This editorial refers to ‘Splice variants of Enigma homolog, differentially expressed during heart development, promote or prevent hypertrophy’ by T. Yamazaki et al., pp. 374–382, this issue.
The Enigma subfamily belongs to the PDZ- (for PSD-95, DLG, ZO-1) and LIM- (for LIN-11, Isl-1, MEC-3) encoding protein family. This family is composed of at least four different members: Enigma (also called PD-LIM7; LMP-1), Enigma homologue (ENH), ZASP/Cypher/Oracle, and LMP-4. All members of the PDZ–LIM Enigma family, including ENH, are cytoplasmic proteins that bind to the cytoskeleton through a direct interaction between their PDZ domain and α-actinin,1 thus localizing the protein to the Z-disk of cardiac myocytes.2 The LIM domain is a cysteine-rich domain, composed of two independent zinc-coordinated fingers, that has been proposed to participate in protein–protein interactions. ENH1 was first identified as an adaptor for protein kinase C (PKC) by using a yeast two-hybrid system. PKC binds to any of the three LIM domains of ENH1 in an isoform-specific manner.3 In neurons, ENH1 recruits PKCε to the N-type voltage-gated Ca2+ channel.4 The LIM2 motif of ENH1 has also been shown to bind protein kinase D1 (PKD1) to regulate the activity of the α1c subunit of the L-type voltage-gated Ca2+ channel.5
Interestingly, four spliced variants of the enh gene have been described, namely ENH1, -2, -3, and -4.1–3 Only ENH1 contains the LIM domains (see Figure 1 in Yamazaki et al.6) and, therefore, should be able to regulate protein kinases and Ca2+ channel activities. Yamazaki et al.6 demonstrate that ENH1 is expressed in the embryonic and neonatal heart but not in the adult, where it is replaced by ENH3 and ENH4. As shown for many other proteins, a shift towards the foetal phenotype and re-expression of ENH1 is observed after pressure overload in vivo. The authors also show that ENH1 is necessary to induce cardiac growth and expression of the foetal programme. Interestingly, overexpression of the LIM-deleted form, ENH4, which is normally present in adult cardiac myocytes, prevents agonist-induced hypertrophy of neonatal cardiac myocytes. It is proposed that the Lim-deleted isoforms have a dominant-negative effect on ENH1.
The signalling pathway linking ENH1 expression to induction of the foetal phenotype was not analysed by Yamazaki et al.6 However, based upon previously published data showing that the LIM domain anchors PKC and PKD and taking into account the well-described molecular pathways implicated in the hypertrophic effect of these kinases, it is tempting to propose that the LIM domains of ENH1 act as a new signalling platform that mediates the PKC and PKD hypertrophic pathways (Figure 1). In the normal adult heart, ENH4 would not recruit PKC and PKD, and, hence, the signalling pathway of these kinases would not be activated. Under stress, ENH1 would be expressed and scaffold PKC and PKD. Then, PKC would phosphorylate PKD1,7 which itself would phosphorylate class II histone deacetylases (HDACs), thereby resulting in their dissociation from MEF2 and their binding to the chaperone protein 14-3-3 that shuttles from the nucleus to the cytoplasm.8,9 This relieves the repressor effect of HDACs on MEF2, leading to activation of the hypertrophic process. HDACs are also phosphorylated by the Ca2+/CaM kinase IIδ, the major cardiac isoform.10–12 What the role, if any, of L-type Ca2+ channel activation by ENH1 is in this process is currently not known.
The LIM domain of ENH1 also binds Id (inhibitor of differentiation/DNA binding) proteins that are natural inhibitors of basic helix–loop–helix transcription factors. In many cell types, differentiation is associated with inhibition of Id function, and it has been postulated that cytoplasmic factors, activated during differentiation, sequester Id proteins and prevent their import to the nucleus. In neurons, ENH1—but not the LIM-deleted isoforms—has been shown to sequester Id2 in the cytoplasm, thus preventing its repressor effect in the nucleus.13 This could be another mechanism by which the alternatively spliced ENH isoforms regulate cardiac phenotype.
In conclusion, the paper by Yamazaki et al. highlights the importance of the LIM domain of ENH1in the induction of cardiac hypertrophy. New elements are provided to explain how the stress signal may be submitted directly to the nucleus via the cytoskeleton. This paper opens an interesting new field of investigation, and additional efforts are now needed to decipher the mechanism linking this anchoring protein to agonists and stress induction of cardiac growth. How the splicing of the enh gene is regulated in the heart is another enigma.
Conflict of interest: none declared.