The image on the cover of the February 4 issue of Biophysical Journal shows a field of cells that have been subjected to centrifugal forces, which in normal cells would remove nearly all the nuclei, here stained blue, leaving behind cytoplasts—remnants filled with the cytoskeleton, here stained green. The difference in these cells is that they have been treated with an inhibitor of a motor protein in the nucleus, the Brahma-related gene (BRG1) ATPase of the Brg/Brm-associated factor (BAF) or SWItch/Sucrose Non-Fermentable (SWI/SNF) complex. Based on the results of this report, the reason these cells do not lose their nuclei as efficiently is that without the BRG1 motor, which binds DNA in nucleosomes and pulls it away from histones to reveal transcription sites, the nucleus cannot transform from its flattened shape in the spread cell to the elongated or spherical nucleus that emerges from the cell. This transformation enables the nucleus to leave the cell with a thin layer of cytosol and a plasma membrane, but no organelles, to form a karyoplast. This method, developed a half century ago, is usually used to prepare cytoplasts, discarding the nuclear fraction. However, we harvest the karyoplasts to perform biophysical studies, by using atomic force microscopy to measure their viscoelastic properties, which are determined by how they change shape in response to compressive forces such as those the nucleus experiences as cells squeeze through tight spaces.

Our lab generally studies the mechanical properties of cytoskeletal or extracellular matrix networks, cells, and tissues and attempts to explain them by using concepts of polymer physics. We had not expected to work on the nucleus, and our lab's previous work on DNA involved measuring the mechanical response of DNA-containing biofilms, which are much softer than the nucleus.  A few years ago, Alison Patteson was studying the differences in the way that the nucleus deforms in cells that either do or do not contain the intermediate filament vimentin, which forms a cage around the nucleus. When cells are made to crawl through tight spaces, such as within a collagen gel or a microfluid chamber, the nucleus deforms significantly, and if there is not a vimentin cage around it, it can break, leading to either cell death, or as others have shown, the accumulation of mutations because of inability to repair locally damaged DNA. We wondered how the nucleus could deform so much, because it is often referred to as the stiffest organelle of the cell. Jeff Byfield undertook to make karyoplasts and measured their deformation under compressive stresses by using the same methods he previously applied to study intact cells and tissues. The way in which the nucleus deformed in response to different levels of stress, applied at different rates, could not be explained by standard viscoelastic models. Although stiff to sudden small perturbations, most of the work of compressing the nucleus from a sphere to a flattened disk was dissipated regardless of compression speed, even though when the stress was removed, the karyoplast returned to its spherical shape, characteristic of an elastic material. These rheological responses resembled those of an active material, a dense viscoelastic object in which motions appear random but are not simply thermally driven, but also driven by much stronger kicks resulting from conformation changes often associated with ATP hydrolysis, characteristic of motor proteins. We therefore inhibited ATP production by the glycolytic enzymes in the karyoplast. The nucleus suddenly became much stiffer and almost completely elastic. The dissipative element of the normal karyoplast was lost. We then looked for the molecular motor responsible for the dissipative motions.

Our first attempts to inhibit such motors as RNA polymerase or DNA polymerase did not affect the response. However, a collaboration 25 years ago with Oliver Rando, then a fellow at Harvard Medical School, pointed the way to explain our result. Rando and his colleagues had isolated all of the subunits identified by genetics as essential elements in a multi-component structure called the “SWI/SNF complex” in yeast and “BAF” in mammalian cells and found, probably to their surprise, that one of the subunits of this authentically nuclear complex was b-actin, and another was an actin-related protein. A third component of this complex (BRG1) contained an amino acid sequence similar to the phosphatidylinositol 4,5-bisphosphate (PIP2) regulatory site of the actin binding protein gelsolin, and Rando wondered if the reconstituted complex could nucleate actin assembly in vitro in a PIP2-dependent manner. Our first attempt to nucleate actin in standard actin assembly buffers failed, as we hoped it would, but then we added the phospholipid PIP2, which activates the protein complex to allow for b-actin, the actin-related protein, and possibly other components to create the template necessary for actin nucleation. Now, actin filaments grew from the reconstituted BAF complex. Later studies have shown that BRG1 is one of the strongest and fastest motors in the animal cell. Luckily, there is a small molecule inhibitor of BRG1, and when we added it to the karyoplast, it stiffened and became almost completely elastic, just as if we had eliminated ATP. Behnaz Eftekhari then noticed that if we added the BRG1 inhibitor first to the intact cell, the nucleus could no longer be so easily removed, suggesting that BRG1 plays a critical role in the deformation required for nucleus expulsion.

These findings suggest that the nucleus might be stiff when removed from the cell and its cytoplasm, or under rapid or localized deformation, but if small persistent forces are imposed on the nucleus, for example by lipid droplets under intracellular stresses as studied by our colleagues Rebecca Wells and Dennis Discher, the nucleus becomes deformed although the lipid droplets stay spherical. The similarity between the nucleus and a droplet of active viscoelastic matter might be a useful way to think about nuclear mechanics, in addition to the more common model of it as a composite of internal chromatin surrounded by a nuclear lamina. Further studies will show how the multiple structural elements in the nucleus conspire to produce the unique mechanical properties that allow it to be, on the one hand, highly compliant, but on the other hand, protective of its chromatin and able to return to its resting state when mechanical stresses are relieved.

— Fitzroy J. Byfield, Behnaz Eftekhari, Kaeli Kaymak-Loveless, Kalpana Mandal, David Li, Rebecca G. Wells, Wenjun Chen, Jasna Brujic, Giulia Bergamaschi, Gijs J.L. Wuite, Alison E. Patteson, and Paul A. Janmey