Gut microbiota-associated leucine elevation promotes cold-induced atherosclerotic plaque formation and instability

Abstract

Background

Environmental factors such as cold exposure have been increasingly recognized as contributors to atherosclerosis progression, yet the underlying mechanisms linking environmental stress to vascular pathology remain incompletely understood. In particular, the role of the gut microbiota and microbiota-associated metabolites in cold-induced atherosclerosis has not been fully elucidated.

Methods

A cold exposure model was established in ApoE⁻/⁻ mice fed a Western diet. Integrated multi-omics analyses were combined with fecal microbiota transplantation (FMT) and mechanistic cellular assays to investigate gut microbiota remodeling, metabolic alterations, and immune regulation during cold-induced atherosclerosis.

Results

In this study, we demonstrate that cold exposure accelerates atherosclerotic plaque growth and instability in parallel with pronounced gut microbiota dysbiosis and alterations in host metabolic profiles. FMT combined with metabolomic analyses showed that cold-associated gut microbiota is closely associated with elevated circulating leucine levels, suggesting that cold-induced microbial remodeling may participate in this process by modulating host systemic leucine availability. Mechanistically, increased leucine suppressed the Zic family member 2 (Zic2) in macrophages, leading to reduced expression of growth arrest–specific 6 (Gas6), a key mediator of efferocytosis. Impaired Gas6-dependent efferocytosis resulted in defective clearance of apoptotic cells, heightened vascular inflammation, and increased plaque instability. Importantly, supplementation with Lactobacillus johnsonii, a commensal bacterium depleted under cold exposure, normalized circulating leucine levels, restored Zic2-Gas6 signaling, enhanced macrophage efferocytosis, and attenuated atherosclerotic plaque progression. Conversely, leucine supplementation recapitulated the effects of cold exposure on plaque development and instability.

Conclusions

Collectively, these findings identify a previously unrecognized cold-microbiota-leucine-Zic2-Gas6-efferocytosis axis that links environmental stress to atherosclerosis progression. Targeting gut microbial regulation of host systemic leucine levels, including L. johnsonii–based interventions, may represent a promising therapeutic strategy for preventing cold-induced atherosclerotic cardiovascular disease.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08492-5.

Background

Atherosclerotic cardiovascular disease (ASCVD) is a leading cause of morbidity and mortality worldwide [1–3]. In addition to the traditional risk factors such as tobacco use, hyperlipidemia, hypertension, diet, and physical activity, cold exposure or low ambient temperature has been identified as a significant environmental risk factor for ASCVD [4–6]. Epidemiological studies have shown that exposure to cold temperatures can lead to increased blood pressure, enhanced sympathetic nervous system activity, and elevated levels of circulating catecholamines, all of which contribute to the acceleration of atherosclerotic plaque formation and instability [7]. Furthermore, cold exposure has been linked to increased incidence of acute coronary syndromes (ACS) and other cardiovascular events, particularly in vulnerable populations such as the elderly and those with pre-existing cardiovascular conditions [8]. Despite these observations, the molecular mechanisms underlying the heightened risk of ASCVD associated with cold exposure remain incompletely understood.

Accumulating evidence suggests that the gut microbiota plays a pivotal role in modulating host cardiometabolic health [9–11]. Microbial dysbiosis has been implicated in the progression of atherosclerosis through alterations in immune responses, lipid metabolism, and the production of bioactive metabolites such as trimethylamine-N-oxide (TMAO) [12], short-chain fatty acids (SCFAs) [13], and secondary bile acids (SBAs) [14]. Among branched-chain amino acids (BCAAs), leucine has attracted growing interest due to its pivotal roles in regulating immune and metabolic functions [15]. Leucine is increasingly recognized as an immunometabolic cue that shapes macrophage programs involved in inflammatory activation and cellular homeostasis, processes that are intimately linked to vascular inflammation and atherosclerotic progression [16]. Among the key functions of macrophages, efferocytosis is essential for maintaining tissue homeostasis and resolving inflammation. Efficient efferocytosis plays a critical role in maintaining atherosclerotic plaque stability by facilitating the clearance of apoptotic cells and preventing secondary necrosis, thereby limiting inflammation and necrotic core formation within lesions [17]. Despite these insights, it remains unclear whether environmental stressors such as cold exposure can influence BCAA metabolism via gut microbiota alterations and how this contributes to atherosclerosis pathogenesis.

In this study, we sought to elucidate the mechanisms by which cold exposure exacerbates atherosclerosis via the gut-metabolism-immunity axis. Using a cold-exposed ApoE^−/− mouse model fed a Western diet, we demonstrate that cold exposure significantly accelerates atherosclerotic plaque growth and instability. Through fecal microbiota transplantation, metabolomic analyses, and targeted molecular assays, we uncover a previously unappreciated cold-microbiota-leucine-Zic2-Gas6 regulatory pathway that mediates macrophage efferocytosis impairment and vascular inflammation. Notably, we show that oral supplementation with L. johnsonii partially restores Gas6 expression, enhances efferocytosis, and mitigates plaque progression. These findings offer new insights into the microbiota-metabolite-immune mechanisms underlying cold-induced atherosclerosis and propose potential microbial or metabolic targets for therapeutic intervention.

Methods

Human studies

The study protocols were conducted in strict accordance with the Declaration of Helsinki and were approved by the Research Ethics Committee of the First Affiliated Hospital of Harbin Medical University (ethical approval number: 2022IIT094). Written informed consent was obtained from all participants prior to sample collection. Participants were recruited at the Medical Center of the First Affiliated Hospital of Harbin Medical University during both winter and summer seasons. Harbin is a northern city in China characterized by sustained low ambient temperatures during winter. Participants recruited between November and January were classified as the winter cohort, whereas those recruited between June and August were classified as the summer cohort. During the winter sampling period, the average outdoor daytime temperature ranged approximately from − 15 °C to -5 °C, based on local meteorological records. Participants in the winter cohort were required to have been continuously exposed to outdoor cold for at least 2 h per day for a minimum of two consecutive weeks prior to fecal sample collection [18]. Outdoor exposure occurred during routine daily activities (e.g., commuting, walking, or daily errands) while wearing standard seasonal clothing. Individuals with occupational cold exposure, use of specialized cold-protective equipment, or exposure to extreme cold environments were excluded to minimize heterogeneity in cold stress intensity. Participants in the summer cohort were recruited between June and August, during which the average outdoor daytime temperature ranged from approximately 20 °C to 28 °C. No participants were exposed to extreme heat conditions, occupational heat exposure, or heat-wave events. Thus, the summer cohort represents a physiological normothermic control group rather than a heat-stress condition. The study population included individuals with ASCVD, encompassing stable coronary artery disease (stable angina pectoris) and acute coronary syndromes (unstable angina and myocardial infarction), as well as age- and sex-matched healthy controls. All participants were sampled during either winter or summer according to the criteria described above. Fecal samples were collected on the day of clinical assessment, immediately snap-frozen in liquid nitrogen, and stored at − 80 °C until further analysis. Baseline demographic and clinical characteristics of the four study groups—healthy controls (winter), ASCVD patients (winter), healthy controls (summer), and ASCVD patients (summer)—are summarized in Supplementary Table 1.

Mouse studies

All animal procedures in this study strictly adhered to the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care Committee at Harbin Medical University to ensure ethical treatment (Ethical approval number: 2020030). Sixty male ApoE^−/− mice (20–22 g, 8 weeks old) were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). The mice were housed at the Experimental Animal Center of Harbin Medical University under a 12:12-hour light-dark cycle and were fed a Western diet containing 20% milk fat and 0.2% cholesterol. After 12 weeks on this diet, they were sacrificed at 20 weeks of age. The mice were randomly assigned to experimental conditions, individually housed, and anesthetized intraperitoneally with Alfaxalone before being euthanized by cervical dislocation. All interventions were initiated at 8 weeks of age concurrently with Western diet feeding and continued for 12 weeks until sacrifice at 20 weeks of age.

Experiment 1. Cold exposure protocol and dietary control

To evaluate the effect of cold exposure on atherosclerosis, male ApoE⁻/⁻ mice were randomly assigned to a room temperature (RT) group or a cold-exposure group using a completely randomized design. Mice in the RT group were housed at 25 °C, whereas mice in the cold group were exposed to 4 °C for 4 h per day in a temperature-controlled chamber and returned to 25 °C for the remaining hours. Cold exposure was performed during the light phase from 09:00 to 13:00 each day to minimize circadian confounding. The intervention lasted for 12 weeks, consistent with the duration required for stable remodeling of gut microbiota and atherosclerotic progression. To minimize the confounding effect of cold-induced hyperphagia, mice in both groups were maintained on the same Western diet throughout the experiment. Food intake was carefully monitored on a daily basis. Water was provided ad libitum. Body weight was monitored at least weekly, and animals were inspected daily to ensure animal welfare.

Experiment 2. FMT

Fresh fecal pellets were collected from donor mice in the RT and cold groups at a consistent time of day (09:00–11:00) to minimize diurnal variation. Samples were processed immediately under sterile conditions. Briefly, feces were resuspended in sterile, pre-reduced PBS at a ratio of 50 mg/mL, homogenized for 5 min, vortexed for 1 min, and centrifuged at 800 g for 3 min to remove large debris [13]. The supernatant was used as the transplant material.

Recipient ApoE⁻/⁻ mice were pre-treated with a broad-spectrum antibiotic cocktail for 14 days to reduce baseline microbiota and facilitate engraftment, followed by a washout period of 24 h before FMT. Recipients were then randomly assigned to receive microbiota from RT donors (RT-FMT) or cold-exposed donors (Cold-FMT). FMT was administered by oral gavage at 200 µL once daily for 3 consecutive days, then twice weekly for 12 weeks. All recipient mice were fed an identical Western diet during the FMT period to control dietary effects. FMT recipient mice were maintained under room-temperature conditions throughout the entire transplantation period.

Experiment 3. Antibiotic-mediated microbiota depletion and cold exposure

To investigate the microbiota-associated contribution to cold-induced atherosclerosis, ApoE⁻/⁻ mice were treated with a broad-spectrum antibiotic (ABX) cocktail to deplete gut microbiota prior to cold exposure. The ABX cocktail consisted of ampicillin (1 g/L), neomycin (1 g/L), metronidazole (1 g/L), and vancomycin (0.5 g/L), administered in drinking water for 2 consecutive weeks as an induction phase. Following this initial depletion period, mice were subjected to the cold exposure protocol as described above. To prevent recovery of the gut microbiota during the prolonged cold exposure period, antibiotic treatment was continued at a reduced maintenance dose (50% of the induction concentration) throughout the cold exposure phase. Specifically, ampicillin (0.5 g/L), neomycin (0.5 g/L), metronidazole (0.5 g/L), and vancomycin (0.25 g/L) were provided in drinking water during the maintenance phase. Control mice received regular drinking water without antibiotics. This induction–maintenance antibiotic regimen was designed to ensure sustained suppression of gut microbiota and to enable assessment of the microbiota-associated effects of cold exposure on atherosclerotic progression.

Experiment 4. L. johnsonii supplementation during cold exposure

To assess the effect of L. johnsonii supplementation on cold-induced atherosclerosis, eight-week-old male ApoE⁻/⁻ mice were exposed to cold conditions as described in Experiment 1 and maintained on a Western diet. Mice were randomly assigned to receive either L. johnsonii or vehicle control. L. johnsonii was administered by oral gavage at a dose of 1 × 10⁹ CFU suspended in 200 µL sterile PBS once per week for 12 weeks. Control mice received an equal volume of sterile PBS on the same schedule. Gavage was performed throughout the cold exposure period.

Experiment 5. Leucine supplementation

To evaluate the effect of leucine on atherosclerosis independent of cold exposure, eight-week-old male ApoE^−/− mice were housed at room temperature and fed a Western diet containing either normal leucine content (1.5% w/w) or high leucine content (5% w/w). Diets were isocaloric and matched for macronutrient composition except for leucine content. Mice were maintained on the assigned diets for 12 weeks, after which atherosclerotic burden and plaque characteristics were assessed.

Experiment 6. Combined L. johnsonii and leucine intervention under cold exposure

To investigate the interaction between L. johnsonii and leucine in cold-induced atherosclerosis, eight-week-old male ApoE⁻/⁻ mice were exposed to cold conditions and maintained on a Western diet. Mice were randomly assigned to one of three groups: cold control, cold + L. johnsonii, or cold + L. johnsonii + leucine. L. johnsonii supplementation was performed as described in Experiment 4. Leucine supplementation was achieved through dietary intervention using a high-leucine Western diet (5% leucine), as described in Experiment 5. All interventions were initiated simultaneously and continued under the daily cold exposure regimen (4 °C, 4 h/day during the light phase) for 12 weeks.

Atherosclerotic plaque analysis

For the staining of atherosclerotic lesions throughout the entire aorta, the aorta extending to the iliac arteries was dissected and longitudinally opened along the ventral axis. The aortas were then mounted on black wax within a pan and stained using Oil Red O (Solarbio, Beijing, China). Initially, the mouse aortas were briefly rinsed in 60% isopropanol for 30 s. Subsequently, they were stained with a freshly prepared Oil Red O solution for 20 min. After staining, the samples were washed with 60% isopropanol for 10 s, followed by distilled water for 10 s. The aortas were then imaged using a digital camera, and lesion quantification was performed using ImageJ software.

The aortic root was fixed with 4% paraformaldehyde and embedded in Tissue Tek OCT compound (Sakura, Tokyo, Japan). The samples were then frozen in liquid nitrogen and stored at -80 °C for future use. A Thermo Scientific CryoStar NX70 low-temperature cryostat was used to section the aortic root embedded in OCT. A series of 6 μm thick cross-sections of the aortic root were obtained, starting from the appearance of the first three aortic valves until their disappearance.

To quantify neutral lipids in the plaques of the aortic root, frozen cross-sections were stained with Oil Red O and counterstained with hematoxylin. In brief, the frozen sections of mouse aortic sinuses were first rinsed with 60% isopropanol for 30 s, followed by staining with a freshly prepared Oil Red O working solution for 20 min. The sections were then washed sequentially with 60% isopropanol for 10 s and distilled water for another 10 s before being counterstained with hematoxylin for 1 min.

Frozen cross sections of the aortic sinus were stained with Hematoxylin and Eosin (HE, Solarbio, Beijing, China) to analyze the necrotic core area. Briefly, the frozen sections of mouse aortic sinuses were stained with hematoxylin solution for 3 min. After hematoxylin staining, the slides were rinsed in running tap water for 5 min, followed by immersion in acid ethanol to destain, and then rinsed with deionized water. Subsequently, the slides were stained with eosin solution for 1 min. The necrotic core in the intima was defined as an acellular/anuclear area (negative for hematoxylin-positive nuclei) that was devoid of HE staining. The quantification of lesion area and necrotic core area size was performed using ImageJ software.

Atherosclerotic plaque stability was assessed by immunohistochemical analysis and Masson’s Trichrome Staining. Cryosections of mouse aortic sinuses underwent initial antigen retrieval treatment. Slides were subsequently blocked with 5% goat serum for 30 min to eliminate nonspecific binding. Following PBS washes, sections were incubated with primary antibodies at 4 °C overnight within a humidified chamber. After additional PBS washes, slides were incubated with species-matched secondary antibodies. Following thorough PBS rinsing, 3,3’-diaminobenzidine (DAB) chromogenic development was performed. Finally, sections were counterstained with hematoxylin, dehydrated through an ethanol series, and permanently mounted with anti-fade medium for microscopic analysis (Zeiss, Jena, Germany). Masson’s trichrome staining was employed to assess collagen deposition and fibrosis in the aortic sinus tissues. To quantify the extent of collagen deposition, the collagen volume fraction was determined by calculating the ratio of the area occupied by collagen fibers to the total area in the visual field. This analysis was performed using Image-Pro Plus software, enabling accurate and standardized measurements.

The cellular composition of atherosclerotic plaques was analyzed by immunofluorescent staining of frozen tissue sections. Briefly, the frozen sections of mouse aortic sinuses were permeabilized with PBS containing 0.25% Triton X-100 and blocked by immersing the slides in 5% goat serum in PBS for 30 min. After washing with PBS, the sections were incubated with primary antibodies in a humid chamber overnight at 4 °C. The slides were then washed with PBS and incubated with secondary antibodies. Finally, after a final wash with PBS, the sections were mounted with antifade mounting medium. Nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI). Images of the stained samples were captured with a fluorescence microscope (Zeiss, Jena, Germany).

The plaque instability index was calculated as (Oil Red O-positive area + MOMA-2-positive area) / (α-SMA-positive area + Masson’s trichrome-positive area) [19, 20]. This composite parameter was used to assess plaque instability. Only samples with complete measurements of all four parameters from the same lesion/section were included in this analysis.

Meso scale discovery (MSD) multi-cytokines measurements

Multiplex cytokine analysis was performed using a MSD electrochemiluminescence platform according to the manufacturer’s instructions. Plasma samples were diluted as recommended and assayed on MSD multi-spot plates. Cytokine concentrations were quantified using MSD Discovery Workbench software based on standard curves generated from manufacturer-provided calibrators and fitted using a four-parameter logistic regression model. All plasma samples were collected in the morning between 08:00 and 10:00 a.m. after an overnight fast to minimize circadian variability in cytokine levels. Blood samples were processed immediately after collection, and plasma was isolated by centrifugation and stored at − 80 °C until analysis. Each sample was measured in technical duplicate. The lower and upper limits of quantification (LLOQ and ULOQ) for each cytokine were defined according to MSD guidelines. Plate-level quality control samples were included on each assay plate. Inter- and intra-assay coefficients of variation (CVs) were < 15%.

Isolation and culture of bone marrow-derived macrophages

To isolate mouse BMDMs, bone marrow was harvested from the femurs and tibias of C57BL/6J mice. First, hind limbs were excised from the mice, and muscle tissue was removed to expose the femur and tibia. The bones were then flushed with RPMI 1640 using syringes to collect bone marrow cells. The cell suspension was passed through a 70 μm filter and centrifuged at 500 g for 10 min. The resulting pellet was resuspended and inoculated into 6-well plates containing RPMI 1640 supplemented with M-CSF (10 ng/mL), 10% (v/v) FBS, and 1% (v/v) penicillin/streptomycin. The cells were incubated at 37 °C in a 5% CO2 atmosphere. After 6 days of incubation, the bone marrow cells had differentiated into macrophages.

Cell culture

Jurkat cells (CL0315) were obtained from the Wuhan Pricella Biotechnology Co., Ltd. and cultured per manufacturer’s instructions.

Efferocytosis assay

For the in vitro efferocytosis assay, primary BMDMs and Jurkat cells were cultured. BMDMs were labeled with CMTPX red cell tracer dye for 30 min, then washed and starved overnight in serum-free medium. Jurkat cells were irradiated with 254 nm ultraviolet light for 1 h, followed by incubation under normal cell culture conditions for 2 h. The apoptotic Jurkat cells were then labeled with CFDA SE green cell tracer. The apoptotic cells were added to the BMDMs at a 1:5 ratio and incubated for 2 h. Imaging was performed using an immunofluorescence microscope (Zeiss, Jena, Germany).

Real-time quantitative PCR (qPCR and qRT-PCR)

For absolute quantification of specific bacterial taxa by qPCR, total DNA was extracted from fecal samples using the TIANamp Stool DNA Kit (Tiangen Biotech, China) according to the manufacturer’s instructions. qPCR was performed to determine the absolute abundance of selected bacterial taxa using taxon-specific primers. Standard curves were generated using serial dilutions of plasmids containing the corresponding target sequences. Results were normalized to fecal sample mass and expressed as copy number per gram of feces. This approach was used to quantify the absolute abundance of specific bacterial taxa rather than total bacterial load or relative abundance derived from 16 S rRNA sequencing.

For host gene expression analysis, total RNA was isolated from tissues or cultured cells using the AxyPrep Multisource Total RNA Miniprep Kit (Axygen). RNA was reverse-transcribed into cDNA using the Yeasen RT Kit following the manufacturer’s protocol. qRT-PCR was performed using SYBR Green Supermix on an ABI 7500 Real-Time PCR System (Applied Biosystems). GAPDH was used as an internal reference gene. The amplification protocol consisted of an initial denaturation at 95 °C for 30 s, followed by 35 cycles of 95 °C for 10 s and 60 °C for 30 s. Relative gene expression levels were calculated using the 2^-ΔΔCt method. Statistical analysis of qRT-PCR data was performed using ΔCt values. Primer sequences are provided in Supplementary Table 2.

Chromatin immunoprecipitation (ChIP)-qPCR assay

ChIP assays were performed to determine the binding of Zic2 to the Gas6 promoter region. Briefly, macrophages were cross-linked with 1% formaldehyde for 10 min at room temperature, and the reaction was quenched with glycine. Cells were then washed with cold PBS and lysed in ChIP lysis buffer supplemented with protease inhibitors. Chromatin was fragmented by sonication to an average size of approximately 200–500 bp. After centrifugation, the clarified chromatin lysates were incubated overnight at 4 °C with an anti-Zic2 antibody (ab150404) or normal rabbit IgG as a negative control. Protein A/G magnetic beads were subsequently added and incubated for 1 h at 4 °C to capture immune complexes. After extensive washing, chromatin complexes were eluted and reverse cross-linked at 65 °C overnight. DNA was purified using a PCR purification kit. Purified DNA was analyzed by qPCR using primers targeting the predicted Zic2-binding region within the Gas6 promoter. ChIP-qPCR enrichment was normalized to input DNA and presented relative to IgG controls. The primer sequences used for ChIP-qPCR were as follows: Gas6-ChIP-F: 5′-CTTGTGGGGTGAGGTTCAGG-3′,

Gas6-ChIP-R: 5′-CAGGTGAGGCAACCTTCGT-3′.

Western blotting

For total protein extraction, frozen tissues were minced and lysed in RIPA buffer (20 µL/mg tissue), homogenized, and centrifuged at 14,000 RPM for 10 min at 4 °C to collect the supernatant. For cell lysates, harvested cells were lysed in RIPA buffer (200 µL/10⁶ cells), pipetted, incubated on ice for 30 min, and centrifuged under the same conditions. Protein lysates were mixed with loading buffer, heated at 95 °C for 5 min, separated by 8%-12% SDS-PAGE, and transferred onto Immobilon-P membranes. After blocking with 5% non-fat milk, membranes were incubated with primary antibodies, followed by HRP-conjugated secondary antibodies. Protein bands were visualized using an enhanced chemiluminescence kit. Antibodies against Gas6 (ab86059), MerTK (ab52968), and CD47 (ab300124) were obtained from Abcam, and GAPDH (60004-1-Ig) was obtained from Proteintech.

Culture of L. johnsonii and Lactobacillus reuteri

L. johnsonii (BNCC186110) and L. reuteri (BNCC254476) were obtained from BeNa Culture Collection and grown anaerobically in brain heart infusion broth. Bacterial concentration (CFU/mL) was assessed by plate counting on mucin media with 1% agar under anaerobic conditions. Bacterial cultures were resuspended in PBS to a final concentration of 10⁹ CFU/200 µL for oral gavage.

Flow cytometry

Cells were stained for 30 min at 4 °C protected from light. The following antibodies were used for flow cytometry analysis: F4/80 FITC (BioLegend, 123107), CD68 PE (BioLegend, 137013), and annexin V (Beyotime). After staining, cells were washed, centrifuged, and resuspended in flow cytometry staining buffer. For apoptosis analysis, propidium iodide (PI) was added to the appropriate samples immediately before acquisition to distinguish between early apoptotic and late apoptotic/necrotic cells. Data were acquired on a BD FACSCanto II flow cytometer (BD Biosciences, USA). Gating strategies and analysis were performed using FlowJo software (Tree Star, USA).

16S rRNA gene sequencing and analysis

For 16S rRNA gene sequencing, total genomic DNA was extracted from fecal samples using the CTAB/SDS method. The V3-V4 regions of the bacterial 16S rRNA gene were amplified using barcoded universal primers, and paired-end sequencing (2 × 250 bp) was performed on an Illumina NovaSeq platform. Raw sequencing data were processed using QIIME2 (version 2021.2). DADA2 was used for quality filtering, denoising, chimera removal, and amplicon sequence variant (ASV) inference. Taxonomic assignment was performed using the SILVA 138 reference database. Alpha and beta diversity analyses based on Bray-Curtis distances were conducted within QIIME2, and principal coordinate analysis (PCoA) was used for visualization.

Metabolomic analysis

Untargeted metabolomic profiling was performed using a Vanquish UHPLC system coupled to a Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific). Chromatographic separation was achieved on a UPLC HSS T3 column using water and acetonitrile as mobile phases, both containing 5 mmol/L ammonium acetate and 5 mmol/L acetic acid. The autosampler temperature was maintained at 4 °C, and the injection volume was 3 µL. Mass spectrometry data were acquired using electrospray ionization in data-dependent acquisition (DDA) mode. Raw data were converted using ProteoWizard and processed using an XCMS-based pipeline implemented in R for peak detection, retention time alignment, and feature extraction. Pooled quality control (QC) samples, prepared by mixing equal aliquots of all study samples, were injected periodically throughout the analytical run to monitor instrument stability. Internal standards were used for signal normalization. Features with a relative standard deviation (RSD) greater than 30% in QC samples or detected in fewer than 50% of samples were excluded. Missing values were imputed using half of the minimum detected value for each metabolite. Multivariate analysis was performed using principal coordinate analysis (PCoA) based on Bray–Curtis distance to visualize global differences in plasma metabolomic profiles between groups. Metabolite annotation was performed based on accurate mass and MS/MS spectral matching against the BiotreeDB MS2 database. with a similarity cutoff of 0.3, which was applied for preliminary annotation. Annotation confidence levels were assigned according to the Metabolomics Standards Initiative (MSI) guidelines. Key metabolites of interest, including leucine, were further validated and quantified using targeted UHPLC–MS/MS analysis with authentic standards, as described below.

Determination of valine, leucine, and isoleucine

Plasma valine, leucine, and isoleucine were quantified using UHPLC-MS/MS. A 20 µL sample was analyzed on an Agilent 1290 Infinity II system with a Waters ACQUITY UPLC HSS T3 column, and detection was performed using an Agilent 6460 triple quadrupole mass spectrometer with an AJS-ESI interface. MRM data were acquired and processed with Agilent MassHunter Software. Calibration standards were serially diluted and analyzed, with LLOD and LLOQ determined based on S/N ratios of 3 and 10, respectively, following US FDA bioanalytical validation guidelines.

Enzymatic colorimetric assay

Plasma lipid parameters, including total cholesterol (TC), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C), were quantified using standard enzymatic colorimetric assays according to the manufacturer’s instructions. Briefly, plasma samples were incubated with enzyme reaction reagents, and absorbance was measured at the appropriate wavelength using a microplate reader. Lipid concentrations were calculated based on standard curves generated from known calibrators.

Construction of luciferase vector and transfection

Gas6-WT and Gas6-MT were cloned into pGL3-basic, while Zic2 was inserted into pcDNA3.1, with all constructs sequence-verified. pGL3-basic carried Firefly luciferase, and pRL-TK encoded Renilla luciferase as an internal control. 293T cells were seeded in 24-well plates (60–70% confluence) and transfected with plasmids (Gas6-WT or Gas6-MT with Zic2-pcDNA3.1 or an empty vector) using Lipofectamine 2000. After 4–6 h, the medium was replaced, and cells were incubated for 48 h. Transfections were performed in quintuplicate. For the dual-luciferase assay, cells were lysed, incubated on ice, and centrifuged, followed by measurement of Firefly and Renilla luciferase activities using the Dual-Luciferase Reporter Assay System (Promega), with Firefly luciferase normalized to Renilla.

Statistical analyses

GraphPad Prism version 9.0 was used for data visualization and statistical analysis. Data are presented as mean ± s.d. Normality was assessed using the Shapiro–Wilk test. For comparisons between two groups, an unpaired two-tailed Student’s t-test was used for normally distributed data, whereas the Mann–Whitney U test was applied for non-normally distributed data. For comparisons among more than two groups, one-way ANOVA followed by Tukey’s multiple-comparison test was used for normally distributed data, whereas the Kruskal-Wallis test followed by Dunn’s multiple-comparison test was used for non-normally distributed data. For multiple testing, P values were adjusted using the Benjamini-Hochberg false discovery rate (FDR) method where applicable. Statistical significance was set at p < 0.05.

Results

Cold exposure accelerates atherosclerotic plaque formation and instability by inducing gut microbiota dysbiosis in ApoE^−/− Mice

Clinical studies have demonstrated that cold exposure increases both the risk and burden of acute myocardial infarction [4, 7]. To determine whether cold exposure contributes to atherosclerotic plaque formation, 8-week-old ApoE^−/− mice fed with a Western diet were subjected to cold exposure (4℃, 4 h/d; 25℃, 20 h/d) for 12 weeks and compared with mice housed at room temperature (RT, 25℃, 24 h/d, Fig. 1A). Throughout the 12-week period, body weight and food intake were monitored, revealing no significant differences between the two groups (Fig. S1A-B). Oil Red O staining revealed a significant increase in plaque formation in both the aorta and aortic sinus of the cold group compared to the RT group (Fig. 1B-C). Moreover, cold exposure led to distinct changes in plaque composition, including increased necrotic core areas (Fig. 1D), decreased Masson’s trichrome-positive areas (Fig. 1E), reduced alpha-smooth muscle actin (α-SMA)-positive areas (Fig. 1F), and increased monocyte/macrophage antigen-2 (MOMA-2)-positive macrophage infiltration (Fig. 1G). These changes were reflected in a higher plaque instability index in the cold group, calculated as [(Oil Red O-positive area) + (MOMA-2-positive area)] / [(α-SMA-positive area) + (Masson’s trichrome-positive area)] (Fig. 1H) [20]. Additionally, cold exposure significantly elevated plasma levels of the pro-inflammatory cytokines tumor necrosis factor-alpha (TNF-α), interleukin-1 beta (IL-1β), and interleukin-6 (IL-6), while reducing the anti-inflammatory cytokine interleukin-10 (IL-10) (Fig. 1I). Taken together, these results indicate that cold exposure significantly accelerates atherosclerotic plaque growth and promotes features of plaque instability in ApoE^−/− mice. Notably, plasma levels of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and triglycerides (TG) remained unchanged between the two groups (Fig. 1J).

Fig. 1.

Cold exposure accelerates atherosclerotic plaque formation and instability in ApoE^−/− mice. A, Model of ApoE^−/− mice exposed to room temperature (RT; 25℃, 24 h/d) or cold temperature (4℃, 4 h/d; 25℃, 20 h/d) for 12 weeks. B, Representative en face photographs of aortas showing Oil Red O-stained atherosclerotic plaques and quantitative data of Oil Red O-positive staining area in the RT group and cold group (n = 6 per group, Scale bar: 2 mm). C, Representative Oil Red O staining of cross-sections of aortic roots and quantitative data of plaque area in each group (n = 4 per group, Scale bar: 500 μm). D, Representative H&E staining of cross-sections of aortic roots and quantitative data of necrotic core area in each group (n = 6 per group, Scale bar: 100 μm). E, Representative MASSON staining of cross-sections of aortic roots and Masson’s trichrome-positive area in each group (n = 6 per group, Scale bar: 100 μm). F, Representative α-SMA immunohistochemistry staining of cross-sections of aortic roots and quantitative data of α-SMA⁺ area in each group (n = 6 per group, Scale bar: 100 μm). G, Representative MOMA-2 immunohistochemistry staining of cross-sections of aortic roots and quantitative data of MOMA-2⁺ area in each group (n = 6 per group, Scale bar: 100 μm). H, Quantification of the instability index in each group (n = 4 per group). I, Levels of inflammatory cytokines TNF-α, IL-1β, IL-6, and IL-10 in each group (n = 6 per group). J, Plasma levels of TC, LDL-C, and TG in each group (n = 6 per group). Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant

An expanding body of research has highlighted the substantial impact of gut microbiota on cardiovascular diseases [11, 21]. However, to date, no studies have directly investigated the role of gut microbiota in cold-induced atherosclerosis. Recent research has shown that cold exposure can induce changes in gut microbiota composition [22–24]. In assessing the crucial role of gut microbiota in the development of cold-induced atherosclerosis, we performed FMT using samples from both RT and cold group mice (Fig. 2A). Notably, microbiota derived from cold group mice led to a significant increase in plaque formation in both the aorta and aortic sinus (Fig. 2B-C). Furthermore, fecal microbiota transplantation induced distinct changes in plaque composition, including increased necrotic core areas (Fig. 2D), decreased Masson’s trichrome-positive areas (Fig. 2E), reduced α-SMA-positive areas (Fig. 2F), and increased MOMA-2-positive macrophage infiltration (Fig. 2G). Collectively, these results suggest that cold-associated microbiota promotes plaque formation and enhances plaque instability (Fig. 2I). In addition, cold-associated microbiota significantly elevated plasma levels of TNF-α, IL-1β, and IL-6, while reducing IL-10 levels in ApoE^−/− mice (Fig. 2J). These findings indicate that alterations in the gut microbiota constitute an essential intermediate linking cold exposure to atherosclerotic plaque progression. To further confirm the role of the gut microbiota, an antibiotic cocktail (ABX) was administered in the drinking water for two weeks prior to cold exposure to deplete the intestinal microbiota. As expected, antibiotic treatment alleviated the formation and instability of atherosclerotic plaques induced by exposure to cold temperatures (Fig. S2A-F), further supporting a critical role of gut microbiota in mediating cold-induced atherosclerotic progression. HE staining revealed morphological alterations in the intestinal structure following cold exposure (Fig. S3A). Consistent with these observations, the expression of tight junction proteins, including zonula occludens-1 (ZO-1) and Occludin, was reduced (Fig. S3B-D), suggesting impaired gut barrier integrity.

Fig. 2.

Cold exposure accelerates atherosclerotic plaque formation and instability by inducing gut microbiota dysbiosis in ApoE^−/− mice. A, Model of ApoE^−/− mice receiving RT-FMT or Cold-FMT for 12 weeks. B, Representative en face photographs of aortas showing Oil Red O-stained atherosclerotic plaques and quantitative data of Oil Red O-positive staining area in the RT-FMT and Cold-FMT groups (n = 6 per group, Scale bar: 2 mm). C, Representative Oil Red O staining of cross-sections of aortic roots and quantitative data of plaque area in each group (n = 4 per group, Scale bar: 500 μm). D-H, Representative H&E staining, MASSON staining, α-SMA immunohistochemistry staining and MOMA-2 immunohistochemistry staining of cross-sections of aortic roots and quantitative data of necrotic core area, collagen-positive area, α-SMA⁺ area, and MOMA-2⁺ area in each group (n = 6 per group, Scale bar: 100 μm). I, Quantification of the instability index in each group (n = 4 per group). J, Levels of inflammatory cytokines TNF-α, IL-1β, IL-6, and IL-10 in each group (n = 6 per group). Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Gut microbiota–associated elevation of circulating leucine promotes cold-induced atherosclerotic plaque formation and instability in ApoE^−/− Mice

The gut microbiota modulates the function of remote organs by regulating the synthesis of host metabolites [25, 26]. To identify the metabolic mediators linking gut microbiota to host physiology under cold exposure, we performed untargeted metabolomic profiling of blood plasma from RT and cold group mice. Principal coordinate analysis (PCoA) revealed a clear separation between groups (Fig. 3A), suggesting a global shift in circulating metabolites. Our functional analysis of the differential metabolites revealed significant disruptions in the biosynthetic pathways of valine, leucine, and isoleucine (Fig. 3B), essential amino acids implicated in various metabolic processes, highlighting BCAA metabolism as a key component of the cold response. Among the individual metabolites contributing to this enrichment, both ketoleucine (an intermediate in leucine metabolism) and pyruvic acid were significantly altered in cold-exposed mice (Fig. S4A), pointing toward active remodeling of BCAA-related fluxes. Notably, targeted quantification using LC-MS/MS revealed a significant increase in plasma leucine levels (Fig. 3C), whereas changes in valine and isoleucine were not statistically significant (Fig. S5A-B). This pattern suggests that leucine, rather than other BCAA, is selectively upregulated during cold exposure, consistent with a directed metabolic shift rather than generalized BCAA accumulation. Consistently, FMT from cold-exposed mice into RT recipients significantly elevated plasma leucine levels (Fig. 3D).

Fig. 3.

Gut microbiota-associated elevation of circulating leucine promotes cold-induced atherosclerotic plaque formation and instability in ApoE^−/− mice. A, PCoA of plasma metabolomic profiles (n = 5 per group). B, Analysis of metabolic pathway enrichment (n = 5 per group). C, The level of leucine in plasma in the RT group and cold group (n = 8 per group). D, The level of leucine in plasma in the RT-FMT group and Cold-FMT group (n = 8 per group). E, Model of ApoE^−/− mice receiving a normal-leucine Western diet (NL) or a high-leucine Western diet (HL) for 12 weeks. F, Representative en face photographs of aortas showing oil red O-stained atherosclerotic plaques and quantitative data of Oil Red O-positive staining area in the normal-leucine group or the high-leucine group (n = 6 per group, Scale bar: 2 mm). G, Representative Oil Red O staining of cross-sections of aortic roots and quantitative data of plaque area in each group (n = 4 per group, Scale bar: 500 μm). H-L, Representative H&E staining, MASSON staining, α-SMA immunohistochemistry staining and MOMA-2 immunohistochemistry staining of cross-sections of aortic roots and quantitative data of necrotic core area, collagen-positive area, α-SMA⁺ area, and MOMA-2⁺ area (n = 6 per group, Scale bar: 100 μm). M, Quantification of the instability index in each group (n = 4 per group). N, Levels of inflammatory cytokines TNF-α, IL-1β, IL-6, and IL-10 (n = 6 per group). Data are presented as mean ± s.d. PCoA was performed based on Bray-Curtis distance and evaluated by PERMANOVA (Adonis). Other P values were determined by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Building upon these insights, we next examined whether dietary elevation of leucine is sufficient to recapitulate key features of cold-associated atherosclerosis. ApoE^−/− mice housed at room temperature were fed isocaloric diets containing either normal or high levels of leucine (Fig. 3E), after which atherosclerotic lesion burden in the aorta and aortic sinus was assessed. Remarkably, mice fed a high-leucine diet exhibited a significant increase in aortic and aortic sinus lesion areas compared with those fed a normal-leucine diet (Fig. 3F-G). In parallel, plaques from the high-leucine diet group displayed enlarged necrotic cores and increased MOMA-2–positive macrophage accumulation, accompanied by reduced α-SMA–positive and Masson’s trichrome–positive areas (Fig. 3H-M), indicating compromised plaque stability. Furthermore, consumption of a high-leucine diet was associated with elevated plasma levels of the pro-inflammatory cytokines TNF-α, IL-1β, and IL-6, together with decreased levels of the anti-inflammatory cytokine IL-10 (Fig. 3N).

Leucine impairs efferocytosis and promotes cold-induced atherosclerotic plaque formation and instability in ApoE^−/− Mice

Given the marked expansion of necrotic cores and heightened inflammatory responses observed under cold exposure and leucine supplementation, we hypothesized that macrophage-mediated efferocytosis might be impaired in this context. Efficient efferocytosis is critical for the clearance of apoptotic cells within atherosclerotic plaques, and its dysfunction is known to promote secondary necrosis, chronic inflammation, and plaque instability [17, 27]. Consistent with this hypothesis, quantitative analysis of aortic lesions revealed a significant reduction in the efferocytosis index [19] —defined as the ratio of macrophage-associated apoptotic cells to free apoptotic cells—in cold-exposed mice compared with RT controls (Fig. 4A-B). Notably, leucine supplementation in RT mice similarly resulted in a marked decrease in efferocytosis, recapitulating the effect of cold exposure (Fig. S6A-B).

Fig. 4.

Leucine impairs efferocytosis and promotes cold-induced atherosclerosis plaque formation and instability in ApoE^−/− mice. A, Representative images of aortic root sections showing macrophages labeled by CD68 (green) and apoptotic cells labeled by cleaved Caspase3 (red) in the RT group and cold group (Scale bar: 50 μm). B, Quantification of the ratio of macrophage-associated cleaved Caspase3-positive to free cleaved Caspase3 cells in aortic roots sections in each group (n = 6 per group). C, Fluorescent images of labeled macrophages (red) with apoptotic Jurkat cells (green) in the vehicle group, leucine group, and leucine + BCH group (scale bar: 50 μm). D, Percent efferocytosis was quantified as the number of macrophages with engulfed apoptotic cells as a percentage of total macrophages in each group (n = 6 per group). E-F, Efferocytic index gating strategy and quantification measured by flow cytometry in each group (n = 6 per group). G, Levels of inflammatory cytokines TNF-α, IL-1β, IL-6, and IL-10 in each group (n = 6 per group). Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test or one-way ANOVA with Tukey’s multiple-comparison test. *p < 0.05, **p < 0.01, ***p < 0.001

To further elucidate the impact of leucine on efferocytosis at the cellular level, we performed in vitro co-culture assays using apoptotic cells labeled with CFDA SE and BMDMs labeled with CMTPX. Confocal microscopy demonstrated that leucine exposure significantly reduced macrophage efferocytic capacity, as reflected by diminished uptake of apoptotic cells. Importantly, this inhibitory effect was effectively reversed by BCH, a leucine transport inhibitor (Fig. 4C-D). Flow cytometric analysis further corroborated these findings, confirming that leucine markedly impaired macrophage efferocytic capacity, whereas BCH treatment restored efferocytosis toward baseline levels (Fig. 4E-F). In parallel, leucine exposure was associated with increased expression of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) and reduced expression of the anti-inflammatory cytokine IL-10, an effect that was largely reversed by BCH treatment (Fig. 4G).

Leucine impairs macrophage efferocytosis via downregulation of Gas6

To elucidate the molecular basis underlying impaired efferocytosis observed under cold exposure and elevated leucine conditions, we first performed a targeted qRT-PCR screen of 12 genes previously implicated in macrophage efferocytosis. Among these candidates, Gas6, CD47, and Mertk exhibited significant transcriptional alterations, with Gas6 showing the most pronounced downregulation (Fig. 5A). To validate these findings at the protein level, we analyzed aortic tissues by Western blotting. While the protein expression of CD47 and MerTK remained largely unchanged following cold exposure (Fig. 5B), Gas6 protein levels were markedly reduced in the cold group (Fig. 5C), suggesting that Gas6 may be particularly responsive to cold-associated metabolic alterations.

Fig. 5.

Leucine impairs macrophage efferocytosis via downregulation of Gas6. A, qRT-PCR was performed to detect the levels of 12 efferocytosis-related genes in aortas in the RT group and cold group (n = 6 per group). B, Representative bands and quantification of MerTK and CD47 expression in each group (n = 6 per group). C, Representative bands and quantification of Gas6 expression in each group (n = 6 per group). D, Fluorescent images of labeled macrophages (red) with apoptotic Jurkat cells (green) in the vehicle group, oeGas6 group, leucine group, and leucine + oeGas6 group (scale bar = 50 μm, n = 6 per group). E, Percent efferocytosis was quantified as the number of macrophages with engulfed apoptotic cells as a percentage of total macrophages in each group (n = 6 per group). F-G, Efferocytic index gating strategy and quantification measured by flow cytometry in each group (n = 6 per group). H-I, Representative bands and quantification of Gas6 expression in each group (n = 6 per group). J, Levels of inflammatory cytokines TNF-α, IL-1β, IL-6, and IL-10 in each group (n = 8 per group). Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test or one-way ANOVA with Tukey’s multiple-comparison test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant

Given that Gas6 is a secreted bridging molecule that facilitates efferocytosis by linking phosphatidylserine on apoptotic cells to MerTK receptors on macrophages, its downregulation provides a plausible mechanistic explanation for defective apoptotic cell clearance under these conditions [28, 29]. To directly assess the functional role of Gas6 in macrophage efferocytosis, we overexpressed Gas6 in BMDMs. Confocal microscopy revealed that Gas6 overexpression (oeGas6) significantly enhanced efferocytic capacity, as evidenced by increased uptake of apoptotic cells. In contrast, leucine treatment alone markedly impaired macrophage efferocytosis. Importantly, co-treatment with oeGas6 largely restored efferocytic capacity despite leucine exposure (Fig. 5D-E). These findings were further corroborated by flow cytometric analysis, which demonstrated a significant increase in the proportion of apoptotic cells engulfed by macrophages in the oeGas6 group compared with control conditions. Consistent with imaging results, leucine reduced macrophage efferocytic capacity, whereas Gas6 overexpression rescued this defect to levels comparable to untreated controls (Fig. 5F-G). Western blot analysis confirmed efficient Gas6 overexpression in BMDMs. Whereas leucine treatment alone decreased endogenous Gas6 protein levels, enforced Gas6 expression effectively restored Gas6 abundance even in the presence of leucine (Fig. 5H-I). Finally, we examined the impact of Gas6 on inflammatory responses associated with defective efferocytosis. Gas6 overexpression significantly suppressed the expression of pro-inflammatory cytokines, including TNF-α, IL-1β, and IL-6, while enhancing the anti-inflammatory cytokine IL-10. In contrast, leucine treatment promoted a pro-inflammatory cytokine profile. Notably, Gas6 overexpression largely mitigated the inflammatory effects induced by leucine and restored IL-10 expression (Fig. 5J). Collectively, these results identify Gas6 as an important molecular mediator linking leucine elevation to impaired macrophage efferocytosis and heightened inflammatory responses, thereby contributing to plaque destabilization under cold exposure.

Zic2 mediates leucine-induced downregulation of Gas6 expression

To investigate the mechanism by which leucine regulates Gas6 expression, we first extracted the 2,000 bp upstream promoter sequence of the Gas6 gene from the NCBI database. We then used the JASPAR, PROMO, and GTRD databases to predict potential transcription factors that could bind to this promoter region. Cross-referencing the results from all three databases identified Zic2 and Nkx2-1 as potential upstream regulators (Fig. 6A). Among the overlapping transcription factors, leucine treatment significantly reduced Zic2 expression, whereas Nkx2-1 remained unchanged (Fig. 6B). To determine whether leucine affects the binding of Zic2 to the Gas6 promoter, ChIP-qPCR analysis was performed in BMDMs. Zic2 occupancy at the predicted Gas6 promoter region was significantly reduced following leucine treatment (Fig. 6C), suggesting that leucine suppresses Gas6 transcription at least in part by reducing Zic2 promoter binding. To further evaluate the functional role of Zic2 in regulating Gas6 expression, Zic2 was silenced in BMDMs using siRNA. Knockdown of Zic2 significantly reduced Gas6 mRNA expression and largely attenuated the suppressive effect of leucine on Gas6 expression (Fig. 6D). Consistently, Gas6 protein levels were decreased following Zic2 knockdown, and leucine treatment no longer produced additional suppression of Gas6 protein expression in these cells (Fig. 6E-F). To assess whether Zic2 directly regulates Gas6 promoter activity, luciferase reporter assays were performed using constructs containing either wild-type or mutant Zic2-binding sites within the Gas6 promoter. Zic2 overexpression significantly enhanced the activity of the wild-type Gas6 promoter, whereas mutation of the predicted binding motif abolished this effect (Fig. 6G), supporting a direct transcriptional regulatory relationship between Zic2 and Gas6. Finally, rescue experiments were performed by overexpressing Zic2 in BMDMs. Forced expression of Zic2 restored Gas6 protein expression under leucine treatment conditions (Fig. 6H-I), indicating that Zic2 functions as an important upstream mediator linking leucine signaling to Gas6 downregulation.

Fig. 6.

Zic2 mediates leucine-induced downregulation of Gas6 expression. A, Venn diagram illustrating the predicted transcription factors of Gas6 obtained from three databases: JASPAR, PROMO, and GTRD, with overlapping factors shown in the intersection. B, qRT-PCR was performed to detect the mRNA expression of Zic2 and Nkx2-1 in the vehicle group and leucine group (n = 8 per group). C, ChIP–qPCR analysis showing the occupancy of Zic2 at the Gas6 promoter region in BMDMs under vehicle or leucine treatment conditions (n = 4 per group).D, qRT–PCR analysis of Gas6 mRNA expression in BMDMs transfected with control siRNA (siCtrl) or Zic2 siRNA (siZic2), followed by vehicle or leucine treatment (n = 6 per group). E–F, Representative Western blot images and quantitative analysis of Gas6 protein expression in BMDMs transfected with siCtrl or siZic2 under vehicle or leucine treatment conditions (n = 4 per group). G, Luciferase reporter assay in BMDMs transfected with wild-type (WT) or mutant (MT) Gas6 promoter reporter constructs together with control vector or Zic2 overexpression plasmid (Zic2-pcDNA3.1) (n = 4 per group). H–I, Representative Western blot images and quantitative analysis of Gas6 protein expression in BMDMs transfected with Zic2 overexpression plasmid (oeZic2) or control vector, followed by vehicle or leucine treatment (n = 4 per group). Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant

Cold exposure is associated with gut microbiota remodeling and reduced L. johnsonii abundance

To investigate the impact of cold exposure on gut microbial community structure and its association with host amino acid metabolism, we performed 16 S rRNA sequencing in RT and cold-exposed mice. At the phylum level, cold exposure induced a marked shift in microbial composition, characterized by a reduction in Firmicutes—which include taxa such as Lactobacillus—and an expansion of Bacteroidota and Verrucomicrobiota (Fig. 7A). Random forest classification further identified Lactobacillus as the top taxon discriminating cold exposure, supporting its sensitivity to environmental temperature changes (Fig. 7B). At the genus level, volcano plot analysis identified Lactobacillus as one of the most significantly decreased taxa in cold-exposed mice, whereas genera such as Edaphobaculum and Flavitalea were enriched under cold conditions (Fig. 7C). We next examined associations between microbial taxa and amino acid–related metabolites. In RT mice, Lactobacillus abundance exhibited positive correlations with branched-chain amino acid–related metabolites, including leucine-associated features (Fig. 7D), consistent with a linkage between Lactobacillus-dominant communities and basal amino acid homeostasis. In contrast, cold exposure markedly altered the microbe–metabolite correlation structure, with diminished associations involving Lactobacillus and strengthened correlations between cold-enriched genera and leucine-related metabolites (Fig. 7E), indicative of a reorganized microbiota-metabolism network.

Fig. 7.

Cold exposure is associated with gut microbiota remodeling and reduced L. johnsonii abundance in mice and humans. A, Stacked bar chart showing the relative abundance of gut microbiota at the phylum level in the RT group and cold group (n = 5 per group). B, Random forest analysis identifying the top 10 discriminatory genera between the RT and cold groups. C, Volcano plot showing differentially abundant genera between two groups. Red points represent enriched genera in the cold group; blue points denote depleted genera. D-E, Spearman correlation heatmaps showing associations between gut microbial genera and BCAA-related metabolites in the RT group and cold group. F, qRT-PCR was performed to detect the abundance of seven common Lactobacillus species in feces from ApoE^−/− mice between two groups. Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test. G, Study design schematic for human fecal sampling across healthy and ASCVD patients in summer and winter (n = 60 per group). H, qRT-PCR validation of L. johnsonii abundance in fecal samples from healthy controls (n = 60) and ASCVD patients (n = 60) collected in winter and summer. Statistical significance was assessed by Kruskal-Wallis test followed by Dunn’s multiple-comparison test with FDR correction. *p < 0.05, ***p < 0.001, ****p < 0.0001

To refine Lactobacillus alterations at the species level, we quantified seven common Lactobacillus species in mice and found that L. johnsonii exhibited the most pronounced reduction following cold exposure (Fig. 7F). To assess clinical relevance, we measured L. johnsonii abundance in fecal samples from healthy controls and ASCVD patients collected in both summer and winter (Fig. 7G). L. johnsonii levels were consistently lower in ASCVD patients than in healthy controls across seasons and displayed a seasonal pattern, with reduced abundance in winter compared with summer in both groups (Fig. 7H). Together, these findings indicate that cold exposure selectively depletes L. johnsonii and is accompanied by a reorganization of gut microbiota–leucine–associated metabolic networks, thereby suggesting that reduced L. johnsonii abundance is associated with altered host metabolic and vascular phenotypes under cold exposure.

L. johnsonii, a probiotic commonly found in the human and animal gut, is known for its anti-inflammatory, lipid-modulating, and immune-regulatory properties [30, 31]. Based on these reported functions, we next examined whether supplementation with L. johnsonii could ameliorate cold-associated atherosclerotic phenotypes. Compared with PBS-treated controls, oral administration of L. johnsonii significantly reduced atherosclerotic lesion burden in the aorta and aortic sinus of cold-exposed ApoE^−/− mice (Fig. 8A-C). This intervention was accompanied by a marked reduction in necrotic core and MOMA-2–positive areas, along with increased α-SMA–positive and Masson’s trichrome–positive areas, indicative of improved plaque stability (Fig. 8D-I). Consistent with these morphological improvements, L. johnsonii supplementation significantly decreased circulating pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) while increasing the anti-inflammatory cytokine IL-10 in the plasma of cold-exposed ApoE^−/− mice (Fig. 8J).

Fig. 8.

L. johnsonii supplementation attenuates cold-induced atherosclerotic plaque progression and instability in ApoE^−/− mice.A, Model of ApoE^−/− mice receiving PBS or L. johnsonii exposed to cold temperature for 12 weeks. B, Representative en face photographs of aortas showing Oil Red O-stained atherosclerotic plaques and quantitative data of Oil Red O-positive staining area in the cold and cold + L. johnsonii group (n = 6 per group, Scale bar: 2 mm). C, Representative Oil Red O staining of cross-sections of aortic roots and quantitative data of plaque area in each group (n = 4 per group, Scale bar: 500 μm). D-H, Representative H&E staining, MASSON staining, α-SMA immunohistochemistry staining and MOMA-2 immunohistochemistry staining of cross-sections of aortic roots and quantitative data of necrotic core area, collagen-positive area, α-SMA⁺ area, and MOMA-2⁺ area in each group (n = 6 per group, Scale bar: 100 μm). I, Quantification of the instability index in each group (n = 4 per group). J, Levels of inflammatory cytokines TNF-α, IL-1β, IL-6, and IL-10 in each group (n = 6 per group). Data are presented as mean ± s.d. P values were determined by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Notably, L. johnsonii supplementation significantly reduced plasma leucine levels in cold-exposed mice (Fig. S7A), supporting a role for this strain in modulating host leucine availability under cold conditions. Intriguingly, L. johnsonii supplementation also markedly increased the efferocytosis index in aortic lesions of cold-exposed mice (Fig. S8A-B), indicating improved macrophage-mediated apoptotic cell clearance. Given that Gas6 is a key regulator of efferocytosis that is suppressed by cold exposure and leucine supplementation, we further examined its expression dynamics. While leucine supplementation downregulated Gas6 expression, L. johnsonii administration conversely increased Gas6 levels in the aortas of cold-treated mice (Fig. S9A-D). Together, these findings suggest that L. johnsonii not only alleviates atherosclerotic lesion severity but is also associated with restoration of the Gas6–efferocytosis pathway disrupted by cold exposure.

To assess the strain specificity of these protective effects, we additionally evaluated the impact of another Lactobacillus strain, L. reuteri, on cold-induced atherosclerosis. In contrast to L. johnsonii, L. reuteri did not significantly attenuate lesion severity under cold conditions (Fig. S10A-B), suggesting a strain-specific effect. Finally, to further examine whether leucine is functionally involved in the atheroprotective effects of L. johnsonii, mice were divided into three groups: cold, cold + L. johnsonii, and cold + L. johnsonii + leucine. Leucine supplementation significantly reversed the protective effects of L. johnsonii on cold-induced atherosclerosis (Fig. S11A-D).

Given that leucine is an essential amino acid for mammals and is not considered a canonical microbial metabolite, we focused our investigation on microbiota-associated regulation of host leucine absorption and clearance rather than direct microbial leucine metabolism [32, 33]. We next examined whether intestinal leucine absorption was altered under cold exposure. Expression of the neutral amino acid transporter SLC6A19 in the small intestine was significantly increased in cold-exposed mice, whereas L. johnsonii supplementation partially normalized SLC6A19 expression (Fig. S12A). These findings suggest that altered intestinal leucine absorption may contribute to elevated circulating leucine under cold exposure. Because the liver is a major site of branched-chain amino acid catabolism, we examined the expression of key BCAA catabolic enzymes in hepatic tissue. Cold exposure was associated with increased Bckdk expression and reduced Ppm1k expression, indicating that branched-chain amino acid catabolic flux may be suppressed. Notably, L. johnsonii supplementation partially reversed these changes (Fig. S12B-C), suggesting that cold exposure may impair hepatic BCAA clearance, which could contribute to leucine accumulation. Consistent with altered intestinal leucine signaling, phosphorylation of mTORC1 downstream targets p-S6 and p-4EBP1 was increased in the small intestine of cold-exposed mice, indicating enhanced mTORC1 activity. Importantly, L. johnsonii supplementation attenuated mTORC1 activation (Fig. S12D-E), suggesting that intestinal leucine sensing is modulated by the gut microbiota under cold exposure.

Discussion

Our findings establish cold exposure as a potent environmental stressor that not only accelerates atherosclerotic plaque growth but also promotes plaque instability in ApoE⁻/⁻ mice, underscoring the vulnerability of cardiovascular homeostasis to environmental challenges. By integrating fecal microbiota transplantation and antibiotic depletion approaches, we implicate the gut microbiota as an important mediator linking cold exposure to atherosclerosis progression. Mechanistically, our data support a model in which alterations in the gut microbiota are associated with changes in host leucine metabolism and elevated circulating leucine levels, which are accompanied by suppression of Gas6 expression and impaired efferocytosis capacity in bone marrow-derived macrophages. This microbiota–leucine–Gas6 axis may represent a potential mechanistic link through which environmental cold stress contributes to defective plaque-resolving immune functions. Importantly, modulation of this axis through oral supplementation with L. johnsonii attenuated both cold-induced atherosclerosis and plaque instability in ApoE⁻/⁻ mice, whereas direct leucine supplementation exacerbated disease severity. Consistent with our experimental findings under cold exposure, we observed a reduced abundance of L. johnsonii in fecal samples from individuals with ASCVD, with the lowest levels detected in patients during winter. These observations suggest that cold-associated disruption of gut microbial homeostasis may be linked to increased cardiovascular risk.

Extensive epidemiological and experimental evidence has identified cold exposure as a major risk factor for CVD [4, 5, 7]. Consistent with these observations, our study demonstrates that cold exposure accelerates both the progression and instability of atherosclerotic plaques in ApoE⁻/⁻ mice, in line with previous reports under similar conditions [20]. Importantly, beyond confirming the deleterious impact of cold stress, our findings extend prior work by providing insight into how cold exposure promotes atherosclerosis—specifically through gut microbiota–associated metabolic alterations and impairment of plaque-resolving immune functions.

Notably, we did not observe significant changes in circulating lipid levels (TC, LDL-C, or TG) between the RT and cold-exposed groups in our model. This finding is consistent with previous studies showing that the effects of cold exposure on circulating lipid parameters are variable and dependent on experimental conditions. For example, cold exposure can induce dynamic and transient changes in lipid metabolism, particularly in triglyceride species, rather than sustained alterations in conventional lipid parameters such as TC and LDL-C [34]. In addition, studies of cold-water immersion have demonstrated that circulating lipid responses are time-dependent and heterogeneous, with some lipid parameters remaining unchanged under certain conditions [35]. Importantly, cold exposure has been shown to promote atherosclerotic progression through mechanisms beyond systemic lipid levels, including enhanced lipolysis, inflammatory activation, and metabolic remodeling [20]. Together, these findings suggest that the pro-atherogenic effects of cold exposure in our study are likely mediated through inflammation- and metabolism-related mechanisms rather than changes in circulating lipid levels.

The involvement of gut microbiota in the onset and progression of cardiovascular diseases is well recognized [9, 10]. Notably, susceptibility to atherosclerosis can be transferred between mouse models via FMT, highlighting a causal role of the gut microbiota in disease development [13]. Building on this foundation, our study demonstrates that gut microbiota critically mediates cold-associated atherosclerosis, as evidenced by the more pronounced plaque growth and instability observed in ApoE⁻/⁻ mice receiving Cold-FMT compared with those receiving RT-FMT. These findings establish gut microbiota as a key intermediary linking environmental cold stress to atherosclerotic progression. Within the cold-altered gut microbial ecosystem, our data highlight L. johnsonii as a protective bacterial species. Supplementation with L. johnsonii not only attenuated the progression of atherosclerosis but also improved plaque stability in ApoE⁻/⁻ mice under cold stress, indicating that its function extends beyond a general probiotic effect to a context-dependent regulator of vascular pathology. Consistent with this notion, previous animal and clinical studies have reported beneficial roles of L. johnsonii in metabolic and inflammatory conditions, including attenuation of renal fibrosis through aryl hydrocarbon receptor (AHR) signaling and improvement of gut microbiota composition in cancer patients [36, 37]. Together, these findings position L. johnsonii as a functionally relevant microbial modulator with potential therapeutic value in cold-related cardiometabolic disorders, potentially through its impact on host metabolic homeostasis.

Recent studies have highlighted that gut microbiota-dependent metabolites, including TMAO, SCFAs, bile acids, and BCAAs such as leucine, play critical roles in cardiovascular disease [12–15]. Our study adds to this growing body of evidence by demonstrating that cold exposure induces a specific increase in circulating leucine levels in ApoE⁻/⁻ mice, which in turn exacerbates atherosclerosis and compromises plaque stability. Unlike previous studies that have primarily linked leucine-induced atherogenesis to high-protein diets and mTORC1 activation [16], our findings suggest that reduced L. johnsonii abundance is associated with elevated plasma leucine under cold exposure, revealing a potential gut microbiota–metabolism–vascular connection independent of dietary intake. Notably, the cardiovascular effects of leucine reported in the literature appear to be highly context dependent. While moderate leucine levels have been associated with improved lipid metabolism and reduced foam cell formation, accumulating evidence suggests that excessive leucine exposure can exert pro-atherogenic effects [38, 39]. In this regard, recent studies have proposed a concentration-dependent “threshold effect” of leucine, whereby elevated leucine levels or sustained high-protein intake activate mTOR signaling in monocytes and macrophages, thereby altering immune tone and promoting inflammation [40]. Together with our findings, these observations support the notion that cold-induced leucine accumulation represents a pathological metabolic signal that contributes to immune dysregulation and atherosclerotic progression under cold stress.

Impaired efferocytosis function can prevent the effective clearance of apoptotic cells within atherosclerotic plaques. Consequently, this impairment promotes arterial inflammation, contributes to the formation of necrotic cores, increases plaque instability, and raises the risk of rupture, ultimately leading to adverse cardiovascular events [17]. Our research reveals that cold exposure diminishes the aortic efferocytosis function in ApoE^−/− mice, consequently promoting atherosclerosis. In contrast, supplementation with L. johnsonii enhances the aortic efferocytosis function in ApoE^−/− mice exposed to cold and alleviates arterial atherosclerotic lesions. In comparison to RT group ApoE^−/− mice, leucine supplementation significantly diminishes aortic root efferocytosis function and exacerbates atherosclerosis. Aligning with these findings, previous research has shown that the absence of bone marrow cell-specific SIRPα reduces inflammation and improves cellular clearance, ultimately alleviating atherosclerosis severity [41]. Similarly, another study showed that augmentation of macrophage glutamine metabolism promotes efferocytosis, leading to reduced inflammation and necrotic core formation, ultimately inhibiting atherosclerosis progression [42]. Taken together, these findings highlight the critical role of efferocytosis in cold-induced atherosclerosis progression. Therefore, targeting efferocytosis through the regulation of L. johnsonii and leucine levels may offer promising therapeutic approaches for treating cold-induced atherosclerosis.

Gas6 serves as a critical molecular link between metabolic cues and efferocytosis by functioning as a key ligand of the TAM receptor family. Through activation of downstream signaling pathways, Gas6 promotes the efficient clearance of apoptotic cells and restrains inflammatory responses, thereby contributing to plaque stability [43]. In the context of our study, Gas6 emerges as an important node connecting cold-induced leucine accumulation to impaired efferocytosis. Consistent with this model, we found that Gas6 expression was significantly reduced in the aortic tissue of cold-exposed ApoE⁻/⁻ mice compared with the RT group. In parallel, Holden et al. [44] reported a negative correlation between Gas6 levels and both maximal plaque thickness and total plaque area, supporting a protective role of Gas6 in atherosclerosis. Importantly, in vitro overexpression of Gas6 effectively reversed the inhibitory effect of leucine on efferocytosis in bone marrow–derived macrophages, providing direct evidence that Gas6 mediates leucine-induced suppression of efferocytotic capacity. The pro-efferocytotic function of Gas6 has also been demonstrated in other disease contexts. For example, administration of recombinant Gas6 in a mouse model of subarachnoid hemorrhage activated Axl and Rac1 signaling, thereby enhancing microglial efferocytosis and ameliorating blood–brain barrier disruption and neurological deficits [45]. Together, these findings position Gas6 as a key effector downstream of leucine that links metabolic perturbations to defective efferocytosis and plaque instability under cold stress.

Despite the strengths of our study, several limitations should be acknowledged. First, only male mice were included in this study, which may limit the generalizability of our findings to females. Second, although we observed an association between reduced L. johnsonii abundance and elevated circulating leucine under cold exposure, our study does not directly establish a causal mechanism by which gut microbiota regulates host leucine metabolism. Future studies using stable isotope tracing or gnotobiotic models will be required to clarify this relationship. Third, while our findings support impaired efferocytosis as a key mechanism underlying plaque instability, we cannot fully exclude the contribution of additional pathways. In particular, apoptosis and ferroptosis were not systematically investigated using dedicated mechanistic assays, and mechanisms beyond the Zic2–Gas6 axis may also participate in the regulation of macrophage efferocytosis and plaque progression.

In conclusion, this study identifies a previously unrecognized cold-associated microbiota–leucine–immune axis through which environmental cold stress accelerates atherosclerosis. We demonstrate that cold exposure disrupts gut microbial homeostasis, characterized by reduced L. johnsonii abundance and elevated circulating leucine levels, which in turn suppress Gas6 expression, impair macrophage efferocytosis, and promote atherosclerotic progression. By mechanistically linking environmental temperature to microbial metabolism and plaque-resolving immune functions, our findings expand the current understanding of how external stressors shape atherosclerotic disease biology. Importantly, this work highlights gut microbiota-associated metabolites and efferocytosis-related pathways as potential targets for risk stratification and therapeutic intervention in ASCVD.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Yongtai Gong and Sen Yan designed the research. Sen Yan, Ning Fang, Long Li, Yun Zhang, Ning Zhang and Qian Liang performed the experiments. Sen Yan, Ning Fang and Yu Duan analyzed data. Yongtai Gong and Yue Li commented on the study and revised the paper. Sen Yan wrote the manuscript with input from all authors.

Funding

This work was supported by grants from the National Key R&D Program of China (2024YFA1307001), the Key Research and Development Plan of Heilongjiang Province (JD2023SJ44), the HMU Marshal Initiative Funding (HMUMIF-22001), the National Natural Science Foundation of China (82370314), and the National Key Research and Development Project of China (2022YFC2503500).

Data availability

Sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject accession number PRJNA1394638 and will be publicly released upon publication. A preliminary version of this work has been made available as a preprint on Research Square (Gong et al. 2024) [46].

Declarations

Ethical approval and consent to participate

The study protocols were conducted in strict accordance with the Declaration of Helsinki and were approved by the Research Ethics Committee of the First Affiliated Hospital of Harbin Medical University (ethical approval number: 2022IIT094). All animal procedures in this study strictly adhered to the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care Committee at Harbin Medical University to ensure ethical treatment (Ethical approval number: 2020030).

Consent for publication

All authors have reviewed and approved the final version of the manuscript and agreed to its publication in Journal of Translational Medicine.

Competing interests

No potential conflict of interest was reported by the author(s).