Content uploaded by Duilio Della Libera
Author content
All content in this area was uploaded by Duilio Della Libera on Oct 19, 2018
Content may be subject to copyright.
ROLE OF CANONICAL AND NON CANONICAL SPLICING MUTATION IN
PATIENTS WITH LYNCH SYNDROME
D. Della Libera, A. D’Urso
UOC Anatomia Patologica - Ospedale S. Maria del Prato, Feltre - ULSS1 Dolomiti.
Introduction.
Hereditary cancer syndromes can predispose affected individuals and their relatives to the early
development of neoplasias belonging to the corresponding spectrum of tumours. Identification of
the causal mutation is essential for a correct diagnosis. Splicing mutations represent a considerable
percentage of disease-causing tumours1,2. Although the pathogenic role of 5’ and 3’ canonical
splice site variants is already evident, the potential impact of exonic mutations on RNA splicing,
by altering exonic splicing enhancer (ESE) and exonic splicing silencer (ESS) sequences, is still
underestimated 2. Published data document that MLH1, a mismatch repair (MMR) gene involved
in Lynch syndrome (LS), the most common form of hereditary colorectal (CRC) and endometrial
cancers (EC), shows a high degree of exon mutations leading to splicing aberrations 1. Based on
these observations, we aimed to review the molecular characterisation offered to the cohort of
patients analysed at the Anatomic Pathology Unit of Feltre Hospital (ULSS1 Dolomiti) for LS,
focusing on mutations not presently listed in the LOVD-InSiGHT database, evaluating their
potential implications in splicing processes.
Materials and Methods
Tumours were subjected to immunohistochemical (ICH) staining for MMR gene products (Roche/
Ventana, Rotkreuz, Switzerland), microsatellite instability (MSI) (HNPCC kit; Experteam, Venice,
Italy), and BRAF gene mutation analysis (Diatech Pharmacogenetics, Jesi (AN) Italy) according
to the manufacturers’ instructions, and MLH1 promoter methylation analysis as described
previously 3. Direct Sanger sequencing was used to analyse MMR genes. The multiplex ligation-
dependent probe amplification (MLPA) technique was used to evaluate large genome
rearrangements (LGRs). RNA analysis was conducted using Extrazol and RT Plus kits (ELITech
Group, Puteaux, France) according to the manufacturer’s instructions.
Results
From 2012 until now, 170 cases of CRC were screened by ICH for MMR genes, MSI and BRAF
mutation analysis, and MLH1 promoter hypermethylation. Candidate patients were subsequently
referred to the second level of molecular testing for MMR genes. The analysis revealed seven
cases harbouring LGRs involving MLH1 and MSH2 genes4 and eight cases characterised by a
small mutation affecting the MLH1, PMS2 and MSH6 genes. Among the latter, four are not listed
in the LOVD-InSiGHT database. The first mutation, c.1558+1G>A5, involves a canonical 5’
splice site consensus sequence of intron 13 of MLH1. Subsequent analysis conducted on patient-
derived RNA demonstrated that it causes exon 13 skipping, clarifying its pathogenic role. The
other three mutations—c.1281delT, c.576dupT and c.1550_1151delTT—affect the PMS2 gene.
These are frameshift mutations that give rise to a premature stop codon and a truncated protein. It
is interesting to note that, although these are exonic mutations, in silico analysis conducted using
Human Splicing Finder6 revealed that c.1281delT creates a new ESS site and alterates an ESE
one, c.576dupT alterates an ESE site, while the third, c.1550_1151delTT, does not seem to involve
splicing regulatory elements. Given the availability of c.1281delT carrier-derived RNA, we
performed further analysis to clarify how the exonic PMS2 mutation could impact on splicing
processes. Tests revealed the presence of multiple aberrant transcripts, the most abundant of which
harbour exon 11 skipping while the minority form is characterised by skipping in exons 11 and 12,
confirming the in silico results.
Discussion
Published data suggest that one-third of disease-associated alleles alter gene splicing7. In
genomes, canonical splice sequences are an essential component of the exon splicing process,
providing a specific molecular signal for the RNA splicing machinery to identify precise splice
points. The c.1558+1G>A mutation we found involves the canonical donor site consensus
sequence in intron 13 of the MLH1 gene and causes exon 13 skipping. It is interesting to note that
the LOVD-InSiGHT database reports variant c.1558+1G>T found by five different authors8,
9,19,11,12. In particular, RNA analysis conducted by Benatti et al. revealed how the c.1558+1G>T
mutation gives rise to an aberrant transcript longer than the wild-type form, including the first 108
bp of intron 13 because of the use of a cryptic acceptor splice site9. Our characterization
contributed to clarifying how the newly identified c.1558+1G>A variant that involves the same
position in the MLH1 sequence of the mutation mentioned above determines a different splicing
defect, probably because of different base substitution. While it is well established how variants in
the consensus donor (5’) and acceptor (3’) splice site regions can abolish or diminish the strength
of canonical splice sites, the potential impact of exonic variants on RNA splicing is not always
considered. Our data emphasise the possible implication of exonic mutations in MMR genes other
than MLH1 in splicing disruption. In particular, RNA analysis of the PMS2 c.1281delT carrier
shows that the exonic mutation, by creating a new ESS site and alterate an ESE one, gives rises to
a predominant aberrant transcript with exon 11 skipping. A previous review of MMR gene
transcripts underlined the existence of several naturally occurring MMR alternative transcripts
with no associated mutations, nine of which derive from PMS2 gene13. No exon 11 skipped
transcript was listed between the naturally occurring alternative mRNA sequences of PMS2,
strengthening the pathogenic role of c.1281delT in disrupting the normal splicing process. In
conclusion, the universal screening strategy for diagnosis of LS, gradually introduced from 2012
until now at our unit has allowed us to identify 8% of LS-related CRCs. The molecular analyses
we conducted proved to be particularly useful in delineating the pathogenetic role of MMR gene
variants not previously described, especially in PMS2 mutation carriers. In fact, both technical
difficulties associated with the presence of pseudogenes that hamper molecular testing, and the
low penetrance of the PMS2 mutation, could give rise to underestimation of its mutation rate,
especially in a clinical setting in which patient selection is made considering only a strong family
history for LS-related tumours. This explains why data on the proportion of PMS2-driven LS and
detailed molecular characterisation of its variants are scarce. Furthermore, our molecular
characterisation, consistent with published reports, underline the potential implication of exonic
mutations in disruption of the splicing process, clarifying one of the possible mechanisms of
MMR gene variant pathogenicity, allowing the formulation of an accurate diagnosis and
subsequent personalised clinical management and surveillance of patients with LS.
Bibliography.
1.Soukarieh O et al., PLoS Genet. (2016) 12(1): e1005756.
2.Van der Klift et al., Molecular Genetics & Genomic Medicine 3.4 (2015): 327–345.
3. Della Libera D et al. Pathologica 03 Vol. 106 (p.287) 2014
4. Della Libera D et al, Pathologica 03, Vol 109 September 2017
5. Della Libera D et al, Pathologica 03, Vol 106 (p.226) 2014
6. Desmet FO et al., Nucleic Acid Research (2009)
7. Soemedi R et al., Nature Genetics (2017) vol 49, 848-855
8. Viel A. et al., Community Genet. (1998);1(4):229-36.
9. Benatti P. et al., Am J Gastroenterol. (1998) 93(11):2219-22.
10. Rossi B.M. et al., Ann Surg Oncol. (2002) 9(6):555-61.
11. Wagner A. et al., Am. J. Hum. Genet. 72:1088–1100 (2003)
12. Pedroni M. et al., Dis Markers. (2007) 23(3):179-87.
13. Thompson B.A. et al., Clin. Genet (2015):87:100-108
