Balb/c mice (6–8 weeks; GemPharmatech Co., Ltd, China) were sensitized by injections (s.c.) of HDM extract (50 μg in 100 μL sterile saline; Greer Laboratories, lot XPB82D3A25) on days 0, 7, and 14, followed by daily intranasal (i.n.) rechallenges (50 μg in 20 μL sterile saline) on days 21, 22, K23, 24, and 25. Nonsensitized mice received sterile saline only (s.c and i.n). Analysis of lung function parameters and organ collection was performed 24 h after the last challenge. All measures were taken to keep animal suffering to a minimum.

The CRISPR/Cas9 design target was designed on exon3 of the transcript GATM-201, causing frame shift mutations to achieve the purpose of inactivating protein function ( Figure S1A ). GATM-sgRNA1 targeting CCAAAGATCATCTGAAGAAGG was used for microinjection. We used PCR to identify GATM mouse genotypes ( Figure S1B ). The GATM knockout F0 generation was obtained by embryo transfer and surrogate conception, and the F1 generation was bred to determine the GATM genotype using T7E1 and sequencing ( Figure S1C ). The pure heterozygotes were further determined using Sanger sequencing results. Mice were housed in a specific pathogen-free animal facility and maintained in isolated ventilated cages on a 12-h light/dark cycle with ad libitum access to food and water.

Tissue was taken from the left lung of each mouse and fixed in 4% paraformaldehyde overnight. The tissues were embedded in paraffin and then cut into 4-μm sections. Lung tissues were stained with H&E, periodic acid-Schiff (PAS), and Masson’s trichrome to observe histological changes including inflammatory-cell infiltration into the peribronchial connective tissues, mucus secretion and fiber formation.

After blood was taken, the lungs were lavaged in situ twice with 0.8 mL sterile saline (0.9% NaCl, prewarmed), and the recovered fluid was pooled. The total number of cells in BALF were counted, and a cytospin sample was prepared. The number of total blood cells, eosinophilic granulocytes, and lymphocytes in BALF were evaluated using a hemacytometer. The BALF was centrifuged at 3000 × rpm at 4°C for 15 min to collect the supernatant. Levels of IL-4 in cell-free BALF were determined using ELISA kits according to manufacturer instruction (ELM-IL-4-1, Raybio).

Mice were sacrificed with an overdose of pentobarbital (100 mg/kg, i.p.) one day after the last airway challenge. Blood samples were taken from the retroorbital plexus/sinus, allowed to rest at room temperature for 1 hour, then centrifuged (3000 × rpm, 20 min), and supernatants were collected for detection of total IgE with an ELISA kit according to manufacturer instructions (ELM-IgE-1, Raybio).

AHR was measured using an animal pulmonary function instrument (Buxco Electronics, US). In brief, an ascending series of methacholine (6.26-50 mg/mL) was administrated into the trachea through the connected atomizer. The baseline airway resistance was evaluated using atomized PBS. The resistance index (RI) of the total lung and airway was determined as per the protocols of the instrument.

For GATM immunohistochemistry, deparaffinized lung sections were subjected to antigen retrieval, and then treated with H₂O₂ for 15 min to block endogenous peroxidase, and then incubated overnight at 4°C in recommended dilutions of anti-GATM antibodies. After washing with PBS, slices were incubated with a secondary antibody for 20 min at room temperature. Signals were visualized with DAB.

The lung tissue was fixed in 4% paraformaldehyde for 1 hour at 4°C, blocked in 20% goat serum for 2 hours at 4°C, and incubated overnight with the antibodies against GATM (GB113430, Servicebio), CD68 (GB113109, Servicebio), and Arginase 1 (GB11285, Servicebio) at 4°C, and then with fluorescein isothiocyanate-labeled secondary antibody (GB22303, Servicebio), at room temperature for 1 hour. We then used a fluorescent triple staining kit (G1236-100T, Servicebio). In addition, DAPI (G1012, Servicebio) was used to stain the nuclei and anti-fluorescence quenching sealer (G1401, Servicebio). The staining was observed under an inverted fluorescence microscope (Servicebio).

Primary AMs were isolated by repeatedly flushing mouse lungs with 1 ml of PBS eight times at room temperature to obtain BAL. Cells were centrifuged at 1,500 rpm for 5 min, washed twice using DMEM complete medium, and then cells were incubated for 3 hours in 24-well plates at a density of 3 x 10⁵ cells/well to allow cells to adhere. Next, the cells were washed twice with PBS to remove non-adherent cells for subsequent processing. To generating peritoneal macrophages, intraperitoneal injections of 3% thioglycolate broth were administered, and then cells were extracted and harvested for assays. To generate BMDMs, bone marrow cells were cultured in macrophage complete medium containing 20 ng/ml M-CSF (CB34, Novoprotein). After 6 days in culture, non-adherent cells were eliminated and adherent cells were harvested for assays. Next, the AMs, PMs and BMDMs were cultured in 1 μg/ml LPS (Sigma-Aldrich, St. Louis, MO, USA) and 20 ng/ml IFN-γ (C746, Novoprotein) for M1 polarization, or 20 ng/mL interleukin-4 (CK15, Novoprotein) for M2 polarization treatment for up to 24 hours. M1 and M2 macrophages were harvested for quantitative PCR and western blotting analysis.

For RNA interference experiments, primary cells were seeded in a 24-well plate in antibiotic-free growth medium. siRNA transfection was performed using Lipofectamine RNAi MAX reagent (13778150, Invitrogen) and Opti-MEM (Gibco, Germany) according to the manufacturer protocol. The sequences for siGATM (RIB-BIO; China) and siSTAT6 (RIB-BIO; China) were GGTCGAAGAGATGTGCAAT and GCTGATCATTGGCTTTATT, respectively; both used at a 50 nM concentration.

Primary BMDMs were incubated in 12-well plates (2 x 10⁵ cells/well) for 48 h to allow siRNA to interfere with GATM. To purify CD4⁺ T cells from the mouse spleen, spleens were dissected from healthy, female C57BL/6 mice (6–8 weeks old) under sterile conditions. A 100 μm cell strainer was placed in a sterile 10-cm dish and lymphocytes were isolated using a mouse lymphocyte separator (Dakewe, China) and CD4⁺ lymphocyte subsets were enriched on T-cell columns (Miltenyi, Germany) by negative selection according to manufacturer instructions and red blood cells lysed with ACK lysis buffer.

Mouse BMDMs and purified CD4⁺ T cells were seeded into 12‐well plates at a density of 2 × 10⁵ cells/well in lower room and 2×10⁶ cells/well and the upper room. Anti-mouse CD3 0.5 mg/l (eBioscience, America) and anti-mouse CD28 0.5 mg/l (eBioscience, America) antibodies were also added into the co‐culture system to stimulate the proliferation of CD4⁺T cells. After 72 hours of co‐culture, T cells were collected, and quantitative PCR was used to detect T-bet and IL-10.

RNA isolation from AMs, PMs, BMDMs were performed via using the Total RNA extraction Kit (B511361-0100, BBI, China). Resulting total RNA was transcribed using via a HiScript^® III RT SuperMix cDNA Synthesis kit (R223-01, Vazyme, China) according to the manufacturer instructions. Quantitative PCR was performed using via ChamQ Universal SYBR quantitative PCR Master Mix (Q711-02, Vazyme, China) on a CFX Connect™ Real-Time System from BIO-RAD (BIO-RAD). β-actin was used as a reference gene. The sequences of primers used were as follows:

Total protein was extracted using a protein extraction kit (CST, USA). The protein concentration was determined by a BCA Protein Assay kit (Beyotime, China). Cells were lysed for 10 min at 95 °C and 20 μg of protein was loaded onto 10% SDS-PAGE gel, then the separated proteins were transferred to a PVDF membrane (Merck Millipore, Bedford, MA, USA) and incubated overnight at 4 °C with primary rabbit monoclonal antibodies against iNOS (1:1000; 13120, CST, USA), rabbit monoclonal antibodies against Arginase-1 (1:1000; 93668, CST, USA), rabbit polyclonal antibodies against GATM (1:1000, PA5-109756, Invitrogen, USA), rabbit monoclonal antibodies against STAT6 (1:1000; 5397, CST, USA), and β-actin-HRP(1:3000; AB2001, ABWays), followed by anti-rabbit-IgG horseradish peroxidase-conjugated secondary antibodies (CST, USA). Signals were detected by enhanced chemiluminescence (Merck Millipore, Bedford, MA, USA) and exposed using a ChemiDoc™ MP Imager (Bio-Rad, USA). The integral optical density of each sample was measured using Image J.

Nitric oxide was detected using the Griess Reagent System (Promega, USA).

Mouse lung tissues were ground into single cell suspensions using a Miltenyi automatic tissue grinder (Miltenyi, Germany). The flow cytometry antibodies used were CD11c (117309, Biolegend, USA) and Siglec F (155509, Biolegend, USA). Cell surface staining was performed at 4°C for 30–40 min in the dark. The purity of sorted populations was >99%, unless otherwise indicated. Data were obtained on a FACSCanto II clinical flow cytometry system (BD Biosciences) and analyzed using FlowJo software (FlowJo LLC, Ashland, Oregon).

30 mg of lung tissue was lysed using CelLytic™ M (sigma, USA), centrifuged at 12000 rpm for 10 min, and the supernatant was collected for detecting 570 nm absorbance, according to the manufacturer instructions for Creatine colorimetric/fluorometric assay kit (BioVision, USA).

Total RNA was extracted from AMs and PMs from WT and GATM^-/- mice using BBI reagent kits (BBI, China). High quality total RNA was used to construct an RNA-seq library. AMs and PMs were cultured in 12-well plates with 500,000 cells/well. PMs were stimulated with LPS and IFN-γ or IL-4 for 24 h. The integrity and total amount of RNA were assessed using the RNA Nano 6000 Assay Kit from the Bioanalyzer 2100 system (Agilent Technologies, CA, USA). RNA-seq library preparation (AMPure XP system) was performed at Novogene, where sequencing was performed on an Illumina NovaSeq 6000 machine.

Data were analyzed using GraphPad Software (Prism 8.00, San Diego, CA, USA). Each experiment was performed independently in at least 3 times. Quantitative data are presented as mean ± S.D., according to the number of comparison groups, and Student’s t-tests or one-way ANOVAs were performed as required. P <0.05 was considered statistically significant. *P <0.05; **P <0.01; ***P <0.001; ****P <0.0001.

The creatine transporter protein Slc6a8 is known to activate M2 polarization by transporting exogenous creatine (22). However, the role of endogenous creatine synthesis (GATM) on macrophage polarization remains unknown. Here, to establish the macrophage polarization model, AMs, PMs and BMDMs were treated for 24 h either with LPS and IFN-γ to induce M1polarization or with IL-4 to induce M2 polarization. It is well known that DNA microarray technology is an established technique for profiling gene expression at the whole genomic level. To investigate potential polarization-relevant genes, we previously performed an Illumina microarray analysis of M1- and M2-polarized AMs (28). As expected, a great number of genes were found to be altered by polarization conditions; among these genes, we identified that GATM together with the M2 marker genes (Arg1, Mrc1, Ccl17 and Retnla) were clearly upregulated in M2-polarized AMs. In this study, we used RNA-sequence technology, which is well established for profiling gene expression of PMs at the whole transcription level, we found a decrease in GATM expression in M1 polarized macrophages. This indicates that M1 activation is negatively regulated by GATM, and is elevated in M2 polarized macrophages ( Figure 1A and Table S1 ). Of the three genes associated with the synthesis and transport of creatine (GATM, GAMT, Slc6a8), quantitative PCR and western blotting analysis showed that only the expression levels of GATM were increased following M2 polarization, indicating that GATM is involved in M2 polarization ( Figures 1B, C ). To determine the effect of creatine on M2 polarization, we added creatine to PMs and quantitative PCR analysis showed elevated expression levels of Arg1, Retnla and Ccl17, whereas there was no change in GATM expression ( Figure 1D ). To further explore the regulatory factors in GATM expression, we used RNAi technology to knockdown STAT6, which a key transcription factor of M2 polarization, and we found that GATM mRNA and protein expression was reduced in IL-4 polarized macrophages ( Figures 1E, F ). Furthermore, we analyzed lipid-metabolism-related genes through transcriptome data and found that GATM knockout significantly reduced Scd1 expression ( Figure 1G and Table S1 ). GATM knockdown also reduced Scd1 mRNA levels in M2 macrophages ( Figure 1H ). Taken together, we conclude that GATM acts as an endogenous proponent of M2 polarization and that GATM expression is regulated by STAT6. In addition, GATM knockdown inhibited lipid metabolism in M2 macrophages to some extent.

To specifically address the role of GATM in macrophage polarization, we generated mice with GATM ablation and obtained GATM^-/- mice via CRISPR/Cas9-mediated gene editing. We found that GATM expressed in M2 polarization ( Figure S2 ) and genetic ablation of GATM did not alter the level of nitric oxide ( Figure 2A ) and did not change transcript levels of Nos2, IL1b and IL12b ( Figure 2B ) or protein expression of iNOS ( Figure 2D ) in BMDMs. However, the expression of Arg1, Mrc1, Ccl17 and Retnla following IL-4-induced M2 polarization was reduced in GATM^-/- mice ( Figure 2C ) and protein expression of GATM and Arg1 ( Figure 2E ) was also reduced. Therefore, the results indicate that GATM had only a small effect on M1 macrophages, however, it did promote M2 polarization.

Creatine is a metabolite involved in cellular energy homeostasis, which facilitates the differentiation of muscle cells and neurons (29, 30). To explore the contribution of GATM in the differentiation and maturation of AMs, we used flow cytometry to analyze the well-known AM markers CD11c and Siglec F. We observed a slight decrease in the number of AMs compared to wild type ( Figure 3A ). Furthermore, we used RNA-seq to compare the difference in transcriptomic analysis between wild type and GATM^-/- AMs. In the comparative transcriptomic analysis, a total of 434 DEGs, including 349 up- and 85 down-regulated genes, were identified from comparing data between the WT and GATM^-/- groups ( Figure 3B and Table S2 ). We then used the DEGs for the bioinformatics analysis, which included Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment, and Ingenuity Pathway Analysis (IPA), to delineate the mechanism underlying the effect of GATM on macrophage polarization. In the comparison of WT and GATM^-/- groups, the GO enrichment analysis showed that the GATM deletion altered multiple biological processes related to T-cell activation, regulation of cell-cell adhesion and activation of immune response, ( Figure 3C ; Table S3 ). The KEGG pathway analysis highlighted cytokine/cytokine-receptor interaction, T-cell differentiation and viral-protein interaction with cytokines and cytokine receptors ( Figure 3D ; Table S4 ). We therefore investigated whether macrophage-intrinsic GATM dampens pro-inflammatory T-cell responses using in vitro co-culture of BMDMs with splenocytes. After 3 days of co-culture, GATM knockdown in BMDMs distinctly promoted the expression of T-bet and IL-10 compared to CD4⁺ T cells ( Figure S3 ). IL-10 production by Treg cells is critical for the control of immune response and increases allergy-induced lung inflammation and hyperreactivity in Il10^flox/flox×Foxp3^YFP-Cre mice (31). T-Bet expression acquisition by Tregs is necessary for the control of type 1 inflammation in vivo ( 32). Furthermore, we found that GATM^-/- inhibited the expression of Arg1, Mrc1, Ccl17, Retnlain AMs ( Figure 3E ), and reduced the expression of Arg1 protein after GATM interference with AMs using RNAi technology ( Figure 3F ). Taken together, our results suggest that GATM promoted M2 polarization and altered clusters of functional genes involved in the immune response.

Asthma is a typical Th2-associated inflammatory disease in which macrophages exhibit M2 polarization (33). To link GATM expression with disease condition, we generated an acute HDM-induced murine model of asthma as a proxy for the acute inflammatory phenotype found in humans (34) ( Figure 4A ). As expected, airway exposure to HDM induced increased airway hyperresponsiveness to methacholine ( Figure 4B ), the majority of serum IgE and BAL IL-4 raised ( Figure 4C ). Examination of lung tissue revealed epithelium injury and smooth muscle thickening of the airway, and peribranchial inflammation with copious amounts of eosinophil and macrophage infiltration in the airway lumen were evident ( Figures 4D-F ). BAL cells had eosinophils and lymphocytes and other factors consistent with other HDM-induced asthma models ( Figure 4G ), indicating that the characteristics of chronic airway inflammation and airway remodeling are well demonstrated in this model. In addition, we detected that the transcriptional level of IL-4, IL-5 and IL-13 were significantly expressed in the lung tissue of asthmatic mice ( Figure 4H ). Studies have reported that creatine aggravates the inflammatory response in asthma (35, 36). Our results showed that the concentration of creatine in the lung tissue of asthmatic mice was higher than in the control group, indicating that creatine may be involved in the development of asthma ( Figure 4I ). Western blotting and immunohistochemistry of GATM showed that there was relatively high expression in lung tissue, and it appeared that HDM sensitization and challenge altered GATM expression in the lungs ( Figure 4J and Figure S4 ). Consistent with this, immunofluorescence revealed that HDM exposure raised GATM expression across the whole lung ( Figure 4K ). Therefore, GATM is involved to some extent in the development of asthma.

To investigate whether the upregulated GATM in asthmatic lung tissue was related to macrophage polarization, we analyzed GATM colocalization with M1 (CD68, iNOS) and M2 (CD68, Arg1) markers in lung tissue using confocal microscopy, and found that GATM was upregulated in M2 macrophages but not M1 polarized macrophages ( Figure 5A and S4A ). Furthermore, we isolated AMs using BAL and cultured them for 3 hours to adhere to the well walls, then observed them under the microscope and found that that the majority of macrophages in the asthma group exhibited a shuttle-like shape, while the control group alveolar macrophages still retained a round shape ( Figure 5B ). We detected that mRNA expression of Arg1, Retnla and GATM was upregulated in AMs, but not in GAMT, Slc6a8, Nos2, IL1b, or IL12b ( Figures 5C, D and Figure S4B ). Moreover, western blotting verified the upregulation of GATM and Arg1 protein levels in these cells (Figure 5E). Overall, our present study suggests that GATM may contribute to the progression of asthma via modulation of M2 macrophage polarization.