Targeting the α Receptor for Platelet-Derived Growth Factor as a Primary or Combination Therapy in a Preclinical Model of Prostate Cancer Skeletal Metastasis

This article is featured in Highlights of This Issue, p. 4907

Although skeletal metastases occur with a significantly high rate in patients affected by advanced prostate carcinoma (1, 2), a limited range of options is currently available for the treatment of metastatic bone lesions (3, 4). Equally frustrating is the evidence that the burden deriving from skeletal dissemination represents the main cause of death for these patients, but the current standard of care for advanced prostate cancer produces merely palliative effects (5). Skeletal metastases cause bone resorption and chronic pain and predispose patients to skeletal-related events such as pathologic fractures and spinal cord compression. This skeletal morbidity is presently treated with the administration of bisphosphonates in prophylactic and adjuvant settings (6). Bisphosphonates are inorganic pyrophosphates that inhibit the enzymatic activities of osteoclasts, the cells responsible for the degradation of bone matrix (7, 8). These compounds are effective in alleviating bone loss and pain, as well as reducing the occurrence of skeletal-related events (9, 10). However, bisphosphonates are not without side effects, including renal toxicity, hypocalcemia, and osteonecrosis of the jaw, which may prevent their long-term administration in bone-metastatic prostate cancer patients (11, 12).

In addition to the management of symptoms produced by skeletal involvement and the prevention of skeletal-related event, the use of bisphosphonates in the clinic to target osteoclasts is also aimed to directly inhibit cancer cell growth and survival. This approach is based on evidence that the degradation of bone matrix releases a variety of trophic factors capable of supporting the disseminated cancer cells in their metastatic growth. The resulting increase in tumor mass will recruit and activate a progressively higher number of osteoclasts, thereby generating a self-sustained vicious cycle (13–15). Unfortunately, results from clinical trials have not been encouraging and seem to indicate that, although successful in reducing bone loss and pain, bisphosphonates are ineffective in arresting disease progression or improving survival in patients with metastatic prostate cancer (6, 16). This scenario is likely determined by the fact that disseminated cancer cells may benefit from additional trophic support that does not derive from bone matrix degradation but is readily available in the bone marrow environment and, therefore, not necessarily dependent on osteoclast activity. Indeed, our previous studies have shown that isolated cancer cells and small foci detected shortly after their initial colonization of the bone are either independent of or much less reliant on spatial interactions with osteoclasts than are larger, more advanced lesions (17). Thus, it seems plausible that other compounds used in combination with bisphosphonates, such as those that target survival and/or proliferative pathways in cancer cells, may result in pronounced curative effects.

The α receptor for platelet-derived growth factor (PDGFRα) is a receptor tyrosine kinase that has been correlated with advanced prostate carcinoma and was found associated with skeletal metastases by others (18–20) and us (17, 21). The activation of this receptor in prostate cancer cells is responsible for the recruitment of the PI3K/Akt cell survival pathway. To target the human PDGFRα, we used the humanized monoclonal antibody IMC-3G3 (ImClone, Eli Lilly). This antibody was developed for the treatment of several pathologic conditions and is currently in phase II clinical trials (22). We found that IMC-3G3 suppresses PDGFRα-mediated activation and downstream Akt stimulation by inducing receptor internalization (23). Interestingly, we have also shown that the soluble fraction of human bone marrow is able to transactivate human PDGFRα through a ligand- and dimerization-independent mechanism (23, 24). Thus, IMC-3G3 could negatively affect the survival of PDGFRα-expressing prostate cancer cells by internalizing the receptor and thereby impeding its transactivation and consequent downstream signaling. Indeed, we have previously observed a strong antimetastatic effect when using IMC-3G3 prophylactically in a preclinical model of bone-metastatic prostate cancer. In this context, the work presented here expands our previous study by addressing any antimetastatic effect of IMC-3G3 administered in a curative protocol, as well as the role played by the simultaneous targeting of PDGFRα on cancer and stromal cells. We also report that IMC-3G3, unlike bisphosphonates, is extremely effective in reducing metastatic growth during the first week following the arrival of cancer cells to the skeleton. Finally, we show that the administration of IMC-3G3, alone or in combination with a bisphosphonate, prolong the survival of mice bearing skeletal metastases from human prostate cancer cells.

The IMC-3G3 and IMC-1E10 antibodies were provided by ImClone, Eli Lilly. The bisphosphonate zoledronic acid (ZA) was obtained from Sigma.

NIH-3T3 cells were obtained from American Type Culture Collection, whereas the highly bone-metastatic PC3-ML subline was derived from the parental PC-3 cells (American Type Culture Collection) as previously described (25). Cells were cultured at 37°C and 5% CO₂ in DMEM supplemented with 10% fetal bovine serum (Hyclone) and 0.1% gentamicin (Invitrogen). The PC3-ML cells were engineered to stably express enhanced variant of green fluorescent protein (GFP) by using a lentiviral vector from America Pharma Source (17).

Cells were washed twice with sterile PBS and then starved from serum for 4 hours before being incubated for 30 minutes with either IMC-1E10 or IMC-3G3 (20 μg/mL) and successively exposed to PDGF-AA (30 ng/mL).

Cell lysates were obtained and SDS-PAGE and Western Blot analysis were done as previously described (24), with few modifications. Membranes were blotted with antibodies targeting phosphor-Akt (Ser-473; Cell Signaling Technology), PDGFRα (R&D Systems), and total Akt (Cell Signaling). Primary antibody binding was detected using a horseradish peroxidase-conjugated secondary antibody (Pierce). Chemiluminescent signals were obtained using SuperSignal West Femto reagents (Pierce) and were detected with the Fluorochem 8900 imaging system and relative software (Alpha Innotech). Densitometry analysis was done using the UN-SCAN IT software (Silk Scientific).

Five-week-old, male, immunocompromised mice (CB17-SCRF) were obtained from Taconic and were housed in a germ-free barrier. At 6 weeks of age, mice were anesthetized with 100 mg/kg ketamine and 20 mg/kg xylazine and successively inoculated in the left cardiac ventricle with PC3-ML cells (5 × 10⁴ in a volume of 100 μL of serum-free DMEM/F12). Although consistently producing skeletal tumors in tibiae and femora of >80% of inoculated mice, PC3-ML cells do not metastasize to any soft-tissue organ with the exception of small tumors produced in the adrenal glands and, very seldom, the liver. Mice were sacrificed at specified time points following inoculation and tissues prepared as described below. For the survival studies, mice were euthanized when in a moribund state as defined by evident and prolonged shivering, extreme prostration, labored breathing, >20% of body weight loss, and/or unresponsiveness to external stimulation. All experiments were conducted in accordance with NIH guidelines for the humane use of animals. All protocols involving the use of animals were approved by the Drexel University College of Medicine Committee for the Use and Care of Animals.

Animals received a loading dose of 214 mg/kg immediately after inoculation with PC3-ML cells, followed by a maintenance dose of 60 mg/kg every 72 hours thereafter. Doses and administration schedule were selected based on pharmacokinetic and tumor-growth studies (26). Control mice were maintained on an identical dosing schedule and received similar injection volumes of either saline or human immunoglobulins of the IgG₁ subclass as the IMC-3G3 and IMC-1E10 antibodies (26).

Tibiae and femora were collected and fixed in 4% formaldehyde solution for 24 hours and then transferred into fresh formaldehyde for additional 24 hours. Bones were decalcified in 0.5 mol/L EDTA for 7 days, followed by incubation in 30% sucrose for cryoprotection or 1% formaldehyde for long-term storage. Tissues were maintained at 4°C during all aforementioned steps and frozen in optimal cutting temperature (OCT) medium (Electron Microscopy Sciences) by placement over dry ice-chilled 2-methylbutane (Fisher). Serial sections of 80 μm in thickness were obtained using a Microm HM550 cryostat (Mikron).

Bright-field and fluorescent images of skeletal metastases were acquired using a SZX12 Olympus stereomicroscope coupled to an Olympus DT70 CCD color camera. Digital images were analyzed with ImageJ software (http://rsb.info.nih.gov/ij/) and calibrated for measurement by obtaining a pixel to millimeter ratio. Morphometric evaluation of skeletal tumors was conducted by analysis of serial cryosections, in which the largest representative tumor section for each metastasis was identified. A freehand tool was used to outline the border of each metastatic lesion, and the area was computed using the ImageJ “measure area” function.

Slides were incubated at 37°C for 5 minutes in a solution containing naphthol 7-Bromo-3-hydroxy-2-naphthoic-o-anisidide phosphate and ethylene glycol monoethyl ether (Sigma). Slides were then transferred to a solution containing sodium nitrite and pararosaniline chloride (Sigma) for ∼3 minutes.

We analyzed number and size of skeletal metastases between groups using a two-tailed Student's t test. Statistical significance between multiple groups was conducted using a one-way ANOVA, followed by Tukey's multiple comparison test. A value of P ≤ 0.05 was considered statistically significant.

Previous studies from our group have shown that bone-metastatic human prostate cancer cells express high levels of PDGFRα (21). The exposure of these cells to the soluble fraction of human bone marrow induces phosphorylation of PDGFRα and its downstream activation of the PI3K/Akt survival pathway (23). The antagonism of human PDGFRα achieved with the IMC-3G3 antibody attenuates the signaling events in vitro and reduces the number and size of skeletal metastases in preclinical animal models (17). These studies show a prophylactic action of the IMC-3G3 antibody on skeletal metastases; here, we sought to determine whether PDGFRα inhibition could also produce a curative effect on metastatic bone lesions that had sufficient time to establish and progress before IMC-3G3 administration. To this end, we used a model of experimental metastasis that uses PC3-ML human prostate cancer cells previously isolated from the PC3 parental cell line for their high bone-metastatic potential (25). These cells were engineered to stably express an enhanced variant of GFP and directly inoculated in the blood circulation of immunocompromised severe combined immunodeficient mice through the left cardiac ventricle. Metastatic tumors ranging from microscopic foci to macroscopic lesions were identified and measured as previously described (17).

A first set of experiments was conducted with mice inoculated with PC3-ML cells and left untreated for either 7 or 14 days, thus providing a time interval for metastatic tumor growth. Following this first period, treatment with IMC-3G3 began and continued until sacrifice at 4 weeks postinoculation. The dosing protocol consisted of a loading dose followed by maintenance doses administered three times weekly through an i.p. route (17). Control mice were maintained on an identical dosing schedule and received either saline or human immunoglobulins of the same IgG₁ subclass as the IMC-3G3 antibody (26), with both controls providing identical results. At the end of the study, mice were euthanized and the area of metastatic lesions was measured by fluorescence microscopy of frozen tissue sections. Animals that were administered IMC-3G3 according to these two different curative protocols presented with skeletal lesions in the tibia and femur that seemed significantly reduced in size from those observed in mice from control groups (Fig. 1A). In addition, delaying administration of IMC-3G3 for 1 or 2 weeks following tumor cell inoculation resulted in nearly identical growth inhibition (Fig. 1B). This strongly suggests that prostate cancer cells that express PDGFRα still depend on this receptor for their growth and survival when tumor foci occupy an average area of the bone marrow <35 ± 6 × 10³ μm² as we determined earlier (17). Interestingly, only at this point in time the bone lesions in our model became spatially associated with the surrounding osteoclasts. In fact, a comparable lack of osteoclast involvement during the initial stages of bone colonization has been also observed in humans (27). Thus, our data seem to point out to a role for PDGFRα in sustaining metastatic progression in the bone before the onset of osteolysis and induction of the vicious cycle.

Multiple studies have shown that reciprocal interactions between cancer cells and the surrounding stroma play a critical role in tumor growth (28–30). For instance, it has been shown that stromal PDGFRα maintains tumor growth and local angiogenesis when stimulated by PDGF-AA, PDGF-CC, or vascular endothelial growth factor-A secreted by cancer cells (31, 32). Based on this evidence, we sought to establish the contribution of the PDGFRα expressed by stromal cells (i.e., osteoblasts and cells of the mesenchymal compartment of the bone marrow) on the skeletal colonization and metastatic progression of PC3-ML cells. The IMC-3G3 antibody has been shown to be highly selective for human PDGFRα (26). We confirmed this specificity by exposing NIH-3T3 mouse fibroblasts to PDGF-AA upon pretreatment with IMC-3G3 and noticed that the phosphorylation of Akt was unaffected as compared with control cells (Fig. 2A). Therefore, to effectively target the stroma, we used IMC-1E10, a humanized antibody selected for binding to murine PDGFRα and otherwise sharing an identical structure with IMC-3G3 (33). The species selectivity of this antibody was confirmed by the complete inhibition of Akt phosphorylation observed in NIH-3T3 cells stimulated with PDGF-AA (Fig. 2A).

In the ensuing studies, mice were inoculated with PC3-ML cells and randomly assigned to three experimental groups, which received saline solution (control group), the IMC-1E10 antibody alone, or IMC-1E10 and IMC-3G3 antibodies in combination. Mice were treated according to a prophylactic protocol we described previously (i.e., loading dose administered immediately following cell inoculation) and maintained on a three times weekly dosing schedule until sacrifice after 4 weeks (17). We show that solely targeting stromal PDGFRα by administration of IMC-1E10 had no effect on reducing the size of metastatic lesions in either the tibiae or femora of treated mice (Fig. 2B). On the other hand, administration of IMC-1E10 in combination with IMC-3G3 retained the ability of IMC-3G3 alone to impair metastatic growth (Fig. 2B). Despite that the activity of 1E10 on stromal PDGFRα was not directly assessed, these results suggest that targeting PDGFRα of the stromal cellular compartment is ineffective in decreasing bone metastatic growth and also fails to provide a synergistic effect when coupled with inhibition of the PDGFRα expressed by metastatic prostate cells. Notably, we found that the 4-week antagonism of stromal PDGFRα did not cause toxicity because animals maintained a healthy weight throughout the experiments and failed to show signs of overt organ damage at the necropsy. These preclinical findings are supported by the recent completion of phase I clinical trials reporting the absence of significant adverse effects by the prolonged administration of IMC-3G3 (22). The absence of toxicity observed in these trials may be explained by the minor physiologic role exerted by PDGFRα in fully developed organisms. Although PDGFRα plays a crucial role in embryonic development, in the adult, its function is mostly overshadowed by that of PDGFRβ in angiogenesis, wound healing, tissue homeostasis, and other systemic effects (34–36). In fact, there is evidence from genetic replacement experiments that PDGFRβ signaling can fully compensate for PDGFRα signaling, but not vice versa (37). The predominant function of PDGFRβ in human physiology strongly supports the rationale of safely targeting PDGFRα to counteract cancer cells that depend on this receptor for growth and survival (22). This seems particularly relevant to the field of prostate cancer because the therapeutic options currently available in the clinic for patients with skeletal metastatic disease are extremely limited, present a limited therapeutic window, and are only minimally effective.

Thus, the next series of experiments aimed to compare IMC-3G3 with a current standard of care for patients presenting with dissemination of prostate cancer to the bone. The presence of significant pain in these patients is caused by alterations in bone turnover deriving from the considerable recruitment and activation of osteoclasts at the site of metastasis (1, 38, 39). The resorption of mineralized bone matrix by these cells followed by deregulated deposition of sclerotic bone by osteoblasts is responsible for the mixed lytic-blastic appearance of metastatic lesions from prostate cancer (40). These histopathologic alterations almost invariably correspond to systemic clinical conditions such as hypercalcemia of variable degrees, as well as skeletal-related event such as spinal nerve compression and pathologic bone fractures (41). Because of the resulting morbidity and reduced quality of life, the administration of inhibitors of osteoclast activity is currently standard of care for patients with bone-metastatic dissemination, including that from prostate cancer (7). ZA is a bisphosphonate with a potent analgesic effect, which has shown significant activity in delaying the time to skeletal-related event in clinical trials conducted with prostate cancer patients (16). Therefore, we decided to investigate whether metastatic bone lesions would respond to the combined administration of ZA and IMC-3G3. For these studies, mice were assigned to groups receiving ZA or IMC-3G3 alone, as well as ZA plus IMC-3G3 combination therapy, and compared with control groups. IMC-3G3 was dosed as reported above, whereas ZA was administered by weekly s.c. injections of 100 μg/kg. At the end of the 4-week study, tissues obtained from sacrificed mice were examined for staining of tartrate-resistant acid phosphatase, a widely used histologic indicator of osteoclast activity (42, 43). As expected, mice that received weekly injections of ZA showed a marked reduction in the number and activity of osteoclasts at the tumor-stroma border (Fig. 3A). We also found that mice receiving either ZA or IMC-3G3 alone had a significant decrease in the size of skeletal lesions. However, there was no synergistic reduction of tumor size when the two therapies were combined (Fig. 3B).

Although ZA and IMC-3G3 were equally effective in delaying the growth of metastatic prostate cancer cells in the bone, we decided to ascertain whether these similar outcomes were the result of different mechanisms of action. This possibility seemed to be supported by the initial evidence presented above that blockade of PDGFRα by IMC-3G3 is particularly effective previous the onset of osteoclast-mediated degradation of the bone matrix, which is the main event inhibited by ZA. To address this point, mice inoculated with PC3-ML cells received either ZA or IMC-3G3 and were sacrificed after only 1 week of bone-metastatic tumor growth. We found that, although both groups of mice had comparable numbers of bone metastases, the lesions in IMC-3G3 treated mice were significantly smaller than the tumors detected in animals treated with ZA, which were indistinguishable from those measured in controls (Fig. 4). These results provide compelling evidence that the antagonism of PDGFRα, achieved in this study with IMC-3G3, effectively impairs the initial colonization of the skeleton by prostate cancer cells. This phase of metastatic progression seems refractory to ZA, which relies on the inhibition of osteoclasts and therefore targets later-stage tumors.

A conclusion that could be drawn from these experiments is that targeting either early metastases with IMC-3G3 or more advanced lesions with ZA leads to a similar outcome in terms of controlling the size of bone metastatic tumors from prostate cancer cells. However, preclinical models of metastatic breast cancer (44) and clinical trials conducted with advanced prostate cancer patients indicate the inability of bisphosphonates to improve overall survival from bone metastatic disease (16). Recent studies with animal models using radiologic and tumor imaging approaches show that reducing the extent of bone destruction and invasion may not fully prevent metastatic breast cancer cells from growing, for example, by infiltrating surrounding soft tissues (44). Based on the evidence that IMC-3G3 reduces metastatic progression by directly impairing cancer cell growth without affecting osteoclasts or bone degradation, we tested the effect on survival of IMC-3G3 and ZA administered either separately or as a combination therapy.

To carry out these studies, mice were inoculated with PC3-ML cells and randomly assigned to one of four groups: control, IMC-3G3, ZA, or IMC-3G3 plus ZA. as described above, mice received IMC-3G3 three times weekly following the initial loading dose, whereas ZA was administered weekly. Animals were kept in the study until moribund. Necroscopic examination did not detect visceral metastases by PC3-ML cells. The only exception was represented by enlarged adrenal glands, frequently bilaterally, consistent with what we have previously reported (45). These small adrenal tumors were only detected by fluorescence stereomicroscopy, always found contained within the gland parenchyma, and therefore very unlikely to cause lethal tumor burden. As shown in Fig. 5, mice treated with ZA alone showed an increase in survival as compared with the control group, although this effect did not reach statistical significance most likely because of the small number of mice used. However, the main conclusion of this set of experiments is that IMC-3G3 used alone prolonged survival and this effect was even more pronounced when this antibody was administered in combination with ZA. Taken together, these results clearly indicate that the blockade of PDGFRα through administration of IMC-3G3 delays the growth of preexisting skeletal metastases from prostate cancer cells and significantly extends survival in animals with bone-metastatic dissemination. These effects were exerted independently from the inhibition of osteolysis, which is the main mechanism of action of bisphosphonates such as ZA. Remarkably, the administration of ZA resulted in an enhanced survival benefit when combined with the direct antitumor action of IMC-3G3 on prostate cancer cells. This indicates that the mobilization of growth factors from the bone matrix by osteolytic events is likely to play a role in supporting prostate cancer growth at the skeletal level. A strategy aimed to contain bone degradation combined with a direct interference of PDGFRα functioning can significantly prolong survival in preclinical models of metastatic prostate cancer.

A. Fatatis is a consultant/advisory board member of ImClone.

We thank Dr. Olimpia Meucci (Drexel University College of Medicine) for the helpful discussion and Dr. Nick Loizos (ImClone, Eli Lilly) for providing the IMC-3G3 and IMC-1E10 antibodies.

Grant Support: W.W. Smith Charitable Trust and Department of Defense Prostate Cancer Program grant PC080987 (A. Fatatis), and Department of Defense Prostate Cancer Predoctoral Fellowship Award (PC094227; M.R. Russell).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.