Medicinal Chemistry

0 downloads 0 Views 3MB Size Report
occurs because of the high bone resorption rate. Estrogen formation is controlled by 17-β hydroxysteroid dehydrogenase 17-β HSD enzymes, where 17-β HSD ...
Future

Review

For reprint orders, please contact [email protected]

Medicinal Chemistry

Type 2 17-β hydroxysteroid dehydrogenase as a novel target for the treatment of osteoporosis

Low estradiol level in postmenopausal women is implicated in osteoporosis, which occurs because of the high bone resorption rate. Estrogen formation is controlled by 17-β hydroxysteroid dehydrogenase 17-β HSD enzymes, where 17-β HSD type 1 contributes in the formation of estradiol, while type 2 catalyzes its catabolism. Inhibiting 17-β HSD2 can help in increasing estradiol concentration. Several promising 17-β HSD2 inhibitors that can act at low nanomolar range have been identified. However, there are some specific challenges associated with the application of these compounds. Our review provides an up-to-date summary of the current status and recent progress in the production of 17-β HSD2 inhibitors as well as the future challenges in their clinical application.

Sex hormones are steroidal molecules derived from cholesterol [1] . They are classified in two essential types: androgens including androstenedione, testosterone and the most active androgen dihydrotestosterone; and estrogens including estriol, estrone and estradiol (E2) which is the most active estrogen. These hormones have a wide spectrum of biological effects on different organs in the human body [2] . In addition to their role in sexual differentiation and reproduction, androgens and estrogens have many other bioeffects in men and women  [3–8] . Indeed, it is found that estrogens have anti-inflammatory effects [9] , vasoprotection effects that help reducing the risk of cardiovascular disease [10] and complex effects on the liver and strong effects on bone [11] . Furthermore, estrogens also have a potent role in the maintenance of mental health [12–14] . Androgens in general have an important role in increasing the skeletal muscle mass [15] , altering the structure of the brain [16] inhibiting fat deposition [17] . Several known disorders are the results of variations in sexual hormones concentrations or expression of their receptors [18] . Consequently, drugs that target sexual hor-

10.4155/FMC.15.74 © 2015 Future Science Ltd

mones regulating and synthetic enzymes can have great clinical value. Estrogens and androgens production is controlled by several enzymes including some cytochromes (such as CYP450 aromatase and CYP450 17A1), some 3-β HSDs (such as 3-β HSD type 1 and 2) and several 17-β hydroxysteroid dehydrogenases (17-β HSDs) such as (17-β HSD type 1 and 2) where these 17-β HSDs are involved in the last step of steroid hormone formation and degradation (see Figures 1 & 2 for more details) [2,4,19–21] . Fourteen isoenzymes of 17-β HSD family were isolated and identified so far. Some of them catalyze the formation of E2 and testosterone (T), by the reducing the weaker precursors of sexual hormones, estrone and androstenedione respectively. Other 17-β HSDs oxidize the active hormones to the inactive forms [22] . Briefly, 17-β hydroxysteroid dehydrogenase type 1, 7 and 12 are the isoenzymes that catalyze the production of E2, and 17-β hydroxysteroid dehydrogenase type 3 and 5 contribute to the production of T, while 17-β hydroxysteroid dehydrogenase type 2, 4, 8 and 10 are the isoenzymes responsible for oxidize the active sex hormones E2 and/or T resulting in the less active forms of these steroids (Figure 1)  [23] .

Future Med. Chem. (2015) 7(11), 1431–1456

Jalal Soubhye*,1, Ibaa Chikh Alard2, Pierre van Antwerpen1,3 & François Dufrasne1 1 Laboratoire de Chimie Pharmaceutique Organique, Faculté de Pharmacie, Université Libre de Bruxelles, Campusplaine, CP 205/5, 1050 Brussels, Belgium 2 Laboratoire de Pharmacie Galénique et Biopharmacie, Faculté de Pharmacie, Université Libre de Bruxelles, Brussels, Belgium 3 Analytical Platform of the Faculty of Pharmacy, Université Libre de Bruxelles, Brussels, Belgium *Author for correspondence: [email protected]

part of

ISSN 1756-8919

1431

Review  Soubhye, Chikh Alard, van Antwerpen & Dufrasne

17-β HSDs

O

Type 1: Breast, ovary, endometrium, placenta Type 7: Liver, lung, thymus Type 12: Breast, liver, placenta, kidney

OH H

H H

H

H

Type 2: Liver, breast, endometrium, placenta, prostate, bones Type 4: Liver, lung, placenta Type 8: Liver, placenta, kidney

HO

Estrone

H

HO

Estradiol

17-β HSDs O H H

Type 3: Testis Type 5: Prostate, liver

OH H

H

O

Androstenedione

H

Type 2: Liver, prostate, bones Type 9: ND Type 10: CNS, brain

H

O

Testosterone

Figure 1. Sex steroids metabolism catalyzed by 17β-HSD enzymes and their distribution in the tissues. ND: Not defined.

Type 2 17-β HSD (EC = 1.1.1.51) is the main oxidative isoenzyme that deactivate the sexual hormones in bones [24] . This isoenzyme was isolated and identified for the first time by Wu et al. in 1993 [25] . In human being, bones always undergo a remodeling process in which mineralized bone tissues are resorbed by osteoclasts followed by their replacement through osteoblasts action causing little change in the overall bone structure [26] . Bone remodeling is very important for regulating the calcium homeostasis, shaping the skeleton during growth and repairing microdamaged bones [27] . Several hormones regulate the function of osteoblasts and osteoclasts in direct or indirect ways. Some of them stimulate the desorption of the bone tissues such as: RANK ligand, macrophage colony stimulating factor, parathyroid hormone (PTH), vitamin D, prostaglandin E2 (PG E2), IL-1 and -6 and TNF-β. Other hormones are responsible of the inhibition of the desorption including osteoprotegerin (OPG), TGF-β, IFN-γ, IL-4 and -18 and E2 [28,29] . To date, the way by which E2 decreases the bone mass loss is not fully understood. However, it has been demonstrated that the E2 decreases or inhibits osteoclasts in several ways. It can decrease the RANK ligand and increase OPG and TGF-β [30] . Moreover, there is a strong link between the low levels of estrogen in postmenopausal women and the progressive loss in bone mass [31] .

1432

Future Med. Chem. (2015) 7(11)

As it is shown above, increasing the concentration of E2, especially in the bone tissues, may be useful for the treatment of osteoporosis. Increasing of E2 can be achieved by two essential strategies: either by increasing the production of E2 or by inhibiting its degradation. As 17-β HSD2 is the main enzyme that catalyzes the degradation of E2 in bone tissue, thus inhibiting this enzyme can increase E2 concentration to help in treating osteoporosis especially in postmenopausal women. Biology of the enzyme 17-β HSD2 (synthesis, structure & function) Synthesis of 17-β HSD2 protein

The gene that encodes the enzyme 17-β HSD2 is located on chromosome 16q24 [32] . The mRNA has been found in a wide variety of tissues including: endometrium, breast, placenta, liver, pancreas, prostate, gastrointestinal, urinary tracts and the adrenals of adults [33–36] . Despite lots of researches on this enzyme, little data about the regulation of the expression of 17-β HSD2 are available. However, many studies indicated that transcription factor specific protein-1(Sp1) and 3 (Sp3) stimulate the expression of primary structure of 17-β HSD2 [37,38] . These specific proteins are regulated by several agents. Further studies demonstrated that the transcription factor specific proteins can be induced by paracrine factors that are stimulated by progesterone receptor. Sp1 and Sp3

future science group

Type 2 17-β hydroxysteroid dehydrogenase as a novel target for the treatment of osteoporosis 

can be stimulated by the highly active sex steroids that are the biological substrates of 17-β HSD2 [37–39] . Furthermore, several studies showed that Sp1 and Sp3 also are regulated by retinoic acid receptors [40] . Actually, it has been found that in addition to its role in oxidation of sex hormones to reduce their activity, 17-β HSD2 might be important for the action of retinoids [41] . On the other hand, like most enzymes, 17-β HSD2 may be regulated by protein translation [42] . In fact, high level of mRNA of 17-β HSD2 inhibits or reduces the expression of the enzyme [43] . Structure 17-β HSD2 protein

The primary structure of 17-β HSD2 consists of 387 amino acids with a molecular weight of 42,782 daltons. The enzyme is located in the endoplasmic reticulum. Although no x-ray structure of the mature enzyme is available, using the hidden Markov model Sonnhammer et al. proposed that the active site of 17-β HSD2 has a cluster of positively charged amino acids followed by about 33 nonpolar amino acids forming a hydrophobic core beside the N-terminus of the

Placenta

Review

Key terms RANK ligand: Also known as osteoclast differentiation factor, is a member of the tumor necrosis factor , it controls bone regeneration and remodeling. It also may have an effect on the immune system. Osteoprotegerin: Member of the tumor necrosis factor; it inhibits the differentiation of the osteoclast precursor to a mature osteoclast by blocking the RANKL-RANK interaction. It is a part of the RANK/RANKL/OPG signaling pathway that regulates osteoclast differentiation. 17-β HSD2: Enzyme that uses nicotinamide adenine dinucleotide as co-enzyme for oxidize estradiol and testosterone in order to deactivate these hormones. Transcription factor specific protein: Controls the rate of transcription of a part of DNA to messenger RNA by binding to a specific DNA sequences.

protein. Close to the N-terminus of 17-β HSD2 two proposed transmembrane helices were predicted [44] . The co-enzyme by which 17-β HSD2 mediates the oxidation of sex hormones is NAD + while for 17-β HSD1 and the other reductive 17-β HSDs NADPH plays this role (Supplementary Figure 1) [45,46] .

Estrogen activity

Estradiol 17b HSD1

Androstenedione

P450 Aromatase

17b HSD2

Estrone

Adrenal cortex

During pregnancy

Ovary

Peripheral target cells Tissues that have ER such as kidney, lung and heart Estrone

Pregnenolone

P450 17a 3b HSD1

Dehydroepiandrosterone

3b HSD2

Progesterone

Dehydroepiandrosterone

17b HSD1

17b HSD2

3b HSD2 3b HSD1 P450 17a

Androstenedione

Androstenedione

P450 aromatase

17b HSD1, 7

Testosterone

P450 aromatase

Other tissues (e.g. brain, skin and adipose cells)

Adult female

Estradiol

Estrogen activity

Estradiol 17b HSD1

Androstenedione

Estradiol

Estrone

17b HSD1, 7

P450 aromatase

Estrogen activity

Adult female and postmenopausal age

17b HSD2

Estrone

Figure 2. Scheme of female sex hormones synthesized in different tissues. The main source of pregnenolone, which is the precursor of sex hormones, is the adrenal cortex [47] . During pregnancy more than 50% of E2 is produced in the placenta. In adult women, most of E2 come from ovary (especially within the follicles) and small amounts from other tissues such as adipose cells, skin, liver, breast and neural tissues. At postmenopause, the ovary stops produce E2 and the production of E2 is limited to small amounts produced in the other tissues. The production of E2 is under control of several enzymes including: 3β-HSDs, CYP17A1 (that has both 17α-hydroxylase and 17, 20-lyase activities), aromatase and 17-β HSDs [2] . ER: Estradiol receptor.

future science group

www.future-science.com

1433

Review  Soubhye, Chikh Alard, van Antwerpen & Dufrasne Function 17-β HSD2

In adult females, E2 concentration changes according to the Phase of the menstrual cycle. Majority of E2 is released from the ovarian granulosa cells [47] . During pregnancy, the E2 level significantly increases and the 50% of E2 amount is produced in the placenta [2] . However, the major source of E2 in postmenopausal women is conversion of androstenedione to estrone by the aromatase which can be found in adipose cells, liver, skin and other tissues. 17-β HSD1 in turn transforms estrone to E2 (Figure 2) . In men, more than 80% of the E2 is produced by transforming T to E2 by the peripheral aromatase [2] . To maintain and control the sex hormones levels, different enzymes inactivate the hormone when its concentration is high (cytochromes [such as Cyp 2C8] and some members of the hydroxysteroiddehydrogenase family) [48,49] . 17-β HSD2 is the main enzyme that transforms the active E2 to the less active estrone which has 10% of the activity of E2. Testosterone and 5α-dihydrotestosterone are also substrates for 17-β HSD2 which converts them to the ketonic inactive form of these hormones (Figure 1) . Furthermore, it was shown that 17-β HSD2 can oxidize the steroid hormone 20α-dihydroprogesterone into progesterone. There is strong evidence that the main role of 17-β HSD2 is to prevent the excessive estrogenic action in tissues and cells [49] . The high amount of 17-β HSD2 in placenta indicates that this enzyme is very important to reduce the E2 level in fetal circulation  [50,51] . It has been suggested that in addition to its role in oxidation of the active sex hormones, 17-β HSD2 has an important role in the physiological action of retinoic acid which is important for growth and other biological functions [41,52] . Pathologies linked to 17-β HSD2 As the enzyme 17-β HSD2 has an important role in sex steroids level control and a broad distribution in a wide variety of tissues, it may be implicated in several diseases. Low activity of 17-β HSD2

Low activity of 17-β HSD2 can contribute to a number of health disorders. It has been reported that 17-β HSD2 cannot be detected in endometriotic lesions [53] . While E2 plays a critical role in its development [54–57] , progesterone inhibits the endometriosis lesions. 17-β HSD2 deficiency contributes to increasing the level of E2 and decreasing the level of progesterone which are the key events responsible for development of endometriosis  [58–61] . Although 17-β HSD2 synthesis may be induced by progesterone, the lack of this enzyme in endometriosis is not due to the low stimulation of progesterone receptors [53] . It has been demonstrated that

1434

Future Med. Chem. (2015) 7(11)

the failure of stromal cells of endometriosis to produce paracrine factors is the main reason leading to endometriosis  [37] . Lack or low levels of 17-β HSD2 were also found in scalp hair of hirsute women. This low level causes an increase in active androgens resulting in high free androgen index, as it occurs in men [62] . In addition, it is found that the complete absence of 17-β HSD2 in the mice causes several manifestation including abnormalities in the kidney, the brain and the placenta resulting in fetal mortality [63] . 17-β HSD2 in tumors

The correlation between 17-β HSD2 and tumors affecting sexual organs has been largely documented. The substrates of this enzyme, E2 and T, can facilitate the development of breast and prostate cancer respectively via stimulation of cell proliferation of the corresponding tissues [64] . Several authors measured the expression of 17-β HSD2 mRNA in tumor tissues. Low level of the enzyme was detected in several breast tumor cell lines [65–68] . Indeed, it has been proposed that the value of the ratio 17-β HSD2/17-β HSD1 gives more valuable data about the prognosis of tumor. It has been found that an increasing ratio correlates with a better prognosis [69] . It must be kept in mind that all of these tumor cell lines that show low level of 17-β HSD2 are from breast cancer with positive estradiol receptor (ER-positive). In contrary, African–American breast tumor cell lines that are ER-negative show an increase in 17-β HSD2 gene expression. Although the incidence of breast cancer at African–American women is 13% lower than at Caucasian women, breast cancer mortality rates is 28% higher among African–American women compared with Caucasian women. It has been suggested that the high level of 17-β HSD2 may contribute to worse clinical outcome among African– American patients who have negative estradiol receptor (ER-negative) [70] . The expression of 17-β HSD2 has been studied in other types of sexual organs tumors. For example, in endometrium tumor cells, the expression of 17-β HSD2 was found to be decreased dramatically, or no 17-β HSD2 was detected. On the contrary, 17-β HSD5 was detected in most of endometrium tumor cells, while no 17-β HSD1 was found in all endometrium tumor cells [71] . These findings indicate that although both breast and endometrium cancers are estrogen-dependent tumors, they display different estrogen metabolic and biosynthetic rates, and 17-β HSD2 and 17-β HSD5 play an important role in the regulation of estrogen production in endometrial carcinoma [71] . 17-β HSD2 was detected with high expression in some (not all) epithelial ovarian tumors. 17-β HSD5 also was present in high concentration in these

future science group

Type 2 17-β hydroxysteroid dehydrogenase as a novel target for the treatment of osteoporosis 

types of tumors indicating that these enzymes have a pivotal role in ovarian tumors [72] . High expression of 17-β HSD2 was found in one tumor cell line of prostate among the studied cell lines but the involvement of the enzyme in disease progression has not been understood yet [32] . It has been demonstrated that the enzymes 17-β HSD1–4 are present in the bone cancer cells but also in normal bone tissues at the same concentrations. Thus in the bone tumor it seems that 17-β HSDs have no clear-cut influence in the development or in the growth inhibition of the tumor [73] . As it was discussed previously, 17-β HSD2 and the other 17-β HSDs are important for regulating the activity of estrogens and androgens that stimulate the proliferation of the cells, so they may contribute in tumor development. Depending on these data, the conditions when these enzymes could be inhibited or stimulated should be determined precisely to maximize the risk–benefit ratio. Osteoporosis

Osteoporosis is a condition of low bone mass and density and microarchitectural disruption resulting in fractures with minimal trauma [27] . The loss of bone mass occurs because of the deficiency in the mineralization of bone rather than the inadequate bone formation. Osteoporosis is divided into three major types, the two first ones representing primary adult osteoporosis and the third one being a secondary disorder. Type 1 osteoporosis is associated with postmenopause. In this type the decrease of E2 accelerates the rate of resorption via increasing the activity of osteoclasts. Associating with aging, the rate of bone loss increases by decreasing the activity of osteoblasts causing low bone formation. This process occurs in type 2 osteoporosis [27] . In type 3 or secondary osteoporosis more than 30% of the cases are induced by drugs (such as corticosteroids, diphenylhydantoin, heparin and warfarin) and diseases (such as anorexia nervosa, coeliac disease and ulcerative colitis) [27] . The most successful strategy of the treatment of secondary osteoporosis is prompt resolution of underlying cause. In the USA, approximately 25% of women who reach the age of 50 will present osteoporotic fractures [27] . While the primary osteoporosis is loss of trabecular bone the secondary osteoporosis is characterized with the loss of cortical and trabecular bone due to the long-term remodeling inefficiency. The high rate of osteoporosis incidence led medicinal chemists to turn their attention to osteoporosis treatment. In order to treat the primary types, the targets of osteoporosis drugs are osteoblasts and osteoclasts via direct or indirect way [27] . The importance of E2 in bone formation is well documented [74–76] . It acts directly on osteo-

future science group

Review

blast  [77,78] . Therefore, the enhancement of E2 in the bone tissues is a feasible strategy to reduce the bone loss [27] . However, several strategies are used for treatment of osteoporosis. Most of these approaches can be classified either under antiresorption agents such as: calcium, vitamin D, estrogens, calcitonin, bisphosphates, denosumab and thiazide diuretics, or under bone-forming agents such as fluorine, androgens and PTH [79–92] . The most important drugs as well as their effects on bone tissues are illustrated in Table 1. Inhibition of 17-β HSD2 Pros & cons of 17-β HSD2 inhibition

One strategy for treating osteoporosis in postmenopausal women is to increase the estrogenic activity in bone tissues. Until now, estrogen replacement therapy (ERT) and selective estrogen receptor modulators (SERMs) have been used for this purpose but with critics about their risk/benefit ratio. Decreasing the metabolism of the most active estrogen (namely E2) by inhibition 17-β HSD2 which is the responsible of this process represents a new strategy for increasing the estrogenic activity in bone tissues. In addition, decreasing of T in men with aging impairs the bone structure. Bagi et al. demonstrated that the inhibition of 17-β HSD2 in both sexes maintains bone formation and bone strength [93] . Indeed, in the last decade, several studies involved the inhibition of this enzyme to acquire a new approach for treating osteoporosis type 1. However, some considerations must be kept in mind when finding useful 17-β HSD2 inhibitors for clinical use. As the change in estrogen concentration may contribute to dangerous disorders in some tissues such as breast and endometrium, inhibition of 17-β HSD2 must take place selectively in bone tissues. The selectivity of the inhibitor for 17-β HSD2 versus the other 17-β HSD isoenzymes is very important because the inhibition of the enzymes involved together in synthesis and catabolism of E2 cannot increase its concentration. On the other hand, 17-β HSD2 is protected due to its location in the endoplasmic reticulum inside the cells, and consequently the inhibitor must have the right hydrophilicity/lipophilicity balance to be able to cross the cell membranes [94] . Key terms Estrogen replacement therapy: Giving one or more estrogen receptor agonists to women who have low levels of estrogen and progesterone. Selective estrogen receptor modulators: Class of molecule that act on the estrogen receptors. This action may be agonist or antagonist according to the tissues.

www.future-science.com

1435

Review  Soubhye, Chikh Alard, van Antwerpen & Dufrasne

Table 1. Most important drugs used in treatment of osteoporosis. Drug

Effect

Ref.

Calcium and vitamin D They increase the bone mineral density and tend to reduce the incidence of vertebral fractures. and its derivatives The effect is small but significant

[79]

Bisphosphonates

They are used as a first-line for the treatment and prevention of osteoporosis. These compounds inhibit the activity and the proliferation of osteoclasts and decrease the rate of bone mass loss

[80,81]

Calcitonins

They are antiresorptive agents. They increase bone mineral density by their effects on Ca +2. They have also analgesic effects which can be achieved within 2 weeks

[82]

ERT

These drugs inhibit osteoclasts, increase calcitonin releasing and vitamin D receptors density. Progesterone must be taken with these drugs in order to prevent breast and endometrium cancers. Combination therapy of this type of drugs with Bisphosphonates has increased bone mineral density but with no effects on fracture reduction

[83]

SERMs

Raloxifene, femarelle and lasofoxifene are SERM drug that are used for osteoporosis treatment. These compounds have agonist activity in some tissues (such as bone and cardiovascular) and antagonist activity in others (such as breast and uterus). Due to these effects, SERM can be used alone for treating the osteoporosis. In bone tissues SERMs have the same effects as ERT. Combination therapy of SERMs with bisphosphonates has increased bone mineral density but with no effects on fracture reduction

[84]

Tibolone

This synthetic steroid has oestrogenic, progestogenic, and androgenic effects. Although this drug has good results in decreasing the risk of fracture they cause an increase in the risk of stroke

[85]

Strontium ranelate

It classified under both anti-resorption and bone-forming agents. It increases the bone mineral density at both spine and hip

[86]

Denosumab

It is a monoclonal antibody of RANKL

[92]

Inhibitors of cathepsin These agents considered as antiresorption, due to their inhibition effects on cathepsin K which is K a lysosomal protease involved in bone resorption. These agents are under investigation now

[88]

Romosozumab and blosozumab

They are humanized antibodies against sclerostin which is a bone morphogenetic protein antagonist

Fluoride

Fluoride sodium stimulates the bone formation due to its direct anabolic effect on the osteoblast

[87]

Parathyroid hormone

It has anabolic effects on bone, and increase bone mineral density

[88]

Androgens

Such as nandrolone. They have the same effects of ERTs on bone mineral density. They are considered as bone-forming agents

[89]

HMG-CoA reductase inhibitors (statins)

It is found that statins have bone-forming properties

[90]

Thiazide diuretics

They reduce urinary calcium excretion

[87]

Ipriflavone

It is a synthetic isoflavone that increases the bone mineral density due to its effect on inhibiting the bone resorption

[87]

Vitamin K

It is found that vitamin K plays a role in bone metabolism

[91]

Anti-semaphorin 4d

They are siRNA molecules that interfere with Sema4d gene expression. Sema4d is a protein that inhibits bone formation. This class of drugs is under investigation now

[131]

[132,133]

Growth hormone and They are considered as potent bone-forming agents IGF-1 OPG

[92]

This protein has antiresorption bone effects. The use of this protein for osteoporosis treatment is under investigation

[134]

ERT: Estrogen replacement therapy; SERM: Selective estrogen receptor modulator.

In vitro assay of 17-β HSD2 inhibition

In order to determine the structure–activity relationship (SAR) of the designed inhibitors, an easy and fast biological in vitro test must be developed to screen numerous compounds in high-throughput conditions. For the enzyme 17-β HSD2 there are several screening

1436

Future Med. Chem. (2015) 7(11)

tests. However, all of these procedures depend on the measurement of the radioactivity of the substrate. The labeled substrate such as [3H]-E2 and [3H] Δ4-dione, with or without unlabeled substrate, is incubated with the enzyme and the inhibitor. After certain time, the reaction is stopped and then the steroids are extracted

future science group

Type 2 17-β hydroxysteroid dehydrogenase as a novel target for the treatment of osteoporosis 

and separated by a chromatographic method. The amounts of oxidized and reduced steroids are measured. In this test, increasing amounts of E2 mean that the activity of the enzyme decreases [95] . Both types 17-β HSD2 and 1 are obtained from human tissues (especially placenta or breast cancer cells) by homogenization and centrifugation. 17-β HSD2 and 17-β HSD1 are found, respectively, in microsomial and cytosolic fractions [96,97] . In fact, when the enzymes were isolated from mice or rats, the results cannot reflect the real inhibition because the 17-β HSD2 of the rat has only 60% homology to the human enzyme. In addition, it must be known that the evaluation of inhibition is different between the free cellular assay (in which the enzyme is isolated) and the cellular assay in which the plasmid that carries the enzyme gene is inserted into a cancer cells such as HEK-293 in order to make the cells express the enzyme. This difference is due to the capacity of the inhibitor to penetrate the cell membranes [98] . 17-β HSD2 inhibitory agents & their structure–activity relationship studies To develop promising therapeutic agents, medicinal chemists must find new hits as starting compounds for designing future lead compounds. Several methods can be applied in order to find these hits. Virtual, random and rational screenings are the most frequently used techniques [99] . However, for 17-β HSD2, it is very difficult to use the virtual screening because of the lack of the x-ray data of this enzyme [44] . Natural compounds

Several authors demonstrated the biological effects of some classes of natural compounds on 17-β HSD2 [100] . Several classes of extracted compounds from plants were evaluated as 17-β HSD2 inhibitors [101–103] . Among all the natural compounds classes (alkaloids not included), only the flavonoids showed an inhibitory effect on 17-β HSD2  [103] . The structure–activity relationship of flavonoids demonstrated that flavone has a good activity, but the introduction of one hydroxyl group on position 3 increases the activity (Figure 3). Introducing other hydroxyl group on another position (except position 7) decreases the activity probably because of the augmentation of hydrophilicity. Hydroxyl on position 3 or position 7 is crucial because these groups are important for binding with active site of the enzyme. These natural compounds have an activity at μM range (Figure 3) [103] . Synthetic compounds Steroid agents Spiro-γ-lactone inhibitors

In 1994, Auger et al. published a paper that describes the first synthetic 17-β HSD2 inhibitor (3-hydroxy-

future science group

5' 1 O

8 7 6 5

4

6' 2 1' 2' 3

Review

4' O

3'

OH O

O

Flavone IC50 = 1.3 µM

HO

3-hydroxyflavone IC50 = 0.4 µM

O OH O

3,7-dihydroxyflavone IC50 = 0.6 µM Figure 3. Activity of flavonoids as 17β-HSD2 inhibitors.

19-nor-17α-pregna-1, 31, 5(10)-triene 21, 17-carbolactone (SP1). This inhibitor is a steroidal molecule which has a structure similar to these of the enzyme substrates and the structure of the natural compound 4-androstene-3,17-dione which inhibits the enzyme at low μM range (IC50 = 1.4 μM) (Figure 4A) . It shows the same structure of E2 with a spiro-γ-lactone on position 17 instead of a hydroxyl group (Figure 4A)  [104] . This inhibitor was showed to be selective versus 17-β HSD1, 17-β HSD3 and 17-β HSD5 [95] . The high potency of SP1 (IC50 = 0.27 μM) led the authors to synthesize other spiro-γ-lactone steroid derivatives. Starting from estrogens and androgens, spiro-γ-lactone was inserted on position 17 to make the backbone of the molecular series. Several compounds were synthesized and tested. All the compounds derived from estrogens (with aromatic group) have activities higher than those derived from androgens. The protection of the hydroxyl group on position 3 (ether or ester), dramatically decreases the activity. This downward trend is proportional to increasing volume of the protective group. This indicates that the hydroxyl is very important for binding to the active site of the enzyme. However, no compound of this series was found to be more active than SP1 (Figure 4A) [105] . In attempt to improve the activity of SP1, the lactone ring was opened to obtain compounds with oxoalkyl groups at position 17, but all the compounds of this series were inactive. Among other compounds with 6- (SP2), 7- (SP3) and 8-membered lactone rings the former was shown to have the best activity with potency at nanomolar range (SP2) (Figure 4A)  [106] . Starting from compound SP2, derived molecules were obtained by the addition of hydrophilic or hydro-

www.future-science.com

1437

Review  Soubhye, Chikh Alard, van Antwerpen & Dufrasne

A

O

O 12

17

11 1 2 3

O

10 4

5

8

14

15

7

R3

R1

HO

6

R2

SP1 IC50 = 0.27 µM

4-androstene-3,17-dione IC50 = 1.4 µM

O

O

O

HO SP4: n = 1 or 2 low activity SP5: n = 3 : IC50 = 200 nM

SP3: IC50 = 150 nM

SP2: IC50 = 6 nM O

O

O

O

O O

O

O

R

O (CH2)n

HO

HO

NS1: R = 3-OH NS2: R = 4-OH

Low activity

O

O

B

R1 = O or lacton R2 = H or O R3 = H or SCOCH3 Double bond position: 1–2 or 4–5

16

13

9

O

O

O

R

R

NS3: R = 3-OH: 25% inhibition at 1 µM NS4: R = 4-OH

C

R

NS5: R = 3-OH NS6: R = 4-OH

NS7: R = 3-OH: 24% inhibition at 1 µM NS8: R = 4-OH

F

F

F

F

O HO

HO

FEST4 17-β HSD2 inhibition 90% 17-β HSD1 inhibition 66%

HO

FEST3 17-β HSD2 inhibition 67% 17-β HSD1 inhibition 42%

HO

F

F

F

F

FEST1 17-β HSD2 inhibition 32% 17-β HSD1 inhibition 39%

FEST2 17-β HSD2 inhibition 32% 17-β HSD1 inhibition 45%

F

F

F O HO

FEST6 17-β HSD2 inhibition 81% 17-β HSD1 inhibition 68%

FEST5 17-β HSD2 inhibition 70% 17-β HSD1 inhibition 60%

F

HO FEST10 17-β HSD2 inhibition 95% 17-β HSD1 inhibition 43%

1438

Future Med. Chem. (2015) 7(11)

HO

HO

HO

FEST7 17-β HSD2 inhibition 60% 17-β HSD1 inhibition 47%

FEST8 17-β HSD2 inhibition 61% 17-β HSD1 inhibition 61% F

F

F

HO FEST9 17-β HSD2 inhibition 85% 17-β HSD1 inhibition 61%

future science group

Type 2 17-β hydroxysteroid dehydrogenase as a novel target for the treatment of osteoporosis 

Review

Figure 4. Structures and activities of the synthetic steroid agents (see facing page). (A) Structures and activities of spiro-γ-lactone steroid derivatives. (B) Structures of nonsteroidal spiro-γ-lactone inhibitors. (C) Inhibition percentage of 17-β HSD types 1 and 2 at 2 μM of the inhibitors. The same concentration of compound FEST7 inhibits 67% of 17-β HSD5.

phobic groups on the lactone ring or introduction of an unsaturation in the lactone ring. However, all of these molecules had lower activities than the compounds containing a 6-membered lactone ring [107] . Although most of the spiro-γ-lactones were selective for 17-β HSD2 versus 17-β HSD1, these inhibitors might have biological estrogenic and/or androgenic effects on their respective receptors because of their structural similarities to sex hormones. However, some spiro-γ-lactones demonstrated antiestrogenic effects on estrogen and androgen receptors [108] . These facts led the medicinal chemists to look for other nonsteroidal inhibitors. Xu et al. has exploited the fact that E2 as well as the steroid spiro-γ-lactone (SP2) had a distance between the two oxygen atoms of approximately 10 Å separated by a hydrophobic scaffold (Supplementary Figure 2) . They designed compounds (NS1–8) which have the same features as SP2 but with phenyldihydroindene and pheyltetrahydronaphtalene (Figure 4B) . All the synthesized compounds were shown to have activity lower than steroid spiro-γ-lactone [109] . Fluorine-substituted estrogen derivatives

As E2 is a substrate for some of 17-β HSDs and a product for the others, it was a logical starting compound from which several inhibitors were designed and synthesized. 17-OH was replaced by a fluorine atom which is an electrowithdrawing atom (FEST1–8). Indeed it is admitted that fluorine may efficiently replace an OH moiety since F is as an electron-rich atom such as O and makes dipolar interactions and it brings also metabolic stability to the new molecules [110] . These compounds were evaluated on 17-β HSD type 1, 2, 4, 5 and 7. All of the compounds derived from 17-fluoroestrogens substituted by hydroxyl and fluorine atom on rings A and D, respectively, had lower activities on 17-β HSD type 1 and 2 (no selectivity observed). They did not show significant inhibitory effects on 17-β HSD type 4, 5 and 7 [111] . Biological results of these compounds (FEST1–10) on 17-β HSD types 1 and 2 (Figure 4C) demonstrated that two fluorine atoms on position 17 led to low inhibition on both 17-β HSD1 and 2. The compounds FEST9 and FEST10 had the best potency on type 2. In these compounds, there is a double bond between C8-C9. This double bond increases the planarity of the molecule. These two compounds are substituted on position 17 by a difluoromethylene group. It has been proposed that the second fluorine atom fills an additional hydrophobic pocket [111] .

future science group

Disubstituted cis-pyrrolidinone

A random screening to look for 17-β HSD isoenzymes resulted in the discovery of a new nonsteroidal inhibitor derived from 4,5-disubstituted cis-pyrrolidinone (PON) (Figure 5)  [112] . The obtained hit PON has a structure very similar to that of clausenamide the natural compound that was extracted from Clausena lansium leaves. PON was used as a lead compound to be the starting point for the development of nonsteroidal 17-β HSD2 inhibitors [112] . The structure of PON was divided into three essential groups (Figure 5) : the aromatic group A, the pyrrolidinone ring B and the aromatic group C. Modifications took place on each group to study the SAR of this series of the compounds. Only compounds having C4-C5 with cis configuration are active inhibitors. Halogens, alkyl groups, amino groups, hydroxyl moieties, aromatic rings and alkoxy groups were placed on phenyl A. The biological tests indicated that adding electron withdrawing groups on position 3 (but not elsewhere) increases the activity. The compound with a phenyl ring on position 3 of the ring A had the best activity among this subgroup (PON9) (Figure 5) . The results also demonstrated that the replacement of the aromatic ring A by aliphatic group dramatically decreases the activity while the replacement of the phenyl by heteroaromatic cycles increases the activity. The compounds with heterocycle of sulfur atom had the best activity (especially the thienyl [PON1–6]) whereas the nitrogen heterocycles caused the loss of the activity. The introduction of halogens or other electron withdrawing group on the thienyl moiety decreased the activity. On the other hand compounds with a thienyl group substituted with a phenyl ring had the best activity among this subgroup (PON2, IC50 = 10 nM) and a bridge of SO2 between thienyl group and phenyl ring slightly decreases the activity (PON4, IC50 = 70 nM and PON5, IC50 = 70 nM) (Figure 5) [113] . Some modifications were done on the pyrrolidinone ring C. Reducing the ketone group or putting a sulfur atom instead of an oxygen atom causes a loss of activity. Any group (except methyl) on the nitrogen decreased the activity whereas the nonsubstituted compounds were devoid of inhibition potency. Ketone or bulky group instead led to inactive compounds, while introducing ether or ester simply decreased the activity. Removing the hydroxyl also caused a decrease in the potency [113] . Then substitutions on ring B were investigated. All the compounds substituted on ring B at para or meta

www.future-science.com

1439

Review  Soubhye, Chikh Alard, van Antwerpen & Dufrasne

C

OH

N

O

A

N

O

OH

HO B Clausenamide Not active

PON IC50 = 5 µM Modification of the aromatic group A and/or B

O

N

OH

Code

R1

R2

IC50 (µM)

PON1

H

H

0.40

2-F

0.01

2-F

0.06

2-F

0.07

H

0.07

H

0.05

PON2 S R1

PON3

F

R2

O PON4

S O O

PON5

O

S N O N

PON6 Introduction of substituents on the aromatic ring A

O

N

Code

R1

R2

IC50 (µM)

PON7

Cl

H

0.70

PON8

Cl

Cl

0.44

H

0.42

OH R1 R2

PON9

Figure 5. Structure–activity relationships of 4,5-disubstituted cis-pyrrolidinone derivatives.

positions had lower activities than molecules without substitutions [113] . Substitutions at the ortho position by small groups improved the potency and fluorine atom gave the best compounds (PON2) (Figure 5)  [98] . It is noteworthy that all of the compounds derived from pyrrolidinone were shown to be reversible inhibitors against 17-β HSD2. It has been demonstrated that Key term Pharmacomodulation: Changing the chemical structure of pharmacologically active molecules by introducing a favorable effect assumed substituents, removing adverse effect groups to improve existing activities or bring up new properties.

1440

Future Med. Chem. (2015) 7(11)

these compounds cannot bind to estrogen and androgen receptors indicating that these inhibitors are selective for 17-β HSDs. Although some inhibitors showed very good activity on the isolated enzyme in vitro, some of these compounds have lower activities in the cellular assay indicating that cell membranes penetration was difficult [98] . Bis(aromatic ring) heterocycle derivatives Class I: di(hydroxyl aromatic ring) heterocycle

Bis(hydroxyphenyl) heterocycles (Figure 6A) has been initially designed as 17-β HSD1 inhibitors. Their selectivity toward 17-β HSD2 has been studied and some compounds were found to have an activity

future science group

Type 2 17-β hydroxysteroid dehydrogenase as a novel target for the treatment of osteoporosis 

against 17-β HSD2 [114] . Pharmacomodulation of this series of compounds demonstrated that the nature of the heterocycle located between the two phenyl moieties was critical. The heterocycles acting as A

Review

H-bond donator such as 2,4-disubstituted imidazole and 3,5-disubstituted pyrazole gave compounds without activity while the heterocycles acting as H-bond acceptors such as oxazole and 1,4-disubstituted imid-

B Heterocycle C, N, S, O

A

Arom -OH atic rin g on p o 3 or sition 4

ing atic r Arom position on -OH or 4 3

H N

R N

N N Ar

N

Ar N

Ar

Ar

Ar

IC50 >10 µM

IC50 >10 µM N Ar

S

H N N Ar

N

Ar

Not active

IC50 >10 µM N N N

HO

OH BHH3

BHH2 17β HSD2 IC50 = 0.27 µM 17β HSD1 IC50 = 1.61 µM CF3

S

Ar

O

OH

BHH1 17β HSD2 IC50 = 0.18 µM 17β HSD1 IC50 = 0.11 µM HO

Ar

N O

OH

17βHSD2 IC50 = 7.28 µM 17βHSD1 IC50 = 0.84 µM

O

HO

N

HO

N

OH

Ar

IC50