Early innate and adaptive immune perturbations determine long-term severity of chronic virus and Mycobacterium tuberculosis coinfection

Summary

Chronic viral infections increase severity of Mycobacterium tuberculosis (Mtb) coinfection. Here we examined how chronic viral infections alter the pulmonary microenvironment to foster coinfection and worsen disease severity. We developed a coordinated system of chronic virus and Mtb infection that induced central clinical manifestations of coinfection, including increased Mtb burden, extra-pulmonary dissemination and heightened mortality. These disease states were not due to chronic virus-induced immunosuppression or exhaustion; rather, increased amounts of the cytokine TNFα initially arrested pulmonary Mtb growth, impeding dendritic cell mediated antigen transportation to the lymph node and subverting immune-surveillance, allowing bacterial sanctuary. The cryptic Mtb replication delayed CD4 T cell priming, redirecting T helper (Th) 1 toward Th17 differentiation and increasing pulmonary neutrophilia which diminished long-term survival. Temporally restoring CD4 T cell induction overcame these diverse disease sequelae to enhance Mtb control. Thus, Mtb co-opts TNFα from the chronic inflammatory environment to subvert immune-surveillance, avert early immune function and foster long-term coinfection.

Graphical Abstract

eTOC Blurb

Pre-existing chronic viral infection promotes subsequent co-infection with Mycobacterim tuberculosis (Mtb). Xu et al. provide insight in to early events of Mtb co-infection, revealing that chronic viral infection alters the timing and development of the Mtb-specific immune response, thus driving the multiple disease sequelae that worsen Mtb co-infection.

Introduction

Mycobacterium tuberculosis (Mtb) and Mtb coinfections are among the greatest global health problems. WHO estimated that one-quarter of the world population is infected with Mtb (Cohen et al., 2019), among whom there were annually ten million active tuberculosis patients and 1.4 million deaths in 2019 (WHO, 2020). Many chronic viral infections, including HIV, Hepatitis B and C viruses (HBV, HCV), human T lymphotropic virus type 1, and Cytomegalovirus (CMV) substantially increase the risk of coinfection by Mtb, heighten disease severity, accelerate disease progression, and decrease overall survival (Chen et al., 2018; Esmail et al., 2018; Grassi et al., 2016; Muller et al., 2019; WHO, 2020; Wu et al., 2015). In particular, pre-existing chronic HIV infection impairs subsequent Mtb control, which is generally attributed to HIV-induced CD4 T cell depletion (Esmail et al., 2018). This as a sole cause is unlikely though since HIV patients with high CD4 T cell counts still have increased Mtb susceptibility and disease progression (Bucsan et al., 2019; Esmail et al., 2018; Gupta et al., 2012), and many chronic viral infections that do not deplete CD4 T cells also lead to more severe Mtb disease (Chen et al., 2018). Further, non-infection driven diseases characterized by chronic inflammation, such as diabetes, can worsen secondary Mtb infections (Baker et al., 2011; Kumar et al., 2013a), suggesting the concept that conserved secondary alterations induced by chronic inflammatory diseases, but not the specific etiologic agent or disease itself per se, underlie the worsened Mtb disease progression.

Chronic viral infections simultaneously induce inflammatory and suppressive molecules that drive widespread immune dysfunction to prevent viral elimination (Bucsan et al., 2019; Lukhele et al., 2019). The production of inflammatory cytokines like TNFα and IFNγ are critical to protect against long-term rampant Mtb growth (Mayer-Barber and Barber, 2015) and loss of these factors leads to heightened Mtb replication and rapid death (Mogues et al., 2001). Consequently, the diminished expression of TNFα and IFNγ and immune exhaustion driven by chronic viral infections is another mechanism proposed to foster Mtb coinfection (Stelekati and Wherry, 2012). On the other hand, chronic infections, like HIV, can sustain pulmonary inflammation, with increased TNFα production (Costiniuk and Jenabian, 2014; Israel-Biet et al., 1991; Jambo et al., 2014; Millar et al., 1991) and smoldering type I interferon (IFN-I) production (Lukhele et al., 2019) potentially driving active Mtb disease and reactivation of latent Mtb (Berry et al., 2010). To further add complexity, chronic viruses enhance multiple immunosuppressive factors, such as IL-10 and PD-L1 that can promote Mtb infection (Mayer-Barber and Barber, 2015). Thus, how the complex interaction of these inflammatory and suppressive factors spread to impede Mtb-specific immunity and promote coinfection remains unclear.

The lung is the initial and primary site for Mtb infection. Mtb first infects and replicates in alveolar macrophages (AMs) and then disseminates to other phagocytes, including dendritic cells (DCs), neutrophils and interstitial macrophages (Cohen et al., 2018). Pulmonary CD11b+ DCs (MHC-II^hiCD11c^hiCD11b+) including monocyte derived DCs, internalize Mtb and migrate to the lung-draining mediastinal lymph nodes (medLNs) to prime Mtb-specific T cells (Samstein et al., 2013; Wolf et al., 2008). Mtb grows in the lungs during the first two weeks prior to the induction of T cell immunity, but is then controlled with the arrival of Mtb-specific CD4 Th1 cells to the lungs (Mogues et al., 2001). In addition to Th1 responses, Mtb infection also induces CD4 Th17 and Treg cells, however their precise roles in Mtb control are controversial, since both can be protective and pathogenic in different conditions (Cardona and Cardona, 2019; Khader et al., 2007; Mourik et al., 2017). To investigate how chronic infection promotes and worsens Mtb coinfection, we established a new model of chronic lymphocytic choriomeningitis virus (LCMV) and Mtb coinfection that recapitulated many clinical manifestations of Mtb coinfection, including elevated pulmonary Mtb burden, increased extra-pulmonary dissemination to spleen and diminished survival compared to Mtb infection alone. Unexpectedly, the long-term disease outcome of Mtb coinfection was determined within the initial days following coinfection. We demonstrated that the heightened levels of TNFα (but not other inflammatory factors) induced by chronic viral infection initially diminished Mtb pulmonary growth, essentially allowing Mtb to go “unsensed” by the immune system and inhibiting Mtb transport to the medLNs. The inhibited antigen arrival to medLNs substantially delayed activation and priming of Mtb-specific T cells, allowing Mtb to replicate to higher set-points before T cell mediated control could be initiated. Once primed, Mtb-specific CD4 T cell differentiation in co-infected mice skewed away from the Th1 responses associated with Mtb control and toward Th17 differentiation. The delayed and redirected T cell differentiation was associated with increased Mtb burden and promoted pulmonary neutrophil infiltration, leading to decreased survival. Therapeutically correcting the timing of CD4 T cell priming re-established CD4 Th1 over Th17 dominance, and diminished pulmonary neutrophilia, providing enhanced Mtb control in the presence of chronic viral coinfection.

Results

Pre-existing chronic viral infection impedes M. tuberculosis control and increases mortality

A major obstacle limiting the understanding of the mechanisms that promote Mtb coinfection during chronic viral infections, and particularly the earliest events in coinfection that set the stage for the long-term disease outcome, has been the lack of a clinically relevant and tractable model system. We developed a coinfection model combining chronic LCMV and Mtb infections. Mice were first chronically-infected with LCMV-Cl13 and then 21 days later co-infected with low dose Mtb H37Rv [100 colony forming units (CFU), aerosol; Fig. 1A]. In parallel, mice infected with Mtb alone served as the control group. In mice infected with Mtb alone there was no mortality until 261 days after infection and they exhibited a median survival time of 301 days, whereas LCMV-Cl13/Mtb co-infected mice began to die at 150 days after Mtb infection and exhibited a median survival time of 241 days (Fig. 1B), demonstrating heightened Mtb mortality in chronic virus coinfection.

Figure 1. Chronic virus infection increases Mtb burden and mortality following Mtb coinfection.

(A) Experimental design for LCMV:Mtb coinfection model.

(B) Survival of mice infected with Mtb alone or co-infected with chronic LCMV and Mtb, as described in (A).

(C) Survival of CD4 depleted/reconstituted mice infected with Mtb alone or confected, as described in Fig S1A.

(D) Mtb burden in the lungs and spleen at indicated time points after Mtb infection.

(E) Mtb burden in the lungs and spleen at the survival endpoints in (C). For the Mtb alone and coinfection groups, mice were sacrificed at the survival endpoints (Mtb: day 228 to 248 after Mtb infection; coinfection: day 36 to 165 after Mtb infection) For Mtb day 259 group, mice were sacrificed at day 259 after Mtb infection before the survival endpoint. See also Figure S1D.

(F) Representative images (top row) of hematoxylin and eosin (H&E)-stained lung sections of Mtb alone and coinfection at day 50 after Mtb infection or at survival endpoint. Bottom row; magnified images o f regions squared in top row. “Coinfection (the endpoint)”: one sick mouse in coinfection group sacrificed at survival endpoint (day 41 after Mtb infection). Scale bars: 2mm (top images); 200μm (bottom images). See also Figure S1E.

(G) Pathological scores of H&E stained lung sections. See also Figure S1E.

(H) LCMV plasma plaque forming units (PFU) in mice infected with LCMV alone or co-infected with LCMV and Mtb at day 50 after Mtb infection (~80 days after LCMV infection). Dashed line: detection threshold 200 PFU.

(I) Lung Mtb burden at day 35 after Mtb infection of mice infected with Mtb alone, LCMV-Cl13 : Mtb coinfection (Cl13+Mtb) or LCMV-Armstrong : Mtb coinfection (Arm+Mtb).

Data represent 2 or more independent experiments with 8-10 mice per group for (B, C), 4-6 mice per group for (D-I). *, p<0.05. NS: not significant. Log rank (Mantel-Cox test) for (B, C), Student’s t-test for (D), One-way ANOVA for (E, I). See also Figure S1.

Chronic LCMV-Cl13 infection is controlled in the majority of tissues (including the lungs) ~60 days after infection (Macleod et al., 2020; Matloubian et al., 1994). This corresponded to 40 days after Mtb infection in the coinfection system, suggesting that the determination of worsened Mtb outcome occurred early after Mtb coinfection. However, to develop a system in which life-long chronic viremia was sustained, we infected mice with LCMV-Cl13 two days after transient CD4 depletion treatment (Matloubian et al., 1994). A group of mice that remain LCMV uninfected were also CD4 depleted in parallel in order to control for any effects of depletion (Fig. S1A). At day 30 after depletion, CD4 T cells similarly reconstituted to 50-60% of their original amount in both naive and chronic LCMV infected mice (Fig. S1B). The CD4 depleted/reconstituted mice exhibited similar survival as non-depleted mice following Mtb infection, and they both had much enhanced survival compared to Mtb infected CD4-deficient mice (Fig. S1C), indicating that the system was not simply characterizing CD4-deficient mice and that the repletion of CD4 T cells functioned to control Mtb infection. In this CD4 depletion/reconstitution system the co-infected mice also succumbed to Mtb infection much more rapidly than mice infected with Mtb alone (Fig. 1C) and exhibited increased pulmonary Mtb burden within 30 days after Mtb infection and elevated extra-pulmonary infection to the spleen by day 50 (Fig. 1D). Mtb burdens then remained higher in the lungs and spleen of co-infected mice compared to the Mtb alone group at both late and endpoint times (Fig. 1D, 1E). There were approximately 20-30% of co-infected mice which reached the survival endpoint before day 50 (Fig. 1C) and these mice developed severe chronic inflammatory infiltration and lesions with large necrotic areas in the lungs at endpoint (Fig. 1F). The remaining co-infected mice exhibited mild to moderate pulmonary lymphocytic infiltration in the alveolar interstitium and structured lesions with similar histological characterization and evaluation to the mice with Mtb infection alone at day 50 (Fig. 1F, 1G). On the other hand, LCMV titers remained high throughout infection and exhibited no significant difference compared with LCMV infection alone (Fig. 1H). Further, previous acute infection with LCMV-Armstrong (which is cleared ~10 days after viral infection), did not alter subsequent Mtb burden in the lungs at day 35 after Mtb infection (Fig. 1I), indicating the heightened Mtb disease was specifically potentiated by chronic viral coinfection. Thus, (1) similar to non-depleted mice, the reconstitution following CD4 depletion is sufficient for the CD4 mediated Mtb control, (2) the increased Mtb CFU and mortality is not due to changes in chronic LCMV titers; and (3) chronic virus coinfection decreases survival compared to Mtb alone group in both the CD4 non-depleted and CD4 depleted/reconstituted mice.

Chronic virus infection remodels the pulmonary immune landscape prior to Mtb coinfection

The observation that co-infected mice had worse Mtb disease following coinfection in mice without CD4 depletion, despite the peripheral control of chronic LCMV replication within 40 days following Mtb infection, suggested that the early events following Mtb coinfection were responsible for the long-term disease consequences. To understand the immune alterations that underlie the inability to control Mtb coinfection, we quantified the pulmonary immune environment using time-of-flight mass cytometry (CyTOF). In total, 29 clusters were identified using ‘PhenoGraph’ (Levine et al., 2015b), comprising 11 major subsets of immune cells (Fig. 2A), as defined based on the protein expression patterns and gating strategy in Figure S2A and S2B. Compared to naive mice, multiple immune subsets were proportionally modulated by chronic viral infection, including increased Foxp3+ CD4 T cells [cluster (c) 3], PD1+ CD8 T cells (c5), monocytes (c21) and activated CD11b+DCs (c1), while NK cells (c22), CD103+ CD8 T cells (c23, c29) and multiple B cell clusters were decreased (Fig. 2B, S2A and S2B). In the presence of chronic LCMV infection, the enhanced monocyte recruitment results in increased monocyte derived dendritic cells (moDCs) which decrease Ly6C expression and upregulate CD11c and MHC-II (Cunningham et al., 2016; Samstein et al., 2013) were likely also included within c1. On the other hand, many populations like neutrophils and eosinophils were not proportionally different in the lungs during chronic infection (Fig. 2B). Interestingly, separate from the canonical Siglec-F^hi AMs (c16), a population of Siglec-F^low AMs (c4) emerged during chronic viral infection and now comprised 25% of the total AM population, whereas it was <1% in naive mice (Fig. 2B, S2A). A similar Siglec-F^low AM population has been associated with pulmonary inflammation and protection against secondary bacterial infection following acute influenza infection (Aegerter et al., 2020), and consistent with this, c4 compared to c16 was characterized by increased expression of multiple antigen presentation and costimulatory proteins, including MHC-I, MHC-II, CD80, CD86 (Fig. S2C).

Figure 2. Chronic viral infection reshapes the pulmonary immune landscape and cell states.

(A) Lung cell clusters in naive and day 30 after chronic LCMV infection. Cluster 2, 10 and 28 overlap with B cell clusters. Clusters are annotated by cell type (top) and cluster number (bottom).

(B) Lung cell clusters with the proportions significantly increased in chronic LCMV group (red) or naive (blue) mice. Clusters in gray: no significant proportion difference. The bottom bar graph shows the Log2 fold proportion change for each cluster between two groups. ***: p<0.0005, **: p<0.005, *: p<0.05.

(C) For indicated clusters, the colors represent the arcsinh ratio value of the change in median spectral indices of the indicated protein between naive and chronic LCMV group. ***: p<0.001, **: p<0.01, *: p<0.05.

(D) Distribution of the 9 myeloid/dendritic cell PhenoGraph clusters in the lungs (shown in Fig. 2C and gated in S2B) and their expression of indicated proteins between naive and chronic LCMV groups. UMAPs for the individual proteins are colored based on expression level, with scales matched between two groups for each protein.

Data represent 2 or more independent experiments with 4-5 mice per group. AMs = alveolar macrophages, Int Mac = interstitial macrophages, Other T = CD4 and CD8 double negative T cells (c24) and TCRβ-Thy1+ T cells (c25). p-values are calculated by the diffcyt R package for (B, C). See also Figure S2.

In addition to changes in the abundance of the distinct immune populations, differential protein expression states were also evident between chronic viral infection and naive mice within the same cell subsets, particularly AMs and CD11b+ DCs, the two main host cells of Mtb infection in vivo (Wolf et al., 2007). In the presence of chronic viral infection, the major AM cluster c16 had increased MHC-I, MHC-II, Tim3 and CD40 expression compared to their counterparts in naive mice (Fig. 2C, 2D, S2C), indicating increased single-cell activation within the same cell populations in the lungs. Similarly, the dominant pulmonary DC population in chronic infection (CD11b+ DCs, c1) exhibited elevated expression of MHC-I, PD-L1, CD80 and CD40, suggesting heightened DC activation, although both CD86 and MHC-II were slightly decreased (Fig. 2C, 2D, S2C). On the other hand, the CD11b+ DC cluster (c1), but not other cell populations, from chronic virus infected mice expressed lower CD209a (DC-SIGN) and CD206 (mannose receptor) (Fig. 2C, 2D, S2C), two receptors critical for Mtb internalization by DCs (Kang and Schlesinger, 1998; Tailleux et al., 2003), suggesting potential diminished capacity of CD11b+ DCs in particular to internalize and to bring Mtb to the draining lymph nodes. Despite the sustained inflammation in the mice with chronic LCMV infection, we did not observe induced bronchial associated lymphoid tissue (iBALT) formed in the lungs (Fig. S2D). Thus, chronic viral infection reshaped the pulmonary immune environment, generating alternative AM types, simultaneously increasing cellular activation and potential for de novo suppression, and altering DC states with the potential to alter Mtb-specific immune fates.

Lymph node transport and Mtb-specific T cell priming is delayed by chronic viral coinfection.

To determine how the pulmonary immune remodeling affected early Mtb growth, we measured Mtb loads in the lungs and lung-draining medLNs. Unlike the increased Mtb CFU at later stages of coinfection (Fig. 1D), Mtb growth was inhibited in co-infected mice at day 15 after Mtb infection, with Mtb loads being decreased by ~1 log in the lungs and by up to 2-logs in the medLNs, with undetectable Mtb CFU in the medLNs of many co-infected mice at this time (Fig. 3A). To better identify cell types harboring and/or transporting Mtb antigen, we utilized a GFP-expressing Mtb-H37Rv strain (Mtb-GFP). At day 18 after Mtb infection, the number of Mtb-positive AMs and CD11b+ DCs were decreased in the lungs of co-infected mice (Fig. 3B and S3A). However, as a proportion of all pulmonary Mtb-positive cells, Mtb-positive CD11b+ DCs were lower in co-infected mice compared to the Mtb alone group, while the frequency of Mtb-positive AMs was increased (Fig. 3C), indicating decreased early internalization of Mtb specifically by pulmonary CD11b+ DCs. Of note, the decreased Mtb internalization was unlikely due to a defect in phagocytic capacity, since both AMs and DCs from naive and chronic LCMV infected lungs had similar ex vivo phagocytic capacity (Fig. S3B).

Figure 3. Mtb transport and immune infiltration into lung-draining lymph nodes is delayed in coinfected mice.

(A) Mtb CFU in the lungs and medLNs at day 15 after Mtb infection. Dashed line: detection threshold 10 CFU.

For (B-D), data were acquired at day 18 after Mtb-GFP infection.

(B) The proportion (FACS plots) and the number (scatter plots) of Mtb-GFP+ AMs and CD11b+ DCs in the lungs.

(C) The proportions of total Mtb-GFP+ cells that are CD11b+DCs or AMs in the lungs.

(D) The proportion (FACS plots) and the number (scatter plots) of total Mtb-GFP+ cells and of Mtb-GFP+ DCs in the medLNs. Dashed line: the detection threshold 4 events.

(E) CFU kinetics and the numerical kinetics of the indicated cell types in the medLNs of Mtb alone (blue) and coinfection group (red).

(F) AlexaFluor488-OVA positive of CD11b+ DCs in the medLNs and lungs at day 18 after Mtb infection. FACS plots show OVA+ percent of CD11b+DCs in medLNs. OVA intranasal transfer was done 24 hours before the sacrifice.

(G) FACS plots show the Ly5.1+ proportion of medLN DCs. The scatter plots show the number of transferred Ly5.1+ bmDC in the lungs and medLNs 2 days after bmDC transfer. Congenic mismatched Ly5.1+ bmDCs were transferred intranasally to naive or chronic LCMV infected Ly5.2+ mice at day 29 after LCMV Cl13 infection, followed 12 hours later by intranasal infection with BCG (rBCG30, 10⁶ CFU per mouse).

Data represent 2 or more independent experiments with 4-6 mice per group. Gating strategy of indicated subsets in (B-D) is shown in Fig. S3A. *, p<0.05. Student’s t-test. See also Figure S3.

Mtb transport from the lungs to the medLNs by CD11b+ DCs is required to initiate T cell responses (Lai et al., 2018; Wolf et al., 2007). Consistent with the decreased Mtb CFU in co-infected mice, overall there were fewer cells harboring Mtb in the medLNs early after Mtb coinfection, with very few Mtb-harboring DCs (Fig. 3D), indicating a delayed Mtb transport to medLNs in chronic viral coinfection. Unlike the timely immune infiltration to the medLNs following Mtb infection alone, the lymphoid expansion and infiltration of many immune subsets was prevented or delayed in the medLNs of co-infected mice (Fig. 3E). The paralysis of immune infiltration to medLNs was not all due to the initially lower Mtb load, since despite the similar Mtb loads in medLNs by day 30, the number of B cells, DCs and neutrophils remained lower in co-infected mice compared to Mtb alone group (Fig. 3E). To evaluate antigen transport to medLNs at the early stage of coinfection, we intranasally administered AlexaFluor488-labeled Ovalbumin (OVA) at day 18 after Mtb infection. While a similar number of CD11b+ DCs uptook OVA in the lungs of both Mtb-alone and co-infected mice, there were significant fewer OVA-positive CD11b+ DCs in the medLNs of co-infected mice (Fig. 3F), suggesting inhibited antigen transport by pulmonary CD11b+ DCs even with sufficient access to antigens in the lungs of co-infected mice. The same numerical deficiency in OVA transport to the medLNs was observed for pulmonary CD103+DCs and for CCR2+ monocytes at day 18 after Mtb infection (Fig. S3C), as well as CD11b+DCs immediately after Mtb infection (Fig. S3D). Both total CD11b+ DCs and Mtb-containing CD11b+ DCs in the lungs expressed comparable levels of the lymph node homing receptor CCR7 (Fig. S3E, S3F), suggesting that the diminished migratory ability to medLNs was not due to CCR7 downregulation. Further, activated bone-marrow derived DCs (bmDCs) transferred intranasally to chronic LCMV infected mice were impaired in their trafficking to medLNs compared to naive mice, whereas the numbers of bmDCs were analogous in the lungs of both naive and co-infected mice after transfer (Fig. 3G), indicating the inhibition of medLN homing was not due to changes in cell death of activated DCs or bmDC “take”. Thus, viral persistence specifically diminishes the trafficking of CD11b+ DCs and Mtb-transport from the lungs to draining lymph nodes.

We next assessed how the delayed Mtb and DC migration in coinfection affected Mtb-specific CD4 T cell priming. To capture the early events in T cell priming we intravenously transferred Mtb Ag85B_240-254 (P25)-specific or Mtb-ESAT6_4-17 (C7)-specific CD4 T cell receptor (TCR) transgenic T cells prior to Mtb-infection and measured their proliferation in the lung-draining medLNs after Mtb infection [the primary site for T cell priming following aerosol Mtb infection (Reiley et al., 2008)]. The initial “take” of Mtb-specific CD4 T cells was not altered by chronic virus infection (Fig. S4A). Yet, whereas Mtb-specific CD4 T cells proliferated extensively and migrated to the lungs within 18 days after Mtb infection alone, their proliferation and pulmonary migration was almost entirely blunted in the presence of chronic virus coinfection (Fig. 4A). The lack of proliferation led to substantially decreased number of Mtb-specific CD4 T cells (Fig. 4A). The diminished T cell priming was analogously observed in coinfected mice that had not been initially CD4 depleted prior to LCMV infection (Fig. S4B), and previous acute LCMV-Armstrong infection did not reduce T cell proliferation or decrease Mtb transit to the medLNs (Fig. S4C, S4D). Importantly, the frequency and number of endogenous MHC tetramer-positive and IFNγ/TNFα-producing Mtb-specific CD4 and CD8 T cells were also decreased in the lungs early following Mtb coinfection (Fig. 4B, S4E). Thus, chronic virus coinfection broadly impedes Mtb-specific T cell priming.

Figure 4. Chronic virus coinfection delays Mtb-specific T cell priming.

(A) Proliferation and number of ESAT6-specific and Ag85B-specific transgenic CD4 T cells in the lungs and medLNs of Mtb alone (blue) and coinfection (red) group at day 18 after Mtb infection.

(B) IFNγ and TNFα expression following ex vivo peptide re-stimulation of endogenous T cells in the lungs at day 21 after Mtb infection.

Data represent 2 or more independent experiments with 3-4 mice per group. *, p<0.05. Student’s t-test. See also Figure S4.

Mtb-specific CD4 T cells are skewed towards Th17 differentiation by chronic virus coinfection

By day 30 after Mtb coinfection, the Mtb-specific CD4 T cells had proliferated and reached analogous numbers with Mtb infection alone in both the medLNs and lungs (Fig. 5A, S5A), indicating that the priming of Mtb-specific CD4 T cells was not permanently inhibited, but instead was delayed by chronic viral coinfection. To understand how the delayed priming affected T cell differentiation, we analyzed ESAT6-specific CD4 T cells by CyTOF at day 30 after Mtb infection. PhenoGraph analysis identified 9 clusters that largely classified into T-bet+ Foxp3- RORγT- Th1 populations (c1, c2, c5); T-bet^dim Foxp3-RORγT^hi Th17 populations (c6, c7), and a Foxp3+ Treg cell population (c8) (Fig. 5B, 5C). Unlike the dominance of CD4 Th1 cells in Mtb infection alone, the Th17 clusters were instead expanded in coinfected mice (Fig. 5B-D, S5B). An increase in c8 (Treg cells) and c9 were also observed, however the large fold change was largely attributable to their low level in mice infected with Mtb alone as opposed to these becoming dominant clusters in co-infected mice (Fig. S5B). Mtb-specific CD4 T cells also exhibited increased IL-17A production, with the amount of RORγT+ and IL-17A+ T cells remaining elevated as coinfection progressed compared to Mtb infection alone (Fig. 5E, 5F). Further, co-infected mice maintained a higher ratio of Mtb-specific IL-17A+ versus IFNγ+IL-17A- CD4 T cells at day 30 and then throughout Mtb infection (Fig. 5G). The pre-existing chronic infection did not diminish long-term Mtb-specific CD4 or CD8 T cell expression of IFNγ or TNFα compared to Mtb infection alone (Fig. S6), indicating that the T cell exhaustion characteristic of the chronic virus-specific T cells did not spread to affect Mtb-specific T cells and that T cell exhaustion itself did not account for the inability to control Mtb coinfection. In the lungs, IL-17A protein levels were also elevated in co-infected mice, while pulmonary IFNγ expression was not different (Fig. S5D), consistent with the similar levels of IFNγ producing CD4 Th1 cells (Fig. S6). Again, the increase in total IL-17A protein expression led to an inversion in the IL-17A:IFNγ ratio in co-infected mice (Fig. S5D), indicating a general environmental shift toward Th17 responses at the cellular and lung tissue level beginning by day 30 after coinfection. Consistent with the increased IL-17A tissue levels, the skewed T cell differentiation was not limited to transgenic CD4 T cells, as the enhanced Th17 response was also observed by endogenous CD4 T cells during coinfection (Fig. S5E, S5F). The elevated Th17 differentiation was not observed in Mtb co-infected mice that previously had an acute LCMV-Armstrong infection (Fig. S5G, S5H), indicating the ongoing chronic viral infection underlies the skewed Mtb-specific Th17 differentiation as opposed to an inherent remodelling of the lung/medLN by previous infection.

Figure 5. Chronic virus coinfection skews Mtb-specific CD4 T cells towards Th17 differentiation and elevates pulmonary neutrophil recruitment.

(A) Proliferation and the number of ESAT6-specific (C7) CD4 T cells in the lungs at day 30 after Mtb infection.

(B-D) CyTOF analysis of ESAT6-specific (C7) CD4 T cells in the lungs at day 30 after Mtb infection.

(B) Left UMAP plots show the distribution and T-bet and RORγT expression of 9 PhenoGraph clusters. Right UMAP plot shows the clusters with proportions significantly increased in coinfection (red) or Mtb alone (blue). The bar graph indicates the log2 fold change of each cluster.

(C) Heatmap depicts protein expression of indicated clusters. Scale is the z-score of log2 transformed median signal intensity.

(D) The proportions of ESAT6-specific Th1 and Th17 clusters. Shown are the median values (midline), interquartile ranges (boxes) and the highest and lowest values (whiskers).

(E) Flow plots show T-bet and RORγT expression of Ag85B-specific (P25) and ESAT6-specific (C7) CD4 T cells in the lungs at day 30 after Mtb infection. Graphs show the numerical kinetics of RORγT+ P25 and C7 T cells in the lungs.

(F) Flow plots show IL-17A and IFNγ expression of Ag85B-specific (P25) and ESAT6-specific (C7) CD4 T cells in the lungs at day 30 after Mtb infection. Graphs show the numerical kinetics of IL-17A producing P25 and C7 T cells in the lungs.

(G) Numerical ratio of IL-17A+ versus IFNγ+IL-17A- CD4 T cells specific to Ag85B and ESAT6 in the lungs.

(H) Colors represent the arcsinh ratio change in medians spectral indices for the indicated protein of each PhenoGraph cluster between Mtb alone and coinfection groups. p-values are calculated by the diffcyt R package. ***, p<0.001. **, p<0.01. *, p<0.05.

(I) Neutrophil (Ly6G+ CD11b+) proportion (FACS plots) and number (scatter plots) in the lungs at day 30 after Mtb infection. Isotype or anti-IL17A antibody treatment: once every other day from day 22 to 30 after Mtb infection.

(J) Survival of mice infected with Mtb alone or co-infected mice treated with isotype antibody or anti-Ly6G antibody (clone 1A8). Yellow zone shows the treatment duration. Log rank (Mantel-Cox test), *, p<0.05. ns, not significant (p = 0.11).

Data represent 2 or more independent experiments with 3-10 mice per group, except panel (I) which is single experiment with 8 mice per coinfection group and 4 mice for the Mtb alone group. Error bars depict SEM. *, p<0.05; ns, not significant. Student’s t-test for (D-G). One-way ANOVA for (I). See also Figure S5 and S6.

In addition to changes in the overall abundance of Th1 and Th17 cells, the Mtb-specific Th17 clusters existed in numerous differential states between Mtb infection alone and coinfection, wherein coinfected mice exhibited increased activation profiles, including elevated expression of PD1, SLAM and CD73 compared to Mtb infection alone (Fig. 5H, S5I), indicating an activated Th17 cell differentiation state in the inflammatory coinfection setting that was normally minimized during Mtb infection alone. Thus, the abundance, cellular activation state and IL-17A production of Mtb-specific Th17 cells are all amplified by chronic viral coinfection.

Early depletion of neutrophils increases long-term survival following Mtb coinfection

The Th17 response is correlated to pulmonary recruitment of neutrophils in infection (Lombard et al., 2016). By day 30 after Mtb infection, neutrophil numbers and the expression of chemokines (G-CSF, CXCL1 and CXCL5) involved in neutrophil recruitment were increased in the lungs of co-infected mice compared to the Mtb alone group (Fig. S5J, S5K). This was particularly evident in the elevated lung neutrophil to lymphocyte ratio (NLR) (Fig. S5J), which is associated with detrimental pulmonary inflammation of tuberculosis in clinical settings (Han et al., 2018; Miyahara et al., 2019; Panteleev et al., 2017). To determine whether the elevated IL-17 response in co-infected mice accounted for the enhanced neutrophil recruitment in the lungs, an anti-IL-17A neutralizing antibody treatment was administered day 22-30 after Mtb infection. Whereas blocking IL-17A during Mtb infection alone did not reduce lung neutrophils, co-infected mice receiving IL-17A blockade exhibited decreased lung neutrophilia (Fig. 5I), indicating that the elevated IL-17A at least in part promoted the excessive neutrophil recruitment during coinfection.

We next determined whether elevated neutrophil recruitment contributed to heightened lethality in co-infected mice. Interestingly, co-infected mice receiving continuous neutrophil depletion from day 27 to day 45 post Mtb infection exhibited improved survival compared with the isotype control group (Fig. 5J, S5L). Further, neutrophil depletion in coinfection extended survival times that were now not significantly different from Mtb infection alone (p = 0.11). The median survival time for isotype treated co-infected mice was 174 days, for Mtb infection alone was 260 days, and for neutrophil depleted mice was 215 days, indicating that lethality of Mtb co-infected mice could be overcome by early neutrophil depletion. Thus, in the presence of chronic viral coinfection, elevated IL-17A induced pulmonary neutrophilia in part mediates the accelerated mortality.

Pre-existing TNFα delays Mtb-specific T cell priming during chronic virus coinfection.

We next sought to understand the mechanisms initially suppressing Mtb growth and inhibiting T cell priming at early phases of Mtb coinfection. Chronic virus infection induces sustained inflammation and increases the expression of multiple inflammatory cytokines, many of which are critical to inhibit Mtb growth such as IFNγ, TNFα and IL-1 (Berry et al., 2010; Dorhoi and Kaufmann, 2014; Manca et al., 2005; Mayer-Barber and Barber, 2015). Consistent with the increased inflammatory state, IFNγ, TNFα and IL-1 protein expression were increased in the lung homogenate of the mice with chronic LCMV infection (Fig. 6A, S7A-C). Type I interferon (IFN-I) signaling is also enhanced by chronic LCMV infection (Wilson et al., 2013) and although generally considered beneficial to Mtb growth in vivo (Dorhoi et al., 2014), its pre-existing heightened expression may affect CD4 T cell priming (Osokine et al., 2014; Snell et al., 2016). Thus, we hypothesized that counter to their long-term beneficial role, the pre-existing levels of these factors could initially restrict (but not eliminate) Mtb growth, leading to decreased Mtb transport to the medLNs, delayed T cell priming and ultimately resulting in the increased Mtb burden. To test this, separately blocked TNFα, IFNγ, IL-1 receptor (IL-1R) and the IFN-I receptor (IFNαR) beginning 1 day before to 13 days after Mtb infection. However, only TNFα blockade restored lung Mtb loads in coinfected mice to the level observed in Mtb infection alone, whereas inhibiting IFNγ, IL-1R or IFNαR had no significant effect (Fig. 6B). TNFα blockade also increased Mtb CFU in the medLNs (Fig. 6B), indicating that the chronic virus-induced TNFα was inhibiting Mtb transport to medLNs. The observed effects of TNFα, IFNγ or IL1 blockade were not due to the loss of their expression, as the amounts of each of these were either increased or the same in the first two weeks of coinfection compared to Mtb infection alone (Fig. S7D). Neither TNFα nor IL-1R blockade increased chronic LCMV titers over this time frame, whereas both IFNγ and IFNαR blockade did increase LCMV titers (Fig. S7E, S7F), indicating the enhanced Mtb transport resulting from TNFα blockade was not due to changes in virus titers.

Figure 6. TNFα blockade restores timely Mtb transport to the medLNs and Mtb-specific T cell priming.

(A) Cytokines in lung homogenate of naïve and chronic LCMV groups (day 30 after LMCV-Cl13 infection). Each column represents one individual mouse.

(B-E) Antibody treatment: once every other day, from day −1 to day 13 (B) or day 17 (C-E) after Mtb infection. Data were acquired at day 14 (B) or day 18 (C-E) after Mtb infection. Iso: isotype antibody; ‘αTNF: anti-TNFα antibody.

(B) Mtb CFU in the lungs and medLNs.

(C) Number of total cells and DCs in medLNs.

(D) Lung Mtb CFU, the number of Mtb-GFP positive AMs and CD11b+ DCs in the lungs, and Mtb-GFP positive DCs in medLNs.

(E) Ag85B-specific CD4 T cell proliferation (histograms) and activation (FACS plots) in the medLNs. The graph indicates the number of activated P25 CD4 T cells (PD-1+ CD44+) in the medLNs.

Data represent 2 or more independent experiments with 4-6 mice per group. Gating strategy in (D) is shown in Fig. S3A. *: p<0.05; ns, not significant. Student’s t-test for (A), One-way ANOVA for (B-E). See also Figure S7.

In conjunction with increasing Mtb transport to the medLNs, TNFα blockade elevated overall cellularity in the medLNs, and specifically the number of DCs (Fig. 6C). While TNFα blockade did not affect the number of Mtb-positive AMs in the lungs, the number of Mtb-positive CD11b+ DCs were increased in both the lungs and medLNs (Fig. 6D), indicating that TNFα decreased Mtb uptake and medLN transport by DCs at early stages of coinfection. Coincident with the increased antigen transport to the medLNs following TNFα blockade, Mtb-specific CD4 T cells were activated and their proliferation enhanced at day 18 after Mtb infection, leading to a greater than 10-fold increase in the number of Mtb-specific CD4 T cells in co-infected mice compared to isotype treated co-infected mice (Fig. 6E). Thus, by initially limiting Mtb growth, chronic virus induced TNFα impedes Mtb transport to the medLNs and consequently, prevents the induction of Mtb-specific T cell immunity.

Temporally correcting Mtb-specific CD4 T cell priming inhibits the elevated Th17 response and enhances Mtb control in co-infected mice.

The association between delayed Mtb-specific T cell priming and increased long-term Mtb burden in co-infected mice suggested a causal relationship. To determine whether the diminished T cell priming could be overcome to the immune aberrations and control Mtb growth, activated bmDCs pulsed either without peptide or with Mtb-Ag85B_240-260 peptide were transferred into mice at day 6 after Mtb infection. While transfer of unpulsed bmDCs did not accelerate T cell proliferation in co-infected mice and transferring Mtb-peptide pulsed bmDCs to mice infected with Mtb alone did not increase CD4 T cell proliferation, Mtb-specific CD4 T cells in co-infected mice receiving Mtb-peptide pulsed bmDCs now proliferated and numerically expanded to the same amount as mice infected with Mtb alone (Fig. 7A). The increased proliferation and numerical expansion occurred in both the medLNs and the lungs, further indicating that the suppression of Mtb-specific T cell priming was not T cell intrinsic and that temporally correcting the induction of Mtb-specific T cells through provision of antigen at the appropriate time enabled their pulmonary trafficking.

Figure 7. Overcoming the CD4 T cell priming delay prevents aberrant Th17 differentiation, rebalances neutrophil recruitment, and enhances Mtb control.

Mice were infected with Mtb alone or were co-infected and then received no bmDC (“no DC”), peptide-unlabeled bmDC (“unlabel DC”), or Mtb Ag85B_240-260 peptide-labeled bmDC + Mtb ESAT6_4-17 peptide-labeled bmDC (“pep-DC)” at day 6 after Mtb infection.

(A) The proliferation (histograms; blue: Mtb alone; red: coinfection group) and the number of Ag85B-specific CD4 T cells in the medLNs and lungs at day 18 after Mtb infection (12 days after bmDC transfer). bmDCs for “pep-DC” group in (A) were pulsed with only Ag85B peptide.

For (B-E), data were acquired at day 35 after Mtb infection (29 days after bmDC transfer).

(B) FACS plots show T-bet and RORγT expression by Ag85B-specific CD4 T cells in the lung. Scatter plots show the number of RORγT+ and IL-17A producing Ag85B-specific CD4 T cells in the lungs.

(C) Numerical ratio of IL-17A+ versus IFNγ+IL-17A- CD4 T cells specific to Ag85B in the lungs.

(D) The neutrophil : lymphocyte ratio in the lungs.

(E) Lung Mtb CFU.

(F) Spleen Mtb CFU of co-infected mice at day 42 after Mtb infection.

Data represent 2 or more independent experiments with 4-6 mice per group. *: p<0.05. ns: not significant. Student’s t-test for (F), One-way ANOVA for (A-E).

Interestingly, the elevated Mtb-specific CD4 Th17 differentiation, the elevated IL-17A to IFNγ production ratio and the increased neutrophil to lymphocyte ratio that normally occurred in co-infected mice, was abrogated when Mtb-specific CD4 T cells were primed at the appropriate time (Fig. 7B-D). This effect toward driving these immune aberrations was specifically due to correcting priming time of Mtb-specific CD4 T cells, since neither reduced Th17 nor NLR were observed when peptide pulsed bmDCs were given to mice infected with Mtb alone (Fig. 7B-D). Further, co-infected mice receiving Ag85B_240-260 and ESAT6_4-17 peptide-pulsed bmDCs exhibited markedly reduced pulmonary Mtb burden and diminished Mtb dissemination to the spleen to the same levels observed following Mtb infection alone (Fig. 7E, 7F). Importantly, Ag85B_240-260 and ESAT6_4-17 are both MHC-II restricted peptides, demonstrating that the delay in CD4 T cell priming underlies many of the Mtb disease manifestations induced by chronic viral coinfection and that despite ongoing viral persistence, temporally correcting CD4 T cell induction restores Mtb control.

Discussion

We developed a coinfection model combining chronic LCMV and Mtb that recapitulated many of the inflammatory and immune dysfunctions observed in chronic viral infections, including increased expression of pro-inflammatory factors such as interferons and TNFα (Costiniuk and Jenabian, 2014; Israel-Biet et al., 1991; Lukhele et al., 2019), the accelerated disease and increased mortality associated with Mtb coinfection during multiple chronic viruses, including HIV, HBV, HCV and CMV (Chen et al., 2018; Muller et al., 2019; WHO, 2020; Wu et al., 2015), as well as in some inflammatory metabolic diseases (Baker et al., 2011). Using this model, we demonstrate the multiplicity of sequentially evolving mechanisms underlying how Mtb co-opted otherwise beneficial immune factors and turned them to potentiate Mtb disease during chronic viral coinfection. Similar to pulmonary HIV infection, chronic LCMV infection drove continual inflammation in the lungs, remodeling the immune landscape and sustaining inflammatory cytokine production, such as TNFα (Costiniuk and Jenabian, 2014; Israel-Biet et al., 1991; Millar et al., 1991; Steffen et al., 1993), IFNs (Lukhele et al., 2019) and IL-1 (Steffen et al., 1993). TNFα is essential for optimal control of Mtb growth in vivo (Mayer-Barber and Barber, 2015). However, we now identified the counterintuitive mechanism whereby initial chronic virus driven TNFα production at the onset of Mtb coinfection led to worsened Mtb disease, diminished survival and many of the sequelae associated with chronic virus and Mtb coinfections. Mechanistically, by initially suppressing, but not eliminating Mtb infection, elevated TNFα production limited Mtb antigen availability for DC uptake, prevented early Mtb transport to the lung-draining lymph node, thereby effectively inhibiting immune infiltration, delaying T cell activation and providing a fertile environment for Mtb multiplication in the absence of immune control. Ultimately, the long-term outcome of these early TNFα-induced events was higher Mtb burden and accelerated death. Interestingly, the same Mtb-suppressive impact was not true for other inflammatory factors IL-1, IFNγ or IFN-I that regulate Mtb infection and immune function. This is particularly surprising for IFN-I, since it drives many of the pathogenic immune dysfunctions in chronic viral infections and has been implicated in the control of secondary infections following an acute influenza infection (Redford et al., 2014). This difference may have to do with the very high levels of IFN-I present in the initial stages of an acute infection (i.e., the time when the influenza coinfection studies were performed) compared to the lower smoldering levels of IFN-I pervasive in chronic viral infections (Lukhele et al., 2019). Regardless, our data demonstrate that the mechanisms increasing bacterial infection during an acute infection are different from those that mediate increased Mtb coinfection during chronic viral infection. This also indicates that not all proinflammatory factors with anti-Mtb activity suppressed early Mtb coinfection, but instead that it was a specific attribute of TNFα in the chronic virus coinfection environment that Mtb used to prevent antigen transport to the lymph nodes and to arrest Mtb-specific immunity.

The innate immune response exhibits limited capacity on its own to control Mtb growth in the initial stages of infection and ultimately, Mtb growth is controlled by the induction of T cells, in particular CD4 Th1 cells (Gallegos et al., 2008). In fact, timely induction of CD4 T cell responses may be of the utmost importance as it has been reported that at later stages of Mtb infection, CD4 T cells primarily prevent further increases in Mtb burden but not actually decrease it (Urdahl et al., 2011). Thus, by delaying T cell priming, Mtb subverted immune control to enable sanctuary-like growth in co-infected mice. When Mtb-specific CD4 T cells were ultimately primed, they were differentially polarized away from the Th1 response critical for Mtb control, toward highly activated Th17 cells. The differential polarization of CD4 Th17 cells due to the priming delay during chronic viral coinfection may be indicative of subsequent changes in antigen presenting cells (Snell et al., 2018; Snell et al., 2016) and/or alterations in the cytokine milieu induced by chronic virus infection, but is of particular interest since Th17 cells can amplify inflammation during Mtb infection, promote pulmonary neutrophilia and are enigmatically associated with both control and with worsened Mtb disease (Khader et al., 2007; Kumar et al., 2013b; Mourik et al., 2017). This same Th17 paradox is consistent with clinical observations associating IL-17 levels to Mtb burden in pulmonary Mtb patients (Kumar et al., 2019), and with patients developing immune reconstitution inflammatory syndrome (Grant et al., 2012). Our data suggest that the ultimate correlation of Th17 cells with Mtb disease control or progression could really be the result of other actual pathogenic disease altering mechanisms such as delayed T cell priming, changes in cytokine balance, and/or neutrophil amplification that themselves promote these distinct disease outcomes.

Interestingly, virus-specific CD4 Th1 responses also progressively diminish with viral persistence (Fahey et al., 2011; Feng et al., 2012; Lindqvist et al., 2012; Petrovas et al., 2012) and the de novo priming of new virus-specific CD4 Th1 cells against the chronic virus is simultaneously abrogated (Snell et al., 2016). Thus, chronic viruses have developed strict mechanisms to impede CD4 Th1 responses that we now show can spread to the induction of coinfection specific CD4 T cell responses. In chronic viral infection, the virus-specific CD4 T cells are heavily skewed toward Tfh responses (Fahey et al., 2011; Osokine et al., 2014; Snell et al., 2016), while priming of Mtb-specific CD4 T cells was instead diverted toward Th17 cells. Yet, whereas IFN-I suppresses chronic virus-specific CD4 Th1 differentiation, inhibiting IFN-I in Mtb coinfection did not promote CD4 T cell priming, indicating distinct mechanisms guiding the differentiation of de novo chronic virus-specific and coinfection specific immune responses. Further, the onset of T cell exhaustion that normally occurs rapidly following chronic viral infection did not spread to the Mtb-specific T cells during coinfection. Despite the pre-existing expression of PD-L1, IL-10 and other suppressive factors induced by chronic virus infection, combined with increased expression of PD1 and CTLA-4 on Mtb-specific CD4 T cells, enhanced exhaustion of the T cell response itself cannot explain the failure to control Mtb growth.

It is not precisely clear how CD4 T cells function to control Mtb growth, particularly in the context of a co-infecting virus. Re-establishing the correct CD4 T cell priming time overcame many of the immune dysfunctions observed following Mtb coinfection. On the other hand, transferring the bmDCs to mice recently infected with Mtb alone had no effect on the T cell priming, differentiation or Mtb control, indicating that the delayed priming in co-infected mice and the lack of timely control of Mtb growth afforded by appropriate induction of CD4 T cell responses was a key driver for the worsened Mtb infection. When the CD4 T cells were ultimately induced, the logarithmic Mtb growth was curtailed, albeit at a much higher set point. The arrested Mtb-specific T cell priming was not due to T cell intrinsic changes, but rather the lack of timely antigen transport to the lymph node resulting from the heightened TNFα expression. A similar scenario is suggested in HIV infection, wherein patients with cART suppressed viral loads and a corresponding decrease in chronic inflammation are better able to handle Mtb coinfection, and the accelerated disease is not observed (Abay et al., 2015; Suthar et al., 2012). Thus, proper and timely induction of CD4 T cell responses is fundamental for subsequent Mtb control and the subversion of TNFα to delay T cell priming drives increased Mtb burden and multiple immune dysfunctions in the face of chronic virus coinfection.

Temporally occurring with the induction of Th17 cells was an increase in pulmonary neutrophilia. Neutrophil accumulation is protective in Mtb in some cases (Martineau et al., 2007), but excessive pulmonary neutrophilia also leads to elevated inflammation and increased mortality (Kimmey et al., 2015; Nandi and Behar, 2011), suggesting a threshold wherein neutrophils are important for protection, but above which leads to excessive disease. Neutrophils are reservoirs for Mtb (Eum et al., 2010) and exhibit relatively lower capacity to control Mtb intracellular growth compared to other cell types such as interstitial macrophages (Huang et al., 2018), suggesting the enhanced infiltration of neutrophils may provide enhanced Mtb growth in more tolerant host cells. When the timing of Mtb-specific CD4 T cells priming was corrected, the hyper Th17 response was reduced, pulmonary neutrophilia was decreased and Mtb burden was controlled. Strikingly, depleting neutrophils in Mtb coinfection enabled a long-term survival advantage that largely overcame the accelerated mortality associated with coinfection, indicating the critical role of early neutrophil amplification in potentiating the severity of Mtb disease. Thus, we now demonstrate the sequential mechanisms and series of early events following Mtb coinfection that then directly dictate the long-term disease outcomes.

Limitation of study

Our study identifies biologic mechanisms that (dys)regulate Mtb-specific immunity in the presence of a chronic viral co-infection to enhance disease severity. Identification of the earliest Mtb-specific infection events in in humans and primate models are precluded by the inability to time infection and to identify Mtb-specific T cells prior to robust expansion. Thus, human translation will require the development of new techniques to understand these earliest infection events that limit Mtb coinfection in the presence of chronic viral infections. In this vane, the C57BL/6 mouse strain used in these studies does not generate a latent infection nor granulomas characteristic of human Mtb infection. Thus, use of additional mouse models (Pedroza-Roldan and Flores-Valdez, 2017; Plumlee et al., 2020) reported to generate these responses will have to be evaluated to understand the role of chronic viral co-infection in (dys)regulating these mechanisms of Mtb disease.

STAR METHODS

RESOURCE AVAILABILITY

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, David Brooks (dbrooks@uhnresearch.ca).

Materials Availability

Further information and material requests should be addressed to David Brooks (dbrooks@uhnresearch.ca).

Data and Code Availability

This study did not generate/analyze [datasets/code]. CyTOF data analysis were done with standard R packages (Cytofkit, UMAP and diffcyt).

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice

C57BL/6J (stock#000664) and B6.129S2-Cd4^tm1Mak/J (CD4−/−) mice (stock#002663) were purchased from Jackson Laboratories or the rodent breeding colony at the Princess Margaret Cancer Center. Ag85B_240-254 specific (P25) transgenic mice were purchased from Jackson Laboratories and bred onto Ly5.1 background. ESAT6_4-17 specific (C7) transgenic mice (Moguche et al., 2017) were generously provided by Dr. Kevin B. Urdahl (Center for Infectious Disease Research, Seattle, WA) and bred onto a Thy1.1 background. All mice were housed under specific pathogen-free conditions. Experiments involving Mycobacterium tuberculosis (Mtb) were conducted in biosafety level III laboratory at the University of Toronto. Animal studies and procedures were approved by the University of Toronto Animal Care Committee or the University Health Network Animal Care Committee at the Princess Margaret Cancer Center/University Health Network. Mice between 8-12 weeks of age were used. Mouse survival endpoints were determined by a score including a 20% decrease in body weight, in combination with diminished activity/lethargy and hunched-back.

Microbe strains

Mtb-H37Rv, Mtb-GFP and Mycobacterium bovis BCG strain rBCG30 (rBCG30-ARMF-II Tice) (Gillis et al., 2014; Horwitz et al., 2000) were prepared and titrated as described previously (Ahn et al., 2018). Mtb strains and rBCG30 strain were grown at 37°C in Middlebrook 7H9 broth (BD) supplemented with 0.2% glycerol, 10% ADC (BD), and 0.05% Tween 80 or on 7H11 agar supplemented with 0.5% glycerol and 10% OADC (BD). The Mtb-GFP strain was constructed by electroporation of Mtb-H37Rv with plasmid pUV15 isolated from strain rBCG30-GFP [unpublished strain provided by Michael Tullius and Marcus Horwitz; this strain was constructed using pUV15 (Ehrt et al., 2005) obtained from Sabine Ehrt].

For bacterial stock preparation, rBCG30 or Mtb culture were harvested when an OD₆₀₀ value of the culture medium reached 0.6~0.8. Bacteria were washed, resuspended in 7H9 broth with 10% glycerol and stored at −80°C. For aerosol infection, 100 colony-forming units (CFU) of Mtb per mouse were delivered using Glas-Col nebulizer Inhalation Exposure System (Ahn et al., 2018). Correct dosing was confirmed by quantifying Mtb CFU in the lung 1 day after Mtb infection. For intranasal infection with rBCG30, 10⁶ CFU rBCG30 were delivered intranasally in 20 μl PBS. For Mtb quantification, tissues or single cell suspensions were homogenized in PBS. Tissue homogenate was serial diluted and coated on 7H11 agar plates. Mtb colonies were counted on the plates after 3-week incubation at 37°C.

LCMV-Clone 13 and LCMV-Armstrong stocks were prepared by a single passage on BHK-21 cells, and viral titers were determined by plaque formation on Vero cells as described previously (Brooks et al., 2005). Stocks. Mice were infected intravenously with 2x10⁶ plaque-forming units (PFU) of LCMV-Clone 13 to generate chronic viral infection or intraperitoneally with 2x10⁴ PFU of LCMV-Armstrong to generate an acute infection. Mtb infection was performed 30 days after the LCMV-Arm infection, correlating to ~18 days after resolution of the acute LCMV-Arm infection.

METHOD DETAILS

In Vivo antibody treatments

For transient depletion of CD4 T cells, mice were intraperitoneally administered 125 μg of anti-CD4 (clone GK1.5) antibody at 5 days before and at 2 days before LCMV infection. CD4 depletion and reconstitution was confirmed in multiple tissues by flow cytometry using an anti-CD4 antibody (clone RM4.4) that is not blocked by the CD4 depleting antibody (Yamada et al., 2015). For neutrophil depletion, 250μg per mouse of anti-Ly6G (clone 1A8) or isotype control (clone 2A3) antibody was intravenously administered every 2 days between day 27 and 45 after Mtb infection. For in vivo cytokine blocking experiments, 500μg per mouse of anti-TNFα (clone TN3-19.12), anti-IFNαR (clone MAR1-5A3), anti-IFNγ (clone H22), anti-IL-1R (clone JAMA-147), or appropriate isotype controls (Armenian hamster IgG) were intravenously administered 1 day prior to Mtb infection and then once every 2 days for 8 treatments. For IL-17A blockade, 500 μg per mouse of anti-IL-17A (clone 17F3) or isotype control (clone MOPC-21) were intraperitoneally administered every 2 days from day 22 to day 30 after Mtb infection. In vivo blocking and depletion antibodies were obtained from BioXcell.

Time-of-Flight mass cytometry (CyTOF)

Details for each antibody is listed in the Key Resource Table. Purified unconjugated antibodies were labeled with metal-tags at the SickKids-UHN Flow and Mass Cytometry Facility using the MaxPar Antibody Labeling kit from Fluidigm. Directly conjugated antibodies were purchased from Fluidigm. All working antibody concentrations were determined by titration.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
89Y-anti-NOS2 (Clone CXNFT)	Purified antibody from ebioscience	Cat#14-5920-82; RRID: AB_2865171
89Y-anti-CD45 (Clone 30-F11)	Fluidigm	Cat#3089005B; RRID: AB_2651152
115In-anti-Ly6C (Clone HK14)	Purified antibody from BioLegend	Cat#128001; RRID: AB_1134213
141Pr-anti-CD44 (Clone IM7)	Purified antibody from BioLegend	Cat#103001; RRID: AB_312952
142Nd-anti-CXCR5 (Clone L138D7)	Purified antibody from BioLegend	Cat#145502; RRID: AB_2561955
143Nd-anti-Siglec-F (Clone E50-2440)	Purified antibody from BD	Cat#552125; RRID: AB_394340
143Nd-anti-CD45.2 (Clone 104)	Purified antibody from BioLegend	Cat#109802; RRID: AB_313439
144Nd-anti-Granzyme B (Clone GB11)	Purified antibody from Invitrogen	Cat#MA1-80734; RRID: AB_931084
145Nd-anti-CTLA-4 (Clone UC10-4B9)	Purified antibody from BioLegend	Cat#106302; RRID: AB_313251
145Nd-anti-MHC-I (Clone AF6-88.5)	Purified antibody from BioLegend	Cat#116501; RRID: AB_313728
146Nd-anti-F4/80 (Clone BM8)	Fluidigm	Cat#3146008B; RRID: AB_2811238
146Nd-anti-EOMES (Clone Dan11mag)	Purified antibody from eBioscience	Cat#14-4875-82; RRID: AB_11042577
147Sm-anti-Thy1.1 (Clone HIS51)	Purified antibody from eBioscience	Cat#14-0900-81; RRID: AB_467373
147Sm-anti-CD45.2 (Clone 104)	Purified antibody from BioLegend	Cat#109802; RRID: AB_313439
148Nd-anti-CD11b (Clone M1/70)	Fluidigm	Cat#3148003B; RRID: AB_2814738
149Sm-anti-CD69 (Clone H1.273)	Purified antibody from BioLegend	Cat#104502; RRID: AB_313105
150Nd-anti-Ly6G (Clone 1A8)	Purified antibody from BioLegend	Cat#127601; RRID: AB_1089179
151Eu-anti-CD25 (Clone 3C7)	Fluidigm	Cat#3151007B; RRID: AB_2827880
152Sm-anti-CD86 (Clone GL-1)	Purified antibody from BioLegend	Cat#105007; RRID: AB_313150
153Eu-anti-CD8α (Clone 53-6.7)	Fluidigm	Cat#3153012B; RRID: AB_2885019
154Sm-anti-CD45.1 (Clone A20)	Purified antibody from BioLegend	Cat#110702; RRID: AB_313491
154Sm-anti-CD73 (Clone TY/11.8)	Fluidigm	Cat#3154019B; RRID: AB_2813854
154Sm-anti-BATF (Clone D7C5)	Fluidigm	Cat#3154012A; RRID: AB_2687838
155Gd-anti-CD103 (Clone M290)	Purified antibody from BD	Cat#553699; RRID: AB_394995
155Gd-anti-CD103 (Clone 2E7)	Purified antibody from BioLegend	Cat#121401; RRID: AB_535944
156Gd-anti-PD-L1 (Clone M1H5)	Purified antibody from eBioscience	Cat#14-5982-82; RRID: AB_467781
158Gd-anti-CD209α (Clone MMD3)	Purified antibody from BioLegend	Cat#833001; RRID: AB_2564962
158Gd-anti-RORɣt (Clone Q31-378)	Purified antibody from BD	Cat#562197; RRID: AB_10894594
159Tb-anti-CD39 (Clone 24DMS1)	Purified antibody from eBioscience	Cat#14-0391-82; RRID: AB_1210501
160Gd-anti-CD4 (Clone rm4-5)	Purified antibody from BioLegend	Cat#100505; RRID: AB_312708
161Dy-anti-T-bet (Clone 4B10)	Fluidigm	Cat#3161014B; RRID: AB_2858233
161Dy-anti-CD40 (Clone HM40-3)	Fluidigm	Cat#3161020B; RRID: AB_2885020
162Dy-anti-Foxp3 (Clone MF-14)	Purified antibody from BioLegend	Cat#126401; RRID: AB_1089120
163Dy-anti-Tim-3 (Clone B8.2C12)	Purified antibody from BioLegend	Cat#134002; RRID: AB_1626128
164Dy-anti-CX3CR1 (Clone SA011F11)	Fluidigm	Cat#3164023B; RRID: AB_2832247
164Dy-anti-CD62L (Clone MEL-14)	Fluidigm	Cat#3164003B; RRID: AB_2885021
164Dy-anti-LCMV Neucleoprotein (Clone VL4)	Purified antibody from Bioxcell	Cat#BE0106; RRID: AB_10949017
165Ho-anti-TCRβ (Clone H57-597)	Purified antibody from BioLegend	Cat#109201; RRID: AB_313424
166Er-anti-B220 (Clone RA3-6B2)	Purified antibody from Invitrogen	Cat#14-0452-82; RRID: AB_467254
167Er-anti-TCF-1 (Clone S33-966)	Purified antibody from BD	Cat#624084; RRID: NA
168Er-anti-Blimp1 (Clone ROS195G)	Purified antibody from Biolegend	Cat#648202; RRID: AB_2300132
169Tm-anti-CD206 (Clone CD68C2)	Fluidigm	Cat#3169021B; RRID: AB_2832249
169Tm-anti-SLAM (Clone TC15-12F12.2)	Purified antibody from BioLegend	Cat#115901; RRID: AB_313680
170Er-anti-CD169 (Clone 3D6.112)	Fluidigm	Cat#3170018B; RRID: AB_2885022
170Er-anti-NK1.1 (Clone PK136)	Fluidigm	Cat#3170002B; RRID: AB_2885023
171Yb-anti-CD80 (Clone 16-10A1)	Fluidigm	Cat#3171008B; RRID: AB_2885024
171Yb-anti-RORɣt (Clone Q31-378)	Purified antibody from BD	Cat#562197; RRID: AB_10894594
172Yb-anti-CD95 (Clone 15A7)	Purified antibody from eBioscience	Cat#14-0951-85; RRID: AB_467393
172Yb-anti-Ki67 (Clone B56)	Purified antibody from BD	Cat#550609; RRID: AB_393778
173Yb-anti-Helios (Clone 22F6)	Purified antibody from BioLegend	Cat#137202; RRID: AB_10900638
174Yb-anti-CD11c (Clone N418)	Purified antibody from BioLegend	Cat#117301; RRID: AB_313770
174Yb-anti-CD127 (Clone A7R34)	Purified antibody from BioLegend	Cat#135002; RRID: AB_1937287
175Lu-anti-PD1 (Clone RMP-30)	Purified antibody from BioLegend	Cat#109101; RRID: AB_313418
176Yb-anti-Thy1.2 (Clone 53-21)	Purified antibody from Thermo Fischer	Cat#14-0902-82; RRID: AB_467379
209Bi-anti-MHC II (Clone M5/114.15.2)	Fluidigm	Cat#3209006B; RRID: AB_2885025
PE/Cy7 anti-mouse CD197 (clone 4B12)	Biolegend	Cat#120123; RRID: AB_2616687
APC anti-mouse/human CD11b (clone M1/70)	Biolegend	Cat#101211; RRID: AB_312794
PerCP/Cyanine5.5 anti-mouse/human CD11b (clone M1/70)	Biolegend	Cat#101227; RRID: AB_893233
APC/Fire™ 750 anti-mouse CD11c (clone N418)	Biolegend	Cat#117351; RRID: AB_2572123
PE/Cy7 anti-mouse CD11c (clone N418)	Biolegend	Cat#117317; RRID: AB_493569
Brilliant Violet 421™ anti-mouse CD11c (clone N418)	Biolegend	Cat#117343; RRID: AB_2563099
APC/Fire™ 750 anti-mouse CD4 (clone GK1.5)	Biolegend	Cat#100459; RRID: AB_2572110
PE/Cy7 anti-mouse CD45.1 (clone A20)	Biolegend	Cat#110729; RRID: AB_1134170
PerCP/Cyanine5.5 anti-mouse CD45.1 (clone A20)	eBioscience	Cat#45-0453-82; RRID: AB_1107003
APC anti-mouse CD45.2 (clone 104)	Biolegend	Cat#109813; RRID: AB_389210
APC/Cyanine7 anti-mouse CD45.2 (clone 104)	Biolegend	Cat#109823; RRID: AB_830788
PE/Cy7 anti-mouse CD62L (clone MEL-14)	Biolegend	Cat#104417; RRID: AB_313102
PerCP/Cyanine5.5 anti-mouse CD8a (clone 53-6.7)	Biolegend	Cat#100733; RRID: AB_2075239
Alexa Fluor® 647 anti-mouse FOXP3 (clone MF-14)	Biolegend	Cat#126407; RRID: AB_1089116
PerCP/Cyanine5.5 anti-mouse Ly-6G/Ly-6C (Gr-1) (clone RB6-8C5)	Biolegend	Cat#108427; RRID: AB_893561
APC anti-mouse IFN-ɣ (clone XMG1.2)	Biolegend	Cat#505809; RRID: AB_315403
PE anti-mouse IL-10 (clone JES5-16E3)	Biolegend	Cat#505007; RRID: AB_315361
FITC anti-mouse IL-17A (clone TC11-18H10.1)	Biolegend	Cat#506907; RRID: AB_536009
PE/Cy7 anti-mouse IL-2 (clone JES6-5H4)	Biolegend	Cat#503831; RRID: AB_2561749
Brilliant Violet 421™ anti-mouse Ly-6G (clone 1A8)	Biolegend	Cat#127627; RRID: AB_10897944
PE anti-mouse I-A/I-E (clone M5/114.15.2)	Biolegend	Cat#107607; RRID: AB_313322
PerCP/Cyanine5.5 anti-mouse I-A/I-E (clone M5/114.15.2)	Biolegend	Cat#107625; RRID: AB_2191072
FITC anti-mouse CD279 (PD-1) (clone J43)	eBioscience	Cat#11-9985-82; RRID: AB_465472
PE/Cy7 anti-mouse CD279 (PD-1) (clone 29f.1.A12)	Biolegend	Cat#135215; RRID: AB_10696422
PE anti-mouse RORγt (clone Q31-378)	BD Biosciences	Cat#562607; RRID: AB_11153137
PE anti-mouse Siglec-F (clone E50-2440)	BD Biosciences	Cat#552126; RRID: AB_394341
PE anti-mouse CD150 (SLAM) (clone TC15-12F12.2)	Biolegend	Cat#115903; RRID: AB_313682
PerCP/Cyanine5.5 anti-mouse T-bet (clone 4B10)	Biolegend	Cat#644805; RRID: AB_1595593
APC anti-mouse CD90.1 (Thy-1.1) (clone H1S51)	eBioscience	Cat#17-0900-82; RRID: AB_469420
Brilliant Violet 421™ anti-mouse TNFα (clone MP6-XT22)	Biolegend	Cat#506327; RRID: AB_10900823
APC anti-mouse/human CD44 (clone IM7)	biolegend	Cat#103011; RRID: AB_312962
PE anti-mouse/human CD45R/B220 (clone RA3-6B2)	biolegend	Cat#103207; RRID: AB_312992
BV786 anti-Mouse CD103 (clone M290)	BD Biosciences	Cat#564322; RRID: AB_2738744
BV711 anti-Mouse CD192 (CCR2) (clone 475301)	BD Biosciences	Cat#747964; RRID: AB_2872425
Anti-mouse IFNγ (clone H22)	Bioxell	Cat#BE0312; RRID: AB_2736992
Anti-mouse TNFα (clone TN3-19.12)	Bioxell	Cat#BE0244; RRID: AB_2687725
Anti-mouse IFNαR (clone MAR1-5A3)	Bioxell	Cat#BE0241; RRID: AB_2687723
Anti-mouse IL-1R (clone JAMA-147)	Bioxell	Cat#BE0256; RRID: AB_2661843
Anti-mouse IL-17A (clone 17F3)	Bioxell	Cat#BE0173; RRID: AB_10950102
Anti-mouse Ly6G (clone 1A8)	Bioxell	Cat#BE0075-1; RRID: AB_1107721
rat IgG2a isotype control (clone 2A3) for anti-mouse Ly6G	Bioxell	Cat#BE0089; RRID: AB_1107769
polyclonal Armenian hamster IgG isotype control for anti-TNFα	Bioxell	Cat#BE0091; RRID: AB_1107773
mouse IgG1 isotype control (clone MOPC-21) for anti-IL-17A	Bioxell	Cat#BE0083; RRID: AB_1107784
Anti-mouse CD4 (clone GK1.5)	Bioxell	Cat#BE0003-1; RRID: AB_1107636
Bacterial and Virus Strains
Mycobacterium tuberculosis H37Rv (Mtb)	Jun Liu (Ahn et al., 2018)	Grew up in house
Mycobacterium tuberculosis H37Rv:pUV15 (Mtb-GFP)	The plasmid pUV15 from Sabine Ehrt (Ehrt et al., 2005)	Grew up in house
Mycobacterium bovis BCG Tice: rBCG30-ARMF-II Tice	Marcus A. Horwitz (Horwitz et al., 2000)	Grew up in house
LCMV Clone 13	Michael Oldstone, Scripps	Grew up in house
LCMV Armstrong	Michael Oldstone, Scripps	Grew up in house
Biological Samples
NA	NA	NA
Chemicals, Peptides, and Recombinant Proteins
Mycobacterium tuberculosis antigen peptide Ag85B240-260 H2N-FQDAYNAAGGHNAVFNFPPNG-ON	New England Peptide	Custom order
Mycobacterium tuberculosis antigen peptide ESAT64-17 H2N-QQWNFAGIEAAASA-OH	New England Peptide	Custom order
Mycobacterium tuberculosis antigen peptide TB10.44-11 H2N-IMYNYPAM-OH	New England Peptide	Custom order
Ovalbumin, Alexa Fluor™ 488 Conjugate	Invitrogen™	Cat#O34781
Cell-ID Cisplatin	Fluidigm	Cat#201064
Cell-ID Intercalator-Irdium—125 μM	Fluidigm	Cat#201192A
Zombie Aqua™ Fixable Viability Kit	Biolegend	Cat#423102
Mouse IL-2 Recombinant Protein	Gibco	Cat#PMC0025
Recombinant Mouse GM-CSF (carrier-free)	Biolegend	Cat#576306
Recombinant Mouse IL-4 (carrier-free)	Biolegend	Cat#574306
Lipopolysaccharides from Escherichia coli O111:B4	Sigma	Cat#L4391-1MG
Brefeldin A	Sigma	Cat#B7651-5MG
Cytochalasin D	Sigma	Cat#C8273-1MG
Hematoxylin Solution, Harris Modified	Sigma	Cat#HHS32-1L
Eosin Y solution, alcoholic	Sigma	Cat#HT-110132
Collagenase from Clostridium histolyticum	Sigma	Cat#C5138-1G
DNase I	Sigma	Cat#DN25-1G
Middlebrook 7H9 Broth	BD	Cat#BD 271310
Middlebrook 7H10 Agar	BD	Cat#BD 262710
BBL™ Middlebrook OADC Enrichment	BD	Cat#B11886
Middlebrook ADC Enrichment	BD	Cat#B11887
Cell lysis buffer	eBioscience	Cat#EPX-99999-000
Fixation Buffer	Biolegend	Cat#420801
Intracellular Staining Permeabilization Wash Buffer (10X)	Biolegend	Cat#421002
Tag-it Violet™ Proliferation and Cell Tracking Dye	Biolegend	Cat#425101
Critical Commercial Assays
EasySep™ Mouse CD4+ T Cell Isolation Kit	STEMCELL	Cat#19852
Foxp3 / Transcription Factor Staining Buffer Set	eBioscience	Cat#00-5523-00
ProcartaPlex Mouse Cytokine & Chemokine Panel 1A 36plex	Invitrogen	Cat#EPX360-26092-901
LEGEND MAX™ Mouse TNF-α ELISA Kit	Biolegend	Cat#430907
LEGEND MAX™ Mouse IFN-γ ELISA Kit	Biolegend	Cat#430807
LEGEND MAX™ Mouse IL-17A ELISA Kit	Biolegend	Cat#432507
KC/CXCL1 Mouse ELISA Kit	Thermofisher	Cat#EMCXCL1
LIX/CXCL5 Mouse ELISA Kit	Thermofisher	Cat#EMCXCL5
G-CSF (CSF3) Mouse ELISA Kit	Thermofisher	Cat#EMCSF3
IL-1 alpha Mouse ELISA Kit	Invitrogen	Cat#BMS611
20-Plex Pd Barcoding Kit	Fluidigm	Cat#201060
Deposited Data
NA	NA	NA
Experimental Models: Cell Lines
NA	NA	NA
Experimental Models: Organisms/Strains
C57BL/6J mice	Jackson laboratory	Cat#000664
Ag85B240-254 specific (P25) TCR transgenic mice	Jackson laboratory	Cat#011005
ESAT64-17 specific (C7) TCR transgenic mice	From Kevin B. Urdahl (Moguche et al., 2017)	NA
CD4 KO mice	Jackson laboratory	Cat#002663
Oligonucleotides
NA	NA	NA
Recombinant DNA
NA	NA	NA
Software and Algorithms
Prism 9	GraphPad Software, Inc	www.graphpad.com
Flow Jo version 10.7.1	BD FLOWJO	www.flowjo.com
BD FACSsuite v1.4.0.7047	BD	www.bdbiosciences.com
Bio-Plex manager version 6.1	BioRad	RRID:SCR_014330
NDP viewer2	Hamamatsu	https://www.hamamatsu.com/us/en/product/type/U12388-01/index.html
Cytobank	Cytobank, Inc	www.cytobank.org
R	R Core Team	www.r-project.org
UMAP	(Melville, 2018)	RRID:SCR_018217
PhenoGraph	(Levine et al., 2015)	RRID:SCR_016919
diffcyt R	(Weber et al., 2019)	https://bioconductor.org/packages/release/bioc/html/diffcyt.html
Cytofkit	(Chen et al., 2016)	https://github.com/JinmiaoChenLab/cytofkit
Other
ESAT64-17/I-Ab class II MHC tetramer (QQWNFAGIEAAASA)	NIH tetramer Core	NA
TB10.44-11/H-2Kb class I MHC, APC labelled monomer (IMYNYPAM)	NIH tetramer Core	NA
HCLIP I-A(b) class II MHC tetramer PVSKMRMATPLLMQA	NIH tetramer Core	NA

Single cell suspensions from individual samples were washed with PBS and pulsed with 12.5μM Cisplatin in PBS for 1min prior to quenching with CyTOF staining media [Mg+/Ca+ HBSS containing 2% FBS (Multicell), 10mM HEPES (Corning), and FBS underlay]. Cells were then fixed for 12min at room temperature with transcription factor fixative (eBiocience, 00-5523-00), permeabilized and individual samples barcoded according to manufactures instructions (Fluidigm 20-Plex Pd Barcoding Kit, 201060), prior to being combined. Combined samples were resuspended in staining media containing metal-tagged surface antibodies and Fc block (CD16/32; Bioxcell) for 30minutes at 4°C. Cells were washed, then permeabilized and stained with metal tagged intracellular antibodies using Transcription Factor Staining Buffer Set. Cells were incubated overnight in PBS containing 0.3% (w/v) saponin, 1.6% (v/v) paraformaldehyde (Polysciences Inc) and 50nM Iridium (Fluidigm). Cells were analyzed on a Helios or Helios2 mass cytometer (Fluidigm) at The Hospital for Sick Children (SickKids) Center for Advanced Single Cell Analysis (CASCA), funded through the SickKids Research Institute and the Canadian Foundation for Innovation. EQ Four Element Calibration Beads (Fluidigm) were used to normalize signal intensity over time using CyTOF software version 6.7. FCS files were manually debarcoded and analyzed using Cytobank6.2 (Cytobank, Inc).

Heatmaps were generated in R using the blue-red, plasma or viridis, color packages and the gplots package. Data was pre-transformed by ‘cytofAsinh’ in R package ‘cytofkit’ or by Arcsinh transformation in software ‘Cytobank’. Phenogragh analysis was performed with R package ‘Cytofkit’(Chen et al., 2016; Levine et al., 2015a) (data transformation = cytofAsinh, seeds = 42, k=30 for analysis in Figure. 2A, K=100 for analysis in Figure. 5B). UMAP analysis was performed with R package ‘UMAP’ [Uniform Manifold Approximation and Projection(Melville, 2018)] with default settings. Differential states and differential abundance were calculated using the limma and edgeR tests respectively, through the ‘diffcyt’ R package (Weber et al., 2019). An adjusted p-value <0.05 was considered significant for the comparisons made. Median signal intensity values were used to generate heatmaps.

Tissue isolation and flow cytometry

Lungs and mediastinal lymph nodes (medLNs) were digested in RPMI 1640 medium with 10% FBS, 1mg/ml Collagenase from Clostridium histolyticum (C5138 Sigma) and 0.15mg/ml DNase I (Sigma) at 37°C for 1 hour (for the lungs) or for 30 mins (for the medLNs), and then were processed in RPMI 1640 medium. Antibodies used for flow cytometry are listed in key resources table. Surface staining was preformed ex vivo using antibodies to CD45 (30-F11), Siglec-F (E50-2440), CD11c (3.9), MHC-II (M5/114.15.2), CD11b (M1/70), CD103 (2E7), CCR2 (475301), Ly6C (HK1.4), B220 (RA3-6B2), CCR7 (4B12), Ly5.1 (A20), Ly5.2 (104), CD8 (53-6.7). Staining for transcription factors T-bet and RORγT was performed using the Foxp3 staining kit (eBioscience). The ESAT6 tetramer and the TB10.4 tetramer was obtained from the NIH Tetramer Core. Samples were run on a FACSVerse and FACSLyric (BD Biosciences) and data analyzed using Flow Jo software (BD FlowJo).

Labeled OVA intranasal transfer

Ovalbumin conjugated with Alexa Fluor™ 488 (labeled OVA) from Invitrogen (cat # 34781) was reconstituted at 2mg/ml in PBS. The labeled OVA stock was diluted in PBS and utilized in intranasal transfer with a dose 20μg in 20μl volume per mouse.

Isolation and labeling of transgenic CD4 T Cells

CD4 T cells were isolated from the spleens of naive P25 or C7 mice using a negative selection mouse CD4 T Cell Isolation Kit (Stem Cell Technologies). Isolated CD4 T cells were labeled in PBS with 2.5mM Tag-it Violet™ proliferation dye (Biolegend) according to manufacturer’s instructions. For in vivo transfer, 10⁶ naive P25 or C7 CD4 T cells were injected intravenously per mouse 5 days before Mtb infection.

Intracellular cytokine re-stimulation

For cytokine staining, lung cells were re-stimulated for 5 hours at 37°C with 5μg/ml of MHC class II-restricted Mtb-peptide Ag85B_240-260 or ESAT6_4-17 peptides, or 2μg/ml of the MHC-I restricted TB10.4_4-11 peptide in the presence of 50U/ml recombinant murine IL-2 and 1mg/ml brefeldin A (Sigma). Following the in vitro re-stimulation, cells were stained with a fixable viability stain, Zombie Aqua (BioLegend) and stained for extracellular proteins. The cells were then fixed with fixation buffer (Biolegend), permeabilized, and stained with intracellular staining perm/wash buffer (Biolegend) and antibodies recognizing IFNγ (XMG1.2), TNFα (MP6-XT22) or IL-17A (TC11-18H10.1) (BioLegend).

Luminex assays and ELISAs

For Luminex assays and ELISAs, the whole lung tissue was harvested and immediately homogenized on ice in Cell Lysis Buffer (EPX-99999-000, eBioscience) with 1 mM Phenylmethylsulfonyl fluoride (Sigma). For Luminex, ProcartaPlex Mouse Cytokine & Chemokine Panel 1A 36plex (Invitrogen) was used according to the user manual to evaluate the amount of 36 targets in the lung homogenate. Data acquisition and calculation was performed using the Bio-Plex® 200 Systems (BioRad). For ELISAs, kits were used according to the user manual for detection of IL-1α (Invitrogen), CXCL1, CXCL5 and G-CSF (Thermofisher), IFNγ, IL-17A and TNFα (Biolegend).

Bone marrow dendritic cell generation and transfer

bmDCs were generated as previously described (Roney, 2013). Briefly, 2x10⁶ bone marrow cells from Ly5.1+ C57BL/6 mice were cultured at 37°C in 10ml bmDC culture medium (RPMI 1640 medium with 10% FBS, penicillin/streptomycin and 2-Mercaptoethanol) with 40ng/mL GM-CSF (BioLegend). At day 3, 10 ml fresh bmDC culture medium with 40ng/ml GM-CSF was added. At day 6 and 8, 10ml media was removed and replaced with fresh bmDC culture medium with 20ng/ml (day 6) or 10ng/ml (day 8) GM-CSF and 20ng/ml IL-4. Loosely adherent cells were harvested at day 10 and stimulated with 1μg/mL LPS (Sigma-Aldrich) for 24 hours. Cells were then separately pulsed with 5μg/ml peptide Ag85B_240-260 peptide or ESAT6 _4-17 peptide in bmDC culture medium for 2 hours at 37°C. For the T cell priming experiment, one million of each Ag85B peptide-pulsed bmDCs and ESAT6 peptide-pulsed bmDCs were combined and injected per mouse intravenously at day 6 after Mtb infection. The number of transferred bmDC was matched for peptide labeled and peptide unlabeled groups.

For the bmDC migration experiments, bmDCs are differentiated and activated the same as above. One million bmDCs were intranasally transferred in 20μl PBS per mouse at day 29 after LCMV Cl13 infection. Twelve hours later (day 30 after LCMV Cl13 infection), intranasal infection with Mycobacterium bovis BCG (rBCG30, 10⁶ CFU per mouse) was conducted to provide mycobacterial stimulation to the pulmonary environment.

Histological analysis

For histology analysis, lungs were harvested and fixed in 10% PFA solution for at least two weeks at room temperature. Lung specimens were processed in an automatic processor (Leica ASP300). Briefly, specimens were dehydrated in an ethanol series, treated with xylene and three paraffin baths before being embedded in paraffin. Sections were cut at 4-4.5μm thickness, dried at 42°C overnight, and subsequently stained with Harris Hematoxylin (Sigma HHS-32) and alcoholic eosin (Sigma HT-110132). After cover-slipping with a xylene based mounting medium (ShurMount, Fisher), sections on slides were digitally scanned at 20x or 40x on a Hamamatsu Nanozoomer (2.0HT). Histopathology scores were evaluated without knowledge of the group information by a pulmonary pathologist with extensive clinical experience. In brief, three grades of the lung lesions were classified and scored by grade. Grade 1 lesions exhibit infiltration of alveolar septa by chronic inflammatory cells; Grade 2 lesions exhibit more extensive alveolar and interstitial infiltration by sheets of histiocytes and lymphocytes, including multinucleated (giant) histiocytes, plus prominent perivascular lymphocytic infiltrate; Grade 3 lesions exhibit large areas of dense lymphocytic and histiocytic infiltrates with areas of necrosis. The lesion area percent (0 to 100%) multiplied by grade score were summed up as histopathology scores for each lung section.

Phagocytosis assay

Lung cells from naive mice and from chronic LCMV infected mice were resuspended in RPMI1640 medium (10% FBS) and then plated in Ultra-Low attachment 96 well plates (corning, REF3474) (2x10⁶ cells in 0.1ml per well). 10μM cytochalasin D (sigma) was added for negative control wells to inhibit the phagocytosis. After 30 mins pre-incubation (37°C, 5% CO₂), 2x10⁶ CFU Mtb-GFP resuspended in 0.1ml PPMI1640 medium were added to each well and mixed gently with lung cells. The incubation (37°C, 5% CO₂) was continued for another 2 hours. Lung cells were then washed with cold PBS twice (centrifuge speed: 200g for 5 mins) to remove extracellular Mtb. At last, lung cells were stained with antibodies labelling alveolar macrophages, dendritic cells and other lung phagocytes to examine the proportion and cell number of Mtb-GFP positive cells.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical parameters and the number of animals per group are described in the figure legends. In all figures, error bars indicate the standard error of the mean (SEM). One-way ANOVA (two-tailed, unpaired, tukey's multiple comparisons test or two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli analysis) for multiple comparisons and Student’s t tests (two-tailed, unpaired) for comparing two conditions were performed using GraphPad Prism software. The calculation of differential states and differential abundance in CyTOF data analysis is described in CyTOF section above. P-value <0.05 was defined as significant. For data shown in figures with log-scale y-axis, data was pre-transformed (log10) to conform to normality. For survival assays, log-rank analysis (Mantel-Cox test) was performed using GraphPad Prism software.

Highlights.

Chronic virus infection induced TNFα decreases early Mtb growth to impede LN transit

Arrested Mtb LN transport delays T cell priming and promotes Th17 differentiation

Increased IL17 promotes pulmonary neutrophilia, impairing survival during coinfection

Overcoming the T cell priming delay restores control of Mtb coinfection