Understanding the role of<i>Iroquoi</i>s homeobox transcription factor 5 (<i>IRX5</i>) in cardiac function: getting to the (human) heart of the matter

This editorial refers to ‘Human model of IRX5 mutations reveals key role for this transcription factor in ventricular conduction’ by Z.R. Al Sayed et al., pp. 2092–2107. The heart structure and function are meticulously controlled by a network of transcription factors (TFs) which ensure appropriate spatial–temporal patterns of gene expression. TFs are classified on the basis of their molecular structure and interaction properties. Of the 100 s of cardiac TFs found in the heart, the Iroquois homeobox (Irx) TF family has been of particular interest because of their prominent roles in regulating and fine-tuning the heart development as well as cardiac electrical patterning.1 All six Irx genes (Irx1–6) are strategically expressed in ventricles with unique and often overlapping relationships, which can both vary during heart development and be modulated in disease states. Animal studies, especially in mice, have revealed that, while Irx2 appears to be dispensable for heart development and function, Irx3 is needed for the rapid and orderly spread of ventricular depolarization by regulating the connexin-dependent coupling between cardiomyocytes in the ventricular conduction system and its postnatal development.2,3 On the other hand, Irx4 defines the ventricular properties by activating ventricular gene expression and suppressing atrial genes while Irx5 ensures normal transmural gradients of ventricular repolarization by controlling the levels of the fast transient outward K+ current (Ito).4 The role of Irx1 and Irx6 in the heart remains unknown. Complexities arising from time-dependent overlapping expression and convergent molecular interactions between Irx TFs are also beginning to emerge using double knockout mice. For example, the ablation of both Irx3 and Irx5 results in developmental structural defects as well as impaired electrical conduction between atria to ventricles.5 Because cardiac arrhythmias are linked to the majority of sudden deaths associated with cardiovascular disease, and because about 20% of all sudden cardiac deaths are associated with inherited disorders, previous mouse studies have led investigators to explore the role of Irx TFs in congenital heart defects as well as the pathogenesis of heart disease and cardiac arrhythmias. For example, mouse studies predicted that Irx5 mutations might contribute to Brudaga syndrome, via Ito-dependent changes in repolarization dispersion.4 However, the implication of IRX5 in the human heart was unknown until loss-of-function mutations of IRX5 were found to cause Hamamy syndrome, an autosomal recessive condition associated with severe bone-related craniofacial defects combined with congenital heart defects and cardiac conduction disturbances.6 In this issue of Cardiovascular Research, Al-Sayed et al. report on the role of IRX5 in human hearts by elegantly combining transcriptome analyses of adult human hearts with studies using cardiomyocytes generated from human-induced pluripotent stem cells (hiPS-CMs) that were derived from two Hamamy syndrome patients and several normal controls.7 After establishing the presence of an expression gradient of IRX5 across the ventricular walls of the human heart, which correlated with the corresponding gradients of the sodium channel NaV1.5 and gap junction connexin 40 (CX40), the authors demonstrated that hiPS-CMs from both Hamamy patients had reduced expression of SCN5A (encoding NaV1.5) and GJA5 (encoding CX40) compared to control hiPS-CMs. Using an impressive technique for in silico injection of background K+ currents (i.e., IK1), which are known to be largely absent in hiPS-CMs, the authors then established that Hamamy hiPS-CMs have reduced voltage-dependent Na+ current (INa) and slowed depolarization rate during evoked action potentials. Taken together, these molecular and cellular changes can explain the reduced conduction velocity seen in Hamamy patients because the spread of depolarization depends both on the rate of depolarization within each cardiomyocyte, which in turn is influenced by the amount of INa and the spread of the depolarization between cardiomyocytes. The authors further examined complex molecular mechanisms whereby IRX5 controls SCN5A transcription and thereby cardiac conduction. These studies are important because many details of how IRX TFs control target genes remain poorly understood, despite their known impact on tissue development, cellular function, and physiology. Neither IRX5 missense mutations (i.e., A150P and N166K), seen in the two Hamamy patients, altered nuclear localization or protein degradation of IRX5, which was expected from the previous work.6 Moreover, Al-Sayed and colleagues demonstrate that native IRX5 directly binds to a conserved IRX binding site in the promoter region of SCN5A as well as to the GJA5 promoter, and this binding ability is lost for Hamamy mutations. Since both the A150P and N166K mutations are located in 3-amino-acid-loop-extension (TALE) homeodomain of IRX5, these findings support the suggestion that the highly conserved, atypical homeodomain of IRX TFs is required for DNA binding capacity. Importantly, their molecular studies further demonstrate that IRX5 needs to form a complex with GATA4 in order to activate SCN5A promoter activity and expression. Since IRX5 was shown by the authors to be expressed in a gradient across the ventricular wall (high in the endomyocardium) while GATA4 expression shows no gradient, the authors conclude that IRX5 and GATA4 together establish the transmural NaV1.5 expression in the human ventricle as summarized in Figure 1. (A) Schematic illustration of endomyocardium (ENDO) to epimyocaridum (EPI) transmural gradients of IRX5 and NaV1.5 across the ventricular free walls.7 No heterogeneity of GATA4 expression was found. (B) IRX5 interacts with GATA4, thereby promoting SCN5A gene expression. (C) Two IRX5 mutations, A150P and N166K in Hamamy patients prevent it from binding to the promoter region of SCN5A. (Created with BioRender.com) Given the glaring difference in the effects between IRX5 missense mutations associated with Hamamy syndrome and Irx5 deletion in mice, a question considered in this study is whether IRX5 in humans and Irx5 in mice function differently in regulating cardiac genes, despite their 93% homology (100% for homeodomain and IRO box). This suggestion is certainly not unexpected from the evolutionary differences between human and mouse hearts, especially related to cardiac electrical properties. Such electrical differences between the species apparently dictate that distinct ion channels be expressed, and it appears that the primary differences between these species involve choosing from the many available K+ channel genes.8 The suggestion that IRX5 has been redeployed for altering depolarizing currents, rather than repolarizing currents, between species, provides important new information on evolutionary strategies for creating electrical diversity. However, the cardiac and craniofacial defects seen in Hamamy syndrome patients are strikingly comparable to those seen in mice when both Irx3 and Irx5 are eliminated. In particular, mouse embryos lacking both Irx3 and Irx5 die in early development with atrioventricular septal defects, whereas postnatal mice lacking Irx3 and Irx5 in cardiomyocytes have conduction disturbances (with reduced NaV1.5 and CX40 expression in the proximal ventricular conduction system).5 Also, mice lacking these genes in osteoblasts develop decreased mineralization in cranial bones.9 This interdependence between Irx3 and Irx5 opens up a possible involvement of IRX3 in the Hamamy syndrome. In this regard, it is not clear whether the conduction defect in Hamamy patients originated from reduced depolarization speeds through the working myocardium or defected electrical propagation within the ventricular conduction system. Parenthetically, it is also worth mentioning that in mouse Purkinje fibres lacking Irx3 have reduced INa (unpublished data by Kyoung-Han Kim, Robert A. Rose, and Peter H. Backx; available upon request). Moreover, the involvement of IRX3 in the Hamamy syndrome phenotype becomes more plausible by recalling that Irx5 can form heterodimers with Irx3 as well as Irx4, and previous Kcnd2-promoter studies have established that Irx4 can convert Irx5 from a transcriptional activator to a repressor. These observations highlight the contextual nature of Irx TFs.10 It seems reasonable to suggest that such subtleties need to be considered when interpreting potential differences in outcomes arising with missense mutations that interfere with DNA binding, as in Hamamy patients, or transcriptional complex formation, vs. mutations leading to a complete loss of the TF. Future studies will be required to test whether the effects of IRX5 missense mutations causing Hamamy syndrome rely on the presence of IRX3, possibly using mice with Hamamy missense mutations in Irx5 or manipulating IRX3 and IRX5 levels in hiPS-CM with or without Hamamy mutations. A major strength of the Al-Sayed et al. study is the combination of bioinformatics analyses of the transcriptome with the hiPS-CM model system to understand the pathophysiology of inherited heart diseases caused by mutations in regulatory genes, such as Irx TFs. This allowed the authors to focus their attention on functional candidate genes and thereafter design appropriate and effective cellular and molecular studies to uncover the mechanisms responsible for the Hamamy syndrome. This study provides an exemplary approach that should prove useful for future investigations into mechanisms of inherited disorders linked to regulatory factors. The approach taken by the authors is expected to be even more effective in the future as additional advancements in stem cell technology allow for the generation of mature chamber-specific hiPS-CMs and more appropriate multicellular preparations that would enable more direct testing of functional properties, such as CX40-mediated cardiac conduction. Conflict of interest: The authors have no conflict of interest to declare related to this article. For this work, K.-H.K. was supported by grants from the Heart and Stroke Foundation of Canada (G-18-0022213) and the Canadian Institutes of Health Research (PJT-173281). P.H.B. holds a Canada Research Chair in Cardiovascular Biology and a Project Grant (MOP 125950) from the Canadian Institutes of Health Research. Support was also provided by a John Evans Leader Award from the Canadian Foundation for Innovation to P.H.B. The opinions expressed in this article are not necessarily those of the Editors of Cardiovascular Research or of the European Society of Cardiology.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.