L-kynurenine reshapes immune microenvironment to alleviate methamphetamine-induced chronic lung injury through gut-lung axis

Abstract

Background

Long-term abuse of methamphetamine (MA) is strongly associated with severe lung injury. Microbiome metabolites are one way to understand the interactions between microbes and disease. Although gut microbes and their metabolites play a crucial role in the gut–lung axis, the microbial mechanism by which MA induces lung injury is unclear. The purpose of this work was to identify the omics characteristic factor associated with MA abuse and explore its immune regulatory mechanism by 16 s rDNA sequencing, LC–MS/MS non-targeted metabolomics analysis, hemodynamics, flow cytometry, and some methods of cellular and molecular biology and morphology.

Results

Based on the joint analysis of the gut microbiome and metabolomics, it was found that MA abuse disrupted the structure of the gut microbiome and drove the reprogramming of metabolites, leading to a reduction in Lactobacillus rhamnosus and its metabolite L-kynurenine (L-KYN). Activated Lactobacillus increased L-KYN level in MA-administrated mice. L-KYN, as a product of Lactobacillus, is a key omics signature factor for MA abuse, which has been further confirmed in vivo. L-KYN induced Treg cells differentiated from CD4⁺ T cells and reshaped the immune microenvironment. L-KYN induced the secretion of IL-10 by Treg cells, mediated the communication between Treg cells and alveolar epithelial cells (AEC) through IL-10, and alleviated MA-induced lung inflammation and alveolar barrier damage through the IL-10/JAK1/STAT3 pathway.

Conclusions

From the perspective of intestinal microbiome–metabolite–immune network regulation, the omics characteristic factor L-KYN reshaped the immune microenvironment and alleviated methamphetamine-induced chronic lung injury through the gut–lung axis, providing a new theoretical and experimental basis for the prevention and treatment of MA-induced chronic lung injury.

Download video file ^{(61.9MB, mp4)}

Video Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s40168-026-02348-2.

Introduction

Methamphetamine (MA), also known as sampling ephedrine and commonly known as “crystal meth”, is a synthetic central nervous system stimulant. In recent years, the use of methamphetamine has been on the rise worldwide, resulting in a large number of morbidity and mortality, and is becoming a social health problem on a global scale [1–3]. Currently, numerous studies have concentrated on the neurotoxicity associated with MA [4–6]. But in fact, cardiopulmonary diseases such as cardiomyopathy, arrhythmia, acute pulmonary edema, and pulmonary hypertension are among the main causes of death associated with MA [7, 8]. Long-term use of MA is associated with chronic lung damage, pulmonary hypertension, atherosclerotic plaque formation, and cardiac arrhythmias [9]. However, the lack of effective pharmacological treatment strategies for MA addiction requires the development of new therapeutic approaches. It has been shown that MA causes enhanced inflammation and increased pro-inflammatory signaling driven by T cells and macrophages [10, 11]. Alveolar epithelial cells (AEC) act as important barrier cells and interact with inflammatory cells by secreting cytokines. Lucas F. Loffredo et al. showed that Treg cells and epithelial cells of a variety of tissues interact to promote tissue repair. CD4⁺CD25⁺ splenic T cells can rescue tissue injury after acute lung injury (ALI) [12]. The evidence suggested that the anti-inflammatory effect of Treg cells may be a breakthrough in MA-induced chronic lung injury. Furthermore, increasing evidence indicates that MA leads to gut barrier impairment and tissue damage, while also disrupting the homeostasis of gut microbiota. The gut microbiota is essential for the bidirectional communication between the gut and the lungs [13]. Disturbances in the gut microbiota have been observed in many lung disease models and drug abuse models [14, 15]. However, it remains unclear whether alterations in the gut microbiota have a causal relationship with MA-induced lung injury.

The genus Lactobacillus is a Gram-positive anaerobic bacterium, including Lactobacillus rhamnosus (LGG), Lactobacillus reuteri, Lactobacillus plantarum, Lactobacillus casei, etc. Among them, certain Lactobacillus strains act as probiotics and play an important role as natural immune probiotics. They have strong immune regulatory ability and anti-pathogen activity in diseases, showing powerful therapeutic effects [16]. More and more evidence indicates that certain Lactobacillus strains not only play a significant role in intestinal diseases but also have the potential to alleviate lung diseases. In preclinical studies, it was observed that oral administration of Lactobacillus rhamnosus CRL1505, Lactobacillus rhamnosus M21, and Lactobacillus rhamnosus could induce immune responses against respiratory syncytial virus, influenza virus, or pneumococcal infections, thereby enhancing pathogen clearance and reducing lung tissue damage [17–20]. It is widely recognized that probiotics can influence the onset and progression of various diseases through direct or indirect mechanisms. Therefore, further exploration of the specific mechanisms by which probiotics exert beneficial effects is particularly important.

In the tryptophan (Trp) metabolic pathway, tryptophan 2, 3-dioxygenase (TDO) and IDO2 metabolize Trp to L-KYN. Research indicates that AhR, activated by L-KYN, is present in both innate and adaptive immune cells and demonstrates anti-inflammatory effects in mice models [21, 22]. Gut microbiota can directly metabolize Trp to produce L-KYN, which is a ligand for AhR. AhR signaling is considered a key component of immune responses at barrier sites, acting on epithelial turnover, barrier integrity, and many immune cell types [23, 24]. It has been demonstrated that Lactobacillus administered with an AhR agonist ameliorated colitis in mice with dysbiosis [25], indicating therapeutic potential in disease. Another study showed that L-KYN-producing Lactobacillus promoted Treg cells’ proliferation through IDO1/Kyn/AHR [26].

Bidirectional communication between the lung and gut can be through various mechanisms, functions, and behaviors, including immune pathway [27–29]. Fecal microbiota transplantation (FMT) and probiotic therapies are widely employed to investigate the mechanisms by which microbiota influence disease pathophysiology, as well as to elucidate the causal relationship between disease and gut microbiota [30, 31]. Here, multi-omics analysis revealed gut microbiota disturbance and metabolic reprogramming, lung inflammation, lung tissue and gut barrier dysfunction, reduction of anti-inflammatory Treg cells, and decreased levels of probiotics Lactobacillus and L-KYN induced by MA. FMT reveals that gut microbiota imbalance and metabolic disorder are the key causes of lung injury. We further investigated the mechanism by which the key metabolite L-KYN regulates Treg cells’ differentiation, thereby inhibiting lung inflammation and damage to the alveolar epithelial barrier. This study aims to provide evidence and potential immunotherapy for MA-induced chronic lung injury from a gut-lung perspective.

Materials and methods

Animals

One hundred and forty-three male BALB/C mice (22 ± 0.2 g) were obtained from the Animal Resource Centre, China Medical University (certificate number: Liaoning SCXK 2022–0007). Each group contained 11 mice. All mice were housed in a temperature (20 ± 2 °C) and humidity (50 − 60%) controlled room, and water and food were provided with unrestricted access in an alternating 12-h light and 12-h dark cycle. All mice were allowed to acclimate for at least one week prior to experimentation. All animal experimental procedures comply with the guidelines of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH), with the approval of the Institutional Animal Care and Use Committee of China Medical University (IACUC Issue No. CMU2022271).

Mice models of chronic lung injury induced by MA

Mice were randomized into two groups based on body weight: control group (Con group) and MA group and were subjected to an escalating dose model of MA administration. In brief, the dose schedule of MA was as follows: 1.0 mg/kg for the first week, 2.0 mg/kg for the second week, 3.0 mg/kg for the third week, 4.0 mg/kg for the fourth week, 5.0 mg/kg for the fifth week, and 6.0 mg/kg for the sixth week, with each dose administered once daily. The mice in the Con group were intraperitoneally injected with the same volume of 0.9% normal saline.

FMT

Notably, all of the mice used in the FMT experiment were treated with an antibiotic mixture (200 mg/kg ampicillin, metronidazole, and neomycin, as well as 100 mg/kg vancomycin) 14 days in advance to eliminate the gut microbiota and create pseudo-germ-free mice (ABX group). Next, ABX mice were used as recipients for FMT for 6 weeks. In short, fresh fecal samples were collected from both Con and MA mice, and all fecal samples from the same group of mice were mixed as donors for fecal transplantation by gavage. Each mouse received 300 μL of fecal microbiota solution once per day for the first 3 days, then once per 2 days till the 6th week. ABX mice were randomized into three groups based on body weight: ABX group, FMT-C group, and FMT-M group. Feces from each group of mice were collected for subsequent experiments (The method for collecting feces is described in the supplementary materials.)

LGG transplantation and L-KYN administration in mice

LGG (ATCC 53103) was purchased from American Type Culture Collection and was cultivated according to the reference [32]. In the LGG transplantation experiment, mice were randomly divided into four groups based on body weight. The mice were divided into the following four groups: the Con group, the MA group, the MA + activated LGG group, and the MA + inactivated LGG group. According to the experimental design, mice were given different doses of the LGG by oral gavage, while the Con group was given the same dose of PBS.

In the L-KYN administration experiment, mice were randomized into four groups based on body weight: Con group, MA group, MA + L-KYN (Sigma-Aldrich, USA) with low dose at 15 mg/kg (MA + Low-L-KYN) and MA + L-KYN with high dose at 30 mg/kg (MA + High-L-KYN). Mice were administered by intraperitoneal injection at a dose of 15 or 30 mg/kg depending on the experimental design [26].

Joint analysis of multi-omics

16 s rDNA sequencing

Total genomic DNA of microorganisms was extracted, and microbiome analysis was performed at Hangzhou Lianchuan Biotechnology Co., LTD. (Hangzhou, China). The resulting sequence data, including α diversity, β diversity, principal coordinate analysis (PCoA) maps, heat maps, linear discriminant analysis binding effect size (LEfSe), and other metrics, were analyzed using the Unichuan Biocloud platform.

LC–MS/MS non-targeted metabolomics analysis

Fecal samples from each group were diluted at a ratio of 100 mg to 1 mL using a methanol–acetonitrile–water mixture in a 2:2:1 volume ratio. The mixture underwent centrifugation at 13,000 rpm and 4 °C for 15 min, after which 100 μL of the supernatant was transferred into a 2 mL EP tube for LC/MS analysis. Metabolic profiles were conducted using the Agilent 1290 Infinity LC system (Agilent, USA).

In vitro co-culture of AEC and Treg cells

AEC were divided into 6-well plates at a concentration of 5 × 10⁵ pieces/well. After incubation for 5 h, half of the medium was discarded, 1 × 10⁵ Treg cells were added to each well, and co-cultured at 37 ℃ and 5% CO₂. 1 µg/mL anti-IL-10RA (R&D Systems, USA) neutralizing antibody was added to AEC 1 h before the addition of Treg cells. Then co-culture was performed with or without anti-IL-10RA antibodies. After co-culture for 48 h, AEC-related indexes were detected.

In addition, AEC were divided into 6-well plates at a concentration of 5 × 10⁵ pieces/well. After incubation for 5 h, half of the medium was discarded, 1 × 10⁵ Treg cells were added to each well and co-cultured at 37 ℃ and 5% CO₂. The co-culture was performed with or without the JAK1 inhibitor ABT-494 (MCE, USA) [33]. After co-culture for 48 h, the JAK1/STAT3-related indicators were detected.

Statistical analysis

All the data are shown as means ± SD. GraphPad Prism 10.0 (GraphPad Software, Inc., San Diego, CA, USA) was used for statistical analysis, except for the 16sRNA sequencing and non-targeted metabolomics data. Student’ s t-test and ordinary one-way ANOVA with Tukey’s test were used for statistical comparisons. Mantel test and Kruskal–Wallis test were used for multi-omics analysis. Spearman’s two-tailed correlation was used for correlation analysis. P < 0.05 was considered to indicate statistical significance (^ns P > 0.05, ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001, ^**** P < 0.0001). All results are fully reproducible and were obtained consistently across multiple independent experiments, and n-values represent independent biological replicates.

Results

Re-programming of gut microbiota and metabolites induced by MA

In order to determine the causal relationship between the gut microbiota and lung injury caused by MA, we performed a multi-omics analysis of gut microbiota and its metabolites. No significant differences in diversity or richness indices between groups were observed for Chao α diversity (Fig. 1A). β diversity, as assessed by PCoA, indicated a difference in microbiota composition between the Con group and MA groups (Fig. 1B). MA administration altered gut microbial abundance. Community bars showed that a higher relative abundance of Actinobacteria was observed in the MA group at the phylum level, and MA exposure reduced the abundance map of Firmicutes (Fig. 1C). At the family level, MA reduced the abundance of Lactobacillaceae (Fig. 1D). Genus-based classification (Top30) showed that the abundance of Lactobacillus in the MA group was lower than the Con group (Fig. 1E). LEfSe differences plots depict that LGG has a higher abundance in the Con group but does not have a dominant presence in the MA group (Fig. 1F).

Fig. 1.

Re-programming of gut microbiota and metabolites induced by MA. A α diversity analysis of Chao index. B Beta diversity NMDS map. The analysis is performed at the OTU level using un-weighted unifrac. n = 6 per group. C Comparison of microbiota composition between the two groups at the phylum level. D Comparison of microbiota composition between the two groups at the family level. E Heat maps of microbiota from phylum to genus top30 (LDA value ≥ 3, all versus all). F LEfSe bar chart from phylum to genus (LDA value ≥ 3, all versus all). G Two sets of volcanic maps. H Bubble diagram of KEGG enrichment pathway. I KEGG metabolic pathway enrichment histogram. J Relative abundance of metabolites of Trp metabolic pathway. K Correlation analysis of Trp metabolites and significantly different microbiota. The Kruskal–Wallis test and Spearman were used for microbial analysis. Other data were analyzed using the unpaired Student’s t test. Data were presented as mean ± SD. Statistical significance was defined as ^ns P > 0.05, and n-values represent independent biological replicates

Next, we performed untargeted metabolomics assays on feces. The volcano map indicated that the composition of fecal metabolites was altered after MA treatment (Fig. 1G). KEGG enrichment results showed that the pathways related to Metabolism were the most enriched (Fig. 1H). KEGG functional prediction indicated that it was related to pathways such as Tyrosine metabolism, Trp metabolism, and Steroid hormone biosynthesis (Fig. 1I). Among them, Trp metabolism ranked high and the number of enriched genes was high. Furthermore, we screened all metabolites of the Trp metabolic pathway and showed that l-kynurenine and 5-hydroxyindole-3-acetic acid were down-regulated in the MA group, while other metabolites were up-regulated (Fig. 1J). The findings indicated a significant positive correlation between L-KYN and Lactobacillus (Fig. 1K). These results suggested that L-KYN is a key omics factor for MA abuse and plays a synergistic role with Lactobacillus.

MA abuse induces lung and gut dysfunction, Treg imbalance, and L-KYN abnormality

To assess the effect of MA on lung function, we first established a model of MA-induced chronic lung toxicity (Fig. 2A). PAVpeak, TAPSE, and RVHI are three commonly used indices for pulmonary function assessment. Compared with the Con group, PAVpeak and RVHI increased, whereas TAPSE decreased in MA-treated mice on echocardiography (all P < 0.0001; Fig. 2B-E). HE staining showed a decrease in the number of alveolar sacs and a thickening of alveolar walls after exposure to MA treatment (P < 0.0001; Fig. 2F, G). HE staining of the colon showed that inflammatory cell infiltration increased, the number of goblet cells decreased, and the overall structure of colon tissue was significantly destroyed (P < 0.0001; Fig. 2F, H). Similarly, MA treatment increased the levels of inflammatory cytokines in the lung, including interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), and IL-6 (all P < 0.0001; Fig. 2I-K). The levels of Zona Occludens 1 (ZO-1) and Epithelial cadherin (E-cadherin) in the MA group were significantly lower than those in the Con group, while the levels of Vimentin were significantly higher than those in the Con group (P < 0.0001; P = 0.0004; P = 0.0015; Fig. 2L-O), indicating lung inflammation and alveolar barrier damage caused by MA. We measured the number of inflammation-suppressing Treg cells in the lung, and the results showed a significant decrease in the MA group compared to the Con group (P < 0.0001; Fig. 2P, Q). Next, we detected L-KYN content in the lung, colon, and plasma, which was consistent with the sequencing results, all of which showed that L-KYN content was reduced in the MA treatment group compared with the Con group, indicating that MA caused the reduction of L-KYN level (P < 0.0038; P < 0.0001; P = 0.0001; Fig. 2R-T). And Spearman correlation analysis showed that L-KYN was negatively correlated with PAVpeak in lung tissues, colon tissues, and plasma of mice in the MA-induced chronic lung toxicity model (R² = 0.6703, P < 0.0038; R² = 0.9111, P < 0.0001; R² = 0.8538, P = 0.0001; Fig. 2U-W). It suggested that the decrease of L-KYN was negatively correlated with lung function indexes, suggesting the protective role of L-KYN in MA lung injury.

Fig. 2.

MA abuse induces lung and gut dysfunction, Treg imbalance, and L-KYN abnormality. A Experimental timeline. B–D PAV peak, TAPSE, and statistical analysis. n = 5. E RVHI. n = 5. F–H HE staining of lung tissue and colon tissue: The number of alveolar sacs in lung tissue was quantified, n = 5–6; Quantification of goblet cells in each crypt of colon tissue. n = 6. (Magnification, × 400; scale bars, 50 μm). I–K Lung tissue inflammatory cytokines TNF-α, IL-6, IL-1β levels. n = 5. L–O Levels of ZO-1, E-cadherin, Vimentin in lung tissue and statistical analysis. n = 5. P, Q Treg cells' content and statistical map in lung tissue. n = 3. R–T L-KYN content in lung, colon, and plasma. n = 5. U–W Correlation analysis of L-KYN and PAVpeak in lung, colon, and plasma tissues. Data were analyzed using the unpaired Student’s t-test and were expressed as mean ± SD. Spearman’s two-tailed correlation was used for U to W. Statistical significance was defined as ^nsP > 0.05, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, and n-values represent independent biological replicates

L-KYN, a metabolite of Lactobacillus, may be a key omics factor for MA abuse by FMT

To search for the key microbiota associated with MA and lung inflammation, we analyzed the gut microbial community of mice after receiving FMT. Analysis of α diversity and β diversity revealed significant differences in microbial communities between groups following colonization (Fig. 3A, B). Analysis of community bar plots indicated differences in microbial composition at the phylum levels across the groups. At the phylum level, FMT-M treatment reduced the relative abundance of Firmicutes and Patescibacteria and increased the relative abundance of Bacteroidetes (Fig. 3D). In addition, at the genus level, after FMT-M treatment, Lactobacillus showed a down-regulated trend (Fig. 3C). LEfSe differences plots indicated that the abundance of LGG was higher in the FMT-C group, but it did not have a dominant position in the FMT-M group either (Fig. 3E). To better understand the key gut microbiota metabolites associated with MA, we performed non-targeted metabolomic assays in stool. KEGG enrichment analysis highlighted Phenylpropanoid biosynthesis, Trp metabolism, ABC transporters, and other 20 related pathways (Fig. 3F). Consistent with the previous result, it was found that Trp metabolism was highly enriched (Fig. 1I). FMT-M reduced the set of Trp metabolites such as 3-methyldioxyindole, 4-hydroxyphenylpyruvic acid, L-KYN, and 5-hydroxyindole-3-acetic acid (Fig. 3J). On the contrary, others were decreased in FMT-M (Fig. 3G). The correlation analysis indicated that L-KYN was significantly positively correlated with Lactobacillus (Fig. 3H). The results further suggested that Lactobacillus and L-KYN may be involved in the pathogenesis of chronic lung toxicity caused by MA.

Fig. 3.

L-KYN, the metabolite of Lactobacillus, may be a key omics factor for MA abuse by FMT. A α diversity analysis of Chao index in different treatment groups. B β diversity in different treatment groups. The analysis is performed at the OTU level using unweighted unifrac. n = 6 per group. C Heatmap of the top 30 differential microbiota at the genus level. D Comparison of microbiota composition between treatment groups at phylum level. E LEfSe bar chart from phylum to genus (LDA value ≥ 3, all versus all). F KEGG metabolic pathway enrichment histogram. G Relative abundance of metabolites of Trp pathway. H Correlation analysis of Trp metabolites and significantly different microbiota. Spearman’s two-tailed correlation was used for U to W. The Kruskal–Wallis test and Spearman were used for microbial analysis. Other data were analyzed using the ordinary one-way ANOVA with Tukey’s test. Data were presented as mean ± SD. Statistical significance was defined as ^nsP > 0.05, ^**P < 0.01, and n-values represent independent biological replicates

FMT from MA-treated mice caused lung and gut dysfunction, and L-KYN abnormality

We subsequently investigated whether the transplantation of fecal microbiota from MA-treated mice into healthy mice could induce MA-like chronic lung injury and compromised gut barrier function (Fig. 4A). After FMT, FMT-M echocardiography in mice showed an increase in PAV peak and RVHI, a decrease in TAPSE (P = 0.0002 vs. ABX; P < 0.0001 vs. FMT-C; all P < 0.0001; Fig. 4B–E). After the FMT-M, the number of alveolar sacs decreased, and the thickness of the alveolar walls increased (P < 0.0001 vs. ABX; P = 0.0004 vs. FMT-C; Fig. 4F, G). Inflammatory cell infiltration increased, the number of goblet cells decreased, and the overall structure of colon tissue was significantly destroyed in the colon (all P < 0.0001; Fig. 4F, H). TNF-α, IL-6, and IL-1β were overexpressed in FMT-M (all P < 0.05; Fig. 4I–K). The levels of ZO-1 and E-cadherin in the FMT-M group were considerably lower than those in the ABX and FMT-C groups, while the levels of Vimentin were much higher than those in the ABX and FMT-C groups (P = 0.0005 vs. ABX, P = 0.0035 vs. FMT-C; P < 0.0001 vs. ABX, P < 0.0001 vs. FMT-C; P < 0.0001 vs. ABX, P < 0.0001 vs. FMT-C; Fig. 4L–O). In addition, L-KYN in the FMT-M group was significantly lower than that in the ABX and FMT-C groups (P < 0.0001; Fig. 4P–R). Spearman correlation showed that L-KYN in the lungs, colon, and plasma was negatively correlated with PAV peak, indicating that the decrease of kynurenine was negatively correlated with the breakdown of lung function (R² = 0.6946, P = 0.0001; R² = 0.7840, P < 0.0001; R² = 0.7607, P < 0.0001; Fig. 4S–U). Our results showed that FMT-M caused MA-like chronic lung injury and intestinal barrier damage, and L-KYN may be a key metabolite of MA-induced chronic lung injury.

Fig. 4.

FMT from MA-treated mice caused lung and gut dysfunction, and L-KYN abnormality. A Experimental timeline. B–D PAV peak, TAPSE, and statistical analysis. n = 5. E RVHI. n = 5. F–H Lung and colon tissues were stained with HE, and the number of alveolar sacs in the lung was quantified, n = 6 per group. Quantification of goblet cells in each crypt of colon tissue. n = 5–6. (× 400 magnification; scale bars,50 μm). I–K Lung tissue inflammatory cytokines TNF-α, IL-6, IL-1β levels. n = 5. L–O Levels of ZO-1, E-cadherin, Vimentin in lung tissue and statistical analysis. n = 5. P–R L-KYN levels in lung tissue, colon, and plasma. n = 5. S–U Correlation between L-KYN and PAV peak in lung tissue, colon tissue, and plasma tissue. Data analyzed using the ordinary one-way ANOVA with Tukey’s test. Data were presented as mean ± SD. Statistical significance was defined as ^ns P > 0.05, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001. n-values represent independent biological replicates

LGG ameliorates MA-induced chronic lung injury

Given the substantial reduction in LGG abundance observed in the MA-induced model, we subsequently investigated the therapeutic potential of LGG supplementation in ameliorating MA-induced chronic lung injury. Therefore, we administered activated and inactivated LGG to mice in the MA groups via oral gavage and evaluated the therapeutic effects (Fig. 5A). The level of L-KYN in the activated LGG medium was significantly higher than that in the inactivated LGG medium (P < 0.0001; Fig. 5B). Compared to the MA group, the MA + activated LGG group and the MA + inactivated LGG group showed reduced infiltration of inflammatory cells in the lung tissue and decreased thickening of the alveolar walls (P < 0.0001; Fig. 5C-I). Moreover, compared with the MA + inactivated LGG group, the MA + activated LGG group demonstrated a greater improvement in MA-induced lung injury (Fig. 5C-I).

Fig. 5.

LGG ameliorates MA-induced chronic lung injury. A Experimental timeline. B L-KYN levels in activated LGG and inactivated LGG. n = 5. C–E PAVpeak, TAPSE, and statistical analysis. n = 5. F RVHI. n = 5. G–I Lung tissue was stained with HE; the number of pulmonary vesicles and the histopathological score were quantified. n = 5–6 (magnification, × 400, scale bars, 50 μm). J–L Lung tissue inflammatory cytokines TNF-α, IL-6, IL-1β levels. n = 5. M–P Levels of ZO-1, E-cadherin, Vimentin in lung tissue and statistical analysis. n = 5. Q–S Absolute Quantification of L-KYN levels in lung, colon, and plasma. n = 5. Data were analyzed by ordinary one-way ANOVA with Tukey’s test. Data were presented as mean ± SD. Statistical significance was defined as ^ns P > 0.05, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, and n-values represent independent biological replicates

Next, we found that compared with the MA group, activated LGG (all P < 0.0001; Fig. 5J-L) and inactivated LGG (P < 0.0001; P = 0.0210; P = 0.0035; Fig. 5J-L) treatment could effectively reduce the expression of TNF-α, IL-6, and IL-1β. The expression of ZO-1 and E-cadherin was increased, while the expression of Vimentin decreased in the MA + activated LGG (P < 0.0004; P < 0.0001; P = 0.0051; Fig. 5M-P). However, the levels of ZO-1 and Vimentin were not improved in the MA + inactivated LGG group (all P > 0.05; Fig. 5M-P). In the MA + activated LGG group and the MA + inactivated LGG group, the L-KYN level in the lung (all P < 0.0001), colon (all P < 0.0001), and plasma (P < 0.0001 vs MA + activated LGG; P = 0.0201 vs MA + inactivated LGG) were significantly increased (Fig. 5Q-S). Meanwhile, compared with the MA + inactivated LGG group, the MA + activated LGG group showed higher levels of L-KYN in the lung (P < 0.0001), colon (P < 0.0001), and plasma (P = 0.0015; Fig. 5Q-S). These results suggested that activated LGG is more effective in alleviating MA-induced chronic lung injury.

In vivo L-KYN ameliorates MA-induced chronic lung injury

Next, we investigated whether L-KYN could ameliorate MA-induced lung inflammation and alveolar epithelial barrier damage in vivo (Fig. 6A). After L-KYN treatment, the results showed that a low dose of L-KYN was unable to alleviate the chronic lung injury induced by MA (all P > 0.05; Fig. 6B-O). In contrast, in the MA + high-L-KYN group, the cardiopulmonary function of the mice was improved. The PAVpeak and RVHI decreased, and the TAPSE was increased (P < 0.0001; P < 0.0001; P = 0.0022; Fig. 6B-E). Compared with the MA group, the number of alveolar sacs increased in the MA + high-L-KYN group, and the histopathological score decreased in the MA + high-L-KYN group (P = 0.0005; P < 0.0001; Fig. 6F, H). In addition, the three inflammatory cytokines’ levels decreased (all P < 0.0001; Fig. 6I–K) and damage to the alveolar epithelial barrier (P = 0.0430; P = 0.0204; P = 0.0007; Fig. 6L–O) was reversed by high doses of L-KYN intervention. Together, these findings highlighted that L-KYN exerts a protective effect in vivo to ameliorate MA-induced lung inflammation and alveolar epithelial barrier damage in a dose-dependent manner.

Fig. 6.

In vivo L-KYN ameliorates MA-induced chronic lung injury. A Experimental timeline. B–D PAVpeak, TAPSE, and statistical analysis: n = 5 (E) RVHI. n = 5. F–H Lung tissue was stained with HE; the number of pulmonary vesicles and the histopathological score were quantified. n = 5–6 (magnification, × 400, scale bars, 50 μm). I–K The levels of inflammatory cytokines TNF-α, IL-6, IL-1β in lungs by ELISA. n = 5. L–O Protein levels of ZO-1, E-cadherin, Vimentin in lungs. n = 5. Data was analyzed by ordinary one-way ANOVA with Tukey’s test. Data was presented as mean ± SD. Statistical significance was defined as ^nsP > 0.05, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, and n-values represent independent biological replicates

L-KYN facilitates Treg cells’ differentiation to reshape the immune microenvironment

To further explore the potential mechanism of L-KYN in improving lung inflammation, we investigated the process and possible mechanism of L-KYN promoting T lymphocyte differentiation into Treg cells. Firstly, we extracted naïve CD4⁺ T cells and determined that 50 µM was the optimal concentration for stimulating the differentiation of Treg cells with L–KYN (Fig. 7A, B). After 50 µM L-KYN treatment, Treg cells were increased by 6.25 times and the IL-10 concentration increased by 20.2 times (all P < 0.0001; Fig. 7C, D). The relative protein expression of FOXP3 and AhR receptors increased significantly (all P < 0.0001; Fig. 7E–G). These results demonstrated the potent promotion effect of L-KYN on Treg differentiation via AhR.

Fig. 7.

L-KYN induced Treg cells’ differentiation to secrete IL-10. A Gate strategy of flow cytometry. B Flow cytometry was used to detect naive T lymphocytes differentiated into Treg cells. C Level of CD4⁺CD25⁺FOXP3⁺ Treg cells’ statistical analysis. n = 3. D Quantitative analysis of IL-10 levels in Treg differentiation medium. n = 5. E–G FOXP3 and AhR receptor expression and statistical analysis. n = 5. Data was analyzed by ordinary one-way ANOVA with Tukey’s test. Data was presented as mean ± SD. Statistical significance was defined as ^nsP > 0.05, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, and n-values represent independent biological replicates

L-KYN inhibits MA-induced epithelial injury via IL-10-mediated Treg–AEC communication

To elucidate the regulatory mechanism of L-KYN on MA-induced alveolar inflammation and alveolar epithelial injury, L-KYN-induced Treg cells were co-cultured with AEC for 24 h to evaluate the expression of inflammatory factors and barrier proteins (Fig. 8A). First, AEC were treated with MA (0 mM,0.1 mM,0.5 mM,1 mM,5 mM) for 6, 12, 24, and 48 h [34]. Among them, the concentration of 1 mM MA for 24 h was determined to be optimal for subsequent investigation (Fig. 8B).

Fig. 8.

L-KYN mediated Treg–AEC communication via IL-10/JAK1/STAT3 pathway. A Co-culture process. B CCK8 experiment. C–F Statistical analysis of ZO-1, E-cadherin, Vimentin, IL-10RA levels in the different groups. n = 5. G–I Levels of inflammatory cytokines TNF-α, IL-6, and IL-1β in AEC. n = 5. J IL-10RA levels in the different groups. n = 5. K–M Representative immunofluorescence staining (× 200 magnification; scale bar, 50 μm) and the average median fluorescent intensity (MFI) of ZO-1 and E-cadherin performed for the different groups. n = 8. N–R Protein levels of p-JAK/JAK1, p-JAK1, p-STAT3/STAT3, p-STAT3 in the different groups and statistical analysis. n = 5. S–U Protein levels of p-JAK1, p-STAT3 in the different groups and statistical analysis. n = 5. Data was analyzed by ordinary one-way ANOVA with Tukey’s test. Data was presented as mean ± SD. Statistical significance was defined as ^nsP > 0.05, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001 and n-values represent independent biological replicates

Compared with AEC without any processing, 1 μg/mL anti-IL-10RA neutralizing antibody alone didn't cause the elevated levels of inflammatory factors and epithlial barrier damage (All P > 0.05; Fig. 8C–I). Compared with the MA-treated AEC group, the protein expression of E-cadherin and ZO-1 was significantly increased, and the protein expression of Vimentin showed a decreasing trend in the co-culture of MA-treated AEC and L-KYN-stimulated differentiated Treg cells for 24 h (P = 0.0224; P = 0.0002; P = 0.0332; Fig. 8C–F). Meanwhile, the levels of TNF-α, IL-6, and IL-1β were significantly reduced (all P < 0.0001; Fig. 8I). After co-culture with L-KYN-induced Treg cells, the relative level of IL-10RA was higher than that of the MA-treated AEC group. However, its expression was significantly decreased after anti-IL-10 treatment (all P < 0.0001; Fig. 8J). Correspondingly, immunofluorescence staining also showed that the contents of ZO-1 and E-cadherin in AEC treated with anti-IL-10 were significantly reduced (all P < 0.0001; Fig. 8K–M). It suggested that Treg cells can significantly inhibit the barrier damage of AEC and inhibit the secretion of inflammatory factors in vitro. However, the above effect could be blocked by the IL-10 neutralizing antibody. Compared with the co-culture of MA-treated AECs with L-KYN-induced Treg cells, the addition of an IL-10 neutralizing antibody to this co-culture system resulted in elevated levels of inflammatory factors (all P < 0.0001), and Vimentin levels were elevated (P < 0.05), and ZO-1 (P < 0.05), E-cadherin (P < 0.001), and IL-10RA levels were reduced (P < 0.0001; Fig. 8C–J). Immunofluorescence corroboration was performed for ZO-1 and E-cadherin (all P < 0.0001; Fig. 8K–M).

Next, we detected changes in JAK1 and STAT3 levels in AEC. Compared with AEC without any processing, the addition of IL-10-neutralizing antibodies to AEC did not cause changes in the JAK1/STAT3 pathway (All P > 0.05; Fig. 8N–R). Compared with the MA-treated AEC group, the JAK1/STAT3 pathway was activated in MA-treated AEC and co-cultured with L-KYN-induced Treg cells, and the expression of p-JAK1/JAK1, p-JAK1, p-STAT3/STAT3, and p-STAT3 was significantly elevated in AEC (P = 0.0075; P = 0.0003; P = 0.0046; P = 0.0017; Fig. 8N–R). The above effects could be blocked by the IL-10 neutralizing antibody. After adding an IL-10 neutralizing antibody to the co-culture system of AEC treated by MA and Treg cells induced by L-KYN, p-JAK1/JAK1, p-JAK1, p-STAT3/STAT3, and p-STAT3 were significantly reduced (P = 0.0007; P < 0.0001; P = 0.0004; P < 0.0001; Fig. 8N–R), indicating that the JAK1/STAT3 pathway was inhibited.

To determine whether the JAK1/STAT3 pathway is involved in MA-induced alveolar inflammation and epithelial barrier injury, we used the JAK1 inhibitor ABT-494 to suppress this pathway. We found that t 2 μM ABT-494 alone markedly reduced the levels of JAK1, p-JAK1, STAT3, and p-STAT3 in AECs. Compared with the MA-treated AEC group, MA-treated AEC co-cultured with L-KYN-induced Treg cells, the levels of p-JAK1 and p-STAT3 elevated (all P < 0.0001; Fig. 8S–U). This activation was effectively blocked by ABT-494: the addition of ABT-494 to the co-culture system resulted in lower p-JAK1 and p-STAT3 levels compared with the co-culture group without the inhibitor (all P < 0.0001; Fig. 8S–U). These findings demonstrated that L-KYN promotes Treg cell differentiation and enhances IL-10 secretion, which in turn activates the JAK1/STAT3 signaling pathway, thereby exerting an immunoprotective effect in MA-induced chronic lung injury.

Discussion

Overall, we provide evidence that L-KYN mediates inter-Treg–AEC communication by stimulating Treg cells’ differentiation to secrete more IL-10, and mechanistically, IL-10 activates the JAK1/STAT3 signaling pathway to inhibit lung inflammation and alveolar epithelial barrier damage.

It is established that MA causes damage to the lungs and results in changes in cardiopulmonary function [35]. Our results showed that MA not only induced an inflammatory response, alveolar epithelial barrier damage, and right heart hypertrophy and other cardiopulmonary dysfunction, but also damaged the gut barrier and altered the composition of the gut microbiota. A significant decrease in the relative abundance of Lactobacillus was observed. It demonstrates that MA may impair gut function and lead to microbial homeostasis dysregulation. At the same time, decreased levels of L-KYN in feces, lung tissue, colon tissue, and plasma were accompanied by a decrease in the relative abundance of Lactobacillus metabolizing L-KYN, suggesting a possible synergistic effect of L-KYN and Lactobacillus in MA-induced chronic lung injury. Moreover, dysbiosis of the gut microbiota and the resulting metabolite imbalance can contribute to chronic lung injury and intestinal barrier impairment.

FMT and probiotics serve as prevalent approaches to manage various disorders by targeting the gut microbiome. The FMT experiments indicated that while the microbiome structure of recipient mice did not precisely mirror that of donor mice, notable similarities were observed in the alterations of the primary bacterial taxa. These observations indicate that the colonization process was successful and confirm the effectiveness of FMT or probiotic therapy. LGG, a well-known probiotic, has been demonstrated to improve osteoporosis by regulating Th17/Treg balance and gut microbiota composition [32], activate hepatic Nrf2 and prevent oxidative liver injury [36]. We found it reduces inflammatory infiltration and repairs MA-induced chronic lung injury. Supplementation with activated LGG more effectively ameliorates MA-induced lung tissue inflammation and epithelial barrier damage. This may be attributed to the fact that inactivated LGG, while retaining the structural integrity and certain components of the bacteria, lacks metabolic activity and the capacity for proliferation, and thus cannot produce sufficient L-KYN. Therefore, probiotic LGG and its metabolite L-KYN are essential factors for improving chronic lung injury caused by MA.

Previous research has demonstrated that gut microbiota plays crucial roles in shaping, regulating, and maintaining healthy immune responses [37, 38]. Notably, microbial metabolites are particularly important in the interplay among the gut microbiome, gut metabolome, and immune therapeutic outcomes [39–42]. Short‑chain fatty acids (SCFAs) and tryptophan metabolites, for example, protect lung function by modulating immune homeostasis, enhancing intestinal barrier integrity, and suppressing pathogen translocation. Butyrate can cross the intestinal wall into the systemic circulation, not only regulating the differentiation and maturation of immune cells in the bone marrow but also reaching the lungs, where it directly regulates Treg cells' activity via G‑protein‑coupled receptors (GPCRs) [43, 44]. Trp metabolites derived from symbiotic bacteria can strengthen the skin barrier and alleviate the inflammation in patients with atopic dermatitis by activating the aryl hydrocarbon receptor [45]. Supplementation with key metabolites can ameliorate microbiota alterations induced by disease and attenuate lung injury. Here, our experiments revealed that L-KYN regulated immune cells and cytokines, activated immune-related pathways, and ameliorated MA-induced chronic lung injury. Studies have reported that the abuse of MA leads to a decrease in the level of Treg cells [35]. We found that MA‑treated mice exhibited elevated systemic levels of pro‑inflammatory cytokines and a marked reduction in anti‑inflammatory Treg cells. Furthermore, our study revealed that L‑KYN mitigates MA‑induced chronic lung injury through immune pathways—by modulating immune cells and cytokines, activating immune‑related signaling, and ultimately alleviating lung injury. In the present study, the lower dose showed limited therapeutic efficacy, whereas the higher dose produced a more pronounced benefit, suppressing relevant inflammatory factors to levels close to those of controls and restoring expression of epithelial barrier markers. A large number of studies have confirmed that L-KYN binds to the AhR receptors on Treg cells, thereby activating the proliferation and differentiation of Treg cells and exerting its immune regulatory function [26, 46, 47]. Based on in vitro binding assays, 50 µM L‑KYN optimally promoted Treg cells' proliferation via the AhR receptor and ameliorated alveolar epithelial barrier injury at the cellular level, underscoring its therapeutic potential. A limitation of this work is that validation was performed only in mice. Future studies should further verify the efficacy and safety of this dosage in larger-scale trials and examine potential functional or morphological effects on other major organs (heart, liver, kidney) over an extended observation period. Future studies should further verify the efficacy and safety of this dosage in larger‑scale trials over an extended observation period.

JAK/STAT serves as a fundamental component of cell signaling, regulating both physiological and pathological processes, including inflammation and stress. Given its established association with immune dysregulation and impaired immune processes, it is pertinent to consider its role in the present context [48, 49]. Mechanistically, in MA disease models, MA causes inflammation, epithelial barrier damage, and apoptosis in neuronal cells, epithelial cells, endothelial cells, and many other cells [50, 51]. We therefore hypothesize that JAK1/STAT3 plays a protective role in AEC. A recent study has demonstrated that IL-10 attenuates perihematomal edema after intracerebral hemorrhage (ICH) through IL-10 receptor/JAK1/STAT3 signaling in central nervous system diseases [52]. These findings supported our hypothesis that MA reduces L-KYN levels, thereby decreasing the abundance of anti-inflammatory Treg cells and mediating inflammation and alveolar barrier damage via IL-10/JAK1/STAT3 signaling. This result demonstrated that L-KYN may be a promising target for treating chronic lung damage caused by MA. Nevertheless, emerging evidence suggests that gut microbiota-derived metabolites may ameliorate disease through multiple mechanisms beyond immune modulation, implying the potential involvement of additional pathways [53, 54]. Furthermore, impaired or suppressed autophagy is known to precipitate the development and progression of tissue inflammation. Considering prior reports that methamphetamine induces oxidative stress and dysregulation of autophagy- and apoptosis-related pathways, and that gut microbiota-derived metabolites can regulate the mitochondrial metabolism of Treg cells [55–57], the additional molecular mechanisms underlying the protective effects of L-KYN and other Lactobacillus-derived mediators warrant further investigation.

In brief, the results of both in vivo and in vitro studies demonstrated that the Lactobacillus metabolite L-KYN interacts with AhR receptors on CD4⁺ T cells, promoting Treg cells’ differentiation and enhancing IL-10 production. This process contributes to protection against MA-induced lung inflammation and alveolar barrier damage via immune remodeling, potentially elucidating a mechanism of the gut-lung axis. These results highlight L-KYN supplementation as a highly promising therapeutic intervention.

However, there are still many unknown mechanisms to be explored. Our future studies will require the use of Treg cell‑deficient mice to better determine the role of T lymphocytes in the MA model. In addition, MA-induced chronic lung injury models have only been explored at the animal level. In the future, further studies are needed to investigate the changes in L-KYN concentration in clinical MA cases and the protective effect of L-KYN on the disease. In particular, the mechanisms by which L-KYN mediates other molecular pathways or components of chronic lung injury species remain unclear. Therefore, answering these questions is the focus of our future research.

Conclusions

Taken together, MA disrupts the structure of the gut microbiota and triggers the reprogramming of metabolites, and MA abuse reduced the levels of Lactobacillus and its end metabolite, L-KYN. L-KYN was closely related to MA-induced lung injury. The protective effects of L-KYN against MA-induced lung injury were confirmed by the experiment in vivo. L-KYN induced differentiation of Treg cells from CD4⁺ T cells. L-KYN induced the secretion of IL-10 from Treg cells and mediated the communication between Treg cells and AEC via IL-10. L-KYN alleviated lung inflammation and damages of the alveolar barrier induced by MA through the IL-10/JAK1/STAT3 pathway. This study not only demonstrates that L-KYN is the key omics characteristic factor of MA-induced lung injury, but also elucidates the new mechanism by which L-KYN, the metabolic product of Lactobacillus rhamnosus, reshapes the immune microenvironment and mediates the communication between Treg–AEC to regulate MA-induced injury from the perspective of the regulation of gut microbiota–metabolites–immune network. The present work establishes an updated theoretical and experimental foundation for the prevention and treatment of MA-induced chronic lung injury.

Authors’ contributions

Pei-Jun Ma and Yun Wang contributed to the study conception and design. Material preparation, data collection and analysis were performed by Pei-Jun Ma, Ming Li, Ying Pan, Wei-Ting Hu, Ding Yang, Ye-Kui Liang, Lei Chen and Xin Wang. The draft of the manuscript was written by Pei-Jun Ma and Ming Li. Yun Wang and Ying Pan revised the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. 81973404) and the Medical Education Research Project of Liaoning Province (No. 2024-N012-07 (YJXM-202448)).

Data availability

The data in this study will be accessible with the following link [http://www.ncbi.nlm.nih.gov/bioproject/1268471](http:/www.ncbi.nlm.nih.gov/bioproject/1268471) (BioProject: PRJNA1268471) and [https://www.cncb.ac.cn/](https:/www.cncb.ac.cn) (Metabolomics Project: PRJCA041728).

Declarations

Ethics approval and consent to participate

All laboratory procedures involving animals follow the guidelines of the National Institutes of Health (NIH) Guidelines for The Care and Use of Laboratory Animals and are guided by the China Medical University Institutional Animal Care and Use Committee (IACUC Issue No. CMU2022271).

Consent for publication

Its publication is approved by all the authors.

Competing interests

The authors declare no competing interests.