Training protocol.

We have used a combined muscle action training protocol (isometric, concentric, eccentric) similar to the one described in Adams et al. (2). Each bout of stimulation lasted 3 s, the 1st s was of the isometric-type contraction followed by a 1-s concentric contraction, followed by a 1-s eccentric mode contraction. A period of 27 s of rest was allowed between each contraction. Rats performed five contractions per set. Five minutes of recovery was allowed between sets. On day 1, the session was limited to five sets, on the subsequent three training days, the animals received six sets of contractions during each training session.

RNA analysis.

Total RNA was extracted from frozen WMG muscle using the Tri Reagent protocol (Molecular Research Center). The extracted RNA was DNase-treated using 1 unit of RQ1 RNase-free DNase (Promega) per microgram of total RNA and was incubated at 37°C for 30 min followed by a second RNA extraction using Tri Reagent LS (Molecular Research Center).

The MHC mRNA isoform distribution was evaluated by reverse transcription (RT) with random primers/oligo(dT) mix, followed by PCR with primers targeting the embryonic, neonatal, I, IIa, IIx, and IIb MHC mRNAs, as described previously (3, 15). In these PCR reactions, each MHC mRNA signal was corrected to an externally added control DNA fragment that was coamplified with the MHC cDNAs using the same PCR primer pair. This provides a means to correct for any differences in the PCR reaction efficiency and/or pipetting of each PCR reaction. A correction factor was used for each control fragment band on the ethidium bromide-stained gel to account for the staining intensity of the variably-sized fragments (224 to 324 bp), as reported previously (15).

While the above RT-PCR can provide information on the MHC mRNA distribution pattern, it does not give information on how each isoform is regulated. Strand-specific RT-PCR was used to analyze the expression of specific MHC pre-mRNAs and mRNAs, as well as antisense RNAs that are of opposite orientation to the MHC genes.

Strand-specific RT-PCR used the one-step RT-PCR kit from Qiagen. These assays were utilized in the determination of the relative level of expression of pre-mRNA, antisense RNA, and mRNA in a known amount of total RNA, in comparing control vs. RE WMG. Also, these analyses were carried out in the mapping analyses of RNA expression across the skeletal MHC gene locus. The provided instruction was followed except in the RT step, only the reverse primer was added to target the sense RNA, whereas only the forward primer was added to target the antisense RNA. At the end of the RT, the reaction was heated to 90°C for 15 min to denature the RT enzyme; then the missing primer was added before initiating the PCR. In these assays, a negative control reaction was carried out to ensure the strand specificity of the product. In the RT step, no primer was added. After denaturing the RT, both the forward and reverse PCR primers were added. Any resulting product from this negative control reaction may be indicative of the formation of a cDNA in a primer-independent fashion or. alternatively. may be indicative of the presence of genomic DNA in the RNA. Our tests show the absence of any products in the negative control reactions, confirming that any amplified product corresponded to the target sense or antisense RNA in question (see Ref. 19 for more details).

For all of the PCR analyses, the primers were designed using DNA-star Primer Select software in which the target sequence was repeat masked using the Repeat Masker Web server at http://www.repeatmasker.org/cgi-bin/WEBRepeatMasker.

5′RACE.

These assays were conducted to identify the TSS of both the antisense IIx RNA and the sense IIb RNA. We adapted the method described by Invitrogen supplied with the kit (5′RACE). The cDNA was synthesized using Superscript II at 50°C and an RT primer reverse complementary to a region of the target RNA near the 5′ end. The cDNA was RNase H treated and purified using the Qiaquick PCR Purification Kit (Qiagen). The pure cDNA was tailed with dCTP using terminal transferase (New England Biolab). The tailed cDNA was amplified using a gene-specific reverse primer that was nested to the RT primer and an adapter primer (5′RACE primer, Invitrogen) (see supplementary Table S1 for the RT and gene-specific primer sequence. The online version of this article contains supplemental data.). The PCR products were cloned into pGEMT Easy (Promega), and the insert of individual clones was sequenced to determine the 5′ end corresponding to the end of the cDNA that is adjacent to the poly dC tail. Sequencing reactions used the ABI Big Dye sequencing kit version 3.1. Sequencing data acquisition was obtained from the UCI Sequencing Facility using an Applied Biosystem 3700 sequencer.

Genomic DNA.

Genomic DNA was extracted from rat tissue using the DNAeasy Tissue Kit (Qiagen). DNA was eluted with water and its concentration was determined by UV absorption at 260 nm (using a factor of 50 μg/ml per OD unit). The obtained genomic DNA was used as template to test all the PCR primer sets targeting pre-mRNA and antisense RNA to ensure that these primers worked with similar efficiency to each other.

Computational analyses of DNA sequence.

Rat genomic sequence (genomic contig number: NW_047334, location: 10q24) was aligned with human (contig number: NT_010718, location: 17p13), mouse (contig number: NT_096135, location: 11 35cM) and monkey (GenBank accession no. AC212180) using Mulan comparative genomics site at http://mulan.dcode.org/ (36). The purpose for this comparative analysis was to identify evolutionarily conserved sequences at 85% in these promoters and to identify conserved transcription factor binding sites (TFBSs) in these conserved regions.

MHC mRNA analysis.

The MHC mRNA isoforms distribution in the WMG and their shift following 4 days of RE were assessed through a competitive PCR approach, and they are shown in Fig. 1. The mRNA of each MHC gene was expressed as the percentage of the total MHC mRNA pool. Confirming previous findings (11, 22), only fast MHC isoforms, IIx and IIb, were expressed in both groups of samples. Since other MHC mRNA isoforms were not detected either in control or in the RE group, they were not reported in the graph. The MHC mRNA distribution in control WMG muscles was represented by 76% of IIb. The acute RE paradigm caused a statistically significant shift in the MHC composition, with a decrease of MHC IIb mRNA to 50% and a consequent increase of MHC IIx mRNA from 24 to 47%. These results were consistent with a previous report on MHC isoform shifts following mechanical loading of fast skeletal muscles (11). Furthermore, these results demonstrate that the RE in these present experiments was effective in inducing MHC isoform shifts, which warrant further analyses of RNA across the gene locus.

Open in a separate window

Fig. 1.

Myosin heavy chain (MHC) mRNA distribution profile as determined by competitive RT-PCR. Also shown are the PCR products separated on the gel and stained with ethidium bromide. MHC is the band amplified from the cDNA, and the control is an external synthetic control added to the cDNA and amplified using the same MHC primers but generate a smaller size product (see Refs. 3 and 15 for details). E, embryonic; Neo, neonatal; CON, control; RE, resistance exercise training. *P < 0.05 RE vs. control.

MHC RNA expression in response to resistance loading.

The IIx and IIb MHC pre-mRNA and mRNA were used as markers of transcriptional and of pretranslational events, respectively. In theses analyses, both the mRNA and the pre-mRNA are expressed as arbitrary units per nanogram of total RNA, whereas in Fig. 1, the MHC mRNA was expressed as relative percentage of the total MHC mRNA pool. Results in Fig. 2 show that for the IIx MHC gene products, both the pre-mRNA and the mature mRNA signals (Fig. 2, A and B) were significantly increased in response to RE (P < 0.0001). The IIx MHC pre-mRNA was approximately fivefold higher in RE muscles than in the control group and the IIx MHC mRNA recorded was 6.5-fold greater (Fig. 2B). For the IIb MHC gene products, the expression of pre-mRNA and the mRNA was significantly decreased in RE muscles (P < 0.001). The RE WMG expressed only 25% and 40% of the IIb pre-mRNA and mRNA, respectively, compared with control muscle (Fig. 2, D and E).

Open in a separate window

Fig. 2.

RNA expression for IIx and IIb MHC in control and RE white medial gastrocnemius (WMG) as determined by strand-specific one-step RT-PCR: IIx MHC pre-mRNA (A), IIx mRNA (B), antisense IIx RNA (xII NAT) (C), IIb MHC pre-mRNA (D), and IIb mRNA (E). F: representative gels are shown for each of the RNA species. Antisense IIb RNA was below detection level; therefore, it is not shown. Sequence of primers used for these analyses is reported in supplementary Table S2. *P < 0.05 RE vs. control.

In addition to the sense RNA species for the IIx and IIb MHC genes, the antisense RNA was analyzed in several locations across the IIx MHC gene and in the IIx-IIb IG region. The IIx MHC antisense RNA (xII NAT), as measured in the IG region (at ~2 kb from the IIx TGA stop codon), was decreased significantly in the RE muscles (RE is 9% of the control group; P < 0.001) (Fig. 2C). Interestingly, the antisense transcription persisted along the entire IIx gene and into its promoter region 5′ upstream of the IIx gene (see below). No signal of antisense RNA was found in the IIb MHC gene, either in control or in RE muscles. The antisense transcript corresponding to the IIb MHC gene was targeted at several locations across the IIb gene, including its 3′ flanking region.

The relationship between xII NAT vs. IIx MHC pre-mRNA and the relationship between xII NAT vs. IIb MHC pre-mRNA were analyzed by linear regression. As presented in Fig. 3, there was a significant inverse correlation between the xII NAT and the IIx MHC pre-mRNA (R = −0.8; P < 0.001). In contrast, there was a significant positive correlation between the xII NAT and that of the IIb pre-mRNA (R = +0.8; P < 0.0001) (Fig. 3). These relationships show that the behavior of the IIx-IIb gene locus transcriptional activity is remarkably similar to those described for the cardiac β- and α-MHC genes (17), as well as for those for the IIa-IIx MHC gene in the soleus muscle (38). The coordinated regulation of the IIx-IIb gene concurs with the previously described models for two linked genes.

Open in a separate window

Fig. 3.

Linear relationship between antisense IIx RNA vs. IIx pre-mRNA (A) and vs. IIb pre-mRNA (B) in the WMG under control or RE condition. Regression lines were generated by linear regression using Graphpad Prism software. R, Pearson correlation coefficient.

Transcription mapping across the IIa-IIx-IIb genes in WMG muscle.

The above-reported results demonstrate that an antisense IIx RNA is expressed in fast WMG muscle fibers, and its expression is downregulated in response to RE. It was of interest to map the extent of sense and antisense transcription along the adult skeletal MHC gene locus (IIa-IIx-IIb) in the fast WMG. Based on the rat genomic DNA sequence in NCBI (contig accession no. NW_047334, Rattus norvegicus chromosome 10 genomic contig, reference assembly; RGSC v3.4), these genes cover a region of ~90 kb of genomic DNA of which the IIa gene is ~27 kb, whereas the IIx and the IIb genes are ~23 kb each in size. The IIa and IIx MHC genes are separated by ~2.7 kb, whereas the IIx-IIb MHC genes are separated by ~14 kb of IG space. Within the 14 kb IG space, there is a gap of ~661 bp of unknown sequence centered at approximately −3 kb relative to the IIb MHC gene TSS. Several DNA repetitive elements are found on the IIx-IIb IG sequence as revealed by using a repeat masker tool (Repeat Masker web server; url: http://www.repeatmasker.org/cgi-bin/WEBRepeatMasker). The masking results showed that these genomic repeats make up ~31% of the total IIx-IIb IG region (Fig. 4).

Open in a separate window

Fig. 4.

Mapping RNA transcription across the IIa-IIx-IIb MHC gene locus. Gray arrows represent RNA detection using strand-specific RT-PCR. Open boxes on the IIx-IIb intergenic (IG) region represent genomic repeats, whereas the hatched box represents a yet unknown sequence (n = 661). Bottom: high-resolution mapping of the IIb proximal promoter that covers 1 kb from the end of the IIx-IIb IG and 1 kb from within the IIb gene at its 5′ end. Bars 1–13 are primer pairs used for mapping RNA expression via 1-step RT-PCR. Both sense (S) and antisense (AS) RNA are shown for each primer pairs for both control and trained (RE) WMG muscles, as well as genomic DNA (G DNA) amplification products. For each reaction, 100 ng of total RNA were used and PCR was run for 30 cycles. Note that for primers 2 thru 9, the gels show a light band corresponding to amplification of S RNA, the light band is barely visible in the high-contrast image. The amplified band is, however, more clearly visible when larger amounts of RNA were used (300 ng/reaction; data not shown). ^aPrimer 10 resulted in an intense band that is less than the predicted size. Sequencing of this lower molecular size band revealed that this is a spliced RNA in which intron 1 of the IIb gene is spliced out. The shown band was 97 bp instead of expected 297 (see supplementary Table S3), and is downregulated with RE. See Table S3 for primer sequence. For genomic DNA amplification, 20 ng of rat genomic DNA was amplified for 27 cycles using Biolase DNA polymerase (Bioline).

A strand-specific RT-PCR-based approach was utilized to target and amplify the sense and antisense RNA in the WMG total RNA at sites corresponding to various regions of the IIa-IIx-IIb MHC gene locus. Strand-specific RT-PCR results demonstrated that the expression of antisense RNA in WMG muscles can be detected across the entire IIx gene, its promoter region consisting of the IIa-IIx IG region (2,677 bp), and in the region corresponding to the 3′ end of the IIa gene (see Fig. 4). Also, antisense RNA was detected corresponding to a large segment of the IG region between the IIx and IIb MHC genes to within approximately −200 bp relative the IIb MHC gene TSS (+1). No sense RNA was detected in the control WMG across either the IIa gene or the IIa-IIx IG region. In a few trained WMG samples, some trace levels of IIa sense RNA was detected. A significant amount of sense RNA was detected across the entire IIx MHC gene, the entire IIb MHC gene, and to ~3 kb past the IIx or IIb gene stop codon (Fig. 4). Furthermore, these mapping analyses detected no antisense RNA that corresponds either to the IIb gene or to its 3′ flanking sequence. To our surprise, a low level of sense RNA was detected across ~400 bp of the IIb MHC proximal promoter. This sense RNA merged with the IIb gene at +1 and it became undistinguishable from the IIb pre-mRNA. This low level of sense RNA corresponding to the IIb proximal promoter could not be attributed to a nonspecific signal or contamination since no bands were observed in the negative control reactions.

A high resolution mapping of sense and antisense transcriptional activity in the proximal IIb promoter (−1 kb to +1 kb vs. IIb TSS) is shown in Fig. 4. These mapping analyses suggest that both sense and antisense IG transcriptions are occurring in the proximal IIb promoter. Both of these sense and antisense IG RNA start within 400 bp from the IIb MHC gene TSS. 5′RACE analyses using primer extension and tailing reactions followed by sequencing, revealed the presence of multiple start sites of the antisense RNA that are clustered between −403 to −271 bp relative to the IIb MHC gene TSS. Similarly, the 5′RACE analyses of the sense RNA targeting sense IIb RNA revealed the presence of multiple start sites with the major site corresponding at +1, but several minor sites clustered between −400 and −100 relative to the major TSS at +1. The close proximity of the antisense TSSs to those of the IIb MHC TSSs confirmed the idea that the two opposite-in-direction transcripts were initiated in the same area of the IIx-IIb IG region from a unique bidirectional promoter or from two adjacent head-to-head opposing promoters. These findings strongly suggest that the IIx-IIb gene regulation is, in a way, similar to the cardiac β- and α-MHC genes, whereby both sense and antisense transcription were detected in the IG region (20). The linear regression analyses showing that the xII NAT was strongly correlated with IIb pre-mRNA in the WMG when subjected to RE (R = +0.8 P < 0.0001) (see Fig. 3B) highlights the intriguing notion of common regulation between the two head-to-head situated promoters.

Study of the locus of the IIX-IIB IG region.

Our data show that the sense and the antisense transcripts share the same proximal promoter region (Fig. 4). Through RNA mapping using a one-step-RT PCR approach and 5′RACE analyses, we could circumscribe the area of transcriptional origin and limit it to a region of approximately −403–271 bp relative to the major IIb TSS located at +1. The TSS analysis carried out with a 5′RACE approach on the rat sequence indicated a cluster of multiple start sites for the antisense RNA in the area situated between −270 and −413 bp upstream of the IIb MHC TSS. This finding is not surprising; it is in agreement with the cardiac antisense β-RNA in which at least two TSSs were identified (17), and the skeletal muscle antisense IIa RNA (aII NAT), whereby more than one TSS was found (38). In addition, mapping analyses and 5′RACE targeting the sense IIb RNA revealed that IIb transcription is carried out with the major start site at +1. However, several minor sites were also found in the IIb promoter, which result in sense RNA that merges with the IIb pre-mRNA and is subjected to similar splicing as suggested by the PCR results using primer 10 (Fig. 4). The observation of the presence of multiple TSS for IIb is not surprising and confirms the finding by Dennehey et al. (14) for the mouse gene showing that the IIb MHC mRNA is expressed in several isoforms that are different in their 5′ untranslated region. Some of these RNA start as far as −1 kb from the IIb TSS (+1) (14). The presence of a bidirectional promoter that coregulates both the sense and antisense RNA transcription is very interesting and is common in human genes. Recently, it has been reported that >10% of human genes are divergent, i.e., controlled by bidirectional promoters (1, 30, 46). Such promoters are defined as IG, < 1-kb long, and flanked by a TSS on both of the DNA strands (32). What makes the bidirectional proximal IIb gene promoter even more interesting in the present study is that its activity is in an inverse relation with the activity of the upstream IIx sense gene (Figs. 3A and ​and6).6). That is, when the bidirectional promoter is active, both the IIb sense and the antisense IIx RNA are transcribed, whereas the IIx MHC gene is shut down. In contrast, when the bidirectional promoter undergoes inhibition, both the IIb gene and the antisense transcript are downregulated, but the IIx gene located upstream is activated. There is a simultaneous ying-yang relationship, facilitated by cross talk through more than 37 kb of genomic DNA. Therefore, the IG region that makes up the proximal bidirectional promoter may act not only as a physical linkage between the two MHC genes (IIx and IIb), but it seems to be essential also for their coordinated regulation. Therefore, the IIx and IIb MHC genes are physically and functionally linked through the IG bidirectional promoter region as suggested by the model presented in Fig. 6. In particular, it appears that in order for the IIx MHC gene to be regulated, a feedback from the IG antisense RNA is needed, especially in states whereby shifts in IIx-IIb expression occurs in response to a change is physiological stimuli.

Comparative sequence analysis.

Genome sequence comparisons are typically used to identify noncoding genomic DNA regions that have been evolutionarily conserved across species, presumably to maintain some critical biological function. Such conserved regions often correspond to TFBS and regulatory regions (33, 35, 48). Thus, in this study, a comparative genomic approach was applied to provide insights on IG IIx-IIb promoter sequence conservation among species. MULAN, a computational tool available at http://mulan.dcode.org, which is able to generate multiple sequence alignments, was used to identify conserved regulatory elements by comparing genomic sequences between related species (36). Four orthologous sequences of 4 kb in length, spanning from ~−1.4 to +2.8 relative to the IIb TSS (+1), were compared. The rat sequence was analyzed against mouse, rhesus monkey, and human DNA sequences. The sequence conservation through evolution presumes a regulatory function of that region. The comparative analysis showed, in fact, that between −1.4 kb and +2.8 kb from the IIb MHC TSS, two regions were highly conserved (85% of similarity) (Fig. 5). The first one (ECR1) was shown to be across the IG region and into the beginning of the IIb MHC gene, located from −294 bp to +47 bp (Fig. 5). The second one (ECR2) was found within the 5′ end of the IIb MHC gene from +211 to +312 bp relative to IIb TSS (+1). Several muscle-specific TFBSs were found on ECR1 upper and lower strands, whereas both ECR1 and ECR2 exhibited several general TFBSs. The ECR1 analysis identified several potential cis regulatory elements such as CAAT box, C/EBP, SRF, and the muscle-specific TFBS for myogenin and MEF-2 (Fig. 5). These conserved regulatory elements were found on the sense strand as well as on the reverse strand. Since ECR1 and ECR2 are highly conserved, their properties may be important promoter regions controlling both the antisense as well as the sense transcription.

The location of the TSS, the presence of conserved cis regulatory elements driving transcription on both the DNA strands and also the position of the very high conserved region across four species provides evidence that the sequence considered (ECR1) is highly probable to function as the common promoter region for the sense IIb and the xII NAT.

It is worth mentioning that previous studies on the mouse IIb promoter have identified some important cis regulatory regions that coincide with some of the ones conserved in ECR1 in the proximal promoter (see Fig. 5). For example Takeda et al. (44) have identified three AT-rich regulatory regions and a proximal E box, and later it was determined that these AT-rich sites interact with MEF-2 and Oct1 (27). Furthermore, the functional role of MEF2, MRF, and SRF transacting factors in the regulation of the IIb proximal promoter in muscle fibers was highlighted in studies by Allen et al. (4, 5). Konig et al. (26) have followed a comparative genomic approach to identify conserved domains in the proximal promoter of mouse and human IIb. However, these previous studies were mostly concerned with the regulation of the IIb promoter activity in the sense direction, whereas in this present study, we unraveled bidirectional activity, and thus we are more concerned in finding that some of these factors can operate for both the sense and antisense transcription.

Relevance to IIx-IIb coexpression in single fibers.

The IIx and IIb MHC genes are closely linked on the skeletal MHC gene locus, and in the WMG are regulated in opposite direction as the muscle undergoes phenotype shift in response to training. Both isoforms are expressed in fast skeletal muscle fibers (9), and therefore may respond to the same transacting factors. In fact, the similarity of TFBS in the proximal promoter among the fast MHC genes, in particular the MEF2-AT-rich sites was noted in several reports (5, 26, 44). Being part of a gene cluster, these genes may be regulated via a common enhancer (26); however, the antisense xII NAT may be important when opposite regulation of the two genes is required. At present, it is not clear whether within the same locus both IIx and IIb are being transcribed at different levels, such that in control WMG 70–80% of the total MHC pool is IIb, whereas the rest is IIx. On the other hand, because the muscle fibers are multinucleated, one might consider that the mosaic nature of MHC isoform expression within a single fiber is because individual myonuclei may be expressing different components of the sense and antisense regulatory circuit differentially at any given moment. On the other hand, because the muscle fibers are multinucleated, one might consider that the mosaic nature of MHC isoform expression within a single fiber is due to individual myonuclei expressing different components of the sense and antisense regulatory circuit differentially at any given moment. Another point to consider is that the dual expression of these two isoforms in many of the individual IIx/IIb fibers is not an all or none process; that is, the antisense model does not operate in a fashion that IIb is fully repressed, while the IIx is optimally expressed. Clearly the regulation is more complex as our data indicate.

Relevance to the differential IIb MHC gene expression between rodent and human.

The results of this study show that xII NAT and the pre-mRNA of the IIb gene are generated by a unique bidirectional promoter located in the IG region flanking the 5′ end of the IIb gene. The IIx-IIb promoter region is highly conserved across mammalian species, even in species phylogenetically distant from each other, such as rat and human. In view of the fact that the IIb MHC gene is rarely expressed in human skeletal muscle (24), it is intriguing to find out why the human IIb gene regulation is different from that of the highly expressed rodent IIb, considering that major regulatory components of the IIb promoter are conserved between the two species. Preliminary studies of the human IIb promoter in a transient reporter assay following direct gene transfer in skeletal muscle show that the human IIb promoter is active in rodent fast muscle and exhibited muscle-type specificity, as the mouse IIb was shown to do previously (5, 43). Thus, it would be useful and intriguing to ascertain the factor/s that really promotes the transcription of the IIb MHC gene in rodent vs. human muscle fibers.

The authors thank Li Ying Zhang, Phuc Tran, Alvin Yu, Bryce Buchowicz, Daniel Jimenez, Sandy Liu, Nkiruka Ojukwu, Jasleen Saini, and Tiffany Yu for excellent technical assistance.