KLK1 as an Epithelial-Specific Brake Inhibits Colorectal Tumorigenesis by Suppressing B1R-Mediated Fibroblast Phenotypic Transition

Abstract

Inflammatory bowel disease (IBD) is increasing worldwide, and the persistence of chronic inflammation may lead to colitis-associated colorectal cancer (CAC). KLK1 expression is reduced in colitis, and its potential role in the intestinal mucosal barrier is still unclear. Here, KLK1 is investigated whether a supplement can reduce colitis and colorectal carcinogenesis. This study investigated KLK1's protective function in intestinal barrier integrity using Dextran Sulfate Sodium Salt (DSS) / Azoxymethane (AOM)-DSS-induced colitis/CAC models, Apc-deficient mice, and human clinical samples. KLK1-AAV2 knockdown mice exhibited exacerbated colitis symptoms, including severe diarrhea and impaired mucosal barrier markers, while KLK1 levels are notably reduced in ulcerative colitis patients and colorectal cancer specimens. Mechanistically, bradykinin receptor B1 (B1R) upregulation in CAC models activated extracellular matrix pathways, driving fibroblast phenotypic shifts that disrupt stromal homeostasis. Crucially, KLK1 supplementation reversed these pathological changes, demonstrating its dual role in maintaining epithelial barrier function and regulating fibroblast-ECM interactions. These findings position KLK1 as a potential therapeutic target for colitis and CRC chemoprevention, offering novel insights into IBD pathogenesis through its modulation of mucosal protection and stromal remodeling processes.

Keywords: bradykinin B1 receptors; cancer‐associated fibroblasts; extracellular matrix; intestinal barrier; tissue kallikrein.

Figure 1

Single‐cell transcriptomic analysis reveals that the expression of KLK1 is downregulated in epithelial cells of human ulcerative colitis tissue. A) The schematics of sample collection, scRNA‐seq transcriptomic analysis of ulcerative colitis, and combined analysis with public datasets. B) Uniform manifold approximation and projection (UMAP) of 356 075 cells analyzed by scRNA‐seq across 50 samples. Data information was placed in Table S1 (Supporting Information). Clusters were annotated by the canonical markers. C) UMAP plot showing the expression levels of the selected markers in all cell subtypes. D) Box plot showing the mean expression of inflammation signatures in cells from different sample groups. The boxes indicate the 25% quantile, median, and 75% quantile; the points indicate the individual signatures. E) The expression levels of the KLKs family genes in different cell subtypes. Dot size indicates the fraction of expressing cells, and the colors represent normalized gene expression levels. F) The expression levels of the KLK1 in epithelial cell subtypes. Dot size indicates the fraction of expressing cells, and the colors represent normalized gene expression levels. G) UMAP plot showing the expression levels of KLK1 and MUC2 in all cells. H,I) Immunofluorescence staining of MUC2⁺ KLK1⁺ cells in human ulcerative colitis and healthy colon with KLK1 (green), MUC2 (red), and DAPI (blue) antibodies. Data are representative of three independent experiments (n=3 per group). J) The relative mRNA expression level of KLK1 between normal colon cells in C57BL/6 mice. Primer sequence information is shown in Table S2 (Supporting Information). Scale bars: 50 µm. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 2

KLK1 is decreased in DSS induced acute colitis model mice. A) Schematic diagram of the experimental design of 2% DSS‐induced acute colitis. B) Representative picture of the colons from different groups (n=5 per group) and C) colon length. D) Body weight curves denote the changes in the mean body weight of mice recorded daily in different groups. E) Photomicrographs show representative images of H&E staining. F) FITC‐Dextran 4 (FD4) was used to detect intestinal permeability in C57BL/6 mice (n=5 per group). G) Detection of mouse serum Klk1 concentration by ELISA (n > 5 per group). H) Relative mRNA level of Klk1 in 2% DSS‐induced acute colitis. All data are expressed as mean ± SEM from three independent experiments. I) Relative mRNA expression after KLK1 knockdown by siRNA. J) Relative mRNA levels of TJP1, OCCLUDIN, and CLAUDIN‐1 after KLK1 knockdown in NCM460. K,L) Immunofluorescence staining of KLK1, Zo‐1 and E‐cadherin in C57BL/6 mice colon with Klk1 (red), Zo‐1 (green), E‐cadherin (green), and DAPI (blue) antibodies (n=3 per group). M,N) Immunohistochemistry of Swiss rolls showing the changes in the expression of KLK1 and intestinal epithelial mucosal barrier markers in an acute colitis model (n=3 per group). Scale bars: 50 µm. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 3

KLK1 reduction by AAV in DSS‐induced inflammation aggravates intestinal mucosal barrier damage. A) Schematic diagram of the experimental design of reducing KLK1 by AAV in the DSS‐induced colitis model. B) High‐resolution endoscopic images of the colon in C57BL/6 mice with AAV‐induced acute colitis treated with H₂O and 2% DSS, respectively. C) Representative picture of the colons from different groups and D) colon length (n=5 per group). E,F) The representative images of H&E staining and histopathological grading of inflammation in C57BL/6 mice colon with AAV‐induced acute colitis treated with H2O and 2% DSS (n=3 per group). G) Detection of mouse serum Klk1 concentration by ELISA (n=5 per group). H–J) Immunofluorescence showing the protein expression of KLK1, and all data are expressed as mean ± SEM from three independent experiments (n=3 per group). K) FITC‐Dextran 4 (FD4) was used to detect intestinal permeability in C57BL/6 mice (n=3 per group). L,M) IHC staining showing the intestinal mucosal barrier‐related indicators after KLK1 reduction by AAV. Scale bars: 20 µm. (n=3 per group). N–R) Protein levels of Klk1 and intestinal epithelial mucosal barrier markers in Klk1 AAV model combined with 2%DSS. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 4

KLK1 can improve the mucosal barrier damage caused by DSS‐induced acute colitis. A) Schematic diagram of the treatment of DSS‐induced acute colitis with rKLK1 and 5‐ASA, respectively. B) High‐resolution endoscopic images of the colon in C57BL/6 mice with DSS‐induced acute colitis treated with rKLK1 and 5‐ASA, respectively. C) Body weight curves denote the changes in the mean body weight of mice recorded daily in different groups. D) Disease activity index (DAI) score of C57BL/6 mice colon with DSS‐induced acute colitis treated with rKLK1 and 5‐ASA.E) Representative picture of the colons from different groups (n=5 per group) and F) colon length. G) Detection of mouse serum Klk1 concentration by ELISA (n > 3 per group). H) Photomicrographs show representative images of H&E staining (n=3 per group). I) The histopathological grading of inflammation in C57BL/6 mice colon with DSS‐induced acute colitis treated with rKLK1 and 5‐ASA was recorded. J) FITC‐Dextran 4 (FD4) was used to detect intestinal permeability in C57BL/6 mice (n > 3 per group). K,L) Immunofluorescence staining of KLK1, Zo‐1, and E‐cadherin in C57BL/6 mice colon with KLK1 (red), Zo‐1 (green), E‐cadherin (green), and DAPI (blue) antibodies (n=3 per group). M,N) Immunohistochemistry of Swiss rolls showing the changes in the expression of KLK1 and intestinal epithelial mucosal barrier markers in the treatment of DSS‐induced acute colitis with rKLK1 and 5‐ASA, respectively (n=3 per group). O, P) Protein levels of Klk1, Zo‐1, E‐cadherin and Occludin in C57BL/6 mice colon with DSS‐induced acute colitis treated with rKLK1 and 5‐ASA, respectively. Scale bars: 50 µm. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 5

KLK1 protects the intestinal mucosal barrier through B1R. A) Schematic diagram of the treatment of DSS‐induced acute colitis with rKLK1 and different receptor inhibitors, including B1R inhibitor SSR240612 and B2R inhibitor Icatibant. B) Body weight curves denote the changes in the mean body weight of mice recorded daily in different groups. C) Relative mRNA level of Klk1 in different C57BL/6 mice colon groups. All data are expressed as mean ± SEM from three independent experiments (n = 3 per group). D) Representative picture of the colons from different groups and E) colon length (n=5 per group). F,G) Immunohistochemistry of Swiss rolls shows changes in the expression of Klk1 and intestinal epithelial barrier markers in B1R and B2R inhibitor models combined with 2% DSS‐induced acute colitis and rKLK1. Scale bars: 50 µm (n = 3 per group). H,I) Immunofluorescence staining of BDKRB1⁺ cells in healthy controls (HC) and ulcerative colitis (UC) patients, BDKRB1 (green) and DAPI (grey) antibodies. Patient information is provided in Table S4 (Supporting Information). Scale bars: 20 µm. (n = 15 per group). J) The immunofluorescence intensity of BDKRB1 is significantly positively correlated with the Mayo score of ulcerative colitis. K) Volcano plot showing differentially expressed genes in KLK1 after NCM460 knockdown (KD) relative to the negative control (NC). L) Bar plot showing GO and KEGG representative pathways enriched by RNA‐Seq after KLK1 knockdown in NCM460 cell lines. M) GSEA enrichment analysis showed that KLK1 significantly downregulated the tight junction pathway and upregulated the ECM pathway after NCM460 knockdown (KD) compared with the negative control (NC). All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 6

KLK1 downregulation promotes fibroblast phenotype transformation. A) Representative picture of the C57BL/6 mice colons from the AOM‐DSS‐induced inflammation‐cancer transformation model and treatment with rKLK1 or rKLK1 combined with SSR240612, respectively (n= 5 per group). B) High‐resolution endoscopic images of the colon in C57BL/6 mice with AOM‐DSS‐induced inflammation‐cancer transformation model and treatment with rKLK1 or rKLK1 combined with SSR240612, respectively. C) Photomicrographs show representative images of H&E staining (n=3 per group). D–F) The colon length (D), number of tumors (E), and histology score (F) in C57BL/6 mice with AOM‐DSS‐induced inflammation‐cancer transformation model and treatment with rKLK1 or rKLK1 combined with SSR240612, respectively. G,H) Immunofluorescence staining of Fibronectin (green), Collagen I (yellow), Collagen III (red), and DAPI (blue) antibodies in AOM‐DSS‐induced inflammation‐cancer transformation model and treatment with rKLK1 or rKLK1 combined with SSR240612, respectively. Scale bars: 20 µm. (n=3 per group). I) Detection of mouse serum Klk1 concentration by ELISA (n=9 per group). J) Detection of human serum KLK1 concentration by ELISA during different group of the development of colorectal cancer. Patient information is provided in Table S3 (Supporting Information). K–M) Representative images of IHC staining for KLK1 and ZO‐1 during different group of the development of colorectal cancer and L,M) Data are representative of five independent experiments. Patient information is provided in Table S4 (Supporting Information) (n=5 per group). N) Pearson correlation analysis of IHC scores of KLK1 and ZO‐1. O) Pearson correlation analysis between the IHC score of KLK1 and the corresponding Mayo score of patients with ulcerative colitis. P) Relative mRNA level of KLK1 in different cell lines. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 7

Analysis of KLK1 expression in AOM‐DSS‐induced inflammation‐cancer transformation scRNA‐seq dataset. A) The schematics of sample collection, scRNA‐seq transcriptomic analysis of the AOM‐DSS‐induced inflammation‐cancer transformation model. B) UMAP of 93 158 cells analyzed by scRNA‐seq across 14 samples. Clusters were annotated by the canonical markers and a feature plot showing the expression of Klk1 in all cells. C) Feature plot showing the expression levels of the selected markers in all cell subtypes. D) The expression levels of Bdkrb1 and Klk1 in different cell subtypes during the scRNA‐seq dataset of C57BL/6 mice transformed by AOM‐DSS‐induced inflammation and cancer. Dot size indicates the fraction of expressing cells, and the colors represent normalized gene expression levels. E) The expression levels of B1R in different cell subtypes and fibroblast subtypes. Dot size indicates the fraction of expressing cells, and the colors represent normalized gene expression levels. F) AOM‐DSS‐induced inflammation‐cancer transformation model scRNA‐seq data analysis of pseudotime differentiation trajectories of fibroblast subsets in the in C57BL/6 mice. G) RNA‐seq data showed FPKM expression of MMP2 and MMP9 in KLK1‐knockdown (KD) relative to negative control (NC). H) Dot plot showing the expression levels of Ctnnb1, Fap, Dcn, Mmp2, and Mmp9 in different groups of AOM‐DSS‐induced inflammation‐cancer transformation model scRNA‐seq data. Dot size indicates the fraction of expressing cells, and the colors represent normalized gene expression levels. I,J) Representative images of IHC staining for Klk1, Mmp2, and Mmp9 during different groups (n=3 per group). K,L) Immunofluorescence staining of β‐catenin (red), Fap (red), Decorin (red), and DAPI (blue) antibodies in AOM‐DSS‐induced inflammation‐cancer transformation model and treatment with rKLK1 and SSR240612, respectively (n=3 per group). Scale bars: 50 µm. All data are shown as the mean ± SEM. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 8

KLK1 supplementation prevents adenomatous carcinogenesis in APC‐deficient intestine. A) High‐resolution endoscopic images of the colon in Apc ^min/+ mice which treatment with rKLK1 or rKLK1 combined with SSR240612, respectively. B) Photomicrographs show representative images of H&E staining and C) The histology score (n = 3 per group). D) Representative picture of the Apc ^min/+ mice colons in treatment with rKLK1 or rKLK1 combined with SSR240612, respectively, and E) Number of tumors (n = 3 per group). F) Detection of mouse serum Klk1 concentration by ELISA (n > 3 per group). G) UMAP of wild type or Apc ^min/+ mice analyzed by scRNA‐seq across 6 samples, and H) Clusters were annotated by the canonical markers. I) UMAP plot showing the expression levels of Klk1 in all cell subtypes. J) The expression levels of Klk1 and Bdkrb1 in different cell subtypes, and K) Ctnnb1, Fap, Dcn, Mmp2, and Mmp9 in different groups. Dot size indicates the fraction of expressing cells, and the colors represent normalized gene expression levels. L,M) Representative images of IHC staining for Klk1, Mmp2, and Mmp9 during different groups (n = 3 per group). N,O) Immunofluorescence staining of β‐catenin (red), Fap(red), Decorin (red), and DAPI (blue) antibodies in Apc ^min/+ mice, which treatment with rKLK1 and SSR240612, respectively (n = 3 per group). Scale bars: 50 µm. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 9

The decrease of KLK1 may be due to the transcriptional regulation by EGR1. A) Heatmaps showing transcription factors enriched in epithelial cells of different groups using SCENIC analysis of the AOM‐DSS‐induced inflammation‐cancer transformation scRNA‐seq dataset (left) and the Apc ^min/+ adenoma carcinoma scRNA‐seq dataset (right). B) The Venn diagram shows the intersection of transcription factors enriched in epithelial cells of different groups by SCENIC analysis of the AOM‐DSS‐induced inflammation‐cancer transformation scRNA‐seq dataset (left) and the Apc ^min/+ adenoma carcinoma scRNA‐seq dataset (right), and the transcription factors predicted in the hTFtarget database that may regulate KLK1 in colon (bottom). The table shows the transcription factors in the Venn diagram. C,D) Schematic diagram of constructing EGR1 overexpression plasmid and transfecting it into goblet cell line HT29‐MTX (C) and qPCR verification of the relative mRNA expression levels of EGR1 and KLK1 under EGR1 overexpression and inflammatory factor stimulation (D). E,F) Schematic diagram of the dual luciferase reporter gene experiment of EGR1 and KLK1 (E) and the relative luciferase activity of KLK1 promoter binding under the conditions of EGR1 overexpression and inflammatory factor stimulation (F). G) The Jaspar database was used to predict the possible binding sites between EGR1 and the pre‐transcriptional start region of KLK1. H) The ChIP method was used to verify the binding site of EGR1 and KLK1. The results were expressed by PCR. I) qPCR showing EGR1 was significantly enriched under the stimulation of inflammatory factors. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.

Figure 10

Role of Bradykinin B1 Receptor in Fibroblast Phenotypic Transition and Inflammatory Progression in Colitis. A) Schematic diagram of the experimental design for co‐culturing the supernatant of NCM460 induced by inflammatory factors with CCD‐18Co colonic fibroblasts. B) Relative mRNA levels of KLK1 and epithelial mucosal barrier genes in NCM460 cells treated with Inflammatory factors. C) After co‐culture of NCM460 supernatant induced by inflammatory factors with CCD‐18Co colonic fibroblasts, the mRNA levels of MMP pathway‐related indicators in CCD‐18Co cells were detected by qPCR. D) Relative mRNA expression of markers related to the ECM pathway and fibroblast phenotype transition (n = 3 per group). E,F) ELISA detection of Il‐1β, Tnf‐α, and Il‐6 levels in mouse serum in the KLK1‐AAV combined with 2% DSS model (E) and the SSR240612 inhibitor model (F) (n = 3 per group). G,H) Western blot detection of key phosphorylated proteins of MAPK and NF‐κB pathways in the KLK1‐AAV combined with 2% DSS model, and quantification of B1r, p‐pErk, p‐p38, and p‐p65 protein levels (n = 3 per group). I) Activation of the KLK1‐[Lys‐des‐Arg⁹‐BK]‐B1R axis under colitis promotes the occurrence of colorectal cancer by activating the MAPK pathway and transforming the phenotypic changes of fibroblasts. All data are shown as the mean ± SEM. Data are representative of three independent experiments. The P value was analyzed by one‐way ANOVA with Tukey's multiple comparisons, and all P values are marked with specific values in the graph.