Posters P1 - CiteSeerX

Posters P1 Structural bases of protein kinase CK2 inhibition Roberto Battistutta1,2, Marco Mazzorana2,3, Elena Papinutto2,3, Stefania Sarno2,3 and Lorenzo A. Pinna2,3 1Department

of Chemical Sciences, 2Venetian Institute of Molecular Medicine (VIMM), 3Department of Biological Chemistry, University of Padua, Padua, Italy e-mail: [email protected] CK2, one of the first protein kinase ever discovered, is an eukaryotic acidophilic Ser/Thr protein kinase. CK2 is considered a quite anomalous protein kinase for the following peculiar properties: i) it is highly pleiotropic, ii) it can use both ATP and GTP as co-substrate, iii) the target serine or threonine must be surrounded by acidic residues (with the minimal consensus sequence Ser/Thr-X-X-Asp/ Glu), and iv) the CK2α catalytic subunit is intrinsically active [1]. More than 300 substrates are known for CK2. The regulatory mechanism of this kinase is still a matter of debate, and it was the subject of extensive investigation. The catalytic subunit of this enzyme is intrinsically active. CK2 is involved in many cellular processes such as cell cycle regulation, circadian rhythms, gene expression, cell growth and differentiation, embryogenesis and apoptosis. CK2 can be considered a valuable drug target for cancer therapy essentially on the basis of the following arguments: a) at protein level, CK2 is elevated in various cancers; b) it is a potent suppressor of apoptosis and strongly promotes the survival of the cell; c) it strengthens the multi-drug resistant phenotype; d) for the previous reasons, it establishes favourable conditions for tumorigenesis [2]. An important CK2 feature that influences the inhibitor design process is constitutive activity, with the consequence that only the active conformation can be targeted. The catalytic site of CK2 displays some unique properties that can be exploited in the design of inhibitors with a high degree of specificity, as indicated by the ability to utilize both ATP and GTP and by the low sensitivity to staurosporine inhibition (IC50 of 19.5 μM versus values in the low nanomolar range for other kinases). Actually, as described below, fairly specific, potent, and cell-permeable inhibitors of CK2 have been successfully developed in the last years [3–7]. From the analysis of the known maize and human CK2α co-crystal structures, it was noted that if a negatively charged moiety is present in a ligand (inhibitor or co-substrate) it tends to cluster in a well specific zone of the ATPbinding cleft, near the salt bridge Lys68-Glu81. A quantitative analysis of the electrostatic potential in the CK2α active site revealed the presence of a positively charged region located in the deeply buried area of the cavity,

between the hydrophobic region I and the salt bridge formed by the fully conserved Lys68 and Asp81, with a mean positive electrostatic potential of 1.5–2.0 kcal/mol. As seen by the systematic analysis of the binding of different classes of CK2 inhibitors, the electrostatic interaction with this area is responsible for the different orientation of the ligands in the active site of CK2. A striking example of this effect is that seen for the different binding modes of the two closely related tetrabromobenzo derivatives TBB and TBI. TBB, with a pKa ~5, binds with the triazole ring inside the positive area, while TBI, with a pKa ~9, is shifted towards the hinge region and forms two halogen bonds with Glu114 and Val116, like all the other tetrabromobenzo-imidazole derivatives analysed so far. In the apo form of CK2α, the positive electrostatic area is occupied by three water molecules. The one in the deepest part of the cavity, called water molecule 1 (W1), is highly conserved in all the known human and maize CK2α crystal structures. It makes hydrogen bonds with the amidic NH of Trp176, with a carboxylic oxygen of Glu81 and with another water molecule (W2), that is present in many structures. When W2 is absent, its position is invariably occupied by a portion of a ligand, as in the case of MNA, MNX, Emodin, IQA or benzamidine, and this suggests that it is directly expelled by the ligand itself, and that this water should be considered a sort of competitor for that position. The third water of the positive area of apoCK2α, W3, is present in only two other structures, namely in the complexes with TBI and K22; in the complexes with DMAT and DRB a chloride ion was found in that position. In the other cases, W3 is usually replaced by atoms of the bound ligand and, most importantly, by functional groups that can carry a negative charge. In other words, ligands carrying an acidic function have a propensity to cluster in a position corresponding to that of waters W2 and W3, in the region with the positive electrostatic potential at about 3.5 Å from Lys68. Ligands without acidic functions prefer to interact with the hinge region, in particular with the backbone carbonyls of Glu114 and Val116. The scaffold of the macrocyclic pyrazolo-triazines is so extended that it occupies almost entirely the CK2 binding pocket; in this case, W3 is substituted by the lactam carbonyl function that anchors the compound to the positive electrostatic area. For many CK2 inhibitors, the main energetic contribution to the binding appears to be due to apolar forces, namely hydrophobic interactions and van der Waals contacts, involving the hydrophobic surface of the CK2 binding cleft formed by residues Leu85, Val95, Leu111, Phe113, and Ile174 (hydrophobic region I), Val53, Ile66, Val116 and Met163 (adenine region) and Val45 and Tyr115 (hydrophobic region II). In particular, for the tetrabromobenzo derivatives, a linear correlation between the log (Ki) and the variation in the accessible surface area (ΔASA) upon binding was identified, indicating that the apolar interactions are ultimately responsible for their rank in potency,

6th International Conference: Inhibitors of Protein Kinases Vol. 56 39

as confirmed by a LIE model. Furthermore, the structureactivity analysis of more than 60 different coumarins and the derived LIE model showed that apolar interactions give the largest contribution to the free energy of binding also for this class of compounds. For the pyrazolotriazine derivatives, the SAR analysis confirmed the important role played by the apolar interactions, involving the extended hydrophobic portions of the inhibitors with hydrophobic region I (alkyl linker), with adenine region (pyrazolo-triazine ring system) and hydrophobic region II (cyclopropyl group). From the analysis of the active sites of different kinases it turned out that the one of CK2α is smaller in size, due to some bulky side chains, which reduce the space available to cofactors and inhibitors. The most important of these residues are Ile66 (maize) or Val66 (human) and Ile174, which in many protein kinases are replaced with less bulky amino acids, namely alanine versus Ile/Val66, alanine, threonine or leucine versus Ile174. Inhibition data on maize CK2α mutants confirmed the importance of Ile66 and Ile174; for the single mutants Ile174Ala or Val66Ala and for the double mutant Ile174Ala/Val66Ala, the TBB IC50 increases from 0.50 to 1.74, 13.0 and 12.5 μM, respectively. The smaller size of the CK2α active site can also account for the unusually modest sensitivity to the large molecular size promiscuous protein kinase inhibitor staurosporine. Very recently (Investigational New Drug Application (IND) submitted on october 2008), Cylene Pharmaceuticals announced that it has initiated a Phase I clinical trial of an orally administered CK2 protein kinase inhibitor, CX-4945, in patients with advanced solid tumors, Castleman’s disease, or multiple myeloma. In preclinical studies, it was able to promote tumor regressions as a single agent, with broad-spectrum anti-proliferative activity against diverse cancer cell lines. References: 1. Pinna LA (2002) Protein kinase CK2: a challenge to canons. J Cell Sci 115: 3873–3878. 2. Sarno S, Pinna LA (2008) Protein kinase CK2 as a druggable target. Mol Biosyst 4: 889–894. 3. Mazzorana M, Pinna LA, Battistutta R (2008) A structural insight into CK2 inhibition. Mol Cell Biochem 316: 57–62. 4. Battistutta R, Mazzorana M, Sarno S, Kazimierczuk Z, Zanotti G, Pinna LA (2005) Inspecting the structure-activity relationship of protein kinase CK2 inhibitors derived from tetrabromo-benzimidazole. Chem Biol 12: 1211–1219. 5. Battistutta R et al. (2007) The ATP-binding site of protein kinase CK2 holds a positive electrostatic area and conserved water molecules. Chembiochem 8: 1804–1809. 6. De Moliner E, Moro S, Sarno S, Zagotto G, Zanotti G, Pinna LA, Battistutta R (2003) Inhibition of protein kinase CK2 by anthraquinone-related compounds. A structural insight. J Biol Chem 278: 1831–1836. 7. Battistutta R, De Moliner E, Sarno S, Zanotti G, Pinna LA (2001) Structural features underlying selective inhibition of protein kinase CK2 by ATP site-directed tetrabromo-2-benzotriazole. Protein Sci 10: 2200–2206.

P2 Plant specific calcium sensor negatively regulates activity of SNF1-related protein kinases 2 Maria Bucholc, Arkadiusz Ciesielski, Grażyna Goch, Anna Anielska-Mazur, Anna Jaworska, Ewa Krzywińska and Grażyna Dobrowolska Institute of Biochemistry and Biophysics Polish Academy of Sciences, A. Pawińskiego St. 5a, 02-106 Warsaw, Poland e-mail: [email protected] SNF1-related protein kinases 2 (SnRK2s) are plant specific enzymes involved in regulation of plant response to environmental stress and abscisic acid-dependent plant development. In Arabidopsis thaliana, as well as in Oryza sativa, there are ten members of the SnRK2 family. It was shown that all of them, except SnRK2.9 from Arabidopsis, are rapidly activated by treatment with different osmolytes, and some of them also by abscisic acid (ABA), suggesting that these kinases are involved in a general response to osmotic stress [1–3]. However, the information concerning mechanism(s) of regulation of their activity is still limited. Results presented by several groups provide proof that phosphorylation in the kinase activation loop is required for their activation [4, 5]. Here, we describe identification of a plant specific calcium sensor, which interacts with the SnRK2 family members and can act as a negative regulator of their activity in plant cells. We screened a Nicotiana plumbaginifolia Matchmaker cDNA library for proteins interacting with Nicotiana tabacum osmotic stress-activated protein kinase (NtOSAK), a member of the SnRK2 family. A putative EF-hand calcium-binding protein was identified as a molecular partner of NtOSAK. The calcium-binding properties of the protein expressed in a bacterial system were characterized. Luminescence spectroscopy using Tb3+ as a spectroscopic probe confirmed that the protein binds calcium. The calcium binding constant of the protein, determined by fluorescence titration of the only Trp protein residue, is K = 2.5 ± 0.9 × 105 M−1. The CD spectrum indicated that the secondary structure of the protein changes significantly in presence of calcium, suggesting its possible function as a calcium sensor in plant cells. To determine whether the identified protein interacts only with NtOSAK or also with other SnRK2s, we cloned cDNA encoding the calcium binding protein orthologue from Arabidopsis thaliana and analyzed its binding with selected Arabidopsis SnRK2s using the yeast two-hybrid system. All studied kinases interacted with the protein. Therefore the protein was named SnRK2 interacting calcium sensor (SCaS). The interactions were confirmed by the Bimolecular Complementation Fluorescence assay, indicating that the binding occurs in planta, exclusively in cytoplasm. In vitro studies revealed that activity of analyzed SnRK2 kinases is inhibited by SCaS in a calcium-dependent manner. The results suggest that SCaS is a negative regulator of SnRK2s activity in response to calcium influx in plant cells. References: 1. Boudsocq M, Barbier-Brygoo H, Lauriere C (2004) Identification of nine SNF1-related protein kinase 2 activated by hyperos-

Abstracts 40 motic and saline stresses in Arabidopsis thaliana. J Biol Chem 279: 41758–41766. 2. KobayashiY, Yamamoto S, Minami H, Kagaya Y, Hattori T (2004) Differential activation of the rice sucrose nonfermenting 1-related protein kinase 2 family by hyperosmotic stress and abscisic acid. Plant Cell 16: 1163–1177. 3. Boudsocq M, Lauriere C (2005) Osmotic signaling in plants. Multiple pathways mediated by emerging kinase families. Plant Physiol 138: 1185–1194. 4. Belin C, de Franco P-O, Bourbousse C, Chaignepain S, Schmitter J-M, Vavasseur A, Giraudat J, Barbier-Brygoo H, Thomine S (2006) Identification of features regulating OST1 kinase activity and OST1 function in guard cells. Plant Physiol 141: 1316–1327. 5. Burza AM, Pękala I, Sikora J, Siedlecki P, Małagocki P, Bucholc M, Koper L, Zielenkiewicz P, Dadlez M, Dobrowolska G (2006) Nicotiana tabacum osmotic stress-activated kinase is regulated by phosphorylation on Ser-154 and Ser-158 in the kinase activation loop. J Biol Chem 281: 34299–34311.

2009

P3 Structural roadmap of cAMP dependent kinase: insights into the mechanism of inhibition and isoform-specific activation by cAMP analogues Cecilia Cheng1, Shelley Phoun1, Simon Brown1 and Susan Taylor1,2,3 1Department

of Chemistry and Biochemistry, and 2Howard Hughes Medical Institute, and 3Department of Pharmacology, University of California, San Diego, USA e-mail: [email protected] Cyclic adenosine monophosphate (cAMP) signaling through cAMP-dependent protein kinase (PKA) is a ubiquitous mammalian signaling pathway involved in metabolism [1], memory [2], and cell growth [3]. While the PKA catalytic (C) subunit has served as a prototype for the protein kinase superfamily, the regulatory (R) subunit defines the mechanism whereby the second messenger, cAMP, translates an extracellular signal into an intracellular biological response. Misregulation of this process is associated with a number of diseases including cancer [4], dilated cardiomyopathy [5], and systemic lupus erythematosus [6, 7]. The goal of this project is to understand the molecular features that govern cAMP-induced activation of PKA in order to develop therapeutic agents that combat disease. Three approaches have been used to achieve this goal: 1) to solve the crystal structure of the R:C heterodimer complex [8]; 2) to elucidate the molecular rules that govern substrate recognition; and 3) to understand the molecular basis for isoform-specific activation of PKA by cAMP analogs. The overall structure of the RIα:C complex consists of an extensive 2300 Å2 contact surface between the C- and R-subunits. The C-subunit adopts a closed conformation with its active site bound to AMP-PNP, two Mn2+ ions, and the pseudosubstrate site of the R-subunit. Surprisingly, the R-subunit undergoes major conformational changes upon binding to the C-subunit. In the cAMP bound conformation, the two cAMP binding domains form a compact globular structure, joined by the kinked αB/C helix. Upon binding to the C-subunit, the R-subunit adopts an extended dumbbell shape due to an extension of the αB/C helix, resulting in a 60 Å movement of domain A. The R:C structure outlined how the different components of the R-subunits bind and interact with the C-subunit. We also took a reductionist approach and assessed whether the inhibitor sequences alone could bind the C-subunits with the endeavor of generating PKA-specific peptide inhibitors. Peptide array analysis was initiated to determine whether these short sequences are sufficient to bind the catalytic subunit with high affinity. There are four isoforms of R-subunits (RIα, RIβ, RIIα, RIIβ) that share the same structural domain organization, but differ in biological function, localization, biochemical properties, and sequence. Surprisingly, only peptides correspondng to RII isoforms demonstrated detectable binding in the presence and absence of ATP and Mg2+. The shortest peptide corresponds to a 13-mer that spans 6 residues before and after the P-site, or the position that is phosphorylated in substrates.


Finally, we aimed to define the structural determinants of cAMP analogs to target specific PKA isoforms. A library of 21 cAMP analogues was screened for isoform specific PKA activation (RIα and RIIβ) using a fluorescence polarization assay designed to measure dissociated Csubunits. Our analysis identified cAMP analogues with substituents placed at the C8 position showed preferential activation of RIα holoenzymes. Second, substitutions placed at the N6 position showed preferential activation of RIIβ holoenzymes. The structures of RIα and RIIβ bound to cAMP show significant differences in the cAMP binding sites. In domain A, RIIβ has a large pocket near the N6 position of cAMP that is absent in RIα. We solved the structure of both RIα and RIIβ bound to HE33, the most RII selective analog. The structure of RIIβ bound to HE33 shows that the space near the N6 position is now occupied by the N6 alkyl substituent, surrounded in a hydrophobic environment. Conversely, RIα lacks this hydrophobic shell and binding of HE33 results in a more open pocket. These structural differences may explain the selectivity of N6 analogues for RIIβ. It is hoped that these studies will address how variations between R-subunit isoforms give rise to the unique and sophisticated mechanisms of PKA regulation in cells and provide a platform for designing isoform-specific inhibitors to combat disease. References: 1. Krebs EG, Beavo JA (1979) Phosphorylation-dephosphorylation of enzymes. Annu Rev Biochem 48: 923–59. 2. Arnsten AF, Ramos BP, Birnbaum SG, Taylor JR (2005) Protein kinase A as a therapeutic target for memory disorders: rationale and challenges. Trends Mol Med 11: 121–128. 3. Chen T, Hinton DR, Zidovetzki R, Hofman FM (1998) Up-regulation of the cAMP/PKA pathway inhibits proliferation, induces differentiation, and leads to apoptosis in malignant gliomas. Lab Invest 78: 165–174. 4. Taimi M, Breitman TR, Takahashi N (2001) Cyclic AMP-dependent protein kinase isoenzymes in human myeloid leukemia (HL60) and breast tumor (MCF-7) cells. Arch Biochem Biophys 392: 137–144. 5. Antos CL, Frey N, Marx SO, Reiken S, Gaburjakova M, Richardson JA, Marks AR, Olson EN (2001) Dilated cardiomyopathy and sudden death resulting from constitutive activation of protein kinase A. Circ Res 89: 997–1004. 6. Kammer GM, Laxminarayana D, Khan IU (2004) Mechanisms of deficient type I protein kinase A activity in lupus T lymphocytes. Int Rev Immunol 23: 225–244. 7. Kammer GM, Khan IU, Malemud CJ (1994) Deficient type I protein kinase A isozyme activity in systemic lupus erythematosus T lymphocytes. J Clin Invest 94: 422–430. 8. Kim C, Cheng CY, Saldanha SA, Taylor SS (2007) PKA-I holoenzyme structure reveals a mechanism for cAMP-dependent activation. Cell 130: 1032–1043.

P4 Structural effects of adenosine mimics on the potency of bisubstrate-analogue inhibitors of protein kinases

Erki Enkvist, Marie Kriisa and Asko Uri University of Tartu, Institute of Chemistry, 2 Jakobi St., 51014 Tartu, Estonia e-mail: [email protected] Bisubstrate-analogue inhibitors are compounds that simultaneously associate with both ATP- and protein-binding domains of protein kinases. The strategy of design of bisubstrate inhibitors could give selective and potent inhibitors of these dual substrate enzymes. Recently we have developed a new series of inhibitors that consist of adenosine-5’-carboxylic acid conjugated with an arginine-rich peptide via a long Ahx-(d-amino acid)-Ahx linker chain, where Ahx denotes 6-aminohexanoic acid [1]. The most potent inhibitors of this series that contained a hexa-(d-arginine) peptide revealed subnanomolar potencies towards several basophilic protein kinases. Even compounds with peptides containing only two d-arginine residues were potent inhibitors of these kinases (PKA, IC50 = 20–100 nM). Here we report on a new series of compounds incorporating different carboxylic acids instead of adenosine-5’-carboxylic acid in these conjugates. The variation included acetic acid, several derivatives of benzoic acid and different heterocyclic structures that were selected on the basis of the previous knowledge about the binding of the fragments to the adenosine pocket of kinases [2, 3]. The conjugate of the peptide with acetic acid revealed no inhibitory activity towards tested protein kinases (PKA, PKB and ROCK) whereas derivatives of benzoic acids were weak to moderate inhibitors (IC50 = 20–200 μM). This points to the requirement for an aromatic moiety that binds to the adenine binding site of the kinase leading to increased affinity of bisubstrate inhibitors towards protein kinases. Conjugation of the peptide part Ahx-(d-Lys)-Ahx-(dArg)2-NH2 with different heterocyclic moieties gave inhibitors with even higher potencies (IC50 = 10–10000 nM) than that of their adenosine counterpart. Conjugates of 5-(2-aminopyrimidin-4-yl)thiophene-2-carboxylic acid [2] with hexa-d-arginine inhibit basophilic protein kinases of the AGC group, PKA, PKB/Akt, PKC (classical and novel isoenzymes), PKG, MSK, ROCK, RSK, with high potency (more than 80% inhibition at 100 nM concentration, as established in Invitrogen’s panel towards 50 PK).

R

R=

HN

Peptide

HN N N

O

O

C

NH2

HN O

O O

HN

NH2

O

H N

H N

N H

O

N N N N

etc.

O N H

C

H N

O C

NH2

O HN

S

N

N

NH2

O NH HN

C

O

H2N

Scheme 1. Structures of the adenosine-mimicking fragments of the conjugates and that of the most active compound.

NH2

Abstracts 42

Compounds of this series have generally higher potency and more general inhibition profile than their adenosine counterparts [1]. References: 1. Lavogina D, Lust M, Viil I, Knig N, Raidaru G, Rogozina J, Enkvist E, Uri A, Bossemeyer D (2009) Structural analysis of ARC-type inhibitor (ARC-1034) binding to protein kinase A catalytic subunit and rational design of bisubstrate analogue inhibitors of basophilic protein kinases. J Med Chem 52: 308–321. 2. Lin X, Murray JM, Rico AC, Wang MX, Chu DT, Zhou Y, Del Rosario M, Kaufman S, Ma S, Fang E, Crawford K, Jefferson AB (2006) Discovery of 2-pyrimidyl-5-amidothiophenes as potent inhibitors for AKT: synthesis and SAR studies. Bioorg Med Chem Lett 16: 4163–168. 3. Sessions EH, Yin Y, Bannister TD, Weiser A, Griffin E, Pocas J, Cameron MD, Ruiz C, Lin L, Schürer SC, Schröter T, LoGrasso P, Feng Y (2008) Benzimidazole- and benzoxazole-based inhibitors of Rho kinase. Bioorg Med Chem Lett 18: 6390–6393.

2009

P5 Histidine phosphorylation affects thymidylate synthase properties Tomasz Frączyk1, Tomasz Ruman2, Joanna Cieśla1, Zbigniew Zieliński1, Elżbieta Wałajtys-Rode2 and Wojciech Rode1,2 1Nencki

Institute of Experimental Biology, Polish Academy of Sciences, Warszawa, Poland; 2Rzeszów University of Technology, Faculty of Chemistry, Rzeszów, Poland e-mail: [email protected] Thymidylate synthase (TS; EC 2.1.1.45), a target in chemotherapy [1], catalyzes the N5,10-methylenetetrahydrofolate (meTHF)-assisted C(5)-methylation of dUMP [2]. Possible phosphorylation of TS, previously reported [3], prompted us to examine this in more detail. TS preparations highly purified [4, 5] in the presence of phosphatase inhibitors, including endogenous TS forms from L1210 parental and FdUrd-resistant cells, and calf thymus, as well as mouse, rat, human and Trichinella spiralis recombinant TSs expressed in bacterial cells, as analyzed with following SDS/ PAGE, contained phosphorylated forms present in a low proportion, except for the calf thymus TS where their apparent content was distinctly higher (Fig. 1). However, MS analysis of the bands revealed no phosphorylated aminoacid residues. Furthermore, MS analysis of IEF fractions of TS preparations from parental and FdUrd-resistant mouse leukemia L1210 cells, whose differing sensitivity to inactivation by FdUMP and its analogues was previously found not due to mutations [4], demonstrated phosphorylation of Ser10 and Ser16 only in the resistant enzyme, although the Pro-Q® Diamond Phosphoprotein Gel Stain indicated also phosphorylation of parental TS.

Figure 1. Phosphorylation of calf thymus endogenous TS: (lanes marked 2 in gels A and B; lanes marked 1 contain MW standards), determined following PAGE under denaturing (SDS/ PAGE) conditions. Gel was stained first for phosphoprotein (ProQ® Diamond Phosphoprotein Gel Stain; A) and later for protein (SYPRO® Ruby Protein Gel Stain; B).

Enrichment of phosphorylated fractions of each of the four recombinant TS preparations using metal oxide/hydroxide affinity chromatography on Al(OH)3 beads [6], yielding always ≈ 1% of the total protein, allowed to demonstrate that TS phosphorylation is responsible for a 3–4fold lower Vmaxapp, with unaltered Kmapp for either substrate or cofactor, and ability to repress in vitro translation of TS cognate, as well as luciferase, mRNA. Surprisingly, while MS analyses did not reveal the presence of phosphorylated residues in any of the fractions investigated,

6th International Conference: Inhibitors of Protein Kinases Vol. 56 43 31P NMR spectroscopy

demonstrated clearly the presence of phosphorylated residues only in the phosphorylated enzyme fractions (Fig. 2). Further analyses of the 31P NMR spectra (including their time-dependent changes following acidification), and comparison with those of synthetic phosphoramidate derivatives of basic amino acids (Lys, Arg and His), and commercially available phospho-amino acids, revealed the presence of phosphorus in a phosphoramidate (acid-labile) bond, pointing to modification of histidine residue(s). As phosphoramidates escape routine MS analysis, the latter may suggest similar modifications in L1210 and calf endogenous TSs. Results of an MS analysis of peptides enriched from the recombinant mouse TS preparation trypsin digest using TiO2 beads (Phos-Trap, Perkin Elmer), the enrichment resulting presumably from phosphohistidine binding by the beads, pointed to His298 being the most probable phosphorylation site. Which protein kinases are responsible for the phosphorylation of TS, remains to be established.

Figure 2. 31P NMR spectrum of the enriched phosphorylated fraction of human recombinant TS with marked positions of the resonances of phosphorylated standards. The insert presents the corresponding spectrum of the non-phosphorylated TS fraction. Acknowledgements: Supported by the Ministry of Science and Higher Education (grant number N401 2334 34). References: 1. Lehman NL (2002) Future potential of thymidylate synthase inhibitors in cancer therapy. Expert Opin Investig Drugs 11: 1775– 1787. 2. Carreras CW, Santi DV (1995) The catalytic mechanism and structure of thymidylate synthase. Annu Rev Biochem 64: 721– 762. 3. Samsonoff WA, Reston J, McKee M, O’Connor B, Galivan J, Maley GF, Maley F (1997) Intracellular location of thymidylate synthase and its state of phosphorylation. J Biol Chem 272: 13281– 13285. 4. Cieśla J, Frączyk T, Zieliński Z, Sikora J, Rode W (2006) Altered mouse leukemia L1210 thymidylate synthase, associated with cell resistance to 5-fluoro-dUrd, is not mutated but rather reflects posttranslational modification. Acta Biochim Polon 53: 189–198. 5. Cieśla J, Gołos B, Wałajtys-Rode E, Jagielska E, Płucienniczak A, Rode W (2002) The effect of Arg 209 to Lys mutation in mouse thymidylate synthase. Acta Biochim Polon 49: 651–658. 6. Wolschin F, Wienkoop S, Weckwerth W (2005) Enrichment of phosphorylated proteins and peptides from complex mixtures using metal oxide/hydroxide affinity chromatography (MOAC). Proteomics 5: 4389–4397.

P6 l(+)-Tartrate and metavanadate are nonlinear competitive inhibitors of cooperative human prostatic phosphatase Magdalena Górny, Natalia Hutyra and Ewa Luchter-Wasylewska Department of Medical Biochemistry, Jagiellonian University, Collegium Medicum, Kopernika 7, 31-034 Kraków, Poland e-mail: [email protected] Human prostatic acid phosphatase (PAP) [EC 3.1.3.2], secreted by the prostate gland into the seminal fluid in great amounts, nonspecifically catalyzes hydrolysis of many phosphoesters, including phosphoproteins, on P-ser, Pthr and P-tyr residues [1, 2]. Tyrosine phosphorylation of c-ErbB-2, involved in regulating the androgen-responsive phenotype of prostate cancer cells, is regulated by PAP, which is therefore also a protein tyrosine phosphatase [2]. Furthermore PAP dephosphorylates semenogelins [3] and specifically proteolyses semenogelins [4] in seminal fluid. Recently it was found that extracellular 5’-AMP is a physiological substrate of PAP [5]. 5’-AMP is dephosphorylated by a 5’-nucleotidase activity of PAP, generating adenosine and activating A1-adenosine receptors in dorsal spinal cord. Moreover intraspinal injection of PAP protein has potent antinociceptive, antihyperalgesic and antiallodynic effects that last longer than the opioid analgesic morphine. Deletion of A1-adenosine receptors eliminates all these biological effects of PAP [5]. A novel PAP-spliced variant mRNA encoding a transmembrane protein (TM-PAP), found in vesicles and membranes, was currently described by Quintero et al. [6]. TM-PAP is widely expressed in nonprostatic tissues like brain, kidney, liver, lung, muscle, placenta, salivary gland, spleen, thyroid, thymus and in fibroblast LNCaP cells, but not in PC-3 prostate cancer cells. In well-differentiated human prostate cancer tissue specimens, the expression of secretory PAP, but not of TM-PAP, is decreased significantly. In previous detailed steady-state studies on phosphoesters’ hydrolysis, we reported that PAP belongs to the regulatory, allosteric enzymes: PAP exhibits positive cooperativity in substrate binding [7, 8]. Substrate saturation curves, described by the Hill rate equation*, are sigmoidal: thus the substrates are homotropic positive effectors (homotropic activators) of PAP. The extent of cooperativity, expressed as the value of the Hill cooperation coefficient (h), grows when enzyme concentration is increased: from 1 at low enzyme concentration to about 4 at a higher one, suggesting that monomeric, dimeric and tetrameric species, respectively, predominate at different PAP concentrations. Degree of cooperativity additionally depends on the chemical nature of the substrate molecule: it increases with growing hydrophobicity, increasing polarizability and decreasing charge. Ligand-induced, concentrationdependent dissociation-association of catalytically active PAP oligomeric forms (monomer-dimer-tetramer) is thus suggested. It was concluded that the cooperativity exhibited by PAP, dependent on its quaternary structure, is described best by models of Frieden, Nichol and Kurganov.

Abstracts 44

Models of Monod, Wyman and Changeaux (MWC), as well as models of Koshland, Nemethy and Filmer (KNF) are thus not adequate for this purpose [7, 8]. In the present research on inhibition of the catalytic activity of allosteric PAP by l(+)-tartrate and metavanadate at several enzyme concentrations, it was found that both inhibitors are competitive but nonlinear. l(+)-tartrate inhibits at millimolar concentrations and metavanadate at nanomolar ones. When concentration of inhibitors is increased, the values of the half-saturation constant (K0,5) rise and of the turnover number (kcat) remains constant. l(+)-Tartrate diminishes the cooperative character of PAP: the values of the Hill cooperation coefficient (h) are decreased when l(+)-tartrate concentration is increased. By contrast, the values of the Hill cooperation coefficient (h) are not changed by metavanadate. Dixon and CornishBowden plots are mostly nonlinear for both inhibitors. Studies on inhibition of the catalytic activity of allosteric human prostatic acid phosphatase (PAP) may describe equilibria between catalytically active enzyme oligomeric forms (monomer-dimer-tetramer) as well as molecules of substrates (homotropic activators) and inhibitors. *Hill rate equation: where: vo is the initial reaction rate, Vmax – the maximal reaction

Vmax([S]0 ) h kcat[E]([S]0 ) h vo = (K ) h + ([S] ) h = (K ) h + ([S] ) h 0.5 0 0.5 0 rate, [E] – the concentration of enzyme, [S]o – the initial concentration of substrate, h – the Hill cooperation coefficient, K0.5 – the half saturation constant, kcat – the catalytic constant (turnover number). References: 1. Wasylewska E, Czubak J, Ostrowski WS (1983) Phosphoprotein phosphatase activity of human prostate acid phosphatase. Acta Biochim Polon 30: 175–184. 2. Meng TC, Lin MF (1998) Tyrosine phosphorylation of c-ErbB-2 is regulated by the cellular form of prostatic acid phosphatase in human prostate cancer cells. J Biol Chem 273: 22096–22104. 3. Ek P, Malm J, Lilija H, Carlsson L, Ronquist G (2002) Exogenous protein kinases A and C, but not endogenous prostasomeassociated protein kinase, phosphorylate semenogelins I and II from human semen. J Androl 23: 806–814. 4. Brillard-Bourdet M, Rehault S, Juliano L, Ferrer M, Moreau T, Gauthier F (2002) Amidolytic acitivity of prostatic acid phosphatase on human semenogelins and semenogelin-derived synthetic substrates. Eur J Biochem 269: 390–395. 5. Zylka MJ, Sowa NA, Taylor-Blake B, Twomey MA, Herrala A, Voikar V, Vihko P (2008) Prostatic acid phosphatase is an ectonucleotidase and suppresses pain by generating adenosine. Neuron 60: 111–122. 6. Quintero IB, Araujo CL, Pulkka AE, Wirkkala RS, Herrala AM, Eskelinen EL, Jokitalo E, Hellstrőm PA, Tuominen HJ, Hirvikoski PP, Vihko PT (2007) Cancer Res 67: 6549–6554. 7. Luchter-Wasylewska E (2001) Cooperative kinetics of human prostatic acid phosphatase. Biochim Biophys Acta 1548: 257–264. 8. Luchter-Wasylewska E, Wasylewski M, Rőhm KH (2003) Concentration-dependent dissociation/association of human prostatic acid phosphatase. J Protein Chem 22: 243–247.

2009

P7 Linker engineering of SF2/ASF splicing factor switches enzymatic activities of human topoisomerase I

Takao Ishikawa, Alicja Czubaty, Krzysztof Staroń Institute of Biochemistry, Faculty of Biology, University of Warsaw ul. Miecznikowa 1, 02-096 Warszawa, Poland e-mail: [email protected] Alternative splicing is a cellular process that enrich the proteome diversity. It is controlled, in part, by two antagonistically working splicing factors, SF2/ASF and hnRNP A1 [1]. SF2/ASF belongs to the SR proteins that are obligatorily equipped with an arginine-serine-rich (RS) domain and one or two RRM (RNA recognition motif) domains. Moreover, both RRM domains are, in case of SF2/ASF, connected by a peptide linker that consists of nine glycine residues (i.e. glycine tract). The other splicing factor, hnRNP A1, is also build from two RRM domains, however, connected by short and rigid linker [2]. The SF2/ASF protein needs to be phosphorylated in the RS domain to function as a splicing factor. Phosphorylation is carried out by SRPK1, Clk/Sty, PRP4 and topoisomerase I (topoI) [3]. The last kinase is particularly interesting because of two mutually exclusive enzymatic activities: DNA nicking which results in its relaxation, and phosphorylation of the SR proteins in the presence of ATP. Interestingly, DNA (a substrate for nicking activity) is known to inhibit the kinase reaction of topoI, whereas ATP and the SF2/ASF protein (substrates for kinase activity) are inhibitors of DNA nicking activity of the enzyme [4]. In our previous work, we have found that both SF2/ASF and hnRNP A1 compete for topoI as each of them binds to the same site in the cap region (residues 215–433) of topoI [5, 6]. However, in contrast to SF2/ASF, hnRNP A1 does not influence DNA nicking activity. Because of their opposed effects on the kinase and relaxation activities, we concluded that SF2/ASF and hnRNP A1 regulate switching of enzymatic activities of topoI. The inhibitory effect on topoI DNA cleavage is linked with the region of SF2/ASF containing both RRM domains [5]. The most pronounced structural dissimilarity between SF2/ASF and hnRNP A1 is the linker located between RRM domains. The former is equipped with flexible glycine tract, while the latter has comparatively short and rigid linker that seems to prevent from unrestricted movements of RRM domains. To find out the role of linkers in switching of topoI activity, we have constructed several recombinant proteins with different linkers between RRM domains. First, we swapped linkers of SF2/ASF and hnRNP A1 to obtain SF2/ASFUP1 and UP1SF2/ASF (UP1 stands for the shortened hnRNP A1 protein commonly used in the in vitro studies) that have native linkers of hnRNP A1 and SF2/ASF, respectively. We found that, unlike the native UP1 protein, the UP1SF2/ASF was not able to fully promote the DNA nicking activity of topoI, which continued to phosphorylate SF2/ASF. On the other hand, using SF2/ASFUP1 protein, we have confirmed that the linker region has no im-


pact either on the interaction of SF2/ASF with topoI, nor on its phosphorylation efficiency. However, substitution of the long linker for a short one in the SF2/ASF protein partly abolished the inhibitory effect of the protein on the nicking activity. Further, we constructed two recombinant SF2/ASF proteins with different characteristics of linker regions: SF2/ASFVal (glycine tract changed to nine valine residues) and SF2/ASFΔL (entirely removed linker). The former is supposed to ensure the rigidness of the linker without changing the distance between RRM domains, while the latter could indicate whether the linker of SF2/ ASF is solely responsible for its effects on topoI or other fragments of the splicing factor additionally contribute to the regulation of enzymatic activities of topoI. We discuss the role of the linker region of the SF2/ASF protein on switching of the enzymatic activities of topo and suggest that the linker is predominantly responsible for inhibition of the relaxation activity, although it far less influences the kinase activity of topo I. References: 1. Dreyfuss G, Kim VN, Kataoka N (2002) Messenger-RNA-binding proteins and the messages they carry. Nat Rev Mol Cell Biol 3: 195–205. 2. Hanamura A, Cáceres JF, Mayeda A, Franza BR Jr, Krainer AR (1998) Regulated tissue-specific expression of antagonistic premRNA splicing factors. RNA 4: 430–444. 3. Kowalska-Loth B, Girstun A, Piekiełko A, Staroń K (2002) ASF/ SF2 protein inhibits camptothecin-induced DNA cleavage by human topoisomerase I. Eur J Biochem 269: 3504–3510. 4. Kowalska-Loth B, Girstun A, Trzcińska AM, PiekiełkoWitkowska A, Staroń K (2005) ASF/SF2 protein binds to the cap region of human topoisomerase I through two RRM domains. Biochem Biophys Res Commun 331: 398–403. 5. Rossi F, Labourier E, Forne T, Divita G, Derancourt J, Riou JF, Antoine E, Cathala G, Brunel C, Tazi J (1996) Specific phosphorylation of SR proteins by mammalian DNA topoisomerase I. Nature 381: 80–82. 6. Trzcińska-Daneluti AM, Górecki A, Czubaty A, KowalskaLoth B, Girstun A, Murawska M, Lesyng B, Staroń K (2007) RRM proteins interacting with the cap region of topoisomerase I. J Mol Biol 369: 1098–1112.

P8 Phosphorylation near nuclear targeting signals regulates nuclear import and export of viral proteins

David A. Jans2, Alex. J. Fulcher1, Daniela M. Roth1, Shadma Fatima2, Dominic J. Glover1 and Gualtiero Alvisi3 1Nuclear

Signalling Laboratory, Department of Biochemistry and Molecular Biology, Monash University, Clayton, Australia; 2ARC Centre of Excellence for Biotechnology and Development; 3Department of Molecular Virology, University of Heidelberg, Heidelberg, Germany e-mail: [email protected] Nucleocytoplasmic trafficking of transcription factors and other signalling molecules is central to eukaryotic cell processes such as differentiation, signal transduction, and transformation, with phosphorylation a common means of regulating the process [1]. Our work [2–8] and that of others is consistent with the idea that phosphorylation also regulates the nucleocytoplasmic trafficking of viral proteins in infected cells, with strong relevance to pathogenicity. Nuclear transport is dependent on nuclear targeting signals (nuclear localisation sequences (NLSs) and nuclear export sequences (NESs) in the nuclear import and export directions respectively), and the cellular transporters that recognise them, the members of the importin/exportin superfamily. Using quantitative confocal laser scanning microscopy (CLSM) in living transfected cells, in vitro reconstituted systems, or immunostained virus-infected cells, as well as in vitro binding assays, we have characterised the nuclear transport pathways of diverse gene products from DNA tumor viruses such as simian virus 40 (SV40) and human cytomegalovirus (HCMV), as well as the ssDNA circovirus chicken anemia virus (CAV), and RNA viruses respiratory syncytial virus, rhinovirus, and Dengue virus [2–8]. A common theme appears to be that phosphorylation close to NLS/NES sequences by cellular kinases plays a key role in modulating recognition by importins/ exportins [1, 2]. In the case of SV40 large tumor antigen (T-ag) and HCMV phosphoprotein ppUL44, the processivity factor for the HCMV DNA polymerase, we have been able to show that phosphorylation by protein kinase CK2 upstream of the importin α/β1-recognised NLS is critical to enhance nuclear import efficiency by increasing the affinity of the importin-NLS interaction [2–4]. Further, specific inhibitors of CK2 activity inhibit nuclear accumulation of both T-ag and ppUL44. Importantly, phosphorylation at other sites regulates nuclear import negatively; in particular, phosphorylation at the cyclin dependent kinase (cdk) or protein kinase C (PKC) sites adjacent to the T-ag and ppUL44 NLSs, respectively, inhibits NLS-dependent nuclear import. We have recently established the mechanism of this inhibition, showing that phosphorylation confers interaction with the novel negative regulator of nuclear import (NRNI) BRAP2, originally isolated as a protein interacting with

Abstracts 46

the breast cancer antigen BRCA1. Ectopic expression of BRAP2 significantly reduces NLS-dependent nuclear accumulation of T-ag and ppUL44, but not of viral proteins that lack a phosphorylation site near their NLSs, such as herpes simplex virus type 1 pUL30 or human immunodeficiency virus Tat. BRAP2 inhibition of nuclear accumulation is specifically dependent on phosphorylation sites flanking the respective NLSs, since substitution of the phosphorylation-site threonines of either T-ag or ppUL44 with non-phosphorylatable or phosphomimetic amino acids prevents or enhances BRAP2 inhibition of nuclear import, respectively. Pulldowns/direct binding assays indicate high affinity binding of BRAP2 to T-ag, strictly dependent on negative charge near the NLS. All results are consistent with BRAP2 being a novel, phosphorylation-regulated NRNI. In the case of CAV VP3, phosphorylation near the C-terminal NES, which appears to occur predominantly in cancer/tumorigenic cells as opposed to isogenic normal/ non-tumorigenic cells, blocks nuclear export; the result is that VP3 accumulates to a significantly higher extent in tumour than in normal cells. We are keen to exploit this cancer cell-specific regulation of nuclear export in drug targeting and gene therapy approaches, to kill tumour cells specifically in clinically relevant settings. An important question is the role in the nucleus of VP3 in CAV infection, with an intriguing aspect in this context being the activity of CAV VP2, which appears to be a dual specificity phosphatase (DSP) that plays a critical role in viral replication and virulence. VP2 has both protein-tyrosine phosphatase (PTPase) and serine-threonine phosphatase (S/T PPase) activity, which appears to regulate the cellular localization of VP3 in infected cells. Our ongoing work suggests that while VP2 action decreases VP3 nuclear localisation, it may not act directly on the key phosphorylation site threonine near the VP3 NES, but rather may modulate phosphorylation at the site by dephosphorylating cellular signalling molecules. Intriguingly, recently discovered CAV-related human anelloviruses encode both a DSP that may have a similar function to CAV VP2, and a cancer cell-specific localizing protein comparable to CAV VP3, implying potential broad medical significance. Since phosphorylation-regulated switching between nuclear import and export of viral proteins appears to be a common regulatory mechanism utilised by diverse viruses, targeting the nuclear import/export pathways is a viable approach to inhibit virus production. In this context, further development of specific CK2 inhibitors appears to be an efficacious anti-viral strategy, whilst the NRNI BRAP2 has potential as an anti-viral agent. References: 1. Jans DA, Lam MHC, Xiao C-Y (2000) Nuclear targeting signal recognition: central control point of nuclear protein transport ? BioEssays 22: 532–544. 2. Alvisi G, Ghildyal R, Rawlinson S, Jans DA (2007) Regulated nucleocytoplasmic trafficking of viral gene products: a therapeutic target? Biochim Biophys Acta Proteins Proteomics 1784: 213–227. 3. Xiao C-Y, Hübner S, Jans DA (1997) SV40 large tumor-antigen nuclear import is regulated by the double-stranded DNA-dependent protein kinase site (serine 120) flanking to the nuclear localization sequence. J Biol Chem 272: 22191–22198.

2009

4. Alvisi G, Jans DA, Guo J, Pinna LA, Ripalti A (2005) A protein kinase CK2 site flanking the nuclear targeting signal enhances nuclear transport of human CMV ppUL44. Traffic 6: 1002–1013. 5. Jans DA, Ackermann M, Bischoff JR, Beach DH, Peters R (1991) cdc2 124 p34 -mediated phosphorylation at T inhibits nuclear import of SV40 T-antigen proteins. J Cell Biol 115: 1203–1212. 6. Poon IKH, Oro C, Dias MM, Jingpu Z, Jans DA (2005) Apoptin nuclear accumulation is modulated by a Crm1-recognised nuclear export signal that is active in normal but not tumor cells. Cancer Res 65: 7059–7064. 7. Rawlinson SM, Pryor MJ, Wright PJ, Jans DA (2009) CRM1mediated nuclear export of dengue virus NS5 RNA polymerase regulates the kinetics of virus production. J Biol Chem in press. 8. Ghildyal R, Ho A, Dias M, Soegiyono L, Bardin PG, Tran KC, Teng M, Jans DA (2009) The respiratory syncytial virus matrix protein possesses a Crm1-mediated nuclear export mechanism. J Virol in press. 9. Ghildyal R, Jordan B, Li D, Bardin PG, Gern JE, Jans DA (2009) Rhinovirus 3C protease can localize in the nucleus and alter active and passive nucleocytoplasmic transport. J Virol in press.


P9 Antileishmanial activity of disubstituted purines and related pyrazolo[4,3-d]pyrimidines Radek Jorda1, Matthew W. Nowicki2, Charles L. Jaffe3, Libor Havlíček1, Vladimír Kryštof1, Miroslav Strnad1 and Malcolm D. Walkinshaw2 1Laboratory

of Growth Regulators, Faculty of Science, Palacký University, Šlechtitelů 11, 783 71 Olomouc, Czech Republic; 2Structural Biochemistry Group, Institute of Structural and Molecular Biology, University of Edinburgh, King’s Buildings, Mayfield Road, Edinburgh EH9 3JR, Scotland; 3Department of Parasitology, Hebrew University, Hadassah Medical School, Post Office Box 12272, Jerusalem 91120, Israel e-mail: [email protected] Trypanosomal and leishmanial cyclin-dependent related kinases (CRKs) are serine/threonine protein kinases which are important in regulation of cell cycle of protozoan parasites [1]. CRK1 and CRK3 are the most investigated kinases and play probably a major role in regulation and coordination of the life cycle of leishmanial species [2–4]. Mammalian cyclin-dependent kinases (CDKs) and leishmanial related kinases display high sequence similarity [3]. We report here screening results directed to find new antileishmanial drugs among disubstituted purines and structurally related disubstituted pyrazolo[4,3-d]pyrimidines that have been previously shown to moderately inhibit human CDKs [5]. Since some compounds blocked the proliferation of axenic amastigotes of Leishmania donovani, we assayed them for interactions with recombinant leishmanial kinase CRK3, an important regulator of the cell cycle of the parasitic protozoan leishmania, using the Thermofluor-based thermal shift assay and surface plasmon resonance. Some compounds from this screen showed promising results and could be used as lead structures for development of new potential antileishmanial drugs. Acknowledgements: The work was supported by grants GA CR 204/08/0511 and MSM 6198959216. References: 1. Naula C, Parsons M, Mottram JC (2005) Protein kinases as drug targets in trypanosomes and Leishmania. Biochim Biophys Acta 1754: 151–159. 2. Tu X, Wang CC (2004) The involvement of two cdc2-related kinases (CRKs) in Trypanosoma brucei cell cycle regulation and the distinctive stage-specific phenotypes caused by CRK3 depletion. J Biol Chem 279: 20519–20528. 3. Mottram JC, Kinnaird JH, Shiels BR, Tait A, Barry JD (1993) A novel CDC2-related protein kinase from Leishmania mexicana, LmmCRK1, is post-translationally regulated during the life cycle. J Biol Chem 268: 21044–21052. 4. Wang Y, Dimitrov K, Garrity LK, Sazer S, Beverley SM (1998) Stage-specific activity of the Leishmania major CRK3 kinase and functional rescue of a Schizosaccharomyces pombe cdc2 mutant. Mol Biochem Parasitol 96: 139–150.

5. Moravcová D, Krystof V, Havlícek L, Moravec J, Lenobel R, Strnad M (2003) Pyrazolo[4,3-d]pyrimidines as new generation of cyclin-dependent kinase inhibitors. Bioorg Med Chem Lett 13: 2989–2992.

Abstracts 48

P10 A tool for homology modeling of kinase targets for drug design Sebastian Kmiecik1,2, Michal Jamroz2 and Andrzej Kolinski2 1Selvita,

Ostatnia 1c, 31-444 Kraków, Poland; 2Faculty of Chemistry, University of Warsaw, L. Pasteura 1, 02-093, Warsaw, Poland e-mail: [email protected] Kinases continue to be hot targets for the pharmaceutical industry. Recently, kinase-targeted structural genomics efforts has significantly increased the number of novel protein structures [1]. This growth of structural data facilitates accurate comparative modeling. Many applications of kinase homology models have been lately described e.g.: binding mode prediction, lead potency and selectivity optimization, virtual screening [2–4]. It was found that in some cases it is even better to use the homology model for docking than the crystal structure of the actual kinase target (when homology model is created from the different kinase bound to related ligand) [2]. Here we introduce Selvita Protein Modeling Platform, an easy to use, web-based tool for accurate homology modeling. The Platform consists of several protocols and tools, where the most beneficial in kinase modeling is homology modeling and ab initio loop modeling. The homology modeling method is driven by CABS ― unique technology which uses spatial restraints derived from a template or many templates in a single modeling run [5–7]. The loop modeling protocol enables easy modeling of insertions in the template sequence. CABS technology allows for experimental-level accuracy of ab initio predictions where the length of the loops could be in the range of 20 residues, which is well beyond the capabilities of the competitive software. There are still a large number of kinases targets of unknown structure that share very low sequence identity with kinases of known structures. In these cases, the Selvita Platform offers a number of strategies like threading or flexible 3D threading for correct fold recognition. The predicted results can be very useful in guiding experimental studies of new targets. References: 1. Marsden B, Knapp S (2008) Doing more than just the structure ― structural genomics in kinase drug discovery. Curr Opin Chem Biol 12: 40–45. 2. Rockey WM, Elcock AH (2006) Structure selection for protein kinase docking and virtual screening: homology models or crystal structures? Curr Protein Pept Sci 7: 437–457. 3. Kairys V, Fernandes MX, Gilson MK (2006) Screening druglike compounds by docking to homology models: a systematic study. J Chem Inf Model 46: 365–379. 4. Muegge I, Enyedy IJ (2004) Virtual screening for kinase targets. Curr Med Chem 11: 693–707. 5. Kolinski A, Bujnicki JM (2005) Generalized protein structure prediction based on combination of fold-recognition with de novo folding and evaluation of models. Proteins 61: 84–90.

2009

6. Kmiecik S, Gront D, Kolinski A (2007) Towards the high-resolution protein structure prediction. Fast refinement of reduced models with all-atom force field. BMC Struct Biol 7: 43. 7. Kmiecik S, Jamroz M, Zwolinska A, Gniewek P, Kolinski A (2008) Designing an automatic pipeline for protein structure prediction. In From computational biophysics to systems biology. Hansmann UHE, Meinke JH, Mohanty S, Nadler W, Zimmermann O, eds. 40: 105–108.


P11 Soluble vascular endothelial growth factor receptor 1 concentration in serum and exudative pleural effusion in breast cancer Ewa Kopczyńska1, Ewelina Bednarczuk1, Maciej Dancewicz2, Janusz Kowalewski2, Agnieszka Kaczmarczyk3, Hanna Kardymowicz3 and Tomasz Tyrakowski1 1Department

of Pathobiochemistry and Clinical Chemistry, and 2Department of Thoracic Surgery and Tumors, Nicolaus Copernicus University in Torun, Collegium Medicum in Bydgoszcz, Poland; 3Department of Laboratory Diagnostics, Oncology Centre in Bydgoszcz, Bydgoszcz, Poland e-mail: [email protected] Soluble vascular endothelial growth factor receptor 1 (sVEGFR1) is a naturally occurring, alternatively spliced form of receptor tyrosine kinase VEGFR1, capable of sequestering ligand (VEGF) or dimerizing with full-length membrane bound receptor (VEGFR2) and preventing signal transduction. sVEGFR1 binds VEGF with high affinity and is able to inhibit VEGF-induced mitogenesis, suggesting that it is a physiological negative regulator of VEGF action [1–3]. sVEGFR1 has been characterized as one of the most important endothelial regulators in tumor angiogenesis. Recombinant sVEGFR1 was found to bind all isoforms of VEGF and to inhibit VEGF-induced endothelial cell proliferation [4–6]. VEGF increases vascular permeability, plays a critical role in the production of malignant pleural effusions and shows high level in this fluid [7]. In the present study, we examined the concentration of sVEGFR1 (negative regulator of VEGF) in serum and exudative pleural effusions in patients with breast cancer. Nineteen patients with exudative pleural effusions due to breast cancer were included in this study. The control group consisted of 16 healthy volunteers. sVEGFR1 concentrations in serum and exudative pleural effusions (EPEs) were measured by an enzyme-linked immunosorbent assay (ELISA). Serum sVEGFR1 concentration was higher in breast cancer than controls (median: 172.0 vs. 117.5 pg/ml; minimum–maximum: 75.0–317.4 vs. 87.1–196.2; P < 0.01). sVEGFR1 levels in EPEs was higher than in serum in patients with breast cancer (median of concentration in EPEs: 511.7 pg/ml; minimum–maximum: 111.4–3024.8; P < 0.001). In two groups of patients: with cancer cells in EPEs and without of them in EPEs, both sVEGFR1 concentration in pleural effusions (median: 548.2 vs. 372.1 pg/ml) and EPEs sVEGFR1/serum sVEGFR1 ratio (4.44 vs. 1.86) did not differ significantly. Rhe higher level of sVEGFR1 in serum of patients with breast cancer than controls and higher level in EPEs than in serum, may suggest that it is involved in tumor-associated disorders. Keywords: sVEGFR1, breast cancer References: 1. Kendall RL, Wang G, Thomas KA (1996) Identification of a natural soluble form of the vascular endothelial growth factor

receptor, FLT-1, and its heterodimerization with KDR. Biochem Biophys Res Commun 226: 324–328. 2. Kendall RL, Thomas KA (1993) Inhibition of vascular endothelial growth factor by an endogenously encoded soluble receptor. Proc Natl Acad Sci USA 90: 10705–10709. 3. Robinson CJ, Stringer SE (2001) The splice variants of vascular endothelial growth factor (VEGF) and their receptors. J Cell Sci 114: 853–865. 4. Toi M, Bando H, Ogawa T, Muta M, Hornig C, Weich HA (2002) Significance of vascular endothelial growth factor (VEGF)/soluble VEGF receptor-1 relationship in breast cancer. Int J Cancer 98: 14–18. 5. Malecki M, Trembacz H, Szaniawska B, Przybyszewska M, Janik P (2005) Vascular endothelial growth factor and soluble FLT-1 receptor interactions and biological implications. Oncol Rep 14: 1565–1569. 6. Takayama K, Ueno H, Nakanishi Y, Sakamoto T, Inoue K, Shimizu K, Oohashi H, Hara N (2000) Suppression of tumor angiogenesis and growth by gene transfer of soluble form of vascular endothelial growth factor receptor into a remote organ. Cancer Res 60: 2169–2177. 7. Tomimoto H, Yano S, Muguruma H, Kakiuchi S, Sone S (2007) Levels of soluble vascular endothelial growth factor receptor 1 are elevated in the exudative pleural effusions. J Med Invest 54: 146–153.

Abstracts 50

P12 Chemical inhibition of CDKs in primary and cancerous cells

Liliana Krasinska, Emilie Cot and Daniel Fisher Institut de Génétique Moléculaire de Montpellier, IGMM, CNRS-UMR 5535, 1919 Route de Mende, 34293 Montpellier cedex 5, France e-mail: [email protected] Although it was found over 20 years ago that cyclin-dependent kinases have essential roles in the cell cycle, their functions are still surprisingly poorly understood at a molecular level. One explanation for this knowledge gap is the functional redundancy of different Cdk complexes. In protozoans, Cdk1 controls passage through a „commitment” point in G1 of the cell cycle, as well as onset both of S-phase and of mitosis. However, metazoans encode one or more additional highly related Cdks, Cdk2 or Cdk3, and other Cdks of several families, whose respective roles are still not well understood [1]. In this work, we employ selective chemical inhibition as a tool to analyse individual Cdk function in vertebrates, using the high-affinity Cdk1/2 inhibitor NU6102 [2]. In Xenopus egg extracts we obtain conditions in which Cdk2 is inhibited but Cdk1 is not, allowing us to definitively demonstrate that Cdk2 is required for efficient firing of replication origins in an embryonic system, but in its absence Cdk1 can compensate [3]. We wished to extend this analysis to study the roles of Cdk2 in both human somatic primary and cancer cells. We find that NU6102 can also discriminate in vitro between human Cdk1 and Cdk2 kinases, although we demonstrate theoretically that Cdk4 is likely to be an in vivo target of NU6102 in spite of the significantly higher Ki of NU6102 for this kinase compared to Cdk1/2. In vivo, in the presence of NU6102, passage through DNA replication is slower in both primary and cancer cells, and entry into mitosis is delayed but not blocked. Passage through mitosis, however, is defective, with chromosome congression defects leading to inefficient cytokinesis, with the majority of cells refusing to form polyploid cells with multilobed nuclei [4]. In the next cell cycle, centrosome number is usually abnormal. Application of inhibitors leads to cell death specifically in cancer cells. These results suggest that mitosis is not an „all or nothing” event occurring once a threshold kinase activity is exceeded, but that correct completion of mitosis requires maintenance of Cdk activity to a sufficient level. Combining NU6102 with a selective Cdk1 inhibitor, RO-3306 [5], causes a much stronger phenotype than with NU6102 alone, suggesting that CDK1 is at least partially active in the presence of NU6102. However, as yet there is no general method allowing to discriminate between roles of individual Cdks in vivo without perturbing the system by knockout or knockdown approaches. We are therefore developing such an approach, using a novel general method for generating inhibitor-resistant kinases, which we apply to Cdk2. Application of NU6102 in the absence of presence of NU6102-resistant Cdk2 should allow us to determine phenotypes caused by inhibition of Cdk2 rather than of other kinases, and may be a useful

2009

tool for improving inhibitor-specificity and avoiding generation of resistance in cancer treatment. References: 1. Sherr CJ, Roberts JM (2004) Living with or without cyclins and cyclin-dependent kinases. Genes Dev 18: 2699–2711. 2. Davies TG et al. (2002) Structure-based design of a potent purine-based cyclin-dependent kinase inhibitor. Nat Struct Biol 9: 745–749. 3. Krasinska L et al. (2008) Cdk1 and Cdk2 activity levels determine the efficiency of replication origin firing in Xenopus. EMBO J 27: 758–769. 4. Krasinska L, Cot E, Fisher D (2008) Selective chemical inhibition as a tool to study Cdk1 and Cdk2 functions in the cell cycle. Cell Cycle 7: 1702–1708. 5. Vassilev LT et al. (2006) Selective small-molecule inhibitor reveals critical mitotic functions of human CDK1. Proc Natl Acad Sci USA 103: 10660–10665.

6th International Conference: Inhibitors of Protein Kinases Vol. 56

P13 Effects of U0126, the inhibitor of mitogenactivated/extracellular signal-regulated protein kinase kinase 1(MEK1), on the first two mitoses in the mouse embryo Jacek Z. Kubiak1, Marta SikoraPolaczek2, Zuzanna Maciejewska2, Aude Pascal1 and Maria A. Ciemerych3 1Institute

of Genetics & Development, CNRS-UMR 6061, “Mitosis & Meiosis” Group, IFR 140 GFAS, University of Rennes 1, Faculty of Medicine, France; 2Department of Embryology, and 3Department of Cytology, Institute of Zoology, Faculty of Biology, University of Warsaw, Warsaw, Poland e-mail: [email protected] The first two mitoses of the mouse embryo differ significantly. Among others, the first mitosis takes much more time than the second one [1, 2]. We have shown recently that the prolongation of the first embryonic mitosis in the mouse embryo depends on a true metaphase arrest which does not include the spindle assembly checkpoint mechanism [3]. The nature of this arrest, taking 30–45 min, remains unknown. MEK1-ERK1/ERK2 MAP kinase signaling pathway plays an important role in regulation of the M-phase progression. It is a key component of the CSF activity arresting oocytes in MII of meiosis and is functional during the embryonic preimplantation period [4].

Figure 1. Effect of U0126 treatment on cleaving mouse embryos. A. First mitotic division is arrested in M-phase in the presence of inhibitor. B. Second mitotic division is not sensitive to inhibitor. Here, one of the blastomeres cleaved in the presence of U0126. Red, tubulin (immunofluorescence); blue, chromatin staining; bar, 20 μm.

We analyzed the effects of U0126, a potent inhibitor of MEK1 kinase [5], on the first and the second mitosis in the mouse embryo cultured in vitro. U0126 perturbs the first mitotic spindle formation, but has no effect on the second mitotic spindle assembly. As a consequence, the one-cell embryos arrest in the first mitosis with condensed chromosomes and disorganized spindle, while the two-cell embryos undergo unperturbed mitosis. This shows another important difference in regulation of the two mitoses and indicates that only the first and not the

51

second embryonic mitosis is inhibited by the drug. This also suggests that the MEK1 pathway could participate in regulation of the first, but not the second mitotic division. However, U0126 may inhibit other enzymes, which could also be involved in mitotic regulation. The signaling pathway involving ERK5 MAP kinase, known to be activated by MEK5, can also be inhibited by U0126 [6, 7]. Since no data are available on the MEK5/ERK5 pathway in mouse oocytes, and very few concern the early embryo, we now study this kinase during early embryonic development. We show that activation/inactivation of ERK5 correlates with the progression of mitotic division. Further studies are being performed to understand the role of this MAP kinase in early mouse cleavages and development. References: 1. Ciemerych MA, Maro B, Kubiak JZ (1999) Control of duration of the first two mitoses in a mouse embryo. Zygote 7: 293–300. 2. Kubiak JZ, Chesnel F, Richard-Parpaillon L, Bazile F, Pascal A, Polanski Z, Sikora-Polaczek M, Maciejewska Z, Ciemerych MA (2008) Temporal regulation of the first mitosis in Xenopus and mouse embryos. Mol Cell Endocrinol 282: 63–69. 3. Sikora-Polaczek M, Hupalowska A, Polanski Z, Kubiak JZ, Ciemerych MA (2006) The first mitosis of the mouse embryo is prolonged by transitional metaphase arrest. Biol Reprod 74: 734–743. 4. Verlhac MH, de Pennart H, Maro B, Cobb MH, Clarke HJ (1993) MAP kinase becomes stably activated at metaphase and is associated with microtubule-organizing centers during meiotic maturation of mouse oocytes. Dev Biol 158: 330–340. 5. Phillips KP, Petrunewich MA, Collins JL, Booth RA, Liu XJ, Baltz JM (2002) Inhibition of MEK or cdc2 kinase parthenogenetically activates mouse eggs and yields the same phenotypes as Mos(−/−) parthenogenotes. Dev Biol 247: 210–223. 6. Kamakura S, Moriguchi T, Nishida E (1999) Activation of the protein kinase ERK5/BMK1 by receptor tyrosine kinases. Identification and characterization of a signaling pathway to the nucleus. J Biol Chem 274: 26563–26571. 7. Nishimoto S, Nishida E (2006) MAPK signalling: ERK5 versus ERK1/2. EMBO Rep 7: 782–786.

Abstracts 52

P14 Variability patterns, intra/intermolecular interactions and correlated mutations suggest potential specific inhibitor binding sites Jacek Kuska1,4, Jacek Leluk2,4 and Bogdan Lesyng3,4 1Faculty

of Biology, Warsaw University, Warsaw, Poland; of Biological Sciences, University of Zielona Gora, Zielona Gora, Poland; 3Department of Biophysics, Faculty of Physics, University of Warsaw, Warsaw, Poland; 4CoE BioExploratorium, University of Warsaw, Warsaw, Poland e-mail: [email protected] 2Faculty

When designing specific inhibitors of target proteins one should take into account the flexible structures of the interacting objects, as well as the sequential variability of the protein macromolecules, which is a challenging task. Phosphotransferases are represented by a few families in the same cell. Therefore the complete and detailed knowledge of kinases’ catalytic subunits is essential for the effective design of specific inhibitors capable to distinguish individual kinases. We elaborated a Biow@re package of applications [1], which was applied to the analysis of four kinase families and their regulatory subunits, amongst others JAK3 kinase [2]. Our results indicate, for example, that peptidomimetic inhibitors, see e.g. [3], are more suitable for the design of highly specific inhibitors than ATP or glucose competitive inhibitors. This is because of diversity of the peptide binding groove which determines narrow specificity of individual kinases. However, for the same reason, the design of inhibitors interacting with the groove is more difficult. This site is open and gives more unpredictable binding possibilities. A)

C)

B)

Figure 1. Spatial representation of the cAMP-dependent protein kinase family. A) The darker grade shows residues which are closer to the centre of mass. B) The light regions are more variable, the dark ones ― more conservative. C) Correlation between residual variability (vertical axis) and the distance from the catalytic subunit mass centre (horizontal axis).

The center of the molecule that forms the ATP-binding pocket is highly conserved and is not a good potential inhibitor target (comp. Fig. 1). The analysis of hydrophobicity patterns revealed neither correlation between a hydropatic property of an amino

2009

acid with its variability measure, nor with its distance from the centre. Another observation refers to the interaction pattern between residues located at a particular distance. We observed that positions of residues distant from each other by less than 5 Å often show very similar variability range. This means that variable residues are in contact with variable ones, and conserved residues interact mainly with conserved ones. The results show consistency of the peptide binding site variability with its intramolecular binding patterns. Such consistency is not observed in a nucleotide binding site which reveals high conservativity. The pattern of such interactions suggests that correlated mutations follow structural as well as functional requirements of the protein [4, 5], however, it is not univocally confirmed by our results. The correlated mutations occurring at the peptide binding site of kinases is much more complex, and is the subject of ongoing studies. Acknowledgements: This work was supported by CoE BioExploratorium, University of Warsaw. References: 1. Kuska J, Leluk J (2007) Biow@re: a package of applications for intra/intermolecular interaction studies. Acta Biochim Polon 54 (Suppl 3): 61–62. 2. Kuska J, Setny P, Lesyng B (2008) Modelling of possible binding modes of caffeic acid derivatives to JAK3 kinase. In: From computational biophysics to systems biology. Hansmann UHE, Meinke JH, Mohanty S, Nadler W, Zimmermann O, eds. NIC Series 40: 297–300 (ISBN: 978-3-9810843-6-8, http://www.fz-juelich. de/nic- series/volume40/volume40.html). 3. Fear G, Komarnytsky S, Raskin I (2007) Protease inhibitors and their peptidomimetic derivatives as potential drugs. Pharmacol Ther 113: 354–368. 4. Halperin I, Wolfson H, Nussinov R (2006) Correlated mutations: advances and limitations. A study on fusion proteins and on the cohesion-dockerin families. Proteins 4: 832–845. 5. Rosen O, Samson AO, Anglister J (2008) Correlated mutations at gp120 positions 322 and 440: implications for gp120 structure. Proteins 71: 1066–1070.


P15 Rational design of ARC-type bisubstrateanalogue inhibitors of basophilic protein kinases Darja Lavõgina1, Marje Lust1, Jevgenia Rogozina1, Erki Enkvist1, Asko Uri1 and Dirk Bossemeyer2 1Institute

of Chemistry, University of Tartu, Estonia; 2Group of Structural Biochemistry, German Cancer Research Centre, Heidelberg, Germany e-mail: [email protected] Conjugates of an adenosine analogue and an argininerich peptide (ARCs) have been developed as bisubstrateanalogue inhibitors for protein kinases (PK) [1, 2]. ARCs were targeted to the catalytic site of the kinase and they were supposed to bind simultaneously to binding sites of both substrates of PKs, ATP and the phosphorylatable protein/peptide. The crystal structure (Fig. 1A) of the complex cAMP-dependent protein kinase catalytic subunit (PKA C) with a representative of ARC-type inhibitors ARC-1034 demonstrated the principal binding pattern of ARC-type inhibitors and paved the way for the rational design of the highly potent ARC-type bisubstrate analogues (Fig. 1B) [3].

A

53

There was no indication of a direct interaction of the arginines of the ARC-1034 with the known substrate recognition residues of PKA C (Glu127, Glu170, and Glu230 [4]). In this respect, ARC-1034 represents only in part the high-affinity ARCs that contain four or six arginines. We presume that the first two d-arginine residues of the longer conjugates provide suitable stereochemical geometry to serve as a joining chain between the linker and the more distal C-terminal arginines, which then are responsible for the interaction with basophilic kinases, including PKA. The elongation of the linker by adding the second Ahx moiety facilitated the interaction of the peptidic part of the bisubstrate-analogue with the amino-acid residues of PKA C responsible for the substrate consensus sequence recognition. The highest affinity towards PKA C was obtained for the conjugate incorporating a d-amino acid as the chiral spacer between the two Ahx-moieties, which oriented the attached oligo-(d-arginine) peptide for its optimal interaction with the kinase. Selectivity testing of the most potent of the novel ARCtype compounds, ARC-1028 was performed in a panel of 50 kinases (Invitrogen SelectScreen Z’-LYTE Assay). As expected, ARC-1028 inhibited most potently basophilic protein kinases of the AGC group and did not inhibit the acidophilic protein kinase CK1 and tyrosine kinase Src. Overall, the most inhibited protein kinases (over 90%

B

Figure 1. A. Top: electron density map within 1.6 Å around ARC-1034 molecule in the active site contoured at 1σ; bottom: structure of ARC-1034. B. Structures of compounds ARC-1028 and ARC-1044.

A prominent and important feature of the ARC molecule is the linker consisting of the 6-aminohexanoic acid moiety (Ahx) and joining the nucleosidic and peptidic part of the inhibitor. Its length, electronic properties and backbone flexibility support multiple favorable interactions with the glycine flap of the kinase (e.g., residues Ser53, Phe54, and Gly55), thus providing an explanation for the good inhibitory potency of ARC-type compounds.

inhibition at 100 nM concentration) were PKA C, PKC (conventional and novel isoforms), ROCK isoforms, and ribosomal S6-kinases (RSK, MSK, p70S6K) [3]. Binding and kinetic assays [5, 6] were used to characterize novel ARCs and establish their selectivity determinants towards kinases of the AGC group, PKA C, ROCK-II and PKBγ. Combining our previous knowledge [7] with the recent results, we designed the compound

Abstracts 54

ARC-1044 (Fig. 1B; Ki < 1 nM towards PKA C) possessing two structural elements supporting PKA C selectivity: the nucleosidic part, represented by the cyclopentane-based carbocyclic analogue of 3’-deoxyadenosine, and the small hydrophobic chiral spacer between the linkers, represented by d-alanine residue. The inhibitory properties of ARC-1044 confirmed our predictions, as the compound exhibited more than 100-fold selectivity toward PKA C over ROCK-II and PKBγ. Finally, the bisubstrate character of the novel inhibitors was confirmed by the fluorescence polarization-based binding/displacement assay [6]. Originating from ARC1028, a fluorescent probe was designed with KD of 0.3 nM towards PKA C. This probe was successfully displaced from its complex with PKA C by both H89 (targeted to ATP-binding site of ATP) and RIIα (regulatory subunit of PKA, known to compete with protein/peptide substrates). Acknowledgements: The work was supported by grants from the Estonian Science Foundation (6710) and the Estonian Ministry of Education and Sciences (SF0180121s08). References: 1. Loog M, Uri A, Raidaru G, Järv J, Ek P (1999) Adenosine-5’-carboxylic acid peptidyl derivatives as inhibitors of protein kinases. Bioorg Med Chem Lett 9: 1447–1452. 2. Enkvist E, Lavogina D, Raidaru G, Vaasa A, Viil I, Lust M, Viht K, Uri A (2006) Conjugation of adenosine and hexa-(d-arginine) leads to a nanomolar bisubstrate-analog inhibitor of basophilic protein kinases. J Med Chem 9: 7150–7159. 3. Lavogina D, Lust M, Viil I, König N, Raidaru G, Rogozina J, Enkvist E, Uri A, Bossemeyer D (2009) Structural analysis of ARC-type inhibitor (ARC-1034) binding to protein kinase A catalytic subunit and rational design of bisubstrate analogue inhibitors of basophilic protein kinases. J Med Chem 52: 308–321. 4. Zheng J, Knighton DR, ten Eyck LF, Karlsson R, Xuong N, Taylor SS, Sowadski JM (1993) Crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MgATP and peptide inhibitor. Biochemistry 32: 2154–2161. 5. Viht K, Vaasa A, Raidaru G, Enkvist E, Uri A (2005) Fluorometric TLC assay for evaluation of protein kinase inhibitors. Anal Biochem 340: 165–170. 6. Vaasa A, Viil I, Enkvist E, Viht K, Raidaru G, Lavogina D, Uri A (2009) High-affinity bisubstrate probe for fluorescence anisotropy binding/displacement assays with protein kinases PKA and ROCK. Anal Biochem 385: 85–93. 7. Enkvist E, Raidaru G, Vaasa A, Pehk T, Lavogina D, Uri A (2007) Carbocyclic 3’-deoxyadenosine-based highly potent bisubstrate-analog inhibitor of basophilic protein kinases. Bioorg Med Chem Lett 17: 5336–5339.

2009

P16 Examination of CK2 inhibitors as potential anticancer and antibacterial agents Małgorzata Makowska1, Justyna Maszkowska1, Aleksandra Bilińska2, Renata Wolinowska3, Stanisław Tyski3, Mirosława Koronkiewicz4 and Maria Bretner1,2 1Institute

of Biochemistry and Biophysics, Polish Academy of Science, Warsaw, Poland; 2Warsaw University of Technology, Warsaw, Poland; 3Medical University of Warsaw, Warsaw, Poland; 4National Medicines Institute, Warsaw, Poland e-mail: [email protected] Protein kinase CK2 (casein kinase 2) play essential roles in many cellular functions, with more than 300 protein substrates identified to date [1]. CK2 is more abundant in tumors as compared to normal tissues and display antiapoptotic effect in cancer cell lines, what make it an important target for antineoplastic drugs (reviewed in [2–4]). The aim of this study was the investigation of the influence of CK2 inhibitors on the cell growth and apoptosis at few neoplastic cell lines. The influence of CK2 inhibitors on a growth of a few bacterial strains was also examined. The first group of compounds were known CK2 inhibitors 4,5,6,7-tetrabromobenzotriazole (TBBt), 4,5,6,7-tetrabromobenzimidazole (TBBi), and their derivatives 3-(4,5,6,7tetrabromo-1H-benzimidazol-1-yl)propan-1-ol (MB001), 3-(4,5,6,7-tetrabromo-1H-benzotriazol-1-yl)propan-1-ol (MB002), and 3-(4,5,6,7-tetrabromo-2H-benzotriazol-2yl)propan-1-ol (MB003) [5]. The second of group of compounds were newly synthesized analogs of benzotriazole and benzimidazole with different substituents in the benzene ring. Experiments were carried out using human neoplastic cell lines: HL-60 (human promyleocytic leukemia), K-562 (human leukemia) and DTA (human colon carcinoma, phorbol esters resistant subline). The tested compounds exhibited various proapoptotic efficacies. In HL-60 cells the most active were MB001 inducing 95% of apoptosis at 25 μM after 48 h incubation time and TBBi ― 80% after 48 h. The viability of DTA cells dropped to 59% after 48 h treatment with MB001 at concentration of 25 μM. The CK2 inhibitors TBBt, TBBi, and MB001 were analyzed for antibacterial potential against five Gram positive bacteria: Bacillus subtilis, Bacillus cereus, Micrococcus luteus, Staphylococcus aureus, Staphylococcus epidermidis. Preliminary screening showed that only TBBt exerted inhibition against all examined bacteria. The MICs (minimum inhibitory concentrations) for TBBt were in the range 12.5–50 μM. Acknowledgements: The study was supported by the Ministry of Science and Higher Education grant PBZ-MNiSW 04/I/2007 and partially by the Warsaw University of Technology. References: 1. Meggio F, Pinna LA (2003) One-thousand-and-one substrates of protein kinase CK2? FASEB J 17: 349–368.

6th International Conference: Inhibitors of Protein Kinases Vol. 56 2. Guerra B, Issinger OG (2008) Protein kinase CK2 in human diseases. Curr Med Chem 15: 1875–1886. 3. Duncan JS, Litchfield DW (2008) Too much of a good thing: the role of protein kinase CK2 in tumorigenesis and prospects for therapeutic inhibition of CK2. Biochim Biophys Acta 1784: 33–47. 4. Pinna LA (2002) Protein kinase CK2: a challenge to canons. J Cell Science 115: 3873–3878. 5. Najda-Bernatowicz A, Łebska M, Orzeszko A, Kopańska K, Krzywińska E, Muszyńska G, Bretner M (2009) Synthesis of new analogs of benzotriazole, benzimidazole and phthalimide ― potential inhibitors of human protein kinase CK2. Bioorg Med Chem 17: 1573–1578.

55

P17 Contribution of pH-dependence to binding of peptide inhibitors by protein kinase A (PKA)

Zofia Piłat, David Shugar and Jan M. Antosiewicz Division of Biophysics, Faculty of Physics, University of Warsaw, Warszawa, ul Żwirki i Wigury 93, 02-089, Poland e-mail: [email protected] Proteins are dynamic systems of electric charges, and electrostatic forces are considered to constitute one of the principal factors of their intra- and inter-molecular interactions. The free energy of a protein, modeled as a particular charge distribution, is traditionally identified with the work necessary to assemble this charge distribution, and it is calculated within the Poisson-Boltzmann model of the given solute-solvent system. Therefore, the electrostatic contribution to the free energy of association of proteins A and B to form the complex AB is as follows: ∆Gbind =

1 nAB 1 nA 1 nB qiφi − ∑ qiφi ∑ qiφi − 2 ∑ 2 i =1 2 i =1 i =1

(1)

Where φi is the electrostatic potential at the location of the charge qi, nx is the number of charges in the species x and each sum represents the work necessary to assemble the nx charges in the appropriate dielectric cavity in aqueous medium. One important factor neglected in Eqn. 1 is related to constant fluctuations in electric charge distribution within proteins resulting from proton exchange by sidechains of some amino acid residues. A protein with M such groups can be found in one of 2M protonation states. Each such state is characterized by the free energy Gm(x1,m,...,xM,m,pH,T), m=1,…,2M with x1,m= 1 or 0, depending on whether group i in the state m is protonated or not, respectively [1]. The probability of finding a given protein in the state m is governed by the Boltzmann law. The total ionization free energy, including these protonation degrees of freedom, reads: 2M

G o = − RT ln ∑ e m =1

−

Gm RT

(2)

and the free energy of association, with protonation degrees of freedom taken into account, is o o (3) ∆Gbind = GAB − GAo − GBo + ∆Gn →n Computation of the absolute value of the free energy of o association ∆Gbind according to Eqn. 3 requires the thermodynamic cycle shown in Fig. 1. All molecules considered in this cycle are fixed in one conformational state. Possible structural changes accompanying the association process are neglected. Associating proteins, with full protonation freedom, initially have all their ionizable groups neutralized, then formation of the complex of the electrostatically neutral protein molecules is carried out, and finally the complex restores full protonation freedom. Details of the calculations are described elsewhere [2]. In the present study, the pH-dependent electrostatic contribution to the free energy of association is calculated for the catalytic subunit of protein kinase A (PKA) and

Abstracts 56

2009

Figure 3. pH-dependence of the free energy of PKA-PKI association.

Figure 1. Thermodynamic cycle used in the present study. Colored structures represent species in pH-dependent protonation equilibrium, able to exchange protons with solvent. Gray structures represent species “frozen” in one protonation state with all titratable groups neutral.

its peptide inhibitor (PKI). Results from this method are compared with those from the traditional approach for two structures of PKA-PKI complexes (1XH9 [3] and 1ATP [4] available from Protein Data Bank). Structures of the complexes are shown in Fig. 2.

Figure 2. PKA–PKI complexes used in the present study. PKA and PKI structures are very similar in both cases, but in the 1ATP structure, beside the inhibitor, there is an ATP and two Mn2+ ions in the kinase binding site.

The electrostatic free energies of association obtained by the traditional and the present approach, respectively, at pH 7 and ionic strength corresponding to 150 mM of monovalent salt, are shown in Table 1. Table 1. Free energy of association obtained by the new method (left) and the traditional approach (right), respectively.

1XH9 1ATP

o ∆Gbind [kcal/mole]

∆Gbind [kcal/mole]

–1.5 –7.2

–9.4 –9.1

It will be noted that the explicit inclusion of the protonation degrees of freedom has a substantial effect on the computed free energies of association. The pH-dependence of the free energy of association, obtained by the new method for both complexes, is shown in Fig. 3. Qualitatively different pH-dependence is observed between the two complexes, apparently related to the presence of the ATP cofactor in the structure described by the 1ATP file and its absence in the 1XH9 file. Unfortunately

no appropriate experimental data are available. However, our study indicates that when one attempts to predict theoretically the free energy of protein-ligand association, and its polar contribution is calculated in the traditional way, an error of several kcal/mole can be made. References: 1. Antosiewicz JM (2008) Protonation free energy levels in complex molecular systems. Biopolymers 89: 262–269. 2. Piłat Z, Antosiewicz JM (2008) Multiple protonation equilibria in electrostatics of protein-protein binding. J Phys Chem B 112: 15074–15085. 3. Breitenlechner C, Friebe W, Brunet E, Werner G, Graul K, Thomas U, Kuenkele K, Schaefer W, Gassel M, Bossemeyer D, Huber R, Engh RA, Masjost B (2005) Design and crystal structures of protein kinase B-selective inhibitors in complex with protein kinase A and mutants. J Med Chem 48: 163–170. 4. Zheng J, Trafny E, Knighton D, Xuong NH, Taylor S, Eyck LT, Sowadski J (1993) 2.2 Å refined crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MnATP and a peptide inhibitor. Acta Crystall Sect D Biol Crystallogr 49: 362–365.


P18 Biochemical studies of a family of atypical protein kinases from the malarial parasite Plasmodium falciparum

Ailsa J. Powell1, Jane Endicott1 and Oliver Billker2 1Laboratory

of Molecular Biophysics, Department of Biochemistry, University of Oxford, Oxford. OX1 3QU. UK; 2Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge. CB10 1SA. UK e-mail: [email protected] The human malarial parasite, Plasmodium falciparum, causes over two million deaths annually in sub-Saharan Africa. P. falciparum is readily developing resistance of to available antimalarials and the lack of a vaccine requires novel approaches to the development of treatment. In this study we are investigating the use of targeting the calcium-dependent protein kinase (CDPK) family from P. falciparum as a novel drug target. CDPKs are present in the genomes of plants, ciliates, green algae and apicomplexan parasites but not in mammals, making them an attractive drug target [1, 2]. The P. falciparum genome encodes six distinct CDPKs, which are expressed in both the human and Anopheles stages of the parasites lifecycle. In this study we are focused on biochemical, biophysical and structural approaches to characterise members of the CDPK family and to support a program to develop ATP-competitive inhibitors for use in functional studies and as leads for drug development. Looking at PfCDPK1, PfCDPK4 and PfCDPK5, which are essential to the survival of the parasite, all share the same domain structure in which a serine/threonine kinase domain (KD) is fused to four C-terminal EF hands that make up the Ca2+ binding domain (CD) by an intervening junction domain (JD) ([3, 4], Fig. 1). 50 L

330 Kinase domain

370 J-region

550 Ca-binding domain

Figure 1. The P. falciparum CDPK family. (A) Domain organisation. The kinase domain (KD) of about 280 amino acids is preceded by a short leader sequence (L). The junction region (J, green), of about 40–50 amino acids, is an autoinhibitory pseudosubstrate domain. Residues numbers given are representative values across the CDPK family. (B) Structure of Cryptosporidium parvum CDPK (SGC, PDB code 2QG5). (C) Structure of an Arabidopsis thaliana junction-4EF hand Ca2+ binding domain construct ([5], PDB code 2AAO). The domain can be divided into N- and C-terminal lobes coloured red and blue, respectively.

Focusing our efforts initially on the kinase domain of PfCDPK5 we have cloned and expressed a number of active constructs. Using these constructs in a Thermafluor thermostability assay we were able to measure the inter-

57

action of ATP-competitive inhibitors with PfCDPK5-KD by measuring differences in melting temperature. Screening of 1585 compounds via this method have identified a family of five-related compounds which also demonstrate evidence of a structure activity relationship. We will use these compounds to determine IC50 values allowing us to evaluate the correlation between the interaction of the compounds with the protein versus their ability to inhibit the enzymatic activity. Cloning of the equivalent kinase domain constructs of PfCDPK1 and PfCDPK4 are underway and we aim to repeat the same inhibitor screens with these proteins. This will generate help generate an inhibitor profile of the ATP-binding sites and will hopefully isolate inhibitors that are specific to the family but also inhibitors that are specific to each individual PfCDPK. References: 1. Doerig C, Billker O, Pratt D, Endicott J (2005) Protein kinases as targets for antimalarial intervention: kinomics, structure-based design, transmission blockade, and targeting host cell enzymes. Biochim Biophys Acta 1754: 132–150. 2. Pattanaik P, Raman J, Balaram H (2002) Perspectives in drug design against malaria. Curr Top Med Chem 2: 483–505. 3. Ward P, Equinet L, Packer J, Doerig C (2004) Protein kinases of the human malaria parasite Plasmodium falciparum: the kinome of a divergent eukaryote. BMC Genomics 5: 79. 4. Harper JF, Harmon A (2005) Plants, symbiosis and parasites: a calcium signalling connection. Nat Rev Mol Cell Biol 6: 555–566. 5. Chandran V, Stollar EJ, Lindorff-Larsen K, Harper JF, Chazin WJ, Dobson CM, Luisi BF, Christodoulou J (2006) Structure of the regulatory apparatus of a calcium-dependent protein kinase (CDPK): a novel mode of calmodulin target recognition. J Mol Biol 357: 400–410.

Abstracts 58

P19 Model selection of the JAK–STAT pathway activation mechanism

Mikołaj Rybiński and Anna Gambin Institute of Informatics, Warsaw University, Warsaw, Poland e-mail: [email protected] Intercellular communication occurs between biomolecules, through a fixed set of reaction channels which create a complicated network. Mathematical modeling of such molecular, signal processing pathways might be very beneficial for acquiring system level understanding about dynamical mechanisms of the cell. The Janus kinase (JAK) and the signal transducer and activator of transcription (STAT) signalling pathway family is highly conserved in eukaryotic organisms and provides a direct route to the nucleus, where in effect gene transcription is altered. This signalling mechanism has coevolved with multiple cellular events, such as antiviral, innate and adaptive immune responses [1] (STAT1 and STAT2 pathways), as well as cell growth and apoptosis processes regulation [2] and embryonic stem cell self-renewal control [3] (STAT3 and STAT5 pathways). Our work focuses on cytokine receptors activation process, in particular, the importance of the dimerization step (cf. [4]). Early experimental methods used to understand the activation of receptors (e.g. immunoprecipitation), did not give precise picture of it’s mechanics because of invasiveness [5]. It has been shown that cytokine receptors assemble on the membrane before the external stimuli is present. Such result was presented for the cytokine receptor activated by type II IFN [5, 6] and subsequently for the interleukin activated receptor [7].

a) “Original” Yamada et al. [9] model and “no JAK” variant.

b) “IFN to dimer” binding case and it’s “no dimerization” variant with preassembled receptors, proposed by Shudo et al. [8]. Figure 1. Schemes of the receptor activation mechanism variants in the JAK–STAT signalling pathway model. Each subfigure represents two model variants: with or without the first reaction.

We review computational models of the JAK–STAT signalling, where the focus was mainly on negative regulation and the core pathway mechanism, i.e. STATs proteins life-cycle. In particular, the version of Shudo et al. [8] differs from the original Yamada et al. [9] model in the receptor activation steps due to the above-mentioned discover-

2009

ies. We present four computational JAK–STAT pathway model variants which capture known differences in receptor activation steps between computational as well as biological models (Fig. 1). Using numerical simulations of the ordinary differential equations (ODE), given by the deterministic semantics of a biochemical reaction network, we investigated the influence of the receptor activation mechanism differences on the model dynamics. We present methodology that allows to state if one model is better than the other. To this end we employed known method of Bayes factor based model selection [10]. For specific dynamics differences driven model selection we used the state-of-the-art global sensitivity analysis [11]. The Bayesian inference gave no evidence in favor of any of the models, but sensitivity analysis of the receptor activation module revealed that the preassembled receptors model is the most flexible in terms of noised data fit and the least robust in terms of original behavior. In context of the range of responsibilities of the JAK–STAT pathways and the parsimony of nature rule of thumb, the “no dimerization” receptor activation variant is recommended. Moreover, all kinetic parameters sensitivity analysis showed that “no JAK” model robustness is slightly more scattered among the network indicating evolutionary preferable mechanism and justification for the constitutive binding of JAKs to the respective receptor chains. References: 1. Aaronson D, Horvath C (2002) A road map for those who don’t know JAK-STAT. Science 296: 1653–1655. 2. Yu H, Jove R (2004) The STATs of cancer ― new molecular targets come of age. Nat Rev Cancer 4: 97–105. 3. Raz R, Lee C-H, Cannizzaro LA, d’Eustachio P, Levy E (1999) Essential role of STAT3 for embryonic stem cell pluripotency. Proc Natl Acad Sci USA 96: 2846–2851. 4. Whitty A, Raskin N, Olson DL, Borysenko CW, Ambrose CM, Benjamin CD, Burkly CC (1998) Interaction affinity between cytokine receptor components on the cell surface. Proc Natl Acad Sci USA 95: 13165–13170. 5. Krause CD, Mei E, Xie J, Jia Y, Bopp MA, Hochstrasser RM, Pestka S (2002) Seeing the light: preassembly and ligand-induced changes of the interferon-γ receptor complex in cells. Mol Cell Proteomics 1: 805–815. 6. Krause CD, Lavnikova N, Xie J, Mei E, Mirochnitchenko OV, Jia Y, Hochstrasser RM, Pestka S (2006) Preassembly and ligandinduced restructuring of the chains of the IFN-γ receptor complex: the roles of Jak kinases, Stat1 and the receptor chains. Cell Res 16: 55–69. 7. Schuster B, Meinert W, Rose-Johns S, Kallen KJ (2003) The human interleukin-6 (IL-6) receptor exists as a preformed dimer in the plasma membrane. FEBS Lett 538: 113–116. 8. Shudo E, Yang J, Yoshimura A, Iwasa Y (2007) Robustness of the signal transduction system of the mammalian JAK/STAT pathway and dimerization steps. J Theor Biol 246: 1–9. 9. Yamada S, Shiono S, Joo A, Yoshimura A (2003) Control mechanism of JAK/STAT signal transduction pathway. FEBS Lett 534: 190–196. 10. Vyshemirsky V, Girolami MA (2008) Bayesian ranking of biochemical system models. Bioinformatics 24: 833–900. 11. Saltelli A, Ratto M, Tarantola S, Campolongo F (2005) Sensitivity analysis for chemical models. Chem Rev 105: 2811–2828.


P20

P21

Resorufin ― a lead for a new protein kinase CK2 inhibitor

Brain glucodeprivation induces DYRK1a overexpression in neurons

Institute for Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, 5230 Odense, Denmark e-mail: [email protected]

1Department

Iben Skjøth Sandholt, Birgitte Brinkmann Olsen, Barbara Guerra and Olaf-Georg Issinger

Screening of a natural compound library led to the identification of resorufin, as a highly selective and potent inhibitor for protein kinase CK2. CK2 is a ubiquitous and essential protein kinase implicated in a wide variety of cellular processes such as proliferation, apoptosis, differentiation and transformation [1–2]. Protein kinase CK2 is composed of two catalytic (α/α’) subunits attached to a dimer of non-catalytic chains (β). Kinetic analysis showed that resorufin is an ATP competitive inhibitor and the holoenzymes were more specifically inhibited than the free catalytic subunits. Testing of 31 serine/threonine, one lipid and 20 tyrosine kinases showed that resorufin, beside CK2, only inhibited SYK (48% inhibition), HIPK2 (31%) and PIM3 (32%). This is in contrast to emodin, a structurally related known CK2 inhibitor [3], which also inhibited nine other kinases up to 90%. To test if resorufin could penetrate the cell membrane and to test the effects of resorufin in established cell lines, four different human cancer cell lines were subjected to resorufin treatment. In the case of the three prostate carcinoma cell lines (PC-3, DU-145, LNCaP) treatment with 40 μM resorufin for 24 h led to 15–20% dead cells, however no caspase-mediated apoptosis was observed. In the case of the colorectal carcinoma HCT116 cell line a similar picture was obtained, yet, when resorufin was administered in cells treated with doxorubicin, apoptosis was induced within 24 h. Endogenous protein kinase CK2 was inhibited by resorufin by about 80% in the three prostate cell lines, however in the case of the HCT116 cells, the inhibition was only 40%. Hence, the discovery of a novel CK2 inhibitor with high selectivity and reasonable potency makes this compound a promising candidate to target protein kinase CK2 in vivo, which in recent years has been coined a druggable kinase [4–5]. References: 1. Guerra B, Issinger OG (2008) Protein kinase CK2 in human diseases. Curr Med Chem 15: 1875–1886. 2. Duncan JS, Litchfield DW (2008) Too much of a good thing: the role of protein kinase CK2 in tumorigenesis and prospects for therapeutic inhibition of CK2. Biochim Biophys Acta 1784: 33–47. 3. Yim H, Lee YH, Lee CH, Lee SK (1999) Emodin, an anthraquinone derivative isolated from the rhizomes of Rheum palmatum, selectively inhibits the activity of casein kinase II as a competitive inhibitor. Planta Med 65: 9–13. 4. Sarno S, Pinna LA (2008) Protein kinase CK2 as a druggable target. Mol Biosyst 4: 889–894. 5. Pagano MA, Cesaro L, Meggio F, Pinna LA (2006) Protein kinase CK2: a newcomer in the “druggable kinome”. Biochem Soc Trans 34: 1303–1306.

59

Dorota Sulejczak1, Michał Fiedorowicz1, Marzena Labak2, Bogusław Tomanek2,3 and Paweł Grieb1 of Experimental Pharmacology, Mossakowski Medical Research Centre, Polish Academy of Sciences, Warsaw, Poland; 2Niewodniczański Institute of Nuclear Physics, Polish Academy of Sciences, Kraków, Poland; 3National Research Council of Canada Institute for Biodiagnostics and Experimental Imaging Centre, Calgary, Canada e-mail: [email protected] DYRK1a is a proline-directed protein kinase which plays a critical role in neurodevelopment [1, 2]. Localized in the Down syndrome critical region of chromosome 21, it is considered a strong candidate culprit gene for the associated learning defects [2]. A recent study showed that DYRK1a alters splicing and leads to hyperphosphorylation of tau protein, suggesting that this kinase is also involved in brain tau phosphorylation and contributes to neurofibrillary degeneration [3, 4]. Moreover, amyloid precursor protein (APP) is also phosphorylated by DYRK1a in vitro and in mammalian cells, and overexpression of this kinase may play a role in accelerating Alzheimer’s disease (AD) pathogenesis through phosphorylation of APP. Here we show that, in rat brain in vivo, both acute and chronic glucodeprivation enhances DYRK1a expression in cortical neurons and astroglia. Acute glucodeprivation was induced by intraperitoneal application of 500 mg/kg of the non-metabolizable glucose derivative 2-deoxyglucose (2DG). Chronic suppression of brain glucose utilization was induced by intracerebroventricular injections of the diabetogenic toxin streptozotocin (3 mg/kg; repeated in day 1 and 3), which is thought to damage insulin receptors. Twenty four hours after the injection of 2DG, or two months following the first application of streptozotocin, the rats were deeply anesthetized and perfused through the ascending aorta with a buffer and an ice-cold fixative. DYRK1a expression was evaluated immunohistochemically with a specific monoclonal antibody (a gift from the late Professor Krystyna Wiśniewska, IBR, Staten Island, NY, USA). In the cerebral cortex from control rats we detected some DYRK1a-immunopositive neurons and astrocytes (mainly in the first cortical layer) with cytoplasmatic localization of staining, and only a few neurons that showed nuclear labeling. 2DG, as well as streptozotocin, markedly elevated expression of DYRK1a and increased the number of immunoreactive neurons displaying nuclear staining. Our observations point to disturbances in brain glucose utilization as a possible trigger of DYRK1a overexpression in cortical neurons. As it is well known that a decrease in brain glucose consumption is an early sign of AD, a possibility shall be considered that consequent DYRK1a induction links failing brain glucose metabolism to the development of dementia, with significant implications in the search for drugs to treat AD.

Abstracts 60 Acknowledgements:

P22

The present study was supported by statutory for the Department of Experimental Pharmacology of Mossakowski Medical Research Centre, Polish Academy of Sciences, and Science Network “Biovision – visualization of biomedical phenomena” (grant of the Ministry of Science and Higher Education).

Evaluation of anti-tumor activity of WP1130, a novel inhibitor of the JAK2/ STAT3 pathway in glioma cells

References: 1. Tejedor F, Zhu XR, Kaltenbach E, Ackermann A, Baumann A, Canal I, Heisenberg M, Fischbach KF, Pongs O (1995) Minibrain: a new protein kinase family involved in postembryonic neurogenesis in Drosophila. Neuron 14: 287–301. 2. Ryoo SR, Cho HJ, Lee HW, Jeong HK, Radnaabazar C, Kim YS, Kim MJ, Son MY, Seo H, Chung SH, Song WJ (2008) Dualspecificity tyrosine(Y)-phosphorylation regulated kinase 1A-mediated phosphorylation of amyloid precursor protein: evidence for a functional link between Down syndrome and Alzheimer’s disease. J Neurochem 104: 1333–1344. 3. Liu F, Liang Z, Wegiel J, Hwang YW, Iqbal K, Grundke-Iqbal I, Ramakrishna N, Gong CX (2008) Overexpression of Dyrk1A contributes to neurofibrillary degeneration in Down syndrome. FASEB J 22: 3224–3233. 4. Wegiel J, Dowjat K, Kaczmarski W, Kuchna I, Nowicki K, Frackowiak J, Mazur Kolecka B, Wegiel J, Silverman WP, Reisberg B, deLeon M, Wisniewski T, Gong CX, Liu F, Adayev T, Chen-Hwang MC, Hwang YW (2008) The role of overexpressed DYRK1A protein in the early onset of neurofibrillary degeneration in Down syndrome. Acta Neuropathol 116: 391–407.

2009

Karolina Swiatek-Machado1, Grzegorz Grynkiewicz2, Bogdan Lesyng3, Piotr Setny3, Wiesław Szeja4, Waldemar Priebe5 and Bozena Kaminska1 1Laboratory

of Transcription Regulation, Nencki Institute, Warsaw, Poland; 2Pharmaceutical Research Institute, Warsaw, Poland; 3Department of Biophysics, University of Warsaw, Warsaw, Poland; 4Syntex, Gliwice, Poland; 5Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas, USA e-mail: [email protected] Signal transducers and activators of transcription (STATs) are crucial regulators of cell proliferation, survival, and differentiation. Aberrantly activated STATs (in particular, STAT3 and STAT5) play a critical role in malignant transformation and tumorigenesis [1]. In many cancerous cell lines and tumors, where growth factor signaling is frequently dysregulated, the STAT3 and STAT5 proteins are persistently tyrosine phosphorylated. Such constitutively activated STAT3 may promote uncontrolled growth and survival through aberrant expression of cyclin D1, c-Myc, Bcl-xL, Mcl-1 and survivin genes contributing to oncogenesis [2]. Constitutive activation of STAT3 results from either deregulation of upstream kinases (JAK1, JAK2, EGFR) or loss of endogenous inhibitors. Antitumor and/ or proapoptotic activity of JAK2 inhibitors have been reported [3, 4]. In the present work, using C6 rat glioma cells, we evaluated activity of WP1130, a novel JAK2/STAT3 pathway inhibitor that is structurally related to WP1066 [5]. Our molecular modeling indicates that WP1130 should bind with a higher affinity to JAK2 than WP1066 does. WP1130 significantly reduced the level of phosphorylated JAK1 and JAK2 as well as phosphorylated STAT3 (Tyr705), at 10 μM concentration. Morphological alterations, reduction of cell survival and appearance of cleaved caspase 3 and PARP were detected 24–48 h after treatment with 5– 10 μM WP1130. The expression of STAT-dependent genes was diminished in WP1130-treated C6 glioma cells. Treatment with WP1130 modulated other signaling pathways, in particular strongly induced phosphorylation of MAP kinases after short incubation. The results described here demonstrate a promising antitumor activity of WP1130 in glioma in vitro models and provide insights into molecular mechanism of its action. References: 1. Yu H, Jove R (2004) The STATs of cancer ― new molecular targets come of age. Nat Rev Cancer 4: 97–105. 2. Alvarez JV, Frank DA (2004) Genome-wide analysis of STAT target genes: elucidating the mechanism of STAT-mediated oncogenesis. Cancer Biol Ther 3: 1045–1050. 3. Duan Z, Bradner J, Greenberg E, Mazitschek R, Foster R, Mahoney J, Seiden MV (2007) 8-Benzyl-4-oxo-8-azabicyclo[3.2.1]oct2-ene-6,7-dicarboxylic acid (SD-1008), a novel Janus kinase 2

6th International Conference: Inhibitors of Protein Kinases Vol. 56 inhibitor, increases chemotherapy sensitivity in human ovarian cancer cells. Mol Pharmacol 72: 1137–1145. 4. Gozgit JM, Bebernitz G, Patil P, Ye M, Parmentier J, Wu J, Su N, Wang T, Ioannidis S, Davies A, Huszar D, Zinda M (2008) Effects of the JAK2 inhibitor, AZ960, on Pim/BAD/BCL-xL survival signaling in the human JAK2 V617F cell line SET-2. J Biol Chem 283: 32334–32343. 5. Iwamaru A, Szymanski S, Iwado E, Aoki H, Yokoyama T, Fokt I, Hess K, Conrad C, Madden T, Sawaya R, Kondo S, Priebe W, Kondo Y (2007) A novel inhibitor of the STAT3 pathway induces apoptosis in malignant glioma cells both in vitro and in vivo. Oncogene 26: 2435–2444.

61

P23 CK2-dependent phosphorylation of actin in acute promyelocytic leukemia (APL) cells induced to differentiation by retinoic acid (RA) Kendra Tosoni1,2, Giorgio Arrigoni1,2, Carmela Gurrieri2,3, Francesco Piazza2,3, Giovanni Di Maira1,2, Laura Quotti Tubi2,3, Gianpietro Semenzato2,3, Lorenzo A. Pinna1,2 and Maria Ruzzene1,2 1Department

of Biological Chemistry, 2Venetian Institute of Molecular Medicine (VIMM), and 3Department of Clinical and Experimental Medicine, University of Padova, Padova, Italy e-mail: [email protected] Acute myeloid leukemia (AML) represents a group of hematopoietic cell disorders characterized by the accumulation of non-functional cells termed myeloblasts, due to failure of differentiation and to overproliferation in the stem cell compartment. The specific AML subtype termed APL (acute promyelocytic leukemia) is typically characterized by the translocation t(15;17), which generates the PML-RARα fusion protein; this is an oncogenic molecule, since it recruits corepressors on retinoic acid (RA)-target promoters, causing their silencing and consequent block of differentiation [1]. RA, used at pharmacological doses, is able to revert PML-RARα transcriptional repression, restoring myeloid differentiation of APL blasts [2]; for this reason, the introduction of RA in therapy has increased the overall survival of APL patients. CK2 is a constitutively active and highly pleiotropic Ser/ Thr kinase, usually present in the cells as an heterotetrameric holoenzyme, composed of two catalytic (α and/or α’) and two regulatory (β) subunits [3]; it plays a crucial role in cell survival and proliferation, and is detectable at high levels in normal proliferating tissues and in all analyzed tumors [4]. Consistently, CK2 level is markedly high in proliferating myeloblasts from patients with AML or with chronic myelogenous leukemia in blast crisis, as well as in APL cell lines, while is almost undetectable in normal granulocytes (Piazza et al., unpublished). We found that the pharmacological inhibition of CK2 prevents the differentiation normally triggered by retinoic acid in NB4 cells, an APL human cell line. We therefore started an investigation aimed at elucidating the role of protein kinase CK2 in the differentiation process of APL cells. We analysed CK2 expression and activity in NB4 cells in response to RA. When we treated cells with 1μM RA for different times, and we analysed total, cytosolic and nuclear extract proteins by SDS/PAGE and Western blot, we found that CK2 α, α' and β subunit levels do not change upon RA treatment at any tested time. The endogenous CK2 activity is also unaffected by RA. However, the phosphorylation degree of many endogenous proteins changes upon RA treatment. With the aim to identify CK2 substrates involved in RA-induced differentiation, we focused in particular on those proteins considered putative CK2 substrates, since their phosphorylation level increas-

Abstracts 62

es in response to RA treatment, and decreases when cells, in addition to RA, are treated with K27 or TBB, which are specific inhibitors of CK2 [5]. Among these proteins, by means of 2D electrophoresis and mass spectrometry analysis, we identified a major protein as β-actin, which was never reported as a CK2 substrate before. β-Actin is a highly conserved protein that forms cytoskeletal microfilaments and exists in equilibrium between monomers (globular G-actin) and polymers (filamentous F-actin), but plays many other functions and is also present in the nucleus, where it is supposed to be involved in several processes, such as transcription and chromatin remodelling [6]. Analysing β-actin by Western blot in NB4 cells treated with RA, alone or in combination with CK2 inhibitors, we found that it co-migrates with one of the major CK2-dependent phosphorylated bands evoked by RA treatment; to confirm that endogenous CK2 is actually involved in the phosphorylation of this band, we performed experiments in the presence of different kinase inhibitors (Fig. 1), and we observed that the phosphorylation of this band is affected by CK2 specific inhibitors and not by staurosporine, a promiscuous inhibitor of many kinases but not of CK2 [7]. We analysed β-actin sequence and we found that it contains some CK2-consensus sites (S/T-XX-E/D). Indeed, we demonstrated that in vitro CK2 phosphorylates β-actin immunoprecipitated from NB4 cell lysates, and a commercial recombinant human β-actin, to a stoichiometry of about 0.2 mol per mol. Interestingly, the monomeric CK2α is more active than the tetrameric holoenzyme α2β2 toward β-actin, as occurs only for few CK2 substrates [8]. staurosporine

in vitro(µM)

-

-

0.1

1

10

TBB 5

K27 5

IQA 5

in vivo

-

RA

RA

RA

RA

RA

RA

RA RA+K27

-

β-actin

Figure 1. Endogenous protein phosphorylation with CK2 inhibitors. Cytosolic proteins from NB4 cells treated in vivo for 48 h as indicated (RA was 1 μM, K27 5 μM) were incubated with a radioactive phosphorylation mixture, in the presence of the indicated protein kinase inhibitors. Proteins were separated by SDS/PAGE, and analysed by autoradiography.

RA treatment induces an increase of β-actin level both in the cytosol and in the nucleus; however, inhibition of CK2 has no effect on β-actin amount in the cytosol, but it prevents its increase in the nucleus, suggesting that CK2 activity is involved in the nuclear localization of β-actin (Fig. 2). In summary, our results indicate that CK2 activity is required for the normal differentiating response of APL cells to RA, and that β-actin is a good candidate for mediating this CK2 function, possibly playing a specific role in the nucleus.

NUCLEAR EXTRACT

CYTOSOL -

RA RA+K27

2009

-

RA RA+K27

Autoradiography WB β-actin WB NPM WB α-tubulin

Figure 2. Effects of RA and K27 on actin level and phosphorylation. Proteins (10 μg) from NB4 cells treated in vivo for 48 h as indicated (RA was 1 μM RA, K27 was 5 μM) were incubated with a radioactive phosphorylation mixture. Proteins were separated by SDS/PAGE, and analysed by autoradiography or Western blot (WB), as indicated. α-tubulin or nucleophosmin (NPM) were used as loading controls. References: 1. Steffen B, Müller-Tidow C, Schwäble J, Berdel WE, Serve H (2005) The molecular pathogenesis of acute myeloid leukemia. Crit Rev Oncol Hematol 56: 195–221. 2. Kambhampati S, Verma A, Li Y, Parmar S, Sassano A, Platanias LC (2004) Signalling pathways activated by all-trans-retinoic acid in acute promyelocytic leukemia cells. Leuk Lymphoma 45: 2175–2185. 3. Pinna LA (2002) Protein kinase CH2: a challenge to canons. J Cell Sci 115: 3873–3878. 4. Duncan JS, Litchfield DW (2008) Too much of a good thing: the role of protein kinase CK2 in tumorigenesis and prospects for therapeutic inhibition of CK2. Biochim Biophys Acta 1784: 33–47. 5. Pagano MA, Andrzejewska M, Ruzzene M, Sarno S, Cesaro L, Bain J, Elliott M, Meggio F, Kazimierczuk Z, Pinna LA (2004) Optimization of protein kinase CK2 inhibitors derived from 4,5,6,7tetrabromobenzimidazole. J Med Chem 47: 6239–6247. 6. Vartiainen MK (2008) Nuclear actin dynamics ― from form to function. FEBS Lett 582: 2033–2040. 7. Meggio F, Donella Deana A, Ruzzene M, Brunati AM, Cesaro L, Guerra B, Meyer T, Mett H, Fabbro D, Furet P, Dobrowolska G, Pinna LA (1995) Different susceptibility of protein kinases to staurosporine inhibition. Kinetic studies and molecular bases for the resistance of protein kinase CK2. Eur J Biochem 234: 317–322. 8. Meggio F, Pinna LA (2003) One-thousand-and-one substrates of protein kinase CK2? FASEB J 17: 349–368.


63

P24 Cellular imaging of membranepermeable bisubstrate-analogue inhibitors of protein kinases

Angela Vaasa1, Marje Lust1, Asko Uri1 and Manuela Zaccolo2

A

1Institute

of Chemistry, University of Tartu, Estonia; 2Department of Neuroscience and Molecular Pharmacology, University of Glasgow, UK e-mail: [email protected] Protein kinases (PKs) play a key role in the regulation of protein functions in living cells. More than 400 human diseases have been linked to aberrant protein kinase signalling. This has made PKs important drug targets [1] especially for different forms of cancer. Currently, 10 smallmolecule compounds have been approved and more than 200 compounds are at various stages of clinical development as drugs against different diseases [2]. All drugs on the market and great majority of the compounds in development are targeted to the ATP-binding site of the kinase. Bisubstrate inhibitor design as a logical approach for the development of potent and selective inhibitors of PK has previously gained limited attention due to polar and charged character of these compounds which has been predicted to lead to their restricted cellular uptake and thus low potential for regulation of cellular protein phosphorylation equilibria. We have previously developed inhibitors for basophilic protein serine/threonine kinases with activities in the subnanomolar region [3, 4]. These ARC-type inhibitors comprise analogues of both substrates, an ATP-binding site targeted adenosine mimics and an arginine-rich peptide targeted to the binding domain for the protein to be phosphorylated. In addition to being an intrinsic component of highly potent bisubstrate inhibitors the latter peptide as a representative of transport peptides converts ARC-type compounds plasma membrane permeable [5]. Now we have started a wider study to establish the structural elements of the nucleoside-peptide conjugates that influence their cellular uptake and intracellular localization. Extracellularly applied compounds that contain six or seven d-arginines in the peptide moiety efficiently enter the cell and accumulate in its cytoplasm and nucleus (Fig. 1A). In nucleus these compounds concentrate to special regions, apparently nucleoli. Inhibitors containing two arginine residues stick to the membrane and do not reach the cytoplasm and nucleus of the cell. Cytosolic diffusion and nuclear accumulation of probes is less sensitive to structure of the adenosine mimics. Cellular uptake and localization of the labelled inhibitors is also influenced by the origin of the fluorescent dye. The high affinity of these inhibitors towards PKA was established in the fluorescence anisotropy binding/displacement assay with fluorescence anisotropy detection (KD = 0.3 nM for ARC-1042) [4, 6]. The ability of fluorescently labelled biligand inhibitors (e.g., ARC-1042) to bind to the catalytic subunit of PKA (PKAc) was shown in live cells. In CHO cells expressing PKAc fused with YFP [7] fluorescence resonance energy transfer between

B Figure 1. A. Cellular uptake of biligand inhibtors with six darginine residues. B. FRET changes between YFP of PKAc and tetramethylrhodamine of ARC-1042 in CHO cells.

fluorescent labels of two interacting partners (YFP of PKAc and tetramethylrhodamine of ARC-1042) was measured. Significant increase in FRET was detected between PKAYFP and ARC-1042 after activation of PKA with forskolin that leads to dissociation of PKA-YFP from the holoenzyme and its interaction with ARC-1042 (Fig. 1B). The effect was reversed by the cell-permeable inhibitor of PKA H89 that displaced ARC-1042 from its complex with PKAc. The ability of ARC-type inhibitors to cross cell membrane and bind to the target kinases with high affinity opens a new avenue for the application of bisubstrate inhibitors for regulation of activity of protein kinases in vivo and for the use of bisubstrate inhibitor-based fluorescent probes for cellular high content screening of PK inhibitors. References: 1. Cohen P (2002) Protein kinases: the major drug targets of the twenty-first century? Nat Rev Drug Discov 1: 309–315. 2. Akritopoulou-Zanze I, Hajduk PJ (2009) Kinase-targeted libraries: the design and synthesis of novel, potent, and selective kinase inhibitors. Drug Discov Today 14: 291–297. 3. Enkvist E, Lavogina D, Raidaru G, Vaasa A, Viil I, Lust M, Viht K, Uri A (2006) Conjugation of adenosine and hexa-(D-arginine) leads to a nanomolar bisubstrate-analog inhibitor of basophilic protein kinases. J Med Chem 49: 7150–7159. 4. Lavogina D, Lust M, Viil I, König N, Raidaru G, Rogozina J, Enkvist E, Uri A, Bossemeyer D (2009) Structural analysis of ARC-type inhibitor (ARC-1034) binding to protein kinase A catalytic subunit and rational design of bisubstrate analogue inhibitors of basophilic protein kinases. J Med Chem 52: 308–321. 5. Uri A, Raidaru G, Subbi J, Padari K, Pooga M (2002) Identification of ability of highly charged nanomolar inhibitors of protein kinases to cross plasma membranes and carry a protein into cells. Bioorg Med Chem Lett 12: 2117–2120. 6. Vaasa A, Viil I, Enkvist E, Viht K, Raidaru G, Lavogina D, Uri A (2009) High-affinity bisubstrate probe for fluorescence anisotropy binding/displacement assays with protein kinases PKA and ROCK. Anal Biochem 385: 85–93. 7. Zaccolo M, De G, Cho CY, Feng L, Knapp T, Negulescu PA, Taylor SS, Tsien RY, Pozzan T (2000) A genetically encoded, fluorescent indicator for cyclic AMP in living cells. Nat Cell Biol. 2: 25–29.

Abstracts 64

P25 The RIO kinases: progress in functional characterization and inhibitor design Nicole LaRonde-LeBlanc

Department of Chemistry and Biochemistry, Center for Biomolecular Structure and Organization, University of Maryland, College Park, MD 20742 e-mail:[email protected] The RIO serine kinases are a family of atypical protein kinases that are distinct from canonical protein kinases in their active sites and substrate recognition [1, 2]. Two RIO family members, Rio1 and Rio2, are essential in yeast, and required for ribosome biogenesis [3–5]. Their involvement in ribosome biogenesis, a process fundamental to cell growth and proliferation, makes them attractive targets for the development of inhibitors. These kinases are universally conserved among archaea and eukaryotes and a homolog is present in some bacterial organisms [1]. Humans and other multicellular organisms have a Rio1 and Rio2 homolog, as well as an additional RIO kinase, Rio3. Little is known about the function of the human RIO kinases, known as RIOK1, RIOK2 and RIOK3. As ribosome biogenesis appears to be largely conserved between yeast and humans, it is assumed that RIOK1 and RIOK2 play a role in ribosome processing in humans as well. However, the function of RIOK3 is largely unknown. RIOK1 is reportedly overexpressed in colon cancers and is a target for a Myc-associated transcription factor, MAPJD, that is overexpressed in a majority of non-small cell lung cancers [6, 7]. A recent report has linked RIOK3 with cell motility and migration in pancreas ductal adenocarcinoma through activation of the Rac GTPase pathway and knockdown of RIOK3 results in decreased invasiveness and mobility in pancreatic ductal cells [8]. As a result, there is interest in the development of inhibitors for the human RIO kinases, and it is important to understand the structure and function of these enzymes in order to develop specific inhibitors. Although we have previously reported high-resolution crystal structures for the Rio1 and Rio2 proteins of Archaeoglobus fulgidus no structural information is currently available for the eukaryotic RIO kinases [9, 10]. The archaeal RIO kinase structures have revealed remarkable structural similarity, despite very limited sequence homology, to canonical eukaryotic protein kinases (EPKs), but several features of EPKs are absent in the RIO kinases, resulting in differences in the active site and substrate binding. In order to characterize the eukaryotic RIO kinases, we have expressed and purified recombinant RIOK1 and RIOK3 in Escherichia coli. Both proteins are capable of autophosphorylation, a characteristic that has been observed in all RIO kinases isolated to date. Using mass spectrometry, we have determined the site of autophosphorylation for RIOK1 and RIOK3, and these sites are surprisingly not conserved among the eukaryotic RIO kinases. We also report the interaction of the archaeal Rio1 kinase with the adenosine analog toyocamycin. Toyocamycin,

2009

which has been studied since the mid-1950’s, is long known to be an inhibitor of ribosomal RNA synthesis, and at high levels both species of rRNA are affected [11, 12]. Since toyocamycin is an adenosine analog, it is reasonable to predict that it would get incorporated though the nucleoside metabolic pathways, or inhibit polymerase function. However, inspection of the archaeal Rio1 active site revealed an empty pocket in the back of the ATP binding site that could accommodate polar substituents at the N7 position of the base. Toyocamycin and its derivative, sangivamycin, are essentially N7 substituted adenosine analogs, and were therefore tested for binding to Rio1 in thermal shift assays. Toyocamycin and sangivamycin both produced larger shifts in the melting temperature of Rio1 than the combination of ATP and magnesium, suggesting that these analogs bind more tightly to Rio1 than ATP/Mg2+. This data is in sharp contrast to the testing of several broad spectrum kinase inhibitors, for which no significant changes in melting temperature were observed. We have solved the crystal structure of the Rio1 kinase bound to toyocamycin. The structure shows that the molecule binds exactly as predicted from structural models with a few conformational adjustments in the active site. Further biochemical data shows the effect of toyocamycin on Rio1 kinase activity. Although toyocamycin and sangivamycin are likely to be non-specific in their intracellular targets, they may provide excellent starting points for the synthesis of derivatives that will specifically target Rio1.

Toyocamycin as a Rio1 inhibitor. A. The structure of toyocamycin. B. Electron density in the active site for the structure of Rio1 bound to toyocmycin. Adenosine is modeled in to show difference density (green) for the nitrile group present in toyocamycin.

Finally, our structural and biochemical data on the structure of Rio1 kinase bound to its autophosphorylation site and ATP will be presented. Mutations in the residues involved in contact in the structure diminish autophosphorylation activity, without affecting phosphorylation of generic kinase substrates. This data supports the idea that this structure represents the substrate-bound complex for the Rio1 kinase. This substrate-enzyme complex is very different from that seen for canonical EPKs and will provide the basis for the design of substrate or substrate/ATP mimics for the inhibition of the Rio1 kinase. References: 1. Laronde-Leblanc N, Wlodawer A (2005) A family portrait of the RIO kinases. J Biol Chem 280: 37297–37300.

6th International Conference: Inhibitors of Protein Kinases Vol. 56 2. Laronde-Leblanc N, Wlodawer A (2005) The RIO kinases: an atypical protein kinase family required for ribosome biogenesis and cell cycle progression. Biochim Biophys Acta 1754: 14–24. 3. Geerlings TH, Faber AW, Bister MD, Vos JC, Raue HA (2003) Rio2p, an evolutionarily conserved, low abundant protein kinase essential for processing of 20 S pre-rRNA in Saccharomyces cerevisiae. J Biol Chem 278: 22537–22545. 4. Vanrobays E, Gelugne JP, Gleizes PE, Caizergues-Ferrer M (2003) Late cytoplasmic maturation of the small ribosomal subunit requires RIO proteins in Saccharomyces cerevisiae. Mol Cell Biol 23: 2083–2095. 5. Vanrobays E, Gleizes PE, Bousquet-Antonelli C, Noaillac-Depeyre J, Caizergues-Ferrer M, Gelugne JP (2001) Processing of 20S pre-rRNA to 18S ribosomal RNA in yeast requires Rrp10p, an essential non-ribosomal cytoplasmic protein. EMBO J 20: 4204–4213. 6. Line A, Slucka Z, Stengrevics A, Silina K, Li G, Rees RC (2002) Characterisation of tumour-associated antigens in colon cancer. Cancer Immunol Immunother 51: 574–582. 7. Suzuki C, Takahashi K, Hayama S, Ishikawa N, Kato T, Ito T, Tsuchiya E, Nakamura Y, Daigo Y (2007) Identification of Mycassociated protein with JmjC domain as a novel therapeutic target oncogene for lung cancer. Mol Cancer Therap 6: 542–551. 8. Kimmelman AC, Hezel AF, Aguirre AJ, Zheng H, Paik JH, Ying H, Chu GC, Zhang JX, Sahin E, Yeo G, Ponugoti A, Nabioullin R, Deroo S, Yang S, Wang X, Mcgrath JP, Protopopova M, Ivanova E, Zhang J, Feng B, Tsao MS, Redston M, Protopopov A, Xiao Y, Futreal PA, Hahn, WC, Klimstra DS, Chin L, Depinho RA (2008) Genomic alterations link Rho family of GTPases to the highly invasive phenotype of pancreas cancer. Proc Natl Acad Sci USA 105: 19372–19377. 9. Laronde-Leblanc N, Guszczynski T, Copeland T, Wlodawer A (2005) Structure and activity of the atypical serine kinase Rio1. FEBS J 272: 3698–3713. 10. Laronde-Leblanc N, Guszczynski T, Copeland T, Wlodawer A (2005) Autophosphorylation of Archaeoglobus fulgidus Rio2 and crystal structures of its nucleotide-metal ion complexes. FEBS J 272: 2800–2810. 11. Cohen MB, Glazer RI (1985) Comparison of the cellular and RNA-dependent effects of sangivamycin and toyocamycin in human colon carcinoma cells. Mol Pharmacol 27: 349–355. 12. Cohen MB, Glazer RI (1985) Cytotoxicity and the inhibition of ribosomal RNA processing in human colon carcinoma cells. Mol Pharmacol 27: 308–313.

65