In our previous study, we have reported that cnn mutant GSCs show high frequency of misoriented spindles, leading to symmetric GSC divisions and thus an increase in GSC number [4].
Our results provide evidence that GSCs express high levels of integrin alpha6, which can serve not only as an enrichment marker but also as a promising antiglioblastoma therapy.
As has been reported previously, our study confirms that GSCs express very low levels of the gap junction protein Cx43.
According to the well-known hypothesis that GSCs are highly invasive and are localised surrounding vessels, the so-called tumour niche (Liu et al, 2006; Calabrese et al, 2007), these data suggest that GSCs express integrin α3 in vivo.
While scoring centrosome orientation in GSCs expressing E-caddCR4h, we noticed that these GSCs had a high frequency of misoriented spindles during mitosis (∼39%, Figure 4A, C), while GSCs expressing E-cadDEFL did not (0%, Figure 4B, C).
Nevertheless, GSCs expressing E-caddCR4h remattachedtoched thethubhub cells, presumably because hub-GSC interactions were supported by endogenous E-cadherin (Figure 1A).
This is a stark contrast to GSCs expressing E-caddCR4h, which have comparable spindle and centrosome misorientation frequencies (Figure 4C).
Furthermore, we observed that overexpression of dCR4h often formed ectopic Baz patches away from the hub-GSC interface, and such a Baz patch was associated with the centrosomes (approximately 10% of GSCs expressing dCR4h, Figure 4C).
Indeed, 25.8% of myt1 1 mosaic mutant germaria contained no myt1 1 mutant GSCs, but did carry myt1 1 mutant germ cell cysts (recognized by the absence of β-gal signals) at 4 days ACI (Fig. 5D); therefore, at least one myt1 1 mutant GSC had been lost from these germaria.
These results show that ball mutant GSCs had lost their capacity to self-renew and left the niche.
