expressed proteins (DEPs), 248 were down regulated, and 115 were up regulated. In the complemented strain, 138 proteins were differentially expressed, consisting of 94 down-regulated and 44 up-regulated proteins (Figure 6B). In the ΔbrbI::brbI strain, complementation of brbI recovered 263 DEPs that were identified in the ΔbrbI strain but induced 38 ΔbrbI::brbI-specific DEPs, and 100 DEPs were overlapped with the ΔbrbI strain (Figure 6C). This suggested that brbI may have extra functions when expressed beyond physiologic levels. Figures 6C–E show a slight discrepancy in the numbers of overlapped DEPs, attributed to two proteins, BrBI and a DUF1127 domain-containing protein (WP_002965878.1), which were up-regulated in the ΔbrbI::brbI strain but down-regulated in the ΔbrbI strain (Supplementary Table S1). Compared with the B. suis S2 strain, the remaining 98 overlapped DEPs were regulated in the same directions in the ΔbrbI and ΔbrbI::brbI strains. Although these 98 proteins were still differently expressed in the ΔbrbI::brbI strain compared with the B. suis S2 strain, 21 of them were significantly complemented (Supplementary Table S2), showing a 78.24% ([21 + 263]/363) complementation effect. Cluster analysis of 401 DEPs identified in the ΔbrbI and ΔbrbI::brbI strains showed a similar result (Figure 6F); these DEPs were divided into ΔbrbI-specific, ΔbrbI::brbI-specific, significantly complemented, and uncomplemented DEPs, indicating an acceptable complementation of the complemented strain.

Gene Ontology analysis was performed to further analyse the effects of brbI deletion in B. suis S2. The 363 DEPs between the ΔbrbI and B. suis S2 strains were enriched to GO terms at the second level in three GO categories: biological process, cellular component and molecular function (Figure 6G and Supplementary Table S3). For biological process, brbI deletion mainly affected proteins associated with “regulation of DNA-templated transcription,” “oxidation-reduction,” “transport,” “proteolysis,” and “protein folding.” For cellular component, BrBI was primarily involved in the “plasma membrane,” “outer membrane-bounded periplasmic space,” “cytoplasm,” “integral component of plasma membrane,” and “cytosol.” For molecular function, BrBI mainly regulated proteins associated with “protein binding,” “ATPase activity coupled to transmembrane movement of substances,” “transporter activity,” “catalytic activity,” and “DNA binding transcription factor activity.” In summary, brbI deletion predominantly affected proteins associated with the membrane, transporter, and transcription.

Given the effect of brbI deletion on transcriptional activities and because 14 of 363 DEPs identified in the ΔbrbI strain were transcriptional factors (Supplementary Table S4), we compared the transcriptomes of the three B. suis S2 strains.

In total, we obtained 14,610,410, 11,764,696, and 12,634,380 clean reads from the B. suis S2 strain, 12,396,722, 9,776,270, and 9,215,932 clean reads from the ΔbrbI strain, and 10,609,342, 18,576,960, and 12,138,906 clean reads from the ΔbrbI::brbI strain. Further, 87.46–93.46% clean reads were mapped to the B. suis S2 reference genome, and 2986 mRNAs were identified in the three B. suis S2 strains (Supplementary Table S5). Compared with the B. suis S2 strain, 677 genes were significantly differentially transcribed in the ΔbrbI strain, including 347 down-regulated and 330 up-regulated genes (Figure 7A and Supplementary Table S6). Compared with the B. suis S2 strain, 144 up-regulated and 170 down-regulated genes were identified in the ΔbrbI::brbI strain (Figure 7B and Supplementary Table S6). In the ΔbrbI::brbI strain, brbI complementation recovered 442 DEGs that were identified in the ΔbrbI strain and induced 79 ΔbrbI::brbI-specific DEGs, and 235 DEGs were overlapped between the ΔbrbI and ΔbrbI::brbI strains (Figures 7C–E). This also suggested that brbI may have extra functions when expressed beyond physiologic status. Compared with the B. suis S2 strain, the 235 overlapped DEGs were regulated in the same directions in the ΔbrbI and ΔbrbI::brbI strains, and 68 of them were significantly complemented in the ΔbrbI::brbI strain (Supplementary Table S7), displaying a 75.33% ([68+442]/677) complementation effect. Cluster analysis revealed that 756 DEGs identified in the ΔbrbI and ΔbrbI::brbI strains could be divided into ΔbrbI-specific, ΔbrbI::brbI- specific, significantly complemented, and uncomplemented DEGs (Figure 7F). Considering the stronger sensitivity of transcriptomics to changes, the 75.33% complementation should be an acceptable effect in the complemented strain.

Gene Ontology enrichment analysis showed that the highly enriched GO terms in the ΔbrbI strain were mainly associated with the biological processes of “metabolic process,” “cellular process,” “organic substance metabolic process,” “cellular metabolic process,” and “primary metabolic process”; the cellular component of “cell part,” “cell,” “plasma membrane,” “cytoplasm,” and “membrane”; the molecular function of “catalytic activity,” “binding,” and “protein binding” (Figure 7G and Supplementary Table S8). These enriched terms were mostly consistent with those in the proteomic profiles with a slight difference in that the predominant genes enriched in the transcriptomic profiles were involved in multiple metabolic processes. This suggested that BrBI played extensive roles in B. suis S2 at both the protein and mRNA levels.

According to the proteomic and transcriptomic profiles, brbI deletion extensively affected the physiology of B. suis S2 at both the protein and mRNA levels. Therefore, we integrated transcriptomic and proteomic analyses to explore the in-depth effects of brbI deletion on B. suis S2.

As mentioned, 2986 mRNAs and 2270 proteins were identified in the transcriptomic and proteomic profiles, respectively, and these identified proteins could all be assigned to the identified mRNAs (Figure 8A). Thus, we focused on the 2270 genes that were identified in both the transcriptomic and proteomic profiles. Of the DEGs identified in the ΔbrbI strain, 422 were included in the 2270 genes (Figure 8B). Therefore, 635 genes were significantly affected at the mRNA and/or protein levels in the ΔbrbI strain (Figure 8C), and we thus targeted these genes going forward. Among the 635 genes, 213 were significantly affected at only the protein level (mRNA_no and protein_sig), 272 were significantly affected at only the mRNA level (mRNA_sig and protein_no), and 150 were significantly affected at both the mRNA and protein levels (mRNA_sig and protein_sig) (Figure 8C and Supplementary Table S4).

The existence of the mRNA_no and protein_sig genes in the ΔbrbI strain indicated that BrBI participated in the processes of protein translation, stability, and degradation. Because BrBI is a transmembrane protein, and its homolog, E. coli YccA, has been characterized as a protease activity-associated protein, we speculated that BrBI may initiate its extensive roles by regulating protein degradation. Consequently, the occurrence of 272 genes being affected at only the mRNA level suggested that some of the DEPs induced by brbI deletion should be transcription factors, which was revealed in the proteomic analysis. However, these mRNA_sig and protein_no genes were unaffected by brbI deletion at the protein level, indicating that these genes may have a powerful and possibly BrBI-independent protein regulation pathway to overcome the altered gene transcriptions and restore the corresponding proteins to a physiological level. Most of the mRNA_sig and protein_sig genes were regulated in the same directions at the mRNA and protein levels, indicating that these genes should be regulated by brbI deletion at only the mRNA level. However, 9 out of 150 mRNA_sig and protein_sig genes showed opposite regulation trends at the mRNA and protein levels. BSS2_RS10645, BSS2_RS10185, and BSS2_RS10195 were down-regulated at the mRNA level but up-regulated at protein level, and BSS2_RS04340, BSS2_RS01465, BSS2_RS00810, BSS2_RS01210, BSS2_RS15335, and BSS2_RS06495 were upregulated at the mRNA level but downregulated at the protein level (Supplementary Table S4), indicating that they should have a powerful, predominant, and BrBI-dependent protein regulation pathway that can reverse the transcription alterations induced by brbI deletion.

According to the known Brucella genome annotations, five groups of genes were summarized from the 635 genes discussed here: membrane protein-, cell-division protein-, transcriptional regulator-, peptidase/protease-, and transporter-associated protein-encoding genes (Supplementary Table S4). In total, 27 membrane protein-, 4 cell-division protein-, 31 transcriptional regulator-, 21 peptidase/protease-, and 81 transporter-associated protein-encoding genes were identified. The peptidase/protease- and transporter-associated protein-encoding genes were generally distributed into the mRNA_sig and protein_sig (4 genes and 26 genes, respectively), mRNA_no and protein_sig (14 genes and 25 genes, respectively), and mRNA_sig and protein_no (3 genes and 30 genes, respectively) groups. The transcriptional factor-encoding genes were distributed into the mRNA_sig and protein_no (14 genes) and mRNA_no and protein_sig (17 genes) groups. The membrane protein-encoding genes were mainly distributed into the mRNA_sig and protein_sig (seven genes) and mRNA_no & protein_sig (20 genes) groups. The cell-division protein-encoding genes were specifically distributed into the mRNA_no and protein_sig group.

These results revealed that BrBI plays extensive roles in B. suis S2 via protein and/or transcriptional pathways. Importantly, the generally distributed transcriptional factor- and peptidase/protease-encoding genes could provide insights into the underlying mechanisms of BrBI. BrBI may initially modulate transcriptional factors or peptidase/proteases at the protein level, which may further trigger regulations on other regulatory or functional genes at the mRNA and/or protein levels. Notably, the transporter-associated protein-encoding genes were the most abundant and generally distributed, which may explain why brbI deletion led to the significantly increased sensitivity of B. suis S2 to environmental stresses and antibiotics. Most of the transporter-associated proteins were membrane-related, indicating that BrBI plays a crucial role in the B. suis S2 membrane. Further, 27 membrane protein-encoding genes were identified, and interestingly, most were mRNA_no and protein_sig genes. The identified cell-division protein-encoding genes were all mRNA_no and protein_sig genes, indicating a specific role of BrBI in B. suis S2 division. Therefore, the role of BrBI in the B. suis S2 division and membrane are discussed in detail in the following sections.

In Gram-negative bacteria, cell division is tightly mediated by a multi-protein complex called the divisome that is assembled to the mid-cell. FtsZ, FtsA, FtsB, FtsI, FtsL, FtsQ, and FtsW are the primary components of the divisome that execute bacterial cytokinesis (Dajkovic and Lutkenhaus, 2006; Szwedziak and Ghosal, 2017). The divisome is assembled in two stages. First, FtsZ polymers are attached to the mid-cell membrane by interacting with FtsA to form a Z-ring scaffold, which provides the constrictive force. Second, the late-division proteins, including FtsE, FtsK, FtsQ, FtsL and FtsB, FtsW, FtsI, and FtsN, are sequentially recruited to the Z-ring to form the complete septum to carry out cell division (Dajkovic and Lutkenhaus, 2006; Szwedziak and Ghosal, 2017). The early (Z-ring) and late divisome components are linked by the FtsQBL complex, which may have a structural role as a scaffold in assembling the divisome (Choi et al., 2018; Condon et al., 2018). In most bacteria, to guarantee that division occurs at the correct site, a Min system comprising MinC, MinD, and MinE prevents establishing the divisome at bacterial poles (Szwedziak and Ghosal, 2017; Ramm et al., 2019). As a spatial modulator of the divisome, defects in the Min system produce small anucleate minicells from the poles of rod-shaped mother cells, increasing the average cell length of the DNA-containing mother cells (Lutkenhaus, 2007).

As mentioned, brbI deletion significantly affected four cell-division protein-encoding genes, indicating a crucial role of BrBI in B. suis S2 division. Therefore, we integrated proteomic and transcriptomic analyses to thoroughly explore how brbI deletion would affect B. suis S2 division. First, brbI deletion did not affect FtsZ and FtsA expression at either the mRNA or protein levels, indicating a minor role of BrBI in Z-ring assembly (Figure 9 and Supplementary Table S9). Further, deleting brbI did not affect the late divisome proteins, FtsE, FtsK, and FtsW, at either the mRNA or protein levels; however, brbI deletion significantly upregulated FtsB, FtsI, FtsL, and FtsQ at only the protein level, possibly indicating disordered cytokinesis septum formation (Figure 9 and Supplementary Table S9). Moreover, brbI deletion did not affect MinC, MinD, and MinE expression at either the mRNA or protein levels, indicating a minor role of BrBI in the Min system (Figure 9 and Supplementary Table S9).

In summary, brbI deletion may affect proper formation of the division septum, which may further lead to aberrant B. suis S2 division. Specifically, the role of BrBI in B. suis S2 division may occur predominantly by regulating the translation/degradation/stability of FtsB, FtsI, FtsL, and FtsQ.

The integrated proteomic and transcriptomic analyses showed that brbI deletion significantly affected the 27 identified membrane protein-encoding genes (excluding ftsB, ftsI, ftsL, and ftsQ); of these, seven were affected at both the mRNA and protein levels, and 20 were affected only at the protein level (Supplementary Table S4). Outer membrane proteins (OMPs) BamA and LptD were significantly down-regulated at only the protein level in the ΔbrbI strain (Figure 10 and Supplementary Table S9). In Gram-negative bacteria, the lipopolysaccharide transport (Lpt), β-barrel assembly machine (Bam), and the localization of lipoproteins (Lol) pathways are required to transport the outer membrane components (Sperandeo et al., 2019). Lpt, Bam, and Lol are necessary for respectively transporting LPS, β-barrel proteins, and lipoproteins across the periplasm to the outer membrane. Thus, down-regulation of BamA and LptD induced by brbI deletion indicated an alteration of the B. suis S2 outer membrane. Consistently, the OMP2a, OMP2b, OMP22, OMP25b, OMP31, and OMPW family protein (WP_002964666.1) and outer membrane β-barrel protein (WP_004689965.1) were downregulated at only the protein level with the exception of OMP25b, which was downregulated at both the mRNA and protein levels in the ΔbrbI strain (Figure 10 and Supplementary Table S9). In Brucella, OMP2a and OMP2b are group 2 OMPs, and OMP22, OMP25b, and OMP31 are group 3 OMPs, which are β-barrel proteins (Goolab et al., 2015). Therefore, down-regulation of OMP25b induced by brbI deletion may occur at the mRNA level; however, regulation of the remaining OMPs by brbI deletion may occur indirectly through BamA at the protein level.

The FtsH protease modulators, HflC and HflK, were also significantly up-regulated at only the protein level via brbI deletion (Figure 10 and Supplementary Table S9). FtsH is a membrane-anchored ATP-dependent metalloprotease that degrades redundant or abnormal membrane proteins to maintain membrane homeostasis in prokaryotes (Langklotz et al., 2012; Bittner et al., 2017). HflC and HflK form a membrane protein complex (HflKC) and further complex with FtsH. It has been reported that HflKC may act as a negative modulator for the proteolytic activity of FtsH (Langklotz et al., 2012). Consequently, we speculate that the upregulated HflC and HflK by brbI deletion may have suppressed the proteolytic activity of FtsH, subsequently leading to an accumulation of membrane proteins or cytosolic regulatory proteins that were meant to be degraded, which eventually perturbed B. suis S2 membrane homeostasis.

Moreover, brbI deletion significantly affected several inner membrane proteins and dozens of membrane-related transporter proteins, but the mechanism by which BrBI regulates these proteins remains unknown. Nevertheless, the integrated proteomic and transcriptomic analyses revealed the extensive roles of BrBI in B. suis S2 membrane homeostasis. Specifically, OMP was regulated by BrBI predominantly at the protein level.