Background

Congestive heart failure (CHF) is an increasing medical problem
that is disabling and costly, and CHF currently affects more than 5 million
Americans. A leading cause of CHF is cardiomyopathy, a disorder with a highly
heritable component. Cardiomyopathies consist of nonischemic dilated (DCM),
hypertrophic (HCM), arrythmogenic, restrictive, and noncompaction types. DCM is
characterized by four chamber enlargement; HCM is characterized by thickening
of the left ventricular wall (Figure 1). DCM and HCM are primarily inherited in
an autosomal dominant pattern with variable expressivity and penetrance
(McNally et al., 2013). Genetic studies of DCM have identified more than 50
responsible genes that encode proteins with a wide variety of cellular
structures including the sarcomere, the cytoskeleton, the nuclear envelope, the
sarcolemma, and the intercellular junction (Figure 2). The genetic
heterogeneity of DCM suggests that several variants may contribute to disease
onset and DCM may have an oligogenic pathogenesis (Puckelwartz and McNally,
2017). Knowledge of the genetic underpinnings of disease provides prognostic
information and can guide clinical decision-making.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

Clinical genetic testing is indicated in
familial DCM and is useful in risk stratification and prognostic assessment. It
employs gene panels for diagnosis, and typical panels include 20-100 genes. The
American College of Medical Genetics Laboratory Quality Assurance Committee
recommends a framework that includes six categories of variants which describe
the likelihood of variant pathogenicity (e.g. benign, likely benign, variant of
unknown significance, likely pathogenic, and pathogenic) (Richards et al.,
2008). The categories are largely based on variant frequency in the population
and the predicted effect on protein structure. While genetic testing can be useful,
it is not sensitive enough. It is reported the sensitivity of genetic testing
in DCM is about 30-40% (Hershberger et al., 2013). Patients with unrevealing
genetic testing currently undergo retesting as genetic panels are expanded, and
the price of whole genome sequencing (WGS) is shrinking; thus, a single
instance of WGS can provide sequence data on both unknown and previously
unreported genes and may be more practical (Golbus et al., 2014). We assess the
viability of whole genome sequencing (WGS) to identify potentially pathogenic cardiomyopathy-associated
genetic variants in established and novel genes.

Co-segregation of a putative mutation with
a trait or phenotype in a multiplex (multiple affected members),
multigenerational family tree is the best genetic evidence supporting causality
(McNally and George, 2015). Here we also analyze genetic segregation patterns of
four genetic variants (two RBM20
variants, one BAG3 variant, one SCN5A variant) in four families with
DCM.  RBM20 encodes a putative
RNA-binding protein and is associated with a clinically aggressive form of DCM
(i.e. young age at diagnosis, end-stage heart failure, and high mortality)
(Brauch et al., 2009). It is responsible for normal splicing of many cardiac
genes and has been shown to be crucial for the formation of a subset of
circular RNAs that originate from the I-band of the TTN gene (Khan et al., 2016). TTN
encodes the large sarcomere protein, titin, and truncations in TTN accounts for 25% of familial DCM (Herman
et al., 2012). BAG3 encodes Bcl2-associated athanogene 3 protein,
a member of BAG family of co-chaperones that interacts with the ATPase domain
of the heat shock protein (Hsp) 70. It is induced by stressful stimuli and modulates
apoptosis, development, cytoskeleton organization and autophagy, mediating cell
adaptive response to stressful stimuli (Rosati et al., 2011). BAG3 is required for myofibrillar integrity and
have been linked to myopathy and cardiomyopathy in humans (Fang et al., 2017). SCN5A
encodes a voltage-gated Na channel subunit implicated in long-QT syndrome.

Missense mutations of SCN5A has also been described in familial DCM with a
higher risk for arrhythmias (McNair et al., 2010; Mann et al., 2012).

 

Methods

WGS was conducted on 91 families (99 individuals) with
cardiomyopathy. In 46/91 (50%) families, there was a prior history of
unrevealing panel-based clinical genetic testing. WGS was performed using
Illumina next generation sequencing. Variants were called and annotated using
MegaSeq and MegaSeq2. MegaSeq2 refines the number of candidate variants from ~4
million to

   Out of the candidate
variants, four variants were picked from four families with dilated
cardiomyopathy to study co-segregation patterns (Table 1). Variant sites were amplified
via PCR in family members including affected members and sequenced. Sequencing
results were compared.

 

Results

MegaSeq2 was applied to 99 cardiomyopathy genomes and in 45/91
families (49%), a potentially pathogenic or likely pathogenic variant was
identified. Here we describe identification and segregation analysis in four
families.

In family DCM-AAM (Figure 3), the proband,
DCM-AAM01, was diagnosed with DCM and muscular dystrophy (MD) at age 16. One of
her sisters and her mother have also been diagnosed with DCM and MD. Her
maternal grandmother was diagnosed with DCM. The family reported they were
Lamin negative in clinical genetic testing. In this family, WGS identified an RBM20
variant, Arg636Gln, caused by an A to G change on chromosome 10. In silico pathogenicity
algorithms, GERP and polyphen-2, predict that this variant is highly conserved
and likely pathogenic.  The global
population frequency of this variant in ExAC, a database of ~60,000 exomes, is
zero. The variant, denoted A/G on the pedigree, segregates with disease in all
affected family members available for testing.

In family DCM-AAJ (Figure 4), the proband,
DCM-AAJ01, has LMNA-negative familial DCM. In the pedigree, 15 family members
are affected by familial DCM. Four of the 15 affected members were available
for testing. In this family, WGS identified an RBM20 variant, Arg634Gln,
caused by an A to G change on chromosome 10. 
In silico pathogenicity algorithms, GERP and polyphen-2, predict
that this variant is highly conserved and likely pathogenic.  The global population frequency of this
variant in ExAC, a database of ~60,000 exomes, is zero.   The variant, denoted A/G on the pedigree,
segregates with disease in all affected family members available for testing.

    In family DCM-AAP
(Figure 5), the proband, DCM-AAP03, was diagnosed with DCM, intermittent PR
prolongation, and bradycardia. His mother was diagnosed with arrhythmia, atrial
fibrillation, and has a pacemaker. His maternal uncle suffered from sudden
cardiac death (SCD). His maternal aunt was diagnosed with DCM and uses a
pacemaker. A distant cousin on his maternal side suffered from SCD. WGS
identified an SCN5A premature stop
variant, Arg1316*, caused by an A to G change on chromosome 3. The in silico
pathogenicity algorithm, GERP, predicts that this variant is conserved and
potentially pathogenic. The global population frequency of this variant in
ExAC, a database of ~60,000 exomes, is zero. The variant, denoted A/G on the
pedigree, segregates with disease in all family affected members available for
testing and in unaffected DCM-AAP02. 

    In family DCM I (Figure
6), multiple family members were diagnosed with DCM and/or myotonic dystrophy.

WGS identified an early termination variant in BAG3, Gln244*, caused by
an C to T change on chromosome 10. The in silico pathogenicity
algorithm, GERP, predicts that this variant is highly conserved and likely
pathogenic.  The global population
frequency of this variant in ExAC, a database of ~60,000 exomes, is zero. The
variant, denoted C/T on the pedigree, segregates with cardiomyopathy in all but
one affected family member. The affected family member who is negative for the
variant, DCM-I7, is affected by myotonic dystrophy without cardiomyopathy.

Other affected family members who are positive for the variant are affected by
DCM, with DCM-I5, exhibiting both DCM and myotonic dystrophy.

 

Discussion

WGS was able to identify potentially pathogenic variants in nearly
half (45/91) of the cohort. While some of the variants identified were in novel
genes, the majority of pathogenic variants were in established cardiomyopathy
genes despite previous panel-based testing. Given the genes being studied in
the segregation analyses have been suggested in previous studies to be
implicated in DCM pathogenesis, the positive segregation results in the four
families support their likely pathogenic role in DCM. The results also support
the viability of WGS in identifying potentially pathogenic
cardiomyopathy-associated genetic variants in both established and novel genes.

Because of the limited sensitivity of clinical genetic testing in
cardiomyopathy, the lowering cost of WGS, and WGS analyses becoming more
sophisticated, WGS as a comprehensive analysis is practical in allowing for
re-evaluation of variation as additional genetic information becomes available.

Most hereditary cardiovascular diseases
show incomplete penetrance and variable expressivity, meaning not all individuals
with a mutation would exhibit the presumed phenotype (incomplete penetrance)
and the disease severity would be different among individuals inheriting the
same mutation (variable expressivity). Epigenetics, genetic modifiers, and
environmental factors can contribute to variable expressivity. These two
factors can confound co-segregation studies. 
Moreover, in large families, there may be more than one rare pathogenic
variant present, also confounding the segregation approach (McNally and George,
2015).

Although co-segregation of a variant in a
multiplex, mutigenerational family strongly supports causality, it does not
establish causality. Experimental studies in
vitro and in vivo, although
potentially time and labor intensive, would help determine biological and
functional effects. Further analysis into the non-coding regions that WGS
provides may yield important clues in gene expression and transcription
modulation and help piece together the larger picture of cardiovascular
genetics.