Cardiac hypertrophy is enlargement of the heart in response to physiological or pathological stimuli, chiefly involving growth of myocytes in size rather than in number. Previous studies have shown that the expression pattern of a group of genes in hypertrophied heart induced by pressure overload resembles that at the embryonic stage of heart development, a phenomenon known as activation of the “fetal gene program”. Here, using a genome-wide approach we systematically defined genes and pathways regulated in short- and long-term cardiac hypertrophy conditions using mice with transverse aortic constriction (TAC), and compared them with those regulated at different stages of embryonic and postnatal development. In addition, exon-level analysis revealed widespread mRNA isoform changes during cardiac hypertrophy resulting from alternative usage of terminal or internal exons, some of which are also developmentally regulated and may be attributable to decreased expression of Fox-1 protein in cardiac hypertrophy. Genes with functions in certain pathways, such as cell adhesion and cell morphology, are more likely to be regulated by alternative splicing. Moreover, we found 3′UTRs of mRNAs were generally shortened through alternative cleavage and polyadenylation in hypertrophy, and microRNA target genes were generally de-repressed, suggesting coordinated mechanisms to increase mRNA stability and protein production during hypertrophy. Taken together, our results comprehensively delineated gene and mRNA isoform regulation events in cardiac hypertrophy and revealed their relations to those in development, and suggested that modulation of mRNA isoform expression plays an importance role in heart remodeling under pressure overload.
Citation: Park JY, Li W, Zheng D, Zhai P, Zhao Y, Matsuda T, et al. (2011) Comparative Analysis of mRNA Isoform Expression in Cardiac Hypertrophy and Development Reveals Multiple Post-Transcriptional Regulatory Modules. PLoS ONE 6(7): e22391. https://doi.org/10.1371/journal.pone.0022391
Editor: Gangjian Qin, Northwestern University, United States of America
Received: April 19, 2011; Accepted: June 20, 2011; Published: July 22, 2011
Copyright: © 2011 Park et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded by National Institutes of Health (NIH) grants HL59139, HL67724, HL69020, HL91469, HL102738, and AG27211 (to J.S.), and the Foundation of Leducq Transatlantic Network of Excellence (to J.S.), and NIH grants HL098802 and GM084089 (to B.T.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The growth of the heart in mammals takes place at embryonic and postnatal developmental stages, but it can also be induced by physiological or pathological stimuli in the adult . While cardiac growth at early developmental stages mostly involves proliferation of cardiac myocytes, expansion of cell size is chiefly responsible for growth of the adult heart. The latter is also known as cardiac hypertrophy, which is a common component of many cardiac diseases. Cardiac hypertrophy is generally considered as an adaptive mechanism in response to increased mechanical load; but hypertrophied heart under sustained pressure can lead to heart failure , .
Studies have shown that the expression levels of many genes are regulated during cardiac hypertrophy , , . A group of genes with low expression in the adult but high expression in the embryo is reactivated in cardiac hypertrophy, a phenomenon widely known as activation of the ‘fetal gene program’. These genes typically play roles in metabolic and contractile functions of the heart , , and are regulated by a set of transcription factors (TFs) which play roles also in embryonic development, such as NFAT, NFκB, MEF2, GATA4, and SRF , . In addition, recent studies have implicated regulation of several microRNAs (miRNAs) in hypertrophy, including miR-1, miR-133, and miR-208 , , . miRNAs are small non-coding RNAs (∼22 nucleotides) that cause mRNA degradation and/or inhibition of translation by binding to their target sites in mRNAs, mostly in the 3′ untranslated region (3′UTR) , . miRNAs have been increasingly found to play important roles in cardiac development and diseases , , , .
Mammalian genes frequently express mRNA isoforms resulting from alternative initiation of transcription, alternative splicing (AS), and alternative cleavage and polyadenylation (APA), which are frequently regulated in temporal- and tissue-specific manners , . Studies have shown that the AS pattern of the heart is distinct from other tissues , , , and dynamic regulation of AS takes place at the embryonic and postnatal developmental stages of the heart , during differentiation of cardiac precursors , and in patients with heart failure . Ablation or overexpression of some splicing regulators, such as SC35, ASF/SF2, SRp38, and CELF, have been shown to cause malfunctions in the heart , , , .
Recent studies have indicated that APA plays an important role in gene regulation. Over half of the human genes have APA isoforms , most of which lead to different 3′UTR lengths. The APA pattern of genes is tissue-specific  and is globally regulated in cell proliferation, differentiation, and development , , . In general, genes tend to express short 3′UTR isoforms more frequently in proliferative or transformed cells than in quiescent or differentiated cells, presumably due to the difference in 3′ end processing activity . Since 3′UTRs typically contain cis elements that inhibit gene expression at the post-transcriptional level, such as AU-rich elements (AREs), GU-rich elements (GREs), and miRNA target sites, regulation of 3′UTR length by APA can impact mRNA metabolism. For example, shortened 3′UTRs in cancer cells have been shown to make mRNAs of a set of oncogenes more stable, leading to higher protein expression .
Here we took a genome-wide approach to systematically define genes and pathways regulated in cardiac hypertrophy induced by transverse aortic constriction (TAC), and compared them with those regulated at different stages of embryonic and postnatal development. By exon-level analysis we found that cardiac hypertrophy involves widespread mRNA isoform changes, a fraction of which is developmentally related. Gene Ontology analysis indicated that regulated AS events are biased to genes with functions in cell adhesion and cell morphology, suggesting an important role of AS in remodeling the heart. Our analysis also indicated that downregulated expression of Fox-1 protein during cardiac hypertrophy may be responsible for a set of regulated AS events. Moreover, we found 3′UTRs of mRNAs are generally shortened through APA in hypertrophy, and microRNA target genes are generally de-repressed, suggesting coordinated mechanisms to increase mRNA stability and protein production during hypertrophy. Taken together, our results comprehensively delineated gene regulatory modules at transcriptional and post-transcriptional levels in cardiac hypertrophy and revealed their relations to those executed in development, and suggested that mRNA isoform regulation plays critical roles in remodeling the heart under pressure overload.
Comparison of gene expression profiles between cardiac hypertrophy and development defined the fetal gene program
To systematically examine gene regulation in short- and long-term cardiac hypertrophy, we used genome-wide exon microarrays to analyze mRNAs expressed in mouse hypertrophied left ventricle (LVH) induced by 1 week (1 W) or 4 W Transverse Aortic Constriction (TAC). To reveal how gene regulation in hypertrophy is related to that in development and to achieve robustness in our analysis, we collected a set of publicly available microarray datasets corresponding to murine heart development at embryonic and postnatal stages, and short- and long-term LVH (Figure 1A and Table S1). Overall, our LVH data included three datasets for 1 W TAC, named LVH1, LVH2, and LVH3, one dataset for 4 W TAC (LVH4), and one dataset for 12 W TAC (LVH5). All TAC samples were compared with sham samples in the same dataset. For embryonic development (ED) of the heart, we divided data into two phases. The first phase, named ED1, was based on comparison of embryonic day (E) 13.5 + E14.5 with E10.5 + E11.5; and the second phase, named ED2, was based on comparison of E16.5 + E18.5 with E13.5 + E14.5. We also obtained a dataset for comparison of embryonic heart at E17 with adult heart, named EA, which can reflect postnatal development. In addition, we included a rat dataset, from which we compared gene expression for postnatal day (P) 20 vs. P1 and P49 vs. P20, named PD1 and PD2, respectively. Expression changes were standardized to make data from heterogeneous sources more comparable (Figure S1).
(A) Schematic of datasets used in our analysis. Top, hypertrophy data with TAC period indicated (W = week). All hypertrophy samples were derived from the left ventricle and thus were named LVH. They were compared with Sham samples. Bottom, development samples. E, embryonic day (for mouse samples); P, postnatal day (for rat samples). ED1 is comparison of E10.5 + E11.5 with E13.5 + E14.5 and ED2 is comparison of E13.5 + E14.5 with E16.5 + E18.5. PD1 is comparison of P1 with P20 and PD2 is comparison of P20 with P49. EA is comparison of E17 with adult. E10.5, E11.5, and EA were based on the whole heart, and others were based on the left ventricle. All ratios were derived from expression level at a later development time point divided by that at an earlier point. (B) Clustering of genes and samples using gene expression changes. Expression changes are standardized log2(ratio) or z, which are shown in a heatmap using the color scale shown at the bottom. A total of 1,945 genes with z >2 in at least one of the samples are shown. Hierarchical clustering is based on Pearson correlation and the average linkage method. (C) Venn diagram of genes regulated in development (DEV) and hypertrophy (LVH). Genes were selected from five development datasets, i.e. ED1, ED2, PD1, PD2, and EA, and three 1 W TAC sets, i.e. LVH1, LVH2, and LVH3. A gene was selected if its z is greater than 2 in at least one sample and the direction of its regulation is consistent across samples. Fisher's exact test was used to assess the significance of bias in genes overlapped between DEV and LVH. (D) Significant Biological Process (BP, top) and Cellular Component (CC, bottom) terms regulated in hypertrophy. Only top 10 terms are shown for BP and CC. The average p-value of three 1 W TAC samples was used to sort entries. P-values in other samples are also shown for comparison. To remove redundancy, we eliminated GO terms that overlapped by more than 50% of associated genes with another term that had a more significant p-value. P-values are presented in a heatmap according to the color scheme shown in the graph.
Using hierarchical clustering, we found, as expected, that development samples were separated from hypertrophy samples, and embryonic development (ED1 and ED2) samples were separated from postnatal development (PD1, PD2 and EA) ones (Figure 1B). Interestingly, gene expression differences between late embryonic stage and adult (EA) in mouse were more similar to those in postnatal development of rat (PD) than to those in embryonic development of mouse (ED), indicating that gene expression changes in rat and mouse at similar developmental stages are well correlated despite different species and array platforms. This result also supports the suitability of using the rat dataset in our analysis.
Among the hypertrophy samples, 1 W TAC samples (LVH1–3) were clustered together and were separated from 4 W and 12 W TAC samples despite that the three 1 W TAC datasets were from different labs and platforms, indicating that genes are distinctly regulated in short- and long-term hypertrophy conditions. As evidenced by the overall clustering pattern (Figure 1B) and Pearson Correlation values (Figure S2), gene expression changes in hypertrophy showed a general negative correlation with those in development, particularly with late embryonic development (ED2) and early postnatal development (PD1). We then identified genes that were significantly regulated in hypertrophy and development. As shown in Figure 1C, more genes were oppositely regulated in hypertrophy vs. development than consistently regulated (261 vs. 89, see Table S3 for the full lists), further supporting the overall inverse correlation between hypertrophy and development (P<2.2×10−16, Fisher's exact test). In addition, a greater number of genes were found to be regulated in development or hypertrophy only.
We next analyzed Gene Ontology (GO) terms in Biological Process (BP) and Cellular Component (CC) categories for the genes regulated in hypertrophy and development (Figure 1D). Using a new method based on comparison of regulation profiles of different sets of genes (Figure S3 and Text S1 for detail), we found that genes with functions in immune response, extracellular matrix (ECM), cell cycle, and cell morphology tended to be upregulated in hypertrophy, whereas those with functions in energy generation, mitochondrion, and peroxisome tended to be downregulated. Some gene pathways were consistently regulated in hypertrophy and development (Figure 1D), for example, immune response; some pathways were oppositely regulated, such as mitochondrion and cell cycle. The relations were also discernable using gene density plots which showed how genes were distributed with respect to regulation in hypertrophy or development (Figure S4).
Taken together, our result systematically defined genes and pathways regulated in hypertrophy and divided them into development-related and nonrelated groups. The former group corresponds to the “fetal gene program”, which is presumably attributed to reversal of the mechanisms involved in development during hypertrophy.
Regulation of mRNA isoform expression in cardiac hypertrophy
Alternative usage of exonic regions leads to mRNA isoforms encoding different proteins and/or UTRs. To understand how mRNA isoform expression is regulated during hypertrophy, we analyzed exon-level probe intensity changes using our exon array data for 1 W and 4 W TAC and Sham samples. Using the splicing index method, which examines exon usage by normalization to gene expression, and the COSIE method for correction of probe hybridization bias (see Materials and Methods for detail), we identified 2,086 and 1,942 exons with significant change of usage (>3*standard deviation) in 1 W and 4 W TAC samples, respectively (Figure 2A). While a significant fraction of regulated exons (about 14–16% in each sample) were commonly regulated in both 1 W and 4 W TAC samples, suggesting continuous regulation in the course of hypertrophy (P<2.2×10−16, Fisher's exact test), a large number of stage-specific regulated events were also identified.
(A) Venn diagram of exon-level regulation in 1 W and 4 W TAC samples (LVH3 and LVH4). Significantly regulated exons are those with splicing index (SI) >3*standard deviation (SD) of all exons. (B) Gene expression changes in 1 W and 4 W TAC samples. Genes were divided into four quartiles based on log2(ratio). (C) Distribution of regulated exons with respect to location and gene expression. Exons were divided into four groups based on gene expression changes as shown in (B), and were further classified into first, internal, and last exons. ‘I’, inclusion; ‘E’, exclusion.
Interestingly, we found a correlation between change of gene expression and regulation of internal exon usage in the 1 W TAC samples (Figures 2B and 2C), i.e., genes upregulated at the expression level were more likely to be regulated at the splicing level as well, for both inclusion and exclusion of internal exons. This trend was not discernable in 4 W TAC samples or for 5′ or 3′ terminal exons (Figure 2B). This result suggests that change of gene expression and regulation of splicing variants are coupled at the early stage of hypertrophy (see Discussion).
We next examined GO terms for genes with significant regulation of alternative usage of exons in hypertrophy. While genes with regulated first and/or last exons did not appear to be biased to any GO terms, those with regulated internal exons were found to be significantly associated with a set of GO terms (Table 1). With respect to BP, ‘cell adhesion’, ‘cell cycle phase’, ‘sensory perception of mechanical stimulus’, and ‘microtubule-based process’ were significant for genes with regulated internal exons in 1 W TAC, and ‘ion transport’ and ‘chromosome organization’ were significant for those with regulated internal exons in 4 W TAC. With respect to CC, ‘cytoskeletal part’ was significant for both 1 W and 4 W TAC conditions, whereas ‘proteinaceous extracellular matrix’ was associated with 1 W TAC only. Since many of these GO terms were also associated with regulation of gene expression level (Figure 2), this result further indicates that there is a connection between transcriptional control and splicing regulation during hypertrophy. In addition, this result suggests that alternative splicing (AS) of internal exons can play an important role in remodeling of the heart under pressure overload.
Correlation of mRNA isoform expression between cardiac hypertrophy and development
We next asked how the isoform regulation in hypertrophy was related to that in development. To this end, we compared exon usage changes in hypertrophy using our exon array data and those in development using the splicing microarray data (EA) . We focused on three types of isoform changes, namely skipping of internal exon (SE), alternative first exon (AFE), and alternative last exon (ALE) (Figure 3).
(A) Comparison of SI for skipped exons (SE). Skipped exons were defined by cDNAs/ESTs. Grey dots are all expressed SE events, and blue dots are the ones that are developmentally regulated and experimentally validated in a previous study . The dotted red line is based on linear regression of blue dots. Pearson correlation coefficient was calculated for each plot, and its E value was estimated based on random sampling of the same number of exons (see Materials and Methods for detail). (B) Venn diagram showing the highly regulated SEs in hypertrophy and development. Highly regulated SEs were selected using 1*SD of SI. (C) and (D) Regulation of alternative first and last exons. U and D are upstream and downstream terminal exons, respectively. Alternative terminal exons were analyzed using difference in log2(U/D) or log2(D/U), where U/D or D/U are ratios of probe set intensities between U and D.
For all the 5,192 exons that can be involved in SE, there was no significant correlation between hypertrophy and development, based on Pearson correlation (r = −0.07 for 1 W TAC and r = 0.02 for 4 W TAC), or on comparison with randomly selected exons (expected value or E = 0.49 for 1 W TAC and E = 0.51 for 4 W TAC, see Materials and Methods for details of the analysis method). However, when we focused on a set of SE events that were previously validated for regulation in development (Table S4) , we detected a significant inverse correlation for 1 W TAC vs. EA (r = −0.46, E = 0.01) but not for 4 W TAC vs. EA (r = −0.17, E = 0.30), indicating a set of splicing events are both regulated in hypertrophy and development but in opposite directions. To further explore this, we used a Venn diagram to examine highly regulated SE events in hypertrophy and development. Indeed, more SE events were inversely regulated than consistently regulated (187 vs. 141, P = 0.014, Fisher's exact test) (Figure 3B).
For the 109 AFE events analyzed, an inverse correlation between development and hypertrophy was clearly discernable (Figure 3C), with 1 W TAC vs. EA (r = −0.23, E = 0.01) being more significant than 4 W TAC vs. EA (r = −0.09, E = 0.15). Several AFE events were highly regulated in LVH and EA with opposite directions, such as A2bp1, Rtn4, Rbm9, and Phactr1, but some appeared to be regulated in LVH only, such as Camk2a and Dusp13. Like AFE events, inverse correlation between development and hypertrophy could also be seen for the 205 ALE events (Figure 3D), and, again, 1 W TAC vs. EA was more significantly correlated (r = −0.12, E = 0.05) than 4 W TAC vs. EA (r = −0.08, E = 0.11). Several ALE events were highly regulated in LVH and EA with opposite directions, such as Asph, Egfr, Pard3, Arpp21, and some were regulated in LVH only, such as Cux1.
Taken together, our result indicates that cardiac hypertrophy involves widespread mRNA isoform changes, and some of the events are oppositely regulated in development, especially for those regulated at the early stage of hypertrophy. These mRNA isoform changes can significantly impact expression of protein isoforms, such as changing protein domains, which are important for adaptation of heart functions under stress overload (See Tables S5, S6, and S7 for top most significantly regulated SE, AFE, and ALE cases in hypertrophy).
Modulation of expression of Fox family genes may play a role in AS regulation in cardiac hypertrophy
The widespread regulation of AS isoforms prompted us to examine expression of splicing factors. To this end, we examined expression changes of genes that were annotated with splicing activities or previously reported to play roles in AS (193 in total), . A number of genes were found to be highly regulated at the mRNA level during hypertrophy, some of which were oppositely regulated in hypertrophy vs. development. Seven genes that were found to be consistently regulated in all 1 W TAC data sets are shown in Figure 4A. We found that Pabpc1, Rbmxrt, and Ptbp1 were upregulated in hypertrophy but downregulated in development, whereas A2bp1 showed the opposite trend. Notably, Rbmxrt, Ptbp1, and A2bp1 all encode proteins (hnRNP G, PTB, and Fox-1, respectively) that have been shown to regulate AS , , .
(A) Gene expression analysis of splicing factors. Splicing factors are those reported in , . Selected genes had fold change >1.5 and P<0.05 (T-test) in at least one of three 1 W TAC samples and had the same direction of regulation across samples. Genes are shown in a heatmap with color representing standardized log2(ratio) based on the color scale shown at the bottom. Grey indicates no detectable expression or not present on microarray. (B) and (C) Regulation of tissue-specific exons in A2bp1 and Rbm9. ‘B’ and ‘M’ represent brain- and muscle-specific exons, respectively, which are expressed in a mutually exclusive manner. Constitutive exons are in blue and alternative ones are in red. Arrows indicate forward (F) and reverse (R) primers for qRT-PCR. Expression changes are based on comparison with Sham samples. Error bars are standard error of the mean (SEM). (D) Splicing regulation for skipped exons with the Fox binding site (TGCATG) located in the upstream or downstream intronic regions adjacent to the exon. Only conserved Fox binding sites were used (see Materials and Methods for detail). Pearson Correlation coefficient and its E value are shown for each graph. (E–F) Experimental validation of skipped exons with Fox binding sites in the downstream intronic region. See Figure S5 for their exon structures. Exon 25 of Fn1 and exon 17 of Itgb1 were analyzed by qRT-PCR and exon 21 of Atp2b1 was examined by semi-quantitative PCR.
Since A2bp1 is highly expressed in the heart  and previous studies implicated its role in heart development , we focused on its expression regulation and impact on AS in hypertrophy. A2bp1 belongs to the gene family encoding Fox proteins, which also includes Rbm9 (Fox-2) and D11Bwg0517e (Fox-3). The Fox proteins all contain an RRM RNA binding motif and Fox-1 and Fox-2 have been shown to have high binding specificity to the consensus sequence UGCAUG , . Both A2bp1 and Rbm9 contain several exons that are differentially used in a tissue-specific manner, including alternative usage of the first exon , , which was also observed in our study (see above). In addition, exons 16 and 17 of A2bp1 and their corresponding exons in Rbm9 (exons 11 and 12) are mutually exclusive exons, which have been found to be highly used in brain and muscle tissues, respectively (Figures 4B and S5). Our exon array data indicated that these two exons were differentially regulated in hypertrophy. To confirm this, we carried out quantitative real-time reverse-transcription PCR (qRT-PCR) using primers targeting two isoforms of A2bp1 or Rbm9. For A2bp1, both isoforms were downregulated, but the muscle isoform was more downregulated than the brain isoform, indicating both transcriptional and AS regulation. For Rbm9, the overall transcript level did not change significantly but the muscle isoform was mildly downregulated and the brain isoform was mildly upregulated, indicating AS regulation only.
In order to examine the role of Fox proteins in regulation of AS in hypertrophy, we selected skipped exons with conserved Fox binding sites (TGCATG) located either upstream (−200 to −1 nt) or downstream (+1 to +200 nt) of the exon, and analyzed their regulation in hypertrophy vs. development (Figure 4D). Interestingly, exons with Fox binding site(s) in the downstream region were significantly regulated in opposite directions between 1 W TAC and EA (r = −0.26, E = 0.05, see Table S8 for the exon list). This trend was not discernable for 4 W TAC samples or for exons with Fox binding sites in the upstream region.
We next focused on several skipped exons with Fox binding sites in the downstream intronic region and carried out validation assays, including exons in Fn1, Itgb1, and Atp2b1. Fn1 and Itgb1 respectively encode fibronectin and an integrin subunit, two cell adhesion molecules whose interaction modulates contraction of cardiac myocytes , . Our microarray data indicated that both genes were upregulated in hypertrophy, and exon 25 of Fn1 and exon b1D of Itgb1 had increased inclusion and exclusion in 1 W TAC, respectively (Figure 4D). Using qRT-PCR with primers targeting exon exclusion and inclusion isoforms (Figure S5), we confirmed microarray results both for their gene expression and AS regulation (Figure 4E). We also examined exon 21 in Atp2b1 (Figure S5), which encodes calcium ion transport ATPase , . By semi-quantitative PCR, we also confirmed its regulation in hypertrophy (Figure 4F).
Taken together, our microarray and validation results indicate that Fox-1 protein is likely to play a role in regulation of AS for a set of exons in both 1 W TAC and development, especially when they bind to the downstream intronic region of a skipped exon.
mRNAs tend to have short 3′UTRs in cardiac hypertrophy
Regulation of 3′UTRs by alternative cleavage and polyadenylation (APA) has recently been found to be associated with cell proliferation, oncogenic transformation, and development , , . To examine how 3′UTRs are regulated in hypertrophy, we took advantage of the fact that a large number of Affymetrix GeneChip probes hybridize to constitutive 3′UTRs (cUTRs) and alternative 3′UTRs (aUTRs) (Figure 5A). cUTRs and aUTRs are defined by polyA sites located in 3′UTRs. We used a score, named relative expression of mRNA isoforms using distal polyA sites (RUD), to indicate the 3′UTR length as illustrated in Figure 5A. RUD was based on comparison of probe intensities for the upstream and downstream regions of a polyA site; a higher score indicates relatively higher abundance of mRNA isoforms resulting from usage of promoter-distal polyA sites, and thus longer 3′UTRs.
(A) Schematic of APA and our method to detect 3′UTR length changes using microarray probes. A hypothetical gene contains three polyA sites, resulting in three 3′UTR isoforms. The 3′UTR regions upstream and downstream of the first polyA site are called constitutive and alternative UTRs, or cUTR and aUTR, respectively. Probes mapped to upstream and downstream of each polyA site were used to calculate the RUD score, which reflects the relative expression of 3′UTR isoforms. The RUD score correlates with 3′UTR length. (B) Comparison of ratio of probe intensity for the downstream region of polyA site to that of upstream region (D/U) of polyA site between Sham and 1 W TAC (LVH3). A total of 4,112 genes were examined. Genes with significant regulation of 3′UTR length (P<0.1, T-test) are colored, with red for genes with 3′UTR lengthened in 1 W TAC and green for those with 3′UTR shortened. (C) Global RUD changes in cardiac development and hypertrophy. The median RUD of all genes in each sample was plotted to represent RUD of the sample. All embryonic samples were compared to E10.5, and TAC samples were compared to Sham. Error bars are standard error of the mean (SEM) based on multiple samples. (D) Venn diagram showing numbers of genes with UTR lengthened or shortened in development and hypertrophy. We selected genes with consistent regulation in at least 4 out of 6 ED samples and in at least 2 out of 3 LVH samples. Fisher's exact test was used to assess significance of the overlap between development and hypertrophy. (E) qRT-PCR analysis of genes with shortened 3′UTRs in TAC. Two sets of PCR primers were designed to target regions upstream and downstream of the first polyA site as shown in (A). The difference in their ratio was used to indicate 3′UTR length changes. PCR primer sequences are listed in Table S2, and gene structures are shown in Figure S6.
As shown in Figure 5B, more genes had higher relative abundance of short 3′UTR isoforms in 1 W TAC as compared to Sham (211 vs. 104), suggesting genes tend to express short 3′UTR isoforms in hypertrophy. Using median RUD score of surveyed genes in each sample, we found 3′UTR shortening took place in all TAC samples (Figure 5C). However, no significant GO terms were found to be associated with genes with shortened 3′UTRs (data not shown), suggesting that 3′UTR isoform regulation is a general feature of hypertrophy affecting all genes.
We next compared 3′UTR regulation in hypertrophy vs. in development. As shown in Figure 5C, compared to E10.5, embryonic development of the heart generally involved more expression of long 3′UTR isoforms, consistent with our previous findings . A significant number of genes were regulated in opposite directions in embryonic development vs. hypertrophy with respect to 3′UTR length regulation (P = 2.4×10−8, Fisher's exact test, Figure 5D), further indicating an inverse correlation of APA isoform regulation between development and hypertrophy.
To validate our microarray results, we selected a number of genes with 3′UTR length changes, including Tia1, Cugbp2, Ube2z, Anapc2, and Cdk1. Tia1 and Cugbp2 encode RNA binding proteins that play roles in various aspects of post-transcriptional regulation , ; Anapc2, and Cdk13 encode proteins involved in cell cycle control; and Ube2z is involved in the ubiquitin-proteasome pathway , , . Using qRT-PCR with primer sets targeting upstream and downstream regions of the first polyA site in 3′UTR (Figure 5A and S6), we confirmed 3′UTR shortening in 1 W TAC in all cases. However, this trend was much less obvious in 4 W TAC except for Cdk13. Taken together, our result indicates that short 3′UTR isoforms are generally more expressed during hypertrophy, particularly at the early stage. Since mRNAs with short 3′UTRs are generally more stable than long 3′UTR isoforms due to evasion of destabilizing elements, this mechanism could lead to stabilization of mRNAs in hypertrophy, resulting in more protein production.
miRNA target genes are globally de-repressed in cardiac hypertrophy
One of the mechanisms regulating mRNA expression via 3′UTRs is miRNA-mediated destabilization of transcripts. To examine how miRNA target genes are regulated in hypertrophy, we compared the regulation profile of miRNA target genes as predicted by TargetScan  with that of other genes for every known miRNA family. As shown in Figure 6A, mRNA targets of several miRNA families were found to be significantly upregulated in hypertrophy (false discovery rate (FDR) <0.05), including those targeted by miR-29, miR-1, miR-9, miR-30, and miR-133. Notably, except for miR-9, these miRNAs are all highly abundant in the heart , indicating that our result has the potential to be physiologically relevant. Interestingly, no miRNAs were predicted to have an enhanced function in hypertrophy, i.e. more repression of target mRNA expression, suggesting that miRNA activity is generally inhibited in cardiac hypertrophy. We also compared hypertrophy with development with respect to these miRNAs. As shown in Figure 6A, mRNA targets of all these miRNAs were downregulated in development, but their significance depended on the developmental stage. For example, miR-29 target genes were more downregulated in postnatal development, whereas miR-1 target genes were mainly downregulated in embryonic development (Figure 6B). This result indicates that miRNAs might play an important role in defining the gene expression program in hypertrophy and contribute to the activation of the ‘fetal gene program’. In addition, combined with our observation that 3′UTRs are generally shortened in cardiac hypertrophy, the miRNA result suggests that inhibition of mRNA stability and/or translation is generally avoided in hypertrophy.
(A) Heatmap of miRNAs whose target genes were significantly regulated in hypertrophy (FDR<0.05). The color represents p-value which was calculated by comparison of CDF curves (see Materials and Methods for detail). The p-values for development datasets were also shown for comparison. miRNAs are sorted by their significance in 1 W TAC. (B) Expression changes of the genes targeted by miR-29 and miR-1. CDF curves of standardized log2(ratio) values are shown for ED, PD, and 1 W TAC samples. ED and PD are based on averaged values across ED and PD samples, respectively. 1 W TAC is based on averaged values of three samples (LVH1–3). P-values indicating difference between CDFs of miRNA target genes and other genes were derived from the Kolmogorov–Smirnov test.
Our comprehensive analysis of microarray data from multiple studies identified genes and pathways significantly and robustly regulated in cardiac hypertrophy. These results, while generally consistent with previous studies , , , , systematically refine and extend the current understanding of regulation of transcriptome in cardiac hypertrophy. Comparison of hypertrophy with embryonic and postnatal development enabled us to systematically define the fetal gene program. In addition, our genome-wide exon-level analysis indicates that regulation of mRNA isoform is widespread during cardiac hypertrophy. While some isoform changes are hypertrophy-specific, other events are negatively correlated with those in development, particularly for the events regulated at the early stage of hypertrophy, suggesting activation of a fetal post-transcriptional program in the heart in response to pressure overload.
Our GO analysis identified a number of pathways that are regulated in both development and hypertrophy. For some pathways, regulation in development and in hypertrophy involves the same set of genes, for example genes related to mitochondrion functions. This can be discerned using gene density plots (Figure S4). However, for some other pathways, regulation of a pathway in hypertrophy involves a different set of genes than in development. For example, despite cell cycle genes as a set are upregulated in hypertrophy and downregulated in development, regulation of individual genes is quite different in these two conditions, as indicated in the gene density plot (Figure S4). This is consistent with the notion that development involves cell proliferation whereas hypertrophy involves expansion of cell size. In addition, gene regulation for some pathways is quite complex, particularly genes related to ECM (for example, proteinaceous extracellular matrix). They are globally upregulated in hypertrophy, but in development only a subset of them is downregulated and another subset is upregulated.
We found that genes upregulated at the expression level are also more likely to be regulated at the splicing level in 1 W TAC, suggesting coupling splicing regulation to transcription. This observation is not due to technical reasons, because this coupling is not discernable in 4 W TAC samples, and we corrected splicing index values for hybridization biases related to gene expression changes . Regulation of AS by transcription can be attributable to difference in recruitment of splicing factors to RNA polymerase II (Pol II), known as recruitment coupling, or change in Pol II elongation rate, known as kinetic coupling , . On this note, Cdk9, which encodes a Pol II elongation factor, has been shown to be regulated in hypertrophy . Further studies are needed to unravel this phenomenon.
Our expression analysis indicates that a number of splicing factors are regulated in hypertrophy, which presumably contributes to global AS regulation. We focused on the gene encoding Fox-1, which, interestingly, is significantly regulated at several levels, including expression, alternative initiation, and AS. While motif analysis did not reveal an enrichment of the Fox binding site near regulated exons (data not shown), we did observe a negative correlation between hypertrophy and development for the skipped exons with the Fox binding site located in downstream of exon (Figure 4D). This result suggests that regulation of Fox-1 may play a role in establishing the fetal splicing program in the hypertrophied heart. Interestingly, PTB, which was previously found to interact with Fox in AS regulation , , is significantly upregulated at the mRNA level. How these two proteins function and interact with each other in hypertrophy is to be further examined. In addition, given that posttranscriptional and posttranslational regulation of splicing factors is very common, it is highly possible that other splicing factors may also play roles in AS regulation in hypertrophy despite that their mRNA levels do not significantly change as found in this study.
Diversity in 3′ end formation of mRNA has recently been highlighted as an important layer of gene regulation , , . Previous studies have reported general 3′UTR shortening in proliferation and oncogenic transformation , . Here we extend this phenomenon to cardiac hypertrophy. It is possible that mechanisms of 3′UTR regulation in hypertrophy and proliferation are related, because 1) a set of cell cycle genes are upregulated in hypertrophy, and 2) we previously found that pre-mRNA processing genes typically have binding sites for cell cycle-related transcription factors in their promoters . Consistent with this notion, we found that mRNAs of CstF proteins, which have been shown previously to play a role in regulation of APA , were upregulated after 3 days of TAC (Figure S7). Interestingly, their upregulation was subdued after 1 week of TAC. Therefore, 3′UTR isoform regulation may be executed at a very early stage of hypertrophy and the effect remains during TAC.
Shortening of 3′UTR can lead to more protein production by avoiding destabilizing cis elements in the 3′UTR, such as miRNA target sites, AREs and GREs. Consistently, despite reports implicating upregulation of some miRNAs in hypertrophy , we found most miRNA target genes were upregulated during hypertrophy, indicating a global de-repression of miRNA targeting. These results suggest coordinated mechanisms to stabilize mRNAs in hypertrophy, which can be important for rapid enlargement of heart in response to mechanical stress.
Materials and Methods
Exon array of TAC samples
All animal work has been conducted according to a protocol approved by the Institutional Animal Care and Use Committee (IACUC) at UMDNJ-New Jersey Medical School. Protocol number: 07120. Transverse Aortic Constriction (TAC) was conducted as described previously . Briefly, aortic constriction was performed by ligation of the transverse aorta with a 28-gauge needle using a 7-0 braided polyester suture. Sham operation was performed without constricting the aorta. Two mice (C57BL/6) were used for 1 W and 4 W TAC and Sham. Total RNA from left ventricles was extracted using the RNeasy Fibrous Tissue Mini Kit (QIAGEN) and was processed by the Whole Transcript (WT) Sense Target Labeling kit (Affymetrix) for exon array analysis (Affymetrix GeneChip Exon Mouse 1.0 ST Array). Our data is deposited in the GEO database of NCBI (GSE: 24242) and is MIAME compliant. See Table S1 for details on the public datasets used in this study.
Microarray data analysis
Microarray data normalization was carried out using the robust multi-chip analysis (RMA) method. To determine whether a gene was expressed, we used the MAS 5.0 method for 3′ arrays and the detection above background (DABG) method for exon arrays. Only genes with detectable expression in >50% of samples in a comparing group were used for further analysis. For the splicing microarray, we mapped probes to exons of gene models based on cDNAs/ESTs, as described previously . Gene expression was calculated using constitutively expressed exons only.
Gene set analysis
Comparison of gene sets for expression changes was based on the cumulative distribution function (CDF)(See Text S1 for detail). Gene Ontology information was obtained from the NCBI Gene database. miRNA target sites were obtained from the TargetScan database (v5.1) . To evaluate FDR for our selection at a given p-value, three datasets (LVH1, LVH2, LVH3) were randomized (sample-wise), and the number of falsely identified entries at a given p-value was calculated.
mRNA isoform analysis
Exon usage was analyzed using the splicing index (SI) method, which is based on expression change of a given exon vs. expression change of its corresponding gene. Since different probes can respond to transcript changes differently, SI values were adjusted by the Corrected Splicing Index for Exon Arrays (COSIE) method . Classification of first, last, and internal exons was based on the RefSeq database. Skipped exons were identified using cDNAs/ESTs . Fox binding sites were analyzed by a method similar to the one previously reported . Expected value (E) for Pearson Correlation for a given set of exons corresponds to the fraction of times (out of 1,000 times) that a randomly selected set including the same number of exons has a better correlation. Analysis of Fox binding sites was based on the method previously developed . Information about alternative polyA sites was retrieved from PolyA_DB . Analysis of APA isoforms was carried out by the RUD method, as previously described .
To validate microarray data, we used real-time reverse transcription PCR (qRT-PCR) or semi-quantitative RT-PCR (see Table S2 for primers used in this study). PCR was carried out using the SYBR Green method with Gapdh mRNA as an internal control. For semi-quantitative RT-PCR, primer sets targeting constitutive exons flanking a skipped exon were used and PCR products were analyzed by ImageJ (http://rsbweb.nih.gov/ij/index.html).
Standardization of log2(ratio) makes data more comparable across datasets. (A) Distribution of gene expression changes (log2(ratio)) of all datasets used in this study. Development samples are shown in green and hypertrophy samples are shown in orange. (B) Distribution of gene expression changes after standardization.
Correlation between hypertrophy and development samples. Pair-wise Pearson Correlation coefficients between samples are shown in a heatmap according to the color scale shown in the graph.
Analysis gene sets using expression change profiles. (A) An example of gene set analysis using cumulative distribution function (CDF). Genes annotated with “proteinaceous extracellular matrix” form a gene set. The black line is CDF curve for genes in the set, and the dotted line is CDF curve for other genes on the microarray. The difference between two CDF curves, or ΔCDF, is indicated by color, with green for negative values and red for positive ones. ΔCDF was used to examine regulation of a gene set, which yielded two p-values, one for positive values or ΔCDF(+), representing upregulation, and one for negative values or ΔCDF(-), representing downregulation (see Supplementary Materials and Methods, Text S1, for detail). (B) Log2(gene set size) vs. log2(SD of ΔCDF). A linear regression line is shown in red, and its R2 is indicated in the graph. The data is derived from randomly sampled genes (10,000 times for each gene set size). (C) Distribution of standardized ΔCDF follows the exponential distribution with rate (λ) = 1.1. The thick black line is distribution of standardized ΔCDF and the dotted red line is exponential distribution. The two distributions are not significantly different when the values are >2 (P = 0.29, Kolmogorov-Smirnov test).
Gene density plots for several significant GO terms. (A) An example of the gene density plot showing correlation between hypertrophy and development. X-axis is log2(ratio) in development, and y-axis is averaged log2(ratio) in three 1W TAC samples. The horizontal and vertical white lines in each plot mark the point with log2(ratio) = 0. (B) Gene density plots for top 5 BP (top) and CC (bottom) GO terms. See Supplementary Materials and Methods (Text S1) for details of the method.
Information about validated genes with AS regulation in hypertrophy. For each gene, a gene structure image is derived from UCSC genome browser. Conservation based on 17 vertebrate species is indicated in the magnified image. A schematic indicating splicing pattern is shown at the bottom. Exon numbers are indicated. Constitutive exons are shown in blue and alternative ones are in red. PCR primers are shown as arrows.
Information about validated genes with APA regulation in hypertrophy. For each gene, a gene structure image is derived from UCSC genome browser. Conservation based on 17 vertebrate species is indicated in the magnified image. PolyA site is shown as pA; PCR primers are shown as arrows.
Regulation of Cstf genes in hypertrophy. qRT-PCR data are shown for Cstf1, Cstf2, and Cstf3 in Sham, 3 day(D) TAC, and 1W TAC. Error bars are standard deviation based on 3 mice.
Supplementary materials and methods and references for supporting information.
Datasets used in this study.
Primers used in this study.
Genes significantly regulated in hypertrophy and/or development.
Forty-five developmentally regulated skipped exons.
Top 50 most significantly regulated exon skipping events in hypertrophy.
Top 5 regulated AFE events in hypertrophy.
Top 5 regulated ALE events in hypertrophy.
Skipped exons with Fox binding sites in the downstream intronic region.
We thank John Castle for sharing the heart development data, and Maha Abdellatif, Pranela Rameshwar, and members of BT lab for helpful discussions.
Conceived and designed the experiments: JYP BT. Performed the experiments: JYP WL DZ PZ YZ TM. Analyzed the data: JYP WL DZ BT. Contributed reagents/materials/analysis tools: SFV JS. Wrote the paper: JYP BT.
- 1. Olson EN (2004) A decade of discoveries in cardiac biology. Nat Med 10: 467–474.EN Olson2004A decade of discoveries in cardiac biology.Nat Med10467474
- 2. Tardiff JC (2006) Cardiac hypertrophy: stressing out the heart. J Clin Invest 116: 1467–1470.JC Tardiff2006Cardiac hypertrophy: stressing out the heart.J Clin Invest11614671470
- 3. Hill JA, Olson EN (2008) Cardiac plasticity. N Engl J Med 358: 1370–1380.JA HillEN Olson2008Cardiac plasticity.N Engl J Med35813701380
- 4. Cerutti C, Kurdi M, Bricca G, Hodroj W, Paultre C, et al. (2006) Transcriptional alterations in the left ventricle of three hypertensive rat models. Physiol Genomics 27: 295–308.C. CeruttiM. KurdiG. BriccaW. HodrojC. Paultre2006Transcriptional alterations in the left ventricle of three hypertensive rat models.Physiol Genomics27295308
- 5. Sheehy SP, Huang S, Parker KK (2009) Time-warped comparison of gene expression in adaptive and maladaptive cardiac hypertrophy. Circ Cardiovasc Genet 2: 116–124.SP SheehyS. HuangKK Parker2009Time-warped comparison of gene expression in adaptive and maladaptive cardiac hypertrophy.Circ Cardiovasc Genet2116124
- 6. Zhao M, Chow A, Powers J, Fajardo G, Bernstein D (2004) Microarray analysis of gene expression after transverse aortic constriction in mice. Physiol Genomics 19: 93–105.M. ZhaoA. ChowJ. PowersG. FajardoD. Bernstein2004Microarray analysis of gene expression after transverse aortic constriction in mice.Physiol Genomics1993105
- 7. Rajabi M, Kassiotis C, Razeghi P, Taegtmeyer H (2007) Return to the fetal gene program protects the stressed heart: a strong hypothesis. Heart Fail Rev 12: 331–343.M. RajabiC. KassiotisP. RazeghiH. Taegtmeyer2007Return to the fetal gene program protects the stressed heart: a strong hypothesis.Heart Fail Rev12331343
- 8. Barry SP, Davidson SM, Townsend PA (2008) Molecular regulation of cardiac hypertrophy. Int J Biochem Cell Biol 40: 2023–2039.SP BarrySM DavidsonPA Townsend2008Molecular regulation of cardiac hypertrophy.Int J Biochem Cell Biol4020232039
- 9. Oka T, Xu J, Molkentin JD (2007) Re-employment of developmental transcription factors in adult heart disease. Seminars in Cell & Developmental Biology 18: 117–131.T. OkaJ. XuJD Molkentin2007Re-employment of developmental transcription factors in adult heart disease.Seminars in Cell & Developmental Biology18117131
- 10. Akazawa H, Komuro I (2003) Roles of cardiac transcription factors in cardiac hypertrophy. Circ Res 92: 1079–1088.H. AkazawaI. Komuro2003Roles of cardiac transcription factors in cardiac hypertrophy.Circ Res9210791088
- 11. Care A, Catalucci D, Felicetti F, Bonci D, Addario A, et al. (2007) MicroRNA-133 controls cardiac hypertrophy. Nat Med 13: 613–618.A. CareD. CatalucciF. FelicettiD. BonciA. Addario2007MicroRNA-133 controls cardiac hypertrophy.Nat Med13613618
- 12. Sayed D, Hong C, Chen IY, Lypowy J, Abdellatif M (2007) MicroRNAs play an essential role in the development of cardiac hypertrophy. Circ Res 100: 416–424.D. SayedC. HongIY ChenJ. LypowyM. Abdellatif2007MicroRNAs play an essential role in the development of cardiac hypertrophy.Circ Res100416424
- 13. van Rooij E, Sutherland LB, Qi X, Richardson JA, Hill J, et al. (2007) Control of stress-dependent cardiac growth and gene expression by a microRNA. Science 316: 575–579.E. van RooijLB SutherlandX. QiJA RichardsonJ. Hill2007Control of stress-dependent cardiac growth and gene expression by a microRNA.Science316575579
- 14. Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116: 281–297.DP Bartel2004MicroRNAs: genomics, biogenesis, mechanism, and function.Cell116281297
- 15. Guo H, Ingolia NT, Weissman JS, Bartel DP (2010) Mammalian microRNAs predominantly act to decrease target mRNA levels. Nature 466: 835–840.H. GuoNT IngoliaJS WeissmanDP Bartel2010Mammalian microRNAs predominantly act to decrease target mRNA levels.Nature466835840
- 16. Cordes KR, Srivastava D (2009) MicroRNA regulation of cardiovascular development. Circ Res 104: 724–732.KR CordesD. Srivastava2009MicroRNA regulation of cardiovascular development.Circ Res104724732
- 17. Rao PK, Toyama Y, Chiang HR, Gupta S, Bauer M, et al. (2009) Loss of cardiac microRNA-mediated regulation leads to dilated cardiomyopathy and heart failure. Circ Res 105: 585–594.PK RaoY. ToyamaHR ChiangS. GuptaM. Bauer2009Loss of cardiac microRNA-mediated regulation leads to dilated cardiomyopathy and heart failure.Circ Res105585594
- 18. Small EM, Frost RJ, Olson EN (2010) MicroRNAs add a new dimension to cardiovascular disease. Circulation 121: 1022–1032.EM SmallRJ FrostEN Olson2010MicroRNAs add a new dimension to cardiovascular disease.Circulation12110221032
- 19. Matkovich SJ, Van Booven DJ, Youker KA, Torre-Amione G, Diwan A, et al. (2009) Reciprocal regulation of myocardial microRNAs and messenger RNA in human cardiomyopathy and reversal of the microRNA signature by biomechanical support. Circulation 119: 1263–1271.SJ MatkovichDJ Van BoovenKA YoukerG. Torre-AmioneA. Diwan2009Reciprocal regulation of myocardial microRNAs and messenger RNA in human cardiomyopathy and reversal of the microRNA signature by biomechanical support.Circulation11912631271
- 20. Wang ET, Sandberg R, Luo S, Khrebtukova I, Zhang L, et al. (2008) Alternative isoform regulation in human tissue transcriptomes. Nature 456: 470–476.ET WangR. SandbergS. LuoI. KhrebtukovaL. Zhang2008Alternative isoform regulation in human tissue transcriptomes.Nature456470476
- 21. Pan Q, Shai O, Lee LJ, Frey BJ, Blencowe BJ (2008) Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet 40: 1413–1415.Q. PanO. ShaiLJ LeeBJ FreyBJ Blencowe2008Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing.Nat Genet4014131415
- 22. Yamazaki T, Walchli S, Fujita T, Ryser S, Hoshijima M, et al. (2010) Splice variants of enigma homolog, differentially expressed during heart development, promote or prevent hypertrophy. Cardiovasc Res 86: 374–382.T. YamazakiS. WalchliT. FujitaS. RyserM. Hoshijima2010Splice variants of enigma homolog, differentially expressed during heart development, promote or prevent hypertrophy.Cardiovasc Res86374382
- 23. Liao P, Yong TF, Liang MC, Yue DT, Soong TW (2005) Splicing for alternative structures of Cav1.2 Ca2+ channels in cardiac and smooth muscles. Cardiovasc Res 68: 197–203.P. LiaoTF YongMC LiangDT YueTW Soong2005Splicing for alternative structures of Cav1.2 Ca2+ channels in cardiac and smooth muscles.Cardiovasc Res68197203
- 24. Lahmers S, Wu Y, Call DR, Labeit S, Granzier H (2004) Developmental control of titin isoform expression and passive stiffness in fetal and neonatal myocardium. Circ Res 94: 505–513.S. LahmersY. WuDR CallS. LabeitH. Granzier2004Developmental control of titin isoform expression and passive stiffness in fetal and neonatal myocardium.Circ Res94505513
- 25. Kalsotra A, Xiao X, Ward AJ, Castle JC, Johnson JM, et al. (2008) A postnatal switch of CELF and MBNL proteins reprograms alternative splicing in the developing heart. Proc Natl Acad Sci U S A 105: 20333–20338.A. KalsotraX. XiaoAJ WardJC CastleJM Johnson2008A postnatal switch of CELF and MBNL proteins reprograms alternative splicing in the developing heart.Proc Natl Acad Sci U S A1052033320338
- 26. Salomonis N, Nelson B, Vranizan K, Pico AR, Hanspers K, et al. (2009) Alternative splicing in the differentiation of human embryonic stem cells into cardiac precursors. PLoS Comput Biol 5: e1000553.N. SalomonisB. NelsonK. VranizanAR PicoK. Hanspers2009Alternative splicing in the differentiation of human embryonic stem cells into cardiac precursors.PLoS Comput Biol5e1000553
- 27. Kong SW, Hu YW, Ho JW, Ikeda S, Polster S, et al. (2010) Heart failure-associated changes in RNA splicing of sarcomere genes. Circ Cardiovasc Genet 3: 138–146.SW KongYW HuJW HoS. IkedaS. Polster2010Heart failure-associated changes in RNA splicing of sarcomere genes.Circ Cardiovasc Genet3138146
- 28. Ding JH, Xu X, Yang D, Chu PH, Dalton ND, et al. (2004) Dilated cardiomyopathy caused by tissue-specific ablation of SC35 in the heart. Embo J 23: 885–896.JH DingX. XuD. YangPH ChuND Dalton2004Dilated cardiomyopathy caused by tissue-specific ablation of SC35 in the heart.Embo J23885896
- 29. Xu X, Yang D, Ding JH, Wang W, Chu PH, et al. (2005) ASF/SF2-regulated CaMKIIdelta alternative splicing temporally reprograms excitation-contraction coupling in cardiac muscle. Cell 120: 59–72.X. XuD. YangJH DingW. WangPH Chu2005ASF/SF2-regulated CaMKIIdelta alternative splicing temporally reprograms excitation-contraction coupling in cardiac muscle.Cell1205972
- 30. Ladd AN, Taffet G, Hartley C, Kearney DL, Cooper TA (2005) Cardiac tissue-specific repression of CELF activity disrupts alternative splicing and causes cardiomyopathy. Mol Cell Biol 25: 6267–6278.AN LaddG. TaffetC. HartleyDL KearneyTA Cooper2005Cardiac tissue-specific repression of CELF activity disrupts alternative splicing and causes cardiomyopathy.Mol Cell Biol2562676278
- 31. Feng Y, Valley MT, Lazar J, Yang AL, Bronson RT, et al. (2009) SRp38 Regulates Alternative Splicing and Is Required for Ca2+ Handling in the Embryonic Heart. Developmental Cell 16: 528–538.Y. FengMT ValleyJ. LazarAL YangRT Bronson2009SRp38 Regulates Alternative Splicing and Is Required for Ca2+ Handling in the Embryonic Heart.Developmental Cell16528538
- 32. Tian B, Hu J, Zhang H, Lutz CS (2005) A large-scale analysis of mRNA polyadenylation of human and mouse genes. Nucleic Acids Res 33: 201–212.B. TianJ. HuH. ZhangCS Lutz2005A large-scale analysis of mRNA polyadenylation of human and mouse genes.Nucleic Acids Res33201212
- 33. Zhang H, Lee JY, Tian B (2005) Biased alternative polyadenylation in human tissues. Genome Biol 6: R100.H. ZhangJY LeeB. Tian2005Biased alternative polyadenylation in human tissues.Genome Biol6R100
- 34. Ji Z, Lee JY, Pan Z, Jiang B, Tian B (2009) Progressive lengthening of 3′ untranslated regions of mRNAs by alternative polyadenylation during mouse embryonic development. Proc Natl Acad Sci U S A 106: 7028–7033.Z. JiJY LeeZ. PanB. JiangB. Tian2009Progressive lengthening of 3′ untranslated regions of mRNAs by alternative polyadenylation during mouse embryonic development.Proc Natl Acad Sci U S A10670287033
- 35. Sandberg R, Neilson JR, Sarma A, Sharp PA, Burge CB (2008) Proliferating cells express mRNAs with shortened 3′ untranslated regions and fewer microRNA target sites. Science 320: 1643–1647.R. SandbergJR NeilsonA. SarmaPA SharpCB Burge2008Proliferating cells express mRNAs with shortened 3′ untranslated regions and fewer microRNA target sites.Science32016431647
- 36. Mayr C, Bartel DP (2009) Widespread shortening of 3′UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells. Cell 138: 673–684.C. MayrDP Bartel2009Widespread shortening of 3′UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells.Cell138673684
- 37. Ji Z, Tian B (2009) Reprogramming of 3′ untranslated regions of mRNAs by alternative polyadenylation in generation of pluripotent stem cells from different cell types. PLoS One 4: e8419.Z. JiB. Tian2009Reprogramming of 3′ untranslated regions of mRNAs by alternative polyadenylation in generation of pluripotent stem cells from different cell types.PLoS One4e8419
- 38. Gabut M, Chaudhry S, Blencowe BJ (2008) SnapShot: The splicing regulatory machinery. Cell 133: 192 e191.M. GabutS. ChaudhryBJ Blencowe2008SnapShot: The splicing regulatory machinery.Cell133192 e191
- 39. Jurica MS, Moore MJ (2003) Pre-mRNA splicing: awash in a sea of proteins. Mol Cell 12: 5–14.MS JuricaMJ Moore2003Pre-mRNA splicing: awash in a sea of proteins.Mol Cell12514
- 40. Kuroyanagi H (2009) Fox-1 family of RNA-binding proteins. Cell Mol Life Sci 66: 3895–3907.H. Kuroyanagi2009Fox-1 family of RNA-binding proteins.Cell Mol Life Sci6638953907
- 41. Heinrich B, Zhang Z, Raitskin O, Hiller M, Benderska N, et al. (2009) Heterogeneous nuclear ribonucleoprotein G regulates splice site selection by binding to CC(A/C)-rich regions in pre-mRNA. J Biol Chem 284: 14303–14315.B. HeinrichZ. ZhangO. RaitskinM. HillerN. Benderska2009Heterogeneous nuclear ribonucleoprotein G regulates splice site selection by binding to CC(A/C)-rich regions in pre-mRNA.J Biol Chem2841430314315
- 42. Xue Y, Zhou Y, Wu T, Zhu T, Ji X, et al. (2009) Genome-wide analysis of PTB-RNA interactions reveals a strategy used by the general splicing repressor to modulate exon inclusion or skipping. Mol Cell 36: 996–1006.Y. XueY. ZhouT. WuT. ZhuX. Ji2009Genome-wide analysis of PTB-RNA interactions reveals a strategy used by the general splicing repressor to modulate exon inclusion or skipping.Mol Cell369961006
- 43. Underwood JG, Boutz PL, Dougherty JD, Stoilov P, Black DL (2005) Homologues of the Caenorhabditis elegans Fox-1 protein are neuronal splicing regulators in mammals. Mol Cell Biol 25: 10005–10016.JG UnderwoodPL BoutzJD DoughertyP. StoilovDL Black2005Homologues of the Caenorhabditis elegans Fox-1 protein are neuronal splicing regulators in mammals.Mol Cell Biol251000510016
- 44. Jin Y, Suzuki H, Maegawa S, Endo H, Sugano S, et al. (2003) A vertebrate RNA-binding protein Fox-1 regulates tissue-specific splicing via the pentanucleotide GCAUG. Embo Journal 22: 905–912.Y. JinH. SuzukiS. MaegawaH. EndoS. Sugano2003A vertebrate RNA-binding protein Fox-1 regulates tissue-specific splicing via the pentanucleotide GCAUG.Embo Journal22905912
- 45. Yeo GW, Coufal NG, Liang TY, Peng GE, Fu XD, et al. (2009) An RNA code for the FOX2 splicing regulator revealed by mapping RNA-protein interactions in stem cells. Nat Struct Mol Biol 16: 130–137.GW YeoNG CoufalTY LiangGE PengXD Fu2009An RNA code for the FOX2 splicing regulator revealed by mapping RNA-protein interactions in stem cells.Nat Struct Mol Biol16130137
- 46. Yang G, Huang SC, Wu JY, Benz EJ Jr (2008) Regulated Fox-2 isoform expression mediates protein 4.1R splicing during erythroid differentiation. Blood 111: 392–401.G. YangSC HuangJY WuEJ Benz Jr2008Regulated Fox-2 isoform expression mediates protein 4.1R splicing during erythroid differentiation.Blood111392401
- 47. Wu X, Sun Z, Foskett A, Trzeciakowski JP, Meininger GA, et al. (2010) Cardiomyocyte contractile status is associated with differences in fibronectin and integrin interactions. Am J Physiol Heart Circ Physiol 298: H2071–2081.X. WuZ. SunA. FoskettJP TrzeciakowskiGA Meininger2010Cardiomyocyte contractile status is associated with differences in fibronectin and integrin interactions.Am J Physiol Heart Circ Physiol298H20712081
- 48. Brancaccio M, Hirsch E, Notte A, Selvetella G, Lembo G, et al. (2006) Integrin signalling: the tug-of-war in heart hypertrophy. Cardiovasc Res 70: 422–433.M. BrancaccioE. HirschA. NotteG. SelvetellaG. Lembo2006Integrin signalling: the tug-of-war in heart hypertrophy.Cardiovasc Res70422433
- 49. Lee JA, Tang ZZ, Black DL (2009) An inducible change in Fox-1/A2BP1 splicing modulates the alternative splicing of downstream neuronal target exons. Genes Dev 23: 2284–2293.JA LeeZZ TangDL Black2009An inducible change in Fox-1/A2BP1 splicing modulates the alternative splicing of downstream neuronal target exons.Genes Dev2322842293
- 50. Strehler EE, Filoteo AG, Penniston JT, Caride AJ (2007) Plasma-membrane Ca(2+) pumps: structural diversity as the basis for functional versatility. Biochem Soc Trans 35: 919–922.EE StrehlerAG FiloteoJT PennistonAJ Caride2007Plasma-membrane Ca(2+) pumps: structural diversity as the basis for functional versatility.Biochem Soc Trans35919922
- 51. Yamasaki S, Stoecklin G, Kedersha N, Simarro M, Anderson P (2007) T-cell intracellular antigen-1 (TIA-1)-induced translational silencing promotes the decay of selected mRNAs. J Biol Chem 282: 30070–30077.S. YamasakiG. StoecklinN. KedershaM. SimarroP. Anderson2007T-cell intracellular antigen-1 (TIA-1)-induced translational silencing promotes the decay of selected mRNAs.J Biol Chem2823007030077
- 52. Barreau C, Paillard L, Mereau A, Osborne HB (2006) Mammalian CELF/Bruno-like RNA-binding proteins: molecular characteristics and biological functions. Biochimie 88: 515–525.C. BarreauL. PaillardA. MereauHB Osborne2006Mammalian CELF/Bruno-like RNA-binding proteins: molecular characteristics and biological functions.Biochimie88515525
- 53. Ye Y, Rape M (2009) Building ubiquitin chains: E2 enzymes at work. Nat Rev Mol Cell Biol 10: 755–764.Y. YeM. Rape2009Building ubiquitin chains: E2 enzymes at work.Nat Rev Mol Cell Biol10755764
- 54. Peters JM (2006) The anaphase promoting complex/cyclosome: a machine designed to destroy. Nat Rev Mol Cell Biol 7: 644–656.JM Peters2006The anaphase promoting complex/cyclosome: a machine designed to destroy.Nat Rev Mol Cell Biol7644656
- 55. Malumbres M, Harlow E, Hunt T, Hunter T, Lahti JM, et al. (2009) Cyclin-dependent kinases: a family portrait. Nat Cell Biol 11: 1275–1276.M. MalumbresE. HarlowT. HuntT. HunterJM Lahti2009Cyclin-dependent kinases: a family portrait.Nat Cell Biol1112751276
- 56. Friedman RC, Farh KK, Burge CB, Bartel DP (2009) Most mammalian mRNAs are conserved targets of microRNAs. Genome Res 19: 92–105.RC FriedmanKK FarhCB BurgeDP Bartel2009Most mammalian mRNAs are conserved targets of microRNAs.Genome Res1992105
- 57. Lehman JJ, Kelly DP (2002) Transcriptional activation of energy metabolic switches in the developing and hypertrophied heart. Clin Exp Pharmacol Physiol 29: 339–345.JJ LehmanDP Kelly2002Transcriptional activation of energy metabolic switches in the developing and hypertrophied heart.Clin Exp Pharmacol Physiol29339345
- 58. Bowers SL, Banerjee I, Baudino TA (2010) The extracellular matrix: at the center of it all. Journal of Molecular and Cellular Cardiology 48: 474–482.SL BowersI. BanerjeeTA Baudino2010The extracellular matrix: at the center of it all.Journal of Molecular and Cellular Cardiology48474482
- 59. Ahuja P, Sdek P, MacLellan WR (2007) Cardiac myocyte cell cycle control in development, disease, and regeneration. Physiol Rev 87: 521–544.P. AhujaP. SdekWR MacLellan2007Cardiac myocyte cell cycle control in development, disease, and regeneration.Physiol Rev87521544
- 60. Taegtmeyer H, Sen S, Vela D (2010) Return to the fetal gene program: a suggested metabolic link to gene expression in the heart. Ann N Y Acad Sci 1188: 191–198.H. TaegtmeyerS. SenD. Vela2010Return to the fetal gene program: a suggested metabolic link to gene expression in the heart.Ann N Y Acad Sci1188191198
- 61. Gaidatzis D, Jacobeit K, Oakeley EJ, Stadler MB (2009) Overestimation of alternative splicing caused by variable probe characteristics in exon arrays. Nucleic Acids Res 37: e107.D. GaidatzisK. JacobeitEJ OakeleyMB Stadler2009Overestimation of alternative splicing caused by variable probe characteristics in exon arrays.Nucleic Acids Res37e107
- 62. de la Mata M, Kornblihtt AR (2006) RNA polymerase II C-terminal domain mediates regulation of alternative splicing by SRp20. Nat Struct Mol Biol 13: 973–980.M. de la MataAR Kornblihtt2006RNA polymerase II C-terminal domain mediates regulation of alternative splicing by SRp20.Nat Struct Mol Biol13973980
- 63. Batsche E, Yaniv M, Muchardt C (2006) The human SWI/SNF subunit Brm is a regulator of alternative splicing. Nat Struct Mol Biol 13: 22–29.E. BatscheM. YanivC. Muchardt2006The human SWI/SNF subunit Brm is a regulator of alternative splicing.Nat Struct Mol Biol132229
- 64. Sano M, Abdellatif M, Oh H, Xie M, Bagella L, et al. (2002) Activation and function of cyclin T-Cdk9 (positive transcription elongation factor-b) in cardiac muscle-cell hypertrophy. Nat Med 8: 1310–1317.M. SanoM. AbdellatifH. OhM. XieL. Bagella2002Activation and function of cyclin T-Cdk9 (positive transcription elongation factor-b) in cardiac muscle-cell hypertrophy.Nat Med813101317
- 65. Das D, Clark TA, Schweitzer A, Yamamoto M, Marr H, et al. (2007) A correlation with exon expression approach to identify cis-regulatory elements for tissue-specific alternative splicing. Nucleic Acids Res 35: 4845–4857.D. DasTA ClarkA. SchweitzerM. YamamotoH. Marr2007A correlation with exon expression approach to identify cis-regulatory elements for tissue-specific alternative splicing.Nucleic Acids Res3548454857
- 66. Jin Y, Suzuki H, Maegawa S, Endo H, Sugano S, et al. (2003) A vertebrate RNA-binding protein Fox-1 regulates tissue-specific splicing via the pentanucleotide GCAUG. Embo J 22: 905–912.Y. JinH. SuzukiS. MaegawaH. EndoS. Sugano2003A vertebrate RNA-binding protein Fox-1 regulates tissue-specific splicing via the pentanucleotide GCAUG.Embo J22905912
- 67. Lutz CS (2008) Alternative polyadenylation: a twist on mRNA 3′ end formation. ACS Chem Biol 3: 609–617.CS Lutz2008Alternative polyadenylation: a twist on mRNA 3′ end formation.ACS Chem Biol3609617
- 68. Neilson JR, Sandberg R (2010) Heterogeneity in mammalian RNA 3′ end formation. Exp Cell Res 316: 1357–1364.JR NeilsonR. Sandberg2010Heterogeneity in mammalian RNA 3′ end formation.Exp Cell Res31613571364
- 69. Millevoi S, Vagner S (2010) Molecular mechanisms of eukaryotic pre-mRNA 3′ end processing regulation. Nucleic Acids Res 38: 2757–2774.S. MillevoiS. Vagner2010Molecular mechanisms of eukaryotic pre-mRNA 3′ end processing regulation.Nucleic Acids Res3827572774
- 70. Takagaki Y, Seipelt RL, Peterson ML, Manley JL (1996) The polyadenylation factor CstF-64 regulates alternative processing of IgM heavy chain pre-mRNA during B cell differentiation. Cell 87: 941–952.Y. TakagakiRL SeipeltML PetersonJL Manley1996The polyadenylation factor CstF-64 regulates alternative processing of IgM heavy chain pre-mRNA during B cell differentiation.Cell87941952
- 71. Hirotani S, Zhai P, Tomita H, Galeotti J, Marquez JP, et al. (2007) Inhibition of glycogen synthase kinase 3beta during heart failure is protective. Circ Res 101: 1164–1174.S. HirotaniP. ZhaiH. TomitaJ. GaleottiJP Marquez2007Inhibition of glycogen synthase kinase 3beta during heart failure is protective.Circ Res10111641174
- 72. Lee JY, Yeh I, Park JY, Tian B (2007) PolyA_DB 2: mRNA polyadenylation sites in vertebrate genes. Nucleic Acids Res 35: D165–168.JY LeeI. YehJY ParkB. Tian2007PolyA_DB 2: mRNA polyadenylation sites in vertebrate genes.Nucleic Acids Res35D165168
- 73. Lewis BP, Burge CB, Bartel DP (2005) Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120: 15–20.BP LewisCB BurgeDP Bartel2005Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets.Cell1201520
- 74. Zhang C, Zhang Z, Castle J, Sun S, Johnson J, et al. (2008) Defining the regulatory network of the tissue-specific splicing factors Fox-1 and Fox-2. Genes Dev 22: 2550–2563.C. ZhangZ. ZhangJ. CastleS. SunJ. Johnson2008Defining the regulatory network of the tissue-specific splicing factors Fox-1 and Fox-2.Genes Dev2225502563