Lung epithelial injury impairs early host immune responses to Mycobacterium tuberculosis

Abstract

Objective

Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), remains a significant global health burden, characterized by complex host–pathogen interactions that drive heterogeneous clinical outcomes. While pulmonary epithelial cells are increasingly recognized as active participants in innate immunity during Mtb infection, how host defense are altered when the epithelial barrier is compromised remains unclear.

Methods

In this study, we developed a murine model combining naphthalene-induced pulmonary epithelial injury with Mtb infection and mapped the pulmonary cells landscape through single-cell RNA sequencing (scRNA-seq), followed by in vitro stimulation assays to validate macrophage functional changes.

Results

Notably, we found a pronounced impairment in pulmonary bacterial clearance. Transcriptomic analysis revealed a widespread suppression of epithelial immune functions and showed that macrophages transitioned from an antimicrobial to an antigen-presenting phenotype, indicating waning pulmonary innate defenses and heightened adaptive immune activation. In vitro experiments further suggested that this macrophage transition may be linked to epithelial cell alterations.

Conclusions

These findings indicate that pulmonary epithelial integrity may influence early host immune responses to Mycobacterium tuberculosis and provide a transcriptomic framework for exploring epithelial–immune crosstalk as a potential therapeutic target.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00011-025-02106-4.

Highlights

Single-cell transcriptomics reveal epithelial–immune crosstalk remodeling during Mtb infection in mice.

Pulmonary epithelial injury weakens mucosal defense and increases bacterial burden.

Macrophages shift from innate immune defense to adaptive immune activation in response to Mtb challenge.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00011-025-02106-4.

Introduction

Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), is a chronic infectious disease with the capacity to establish long-term latency in the host [1]. According to the World Health Organization's 2024 Global Tuberculosis Report, TB affected an estimated 10.8 million people in 2023, resulting in approximately 1.25 million deaths. Despite a gradual decline in incidence over recent years, TB remains the leading cause of mortality from a single infectious agent worldwide [2]. The clinical outcomes of Mtb infection vary considerably, typically progressing along one of three paths: complete eradication of the pathogen (“resisters”), development of active tuberculosis (ATB) with high transmissibility, or establishment of latent tuberculosis infection (LTBI) under immune containment [3]^. Among the roughly one-quarter of the global population with LTBI, 5–15% are expected to eventually develop ATB, contributing to sustained transmission [4, 5]. The divergent trajectories of Mtb infection are thought to arise from a combination of host immune dysregulation and intricate host–pathogen interactions. However, the precise cellular and molecular mechanisms that govern these outcomes remain incompletely understood [6, 7].

Macrophages are pivotal effector cells in host immunity against Mtb, functioning both as immune sentinels and primary intracellular reservoirs for the pathogen [8]. Upon infection, they initiate antimicrobial responses through phagolysosomal fusion and the production of reactive oxygen and nitrogen species [9]. In parallel, macrophages coordinate adaptive immunity by presenting antigens to T cells, which in turn secrete interferon-gamma (IFN-γ), reinforcing macrophage activation in a positive feedback loop [10]. However, Mtb employs multiple immune evasion strategies to circumvent macrophage-mediated bactericidal mechanisms, including blocking phagosome maturation, resisting oxidative stress, and reprogramming host signaling pathways [11]. Moreover, by establishing an acidic inflammatory microenvironment within lesions, Mtb modulates macrophage function, attenuating antimicrobial activity and exacerbating tissue damage [12]. Overall, macrophage-dominated immune responses play a central regulatory role in the initiation and progression of Mtb infection.

Beyond immune cells, pulmonary epithelial cells have emerged as critical players in the host response to Mtb [13]. Traditionally viewed as passive physical barriers, epithelial cells are now recognized for their immunoregulatory functions, including pathogen sensing via pattern recognition receptors (e.g., TLRs and NOD-like receptors) [14–16], secretion of antimicrobial peptides such as human β-defensin-1 (hBD-1) and lysozymes [17, 18], and modulation of iron homeostasis through hepcidin expression [19]. In response to infection, pulmonary epithelial cells also rapidly release cytokines (e.g., IL-6, IL-8, TNF-α) [20] and chemokines (e.g., CXCL5, CCL2) [21] that facilitate the recruitment of innate immune cells to sites of infection, thereby initiating localized antimicrobial responses [22–25]. Importantly, epithelial cells engage in bidirectional communication with immune cells, contributing to both innate and adaptive immunity. For instance, pulmonary epithelial antigen presentation via major histocompatibility complex class I (MHC-I) and MHC-related protein 1 (MR1) can activate CD8⁺ T cells to produce IFN-γ, while crosstalk with infected macrophages can amplify IL-1β and type I IFN signaling [26, 27]. These interactions are pivotal in orchestrating mucosal immune responses and maintaining host homeostasis during Mtb infection.

Despite growing interest in the role of the pulmonary epithelium in host defense, a comprehensive understanding of how epithelial barrier integrity influences early immune responses to Mtb is lacking. In particular, the immune consequences of pulmonary epithelial injury during the initial stages of infection remain poorly characterized. In this study, we established a murine model combining naphthalene-induced pulmonary epithelial injury with Mtb infection and employed single-cell RNA sequencing (scRNA-seq) to delineate the remodeling of the pulmonary immune landscape [28]. We found that epithelial dysfunction in the lungs leads to impaired epithelial immune activity and drives a heterogeneous transition of macrophages from innate immune defense toward adaptive immune activation, ultimately weakening early pathogen clearance. Our findings highlight the critical role of pulmonary epithelial cells in orchestrating immune regulation during Mtb infection, shed light on the regulatory role of epithelial–immune interactions in Mtb infection and offer potential new directions for therapeutic intervention.

Methods

Mice

Eight-week-old male C57BL/6 mice were obtained from Shanghai Model Organisms Center, Inc. All procedures were conducted in an ABSL-3 facility at the Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences, under controlled environmental conditions (12-h light/dark cycle, 22–25 °C, 50–60% humidity). Mice had ad libitum access to food and water. All experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the Institute (approval No. ZLJ23006).

Preparation of Mycobacterium tuberculosis (Mtb) H37Rv inoculum

The frozen stock of Mtb H37Rv (− 80 °C) was thawed and streaked onto Middlebrook 7H10 agar plates (Becton Dickinson and Company). After incubation at 37 °C for 4–6 weeks, a single colony was transferred into 10 mL of Middlebrook 7H9 complete liquid medium (Gene-Optimal Science & Technology) supplemented with Oleic acid–Albumin–Dextrose–Catalase (OADC) enrichment (BD) and cultured statically at 37 °C for 2–3 weeks. The culture was gently pipetted, filtered through a 40-μm strainer (Biosharp) to obtain a single-cell bacterial suspension, and centrifuged at 3000 rpm for 10 min. Bacterial pellets were resuspended in phosphate-buffered saline (PBS), and the suspension was adjusted to the desired concentration based on OD600 measurement with a spectrophotometer (Ambion, Thermo Fisher Scientific) and colony-forming unit (CFU) determination by plating on Middlebrook 7H10 agar at 37 °C for 4 weeks [29], ensuring that the final inoculum reached 5 × 10⁶CFU/mL for in vivo infection and 1 × 10⁷ CFU/mL for in vitro experiments. Fresh bacterial suspensions were prepared 1 h before infection, kept on ice, and gently inverted before use to maintain a uniform single-cell suspension.

Murine infection with Mtb

Mice were randomly assigned into four groups: Control group (no treatment), Naph group (naphthalene treatment only), H37Rv group (Mtb infection only), and Naph + H37Rv group (Naphthalene-pretreated followed by Mtb infection). In the study, there are three control groups: Control group (no treatment), Naph group (naphthalene treatment only), H37Rv group (Mtb infection only). These groups were used to systematically dissect the effects of pulmonary epithelial injury, Mtb infection, and their combined impact on lung immune responses. The experimental timeline consisted of intraperitoneal injection of naphthalene on day 0, Mtb infection on day 4, and lung tissue collection on day 18. To induce selective pulmonary epithelial injury, mice received a single intraperitoneal injection of naphthalene (250 mg/kg; Sigma-Aldrich) dissolved in corn oil. On day 4, mice were anesthetized with sodium pentobarbital (50 mg/kg, intraperitoneally; Sigma-Aldrich), followed by intravenous injection of 100 µL of a logarithmic-phase Mycobacterium tuberculosis H37Rv suspension (5 × 10⁶ CFU/mL in sterile PBS) via the tail vein. All mice were housed in an ABSL-3 facility and monitored daily for health and activity. On day 18 post-infection, mice were euthanized under deep anesthesia with pentobarbital (100 mg/kg), and lung tissues were harvested for downstream analyses. For each group: Six mice were used for bilateral lung collection for CFU enumeration. At least five mice were allocated for histological and molecular analyses: left lungs were fixed, embedded, and stained; right lungs were used for RNA and protein extraction. Three mice were used for preparation of single-cell suspensions from both lungs, followed by single-cell RNA sequencing (scRNA-seq).

Colony-forming unit (CFU) assay for Mtb

Lungs were rinsed with 2% sulfuric acid and physiological saline, and then homogenized in 1 mL of physiological saline. After thorough grinding, the homogenate was serially diluted in a tenfold dilution series. 50 µL of each dilution was plated onto 7H10 solid agar plates and incubated at 37 °C for 4 weeks. CFU counts were performed to determine the bacterial load in the pulmonary tissue.

Histological and acid-fast staining

Lung lobes were fixed in 10% neutral-buffered formalin for 48 h, dehydrated, embedded in paraffin, and sectioned at 5 μm. Sections were stained with hematoxylin and eosin (H&E) or acid-fast stain (Boerfu Biotechnology) following standard protocols. Digital slide scanning was performed using a NanoZoomer system.

Immunofluorescence staining

After deparaffinization and antigen retrieval, sections were permeabilized with 0.2% Triton X-100 and blocked with 5% normal goat serum. Primary antibodies were incubated overnight at 4 °C, followed by fluorophore-conjugated secondary antibodies for 2 h at room temperature. Nuclei were counterstained with DAPI. Antibodies used are detailed in the supplementary information (Table S1). For quantification, epithelial cells were identified by marker expression and DAPI staining in at least five random fields per group. Data were expressed as the proportion of marker-positive cells among total epithelial cells.

RNA extraction and RT-qPCR

Total RNA was extracted from homogenized lung tissues using TRIzol reagent (Ambion, Thermo Fisher Scientific), followed by reverse transcription using Vazyme kits. Quantitative PCR was performed using SYBR Green master mix on a LightCycler system (Vazyme). Expression levels of Ifng, Tnf, Il6, Il1b, Hamp, and Il17a were normalized to β-actin, and calculated using the 2^^–ΔCt. Primer sequences are listed in Table S2.

Western blot analysis

Lung tissues were homogenized, and the lysates were prepared using lysis buffer. Protein concentrations were determined with a BCA protein assay kit. Equal amounts of protein (10 μg) were separated via 10% SDS-PAGE, transferred to PVDF membranes, and blocked in 5% non-fat milk for 1 h. Membranes were incubated overnight at 4 °C with primary antibodies against MHC-II and Tubulin (Santa Cruz Biotechnology), followed by HRP-conjugated secondary antibodies (Proteintech) for 2 h at room temperature. Protein bands were visualized using enhanced chemiluminescence (ECL) and quantified with ImageJ software (version 1.54).

Single-cell RNA sequencing and data processing

Lung tissues were dissociated using a commercial enzymatic dissociation kit (SeekGene BioTech), and cell suspensions were filtered, stained, and assessed for viability using a fluorescence cell analyzer (Countstar Rigel S2). High-quality cells were enriched by magnetic separation and used to generate 3′-end droplet-based scRNA-seq libraries using the SeekOne^® platform. Libraries were sequenced on an Illumina NovaSeq 6000 with paired-end 150-bp reads. Raw FASTQ files were processed using fastp for quality trimming, and aligned to the mouse reference genome using STAR. Gene expression matrices were generated using featureCounts and filtered for cell quality using Seurat and EmptyDrops. Cells with < 300 UMIs or high mitochondrial content were excluded. Normalization, scaling, and dimensionality reduction were performed using Seurat v4.0. Highly variable genes were selected, followed by PCA and UMAP visualization. Cell clustering was conducted using a graph-based approach, and cell types were annotated using canonical marker genes (e.g., Epcam for epithelial cells, Col1a1 for fibroblasts, Ptprc for immune cells). Immune subsets were further annotated using established lineage markers (Table S3). Differential expression analysis between groups was performed using the FindMarkers function. The differentially expressed genes (DEGs) among the two groups were subjected to further analysis using the FindMarkers function of the Seurat package under the conditions of fold change > 1.5 and adjusted p-value < 0.05. Subsequently, enrichment analysis of the DEGs was conducted using the online tool on the DAVID website (https://david.ncifcrf.gov/).

Macrophage culture and H37RV infection

RAW264.7 macrophages (kindly provided by Prof. Chunquan Cai, Tianjin Pediatric Research Institute) were cultured in complete medium (Zhong Qiao Xin Zhou Biotechnology) at 37 °C in a humidified incubator with 5% CO₂. Cells in good condition were seeded into 6-well plates at a density of 1 × 10⁶ cells/well and allowed to adhere for 24 h. The culture medium was then replaced with fresh medium containing recombinant murine proteins C3, C5, or Gas6 (2000 ng/mL). Control wells received medium without recombinant proteins. After 24 h of stimulation, supernatants and cell lysates were collected for subsequent analyses. To assess the functional effect of epithelial ligands on macrophage phagocytic capacity, RAW264.7 macrophages were infected with Mtb H37Rv (1 × 10⁷ CFU/mL, MOI = 10), either alone or in combination with the aforementioned recombinant proteins (final concentration: 2000 ng/mL). After 24 h, cells were washed thoroughly with PBS to remove extracellular bacteria and permeabilized with 0.1% Triton X-100. Intracellular Mtb-specific bacterial rRNA was detected using the SAT (Simultaneous Amplification and Testing) assay (Rendu Biotechnology) according to the manufacturer’s instructions. Results were expressed as SAT Detection Time (DT, min), with shorter DT values generally suggesting a higher initial rRNA abundance and a potentially greater bacterial burden. All infection assays were conducted in triplicate.

Enzyme-linked immunosorbent assay (ELISA)

Supernatants in macrophage cultures were centrifuged at 2,500 rpm for 5 min at 4 °C to remove debris. Concentrations of TNF-α, IFN-γ, and Camp were measured using commercial ELISA kits (Fine BioTech) according to the manufacturer’s protocols. Absorbance at 450 nm was recorded, and cytokine levels were calculated using standard curves. All measurements were performed in at least three independent biological replicates.

Statistical analysis

Data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 10.1.2. Comparisons between two groups were assessed by unpaired two-tailed Student’s t-tests. A p-value < 0.05 was considered statistically significant. scRNA-seq data analysis (including DEGs, GO enrichment, and CellChat interaction networks) was performed using R 4.1.2 and relevant packages as specified in figure legends.

Results

Mtb infection activates pro-inflammatory and antimicrobial programs in the murine lung

To investigate cellular and molecular alterations in the lung during early Mtb infection, we established a murine infection model and performed histological staining and scRNA-seq on lung tissues collected 14 days post-infection (Fig. 1A). H&E staining revealed significant alveolar architectural disruption and pronounced infiltration of inflammatory cells in the lungs of the H37Rv-infected group (H37Rv group) compared with the uninfected controls (Fig. 1B). Acid-fast staining confirmed the localization of Mtb bacilli within inflammatory foci (Fig. 1C), suggesting that infection triggers local chemotactic immune responses.

Fig. 1.

Single-cell profiling reveals immune activation following Mtb infection in mice. A Experimental workflow of H37Rv infection in mice. Lung tissues were collected on day 14 post-infection for histology and scRNA-seq. B Representative H&E-stained lung sections showing alveolar destruction and inflammatory infiltration in H37Rv-infected mice, scale bar = 100 µm. C Acid-fast staining showing Mtb localization within inflammatory lesions, scale bar = 25 µm. D UMAP plot displaying major cell types in the control and H37Rv groups based on scRNA-seq clustering. E Comparison of global signaling pathway activity using CellChat. Significantly upregulated pathways in the H37Rv group are highlighted. F RT-qPCR analysis of gene expression in lung tissues. *p < 0.05, **p < 0.01, ***p < 0.001. G Differential analysis of intercellular communication networks between groups. Left: number of interactions; Right: interaction strength. H IFN-γ signaling network visualized by CellChat. each node represents a distinct cell type, edge thickness indicates signal strength

scRNA-seq was subsequently conducted on lung tissues from both control and infected mice, yielding 25,932 high-quality cells. After dimensionality reduction and unsupervised clustering, four major cellular compartments were identified based on canonical marker genes: epithelial cells, immune cells, endothelial cells, and stromal cells (Fig. 1D and Supplementary Fig. 1A). Further sub-clustering of the immune compartment revealed diverse immune subsets including monocytes, macrophages, T cells, NKT cells, NK cells, B cells, dendritic cells, neutrophils, and basophils (Supplementary Fig. 1B). To examine changes in intercellular communication across the entire lung during infection, we employed CellChat analysis to infer ligand–receptor interactions. Compared with the control group, the H37Rv-infected group exhibited increased signaling activity in multiple pro-inflammatory pathways, notably those associated with IFN-γ and IL-16 (Fig. 1E). Consistent with this, RT-qPCR analysis of lung tissue revealed significantly elevated mRNA levels of pro-inflammatory cytokines (Ifng, Tnf, Il6, Il1b) and a marked reduction in Hamp expression (Fig. 1F), indicating heightened inflammation and altered iron homeostasis. Quantitative analysis of cell–cell communication revealed that although the overall number of interactions remained relatively stable between groups, the signaling strength between macrophages and T cells was notably enhanced in the H37Rv group. This suggests that Mtb infection promotes macrophage-derived immunoregulatory signaling that enhances T cell activation (Fig. 1G). In particular, increased IFN-γ signaling from T cells to macrophages was observed, consistent with a coordinated amplification of adaptive immunity during early infection (Fig. 1H). These data indicate that Mtb infection is associated with increased immune cell crosstalk—particularly between macrophages and T cells—which may contribute to coordinated innate and adaptive immune responses. Collectively, these results show that Mtb infection triggers a pro-inflammatory immune response in the lung, characterized by immune cell recruitment and activation of effector cells such as macrophages, which play a critical role in host defense.

Macrophages dominate the early immune landscape in Mtb-infected lungs

To further elucidate the cellular drivers of the pulmonary immune response during Mtb infection, we performed an in-depth analysis of immune subsets from the scRNA-seq dataset. A marked expansion of immune cells was observed following infection: immune cells comprised 86.18% of all pulmonary cells in the H37Rv group, compared to 73.40% in the control group, reflecting a 12.78% increase (Fig. 2B). Within this compartment, macrophages exhibited the most prominent increase, with a 17.32% rise in abundance relative to uninfected controls (Fig. 2C), indicating robust macrophage recruitment and accumulation in response to Mtb infection. Differential gene expression analysis identified 3981 significantly DEGs across all cells between the H37Rv and control groups (fold change > 1.5 and adjusted p-value < 0.05), including 2158 upregulated and 1823 downregulated genes. Specifically in macrophages, 214 genes were upregulated and 207 were downregulated (Fig. 2D). Notably, several pro-inflammatory and antimicrobial genes were upregulated in macrophages following infection. These included Il1b (encoding IL-1β), as well as the chemokines Cxcl10, Ccl4, and Ccl5, which mediate the recruitment of T cells, monocytes, and NK cells. Upregulation of Camp, Ltf, and Itgam indicated enhanced bactericidal and phagocytic functions, while elevated expression of Csf1r suggested increased macrophage proliferation and survival (Fig. 2E). Gene Ontology Biological Process (GOBP) enrichment analysis of the 214 upregulated macrophage DEGs revealed significant activation of pathways related to innate immune regulation, bacterial defense, leukocyte cytotoxicity, and interferon signaling (Fig. 2F). Further categorization of tuberculosis-related GOBP terms highlighted enrichment in phagocytosis, natural killer (NK) cell activation, TNF production, granulocyte chemotaxis, and iron ion transport regulation (Fig. 2G), underscoring a broad-spectrum activation of macrophage antimicrobial programs.

Fig. 2.

Transcriptomic activation of lung macrophage by Mtb infection. A UMAP showing immune cell subsets of lung tissues in the control and H37Rv groups. B Pie charts illustrating the proportion of immune cells among total cells in each group. C Pie charts depicting macrophage proportions among total cells. D Summary of differentially expressed genes (DEGs) across all cell types (fold change > 1.5 and adjusted p-value < 0.05). E Violin plots showing representative inflammation-related DEGs upregulated in macrophages. F Top 10 Gene Ontology Biological Process (GOBP) terms in macrophages. G Selected TB-related functional pathways significantly enriched in macrophages

Beyond macrophages, more than 200 DEGs were identified in other cell subsets—including monocytes, T cells, dendritic cells, neutrophils, basophils, and endothelial cells (Fig. 2D)—reflecting widespread immune remodeling. In monocytes, DEGs were enriched in ribonucleoprotein biogenesis and translation-associated pathways, suggestive of active biosynthetic states supporting immune reprogramming (Supplementary Fig. 2A). T cell DEGs were associated with T cell activation, differentiation, and intercellular adhesion, indicating heightened adaptive immune engagement (Supplementary Fig. 2B) [30]. Neutrophils exhibited signatures of both immune activation and regulation, including enrichment in positive and negative regulators of the innate immune response (Supplementary Fig. 2C) [31]. Dendritic cells showed increased expression of genes related to lymphocyte activation and antigen presentation (Supplementary Fig. 2D), while basophils suggested enrichment in antibody-dependent cytotoxicity and MHC class Ib antigen presentation (Supplementary Fig. 2E). Concurrently, endothelial cells were also activated, with upregulated genes involved in IFN-γ signaling and immune regulation (Supplementary Fig. 2F). In summary, Mtb infection is associated with a robust, macrophage-enriched immune landscape, along with transcriptional signatures suggestive of coordinated activation across both innate and adaptive immune compartments.

Mtb infection enhances epithelial immune responses while compromising barrier integrity

Given the observed tissue disruption and inflammatory cell infiltration upon Mtb infection (Fig. 1B), we next investigated the impact of infection on epithelial barrier integrity and function. Immunofluorescence staining was performed to quantify major epithelial subtypes in lung tissue sections (Fig. 3A). Compared with the control group, no significant differences were observed in the numbers of club cells (CYP2F2⁺), ciliated cells (TUBA1A⁺), or alveolar type II (AT2) cells (SPC⁺). The expression of SCGB1A1, a key secretory protein of club cells, remained unchanged. However, a significant increase in goblet cells (CLCA1⁺) was detected in the H37Rv group (Fig. 3B), suggesting enhanced differentiation of club cells into goblet cells. These results indicate that Mtb infection drives a transition of epithelial cells from a homeostatic state to an immune defense state in response to pathogen invasion.

Fig. 3.

Enhanced anti-Mtb immunity and barrier damage in lung epithelium. A Immunofluorescence images of pulmonary epithelial cells: Club (CYP2F2⁺), Goblet (CLCA1⁺), Ciliated (TUBA1A⁺), AT1 (PDPN⁺), and AT2 (SPC⁺) cells. B Quantification of pulmonary epithelial cell proportions. C) Violin plots of representative DEGs upregulated in pulmonary epithelial cells. D Top 10 enriched GOBP terms from upregulated and downregulated DEGs in pulmonary epithelial cells. E Tuberculosis-related pathways in pulmonary epithelial cells

To further explore the transcriptional reprogramming of epithelial cells during infection, we conducted DEG analysis on epithelial subsets from the scRNA-seq data. Compared with controls, the H37Rv group exhibited 95 upregulated and 235 downregulated genes in epithelial cells (Fig. 2D). Upregulated genes included Lyz1 and Lyz2 (lysozymes), Slpi and Scgb1a1 (barrier protective factors), and Cxcl15 (a chemokine mediating immune cell recruitment), indicating enhanced mucosal defense functions. Additionally, the surfactant-associated genes Sftpd, Sftpb, and Sftpa1 were significantly elevated, which likely contributes to maintaining alveolar surface stability and enhancing pathogen recognition and clearance (Fig. 3C). GOBP enrichment analysis revealed that upregulated genes were significantly enriched in biological processes such as regulation of autophagy and T cell–mediated cytotoxicity (Fig. 3D), suggesting that epithelial cells may facilitate microbial clearance by both intrinsic defense mechanisms and cooperative immune signaling. In contrast, downregulated genes were associated with extracellular matrix organization, cell–cell adhesion, angiogenesis, and structural remodeling-indicating that infection compromises epithelial regenerative and structural integrity. Further functional classification highlighted that upregulated genes were enriched in immune-related processes including innate immune response, defense against bacteria, response to type II interferon, phagocytosis promotion, and inflammatory regulation (Fig. 3E). Enrichment in surfactant homeostasis pathways also suggested adaptive epithelial responses aimed at maintaining alveolar functionality under stress. Meanwhile, downregulated genes were significantly involved in processes related to lung development, alveolar morphogenesis, epithelial migration, tight junction formation, and barrier maintenance, further indicating that infection may compromise epithelial structural integrity and impair the regenerative capacity of the pulmonary. In summary, Mtb infection is associated with transcriptional changes in pulmonary epithelial cells that may reflect enhanced participation in local mucosal immune responses.

Pulmonary epithelial injury exacerbates early bacterial burden

To evaluate the impact of epithelial injury on host immune responses to Mtb, we employed a dual-injury murine model, in which mice were pretreated with naphthalene to induce pulmonary epithelial cell injury prior to infection with H37Rv (the Naph + H37Rv group; Fig. 4A). H&E staining of lung tissues revealed pronounced alveolar destruction and inflammatory infiltration in both the H37Rv and Naph + H37Rv groups (Fig. 4B). In contrast, naphthalene treatment alone (the Naph group) did not result in significant pathological changes by day 18 post-injury, indicating that epithelial regeneration had largely occurred by this time (Supplementary Fig. 3A). Acid-fast staining confirmed Mtb accumulation within inflammatory lesions in both infected groups (Fig. 4C). Quantification of granulomatous regions showed no significant difference in the percentage of affected lung tissue between the H37Rv and Naph + H37Rv groups (Fig. 4D). However, bacterial burden, as determined by CFU assays, was significantly higher in the Naph + H37Rv group (Fig. 4E), suggesting impaired early bacterial clearance in the setting of prior epithelial injury.

Fig. 4.

Epithelial injury enhances Mtb burden following infection. A Schematic of naphthalene-induced epithelial injury followed by Mtb infection. B H&E-stained lung sections from the H37Rv and Naph + H37Rv groups. C Acid-fast staining of lung tissues. D Quantification of granuloma area as a percentage of total lung area. E Colony-forming unit (CFU) counts in lungs from the H37Rv and Naph + H37Rv groups. F UMAP showing cell type distribution from the H37Rv and Naph + H37Rv groups. G CellChat-based comparison of signaling pathway activity between two groups. H RT-qPCR analysis of inflammation-related cytokine mRNA expression in lung tissues. *p < 0.05; **p < 0.01; ***p < 0.001. I Visualization of intercellular communication differences between two groups. Left: number of interactions; Right: interaction strength. J Intercellular communication networks of the IFN-γ signaling pathway in the two groups, each node represents a distinct cell type, edge thickness indicates signal strength

To assess the cellular basis of this altered host response, we compared cell-type distributions between groups using scRNA-seq. Spatial distributions of lung cell subsets were not significantly altered between the H37Rv and Naph + H37Rv groups (Fig. 4F), nor between the Naph and control groups (Supplementary Fig. 3B), suggesting that epithelial injury does not overtly affect pulmonary cellular architecture at this time point. CellChat analysis revealed that epithelial injury exacerbated immune signaling alterations during Mtb infection. Compared to the H37Rv group, the Naph + H37Rv group exhibited heightened activation of multiple immune-related pathways, including CX3C, TNF, and notably IFN-γ signaling (Fig. 4G). These changes were not observed in the Naph group alone, confirming that the effects were infection-specific rather than injury-induced (Supplementary Fig. 3C). RT-qPCR analysis supported these findings, showing significantly elevated expression of Ifng and Tnf, but reduced Il6 and Il1b expression in the Naph + H37Rv group relative to the H37Rv group (Fig. 4H). In the Naph group, only Hamp expression was elevated, indicating perturbation of iron homeostasis without active inflammation (Supplementary Fig. 3D). Further intercellular communication analysis showed an increased number and intensity of cell–cell interactions in the Naph + H37Rv group compared to the H37Rv group, particularly in macrophage-T cell signaling and macrophage autocrine loops (F4g. 4I). Specifically, IFN-γ ligand output from T cells to macrophages was markedly enhanced, pointing to elevated T cell-driven activation of macrophages (Fig. 4J). In contrast, epithelial cell-stromal cell/endothelial cell/macrophage communication was increased in the Naph group versus control (Supplementary Fig. 3E), which may be associated with epithelial regeneration and repair. This process appeared to be independent of IFN-γ signaling regulation (Supplementary Fig. 3F). Taken together, these findings suggest that epithelial injury may contribute to impaired Mtb clearance and is associated with increased immune heterogeneity in the lung microenvironment.

Transcriptomic changes in recovered lung epithelium following Mtb infection

To further investigate how prior epithelial injury impacts barrier function during early Mtb infection, we analyzed epithelial cell regeneration and immune competence in the Naph + H37Rv group. Immunofluorescence staining and quantitative analysis revealed no significant differences in the abundance of club cells (CYP2F2⁺), ciliated cells (TUBA1A⁺), goblet cells (CLCA1⁺), or alveolar type II (AT2) cells (SPC⁺) between the Naph + H37Rv and H37Rv groups (Fig. 5A–B, Supplementary Fig. 4A–B). SCGB1A1 secretion by club cells was also unaffected. Similarly, epithelial cell numbers and distribution in the Naph group were comparable to those in the control group, suggesting that epithelial regeneration was largely complete by day 18 post-injury and that Mtb infection did not impair this process. However, transcriptomic analysis revealed significant alterations in epithelial gene expression profiles. Compared to the H37Rv group, epithelial cells from the Naph + H37Rv group exhibited 213 upregulated and 128 downregulated genes (Fig. 5C). Notably, several key antimicrobial and immunoregulatory genes (Lyz2, Slpi, Cxcl15, Sftpd, Sftpb, and Sftpa1) were markedly downregulated, suggesting a reduction in epithelial-mediated pathogen defense.

Fig. 5.

Transcriptomic changes in recovered lung epithelium following Mtb infection. A Immunofluorescence images of pulmonary epithelial cells: Club cells (CYP2F2 +), goblet (CLCA1 +) and ciliated cells (TUBA1A +), AT1 (PDPN +) and AT2 cells (SPC +). B Quantification of epithelial cell proportions in the Naph + H37Rv and H37Rv groups. C Violin plots showing the DEGs in pulmonary epithelial cells (fold change > 1.5 and adjusted p-value < 0.05). D Top 10 enriched GOBP terms from upregulated and downregulated DEGs in pulmonary epithelial cells. E Tuberculosis-related pathways in pulmonary epithelial cells

GOBP enrichment analysis of upregulated genes revealed significant enrichment in pathways related to ciliary structure and motility, which may facilitate mucus clearance but do not directly contribute to bacterial killing (Fig. 5D). Conversely, downregulated genes were enriched in pathways associated with protein folding, cytoplasmic translation, and ribosomal function, indicating impaired protein synthesis and cellular homeostasis. Further analysis of immune-related biological processes showed that downregulated genes were involved in antibacterial defense, phagocytosis promotion, IL-1–mediated signaling, fibroblast growth factor response, and chemokine-mediated signaling (Fig. 5E). These changes reflect compromised epithelial contributions to innate immunity and immune cell recruitment. In addition, downregulated pathways included those regulating epithelial proliferation and autophagy, suggesting that regenerative capacity and intracellular pathogen clearance were also impaired. Collectively, these findings indicate that while epithelial structure and cell-type composition are restored after injury, their functional competence in immune defense is substantially reduced. These transcriptional changes may reflect reduced epithelial immune competence during early infection, potentially limiting pathogen clearance and shaping the local immune response.

Epithelial injury influences macrophage immune functions during Mtb infection

To elucidate how prior epithelial injury influences immune cell functions during Mtb infection, we examined immune cell responses using the scRNA-seq dataset. While overall immune cell proportions and spatial distributions were similar between the Naph + H37Rv and H37Rv groups, a modest increase in macrophage abundance (5.97%) was observed in the Naph + H37Rv group (Fig. 6A–C). Differential gene expression analysis across all immune cells identified 796 upregulated and 827 downregulated genes, including 22 upregulated and 26 downregulated genes specific to macrophages (Fig. 6D). Within macrophages, pro-inflammatory gene Il1b was further upregulated in the Naph + H37Rv group, suggesting sustained inflammatory signaling. Notably, genes encoding MHC molecules (H2-Aa, H2-Ab1, H2-Eb1 (MHC-II), and H2-DMa, H2-DMb1 (MHC-I)) were also significantly upregulated, indicating enhanced antigen presentation capacity. In contrast, antimicrobial effectors Camp and Ltf, previously upregulated in Mtb infection alone, were downregulated following epithelial injury (Fig. 6E). These results suggest that macrophages undergo functional reprogramming—shifting from a microbicidal, innate-dominant phenotype to one characterized by antigen presentation and adaptive immune activation. Control analyses of the Naph group (without Mtb infection) showed no significant changes in immune cell distribution or gene expression compared to the uninjured control group (Supplementary Fig. 5A–D), confirming that the immune reprogramming observed in the Naph + H37Rv group was specific to the interaction between epithelial injury and infection.

Fig. 6.

Epithelial injury influences macrophage immune functions during Mtb infection. A UMAP showing immune cell subset distributions in the H37Rv and Naph + H37Rv groups. B Pie charts showing immune cell proportions among total cells. C) Pie charts showing macrophage proportions among total cells. D Summary of DEGs across all cell types (fold change > 1.5 and adjusted p-value < 0.05). E Violin plots of representative DEGs in macrophages from the H37Rv and Naph + H37Rv groups. F Top 10 enriched GOBP terms from upregulated and downregulated DEGs in macrophages. G Tuberculosis-related pathways in macrophages

To explore the functional implications of this shift, GOBP enrichment analysis was performed on DEGs derived from macrophages. Upregulated genes in the Naph + H37Rv group primarily related to antigen processing and presentation, T cell activation, and adaptive immune regulation (Fig. 6F). Conversely, downregulated genes primarily related to humoral immune responses, antimicrobial peptide production, ribosome biogenesis, and protein translation—hallmarks of macrophage innate immunity. Classification of tuberculosis-related GOBP terms (Fig. 6G) further reinforced this dichotomy: upregulated genes were associated with effector activation and adaptive immune enhancement, while downregulated genes mapped to pathways involved in NK cell differentiation, iron ion transport, pro-inflammatory signaling, and humoral antimicrobial defense. Together, these findings suggest that epithelial barrier injury may alter pulmonary immunity to Mtb, potentially contributing to macrophage transcriptional shifts from an innate antimicrobial to an antigen presentation–oriented phenotype.

Western blot analysis of lung tissues revealed elevated MHC-II protein levels in the Naph + H37Rv group relative to the H37Rv group. Combined with macrophage DEG profiles, this suggests an enhancement of macrophage antigen-presenting capacity in the Naph + H37Rv group (Fig. 7A). To further evaluate the functional relevance of epithelial–macrophage interactions predicted by scRNA-seq analysis, we performed in vitro stimulation and infection assays using RAW264.7 murine macrophages. Cells were treated with recombinant murine proteins C3, C5, or Gas6, which were among the epithelial ligands downregulated after epithelial injury (Fig. 7B, Online Resource 1). ELISA assays showed that stimulation with C5 significantly reduced IFN-γ and TNF-α secretion by macrophages, while treatment with C3 and Gas6 led to a mild decrease (Fig. 7C). These findings suggest that C3, C5, and Gas6 exert certain effects on regulating macrophage inflammatory responses. In addition, stimulation with C5 resulted in an upward trend in Camp secretion by macrophages, indicating that C5 may enhance macrophage antibacterial function by upregulating antimicrobial peptide levels. These results align with in vivo transcriptomic signatures of pulmonary immune dysregulation (Fig. 4H). To directly assess bacterial phagocytic capacity, we infected macrophages with Mtb H37Rv and quantified the levels of intracellular bacterial rRNA using the Simultaneous Amplification and Testing (SAT) assay. The observed reduction in DT values following stimulation with C3, C5, and Gas6 may reflect higher initial rRNA abundance, which could be consistent with a trend toward increased intracellular bacterial burden (Fig. 7D). These results support a model in which epithelial-derived signals modulate macrophage immune functions, consistent with the transcriptional shifts observed in vivo. Collectively, these findings provide preliminary functional evidence supporting the immunomodulatory role of epithelial-derived ligands on macrophage antigen-presenting capacity and antimicrobial capacity during early Mtb infection.

Fig. 7.

Regulation of macrophage function by epithelial secretory proteins. A Immunoblot and quantification of MHC-II in lung tissues from the H37Rv and Naph + H37Rv groups, with Tubulin serving as the loading control. *p < 0.05. B Violin plots of representative epithelial-derived ligands involved in epithelial–macrophage signaling and differentially expressed between the H37Rv and Naph + H37Rv groups (fold change > 1.2 and p-value < 0.05). C ELISA analysis of IFN-γ, TNF-α, and Camp levels in the supernatants of RAW264.7 macrophages stimulated with recombinant C3, C5, or Gas6 for 24 h. *p < 0.05; **p < 0.01; ***p < 0.001. D Simultaneous Amplification and Testing (SAT) detection time (DT) for intracellular Mtb rRNA in RAW264.7 macrophages infected with H37Rv for 24 h, either alone (control) or in combination with recombinant C3, C5, or Gas6 (final concentration, 2000 ng/mL). Shorter DT values generally indicate a higher initial rRNA abundance and a greater bacterial burden. (E) In mice pretreated with naphthalene to induce epithelial injury prior to Mtb infection, epithelial immune gene expression was markedly reduced, indicating barrier dysfunction. Macrophages showed increased expression of antigen presentation–related genes (MHC-I and MHC-II) and decreased expression of the antimicrobial effector Camp, reflecting a phenotypic shift from innate antibacterial activity toward adaptive immune modulation, likely regulated by alterations in the epithelial secretome (C3, C5, Gas6). These coordinated changes in the pulmonary immune landscape may contribute to impaired early pathogen clearance.

Discussion

The clinical outcome of Mtb infection is shaped by the dynamic interplay between host immune defenses and pathogen-mediated immune modulation [32, 33]. Our findings reveal that epithelial barrier integrity is crucial for maintaining effective host immunity and limiting pathogen proliferation during early Mtb infection. Disruption of this barrier not only weakens epithelial antimicrobial defenses but also alters macrophage functional states-favoring adaptive immune activation over innate bactericidal responses. This immune reprogramming impairs host's early bacterial clearance and may facilitate immune evasion.

Consistent with previous reports [34, 35], our data indicate that Mtb infection elicits a macrophage-centered inflammatory response, characterized by enhanced recruitment and upregulation of pro-inflammatory and antimicrobial genes such as Il1b, Cxcl10, and Ltf. CellChat analysis further indicated intensified signaling between macrophages and T cells, particularly through the IFN-γ axis, suggesting active coordination between innate and adaptive arms of immunity [36]. In parallel, pulmonary epithelial cells exhibited upregulated expression of immune-regulatory genes (Lyz1, Scgb1a1, Sftpd, Cxcl15), implicating them in mucosal defense and immune recruitment [37–41]. However, epithelial cells also showed downregulation of pathways involved in tissue repair, adhesion, and morphogenesis [42, 43], indicating compromised barrier restoration under infection-induced stress.

To probe the functional consequences of barrier dysfunction, we introduced naphthalene-induced pulmonary epithelial injury prior to Mtb infection [44]. This dual insult resulted in elevated bacterial burden despite histological evidence of epithelial regeneration. Transcriptomic analysis revealed diminished expression of antimicrobial and immune-activating genes (Slpi, Lyz2, Sftpa1) in epithelial cells, accompanied by impaired autophagy and phagocytosis-related pathways. These findings indicate that epithelial cells, while structurally restored, are functionally compromised in their immune roles during early infection. Importantly, epithelial injury also reprogrammed macrophage responses in the lung. Compared to Mtb infection alone, macrophages in the Naph + H37Rv group displayed reduced expression of innate effectors (Camp, Ltf) and increased expression of MHC-related antigen presentation genes (H2-Aa, H2-Ab1), reflecting a shift toward adaptive immune facilitation. GOBP enrichment analysis confirmed suppression of humoral and antimicrobial pathways and enrichment of adaptive immune and antigen-processing functions [45]. These changes were not observed in epithelial injury alone, suggesting a synergistic effect of infection and barrier disruption. Our results align with previous findings that immune cell fate can be redirected by environmental cues during infection [46–49]. Here, we provide single-cell evidence that epithelial dysfunction serves as such a cue, reshaping immune landscapes in the lung and reducing host clearance capacity (Fig. 7E).

Despite the comprehensive single-cell transcriptomic profiling conducted in this study, we acknowledge that our findings are primarily correlative in nature. Although we observed robust transcriptional changes indicative of suppressed epithelial immune responses and altered macrophage functional states, these patterns alone do not confirm changes in biological activity or establish causal relationships. To partially address this limitation, we conducted an in vitro validation experiment in which recombinant epithelial ligands were used to stimulate macrophages. The resulting phenotypic changes were consistent with predictions from our scRNA-seq data, but represent only preliminary functional evidence. Further investigations are needed to substantiate the mechanistic link between epithelial injury and impaired pathogen clearance. Such studies may include phagocytosis and bacterial killing assays, cytokine secretion profiling, epithelial–macrophage co-culture systems [50], and in vivo loss-of-function models targeting epithelial-derived signaling pathways. It is important to emphasize that our findings, while supported by high-resolution transcriptomic data, are exploratory and hypothesis-generating. The observed transcriptional shifts in epithelial and immune cell populations offer insight into potential mechanisms of epithelial–immune crosstalk but do not demonstrate functional consequences. As our conclusions are based primarily on scRNA-seq-derived associations, further validation through functional experiments and perturbation models will be essential to confirm causality and therapeutic relevance.

In conclusion, our study reveals that pulmonary epithelial injury is associated with widespread alterations in the immune microenvironment during early Mycobacterium tuberculosis infection. Specifically, we observed weakened epithelial antimicrobial signatures and a transcriptional shift in macrophages toward an antigen-presenting phenotype. These changes coincided with impaired bacterial clearance in vivo. While our findings provide important single-cell transcriptomic insights into epithelial–immune crosstalk, we acknowledge that further functional studies are needed to confirm causality. Collectively, this work offers a foundation for future investigations into the immunomodulatory role of the epithelium in tuberculosis and highlights potential avenues for therapeutic exploration.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

XM, XL, and HC designed and performed all experiments. XM, ZH, and GX helped with animal studies; XM, XL, YW, JW and YL analyzed the data; XM, XL, QW and HC drafted the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

The study was funded by the Natural Science Foundation of Tianjin, China (23JCYBJC01370, 24JCYBJC01420), Tianjin Health Research Project (TJWJ2022XK035) and Tianjin Key Medical Discipline Construction Project (TJYXZDXK-3-018B).

Data availability

The single-cell RNA sequencing (scRNA-seq) data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE298041 and are publicly available at: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE298041.

Declarations

Conflict of interest

The authors declare no conflict interests.

Institutional review board statement

All experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the Institute (approval No. ZLJ23006).