However, phosphorylation was less intense in the GDC-0879 pellet fraction of menadione-treated R120G cells. In menadione-treated WT cells, but not in R120G cells, a preferential phosphorylation of .gif)
HspB5 in the pellet fraction was observed. Moreover, we noticed that mutant HspB5 in the pellet fraction of untreated R120G cells was less phosphorylated than its soluble counterpart. Further analysis of the phosphorylation of the different oligomeric ALK5 Inhibitor II structures of HspB1 and HspB5 confirmed the complex nature of these modifications. As already described, in HeLa as well as in control neo cells HspB1 is characterized by three subpopulations based on their serine sites specific phosphorylation and native size. However, in WT cells this structural organization was no more observed since most HspB1 interacted with HspB5 resulting in the formation of a large oligomeric complex which surprisingly contained only one HspB1 phosphorylated isoform. Hence, in spite of the fact that the N-terminal domain of HspB1 consisting of amino acids 1�C124, which bears the phosphoserine sites, has been reported to not interact with HspB5, our observations suggest that its recognition by MAPKAPK2,3 kinase is potentially impaired. Contrasting with these observations, the partial dissociation of HspB1-HspB5 complex in response to menadione was associated with the presence of the three HspB1 phosphoisoforms in the remaining complex, hence suggesting a profound reorganization of this chimeric complex. In both normal and oxidative conditions, HspB5 phosphoisoforms were mainly localized in the HspB1HspB5 complex. This suggests that HspB1 does not interfere with the kinase accessibility of HspB5 N-terminal domain. Of interest, in untreated WT cells, the small fraction of HspB1 that was not interacting with HspB5 was recovered in small oligomers that differed in their phosphorylation from those observed in control Neo cells devoid of HspB5 expression. Consequently, the formation of HspB1-HspB5 complex indirectly generated the formation of a new-type of highly phosphorylated small HspB1 homo-oligomers that could play a role in the enhanced oxidoresistance of WT cells through their ability to interact with G6PDH. The R120G mutant altered HspB1-HspB5 structural organization in such a way that most of the cellular content of HspB1 phosphoserine 15 was now recovered in the complex while this modification was not present in the complex formed with wild type HspB5. In contrast, no major changes in HspB1 phosphoserines 78 and 82 distribution were noticed. In spite of the fact that HspB1-mutant HspB5 complex was almost completely disrupted by oxidative stress, it is interesting to note that, as in WT cells, the remaining complex contained the three phosphoisoforms of HspB1. This suggests that in oxidative conditions, the N-terminal part of HspB1 can be freely recognized by the corresponding kinase. Moreover, the partial disruption of the mutant complex in cells transiently expressing HspB1 serine 15 non-phosphorylatable mutant suggests that serine 15 phosphorylation could stabilize HspB1 interaction with mutant HspB5. It is not known whether the absence of serine 15 phosphorylation in the wild type complex is due to the masking of this serine site or reflects its uselessness nature for stabilization. Within the limitation that this study has been made in a single set of HeLa-derived cell lines, the observations reported here illustrate the complex nature of the interaction between HspB1 and HspB5. It is also not known if similar interacting behavior occurs in other types of cells. Moreover, we cannot exclude that the low level of HspB6 expression detected in HeLa cells may modulate, at least to a certain extent, the structural organization of the complexes formed by HspB1 and HspB5.