Actin is abundant in the nucleus and has been implicated in transcription; however, the nature of this involvement has not been established. Here we demonstrate that beta-actin is critically involved in transcription because antibodies directed against beta-actin, but not muscle actin, inhibited transcription in vivo and in vitro. Chromatin immunoprecipitation assays demonstrated the recruitment of actin to the promoter region of the interferon-gamma-inducible MHC2TA gene as well as the interferon-alpha-inducible G1P3 gene. Further investigation revealed that actin and RNA polymerase II co-localize in vivo and also co-purify. We employed an in vitro system with purified nuclear components to demonstrate that antibodies to beta-actin block the initiation of transcription. This assay also demonstrates that beta-actin stimulates transcription by RNA polymerase II. Finally, DNA-binding experiments established the presence of beta-actin in pre-initiation complexes and also showed that the depletion of actin prevented the formation of pre-initiation complexes. Together, these data suggest a fundamental role for actin in the initiation of transcription by RNA polymerase II.

No abstract

The muscle actins in higher vertebrates display highly conserved amino acid sequences, yet they show distinct expression patterns. Thus, cardiac ␣-actin, skeletal ␣-actin, vascular smooth muscle ␣-actin, and enteric smooth muscle ␥-actin comprise the major actins in their respective tissues. To assess the functional and developmental significance of cardiac ␣-actin, the murine (129͞SvJ) cardiac ␣-actin gene was disrupted by homologous recombination. The majority (Ϸ56%) of the mice lacking cardiac ␣-actin do not survive to term, and the remainder generally die within 2 weeks of birth. Increased expression of vascular smooth muscle and skeletal ␣-actins is observed in the hearts of newborn homozygous mutants and also heterozygotes but apparently is insufficient to maintain myofibrillar integrity in the homozygous mutants. Mice lacking cardiac ␣-actin can be rescued to adulthood by the ectopic expression of enteric smooth muscle ␥-actin using the cardiac ␣-myosin heavy chain promoter. However, the hearts of such rescued cardiac ␣-actin-deficient mice are extremely hypodynamic, considerably enlarged, and hypertrophied. Furthermore, the transgenically expressed enteric smooth muscle ␥-actin reduces cardiac contractility in wild-type and heterozygous mice. These results demonstrate that alterations in actin composition in the fetal and adult heart are associated with severe structural and functional perturbations.

Two IgG1, kappa monoclonal antibodies (Mab) against actin have been obtained from a fusion in which chicken gizzard actin was used as the immunogen. One Mab, designated B4, shows a preferential reactivity toward enteric smooth muscle actin but also cross-reacts with skeletal, cardiac, and aorta actins on the basis of immunoblots, ELISA assays, and indirect immunofluorescence. However, this antibody does not react with either cytoplasmic actin in any of these assay systems. A second Mab, designated C4, reacts with all six known vertebrate isoactins as well as Dictyostelium discoideum and Physarum polycephalum actins. Thus B4 Mab appears to react with an epitope that is at least partially shared among the muscle actins but not found in cytoplasmic actins, while C4 Mab binds to an antigenic determinant that has been highly conserved among the actins. The binding sites of both Mabs on skeletal actin overlap that of pancreatic DNase I. Both antibodies bind a SV8 proteolytic product comprising the amino-terminal two-thirds of the actin molecule, and their epitopes appear to overlap since C4 can compete for the binding of B4 to skeletal actin. Neither antibody is able to prevent actin polymerization.