An Inducible Cre-lox System to Analyze the Role of LLO in Listeria monocytogenes Pathogenesis

Listeriolysin O (LLO) is a pore-forming cytolysin that allows Listeria monocytogenes to escape from phagocytic vacuoles and enter the host cell cytosol. LLO is expressed continuously during infection, but it has been a challenge to evaluate the importance of LLO secreted in the host cell cytosol because deletion of the gene encoding LLO (hly) prevents localization of L. monocytogenes to the cytosol. Here, we describe a L. monocytogenes strain (hlyfl) in which hly is flanked by loxP sites and Cre recombinase is under the transcriptional control of the L. monocytogenes actA promoter, which is highly induced in the host cell cytosol. In less than 2 h after infection of bone marrow-derived macrophages (BMMs), bacteria were 100% non-hemolytic. hlyfl grew intracellularly to levels 10-fold greater than wildtype L. monocytogenes and was less cytotoxic. In an intravenous mouse model, 90% of bacteria were non-hemolytic within three hours in the spleen and eight hours in the liver. The loss of LLO led to a 2-log virulence defect in the spleen and a 4-log virulence defect in the liver compared to WT L. monocytogenes. Thus, the production of LLO in the cytosol has significant impact on the pathogenicity of L. monocytogenes.


Introduction
The field of microbial pathogenesis and the study of virulence factors has been guided for decades by Molecular Koch's Postulates, which stipulate that inactivation of a gene encoding a suspected virulence factor should lead to measurable loss of virulence, and replacement of the gene should restore pathogenicity [1]. Although targeted gene deletions are invaluable in determining the function of genes and pathways, there remain circumstances in which it is not possible to generate viable deletion mutants, or deletion of a gene encoding multiple functions precludes analysis of later functions. The latter is the case for the gene encoding Listeriolysin O (hly) of Listeria monocytogenes.
L. monocytogenes is a Gram-positive facultative intracellular pathogen that specifically replicates in the cytosol of host cells. In order to reach the host cell cytosol, L. monocytogenes must first escape from the phagocytic entry vacuole, which requires the secreted pore-forming cytolysin Listeriolysin O (LLO) [2,3]. In the cytosol, L. monocytogenes replicates and produces an actin nucleation factor (ActA) to move intracellularly and form protrusions that are engulfed by neighboring cells and resolved into double-membraned vacuoles. Again, LLO is required for escape from these secondary vacuoles [4,5].

Cre-lox Allows for Rapid Excision of hly during Infection of Macrophages
To study the role of LLO secreted in the cytosol during infection, we engineered a strain of L. monocytogenes, called hly fl , to excise the gene encoding LLO, hly, following escape of L. monocytogenes from the phagocytic vacuole. Specifically, loxP sites were inserted into the L. monocytogenes chromosome to flank hly and an adjacent gene, tetL, which provides tetracycline resistance. Cre recombinase, which mediates DNA recombination between loxP sites, was inserted into the chromosome using the pPL2 integrative vector, and expressed under the control of the L. monocytogenes actA promoter, which is relatively inactive prior to vacuolar escape of L. monocytogenes and becomes highly expressed in the cytosol ( Figure 1A). Thus, this strain is able to produce LLO initially to facilitate escape from the phagocytic vacuole but, once in the cytosol, hly is excised and LLO production ceases. To determine the efficiency of the system, BMMs were infected with hly fl L. monocytogenes and bacteria from the infected cells were recovered at different time points and plated on blood-agar media. Secreted LLO causes rapid β-hemolysis and L. monocytogenes colonies that secrete LLO can be easily identified ( Figure 1B). Prior to infection, hly fl L. monocytogenes were grown in broth containing tetracycline to select against low-level excision of hly and tetL. By 30 min post-infection almost 90% of recovered colony-forming units (CFU) were non-hemolytic ( Figure 1C). By 60 min post-infection, 98% of recovered CFU were non-hemolytic and by 90 min post-infection all colonies were non-hemolytic. Therefore, the excision of hly is rapid and complete during infection of BMMs.
Toxins 2020, 12, 38 3 of 14 initially to facilitate escape from the phagocytic vacuole but, once in the cytosol, hly is excised and LLO production ceases. To determine the efficiency of the system, BMMs were infected with hly fl L. monocytogenes and bacteria from the infected cells were recovered at different time points and plated on blood-agar media. Secreted LLO causes rapid β-hemolysis and L. monocytogenes colonies that secrete LLO can be easily identified ( Figure 1B). Prior to infection, hly fl L. monocytogenes were grown in broth containing tetracycline to select against low-level excision of hly and tetL. By 30 min postinfection almost 90% of recovered colony-forming units (CFU) were non-hemolytic ( Figure 1C). By 60 min post-infection, 98% of recovered CFU were non-hemolytic and by 90 min post-infection all colonies were non-hemolytic. Therefore, the excision of hly is rapid and complete during infection of BMMs.  To determine whether secretion of LLO by L. monocytogenes in the cytosol affects the growth of the bacteria in cells, intracellular growth was evaluated in BMMs. During the first five hours of infection, hly fl L. monocytogenes grew identically to WT L. monocytogenes and ΔactA, which is defective in actin-based motility and therefore defective in cell-to-cell spread ( Figure 2A). However, after five hours of infection, the growth of the strains diverged. Between five and twenty-four hours of To quantify the excision of hly, BMMs were infected with hly fl and both hemolytic and non-hemolytic colonies were enumerated by plating bacteria on blood-agar media at different timepoints. Mean and SD of data pooled from three independent experiments are shown.

LLO Secreted in the Cytosol Affects Intracellular Growth and Contributes to Cytotoxicity
To determine whether secretion of LLO by L. monocytogenes in the cytosol affects the growth of the bacteria in cells, intracellular growth was evaluated in BMMs. During the first five hours of infection, hly fl L. monocytogenes grew identically to WT L. monocytogenes and ∆actA, which is defective in actin-based motility and therefore defective in cell-to-cell spread ( Figure 2A). However, after five hours of infection, the growth of the strains diverged. Between five and twenty-four hours of infection, the number of WT L. monocytogenes plateaued and then declined, but the number of recovered hly fl L. monocytogenes increased to 10-fold more than the maximum of WT and remained elevated, suggesting that secretion of LLO in the cytosol negatively impacts growth of WT. infection, the number of WT L. monocytogenes plateaued and then declined, but the number of recovered hly fl L. monocytogenes increased to 10-fold more than the maximum of WT and remained elevated, suggesting that secretion of LLO in the cytosol negatively impacts growth of WT.  WT L. monocytogenes has the ability to spread to and replicate in neighboring cells. If hly fl has a defect in escape from secondary vacuoles that would limit its ability to replicate in neighboring cells, the difference in growth between WT and hly fl may reflect both the effects of LLO secreted in the cytosol and growth following cell-to-cell spread. To analyze the effects of LLO in the cytosol without complication by cell-to-cell spread, we compared growth of hly fl in a ∆actA background (∆actA hly fl ) to ∆actA, which is defective in actin-based motility and therefore defective in cell-to-cell spread. During the first eight hours of infection, ∆actA and ∆actA hly fl grew similarly to WT. However, between eight and twenty-four hours, the number of ∆actA bacteria decreased dramatically, whereas ∆actA hly fl decreased much less-having as much as 100-fold more bacteria than ∆actA. The rapid loss of ∆actA CFU, which is due to the influx of gentamicin [20], was partially rescued by deletion of hly in the cytosol, indicating that the decline of ∆actA in cells is partly LLO dependent.
The ability of LLO to form pores in cholesterol-containing cell membranes is well documented. We hypothesized that the growth of WT L. monocytogenes could be restricted by LLO-induced cytotoxicity because LLO has the potential to bind to the cell membrane and cause cell death. To quantify the amount of cell death caused by L. monocytogenes infection, cytotoxicity was measured by lactate dehydrogenase (LDH) release assay ( Figure 2B). After a 24 h infection of BMMs, 43% of cells were killed by WT L. monocytogenes infection. Only 22% of cells were killed by hly fl L. monocytogenes infection, indicating that LLO secreted in the cytosol contributes significantly to cytotoxicity during infection.
Previously, a requirement for LLO in escaping double-membraned phagocytic vacuoles after cell-to-cell spread was demonstrated using IPTG-inducible LLO [5]. Because hly fl should delete hly in the cytosol of the first cell it infects, we expected it to also be defective in escape from secondary vacuoles. To examine the fate of hly fl , we performed cell spreading and plaque assays to measure cell-to-cell spread ( Figure 2C-E). The number of cells in infectious foci were quantified for BMMs infected with WT and hly fl for 5 h. On average, WT L. monocytogenes spread to two to three neighboring cells over a 5 h period, while hly fl was only found in an average of two neighboring cells ( Figure 2C). Furthermore, WT L. monocytogenes was often well distributed in cells of an infectious focus, indicating that bacterial replication had occurred in neighboring cells following cell-to-cell spread. In contrast, infectious foci of hly fl were often observed with one macrophage harboring the majority of bacteria and limited numbers of bacteria in the surrounding cells, suggesting that hly fl spread to neighboring cells, but had a subsequent vacuolar escape defect ( Figure 2E). To assess the impact on cell-to-cell spread resulting from hly deletion over a longer time period, L2 cells were infected with WT and hly fl L. monocytogenes. WT L. monocytogenes forms plaques in a monolayer of L2 cells by spreading from cell-to-cell. Both hly fl and ∆hly L. monocytogenes were unable to form plaques, indicating that they cannot efficiently spread cell-to-cell. ∆hly that was complemented with hly fl but missing Cre recombinase, and therefore unable to excise hly, was restored to WT L. monocytogenes levels of plaque formation ( Figure 2D). These results suggest that though LLO is not required for growth in cells, its continued production contributes to cell-to-cell spread.

hly is Excised In Vivo and Its Excision Reduces Virulence
To quantify the efficiency of the hly fl Cre-lox system in vivo, C57BL/6J mice were infected intravenously with 10 5 CFU of hly fl L. monocytogenes ( Figure 3). Hemolytic capacity of the inoculum was verified by plating on blood agar. At 1, 2, 3, 5, 8, and 24 h post-infection, bacteria were recovered from the spleen and liver and plated on blood agar and both hemolytic and non-hemolytic CFU were enumerated. In the spleen ( Figure 3A), less than 20% of bacteria were hemolytic one hour post-infection. Hemolytic bacteria represented less than 2% of the population three hours post-infection and were nearly undetectable five hours post-infection. Then, 24 h post-infection, a small population of hemolytic bacteria were detected in the spleen (see discussion). In the liver ( Figure 3B), excision of hly was slower than in the spleen. One hour post-infection, only 35% of bacteria were non-hemolytic; eight hours post-infection 95% of bacteria were non-hemolytic and hemolytic colonies were undetectable by 24 h post-infection.  To determine the importance of LLO after escape of L. monocytogenes from the initial vacuole, we characterized hly fl L. monocytogenes using a mouse model of virulence. CD-1 mice were infected intravenously with 10 5 CFU of L. monocytogenes and 48 h post-infection, CFU from the spleen and liver were enumerated ( Figure 3). WT L. monocytogenes grew to 10 7 CFU in both the spleen and liver, while Δhly L. monocytogenes was extremely attenuated with bacteria from the spleen and liver near or below the limit of detection in most mice. Interestingly, hly fl L. monocytogenes had a moderate level of attenuation. 10 4 CFU were recovered from the spleen, representing a statistically-significant 3-log reduction in virulence compared to WT, although a smaller reduction in virulence than that of Δhly. In the liver, less than 10 3 CFU were recovered.

Vaccination with hly fl Confers Protective Immunity
The two requirements for a vaccine are safety and efficacy. In Figure 3, we showed that hly fl is highly attenuated, and thus satisfies the safety requirement. To test the efficacy of hly fl L. monocytogenes as a vaccine, a protection study was performed. Protection of hly fl was compared to protection conferred by ΔactA, which is well established as an attenuated and effective vaccine strain [21], and Δhly, which does not confer strong protection. C57BL/6J mice were vaccinated with either 10 3 or 10 5 L. monocytogenes. Four weeks post-vaccination, the mice were challenged with a lethal dose of WT L. monocytogenes. Three days post-challenge, CFU from the spleen were enumerated ( Figure  4A). Vaccination with 10 5 CFU of hly fl provided 5-logs of protection, albeit less protection than ΔactA, To determine the importance of LLO after escape of L. monocytogenes from the initial vacuole, we characterized hly fl L. monocytogenes using a mouse model of virulence. CD-1 mice were infected intravenously with 10 5 CFU of L. monocytogenes and 48 h post-infection, CFU from the spleen and liver were enumerated (Figure 3). WT L. monocytogenes grew to 10 7 CFU in both the spleen and liver, while ∆hly L. monocytogenes was extremely attenuated with bacteria from the spleen and liver near or below the limit of detection in most mice. Interestingly, hly fl L. monocytogenes had a moderate level of attenuation. 10 4 CFU were recovered from the spleen, representing a statistically-significant 3-log reduction in virulence compared to WT, although a smaller reduction in virulence than that of ∆hly. In the liver, less than 10 3 CFU were recovered.

Vaccination with hly fl Confers Protective Immunity
The two requirements for a vaccine are safety and efficacy. In Figure 3, we showed that hly fl is highly attenuated, and thus satisfies the safety requirement. To test the efficacy of hly fl L. monocytogenes as a vaccine, a protection study was performed. Protection of hly fl was compared to protection conferred by ∆actA, which is well established as an attenuated and effective vaccine strain [21], and ∆hly, which does not confer strong protection. C57BL/6J mice were vaccinated with either 10 3 or 10 5 L. monocytogenes. Four weeks post-vaccination, the mice were challenged with a lethal dose of WT L. monocytogenes. Three days post-challenge, CFU from the spleen were enumerated ( Figure 4A).
Vaccination with 10 5 CFU of hly fl provided 5-logs of protection, albeit less protection than ∆actA, while vaccination with 10 5 CFU of ∆hly did not confer protection. However, vaccination with 10 3 hly fl was not as protective as 10 3 ∆actA.
Toxins 2020, 12, 38 7 of 14 while vaccination with 10 5 CFU of Δhly did not confer protection. However, vaccination with 10 3 hly fl was not as protective as 10 3 ΔactA. As vaccination with vacuole-confined Δhly is inhibited by the secretion of the immunosuppressive cytokine IL-10 [22], we hypothesized that hly fl is also inhibited by IL-10 because it is defective in cell-to-cell spread and remains confined in secondary vacuoles. To test this hypothesis, we performed protection studies using an IL-10 receptor blocking antibody (αIL-10R), which improves the protective capacity of Δhly [22]. Indeed, administration of αIL-10R improved the protective capacity of Δhly and hly fl but not ΔactA L. monocytogenes ( Figure 4B). Furthermore, the protection conferred by hly fl was improved to levels similar to those conferred by vaccination with ΔactA. Thus, hly fl is a highly attenuated strain of L. monocytogenes capable of inducing protective immunity, though its protection is reduced compared to ΔactA likely due to IL-10 induction.

Discussion
As a member of the cholesterol-dependent cytolysin (CDC) family of pore-forming toxins, LLO is unique as the only CDC that is secreted by an intracellular pathogen and therefore the only CDC that primarily acts on cells from within. Therefore, LLO likely affects cells in ways that extracellular CDCs do not. Yet, because deletion of hly prevents study of the effects of LLO in the cytosol, other studies have utilized the application of exogenous purified LLO to study its effects [15][16][17][18][19]. Here, we described a L. monocytogenes strain that uses a Cre-lox system to delete hly following vacuolar escape. This strain, hly fl , became 100% non-hemolytic less than 1.5 h after infection of macrophages ( Figure  1C) and replicated to high numbers in the cytosol (Figure 2A,E). hly excision after vacuolar escape allows the study of LLO functions in the cytosol separate from its role in vacuolar escape.

Insights into the Effects of LLO during Infection
LLO is a pore-forming toxin that oligomerizes and forms pores in cholesterol-containing membranes, including the host cell vacuolar and plasma membranes [23,24]. Secretion of LLO in the cytosol has the potential to damage the host cell plasma membrane. Although multiple mechanisms As vaccination with vacuole-confined ∆hly is inhibited by the secretion of the immunosuppressive cytokine IL-10 [22], we hypothesized that hly fl is also inhibited by IL-10 because it is defective in cell-to-cell spread and remains confined in secondary vacuoles. To test this hypothesis, we performed protection studies using an IL-10 receptor blocking antibody (αIL-10R), which improves the protective capacity of ∆hly [22]. Indeed, administration of αIL-10R improved the protective capacity of ∆hly and hly fl but not ∆actA L. monocytogenes ( Figure 4B). Furthermore, the protection conferred by hly fl was improved to levels similar to those conferred by vaccination with ∆actA. Thus, hly fl is a highly attenuated strain of L. monocytogenes capable of inducing protective immunity, though its protection is reduced compared to ∆actA likely due to IL-10 induction.

Discussion
As a member of the cholesterol-dependent cytolysin (CDC) family of pore-forming toxins, LLO is unique as the only CDC that is secreted by an intracellular pathogen and therefore the only CDC that primarily acts on cells from within. Therefore, LLO likely affects cells in ways that extracellular CDCs do not. Yet, because deletion of hly prevents study of the effects of LLO in the cytosol, other studies have utilized the application of exogenous purified LLO to study its effects [15][16][17][18][19]. Here, we described a L. monocytogenes strain that uses a Cre-lox system to delete hly following vacuolar escape. This strain, hly fl , became 100% non-hemolytic less than 1.5 h after infection of macrophages ( Figure 1C) and replicated to high numbers in the cytosol (Figure 2A,E). hly excision after vacuolar escape allows the study of LLO functions in the cytosol separate from its role in vacuolar escape.

Insights into the Effects of LLO during Infection
LLO is a pore-forming toxin that oligomerizes and forms pores in cholesterol-containing membranes, including the host cell vacuolar and plasma membranes [23,24]. Secretion of LLO in the cytosol has the potential to damage the host cell plasma membrane. Although multiple mechanisms limit LLO damage to the cell plasma membrane, it has been difficult to establish whether these mechanisms are entirely effective in preventing cell death [7]. Here, we showed that L. monocytogenes that does not produce LLO in the cytosol replicates to greater numbers in macrophages and is less cytotoxic. The revelation that LLO secreted in the cytosol is cytotoxic makes it curious that L. monocytogenes continuously secretes LLO. It is possible that the continuous production of LLO is necessary to ensure a rapid escape from secondary vacuoles, as defects that reduce cell-to-cell spread are highly attenuated.
It is also possible that the innate immune response to LLO-induced cell death contributes to pathogenesis. In the host cell cytosol, infrequent lysis of L. monocytogenes induces pyroptotic cell death via the AIM2 inflammasome [25]. Some forms of cell death are thought to inhibit the generation of protective immunity. Strains of L. monocytogenes engineered to induce pyroptosis, necrosis or apoptosis inhibit the generation of protective immunity [26]. Thus, it is likely that LLO-induced cell death also affects the immune response to L. monocytogenes infection. The type of cell death induced by LLO and its effects on the innate and adaptive immune responses to L. monocytogenes merit further study.
Analysis of the kinetics of hly deletion, specifically the more rapid deletion in the spleen compared to the liver, revealed differences in the environments experienced by L. monocytogenes. CD169+ macrophages that are localized to the marginal zone of the spleen and dendritic cells are thought to be the first splenic cell types infected by L. monocytogenes upon intravenous inoculation [27,28]. Three hours post-infection, the majority of L. monocytogenes in the spleen are trapped within CD169+ macrophages [27]. Following infection with hly fl , bacteria in the spleen became non-hemolytic at a rate similar to in BMMs, with almost complete loss of hemolytic capacity three hours post-infection, suggesting that infection of BMMs closely models infection in the spleen with respect to activation of virulence genes. However, bacteria became non-hemolytic at a much slower rate in the liver, with greater than 20% of bacteria remaining hemolytic at 5 h post-infection and complete loss of hemolysis between eight and 24 h post-infection. In the liver, four hours after infection with L. monocytogenes 100% of infected cells are tissue-resident macrophages, known as Kupffer cells [29]. Bacteria that are not killed by the Kupffer cells can transfer to hepatocytes, which become heavily infected [30]. Over the next couple days, infected Kupffer cells die by necroptosis; infiltrating neutrophils and monocyte-derived macrophages lyse infected hepatocytes and become the primarily infected cells [30][31][32]. It is possible that hly fl became non-hemolytic slowly in the liver because actA expression was not efficiently upregulated in Kupffer cells, and Cre-lox recombination only occurred after the bacteria were transferred to hepatocytes, neutrophils and/or monocyte-derived macrophages. If this is the case, it would be interesting to understand why the intracellular environment of a Kupffer cell does not activate actA expression like other cells. Alternatively, the bacteria that became non-hemolytic during the first 8 h of infection could represent the population of bacteria inside Kupffer cells, hepatocytes and/or infiltrating neutrophils and monocyte-derived macrophages, and the bacteria that remained hemolytic may represent an extracellular population of bacteria in the liver. A small population of extracellular bacteria associated with nonparenchymal cells in the liver six hours after infection has been previously identified, though it is not clear whether these bacteria became extracellular following lysis of infected hepatocytes, or whether they never infected cells [33].

Limitations of hly fl
Although we have successfully employed hly fl to demonstrate the contribution of LLO to cytotoxicity in macrophages, this system is limited by the fact that loss of LLO is permanent and not conditional to the environment of the cytosol. As a result, hly fl became trapped in secondary vacuoles and was defective in cell-to-cell spread ( Figure 2C-E). Bacteria that are released from the cytosol upon cell lysis also behave like LLO-minus mutants. The inability to escape subsequent vacuoles likely explains the attenuation of hly fl in mice ( Figure 3C,D), and why hly fl vaccination was improved by αIL-10R antibody ( Figure 4B).
An additional complication of hly fl is that Cre-lox recombination is susceptible to inactivation. A previous study in which a transposon library was generated in a strain of L. monocytogenes with Cre-lox identified transposon insertions in the actA promoter driving cre expression and loxP sites that prevented recombination. We observed a population of hemolytic bacteria in the spleen that expanded between eight-and 24 h post-infection. We isolated several colonies of hemolytic bacteria 24 h post-infection and reinfected BMMs, and bacteria were 100% hemolytic five hours post-infection (data not shown). It is possible that these hemolytic bacteria are the progeny of a founding bacterium that had a mutation in its Cre-lox machinery that prevented recombination from occurring.
In the future, the ideal tool to study the cytosolic effects of LLO in mice would not secrete LLO in the host cell cytosol but could secrete LLO upon entry into secondary cells to continue the life cycle. Nevertheless, we believe hly fl is well suited for studying LLO secreted in the cytosol in cells, and insights gained from study of hly fl in cells can be translated to the whole animal setting using various mouse models.

Conclusions
Cre-lox recombination has been a popular tool for the study of plants and mice for many decades [34][35][36]. In bacteria, its use has been more limited, and it has only been used a few times in L. monocytogenes. Previously in L. monocytogenes, Cre-lox was used to generate a strain that cannot replicate following activation of the actA promoter by flanking essential genes near the origin of replication with loxP sites and driving Cre expression with the actA promoter [37]. This strain is highly attenuated and potently activates the CD8+ T-cell response and thus is a candidate vaccine-delivery system [38]. In another instance, a strain of L. monocytogenes that deletes actA in the host cell cytosol was used to show that ActA expressed in the host cell cytosol contributes to cell-to-cell spread and simultaneously allows L. monocytogenes to avoid xenophagy [39]. This works represents the first use of Cre-lox recombination to study the function of a virulence factor that is active at temporally and spatially distinct periods. We believe that this system has the promise to uncover many effects of LLO secreted in the cytosol and could also uniquely contribute to better understanding the cellular responses to membrane damage from an intracellularly secreted pore-forming toxin.

Construction of hly fl
hly fl was constructed by integrating two plasmids, one encoding hly and tetL flanked by loxP sites and the other encoding cre downstream of the actA promoter, into a ∆hly strain of L. monocytogenes. The hly and tetL genes were cloned into pPL1 and flanked by lox66/lox71 loxP sites such that Cre expression resulted in excision of the region flanked by loxP sites (fl). The resulting plasmid (pPL1-hly fl ) was transformed into SM10 E. coli. The cre recombinase gene was previously engineered downstream of the actA promoter in pPL2e, yielding the plasmid pPL2e-actA-cre, which was also transformed into SM10 E. Coli. Transconjugation was performed to integrate both plasmids into ∆hly L. monocytogenes in a stepwise manner. First, pPL1-hly fl was transconjugated with ∆hly L. monocytogenes and transconjugate colonies that were resistant to streptomycin (200 µg/mL) and chloramphenicol (7.5 µg/mL) were selected. Second, ∆hly pPL1-hlyfl was transconjugated with pPL2e-actA-cre, and transconjugate colonies resistant to streptomycin, chloramphenicol, and erythroymycin (1 µg/mL) were selected. Similarly, ∆actA hly fl was engineered by transconjugating pPL1-hly fl and pPL2e-actA-cre into ∆actA∆hly. The hly fl complement strain was engineered by transconjugating ∆hly with pPL1-hly fl .
A control pPL1 plasmid (pPL1-tetL fl ) encoding tetL, but not hly, flanked by loxP sites was engineered by excising hly from pPL1-hly fl . The ∆hly control strain was engineered by transconjugating ∆hly with pPL1-tetL. The WT control strain was engineered by transconjugating WT with pPL1-tetL fl . The ∆actA control strain was engineered by transconjugating ∆actA with pPL1-tetL fl .

Bacterial Culture
Strains used in this study are listed in Table S1. Bacteria were grown overnight at 37 • C in Brain-Heart Infusion (BHI; BD, Sparks, MD, USA) containing 200 µg/mL streptomycin (GoldBio, St. Louis, MO, USA), and bacteria with tetL were additionally grown in 2 µg/mL tetracycline (GoldBio, St. Louis, MO, USA). Overnight cultures were diluted 1:200 and grown in BHI containing streptomycin (for bacteria without tetL) or streptomycin and tetracycline (for bacteria with tetL) at 37 • C, shaking, to an optical density of 0.5. These cultures were then pelleted by centrifugation and resuspended in phosphate-buffered solution (PBS; Gibco, Paisley, UK) containing 9% glycerol. These cultures were then aliquoted and frozen at −80 • C. Aliquots were thawed and used directly for experiments.

Preparation of M-CSF
The 3T3 cell media was prepared using Dulbecco's Modified Eagle Media (DMEM, Gibco, Grand Island, NY, USA) with 10% FBS (Seradigm, US Origin), 1% L-Glutamine (Corning, Manassas, VA, USA), and 1% Sodium pyruvate (Corning, Manassas, VA, USA), with or without 1× Penicillin Streptomycin Solution ("Pen/Strep"; Corning, Manassas, VA, USA). The 10 7 M-CSF-producing 3T3 cells were seeded into a T75 flask with 20mL media containing Pen/Strep and grown at 37 • C 5% CO 2 (Day 1). To split cells, media was aspirated, cells were washed with warm PBS, and incubated with 0.05% Trypsin-EDTA (Gibco, Grand Island, NY, USA) for five minutes at 37 • C. On day 4, cells were split to a T225 in 50 mL media containing Pen/Strep. On day 7, cells were split to five T225 flasks in media without Pen/Strep. On day 9 or 10, when cells when cells were 100% confluent, the five T225 flasks were split into 25 T225 flasks and grown until 100% confluent (about 3 days) and an additional two days (about 5 days total) in media without Pen/Strep. On day 14 or 15, supernatant was removed from all flasks, filter sterilized with a 0.2 µM bottle filter, and stored at 4 • C. Then, 50 mL fresh media without Pen/Strep was added back to each T225 and flasks were incubated an additional 3 days. On day 17 or 18, supernatants were collected as before, and combined with the previous supernatants. Supernatants were stored at −20 • C and used as the source of M-CSF for bone marrow-derived macrophage preparation and culture.

Bone Marrow-Derived Macrophage Culture
BMM growth media was prepared using high glucose DMEM (Thermo Fisher Scientific) with 20% Fetal Bovine Serum (Seradigm), 1% L-glutamine (Corning), 1% Sodium pyruvate (Corning), 14mM 2-Mercaptoethanol (Gibco; Grand Island, NY, USA), and 10% 3T3 cell supernatant (from M-CSF-producing 3T3 cells). Macrophages were prepared from the femurs of C57BL/6J mice. Femurs were isolated, sterilized with 70% ethanol, and crushed with a mortar and pestle in BMM growth media. Cells were strained through a 70µM filter and distributed into ten 150-mm non-TC dishes in 30mL BMM culture medium. An additional 30mL BMM culture medium was added at day 3. After cells were incubated for a total of seven days at 37 • C with 5% CO 2 , cells were harvested and frozen at −80 • C in BMM culture medium with 10% Fetal Bovine Serum (Seradigm) and 10% DMSO (Sigma, St. Louis, MO, USA) added.

Hemolysis in BMMs
A 60 mm non-TC dish with 15 12 mm glass coverslips was seeded with 3 × 10 6 BMMs. The following day, the cells were infected with 2 × 10 5 CFU hly fl for 30 min. Then, cells were washed with PBS and BMM media with 50 µg/mL gentamicin was added. At each timepoint, coverslips were removed from the dish and placed in water to lyse the cells. Bacteria were plated on blood-agar media. Plates were incubated overnight at 37 • C and then transferred to 4 • C until halos surrounding hemolytic colonies were clear. Hemolytic and ahemolytic colonies were enumerated.

Hemolysis in Mice
C57BL/6J (The Jackson Laboratory, Bar Harbor, ME, USA) mice were infected with 1 × 10 5 CFU of hly fl L. monocytogenes. At 0, 1, 2, 3, 5, 8, and 24 h post-infection mice were euthanized (3 mice per timepoint) and spleens and livers were harvested, homogenized, and plated on blood-agar media. Plates were incubated overnight at 37 • C and then transferred to 4 • C until halos surrounding hemolytic colonies were clear. Hemolytic and ahemolytic colonies were enumerated.

Lactate Dehydrogenase (LDH) Assay
BMMs were seeded into a 24-well plate with 5 × 10 5 BMMs/well. The following day, cells were infected with 2 × 10 6 CFU L. monocytogenes for 30 min. Then, cells were washed with PBS and BMM media containing 5% FBS and 50µg/mL gentamicin was added to wells. Then, 24 h post-infection LDH assay was performed as previously described [25].

Intracellular Growth Curves
In total, 3 × 10 6 BMMs were plated in 60 mm non-TC-treated Petri dishes with 15 12 mm glass coverslips in each dish. The following day, two dishes per strain were infected with 5 × 10 5 CFU (MOI = 0.17) and intracellular growth curves were performed as described previously [40].

Plaque Assay
Six-well plates were seeded with 1.2 × 10 6 L2 cells per well. The plaque assay was performed as previously described [41].

Cell Spreading Assay
BMMs were seeded into a 24-well plate with glass coverslips in each well, at 3.5 × 10 5 cells/well. Three coverslips were infected for each strain. Then, 30 min post-infection, cells were washed with PBS and BMM media containing 50 µg/mL gentamicin (Sigma-Aldrich, St. Louis, MO, USA) was added. At five hours post-infection, coverslips were removed, and cells were fixed with 100% methanol and stained using Diff-Quik stain. Coverslips were mounted onto glass slides using Permount (Fisher Chemical). Light microscopy was performed using a BZ-X700 microscope (KEYENCE, Osaka, Japan) and a 60× objective lens. Cell spread index was calculated by counting the number of cells in an infectious focus containing 5 or more bacteria and subtracting one cell so that the cell spread index represents the number of cells that were spread to from the initially infected cell. Analysis was performed blindly. In total, 60 hly fl and 67 WT infectious foci were analyzed.

Animal Use Ethics Statement
All animal work was done in strict accordance with university regulations. Protocols were reviewed and approved by the Animal Care and Use Committee at the University of California, Berkeley AUP-2016-05-8811. Date of approval: 11 February 2016.

Virulence in Mice
Eight-week-old female CD-1 mice (Charles River Laboratories, Wilmington, MA, USA) were infected intravenously with 1 × 10 5 CFU in 200 µL PBS. Forty-eight hours post-infection, the mice were euthanized, and spleens and livers were harvested, homogenized in 0.1% IGEPAL CA-630 (Sigma, St. Louis, MO, USA) in water, and plated on LB agar (Fisher, Fair Lawn, NJ, USA) with 200 µg/mL streptomycin for enumeration of bacterial burdens.

Vaccination of Mice
Eight-to-ten-week-old female C57BL/6J mice (The Jackson Laboratory) were vaccinated by intravenous injection of L. monocytogenes in 200 µL PBS. Four weeks post-vaccination, mice were challenged with 5 × 10 4 CFU WT L. monocytogenes injected intravenously in 200 µL PBS. Three days post-challenge, mice were euthanized, and spleens and livers were harvested, homogenized in 0.1% IGEPAL CA-630 (Sigma), and plated on LB Agar with 200 µg/mL streptomycin for enumeration of bacterial burdens. Mice treated with αIL-10R antibody (Clone 1B1.3A, Bio X Cell, West Lebanon, NH, USA) were injected with 250 µg of antibody in 100 µL PBS two hours prior to vaccination.