A therapeutic target for prostate cancer based on angiogenin-stimulated angiogenesis and cancer cell proliferation

Abstract

Human angiogenin is progressively up-regulated in the prostate epithelial cells during the development of prostate cancer from prostate intraepithelial neoplasia (PIN) to invasive adenocarcinoma. Mouse angiogenin is the most up-regulated gene in AKT-induced PIN in prostate-restricted AKT transgenic mice. These results prompted us to study the role that angiogenin plays in prostate cancer. Here, we report that, in addition to its well established role in mediating angiogenesis, angiogenin also directly stimulates prostate cancer cell proliferation. Angiogenin undergoes nuclear translocation in PC-3 human prostate cancer cells grown both in vitro and in mice. Thus, knocking down angiogenin expression in PC-3 human prostate adenocarcinoma cells inhibits ribosomal RNA transcription, in vitro cell proliferation, colony formation in soft agar, and xenograft growth in athymic mice. Blockade of nuclear translocation of angiogenin by the aminoglycoside antibiotic neomycin inhibited PC-3 cell tumor growth in athymic mice and was accompanied by a decrease in both cancer cell proliferation and angiogenesis. These results suggest that angiogenin has a dual effect, angiogenesis and cancer cell proliferation, in prostate cancer and may serve as a molecular target for drug development. Blocking nuclear translocation of angiogenin could have a combined benefit of antiangiogenesis and chemotherapy in treating prostate cancer.

Angiogenin is a 14-kDa angiogenic ribonuclease originally isolated from HT-29 colon adenocarcinoma cells (1). Its expression is up-regulated in various types of human cancers, including breast (2), cervical (3), colon (4), colorectal (5), endometrial (6), gastric (7), liver (8), kidney (9), ovarian (10), pancreatic (11), prostate (12), and urothelial (13) cancers, as well as astrocytoma (14), leukemia (acute myeloid leukemia and myelodysplastic syndrome) (15, 16), lymphoma (non-Hodgkin's) (17), melanoma (18), osteosarcoma (19), and Wilms' tumor (20). Among them, prostate cancer, in which angiogenin expression is positively correlated with disease progression (12), is of particular interest. Majumder et al. (21) reported that the angiogenin protein content in the serum of patients with hormone refractory prostate cancer (40 patients), in newly diagnosed prostate cancer patients (39 patients), and in control patients with no evidence of prostate cancer (37 patients) was 436 ± 24, 392 ± 17, and 328 ± 20 ng/ml, respectively. There is a statistically significant difference in serum angiogenin levels between the controls and untreated, hormone-naïve prostate cancer patients (P < 0.01) and between controls and hormone refractory prostate cancer patients (P < 0.001). There is also a trend toward higher levels of angiogenin in hormone refractory patients compared with newly diagnosed patients.

It is known that circulating angiogenin in normal plasma is mainly produced by the liver (22) and is at a concentration of 250–350 ng/ml (11, 13). Therefore, if the elevated serum angiogenin level in prostate cancer patients results from an up-regulation of prostatic expression of angiogenin, it reflects a dramatic increase in angiogenin expression in the prostate. Katona et al. (12) analyzed a large cohort of 107 radical prostatectomy specimens by immunohistochemistry (IHC) and found that angiogenin expression increases progressively as prostatic epithelial cells evolve from a benign phenotype to an invasive phenotype. This finding was in agreement with an earlier report in which Olson et al. (23) showed that angiogenin protein is barely detected in normal prostate tissue (7 patients) and is dramatically increased in prostate adenocarcinoma (10 patients).

The mouse has four angiogenin isoforms that show 76%, 39%, 62%, and 46% identity to human angiogenin (24). Mouse angiogenin-3 is prostate-specific (25) and is the most up-regulated gene detected in prostate epithelial cells in the AKT-induced prostate intraepithelial neoplasia (PIN) lesions observed in murine prostate-restricted Akt kinase transgenic (MPAKT) mice (21). Gene microarray analysis showed that mouse angiogenin-3 had the highest signal-to-noise score and was induced 32-fold in MPAKT mice (21). Mouse angiogenin-1 and -2 were also overexpressed with a 3- and 10-fold induction, respectively. Moreover, mTOR (mammalian target of rapamycin) inhibition reversed Akt-induced PIN in MPAKT mice and partially restored angiogenin-3 expression (26). These results suggest that angiogenin is involved in prostate cancer pathogenesis. In the present study, we examined the effect that knocking down angiogenin expression has on the growth of PC-3 cells both in vitro and in vivo. We have also examined the antitumor activities of neomycin, which blocks nuclear translocation of angiogenin in both endothelial and cancer cells. Our results suggest that angiogenin plays a role in both tumor angiogenesis and cancer cell proliferation during prostate cancer development.

Results

Enhanced Expression and Nuclear Translocation of Angiogenin in Prostate Cancer.

We have examined angiogenin expression levels in human prostate tissues from 23 prostate adenocarcinoma, 20 benign prostate hyperplasia (BPH), and 10 control patients. Fig. 1 shows representative images of IHC staining with the antiangiogenin mAb 26-2F. In normal prostate tissue, angiogenin is detectable in the stroma between the secretory glands (Fig. 1A). Angiogenin expression is significantly higher in BPH tissue (Fig. 1B) and even higher in prostate cancer tissue (Fig. 1C). No angiogenin was detected in the nucleus of the glandular epithelial cells in all 10 normal prostate tissue specimens. But strong nuclear staining was detected in the glandular epithelial cells in BPH samples (Fig. 1B) and in the invasive cancer cells of prostate cancer samples (Fig. 1C).

Fig. 1.

IHC staining of human angiogenin in prostate tissue samples. Prostate tissue samples from normal (A), BPH (B), and cancer (C) patients were stained with mAb 26-2F (30 μg/ml) and visualized with Dako's Envision kit. (A) Weak staining was observed in the stroma in the normal prostate tissue. (B and C) Enhanced staining was observed in BPH (B) and prostate cancer (C) tissues, with strong cytoplasmic and nuclear staining. (Magnification: ×400.)

Although angiogenin expression is elevated in many types of human cancer (2–7, 9–20), nuclear angiogenin has so far only been reported in breast cancer tissue samples (27). The strong nuclear angiogenin staining in prostate cancer samples (Fig. 1C) prompted us to survey nuclear angiogenin by IHC with a tissue microarray slide containing 35 different human cancer samples. Table 1lists the categorical results of angiogenin staining in the extracellular matrix (ECM), cytoplasm, and nucleus. Glioma, peripheral nerve sheath tumor, prostate cancer, and cervical cancer tissues have the strongest nuclear staining of angiogenin.

Table 1.

IHC staining of angiogenin in human cancer tissues

Cancer type	ECM	Cytosol	Nucleus
Glioma	+	+++	++++
Peripheral nerve sheath	+	+++	++++
Prostate	+++	+++	++++
Cervical	+++	+++	++++
Breast	+++	+++	+++
Colon	+++	+++	+++
Seminoma	+++	+++	+++
Renal cell	+++	+++	++
Skin squamous cell	+	+++	++
Rhabdomyosarcoma	+	++	++
Mesothelioma	−	++	++
Ewing's sarcoma	++	+	++
Endometrial	+	+	++
Ovarian yolk sac	++	−	++
Osteosarcoma	+++	+	++
Lung squamous	−	−	++
Thyroid	−	+++	+
Lung adenocarcinoma	++	++	+
Esophagus	+	++	+
Gastric	+++	+	+
Ovarian	+	+	+
Skin basal cell	+	+	+
Synovial sarcoma	−	+	+
Medulloblastoma	−	−	+
Liver	++	+	−
Gastrointestinal stroma	+	+	−
Leiomyosarcoma	−	+	−
Fibrosarcoma	−	+	−
Melanoma	−	+	−
Bladder	++	+	−
Neuroblastoma	++	−	−
Liposarcoma	−	−	−
Hodgkin's disease	−	−	−
B cell lymphoma	−	−	−
T cell lymphoma	−	−	−

The tumor types are listed according to the relative abundance of nuclear angiogenin. +, low; ++, moderate; +++, high; ++++, very high; −, not detectable.

To determine whether nuclear angiogenin observed in these cancer cells is associated with malignant transformation, we surveyed the distribution of angiogenin in a tissue microarray slide containing 30 normal human tissues. Table 2 shows that nuclear angiogenin was detectable in the normal cerebral cortex, peripheral nerve, pancreas, testis, tonsil, and other tissues. No appreciable nuclear angiogenin was detected in normal prostate and cervix samples, although angiogenin was detected in the ECM. Therefore, the prostate and cervix are the two histological sites where nuclear translocation of angiogenin is most significantly increased in tumorigenesis. Consistently, we have shown previously that knocking down angiogenin expression in HeLa cervical cancer cells inhibits cell proliferation and tumor growth in mice (28).

Table 2.

IHC staining of angiogenin in normal human tissues

Tissue type	ECM	Cytosol	Nucleus
Tonsil	++++	++	++
Liver	++++	−	−
Breast	++++	−	−
Testis	+++	+	++
Kidney	+++	++	++
Peripheral nerve	+++	+	+++
Esophagus	+++	−	−
Intestine	+++	−	−
Colon	+++	−	−
Prostate	+++	−	−
Cervix	+++	−	−
Ovary	+++	−	−
Thyroid	+++	−	−
Skin	++	+	++
Lung	++	+	−
Heart muscle	++	++	−
Stomach	++	+	+
Cerebellum	++	−	+
Skeleton muscle	++	−	−
Parathyroid	++	−	−
Cerebral cortex	+	+	+++
Pancreas	+	+	+++
Bone marrow	+	+	++
Uterus	+	+	−
Adrenal	+	−	+
Pituitary	+	−	−
Spleen	+	−	−
Omentum	+	−	−
Salivary	−	+++	+
Thymus	−	−	−

The tissues types are listed according to the relative abundance of angiogenin in the ECM because nuclear angiogenin was not significant in most of the samples. +, low; ++, moderate; +++, high; ++++, very high; −, not detectable.

Down-Regulating Angiogenin Expression in PC-3 Cells Inhibits rRNA Transcription and Cell Proliferation.

To examine the role of angiogenin in prostate cancer cell proliferation, we knocked down angiogenin expression in PC-3 cells by means of plasmid-mediated RNAi and measured the resultant changes in cell proliferation and tumorigenesis. pANG-RNAi targets human angiogenin mRNA at nucleotides 381–401 and has been used successfully in our previous work to knock down angiogenin expression in HeLa cells (28). As shown in Fig. 2A, stable transfection of pANG-RNAi in PC-3 cells decreased angiogenin expression from 0.65 (vector control pBS/U6 transfectants) to 0.15 ng per 10⁶ cells per day, representing a 77% reduction.

Fig. 2.

Down-regulation of angiogenin in PC-3 cells inhibits rRNA transcription and cell proliferation. PC-3 cells were transfected with an angiogenin RNAi plasmid, pANG-RNAi, or with the vector control pBS/U6. Stable transfectants were selected with 0.5 μg/ml puromycin for 2 weeks. (A) Secreted angiogenin levels determined by ELISA. (B) The steady-state level of 47S rRNA determined by Northern blotting with actin mRNA as the loading control. (C) Cell proliferation as determined with a Coulter counter. When present, exogenous angiogenin was 1 μg/ml.

We have observed previously that the function of nuclear angiogenin is related to rRNA transcription (29), a rate-limiting step in ribosome biogenesis. Knocking down angiogenin expression in endothelial cells (30) and in HeLa cells (28) reduces rRNA transcription, thereby inhibiting cell proliferation. Fig. 2B shows that down-regulation of angiogenin expression in PC-3 cells decreases the steady-state level of 47S rRNA (Fig. 2B, center lane) as determined by Northern blotting analysis. Exogenous angiogenin (1 μg/ml) was able to restore the 47S rRNA level to that of the vector control (Fig. 2B, right lane), indicating that the inhibitory effect was mediated by down-regulating angiogenin rather than by a nonspecific effect of plasmid transfection.

Because of the central importance of rRNA transcription in cell growth, decreased rRNA transcription should attenuate cell proliferation; we confirmed this in a cell proliferation assay (Fig. 2C). Cell number counting showed that pANG-RNAi transfectants have a reduced proliferation rate compared with that of the pBS/U6 vector transfectants. Again, exogenous angiogenin was able to restore cell proliferation to the level measured for the vector control.

Knocking Down Angiogenin Expression Decreases Tumorigenicity of PC-3 Cells.

The anchorage-independent growth of pANG-RNAi transfectants was analyzed by a colony formation assay in soft agar (Fig. 3A). Angiogenin RNAi transfection decreased colony number by 42%, from 1,397 ± 27 to 812 ± 42 (P < 0.0001). The colony size was also decreased from an average diameter of 111 ± 9 to 70 ± 10 μm (P < 0.001). A complete recovery in both the colony number and size was obtained when exogenous angiogenin (0.1 μg/ml) was added (Fig. 3A Right), indicating a specific role for angiogenin in anchorage-independent growth of PC-3 cells.

Fig. 3.

Knocking down angiogenin expression in PC-3 cells decreases tumorigenicity. (A) Soft agar assay in which cells were seeded at a density of 4 × 10³ cells per 35-mm dish and cultured in 0.35% soft agar in DMEM plus 10% FBS at 37°C for 7 days. When present, angiogenin was added to both the soft agar and the medium at 0.1 μg/ml. The colonies were stained with 0.05% crystal violet. Colony numbers in the entire dish were counted. The average colony size was determined by measuring the diameters of colonies in 10 microscope fields with a microcaliper. (B and C) Xenograft growth of PC-3 tumors in nude mice. The vector control (pBS/U6) and the angiogenin RNAi (pANG-RNAi) transfectants (1 × 10⁶ cells per mouse) were injected s.c. (eight mice per group) into 6-week-old male athymic mice. (B) Mice were checked daily for tumor appearance by palpation, and tumor volume was measured every 3 days. (C) Tumors were removed on day 31 and weighed.

The effect of knocking down angiogenin on in vivo growth of PC-3 cells was examined in a xenograft model in nu/nu mice. No palpable tumors were detected after 19 days in animals inoculated with angiogenin RNAi transfectants, whereas all of the animals inoculated with the vector control transfectants had tumors. Eventually, all of the animals in both groups developed tumors. However, the growth rate in the knockdown group was substantially lower than that in the control group (Fig. 3B). The animals were killed on day 31, and the tumors were removed. Fig. 3C shows that the average tumor weight in the pANG-RNAi mice was 96 ± 51 mg, representing an 82% reduction from that in the pBS/U6 mice (518 ± 52 mg) (Fig. 3C). These results demonstrate that down-regulation of angiogenin expression decreases the tumorigenicity of PC-3 cells.

IHC staining with mAb 26-2F shows that tumor tissue derived from the pANG-RNAi transfectants (Fig. 4B) has a significantly lower angiogenin protein level than that from the vector control transfectants (Fig. 4A). Nuclear accumulation of angiogenin is more obvious in the control tumors than in the RNAi transfectant tumors (Fig. 4 A and B). IHC staining with an anti-PCNA (proliferating cell nuclear antigen) antibody (Fig. 4 C and D) showed that the percentage of PCNA-positive cells decreased from 74 ± 5 to 37 ± 13 after angiogenin RNAi transfection. Neovessel densities in the vector control and RNAi transfected tumors, as shown by anti-VWF (von Willebrand factor) staining (Fig. 4 E and F), are 36 ± 3 and 15 ± 2 vessels per mm², respectively. These results indicate that both cell proliferation and tumor angiogenesis are decreased in the tumors derived from the pANG- RNAi transfectants.

Fig. 4.

IHC staining of angiogenin, PCNA, and neovessels. Thin sections (4 μm) from formalin-fixed, paraffin-embedded tumor tissues derived from vector-transfected PC-3 cells (A, C, and E) and from angiogenin RNAi-transfected PC-3 cells (B, D, and F) were stained with antiangiogenin (A and B), anti-PCNA (C and D), and anti-VWF (E and F) antibodies. The bound primary antibodies were visualized with Dako's Envision system. VWF-positive vessels in each tumor were counted in the five most vascularized areas at ×200 magnification, and the numbers were averaged. Vessel density (vessels per field) is shown as mean ± SD for each group. PCNA-positive and total numbers of cells were counted in five randomly selected areas at ×200 magnification. Images shown were from a representative animal of each group.

Prevention of Nuclear Translocation of Angiogenin Inhibits PC-3 Xenograft Tumor Growth in Mice.

We have shown that neomycin, an aminoglycoside antibiotic, prevents nuclear translocation of angiogenin in endothelial cells and inhibits its mitogenic and angiogenic activity (31). To determine whether blocking nuclear translocation of angiogenin will inhibit PC-3 cell tumor growth, we used an ectopic model to test the effect of neomycin on the growth of PC-3 cells in athymic mice. First, we confirmed that nuclear translocation of angiogenin in PC-3 cells is indeed blocked by neomycin (Fig. 5 A and B). Fig. 5C shows that treatment with neomycin s.c. at a dose of 60 mg/kg of body weight significantly delayed the establishment of PC-3 cell tumors in nude mice. By day 20, all of the animals in the control group had developed palpable tumors, whereas 7 of the 12 mice in the neomycin-treated group were still tumor-free. At the end of the experiment (56 days), half of the mice in the neomycin-treated group remained tumor-free. In those animals that did develop tumors, the average tumor weight was 58 ± 34 mg, which is 23% of that of the control group (253 ± 56 mg) (Fig. 5C). PCNA-positive cells in the tumor tissues from control and neomycin-treated animals were 75 ± 5% and 30 ± 6%, respectively, indicating that proliferation of PC-3 cells was inhibited by neomycin (Fig. 5 E and F). There was a concomitant decrease in tumor angiogenesis after neomycin treatment, as indicated by the neovessel densities determined by IHC staining with an anti-VWF antibody (Fig. 5 G and H). Neomycin treatment decreased the vessel density from 91 ± 4 to 26 ± 11, representing an 81% inhibition (the relatively high neovessel density in the control was due to the use of Matrigel in these experiments). These results demonstrated that blocking nuclear translocation of angiogenin effectively inhibits PC-3 cell tumor growth in mice, presumably through inhibition of both tumor cell proliferation and angiogenesis.

Fig. 5.

Effect of neomycin on PC-3 cell tumor growth in athymic mice. (A and B) Inhibition of nuclear translocation of angiogenin. PC-3 cells were cultured in DMEM plus 10% FBS for 24 h and then incubated with 1 μg/ml angiogenin in the absence (A) or presence (B) of 100 μM neomycin at 37°C for 30 min. Angiogenin was visualized with 26-2F and Alexa Fluor 488-labeled goat anti-mouse IgG. (C–H) Inhibition of tumor growth. PC-3 cells (5 × 10⁵ in 67 μl of Hanks' balanced salt solution) were mixed with 33 μl of Matrigel. The mixture was injected into the left shoulder of the mice. The mice then received s.c. injections of neomycin at a dose of 60 mg/kg of body weight or PBS daily for 2 weeks, followed by injections every other day for another 6 weeks. Twelve mice were used per group. (C) Mice were examined daily by palpation for tumor appearance. (D) At day 56, mice were killed, and tumor tissues were removed and weighed. (E and F) Tissue specimens were fixed in 10% formalin, and 4-μm paraffin sections were cut. Proliferating cells were stained with an anti-PCNA mAb. PCNA-positive cells and total numbers of cells were counted in five randomly selected areas at ×200 magnification. (G and H) Neovessels were stained with an anti-VWF antibody, and neovessels in each tumor were counted in the five most vascularized areas at ×200 magnification.

Discussion

The results presented here indicate that angiogenin plays a dual role in prostate cancer. In addition to a function in tumor angiogenesis, several lines of evidence show that angiogenin also mediates prostate cancer cell proliferation directly. Knocking down angiogenin expression in PC-3 cells inhibited cell proliferation by 65% (Fig. 2C). It is of note that the RNAi construct used in these experiments decreased angiogenin expression only by 82% (Fig. 2A). We have tested a number of RNAi sequences targeting different regions of angiogenin mRNA and found that, among stably transfected cells, an ≈80% reduction in angiogenin production was the maximum that could be obtained. Two of the five RNAi sequences failed to produce any transfectants because none of the cells survived. The reasons are not clear at present. One possibility is that angiogenin-mediated rRNA transcription is essential, so cells do not survive when angiogenin is inhibited beyond a certain degree. We have previously shown that inhibition of angiogenin expression beyond a certain level in endothelial cells resulted in cell death (30). We are currently generating prostate-specific angiogenin knockout/knockdown mice to study in detail the function of angiogenin in prostate development and in prostate cancer.

Angiogenin RNAi transfectants have reduced capacities to form colonies and to grow in soft agar (Fig. 3A). Thus, both anchorage-dependent and anchorage-independent growth of PC-3 cells was inhibited by down-regulating angiogenin expression. This inhibition can be alleviated completely by exogenous angiogenin (Figs. 2 B and C and 3A). These results indicate that the decreases in cell proliferation and colony formation result from decreased angiogenin expression rather than from a nonspecific action of RNAi. In addition, these results imply that exogenous angiogenin can functionally replace endogenous angiogenin in PC-3 cells. Angiogenin has a signal peptide, so the majority of the protein is secreted (32). To determine whether secretion is a necessary step for angiogenin to mediate cell proliferation, we examined the effect of 26-2F, a neutralizing mAb of angiogenin, on PC-3 cell proliferation. Addition of 26-2F to the culture medium inhibited PC-3 cell proliferation in a dose-dependent manner, whereas a subtype-matched nonimmune IgG had no effect (data not shown). These results suggest that an autocrine action of angiogenin accounts for at least part of the activity of angiogenin in mediating PC-3 cell proliferation.

The effect of angiogenin on PC-3 cell proliferation was confirmed by the in vivo results obtained in the xenograft tumor model (Fig. 3 B and C). The appearance of PC-3 cell tumors from angiogenin RNAi transfectants was delayed compared with that from the vector control transfectants. IHC staining for PCNA and VWF demonstrated that both cancer cell proliferation and tumor angiogenesis were substantially decreased in the tumors derived from the pANG-RNAi transfectants (Fig. 4). The role of angiogenin in angiogenesis induced by PC-3 cells in xenograft tumors was reported previously. Olson et al. (23, 33) showed that an anti-human angiogenin mAb and human angiogenin-specific antisense oligonucleotide inhibited the growth of PC-3 cells injected into the prostate of athymic mice. Based on the fact that angiogenin is a tumor angiogenic protein, the researchers concluded that the observed antitumor effects were due to inhibition of angiogenin-induced tumor angiogenesis. However, our results now suggest that inhibition of cancer cell proliferation may also contribute to the anticancer activity of angiogenin antagonists. We have repeated the experiments with the same antisense oligonucleotide (JF-2S) reported in these studies and found that it actually significantly inhibited PC-3 cell proliferation in vitro (data not shown).

IHC staining with 26-2F confirmed that the level of human angiogenin protein is decreased in pANG-RNAi tumors (Fig. 4 A and B). Consistent with our results with human prostate cancer tissues (Fig. 1), angiogenin was detected in the nucleus of the cancer cells derived from pBS/U6 vector control transfectants. Angiogenin staining is dramatically enhanced in prostate cancer tissues as compared with that in normal prostate tissues and in BPH. The other striking difference in the staining pattern of angiogenin is that there is strong nuclear staining in prostate cancer tissues. Tables 1 and 2 show that, among the various cancer types that we have surveyed, prostate and cervical cancers are the two types where nuclear accumulation of angiogenin is increased most significantly compared with normal tissue. Nuclear angiogenin is prominent in glioma and in malignant peripheral nerve sheath tumors. However, the normal cerebral cortex and normal peripheral nerve contain significant amounts of nuclear angiogenin. We believe our report to be unique in its indication that nuclear angiogenin is detectable in clinical prostate cancer tissue samples. In fact, breast cancer is the only other cancer type that has been reported to contain nuclear angiogenin (27).

The function of nuclear angiogenin is most likely related to rRNA transcription. In endothelial cells, angiogenin undergoes rapid translocation to the nucleus, where it binds to the promoter region of rRNA and stimulates rRNA transcription. An angiogenin-binding element has been identified and has been shown to have angiogenin-dependent promoter activity in a luciferase reporter assay. Consistent with the hypothesis that stimulation of rRNA transcription is one of the mechanisms by which angiogenin mediates prostate cancer growth, pANG-RNAi transfectants have reduced rRNA transcription that can be restored by adding exogenous angiogenin (Fig. 2B).

To investigate whether blocking nuclear translocation of angiogenin may have therapeutic value against prostate cancer, we examined the inhibitory activity of neomycin against xenograft growth of PC-3 cells in athymic mice. We have shown that neomycin blocks nuclear translocation of angiogenin, inhibits angiogenin-induced endothelial cell proliferation and angiogenesis (31), and suppresses xenograft growth of some human cancer cells (HT-29, MD-435, and A432) in athymic mice (34). Data presented in Fig. 5 demonstrate that neomycin blocks nuclear translocation of angiogenin in PC-3 cells and inhibits tumor establishment and growth in athymic mice, accompanied by a marked decrease in cancer cell proliferation and tumor angiogenesis.

Prevailing evidence suggests that the function of angiogenin in mediating both endothelial cell and cancer cell proliferation is related to rRNA transcription and depends on nuclear translocation (28, 30). Nuclear translocation of angiogenin in endothelial cells depends strictly on cell density. It decreases as the cell density increases and diminishes when cells are confluent (35). However, nuclear translocation of angiogenin in cancer cells seems to be density-independent (28). We have proposed that one of the reasons for this difference is that cancer cells are able to proliferate independently of cell density and thus require a constant supply of ribosomes. Constitutive nuclear translocation of angiogenin would meet this high metabolic demand of cancer cells. Blocking nuclear translocation of angiogenin in endothelial and cancer cells will inhibit tumor angiogenesis and prostate cancer cell proliferation, respectively, and thus will have a combined benefit of chemotherapy and antiangiogenesis therapy. We are currently evaluating the therapeutic value of neomycin and neamine, the nontoxic derivative of neomycin, against spontaneous prostate cancers in mice.

Materials and Methods

Cells and Transfectants.

PC-3 cells were cultured in DMEM plus 10% FBS. The cells were transfected with pANG-RNAi, an angiogenin RNAi plasmid that has been shown to knock down angiogenin expression in HeLa cells (28). This plasmid or the control vector pBS/U6 was cotransfected with pBabe-puro, a plasmid containing the puromycin resistance gene, into PC-3 cells in the presence of Lipofectin. Stable transfectants were selected with 0.5 μg/ml puromycin for 2 weeks. Pooled populations of the vector and the RNAi transfectants were used in this study.

Cell Proliferation Assays.

Anchorage-dependent cell proliferation was determined by counting the cell numbers with a counter from Coulter (Hialeah, FL). Anchorage-independent cell proliferation was determined by a soft agar assay. Cells were seeded at a density of 4 × 10³ cells per 35-mm cell culture dish in 0.35% agar and cultured for 7 days at 37°C under 5% CO₂. Dishes were stained with 0.05% crystal violet overnight at 4°C. Colonies were counted in the entire dish, and the colony size was determined by a microcaliper. The two-tailed Student t test was used to determine the differences between the groups. When exogenous angiogenin was present, it was added when the cells were seeded and was replenished every time the medium was changed.

ELISA Detection of Human Angiogenin.

A double-antibody ELISA method (36) was used to measure the angiogenin content in the medium. Cell culture media were collected, and the volumes were normalized to the cell numbers. ELISA plates were coated with 1 μg of antiangiogenin mAb 26-2F per well and blocked with 5 mg/ml BSA in PBS. Samples (100 μl) were added to the wells, and the plates were incubated at 4°C overnight, washed with PBS five times, and incubated with 100 μl of antiangiogenin polyclonal antibody R112 per well (1:4,000) at room temperature for 2 h. The plate was washed four times with PBS and incubated with an alkaline phosphatase-labeled goat anti-rabbit antibody (1.25 μg/ml) at room temperature for 1 h. After washing four times with PBS, 100 μl of 5 mg/ml p-nitrophenyl phosphate in 0.1 M diethanolamine containing 10 mM MgCl₂ (pH 9.8) was added, and the absorbance at 410 nm was measured. A standard curve of recombinant human angiogenin at concentrations ranging from 50 to 1,000 pg per well was performed each time on every plate.

Xenograft Growth of PC-3 Cell Tumors in Athymic Mice.

Outbred male athymic mice (nu/nu) were obtained from Charles River Laboratories (Wilmington, MA). The vector control and the RNAi transfectants (1 × 10⁶ cells per mouse) were injected s.c. in the right shoulder. Eight mice per group were used. Tumor sizes were measured every 3 days and recorded in mm³ (length × width²). Mice were killed on day 31, and the wet weight of the PC-3 tumor was recorded. For tumor therapy experiments, the mice were inoculated s.c. with 100 μl of a mixture containing 5 × 10⁵ PC-3 cells and 33 μl of Matrigel. The mice were treated s.c. with neomycin (60 mg/kg) or PBS daily for 2 weeks and then every other day until day 56, at which time the animals were killed and the tumors were removed. Twelve mice were used in each group.

IHC.

Microarray slides of human prostate cancer and BPH tissues, human multitumor tissues, and human normal tissues were from Zymed Laboratories (South San Francisco, CA). PC-3 cell tumors were fixed in 10% formalin and embedded in paraffin, and thin sections (4 μm) were cut. The tissue microarray slides and the thin sections of PC-3 cell tumor were deparaffinized with xylene, rehydrated in ethanol, and microwaved for 15 min in 10 mM citrate buffer (pH 6.0). Endogenous peroxidase was blocked by treatment with 0.3% H₂O₂ in methanol for 30 min. The slides were blocked in 5% dry milk for 10 min, incubated with the primary antibodies at 4°C overnight, and visualized with the Envision system (Dako, Carpinteria, CA). The slides were counterstained with hematoxylin. Negative controls were obtained by omission of the primary antibodies. Angiogenin staining was performed with 26-2F as the primary antibody at a concentration of 30 μg/ml. Proliferating cells were stained with an anti-PCNA mAb (PC10, Dako) at 1:200 dilution. The number of PCNA-positive cells and the total number of cells were counted in five randomly selected areas at ×200 magnification. Neovessels were stained with a polyclonal anti-human VWF antibody (Dako) at 1:200 dilution. VWF-positive vessels in each tumor were counted in the five most vascularized areas at ×200 magnification (i.e., with a ×20 objective lens and a ×10 ocular lens; 0.785 mm² per field).

Abbreviations

IHC

immunohistochemistry

BPH

benign prostate hyperplasia

ECM

extracellular matrix

PCNA

proliferating cell nuclear antigen

VWF

von Willebrand factor.