Amish nemaline myopathy and dilated cardiomyopathy caused by a homozygous contiguous gene deletion of TNNT1 and TNNI3 in a Mennonite child

Abstract

Amish nemaline myopathy (ANM) is a severe congenital form of NM, known to be fatal in early childhood due to pulmonary insufficiency. Homozygous mutations in TNNT1 were originally ascertained in an Older Amish community in 2000. To date, only five reports with six pathogenic variants in TNNT1 have been described in both Amish and non-Amish consanguineous families. Here, we describe a 16-month old female from a small Mennonite community from Mexico, presenting with congenital hypotonia and dilated cardiomyopathy, with a novel homozygous deletion of 19q13.42 of about 11 kb in size, encompassing TNNT1 and TNNI3. Cardiomyopathy has not been observed in association with ANM in previous reports and systolic function is usually normal in children in this form of NM. Conversely, homozygous mutations in TNNI3 have been associated with dilated cardiomyopathy. Our report underscores the consideration of contiguous gene deletion in children with ANM who present with congenital hypotonia and cardiomyopathy. The report also expands the known spectrum of non-Amish related ANM mutations to include homozygous multi-exonic TNNT1 deletion.

INTRODUCTION

Nemaline myopathy (NM) is a genetically heterogeneous disorder characterized by early onset muscle weakness and characteristic rod-like inclusions, or nemaline bodies in skeletal muscle fibers. There are at least 10 genes described to date that cause NM (ACTA1, NEB, TMP3, TMP2, TNNT1, CFL2, KBTBD13, KHLH40, KHL41, and LMOD3), with 6 currently accepted clinical subtypes established according to the age of onset of disease and severity of presentation. (Sanoudou and Beggs, 2001; Wallgren-Pettersson and Laing, 2000; Wallgren-Pettersson et al., 2011). Recently, TNNT3 has been described as another gene likely associated with autosomal recessive nemaline myopathy with distal arthrogryposis (Sandaradura et al., 2018). The majority of cases of NM are attributed to pathogenic variants in ACTA1 and NEB, but it is difficult to clinically distinguish between causative genes. There is substantial clinical overlap among the subtypes of NM with congenital or early childhood onset. However, Amish NM (ANM), a fatal form of NM has a unique and specific phenotype, defined by congenital hypotonia, tremors at birth with spontaneous resolution, failure to thrive, pectus carinatum, progressive contractures, and death usually by second year of life due to respiratory insufficiency (Johnston et al., 2000). Without mechanical ventilation and gastrostomy tube feedings, survival beyond that is indeterminate. ANM, also called NM type 5 (NEM5) was originally described in an Older Amish community with a homozygous pathogenic variant, NM_003283.5: c.538G>T; p.(Glu180*) in TNNT1. Pathogenic variants in TNNT1 cause complete loss of slow skeletal muscle troponin T (TnT), representing a class of sarcomeric thin filament disease (Amarasinghe et al., 2016; Jin et al., 2003; Johnston et al., 2000). TNNT1 is essential for striated muscle contraction; truncating variants have repeatedly been demonstrated to prevent the incorporation of TnT into the myofilament leading to progressive muscle degeneration and weakness (Amarasinghe et al., 2016; Jin et al., 2003; Johnston et al.; 2000; Wang et al., 2005). TNNT1 is preferentially expressed in slow-contracting type 1 fibers. TNNT1 mutations have been reported in a limited number of non-Amish populations. A splice site variant, NM_003283.4; c.309+1G>A, and an exon 14 deletion in compound heterozygous form was reported in a Dutch pedigree (van der Pol et al., 2014). A homozygous novel nonsense variant, NM_003283.5; c.323C>G (p.S108*), was described in a Hispanic male with ANM (Marra et al., 2015). More recently, seven unrelated Palestinian families with nine children were reported to have ANM caused by a novel homozygous rearrangement, NM_003283.5; c.574_577delinsTAGTCGTGT in TNNT1 (Abdulhaq et al., 2016). Another family was identified in a study of 45 children with neuromuscular disease with a homozygous variant, NM_003283.5; c.606_607insTAGTG (p. Leu203*) (Fattahi et al., 2017). Altogether, five reports with six pathogenic variants in TNNT1 have been described (Table 1).

Table 1:

Summary of mutations in TNNT1 reported in ANM

	Johnston et al., 2000	van der Pol et al., 2014	Marra et al., 2015	Abdulhaq et al., 2016	Fattahi et al., 2017	This report
Nucleotide substitution	c.538G>T	c.309+1G>A	c.323C>G	c.574_577delins TAGTCGTGT	c.606_607insT AGTG	-
Amino acid change	p.E180*	-	p.S108*	-	p. Leu203*	-
Exon deletion	-	exon 14	-	-	-	exons 1–9 of TNNT1 and exon 8 of TNNI3
Mutation	Homozygous	Compound heterozygous	Homozygous	Homozygous	Homozygous	Homozygous
Known to be consanguineous	Yes	No	Yes	7 Palestinian families (consanguineous and non- consanguineous)	Yes	No
Ethnic population	Amish	Dutch	Hispanic	Palestinian	Unknown	Mennonite
Dilated cardiomyopathy	-	-	-	-	-	Yes

Dilated cardiomyopathy (DCM), characterized by left ventricular dilation and reduced systolic function, is the most common form of cardiomyopathy and is predicted to be hereditary in a significant number of cases. Troponin I encoded by TNNI3 is a thin filament protein forming part of the troponin complex triad that is crucial for the interaction of thick and thin filaments for striated muscle contraction (Bhavsar et al, 1996). TNNI3 has been implicated in various forms of cardiomyopathy including hypertrophic cardiomyopathy 7 (OMIM #613690), familial restrictive cardiomyopathy 1 (OMIM #115210), dilated cardiomyopathy 1FF (OMIM #613286), and dilated cardiomyopathy 2A (OMIM #611880). TNNI3 has been reported to cause DCM in rare instances. It was first described with a homozygous missense variant in a family with DCM in two children of a distant consanguineous union who presented with DCM and progressive cardiac failure in early adulthood. Functional protein studies, histology, and variant segregation analysis were consistent with the disease (Murphy et al., 2004; Mogensen et al., 2015). The rarity of TNNI3 variants in both dominant and recessive DCM have made genotype-phenotype correlations challenging, but functional characterization supports the broad spectrum of cardiomyopathy observed in individuals with apparent pathogenic TNNI3 variants (Carballo et al., 2009; Mogensen et al., 2015).

Here, we describe a novel homozygous deletion of 19q13.42 of about 11 kb in size, encompassing exons 1–9 of the TNNT1 gene and exon 8 of TNNI3 in a female with congenital hypotonia and cardiomyopathy from a small Mennonite community in Mexico.

CLINICAL REPORT

The patient was the product of a full term pregnancy complicated by decreased fetal movements noted around 7 months gestation, requiring weekly stress tests. Delivery at full term was uncomplicated via normal spontaneous vaginal delivery. She weighed 2.75 kg (9 %ile) at birth and had length of 45.7 cm (1%ile). Neonatal hypotonia was recognized at birth, however no significant feeding or respiratory problems were noted. Hand tremors were recognized daily during the first week of life, and by 2–3 months of age, extended to her arms, legs, and jaw, and spontaneously resolved by 5 months of age. She was able to lift her arms and hands to face in infancy but was never able to roll over or sit unassisted. At 14 months, she presented with significant failure to thrive and a gastrostomy tube was placed. She continued to have recurrent emesis and was unable to tolerate enteral feeds. Failure to thrive was persistent and her growth parameters remained <3^rd percentile. An echocardiogram was completed at 14 months and she was found to have DCM. The left atrium and left ventricle were moderately dilated with left ventricular end diastolic dimension measuring 3.3 cm (z score = 4). There was moderate left ventricular systolic dysfunction (EF 33%). The dimensions of the left ventricle wall dimensions and interventricular septum were normal with no evidence of hypertrophy. The E to A wave ratio was normal suggesting diastolic function was normal. The coronary origins were normal and there were no shunting lesions. A 24-hour Holter monitor demonstrated no arrhythmias to account for the dilation or dysfunction. Muscle biopsy was completed at 14 months, which showed atrophic type 1 fibers and mild to moderate increase in endomysial fibrous connective tissue, indicative of congenital myopathy. Electron microscopy was not performed. On evaluation at 16 months, she was noted to be socially interactive, but had regressed in her motor skills and was unable to left arms or legs against gravity. She was receiving occupational and physical therapies by then without much improvement. There were no clinical signs of respiratory distress or pulmonary insufficiency at that time. She had significant and diffuse hypotonia with head lag. She had a myopathic face with tented upper lip, severe pectus carinatum, and bilateral elbow contractures. Deep tendon reflexes were absent at patellae. She was hospitalized for failure to thrive and received a jejunal tube for continuous feeds. She was discharged home with care coordinated at local hospitals.

She had a normal brain MRI and nerve conduction studies; EMG showed neuropathic motor unit remodeling and recruitment in areas of the left biceps, anterior tibialis and vastus medialis. Her prior laboratory work up included normal plasma amino acids, lactic acid, pyruvic acid, ammonia, acylcarnitine panel, and urine organic acids. Creatine kinase was low (36 U/L; normal reference range 60 – 305 U/L). She had normal high resolution karyotype study (46,XX), and normal dosage and sequencing for spinal muscular atrophy (2 copies SMN1, 2 copies SMN2).

She developed hypoventilation around 18 months and required supplemental oxygen by nasal cannula. At almost 19 months of age, she was reported to fall ill with fever and vomiting and was treated at home with antibiotics and morphine. She subsequently passed away during sleep. An autopsy for the cause of death was not performed.

METHODS

The study was performed in accordance with the ethical standards of the Institutional and Review Board of Baylor College of Medicine and informed consent was obtained from the family. Chromosomal microarray (single nucleotide polymorphism and oligo array) (CMA) and rapid trio Whole Exome Sequencing (WES) were sent for her evaluation. The patient was studied by V11.2 chromosomal microarray analysis designed by Baylor Genetics Laboratories and manufactured by Agilent Technology (Santa Clara, CA, USA). The 400k array targeted over 4,200 genes at the exon level and included 60,000 probes used for SNP analysis for the detection of uniparental disomy (UPD) and absence of heterozygosity (AOH). Further details are available at https://www.bcm.edu/geneticlabs/. Peripheral blood sample from the patient was submitted to Baylor Genetics. DNA was extracted from whole blood using the Puregene DNA Blood Kit (Gentra, Minneapolis, MN, USA) according to the manufacturer’s instructions. The procedures for DNA digestion, labeling, and hybridization for the oligo arrays were performed as described previously. Slides were scanned into image files using the Agilent G2565 Microarray Scanner. Scanned images were quantified using Agilent Feature Extraction software, then analyzed for copy-number changes using our in-house analysis package, as described previously (Boone et al., 2010; Wiszniewska et al., 2014).

Exome Sequencing and Analysis

Rapid WES was performed at Baylor Genetics as described (Yang et al., 2013; Yang et al., 2014). The mean depth of coverage was 154X, with 97.5% of the targeted regions (exonic regions of all nuclear genes plus ±5 base pairs of exon-intron boundaries) sequenced at 20 times and higher. The samples were concurrently analyzed by HumanExome-12 v1 array (Illumina) for quality control and for detecting copy number variations (CNVs), regions of absence of heterozygosity (AOH), and uniparental disomy (UPD). CNVs were also characterized using the normalization of exome read depths as previously described (Lalani et al., 1993; Lalani et al., 2016). Exome data were interpreted according to the American College of Medical Genetics and Genomics guidelines and variant interpretation guidelines of Baylor Genetics.

RESULTS

Both CMA and WES identified an 11 kb homozygous deletion of 19q13.42 including exons 1–9 of TNNT1 and exon 8, the last exon of TNNI3 (arr[GRCh37] 19q13.42(55652193_55663445)x3). TNNT1 is flanked byTNNI3 in the telomeric region of 19q13.42. Single nucleotide polymorphism (SNP) genotyping showed that both parents were heterozygous for the deletion (Supplementary figure 1). CMA also identified multiple regions of absence of heterozygosity, of about 31 Mb, suggesting shared ancestry. A diagnosis of ANM due to homozygous TNNT1 deletion and TNNI3-related dilated cardiomyopathy was made.

DISCUSSION

Here we report a homozygous deletion involving TNNT1 and the last exon of TNNI3 as part of 19q13.42 contiguous gene deletion, causing NM and DCM in a female from a small Mennonite community in Mexico. Our patient presented with most of the classic features associated with ANM including transient tremors in the jaw and limbs, severe pectus carinatum, progressive contractures, and profound, diffuse hypotonia, requiring respiratory support by 19 months of age. She did not have spinal rigidity with kyphosis or scoliosis nor severely limited hip abduction reported in some individuals (Abdulhaq et al., 2016). She passed away before second year of life, as the patients in previous studies. Since its first description in 2000, five reports of ANM with six pathogenic variants in TNNT1 have been described. None of them were reported to have cardiomyopathy. This report elucidates another variant related to ANM outside of the originally described Older Amish population and highlights novel pathogenic variants in TNNT1 and TNNI3 related to CNV.

Further review of the muscle biopsy results confirmed that histopathology findings were consistent with congenital myopathy. However, Gomori trichrome stain did not show inclusions typical of nemaline rods (Supplementary figure 2). Electron microscopy which would best reveal the presence of nemaline rods, was not performed. Histopathological findings of ANM have been reported in a limited number of patients to date. In a previous cohort, muscle biopsy of 7 individuals with confirmed, pathogenic TNNT1 mutations was performed and nemaline rods were detected in only 4/7 biopsies (Abdulhaq et al., 2016). Fiber type disproportion with small type I fibers and endomyseal fibrosis were seen, as in our patient. When nemaline rods were present, they were subtle and required close review without electron microscopy (Abdulhaq, et al. 2016). These data, along with our patient’s findings emphasize the utility of comprehensive genomic testing for precise diagnosis in severe presentation of congenital myopathy.

There have been only a few reports of DCM presenting in childhood in patients with NM (Gatayama et al., 2013; Ishibashi-Ueda et al., 1990). Cardiac function is usually normal in the pediatric age group; when cardiomyopathy is seen, it usually occurs in adulthood and is observed in the form of hypertrophic cardiomyopathy. A systematic literature review in 2015 (Finsterer and Stollberger, 2015) identified 35 adults and children with NM and cardiac disease. Only 9 of these patients presented with dilated cardiomyopathy (DCM); of these, only 4 had a childhood onset. To date, ANM has not been reported with cardiomyopathy or other cardiac disease. On the other hand, both dominant (Carballo et al., 2009) and recessive (Murphy et al., 2004) inheritance pattern have been implicated in DCM related to TNNI3. The gene encodes cardiac troponin I, one of the thin filament proteins and can cause a broad phenotypic spectrum of cardiac disease including DCM. The majority of reported mutations (85%) in TNNI3 have been identified in exons 7 and 8 (Wallgren-Pettersson et al., 2011; van den Wijngaard et al., 2011). In large public datasets including the Database of Genomic Variants (http://dgv.tcag.ca/dgv/app/home) and ExAC (http://exac.broadinstitute.org), there are no individuals reported with either heterozygous or homozygous loss of exon 8 of TNNI3. Based on these data and the fact that ANM has not previously been reported with cardiomyopathy, DCM observed in this patient is likely secondary to homozygous TNNI3 deletion flanking TNNT1. This novel homozygous contiguous gene deletion is an example of classic presentation of two disorders likely observed as a blended phenotype. The differential diagnosis of congenital myopathies is very broad and there is significant phenotypic overlap among different types.

Nemaline myopathy can usually be distinguished with identification of nemaline bodies on muscle biopsy, however with advances in genetic testing and ubiquitous use of exome sequencing, invasive procedures such as muscle biopsy are utilized less often for diagnostic reasons. Histology of AMN and other congenital myopathies can also be non-specific, as in our patient, further highlighting the benefits of molecular diagnosis. Even when NM can be distinguished, significant overlap among the subtypes still exists, especially the severe congenital (neonatal) NM, ANM, and intermediate congenital NM. Genotype-phenotype correlations remain poorly defined because of this clinical overlap and the large proportion of cases of NM that do not have a known underlying genetic etiology. Previously, ANM was only reported to occur in the Older Amish, but there have now been 6 different pathogenic variants implicated in ANM across a variety of ethnicities and populations. During the course of review of this paper, Fox et al. described natural history of ANM associated with the Amish founder mutation (Fox et al., 2018). This report provides further supportive evidence that ANM should be suspected regardless of ethnicity in infants with severe form of NM.

Figure 1:

Contiguous homozygous deletion of exons 1–9 of TNNT1 and exon 8 of TNNI3 on 19q13.42 of about 11 kb is shown by CMA, SNP, and Integrative Genomics Viewer (IGV) plots. Panel A shows CMA plot for the entire chromosome 19; the homozygous deletion is indicated by a thin arrow. Panel B depicts ~4.012 Mb region of absence of heterozygosity (AOH) in the distal region of the long arm of chromosome 19, marked by a thick arrow. The homozygous deletion is within the AOH interval. Panel C shows the homozygous loss detected by the SNP array. The IGV plot is shown in panel D. Note the absence of reads for the 11 kb region encompassing exons 1–9 of TNNT1 and exon 8 of TNNI3.