The mitotic phenotype of clathrin removal in mammalian cells is a delay in progression through mitosis that is apparently the result of defects in congression during prometaphase. The cells make it through mitosis, but do so in a delayed manner due to prolonged signalling of the spindle checkpoint. It has never been suggested that clathrin is essential for mitosis, just that it is necessary for normal mitosis.
The second problem with this paper is the inference that clathrin localisation at the mitotic spindle is not linked to a mitotic function throughout evolution on the basis of the experiments reported here. There are several reasons why Borlido et al. found no mitotic phenotype upon clathrin removal.
1. The DKO-R cell line is not representative of chicken cells. When clathrin is removed from DT40 cells this results in cell death by apoptosis (Wettey et al., 2002). Wettey et al. isolated a variant cell line (DKO-R) that had adapted to the loss of clathrin and was resistant to cell death, allowing the study of clathrin function in membrane trafficking. The DKO-R cell line was also used in the present study, leading to the possibility that the adaptation that this cell line has undergone prevents the mitotic phenotype upon clathrin removal to be evident.
2. Functional substitution by CHC22. DKO-R cells have significant CHC22 expression (Stoddart et al., Mol Biol Cell, 2005) and in the absence of good evidence that CHC22 cannot functionally substitute for CHC17 repression, it remains possible that the present results could be explained by functional substitution of CHC17 removal by CHC22.
3. Is there actually a mitotic phenotype in the DKO-R cells? The experiments showing the proportion of cells in G2/M following release from nocodazole suggest that there is a small population of cells delayed in G2/M (Fig 1D). It is difficult to ascertain from what is reported whether this population difference is statistically different. The proportions were analysed by ANOVA, an analysis of variance statistical test that is suited to normally distributed data. Wouldn’t a non-parametric test such as Kruskal-Wallis would have been more appropriate?
Borlido et al. use HEK293 cells to confirm our previous observations of mitotic defects following clathrin depletion in HEK293 cells and NRK cells. Similar defects have been reported in murine cells with reduced clathrin at the mitotic spindle (Yamauchi et al. EMBO J., 2008). So is it possible that the location but not the function of clathrin at the mitotic spindle is evolutionarily conserved? The phenotype of clathrin (T20G5.1) removal in C. elegans oocytes is a ‘jumpy’ phenotype (Gönczy et al., Nature, 2000), suggesting that in worms, clathrin has a role in cell division, whether this is related to its localisation on the mitotic spindle remains to be determined.
This paper tells us very little about the role of clathrin during mitosis beyond the fact that the DKO-R variant cell line is not suitable for studying this function of clathrin.