The Multiple Ways Nuclei Scale on a Multinucleated Muscle Cell Scale

Abstract

In mononucleated cells, nuclear size scales with cell size. Whether this relationship extends to multinucleated cells was unknown. In this issue of Developmental Cell, Windner et al. (2019) examine scaling of nuclei in multinucleated Drosophila muscle fibers. They find that nuclear size is affected by global, regional and local factors.

Nuclear size scaling is a phenomenon conserved from single-celled to multicellular eukaryotic organisms: within a given cell type, there is a relatively constant nucleus:cell volume ratio, and this ratio is disrupted in many cancers (Jevtić and Levy, 2014). The significance of this ratio and its underlying mechanism are poorly understood. In this issue of Developmental Cell, Windner et al. explored nuclear scaling in multinucleated Drosophila muscle fiber cells (Windner et al., 2019). Given that multinuclear muscle fibers have distinct functional domains, this experimental approach could be used to explore the relationship between nuclear size and nuclear function. Such a link might also explain the distinct nuclear:cell volume ratios observed in different cell types. Moreover, the study of multinucleated cells allows for the assessment of global (i.e. cell-wide), regional (at a functional cellular domain), and local (nuclear surroundings) factors that might affect nuclear scaling. Indeed, the findings of this study show that all three contribute to nuclear size.

As in humans, Drosophila muscle fibers form by the fusion of diploid myoblasts, creating one cell with multiple nuclei- the myotube or muscle fiber (Figure 1A). The number of fusing myoblasts determines the number of nuclei per cell. Those nuclei then increase in ploidy via endoreplication, leading to an increase in gene expression and resulting in an increase in cell size (Figure 1B). Windner et al. focused on the Drosophila VL3 and VL4 larval muscle cells. Due to their flat configuration, 2D measurements accurately reflect 3D values, facilitating the quantification of different parameters that may contribute to nuclear scaling (e.g. cell size, nuclear size, internuclear distances, etc). Moreover, being in Drosophila, this system has the advantage that it is genetically manipulatable. VL3 muscle fibers are larger than VL4 cells and contain more nuclei. In both cells, nuclei are evenly spaced, with predominantly one row of nuclei in VL4 cells and two rows in VL3.

FIGURE 1:

(A) The formation of muscle fibers (myofibers). Red: cytoplasm, blue: nuclei. Note the concurrent scaling of nuclear and cell sizes.

(B) The relationship between DNA content, gene expression, cell size and nuclear size in global nuclear size scaling. see text for more details. Inset:

(C) Regional and local factors that affect nuclear size according to their position (top) or abundance (bottom) along the anterior-posterior axis.

In an elegant set of experiments, Windner et al. observed a range of nuclear sizes within a given cell. Nonetheless, as in mononucleated cells, total nuclear size correlated strongly with cell size in both VL3 and VL4 cells. This indicated that there are one or more global factors dispersed throughout the cell that determine the collective volume of nuclei. Interestingly, however, larger nuclei tended to be towards the center of the cell (Figure 1C). Thus, while there was a strong global influence on overall nuclear size, the asymmetric size distribution of individual nuclei along the muscle fiber indicated that regional and/or local factors could also be involved.

What local factor(s) might affect nuclear size? It was recently shown that the nucleus:cell volume ratio is not determined by the availability of nuclear envelope components, supporting the idea that nuclear size is determined by non-membranous nuclear and/or cytoplasmic factors (Walters et al., 2019). The size of nuclei in multinucleated fission yeast cells is proportional to the volume of the surrounding cytoplasm (Neumann and Nurse, 2007), and Xenopus experiments demonstrating that specific cytoplasmic components may influence nuclear size (Levy and Heald, 2010). It was already known that the cytoplasm of muscle fiber cells is not uniform: the myonuclear domain hypothesis (reviewed in Hall and Ralston, 1989) posits that each nucleus has a distinct region with which it interacts. This idea is supported by the observations that in muscle fibers, mRNA and certain proteins made thereof remain associated with their nucleus of origin (Ralston and Hall, 1992 and references within). Based on this, Windner et al. assigned each nucleus a Voronoi domain—a geometric area of cytoplasm surrounding each nucleus, in which each point in this domain is closer to that nucleus than to any other nucleus (Du et al., 2010). Individual nuclear size scaled with their Voronoi domain, albeit much weaker than the observed global scaling. Thus, in the muscle fiber, something in addition to the immediate cytoplasm could affect nuclear size.

In muscle fiber cells, distinct innervating neuromuscular junctions (NMJs) are found in a narrow region towards the anterior half of both VL3 and VL4 cells (Figure 1C). This positioning correlated with the presence of larger nuclei towards the center of the cells. Moreover, Windner et al. found that nucleolar size and the K9-acetylation of histone H3, both a proxy for synthetic activity, peaked towards the anterior half of cells (Figure 1C). At this point, the association between NMJs, larger nuclei and higher synthetic activity remains a correlation, as the authors did not examine the effect of changing nuclear positioning relative to the NMJs on nuclear size. Nonetheless, this raises the possibility for a regional effect on nuclear size mediated by NMJs, perhaps to increase the production of proteins needed for NMJ-related activities.

As noted above, part of the increase in muscle fiber size is attributed to endoreplication, a process in which DNA replication is not followed by nuclear division, resulting in increased ploidy. In VL3 and VL4 muscle cells, the ploidy of nuclei ranges from 16C to 64C. Windner et al. observed that 64C nuclei were more frequently found towards the center of these cells (Figure 1C) and 16C nuclei preferentially at the cells’ ends, correlating with nuclear size. This raises an interesting question: in a common cytoplasm, how do some nuclei replicate more than others, and why do nuclei with different ploidies preferentially localize to different domains? Does the position of the NMJs drive this asymmetry, and do the myonuclear domains provide an environment for a positive feedback loop for endoreplication potential?

To begin to address the mechanism linking ploidy with nuclear size, Windner et al manipulated the ploidy in VL3 and VL4 cells by Myc over-expression (OE) (producing higher ploidy than control), or Cdt1 knockdown (KD) (lower ploidy). Despite dramatic changes in ploidy, cell size stayed relatively similar to control. In the Cdt1 KD mutants, nuclear size decreased relative to control, and the Myc OE mutants exhibit a dramatic increase in nuclear size, with an even more dramatic increase in nucleolar size. Because the amount of DNA per se does not drive nuclear scaling (for example, see Neumann and Nurse, 2007), it is likely that nuclear size changed because of reduced (Cdt1 KD) or increased (Myc OE) nuclear activity, leading to a respective change in the availability of factors that determines nuclear size. It would be interesting to examine whether the gradient in ploidy across the cell length (Figure 1C) is important for muscle fiber function. Moreover, because the increase in ploidy was not accompanied by an increase in cell size, this suggests that nuclear size can be directly affected by gene expression (Figure 1B, blue arrow), rather than through a mechanism that gauges cell size and adapts nuclear size accordingly. If this is the case, then nuclear size scaling could be a passive process, in which nuclear size and cell size scale independently of each other but proportionately with nuclear activity. It also remains to be determined whether the size of the nucleus affects gene expression (Figure 1B, green arrow). The identification of local, regional and global factors that determine nuclear size will help establish the precise mechanism the determines the nuclear:cell volume ratio and is relationship to nuclear function.

Previous studies in Drosophila have linked muscle dysfunction with defects in nuclear positioning, a phenomenon also frequently seen in human muscular disorders such as muscular dystrophies (Metzger et al., 2012 and Folker and Baylies, 2013). Other types of multinucleated cells exist (e.g. trophoblasts), and multinucleated model systems recapitulating human musculature, such as the one employed here, provide an opportunity to study elements that link nuclear size with nuclear function and their contribution to disease.