Noteworthy the transient downregulation of
Noteworthy, the transient downregulation of the regulatory CK2β subunit reduces HSP27 expression in all the three different human cell lines (Fig. 2b). Similarly, the downregulation of one of the two catalytic subunits, CK2α or CK2α′, is sufficient to affect HSP27 expression without affecting other HSPs, with the exception of HSP105 that is instead strongly downregulated in HeLa FUT-175 (Fig. S2). The link between CK2 and HSP27 could be direct or mediated by a signalling cascade. Firstly, it has been checked if CK2 physically interacts with or participates to the same molecular complexes of HSP27 using co-immunoprecipitation, confocal immunofluorescence microscopy, and protein complexes fractionation by glycerol gradient sedimentation. C2C12 and HeLa cell lysates were subjected to immunoprecipitation experiments using CK2α antibody. HSP27 co-immunoprecipitates with CK2α in both cell lysates while no protein was detected using a pre-immune serum (Ctrl), suggesting an interaction between these two proteins (Fig. 3a). Similarly, both the CK2 catalytic subunits, CK2α and CK2α′ co-purify with HSP27 after HSP27 immunoprecipitation from C2C12 and HeLa lysates (Fig. 3a). To further verify the complex formation between CK2 and HSP27, co-localization experiments have been performed using double label fluorescence confocal microscopy. Overlay of images obtained by confocal microscopy of C2C12 cells co-labelled with antibody against HSP27 and CK2, confirms a partial co-localization of the two signals (Fig. 3b). Furthermore, a quantitative measurement of the amount of HSP27 interacting with CK2 was obtained separating molecular complexes by fractionation of C2C12 cell lysate into a glycerol gradient sedimentation experiment. Fig. 3c shows that the peaks of distribution of HSP27 and CK2α/α′ are partially overlapped, mainly in complexes of high molecular weight. CK2 substrate selection is strongly dependent to the presence of acidic residues surrounding the phosphoacceptor site [10,21] and, indeed, potential CK2 targets may be suggested relying only on the primary structure of the protein . For this reason, HSP27 does not represent a potential canonical CK2 substrate as its sequence does not contain any Ser/Thr residue fulfilling the minimal CK2 consensus sequence, i.e. S/TxxD/E. However, since about the 20% of the CK2 bona fide substrates do not conform to the canonical CK2 consensus sequence [9,10], an in vitro phosphorylation assay of HSP27 purified from bacteria (Fig. 4a) or immunoprecipitated from transfected eukaryotic cells (Fig. 4b) has been performed in presence of recombinant tetrameric CK2 and [γ-33P]ATP. As displayed in Fig. 4a, recombinant HSP27 is not phosphorylated at all in vitro by CK2. However, we cannot exclude that previous posttranslational modifications present in eukaryotic cells but not in a recombinant protein purified from bacteria, may be required for the phosphorylation by CK2 . To assess this possibility, FLAG-HSP27, immunoprecipitated from previously transfected HeLa cells, has been subjected to in vitro phosphorylation by recombinant tetrameric CK2. Fig. 4b shows that even immunoprecipitated FLAG-HSP27 does not undergo CK2-dependent phosphorylation. Summarizing, these experiments demonstrate that HSP27 is not a CK2 substrate and therefore the mechanism by which CK2 affects HSP27 expression must be indirect. A priori, CK2 might influence the level of HSP27 either by affecting its stability or by modulating the transcription of its gene. However, taking into account that in C2C12 HSP27 is selectively regulated by CK2 while other HSPs, namely HSP70, HSP90 and HSP105, are not (see Fig. 1a), and that these chaperones are all under the control of the same transcription factor (HSF-1), the latter hypothesis looks very unlike as it would imply an HSF-1 independent mechanism. Accordingly, relative mRNA expression of HSP27 quantified by Real-time PCR in wilt-type and CK2 knockout cell lines shows no or small differences according to the housekeeping gene used for normalization (Fig. 5a). On the other hand, it is well known that CK2 regulates the turnover of many proteins by both direct phosphorylation and indirect mechanisms . The former hypothesis therefore has been checked. Since proteins are mainly degraded by either the lysosomal or proteasomal pathways, we examined the effect of inhibitors specific for the two proteolytic pathways on the HSP27 protein levels, bearing in mind, however, that proteasomal inhibition might induce HSP27 transcription . Wild-type, CK2α/α′(−/−) or CK2β(−/−) cells have been incubated with either the autophagy inhibitor bafilomycin A1 or the proteasomal inhibitor MG132. The protein level of HSP27 significantly increases when the cells have been treated with MG132 in both knockout but not in wild-type cells, whereas there is no significant difference with bafilomycin A1 treatment (Fig. 5b). Since the increase of HSP27 expression is observed only in knockout cells, we can conclude that the MG132 effect after 6 h long treatment is mainly due to a block of protein degradation rather than a transcription stimulation. Indeed, an effect related to the increase in transcription would require a longer incubation. These data suggest that the reduced HSP27 expression might be a result of an enhanced proteasomal degradation and that CK2 activity is required to maintain the stability of this chaperone. HSP27 proteasomal degradation is under the control of SMURF2 ubiquitin ligase , an HECT (homologous to the E6-accessory protein C-terminus)-type E3 ubiquitin ligase also involved in TGF-β and BMP signalling. Therefore, it has been checked if CK2 regulates HSP27 stability affecting its ubiquitin ligase. Fig. 5c shows that indeed SMURF2 protein expression is up-regulated in both CK2α/α′(−/−) and CK2β(−/−) cells. Again, we wondered if also a transient CK2 downregulation or an inhibition of its enzymatic activity by small molecule inhibitors treatment are able to increase SMURF2 expression. In both experimental conditions, i.e. siRNA CK2β downregulation (Fig. 5d) and CX-4945 treatment (Fig. 5e), the increment of SMURF2 expression inversely correlates with a lower HSP27 protein level.