Bio-Catalytic Structural Transformation of Anti-cancer Steroid, Drostanolone Enanthate with Cephalosporium aphidicola and Fusarium lini, and Cytotoxic Potential Evaluation of Its Metabolites against Certain Cancer Cell Lines

In search of selective and effective anti-cancer agents, eight metabolites of anti-cancer steroid, drostanolone enanthate (1), were synthesized via microbial biotransformation. Enzymes such as reductase, oxidase, dehydrogenase, and hydrolase from Cephalosporium aphidicola, and Fusarium lini were likely involved in the biotransformation of 1 into new metabolites at pH 7.0 and 26°C, yielding five new metabolites, 2α-methyl-3α,14α,17β-trihydroxy-5α-androstane (2), 2α-methyl-7α-hydroxy-5α-androstan-3,17-dione (3), 2-methylandrosta-11α-hydroxy-1, 4-diene-3,17-dione (6), 2-methylandrosta-14α-hydroxy-1,4-diene-3,17-dione (7), and 2-methyl-5α-androsta-7α-hydroxy-1-ene-3,17-dione (8), along with three known metabolites, 2α-methyl-3α,17β-dihydroxy-5α-androstane (4), 2-methylandrosta-1, 4-diene-3,17-dione (5), and 2α-methyl-5α-androsta-17β-hydroxy-3-one (9), on the basis of NMR, and HREI-MS data, and single-crystal X-ray diffraction techniques. Interestingly, C. aphidicola and F. lini were able to catalyze hydroxylation only at alpha positions of 1. Compounds 1–9 showed a varying degree of cytotoxicity against HeLa (human cervical carcinoma), PC3 (human prostate carcinoma), H460 (human lung cancer), and HCT116 (human colon cancer) cancer cell lines. Interestingly, metabolites 4 (IC50 = 49.5 ± 2.2 μM), 5 (IC50 = 39.8 ± 1.5 μM), 6 (IC50 = 40.7 ± 0.9 μM), 7 (IC50 = 43.9 ± 2.4 μM), 8 (IC50 = 19.6 ± 1.4 μM), and 9 (IC50 = 25.1 ± 1.6 μM) were found to be more active against HeLa cancer cell line than the substrate 1 (IC50 = 54.7 ± 1.6 μM). Similarly, metabolites 2 (IC50 = 84.6 ± 6.4 μM), 3 (IC50 = 68.1 ± 1.2 μM), 4 (IC50 = 60.4 ± 0.9 μM), 5 (IC50 = 84.0 ± 3.1 μM), 6 (IC50 = 58.4 ± 1.6 μM), 7 (IC50 = 59.1 ± 2.6 μM), 8 (IC50 = 51.8 ± 3.4 μM), and 9 (IC50 = 57.8 ± 3.2 μM) were identified as more active against PC-3 cancer cell line than the substrate 1 (IC50 = 96.2 ± 3.0 μM). Metabolite 9 (IC50 = 2.8 ± 0.2 μM) also showed potent anticancer activity against HCT116 cancer cell line than the substrate 1 (IC50 = 3.1 ± 3.2 μM). In addition, compounds 1–7 showed no cytotoxicity against 3T3 normal cell line, while compounds 8 (IC50 = 74.6 ± 3.7 μM), and 9 (IC50 = 62.1 ± 1.2 μM) were found to be weakly cytotoxic.


INTRODUCTION
Microbial transformation is one of the most important approaches for the structural transformation of various classes of organic compounds. This technique has been successfully employed in green chemistry, i.e. the drug discovery and development programs, providing an excellent source of compounds around core structures, followed by screening for various biological activities. In several instances, microbial transformation has been used as an important tool for the regio-, chemo-, and stereo-selective conversions of organic compounds which are difficult to achieve by conventional methods (Holland and Weber, 2000;Fernandes et al., 2003;Mihovilovic et al., 2003;Yildirim et al., 2003;Bartmanska et al., 2005;Borges et al., 2009;Choudhary et al., 2011;Ravindran et al., 2012). Due to the inactive nature of hydrocarbon skeleton of steroids, they are often difficult to be derivatized by conventional synthetic methods. Therefore, microbial transformation is often used for the structural alteration of steroids. The presence of P450 cytochrome enzyme systems in fungi, makes whole-cell biocatalysis an efficient tool for stereo-, and regio-specific hydroxylation (Choudhary et al., 2005a;Tong and Dong, 2009;Kristan and RiŽner, 2012;Baydoun et al., 2014).
Drostanolone, and its esters derivatives, such as drostanolone propionate, drostanolone pentanoate, and drostanolone enanthate (1) are anabolic-androgenic steroids (AASs) used by athletes to strengthen their muscles without gaining fat. In addition, drostanolone, and its esters derivatives have the ability to inhibit the production of estrogen. Propionate ester of drostanolone is also used for the treatment of breast cancer, under the brand name of Masteron (Chowdhury et al., 1976;Clavel et al., 1982;Marinov et al., 1986;Vardar et al., 2002;Bahrke and Yesalis, 2004).
Cancer is currently recognized as a major public health challenge. Cancers are the second leading cause of death in the United States, and all across the world. According to the World Health Organization (WHO), the prevalence of cancer is exceeding 6 million cases per year. Cancer cells have high proliferation rate. They spread rapidly in the living system, and can survive against strong chemotherapeutics and DNA damaging agents. Similarly, cytotoxic drugs have many adverse effects on normal cells, and thus their use in cancer chemotherapy is a therapeutic challenge. Due to this, development of safe, effective and selective chemotherapeutic agents is urgently needed against various cancers (Munoz-Pinedo et al., 2012;Su et al., 2015;Swadogo et al., 2015;Rebecca et al., 2016).
Breast cancer is one of the most common cancers in 140 countries of the world. It is a major cause of cancer-related death in females all across the world, characterized by the abnormal growth of cells in the breast lobules or ducts with the high proliferation rate (Hanahan and Weinberg, 2000;Ferlay et al., 2015).
Cervical cancer is the second most predominant cancer in females all across the world, after the breast cancer. The main cause of cervical cancer is the formation of malignant cells in tissues of the cervix (Wang et al., 2013;Hafiza and Latifah, 2014;Pariente et al., 2016). The HeLa cancer cell line, obtained from human cervical cancer cells, is a common cellular model to evaluate the cytotoxic potential of test compounds.
Prostate cancer is the second leading cancer in male worldwide after the bronchus cancer, and the third most common cause of cancer death. It is the most common reason of malignancy in men. Its incidence increases with age, more common over the age of 50 years (Henry and Omahony, 1999;De-Bono et al., 2010;Wenbin et al., 2015). The PC-3 cancer cell line, obtained from male prostate cancer cells, is a widely used model to investigate the toxicity of test compounds.
Lung cancer is a highly prevalent cancer among men in the United States since the mid-1950s and among women, since the late 1980s. It is a leading cause of cancer related death. Lung cancer is mostly attributed to smoking (Travis et al., 2002;Villanti et al., 2013;Mishra et al., 2016). The H460 cancer cell line, obtained from human lung cancer cells (lymphogenous metastatic subline of human large cell lung carcinoma), are often used to evaluate cytotoxicity of test compounds.
After the breast, lung, and prostate cancers, colon cancer is the fourth most prevalent cancer all across the world. It is the cancer found in approximately equal frequency in males and females. It is the second most common cause of cancer-related deaths in western countries (Levin et al., 2008;Andre et al., 2009;Ahearn et al., 2012). The HCT116 cancer cell line, obtained from human colon cancer cells, is commonly used to assess the cytotoxicity of test compounds.
In continuation of our studies on the fungal transformation of bioactive steroids (Choudhary et al., 2005b(Choudhary et al., ,c, 2007(Choudhary et al., , 2010Ahmad et al., 2014;Siddiqui et al., 2017), we synthesized analogs of drostanolone enanthate (1) through its transformation with Cephalosporium aphidicola, and Fusarium lini. Metabolites 2-9, as well as substrate 1, were evaluated against HeLa (cervical carcinoma), PC-3 (prostate carcinoma), H460 (lung cancer), HCT116 (colon cancer), and 3T3 (mouse fibroblast normal) cell lines, using high-throughput cell-based assay, the most efficient and convenient laboratory method, the MTT assay to predict the response of test compounds in malignancies where they showed specificity against the cancer cells. This study has thus identified anti-cancer metabolites of drostanolone enanthate (1) for further studies.

Instrumental Analysis
Thin layer chromatography (TLC) (silica gel, 20 × 20, 0.25 mm thick, PF 254 , Merck, Germany) was used for the analysis of degree of transformation and purity. Silica gel column chromatography (70-230 mesh, E. Merck, Germany) was used for the initial purification of metabolites. In addition, preparative reverse phase recycling (JAI LC-908W, Japan), equipped with YMC L-80 (4-5 µm, 20-50 mm i.d.), was used for the final purification of metabolites. Structures of metabolites were elucidated with the help of 1 H-(400, 500, and 600 MHz), and 13 C-NMR (100, 125, and 150 MHz) spectra, which were recorded on Bruker Avance-NMR spectrometers (France) in CD 3 OD, CD 3 COCD 3 or DMSO-d 6 . HREI-MS and IR spectra were performed on JEOL JMS-600H (Japan) (doublefocusing magnetic sector mass analyzer) mass spectrometer (EI, electron impact ionization), and Bruker Vector 22 FT-IR spectrophotometer, respectively. Optical rotations of metabolites were recorded on JASCO P-2000 polarimeter (Japan). Evolution 300 UV-visible spectrophotometer was used to record the UV spectra. The melting points of the transformed products were measured by using Buchi M-560 (Switzerland) instrument. Single-crystal X-ray diffraction data was collected on Bruker APEXII D8 Venture diffractometer, fitted with PHOTON 100 detector (CMOS Technology) and fine-focus sealed tube having X-ray source (Cu Kα radiations α = 1.54178 Å). Reflection intensities were integrated using SAINT software. Absorption correction was done on Multi-scan, and structure was solved by SHELXTL program (Gerlier and Thomasset, 1986;Sheldrick, 2008;Spek, 2009).

Microbial Cultures
Microbial culture of C. aphidicola ATCC 28300 and F. lini NRRL 2204 were obtained from the American Type Culture Collection (ATCC) and Northern Regional Research Laboratories (NRRL), respectively. Cultures were grown on Sabouraud dextrose agar (SDA) slant, and maintained at 4 • C.

Media Preparation
Four liter of media for the growth of C. aphidicola ATCC-28300 was prepared by mixing 200 g of glucose, 4 g of KH 2 PO 4 , 8 g of glycine, 4 g of KCl, 8 g of MgSO 4 .7H 2 O, and 8 mL of trace elements in 4 L of distilled water. Similarly, four liter of media was prepared for F. lini NRRL 2204 by mixing 40 g of glucose, 20 g of yeast extract, 20 g of peptone, 20 g of KH 2 PO 4 , 20 g of NaCl, and 40 mL of glycerol in 4 L of distilled water.

Fermentation of Drostanolone Enanthate
(1) with C. aphidicola ATCC 28300, and F. lini NRRL 2204 Medium for the growth of C. aphidicola and F. lini was prepared by mixing above mentioned ingredients in 4 L of distilled water for each fungus, which was then dispensed equally into 40 Erlenmeyer flasks of 250 mL (100 mL in each flask). All these flasks were autoclaved at 121 • C, and were inoculated with SDA slants of C. aphidicola ATCC 28300 and F. lini NRRL 2204 under sterilized conditions. These fungal culture containing flasks were left on the rotary shaker (121 rpm) at 26 ± 2 • C for 3-4 days to obtain maximum growth of C. aphidicola and F. lini. Compound 1 (1 g) was dissolved in 20 mL of acetone for each fungus, and was dispensed 0.5 mL in each flask, and again placed on rotary shaker for 12 days. A negative control (medium + fungal culture), as well as positive control (medium + substrate) were also prepared analogously to evaluate the fungal metabolites and degradation of compound 1, respectively. After incubation of 12 days, all the flasks were filtered and extracted with 20 L of dichloromethane (DCM), organic phase was separated, and then was evaporated under reduced pressure on rotary evaporator. This yielded 1.5 g of pale yellowish solid crude. This crude material was fractionated over silica gel column chromatography by elution with gradient hexanes and acetone solvent systems. Column chromatography yielded four main fractions (1-4), which were then purified by recycling HPLC. Metabolites 2 (t R = 16 min, 8 mg, 0.8%), 3 (t R = 19 min, 11 mg, 1.1%), 4 (t R = 17 min, 35 mg, 3.5%), 5 (t R = 23 min, 42 mg, 4.2%), 6 (t R = 20 min, 9 mg, 0.9%), 7 (t R = 21 min, 12 mg, 1.2%), and 9 (t R = 18 min, 5 mg, 0.5%) were purified from fractions 1-7, respectively, through reverse phase recycling HPLC using methanol-water as solvent system (MeOH: H 2 O; 70: 30). Whereas, metabolite 8 was obtained from fraction 8 through normal phase recycling HPLC (t R = 32 min, CHCl 3 : MeOH; 95: 5, 7.5 mg, 0.75%).
The colorimetric assay MTT [3-(4, 5-dimethyl thiazol-2yl)-2, 5-diphenyl tetrazolium bromide] was used for the evaluation of cell metabolic activity. In this assay, the yellow MTT is reduced to purple formazon in the mitochondria of the living cells (Gerlier and Thomasset, 1986). The more the living cells, the more the color and thus more absorbance is observed by colorimeter (Fesahat et al., 2015). We have used this method to analyze the effect of our transformed products on cellular cytotoxicity against cancer cells.  Danihelová et al., 2013). A broad range of epithelial carcinoma cell lines were used in this experiment where cisplatin was a common drug of choice. It targets DNA, interferes with cell division during mitosis and induces apoptosis (Eastman, 1999). IC 50 values were calculated to determine the minimum concentration required to kill 50% of the cells.
Whereas, At = Absorbance value of test compound Ab = Absorbance value of blank Ac = Absorbance value of control six metabolites 2-7 (Figure 1), while with F. lini yielded five metabolites 3-5, and 8-9 (Figure 2). Detail descriptions of new metabolites 2, 3, 6, 7, and 8 are narrated below. Metabolite 2 was isolated as white crystals. The molecular formula C 20 H 30 O 3 was based on HREI-MS which showed the [M + ] at m/z 322.2517 (calcd. 322.2508, C 20 H 34 O 3 ), suggesting the addition of three hydrogen, and an oxygen atom, reduction of C-3 ketonic carbonyl, and hydrolytic cleavage of heptanoate side chain in substrate 1. The IR absorbance at 3,531, and 3,423 cm −1 were due to presence of hydroxyl groups. The 1 H-NMR spectrum showed an additional methine proton signal at δ 3.69 (Table 1) (Data sheet 2). The signals for heptanoate side chain were also found missing in 1 H-NMR spectrum of compound 2. The 13 C-NMR spectrum showed signals for a new methine carbon at δ 71.4, and a new quaternary carbon at δ 84.8. Carbon signals for heptanoate chain and ketonic carbonyl were also found missing in the 13 C-NMR spectrum. This suggested reduction of C-3 carbonyl, hydrolytic loss of heptanoate chain, and addition of an OH at C-14 ( Table 1). The OH group was placed at C-3 on the basis of HMBC correlations of H 3 -20 (δ 0.91, d) with C-3 (δ 71.4) (Figure 3). This resulted from the reduction of C-3 carbonyl. Second OH group was placed at C-14, based on the HMBC correlation of H 3 -18 (δ 0.81, s) with C-14 (δ 84.8). The third OH was at C-17, resulted from hydrolytic loss of heptanoate moiety. The O-H group at C-3 was deduced to be α-oriented based on the NOESY correlations of H-3 (δ 3.69, br. s) with β-oriented H-2 (δ 1.64, overlap), and α-oriented H 3 -20 (δ 0.91, d) (Figure 4). The OH-14 was deduced to be α-oriented, based on NOESY correlations of H-9 (δ 1.32, m) with OH-14 (δ 2.66, s) (acetoned 6 ). Single-crystal X-ray diffraction analysis further supported the structure of metabolite 2, comprised of three rings in chair conformation (A, B, and C), and one in envelop conformation (D). Three OH groups at C-3, C-14, and C-17 were assigned, α-, α-, and β-orientation (Figure 5). Single-crystal diffraction data of metabolite 2 was submitted to Cambridge Crystallographic Data Collection (CCDC 1500706). The structure of metabolite 2 was thus deduced as 2α-methyl-3α,14α,17β-trihydroxy-5αandrostane.

Structure Elucidation
Metabolite 3 was obtained as a white solid. The HREI-MS displayed the [M + ] at m/z 318.2192 (calcd. 318.2195, C 20 H 30 O 3 ). This represented the loss of heptanoate side chain, and addition of an oxygen atom in substrate 1. IR spectrum showed absorbances for OH (3,436 cm −1 ) and ketonic carbonyls (1,711 cm −1 ). In the 1 H-NMR spectrum, signals of heptanoate chain and C-17 methine proton were found missing, whereas a new downfield methine proton signal was observed at δ 3.92 (Table 1) (Data sheet 3). This suggested hydroxylation of steroidal skeleton and oxidative cleavage of heptanoate moiety. The 13 C-NMR spectrum also supported the above inferences. Carbon signals for heptanoate moiety were found missing, whereas a new ketonic carbonyl and an oxy-methine signals appeared in 13 C-NMR spectrum ( Table 1). This indicated oxidative cleavage of heptanoate chain and hydroxylation in substrate 1. The newly appeared methine proton at δ 3.92 (H-7) showed HMBC correlations with C-5, and C-8, suggesting an OH at C-7 (Figure 3). The HMBC correlations of H 3 -18 and H 2 -16 with a carbon at δ 223.8 suggested a ketonic carbon at C-17. OH at C-7 was further supported by COSY correlations of H-7 with H 2 -6 and H-8. H-7 (δ 3.92, s) showed NOESY correlations with axially oriented H-8 (δ 1.70, overlap), which suggested an α-orientation of OH at C-7 (Figure 4). Thus, the structure of metabolite 3 was deduced as 2α-methyl-7α-hydroxy-5α-androstan-3,17-dione.
Single-crystal X-ray diffraction analyses were also carried out on compounds 1, and 5. Compound 1 was comprised of four rings i.e., A, B, C, and D with chair, chair, chair, and envelope conformations. A methyl was present at C-2 with equatorial orientation. Heptanoate ester chain was present at C-17 of ring D (Figure 6). Whereas, metabolite 5 contains rings A, B, C, and D in planer, chair, chair, and envelop conformations. Heptanoate side chain was absent in metabolite 5 (Figure 7). Single-crystal diffraction data of metabolites 1, and 5 were submitted to Cambridge Crystallographic Data Collection (CCDC 1500705, and CCDC 1500707, respectively). Singlecrystal X-ray diffraction analysis data of compounds 1, 2, and 5 is presented in Table 4. Cytotoxicity against HeLa,PC3,H640,and HCT116 Cancer,and 3T3 Normal Cell Lines Natural products,e.g.,flavonoids,steroids,terpenes,etc. can serve as an excellent source for the production of many drugs, as they are reported for many biological activities such as anti-cancer, anti-inflammatory, anti-leishmanial, anti-bacterial activities, etc. Therefore, it is necessary to derivatize them, to study structure-activity relationship which might be helpful for the synthesis of safe and effective drugs. At present, many pharmaceutical products are efficiently synthesized via microbial biotransformation due to presence of a range of enzymes in them. Based on reported anti-cancer activity of drostanolone (9), drostanolone propionate, and drostanolone heptanoate (1), compounds 1-9 were evaluated for cytotoxicity against HeLa, PC-3, H460, and HCT116 cancer cell lines. Interestingly, all compounds showed a moderate to potent anti-cancer activity. The anti-cancer activity of compounds 1-9 were inferred from the IC 50 values against HeLa (human cervical carcinoma), PC-3 (human prostate cancer), H460 (human lung cancer), and HCT116 (human colon cancer) ( Table 5). Based on these results, the cytotoxicity of these compounds was also evaluated against 3T3 (mouse fibroblast) normal cell line, where all compounds,

Structure-Activity Relationship
Variations in the structures of metabolites effected their anticancer activity. Metabolite 2 with the absence of ester moiety, and two αO-H groups at C-3, and C-14, and a βO-H group at C-17 showed a lower cytotoxicity against HeLa, H460, and HCT116 cancer cell lines, but its anti-cancer activity was higher against PC-3 cancer cell line, as compared to compound 1.
Metabolite 3 with the loss of ester group and two keto groups at C-3, and C-17, and αO-H group at C-7 also showed a lower activity against HeLa, H460, and HCT116 cancer cell lines, but increased activity against PC-3 cancer cell line in comparison to substrate 1. Reduction of C-3 carbonyl to O-H and hydrolytic cleavage of ester moiety in compound 4 increased its activity against both HeLa and PC-3 cancer cell lines to some extent, but its anti-cancer activity against H460, and HCT116 cancer cell lines decreased as compared to substrate 1. Presence of C = C between C-1/C-2 and C-4/C-5, along with C = O at C-3, and C-17 made metabolite 5 a potent anti-cancer agent against HeLa cancer cell line, as compared to substrate 1 and the standard drug (cisplatin). Moreover, its anti-cancer activity against PC-3 cancer cell line was also increased as compared to substrate 1. Interestingly, its anti-cancer activity against H460, and HCT116 cancer cell lines was lower in comparison to compound 1. Presence of C = C between C-1/C-2 and C-4/C-5, C = O at C-3, and C-17, along with αO-H group at C-11 in metabolite 6 increased its anti-cancer activity against HeLa, and PC-3 cancer cell lines in comparison to substrate 1 and showed similar anticancer potential against HeLa cancer cell line as standard drug (cisplatin). However, its anti-cancer activity against H460, and HCT116 cancer cell lines decreased in comparison to compound 1. Presence of C = C between C-1/C-2 and C-4/C-5, C = O at C-3, and C-17, along with αO-H group at C-14 in metabolite 7 again increased its anti-cancer activity against HeLa, and PC-3 cancer cell lines, and decreased its anti-cancer activity against H460, and HCT116 cell lines in comparison to substrate 1. Presence of C = C at C-1/C-2, C = O at C-3, and C-17, and αOH at C-7 made metabolite 8 the most potent anti-cancer against HeLa cancer cell line in comparison to substrate 1 and the standard drug (cisplatin). It also showed a good activity against PC-3 cancer cell line but decreased anti-cancer activity against H460, and HCT116 cancer cell lines as compared to substrate 1. Metabolite 9 with only hydrolysis at C-17 showed increased anti-cancer activity against HeLa, PC-3, and HCT116 cancer cell lines but decreased cytotoxicity against H460 cancer cell line in comparison to substrate 1 and the standard drug (cisplatin). On the basis of above discussion, it can be concluded that increased anti-cancer activity of compounds 5-8 against HeLa, and PC-3 cancer cell lines was probably due to the presence of C = C at C-1/C-2, and carbonyl ketone at C-3, and C-17 in comparison to substrate 1.

AUTHOR CONTRIBUTIONS
The concept of presented research was developed and designed by MIC. MIC was also the project supervisor. All the data were analyzed and interpreted by MIC, A-t-W, and HC. MS, SY, MA, and NF performed the experiments. MIC, A-t-W, and MS wrote the manuscript. All the authors revised the manuscript. The final version of manuscript was also approved by MIC before submission.