Multiple TP53 p.R337H haplotypes and implications for tumor susceptibility

Summary The germline TP53 p.R337H mutation is reported as the most common germline TP53 variant. It exists at a remarkably high frequency in the population of southeast Brazil as founder mutation in two distinct haplotypes with the most frequent co-segregating with the p.E134∗ variant of the XAF1 tumor suppressor and an increased cancer risk. Founder mutations demonstrate linkage disequilibrium with neighboring genetic polymorphic markers that can be used to identify the founder variant in different geographic regions and diverse populations. We report here a shared haplotype among Brazilian, Portuguese, and Spanish families and the existence of three additional distinct TP53 p.R337H alleles. Mitochondrial DNA sequencing and Y-STR profiling of Brazilian carriers of the founder TP53 p.R337H allele reveal an excess of Native American haplogroups in maternal lineages and exclusively European haplogroups in paternal lineages, consistent with communities established through male European settlers with extensive intermarriage with Indigenous women. The identification of founder and independent TP53 p.R337H alleles underlines the importance for considering the haplotype as a functional unit and the additive effects of constitutive polymorphisms and associated variants in modifier genes that can influence the cancer phenotype.


Introduction
TP53 is a tumor suppressor that regulates the cell cycle, senescence, DNA repair, metabolism, and apoptosis in response to DNA damage and other cellular stresses.Given this central role in maintaining cellular homeostasis, it is not surprising that p53 is frequently selected against in human cancer by mutation, thus rendering this tumor suppressor inactive.Although most TP53 mutations are somatically acquired, individuals who inherit an inactivating TP53 variant are highly susceptible to developing cancer at an early age with diverse tumor types and multiple malignancies.
3][4] It is now recognized that the TP53 p.R337H variant is a founder mutation 5 frequently found throughout south and southeastern Brazil, including the major states of Sa ˜o Paulo and Parana (total population of 60 million) being detected in one in every 300 individuals. 6,7Carriers of the founder TP53 p.R337H variant share an identical DNA sequence for the TP53 locus, 8 with the majority retaining homology that extends to the telomeric region of chromosome 17 encompassing a nonsense mutation (p.E134*) in XAF1 (X-linked inhibitor of apoptosis-associated factor 1). 8 Previous studies have established XAF1 as a proapoptotic tumor suppressor that functions within a positive autoregulatory loop with TP53. 9 Consistent with these findings, individuals who inherit the compound mutant haplotype (TP53 p.R337H; XAF1 p.E134*) are at an increased risk for cancer in general, sarcomas, and multiple tumors compared with those carriers with only the p.R337H mutation. 8aplotype analysis makes it possible to discriminate between a founder mutation 10 or a variant resulting from an independent mutational event.This is particularly important in the context of hypomorphic TP53 variants, as specific polymorphisms can impact p53 expression and function, and thus influence cancer risk. 11,12By integrating genotype, haplotype, and ancestry analyses we have assessed 38 TP53 p.R337H carriers from South America, North America, Europe, South Africa, and Asia to determine the diversity of the TP53 p.R337H alleles.

TP53 p.R337H carriers
Genomic DNA samples from 38 unrelated individuals (33 patients with cancer, including 25 females) harboring the TP53 p.R337H variant from 11 different countries (Brazil, Portugal, Spain, France, Germany, England, United States, Japan, American Samoa, South Africa, and Argentina) were included in this study (Table 1).Personal and family histories of cancer were collected by the genetic counselor or physician at each participating institution.Genetic testing was offered as a TP53 single gene analysis or as part of a multi-gene hereditary cancer panel (Ambry Genetics, OvaNext; Myriad, Myrisk Hereditary Cancer Test; Illumina, TruSight Hereditary Cancer Panel) with additional genetic findings included in Table 1.The validation cohort included 86 unrelated (documented for at least three generations) and unaffected Brazilian males who carry the TP53 p.R337H allele.Appropriate institutional informed consent guidelines were followed for all participants or caregivers.This study was approved by local ethics committees and the institutional review board at St. Jude Children's Research Hospital.

Haplotype determination of recurrent TP53 p.R337H alleles
Genomic DNA was isolated and used to determine TP53 sequence and the XAF1 p.E134* variant status as previously reported. 8Additionally, fluorescent-labelled PCR products for 10 polymorphic microsatellite markers located within (VNTRp53) or telomeric to the TP53 locus (Table S1) were genotyped by using the ABI 3500 sequencer (ThermoFisher Scientific) and sizes assigned to the different fragments using GeneMapper v 6.0 (ThermoFisher Scientific). 8,13Haplotypes were determined by segregation analysis of genomic DNA from parents and descendants, loss of heterozygosity in tumor samples, or inferred.

Mitochondrial DNA control region
The nucleotide sequence of mitochondrial DNA (mtDNA) hypervariable segment 1 (HVI; between nucleotide positions 16024 and 16365), segment 2 (HV2; between nucleotide positions 73 and 340), and segment 3 (HV3, covering the position 438 to 574) of mtDNA control region was determined for 37 of 38 probands, and available family members (n ¼ 26).We have also determined the mtDNA haplotype for all Brazilian male individuals harboring the p.R337H allele (validation cohort) inherited from their mother (n ¼ 45).

Y chromosome markers
A set of 23 Y-STRs (short tandem repeats) was analyzed for 13 unrelated male p.R337H carriers and four first-degree male relatives in the main cohort of 38 individuals using the PowerPlex Y23 System (Promega) as described by the manufacturer.Y-STR haplotypes were also determined for Brazilian male individuals harboring the p.R337H allele (validation cohort) that was paternally inherited (n ¼ 41).Alleles were separated and detected on an ABI 3500 sequencer (ThermoFisher Scientific) and sizes were assigned to the different fragments using GeneMapper v 6.0 (ThermoFisher Scientific).The alleles were named according to the number of repeated units, based on the sequenced allelic ladder following International Society for Forensic Genetics recommendations. 15,16he classification of Y chromosome haplogroup was done using the Haplogroup Predictor program FTDNA 2.0 (http://www.hprg.com/hapest5/index.html) and Y chromosome Haplotype Reference Database (https://yhrd.org/search).
A third TP53 p.R337H haplotype containing the R72/A1 polymorphisms but distinct from Hap1 and Hap2 based on allelic differences in the VNTRp53(n) and p53CA (n) polymorphic markers (Hap3) was observed in 10 probands from Europe and the United States (Figure 1; Table S2).A TP53 p.R337H allele with P72/A2 (Hap4) was observed in six individuals from the United States.A documented de novo TP53 p.R337H variant with P72/A2 (Hap5) was identified in a pediatric ACT patient from an American Samoa family (Table 1). 20In addition, a heterozygous TP53 p.R337H proband with ACT from South Africa (#38) 21 was identified, but the haplotype, and whether it was inherited or de novo, was not determined due to unavailability of parental DNA or tumor samples.Nonetheless, microsatellite analysis excluded this germline variant as being Hap 1 or Hap2.
The country of birth and migration status when known for each proband was annotated (Figure 2; Table 1).Carriers of Hap1 (n ¼ 3) and Hap2 (n ¼ 2) identified in the United States and Japan are known to have immigrated from Brazil (Table 1).However, seven of the eight families from the Iberian Peninsula (Spain and Portugal) report no connection to Brazil.
Interestingly, the TP53 p.R337H variant has also been observed in two ancient genomes (European Nucleotide Archive; ENA Browser [www.ebi.ac.uk]), one carrier from Mongolia and the other from Italy.Whether these ancient alleles are fixed in the human lineage or independent events was not determined. 24mily histories of cancer The cancer history for the 38 p.R337H carriers in this study was collected (Table 1).All six Hap4 probands developed breast cancer at an early age (Table 1; range from 29 to 46 years old) with five developing multiple LFS core tumor types.Except for one proband (#22), all cases tested negative for BRCA1/2 mutations.
The cancer phenotype for Hap3 probands (n ¼ 10) appears to be milder than Hap4, with four adults currently unaffected by cancer and the remaining six carriers having developed non-LFS core tumors with late onset (Table 1).Hap3 proband #24 (39 years old), presently unaffected by cancer, was self-reported to be of Ashkenazi Jewish descendant with no remarkable family history of cancer, except for a nephew with a diagnosis of choroid plexus carcinoma at age 2. Excluding two carriers who developed childhood ACT (proband #37, de novo p.R337H variant; proband #38, de novo or inherited not determined), nearly all pediatric tumors in the current cohort (n ¼ 14) were associated with the Hap1 and Hap2 founder p.R337H mutation (n ¼ 13; 93%).

Discussion
In the present study we have identified five distinct TP53 p.R337H haplotypes, with two (Hap1 and Hap2) sharing an identical Caucasian founder allele widespread in the Brazilian population. 5,8The TP53 p.R337H founder variant co-segregates with XAF1 p.E134* in the majority of Brazilian cases (Hap1) and as a single mutation (Hap2). 8Hap1 was also observed in the majority of cases from Spain and Portugal.The remaining three TP53 p.R337H haplotypes (Hap3-5) that have been identified outside of Brazil are distinct from the founder allele based on internal polymorphisms and polymorphic markers along chromosome 17p.
The p.R337H variant has been functionally characterized by studies in vitro and in vivo as hypomorphic, 2,[25][26][27] with individuals exibiting a wide range of cancer phenotypes.In many instances, however, the p.R337H variant has been classified as pathogenic and carriers, independent of haplotype and other possible pathogenic variants, are considered to be associated with classic LFS. 28he TP53 p.R337H haplotypes identified here diverge in internal TP53 polymorphisms, such as codon 72 (Pro72Arg) and PIN3 (intron 3), reported to affect p53 expression and function. 11,12Notably, codon 72 is part of the p53 proline-rich motif that is critical for interactions with MDM-2, 29 ASPPs, 30 XAF1, 9 NF-kB, 31 and proteins containing the SH3 domain. 32Specifically, the R72 variant is more active at stimulating cellular apoptosis and suppressing cellular transformation. 33In addition, the PIN3 variant (A2) has been associated with decreased p53 expression 34 and several case-control studies have reported an increased risk of various cancer types associated with the A2 allele in Caucasians, with the most consistent association reported for breast and colorectal cancers. 35It is well established that the combination of specific variants in cis can potentially modify the penetrance of coding variants, 36 with these variants exhibiting much smaller effects individually but with additive haplotypic effects.Phenotypic variability associated with another variant in cis is exemplified by the presence of XAF1 p.E134* in a subset of TP53 p.R337H carriers (Hap1) that are at higher risk of cancer in general, sarcomas, and multiple tumors compared with those with TP53 p.R337H alone (Hap2). 8lthough the number of carriers of Hap3 and Hap4 were not sufficient to establish a clear haplotype-cancer phenotype association, we did observe a pattern whereby Hap4 carriers more closely resembled classical LFS, including breast cancers, sarcomas, and multiple primary tumors.In contrast, Hap3 carriers present generally with tumors not seen typically in LFS and at late onset with incomplete penetrance, suggestive of a lower risk of cancer.
To determine the ancestry and occurrence of TP53 p.R337H variant within the framework of the history of Brazil, our analysis of mitochondrial haplogroups of carriers defined an excess of Native American haplogroups (A-D; 73%) in maternal lineages.Although many native groups in Brazil have vanished, particularly in southeast regions, 37 our observed 73% of Brazilian p.R337H carriers with the Native American matrilineal lineages is in clear contrast with present day Brazilians in the same region that still carry the genetic imprinting of early colonization phase but with much more uniform distribution: 39% European, 33% Amerindian, and 28% African lineages. 37Y-STR profiling of carriers from the paternal lineage revealed exclusively European haplogroups, particularly R1b.This composition indicates a remarkable signature among communities established by European male settlers with extensive intermarriage with Indigenous women. 38We postulate that Iberian Peninsula Europeans introduced the TP53 p.R337H variant in the Brazilian population in early colonial times.Our hypothesis is based on the knowledge that the European-Amerindian admixture started soon after the arrival of the first colonizers. 38In addition, we observed a high frequency of Y-STR haplogroup R1b in Brazilian carriers as well as in TP53 p.R337H carriers from Portugal and Spain.It is well established that R1b haplogroup is the most common Y chromosome branch of Atlantic Europe, the most frequent haplogroup in the Iberian Peninsula and with great representation among Sephardic Jews. 39Of note, our Spanish p.R337H carriers are from Extremadura, a Spanish border region adjacent to Portugal, with large Sephardic communities and a prominent source for migrants to the Americas during the 16th century. 40lthough we cannot determine whether the p.R337H founder allele first arose within the context of Hap1 or Hap2, we have observed the uncoupling of XAF1 p.E134* and TP53 p.R337H in several Hap1 Brazilian families. 8These findings suggest that the region separating XAF1 and TP53 on chromosome 17 is susceptible to recombination and that gene conversion could promote haplotype diversity.These findings also highlight the importance of determining the haplotype of this hypomorphic mutation in each carrier even when occurring within the same family. 8n conclusion, we have identified five distinct TP53 p.R337H haplotypes that differ in specific intragenic and extragenic polymorphisms, as well as associated modifiers.Given the hypomorphic nature of p.R337H, in contrast to pathogenic DNA binding domain TP53 mutants, its tumor suppressor activity can be significantly influenced by associated polymorphisms and variants.One such example is the p.R337H founder mutation that is common throughout southern Brazil, but in the context of two haplotypes that differ in the status of XAF1 and tumor susceptibility (Hap1 and Hap2). 8,41,42As shown here, additional non-founder independent p.R337H alleles (Hap3-5) have been identified that also likely influence cancer risk.Therefore, genetic counselors and health care providers must be aware of these distinct p.R337H haplotypes and the potential implications of their associated differences on p53 expression, structure, and function, and consequently cancer risk.

Figure 2 .
Figure 2. World map showing the distribution of TP53 p.R337H alleles and their constitutive haplotypes Known migration routes of carriers from Brazil to the United States, Portugal, and Japan are highlighted by purple (Hap1) and blue (Hap2) arrows with dotted lines.

Table 1 .
Geographic origin and demographics of 38 TP53 p.R337H carriers

Table 1 .
Continued a Gender defined by biological attributes.