Towards an understanding of the genetic aspects of hypoplastic left heart syndrome (HLHS)

Abstract

Hypoplastic left heart syndrome (HLHS) is a rare congenital heart defect in which the left side structures of the heart are severely underdeveloped and major, repeated surgeries throughout childhood are required for survival. Although the specific features can vary considerably, a series of abnormalities involving a diminutive left ventricle, a narrow ascending aorta and atretic or stenotic mitral and aortic valves, are always noted. Based on inheritance of HLHS in families, a strong genetic component to HLHS has been suggested. Despite this, genetic studies have failed to reveal mutated genes that cause HLHS. A novel approach was taken in this study where it was hypothesised that there might be distinct anatomical subgroups of HLHS that differed from the current clinical classification. Thus, performing analyses on specific and homogeneous subgroups might be more amenable to understand which defects in HLHS are potentially primary and which secondary, and ultimately finding the causative gene(s) for HLHS. To accomplish this, archival formalinfixed HLHS hearts, obtained from Birmingham Children's Hospital, were carefully phenotyped and re-classified based on their ventricular morphology. It was suggested that genomic analysis of these hearts might reveal causative mutations for HLHS, but it is also recognised that formalin fixation causes damage to DNA. Consequently, a method was developed for extracting high quality DNA from formalin fixed tissue that could be used for PCR or Next Generation Sequencing (NGS). Through comparison of NGS whole exome sequencing data from formalin-fixed DNA with blood derived DNA, it was possible to determine bioinformatics processes to maximise the detection of variants with high specificity and moderate sensitivity from formalinfixed tissue. Finally, as a prelude to future genomic studies, the expression patterns of genes previously implicated in HLHS genes were analysed, focussing on the structures affected by HLHS and considering how they might link to the previously identified HLHS phenotypic subgroups. Although some HLHSimplicated genes, including Gja1 and Sap130, showed very broad expression patterns, and thus it was hard to associate them with the HLHS phenotypes, other genes showed specific expression within the structures affected by HLHS. It was speculated that Notch3, Hand1, Foxc2, and Pcdha9 would be more likely to correlate with, based on their expression patterns, the abnormalities associated with various subgroups of HLHS. Together, these data suggest that the broad clinical spectrum of HLHS is made up of distinctive anatomical subgroups. DNA obtained from archival hearts should not be used for primary variant detection but could be used for corroboration of findings for example from family studies. Some postulated genes appear to be expressed in structures affected in HLHS and are more interesting candidates for the malformations, whilst others are more broadly expressed and may represent modifier genes or false positive findings.