In Vitro Biotransformation, Safety, and Chemopreventive Action of Novel 8-Methoxy-Purine-2,6-Dione Derivatives

Metabolic stability, mutagenicity, antimutagenicity, and the ability to scavenge free radicals of four novel 8-methoxy-purine-2,6-dione derivatives (compounds 1–4) demonstrating analgesic and anti-inflammatory properties were determined. Metabolic stability was evaluated in Cunninghamella and microsomal models, mutagenic and antimutagenic properties were assessed using the Ames and the Vibrio harveyi tests, and free radical scavenging activity was evaluated with 2,2-diphenyl-1-picrylhydrazyl radical scavenging assay. In the Cunninghamella model, compound 2 did not undergo any biotransformation; whereas 3 and 4 showed less metabolic stability: 1–9 and 53–88% of the parental compound, respectively, underwent biotransformation reactions in different Cunninghamella strains. The metabolites detected after the biotransformation of 3 and 4 were aromatic hydroxylation and N-dealkylation products. On the other hand, the N-dealkylation product was the only metabolite formed in microsome assay. Additionally, these derivatives do not possess mutagenic potential in microbiological models (Vibrio harveyi and Salmonella typhimurium) considered. Moreover, all compounds showed a strong chemopreventive activity in the modified Vibrio harveyi strains BB7X and BB7M. However, radical scavenging activity was not the mechanism which explained the observed chemopreventive activity.


Introduction
During the biotransformation processes, drugs and chemicals are structurally modified by various enzymatic systems to form more polar substances, which can be excreted more easily than the original compounds. Problems arise when these modifications generate toxic products [1,2]. Traditionally, drug metabolism studies use in vivo experiments in mice, rat or guinea pig, or chimeric mouse models with transplanted human hepatocytes [3,4]. However, these models can create ethical dilemmas; and the experiments are expensive and time consuming [3]. In addition, metabolites are sometimes produced in low amounts, thus hindering their identification [4].
Enzymatic systems involved in the metabolism of exogenous organic compounds are similar in mammals and in certain fungal microorganisms, like Cunninghamella. Therefore, Cunninghamella fungi can be used as an alternative to in vivo metabolism models [1][2][3][5][6][7][8]. The use of these microorganisms enables the reduction of research costs and does not arise ethical dilemmas. These alternative tests are simple and reliable as metabolites can be easily extracted to be identified later using instrumental analytical techniques [3]. Cunninghamella echinulata contains a rich set of microsomal cytochrome P450, and it can conduct phase I (oxidative) and phase II (conjugative) biotransformation under certain conditions. Moreover, the fungus possesses the ability to metabolize a wide variety of xenobiotics in regio-and stereo-selective manners [9].
A second alternative to in vivo metabolism assays of chemical compounds are in vitro screening assays with liver microsomes: they can predict the metabolic routes of tested compounds, as well as kinetics of these transformations [10]. Cytochrome P450 enzymes are involved in phase I reactions, whereas phase II reactions are commonly carried out by different transferase enzymes such as glutathione transferase and N-acetyltransferase [11]. As microsomal assays have a good reproducibility and performing them is straightforward, they are becoming a standard in vitro screening in drug discovery processes [12][13][14]. Finally, liver microsomes from different species (such as rat, mouse, dog, monkey, and human) can be used to provide more accurate data [11].
Zygmunt et al. [15,16] reported in previous studies that some novel derivatives of 8methoxy-purine-2,6-diones substituted at 7-position of purine core by carboxylic, ester, or amide moieties showed potent anti-inflammatory and analgesic activity, superior to the reference drug, in preliminary pharmacological studies [15,16]. Their analgesic properties were evaluated in two pharmacological in vivo models: formalin and writhing syndrome tests. On the other hand, anti-inflammatory potential of the aforesaid compounds was determined using zymosan-induced peritonitis and carrageenan-induced edema models. Derivatives containing an amide substituent showed very strong analgesic and anti-inflammatory (antiedematous) activities. The most active compound (4) showed a 23-fold and a 36-fold higher activity than acetylsalicylic acid (the analgesic used in clinical practice considered as reference) in the writhing syndrome test in mice and in the formalin test in mice, respectively [15,16]. Thus, these newly synthesized derivatives represent a new class of analgesic and antiinflammatory agents with possible future therapeutic applications in the treatment of inflammatory diseases and in the attenuation of pain. However, once verified their promising activity, more in-depth studies need to be performed in their pharmacology, pharmacokinetics, safety, and toxicology, among other fields, to evaluate the feasibility of the possible clinical applications of these compounds, as well as to know better how they exert their action.

Tested Compounds
Four 8-methoxy-purine-2,6-dione derivatives (1-4) described in previous publications [15,16] were provided by Dr. Grażyna Chłoń-Rzepa from the Department of Medicinal Chemistry, Jagiellonian University Medical College. The structures of these compounds were established on the basis of CHN elemental analysis and spectral data (IR, 1 HNMR, and mass spectra). Additionally, qualitative analysis was performed using TLC and HPLC.

Cunninghamella Strains and Culture Conditions
Three different strains of Cunninghamella were used: Cunninghamella echinulata NRRL 1384, Cunninghamella blakesleeana 1908 DMS, and Cunninghamella elegans 1906 DMS. The first was provided by Dr. A.J. Carnell (University of Liverpool, UK), whereas the two remaining were purchased from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (DMSZ, Braunschweig, Germany).
Cunninghamella strains were cultivated on solid potato dextrose agar (PDA) plates at 28°C. After 4 days, a liquid fermentation basal medium was prepared by the addition of 6 g of D-glucose, 1.5 g of NaCl, 1.5 g of yeast extract from Saccharomyces cerevisiae, and 1.5 g of K 2 HPO 4 to 300 mL of distilled water. Aliquots of 24.5 mL of medium were inoculated with 300 μL of suspension of fungal spores obtained after wetting colonies in PDA. The fungi were incubated with shaking until the formation of spherical clumps.

Cunninghamella Biotransformation Assay
Compounds 2-4 were evaluated in this experiment. Once microorganisms formed spherical clumps in the preliminary incubation, the corresponding tested compound was added to the broth and the mixture was incubated for 7 days. Two control experiments were also performed, in the absence of compound and fungus, respectively [3,9]. Biotransformation progress was monitored by TLC and by liquid chromatography electrospray ionization-tandem mass spectrometry (LC-MS/MS). Monitoring was performed at four different incubation times: 30 min, 3 days, 5 days, and 7 days.
To analyze metabolites in LC-MS/MS, an aliquot of 500 μL of filtered sample was taken and extracted with 500 μL of dichloromethane. The organic fraction was dried with sodium sulfate, filtered, and evaporated. The obtained residue was analyzed in LC-MS/MS. At the end of experiment, all the remaining medium was extracted with 25 mL of dichloromethane after separating the fungal clumps by filtration over cellulose filters [3,17].
The final protein concentration in the assay was 0.4 mg/mL. Compounds 1 and 3 were tested at a final concentration of 20 μM.
The amount of a parental compound remaining in a solution after biotransformation was monitored using LC-MS/MS as described previously in BCunninghamella Biotransformation Assay.^Graphs representing ln of percentage of parent compound remaining versus incubation time were drawn to calculate in vitro half-time (t 1/2 ) from the slope of linear regression. To obtain intrinsic clearance (Cl int ) of a tested compound in rat liver microsomal assay, t 1/2 previously calculated was substituted at the equation Cl int = [(V mic /P mic ) × 0.693/t 1/2 ], being V mic and P mic the volume of incubation in microliters and the amount of incubated protein in milligrams, respectively [23,24].

Analytical Methods
TLC experiments were performed using 0.2-mm silica-coated aluminum plates, being a 95:5 dichloromethane/methanol mixture the mobile phase. The plates were observed under UV light. Agilent 1100 HPLC system (Agilent Technologies, Waldbronn, Germany) equipped with a Xbridge™ C18 analytical column (2.1 × 30 mm, 3.5 μM; Waters, Dublin, Ireland) was used in the chromatographic separation of LC-MS/MS analysis. An elution gradient of acetonitrile and water, with 0.1% of formic acid, was used as a mobile phase. MS spectra were taken at Applied Biosystems MDS Sciex API 200 triple quadrupole mass analyzer (Concord, Ontario, Canada) with an electrospray ionization (ESI) interface between HPLC and MS.

Bacterial Strains
Four Vibrio harveyi strains were used in this study: a wild-type BB7 (naturally found in the Baltic Sea) and three genetically modified strains, namely BB7M (obtained by the introduction of pAB1273 plasmid containing the genes mucA and mucB to enhance the error-prone DNA repair), BB7X (hypersensitive to UV radiation after genetic modification with Tn5TpMCS), and BB7XM (containing the two abovementioned modifications) [25][26][27][28]. These strains were kindly provided by Prof. G. Węgrzyn (Department of Molecular Biology, University of Gdańsk, Poland).
In the Ames test, Salmonella typhimurium TA100 strain was used. This strain has a base-pair substitution that can be reverted by mutations at GC pairs. In addition, it contains mutations at uvrB-bio and rfa genes to eliminate excision repair mechanisms and to make the bacteria more permeable to chemicals, respectively. Finally, TA100 also includes pKM101 plasmid, which enhances both chemical and UV-induced mutagenesis via an increase in the error-prone recombinational DNA repair pathway [29,30]. TA100 strain of S. typhimurium was provided by Dr. T. Nohmi (Division of Genetics and Mutagenesis, National Institute of Hygienic Sciences, Tokyo, Japan). V. harveyi and S. typhimurium bacterial cultures were maintained at 30 and 37°C, respectively.

Mutagenic Agent
A standard mutagen NQNO was used as a positive control in mutagenicity assays and as a mutagen added to all probes (except negative controls) in antimutagenicity tests. It causes point mutations at the genome as it induces G:C → A:T transitions in Vibrio harveyi strains [27,31] and in other bacteria such as Escherichia coli and S. typhimurium [32]. NQNO was initially dissolved in DMSO. Next, its working solution was prepared in sterile water, being 40 ng/mL its final concentration.

Vibrio harveyi and Ames Mutagenicity Assays
To perform the V. harveyi mutagenicity test, 10 μL of a working solution of the tested compound (100 μg/mL) were added to V. harveyi bacterial culture (OD 600 = 0.1) in NaCl-containing BOSS liquid medium, being 40 ng/mL the final concentration of the tested compound. The samples were left incubating till OD 600 increased to 0.3-0.4. Then, an inoculum containing 5 × 10 6 cells was added to solid BOSS agar plate supplemented with neomycin sulfate at a final concentration of 100 μg/mL. Samples were left incubating 48 h at 30°C, and revertant colonies were counted manually. A positive control with NQNO and two negative controls (DMSO and blank water probe) were also tested. Each experiment was performed in triplicate, and results were given in terms of mutagenic index (MI), which is the quotient between the number of revertant colonies induced in a test sample and the number of revertants in a negative control [33][34][35]. A compound is considered mutagenic if MI is above 2.0 or if the number of revertants is higher in compound samples that in NQNO ones [26,31].
Additionally, two different final concentrations of compounds (40 and 500 ng/mL) were evaluated in the Ames test to confirm the results obtained in the Vibrio harveyi assay using a second independent method. Positive (NQNO, concentration of 40 ng/mL) and negative (DMSO) controls were also tested. S. typhimurium was cultivated in a fresh medium 12 h prior to a test. To perform mutagenicity assay, 100 μL of an overnight culture (containing approx. 1 × 10 8 to 2 × 10 8 bacteria) were inoculated along with 50 μL of a tested compound in 2 mL of top agar supplemented with traces of histidine and of biotin. Then the mixture was added over the surface of GM agar plates and was left incubating 48 h at 37°C. Afterwards, the colonies were counted manually [25,28]. A compound is considered mutagenic if MI >2.0 or if compound samples contain more revertants than NQNO controls [33][34][35][36].

Vibrio harveyi Antimutagenicity Assay
Procedure followed in the V. harveyi antimutagenicity assay [34,35] was similar to the mutagenicity method described before. Differential point is that 10 μL of NQNO (final concentration of 40 ng/mL) were added to V. harveyi bacterial culture in BOSS liquid medium 15 min after the addition of tested compound to the culture. The results were expressed as the inhibition percentage of mutagenic effect, which was calculated with the formula: 100 − [(R 1 /R 2 ) × 100], where R 1 is the number of revertants per plate in the presence of both mutagen and tested substance, and R 2 is the number of revertants per plate in the presence of mutagen [34,36]. The antimutagenic effect was considered weak/absent, moderate, or strong when the inhibition percentage was up to 25%, from 25 to 40%, and higher than 40%, respectively [34,37]. S. typhimurium antimutagenicity assay was based on similar assumptions as the V. harveyi antimutagenicity test. In this case, the chemopreventive activity was determined at two concentrations of a tested compounds and a standard mutagen, i.e., 40 and 500 ng/mL.

DPPH Assay
The capacity of selected 8-methoxy-purine-2,6-dione derivatives to scavenge stable free radical DPPH was assessed [22,[38][39][40][41][42]. A blank and a triplicate of 10 different concentrations of each compound or of each positive control (gallic and ascorbic acids) were evaluated in the experiment. Aforesaid concentrations of compounds were obtained by dilution from a 1000 μM stock solution in methanol. One hundred microliters of DPPH stock solution were added to each well; later, 100 μL of the corresponding compound dilution were added. Consequently, the final concentration of DPPH radical was 250 μM, whereas the final concentrations of the tested compounds or positive controls were 500, 375, 250, 125, 62.5, 50, 37.5, 25, 12.5, and 6.25 μM. Afterwards, the 96-plate was wrapped with aluminum and stored in the dark 30 min prior to the spectrophotometric measurement at a 517-nm wavelength. The obtained absorbances were used to calculate the percentage of radical scavenged by the action of a compound or of a positive control at each concentration using a formula 100 × (A 0 − A m )/A 0 . In this equation A m is the average absorbance of the three triplicates measured at the corresponding concentration and A 0 is the average absorbance of DPPH radical in the absence of the tested substance. Knowing inhibition data at the different concentrations, IC 50 was calculated when applicable [40].

Cunninghamella Biotransformation
The metabolic degradation of compounds 2-4 was studied in the Cunninghamella biotransformation assay, being these derivatives tested in three different Cunninghamella species: C. echinulata, C. blakesleeana, and C. elegans. Compound 1 was not evaluated within this experiment due to its structural similarity to compound 2. Metabolites were determined through the study of the corresponding LC-MS/MS spectra. Unless stated otherwise, the metabolites discussed onwards are the ones found at the end of experiment after 7-day incubation of investigated purine-2-6-dione derivatives with the corresponding Cunninghamella strain.
None of the three Cunninghamella strains evaluated generated detectable metabolites of 2 in a biotransformation assay, according to the LC-MS/MS spectra recorded. The main peak observed in LC-MS/MS corresponded with protonated molecule [M + H] + of 2 (m/z 283) and was eluted in LC at a retention time of 3.52 min.
In case of 3, a protonated molecule [M + H] + (m/z 386) of a parental compound (Fig. 2) remained to be the most important derivative ion present in the extract, as only a 0.6, 4.6, and 7.6% of a tested compound was metabolized at the end of experiment in C. elegans, C. blakesleeana, and C. echinulata, respectively ( Table 1). The main metabolite (3-M4) in three strains corresponded to the product of aromatic hydroxylation (Fig. 3) of a parental compound, and was eluted at a retention time of 3.62 min (m/z 402). The percentages of the metabolite observed were 0.6, 4.6, and 4.2% in C. elegans, C. blakesleeana, and C. echinulata, respectively. No more metabolites were detected in the first two strains. C. echinulata generated three additional metabolites in the biotransformation assay. The first of these three metabolites (3-M1) had a retention time of 2.75 min, an abundance of 1.5%, and an m/z of 296. On the other hand, the main peak in MS of the two remaining metabolites had a molecular weight of 402 (Fig. 3). The percentages of appearance of these two metabolites and their retention times were 3.79 min and 0.3% for the first metabolite (3-M3) and 4.29 min and 1.6% for the second (3-M2) (Table 1, Figs. 3 and 4).
A protonated molecule [M + H] at m/z 441 was detected in mass spectra of four after the biotransformation processes, corresponding to a molecular weight (m/z 440) of a parental compound. In all Cunninghamella strains, only one metabolite (4-M1), the product of aromatic hydroxylation, was observed, at a retention time of 3.37 min and with m/z 457. The calculated percentages of the parental compound metabolized in each case were 53.0, 66.6, and 87.6% in C. elegans, C. blakesleeana, and C. echinulata, respectively (Table 1, Fig. 4).

Microsomal Biotransformation
The metabolic stability of compounds 1 and 3 was measured by the incubation with rat liver microsomes and subsequent monitoring of the metabolites formed by LC-MS/MS. Compound 2 was not examined with microsomes due to its metabolic stability in a former Cunninghamella assay. During the microsomal biotransformation, compound 1 was metabolically stable as no metabolites were observed after 1 h of its incubation with microsomes: only a protonated paternal compound was detected in LC-MS/MS spectra.
On the other hand, only one metabolite (3-M1) of compound 3 ( Table 2) was observed after rat microsomal incubation, being its molecular weight and its retention time 296 g/mol and 2.78 min, respectively. The metabolite found was the product of the N-dealkylation of a   (Fig. 4). Calculated in vitro half-time (t 1/2 ) for 3 was 17.7 min, and intrinsic clearance (Cl int ) was 97.88 μL/min/mg protein (Fig. 5).

Mutagenicity
The results of mutagenicity assay in four Vibrio harveyi strains are shown in Table 3. None of the evaluated four 8-methoxy-purine-2,6-dione derivatives displayed mutagenicity at the concentration of 40 ng/mL, as MI of four investigated compounds was always below 2.0 in  (Table 3) using the Salmonella typhimurium TA100 strain.

Antimutagenicity
According to data obtained in the antimutagenicity assay (Table 4), four 8-methoxy-purine-2,6-dione derivatives inhibited the mutagenic action of a standard mutagen NQNO in four Vibrio harveyi strains evaluated. All derivatives exhibited a very strong chemopreventive activity in two modified V. harveyi strains (BB7X and BB7M). The antimutagenic percentages calculated ranged from 56 to 74%, and from 39 to 54% in BB7X and BB7M strains, respectively. In contrast, 1 (20%), 2 (38%), and 3 (29%) exerted a moderate or a close to moderate chemopreventive effect in a wild BB7 strain, whereas 4 (53%) was a strong chemopreventive agent in the same strain. Finally, chemopreventive activity exerted by the evaluated 8-methoxy-purine-2,6-dione derivatives was much lower in the BB7XM strain, in which chemopreventive percentages ranged from 9 to 15%. Moreover, all derivatives showed a very potent antimutagenic activity in S. typhimurium TA100 strain. The chemopreventive percentages ranged from 88 to 96% and from 88 to 95% for concentrations of 40 and of 500 ng/mL, respectively (Table 4).    Mean values from experiments ± standard deviation (SD) are presented (with inhibition of NQNO mutagenicity indicated in parentheses). Number of colonies is the average number of revertants per plate CTR1 control experiment without mutagen, CTR2 control experiment without mutagen, but with DMSO, NQNO experiment with 4-nitroquinoline N-oxide as a single compound.
scavenging activity, as they showed a low or non-existent capacity to quench DPPH radical in DPPH assay performed. Thus, further research should be performed in the future to ascertain the exact mechanism responsible for this chemopreventive activity [43,44].
To sum up, within the study, metabolic profile and selected biological properties of some novel 8-methoxy-purine-2,6-dione derivatives (compounds 1-4) with analgesic and antiinflammatory properties were evaluated. It was demonstrated that in the Cunninghamella model, compound 2 did not undergo any biotransformation; whereas 3 and 4 showed less metabolic stability as they underwent biotransformation reactions. Moreover, the new derivatives of purine-2,6-dione do not possess mutagenic potential in the microbiological models considered and showed a strong chemopreventive activity in the modified Vibrio harveyi strains BB7X and BB7M.
Therefore, the obtained results might represent an important step in designing and planning future studies with purinediones.