Abstract
This study aimed to evaluate Keratin 19 (KRT19) as a potential prognostic biomarker for the diagnosis and prognosis of pancreatic adenocarcinoma (PAAD). Using data from The Cancer Genome Atlas and the Gene Expression Omnibus, we analyzed KRT19 expression in PAAD tissues and identified differentially expressed genes associated with KRT19. Gene ontology (GO) and gene set enrichment analysis were performed to explore the underlying mechanisms of KRT19 in PAAD progression. Spearman correlation analysis was used to assess the relationships between KRT19 expression and immune cell infiltration, immune checkpoint genes, and TP53 expression. Logistic regression was employed to examine the association between KRT19 expression and clinicopathological features. Kaplan–Meier survival curves, receiver operating characteristic curves, a nomogram model, and Cox regression analyses were used to evaluate the diagnostic and prognostic value of KRT19. KRT19 expression was significantly higher in tumor tissues than in adjacent non-tumor tissues and was closely associated with immune cell infiltration, HAVCR2 expression, and TP53 status. KRT19 levels correlated with T stage, overall survival (OS), disease-specific survival, and histologic grade. Cox regression and receiver operating characteristic analysis further indicated that KRT19 expression is an independent risk factor for OS, disease-specific survival, and progression-free interval in PAAD patients and can effectively distinguish tumor from normal tissue. In conclusion, the findings suggest that KRT19 could serve as a potential biomarker for the diagnosis and prognosis of PAAD, and it may be implicated in the regulation of the immune microenvironment.
Keywords: biomarker, immune cell infiltration, immune checkpoints, KRT19, pancreatic adenocarcinoma
1. Introduction
Pancreatic adenocarcinoma (PAAD) is one of the most troublesome malignancies in clinical practice. Its onset is concealed, develops rapidly, and the mortality rate is extremely high.[1,2] The latest data released by the National Cancer Center of China in 2022 shows that pancreatic cancer ranks 8th in the incidence rate of malignant tumors in men, 12th in women and 6th in the mortality rate of malignant tumors in China.[3] PAAD is difficult to diagnose in the early stage, has low surgical resection rate, and has highly malignant biological behavior, with poor prognosis. Although the 5-year survival rate of PAAD has broken through in the past decade or so, the survival rate is still the lowest among all malignant tumors.[4,5] The lack of effective early diagnostic biomarkers is part of the cause of high mortality in PAAD patients. Therefore, it is urgent to explore the molecular mechanism of PAAD development and screen highly sensitive and specific biomarkers to improve the diagnosis, treatment and prognosis of PAAD.
Keratin 19 (KRT19) is a protein-coding gene located on chromosome 17q21-1 and is a member of the keratin family. Studies have shown that KRT19 regulates breast cancer properties by activating AKT signaling pathways. In addition, recent studies have shown that regulation of KRT19 expression has different effects on cell proliferation, survival, invasion, migration and apoptosis, depending on the type of cancer cell.[6,7] These reports suggest that the abnormal expression of KRT19 is closely related to the occurrence and development of cancer, which is of great significance in targeting research and clinical diagnosis.
Previous studies have established KRT19 as a pivotal hub gene in PAAD, demonstrating significant overexpression at both the RNA and protein levels.[8,9] However, research remains limited regarding the specific roles of KRT19 in PAAD, particularly its associations with immune cell infiltration, immune checkpoint expression, genetic mutation profiles, and its utility in diagnosis and prognosis. Therefore, this study aims to employ bioinformatic approaches to systematically analyze the expression profile of KRT19 and its correlation with immune infiltration in PAAD. Furthermore, we seek to evaluate its potential value for diagnosis and prognostic assessment, with the goal of providing novel molecular markers and a theoretical foundation for the clinical management of PAAD.
2. Materials and methods
2.1. Data collection and processing
The Transcriptome data and clinical data of patients with PAAD involved in this study are from the Cancer Genome Alas (TCGA) and Gene Expression Omnibus (GEO) databases, and data processing using numerous bioinformatics methods.
2.2. Bioinformatics analysis of KRT19 expression in pancreatic adenocarcinoma and normal tissue samples
Transcriptomic and clinical data utilized in this study were systematically curated from multiple public databases. Transcriptomic profiles across 33 cancer types were obtained from TCGA database. For PAAD, clinical and molecular data from 179 tumor tissue samples and 4 normal pancreatic tissue samples were extracted. To validate the expression pattern of KRT19 in PAAD, mRNA expression profiles from 3 independent datasets (GSE101448, GSE16515, and GSE15471) were retrieved from the GEO database. Based on the median expression value of KRT19 in the TCGA-PAAD cohort, the 179 PAAD samples were stratified into KRT19-high and KRT19-low expression groups. Differential expression analysis was performed using the DESeq2 (v1.26.0) R package,[10] which employs a negative binomial generalized linear model to identify differentially expressed genes (DEGs). The analysis incorporated read depth normalization and dispersion estimation to ensure robust detection of genes with significant expression differences between groups. For visualization of the results, the ggplot2 (v3.3.3) R package was utilized to generate 2 primary types of plots:
Volcano plots to display the global distribution of gene expression changes, represented by log2 fold change values, against their statistical significance (-log10 adjusted P-values), highlighting genes that met the significance threshold (adjusted P-value < .05).
Heatmaps to depict the expression patterns of significantly dysregulated genes across individual samples, providing a clear overview of transcriptional heterogeneity and group-specific signatures. This integrated approach ensures rigorous analysis and intuitive visualization of multi-source data, aligning with best practices in bioinformatic research.
2.3. Enrichment analysis of KRT19 related differentially expressed genes in PAAD
In this study, Gene Set Enrichment Analysis (GSEA) was performed on differentially expressed genes associated with KRT19 using the R package ClusterProfiler (v3.14.3).[11] The analysis utilized the reference gene set “c2.cp.v7.2.symbols.gmt.” Significantly enriched gene sets were identified based on the following criteria: absolute normalized enrichment score |NES| > 1, false discovery rate (FDR) < 0.05, and adjusted P-value (p.adjust) < .05.
2.4. Comprehensive analysis of immune cell infiltration in pancreatic cancer
To comprehensively evaluate the immune cell infiltration levels in PAAD and control samples, this study employed 2 computational methodologies: single-sample Gene Set Enrichment Analysis (ssGSEA) and the CIBERSORT deconvolution algorithm. The combined use of these approaches aimed to provide complementary and validated insights into the tumor immune microenvironment.
Quantification of immune cell infiltration degrees was performed using the ssGSEA algorithm implemented in the R package “GSVA” (v1.34.0).[12] The analysis utilized a precompiled gene set comprising marker genes for 28 immune cell subtypes. This method transformed the gene expression profile of each individual sample into an immune cell enrichment score (ES) matrix. These enrichment scores represent the relative degree of enrichment for each immune cell type within a single sample, providing a normalized measure of cellular abundance.
To complement the ssGSEA results and obtain estimates of immune cell composition proportions, this study concurrently applied the CIBERSORT algorithm (https://cibersort.stanford.edu/). CIBERSORT is a deconvolution algorithm based on linear support vector regression (linear SVR). It leverages the LM22 signature matrix, a predefined gene expression signature encompassing 22 human immune cell subtypes, to dissect bulk tissue transcriptomic data and infer the relative proportions of various immune cell populations within each sample.
Following the acquisition of immune cell infiltration scores (ssGSEA) or proportion estimates (CIBERSORT), Spearman rank correlation analysis was employed to investigate potential associations between the expression level of KRT19 (represented as log2-transformed TPM values) and the infiltration levels of various immune cells. This analysis computed the Spearman correlation coefficient (ρ) and its corresponding significance level (FDR) between the KRT19 expression vector and the infiltration score/proportion vector for each immune cell type. A significance threshold of FDR < 0.05 was applied for all statistical tests. All statistical analyses were conducted within the R programming environment (v4.3.0).
2.5. Correlation analysis of expression of KRT19 in PAAD with some immune checkpoints and TP53
Spearman rank correlation analysis, a non-parametric method suitable for assessing monotonic relationships without assuming normal distribution of data, was employed to evaluate the associations between KRT19 expression levels and the expression of immune checkpoint genes (including HAVCR2, CTLA4, and PDCD1), as well as TP53 expression, in PAAD samples from the TCGA database. This method is robust against non-normal distributions and outliers, making it appropriate for exploring potential monotonic trends in molecular expression data.
All statistical analyses and visualizations were conducted in the R programming environment (version 4.3.0). The ggplot2 package (version 3.3.3) was used to generate scatter plots for visualizing the correlation patterns.
2.6. Correlation between KRT19 expression levels and clinicopathological characteristics in patients with PAAD
We retrieved clinicopathological data from The Cancer Genome Atlas pancreatic cancer (TCGA-PAAD) cohort and integrated previously published research data on PAAD patients. Based on KRT19 expression levels, patients were stratified into high-expression and low-expression groups. A comparative analysis was then conducted to assess differences in key clinicopathological parameters between these 2 groups, including race, disease-specific survival (DSS) events, overall survival (OS) events, T stage, and histological grade, to determine the significance of the correlation between KRT19 expression levels and these clinicopathological features.
Comparisons across different clinicopathological features and gene expression groups were performed using contingency table analyses, with the Pearson chi-square or Fisher exact test applied depending on the sample size and expected frequencies (latter used when any expected count was < 5).
2.7. Prognostic evaluation and Kaplan–Meier survival analysis
This study utilized data from TCGA-PAAD to evaluate the association between KRT19 expression levels and the prognosis of pancreatic cancer patients. The analytical process included data acquisition and preprocessing, Kaplan–Meier survival curve analysis, along with univariate and multivariate Cox proportional hazards regression analyses. All statistical analyses were performed using R language (version 4.3.0) and relevant packages, with the “survival” package (version 3.2-10) employed for model construction and hypothesis testing in survival analysis, and the “survminer” package (version 0.4.9) utilized for survival curve visualization.
3. Results
3.1. Pan-cancer analysis reveals differential KRT19 expression patterns with consistent upregulation in pancreatic adenocarcinoma
Comprehensive analysis of KRT19 expression patterns across 33 cancer types from TCGA revealed significant heterogeneity among different malignancies. Specifically, KRT19 demonstrated statistically significant upregulation in 11 cancer types (P < .05) while showing marked downregulation in 7 cancer types (P < .05) (Fig. 1A), suggesting its context-dependent biological functions within distinct tumor microenvironments.
Figure 1.
KRT19 expression was significantly upregulated in PAAD tissue. (A) KRT19 expression in 33 types of cancer tissues and adjacent normal tissues from the TCGA database. (B, C, D, E) KRT19 was significantly increased in PAAD tissues from GSE101448 (B), GSE16515 (C), GSE15471(D), TCGA-PAAD (E). (F, G) The volcano plots (F) and the heat maps (G) show the expression levels of specific mRNAs in the PAAD patients with high- and low KRT19 expression from the TCGA-PAAD project.
Further investigation in PAAD demonstrated consistent overexpression of KRT19 in tumor tissues compared to adjacent normal controls. This observation was robustly validated across multiple independent datasets (GSE101448, GSE16515, GSE15471, and TCGA-PAAD) (Figs. 1B-E), with all expression differences reaching statistical significance (P < .05). These consistent findings indicate that KRT19 likely plays a critical oncogenic role in pancreatic cancer pathogenesis and progression.
3.2. Identification of differentially expressed genes between KRT19 high- and low-expression groups
PAAD patients (n = 183) were stratified into KRT19 high-expression and low-expression groups based on median KRT19 expression levels. Differential expression analysis identified 1387 significantly dysregulated genes using threshold parameters of |log₂FC| > 1.5 and adjusted P < .05. Among these, 242 genes were significantly upregulated while 1145 genes were downregulated in the high-expression group compared to the low-expression group(Fig. 1F). A single-gene co-expression heatmap (Fig. 1G) displays the top 10 most significantly differentially expressed genes, revealing distinct expression patterns between the 2 groups.
3.3. Functional enrichment analysis of KRT19 related differentially expressed genes in PAAD
To investigate the biological functions of KRT19-related DEGs in PAAD, Gene Ontology (GO) enrichment and Gene Set Enrichment Analysis (GSEA) were performed. As illustrated in Figure 2A, the DEGs were significantly enriched in several key biological processes, including “signal release,” “regulation of membrane potential,” “regulation of trans-synaptic signaling,” and “modulation of chemical synaptic transmission.” Within the cellular component category, the most prominent terms were “presynapse,” “transmembrane transporter complex,” “ion channel complex,” and “cation channel complex.” For molecular function, the highly enriched entries comprised “channel activity,” “gated channel activity,” “ion gated channel activity,” and “cation channel activity.”
Figure 2.
Functional enrichment analysis of differentially expressed genes (DEGs) based on the expression level of KRT19 in PAAD. (A) GO enrichment analysis revealed enriched biological functions (BP), cellular components (CC) and molecular functions (MF). (B-I) GSEA analysis was performed to further screen the significant pathway base on the KRT19 associated DEGs in PAAD.
Furthermore, GSEA revealed significant enrichment of KRT19-linked DEGs in multiple critical pathways, including muscle contraction, FCGR activation, complement cascade, GABA receptor activation, potassium channels, FCGR3A-mediated IL-10 synthesis, costimulation by the CD28 family, and type 2 diabetes mellitus (Figs. 2B–I). These results suggest that KRT19 may influence PAAD progression through mechanisms involving synaptic signaling, ion channel activity, and immune-inflammatory pathways.
3.4. The expression level of KRT19 is related to the infiltration of multiple immune cell types in the PAAD Tissues
To investigate the role of KRT19 in the tumor immune microenvironment of PAAD, we evaluated the Spearman correlation between KRT19 expression and immune cell infiltration levels in tumor tissues using the ssGSEA and CIBERSORT algorithms. The results demonstrated that KRT19 expression was significantly correlated with the infiltration levels of various immune cells, revealing a complex immunoregulatory pattern.
Notably, as a marker of epithelial cell characteristics, cytokeratin 19 (KRT19) expression is closely associated with a decrease in the infiltration levels of multiple key anti-tumor immune cells. In the results of single-sample Gene Set Enrichment Analysis (ssGSEA) and CIBERSORT (as shown in Figures 3A–B), this correlative trend is particularly consistent in CD8 + T cells (ssGSEA: rho = -0.258, adjusted P-value (p_adj) = 0.001; CIBERSORT: rho = -0.279, p_adj < 0.001) and monocytes (ssGSEA: rho = -0.362, p_adj < 0.001; CIBERSORT: rho = -0.293, p_adj < 0.001) (see Figures 3C–D). Furthermore, the reduced infiltration levels of other immune cells with effector functions, such as effector memory CD4 + T cells, follicular helper T cells (Tfh cells), and natural killer (NK) cells, are also associated with KRT19 expression. These findings suggest that KRT19 may play an immunosuppressive role in the tumor microenvironment of PAAD.
Figure 3.
The relationship between KRT19 expression and immune cell infiltration in PAAD patients. (A) Bubble plot showing the Spearman correlation between KRT19 expression and the infiltration levels of 28 immune cell types, as assessed by the ssGSEA algorithm. (B) Bubble plot of the Spearman correlation between KRT19 expression and the infiltration levels of 22 immune cell types, analyzed using the CIBERSORT algorithm (Bubble size represents the absolute value of rho (│rho│), and color indicates the adjusted P-value).(C) Box plots comparing the infiltration levels of CD8 + T cells and monocytes between KRT19-high and KRT19-low expression groups, as estimated by the ssGSEA algorithm (*FDR < 0.05). (D) Box plots comparing the infiltration levels of CD8 + T cells and Monocytes between KRT19-high and KRT19-low expression groups, as estimated by the CIBERSORT algorithm (*FDR < 0.05).
However, the analysis also reveals the complexity of its role: KRT19 expression is closely associated with the increase in certain immune cell subsets, with the most prominent associations observed in M0 macrophages (rho = 0.412, p_adj < .001) and regulatory T cells (Tregs) (rho = 0.249, p_adj = .004). These 2 cell types are typically linked to immunosuppression and tumor progression.
These results indicate that high expression of KRT19 is significantly associated with an immunosuppressive microenvironment in PAAD, characterized by reduced infiltration of cytotoxic immune cells and an increase in certain immunosuppressive cell types. This provides important insights into the potential mechanisms by which KRT19 may promote tumor immune evasion.
3.5. Assessment of the correlation between KRT19 expression level and immune checkpoint gene and TP53 in the PAAD Tissues
We next evaluated the correlation between KRT19 expression and key immunomodulatory factors, including the checkpoint genes HAVCR, CTLA-4, and PDCD-1, as well as the tumor suppressor TP53, which is frequently dysregulated in malignancies. Our analysis in the TCGA PAAD cohort revealed that while KRT19 levels showed no significant correlation with the expression of CTLA-4 or PDCD-1, they were positively correlated with both HAVCR (TIM-3) and TP53 (Fig. 4A–D). These findings suggest that KRT19 may contribute to an immunosuppressive microenvironment in PAAD not through the CTLA-4/PD-1 axis but potentially via the HAVCR pathway, while also aligning with TP53-associated oncogenic processes.
Figure 4.
The correlation analysis between the expression levels of KRT19, HAVCR, CTLA-4, PDCD-1, and TP53 in PAAD. (A-D) The correlation analysis results between the expression levels of EXO1 and the expression levels of (A) HAVCR, (B) CTLA-4, (C) PDCD-1, and (D) TP53 in the TCGA-PAAD dataset.
3.6. KRT19 expression level is related to multiple clinicopathological characteristics of PAAD
This study evaluated the relationship between clinicopathological characteristics and KRT19 expression levels in patients with PAAD (Table 1). The results demonstrated significant differences in histological grade, OS, and DSS between PAAD patients with high and low KRT19 expression. Additionally, no significant differences in KRT19 expression levels were observed among different ethnic groups (Fig. 5A). Regarding clinical staging, KRT19 expression was significantly higher in T3 and T4 stages compared to T1 and T2 stages (Fig. 5B). In terms of tumor differentiation, KRT19 expression levels were significantly elevated in G3 and G4 grades relative to G1 and G2 grades (Fig. 5C). Further analysis revealed that within the deceased patient group, KRT19 expression was significantly increased in both DSS and OS periods (Figs. 5D, 5E). These findings indicate that KRT19 expression levels are closely associated with histological features, clinical stage, and survival prognosis in PAAD, suggesting its potential role as a prognostic biomarker.
Table 1.
The clinicopathological characteristics of PAAD patients with high and low expression levels of KRT19.
| Characteristic | Low expression of KRT19 | High expression of KRT19 | P |
|---|---|---|---|
| Total number of patients | 89 | 89 | |
| T stage, n (%) | .493 | ||
| T1 | 3 (1.7%) | 4 (2.3%) | |
| T2 | 15 (8.5%) | 9 (5.1%) | |
| T3 | 67 (38.1%) | 75 (42.6%) | |
| T4 | 2 (1.1%) | 1 (0.6%) | |
| Pathologic stage, n (%) | .850 | ||
| Stage I | 11 (6.3%) | 10 (5.7%) | |
| Stage II | 71 (40.6%) | 75 (42.9%) | |
| Stage III | 2 (1.1%) | 1 (0.6%) | |
| Stage IV | 3 (1.7%) | 2 (1.1%) | |
| Race, n (%) | 1.000 | ||
| Asian | 6 (3.4%) | 5 (2.9%) | |
| Black or African American | 3 (1.7%) | 3 (1.7%) | |
| White | 78 (44.8%) | 79 (45.4%) | |
| Histologic grade, n (%) | .045 | ||
| G1 | 20 (11.4%) | 11 (6.2%) | |
| G2 | 48 (27.3%) | 47 (26.7%) | |
| G3 | 18 (10.2%) | 30 (17%) | |
| G4 | 2 (1.1%) | 0 (0%) | |
| OS event, n (%) | .004 | ||
| Alive | 53 (29.8%) | 33 (18.5%) | |
| Dead | 36 (20.2%) | 56 (31.5%) | |
| DSS event, n (%) | .014 | ||
| Alive | 59 (34.3%) | 41 (23.8%) | |
| Dead | 28 (16.3%) | 44 (25.6%) | |
| Age, mean ± SD | 64.44 ± 9.7 | 65.06 ± 11.84 | .704 |
PAAD = prognosis of pancreatic adenocarcinoma.
Figure 5.
KRT19 expression levels correlate with multiple clinicopathological characteristics of PAAD patients. (A-E) The correlation analysis between KRT19 expression levels and (A) race, (B) T stages, (C) histologic grades, (D) DSS, (E) OS of PAAD patients. *P < .05.
3.7. KRT19 is a potential prognostic and diagnostic biomarker for PAAD
K–M survival curve analysis presented that PAAD patients with a low-KRT19 expression were associated with significantly higher OS (P = .001) (Fig. 6A), DSS (P = .003) (Fig. 6B), and PFI (P = .002) (Fig. 6C) compared to those with high-KRT19 expression levels. Multivariate Cox regression analysis indicated that KRT19 was an independent risk factor for predicting OS (hazard ratio [HR]: 1.638, P = .025), DSS (HR: 1.710, P = .028), and PFI (HR: 1.522, P = .045). T stage was also an independent predictor of OS, DSS, and PFI (Table 2). The KRT19 expression levels could differentiate PAAD tissues from adjacent peritumoral tissues with an AUC of 0.974 (Fig. 7A). The time-dependent receiver operating characteristic curve analysis showed that AUC values were higher than 0.6 (Fig. 7B) for 1, 3, and 5year survival rates of PAAD patients based on KRT19 expression levels. We constructed a nomogram model with T stage, pathological stage and KRT19 expression level as parameters. Based on multivariate Cox regression analysis, these factors were identified as highly significant prognostic factors (Fig. 7C). Nomogram has certain clinical value in predicting 1-year, 3-year, and 5-year survival probability of PAAD patients.
Figure 6.
KRT19 shows a high prognostic prediction value in PAAD patients. (A) overall survival, (B) disease-specific survival, (C) progression-free interval.
Table 2.
Cox regression analysis of clinical outcomes in PAAD patients.
| Characteristics | HR for overall survival (95% CI) |
HR for disease-specific survival (95% CI) | HR for progression-free interval (95% CI) | |||
|---|---|---|---|---|---|---|
| Univariate | Multivariate | Univariate | Multivariate | Univariate | Multivariate | |
| T stage (T1&T2 vs T3&T4) | 2.023* | 1.967* | 3.119** | 2.762* | 2.414** | 2.051* |
| Pathologic stage (Stage I & Stage II vs Stage III & Stage IV) | 0.673 | 0.810 | 1.109 | |||
| Histologic grade (G1&G2 vs G4&G3) | 1.538 | 1.265 | 1.616 | 1.326 | 1.684* | 1.346 |
| Race (Asian vs White) | 1.257 | 1.704 | 1.058 | |||
| KRT19 (Low vs High) | 1.988*** | 1.638* | 2.073** | 1.710** | 1.862** | 1.522* |
CI = confidence interval, HR = hazard ratio, PAAD = prognosis of pancreatic adenocarcinoma.
P < .05.
P < .01.
P < .001.
Figure 7.
The KRT19 shows a good diagnostic and prognostic performance in PAAD. (A) Diagnostic ROC curve differentiating PAAD tissues from normal tissues based on KRT19 expression levels. (B) Time-dependent survival ROC curves to predict 1-, 3-, and 5-year survival rates of PAAD patients based on the KRT19 expression levels. (C) ROC curve analysis to evaluate the prediction efficacy of the nomogram model that includes clinicopathological factors (T stages, pathologic stages) and KRT19 expression levels to predict the 1-, 3-, and 5-year survival rates of PAAD patients. ROC = receiver operating characteristic.
3.8. Prognostic performance of KRT19 in clinical pathological subgroups of PAAD patients
Cox regression analysis based on clinicopathological parameters was used to determine the predictive value of KRT19 in specific subgroups of PAAD patients. High KRT19 expression levels were associated with unfavorable OS in the PAAD patients of clinical T1 and T4 stages (HR = 5.75, P = .029), clinical histologic grades, G1 and G2 (HR = 2.52, P = .003) (Fig. 8A), and DSS. High KRT19 expression levels were also closely related to lower PFI in clinical T1 and T2 stages (HR = 0.022, P = .001) (Fig. 8B) and clinical histologic grades, G1and G2 (HR = 1.93, P = .001) (Fig. 8C).
Figure 8.
Forest plot of the association between KRT19 expression and survival in PAAD patients from Cox regression analysis. (A–C) The Cox regression analysis results show the prognostic performance of KRT19 expression levels regarding (A) overall survival, (B) disease-specific survival, and (C) progression-free interval in different subgroups of PAAD patients. The results are represented by the hazard ratio (HR).
4. Discussion
PAAD is one of the most malignant tumors in the digestive tract.[13,14] Carbohydrate antigen 19 to 9 (CA199) is the most commonly used diagnostic marker for PAAD. However, CA199 does not increase significantly in the early stages of the disease, and its sensitivity and specificity are only 70% to 80%.[15] Therefore, in-depth study on the molecular mechanism of PAAD progression and excavation of new biomarkers with high specificity and sensitivity is one of the important measures to improve the survival and prognosis of PAAD patients.
In this study, we found that KRT19 was significantly overexpressed in 11 out of 33 human cancer tissues. Additionally, high expression of KRT19 was observed in PAAD tissues from TCGA and GEO databases. This suggests that KRT19 may be a potential biomarker for PAAD. GO analysis revealed that KRT19-associated DEGs were significantly enriched in biological pathways related to synaptic signaling and ion channel transport. Although not considered classical hallmarks of PAAD, these processes are increasingly recognized for their roles in tumor biology. For example, dysregulation of ion channels (e.g., potassium and calcium channels) has been linked to cancer proliferation, migration, and therapy resistance. In PAAD specifically, certain ion channels modulate cellular electrophysiology, thereby enhancing cancer cell invasiveness.[16] Notably, PAAD exhibits a high propensity for perineural invasion, a unique pattern of tumor-nerve interaction. This close association suggests that tumor cells may co-opt neuronal signaling mechanisms, including synaptic transmission and membrane potential regulation, to promote proliferation, invasion, and communication within the tumor microenvironment.[17] GSEA results showed that PAAD tissues with high KRT19 expression were enriched in DEGs associated with FCGR activation, FCGR3A mediated IL-10 synthesis, complement cascade, and costimulation by the CD28 family activation. The accepted concepts of inflammatory pathways include immune complexes (IC) mediated complement activation and Fcgamma receptor (FcγR) triggering on Phagocyte.[18,19] Costimulation signals are mediated by cell surface protein families such as CD28, and T cell activation signals that occur in the absence of CD28 co stimulation can lead to T cell anergy.[20] Many scholars also believe that tumor microenvironment plays an important role in the occurrence, development and prognosis of tumors, and immune cells, as an important part of tumor microenvironment, participate in the guidance of disease prognosis and clinical treatment. Thus, we calculated an overview of immune cell infiltration in PAAD and adjacent tissues, and our study suggests a strong link between KRT19 expression and tumor-induced immune response. In PAAD, KRT19 expression was associated with lower infiltration levels of CD8 + T cells and monocytes. CD8 + T cells are key effector cells that mediate cytotoxic immune responses against tumors, and their density in the tumor microenvironment (TME) is positively correlated with patient prognosis and responsiveness to immunotherapy. Insufficient infiltration of CD8 + T cells compromises effective tumor cell recognition and elimination,[21,22] representing a critical mechanism of immune evasion in PAAD.
Monocytes serve as primary precursors of antigen-presenting cells, such as dendritic cells (DCs) and M1 macrophages, within the TME. Their infiltration level critically influences immune initiation efficiency. These cells can differentiate into DCs to activate CD8 + T cells via antigen presentation or into M1 macrophages that enhance anti-tumor immunity by secreting pro-inflammatory cytokines.[23] In this study, the correlation coefficient between KRT19 and monocytes was the highest among all analyzed immune cell types, suggesting that KRT19 exerts a particularly significant inhibitory effect on monocyte infiltration.
Furthermore, KRT19 expression was negatively correlated with the levels of effector memory CD4 + T cells, follicular helper T cells (Tfh), and natural killer (NK) cells, further supporting its role in suppressing broad-spectrum anti-tumor immunity in PAAD. Effector memory CD4 + T cells facilitate CD8 + T cell activation through cytokine secretion,[24] while Tfh cells enhance CD8 + T cell-mediated anti-tumor immunity indirectly via IL-21 production.[25,26] NK cells directly eliminate tumor cells.[27] The reduced infiltration of these immune subsets collectively impairs anti-tumor immune responses in PAAD through multiple mechanisms.
Additionally, our data demonstrate that KRT19 expression is positively correlated with increased infiltration of M0 macrophages and regulatory T cells (Tregs) in PAAD tissues. M0 macrophages represent unpolarized precursors that are predisposed to M2 polarization, an immunosuppressive phenotype-within the TME. M2-polarized macrophages suppress effector T cell function by secreting anti-inflammatory factors such as IL-10 and TGF-β, while also promoting tumor angiogenesis and metastatic invasion.[28,29] The increased abundance of M0 macrophages in this context provides a substantial reservoir of immunosuppressive precursor cells in the PAAD TME, which can rapidly polarize into M2 macrophages upon TME signaling, thereby fostering a tumor-promoting microenvironment.
Tregs are a specialized T cell subset that functionally suppress effector immune responses. While they play a physiological role in preventing autoimmunity by restraining autoreactive lymphocytes,[30] Tregs also exert pro-tumorigenic effects by mediating immunosuppressive surveillance within the TME.[31] Our findings suggest that elevated KRT19 expression may reinforce immunosuppression by promoting Treg recruitment to the tumor site.
We analyzed the relationship between KRT19 expression levels and immune checkpoint genes including CTLA-4, PDCD-1, HAVCR2, and TP53. HAVCR2 (aka TIM3), is another important key protein associated with tumor immune escape. HAVCR2 and PD-1 are co expressed in various tumors, and HAVCR2 expression levels increase in PD-1 resistant tumors. Increasing the expression of HAVCR2 may be one of the resistance mechanisms of PD-1 inhibitors.[32] In addition, TP53 is highly expressed in malignant tumors, and TP53 mutations are also markers of poor prognosis in many cancers. TP53 mutation inhibits anti-tumor immunity and reduces the efficacy of cancer immunotherapy. Our results suggest that KRT19 may provide a new approach for improving the efficacy of immunotherapy in patients with PAAD.[33,34]
Our investigation also showed that KRT19 expression levels in the PAAD tissues significantly correlated with the T stage, OS, DSS, and Histologic grade. In addition, the K-M survival curve showed that patients with higher KRT19 expression levels had lower OS, DSS, and PFI rates than patients with lower KRT19 expression levels. The AUC value of KRT19 expression level for differential diagnosis of PAAD is 0.974. Moreover, the AUC values for 1-year, 3-year, and 5-year predicted survival rates were all higher than 0.65. These data suggest that KRT19 is a potential diagnostic and prognostic biomarker for PAAD. We also established a nomogram model based on the results of multivariate Cox regression analysis, which showed that the expression level of KRT19 helps improve the prognosis evaluation of PAAD patients. Nevertheless, this study has several limitations. First, the analysis was solely based on bioinformatic computational methods, and the expression of KRT19 at the protein level, along with its specific functions, has not been experimentally validated. Second, the assessment of immune cell infiltration relied on a single algorithm. Future studies should incorporate multiple computational approaches or experimental techniques-such as immunohistochemistry or flow cytometry-for cross-validation. Most importantly, while our work primarily reveals a correlation between KRT19 and the immune microenvironment, the specific molecular mechanisms underlying these statistical associations require further elucidation through functional experiments.
In conclusion, this study is a good indication of the value of KRT19 in the diagnosis and prognosis of PAAD. KRT19 is associated with infiltration of immune cells. Meanwhile, KRT19 is closely related to the expression of immune checkpoint gene in PAAD, suggesting that PAAD may be a target for potential immune-related biomarker.
Author contributions
Conceptualization: Zhiqiang Chen.
Data curation: Zhiqiang Chen.
Writing – original draft: Zhiqiang Chen, Ling He.
Methodology: Ling He.
Software: Ling He, Li Li.
Writing – review & editing: Li Li.
Abbreviations:
- DGEs
- differentially expressed genes
- DSS
- disease-specific survival
- GEO
- Gene Expression Omnibus
- GO
- gene ontology
- GSEA
- gene set enrichment analysis
- KRT19
- Keratin 19
- OS
- overall survival
- PAAD
- pancreatic adenocarcinoma
- PFI
- progression-free interval
- TCGA
- The Cancer Genome Atlas
The authors have no conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are publicly available.
How to cite this article: He L, Li L, Chen Z. Keratin 19 is a potential diagnostic and prognostic biomarker in pancreatic adenocarcinoma. Medicine 2025;104:48(e46233).
Contributor Information
Ling He, Email: 1570547184@qq.com.
Li Li, Email: 443157498@qq.com.
References
- [1].Mizrahi JD, Surana R, Valle JW, Shroff RT. Pancreatic cancer. Lancet. 2020;395:2008–20. [DOI] [PubMed] [Google Scholar]
- [2].Wang Q, Sabanovic B, Awada A, et al. Single-cell omics: a new perspective for early detection of pancreatic cancer? Eur J Cancer. 2023;190:112940. [DOI] [PubMed] [Google Scholar]
- [3].Zheng RS, Zhang SW, Sun KX, et al. Cancer statistics in China, 2016. Zhonghua Zhong Liu Za Zhi. 2023;45:212–20. [DOI] [PubMed] [Google Scholar]
- [4].Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021. CA Cancer J Clin. 2021;71:7–33. [DOI] [PubMed] [Google Scholar]
- [5].Ju JH, Yang W, Lee KM, et al. Regulation of cell proliferation and migration by keratin19-induced nuclear import of early growth response-1 in breast cancer cells. Clin Cancer Res. 2013;19:4335–46. [DOI] [PubMed] [Google Scholar]
- [6].Govaere O, Komuta M, Berkers J, et al. Keratin 19: a key role player in the invasion of human hepatocellular carcinomas. Gut. 2014;63:674–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Saha SK, Choi HY, Kim BW, et al. KRT19 directly interacts with beta-catenin/RAC1 complex to regulate NUMB-dependent NOTCH signaling pathway and breast cancer properties. Oncogene. 2017;36:332–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Xia W, Zhang L, Liu Y. Weighted gene co-expression network analysis reveals six hub genes involved in and tight junction function in pancreatic adenocarcinoma and their potential use in prognosis. Genet Test Mol Biomarkers. 2019;23:829–36. [DOI] [PubMed] [Google Scholar]
- [9].Li Y, Zhu YY, Dai GP, et al. Screening and validating the core biomarkers in patients with pancreatic ductal adenocarcinoma. Math Biosci Eng. 2019;17:910–27. [DOI] [PubMed] [Google Scholar]
- [10].Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16:284–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Hanzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinf. 2013;14:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Ou ZL, Luo Z, Lu YB. Long non-coding RNA HULC as a diagnostic and prognostic marker of pancreatic cancer. World J Gastroenterol. 2019;25:6728–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Murphy JE, Wo JY, Ryan DP, et al. Total neoadjuvant therapy with FOLFIRINOX in combination with losartan followed by chemoradiotherapy for locally advanced pancreatic cancer: a phase 2 clinical trial. JAMA Oncol. 2019;5:1020–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Goonetilleke KS, Siriwardena AK. Systematic review of carbohydrate antigen (CA 19-9) as a biochemical marker in the diagnosis of pancreatic cancer. Eur J Surg Oncol. 2007;33:266–70. [DOI] [PubMed] [Google Scholar]
- [16].Auwercx J, Neve B, Vanlaeys A, et al. The kinase domain of TRPM7 interacts with PAK1 and regulates pancreatic cancer cell epithelial-to-mesenchymal transition. Cell Death Dis. 2025;16:335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Weitz J, Garg B, Martsinkovskiy A, Patel S, Tiriac H, Lowy AM. Pancreatic ductal adenocarcinoma induces neural injury that promotes a transcriptomic and functional repair signature by peripheral neuroglia. Oncogene. 2023;42:2536–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Pokrass MJ, Liu MF, Lindorfer MA, Taylor RP. Activation of complement by monoclonal antibodies that target cell-associated beta(2)-microglobulin: implications for cancer immunotherapy. Mol Immunol. 2013;56:549–60. [DOI] [PubMed] [Google Scholar]
- [19].Kim JM, Ashkenazi A. Fcgamma receptors enable anticancer action of proapoptotic and immune-modulatory antibodies. J Exp Med. 2013;210:1647–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Picarda E, Ohaegbulam KC, Zang X. Molecular pathways: targeting B7-H3 (CD276) for human cancer immunotherapy. Clin Cancer Res. 2016;22:3425–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Dai S, Zeng H, Liu Z, et al. Intratumoral CXCL13(+)CD8(+)T cell infiltration determines poor clinical outcomes and immunoevasive contexture in patients with clear cell renal cell carcinoma. J ImmunoTher Cancer. 2021;9:e001823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Li B, Zhu X, Sun L, et al. Induction of a specific CD8+ T-cell response to cancer/testis antigens by demethylating pre-treatment against osteosarcoma. Oncotarget. 2014;5:10791–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Mantovani A, Sozzani S, Locati M, Allavena P, Sica A. Macrophage polarization: tumor-associated macrophages as a paradigm for polarized M2 mononuclear phagocytes. Trends Immunol. 2002;23:549–55. [DOI] [PubMed] [Google Scholar]
- [24].Ramesh P, Shivde R, Jaishankar D, Saleiro D, Le Poole IC. A palette of cytokines to measure anti-tumor efficacy of T cell-based therapeutics. Cancers (Basel). 2021;13:821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Mao X, Tang X, Pan H, et al. B cells and IL-21-producing follicular helper T cells cooperate to determine the dynamic alterations of premetastatic tumor draining lymph nodes of breast cancer. Research (Wash D C). 2024;7:346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Gu-Trantien C, Loi S, Garaud S, et al. CD4(+) follicular helper T cell infiltration predicts breast cancer survival. J Clin Invest. 2013;123:2873–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Tuhin IJ, Zhu HJ, Monty MA, et al. From innate power to intelligent design: the evolution of NK cell-based cancer immunotherapy. Crit Rev Oncol Hematol. 2025;216:104972. [DOI] [PubMed] [Google Scholar]
- [28].Dehne N, Mora J, Namgaladze D, Weigert A, Brüne B. Cancer cell and macrophage cross-talk in the tumor microenvironment. Curr Opin Pharmacol. 2017;35:12–9. [DOI] [PubMed] [Google Scholar]
- [29].Yang Y, Yang Y, Yang J, Zhao X, Wei X. Tumor microenvironment in ovarian cancer: function and therapeutic strategy. Front Cell Dev Biol. 2020;8:758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].de Aquino MTP, Malhotra A, Mishra MK, Shanker A. Challenges and future perspectives of T cell immunotherapy in cancer. Immunol Lett. 2015;166:117–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Lindau D, Gielen P, Kroesen M, Wesseling P, Adema GJ. The immunosuppressive tumour network: myeloid-derived suppressor cells, regulatory T cells and natural killer T cells. Immunology. 2013;138:105–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Das M, Zhu C, Kuchroo VK. Tim-3 and its role in regulating anti-tumor immunity. Immunol Rev. 2017;276:97–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Jiang Z, Liu Z, Li M, Chen C, Wang X. Immunogenomics analysis reveals that TP53 mutations inhibit tumor immunity in gastric cancer. Transl Oncol. 2018;11:1171–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Lyu H, Li M, Jiang Z, Liu Z, Wang X. Correlate the TP53 mutation and the HRAS mutation with immune signatures in head and neck squamous cell cancer. Comput Struct Biotechnol J. 2019;17:1020–30. [DOI] [PMC free article] [PubMed] [Google Scholar]