Nature Cell Biology
6, 1013 - 1014 (2004)
doi:10.1038/ncb1104-1013
Actin's many actions start at the genesWerner W. FrankeWerner W. Franke is in the Division of Cell Biology, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. w.franke@dkfz.de Over more than three decades, sporadic and circumstantial evidence has accumulated to suggest that nuclear actin has crucial functions in RNA polymerase II-based transcription. Now, using a biochemical approach,  -actin has been identified as a highly specific, constitutive component of the active transcriptional complex required for formation of the pre-initiation transcription complex.For more than three decades it was considered cell-biological heresy to suggest that actin might exist and function in the nucleus. Actin was understood to mean filaments, cables, contraction, motility and so on — all cytoplasmic functions for actin; the nucleus was off limits; after all, in immunolocalization experiments actin was demonstrably absent from the nucleus, and in fact, it was often used as a negative control. So, not surprisingly, authors who dared to report results indicating that this famous structural protein also exists in the nucleus, mostly in a soluble form, and — even more heretical — suggest that it might be involved in classic nuclear functions such as transcription, processing of transcripts and ribonucleoprotein transport to the cytoplasm, had a difficult time. Nuclear actin was also a notorious grant proposal poison. Nevertheless, at intervals there were isolated reports on 'the nuclear actin problem', as it was often called, and one prominent developmental biologist classified it as a typical "axolotl": strange and interesting, but never metamorphosing into a mature entity.
The idea that actin may be in the nucleus and may be involved in transcription emerged in 1969 as a possible explanation for the observation1 that inhibition of transcription in amphibian oocytes by actinomycin D resulted in the formation of intranuclear bundles of filaments resembling cytoplasmic microfilaments. This observation was then followed by reports showing not only the presence of actin in isolated nuclei but also the association of actin with RNA polymerase II during purification2,
3, which was supported and extended by the in vivo and in vitro experiments of others4,
5. Only slowly did the presence of sizeable amounts of actin, actin-related-, actin-binding- and actin-regulatory proteins in the nucleus become accepted facts6,
7. More recently, however, cell biologists have become more open-minded — considering the possibility that actin is also involved in diverse nuclear functions, including transcription, ribonucleoprotein packaging and transport, chromatin remodelling and the formation of karyoskeletal elements8,
9,
10,
11 (Fig. 1).
 | | |
Now finally, Hofmann et al. report on page 1094 of this issue12 the results of a comprehensive study of a consortium of a dozen scientists from five institutions, led by Primal de Lanerolle, which leaves little doubt that at least one member of the actin family of proteins is indeed needed as a constitutive and functionally crucial component of the RNA polymerase II-based transcriptional machinery. These authors conclude that it is exclusively  -actin, the non-muscle isoform that is involved in this process. However, their finding that all other isoforms, including  -actin, another non-muscle actin, and the four muscle-specific actins, are nearly inactive in their transcription assay, is remarkable in view of the few amino-acid exchanges between these diverse actins. After all, -actin differs only in four amino acids clustered at the amino terminus, and all four exchanges are conservative in nature.
The central aspect of this report is a series of biochemical experiments, using HeLa cells to demonstrate that it is the -actin isoform that selectively associates with RNA polymerase II in vitro and in vivo, and that only the -actin-specific antibodies inhibit transcriptional activity in assays using purified RNA polymerase II and pure transcription factors. The authors report that the addition of recombinant -actin can markedly stimulate transcription. The actin-containing transcription complex associates with the gene promoter in a manner that is strictly dependent on the induction of gene expression. The assembly and activity of this complex does not depend on chromatin remodelling, an activity that itself depends on actin. They also demonstrate that the formation of the initiation complex and the continuation of transcription is absolutely dependent on -actin. In particular, Hofmann et al. emphasize a requirement for actin in the formation of the pre-initiation and the TATA-box complexes. Co-immunoprecipitation experiments identified a specific complex containing -actin, RNA polymerase II and the TATA-box-binding protein.
The report by Hofmann et al.12 is highly provocative in several ways. Certainly, the specificity claim — that a single isoform of non-muscle -actin is involved — will be challenged by assays using alternative isoforms and mutational analyses13. Moreover, this claim also raises the more general question of whether and how the four muscle-type actins are excluded from the transcriptional process or possibly from the nucleus in the diverse cells of myogenic lineage. After all, nuclear actin that is recruited to the transcriptional complex represents only a minute fraction of nuclear actin14, and the mechanism by which transcriptionally involved actin molecules could be sorted from the bulk of nuclear actin is at present an enigma.
The present paper12 has to be seen in the wider perspective, of recent reports that actin, accompanied by myosin, is an essential constituent of the RNA polymerase I complex responsible for the nucleolar transcription of rDNA15. Here too, actin is needed throughout the entire process, from initiation to the final release of the large RNA polymerase−pre-rRNP complex (I. Grummt, personal communication). And to complete the trio, a role for -actin has concurrently been found in RNA polymerase III transcription16. Certainly, examination of possible requirements of other cytoskeletal proteins in the regulation of RNA polymerases is now generally mandatory, given that transient nucleoplasmic storage particles of RNA polymerase III contain plakophilin-2, a 'hard-core' cytoskeletal protein originally identified in desmosomal plaques17.
Finally, much remains to be learnt about how actin complexes with the RNA polymerse and how this complex is stored and disassembled. In this regard, the Cajal body — the main nuclear substructure known as the assembly site and transitory stockpile of all three types of RNA polymerase complexes — is obviously of special interest14. Remarkably, these prominent nuclear bodies also harbour certain actin-binding proteins, such as profilin18 — a protein involved in the regulation of actin dynamics. Most of what we have learnt in recent years about the functions of actin in the cytoplasm has come from the study of the functions of its partners, and it does not need a prophecy to predict that this approach will also be successful in the elucidation of the nuclear functions of actin.
REFERENCES
-
Lane, N. J. J. Cell Biol. 40, 286−291 (1969). | Article | PubMed | ISI | ChemPort |
-
Jockusch, B. M., Becker, M., Hindennach, I. & Jockusch, H. Exp. Cell Res. 89, 241−246 (1974). | PubMed | ISI | ChemPort |
-
Smith, S. S., Kelly, K. H. & Jockusch, B. M. Biochem. Biophys. Res. Commun. 86, 161−166 (1979). | PubMed | ISI | ChemPort |
-
Egly, J. M., Miyamoto, N. G., Moncollin, V. & Chambon, P. EMBO J. 3, 2363−2371 (1984). | PubMed | ISI | ChemPort |
-
Scheer, U., Hinssen, H., Franke, W. W. & Jockusch, B. M. Cell 39, 111−122 (1984). | Article | PubMed | ISI | ChemPort |
-
Ankenbauer, T., Kleinschmidt, J. A., Walsh, M. J., Weiner, O. H. & Franke, W. W. Nature 342, 822−825 (1989). | Article | PubMed | ISI | ChemPort |
-
Rimm, D. L. & Pollard, T. D. J. Cell Biol. 109, 585−591 (1989). | Article | PubMed | ISI | ChemPort |
-
Olave, I. A., Reck-Peterson, S. L. & Crabtree, G. R. Annu. Rev. Biochem. 71, 755−781 (2002). | Article | PubMed | ISI | ChemPort |
-
Pederson, T. & Aebi, U. J. Struct. Biol. 140, 3−9 (2003). | Article | ISI |
-
Bettinger, B. T., Gilbert, D. M. & Amberg, D. C. Nature Rev. Mol. Cell Biol. 5, 410−415 (2004). | Article | PubMed | ISI | ChemPort |
-
Kiseleva, E. et al. J. Cell Sci. 117, 2481−2490 (2004). | Article | PubMed | ISI | ChemPort |
-
Hofmann, W. et al. Nature Cell Biol. 6, 1094−1101 (2004). | Article |
-
Posern, G., Sotiropoulos, A. & Treisman, R. Mol. Biol. Cell 13, 4167−4178 (2002). | Article | PubMed | ISI | ChemPort |
-
Doyle, O., Corden, J. L., Murphy, C. & Gall, J. G. J. Struct. Biol. 140, 154−166 (2002). | Article | PubMed | ISI | ChemPort |
-
Fomproix, N. & Percipalle, P. Exp. Cell Res. 294, 140−148 (2004). | Article | PubMed | ISI | ChemPort |
-
Hu, P., Wu, S. & Hernandez, N. Genes Dev. (in the press).
-
Mertens, C. et al. Proc. Natl Acad. Sci. USA 98, 7795−7800 (2001). | Article | PubMed | ChemPort |
-
Skare, P., Kreivi, J.-P., Bergström, Å. & Karlsson, R. Exp. Cell Res. 286, 12−21 (2003). | Article | PubMed | ISI | ChemPort |
|
