Classification of Genetic Variants

  • Maurizio Genuardi
  • Elke Holinski-Feder
  • Andreas Laner
  • Alexandra Martins


Widespread resequencing for research and diagnostic purposes has disclosed a huge amount of genetic variability in the human genome, including the genes associated with inherited predisposition to colorectal cancer. The functional and clinical consequences of the gene variants identified are often difficult to predict. Therefore, it has becoming increasingly evident that standardized approaches for the clinical interpretation of gene variants are needed in order to maximize the clinical utility of molecular testing. In this chapter, we discuss strategies for variant classification, with special reference to hereditary colorectal cancer genes and to the functional and clinical points of evidence that are available for their interpretation.


mRNA functional studies Alternative splicing Nonsense- mediated mRNA decay In vitro protein assays In silico prediction tools Multifactorial Bayesian analysis Variants of uncertain significance (VUS) 

1 General Principles

High-throughput genetic technologies have revealed the extent of DNA sequence variation in humans. While two individuals differ on average by 1 nucleotide per kilobase in their coding sequences, the differences may involve a fraction as high as 0.5% of the whole genome [1].

The functional significance of such a vast genetic variability is largely unknown. Consequently, sequencing entire genomes, exomes or large gene panels yields a huge amount of data on variants of unknown consequences. In current medical practice, specific multigene molecular diagnostic tests are often preferred over exome or full genome sequencing. The former are typically focused on functionally relevant regions of specific genes, (mostly exons and exon-intron boundaries, occasionally regulatory 5′ and 3′ sequences, or other regions, depending on the underlying molecular mechanisms known for the condition tested). While more is known about the organization and function of these specific sequences (as compared to the entire genome), the effects of DNA changes identified must still be determined very carefully, given the implications for the genetic counseling of the tested individuals and their families. Therefore, clinical interpretation of DNA variants must be performed by adopting well-defined procedures that take into account multiple lines of evidence in favor or against pathogenicity.

In the case of hereditary colorectal cancer genetic screens, the data useful for variant interpretation pertain to three different domains (Table 18.1): (1) characteristics inherent to the DNA sequence; (2) clinical information on the patients and their families, including pathology and molecular tumor studies; and (3) functional data derived from studies assessing the consequences on either RNA synthesis/processing or the protein(s) encoded by the variant allele.
Table 18.1

Types of evidence used for DNA variant classification

Categories of evidence

DNA sequence



Variant location (i.e., coding, noncoding, mutation hotspot, functional domain)

Phenotype (type of cancer, tumor features, age of onset)

mRNA splicing level, allelic-specific expression and splicing pattern (in vitro and in vivo assays)

Predicted effects (i.e., truncating, splicing, missense, synonymous, intronic)

Co-segregation with the phenotype

Quantitative mRNA expression

For missense changes: Involvement of a codon previously affected by pathogenic amino acid substitutions

Type of inheritance (for semidominant and recessive conditions: correlation with phenotype if co-occurrence in trans with a known pathogenic variant)

Protein expression level, localisation and function (in vitro and in vivo assays)

Population frequency

De novo variant with negative family history

Tumor molecular studies (i.e., microsatellite instability (MSI), specific mutational signatures)

In silico predictions (RNA, protein)

2 Characteristics of the DNA Sequence

The type and location of the sequence change are the first important elements to consider. Importantly, these information are known for most variants, notable exceptions being some large rearrangements (inversions, amplifications) while clinical and/or functional data may not be available.

Some types of variants have a very high a priori likelihood of pathogenicity. For tumor suppressor or “mutator” genes, such as those involved in colorectal cancer (CRC) predisposition, these include the overwhelming majority of variants leading to the introduction of premature stop codons (i.e., nonsense, frameshift, and some splice site variants). Notable exceptions are changes that are not predicted to disrupt important protein functional domains, e.g. those that introduce stop codons in the most 3′ portion of a gene or small in-frame alterations.

Other variants, such as deep intronic and synonymous changes, have a lower likelihood of disrupting gene function. However, since they can occasionally have consequences on RNA processing, their clinical effects cannot be established in the absence of other types of evidence.

Nucleotide substitutions that cause potential missense changes are often the most problematic variants for clinical assessment. Their effects depend on a number of factors: (1) functional relevance of the affected amino acid. This can be also assessed indirectly by comparing the amino acids present in the equivalent position in orthologous sequences from other species. Also, previous involvement of the same codon in a pathogenic missense change suggests a relevant role for the wild-type amino acid, although it does not automatically imply that any substitution in that position is deleterious. (2) Type of amino acid substitution (i.e., conservative vs. non-conservative change, based on chemical and physical characteristics of the wild type and of the variant amino acid). (3) Lastly, but importantly, potential consequences on RNA, namely, splicing alterations induced by the nucleotide change; in the latter case, the amino acid change does not occur at all in the mRNA or is associated only with a fraction of the transcripts produced by the variant allele. The impact of nucleotide variants on RNA and protein integrity/function can be assessed by in vitro and/or in vivo assays (see below).

Changes in 5′ and 3′ regulatory regions can theoretically have effects on RNA processing and stability, which can be determined by RNA studies (see below).

3 Clinical Evidence: Phenotype Information

Information on the phenotype of the patient/family should be provided by referring physicians. For genes with (nearly) complete penetrance that are usually associated with highly characteristic phenotypes, such as APC, the detection of a variant in an individual who is healthy or who does not show the typical manifestations (i.e., late-onset CRC in the absence of multiple adenomas) is a clue in favor of non-pathogenicity. On the other hand, the consistent association of an APC variant with classical adenomatous polyposis in multiple unrelated families is suggestive of its pathogenicity.

The same applies to hereditary CRC syndromes associated with less specific phenotypes, such as Lynch syndrome (LS). However, in this case, the value of clinical information is lower and must be weighed against reduced penetrance and variable phenotypic expression.

The principles of traditional Bayesian linkage analysis can be very useful to assess the pathogenicity of a variant. If multiple family members are available for analysis, co-segregation of the variant with gene-specific phenotypic manifestations can be investigated, and odds ratios in favor of causality can be determined [2]. Gene-specific penetrance values must be considered, since MSH6 and PMS2 are associated with significantly lower disease risks compared to MLH1 and MSH2 [3, 4]. For diseases with reduced penetrance, it is particularly important to obtain information on cancer-affected family members, since unaffected individuals have a relatively high chance of being carriers and are therefore less informative.

Segregation analysis is also important to verify the phase when the variant of interest is found in an individual who also has a bona fide pathogenic variant, if the associated condition is autosomal recessive or autosomal dominant and shows semidominance (i.e., a more severe phenotype in individuals who are compound heterozygotes or homozygotes for pathogenic variants compared to simple heterozygotes). Among hereditary CRC conditions, the latter phenomenon is well documented for LS, where biallelic constitutional inactivation of a mismatch repair (MMR) locus is associated with a condition characterized by early-onset pediatric cancers and manifestations of type 1 neurofibromatosis [5]. Co-occurrence in trans of two pathogenic variants in the same MMR gene is expected to lead to this phenotype, named constitutional mismatch repair deficiency (CMMRD), which is more severe than LS. On the other hand, if the two variants are in cis, no inference on the sequence change under scrutiny can be made, since the phenotype could be caused by the associated pathogenic change alone. Therefore, if two variants, one pathogenic and one of unknown significance, are detected in a CMMRD patient, and segregation studies show that each one is inherited from a different parent, this is considered evidence in favor of pathogenicity. On the contrary, their detection in trans in an individual with a diagnosis of LS, for whom CMMRD can be excluded, provides evidence against pathogenicity [6]. The same principle underlies the use of co-occurrence analysis for the interpretation of variants identified in genes causing autosomal recessive conditions [7], such as MUTYH-associated polyposis (MAP): in a patient with attenuated or classical colorectal polyposis, the finding of a pathogenic variant and an unclassified MUTYH variant in trans supports pathogenicity for the latter.

Although not strictly pertaining to the clinical setting, allele population frequencies, derived from the analyses of biological samples of control subjects or stored in public genetic databases, such as gnomAD [8], are an important source of information for variant interpretation. In principle, the higher the frequency of a variant allele, the lower the likelihood of pathogenicity. However, some pathogenic alleles may attain polymorphic or nearly polymorphic frequencies in specific ethnic groups due to a founder effect. In addition, alleles causing autosomal recessive conditions, such as MAP, tend to be more frequent than dominant disease alleles. Hence when using allele frequency data as evidence for variant interpretation , one should take into account both the type of inheritance and disease prevalence, and disease-specific thresholds should be set [7].

4 Clinical Evidence: Tumor Pathology

Tumors associated with hereditary cancer syndromes often have characteristics that are unusual in the nonhereditary counterparts. For instance, medullary or triple-negative ductal breast carcinoma is significantly more frequent in carriers of BRCA1 pathogenic variants [9, 10]. In hereditary polyposis syndromes, the histology of intestinal polyps is a very important diagnostic clue: hamartomatous polyps of the juvenile or Peutz-Jeghers types are characteristic of juvenile polyposis and Peutz-Jeghers syndrome, respectively [11]. In LS patients, CRCs tend to develop in the right colon and are often mucinous with a prominent lymphocytic infiltration.

The most useful type of information for LS is derived from molecular studies. The molecular tests commonly used to identify markers of LS, microsatellite instability (MSI) and immunohistochemistry (IHC) of MMR proteins, can be equated to a sort of in vivo functional test [12]. A high degree of instability (MSI-H) or the absence of one or more MMR proteins in a tumor is indicative of MMR deficiency. Therefore, the consistent association of a MMR gene variant with MSI-H and/or with loss of the protein encoded by the variant allele is evidence for its pathogenicity. Conversely, if the tumor is microsatellite stable (MSS) or shows normal expression of the protein encoded by the variant gene and by its heterodimeric partner, the likelihood of pathogenicity is lower.

Tumors arising in subjects with other types of hereditary CRC predisposition, especially those caused by impairment of DNA repair genes, are often associated with specific molecular alterations. The base excision repair (BER) protein MUTYH is involved in the repair of oxidative damage that leads to the production of 8-oxo-guanine, which mispairs with adenine. Hence, somatic G > T transversions in driver genes, such as KRAS and APC, are more frequent in MAP-associated tumors compared to non-MAP CRCs [13]. Likewise, the most common type of mutations in tumors from biallelic carriers of pathogenic variants in NTHL1, another BER gene with different repair specificity, are C > T transitions [14]. Finally, CRCs from patients with PPAP (polymerase proofreading-associated polyposis) tend to have an ultramutated phenotype, often associated with inactivation of the MMR system [15, 16, 17]. In principle, these molecular signatures could be useful for the clinical interpretation of sequence variants identified in MUTYH, NTHL1, POLD1, and POLE, and further studies on larger number of samples are needed to establish how they can be incorporated in the classification algorithms.

In principle, clinically useful information could also derive from molecular tumor studies for the search of second hits in tumor suppressor or genome integrity maintenance genes, such as APC and the MMR genes. While it has been shown that loss of heterozygosity is not a useful marker of pathogenicity for MMR genes [18], the value of somatic MMR gene mutations ascertained by sequencing of tumor DNA has yet to be determined.

5 The Role of RNA Studies

It is now widely accepted that every nucleotide variant can potentially affect RNA expression by directly altering either the level of transcription of a gene of interest (e.g. modifications of promoter or enhancer sequences), mRNA maturation (e.g. disruption of splicing or polyadenylation signals) or mRNA stability.

Mutations affecting pre-mRNA splicing are a major cause of genetic disease, including hereditary CRC [19]. The biological and clinical interpretation of sequence variations should therefore always take into consideration a potential impact on RNA splicing. Nucleotide variants may alter the splicing pattern of the genes to which they map to, either partially (leaky variants) or totally (complete loss of the reference full-length transcript). These alterations may be due to simple events resulting in a single aberrant transcript or, less frequently, to complex anomalies yielding multiple abnormal RNAs [20, 21, 22]. As illustrated in Fig. 18.1, simple events include (i) exon skipping, (ii) deletion of a portion of an exon, (iii) retention of a contiguous intronic fragment, (iv) retention of an entire intron (no splicing), and (v) exonization of an intronic fragment located in a region noncontiguous to reference exons (e.g., inclusion of a so-called pseudoexon). Complex anomalies are often due to a combination of different simple events arising simultaneously. Moreover, some RNA splicing alterations do not necessarily result in aberrant RNA species but in changes in the ratio of normal alternative transcripts [23, 24].
Fig. 18.1

Examples of variant-induced RNA splicing defects

In most cases, variant-induced splicing defects are due to modifications of cis-acting signals that are crucial for proper RNA splicing (Fig. 18.2). The best known signals include the sequences that directly define the splice sites, i.e., the splice sites themselves (donor sites and acceptor sites) and the branch sites, as well as sequences that contribute to the recognition of the splice sites and help regulating the splicing pattern of constitutive and alternative exons [21]. The latter are generally referred to as splicing regulatory elements (SRE), can be exonic or intronic (ESR or ISR for exonic or intronic splicing regulators), and either have an enhancer or silencer role in exon inclusion (ESE and ISR, or ESS and ISS, respectively). Whereas variant-induced alterations of splice sites are relatively easy to foretell by using computational tools, those affecting potential SRE are still difficult to predict, though recent studies highlighted the promising value of new ESR-dedicated in silico tools [25, 26, 27, 28, 29, 30]. Besides the abovementioned major splicing signals, other sequence features can influence RNA splicing patterns, such as chromatin conformation, promoter strength, and RNA secondary structure. Importantly, it is currently estimated that splicing alterations may account for at least 15% up to 50–70% of all described pathogenic variants [19, 30, 31, 32, 33].
Fig. 18.2

RNA splicing signals

Several cases of splicing alterations caused by nucleotide variants mapping to genes implicated in hereditary colorectal cancer (CRC) have been reported to date, including variants in APC, MLH1, MSH2, MSH3, MSH6, MUTYH, PMS2, POLE, PTEN, and STK11. A few examples are described in Table 18.2, among which one can find nucleotide variants mapping at exon-intron junctions, within the body of the exons, or deep in the introns. As shown, splicing defects may include noncoding variants and also variants otherwise considered as missense, nonsense, or even translationally silent (synonymous). These examples illustrate why any nucleotide variant, independently of their position or coding potential, should be investigated for their eventual impact on RNA splicing, CRC-associated variants being no exception.
Table 18.2

Examples of RNA splicing anomalies caused by variants identified in genes implicated in hereditary colorectal cancer

RNA splicing anomalies



RNA data


Exon skipping



Skipping of exon 4




Skipping of exon 10




Skipping of exon 5




Skipping of exon 17




Skipping of exon 9




Skipping of exon 8




Skipping of exon 10




Skipping of exon 10




Skipping of exon 34




Skipping of exon 3




Skipping of exon 6




Skipping of exon 4


Partial deletion of an exon



Deletion of first nt of exon 11




Deletion of first 12 nt of exon 15




Deletion of first 4 nt of exon 8




Deletion of last 92 nt of exon 12




Deletion of first 8 nt of exon 3




Deletion of first 22 nt of exon 8




Deletion of last 5 nt of exon 2




Deletion of last 159 nt of exon 5


Retention of a contiguous intronic fragment



Retention of 6 last nt of intron 4




Retention of first 88 nt of intron 14




Retention of last 24 nt of intron 3




Retention of last 7 nt of intron 3




Retention of first 75 nt of intron 7


Retention of an entire intron



Retention of full intron 10




Retention of full intron 4


Pseudoexon inclusion


c.[532-941G>A(;) c.532-845A>G]

167 nt pEx (intron 4)




127 nt pEx (intron 5)




83 nt pEx (intron 10)




83 nt pEx (intron 10)




75 nt pseudoexon (intron 1)


Complex splicing anomalies



Skipping of exon 6 and partial deletion of the first 49 nt of exon 8




Skipping of exon 10 and partial deletion of first 27 nt of exon 10


APC (NM_001127510.2), MLH1 (NM_000249.2), MSH2 (NM_000251.2), MSH3 (NM_002439.4), MSH6 (NM_000179.1), MUTYH (NM_001128425.1), PMS2 (NM_000535.5), POLE (NM_006231.2), PTEN (NM_000314.4), STK11 (NM_00455)

nt nucleotides, pEx pseudoexon

Contrary to RNA splicing mutations, knowledge on variants susceptible of affecting the transcription level of CRC-implicated genes is currently scarce. Most variants identified within the minimal promoters of the MMR genes remain unstudied, only a few having been classified as non-pathogenic (Class 1), likely non-pathogenic (Class 2) or of unknown significance (Class 3). The exception is MSH2 c.-78_-77del, a promoter variant that is now considered as probably pathogenic (Class 4) in the context of Lynch syndrome [59, 60]. Another example of a genomic deletion in a promoter region which leads to a clinically relevant reduction of the expression level of the corresponding gene is the deletion of promoter 1B in the APC gene, which has been detected in several families with familial polyposis [61, 62]. Further studies are needed to assess the functional impact of variants mapping to the promoter regions of CRC-genes, for instance by measuring endogenous allele-specific expression, performing luciferase reporter assays, and determining alterations in transcription factor binding [59]. Moreover, one has to keep in mind that promoter function can also be affected by distant changes. An example of such situation can be found in Lynch syndrome patients carrying germline deletions in the 3’ portion of the EPCAM gene. These deletions lead to EPCAM transcription readthrough, causing silencing of the downstream MSH2 promoter [63].

Finally, variants which alter the stability and the turnover of mRNA have been demonstrated to modulate expression levels of cancer genes and are implicated in tumorigenesis. These variants may affect the secondary structure of 5′ or 3′ untranslated regions, miRNA binding sites, or the polyadenylation site [64]. Interestingly, even if DNA or RNA analyses do not identify a causative variant, the confirmation of a functional impact, such as dramatically reduced expression of a MMR gene, may nevertheless warrant specific clinical recommendations.

6 Strategies for RNA Analyses

Variant-induced splicing anomalies are usually detected upon performing experimental work that is in many cases motivated by preceding bioinformatic analyses. For obvious reasons, most molecular diagnostic laboratories rely on the availability of patients’ RNA samples and on conventional gene-specific RT-PCR approaches to conduct RNA splicing analyses. This type of strategy, which allowed the identification of a large number of splicing mutations in genes implicated in hereditary CRC, such as most of those described in Table 18.1, implies a deep knowledge of the normal/alternative splicing pattern of the genes of interest, and the analysis of patients’ RNA in parallel to those of several control individuals [65, 66]. Moreover, patients’ RNA studies need to take into consideration that, in the absence of a co-occuring exonic variant, it is difficult to trace intronic splicing mutations especially if abnormal frameshift transcripts are produced and degraded by nonsense-mediated decay (NMD). Treatment of cellular cultures with NMD inhibitors such as puromycin or cycloheximide can be used to outwit this limitation [30, 35, 66]. Complementary strategies include: (i) functional in cellulo assays based on the use of minigenes and (ii) massively parallel high throughput RNA sequencing (RNA-seq), each method having its advantages and limitations [18, 19, 20, 30, 35, 45, 56, 66, 67, 68]. For instance, minigene-based assays allow to both circumvent the need for patients RNA samples and to establish a direct causality effect, but depending on the type of construct, may miss complex splicing anomalies involving multiple exons. RNA-seq analyses have the advantage of interrogating multiple transcripts in parallel and at high resolution, but remain expensive for routine diagnostic applications [68, 69].

RT-PCR analyses of RNA samples from patients suspected of hereditary CRC have thus far been very useful for the identification of: (i) deleterious splicing mutations (for examples please see Table 18.1 and [6]), (ii) genetic inversions [70], (iii) imbalanced allelic expression (RT-PCR in combination with other techniques such as pyrosequencing, SNuPE or SNaPshot) [18, 56, 71, 72, 73, 74, 75] and (iv) and to discriminate PMS2 variants from those mapping to PMS2 pseudogenes [76, 77]. RNA splicing analysis may also help illuminating genotype-phenotype correlations as reported by Sjursen and colleagues who described the identification of a Turcot syndrome patient homozygous for a PMS2 splicing mutation (PMS2 c.989-1G>T) but having a phenotype milder than expected [58]. RT-PCR analysis revealed that PMS2 c.989-1G>T caused the production of two aberrant transcripts, one lacking the 156 nucleotide-long exon 10, and the other merely lacking its first 27 nucleotides (judged as probably less detrimental, which provided a rational explanation for the atypical phenotype).

It is anticipated that in the future, whole or targeted RNA-seq, will become a reality in molecular diagnostic laboratories, especially with the implementation of new approaches allowing long sequencing reads and single-cell analysis [68, 69]. Still, important efforts are expected in the field of bioinformatics in order to improve both RNA splicing predictions and RNA-seq data processing and analysis (including qualitative and quantitative aspects of these approaches). Moreover, further studies will be needed to determine the sensitivity and specificity of the different RNA splicing-dedicated methods. Other open questions relate to the characterization of alternative splicing patterns in different tissues or in a same tissue exposed to different external stimuli, the biological role of alternative isoforms, and the choice of the most relevant tissues to be analyzed for detecting disease-causing RNA splicing defects, such as those increasing genetic predispositions to CRC. Adequate RNA can in principle be extracted directly from cell culture, blood (heparin, citrate, or ethylenediaminetetraacetic acid), or tissue samples, provided that these are not kept under nonphysiological conditions and are processed on the same day. If the laboratory setup or other circumstances prevent the long-term or short-term culture of lymphocytes, RNA stabilizing agents or commercial kits like RNAlater (QIAGEN) or PAXgene (PreAnalytiX) are recommended. Although these approaches are not as costly or time-consuming as cell cultures, the degradation of mRNA transcripts harboring premature termination codons (PTCs) caused by the sophisticated cellular quality control mechanism called NMD is likely to obscure results. Briefly, PTC-harboring transcripts are recognized during the “pioneer round of translation” and subsequently degraded, which helps preventing dominant-negative effects such as the incorporation of a misfolded protein in a multi-protein complex or gain of function effects.

Even if this cellular surveillance mechanism is not perfect in recognizing all PTCs, the vast majority of PTC-carrying transcripts are degraded and thus not detectable in patient-derived RNA. Consequently, NMD can be a major source of error, a fact which, depending on the strategy used, must be taken into account when analyzing RNA.

The long-term or short-term cultivation of mononuclear peripheral blood cells, although comparatively time-consuming and laborious, has many technical advantages over the direct preparation of RNA from tissue samples. Short-term lymphocyte cultures and long-term cell cultures from Epstein-Barr virus (EBV)-immortalized white blood cells can be used with equal results; however, the latter is likely to be found in only a few laboratories, due to regulatory restrictions and the considerable operating expenses involved. Compared to sampling fresh blood or tissue, the quantity and quality of RNA from cell cultures is generally higher, and NMD inhibition can be performed. In principle, there are two equally efficient substances which have been demonstrated to reliably block NMD in cell cultures: cycloheximide and puromycin. Both of these antibiotics inhibit translation at the eukaryotic ribosome, including the pioneer round of translation, with an impact on NMD [78].

In analyzing RNA from fresh blood, cultivated white blood cells, or minigene constructs transiently expressed in cell lines for genes involved in CRC syndromes, one can wonder if they recapitulate the effects produced in disease-affected tissues. By virtue of their integral role in DNA repair, cell cycle control, cellular differentiation, and genome maintenance, most genes associated with CRC syndromes are expressed ubiquitously, especially in quickly dividing tissues like white blood cells. In this regard, peripheral blood mononuclear cells (PBMCs) are well suited for RNA studies of these genes. It is nevertheless crucial to bear in mind that this is valid only for splicing events at constitutive exons, where the splicing machinery is guided and regulated primarily by the consensus DNA motifs of splice sites [20]; these exons are generally expected to be equally spliced in all tissues.

By contrast, exons affected by alternative splicing—that is, cassette exons skipped in certain isoforms—are critically regulated by a finely tuned system of tissue-specific splicing factors and splicing regulatory proteins such as the aforementioned exonic or intronic splicing enhancers and silencers. Tissue-specific differences in splicing patterns are well documented where RNA studies have been performed in different cell lines or tissues, but these differences mainly affect alternatively spliced exons. Consequently, these studies suggest a high overall consistency of results as defined by the joint detection of the main aberrant transcript [26, 67, 79]. The differences observed between cell lines, patient-derived RNA, and minigene constructs mainly affect additional alternative/aberrant transcripts or variations of intensity between distinct alternative splicing products. The European Mismatch Repair Working Group is currently formulating consensus proposals for a standardized protocol regarding cDNA analysis and the investigation of the effect of MMR variants on RNA splicing.

Expert groups have formulated recommendations for the interpretation and classification of sequence variants for LS-associated genes (International Society for Gastrointestinal Tumors, InSiGHT) [6] and for the two breast cancer susceptibility genes BRCA1 and BRCA2 (Evidence-Based Network for the Interpretation of Germline Mutant Alleles (ENIGMA) Consortium) [65], which advocate standardized RNA testing of all suspected splice-altering variants in order to unequivocally assess the nature of the pathogenic effect.

Many laboratories routinely use reverse transcriptase (RT) PCR-amplified cDNA fragments from fresh blood collected into PAXgene tubes (PAXgene RNA), or lymphocyte cultures to assess mRNA expression levels and to detect splicing aberrations. RT-PCR products are visualized on agarose gels to determine splicing patterns and to estimate the level of mRNA expression. Subsequent cloning and sequencing of alternate transcripts can verify splicing aberrations. This is a convenient strategy for confirming or excluding pathogenic splice defects. If the variant allele tested is found to produce only transcripts with a PTC or an in-frame deletion that disrupts a validated physiologically important domain, the variant can be reliably classified as pathogenic [6, 7, 65].

Promoter variants which are suspected to disrupt a regulatory element, like a transcription factor binding site, can reliably be analyzed by quantitative reverse transcription PCR (RT-qPCR) or allele-specific expression (ASE; see next paragraph). RT-qPCR enables reliable detection and quantitative measurement of products generated during each cycle of PCR process and is a valuable tool to assess the expression levels of a gene.

Reporter assays measure the activity of a promoter and are commonly used to study gene expression at the transcriptional level. Typically a wild type (WT) promoter sequence and the corresponding sequence containing the variant of interest are cloned in an expression vector which is thereafter transiently expressed in cell culture. The activity of the promoter can be assessed by measurement of reporter expression (usually luciferase) [80, 81].

The determination of allele-specific expression (ASE) is a powerful tool for assessing the relevance of suspected pathogenic alleles and can be performed using RNA isolated from fresh blood, PAX RNA, or cultivated lymphocytes. This approach is useful without prior knowledge of the underlying cause of a pathogenic effect, e.g., if no causative variant was identified by DNA testing due to deep intronic localization, promotor methylation, genomic inversion, translocation, etc. In single-nucleotide extension assays such as SNuPE, SNaPshot, and pyrosequencing or in MALDI-ToF mass spectrometry, ASE analysis takes advantage of a previously detected germline exonic single-nucleotide variant (SNV) as a proxy for allelic expression. This method can determine if both alleles are expressed equally or differently compared to WT control samples.

Minigene constructs are ex vivo systems in which variants can be functionally tested in a monoallelic manner, even without patient RNA. This involves the PCR amplification of patient DNA or, alternatively, the de novo construction via site-directed mutagenesis of a genomic fragment encompassing the variant of interest (preferably the entire exon) along with flanking intronic sequences, which is then cloned into a minigene system. After transient expression of these vectors in cell culture, possible differences in splicing patterns between WT transcripts and transcripts derived from the vector carrying the variant can be assessed by RT-PCR and sequencing [30, 67, 73, 79, 82]. Although potentially prone to the influence of tissue-specific splicing factors expressed in the corresponding cell lines, minigene analysis shows excellent conformity with analysis of patient-derived RNA [45]. Minor differences are observed in alternatively spliced exons which do not affect the general interpretation.

7 In Vitro Protein Functional Assays

Ideally, demonstrating that the protein encoded by the variant allele either maintains or loses the functional properties of the WT isoform should be compelling evidence for its clinical interpretation. The reality is that there are no standardized functional assays for hereditary CRC. In addition, most of the genes involved in CRC predisposition have multiple functional domains and can be involved in different cellular pathways, some of which may not be related to tumorigenesis.

In vitro assays have been developed for DNA repair genes, namely, those involved in MMR and base excision repair (BER), originally in prokaryotes, and subsequently in different eukaryotic species.

There are a number of different MMR assays that test repair activity, protein stability, interaction with partner proteins, cellular localization, resistance to alkylating agents, as well as other functions specific to single components of the MMR machinery. Some of these these tests use different artificial substrates and recombinant MMR proteins and can be performed in different cell types, including S. cerevisiae and human cells, or in cell-free systems [6, 83, 84]. They are available only in highly specialized laboratories and are currently not incorporated in routine diagnostic activities. In addition, their output is on a quantitative scale (i.e., % of repair compared to wild type), and there is interlaboratory variability in the results of assays testing the same properties [6].

Even less is known about the accuracy of in vitro assays for BER and polymerase proofreading activity. Although tests have been developed for MUTYH, they have been applied to a limited number of variants. Like MMR genes, a number of assays are available in different laboratories, and different functional properties can be investigated (Table 18.3).
Table 18.3

Functional assays for mismatch repair and base excision repair proteins



Mismatch repair (MLH1, MSH2, MSH6, PMS2)

Complementation of repair activity on artificial substrates: a. in yeast; b. in mammalian cells

[85, 86, 87, 88, 89, 90]

Repair activity in cell-free systems


Complementation of repair activity measuring mutation rates at endogenous loci (HPRT; microsatellites) in cell lines


Cellular tolerance to methylating agents

[93, 94, 95]

Protein expression and stability

[96, 97, 98, 99, 100, 101, 102]

Cellular localization

[97, 102, 103, 104, 105]

Protein-protein interactions

[106, 107, 108]

Protein-DNA binding

[109, 110]

ATPase activity, ATP/ADP cycling, ATP-induced conformational changes

[111, 112]

Base excision repair (MUTYH)

Complementation of repair activity: a. in E. coli; b. in mammalian cells

[113, 114, 115]

In vitro DNA glycosylase activity

[114, 115, 116, 117]

Protein expression

[113, 116]

Cellular localization

[113, 116]

Sensitivity to oxidative damage


Therefore, there is currently no single assay that can be used for the purpose of clinical interpretation of genetic variations in the field of hereditary CRC predisposition. In general, tests performed on mammalian systems are preferred over those in yeast or bacteria, since the conditions are more similar to those occurring in vivo in human cells. For MMR genes, it has been recommended that an assay be considered as evidence for variant classification when concordant results are obtained from two independent laboratories assessing the same function. Furthermore, multiple properties must be examined: for instance, a protein can be unstable but able to repair mismatches in vitro when expressed at levels that are, however, presumably much higher than in vivo [6]. Hence, in order to consider a variant proficient, values corresponding to the WT range must be obtained for all its different functional properties.

8 In Silico Prediction Tools

In the last two decades, a number of tools for the prediction of functional consequences have been developed to assist in the biological and clinical interpretation of DNA variant significance [7]. In this section we refer to the programs that assess the potential effects of amino acid substitutions. Importantly, their accuracy, tested versus a set of well-defined controls (i.e., variants of established pathogenicity or neutrality), is estimated between 65 and 80% [118], and specificity is particularly low, causing an excess of false positives (i.e., neutral missense changes predicted as deleterious) [119].

Due to these caveats, they should be considered as an accessory supporting source of evidence for clinical interpretation, when classification is achieved using other data. In addition, since these tools can perform differently with the same variant, depending on the gene and the protein sequence, it is advisable to use more than one program and to consider the outputs from different programs as a single piece of evidence: concordant results for pathogenicity or neutrality across two or more softwares can be used as supporting information, while discordant results are not informative [7, 120].

As discussed below, results from in silico predictions can be incorporated in multifactorial Bayesian models .

9 Multifactorial Analysis

All of the abovementioned characteristics provide qualitative clues for variant interpretation. They can also be incorporated in multifactorial Bayesian algorithms when their specificities and sensitivities have been calibrated against a robust set of reference variants of established significance. These algorithms are based on estimates of likelihood ratios (LRs) that compare for each component the probability of the observed data assuming that the variant is pathogenic versus the hypothesis of non-pathogenicity (i.e., for the MSI-H status: % MSI-H in tumors from pathogenic variant carriers/% MSI-H in tumors from noncarriers). LRs derived for each point of evidence are then cumulated, and a posterior probability of pathogenicity is calculated.

This approach has been initially developed for the BRCA1 and BRCA2 genes [2] and subsequently applied, with appropriate modifications, to the MMR genes. The MMR multifactorial model uses prior probabilities derived either from in silico predictions with the programs MAPP [121] and PolyPhen-2.1 [122] for potential missense changes or from values obtained on BRCA1/BRCA2 for intronic substitutions. LR calculations are then performed for segregation analysis, family history, and tumor molecular pathology data, specifically MSI, IHC, and BRAF mutation status [51].

Multifactorial analysis has the advantage of providing quantitative estimates that can be easily used for clinical decisions. Although this is very useful for clinicians, the model still needs improvements to obtain more accurate assessments for the components that are already incorporated and LRs from additional datasets currently not included, such as functional studies.

10 Systems for Variant Classification

The above qualitative and quantitative evidences are used to classify gene variants for use in the clinical setting. There are two main classification systems available today: one has been devised specifically for cancer predisposition genes by a working group of the International Agency for Cancer Research (IARC) [123]; the other one is the general system developed by a joint effort of the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) [7]. Both use five categories, defined by qualitative terms only for ACMG/AMP and also by numbers for IARC, ranging from pathogenic (IARC Class 5) to benign/nonpathogenic (IARC Class 1). Intermediate classes include the following categories: likely pathogenic (Class 4), variants of uncertain significance (VUSs; Class 3), and likely benign/nonpathogenic (Class 2). For Class 4 and Class 2 variants, clinical advice is the same as for Class 5 and Class 1, respectively. Class 3 includes variants for which available information has not been sufficient to establish their clinical relevance and are thus non-actionable clinically.

Classification can be achieved either by a combination of qualitative data or by multifactorial analysis. The latter is the most reliable approach, but quantitative models have been built only for a very limited number of genes so far. In the field of hereditary CRC, a Bayesian algorithm has been developed for the MMR genes only.

In the ACMG/AMP system , qualitative components are subdivided based on the strength of evidence: stand-alone, very strong, strong, moderate, or supporting (in decreasing order of strength). The only stand-alone criterion is allele population frequency > 5%, allowing classification as benign.

The IARC system does not provide qualitative criteria and refers to gene-specific recommendations, such as those devised by the InSiGHT Variant Interpretation Committee (VIC) [6]. The InSiGHT MMR rules require that concordant evidence be available pertaining to both the clinical and the functional components of classification in order to assign a variant to a clinically actionable class (5, 4, 2, or 1). Sequence-based information may be used as a stand-alone criterion for variants that have a very high prior probability of pathogenicity (i.e., nonsense substitutions, with the caveats mentioned above in “Characteristics of the DNA sequence”).

The InSiGHT criteria for MMR genes have been developed by an international multidisciplinary panel of experts and are subject to periodic revisions based on novel findings that may lead to a refinement of the interpretation rules. Their use for classification of MMR variants is therefore strongly recommended.

In the absence of specific criteria, ACMG/AMP recommendations can be used for other hereditary CRC genes, having in mind that, since these are nonspecific and less robust, the likelihood of misclassification may be higher.

From the practical standpoint, health professionals who want to have information on the clinical significance of DNA sequence variants can consult online databases. The InSiGHT website contains information on hereditary CRC genes, including the classifications of >2400 MMR gene variants and the underlying evidence for each of them ( ClinVar, hosted by the National Center for Biotechnology Information (, is a public archive of classification reports of different complexity, ranging from the representation of an allele and its interpretation to the classifications of expert panels, including the InSiGHT MMR VIC. It archives interpretations on any gene, but it does not curate submitted information nor does it perform interpretations. Therefore, the submissions must be critically assessed, checking whether they are concordant or not when there are multiple contributions for the same variant and verifying the level of evidence (i.e., single variant with no clinical information or classification of expert panel).


  1. 1.
    Gonzaga-Juaregui C, Lupski JR, Gibbs RA. Human genome sequencing in health and disease. Annu Rev Med. 2012;63:35–61. Scholar
  2. 2.
    Goldgar DE, Easton DF, Deffenbaugh AM, Monteiro AN, Tavtigian SV, Couch FJ, Breast Cancer Information Core (BIC) Steering Committee. Integrated evaluation of DNA sequence variants of unknown clinical significance: application to BRCA1 and BRCA2. Am J Hum Genet. 2004;75:535–44.CrossRefGoogle Scholar
  3. 3.
    Møller P, Seppälä T, Bernstein I, Holinski-Feder E, Sala P, Evans DG, et al. Cancer incidence and survival in lynch syndrome patients receiving colonoscopic and gynaecological surveillance: first report from the prospective lynch syndrome database. Gut. 2017;66:464–72. Scholar
  4. 4.
    Møller P, Seppälä T, Bernstein I, Holinski-Feder E, Sala P, Evans DG, et al. Incidence of and survival after subsequent cancers in carriers of pathogenic MMR variants with previous cancer: a report from the prospective lynch syndrome database. Gut. 2017;66:1657–64. Scholar
  5. 5.
    Wimmer K, Kratz CP, Vasen HF, Caron O, Colas C, Entz-Werle N, et al. Diagnostic criteria for constitutional mismatch repair deficiency syndrome: suggestions of the European consortium ‘care for CMMRD’ (C4CMMRD). J Med Genet. 2014;51:355–65. Scholar
  6. 6.
    Thompson BA, Spurdle AB, Plazzer JP, Greenblatt MS, Akagi K, Al-Mulla F, et al. Application of a five-tiered scheme for standardized classification of 2,360 unique mismatch repair gene variants lodged on the InSiGHT locus-specific database. Nat Genet. 2014;46:107–15. Scholar
  7. 7.
    Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. ACMG Laboratory Quality Assurance Committee. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17:405–24.CrossRefGoogle Scholar
  8. 8.
    Genome Aggregation Database (gnomAD). 2016. Accessed 17 Mar 2018.
  9. 9.
    Stevens KN, Vachon CM, Couch FJ. Genetic susceptibility to triple-negative breast cancer. Cancer Res. 2013;73:2025–30. Scholar
  10. 10.
    Zugazagoitia J, Pérez-Segura P, Manzano A, Blanco I, Vega A, Custodio A, et al. Limited family structure and triple-negative breast cancer (TNBC) subtype as predictors of BRCA mutations in a genetic counseling cohort of early-onset sporadic breast cancers. Breast Cancer Res Treat. 2014;148:415–21. Scholar
  11. 11.
    Lucci-Cordisco E, Risio M, Venesio T, Genuardi M. The growing complexity of the intestinal polyposis syndromes. Am J Med Genet A. 2013;161A:2777–87. Scholar
  12. 12.
    Lucci-Cordisco E, Boccuto L, Neri G, Genuardi M. The use of microsatellite instability, immunohistochemistry and other variables in determining the clinical significance of MLH1 and MSH2 unclassified variants in lynch syndrome. Cancer Biomark. 2006;2:11–27.CrossRefGoogle Scholar
  13. 13.
    Van Puijenbroek M, Nielsen M, Tops CM, Halfwerk H, Vasen HF, Weiss MM, et al. Identification of patients with (atypical) MUTYH-associated polyposis by KRAS2 c.34G > T prescreening followed by MUTYH hotspot analysis in formalin-fixed paraffin-embedded tissue. Clin Cancer Res. 2008;14:139–42. Scholar
  14. 14.
    Weren RD, Ligtenberg MJ, Kets CM, de Voer RM, Verwiel ET, Spruijt L, et al. A germline homozygous mutation in the base-excision repair gene NTHL1 causes adenomatous polyposis and colorectal cancer. Nat Genet. 2015;47:668–71. Scholar
  15. 15.
    Briggs S, Tomlinson I. Germline and somatic polymerase ε and δ mutations define a new class of hypermutated colorectal and endometrial cancers. J Pathol. 2013;230:148–53. Scholar
  16. 16.
    Jansen AM, van Wezel T, van den Akker BE, Ventayol Garcia M, Ruano D, Tops CM, et al. Combined mismatch repair and POLE/POLD1 defects explain unresolved suspected lynch syndrome cancers. Eur J Hum Genet. 2016;24:1089–92. Scholar
  17. 17.
    Howitt BE, Shukla SA, Sholl LM, Ritterhouse LL, Watkins JC, Rodig S, Stover E, Strickland KC, D’Andrea AD, Wu CJ, Matulonis UA, Konstantinopoulos PA. Association of olymerase e-mutated and microsatellite-instable endometrial cancers with neoantigen load, number of tumor-infiltrating lymphocytes, and expression of PD-1 and PD-L1. JAMA Oncol. 2015;1:1319–23. Scholar
  18. 18.
    Tricarico R, Kasela M, Mareni C, Thompson BA, Drouet A, Staderini L, et al. Assessment of the InSiGHT interpretation criteria for the clinical classification of 24 MLH1 and MSH2 gene variants. Hum Mutat. 2017;38:64–77. Scholar
  19. 19.
    Baralle D, Lucassen A, Buratti E. Missed threads. The impact of pre-mRNA splicing defects on clinical practice. EMBO Rep. 2009;10:810–6. Scholar
  20. 20.
    Baralle D, Buratti E. RNA splicing in human disease and in the clinic. Clin Sci (Lond). 2017;131:355–68. Scholar
  21. 21.
    Cartegni L, Chew SL, Krainer AR. Listening to silence and understanding nonsense: exonic mutations that affect splicing. Nat Rev Genet. 2002;3:285–98.CrossRefGoogle Scholar
  22. 22.
    Ward AJ, Cooper TA. The pathobiology of splicing. J Pathol. 2010;220:152–63. Scholar
  23. 23.
    Klamt B, Koziell A, Poulat F, Wieacker P, Scambler P, Berta P, Gessler M. Frasier syndrome is caused by defective alternative splicing of WT1 leading to an altered ratio of WT1 +/-KTS splice isoforms. Hum Mol Genet. 1998;7:709–14.CrossRefGoogle Scholar
  24. 24.
    Park SA, Ahn SI, Gallo J-M. Tau mis-splicing in the pathogenesis of neurodegenerative disorders. BMB Rep. 2016;49:405–13.CrossRefGoogle Scholar
  25. 25.
    Spurdle AB, Couch FJ, Hogervorst FBL, Radice P, Sinilnikova OM, IARC Unclassified Genetic Variants Working Group. Prediction and assessment of splicing alterations: implications for clinical testing. Hum Mutat. 2008;29:1304–13. Scholar
  26. 26.
    Houdayer C, Caux-Moncoutier V, Krieger S, Barrois M, Bonnet F, Bourdon V, et al. Guidelines for splicing analysis in molecular diagnosis derived from a set of 327 combined in silico/in vitro studies on BRCA1 and BRCA2 variants. Hum Mutat. 2012;33:1228–38. Scholar
  27. 27.
    Di Giacomo D, Gaildrat P, Abuli A, Abdat J, Frébourg T, Tosi M, Martins A. Functional analysis of a large set of BRCA2 exon 7 variants highlights the predictive value of hexamer scores in detecting alterations of exonic splicing regulatory elements. Hum Mutat. 2013;34:1547–57. Scholar
  28. 28.
    Erkelenz S, Theiss S, Otte M, Widera M, Peter JO, Schaal H. Genomic HEXploring allows landscaping of novel potential splicing regulatory elements. Nucleic Acids Res. 2014;42:10681–97. Scholar
  29. 29.
    Xiong HY, Alipanahi B, Lee LJ, Bretschneider H, Merico D, Yuen RK, et al. RNA splicing. The human splicing code reveals new insights into the genetic determinants of disease. Science. 2015;347:1254806. Scholar
  30. 30.
    Soukarieh O, Gaildrat P, Hamieh M, Drouet A, Baert-Desurmont S, Frébourg T, et al. Exonic splicing mutations are more prevalent than currently estimated and can be predicted by using in Silico tools. PLoS Genet. 2016;12:e1005756. Scholar
  31. 31.
    Teraoka SN, Telatar M, Becker-Catania S, Liang T, Onengüt S, Tolun A, et al. Splicing defects in the ataxia-telangiectasia gene, ATM: underlying mutations and consequences. Am J Hum Genet. 1999;64:1617–31.CrossRefGoogle Scholar
  32. 32.
    Ars E, Serra E, García J, Kruyer H, Gaona A, Lázaro C, Estivill X. Mutations affecting mRNA splicing are the most common molecular defects in patients with neurofibromatosis type 1. Hum Mol Genet. 2000;9:237–47.CrossRefGoogle Scholar
  33. 33.
    Lim KH, Ferraris L, Filloux ME, Raphael BJ, Fairbrother WG. Using positional distribution to identify splicing elements and predict pre-mRNA processing defects in human genes. Proc Natl Acad Sci U S A. 2011;108:11093–8. Scholar
  34. 34.
    Aretz S, Uhlhaas S, Sun Y, Pagenstecher C, Mangold E, Caspari R, et al. Familial adenomatous polyposis: aberrant splicing due to missense or silent mutations in the APC gene. Hum Mutat. 2004;24:370–80.CrossRefGoogle Scholar
  35. 35.
    Auclair J, Busine MP, Navarro C, Ruano E, Montmaigne G, Desseigne F, et al. Systematic mRNA analysis for the effect of MLH1 and MSH2 missense and silent mutations on aberrant splicing. Hum Mutat. 2006;27:145–54.CrossRefGoogle Scholar
  36. 36.
    Adam R, Spier I, Zhao B, Kloth M, Marquez J, Hinrichsen I, et al. Exome sequencing identifies Biallelic MSH3 Germline mutations as a recessive subtype of colorectal adenomatous polyposis. Am J Hum Genet. 2016;99:337–51. Scholar
  37. 37.
    Plaschke J, Linnebacher M, Kloor M, Gebert J, Cremer FW, Tinschert S, et al. Compound heterozygosity for two MSH6 mutations in a patient with early onset of HNPCC-associated cancers, but without hematological malignancy and brain tumor. Eur J Hum Genet. 2006;14:561–6.CrossRefGoogle Scholar
  38. 38.
    Dallosso AR, Dolwani S, Jones N, Jones S, Colley J, Maynard J, et al. Inherited predisposition to colorectal adenomas caused by multiple rare alleles of MUTYH but not OGG1, NUDT1, NTH1 or NEIL 1, 2 or 3. Gut. 2008;57:1252–5. Scholar
  39. 39.
    Pin E, Pastrello C, Tricarico R, Papi L, Quaia M, Fornasarig M, et al. MUTYH c.933+3A>C, associated with a severely impaired gene expression, is the first Italian founder mutation in MUTYH-associated polyposis. Int J Cancer. 2013;132:1060–9. Scholar
  40. 40.
    Borràs E, Pineda M, Cadiñanos J, Del Valle J, Brieger A, Hinrichsen I, et al. Refining the role of PMS2 in lynch syndrome: germline mutational analysis improved by comprehensive assessment of variants. J Med Genet. 2013;50:552–63. Scholar
  41. 41.
    Pachlopnik Schmid J, Lemoine R, Nehme N, Cormier-Daire V, Revy P, Debeurme F, et al. Polymerase ε1 mutation in a human syndrome with facial dysmorphism, immunodeficiency, livedo, and short stature (“FILS syndrome”). J Exp Med. 2013;209:2323–30. Scholar
  42. 42.
    Agrawal S, Pilarski R, Eng C. Different splicing defects lead to differential effects downstream of the lipid and protein phosphatase activities of PTEN. Hum Mol Genet. 2005;14:2459–68.CrossRefGoogle Scholar
  43. 43.
    Suphapeetiporn K, Kongkam P, Tantivatana J, Sinthuwiwat T, Tongkobpetch S, Shotelersuk V. PTEN c.511C>T nonsense mutation in a BRRS family disrupts a potential exonic splicing enhancer and causes exon skipping. Jpn J Clin Oncol. 2006;36:814–21.CrossRefGoogle Scholar
  44. 44.
    Papp J, Kovacs ME, Solyom S, Kasler M, Børresen-Dale A-L, Olah E. High prevalence of germline STK11 mutations in Hungarian Peutz-Jeghers syndrome patients. BMC Med Genet. 2010;11:169. Scholar
  45. 45.
    van der Klift HM, Jansen AML, van der Steenstraten N, Bik EC, Tops CMJ, Devilee P, Wijnen JT. Splicing analysis for exonic and intronic mismatch repair gene variants associated with lynch syndrome confirms high concordance between minigene assays and patient RNA analyses. Mol Genet Genomic Med. 2015;3:327–45. Scholar
  46. 46.
    Trojan J, Plotz G, Brieger A, Raedle J, Meltzer SJ, Wolter M, Zeuzem S. Activation of a cryptic splice site of PTEN and loss of heterozygosity in benign skin lesions in Cowden disease. J Invest Dermatol. 2001;117:1650–3.CrossRefGoogle Scholar
  47. 47.
    Reifenberger J, Rauch L, Beckmann MW, Megahed M, Ruzicka T, Reifenberger G. Cowden’s disease: clinical and molecular genetic findings in a patient with a novel PTEN germline mutation. Br J Dermatol. 2003;148:1040–6.CrossRefGoogle Scholar
  48. 48.
    Kaufmann A, Vogt S, Uhlhaas S, Stienen D, Kurth I, Hameister H, et al. Analysis of rare APC variants at the mRNA level: six pathogenic mutations and literature review. J Mol Diagn. 2009;11:131–9. Scholar
  49. 49.
    Sharp A, Pichert G, Lucassen A, Eccles D. RNA analysis reveals splicing mutations and loss of expression defects in MLH1 and BRCA1. Hum Mutat. 2004;24:272.CrossRefGoogle Scholar
  50. 50.
    Yan H, Papadopoulos N, Marra G, Perrera C, Jiricny J, Boland CR, et al. Conversion of diploidy to haploidy. Nature. 2000;403:723–4.CrossRefGoogle Scholar
  51. 51.
    Thompson BA, Goldgar DE, Paterson C, Clendenning M, Walters R, Arnold S, et al. A multifactorial likelihood model for MMR gene variant classification incorporating probabilities based on sequence bioinformatics and tumor characteristics: a report from the Colon Cancer family registry. Hum Mutat. 2013;34:200–9. Scholar
  52. 52.
    Celebi JT, Wanner M, Ping XL, Zhang H, Peacocke M. Association of splicing defects in PTEN leading to exon skipping or partial intron retention in Cowden syndrome. Hum Genet. 2000;107:234–8.CrossRefGoogle Scholar
  53. 53.
    Tao H, Shinmura K, Hanaoka T, Natsukawa S, Shaura K, Koizumi Y, et al. A novel splice-site variant of the base excision repair gene MYH is associated with production of an aberrant mRNA transcript encoding a truncated MYH protein not localized in the nucleus. Carcinogenesis. 2004;25:1859–66.CrossRefGoogle Scholar
  54. 54.
    Abed AA, Günther K, Kraus C, Hohenberger W, Ballhausen WG. Mutation screening at the RNA level of the STK11/LKB1 gene in Peutz-Jeghers syndrome reveals complex splicing abnormalities and a novel mRNA isoform (STK11 c.597(insertion mark)598insIVS4). Hum Mutat. 2001;18:397–410.CrossRefGoogle Scholar
  55. 55.
    Spier I, Horpaopan S, Vogt S, Ulhaas S, Morak M, Stienen D, et al. Deep intronic APC mutations explain a substantial proportion of patients with familial or early-onset adenomatous polyposis. Hum Mutat. 2012;33:1045–50. Scholar
  56. 56.
    Nieminen TT, Pavicic W, Porkka N, Kankainen M, Järvinen HJ, Lepistö A, Peltomäki P. Pseudoexons provide a mechanism for allele-specific expression of APC in familial adenomatous polyposis. Oncotarget. 2016;7:70685–98. Scholar
  57. 57.
    Clendenning M, Buchanan DD, Walsh MD, Nagler B, Rosty C, Thompson B, et al. Mutation deep within an intron of MSH2 causes lynch syndrome. Familial Cancer. 2011;10:297–301. Scholar
  58. 58.
    Sjursen W, Bjørnevoll I, Engebretsen LF, Fjelland K, Halvorsen T, Myrvold HE. A homozygote splice site PMS2 mutation as cause of Turcot syndrome gives rise to two different abnormal transcripts. Familial Cancer. 2009;8:179–86. Scholar
  59. 59.
    Liu Q, Thompson BA, Ward RL, Hesson LB, Sloane MA. Understanding the pathogenicity of noncoding mismatch gepair gene promoter variants in lynch syndrome. Hum Mutat. 2016;37:417–26. Scholar
  60. 60.
    Yan H, Jin H, Xue G, Mei Q, Ding F, Hao L, Sun SH. Germline hMSH2 promoter mutation in a Chinese HNPCC kindred: evidence for dual role of LOH. Clin Genet. 2007;72:556–61.CrossRefGoogle Scholar
  61. 61.
    Kadiyska TK, Todorov TP, Bichev SN, Vazharova RV, Nossikoff AV, Savov AS, et al. APC promoter 1B deletion in familial polyposis--implications for mutation-negative families. Clin Genet. 2014;85:452–7. Scholar
  62. 62.
    Lin Y, Lin S, Baxter MD, Lin L, Kennedy SM, Zhang Z, et al. Novel APC promoter and exon 1B deletion and allelic silencing in three mutation-negative classic familial adenomatous polyposis families. Genome Med. 2015;7:42. Scholar
  63. 63.
    Tutlewska K, Lubinski J, Kurzawski G. Germline deletions in the EPCAM gene as a cause of lynch syndrome - literature review. Hered Cancer Clin Pract. 2013;11:9. Scholar
  64. 64.
    Diederichs S, Bartsch L, Berkmann JC, Fröse K, Heitmann J, Hoppe C, et al. The dark matter of the cancer genome: aberrations in regulatory elements, untranslated regions, splice sites, non-coding RNA and synonymous mutations. EMBO Mol Med. 2016;8:442–57. Scholar
  65. 65.
  66. 66.
    Thompson BA, Martins A, Spurdle AB. A review of mismatch repair gene transcripts: issues for interpretation of mRNA splicing assays. Clin Genet. 2015;87:100–8. Scholar
  67. 67.
    Tournier I, Vezain M, Martins A, Charbonnier F, Baert-Desurmont S, Olschwang S, et al. A large fraction of unclassified variants of the mismatch repair genes MLH1 and MSH2 is associated with splicing defects. Hum Mutat. 2008;29:1412–24. Scholar
  68. 68.
    Conesa A, Madrigal P, Tarazona S, Gomez-Cabrero D, Cervera A, McPherson A, et al. A survey of best practices for RNA-seq data analysis. Genome Biol. 2016;17:13. Scholar
  69. 69.
    Saliba A-E, Westermann AJ, Gorski SA, Vogel J. Single-cell RNA-seq: advances and future challenges. Nucleic Acids Res. 2014;42:8845–60. Scholar
  70. 70.
    Liu Q, Hesson LB, Nunez AC, Packham D, Williams R, Ward RL, Sloane MA. A cryptic paracentric inversion of MSH2 exons 2-6 causes lynch syndrome. Carcinogenesis. 2016;37:10–7. Scholar
  71. 71.
    De Lellis L, Aceto GM, Curia MC, et al. Integrative analysis of hereditary nonpolyposis colorectal cancer: the contribution of allele-specific expression and other assays to diagnostic algorithms. PLoS One. 2013;8:e81194. Scholar
  72. 72.
    Hesson LB, Packham D, Kwok C-T, Nunez AC, Ng B, Schmidt C, et al. Lynch syndrome associated with two MLH1 promoter variants and allelic imbalance of MLH1 expression. Hum Mutat. 2015;36:622–30. Scholar
  73. 73.
    Lastella P, Surdo NC, Resta N, Guanti G, Stella A. In silico and in vivo splicing analysis of MLH1 and MSH2 missense mutations shows exon- and tissue-specific effects. BMC Genomics. 2006;7:243.CrossRefGoogle Scholar
  74. 74.
    Kwok C-T, Ward RL, Hawkins NJ, Hitchins MP. Detection of allelic imbalance in MLH1 expression by pyrosequencing serves as a tool for the identification of germline defects in lynch syndrome. Familial Cancer. 2010;9:345–56. Scholar
  75. 75.
    Santibanez Koref M, Wilson V, Cartwright N, Cunnington MS, Mathers JC, et al. MLH1 differential allelic expression in mutation carriers and controls. Ann Hum Genet. 2010;74:479–88. Scholar
  76. 76.
    Etzler J, Peyrl A, Zatkova A, Schildhaus H-U, Ficek A, Merkelbach-Bruse S, et al. RNA-based mutation analysis identifies an unusual MSH6 splicing defect and circumvents PMS2 pseudogene interference. Hum Mutat. 2008;29:299–305.CrossRefGoogle Scholar
  77. 77.
    van der Klift HM, Mensenkamp AR, Drost M, Bik EC, Vos YJ, et al. Comprehensive mutation analysis of PMS2 in a large cohort of probands suspected of lynch syndrome or constitutional mismatch repair deficiency syndrome. Hum Mutat. 2016;37:1162–79. Scholar
  78. 78.
    Nasif S, Contu L, Mühlemann O. Beyond quality control: the role of nonsense-mediated mRNA decay (NMD) in regulating gene expression. Semin Cell Dev Biol. 2018;75:78–87. Scholar
  79. 79.
    Bonnet C, Krieger S, Vezain M, Rousselin A, Tournier I, Martins A, et al. Screening BRCA1 and BRCA2 unclassified variants for splicing mutations using reverse transcription PCR on patient RNA and an ex vivo assay based on a splicing reporter minigene. J Med Genet. 2008;45:438–46. Scholar
  80. 80.
    Ward RL, Dobbins T, Lindor NM, Rapkins RW, Hitchins MP. Identification of constitutional MLH1 epimutations and promoter variants in colorectal cancer patients from the Colon Cancer family registry. Genet Med. 2013;15:25–35. Scholar
  81. 81.
    Waidmann MS, Bleichrodt FS, Laslo T, Riedel CU. Bacterial luciferase reporters: the Swiss army knife of molecular biology. Bioeng Bugs. 2011;2:8–16. Scholar
  82. 82.
    Gaildrat P, Krieger S, Théry JC, Killian A, Rousselin A, Berthet P, et al. The BRCA1 c.5434C->G (p.Pro1812Ala) variant induces a deleterious exon 23 skipping by affecting exonic splicing regulatory elements. J Med Genet. 2010;47:398–403. Scholar
  83. 83.
    Shimodaira H, Filosi N, Shibata H, Suzuki T, Radice P, Kanamaru R, et al. Functional analysis of human MLH1 mutations in Saccharomyces cerevisiae. Nat Genet. 1998;19:384–9.CrossRefGoogle Scholar
  84. 84.
    Nyström-Lahti M, Perrera C, Räschle M, Panyushkina-Seiler E, Marra G, Curci A, et al. Functional analysis of MLH1 mutations linked to hereditary nonpolyposis colon cancer. Genes Chromosomes Cancer. 2002;33:160–7.CrossRefGoogle Scholar
  85. 85.
    Trojan J, Zeuzem S, Randolph A, Hemmerle C, Brieger A, Raedle J, Plotz G, Jiricny J, Marra G. Functional analysis of hMLH1 variants and HNPCC-related mutations using a human expression system. Gastroenterology. 2002;122:211–9.CrossRefGoogle Scholar
  86. 86.
    Kondo E, Suzuki H, Horii A, Fukushige S. A yeast two-hybrid assay provides a simple way to evaluate the vast majority of hMLH1 germ-line mutations. Cancer Res. 2003;63:3302–8.PubMedGoogle Scholar
  87. 87.
    Plotz G, Welsch C, Giron-Monzon L, Friedhoff P, Albrecht M, Piiper A, et al. Mutations in the MutSalpha interaction interface of MLH1 can abolish DNA mismatch repair. Nucleic Acids Res. 2006;34:6574–86.CrossRefGoogle Scholar
  88. 88.
    Ollila S, Dermadi Bebek D, Jiricny J, Nystrom M. Mechanisms of pathogenicity in human MSH2 missense mutants. Hum Mutat. 2008;29:1355–63. Scholar
  89. 89.
    Drost M, Zonneveld JB, van Hees S, Rasmussen LJ, Hofstra RM, de Wind N. A rapid and cell-free assay to test the activity of lynch syndrome-associated MSH2 and MSH6 missense variants. Hum Mutat. 2012;33:488–94. Scholar
  90. 90.
    Wielders EA, Dekker RJ, Holt I, Morris GE, te Riele H. Characterization of MSH2 variants by endogenous gene modification in mouse embryonic stem cells. Hum Mutat. 2011;32:389–96. Scholar
  91. 91.
    Marra G, D’Atri S, Yan H, Perrera C, Cannavo E, Vogelstein B, Jiricny J. Phenotypic analysis of hMSH2 mutations in mouse cells carrying human chromosomes. Cancer Res. 2001;61:7719–21.Google Scholar
  92. 92.
    Blasi MF, Ventura I, Aquilina G, Degan P, Bertario L, Bassi C, et al. A human cell-based assay to evaluate the effects of alterations in the MLH1 mismatch repair gene. Cancer Res. 2006;66:9036–44.CrossRefGoogle Scholar
  93. 93.
    Mastrocola AS, Heinen CD. Lynch syndrome-associated mutations in MSH2 alter DNA repair and checkpoint response functions in vivo. Hum Mutat. 2010;31:E1699–708. Scholar
  94. 94.
    Takahashi M, Shimodaira H, Andreutti-Zaugg C, Iggo R, Kolodner RD, Ishioka C. Functional analysis of human MLH1 variants using yeast and in vitro mismatch repair assays. Cancer Res. 2007;67:4595–604.CrossRefGoogle Scholar
  95. 95.
    Belvederesi L, Bianchi F, Galizia E, Loretelli C, Bracci R, Catalani R, Amati M, Cellerino R. MSH2 missense mutations and HNPCC syndrome: pathogenicity assessment in a human expression system. Hum Mutat. 2008;29:E296–309. Scholar
  96. 96.
    Perera S, Bapat B. The MLH1 variants p.Arg265Cys and p.Lys618Ala affect protein stability while p.Leu749Gln affects heterodimer formation. Hum Mutat. 2008;29:332. Scholar
  97. 97.
    Borràs E, Pineda M, Blanco I, Jewett EM, Wang F, Teulé A, et al. MLH1 founder mutations with moderate penetrance in Spanish lynch syndrome families. Cancer Res. 2010;70:7379–91. Scholar
  98. 98.
    Hardt K, Heick SB, Betz B, Goecke T, Yazdanparast H, Küppers R, et al. Missense variants in hMLH1 identified in patients from the German HNPCC consortium and functional studies. Familial Cancer. 2011;10:273–84. Scholar
  99. 99.
    Hinrichsen I, Brieger A, Trojan J, Zeuzem S, Nilbert M, Plotz G. Expression defect size among unclassified MLH1 variants determines pathogenicity in lynch syndrome diagnosis. Clin Cancer Res. 2013;19:2432–41. Scholar
  100. 100.
    Borràs E, Pineda M, Brieger A, Hinrichsen I, Gómez C, Navarro M, et al. Comprehensive functional assessment of MLH1 variants of unknown significance. Hum Mutat. 2012;33:1576–88. Scholar
  101. 101.
    Raevaara TE, Korhonen MK, Lohi H, Hampel H, Lynch E, Lönnqvist KE, et al. Functional significance and clinical phenotype of nontruncating mismatch repair variants of MLH1. Gastroenterology. 2005;129:537–49.PubMedGoogle Scholar
  102. 102.
    Gassman NR, Clodfelter JE, McCauley AK, Bonin K, Salsbury FR Jr, Scarpinato KD. Cooperative nuclear localization sequences lend a novel role to the N-terminal region of MSH6. PLoS One. 2011;6:e17907. Scholar
  103. 103.
    Belvederesi L, Bianchi F, Loretelli C, Bracci R, Cascinu S, Cellerino R. Sub-cellular localization analysis of MSH6 missense mutations does not reveal an overt MSH6 nuclear transport impairment. Familial Cancer. 2012;11:675–80. Scholar
  104. 104.
    Guerrette S, Wilson T, Gradia S, Fishel R. Interactions of human hMSH2 with hMSH3 and hMSH2 with hMSH6: examination of mutations found in hereditary nonpolyposis colorectal cancer. Mol Cell Biol. 1998;18:6616–23.CrossRefGoogle Scholar
  105. 105.
    Guerrette S, Acharya S, Fishel R. The interaction of the human MutL homologues in hereditary nonpolyposis colon cancer. J Biol Chem. 1999;274:6336–41.CrossRefGoogle Scholar
  106. 106.
    Yuan ZQ, Gottlieb B, Beitel LK, Wong N, Gordon PH, Wang Q, et al. Polymorphisms and HNPCC: PMS2-MLH1 protein interactions diminished by single nucleotide polymorphisms. Hum Mutat. 2002;19:108–13.CrossRefGoogle Scholar
  107. 107.
    Clark AB, Cook ME, Tran HT, Gordenin DA, Resnick MA, Kunkel TA. Functional analysis of human MutS alpha and MutS beta complexes in yeast. Nucleic Acids Res. 1999;27:736–42.CrossRefGoogle Scholar
  108. 108.
    Drotschmann K, Clark AB, Tran HT, Resnick MA, Gordenin DA, Kunkel TA. Mutator phenotypes of yeast strains heterozygous for mutations in the MSH2 gene. Proc Natl Acad Sci U S A. 1999;96:2970–5.CrossRefGoogle Scholar
  109. 109.
    Iaccarino I, Marra G, Dufner P, Jiricny J. Mutation in the magnesium binding site of hMSH6 disables the hMutSalpha sliding clamp from translocating along DNA. J Biol Chem. 2000;275:2080–6.CrossRefGoogle Scholar
  110. 110.
    Heinen CD, Wilson T, Mazurek A, Berardini M, Butz C, Fishel R. HNPCC mutations in hMSH2 result in reduced hMSH2-hMSH6 molecular switch functions. Cancer Cell. 2002;1:469–78.CrossRefGoogle Scholar
  111. 111.
    Molatore S, Russo MT, D’Agostino VG, Barone F, Matsumoto Y, Albertini AM, et al. MUTYH mutations associated with familial adenomatous polyposis: functional characterization by a mammalian cell-based assay. Hum Mutat. 2010;31:159–66. Scholar
  112. 112.
    Komine K, Shimodaira H, Takao M, Soeda H, Zhang X, Takahashi M, Ishioka C. Functional complementation assay for 47 MUTYH variants in a MutY-disrupted Escherichia coli strain. Hum Mutat. 2015;36:704–11. Scholar
  113. 113.
    Shinmura K, Kato H, Goto M, Yamada H, Tao H, Nakamura S, Sugimura H. Functional evaluation of nine missense-type variants of the human DNA Glycosylase enzyme MUTYH in the Japanese population. Hum Mutat. 2016;37:350–3. Scholar
  114. 114.
    Ali M, Kim H, Cleary S, Cupples C, Gallinger S, Bristow R. Characterization of mutant MUTYH proteins associated with familial colorectal cancer. Gastroenterology. 2008;135:499–507. Scholar
  115. 115.
    D’Agostino VG, Minoprio A, Torreri P, Marioni I, Bossa C, Petrucci TC, et al. Functional analysis of MUTYH mutated proteins associated with familial adenomatous polyposis. DNA Repair. 2010;9:700–7. Scholar
  116. 116.
    Ou J, Niessen RC, Lützen A, Sijmons RH, Kleibeuker JH, de Wind N, et al. Functional analysis helps to clarify the clinical importance of unclassified variants in DNA mismatch repair genes. Hum Mutat. 2007;28:1047–54.CrossRefGoogle Scholar
  117. 117.
    Rasmussen LJ, Heinen CD, Royer-Pokora B, Drost M, Tavtigian S, Hofstra RM, de Wind N. Pathological assessment of mismatch repair gene variants in lynch syndrome: past, present, and future. Hum Mutat. 2012;33:1617–25. Scholar
  118. 118.
    Thusberg J, Olatubosun A, Vihinen M. Performance of mutation pathogenicity prediction methods on missense variants. Hum Mutat. 2011;32:358–68. Scholar
  119. 119.
    Choi Y, Sims GE, Murphy S, Miller JR, Chan AP. Predicting the functional effect of amino acid substitutions and indels. PLoS One. 2012;7:e46688. Scholar
  120. 120.
    Chan PA, Duraisamy S, Miller PJ, Newell JA, McBride C, Bond JP, et al. Interpreting missense variants: comparing computational methods in human disease genes CDKN2A, MLH1, MSH2, MECP2, and tyrosinase (TYR). Hum Mutat. 2007;28:683–93.CrossRefGoogle Scholar
  121. 121.
    Stone EA, Sidow A. Physicochemical constraint violation by missense substitutions mediates impairment of protein function and disease severity. Genome Res. 2005;15:978–86.CrossRefGoogle Scholar
  122. 122.
    Adzhubei I, Jordan DM, Sunyaev SR. Predicting functional effect of human missense mutations using PolyPhen-2. Curr Protoc Hum Genet. 2013.
  123. 123.
    Plon SE, Eccles DM, Easton D, Foulkes WD, Genuardi M, Greenblatt MS, et al. Sequence variant classification and reporting: recommendations for improving the interpretation of cancer susceptibility genetic test results. Hum Mutat. 2008;29:1282–91. Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Maurizio Genuardi
    • 1
    • 2
  • Elke Holinski-Feder
    • 3
    • 4
  • Andreas Laner
    • 3
  • Alexandra Martins
    • 5
  1. 1.Institute of Genomic Medicine, Catholic University of the Sacred HeartRomeItaly
  2. 2.Fondazione Policlinico Universitario “A. Gemelli”RomeItaly
  3. 3.Medizinische Klinik und Poliklinik IV, Campus Innenstadt, Klinikum der Universität MünchenMunichGermany
  4. 4.MGZ – Medizinisch Genetisches ZentrumMunichGermany
  5. 5.Inserm-U1245-IRIB, Normandy Centre for Genomic and Personalized Medicine, University of RouenRouenFrance

Personalised recommendations