- Email: [email protected]
P21-activated kinase 1 (Pak1) signaling influences therapeutic outcome in pancreatic cancer
Annals of Oncology Advance Access published April 26, 2016 1
p21 activated kinase 1 (Pak1) signaling influences therapeutic outcome in pancreatic cancer
S. Jagadeeshan1,5,#, A. Subramanian1,#, S. Tentu1,#, S. Beesetti1, M. Singhal1, S. Raghavan1, R. P. Surabhi2, J. Mavuluri1, H. Bhoopalan2, J. Biswal6, R. S. Pitani3, S. Chidambaram3, S. Sundaram4, R. Malathi5, J. Jayaraman6, A. S. Nair7, G.
1
Department of Biotechnology, Indian Institute of Technology Madras (IITM), Chennai,
India 2
Department of Human Genetics, Sri Ramachandra University, Porur, Chennai, India
3
Department of CEFT, Sri Ramachandra University, Porur, Chennai, India
4
Department of Pathology, Sri Ramachandra University, Porur, Chennai, India
5
Department of Genetics, University of Madras, Chennai, India
6
Department of Bioinformatics, Alagappa University, Karaikudi, India
7
Rajiv Gandhi Centre for Biotechnology (RGCB), Thiruvananthapuram, Kerala, India
*
To whom the correspondence should be addressed: Dr. Rayala Suresh Kumar,
Department of Biotechnology, IIT MADRAS, Chennai-600036, INDIA. email:[email protected] or Dr.V.Ganesh, Sri Ramachandra University, Chennai, INDIA. e-mail: [email protected] © The Author 2016. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Venkatraman2,* and S. K. Rayala1,*
2 #
Equally contributed to the work
ABSTRACT Background: Therapeutic resistance to Gemcitabine in PDAC is attributed to various cellular mechanisms and signaling molecules that influence as a single factor or in combination.
along with in vivo athymic mouse tumor xenograft models and clinical samples, we demonstrate that Pak1 is a crucial signaling kinase in Gemcitabine resistance. Results: Pak1 kindles resistance via modulation of EMT and activation of pancreatic stellate cells (PSCs). Our results from Gemcitabine resistant and sensitive cell line models showed that elevated Pak1 kinase activity is required to confer Gemcitabine resistance. This was substantiated by elevated levels of phosphorylated Pak1 and RRM1 levels in majority of human PDAC tumors as compared to normal. Delineation of the signaling pathway revealed that Pak1 confers resistance to Gemcitabine by preventing DNA damage, inhibiting apoptosis and regulating survival signals via NF-B. Further, we found that Pak1 is an upstream interacting substrate of Tak1 - a molecule implicated in Gemcitabine resistance. Molecular mechanistic studies revealed that Gemcitabine docks with active site of Pak1, furthermore Gemcitabine treatment induces Pak1 kinase activity both in vivo and in cell free system. Finally, results from athymic mouse tumor models illustrated that Pak1 inhibition by IPA-3 enhances the cytotoxicity of Gemcitabine and brings about pancreatic tumor regression.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Experimental Design: In this study, utilizing in vitro Pak1 OE and KD cell line models
3
Conclusion: To our knowledge, this is the first study illustrating the mechanistic role of Pak1 in causing Gemcitabine resistance via multiple signaling crosstalks and hence Pak1 specific inhibitors will prove to be a better adjuvant with existing chemotherapy modality for PDAC. Key Words: Gemcitabine, Pak1, Desmoplastic, IPA-3, Resistance
Knockdown; PC, Pancreatic Cancer; TAK1, Transforming growth factor beta-Activated Kinase 1; HCAM, Homing Association Cell Adhesion Molecule; RRM1, Ribonucleotide Reductase M1; SMA, Smooth Muscle Actin; I-TASSER, Iterative Threading ASSEmbly Refinement; RMSF, Root Mean Square Fluctuation; RMSD, Root Mean Square Deviation.
Key Message: "Pak1 modulates drug resistance in pancreatic cancer"
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Abbreviations: Pak1, p21 activated kinase 1; WT, Wild Type; OE, Overexpression; KD,
4
INTRODUCTION Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive cancers, with low 5-year survival rate (Wolfgang et al. [1]). Currently, Gemcitabine-based chemotherapy or adjuvant therapy remains the standard form of treatment for patients with advanced PDAC, but exhibits resistance to Gemcitabine over a period of time thus impeding treatment (Wolfgang et al. [1]). Imperative genes contributing to PC
Das et al. [2]) and these pathways important for normal pancreatic development are deregulated in PDAC cells (Cowan and Maitra et al. [3]). More than 90% of PCs harbor oncogenic KRAS mutations at codon 12 (Cowan and Maitra et al. [3]), which results in constitutive activation of the G protein RAS, leading to abnormal activation of several downstream signaling proteins including serine/threonine kinase Pak1, that functions as a downstream activator for various oncogenic signaling pathways (Eswaran et al. [4]). Recently, we reported that Pak1 plays an important role in pancreatic tumorigenesis via transcriptional regulation of fibronectin (Jagadeeshan et al. [5]). Pak1 plays a significant role in Ionizing Radiation (IR) induced DNA damage response pathway via phosphorylating its substrate MORC-2 (Li et al. [6]). A recent study stated that Pak1 promotes epithelial-mesenchymal transition (EMT) and radio-resistance in lung cancer cells (Kim et al. [7]). Hence, the increasing number of scientific reports implicating the role of Pak1 in tumorigenesis, DNA repair and subsequently chemotherapeutic resistance, have converted Pak1 as an attractive therapeutic target for multiple carcinomas, and in turn instigating the development of small-molecule inhibitors. Previous work has shown that IPA-3 is a highly selective inhibitor of Pak1 (Deacon et al.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
progression, encode proteins that regulate signal transduction pathways (Macgregor-
5
[8]). Reports also indicate the preclinical efficacy of IPA-3 in targeting Pak1 in hepatocellular carcinoma model (Wong et al. [9]). The current study focuses on understanding the molecular mechanism by which Pak1 modulation contributes to Gemcitabine resistance in PC.
Methods-Patients :
Immunohistochemistry for phospho-Pak1 expression was evaluated using an indirect immunoperoxidase procedure (ABC-Elite, Vector Laboratories). Statistical analysis was carried out using SPSS software version 20 (IBM Corp. India) Human Pancreatic cancer lysate preparation: Pancreatic cancer tissues with adjacent normal were collected from 12 patients. Lysates were analyzed for phosphoPak1 (Thr212), Pak1, RRM1 and β-actin . Immunocytochemistry was performed on human pancreatic stellate cells were plated on glass cover-slips in 6-well culture plates. After 24 h, cells were rinsed in PBS, and incubated with conditioned media from pancreatic cancer cell line clones for 48 h. After 48 h the cells were washed with PBS and fixed in 4% paraformaldehyde for 15 minutes, followed by a methanol fixation for 10 minutes. Fixed cells were then incubated with αSMA antibody (Biogenex, San Ramon, CA) at 37°C for 30 min followed by staining with anti-mouse HRP-DAB and counter stained with haemtoxylin. Slides were scored by the pathologist using the Quick score method adapted from a protocol previously used (Charafe-Jauffref et al. [2]).
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Immunohistochemistry, Western blotting and Immunocytochemistry:
6
Materials and Methods: Cell Lines: The STR-profiling validated PDAC cell lines MiaPaCa2 and MDAPanc-28 were maintained in DMEM/RPMI medium. BxPC3 cells were provided as gift from Dr Ajay P Singh, Mitchell Cancer Institute, University of Southern Alabama, USA. Gemcitabine resistant (GR) and sensitive (GS) cells were kindly gifted by Dr.Fazlul H. Sarkar (Wayne State University, Detroit, MI).
Medical Center, Houston, Texas, US). pCMV6-Myc-Pak1, pCMV6-Myc-Pak1-T423E and pCMV6-Myc-Pak1-K299R from Addgene, Cambridge, MA. pGEX-GST-Pak1 were cloned from pCMV6-myc-Pak1. pGEX-GST-Tak1 and pCDNA6-V5-Tak1 were cloned from Tak1 plasmid. Generation of Stable clones: Stable Pak1 overexpression and knock down clones were generated as described previously (Jagadeeshan et al. [1]). Determination of cell viability and survival: Cell viability was determined by MTT assay. IC50 was calculated using GraphPad Prism 5 Software. DNA damage detection assays: Determining of DNA strand breaks after Gemcitabine alone/ IPA-3 alone or in combination treatment, alkaline comet assays (single-cell gel electrophoresis) were performed for parental and Pak1 modulated clones. Tail Moment (TM) values of one hundred cells were scored at random per slide using fluorescence microscope with CometScore software (TriTek Corp., Sumerduck, VA).
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Plasmids: Tak1 plasmid from Dr. Jianhua Yang (Bayor College of Medicine, Texas
7
mRNA extraction and cDNA synthesis : mRNA extraction, cDNA synthesis, conventional and quantitative real-time PCR (RT –PCR were performed as previously described (Jagadeeshan et al. [1]). Confocal Microscopy: Cells were plated on glass cover-slips in 6-well culture plates. After 24 h, the cells were rinsed in PBS, fixed in 4% paraformaldehyde for 15 minutes and methanol fixation for 10mins. Fixed cells were blocked with 1%BSA-PBST,
incubated with secondary antibodies conjugated to Alexa Fluor 546 (1:200). The DNA was visualized by counter-staining with DAPI.
In vitro Kinase assay: Kinase assays were performed as described previously (Jagadeeshan et al. [1]). In vitro binding and GST Pull down assays: In vitro translation of Tak1 was performed with TNT T7 Coupled Reticulocyte Lysate System (Promega, Madison, WI, USA) and S35–methionine (BRIT, Mumbai, India) according to the manufacturer’s instructions. Immunoprecipitation assays: 2mg cell lysate was incubated with 2μg of specific antibody overnight at 4°C, to which 30μl of Protein A/G PLUS agarose beads were added and incubated for an hour at 4°C. The pellet was washed in ice-cold Nonidet P-40 buffer (Tris-HCl buffer [50mmol/L (pH 8)] containing 150mmol/L NaCl and 1% Nonidet P-40). The precipitated components were eluted by boiling in SDS loading buffer, resolved by SDS-PAGE, and analyzed by immunoblotting.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
incubated with the indicated antibodies overnight. The cells were washed with PBS,
8
Molecular Modeling and Docking Studies: The 3D model of PAK1 protein generated was validated using SAVS server. Procheck was used to analyse the stereo chemical quality of the protein based on residue by residue approach. The Ramachandran plot for PAK1 protein showed the overall quality factor of the protein. The protein structure modeled using the I-TASSER server was further subjected to molecular dynamics simulation after the structural analysis. In the Molecular Dynamics Simulation, the
cutoff and the smart minimizer algorithms with max steps of 200. The GROMACS 4.6.1 molecular dynamics package and GROMOS 43A6 force field was used to analyze the modeled Pak1 protein stability. The modeled PAK1 and complex simulation was performed using same force field and parameters. The protein system was solvated with SPC216 water model with 1.30 nm cubic box respectively from the molecule and the edge of the box. The lowest-energy minimized protein structure of Pak1 through Molecular Dynamics Simulation was selected for docking studies. The protein was prepared by a multistep process using the Protein Preparation Wizard of the Schrödinger 2013 suite (Schrödinger LLC, New York, USA). The structure of Gemcitabine was retrieved from PubChem database and prepared using LigPrep module (Schrödinger LLC) for docking. In Vivo Mouse experiments and analyses: 4-6-week old nude mice were injected with Mia Pa Ca 2 cells (6 X 10 6) subcutaneously in both the flanks and randomized into four groups (5 per group): Control (DMSO/Saline), IPA-3 (4 mg/kg), Gemcitabine (25mg/kg) and Gemcitabine + IPA-3. IPA-3 was formulated in DMSO and administrated three times weekly (4 mg/kg) i.p and the tumor size was recorded.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
structure was minimized using GROMOS 43A6 force field with electrostatics spherical
9
Statistical analysis: Data are expressed as the mean ± S.E.M and analyzed by Tukey’s test after one way ANOVA using Graph Pad Prism 5 Software (San Diego, CA). Each experiment was repeated thrice. The P values less than 0.05 were considered to denote statistical significance. Refer supplementary materials and methods for a more detailed procedures.
Pak1 modulates EMT markers in PDAC cells: Our previous study (Jagadeeshan et al. [6]) raised the possibility of Pak1 modulating EMT in PC cells. To test this, we checked for typical EMT markers - E-Cadherin and Vimentin in Pak1 stable OE clones of MiaPaCa2 cells and Pak1 KD clones of MDAPanc-28 by Western blotting, qPCR and confocal microscopy. Convincingly, we noticed that E-cadherin expression was down regulated in Pak1 OE clones and conversely, up-regulated in Pak1 KD clones. Correspondingly, the expression of mesenchymal marker Vimentin, was up-regulated in Pak1 OE clones and inversely in Pak1 KD clones (Fig. 1A-C and Supplementary S1AB). Supernatants from Pak1 modulated cells activate human normal Pancreatic Stellate Cells: Since Pak1 modulates EMT markers, we speculated that Pak1 modulated cells may secrete factors which might transform normal PSCs from a quiescent to activated phenotype. We tested this and observed that supernatants
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
RESULTS
10
derived from Pak1 OE clones activated PSCs as revealed by increased cytoskeletal protein α-SMA in comparison to supernatants from Pak1 KD clones (Figure 1D). These results indicate that Pak1 might modulate desmoplastic reaction in PC via activation of PSCs. Pak1 expression and its phosphorylation are associated with Gemcitabine resistance in PC Cells: Since activated PSCs have been shown to be associated with
PDAC cells is related to Gemcitabine resistance by using three cell lines BxPC3, MiaPaCa2 and MDAPanc-28 which expressed different basal levels of Pak1, pPak1 and its kinase activity (Supplementary Figure S2A). Growth inhibition with Gemcitabine was inversely related to the level and activity of Pak1 (Supplementary Figure S2B). Further, to directly test the possibility of whether Pak1 expression is modulated during the process of acquiring resistance, we used a previously characterized Gemcitabine Sensitive (GS) and Resistant (GR) MiaPaCa2 cell lines (Ali et al. [11]). The levels of Pak1, its phosphorylation, kinase activity and IC50 were higher in GR cells as compared to GS (Supplementary Figure S2C and S2D). Consistently, siRNA mediated transient silencing of Pak1 sensitized PC cells to Gemcitabine (Supplementary Figures S3A-B). pPak1 expression in PDAC clinical samples: To investigate whether elevated Pak1 expression levels and its phosphorylation status play a role in acquiring gemcitabine resistance in patients, we evaluated the status of pPak1 (Threonine 212) in 12 paired samples of human PDAC tumors and adjacent normal pancreatic tissues by Western blotting. We observed that pPak1 is significantly up-regulated in 10 out of 12 PDAC samples as compared with adjacent normal tissues (Figure 2A). RRM1, a well-known
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
chemoresistance in PDAC (Haqq et al. [10]), we checked if elevated Pak1 expression in
11
Gemcitabine resistance marker (Nakahira et al. [12]) was also increased in 9 out of 10 pPak1 dysregulated PDAC samples. We also examined pPak1 expression levels in human PDAC tissue microarray containing 20 normal cases and 54 tumor cases by immunohistochemistry (IHC). As anticipated all the normal pancreatic acini (20/20) were negative for pPak1, while 81.4% of PDACs showed positivity for pPak1 (Figure 2B and C). Using the Q scoring data, we performed a 1000- Bootstrap replication analysis to
Additionally ROC curve analysis showed that a pPak1 score of 0.5 could be indicative of malignancy with a sensitivity of 81% and 100% specificity (Figure 2D). These findings suggest that elevated pPak1 level is an important indicator of chemoresistance in PC. Effects of Pak1 modulation on Chemoresistance of PC cells: MTT assay and clonogenic results showed s Pak1 KD clones were sensitized to Gemcitabine (Figure 3A-B and Supplementary Figure S4A-B) and a 10 fold increase in the IC50 value with significant number of colonies in Pak1 OE clones in comparison with the empty vector clone under 0.5% serum condition (Figure 4A-B and Supplementary Figure S5A-B), indicating a strong correlation between an elevated Pak1 protein level and increased chemoresistance to Gemcitabine. Pak1 modulation of DNA damage and apoptosis induced by Gemcitabine: As Gemcitabine is known to induce apoptosis (Bold et al. [13]), we performed nuclear condensation, Annexin V apoptosis assay, TUNEL assay, Comet assay, DNA fragmentation and PARP-1 cleavage upon exposure. Results showed increase in apoptosis in Pak1 KD clones as compared to NT clones and vice versa in Pak1 OE clones as compared to Vector clones (Figure 3C-H and 4C-H, and Supplementary
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
arrive at a mean pPak1 scoring of 2.64 with a range from 2.08-3.32 (95% CI).
12
Figure S4C and S5C). These results suggest the pivotal role of Pak1 in preventing the DNA damage induced upon chemotherapy. IPA-3 mediated Inhibition of Pak1 Kinase Activity induces Chemosensitization of PDAC Cells: To establish Pak1 as a pharmacologically relevant target for the chemosensitization of PC, we first tested the activity of IPA-3 in vitro. Results showed a significant decrease in the Pak1 kinase activity upon treatment of MDAPanc-28 with
and MDAPanc-28 cells to increasing doses of IPA-3 for 72 h showed that IPA-3 IC50 directly correlated with the Pak1 levels of these cell lines (Supplementary Figure S6B). Next we determined the effect of IPA-3 addition on in vitro cytotoxicity of Gemcitabine by treating BxPC3, MiaPaCa2 and MDAPanc-28 cells with increasing doses of Gemcitabine in combination with IPA-3 at doses 2.5-20µM for each cell line. The IPA-3 addition potentiated the cytotoxicity of Gemcitabine in all three cell lines (Figure 5A and Supplementary Figure S6C). Consistent with the above results, IPA-3 significantly primed for the cytotoxicity of Gemcitabine in Pak1 OE system and displayed a 1000 fold reduction in Gemcitabine IC50 value on combination (Figure 5B-G and Supplementary Figure S6D-F).
PAK1 MODULATES GEMCITABINE RESISTANCE via TAK1 Tak1 is a novel direct substrate of Pak1: To gain a deeper insight into the molecular mechanism involved in Pak1 mediated Gemcitabine resistance; we looked at the human kinome for reported kinases involved in Gemcitabine resistance and arrived on Tak1 (MAP3K7) (Melisi et al. [14]). Analysis using the Group based prediction 3.0 (GPS 3.0)
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
IPA-3 (5 - 20µM) (Supplementary Figure S6A). Further, exposing BxPC3, MiaPaCa2
13
software also predicted a Pak1 phosphorylation of Tak1 at serine 412 (S412) (Supplementary Figure S7). It was also noteworthy that Tak1 site identified by GPS software (NGQPRRRSIQDLTVT) is within the potential consensus sequence for Pak1 phosphorylation consensus motif (K/RXXS/T) (Rennefahrt et al. [15]), thus strengthening the hypothesis. Based on the above clues, to directly test this hypothesis, we performed in vitro Pak1 kinase assay using GST-Tak1 as a substrate. The Pak1
expression of pTak1 as well as total Tak1 in Pak1 OE model and vice-versa in Pak1 KD model system (Figure 6B) directly demonstrated the modulation of S412 phosphorylation on Tak1 by Pak1 under physiological conditions. While no significant changes in the mRNA levels of Tak1 with Pak1 modulation was observed (Supplementary Figure S8). We also observed a significant increase in the levels of RRM1 upon Pak1 OE and viceversa with KD clones (Figure 6B) implying that Pak1 modulation could be linked to Gemcitabine resistance. Further, to determine if Pak1 kinase activity was essential for the phosphorylation of Tak1, we treated MDAPanc-28 cell lines with IPA-3 and checked for the phosphorylated and total levels of Tak1. It was observed that Tak1 phosphorylation on S412was reduced by IPA-3 whereas total Tak1 remains unchanged (Figure 6C). This was supported by our observation that co-transfection of kinase active-Pak1 (Pak1-T423E) with V5-Tak1 increased the S412 phosphorylation of Tak1 in 293 cells, where as co-transfection of kinase dead-Pak1 (K299R) did not increase the phosphorylation of Tak1 (Supplementary Figure S9). Collectively, these observations confirmed that Tak1 is a physiological substrate of Pak1 and that S412 on Tak1 is the Pak1 phosphorylation site.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
enzyme efficiently phosphorylated GST-Tak1 (Figure 6A)., A significant increase in the
14
Tak1 interacts with Pak1 in vitro and in vivo: To verify the specificity of the Tak1Pak1 association, we evaluated the ability of in vitro translated Tak1 protein to bind with the Pak1-GST fusion protein. The Pak1-GST fusion protein, efficiently interacted with 35S-labeled full-length Tak1 protein (Figure 6D). The in vivo interaction of the endogenous Tak1 with endogenous Pak1 was confirmed by immunoprecipitation and results showed that Pak1 interacted with Tak1 (Figure 6E).
Pak1: To substantiate Pak1 modulates chemoresistance via Tak1, we inhibited Tak1 activity with 5Z-7-oxozeaenol / Tak1 siRNA in Pak1 OE clones. Results from MTT and clonogenic cell survival assays illustrated that Tak1 inhibition significantly attenuated the chemoresistance induced by Pak1 overexpression (Figure 6F-G; Supplementary Figure S10A-C). Gemcitabine treatment regulates Pak1 kinase activity; evidence from in vivo Pak1 and Cell free system: Recently, it was shown that IR which induces DNA DSBs promotes phosphorylation and activation of Pak1 kinase (Li et al. [6]). Since Gemcitabine also induces DNA DSBs, we hypothesized that Gemcitabine treatment might also regulate phosphorylation and activation of Pak1 kinase. Lysates collected at different time intervals of post-Gemcitabine treatment showed that there is a significant increase in the pPak1 (Thr212) levels and kinase activity by about 60 min (Figure 7A). It was interesting to note that the observed increase in the Pak1 kinase activity by postGemcitabine treatment was abrogated upon treatment with IPA-3 (Figure 7B). Cell free system results from in vitro Pak1 kinase assay done after incubating WT Pak1 with
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Tak1 inhibitor 5Z-7-oxozeaenol attenuates Gemcitabine resistance induced by
15
Gemcitabine at different concentrations showed that there is a significant increase in the Pak1 activity upon incubating with Gemcitabine (Figure7C). Molecular modeling studies reveal Gemcitabine docks at the active site of Pak1 Protein: To further explore the mechanism of how Gemcitabine activates Pak1, an in silico approach was utilized to determine the existence of interaction between Gemcitabine and Pak1. The three dimensional model of Pak1 was constructed using I-
efficiency of ligand to Pak1 protein, Gemcitabine was docked and referenced against multiple known ligands namely; ATP (Adenosine Triphosphate), ANP (Phosphoaminophosphonic acid-adenylate ester), Octahedral Ruthenium Pyridocarbazole, Ruthenium Pyridocarbazole and 1ST (3-Hydroxystaurosporine). The studies reveal the binding affinity of the referenced compounds with glide scores ranging from -10.936 kcal/mol to -6.307 kcal/mol and their energies varied from -65.483 kcal/mol to -40.941 kcal/mol. Gemcitabine docked into the active site of Pak1 with a glide score of -6.271 kcal/mol and a glide energy of -40.333 kcal/mol. The mechanism involves the abstraction and transfer of H-atom from 3’OH group of deoxycytidine to the carbonyl group of Tyr142 at a distance of 1.98 Å. The deviation in the RMSD trajectory for the gemcitabine complexed Pak1 was around 0.2-0.4 nm compared to the 0.8-0.9 nm of native and reference ligand complexed Pak1. The RMSF for the Gemcitabine docked protein is observed within the region of 170-200, as against the region of 60 and 150-160 which is for Pak1. The values of RMSD and RMSF show that the stability is acquired in the kinase domain and this could be favorable to increase the activity of Pak1 by the transfer of phosphoryl groups onto active site residues Serine, Proline,
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
TASSER and it spans about 23 helices and 6β-sheets. To determine the binding
16
Aspartate and Aspargine which are important in the phosphorylation process (Figure 7D and Supplementary Figure S11). Thus, the Pak1 protein is found to be highly stable when bound with Gemcitabine and the interactions favors the activity of Pak1. These results suggest that Gemcitabine promotes Pak1 kinase activation, which in turn might trigger various downstream signaling survival pathways, helping the cells to evade apoptosis and resist chemotherapy.
In vitro evidences from the above results suggested that Pak1 could be a potential target for the chemosensitization of PC in vivo. Results from in vivo mice experiments on MiaPaCa2 tumor xenografts showed that IPA-3 and Gemcitabine alone did not show tumor regression, whereas the combination, showed efficient tumor regression after 4 cycles of therapy, as evident by the tumor volume measurements. Our results showed that, the mean tumor volume was 1518 mm3 for the vehicle control group, 1317 mm3 for the IPA-3, 1364 mm3 in the Gemcitabine group and it was 897 mm3 for the combination therapy group (Figure 7E). Thus our results show that Pak1 inhibition by IPA-3 enhances the cytotoxicity of Gemcitabine and brings about pancreatic tumor regression. The in vivo results substantiate the in vitro data, that inhibiting Pak1 using IPA-3 enhances the susceptibility of pancreatic tumor cells to chemotherapeutic agent like Gemcitabine. DISCUSSION Several studies have established the role of PAK family - especially Pak1 in pancreatic cancer (Baker et al. [16], Radu et al. [17], He and Baldwin [18], Yeo et al. [19], Zhou et al. [20]) signaling. In this study, it is clearly evident that Pak1 modulation regulates EMT
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Pak1 inhibition modulates chemosensitivity in athymic mouse tumor xenografts:
17
and causing significant alteration in the tumor microenvironment leading to activation of PSCs which in turn contributes to Gemcitabine resistance. Our data from GR and GS models clearly demonstrated that elevated Pak1 kinase activity is required to confer Gemcitabine resistance. This was supported by elevated levels of pPak1 and RRM1 levels in majority of clinical PDAC tumors as compared to normal pancreatic tissue. Further, delineation of the signaling pathway revealed that Pak1 modulates DNA
interaction with and phosphorylating Tak1. This was substantiated by the observation that Tak1 inhibitor 5Z-7-oxozeaenol attenuated Gemcitabine resistance induced by Pak1. Further, mechanistic studies revealed that Gemcitabine docks with active site of Pak1 and Gemcitabine treatment induces Pak1 kinase activity. Finally, results from xenograft animal tumor models showed that Pak1 inhibition by IPA-3 enhances the cytotoxicity of Gemcitabine and brings about tumor regression. This study clearly demonstrated the connecting link between Pak1-Tak1-NF-κB-EMT pathways in contributing to Gemcitabine resistance. Pak1 being upstream effector molecule regulates this via multiple effects like DNA damage, apoptosis, EMT and finally the outcome of the drug response. It was recently shown that Glaucarubinone and Gemcitabine synergistically reduce PC growth via down-regulation of Paks (Yeo et al. [21]). Further, it was reported that therapy with anti-Pak1 drugs - PP1/GL-2003 combination has proven to be most effective in human PC xenograft models (Hirokawa et al. [22]). Owing to all this, it is possible that inhibition of Pak1 signaling with more specific small molecule inhibitors could enhance drug sensitivity either alone or with chemotherapy combination and may represent a promising therapeutic strategy for PC.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
damage and apoptosis and also Pak1 kinase activity inhibition sensitizes PC cells via
18
However, further work with orthotropic models in Pak1-knockout mouse and validated Pak1 specific inhibitors are needed to provide more insights into the better therapeutic regimen for PDAC. ACKNOWLEDGEMENTS: We acknowledge the many helpful suggestions and support of the members of our laboratories. We thank Mr. Y R Krishnamoorthy for the generation of Pak1 modulated
Funding: We thank the Department of Biotechnology (DBT), Government of India for the financial support to SKR (grant no.: BT/PR13559/Med/30/283/2010) and Indian Institute of Technology Madras (IITM) for all other facilities. Disclosure: The authors have declared no conflicts of interest.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
clones, Mr. Anuj Kaushik, Ms. Monisha with FACS and Blotting.
19
REFERENCES: 1. Wolfgang CL, Herman JM, Laheru DA, et al. Recent progress in pancreatic cancer. CA Cancer J Clin. 2013; 63(5): 318-348. 2. Macgregor-Das AM, and Iacobuzio-Donahue CA. Molecular pathways in pancreatic carcinogenesis. J Surg Oncol. 2013; 107(1): 8-14. 3. Cowan RW, Maitra A. Genetic progression of pancreatic cancer. Cancer J. 2014;
4. Eswaran J, Li DQ, Shah A, et al. Molecular pathways: targeting p21-activated kinase 1 signaling in cancer--opportunities, challenges, and limitations. Clin Cancer Res. 2012; 18(14): 3743-3749. 5. Jagadeeshan S, Krishnamoorthy YR, Singhal M, et al. Transcriptional regulation of fibronectin by p21-activated kinase-1 modulates pancreatic tumorigenesis. Oncogene. 2015; 34(4): 455-464. 6. Li DQ, Nair SS, Ohshiro K, et al. MORC2 signaling integrates phosphorylationdependent, ATPase-coupled chromatin remodeling during the DNA damage response. Cell Rep. 2012; 2(6): 1657-1669. 7. Kim E, Youn H, Kwon T, et al. PAK1 tyrosine phosphorylation is required to induce epithelial-mesenchymal transition and radioresistance in lung cancer cells. Cancer Res. 2014; 74(19): 5520-5531. 8. Deacon SW, Beeser A, Fukui JA, et al. An isoform-selective, small-molecule inhibitor targets the autoregulatory mechanism of p21-activated kinase. Chem Biol. 2008; 15(4): 322-331.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
20(1): 80-84.
20
9. Wong LL, Lam IP, Wong TY, et al. IPA-3 inhibits the growth of liver cancer cells by suppressing PAK1 and NF-κB activation. PLoS One. 2013; 8(7):e68843. 10. Haqq J, Howells LM, Garcea G, et al. Pancreatic stellate cells and pancreas cancer: current perspectives and future strategies. Eur J Cancer. 2014; 50(15): 2570-2582. 11. Ali S, Ahmad A, Banerjee S, et al. Gemcitabine sensitivity can be induced in
by curcumin or its analogue CDF. Cancer Res. 2010; 70(9):3606-3617. 12. Nakahira S, Nakamori S, Tsujie M, et al. Involvement of ribonucleotidereductase M1 subunit overexpression in gemcitabine resistance of human pancreatic cancer. Int J Cancer. 2007; 120(6): 1355-1363. 13. Bold RJ, Chandra J and McConkey DJ. Gemcitabine-induced programmed cell death (apoptosis) of human pancreatic carcinoma is determined by Bcl-2 content. Ann SurgOncol. 1999; 6(3): 279-285. 14. Melisi D, Xia Q, Paradiso G, et al. Modulation of pancreatic cancer chemoresistance by inhibition of TAK1. J Natl Cancer Inst. 2011; 103(15): 11901204. 15. Rennefahrt UE, Deacon SW, Parker SA, et al. Specificity profiling of Pak kinases allows identification of novel phosphorylation sites. J Biol Chem. 2007; 282(21): 15667-15678. 16. Baker NM, Yee Chow H, Chernoff J, et al. Molecular pathways: targeting RACp21-activated serine-threonine kinase signaling in RAS-driven cancers. Clin Cancer Res. 2014; 20(18): 4740-4746.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
pancreatic cancer cells through modulation of miR-200 and miR-21 expression
21
17. Radu M, Semenova G, Kosoff R, et al. PAK signalling during the development and progression of cancer. Nat Rev Cancer. 2014; 14(1): 13-25. 18. He H and Baldwin GS. p21-activated kinases and gastrointestinal cancer. Biochim Biophys Acta. 2013; 1833(1): 33-39. 19. Yeo D, He H, Baldwin GS, Nikfarjam M. The role of p21-activated kinases in pancreatic cancer. Pancreas. 2015; 44(3): 363-369.
and resistance to MET inhibition. J Pathol. 2014; 234(4): 502-513. 21. Yeo D, Huynh N, Beutler JA, et al. Glaucarubinone and gemcitabine synergistically reduce pancreatic cancer growth via down-regulation of P21activated kinases. Cancer Lett. 2014; 346(2): 264-272. 22. Hirokawa Y, Levitzki A, Lessene G, et al. Signal therapy of human pancreatic cancer and NF1-deficient breast cancer xenograft in mice by a combination of PP1 and GL-2003, anti-PAK1 drugs (Tyr-kinase inhibitors). Cancer Lett. 2007; 245(1-2): 242-251.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
20. Zhou W, Jubb AM, Lyle K, et al. PAK1 mediates pancreatic cancer cell migration
22
FIGURE LEGENDS: Figure 1: Pak1 expression modulates expression of EMT markers. A, B & CWestern blot analysis, qPCR analysis and confocal images for EMT markers in Pak1 modulated clones. Each value represent the mean ± SEM. * p<0.05, ** p<0.005 compared with Vector clones and * p<0.05, *** p<0.001 compared with NT1 clones by Tukey’s test after one way ANOVA. # represents marginal significance. D-
after incubating for 48 h in conditioned media obtained from Pak1 modulated clones. Table showing the Q-score for α-SMA positivity under each condition. Figure 2: Pak1 expression is associated with gemcitabine resistance in PDAC tumor samples. A- Pak1, phospho-Pak1 (pPak1) and RRM1 expression in human PDAC tumors and adjacent normal pancreas by Western blotting. B- Representative immunostained paraffin-embedded tissue sections (100X magnification) of paired normal and PDAC tumor tissue array against pPak1. C-Table showing pPak1 positivity at various grades of PDAC tumor. D- Histogram of ROC curve for sensitivity and specificity of pPak1 in PDAC (95% CI). Figure 3: Pak1 knockdown enhances gemcitabine sensitivity in vitro. A- Table showing differences in IC50 for Gemcitabine in NT and Pak1 KD clones. B, C, D, E & FPercentage survival, nuclear condensed, Annexin V FITC positive, TUNEL positive and Comet tail moment positive MDAPanc-28 Pak1 KD clones. Each value represent the mean ± SEM. * p<0.05, ** p<0.005, *** p<0.001 compared with untreated NT clones by Tukey’s test after one way ANOVA. G- DNA fragmentation analysis in MDAPanc-28 KD
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Immunocytochemical analysis of PSC through α-SMA staining (100X magnification),
23
clones after Gemcitabine treatment for 24 h. H- Western blot showing PARP cleavage in Pak1 KD clones on Gemcitabine treatment. Figure 4: Pak1 overexpression enhances gemcitabine chemoresistance in vitro. A- Table showing differences in IC50 for Gemcitabine in pancreatic cancer cells with stable Pak1 over expression. MTT assay for Pak1 overexpression clones were conducted under 0.5% serum condition. B, C, D, E & F- Percentage survival, nuclear
cells of MiaPaCa2 Pak1 OE clones . Each value represent the mean ± SEM. * p<0.05, ** p<0.005, *** p<0.001 compared with untreated Vector clones by Tukey’s test after one way ANOVA. G- DNA fragmentation analysis in MiaPaCa2 Pak1 OE clones after Gemcitabine treatment for 24 h. H- Western blot showing PARP cleavage in Pak1 overexpressing MiaPaCa2 clones. Figure 5: Inhibition of Pak1 kinase activity and chemo sensitization of human pancreatic cancer cells in vitro. A & B-Table showing differences in IC50 for Gemcitabine in pancreatic cancer cells and Pak1 overexpressed clones. C, D, E & FPercentage survival, nuclear condensed, TUNEL positive and Comet tail moment of MiaPaCa2 Pak1 overexpression clones after treatment with Gemcitabine alone or in combination with IPA-3 at doses corresponding with the 24h IC50 measured for each cell line. Each value represent the mean ± SEM. * p<0.05, ** p<0.005, *** p<0.001 compared with untreated Vector clones by Tukey’s test after one way ANOVA. GWestern blot showing PARP cleavage in Pak1 overexpressing MiaPaCa2 clones on Gemcitabine and IPA-3 combined treatment.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
condensed, Annexin V FITC positive, TUNEL positive and Comet tail moment positive
24
Figure 6: Pak1 interacts with Tak1 and phosphorylatesTak1 at serine residue 412. A- In vitro kinase assay using GST and GST-Tak1 as substrates and purified human Pak1 as an enzyme. The upper panel shows the autoradiogram indicating the phosphorylation of GST-Tak1 and the lower panel shows ponceau-S stained membrane. B- Western blot analysis showing phospho-Tak1, Tak1 and RRM1 in Pak1 modulated clones. C- Western blot analysis showing the Tak1 and pTak1 on IPA - 3
GST-Pak1 constructs. Bound proteins were separated and analyzed by SDS-PAGE and autoradiography. The lower panels are ponceau-S stained blots showing either GST Pak1 and GST. * denotes the presence of the GST-Pak1 fusion proteins. E- Protein extracts made from MDAPanc-28 cells were immunoprecipitated with control IgG or Pak1/ Tak1 antibody and were immunoblotted with the indicated antibodies. The input represents around 10% of the total amount of protein used for the immunoprecipitation. F- Differences in IC50 for Gemcitabine, 5Z-7-Oxozeanol and their combination in pancreatic cancer cells with stable Pak1 over expression. G- Percent survival of Pak1 over expression clones of pancreatic cancer cells after treatment with Gemcitabine/5Z7-Oxozeanol alone or in combination at doses corresponding with the IC50 measured for each cell line or dimethyl sulfoxide (DMSO) as control, for 24 h. Each value represent the mean ± SEM. *** p<0.001 compared with untreated vector clones by Tukey’s test after one way ANOVA. Figure 7: Gemcitabine docks to Pak1 augment Pak1 kinase activity and its inhibition by IPA-3 attenuates tumor growth in vivo. A- Western blot analysis showing pPak1 upregulation and corresponding kinase activity after post-Gemcitabine
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
treatment. D- In vitro translated Tak1 protein (35S-labeled) was incubated with GST or
25
wash. B- MDAPanc-28 cells were treated with/ without IPA-3 (20µM) or Gemcitabine (1µM) or in combination and the lysates were collected and subjected for Pak1 kinase activity using MBP as substrate. C- In vitro kinase assay using MBP as substrate showing the increase in phosphorylation of recombinant Pak1 enzyme in the presence and absence of gemcitabine. D- Gemcitabine docks with Pak1. Modeled PAK1 structure generated by I-Tasser server. Red color indicates α helices, yellow color indicates the
Gemcitabine and the interactions along with distance of residues. See also Figure S11. E- Representative image of MiaPaCa2 cell tumor xenografts. MiaPaCa2 cells were subcutaneously implanted in nude mice. Tumor regression analysis of mice tumors after treatment with Vehicle (DMSO), Gemcitabine, IPA-3 and combination of Gemcitabine and IPA-3 on day 0 and day 24 of treatment. Each value represents the mean ± SEM.* p<0.05 compared with day 0 for respective group by Unpaired t-test. ns represents nonsignificant.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
β-sheets and the green color indicates the loop region. Lig Plot Diagram of PAK1 with
26
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
27
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
28
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
29
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
30
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
31
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
32
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
p21 activated kinase 1 (Pak1) signaling influences therapeutic outcome in pancreatic cancer
S. Jagadeeshan1,5,#, A. Subramanian1,#, S. Tentu1,#, S. Beesetti1, M. Singhal1, S. Raghavan1, R. P. Surabhi2, J. Mavuluri1, H. Bhoopalan2, J. Biswal6, R. S. Pitani3, S. Chidambaram3, S. Sundaram4, R. Malathi5, J. Jayaraman6, A. S. Nair7, G.
1
Department of Biotechnology, Indian Institute of Technology Madras (IITM), Chennai,
India 2
Department of Human Genetics, Sri Ramachandra University, Porur, Chennai, India
3
Department of CEFT, Sri Ramachandra University, Porur, Chennai, India
4
Department of Pathology, Sri Ramachandra University, Porur, Chennai, India
5
Department of Genetics, University of Madras, Chennai, India
6
Department of Bioinformatics, Alagappa University, Karaikudi, India
7
Rajiv Gandhi Centre for Biotechnology (RGCB), Thiruvananthapuram, Kerala, India
*
To whom the correspondence should be addressed: Dr. Rayala Suresh Kumar,
Department of Biotechnology, IIT MADRAS, Chennai-600036, INDIA. email:[email protected] or Dr.V.Ganesh, Sri Ramachandra University, Chennai, INDIA. e-mail: [email protected] © The Author 2016. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Venkatraman2,* and S. K. Rayala1,*
2 #
Equally contributed to the work
ABSTRACT Background: Therapeutic resistance to Gemcitabine in PDAC is attributed to various cellular mechanisms and signaling molecules that influence as a single factor or in combination.
along with in vivo athymic mouse tumor xenograft models and clinical samples, we demonstrate that Pak1 is a crucial signaling kinase in Gemcitabine resistance. Results: Pak1 kindles resistance via modulation of EMT and activation of pancreatic stellate cells (PSCs). Our results from Gemcitabine resistant and sensitive cell line models showed that elevated Pak1 kinase activity is required to confer Gemcitabine resistance. This was substantiated by elevated levels of phosphorylated Pak1 and RRM1 levels in majority of human PDAC tumors as compared to normal. Delineation of the signaling pathway revealed that Pak1 confers resistance to Gemcitabine by preventing DNA damage, inhibiting apoptosis and regulating survival signals via NF-B. Further, we found that Pak1 is an upstream interacting substrate of Tak1 - a molecule implicated in Gemcitabine resistance. Molecular mechanistic studies revealed that Gemcitabine docks with active site of Pak1, furthermore Gemcitabine treatment induces Pak1 kinase activity both in vivo and in cell free system. Finally, results from athymic mouse tumor models illustrated that Pak1 inhibition by IPA-3 enhances the cytotoxicity of Gemcitabine and brings about pancreatic tumor regression.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Experimental Design: In this study, utilizing in vitro Pak1 OE and KD cell line models
3
Conclusion: To our knowledge, this is the first study illustrating the mechanistic role of Pak1 in causing Gemcitabine resistance via multiple signaling crosstalks and hence Pak1 specific inhibitors will prove to be a better adjuvant with existing chemotherapy modality for PDAC. Key Words: Gemcitabine, Pak1, Desmoplastic, IPA-3, Resistance
Knockdown; PC, Pancreatic Cancer; TAK1, Transforming growth factor beta-Activated Kinase 1; HCAM, Homing Association Cell Adhesion Molecule; RRM1, Ribonucleotide Reductase M1; SMA, Smooth Muscle Actin; I-TASSER, Iterative Threading ASSEmbly Refinement; RMSF, Root Mean Square Fluctuation; RMSD, Root Mean Square Deviation.
Key Message: "Pak1 modulates drug resistance in pancreatic cancer"
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Abbreviations: Pak1, p21 activated kinase 1; WT, Wild Type; OE, Overexpression; KD,
4
INTRODUCTION Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive cancers, with low 5-year survival rate (Wolfgang et al. [1]). Currently, Gemcitabine-based chemotherapy or adjuvant therapy remains the standard form of treatment for patients with advanced PDAC, but exhibits resistance to Gemcitabine over a period of time thus impeding treatment (Wolfgang et al. [1]). Imperative genes contributing to PC
Das et al. [2]) and these pathways important for normal pancreatic development are deregulated in PDAC cells (Cowan and Maitra et al. [3]). More than 90% of PCs harbor oncogenic KRAS mutations at codon 12 (Cowan and Maitra et al. [3]), which results in constitutive activation of the G protein RAS, leading to abnormal activation of several downstream signaling proteins including serine/threonine kinase Pak1, that functions as a downstream activator for various oncogenic signaling pathways (Eswaran et al. [4]). Recently, we reported that Pak1 plays an important role in pancreatic tumorigenesis via transcriptional regulation of fibronectin (Jagadeeshan et al. [5]). Pak1 plays a significant role in Ionizing Radiation (IR) induced DNA damage response pathway via phosphorylating its substrate MORC-2 (Li et al. [6]). A recent study stated that Pak1 promotes epithelial-mesenchymal transition (EMT) and radio-resistance in lung cancer cells (Kim et al. [7]). Hence, the increasing number of scientific reports implicating the role of Pak1 in tumorigenesis, DNA repair and subsequently chemotherapeutic resistance, have converted Pak1 as an attractive therapeutic target for multiple carcinomas, and in turn instigating the development of small-molecule inhibitors. Previous work has shown that IPA-3 is a highly selective inhibitor of Pak1 (Deacon et al.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
progression, encode proteins that regulate signal transduction pathways (Macgregor-
5
[8]). Reports also indicate the preclinical efficacy of IPA-3 in targeting Pak1 in hepatocellular carcinoma model (Wong et al. [9]). The current study focuses on understanding the molecular mechanism by which Pak1 modulation contributes to Gemcitabine resistance in PC.
Methods-Patients :
Immunohistochemistry for phospho-Pak1 expression was evaluated using an indirect immunoperoxidase procedure (ABC-Elite, Vector Laboratories). Statistical analysis was carried out using SPSS software version 20 (IBM Corp. India) Human Pancreatic cancer lysate preparation: Pancreatic cancer tissues with adjacent normal were collected from 12 patients. Lysates were analyzed for phosphoPak1 (Thr212), Pak1, RRM1 and β-actin . Immunocytochemistry was performed on human pancreatic stellate cells were plated on glass cover-slips in 6-well culture plates. After 24 h, cells were rinsed in PBS, and incubated with conditioned media from pancreatic cancer cell line clones for 48 h. After 48 h the cells were washed with PBS and fixed in 4% paraformaldehyde for 15 minutes, followed by a methanol fixation for 10 minutes. Fixed cells were then incubated with αSMA antibody (Biogenex, San Ramon, CA) at 37°C for 30 min followed by staining with anti-mouse HRP-DAB and counter stained with haemtoxylin. Slides were scored by the pathologist using the Quick score method adapted from a protocol previously used (Charafe-Jauffref et al. [2]).
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Immunohistochemistry, Western blotting and Immunocytochemistry:
6
Materials and Methods: Cell Lines: The STR-profiling validated PDAC cell lines MiaPaCa2 and MDAPanc-28 were maintained in DMEM/RPMI medium. BxPC3 cells were provided as gift from Dr Ajay P Singh, Mitchell Cancer Institute, University of Southern Alabama, USA. Gemcitabine resistant (GR) and sensitive (GS) cells were kindly gifted by Dr.Fazlul H. Sarkar (Wayne State University, Detroit, MI).
Medical Center, Houston, Texas, US). pCMV6-Myc-Pak1, pCMV6-Myc-Pak1-T423E and pCMV6-Myc-Pak1-K299R from Addgene, Cambridge, MA. pGEX-GST-Pak1 were cloned from pCMV6-myc-Pak1. pGEX-GST-Tak1 and pCDNA6-V5-Tak1 were cloned from Tak1 plasmid. Generation of Stable clones: Stable Pak1 overexpression and knock down clones were generated as described previously (Jagadeeshan et al. [1]). Determination of cell viability and survival: Cell viability was determined by MTT assay. IC50 was calculated using GraphPad Prism 5 Software. DNA damage detection assays: Determining of DNA strand breaks after Gemcitabine alone/ IPA-3 alone or in combination treatment, alkaline comet assays (single-cell gel electrophoresis) were performed for parental and Pak1 modulated clones. Tail Moment (TM) values of one hundred cells were scored at random per slide using fluorescence microscope with CometScore software (TriTek Corp., Sumerduck, VA).
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Plasmids: Tak1 plasmid from Dr. Jianhua Yang (Bayor College of Medicine, Texas
7
mRNA extraction and cDNA synthesis : mRNA extraction, cDNA synthesis, conventional and quantitative real-time PCR (RT –PCR were performed as previously described (Jagadeeshan et al. [1]). Confocal Microscopy: Cells were plated on glass cover-slips in 6-well culture plates. After 24 h, the cells were rinsed in PBS, fixed in 4% paraformaldehyde for 15 minutes and methanol fixation for 10mins. Fixed cells were blocked with 1%BSA-PBST,
incubated with secondary antibodies conjugated to Alexa Fluor 546 (1:200). The DNA was visualized by counter-staining with DAPI.
In vitro Kinase assay: Kinase assays were performed as described previously (Jagadeeshan et al. [1]). In vitro binding and GST Pull down assays: In vitro translation of Tak1 was performed with TNT T7 Coupled Reticulocyte Lysate System (Promega, Madison, WI, USA) and S35–methionine (BRIT, Mumbai, India) according to the manufacturer’s instructions. Immunoprecipitation assays: 2mg cell lysate was incubated with 2μg of specific antibody overnight at 4°C, to which 30μl of Protein A/G PLUS agarose beads were added and incubated for an hour at 4°C. The pellet was washed in ice-cold Nonidet P-40 buffer (Tris-HCl buffer [50mmol/L (pH 8)] containing 150mmol/L NaCl and 1% Nonidet P-40). The precipitated components were eluted by boiling in SDS loading buffer, resolved by SDS-PAGE, and analyzed by immunoblotting.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
incubated with the indicated antibodies overnight. The cells were washed with PBS,
8
Molecular Modeling and Docking Studies: The 3D model of PAK1 protein generated was validated using SAVS server. Procheck was used to analyse the stereo chemical quality of the protein based on residue by residue approach. The Ramachandran plot for PAK1 protein showed the overall quality factor of the protein. The protein structure modeled using the I-TASSER server was further subjected to molecular dynamics simulation after the structural analysis. In the Molecular Dynamics Simulation, the
cutoff and the smart minimizer algorithms with max steps of 200. The GROMACS 4.6.1 molecular dynamics package and GROMOS 43A6 force field was used to analyze the modeled Pak1 protein stability. The modeled PAK1 and complex simulation was performed using same force field and parameters. The protein system was solvated with SPC216 water model with 1.30 nm cubic box respectively from the molecule and the edge of the box. The lowest-energy minimized protein structure of Pak1 through Molecular Dynamics Simulation was selected for docking studies. The protein was prepared by a multistep process using the Protein Preparation Wizard of the Schrödinger 2013 suite (Schrödinger LLC, New York, USA). The structure of Gemcitabine was retrieved from PubChem database and prepared using LigPrep module (Schrödinger LLC) for docking. In Vivo Mouse experiments and analyses: 4-6-week old nude mice were injected with Mia Pa Ca 2 cells (6 X 10 6) subcutaneously in both the flanks and randomized into four groups (5 per group): Control (DMSO/Saline), IPA-3 (4 mg/kg), Gemcitabine (25mg/kg) and Gemcitabine + IPA-3. IPA-3 was formulated in DMSO and administrated three times weekly (4 mg/kg) i.p and the tumor size was recorded.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
structure was minimized using GROMOS 43A6 force field with electrostatics spherical
9
Statistical analysis: Data are expressed as the mean ± S.E.M and analyzed by Tukey’s test after one way ANOVA using Graph Pad Prism 5 Software (San Diego, CA). Each experiment was repeated thrice. The P values less than 0.05 were considered to denote statistical significance. Refer supplementary materials and methods for a more detailed procedures.
Pak1 modulates EMT markers in PDAC cells: Our previous study (Jagadeeshan et al. [6]) raised the possibility of Pak1 modulating EMT in PC cells. To test this, we checked for typical EMT markers - E-Cadherin and Vimentin in Pak1 stable OE clones of MiaPaCa2 cells and Pak1 KD clones of MDAPanc-28 by Western blotting, qPCR and confocal microscopy. Convincingly, we noticed that E-cadherin expression was down regulated in Pak1 OE clones and conversely, up-regulated in Pak1 KD clones. Correspondingly, the expression of mesenchymal marker Vimentin, was up-regulated in Pak1 OE clones and inversely in Pak1 KD clones (Fig. 1A-C and Supplementary S1AB). Supernatants from Pak1 modulated cells activate human normal Pancreatic Stellate Cells: Since Pak1 modulates EMT markers, we speculated that Pak1 modulated cells may secrete factors which might transform normal PSCs from a quiescent to activated phenotype. We tested this and observed that supernatants
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
RESULTS
10
derived from Pak1 OE clones activated PSCs as revealed by increased cytoskeletal protein α-SMA in comparison to supernatants from Pak1 KD clones (Figure 1D). These results indicate that Pak1 might modulate desmoplastic reaction in PC via activation of PSCs. Pak1 expression and its phosphorylation are associated with Gemcitabine resistance in PC Cells: Since activated PSCs have been shown to be associated with
PDAC cells is related to Gemcitabine resistance by using three cell lines BxPC3, MiaPaCa2 and MDAPanc-28 which expressed different basal levels of Pak1, pPak1 and its kinase activity (Supplementary Figure S2A). Growth inhibition with Gemcitabine was inversely related to the level and activity of Pak1 (Supplementary Figure S2B). Further, to directly test the possibility of whether Pak1 expression is modulated during the process of acquiring resistance, we used a previously characterized Gemcitabine Sensitive (GS) and Resistant (GR) MiaPaCa2 cell lines (Ali et al. [11]). The levels of Pak1, its phosphorylation, kinase activity and IC50 were higher in GR cells as compared to GS (Supplementary Figure S2C and S2D). Consistently, siRNA mediated transient silencing of Pak1 sensitized PC cells to Gemcitabine (Supplementary Figures S3A-B). pPak1 expression in PDAC clinical samples: To investigate whether elevated Pak1 expression levels and its phosphorylation status play a role in acquiring gemcitabine resistance in patients, we evaluated the status of pPak1 (Threonine 212) in 12 paired samples of human PDAC tumors and adjacent normal pancreatic tissues by Western blotting. We observed that pPak1 is significantly up-regulated in 10 out of 12 PDAC samples as compared with adjacent normal tissues (Figure 2A). RRM1, a well-known
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
chemoresistance in PDAC (Haqq et al. [10]), we checked if elevated Pak1 expression in
11
Gemcitabine resistance marker (Nakahira et al. [12]) was also increased in 9 out of 10 pPak1 dysregulated PDAC samples. We also examined pPak1 expression levels in human PDAC tissue microarray containing 20 normal cases and 54 tumor cases by immunohistochemistry (IHC). As anticipated all the normal pancreatic acini (20/20) were negative for pPak1, while 81.4% of PDACs showed positivity for pPak1 (Figure 2B and C). Using the Q scoring data, we performed a 1000- Bootstrap replication analysis to
Additionally ROC curve analysis showed that a pPak1 score of 0.5 could be indicative of malignancy with a sensitivity of 81% and 100% specificity (Figure 2D). These findings suggest that elevated pPak1 level is an important indicator of chemoresistance in PC. Effects of Pak1 modulation on Chemoresistance of PC cells: MTT assay and clonogenic results showed s Pak1 KD clones were sensitized to Gemcitabine (Figure 3A-B and Supplementary Figure S4A-B) and a 10 fold increase in the IC50 value with significant number of colonies in Pak1 OE clones in comparison with the empty vector clone under 0.5% serum condition (Figure 4A-B and Supplementary Figure S5A-B), indicating a strong correlation between an elevated Pak1 protein level and increased chemoresistance to Gemcitabine. Pak1 modulation of DNA damage and apoptosis induced by Gemcitabine: As Gemcitabine is known to induce apoptosis (Bold et al. [13]), we performed nuclear condensation, Annexin V apoptosis assay, TUNEL assay, Comet assay, DNA fragmentation and PARP-1 cleavage upon exposure. Results showed increase in apoptosis in Pak1 KD clones as compared to NT clones and vice versa in Pak1 OE clones as compared to Vector clones (Figure 3C-H and 4C-H, and Supplementary
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
arrive at a mean pPak1 scoring of 2.64 with a range from 2.08-3.32 (95% CI).
12
Figure S4C and S5C). These results suggest the pivotal role of Pak1 in preventing the DNA damage induced upon chemotherapy. IPA-3 mediated Inhibition of Pak1 Kinase Activity induces Chemosensitization of PDAC Cells: To establish Pak1 as a pharmacologically relevant target for the chemosensitization of PC, we first tested the activity of IPA-3 in vitro. Results showed a significant decrease in the Pak1 kinase activity upon treatment of MDAPanc-28 with
and MDAPanc-28 cells to increasing doses of IPA-3 for 72 h showed that IPA-3 IC50 directly correlated with the Pak1 levels of these cell lines (Supplementary Figure S6B). Next we determined the effect of IPA-3 addition on in vitro cytotoxicity of Gemcitabine by treating BxPC3, MiaPaCa2 and MDAPanc-28 cells with increasing doses of Gemcitabine in combination with IPA-3 at doses 2.5-20µM for each cell line. The IPA-3 addition potentiated the cytotoxicity of Gemcitabine in all three cell lines (Figure 5A and Supplementary Figure S6C). Consistent with the above results, IPA-3 significantly primed for the cytotoxicity of Gemcitabine in Pak1 OE system and displayed a 1000 fold reduction in Gemcitabine IC50 value on combination (Figure 5B-G and Supplementary Figure S6D-F).
PAK1 MODULATES GEMCITABINE RESISTANCE via TAK1 Tak1 is a novel direct substrate of Pak1: To gain a deeper insight into the molecular mechanism involved in Pak1 mediated Gemcitabine resistance; we looked at the human kinome for reported kinases involved in Gemcitabine resistance and arrived on Tak1 (MAP3K7) (Melisi et al. [14]). Analysis using the Group based prediction 3.0 (GPS 3.0)
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
IPA-3 (5 - 20µM) (Supplementary Figure S6A). Further, exposing BxPC3, MiaPaCa2
13
software also predicted a Pak1 phosphorylation of Tak1 at serine 412 (S412) (Supplementary Figure S7). It was also noteworthy that Tak1 site identified by GPS software (NGQPRRRSIQDLTVT) is within the potential consensus sequence for Pak1 phosphorylation consensus motif (K/RXXS/T) (Rennefahrt et al. [15]), thus strengthening the hypothesis. Based on the above clues, to directly test this hypothesis, we performed in vitro Pak1 kinase assay using GST-Tak1 as a substrate. The Pak1
expression of pTak1 as well as total Tak1 in Pak1 OE model and vice-versa in Pak1 KD model system (Figure 6B) directly demonstrated the modulation of S412 phosphorylation on Tak1 by Pak1 under physiological conditions. While no significant changes in the mRNA levels of Tak1 with Pak1 modulation was observed (Supplementary Figure S8). We also observed a significant increase in the levels of RRM1 upon Pak1 OE and viceversa with KD clones (Figure 6B) implying that Pak1 modulation could be linked to Gemcitabine resistance. Further, to determine if Pak1 kinase activity was essential for the phosphorylation of Tak1, we treated MDAPanc-28 cell lines with IPA-3 and checked for the phosphorylated and total levels of Tak1. It was observed that Tak1 phosphorylation on S412was reduced by IPA-3 whereas total Tak1 remains unchanged (Figure 6C). This was supported by our observation that co-transfection of kinase active-Pak1 (Pak1-T423E) with V5-Tak1 increased the S412 phosphorylation of Tak1 in 293 cells, where as co-transfection of kinase dead-Pak1 (K299R) did not increase the phosphorylation of Tak1 (Supplementary Figure S9). Collectively, these observations confirmed that Tak1 is a physiological substrate of Pak1 and that S412 on Tak1 is the Pak1 phosphorylation site.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
enzyme efficiently phosphorylated GST-Tak1 (Figure 6A)., A significant increase in the
14
Tak1 interacts with Pak1 in vitro and in vivo: To verify the specificity of the Tak1Pak1 association, we evaluated the ability of in vitro translated Tak1 protein to bind with the Pak1-GST fusion protein. The Pak1-GST fusion protein, efficiently interacted with 35S-labeled full-length Tak1 protein (Figure 6D). The in vivo interaction of the endogenous Tak1 with endogenous Pak1 was confirmed by immunoprecipitation and results showed that Pak1 interacted with Tak1 (Figure 6E).
Pak1: To substantiate Pak1 modulates chemoresistance via Tak1, we inhibited Tak1 activity with 5Z-7-oxozeaenol / Tak1 siRNA in Pak1 OE clones. Results from MTT and clonogenic cell survival assays illustrated that Tak1 inhibition significantly attenuated the chemoresistance induced by Pak1 overexpression (Figure 6F-G; Supplementary Figure S10A-C). Gemcitabine treatment regulates Pak1 kinase activity; evidence from in vivo Pak1 and Cell free system: Recently, it was shown that IR which induces DNA DSBs promotes phosphorylation and activation of Pak1 kinase (Li et al. [6]). Since Gemcitabine also induces DNA DSBs, we hypothesized that Gemcitabine treatment might also regulate phosphorylation and activation of Pak1 kinase. Lysates collected at different time intervals of post-Gemcitabine treatment showed that there is a significant increase in the pPak1 (Thr212) levels and kinase activity by about 60 min (Figure 7A). It was interesting to note that the observed increase in the Pak1 kinase activity by postGemcitabine treatment was abrogated upon treatment with IPA-3 (Figure 7B). Cell free system results from in vitro Pak1 kinase assay done after incubating WT Pak1 with
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Tak1 inhibitor 5Z-7-oxozeaenol attenuates Gemcitabine resistance induced by
15
Gemcitabine at different concentrations showed that there is a significant increase in the Pak1 activity upon incubating with Gemcitabine (Figure7C). Molecular modeling studies reveal Gemcitabine docks at the active site of Pak1 Protein: To further explore the mechanism of how Gemcitabine activates Pak1, an in silico approach was utilized to determine the existence of interaction between Gemcitabine and Pak1. The three dimensional model of Pak1 was constructed using I-
efficiency of ligand to Pak1 protein, Gemcitabine was docked and referenced against multiple known ligands namely; ATP (Adenosine Triphosphate), ANP (Phosphoaminophosphonic acid-adenylate ester), Octahedral Ruthenium Pyridocarbazole, Ruthenium Pyridocarbazole and 1ST (3-Hydroxystaurosporine). The studies reveal the binding affinity of the referenced compounds with glide scores ranging from -10.936 kcal/mol to -6.307 kcal/mol and their energies varied from -65.483 kcal/mol to -40.941 kcal/mol. Gemcitabine docked into the active site of Pak1 with a glide score of -6.271 kcal/mol and a glide energy of -40.333 kcal/mol. The mechanism involves the abstraction and transfer of H-atom from 3’OH group of deoxycytidine to the carbonyl group of Tyr142 at a distance of 1.98 Å. The deviation in the RMSD trajectory for the gemcitabine complexed Pak1 was around 0.2-0.4 nm compared to the 0.8-0.9 nm of native and reference ligand complexed Pak1. The RMSF for the Gemcitabine docked protein is observed within the region of 170-200, as against the region of 60 and 150-160 which is for Pak1. The values of RMSD and RMSF show that the stability is acquired in the kinase domain and this could be favorable to increase the activity of Pak1 by the transfer of phosphoryl groups onto active site residues Serine, Proline,
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
TASSER and it spans about 23 helices and 6β-sheets. To determine the binding
16
Aspartate and Aspargine which are important in the phosphorylation process (Figure 7D and Supplementary Figure S11). Thus, the Pak1 protein is found to be highly stable when bound with Gemcitabine and the interactions favors the activity of Pak1. These results suggest that Gemcitabine promotes Pak1 kinase activation, which in turn might trigger various downstream signaling survival pathways, helping the cells to evade apoptosis and resist chemotherapy.
In vitro evidences from the above results suggested that Pak1 could be a potential target for the chemosensitization of PC in vivo. Results from in vivo mice experiments on MiaPaCa2 tumor xenografts showed that IPA-3 and Gemcitabine alone did not show tumor regression, whereas the combination, showed efficient tumor regression after 4 cycles of therapy, as evident by the tumor volume measurements. Our results showed that, the mean tumor volume was 1518 mm3 for the vehicle control group, 1317 mm3 for the IPA-3, 1364 mm3 in the Gemcitabine group and it was 897 mm3 for the combination therapy group (Figure 7E). Thus our results show that Pak1 inhibition by IPA-3 enhances the cytotoxicity of Gemcitabine and brings about pancreatic tumor regression. The in vivo results substantiate the in vitro data, that inhibiting Pak1 using IPA-3 enhances the susceptibility of pancreatic tumor cells to chemotherapeutic agent like Gemcitabine. DISCUSSION Several studies have established the role of PAK family - especially Pak1 in pancreatic cancer (Baker et al. [16], Radu et al. [17], He and Baldwin [18], Yeo et al. [19], Zhou et al. [20]) signaling. In this study, it is clearly evident that Pak1 modulation regulates EMT
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Pak1 inhibition modulates chemosensitivity in athymic mouse tumor xenografts:
17
and causing significant alteration in the tumor microenvironment leading to activation of PSCs which in turn contributes to Gemcitabine resistance. Our data from GR and GS models clearly demonstrated that elevated Pak1 kinase activity is required to confer Gemcitabine resistance. This was supported by elevated levels of pPak1 and RRM1 levels in majority of clinical PDAC tumors as compared to normal pancreatic tissue. Further, delineation of the signaling pathway revealed that Pak1 modulates DNA
interaction with and phosphorylating Tak1. This was substantiated by the observation that Tak1 inhibitor 5Z-7-oxozeaenol attenuated Gemcitabine resistance induced by Pak1. Further, mechanistic studies revealed that Gemcitabine docks with active site of Pak1 and Gemcitabine treatment induces Pak1 kinase activity. Finally, results from xenograft animal tumor models showed that Pak1 inhibition by IPA-3 enhances the cytotoxicity of Gemcitabine and brings about tumor regression. This study clearly demonstrated the connecting link between Pak1-Tak1-NF-κB-EMT pathways in contributing to Gemcitabine resistance. Pak1 being upstream effector molecule regulates this via multiple effects like DNA damage, apoptosis, EMT and finally the outcome of the drug response. It was recently shown that Glaucarubinone and Gemcitabine synergistically reduce PC growth via down-regulation of Paks (Yeo et al. [21]). Further, it was reported that therapy with anti-Pak1 drugs - PP1/GL-2003 combination has proven to be most effective in human PC xenograft models (Hirokawa et al. [22]). Owing to all this, it is possible that inhibition of Pak1 signaling with more specific small molecule inhibitors could enhance drug sensitivity either alone or with chemotherapy combination and may represent a promising therapeutic strategy for PC.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
damage and apoptosis and also Pak1 kinase activity inhibition sensitizes PC cells via
18
However, further work with orthotropic models in Pak1-knockout mouse and validated Pak1 specific inhibitors are needed to provide more insights into the better therapeutic regimen for PDAC. ACKNOWLEDGEMENTS: We acknowledge the many helpful suggestions and support of the members of our laboratories. We thank Mr. Y R Krishnamoorthy for the generation of Pak1 modulated
Funding: We thank the Department of Biotechnology (DBT), Government of India for the financial support to SKR (grant no.: BT/PR13559/Med/30/283/2010) and Indian Institute of Technology Madras (IITM) for all other facilities. Disclosure: The authors have declared no conflicts of interest.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
clones, Mr. Anuj Kaushik, Ms. Monisha with FACS and Blotting.
19
REFERENCES: 1. Wolfgang CL, Herman JM, Laheru DA, et al. Recent progress in pancreatic cancer. CA Cancer J Clin. 2013; 63(5): 318-348. 2. Macgregor-Das AM, and Iacobuzio-Donahue CA. Molecular pathways in pancreatic carcinogenesis. J Surg Oncol. 2013; 107(1): 8-14. 3. Cowan RW, Maitra A. Genetic progression of pancreatic cancer. Cancer J. 2014;
4. Eswaran J, Li DQ, Shah A, et al. Molecular pathways: targeting p21-activated kinase 1 signaling in cancer--opportunities, challenges, and limitations. Clin Cancer Res. 2012; 18(14): 3743-3749. 5. Jagadeeshan S, Krishnamoorthy YR, Singhal M, et al. Transcriptional regulation of fibronectin by p21-activated kinase-1 modulates pancreatic tumorigenesis. Oncogene. 2015; 34(4): 455-464. 6. Li DQ, Nair SS, Ohshiro K, et al. MORC2 signaling integrates phosphorylationdependent, ATPase-coupled chromatin remodeling during the DNA damage response. Cell Rep. 2012; 2(6): 1657-1669. 7. Kim E, Youn H, Kwon T, et al. PAK1 tyrosine phosphorylation is required to induce epithelial-mesenchymal transition and radioresistance in lung cancer cells. Cancer Res. 2014; 74(19): 5520-5531. 8. Deacon SW, Beeser A, Fukui JA, et al. An isoform-selective, small-molecule inhibitor targets the autoregulatory mechanism of p21-activated kinase. Chem Biol. 2008; 15(4): 322-331.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
20(1): 80-84.
20
9. Wong LL, Lam IP, Wong TY, et al. IPA-3 inhibits the growth of liver cancer cells by suppressing PAK1 and NF-κB activation. PLoS One. 2013; 8(7):e68843. 10. Haqq J, Howells LM, Garcea G, et al. Pancreatic stellate cells and pancreas cancer: current perspectives and future strategies. Eur J Cancer. 2014; 50(15): 2570-2582. 11. Ali S, Ahmad A, Banerjee S, et al. Gemcitabine sensitivity can be induced in
by curcumin or its analogue CDF. Cancer Res. 2010; 70(9):3606-3617. 12. Nakahira S, Nakamori S, Tsujie M, et al. Involvement of ribonucleotidereductase M1 subunit overexpression in gemcitabine resistance of human pancreatic cancer. Int J Cancer. 2007; 120(6): 1355-1363. 13. Bold RJ, Chandra J and McConkey DJ. Gemcitabine-induced programmed cell death (apoptosis) of human pancreatic carcinoma is determined by Bcl-2 content. Ann SurgOncol. 1999; 6(3): 279-285. 14. Melisi D, Xia Q, Paradiso G, et al. Modulation of pancreatic cancer chemoresistance by inhibition of TAK1. J Natl Cancer Inst. 2011; 103(15): 11901204. 15. Rennefahrt UE, Deacon SW, Parker SA, et al. Specificity profiling of Pak kinases allows identification of novel phosphorylation sites. J Biol Chem. 2007; 282(21): 15667-15678. 16. Baker NM, Yee Chow H, Chernoff J, et al. Molecular pathways: targeting RACp21-activated serine-threonine kinase signaling in RAS-driven cancers. Clin Cancer Res. 2014; 20(18): 4740-4746.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
pancreatic cancer cells through modulation of miR-200 and miR-21 expression
21
17. Radu M, Semenova G, Kosoff R, et al. PAK signalling during the development and progression of cancer. Nat Rev Cancer. 2014; 14(1): 13-25. 18. He H and Baldwin GS. p21-activated kinases and gastrointestinal cancer. Biochim Biophys Acta. 2013; 1833(1): 33-39. 19. Yeo D, He H, Baldwin GS, Nikfarjam M. The role of p21-activated kinases in pancreatic cancer. Pancreas. 2015; 44(3): 363-369.
and resistance to MET inhibition. J Pathol. 2014; 234(4): 502-513. 21. Yeo D, Huynh N, Beutler JA, et al. Glaucarubinone and gemcitabine synergistically reduce pancreatic cancer growth via down-regulation of P21activated kinases. Cancer Lett. 2014; 346(2): 264-272. 22. Hirokawa Y, Levitzki A, Lessene G, et al. Signal therapy of human pancreatic cancer and NF1-deficient breast cancer xenograft in mice by a combination of PP1 and GL-2003, anti-PAK1 drugs (Tyr-kinase inhibitors). Cancer Lett. 2007; 245(1-2): 242-251.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
20. Zhou W, Jubb AM, Lyle K, et al. PAK1 mediates pancreatic cancer cell migration
22
FIGURE LEGENDS: Figure 1: Pak1 expression modulates expression of EMT markers. A, B & CWestern blot analysis, qPCR analysis and confocal images for EMT markers in Pak1 modulated clones. Each value represent the mean ± SEM. * p<0.05, ** p<0.005 compared with Vector clones and * p<0.05, *** p<0.001 compared with NT1 clones by Tukey’s test after one way ANOVA. # represents marginal significance. D-
after incubating for 48 h in conditioned media obtained from Pak1 modulated clones. Table showing the Q-score for α-SMA positivity under each condition. Figure 2: Pak1 expression is associated with gemcitabine resistance in PDAC tumor samples. A- Pak1, phospho-Pak1 (pPak1) and RRM1 expression in human PDAC tumors and adjacent normal pancreas by Western blotting. B- Representative immunostained paraffin-embedded tissue sections (100X magnification) of paired normal and PDAC tumor tissue array against pPak1. C-Table showing pPak1 positivity at various grades of PDAC tumor. D- Histogram of ROC curve for sensitivity and specificity of pPak1 in PDAC (95% CI). Figure 3: Pak1 knockdown enhances gemcitabine sensitivity in vitro. A- Table showing differences in IC50 for Gemcitabine in NT and Pak1 KD clones. B, C, D, E & FPercentage survival, nuclear condensed, Annexin V FITC positive, TUNEL positive and Comet tail moment positive MDAPanc-28 Pak1 KD clones. Each value represent the mean ± SEM. * p<0.05, ** p<0.005, *** p<0.001 compared with untreated NT clones by Tukey’s test after one way ANOVA. G- DNA fragmentation analysis in MDAPanc-28 KD
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
Immunocytochemical analysis of PSC through α-SMA staining (100X magnification),
23
clones after Gemcitabine treatment for 24 h. H- Western blot showing PARP cleavage in Pak1 KD clones on Gemcitabine treatment. Figure 4: Pak1 overexpression enhances gemcitabine chemoresistance in vitro. A- Table showing differences in IC50 for Gemcitabine in pancreatic cancer cells with stable Pak1 over expression. MTT assay for Pak1 overexpression clones were conducted under 0.5% serum condition. B, C, D, E & F- Percentage survival, nuclear
cells of MiaPaCa2 Pak1 OE clones . Each value represent the mean ± SEM. * p<0.05, ** p<0.005, *** p<0.001 compared with untreated Vector clones by Tukey’s test after one way ANOVA. G- DNA fragmentation analysis in MiaPaCa2 Pak1 OE clones after Gemcitabine treatment for 24 h. H- Western blot showing PARP cleavage in Pak1 overexpressing MiaPaCa2 clones. Figure 5: Inhibition of Pak1 kinase activity and chemo sensitization of human pancreatic cancer cells in vitro. A & B-Table showing differences in IC50 for Gemcitabine in pancreatic cancer cells and Pak1 overexpressed clones. C, D, E & FPercentage survival, nuclear condensed, TUNEL positive and Comet tail moment of MiaPaCa2 Pak1 overexpression clones after treatment with Gemcitabine alone or in combination with IPA-3 at doses corresponding with the 24h IC50 measured for each cell line. Each value represent the mean ± SEM. * p<0.05, ** p<0.005, *** p<0.001 compared with untreated Vector clones by Tukey’s test after one way ANOVA. GWestern blot showing PARP cleavage in Pak1 overexpressing MiaPaCa2 clones on Gemcitabine and IPA-3 combined treatment.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
condensed, Annexin V FITC positive, TUNEL positive and Comet tail moment positive
24
Figure 6: Pak1 interacts with Tak1 and phosphorylatesTak1 at serine residue 412. A- In vitro kinase assay using GST and GST-Tak1 as substrates and purified human Pak1 as an enzyme. The upper panel shows the autoradiogram indicating the phosphorylation of GST-Tak1 and the lower panel shows ponceau-S stained membrane. B- Western blot analysis showing phospho-Tak1, Tak1 and RRM1 in Pak1 modulated clones. C- Western blot analysis showing the Tak1 and pTak1 on IPA - 3
GST-Pak1 constructs. Bound proteins were separated and analyzed by SDS-PAGE and autoradiography. The lower panels are ponceau-S stained blots showing either GST Pak1 and GST. * denotes the presence of the GST-Pak1 fusion proteins. E- Protein extracts made from MDAPanc-28 cells were immunoprecipitated with control IgG or Pak1/ Tak1 antibody and were immunoblotted with the indicated antibodies. The input represents around 10% of the total amount of protein used for the immunoprecipitation. F- Differences in IC50 for Gemcitabine, 5Z-7-Oxozeanol and their combination in pancreatic cancer cells with stable Pak1 over expression. G- Percent survival of Pak1 over expression clones of pancreatic cancer cells after treatment with Gemcitabine/5Z7-Oxozeanol alone or in combination at doses corresponding with the IC50 measured for each cell line or dimethyl sulfoxide (DMSO) as control, for 24 h. Each value represent the mean ± SEM. *** p<0.001 compared with untreated vector clones by Tukey’s test after one way ANOVA. Figure 7: Gemcitabine docks to Pak1 augment Pak1 kinase activity and its inhibition by IPA-3 attenuates tumor growth in vivo. A- Western blot analysis showing pPak1 upregulation and corresponding kinase activity after post-Gemcitabine
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
treatment. D- In vitro translated Tak1 protein (35S-labeled) was incubated with GST or
25
wash. B- MDAPanc-28 cells were treated with/ without IPA-3 (20µM) or Gemcitabine (1µM) or in combination and the lysates were collected and subjected for Pak1 kinase activity using MBP as substrate. C- In vitro kinase assay using MBP as substrate showing the increase in phosphorylation of recombinant Pak1 enzyme in the presence and absence of gemcitabine. D- Gemcitabine docks with Pak1. Modeled PAK1 structure generated by I-Tasser server. Red color indicates α helices, yellow color indicates the
Gemcitabine and the interactions along with distance of residues. See also Figure S11. E- Representative image of MiaPaCa2 cell tumor xenografts. MiaPaCa2 cells were subcutaneously implanted in nude mice. Tumor regression analysis of mice tumors after treatment with Vehicle (DMSO), Gemcitabine, IPA-3 and combination of Gemcitabine and IPA-3 on day 0 and day 24 of treatment. Each value represents the mean ± SEM.* p<0.05 compared with day 0 for respective group by Unpaired t-test. ns represents nonsignificant.
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
β-sheets and the green color indicates the loop region. Lig Plot Diagram of PAK1 with
26
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
27
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
28
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
29
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
30
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
31
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016
32
Downloaded from http://annonc.oxfordjournals.org/ at Laurentian University on April 28, 2016