Novel type 1 5a-reductase Inhibitors with Antiproliferative Potential on LNCaP cells

This study demonstrated the antiproliferative potential of several dehydroepiandrosterone derivatives 2a-b, 3a-f, and 4a-f in LNCaP cells. LNCaP cells were cultured in the presence of the vehicle, dihydrotestosterone, or testosterone plus 2a-b, 3a-f, 4a-f, or finasteride. At 24 h the 3-(4,5-dimethylthiazol-2-yl)-2,5-di phenyltetrazolium bromide-test was performed to determine cell proliferation. In addition, the kinetic of SRD5A1 in these cells was studied in the presence or absence of different concentrations of 2a. Testosterone and dihydrotestosterone increased the proliferation of LNCaP compared to the vehicle-treated cells. Furthermore, steroids 2a-b, 3a-f, and 4a-f decreased the number of viable cells compared to testosterone treatment. However, finasteride did not affect viability. LNCaP cells converted radiolabeled testosterone into dihydrotestosterone. This conversion was inhibited by high concentrations of 2a, while at a pM concentration, the conversion increased slightly, suggesting the presence of allosteric sites in SRD5A1. In conclusion, the three series of derivatives of dehydroepiandrosterone significantly decreased the number of viable LNCaP cells, therefore, showing therapeutic potential to treat metastatic prostate cancer.


INTRODUCTION
While countless research groups and pharmaceutical companies have focused their knowledge and effort on early prostate cancer (PC) detection to prevent its spread, most patients with scattered metastases have a poor prognosis at the end of their diagnosis. Worldwide, out of the 21% of men diagnosed with prostate cancer, 16.5% die from metastasis. (Global Cancer Observatory*).
Previous reports have shown that metastatic bone tumor growth can start from malignant migratory cells produced in the prostate. This new tumor can establish itself in the bones [1][2][3][4] and begin synthesizing androgens to continue developing. 5 Adaptive androgen signaling in tumor cells may occur because intratumoral androgen biosynthesis or also to the somatic androgen receptor (AR) mutations are carried out. [6][7] It could also be to the AR gene amplification 8 or changes in nuclear coregulators, which make AR constitutively active. [9][10][11] The 5 a-reductase enzyme (SRD5A) catalyzes the conversion of testosterone (T) to its most active metabolite, which is dihydrotestosterone (DHT, Fig. 1) in androgen-dependent tissues. 12 DHT binds to its cytoplasmic receptor, which shows greater affinity than T. Then the formed complex translocates to the cell nucleus, which binds to the DNA to encode gene expression [13][14] .
incubated with 1 nM T. In addition, they accumulate up to 15 nM of T in the first hour after treatment. However, after 24 h the T forms glucuronides, and undetectable levels of T could be determined. 16 In the prostate, SRD5A1 and SRD5A2 have been well characterized. These isoenzymes possess different kinetic properties. 17,18 SRD5A2 is the predominant isoenzyme in the normal prostate and shows a higher T-binding affinity than SRD5A1. 18 While in prostatic intraepithelial neoplasms, there has greater activity of SRD5A1 instead of SRD5A2. In addition, SRD5A1 is largely overexpressed in cancers that be recurrent castration. 19 However, in benign prostatic hypertrophy, there is a decrease in SRD5A1 activity, while an increase in SRD5A2 levels has been found. 20 The fact that SRD5A1 activity is increased in castration-recurrent prostrate tumors would explain their ability to survive and grow in an environment with low androgen content. A very low concentration of T would be sufficient to induce cell proliferation and secretion of prostate antigen (PSA). 19,20 This evidence allows us to hypothesize that specific inhibitors of SRD5A1 could be used as a beneficial treatment for recurrent PC after castration and metastatic prostate cancer.
Previously, our group reported the synthesis and activity of different derivatives of dehydroepiandrosterone as specific inhibitors of SRD5A1. [20][21][22] Therefore, this research has focused on demonstrating that the particular inhibitors of SRD5A1 shown in Table 1 decrease the proliferative activity of LNCaP cells. 12,19,24 Three SRD5A isoenzymes (1,2, and 3) have been described 15,16 . SRD5As of types 1 and 3 are present in the LNCaP metastatic prostate cancer cells. These cells take up 4 to 5 pg of T/mg of tissue when

Determination of the effect of 2a-b, 3a-f and 4a-f on the proliferation of LNCaP cells
LNCaP cells (passage 25-50) were used at 4X10 4 cells/well density in 96 well container plates. The cells were allowed to adhere to the plaque for 48 h in a Roswell Park Memorial Institute medium (RPMI-1640, phenol-free). The medium containing a phenol-free 10% fetal bovine serum (Glibco, Mexico) and 2 mM glutamine was supplemented. The plates were incubated at 37°C in the humidified atmosphere, with 5% CO 2 -95% air in a VWR incubator (Symphony 1.4 A, VWR, CDMX).
The plate was incubated for 24 h under the same conditions described above. The proliferation of LNCaP cells was determined by the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-di phenyltetrazolium bromide) method. 25,26 The formazan crystals formed from the reaction with 12 mM MTT were dissolved in 100 μL of DMSO. The resulting absorbance was measured by spectrophotometry at 570nm. (Biotek™ Epoch™ Microplate Spectrophotometer, Thermo Fisher Scientific, USA). Finasteride was used as an inhibitor of SRD5A2 activity in the trial. The FDA has previously approved this steroid to treat benign prostatic hyperplasia.

Preparation of solubilized microsomes as a source of SRD5A1
LNCaP cells were cultured to confluence, the medium was removed, and the culture was rinsed with buffer A (sodium phosphate 20 mM, pH 6.5, sucrose 0.3 M, and dithiothreitol (DTT) 1 mM). The cell monolayer was scraped, and the cells were homogenized in a buffer volume A 1:10. Subsequently, the tube containing this mixture was centrifuged at 100,000 xg for 60 min to obtain the microsomes. All the previous procedures were performed at 0-4°C.
The tube sediment was solved in a volume of buffer B (sodium citrate 10 mM, potassium chloride 100 mM, DTT 1.0 mM, Lubrol PX 0.1%, glycerol 20, and NADPH 100 μM adjusting the pH=6.6) according to Levy's method 27 and incubated in ice for 30 minutes. The tube was then centrifuged at 100,000 xg for an hour. A supernatant aliquot was stored to determine proteins by the Bradford method, and the remainder was kept at -80°C until further trials with inhibitor 2a. The microsome solution (RS) was diluted in cold 20 mM phosphate buffer, pH 6.6, containing 1 μM NADPH, to a protein concentration of 60 μg/10 μL.

Determination of the optimal incubation time
Subsequently, we conducted incubations at different times using 2 nM T[ 3 H], 200 μM NADPH [27], and 1 mM DTT at a final volume of 0.5 mL of 20 mM sodium phosphate buffer, pH 7. 5 and a temperature of 37°C. The reaction began by adding 60 μg of protein RS as a source of SRD5A1 to determine the optimal conversion time from [ 3 H]T to [ 3 H]DHT. All the trials were prepared in advance and repeated on four separate occasions.
The resulting steroids were extracted with an equal volume of 5X dichloromethane. The solvent volume was collected in a single tube and evaporated to dryness. The extract was diluted in a mixture of 100 μL chloroform/methanol=2:1. This mixture was applied on an HPTLC plate (Keisel gel 60 F254 HPTLC plates) [Arellano et al., 2018] to separate T from DHT. Different steroid standards (T and DHT) were placed on separate lanes on both sides of this plate to identify the steroids produced. This plaque was then developed into a chloroform/acetone mixture=9:1 23 and dried at room temperature. Radioactivity in HPTLC was analyzed with a Bioscanner (AR-2000 Imaging Scanner, Washington, DC). The DHT standard was identified by spraying 8% phosphomolybdic acid in methanol on the corresponding plate rails. At the same time, the T standard was detected with a UV lamp (254 nm).

Characterization of 2a inhibition on SRD5A1 activity
Samples containing different concentrations of unlabeled T (2, 5, 7, 11, 16, 20 μM), 2 nM T[ 3 H], 200 μM NADPH [27], and in the presence or absence of 2a (0.1 nM, 1 nM, and 1 μM) in a final volume of 0. 5 mL of 20 mM sodium phosphate buffer, pH 7.5 were prepared. The reaction began by adding 30 μg of RS (the amount of protein that produces the optimal conversion to DHT, data not shown) and incubating the samples at 37 °C for 17 min (Fig. 6). As a control, a mixture identical to the previous one was prepared without the SR in the tube. All the assays were repeated on four separate occasions in triplicate.
The formed steroids (T or DHT labeled or not) were separated and identified, as explained in section 2.3.

Prostatic cytosol isolation
The Internal Committee for the Care and Use of Laboratory Animals (CICUAL) of the Autonomous Metropolitan University (UAM, CDMX) previously approved all rat procedures according to the NOM-062-ZOO 1999.
The prostates of the gonadectomized rats, 48 h before the experiment were dissected, separated, and homogenized in a volume of TEMD buffer (Tris-HCl 40 mM, EDTA 3 mM, and sodium molybdate 20 mM, dithiothreitol 0.5 mM, glycerol 10% at pH 8) plus protease inhibitors (PMSF 2 mM, 10 mg/mL antipain, leupeptin 5 mM). 28 All these procedures were carried out at 4°C on ice. The homogenates were centrifuged at 140,000 xg for 60 min, and an aliquot of the supernatant (cytosol) was aliquoted for protein determination. The cytosol was then stored at -70°C until use.

Androgen receptor (AR) competition tests
For competitive binding tests, we follow the method reported above. 29

Statistical analysis
The statistical analysis was accomplished using ANOVA One Way and Dunnett's test to compare the means with JMP version 8 software. P<0.05 value was accounted as a significant difference. Figure 2 shows the number of viable LNCaP cells in different concentrations of T or DHT after 24 horus. of treatment. The results indicated that a concentration of 1 μM T increased the proliferation of LNCaP cells. However, the dose-response effect was only remarkable when the T concentration differences were between 1000 units. (see Figure 2).

Effect of T and DHT on LNCaP cell proliferation
On the other hand, 1 nM of DHT (Fig. 2) significantly increased these cells' viability compared with the vehicle-treated control (V) at 24 hours. However, higher concentrations of DHT decreased cell proliferation. These data are according to the previously publicized by Horoszewicz et al., 1983. 31 Results also indicated that after 48 and 72 h of the T or DHT treatment, there were no significant differences in the number of viable LNCaP cells compared to vehicle-treated cells (data not shown). These facts could be explained because both steroids undergo glucuronidation reactions inside the LNCaP cells after 24 hours. 16,32,33 Due to these results, the following experiments were performed for 24 h in the presence of 1 μM T and 1 nM DHT. However, FIN 15,16,34 , in combination with T, showed no effect vs. T (P>0.05) on LNCaP cell proliferation, Fig. 3. Lazier et al., previously reported that FIN did not induce the death of LNCaP cells. In contrast, dutasteride, a dual inhibitor of SRD5A1 and SRD5A2 activity, resulted in significant death of these cells at a concentration of 50 μM. 35 In addition, the LNCaP cell line does not express SRD5A2 16 , which explains the behavior of FIN in LNCaP cells in our experiments.

Antiproliferative activity of 3a-f derivatives in LNCaP cells
The antiproliferative activity of 3a-f steroids in combination with T is shown in Fig. 4. Except for steroid 3b, all other series 3-compounds decreased the number of viable cells vs. T-treatment. The 3c derivative showed high dose-dependent activity, and 3d showed increased activity at all studied concentrations. However, 3a and 3f only showed activity at 0.1 nM and 1 nM vs. T concentrations.

Antiproliferative activity of 4a-f derivatives in LNCaP cells
The proliferative activity of LNCaP cells treated with T or T plus 4 a-f ±standard error is shown in Fig. 5. All products of series 4 decreased the number of viable cells, being 4f the one that showed the greatest potency and efficiency.
The optimal activity time of SRD5A1, obtained from the RS of LNCaP cells, was determined from the graph in Fig. 6 and was 17 minutes. Contrary to expectations, the same amount of DHT was formed with or without the inhibitor (368 ng/mg protein/h). These data suggest that, at the 0.1 nM concentration, 2a acted as a positive cooperator for SRD5A1. The existence of heterotropic positive allosteric cooperators has been previously reported for cytochrome P-450. [36][37][38] Harlow et al., 1997, demonstrated that progesterone underwent a hydroxylation reaction in a range of a-NF concentrations between 0-150 μM. While at higher concentrations of a-NF, the response did not occur. These data indicated that the group enzymes of the P-450 family have allosteric sites that can act as positive or negative cooperators depending on the concentration of the substrate.  In this sense, our group demonstrated 39 that the activity of SRD5A2 (obtained from prostate nuclear membrane fractions) increased in the presence of its reaction products, such as DHT or 5-androstanedione. These results evidenced the presence of allosteric sites in this enzyme, which can act as positive cooperators.

Competition of series 1,2 and 3 derivatives for androgen receptor (AR) binding sites
IC 50 values for series 1,2, and 3 derivatives were determined according to the plots of concentration vs. percent binding of AR, as exemplified in Fig. 8

CONCLUSION
This study showed that most of the derivatives of dehydroepiandrosterone (2a-b, 3a-f, and 4a-f) decreased the number of viable LNCaP cells in the presence of T. While T and DHT increased the proliferation of these cells. However, FIN failed to decrease the number of viable cells under experimental conditions. 2a-b, 3a-f and 4a-f were previously identified by our group as SRD5A1 inhibitors. 22,23 Of the three studied series, the 4a-f derivatives showed the highest efficacy and potency in inhibiting the proliferation of LNCaP cells. These steroids did not bind to AR present in rat prostate cytosol.
The microsomal fraction of LNCaP cells could convert T into DHT in the presence of NADPH at 17 min of incubation at 37°C, indicating the activity of SRD5A1. 27 In addition, 2a at concentrations of 1 nM and 1 μM blocked the conversion of T to DHT. However, in the presence of 0.1 nM of 2a, DHT production was the same as in the absence of the inhibitor. This result suggests the allosteric nature of SRD5A1. 38 Previous pharmacological studies carried out by our group in hamsters showed that the derivatives of the three studied series did not cause toxicity during the six days of treatment. 22,23 These steroids could potentially ameliorate metastatic prostate tumors.

ACkNoWleDgMeNT
This research received no external funding.

Conflict of interest
The authors declare no conflict of interest The results indicated that 4a-f did not bind to AR in rat prostate cytosol. However, it is necessary to consider that the AR of LNCaP cells displays a mutation in the ligand-binding domain, known as T877A. This mutation could affect the binding affinity of several of the hormonal or antihormonal compounds to AR. 40,41