Pivotal role of cardiac lineage protein-1 (CLP-1) in transcriptional elongation factor P-TEFb complex formation in cardiac hypertrophy

Abstract

Objective: Our aim was to determine if the expression pattern of CLP-1 in developing heart is consistent with its role in controlling RNA transcript elongation by transcriptional elongation factor b (P-TEFb) and if the inhibitory control exerted over P-TEFb by CLP-1 is released under hypertrophic conditions.

Methods: We performed immunoblot and immunofluorescence analysis of CLP-1 and the P-TEFb components cdk9 and cyclin T in fetal mouse heart and 2 day post-natal mouse cardiomyocytes to determine if they are co-localized. We induced hypertrophy in rat cardiomyocytes either by mechanical stretch or treatment with hypertrophic agents such as endothelin-1 and phenylephrine to determine if CLP-1 is released from P-TEFb in response to hypertrophic stimuli. The involvement of the Jak/STAT signal transduction pathway in this process was studied by blocking this pathway with the Jak2 kinase inhibitor, AG490, and assessing the association of CLP-1 with P-TEFb complexes.

Results: We found that CLP-1 is expressed along with P-TEFb components in developing heart during the period in which knockout mice lacking the CLP-1 gene develop cardiac hypertrophy and die. Under conditions of hypertrophy induced by mechanical stretch or agonist treatment, CLP-1 dissociates from the P-TEFb complex, a finding consistent with the de-repression of P-TEFb kinase activity seen in hypertrophic cardiomyocytes. Blockage of Jak/STAT signaling by AG490 prevented release of CLP-1 from P-TEFb despite the ongoing presence of hypertrophic stimulation by mechanical stretch.

Conclusions: CLP-1 expression in developing heart and isolated post-natal cardiomyocytes colocalizes with P-TEFb expression and therefore has the potential to regulate RNA transcript elongation by controlling P-TEFb cdk9 kinase activity in heart. We further conclude that the dissociation of CLP-1 from P-TEFb is responsive to hypertrophic stimuli transduced by cellular signal transduction pathways. This process may be part of the genomic stress response resulting in increased RNA transcript synthesis in hypertrophic cardiomyocytes.

Figure 1

Expression of CLP-1, cyclin T and cdk9 proteins in developing and post-natal heart (A), kidney (B), and liver (C). One hundred micrograms of protein lysate was loaded per lane and expression of CLP-1, cyclin T, cdk9, and GAPDH proteins detected by immunoblot analysis. GAPDH expression served as loading control. The E15 to E17 pre-natal period encompasses the time period in which CLP-1 knockout fetuses develop hypertrophy and die (24). Definitive liver and kidney tissues were not identifiable in E13 embryos and thus not harvested.

Figure 2

Expression of cdk9 and CLP-1 in E17 and E19 heart. A. Serial paraffin sections from E17 mouse heart showing co-localization of cdk9 and CLP-1 in heart cells. Boxed regions of hearts show area depicted in images (a transected vein (white area in image) was used to co-align images for comparing overlap in nuclear stainings). B. Immunoprecipitation of E19 heart lysates showing co-immunoprecipitation of CLP-1 with P-TEFb components. Input refers to the non-precipitated lysate.

Figure 3

Immunocytochemical co-localization of CLP-1, cyclin T and cdk9 in isolated 2 day post natal rat cardiomyocytes. A–C: (A) CLP-1 is in the nucleus of MF20-positive cardiomyocytes; (B) nuclear co-localization of CLP-1 and cyclin T; (C) control (no 1° antibody). (DIC, Differential Interference Contrast microscopy). D,E: Confocal microscopy of cardiomyocytes. D and E: CLP-1 and cdk9 are co-localized to the nucleus of MF20-positive cardiomyocytes. Panels i-k, magnified images of nuclei showing speckled type nuclear staining of cdk9 and CLP-1.

Figure 4

A. Stretched cardiomyocytes express markers of hypertrophy. Northern blot showing increased ANF (atrial natriuretic factor) mRNA levels in stretched cardiomyocytes. GAPDH is loading control. B. Phosphorylation of STAT 3 increases with continued stretch of cardiomyocytes (1 hr and 24 hr); overall STAT3 protein levels remain constant. C. CLP-1 is released from p-TEFb complexes in stretch-induced hypertrophic cardiomyocytes. Immunoblot of p-TEFb components in stretched and non-stretched cardiomyocytes showing markedly reduced levels of CLP-1 in P-TEFb complexes from cardiomyocytes stretched for 48 hours; cyclin T1 and cdk9 levels remain unchanged. D. Total levels of p-TEFb components (free and complexed) remain unchanged with stretch. GAPDH is loading control.

Figure 5

Inhibition of the Jak2 kinase results in decreased dissociation of CLP-1 under hypertrophic conditions. A. Mechanical stretch induces phosphorylation of STAT3; addition of AG490 prevents STAT3 phosphorylation. B. AG490 blocks the dissociation of CLP-1 from the p-TEFb complex. CLP-1 levels are reduced in cyclin T1 immunoprecipitates from stretched cardiomyocytes relative to non-stretched controls, but remain unchanged in stretched cardiomyocytes treated with AG490. Cardiomyocytes not receiving AG490 received vehicle alone. C. Total levels of P-TEFb (free and complexed) remain unchanged in 48 hour stretched and non-stretched cardiomyocytes.

Figure 6

CLP-1 dissociates from P-TEFb in cardiomyocytes treated with hypertrophic agents. A. Immunoprecipitated cyclin T1 shows less associated CLP-1 in endothelin-1-treated cardiomyocytes than untreated cells (control). B. Immunoprecipitated cyclin T1 shows less associated CLP-1 in phenylephrine-treated cardiomyocytes than in untreated cells. Input is non-immunoprecipitated cell lysates.