for over 40 years, our primary understanding of how eukaryotic gene transcription is regulated has centered on the promoter concept. In this model, transcriptional activators bind to consensus sequences in the proximal upstream regulatory region of a gene, altering chromatin structure to allow recruitment of the RNA polymerase complex to a promoter sequence, which then creates a complementary transcript of the negative strand (Fig. 1A).
However, many genes in eukaryotes exist in gene clusters formed via gene duplication, and a key question centers on how coordinated regulation of expression of genes within these clusters occurs. Specifically, gene isoform switching within a given cell or tissue in response to developmental, physiological, or pathological stimuli often requires sequential temporal and/or spatial expression of adjacent genes; thus, expression of two or more genes must be coordinated, which requires an additional level of integration. Locus control regions (LCRs), or groups of regulatory elements upstream of the gene cluster that regulate the sequential activation of the genes in the cluster, represent one mechanism by which this occurs (Fig. 1B). LCRs have been identified for several key gene clusters, including the globin and interleukin genes (5).
Rinaldi et al. (14) provide further evidence for another mechanism regulating isoform switching in a set of clustered genes, the striated muscle myosin heavy chain (MHC) genes. Eight MHC genes are expressed in cardiac and skeletal muscle and are clustered in two groups: the predominantly cardiac-expressed β- and α-MHC genes are found in this order on chromosome 14 in the mouse and humans and chromosome 15 in the rat, while six skeletal-specific isoforms are found clustered on chromosome 10 in the rat, chromosome 11 in the mouse, and chromosome 17 in humans (18). Differential expression of these isoforms in both heart and skeletal muscle occurs in response to a number of developmental, neural, hormonal, and mechanical stimuli, but the mechanisms coordinating this transcriptional switching have remained obscure.
In previous work (7), this group demonstrated that switches between the predominantly cardiac-expressed β- and α-MHC genes occur via an intergenic bidirectional promoter (Fig. 1C) that simultaneously produces a sense α-transcript, as well as an antisense β-transcript that decreases β-MHC expression. Previous investigators had identified the presence of antisense MHC transcripts in the heart (4, 10–12), but it was not clear what, if any, function these had. The paper by Haddad et al. (7) demonstrated that increased activity of this bidirectional promoter increases expression of α- while decreasing expression of β-MHC by increasing the sense and antisense transcripts for these loci, respectively. These authors have subsequently shown that this mechanism contributes to the α- and β-MHC switching occurring in the heart in response to a wide variety of physiological and pathological conditions, including development (9), hypo- and hyperthyroidism (6, 7, 9), pressure overload (8), and diabetes (6, 7).
The present work extends this work by examining whether a bidirectional intergenic sense/antisense promoter exists between the IIx and IIb skeletal muscle MHC genes. Three “fast” MHC isoforms, IIa, IIx and IIb, are expressed in adult skeletal muscle in the rodent. Switching between these three MHC isoforms can occur in response to changes in activity levels, with increased activity favoring a switch toward type IIa and reduced activity favoring a switch toward type IIb; these shifts always occur sequentially through the three isoforms in the order IIa-IIx-IIb, which is the order of these genes in the cluster. In a previous paper, this group showed that an intergenic bidirectional promoter exists between the IIa and IIx MHC genes that inactivates IIa and activates IIx MHC expression during conditions of muscle atrophy in response to decreased usage (13).
The present paper expands upon this previous work in two ways: it examines whether a second bidirectional intergenic promoter exists between the IIx and IIb MHC genes, and if so, whether it can account for the shift from IIb to IIx occurring in response to models of increased muscle activity. Using a model of muscle stimulation simulating a 4-day regimen of resistance exercise training in rats, Rinaldi et al. (14) show that levels of IIx pre-mRNA and mature mRNA are increased while levels of IIb pre-mRNA and mature mRNA are decreased. Moreover, using strand-specific primers, they demonstrate that an antisense RNA spanning all of the IIx MHC gene and extending into the 3′ end of the IIa MHC gene is present in control rats but is suppressed following resistance training. Sequence analysis identified an evolutionarily conserved region between the IIx and IIb MHC genes that contains several putative consensus binding sequences that may function as the sense/antisense promoter for producing these two transcripts (14).
Thus this work supports the hypothesis that intergenic bidirectional promoter activity is a strong contributor to regulation of the skeletal MHC gene cluster as a whole in response to both increased and decreased muscle activity and may be another mechanism for regulating gene switching among clusters in general. Other examples of antisense RNA production or “opposite strand transcription” have been reported in other systems (17), including the hox genes (16), and thus, this may represent a fairly common mechanism for regulating expression of genes and/or switching between adjacent isoform genes.
However, several questions have yet to be addressed. First, the mechanisms by which the antisense RNA interferes with gene expression have yet to be elucidated. One possibility is that the antisense RNA base pairs with the sense transcript and interferes with pre-mRNA processing, export, or stability. A second possibility is that either the antisense RNA or the process of transcribing it interferes with transcription of the sense strand, either through interference of sense strand polymerase function and strand progression, or possibly through epigenetic mechanisms (14). Second, it is not yet clear whether this bidirectional sense/antisense mechanism is solely responsible for coordinating expression of genes in the MHC cluster. To date, no LCR has been identified for either the cardiac or skeletal MHC cluster, but this does not rule out the possibility that locus control may exist for either cluster. Third, this model needs to be reconciled with the frequent coexpression of IIx and IIb MHC in single fibers of control rats (15). Because muscle fibers are multinucleated, it may be that different myonuclei within the muscle fiber syncytium activate the bidirectional IIx/IIb promoter at different levels. Fourth, it is not yet clear whether other muscle contractile or metabolic gene isoforms that exist in gene clusters are regulated in a similar manner during muscle adaptation; a prevailing question in muscle adaptation research is how expression of dozens if not hundreds of different genes and gene isoforms are coordinately regulated during states of plasticity to produce defined fiber types once a new steady-state has been reached. Finally, the role of previously identified consensus elements in the MHC gene proximal promoter regions (1–3) in regulating dual expression of the sense and antisense transcripts has yet to be addressed.
Nevertheless, the exciting findings of Rinaldi et al. confirm that the bidirectional intergenic antisense mechanism is a potent means for regulating gene expression and isoform switching in the MHC family. This novel mechanism joins micro-RNA and small inhibitory RNA as new and exciting mechanisms by which RNA can modulate gene expression.
- Copyright © 2008 the American Physiological Society