Characterization of renal cell carcinoma‐associated constitutional chromosome abnormalities by genome sequencing

Abstract Constitutional translocations, typically involving chromosome 3, have been recognized as a rare cause of inherited predisposition to renal cell carcinoma (RCC) for four decades. However, knowledge of the molecular basis of this association is limited. We have characterized the breakpoints by genome sequencing (GS) of constitutional chromosome abnormalities in five individuals who presented with RCC. In one individual with constitutional t(10;17)(q11.21;p11.2), the translocation breakpoint disrupted two genes: the known renal tumor suppressor gene (TSG) FLCN (and clinical features of Birt‐Hogg‐Dubé syndrome were detected) and RASGEF1A. In four cases, the rearrangement breakpoints did not disrupt known inherited RCC genes. In the second case without chromosome 3 involvement, the translocation breakpoint in an individual with a constitutional t(2;17)(q21.1;q11.2) mapped 12 Kb upstream of NLK. Interestingly, NLK has been reported to interact indirectly with FBXW7 and a previously reported RCC‐associated translocation breakpoint disrupted FBXW7. In two cases of constitutional chromosome 3 translocations, no candidate TSGs were identified in the vicinity of the breakpoints. However, in an individual with a constitutional chromosome 3 inversion, the 3p breakpoint disrupted the FHIT TSG (which has been reported previously to be disrupted in two apparently unrelated families with an RCC‐associated t(3;8)(p14.2;q24.1). These findings (a) expand the range of constitutional chromosome rearrangements that may be associated with predisposition to RCC, (b) confirm that chromosome rearrangements not involving chromosome 3 can predispose to RCC, (c) suggest that a variety of molecular mechanisms are involved the pathogenesis of translocation‐associated RCC, and (d) demonstrate the utility of GS for investigating such cases.

the FHIT TSG (which has been reported previously to be disrupted in two apparently unrelated families with an RCC-associated t(3;8)(p14.2;q24.1). These findings (a) expand the range of constitutional chromosome rearrangements that may be associated with predisposition to RCC, (b) confirm that chromosome rearrangements not involving chromosome 3 can predispose to RCC, (c) suggest that a variety of molecular mechanisms are involved the pathogenesis of translocation-associated RCC, and (d) demonstrate the utility of GS for investigating such cases.  1), such that the risk of RCC in translocation carriers was estimated to be 80% at age 60 years. 5 Subsequently somatic deletions of the short arm of chromosome 3 (3p) were found to be the most common cytogenetic abnormality in sporadic clear cell RCC suggesting the presence of critical renal tumor suppressor genes (TSGs) on 3p. 6 These developments led to the suggestion that identification of individuals with suspected inherited forms of RCC should be screened for constitutional translocations involving 3p and that the characterization of RCC-associated translocation breakpoints might lead to the identification of novel inherited RCC genes. 7 Subsequent research studies have confirmed that the short arm of chromosome 3 does indeed harbor several TSGs that are frequently inactivated in sporadic RCC (eg, VHL, PBRM1, BAP1, and RASSF1A). [8][9][10][11][12][13][14][15] In a review of previously published reports, we identified 17 RCC-associated constitutional translocations (15 of which involved a chromosome 3 breakpoint) of constitutional chromosome abnormalities associated with RCC (Table 1). 7,[16][17][18][19][20][21][22][23][24][25][26][27][28][29] Molecular characterization of the translocation breakpoints in individual cases have identified a series of candidate TSGs disrupted (or nearby) the translocation breakpoints but none of the 15 cases with chromosome 3 breakpoints was found to disrupt either 3p genes that are frequently mutated in sporadic RCC or known familial RCC genes that map outside of 3p (eg, FLCN, FH, and SDHB). The observation that the chromosome 3 breakpoints in RCC-associated translocations were heterogeneous led to the suggestion that RCC predisposition in such cases might not necessarily involve disruption of a TSG but might confer susceptibility because of instability of the derivative chromosome 3 leading to loss at an early stage of tumorigenesis. 30 Assessment and characterization of further families and individuals carrying translocations associated with predisposition to RCC may help elucidate the genetic features and mechanisms that lead to disease onset in these patients. Here, we report the results of performing genome sequencing (GS) to characterize five constitutional rearrangements detected in individuals with RCC and interpret the results in the context of previous reports of RCC-associated constitutional translocations.

| Literature review
Reports of cases of RCC with a constitutional chromosome rearrangement were identified through a search of PubMed using the search terms "renal cell carcinoma" or "renal cancer" or "kidney cancer/ tumor" and "rearrangement/inversion/translocation or chromosome" and by searching of previously published reports (performed January 2019). When previous reports had suggested candidate genes that were either close to or disrupted by the relevant chromosomal breakpoints, evidence to suggest that the genes were implicated in human cancer was sought by reviewing curated data from the Network of Cancer Genes data portal (NCG; http://ncg.kcl.ac.uk/ version 6) 31 (performed January 2019) where genes were classified as either "known cancer genes," "candidate cancer genes," or "non-cancer genes." Genes flagged as "false positive cancer genes" were designated as "non-cancer genes."

| Clinical studies
Individuals presenting with RCC and with constitutional rearrangements were ascertained through Regional Clinical Genetics Units in the United Kingdom. DNA was extracted from whole blood according to standard 3 | RESULTS

| Literature review of previously reported cases
A total of 17 previously published distinct constitutional chromosome rearrangements were identified from searches of the biomedical literature (Table 1). In 15 cases (88%), chromosome 3 was involved (all of which were reciprocal translocations) and there were a variety of partner chromosomes in the 15 translocation cases (eg, three with chromosome 6, three with chromosome 8- Table 1 and Figure 1). For the RCC-associated chromosome 3 translocation cases, the breakpoints were almost evenly distributed between the long arm (3q, n = 8) and short arm (3p; n = 7) and were heterogeneous ( Figure 2). The chromosomal rearrangement breakpoints had been mapped in 15 of 17 previously reported cases and a total of 10 candidate genes had been reported to be disrupted by the relevant rearrangement breakpoints (Table 2). Additionally, 21 genes found to be in the vicinity of translocation breakpoints and cited as relevant genes by the authors of the original report were also assessed ( Table 3). The evidence for implicating the various genes in RCC predisposition was assessed using NCG data portal (Tables 2 and 3). Of the 10 genes directly disrupted by translocation breakpoints, two are classified as known cancer genes, with all remaining genes having no evidence supporting their role in cancer. With regards to 21 genes stated to be in the vicinity of a translocation breakpoint, two were designated as known cancer genes and four were classified as candidate cancer genes.

| Clinical features of previously unreported cases
Five previously unreported constitutional chromosomal rearrangements ascertained through a patient presenting with RCC were identified through UK genetics services. The cytogenetic, clinical features, and pathological features of the five probands and (where relevant) their affected relatives are described in Table 4. There were four translocations (involving chromosome 3 in two cases) and a pericentric inversion of chromosome 3 (Table 4 and Figure 1). Two or more individuals developed RCC in three kindreds. presence of this inversion, though the number of reference spanning reads was only 2 (Table S3) (Table S3). As with other cases, differences between breakpoints on chromosome 10 and 17 from both karyotyping and GS were found with 10q11.22 mapping to 10q11.21  Table 5. Translocations will be referred to by the shortened nomenclature system as described by Ordulu et al 44 in both the text and tables hereafter.

| Computational evaluation of breakpoint-related genes
The five constitutional rearrangements were confirmed or postulated to disrupt three protein coding genes (FHIT, FLCN, and RASGEF1A). q11.2) occurring within "unorganized chromatin" regions (Table 6 and Figures S7-S16). TADs which harbored a breakpoint were assessed for encapsulated genes and the subsequently identified genes assessed for relevance to cancer via the NCG (Table 6). Analysis demonstrated two known cancer genes (NCOA4 and RET) and a further five candidate cancer genes (LLGL1, LRIG1, LYRM9, ST3GAL6, and TMEM199) were within breakpoint-containing TADs.      27,28 and the translocation breakpoints were characterized in only one of these cases. It is entirely possible that non-chromosome 3 constitutional translocations and RCC might occur coincidentally and we note that, though there was an early age at onset (37 years) in the proband with t(2;17)(q21.1;q11.2) and an unconfirmed family history of RCC in his paternal grandfather, the translocation was also found in his mother and two siblings who were unaffected at ages 58, 40, and 31 years. However, identification of a translocation breakpoint that disrupted the FLCN gene in a patient with a t(10;17) (q11.21;p11.2) illustrated the value of characterizing all RCCassociated constitutional rearrangements. Inactivating mutations in FLCN cause BHD syndrome which is characterized by facial fibrofolliculomas, pulmonary cysts, and pneumothorax and RCC. 45,59 The occurrence of fibrofolliculomas is age-dependent and pneumothorax occurs in a minority of cases and so BHD may present with non-syndromic RCC. 60 However in the family reported herein, the t (10;17)(q11.21;p11.2) was associated with other evidence of BHD syndrome. To our knowledge, this is the first description of a constitutional translocation causing BHD syndrome.

| Tumor analysis
The other novel translocation case did not disrupt a known cancer gene but occurred close to NLK (Nemo-like kinase), a serine/threonine-protein kinase, which has been associated with the noncanonical Wnt and MAPK signaling pathways. Although NLK is currently not designated as a known cancer gene, evidence of tumor suppressor activity has been reported 61-63 and a role for NLK in the stabilization of TP53 has been suggested. 64 Interestingly, NLK appears to collaborate with FBXW7 in the ubiquitination of MYB by enhancing ligation of additional ubiquitin molecules via NLK phosphorylation, leading to downregulation of cellular proliferation 65 and, previously, an RCCassociated constitutional translocation, t(3;4)(q21;q31), was demonstrated to interrupt FBXW7. 25 Furthermore, FBXW7 is a designated TSG, that is, mutated in multiple types of primary cancers and encodes an F-box protein that is part of a SCF complex thought to target cyclin E and mTOR for ubiquitin-mediated degradation. 66,67 Very recently, FBXW7 has been identified as a novel cancer predisposition gene following an analysis of individuals with Wilms tumor. 68 Additionally, it was demonstrated that FBXW7 interacts with ubiquitin-conjugating enzyme E2Q-like protein 1; this gene is known to be disrupted in a previously reported RCC translocation case, 28 suggesting an interesting connection between multiple interacting gene products in translocation-related RCC.
In conclusion, we report five new cases of RCC-associated consti- For chromosome 3 translocations, it is unclear why most cases that are not ascertained because of a personal or family history of RCC appear to be associated with a very low risk of RCC. 55 In those translocations that do predispose to RCC, there may be a combination of factors involved including instability of the translocated chromosome during cell division together with disruption of a TSG (eg, FHIT) and/or polygenic effects that increase RCC susceptibility.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.