Deciphering and verifying cuproptosis-associated hub genes in diabetic foot ulcer by combining single-cell and bulk RNA-sequencing

Deciphering and verifying cuproptosis-associated hub genes in diabetic foot ulcer by combining single-cell and bulk RNA-sequencing | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Deciphering and verifying cuproptosis-associated hub genes in diabetic foot ulcer by combining single-cell and bulk RNA-sequencing Mingwei Chen, Jianran Sun, Yichang Liu, Ying Tang, Xiaotong Zhao, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5336506/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Prior studies have indicated elevated plasma copper levels in patients with diabetes compared with healthy controls. However, cuproptosis remains unexplored in diabetic foot ulcers (DFUs). DFU-associated datasets (GSE165816, GSE80178, and GSE134431) were obtained from the Gene Expression Omnibus (GEO) database. Cuproptosis-associated genes (CRGs) were identified by combining single-cell and bulk RNA-sequencing data. Differentiation, enrichment, network, pseudotime, immune infiltration, cellular communication, drug sensitivity, and molecular docking analyses were performed. The CRGs were verified in a Wistar DFUs rat model. Four hub genes were obtained (DCN, IGF1, CXCL12, and CXCL8). Enrichment analysis indicated that these genes were involved primarily in cytokine storms. Moreover, network analysis revealed the relationships among competing endogenous RNAs, transcription factors, single nucleotide polymorphisms, and hub genes. In addition, pseudotime analysis revealed greater numbers of plasma cells, naive B cells, and CD4 + T cells in DFUs than controls. Furthermore, immune infiltration analysis indicated immune cells dysregulation in DFUs, characterized by lower numbers of activated mast cells, activated NK cells, and M1 macrophages than those in controls. In addition, cellular communication analysis revealed that mesenchymal stem cells frequently interacted with fibroblasts, keratinocytes, endothelial cells, and T cells. The nomogram indicated that four hub genes were included for diagnosis of DFUs and the DFU risk was approximately 0.86. Finally, drug sensitivity analysis and molecular docking demonstrated that sirolimus was an effective drug for DFU treatment. Together, our findings link IGF1 and CXCL12 to cuproptosis, thus providing novel insights for DFU diagnosis and treatment. Health sciences/Endocrinology/Endocrine system and metabolic diseases/Diabetes/Diabetes complications Health sciences/Diseases/Endocrine system and metabolic diseases/Thyroid diseases Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 INTRODUCTION Diabetic foot ulcers (DFUs), which affect approximately 18.6 million people worldwide, are a concerning condition in patients with diabetes mellitus (DM) and may result in vascular lesions or neuropathic disorders. Indeed, greater than 15% of patients with DM have been estimated to develop DFUs within their lifetimes [ 1 ]. The 5-year mortality rate for DFUs is nearly 30%, which is higher than many cancers. In addition, more than 70% of patients are hospitalized for amputations, which negatively impacts their physical fitness and quality of life [ 2 ]. With the increasing prevalence of DM, DFUs pose a considerable burden on universal healthcare systems. Disbursements for DFU treatment have been estimated to be 9–13 billion USD/year; thus, DFU may be one of the most exorbitant diabetic complications [ 3 ]. Single-cell RNA-sequencing (scRNASeq) aids in discerning cellular function and pathogenesis, capturing the transcriptomes of discrete cells in diverse tissues [ 4 ]. scRNASeq is widely used to delineate the neoplastic microenvironment in the complex ecosystems of heterogeneous cancers, and to reveal molecular mechanisms as well as drug targets [ 5 ]. The Human Cell Atlas Project aims at describing all cell types in humans through diverse molecular characterization methods, by coupling gene expression profiles with cellular orientation and morphology using scRNASeq. The micronutrient copper (Cu), a crucial catalytic cofactor in extensive biological processes, is necessary in microorganisms and mammals. Cu acts as a catalyst in mitochondrial respiration, cellular metabolism, and oxygen transportation. Yet, intracellular Cu levels are maintained at extremely low concentrations. The average adult man incorporates only 100 mg of Cu through homeostatic mechanisms to prevent the aggregation of free intracellular Cu, which is detrimental to cells. Cuproptosis, a ground-breaking mode of cell death relying on Cu, is provoked by excess Cu 2+ , unlike ferroptosis and autophagy. The mechanism underlying cuproptosis has been reported that copper directly binds to lipoylated constituents of the tricarboxylic acid cycle [ 6 ]. The relationship between diabetic complications and plasma Cu is controversial according to prior studies. An investigation indicates lower concentrations of plasma Cu in patients with DFUs than patients with diabetes but without ulcers. A deficiency of Cu exacerbates glycemic control, which triggers temporized healing of DFUs [ 7 ]. Another study treats Cu overload with the chelator trientine, thereby expediting urinary excretion of Cu through formation of a Cu 2+ -trientine complex, and mitigating diabetic cardiovascular disease [ 8 ]. However, whether cuproptosis participates in DFUs is unclear. Therefore, we combined scRNASeq with bulk RNASeq to identify relevant genes to develop an innovative treatment for DFU. MATERIALS AND METHODS The Materials section is described in detail in the Supplementary Materials. Data extraction DFU-associated datasets were extracted from Gene Expression Omnibus (GEO) records GSE165816, GSE80178, and GSE134431. Other cuproptosis-associated genes (CRGs) were identified from the literature, such as ATP7A, ATP7B, CDKN2A, DBT, DLAT, DLD, DLST, DLTA, FDX1, GCSH, GLRX5, GLS, GSS, ISCA2, LIAS, LIPT1, MTF1, NDUFA1, NDUFB2, NDUFC1, NFE2L2, PDHA1, PDHB, SLC31A1, and TIMMDC1 [ 9 – 12 ]. Identification of hub genes The Seurat package was used to filter low-quality cells, and cell subpopulations were annotated with the SingleR package to identify dissimilar cell types. The marker genes for each cell type were analyzed to identify differentially expressed genes (DEGs), denoted DEGs1. Subsequently, DEGs from diverse cell types were aggregated according to CRG scores and denoted DEGs2. GSE80178 was analyzed, and the DEGs were denoted DEGs3. Subsequently, an intersection of DEGs1, DEGs2, and DEGs3 was generated to determine candidate genes. Hub genes were identified through analysis of the protein-protein interaction network. Enrichment analysis GO and KEGG enrichment analyses were performed for 79 candidate genes. GeneMANIA, gene set enrichment analysis (GSEA), gene set variation analysis (GSVA), and ingenuity pathway analysis (IPA) were conducted for four hub genes. Creation of the regulatory network To further examine the underlying mechanisms of crux genes participating in DFUs, transcription factor (TF)-gene, competing endogenous RNA (ceRNA), and miRNA-single nucleotide polymorphism (SNP)-mRNA networks were built. The Drug-Gene Interaction Database and Comparative Toxicogenomics Database was used to predict drugs and diseases correlating with hub genes. Single cell trajectory analysis Initially, cell types were annotated with UMAP for dimensionality reduction. The monocle2 package was used to analyze the pseudotime trajectory and cell distribution of four hub genes. The Differentialgenetest function was used to explore gene dynamics in cell differentiation. Immune infiltration and cell-cell communication The correlation between CRGs and immune infiltration was assessed with the psych and ggcorrplot packages. The ggplot2 package was used for visualization. With the rms package (version 6.5.0), a nomogram was established. The CellChat package (version 1.6.1) was used to analyze signaling interactions. Drug sensitivity analysis and molecular docking The molecular structures were searched in the PubChem database. Potential drug responses were predicted with the pRRophetic package. Creation of a DFU model A total of 18 male Wistar rats (age approximately 4 weeks, weight 230–250 g) were purchased from Jinan Pengyue Experimental Animal Breeding Ltd. (license ID: SCXK2022-0006, Jinan, China). To create DM and control models, rats were injected with streptozotocin and citrate buffer, respectively. Rats with a fasting blood glucose concentration > 16.7 mmol/L for 3 days and DM symptoms were considered as type 2 diabetes (T2DM) model. After 1 week, rats with T2DM were anesthetized with 10% chloral hydrate. DFU and non-diabetic foot ulcer (NFU) models were established by removal of a layer of skin on the instep of the foot. Hematoxylin-eosin and Masson’s trichrome staining Tissues were fixed with 4% paraformaldehyde. Paraffin-embedded sections were cut (4.0 µm thick) for further staining, then stained with hematoxylin for 3–5 minutes and eosin for 2–5 minutes. Masson’s trichrome staining was performed according to the manufacturer's instructions (Solarbio, Beijing, China). Immunohistochemistry Slides were incubated with primary antibodies overnight at 4°C, after which slices were incubated with secondary antibody. RT-qPCR and WB RT-qPCR and WB were performed (primer sequences in Table S1). Details are provided in the Supplementary Materials. Statistical analysis R software and SPSS 25.0 were used for statistical analysis. The skew distributional data were expressed as M [ P 25 , P 75 ] from three separate experiments. Differences between two groups were assessed with Wilcoxon rank sum test when the distribution of two samples was skewed. Mean optical density (MOD) = sum integrated optical density/sum area. A two-tailed P < 0.05 indicated a statistically significant difference. RESULTS Validation of DEGs The research flowchart was illustrated in Fig. 1. After quality control (Fig. S1A-B), the number of cells was 57,302. After raw data were processed with normalization, the top 2,000 HVGs were determined (Fig. S2A). Consequently, 30 principal components were identified (Fig. S2B). A total of 24 cell clusters were classified (Fig. S2C-D). Through difference analysis, a total of 1,273 DEGs1 were discovered, in which cell types were labeled with marker genes (Fig. S2E). Furthermore, CRG scores between DFUs and controls were distinct in the remaining ten cell types (Fig. S2F). In addition, the cells were allocated to high and low CRG score groups according to the median CRG score (Fig. S2G), and 685 DEGs2 were identified (Fig. S1C-L). In analysis of the training set, 2,652 DEGs3 were identified (Fig. S3A-B). By intersecting DEGs1, DEGs2, and DEGs3, 79 candidate genes were compiled (Fig. S3C). The top eight GO terms are shown in Fig S3D. In total, 256 GO terms were identified (Table S2). The top seven KEGG pathways are shown in Fig. S3E, and the total terms are indicated in Table S3. Finally, a protein-protein interaction network was constructed (Fig. S3F). Discovery of hub genes With seven algorithms, four hub genes were detected (Fig. 2A). Subcellular localization suggested that the proteins translated by hub genes were located primarily in the cytoplasm (Fig. 2B). GeneMANIA suggested that hub expression genes and co-expression genes (CXCL8, PF4, and CCL2) had mutual functions (Fig. 2C). GSEA revealed that CXCL2, DCN, and IGF1 were enriched in spliceosome, RNA degradation, and cell cycle in the high expression group, whereas CXCL8 participated in the above-mentioned pathways in the low expression group (Fig. 2D-F). GSVA indicated that CXCL2, DCN, and IGF1 were associated with graft versus host disease and asthma in the high expression group, whereas CXCL8 had roles in those pathways in the low expression group (Fig. 2G-I). IPA demonstrated that hub genes were markedly enriched in five pathways, three of which were activated and two of which were inhibited (Fig. 2J). Regarding disease and function, the hub genes were associated with functions including extracranial solid tumor, tumor incidence, and apoptosis (Fig. 2K). Network and expression levels of hub genes Initially, the TF-gene network contained 161 nodes and 209 edges (Fig. S4A). Moreover, the ceRNA network consisted of 3 mRNAs, 11 miRNAs, and 24 lncRNAs (Fig. S4B-C). Furthermore, the miRNA-SNP-mRNA network included 22 SNP locations (Fig. S4D; details in Table S4). A disease-hub gene-drug network comprised 105 nodes and 278 edges (Fig. S4E). The training and validation sets indicated that CXCL12, DCN, and IGF1 were downregulated, whereas CXCL8 was upregulated in the DFU group compared with the control group (Fig. S4F-G). Pseudotime analysis In B cells, two distinct subpopulations existed (Fig. 3A). Principal component analysis revealed that the value of characteristic number was 30, after which a plateau was reached (Fig. 3B). In addition, B cell differentiation patterns showed similar trajectories between the DFU and control groups, although slight deviations were observed at several crucial points, such as bifurcation and branch ends. In the middle and late stages, B cells were far more abundant in DFU samples than control samples (Fig. 3C). Three stages of differentiation of B cells were observed (Fig. 3D). The expression of key genes during development suggested that the CXCL8 and DCN genes showed a rising trend, then reverted to stable expression levels, whereas the expression of CXCL12 and IGF1 remained generally stable (Fig. 3E). Among T cells, five subpopulations were observed (Fig. 3F). The characteristic number value was nearly 30 (Fig. 3G). The trajectory of T cells was like an arc during cellular differentiation. Initially, T cells were assembled on the left side of this plot. As the pseudotime progressed, T cells gradually gathered towards the right side of the picture. In the late period, the number of T cells in DFU samples was greater than the controls samples (Fig. 3H). Similarly, T cells evolved into five stages in both groups (Fig. 3I). CXCL12 and DCN expression declined initially, then became constant, whereas CXCL8 and IGF1 expression did not change (Fig. 3J). Subtype analysis revealed that the numbers of plasma cells and naive B cells were higher, and the numbers of memory B cells were lower in DFUs than controls (Fig. 4A-F). In the late stage of differentiation, the number of CD4 + T cells was significantly greater in DFUs than controls (Fig. 4G-L). Moreover, DCN was highly expressed in fibroblasts, chondrocytes, and tissue stem cells; CXCL12 was highly expressed in endothelial cells and MSCs; CXCL8 was abundant in monocytes; and IGF1 was abundant in fibroblasts (Fig. S5). Immune infiltration analyses and cellular communication The CIBERSORT approach indicated higher numbers of naive B cells, monocytes, activated myeloid dendritic cells, and resting NK cells, but lower numbers of activated mast cells, activated NK cells, and M1 macrophages in DFUs than controls (Fig. 5A-B). Furthermore, Spearman correlation analysis indicated that hub genes and differentially expressed (DE) cells were correlated (Fig. 5C), verified CRGs and DE cells were associated (Fig. 5D), in addition, the relationship between our CRGs and verified CRGs (Fig. 5E). In the nomogram (Fig. 5F), four hub genes were included for diagnosis of DFUs; the total point values were nearly 120; and the DFU risk was approximately 0.86. Our findings suggested that multiple cells communicated with each other continuously (Fig. 5G-H). MSCs frequently interacted with fibroblasts, keratinocytes, endothelial cells, and T cells (Fig. 5I). The ligands and receptors were MIF and CD74 + CXCR4, and MIF and CD74 + CD44, respectively. The corresponding cells were macrophages and T cells (Fig. 5J). Drug sensitivity analysis and molecular docking Rapamycin was predicted to be the most effective drug for patients with DFUs (Fig. 6A). The binding energy of sirolimus to DCN was − 7.5 kcal/mol (Fig. 6B-D), that of chlorambucil to CXCL12 was − 5.2 kcal/mol (Fig. 6E-G), and that of foscarnet to CXCL8 was − 4.0 kcal/mol (Fig. 6H-J; details in Table S5). Experimental validation Compared with the control group, DFU rats had lower weight, higher blood glucose levels, and poorer wound healing (Fig. 7A-D), RT-qPCR and WB demonstrated lower CXCL12 and IGF1 expression (Fig. 7E-I). Because the RT-qPCR findings for DCN and CXCL8 in the preliminary animal experiment were not consistent with the bioinformatics findings, subsequent experiments were not performed. HE staining indicated that rats with DFUs, compared with controls, had greater aggregation of neutrophils and lymphocytes along with nascent capillaries (Fig. 8A-B). Masson’s trichrome staining revealed a few collagen fibers in DFU rats (Fig. 8C-D). IHC showed lower protein expression of CXCL12 and IGF1 in DFUs than controls (Fig. 8E-J). DISCUSSION Our study demonstrated diminished DCN in DFUs, thus corroborating that DCN facilitates angiogenesis [ 13 ]. DCN, also known as decorin, a molecule in the small leucine-rich proteoglycan family, has known functions in fibrosis and malignancy. Initially, decorin is discovered as a collagen-bound member that constrains fibrillogenesis as a configurational constituent of the substrate. Subsequent research shows that decorin is oncosuppressive during cancer formation, progression, and metastasis [ 14 ]. Moreover, we observed that DFUs had lower levels of IGF1 than controls, thus supporting evidence that deficiency of IGF1 delays wound healing [ 15 ]. IGF1 is one of the two isoforms of IGF in mammals. This chemokine in endothelial cells facilitates keratinocyte colonization and re-epithelialization. In addition, CXCL8, also known as IL-8, is a pro-inflammatory cytokine that we observed to be upregulated in DFUs. This cytokine attracts neutrophils to injury sites, thus facilitating healing, such as by secreting antimicrobial peptides and creating neutrophil extracellular traps that kill or immobilize bacteria [ 16 ]. Furthermore, we found that CXCL12, also designated as SDF-1, was down-regulated, in line with findings suggesting that DFU-originating fibroblasts excrete a lower concentration of CXCL12 than natural fibroblasts [ 17 ]. Our enrichment analysis revealed that CRGs activated cytokine storm signaling pathways. A critical function of growth factors (GFs [e.g., PDGF, FGF, and VEGF]), which is a classification of cytokines, in wound healing is their stimulation of angiogenesis, thereby triggering perpetual vascularization to revitalize ischemic tissues. Early in wound closure, PDGF attracts fibroblasts to lesion sites, and acts as a mitogen stimulating the transformation of fibroblasts into myofibroblasts, thus resulting in shrinkage of the vulnus. PDGF is the first topical agent gaining regulatory approval to boost wound healing [ 18 ]. FGF has stronger effects than PDGF and VEGF in expediting angiogenesis and generating granulation tissue that pads gap of wound; however, FGFs have several drawbacks, including degradation due to the proteolytic conditions in wounds [ 19 ]. VEGF, which is secreted by endothelial cells, pericytes, is regulated by the HIF-1α pathway. In addition, VEGF stimulates collagenases to degrade the basement membrane and simultaneously potentiates epithelialization [ 20 ]. Our TF network indicated that CXCL12 interacted with the TF ID1, in agreement with a study that signals for fibrotic transformation of the hepatic vascular niche are derived from CXCL12 and its receptors (CXCR7 and CXCR4) in liver sinusoidal endothelial cells (LSECs). After acute impairment, CXCR7 and CXCR4 activate ID1 to prompt LSECs to secrete paracrine growth regulators to restore liver. Moreover, our ceRNA network results demonstrated that the lncRNA TUG1 acted as a ceRNA for miR-1-3p, whose target gene was IGF1, in agreement with prior literature [ 21 ]. TUG1 acts as a ceRNA, as substantiated by a host of studies in tumors, with roles including regulating the miR-142/ZEB2 axis in bladder cancer [ 22 ], miR-221/PTEN axis in lung cancer [ 23 ], and miR-219/FMNL2 axis in oral squamous cell carcinoma [ 24 ]. Our pseudotime analysis indicated far more mid-late B cells in DFU samples than controls, in agreement with B cells expediting wound healing [ 25 ]. Because the skin is a protective barrier facing multiple injuries daily, an urgent need exists to repair wounds in which B cells reside. For example, mice lacking the B cell marker, CD19, have delayed wound healing, whereas upregulation of CD19 is beneficial for wound closure [ 26 ]. In the late period, the number of T cells in DFU samples was greater than control samples in contrast to a study in which patients with chronic DFUs retain a similar number of naive and effector T cells compared to controls [ 27 ]. Skin-resident T cells, which are comprised of αβ and γδ T cells, have a role in would repair and are involved in cutaneous infections preceding adaptive immunity [ 28 ]. Epidermal T cells interact with keratinocytes to generate an IL-15-IGF anulus, thereby potentiating IGF-1 expression and re-epithelialization [ 29 ]. In addition, our T cell subtype analysis indicated that the numbers of CD4 + T cells were significantly greater in DFUs than controls in the late stage of differentiation. However, the effect of CD4 + T cells on wound healing is disputed: one article has indicated that depletion of CD4 + T cells is detrimental for healing, whereas another study has reported that CD4 + T cell depletion has no influence on wound closure [ 30 ]. Our immune infiltration analysis indicated fewer activated mast cells, activated NK cells, and M1 macrophages in DFUs than controls, thus supporting that DFUs involve immune cells dysregulation due to chronic inflammation [ 31 ]. In the proliferative stage, macrophages transform from a pro-inflammatory to a pro-healing phenotype. However, DFUs are deficient in the M1-to-M2 switch, thus resulting in a paucity of GFs, which are indispensable for remodeling. Furthermore, mast cells aid in scar formation, by modulating the conversion from scarless to fibrotic closure [ 32 ]. In addition, activated NK cells secrete perforins, granzymes, and interferon-γ, thereby intensifying M1 macrophage polarization [ 33 ]. Notably, our cellular communication experiments indicated that MSCs frequently interacted with fibroblasts, keratinocytes, endothelial cells, and T cells. The MSCs arise from the bone marrow, umbilical cord, adipose tissue, and placenta, which secrete GFs that support wound healing [ 34 ]. Intriguingly, the MSCs and fibroblasts appear to have overlapped immunophenotypes and differentiation capabilities, MSCs might be immature fibroblasts [ 35 ]. In addition, MSCs can be converted into keratinocytes and endothelial cells, particularly in ulcerated wounds, thus leading to the secretion of cytokeratin 19 by keratinocytes and the induction of angiogenesis [ 36 ]. Moreover, MSCs regulate immunity by decreasing the ratios of pro-inflammatory T cells, such as Th17 and Th1; increasing the numbers of immunosuppressive T cells, such as Treg cells; and stimulating the formation of M2 macrophages, thereby alleviating inflammation and accelerating wound healing [ 37 ]. Our drug sensitivity analysis and molecular docking demonstrated that foscarnet, chlorambucil, and sirolimus were effective drugs for DFU treatment. Notably, rapamycin, also termed as sirolimus, is an autophagy activator and immunosuppressant. It is reported that Pseudomonas aeruginosa curb autophagy in wounds. However, rapamycin activates autophagy, thereby increasing microorganism removal [ 38 ]. Herein, a conundrum may exist. On the one hand, the molecular structure of IGF1 does not match to a suitable drug, although trofinetide, an analog of IGF1, has class I evidence for Rett syndrome treatment [ 39 ]. On the other hand, no literature basis exists for DFU therapy with the other two drugs. Molecular docking largely considers binding energies, based on the reference drug set. If any drugs are used for molecular docking, the binding energy will exceed 5 kcal/mol. Second, molecular docking forecasts a theoretical scenario, and clinical experiments are required to verify the true efficacy. Finally, compared with traditional remedies for DFUs, including debridement and wound dressing, MSCs are a more practical treatment approach for the supervision of persistent wounds because MSCs restrain pro-inflammatory cytokines and stimulate the secretion of GFs, such as FGF and VEGF [ 34 ]. Furthermore, ON101 is the first ratified macrophage-modulation drug for DFUs treatment, expediting the M1-to-M2 transition [ 40 ]. Although examining CRGs in DFUs is highly challenging, this research has several limitations. First, because of the finite sample size of the DFU transcriptome datasets, ROC curve analyses and machine learning were not performed. In addition, the roles of CRGs in DFUs remain recondite and will be accurately elucidated in our future mechanistic research. CONCLUSIONS Multiple genes were screened through bioinformatics and animal experiments. CRGs in DFUs were identified, notably DCN, IGF1, CXCL12, and CXCL8. Findings in a rat model further validated two downregulated genes: CXCL12 and IGF1. The nomogram proves diagnostic value of CRGs. Additionally, rapamycin is a predictive drug for DFU therapy. Declarations DATA AVAILABILITY The data used and/or analyzed during this research are accessible from the corresponding author on reasonable request. ACKNOWLEDGEMENTS None. AUTHOR CONTRIBUTIONS JS conceived this study. JS, LY, YT, XZ, and DD performed the experiments and analyzed the data. JS wrote the draft. SY revised the manuscript. MC participated in supervision. All authors approved the manuscript for publication. FUNDING This research was funded by the Natural Science Foundation of Anhui Province (2108085MH269) and the Natural Science Research Project of Colleges and Universities (KJ2021A0274). COMPETING INTERESTS The authors declare no competing interests. ETHICS APPROVAL Animal experiments complied with the ARRIVE guidelines 2.0, and the research protocol was approved by the Institutional Animal Ethics Committee (ethical number: 2023-NA-44). CONSENT FOR PUBLICATION All authors have authorized publication. References Mariadoss AVA, Sivakumar AS, Lee CH, Kim SJ (2022) Diabetes mellitus and diabetic foot ulcer: etiology, biochemical and molecular based treatment strategies via gene and nanotherapy. Biomed Pharmacother 151:113134 Armstrong DG, Tan TW, Boulton AJM, Bus SA (2023) Diabetic foot ulcers: a review. JAMA 330:62–75 Rice JB, Desai U, Cummings AK, Birnbaum HG, Skornicki M, Parsons NB (2014) Burden of diabetic foot ulcers for medicare and private insurers. Diabetes Care 37:651–658 Theocharidis G, Thomas BE, Sarkar D, Mumme HL, Pilcher WJR, Dwivedi B et al (2022) Single cell transcriptomic landscape of diabetic foot ulcers. Nat Commun 13:181 Stubbington MJT, Rozenblatt-Rosen O, Regev A, Teichmann SA (2017) Single-cell transcriptomics to explore the immune system in health and disease. Science 358:58–63 Chen L, Min J, Wang F (2022) Copper homeostasis and cuproptosis in health and disease. Signal Transduct Target Ther 7:378 Yadav C, Srikantiah RM, Manjrekar P, Shenoy MT, Chaudhury D (2020) Assessment of mineral pathophysiology in patients with diabetic foot ulcer. Biol Trace Elem Res 195:366–372 Cooper GJ, Chan YK, Dissanayake AM, Leahy FE, Keogh GF, Frampton CM et al (2005) Demonstration of a hyperglycemia-driven pathogenic abnormality of copper homeostasis in diabetes and its reversibility by selective chelation: quantitative comparisons between the biology of copper and eight other nutritionally essential elements in normal and diabetic individuals. Diabetes 54:1468–1476 Zhu Y, Chen X, Geng S, Li Q, Li Y, Yuan H et al (2023) Identification of the cuproptosis-related hub genes and therapeutic agents for sarcopenia. Front Genet 14:1136763 Chen Y, Zou X, Ma M, Liu Y, Wang R, Dai Z et al (2023) Expression profiles of cuproptosis-related genes determine distinct subtypes of pancreatic ductal adenocarcinoma. Curr Oncol 30:1648–1662 Chen Y, Tang L, Huang W, Abisola FH, Zhang Y, Zhang G et al (2023) Identification of a prognostic cuproptosis-related signature in hepatocellular carcinoma. Biol Direct 18:4 Guo Z, Liu J, Liang G, Liang H, Zhong M, Tomlinson S et al (2024) Identification and validation of cuproptosis-related genes in acetaminophen-induced liver injury using bioinformatics analysis and machine learning. Front Immunol 15:1371446 Chen F, Lai J, Zhu Y, He M, Hou H, Wang J et al (2020) Cardioprotective effect of decorin in type 2 diabetes. Front Endocrinol (Lausanne) 11:479258 Qin P, Chen H, Wang Y, Huang L, Huang K, Xiao G et al (2023) Cancer-associated fibroblasts undergoing neoadjuvant chemotherapy suppress rectal cancer revealed by single-cell and spatial transcriptomics. Cell Rep Med 4:101231 Yu F, Gong D, Yan D, Wang H, Witman N, Lu Y et al (2023) Enhanced adipose-derived stem cells with IGF-1-modified mRNA promote wound healing following corneal injury. Mol Ther 31:2454–2471 Mohsin F, Javaid S, Tariq M, Mustafa M (2024) Molecular immunological mechanisms of impaired wound healing in diabetic foot ulcers (DFU), current therapeutic strategies and future directions. Int Immunopharmacol 139:112713 Sorooshian P, Metcalfe AD, Lali FV (2021) In vitro modelling of disease-induced changes in the diabetic wound fibroblast. J Wound Care 30:300–303 Gu X, Han YY, Yang CY, Ji HM, Lan YJ, Bi YQ et al (2021) Activated AMPK by metformin protects against fibroblast proliferation during pulmonary fibrosis by suppressing FOXM1. Pharmacol Res 173:105844 Liu Y, Liu Y, Deng J, Li W, Nie X (2021) Fibroblast growth factor in diabetic foot ulcer: Progress and Therapeutic Prospects. Front Endocrinol (Lausanne) 12:744868 Kerstan A, Dieter K, Niebergall-Roth E, Klingele S, Junger M, Hasslacher C et al (2022) Translational development of ABCB5(+) dermal mesenchymal stem cells for therapeutic induction of angiogenesis in non-healing diabetic foot ulcers. Stem Cell Res Ther 13:455 Tang K, Lv D, Miao L, Mao Y, Yu X (2022) LncRNA TUG1 functions as a ceRNA for miR-1-3p to promote cell proliferation in hepatic carcinogenesis. J Clin Lab Anal 36:e24415 Liu Q, Liu H, Cheng H, Li Y, Li X, Zhu C (2017) Downregulation of long noncoding RNA TUG1 inhibits proliferation and induces apoptosis through the TUG1/miR-142/ZEB2 axis in bladder cancer cells. Onco Targets Ther 10:2461–2471 Guo S, Zhang L, Zhang Y, Wu Z, He D, Li X et al (2019) Long non-coding RNA TUG1 enhances chemosensitivity in non-small cell lung cancer by impairing microRNA-221-dependent PTEN inhibition. Aging 11:7553–7569 Yan G, Wang X, Yang M, Lu L, Zhou Q (2017) Long non-coding RNA TUG1 promotes progression of oral squamous cell carcinoma through upregulating FMNL2 by sponging miR-219. Am J Cancer Res 7:1899–1912 Moore EM, Maestas DR Jr., Cherry CC, Garcia JA, Comeau HY, Davenport Huyer L et al (2021) Biomaterials direct functional B cell response in a material-specific manner. Sci Adv 7:eabj5830 Iwata Y, Yoshizaki A, Komura K, Shimizu K, Ogawa F, Hara T et al (2009) CD19, a response regulator of B lymphocytes, regulates wound healing through hyaluronan-induced TLR4 signaling. Am J Pathol 175:649–660 Moura J, Rodrigues J, Goncalves M, Amaral C, Lima M, Carvalho E (2017) Impaired T-cell differentiation in diabetic foot ulceration. Cell Mol Immunol 14:758–769 Reider IE, Lin E, Krouse TE, Parekh NJ, Nelson AM, Norbury CC (2024) gammadelta T Cells mediate a requisite portion of a wound healing response triggered by cutaneous poxvirus infection. Viruses 16:425 Wang Y, Bai Y, Li Y, Liang G, Jiang Y, Liu Z et al (2017) IL-15 Enhances Activation and IGF-1 production of dendritic epidermal t cells to promote wound healing in diabetic mice. Front Immunol 8:1557 Davis PA, Corless DJ, Aspinall R, Wastell C (2001) Effect of CD4(+) and CD8(+) cell depletion on wound healing. Br J Surg 88:298–304 Qi X, Cai E, Xiang Y, Zhang C, Ge X, Wang J et al (2023) An immunomodulatory hydrogel by hyperthermia-assisted self-cascade glucose depletion and ROS scavenging for diabetic foot ulcer wound therapeutics. Adv Mater 35:e2306632 Dong J, Chen L, Zhang Y, Jayaswal N, Mezghani I, Zhang W et al (2020) Mast cells in diabetes and diabetic wound healing. Adv therapy 37:4519–4537 Deng H, Li B, Shen Q, Zhang C, Kuang L, Chen R et al (2023) Mechanisms of diabetic foot ulceration: a review. J Diabetes 15:299–312 Yu X, Liu P, Li Z, Zhang Z (2023) Function and mechanism of mesenchymal stem cells in the healing of diabetic foot wounds. Front Endocrinol (Lausanne) 14:1099310 Soundararajan M, Kannan S (2018) Fibroblasts and mesenchymal stem cells: Two sides of the same coin? J Cell Physiol 233:9099–9109 Rautiainen S, Laaksonen T, Koivuniemi R (2021) Angiogenic effects and crosstalk of adipose-derived mesenchymal stem/stromal cells and their extracellular vesicles with endothelial cells. Int J Mol Sci 22:10890 Luz-Crawford P, Kurte M, Bravo-Alegria J, Contreras R, Nova-Lamperti E, Tejedor G et al (2013) Mesenchymal stem cells generate a CD4 + CD25+Foxp3 + regulatory T cell population during the differentiation process of Th1 and Th17 cells. Stem Cell Res Ther 4:65 Xu J, Ma Y, Zhu X, Zhang J, Cheng Z, Wu W et al (2020) Enhanced autophagy promotes the clearance of Pseudomonas aeruginosa in diabetic rats with wounds. Annals translational Med 8:1362 Glaze DG, Neul JL, Kaufmann WE, Berry-Kravis E, Condon S, Stoms G et al (2019) Double-blind, randomized, placebo-controlled study of trofinetide in pediatric Rett syndrome. Neurology 92:e1912–e1925 Huang YY, Lin CW, Cheng NC, Cazzell SM, Chen HH, Huang KF et al (2021) Effect of a novel macrophage-regulating drug on wound healing in patients with diabetic foot ulcers: a randomized clinical trial. JAMA Netw open 4:e2122607 Additional Declarations (Not answered) Supplementary Files FigS1.tif Figure S1 5Originalfulllengthwesternblots.docx Original full length western blots TableS3DetailedcontentofKEGGenrichmentanalysis.docx Table S3 FigS5.tif Figure S5 4SupplementalMaterials.docx Supplemental Materials TableS2DetailedcontentofGOenrichmentanalysis.docx Table S2 6TheARRIVEguidelines2.0authorchecklist.pdf The ARRIVE guidelines 2.0 author checklist TableS5Detailsofmoleculardocking.docx Table S5 TableS4InfluenceofSNPmutationsonmiRNAhubgenes.docx Table S4 TableS1Primersequences.docx Table S1 FigS3.tif Figure S3 FigS4.tif Figure S4 FigS2.tif Figure S2 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5336506","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":617917223,"identity":"ea62a59f-3d71-465e-ad32-c7e5b7e09025","order_by":0,"name":"Mingwei Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAl0lEQVRIiWNgGAWjYDACCSBOqJCQkydNy4MzFsaGDaRoYXzYVpHIcIBYHfyzu9MkEudJJDA2MD98dIMoS+6c3SaRuE0ij52Bzdg4hxgtBhK5224AtRQzNvCwSZOgZY5EYsMB0rQ0kKJF4kbu9h8JxySMDZuJ9Qv/jNzNhj9q6uTk2ZsfPiZKCwIwk6Z8FIyCUTAKRgE+AAABUS+HtJWapAAAAABJRU5ErkJggg==","orcid":"","institution":"The First Affiliated Hospital of Anhui Medical University","correspondingAuthor":true,"prefix":"","firstName":"Mingwei","middleName":"","lastName":"Chen","suffix":""},{"id":617917224,"identity":"c82f321c-088d-4e18-be40-c7a9e8c21fd0","order_by":1,"name":"Jianran Sun","email":"","orcid":"","institution":"The First Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Jianran","middleName":"","lastName":"Sun","suffix":""},{"id":617917225,"identity":"fe837637-0355-4851-9321-ee0d874e3919","order_by":2,"name":"Yichang Liu","email":"","orcid":"","institution":"The First Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yichang","middleName":"","lastName":"Liu","suffix":""},{"id":617917226,"identity":"af071842-e107-44eb-8979-fa1706b3e35c","order_by":3,"name":"Ying Tang","email":"","orcid":"","institution":"The First Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"Tang","suffix":""},{"id":617917227,"identity":"dc1d9f7f-3e1c-4766-b82d-135fa477b997","order_by":4,"name":"Xiaotong Zhao","email":"","orcid":"","institution":"The First Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xiaotong","middleName":"","lastName":"Zhao","suffix":""},{"id":617917228,"identity":"ca46c674-8ff3-417f-a453-ce9a07439f2e","order_by":5,"name":"Datong Deng","email":"","orcid":"","institution":"The First Affiliated Hospital of Anhui Medical University","correspondingAuthor":false,"prefix":"","firstName":"Datong","middleName":"","lastName":"Deng","suffix":""},{"id":617917229,"identity":"a88fc17c-841c-48ca-bc37-8481c1eed07c","order_by":6,"name":"Shandong Ye","email":"","orcid":"","institution":"University of Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Shandong","middleName":"","lastName":"Ye","suffix":""}],"badges":[],"createdAt":"2024-10-26 08:50:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5336506/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5336506/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106405030,"identity":"afe4def8-f2c1-45f4-955f-748bc066a3d9","added_by":"auto","created_at":"2026-04-08 09:20:40","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":219444,"visible":true,"origin":"","legend":"\u003cp\u003eWorkflow of the study.\u003c/p\u003e","description":"","filename":"Fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/c6d764d35d6a09e09f043691.png"},{"id":106405634,"identity":"641447af-6845-4327-90d7-5138ccbb2348","added_by":"auto","created_at":"2026-04-08 09:27:54","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":746827,"visible":true,"origin":"","legend":"\u003cp\u003eFour hub genes.\u003c/p\u003e\n\u003cp\u003eA-C Selection (A), subcellular localization (B), and co-expression networks (C). D-F CXCL2 and DCN (D), IGF1 (E), CXCL8 (F) from GSEA. G-I CXCL2 and DCN (G), IGF1 (H), CXCL8 (I) from GSVA. J-K Prominent pathway enrichment (J) and influence on disease and function (K) from IPA.\u003c/p\u003e","description":"","filename":"Fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/e9d4c0eca55aa84b158971ca.png"},{"id":106405026,"identity":"c0b2719d-4588-47e3-a034-35b425dbacce","added_by":"auto","created_at":"2026-04-08 09:20:37","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":350389,"visible":true,"origin":"","legend":"\u003cp\u003ePseudotime trajectory inferences for B and T cells.\u003c/p\u003e\n\u003cp\u003eA-E UMAP (A), elbow plot (B), polarization trajectories (C), three disparate conditions between DFUs and controls (D), and hub gene expression during development (E) of B cells. F-J UMAP (F) and elbow plot (G), one trajectory (H), and five distinct stages (I) between DFUs and controls, and expression of key genes during development (J) of T cells.\u003c/p\u003e","description":"","filename":"Fig3.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/56298faca016c9faf9f5b662.png"},{"id":106405885,"identity":"f99d3a98-e157-4682-8318-7dbfb86a4581","added_by":"auto","created_at":"2026-04-08 09:28:54","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":297322,"visible":true,"origin":"","legend":"\u003cp\u003eSubtypes of B and T cells with\u003cstrong\u003e \u003c/strong\u003epseudotime analysis\u003c/p\u003e\n\u003cp\u003eA-F UMAP (A), dot plot (B), annotation of subtypes (C), proportions of each cell type (D), three discrepant phases (E), and subtypes (F) between DFU and control groups of B cells. G-L UMAP (G), dot plot (H), annotation of subtypes (I), proportions of each cell type (J), three diacritical stages (K), and two subtypes (L) between DFU and control groups of T cells.\u003c/p\u003e","description":"","filename":"Fig4.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/d43a9c0ace1a1bbc05d07019.png"},{"id":106723936,"identity":"58894f0c-0249-4fb4-8688-f9dedbdabfd9","added_by":"auto","created_at":"2026-04-12 18:21:48","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":596617,"visible":true,"origin":"","legend":"\u003cp\u003eImmune infiltration analyses and intercellular communication\u003c/p\u003e\n\u003cp\u003eA Stacked bar chart. B Violin plot. C-D Correlation analysis between hub genes and DE cells (C), as well as verified CRGs and DE cells (D). E Association analysis between our CRGs and verified CRGs. F Nomogram. G-H Number (G) and strength (H) of interactions. I Heat map. J Bubble plot between ligands and receptors.\u003c/p\u003e","description":"","filename":"Fig5.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/35ab4e5d34d7598ef7064149.png"},{"id":106405025,"identity":"05a64ad2-003d-4c84-b341-61cbea6c7048","added_by":"auto","created_at":"2026-04-08 09:20:37","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":634596,"visible":true,"origin":"","legend":"\u003cp\u003eDrug sensitivity analysis and molecular docking\u003c/p\u003e\n\u003cp\u003eA Drug sensitivity analysis. B-D Overall (B), sectional (C), and detailed (D) diagram of DCN. E-G Whole (E), localized (F), and detailed (G) graph of CXCL12. H-J Comprehensive (H), partial (I), and exhaustive (J) illustration of CXCL8.\u003c/p\u003e","description":"","filename":"Fig6.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/6b185c67946cf1c769b0caac.png"},{"id":106405700,"identity":"5e123ae4-e6c0-471e-abad-59df43706cc8","added_by":"auto","created_at":"2026-04-08 09:28:17","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":315229,"visible":true,"origin":"","legend":"\u003cp\u003eValidation of\u003cstrong\u003e \u003c/strong\u003eCXCL12 and IGF1 in the DFU model.\u003c/p\u003e\n\u003cp\u003eA Body weight. B Blood glucose. C Wound healing rate. D Morphological illustration of rat paws. E-I Expression levels of CXCL12 and IGF1 mRNA (E-F) and protein (G-I). Data are expressed as \u003cem\u003eM\u003c/em\u003e [\u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003e25\u003c/em\u003e\u003c/sub\u003e, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003e75\u003c/em\u003e\u003c/sub\u003e] from three separate assays.\u003c/p\u003e\n\u003cp\u003e* \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, ** \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01, **** \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001 vs. controls.\u003c/p\u003e","description":"","filename":"Fig7.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/115b4a389fdab1f2832d625b.png"},{"id":106405072,"identity":"c6524947-0166-4cfd-a777-8435aeb795fd","added_by":"auto","created_at":"2026-04-08 09:20:52","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1080500,"visible":true,"origin":"","legend":"\u003cp\u003ePathological changes in ulcerated tissues.\u003c/p\u003e\n\u003cp\u003eA-B HE staining (magnification ´200, bars = 50 μm) in NFUs (A) and DFUs (B). C-D Masson’s trichrome staining (magnification ´200, bars = 50 μm) in NFUs (C) and DFUs (D). E-J Expression of CXCL12 (E-G) and IGF1 (H-J) in NFUs and DFUs, detected by IHC (magnification ´200, bars = 50 μm). Data are expressed as mean ± standard deviation from three separate experiments.\u003c/p\u003e\n\u003cp\u003e* \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05, *** \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001 vs. controls.\u003c/p\u003e","description":"","filename":"Fig8.png","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/34360d5ba471f3e45eaaa54e.png"},{"id":106725441,"identity":"abe6b2c7-4b90-4d08-b279-ea1285c1cd38","added_by":"auto","created_at":"2026-04-12 18:32:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4463067,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/2d2b6fe2-8bba-4539-950a-fe4dff989f4c.pdf"},{"id":106405593,"identity":"23712ff7-693c-4c00-a5f6-b51bcc512f57","added_by":"auto","created_at":"2026-04-08 09:27:43","extension":"tif","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1916488,"visible":true,"origin":"","legend":"Figure S1","description":"","filename":"FigS1.tif","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/4409f7e78e3ce0b3c628aa94.tif"},{"id":106405029,"identity":"f958f4af-3f4c-41e2-a641-920833040509","added_by":"auto","created_at":"2026-04-08 09:20:40","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":250258,"visible":true,"origin":"","legend":"Original full length western blots","description":"","filename":"5Originalfulllengthwesternblots.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/68ab2c6bc7d3580acda7d76d.docx"},{"id":106405071,"identity":"44751b26-5ca8-4dcb-a502-6a71389ad00d","added_by":"auto","created_at":"2026-04-08 09:20:52","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":17213,"visible":true,"origin":"","legend":"Table S3","description":"","filename":"TableS3DetailedcontentofKEGGenrichmentanalysis.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/ac38f09da755be79a159e8fe.docx"},{"id":106405549,"identity":"df0a0c26-ad5e-4ed3-a3d8-d25317822834","added_by":"auto","created_at":"2026-04-08 09:27:21","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":4717052,"visible":true,"origin":"","legend":"Figure S5","description":"","filename":"FigS5.tif","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/af5b16a93d46b1ad5ddcb05d.tif"},{"id":106405028,"identity":"80806c14-64bc-4e6f-98e9-9c971f293c3c","added_by":"auto","created_at":"2026-04-08 09:20:39","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2945501,"visible":true,"origin":"","legend":"Supplemental Materials","description":"","filename":"4SupplementalMaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/428f2a64d2395ecef6dc4079.docx"},{"id":106405681,"identity":"54162376-1670-493e-a881-db2526c1abd9","added_by":"auto","created_at":"2026-04-08 09:28:09","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":84275,"visible":true,"origin":"","legend":"Table S2","description":"","filename":"TableS2DetailedcontentofGOenrichmentanalysis.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/9c790c7aecee3df0b4f1fb6d.docx"},{"id":106405027,"identity":"4857b228-26cc-442c-8364-4b9fe1989ebf","added_by":"auto","created_at":"2026-04-08 09:20:38","extension":"pdf","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":955851,"visible":true,"origin":"","legend":"The ARRIVE guidelines 2.0 author checklist","description":"","filename":"6TheARRIVEguidelines2.0authorchecklist.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/d0d5dce5e561abde9d9749ec.pdf"},{"id":106405068,"identity":"9836bde4-94b0-4f4a-b7a4-ee30e52098a7","added_by":"auto","created_at":"2026-04-08 09:20:50","extension":"docx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":16550,"visible":true,"origin":"","legend":"Table S5","description":"","filename":"TableS5Detailsofmoleculardocking.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/a94128e09aa493257b25766b.docx"},{"id":106405073,"identity":"7b72b555-c66e-4fea-9216-2b9bbab6ed76","added_by":"auto","created_at":"2026-04-08 09:20:52","extension":"docx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":21527,"visible":true,"origin":"","legend":"Table S4","description":"","filename":"TableS4InfluenceofSNPmutationsonmiRNAhubgenes.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/5eca83ef26cdf1febc7ef847.docx"},{"id":106406789,"identity":"0e41d848-a95f-44ab-b35a-50cf4c4cf430","added_by":"auto","created_at":"2026-04-08 09:34:04","extension":"docx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":15740,"visible":true,"origin":"","legend":"Table S1","description":"","filename":"TableS1Primersequences.docx","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/53c226b8d05f8af4ab8704ee.docx"},{"id":106405069,"identity":"1b793f8f-b088-4711-996a-ab7963624da9","added_by":"auto","created_at":"2026-04-08 09:20:51","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":9604484,"visible":true,"origin":"","legend":"Figure S3","description":"","filename":"FigS3.tif","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/374c209ccd754e561a606a66.tif"},{"id":106405619,"identity":"0170426f-8862-457c-b42e-6261b0e878c3","added_by":"auto","created_at":"2026-04-08 09:27:48","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":7203732,"visible":true,"origin":"","legend":"Figure S4","description":"","filename":"FigS4.tif","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/071d7aaf7cb6d60a574fa6bd.tif"},{"id":106405784,"identity":"1d6dd795-8fba-40bf-97f5-52396b030706","added_by":"auto","created_at":"2026-04-08 09:28:28","extension":"tif","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":10114172,"visible":true,"origin":"","legend":"Figure S2","description":"","filename":"FigS2.tif","url":"https://assets-eu.researchsquare.com/files/rs-5336506/v1/17c11a59a8adb8d0e262cd4b.tif"}],"financialInterests":"(Not answered)","formattedTitle":"Deciphering and verifying cuproptosis-associated hub genes in diabetic foot ulcer by combining single-cell and bulk RNA-sequencing","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eDiabetic foot ulcers (DFUs), which affect approximately 18.6\u0026nbsp;million people worldwide, are a concerning condition in patients with diabetes mellitus (DM) and may result in vascular lesions or neuropathic disorders. Indeed, greater than 15% of patients with DM have been estimated to develop DFUs within their lifetimes [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The 5-year mortality rate for DFUs is nearly 30%, which is higher than many cancers. In addition, more than 70% of patients are hospitalized for amputations, which negatively impacts their physical fitness and quality of life [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. With the increasing prevalence of DM, DFUs pose a considerable burden on universal healthcare systems. Disbursements for DFU treatment have been estimated to be 9\u0026ndash;13\u0026nbsp;billion USD/year; thus, DFU may be one of the most exorbitant diabetic complications [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSingle-cell RNA-sequencing (scRNASeq) aids in discerning cellular function and pathogenesis, capturing the transcriptomes of discrete cells in diverse tissues [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. scRNASeq is widely used to delineate the neoplastic microenvironment in the complex ecosystems of heterogeneous cancers, and to reveal molecular mechanisms as well as drug targets [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. The Human Cell Atlas Project aims at describing all cell types in humans through diverse molecular characterization methods, by coupling gene expression profiles with cellular orientation and morphology using scRNASeq.\u003c/p\u003e \u003cp\u003eThe micronutrient copper (Cu), a crucial catalytic cofactor in extensive biological processes, is necessary in microorganisms and mammals. Cu acts as a catalyst in mitochondrial respiration, cellular metabolism, and oxygen transportation. Yet, intracellular Cu levels are maintained at extremely low concentrations. The average adult man incorporates only 100 mg of Cu through homeostatic mechanisms to prevent the aggregation of free intracellular Cu, which is detrimental to cells. Cuproptosis, a ground-breaking mode of cell death relying on Cu, is provoked by excess Cu\u003csup\u003e2+\u003c/sup\u003e, unlike ferroptosis and autophagy. The mechanism underlying cuproptosis has been reported that copper directly binds to lipoylated constituents of the tricarboxylic acid cycle [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe relationship between diabetic complications and plasma Cu is controversial according to prior studies. An investigation indicates lower concentrations of plasma Cu in patients with DFUs than patients with diabetes but without ulcers. A deficiency of Cu exacerbates glycemic control, which triggers temporized healing of DFUs [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Another study treats Cu overload with the chelator trientine, thereby expediting urinary excretion of Cu through formation of a Cu\u003csup\u003e2+\u003c/sup\u003e-trientine complex, and mitigating diabetic cardiovascular disease [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. However, whether cuproptosis participates in DFUs is unclear. Therefore, we combined scRNASeq with bulk RNASeq to identify relevant genes to develop an innovative treatment for DFU.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cp\u003eThe Materials section is described in detail in the Supplementary Materials.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData extraction\u003c/h2\u003e \u003cp\u003eDFU-associated datasets were extracted from Gene Expression Omnibus (GEO) records GSE165816, GSE80178, and GSE134431. Other cuproptosis-associated genes (CRGs) were identified from the literature, such as ATP7A, ATP7B, CDKN2A, DBT, DLAT, DLD, DLST, DLTA, FDX1, GCSH, GLRX5, GLS, GSS, ISCA2, LIAS, LIPT1, MTF1, NDUFA1, NDUFB2, NDUFC1, NFE2L2, PDHA1, PDHB, SLC31A1, and TIMMDC1 [\u003cspan additionalcitationids=\"CR10 CR11\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eIdentification of hub genes\u003c/h3\u003e\n\u003cp\u003eThe Seurat package was used to filter low-quality cells, and cell subpopulations were annotated with the SingleR package to identify dissimilar cell types. The marker genes for each cell type were analyzed to identify differentially expressed genes (DEGs), denoted DEGs1. Subsequently, DEGs from diverse cell types were aggregated according to CRG scores and denoted DEGs2. GSE80178 was analyzed, and the DEGs were denoted DEGs3. Subsequently, an intersection of DEGs1, DEGs2, and DEGs3 was generated to determine candidate genes. Hub genes were identified through analysis of the protein-protein interaction network.\u003c/p\u003e\n\u003ch3\u003eEnrichment analysis\u003c/h3\u003e\n\u003cp\u003eGO and KEGG enrichment analyses were performed for 79 candidate genes. GeneMANIA, gene set enrichment analysis (GSEA), gene set variation analysis (GSVA), and ingenuity pathway analysis (IPA) were conducted for four hub genes.\u003c/p\u003e\n\u003ch3\u003eCreation of the regulatory network\u003c/h3\u003e\n\u003cp\u003eTo further examine the underlying mechanisms of crux genes participating in DFUs, transcription factor (TF)-gene, competing endogenous RNA (ceRNA), and miRNA-single nucleotide polymorphism (SNP)-mRNA networks were built. The Drug-Gene Interaction Database and Comparative Toxicogenomics Database was used to predict drugs and diseases correlating with hub genes.\u003c/p\u003e\n\u003ch3\u003eSingle cell trajectory analysis\u003c/h3\u003e\n\u003cp\u003eInitially, cell types were annotated with UMAP for dimensionality reduction. The monocle2 package was used to analyze the pseudotime trajectory and cell distribution of four hub genes. The Differentialgenetest function was used to explore gene dynamics in cell differentiation.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eImmune infiltration and cell-cell communication\u003c/h2\u003e \u003cp\u003eThe correlation between CRGs and immune infiltration was assessed with the psych and ggcorrplot packages. The ggplot2 package was used for visualization. With the rms package (version 6.5.0), a nomogram was established. The CellChat package (version 1.6.1) was used to analyze signaling interactions.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDrug sensitivity analysis and molecular docking\u003c/h3\u003e\n\u003cp\u003eThe molecular structures were searched in the PubChem database. Potential drug responses were predicted with the pRRophetic package.\u003c/p\u003e\n\u003ch3\u003eCreation of a DFU model\u003c/h3\u003e\n\u003cp\u003eA total of 18 male Wistar rats (age approximately 4 weeks, weight 230\u0026ndash;250 g) were purchased from Jinan Pengyue Experimental Animal Breeding Ltd. (license ID: SCXK2022-0006, Jinan, China). To create DM and control models, rats were injected with streptozotocin and citrate buffer, respectively. Rats with a fasting blood glucose concentration\u0026thinsp;\u0026gt;\u0026thinsp;16.7 mmol/L for 3 days and DM symptoms were considered as type 2 diabetes (T2DM) model. After 1 week, rats with T2DM were anesthetized with 10% chloral hydrate. DFU and non-diabetic foot ulcer (NFU) models were established by removal of a layer of skin on the instep of the foot.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eHematoxylin-eosin and Masson\u0026rsquo;s trichrome staining\u003c/h2\u003e \u003cp\u003eTissues were fixed with 4% paraformaldehyde. Paraffin-embedded sections were cut (4.0 \u0026micro;m thick) for further staining, then stained with hematoxylin for 3\u0026ndash;5 minutes and eosin for 2\u0026ndash;5 minutes. Masson\u0026rsquo;s trichrome staining was performed according to the manufacturer's instructions (Solarbio, Beijing, China).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eImmunohistochemistry\u003c/h2\u003e \u003cp\u003eSlides were incubated with primary antibodies overnight at 4\u0026deg;C, after which slices were incubated with secondary antibody.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eRT-qPCR and WB\u003c/h2\u003e \u003cp\u003eRT-qPCR and WB were performed (primer sequences in Table S1). Details are provided in the Supplementary Materials.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eR software and SPSS 25.0 were used for statistical analysis. The skew distributional data were expressed as \u003cem\u003eM\u003c/em\u003e [\u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003e25\u003c/em\u003e\u003c/sub\u003e, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003e75\u003c/em\u003e\u003c/sub\u003e] from three separate experiments. Differences between two groups were assessed with Wilcoxon rank sum test when the distribution of two samples was skewed. Mean optical density (MOD)\u0026thinsp;=\u0026thinsp;sum integrated optical density/sum area. A two-tailed \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicated a statistically significant difference.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eValidation of DEGs\u003c/h2\u003e \u003cp\u003eThe research flowchart was illustrated in Fig.\u0026nbsp;1. After quality control (Fig. S1A-B), the number of cells was 57,302. After raw data were processed with normalization, the top 2,000 HVGs were determined (Fig. S2A). Consequently, 30 principal components were identified (Fig. S2B). A total of 24 cell clusters were classified (Fig. S2C-D). Through difference analysis, a total of 1,273 DEGs1 were discovered, in which cell types were labeled with marker genes (Fig. S2E). Furthermore, CRG scores between DFUs and controls were distinct in the remaining ten cell types (Fig. S2F). In addition, the cells were allocated to high and low CRG score groups according to the median CRG score (Fig. S2G), and 685 DEGs2 were identified (Fig. S1C-L). In analysis of the training set, 2,652 DEGs3 were identified (Fig. S3A-B). By intersecting DEGs1, DEGs2, and DEGs3, 79 candidate genes were compiled (Fig. S3C). The top eight GO terms are shown in Fig S3D. In total, 256 GO terms were identified (Table S2). The top seven KEGG pathways are shown in Fig. S3E, and the total terms are indicated in Table S3. Finally, a protein-protein interaction network was constructed (Fig. S3F).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eDiscovery of hub genes\u003c/h2\u003e \u003cp\u003eWith seven algorithms, four hub genes were detected (Fig.\u0026nbsp;2A). Subcellular localization suggested that the proteins translated by hub genes were located primarily in the cytoplasm (Fig.\u0026nbsp;2B). GeneMANIA suggested that hub expression genes and co-expression genes (CXCL8, PF4, and CCL2) had mutual functions (Fig.\u0026nbsp;2C). GSEA revealed that CXCL2, DCN, and IGF1 were enriched in spliceosome, RNA degradation, and cell cycle in the high expression group, whereas CXCL8 participated in the above-mentioned pathways in the low expression group (Fig.\u0026nbsp;2D-F). GSVA indicated that CXCL2, DCN, and IGF1 were associated with graft versus host disease and asthma in the high expression group, whereas CXCL8 had roles in those pathways in the low expression group (Fig.\u0026nbsp;2G-I). IPA demonstrated that hub genes were markedly enriched in five pathways, three of which were activated and two of which were inhibited (Fig.\u0026nbsp;2J). Regarding disease and function, the hub genes were associated with functions including extracranial solid tumor, tumor incidence, and apoptosis (Fig.\u0026nbsp;2K).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eNetwork and expression levels of hub genes\u003c/h2\u003e \u003cp\u003eInitially, the TF-gene network contained 161 nodes and 209 edges (Fig. S4A). Moreover, the ceRNA network consisted of 3 mRNAs, 11 miRNAs, and 24 lncRNAs (Fig. S4B-C). Furthermore, the miRNA-SNP-mRNA network included 22 SNP locations (Fig. S4D; details in Table S4). A disease-hub gene-drug network comprised 105 nodes and 278 edges (Fig. S4E). The training and validation sets indicated that CXCL12, DCN, and IGF1 were downregulated, whereas CXCL8 was upregulated in the DFU group compared with the control group (Fig. S4F-G).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003ePseudotime analysis\u003c/h2\u003e \u003cp\u003eIn B cells, two distinct subpopulations existed (Fig.\u0026nbsp;3A). Principal component analysis revealed that the value of characteristic number was 30, after which a plateau was reached (Fig.\u0026nbsp;3B). In addition, B cell differentiation patterns showed similar trajectories between the DFU and control groups, although slight deviations were observed at several crucial points, such as bifurcation and branch ends. In the middle and late stages, B cells were far more abundant in DFU samples than control samples (Fig.\u0026nbsp;3C). Three stages of differentiation of B cells were observed (Fig.\u0026nbsp;3D). The expression of key genes during development suggested that the CXCL8 and DCN genes showed a rising trend, then reverted to stable expression levels, whereas the expression of CXCL12 and IGF1 remained generally stable (Fig.\u0026nbsp;3E).\u003c/p\u003e \u003cp\u003eAmong T cells, five subpopulations were observed (Fig.\u0026nbsp;3F). The characteristic number value was nearly 30 (Fig.\u0026nbsp;3G). The trajectory of T cells was like an arc during cellular differentiation. Initially, T cells were assembled on the left side of this plot. As the pseudotime progressed, T cells gradually gathered towards the right side of the picture. In the late period, the number of T cells in DFU samples was greater than the controls samples (Fig.\u0026nbsp;3H). Similarly, T cells evolved into five stages in both groups (Fig.\u0026nbsp;3I). CXCL12 and DCN expression declined initially, then became constant, whereas CXCL8 and IGF1 expression did not change (Fig.\u0026nbsp;3J).\u003c/p\u003e \u003cp\u003eSubtype analysis revealed that the numbers of plasma cells and naive B cells were higher, and the numbers of memory B cells were lower in DFUs than controls (Fig.\u0026nbsp;4A-F). In the late stage of differentiation, the number of CD4\u003csup\u003e+\u003c/sup\u003e T cells was significantly greater in DFUs than controls (Fig.\u0026nbsp;4G-L). Moreover, DCN was highly expressed in fibroblasts, chondrocytes, and tissue stem cells; CXCL12 was highly expressed in endothelial cells and MSCs; CXCL8 was abundant in monocytes; and IGF1 was abundant in fibroblasts (Fig. S5).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eImmune infiltration analyses and cellular communication\u003c/h2\u003e \u003cp\u003eThe CIBERSORT approach indicated higher numbers of naive B cells, monocytes, activated myeloid dendritic cells, and resting NK cells, but lower numbers of activated mast cells, activated NK cells, and M1 macrophages in DFUs than controls (Fig.\u0026nbsp;5A-B). Furthermore, Spearman correlation analysis indicated that hub genes and differentially expressed (DE) cells were correlated (Fig.\u0026nbsp;5C), verified CRGs and DE cells were associated (Fig.\u0026nbsp;5D), in addition, the relationship between our CRGs and verified CRGs (Fig.\u0026nbsp;5E). In the nomogram (Fig.\u0026nbsp;5F), four hub genes were included for diagnosis of DFUs; the total point values were nearly 120; and the DFU risk was approximately 0.86.\u003c/p\u003e \u003cp\u003eOur findings suggested that multiple cells communicated with each other continuously (Fig.\u0026nbsp;5G-H). MSCs frequently interacted with fibroblasts, keratinocytes, endothelial cells, and T cells (Fig.\u0026nbsp;5I). The ligands and receptors were MIF and CD74\u0026thinsp;+\u0026thinsp;CXCR4, and MIF and CD74\u0026thinsp;+\u0026thinsp;CD44, respectively. The corresponding cells were macrophages and T cells (Fig.\u0026nbsp;5J).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eDrug sensitivity analysis and molecular docking\u003c/h2\u003e \u003cp\u003eRapamycin was predicted to be the most effective drug for patients with DFUs (Fig.\u0026nbsp;6A). The binding energy of sirolimus to DCN was \u0026minus;\u0026thinsp;7.5 kcal/mol (Fig.\u0026nbsp;6B-D), that of chlorambucil to CXCL12 was \u0026minus;\u0026thinsp;5.2 kcal/mol (Fig.\u0026nbsp;6E-G), and that of foscarnet to CXCL8 was \u0026minus;\u0026thinsp;4.0 kcal/mol (Fig.\u0026nbsp;6H-J; details in Table S5).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eExperimental validation\u003c/h2\u003e \u003cp\u003eCompared with the control group, DFU rats had lower weight, higher blood glucose levels, and poorer wound healing (Fig.\u0026nbsp;7A-D), RT-qPCR and WB demonstrated lower CXCL12 and IGF1 expression (Fig.\u0026nbsp;7E-I). Because the RT-qPCR findings for DCN and CXCL8 in the preliminary animal experiment were not consistent with the bioinformatics findings, subsequent experiments were not performed.\u003c/p\u003e \u003cp\u003eHE staining indicated that rats with DFUs, compared with controls, had greater aggregation of neutrophils and lymphocytes along with nascent capillaries (Fig.\u0026nbsp;8A-B). Masson\u0026rsquo;s trichrome staining revealed a few collagen fibers in DFU rats (Fig.\u0026nbsp;8C-D). IHC showed lower protein expression of CXCL12 and IGF1 in DFUs than controls (Fig.\u0026nbsp;8E-J).\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eOur study demonstrated diminished DCN in DFUs, thus corroborating that DCN facilitates angiogenesis [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. DCN, also known as decorin, a molecule in the small leucine-rich proteoglycan family, has known functions in fibrosis and malignancy. Initially, decorin is discovered as a collagen-bound member that constrains fibrillogenesis as a configurational constituent of the substrate. Subsequent research shows that decorin is oncosuppressive during cancer formation, progression, and metastasis [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Moreover, we observed that DFUs had lower levels of IGF1 than controls, thus supporting evidence that deficiency of IGF1 delays wound healing [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. IGF1 is one of the two isoforms of IGF in mammals. This chemokine in endothelial cells facilitates keratinocyte colonization and re-epithelialization. In addition, CXCL8, also known as IL-8, is a pro-inflammatory cytokine that we observed to be upregulated in DFUs. This cytokine attracts neutrophils to injury sites, thus facilitating healing, such as by secreting antimicrobial peptides and creating neutrophil extracellular traps that kill or immobilize bacteria [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Furthermore, we found that CXCL12, also designated as SDF-1, was down-regulated, in line with findings suggesting that DFU-originating fibroblasts excrete a lower concentration of CXCL12 than natural fibroblasts [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOur enrichment analysis revealed that CRGs activated cytokine storm signaling pathways. A critical function of growth factors (GFs [e.g., PDGF, FGF, and VEGF]), which is a classification of cytokines, in wound healing is their stimulation of angiogenesis, thereby triggering perpetual vascularization to revitalize ischemic tissues. Early in wound closure, PDGF attracts fibroblasts to lesion sites, and acts as a mitogen stimulating the transformation of fibroblasts into myofibroblasts, thus resulting in shrinkage of the vulnus. PDGF is the first topical agent gaining regulatory approval to boost wound healing [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. FGF has stronger effects than PDGF and VEGF in expediting angiogenesis and generating granulation tissue that pads gap of wound; however, FGFs have several drawbacks, including degradation due to the proteolytic conditions in wounds [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. VEGF, which is secreted by endothelial cells, pericytes, is regulated by the HIF-1α pathway. In addition, VEGF stimulates collagenases to degrade the basement membrane and simultaneously potentiates epithelialization [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOur TF network indicated that CXCL12 interacted with the TF ID1, in agreement with a study that signals for fibrotic transformation of the hepatic vascular niche are derived from CXCL12 and its receptors (CXCR7 and CXCR4) in liver sinusoidal endothelial cells (LSECs). After acute impairment, CXCR7 and CXCR4 activate ID1 to prompt LSECs to secrete paracrine growth regulators to restore liver. Moreover, our ceRNA network results demonstrated that the lncRNA TUG1 acted as a ceRNA for miR-1-3p, whose target gene was IGF1, in agreement with prior literature [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. TUG1 acts as a ceRNA, as substantiated by a host of studies in tumors, with roles including regulating the miR-142/ZEB2 axis in bladder cancer [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], miR-221/PTEN axis in lung cancer [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], and miR-219/FMNL2 axis in oral squamous cell carcinoma [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOur pseudotime analysis indicated far more mid-late B cells in DFU samples than controls, in agreement with B cells expediting wound healing [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Because the skin is a protective barrier facing multiple injuries daily, an urgent need exists to repair wounds in which B cells reside. For example, mice lacking the B cell marker, CD19, have delayed wound healing, whereas upregulation of CD19 is beneficial for wound closure [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. In the late period, the number of T cells in DFU samples was greater than control samples in contrast to a study in which patients with chronic DFUs retain a similar number of naive and effector T cells compared to controls [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Skin-resident T cells, which are comprised of αβ and γδ T cells, have a role in would repair and are involved in cutaneous infections preceding adaptive immunity [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Epidermal T cells interact with keratinocytes to generate an IL-15-IGF anulus, thereby potentiating IGF-1 expression and re-epithelialization [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. In addition, our T cell subtype analysis indicated that the numbers of CD4\u003csup\u003e+\u003c/sup\u003e T cells were significantly greater in DFUs than controls in the late stage of differentiation. However, the effect of CD4\u003csup\u003e+\u003c/sup\u003e T cells on wound healing is disputed: one article has indicated that depletion of CD4\u003csup\u003e+\u003c/sup\u003e T cells is detrimental for healing, whereas another study has reported that CD4\u003csup\u003e+\u003c/sup\u003e T cell depletion has no influence on wound closure [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOur immune infiltration analysis indicated fewer activated mast cells, activated NK cells, and M1 macrophages in DFUs than controls, thus supporting that DFUs involve immune cells dysregulation due to chronic inflammation [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. In the proliferative stage, macrophages transform from a pro-inflammatory to a pro-healing phenotype. However, DFUs are deficient in the M1-to-M2 switch, thus resulting in a paucity of GFs, which are indispensable for remodeling. Furthermore, mast cells aid in scar formation, by modulating the conversion from scarless to fibrotic closure [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. In addition, activated NK cells secrete perforins, granzymes, and interferon-γ, thereby intensifying M1 macrophage polarization [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Notably, our cellular communication experiments indicated that MSCs frequently interacted with fibroblasts, keratinocytes, endothelial cells, and T cells. The MSCs arise from the bone marrow, umbilical cord, adipose tissue, and placenta, which secrete GFs that support wound healing [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Intriguingly, the MSCs and fibroblasts appear to have overlapped immunophenotypes and differentiation capabilities, MSCs might be immature fibroblasts [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. In addition, MSCs can be converted into keratinocytes and endothelial cells, particularly in ulcerated wounds, thus leading to the secretion of cytokeratin 19 by keratinocytes and the induction of angiogenesis [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Moreover, MSCs regulate immunity by decreasing the ratios of pro-inflammatory T cells, such as Th17 and Th1; increasing the numbers of immunosuppressive T cells, such as Treg cells; and stimulating the formation of M2 macrophages, thereby alleviating inflammation and accelerating wound healing [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOur drug sensitivity analysis and molecular docking demonstrated that foscarnet, chlorambucil, and sirolimus were effective drugs for DFU treatment. Notably, rapamycin, also termed as sirolimus, is an autophagy activator and immunosuppressant. It is reported that Pseudomonas aeruginosa curb autophagy in wounds. However, rapamycin activates autophagy, thereby increasing microorganism removal [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Herein, a conundrum may exist. On the one hand, the molecular structure of IGF1 does not match to a suitable drug, although trofinetide, an analog of IGF1, has class I evidence for Rett syndrome treatment [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. On the other hand, no literature basis exists for DFU therapy with the other two drugs. Molecular docking largely considers binding energies, based on the reference drug set. If any drugs are used for molecular docking, the binding energy will exceed 5 kcal/mol. Second, molecular docking forecasts a theoretical scenario, and clinical experiments are required to verify the true efficacy. Finally, compared with traditional remedies for DFUs, including debridement and wound dressing, MSCs are a more practical treatment approach for the supervision of persistent wounds because MSCs restrain pro-inflammatory cytokines and stimulate the secretion of GFs, such as FGF and VEGF [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Furthermore, ON101 is the first ratified macrophage-modulation drug for DFUs treatment, expediting the M1-to-M2 transition [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAlthough examining CRGs in DFUs is highly challenging, this research has several limitations. First, because of the finite sample size of the DFU transcriptome datasets, ROC curve analyses and machine learning were not performed. In addition, the roles of CRGs in DFUs remain recondite and will be accurately elucidated in our future mechanistic research.\u003c/p\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eMultiple genes were screened through bioinformatics and animal experiments. CRGs in DFUs were identified, notably DCN, IGF1, CXCL12, and CXCL8. Findings in a rat model further validated two downregulated genes: CXCL12 and IGF1. The nomogram proves diagnostic value of CRGs. Additionally, rapamycin is a predictive drug for DFU therapy.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\n \u003cp\u003e\u003cstrong\u003eDATA AVAILABILITY\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eThe data used and/or analyzed during this research are accessible from the corresponding author on reasonable request.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eNone.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eAUTHOR CONTRIBUTIONS\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eJS conceived this study. JS, LY, YT, XZ, and DD performed the experiments and analyzed the data. JS wrote the draft. SY revised the manuscript. MC participated in supervision. All authors approved the manuscript for publication.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eFUNDING\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eThis research was funded by the Natural Science Foundation of Anhui Province (2108085MH269) and the Natural Science Research Project of Colleges and Universities (KJ2021A0274).\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eCOMPETING INTERESTS\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eETHICS APPROVAL\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eAnimal experiments complied with the ARRIVE guidelines 2.0, and the research protocol was approved by the Institutional Animal Ethics Committee (ethical number: 2023-NA-44).\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eCONSENT FOR PUBLICATION\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eAll authors have authorized publication.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMariadoss AVA, Sivakumar AS, Lee CH, Kim SJ (2022) Diabetes mellitus and diabetic foot ulcer: etiology, biochemical and molecular based treatment strategies via gene and nanotherapy. Biomed Pharmacother 151:113134\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArmstrong DG, Tan TW, Boulton AJM, Bus SA (2023) Diabetic foot ulcers: a review. JAMA 330:62\u0026ndash;75\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRice JB, Desai U, Cummings AK, Birnbaum HG, Skornicki M, Parsons NB (2014) Burden of diabetic foot ulcers for medicare and private insurers. Diabetes Care 37:651\u0026ndash;658\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTheocharidis G, Thomas BE, Sarkar D, Mumme HL, Pilcher WJR, Dwivedi B et al (2022) Single cell transcriptomic landscape of diabetic foot ulcers. Nat Commun 13:181\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStubbington MJT, Rozenblatt-Rosen O, Regev A, Teichmann SA (2017) Single-cell transcriptomics to explore the immune system in health and disease. Science 358:58\u0026ndash;63\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen L, Min J, Wang F (2022) Copper homeostasis and cuproptosis in health and disease. Signal Transduct Target Ther 7:378\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYadav C, Srikantiah RM, Manjrekar P, Shenoy MT, Chaudhury D (2020) Assessment of mineral pathophysiology in patients with diabetic foot ulcer. Biol Trace Elem Res 195:366\u0026ndash;372\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCooper GJ, Chan YK, Dissanayake AM, Leahy FE, Keogh GF, Frampton CM et al (2005) Demonstration of a hyperglycemia-driven pathogenic abnormality of copper homeostasis in diabetes and its reversibility by selective chelation: quantitative comparisons between the biology of copper and eight other nutritionally essential elements in normal and diabetic individuals. Diabetes 54:1468\u0026ndash;1476\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Y, Chen X, Geng S, Li Q, Li Y, Yuan H et al (2023) Identification of the cuproptosis-related hub genes and therapeutic agents for sarcopenia. Front Genet 14:1136763\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Y, Zou X, Ma M, Liu Y, Wang R, Dai Z et al (2023) Expression profiles of cuproptosis-related genes determine distinct subtypes of pancreatic ductal adenocarcinoma. Curr Oncol 30:1648\u0026ndash;1662\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Y, Tang L, Huang W, Abisola FH, Zhang Y, Zhang G et al (2023) Identification of a prognostic cuproptosis-related signature in hepatocellular carcinoma. Biol Direct 18:4\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo Z, Liu J, Liang G, Liang H, Zhong M, Tomlinson S et al (2024) Identification and validation of cuproptosis-related genes in acetaminophen-induced liver injury using bioinformatics analysis and machine learning. Front Immunol 15:1371446\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen F, Lai J, Zhu Y, He M, Hou H, Wang J et al (2020) Cardioprotective effect of decorin in type 2 diabetes. Front Endocrinol (Lausanne) 11:479258\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin P, Chen H, Wang Y, Huang L, Huang K, Xiao G et al (2023) Cancer-associated fibroblasts undergoing neoadjuvant chemotherapy suppress rectal cancer revealed by single-cell and spatial transcriptomics. Cell Rep Med 4:101231\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu F, Gong D, Yan D, Wang H, Witman N, Lu Y et al (2023) Enhanced adipose-derived stem cells with IGF-1-modified mRNA promote wound healing following corneal injury. Mol Ther 31:2454\u0026ndash;2471\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMohsin F, Javaid S, Tariq M, Mustafa M (2024) Molecular immunological mechanisms of impaired wound healing in diabetic foot ulcers (DFU), current therapeutic strategies and future directions. Int Immunopharmacol 139:112713\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSorooshian P, Metcalfe AD, Lali FV (2021) In vitro modelling of disease-induced changes in the diabetic wound fibroblast. J Wound Care 30:300\u0026ndash;303\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu X, Han YY, Yang CY, Ji HM, Lan YJ, Bi YQ et al (2021) Activated AMPK by metformin protects against fibroblast proliferation during pulmonary fibrosis by suppressing FOXM1. Pharmacol Res 173:105844\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y, Liu Y, Deng J, Li W, Nie X (2021) Fibroblast growth factor in diabetic foot ulcer: Progress and Therapeutic Prospects. Front Endocrinol (Lausanne) 12:744868\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerstan A, Dieter K, Niebergall-Roth E, Klingele S, Junger M, Hasslacher C et al (2022) Translational development of ABCB5(+) dermal mesenchymal stem cells for therapeutic induction of angiogenesis in non-healing diabetic foot ulcers. Stem Cell Res Ther 13:455\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang K, Lv D, Miao L, Mao Y, Yu X (2022) LncRNA TUG1 functions as a ceRNA for miR-1-3p to promote cell proliferation in hepatic carcinogenesis. J Clin Lab Anal 36:e24415\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Q, Liu H, Cheng H, Li Y, Li X, Zhu C (2017) Downregulation of long noncoding RNA TUG1 inhibits proliferation and induces apoptosis through the TUG1/miR-142/ZEB2 axis in bladder cancer cells. Onco Targets Ther 10:2461\u0026ndash;2471\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo S, Zhang L, Zhang Y, Wu Z, He D, Li X et al (2019) Long non-coding RNA TUG1 enhances chemosensitivity in non-small cell lung cancer by impairing microRNA-221-dependent PTEN inhibition. Aging 11:7553\u0026ndash;7569\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYan G, Wang X, Yang M, Lu L, Zhou Q (2017) Long non-coding RNA TUG1 promotes progression of oral squamous cell carcinoma through upregulating FMNL2 by sponging miR-219. Am J Cancer Res 7:1899\u0026ndash;1912\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore EM, Maestas DR Jr., Cherry CC, Garcia JA, Comeau HY, Davenport Huyer L et al (2021) Biomaterials direct functional B cell response in a material-specific manner. Sci Adv 7:eabj5830\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIwata Y, Yoshizaki A, Komura K, Shimizu K, Ogawa F, Hara T et al (2009) CD19, a response regulator of B lymphocytes, regulates wound healing through hyaluronan-induced TLR4 signaling. Am J Pathol 175:649\u0026ndash;660\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoura J, Rodrigues J, Goncalves M, Amaral C, Lima M, Carvalho E (2017) Impaired T-cell differentiation in diabetic foot ulceration. Cell Mol Immunol 14:758\u0026ndash;769\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReider IE, Lin E, Krouse TE, Parekh NJ, Nelson AM, Norbury CC (2024) gammadelta T Cells mediate a requisite portion of a wound healing response triggered by cutaneous poxvirus infection. Viruses 16:425\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y, Bai Y, Li Y, Liang G, Jiang Y, Liu Z et al (2017) IL-15 Enhances Activation and IGF-1 production of dendritic epidermal t cells to promote wound healing in diabetic mice. Front Immunol 8:1557\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavis PA, Corless DJ, Aspinall R, Wastell C (2001) Effect of CD4(+) and CD8(+) cell depletion on wound healing. Br J Surg 88:298\u0026ndash;304\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQi X, Cai E, Xiang Y, Zhang C, Ge X, Wang J et al (2023) An immunomodulatory hydrogel by hyperthermia-assisted self-cascade glucose depletion and ROS scavenging for diabetic foot ulcer wound therapeutics. Adv Mater 35:e2306632\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong J, Chen L, Zhang Y, Jayaswal N, Mezghani I, Zhang W et al (2020) Mast cells in diabetes and diabetic wound healing. Adv therapy 37:4519\u0026ndash;4537\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng H, Li B, Shen Q, Zhang C, Kuang L, Chen R et al (2023) Mechanisms of diabetic foot ulceration: a review. J Diabetes 15:299\u0026ndash;312\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu X, Liu P, Li Z, Zhang Z (2023) Function and mechanism of mesenchymal stem cells in the healing of diabetic foot wounds. Front Endocrinol (Lausanne) 14:1099310\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoundararajan M, Kannan S (2018) Fibroblasts and mesenchymal stem cells: Two sides of the same coin? J Cell Physiol 233:9099\u0026ndash;9109\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRautiainen S, Laaksonen T, Koivuniemi R (2021) Angiogenic effects and crosstalk of adipose-derived mesenchymal stem/stromal cells and their extracellular vesicles with endothelial cells. Int J Mol Sci 22:10890\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuz-Crawford P, Kurte M, Bravo-Alegria J, Contreras R, Nova-Lamperti E, Tejedor G et al (2013) Mesenchymal stem cells generate a CD4\u0026thinsp;+\u0026thinsp;CD25+Foxp3\u0026thinsp;+\u0026thinsp;regulatory T cell population during the differentiation process of Th1 and Th17 cells. Stem Cell Res Ther 4:65\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu J, Ma Y, Zhu X, Zhang J, Cheng Z, Wu W et al (2020) Enhanced autophagy promotes the clearance of Pseudomonas aeruginosa in diabetic rats with wounds. Annals translational Med 8:1362\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlaze DG, Neul JL, Kaufmann WE, Berry-Kravis E, Condon S, Stoms G et al (2019) Double-blind, randomized, placebo-controlled study of trofinetide in pediatric Rett syndrome. Neurology 92:e1912\u0026ndash;e1925\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang YY, Lin CW, Cheng NC, Cazzell SM, Chen HH, Huang KF et al (2021) Effect of a novel macrophage-regulating drug on wound healing in patients with diabetic foot ulcers: a randomized clinical trial. JAMA Netw open 4:e2122607\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5336506/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5336506/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePrior studies have indicated elevated plasma copper levels in patients with diabetes compared with healthy controls. However, cuproptosis remains unexplored in diabetic foot ulcers (DFUs). DFU-associated datasets (GSE165816, GSE80178, and GSE134431) were obtained from the Gene Expression Omnibus (GEO) database. Cuproptosis-associated genes (CRGs) were identified by combining single-cell and bulk RNA-sequencing data. Differentiation, enrichment, network, pseudotime, immune infiltration, cellular communication, drug sensitivity, and molecular docking analyses were performed. The CRGs were verified in a Wistar DFUs rat model. Four hub genes were obtained (DCN, IGF1, CXCL12, and CXCL8). Enrichment analysis indicated that these genes were involved primarily in cytokine storms. Moreover, network analysis revealed the relationships among competing endogenous RNAs, transcription factors, single nucleotide polymorphisms, and hub genes. In addition, pseudotime analysis revealed greater numbers of plasma cells, naive B cells, and CD4\u003csup\u003e+\u003c/sup\u003e T cells in DFUs than controls. Furthermore, immune infiltration analysis indicated immune cells dysregulation in DFUs, characterized by lower numbers of activated mast cells, activated NK cells, and M1 macrophages than those in controls. In addition, cellular communication analysis revealed that mesenchymal stem cells frequently interacted with fibroblasts, keratinocytes, endothelial cells, and T cells. The nomogram indicated that four hub genes were included for diagnosis of DFUs and the DFU risk was approximately 0.86. Finally, drug sensitivity analysis and molecular docking demonstrated that sirolimus was an effective drug for DFU treatment. Together, our findings link IGF1 and CXCL12 to cuproptosis, thus providing novel insights for DFU diagnosis and treatment.\u003c/p\u003e","manuscriptTitle":"Deciphering and verifying cuproptosis-associated hub genes in diabetic foot ulcer by combining single-cell and bulk RNA-sequencing","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-07 16:50:33","doi":"10.21203/rs.3.rs-5336506/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"531dc147-5f34-4055-bc45-085299363052","owner":[],"postedDate":"April 7th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":65756192,"name":"Health sciences/Endocrinology/Endocrine system and metabolic diseases/Diabetes/Diabetes complications"},{"id":65756193,"name":"Health sciences/Diseases/Endocrine system and metabolic diseases/Thyroid diseases"}],"tags":[],"updatedAt":"2026-04-07T16:50:33+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-07 16:50:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5336506","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5336506","identity":"rs-5336506","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}