The somatic genomic landscape of chromophobe renal cell carcinoma

Authors

Caleb F. Davis

Christopher Ricketts

Min Wang

Lixing Yang

Andrew D. Cherniack

Hui Shen

W. Kimryn Rathmell

Chad J. Creighton

The Cancer Genome Atlas Research Network

Doi

10.1016/j.ccr.2014.07.014

PMID: 25155756 · DOI: 10.1016/j.ccr.2014.07.014 · Journal: Cancer Cell (2014)

TL;DR

The TCGA chromophobe renal cell carcinoma (ChRCC) project characterized 66 primary ChRCC tumors using whole-exome sequencing, whole-genome sequencing (n=50), mtDNA sequencing (n=61), RNA-seq, miRNA-seq, DNA methylation (HM450), and Affymetrix SNP6 arrays. ChRCC has a low exonic mutation rate (~0.4/Mb, ~3-fold lower than ccRCC), with only two MutSig-significant genes (TP53 at 32% and PTEN at 9%) and a characteristic loss of whole chromosomes 1, 2, 6, 10, 13, and 17 in 86% of cases. DNA methylation and gene-expression patterns place ChRCC origin in the distal nephron, distinct from ccRCC’s proximal-tubule origin. Recurrent mtDNA mutations (notably MT-ND5 in 6/61 cases) and increased mitochondrial biogenesis indicate elevated oxidative phosphorylation — counter to the Warburg pattern of ccRCC. Most strikingly, WGS uncovered recurrent genomic rearrangement breakpoints within the TERT promoter region in 6/50 cases, associated with massively elevated TERT expression and kataegis — a novel TERT-upregulation mechanism distinct from C228T/C250T point mutations or amplification (PMID:25155756).

Cohort & data

Key findings

  • Aneuploidy signature. 86% (57/66) of ChRCC showed loss of one copy of whole chromosomes 1, 2, 6, 10, 13, and 17; additional whole-chromosome losses of 3, 5, 8, 9, 11, 18, and 21 at 12–58% frequency. No focal copy-number events by GISTIC, indicating a simpler chromosomal landscape than ccRCC. All 47 classic ChRCC showed this pattern; only 10/19 eosinophilic cases did, with 4 eosinophilic cases having no CNAs at all (PMID:25155756).
  • Low mutation burden, two MutSig-significant genes. Median exonic somatic mutation rate ~0.4/Mb (~3-fold below ccRCC). One outlier case had >10/Mb mutations with a mismatch-repair-deficiency signature. Only TP53 (q<0.1, mutated in 21/66 = 32%) and PTEN (6/66 = 9% nonsilent, plus 2 homozygous deletions) reached MutSig significance (PMID:25155756).
  • mTOR-pathway involvement. Beyond PTEN, MTOR (2 cases), TSC1/TSC2 (4 cases combined), and one activating NRAS mutation contributed; in total mTOR-pathway alterations spanned 15/66 (23%) of ChRCC (PMID:25155756).
  • Distinct DNA methylation and cell of origin. Differential methylation at >64,000 of ~450,000 HM450 loci between ChRCC and ccRCC (p<0.001, beta diff >0.1). ChRCC showed more hypomethylation. CDKN2A/p16 was epigenetically silenced in 4 ChRCC cases. Gene-expression correlation against a nephron atlas (Cheval et al., 2012) placed ChRCC’s origin in the distal nephron and ccRCC’s in the proximal tubule (PMID:25155756).
  • Mitochondrial activation. Nuclear-encoded Krebs cycle and electron transport chain genes were broadly upregulated in ChRCC vs normal kidney, with elevated PPARGC1A (mitochondrial biogenesis regulator, p<1E-5) and ~4× mitochondrial genome copy number in ChRCC vs normal — opposite the transcriptional suppression of mitochondrial genes seen in ccRCC (PMID:25155756).
  • Recurrent mtDNA mutations, especially in Complex I. Across 61 ChRCC cases, 142 somatic mtDNA events were called (75 in the D-Loop); 35 events in 27 cases at >50% heteroplasmy. Complex I gene mutations occurred in 11/61 (18%) cases. MT-ND5 was the most frequently altered mtDNA gene (6 cases, all >70% heteroplasmy); 5 of these 6 were histologically eosinophilic (p<0.01, Fisher’s exact) and 3 had no CNAs (p<0.002). MT-ND5-mutated cases showed a distinct 719-gene transcriptional signature (FDR<0.05) enriched for mitochondrial genes including SDHB and NDUFS1 (PMID:25155756).
  • Genomic rearrangements and kataegis. WGS (Meerkat) detected an average of 16 rearrangements per case (range 0–207) without recurrent gene fusions. A subset of cases showed kataegis (highly localized C>T or C>G mutation clusters) co-occurring with rearrangements. APOBEC-pattern mutagenesis was significantly enriched in kataegis regions and across the whole genome in 6 cases; APOBEC3B mRNA was elevated in ChRCC vs normal kidney (PMID:25155756).
  • TERT-promoter structural rearrangements (novel mechanism). Genomic rearrangement breakpoints fell within ~10 kb upstream of the TERT transcription start site in 6/50 ChRCC by WGS (confirmed by PCR + Sanger in 6 cases). These cases had the highest TERT expression (average >500 units; p<1E-20 by t-test), well above the 3 cases harboring TERT-promoter C228T point mutations (~1 unit average). 5 of 6 rearrangements were intrachromosomal (one involved PDCD6); the sixth involved NEK5 on chromosome 13. 3 of the 6 also showed the strongest kataegis (p=0.001, Fisher’s exact). The variants were estimated to be present in nearly all tumor cells, suggesting they are early/driver events (PMID:25155756).
  • Survival correlates. mRNA, miRNA, and DNA methylation correlates of patient survival in ChRCC overlapped with cell-cycle correlates previously identified in ccRCC, but did not include the Warburg-effect-like signature seen in aggressive ccRCC (PMID:25155756).

Genes & alterations

  • TP53 — somatic mutation in 21/66 (32%); MutSig q<0.1. Mutations correlated with decreased expression of p53 transcriptional targets (PMID:25155756).
  • PTEN — nonsilent somatic mutation in 6/66 (9%) plus 2 homozygous deletions; MutSig q<0.1. Germline PTEN mutations underlie Cowden syndrome (predisposes to ChRCC) (PMID:25155756).
  • TERT — recurrent genomic rearrangement breakpoints in 6/50 ChRCC within ~10 kb upstream of the TSS, associated with >500-unit TERT expression — a novel TERT up-regulation mechanism distinct from C228T/C250T promoter point mutations (also seen here in 3 cases at low expression) and from amplification (PMID:25155756).
  • MTOR — somatic mutation in 2/66 (PMID:25155756).
  • TSC1, TSC2 — somatic mutations in 4/66 (combined). Germline TSC1/TSC2 (tuberous sclerosis complex) predisposes to ChRCC (PMID:25155756).
  • NRAS — one activating somatic mutation (PMID:25155756).
  • FLCN — germline mutations cause Birt-Hogg-Dubé syndrome; 34% of BHD-associated kidney tumors are ChRCC. FLCN is not somatically recurrent in this sporadic cohort but is invoked as a pathway-convergence point (PMID:25155756).
  • CDKN2A — epigenetically silenced in 4 ChRCC cases (PMID:25155756).
  • PPARGC1A — significantly elevated mRNA in ChRCC vs normal kidney (p<1E-5), consistent with increased mitochondrial biogenesis (PMID:25155756).
  • APOBEC3B — elevated mRNA in ChRCC vs normal; APOBEC mutational signature enriched in kataegis-bearing cases (PMID:25155756).
  • PDCD6 — partner in one intrachromosomal TERT-promoter rearrangement (PMID:25155756).
  • NEK5 — partner in the one inter-chromosomal TERT rearrangement on chromosome 13 (PMID:25155756).
  • SDHB, NDUFS1 — among 43 mitochondrial-GO genes elevated in MT-ND5-mutated cases (PMID:25155756).
  • MT-ND5 (mitochondrially encoded NADH dehydrogenase 5; not a HUGO nuclear gene) — somatic mtDNA mutations in 6/61 cases at >70% heteroplasmy; most frequently altered Complex I gene (PMID:25155756).
  • PBRM1 — invoked as a precedent for discovering frequent kidney-cancer driver mutations from modest sample sizes (originally identified in ccRCC from 25 tumors); not somatically recurrent in this ChRCC cohort (PMID:25155756).

Clinical implications

  • ChRCC is biologically distinct from ccRCC and warrants disease-specific therapeutic strategies rather than adopting ccRCC regimens — driven by different cell of origin, opposite mitochondrial-metabolism phenotype (non-Warburg), and largely non-overlapping driver genetics (PMID:25155756).
  • mTOR-pathway alterations in 23% of ChRCC (PTEN, MTOR, TSC1/2, NRAS combined) suggest a subset of ChRCC patients may be candidates for mTOR-pathway-directed therapy, though the paper does not test agents directly (PMID:25155756).
  • TERT-promoter structural rearrangements define a previously unrecognized telomerase-activation mechanism; the authors suggest comparable structural events may operate at TERT or other drivers in other cancer types and should be sought by WGS analysis (PMID:25155756).
  • Histologic subtype tracks with genomics. Eosinophilic ChRCC is enriched for MT-ND5 mutations and includes a subset with no CNAs, suggesting subtype-aware classification could be diagnostically useful (PMID:25155756).

Limitations & open questions

  • Single biopsy per tumor — intra-tumor and inter-biopsy heterogeneity not addressed; sub-clonal/temporal ordering of mtDNA, structural, and TERT-promoter events not resolved (the authors flag this explicitly) (PMID:25155756).
  • N=66 limited statistical power for genes mutated below ~10% frequency; many cancer-relevant low-frequency mutations (e.g., MTOR n=2, NRAS n=1) cannot be called recurrent in this cohort (PMID:25155756).
  • The precise mechanism by which TERT-promoter rearrangements upregulate TERT (cis-regulatory element relocation vs escape from condensed chromatin) remains unresolved (PMID:25155756).
  • Whether MT-ND5/Complex I mutations cause compensatory mitochondrial biogenesis, or alternative metabolic adaptations, is open — the paper rules out a simple Warburg-like loss-of-OXPHOS interpretation but does not establish the causal alternative (PMID:25155756).
  • ~47% of cases lack alterations in either the p53 or PTEN pathways, suggesting additional driver mechanisms (possibly in non-exonic regions) remain to be discovered (PMID:25155756).

Citations from this paper used in the wiki

  • “We describe the landscape of somatic genomic alterations of 66 chromophobe renal cell carcinomas (ChRCCs) based on multidimensional and comprehensive characterization, including mitochondrial DNA (mtDNA) and whole genome sequencing.” (Summary)
  • “Only two significant genes were thus identified (MutSig q<0.1): TP53 and PTEN.” (Results — Copy Number and Whole Exome Analysis)
  • “TP53 was frequently mutated in 32% of cases (21 of the 66 profiled) … PTEN was the next most frequently mutated, with 9% (6 of 66) nonsilent mutations” (Results)
  • “genomic targeting of the mTOR pathway occurred overall in 15 (23%) of 66 ChRCC” (Results)
  • “Electron transport chain Complex I genes were altered in 18% of cases (n=11) … the most frequently altered gene was MT-ND5, in six cases (all with >70% heteroplasmy), with five of these being histologically classified as eosinophilic ChRCC (p<0.01)” (Results — Pathway and Mitochondrial DNA Analysis)
  • “Subsequent WGS analysis identified genomic rearrangements involving the TERT promoter region, leading to breakpoints within the region in six out of 50 ChRCC cases … these cases also had the highest levels of TERT expression (average>500 units, p<1E-20, t-test)” (Results — Whole Genome Analysis)
  • “In five ChRCC cases, the TERT-associated rearrangements were intrachromosomal (one involving part of PDCD6), while the sixth case involved NEK5 on chromosome 13.” (Results)
  • “Our finding of recurrent DNA rearrangement breakpoints within the TERT promoter region in over 10% of evaluated cases represents a mechanism for increased TERT expression in cancer different from point mutations … gene amplification … and germline polymorphisms” (Discussion)

This page was processed by crosslinker on 2026-05-11.