Role of Antisense RNA in Coordinating Cardiac Myosin Heavy Chain Gene Switching*

A novel mechanism of regulation of cardiac α and β myosin heavy chain gene by naturally occurring antisense transcription was elucidated via pre-mRNA analysis. Herein, we report the expression of an antisense β myosin heavy chain RNA in the normal rodent myocardium. The pattern of expression of the antisense βMHC RNA (β RNA) under altered thyroid state and in diabetes directly correlates with that of the α pre-mRNA/mRNA, whereas it negatively correlates with the β mRNA expression. Rapid amplification of the 5′ end shows that this antisense transcript originates 2 kb downstream of the β gene, and it is transcribed across the entire β gene from the opposite strand. Our results demonstrate that the β-α myosin heavy chain intergenic DNA possesses a bidirectional transcriptional activity, one direction transcribing the α gene, and the opposite direction transcribing the antisense β RNA. This process turns on the α expression, and it simultaneously turns off that of the β and thus coordinates α and β expression in an opposite fashion. Comparative analyses of the intergenic DNA sequence across five mammalian species revealed a conserved region that is proposed to be a common regulatory region for the α and antisense β promoter. This finding unravels the mechanism of cardiac α-β gene switching and implicates the role of cardiac myosin gene organization with their function.

Cardiac muscle expresses two myosin heavy chain (MHC) 1 isoforms designated as ␣ (high ATPase) and ␤ (low ATPase) that are encoded by two distinct genes located in close proximity on the same chromosome (1)(2)(3)(4). The MHC is the molecular motor driving muscle contraction, and its phenotypic composition regulates the intrinsic contraction properties of the heart (5,6). Cardiac MHC isoform expression is developmentally regulated (5), and it can change totally in either direction under certain pathophysiological states (7)(8)(9). For example during the first 3 weeks of postnatal life of rodents there is a complete switch from a predominant ␤MHC expression at birth (Ͼ90%) to a predominant ␣MHC expression at 3 weeks of age (Ͼ95%). Throughout adult life, the ␣MHC expression predominates in a normal rodent heart, with ␤MHC expression gradually increasing as the animal gets older. At any time during life, the pattern of MHC expression can be altered. Hypothyroidism and diabetes are associated with a switch in the cardiac MHC gene expression from a predominant ␣MHC to a predominant ␤MHC. In contrast, thyroid hormones treatment increases the ␣MHC expression while down-regulating the ␤MHC expression. The exact molecular mechanisms causing this tightly coordinated regulation of these two genes remains unclear. Thyroid hormone has been shown to be a major regulator of MHC gene expression, and its regulation is thought to occur mainly via transcriptional processes regulating each gene independently in a well coordinated fashion (8). Several thyroid responsive elements have been located on the promoter of the ␣MHC gene, whereas the localized action of thyroid hormone on the ␤MHC promoter remains poorly defined (10).
In an effort to further characterize the mode of regulation of cardiac MHC expression in response to hypothyroidism and to diabetes, we first examined the MHC primary transcript expression in the myocardium under control, as well as in the diabetic and hypothyroid states, in which both states are associated with a rapid shift from predominant ␣MHC to predominant ␤MHC mRNA and protein expression. Changes in the primary transcript level can be used as an indicative measure of changes in transcriptional activity of a gene (11)(12)(13). Primary transcript expression was determined by RT-PCR techniques using random primers for the reverse transcription reaction and intronic primers for the PCR. Using intronic primers ensure targeting only nascent primary transcripts as opposed to processed mature mRNA. Thus we used this approach to monitor primary transcripts expression and correlate these with the corresponding MHC mRNA expression. Our results 2 showed that the pattern of ␣MHC mRNA expression directly correlates with its nascent pre-mRNA and agrees with the concept that this gene is regulated at the transcriptional level in response to altered thyroid state and in diabetes. In contrast, the ␤MHC pre-mRNA expression pattern and its relationship to the mRNA were paradoxical. For example, in the hypothyroid heart, the ␤MHC mRNA increased significantly severalfold whereas the ␤MHC pre-mRNA did not change. Also, diabetic heart RNA analyses showed that although the ␤MHC mRNA increased severalfold, the ␤ pre-mRNA levels actually were decreased. These results triggered us to design primers specific to the strands in question. Using random primers for cDNA synthesis can target all RNA populations and may not be specific to the sense strand in question. When RT primers are designed to specifically target either the sense or the antisense primary transcript, then the coupled PCR will amplify only the products of a specific strand. When this approach was followed, it was found that the normal rodent heart expresses both sense and antisense ␤MHC transcript, and the effect of thyroid hormones and diabetes is more pronounced on the antisense reg-ulation. This antisense regulation is directly linked to the regulation of the mature ␤MHC mRNA transcript accumulation. Thus, the primary purpose of this study was to characterize the antisense ␤MHC RNA (AS ␤ RNA) and determine its potential significance in the mammalian hearts.

EXPERIMENTAL PROCEDURES
Animal Model-Adult female Sprague-Dawley rats (Ϸ150 -180-g body weight) were used for all experiments. Hypothyroidism was induced using daily propylthiouracil (PTU) treatment via intraperitoneal injections at a dose of 12 mg/kg body weight (10), whereas diabetes was induced with a single streptozotocin IV injection at 75-100 mg/kg (14). This study followed the NIH Animal Care Guidelines and was approved by the University of California, Irvine Animal Care and Use Committee.
RNA Analyses-Total RNA was extracted from frozen myocardial tissue using the Trireagent protocol (Molecular Research Center). Extracted RNA was DNase-treated using 1 unit of RQ1 RNase-free DNase (Promega) per g of total RNA and was incubated at 37°C for 30 min followed by a second extraction using Trireagent LS.
1 g of total RNA was reverse-transcribed using specific primers (Table I). After reverse transcription, PCR was carried out with specific PCR primers targeting either the pre-mRNA or mRNA (Table II). The cDNA dilutions were in the order of 1 to 5 for all pre-mRNA amplifications, and these were run for 29 cycles. For mRNA amplification, cDNA was diluted 50-fold, and PCR was carried out for 24 cycles. PCR products were separated by electrophoresis on agarose gels and stained with ethidium bromide. The ultraviolet light-induced fluorescence of stained DNA was captured on a photographic film, and band intensities were quantified using scanning densitometry and ImageQuant software (Molecular Dynamics).
Amplification of the ␤-␣MHC Intergenic DNA by PCR-To elucidate the origin of the antisense RNA, we needed some sequence information on the 3Ј flanking region of the ␤MHC gene. The intergenic DNA between ␣ and ␤MHC gene was amplified from rat genomic DNA by high fidelity PCR (Pfu Ultra; Stratagene) using ␤ I39 for the forward primer and ϩ420 ␣ for the reverse primer. The ␤ I39 sequence is from the last intron of the ␤MHC gene (Table I). The ϩ420 ␣ is complementary to 20 bp from the ␣MHC gene extending from ϩ400 to ϩ420 relative to the ␣MHC TSS. The PCR resulted in a ϳ5-kb PCR fragment that was purified by gel electrophoresis and extraction (using the Qiagen gel extraction kit). This intergenic fragment was ligated into the multicloning site of a pGEM-T cloning vector (Promega) using the supplier's recommendation. After ligation, DH5␣ bacteria (Invitrogen) were transformed, and several clones containing the insert were obtained.
The ϳ5-kb intergenic insert was fully sequenced in six different clones using the BigDye Terminator mix (ABI) and an automatic ABI 3700 sequencer (University of California, Irvine DNA Core facility).
Rapid Amplification of the cDNA 5Ј Ends (5Ј RACE)-Specific RT-PCR results indicate that the AS ␤ RNA originates in the middle of the intergenic region between the ␣ and ␤MHC genes. To characterize the 5Ј end of the AS ␤ RNA, we used the 5Ј RACE system according to the manufacturer's recommendations (Invitrogen). The RT primer sequence was 5Ј-CAGAATGGGTGAGGAGA-3Ј (located at position 2239 -2255 bp on the intergenic fragment GenBank TM accession number AY191158). The RT reaction was followed by two rounds of PCR using the nested gene-specific (GSP) reverse primers GSP-1, 5Ј-GATATG-AGCGCCGGAACAGCAGAG-3Ј; and GSP-2, 5Ј-AGACGGGGGATCC-AGGTAACAAAG-3Ј. The obtained PCR products were extracted from the gel (Qiagen) and directly sequenced using GSP-2 as the sequencing primer.
DNA Injection in the Myocardium and Reporter Gene Assay-The reporter gene assay approach was used to test the antisense promoter activity in vivo. The entire ␤-␣ intergenic sequence was ligated in the antisense orientation upstream of the Renilla luciferase (R-Luc) coding domain of the pRL null vector (Promega). We tested the promoter activity in vivo using direct plasmid transfer via an intramuscular injection (15). The DNA injection into the myocardium was performed via a subdiaphragmatic approach as described previously (16). Seven days post-plasmid injections, rats were euthanized, and tissue was processed as described previously (10) for reporter gene assays. The firefly luciferase and Renilla luciferase activities were determined using the dual luciferase kit from Promega according to the manufacturer's instruction. Reaction chemiluminescence was determined using an analytical luminometer (monolight 2010C). A myosin light chain 2 promoter fragment (a gift from Dr. Karyn Esser, University of Illinois, Chicago, IL) that is constitutively active in cardiac myocytes was used to correct for transfection efficiency. myosin light chain 2 drives the expression of firefly luciferase in pGL3 vector (17) and was co-injected with the test promoter in equimolar amounts.

Cardiac MHC RNA Expression in Hypothyroid State and in
Diabetes-Analyses of the rat cardiac MHC RNA species under normal control, hypothyroidism, and diabetes demonstrate the following observations. 1) In the normal control ventricle, simultaneous expression of the AS ␤ RNA, the sense ␤ pre-mRNA (S ␤ pre-mRNA), and the sense ␣ pre-mRNA (S ␣ pre-mRNA) is observed (Fig. 1A). In these same control ventricles, ␣ mRNA is almost exclusively expressed (Fig. 1A), with only small amounts (Ͻ5%) of ␤ mRNA that can be detected. 2) Hypothyroidism (PTU treatment) and diabetes (D) are associated with an increased expression of the S ␤ pre-mRNA and a dramatically decreased expression of both the AS ␤ RNA and S ␣ pre-mRNAs. These are associated with a shift from ␣ to ␤ predominance at the mature mRNA level (Fig. 1A). 3 a Abbreviations used are as in Table I. the above conditions only trace levels of antisense ␣ RNA can be detected (Fig. 1A). The observed shifts in ␤ mRNA under hypothyroid and diabetic states far exceed those in the S ␤ pre-mRNA (Fig. 1B) and thus cannot be explained by changes in the ␤ gene sense transcription alone. Rather, these findings suggest that cardiac ␤MHC gene expression is subjected to complex transcriptional and post-transcriptional processes impacting its expression at the mRNA level. In contrast, ␣ mRNA shifts corresponded well to changes in the S ␣ pre-mRNA levels thus supporting a classical transcriptional regulation of this gene. Analyses of the relationships between pre-mRNA species and mRNA products demonstrate a strong direct correlation between AS ␤ RNA and S ␣ pre-mRNA (Fig. 1C), whereas a significant inverse correlation was observed between the AS ␤ RNA levels and the expressed ␤ mRNA (Fig. 1D). These relationships suggest a direct co-regulation of transcription of sense ␣ and antisense ␤ via a common promoter region located between the sense ␣ and the antisense ␤ basal promoters on the ␤-␣ intergenic segment and that the AS ␤ RNA inhibits the processing of the S ␤ pre-mRNA into mature mRNA. This novel regulatory mechanism thus provides a working model (shown in Fig. 2), which can be used to explain the remarkable malleability in cardiac MHC gene expression thereby providing efficient antithetical regulation of the two cardiac genes during development and under altered hormonal status. Isolation of the Rat Cardiac ␤-␣MHC Intergenic Region and Sequence Analyses-Based on the above findings highlighting the role of the intergenic sequence in regulating the ␤ and ␣MHC gene locus, our goal was to isolate and characterize the intergenic DNA in terms of full sequence and bidirectional transcriptional activity. A 5-kb DNA fragment, corresponding to the full-length ␤-␣MHC intergenic fragment, was amplified from rat genomic DNA by high fidelity PCR and was fully sequenced. To gain insight on regulatory regions, the entire intergenic sequence was subjected to phylogenetic footprinting analysis, which is a powerful computational method to identify high probability DNA regulatory sites corresponding to well conserved regions in orthologous gene promoters (18,19). Analyses of the rat sequence against the cardiac MHC intergenic sequences from four other mammalian species (extracted from the GenBank TM via Blast searches) show that regulatory elements are clustered mainly in two domains, a proximal module consisting of the already characterized ␣ promoter (20) and a distal module located between 1.2 and 1.6 kb relative to ␣ TSS (Fig. 3). This distal module could represent the proximal regulatory region of the antisense promoter. These regulatory domains correspond to high percent sequence identity across the five species (Fig. 3A). The close physical proximity between the AS ␤ promoter and the ␣MHC promoter places them in an ideal strategic location to be tightly co-regulated via a common promoter region (CPR in Fig. 2A).

) Under
Antisense Promoter Activity in Hypothyroid Hearts-To test the function of the antisense transcriptional activity of the intergenic fragment, the entire 5-kb intergenic fragment was ligated into the multicloning site of the pRL null vector (Promega) in the antisense direction in front of the Renilla luciferase gene (Fig. 4A). This construct was tested for activity in driving reporter expression in the intact rat myocardium using a direct gene transfer approach. It was found that under normal condition, the antisense intergenic fragment drives expression to a level of 300% that of a promoterless vector. This expression was significantly reduced by 85% under the hypo- thyroid state (Fig. 4B). The responsiveness to PTU of this antisense transcriptional activity mimics the response that is found in vivo based on pre-mRNA analysis (Fig. 1A).
Where Does This Antisense RNA Originate from, and How Far Does It Extend Relative to the ␤MHC Gene?-To map the 5Ј end of AS ␤ RNA, 5Ј RACE (Invitrogen) was performed. Sequencing of the 5Ј RACE products found two transcription start sites at positions 2,195 and 2,158 bp upstream from the TSS of the ␣MHC gene. Reverse transcription using a primer from the first intron of the ␤MHC gene in the sense orientation, followed by PCR using intergenic primers targeting the cDNA corresponding to an intergenic region located at position Ϫ2753 to Ϫ2218 bp from the ␣MHC gene TSS was successful in amplifying a cDNA product that was responsive to thyroid state (Fig.   FIG. 3. Phylogenetic footprinting: cross-species comparison of the entire cardiac MHC intergenic region. A, schematic representation of the intergenic region between the ␤ and ␣MHC genes and a corresponding high resolution percent identity plot across five mammalian species: human, rat, mouse, hamster, and rabbit. Each data point represents percent identity in a window of 10 nt on the alignment consensus. Regions that are highly conserved may imply a probable regulatory role. B, distal regulatory module. Sequence alignment of the distal region corresponding to high percent identity across species from 3100 to 3400 bp of the aligned consensus (Ϫ1582 to Ϫ1290 bp relative to the rat ␣MHC TSS). Boxes and labels delineate the putative regulatory regions obtained via phylogenetic footprinting analysis. Note the presence of three putative thyroid responsive elements (TRE), which could be the site of action of thyroid hormones. C, proximal regulatory module. Sequence alignment of the proximal region corresponding to the ␣MHC promoter showing high percent identity on the plot. Labeled boxes show the pre-defined regulatory sites (20). For illustrative purposes, the Multalin multiple alignment program (45) was used to align all five sequences as shown in B and C, and TESS (cbil.upenn.edu/tess) was used to identify transcription factors associated with the conserved sites. In B and C, shaded bases designate 100% identity across the five species. GenBank TM accession numbers for aligned DNA are as follows: rat ␤-␣MHC intergenic region, AY191158; human ␤-␣MHC intergenic sequence, Z20656; mouse ␤-␣MHC intergenic sequence, U71441; hamster ␤-␣MHC intergenic sequence, L15351; and rabbit ␤-␣MHC intergenic sequence, AF192305.

FIG. 2. Schematic of cardiac MHC gene locus.
Cardiac MHC genes are located on the same chromosome. Each gene is ϳ25 kb, and they are separated by 4 kb. ␤P is the basal promoter of the ␤MHC gene transcribing the sense ␤MHC pre-mRNA. ␣P is the basal promoter of the ␣MHC gene transcribing the sense ␣MHC pre-mRNA. AS ␤P is an intergenic promoter transcribing the antisense ␤ RNA, which is proposed to inhibit the processing of the sense ␤ pre-mRNA into mature ␤ mRNA. We propose that both sense ␣ and antisense ␤ transcriptions are regulated by a common promoter region (CPR). 4C). In contrast, using PCR primers targeting a position from Ϫ2183 to Ϫ1897 relative to the ␣MHC TSS did not amplify a cDNA product from either control or PTU hearts. Both primer sets amplified a product when the 5-kb AS IG pRL plasmid DNA was used as a template (Fig 4C). These findings provide evidence that the AS ␤ RNA originates in the intergenic region between the ␣ and ␤MHC genes and extends across the fulllength ␤MHC gene.
Evidence That the Intergenic Bidirectional Promoter Is Active in the Human Heart-If the bidirectional ␣-␤ intergenic promoter is conserved across species (as predicted based on Fig. 3), one should be able to demonstrate its activity in the human heart. The intergenic promoter activity can be probed via analysis of the AS ␤ RNA. Three separate batches of human heart total RNA (Stratagene) were analyzed for the expression of ␣ and ␤ mRNA, S pre-mRNA, and AS RNA. RT-PCR analyses of the various MHC RNA species are consistent with previous findings that normal human heart expresses predominantly the ␤MHC isoform (21,22). It was found that although the ␤ mRNA/pre-mRNA is predominant in human ventricles, the ␣ mRNA/pre-mRNA can also be detected (Fig. 5). Furthermore, using specific RT-PCR, we were able to detect the AS ␤ RNA (Fig. 5B, lane 5). Thus, this regulatory mechanism is functional in human hearts and may contribute to pathologically associated shifts in cardiac MHC composition that can lead to altered cardiac function (21,(23)(24)(25). DISCUSSION Naturally Occurring Antisense RNA in Eukaryotic Cells-Antisense transcription has been reported for an increasing number of eukaryotic genes, and this process is believed to play an important role in the regulation of gene expression (26). Cis-acting antisense RNAs overlapping with corresponding sense RNAs have been reported for several eukaryotic genes (see review in Ref. 27), and they have been implicated in imprinting (28), alternative splicing regulation (29), and regulating the sense RNA expression (30,31).
It is relevant to note that the existence of an antisense cardiac ␤MHC RNA has been reported previously, but it was neither well characterized nor linked to a particular function such as coordinating the cardiac MHC antithetical shifts. In nuclear run-on assays aimed to study transcriptional activity of cardiac MHC genes in the rat heart, Boheler et al. (32) have reported a strong transcription signal in normal hearts when using a sense ␤MHC gene probe consisting of the third exon and some parts of flanking intron 2 and intron 3 sequences. To our knowledge, they have not studied the expression of this transcript under altered MHC expression. Furthermore, this report of the antisense expression has not linked its expression with a functional significance. More recently, Luther et al. (33)(34)(35) reported the expression of antisense MHC mRNA in both the rat and human myocardium, and it was proposed that this antisense MHC regulates translation of the sense mRNA into protein. In these studies (33)(34)(35), there was no differentiation between primary transcript versus processed mature mRNA. The primers used were located on the same exon, and thus both pre-mRNA and mRNA can be amplified without a difference, which could lead to misinterpretation of the results.
Role of Antisense ␤MHC RNA Is Elucidated-In this study, using a specific RT-PCR approach, we were able to detect significant levels of a AS ␤ RNA in rodent hearts. Based on the FIG. 4. Transcriptional repression of the antisense ␤-␣MHC intergenic promoter by PTU treatment. A, schematic of the reporter gene construct (5kb AS IG pRL) containing the entire ␤-␣ intergenic sequence ligated upstream of the R-Luc coding domain in the antisense orientation. B, reporter activity at 7 days following direct gene injection in the myocardium. Activities were expressed as a ratio of R-Luc activity to myosin light chain 2 firefly luciferase activity (F-Luc) and were reported as percent of the pRL null (promoterless) vector activity. Data presented are means Ϯ S.E. of n ϭ 8/group. *, p Ͻ 0.05 versus pRL null; #, p Ͻ 0.05 versus NC. C, the antisense ␤ RNA starts in the intergenic region between ␣ and ␤ and extends to the first intron of the ␤MHC gene. An antisense-specific cDNA was synthesized using a sense primer corresponding to the first intron (I1) of the ␤ gene (see Table I for the primer sequence). A product was amplified by PCR when the I1 cDNA was targeted with intergenic primers located at position between Ϫ2753 and Ϫ2218 bp from the ␣MHC gene TSS. The obtained product was almost undetectable in PTU hearts reflecting the target (AS ␤ RNA) sensitivity to hypothyroid state. In contrast, primers targeting position between Ϫ2183 and Ϫ1897 were not able to amplify cDNA product, but they worked well when 5-kb AS IG pRL was used as template.
relationship between the different forms of MHC RNA expressed in the rodent heart, the antisense RNA is proposed to inhibit the S ␤MHC pre-mRNA processing into functional mRNA. Furthermore, these data indicate that this AS ␤ RNA is a key player in the up-regulation of ␤ mRNA in both diabetes and hypothyroidism. The close proximity between the AS ␤ and the S ␣ TSSs and phylogenetic footprinting analysis together provide evidence that the intergenic region between the ␣ and ␤MHC genes plays a role in coordinating the cardiac MHC gene expression pattern in vivo, especially when a rapid switch of ␤ to ␣ or ␣ to ␤MHC expression occurs (see below).
Potential Cis-Regulatory Elements Involved in the Antisense Promoter Regulation-Based on pre-mRNA analyses reported in Fig. 1, the AS ␤ RNA transcription is co-regulated with the ␣MHC gene transcription, and both are turned off in the hypothyroid and diabetic states. The regulatory action of thyroid hormone is thought to be via direct interaction with its receptor, which is a transcription factor that interacts with thyroid responsive elements located on the promoter of the target genes (36,37). In contrast, the action of diabetes is less defined and may be the results of combinatory effects. Based on phylogenetic footprinting of the intergenic sequence, several cisregulatory elements including thyroid responsive elements and MEF-2 binding sites were elucidated (Fig. 3B) in a region proposed to be a common regulatory region between the two promoters. Although the thyroid responsive elements are likely the site of action of thyroid hormone, MEF-2 could be the site of action of diabetes. In striated muscle, it has been shown that insulin activates glucose transporter (Glut 4) gene transcription via a MEF-2 regulatory site. MEF-2 is a positive regulator of ␣MHC gene expression, and its levels are shown to be downregulated in diabetes (38). Furthermore, a synergistic interaction was reported between MEF-2 and thyroid hormone in regulating the ␣MHC gene. This synergy could also be a player in diabetes considering that diabetes is often associated with a mild hypothyroid state. Future studies should focus on analyzing promoter regulatory elements via footprinting and gel mobility shifts assays to gain insight into the mechanism of this dual regulation.
␤MHC Promoter Activity in Transgenic Mice Reveals Abnormalities during Development-During the first 3 weeks postbirth in rodents, the endogenous ␤MHC mRNA expression is down-regulated from Ͼ90% to trace levels, whereas the ␣MHC expression is up-regulated to become the predominant isoform expressed. This developmentally regulated antithetical shift in cardiac MHC expression is thought to be regulated by the surge in circulating thyroid hormones that occurs in the animal during this time frame of development (5,9,39). Interestingly, it has been reported that the ␤MHC promoter does not respond well to this developmental regulation when studied in transgenic animals (40,41). Specifically, these investigators stated that for the majority of the generated lines, reporter expression driven by the ␤MHC promoter transgene remained high during this stage of development even though the endogenous ␤MHC mRNA was rapidly down-regulated. Our findings reported in this paper explain these observations and support the hypothesis that the AS ␤ RNA generated from the intergenic region is important in directing the coordinated ␤-␣ antithetical regulation. In other words, it appears that the down-regulation of the ␤MHC expression during development may not be solely regulated at the transcriptional level via the gene promoter activity. Instead, for this regulation to occur it needs the AS ␤ RNA. It is not clear how the AS RNA interferes with the ␤MHC gene expression. It could be interference with its transcription or interference with its processing into mature mRNA. Clearly, more research is needed on this important topic.
Can Different Levels of AS ␤ RNA Explain Differences in Cardiac MHC Predominance among Species?-Previous studies show that large size mammals and those mammals with inherently slower resting heart rates such as humans, cows, guinea pigs, and rabbits have a predominance of the ␤MHC expression, whereas small mammals with resting heart rates above 300 beats/min such as mice and rats have a predominance of the ␣MHC expression (42,43). These MHC profiles appear to be optimal for cardiac function and energetics under normal condition for each species. Although we have not analyzed the expression of the AS ␤ RNA in other species besides the rat and human, the comparative analysis among the five  Tables I and II for primer information. B, representative gel showing different MHC mRNA and pre-mRNA PCR products obtained from RNA analysis. In lanes 1-4, the RT reaction was performed with oligo(dT)/random primers (200 ng) and 2 g of RNA in a 10-l reaction. In lane 5, the RT reaction was performed with 2.5 pmols of the all primer and 2 g of total RNA in a 10-l reaction. Detection of mRNA was accomplished after 24 PCR cycles using 1 l of 1:40 diluted cDNA template. Pre-mRNA and AS RNA detection was possible after 30 cycles using 1 l of undiluted cDNA. No signal was detected when non-RT RNA was used as template for similar PCR conditions. species suggest that the AS ␤ promoter is well conserved among small size and large size species. Conserved elements are more likely associated with function. Furthermore, based on our findings that the antisense ␤ RNA expression is coregulated along with the ␣MHC gene, we propose that the abundance of the AS ␤ RNA in different species would be directly correlated with the ␣MHC abundance. Thus hearts with low ␣MHC expression will have a low level of the AS ␤ RNA expression. Importantly, we propose that the role of this AS ␤ RNA is to coordinate simultaneous shifts between ␣ and ␤MHC gene expression and to effectively maintain the balanced stoichiometry for cardiac MHC protein. Thus, in all species, the expression of AS ␤ RNA will be important to coordinate these simultaneous shifts in phenotype that occur as the heart adapts to altered functional demands.
Significance of the MHC Regulatory Mechanism for Human Heart Function-Although earlier studies suggested that only the ␤MHC gene is expressed in the human myocardium, recent studies clearly show that a significant amount of the ␣MHC gene product is expressed in the human heart (21,22). The expression of the ␣ isoform is severely down-regulated in some forms of cardiac myopathy and in failing hearts (21,22). Furthermore, recent studies suggest that a small proportion of the ␣MHC expression in cardiac cells is important and critical for increased cell function and contractility (44). Improvement of left ventricular function in humans with dilated cardiac myopathy was directly related to increases in ␣MHC expression (23). Thus, being able to increase the ␣MHC expression in failing hearts by only small amounts could be beneficial in terms of increasing the myocardium contractility and power output, which should have a significant functional impact. Therefore, from a clinical perspective it becomes important to fully understand the mechanism(s) regulating cardiac ␣ and ␤MHC gene expression. The discovery of the AS ␤ RNA and its important regulatory role will add a new dimension to this understanding.
In this report we identify a novel mechanism of regulation for the tandemly linked cardiac ␤ and ␣MHC genes. This mechanism involves the antisense ␤MHC RNA. Although this antisense expression is shown to be important for diabetic and hypothyroid state-associated MHC shifts, its involvement remains to be determined during cardiac development, as well as in other models of cardiac remodeling such as during compensatory hypertrophy or in response to caloric restriction in which the ␤MHC gene is up-regulated. Furthermore better understanding the regulation of the antisense promoter would be crucial to fully understand cardiac MHC gene regulation in the clinical settings. In addition, we show evidence for a relationship between the organization of the cardiac MHC genes and the molecular basis for their antithetical regulation through an antisense transcription mechanism. This mode of gene regulation via bidirectional intergenic transcriptional activity has broader implications that could provide insight on regulatory mechanisms for other tandemly organized isogenes, i.e. skeletal muscle MHC isoforms and other clustered gene families. Furthermore, this report illustrates the importance of studying gene regulation in the context of the locus as a functional unit rather than studying each gene independently.