Keloid disease (KD) is a benign skin fibroplasia caused by abnormal wound healing after skin injury (1) leading to hyperplasic invasive growth, and has a high recurrence rate (2). The occurrence of KD involves trauma, chronic inflammation, and fibrosis tumor inheritance (1, 3, 4). Keloids can grow on all parts of the body (5), and are accompanied by unbearable itching and pain which seriously affects quality of life. Keloids, especially on the face, can also have a serious impact on mental health (6, 7). Although there are many studies on KD the pathogenesis is still not completely clear (8); improved understanding of the pathogenesis will likely lead to new treatments. Several studies have shown that inflammation is involved in regulating KD collagen synthesis, and the intensity of inflammation is positively correlated with the final scar size (9, 10). Therefore, study of the inflammation-related molecular pathogenesis of KD may lead to new KD prevention and treatment strategies.

It is well known that scars are the result of both inflammation and fibrosis after injury repair (1, 8, 11). In the early stage of repair, inflammatory cells play a pro-inflammatory role through cytokines. It usually enters the repair and healing stage after 72 hours and finally completes the remodeling of collagen (12). Pro-inflammatory factors such as IL-1α, IL-1β, IL-6 and TNF-α are up-regulated in KD tissue (11). It has been speculated that chronic inflammation persists in KD causing excessive deposition of extracellular matrix which is an important cause of keloid formation (13, 14). This indicates that KD is an inflammatory disease of the skin (12). In addition, Shi et al. demonstrated that IL-10 can negatively regulate collagen synthesis, thereby reducing scar formation (13, 15). Nishiguchi et al. reported that the chemokine CXCL12 can promote scar formation in mice (12, 16). A large number of studies have shown that KD is correlated with chronic inflammation (11, 12, 17). However, few studies have explored of inflammation-related genes IRGs in KD and the specific mechanism of action in KD pathogenesis. Therefore, we identified and analyzed differentially-expressed IRGs in KD in order to discover new genes that might be important in KD pathogenesis, both as biomarkers for early diagnosis and as novel drug targets.

KD is a benign skin tumor caused by abnormal hyperplasia of connective tissue in the skin, that occurs during prolonged abnormal wound healing. The mechanisms by which keloids form are currently unclear. Some scholars believe that the abnormal response of fibroblasts to inflammation is causes keloid formation. We propose that the inflammatory response is a significant factor in keloid pathogenesis (13–15). However, most of the current research on keloids focuses on fibroblasts and collagen with little emphasis on the importance of inflammatory genes. Therefore, finding key inflammatory genes associated with KD may help to identify new diagnostic biomarkers and drug targets.

In this study we explored the differentially expressed IRGs in two KD datasets, conducted multiple functional enrichment analyses, constructed a PPI network, and explored immune infiltration in the KD microenvironment. Finally, three keloid biomarkers were identified: LPAR1, FOXF1 and TRIM32. In the RT-qPCR data collected from our clinical samples LPAR1 and TRIM32 were differentially expressed in KD samples (P<0.05) whereas FOXF1 was not (P>0.05).

The protein encoded by TRIM32 is a member of the tripartite motif-containing family. This protein is located in the cytoplasm and nucleus and has E3 ubiquitin ligase activity (23). TRIM32 can ubiquitinate PIAS4/PIASY and promote its degradation in UVB and TNF-α stimulated keratinocytes. In our study, the GSEA results indicated that TRIM32 was mainly enriched in IFN-α reactions, cell apoptosis, and hypoxia. Chaudhuri et al. reported that knocking down TRIM32 inhibited glucose-induced podocyte apoptosis, oxidative stress, and inflammatory response (24). Liu et al. reported that the gene manipulation of Trim32 can regulate Th17 vs. Th2 immunity in response to TLR activation, suggesting that atopic dermatitis is a result of TRIM32 protein deficiency in the skin. It was speculated that TRIM32 plays a crucial role in inflammatory diseases and congenital immunodeficiency diseases (25). Our analysis found that TRIM32 was upregulated in publicly available KD microarray data, and RT-qPCR from our clinical samples confirmed this (P < 0.05). We speculate that TRIM32 is closely involved in the formation of keloids. Further research on the inflammatory regulation of scarring by TRIM32 may establish TRIM32 as a potential treatment target for keloids.

The protein encoded by LPAR1 is an integral membrane protein in the family of lysophosphatidic acid receptors also known as EDG receptors (26, 27). LPAR1 is involved in the reorganization, migration, differentiation, and proliferation of actin cytoskeleton, as well as its response to tissue damage and infection (28–30). LPAR1 promotes the formation of lamellar pseudopodia at the anterior edge of migrating cells by activating RAC1. This activation plays a role in chemotaxis and cell migration, which are important in injury responses (31–33). Wu et al. reported that LPAR1 can mediate various biological functions of tumors (34) and participate in the activation, proliferation differentiation, and migration of immune cells (32). Our correlation analysis between the ssGSEA scores of the differentially abundant immune cells and biomarkers in this study showed that LPAR1 was positively correlated with activated CD4 T cells and effector memory CD4 T cells. LPAR1 expression was reported to be positively correlated with the expression of chemokines and chemokine receptors, suggesting that LPAR1 may regulate immune cell migration (35). The E2F family of transcription factors regulate cell function via gene transcription. E2F was reported as a novel fibrotic gene regulating pulmonary fibrosis (36). The enrichment of single gene GSEA in this study indicated that LPAR1 is significantly enriched in the “E2F target” pathway. LPAR1 is most highly expressed in endothelial cells and fibroblasts in skin and soft tissues. Our analysis showed that LPAR1 was downregulated in KD samples, and this was confirmed by our RT-qPCR data from clinical samples. We therefore speculate that LPAR1 plays an important inflammatory and immune regulatory role in the formation of keloids.

FOXF1 belongs to the forkhead transcription factor family and is characterized by a unique forkhead domain (37). In an immune cell analysis of infantile angiomatosis, FOXF1 was found to be positively correlated with the degree of monocyte infiltration (38). Recent studies have shown that overexpression of FOXF1 can inhibit the production of α-SMA, fibronectin, and type IV collagen, thereby alleviating TGF-β1-induced fibrosis (39). In addition, overexpression of FOXF1 can promote the proliferation of BEAS-2B cells, inhibit apoptosis, and inhibit inflammation in response to TGF-β1. Fenghua et al. reported that increasing FOXF1 expression in endothelial cells could alleviate pulmonary fibrosis (40). FOXF1 is highly expressed in both endothelial cells and fibroblasts, suggesting that FOXF1 is involved in chronic inflammation following tissue injury and inhibits collagen deposition and fiber proliferation in keloid formation. In this study, GSEA results showed that FOXF1 was mainly enriched in E2f targets, G2M checkpoints and myogenesis. However, in our RT-qPCR experiment, we found no significant difference in FOXF1 expression between KD and normal samples (P > 0.05). This may be due to the smaller number of samples in the verification set (5 vs. 5) compared to the microarray data (10 vs. 9).

Our clinical predictive model predicts that the risk of developing KD increases as the expression of FOXF1 and LPAR1 decrease and the expression of TRIM32 increases. Previous studies on these three genes support this prediction. FOXF1 is associated with tissue development and inhibition of FOXF1 may cause abnormalities in the cell cycle of wound tissue leading to impaired wound healing. The inhibition of LPAR1 leads to a decrease in chemotaxis which is crucial for the inflammatory response around the wound. Moderate migration of inflammatory cells such as macrophages, mast cells, and granulocytes helps to remove necrotic cell debris and repair fibers during wound healing. Decreased expression of LPAR1 inhibits the formation of lamellar pseudopodia at the leading edge of migrating cells which also slows down wound healing. TRIM32 promotes the degradation of PIAS4 in keratinocytes. Therefore, increased TRIM32 expression reduces the inhibitory effect of PIAS4 on the formation of keratinocytes, resulting in a large accumulation of keratinocytes around the wound, which secrete keratin fibers that are the main components of scar tissue.

In this study, immuno-infiltration analysis showed significant differences between keloid and normal tissue in CD4+ effector T cells, myeloid-derived suppressor cells, activated dendritic cells, immature dendritic cells, follicular helper T cells, and monocytes. The levels of CD4+ effector T cells, myeloid-derived suppressor cells, activated dendritic cells, and immature dendritic cells were significantly lower in KD tissues than in control tissues while the levels of follicular helper T cells and monocytes were significantly higher. It has been confirmed that the Th2 characteristic is possessed by KD (41). Our analysis suggested that FOXF1 and LPAR1 were significantly negatively correlated with monocytes and follicular helper T cells, and significantly positively correlated with myeloid-derived suppressor cells, and that the levels of monocytes and follicular T helper cells at the wound were significantly increased. Henderson et al. analyzed more than 100,000 human hepatocytes and identified a subset of macrophages associated with scarring. This group of macrophages express TRIM32 and CD9, are differentiated from circulating monocytes, and are known to promote fibrosis (42). Previous studies have reported that monocytes and macrophages are key components of the immune system and participate in the regulation of inflammatory immunity and tissue repair by activating T and B lymphocytes (43). Follicular helper T cells are involved in the humoral immune regulation of inflammation and play a crucial role in autoimmunity and tumor-related immunity (44). Myeloid-derived suppressor cells are a group of suppressor cells of bone marrow origin which are precursors of dendritic cells, macrophages, and granulocytes, and have the ability to significantly inhibit immune cell responses (45). Chronic inflammation and fibrosis may be caused by improper activation of the immune response mediated by macrophages, an example of which is the development of fibrosis in systemic sclerosis (46). The biomarkers we identified are related to monocytes, myeloid-derived suppressor cells, and follicular helper T cells, which may all play an important role in the formation of keloids.

In the immune infiltration analysis, we found that there were different degrees of correlation between the biomarkers and the infiltration of immune cells. In order to further explore the role of immune cells in the development of KD, we used the HPA database to explore the expression of the biomarkers in different cell types. LPAR1 was enriched in macrophages, T-cells and mast cells, TRIM32 was enriched in T cells and mast cells, while FOXF1 was not significantly expressed in any immune cells. In a mouse model of multiple sclerosis, Choi et al. found that LPAR1-3 antagonists increased cell infiltration and immune cell activation (including macrophages) (PMID:34666785). In addition, Choi et al. demonstrated that in the immune microenvironment of tumors, different LPA receptors promoted metastasis, which helped create a T cell rejection and pro-tumor microenvironment suitable for therapeutic intervention (PMID: 34788605). Wang et al. reported that in a mouse model of atopic dermatitis (AD), TRIM32 acted as a regulator of PKCζ and could control the differentiation of Th2 cells, which are very important for the pathogenesis of AD (PMID: 33096083). We believe that immune cells, in particular T cells, play an important role in the development of KD, and are expected to become a new target for KD immunotherapy. However, the molecular mechanisms involved need further investigation.

In this study the transcriptional regulatory network analysis indicated that FOXF1 and LPAR1 share two transcriptional regulatory factors, E2F1 and SP1. In addition, the two share three miRNAs, hsa-miR-200c-3p, hsa-miR-200b-3p, and hsa-miR-429. The downregulation of FOXF1 and LPAR1 in keloid patients could be caused by the inactivation of E2F1 and SP1 due to mutations or other factors, or by the effect of the three miRNAs. TRIM32 did not share any miRNAs with the other two genes. JUND and TP53 were predicted to target TRIM32, which may contribute to its upregulation.

This study has several limitations. First, our analysis was based on a limited number of clinical samples from public databases, and may suffer from poor statistical power due to the small sample size. In addition, our analysis of the expression patterns of the identified biomarkers was based on public databases, and further validation is necessary, which would need to be done by collecting a larger number of clinical samples or conducting animal experiments. Given these limitations, larger datasets are needed to support further research and validation of the genes and molecular mechanisms that we identified.

In this article we analyzed IRGs in KD, leading to the identification of two new biomarkers of keloid tissue. Further studies on IRGs in KD may lead to new tools for early diagnosis as well as the identification of novel drug targets for treatment of KD.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material .

The studies involving humans were approved by Research Ethics committee of the second hospital of shandong university. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

XCW: Data curation, Methodology, Validation, Writing – original draft, Writing – review & editing. XYW: Data curation, Validation, Writing – review & editing. ZL: Methodology, Resources, Writing – review & editing. LL: Data curation, Resources, Writing – review & editing. JZ: Methodology, Resources, Writing – review & editing. DJ: Data curation, Writing – review & editing, Methodology. GH: Methodology, Writing – review & editing, Resources.