anti-androgenic activity of fatty acids

Anti-Androgenic Activity of Fatty Acids

by Jie Liu, Kuniyoshi Shimizu, and Ryuichiro Kondo*

Department of Forest and Forest Products Science, Faculty of Agriculture, Kyushu University, Fukuoka812-8581, Japan (phone: þ81-92-642-2811; fax: þ81-92-642-2811;

e-mail: [email protected])

In this study, we show that 5a-reductase derived from rat fresh liver was inhibited by certain aliphaticfree fatty acids. The influences of chain length, unsaturation, oxidation, and esterification on the potencyto inhibit 5a-reductase activity were studied. Among the fatty acids we tested, inhibitory saturated fattyacids had C12 –C16 chains, and the presence of a C¼C bond enhanced the inhibitory activity. Esterificationand hydroxy compounds were totally inactive. Finally, we tested the prostate cancer cell proliferationeffect of free fatty acids. In keeping with the results of the 5a-reductase assay, saturated fatty acids with aC12 chain (lauric acid) and unsaturated fatty acids (oleic acid and a-linolenic acid) showed a proliferationinhibitory effect on lymph-node carcinoma of the prostate (LNCaP) cells. At the same time, thetestosterone-induced prostate-specific antigen (PSA) mRNA expression was down-regulated. Theseresults suggested that fatty acids with 5a-reductase inhibitory activity block the conversion oftestosterone to 5a-dihydrotestosterone (DHT) and then inhibit the proliferation of prostate cancer cells.

Introduction. – Benign prostatic hyperplasia (BPH) is a ubiquitous condition inaging males, such that the incidence of BPH detected at autopsy increases from ca. 30%at age 50 to >80% at age 80 [1]. Unfortunately, the urinary symptoms attributed toBPH lead to significant erosion in the quality of life for affected men, and manyundergo surgery as a result [2]. Prostate cancer is one of the most frequently diagnosedmalignancies and is the second leading cause of cancer death in American men.

In several androgen targets, like the prostate, testosterone is converted to 5a-dihydrotestosterone (DHT), which is the most potent natural androgen. Upon bindingof the native ligand, DHT, the androgen receptor activates target genes such as theprostate-specific antigen (PSA) gene [3] [4]. PSA is the most commonly used serummarker for the diagnosis and progression of prostate cancer. The microsomal enzymesteroid 5a-reductase [EC 1.3.99.5] catalyzes the NADPH-dependent reduction of theD4,5-double bonds of a variety of 3-oxo-D4-steroids [5]. Although the etiology of BPH isunclear, the permissive role of DHT in the hyperplastic growth of the prostate is well-established. Patients with BPH have been reported to have higher than normalconcentrations of DHT in the prostate [6] [7]. Androgen-ablation therapy of prostatecancer reduces the levels of circulating androgens, and then inhibits tumor proliferationand induces apoptosis of tumor cells [8]. Since DHT promotes the development of acne[9], male pattern alopecia [10], BPH, and female hirsutism [11], inhibitors of 5a-reductase may be useful for the treatment of these conditions. In efforts to develop 5a-reductase inhibitors for the treatment of human diseases such as BPH and prostatecancer, pre-clinical evaluation of 5a-reductase inhibitors depends on in vitro studies.

CHEMISTRY & BIODIVERSITY – Vol. 6 (2009) 503

� 2009 Verlag Helvetica Chimica Acta AG, Z�rich

In this study, we show that 5a-reductase derived from rat fresh liver was inhibited bycertain aliphatic free fatty acids. To determine the structure – enzyme inhibitionrelationships for free fatty acids, we studied the influence of chain length, unsaturation,oxidation, and esterification on the potency to inhibit 5a-reductase activity. Among thefatty acids we have tested, inhibitory saturated fatty acids have C12 – C16 chains, whichshowed more than 50% inhibition at 1.3 mm. The presence of a C¼C bond in inhibitoryunsaturated fatty acids is helpful for inhibitory activity. Esterification and hydroxycompounds were totally inactive. Finally, we tested the prostate cancer cellproliferation effect of free fatty acids. Consistent with the results of the 5a-reductaseassay, both lauric acid (saturated fatty acid), and a-linolenic acid and oleic acid(unsaturated fatty acids) showed a proliferation inhibitory effect on LNCaP cells. Atthe same time, the expression of PSA mRNA was down-regulated. These resultssuggested that fatty acids with 5a-reductase inhibition, such as a-linolenic acid, lauricacid, and the like, block the conversion of testosterone to DHT and then inhibit theproliferation of prostate cancer cells.

Results and Discussion. – In our previous study [12], fatty acids such as oleic acid,linoleic acid, and palmitic acid, which were isolated from the EtOH extracts ofGanoderma lucidum, showed high inhibitory activity for 5a-reductase, which is anuclear membrane-associated enzyme. Many fatty acids are essential components ofmammalian membranes. Interestingly, as previously described, 5a-reductase seems tobe very sensitive to the composition of the molecular environment, since accessibility ofthe enzyme to its cofactor depends on the conformational state of the protein, the latterdepending on the composition of the membrane. Specific aliphatic unsaturated fattyacids have been shown to inhibit human or rat microsomal 5a-reductase activity. Thefact that only unsaturated fatty acids with specific configurations were potent inhibitorsof 5a-reductase, and that two other microsomal enzymes examined were not affectedsuggested that the inhibition was selective [13]. These results let us study the influenceof chain length, unsaturation, oxidation, and esterification of fatty acids moiety on thepotency of inhibiting 5a-reductase activity.

5a-Reductase Inhibitory Activity. Fig. 1 shows that only certain saturated fatty acidshave inhibitory activity. Among the saturated fatty acids, the inhibitory acids (morethan 50% inhibition at 1.3 mm) have C12 – C16 chains. The relative inhibitory potenciesfor the 5a-reductase of saturated fatty acids are, in decreasing order, C12>C14>C15>C16>C18. This result suggested that the 5a-reductase inhibitory activity of saturatedfatty acids depends on the hydrocarbon-chain length. This result was similar to thosereported by Niederpr�m et al. [14]. When C18 fatty acids were investigated, only theunsaturated compounds showed 5a-reductase inhibitory activity of greater than 50%inhibition at the concentration of 1.3 mm (Figs. 1 and 2, and Table). The presence of aC¼C bond enhanced the inhibitory activity (Figs. 1 and 2, and Table). The number andposition of the C¼C bonds also affected the potency. For example, the C18 fatty acidscan be listed in decreasing order of their inhibitory potencies as follows: a-linolenicacid, linoleic acid, oleic acid, and stearic acid. Compared with the active fatty acids,methyl ester and alcohol analogues showed a lower level of 5a-reductase inhibitoryactivity. The free COOH group is important, since the methyl ester and alcoholanalogues of these inhibitory unsaturated fatty acids were either inactive or only

CHEMISTRY & BIODIVERSITY – Vol. 6 (2009)504

slightly active. This result, together with the inhibitory activity of saturated fatty acids,suggests that certain hydrophobicity is a prerequisite for 5a-reductase inhibitoryactivity, which also requires strongly polar end-groups. 5a-Reductase inhibitory activityby fatty acids has already been reported by Liang and Liao [13]. In the present study,


Fig. 1. Inhibition of 5a-reductase by fatty acids. Each column represents the mean�S.D., n¼4. Sampleconcentration is 1.3 mm.

Fig. 2. Inhibition of 5a-reductase by methyl ester and alcohol analogues of fatty acids. Each columnrepresents the mean�S.D., n¼4. Sample concentration is 1.3 mm.

we showed that lauric acid, which was inactive in the assay system of Liang and Liao,was active in reducing the activity of 5a-reductase. However, a precise explanation ofthe contradiction remains elusive.

Inhibitory Effect on Prostate Cancer Cells. The LNCaP (lymph node carcinoma ofthe prostate) human prostate cancer cell line is a well-established and androgen-dependent cell line [15]. LNCaP Cells retain most of the characteristics of humanprostate carcinoma, like the dependence on androgens, the presence of the androgenreceptor, and the production of PSA. For these reasons, the LNCaP cell line is anattractive model for in vitro studies of the biology of human prostate cancer [16].

Phospholipids represent two-thirds and cholesterol nearly one-third of total lipidsin BPH homogenates [17]. The main fatty acids were stearic (C18:0) and linoleic (C18:2)acid in epithelium. In the epithelium, the amount of oleic acid was significantly higherthan in the stroma [18]. These results led us to investigate the effect of a-linolenic acid,oleic acid, stearic acid, and lauric acid, as well as their methyl esters, on the proliferationinhibition effect of LNCaP. Androgen-enhanced LNCaP cell proliferation in a mediumwith 100 nm DHT is two times greater than that of vehicle-treated cells in androgen-free medium, which confirms the cell-growth response to androgen stimulation.Without the testosterone and DHT, a-linolenic acid, oleic acid, and lauric acid cancause the inhibition of cell growth for LNCaP cells. For a-linolenic acid, oleic acid, andlauric acid, the inhibition of cell growth occurred initially at 100 mm and increased as thedose was increased. In contrast, under the same conditions, stearic acid and the methylester of these compounds showed no measurable increase in cell death. These resultssuggest that the 5a-reductase inhibitory activity of a-linolenic acid, oleic acid, andlauric acid may be involved in cell proliferation, which is mediated by androgen, 5a-reductase, and the androgen receptor signaling pathway.

As shown in Fig. 3, the proliferation inhibitory effects of lauric acid, oleic acid, anda-linolenic acid with testosterone may come from 5a-reductase inhibitory activities.But the proliferation inhibitory effects of lauric acid, oleic acid, and a-linolenic acidwith DHT cannot arise from 5a-reductase inhibitory activities of fatty acids. The directcytotoxic effects must, of course, be considered. It was reported that long-chainpolyunsaturated fatty acids have a significant inhibitory effect on cell growth andproliferation [19]. The induction of programmed cell death in cancer cells is a criticalmechanism for many treatment modalities [20]. When both DHT and each fatty acidare added, cell death may be caused by the inhibition of cell growth without androgen


Table. IC50 Values of 5a-Reductase Inhibitory Activity of Some Fatty Acids

Fatty acid IC50�SD [mm]

Lauric acid (C12H24O2) 470�3Myristic acid (C14H28O2) 500�2Pentadecanoic acid (C15H30O2) 630�1Palmitic acid (C16H32O2) 870�4Stearic acid (C18H36O2) >1300Oleic acid (C18H34O2(18 : 1)) 140�5Linoleic acid (C18H32O2(18 : 2)) 130�3a-Linolenic acid (C18H30O2(18 : 3)) 120�2

signaling. Thus, we designed the following experiment to clarify whether a directcytotoxic effect of fatty acids could be observed.

Expression of the Serum Marker of Prostate Cancer – PSA. We examined theexpression of PSA mRNA, because PSA is an androgen-regulated gene that isfrequently up-regulated in proliferating prostate cells. The expression of PSA is mainlyregulated by the androgen receptor (AR). Inhibition of AR signaling blocks theexpression of PSA. If the inhibition of cell growth was not caused by androgensignaling, the PSA mRNA would not be affected by the addition of each fatty acid.Otherwise, a change in the PSA mRNA expression would be observed when the fattyacids were added. As shown in Fig. 4, the lauric acid, oleic acid, and a-linolenic acid caninhibit cell proliferation. The inhibitory effects of these fatty acids on cell proliferationare similar to each other, so we choose lauric acid (as a saturated fatty acid) and a-linolenic acid (as an unsaturated fatty acid) for the PSA mRNA experiment. Theeffects of lauric acid and a-linolenic acid on gene expression in LNCaP cells were testedwith real-time polymerase chain reaction (PCR). It was reported that, when well-characterized, hormonally responsive LNCaP human prostate cancer cells were used tocompare the effects of testosterone and DHT on cell proliferation, and gene andprotein expression of PSA, testosterone and DHT increase cell proliferation, andmRNA and/or protein expression of PSA in the same way [21]. The same results wereobserved in our experiments. Control cells cultured in androgen-free mediumconstitutively expressed a significantly low level of PSA. With androgen testosteroneor DHT, the PSA level in these cells increased ca. 2.5-fold. The testosterone-regulatedPSA mRNA expression was markedly suppressed by co-culturing with fatty acids for72 h. These findings are consistent with our hypothesis that the cell growth inhibited byfatty acids is involved in inhibiting the conversion of testosterone to DHT. In contrast,the DHT-regulated PSA mRNA expression was not significantly affected. These resultsalso suggested that lauric acid and a-linolenic acid cannot affect the steroid-receptorbinding of steroid hormones in cell proliferation. The inhibition of cell growth of thefatty acids co-cultured with testosterone at least partially comes from the inhibition offatty acids on 5a-reductase, which affects the AR signaling pathway. The inhibition ofcell growth of the fatty acids co-cultured with DHT cannot be related to androgensignaling; more specifically, the inhibition may be caused by the direct cytotoxic effectsof each fatty acid.

When activated, the transcription nuclear factor kB (NFkB) blocks programmedcell death or apoptosis [22]. NFkB is often up-regulated in cancer cells, resulting in cellsthat are resistant to chemotherapy drugs or radiation, and that do not die in response tothe genetic damage that has occurred. The Bcl-2 family of genes and COX 2 expressioncan also block apoptosis, resulting in cells that do not die at the appropriate time. Theunsaturated fatty acids can restore functional apoptosis by down-regulating NFkB [22],which, in turn, down-regulates COX 2 expression [23] [24], and by down-regulating theexpression of Bcl-2 family genes [25] [26]. Further studies are needed to elucidate theeffects of fatty acids on the transcriptional activity of NFkB in LNCaP cells.

Since DHT promotes the development of acne, male pattern alopecia, BPH,prostate cancer, and female hirsutism, inhibitors of 5a-reductase may be useful for thetreatment of these conditions. Clinical observations of patients with 5a-reductasedeficiency [27] and studies of the effects of 5a-reductase inhibitors on experimental


animals [20] [28] have indicated that spermatogenesis, the maintenance of libido, sexualbehavior, and the feedback inhibition of gonadotropin secretion do not require theconversion of testosterone to DHT. Thus, treatments of androgen-dependent skin andprostatic diseases by 5a-reductase inhibitors would be expected to produce fewer side-effects than the presently available hormonal therapies that involve castration and/orthe administration of oestrogens, androgen receptor antagonists, or gonadotropin-


Fig. 3. The inhibition effect of a-linolenic acid, oleic acid, lauric acid, a-linolenic acid methyl ester, oleicacid methyl ester, lauric acid methyl ester, and stearic acid methyl ester on the LNCaP cell growth. Eachcolumn represents the mean�S.D., n¼3. T: testosterone; DHT: dihydrotestosterone. **: p<0.01 against

corresponding control, T, or DHT.

releasing hormone agonists. In efforts to develop 5a-reductase inhibitors for thetreatment of human diseases such as BPH and prostate cancer, the pre-clinicalevaluation of 5a-reductase inhibitors depends on in vitro studies.

Conclusions. – Our results show that 5a-reductase derived from rat fresh liver wasinhibited by certain aliphatic free fatty acids. The influences of chain length,unsaturation, oxidation, and esterification on the potency to inhibit 5a-reductaseactivity were studied. Among the fatty acids we tested, inhibitory saturated fatty acidshad C12 –C16 chains, and the presence of a C¼C bond enhanced the inhibitory activity.


Fig. 3. (cont.)

Esterification and hydroxy compounds were totally inactive. a-Linolenic acid, oleicacid, and lauric acid also showed an ability to inhibit the testosterone-inducedproliferation of LNCaP cells, and this inhibition at least partially came from the 5a-reductase inhibitory activity. This is the first time we have demonstrated that thetestosterone-regulated PSA mRNA expression was markedly suppressed by co-culturing with each fatty acid. This suppression was the result of 5a-reductase inhibitoryactivity, which, at least partially, inhibits the conversion of DHT from testosterone.

Experimental Part

Materials. Unless otherwise specified, chemicals were obtained from Sigma Aldrich Japan Co., Ltd.(Tokyo, Japan). Org. solvents were purchased from Wako Pure Chemical Industries Co. (Osaka, Japan).[4-14C]Testosterone was obtained from Perkin-Elmer Japan Co., Ltd. (Kanagawa, Japan).

Preparation of Rat Microsomes. Microsomes of rat liver from female SD rats (7 weeks of age) wasprepared by a method previously reported by Shimizu et al. with some modifications [29]. From twomature SD female rats, the liver was removed, and minced tissue was homogenized in 4 tissue volumes ofmedium A (0.32m sucrose, 1 mm dithiothreitol, and 20 mm sodium phosphate, pH 6.5). The resulting


Fig. 4. The PSA mRNA expression of a-linolenic acid (a), and lauric acid (b) on the LNCaP cell. Eachcolumn represents the mean�S.D., n¼3. T: testosterone; DHT: dihydrotestosterone. **: p<0.01 against

corresponding T or DHT.

supernatant from the centrifugations was further centrifuged at 105,000�g for 1 h twice. The washedmicrosomes were suspended in 1 pellet volume of medium A, and the dispersion of microsomes wasachieved using a syringe with 18 G, 23 G, and 26 G needles in succession. The microsome suspension wasstored at �808 just before use.

Measurement of 5a-Reductase Inhibitory Activity. The complete reaction mixture included 1 mm

dithiothreitol, 20 mm phosphate buffer (pH 6.5), 1.9 nCi [4-14C]testosterone (0.7 mm), 150 mm testoster-one, 167 mm NADPH, and the microsomes (0.6 mg of protein) in a final volume of 0.3 ml. Theconcentration of testosterone contributed by [4-14C]testosterone was negligible. The sample was added tothe soln. at each concentration. The incubation was carried out for 10 min at 378. It was started by theaddition of aliquot microsomes to the pre-heated reaction soln. in each tube. After 10 min, the incubationwas terminated by adding 10 ml of 3m NaOH. To extract metabolites, 1 ml of Et2O was added, and thetubes were capped and shaken. The org. phase was applied to a silica-gel plate (Kieselgel 60 F254), and theplate was developed in AcOEt/hexane 7 : 3 at r.t. The radioactivity profile was determined with animaging analyzer (FLA-5000 RF, Fuji Film Co., Ltd. Tokyo, Japan). The 5a-reductase activity wascalculated from the percentage of the extent of the conversion of [4-14C]testosterone to [4-14C]dihy-drotestosterone.

Inhibitory Effect on Prostate Cancer Cells. Androgen receptor-positive human prostate cancerLNCaP cells were obtained from American Type Culture Collection. The cells were maintained in RPMI1640 supplemented with 10% fetal bovine serum (FBS) at 378 in a 5% CO2, 95% air-humidifiedatmosphere incubator. The cells were used between passages 5–30 at a split ratio of 1 : 4 at each passage.The cells were plated into 24-well plates with 2�105/well density supplemented with 5% steroid-depleted(charcoal-stripped) FBS. Twenty-four h later, the cells were treated with either vehicle control orandrogens (testosterone or DHT) in the presence or absence of each concentration of sample for another3 d. Cell proliferation was determined by the 3-amino-7-(dimethylamino)-2-methylphenazine (NR)method. The NR soln. was made at 5 mg/ml, and diluted by culture medium to 5 mg/ml. The NR extractsoln. consisted of 49% H2O, 50% EtOH, and 1% AcOH (v/v). The culture medium was changed to NRsoln. and incubated for 3 h at 378, and then the NR soln. was aspirated, and the cells were washed by PBStwice. The NR extract soln. (500 ml) was added to each well to extract for 20 min at r.t. The absorbance ofeach well was measured at 540 nm.

Real-Time Polymerase Chain Reaction (PCR) Analysis. Total RNA was extracted using Isogen(NipponGene, Toyama, Japan) from the LNCaP cells after they were cultured with or without fatty acids.Complementary DNA was synthesized in a final volume of 20 ml that included 1 mg of total RNA (4–5 mlof 0.2 –0.3 mg/ml total RNA), 1 mm oligo-dT 18-mer primer, 10 units of Rnase inhibitor, and 10 units ofAMV Reverse Transcriptase (Takara, Japan) according to the manufacturer�s instructions. Real-timePCR was performed in a final volume of 10 ml with a Line Gene (Bioflux Corporation, Japan). The SYBRPremix Ex Taq kit (Takara, Japan) was used according to the manufacturer�s instructions with a finalconcentration of 0.2 mm for each primer. PCR Amplification was performed as follows: i) an initialdenaturation at 948 for 1 min, ii) 35 cycles, with 1 cycle consisting of denaturation at 948 for 30 s,annealing at 548 at 1 min, and elongation at 728 for 2 min. The following primers were used: for the PSA,the sense primer was 5’-GAGGTCCACACACTGAAGTT-3’, and the anti-sense primer was 5’-CCTCCGAAGAATCGATTCCT-3’; for the b-actin, the sense primer was 5’-CACTGTGTTGGCGTA-CAGGT-3’, and the anti-sense primer was 5’-TCATCACCATTGGCAATGAG-3’. The ratio of gene-specific expression was defined as the relative expression to the actin expression. The data were threeindividual runs�SD.

Statistics. Results were expressed as means�SD. Statistical significance of the cell proliferation wasdetermined by the t-test.

REFERENCES

[1] J. Geller, J. Am. Geriatr. Soc. 1991, 39, 1208.[2] H. M. Arrighi, E. J. Metter, H. A. Guess, J. L. Fozzard, Urology 1991, 38(1 Suppl.) , 4.[3] J. Gobinet, N. Poujol, C. Sultan, Mol. Cell. Endocrinol. 2002, 198, 15.


[4] E. P. Gelmann, J. Clin. Oncol. 2002, 20, 3001.[5] D. W. Russell, J. D. Wilson, Annu. Rev. Clin. Biochem. 1994, 63, 25.[6] P. K. Siiteri, J. D. Wilson, J. Clin. Invest. 1974, 38, 113.[7] J. Geller, J. Albert, D. Lopez, S. Geller, G. Niwayama, J. Clin. Endocrinol. Metab. 1976, 43, 686.[8] N. Kyprianou, H. F. English, J. T. Isaacs, Cancer Res. 1990, 50, 3748.[9] G. Sansone, R. M. Reisner, J. Invest. Dermatol. 1971, 56, 366.

[10] K. D. Bingham, D. A. Shaw, J. Endocrinol. 1973, 57, 111.[11] F. Kuttenn, I. Mowszowicz, G. Schaison, P. Mauvais-Jarvis, J. Endocrinol. 1977, 75, 83.[12] J. Liu, K. Shimizu, F. Konishi, K. Noda, S. Kumamoto, R. Kurashiki, R. Kondo, Food Chem. 2007,

100, 1691.[13] T. Liang, S. Liao, Biochem. J. 1992, 285, 557.[14] H.-J. Niederpr�m, H.-U. Schweikert, J. W. Th�roff, K. S. Z�nker, Ann. N.Y. Acad. Sci. 1995, 768, 227.[15] J. S. Horoszewicz, S. S. Leong, E. Kawinski, J. P. Karr, H. Rosenthal, T. M. Chu, E. A. Mirand, G. P.

Murphy, Cancer Res. 1983, 43, 1809.[16] P. Negri-Cesi, M. Motta, J. Steroid Biochem. Mol. Biol. 1994, 51, 89.[17] H. Weisser, M. Krieg, Prostate 1997, 30, 41.[18] H. Weisser, M. Krieg, Prostate 1998, 36, 235.[19] M. E. Begin, Nutrition 1989, 5, 258.[20] Z. Y. Chen, N. W. Istfan, Prostaglandins, Leukotrienes Essent. Fatty Acids 2000, 63, 301.[21] L. C. M. Chiu, J. M. F. Wan, Cancer Lett. 1999, 145, 17.[22] S. A. Schwartz, A. Hernandez, B. M. Evers, Surg. Oncol. 1999, 8, 143.[23] J. M. Connolly, D. P. Rose, Cancer Lett. 1998, 132, 107.[24] M. Tsujii, R. N. DuBois, Cell 1995, 83, 493.[25] B. A. Narayanan, N. K. Narayanan, B. S. Reddy, Int. J. Oncol. 2001, 19, 1255.[26] J. T. Arnold, H. Le, K. K. McFann, M. R. Blackman, Am. J. Physiol. Endocrinol. Metab. 2005, 288,

E573.[27] J. Imperato-McGinley, R. E. Peterson, T. Gautier, E. Sturla, J. Steroid Biochem. 1979, 11(1B) , 637.[28] T. Liang, E. J. Brady, A. Cheung, R. Saperstein, Endocrinology 1984, 115, 2311.[29] K. Shimizu, K. Kondo, K. Sakai, Y. Shoyama, H. Sato, T. Ueno, Biosci. Biotechnol. Biochem. 2000,

64, 875.

Received March 31, 2008


anti-androgenic activity of fatty acids

Documents