A novel small molecule inhibitor of protein kinase D blocks pancreatic cancer growth in vitro and in vivo

Abstract

Protein kinase D (PKD) family members are increasingly implicated in multiple normal and abnormal biological functions, including signaling pathways that promote mitogenesis in pancreatic cancer (PaCa). However, nothing is known about the effects of targeting PKD in PaCa. Our PKD-inhibitor discovery program identified CRT0066101 as a specific inhibitor of all PKD isoforms. The aim of our study was to determine the effects of CRT0066101 in PaCa. Initially, we showed that autophosphorylated PKD1 and PKD2 (activated PKD1/2) are significantly upregulated in PaCa and that PKD1/2 are expressed in multiple PaCa cell-lines. Using Panc-1 as a model system, we demonstrated that CRT0066101 reduced BrdU incorporation, increased apoptosis, blocked neurotensin (NT)-induced PKD1/2 activation, reduced NT-induced PKD-mediated Hsp27 phosphorylation, attenuated PKD1-mediated NF-κB activation, and abrogated expression of NF-κB-dependent-dependent proliferative and pro-survival proteins. We showed that CRT0066101 given orally (80 mg/kg/day) for 28 days significantly abrogated PaCa growth in Panc-1 subcutaneous xenograft model. Activated PKD1/2 expression in the treated tumor-explants was significantly inhibited with peak tumor concentration (12 µM) of CRT0066101 achieved within 2 h after oral administration. Further, we showed that CRT0066101 given orally (80 mg/kg/day) for 21 days in Panc-1 orthotopic model potently blocked tumor growth in vivo. CRT0066101 significantly reduced Ki-67⁺ proliferation index (p< 0.01), increased TUNEL⁺ apoptotic cells (p<0.05), and abrogated expression of NF-κB-dependent proteins including cyclin D1, survivin, and cIAP-1. Our results demonstrate for the first time that a PKD-specific small molecule inhibitor CRT0066101 blocks PaCa growth in vivo and show that PKD is a novel therapeutic target in PaCa.

INTRODUCTION

Protein kinase D (PKD) is a new family of serine/threonine kinases comprised of PKD1, PKD2, and PKD3 and is characterized by distinct structural features and enzymological properties [reviewed in (1)]. PKD family members are known effectors of diacylglycerol-regulated signal transduction pathways and are activated by activation loop phosphorylation through protein kinase C (PKC)-dependent (1, 2) and PKC-independent (3) pathways. We, and others, showed that activated PKD1 and PKD2 autophosphorylate at the C-terminal end, corresponding to Ser-910 in human PKD1 (Ser-916 in murine PKD1) and Ser-876 in human PKD2 (1). Accumulating evidence suggests that PKD family members play a critical role in the regulation of several cellular processes and activities, including chromatin organization, Golgi function, gene expression, cell survival, adhesion, motility, differentiation, DNA synthesis and proliferation [reviewed in (1)]. PKD1 activation also initiates the NF-κB signaling pathway, triggering cell survival responses (4). Over-expression of PKD1 or PKD2 enhanced cell cycle progression and DNA synthesis in Swiss 3T3 fibroblasts (5). PKD is implicated in multiple pathological conditions including regulation of cardiac gene expression and contractility (6). Consequently, the development of specific PKD family inhibitors would be useful for defining the physiological roles of PKD as well as for developing novel therapeutic approaches in a variety of pathological conditions.

Neuropeptides, including neurotensin (NT), and growth factors promote activation of PKD family members in multiple neoplasias including pancreatic cancer (PaCa), a devastating disease with an overall 5-year survival rate of only 3–5% (7, 8). We showed that G protein-coupled receptor (GPCR) agonists including NT stimulated PKD-dependent mitogenic signaling pathways in PaCa (9) and more recently that PKD1 over-expression facilitated DNA synthesis and proliferation in PaCa cells (10). PKD1 significantly induced resistance to CD95-dependent apoptosis (11) and phosphorylated Hsp27 in PaCa (12), which is implicated in drug resistance in these cells (13). PKD also plays a potential role in cancer cell invasion and motility (14) and is necessary for tumor-associated angiogenesis (2). As PKD plays a crucial role in tumorigenesis including PaCa, we initiated a PKD inhibitor discovery program to further unravel its biological functions.

Here, we describe anti-tumor activities of a small molecule PKD family specific inhibitor CRT0066101 in PaCa. We showed that activated PKD1/2 (i.e. autophosphorylated at the C-terminal end) are over-expressed in PaCa as compared to normal pancreatic ducts and that these PKD family members are also abundantly expressed in multiple PaCa cell lines as compared with immortalized human pancreatic duct epithelial (HPDE) cell line. Using Panc-1 cells as our model system, we demonstrated that CRT0066101 significantly blocked proliferation, induced apoptosis, reduced NT-induced PKD1/2 activation, abrogated activation of PKD1/2-induced NF-κB, and blocked NF-κB-dependent gene products essential for cell proliferation and survival. Further, CRT0066101 blocked Panc-1 cell proliferation and growth in multiple xenograft models. CRT0066101 reduced proliferation index (Ki-67⁺ cells), increased apoptosis (TUNEL⁺ cells), and abrogated expression of several NF-κB dependent pro-survival proteins in tumor explants. Our results showed that CRT0066101 is a novel PKD-specific inhibitor that blocks PaCa growth both in vitro and in vivo.

MATERIALS AND METHODS

Please see Supplementary Materials and Methods for further details.

Cell cultures and reagents

PaCa cell lines including Panc-1 were obtained either from ATCC (American Type Culture Collection, Manassas, VA) or from Cancer Research UK (CR-UK), London, UK. They were cultured either in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), 100 units/mL penicillin, and 100 µg/mL streptomycin or in Dulbecco’s modified Eagle’s medium (DMEM) from CR-UK (London, UK) supplemented with 10% fetal calf serum (PAA, Pasching, Austria). The human pancreatic duct epithelial (HPDE) cells were generous gifts from Dr. Ming-Sound Tsao (University of Toronto, Ontario, Canada) (15, 16). These cells were cultured in keratinocyte serum-free (KSF) medium supplied with 5 ng/mL epidermal growth factor (EGF) and 50 µg/mL bovine pituitary extract (Invitrogen, Carlsbad, CA). Cells were regularly tested for Mycoplasma and were found to be negative. Antibodies to Hsp27, pS82-Hsp27, pS152/156-MARCKS, and pS916-PKD1/2 antibodies were purchased from Cell Signaling Technology (Danvers, MA). Survivin and β-actin antibodies were obtained from R&D Systems (Minneapolis, MN) and Sigma-Aldrich (St. Louis, MO), respectively. Antibodies to PKD-1/2 (total), cyclin D1, cIAP1, Bcl-xL, and Bcl-2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).

Immunohistochemistry (IHC)

Formalin-fixed paraffin-embedded PaCa tissue micro-arrays (US Biomax, Rockville, MD) were stained with monoclonal pS916-PKD1/2 antibody (Epitomics, Burlingame, CA) at 1:10 dilution for overnight at 4⁰C as previously described (17). This monoclonal antibody specifically recognizes activated PKD1/2.

Biochemical IC₅₀ determination and specificity

Specificity of CRT0066101 was performed by in vitro kinase assays using a commercial kinase profiling service (Millipore UK Ltd, Herts, UK). The IC₅₀ of CRT0066101 on PKD1-3 was determined in vitro using IMAP (Immobilized metal ion affinity-based fluorescence polarization; MDS Analytical Technologies Ltd, Berks, UK).

FACE assay

The Fast Activated Cell-based ELISA (FACE) assay (Active motif, Rixensart, Belgium) was used to measure effects of CRT0066101 on pS916-PKD expression in vitro.

BrdU incorporation assay

Cell proliferation was measured using BrdU proliferation kits (Thermo Scientific Cellomics Europe, Reading, UK) in accordance with manufacturer’s instructions.

Apoptosis assay

The Meso Scale Discovery (MSD) and the Multiplex Apoptosis Panel kit were used to determine levels of cleaved caspase-3 (MSD, Gaithersburg, Maryland).

Cell viability assay

Cell viability was measured using the CellTiter Aqueous One Solution Cell Proliferation Assay kit (Promega, Madison, WI) as described previously (17).

Electrophoretic mobility shift assay

To determine activation of NF-κB, electrophoretic mobility shift assay (EMSA) was performed using nuclear extracts (NE) of Panc-1 cells either transfected with control (Panc-1) or PKD-1 over-expressing (Panc-1-PKD1) vectors prior to treatment with DMSO or CRT0066101 as described previously (18).

NF-κB-luciferase reporter assay

1×10⁵ Panc-1 cells were plated in 24-well plates 16 h prior to transfection. A total of 0.76 µg of pGL3-Luc-NF-κB or pGL3-Luc DNA was then co-transfected with 0.4 µg of Renilla luciferase vector DNA (internal control, Promega, Madison, WI) into Panc-1 cells using lipofectamine 2000 according to the manufacturer’s instruction (Invitrogen, Carlsbad, CA). Luminescence values were normalized with Renilla activity and the reporter assays were performed in triplicate.

Animals

Male athymic nu/nu mice (4 weeks old) and CD-1 mice were either obtained from the breeding colony of the Department of Experimental Radiation Oncology at UTMDACC or from Cancer Research Technology, UK (CR-UK). Before initiating experiments, we acclimatized all mice to a pulverized diet for 3 days. All the mice were weighed daily during the course of experiments. Our experimental protocol was reviewed and approved by the Institutional Animal Care and Use Committee at UTMDACC or at CR-UK.

Maximum Tolerated Dose (MTD) assay

Prior to examining the efficacy of the PKD inhibitor, CRT0066101, the maximum tolerated dose (MTD) of the drug was established in CR-UK nu/nu mice and determined to be 80 mg/kg/day.

Heterotopic (subcutaneous) pancreatic cancer model

CR-UK nu/nu mice were inoculated subcutaneously into the left and right flanks with 5 × 10⁶ cells of Panc-1 cells in 100 µL PBS. Tumors were allowed to grow till their optimal size as measured by calipers.

Determination of mouse pharmacokinetics and efficacy of CRT0066101 in a heterotopic pancreatic cancer model

CD-1 mice were administered with a daily oral dose of 80 mg/kg for 5 days and blood removed by cardiac puncture under terminal anesthesia for measurement of plasma concentration. For the heterotopic xenograft model, CR-UK nu/nu mice were subcutaneously injected with 5 × 10⁶ Panc-1 cells. Nineteen days after implantation of Panc-1 cells, tumor areas were, on average 0.3 cm² and were randomized into the following groups (n = 8 mice per group): (a) vehicle (control) 5% dextrose administered by oral gavage once daily and (b) 80 mg/kg CRT0066101 dissolved in 5% dextrose administered by oral gavage once daily. Tumors were measured in 2 dimensions every 2–3 days by calipers and area was calculated by multiplying length by width. Therapy was given until tumors reached their designated size limits (1.44 cm²) or until day 24 in CRT0066101 treated group. Final tumor areas were compared among groups using a Student's t test and Fisher’s exact test with p < 0.05.

Measurement of active PKD (pS916-PKD) in heterotopic tumor explants

CR-UK nu/nu mice were inoculated with 5 × 10⁶ cells per flank and left to grow until the tumors reached about 0.3 cm² in size. Mice were treated as indicated for 5 days and sacrificed at 0, 2, 6, and 24 h after the last treatment. Tumors were excised and lysates were prepared for analysis by western blot using pS916-PKD1/2 antibody (Cell Signaling Technology, Danvers, MA).

Measurement of CRT0066101 in heterotopic tumor tissues and plasma

Drug accumulation was measured in both plasma and tumor samples (n = 4 mice per group). Tumors were homogenized in ice-cold PBS and quantitative analysis was carried out using a Waters Quattro micro triple quad LCMS system with a Phenomenex Gemini-NX C18 as per manufacturer’s instructions.

Orthotopic pancreatic cancer model

Panc-1 PaCa cells were stably transduced with luciferase and the orthotopic model was established as previously described (19).

Experimental protocol for orthotopic pancreatic cancer model and bioluminescence imaging

One week after implantation of Panc-1 cells, mice were randomized into the following groups (n = 7 mice per group) based on the bioluminescence measured after the first IVIS imaging: (a) untreated control (vehicle; 5% dextrose orally) and (b) CRT0066101 (80 mg/kg dissolved in 5% dextrose) given orally (by gavage) once daily. Tumor volumes were monitored weekly by the bioluminescence IVIS Imaging System 200 as previously described (19). Therapy was given for 3 weeks and animals were sacrificed on day 35 after tumor implantation. Primary tumors in the pancreas were excised and the final tumor volume was measured as V = 2/3πr³, where r is the mean of the three dimensions (length, width, and depth). The final tumor volumes were initially subjected to one-way ANOVA and then later compared among groups using unpaired Student's t test. Half of the tumor tissue was fixed with formalin and paraffin embedded (FFPE) for immunohistochemistry and routine H&E staining. The other half was snap frozen in liquid nitrogen and stored at −80°C. H&E staining confirmed the presence of tumor in each group.

Western blots

Pancreatic tumor tissues (75–100 mg/mouse; n = 2 per group) from control and CRT0066101 treated mice were minced, homogenized using a Dounce homogenizer, centrifuged at 16,000 × g at 4°C for 10 min, and treated with ice-cold lysis buffer for preparation of cellular lysates. Similarly, whole-cell lysates were prepared from indicated PaCa and HPDE cell line as mentioned. Whole-cell lysates or cytoplasmic lysates (CE) were prepared from Panc-1 cells either transfected with control (Panc-1) or PKD1 over-expressing vectors (Panc-1-PKD1) using ice-cold lysis buffer as mentioned. The proteins were then fractionated by SDS-PAGE, electrotransferred to nitrocellulose membranes, blotted with each antibody, and detected by enhanced chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ).

Ki-67 analysis

FFPE sections (5 µm) were stained with Ki-67 (rabbit monoclonal clone SP6, NeoMarkers, Fremont, CA) antibody (20). Results were expressed as percent of Ki-67⁺ cells ± SE per 200× magnification. A total of 10 HPFs were examined and counted from each treatment groups (n = 7 per group). The values were compared both using unpaired Student’s t-test and Fisher’s exact test with p < 0.05.

In situ terminal deoxynucleotidyl transferase–mediated nick end labeling (TUNEL) assay

Cryostat sections (5 µm) were fixed in 4% paraformaldehyde (in PBS, pH 7.4), and in situ terminal deoxynucleotidyl transferase–mediated nick end labeling (TUNEL) assay (Roche Diagnostics, Germany) was done as per the manufacturer’s instructions described previously (20). A total of 5 HPFs (400×) were examined and counted for TUNEL⁺ and total number of cells from each treatment groups (n = 7 per group). Results were expressed as the percent of TUNEL⁺ cells ± SE per 400× magnification (HPF). The values were compared both using unpaired Student’s t-test and Fisher’s exact test with p < 0.05.

Microvessel density

Cryostat sections (5 µm) were stained with rat anti-mouse CD31 monoclonal antibody (BD Biosciences, San Jose, CA). Areas of greatest vessel density were then examined under higher magnification (200×) and counted (20). Results were expressed as the mean number of vessels ± SE per high power field (HPF). A total of 10 HPFs were examined and counted from each treatment groups (n = 7 per group). The values were compared both using unpaired Student’s t-test and Fisher’s exact test with p < 0.05.

RESULTS

Active and total PKD family expression in PaCa

We demonstrated for the first time autophosphorylated (active) PKD family expression by IHC analysis of a human PaCa tissue microarray (Figure 1A). The monoclonal antibody used for IHC detects autophosphorylated PKD1 and PKD2 on Ser-910 and Ser-876, respectively. PKD3 does not contain a residue that can be phosphorylated at the equivalent C-terminal position. As shown in Figure 1B, active PKD1/2 was significantly expressed in PaCa tissues as compared with normal ducts (91% vs 22%). The intensity of IHC staining was independently graded by a pathologist, who was blinded to this study. Representative photomicrographs of IHC staining intensity are shown in Figure 1A and Supplementary Figure 1A. Furthermore, the majority of PaCa tissues showed increased IHC staining intensity compared to normal ducts (Supplementary Figure 1B). We also examined expression of total PKD1/2 in multiple PaCa cell lines and in immortalized human pancreatic duct epithelial cells (HPDE). The antibody used (C-20, Santa Cruz Biotechnology, Santa Cruz, CA) recognizes both, PKD1 and PKD2 (but not PKD3). As shown in Figure 1C, a majority of the PaCa cell lines examined exhibited significant expression of PKD1/2 in comparison with HPDE.

Figure 1.

Increase of activated PKD1/2 in pancreatic ductal adenocarcinoma (PaCa) and PKD family expression in PaCa cell lines. A. Representative immunohistochemistry (IHC) photomicrographs of activated PKD1/2 in PaCa tissue microarrays depicting normal pancreatic ducts and PaCa as described in Materials and Methods. B. Summary of activated PKD1/2 IHC grading in PaCa tissue microarrays as described in Materials and Methods. C. Western blot analysis of PKD1/2 in multiple human PaCa cell lines and in immortalized human pancreatic duct epithelial cells (HPDE) as described in Materials and Methods.

Discovery of CRT0066101 as a PKD specific small-molecule inhibitor

To validate PKD as a potential anti-cancer target, we screened a diverse compound library against purified PKD family enzymes to identify novel inhibitors against this protein kinase family. The backbone structure of the novel PKD inhibitors is depicted in Figure 2A. A lead member of this family, CRT0066101, was used in vitro to quantitate its effects on the catalytic activity of PKD family members, as determined by inhibition of peptide substrate phosphorylation. The biochemical IC₅₀ values were 1, 2.5 and 2 nM for PKD1, 2, and 3 respectively (data not shown). The specificity of CRT0066101 for PKD family members was also confirmed in an in vitro kinase assay comprising a panel of > 90 protein kinases including PKCα/PKBα/MEK/ERK/c-Raf/c-Src/c-Abl that have a role in cancer promotion or progression. Using intact Panc-1 cells as our model system and a high-throughput FACE assay, the IC₅₀ value was 0.5 µM (Supplementary Figure 2A). CRT0066101 specifically blocked PKD1/2 activity and did not suppress PKCα/PKCβ/PKCε activity in multiple cancer cell types including A549 (lung) and MiaPaCa-2 (pancreas).

Figure 2.

Effects of CRT0066101 on Panc-1 cells in vitro. A. Chemical structure of a typical amino-ethyl-aryl compound, where R1, R2, and R3 represent functional groups. This represents the backbone structure of CRT0066101 and its integrity confirmed by LC-MS and NMR. B. Panc-1 cells were treated with CRT0066101 prior to stimulation with 20% serum and BrdU incorporation was measured by ArrayScan™ as described in Materials and Methods. CRT0066101 inhibited both serum stimulated and basal Panc-1 proliferation (as measured by BrdU incorporation); thereby leading to 200% inhibition at the maximal dose tested. C. Panc-1 cells were serum starved for 6 h and levels of cleaved caspase-3 determined using the MSD multiplex apoptosis panel kit as described in Materials and Methods. Data represents mean ± SE with p < 0.05 as compared with 0.5 µM. D. Colo357 and Capan-2 cells were either treated with control vehicle DMSO (0 µM) or CRT0066101 (5 µM) for 24 h and 48 h, respectively. Cell proliferation was determined by using the CellTiter Aqueous One Solution Cell Proliferation Assay kit as described in Materials and Methods. All the experiments were performed in triplicate. Data represents mean ± SE and is presented as fold change over untreated control (24 h) with p < 0.01.

CRT0066101 reduced proliferation, increased apoptosis, and reduced cell viability of PaCa cells

Based on our results in Figure 1C, we have used Panc-1 cells expressing moderate levels of PKD1/2 as our model system. As shown in Figure 2B, CRT0066101 significantly inhibited Panc-1 cell proliferation, as revealed by inhibition of BrdU incorporation, with an IC₅₀ value of 1 µM. Treatment with CRT0066101 resulted in a 6–10 fold induction of apoptosis in Panc-1 cells, as judged by the levels of cleaved caspase-3 (Figure 2C). Next, we evaluated the sensitivity to CRT0066101 of PaCa cells with varying endogenous levels of PKD1/2 expression. Specifically, we measured cell proliferation of PaCa cells either expressing moderate-high (Colo357, Panc-1, MiaPaCa-2, and AsPC-1) or low (Capan-2) levels of PKD1/2 after treatment without or with CRT0066101 (5 µM) for 24 h and 48 h. As shown in Figure 2D and Supplementary Figure 2B, CRT0066101 significantly reduced cell proliferation of Colo357, Panc-1, MiaPaCa-2, and AsPC-1 cells but had a modest effect in Capan-2 cells. These results suggest that CRT0066101 significantly blocked proliferation of PaCa cells that express moderate to high endogenous levels of PKD1/2.

CRT0066101 reduced agonist-induced PKD activation and PKD-dependent phosphorylation of Hsp27 in Panc-1

To determine specificity of CRT0066101 inhibition of PKD1/2 activity induced by a physiological agonist through receptor-mediated pathways, we examined whether it prevents PKC-dependent PKD1/2 activation induced by the GPCR agonist neurotensin (NT) in PaCa cells. As shown in Figure 3A, CRT0066101 (5 µM) blocked both the basal and NT-induced pS916-PKD1/2 (activated PKD1/2) in Panc-1 and Panc-28 cells. Next, we showed that CRT0066101 in Panc-1 cells abrogated NT-induced phosphorylation of Hsp27 (pS82-Hsp27), which is a physiological substrate of PKD1/2 (12), in a dose-dependent manner (Figure 3B). Interestingly, CRT0066101 did not block phosphorylation of MARCKS (pS152/156-MARCKS), which is a known PKC substrate (Figure 3B). In contrast, GF-1, a conventional and novel PKC isoform-selective inhibitor, potently blocked expression of pS152/156-MARCKS, pS916-PKD1/2, and pS82-Hsp27 (Figure 3B). These results strongly suggest that CRT0066101 is a PKD-specific inhibitor.

Figure 3.

Effects of CRT0066101 on agonist-mediated PKD activation, PKD-mediated substrate phosphorylation, NF-κB activation, and expression of NF-κB-dependent gene products in PaCa cells. A. Serum-starved cultures of Panc-1 and Panc-28 cells were treated either with control vehicle DMSO (0 µM) or CRT0066101 (5 µM) for 1 h prior to stimulation without (−) or with 50 nM of NT for 10 min. Cell lysates were prepared and western blots were performed with antibodies against pS916-PKD1/2 and total PKD1/2 as described in Materials and Methods. The membrane was re-probed with β-actin to verify equal loading. B. Serum-starved cultures of Panc-1 cells were incubated in the absence (−) or in the presence of 5 µM or 1 µM CRT0066101 or 3.5 µM GF 109203X (GF1) for 1 h prior to stimulation without (−) or with 5 nM NT for 10 min. Cell lysates were prepared and western blots were performed with antibodies against pS82-Hsp27pS82, pS916-PKD1/2, and pS152/156-MARCKS. The membrane was re-probed with total Hsp27 antibody to verify equal loading. C. Panc-1 cells were either transfected with control (Panc-1) or PKD1 over-expressing (Panc-1-PKD1) vectors and 23 h post-transfection were treated with DMSO (−) or 5 µM CRT0066101 (+) for 1 h prior to preparation of nuclear extracts (NE) for measuring NF-κB activation by EMSA as described in Materials and Methods; p.c. stands for positive control (nuclear extracts from KBM-5 cells stimulated with TNF-α). Results in the bottom of EMSA are expressed as fold activity of the untreated control (Panc-1 cells transfected with mock vector). D. Panc-1 cells were either transfected with control (Panc-1) or PKD1 over-expressing (Panc-1-PKD1) vectors and 23 h post-transfection were treated with DMSO (−) or 5 µM CRT0066101 (+) for 1 h prior to preparation of whole cell lysates to analyze effects on NF-κB-dependent gene expression. Western blots were then performed with antibodies against cyclin D1, survivin, and cIAP-1 as described in Materials and Methods. The membrane was re-probed with β-actin to verify equal loading. All the results shown are representatives of three independent experiments.

CRT0066101 reduced PKD-dependent NF-κB activation and NF-κB-dependent gene expressions in Panc-1

To further dissect the mechanism by which PKD1/2 enhances cell proliferation and pro-survival signaling, we investigated the effect of its inhibition on NF-κB, a pivotal transcription factor that regulates these biological processes (21–23). We, and others, have shown that GPCR-mediated signaling through PKD1/2 regulates NF-κB gene products in multiple cell types (4, 24, 25). Specifically, we examined whether CRT0066101 could block PKD1/2 over-expression induced NF-κB activity in PaCa cells. Using Panc-1 cells as our model system, we showed that PKD1 over-expression induced NF-κB activity by 60% and this effect was blunted by CRT0066101 (5 µM), as demonstrated by EMSA of nuclear extracts (NE) in Figure 3C. In agreement with previous reports (19), we observed constitutive NF-κB activation in control-vector transfected Panc-1 cells. As noted in Figure 1C, Panc-1 cells express moderate levels of PKD1/2 that can partly explain constitutive NF-κB activation in these cells. In addition, ligand-induced physiological activation of PKD1/2 potently stimulated NF-κB dependent reporter gene activation over the constitutive basal level. As shown in Supplementary Figure 3A, CRT0066101 reduced TNF-α induced NF-κB dependent reporter gene (luciferase) activity in Panc-1 cells in a dose-dependent manner at 6 h and 24 h after stimulation. Similar results were obtained in PaCa cells including AsPC-1 and Panc-28. Furthermore, we showed that CRT0066101 significantly inhibited PKD2-mediated NF-κB dependent reporter gene (luciferase) activity (Supplementary Figure 3B). Similar results were obtained with PKD1 over-expression in Panc-1 cells. To confirm that CRT0066101 abrogated NF-κB-dependent gene products that are critical for cell proliferation and survival (21–23), we showed CRT0066101 potently inhibited expression of cyclin D1, survivin, and cIAP-1 in both Panc-1 and Panc-1-PKD1 over-expressing cells (Figure 3D). Interestingly, we observed an increased expression of survivin after PKD1 over-expression in Panc-1 cells that was significantly attenuated by CRT0066101. Thus, our results suggest that PKD1/2 promote activation of NF-κB and its gene products.

CRT0066101 inhibited growth of tumors in a subcutaneous PaCa model

Pharmacokinetic parameters were determined after a single bolus dose of CRT0066101. The terminal half life and bioavailability were determined to be 60 min and ~100% respectively (data not shown). Furthermore, plasma concentrations of CRT0066101 were evaluated following administration of daily oral doses of 80 mg/kg for 5 days in CD-1 mice and as shown in Supplementary Figure 4A, optimal therapeutic concentrations (8 µM) of CRT0066101 were detectable 6 h after oral administration of this drug. These results prompted us to further evaluate detailed pharmacological properties of CRT0066101 in vivo. We treated established tumors in subcutaneous Panc-1 xenograft models with a dose of 80 mg/kg CRT0066101, given orally once daily, or vehicle control. There were no signs of toxicity throughout the treatment period as shown by stable bodyweights during the entire treatment period (Supplementary Figure 4B). We observed a statistically significant reduction (by 48.5%) in tumor volume at day 9 in the treated group (Figure 4A). To substantiate that the effects on tumor growth were mediated via inhibition of PKD activity, PKD1/2 autophosphorylation was measured in subcutaneous Panc-1 tumor explants following treatment with 5 oral doses of CRT0066101 or vehicle control. We observed a statistically significant, but modest reduction of PKD1/2 autophosphorylation 2 h after cessation of dosing (Figure 4B and Supplementary Figure 4C) and this effect was not quantifiable at subsequent time points (data not shown). Concentrations above those required to elicit effects in vitro were detectable in plasma (data not shown) and in Panc-1 tumor explants up to 6 h following this treatment regimen (Figure 4C).

Figure 4.

Effects of oral administration of CRT0066101 in vivo (heterotopic Panc-1 xenograft model). A. Mean tumor area depicted as tumor area (percent or %) of Panc-1 heterotopic xenografts in nu/nu mice following treatment with control (5% Dextrose) or CRT0066101 (n = 8) as described in Materials and Methods. All the mice in the control group were sacrificed at day 9 as the mean tumor area reached the cut-off value (> 1.44 cm²) per CRUK IACUC guidelines. Data represents mean ± SE. B. Quantified expression of activated PKD1/2 in Panc-1 tumors (n = 4 per group) 2 h after last treatment with CRT0066101 as described in Materials and Methods. Data represents mean ± SE with * p < 0.05 vs control. C. Tumor concentrations of CRT0066101 (n = 4) as described in Materials and Methods. Samples were collected 2, 6, and 24 h after mice received their 5^th dose. In A and B/C nu/nu mice were administered with CRT0066101 (80 mg/kg, once daily, oral) for 24 or 5 days, respectively. Data represents mean ± SE with * p < 0.05 vs 2 h tissue concentration.

CRT0066101 reduced growth of tumors in an orthotopic PaCa model

The effects of CRT0066101 in orthotopically implanted Panc-1 cells in nude mice were investigated as shown in Figure 5A (26). Bioluminescence was measured as a surrogate marker to assess tumor volume. The bioluminescence imaging (Fig 5B, left panel) indicated that there is an increase of tumor volume in the control group compared to the treated group. The normalized photon counts showed statistically significant reduction after 18 days of treatment with CRT0066101 and continued for 5 days beyond the cessation of treatment (Figure 5B, right panel). As shown in Figure 5C, there was a 58% reduction of final tumor volume at the end of trial in the treated group.

Figure 5.

Effects of oral administration of CRT0066101 in an orthotopic Panc-1 xenograft model. A. Schematic representation of experimental protocol. Groups I and II were treated with vehicle (5% Dextrose) or CRT0066101 (80 mg/kg, once daily, oral, n = 7), respectively. B. Left panel, IVIS bioluminescence images of orthotopically implanted pancreatic tumors in live anesthetized mice as described in Materials and Methods. B. Right panel, measurements of photons per second depicting tumor volumes of Gp I (white) and Gp II (yellow) using live IVIS imaging were plotted at the indicated times (n = 7) as described in Materials and Methods. Points, mean; bars, SE. *p < 0.01. C. Gp I (white) and Gp II (yellow) in each mice measured during autopsy using Vernier calipers and calculated using the formula V = 2/3πr³ (n = 7) as described in Materials and Methods. Columns, mean; bars, SE. *p < 0.01.

To understand the mechanism of reduction in tumor volume by CRT0066101, we examined for proliferation and apoptosis markers in these tumor explants. Ki-67 (proliferation marker) expression was significantly reduced (by 52%) in the treated group (Figure 6A) and the in situ TUNEL assay demonstrated 58% increased apoptosis in tumor explants of the treated group (Figure 6B) as compared with the control. CRT0066101 inhibited expression of cyclin D1 (Figure 6C), which is involved in tumor cell proliferation (27). Over-expression of NF-κB dependent gene products including survivin, Bcl-2, Bcl-xL, and IAP-1 are linked to tumor survival, chemoresistance, and radioresistance (28). CRT0066101 treatment abrogated expression of these proteins in tumor explants (Figure 6C). We also examined the effects of CRT0066101 on tumor-associated angiogenesis as this process is critical for tumor survival and proliferation (2, 20). CRT0066101 treatment reduced CD31⁺ microvessel density by 60% in the tumor explants (Supplementary Figure 5A) and abrogated expression of COX-2 and VEGF (Supplementary Figure 5B), important mediators of PaCa-associated angiogenesis.

Figure 6.

Effects of CRT0066101 on proliferation, apoptosis, proliferative, and pro-survival gene products in orthotopic PaCa tumor explants. A. Quantification of Ki-67⁺ cells or proliferation index as described in Materials and Methods. Values are means (n = 7) ± SE. *p < 0.01. Representative Ki-67⁺ IHC are shown for Group I or Gp I (control) and Group II or Gp II (CRT0066101) at 200×. B. Quantification of in situ TUNEL⁺ cells as described in Materials and Methods. Values are means (n = 7) ± SE. *p < 0.05. Representative TUNEL⁺ cells are shown for Group I or Gp I (control) and Group II or Gp II (CRT0066101) at 400×. C. Western blot analysis of proliferative (cyclin D1) and pro-survival proteins (survivin, Bcl-2, Bcl-xL, and cIAP-1) in cellular lysates from Group I or Gp I (control) and Group II or Gp II (CRT0066101) tumor explants as described in Materials and Methods. Cell lysates were prepared from each tumor explants obtained from 2 distinct animals in vehicle (1 and 2) and CRT0066101 (3 and 4) treated groups. The membrane was re-probed with β-actin to verify equal loading.

DISCUSSION

The PKD family regulates multiple key cellular functions and is implicated in the pathogenesis of neoplastic disorders, including PaCa (29). One of the salient features presented here is that activated PKD1/2 is significantly upregulated in PaCa as compared with normal pancreatic ducts. Immunohistochemistry (IHC) of a PaCa tissue microarray showed strongly positive staining in the majority of PaCa neoplastic epithelia while most of the normal ductal epithelia had negative or minimal staining.

Despite the evidence supporting the fundamental role of PKD in cancer and consequentially as a putative target, there has been limited progress in the development of selective inhibitors. Such inhibitors are required for detailed pharmacological investigations of inhibition of the molecular target. Consequently, we established a PKD-specific small molecule discovery program and identified a prototype compound CRT0059359 with activity in vitro. To obtain a bioavailable pharmacologically active compound, medicinal chemistry and biological optimization were utilized to identify CRT0066101 as our lead product. Our in vitro studies noticeably demonstrated that CRT0066101 is a PKD-specific inhibitor as it potently blocked GPCR agonist-induced PKD activation, abrogated PKD-specific phosphorylation of a physiological substrate (Hsp27) involved in chemoresistance, and reduced PKD-mediated NF-κB activation in PaCa cells. Indeed, CRT0066101-mediated inhibition of PKD1/2 decreased NF-κB activity and expression of NF-κB-dependent gene products essential for cell proliferation and survival in PaCa cells. In two distinct murine PaCa models, CRT0066101 significantly abrogated tumor growth without any detectable systemic toxicity and increased survival period (a surrogate marker) in the subcutaneous model. Further, CRT0066101 reduced proliferation, increased apoptosis, and reduced angiogenesis in our orthotopic PaCa model.

There are several reports of PKD inhibitors in the literature including Go6976 (30), H-89 (31), and the polyphenolic cancer chemopreventive agent trans-3-4’,5-trihydroxystilbene (resveratrol) (32). However, the therapeutic potential of these agents may be hindered by their poor selectivity or high concentrations that would be required to achieve pharmacological effects (32). None of these compounds have been tested as therapeutic agents in vivo. Recently, a non-competitive selective PKD inhibitor CID755673 was identified from the NIH repository of small molecules (33). CID755673 significantly reduced proliferative, migratory, and invasive phenotypes in prostate cancer cells with IC₅₀ values between 10–30 µM (33) but also induced unexpected mitogenic effects (34). It will be important to establish the therapeutic potency and toxicity profile of CID755673 in vivo. CRT0066101 is structurally distinct from CID755673 and inhibits the activity of all PKD isoforms (PKD1, PKD2, and PKD3). More importantly, CRT0066101 has significant activities in vitro in the range of 0.5–5 µM and we show the feasibility of achieving these concentrations of drug in mice following oral dosing.

In conclusion, this study identifies CRT0066101 as a potent inhibitor of the PKD family and validates the role of PKD1/2 in PaCa tumorigenesis. We showed, for the first time, that this inhibitor was orally bioavailable and blocked tumor growth in vivo. Our data supports further evaluation of this series of novel anti-cancer therapeutics.

Abbreviations

PKD

protein kinase D

PaCa

pancreatic cancer

PKD1

protein kinase D1

PKD2

protein kinase D2

PKD3

protein kinase D3

PKC

protein kinase C

NT

neurotensin