Combination of Whole Genome Sequencing, Linkage, and Functional Studies Implicates a Missense Mutation in Titin as a Cause of Autosomal Dominant Cardiomyopathy With Features of Left Ventricular Noncompaction

Abstract

Background: High throughput next-generation sequencing techniques have made whole genome sequencing accessible in clinical practice; however, the abundance of variation in the human genomes makes the identification of a disease-causing mutation on a background of benign rare variants challenging.

Methods and results: Here we combine whole genome sequencing with linkage analysis in a 3-generation family affected by cardiomyopathy with features of autosomal dominant left ventricular noncompaction cardiomyopathy. A missense mutation in the giant protein titin is the only plausible disease-causing variant that segregates with disease among the 7 surviving affected individuals, with interrogation of the entire genome excluding other potential causes. This A178D missense mutation, affecting a conserved residue in the second immunoglobulin-like domain of titin, was introduced in a bacterially expressed recombinant protein fragment and biophysically characterized in comparison to its wild-type counterpart. Multiple experiments, including size exclusion chromatography, small-angle x ray scattering, and circular dichroism spectroscopy suggest partial unfolding and domain destabilization in the presence of the mutation. Moreover, binding experiments in mammalian cells show that the mutation markedly impairs binding to the titin ligand telethonin.

Conclusions: Here we present genetic and functional evidence implicating the novel A178D missense mutation in titin as the cause of a highly penetrant familial cardiomyopathy with features of left ventricular noncompaction. This expands the spectrum of titin's roles in cardiomyopathies. It furthermore highlights that rare titin missense variants, currently often ignored or left uninterpreted, should be considered to be relevant for cardiomyopathies and can be identified by the approach presented here.

Keywords: cardiomyopathy; left ventricular noncompaction; missense mutation; telethonin; titin; whole genome sequencing.

Figure 1.

A, Pedigree of the family; males depicted as squares; females, circles; slanted symbols, deceased individuals. Clinically affected individuals are marked in gray, unaffected are shown in white, and ? means unclassified clinical status. The presence of the TTN p.A178D mutation is indicated (+ indicates present; −, absent; ND, not determined.) Individuals selected for whole genome sequencing (WGS) are marked with thicker symbols (III-1 and III-4). B, Echocardiogram images showing the characteristic spongy appearance of noncompaction in individual II-2 with and without contrast. C, Echocardiogram image from individual II-4 showing significant dilatation, but maintaining a thickened myocardium and preserved ejection fraction.

Figure 2.

A, Position of the TTN p. A178D on a structural model (pdb: 1YA5) of the titin Z1Z2 domains (purple) in complex with telethonin (pink). B, Close-up of the site of mutation. The red discs show van der Waals overlaps or steric clashing that A178D is predicted to cause with valine127 and proline133. The figures of the crystal structure were generated by Pymol (http://www.pymol.org).

Figure 3.

A, Circular dichroism (CD) spectroscopy of purified titin Z1Z2 fragments (WT solid line and A178D dashed line). B, Size exclusion chromatography for titin Z1Z2 fragments (WT solid line and A178D dashed line). Z1Z2 WT elutes as monomeric protein ($), whereas peaks corresponding to dimer (*) and higher molecular aggregates (#) are observed for Z1Z2 A178D.

Figure 4.

Destabilization of the titin Z1Z2 fragment in the presence of the A178D mutation. A, Denaturing gel-electrophoresis of purified titin Z1Z2 fragments (WT and A178D) expressed in Escherichia coli. The WT fragment is detected as a single band of 23 kDa (arrow). Only for Z1Z2 A178D, a degradation product (arrowhead) is observed. The position of marker proteins and their size (in kD) is indicated. B, Titin Z1Z2 protein fragments (left, WT; right, A178D) were incubated with protease thermolysin for the length indicated. Control: titin Z1Z2 protein sample without thermolysin. A stable degradation fragment of ≈15 kDa is observed for the mutant titin Z1Z2. The position of marker proteins and their size (in kD) is indicated. C, Decreased stability of titin Z1Z2 A178D in neonatal rat cardiomyocytes (NRC). NRC were infected in duplicates with adenoviral particles for hemagglutinin (HA)-tagged titin Z1Z2 (WT or A178D, MOI 5). Infection with parental empty vector (GFP) and noninfected cells (NI) served as controls. Steady-state titin fragment protein amount was assayed by Western blotting for the HA tag. Probing for humanized Renilla reniformis green fluorescent protein (hrGFP) served as infection control, and probing for endogenous GAPDH served as loading control. Despite equal infection rates (for confirmatory control experiments see Figure VIA in the Data Supplement), less titin A178D protein fragment was detected, indicating reduced stability in NRC. D, Localization of titin Z1Z2 in NRC. Cells were transfected with constructs coding for HA-tagged titin Z1Z2 WT (top, first row) or titin Z1Z2 A178D (bottom, first row) mutant protein fragment and counterstained for endogenous titin with T12 antibody (middle row, Z-disk proximal epitope, but not recognizing the transfected titin Z1Z2 protein fragment). Merged images are shown in the third row; HA, in red; endogenous titin, in green. Scale bar represents 10 μm.

Figure 5.

Functional implications of the titin Z1Z2 A178D mutation. A, Semiquantitative Glutathione S-transferase (GST) pulldown assays using telethonin fragments fused to GST (left aa 1–90, right full length) and titin Z1Z2 fragments (WT and A178D as indicated) as GFP fusions expressed in COS-1 cells. Bound titin-GFP fragments are detected by Western blotting (top row); input lysate controls are shown in the second row. Pulled down GST-telethonin fragments are shown in row three, as well as lysate controls (bottom row). B, Quantification of GST pulldown experiments from panel A visualizes reduced binding of titin Z1Z2 to telethonin in the presence of the A178D mutation; values are expressed as bound protein relative to lysate, with the first WT experiment set to 100% (n=2 per group, values expressed as mean with standard deviation for error bars; one representative experiment of 3 independent ones is shown). Because of the semiquantitative nature of the experiments, no statistical test was performed. C, Förster Resonance Energy Transfer (FRET) experiments using COS-1 cells cotransfected with telethonin (aa 1–90) fused to yellow fluorescent protein (YFP; first row) and titin Z1Zr3 fused to cyan fluorescent protein (CFP; second row). Images pre and post bleach are shown. FRET ratios are shown in the third row. Insets show magnification of indicated area. Scale bar represents 10 μm. D, Quantification of FRET efficiency for titin Z1Zr3 WT/A178D and telethonin pairs. The A178D mutation reduces the FRET efficiency from ≈0.15 to 0.01, indicating a dramatic loss of binding ability (WT n=20 and A178D n=26 cells; *P<0.0001 unpaired Student’s t test).