We therefore tested HspB1 and HspB5 level of phosphorylation in Neo, WT and R120G cells exposed or not to 60 mM menadione for 2 h, a duration that corresponded to the maximal level of phosphorylation induced by this drug. It is seen in Fig. 4A and 4Ba,b that, in non-treated cells, expression of wild type or mutant HspB5 did not significantly change the level of phosphorylation of HspB1 as demonstrated by calculating the R120G/WT ratio, which is representative of the modulation of the phosphorylation of a define serine site in response to the mutation. In contrast to HspB1, the phosphoserine sites of HspB5 were stimulated by the R120G mutation. In response to menadione treatment, the phosphorylation of HspB1 serine sites was increased by about 4-fold. In the case of HspB5, the effect was less intense and showed a decreased intensity depending on the N-terminal position of the serine sites. However, in oxidative conditions the R120G mutation still strongly enhanced HspB5 phosphorylation. A similar observation was made when cells were exposed to hydrogen peroxide treatment. Phosphorylation was Nutlin-3 further analyzed following cell fractionation in a 10,0006g supernatant and pellet. As expected, in non-treated Neo and WT cells, all the phosphorylated proteins were recovered in the soluble fraction. Analysis of 
the fraction of HspB1 and HspB5 PLX-4720 Raf inhibitor present in the pellet fraction of R120G cells was performed by comparing, in the immunoblots, the ratios between the signals given by the percentage of the phosphorylated protein in the pellet to that of the percentage of the protein in that particular fraction. In non-treated R120G cells, the fraction of HspB1 in the pellet fraction had a phosphorylation index close to 1.0 while that of mutant HspB5 was in the range, depending on the serine site, of 0.2 to 0.5. This indicates that mutant HspB5 in the pellet fraction is less phosphorylated than its soluble counterpart and that HspB1 level of phosphorylation is less affected by the redistribution of this protein in the pellet fraction. Analysis of menadione treated Neo cells first showed that a large fraction of the total cellular content of HspB1 was recovered in the pellet fraction. The phenomenon was less intense in WT cells and more drastic in R120G cells. A similar observation was made for HspB5 in WT and R120G cells. Hence, the presence of these proteins in the pellet fraction correlated with the different sensitivity of these cells to oxidative stress. In response to menadione, HspB1 phosphorylation in Neo and WT cells was roughly proportional to the level of HspB1 in the fractions while it was slightly decreased in R120G cells. In contrast, HspB5 in WT cells showed a preferential phosphorylation in the pellet fraction. The phenomenon was not observed in R120G cells since the high level of menadione-induced phosphorylation was close to be proportional to the level of HspB5 present in the soluble and pellet fraction. To study the interaction between HspB1 and HspB5 in a defined cellular environment, HeLa cells were used since they constitutively express a high level of HspB1 but do not, or only weakly, express other interacting sHsps, particularly HspB5 and HspB6, which are known to form hetero-oligomeric complexes with HspB1. Genetically modified cells that stably express similar levels of either wild type or R120G mutated HspB5 were obtained. Of interest, HspB5 level of expression was close to that of endogenous HspB1. Another interesting characteristic was that more than 60% of HspB5 mutant expressed in these cells was recovered in a soluble form. This rather low level of aggregation suggests that the selected clones are probably adapted to the presence of the mutant protein. Indeed, R120G HspB5 polypeptide is well known for its drastic aggregation prone property; a phenomenon particularly intense in cells devoid of HspB1 expression.