Abstract
Renal transplantation is the only efficacious treatment for end‐stage kidney disease. However, some people have developed renal insufficiency after transplantation, the mechanisms of which have not been well clarified. Previous studies have focused on patient factors, while the effect of gene expression in the donor kidney on post‐transplant renal function has been less studied. Donor kidney clinical data and mRNA expression status were extracted from the GEO database (GSE147451). Weight gene co‐expression network analysis (WGCNA) and differential gene enrichment analysis were performed. For external validation, we collected data from 122 patients who accepted renal transplantation at several hospitals and measured the level of target genes by qPCR. This study included 192 patients from the GEO data set, and 13 co‐expressed genes were confirmed by WGCNA and differential gene enrichment analysis. Then, the PPI network contained 17 edges as well as 12 nodes, and four central genes (PRKDC, RFC5, RFC3 and RBM14) were identified. We found by collecting data from 122 patients who underwent renal transplantation in several hospitals and by multivariate logistic regression that acute graft‐versus‐host disease postoperative infection, PRKDC [Hazard Ratio (HR) = 4.44; 95% CI = [1.60, 13.68]; p = 0.006] mRNA level correlated with the renal function after transplantation. The prediction model constructed had good predictive accuracy (C‐index = 0.886). Elevated levels of donor kidney PRKDC are associated with renal dysfunction after transplantation. The prediction model of renal function status for post‐transplant recipients based on PRKDC has good predictive accuracy and clinical application.
Keywords: bioinformatic analysis, external validation, kidney transplantation, PRKDC, renal dysfunction
1. INTRODUCTION
Renal transplantation is a beneficial therapy for many patients with renal failure as well as some chronic renal diseases. However, in the early stages of transplantation, delayed graft function (DGF), or malfunction of a transplanted kidney, is a severe complication. DGF is a severe dysfunction that can give rise to a prolonged hospital stay, a high risk of acute and chronic acute graft‐versus‐host disease (GVHD), and even graft failure. 1 DGF occurs primarily in a lack of donors and is influenced by a variety of recipients 2 and donor factors, 3 as well as cold ischemic time. 4 The incidence of DGF varies considerably between countries and may be related to ethnicity, with incidence rates ranging from 21% to 70%. 5 The most accepted definition of DGF is at least one dialysis session in the first week after transplantation. 6 Even though a few risk factors for DGF have been identified through recent studies, the pathogenesis of the disease is not fully understood.
The PRKDC gene is located on chromosome 8q11. The PRKDC gene encodes the DNA‐dependent protein kinase catalytic subunit (DNA‐PKcs), a member of the phosphatidylinositol 3‐kinase‐related kinase family, a serine/threonine protein phosphorylation kinase. 7 The mutation rates of PRKDC and the expression levels of DNA‐PKcs vary significantly in various tumours. Chen et al. performed a statistical analysis of the mutation rates of PRKDC reported in The Cancer Genome Atlas and Chinese population databases; the authors found that PRKDC had high mutation rates in several tumours, including gastric, colorectal and endometrial cancers, with a high correlation with microsatellite instability‐high correlation with microsatellite instability. 8 However, the role of PRKDC in the donor kidney and its effect after transplantation has not been reported.
With the technological advances in gene high‐throughput technology and proteomics, combined with bioinformatics approaches, access to validated biomarkers for diagnosis and patient survival has been facilitated. Bioinformatics‐based multi‐gene co‐expression analysis is critical. WGCNA is a multigene analysis method first proposed in 2008. 9 Instead of individual genes or isolated biomarkers, WGCNA focus on gene co‐expression modules. It associates them with specific features, improving the efficacy of identifying potentially valuable bioregulatory targets. 9 Gene differential expression analysis is an effective way to find the driver genes of a disease. 10 In summary, the simultaneous use of differential gene expression analysis and WGCNA enables the search for potential biomarkers that affect the functional status of specific organs.
In this project, we obtained high‐throughput data from the GEO database for 192 kidney transplant donor kidneys. We used differential gene expression analysis and WGCNA to search for potential genes that affect post‐transplant kidney function. In addition, we used differential gene enrichment analysis and PPI networks to screen out central genes. We proceeded to examine the potential role of PRKDC in influencing post‐transplant renal function through comprehensive bioinformatics analysis and validated the results with 122 patients in our medical centre. We performed a logistic regression analysis based on the screened target genes to screen for high‐risk factors and constructed a predictive model for post‐transplant renal functional status based on these genes for clinical application.
2. METHODS
2.1. Data filtering
Clinical data and gene data related to kidney transplant donor kidneys were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/gds). Data filtering was performed using a function ‘rpmk’ in the edgeR package. 11 After searching the database, this study used the R function ‘GEOquery’ to download a collection of clinical data such as renal functional status after kidney transplantation (GSE147451) 12 for a total of 192 patients.
2.2. WGCNA analysis
To increase the accuracy of the joint analysis of multiple genes, genes used for WGCNA were filtered. We used the R function ‘WGCNA’ to perform multiplex analysis between gene expression data profiles of GEO, grouping genes that are co‐highly expressed into different modules. 13 ‘PickSoftThreshold’ was used to build a scale‐free network. After performing Pearson correlation analysis, we made a similarity matrix. We then correlated the previously filtered modules with their corresponding clinical features to identify the needed functional modules.
2.3. Differentially expressed genes (DEG) analysis
To clarify donor DEGs between different renal functional states in patients after renal transplantation, we applied the R function ‘limma’ in the database. the criteria for screening DEGs were adj. p < 0.05 and |logFC| ≥ 1.0. the R packages VennDiagram and ggplot2 were applied to plot Venn and Volcano plots. 14
2.4. Pathway and function enrichment analysis
We enriched specific genes for specific functional parts or signalling pathways to clarify the function of GEO differential genes. Genome Encyclopedia (KEGG) and Gene Ontology (GO) pathway analysis were performed by the cluster profile R package. 15
2.5. Identification of hub genes and construction of PPI and
We have constructed the PPI network through the STRING database (https://string‐db.org/). The network was analysed and established with Cytoscape software (v 3.8.1) using ≥0.4 points as the extraction value. In the network, edges represent protein interactions, and nodes represent proteins. To specify the hub genes in the network, we used the maximum cluster centrality algorithm calculated by CytoHubba's plug‐in. The top four genes were set as hub genes.
2.6. Analysis of the expression levels of hub genes
To further verify the accuracy of the screened genes, we analysed the expression of these critical genes in different renal functional states of patients after renal transplantation. A bar graph represented the expression of each gene in other renal functional forms. The statistical significance between the two groups was determined by Student's t‐test subject to normality (p < 0.05).
2.7. RT‐qPCR analysis
RT‐qPCR was performed using QuantStudio 3 platform with SYBR Green. qPCR was started at 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of amplification at 95°C for 15 s and 60°C for the 60 s. Reverse transcription was performed at 25°C for 10 min, 42°C for 15 min, and 85°C for 5 min. mRNA levels were standardized with GAPDH as an internal control and determined according to the 2(−ΔΔCt) method. The sequences used were as follows: RPL35A, forward, 5′‐TTGAAGGTGTTTACGCCCGAG‐3′ and reverse, 5′‐TGCTTCGGAATTTGGCACGA‐3′; UBQLN4, forward, 5′‐ATTCGGGTCACCGTCAAGAC‐3′ and reverse, 5′‐GCCTTAAACCTCCGGGAGATTT‐3′; GAPDH, forward, 5′‐ GAGAAGGCTGGGGCTCATTT‐3′ and reverse, 5′‐ ATGACGAACATGGGGGCATC‐3′.
2.8. Patients
This study included data from 122 kidney transplant patients collected from the Ninth People's Hospital of Suzhou and the First Affiliated Hospital of Soochow University between 1 February 2018, and 31 January 2021, as external validation. We collected the data on 31 February 2021. The median age was 44 (IQR 24–52), and the project included 74 (61.0%) men. Patient data, including gender, age, body mass index, primary disease, chronic disease history, dialysis method, immunosuppressant use, aGVHD, infection, preoperative and postoperative creatinine, Ccr and PRKDC expression levels were selected for analysis. Glomerular filtration rate ≤ 50 mL/min was regarded as poor renal function group. Informed consent was obtained from all patients' immediate family members or patients' own. All research projects conformed to the guidelines of the Soochow University Ethics Committee and followed the Declaration of Helsinki (ID202214523). The subjects included in the external validation portion of this study were subjected to the same conditions as the cases in the GSE147451 data set.
2.9. Statistical analysis
Missing values, accounting for ≤5.0% of the data, were estimated using the ‘mice’ package 16 in RStudio (R version 3.5.4) 17 in RStudio, version 3.5.4. The random forest method was utilized for this estimation. Categorical variables were analysed using proportions and χ2 tests. Skewed variables were expressed as medians with interquartile ranges. Kruskal‐Wallis tests or one‐way anova were used for comparisons between groups. COX regression models were employed for both univariate and multivariate post‐transplant renal functional status analyses, which were adjusted accordingly. COX regression we utilized is consistent with the pH hypothesis. A multifactorial regression model was utilized to confirm risk factors and predict renal functional status after kidney transplantation. To ensure the validity of our analysis, we included significant variables from the multifactorial regression in the nomogram. The weights of each variant were confirmed and nomograms were generated and internally validated via a 1001‐bootstrap method18. The C statistic was calculated using the ‘rms’ package in R software. Statistical analyses were performed using several R packages, namely ‘PredictABEL’, ‘rms’, ‘risk regression’, ‘survminer’, and ‘ggplot2’19. A significance level of p < 0.05 was used throughout the analysis. 18
2.10. Patient and public involvement
The study was a comprehensive study based on a public database and local hospital cases, and the project evaluated renal reconstruction and function after kidney transplantation. The public—that is a statutory health insurer, physicians, local medical managers and patient representatives—was involved in the design and implementation of the entire project. In addition, practitioners (physicians, local health care workers, health care administrators) and scientific experts were involved in the discussions of this study. The results of the study and the project as a whole will be disseminated to participants through practise‐oriented publications and newsletters.
3. RESULTS
The flow chart for this project is shown in Figure 1. We performed a WGCNA analysis from the GEO database (GSE147451) to identify the critical genomes of donor kidneys that affect renal function after kidney transplantation. Each colour represents a module. Here, 13 modules in the data set (Figure 2A) were confirmed. The results of the analysis do not indicate any significant outliers. Figure 2B shows module–trait relationships, indicating that the turquoise module (Figure 2C) and the lightcyan module (Figure 2D) were highly discriminatory between different post‐transplant renal functional states (turquoise module: cor = 0.04, p = 0.011, lightcyan module: cor = 0.17, p = 0.069). Other modules are displayed in the supplemental materials.
FIGURE 1.
Project design and flow chart of this study.
FIGURE 2.
Identifying modules related to clinical information in the GEO data set (GSE147451). Cluster dendrograms of co‐expression network modules are sorted by hierarchical clustering of genes GEO data set (A). Each module was assigned a different colour. Module–trait relationships in the GEO data set (B). Each row corresponds to a coloured module, and each column corresponds to a clinical trait (normal and abnormal renal function after transplantation): Midnight blue module (C) and the turquoise module (D) corresponding correlations and p‐values.
We have found that 2211 DEGs in the GEO data set (Figure 3A) were selected via the ‘limma’ package, using adj. p < 0.05 and |logFC| ≥ 1.0 as the cut‐off criteria. We decided on the top 50 genes based on expression differences and statistical significance to plot the heat map (Figure 3B). Sixty‐six genes were obtained from these modules according to WGCNA. Finally, 690 genes were obtained at the intersection of the DEGs and midnight‐blue modules involving genes obtained from the GEO database (Figure 3C).
FIGURE 3.
DEGs were identified in the kidney transplant donor data set of GSE147451 with logFC| ≥ 1.0 and adj. p < 0.05 as the cut‐off criteria. (A) Volcano map of DEGs in the GEO data set. (B) Heat map of the top 50 DEGs in the GSE147451 data set. (C) Venn diagram of genes between the DEG list and the co‐expression module.
To fully understand the functions of these genes in the co‐expression module, GO (Figure 4A) and KEGG (Figure 4B) analyses were performed on the differential genes in GEO using the R package ‘ClusterProfiler’. GO investigation classified genes into three categories: cellular components (CC), molecular functions (MF), and biological processes (BP). The genes involved in T‐cell activation and neutrophil activation were enriched in BP. The CC category involved several genes in the mitochondrial matrix and immunological synapse. In addition, the amide binding and ubiquitin‐like protein ligase binding were mainly engaged in MF. KEGG displayed that genes in Graft versus host disease were enriched.
FIGURE 4.
GO (A) and KEGG (B) enrichment analysis of genes in the GEO database. The colours represent adjusted p values, and the size of the spots represents the number of genes. Visualization of PPI networks and candidate hub genes. (C) PPI network of genes between the DEG list and the two co‐expression modules. Blue nodes represent genes. Edges indicate the interaction relationships between the nodes. (D) Identification of central genes in the PPI network using the MCC algorithm.
We analysed the STRING database PPI to investigate the relationship between co‐expression modules and DEGs. We plotted a PPI network with 320 edges and 219 nodes in Figure 4C. In addition, the top four hub genes with MCC scores in the PPI were calculated using the CytoHubba plugin. These genes are PRKDC, RFC5, RFC3 and RBM14 (Figure 4D). All four essential genes were highly expressed in patients with poor post‐transplant renal function (Figure 5).
FIGURE 5.
Histogram of the differential expression of PRKDC (A), RBM14 (B), RFC3 (C) and RFC5 (D) in different renal functional states after renal transplantation.
To further clarify the effect of the above genes on renal function after kidney transplantation, we collected clinical data from 122 kidney transplant patients at the medical centre. We examined the expression levels of the above genes. The clinical baseline data for these patients are shown in Table 1. Specifically, 122 patients who underwent kidney transplantation between 1 February 2018, and 31 January 2021, to the follow‐up endpoint (1 December 2021) were included in this study. Among them, the poorer renal function was 30 (GFR < 50 min/L). Among these patients, the median BMI was 21.52. 92 (75%) patients had the hypertensive disease. 66 (54%) patients received preoperative haemodialysis, while 48 (39%) patients received peritoneal dialysis. The overall mean level of PRKDC was 13.98 ± 7.88. aGVHD occurred in 39 (32%) patients and infection occurred in 58 (48%) patients.
TABLE 1.
Study participant characteristics at enrolment.
| Variables | Total (n = 122) | Glomerular filtration rate | ||
|---|---|---|---|---|
| >50 mL/min (n = 92) | ≤50 mL/min (n = 30) | p‐value | ||
| Age, median (Q1, Q3) | 44 (34, 52) | 40.5 (34, 51) | 47 (43.25, 52.75) | 0.027* |
| Gender, n (%) | 0.465 | |||
| Male | 74 (61) | 58 (63) | 16 (53) | |
| Female | 48 (39) | 34 (37) | 14 (47) | |
| Chronic disease, n (%) | 0.321 | |||
| None | 18 (15) | 13 (14) | 5 (17) | |
| Hypertension | 92 (75) | 72 (78) | 20 (67) | |
| Diabetes | 6 (5) | 4 (4) | 2 (7) | |
| Others | 6 (5) | 3 (3) | 3 (10) | |
| Primary disease, n (%) | 0.644 | |||
| Chronic renal failure | 78 (64) | 57 (62) | 21 (70) | |
| Chronic glomerulonephritis | 21 (17) | 16 (17) | 5 (17) | |
| Others | 23 (19) | 19 (21) | 4 (13) | |
| Dialysis method, n (%) | 0.944 | |||
| None | 8 (7) | 6 (7) | 2 (7) | |
| Haemodialysis | 66 (54) | 49 (53) | 17 (57) | |
| Peritoneal dialysis | 48 (39) | 37 (40) | 11 (37) | |
| BMI, median (Q1, Q3) | 21.52 (19.41, 24.03) | 21.4 (19.46, 24.07) | 21.76 (19.44, 23.79) | 0.912 |
| FK506 use, n (%) | 0.22 | |||
| No | 25 (20) | 16 (17) | 9 (30) | |
| Yes | 97 (80) | 76 (83) | 21 (70) | |
| MMF use, n (%) | 0.523 | |||
| No | 57 (47) | 45 (49) | 12 (40) | |
| Yes | 65 (53) | 47 (51) | 18 (60) | |
| CSA use, n (%) | 0.106 | |||
| No | 108 (89) | 84 (91) | 24 (80) | |
| Yes | 14 (11) | 8 (9) | 6 (20) | |
| ATG use, n (%) | 1 | |||
| No | 120 (98) | 90 (98) | 30 (100) | |
| Yes | 2 (2) | 2 (2) | 0 (0) | |
| Cr before surgery, median (Q1, Q3) (μmol/L) | 934.05 (790.95, 1180.45) | 976 (796.42, 1183.92) | 842 (747.65, 1061.2) | 0.248 |
| aGVHD, n (%) | <0.001*** | |||
| No | 83 (68) | 72 (78) | 11 (37) | |
| Yes | 39 (32) | 20 (22) | 19 (63) | |
| Infection, n (%) | 0.002** | |||
| No | 64 (52) | 56 (61) | 8 (27) | |
| Yes | 58 (48) | 36 (39) | 22 (73) | |
| Cr after surgery, median (Q1, Q3) (μmol/L) | 99.85 (87.25, 126.95) | 96.05 (77.3, 112.92) | 141 (108.4, 189.52) | <0.001*** |
| Ccr after surgery, median (Q1, Q3) (mL/min) | 63.15 (50.13, 80.43) | 69.61 (61.5, 87.16) | 40.3 (25.43, 45.2) | <0.001*** |
| PRKDC expression, mean ± SD | 13.98 ± 7.88 | 11.7 ± 6.13 | 20.98 ± 8.59 | <0.001*** |
Abbreviations: aGVHD, acute graft‐versus‐host disease; ATG, antithymocyte globulin; BMI, body mass index; Ccr, creatinine clearance rate; Cr, creatinine; CSA, Ciclosporin A; MMF, mycophenolate mofetil.
p < 0.05
p < 0.01
p < 0.001.
Subsequently, by univariate logistics regression, we included variables (p < 0.2) in the multifactorial analysis. It showed that aGVHD [HR = 5.66; 95% CI = [2.05, 16.73]; p < 0.001], infection [HR = 5.10; 95% CI = [1.85, 15.67]; p = 0.003] and PRKDC [HR = 4.44; 95% CI = [1.60, 13.68]; p < 0.006] were all independent risk factors for post‐transplant renal insufficiency (Table 2). Then, independent risk factors screened by the multifactorial analysis were selected to construct a predictive model of renal functional status after kidney transplantation using nomogram (Figure 6A). Nomogram predicted the risk of renal dysfunction with a C‐statistic of 0.886. A fitted calibration curve is in the internal validation in Figure 6B.
TABLE 2.
Cox regression analysis of hazard ratio on renal function after renal transplantation.
| Variables | Univariate analysis | Multivariate analysis | ||
|---|---|---|---|---|
| Hazard ratio (95% CI) | p‐value | Hazard ratio (95% CI) | p‐value | |
| Gender, female versus male | 1.49 [0.64, 3.44] | 0.346 | ‐ | ‐ |
| FK506 use, yes versus no | 0.49 [0.19, 1.30] | 0.142 | 1.14 [0.20, 8.03] | 0.885 |
| MMF use, yes versus no | 1.44 [0.63, 3.38] | 0.397 | ‐ | ‐ |
| CSA use, yes versus no | 2.62 [0.80, 8.30] | 0.1 | 3.26 [0.38, 33.81] | 0.295 |
| ATG use, yes versus no | ‐ | 0.989 | ‐ | ‐ |
| aGVHD, yes versus no | 6.22 [2.59, 15.64] | <0.001*** | 5.66 [2.05, 16.73] | 0.001** |
| Infection, yes versus no | 4.28 [1.78, 11.20] | 0.002** | 5.10 [1.85, 15.67] | 0.003** |
| Age, >60 versus ≤60 (year) | 0.76 [0.04, 5.39] | 0.808 | ‐ | ‐ |
| Chronic disease, hypertension versus none | 0.56 [0.23, 1.41] | 0.204 | ‐ | ‐ |
| Primary disease, chronic renal failure versus others | 1.43 [0.60, 3.62] | 0.427 | ‐ | ‐ |
| Dialysis method, haemodialysis versus others | 1.15 [0.50, 2.67] | 0.745 | ‐ | ‐ |
| BMI, >24 versus ≤24 | 0.97 [0.37, 2.41] | 0.957 | ‐ | ‐ |
| Cr before surgery, >934 versus ≤934 (μmol/L) | 0.49 [0.20, 1.12] | 0.096 | 0.58 [0.20, 1.60] | 0.293 |
| PRKDC expression, >12.81 versus ≤12.81 | 3.74 [1.56, 9.76] | 0.004** | 4.44 [1.60, 13.68] | 0.006** |
Abbreviations: aGVHD, acute graft‐versus‐host disease; ATG, antithymocyte globulin; BMI, body mass index; Ccr, creatinine clearance rate; Cr, creatinine; CSA, Ciclosporin A; MMF, mycophenolate mofetil.
p < 0.01
p < 0.001.
FIGURE 6.
Nomogram predicting the risk of bleeding renal insufficiency in patients after renal transplantation. (A) Nomogram for predicting renal insufficiency in patients. (B) Validity of calibration curves in estimating patient prognosis.
4. DISCUSSION
Renal transplantation is the therapy of choice for patients with renal failure. However, DGF is a severe post‐transplant complication. It can give rise to increased risk of acute and chronic rejection, prolonged hospitalization and graft failure. 19
In this study, we analysed 192 patients in the GEO data set. Based on a comprehensive bioinformatics analysis, we obtained 690 lists of DEGs, and modules crossed by overlapping genes from the GEO data set (GSE147451). GO and KEGG analysis were performed using the function “ClusterProfiler” of R software. Immediately after, the four genes with the highest MCC scores in PPI were screened, including PRKDC, RFC5, RFC3 and RBM14. These essential genes impact the renal functional status of patients after kidney transplantation. PRKDC, whose high expression was associated with post‐transplant renal insufficiency. In the validation phase, we collected clinical data and gene expression assays from 122 kidney transplant patients in our hospital. Four independent risk factors, aGVHD, infection and PRKDC, were screened by multifactorial analysis, and a prediction model was established focusing on these genes, which showed good prediction accuracy.
Recently, gene networks have been increasingly used for bioinformatics analysis. WGCNA is a efficacy method to analyse the expected effects of multiple genes among revealed genomic data. 20 WGCNA has been applied in various biological studies to confirm therapeutic targets or candidate biomarkers, especially in tumours. 21 , 22 , 23 However, the analysis of WGCNA in transplantation‐related, especially renal functional status after kidney transplantation, is currently less studied.
In the project, we focused on the hub gene PRKDC and combined it with the external validation of 122 patients in our medical centre. Both PRKDC and DNA‐PKcs affect tumorigenesis, progression, invasion and metastasis through different pathophysiological processes and further prognostic analyses have shown that DNA‐PKcs expression can help predict patient prognosis.Lu et al. 24 after studying the effect of DNA‐PKcs expression levels on clinical staging and lymphatic and distant metastasis of colorectal cancer, further demonstrated that DNA‐PKcs expression levels were negatively correlated with the 5‐year survival rate of patients. Xing et al. 25 analysed the gene expression profile of NSCLC cells and showed that patients with high DNA‐PKcs or ATM expression levels in the tumour sample/normal tissue sample (T/N) ratio had a significantly increased risk of death. In other cancers such as gastric, hepatocellular, breast and nasopharyngeal cancers, the expression level of DNA‐PKcs was also negatively correlated with prognosis. 26 , 27 , 28 , 29 However, there are no reports on the functional aspects of the post‐transplant kidney and our study expands on this part. And we may predict the prognosis of kidney transplantation recipients on PRKDC.
The present study has its limitations. The effect of PRKDC should be verified in vitro experiments. Then, studies on the mechanisms affecting renal insufficiency in recipients after renal transplantation should also be continued. Prediction maps also need to be confirmed using larger sample sizes. Moreover, a prospective investigation should be initiated before the clinical application of the prediction model. We are preparing to conduct further prospective clinical trials in our follow‐up studies to validate the findings of this study.
In conclusion, we identified four essential genes affecting post‐transplant renal function through comprehensive bioinformatics analysis. By differential gene expression analysis and WGCNA, PRKDC was closely associated with post‐transplant renal function. External validation demonstrated that PRKDC expression levels are an independent risk factor for predicting renal functional status in patients after renal transplantation. Furthermore, nomograms based on aGVHD, infection and PRKDC proved more effective in assessing kidney function after transplantation.
STRENGTHS AND LIMITATIONS OF THIS STUDY
This study found that infection and aGVHD are related to renal dysfunction after transplantation. Elevated levels of donor kidney PRKDC are associated with renal dysfunction after transplantation. The prediction model of renal function status for post‐transplant recipients based on PRKDC has good predictive accuracy and clinical application. The present study has its limitations. The effect of PRKDC should be verified in vitro experiments. Then, studies on the mechanisms affecting renal insufficiency in recipients after renal transplantation should also be continued. Prediction maps also need to be confirmed using larger sample sizes. Moreover, a prospective investigation should be initiated before the clinical application of the prediction model.
AUTHOR CONTRIBUTIONS
Zhijun Cao: Formal analysis (equal); investigation (equal); writing – original draft (equal); writing – review and editing (equal). Hao Jiang: Data curation (equal); investigation (equal); methodology (equal); visualization (equal). Chunchun Zhao: Resources (equal); software (equal). Huifeng Zhou: Data curation (equal); formal analysis (equal). Zheng Ma: Visualization (equal); writing – original draft (equal). Chen Xu: Formal analysis (equal); investigation (equal). Jianglei Zhang: Project administration (equal); supervision (equal). Minjun Jiang: Conceptualization (lead); data curation (lead); project administration (lead); writing – original draft (equal); writing – review and editing (equal). Zhenfan Wang: Data curation (equal); formal analysis (equal); methodology (equal).
FUNDING INFORMATION
This work was supported by the Natural Science Foundation of Jiangsu Province (Grants Number BK20210094), the Suzhou Gusu Medical Youth Talent (GSWS2021012), Suzhou Youth Science and Technology Program KJXW2022076 and the Jiangsu Innovative and Entrepreneurial Talent Programme(JSSCBS20211560).
CONFLICT OF INTEREST STATEMENT
The authors declare no competing financial interests.
Supporting information
Appendix S1: Supporting Information
Cao Z, Jiang H, Zhao C, et al. Up‐regulation of PRKDC was associated with poor renal dysfunction after renal transplantation: A multi‐centre analysis. J Cell Mol Med. 2023;27:1362‐1372. doi: 10.1111/jcmm.17737
Zhijun Cao, Hao Jiang, and Chunchun Zhao have contributed equally to this work and share the first authorship.
Contributor Information
Chen Xu, Email: dr_xuchensz@sina.com.
Jianglei Zhang, Email: zhangjianglei82@tom.com.
Minjun Jiang, Email: jiangminjun70@163.com.
Zhenfan Wang, Email: wangzhenfan162@163.com.
DATA AVAILABILITY STATEMENT
Data may be obtained from a third party and are not publicly available. The data are, in large part, owned by Soochow University. The GEO data could be download in the website as follows https://www.ncbi.nlm.nih.gov/gds.
REFERENCES
- 1. Matas AJ, Gillingham KJ, Elick BA, et al. Risk factors for prolonged hospitalization after kidney transplants. Clin Transplant. 1997;11(4):259‐264. [PubMed] [Google Scholar]
- 2. Doshi MD, Garg N, Reese PP, Parikh CR. Recipient risk factors associated with delayed graft function: a paired kidney analysis. Transplantation. 2011;91(6):666‐671. [DOI] [PubMed] [Google Scholar]
- 3. Humar A, Ramcharan T, Kandaswamy R, Gillingham K, Payne WD, Matas AJ. Risk factors for slow graft function after kidney transplants: a multivariate analysis. Clin Transplant. 2002;16(6):425‐429. [DOI] [PubMed] [Google Scholar]
- 4. Kayler LK, Magliocca J, Zendejas I, Srinivas TR, Schold JD. Impact of cold ischemia time on graft survival among ECD transplant recipients: a paired kidney analysis. Am J Transplant. 2011;11(12):2647‐2656. [DOI] [PubMed] [Google Scholar]
- 5. Hagenmeyer EG, Haussler B, Hempel E, et al. Resource use and treatment costs after kidney transplantation: impact of demographic factors, comorbidities, and complications. Transplantation. 2004;77(10):1545‐1550. [DOI] [PubMed] [Google Scholar]
- 6. Lim WH, Johnson DW, Teixeira‐Pinto A, Wong G. Association between duration of delayed graft function, acute rejection, and allograft outcome after deceased donor kidney transplantation. Transplantation. 2019;103(2):412‐419. [DOI] [PubMed] [Google Scholar]
- 7. Sibanda BL, Chirgadze DY, Ascher DB, Blundell TL. DNA‐PKcs structure suggests an allosteric mechanism modulating DNA double‐strand break repair. Science. 2017;355(6324):520‐524. [DOI] [PubMed] [Google Scholar]
- 8. Chen Y, Li Y, Guan Y, et al. Prevalence of PRKDC mutations and association with response to immune checkpoint inhibitors in solid tumors. Mol Oncol. 2020;14(9):2096‐2110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Yang R, Du Y, Wang L, Chen Z, Liu X. Weighted gene co‐expression network analysis identifies CCNA2 as a treatment target of prostate cancer through inhibiting cell cycle. J Cancer. 2020;11(5):1203‐1211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Liu W, Gan CY, Wang W, et al. Identification of lncRNA‐associated differential subnetworks in oesophageal squamous cell carcinoma by differential co‐expression analysis. J Cell Mol Med. 2020;24(8):4804‐4818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139‐140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Davis S, Meltzer PS. GEOquery: a bridge between the gene expression omnibus (GEO) and bioConductor. Bioinformatics. 2007;23(14):1846‐1847. [DOI] [PubMed] [Google Scholar]
- 13. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Chen H, Boutros PC. VennDiagram: a package for the generation of highly‐customizable Venn and Euler diagrams in R. BMC Bioinformatics. 2011;12:35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics. 2012;16(5):284‐287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Schlosser P, Knaus J, Schmutz M, et al. Netboost: boosting‐supported network analysis improves high‐dimensional omics prediction in acute myeloid leukemia and Huntington's disease. IEEE/ACM Trans Comput Biol Bioinform. 2021;18(6):2635‐2648. [DOI] [PubMed] [Google Scholar]
- 17. Kleinendorst RWD, Barzaghi G, Smith ML, Zaugg JB, Krebs AR. Genome‐wide quantification of transcription factor binding at single‐DNA‐molecule resolution using methyl‐transferase footprinting. Nat Protoc. 2021;16(12):5673‐5706. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Min SH, Zhou J. Smplot: an R package for easy and elegant data visualization. Front Genet. 2021;12:802894. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Schnuelle P, Gottmann U, Koppel H, et al. Comparison of early renal function parameters for the prediction of 5‐year graft survival after kidney transplantation. Nephrol Dial Transplant. 2007;22(1):235‐245. [DOI] [PubMed] [Google Scholar]
- 20. Zhang B, Horvath S. A general framework for weighted gene co‐expression network analysis. Stat Appl Genet Mol Biol. 2005;4. [DOI] [PubMed] [Google Scholar]
- 21. Tang J, Kong D, Cui Q, et al. Prognostic genes of breast cancer identified by gene co‐expression network analysis. Front Oncol. 2018;8:374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Wang D, Liu J, Liu S, Li W. Identification of crucial genes associated with immune cell infiltration in hepatocellular carcinoma by weighted gene co‐expression network analysis. Front Genet. 2020;11:342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Qiu X, Cheng SH, Xu F, Yin JW, Wang LY, Zhang XY. Weighted gene co‐expression network analysis identified MYL9 and CNN1 are associated with recurrence in colorectal cancer. J Cancer. 2020;11(8):2348‐2359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Lu Y, Zhang HL, Li YZ, Zhao P. Clinicopathological significance of expressions of DNA dependent protein kinase catalytic subunit and P16 in colorectal carcinoma. Zhonghua Yi Xue Za Zhi. 2008;88(29):2025‐2029. [PubMed] [Google Scholar]
- 25. Xing J, Wu X, Vaporciyan AA, Spitz MR, Gu J. Prognostic significance of ataxia‐telangiectasia mutated, DNA‐dependent protein kinase catalytic subunit, and Ku heterodimeric regulatory complex 86‐kD subunit expression in patients with nonsmall cell lung cancer. Cancer. 2008;112(12):2756‐2764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Zhang Y, Wen GM, Wu CA, et al. PRKDC is a prognostic marker for poor survival in gastric cancer patients and regulates DNA damage response. Pathol Res Pract. 2019;215(8):152509. [DOI] [PubMed] [Google Scholar]
- 27. Yan SS, Liu L, Liu ZG, Zeng MS, Song LB, Xia YF. Expression and clinical significance of DNA‐PKcs in nasopharyngeal carcinoma. Ai Zheng. 2008;27(9):979‐983. [PubMed] [Google Scholar]
- 28. Evert M, Frau M, Tomasi ML, et al. Deregulation of DNA‐dependent protein kinase catalytic subunit contributes to human hepatocarcinogenesis development and has a putative prognostic value. Br J Cancer. 2013;109(10):2654‐2664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Zhang Y, Yang WK, Wen GM, et al. High expression of PRKDC promotes breast cancer cell growth via p38 MAPK signaling and is associated with poor survival. Mol Genet Genomic Med. 2019;7(11):e908. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1: Supporting Information
Data Availability Statement
Data may be obtained from a third party and are not publicly available. The data are, in large part, owned by Soochow University. The GEO data could be download in the website as follows https://www.ncbi.nlm.nih.gov/gds.