Genomic heterogeneity of translocation renal cell carcinoma

Abstract

Purpose: Translocation renal cell carcinoma (tRCC) is a rare subtype of kidney cancer involving the TFEB/TFE3 genes. We aimed to investigate the genomic and epigenetic features of this entity.

Experimental design: Cytogenomic analysis was conducted with 250K single-nucleotide polymorphism microarrays on 16 tumor specimens and four cell lines. LINE-1 methylation, a surrogate marker of DNA methylation, was conducted on 27 cases using pyrosequencing.

Results: tRCC showed cytogenomic heterogeneity, with 31.2% and 18.7% of cases presenting similarities with clear-cell and papillary RCC profiles, respectively. The most common alteration was a 17q gain in seven tumors (44%), followed by a 9p loss in six cases (37%). Less frequent were losses of 3p and 17p in five cases (31%) each. Patients with 17q gain were older (P=0.0006), displayed more genetic alterations (P<0.003), and had a worse outcome (P=0.002) than patients without it. Analysis comparing gene-expression profiling of a subset of tumors bearing 17q gain and those without suggest large-scale dosage effects and TP53 haploinsufficiency without any somatic TP53 mutation identified. Cell line-based cytogenetic studies revealed that 17q gain can be related to isochromosome 17 and/or to multiple translocations occurring around 17q breakpoints. Finally, LINE-1 methylation was lower in tRCC tumors from adults compared with tumors from young patients (71.1% vs. 76.7%; P=0.02).

Conclusions: Our results reveal genomic heterogeneity of tRCC with similarities to other renal tumor subtypes and raise important questions about the role of TFEB/TFE3 translocations and other chromosomal imbalances in tRCC biology.

Figure 1

Genomic subtypes of Xp11.2 tRCC. TFE3 tRCC could have different RCC profiles consistent with ccRCC (n = 5), pRCC (n = 3), or novel RCC karyotypes (n = 5).

Figure 2

A, unsupervised clustering analysis of most frequently gained and lost regions in patients with tRCC showing two main subtypes. B, more alterations occurred in patients with 17q gain than in patients without the gain. C, Kaplan-Meier survival estimates for patients with and without 17q gain.

Figure 3

A, karyotype for the novel MDA-92 cell line, derived from patient T3, shows the 3p loss and 5 different translocations enumerated as follows: M1 = t(1;3), M2= t(9;17), M3= t(6;10), M4= t(18;21) and M5= different translocations involving t(X;17) B, spectral karyotyping results for the novel HCR-59 cell line derived from patient T1. C, classified karyotyping results for the novel HCR-59 cell line derived from patient T1. B and C show 6 different translocations enumerated as follows: t(1;3); t(5;17); t(17;22); t(18;21), t(6;10) and t(X;17).

Figure 4

Partial 17q gain and gene expression. GSEA of genes expressed differently between selected cases with partial 17q gain and without was performed by RNA-seq. A, significant correlation was observed between chromosomal 17q gain and expression of genes located in the 17q arm (e.g., 17q21 and 17q25). B) C6 oncogenic signature in the Molecular Signatures Database showing enrichment for genes down-regulated by TP53 in NCI-60 panel of cell lines with mutated TP53 as compared to those classified as normal. C) Ingenuity Pathway Analysis showing the network of genes consistent with TP53 inactivation. Overall there were 51 out of 89 genes which have expression direction consistent with inhibition of TP53.

Figure 5

Genomic aberrations and LINE-1 methylation. A, Spearman’s correlation between the number of chromosomal arms with gain or loss in tRCC samples and LINE-1 methylation (n = 12). B, LINE-1 methylation level is lower in tumors of adult as compared to young patients.