http://informahealthcare.com/gyeISSN: 0951-3590 (print), 1473-0766 (electronic)

Gynecol Endocrinol, Early Online: 1–5! 2014 Informa UK Ltd. DOI: 10.3109/09513590.2014.943727

ORIGINAL ARTICLE

Antioxidant effect of the active metabolites of tibolone

Julia Stark1*, Szabolcs Varbiro2*, Miklos Sipos2, Zsolt Tulassay1, Levente Sara2, Ildiko Adler1, Elek Dinya3,Zoltan Magyar4, Bela Szekacs1, Istvan Marczell1, Helenius J. Kloosterboer5, Karoly Racz1, and Gabor Bekesi1

12nd Department of Internal Medicine, Faculty of Medicine, 22nd Department of Obstetrics and Gynecology, Faculty of Medicine, Semmelweis

University, Budapest, Hungary, 3Faculty of Health and Public Services, Institute of Health Informatics Development and Further Training,

Semmelweis University, Budapest, Hungary, 41st Department of Obstetrics and Gynecology, Faculty of Medicine, Semmelweis University, Budapest,

Hungary, and 5KC2, Oss, The Netherlands

Abstract

Certain steroidal compounds have an antioxidant effect in humans. Our aim was to testwhether the synthetic steroid tibolone and its metabolites are also able to display such aproperty. For this, granulocytes from healthy men and women were incubated for twohours with different concentrations (10�7, 10�8, 10�9 M) of either estradiol, tibolone, 3a-hydroxytibolone, 3b-hydroxytibolone, D4-tibolone, 3a-sulfated-tibolone, 3a-17b-disulfated-tibolone, 3b-sulfated-tibolone or 3b-17b-disulfated-tibolone. Superoxide anion generation ofneutrophils was measured by photometry. Results of different steroids were given aspercentages of their controls. A more simple superoxide generating system, the xanthine–xanthine oxidase reaction was also tested. We found that granulocyte superoxide productiondid not differ from the control using 10�9 M of steroids. Using 10�8 M concentration: estradiol(80.9 ± 2.5%); 3b-sulfated-tibolone (83.3 ± 4.7%); 3b-17b-disulfated-tibolone (81.0 ± 4.2%)caused a significant decrease in superoxide production, compared to the control. In additionat 10�7 M, 3b-hydroxytibolone and 3a-sulfated-tibolone also showed antioxidant effects.In the xanthine–xanthine oxidase system estradiol (67.4 ± 1.0%), 3a-sulfated-tibolone(85.8 ± 5.3%), 3a-17b-disulfated-tibolone (71.9 ± 2.5%), 3b-sulfated-tibolone (73.9 ± 5.0%),and 3b-17b-disulfated-tibolone (65.8 ± 3.4%) caused a significant decrease in superoxideproduction. Conclusively, although tibolone itself did not show significant antioxidant capacity,most of its active metabolites have antioxidant effects.

Keywords

Antioxidant effect, estradiol, superoxide anioninhibition, tibolone, tibolone metabolites

History

Received 9 April 2014Revised 5 June 2014Accepted 8 July 2014Published online 23 July 2014

Introduction

Reactive oxygen species (ROS) are highly reactive oxygen-derived metabolites that react with lipids, proteins, peptides andnucleic acids. ROS from activated macrophages and neutrophilsenhance the intracellular signaling pathways of lymphocytesand contribute to activation of the antigen-specific immuneresponse. Intracellular ROS are under the control of enzymeslike superoxide-dismutase, catalase and glutathione peroxidase.If ROS reach extracellular space, the balance of free radicalsand antioxidant compounds is disturbed, because extracellularscavengers are weaker than intracellular ones. There is a growingawareness that oxidative stress plays a major role in variousclinical conditions like aging, diabetes, atherosclerosis, chronicinflammation, malignant diseases, neurodegenerative diseasesand disorders associated with menopause [1,2].

Estradiol has a known antioxidant effect through a directdecrease in superoxide anion production, neutralization of excessROS and enhancement of cellular antioxidative defense molecules(for review see [3]). Neutralization of excess ROS can be achievedwith certain structural features, such as the phenolic OH groupat the C-3 position of the A ring of the steroid molecule, and bythe activation of nitric oxide synthase (NOS) and nitric oxide(NO) generation [3]. In our earlier study, as a result of the effectof estrogen, increased myeloperoxidase activity and its negativefeedback, decreased superoxide anion levels were found [4].Endogenous estrogen plays a major role in the protection of theendothelial function against the effect of aging in premenopausalwomen by preserving NO availability and eliminating oxidativestress [5]. These effects may contribute to the well-knownadvantages of premenopausal women compared to menopausalwomen in respect of antioxidant status and prevention of the freeradical-mediated diseases mentioned above.

Upon the antioxidant effect of estrogen, we previously testedthe antioxidant capacity of many natural steroid hormonesand metabolites. In addition to estradiol, DHEA, DHEAS,testosterone, cortisol, progesterone, estrone and estriol haveproven antioxidant activity. On the other hand, certain com-pounds (such as 11b-hydroxyprogesterone) had pro-oxidantproperty [6,7].

After a systematic examination of the antioxidant effect ofnatural steroid structures our aim is now to test synthetic steroid

*These authors contributed equally to this work.

Address for correspondence: Gabor Bekesi, MD, PhD, 2nd Department ofInternal Medicine, Faculty of Medicine, Semmelweis University,Szentkiralyi u. 46, H-1088, Budapest, Hungary. Tel: +36/30/9324008.Fax: +36/1/2660816. E-mail: bekesi.gabor@med.semmelweis-univ.huSzabolcs Varbiro, MD, PhD, 2nd Department of Obstetrics andGynecology, Faculty of Medicine, Semmelweis University, Ulloi ut 78/A H-1082, Budapest, Hungary. Tel: +36/20/3359099. Fax: +36/1/3346616. E-mail: varbiro.szabolcs@med.semmelweis-univ.hu

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/24/

14Fo

r pe

rson

al u

se o

nly.

structures commonly used in clinical practice. Tibolone is awell known drug in menopausal medicine, relieves climactericsymptoms and prevents postmenopausal bone loss. It is classifiedas a selective tissue estrogenic activity regulator (STEAR) dueto its tissue-selective effects [8,9]. There is little informationso far about the effects of tibolone on superoxide production.Some data are available about the influence of tibolone onthe NOS–NO system: tibolone induced a sustained increase ofNO plasma levels in postmenopausal women, mainly throughactivating NO synthesis in human endothelial cells [10,11].In another study, tibolone was effective as an antioxidant in thebrain cortex, tibolone-treated animals showed lower lipidhydroperoxide levels compared to the control ovariectomizedrats [12]. Tibolone and two of its metabolites, D4-tibolone and3b-hydroxytibolone in a high concentration (10�6 M) increasedcatalase activity in human breast cancer cells, thus the detoxifi-cation of hydrogen peroxide was enhanced [13].

The aim of our present study was to identify specific tibolonemetabolites, which may have an active effect on oxidative stressand compare the effect of these compounds with knownantioxidant steroid structures.

Methods

Blood samples were obtained in EDTA tubes from 10 healthyvolunteers (women and men, aged 28 to 46). Volunteers weremedical doctors, all of whom gave informed consent. The studywas approved by the Semmelweis University ethical committee.All volunteers had a negative medical history and none ofthem was taking any medication. The blood was applied toFicoll in layers for the sedimentation of red blood cells and putaside for an hour. Then it was transferred to 63 and 72% Percolland centrifuged for 25 min at a rate of 300� g at 20 �C.Granulocytes were separated as follows: they were buffer-washed twice and centrifuged with 220� g. Cells aggregated atthe bottom of the tube were re-suspended in a few milliliters ofbuffer (Hank’s salt solution, Biochrom KG, Berlin), then countedusing Turk’s solution. The cell concentration of the suspensionwas adjusted to 5 million/ml. The granulocyte suspensionswere incubated for two hours at 37 �C with different con-centrations (10�7, 10�8 and 10�9 M) of steroid hormones. Thefollowing steroids were tested in this study: estradiol (Sigma, StLouis, MO), tibolone, 3a-hydroxytibolone, 3b-hydroxytibolone,D4-tibolone, 3a-sulfated-tibolone, 3a-17b-disulfated-tibolone,3b-sulfated-tibolone, 3b-17b-disulfated-tibolone (tibolone metab-olites were provided by the manufacturer).

The superoxide anion generation of neutrophils was measuredusing Guarnieri’s method [14], as modified by Jansson [15].The quantity of free radicals released was assessed byphotometry; the reduction of ferricytochrome-C (Sigma, StLouis, MO) by the superoxide was measured as optical densityat 550 nm. The maximum superoxide anion production wasassessed as change of optical density (DOD), five minutes afterstimulation with N-formyl-Met-Leu-Phe (FMLP, Sigma). Thefinal concentration of FMLP was 10�6 M. The result of the firstmeasurement just after adding FMLP to the cell suspensionserved as the starting point (zero value). Further measurementswere carried out every minute (0–1–2–3–4–5 min). Using themolar extinction coefficient of cytochrome and standardizing theoutput per 106 cells, results are given as nmol superoxide anion/106 cells. Results for different steroids were given as percentagesof their controls in order to improve comparability. All com-pounds were tested with the same assay.

To control our results, we also tested the compounds in a non-cellular superoxide generating system, the xanthine–xanthineoxidase reaction. During this additional experiment, we used 0.05units/ml of xanthine oxidase (Sigma) in 100 mM K-phosphatebuffer, pH 7.8, containing 0.2 mM EDTA and 0.5 mM xanthine(Sigma). Steroidal compounds were tested at 10�7 M.Measurements of the reduction of ferricytochrome-C byGuarnieri’s method (mentioned above) were carried out for6 min at a wavelength of 550 nm [14].

Statistical analysis was performed using General LinearModels (SAS 8.2 statistical software, Procedure GLM, Cary,NC) and Dunnett’s ‘‘t’’ test to compare each treatment versuscontrol. The comparisons for each compound were carried out onthe three concentration levels mentioned above. Results wereconsidered significant at the level of p50.05. Results are given asmeans ± SEM. The error bars on the diagrams represent errorsbetween individuals.

Results

Ex vivo superoxide anion production after incubationwith different steroid concentrations

Granulocyte superoxide production was not different from thecontrol using 10�9 M of the various tibolone metabolites and thereference estradiol.

At 10�8 M, estradiol (80.9 ± 2.5%, p50.05), 3b-sulfated-tibolone (83.3 ± 4.7%, p50.05), and 3b-17b-disulfated-tibolone(81.0 ± 4.2%, p50.05) caused a significant decrease in ex vivosuperoxide production (Figure 1).

Figure 1. Superoxide anion production of human neutrophil granulocytes in the percentages of controls in the presence of 10�8 M of different steroidslabeled in the figure. Results are given as means ± SEM. The error bars on the diagrams represent errors between individuals. White columns with anasterisk show significant alterations compared to the controls (p50.05). Grey columns represent the non-significant groups.

2 J. Stark et al. Gynecol Endocrinol, Early Online: 1–5

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/24/

14Fo

r pe

rson

al u

se o

nly.

At 10�7 M the above-mentioned compounds produced ahigher inhibition of superoxide formation and now the effectof 3b-hydroxytibolone and 3a-sulfated-tibolone also becamesignificant: estradiol (76.4 ± 4.2%, p50.001); 3b-hydroxytibolone(82.9 ± 5.3%, p50.05); 3a-sulfated-tibolone (81.1 ± 4.4%,p50.05); 3b-sulfated-tibolone (79.2 ± 5.7%, p50.01);3b-17b-disulfated-tibolone (74.6 ± 5.1%, p50.0001) (Figure 2).

Tibolone, 3a-hydroxytibolone, D4-tibolone and 3a-17b-disulfated-tibolone did not cause any significantchange in the superoxide production of neutrophil gran-ulocytes at the concentrations used.

Superoxide anion production in the xanthine–xanthineoxidase system

In the xanthine–xanthine oxidase system estradiol (67.4 ± 1.0%,p50.05), 3a-sulfated-tibolone (85.8 ± 5.3%, p50.05), 3a-17b-disulfated-tibolone (71.9 ± 2.5%, p50.05), 3b-sulfated-tibolone(73.9 ± 5.0%, p50.05) and 3b-17b-disulfated-tibolone(65.8 ± 3.4%, p50.05) decreased superoxide productionsignificantly (Figure 3). Tibolone, 3a-hydroxytibolone, 3b-hydroxytibolone and D4-tibolone did not cause any significantchange in superoxide production.

Discussion

In our earlier studies we found decreased superoxide anionproduction after using several natural steroid hormones and their

intermediate metabolites [4,6,7,16]. Tibolone is rapidly metabo-lized into three steroid receptor activating compounds whichdetermine highly the biological activity in various tissues.Tibolone metabolites are to a large degree sulfated and areinvolved in the tissue selective action of tibolone [8,11,17].High concentrations of sulfated metabolites are present in thecirculation (see references in [8]).

Sulfated metabolites are not active at the steroid receptorlevel, but may have antioxidant properties, as we have shown.Sulfconjugated catechol estrogens also have antioxidant activity[18], and sulfated flavonoids possess this property as well [19].DHEAS, the sulfated metabolite of DHEA also has antioxidantproperties [20]. The mechanism of action via a non-receptor pathis still unclear. Iwasaki et al. [20] suggested that the mechanismmay set in via the inhibition of proinflammatory cytokine-stimulated, NF-kappa B-mediated transcription.

Some data are available on the improvement of free radical-mediated diseases after tibolone treatment. It has beneficial effecton inflammation and oxidative state in postmenopausal women[21]. Progression of Sjogren’s syndrome might be prevented bytibolone treatment [22]. The effect of tibolone on atherosclerosisis somewhat contradictory: unlike estradiol, it decreases high-density lipoprotein (HDL) cholesterol and triglycerides; however,animal models and clinical studies did not confirm an increasedrisk of plaque formation and cardiovascular diseases [8]. In astudy by Zandberg et al. [23] tibolone protected ovariectomizedand cholesterol-fed rabbits from atherosclerosis: it reduced the

Figure 2. Superoxide anion production of human neutrophil granulocytes in the percentages of controls in the presence of 10�7 M of different steroidslabeled in the figure. Results are given as means ± SEM. The error bars on the diagrams represent errors between individuals. White columns with anasterisk show significant alterations compared to the controls (p50.05). Grey columns represent the non-significant groups.

Figure 3. Superoxide anion production in the xanthine–xanthine oxidase system in the percentage of the controls in the presence of 10�7 M of differentsteroids labeled in the figure. Results are given as means ± SEM. The error bars on the diagrams represent errors between individuals. White columnswith an asterisk show significant alterations compared to the controls (p50.05). Grey columns represent the non-significant groups.

DOI: 10.3109/09513590.2014.943727 Tibolone as antioxidant 3

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/24/

14Fo

r pe

rson

al u

se o

nly.

accumulation of cholesterol, fatty streak formation, impairment ofendothelium-dependent response, and advanced lesion formationafter endothelial damage. The tibolone-derived prevention ofatherosclerotic lesion formation is in part carried out throughreducing leukocyte adhesion molecule expression on humanendothelial cells [24]. Beyond the effects on the NOS–NO system[10–12], the antioxidant activity verified by this study mayexplain the blocking effect of tibolone on these partially freeradical-mediated pathomechanisms.

Based on the results of the present study, we can concludethat some of the tibolone metabolites have pronounced antioxi-dant activity. Beyond the steroid structure, we observed otherchemical specificities in this action. Almost all of the sulfatedmetabolites display antioxidant activity, the strongest one was 3b-17b-disulfated-tibolone in both tests used in this study. Thismetabolite shows an activity which is comparable with thereference compound estradiol and is present in large amounts inthe circulation of women who use tibolone [17]. Apparently,phenolic A-ring and steroid receptor activity are not absoluterequirements for the antioxidant action of a compound. There isno indication of any pro-oxidant activity. Both model systems – exvivo and in vitro – demonstrated similar effects of tibolonederivatives. The concentrations used in this study are similarto plasma levels of tibolone metabolites after tibolone treatmentof postmenopausal women [8,17], and were commonly used inin vitro studies in the literature and in our own previousmeasurements.

As mentioned earlier, oxidative stress plays a major role inaging and subsequent disorders [1]. Diseases such as diabetes,metabolic syndrome, atherosclerosis, hypertension and neurode-generative diseases have well-established therapy. Some of thesemedications have been found to have an antioxidant effect, such asselegiline in Parkinson’s disease [25], the angiotensin-convertingenzyme inhibitors captopril and ramipril [26,27], the non-selective beta-blocker carvedilol [28], and also some of thestatins and fibrates [29,30]. Tibolone has been used for thetreatment of menopausal complaints, and according to ourresults, apart from these beneficial effects it also has anantioxidant capacity through the active metabolites. Thus, itmight have some additive action combined with other first-linedrugs used for menopausal patients in different free radical-mediated illnesses. Our data may benefit the development of newfree radical scavengers/blockers using molecular design on thetarget molecule.

Acknowledgements

The authors wish to express their gratitude to OrganonInternational Ltd. for their technical, material and methodologicalsupport and for providing us with the different metabolites of tibolone,to Balazs Gerecz for his administrative and technical support and toKrisztina Nagy for her devoted efforts in solving technical problems inlaboratory work.

Declaration of interest

This study was sponsored by Organon International Ltd. andMaecenator Foundation.

References

1. Gil del Valle L. Oxidative stress in aging: theoretical outcomes andclinical evidences in humans. Biomed Aging Pathol 2011;1:1–7.

2. Babior BM. Phagocytes and oxidative stress. Am J Med 2000;109:33–44.

3. Kumar S, Lata K, Mukhopadhyay S, Mukherjee TK. Role ofestrogen receptors in pro-oxidative and anti-oxidative actions ofestrogens: a perspective. Biochim Biophys Acta 2010;1800:1127–35.

4. Bekesi G, Kakucs R, Varbiro S, et al. In vitro effects of differentsteroid hormones on superoxide anion production of humanneutrophil granulocytes. Steroids 2000;65:889–94.

5. Viridis A, Taddei S. Endothelial aging and gender. Maturitas 2012;71:326–30.

6. Bekesi G, Racz K, Hrabak A, et al. Systematic investigation ofdifferent steroid precursors with respect to their effect on superoxideanion production by human neutrophil granulocytes. Horm MetabRes 2004;36:155–63.

7. Bekesi G, Tulassay Z, Racz K, et al. The effect of estrogens onsuperoxide anion generation by human neutrophil granulocytes:possible consequences of the antioxidant defense. GynecolEndocrinol 2007;23:451–4.

8. Kloosterboer HJ. Historical milestones in the development oftibolone (Livial�). Climacteric 2011;14:609–21.

9. Biglia N, Maffei S, Lello S, Nappi RE. Tibolone in postmenopausalwomen: a review based on recent randomised controlled clinicaltrials. Gynecol Endocrinol 2010;26:804–14.

10. Cicinelli E, Ignarro LJ, Galantino P, et al. Effects of tibolone onplasma levels of nitric oxide in postmenopausal women. Fertil Steril2002;78:464–8.

11. Simoncini T, Mannella P, Fornari L, et al. Tibolone activates nitricoxide synthesis in human endothelial cells. J Clin Endocrinol Metab2004;89:4594–600.

12. de Aguiar RB, Dickel OE, Cunha RW, et al. Estradiol valerateand tibolone: effects upon brain oxidative stress and bloodbiochemistry during aging in female rats. Biogerontology 2008;9:285–98.

13. Petit E, Courtin A, Kloosterboer HJ, et al. Progestins induce catalaseactivities in breast cancer cells through PRB isoform: correlationwith cell growth inhibition. J Steroid Biochem Mol Biol 2009;115:153–60.

14. Guarnieri C, Melandri G, Caldarera I. Reduced oxidative activity ofcirculating neotrophils in patients after myocardial infarction. CellBiochem Funct 1990;8:157–62.

15. Jansson G. Oestrogen-induced enhancement of myeloperoxidaseactivity in human polymorphonuclear leukocytes – a possible causeof oxidative stress in inflammatory cells. Free Radic Res Commun1991;14:195–208.

16. Bekesi G, Kakucs R, Sandor J, et al. Plasma concentrationof myeloperoxidase enzyme in pre- and postclimacterialpeople: related superoxide anion generation. Exp Gerontol 2001;37:137–48.

17. Vos RM, Krebbers SF, Verhoeven CH, Delbressine LP. The in vivohuman metabolism of tibolone. Drug Metab Dispos 2002;30:106–12.

18. Takanashi K, Watanabe K, Yoshizawa I. On the inhibitory effect ofC17-sulfoconjugated catechol estrogens upon lipid peroxidation ofrat liver microsomes. Biol Pharm Bull 1995;18:1120–5.

19. Lodi F, Jimenez R, Moreno L, et al. Glucuronidated and sulfatedmetabolites of the flavonoid quercetin prevent endothelial dysfunc-tion but lack direct vasorelaxant effects in rat aorta. Atherosclerosis2009;204:34–9.

20. Iwasaki Y, Asai M, Yoshida M, et al. Dehydroepiandrosterone-sulfate inhibits nuclear factor-kappaB-dependent transcription inhepatocytes, possibly through antioxidant effect. J Clin EndocrinolMetab 2004;89:3449–54.

21. Vassalle C, Cicinelli E, Lello S, et al. Effects of menopause andtibolone on different cardiovascular biomarkers in healthy women.Gynecol Endocrinol 2011;27:163–9.

22. Sartore A, Grimaldi E, Guaschino S. The treatment of Sjogren’ssyndrome with tibolone: a case report. Am J Obstet Gynecol 2003;189:894.

23. Zandberg P, Peters JLM, Demacker PNM, et al. Tiboloneprevents atherosclerotic lesion formation in cholesterol-fed, ovar-iectomized rabbits. Arterioscler Thromb Vasc Biol 1998;18:1844–54.

24. Sator K, Sator MO, Sator PG, et al. Effects of tibolone on selectinsin postmenopausal women. Maturitas 2006;53:166–70.

25. Foley P, Gerlach M, Youdim MB, Riederer P. MAO-B inhibitors:multiple roles in the therapy of neurodegenerative disorders?Parkinsonism Relat Disord 2000;6:25–47.

26. Benzie IF, Tomlinson B. Antioxidant power of angiotensin-converting enzyme inhibitors in vitro. Br J Clin Pharmacol 1998;45:168–9.

4 J. Stark et al. Gynecol Endocrinol, Early Online: 1–5

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/24/

14Fo

r pe

rson

al u

se o

nly.

27. Keles MS, Bayir Y, Suleyman H, Halici Z. Investigation of effects ofLacidipine, Ramipril and Valsartan on DNA damage and oxidativestress occurred in acute and chronic periods following isoproterenol-induced myocardial infarct in rats. Mol Cell Biochem 2009;328:109–17.

28. Gasparova Z, Ondrejickova O, Gajdosıkova A, et al. Oxidative stressinduced by the Fe/ascorbic acid system or model ischemia in vitro:

effect of carvedilol and pyridoindole antioxidant SMe1EC2 in youngand adult rat brain tissue. Interdiscip Toxicol 2010;3:122–6.

29. Murrow JR, Sher S, Ali S, et al. The differential effect of statins onoxidative stress and endothelial function: atorvastatin versuspravastatin. J Clin Lipidol 2012;6:42–9.

30. Heller FR, Descamps O, Hondekijn JC. LDL oxidation: therapeuticperspectives. Atherosclerosis 1998;137:S25–31.

DOI: 10.3109/09513590.2014.943727 Tibolone as antioxidant 5

Gyn

ecol

End

ocri

nol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y M

ichi

gan

Uni

vers

ity o

n 10

/24/

14Fo

r pe

rson

al u

se o

nly.