Protein & Peptide Letters, 2010, 17, 151-163
151
Interaction and Structural Modification of Topoisomerase II by Peptidyl Prolyl Isomerase, pin1: An In Silico Study Rohit Mathur1,2, Shubhanker Suman1, Nicolas Beaume3, Mashook Ali2, Anant N. Bhatt1, Madhu Chopra2, Daman Saluja2, Anil K. Mishra1, Sudhir Chandna1, Pramesh N. Kapoor4 and Bilikere S. Dwarakanath1,* 1
Institute of Nuclear Medicine and Allied Sciences, Brig. S. K. Mazumdar Marg, Delhi- 110054, India; 2Dr B R Ambedkar Center for Biomedical Research, University of Delhi, Delhi-110007, India; 3U-601 INSERM, 9 Quai Moncousu, Nantes Cedex, FR-44093, France; 4Department of Chemistry, University of Delhi, Delhi-110007, India Abstract: The peptidyl prolyl isomerase (Pin1) that catalyzes the isomerization of peptide bonds involving proline and phosphorylated serine/threonine/tyrosine and alters the conformation and differential folding has been implicated in the regulation and function of phosphorylated proteins including mitotic and cell cycle proteins viz. Cdc25c, Bcl2, p53 etc. DNA topoisomerase II is one of the nuclear enzymes that maintain the DNA topology and regulates nuclear transactions like chromatin segregation and mitosis. In the present studies, we have carried out in-silico investigations on the possibilities of pin1 interaction with topo II and its functional regulation. We found ten potential pin1 interacting sites within topo II, which were part of loop and/or low complexity regions except helix at S802 within the catalytic domain. Proline directed phosphorylation was found to be possible at 1354, 1361, 1393 positions by cdk. Change in dihedral angle () to 0 degree at all potential pin1 interacting sites at 575, 602, 802 and 950 for cis conformation of peptide bond introduced significant structural change with higher potential energy. All-cis-topo II structure reveals that potential pin1 sites come closer to each other, perhaps forming a motif, thereby suggesting cooperative phenomenon to maintain higher potential energy conformation. The bio-informatic analysis of topo II showed that multisite interaction of pin1 is possible at all the predicted sites. However, a strong possibility of pin1 interaction exist within c-terminal at 1213, 1247, 1354, 1361, 1393 sites, which may lead to either alterations in localization or modification in the activity and perhaps stability of the enzyme.
Keywords: Pin1, Post-Phosphorylation, Topo II, Interaction prediction. INTRODUCTION DNA topoisomerases are the essential nuclear enzymes required for maintenance of topological state of DNA by passing an intact DNA strand through a transient single (topo I) or double (topo II) strand break, facilitating various nuclear processes like transcription, replication, chromosome segregation and mitotic division [1-4]. Topoisomerases are well known markers of cellular proliferation and their levels elevated in many cancers [5]. Topo I is predominantly required for DNA replication and maintenance of S phase, while topo II is known to regulate the chromosome strand segregation during mitosis [6]. Levels and activity of topo II increases nearly tenfold during mitosis (G2/M phase) whereas, topo II and topoI levels remain essentially similar throughout the cell cycle [3,7,8]. This large increase in the level of topo II during mitosis cannot be reconciled with a two fold increase in transcription and mRNA stability, implying that posttranslational modifications like phosphorylation etc contribute to the increased activity [9-12]. Consistent with this, quantitative increase in the extent of phosphorylation of chicken, mouse and hamster topo II has been observed during progress from the S-phase to G2/M phases of Address correspondence to this author at the Institute of Nuclear Medicine and Allied Sciences, Brig. S. K. Mazumdar Marg, Delhi- 110054, India; Tel: (+91) 11-2391-8838; Fax: (+91) 11-2391-9509; E-mail:
[email protected] 0929-8665/10 $55.00+.00
the cell cycle [13-15]. Further, studies using tryptic phosphopeptide analysis have identified some exclusive sites of phosphorylation during mitosis [16,17]. Treatment with calf intestinal phosphatase reduces the activity of topo II [18], implying that phosphorylation of topo II is indeed essential for its catalytic activity. A number of protein kinases, including casein kinase II [19], protein kinase C [20,21], Ca2+/calmodulin-dependent protein kinase [8,21] and p34cdc2 kinase [8] can phosphorylate topo II in vitro, that generally stimulate the ATPdependent decatenation activity. Protein kinase C has been shown to activate topo activity presumably via increase in hydrolysis of ATP and thereby facilitating the increased turnover of the enzyme [22]. Topo II can associate physically with both protein kinase CK2 (casein kinase II) and Cdc2 kinase [23]. While, CK2 binds to topo II, its interaction is interrupted in early mitosis, when protein kinase CK2 is excluded from the nucleus [24] and at the same time as Cdc2 is recruited to the chromatin. Interestingly, studies with purified enzymes demonstrate that the two kinases compete in a phosphorylation- independent manner during interphase for the binding to topo II [22,25]. These results suggest that the cellular functions of topo II may, at least in part, be mediated by direct competition between the two kinases. The recent demonstration that topo II is required for recruitment of Cdc2 kinase to the nucleus suggests that the Cdc2 © 2010 Bentham Science Publishers Ltd.
152 Protein & Peptide Letters, 2010, Vol. 17, No. 2
kinase/topo II interaction may not only represent the first step in mitotic onset, but also the last step which can be modulated to prolong G2 [26]. Cyclin dependent kinases (Cdks) are proline-directed kinases, which phosphorylate serine or threonine residue preceding a proline, play an important role in the regulation of G2/M phase of cell cycle [27-29]. Proline, an amino acid with nitrogen of amide peptide bond within imidazole ring, plays an important role in recognition of such sites by proline directed kinases and regulation of protein conformation. These kinases play a specific and regulatory role in the maintenance of cell cycle [28,30]. Recently, various computational approaches have been used to analyze the structure and function of various proteins by studying phosphorylation [31,32], binding interactions amongst proteins [33,34] as well as protein and other ligands [35-38], particularly involving cyclin dependent kinases [39]. At the onset of mitosis, which involves the major structural rearrangement during cell division, phosphoepitopes like MPM-2 are generated followed by activation of mitotic kinases like Cdc2, Plk-1 and NIMA [40-42]. These kinases organize the temporal regulation of mitosis by sequential phosphorylation but how the structural changes are brought about by these kinases remains to be elucidated. Proline residues introduce a backbone switch into a polypeptide chain by rate limiting cis/trans isomerization about the prolyl bond [43,44]. Recently, peptidyl prolyl isomerase (Pin1) protein has been unraveled that catalyzes the cis/trans isomerization in vivo. Pin1 specifically binds and effectively catalyzes the prolyl isomerization of phosphorylated Ser/Thr-Pro motifs that are present in mitosis specific phosphoproteins [45-47]. Pin1 was first isolated as NIMA interacting protein in yeast two hybrid screen [48]. Therefore, Pin1 is believed to regulate cell cycle progression via a novel mechanism that is sequence specific and involves phosphorylation dependent prolyl isomerization [45], which corroborates well with the increase in its level during G2/M phase of the cell cycle [48]. Conformational changes brought about by interaction with pin1 have profound effects on the functional status of pin1 substrates by way of modulating catalytic activity, phosphorylation status, protein-protein interactions, subcellular localization and protein stability [49]. Interestingly and notably, Pin1 has been reported to be highly overexpressed in a number of human malignancies including breast and prostate cancers [50-52]. Most importantly, Pin1 enhances several oncogenic signaling pathways and facilitate both cellular proliferation and oncogenesis [46,53-55]. Topo II is deeply involved in the maintainance of DNA topology and is specifically activated to resolve various structural complications. Topo II has also been found to interact with HDAC1 and HDAC2 and has a role in chromatin remodeling [56-58]. Although certain biochemical evidences [59,60] point towards a higher degree of posttranslational modifications of topo II, perhaps involving post-phosphorylation events like pin1 mediated isomerization, underlying its activity, an unequivocal demonstration and its functional relevance is still lacking. We have initiated systematic studies involving in-silico analysis, proteinprotein interactions in vitro as well as cellular responses to
Mathur et al.
pin1 inhibitors and topo II inhibitors. Various in-silico methods such as protein 3D structure prediction [61-63], molecular docking [64], molecular packing [65,66], pharmacophore modeling [67], Mote Carlo simulated annealing approach [68], diffusion-controlled reaction simulation [69], protein subcellular location prediction [70,71] have been in use for rational drug designing. In the present study we have retrieved 3-D structures from PDB database and then simulated the docking of both the proteins under different conditions using Hex 5.1 software, which utilized Monte Carlo simulations and analyzed the site specific interaction of topo II with pin1. Moreover, we have also examined the possibility and relevance of structural and functional regulation of topo II by Pin1. MATERIAL AND METHODS Multiple Sequence Alignment Conservation of topo II sequence as well as potential pin1 interacting sites among various species was studied using multiple sequence alignment. A list of all evolutionary related proteins was created. Detailed full length sequences for all putative proteins were retrieved using Expasy server (http://www.expasy.ch). All these sequences were subjected to Multiple sequence alignment. Multiple sequence alignment was performed using T-Coffee, an improved CLUSTAL-W program [72]. A gap open penalty of 10 and gap extension penalty of 0.2 was used for multiple sequence alignment. Phylogenetic tree was also generated, 1000 cycles were run to get the accurate prediction. Binding of Consensus Sequence All the potential pin1 interacting sites were matched for the similarity with the best model substrate sequence (derived from wet lab experiments), using EMBOSS. A gap penalty of 10 and extend penalty of 0.5 was set for local sequence alignment tool WATER [73]. Local sequence alignment predicts similarity in sequences rather than weighted amino acid preferences generally used. Prediction of Specific Kinase Induced Phosphorylation Prediction for serine, threonine and tyrosine phosphorylation sites was carried out using Net Phos 2.0 server (http://www.cbs.dtu.dk/services/NetPhos) [74]. While, Prediction of phosphorylating kinase involved for phosphorylation at specific sites within topo II was made using scansite.mit.edu server (http://www.scansite.mit.edu) [75]. Topo II protein sequence was given as input and default settings were used to predict the site specific phosphorylation score and probability of respective kinase phosphorylating them. Accessibility Surface accessibility of potential pin1 interacting sites within topo II was estimated using WESA (weighed ensemble solvent accessibility) algorithm [76]. Final output of wesa is the average output from the five different methods BS, MLR, DT, NN and SVM. Wesa output is given in terms of a score of 0 or 1, which referred to Not accessible (0) and accessible (1) respectively. All these methods employed used primary sequence for prediction.
Interaction of Topoisomerase II and Pin1
Structural Analysis of Topo II and its Alteration Secondary structure prediction was performed using Psipred server [77]. Topo II protein sequence was given as input and default settings were used to predict the position specific prediction of secondary structure. Tertiary structure of full length topo II is not yet elucidated. X-ray structure of N-terminal sequence (29-430; Pdb: 1zxm) of human topo II as well as theoretical structure of catalytic domain sequence (402-1200; Pdb: 1lwz) based on X-ray structure of yeast topo II (402-1200; Pdb: 1bjt) has been elucidated. Possible structural alteration induced by cis-trans isomerization by pin1 were simulated by alteration of dihedral angle (). Dihedral angle was modified and was set to 0 degree for cis conformation. Energy minimization was performed using GROMOS linked in spdvb for cis conformer and stability of structure was evaluated. Further, docking of pin1 with topo II was performed using HEX5.1 software [78] using solid surface docking simulations. In docking calculations, each molecule is modeled using 3D expansions of real orthogonal spherical polar basis functions to encode both surface shape and electrostatic charge and potential distributions. All docking experiments were replicated 1000 times and best Etotal for the interaction was taken as key parameter to evaluate the significance of the interaction. Etotal was represented in terms of KJ/mol. RESULTS Potential Sites of Interaction within Topo II with pin1 Pin1 protein consists of two domains, a ww domain and a PPIase or rotamase domain [79,80]. Pin1 interaction with many proteins analysed earlier have shown that ww domain is required for binding of pin1 to its substrate proteins, while rotamase domain is required for cis-trans isomerization of the peptide bond. Since ww domain of pin1 recognizes substrates with specific sequences containing phosphorylated Ser/Thr/Tyr immediately preceding proline [47,81], we searched for such sites in topo II . The primary amino acid sequence of topo II retrieved from Expasy server (www.expasy.ch) swiss prot no. p11388 is composed of 1531 amino acids. There are 61 serine, 25 threonine and 12 tyrosine with a total of 98 sites, retrieved from Netphos 2.0 server. Each of these sites is significant in the regulation of activity and function of topo II. Amongst these, ten sites (nearly 10 percent of total sites) are found to contain the minimum consensus sequence i.e. pS/pT-P for interaction with pin1 (Fig. 1a). While four (40 percent of potential pin1 interactive sites) such sites were found in DNA binding catalytic domain, five sites (50 percent of potential pin1 interactive sites) were identified in c-terminal regulatory domain in addition to a single site in the beginning of the (10 percent of potential pin1 interactive sites) Nterminal of topo II. This data suggest maximum potential pin1 interacting sites in c-terminal of topo II, thereby implicating its role in structural modification and regulation of topo II activity. Conserved Sequences To understand the regulation of topo II and emergence of pin1 in regulating activity of topo II during evolution, we
Protein & Peptide Letters, 2010, Vol. 17, No. 2
153
compared the primary sequences of topo II among various species by performing multiple sequence alignment using Tcoffee, an improved clustal-w algorithm. Catalytic domain is highly conserved; however c-terminal is non-conserved among different species (Fig. 1b). Sequences among the ten predicted pin1 interaction sites in topo II, were similar only in closely related mammalian species like pig, mouse and rat (Fig. 1b), which correlates with the phylogenetic distance from human during evolution (Fig. 1c). Two potential pin1 binding sites at T575 and S802 within catalytic domain were found to be conserved, however rest of the potential pin1 interacting sites were non-conserved among different species including yeast, pea, zebra fish, drosophila, C. elegans etc. Most of the potential pin1 interacting sites are in c-terminal region, which is non-conserved and perhaps has evolved during evolution in species specific manner (Fig. 1b, c). Binding of pin1 Consensus Sequence Structure of pin1 binding pocket of WW-domain reveals a basic cluster near the anionic binding site for binding to phosphorylated serine. Therefore, pin1 binding prefers acidic amino acids just preceding proline. To identify the best possible potential pin1 binding site among the ten identified sites containing the essential pin1 binding motif, we used EMBOSS with local sequence alignment algorithm WATER [73]. Apart from the essential sequence requirement for interaction i.e. phosphorylated-Ser followed by proline, a better binding consensus sequence (WFYS*PR) has also been worked out using a library screening of phosphorylated peptides [45]. The best model peptide substrate WFYS*PR determined experimentally was aligned with each potential interacting site in topo II. Perfectly matched amino acid contacts were referred as identity, while amino acids of similar type (acidic/ basic) were referred to similarity. A gap penalty of 10 as well as extend penalty of 0.5 was set to obtain the score for alignment. Site at S802 as well as S1213 had the highest score of 16 representing the best possible match (Table 1). EMBOSS analysis revealed that sequences with phosphorylation at T575, S802 and S1213 (nearly 30 percent of potential pin1 binding sites) matched best and could be probable binding sites for pin1. Analysis of Accessibility of Potential pin1 Interacting Sites in Topo II Accessibility of potential pin1 interacting sites to various kinases as well as to different proteins is a limiting factor for functional interaction of proteins. Therefore, we analyzed all the predicted sites of pin1 interaction in topo II for their accessibility to kinases responsible for phosphorylation and interaction with other proteins. Prediction was performed using WESA algorithm [76] and a score for accessibility was obtained (Table 2). All the potential pin1 interacting sites in topo II were fairly accessible to kinases except at T575 and S802 positions (Table 2). These prediction analyses correlated with Scansite prediction as well as phosphorylation potential as predicted from Netphos server.
154 Protein & Peptide Letters, 2010, Vol. 17, No. 2
Mathur et al.
Figure 1. a) Domain structure analysis of topos II showing the potential pin1 interaction phosphorylation sites. b) Multiple sequence alignment of topo II sequences present in various species using T-coffee, an improved clustal-w algoritm. Colour representation with dark red representing the conserved vary in gradation to dark purple representing the least conserved region. MSA has been shown for potential pin1 interacting sites only. c) Phylogenetic tree of topo II representing its distance from various species during evolution.
Interaction of Topoisomerase II and Pin1
Table 1.
Protein & Peptide Letters, 2010, Vol. 17, No. 2
Pairwise Sequence Alignment with pin1 Binding Model Sequence WFYS*PR Using EMBOSS with Local Sequence Alignment Algorithm Water. Gap Penalty=10, Extend Penalty=0.5, Matrix- Blossum 62
Position
Sequence
Match
Identity
Similarity
Gaps
Score
4
MEVSPLQP
SP
2
2
0
11
575
EEFITPIVK
F-TP
2
3
0
13
602
WKSSTPNHK
TP
1
2
0
8
802
KDSASPRYI
SPR
3
3
0
16
950
GTEKTPPLI
TP
1
2
0
8
1213
EVLPSPRYI
SPR
3
3
0
16
1247
NTEGSPQED
SPQ
2
3
0
12
1354
PSDASPPKT
SP
2
2
0
11
1361
KTKTSPKLS
SPK
2
3
0
13
1393
PLSSSPPAT
SP
2
2
0
11
Table 2.
155
Prediction of Surface Accessibility of pin1 to its Potential Interacting Sites in Topo II Using WESA Algorithm, Which Calculate the Possibility of Burial or Exposed Amino Acid Based on Score from Five Different Methods by BS, MLR, DT, NN, SVM. Accessible (A)/ Non-Accessible (NA) were Predicted Based on the WESA Score
Position
Amino Acid
BS
MLR
DT
NN
SVM
WESA
4
Serine
A
A
A
A
A
1
A
5
Proline
NA
A
A
A
A
1
A
575
Threonine
NA
NA
A
NA
NA
0
NA
576
Proline
NA
A
NA
NA
NA
0
NA
602
Threonine
A
A
A
A
A
1
A
603
Proline
A
A
A
A
A
1
A
802
Serine
NA
NA
NA
NA
NA
0
NA
803
Proline
NA
NA
NA
NA
NA
0
NA
950
Threonine
A
A
A
A
A
1
A
951
Proline
A
A
A
A
A
1
A
1213
Serine
A
A
A
A
A
1
A
1214
Proline
A
A
A
A
A
1
A
1247
Serine
A
A
A
A
A
1
A
1248
Proline
A
A
A
A
A
1
A
1354
Serine
A
A
A
A
A
1
A
1355
Proline
A
A
A
A
A
1
A
1361
Serine
A
A
A
A
A
1
A
1362
Proline
A
A
A
A
A
1
A
1393
Serine
A
A
A
A
A
1
A
1394
Proline
A
A
A
A
A
1
A
156 Protein & Peptide Letters, 2010, Vol. 17, No. 2
Mathur et al.
Prediction of Proline Directed Phosphorylation of Topo II Since phosphorylation at Ser/Thr preceding proline is essential for interaction, we investigated the potential of various kinases and probably regulating aspect of site specific phosphorylation in maintenance of active conformation of the enzyme. Various domains highlighting the sites of phosphorylation, which may interact with Pin1 has been illustrated (Fig. 1). Possibility of phosphorylation by various kinases was determined by motif scan at high (p
