*2.2. Viral Propagation and Titration*

Human Epidermoid carcinoma strain 2 (Hep-2) cell line (American Type Culture Collection, CCL-23TM) (American Type Culture Collection, CCL-7TM) was used to propagate hRSV serogroup A2, strain 13018–8, a clinical isolate from Instituto de Salud Pública de Chile [28]. Hep-2 monolayers were grown in T75 flasks with Dulbecco's Modified Eagle Medium (DMEM) (Life Technologies Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco Invitrogen Corp, Carlsbad, CA, USA) until 80–90% confluency. Flasks containing 5 mL of DMEM 1% FBS for infection with hRSV, the viral inoculum with 2 × <sup>10</sup><sup>5</sup> plaque formation units (PFU), were incubated at 37 ◦C. After 2 h of virus adsorption, supernatants were replaced with fresh DMEM 1% FBS medium and incubated for 48 h (i.e., until cytopathic effects were detectable). For harvesting, cells were scraped, and the flask content was pooled and centrifuged first at 300× *g* for 10 min and then at 500× *g* for 10 min to remove cell debris. Using the same harvesting protocol, supernatants of non-infected cells were collected and used as the non-infectious control (referred to from here on as Mock). Viral titers of supernatants were determined by immunocytochemistry in 96-well plates with Hep-2 cells, as described previously [29,30].

### *2.3. Mycobacterium Bovis-BCG-Culture, and Storage*

The BCG Danish 1331 strain was grown in medium 7H9 (Sigma-Aldrich, Saint Louis, MO, USA, M0178-500G), a specific mycobacteria broth [30], supplemented with 10% Middlebrook oleic acid, albumin, dextrose, and catalase (OADC) Growth Supplement (Sigma-Aldrich, M0678-1VL), with constant stirring at 120 rpm until reaching an OD600 nm equal to 0.8. At this point, the mycobacteria culture was washed three times with 1X PBS-0.05% Tween 80, resuspended with 1X PBS-glycerol 50% at a final concentration of 1 × 106 colonyforming units (CFU)) per vial and frozen at −80 ◦C until their use. For the infection, BCG vials were centrifuged at 14,000× *g* and resuspended in PBS for intranasal administration.

## *2.4. Mouse Immunization and Viral Infection*

The effect of a recombinant BCG (rBCG) strain in the modulation of HO-1 was evaluated by immunization with rBCG expressing the nucleoprotein of hRSV (rBCG-N-hRSV) as described next. Six to eight-week-old BALB/cJ mice were immunized by sub-cutaneous <sup>1</sup> × <sup>10</sup><sup>8</sup> CFU of BCG WT or rBCG-N-hRSV in a final volume of 100 <sup>μ</sup>L per dose at days 0 and 14. Twenty-one days after immunization, mice were intraperitoneally anesthetized and challenged by intranasal infection with ~1 × <sup>10</sup><sup>7</sup> PFU of hRSV A2, strain 13018-8. On days 7 and 14 post-infection, mice were euthanized. rBCG production was performed as described previously [30].

#### *2.5. Mouse Viral and Mycobacterial Infections*

Two infection schemes were conducted to determine the consequences of hRSV infection and the effect of a subsequent infection. The short scheme was performed to evaluate if the inoculation with mycobacteria could infect and damage the pulmonary tissue a few days after the clearance of hRSV (day 10 post-infection with hRSV). Euthanasia was performed 11 days after the inoculation with mycobacteria since the alveolar tissue repair was not complete on this day. For this reason, it would be expected to find an effect induced by the administration of mycobacteria. The choice of this day was mainly due to the slow replicative cycle of this bacterium, which makes it difficult to carry out these tests at earlier times. The long scheme was performed to evaluate if, a significant number of days after the clearance of hRSV (day 21 post-infection with hRSV), the inoculation with mycobacteria could infect and damage the pulmonary tissue. During the day of the inoculation, the cell target of mycobacteria alveolar macrophages was replaced, allowing the mycobacteria more time to proliferate and cause an effect on the lung. Euthanasia was performed 21 days after the inoculation with mycobacteria since, on this day, the alveolar tissue repair is not complete. Six to eight-week-old C57BL/6J mice received an intranasal infection with 1 × <sup>10</sup><sup>7</sup> PFU of hRSV A2 strain 13018-8 of 100 <sup>μ</sup>L per mouse. After 10- or 21-days post-infection (dpi), mice were intranasally instilled with 1 × <sup>10</sup><sup>6</sup> CFU/mice of BCG. After 11- and 21-days post inoculation with BCG, mice were euthanized for collection of lung samples, bronchoalveolar lavage (BAL), and mediastinal lymph nodes. The controls were mock and vehicle (Sauton diluent) for the inoculation with hRSV and BCG, respectively. The administration of the first and second inoculation and their description are described in Table 1.

**Table 1.** Schemes of infection and their times.


#### *2.6. Evaluation of hRSV-Associated Disease Parameters*

To determine the infiltration of polymorphonuclear cells, BAL was collected as previously described [29] and stained with anti–CD11b PerCP-Cy5.5 (clone M1/70, BD Pharmingen, San José, CA, USA), anti-CD11c APC (clone HL3, BD Pharmingen), anti-IA/IE APCcy7 (clone M5/114.15.2, Biolengend), anti-Singlec F PE CF594 (clone E50-2240, BD Bioscience, San José, CA, USA), anti-Ly6C BV605 (clone HK1.4, Biolegend, San José, CA, USA) and anti-Ly6G FITC (Clone 1A8, BD Pharmingen) antibodies. As previously described, viral loads were detected in the lungs by real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) [28,30,31]. In addition, lung samples were stored in 4% paraformaldehyde solution (PFA), maintained at 4 ◦C, embedded in paraffin, cut, and stained with H&E as previously described [29]. The BCG count after 21 dpi was performed by seeding 1000 μL of lung homogenate in 7H10 plates and incubation for 14 to 21 days at 37 ◦C with 5% CO2. To evaluate the presence of HO-I in the lungs by flow cytometry, the lungs were incubated with collagenase IV for 30 min at 37 ◦C with agitation (120 rpm). Then the cells were homogenized using a 70 μm cell strainer. The cells were incubated with ammonium-chloride-potassium (ACK) lysis buffer for 5 min and centrifugated 300 g for 5 min at 4 ◦C, and then stained with α-CD45-BV510 (clone 30-FL1, BD Horizon, San José, CA, USA), α-CD11c APC (clone HL3, BD Pharmingen), α-IA/IE-V500 (clone M5/114.15.2, BD Pharmigen), α- CD326(Ep-CAM) PE (clone G8.8, Biolegend). Then, for HO-1 intracellular staining, fixed cells were incubated with anti-mouse HO-1 monoclonal antibody (mAb) (Abcam, UK) in permeabilization buffer (1% saponin, 10% FBS in PBS) for 45 min

at 4 ◦C. For all antibody dilutions, 1 μL of antibody was diluted in 500 μL of a buffer PEB (1X Phosphate Buffered Saline (PBS)-0.5% Bovine Serum Albumin (BSA)-0.2mM Ethylenediaminetetraacetic acid (EDTA)). Data were acquired in an LSRFortessa X2-0 cytometer (BD Biosciences) and analyzed using FlowJo v10.0.7 software (BD Biosciences). The gating strategy for detecting immune cells is shown in Supplementary Figure S1.

#### *2.7. Lung Histopathology Analyses*

Before proceeding with BAL collection, the major bronchus of the left lung was clamped using 10 cm Kelly hemostatic forceps to perform histopathology analyses without significantly altering tissue architecture. After obtaining the BAL from the right lung, the left lung was fixed with 4% paraformaldehyde, then paraffin-embedded using a Leica ASP300S enclosed, automatic tissue processor (Leica Microsystems, Wetzlar, Germany). Then, 5 μm-thick tissue sections were obtained using a Microm HM 325 Rotary Microtome (Thermo Scientific, Waltham, MA, USA) before being mounted and stained for histopathology analyses using H&E stain. A histopathological score was used to measure structural alterations in lung sections of control and infected animals [32,33]. The histopathological score was as follows: 0 = normal tissue morphology, normal alveolar architecture, connective tissue associated with bronchi (slight presence of immune cells); 1 = alveolar spaces are reduced, and there are immune cells present; 2 = bronchoalveolar involvement is defined as reduced alveolar spaces and high infiltration of immune cells (neutrophils and lymphocytes) within and surrounding the airways, including bronchi; 3 = pulmonary consolidation is evidenced by loss of alveolar spaces, bronchial walls thickening or bronchial collapse, and high cellular infiltration. In addition, Ziehl-Neelsen (ZN) staining was performed according to standard protocols [34,35]. In brief, the major bronchus of the left lung section was dewaxed by washing with decreasing alcohol concentrations (from 96% to 70% ethanol), heat-fixed, then stained with carbol-fuchsin (Bacto TB Carbolfuchsin KF, Becton-Dickinson, Sparks, MD, USA) for 4 min, washed, and incubated with 3% hydrochloric acid (HCl) until the stain was completely dissolved. Counterstaining was performed with brilliant green (Bacto TB Brilliant Green K, Becton- Dickinson, Sparks, MD, USA) for 20 s. Sections were air-dried after thorough washing.

#### *2.8. Relative Expression by RT-qPCR*

Quantitative real-time RT-qPCR total RNA was isolated from lung tissues collected using the Trizol reagent according to the manufacturer's instructions (Thermo Fisher Scientific). Complementary DNA (cDNA) synthesis from total RNAs was performed using the iScriptTM Reverse Transcription Supermix for RT-qPCR (Bio-Rad, Hercules, CA, USA) and random primers. RT-qPCR reactions were carried out using a StepOne plus thermocycler (Applied Biosystems, Waltham, MA, USA). The abundance of *ho-1* and *nrf2* mRNAs were determined by relative expression to the respective housekeeping gene (*β-actin* gene) by the 2-ΔΔ threshold cycle (ΔΔCt) method [36]. The RT-qPCR assays performed had 100% efficacy. For *n-hRSV* gene expression, absolute quantification data were expressed as the number of hRSV N-gene copies for every 5 × 103 copies of the *<sup>β</sup>-actin* transcript, as previously described [28,37]. The choice of using the *β-actin* gene for a single reference gene is based on present high stability [31,38]. The primers used can be found in Table 2.

#### *2.9. Statistical Analyses*

All statistical analyses were performed using GraphPad Prism version 6.0 Software. Statistical significance values and analyses are detailed in each figure legend. For Figure 1, only a normal distribution was observed for the data in Figure 1B. The data from Figure 1C–F showed a non-normal distribution, so non-parametric Mann-Whitney tests were performed. One-way ANOVA tests with a post hoc Tukey test were performed in Figures 2–5 and Supplementary Figures S2–S4 since the data showed a normal distribution. Two-way Anova with a post hoc Dunnett's multiple comparisons test were performed to compare

the kinetics of the weight curves of Figure 2A and Supplementary Figure S2A. Figure 1A used a two-way ANOVA with a post hoc Šídák's multiple comparisons test.

**Table 2.** List of primers used for real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis.


**Figure 1.** Evaluation of infection, inflammation, and immunomodulatory parameters from human respiratory syncytial virus (hRSV)-infected mice. (**A**) Body mass loss of C57BL/6 mice infected with <sup>1</sup> <sup>×</sup> <sup>10</sup><sup>7</sup> plaque formation units (PFU) of hRSV A2 for ten days. All the following parameters were measured at day 10 post hRSV infection. (**B**) Determination of viral load through specific real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) for hRSV. (**C**) Neutrophils (**D**) Heme Oxygenase (HO)-1, (**E**) OX-2 glycoprotein membrane (CD200), and (**F**) nuclear factor erythroid 2-related factor (NRF2). Data are shown as median ± interquartile range of at least two independent experiments with three animals per group. (**B**) One-way ANOVA was performed with a post-hoc Tukey test. (**C**–**F**) t-student was performed with the Mann-Whitney U test (\* *p* < 0.05; \*\*\* *p* ≤ 0.001). Created with BioRender.com.

**Figure 2.** Evaluation of infection parameters from primary infection with a human respiratory syncytial virus (hRSV) and 11 days post-infection with Bacillus Calmette-Guerin (BCG). (**A**) Scheme of infection in mice. (**B**) Body mass loss of C57BL/6 mice infected with 1 <sup>×</sup> 107 plaque formation units (PFU) of hRSV A2 and a subsequent challenge with BCG intranasal at day 10 post-infection with hRSV. The significant difference corresponds to a two-way ANOVA of multiple comparisons between *hRSV-BCG* versus *mock-vehicle* and hRSV-BCG versus hRSV-vehicle. (**C**) Determination of viral load through specific real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) for hRSV. (**D**) Lung tissue sections were stained with Hematoxylin and Eosin (10× magnification). (**E**) Histopathological score. (**F**) Bacterial load in the BCG-groups. (**G**–**K**) Flow cytometry analyses of bronchoalveolar lavages (BAL) from mice infected with *Mycobacterium bovis (M. bovis)*. The figure shows the absolute cell numbers for neutrophils (**G**), monocytes (**H**), eosinophils (**I**), alveolar macrophages (**J**), and interstitial macrophages (**K**) in BAL of *M. bovis*-infected mice. Data are shown as means ± SEM of three independent experiments with 3–4 animals per group. One-way ANOVA was performed with a post-hoc Tukey test. (\* *p* <0.05; \*\* *p* < 0.01; \*\*\* *p* ≤ 0.001; \*\*\*\* *p* ≤ 0.0001).

**Figure 3.** Determination of mycobacteria and Heme Oxygenase (HO)-1 activity during a short infection scheme with the human respiratory syncytial virus (hRSV) and Bacillus Calmette-Guerin (BCG). (**A**) Acid-fast staining of *Mycobacterium bovis*-infected mouse lungs collected on day 11 post-inoculation with BCG (100× magnification). Quantification by real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) of *ho-1* (**B**) and the nuclear factor E2-related factor 2 (*nrf2*) (**C**). Data are shown as means ± SEM of three independent experiments with 3–4 animals per group. One-way ANOVA was performed with a post-hoc Tukey test (\* *p* < 0.05).

### **3. Results**

#### *3.1. HRSV Infection Induces de Expression of Immunomodulatory Molecules*

Our study model evaluated different disease and cellular infiltration parameters to characterize the inflammatory immune response induced by hRSV. A significant reduction in viral loads was observed during previous studies using a mouse infection model by day 7 post hRSV instillation [39]. Therefore, we measured weight loss and viral load by quantitative RT-PCR from the lungs of mice at 10 dpi and used as positive controls (C+) the viral load at 3 dpi (Figure 1A,B). No significant differences were found in the weight loss after the infection compared to mock-treated mice (Figure 1A). As shown in Figure 1B, animals showed levels of *n-hRSV* RNA close to non-detectable, indicating an adequate viral clearance by day 10 post-challenge, making it an appropriate inoculation time point for subsequent BCG challenge. As expected, only a significant difference was found in the positive control with respect to the groups evaluated (*p* = 0.0001). These results correlated with a low percentage of neutrophils recruited in the lungs (Figure 1C). Additionally, to evaluate whether the resolution of the infection was affected by anti-inflammatory modulators, the expression of *ho-1*, nrf2 related to oxidative stress, and *cd200* genes were measured by RT-qPCR. At 10 dpi, we observed a significant increase in *ho-1* (*p*= 0.0152) and *cd200* (*p* = 0.0244) relative expression in convalescent mice lungs compared to control animals. However, no differences were observed in the transcription factor controlling *ho-1* gene expression *nrf2* between the groups (Figure 1F). We have recently shown that HO-1 upregulation has an antiviral protective effect in the lungs by limiting viral replication in the tissue and epithelial cells, as well as by modulating the immunogenicity of antigenpresenting cells (APC), such as dendritic cells (DC) and alveolar macrophages [20,40]. Given these data, it could be suggested that upon viral clearance, the high levels of *ho-1* expression in the lower respiratory tract might play a role in modulating the immunity to secondary infections.

#### *3.2. Previous hRSV Infection Causes Further Long-Term Susceptibility to Mycobacterium Bovis-Driven Pneumonia*

To further evaluate whether animals that have cleared hRSV from the lungs remain immunocompetent to clear secondary infections, control (mock-instilled) and hRSVconvalescent animals were challenged with *M. bovis* BCG after 10 dpi. BCG inoculation was given intranasally to mice using a saline solution as a control (Figure 2A). As shown in Figure 2B, hRSV-convalescent mice were inoculated with BCG (*hRSV-BCG*), but not their relevant controls (*Mock-BCG*), showed increased weight loss at day 13 post-infection. Statistical differences were found along the kinetics for *hRSV-BCG* vs. *Mock-Vehicle* or *Mock-BCG* groups (*p* ≤ 0.0001). This weight loss can be associated with the infection with BCG rather than hRSV persistence, as evidenced by the absence of detectable viral loads at day 21 post-infection (Figure 2C). As expected, only a significant difference was found in the positive control with respect to the groups evaluated (*p* = 0.0002). Additionally, this was associated with more severe histopathological scores in both groups infected with BCG. Careful histopathological scoring did not show any significant difference in the semiquantitative blinded scoring of both BCG-infected groups despite a more marked thickening of alveolar walls (Figure 2D,E). The more severe lung pathology observed at 21 dpi in the *hRSV-BCG* group was characterized by the thickening of the lung parenchyma due to interstitial inflammation in discrete and well-defined foci developed around bronchi and alveolar sacs (Figure 2E), which showed a significant difference compared to *Mock-Vehicle* (*p* = 0.00219). Although less severe, a mild inflammation of the lung parenchyma was observed with foci in the bronchi in mice pre-treated with Mock and then inoculated with BCG (*Mock-BCG*) (Figure 2D,E). Additionally, the maintenance of the pulmonary architecture and no significant lung inflammation were observed in both vehicle-inoculated controls (*Mock-Vehicle* and *hRSV-Vehicle* groups). Mycobacteria CFUs were detected in the lungs of animals previously inoculated with either mock or hRSV (Figure 2F). Consistent with the histopathological score, only the *hRSV-BCG* group showed a significant

increase of BAL neutrophils (Ly-6C<sup>−</sup> CD11b<sup>+</sup> Ly-6Ghi) compared to the *Mock-Vehicle* group (*p* = 0.0445) (Figure 2G). Accompanying neutrophils, we observed no significant differences in BAL monocytes (Ly-6C+ CD11b<sup>+</sup> Ly-6G−) and eosinophils (CD11b+ Siglec F+) in all groups of mice (Figure 2H,I, respectively). Next, we analyzed other lung immune cell populations by flow cytometry, such as interstitial and alveolar macrophages. Consistent with the infiltration of neutrophils in BAL, alveolar macrophages, defined as CD11b<sup>−</sup> CD11c+ Siglec-F+ [41], showed a significant increase only in *hRSV-BCG* compared to both *Mock-Vehicle* (*p* = 0.0062) and *Mock-BCG* (*p* = 0.0159) groups (Figure 2J). A slight, non-significant increase was observed in the *hRSV-Vehicle* group (Figure 2J). No significant increase in interstitial macrophages (CD11b+CD11c+Siglec-F−) was detected between groups, but a slight increase was found in the *hRSV-BCG* group. (Figure 2K).

To assess whether the increased susceptibility to *M. bovis* BCG induced by a previous hRSV infection is short-lived, we evaluated a long scheme in which the second inoculation and the euthanasia were performed 21 and 42 days after the first inoculation, respectively (Table 1). Interestingly, the results obtained from these trials showed that there were no significant changes between those reflected in the parameters of disease or inflammation evaluated in the different groups of animals (Supplementary Figure S2), except for more significant damage related to the histological score (Supplementary Figure S2E).

Following our detection of viable bacilli in the lungs of BCG-infected animals, we performed ZN staining of the lung sections shown in Figure 3A to assess whether the significant increase in pathology and infiltration of neutrophils correlated with the increased presence of dormant bacilli, which might not be detected by ex vivo culture of lung homogenates. In agreement with the low numbers of CFU detected in BCG-infected animals, no active *M. bovis* was found in the lungs of both BCG-infected groups of mice (Figure 3A). However, a population of macrophages that reacted with the stain was observed only in the lungs of *hRSV-BCG* treated mice but not in the other treatments (red arrows). Since ZN staining binds to mycolic acids, the observation of weak ZN staining suggests the existence of either myeloid cells processing remainder mycobacterial components, or macrophages containing bacilli with a low metabolic rate (i.e., under dormancy) [42].

Since HO-1 induction has been associated with both impairment of antigen processing by APC [40] and bacilli dormancy [23], we sought to evaluate whether the detection of ZN-stained cells was associated with the modulation of HO-1 expression in the lungs. A significant increase of almost 50% in the relative expression of HO-1 was observed only in *hRSV-BCG*, as compared to *mock-vehicle* (*p* = 0.0357), *hRSV-vehicle* (*p* = 0.0263) or *mock-BCG* (*p* = 0.0239) control groups (Figure 3B). No significant differences were observed in the expression of Nrf2, suggesting that HO-1 up-regulation might imply other transcription factors [43] (Figure 3C).

No significant changes were observed in the determination of dominant BCG in the lungs of the long scheme animals, nor were differences in the expression levels of HO-I or NRF2 (Supplementary Figure S3).

#### *3.3. Characterization of the Inflammatory Landscape of the Lungs*

As mentioned above, CD200 can induce an inhibitory signal in alveolar macrophages, among other APC, downregulating the inflammation in the pulmonary tissue [25]. No significant differences in the expression of the epithelial anti-inflammatory molecule CD200 were observed between the groups of the short scheme (Figure 4A), suggesting that the absence of CD200 upregulation might support the observed lung pathology [43]. Additionally, the relative expression of pro-inflammatory cytokines IFN-γ, IFN-β, and IL-6 was evaluated, but no significant differences were found for any of these molecules (Figure 4B–D). Only a slight increase in the relative expression of IFN-γ was detected in the *hRSV-BCG* compared to the other groups (Figure 4B). For the long scheme, an increase in the relative expression was observed for all molecules evaluated. Still, only significant differences were found for the expression of *ifn-γ* for the *hRSV-BCG* group compared to *mock-vehicle* (*p* = 0.0094) (Supplementary Figure S4B).

**Figure 4.** Determination of the relative expression of immunomodulatory molecules and cytokines in infection with Bacillus Calmette-Guerin (BCG) at 10 days post-human respiratory syncytial virus (hRSV)-infection. Quantification by real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) of OX-2 glycoprotein membrane (*cd200*) at day 10 post-infection (**A**), interferon-gamma (*ifn-γ*) (**B**), interferon beta (*ifn-β*) (**C**), and interleukin (IL)-6 (**D**). Data are shown as means ± SEM of three independent experiments with 3–4 animals per group. One-way ANOVA was performed with a post-hoc Tukey test.
