Germline MLH1, MSH2 and MSH6 variants in Brazilian patients with colorectal cancer and clinical features suggestive of Lynch Syndrome

Abstract

Lynch syndrome (LS) is the most common hereditary colorectal cancer syndrome, caused by germline mutations in one of the major genes involved in mismatch repair (MMR): MLH1, MSH2, MSH6 and more rarely, PMS2. Recently, germline deletions in EPCAM have been also associated to the syndrome. Most of the pathogenic MMR mutations found in LS families occur in MLH1 or MSH2. Gene variants include missense, nonsense, frameshift mutations, large genomic rearrangements and splice-site variants and most of the studies reporting the molecular characterization of LS families have been conducted outside South America. In this study, we analyzed 60 unrelated probands diagnosed with colorectal cancer and LS criteria. Testing for germline mutations and/or rearrangements in the most commonly affected MMR genes (MLH1, MSH2, EPCAM and MSH6) was done by Sanger sequencing and MLPA. Pathogenic or likely pathogenic variants were identified in MLH1 or MSH2 in 21 probands (35.0%). Of these, approximately one-third were gene rearrangements. In addition, nine variants of uncertain significance (VUS) were identified in 10 (16.6%) of the sixty probands analyzed. Other four novel variants were identified, only in MLH1. Our results suggest that MSH6 pathogenic variants are not common among Brazilian LS probands diagnosed with CRC and that MMR gene rearrangements account for a significant proportion of the germline variants in this population underscoring the need to include rearrangement analysis in the molecular testing of Brazilian individuals with suspected Lynch syndrome.

Keywords: Colorectal cancer; Lynch syndrome; MMR genes.

Figure 1

Chromosome microarray results of the MLPA positive patients. Each patient is represented by a gray line (numbers 1, 2, 3, 4, 5 and 6). The dots represent the Weighted Log2 Ratio measured to each marker in chromosome microarray. Dots on the gray line represent markers with a Weighted Log2 Ratio = 0, indicating that no copy number variation occurs; markers with a reduced signal intensity are represented by dots under the line; Weighted Log2 Ratio = −0.5 represents a heterozygous deletion; Weighted Log2 Ratio = −1 represents a homozygous deletion; positive values for Weighted Log2 Ratio represent duplications. Arrows represent the extension of the corresponding deletions. (A) Three patients showed MLH1 heterozygous deletion of exons 17, 18 and 19. In chromosome microarray, we observed that this deletion extended to the neighbor gene LRRFIP2, exons 16–19. The deletion has the same extension in the three patients, encompassing 12,050 base pairs (the first deleted marker is at position chr3g.37090000 and the last is at position chr3g.37102050). (B) Two patients showed heterozygous EPCAM/MSH2 deletion in MLPA analysis. In chromosome microarray, the extension of the heterozygous deletion is different between the two patients. The first patient (number 4) has EPCAM complete deletion and MSH2 deletion of exons 1–4, encompassing 58,000 base pairs (the first deleted marker is at position chr2g.47582000 and the last is at position chr2g.47640000). The second patient (number 5) has EPCAM complete deletion and MSH2 deletion of exons 1–8, encompassing 133,920 base pairs (the first deleted marker is at position chr2g.47547920 and the last is at position chr2g.47682000). (C) One patient harbored a deletion of MSH2 exon 16 in MLPA analysis. Chromosome microarray is designed to detected larger deletions, since many exons may not be covered by a marker; a zoom in MSH2 exon 16 is indicated by the arrow, and no marker is present to detect copy number variation for this exon, as observed in the gray line above it. For this reason, this deletion was not confirmed by microarray technique, but we confirmed that no point mutations are present in this exon, which could result in MLPA probe hybridization failure. No other significant alterations were found in other chromosome regions in the six analyzed patients.

Figure 2

ROC Curve the MMR mutation risk prediction model PREMM1,2,6 in a sample of Brazilian Lynch syndrome probands.