, , , , , , , ,

(1.National Demonstration Center for Experimental (Pharmacy) Education,Zunyi Medical University,Zunyi Guizhou 563099,China; 2.Department of Endocrinology,The Third Affiliated Hospital of Zunyi Medical University,Zunyi Guizhou 563000,China)

基础医学研究

Speciesdifferenceontestosteronemetabolisminhumansandanimals

HeYuqi1,ZengYao1,LingLei1,LuAnjing1,DuYimei1,YangTao1,LiBoda1,PengJie2,WuQing1

(1.National Demonstration Center for Experimental (Pharmacy) Education,Zunyi Medical University,Zunyi Guizhou 563099,China; 2.Department of Endocrinology,The Third Affiliated Hospital of Zunyi Medical University,Zunyi Guizhou 563000,China)

ObjectiveTo investigate the species difference of testosterone metabolism among human and 7 animal species including mice,rats,guinea pigs,rabbits,dogs,pigs,sheep,and cattle.MethodsTestosterone was incubated with liver microsomes of humans and investigated animals,together with CYP-mediated metabolic reaction factors including NADP Na,glucose-6-phosphate,and glucose-6-phosphate dehydrogenase.Incubation samples were introduced into HPLC to determinate testosterone and its metabolites.Heatmap was used to visualize the metabolic profile while principle component analysis was used to tell the difference of metabolic profile among species.High resolution mass spectrometry was used to assist the identification of metabolite structures.Results6β-hydroxylation is not the unique metabolic pathway of testosterone.Eighteen NADP-dependent metabolites were observed in humans or other investigated animals.Although 6β-hydroxylation showed positive correlation with the substrate elimination,this specific metabolic pathway could not compensate all substrate consumption.Guinea pig generates the most metabolites while sheep showed the weakest capacity on testosterone metabolism.Mice,dogs,and rabbits showed the most similar metabolic profile to humans.Conclusion6β-hydroxylation metabolites of testosterone was not applicable to evaluate the CYP 3A activity of all species.Mice,dogs,and rabbits are recommended in pre-clinical trial if the target is proved to be metabolized by human CYP 3A4.

testosterone; species difference; CYP 3A; metabolism

A right animal model is extremely important for drug discovery and development in particularly in pre-clinical trial stage.Thus,investigation of species difference has to be noted beforeinvivoexperiments.

Metabolic stability is an important factor leading failure of drug development[1-2].Weaker metabolic stability of drug candidate in human than in experimental animals,may lead lower exposure of the therapeutic target in drug molecules.Eventually,the significant drug effect observed in animals disappeared in humans.To prevent from misused animal model,understanding the metabolic profile and capacity of specific drug candidates in animals is necessary.The most effective and popular method to understand the metabolic capacities of animals is to check the activity of metabolic enzymes via specific probe substrates[3].

CYP 3A is the most important cytochrome P450 isozyme[4].This isozyme is responsible for metabolisms of around half market drugs.Testosterone is a famous CYP 3A4 substrate,usually was used to evaluate CYP 3A4 activity of individual human livers[5].Although testosterone 6β-hydroxylation is proved to be a selective reaction catalyzed by CYP 3A4 in humans[5],whether it’s metabolism in animals is the same or similar as human is still not clear.It means that comparison of CYP 3A activities between human and animals based on testosterone 6β-hydroxylation rate may not or may not fully reflect the real facts.

Regarding species difference on testosterone metabolism,comparison between humans and rodents has been discussed[6-7].However,in case of other normally used experimental animals such as rabbit,guinea pig,and dogs,species difference on testosterone metabolism is still not known.In special cases,some large animals,such as pigs,sheep,and cattle,are also used in pharmacology and pharmacokinetics study.Difference on testosterone metabolism between those animals and humans is also worth involving.

In the present study,testosterone metabolism in liver microsomes from human and various animal species were studied.This was the first time that species difference on testosterone metabolism was studied systematically.The metabolic profiles would indicate how similar was each investigated animals to human.The result may help to decide the right animal models in case the investigated drugs were potentially metabolized by CYP 3A.

1 Materials and Methods

1.1 Instruments and Materials An HPLC system (1100,Agilent Co.Ltd.,USA) was used to quantify testosterone,testosterone 6β-hydroxylate,and other testosterone metabolites.A Q-Exactive ultra-performance liquid chromatography coupled with high resolution mass spectrometry (UPLC-HiMS,Thermo Fisher Co.Ltd.,USA) was used to assist the identification of the structure of 6β-hydroxylated testosterone.A C18column (Zorbax C18,Agilent Co.Ltd.,USA) was used for separation.Testosterone,HPLC-grade acetonitrile and methanol,NADP Na+,glucose-6-phosphate,and glucose-6-phosphate dehydrogenase were purchased from Sigma-Aldrich (St.Louis,MO,USA).Human liver microsomes were purchased from Shanghai RILD Co.Ltd.(Shanghai,China).Mice,rats,guinea pigs,rabbits were purchased from the Animal Experimental Center of Zunyi Medical University,and livers of those animals were harvested immediately after euthanasia.Dog (beagle) livers were got from Shanghai University of Traditional Chinese Medicine.Cattle and sheep livers were harvested immediately after death of animals in food market.

1.2 Preparation of animal liver microsomes After liver homogenation,microsomes were prepared from fresh livers following our previously established method[8].All microsomes products were checked with protein concentration determination kit and diluted to the protein concentration of 10 mg/ml and stored at -80oC.

1.3 Liver microsomes incubation system for CYP-mediated metabolisms Incubation system was set following our previously published method[9].Liver microsomes (0.5 mg/ml) were pre-incubated with testosterone (50 μM),glucose-6-phosphate (1 mM),glucose-6-phosphate dehydrogenase (1 Unit/ml),and MgCl2(4 mM).Reaction start from adding of NADP Na+(1 mM).The incubation system was normalized to 100 μl with phosphate buffer saline (PBS,100 mM,pH 7.4).10 min after incubation,100 μl of acetonitrile was added to stop the reaction and precipitate proteins.All tubes were shaken thoroughly and then centrifuged at 10,000×g and 4oC for 10 min.The supernatant was used for HPLC analysis.A reaction group (n=3) involving all reaction factors,a negative group (n=3) without NADP Na+,and a blank group (n=3) without testosterone were set up.

1.4 Determination of testosterone and its metabolites To quantify metabolites of testosterone,supernatant of each incubations from 1.3 were injected to an Agilent HPLC system,eluted by Water (A) - Methanol (B) on a C18column (Agilent Zorbax C18,4.6×250 mm,5 μm).Elution gradient was shown below:0～15 min,48%～30% of A; 15～22 min,30%～20% of A;22～22.5 min,20%～5% of A,followed by a column wash program.Testosterone and its metabolites were detected at UV 254 nm.To identify which one is testosterone 6β-hydroxylate,the chromatogram of the test group was compared with that of blank and negative controls.The tentatively assigned metabolites was introduced to UPLC-HiMS to obtain the molecular weight and MS/MS spectrums.

1.5 Statistics Correlation analysis was done in SPSS 18.0 (IBM Co.Ltd.,USA) program by using the function of linear correlation (pearson).Principle component analysis and visualization at heatmap were done in R program by using mixOmics and gplots package.

2 Results

2.1 CYP-mediated metabolic profile of testosterone in humans and animals Normally,we assumed 6β-hydroxylationIn is the most common metabolic pathway of testosterone.However,in the present study,in human and animal liver microsomes incubation system,we observed 18 NADP-dependent metabolites.Metabolites were assigned with numbers based on their retention time on chromatograms (Fig1).Guinea pigs generates almost all metabolites observed in the present study,showing the highest metabolic capacity among all species investigated in the present study.Sheep generated fewest metabolites.Viewing the whole metabolic profile,obvious species difference is available.

M1～M18 are NADP-dependent metabolites of testosterone. Fig 1 Metabolic profiles of testosterone in human and animal liver microsomes incubation system specific for CYP-mediated reaction

2.2 6β-hydroxylated testosterone identification in human liver microsome incubations In total 4 peaks were observed in human liver microsome incubation system.All of the 4 peaks were just obviously observed in reaction group but not negative or blank group (Fig 2A).Based on the retention time,we assigned them as M6,M8,M16,and M18.Among the 4 metabolites,M8 showed the predominant amount.As in most publication,6β-hydroxylation is the major and specific metabolic pathway,M8 was tentatively assigned as testosterone 6β-hydroxylate.Moreover,we introduced M8 into mass spectrometer for MS2 fragment analysis.In MS2 spectrum of M8 (Fig 2B),M+H peak showed consistent value with testosterone 6β-hydroxylate.The fragments 97.07 and 109.07 were successfully aligned to structure of testosterone 6β-hydroxylate.

2.3 Correlation analysis between testosterone 6β-hydroxylation and testosterone elimination In humans and 8 animal species,generation of 6β-hydroxylated testosterone and elimination of the substrate (testosterone) were compared with each other.A significant correlation with coefficient (R) of 0.8 (0 ～ 1) was observed (Fig 3).Except for rat,guinea pig,human,and rabbit,testosterone 6β-hydroxylation and testosterone elimination generally showed the same correlation trend.In rats and guinea pigs,the spots shift away the trend line to y-axis.It means that in these two animal species,metabolic pathways other than testosterone 6β-hydroxylation play relative important role for testosterone metabolism.In contrast,in humans and rabbits,the spots shift away the trend line to x-axis,indicating the 6β-hydroxylation play the predominant role in testosterone elimination in these two animals.

A：Chromatograms of incubations from reaction,negative,and blank groups.Serial number of metabolites is the same as that shown in fig.1;B：MS2 spectrum of M8. Fig 2 Identification of the predominant metabolites in human liver microsomes

Fig 3 Correlation analysis between testosterone 6β-hydroxylation and testosterone elimination

2.4 Quantitative visualization of testosterone metabolic profile Heatmap visualization divided all metabolites into three groups (Fig 4).The first group includes M8 and M18 which were generated by all investigated species.M2,M3,M12,M17,M7,M4,M1,M10,M15,and M11 were classified into the second group located in middle part of the heatmap.These metabolites just were observed in a few species,and the amount of these metabolites are relatively low.The third class metabolites includes M6,M5,M13,M9,M16,and M14.These metabolites were not observed in all species but in most species or have relative large amount in some species.Regarding individual species,Guinea pig has the most metabolites while sheep generates the fewest metabolites.Heatmap showed clearer information for metabolic profile of testosterone than chromatograms.

From green to red represent low to high amount of metabolites. Fig 4 Quantitative visualization of metabolic profile of testosterone in humans and 7 animal species by heatmap

2.5 Principle component analysis Metabolites data matrix was inputted into R program for principle component analysis.Score plots (Fig5A) showed that metabolic profile of testosterone in guinea pigs and rats are far away from other species.However,guinea pigs shifted away to right side while rats shifted away to downside.Human generally located in the center of the score plots.Mice and rabbits located close to humans.All of these phenomenon observed in the score plots indicated that the metabolic profile of testosterone in rabbits and mice are more similar to human than other animals.Guinea pigs and rats showed so different metabolic capacities from human,however,they are also different from each other.M16,M13,and M8 got larger loading values at PC1 (Fig5B).As guinea pigs shift to right side along PC1,it means that M16,M13,and M8 were generated more in guinea pigs than in other species.With the same method,we identified that M15 and M11 have higher amount in rats than in other species.All of those findings are consistent with the directly observation on the heatmap.

Fig 5 Score plots (A) and loading plots (B) for principle component analysis of testosterone metabolisms in humans and anmials

3 Discussion

For convenience and experimental cost saving,in pharmacokinetics study,rodents and rabbits are usually selected.For sampling,in order to get the blood samples at various time points from a single animal,rats were used more frequently than mice.In some cases,large animals were also used,such as dogs,pigs,as well as sheep.Although rats were used more frequently,at genomic DNA level,rats showed different sequence of CYP 3A from humans[10-11].In rats,the enzyme with similar function with human CYP 3A4,is called as CYP 3A1[12].In mice,this enzyme is called as CYP 3A11[13].At the function level,testosterone is always used to probe the activity of CYP 3A in various animal species[14-16].However,we found that,not like usually assumed,rats showed so different metabolic profile on testosterone metabolism from humans.In contrast,testosterone metabolic profile in mice close to human more than rats.Other than rodents,in relatively large animals,rabbits and dogs showed more similar metabolic profile of testosterone to human more than sheep,cattle,and pigs.In some animal model for special disease modulation,guinea pigs were used[17].However,in the present study,regarding CYP 3A activity,guinea pigs showed so higher capacity than human and other animal species.It implies that,once a drug candidate is proved to be metabolized by CYP 3A4 in human liver microsomes or human primary hepatocyte cultures,in the pre-clinical trial stage,to make sure the same drug exposureinvivo,mice should be used in the early stage,rats and guinea pigs should be avoided.If larger animals are needed,rabbits and dogs were recommended.

Taken together,there is huge species difference in testosterone metabolism mediated by CYP3A among human and animals.Mice,dogs,rabbits are recommended to be used as animal model to promise the similar drug exposure,if a drug has already been proved to be metabolized by human CYP 3A4.

[1] Baranczewski P,Stanczak A,Sundberg K,et al.Introduction to in vitro estimation of metabolic stability and drug interactions of new chemical entities in drug discovery and development [J].Pharmacol Rep,2006,58(4):453-472.

[2] Eddershaw P J,Beresford A P,Bayliss M K.ADME/PK as part of a rational approach to drug discovery [J].Drug Discov Today,2000,5(9):409-414.

[3] Streetman D S,Bertino J S Jr,Nafziger A N.Phenotyping of drug-metabolizing enzymes in adults:a review of in-vivo cytochrome P450 phenotyping probes [J].Pharmacogenetics,2000,10(3):187-216.

[4] Quattrochi L C,Guzelian P S.Cyp3A regulation:from pharmacology to nuclear receptors [J].Drug Metab Dispos,2001,29(5):615-622.

[5] Wang R W,Newton D J,Scheri T D,et al.Human cytochrome P450 3A4-catalyzed testosterone 6 beta-hydroxylation and erythromycin N-demethylation.Competition during catalysis [J].Drug Metab Dispos,1997,25(4):502-507.

[6] Maenpaa J,Syngelma T,Honkakoski P,et al.Comparative studies on coumarin and testosterone metabolism in mouse and human livers.Differential inhibitions by the anti-P450Coh antibody and metyrapone [J].Biochem Pharmacol,1991,42(6):1229-1235.

[7] Swales N J,Johnson T,Caldwell J.Cryopreservation of rat and mouse hepatocytes.II.Assessment of metabolic capacity using testosterone metabolism [J].Drug Metab Dispos,1996,24(11):1224-1230.

[8] He Y Q,Liu Y,Zhang B F,et al.Identification of the UDP-Glucuronosyltransferase Isozyme Involved in Senecionine Glucuronidation in Human Liver Microsomes [J].Drug Metabolism and Disposition,2010,38(4):626-634.

[9] 鲁艳柳,邓红,潘虹,等.山姜素对人肝微粒体中CYP1A2的选择性抑制研究[J].遵义医学院学报,2015,38(5):454-459.

[10]Hashimoto H,Toide K,Kitamura R,et al.Gene structure of CYP3A4,an adult‐specific form of cytochrome P450 in human livers,and its transcriptional control [J].The FEBS Journal,1993,218(2):585-595.

[11]Nagata K,Ogino M,Shimada M,et al.Structure and expression of the rat CYP3A1 gene:isolation of the gene (P450/6betaB) and characterization of the recombinant protein [J].Arch Biochem Biophys,1999,362(2):242-253.

[12]Kuzbari O,Peterson C M,Franklin M R,et al.Comparative analysis of human CYP3A4 and rat CYP3A1 induction and relevant gene expression by bisphenol A and diethylstilbestrol:implications for toxicity testing paradigms [J].Reprod Toxicol,2013,37 (3):24-30.

[13]Zimmermann C,Van Waterschoot R A,Harmsen S,et al.PXR-mediated induction of human CYP3A4 and mouse Cyp3a11 by the glucocorticoid budesonide [J].Eur J Pharm Sci,2009,36(4-5):565-571.

[14]Piver B,Berthou F,Dreano Y,et al.Inhibition of CYP3A,CYP1A and CYP2E1 activities by resveratrol and other non volatile red wine components [J].Toxicol Lett,2001,125(1-3):83-91.

[15]Huan J Y,Miranda C L,Buhler D R,et al.The roles of CYP3A and CYP2B isoforms in hepatic bioactivation and detoxification of the pyrrolizidine alkaloid senecionine in sheep and hamsters [J].Toxicol Appl Pharmacol,1998,151(2):229-235.

[16]Li C C,Yu H F,Chang C H,et al.Effects of lemongrass oil and citral on hepatic drug-metabolizing enzymes,oxidative stress,and acetaminophen toxicity in rats [J].Journal of Food and Drug Analysis,2017.DOI:10.1016/j.jfda.2017.01.008.

[17]Chang M Y,Gwon T M,Lee H S,et al.The effect of systemic lipoic acid on hearing preservation after cochlear implantation via the round window approach:A guinea pig model [J].Eur J Pharmacol,2017,799:67-72.

[收稿2017-06-02;修回2017-08-10]

(编辑：谭秀荣)

国家自然科学基金资助项目(NO：81402985；NO:81560673；NO:81660685)；贵州省科技重大专项(NO：黔科合重大专项字NO：[2015]6010)；贵州省科学技术基金资助项目(NO：黔科合J字[2015]2158；黔科合JZ字[2015]2010)；贵州省出国留学人员择优资助计划项目(NO：[2015]03)；国家级/省级大学生创新创业项目(NO：201510661002)；贵州省药剂学研究生卓越人才培养计划(黔教研合ZYRC字[NO：2014]019)；遵义医学院博士启动基金；遵义市科技局联合基金(遵市科合社字NO：[2016]15)。

吴庆，女，本科，教授，研究方向：药物分析技术，E-mail:68430978@qq.com；彭捷，女，本科，副主任医师，研究方向：内分泌学，E-mail:2219282718@qq.com。

睾酮代谢的种属差异研究

何芋岐1，曾 瑶1，凌 蕾1，陆安静1，杜艺玫1，杨 滔1，李伯达1，彭 捷2，吴 庆1

(1.遵义医学院 药学国家级实验教学示范中心(药学实验室)，贵州 遵义 563099；2.遵义医学院第三附属医院 内分泌科，贵州 遵义 563000)

目的研究睾酮在人及小鼠、大鼠、犬、兔、猪、豚鼠、牛、羊等实验动物肝微粒体代谢中的差异。方法睾酮分别于上述种属肝微粒体进行孵育，加入NADP Na、葡萄糖-6-磷酸、葡萄糖-6-磷酸脱氢酶等试剂启动体外P450酶反应。样品采用高效液相色谱紫外检测器及高分辨质谱进行分析。代谢轮廓用热图进行可视化，并用主成分分析进行差异研究。结果6β-羟基化不是睾酮唯一的代谢途径。本研究在不同种属中共发现了18种睾酮的代谢产物。尽管6β-羟基化在不同种属中与睾酮的底物消除基本呈正相关，但该途径无法补偿其他途径造成的底物消除。豚鼠具有最强的睾酮代谢能力，而羊对睾酮的代谢能力最弱。小鼠、犬、兔肝微粒体中，睾酮的代谢轮廓与人最相似。结论6β羟基化途径的定量不能用作所有物种CYP 3A活性评判的精准依据。如果药物证明由CYP 3A4进行代谢，在临床前期研究中，应结合药理学模型的需要，尽量采用小鼠、犬、兔作为动物模型，以保证药物体内暴露与人体实验尽量接近。

睾酮；种属差异；CYP 3A；代谢

R966

A

1000-2715(2017)05-0463-06