BBRC Biochemical and Biophysical Research Communications 317 (2004) 703–707 www.elsevier.com/locate/ybbrc
KIBRA is a novel substrate for protein kinase Cf Katrin B€ uther, Christian Plaas, Angelika Barnekow, and Joachim Kremerskothen* Department for Experimental Tumorbiology, University Muenster, Badestrasse 9, D-48149 Muenster, Germany Received 5 March 2004
Abstract WW domain-containing proteins are found in all eukaryotic cells and they are involved in the regulation of a wide variety of cellular functions. We recently identified the neuronal protein KIBRA as novel member of this family of signal transducers. In this report, we describe the identification of protein kinase C (PKC) f as a KIBRA-interacting protein. PKCf is known to play an important role in synaptic plasticity and memory formation but its specific targets are not well known. Our studies presented here revealed that KIBRA is a novel substrate for PKCf and suggest that PKCf phosphorylation may regulate the cellular function of KIBRA. Ó 2004 Elsevier Inc. All rights reserved. Keywords: KIBRA; WW domain; PKCf; Synaptic plasticity; Phosphorylation
Cellular processes are regulated by specific protein– protein interactions. Frequently, such interactions are mediated by conserved modular domains and lead to posttranslational protein modifications or the recruitment of globular complexes to specific cell compartments (reviewed in [1]). WW domains, that bind to proline-rich motifs in target molecules, have been identified in a variety of eukaryotic proteins involved in cellular processes like growth, differentiation or transcription control [2–5]. Recently, we identified KIBRA as a novel WW domaincontaining protein mainly expressed in brain and kidney [6]. In addition to the two WW domains at the aminoterminus of the protein, KIBRA possesses a C2-like domain that may mediate Ca2þ sensitivity [7]. In kidney cells from monkey (CV1) KIBRA displays a cytosolic distribution with a perinuclear enrichment [6]. To gain insight into the cellular function of KIBRA we performed a yeast two hybrid screen and isolated a fragment of PKCf as a KIBRA interacting protein. PKCf is a member of the family of atypical PKCs that are distinguished structurally by the presence of only one single zinc finger module and their inability to bind and to respond to phorbol esters and diacylglycerol [8–10]. Many studies have shown that PKCf is involved in signal transduction processes which regulate various cell pro* Corresponding author. Fax: +49-251-832-8390. E-mail address:
[email protected] (J. Kremerskothen).
0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.03.107
cesses like growth and differentiation [9,10]. The major activation pathway of PKCf depends on phosphatidylinositol(PI)-3,4,5-triphosphate (PIP3 ), which is mainly produced by PI-3 kinase [9,10]. Upon activation, PKC zeta associates with other signaling molecules, as well as scaffolding proteins, to form large complexes [8–10]. Recent studies have shown that PKCf activity is critical for the consolidation of long-term potentiation and the formation of memory [11,12]. Although the role of PKCf in cellular signaling cascades has been extensively studied, there is only little information about the specific substrates of this PKC isoform. The findings reported here indicate that PKCf directly associates through its catalytic domain with KIBRA and phosphorylates the WW domain-containing protein at its carboxyterminus. Materials and methods Reagents. All culture media, fetal calf serum (FCS), and supplements were obtained from Invitrogen (Karlsruhe, Germany). Standard reagents (highest grade) were from Sigma–Aldrich (Deisenhofen, Germany). Purified PKCf and the lipid activator mixture were from Upstate (Lake Placid, USA). Plasmids. Fragments of human KIBRA and PKCf were cloned into yeast two hybrid vectors (pPGBT9, pAS2-1, pACTII, Clontech) and bacterial expression vectors (pGEX-KG). For localization studies, the coding region of full-length KIBRA (aa 1–1113) was cloned into the NcoI/XhoI sites of a modified pSV-SPORT vector (pSV42-mycKIBRA). Details are available from J.K.
704
K. Bu¨ther et al. / Biochemical and Biophysical Research Communications 317 (2004) 703–707
Antibodies. Myc-tagged KIBRA was detected with the monoclonal antibody 9E10 [6]. Antibodies directed against PKCf were from Sigma–Aldrich (Deisenhofen, Germany). Peptide-specific antibodies against PKCa and PKCe were a generous gift from Todd Sacktor (SUNY Brooklyn, USA) and were described earlier [11]. Horseradish peroxidase and fluorochrome conjugated secondary antibodies were purchased from Amersham Bioscience (Freiburg, Germany) and Molecular Probes (Eugene, USA), respectively. Yeast two hybrid screen. Using full-length KIBRA as bait, a human brain library (Clontech, >5 106 independent clones) was screened in yeast strain Y190 according to the manufacturer’s instructions. Individual colonies growing on selective media and containing potentially positive clones were isolated, re-tested for interaction with KIBRA, and false-positive clones were discarded. The inserts from the positive clones were sequenced. For co-transformations, Y190 yeast cells were simultaneously transformed with the corresponding bait/prey constructs and the interaction was analyzed by plating on selective media and b-galactosidase filter assays [13]. Cell culture, transfections, and immunocytochemistry. Cultivation, transfection, and immunocytochemistry of HeLa cells were described previously [6]. Cell extracts. HeLa cells were washed with ice-cold PBS (phosphate-buffered saline) and scraped into lysis buffer containing 10 mM _ Tris/HCl, pH 7.4, 150 mM NaCl, 1mM MgCl2 , 1 mM CaCl2 , 0.2% Triton X-100, and protease inhibitor mix Complete (Roche). Rat brain extracts were prepared in lysis buffer as described previously [14]. Lysates were centrifuged at 12,000g for 60 min at 4 °C and supernatant fractions were stored in aliquots at )70 °C. Recombinant proteins and GST pull downs. pGEX-KG constructs containing KIBRA fragments were used for the IPTG-induced expression of GST fusion proteins in Escherichia coli BL21. Recombinant proteins were isolated from bacterial lysates with glutathione–Sepharose beads (Amersham-Bioscience, Freiburg, Germany) according to the manufacturer‘s instructions. Tissue or cell lysates were combined with approximately 5 lg of GST fusion proteins immobilized on glutathione–Sepharose beads and incubated overnight at 4 °C. Beads were washed extensively with lysis buffer, bound proteins were eluted by boiling in SDS sample buffer, subjected to SDS– PAGE, and analyzed by Western blotting as described before [6]. Phosphorylation analysis. For the phosphorylation of potential substrates, 20 ng of purified PKCf (1 ll) was mixed with 100 ng of recombinant proteins diluted in 9 ll ABD III buffer (20 mM Mops, pH 7.2, 25 mM b-glycerolphosphate, 1 mM sodium orthovanadate, and 1 mM DDT). After addition of 10 ll sonicated lipid activator (phosphatidylserine/diaglycerol) and 10 ll [32 P]ATP, samples were incubated at 30 °C for 10 min. Phosphorylated proteins were separated by SDS– PAGE and dried gels were exposed to X-ray films using intensifying screens. Filter assays were done with spotted KIBRA peptides (1.3 mg/spot) that contain potential PKC phosphorylation motifs. The filters were activated in methanol for 5 min, washed three times in ABD III buffer, and blocked in ABD III containing 3% BSA. After washing with ABD III, 100 ng purified PKCf diluted in 320 ll ABD III, 40 ll lipid activator, and 40 ll diluted [32 P]ATP were added and filters were incubated for 60 min at 30 °C. After washing with ABDIII containing 1 M NaCl, filters were dried and exposed to X-ray films.
Results and discussion KIBRA interacts with the catalytic domain of PKCf Using full-length KIBRA as bait in a yeast two hybrid screen with a human brain cDNA library we isolated several clones that bound specifically to KIBRA
but not to control baits. One of the clones (K16) contains the sequence encoding the catalytic kinase domain of the atypical PKC isoform f containing the ATP binding region and the activation loop but lacks regulatory motifs such as the PB1 domain or the pseudosubstrate sequence (PS, Fig. 1A). The kinase activity of PKCf is known to be controlled by an autoinhibition mechanism through an interaction of the pseudosubstrate sequence and the substrate-binding cavity of the kinase domain [8]. In addition, the ATP-binding region inside the catalytic domain contains a Lys residue (K281), which is crucial for the kinase activity [10]. Transformation studies using the yeast two hybrid system revealed that full-length PKCf does not interact with KIBRA when co-expressed in yeast (Fig. 1A). This finding can be explained by the autoinhibition mechanism and was already described for the interaction of PKCf and its substrate heterogeneous ribonucleoprotein A1 (hnRNP A1 [15]). The interaction with KIBRA needs the complete PKCf kinase domain because overlapping fragments containing the aminoterminus or the carboxyterminus, respectively, do not bind to KIBRA
Fig. 1. Mapping of binding sites on PKCf and KIBRA. Deletion mutants of PKCf (A) and KIBRA (B) as shown were co-expressed in Y190 yeast cells. Growth on selective media and b-galactosidase activity indicated an interaction (+). No interaction is indicated by ()). Numbers represent amino acid positions. (A) The complete catalytic domain of PKCf, but not the full-length protein containing the regulatory domain with the PB1, the pseudosubstrate (PS), and the C1 motif, is sufficient and necessary for an interaction with KIBRA. (B) PKCf interacts with a small KIBRA fragment of 44 aa (953–996) that contains four potential PKC phosphorylation sites (S967 , S975 , S978 , and S981 ).
K. Bu¨ther et al. / Biochemical and Biophysical Research Communications 317 (2004) 703–707
(Fig. 1A). Mapping of the binding site in the KIBRA sequence revealed that a short fragment of 44 amino acids containing four potential PKC phosphorylation consensus motifs [8] was sufficient for an interaction with PKCf (Fig. 1B). KIBRA associates specifically with the PKCf and the PKMf isoform
705
rable to the yeast co-transformation studies, no binding was observed between GST–KIBRA661–1113 and the classical PKC isoform alpha as well as the novel PKC isoform e although these proteins were abundant in the used brain extract (Fig. 2, lanes 5–8). These findings confirm the specificity of the interaction between KIBRA and the atypical PKCf isoform. KIBRA and PKCf colocalize in HeLa cells
The binding of KIBRA to PKCf was isoform-specific because PKCa as a member of the classical PKC family as well as PKCe, which represents the novel PKC family, did not interact with KIBRA in the yeast two hybrid system (data not shown). The specificity of the PKC zeta–KIBRA interaction was further controlled by GST pull-down experiments. When we used the carboxyterminal part of KIBRA (KIBRA661–1113 ) coupled to GST we were able to pull down PKCf from HeLa cell extracts as well as from rat brain extract (Fig. 2, lanes 2 and 4). No binding was found with GST (data not shown). Interestingly, we also observe an interaction of KIBRA with PKMf (55 kDa), a constitutively active kinase that is exclusively expressed in brain [16]. Antibodies against the carboxyterminus of PKCf cross-react with PKMf which is homolog to the catalytic domain of PKCf representing not a proteolytic fragment of PKCf but alternatively transcribed PKC isoform [16]. Additionally, we could detect a third unknown protein (62 kDa) in HeLa cell extract and in rat brain extract that binds specifically to the GST–KIBRA661–1113 fragment but not to GST and that was recognized by the anti-PKCf antibodies (Fig. 2, lanes 2 and 4). Compa-
To further analyze the KIBRA–PKCf interaction in vivo we determined their localization in eukaryotic cells. As our anti KIBRA antibodies were inapplicable for immunohistochemistry, HeLa cells were transfected with a plasmid encoding myc-tagged KIBRA and the recombinant protein was detected with anti-myc antibodies. Like PKCf, myc-KIBRA displayed a cytoplasmic staining and especially at perinuclear regions, both proteins showed a distinct colocalization (Fig. 3). Overexpression of KIBRA did not change distribution of PKCf (data not shown). These experiments revealed that KIBRA and PKCf are localized in the same subcompartments of HeLa cells and further support their putative interaction. KIBRA is a novel substrate for PKCf The KIBRA sequence contains several potential PKC phosphorylation sites with the consensus motif S/TXR/ K [8]. To test whether KIBRA is not only an interacting protein but a novel substrate of PKCf we performed phosphorylation assays with a GST–KIBRA661–1113
Fig. 2. Isoform-specific binding of PKCf/PKMf and KIBRA. Recombinant GST–KIBRA661–1113 was used for pull-down experiments. Proteins from a HeLa cell extract or rat brain extract were analyzed by Western blotting using PKC isoform-specific antibodies. Only PKCf (lanes 1–4) and PKMf (lanes 3 and 4) bound to GST–KIBRA661–1113 . No interaction was detected with classical (PKCa, lanes 5 and 6) or novel (PKCe, lanes 7 and 8) PKCs. E, cell or tissue extract (15% of the input); PD, isolated proteins after pull down with GST–KIBRA661–1113 . All experiments were performed at least two times with similar results.
Fig. 3. PKCf and KIBRA are colocalized in HeLa cells. HeLa cells were transiently transfected with a construct encoding myc-tagged KIBRA and were fixed after 24 h. Endogenous PKCf and recombinant myc-KIBRA were detected by indirect immunohistochemistry.
706
K. Bu¨ther et al. / Biochemical and Biophysical Research Communications 317 (2004) 703–707
Fig. 4. KIBRA serine residues S975/S978 are phosphorylated by PKCf. (A) GST, GST–KIBRA432–661 , GST–KIBRA661–113 , and myelin basic protein (MBP) were incubated in a phosphorylation reaction with purified PKCf and [32 P]ATP. Phosphorylated proteins were separated by SDS–PAGE and analyzed by autoradiography. Only GST–KIBRA661–1113 and MBP are in vitro substrates for PKCf. (B) Phosphorylation reactions were performed with PKC zeta and immobilized KIBRA peptides corresponding to putative consensus motifs for PKC phosphorylation. In this assay, only P2 was a substrate for PKCf whereas P1 and P3 were not phosphorylated. Successive replacement of individual serine residues (P4–P9) demonstrated that S975 and S978 were independently phosphorylated by PKCf. All phosphorylation assays were repeated two or three times.
fusion protein and purified PKCf. Autoradiography of the phosphorylated proteins separated by SDS–PAGE revealed that KIBRA661–1113 (Fig. 4A, lane 3) and the known PKC substrate myelin basic protein (Fig. 4, lane 4) but not GST–KIBRA432–661 (Fig. 4A, lane 2) or GST (Fig. 4A, lane 1) were efficiently phosphorylated by PKCf. These data indicate that in vitro the carboxyterminal part of KIBRA (aa 661–1113) is a specific substrate for PKCf. Because the fusion protein GST– KIBRA661–1113 was higly sensitive to proteolysis, we got multiple fragments of the recombinant protein that were phosphorylated by PKCf (Fig. 4A, lane 3). PKCf phosphorylates KIBRA at serine 975 and 978 The 44 aa KIBRA fragment that was sufficient for an interaction with KIBRA in the yeast co-transformations (Fig. 1A) contains four potential PKC phosphorylation sites (S967 , S975 , S978 , and S981 ). To determine which of these serine residues can be phosphorylated by PKCf we performed peptide phosphorylation assays. For that purpose, the spotted peptides LERRS967 VRMKRP (P1), MKRPSS975 VKS978 LR (P2), and VKS978 LRS981
ERLIR (P3), respectively, were incubated with purified PKCf in a phosphorylation reaction and incorporation of radioactive [32 P]phosphate was determined thereafter by autoradiography. In this test, only P2 was efficiently phosphorylated whereas P1 and P3 showed only background labeling (Fig. 4B). Using mutated peptide sequences with individual serine to alanine point mutations, we could prove that S975 and S978 are independently phosphorylated by PKCf. A replacement of both serine residues (S975 and S978 ) in P8 resulted in a very weak phosphorylation whereas the exchange of only one residue (P4–P7) displayed only minor effects (Fig. 4B). Interestingly, the P3 peptide, that also contained the S978 residue, was not phosphorylated by PKCf, indicating that the amino acids in position )3 or further upstream of the phosphorylation site were important for the substrate recognition. In summary, our data indicate that the WW domaincontaining protein KIBRA interacts specifically with the atypical PKC isoformf as well as with the brain-specific PKMf and that at least in vitro, KIBRA is a substrate for PKC/PKMf phosphorylation. The biochemical consequence of the phosphorylation for the cellular function of KIBRA is unknown so far. For hnRNP A1 it was shown that phosphorylation by PKC zeta affects its cytoplasmic accumulation as well as its RNA-binding activity [15]. Because overexpression of PKCf or treatment of HeLa cells with PKCf-specific inhibitors does not influence the cellular distribution of KIBRA (data not shown), we conclude that KIBRA localization is not regulated by PKC phosphorylation. KIBRA is able to form antiparallel homodimers and the 44 aa fragment that contains the PKCf phosphorylation sites is part of the dimerization domain (Kremerskothen et al., unpublished results). Therefore, we speculate that PKCf phosphorylation may regulate KIBRA dimer formation as it was shown recently for the polycomb-group protein Mel-18 [17]. We already demonstrated that KIBRA interacts via its WW domains with the postsynaptic protein Dendrin, which is translated from a dendritic mRNA and is thought to play a role in synaptic plasticity [6,18]. Phosphorylation of KIBRA by PKCf or PKMf, those PKC isoforms that are also involved in the regulation of synaptic transmission, may affect the binding capacity or specificity of the WW domains and therefore the association of KIBRA with other signaling molecules. Undoubtedly, further work is necessary to get more insights into the cellular function of KIBRA and the effects of PKCf phosphorylation. Acknowledgments The authors thank Todd Sacktor, Jorge Moscat, Axel Puls, Graeme Guy for their generous gifts, and Wolfgang Nastainczyk for the synthesis of the peptide libraries.
K. Bu¨ther et al. / Biochemical and Biophysical Research Communications 317 (2004) 703–707
References [1] T. Pawson, M. Raina, P. Nash, Interaction domains: from simple binding events to complex cellular behavior, FEBS Lett. 513 (2002) 2–10. [2] M. Sudol, Structure and function of the WW domain, Prog. Biophys. Mol. Biol. 65 (1996) 113–132. [3] Einbond, M. Sudol, Towards prediction of cognate complexes between the WW domain and proline-rich ligands, FEBS Lett. 384 (1996) 1–8. [4] M. Sudol, K. Sliwa, T. Russo, Functions of WW domains in the nucleus, FEBS Lett. 490 (2001) 190–195. [5] J.L. Ilsley, M. Sudol, S.J. Winder, The WW domain: linking cell signalling to the membrane cytoskeleton, Cell Signal. 14 (2002) 183–189. [6] J. Kremerskothen, C. Plaas, K. Buther, I. Finger, S. Veltel, T. Matanis, T. Liedtke, A. Barnekow, Characterization of KIBRA, a novel WW domain-containing protein, Biochem. Biophys. Res. Commun. 300 (2003) 862–867. [7] J. Rizo, T.C. Sudhof, C2-domains, structure and function of a universal Ca2þ -binding domain, J. Biol. Chem. 273 (1998) 15879– 15882. [8] A. Toker, Signaling through protein kinases C, Front. Biosci. 3 (1998) 1134–1147. [9] J. Moscat, M.T. Diaz-Meco, The atypical protein kinase Cs, EMBO Rep. 1 (2000) 399–403. [10] T. Hirai, K. Chida, Protein kinase C zeta: activation mechanisms and cellular functions, J. Biochem. (2003) 1–7. [11] T.C. Sacktor, P. Osten, H. Valsamis, X. Jiang, M. Naik, E. Sublette, Persistent activation of the zeta isoform of protein kinase
[12]
[13]
[14]
[15]
[16]
[17]
[18]
707
C in the maintenance of long term potentiation, Proc. Natl. Acad. Sci. USA 90 (1993) 8342–8346. E.A. Drier, M.K. Tello, M. Cowan, P. Wu, N. Blace, T.C. Sacktor, J.C.P. Yin, Memory enhancement and formation by atypical PKM activity in Drosophila melanogaster, Nat. Neurosci. 5 (2002) 316–324. T. Weide, M. Bayer, M. Koster, J.P. Siebrasse, R. Peters, A. Barnekow, The Golgi matrix protein GM130: a specific interacting partner of the small GTPase rab1b, EMBO Rep. 2 (2001) 336–431. J. Kremerskothen, M. Nettermann, A. op de Bekke, M. Bachmann, J. Brosius, Identification of human autoantigen La/SS-B as BC1/BC200 RNA-binding protein, DNA Cell Biol. 17 (1998) 751– 759. M.M. Municio, J. Lozano, P. Sanchez, J. Moscat, M.T. DiazMeco, Identification of heterogenous ribonucleoprotein A1 as a novel substrate for protein kinase C zeta, J. Biol. Chem. 270 (1995) 15884–15891. A.I. Hernandez, N. Blace, J.F. Crary, P.A. Serrano, M. Leitges, J.M. Libien, G. Weinstein, A. Tcherapanov, T.C. Sacktor, Protein kinase C zeta synthesis from a brain mRNA encoding an independent protein kinase C zeta catalytic domain, J. Biol. Chem. 278 (2003) 40305–40316. S. Fujisaki, Y. Ninomya, H. Ishihara, M. Miyazaki, R. Kanno, T. Asahara, M. Kanno, Dimerization of the Polycomb protein Mel18 is regulated by PKC phosphorylation, Biochem. Biophys. Res. Commun. 300 (2003) 135–140. A. Herb, W. Wisden, M.V. Catania, D. Marechal, A. Dreese, P. Seeburg, Prominent localization in forebrain neurons of a novel mRNA and its product, Dendrin, Mol. Cell Neurosci. 8 (1997) 367–374.
