Background
Every year prostate cancer is diagnosed in more than 500 000 men worldwide [
1]. Approximately 30 – 50% of them have evidence of metastatic disease, which results in a high morbidity and mortality. In advanced prostate cancer, 25 – 42% of the patients have metastases in regional lymph nodes [
2] and about 70% have metastases in bone [
3]. The mechanisms behind invasive prostate cancer are still largely unknown and the metastatic disease is basically incurable.
Only a few experimental models of spontaneous prostate cancer exist, since this malignancy is uncommon in nonhuman mammals. To study tumor growth
in vivo, immunodeficient mice are commonly used as models [
4,
5]. In athymic nude mice, human prostate cancer cells can be grown subcutaneously or tumor cells can be inoculated orthotopically into the prostate [
5]. In order to mimic a metastatic disease, cancer cells can also be inoculated into the circulation either via the tail vein [
6] or the left cardiac ventricle [
7]. However, considering the structure and function of stroma, vasculature and lymphatic vessels orthotopic tumors provide the most relevant environment for tumor growth and invasion.
Bisphosphonates are used in prevention and treatment of metastatic and myelomatous bone disease [
8‐
10] and osteoporosis [
11]. They primarily inhibit osteoclast-mediated bone resorption but in addition to this, they have been shown to have direct effects on different types of tumor cells and tumors. Our previous results [
12] showed that the aminobisphosphonate alendronate is able to strongly reduce invasion and migration of PC-3 prostate cancer cells at low, non-toxic concentrations and in a dose-dependent manner. Alendronate also inhibits PC-3 cell adhesion to extracellular matrix proteins in vitro [
13,
14] and secretion of MMP:s by tumor cells [
15]. Other bisphosphonates (clodronate, pamidronate, ibandronate and zoledronate) have been shown to induce apoptosis of breast cancer cells in vitro [
16,
17], inhibit angiogenesis [
18] and, at high concentrations, to inhibit cell proliferation [
19].
Interesting results have also been reported in animal models. They show that in models of bone metastasis of breast cancer, bisphosphonates effectively reduce tumor burden in bone [
20,
21], which has primarily been explained by inhibition of osteoclast-mediated bone resorption and decreased release of bone matrix-bound growth factors. Besides these indirect effects on bone metastasis, bisphosphonates have been shown to decrease tumor spread and formation of visceral metastases and increase survival in tumor-bearing mice. Alendronate has been shown to inhibit intraperitoneal dissemination of ovarian cancer cells [
22] and the newer aminobisphosphonates in particular, such as zoledronate and minodronate, have been reported to inhibit growth in 4T1/luc breast tumors [
23], melanoma [
24], cervical carcinoma [
25] and mesothelioma tumors [
19], as well as to suppress lung metastasis in osteosarcoma-bearing mice [
26]. In addition, zoledronate can inhibit angiogenesis in subcutaneous matrix implants [
27], in the prostates of testosterone-treated castrated rats [
18] and in a mouse model of cervical carcinogenesis [
19].
In clinical trials, Kanis
et al. [
8] and Powles
et al. [
28] showed that clodronate decreased the tumor burden of breast cancer in bone. The overall survival of the patients was better but there was no difference in visceral metastases. In a study by Diel
et al. [
29], clodronate reduced both bone and visceral metastases in breast cancer patients, although Saarto
et al. [
30] have reported the conflicting results as regards visceral metastasis.
The aim of the present study was to examine the effect of alendronate treatment on prostate tumor growth, invasion and lymph node metastasis. For this purpose we improved and exploited the model of orthotopic PC-3 prostate tumors in nude mice. The effects of the aminobisphosphonate alendronate on prostate tumor growth and spread to prostate-draining sacral and iliac lymph nodes were investigated by measuring tumor size and analyzing the tumors by means of histomorphometry and by carrying out immunohistochemical staining of markers of proliferation, angiogenesis, lymphangiogenesis and apoptosis. Using this model we demonstrate that alendronate decreases PC-3 prostate tumor growth and size of metastases in prostate-draining iliac and sacral lymph nodes in nude mice.
Methods
Animals
Eight-week-old male athymic nu/nu mice (Harlan Winkelman GmbH, Borchen, Germany) were maintained in a pathogen-free environment, under controlled conditions (20–21°C, 30–60% relative humidity and 12-hour lighting cycle). They were fed with small-animal food pellets (RM3 ESQC, Special Diet Services, Witham, England) and supplied with autoclaved tap water ad libitum. Animal welfare was monitored daily, looking for clinical signs, and the animals were weighed twice a week. The animal experiments were carried out according to the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes, and statutes 1076/85 and 1360/90 of The Animal Protection Law in Finland and EU Directive 86/609. The experimental procedures were reviewed by the local Ethics Committee on Animal Experimentation of the University of Turku and approved by the local Provincial State Office of Western Finland.
Cell culture
The human prostate cancer cell line PC-3 was obtained from the American Tissue Type Culture Collection (Rockville, MD). At near confluence, the cells were harvested and suspended at a concentration of 5 × 105/20 μl in a sterile dye solution consisting of phosphate-buffered saline (PBS; Biochrom AG, Berlin, Germany) with green food color 33022 (0.5 g/ml; Roberts Oy, Turku, Finland). The cells were kept on ice until used for inoculation. Cell viability was 97% or above at the time of inoculation, as determined by trypan blue staining of the cell suspension.
Orthotopic inoculation into the prostate
Half an hour before inoculation of tumor cells an analgesic drug (Temgesic, 0.3 μg/g, Schering-Plough Nv, Brussels, Belgium) was injected subcutaneously. The mice were anesthetized by means of isofluran inhalation (1.5–3%, air flow 200 ml/min, Univentor 400 anesthesia unit, Univentor Ltd., Zejtun, Malta) and placed in a supine position under a sterile cover. An incision was made 3 mm above the pubic symphysis and the bladder and seminal vesicles were carefully lifted to expose the dorsal prostate. To improve visual control of correct inoculation into the prostate, the cell suspension was supplemented with nontoxic green food color. Then, 5 × 105 cells with the dye, were slowly inoculated into the ventral prostate through the dorsal prostate at a 45° angle, avoiding the urethra. The success of inoculation could be verified by means of the green color. If leakage into the peritoneal cavity or urethra/bladder was observed, the mice were not included in the experiment. Inoculation was performed with a 30 G needle attached to a 25 μl glass syringe (both Hamilton Bonaduz AG, Bonaduz, Switzerland). After inoculation, the abdominal muscle layer was closed with a 4-0 absorbable suture (Bondek plus, polyglycolic acid-coated suture, Genzyme GmbH, Neu-Isenberg, Germany) and the skin with a 4-0 nonabsorbable suture (monofilament, polyamide suture, Genzyme GmbH, Neu-Isenberg, Germany).
In order to study the effects of alendronate on prostate tumor growth and invasion, mice were randomized according to weight into 2 groups. The orthotopic experiments were performed using the total numbers of 15 and 16/group considered to be large enough for the analyses of the statistical significance of the results. Mice in the alendronate group were daily treated with alendronate (0.05 mg/kg s.c., provided by Merck&Co., Inc., Whitehouse Station, NJ) in 100 μl PBS and mice in the control group were injected daily with 100 μl PBS. Alendronate and control treatments were started on the day of orthotopic inoculation of 5 × 10
5 PC-3 cells. The mice were sacrificed 4 weeks after inoculation. Prostate (including tumor) size was measured with calipers and the volume calculated according to the method described by Janik et al. [
31] as length × width × depth × π/6. The prostate lobes, selected internal organs (lungs, kidneys, adrenal glands, liver and spleen), regional lymph nodes (iliac and sacral), distant lymph nodes (inguinal, sciatic, axillary and brachial), hind limbs and vertebrae were macroscopically examined for the occurrence of tumors, excised and immersed in 4% neutral-buffered formalin. The hind limbs and vertebrae were radiographed and the bone samples were decalcified in 10% EDTA solution for two weeks before further processing.
Histology and immunohistochemistry
Formalin-fixed tissue samples were embedded in paraffin and 5-μm sections were cut and stained with hematoxylin and eosin (H&E) using standard techniques. The relative areas of necrosis in tumors and the relative area of tumor metastases in iliac and sacral lymph nodes was determined from H&E-stained sections using an analysis program for histomorphometry (AxioVision software, Carl Zeiss Microimaging GmbH, Oberkochen, Germany). The lymph nodes were cut through and 8 to 10 sections/lymph node were analyzed in order to determine metastatic area and the results were expressed as percentage of the total area of the lymph node.
The tumor sections were treated with antibodies: mouse monoclonal anti-LYVE-1 (a gift from Dr. David Jackson) [
32], rat monoclonal anti-CD34 (Santa Cruz Biotechnology Inc., Santa Cruz, CA) or mouse monoclonal anti-Ki-67 (Novocastra Laboratories Ltd., Newcastle upon Tyne, UK) o/n at +4°C. The samples were then treated with biotin-labeled goat anti-mouse secondary antibodies (Vector Laboratories, Burlingame, USA), biotin-labeled rabbit anti-rat secondary antibodies (DAKO Denmark A/S, Glostrup, Denmark), or biotin-labeled rabbit anti-mouse secondary antibodies [
33], respectively. A mouse-on-mouse kit (Vector Laboratories, Burlingame, CA) was used with Ki-67 antibody staining to inhibit non-specific staining of anti-mouse secondary antibody. The in situ end labeling (ISEL) method was used to detect apoptotic cells. Negative controls (sections of every sample stained without the primary antibody) were used to verify the specificity of staining. The relative numbers of Ki-67- and ISEL-positive cells were counted by means of a 10 × 10 grid using three sections at 500 μm intervals to make sure that the same cell was not counted twice. Altogether 3.000 cells/tumor from 3 different levels were counted. The length of the CD34-positive vessels was counted from the 3 most vascularized fields of each tumor by drawing lines following stained vessels and measuring the length of the lines using AxioVision software (Carl Zeiss Microimaging GmbH, Oberkochen, Germany). The sections of prostate-draining lymph nodes were studied for Ki-67, CD34 and ISEL immunostaining using the same method as described above. The results were blind-tested by three independent analyzers.
RNA isolation and Northern blot analysis
Total RNA was extracted from PC-3 cells using the guanidinium isothiocyanate method Chomczynski and Sacchi [
34]. Subsequently, 20 μg of RNA was separated by electrophoresis, stained with ethidium bromide and blotted on a GeneScreen Plus nylon membrane (NEN Research Products, Boston, MA), using standard conditions suggested by the manufacturer. A cDNA insert (kindly provided by Professor Kari Alitalo, University of Helsinki, Finland) [
35] of human VEGF-C was [
32P]-dCTP-labeled by the random priming method (Ready-to-go DNA labeling Beads, Pharmacia Biotech, Piscataway, NJ). Hybridization and exposure to X-ray film were carried out using conditions suggested by the manufacturer.
Western blot analysis
Serum-free DMEM conditioned by PC-3 cells for 2 days was harvested from the cultures. The conditioned medium was centrifuged to remove cell debris. Heparin-binding proteins were isolated from the supernatant with 100 μl of heparin-sepharose (1:1 slurry, Amersham Pharmacia Biotech., Piscataway, NJ) overnight at +4°C. Heparin-sepharose beads were sedimented by centrifugation and washed three times with 20 mM Tris-HCl, pH 7.4, 300 mM NaCl. Heparin-sepharose-bound proteins were extracted by means of 5-min incubation in Laemmli sample buffer at +95°C and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Medium conditioned by 3 × 106 cells was used per well. After transfer to nitrocellulose membranes (Bio-Rad), VEGF-C was detected using a 1:500 dilution of a goat polyclonal antihuman VEGF-C antibody (Santa Cruz Biotechnology, USA) and a secondary antibody, horseradish peroxidase-labeled anti-goat IgG (DAKO, Denmark), which was used at 1:2000 dilution. Protein bands were visualized using the ECL chemiluminescence detection system (Amersham Corp.). Conditioned media from human breast cancer cells (MCF-7 cells; 1 × 106/well), stably transfected with VEGF-C, were used as positive controls and normal goat IgG (R&B-Systems Inc., Minneapolis, MN) was used instead of anti-VEGF-C antibody as a negative control.
Statistics
Results were expressed as mean value ± SD. Statistical significance was taken to be p < 0.05, using two-tailed Student's t-test.
Discussion
In prostate cancer, 25–42% of patients have metastases in regional lymph nodes at the time of diagnosis [
2,
39]. Our modified orthotopic PC-3 tumor model, which invades to lymph nodes [
5], enables studies on this stage of prostate cancer metastasis. On the other hand, prostate cancer patients often have bone metastases and they are treated with bisphosphonates. Alendronate and other bisphosphonates have been reported to decrease the spread of various other tumors
in vivo. Our own experiments have also previously shown alendronate inhibition of prostate cancer cell invasion
in vitro. This prompted us to study the
in vivo growth and invasion of orthotopic prostate tumors modeling primary prostate cancer.
Alendronate is an aminobisphosphonate that is principally used in treatment of osteoporosis. Detailed molecular mechanisms have not been studied until recently. Alendronate has been shown to have inhibitory effects on osteoclast generation and maturation as well as on osteoclast activity [
40]. Although bone metastases of prostate cancer are usually more osteosclerotic than osteolytic, there are indications that bisphosphonates relieve bone pain and inhibit fractures in prostate cancer patients [
41‐
44]. Bisphosphonates thus seem to be beneficial in treatment of prostate cancer bone metastases, at least in subgroups with severe bone pain [
44]. In experimental models bisphosphonates have been shown to inhibit tumor invasion when used as a combination therapy with chemotherapeutics (taxol) [
45].
Our present results demonstrate that alendronate is able to inhibit prostate tumor growth and invasion to local lymph nodes
in vivo in nude mice. Surprisingly, a major effect of alendronate, besides being associated with a decreased size of lymph node metastases, was a decrease of tumor size. To our knowledge, alendronate has not earlier been reported to inhibit the growth of primary prostate cancer
in vivo. Alendronate treatment did not, however, decrease the rate of proliferation as shown by Ki-67 immunohistochemical staining of alendronate
vs. control-treated prostate tumors. This is in line with the results of our previous studies
in vitro, in which alendronate did not affect the proliferation of PC-3 cells over a wide range of concentrations that strongly inhibited invasion [
12]. Other studies have, however, revealed a decreased rate of proliferation
in vitro in the presence of alendronate but in those studies very high (micromolar) concentrations of alendronate were used [
19,
46,
47].
In our
in vivo study alendronate treatment did, however, increase apoptosis in prostate tumors, which could contribute to decreased tumor growth. This is in agreement with the results of some other studies in which
in vitro cultures [
16,
17] and
in vivo tumor models [
48] have been used. Relatively high concentrations of bisphosphonates have usually been needed to trigger apoptosis
in vitro but
in vivo bisphosphonate treatments with doses and schedules comparable to those used in clinical treatments have been reported to induce apoptosis in several models as regards both bone [
20] and visceral metastases [
19,
22‐
26]. This suggests that besides direct cellular effects, other mechanisms may also contribute to decreased cell survival
in vivo.
Another effect that may contribute to decreased tumor size and tumor cell survival in alendronate-treated mice is inhibition of angiogenesis, as suggested by a decreased density of CD34-positive blood vessel capillaries. This observation is also in accordance with earlier reports on the capacity of various bisphosphonates to inhibit angiogenesis
in vitro and
in vivo in the context of some physiological and pathological conditions. The decreased angiogenesis associated with bisphosphonate treatment has been related to modulation of endothelial cell proliferation, adhesion, migration [
49] and apoptosis [
19]. Zoledronate as well as alendronate has been found to inhibit secretion and activity of MMP-2 and MMP-9 by tumor cells and tumor-infiltrating macrophages in mouse models of cervical cancer [
19] and prostate cancer [
50]. In addition, both minodronate and zoledronate affect VEGF interaction with VEGF receptors and signaling in endothelial cells [
19,
23]. In our prostate tumors alendronate facilitated apoptosis and inhibited angiogenesis, the latter leading to restricted nutrition and oxygenation, thereby affecting tumor growth. There was, however, no statistically significant increase in the proportion of necrotic areas in alendronate-treated tumors (data not shown).
Orthotopic PC-3 tumors metastasize to sacral and iliac lymph nodes that drain the prostate [
5]. Immunostaining for LYVE-1 demonstrated a dense network of lymphatic capillaries in the periphery of tumors as well as in the peritumoral area in our PC-3 prostate tumors. A high level of expression of lymphangiogenic VEGF-C has been shown to correlate with the density of lymphatic capillaries [
49,
51‐
53] and increased lymphangiogenesis has been shown to facilitate tumor metastasis to lymph nodes [
51‐
55] although conflicting results have also been published [
56]. Our previous studies with orthotopic mammary tumors in nude mice, and those of others, have demonstrated that VEGF-C stimulates lymphangiogenic tumor spread
in vivo [
54,
55]. High expression level of VEGF-C has been detected in about 50% of human cancers [
49]. PC-3 cells and tumors express VEGF-C at a high level [
35,
37,
38], and it is thus conceivable that the spread of orthotopic PC-3 tumors to lymph nodes occurs primarily via lymphatic vessels. Demonstration of tumor growth inside lymphatic capillaries in our orthotopic tumors also speaks for a lymphatic route of metastasis. Treatment with alendronate reduced the metastatic area in lymph nodes but it did not, however, affect the density of the lymphatic vessel network or the number of lymph node metastases. Therefore, it is possible that the decreased metastatic growth in lymph nodes is explained by alendronate regulation of tumor growth via effects on tumor cells themselves or on angiogenesis, rather than via effects on lymphangiogenesis [
18,
49]. According to our results increased apoptosis of tumor cells and/or endothelial cells in both primary tumors and lymph node metastases could be a major mechanism by which alendronate decreased metastatic area in iliac and sacral lymph nodes. Our previous results demonstrated a strong inhibitory effect of alendronate on PC-3 cell invasion
in vitro [
12] and other investigations have shown inhibition of MMP secretion and associated invasion capacity of prostate cancer cells [
15]. These results are in line with the decreased cell invasion to and limited growth of PC-3 tumors in local lymph nodes of alendronate-treated mice. Decreased angiogenesis is also associated with decreased metastatic potential [
57] and could as such contribute to the decreased metastatic growth in lymph nodes.
Our results were based on analyses at an end-point of 4 weeks, which limits the estimation of alendronate effects on tumor formation and progression to this time point. Future experiments with noninvasive imaging methods will allow following growth and spreading of orthotopic tumors at several intervening time points over an optimal time period without markedly expanding the numbers of experimental animals.
There is evidence that the inhibitory effects of aminobisphosphonates are largely mediated via the mevalonate pathway [
12,
13,
40]. Although the precise mode of action of aminobisphosphonates is not understood, the mechanisms that have been recognized are based on their ability to impair post-translational prenylation of Ras, Rac and Rho [
12,
58]. Aminobisphosphonates have been found to target farnesylpyrophosphate and/or geranylgeranylpyrophosphate synthetase, which leads to decreased generation of farnesyl diphosphate and geranylgeranyl diphosphate. These intermediates are needed for post-translational prenylation of small GTP-binding proteins (Ras, Rac and Rho) which are essential for many cellular functions such as proliferation, survival and invasiveness. It is probable that interference with prenylation reactions is involved in many effects of aminobisphosphonates such as decreased angiogenesis and cell migration as well as increased apoptosis [
58].
Recent reports have also shown evidence that aminobisphosphonates can induce apoptosis and/or cause growth arrest to G2/M phase in the same way than pyrophosphate resembling bisphosphonates by production of an endogenous ATP analog [
59]. Especially newer aminobisphosphonates can directly inhibit proliferation [
19,
60]. However, these growth inhibitory effects can be overcome with excess geranylgeranyl pyrophosphate, which links this mechanism to the bisphosphonate inhibition of the mevalonate pathway.
Both aminobisphosphonates and pyrophosphate resembling bisphosphonates have also shown to increase resistance to the apoptotic and growth inhibitory effects in some cancer cells. These effects are associated with activation of p38 mitogen activated protein kinase pathway and may support cell survival and promote proliferation. Interestingly, the p38-mediated effects may not be dependent on the inhibition of the mevalonate pathway [
61]. It is possible that the p38-mediated survival effects balance the growth inhibitory effects in the tumor models as ours in which alendronate does not inhibit cell proliferation.
Bisphosphonates bind to calcified bone matrix and accumulate in bone very rapidly, which keeps blood levels of the drug low
in vivo. Studies carried out by Fournier
et al. [
18] have revealed, however, that in rats at least, some bisphosphonates including clodronate, zoledronate and ibandronate, transiently accumulate in the prostate rather than in several other non-calcified tissues before being trapped by bone. Prostatic levels of bisphosphonates reached a peak at 30–60 minutes after administration of the drug and then declined. The mechanism of bisphosphonate accumulation in tumors is not understood. The results of some studies have suggested that tumor calcification, necrosis or activated macrophages have a role in the uptake process [
58,
61‐
64]. Preferential prostatic accumulation may partly explain the anticancer effects of alendronate on prostatic tumors, and increase the possibilities of exploiting bisphosphonates in the development of therapies against prostate cancer.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JT carried out the cell cultures, RNA isolation, Northern Blot and Western Blot analyses, analysis of morphology and morphometry and statistical analysis. Orthotopic tumor experiments were done by JT and MV. PH and KV participated in the design of the study. JT wrote the first version of the manuscript and all authors helped to process it. All authors have read and approved the final manuscript. PH gave final approval for the manuscript to be submitted.