The absence of Tax1bp1 reduces Mtb growth during acute and chronic stages of animal infection.

To assess Tax1bp1’s contribution to controlling Mtb infection in vivo we exposed wild-type and Tax1bp1-deficient mice to aerosolized, virulent Mtb. We first measured colony forming units (CFU) one day after exposing male and female mice to a low dose. We saw no statistically significant difference in bacterial uptake between wild-type and mutant mice of the same sex (Fig 1A). However, female Tax 1bp1^-/- mice had decreased bacterial uptake compared to male Tax 1bp1^-/- mice (Fig 1A). To control for this variable, subsequent animal infections were performed with sex-matched wild-type and Tax 1bp1^-/- mice.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Tax1bp1-deficiency abrogates M. tuberculosis virulence and inflammatory cytokine responses during mouse aerosol infection.

(A) Age and sex-matched male and female mice were infected by the aerosol route with a mean Mtb CFU of 240 as determined by CFU enumeration from lung homogenates at 1-day post-infection. (B) Male mice were euthanized at 11 and 21-days post-infection for CFU enumeration. Female mice were euthanized at 50-days post-infection. Results are the mean ± SEM from lung homogenates of 4-5 infected mice per genotype. (C) Cytokine levels from infected lung homogenates at 11-, 21-, and 50-days post-infection were measured by ELISA. Results are the mean ± SEM from five samples. (D) Levels of type I and II interferon-induced JAK/STAT signaling were measured by luminescence in relative light units (RLUs) from infected lung homogenates by the ISRE assay. Results are the mean ± SEM from five samples. Brackets indicate p values from t-test comparisons. (E) Infected mice were monitored for death or 15% loss of maximum body weight, at which point they were euthanized. Log-rank (Mantel-Cox) test and Gehan-Breslow-Wilcoxon comparison test p values for survival were 0.0008 and 0.0047, respectively. (F) The weight of the infected mice is displayed. The connecting line denotes the mean weight of the mice alive at each time point.

https://doi.org/10.1371/journal.ppat.1012829.g001

To examine the role of Tax1bp1 in controlling Mtb growth, we harvested the lung, spleen, and liver of wild-type and Tax 1bp1^-/- mice on days 9 or 11, 21, and 50 post-infection and quantified bacterial CFUs (Figs 1B and S1). Contrary to our previous results in BMDMs infected ex vivo [45], we found that Mtb growth decreased in the lungs of Tax1bp1-deficient mice (Figs 1B and S1). This unexpected difference manifests even after 11 days of infection, during the acute stage when Mtb replicates within AMs. Mtb’s restricted growth was magnified in both peripheral organs, liver, and spleen, consistent with the differences in lung CFUs. These results contrast with the autophagy receptor function of Tax1bp1 since Tax1bp1 contributes to targeting Mtb to selective autophagy [46], a host response that can contribute to the control of Mtb growth in vivo and ex vivo to a variable degree [24,47–55].

Tax1bp1-deficiency abrogates pro-inflammatory cytokine production during Mtb infection in vivo.

We reasoned, given Tax1bp1’s known role in inflammation, that differences in Mtb growth between wild-type and Tax1bp11-deficient mice might reflect differences in inflammatory responses to Mtb infection. Indeed, consistent with this idea, the lungs of Tax1bp1-deficient mice exhibited decreased levels of IL-6, TNF-α, IL1-β, and IL-12/IL-23 p40 (Figs 1C and S1).

Type I and II IFNs are particularly important for controlling Mtb infection [22, 29, 47]. Therefore, we measured interferon levels using a type I and II IFN reporter cell line (ISRE) and type II IFN (IFN-γ) by ELISA (Fig 1C and 1D). Consistent with other pro-inflammatory cytokines, we found that Tax1bp1-deficient mice had significantly lower levels of type I and II IFN in the lungs compared to wild-type mice (Fig 1C and 1D).

Although Tax1bp1-deficiency decreased inflammatory cytokine synthesis during Mtb infection, microscopic examination of infected lung tissue did not reveal any significant differences in the cellular infiltrate of the lungs as reflected by lesion severity or tissue necrosis (S2 Fig). There was a trend that was not statistically significant towards decreased neutrophil recruitment in Tax1bp1-deficient lung samples as reflected by myeloperoxidase staining (S2 Fig).

During infection of BMDMs, we previously observed that Tax1bp1-deficiency enhanced ubiquitin colocalization with Mtb [44]. Increased ubiquitylation was also observed in the brains of Tax 1bp1^-/- mice [46]. Thus, we sought to determine whether ubiquitin recruitment was regulated during infection in vivo. While Tax1bp1 led to a slight decrease in ubiquitin and Mtb colocalization in infected lung tissue samples at 50 days post-infection, this did not reach statistical significance (S3 Fig).

Finally, we carried out survival studies and found that Tax1bp1-deficiency contributes to reduced mortality during Mtb infection (Fig 1E) and enhances weight gain (Fig 1F). Collectively, our results suggest that Tax1bp1-deficiency limits host-detrimental systemic inflammatory responses and reduces host susceptibility and mortality.

Tax1bp1-deficiency impairs the recruitment of neutrophils and CD8⁺ T cells to the lungs during Mtb infection.

Having shown that Tax1bp1-deficiency restricts bacterial growth and inflammatory cytokine production during Mtb and Listeria monocytogenes animal infection, we next sought to further explore how Tax1bp1-deficiency impacts recruitment of inflammatory immune cells during Mtb infection. First, we turned to a more sensitive approach compared to histopathology to examine immune cell populations in the lungs. To precisely quantify immune cell types in the lungs during infection with green-fluorescent Mtb, we performed analytical flow cytometry of lung homogenates (Fig 3). We quantified two sets of immune cells after gating for live single cells: (1) myeloid cell subsets including CD11c^lo (MNC1) and CD11c^hi (MNC2) mononuclear cells, neutrophils, and AMs (S6 Fig); and (2) B, T, and NK cells (S7 Fig). Using this more sensitive approach revealed (i) decreased neutrophil recruitment and (ii) increased proportions of live AMs at 28 days post-infection in mice lacking Tax1bp1 (Fig 3). One possible explanation for this result is that Tax1bp1-deficiency increased the survival of AMs during Mtb infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Tax1bp1-deficiency impairs neutrophil and CD8⁺-T cell recruitment to the lungs during Mtb infection.

Lung homogenates from female mice (day 14, 21 post-infection) and male mice (day 28 post-infection) infected with ZsGreen-expressing Mtb were stained for antibodies to detect immune cells and analyzed by flow cytometry. (A and C) Displayed are the immune cell subtype proportions of the single cell population from individual mouse lung samples. To account for differences in the number of cells analyzed between genotypes, data were normalized by dividing the proportion of each single cell population by the total number of each live cell type analyzed. (B and D) Displayed are the ZsGreen+ counts quantified in each immune cell subtype from individual mouse lung samples. Data were normalized to the number of cells sorted. Mean and SEM are displayed. The p-values from t-test comparisons are shown.

https://doi.org/10.1371/journal.ppat.1012829.g003

In addition to decreasing neutrophil recruitment, Tax1bp1-deficiency increased MNC1 recruitment at 28 days post-infection but had no impact on MNC2 recruitment (Fig 3C). Tax1bp1-deficiency reduced CD8⁺ T cell recruitment by 2 and 2.5-fold at 21- and 28-days post-infection, respectively (Fig 3A). This result is consistent with Tax1bp1-deficiency limiting systemic inflammatory cytokine responses, as CD8⁺ T cells are associated with the production of cytokines including IFN-γ [57] and TNF-α [58]. In contrast, CD4⁺ T cell, NK cell, and B cell levels in the lungs were not affected by Tax1bp1-deficiency (Fig 3A). Together with the decreased cytokine production and lack of pulmonary histopathologic changes mediated by Tax1bp1-deficiency, the reduced CD8⁺ T cell and neutrophil recruitment indicates that Tax1bp1-deficiency abrogates systemic inflammatory responses during infection.

Tax1bp1-deficiency has a cell type-specific impact on Mtb growth.

Having shown that Tax1bp1-deficiency impacts the recruitment of immune cell subpopulations during Mtb infection, we then determined the effect of Tax1bp1-deficiency on Mtb growth in these same immune cells. As a reflection of Mtb burden, we analyzed the number of ZsGreen⁺ cells normalized to the number of each cell type analyzed. As expected, most of the green fluorescence signal from Mtb was detected in the myeloid cells (Fig 3B-D) that serve as a replicative niche for Mtb[4,9,48]. There were no significant differences in ZsGreen counts in T, B, or NK cells except at day 14 in CD8⁺ T cells (Fig 3B). Tax1bp1-deficiency reduced ZsGreen+ Mtb counts in AMs, MNC1, and MNC2 at 14 days post-infection and in AMs at 28 days post-infection (Fig. 3D). We interpret these results to suggest that Tax1bp1-deficiency restricts Mtb growth in AMs, MNC1, and MNC2 at specific time points.

To further investigate the hypothesis that Tax1bp1-deficiency restricts Mtb growth in a cell type-specific manner, we performed Mtb CFU analysis in sorted immune cells in addition to measurement of bacterial fluorescence as a proxy for Mtb burden. CFU analysis enables measurement of specifically live bacteria, whereas fluorescence counts can represent both live and dead bacteria. As opposed to the previous experiments analyzing immune cells from individual mice (Fig. 3), here we pooled lung cell suspensions from five mice of each genotype to obtain enough cells for downstream CFU analysis from sorted cell populations. In the rederived Tax 1bp1^-/- mice, as observed previously, Tax1bp1-deficiency restricted Mtb growth in the lungs and spleens (S8 Fig). Tax1bp1-deficiency also restricted Mtb CFU1bp1in innate immune cells except for neutrophils and MNC1 on day 21 (S8 Fig). To determine whether the cell type-specific phenotype of Tax1bp1-deficiency on Mtb growth was consistent across independent experiments, the ZsGreen+ Mtb counts from the sorted cells were quantified (experiment 1, S8 Fig) and analyzed together with a second independent experiment using pooled samples (experiment 2, S8 Fig), and the mean ZsGreen+ counts from mouse samples analyzed individually (experiment 3, S8 Fig) as previously displayed in Fig 3D. This analysis of three independent experiments indicates that Tax1bp1-deficiency can enhance Mtb growth in MNC1 and neutrophils at 21-days post-infection (S8 Fig), which is consistent with our previous CFU analysis in BMDMs infected ex vivo. In contrast, Mtb growth was reduced in Tax 1bp1^-/- AMs at 14- and 21-days post-infection and in MNC2 at 14-days post-infection (S8 Fig).

In summary, Tax1bp1-deficiency has a cell type-specific impact on Mtb growth, with an overall effect of decreasing Mtb growth in the major organs. We conclude that Tax1bp1-deficiency restricts Mtb growth in AMs and MNC2 while simultaneously enhancing Mtb growth in MNC1 and neutrophils at specific time points.

Tax1bp1-deficiency reduces Mtb growth during AM infection ex vivo.

We were curious, given our previous ex vivo results in BMDMs [44] whether Tax1bp1-deficiency also abrogates Mtb growth in AMs infected ex vivo. Therefore, we obtained AMs from uninfected mice by bronchoalveolar lavage (BAL) that were isolated by simple adherence of the AMs in the BAL fluid. When infected with a luminescent strain of Mtb, Mtb grew more slowly in AMs from mice lacking Tax1bp1 relative to wild-type AMs both in the presence or absence of IFN-γ (Fig 4A),a cytokine that activates macrophage antibacterial host responses [56]. Similarly, luminescent and wild-type Mtb grew more slowly in Tax 1bp1^-/- AMs as measured by CFU (Fig 4B-C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Mtb growth was restricted in Tax1bp1^-/- AMs infected ex vivo.

AMs were infected ex vivo with luciferase-expressing Mtb H37Rv (A, B) or wild-type Mtb Erdman (C) at a M.O.I. of 1 in the presence or absence of IFN-γ added at the time of infection. (A) The media was replaced with fresh media daily. Monolayer luminescence was measured daily. (B, C) CFU were measured immediately after infection (day 0) or 4 days post-infection. Displayed are the mean, SEM, and FDR-adjusted p values from the t-test.

https://doi.org/10.1371/journal.ppat.1012829.g004

Because adherent BAL cells may contain other cell types besides AMs that could influence microbial growth, we next characterized the cells in the BAL by flow cytometry and determined whether these growth phenotypes also occur in live CD11c⁺SilgecF⁺ AMs sorted from the BAL fluid. Consistent with a previous report that >95% of BAL cells from uninfected mice are AMs [57], 95.6% of live single cells from wild-type animals, and 96.8% of the cells from Tax 1bp1^-/- mice, expressed the AM signatures CD11c and SiglecF (S9 Fig). Tax1bp1-deficiency reduced Mtb growth in the sorted AMs (S9 Fig). Together, this data implicates Tax1bp1-deficiency in reducing Mtb growth during AM infection and is consistent with our in vivo studies, although it is in sharp contrast to our results in Mtb-infected BMDMs ex vivo [44]. Therefore, we focused on understanding Tax1bp1’s function in AMs as this model better represents the overall impact of Tax1bp1 on real-world infections. Even though Tax1bp1’s phenotype in BMDMs does not model the phenotype in several other cell types (i.e., AMs, MNC2), we took advantage of Tax1bp1 phenotype in BMDMs by using BMDMs as a control.

Tax1bp1-deficiency reduces autophagy flux and colocalization of LC3 with Mtb.

Tax1bp1 has an established role in antimicrobial selective autophagy [44, 45]. To determine whether this mechanism might contribute to differences in the growth of Mtb in ex vivo AMs and BMDMs from Tax1bp1-deficient mice relative to wild-type mice we investigated biogenesis of macrophage autophagosomes, which are double-membrane structures that form around cargo destined for lysosomal fusion. As an indicator of autophagosome formation, we measured the conversion of the precursor protein LC3-I to LC3-II [49]. LC3-II is the activated and cleaved form of LC3 that embeds in the autophagosome membrane [50]. Tax1bp1-deficiency led to a reduction in the LC3-II:-LC3-I ratio in unstimulated AMs (S10 Fig) but not BMDMs (S10 Fig), indicating that Tax1bp1-deficiency abrogates autophagosome formation in AMs. To assess autophagic turnover (flux), we increased the sensitivity of the assay by inducing LC3-I to LC-II conversion through starvation with nutrient-deficient media (EBSS) and blocked LC3-II degradation by treating the cells with bafilomycin (S10 Fig). Tax1bp1-deficiency led to a more dramatic decrease in the LC3-II:LC3-I ratio in AMs (S10 Fig) compared to BMDMs (S10 Fig), revealing that Tax1bp1-deficiency leads to a reduction in autophagy flux in AMs and BMDMs. These results are consistent with a previous report that Tax1bp1 promotes autophagy flux in HeLa cells [51].

Since we previously showed that Tax1bp1-deficiency impaired selective autophagy of Mtb in BMDMs [44] and Tax1bp1-deficiency affects autophagy flux in macrophages (S10 Fig), we questioned whether Tax1bp1 would also reduce selective autophagy of Mtb in AMs. To test this, we performed confocal fluorescence microscopy of wild-type and Tax 1bp1^-/- AMs infected with fluorescent Mtb and assessed the colocalization of Mtb and autophagy markers. At 24 hours post-infection, Tax1bp1-deficiency slightly increased Mtb colocalization with ubiquitin by 16% (S11 Fig), whereas Tax1bp1-deficiency decreased Mtb colocalization with LC3 by 26% (S11 Fig). These results indicate that Tax1bp1-deficiency inhibits the progression of the ubiquitin+ Mtb phagosome to the LC3 + autophagosome. This is the same pattern of autophagy marker staining as we previously observed in Mtb-infected wild-type and Tax 1bp1^-/- BMDMs [44]. Since the impact of Tax1bp1-deficiency on Mtb growth, but not autophagy targeting, was cell type-specific, we reasoned that the lack of Tax1bp1 restricted Mtb growth in AMs by a different mechanism.

Tax1bp1 protein level does not change during Mtb infection of AMs.

We next sought to identify a mechanistic explanation for the cell type-specific role of Tax1bp1-deficiency during Mtb infection of BMDMs and AMs. Based on previous literature that showed Tax1bp1 protein levels in peripheral blood mononuclear cells correlate with the presence of autoimmune disease [52], 8; and that Tax1bp1 protein expression is partially regulated at the post-translational level [58], we hypothesized that differences in Tax1bp1 protein abundance during Mtb infection may, in part, explain the different phenotypes in BMDMs and AMs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Global protein abundance changes during Mtb infection of AMs.

In five independent biological replicate experiments, wild-type AMs were mock- or Mtb-infected at a M.O.I. of 10. At 24-hours post-infection, AM lysates were harvested, and the samples were digested with trypsin for LC-MS/MS analysis. The amino acid sequence, the relative levels of peptides, and the protein ID for the peptides were determined by a database search and statistical analysis in the infected vs. mock-infected samples. (A) Volcano plot displaying the proteins that significantly changed in abundance. The grey line denotes an adjusted p value of 0.05. Statistically significant hits with a log₂fold change > 1 (yellow) or <-1 (blue) and an adjusted p value < 0.05 are highlighted. (B) Area-proportional Venn diagram displaying the number of statistically significant proteins that changed in abundance during Mtb infection of AMs and BMDMs. (C) Gene ontology enrichment analysis of statistically significant hits from global protein abundance analysis of the changes during Mtb infection of AMs, BMDMs, or both cell types. The enrichment analysis was performed with Metascape [59].

https://doi.org/10.1371/journal.ppat.1012829.g005

Our previous global protein abundance measurements from cell lysate samples revealed that Tax1bp1 protein abundance increased by >2.5 fold in Mtb-infected wild-type BMDMs relative to mock-infected BMDMs after 24 hours [45]. Therefore, to compare with our previous analysis in BMDMs, we performed a similar proteomic analysis of five independent biological replicate wild-type AM cell lysate samples with the same Mtb strain and at the same multiplicity of infection (M.O.I) that were harvested at the same time point. Principal component analysis (PCA) revealed clustering of the data from the mock- or Mtb-infected replicate samples, indicating consistent changes in the data (S12 Fig). Using a log₂(fold change Mtb-infected vs. mock-infected) threshold of >1 or <-1 and adjusted p value cut-off of 0.05, 144 proteins changed in protein abundance during Mtb infection of AMs (Fig 5A). As displayed in the Venn diagram, 70 of the proteins that changed in abundance in a statistically significant manner during Mtb infection of AMs were also identified during infection of BMDMs (Fig 5B). In contrast to our results in BMDMs, Tax1bp1 protein level did not significantly change during Mtb infection of wild-type AMs (log₂(fold change) of 0.06, adjusted p value of 0.78; Fig 5A).

To determine the functional pathways that changed during Mtb infection, we performed gene ontology biological process enrichment analysis of the 70 proteins that changed in abundance significantly during infection in both AMs and BMDMs (Fig 5B). Then, to probe cell type-specific responses, we analyzed the 74 and 194 proteins that changed in abundance only in AMs and BMDMs, respectively (Fig 5B). The top six functional pathways identified in each group from the gene ontology enrichment analysis are displayed (Fig 5C). Consistent with Mtb’s ability to elicit anti-viral responses in macrophages [21], the response to virus pathway was the most significantly enriched pathway in both AMs and BMDMs (Fig 5B) and included proteins encoded by key interferon stimulated genes including Isg15 [53], Oasl2 [54] (analogous to human Oasl [55]), Ifit1, Ifit2, and Ifit3 [60]. Pathways involved in DNA damage repair (mismatch repair directed by MSH2:MSH6), DNA replication, and ribosome biogenesis (parvulin-associated pre-rRNP complex) were significantly enriched in the analysis of proteins that changed in abundance during Mtb infection of BMDMs exclusively (Fig 5C). The proteins that changed in abundance only in AMs were encoded by genes enriched in viral and interferon response pathways, similar to those that significantly changed in both AMs and BMDMs (Fig 5C). In summary, although gene ontology enrichment analysis suggests that there are overlapping pathways in the host response to Mtb in both cell types, at the individual protein level there are differences in the responses between AMs and BMDMs. In particular, the increase in Tax1bp1 protein abundance during Mtb infection is cell-type specific, occurring in BMDMs but not AMs. The proteomic analysis highlights mechanistic differences in the host responses to Mtb in these two cell types.

Host and pathogen expression analysis of Mtb-infected Tax 1bp1^-/- AMs ex vivo.

We exploited the observation that Mtb infection of AMs ex vivo recapitulates the phenotype seen in vivo in Tax1bp1-deficient mice by using an unbiased transcriptomics approach to develop hypotheses about host responses regulated by Tax1bp1. To broadly query host effector responses in Tax1bp1-deficient cells and simultaneously determine if the lack of Tax1bp1 triggers downregulation of Mtb genes that support intracellular Mtb replication, we performed host and pathogen dual transcriptional profiling of Mtb-infected wild-type and Tax 1bp1^-/- AMs. Differential gene expression analysis of Mtb transcripts showed no significant changes in gene expression using an adjusted p value cutoff of 0.05. However, using a less stringent p value threshold for statistical significance of an unadjusted p value of 0.05, similar to other reports of Mtb transcriptional profiling [61], Tax1bp1-deficiency downregulated two Mtb genes, mmpL4 and mbtE, following Mtb infection (S13 Fig). While the former is required for intracellular Mtb replication [62], we conclude that Tax1bp1-deficiency does not lead to major changes in the Mtb transcriptional profile at this time point of 36 hours post-infection. One possible explanation for the discrepancy between changes in Mtb growth caused by Tax1bp1-deficiency and the lack of major Mtb transcriptional changes is that our transcriptional profiling was performed before significant changes in Mtb growth were observed, beginning at 48 hours post-infection (Fig 4). Another possibility is that Tax1bp1-deficiency led to changes in host responses that increase the killing of Mtb by host cells because this would not be expected to produce dramatic changes in the bacterial transcriptional profile.

Consistent with the idea that Tax1bp1-deficiency significantly alters host responses to Mtb, we identified 140 differentially expressed host genes with a log₂(fold change) of greater than 1 or less than -1 and an adjusted p value <0.05 comparing wild-type and Tax 1bp1^-/- AMs infected with Mtb (S13 Fig). Gene ontogeny enrichment analysis of the up- and down-regulated genes identified cytokines and inflammatory response (20 genes, -log (p value) of 4.8) as a significant functional pathway controlled by Tax1bp1 in Mtb-infected AMs. Genes in the cytokines and inflammatory response pathway included genes downregulated (e.g., Cxcl1 [63]) and upregulated (e.g., Cd4 [64], Pf4 [65], Pdgfa [56], Kitl [57], and Cxcl3 [66]) in Tax 1bp1^-/- AMs (Figs 6 and S13). Gene ontogeny enrichment analysis also identified positive regulation of the intrinsic apoptotic signaling pathway (16 genes, -log (p value) of 4.1), including the genes encoding for Bok, an apoptotic regulator, and prostaglandin-endoperoxide synthase 2 (Ptgs2, also known as COX-2; S13 Fig). Most of the genes in this pathway, including Mmp2 [67], Nck2 [67], Bok [68], Gsdme [69], Cam2kb [58], Eya4 [70], Pmp22 [71], and Rgcc [72], were upregulated in Tax 1bp1^-/- AMs. However, Ptgs2 [73], Clu [61], Ctgf [62], and Cxcl1 [63] were four other genes in this pathway that were negatively regulated in Tax 1bp1^-/- AMs. The differentially expressed gene of the greatest magnitude upregulated in Mtb-infected Tax 1bp1^-/- AMs compared to wild-type AMs was Sox7 (S13 Fig; log₂(fold change) of 5.4, adjusted p value of 2.94 x 10^-12). Sox7 is a transcription factor that induces apoptosis through the MAP kinase ERK-BIM (BCL2-interacting mediator of cell death) pathway [74]. In summary, gene ontogeny enrichment analysis suggests that Tax1bp1-deficiency regulates the expression of inflammatory signaling and host cell death genes, both of which can contribute to the control of Mtb growth. Therefore, we further tested the hypothesis that Tax1bp1-deficiency impacts these two host responses during Mtb infection of AMs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Tax1bp1-deficiency results in differential expression of inflammatory response and apoptotic signaling pathway genes during Mtb infection of AMs.

Wild-type and Tax1bp1^-/- AMs were infected in biological triplicate with Mtb at a M.O.I. of 2. The RNA was harvested at 36-hours post-infection for differential pathogen and host gene expression analysis by RNAseq. Gene ontogeny enrichment analysis of statistically significant differentially expressed host genes (log₂(fold change) >1 or <-1, adj. p values <0.05) during Mtb infection of wild-type and Tax1bp1^-/- AMs was performed with Metascape [59]. Among this analysis of the up- and down-regulated genes triggered by Tax1bp1-deficiency, the top ten enriched pathways and the number of genes in each functional pathway are displayed.

https://doi.org/10.1371/journal.ppat.1012829.g006

Tax1bp1-deficiency reduces inflammatory cytokine signaling during AM infection ex vivo.

Given that inflammatory responses are intricate and the impact of particular genes or proteins can vary based on the cell type and immune environment [75], we next investigated whether the effects of Tax1bp1-deficiency on cytokines and inflammatory responses during AM infection ex vivo aligned with our observations in vivo (Figs 1C, 1D, and S1). In our RNAseq analysis, there were trends towards increased gene expression of IL-1β(2.3 fold change in wild-type vs. Tax 1bp1^-/- AMs, adjusted p value of 0.073) and interleukin-1 receptor type 1 (Il1r1; 2.1 fold change in wild-type vs. Tax 1bp1^-/- AMs, adjusted p value of 0.053; S13 Fig) in wild-type AMs.

To determine if Tax1bp1-deficiency impacts cytokine responses at the protein level and to elicit more robust cytokine responses, we next performed Mtb infections at a higher M.O.I. of 5. Tax1bp1-deficiency led to decreased levels of several inflammatory cytokines, including IL-1β, but not IFN-β, in infected cell supernatants (Fig 7A). As previously noted, Tax1bp1-deficiency also led to decreased expression of Ptgs2 (prostaglandin-endoperoxide synthase 2; 4.2 fold change in wild-type vs. Tax 1bp1^-/- AMs, adjusted p value of 0.01; S13 Fig), which is involved in the production of inflammatory prostaglandins. Notably, prostaglandin E₂ (PGE₂) is an inflammatory eicosanoid that decreases Mtb killing by blocking efferocytosis [76,77]. In addition to reducing levels of cytokines regulated by NF-κB, Tax1bp1-deficiency also reduced the production of PGE₂ during Mtb infection of AMs (Fig 7A). To our knowledge, PGE₂ was not previously known to be regulated by Tax1bp1. Importantly, these cytokine and eicosanoid levels were measured early after infection (24 hours) when the Mtb CFU were the same in wild-type and Tax 1bp1^-/- AMs (Fig 7B), before any differences in Mtb growth were observed at later time points. Therefore, these findings suggest that the increased cytokine and PGE₂ production mediated by Tax1bp1-deficiency happens independently of bacterial burden. In summary, Tax1bp1-deficiency decreased the production of proinflammatory cytokines and eicosanoids during Mtb infection of AMs. These findings are consistent with our cytokine analysis during in vivo infection (Figs 1 and 2) and in contrast to our previous report in BMDMs, in which Tax1bp1-deficiency did not impact inflammatory cytokine production during Mtb infection [45].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. IL-6, IL-1β, and PGE₂ production was abrogated during AM infection of Tax1bp1^-/- AMs.

AMs were seeded at 100,000 cells/well and infected in triplicate wells with Mtb at a M.O.I. of 5. At 24 hours post-infection, (A) the supernatants were collected for cytokine measurement by ELISA, and (B) AM monolayers were lysed and plated for Mtb CFU. Mean, SEM, and p values from the t-test are displayed.

https://doi.org/10.1371/journal.ppat.1012829.g007

Tax1bp1 enhances necrotic-like cell death and delays apoptosis of Mtb- infected AMs.

In addition to regulating cytokine and inflammatory responses, our gene expression analysis suggested that Tax1bp1-deficiency regulates apoptotic gene expression during AM infection. Tax1bp1 was previously shown to negatively regulate apoptosis following cytokine stimulation [78] and viral infection [43]. Furthermore, apoptosis is a mode of cell death that leads to killing of Mtb through efferocytosis, in which Mtb-infected apoptotic cells are phagocytosed by neighboring macrophages [79, 80]. This is in contrast to necrosis, which is a form of uncontrolled cell death that is highly immunostimulatory and enhances Mtb growth [81–85].

To test if Tax1bp1-deficiency impacts cell death during Mtb infection of AMs and BMDMs, we analyzed uninfected control cells and Mtb-infected cells by live cell fluorescence microscopy using CellEvent caspase 3/7 to detect apoptotic cells and propidium iodide (PI) for necrotic/late apoptotic cells. In the absence of IFN-γ stimulation, Tax1bp1-deficiency led to a modest decrease in necrotic-like cell death of AMs on days 1–4 post-infection (Figs 8A, 8B and S14). In IFN-γ stimulated AMs, Tax1bp1-deficiency accelerated apoptosis on day 3 post-infection (Figs 8C, 8D and S15). In contrast to AMs, Tax1bp1-deficiency did not impact the amount of apoptotic or necrotic-like cell death in unstimulated BMDMs (S15A-S15C, S16A and S16B Figs). Only when stimulated with IFN-γ, Tax1bp1-deficiency accelerated apoptotic cell death during Mtb infection of BMDMs on day 4 post-infection (S15 Fig). Together, these results show that Tax1bp1-deficiency reduces necrotic-like cell death of Mtb-infected AMs (Fig 8E) but not BMDMs, and Tax1bp1-deficiency advances the timing of apoptosis during Mtb infection of IFN-γ-stimulated AMs and BMDMs. Only the latter phenotype in IFN-γ-stimulated Tax 1bp1^-/- AMs was not specific to Mtb-infection because we also observed enhanced apoptosis in the uninfected IFN-γ-stimulated Tax 1bp1^-/- AMs (S16 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Necrotic-like cell death was inhibited and apoptosis accelerated in Mtb-infected Tax1bp1^-/- AMs.

AMs were infected with Mtb at a M.O.I. of 1 in the presence of CellEvent caspase 3/7 to detect apoptotic cells and propidium iodide (PI) for necrotic/late apoptotic cells without (A, B) or with (C, D) IFN-γ added to the media. Fluorescence images were obtained at 20X magnification in two positions per well in three replicate wells. (A, C) Representative fluorescence and brightfield microscopy images were merged, cropped, and scaled. (B, D) The number of fluorescent cells in each field was quantified in the green (CellEvent) and red fluorescence (PI) channels. Fluorescent cell numbers were normalized by the total number of cells in the phase contrast image for each field. Mean, SEM, and statistically significant FDR-adjusted p values from t-test comparisons are displayed. For clarity, only statistically significant p values (p < 0.05) are shown. The white bar depicts 100 μm. Data are representative of three independent experiments. (E) Tax1bp1-deficiency restricts Mtb growth, inflammatory cytokine synthesis, PGE₂ production, and necrotic-like host cell death in AMs.

https://doi.org/10.1371/journal.ppat.1012829.g008

Cytokine measurements.

Cytokines from lung lysates were measured using DuoSet ELISA kits (R&D systems) following the manufacturer’s protocol. Interferon levels were measured from the Mtb-infected lung lysates using L929 ISRE-luciferase reporter cells as previously described [110]. Luciferase reporter cells were seeded in a 96-well plate for 24 hours. Lung lysates were incubated with the reporter cells for 8 hours, and luciferase activity was measured with the Luciferase Assay Report Assay (Promega) using the manufacturer’s protocol.

Histology sample processing and quantitative analysis.

Formalin-fixed specimens were washed three times in PBS and stored in 70% ethanol. Histologic processing was performed by Histowiz. Serial ultrathin sections were stained for hematoxylin & eosin (H&E), ubiquitin (anti-ubiquitylated antibody, AB1690, EMD-Millipore, 1:100 dilution), tuberculosis (Abcam ab214721, 1:1000 dilution), or myeloperoxidase (Abcam ab9535, 1:50 dilution). Primary antibodies were detected with 3,3’-diaminobenzidine (DAB) staining. The ubiquitin and tuberculosis IHC images were aligned and combined using image registration scripts in QuPath. The MPO staining analysis was analyzed using Indica Labs Halo image analysis software. Cells were segmented using the Multiplex IHC algorithm v3.1.4 in Halo and MPO-positive cells were determined by thresholding the DAB channel. Positive cells were further sub-divided into Low intensity, Medium Intensity and High intensity bins to allow for subsequent calculation of a H-score based on percentage of cells positive for Low, Medium and High using the following formula: H-Score = (1 x % positive cells low) + (2 x % positive cells medium) + (3 x % positive cells high). Cells in all the intensity bins were considered positive for myeloperoxidase staining.

Images of tuberculosis lesions were exported from QuPath into ImageJ. A minimum threshold of two standard deviations above the mean signal was applied to filter positive pixels for ubiquitin and tuberculosis immunohistochemistry images. Colocalization was calculated from pixel overlap in images of the tuberculosis and ubiquitin immunohistochemistry. A veterinary pathologist analyzed the H&E histopathology images.

M. tuberculosis mouse infections at UC San Francisco.

Tax1bp1^-/- mice were imported from UC Berkeley and rederived by the UC San Francisco Rederivation Core to eliminate the potential for any interinstitutional murine pathogen transmission. Low-dose aerosol infections (100–200 CFU) of age- and sex-matched wild-type and Tax1bp1^-/- mice were performed with a Glass-col inhalation chamber. At the indicated time points, mice were euthanized with CO₂, and their lungs were minced with scissors and digested in 3 ml of RPMI-1640 with 5% heat-inactivated FBS containing 1 mg/ml collagenase D (Sigma) and 30 µg/ml DNAseI (Sigma) for 30 min at 37°C. Cells were processed with a gentleMACS dissociator (Miltenyi Biotec, lung program 2) and filtered through a 70 μm strainer. The samples were rinsed with 1 ml of FACS buffer (PBS with 3% heat-inactivated FBS, 2 mM EDTA). Residual tissue on the cell strainer was further processed using a syringe plunger and rinsed with 1 ml of FACS buffer. The cell suspension was then centrifuged at 650 × g for 3 minutes at 4°C, and the supernatant was discarded. The cell pellet was resuspended in 3 ml of ACK lysis buffer (Gibco) to lyse the RBCs, and lysis was quenched with 3 ml of FACS buffer solution. After centrifuging the cell suspension at 650 × g for 3 minutes at 4°C, the supernatant was removed, and the cells were resuspended in 1 ml of FACS buffer. Each cell suspension was pooled (from 5 mice) and passed through a 50-µm strainer.

Single-cell lung suspensions were stained with the Zombie Aqua Fixable Viability kit (1:200 dilution, BioLegend, #423101) and treated with CD16/CD32 Fc block (1:100 dilution, BD 553142) in PBS (1 ml) for 15 min. The samples were centrifuged at 650 × g for 3 minutes at 4°C, and the supernatants were removed. Cells were stained with 2 ml of antibody mixture diluted in Brilliant Stain Buffer (Table 1; Invitrogen #00-4409-42) for 30 min at 4°C. Antibodies diluted in Brilliant Stain Buffer (BD, #566349) were added to the cells. Antibody staining was performed for 30 minutes at 4°C. Subsequently, the cell suspensions were centrifuged at 650 × g for 3 minutes at 4°C, washed with 1 ml of FACS buffer solution, resuspended in 3 ml of FACS buffer, and passed through a 50-μm strainer. Cell subsets were sorted using a BD Aria Fusion Sorter through a 100 µm nozzle using the 4-way purity mode.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Key Resources.

https://doi.org/10.1371/journal.ppat.1012829.t001

Bacteria were quantified from sorted cells by serial dilution in PBS containing 0.05% Tween80 and plated on 7H10 agar plates supplemented with 10% Middlebrook OADC, 0.5% glycerol, and PANTA antibiotic mixture at a 1:500 dilution to reduce contamination risk from non-mycobacteria during organ dissection. BD PANTA antibiotic mixture (BD, B4345114) containing polymyxin B, amphotericin B, nalidixic acid, trimethoprim, and azlocillin was prepared by dissolving the contents of 1 lyophilized vial in 3 ml of OADC.

Bone marrow-derived macrophage infection.

Bone marrow-derived macrophage infections with Listeria monocytogenes 10403S were performed as previously described [112–114].

Peritoneal cell exudate infection.

Approximately 7 ml of ice-cold PBS was injected into the peritoneum of euthanized mice. Peritoneal exudate cells were treated with ACK lysis buffer, resuspended in tissue culture cell media (RPMI supplemented with 1 mM L-glutamine and 10% fetal bovine serum), and seeded into 24-well plates with glass coverslips at a density of 1.5 × 10⁶ cells/ml for 24-hours prior to infection. Prior to infection, non-adherent cells were removed by replacement of the tissue culture media.

Listeria monocytogenes 10403S was inoculated from a single colony into BHI media. Following overnight incubation at 30 ˚C, bacteria were washed in PBS and resuspended to an OD of 1.5 in PBS. Bacteria were diluted 1:1000 in tissue culture cell media for infection of peritoneal exudate cells. 30 minutes post-infection, cells were rinsed twice with PBS, and fresh media was replaced. At 1 hour post-infection, gentamicin sulfate was added at a final concentration of 50 µg/ml to kill extracellular bacteria. At 2- and 8 hours post-infection, coverslips were placed in 5 ml of water and vortexed. Serial dilutions were plated on LB agar supplemented with streptomycin (200 µg/ml). CFU were enumerated after 24 hours of incubation at 37 ˚C. Infections were performed with 3 coverslips for each experimental condition.

L. monocytogenes mouse infections.

Age and sex-matched male and female mice were infected with L. monocytogenes. A 2 ml overnight culture of L. monocytogenes grown in brain heart infusion media at 30 ˚C slanted. L. monocytogenes was subcultured in 5.5 ml of BHI media and incubated with shaking at 37 ˚C until reaching an optical density between 0.4-0.8. The bacteria were washed and diluted in PBS to achieve an inoculum of approximately 5 × 10⁵ CFU/ml. 200 μl of this suspension was injected into the tail vein of the mice. For the intraperitoneal infection, the bacterial cells were prepared as described, and 200 µl of bacterial suspension was injected into the peritoneum at a dose of 3.74 × 10⁵ CFU/mouse. At the indicated time points, the spleen and liver were harvested in water containing 0.1% NP-40. Organs were homogenized, and CFU were enumerated on LB agar supplemented with streptomycin (200 µg/ml).

Alveolar macrophage (AM) isolation and culture.

AMs were harvested from naïve mice by bronchoalveolar lavage with 10 ml of PBS containing 2 mM EDTA, 0.5% fetal bovine serum (FBS) pre-warmed to 37°C as described previously [115, 116]. BAL cells were seeded at a density of 100,000 cells/well in 96-well plates. For experiments with sorted AMs, live, CD11c+SiglecF + AMs were sorted at a density of 50,000 cells/well in 96-well plates using the previously described protocol for antibody staining and gating [117]. For short-term cultivation up to 4 days, AMs were cultured in RPMI-1640 medium supplemented with 10% (v/v) FBS, 2 mM GlutaMAX, 10 mM HEPES and 100 U ml–1 penicillin–streptomycin. After allowing at least 2 hours for adhesion, the media was replaced with fresh media without antibiotics. For cultivation >4 days, AMs were cultured in RPMI-1640 medium supplemented with 10% (v/v) FBS, 2 mM GlutaMAX, 1 mM sodium pyruvate, and 2% (v/v) GM-CSF supernatant produced by a B16 murine melanoma cell line.

Macrophage infections with Mtb.

Mtb H37Rv strain was transformed with pMV306hsp-LuxG13 for expression of LuxCDABE. Logarithmic phase cultures of Mtb (H37Rv-Lux or wild-type Erdman strain) were grown in 7H9 media supplemented with 10% Middlebrook OADC, 0.5% glycerol, 0.05% Tween80 in inkwell bottles at 37°C with rotation at 100 rpm. Mtb cell pellets were washed twice with PBS followed by centrifugation for 5 min at 1462 × g and sonication to remove and disperse clumps. Mtb was resuspended in RPMI with 10% horse serum. Media was removed from the macrophage monolayers, the bacterial suspension was overlaid, and centrifugation was performed for 10 min at 162 × g. Following infection, the media was replaced with cultivation media with 15 ng/ml IFN-γ (Peprotech) or without IFN-γ. In experiments performed with luminescent Mtb, luminescence measurements were obtained daily following media changes daily using a GloMax microplate reader (Promega). For CFU measurements, the monolayers were washed with PBS, lysed in PBS with 0.05% Tween80, serially diluted, and spread on 7H10 agar plates supplemented with 10% Middlebrook OADC and 0.5% glycerol. CFU were enumerated after 21 days of incubation at 37°C.

Gene expression analysis during Mtb infection of AMs.

AMs were infected with wild-type Mtb Erdman at a M.O.I. of 2. At 36-hours post-infection, the monolayers were washed with PBS, and the AMs were lysed in 200 µl of Trizol reagent. Samples were pooled from four technical replicate wells. The experiment was performed independently three times (i.e., three independent biological replicates). Mtb was centrifuged, the supernatant containing host RNA was removed, and the Mtb pellet was resuspended in 400 µl of fresh Trizol and 0.1 mm zirconia/silica beads. Mtb was mechanically disrupted with the Mini Bead-Beater Plus (Biospec Products) as previously described [118]. 70% of the sample containing host RNA was pooled with the sample containing Mtb RNA, the samples were treated with 200 µl of chloroform, and RNA was purified with the Trizol Plus RNA purification kit (Ambion). Purified total RNA was treated with DNAseI (ThermoFisher) and dried by rotary evaporation in RNA stabilization tubes (Azenta US, Inc.; South Plainfield, NJ, USA). Sample QC, dual rRNA depletion for bacteria and mouse, library preparation, Illumina sequencing (2x150 bp; 30M reads per sample), and differential gene expression analysis were performed by Azenta Life Sciences US, Inc.

Cytokine analysis of Mtb-infected AMs.

AMs were infected with wild-type Mtb Erdman strain at a M.O.I. of 5. At 24 hours post-infection, the supernatants were filtered through a 0.2 μm syringe filter and analyzed by ELISA for IFN-β (PBL Assay Bioscience), TNF-α, and IL-1β (R&D systems) as previously described, and prostaglandin E₂ (Cayman Chemicals).

Live cell imaging.

Uninfected control macrophages and macrophages infected with Mtb at a M.O.I. of 1 were incubated in the presence of 0.1 µg ml⁻¹ of propidium iodide (LifeTechnologies) and two drops per milliliter of CellEvent Caspase-3/7Green ReadyProbes reagent (Invitrogen) to measure necrosis/late apoptosis and apoptosis, respectively. After the initial media change on the day of Mtb infection, as previously performed [85], the cell culture media was not changed during the remainder of the experiment to avoid disruption of the macrophage monolayers. Fluorescence and phase contrast images were obtained at 20x magnification with a Keyence BZ-X 700 microscope. Images were obtained daily in three technical replicate wells per condition and at two positions in each well. Quantification of the number of necrotic and apoptotic cells was performed with ImageJ version 1.54f as described previously [85]. Images were converted to 8-bit (grayscale), binarized, and enumerated using the analyze particles module (size threshold 0.001-infinity). Cell numbers from phase contrast microscopy images were enumerated using Cell Profiler.

Immunofluorescence microscopy.

AMs were infected with fluorescent Mtb at a M.O.I. of 2. At 8- and 24-hours post-infection, monolayers were washed with PBS, fixed with 4% PFA for 20 minutes, washed with PBS, and stained with anti-LC3 or anti-ubiquitin primary antibodies and AlexaFluor-647 conjugated secondary antibodies as previously described [44]. Images were obtained at 63x magnification from quadruplicate wells per condition, in 69 x/y positions, and 4 z positions (0 µm, 0.5 µm, 1 µm, and 1.5 µm) with an Opera Phenix microscope (Perkin Elmer). Colocalization analysis of LC3, ubiquitin, and Mtb was performed with Harmony version 4.9 (Perkin Elmer) using the following analysis parameters. The four z stack images in each x/y position were processed into a maximum intensity projection. Nuclei were identified in the DAPI channel using Method B with a common threshold of 0.07 and an area threshold of > 20 µm². Cytoplasm was identified in the AlexaFluor 647 channel using Method A with an individual threshold of 0.06. The find spot module was used to identify LC3 or ubiquitin “spots” in the AlexaFluor 647 channel using method C with a contrast setting of 0.42, uncorrected spot to region intensity of 3.8, and default radius. Mtb were identified in the AlexaFluor 488 channel using the find spot module method B with a detection sensitivity of 0.5 and splitting sensitivity of 0.5. To identify Mtb that colocalized with LC3 or ubiquitin “spots”, the select population module was used for the Mtb population with the select by mask method. The percent colocalization was calculated for each well from all the images obtained in each well using the evaluation module.

LC3 immunoblot.

AMs and BMDMs were either incubated in fresh culture media or Earle’s Balanced Salt Solution (EBSS) for 8 hours following a previously described protocol [85]. For cells treated with EBSS, 300 nM bafilomycin was added for 2 hours prior to the completion of treatment. Cell lysates were generated by adding RIPA buffer containing mini-complete protease inhibitor cocktail (Roche). Lysate protein concentration was determined by BCA assay. 15 μg of protein from each sample was separated by SDS-polyacrylamide gel electrophoresis (NuPage 4–12% Bis-Tris Mini-protein gels) and transferred to PVDF membranes. After blocking with Intercept (PBS) blocking buffer (LicorBio #927–70001), membranes were probed with anti-LC3 antibodies (Sigma L7543, dilution 1:1000) followed by IRDye 800CW Goat anti-rabbit IgG secondary antibody (LicorBio #926–32211, dilution 1:10,000). The membrane was then probed anti-pan actin antibodies (Cell Signaling #4968, dilution 1:1000) followed by IRDye 800CW Goat anti-rabbit IgG secondary antibody. Membrane images were obtained on a Li-Cor Odyssey system. Band densities for LC3-I and LC3-II were measured using ImageJ (National Institutes of Health).

AM infections for global protein abundance analysis.

BAL was performed on 14 mice per biological replicate experiment. BAL cells were pooled and seeded at a density of 40e⁴-60e⁴ cells/well in 4-well Omniwell plates in AM media containing GMCSF. Media was changed daily. After 7 days of incubation at 37°C with 5% CO₂, half of the AMs were infected with Mtb Erdman strain at a M.O.I. of 10. The other half of the AMs were mock-infected by changing the media with RPMI-1640 with 10% heat-inactivated horse serum and centrifuged simultaneously with the Mtb-infected AMs. After centrifugation, the media was changed to AM media. After 24-hours incubation at 37°C with 5% CO₂, the monolayers were washed with PBS, and AMs fixed on the plates with 100% methanol for 20 minutes. The monolayers were washed twice with PBS. AMs were lysed with 6 ml of lysis buffer (8 M urea, 150 mM NaCl, 100 mM ammonium bicarbonate, pH 8; added per 10 ml of buffer: 1 tablet of Roche mini-complete protease inhibitor EDTA free and 1 tablet of Roche PhosSTOP tablet) prepared fresh before each replicate experiment. Lysates were stored at −80°C until further processing.

Trypsin digest and desalting.

A bicinchoninic acid assay (Pierce) was performed to measure protein concentration in cell lysate supernatants. 1 mg of each clarified lysate was reduced by the addition of dithiothreitol (DTT) to a 5 mM final concentration for 30 min at room temperature and alkylated by the addition of iodoacetamide to 10 mM final concentration for 30 min at room temperature in the dark. Remaining alkylating agent was quenched by the addition of DTT to 10 mM final concentration for 30 min at room temperature in the dark. The samples were diluted with 100 mM ammonium bicarbonate, pH 8.0, to reduce the urea concentration to below 2M. Samples were incubated with sequencing grade modified trypsin (Promega) at a 1:25 enzyme:protein ratio and LysC (Wako) at a 1:100 enzyme:protein ratio overnight at 37°C with rotation. Prior to desalting, the sample pH was reduced to approximately 2.0 by the addition of 10% trifluoroacetic acid (TFA) to a final concentration of 1% trifluoroacetic acid. Peptides were desalted using 50 mg SepPak C18 solid-phase extraction cartridges (Waters). The columns were activated with 3 ml of 80% acetonitrile (ACN) 0.1% TFA, and equilibrated with 3 ml of 0.1% TFA. Peptide samples were applied to the columns, and the columns were washed with 3 ml of 0.1% TFA. Peptides were eluted with 1.1 ml of 40% ACN, 0.1% TFA. 10 µg of peptides analyzed by LC-MS/MS for global protein abundance measurement.

Liquid chromatography and mass spectrometry.

Following digestion, peptides were separated on a PepSep reverse-phase C18 column (1.9 μm particles, 15 cm, 150 mm ID) (Bruker) with a gradient of 3–28% buffer B (0.1% formic acid in 80% acetonitrile) over buffer A (0.1% formic acid in water) over 67 minutes, an increase to 40% B in 5 minutes, and held at 95% B for 8 minutes. Eluting peptide cations were analyzed by electrospray ionization on an Orbitrap Exploris (Thermo Fisher Scientific). For DIA analysis, MS1 scans of peptide precursors were performed at 120,000 resolution (200 m/z) over a scan range of 350–1050 m/z, with an AGC target of 250% and a max injection time of 100 ms. MS2 scans were collected over 350–950 m/z in 15 m/z isolation windows with a 0.5 m/z overlap. Maximum injection time was set to auto, with AGC set to standard, and an MS2 resolution of 15,000. Higher energy collisional dissociation (HCD) was performed at a NCE of 28%. Six gas-phase fractions (GPF)-DIA runs were collected from a pooled sample of all conditions [119]. tSIM MS1 scans were performed at 120,000. tMS2 scans were collected at 30,000 fragment resolution, an AGC target of 1e6, a maximum ion injection time of 60 ms, and a NCE of 26. Data was collected using 4m/z precursor isolation windows in a staggered-window pattern, with each GPF fraction covering an approximately 100 m/z range.

Global protein abundance analysis.

The GPF mass spectrometry data was searched against the Uniprot mouse database appended with the Erdman Mtb strain in Spectronaut to build a spectral library of detected peptides and proteins (Biognosys, version 19.0). Default settings, including trypsin digestion, variable modifications of methionine oxidation and N-termini acetylation, and fixed modification of cysteine carbamidomethylation, were used. DIA mass spectrometry data was then searched in Spectronaut against the GPF-generated spectral library. DIA runs were filtered to obtain a false discovery rate of 1% at the peptide spectrum match and protein level [120]. Quantitative analysis was performed in the R statistical programming language using the artMS package (release 3.20) and quantitative packages from MSstats (version 3.20) [121]. For MSstats analysis, the normalization method was set to equalizeMedians, and censored missing values were imputed by the Accelerated Failure Time mode. Statistically significant changes were selected by applying a log2-fold-change (>1.0 or<−1.0) and an adjusted p value (<0.05) [122].

Statistical analysis.

GraphPad Prism (v.10.3.1) was used for statistical analysis. Unpaired t-test comparisons were calculated assuming Gaussian distributions, and the p values were reported. In experiments with more than two experimental conditions, p values from the t-test comparison between two groups were adjusted for the FDR (multiple comparisons) using the two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli (Figs 4, 8, S15, and S16).