The eicosanoids PGE₂, 5-HETE, 12-HETE, and 15-HETE, and the corresponding deuterated eicosanoid standards were purchased from Cayman Chemical. AA, butylated hydroxytoluene (BHT), citric acid, and EDTA were obtained from Sigma Chemical. All high-pressure liquid chromatography (HPLC) solvents used for analyses of eicosanoids were purchased from Fisher Scientific. Anti-COX-2 and anti-15-LOX antibodies were obtained from Cayman Chemical. Anti-5-LOX antibody was purchased from Research Diagnostics and anti-β-actin antibody was purchased from Sigma Chemical. Anti-platelet-type 12-LOX antibody was obtained from Oxford Biomedical Research.

Human prostate cancer cell lines LNCaP, PC-3, and DU145 were obtained from the American Type Culture Collection and maintained in a humidified atmosphere containing 5% CO₂ at 37°C. LNCaP and PC3 cells were routinely cultured in RPMI-1640 medium (Invitrogen), supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Hyclone Laboratories) containing 50 IU/ml penicillin, 50 μg/ml streptomycin, and 2 mM L-glutamine [all from Gibco (Invitrogen)]. DU145 cells were cultured under similar conditions except that cells were grown in DMEM medium (Mediatech).

The human prostate cancer cell lines MDA PCa 2a and MDA PCa 2b were gifts from Dr Nora Navone (28). These cell lines were grown in BRFF-HPC1 medium (AthenaES) supplemented with 20% FBS in tissue culture flasks coated with FNC coating mix (AthenaES). All cell lines were authenticated via microscopic morphology check and DNA analysis.

All animal experiments were approved by the Institutional Animal Care and Use Committee at MD Anderson. Xenograft models of PC-3, LNCaP, DU145, and MDA PCa 2b cells were developed as previously described (28). In brief, PC-3 and DU145 (1×10⁶), LNCaP (5×10⁶), and MDA PCa 2b (1×10⁷) cells were suspended in 50% Matrigel (BD Biosciences) and subcutaneously implanted into the flanks of 8-week-old male BALB/c athymic (Nu/Nu) mice. Mice were euthanized with CO₂ approximately 1 month later, when the tumor volume reached 1 cm³. A portion of each tumor was snap frozen and stored at −80°C until measurements of eicosanoid profiles and protein expression of eicosanoid enzymes were obtained.

Ex vivo biopsy specimens were obtained from eight patients undergoing prostatectomy who consented to the use of their tissues in research, according to the Institutional Tissue Banking protocol at MD Anderson. Whenever possible, multiple core biopsy specimens were obtained from each prostatectomy specimen. The presence of tumor in each core was determined by making a touch prep of the cores before the specimens were snap frozen in liquid nitrogen, placed in cryovials, and archived at −80°C in the Prostate Tissue Bank.

Technicians ensured conformance with the prostatectomy specimen map delineating individual tumor foci. Seven core biopsy specimens containing tumor and three without tumor were used to test for the expression of COXs and LOXs; seven core biopsy specimens without tumor and 19 with tumor were used for evaluating eicosanoid profiles. In this proof-of-principle analysis, the limited amount of tissue prevented the testing of matching specimens across all two-sample types and evaluating intrapatient as well as interpatient variabilities.

Total RNA was extracted from the five prostate cancer cell lines using TRI reagent (Molecular Research Center). The RNA from each sample was reverse transcribed and then measured quantitatively using real-time polymerase chain reaction (PCR) and a comparative threshold cycle method, as previously described (29,30).

Cells growing in log phase were washed with cold PBS and scraped free in the presence of a lysis buffer (Invitrogen) with protease inhibitor cocktail (Sigma). Cell lysates were then sonicated on ice for 3 min, incubated for 10 min on ice, and centrifuged at 14,000 × g for 10 min at 4°C.

Protein levels were quantified via the detergent-compatible protein assay (Bio-Rad). Equal levels of protein (60 μg) were fractionated on precast gels (Bio-Rad) and then transferred onto polyvinylidene diflouride membranes, according to standard methods. After a 1- to 2-h incubation in 5% nonfat dry milk blocking buffer prepared in Tris-buffered saline with 0.1% Tween 20, the membranes were probed with primary antibodies diluted 1:2,000 in blocking buffer. Protein bands were visualized via chemiluminescence, using the ECL+ detection kit (Amersham Biosciences) and hyper-film (Amersham Biosciences). Equal loading of samples was illustrated by Western blotting for the presence of β-actin.

The levels of eicosanoids in the prostate cancer cells, xenograft tissues, and human biopsy samples were determined according to the methods of Kempen et al and Yang et al (31–33). In brief, the human prostate cancer cells (5×10⁶) were harvested by trypsinization, washed with PBS, and then resuspended in 0.5 ml of PBS containing 1 mM CaCl₂. For exogenous eicosanoid analysis, samples were incubated with 2.5 μl of calcium ionophore A23187 (1 mM; Sigma) for 2 min at 37°C, followed by addition of an aliquot of 2.5 μl of 10 mM AA. Samples were then incubated for a further 10 min under similar conditions. The reaction was terminated by the addition of aliquots of 40 μl of 1 N citric acid and 5 μl of 10% BHT. An aliquot of 10 μl of the relevant deuterated eicosanoids (PGE₂-d₄, 15-HETE-d₈, 12-HETE-d₈, and 5-HETE-d₈) per 100 ng/ml) was added to the reaction mixtures as internal standards. The eicosanoids were extracted three times with 2 ml of hexane-ethyl acetate (1:1, v/v). The upper organic phases were then pooled and evaporated to dryness under a stream of nitrogen at room temperature. All extraction procedures were performed under conditions of minimal light. Samples were then reconstituted in 100 μl of methanol with 10 mM ammonium acetate buffer (70:30, v/v; pH, 8.5) before analysis by liquid chromatography/tandem mass spectrometry (LC/MS/MS). For the measurement of endogenous eicosanoids, the resuspended cells were acidified by the addition of an aliquot of 40 μl of 1 N citric acid followed by the addition of the deuterated internal standards and extraction of eicosanoids as previously described (33).

For analysis of eicosanoids in the xenograft tissues, each frozen tissue specimen (25–50 mg) was ground to a fine powder in a liquid nitrogen-cooled mortar (Fisher). Samples were then transferred to microcentrifuge tubes. Ice-cold PBS buffer, triple each sample’s volume and containing 0.1% BHT and 1 mM EDTA, was added, and the tubes were sealed. The mixtures were then homogenized using an ultrasonic processor (Misonix) at 0°C for 3 min. A 100-μl aliquot of each homogenate was transferred to a glass tube (13×100 mm), and the eicosanoids were extracted using the method of Yang et al (33).

We next measured the AA metabolites in the human prostate core biopsy samples using a modification of the method of Yang et al (33). In brief, the frozen specimens were mixed with 150 μl of homogenization buffer [500 mM Tris-HCl (pH 7.2), 0.5 M sucrose, 200 M EDTA, 100 mM EGTA, 0.4 M NaF, 10% Triton X-100, and 10 mM sodium orthovanadate (Sigma)] and incubated at 0°C for 30 min. The samples were then homogenized and processed for eicosanoid extraction as described above.

Reverse-phase HPLC electrospray ionization MS was used to measure eicosanoid (PGE₂, 5-HETE, 12-HETE) levels in cells using our previously published method (31,34). A Micromass Quattro Ultima tandem mass spectrometer (Waters) was equipped with an Agilent 1100-HP binary pump HPLC inlet for use in these studies. Eicosanoids were separated by a Luna 3-μ phenyl-hexyl (2×150 mm) LC column (Phenomenex). The mobile phase consisted of 10 mM ammonium acetate (pH 8.5) and methanol; the flow rate was 250 μl/min, the column temperature was 50°C, and the sample injection volume was 25 μl. Samples were kept at 4°C in an autosampler before their injection into the analytical column. The mass spectrometer was operated in the electrospray negative-ion mode with a cone voltage of 100 V. Fragmentation of all compounds was performed using argon as the inert collision gas at a cell pressure of 2.1×10⁻³ torr. The collision energy was 19 V. All eicosanoids were detected using negative ionization and multiple-reaction monitoring of the transition ions for eicosanoid products and their internal standards (33).

Descriptive statistics were calculated and exploratory data analyses were performed. Categorical data were summarized by frequency counts, proportions, or percentages. Continuously scaled measures were summarized as means and standard deviations or as medians with ranges. Gene expression image plots, bar graphs, and scatter plots were used to graphically present the data. Student’s t-test was used for analyzing continuous variables. A mixed-effects model was used in the analysis of repeated measurements from the same patients. All statistical tests were two sided, and p-values of <0.05 were considered to indicate statistically significant differences. Statistical analyses were performed with S-PLUS software (TIBCO Software).

The data from the quantitative analysis of mRNA of eicosanoid enzymes in the five different human prostate cancer cell lines are shown in Fig. 1. COX-2 mRNA was detectable only in PC-3 and DU145 cells, with PC-3 cells containing 10 times more than DU145 cells (Fig. 1A). By contrast, levels of 12-LOX mRNA were observed only in the MDA PCa 2b metastatic prostate cancer cells (Fig. 1B). Notably, the level of 5-LOX mRNA in LNCaP cells was almost five times higher than that in PC-3 cells, whereas very limited expression of 5-LOX was detected in MDA PCa 2a and MDA PCa 2b cells (Fig. 1C). The mRNA level of 15-LOX-2 was also highest in the LNCaP cells, relative to that in the other cells tested (Fig. 1D).

The protein expression of COX-2 and LOX enzymes was examined in the five different prostate cancer cell lines by Western blotting to confirm that the RNA expression of these enzymes had resulted in protein expression. Although the protein expression of COX-2 in PC-3 cells was similar to that of mRNA in this cell line, the protein expression levels of 5-LOX, 12-LOX, and 15-LOX-2 were somewhat different from that of mRNA (Fig. 2). This could be reflective of either assay differences or, more importantly, differences in mRNA translation efficiency between the various cancer cell lines.

The mRNA and protein levels of each eicosanoid enzyme were differently regulated in the five prostate cancer cell lines we tested, which suggested that the relevant metabolites of the eicosanoids may also differ. Therefore, we measured both endogenous and exogenous eicosanoid levels in those cell lines. As shown in Fig. 3A, levels of both endogenous and exogenous PGE₂, a COX-2 metabolite, were highest in the PC-3 cells, suggesting a high capacity to produce this pro-inflammatory eicosanoid. The level of exogenous PGE₂ in PC-3 cells was almost 15 times higher than that produced endogenously.

By comparison, the levels of endogenous 12-HETE in the metastatic prostate cancer cells MDA PCa 2a (0.48±0.10 ng/million cells) and MDA PCa 2b (0.41±0.19 ng/million cells) were almost twice those in the PC-3, DU145, and LNCaP cells (0.16–0.25 ng/million cells) (Fig. 3B). When AA was used as a supplement for the cells, the formation of 12-HETE was markedly increased in the MDA PCa 2b and PC-3 cells. The level of exogenous 12-HETE was ranked from highest to lowest in the MDA PCa 2b, PC-3, MDA PCa 2a, DU145, and LNCaP cells.

The level of endogenous 5-HETE was highest in the LNCaP cells, but the exogenous level of 5-HETE was noticeably higher in the PC-3 cells, in fact, more than 1.5 times the levels in the other cell lines (Fig. 3C). This suggests that the capacity to produce 5-LOX is greater in PC-3 cells than in the other cancer cell lines tested, and it correlated with the highest expression of the 5-LOX protein in this particular cell line (Fig. 2). The level of endogenous 15-HETE was also highest in LNCaP cells, but after exposure of the cell lines to AA, the MDA PCa 2a cells showed the highest level of 15-HETE formation (Fig. 3D).

To explore the role of eicosanoids in prostate cancer progression, we further investigated the expression of COX-2 and various LOX proteins in human prostate xenograft tissues (Fig. 4, top panel). As the tumorigenicity of the MDA PCa 2a cells was extremely low, we evaluated the protein expression of eicosanoid enzymes in only the other four human prostate cancer xenograft tissues. The protein expression of the COX-2 and 5-LOX enzymes was highest in the PC-3 xenografts, whereas the protein levels of 12-LOX were higher in both PC-3 and PCa 2b cells. In line with the protein expression of 15-LOX-2 in the cells in vitro, its expression was consistently highest in the two tumor tissues derived from MDA PCa 2b xenografts (Fig. 4, top panel).

AA metabolites were further examined in xenografts of the DU145, LNCaP, PC-3, and MDA PCa 2b human prostate cancer cells (Fig. 4, bottom panel). In a pattern similar to that observed in the cultured prostate cancer cell lines, the concentration of endogenous PGE₂ was significantly higher in PC-3 xenograft tissues (10.5±2.17 ng/mg protein) than it was in the other three xenograft tissues (range, 0.07–1.1 ng/mg of protein) (p<0.001) (Fig. 4A, bottom panel). The levels of 12-HETE in both PC-3 and MDA PCa 2b xenograft tissues were also much higher (13.9 and 10.9 ng/mg protein, respectively) than those in the LNCaP xenograft tissues (3.9 ng/mg protein) (p<0.01) (Fig. 4B, bottom panel). The highest levels of both 5- and 15-HETE in PC3 and MDA PCa 2b tumor tissues were only about one third or one fourth (2.07±0.46 for 5-HETE and 3.24±0.45 ng/mg protein) the levels of PGE₂ and 12-HETE (Fig. 4C and D, bottom panel). These results appear to correspond with the endogenous levels of these AA metabolites in the same prostate cancer cells in vitro.

To test the role of AA metabolism in prostate cancer, the expression of COX and LOX proteins in human prostate core biopsy specimens containing tumor was compared with that in normal, nontumorous prostate tissue specimens. As shown in Fig. 5, the expression of COX-1 and COX-2 enzymes did not differ between the tumorous and normal tissues. In almost all of the tumorous specimens, the expression of 12-LOX was moderately greater than it was in the specimens of normal tissue. The protein levels of 5-LOX were also greater in the cores containing tumor than in those without, but the magnitude of the increases was not as remarkable as that in the case of 12-LOX. We found that the expression of 15-LOX-2 was slightly higher in most of the tumorous cores than it was in the normal cores.

AA metabolism was also examined in seven specimens without tumor and 19 specimens containing tumor obtained from core biopsies of prostatectomy specimens from eight patients. A mixed-effects linear model was fitted to assess the differences between the tumorous and normal specimens. The mean PGE₂ levels in the non-tumor-containing and tumorous specimens were very similar, as shown in Fig. 6A (0.335 ng/mg and 0.314 ng/mg protein, respectively; p=0.703). The mean level of endogenous 12-HETE was significantly higher in the cores containing tumor than it was in those without (0.094 ng/mg and 0.010 ng/mg protein, respectively; p=0.019) (Fig. 6B).

As with the PGE₂ levels, the mean 5-HETE levels were very similar in tumorous and normal tissues (0.062 ng/mg and 0.073 ng/mg protein, respectively; p= 0.483) (Fig. 6C). However, the mean level of the 15-LOX-2 metabolite 15-HETE in the tumorous cores was more than twice that in the normal cores, although the difference was not statistically significant (0.269 ng/mg and 0.129 ng/mg protein, respectively; p=0.110) (Fig. 6D).

The changes in AA metabolism in the core biopsy specimens containing malignant cells were similar to those observed both in prostate cancer cells in vitro and in prostate cancer xenograft tissues, suggesting that 12-LOX may play an important role in the progression of human prostate cancer.