Burkholderia pseudomallei BicA protein promotes pathogenicity in macrophages by regulating invasion, intracellular survival, and virulence

ABSTRACT Burkholderia pseudomallei (Bpm) is the causative agent of melioidosis disease. Bpm is a facultative intracellular pathogen with a complex life cycle inside host cells. Pathogenic success depends on a variety of virulence factors with one of the most critical being the type 6 secretion system (T6SS). Bpm uses the T6SS to move into neighboring cells, resulting in multinucleated giant cell (MNGC) formation, a strategy used to disseminate from cell to cell. Our prior study using a dual RNA-seq analysis to dissect T6SS-mediated virulence on intestinal epithelial cells identified BicA as a factor upregulated in a T6SS mutant. BicA regulates both type 3 secretion system (T3SS) and T6SSs; however, the extent of its involvement during disease progression is unclear. To fully dissect the role of BicA during systemic infection, we used two macrophage cell lines paired with a pulmonary in vivo challenge murine model. We found that ΔbicA has a distinct intracellular replication defect in both immortalized and primary macrophages, which begins as early as 1 h post-infection. This intracellular defect is linked with the lack of cell-to-cell dissemination and MNGC formation as well as a defect in T3SS expression. The in vitro phenotype translated in vivo as ΔbicA was attenuated in a pulmonary model of infection, demonstrating a distinct macrophage activation profile and a lack of pathological features present in the wild type. Overall, these results highlight the role of BicA in regulating intracellular virulence and demonstrate that specific regulation of secretion systems has a significant effect on host response and Bpm pathogenesis. IMPORTANCE Melioidosis is an understudied tropical disease that still results in ~50% fatalities in infected patients. It is caused by the Gram-negative bacillus Burkholderia pseudomallei (Bpm). Bpm is an intracellular pathogen that disseminates from the infected cell to target organs, causing disseminated disease. The regulation of secretion systems involved in entry and cell-to-cell spread is poorly understood. In this work, we characterize the role of BicA as a regulator of secretion systems during infection of macrophages in vitro and in vivo. Understanding how these virulence factors are controlled will help us determine their influence on the host cells and define the macrophage responses associated with bacterial clearance.

Bpm has been routinely imported (3,4), but evidence is beginning to suggest endemicity in the Gulf Coast regions, particularly Mississippi (5,6).Combating Bpm is particularly difficult due to the lack of a licensed vaccine (7), extensive antibiotic resistance, ability to generate persistent infections (8,9), and its intracellular lifestyle (1).As a facultative intracellular pathogen, Bpm utilizes a myriad of virulence factors to promote replication including the type 6 secretion system (T6SS) (10).The T6SS is a contractile nanomachine widely distributed across Gram-negative species that is primarily used to deliver effector proteins for interbacterial competition (11).There is a small subset of T6SSs that have eukaryotic targets, including Bpm which uses the T6SS to spread from cell to cell via cell fusion events, resulting in multinucleated giant cell (MNGC) formation (10,12).MNGC formation is the keystone event during the intracellular pathogenesis process, but the mechanisms behind the membrane fusion events are largely unknown, from both the bacterial and host sides.
Recently, we began investigating the mechanisms of T6SS-mediated virulence within the context of the understudied gastrointestinal (GI) route of infection (13) and performed a dual RNA-seq analysis with primary murine intestinal epithelial cells infected with wild-type (WT) Bpm or a T6SS structural mutant ∆hcp1 (BPSS1498) (14).In that analysis, bicA (BPSS1533) was identified as significantly upregulated in ∆hcp1, and this was particularly interesting as BicA has been implicated in the regulatory network controlling both type 3 secretion system (T3SS) and T6SS expression (15).It has been suggested that BicA is the chaperone/co-activator of BsaN and, together, they act to coordinate the timely expression of T3SS effectors and downstream T6SS regulator genes bprC and virAG (16).Interestingly, bsaN was significantly downregulated in ∆hcp1, which suggests a different involvement of BicA in the regulation of T3SS and T6SS.We evaluated the contribution of BicA within the GI model of infection and determined that it is critical for intracellular replication, cell-to-cell spread, and lethality (14).As the dynamics of the GI model of infection are unclear, we began interrogating the mecha nisms of ∆bicA in the well-characterized and systemic macrophage model.Macrophages are multi-function immune cells that are present as specialized tissue resident cells or circulating undifferentiated monocytes.They are capable of a variety of jobs including pathogen clearance, antigen presentation, and immune coordination via secretion of cytokines and chemokines, which greatly influence the inflammatory landscape (17).This makes macrophages an attractive target for subversion by intracellular pathogens like Bpm, which uses macrophages as a replicative niche and a dissemination "trojan horse" (18,19).It has been hypothesized that Bpm modulates the activation state of macro phages to promote replication versus clearance, but the mechanism of this modulation is unclear (20,21).
In this work, we demonstrate that BicA is necessary for successful replication inside macrophages using both immortalized and primary macrophages and this intracellular defect starts upon entry into the cell.Furthermore, we characterized the expression profile of critical virulence factors in ∆bicA versus WT Bpm strains and found a dis tinct dysregulation of multiple systems.Finally, we examined the role of BicA during inhalational melioidosis and the phenotype of pulmonary macrophages in response to infection.Collectively, we provide evidence to establish BicA as a major regulator of virulence and that its absence differentially activates macrophages and leads to clearance of the bacteria.

BicA required for proficient intracellular replication in macrophages
We previously reported that ∆bicA demonstrated a profound intracellular replication defect in intestinal epithelial cells (14), so we began our investigation into the role of BicA in the macrophage model by examining intracellular replication in two in vitro models: RAW 264.7 (RAW) cells and primary bone marrow-derived macrophages (BMDMs).RAW 264.7 cells are an immortalized macrophage-like cell line originally created from BALB/c mice, so we chose to use BALB/c mice as bone marrow donors for the primary BMDM model.We evaluated the intracellular replication of Bpm K96243 WT, ∆bicA, and ∆bicA::bicA at 3, 6, and 12 h post-infection (hpi) in both RAW cells and BMDMs (Fig. 1A and B) and found that ∆bicA replicates at a lower rate as early as 3 hpi and that defect persists through 12 hpi.The replication profiles were nearly identical between macrophage models, which led us to continue the characterization using both cell models.Since we saw decreased replication at the earliest timepoint of 3 hpi, we wanted to ensure this was not due to differential phagocytosis and to look at the very early stages of infection before the bacteria starts to replicate in the cytoplasm.We measured phagocytosis rates among WT, ∆bicA, and ∆bicA::bicA and found no significant differences after 1 h of internalization (Fig. 1C).We also looked at intracellular survival 1 h post-phagocytosis or 2 h after initial contact of the macrophages with the bacterial inoculum and found that ∆bicA exhibited a significant decrease in intracellular survival when compared to WT and ∆bicA::bicA (Fig. 1D).Since bacterial enumeration does not tell a complete story, we visualized BMDMs infected with WT, ∆bicA, or ∆bicA::bicA or mock infected at 3, 6, and 12 hpi via immunofluorescence (Fig. 2).The WT-and ∆bicA::bicA-infected cells showed robust replication at 3 hpi, MNGC formation at 6 hpi, and a sharp decrease in cell viability at 12 hpi.The ∆bicA-infected cells revealed few intracellular bacteria that, qualitatively, seemed static compared to WT and ∆bicA::bicA.These cells also showed no MNGC formation and retained mock levels of viability later into the time course (Fig. 2).Together, these data suggest that BicA is critical for intracellular replication and progression through the pathogenesis process.The

ΔbicA strain demonstrates disrupted expression of critical virulence factors
Because BicA is predicted to be involved in the regulation cascade of both the T3SS and T6SS machineries, we evaluated the expression of an array of virulence genes during macrophage infection.We selected a panel of genes important during intracellular survival, which include the T3SS, T6SS, actin motility proteins, and some of the regulators of these virulence mechanisms (Table 1).RAW 264.7 cells were infected with WT, ∆bicA, or ∆bicA::bicA, and bacterial RNA was extracted at 3 and 6 hpi and semi-quantitative PCR was performed on cDNA libraries generated from bacterial RNA.Beginning at 3

FIG 2
The ΔbicA strain appears to exhibit reduced motility and lacks MNGC formation.BALB/c BMDMs were infected at an MOI of 10 with Bpm K96243 WT, ΔbicA, or ΔbicA::bicA for 3, 6, or 12 h before being fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.25% Triton-X 100.Bacterial cells were washed and stained with sera from mice vaccinated with a live-attenuated Bpm followed by anti-mouse IgG, IgM, and IgA (H + L) secondary antibody conjugated to Alexa Fluor 488.Actin and DNA were visualized using rhodamine phalloidin and DAPI, respectively.Images were visualized under 100× magnification using Olympus BX51 upright fluorescence microscope and further analyzed using ImageJ software.
hpi, ∆bicA exhibits slight increased expression of T3SS effectors with downregulation of T6SS structural loci and actin motility when compared to WT (Fig. 3; Fig. S1).This trend continues to progress at 6 hpi with elevated levels of T3SS effector expression and greater downregulation of the T6SS.Actin motility genes are slowly recovering expression as bimA is on par with WT at 6 hpi; however, bimC is still downregulated.Interestingly, at both 3 and 6 hpi, there appears to be a slight increase in activity at the tssA-virAG operon but with no evident increase in expression of T6SS structural genes.This operon contains the primary regulators (virAG) and structural protein (tssA), so the production of these proteins should lead to the production of T6SS machinery, but in ∆bicA, this dynamic is interrupted through an unknown regulatory event.Expression levels of ∆bicA::bicA were also compared to WT, but no differences were detected (Table S1).The difference in expression of critical virulence factors sheds light on the regulatory impact of BicA and begins to explain the intracellular behavior of ∆bicA.The mechanism that restricts the intracellular survival of the ΔbicA strain is unknown, but it likely tied to alteration of T3SS activity, as this secretion system is responsible for early intracellular success of Bpm.

The ΔbicA strain is attenuated in an intranasal challenge model and results in differential macrophage recruitment and activation
Previously, we examined the virulence of ΔbicA in both acute and chronic models of gastrointestinal infection (14) and found that although ΔbicA demonstrated attenuation, those infection models are not able to discern the specific role of macrophages during infection.To fully investigate the virulence of ΔbicA, we intranasally challenged BALB/c mice with 3-5 LD 50 Bpm K96243, ΔbicA, or ΔbicA::bicA and monitored their survival for 21 days (Fig. 4A).All animals challenged with WT K96243 and ΔbicA::bicA succum bed to infection or reached the humane endpoint on day 4 post-infection.Inversely, animals challenged with ΔbicA demonstrated 100% survival and showed no outward signs of infection besides minor body weight loss that was eventually recovered to near pre-challenge levels (Fig. 4A and B).It should be noted that the body weight loss observed in a subset of ΔbicA-challenged animals was delayed when compared to WT and ΔbicA::bicA.The reason behind the delayed onset weight loss is unclear, but it could be related to the decreased intracellular replication phenotype seen in vitro (Fig. 1A and  B).The survivors of the ΔbicA challenge were euthanized 21 days post-infection, and lungs, liver, and spleen were collected for bacterial enumeration.Moderate amounts of bacteria were found in the lungs with very low titers in the liver and spleen (Fig.  4C).One animal had high titers in both lungs and spleen, and the spleen was enlarged with visible abscesses; this gross pathology is common in spleens chronically infected with WT Bpm (22).Once the attenuation of ΔbicA was established in this model, we sought to investigate the role of macrophages during pulmonary infection.Another set of BALB/c mice were challenged with Bpm K96243, ΔbicA, or ΔbicA::bicA, and at 48 h post-challenge, the lungs were harvested and divided to be processed for flow cytometric analysis and sectioned for pathological scoring.We designed a panel that allowed us to examine the macrophage populations within the lungs during infection and the basic activation state of those macrophages (Table 2).The gating strategy used to identify the total pulmonary macrophage populations in the lungs was modified from (23), but extra markers were added to assess the basic activation state of the populations identified (Fig. S2).Macrophages have the ability to polarize to M1 and M2 subsets; M1 is the classic inflammatory profile associated with pathogen clearance, while M2 subsets have immunoregulatory roles connected to tissue healing and limitation of inflammatory damage (20).Macrophage polarization is a complex system, but for our purposes, we have simplified the populations to M1-like expressing CD80 and CD86, while M2-like populations express CD163 and Arginase-1.These markers have allowed us to assess the basic activation states and gain insights into the dynamics of infection through the macrophage lens.Animals infected with WT K96243 and ΔbicA::bicA displayed a trend of having higher total level of macrophages when compared to ΔbicA-infected animals (Fig. 5A and B).When those macrophage populations were examined for M1and M2-like phenotypes, we found that while WT and ΔbicA::bicA had higher total numbers of M1-polarized macrophages (Fig. 5C), all groups had similar rates of M1-like activated macrophages (Fig. 5D).WT and ΔbicA::bicA generated a small but distinctive M2-like population that was absent from ΔbicA (Fig. 5E and F).Lung sections collected for pathological scoring were fixed in formalin, and hematoxylin and eosin (H&E) stained before being scored blind by a pathologist.Slides were scored on a 1-3+ system based on four criteria: nodules of inflammation, karyorrhectic debris/apoptosis, hemorrhage and congestion, and alveolar collapse.Pathology in the lungs of WT-and ΔbicA::bicAinfected animals was characterized by large nodules of inflammation that were densely populated by mononuclear cells and apoptotic debris with moderate amounts of congestion and alveolar collapse.Conversely, ΔbicA-infected lungs were characterized by the complete lack of inflammatory nodules but slight to moderate congestion and hemorrhaging (Fig. 6A and B).Interestingly, higher levels of infiltrating macrophages and M2-like populations found in WT and ΔbicA::bicA correlated with the presence of large inflammatory nodules and apoptotic debris, but all conditions contain slight to moderate congestion and relatively equal M1-like populations.It is unclear whether this M2-like population elicited by the WT and ΔbicA::bicA is a response to uncontrolled bacterial replication and tissue damage or an intentional mechanism utilized by Bpm.
Overall, these data demonstrate that ΔbicA is attenuated in an inhalational model and elicits a distinct pulmonary macrophage profile that correlates with pathological features disparate from WT and ΔbicA::bicA.

DISCUSSION
Melioidosis is a neglected tropical disease (24) that is a growing threat to nearly every continent on the globe.With recent findings demonstrating that Bpm is endemic to areas thought to be low burden or even non-endemic (5,6,25) and high incidence of antibacterial resistance, it is paramount that the underlying host-pathogen interactions are understood.The T6SS-mediated virulence is poorly characterized, but previously, we performed dual RNA-seq on intestinal epithelial cells infected with WT or ΔT6SS Bpm to tease apart what bacterial and host factors are influenced by the T6SS.BicA was identified as differentially upregulated in the ΔT6SS mutant and then selected for further characterization due to its implication in the T3SS and T6SS regulatory cascade.We found that a ΔbicA mutant was attenuated in both in vitro and in vivo models of gastrointestinal infection (14).To fully characterize the role of BicA during infection, we investigated its contribution during macrophage and pulmonary infection.As "profes sional phagocytes, " macrophages bear the responsibility of being first responders to infection, pathogen uptake and clearance, antigen presentation, and immune modula tion/coordination (26).The behavior of macrophages has a profound impact on the microenvironment and systemic recruitment of other immune cells.For these reasons, macrophages are an attractive target for Bpm, as manipulation can skew the immune response to be advantageous for replication.We chose two macrophage models, RAW 264.7 cells and BMDMs from BALB/c mice.RAW 264.7 cells are an immortalized cancerous cell line, so we wanted to compare any phenotype with primary cells from the same murine background to ensure that the phenotypes were not an artifact of these immortalized cells.The ΔbicA strain demonstra ted an intracellular replication defect, compared to WT and ΔbicA::bicA in both RAW 264.7 and BMDMs at 3, 6, and 12 h (Fig. 1A and B).This defect was not due to a differen tial rate of phagocytosis (Fig. 1C) but decreased intracellular survival starting as early as 1 h post-infection (Fig. 1D).ΔbicA appeared to remain trapped within macrophages and did not form MNGCs (Fig. 2), while the complemented strain recapitulated the WT features of actin motility and MNGC formation.Expression of some virulence factors was measured in ΔbicA and compared to WT at 3 and 6 h post-infection in RAW 264.7 cells (Fig. 3), and it confirmed qualitative observations from Fig. 2. Actin motility genes (bimA and bimC) and T6SS structural proteins (hcp1 and tssB) were generally repressed in ΔbicA at 3 and 6 h; however, T3SS effector genes (bopA, bapA, and bapB) were upregulated at 6 h.Interestingly, there was activity on the tssA-virAG operon that was on par with WT, but the downstream activity of T6SS genes hcp1 and tssB was repressed.This suggests that there might be a secondary signal for production of the T6SS that is lacking during ΔbicA intracellular infection.One plausible explanation is that the lack of BicA is interrupting the co-activation/chaperoning of BsaN which feeds forward to activate the T6SS and actin motility.However, at 3 h, bsaN and bprC expression was on par with WT, which translated downstream to virAG activity even at 6 h when bsaN expression dips below WT levels.Downstream signaling events that should lead to T6SS activity occurred with diminished expression of hcp1 and tssB so the lack of co-activator/chaperone activity is an unlikely source of the phenomenon.Alternatively, another explanation is that the ΔbicA strain lacks access to host cytoplasmic molecules that aid in expression of virulence factors.It has been shown that host glutathione aids in upregulating T6SS genes like hcp1, so it is possible that ΔbicA is being sequestered away from these molecules (27).The upregulation of T3SS effectors in ΔbicA could suggest that T3SS activity is upregulated, and the restrictive replication environment is potentially caused by increased expression T3SS proteins that are sensed by pattern recognition receptors like NLR apoptosis inhibitory proteins and results in inflamma some activation (28).However, the BicA homolog in Salmonella, SicA, is responsible for stabilizing or preventing degradation of SipB and SipC in the bacterial cytoplasm, and ΔsicA has been shown to be partially complemented with the addition of bicA (29,30).SipB and SipC form the translocon pore of the T3SS, so it is possible that ΔbicA is unable to produce an active T3SS.This notion is supported by reports that BicA is required for secretion of BopA and BopE (31).If this is the case, then ΔbicA is likely trapped in the phagosome for extended periods of time as the T3SS is responsible for escape, but mutants do exhibit an independent, albeit delayed, mechanism of escape (32).In Fig. 2, ΔbicA appears primarily clustered at 3 and 6 h post-infection, which might suggest they might be trapped in the phagosome as WT and ΔbicA::bicA replicate in a more diffuse pattern within the cytoplasm.Being trapped in the phagosome would restrict access to cytoplasmic host molecules like glutathione and at least partially explain the repression of the T6SS in the presence of regulatory events that should promote expression.
BicA has previously been implicated during inhalational melioidosis (33), but this was done with a transposon-based interruption of the gene, so we sought to confirm the importance of BicA using our ΔbicA isogenic mutant.When BALB/c mice were intranasally challenged with 3-5 LD 50 of WT, ΔbicA, or ΔbicA::bicA, we observed 100% survival in ΔbicA-challenged animals, whereas all animals succumbed to infection on day 4 post-infection in the WT and ΔbicA::bicA groups (Fig. 4A).The ΔbicA-infected mice only presented a delayed but slight decrease in body weight that recovered to pre-challenge levels suggesting they were able to effectively control the infection (Fig. 4B).On day 21, the ΔbicA survivors were euthanized, and bacterial loads were assessed in the lungs, liver, and spleen.Low levels of bacteria were detected in all three organs with the lungs being the location of higher bacterial numbers (Fig. 4C).It should be noted that all animals exhibited dissemination from the lungs to the liver and spleen, but very few bacteria were recovered from these sites.One animal had robust replication in both lungs and spleen with visible abscesses on the organs, which is a feature that is common in WT-infected organs, but this is likely an outlier event (22).The factors that created an environment conducive to ΔbicA replication in this animal are currently unknown.
Characterizing how intracellular pathogenesis events influence the host response is paramount to identifying avenues that can be targeted to combat the pathogen.To begin assessing this, we designed a flow cytometry panel to explore macrophage populations in the lungs during infection.We chose 48 h post-challenge to assess macrophage activity due to the stark contrast between WT/ΔbicA::bicA compared to ΔbicA; in our in vivo survival study at this timepoint, the groups started to diverge in disease severity (Fig. 4C).We reasoned that any differences at this timepoint would provide useful insight to the dynamics of the immune response to infection.A compre hensive gating strategy was devised using (23) as a guide with added polarization markers to delineate this portion of the inflammatory landscape (Fig. S2).Although WTand ΔbicA::bicA-infected animals tended to elicit higher numbers of total pulmonary macrophages (Fig. 5B), all three experimental groups had equal rates of M1 polariza tion (Fig. 5D).It should be noted that ΔbicA::bicA had significantly more M1-polarized macrophages than in both WT and ΔbicA (Fig. 5C) but the rate or proportion at which macrophages polarized was not different.Interestingly, WT and ΔbicA::bicA also had distinct M2-like populations (Fig. 5E and F).This suggests that the presence of proinflammatory M1 macrophages aids in controlling bacterial replication, but there is a larger, negative contribution by M2 macrophages to promote replication.The phenomenon of skewing both M1 and M2 subsets is not unique to Bpm, it is a trait shared by Mycobacterium tuberculosis, Mycobacterium leprae, and Coxiella burnetti (20).Both subsets can lead to downstream pathogenic effects on the host as uncontrolled inflammation from M1 can create excess tissue damage and, conversely, M2 can create an antiinflammatory environment that allows pathogens to replicate undetected (20,26).In conjunction with flow cytometry, we analyzed the tissue histopathology of a select number of mice in this cohort, and the lungs were scored by a pathologist based on multiple criteria: nodules of inflammation, karyorrhectic debris/apoptosis, hemor rhage and congestion, and alveolar collapse.WT-and ΔbicA::bicA-infected lungs were characterized by large, discrete nodules of inflammation full of infiltrating mononuclear cells and apoptotic debris plus moderate amounts of congestion and alveolar collapse.The ΔbicA-infected lungs lack the pronounced nodules but exhibit slight to moderate amounts of hemorrhage and congestion (Fig. 6A and B).The nodules of mononuclear cells match the increase in pulmonary macrophages and are centered on replication hotspots.The presence of M2-like macrophages cannot be directly mapped to these foci, but the centers being full of apoptotic debris increase the likelihood as M2 is more readily able to clear this debris through efferocytosis (34,35).
In summary, we have explored the role of BicA in both immortalized and primary macrophages, demonstrating that ΔbicA has an intracellular survival defect.This defect is the result of a disruption in virulence factor expression defined by repression of the T6SS and actin motility; however, the cause of this repression is not fully understood.The ΔbicA mutant is highly attenuated in an inhalation model of melioidosis, inducing a differential macrophage recruitment and polarization profile paired with less severe histopathology.WT-and ΔbicA::bicA-infected mice recruit higher numbers of macro phages and promote a distinct M2 population that is absent in ΔbicA, suggesting that M2 polarization might be deleterious during infection.

Bacterial strains and growth conditions
All experiments were conducted with the prototypical wild-type strain of Burkholderia pseudomallei K96243 or derivative strains (K96243 ∆bicA and K96423 ∆bicA::bicA).K96243 ∆bicA and K96423 ∆bicA::bicA were constructed in Sanchez-Villamil et al. (14).All strains were routinely grown at 37°C on Luria Bertani (LB) agar plates and in LB broth with shaking.
RAW 264.7 cells or BMDMs were seeded at 5 × 10 5 per well in complete DMEM or RPMI without antibiotics into 24-well plates and allowed to adhere overnight.Bpm strains were streaked on LB agar plates and grown at 37°C for 48 h; the LB broth was inoculated and grown at 37°C with shaking for 12 h.Bacterial culture was diluted to 5 × 10 6 CFU/mL in antibiotic-free complete DMEM or RPMI and added to the cells for an MOI of 10.Cells were incubated with inoculum for 1 h for internalization and washed with PBS, and then, the media containing 1 mg/mL kanamycin were added for 1 h to kill off extracellular bacteria (13,14).For bacterial enumeration, cells were washed twice with PBS to remove any extracellular bacteria, lysed with 0.1% TritonX-100, serially diluted in PBS, and plated on LB agar plates.

Immunofluorescence assay
Infected BALB/c murine BMDM cells were fixed with 4% paraformaldehyde for 30 min following the select agent inactivation protocol approved by University of Texas Medical Branch (UTMB) Environment Health and Safety.Cells were permeabilized with 0.25% Triton X-100 in PBS for 7 min at RT before incubation with serum (1:1,000) from Bpm ΔtonBΔhcp1 (PBK001) live-attenuated vaccine-immunized mice (36).Cells were washed then incubated with 1:5,000 goat anti-mouse IgG, IgM (H + L) secondary antibody conjugated to Alexa 488 (Invitrogen) followed by actin and DNA staining using rhodamine phalloidin (Invitrogen) and DAPI (Sigma) together at 1: 10,000 dilutions.The coverslips were mounted using Prolong gold antifade (Molecular Probes, Life Technol ogy) and sealed with nail polish.Stained cells were visualized using an Olympus BX51 upright fluorescence microscope and analyzed using ImageJ software from National institutes of Health (NIH).

Measurement of gene expression by qRT-PCR
RNA for qRT-PCR analysis was prepared from two independent experiments of infected RAW 264.7 cells at 3 and 6 h post-infection at MOI 10.Bacterial RNA was isolated using Ambion TRIzol reagent (Life technologies) and Direct-zol RNA Miniprep kit (Zymo Research).The cDNA was synthesized using the iScript cDNA synthesis kit (Bio-Rad) following the manufacturer's protocol (priming: 25°C, 5 min; reverse transcription: 42°C, 30 min; RT inactivation: 85°C, 5 min; and store temperature: 4°C).The concentration and purity of cDNA were measured and normalized to 100 ng/mL for qRT-PCR step.The primers for qRT-PCR indicated in Table 1 were designed and then evaluated for specificity by conventional PCR using Q5 Highfidelity DNA polymerase (New England Biolab).Gene expression was quantified using QuantiNova SYBR green (Qiagen) following the PCR cycling program as follow: initial heat activation step at 95°C for 2 min; two-step 40 cycles of 5 s at 95°C and 30 s at 60°C.The threshold cycle and melting curve of each gene were automatically established and recorded by the software CFX Maestro Software (version 4.0).Relative gene expression level of each gene in ΔbicA mutant was normalized to wild-type strain using the 2 −ΔΔ Ct method with 16S rRNA as reference gene.

Intranasal challenge and survival studies
Female 6-to 8-week-old BALB/c mice (n = 5 per group) (Jackson Laboratories) were intranasally (i.n.) challenged with 3-5 LD 50 Bpm K96243, ΔbicA, or ΔbicA::bicA in 50 µL (25 µL per nare).One LD 50 is equal to 312 CFU.Infected mice were monitored for survival and weight loss for 21 days post-infection and euthanized if the animal reached the threshold for humane endpoint.On day 21 post-infection, survivors were humanely euthanized, and lungs, liver, and spleen were collected for bacterial enumeration.

Flow cytometry
Female 6-to 8-week-old BALB/c mice (n = 5 per group) (Jackson Laboratories) were i.n.challenged with 3-5 LD 50 Bpm K96243, ΔbicA, or ΔbicA::bicA; at 48 hpi, animals were euthanized, and lungs were harvested for processing.Lung tissue was cut into small pieces and dissociated via incubation for 30 min at 37°C with slight rocking in RPMI plus 0.5 mg/mL collagenase IV and 30 µg/mL DNase I.The dissociated tissue was homogen ized through a 100-µm cell strainer, and fibroblasts and debris were pelleted via a 60 × g centrifugation for 1 min.Supernatant was collected, and RBCs were lysed for 5 min at RT.Following washes, pulmonary cells were adjusted to 1 × 10 6 cells and stained using the reagents in Table 2. Briefly, cells were incubated with Zombie NIR (BioLegend) for 5 min in PBS, washed, and incubated with TruStain X plus (BioLegend) for 30 min followed by the extracellular antibodies (Table 2).Cells were fixed and permeabilized using Cytofix/Cytoperm (BD Biosciences) and stained for intracellular markers.Fully stained cells were resuspended in 4% ultrapure formaldehyde in PBS for 48 h in accordance with the inactivation protocol approved by UTMB Department of Biosafety before removal from BSL3 laboratory for analysis via BD Symphony full spectrum flow cytometer.Data were analyzed using FlowJo software.

Evaluation of lung pathology
Lungs were collected from mice after humane euthanasia 48 h post-infection and fixed in 10% formalin for 48 h.Formalinfixed lung samples were submitted to the UTMB Anatomical Pathology core for paraffin embedding, mounting, and H&E staining.Stained slides were analyzed and scored by a pathologist (H.L.S.) on a 1-3+ system.Slides were scanned and images taken using Aperio ImageScope.

Statistical analysis
All statistical analyses were done using GraphPad Prism software (v9.0).P-values of <0.05 are considered statistically significant.Survival differences were assessed via Kaplan-Meier survival curve followed by a log-rank test.An ordinary one-way analysis of variance followed by Sidak's multiple comparison test was used to analyze differences in intracellular replication and flow cytometry populations.

FIG 1
FIG 1 The bicA strain demonstrates an intracellular survival defect in macrophages.Macrophages, RAW 264.7 cells (A) or BALB/c BMDMs (B), were infected at a multiplicity of infection (MOI) of 10 with Bpm K96243 WT, bicA, or bicA::bicA and bacteria enumerated at 3, 6, and 12 hpi to assess intracellular replication.(C) Rate of internalization was assessed by incubating bacteria with RAW 264.7 cells for 1 h before enumeration.(D) Early intracellular survival was assessed at 1 hpi.Bars represent an average of two independent experiments performed in triplicate ± SD.Significant differences were assessed via one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test.*P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.0001.

FIG 3
FIG 3The ΔbicA strain displays disrupted expression of virulence factors during intracellular infection.RAW 264.7 cells were infected at an MOI of 10 with Bpm K96243 WT, ΔbicA, or ΔbicA::bicA for 3 or 6 h before being lysed and intracellular bacteria collected via differential centrifugation.Total bacterial RNA was collected from two independent experiments and used as a template for cDNA synthesis.Gene expression was measured by qPCR and the relative gene expression evaluated using the 2 −ΔΔCt method using WT as a control.The heatmap shows expression of ΔbicA relative to WT, meaning anything on the red spectrum is upregulated in ΔbicA and green is repressed in ΔbicA.

FIG 4 The
FIG 4 The ΔbicA strain is attenuated in the intranasal challenge model.BALB/c mice (n = 5 per group) were intranasally challenged with 3-5 LD 50 of Bpm K96243 WT, ΔbicA, or ΔbicA::bicA (1 LD 50 ~312 CFU) and monitored for 21 days post-infection for survival (A) and weight loss (B).Animals were euthanized once the humane endpoint threshold was reached.On day 21 post-infection, ΔbicA survivors were euthanized, and lungs, liver, and spleen were homogenized for bacterial enumeration (C).Error bars in B represent SEM, and lines in C represent median value.

TABLE 1
Primers for qPCR used in this study

TABLE 2
Flow cytometry antibodies and reagents