Previously published data showed an interaction

Previously published data showed an interaction of α-tubulin and CK1δ [15] that was validated in the present study using SPR (Fig. 1A). The dominant motifs involved in the CK1δ/α-tubulin interface were defined using a peptide library of peptides covering the CK1δ amino galactoside sequence (Supplementary Table S1). Our analysis identified peptide 39 (P39) as prominent binding partner of α-tubulin in SPR (Fig. 1B) and ELISA (Fig. 1C). While SPR identified CK1δ-peptides 1, 5, 9, 18, and 39 as interactors, peptides 9 and 18 could not be confirmed in ELISA. The observed differences could rely on the distinct sensitivity of both assays, on three-dimensional differences in binding or altered binding energetics by the use of different buffers. P39 encompasses the CK1δ sequence aa361 to aa375 and entails a serine at position 370 known to be phosphorylated by Chk1, PKC, Clk2, Akt and PKA (reviewed in Ref. 4). These findings are of particular interest since previous studies showed a tremendous influence of the phylogenetically highly conserved S370 on the modulation of CK1δ activity [30], [41]. Peptide therapeutics are generally recognised as potential selective drugs reproducing “natural” motifs, thereby affecting kinase activity only towards specific target(s) by interfering with the kinase/substrate interface [42]. Therefore, the effect of P39 on CK1δ-mediated phosphorylation of α-tubulin was investigated by performing kinase assays in the presence of P39 (Fig. 2). The addition of P39 inhibits CK1δ mediated phosphorylation of α-tubulin seemingly in a dose dependent manner, presumably by the recruitment of P39 to the α-tubulin docking site of CK1δ thus disrupting the interaction surface between CK1δ and α-tubulin. To support our in vitro data an in silico approach investigated the interaction of P39 and α-tubulin (Fig. 3). Potential interfaces between CK1δ peptide and α-tubulin 4A were modelled. Energy calculations of all three tested complexes (S370pS/P39/S370A – α-tubulin 4A) showed that interfaces are energetically stable. Subsequent docking calculations identified differences among the binding modes of the peptides, the most outstanding of which were (a) increased electrostatic energy for the phosphorylated peptide, (b) increased cluster number, and therefore lower specificity for the Ala substituted peptide, and (c) lower, overall energetic contributions for the Ala substituted peptide. Our results portend a direct influence of P39 on α-tubulin, on an interface which not necessarily disrupts tubulin polymerisation. Our predictions are in agreement with fluorescence thermal shift (FTS) assays, where the thermal stability of α-tubulin in the presence of P39 was investigated (Fig. 4). α-tubulin melting temperature was decreased by addition of CK1δ-peptide P39, an effect which was even more pronounced by replacing S370 with phosphoserine or with glutamic acid, expected to behave as a phospho-mimic. Therefore it can be concluded that phosphorylation of S370 increases binding affinity of P39 to α-tubulin, corroborated by comparative docking calculations (Table 1), implying a coordinated involvement of CK1 with other cellular kinases such as Chk1 in the regulation of α-tubulin and mitotic progression [30]. A disturbance within these processes by addition of P39 should impair mitotic function in cell culture. Thus, CV-1 cells, stably transfected with EYFP-tubulin, were treated with P39 using a C. botulinum derived transport system. While no effect on the microtubule phenotype could be observed for transport components, cells incubated with biotinylated P39, either with or without C2IN-Str and C2IIa, showed vesicular structures (Fig. 6). To exclude a shuttling effect of the Biotin-tag, CV-1 cells, stably transfected with EYFP-tubulin and HeLa cells, transiently transfected with RITA were treated with unbiotinylated P39 and monitored via epifluorescence microscopy live-cell imaging. Although peptides often do not cross the cell membrane (reviewed in Ref. 43), we could observe an inhibition in mitotic progression and disruption of cells entering mitosis in P39 treated cells (Fig. 7). These results are in accordance with a scenario whereby P39 perturbs interactions between α-tubulin and CK1δ, hampering a proper phosphorylation of α-tubulin by CK1δ, thereby leading to a reduction of microtubule stability during mitosis. It is likely that phosphorylation of P39 by cellular kinases is necessary to increase its binding affinity towards α-tubulin, in accordance with that hypothesis, S370A, unable to being targeted for phosphorylation, showed no effect in cell culture while S370pS displayed effects comparable to P39. Since it is known that the CK1δ/α-tubulin association significantly increases after induction of DNA-damage [15], a scenario would be likely where DNA-damage increases activation of cellular kinases, such as Chk1, known to phosphorylate S370 of CK1δ [30] thereby enhancing the association of CK1δ and α-tubulin allowing subsequent phosphorylation of tubulin. In addition to our results, showing that α-tubulin stability decreases in the presence of P39 (Fig. 4), it could already been shown that phosphorylation of tubulin decreases MT stability [44]. Therefore, CK1δ/α-tubulin interaction could be crucial for initiation of the microtubule catastrophe upon DNA-damage. In summary, the novel information provided by our work extends the current knowledge about the CK1δ/α-tubulin interaction that could be used for the development of new pharmacological tools in the promising field of peptide therapeutics [43]. These peptides are expected to display a much higher specificity, reducing off-target effects without suppressing the whole array of signals generated from CK1; they may prove also useful for patients developing resistances against conventional kinase inhibitors. In perspective it will be crucial to obtain further information about the role of CK1δ in α-tubulin regulation, especially in mitotic cells, and the precise mechanism of action of P39.