Genomic Analysis Reveals Novel Specific Metastatic Mutations in Chinese Clear Cell Renal Cell Carcinoma

Clear cell renal cell carcinoma (ccRCC) accounts for more than 75% of renal cell carcinoma. Nearly 25% of ccRCC patients were diagnosed with metastasis. Though the genomic profile of ccRCC has been widely studied, the difference between localized and metastatic ccRCC was not clarified. Primary tumor samples and matched whole blood were collected from 106 sporadic patients diagnosed with renal clear cell carcinoma at Qilu Hospital of Shandong University from January 2017 to November 2019, and 17 of them were diagnosed with metastasis. A hybridization capture-based next-generation sequencing of 618 cancer-related genes was performed to investigate the somatic and germline variants, tumor mutation burden (TMB), and microsatellite instability (MSI). Five genes with significantly different prevalence were identified in the metastatic group, especially TOP1 (17.65% vs. 0%) and SNCAIP (17.65% vs. 0%). The altered frequency of PBRM1 (0% vs. 27%) and BAP1 (24% vs. 10%) differed between the metastatic and nonmetastatic groups, which may relate to the prognosis. Of these 106 patients, 42 patients (39.62%) had at least one alteration in DNA damage repair (DDR) genes, including 58.82% of metastatic ccRCC patients and 35.96% of ccRCC patients without metastasis. Ten pathogenic or likely pathogenic (P/LP) variants were identified in 11 sporadic clear cell renal cell carcinoma patients (10.38%), including rarely reported ATM (n=1), MUTYH (n=1), NBN (n=1), RAD51D (n=1), and BRCA2 (n=1). No significant difference in the ratio of P/LP variant carriers or TMB was identified between the metastatic and nonmetastatic groups. We found a unique genomic feature of Chinese metastatic ccRCC patients with a higher prevalence of alterations in DDR, TOP1, and SNCAIP. Further investigated studies and drug development are needed in the future.


Introduction
Renal cell carcinoma (RCC) is the seventh most common cancer in men and the ninth most common cancer in women, accounting for 2 to 3 percent of all adult malignancies. The estimated incidence and mortality of RCC in China in 2015 were 66,800 and 23,400, respectively [1]. RCC comprises more than 10 histological and molecular subtypes, with differences in the histological pattern, clinical course, and genomic feature. Of these types, clear cell renal cell carcinoma (ccRCC) accounts for the most proportion (75%-80%) and has a worse prognosis [2]. To date, surgery, especially radical nephrectomy, is still the most effective treat-ment for localized ccRCC patients. However, nearly 25% of RCC patients were diagnosed with metastasis and unable to take surgery to remove the tumor [3]. Besides, RCC is commonly resistant to chemotherapy and radiotherapy, so the 5-year survival was poor. A clear understanding of the genomic features of metastatic ccRCC will help us to select personalized treatments and develop new effective drugs.
With the widespread next-generation sequencing technology, the genomic landscape of ccRCC has been revealed by many studies [4]. ccRCC was mainly characterized by the loss of chromosome 3p, and alterations of genes involved in this location, especially VHL, SETD2, PBRM1, and BAP1, are suggested as the driver events of ccRCC [5]. Research on the genomic feature of Chinese ccRCC patients is deficient, and limited results showed a similar driver role of VHL, SETD2, PBRM1, and BAP1 genes and unique genes involved in the ubiquitin-mediated proteolysis pathway [6]. A comparison of genetic differences between localized and metastatic ccRCC has not been well depicted yet. A significant higher mutated frequency of TP53 was found in the metastatic ccRCC, but no difference in other genes or tumor mutation burden was found [7].
We conducted a genomic study to determine the difference in the gene prevalence between Chinese nonmetastatic and metastatic ccRCC patients and discovered potential related to metastasis.

Biospecimen Collection and Clinical Data.
Primary tumor samples and matched whole blood were collected from 106 sporadic patients diagnosed with renal clear cell carcinoma at Qilu Hospital of Shandong University from January 2017 to November 2019. 17 of them were diagnosed with metastasis, and the rest was localized ccRCC. Written informed consent was obtained from each patient for collection and further publication of their clinical details and tumor genomic profiles. All enrolled patients did not receive any neoadjuvant treatment before sample collection, including chemotherapy, anti-VEGFR or mTOR therapy, and immunotherapy. All tumor samples were pathologically reviewed to have at least 60% tumor cells. Clinical data were collected, including patient ID, gender, age, and metastatic status at diagnosis (Supplementary Table S1).

Target
Next-Generation Sequencing. For the matched blood and tumor samples, germline DNA (gDNA) was isolated using the DNeasy Blood & Tissue Kit (Qiagen, 69504) according to the manufacturer's instruction. 100 ng of gDNA was sheared with a Covaris E210 system (Covaris, Inc.) to target fragment sizes of 200 bp. We performed library preparation for tumor gDNA and matched germline gDNA using the Accel-NGS 2S DNA Library Kit (Swift Biosciences, Inc.) and target enrichment using the xGen Lockdown Probe kit (IDT, Inc.). The custom xGen Lockdown Probe was synthesized by IDT, Inc. for the exons and parts of introns of 618 genes of interest (Supplementary Table S2). Samples underwent paired-end sequencing on an Illumina Nextseq CN500 platform (Illumina Inc.) with a 150 bp read length and mean coverage of 1000x at a 95% capture rate and 40% dup rate.

Data Analysis.
Raw sequencing data were aligned to the reference human genome (UCSC hg19) through Burrows-Wheeler Aligner and producing a binary alignment/map (BAM) file. After the duplicate removal and local realignment, the Genome Analysis Toolkit (GATK) was used for single-nucleotide variation (SNV) and short insertion/deletion (indel) calling. Variants were annotated using the ANNOVAR software tool. Variants identified in gDNA from buffy coat fraction aliquots with allele fraction (AF) beyond 25% were determined as germline variants. The interpreta-tion of germline variants followed the standards and guidelines of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) and was independently reviewed by two genetic consultants [8]. Somatic variants with AF beyond 1% were generated from each tumor gDNA by removing the germline variants and further annotated according to the Catalog of Somatic Mutations in Cancer (COSMIC) database. The functional classification of each somatic mutation followed the interpretation and reporting standards and guidelines recommended by the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists (ASCO/CAP) [9]. The tumor mutation burden of each sample was calculated according to a published and widely applied method [10].

Statistical
Analysis. Differential mutation analysis was performed using the Chi-squared test or Fisher exact test under a dominant model. Two-sided P values less than 0.05 were considered to be statistically significant. All analyses were performed using SPSS 25.0 software.

Somatic Mutation Landscape of ccRCC with/without
Metastasis. The clinical data of 106 enrolled ccRCC patients of primary kidney tissues (comprising of 89 localized and 17 metastatic ccRCC) is shown in Table 1.
We used the Circos plot to display the gene mutations on the chromosome and found that the somatic mutations were ). The prevalence of altered genes of the total of 106 ccRCC patients was compared with The Cancer Genome Atlas (TCGA) data. Four genes with both high mutation frequency and low P value were identified, including ZFHX3 (our cohort versus TCGA, 6.60% vs. 1.17%), NOTH3 (6.60% vs. 1.41%), ARID1B (5.66% vs. 0.94%), and TSC2 (5.66% vs. 0.94%) (Figure 2(c)). Notably, a significantly higher frequency of altered E2F3 was identified in our cohort (7.55% vs. 0%) (Supplementary Table S3).

Differences in the Prevalence of Altered Genes between
ccRCC with or without Metastasis. Next, we investigated the different frequencies of altered genes between the metastatic and nonmetastatic groups (Figure 3(a)). The metastatic and nonmetastatic groups harbored 191 and 40 genes with different mutated frequencies, respectively. The top ten unique mutated genes involved in the nonmetastatic cohort ( Figure 3(b)) were PBRM1, NOTCH3, MTOR, TSC2, FAT4, KMT2A, ELOC, TSC1, SPEN, and PTPRT. On the contrary, a different prevalence of TOP1, SNCAIP, PRDM1, CDKN1A, WRN, TGFBR1, TEK, TCF7L2, SPOP, and SOX9 was involved in the metastatic cohort genes (Figure 3(b) and Supplementary Table S4). Reactome pathway analysis of the genes with different prevalence showed that the nonmetastatic group was characterized by gene sets associated with developmental biology, signal transduction signaling, PI3K/AKT activation, and cancer classic signaling pathways (Figure 3(c)), while the metastatic group was characterized by gene sets associated with homology recombination repair, DNA repair, and DNA double-strand break signaling pathway (Figure 3(c)).

Details on Genes with Different Prevalence in ccRCC with
or without Metastasis. Specifically, we identified 101 genes with different prevalence between the metastatic and nonmetastatic cohorts. Lollipop plots showed the distribution of VHL, PBRM1, and BAP1 mutations, and truncating was the most common mutation types in three genes, followed by missense and inframe indels (Figure 4(a)). Notably, TOP1 and SNCAIP variants were only identified in the metastatic group (P = 0:0035, Table 2). The distribution of TOP1 and SNCAIP variants is illustrated in Figure 4(b). On the contrary, PBRM1 was only mutated in the nonmetastatic group ( Figure 4(c)). No significant difference was found in VHL and BAP1 between the two groups.

Germline Mutations in ccRCC.
Ten pathogenic or likely pathogenic (P/LP) variants were identified in 11 sporadic clear cell renal cell carcinoma patients (10.38%). None of   BioMed Research International the P/LP carriers was with bilateral or multiple renal masses (Table 3). Three patients were with a family cancer history, but none had familial renal cell cancer. The mean age and median age at diagnosis of identified P/LP carriers were 50.72 and 50 years, respectively. Only 4 P/LP carriers (36.36%) were diagnosed below 46 years old. There were no significant differences in the age at diagnosis between P/LP carriers and noncarriers. Except VHL and FH genes, which have been widely identified in hereditary kidney cancer, we found P/LP variants in rarely reported ATM (n = 2), MUTYH (n = 2), NBN (n = 2), RAD51D (n = 1), and BRCA2 (n = 1) genes in our cohort. No significant difference in the P/LP variant carriers' ratio was identified between the metastatic and nonmetastatic groups. # Patients 022, 024, and 092 were ccRCC with metastasis. * Two sporadic patients had the same ATM variants.

Discussion
The genomic differences between Chinese metastatic and nonmetastatic ccRCC have not been revealed. Compared with TCGA, a higher prevalence of ZFHX3, NOTH3, ARID1B, TSC2, and E2F3 was identified in our cohort, suggesting a difference in the genomic feature between Chinese and Caucasian ccRCC patients. A past research found a prognostic value of E2f3 expression in ccRCC, which was significantly associated with tumor size, metastasis, lymph node metastasis, and tumor stage [11].
The genomic profile of ccRCC has been widely studied and is mainly characterized by the loss of chromosome 3p [5]. Alterations of genes involved in this location, especially VHL, SETD2, PBRM1, and BAP1, are suggested as the driver events of ccRCC (Supplementary Table S5). Interestingly, we  BioMed Research International only identified PBRM1 variants in the nonmetastatic ccRCC group (27%); on the contrary, a higher BAP1 prevalence in the metastatic ccRCC group (24% vs. 10%). BAP1 mutation is associated with worse survival in ccRCC, and PBRM1mutated patients had a favorable survival instead [12]. Differences in the prevalence of these two genes between localized and metastatic ccRCC may contribute to the different prognosis effects on survival. Meanwhile, PBRM1 alterations may correlate with the sensitivity of immune checkpoint therapy in ccRCC, though there are controversial results on its immunogenic effects [13,14].
We only identified TOP1 and SNCAIP alterations in the metastatic ccRCC, which was not reported before. TOP1 encodes topoisomerase I (Top1), which is involved in the process of DNA replication and chromosomal recombination. As depicted by the immunochemistry analysis, Top1 expression was elevated in 23.5% of RCC, and increased expression was associated with a higher grade (grades 3 and 4) [15]. Only 0.7% of ccRCC patients had TOP1 alterations in the TCGA database, and all alterations were missense both in TCGA and our cohort. Camptothecin and related drugs, including irinotecan, were designed to inhibit tumor cell

BioMed Research
International proliferation by trapping Top1 on the DNA to stop the DNA replication process. However, the TOP1 alterations we identified (including p.His399Arg, p.Thr154Lys, and p.Gln713Glu) were not function clarified yet, so the potential response to the camptothecin-related drugs could not be predicted. The other unique marker of metastasis we identified was SNCAIP, a gene encoded synuclein alpha interacting protein, which was mainly related to Parkinson disease. Due to the deficiency of research, the oncogenic or tumor suppressor role of SNCAIP was not well depicted [16]. Promoter hypermethylation of SNCAIP could serve as a marker for colorectal cancer identification [17]. A previous study revealed that SNCAIP mutations were uniquely found in the diabetic group in pancreatic cancer, mainly involved in immune-related pathways [18]. In bladder cancer, SNCAIP was dysregulated and identified as a hub gene for predicting disease progression and prognosis [19]. Meanwhile, SNCAIP was rarely mutated (0.44%) but highly amplified (11.61%) in ccRCC in the TCGA database. So, it is worthy of additional studies on SNCAIP function and potentially drug development in ccRCC. We also identified CDKN1A alterations in the metastatic groups, and loss of CDKN1A had been proved to be an unfavorable predictor of prognosis in chromophobe renal cell carcinoma patients [20].
The application of immune checkpoint inhibitors (ICIs) in ccRCC has been rapidly progressing in the past years, so  TOP1  SNCAIP  PBRM1  PBRM1  CDKN1A  SMAD4  PTK2  LATS2  KDM6A  TCF3  WRN  TGFBR1  TEK  TCF7L2  SPOP  SOX9  SMARCD1  RPTOR  RAD50  PTPRS  POLE  PGR  MAP3K13  KNSTRN  KEAP1  IFNGR1  HIST1H2BK  GREM1  FOXA1  FGFR3  FGFR2  FANCM  ERCC4  ERCC2  DNMT3B  CUL3  CTCF  CDK2  CDH1  CDC42  CDA  BTG1  BRCA1  AXIN2  ALOX12B  ALK  TET2  SPTA1  NOTCH1  EP300  BAP1  FAT1  CD276  BIRC5  WISP3  RXRA  RET  PRKCG  PEG3  PDCD1LG2  NOTCH2  NFKBIA  MUTYH  MAGI2  KDM5A  IRS1  INPPL1  HRAS  FGFR4  FANCA  CASP8  CARD11  ATR  ASXL1  ARID5B  ADGRA2  ABL1  RNF43  PPM1D  POLD1  NFE2L2  MAP2K2  LZTR1  KMT2B  GLI2  GATA6  ERRFI1  ATRX  PIK3CB  KMT2C  BRD4  ARID2  TSC2  FAT4  NOTCH3  MTOR  VHL  E2F3  KMT2D     the potential response biomarkers are arresting significant interests. The primary known molecular biomarkers for ICIs were MSI-H and TMB, respectively. Previous research had found that MSI-H was rare (0.6%) in RCC, which was also proved in our cohort as no MSI-H tumor was identified in the 106 ccRCC [21]. Research based on unmatched primary and metastatic non-small-cell lung carcinoma found a higher TMB trend in metastatic tumors [22]. However, no significant difference was found between metastatic and nonmetastatic tumors in our study. Limited by the sample sizes, we suggested that further research would be needed to reveal the difference in TMB between metastatic and nonmetastatic ccRCC. Furthermore, recent research found a correlation between DDR gene alterations and ICIs [23]. A retrospective analysis found that 5% and 16% ccRCC patients had germline and tumor variants in DDR genes, respectively. Meanwhile, oncogenic DDR gene variants were associated with better overall survival in immune-oncology but not in the tyrosine kinase inhibitor cohort [24]. A relatively higher incidence of DDR mutation was identified in the metastatic groups (58.82% versus 35.96%) in our research. Whether metastatic ccRCC patients with DDR mutation would benefit from ICIs, PARP inhibitor, or platinum needed further investigation.
Hereditary kidney cancer was estimated to account for 5%-8% of newly diagnosed kidney cancer. A recently published study found that 9.5% of Chinese sporadic earlyonset (below 46 years old at diagnosis) had P/LP variants in 10 genes and a significant correlation to family cancer history in second-degree relatives [25]. In our cohort, 10.38% of sporadic ccRCC patients had P/LP germline variants, and the main reason for the higher frequency may be because of the more widely analyzed genes involved. Notably, 81.82% (9/11) P/LP variants were in genes related to the DDR repair pathway. Of patients with P/LP variants, 7 (63.64%) would not have been referred for genetic evaluation according to NCCN and Chinese kidney cancer guidelines.

Conclusions
This present study demonstrated the genetic landscape of Chinese nonmetastatic and metastatic ccRCC patients. Different prevalence of genes, especially TOP1 and SNCAIP, was suggested to associate with metastasis. On the contrary, PBRM1 was correlated with nonmetastatic disease. A higher incidence of DNA damage repair gene alteration was identified in the metastatic disease. Further development of target drugs on TOP1, SNCAIP, and DDR might be useful for Chinese metastatic ccRCC patients.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.