Background

MicroRNAs (miRNAs) are short (20 to 23-nucleotide), endogenous, single-stranded RNA molecules that regulate gene expression by hybridization to messenger RNAs (mRNAs) with the consequence of mRNA degradation or translational inhibition of targeted transcripts. Genes that encode for miRNA are transcribed by either RNA polymerase II or RNA polymerase III into primary miRNA (pri-miRNA) transcripts, which are then cleaved by the nuclear microprocessor complex formed by the RNase III enzyme Drosha (RNASEN) and the DGCR8 (DiGeorge critical region 8) protein. The RNase III Dicer cleaves off the loop of the pre-miRNA to generate a roughly 22-nucleotide miRNA duplex [1].

Although insights into the regulatory function of miRNAs toward their mRNA targets are beginning to emerge, less is known about the regulation of miRNA gene expression and miRNA biogenesis [1]. For instance, it has been shown that miRNAs participate in the control of cardiac development, and a number of miRNAs play a role in cardiac hypertrophy [2–5]. In addition, serum response factor (SRF), an important transcription factor, participates in the regulation of several cardiac enriched miRNAs, including mir-1 and mir-133a [4, 6]. However, it is unclear at what specific stage SRF regulates the biogenesis of miRNA.

SRF is a member of the MADS (MCM1, agamous, deficiens, SRF) family of transcriptional activators that has been implicated in the regulation of a number of genes that are important in cell proliferation and differentiation. SRF regulates its target genes by binding to the serum response element (SRE), which contains a consensus CC(A/T6)GG (CArG) motif [7–9]. This cognate response site of SRF is found in the promoter region of certain immediate-early genes and many muscle-specific genes [9–30]. The miRNA profile has been shown to serve as an impressive phenotypic signature. Therefore, miRNA has the potential to be developed as diagnostic and/or prognostic tools [31]. On the other hand, since individual miRNAs are dysregulated either negatively or positively with disease, strategies are needed to counteract or reverse the altered expression, potentially through reconstituting and/or targeting regimens, respectively [32–34]. A deeper understanding of miRNA biogenesis, as well as the regulation of each step in the miRNA maturation process and the involvement of signaling pathways and transcription factors is needed.

It has been reported that miRNAs undergo multiple steps during biogenesis, in which three forms of miRNAs are produced: pri-miRNA, pre-miRNA and mature miRNA [1, 35]. Multiple proteins are involved in this process. RNA polymerase II or RNA polymerase III transcribes the transcripts of pri-miRNA; a microprocessor complex cleaves pri-miRNA to generate pre-miRNA; and another endonuclease, Dicer, cleaves the pre-miRNA to release a miRNA-miRNA duplex. The miRNA strand is then stripped away from the duplex to leave a mature, approximately 22-nucleotide miRNA [36]. These individual steps can be targeted to increase or decreaase the miRNA level. Our data reveal that SRF executes it regulatory role in miRNA biogenesis through transcriptional regulation where it controls the amount of pri-miRNA available for downstream miRNA maturation process. We have proposed a model of miRNA regulation by SRF and its cofactors (Figure 10).

Figure 10
figure 10

A model of miRNA regulation mediated by SRF and its cofactors, including myocardin, p49/Strap, Zipzap and other proteins. Generally, the pri-miRNA transcript contains one miRNA (e.g pri-mir-21), but it can also contain more than one miRNAs (e.g. mir-1 and mir-133a). SRF regulates the transcription activity of both types of pri-miRNAs, thereby affecting the downstream mature miRNA level. The length of three representative primary transcripts: pri-mir-21 is over 3 kb, "pri-mir-1-1 and pri-mir-133a2" is 10 kb, and "pri-mir-1-2 and pri-mir-133a1" is 6 kb.

Full size image

Mir-21 is known to be up-regulated in many forms of cancer as well as in the heart during cardiac hypertrophic growth and heart failure [37–40]. In the failing heart, miR-21 levels are increased in fibroblasts where it inhibits sprouty homologue 1 (Spry1) and promotes fibrosis [16]. In vivo, the silencing of miR-21 by a specific "antagomir" in a mouse pressure-overload-induced disease model was reported to reduce cardiac ERK-MAP kinase activity, inhibit interstitial fibrosis and attenuate cardiac dysfunction [16]. It has been reported that TGF-β and BMP signaling promote a rapid increase in the expression of mature miR-21 at the post-transcriptional level through accelerating the process of pri-miR-21 into precursor miR-21 (pre-miR-21) by the DROSHA complex [41]. Mir-21 is one of the miRNAs that was significantly increased in the SRF-Tg heart, but was decreased in the anti-SRF-Tg heart. However, the mir-21 gene promoter was repressed by SRF and p49/Strap in transfection assay, and the promoter was strongly induced by myocardin. Similarly, we observed that SRF activates both cardiac α-actin and MLC-2v promoter-luciferase DNA constructs in vitro assay, but SRF repressed the expression of both mRNA in the mouse heart [18, 22]. These data indicate that the regulation of mir-21 gene expression is complex, and involves SRF in coordination with multiple SRF cofactors, including myocardin, p49/Strap and Zipzap. The fact that a "knock-down" of SRF level could decrease the mir-21 level provides an opportunity for future studies manipulating mir-21 level through the SRF-mediated signaling pathway.

Both miR-1 and miR-133a are produced from the same polycistronic transcripts, which are encoded by two separate genes in the mouse and the human genomes [42]. The mouse pri-mir-1-1 and pri-mir-133a-2 are transcribed into a polycistronic transcript that is 10 kb in length, and the pri-mir-1-2 and pri-mir-133a-1 are transcribed into another polycistronic transcript that is 6 kb in length [42]. Our data revealed that the down-regulation of miR-1 correlates closely with that of miR-133a in the SRF-Tg at various time points from 7 days to 6 months of age (Figure 7B). These findings suggest that SRF may regulate these two miRNAs at the level of polycistronic transcription, rather than at each individual miRNA (pri-mir-1 or pri-mir-133a) transcription, thereby kee** the expression of both miRNAs closely correlated. Since mir-1-1 and mir-1-2 genes are located on two different chromosomes, their expression is divergent. The pri-mir-1-1 is expressed at 6-fold higher than pri-mir-1-2. Therefore, the contribution of pri-mir-1-1 to the mature miR-1 pool may be greater than that of pri-mir-1-2. Given the fact that targeted mutation of mir-1-2 gene resulted in embryonic myocardial dysfunction and half of the mutant mice suffered early death due to ventricular septal defect (VSD) [4], one might speculate that a targeted mutation of mir-1-1 gene would also cause equally (or more) severe consequences.

The miR-1 is the most abundant miRNA that is expressed in the heart. Our present study revealed that miR-1 accounted for 7% of all the 578 miRNAs detected by the microarray. Mir-1 and mir-133a are down-regulated in cardiac hypertrophy and cardiac failure, suggesting that they may play a role in the underlying pathogenesis [14, 43]. It is plausible that increasing mir-1 and mir-133a level at a specific time point may have potentially beneficial effects against the pathological conditions. Matkovich et al reported that an increase of mir-133a level in the postnatal heart has beneficial effects against cardiac fibrosis after transverse aortic constriction [44].

In conclusion, our current study demonstrates that cardiac-specific overexpression of SRF leads to altered expression of cardiac miRNAs, especially the down-regulation of miR-1 and miR-133a, and up-regulation of miR-21, the dysregulation of which is known to contribute to cardiac hypertrophy. We observed that SRF plays a role in the regulation of miRNA biogenesis, specifically at the level of transcription of pri-miRNA. Reducing cardiac SRF level using the antisense-SRF transgenic approach led to the expression of miR-1, miR-133a and miR-21 in the opposite direction to that of SRF overexpression. Our findings may help in the future development of therapeutic interventions through targeting of the SRF-mediated signaling pathways.