A protective role for type I interferon signaling following infection with Mycobacterium tuberculosis carrying the rifampicin drug resistance-conferring RpoB mutation H445Y

Interleukin-1 (IL-1) signaling is essential for controlling virulent Mycobacterium tuberculosis (Mtb) infection since antagonism of this pathway leads to exacerbated pathology and increased susceptibility. In contrast, the triggering of type I interferon (IFN) signaling is associated with the progression of tuberculosis (TB) disease and linked with negative regulation of IL-1 signaling. However, mice lacking IL-1 signaling can control Mtb infection if infected with an Mtb strain carrying the rifampin-resistance conferring mutation H445Y in its RNA polymerase β subunit (rpoB-H445Y Mtb). The mechanisms that govern protection in the absence of IL-1 signaling during rpoB-H445Y Mtb infection are unknown. In this study, we show that in the absence of IL-1 signaling, type I IFN signaling controls rpoB-H445Y Mtb replication, lung pathology, and excessive myeloid cell infiltration. Additionally, type I IFN is produced predominantly by monocytes and recruited macrophages and acts on LysM-expressing cells to drive protection through nitric oxide (NO) production to restrict intracellular rpoB-H445Y Mtb. These findings reveal an unexpected protective role for type I IFN signaling in compensating for deficiencies in IL-1 pathways during rpoB-H445Y Mtb infection.


Introduction
Mycobacterium tuberculosis (Mtb) is estimated to cause 10 million active cases of tuberculosis (TB) annually [1].Global dissemination of drug resistance to front and second-line antibiotics limits efforts to contain and combat this pervasive disease [2][3][4][5][6][7].Alternatively, to avoid the toxicity associated with anti-TB chemotherapies, researchers are developing host-directed therapeutics to better augment the host immune response to Mtb infection and effective vaccines that protect against TB disease.These strategies require a thorough understanding of the host-pathogen interactions and the immune events correlated with protective responses to Mtb infection.
Substantial efforts have been made to understand important immune cell populations and signaling pathways necessary for host protection and survival during Mtb infection [8][9][10][11][12][13][14].Interleukin-1 (IL-1) signaling is essential for protection during Mtb infection, as hosts with defective IL-1 signaling experience immunopathology and succumb to infection [15,16].Conversely, type I interferon (IFN) signaling is thought to be detrimental during Mtb infection through antagonism of IL-1 signaling and induction of maladaptive inflammatory processes [17][18][19].In humans, type I IFN signaling is associated with TB disease progression from controlled Mtb infection [20][21][22][23].We previously found that the presence of a drug resistance-conferring mutation, rpoB-H445Y, can alter host-pathogen interactions such that IL-1 signaling is not necessary to control drug-resistant (rpoB-H445Y) Mtb infection [7,24].Additionally, cytokine production is altered during infection of macrophages with rpoB-H445Y Mtb in vitro and in vivo [25], with heightened type I IFN and limited IL-1 production.The rpoB-H445Y Mtb strain contained an increased abundance of the long-chained lipid phthiocerol dimycocerosate (PDIM) in its cell wall.Treatment of wildtype (wt) Mtb-infected BMDMs with purified PDIM from rpoB-H445Y Mtb was sufficient to drive altered cytokine production [24].However, we have not determined the IL-1-independent protective pathways that led to the control of this rpoB-H445Y Mtb infection.
In the current study, we demonstrate that, in contrast to what is seen with wild type (wt) Mtb infection, type I IFN signaling has a protective role during rpoB-H445Y Mtb infection.While a global loss of either IL-1 or type I IFN signaling alone did not affect rpoB-H445Y Mtb replication in vivo, mice with deficient IL-1 signaling were susceptible to rpoB-H445Y Mtb infection when they also lacked type I IFN signaling.Increased rpoB-H445Y Mtb burden was associated with an increase in inflammatory cell influx and inflammation.Furthermore, recruited myeloid cells, such as monocytes and macrophages, were the source of type I IFN during rpoB-H445Y Mtb infection.Interestingly, the role of type I IFN signaling in host protection depended upon cell type, such that it was protective in LysM + cells but detrimental in CD11c + cells.In vitro experiments suggested that both type I IFN and IL-1 signaling led to nitric oxide (NO) production, which limited rpoB-H445Y Mtb growth.Our results highlight the complex role of type I IFN signaling and that host protective mechanisms can be markedly different during infection with two closely related bacterial strains.Insights from this work could be used to suggest potential targets for future avenues of study in the development of host-directed therapies against drug-resistant Mtb.

Type I IFN signaling mediates protection in the absence of IL-1 signaling in rpoB-H445Y Mtb infection
IL-1 signaling is essential in controlling wt Mtb infection [15,18].However, this signaling pathway is dispensable when mice are infected with rpoB-H445Y Mtb [24], a drug-resistant but non-isogenic Mtb strain derived from wt Mtb [25], suggesting that there are alternate signaling pathways mediating protection.To determine if type I IFN signaling was protective in the absence of IL-1 signaling against rpoB-H445Y Mtb in vivo, we generated mice lacking both IL-1 and type I IFN receptors and evaluated their susceptibility to wt or rpoB-H445Y Mtb infection.As expected, Il1r1-/-mice were unable to control wt Mtb infection, with elevated lung burden at 14-and 30-days post-infection (dpi), relative to C57BL/6 (B6) mice (Fig 1A), whereas loss of Ifnar1 alone has no impact on Mtb burden.As before, following wt Mtb infection [18,20], the increase in lung burden caused by loss of Il1r1 was reversed in mice lacking both Ifnar1 and Il1r1 [double knockout (DKO)] (Fig 1A).An increase in the susceptibility of Il1r1 -/-mice was accompanied by an increase in inflammatory damage in the lungs of wt Mtb infected (S1A Fig), but this inflammatory damage did not significantly decrease in the DKO mice where Mtb control was evident.We also observed increased numbers of recruited macrophages (RMs) expressing high levels of Major Histocompatibility Complex Class II (MHCII hi ), a surrogate marker for activation, of neutrophils, and of MHCII hi monocytes in the lungs of wt Mtb-infected Il1r1 -/-mice (S1B Fig) .In the infected DKO mice where susceptibility was reversed, only the numbers of neutrophils were significantly decreased.We found no differences in the numbers of cytokine-producing, activated (CD44 hi ) T lymphocyte populations between the infected groups (S1C and S1D Fig) .A different pattern of susceptibility was apparent when we infected the mice with rpoB-H445Y Mtb.We observed higher bacterial burdens only in the DKO mice at 30 dpi (Fig 1B), while loss of either IL1 or type I IFN signaling alone did not impact lung rpoB-H445Y Mtb burden.The increased bacterial load in the DKO rpoB-H445Y Mtb-infected lungs also coincided with increased lung inflammation (Fig 1C) and increased numbers of lung MHCII hi RMs and MHCII hi myeloid dendritic cells (mDCs) in the DKO mice at 30 dpi (Fig 1D).However, there were no differences in the numbers of neutrophils, MHCII hi monocytes, MHCII hi alveolar macrophages (AMs), and cytokine-producing, CD44 hi T lymphocyte populations between the DKO mice and B6 and Il1r1 -/-mice (S2A-S2E Fig) .As previously noted [25], loss of Ifnar1 alone was sufficient to drive an increase in the numbers of these cellular populations and the amount of inflammation during rpoB-H445Y Mtb infection.Thus, while type I IFN signaling is dispensable globally for control of wt and rpoB-H445Y Mtb infection, in the absence of IL-1 signaling, type I IFN signaling is necessary for immunity against rpoB-H445Y Mtb.

Similar type I IFN transcriptional pathways are induced during wt and rpoB-H445Y Mtb infections but result in divergent infection outcomes
Type I IFN signaling has been thought to be detrimental during wt Mtb infection based upon its association with the progression of TB patients towards active disease and its ability to antagonize the protective IL-1 signaling pathway [18,20,22].To discern the roles of type I IFN signaling during both Mtb infections, we determined the lung transcriptional landscape in response to either infection at 14 days post-infection, an early timepoint before significant differences in lung rpoB-H445Y Mtb burden were apparent in the gene-deficient mice.We identified the most differentially expressed genes (DEGs) across groups and categorized them by pathways (Fig 2).In the lungs of mice infected with wt Mtb, genes related to IFN signaling were upregulated in the B6 and Il1r1 -/-mice compared with Ifnar1 -/-and DKO mice (Fig 2A).Specific analysis of the type I IFN transcriptional signature revealed downregulation of genes in this pathway in the Ifnar1 -/-and DKO mice when compared to the B6 and Il1r1 -/-mice during wt Mtb infection (S3A Fig) .Additionally, genes involved in antigen presentation via MHC I, such as H2-q4 and H2-t22, were induced by type I IFN signaling during wt Mtb infection at this early timepoint.We also observed induction of genes involved in branched chain amino acid catabolism (Aldh6a1 and Ivd), phagocytosis [26] (Snap23 and Stx4a), endoplasmic reticulum (ER) stress [27,28] (Hspa5 and Calr), and cellular response to heat stress (Hsp90aa1 and St13) specifically in the B6 mice infected with wt Mtb relative to the single knockout and DKO mice (Fig 2A).We observed enrichment of DEGs in pathways involved in type I IFN signaling during wt Mtb infection in both B6-and Il1r1 -/-infected mouse groups, although they differed in their ability to control infection.Together, our data suggest that decreased co-expression of type I IFN-induced genes with genes involved in metabolism and stress could drive detrimental does not determine the outcomes of wt Mtb infection.
In the lungs of mice infected with rpoB-H445Y Mtb, we also observed that DEGs related to IFN signaling were upregulated in B6 and Il1r1 -/-mice, whereas these pathways were downregulated in the Ifnar1 -/-and DKO mice (Fig 2B).Indeed, the type I IFN transcriptional signature was downregulated in both the Ifnar1 -/-and DKO mouse groups when compared with either wt or Il1r1 -/-mice following rpoB-H445Y Mtb infection (S3B Fig) .Interestingly, expression of some interferon-stimulated genes (ISGs), such as Abce1 and Ptpn11, was higher in the lungs of Il1r1 -/-mice than in B6 mice after rpoB-H445Y Mtb infection.Another IFN-related gene, Ube2l6, decreased in expression in the rpoB-H445Y Mtb infected Il1r1 -/-mice lungs relative to infected B6 counterparts (Fig 2B).These findings suggest that IL-1 signaling plays a complex role in the regulation of IFN-related genes.Coincident with the upregulation of IFNrelated genes in B6 and Il1r1 -/-rpoB-H445Y Mtb-infected mice, genes involved in metabolic pathways, such as fatty acid metabolism and cholesterol biosynthesis were upregulated.In C57Bl/6 (B6), Ifnar1 -/-, Il1r1 -/-, and Ifnar -/-Il1r1 -/-[double knockout (DKO)] mice were aerosol infected with a low dose of wt or rpoB-H445Y Mtb and sacrificed at 14 dpi.RNA was extracted from homogenized lung tissue and sequenced.Heatmaps displaying average expression levels of top differentially expressed genes (DEGs) across mouse genotypes after A) wt or B) rpoB-H445Y Mtb infection with annotated pathways (four to six mice per group).Significant DEGs related to the corresponding reactome pathways are visualized using heatmaps(|log 2 FC| >0, adjusted p-value < 0.05).Average expression (normalized) of selected genes on log 2 scale are shown.https://doi.org/10.1371/journal.ppat.1012137.g002contrast, mice lacking IL-1 signaling exhibited decreased expression of genes, such as Slc2a3 and Slc4a1, that are part of the solute carrier (SLC)-mediated transmembrane transport family.Therefore, our data indicate that the loss of rpoB-H445Y Mtb control observed in DKO mice was associated with the downregulation of various cellular processes that could be triggered by either type I IFN or IL-1 signaling.Many IFN-related genes, such as Isg15, Ifnar1, and Mx2, were induced following either wt or rpoB-H445Y Mtb infection.In contrast, genes such as Abce1 and Ptpn11 were only upregulated following rpoB-H445Y Mtb infection, whereas Uba7 was upregulated only during wt Mtb infection.Furthermore, Ube2l6, which exhibited differential regulation in Il1r1 -/-and B6 mice during rpoB-H445Y Mtb infection, was upregulated in both groups of mice during wt Mtb infection.So, while we observed broad induction of many IFN signaling-related genes during both Mtb infections, some IFN-related genes were differentially induced across wt and rpoB-H445Y Mtb infections.

CD11b + myeloid cells produce type I IFN, while LysM + cells respond to type I IFN following rpoB-H445Y Mtb infection
Type I IFN is rapidly produced by many cell types in response to infection [21,29].We used IFNβ-YFP reporter mouse strain, in which cells that produce IFN β also co-express YFP and are readily detectable [30][31][32] (Fig 3A ), and found that type I IFN production was induced in various myeloid lung populations following infection with rpoB-H445Y Mtb (Fig 3B).As early as 7 dpi, rpoB-H445Y Mtb infection lead to significantly higher accumulation of IFN-producing lung monocytes and RMs as compared with uninfected controls.No discernable production of type I IFN in AMs, neutrophils, or any non-immune (CD45 -) population was detected at this timepoint.Plasmacytoid dendritic cells (pDCs), a major producer of type I IFN during bacterial or viral infection [14,32], were not readily detected in our analysis.We then tracked type I IFN production in rpoB-H445Y Mtb infection at later timepoints (Fig 3B).Lung monocytes were also the primary cell type expressing type I IFN production at 14 dpi, and RMs were the major cellular population that was producing type I IFNs at 28 dpi.Taken together, our data indicate that the main sources of type I IFN production in rpoB-H445Y Mtb infection are CD11b + myeloid cells, namely monocytes and RMs.
Loss of control of rpoB-H445Y Mtb infection in mice lacking type I IFN and IL-1 signaling was accompanied by an influx of inflammatory myeloid cells, and the primary producers of type I IFN throughout rpoB-H445Y Mtb infection were myeloid cells.To test whether myeloid cells were also responding to type I IFN in rpoB-H445Y Mtb infection in the absence of IL-1 signaling, we generated LysM-Cre Ifnar1 fl/fl mice (deletion of the type I IFN receptor on LysMexpressing myeloid cells [33]), infected these mice with rpoB-H445Y Mtb, and also treated them with IL-1 blocking antibodies or isotype controls (Fig 3C).As we observed in the rpoB-H445Y Mtb-infected Il1r1 -/-mice, blocking of IL-1 signaling with antibody did not impact lung bacterial burden in the Ifnar1  H445Y Mtb infection relative to littermate controls.Thus, during rpoB-H445Y Mtb infection, type I IFN signaling is protective in LysM-expressing cells, especially in the absence of IL-1 signaling, but detrimental in CD11c-expressing cells.

NO production restricts intracellular rpoB-H445Y Mtb growth in murine macrophages
To address the importance of macrophages in producing and responding to type I IFN signals, we utilized bone marrow-derived macrophages (BMDMs) to elucidate the mechanistic role of type I IFN signaling in the absence of IL-1 signaling during rpoB-H445Y Mtb infection.As before [24], BMDMs lacking IL-1 signaling were more susceptible to wt Mtb infection at 6 dpi (  necrotic (AnnexinV hi 7AAD hi ) stages.While wt Mtb drove high levels of necrotic cell death early in infection, the rpoB Mtb mutant induced high levels of necrosis later during infection.Similarly, high levels of apoptosis were observed in Il1r1 -/-BMDMs only later during rpoB mutant Mtb infection.As both infections induced high levels of cell death in the Il1r1 -/- BMDMs, but not in the DKO BMDMs, these responses could be a general response to Mtb infection that is mediated by type I IFNs [34].While these increased frequency of cell death in Il1r1 -/-BMDMs coincided with their inability to control wt Mtb infection, the same does not appear to be the case during rpoB Mtb mutant infection, suggesting a role for other host cellular processes to mediate protection against rpoB-H445Y Mtb.
Type I IFN signaling can drive nitric oxide (NO) production as part of its anti-bacterial functions [29].Thus, we assessed the abundance of nitrite, the byproduct of NO generation, in the supernatants of wt or rpoB-H445Y Mtb infected BMDMs (Fig 4C).BMDMs from mice with functional type I IFN and IL-1 signaling produced increased NO in response to rpoB-H445Y Mtb infection when compared with other infected BMDMs.While there was a reduction in NO production in rpoB-H445Y Mtb infected BMDMs lacking either IL-1 or type I IFN signaling relative to infected B6 BMDMs, BMDMs from these two groups still produced more NO than their infected DKO counterparts.The stepwise decrease in NO production and the stepwise increase in CFU between B6 BMDMs to Ifnar1 -/-and Il1r1 -/-BMDMs to DKO BMDMs support the idea that NO production is associated with protection against rpoB-H445Y Mtb infection.To assess the role of NO in controlling rpoB-H445Y Mtb infection, we used carboxy-PTIO (C-PTIO), an NO scavenger, at varying concentrations (Fig 4D).Relative to untreated controls, the addition of 100 uM C-PTIO resulted in a loss of control in B6 BMDMs during rpoB-H445Y Mtb infection.In contrast, the addition of 1 uM and 10 uM of C-PTIO was sufficient to increase the susceptibility of Ifnar1 -/-and Il1r1 -/-BMDMs, respectively, while adding any amount of C-PTIO did not impact the course of infection in DKO BMDMs.This "minimum susceptibility" analysis suggests a stoichiometric role for NO in limiting rpoB-H445Y Mtb during infection.To more specifically assess the role of NO production in controlling rpoB-H445Y Mtb infection, we generated BMDMs from Nos2 -/-bone marrow and compared their ability to control bacterial replication relative to B6 BMDMs (Fig 4E).Similar to B6 BMDMs where NO production was inhibited by C-PTIO addition, BMDMs that were unable to produce NO due to lack of the inducible nitric oxide synthase were unable to control mycobacterial growth.Collectively, our data show that the protective mechanisms underlying type I IFN signaling are partially through NO production to restrict intracellular rpoB-H445Y Mtb replication.

Discussion
IL-1 signaling is essential in the early immune response to Mtb infection [15,16].However, the rifampin-resistant rpoB-H445Y Mtb bypass the need for IL-1 signaling, and the protective immune pathways in play have not been understood [24].In this study, we show a protective role for type I IFN signaling in the absence of IL-1 signaling during infection with rpoB-H445Y Mtb.Additionally, the producers of type I IFN during infection are lung monocytes and recruited macrophages.During rpoB-H445Y Mtb infection, while LysM-expressing cells respond to the type I IFN signals to confer protection, type I IFN signaling in CD11c-expressing cells has a detrimental role.Finally, we show that type I IFN and IL-1 signaling cooperate to drive NO production to limit rpoB-H445Y Mtb replication.Together, our results demonstrate a protective role for type I IFN signaling during rpoB-H445Y Mtb infection, in contrast to its proposed detrimental role during wt Mtb infection.These results highlight fundamentally divergent host-pathogen interactions during closely related Mtb infections.
The impacts of type I IFN and IL-1 signaling in wt Mtb infection and on TB pathogenesis have been well studied [20,35,36].In susceptible mouse models of TB such as C3Heb/FeJ and 129S2, type I IFNs facilitate the recruitment of Mtb-permissive cell types, drive neutrophil extracellular traps that impair Mtb control, and promote exacerbated immunopathology and disease [19,23,34,37].Indeed, in human TB, type I IFNs and neutrophils are associated with and can predict the progression toward clinical TB diagnosis [22,38,39].In models utilizing resistant mice such as B6, IL-1 signaling is necessary to control Mtb infection, especially through the production of eicosanoids, such as PGE2 [15,16,18].In these models, type I IFN signaling plays a detrimental role through antagonizing IL-1 production and signaling, limiting effective T cell immunity, and sensitizing infected macrophages towards death [17,18,34,37,40,41].Our data also shows that lack of IL-1 signaling during wt Mtb infection results in increased frequencies of macrophage death in vitro as well as increased accumulation of neutrophils in the lungs in vivo and how these processes could be driven by type I IFN signaling.Heightening type I IFN induction through poly-IC treatment in B6 mice also led to increased Mtb lung burdens through recruitment of Mtb-permissive immune cells [42].We have recently shown that type I IFN production is heightened during rpoB-H445Y Mtb infection, with type I IFN signaling leading to a suppressed immune response in the lungs of rpoB-H445Y Mtb-infected mice [25].Correspondingly, lack of type I IFN signaling during rpoB-H445Y Mtb infection led to increased immune cell infiltration and activation and more inflammation, which is shown here as well.But, this more robust immune response in Infar -/- mice led to better control of rpoB-H445Y Mtb and limited inflammation during the chronic stages of infection [25].However, some studies have found a protective role for type I IFN signaling during wt Mtb infection [35,43].For example, autocrine type I IFN signaling can promote macrophage control of wt Mtb infection, but Mtb has been shown to modulate host macrophage responses to limit potential protective autocrine type I IFN signaling, while promoting detrimental paracrine signaling [43].Also, in a context without protective type II IFN signaling, additional loss of type I IFN signaling led to the double IFN knockout mice succumbing to Mtb infection earlier than either single IFN knockout mice [35].Together, these studies suggest a complex role for type I IFN signaling where wt Mtb could be hijacking this host signal during infection to subvert protective immune responses such as IL-1 signaling.
In our work with the DR Mtb strains containing the rpoB-H445Y mutation, we have found that the impact of type I IFN and IL-1 signaling on Mtb infection are drastically changed.Previously, we found antagonism of IL-1 production by type I IFN signaling and heightened type I IFN production during infection with rpoB-H445Y Mtb.Yet, rpoB-H445Y Mtb-infected B6 mice were able to control infection, even without requiring IL-1 signaling [24].In the current study, we find that type I IFN signaling pathway genes are induced in the lungs following wt or rpoB-H445Y Mtb infection.However, while decreased expression of genes induced by type I IFN was seen in infected DKO mice relative to their Il1r1 -/-counterparts, these transcriptional changes were associated with protective or detrimental host responses during wt or rpoB-H445Y Mtb infection, respectively.So, the difference in host control of these two infections could be driven by differences in which specific ISGs were induced as well as how those ISGs function in tandem with other upregulated cellular processes, such as different metabolic pathways and stress-related responses [24,44].Thus, contextual cues could lead to diverse outcomes of type I IFN signaling during Mtb infections.Further work is needed to study how the wt and rpoB-H445Y Mtb strains drive differential expression of host genes and processes and how those pathways contribute to protective or detrimental effects of type I IFN signaling during infection.
Type I IFNs can be produced by many cell types, especially in response to microbial infection [36].Following rpoB-H445Y Mtb infection, monocytes and RMs were the major producers of type I IFNs.Kotov et al. also recently investigated sources of type I IFNs during wt Mtb infection and found interstitial macrophages and pDCs to be dominant producers [45].While we also found recruited monocyte and macrophage populations to be contributing to type I IFN production, we did not observe a significant population of pDCs in the lungs during infection, a cell type known to be a potent producer of type I IFNs [14,46], nor did we find IFN-producing non-immune epithelial cells [47].These findings could be attributed to pDCs and epithelial cells acting earlier than 7 dpi or our approach not being sensitive enough to capture the pDC and epithelial cell populations.For example, Kotov et al. used a double fluorescent reporter sytem to identify all cells that have expressed Ifnb1 at any point during infection, which would capture early or transient expression of type I IFNs.Thus, we cannot rule out a role for pDCs in our rpoB-H445Y Mtb infection model, which could be studied in follow-up work.Also, Kotov et al. carried out their studies using the Mtb Erdman strain, while we used the Mtb HN878 strain as our reference wt strain, and differences between these strains could impact the dynamics of type I IFN production by different cell types.Future experiments could also test whether myeloid cells directly infected with Mtb are type I IFN producers and how the kinetics of infection influences the kinetics of IFN production found in this study [48,49].Additional insights can also be gained by utilizing newly emerging technologies such as spatial transcriptomics, to study localization of the IFN-producing cell types with respect to infected and other responding cell types [50][51][52].
Type I IFNs signal through the IFNAR1/IFNAR2 heterodimeric receptor, which is ubiquitously expressed [36].Thus, depending on the cell type and other signaling processes, a wide variety of cellular responses can be induced, some that are protective and others that are detrimental.In the rpoB-H445Y Mtb infection model, even with functional IL-1 signaling, LysMexpressing cells required type I IFN signaling to drive a protective immune response, while type I IFN signaling was detrimental in CD11c-expressing cells.It should be noted that LysM is broadly expressed in a variety of immune cell types such as monocytes, macrophages, and granulocytes as well as some non-immune type II alveolar epithelial cells [33,[53][54][55].Our results show cell-specific effects of type I IFN signaling during rpoB-H445Y Mtb infection, but we can only speculate on the specific contribution of each cell type during infection.Neutrophils are a major cell type targeted by the LysM-Cre model and not affected in the CD11c-Cre model and, as noted above, have been associated with type I IFN signaling in other Mtb infections and shown here in wt HN878 Mtb infection.However, we saw no increases in numbers of neutrophils in either of the susceptible DKO mice or the LysM-Cre + Ifnar1 fl/fl mice.While there could be differences in functionality between the neutrophils from protected and susceptible mice, such possibilities require follow-up investigation in the future.The consistent findings of increased abundance of activated RMs across susceptible mouse models, RMs as a source of type I IFN production, and a susceptible BMDM infection model suggest that these infiltrating macrophages are likely an immune subset contributing to a deficient immune response to rpoB-H445Y Mtb in the absence of type I IFN signaling.
On the other hand, CD11c expression in the lungs is restricted to AMs and DCs.As AMs are a common subset targeted by both Cre mouse models, it is likely that loss of type I IFN signaling in this subset does not play a role in rpoB-H445Y Mtb infection.However, our findings suggest a likely detrimental role for type I IFN signaling on DCs, which is supported by our previous study where we observed reduced BMDC migration in vitro and limited trafficking of Mtb to the draining lymph nodes in vivo during rpoB-H445Y Mtb infection [25].We also previously found that global loss of Ifnar does not appear to impact rpoB-H445Y Mtb bacterial burden by 30 dpi, but Ifnar -/-mice carry reduced bacterial loads during chronic stages of infection (� 50 dpi) [25].Together, our data suggest that, in infected Ifnar -/-mice, the deficiency in the ability of RMs to control rpoB-H445Y Mtb growth might be offset by a more robust adaptive immune response triggered by DCs that are not limited by type I IFN signaling, which also improves long-term control of rpoB-H445Y Mtb infection [56,57].The susceptibility of the DKO mice to rpoB-H445Y Mtb infection and limited adaptive responses, relative to infected Ifnar -/-mice, suggest that IL-1 signaling might play a role in the DC priming of the adaptive responses and is a mechanism contributing to control.The cell-cytokine specific contributions to protection against rpoB-H445Y Mtb should be explored in future studies.
In this study, we observed a role for NO in limiting rpoB-H445Y Mtb growth, with high levels of NO associated with control of rpoB-H445Y Mtb infection and limiting NO production resulted in increased rpoB-H445Y Mtb growth.This role for NO might be Mtb strain specific and not play a significant role during wt Mtb infection.Indeed, infection with rpoB-H445Y Mtb induced higher levels of NO than with the wt Mtb strain.Although we have not determined the mechanism behind the NO production, type I IFN and IL-1 signaling could function in tandem with another signal, such as TNF, to promote Nos2 expression and NO production [58].Similarly, on the bacterial side, PDIM or another virulence factor might contribute to suppression of NO production during wt Mtb infection, and this suppression might be lost during rpoB-H445Y Mtb infection due to differences in virulence factors between wt Mtb and and rpoB-H445Y Mtb strains [24].Future work can aim to dissect the roles of host and pathogen factors in contributing to NO production.Studies in murine models suggest that NO contributes to limiting Mtb replication directly or indirectly, especially in macrophage subsets [59,60].In primary human AM studies, NO induction has been mechanistically implicated in the killing of Mycobacterium species in vitro [61,62].Thus, while NO production is elevated in murine cells it may not be reflective of Mtb control in human macrophages [63].Our results show that both type I IFN and IL-1 signaling cooperate to induce production of NO in an additive manner.While type I IFN signaling has been more closely linked with NO production through the inducible nitric oxide synthase (Nos2) [43,64], it is possible that IL-1 could also drive NO production as in the case of Brucella infection [65].Collectively, the data suggest that the dispensability of either signaling pathway during in vivo rpoB-H445Y Mtb infection is tied to compensatory NO production that is sufficient for protection.
An important question arising from this work is why wt and rpoB-H445Y Mtb are driving different infection outcomes.As we observe induction of similar pathways, albeit at different magnitudes, we hypothesize that differences in the expression of virulence factors between wt and rpoB-H445Y Mtb could influence the magnitude and timing of host cell responses.Concurrently, differential expression of virulence factors could impact the physiology of rpoB-H445Y Mtb and alter the response of the mutant Mtb to the pathways induced by type I IFN signaling, such as susceptibility to NO.We speculate that the location of the mutation in rpoB, which is a critical component of the bacterial transcription machinery, may result in broad transcriptional changes and expression of virulence factors [24,66].We hypothesize that the increased production of long-chained PDIM in the rpoB-H445Y Mtb strain could be affecting the integrity of host phagosomes and lead to increased bacterial escape into the cytosol [67][68][69][70].But, we have not thoroughly assessed all of the changes these DR mutations drive and how those changes are linked with the divergent host-pathogen interactions.A caveat of this study is that we cannot establish a causal link between the rpoB-H445Y mutation and the immune phenotypes observed in the study.The rpoB-H445Y Mtb strain used in this study was derived from the wt Mtb strain through antibiotic selection, and multiple independent DR Mtb isolates with the rpoB-H445Y mutation drove the same responses in vitro and in vivo [24].However, a few additional mutations exist that separate the two strains outside of the rpoB-H445Y mutation, and we cannot rule out the possibility that additional mutations contribute to the differential host response [25].The question of causality remains open and must be established in future work.In summary, our studies have mechanistically shown how type I IFN signaling can drive protective or detrimental outcomes during Mtb infections and impact TB disease pathogenesis.Additionally, we uncover novel immune circuits through the use of a rpoB-H445Y Mtb strain with two mutually antagonistic signaling pathways acting in a compensatory but not redundant manner.Thus, our work highlights the necessity for greater investigation of hostpathogen interactions using a broader repertoire of Mtb strains and including drug-resistant isolates, which can reveal novel targets for future development of host-directed therapeutics.

Ethics statement
Protocols involving the use of animals were approved by the Washington University in St. Louis or The University of Chicago Institutional Animal Care and Use Committee (IACUC) guidelines.All of the experiments were performed in accordance with the protocols.

Mice
C57BL/6J (B6), Ifnar1 −/− , Il1r1 -/-, Nos2 -/-, and Ifnb mob mice on the B6 background were purchased from The Jackson Laboratory (Bar Harbor, ME).Ifnar1 -/-Il1r1 -/-mice were generated by crossing Ifnar1 −/− and Il1r1 -/-mice.LysM-Cre Ifnar1 fl/fl and Cd11c-Cre Ifnar1 fl/fl mice were generously provided by the Baldridge Lab (Washington University in St. Louis).In some experiments, LysM-Cre + and LysM-Cre -Ifnar1 fl/fl mice were administered sterile PBS or a-IL1R blocking antibody (BioXCell) at 100 ug/mouse in 200uL every other day intraperitoneally (IP) starting at -1 dpi.All mice were maintained in the animal facility at Washington University in St. Louis and bred in house.Experimental mice were age and sex matched and infected between the ages of 6 and 8 weeks.All mice were maintained and used in accordance with the approved Washington University in St. Louis Institutional Animal Care and Use Committee guidelines.

Generation of BMDMs
Bone marrow derived macrophages (BMDMs) were generated as previously described [11,24].Briefly, bone marrow cells were collected from the tibia and femur of B6 mice and cultured in complete Dulbecco's modified eagle's medium (cDMEM) with 20 ng/mL of recombinant granulocyte-macrophage colony-stimulating factor (GM-CSF).Cells were cultured at 37˚C in 5% CO 2 and supplemented with media on day 3. On day 7, floating cells were taken up and discarded, and adherent cells were collected as BMDMs.

Mtb infections
Mtb strain HN878 (wt) was obtained from BEI resources (Manassas, VA) under National Institutes of Health contract AI-75320.Independent rifampicin resistant Mtb colonies (biological replicates) were selected from rifampicin (2 μg/ml) containing 7H11 agar plates [3].The sequences of rpoB were confirmed by Sanger sequencing (Genewiz), and Mtb stocks were created for further experimentation.The whole genomes of all Mtb strains were previously sequenced, with all SNPs between both, wt and rpoB-H445Y, Mtb strains are published [25].All Mtb strains were cultured in Proskauer Beck medium supplemented with 0.05% Tween 80 and frozen at −80˚C while in mid-log phase prior to use in mouse and macrophage infections.Colony forming units (CFU) of the bacterial stocks were determined through serial dilutions on 7H11 agar plates.Mice were aerosol infected with low doses (~100 CFU) of indicated Mtb strains in sterile PBS using a Glas-col nebulizer [71].Mtb bacterial burden/organ was quantitated by plating serial dilutions of homogenized lung tissue on 7H11 agar plates (BD Biosciences, Franklin Lakes, NJ).Plates were incubated for 2-3 weeks at 37˚C, and the number of colonies were counted.In vitro infections were carried out at a multiplicity of infection (MOI) of 1 in antibiotic free media for six days.In some experiments, macrophages were treated with indicated amounts of carboxy-PTIO (C-PTIO) four hours after the start of infection.Bacterial burden in each well was determined through serial plating of supernatant and lysed macrophages after six days post infection.
For the determination of frequencies of apoptotic and necrotic cells, BMDMs were collected at the indicated timepoints and stained with Annexin V (PE) and 7AAD as per the instructions of the PE Annexin V Apoptosis Detection Kit (559763, BD Pharmigen), prior to processing on the X-20 flow cytometer.

Histology
Lung lobes were perfused with 10% neutral buffered formalin and embedded in paraffin.Formalin-fixed paraffin-embedded (FFPE) lung sections were stained with hematoxylin and eosin (H&E) and inflammatory features were evaluated by light microscopy.Inflammatory lesions were outlined with the automated tool of the Zeiss Axioplan 2 microscope (Carl Zeiss) and the total inflammatory area in each lung lobe was measured.

Protein determination
BMDMs were either stimulated with Mtb antigens for 1 day or left unstimulated, lysed in RIPA Lysis Buffer (Santa Cruz Biotechnology), and heat denatured in reducing sample buffer (Thermo Fisher Scientific).Proteins were separated in 4% to 20% polyacrylamide gradient gels (Bio-Rad) and transferred onto PVDF membranes.Membranes were probed with primary antibodies against IFNAR1 (R&D, Cat#AF3039), and GAPDH (Santa Cruz Biotechnology, Cat# SC-25778) followed by incubation with HRP-conjugated anti-rabbit IgG secondary antibody (Santa Cruz Biotechnology) and developed by Clarity Western ECL Substrate (Bio-Rad).

RNA isolation, sequencing, and processing
In brief, lung tissue was homogenized in RLT buffer, and total RNA was extracted with RNeasy Mini Kit (Qiagen) as previously described [11].Total RNA integrity was determined using Agilent Bioanalyzer or 4200 Tapestation.Library preparation was performed with 5 to 10ug of total RNA with a Bioanalyzer RIN score greater than 8.0.Ribosomal RNA was removed by poly-A selection using Oligo-dT beads (mRNA Direct kit, Life Technologies).mRNA was then fragmented in reverse transcriptase buffer and heating to 94 degrees for 8 minutes.mRNA was reverse transcribed to yield cDNA using SuperScript III RT enzyme (Life Technologies, per manufacturer's instructions) and random hexamers.A second strand reaction was performed to yield ds-cDNA. cDNA was blunt ended, had an A base added to the 3' ends, and then had Illumina sequencing adapters ligated to the ends.Ligated fragments were then amplified for 12-15 cycles using primers incorporating unique dual index tags.Fragments were sequenced on an Illumina NovaSeq-6000 using paired end reads extending 150 bases.RNA-seq reads were then aligned and quantitated to the Ensembl release 101 primary assembly with an Illumina DRAGEN Bio-IT on-premise server running version 3.9.3-8software (GTAC MGI).
The read count data of each gene were generated using featureCounts from Rsubread package [73] with default parameter settings.Significantly differentially expressed genes across different groups were identified using DESeq2, version 1.30.1 [74], with default settings and a minimum p-value significance threshold of 0.05 (after FDR correction for the number of tests) [75].Pathway enrichment analysis among differentially expressed gene sets of interest was performed for Reactome pathways using ReactomePA [76] with p-value � 0.05 after FDR correction.Gene set enrichment analysis was also performed among differentially expressed gene sets of interest with hallmark gene sets for Mus musculus using fgsea [77] with eps set to zero to estimate p-value more accurately and p-value � 0.05 after FDR correction.Additionally, several R packages were used including stringr [78], pheatmap [79], ggplot2 [80], and org.Hs.eg.db [81] for data processing and generate figures.

Statistical analysis
All datasets were evaluated for normality using the Shapiro-Wilk Test.Differences between the means of more than two groups were analyzed using One-way ANOVA or Two-way ANOVA, where applicable, with Tukey's post-tests for normally distributed distributions or Kruskal Wallis Test with Dunn's multiple comparisons tests.Differences between the means of two groups were analyzed by the unpaired two-tailed student's t-test.All statistical analyses were done in GraphPad Prism 9. A p-value < 0.05 was considered significant.The data points across the figures represent the mean (±SD or +SEM) of values as noted.*, p�0.05; **, p�0.01; ***, p�0.001; ****, p�0.0001; ns-not significant (p>0.05).All experiments were replicated for reproducibility.
Mtb and sacrificed after 30 dpi.A) Formalin-fixed, paraffin embedded (FFPE) lung sections were stained with hematoxylin and eosin (H&E), and the total inflammatory areas were quantified after wt Mtb infection at 30 days post infection (dpi).B) Total numbers of Major Histocompatibility Complex Class II hi (MHCII hi ) recruited macrophages (RMs), myeloid dendritic cells (mDCs) MHCII hi , neutrophils, MHCII hi alveolar macrophages (AMs), and MHCII hi monocytes were determined via flow cytometry.The numbers of C) CD4 + CD44 hi IFNg + T cells and D) CD8 + CD44 hi IFNg + T cells were also determined by flow cytometry following ex vivo stimulation with purified Mtb antigens ESAT6 and Ag85B.The data shown represent the means ± SD of four to seven mice per experiment.The data were evaluated for normality using the Shapiro-Wilk Test and passed (p-value > 0.05).One-way ANOVA, with Tukey's multiple comparisons tests, were used for A-D.Ifnar1 fl/fl mice and littermate controls were aerosol infected with a low dose of rpoB-H445Y Mtb and administered a-IL1R or isotype antibodies i.p. every other day, starting one day before infection, and sacrificed at 28 dpi.A) FFPE lung sections that were H&E stained were prepared and the inflammatory areas were quantified at 28 dpi.Representative histological images are shown, with quantification of total inflammatory area divided by total lung area for each lobe depicted below.B) Through flow cytometry, total numbers of MHCII hi RMs and mDCs were calculated in the lungs at 28 dpi with the rpoB-H445Y Mtb strain.The data shown represent the means ± SD of five to nine mice per experiment.The data were evaluated for normality using the Shapiro-Wilk Test and passed (p-value > 0.05).One-way ANOVA, with Tukey's multiple comparisons tests, were used for A and B.
fl/fl mice (Fig 3D).But, similar to the rpoB-H445Y Mtbinfected DKO mice, loss of the type I IFN receptor in LysM-Cre + Ifnar1 fl/fl mice treated with IL-1 signaling blocking antibody led to an increase in lung Mtb burden.Unexpectedly, loss of the type I IFN receptor in the LysM-expressing cells alone led to increased rpoB-H445Y Mtb lung burden (Fig 3D).The increased susceptibility in LysM-Cre + mice infected with rpoB-H445Y Mtb correlated with increased pulmonary inflammation (S4A Fig) and increased accumulation of lung MHCII hi RMs and mDCs (S4B Fig).Together, our results suggest that type I IFN signaling is necessary in LysM + cells to mediate control of rpoB-H445Y Mtb infection but dispensable when deleted globally as in the Ifnar1 -/-mice.To better understand this apparent discrepancy, we assessed the impact of the loss of type I IFN signaling in CD11c-expressing cells, which include AMs and DCs (Fig3E).Cd11c-Cre + Ifnar1 fl/fl mice better controlled rpoB-

Fig 3 .
Fig 3. Type I IFN producers are CD11b + myeloid cells, while LysM + cells respond to type I IFN following rpoB-H445Y Mtb infection.IFNβ-YFP mice were aerosol infected with rpoB-H445Y Mtb and sacrificed at the indicated timepoints along with uninfected controls.A) Representative histograms for various subsets [Monocytes (left), Recruited Macrophages (RMs) (middle left), Neutrophils (middle right), and Alveolar Macrophages (AMs) (right)] depicting YFP expression by intracellular antibody staining (green) compared to controls [uninfected IFNβ-YFP mice stained with intracellular antibody (orange), infected IFNβ-YFP mice stained with intracellular isotype antibody (blue), and infected B6 non-reporter mice stained with intracellular antibody (red)].B) The frequencies of YFP + cells were determined by flow cytometry following sacrifice of rpoB-H445Y Mtb-infected IFNβ-YFP mice at 7, 14, and 28 dpi, along with uninfected controls.C) Experimental scheme for the LysM-Cre Ifnar1 fl/fl mice with the purple arrow indicating aerosol infection with rpoB-H445Y Mtb and the pink arrow indicating administration of a-IL1R or isotype antibodies i.p.D) Lung Mtb burden was determined at 28 dpi.E) LysM-Cre + Ifnar1 fl/fl (semi-closed green circles) and Cd11c-Cre + Ifnar1 fl/fl mice (semi-closed orange circles), along with littermate controls (open blue circles), were infected with rpoB-H445Y Mtb, and bacterial burden was determined in the lungs at 28 dpi.The data shown represent the means ± SD of three to eight mice per experiment.The data were evaluated for normality using the Shapiro-Wilk Test and passed (p-value > 0.05).Two-way ANOVA and One-way ANOVA, with Tukey's multiple comparisons tests, were used for B and D-E, respectively.Significant differences are indicated with asterisks (*, pvalue � 0.05; **, p-value � 0.01; ***, p-value � 0.001; ****, p-value � 0.0001) by appropriate statistical tests.One-way ANOVA.Two independent experiments shown.https://doi.org/10.1371/journal.ppat.1012137.g003 Fig 4A).Additional loss of type I IFN signaling rescued this defect in the DKO BMDMs.In contrast, during rpoB-H445Y Mtb infection, loss of either type I IFN or IL-1 signaling led to an increase in intracellular Mtb growth, with loss of both (DKO) causing a further increase in Mtb growth (Fig 4B).We also assessed cell death dynamics during the BMDM infections (S5 Fig).Across the four genotypes, we find that BMDMs lacking only Il1r1 exhibited increased cell death as seen with increased frequencies of cells at apoptotic (AnnexinV hi 7AAD lo ) and

Fig 4 .
Fig 4. NO production restricts intracellular rpoB-H445Y Mtb growth in macrophages.Bone marrow-derived macrophages (BMDMs) from B6, Ifnar1 -/-, Il1r1 -/-, DKO mice were infected with either A) wt Mtb or B) rpoB-H445Y Mtb, and Mtb CFU was determined at 6 dpi.C) The concentration of nitrite in the supernatants of wt or rpoB-H445Y Mtb infected BMDMs was determined, along with uninfected controls.D) B6, Il1r1 -/-, Ifnar -/-and DKO BMDMs were infected with rpoB-H445Y Mtb and then treated with the indicated concentrations of an NO scavenger, Carboxy-PTIO (C-PTIO).Bacterial burden was determined on 6 dpi.Fold change in Mtb growth was determined relative to untreated and infected BMDMs.E) B6 and Nos2 -/-BMDMs were infected with rpoB-H445Y Mtb, and Mtb CFU was determined at 6 dpi.The data shown represent the means ± SD of three to four biological replicates per experiment.The data were evaluated for normality using the Shapiro-Wilk Test and passed (p-value > 0.05).One-way ANOVA and Two-way ANOVA, with Tukey's multiple comparisons tests, were used for A-B and C-D, respectively.Unpaired two-tailed student's t-test was used for E. Significant differences are indicated with asterisks (*, p-value � 0.05; **, p-value � 0.01; ***, pvalue � 0.001; ****, p-value � 0.0001) by appropriate statistical tests.One of three independent experiments shown.https://doi.org/10.1371/journal.ppat.1012137.g004 Significant differences are indicated with asterisks (*, p-value � 0.05; ns, p-value > 0.05) by appropriate statistical tests.One of three independent experiments shown.(TIF) S2 Fig. Loss of control of rpoB-H445Y Mtb infection is not associated with changes in other myeloid or T cell subsets.C57Bl/6 (B6), Ifnar1 -/-, Il1r1 -/-, and Ifnar1 -/-Il1r1 -/-[double knockout (DKO)] mice were aerosol infected with a low dose of rpoB-H445Y Mtb and sacrificed after 30 dpi.Total numbers of A) neutrophils, B) Major Histocompatibility Complex Class II hi (MHCII hi ) monocytes, and C) MHCII hi alveolar macrophages (AMs) were determined via flow cytometry.The numbers of D) CD4 + CD44 hi IFNg + T cells and E) CD8 + CD44 hi IFNg + T cells were also determined by flow cytometry following ex vivo stimulation with purified Mtb antigens ESAT6 and Ag85B.The data shown represent the means ± SD of five to nine mice per experiment.The data were evaluated for normality using the Shapiro-Wilk Test and passed (p-value > 0.05).One-way ANOVA, with Tukey's multiple comparisons tests, were used for A-E.Significant differences are indicated with asterisks (*, p-value � 0.05; **, p-value � 0.01; ***, p-value � 0.001; ****, p-value � 0.0001) by appropriate statistical tests.One of three independent experiments shown.(TIF) S3 Fig. Type I IFN signatures are upregulated in B6 and IL-1 knockout mice infected with either wt or rpoB-H445Y Mtb.C57Bl/6 (B6), Ifnar1 -/-, Il1r1 -/-, and Ifnar1 -/-Il1r1 -/-(DKO) mice were aerosol infected with a low dose of wt or rpoB-H445Y Mtb and sacrificed at 14 dpi.RNA was extracted from homogenized lung tissue and sequenced.The expression of genes in the type I IFN pathway in bulk lung cells from DKO mice over bulk lung cells from the wt mice is shown (top), DKO mice over Il1r1-/-mice (middle), and Ifnar1-/-over wt mice (bottom) after A) wt or B) rpoB-H445Y Mtb infection with annotated pathways (n = 4-6).Gene set enrichment analysis was done using an fgsea R package as noted in Methods.The pathways are enriched in the second group of each comparison.(TIF) S4 Fig. Increased susceptibility of LysM-Cre + Ifnar1 fl/fl mice to rpoB-H445Y Mtb infection is associated with exacerbated immunopathology.As described in Fig 3C, LysM-Cre + Significant differences are indicated with asterisks (*, p-value � 0.05; **, p-value � 0.01; ***, p-value � 0.001; ****, p-value � 0.0001) by appropriate statistical tests.One of two independent experiments shown.(TIF) S5 Fig. Cell death dynamics of rpoB-H445Y Mtb-infected BMDMs are influenced by genotype but are not directly correlated with susceptibility.BMDMs from B6, Ifnar1 -/-, Il1r1 -/-, DKO mice were infected with either wt Mtb (circles) or rpoB-H445Y Mtb (squares).At indicated timepoints post infection, BMDMs were collected, stained with Annexin V and 7AAD, per manufacturer's instruction, and processed using flow cytometry.Uninfected controls were also included.The frequencies of A) apoptotic (7AAD lo AnnexinV hi ) and B) necrotic (7AAD hi AnnexinV hi ) BMDMs were quantified.The data shown represent the means ± SD of three to four biological replicates per experiment.The data were evaluated for normality using the Shapiro-Wilk Test and passed (p-value > 0.05).Two-way ANOVA, with Tukey's multiple comparisons tests, was used for A and B. Significant differences are indicated with asterisks (*, pvalue � 0.05; **, p-value � 0.01; ***, p-value � 0.001; ****, p-value � 0.0001) by appropriate statistical tests.One of two independent experiments shown.(TIF) S1 Data.This file contains the source data used to generate every figure in this manuscript.Each tab of the Excel file corresponds to a specific figure panel (main and supplemental) and is labeled as such.(XLSX)