Limits...
Binding of Kif23-iso1/CHO1 to 14-3-3 is regulated by sequential phosphorylations at two LATS kinase consensus sites.

Fesquet D, De Bettignies G, Bellis M, Espeut J, Devault A - PLoS ONE (2015)

Bottom Line: This S814 phosphosite was previously reported to constitute a 14-3-3 binding site, which plays a role in Kif23 clustering during cytokinesis.Surprisingly, we found that phosphorylation of the upstream S716 NDR/LATS consensus site, present only in the longest Kif23 isoform, is required for efficient phosphorylation at S814, thus revealing sequential phosphorylation at these two sites, and differential regulation of Kif23-14-3-3 interaction for the two Kif23 isoforms.Finally, we provide evidence that Kif23 is largely unphosphorylated on S814 in post-abscission midbodies, making this Kif23 post-translational modification a potential marker to probe these structures.

View Article: PubMed Central - PubMed

Affiliation: CRBM UMR 5237, CNRS and Université de Montpellier, Montpellier, France.

ABSTRACT
Kif23 kinesin is an essential actor of cytokinesis in animals. It exists as two major isoforms, known as MKLP1 and CHO1, the longest of which, CHO1, contains two HXRXXS/T NDR/LATS kinase consensus sites. We demonstrate that these two sites are readily phosphorylated by NDR and LATS kinases in vitro, and this requires the presence of an upstream -5 histidine residue. We further show that these sites are phosphorylated in vivo and provide evidence revealing that LATS1,2 participate in the phosphorylation of the most C-terminal S814 site, present on both isoforms. This S814 phosphosite was previously reported to constitute a 14-3-3 binding site, which plays a role in Kif23 clustering during cytokinesis. Surprisingly, we found that phosphorylation of the upstream S716 NDR/LATS consensus site, present only in the longest Kif23 isoform, is required for efficient phosphorylation at S814, thus revealing sequential phosphorylation at these two sites, and differential regulation of Kif23-14-3-3 interaction for the two Kif23 isoforms. Finally, we provide evidence that Kif23 is largely unphosphorylated on S814 in post-abscission midbodies, making this Kif23 post-translational modification a potential marker to probe these structures.

Show MeSH
Phosphorylation dependant interaction between Kif23 and 14–3–3.A. WT and mutant Flag-tagged Kif23-iso1 were expressed with myc-tagged 14–3–3 in HEK293T cells. Material immunoprecipitated with anti-myc antibodies was analysed by Western blot for the presence of Kif23. B. Histograms show amounts of Flag-Kif23 present on anti-myc immunoprecipitates corrected by the amount of Flag-Kif23 in extracts, calculated from A and three other experiments. P values refer to comparisons between WT and specified mutant Kif23.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4320110&req=5

pone.0117857.g005: Phosphorylation dependant interaction between Kif23 and 14–3–3.A. WT and mutant Flag-tagged Kif23-iso1 were expressed with myc-tagged 14–3–3 in HEK293T cells. Material immunoprecipitated with anti-myc antibodies was analysed by Western blot for the presence of Kif23. B. Histograms show amounts of Flag-Kif23 present on anti-myc immunoprecipitates corrected by the amount of Flag-Kif23 in extracts, calculated from A and three other experiments. P values refer to comparisons between WT and specified mutant Kif23.

Mentions: Unexpectedly, mutation of S716 to alanine not only impeded phosphorylation at S716 (Fig. 2B), it also strongly reduced that of S814 on isoform 1 (Fig. 4). On the opposite, mutation of S814 in Kif23-iso1 did not affect phosphorylation of S716 (Fig. 2B). Meanwhile, the shorter Kif23-iso2, lacking the S716 phosphosite, exhibited similar levels of S814 phosphorylation to that of Kif23-iso1 (S7A Fig.), comforting the conclusion that phosphorylation of S716 is necessary for phosphorylation on S814, rather than allowing hyper-phosphorylation of S814. This suggests that an ordered sequence of phosphorylation is taking place, with that on S716 occurring before S814. This could be due to unphosphorylated domain around S716 interacting with and masking S814. It was previously demonstrated that Kif23-iso2 interacts with 14–3–3 in a phospho-S814 dependent manner and that the peptide sequence surrounding phospho-S814 has 14–3–3 binding activity [25]. We could confirm this interaction by immunoprecipitation using tagged versions of Kif23-iso1 and 14–3–3-γ expressed in HEK293T cells (Fig. 5 and S7A Fig.), as well as with both endogenous Kif23 isoforms (S7B Fig.). Mutation of S814 to alanine led, as expected, to a strong, 6 fold decrease in 14–3–3 binding (Fig. 5). Mutation of the upstream H809 but not that of H711, significantly reduced binding to 14–3–3, although more modestly. We then asked if Kif23-iso1 S716A mutant could influence formation of Kif23/14–3–3 complexes. In keeping with its ability to negatively regulate S814 phosphorylation, S716A mutant strongly reduced binding to 14–3–3 (Fig. 5). It ensues that binding of Kif23-iso1 to 14–3–3 depends on the phospho-S814 binding site, which in turn is dependent on prior phosphorylation of the upstream S716. As Kif23-iso1 S814A mutant retained only residual binding capacity to 14–3–3, phospho-S716 is probably not a 14–3–3 binding site per se, and S716D mutation was unable to rescue 14–3–3 binding of the S814 mutant (S8 Fig.). Altogether, these results show that phospho-S814 is the main determinant of 14–3–3/Kif23 complex formation, while phosphorylation of S716 is necessary for that on S814. Since S716 phosphosite is only present in isoform 1, this raises the possibility that the two Kif23 isoforms are differentially regulated.


Binding of Kif23-iso1/CHO1 to 14-3-3 is regulated by sequential phosphorylations at two LATS kinase consensus sites.

Fesquet D, De Bettignies G, Bellis M, Espeut J, Devault A - PLoS ONE (2015)

Phosphorylation dependant interaction between Kif23 and 14–3–3.A. WT and mutant Flag-tagged Kif23-iso1 were expressed with myc-tagged 14–3–3 in HEK293T cells. Material immunoprecipitated with anti-myc antibodies was analysed by Western blot for the presence of Kif23. B. Histograms show amounts of Flag-Kif23 present on anti-myc immunoprecipitates corrected by the amount of Flag-Kif23 in extracts, calculated from A and three other experiments. P values refer to comparisons between WT and specified mutant Kif23.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4320110&req=5

pone.0117857.g005: Phosphorylation dependant interaction between Kif23 and 14–3–3.A. WT and mutant Flag-tagged Kif23-iso1 were expressed with myc-tagged 14–3–3 in HEK293T cells. Material immunoprecipitated with anti-myc antibodies was analysed by Western blot for the presence of Kif23. B. Histograms show amounts of Flag-Kif23 present on anti-myc immunoprecipitates corrected by the amount of Flag-Kif23 in extracts, calculated from A and three other experiments. P values refer to comparisons between WT and specified mutant Kif23.
Mentions: Unexpectedly, mutation of S716 to alanine not only impeded phosphorylation at S716 (Fig. 2B), it also strongly reduced that of S814 on isoform 1 (Fig. 4). On the opposite, mutation of S814 in Kif23-iso1 did not affect phosphorylation of S716 (Fig. 2B). Meanwhile, the shorter Kif23-iso2, lacking the S716 phosphosite, exhibited similar levels of S814 phosphorylation to that of Kif23-iso1 (S7A Fig.), comforting the conclusion that phosphorylation of S716 is necessary for phosphorylation on S814, rather than allowing hyper-phosphorylation of S814. This suggests that an ordered sequence of phosphorylation is taking place, with that on S716 occurring before S814. This could be due to unphosphorylated domain around S716 interacting with and masking S814. It was previously demonstrated that Kif23-iso2 interacts with 14–3–3 in a phospho-S814 dependent manner and that the peptide sequence surrounding phospho-S814 has 14–3–3 binding activity [25]. We could confirm this interaction by immunoprecipitation using tagged versions of Kif23-iso1 and 14–3–3-γ expressed in HEK293T cells (Fig. 5 and S7A Fig.), as well as with both endogenous Kif23 isoforms (S7B Fig.). Mutation of S814 to alanine led, as expected, to a strong, 6 fold decrease in 14–3–3 binding (Fig. 5). Mutation of the upstream H809 but not that of H711, significantly reduced binding to 14–3–3, although more modestly. We then asked if Kif23-iso1 S716A mutant could influence formation of Kif23/14–3–3 complexes. In keeping with its ability to negatively regulate S814 phosphorylation, S716A mutant strongly reduced binding to 14–3–3 (Fig. 5). It ensues that binding of Kif23-iso1 to 14–3–3 depends on the phospho-S814 binding site, which in turn is dependent on prior phosphorylation of the upstream S716. As Kif23-iso1 S814A mutant retained only residual binding capacity to 14–3–3, phospho-S716 is probably not a 14–3–3 binding site per se, and S716D mutation was unable to rescue 14–3–3 binding of the S814 mutant (S8 Fig.). Altogether, these results show that phospho-S814 is the main determinant of 14–3–3/Kif23 complex formation, while phosphorylation of S716 is necessary for that on S814. Since S716 phosphosite is only present in isoform 1, this raises the possibility that the two Kif23 isoforms are differentially regulated.

Bottom Line: This S814 phosphosite was previously reported to constitute a 14-3-3 binding site, which plays a role in Kif23 clustering during cytokinesis.Surprisingly, we found that phosphorylation of the upstream S716 NDR/LATS consensus site, present only in the longest Kif23 isoform, is required for efficient phosphorylation at S814, thus revealing sequential phosphorylation at these two sites, and differential regulation of Kif23-14-3-3 interaction for the two Kif23 isoforms.Finally, we provide evidence that Kif23 is largely unphosphorylated on S814 in post-abscission midbodies, making this Kif23 post-translational modification a potential marker to probe these structures.

View Article: PubMed Central - PubMed

Affiliation: CRBM UMR 5237, CNRS and Université de Montpellier, Montpellier, France.

ABSTRACT
Kif23 kinesin is an essential actor of cytokinesis in animals. It exists as two major isoforms, known as MKLP1 and CHO1, the longest of which, CHO1, contains two HXRXXS/T NDR/LATS kinase consensus sites. We demonstrate that these two sites are readily phosphorylated by NDR and LATS kinases in vitro, and this requires the presence of an upstream -5 histidine residue. We further show that these sites are phosphorylated in vivo and provide evidence revealing that LATS1,2 participate in the phosphorylation of the most C-terminal S814 site, present on both isoforms. This S814 phosphosite was previously reported to constitute a 14-3-3 binding site, which plays a role in Kif23 clustering during cytokinesis. Surprisingly, we found that phosphorylation of the upstream S716 NDR/LATS consensus site, present only in the longest Kif23 isoform, is required for efficient phosphorylation at S814, thus revealing sequential phosphorylation at these two sites, and differential regulation of Kif23-14-3-3 interaction for the two Kif23 isoforms. Finally, we provide evidence that Kif23 is largely unphosphorylated on S814 in post-abscission midbodies, making this Kif23 post-translational modification a potential marker to probe these structures.

Show MeSH