PUS1 is a novel biomarker for evaluating malignancy of human renal cell carcinoma

Data collection

The gene expression, mRNA sequencing (mRNA) expression and clinical information of 532 kidney renal clear cell carcinoma samples were retrieved from the Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). The normal kidney composed of 72 TCGA samples and 89 GTEx database. GTEx datasets are available at the GTEx data portal (https://www.gtexportal.org/home/datasets) for the current release (V8). The four GEO clear cell renal cancer cohorts (GSE15641, GSE17818, GSE17895, and GSE40435) were obtained from the Gene-Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) as MINiML files. In GSE15641, 23 control samples and 32 clear cell renal samples were enrolled; in GSE17818, 13 normal patients and 103 renal clear cell carcinoma patients; in GSE17895, the data contained 22 health renal samples and 138 patients; and the GSE40435 dataset had included 101 health kidney and 101 kidney clear cell carcinoma.

Biological data processing

For 14 TCGA cancer types with over ten paired tumor and normal samples (THCA, KIRP, BLCA, LIHC, HNSC, BRCA, LUAD, PRAD, ESCA, KICH, LUSC, KIRC, STAD, COAD), we performed GSCA differential expression analysis using normalized and batch corrected RSEM mRNA expression. We calculated the fold change by dividing the mean of the tumor by the mean of the normal, and our p-value is estimated by a t-test and then adjusted by the Fisher’s exact test. The data from GEO contain the platform, samples and GSE records. The microarray data were normalized by the R (version 3.4.1) preprocess core package named “Normalize. Quantiles” following the log₂ transformation. Gene symbols were generated from probes based on the annotation information in the normalized data. The Kaplan-Meier (KM) survival analysis with log-rank test was applied for comparing the survival difference between the above two groups. For KM curves, p-values and hazard ratio (HR) with 95% confidence interval (CI) were generated by log-rank tests and univariate Cox proportional hazards regression as described previously [20]. All analytical methods above and R packages were performed using R software. P < 0.05 was considered as statistically significant.

Enrichment analysis

After using limma package to study the differential expression of the mRNA from TCGA, a functional enrichment analysis was conducted to further confirm the underlying function of potential targets. The Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed to annotate genes with functions. The ClusterProfiler package in R was used to analyze the GO function and enrich the KEGG pathway. The R software ggplot2 package was employed to draw boxplot, and the pheatmap package was employed to draw heatmap.

Construction and assessment of the nomogram

We further conducted Univariate and multivariate cox regression analysis to identify the proper terms to build the nomogram as described previously [21]. The forest showed the P value, HR and 95% CI of PUS1, age, gender, TNM stage and grade by “forestplot” R package. Multivariate cox proportional hazards analysis was used to develop a nomogram to predict 1-year, 3-year and 5-year overall recurrence. Based on the nomogram’s points associated with each risk factor, an individual patient’s risk of recurrence can be calculated by “RMS” R package.

The renal cancer cell line analysis

The mRNA expression matrix of 33 renal cancer cell lines was downloaded from the Cancer Cell Line Encyclopedia (CCLE) portal (https://www.broadinstitute.org/ccle) and the analysis was generated by the R ggplot2 package.

Cell culture

Human embryonic kidney 293T (HEK293T) cells and human RCC cell lines, including Caki-1, 786-O, ACHN, OSRC2, A498, Caki-2 and 769P cells, were purchased from the Chinese Academy of Sciences (Shanghai, China). HEK293T cells were maintained in high glucose DMEM (Hyclone, Logan, UT, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, Life Technologies, Inc., Grand Island, NY). ACHN cells were cultured in MEM medium (Hyclone, USA) supplemented with 10% FBS, and Caki-1, Caki-2, 786-O, OSRC2, A498 and 769P cells were all incubated in RPMI-1640 medium (Hyclone, USA) containing 10% FBS as described previously [22]. All the cells were grown as a monolayer on plastic cell culture dishes at 37°C in 5% CO₂.

Antibodies

Anti-PUS1 (#db3219) antibody was from DiagBio (Hangzhou, China) and HA-tag (AF2305) antibody was obtained from Beyotime Biotechnology (Shanghai, China). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH, sc-32233) antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

Plasmids, viruses and infections

The PUS1 lentiviral expression vector, pLV3-CMV-PUS1-3×HA-CopGFP-Puro, was constructed by Mr. Lin Li, and the pLV3-CMV-3×HA-CopGFP-Puro empty vector was used as a control. The lentiviral shRNA-expressing vector, pLKO.1-EGFP-PUS1-shRNA-puro, was constructed by Dr. Shouying Xu, and the pLKO.1-EGFP-puro empty vector carrying a scrambled sequence was used as a control. Lentiviruses were generated by co-transfecting the HEK293T packaging cells with lentiviral plasmids and packaging plasmids (PsPAX and PMD2.0G) as described previously [23]. The lentiviruses-containing supernatants with the titers greater than 1 × 10⁶ cfu/ml were used for subsequent infection in the presence of 8 μg/ml polybrene (Sigma-Aldrich, St. Louis, MO, USA). After infection, cells were selected with puromycin at a dose of 2 μg/mL for seven days, and then the cells were cultured with puromycin at a low dose and were subsequently prepared for following experiments.

Western blotting

Total protein from human RCC Caki-1, 786-O, ACHN, OSRC2, A498, Caki-2 and 769P cells were obtained using RIPA lysis buffer (Beyotime, Shanghai, China), and protein concentrations were determined by standard Bradford assay with a commercial kit (Beyotime, Shanghai, China), followed by western blotting as described previously [24–26]. Briefly, 50 μg of total protein was subjected to SDS-PAGE followed by a transfer onto polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA, USA). Then, membranes were incubated with different primary antibodies at 4°C overnight, and were further incubated with corresponding secondary antibodies (Beyotime, Shanghai, China) the next day. Immunosignals were developed by using the Enhanced Chemiluminescence System.

Cell counting kit-8 (CCK-8) assays

CCK-8 assays were conducted as described previously [26, 27]. Briefly, cells were seeded into 96-well plates and were infected with PUS1-expressing lentiviruses, control lentiviruses, PUS1-shRNA-expressing lentiviruses, or lentiviruses with scrambled shRNA. After continued culture for 48 h, cells in each well were subjected to 20 μl CCK-8 solution (Beyotime, Shanghai, China) and were cultured for 2 h. Cell viability was subsequently examined by measuring the absorbance at 450 nm via a multi-mode plate reader (Bio-Tek Instruments, Hopkinton, MA, USA).

Wound healing assays

Wound healing assays were performed as previously described [28]. Briefly, 4 × 10⁵ cells were cultured in six-well plates and were infected with PUS1-expressing lentiviruses, control lentiviruses, PUS1-shRNA-expressing lentiviruses, or lentiviruses with scrambled shRNA, and were cultured for 24 h. Then, the cells were subjected to serum starvation for 12 h. After that, cells were rinsed with medium and the confluent cell layer was scratched with a sterile tip to generate an artificial wound. After 12 h, migrated cells were observed by microscopy and the distance between the edges of the “wound” was consequently analyzed.

Matrigel invasion assays

Cell invasive capacity was determined by using Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) coated Transwell inserts (6.5 μm, Costar, Cambridge, MA, USA) containing polycarbonate filters with pores (8 μm) as detailed previously [28]. Briefly, a mixture of Matrigel and culture medium at the volume proportion of 1:2 at 50 μl was enclosed by each membrane in Transwell inserts, respectively. After infection and culture for 48 h, 2 × 10⁵ cells in a total volume of 200 μl of serum-free medium were added in the upper chamber, while 500 μl of culture medium with 10% FBS was loaded to the corresponding lower well. After culture for another 24 h, non-invaded cells were gently removed from the upper surface of the filter, and the cells that had moved through the filter and stayed at the bottom of the Transwell membrane were fixed in ice-cold methanol and were then stained with 0.2% crystal violet-PBS solution (m/v) for 10 min at room temperature. Numbers of the invasive cells from the bottom of the Transwell membrane in six randomly selected fields were counted in each experiment, and the invasive cells were consequently observed and counted using a light microscope.

Colony formation assays

Colony formation assay was performed as previously described [29]. Briefly, cells were infected with PUS1-expressing lentiviruses, control lentiviruses, PUS1-shRNA-expressing lentiviruses, or lentiviruses with scrambled shRNA. At 24 h after infection, cells were seeded in 6-well plates at a density of 1000 cells/well in 2 ml culture medium. After growing for two weeks, colonies were stained with 0.5% crystal violet solution at room temperature, and colonies were subsequently counted, photographed and statistically analyzed.

RCC tissue sample obtainment

A total of 10 pairs of RCC tissues and adjacent non-tumor tissues were collected from patients undergoing surgery at the Third Affiliated Hospital of the Second Military Medical University from 2021 to 2022. For inclusive criteria, patients didn’t receive any chemotherapy or radiotherapy before operation. All the tumor tissues and paired adjacent non-tumor tissues were identified by pathologists. A protocol associated with this study was approved by the Third Affiliated Hospital of the Second Military Medical University institutional review board, and informed consent was obtained from each patient.

Statistical analysis

Data were expressed as mean ± SD, and were analyzed by one-way ANOVA and Tukey-Kramer multiple comparison test (SPSS 13.0J software; SPSS, Inc., Chicago, IL). P < 0.05 and p < 0.01 were considered with statistical significance.

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

Consent for publication

The author declares that all work described here has not been published before and that its publication has been approved by all co-authors.

The expression of PUS1 is increased in renal cancer

Firstly, we investigated the expression of pseudouridylate synthases family in pan-cancer using The Cancer Genome Atlas (TCGA) database online. Totally 14 TCGA cancer types were enrolled, including thyroid carcinoma (THCA), kidney renal papillary cell carcinoma (KIRP), bladder urothelial carcinoma (BLCA), liver hepatocellular carcinoma (LIHC), head and neck squamous cell carcinoma (HNSC), breast invasive carcinoma (BRCA), lung adenocarcinoma (LUAD), prostate adenocarcinoma (PRAD), esophageal carcinoma (ESCA), kidney chromophobe (KICH), lung squamous cell carcinoma (LUSC), kidney renal clear cell carcinoma (KIRC), stomach adenocarcinoma (STAD) and colon adenocarcinoma (COAD), which have over ten paired tumor and normal samples, correspondingly. According to our findings, the expression of PUS1, PUS7, DKC1 and RPUSD1 is highly expressed in most cancer types, and among those genes, the expression of PUS1, PUS7 and RPUSD1 is markedly upregulated in renal cancers, including KICH, KIRC and KIRP (Figure 1A, 1B). In addition, we also examined the PUS1 gene expression in four Gene Expression Omnibus (GEO) datasets, including GSE15641, GSE17878, GSE17895 and GSE40435. As expected, the results exhibited the increase of PUS1 expression in renal cancer (Figure 1C–1F). Consistently, the overall survival of KIRC between high levels of PUS1 and low levels of PUS1 revealed statistical significance (Figure 1G), indicating the higher expression of PUS1 may be associated with prognosis of renal cancer.

Figure 1. The expression of PUS1 is increased in renal cancer. (A) Expression of PUS family gene in pan-cancer. The color of bubbles represents the difference in expression of PUS family genes between tumor tissues and paired normal para-cancer tissues. Red: up-regulated expression in tumor tissues; blue, down-regulated expression in tumor tissue; bubble size indicates significant difference. (B) The expression of PUS1 in KIRC. Red: tumor tissue; blue: paired para-cancer normal tissue. (C–F) The PUS1 gene expression in four GEO datasets, including GSE17818 (C), GSE17895 (D), GSE40435 (E) and GSE15641 (F). (G) The overall survival (OS) of PUS1 in KIRC. Red: high expression group of PUS1; blue: low expression group of PUS1.

The differential mRNA expression between clear cell renal cancer and normal tissue

We next analyzed the KIRC data from TCGA database and conducted the heatmap to compare the differentially expressed genes (DEGs) (Figure 2A). The volcano plots of DEGs were visualized between the cancer and normal tissues, and the value of log₂(fold changes) above 1.2 or below −1.2 was defined as the threshold for the differential mRNA expression (Figure 2B). In total, 9708 DEGs were picked up ultimately, with 7324 up-regulated and 2384 down-regulated (table not shown). Noticeably, the results showed that the expression of genes, including PUS1, PUS7L, RPUSD2, PUS10, RPUSD4, TRUB1, and DKC1, was increased, while conversely, the expression of some genes was decreased, such as PUSL1 (Figure 2B). To uncover the potential biological processes involved, we next performed the Gene Ontology (GO) analysis and obtained the enrichment for several cancer pathways of DEGs, which particularly focused on cell-cell adhesion and cytokine production, and intriguingly, results also suggested that pathways on immune related activity in T-cell activation were involved (Figure 2C, 2D). Similarly, we further conducted the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis with the figure portraying several notable signal pathways, which included PD-1 checkpoint, PD-L1 expression and cell cycle (Figure 2E, 2F).

Figure 2. The differential mRNA expression between clear cell renal cancer and normal tissue. (A) Heatmap color indicates the differentially expressed genes (DEGs) by analyzing the KIRC data from TCGA database. (B) The volcano plots of DEGs in (A) were visualized. (C, D) GO analysis for enrichment of upregulated (C) and downregulated (D) signaling pathways significantly associated with PUS1 expression in KIRC. (E, F) KEGG analysis for enrichment of upregulated (E) and downregulated (F) signaling pathways significantly associated with PUS1 expression in KIRC.

Construction of a predictive nomogram

Considering a vital risk of PUS1 for renal cancer patients, we further developed a clinically predictive model that was workable and may contribute to estimating the prognosis of renal cancer in clinic. We initially calculated the univariate and multivariate analysis of overall survival (OS) rate, including the expression of genes belonging to the PUS family (PUS1, PUS3, PUS7, PUS10, PUS7L, PUSL1), patients’ age, gender, race and Tumor Node Metastasis (TNM)-stage, and set grade as co-variate (Figure 3A, 3B). Intriguingly, our results provided a strong evidence that, similar as TNM-stage and tumor grade, PUS1 could be a potentially high risk for renal cancer, with a univariate cox regression of (HR 2.46131, 95% CI = 1.81506–3.33766, P < 0.0001) and a multivariate cox regression of (HR 1.9173, 95% CI = 1.28641–2.8576, P = 0.00139), which had already been accepted as predictive standards for renal cancer prognosis (Figure 3C). Next, we generated a nomogram for 1-year, 3-year, and 5-year renal clear cancer patients OS rate model in the discovery group. The elements of the model contained tumor grade and PUS1 expression, with the nomogram showing a promising concordance with 1-year, 3-year, and 5-year OS (Figure 3D).

Figure 3. Construction of a predictive nomogram. (A, B) Univariate (A) and multivariate (B) Cox regression analyses were performed in renal cancer. (C) The nomogram gives 1-, 3- and 5-year overall survival in renal cancer. (D) The survival time of renal cancer dataset is displayed.

Expression levels of PUS1 are associated with RCC cell malignancy

To figure out whether dysregulation of PUS1 contributes to RCC progression, we examined the PUS1 protein expression in seven different RCC cell lines, including BCaki-1, 786-O, OSRC2, ACHN, 769P, A498 and Caki-2. Our results revealed that the expression of PUS1 in 786-O, OSRC2, ACHN cells was higher than other cell lines (Figure 4A), and the expression was particularly low in A498 cells (Figure 4A), which was consistent with the data of the renal cancer cell line analysis from the CCLE portal (Supplementary Figure 1). In addition, the mRNA expression of PUS1 was significantly increased in collected RCC tissues compared to that of paired normal tissues, reinforcing the notion that PUS1 may accelerate RCC progression (Figure 4B).

Figure 4. Alterations in PUS1 expression affect A498 cell and ACHN cell viability, migration, invasion, colony formation, and the expression of E-Cadherin and Vimentin. (A) The protein expression of PUS1 in different RCC cell lines. (B) The mRNA expression levels of PUS1 in cancer tissues and adjacent normal epithelial tissues from patients with RCC. N = 10. (C) A498 cells are infected with PUS1-LV and PUS1 expression was detected by western blot. (D) ACHN cells are infected with PUS1-shRNA-LV and PUS1 expression was detected by western blot. (E) A498 cells are infected with PUS1-LV and cell viability was measured. (F) ACHN cells are infected with PUS1-shRNA-LV and cell viability was measured. (G) A498 cells are infected with PUS1-LV and cell migration was determined by wound healing assay. (H) Quantitative statistics of (G). (I) A498 cells are infected with PUS1-LV and cell invasion was detected by matrigel-transwell assay. (J) Quantitative statistics of (I). (K) ACHN cells are infected with PUS1-shRNA-LV and cell migration was determined by wound healing assay. (L) Quantitative statistics of (K). (M) ACHN cells are infected with PUS1-shRNA-LV and cell invasion was detected by matrigel-transwell assay. (N) Quantitative statistics of (M). (O) A498 cells are infected with PUS1-LV and cell colony formation was performed. (P) Quantitative statistics of (O). (Q) ACHN cells are infected with PUS1-shRNA-LV and cell colony formation was performed. (R) Quantitative statistics of (Q). (S) Protein levels of E-Cadherin and Vimentin in A498 cells infected with PUS1-LV. (T) Protein levels of E-Cadherin and Vimentin in ACHN cells infected with PUS1-shRNA-LV. SD, error bar; ^**p < 0.01; ^*p < 0.05.

We next cloned human PUS1 coding DNA sequence (CDS) sequence and inserted it into the lentivirus vector with 3×HA tags to obtain the pLV3-CMV-PUS1-3×HA-CopGFP-Puro plasmid and generated 3×HA-PUS1 lentiviruses (PUS1-LV), whose expression in A498 cells was determined by western blot of detecting PUS1 and HA tag, respectively (Figure 4C). Meanwhile, we constructed the pLKO.1-EGFP-puro vector containing shRNA sequences targeting PUS1 to inhibit endogenous PUS1 expression and generated PUS1-shRNA lentiviruses (PUS1-shRNA-LV) that significantly suppressed PUS1 protein expression in ACHN cells (Figure 4D). Infection of PUS1-LV significantly potentiated viability in a time dependent-manner, promoted migration, enhanced invasion, and increased colony numbers in A498 cells (Figure 4E, G–4J, 4O, 4P). Conversely, infection of PUS1-shRNA-LV resulted in the opposite effects in ACHN cells, including lowered viability at different time points, particularly at 48 h, 72 h and 96 h, down-regulated migration, suppressed invasive capability, and decreased colony numbers (Figure 4F, 4K–4N, 4Q, 4R). In consistence with the migration and invasion data, the expression of epithelial-mesenchymal transition (EMT)-related markers was altered simultaneously. Noticeably, overexpression of PUS1 by PUS1-LV decreased E-Cadherin protein expression but increased Vimentin protein levels in A498 cells (Figure 4S), whereas silencing of PUS1 by PUS1-shRNA-LV induced E-Cadherin expression but repressed Vimentin protein levels in ACHN cells contrarily (Figure 4T), suggesting PUS1 expression levels were associated with RCC cell mobility and metastasis. These results, which were in agreement with our bioinformatic analysis of the RNA sequencing data, prove that, PUS1 participates in RCC progression, indicating that PUS1 might have potential predictive value for prognosis of RCC in clinic.