Actin's role in the nucleus comes packed in its very own black box. The cell nucleus is a very small and complex place but given how much is known about actin in the cytoplasm, it can be startling to hear how little agreement there is about actin in the nucleus. Indeed, only in recent years has actin's presence in the nucleoplasm been widely conceded along with some—still to be delineated—role as an RNA processing factor. Now comes a new study in Molecular Biology of the Cell (MBoC) that sheds direct light on what actin is up to in the nucleus of somatic cells. But the work also underscores the difficulty of shedding light in very small black boxes.
One of the problem facing researchers has been crafting the proper reporter probes for tracking nuclear actin, says Brittany Belin, the first author on the paper with colleagues Beth Cimini, Elizabeth Blackburn, and R. Dyche Mullins, all of the University of California, San Francisco (USCF). Actin-binding domains (ABDs) have to be selected to differentiate between monomeric and filamentous actin. Nor can the ABD be too powerful because of another quirk of nuclear actin—it is found at very high levels in the oocyte nucleus but at very low levels in the somatic nucleus. Working in human somatic osteosarcoma cells, the design of a monomeric actin-binding probe or G (for globular) ABD was fairly straightforward, Belin explains. Designing a filamentous actin-binding domain probe or FABD was not. At such low concentrations, an overly powerful FABD might create its own artifacts by pulling filamentous actin into artificial clumps.
To pass the "three bears porridge test" for a nuclear actin FABD—neither too strong nor too weak—Belin screened mutants of the human UTRN gene that produces a filamentous actin-only binding protein called utrophin, to come up with truncated utrophin molecules. Belin found the sweet spot in Utr230, a molecule with 230 amino acids instead of 261.
Inside the nucleus, the GABD probe localized to nuclear speckles, nuclear domains known to be rich in pre-mRNA splicing factors, thus cementing monomeric actin's connection to mRNA transcription. The Utr230-loaded FABD probes grabbed onto small, distributed actin "puncta," pointy filament shards, which didn't seem to associate directly with any of the nuclear landmarks including speckles, lamins, chromatin, or telomeres.
These filamentous actin puncta exhibited two different movement patterns, says Belin, either relatively quick "diffusive" motion or much slower "subdiffusive" motion. Both fast and slow particles, however, moved backward and forward as if bouncing off weak springs. "Our best hypothesis is that actin is interacting with or is part of a viscoelastic polymer mesh," says Belin. "We think it's not DNA because it's not in the chromatin rich areas of the nucleus. Possibly it's some sort of scaffold for locally organized events (in the nucleoplasm) but we're not sure."
Even if the exact purpose of this viscoelastic polymer mesh is still to be determined, Dyche Mullins says the results are exciting because the data start to answer some very basic questions. "What's it like to be in the nucleus? What's the physical geography of the nucleus for one thing? What does it feel like? What's the texture?" Based on Belin's work, it's both a viscous and an elastic medium, says Mullins. He explains that the collaborators on this paper chose the term "viscoelastic mesh" very carefully to avoid being drawn into earlier battles about a nuclear matrix. Whatever you call it, says Mullins, "It's clear that there's some kind of meshwork that this actin is either forming part of or interacting with. For us, it's like opening the first window into what's it's like to be in the nucleus."
More light is needed, says Mullins. "The nucleus is such a black box and for most people who work on the cytoskeleton, the nuclear pore might as well have 'Abandon hope, all ye who enter here' written over it. Brittany was very brave to take on this project."
CellTweet #9. Citation: Belin BJ, Cimini BA, Blackburn EH, Mullins RD. (2013). Visualization of actin filaments and monomers in somatic cell nuclei. Mol Biol Cell 24(7), 982-994