Asymmetric Mbc, active Rac1 and F-actin foci in the fusion-competent myoblasts during myoblast fusion in Drosophila

Abstract

There was an error published in Development138, 1551-1562.We have been unable to reproduce the data in Figure 1H, reporting that the myoblast fusion defect in mbc mutant embryos is rescued by expression of constitutively activated Rac1. Although we are unable to confirm the source of the error, all stocks have been re-confirmed and all other results in the original publication independently confirmed by two of the authors.Revised data pertaining to Figure 1 of the original paper are provided in Fig. 1 below. All are lateral views of stage 16 embryos stained as in the original publication.Results shown in Fig. 1A,B,E are as originally published (original Figure 1 panels A, B and I, respectively). Of note, the rescued embryo in Fig. 1B exhibits dorsal closure defects characteristic of mbc mutants, confirming that rescue is only occurring in the musculature as expected. It is apparent in Fig. 1C,D, however, that constitutively active Rac1 does not rescue fusion in an mbc mutant embryo (as originally shown in Figure 1H of the original publication). We do note that FCMs expressing activated Rac1 exhibit an increased tendency to cluster around founder cells, suggesting that migration is increased in these FCMs. These cells often also adopt a more elongated appearance with more pronounced filopodia (Fig. 1D). Additionally, the data in Fig. 1F confirm that pan-mesodermal expression of activated Rac does not rescue the myoblast fusion defect and, moreover, causes severe perturbations owing to higher and earlier expression of activated Rac1.To confirm that the bipartite Mbc-Elmo guanine nucleotide exchange factor does function in the Rac1 pathway, we examined the consequences of their overexpression in the musculature (Fig. 2, below).As anticipated, this overexpression caused defects in myoblast fusion reminiscent of those described for overexpression of activated Rac1 itself (Luo et al., 1994). Notably, this myoblast fusion defect is suppressed by loss of one copy of Rac1 and Rac2 (Fig. 2B), consistent with their activation by Mbc and Elmo.In summary, the primary conclusions of the original publication remain intact. However, our current results do not support the additional conclusion that the only role of mbc is to activate Rac1. Although Mbc and Elmo appear to be capable of activating this small GTPase, the inability of constitutively activated Rac1 to rescue fusion in mbc mutant embryos mutants suggest that Mbc is required for more than the simple activation of Rac1, perhaps controlling its localization or activating another small GTPase.The authors apologize to readers and anyone who experimentally followed up on these results for this mistake.