C57BL/6j mice were bred under specific-pathogen-free (SPF) conditions at the Research Center Borstel. IL-10 reporter mice (Vert-X) which express enhanced green fluorescent protein (eGFP) under the control of the IL-10 promoter (22) were bred under SPF conditions at the animal facility of the University of Lübeck. IL-10 flox/flox mice (B6.129P2(C)-Il10tm1Roer/MbogJ), with loxP sites that flank exon 1 of IL-10, were bred under SPF conditions with CD19-Cre mice (B6.129P2(C)-Cd19tm1(cre)Cgn/J) at the animal facility of the University of Lübeck, resulting in the generation of mice with B cell-specific inactivation of IL-10 (B cell IL-10 KO), and IL-10 competent littermate controls (23). For infection experiments, male and female mice aged between 11-20 weeks were maintained under barrier conditions in the BSL 3 facility at the Research Center Borstel in individually ventilated cages. Animal experiments were in accordance with the German Animal Protection Law and approved by the Ethics Committee for Animal Experiments of the Ministry of Agriculture, Rural Areas, European Affairs and Consumer Protection of the State of Schleswig-Holstein.

Mtb HN878 was grown in Middlebrook 7H9 broth (BD Biosciences, Franklin Lakes, USA) supplemented with 10% v/v OADC (Oleic acid, Albumin, Dextrose, Catalase) enrichment medium (BD Biosciences, Franklin Lakes, USA) to logarithmic growth phase (OD₆₀₀ 0.2 - 0.4) and aliquots were frozen at -80°C. Viable cell number in thawed aliquots were determined by plating serial dilutions onto Middlebrook 7H11 agar plates supplemented with 10% v/v OADC followed by incubation at 37°C for 3-4 weeks. For infection of experimental animals, Mtb stocks were thawed and diluted in sterile distilled water to a concentration providing an uptake of approximately 50 viable bacilli per lung. The bacterial suspension was used directly after preparation, with infection carried out via the respiratory route by using an aerosol chamber (Glas-Col, Terre-Haute, USA) as described previously (6). The inoculum size was quantified 24 h after infection in the lungs and bacterial loads in lung and spleen were evaluated at different time points after aerosol infection by determining colony forming units (CFU). Organs were removed aseptically, weighed, and homogenized in 0.05% v/v Tween 20 in PBS containing a proteinase inhibitor cocktail (Roche, Basel, Switzerland) prepared according to the manufacturer’s instructions for subsequent quantification of cytokines (see below). Tenfold serial dilutions of organ homogenates in sterile water/1% v/v Tween 80/1% w/v albumin were plated onto Middlebrook 7H11 agar plates supplemented with 10% v/v OADC and incubated at 37°C for 3–4 weeks.

Clinical score was used to indicate severity of disease progression (7). Animals were scored in terms of general behavior, activity, feeding habits, and weight gain or loss. Each of the criteria is assigned score points from 1 to 5 with 1 being the best and 5 the worst. The mean of the score points represents the overall score for an animal. Animals with severe symptoms (reaching a clinical score of ≥3.5) were euthanized to avoid unnecessary suffering, and the time point was recorded as the end point of survival for that individual mouse.

Mice were sacrificed and perfused intracardially with sterile PBS before organ harvest. Lungs were digested in 100 µg/ml DNase I (Roche, Basel, Switzerland) and 50 µg/ml Liberase TL (Roche, Basel, Switzerland) in RPMI for 60 minutes and passed through a 100 µm pore size cell strainer to obtain a single cell suspension. Single-cell suspensions were blocked with anti-CD16/CD32 (clone 93, Biolegend, Fell, Germany) and sera (rat, rabbit, mouse (PAN Biotech, Aidenbach, Germany) and hamster (abcam, Cambridge, UK)) for 30 min at 4°C. Subsequently, cells were washed and incubated for 30 min at 4°C with the following antibodies: BV510 anti-CD45 (clone 30-F11); BV421 or PerCP-Cy5.5 anti-CD138 (clone 281-2); APC anti-GL7 (clone GL7) and BV421 anti-CD8a (clone 53-6.7) from BioLegend, Fell, Germany and PE-Cy7 anti-B220 (clone RA3-6B2) and APC-H7 anti-CD4 (clone GK1.5) from BD Biosciences, Franklin Lakes, USA. After washing, cells were resuspended in FACS buffer (3% fetal calf serum heat-inactivated, 0.1% NaN₃, 2 mM EDTA) and analyzed on a MACSQuant Analyzer 10 (Miltenyi Biotec B.V. & Co. KG, Bergisch Gladbach, Germany). Imaging of IL-10 producing B cells were performed on a BD FACSDiscover™ S8 Cell Sorter with BD CellView^™ Image Technology and BD SpectralFX^™ Technology (BD Biosciences, Franklin Lakes, USA). Data were analyzed with the FCSExpress software (DeNovo™ Software, Pasadena, USA) and the Cytolution software (Cytolytics GmbH, Tübingen, Germany).

The concentrations of cytokines and chemokines in lung homogenates were determined by LEGENDplex™ (Mouse Inflammation panel (IL-23, IL-1α, IFN-γ, TNF-α, CCL2, IL-12p70, IL-1β, IL-10, IL-6, IL-27, IL-17A, IFN-β, GM-CSF) and Mouse Proinflammatory Chemokine Panel (CXCL1, CCL5, CCL20, CCL11, CCL17, CCL2, CXCL9, CXCL10, CCL3, CCL4, CXCL13, CXCL5, CCL22) respectively, BioLegend, Fell, Germany) according to the manufacturer’s protocol.

High-binding 96-well ELISA plates (SARSTEDT AG & Co. KG, Nümbrecht, Germany) were coated with 20 µg/ml HN878 whole-cell lysate (BEI resources, Manassas, United States) for 1 h at 37°C and blocked with 3% bovine serum albumin (BSA) in Tris-buffered saline (TBS) overnight at 4°C. After washing 3 times with TBS + 0.1% Tween 20 (TBST), lung homogenates (1:10 dilution in TBS/1% BSA) or TBS as control were added and incubated for 1 h at 37°C. After an additional washing step, alkaline phosphatase (AP)-conjugated goat anti-mouse IgG, IgM and IgA (SouthernBiotech, Birmingham, USA) were added at 1:1000 dilution in TBS/1% BSA and incubated for 1 h at 37°C. After washing four times with TBST, p-Nitrophenylphosphate was added. The reaction was incubated at RT for 20 minutes, then stopped with 1 N NaOH. Absorbance was measured at 405 nm.

Superior lung lobes from infected mice were fixed with 4% w/v paraformaldehyde for 24 h, embedded in paraffin, and sectioned (4 μm). Sections were stained with hematoxylin and eosin (H&E; Merck) or carbolfuchsin (Merck) followed by decolorization with acid-alcohol to visualize mycobacteria in the lungs. Slides were imaged with a PhenoImager™ HT instrument (Akoya Biosciences, Marlborough, USA). The quantitative analysis of inflammatory area was performed by determining the total lung area and the inflammatory area in H&E-stained sections using the software QuPath (https://qupath.github.io/) (24).

Formalin-fixed, paraffin-embedded (FFPE) samples of mouse lung tissues were deparaffinized by immersion into xylene (2x 10 min) and rehydrated by graded series of ethanol (2x 2 min 100%, 2 min 96%, 2 min 90%, 2 min 80%, 2 min 70%, 2 min distilled water). Rehydrated samples were transferred to 1x TBST (50 mM Tris, pH 7.6, 0.05% Tween-20) and subsequently forwarded to mIF staining (25). The Opal 3-Plex Anti-Rb Manual Detection Kit (NEL840001KT, Akoya Biosciences, Marlborough, USA) was used with additional fluorophores as indicated in Supplementary Table S1 . Overall, mIF staining consisted of several, subsequent cycles of IF staining targeting one antigen per time: Before antigen detection, endogenous peroxidases were blocked once by incubation for 10 min at RT in 3% H₂O₂ followed by washing with 1x TBST. One IF cycle consisted of 10 min protein blocking using antibody diluent (Akoya Biosciences, Marlborough, USA) followed by 3x 2 min washing with 1x TBST. All tissue underwent heat-induced epitope retrieval (HIER) with citric acid buffer pH6 (Akoya Biosciences, Marlborough, USA) prior to protein blocking and incubation with primary antibody. Primary antibodies were diluted using antibody diluent (Akoya Biosciences, Marlborough, USA) and incubated for 45 min at RT followed by 3x 3 min washing with 1x TBST. Primary antibodies were detected by incubation for 10 min with HRP-Polymer (Akoya Biosciences, Marlborough, USA) and subsequent TBST washing for 3x 3 min. Signals were visualized by OPAL-TSA reaction for 10 min at RT and stopped by washing 3x for 3 min with 1x TBST. HIER for 1 min at 1000 W and 10 min at 100 W was used to remove primary antibodies and HRP polymer before entering the next staining cycle. Finally, nuclei were stained using spectral DAPI (Akoya Biosciences, Marlborough, USA) and slides were mounted with cover slips using ProLong Gold (Invitrogen, Carlsbad, USA). Please see Supplementary Table S1 for a detailed pipetting protocol for the mIF panel. The panel was developed using positive control tissues and antigen stability was tested for each antibody by increasing rounds of AR using citric acid buffer pH6 to determine optimal positioning within each mIF panel. Single-plex stains for each target of the multiplex panel from the respective cycle were conducted and their results qualitatively compared to multiplex results for concordance of overall staining pattern, intensity and frequency of detected cells.

Whole-slide imaging (WSI) was performed using a PhenoImager HT multi-spectral slide scanner (Akoya Biosciences, Marlborough, USA) at 20x magnification (0.5 µm/pixel) with exposure times held constant for all analyzed samples (DAPI: 1.39ms, Opal 480: 10 ms, Opal 520: 20 ms, Opal 570: 55 ms, Opal 780: 15 ms). Phenochart software 1.1.0 (Akoya Biosciences, Marlborough, USA) was used to select regions of interest for generation of image analyses algorithm as well as batch analyses across all tissues. For this, an analysis algorithm was built using InForm software 2.6.0 (Akoya Biosciences, Marlborough, USA) with representative regions across different samples. The image analysis algorithm consisted of several steps including spectral unmixing based on proprietary spectral libraries of designated fluorochromes as supplied by Akoya Biosciences, Marlborough, USA, pixel-based tissue segmentation, cell segmentation and cell classification. Representative regions from several different WSIs were imported into InForm software, spectrally unmixed and forwarded to user-guided, machine-learning tissue segmentation based on manual annotation for areas consisting either of empty glass/no tissue, B cell enriched tissue and tissue without B cell enrichment. Tissue segmentation accuracy was tested on randomly selected WSIs and re-trained on areas where initial tissue segmentation failed to correctly identify designated areas. Once the tissue segmentation reached satisfactory level, individual cells were segmented based on their DAPI signal and accessory information from antibody staining. A cell classification algorithm was trained to identify each antigen in a binary approach (antigen positive vs negative). For analysis of WSIs, the complete tissue was deconstructed into boxes and subjected to batch analyses. All images were individually reviewed and areas containing imaging artifacts (dust, fibers, tissues folds, out-of-focus) as well as pleura with excess autofluorescence removed. Resulting cell segmentation files of 828 images from 18 WSIs were merged and consolidated into a single data file using R package phenoptr reports 0.3.3 (26). Phenoptr reports was further used to determine the cell density (cells/mm²) for selected combinations of detected antigens across the complete tissue of each animal. The slides were analyzed in a blinded manner.

Superior lung lobes from infected mice were fixed with 4% w/v paraformaldehyde for 24 h and embedded in paraffin. RNA was extracted with RNeasy FFPE Kit (Qiagen, Germantown, USA) from tissue sections, dissolved in nuclease-free water and RNA integrity analyzed using a 2100 Bioanalyzer (RNA 6000 Nano Kit, Agilent Technologies, Waldbronn, Germany). Samples with DV200 values >30% were included for gene expression analysis. The NanoString nCounter (NanoString Technologies, Inc., Seattle, USA) platform was used to assess the effects of B cell derived IL-10 on changes in immune gene expression. Briefly, RNA (n = 3/group) was analyzed using the nCounter Host response panel (catalog # 115000486), which includes 773 immune-related genes and 12 internal reference genes for data normalization. Assays were performed using a nCounter Prep Station 5s (NanoString Technologies, Inc., Seattle, USA) according to the manufacturer’s instructions, and the proceeded cartridge was sent for digital analysis using the nCounter^® Sprint Profiler system (performed at the Institute of Pathology, Hannover Medical School, Hannover, Germany). Briefly, RNA is directly tagged with a capture probe and a reporter probe specific to the genes of interest. After hybridization, samples were washed and the purified target-probe complexes were aligned and immobilized onto the nCounter cartridge. The transcripts were counted via detection of the fluorescent barcodes within the reporter probe. Raw gene expression data were analyzed using NanoString’s software nSolver v4.0 with the Advanced Analysis Module v2.0.134 and ROSALIND^® (https://nanostring.rosalind.bio).

All data were analyzed using GraphPad Prism 10 (GraphPad Software, La Jolla, USA) or Excel 2016 (Microsoft Corporation, Redmond, USA). Statistical tests are indicated in the individual Figure legends. Correlation was determined by calculating Pearson’s coefficient using a two-tailed analysis. A correlation was taken into account as of r≥ 0.60 (defined as strong correlation). Principal Component Analysis (PCA) was performed in BioVinci version 3.0.9 (BioTuring Inc., San Diego, USA) – a machine learning aided analysis platform for biological datasets.

All data generated or analyzed during this study are included in this published article [and its Supplementary Information Files ].

It has previously been shown that B cells are a source of IL-10 during Mtb HN878 infection in female mice (18). Given that sex-based differences in the frequencies of IL-10-producing B cells have been reported in mouse models of arthritis (27), our goal was to assess B cell-specific IL-10 expression following Mtb infection in a sex-specific manner. To define IL-10 production in vivo, we utilized male and female IL-10 transcriptional reporter (Vert-X) mice, which express enhanced green fluorescent protein (eGFP) under the control of the IL-10 promoter (22). They are bred on a C57BL/6 background and exhibited increased male susceptibility in the late stage of Mtb HN878 infection ( Supplementary Figure S1 ), consistent with previous findings reported by our group for C57BL/6 mice (7).

Vert-X mice were infected via the aerosol route with Mtb HN878, and at the indicated time points, single-cell suspensions of lung tissues were prepared and analyzed for GFP expression ( Figure 1A ; gating strategy in Supplementary Figure S2 ). Consistent with previous studies (18), T cells were the main producers of IL-10, and furthermore, we did not determine differences between the sexes ( Figure 1B ). IL-10 expressing B cells, defined as GFP⁺CD138^-B220⁺ B cells and GFP⁺CD138⁺B220^var plasma B cells (where B220^var refers to variable B220 expression, as plasma cells can downregulate B220), were detected in the lungs of both sexes at all-time points investigated. Even though the proportion of these cells within the overall population of IL-10 producers remained constant ( Figure 1C ), the total number of IL-10-producing B cells increased over time ( Figure 1D ). While UMAP analysis revealed a consistent increase in the proportion of GL7⁺ germinal center B cells, CD138⁺ plasma cells (PCs) remained a relatively small population throughout the infection ( Figures 1E, F ). However, PCs were by far the most significant producers of IL-10 at all-time points analyzed ( Figures 1G–K ). This aligns with established literature that identifies PCs as the major source of B cell-derived IL-10 in vivo across various disease models, including autoimmune, malignant and infectious conditions (28). Statistical analysis revealed that there were no significant differences in the proportion of IL-10⁺ PCs and the expression of IL-10 on a per cell basis in males compared to females after Mtb infection ( Figures 1I, J ).

Overall, B cells contribute to IL-10 production in the Mtb infected lung in both sexes, with CD138⁺ PCs representing the primary source of B cell-derived IL-10.

To determine the actual role of B cell-derived IL-10 during Mtb infection we evaluated disease progression in mice that were deficient in IL-10 expression specifically in B cells (Cd19-cre^+/-Il10^fl/fl; B cell IL-10 KO) ( Figure 2A ). B cell IL-10 KO mice were more resistant to the Mtb infection than their littermates as reflected by reduced body weight loss and a prolonged survival ( Supplementary Figure S3 , Figure 2B ). Intriguingly, segregation of survival data by sex revealed that the survival benefit in the absence of B cell-derived IL-10 was more pronounced in males compared to females ( Figures 2C, D ). Together, these data clearly argue against a protective role for B cell-derived IL-10 but instead suggest it increases susceptibility to TB, particularly in males.

We next investigated whether the improved survival observed in mice lacking B cell-derived IL-10 resulted from a better ability to control bacterial loads after HN878 infection. Day 82 analysis of lung tissue revealed no significant difference in bacterial burden between groups. However, by day 103, male B cell IL-10 KO mice exhibited a significantly lower lung bacterial load compared to male littermates ( Figure 2E ), who displayed initial signs of disease around this time ( Supplementary Figure S3C ). In support of our CFU data, analysis of acid-fast bacteria in the lungs of moribund animals revealed that male littermates frequently harbored clusters of mycobacteria, while the KO mice more commonly contained single bacteria ( Figure 2G ). B cell IL-10 deficiency did not affect bacterial burden in the spleen ( Figure 2F ), suggesting distinct roles for B cell-derived IL-10 in the lung and spleen during Mtb infection.

To examine the immunological environment in the lungs that accompanied improved bacterial control in the absence of B cell IL-10, we quantified the levels of inflammatory cytokines and chemokines known to play a role in Mtb infection. While absence of B cell IL-10 did not lead to overall changes in the levels of cyto- or chemokines in female lungs ( Figures 2H, I ), it had profound effects on certain cytokines in male lungs ( Figure 2H ). Specifically, we observed significant changes in IFN-γ and IL-27 levels, with differences being evident not only between littermate and B cell IL-10 KO mice, but also between males and females ( Figures 2J, K ). Interestingly, male littermates displayed significantly lower levels of IFN-γ compared to females ( Figure 2J ), which increased significantly in the absence of B cell-derived IL-10. In contrast to IFN-γ, IL-27 levels were significantly higher in male littermates compared to females ( Figure 2K ). Notably, the absence of B cell IL-10 led to a significant decrease in IL-27 levels specifically in male mice. This reduction brought their IL-27 levels down to a level comparable to that observed in female mice. This difference in how B cell IL-10 regulates IFN-γ and IL-27 between the sexes is further supported by correlation analysis ( Figure 2L ). A strong negative correlation between IL-27 and IFN-γ levels was observed in male mice, while this negative correlation was absent in female mice, indicating a possible different regulatory pathway for these cytokines in female mice compared to male mice.

IL-27, like IL-10, is an important anti-inflammatory cytokine that has a dual role in TB. On the one hand, IL-27 limits protective immunity to Mtb but on the other hand it prevents from excessive inflammation and immunopathology (29). Thus, we next evaluated the granulomatous response in absence of B cell-derived IL-10, which led to a significant reduction in IL-27 levels in males (see Figure 2K ). To this end, H&E-stained lung sections from B cell IL-10 KO mice and littermates were examined 82 days after HN878 infection ( Figure 3A ). No significant difference in cellular infiltration was observed in the lungs of female B cell IL-10 KO mice compared to their female littermates. In contrast, male B cell IL-10 KO mice showed a trend towards greater cellular infiltration compared to their male littermates, which, although not statistically significant, is still noteworthy ( Figure 3B ). In fact, the absence of B cell-derived IL-10 in males led to levels of cellular infiltration in the lungs comparable to those observed in females. This suggests that B cell-derived IL-10 plays a specific role in down-regulating the inflammatory response within the lungs of male mice during Mtb infection. The results from the histological examination align remarkably well with the cytokine analysis, which is further supported by the observed correlations ( Figures 3C, D ). In males, a strong positive correlation exists between the area of inflammation and IFN-γ levels ( Figure 3C ), while conversely, a negative correlation was observed between inflammation and IL-27 levels ( Figure 3D ). Notably, these correlations were absent in females, indicating a potentially distinct inflammatory pathway in females that is not influenced by B cell-derived IL-10.

The sex-specific differences in cellular infiltration and the potential role of B cell-derived IL-10 in males prompted a more detailed examination of immune cell populations within the infected lung using mIF to identify CD19⁺ B cells, CD3⁺ T cells, CD68⁺ macrophages, and neutrophil elastase positive (NE⁺) neutrophils ( Supplementary Figure 4A ). Consistent with our previous findings (7), the total area and density of CD19⁺ B cells within the lesions were lower in male compared to female littermates, although not statistically significant ( Supplementary Figures S4B, C ). Interestingly, B cell IL-10 deficiency did not significantly affect B cell area or density in either sex. Likewise, our analysis of CD3⁺ T cells revealed no major difference between littermates and B cell IL-10 KO mice, or – unlike CD19⁺ B cells - between the sexes ( Supplementary Figure 4D ). In contrast, B cell IL-10 deficiency significantly impacted macrophages and neutrophils in males ( Figure 3E ). Although the overall abundance was low, male littermates exhibited a significantly higher density of NE⁺ neutrophils within their lesions compared to females, and B cell IL-10 deficiency led to a reduction in neutrophils specifically in males, but not in females ( Figure 3F ). Conversely, male B cell IL-10 KO mice had a marked increase in CD68⁺ macrophages compared to littermates, while macrophage density in females remained similar regardless of B cell IL-10 status ( Figure 3G ).

In order to understand the combined influence of the measured parameters described, we employed Principal Component Analysis (PCA) on our comprehensive dataset of day 82 post infection ( Supplementary Table S2 ). The first three principal components (PCs) captured a significant portion (61%) of the total variance within the dataset ( Figure 3H ). The PCA plot revealed a distinct separation between male littermates and all other groups. Interestingly, male B cell IL-10 KO clustered closer to the combined group of female littermates and female B cell IL-10 KO mice. This suggests a sex-based difference in the overall immune response to Mtb infection. Moreover, the closer proximity of male B cell IL-10 KO mice to females aligns with their improved survival compared to male littermates.

Overall, these data suggest that the absence of B cell-derived IL-10 has a more pronounced effect on the immune response in males, making them more susceptible to Mtb in its presence.

To better understand the molecular basis of the observed sex differences in susceptibility and lung inflammation following Mtb infection in mice that lack B cell-derived IL-10, we employed the nCounter platform to analyze immune-related gene expression profiles in the lungs that we had analyzed by mIF before ( Supplementary Table S3 ). Sex-wise gene expression analysis revealed a striking disparity: Only few genes were differentially expressed in female B cell IL-10 KO mice compared to female controls ( Figure 4A ). In stark contrast, male B cell IL-10 KO mice displayed significant changes, with 114 genes being differentially regulated, most of which were downregulated ( Figure 4B ). Quantitative analysis of pathway scores revealed a striking sex difference in response to B cell IL-10 deficiency. Pathway scores in males showed greater variation compared to their female counterparts, where scores remained relatively unchanged by B cell IL-10 depletion ( Figures 4C, D ). Notably, the JAK-STAT pathway, known to play a central role in mediating the effects of IL-10, exhibited the most significant downregulation in male B cell IL-10 KO mice ( Figure 4E ). Additionally, pathways related to B cell receptor (BCR) and IL-2 signaling were also downregulated in males, while the host defense peptides pathway showed a contrasting increase ( Figures 4F–H ).

Given the downregulation of BCR in male B cell IL-10 KO mice, we aimed to assess the impact on Mtb-specific B cell responses. To do this, we measured immunoglobulins in lung homogenates that recognized antigens in an Mtb whole-cell lysate (WCL). Although we observed no differences in the titer of anti-mycobacterial IgG and IgM antibodies between sexes or between littermates and B cell IL-10 KO ( Figures 4I, J ), anti-mycobacterial IgA levels were reduced in B cell IL-10 KO mice, with the reduction reaching significance only in male mice ( Figure 4K ). This finding is consistent with previous reports indicating that IL-10 promotes the generation of plasma cells and enhances IgA production (30–33).

Collectively, these findings provide compelling evidence for our hypothesis that B cell-derived IL-10 plays a far more critical role in regulating the male immune response during Mtb infection compared to females.