Animal care and procedures were performed by following the guidelines of good experimental practices according to Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes [amended by the Regulation (EU) 2019/1010)] and Spanish laws (RD 53/2013). The protocol was approved by the Community of Madrid Ethics Committee (reference PROEX 180.2/22). K18-hACE2 mice strain, expressing angiotensin-converting enzyme 2 (hACE2), necessary for SARS-CoV-2 to enter the host cells, was used in order to evaluate the effectiveness of dpB in combination with specific SARS-CoV-2 vaccine against SARS-CoV-2 infection. K18-hAC2 female mice (n=108) aged 4–6 weeks were obtained from Charles River Laboratories España S.A (Sant Cugat del Vallès, Barcelona, Spain). The cages were equipped with Altromin-LASQCdiet^® Rod 14-H (Altromin Lage, Germany) for feeding. Both food and water were available ad libitum. Wheeled houses for environmental enrichment were also included (Ref: K3327 + K3250, Sodispan Research, Madrid, Spain). During the initial 2-week period, they were given time to acclimate to their new cages and socialize with their partners. After that period, the animals were marked with an ear tagger for individual identification (in many cases, it was not necessary since mice were easily identifiable by white areas on their tails). DpB and vaccine administration procedures were carried out in the Biosafety Level 2 + (BSL2+) area at the VISAVET Health Surveillance Centre (Complutense University, Madrid). Afterwards, animals were moved to the Biosafety Level 3 (BSL3) area at the VISAVET Centre for SARS-CoV-2 infection studies.

SARS-CoV-2 MAD6 was used for experimental infection assay. Calu-3 cells were prepared to reproduce stocks of SARS-CoV-2 (29). The cells were incubated at 37 °C under 5% CO₂ in Eagle’s Minimum Essential Medium (EMEM) with L-glutamine (Merck KGaA, Darmstadt, Germany) and supplemented with 100 IU/mL penicillin, 100 μg/mL streptomycin, and 10% fetal bovine serum (FBS) (Merck KGaA, Darmstadt, Germany).

For viral growth in Calu-3 cells, a multiplicity of infection (MOI) of 0.0001 was used. After the virus was absorbed for one hour at 37°C, the viral growth media EMEM supplemented with 2% fetal bovine serum was added. The cell lysate and supernatant were harvested after 3 days of incubation at 37 °C with 5% CO₂.

Vero E6 cells, provided by the Carlos III Health Institute (Madrid, Spain), or ATCC^® (Manassas, Virginia, USA), were prepared to titrate SARS-CoV-2 stocks by determining the amount of virus causing cytopathic effects in 50% of tissue culture infectious dose (TCID50/mL). Additionally, this cell line was used to verify the viability of the SARS-CoV-2 inoculum used for infecting the animals after the procedure.

A frozen vial of dpB was inoculated in a starter liquid culture medium and incubated at 37 °C in aerobiosis for 4 weeks and then inoculated in a liquid culture medium at 37 °C in aerobiosis for another 4 weeks. At the end of the culture, the growth was collected with a pipette, centrifuged, washed with PBS and given a treatment of physical breakage with glass beads, and diluted with PBS until a homogeneous product was at approximately 1 McFarland [equivalent to approximately 10⁶ colony forming units (CFU)/mL]. The final dose for inoculation obtained was 6.3x10⁵ CFU/mL.

For dpB inactivation, formaldehyde (37%) is added until achieving a final concentration of 0.1%. The growth is immersed in a bath of 75 °C for 70 minutes and afterwards placed on a sonicator (Misonix Sonicator XL2020 Ultrasonic Liquid Processor), conducting during 10 cycles that included 3 minutes of sonication and 1 minute of pause. The dpB concentration of the final solution was then adjusted to 1.2 McF (10⁶ CFU/mL) with PBS.

Vaccine employed in the first administration was combined with a complete Freund adjuvant (inactivated mycobacterium). Vaccine employed in the second administration (boost) was combined with incomplete Freund adjuvant (30).

Animals were divided into 6 different groups ( Figure 1 ) each of them with a different treatment: 1) Inactivated dpB, 2) Inactivated dpB - SARS-CoV-2 vaccine, 3) dpB, 4) dpB - SARS-CoV-2 vaccine, 5) Infection control, and 6) SARS-CoV-2 vaccine. Inoculation of inactivated dpB and live dpB took place on day 0, followed by SARS-CoV-2 vaccine on day 35 and its boost on day 58 of the experiment. All the animals were then challenged with SARS-CoV-2 on day 76 or 90. All the groups were divided into two subgroups following-up duration: i) animals sacrificed 3–4 days post infection (dpi) and ii) animals that developed the disease. Animals were sacrificed using cervical dislocation when reaching the endpoint criteria described in the following section (5, 6, 7, 8 dpi), or animals left until the 9 dpi (end of the experiment) that did not develop COVID-19 or did not reach endpoint criteria.

DpB both alive (6.3 x 105 CFU/mL) and inactivated (1.2 McF) were intravenously (IV) administered (100 µl per mice) 2 weeks after the arrival of the mice, using a 30 G needle through the caudal vein of the tail (groups 1, 2, 3 and 4). A restrainer was used to immobilize the mice and the tail was pre-warmed slightly using a heat lamp for enhanced vasodilation for dpB administration.

SARS-CoV-2 vaccine was administered intraperitoneally (IP) (200 µL) in two different times. First inoculation was carried out 35 days after dpB-stimulation and the vaccine boost was administered 22 days after the first inoculation.

A blood sampling in 8–9 animals from each group was carried out 2 months after dpB stimulation and before SARS-CoV-2 infection. A total blood volume of 200-250 µL was obtained throughout venipuncture with a 23 G needle. Gauze was used for compression to stop the bleeding and once the bleeding stopped, IP physiological serum was administered to replace fluids. Subsequently, plasma was obtained for the detection of antibodies against SARS-CoV-2, inflammation and coagulation markers and the determination of cytokines ( Figure 1 ).

Animals were challenged with SARS-CoV-2 virus 76 to 90 days after the dpB administration ( Figure 1 ). For infection, the 101 mice (groups 1-6) were sedated with xylazine (20 mg/mL) [Xilagesic 20MG 25ML, CALIER, Les Franqueses del Vallès, Barcelona] at a dose of 2 mg/kg and ketamine 100 mg/mL [Ketamidor 100 mg/ml, 50 ml, richter pharma, Austria] at a dose of 20 mg/kg intraperitoneally (IP). Afterwards, animals were inoculated intranasally (NAS) with SARS-CoV-2 at a dosage of 5x104 PFUs or 3.5x104 TCID50 per mouse. The total volume inoculated was 25 μL alternating nostrils in volumetric fractions of 5 μL ( Figure 1 ).

To confirm that animals were successfully infected, oropharyngeal swab samples were collected at 2 dpi [DeltaSwab^® Virus with viral transport media (MTV) (Deltalab S.L., Cataluña, Spain)].

Following the challenge, animals were weighed and monitored daily for clinical signs. A clinical scoring table ( Table 1 ) was prepared and utilized to document each clinical sign on a scale from 0 to 2 by two different operators, so the subjectivity was lowed to the minimum. The animals were sacrificed upon reaching a cumulative clinical score (sum of scores for each evaluated clinical sign) of 5 or a loss of weight higher than 20%. Once the designated sacrifice days were reached (3 and 4 dpi) or endpoint criteria were fulfilled, blood samples in heparin were obtained via cardiopuncture from the animals after sedation IP with xylazine (20 mg/mL) at a dose of 4 mg/kg and ketamine (100 mg/mL) at a dose of 40 mg/kg, before sacrifice. The animals that did not meet the endpoint criteria were euthanized on 9 dpi, marking the end of the experiment. Subsequently, comprehensive necropsy was conducted on the animals. In this study, brain, upper respiratory tract (trachea and nasal turbinates), lower respiratory tract (lungs) and heart samples were collected in AllProtect Tissue Reagent (Qiagen, Venlo, Netherlands) for viral RNA detection and quantification. The former tissues were also collected in 10% neutral formalin for further histopathological studies.

MagMAX™ CORE Nucleic Acid Purification Kit was used for RNA extraction of tissues according to the manufacturer’s instructions with the KingFisher Flex (Thermofisher) automatized system. In addition, the detection and quantification of SARS-CoV-2 loads from tissues and oropharyngeal swabs was performed using the CoVID19 dtec RT qPCR Test (Genetic PCR Solutions™, Alicante, Spain) including the respective positive and negative controls.

SARS-CoV-2 neutralizing antibodies were detected in all samples using the GenScript cPass™ SARS-CoV-2 Neutralizing Antibody Detection Kit (GenScript Inc., Piscataway, NJ, United States). In this procedure, the protein-protein interaction between HRP-RBD (horseradish peroxidase-receptor binding domain) and human angiotensin-converting enzyme II (hACE2) can be blocked by neutralizing antibodies against SARS-CoV-2 RBD, which was previously validated (31). Samples presenting a cutoff higher or equal to 30% were considered positive, indicating the presence of SARS-CoV-2 neutralizing antibodies, and samples with a cut-off lower than 30% were considered negative according to the manufacturer’s instructions.

Complete heparinized blood was centrifugated and plasma was stored at -80 °C until the analyses of the following biomarkers: CRP (C-reactive protein) (Mouse CRP ELISA Kit. Invitrogen, Massachusetts, USA), D-dimer [Mouse D2D (D-Dimer) ELISA Kit. FineTest^®, Boulder, USA] and iNOS (nitric oxide synthase) (iNOS ELISA Kit. MyBioSource, San Diego, USA).

Proinflammatory cytokines as interleukin 1β (IL-1β), interleukin 6 (IL-6), tumor necrosis factor alpha (TNF-α), interferon gamma (IFN-γ), and anti-inflammatory as interleukin 10 (IL-10) and transforming growth factor beta (TGF-β) were measured in plasma with the ProcartaPlex™ System (Invitrogen, Waltham, MA, USA) and Bio-Plex 200 equipment (Bio-Rad Laboratories, Alcobendas, Madrid, Spain). These measurements were made both before SARS-CoV-2 infection in 8–9 animals per group 2 months after BCG stimulation and in all the animals post-SARS-CoV-2 infection in the moment of the sacrifice of each animal.

Tissues were fixed in 10% neutral formalin (Panreac AppliChem ITW Reagents, Barcelona, Spain) for 48 hours. The samples were automatically processed (Citadel 2000 Tissue Processor, Thermo Fisher Scientific, Waltham, MA, USA) and embedded in paraffin (HistoStar Embedding Workstation, Thermo Fisher Scientific, Waltham, MA, USA). Serial sections of 4 µm thickness were obtained using a microtome (FinesseMe+, Thermo Fisher Scientific, Waltham, MA, USA), and stained with hematoxylin-eosin (HE) (Gemini AS Automated Slide Stainer, Thermo Fisher Scientific, Waltham, MA). Finally, samples were mounted (CTM6 Coverslipper, Thermo Fisher Scientific) and evaluated for histopathological alterations under a Leica DM2000 microscope (Leica Microsystems, Wetzlar, 162 Germany). A single-blind study was conducted by the pathologist in which different lesions of brain and lungs (indicated in Supplementary Table S1 ) were evaluated using a histological score, which assigned a value between 0–3 based on severity.

Initially, normality tests (Shapiro-Wilk) were conducted, revealing that the distribution was not normal for most variables. Differences between groups for each of the variables were evaluated using Kruskal-Wallis and Mann-Whitney test in SPSS (version 28.0.1.1). Mann-Whitney p-values were adjusted by Borferroni method (p_b). The results were considered statistically significant when the p_b-value was less than 0.05. Furthermore, a correlation analysis of the different variables was carried out using Spearman’s test.

All graphical representations included in this study were created using R (version [4.4.2]) within the RStudio integrated development environment (IDE) (32). The ggplot2 package (33) was used to generate scatter plots, boxplots, and jittered point overlays, while the heatmap package (34) was employed for heatmap visualizations ( Figure 2 ). Custom color palettes and additional annotations were applied to ensure clarity and high-quality presentation of the data.

Mortality was 0% in all groups vaccinated against SARS-CoV-2 (groups 2, 4, 6), regardless of prior dpB stimulation, with only minimal clinical signs observed, such as slight weight loss or tousled hair ( Figure 2 ). This underscores the efficacy of the SARS-CoV-2 vaccine employed in this study in preventing severe disease, even without additional immunomodulation. Thus, those animals that had been vaccinated were positive for neutralizing antibodies in the first sampling before infection (31 days after first dose of SARS-CoV-2 vaccine), as well as in post-infection sampling. However, the animals with no SARS-CoV-2 vaccine, started to develop antibodies after 6–7 dpi (data not shown).

In pre-infection samples (day 66), animals receiving the combination of live dpB and SARS-CoV-2 vaccination (group 4) showed statistically higher levels of IL-6 and IL-10, than inactivated dpB-SARS-CoV-2 vaccine group (group 2) (M-W, p_b< 0.001; M-W, pb=0.040) ( Figures 3A, E ). Also, group 4 exhibited higher levels of IFN-γ cytokine compared to group 2 (M-W, p_b<0.001) and to SARS-CoV-2 vaccinated-only animals (group 6) (M-W, p_b=0.003) ( Figure 3D ). Similarly, for almost all cytokine levels (IL-6, TNF-α, IL-1β and TGF-β) significant differences with higher values were observed between animals with SARS-CoV-2 vaccine (groups 2, 4, 6) and the respective non-vaccinated groups (groups 1, 3 and 5) (M-W, p<0.05). Significant differences were observed for animals with inactivated dpB and SARS-CoV-2 vaccine versus inactivated dpB, live dpB and SARS-CoV-2 vaccinated group versus only live dpB group and SARS-CoV-2 vaccine group versus the infection control group (M-W, p<0.05).

For assessing the cytokine response, negative correlations were observed (Spearman test) between pre-infection levels of pro-inflammatory cytokines (IL-1β, TNF-α, IL-6) and viral loads in the moment of sacrifice in multiple organs, including the brain, lungs, and heart ( Table 2 ). Interestingly, a similar negative correlation was found for anti-inflammatory cytokine IL-10 and viral loads in the lungs, and for IFN-γ in both lungs and heart, reinforcing the protective role of these cytokines in viral clearance ( Table 2 ). Conversely, pre-infection TGF-β levels positively correlated with higher viral loads in the brain, lungs, and heart, potentially suggesting a suppressive effect on immune activation that limits effective viral control ( Table 2 ).

The levels for each of the cytokines at 3–4 dpi obtained were similar to the concentration obtained in the pre-infection sampling. Those receiving the live form of dpB with the SARS-CoV-2 vaccine (group 4) showed elevated IFN-γ and IL-6 levels when comparing to inactivated dpB - SARS-CoV-2 vaccine (group 2) (M-W; p_b<0.001, p_b=0.033) and only-SARS-CoV-2 vaccinated animals (group 6) (M-W; p_b<0.001, p_b=0.014) ( Figures 4A, D ). For IL-1β and IL-10, animals receiving both live dpB and SARS-CoV-2 vaccine showed significantly higher levels than SARS-CoV-2 vaccinated-only animals (M-W, p_b = 0.017; M-W, p_b=0.007) ( Figures 4C, E ) reflecting an enhanced cytokine response likely associated with the synergistic effects of trained and adaptive immunity. In this case, for TNF-α cytokine, the highest levels were obtained for both groups combining immunomodulator and SARS-CoV-2 vaccine (groups 2 and 4) ( Figure 4B ).

Regarding to inflammatory markers, plasma samples collected on 3–4 dpi showed that animals from the group receiving both live dpB and SARS-CoV-2 vaccine showed higher values (mg/mL) of D-dimer, CRP and iNOS when compared to infection SARS-CoV-2 vaccinated group (M-W; p_b=0.042, p_b=0.021, p_b=0.003) ( Figures 5A–C ). For animals developing the disease no significant differences were obtained comparing these experimental groups.

In brain tissue, although no significant differences were obtained (K-W; p>0.05), the lowest viral loads were found in the live dpB + SARS-CoV-2 vaccine group, followed by the inactivated dpB - SARS-CoV-2 vaccine group, and finally the vaccinated animals ( Figure 6A ). In pulmonary and cardiac tissues, all vaccinated groups exhibited undetectable viral loads ( Figures 6B, C ). With respect to the upper respiratory tract including turbinates and trachea, a similar trend was observed as in brains, but with comparatively lower viral loads, i.e., vaccinated animals had the lowest viral loads although non-statistical differences were observed (K-W, p>0.05) ( Figure 6D ).

In line with viral loads results in tissues, further histopathological examination highlighted the higher protective effects of combining dpB with the SARS-CoV-2 vaccine. In the brain, acute lesions characteristic of SARS-CoV-2 (primarily consisted of moderate to severe non-suppurative meningoencephalitis and neuronal degeneration, characterized by cytoplasmic vacuolization with a pyknotic and eccentric nucleus) were observed. Additionally, moderate white matter tract myelin sheath vacuolation, increased glial cell proliferation, vasculitis, and perivascular hemorrhages were frequently observed.

DpB-SARS-CoV-2 vaccine (group 4) [median histopathological score (MHS)=2.0] and inactivated dpB-SARS-CoV-2 vaccine group (group 2) (MHS=1.5) showed an absence of brain lesions characteristic of SARS-CoV-2 infection at 7–9 dpi ( Figure 7A ). In comparison, SARS-CoV-2 vaccinated group (group 6) also did not present typical lesions (MHS=3.75), except for one case (HS=9.5) ( Figure 7A ), which had moderate perivascular lymphocytic cuffs and a moderate increase of shrunken neurons and glial cells, at 9 dpi. The total score in group 6 was significantly higher than groups 4 and 2 (M-W, p= 0.017, p=0.009) ( Figures 8A–C ).

In the lungs, acute lesions characteristic of SARS-CoV-2 included bronchointerstitial pneumonia, characterized by infiltration of mononuclear cells (mainly macrophages plasma cells and lymphocytes) and occasional neutrophils around bronchioles and blood vessels, mild to moderate patchy septal thickening; type II pneumocyte hyperplasia, perivascular lymphocytic cuffs, microhemorrhages and alveolar edema.

Similarly, reduced severity of lung lesions at 9 dpi was observed in groups that combine the vaccine with dpB. Live dpB +SARS-CoV-2 vaccine (group 4) (MHS=3.0) and inactivated dpB-SARS-CoV-2 vaccine (group 2) (MHS=3.5) animals characterized by minimal interstitial pneumonia and thickening of the alveolar septa; whereas SARS-CoV-2 vaccine group (group 6) (MHS=6.25) showed the same lesions, but slightly more pronounced mainly when compared to groups 4 and 2 (M-W, p_b=0.004, p_b=0.015) ( Figures 7B , 9A–C ). Group 4 was notable for presenting hyperplasia of the bronchus and bronchioles-associated lymphoid tissue (BALT). Additionally, it exhibited round lymphoid formations, characterized by the aggregation of lymphoplasmacytic cells and macrophages, associated with small blood vessels and located in the alveolar interstitium, not adjacent to the bronchi or bronchioles [compatible with inducible BALT (iBALT) or tertiary lymphoid organs (TLO)] ( Figure 9A ).