Expression of lnterleukin-8 Correlates with Angiogenesis, Tumorigenicity, and Metastasis of Human Prostate Cancer Cells Implanted Orthotopically in Nude Mice

Abstract

We determined whether the expression of interleukin-8 (IL-8) by human prostate cancer cells correlates with induction of angiogenesis, tumorigenicity, and production of metastasis. Low and high IL-8-producing clones were isolated from the heterogeneous PC-3 human prostate cancer cell line. The secretion of IL-8 protein correlated with transcriptional activity and levels of IL-8 mRNA. All PC-3 cells expressed both IL-8 receptors, CXCR1 and CXCR2. The low and high IL-8-producing clones were injected into the prostate of nude mice. Titration studies indicated that PC-3 cells expressing high levels of IL-8 were highly tumorigenic, producing rapidly growing, highly vascularized prostate tumors with and a 100% incidence of lymph node metastasis. Low IL-8-expressing PC-3 cells were less tumorigenic, producing slower growing and less vascularized primary tumors and a significantly lower incidence of metastasis. In situ hybridization (ISH) analysis of the tumors for expression of genes that regulate angiogenesis and metastasis showed that the expression level of IL-8, matrix metalloproteinases, vascular endothelial growth factor (VEGF), and E-cadherin corresponded with microvascular density and biological behavior of the prostate cancers in nude mice. Collectively, the data show that the expression level of IL-8 in human prostate cancer cells is associated with angiogenesis, tumorigenicity, and metastasis.

Introduction

Prostate cancer is the most prevalent cancer in men in North America and the second most deadly [1,2]. Its incidence is continuing to increase in the United States, and over 180,100 newly diagnosed cases and 31,900 cancer-related deaths are estimated for the year 2000 [1]. The major cause of death from prostate cancer is metastases that are resistant to therapy. These metastases are diagnosed in approximately 24% of patients at initial presentation and are later discovered in an additional 30% of patients as a result of surgical staging. That the currently used systemic therapy is inefficient is evident in that only 20% of patients with distant (stage D2) metastases survives for 5 years. Further, despite earlier and better diagnosis, metastasis may develop in a significant number of patients with clinically localized disease treated with radical prostatectomy [3–6].

The process of metastasis selects for cells that invade, embolize, survive in the circulation, arrest in a distant capillary bed, and extravasate into, and multiply within the organ parenchyma [7–9]. The outcome of metastasis depends on multiple interactions (cross talk) between metastatic cells and homeostatic mechanisms, which the tumor cells can usurp [8–10]. A crucial step in the continuous growth of tumors and development of metastases is the recruitment of new blood vessels in and around tumors [11–13]. A tumor mass that is <0.2 mm in diameter can receive oxygen and nutrients by diffusion [12], but to grow further, the tumor requires angiogenesis. Angiogenesis is triggered by a change in the local equilibrium between positive and negative regulatory molecules [14]. Some of the major proangiogenic factors include basic fibroblast growth factor (bFGF), vascular endothelial growth factor/vascular permeability factor (VEGF/VPF), and interleukin-8 (IL-8) [15,16]. Some of the major antiangiogenic molecules include interferon (IFN), angiostatin, endostatin, and thrombospondin [16,17]. We recently reported that the expression of metastasis- and angiogenesis-regulating genes directly correlates with the metastatic potential of human prostate cancer cells (and clones) implanted into the prostates of nude mice. Specifically, the expression of two genes that control invasion and angiogenesis, MMP-9 and IL-8, showed excellent correlation with metastasis [18].

IL-8, which belongs to the superfamily of CXC chemokines, has a wide range of proinflammatory effects and is produced by a wide range of cells, including lymphocytes, monocytes, endothelial cells, fibroblasts, hepatocytes, and keratinocytes [19]. Recent data also demonstrate that IL-8 is produced by various tumor cells, including human ovarian cancer cells [20,21], prostate cancer cells [28], and gastric cancer cells [22,23]. Two receptors for IL-8, CXCR1 and CXCR2 have been identified, and cDNAs for each receptor have been cloned [24,25].

IL-8 was initially described as a neutrophil chemoattractant [26]. It has also been shown to be an autocrine growth factor for keratinocytes, melanoma cells, and human liver and pancreatic cancer cells [27]. IL-8 expression correlates with angiogenesis of human gastric carcinomas [22,23] and colon cancer [28]. Moreover, IL-8 has been shown to enhance production and secretion of collagenase type IV by tumor cells [29], suggesting that it can modulate invasiveness and/or extracellular matrix remodeling in the tumor environment. As cell proliferation, angiogenesis, migration, and invasion are all important components of the metastatic process, IL-8 expression by tumor cells can influence their metastatic capabilities [30]. Indeed, the expression of IL-8 has been shown to correlate with the metastatic potential of human melanoma cells [30], human ovarian cancer cells [20,21,31], human gastric carcinoma cells [23], and human prostate cancer cells [18,32].

The major aim of this study was to determine whether the expression of IL-8 by clonal populations of human prostate cancer cells correlates with induction of angiogenesis, tumorigenicity, and metastasis subsequent to implantation into the prostate of nude mice. The data show that endogenous IL-8 modulates expression of several metastasis-regulating genes, including MMP-2, MMP-9, E-cadherin, and VEGF in tumor cells, and consequently influences angiogenesis, invasion, and metastasis.

Materials and Methods

PC-3 Human Prostate Cell Line, Metastatic Variant, and Clonal Populations

The PC-3 human prostate cancer cell line was originally obtained from the American Type Culture Collection (Rockville, MD). The PC-3M cell line was derived from a liver metastasis produced by the parental PC-3 cells growing in the spleen of a nude mouse. PC-3M cells were implanted orthotopically into the prostate of nude mice, and after several cycles of in vivo selection, the highly metastatic PC-3 MM2 line was isolated [33].

The PC-3 parental (PC-3P) and PC-3 MM2 lines were maintained as monolayer cultures in RPMI-1640 (Celox Corp., Hopkins, MN) supplemented with 10% fetal bovine serum, sodium pyruvate, nonessential amino acids, l-glutamine, a two-fold vitamin solution (Gibco, Grand Island, NY), and penicillin-streptomycin (Flow Laboratories, Rockville, MD). Cell cultures were maintained on plastic and incubated in 5% CO₂/95% air at 37°C. Cultures were free of Mycoplasma and the following murine viruses: reovirus type 3, pneumonia virus, K virus, Theiler's encephalitis virus, Sendai virus, minute virus, mouse adenovirus, mouse hepatitis virus, lymphocytic choriomeningitis virus, ectromelia virus, and lactate dehydrogenase virus (assayed by M.A. Bioproducts, Walkersville, MD). The PC-3P line was cloned by a limited dilution technique, and clonal populations were maintained in culture exactly as described above.

In Vitro Production of IL-8

The production and secretion of IL-8 by PC-3P, PC-3 MM2, and different clones were determined 24 hours after plating 1x10⁵ cells in 300 µl of medium into 38-mm² wells (96-well plates). The supernatants of four wells from each plate were collected and analyzed for level of IL-8 using a commercially available enzyme-linked immunosorbent assay (ELISA) kit from R&D Systems (Minneapolis, MN). The concentration of IL-8 was standardized by cell number. The cells were defined as low or high IL-8 producers. Three clones expressing the highest level of IL-8 protein (>20 ng/ml) were pooled and three clones expressing the lowest level of IL-8 protein (<1.0 ng/ml) were pooled. The expression and production of IL-8 by the PC-3P, PC-3 MM2, and low- and high-expressing PC-3 lines were confirmed by both ELISA and Northern blot analysis.

RNA Isolation and Northern Blot Analysis

Total RNA was extracted from each cell line using TRI Reagent (Invitrogen Co., San Diego, CA). Twenty micrograms of total RNA was electrophoresed on a 1 % denaturing formaldehyde agarose gel, transferred to a GeneScreen nylon membrane (DuPontCo., Boston, MA), and UV cross-linked with 120,000 µJ/cm² using a UV Stratalinker 1800 (Stratagene, La Jolla, CA). Hybridizations were performed as described previously [21]. Nylon filters were washed at room temperature and 65°C with washing solution. Membranes were exposed to Kodak X-ray film.

Reverse Transcription Polymerase Chain Reaction (RT-PCR) for IL-8 Receptors

The presence of the IL-8 receptors, CXCR1 and CXCR2, was determined by RT-PCR using a commercial kit (Access RT-PCR System; Promega, Madison, Wl). One microgram of total RNA extracted from PC-3P, PC-3 MM2, and high and low IL-8-expressing lines was used for RT-PCR. RT was done with antisense primer at 48°C for 45 minutes to synthesize the first strand of cDNA followed by inactivation of RT at 94°C for 2 minutes. PCR was performed with the following conditions: denaturation at 94°C for 1 minute, annealing at 58°C for 1 minute, and extension at 68°C for 2 minutes. The reaction was cycled 40 times and completed with an additional extension after the cycles at 68°C for 7 minutes. The primers were sense 5′-AGT TCT TGG ACA GTC ATC G-3′ and antisense 5′-CTT GGA GGT ACC TCA ACA GC-3′ for CXCR1 and sense 5′-ACA TTC CTG TGC AAG GTG G-3′ and antisense 5-CAG GGT GAA TCC GTA GCA GA-3′ for CXCR2.

Mitogenic Activity of IL-8

To determine whether IL-8 is an autocrine growth factor, 2x10³ PC-3 cells (PC-3P, PC-3 MM2, PC-3 low IL-8, and PC-3 high IL-8) were seeded into 38-mm² wells of 96-well plates in triplicate and allowed to adhere overnight. The spent medium was removed and replaced with medium alone, medium containing 100 ng/ml of recombinant human monocyte IL-8 (Biosource, Camarillo, CA), or medium containing 50 µg/ml of murine monoclonal anti-human IL-8 neutralizing antibody (Biosource). At different time points, the number of metabolically active cells was determined by tetrazolium salt (MTT) assay [34]. Tetrazolium (MTT, M2128) was purchased from Sigma Chemical Co. (St. Louis, MO), and a stock solution was prepared by dissolving 5 mg of MTT in 1 ml phosphate-buffered saline (PBS) and filtering the solution to remove particulates. The solution was protected from light and stored at 4°C. Following a 2- to 4-hour incubation period in medium containing 0.42 mg/ml of MTT, the cells were lysed in dimethyl sulfoxide. An Mr-5000 96-well microtiter plate reader at 750 nm (Dynatech Inc., Chantilly, VA) monitored the conversion of MTT to formazan by metabolically viable cells. Growth inhibition was calculated from the formula:

$$\text{Cytostasis (\%) = [1}-(\text{A/B})]\times 100$$ 1

where A is the absorbance of treated cells and B is the absorbance of the control cells.

Nuclear Run-On Assay

PC-3 cells (low and high IL-8 expressors) (1x10⁷) were seeded into separate 150-mm tissue culture dishes and incubated overnight at 37°C. The nuclei were isolated and aliquoted. For the in vitro transcription reaction, 100 µl of nuclei from each sample was mixed with an equal volume of 2x reaction buffer containing 5 mM MgCl₂, 5 mM DTT, 150 mM KCl, 50 mM HEPES (pH 7.4), 0.7 mM concentrations of ATP, CTP, and GTP, and [α-³²P]UTP (total 3000 Ci/mmol; Amersham Corp., Arlington Heights, IL). The reaction was incubated at 30°C for 30 minutes, and ³²P-labeled RNA was then isolated and precipitated with ethanol. Labeled nuclear RNA was hybridized with dot blots containing the IL-8 and α-tubulin inserts at 65°C for 72 hours. The filters were washed twice with 2x SSC and then exposed to X-ray film at -80°C for 1 day. Quantitative analysis was done by densitometry and standardized to α-tubulin. Probes used for hybridization were 0.5-kb EcoRI cDNA fragment corresponding to human IL-8 (a generous gift from Dr. K. Matsushima, Kanazawa University, Kanazawa, Japan), and 1.4-kb cDNA fragment corresponding to α-tubulin (Stratagene, 936205).

mRNA Stability Assay

Stability of mRNA was compared by treating cells with actinomycin D as described by Lindholm et al. [35]. PC-3 cells were cultured overnight. Fresh medium containing 5 µg/ml actinomycin D (Sigma) was added to the cultures to block transcription. After 12 and 24 hours, total RNA was extracted and Northern blot analysis was performed as described. The amounts of IL-8 mRNA and α-tubulin were quantified by densitometry.

Animals

Male athymic BALB/c nude mice were purchased from the Animal Production Area of the National Cancer Institute, Frederick Cancer Research Facility (Frederick, MD). The mice were housed in laminar flow under specific pathogen-free conditions and used at 8 weeks of age. Animals were maintained in facilities approved by the American Association for Accreditation of Laboratory Animal Care in accordance with current regulations and standards of the United States Department of Agriculture, Department of Health and Human Services, and National Institutes of Health.

Orthotopic Implantation of PC-3 Cells in Nude Mice

For all in vivo experiments, tumor cells in exponential growth phase were harvested after a brief exposure to 0.25% trypsins:0.1% EDTA solution (w/v). The dishes were tapped sharply to dislodge the cells, 10% minimum essential medium was added, and the cell suspension was pipetted to obtain single-cell suspensions and counted. The cells were washed and resuspended in Ca²⁺- and Mg²⁺-free HBSS at 5x10⁵ cells/50 µl. Cell viability was determined by trypan blue exclusion, and only single-cell suspensions of >95% viability were used as implants in the prostate of the mice as described previously [33]. In brief, nude mice were anesthesized with Nembutal (Abbott Laboratories, North Chicago, IL) and placed in a supine position. A low midline incision was made and the prostate was exposed. Fifty microliters of HBSS containing 5x10⁵ cells was injected into a lateral lobe of the prostate. The wound was closed with surgical metal clips.

Necropsy Procedures and Histological Studies

The mice were euthanized at 3, 4, and 5 weeks after tumor cell injection, and the body weights were determined. Primary tumors in the prostate were excised, measured, and weighed. For immunohistochemistry and H&E staining procedures, one part of the tumor was formalin-fixed and paraffin-embedded, another part was embedded in OCT compound (Miles Inc., Elkhart, IN), rapidly frozen in liquid nitrogen, and stored at -70°C. Macroscopically enlarged regional lymph nodes were harvested and the presence of metastatic disease was confirmed by histology.

In Situ Hybridization (ISH)

To determine the expression of metastasis-regulating genes (IL-8, MMP-2, MMP-9, VEGF/VPF, E-cadherin), prostate tumors from mice injected orthotopically with PC-3 cells that express low or high levels of IL-8 were harvested, fixed in formalin, and embedded in paraffin. Tumor sections (4–5 µm) were used for ISH to identify specific mRNA using the Microprobe manual staining system (Fisher Scientific, Pittsburgh, PA) as described previously [18,36]. The probes used in our study were: 5′-CTC CAC CCA CCT CTG CAC CC-3′ (1:200) for IL-8; 5′-CCG GTC CAC CTC GCT GGC GCT CCG GA-3′ (1:200) for MMP-9; 5′-TGG TGA TGT TGG ACT CCT CAG TGG GCU-3′ (1:200) for VEGF; 5′-CGG GAA GCC GCC GCT GCC GCC-3′ (1:200) for bFGF; and 5′TCC AGC GGG CTG GAG TCT GAA CTG-3′ and 5′-GAC GCC GGC GG C CCC TTC ACA GTC-3′ for E-cadherin. Hybridization (of each probe) was carried out for 45 minutes. The samples were then washed three times for 2 minutes with 2x SSC at 45°C and incubated for 30 minutes at 45°C with alkaline phosphatase-labeled avidin, rinsed in 50 mmol/l Tris buffer (pH 7.6), rinsed for 1 minute with alkaline phosphatase enhancer, and incubated for 15 minutes at 45°C with a chromogen substrate. A positive reaction in this assay stained red. A poly d(T)₂₀ oligonucleotide was used to verify the integrity and lack of degradation of mRNA in each sample, and controls for endogenous alkaline phosphatase included treatment of the sample in the absence of the biotinylated probe and use of chromogen alone. The identical procedure without probes was carried out on each sample to provide a background for densitometry quantification. The intensity of ISH reaction was evaluated in five x100 fields at the periphery of the prostate tumors representing the areas of most staining. Each x100 field was evaluated using the ImageQuant analyzer and Optima software program (Bioscan, Edmonds, WA). The results were normalized for background and poly d(T) expression, and then the intensities of tumor sections produced by low or high IL-8-producing cells were measured.

Quantification of Microvascular Density in Prostate Cancers

Cryostat sections of tumors were fixed for 10 minutes at room temperature with 2% paraformaldehyde in PBS (pH 7.5), washed twice with PBS, and treated with 1% Triton X-100 for 5 minutes. The sections were then washed three times for 20 minutes at room temperature with PBS containing 1% normal goat serum and 1% horse serum. Excess blocking reagent was drained off and the samples were incubated for 18 hours at 4°C with anti-CD31 antibody [37]. The samples were rinsed three times with PBS and incubated for 1 hour at room temperature with a secondary antibody (goat anti-rat; BD Pharmingen, CA). The samples were rinsed four times with PBS, then rinsed with distilled water, and incubated for 20 minutes at room temperature with diaminobenzidine (Research Genetics, Huntsville, AL). The sections were then washed three times with distilled water, counterstained with aqueous hematoxylin, washed, mounted with Permount (Research Genetics), and examined using a bright-field microscope. A positive reaction was indicated by a reddish brown precipitate in the cytoplasm. Any brown-staining endothelial cell cluster distinct from adjacent microvessels, tumor cells, or other stromal cells was considered to be a single microvessel [38]. Negative controls consisted of samples immunostained with nonspecific IgG. Areas containing the highest number of capillaries and small venules were identified by scanning tumor sections at low power (x40). After at least five areas of dense vascularization were identified, individual vessels were counted in x100 fields (x10 objective and x10 ocular; 0.14 mm²/field) and standardized to number of vessels per 1 mm² (number/mm²). On the basis of the criteria described by Weidner et al., [38] vessel lumens were not required for a structure to be classified as a vessel. All vessel counts were performed on coded samples by two investigators.

Statistical Analysis

The significance of the in vitro data was analyzed using the Student's t test (two-tailed), and the in vivo data were analyzed using the Mann-Whitney U test.

Results

Isolation of PC-3 Clones with Different Levels of IL-8 Expression

The PC-3P was cloned by limited dilution. Approximately 100 clones were isolated. The production and release of IL-8 into culture supernatants were determined by ELISA. The range of IL-8 production by the clones varied from <1.0 to >20 ng/ml (Figure 1). Three clones producing <1.0 ng/ml IL-8 were combined into one cell line designated as low IL-8. Three clones producing >20 ng/ml of IL-8 were combined into one cell line designated as high IL-8. The high IL-8 line expressed a greater than 20-fold level of IL-8 protein (Figure 1) and a 15-fold level of IL-8 mRNA as compared to the low IL-8 cells (Figure 2A).

Figure 1.

Production of IL-8 protein by clones isolated from the PC-3 parental line. Clones were isolated by a limited dilution technique and expanded in culture. Cells were plated into wells and culture supernatants were harvested 48 hours later. Level of IL-8 was determined by ELISA. Arrows identify clones producing low or high levels of IL-8 protein.

Figure 2.

IL-8 expression in PC-3 lines. (A) Northern blot analysis. PC-3 IL-8 high cells had 15.3-fold higher expression of IL-8 mRNA than PC-3 IL-8 low cells. (B) Nuclear run-on assays. IL-8 transcriptional activity in the PC-3 IL-8 high cells was 4.4-fold higher than in the PC-3 IL-8 low cells.

The highly metastatic PC-3 MM2 cell line was also cloned by a limited dilution method. Fifty clones were isolated. The production of IL-8 protein by all the clones ranged from 10 to 20 ng/ml (data not shown), suggesting that this line is homogeneous for production of IL-8.

Transcriptional Regulation of IL-8

Next, using a nuclear run-on assay, we determined whether the different levels of IL-8 production by the low and high IL-8 lines were due to transcriptional regulation. The data shown in Figure 2B demonstrate that, consistent with the ELISA data, the transcription rate of IL-8 in the PC-3 IL-8 high cells was 4.4-fold higher than in the PC-3 IL-8 low cells. To determine whether the increased production of IL-8 in the PC-3 IL-8 high cells was due to increased IL-8 mRNA stability, we compared the half-life of mRNA in cultures of PC-3 IL-8 low and IL-8 high cells. Cells were treated with actinomycin D to stop transcription. mRNA was collected 12 and 24 hours later and Northern blot analysis was performed. No discernible differences were found in stability of mRNA between the PC-3 IL-8 low and high cells (data not shown).

Expression of IL-8 Receptors

The RT-PCR technique was used to detect the presence of the two IL-8 receptors, CXCR1 and CXCR2. The data shown in Figure 3 demonstrate that all cells examined (PC-3P, PC-3 MM2, PC-3 IL-8 low, and PC-3 IL-8 high) expressed CXCR1 (265 bp) and CXCR2 (361 bp). The bands were purified from the gel, and their identify confirmed by sequencing analysis.

Figure 3.

Expression of IL-8 receptors in PC-3 cells. The expression of the IL-8 receptors, CXCR1 (265 bp) and CXCR2 (361 bp), was determined by RT-PCR carried out in the absence of reverse transcriptase as negative control to rule out DNA contamination. Note that all PC-3 cells expressed both receptors for IL-8 receptors. P, PC-3P; M, PC-3 MM2; H, PC-3 IL-8 high; L, PC-3 IL-8 low; N, negative control.

Mitogenic Activity of IL-8

Since all PC-3 cells expressed the receptors for IL-8, we next determined whether IL-8 is an autocrine growth factor for the cells. In this set of in vitro experiments, PC-3P, PC-3 MM2, PC-3 IL-8 low, and PC-3 IL-8 high cells were incubated in medium alone, in medium containing 100 ng/ml recombinant IL-8, or in medium containing 50 µg/ml mouse antihuman IL-8 neutralizing antibody. Cell proliferation was determined after 24, 48, and 72 hours of incubation using the MTT assay [34]. The increase in cell number was expressed as percent proliferation over control. The data shown in Table 1 do not demonstrate that the addition of r-IL-8 to any of the cells significantly enhanced proliferation. Moreover, the presence of neutralizing antibodies did not decrease proliferation of any of the PC-3 lines. The data, therefore, do not support the possibility that IL-8 is an autocrine growth factor for the PC-3 cells.

Table 1.

Mitogenic Activity of IL-8.

Cells*	Medium Content
	Control	r-IL-8 (100 ng/ml)	Neutralizing Antibodies (50 µg/ml)
PC-3P	365±1†	414±28	408±22
PC-3 MM2	374±13	445±13	367±9
PC-3 IL-8 low	333±25	369±35	379±15
PC-3 IL-8 high	365±15	382±10	337±14

^*Two thousand cells were seeded (in triplicate) into 38-mm² wells. Cell proliferation at 24, 48, and 72 hours was determined by the MTT assay. The data shown are for 72 hours.

^†Mean±SD of triplicate cultures. The numbers represent percent increase over cells plated per well. No discernible differences were found among the groups.

Tumorigenicity and Production of Metastasis

To determine the relative tumorigenic potential of the PC-3 IL-8 low and PC-3 IL-8 high cells, we injected increasing numbers of cells into the prostate of groups of nude mice (n=5). As shown in Table 2, the minimal number of PC-3 IL-8 low cells necessary to produce 100% take was 2x10⁵. In contrast, the minimal number of PC-3 IL-8 high cells necessary to produce 100% takes was 5x10⁴.

Table 2.

Tumorigenic Potential of PC-3 Cells with Low or High Expression of IL-8.

Number of Cells Injected*	PC-3 IL-8 low		PC-3 IL-8 high
	Incidence†	Tumor Weight Median (range)	Incidence†	Tumor Weight Median (range)
1.25x104	0/5	0	0/5	0
2.5x104	0/5	0	0/5	0
5x104	1/5	120	4/4	170 (110–180)
1x105	1/5	100	5/5	220 (100–240)
2x105	4/4	180 (140–200)	5/5	670 (420–780)
5x105	5/5	230 (190–270)	5/5	860 (560–1040)

^*Nude mice (n=5) were given prostate injections of the indicated number of PC-3 IL-8 low or high cells. The mice were killed 5 weeks later, at which point the prostate and tumors were removed and weighed (tumor weight is in milligrams).

^†Number of positive mice/number of mice injected.

In the next set of experiments, PC-3 cells were injected into the prostate of nude mice at the inoculum dose of 2x10⁵. The mice were killed when moribund (PC-3 MM2) or 3, 4, and 5 weeks after implantation. At each time point, each group consisted of five mice, i.e., total of 20 mice per time point. The prostates (with tumors) were weighed and regional lymph nodes were harvested for histological analysis. This experiment was repeated three times with similar results. The combined data are therefore summarized in Figure 4A and B. All mice implanted with PC-3 MM2 cells developed prostate tumors exceeding 1.5 g by week 3 after implantation. The mice were therefore killed. Prostate tumors produced by PC-3P cells reached 0.5 g in weight by week 3, and by week 5, the tumors exceeded 1.5 g in weight (Figure 4A). PC-3 IL-8 low cells produced slow-growing tumors, whereas PC-3 IL-8 high cells produced rapidly growing prostate tumors. By week 4, the mean tumor weights of PC-3 IL-8 high and PC-3 and IL-8 low cells were 0.68±0.2 and 0.21 ±0.07 g, respectively (P<.01). By week 5, the mean tumor weights were 1.52±0.4 and 0.59±0.2 g, respectively (P<.01). PC-3 IL-8 high cells produced early lymph node metastasis in all injected mice, whereas PC-3 IL-8 low cells produced later metastasis (Figure 4S). The incidence (percent of positive mice) of regional lymph node metastasis is also shown in Figure 2. All (100%) mice injected with PC-3P or PC-3 MM2 cells had regional lymph node metastasis (Figure 4A). PC-3 IL-8 high cells produced lymph node metastasis in all injected mice (100%) by week 5, whereas PC-3 IL-8 low cells produced lymph node metastasis in 64% of the injected mice (Figure 4B).

Figure 4.

Tumorigenic and metastatic potential. (A) PC-3 parental and PC-3 MM2 and (B) PC-3 IL-8 low and PC-3 IL-8 high cells were injected into the prostate of nude mice. At 3, 4, or 5 weeks after tumor cell injection, groups of mice (n=5) were killed and necropsied. All mice with large palpable prostate (>1.5 g) were killed and necropsied, regardless of the time of injection. Mean tumor weight ±SD is shown in the bar graph. Percent regional lymph node metastasis is shown over the bars.

Expression of Angiogenesis and Metastasis-Regulating Genes

The integrity of mRNA in each sample was first verified using a poly d(T)₂₀ probe [18,36–39]. All samples had an intense reaction, indicating that the mRNA was intact. Normalization of mRNA expression intensities for poly d(T)₂₀ probe intensity and also for uninvolved prostate gland allowed for comparison of expression intensities of multiple samples analogous for loading controls used in other assays. Because the expression levels of bFGF, MMP-2, MMP-9, and E-cadherin vary between the periphery and center of neoplasms [18,36–39], we concentrated on gene expression level at the periphery (actively growing) of the tumors.

The images shown in Figure 5 demonstrate that tumors produced by PC-3 IL-8 high cells expressed a significantly higher level of IL-8 mRNA than tumors produced by PC-3 IL-8 low cells (expression intensities of 99 vs 66, P<.01). The more aggressive PC-3 IL-8 high tumors also expressed higher levels of VEGF, higher levels of MMP-2, and lower levels of E-cadherin than the PC-3 IL-8 low tumors. No discernible differences were found among the tumors for expression of bFGF.

Figure 5.

ISH analysis for angiogenesis- and metastasis-regulating genes in prostate tumors. Hybridization with poly d(T)₂₀ probe confirmed mRNA integrity. The numbers indicate expression intensities as compared with poly d(T)₂₀, which was assigned the value of 100. All analyses were carried out at the periphery of the neoplasms.

Induction of Vascularization

In the final set of experiments, we determined the relative microvessel density (MVD) at the periphery of the prostate tumors produced by the PC-3 IL-8 low or IL-8 high cells (Figure 6). Consistent with expression of IL-8, VEGF, and MMP-9, the tumors formed by PC-3 IL-8 cells had an MVD of 587±66, whereas PC-3 IL-8 low tumors had an MVD of 247±57 (P<.001).

Figure 6.

Relative MVD in prostate tumors. Prostate tumors produced by PC-3 IL-8 high and PC-3 IL-8 low cells were harvested and prepared for immunostaining with anti-CD31 antibodies. The mean number of blood vessels was 587±66 in tumors produced by PC-3 IL-8 high and 247±57 in tumors produced by PC-3 IL-8 low cells (P <.001).

Discussion

The present results demonstrate that the expression level of IL-8 in human prostate cancer cells is associated with tumorigenicity, neovascularization, and production of lymph node metastasis. The progressive growth and production of metastasis are dependent on the induction of blood supply, i.e., angiogenesis, which is mediated by a change in the balance between proangiogenic (e.g., IL-8) and antiangiogenic (e.g., IFN) molecules [11–13,40–42]. IL-8 is a multifunctional CXC chemokine with an ELR motif (Glu-Leu-Arg) preceding the first conserved cysteine in the NH₂ terminus. IL-8 has been reported to stimulate proliferation of keratinocytes [27], melanoma cells [27,30], and human ovarian cancer cells [40] to induce hepatotactic migration [43] and angiogenesis [44,45], probably by enhancing expression of MMP-2 and MMP-9 [29,46], which are necessary for degradation of the extracellular matrix by angiogenic endothelial cells. The expression level of IL-8 is elevated in highly vascularized human non-small cell lung cancer [47,48] and human gastric cancer [23,28]. High IL-8 expression level has also been reported for human ovarian cancer [40], bladder cancer [49], and prostate cancer [18,36]. Moreover, the expression level of IL-8 directly correlates with the metastatic potential of human melanoma cells implanted into the subcutis of nude mice [30,50].

By the time of diagnosis, neoplasms are heterogeneous for many biological properties [3,7]. Our present study demonstrates that the parental PC-3 prostate cancer cell line is heterogeneous for expression of IL-8. The relatively low and high expression of IL-8 (in different clones) was due to differences in transcriptional activity rather than mRNA stability. Recent data from our laboratory concluded that the differential expression of IL-8 by different clonal populations of human melanoma cells is due to different expression levels of constitutive NF-κB activity [51]. Preliminary data suggest that this is also the case for the PC-3 cells.

Several low IL-8-expressing clones were combined into a single line, as were several high IL-8-expressing clones. The PC-3 IL-8 low cells were less tumorigenic in the prostate of nude mice than the PC-3 IL-8 high cells. Moreover, the tumors formed by the PC-3 IL-8 low cells grew slower and produced a lower incidence of metastasis than the PC-3 IL-8 high cells. To determine the mechanism by which IL-8 promotes tumorigenicity and metastasis, we examined tumor cell proliferation in vitro and in vivo. Since IL-8 has been shown to stimulate proliferation of both normal cells and tumor cells, we first determined whether it is an autocrine or paracrine growth factor for the PC-3 cells. Regardless of IL-8 expression level, all PC-3 cells tested expressed both receptors for IL-8. Our in vitro data, however, did not demonstrate that IL-8 enhances proliferation of the PC-3 cells.

The metastatic potential of human neoplasms has been shown to correlate directly with the expression levels of several independent genes that regulate the following: angiogenesis (VEGF/VPF, bFGF, IL-8) [39,52], and invasion (MMP-2/MMP-9 genes) [53]. There have also been several reports that the expression of E-cadherin, which is involved in cell-to-cell cohesion, inversely correlates with tumor progression and metastasis [36,54]. Most of these correlative studies reached the inevitable conclusion that the expression of a given gene is necessary, but insufficient, for a tumor cell to complete the multistep process of metastasis. Because each of the discrete links in the pathogenesis of a metastasis is regulated by one or several independent genes, the identification of cells with metastatic potential in heterogeneous neoplasms requires multiparametric/multivariate analysis of gene expression.

We have developed a rapid colorimetric mRNA ISH technique for detecting the activity of the different steps of angiogenesis and metastasis. The growth and metastasis of the PC-3 tumors in the prostate of nude mice correlated with expression of several genes that regulate different steps of angiogenesis and metastasis. The rapidly growing tumors produced by PC-3 IL-8 high cells expressed high levels of IL-8, MMP-2, MMP-9, and VEGF mRNA, but lower levels of E-cadherin. Once cells detach from the primary tumor, they must invade the host stroma if they are to metastasize [55]. Degradation of blood vessel basement components, especially type IV collagen, is one of the necessary steps in metastasis. The levels of Mr 72,000 and Mr 92,000 type IV collagenase in human and rodent neoplasms directly correlate with invasion and metastasis, and specific inhibitors of MMPs have been shown to inhibit tumor cell invasion [53,55]. Thus, a decrease in the expression of E-cadherin and increase in collagenase type IV activity enhance tumor cell invasion and metastasis.

Consistent with the expression level of IL-8, MMP-2/-9, and VEGF, the faster-growing PC-3 tumors were highly vascularized as detected by immunostaining with anti-CD31 antibodies. Increased MVD has been reported to directly correlate with clinical outcome of many human neoplasms, including prostate cancer [38,42].

Conclusion

This study demonstrates that expression of IL-8 by human prostate cancer cells is associated with increased expression of MMP-2/-9 and VEGF and a decrease in the expression of E-cadherin. Since the expression of IL-8 by prostate cancer cells correlates with enhanced angiogenesis, tumorigenicity, and metastasis, IL-8 is an attractive target for therapeutic development.

Abbreviations

bFGF

basic fibroblast growth factor

HBSS

Hanks' balanced salt solution

IFN

interferon

IHC

immunohistochemistry

IL-8

interleukin-8

MMP

matrix metalloproteinase

MTT

tetrazolium salt

MVD

microvessel density

PBS

phosphate-buffered saline

PNCA

proliferating cell nuclear antigen

VEGF/VPF

vascular endothelial growth factor/vascular permeability factor

VEGF-R

vascular endothelial growth factor receptor