中国P站

 TNNI3; Oncogene; papillary Renal Cell Carcinoma (pRCC); The Cancer Genome Atlas (TCGA)

CCK-8: Cell Counting Kit-8; ccRCC: clear cell Renal Cell Carcinoma; CTLs: Cytotoxic T Lymphocytes; cTnI: cardiac Troponin I; DCs: Dendritic cells; GO: Gene Ontology; ICB: Immune Checkpoint Blockade; KEGG: Kyoto Encyclopedia of Genes and Genomes; LASSO: Least Absolute Shrinkage and Selection Operator; OS: Overall Survival; pRCC: papillary Renal Cell Carcinoma; RCC: Renal Cell Carcinoma; ROC: Receiver Operating Characteristic; TCGA: The Cancer Genome Atlas; TIDE: Tumor Immune Dysfunction and Exclusion; TIMER: Tumor Immunity Estimation Resource.

TNNI3-encoded cardiac Troponin I (cTnI) is a key and wellcharacterized phosphoprotein that binds to actin-tropomyosin and regulates cardiac muscle contraction and ATPase activity of actomyosin. It is traditionally considered as a cytoplasm-localized sarcomeric thin filament protein and has received extensive attention as a serum biomarker for myocardial tissue damage. Recent studies have found that cTnI also exists in the nucleus. It plays a role both in the regulation of chromosome stability and cell polarity in Drosophila, as well as the directed differentiation of mammalian stem cells into cardiomyocytes, suggesting that cTnI may participate in new functions as a regulatory molecule. Although the function of intranuclear cTnI and its transport mechanism into nucleus are largely unknown, these studies suggest novel functions of cTnI besides serving as a contraction protein. Previous bioinformatics analysis found that cTnI may be involved in biological processes such as cell proliferation, immunity, fatty acid metabolism and mitochondrial membrane stability. The mutation of TNNI3 has been found to impede the expression of PG1-1α, ND5, LRPPRC and other genes associated with energy metabolism. Additionally, it hinders the activity of mitochondrial respiratory chain complex 1, ultimately leading to the disruption of mitochondrial structure and function. Given the importance of mitochondria in cancer research, it makes sense to examine the relationship between TNNI3 and tumor. Although aberrant expression of cTnI have been reported in several human cancer cells, the roles of cTnI in cancers are still largely unknown [1].

Renal Cell Carcinoma (RCC) has now upgraded to be a commonly diagnosed cancer type worldwide. There are nearly 380,000 diagnosed cases and over half died from it per year across the world. The second most common RCC, papillary RCC (pRCC), lacks effective therapeutic targets and prognostic molecular biomarkers compared to the most common pathological type, clear cell RCC (ccRCC). pRCC is a heterogeneous disease with two histological subtypes that present significant variations in disease progression and survival outcomes. Despite the identification of several gene mutations in pRCC tissues, the underlying molecular mechanisms are not well understood. At present, no consensus has been reached on the optimal risk gene signature for prognostic evaluation of patients with pRCC. Hence, there is an urgency to identify novel prognostic biomarkers and new therapeutic targets of pRCC to ameliorate the survival outcomes of its sufferers [2].

The Cancer Genome Atlas (TCGA), the largest publicly accessible cancer genomics database, profiles nearly 300 pRCC samples and provides a comprehensive genomic landscape with clinical annotations. In this study, we investigated the expression profiling of TNNI3 in 289 pRCC samples obtained from the TCGA database for a better understanding of TNNI3 in pRCC. Through our analyses, we identified three potential prognostic genes that are closely linked to TNNI3 and found their association with the Wnt signaling pathway. Additionally, our findings revealed that TNNI3 has the capability to enhance the proliferation and migration of Caki-2 RCC cells. Moreover, we conducted an assessment to determine the potential of TNNI3 as a predictive biomarker for immunotherapy in patients diagnosed with pRCC. Our analyses suggest that TNNI3 may act as a novel oncogene in pRCC and high TNNI3 expression might be an indicator of unfavorable prognosis in pRCC [3].

Datasets and Differentially expressed genes (DGgenes) screening

The detailed clinical information and transcriptomics of pRCC cohort were based on the TCGA KIRP dataset. The mRNA expression profiles (HTSeq-Counts) of 289 KIRP tissues and 32 non-tumor tissues were downloaded via the TCGA data portal (https:// portal.gdc.cancer.gov/). Genes were annotated by human gene annotation files (GRCh38.p12). The expression of TNNI3 was analyzed first in tumor and non-tumor tissues. Tumor tissues were then divided into TNNI3-low group and TNNI3-high group. Differential expression analysis of genes was performed using R/Bioconductor package DESeq2. Adjusted p-value less than 0.05 and the absolute value of log2 fold-change of gene expression (|log2 (Fold-Change) |) greater than the mean value of normalized counts (mean (abs (log (Fold-Change))) + 2 × sd (abs (log (Fold-Change))), were applied for primary filtering of DEgenes. Correlation analysis was conducted between TNNI3 and the DEgenes. Correlated genes were selected out when a p-value of 0.05 was set as the cutoff. To validate the prognostic risk score model based on TCGA dataset, the independent cohort GSE2748 from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) was employed [4].

Characterization of genes associated with Overall Survival (OS)

All the 289 KIRP patients were put into the training set. Univariate Cox analysis and log-rank test were used to examine the association between the expression levels of the 361 positively correlated genes and OS. Intersections of genes with p-value<0.05 in both analyses were selected for further analysis. The Least Absolute Shrinkage and Selection Operator (LASSO) analysis was then performed using glmnet package in R-software to find out vital genes from the prognostic genes.

Establishment of prognostic risk score formula

Multivariable Cox regression analysis was performed for establishing a prognostic risk score formula. The random forest plot was developed using the R package survminer. Risk scores of each included patient were calculated via the formula as mentioned above. Afterwards, all patients were classified into high-risk and low-risk groups by setting the median risk score as the cutoff value [5].

Assessment of the prognostic risk score model

Kaplan-Meier survival curve was used to compare the prognosis between the low-risk and high-risk groups. Additionally, a timedependent Receiver Operating Characteristic (ROC) curve, based on the risk scores for 3 and 5-year OS probability, was employed to assess the diagnostic accuracy. A p-value<0.05 indicates statistical significance. Meanwhile, concordance probability (C-index) was calculated using the R package survcomp to assess the value of prognostic risk formula. Then risk heatmap was applied to plot the expression profiles of key genes in the low-risk and high-risk groups.

Functional prediction

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was performed to annotate the potential functions of DEgenes using the clusterProfiler package in the R software.

Analysis of immune infiltrating and immunotherapy response

We utilized the TIMER2.0 database to assess the correlation between TNNI3 mRNA expression and infiltrating immune cells, encompassing B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages and myeloid Dendritic Cells (DCs) in the TCGA KIRP dataset. Subsequently, the expression profiles of immune checkpointrelated genes, specifically SIGLEC15, TIGIT, CD274, HAVCR2, PDCD1, CTLA4, LAG3, and PDCD1LG2, were examined in the TNNI3 high and low-expression groups. Ultimately, the potential response to Immune Checkpoint Blockade (ICB) was predicted using the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm [6].

Tissue microarray and immunohistochemistry

A pRCC tissue microarray was purchased from Shanghai Outdo Biotech Co. Ltd. (Shanghai, China). The immunohistochemical protocols were performed in accordance with a conventional protocol. An anti-TNNI3 antibody (Invitrogen, USA) was used as the primary antibody in the immunohistochemical analysis.

Cell culturing and transfection

The human RCC cell line Caki-2 were cultured in McCoy's 5A (modified) (Gibco, USA) supplemented with 100 U/ml of penicillin/ streptomycin and 10% Fetal Bovine Serum (FBS) at 37°C in an atmosphere containing 5% CO2 . Adenovirus transfections were performed as per previous description. Briefly, the cells were transfected with empty vector or adenovirus expressing TNNI3. Transfected cells were collected for further analyses after 48 h.

In vitro wound-healing assay

Transfected Caki-2 cells were removed using 0.5 mM EDTA and plated on 6-well plates at a density of 1 × 10⁶ cells/mL for overnight incubation. Wounds were made with a 10 μL pipette tip and images were taken immediately and at 24 h after wounding respectively. All assays were conducted in triplicate and repeated at least three times [7].

Cell proliferation assay

A Cell Counting Kit-8 (CCK-8) assay (K1018, APExBIO) was used for the analysis of cell proliferation. Caki2 cells were plated into 96-well plates at a density of 1.5 × 103 cells/well. Cell viability was detected at the indicated time points according to the manufacturer’s protocol. Briefly, 10 μL of CCK-8 solution was added to each well and incubated at 37°C for 2 h. The absorbance at 450 nm was then measured using a microplate reader.

Western blot assay

Western blot was carried out as per previous description. Proteins bound to the Polyvinylidene Fluoride (PDVF) membrane were analyzed using primary antibodies against LGR5 (Abcam, USA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Arigo, Taiwan), β-catenin (CST, USA) and DYKDDDK Tag (GenScript, China). Band intensity was quantified using a G: BOX imaging system (Syngene, UK).

Statistical analysis

All statistical analyses were carried out using R software or Statistical Package for the Social Sciences (SPSS) software (IBM, USA). p-values less than 0.05 indicate statistical significance.

TNNI3 is overexpressed in pRCC

Initially, we assessed the mRNA expression levels of TNNI3 in pRCC samples obtained from the TCGA database. Our analysis revealed a significant increase in TNNI3 transcripts in pRCC samples compared to their corresponding adjacent normal tissues (p<0.002, Figure 1A). Additionally, TNNI3 exhibited an upregulation trend when comparing tumor samples with unpaired normal samples (p<0.001, Figure 1B). To further corroborate these findings, we performed immunohistochemical staining for TNNI3 on pRCC tissue microarrays. The results demonstrated a significantly higher expression of TNNI3 protein in pRCC tissues compared to adjacent nontumor tissues (Figure 1C). Subsequently, tumor samples from TCGA were divided into TNNI3-low and TNNI3-high groups for screening DEgenes. A total of 835 DEgenes, with a cutoff of fold change > | abs (logFold-Change) + 2 × sd (abs (logFold-Change)) | and adjusted p<0.05, were gained using the R/Bioconductor package DESeq2. Among these DEgenes, 122 were up-regulated and 713 were down-regulated, among which 361 genes were correlated with TNNI3 (Figure 1D) [8].

Figure 1: (A-B) Comparison of TNNI3 mRNA expression between normal and tumor samples in TCGA pRCC dataset (KIRP). (C) Representative immunohistochemical staining of TNNI3 in matched normal (upper) and tumor (lower) samples (scale bar=200 nm). (D) Volcano plot of TNNI3 related DEgenes between TNNI3 low expressed and high-expressed pRCC groups.

Recognition of key genes correlated with OS

To assess the prognostic association between gene expression profiles and patient OS, we initially conducted univariate Cox regression analysis and log-rank test. Among the 361 genes associated with TNNI3, 59 and 45 genes were found to have a p-value<0.05 in the respective analyses. These genes were then intersected and subjected to the LASSO method, leading to the identification of 8 crucial genes (Figure 2A, B). Subsequently, through multivariable Cox regression analysis, PTPRH, LGR5 and DMRT3 were determined as integrated prognostic biomarkers for pRCC (Figure 2C). The expression of the three genes were significantly associated with OS (Figure 3A-C), and their correlation with TNNI3 are shown in Figure 3D-F [9].

Figure 2: (A) Wayne diagram indicates the overlap genes of two analyses. (B) LASSO Cox regression analysis shows the identification of the 8 potential prognostic risk signature genes. (C) Multivariable regression forest plot of 8 intersection genes.

Figure 3: (A-C) Kaplan-Meier survival curves of PTPRH, LGR5 and DMRT3. (D-F) Correlations between TNNI3 and the three prognostic genes.

Establishment and validation of prognostic risk score formula

A prognostic risk score formula was then developed based on PTPRH, LGR5 and DMRT3 expression profiles and their coefficients of regression. All patients were classed into high-risk and low-risk groups by applying the median risk score as the cutoff value. The distribution of risk scores, survival status of the included patients, as well as the expression profiles of the 3 genes are displayed in Figure 4A. Subsequently, the prognostic value of the above risk formula was assessed using Kaplan-Meier analysis. It was observed that the lowrisk group exhibited a significantly superior OS in comparison to the high-risk group (p<0.0001, Figure 4B). Furthermore, the prognostic capability of the risk formula was evaluated by means of timedependent ROC analysis. The calculated areas under the ROC curve for 3 and 5-year OS were determined to be 0.75 and 0.74, respectively, indicating a good predictive value for patient survival (Figure 4C). To validate our risk model, we conducted parallel analyses utilizing the GSE2748 dataset. A total of 28 patients diagnosed with pRCC were stratified into high and low-risk subgroups based on their respective risk score values. The analysis yielded expected outcomes, which demonstrated that individuals belonging to the high-risk subgroup exhibited significantly worse OS compared to those in the low-risk subgroup (p<0.05, Figure 4D).

Figure 4: (A) Distribution of risk scores, survival status and risk heatmap of the three prognostic genes (PTPRH, LGR5 and DMRT3). (B) Kaplan-Meier survival curve for the high- and low-risk groups of the training set. (C) Time-dependent ROC curve analysis of the prognostic model. (D) Kaplan-Meier survival curve for the high and low-risk groups of the validation set.

TNNI3 is associated with canonical Wnt signaling pathway

To search for candidate genes interacted with PTPRH, LGR5 and DMRT3, we further used the STRING (http://string-db.org) database to perform a gene-gene interaction analysis for each gene (Figure 5A). Target genes were then chosen for GO and KEGG analysis to explore the potential mechanisms involved in the progression of pRCC. Both GO and KEGG analysis revealed a significant enrichment of Wnt signaling pathway (Figure 5B, C). In addition, cell proliferation and metabolic process were also shown in the GO analysis, while in the KEGG analysis, most of the processes were related to cancers.

Figure 5: (A) Genes interacted with PTPRH, LGR5 and DMRT3 studied by STRING. (B) The top 12 functional GO terms of the target genes. (C) The top 15 terms of the enriched KEGG pathway for target genes.

TNNI3 promotes Caki-2 cells proliferation and migration

Considering the involvement of Wnt signaling pathway in multiple processes of tumor progression, cell proliferation and migration assays were carried out for further exploration. CCK-8 assays were first performed to examine the effect of TNNI3 on cell proliferation. A significant increase in cell proliferation was observed after transfection with TNNI3 compared with control group (Figure 6A). Wound healing assays were then conducted on Caki-2 cells to explore the effect of TNNI3 on migration and the results Page 9/24 showed that TNNI3 over-expressing cells closed the wound area faster than the control cells (Figure 6B). These findings indicated that TNNI3 could promote proliferation and migration of Caki-2 cells.

TNNI3 activates Wnt signaling pathway by up-regulating LGR5 and β-catenin expressions

To determine whether TNNI3 promoted the proliferation and migration of Caki-2 cells through the Wnt signaling pathway, western blot assays were conducted. The results showed that the expression of LGR5, the receptor responsible for triggering the Wnt signal cascade and β-catenin, the core component of Wnt pathway were significantly higher in the TNNI3 overexpression group compared to the control group (Figure 6C). Taken together, these results suggested that TNNI3 may enhance the proliferation and migration of Caki-2 cells by activating Wnt signaling pathway.

Figure 6: (A) Cell proliferation was assessed using CCK8 assay. (B) Cell migration was assessed using wound healing assay. (C) Western blot analysis of LGR5, β-catenin and flag. GAPDH was used as a loading control.

TNNI3 may act as an immunotherapy predictor in pRCC

Immunotherapy, an emerging modality for cancer treatment, has been recommended as the initial treatment for ccRCC. However, its approval for pRCC is currently lacking due to the absence of reliable validity assessment. Therefore, in order to assess the potential of TNNI3 as a predictor of immunotherapy response in pRCC, we first conducted an analysis using the Tumor Immunity Estimation Resource (TIMER) to examine the correlations between TNNI3 expression and immune score. The results demonstrated a positive correlation between TNNI3 expression and the infiltration of various immune cells, including B cells, CD4+ T cells, CD8+ T cells, neutrophils and myeloid dendritic cells (Figure 7). Immune checkpoint molecules exert an inhibitory role in the immune system and serves as targets for ICB therapy. Subsequently, the expression levels of immune checkpoint-related genes were compared between the TNNI3 high and low-expression groups. The results indicated a decrease in the expression levels of CD274 (PD-L1), HAVCR2 (TIM-3), PDCD1LG2 (PD-L2) and TIGHT in the TNNI3 high-expression group (Figure 8A). Additionally, a correlation analysis between TNNI3 and these immunotherapy targets provided further support for these findings (Figure 8B). Finally, the TIDE algorithm was employed to assess the predictive significance of TNNI3 expression in the context of ICB therapy. Remarkably, the TNNI3 high-expression cohort demonstrated elevated TIDE scores, which are indicative of suboptimal response to ICB therapy (Figure 8C). These findings propose the potential utility of TNNI3 as a predictive biomarker for immunotherapy in patients with pRCC.

Figure 7: Relationship between TNNI3 expression and six types of immune cell infiltration in pRCC.

Figure 8: (A) The expression distribution of immune checkpoints genes in TNNI3 high and low-expression groups. (B) Correlations between TNNI3 and immune checkpoint-related genes. (C) The distribution scores of ICB therapy response in TNNI3 high and lowexpression groups.

TNNI3-encoded cTnI is one of the structural proteins of the myocardial contractile apparatus. It has been well demonstrated to be one of the most specific and sensitive biomarkers for myocardial injury. Therefore, it can serve as a valid diagnostic tool for clinical screening, early identification, timely assessment and monitoring of cardiotoxicity induced by chemo, radio, targeted, or immuno-therapy for cancers. Other than that, the relationship between cancer and cTnI is rarely reported. Although cTnI has been reported to present abnormal expression in several cancer cells, its underlying roles in cancers are poorly understood. We have innovatively demonstrated that cTnI could transport into nucleus in adult mouse and human fetal hearts. Furthermore, we have observed an association between cTnI and the regulation of energy metabolism-related genes such as PDEs, PGC-1a, ND5 and ATP2A2. These results provided new directions for us to understand the relationship between TNNI3 and cancer. As the second most common kidney cancer following ccRCC, pRCC carries a high mortality rate and an unfavorable prognosis, particularly in cases of the more aggressive type II pRCC. Patients with pRCC, however, are often excluded from molecular investigation due to limited number of cases. Therefore, there is an urgency to identify effective diagnostic and prognostic biomarkers for improving survival outcomes of patients with pRCC. This study aimed to address this issue by examining the TCGA datasets, where the expression of TNNI3 in pRCC samples was revealed a significant increase compared to normal samples. Initially, a total of 835 DEgenes were selected from TNNI3-low and TNNI3-high pRCC patients. Among these DEgenes, 361 genes exhibited a positive or negative correlation with TNNI3 expression. Subsequently, employing univariate Cox regression analysis, log-rank test, and LASSO analysis, eight genes, namely DMRT3, PTPRH, RYR2, ACTG2, PLA2G5, SLC7A11, LGR5 and F2RL2 were identified as being associated with survival outcomes. Further, PTPRH, LGR5, and DMRT3 were selected as signature genes through multivariable Cox regression analysis. Although TNNI3 was not directly correlated with OS in pRCC, we demonstrated the expression of the three TNNI3-related genes were closely associated with survival outcomes in pRCC. Based on this, a prognostic risk score formula was constructed. Patients with high-risk signatures had shorter OS than those with low-risk signatures in training and validation sets, respectively. Collectively, these findings confirmed the predictive value of PTPRH, LGR5 and DMRT3 for assessing the prognosis for patients with pRCC.

To obtain a better understanding of the molecular mechanisms underlying the three TNNI3-related genes, we utilized STRING to construct gene-gene interaction networks for PTPRH, LGR5 and DMRT3, respectively. This resulted in a total of 31 genes, which were subsequently analyzed using GO and KEGG methods. The results of the GO analysis revealed that the three genes were predominantly linked to the canonical Wnt signaling pathway, which was further supported by the KEGG pathway analysis. Additionally, the KEGG analysis suggested that these genes may be implicated in the development of several types of cancer, including breast cancer, gastric cancer, endometrial cancer and prostate cancer. The Wnt protein family has been observed to exert a wide range of effects on cellular processes, including organogenesis, stem cell regeneration and cell survival. It is noteworthy that the activation of Wnt signaling is essential for tumor survival and progression, even in the presence of anti-neoplastic agents. To date, the involvement of the Wnt signaling pathway has been reported in various cancers, such as colorectal cancer, endometrial cancer, breast cancer and kidney cancer. These findings provide substantial support for the GO and KEGG results obtained in this study.

The activation of Wnt pathway signaling is contingent upon the accumulation of cytoplasmic β-catenin, which will later translocate into nucleus and bind to the TCF/LEF transcription factor family to initiate transcription of downstream target genes. On the other hand, the Wnt target gene LGR5 is recognized as a stem cell marker in intestinal epithelia, hair follicles and hepatocytes. Notably, LGR5 has been found to be upregulated in human colorectal cancer, basal cell carcinoma, hepatocellular carcinoma and neuroblastoma. Moreover, LGR5 expression has been shown to be a predictor of unfavorable prognosis in patients with colorectal cancer. The present study observed an upregulation of βcatenin and LGR5 in Caki-2 cells overexpressing TNNI3, indicating an activation of the Wnt signaling pathway. Furthermore, TNNI3 was found to promote cell proliferation and migration, as evidenced by CCK-8 and wound healing assays. These findings suggest that TNNI3 may serve as a detrimental gene in the context of pRCC. The management of advanced or metastatic pRCC has presented a persistent difficulty in the field. Several targeted agents have been employed in the treatment of pRCC; however, their outcomes have proven unsatisfactory when compared to ccRCC. Presently, immunotherapy emerges as a potential solution to address this predicament. The infiltration of immune cells serves as a prerequisite for the successful implementation of immunotherapy. To evaluate the abundance of tumor-infiltrating immune cells, we utilized the tissue-specific TIMER database in pRCC samples exhibiting varying levels of TNNI3 expression and the results showed that immune cell infiltration was higher in the TNNI3 high-expression group. However, the efficacy of immunotherapy depends on the stimulation of immune activity. Immune checkpoint molecules, which act as inhibitors of immune cells, impeding the body's ability to mount a potent anti-tumor immune response and facilitating the tumor's evasion of immune surveillance. The correlation between immune checkpoint-related gene expression and the clinical benefit of ICB therapy has been documented. On the other hand, the TIDE algorithm employs a collection of gene expression markers to evaluate two distinct mechanisms of tumor immune evasion, namely the impairment of tumor-infiltrating Cytotoxic T Lymphocytes (CTLs) and the rejection of CTLs by immunosuppressive factors. Elevated TIDE scores are indicative of reduced effectiveness of ICB therapy and decreased survival rates. The findings of our analyses indicate that the TNNI3 low-expression group exhibited a reduced presence of infiltrating immune cells, yet demonstrated elevated expression of immune checkpoint-associated genes and lower TIDE scores. These observations suggest that the TNNI3 low-expression group may exhibit a more favorable response to ICB therapy. Consequently, these results may serve as a valuable reference for the selection of participants in future clinical trials investigating immunotherapy for pRCC.

Collectively, our findings indicate that the expression of the cardiac-specific gene TNNI3 is elevated in individuals diagnosed with pRCC. Additionally, we developed a gene signature (PTPRH, LGR5, and DMRT3) related to TNNI3 that exhibited a significant correlation with OS in pRCC patients. GO and KEGG analysis identified that the three genes are linked to the canonical Wnt signaling pathway. Moreover, our cellular experiments confirmed that TNNI3 can activate the Wnt pathway by upregulating LRG5 and βcatenin, consequently facilitating cell proliferation and migration in Caki-2 cells. Finally, analyses of immune-related functions demonstrated a positive correlation between heightened expression of TNNI3 in pRCC and a diminished response to immunotherapy. These findings suggest that TNNI3 may function as an oncogene in pRCC, thereby filling part of the knowledge gap of the association between TNNI3 and cancer. While additional investigations are necessary, this study has provided valuable insights and has paved the way for future research endeavors exploring the involvement of TNNI3 in other types of cancer.

The pRCC tissue microarray was purchased from Shanghai Outdo Biotech Co., Ltd. (Shanghai, China) and obtained the ethical committee approval.

Not applicable.

The datasets analyzed in the current study were downloaded from the following public domain resources: https://portal.gdc.cancer.gov/ repository and https://www.ncbi.nlm.nih.gov/geo/.

The authors declare that they have no competing interests.

This study was supported by research grants from National Natural Science Foundation of China (Grant No. 82200263).

WAZ and WQH designed the study. BC conducted the experiments. XXC performed the statistical analyses. BC and LY wrote the manuscript. All authors read and approved the final manuscript.

We thank Dr. Jianming Zeng (University of Macau) and all the members of his bioinformatics team, for generously sharing their experience and codes.

1. Tobacman LS (1996) . Annu Rev Physiol 58: 447-481.

   [] [] []

2. Apple FS (2002) Cardiac troponin: Redefining the detection of myocardial infarction. Am Clin Lab 21: 32-34.

   [] []

3. Bodor GS, Porter S, Landt Y, Ladenson JH (1992) Development of monoclonal antibodies for an assay of cardiac troponin-I and preliminary results in suspected cases of myocardial infarction. Clin Chem 38: 2203-2214.

   [] []

4. Reichlin T, Hochholzer W, Bassetti S, Steuer S, Stelzig C, et al. (2009) . N Engl J Med 361: 858-867.

   [] [] []

5. Kehat I, Kenyagin-Karsenti D, Snir M, Segev H, Amit M, et al. (2001) . J Clin Invest 108: 407-414.

   [] [] []

6. Asumda FZ, Chase PB (2012) . Differentiation 83: 106-115.

   [] [] []

7. Sahota VK, Grau BF, Mansilla A, Ferrus A (2009) . J Cell Sci 122: 2623-2631.

   [] [] []

8. Moksnes MR, Rosjo H, Richmond A, Lyngbakken MN, Graham SE, et al. (2021) . Hum Mol Genet 30: 2027-2039.

   [] [] []

9. Lu Q, Pan B, Bai H, Zhao W, Liu L, et al. (2022) . Genes Dis 9: 1689-1700.

   [] [] []