Among the benzoquinolinones LY 191704 and LY 266111 (Eli Lilly) (Fig. 21A and B) possess an ideal 4-N-methyl, 8-chloro, trans, angular hydrogen arrangement (6) leading to a high inhibition potency against type 1 isoenzyme (29). Analysis on human 5a-reductases expressed from transfected cDNAs in simian COS cells indicate that LY 191704 is a specific noncompetitive inhibitor of the human type 1 isoenzyme with an IC50 of 8 nM (37). LY 266111 showed a corresponding IC50 of
9 nM on recombinant CHO cells (35). These benzoquinolinone derivatives demon-
strated that potent inhibition of the type 1 5a-reductase could be realized by a compact rigid tricyclic nucleus that does not require the full steroid ring system (6).
LY 320236 (Eli Lilly) (Fig. 21C) is a potent dual inhibitor with differing modes of activity against the two human 5a-reductase isoenzymes. LY 320236 inhibits
5a-reductase activity in human scalp skin competitively (Ki type 1 ¼ 28.7 nM) and prostatic homogenates noncompetitively (Ki type 2 ¼ 10.6 nM) (38). LY 320236 is currently under phase 1 clinical trials against prostate cancer.
Another interesting class of nonsteroidal compounds are the 6-substituted 1H- quinolin-2-ones (Fig. 21D) (39).
The 1H-benzoquinolizine-3-one derivatives are selective inhibitors of human type 1
5a-reductase. IC50 of AS-97004 (Ares Serono) (Fig. 22) determined on recombinant
CHO cells accounts for approximately 300 nM.
Serono’s antiacne program focused on the preclinical lead compound AS-
601811. Phase 1 trials began in the second quarter of 2001. The agent was orally bioavailable in rodents and dogs and generally well tolerated in animal models. At doses up to 15 mgkg21d21 no toxicity was observed at 15 days after dosing.
Phenylethyl Cyclohex-1-ene Carboxylic Acids
A novel approach to the synthesis of conformationally flexible nonsteroidal mimics of Epristeride is cyclohex-1-ene carboxylic acids (Fig. 23A). The compounds turned out to be good inhibitors of type 2 isoenzyme (e.g., R ¼ CON(i-propyl)2, IC50 ¼ 760 nM). Type 1 isoenzyme was only slightly inhibited. The compounds were tested for inhibition of human 5a-reductase isoenzymes, using prostatic tissue for type 2 and DU 145 cells for type 1 isoenzyme. This cell line, deriving from brain metastases of an epithelial human prostate adenocarcinoma, has been demonstrated to express only type 1 isoenzyme and has been recently used to test inhibitors toward this 5a-reductase isoform (40).
It was recently demonstrated that in the class of A-C ring steroidmimetics the biphenyl-4-carboxylic acids were appropriate 5a-reductase inhibitors. As a conse- quence of structure-activity studies, 3,4-dihydro-naphtalene-2-carboxylic acids and naphthalene-2-carboxylic acids (Figs 23B and C), which are structurally based on the steroidal inhibitor Epristeride, were synthesized and evaluated for
5a-reductase inhibitory activity. Human prostate homogenates and the DU 145 cell line served as enzyme sources.
Apparently, the 3,4-dihydronaphtalene-2-carboxylic and naphthalene-2- carboxylic acid moiety mimicking the steroidal A-B ring is appropriate for gen- erating potent transition state analogs of the steroidal substrate testosterone and androsterone. The most potent inhibitors of the human enzymes show IC50 values in the nM range. These naphthalene derivatives exceeded similarly substituted 1H-quinolin-2-ones in their inhibitory activities. This was to be expected, since in the class of biphenyl type inhibitors, a benzoic acid moiety is superior to a 2-pyridone structure as a steroidal A ring mimetic.
The strong influence of the type and the position of the substituents at the
6-phenyl group as well as the absence or presence of a double bond in 1 position on the activity of the compounds are remarkable. The big difference in their inhibi- tory activities toward type 1 and type 2 isoenzyme in the case of single compounds demonstrates that they are appropriate to elaborate the structural requirements for a selective or a dual inhibition of type 1 or type 2 isoenzyme inhibition (39).
Indolizine- and Indol-Butanoic Acids
FR146687 (Fujisawa Pharmaceutical) (Fig. 24A) shows a noncompetitive inhibition of both isoenzymes in vitro and has no inhibitory effects on other steroid oxido- reductases (41).
FK143 (Fujisawa Pharmaceutical) (Fig. 24B) was shown to inhibit both types of human 5a-reductase equally. The mode of inhibition against both the isoenzymes is noncompetitive. The inhibition constants for the human type 1 and type 2 isoen- zymes were 27 and 20 nM, respectively. Species selectivity between human and rat on the inhibitory activity of FK143 was not found for both the isoenzymes. No other inhibitory effects were shown to other oxidoreductases (42). After a single oral administration of FK143 (100 – 500 mg) on healthy volunteers, DHT plasma concen- trations decreased to about 65% of a predose value (43).
Zinc has long been used in dermatology and is reported to reduce sebum secretion. Metal ions have numerous roles in protein function, especially at the active site of enzymes. They could be complexed to a substrate or modify the structure of the protein. The two types of human 5a-reductases were transfected into SW-13 cells, a human adrenal carcinoma cell line, containing negligible endogeneous
Note: Transfected SW13 human adrenal carcinoma cells were homogenized and incubated with 200 nM [14C]androstenedione in the presence of 50 mM sodium phosphate buffer, pH 7.5, and 400 mM NADPH. Ki values were deter- mined by the Cornish-Bowden method and are expressed as mean + SD. “n.i.” indicates that no significant inhibition was observed.
Source: Adapted from Ref. 44.
5a-reductase activity. The expressed 5a-reductases were analyzed for their sensi- tivity to heavy metal ions using 0.2 mM (14C)androstenedione as a substrate.
The addition of CuCl2, CdCl2, and ZnCl2 led to a potent inhibition of type 1 isoenzyme, with Ki values of 3.47, 0.61, and 1.51 mM, respectively (Table 4). NiCl2 and FeCl3 exerted weaker inhibition, with Ki values of 176 and 250 mM. MnCl2, MgCl2, CaCl2, and LiCl, however, did not significantly inhibit either type 1 or type 2 5a-reductase activities. It is somewhat surprising that type 2 5a-reductase was not inhibited by ZnCl2 or CdCl2, as it contains a triple stretch of histidine resi- dues at amino acid positions 229 to 231, and thus is more likely to bind zinc and its analogs than is type 1. Of the cations studied, only CuCl2 significantly inhibited type 2 5a-reductase activity with a Ki value of 11 mM.
Anions only showed a moderate influence on the type 1 isoenzyme inhibition.
ZnSO4 has an effect similar to ZnCl2, whereas Zn(CH3COO)2 showed a slightly decreased inhibition (44).
Phytotherapeutics are generally well accepted by consumers and, therefore, received considerable attention from the pharmaceutical industry. However, a limiting factor for the use and effectiveness of phytotherapy is the lack of standardi- zation of most of these products.
The berries of saw palmetto (Serenoa repens), a small palm native to the south east United States, possess a dual 5a-reductase inhibition activity, due to their high content of phytosterols (b-sitosterol (Fig. 25A), stigmasterol (Fig. 25B), lupeol (Fig. 25C), lupenone (Fig. 25D), and cycloartenol (Fig. 25E). Saw palmetto was used by native Americans as herbal treatment for enuresis, nocturia, atrophy of testes, impotence, inflammation of the prostate, and low libido in men. Standardized lipo- philic saw palmetto extracts are commercially available: e.g., Reguw-Seb (Penta- pharm) for topical treatment of seborrheic skin conditions and Permixonw (Pierre Fabre Medicament) for systemic application to treat benign prostatic hyperplasia. A
90-day topical treatment of a Reguw-Seb, containing emulsion lead to a statistically significant decrease of 46% (p , 0.01) in severity of seborrheic skin conditions
(Fig. 26). Permixonw was launched in Europe in 1984 but has no FDA approval. The lipido-sterol extract markedly inhibits both the human isoenzymes. Type 1 isoenzyme is noncompetitively (Ki ¼ 7.2 mg/mL) and type 2 isoenzyme uncompetitively (Ki ¼ 4.9 mg/mL) inhibited (45).
Further dual phytotherapeutic 5a-reductase inhibitors include, among other extracts of Pygeum africanum, Artocarpus incisus (14), Thuja orientalis, Lami- naria saccharina, Arnica Montana, Cinchona succirubra, Eugenia caryophyllata, Humulus lupulus, Hypericum perforatum, Mentha piperata, Rosmarinus officinalis, Salvia officinalis and Thymus officinalis (46); furthermore, diterpens (29), flavins (25), and isoflavonoids such as genistein (Fig. 27A) and daidzein (Fig. 27B), lignans, resveratrol (Fig. 28) (14), curcumin (Fig. 29) (IC50 type 1 ¼ 3 mM, IC50 type 2 ¼ 5 mM) (47), and certain polyunsaturated fatty acids. The relative inhibitory potencies of unsaturated fatty acids are, in decreasing order: g-lino- lenic acid (Fig. 30A), a-linolenic acid (Fig. 30B), linoleic acid (Fig. 30C), palmi- toleic acid (Fig. 30D), oleic acid (Fig. 30E), and myristoleic acid (Fig. 30F). Other unsaturated fatty acids such as undecylenic acid, erucic acid, and nervonic acid, are inactive (48).
inhibitory effect of zinc sulfate and azelaic acid on 5a-reductase activity was shown (49). Azelaic acid also possesses bacteriostatic activity to both aerobic and anaerobic bacteria including Propionibacterium acnes. Moreover, azelaic acid is an antikeratinizing agent, displaying antiproliferative cytostatic effects on keratinocytes, and modulating the early and terminal phases of epidermal differentiation. Therefore, several topical dermatological antiacne preparations containing a combination of a zinc salt and high concentrations of azelaic acid are available.
The most challenging and urgent work to achieve rigorous progress in the develop- ment of novel drugs, based on the inhibition of 5a-reductase, will be the purifi- cation of these enzymes, for which all efforts have failed so far because of their unstable nature during purification. Crystallization and X-ray diffraction structural analysis of the purified enzymes will be another important topic. Computer- supported molecular modeling, then, should provide directory data for devel- oping more potent and selective inhibitors. On the other hand, the mechanism of action of enzymes and inhibitors needs to be further investigated. Steroidal inhibitors have already been extensively studied. However, developing of more efficient nonsteroidal compounds with simpler molecular structures and higher selectivity at lower costs is needed. This will also provide new information about enzyme and inhibitor interactions, and might lead to novel drugs that act differently from steroidal derivatives (6).
The prevalence of symptomatic benign prostatic hyperplasia is around 14% in men in their 40s, 24% in men in their 50s, and 43% in men beyond the age of 60. The number of patients undergoing operations for benign prostatic hyperplasia has decreased by approximately 60% in the United States over the past decade, due to the introduction of 5a-reductase inhibitors. Moreover, prostate cancer is the second leading cause of cancer death in males in the United States (20). Although acne is the most common skin disorder treated by dermatologists, it is, therefore, not surprising that the majority of the publications focus on the investigation of novel type 2 and dual acting 5a-reductase inhibitors. The contribution of DHT in the serum, which is partially derived from type 1 5a-reductase in the liver and the small amount of type 1 5a-reductase in the prostate, may play a role in main- taining prostatic enlargement. Thus, it is an effort to increase efficacy of treatment for benign prostatic hyperplasia that inhibits both types of 5a-reductase.
Beside the phytotherapeutic preparation Permixonw, to date Finasteride and Dutasteride are still the only synthetic 5a-reductase inhibitors, which are on the market as active ingredients in drugs. Anyhow, some inhibitors of type 1 isoenzyme have, meanwhile, been developed and studied in men with acne. To our knowl- edge, however, no successful outcomes have been published yet. The potential of drugs, such as Dutasteride that inhibits both the isoenzymes, for use in acne treat- ment remains to be determined.
Though systemically applied inhibitors of 5a-reductase could prove to be safe in women, they are currently not used because of potential fetal feminization and teratogenic risks (1). Currently published clinical results on acne treatment are rather disappointing (50,51). Therefore, effective topical applications may be a useful alternative. In humans, clinical responses to topical 5a-reductase inhibitors have been moderate, yet measurable. Actually, there exist no topical
pharmaceuticals claiming exclusively 5a-reductase inhibition activity, although there are antiacne products on the market containing, for example, zinc acetate and azelaic acid, respectively. Most of the synthetic or phytochemical 5a-reductase inhibitors are still under in vitro investigation, animal studies, or clinical trials. Finally, it nevertheless remains to be pointed out that current research in acne therapy favors investigations of topical androgen receptor blockers over topical