Female and Male K18-hACE2 mice [strain: 2B6.Cg-Tg(K18-ACE2)2Prlmn/J], 6–8 weeks of age were purchased from Jackson Laboratory (Bar Harbor, ME). Mice were housed under pathogen-free conditions at Seattle Children’s Research Institute (SCRI) biosafety level 3 animal facility and were handled in accordance with the specific guidelines of SCRI’s Institutional Animal Care and Use Committee (IACUC). Mice were infected with a low dose (50–100 bacteria) aerosol (LDA) of M.tb HN878 using a Glas-Col whole-body exposure chamber (Glas-Col, Terre Haute, IN). Twenty-four hours post challenge the lungs of three mice were homogenized and plated on Mitchison 7H11 agar (Thermo Fisher Scientific, Waltham, MA) to confirm delivery of 50–100 CFU per mouse. For SARS-CoV-2 infection mice were first put under anesthesia with intraperitoneal (i.p.) administration of Ketamine (Patterson Veterinary, Loveland, CO) and Xylazine (Patterson Veterinary). SARS-CoV-2 clinical isolates were administered at 200 PFU via intranasal installation of 40µL per nare. Following SARS-CoV-2 infection mice were weighed daily. Animals that reached 20% weight loss and/or exhibited physical signs of morbidity were humanely euthanized.

Vero TMPRSS2 (National Institute for Biological Standards and Control (NIBSC), Hertfordshire, England), Vero E6 (ATCC, Manassas, VA), and Calu-3 epithelial cells (ATCC) were maintained at 37 °C + 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 1% penicillin/streptomycin (cDMEM). Cells were tested regularly for mycoplasma with Mycoplasma PCR detection kit (MilliporeSigma, Burlington, MA).

SARS-CoV-2 Beta (hCoV-19/SouthAfrica/KRISP-EC-K005321/2020) was obtained from BEI Resources and housed under standard BSL-3 laboratory conditions. SARS-CoV-2 virus was propagated and titered by plaque assay in Vero E6 cells. Cultured cells were infected with the original stock at a MOI of 0.1 and incubated at 37°C + 5% CO₂ for 72 h. Supernatants were harvested, centrifuged to remove debris, aliquoted and frozen at −80°C.

At the indicated time points harvested organs were homogenized in DMEM using gentleMACS Octo Tissue Dissociator (Miltenyi, Bergisch Gladbach, Germany). Serial dilutions of organ homogenates were made in PBS with 0.05% Tween80, and aliquots of dilutions were plated on Middlebrook 7H10 agar tri-plates (Molecular Toxicology, Boone, NC), as previously described (46, 47). After 3–4 weeks of incubation at 37°C + 5% CO₂, colony counts were recorded. Bacterial burden, in colony forming units (CFU) per organ, was calculated, and expressed as Log10.

Viral burden was measured with the plaque forming assay (PFA) using similar techniques described previously (48). Vero TMPRSS2 cells were plated in 6-well plates one day prior to titers at 4.8x10⁵ cells/mL in 2mL of cDMEM per well. Harvested organs were homogenized in DMEM containing 1% FBS (D1 media) using the gentleMACs Octo Dissociator. Organ homogenates were serially diluted ten-fold using D1 media and added dropwise to the plated Vero cells. Plates were incubated at 37 ˚C + 5% CO₂ for 60 minutes, with 15-minute intervals of rocking plates in all directions. After 60 minutes, 2mL of overlay media comprised of D1 media and 0.2% agarose was added to each well and incubated at 37°C + 5% CO₂ for 48 hours. Cells were then fixed with 2 mL of 10% Formaldehyde solution and incubated at room temperature for 30 minutes. The overlay was removed, and cells stained with 1mL of Crystal Violet (BD Biosciences, Franklin Lakes, NJ) per well for 20 minutes. Lastly, each well was washed with 1mL of PBS and the number of plaques in each well were recorded.

At the indicated time points, three whole lung and accessory lobes were collected per group and fixed in 10% Neutral Buffer Formalin (NBF) for 24 hours. The fixed lung samples were embedded in paraffin and sectioned by the University of Washington histology core. Blinded slides were sent to Colorado State University and stained with hematoxylin and eosin (H&E) then analyzed by veterinary pathologist Dr. Brendan Podell as previously published (46, 47). H&E stained sections were scanned at 20X magnification using an Olympus VS120 microscope, Hamamatsu ORCA-R2 camera, and Olympus VS-ASW 2.9 software. Visiopharm software was used for image analysis. For each tissue section, a region of interest (ROI) was generated at a low magnification with a custom tissue detecting algorithm using decision forest training and classification to differentiate tissue versus background based on color and area. Lesions were identified within tissue ROIs at a high magnification with an additional custom-made algorithm using decision forest training and classification based on staining intensity, color normalization and deconvolution, area, and morphological features. Percent lesion calculations were integrated into the same algorithm and calculated from tissue area and lesion area as designated by the ROI and lesions detected. Lesion identification and quantification were then reviewed by Dr. Podell (46, 47).

Cell populations within the lung were measured kinetically utilizing methods previously published (47). Briefly, lung homogenates were incubated in RBC lysis buffer (eBioscience/Thermo Fischer Scientific), washed and resuspended in RPMI 1640 + 10% FBS, and then evenly dispensed into 96-well round bottom plates. Cells were stained for surface markers with fluorochrome-conjugated monoclonal antibodies against mouse Ly6G (FITC, clone 1A8, Biolegend), Ly6C (PerCP-Cy5.5, clone HK1.4, eBioscience), MHCII I-A/I-E (eF450, clone M5/114.15.2, Invitrogen), CD11c (Bv510, clone N418, Biolegend), CD3 (Bv650, clone 17A2, Biolegend), CD19 (APC, clone 6D5, Biolegend), CD11b (Alexa700, clone M1/70, eBioscience), NK1.1 (PE, clone PK136, eBioscience), CD64 (PE-Cy7, clone X54–5/7.1, Biolegend) and 1 μg/mL of Fc receptor block anti-CD16/CD32 (clone 93, eBioscience) in PBS with 1% bovine serum albumin (BSA) for 15 minutes at room temperature. Samples were washed and before removing samples from the BSL3, samples were incubated in 4% paraformaldehyde for 30 minutes. After wash and resuspension in PBS + 1% BSA, cells were acquired on a BD Bioscience LSRII flow cytometer (BD Bioscience) and analyzed using FlowJo version 10.8.1 (BD Bioscience).

Bronchoalveolar lavage fluid (BALF) was collected by flushing lungs with 1X PBS, then centrifuged at 400g for 7 minutes to remove cellular debris and filtered. The processed BALF was used in the Meso Scale Discovey (MSD) V-PLEX Proinflammatory Panel 1 Mouse kit (#K15048D), V-PLEX Cytokine Panel 1 Mouse Kit (#K15245D) and U-PLEX Interferon Combo 1 (#K15320K) to measure cytokine levels on the MESO QuickPlex SQ 120MM (Meso Scale Diagnostics, Rockville, MD). For similar in vitro endpoints from cultured Calu-3 supernatants, an MSD human U-PLEX Viral Combo 1 kit was used (#K15343K-1).

Accessory lung lobes from mice at specified time points post SARS-CoV-2 co-infection were harvested and homogenized in 900µL of Qiazol, followed by RNA extraction using the QIAGEN RNeasy Plus Universal mini kit according to the manufacturer’s protocol (QIAGEN, Hilden, Germany). RNA was eluted into 30μl. RNA concentration and quality was determined using the NanoDrop 8000 (Thermo Fisher Scientific) and stored at -80°C until assayed. The obtained RNA was then utilized in the High-Capacity RNA-to-cDNA kit for cDNA synthesis using SuperScript™ IV Reverse Transcriptase (Thermo Fisher Scientific), containing a reverse transcriptase with a high-fidelity enzyme following manufacture protocol.

For Fluidigm Real-Time PCR and Dynamic Array IFC (Integrated Fluidic Circuit) Setup, specific target amplification (STA) was done as per the manufacturer’s recommendations as the initial step (pre-amplification of cDNA) for the Biomark HD system (Standard BioTools, formerly Fluidigm) carried out on the Standard BioTools 48.48 Gene Expression (GE) Dynamic Array integrated fluidic circuit (IFC) (49). Assay-sets (primers only) were combined as a delta gene multiplex pool (see Supplementary Table 1 ). Preamplification was carried out for each cDNA sample against a reaction-set. Exonuclease I was then used to clean up the preamplification reactions.

Subsequently, the Biomark Chip was primed, and assay premix for each target was aspirated into the IFC assay inlets for a final concentration of 9 µM primers and 2.5 µM probe per reaction, and pre-amplified samples were aspirated into sample inlets. The IFC was then run in the Biomark HD thermocycler, using the manufacturer-supplied thermal cycling conditions. Results were analyzed using the Fluidigm Real-time PCR Analysis software, where thresholds were manually defined, the baseline was automatically assigned, and a Cycle of quantification (Cq) cut-off value of 38 was applied. The cycle threshold (Ct) values for the candidate housekeeping gene, RPL13, and target genes were obtained, and delta-delta CT values were calculated.

For the in vitro experiments, frozen human peripheral blood mononuclear cells (PBMCs) and whole blood were procured from Bloodworks Northwest (Seattle, WA). PBMCs were thawed, counted, and resuspended to a concentration of 2x10⁶ cells/mL, then rested overnight in RPMI media supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C + 5% CO₂. Cells were counted the next day and viability was assessed before being adjusted to a concentration of 1.5x10⁶ viable cells/mL. Subsequently, the cells were infected with M.tb HN878 at a multiplicity of infection (MOI) of 1 and incubated for 96 hours. Following infection, the cells were centrifuged at 700g for 3 minutes, and the supernatants were collected and filtered through a 0.22-micron filter. Vero or Calu-3 cells were plated and treated with supernatants for 24 hours at 37°C + 5% CO². Media-only treated cells were used as controls. Post-treatment, the cells were challenged with 75 PFU of SARS-CoV-2 Beta, and plaques were recorded 48 hours post-infection using the viral titer PFA described above.

To assess mRNA expression in cultured cells, RNA was extracted from cultured Calu-3 cells using the QIAGEN RNeasy Plus Universal mini kit. The cells were harvested using 900µL of Qiazol, followed by RNA isolation and cDNA synthesis employing the High-Capacity RNA-to-cDNA kit. RNA and cDNA concentration and quality was determined using the NanoDrop 8000 (Thermo Fisher Scientific) and stored at -80°C until assayed. Quantification of mRNA levels was performed using the GoTaq qPCR and RT-qPCR Systems kit from Promega, following the manufacturer’s protocol, and the StepOne Plus Real-Time PCR System (Thermo Fisher Scientific). The mRNA expression levels of Calu-3 cells are presented as Log2 fold change (FC) compared to media-only treated cells and normalized to the housekeeping gene, Beta Actin. Primers used were selected from published sequences in PrimerBank (RRID: SCR_006898) (see Supplementary Table 2 ).

For neutralization studies, PBMCs were thawed and counted as described previously. PBMCs were plated in 12-well plates at 2.25x10⁶ cells/mL in one mL of RPMI media supplemented with 10% FBS and 1% penicillin/streptomycin. Neutralizing antibodies for human CD4 (BE0351, BioXcell), Lebanon, NH), CD8α (BE0004–2, BioXcell), CD314 (BE0288, BioXcell), and relevant isotype controls, mouse IgG1(BE0083, BioXcell), and mouse IgG2 (BE0085, BioXcell) were then administered at 10µg/mL, 20µg/mL and 100µg/mL in 1 mL. PBMCs and neutralizing antibodies were incubated for 1 hour at 37°C + 5% CO₂ prior to infection with M.tb HN878 at MOI of 1 for 96 hours. For IFNγ neutralization, the IFNγ antibody (BE0235, BioXcell) was added directly to the supernatant from PBMCs infected with M.tb HN878 at 10µg/mL, 20µg/mL and 100µg/mL escalating doses. After infection, supernatants were filtered through a 0.22-micron filter. Vero cells were treated with filtered PBMC supernatants for 24 hours, then challenged with 75 PFU of SARS-CoV-2 Beta, and plaques were recorded 48 hours post-infection using the viral titer PFA described above.

Survival analysis was based on the Mantel-Cox log-rank test with Bonferroni correction for multiple comparisons and carried out using GraphPad Prism 9.3.1 (GraphPad Software, San Diego, CA). Bacterial burden, viral titers, cytokine levels, and cell populations (percent frequency and counts) were assessed at a single time point using one-way ANOVA with Tukey’s multiple comparison test to compare between infection groups. Flow cytometry data was assessed using FlowJo v10.8.1 (BD) and statistical analyses were performed using GraphPad Prism 9.3.1 software. The graphics were made with Biorender. Heat maps of in vivo mRNA expression were created with RStudio using ‘pheatmap’ function. P values < 0.05 were considered significant and labeled accordingly in each of the figures (* = P<0.05, **= P<0.01, ***=P<0.001, ****= P<0.0001). Outliers were identified using Grubbs’ test at alpha 0.05.

We hypothesized that infecting mice first with M.tb to induce an active infection and subsequent immune response, followed by co-infection with SARS-CoV-2, could potentially alter disease outcomes and affect survival endpoints. Highly virulent W-Beijing clinical strain, M.tb HN878, was delivered as a low dose aerosol challenge (LDA, 50–100 bacteria) to female and male K18-hACE2 mice. Three weeks post-M.tb infection, mice were challenged with 200 plaque forming units (PFU) of SARS-CoV-2 Beta. ( Figure 1A ). Male and Female mouse cohorts (n=10 per sex) were assessed for survival following infection with M.tb, SARS-CoV-2, or co-infection. Mice singularly infected with SARS-CoV-2 had significantly lower survival rates compared to those in the saline or M.tb infected groups. However, the M.tb and SARS-CoV-2 co-infected group showed a significantly higher survival rate than the group infected solely with SARS-CoV-2 ( Figure 1B ). The increased survival amongst the co-infection group versus the singular SARS-CoV-2 infection group suggests prior M.tb infection may provide partial protection from SARS-CoV-2 challenge, in alignment with our hypothesis.

Viral titers of SARS-CoV-2 were evaluated locally and systemically to determine if co-infected mice have differences in viral load magnitude or distribution. In alignment with prior work (43–45), the co-infected group exhibited decreased lung viral titers at day 1 post-co-infection and significantly decreased lung and lymph node viral titers at day 3 post-co-infection, the anticipated viral peak of our collection timeline, compared to SARS-CoV-2 alone cohorts ( Figure 1C ). There was no difference in viral burden in brain samples when comparing the two infection groups ( Figure 1C ). There was no difference in CFU between the groups for all organs and time points, suggesting the exhibited protection from morbidity was not due to a change in M.tb burden ( Figure 1D ).

We have previously seen that bacterial burden can be uncoupled from pulmonary pathology in mouse models of TB, where pulmonary disease and morbidity endpoints may be driven more by host factors (46). Interestingly, when assessing the lung pathology in these co-infection studies ( Figure 2A ) there was no significant difference in percent lesion area between co-infected animals and the SARS-CoV-2-only infected group at day 1 and 3 ( Figures 2B, C ). However, by day 7 there was a trend towards decreased lesion scores in the co-infected groups compared to the M.tb only infection group ( Figure 2D ), which became significant by day 14 ( Figure 2E ). This trend has been published previously (44) and speaks to the complexities of lung pathology in the context of co-infection. Conducting additional analysis which more clearly defines the differences between TB and COVID-19 pathology is worth further exploration. While these data suggest infection with SARS-CoV-2 may help resolve acute lesions from existing M.tb infection ( Figure 2E ), the primary focus of this work is to determine how infection with M.tb establishes an inhospitable pulmonary environment for SARS-CoV-2.

Given the partial protective phenotype displayed by the co-infection model, the immune profiles of M.tb-infected, SARS-CoV-2-infected and co-infected animals were evaluated. Using flow cytometry, the kinetic influx of immune cells to the lung following co-infection were compared to the other cohorts. Both M.tb-infected mice (21 days post infection) and M.tb-infected mice subsequently infected with SARS-CoV-2 (co-infected group) showed elevated levels of neutrophils and macrophages at the day 1 and 3 post-virus challenge time points, and increased influx of T cells and NK cells at day 3 compared to the group infected with SARS-CoV-2 only ( Figure 3 ). Absolute cell counts mirroring these trends were also observed ( Supplementary Figure 3 ). This showcases the influence chronic M.tb infection has on the inflammatory environment of the lung. Bronchioalveolar lavage fluid (BALF) from M.tb-infected and co-infected mice contained significantly elevated IL-6, TNF-α, IFNγ, IP-10, MIP-1a, MCP-1, and KC-GRO at days 1 and 3 post-co-infection, while the SARS-CoV-2 infection group displayed delayed induction of these same effector molecules until day 7 ( Figure 4 ). Interestingly, there was an absence of strong kinetic patterns of increases or persistent decreases in inflammatory gene expression within the lung across infection groups ( Figures 5A-C ). On day 3 there were trends of increased expression in certain ISGs, PRRs and inflammatory pathway genes in M.tb and co-infected groups ( Figure 5B , Supplementary Figure 4 ).

The increased inflammatory response observed in the co-infected group was initially surprising since the robust pro-inflammatory response has been identified to be detrimental in SARS-CoV-2 infection (13). However, the timing and type of immune response induced may be important for priming the lung to combat SARS-CoV-2 infection. Mechanistic in vitro studies were next used to evaluate which innate immune response induced by the primary M.tb infection provide protection against subsequent SARS-CoV-2 infection.

To model co-infection in vitro, PBMCs were infected with M.tb, as they serve as niche host cells and are responsive to infection. Conversely, SARS-CoV-2 infection was modeled using permissive epithelial cells, which fulfill a similar role. While lung epithelial cells show limited direct responsiveness to M.tb, they exhibit heightened reactivity and transcriptional changes when exposed to M.tb-infected myeloid cells (50). Our investigation aimed to determine whether cytokines generated during initial M.tb infection of immune cells could confer protection against secondary SARS-CoV-2 infection in susceptible bystander epithelial cells.

We used PBMCs from healthy male and female donors, collected before and after 2019, as well as from Bacille Calmette-Guérin (BCG)-immunized donors ( Supplementary Table 3 ), to investigate the effects of prior SARS-CoV-2 exposure or BCG immunization on immune responses. There was additional interest in investigating prior BCG immunizations given the attenuated M. bovis vaccine is currently the only licensed TB vaccine and regularly administered in TB endemic regions. While early in the pandemic there were hypotheses that prior BCG immunization may provide protection against SARS-CoV-2 (51), these claims were later dispelled in clinical studies (52, 53). Frozen PBMCs were thawed and either mock-infected or infected with M.tb HN878 at a MOI of 1 for 96 hours ( Figure 6A ). Supernatants were harvested, filtered, and applied to Vero cells, which are highly permissive to SARS-CoV-2 infection, to assess whether cytokines alone could confer protection against SARS-CoV-2 infection. Treatment of Vero cells with supernatants from M.tb-infected PBMCs resulted in significantly reduced viral titers, with no significant differences observed among PBMC groups defined by date or vaccination history ( Figure 6B ). These findings were confirmed using more physiologically relevant Calu-3 human airway epithelial cells (54–56) where diminished viral titers were observed in samples pre-treated with supernatants from M.tb-infected PBMCs ( Figure 6C ). To define the essential elements of protection, the cytokine levels within the supernatants were quantified, revealing increased production of several proinflammatory cytokines, including G-CSF, GM-CSF, TNF-α, IL-1β, IL-6, and IFNγ, following M.tb infection compared to mock-infected PBMCs ( Figure 6D ). Given the absence of significant differences between PBMC groups, subsequent experiments were conducted using PBMCs collected prior to 2019.

Supernatant-treated Calu-3 cells were then used in RT-qPCR analysis to determine if cells underwent transcriptional changes upon treatment with supernatants. Treated Calu-3 cells showed significantly increased expression of ISGs such as OAS1, OAS3, MX2 and notably, IFIH1, the gene encoding MDA5, a primary PRR for SARS-CoV-2, compared to media-treated cells ( Figure 7A ). Subsequently, 24 hours post-SARS-CoV-2 infection, the expression of these ISGs increased in both control and M.tb-infected PBMC supernatant-treated cells, with a significant increase in expression sustained in the supernatant-treated cells ( Figure 7B ). These findings support our hypothesis that prior M.tb infection primes epithelial cells towards resisting viral infection by inducing ISG expression.

While pre-treatment led to transcriptional changes and protection from SARS-CoV-2 after 48 hours of infection, pinpointing the stage of the viral infection cycle that may be affected was of interest. To align with our transcriptional data, Calu-3 cells were treated with supernatants or media (positive control) and infected with SARS-CoV-2 for 1, 6, 24, and 48 hours, then assessed for viral load. Following 1 hour of infection, no significant differences in viral load were observed ( Figure 8A ), suggesting no influence or perturbations in viral entry pathways—a result consistent with the expected SARS-CoV-2 doubling time of around 6 hours (57). However, after 6 hours of infection, there was a noticeable trend toward decreased titers in treated Calu-3 cells ( Figure 8B ). By 24 hours, treated cells displayed no plaques likely reaching the limit of detection ( Figure 8C ), suggesting that treated cells were not permissive to replication and actively eliminated the virus, thereby conferring protection.

To explore the mechanism of protection, an investigation into the involvement of specific cell types and cytokines was conducted. While type I IFNs are normally associated as the predominant anti-viral response, we did not see significant levels within our measurements. However, we did detect significant levels of IFNγ in both in vivo and in vitro models, and wanted to determine if blocking IFNγ would attenuate the observed protection. Accordingly, major cell types known to induce IFNγ were targeted.

Human PBMCs were co-incubated with neutralizing antibodies against CD4+ T cells, CD8α+ T cells, CD314+ NK cells, and IFNγ at increasing concentrations, and then infected with M.tb for 96 hours. Mouse IgG1 and IgG2 antibodies were used as isotype negative controls, while media-only treated cells served as a control for viral replication. After the 96-hour incubation period, supernatants were collected, as previously described, and used to treat permissive Vero cells to measure PFU following SARS-CoV-2 infection. Blocking of CD314+ NK cells did not result in a significant increase in viral titer, and CD8α+ T cells reached significance only at the highest concentration ( Figure 9 ). Neutralization of IFNγ led to diminished protection at 20 and 100 μg/mL, as evidenced by an increase in viral titer, highlighting its importance in conferring protection against SARS-CoV-2 ( Figure 9 ).

Interestingly, neutralization of CD4+ T cells resulted in an increased viral load with escalating antibody concentrations, suggesting that protection could be dependent on IFNγ and CD4+ T cell activity.