Inhibition of thyroxine binding to serum proteins by fenclofenac and related compounds

Clinica Chimica Acta, 112 (1981) 77-83 0 Elsevier/North-Holland Biomedical Press

CCA

77

1730

Inhibition of thyroxine binding to serum proteins by fenclofenac and related compounds S.J. Capper a,*, M.J. Humphrey b and A.B. Kurtz a a Institute of Nuclear Medicrne, The Middlesex Hospital Medical School, Moriimer Streei,

London

WIN 8AA (U.K.) and h Department of Drug Metabolism, Reckitt and Colman Lid, Pharmaceutical Division, Hull HlJ8 7DS (UK.) (Received

September

29th, 1980)

Summary

The anti-inflammatory drug fenclofenac lowers the serum concentrations of thyroxine (T4) and triidothyronine (T3) in subjects taking a regular dose of 600 mg twice daily. In eight male volunteers the concentration of T4 in serum was reduced to almost a third of the pre-treatment value while the concentration of T3 fell by about a half. The interaction of fenclofenac, and related compounds, with the thyroid hormone binding proteins of serum was studied in vitro by equilibrium dialysis with measurement of the concentrations of free T4 and T3 in the dialysate by immunoassay. Fenciofenac, at a concentration of 100 mg/l (of the order achieved therapeutically), increased the concentrations of free T4 by 131% and free T3 by 62%. The related compounds which increased the free hormone concentrations (although at higher concentrations than achieved during therapy) include diclofenac, monohydroxyfenclofenac and dihydroxyfenclofenac. A second in vitro method was used to study the interaction of these drugs with isolated thyroxine-binding globulin (TBG). An antiserum to TBG, coupled to Sephadex particles, was used to isolate TBG from other serum proteins. The effect of the different drugs on the binding of T4 by the immobilised TBG was then measured. These compounds competitively inhibit the binding of T4 by TBG, and for each an inhibitor constant (Ki) was determined.

Introduction

A number of drugs are known to affect the binding of thyroid hormones by serum proteins [ 1- 111; the most well known are salicylate and diphenylhydantoin. Salicylate competitively displaces T3 from serum binding proteins [4], and during chronic therapy diphenylhydantoin lowers the concentration of T4 in serum [6]. Fenclofenac is a non-steroidal anti-inflammatory agent which possesses some

* To whom correspondence

should be addressed.

78

structural resemblance to T4; it is a chlorinated phenoxyphenyl acetic acid which is largely protein bound in the circulation. It has been found to cause profoundly altered thyroid function tests without producing any clinical evidence of ill health [2,3,10]. This study investigates the effect of fenclofenac on the serum concentrations of T4 and T3 during therapy and, in in vitro studies, the effect of fenclofenac, fenclofenac metabolites and the related compound diclofenac on serum protein binding of T4 and T3 in an equilibrium dialysis system and on T4 binding by solid phase antibody-bound TBG. Materials and methods Volunteers Eight male volunteers took fenclofenac 600 mg twice daily for three weeks. This study received ethical committee approval. Blood samples were collected at 09:30; three samples were collected from each subject before taking the drug and, for the determination of steady state concentrations, a further three samples were collected after taking the drug for 10, 14 and 21 days (the fenclofenac dose in the morning was taken after sampling). The separated serum was stored deep frozen. Fenclofenac and related compounds Fenclofenac [2 - (2,4 - dichlorophenoxy) - phenyla~etic acid], monohydroxyfen clofenac [2-(2,4-dic~orophenoxy)-5-hydrox~henyl acetic acid], dihydroxyfenclofenac [2-(2,4-dichlorophenoxy)-4,5_dihydroxyphenyl acetic acid] and diclofenac [2-(2,6dichlorophenylamino)-phenyl acetate sodium] were supplied by the research laboratories of Reckitt and Colman Ltd. They were dissolved in a very small volume of sodium hydroxide solution, 0.5 mol/l, and were then diluted in buffer and stored at 4°C. The concentrations of fenclofenac and monohydroxyfen~lofenac in serum were determined by high pressure liquid chromatography. Thyronines L-Thyroxine and L-triiodothyronine, from Sigma Chemical Corp., were standardised by sp~trophotomet~ and stored in sealed ampoules in alkaline propylene glycol. L-[3’,5‘-‘251]thyroxine, with a specific activity of more than 1200 pCi/mg, and L-[3‘-i2”I]triiodothyronine, with a specific activity of more than 1700 pCi/mg, were obtained from the Radiochemical Centre, Amersham, U.K. Buffer

The buffer used in these studies was Hepes (~-2-hydroxyethylpiper~ne-~‘-2ethanesulphonic acid) 0.01 mol/l, pH ‘7.4, containing sodium chloride 6.2 g/l and sodium azide 0.65 g/l. Where indicated gelatin 1 g/l and Tween 20 0.75 ml/l were added. Solid phase thyroxine-birding g~o~lin The antiserum to human TBG (obtained from Seward Laboratories) was raised in sheep. It was coupled to cyanogen bromide activated ultrafine Sephadex using 500 ~1 of antiserum for each gram of Sephadex [12]. 100 mg of the Sephadex-coupled anti-TBG was incubated with 125 ~1 of normal pooled human serum in 50 ml buffer

79

for 18 h at 4°C. The immobilised antibody with bound TBG was then separated by centrifugation at 1500 X g, 4°C and washed three times in buffer containing gelatin and Tween 20. Radioimmunoassay

of thyroxine

and triiodothyronine

The total concentrations of T4 and T3 in serum were determined radioimmunoassays using solid phase antisera. Equilibrium

by routine

dialysis studies

Aliquots of pooled normal serum were mixed with equal volumes of buffer containing varying amounts (O-500 mg/l) of fenclofenac or one of the related compounds. 400 ~1 of the diluted serum was dialysed, using Visking tubing, for 18 h at 37°C against 4.8 ml of buffer and the concentrations of the T4 and T3 in the dialysate determined by immunoassay [ 131. Studies with thyroxine-binding

globulin

To a series of polystyrene tubes the following reagents were added in buffer containing gelatin and Tween 20: (1) 200 ~1 of buffer containing T4 at concentrations from O-3.45 nmol/l (O-2.68 ng/ml), (2) 100 ~1 of buffer containing labelled T4, 5500 d.p.s./ml, (3) 100 ~1 of buffer with or without drug (the various compounds were added at the following concentrations: fenclofenac 0.02 mg/l, diclofenac 0.02 mg/l, monohydroxyfenclofenac 0.1 mg/l, and dihydroxyfenclofenac 1 mg/l), (4) 250 ~1 of buffer containing Sephadex-immobilised TBG, 2 mg/ml, with a T4 binding site concentration of 184 fmol/l (143 pg/l). The tubes were incubated on a rotator for 24 h at room temperature and then centrifuged for 10 min at 1500 X g and 4°C. The supernatant was aspirated and the pellet was washed twice with cold buffer. The bound radioactivity was then counted; non-specific binding was determined in tubes containing Sephadex coupled antibody without TBG. The results were expressed as a double reciprocal plot of the amount of bound T4 (y-axis) against the concentration of T4 (x-axis). This procedure is analogous to a Lineweaver-Burk plot [ 141; the intercept on the y-axis gave the reciprocal of the T4 binding site concentration and the intercept on the x-axis the affinity constant. This data was analysed using the least squares method of linear regression. For competitive inhibitors the inhibitor constant (Ki) was calculated from:

where i is the inhibitor concentration, K is the affmity constant of TBG without inhibitor and K, the affinity constant of TBG in the presence of inhibitor. Results

Eight volunteers took fenclofenac 600 mg twice daily. Before taking the drug the mean concentration (k S.E.M.) of T4 in serum was 112.4 2 4.4 nmol/l (87.3 * 3.4 pgg/l) and of T3 2.09 I+ 0.09 nmol/l (1.36 + 0.06 hg/l). A steady state concentration of fenclofenac in the circulation was reached after 10 days and after this the mean concentration of T4 was 39.4 + 2.8 nmol/l (30.6 t 2.2 pg/l) and that of T3 was

80

TABLE

I

THE EFFECT OF FENCLOFENAC OR OTHER COMPOUNDS ON THE SERUM TION OF FREE T4 IN VITRO AS MEASURED BY EQUILIBRIUM DIALYSIS Values in pmol/l results are given.

“w

as mean?

S.E.M.; n in parentheses.

Fenclofenac

(mg/l) 0

18.35 28.6 k

50

4.5 (6) 42.4? 60.5_t 7.1 (6) 78.6k 10.7 (4)

100 150 200

TABLE

1.8 (6) 3.5 (6)

Where only two experiments

Monohydroxyfenclofenac

Dihydroxyfenclofenac

26.7% 32.4&

20.222.8 29.827.1 38. I k 9.6 35.5 k 8.4 30.622.2

1.8 (4) 2.1 (4)

45.4, 49.6 (2) 55.6.63.0 (2) 91.7, 101.4 (2)

were performed

both

Diclofenac

(3) (4) (3) (4) (4)

24.Ok 41.9i 57.5 _t 66.92 99.7 k

2.8 4.3 3.2 11.9 6.0

(5) (6) (4) (4) (4)

II

THE EFFECT OF FENCLOFENAC OR OTHER COMPOUNDS ON THE SERUM TION OF FREE T3 IN VITRO AS MEASURED BY EQUILIBRIUM DIALYSIS Values in pmol/l

Drug

as meankS.E.M.;

Fenclofenac

(mg/l) 0

6.9t0.2

(IO)

8.6k0.5 (10) 11.2kO.5 (IO) 12.9kO.5 (10) 15.5~1.0(10)

50 100 150 200

TABLE

CONCENTRA-

CONCENTRA-

n in parentheses

Monohydroxyfenclofenac

Dihydroxyfenclofenac

7.4kO.7 8. I t 0.6 IO.82 1.0 12.9kO.8 17.4io.3

7.120.7 7.5 i0.6 8.6i0.5 10.3 kO.5 10.0t0.7

(4) (4) (4) (4) (4)

Diclofenac

(3) (4) (3) (4) (4)

6.7ir0.3 9.6 20.3 10.4kO.9 15.3+0.7 17.411.4

(5) (6) (6) (6) (6)

III

COMPETITIVE DISPLACEMENT OTHER COMPOUNDS The equilibrium

constant

Drug

OF T4 FROM

of TBG for T4 in the absence

Concentration drug (mg/B

Fenclofenac Diclofenac Monohydroxyfenclofenac Dihydroxyfenclofenac

SOLID

0.02 0.02 0.10 I.0

PHASE

TBG BY FENCLOFENAC

of drugs in this system was

of

Inhibitor constant

AND

I. 15 X IO9 l/mol.

K,

(mg/l) 0.055 0.028 0.127 20.4

1.15 2 0.06 nmol/l (0.75 & 0.05 pg/l). The mean steady state concentration of fenclofenac in serum was 77.7 2 5.4 mg/l and of monohydroxyfenclofenac 3.49 k 0.23 mg/l. The addition of fenclofenac up to a concentration of 250 mg/l, or of the other compounds up to concentrations of 50 mg/l, did not cause any displacement of T4 or T3 from antibodies in any of the assays used.

6

Fig. 1. The Lineweaver-Burk plot for the interaction of thyroxine preparation of thyroxine-binding globulin (duplicate values).

and fenclofenac

with a solid phase

Equilibrium dialysis demonstrated that all of the compounds studied displaced T4 and T3 from serum protein binding sites with dose-related increases in the concentrations of T4 and T3 in the dialysate. The results are given in Tables I and II. Fenclofenac and the other compounds caused competitive inhibition of T4 binding by immobilised TBG. The Lineweaver-Burk plot for fenclofenac is shown in Fig. 1. The inhibitor constants ( Ki) for these compounds are shown in Table III. Discussion

Fenclofenac profoundly altered the concentrations of T4 and T3 in the circulation; there was a reduction during therapy, of the T4 concentration to 35% of the pre-treatment level and for the T3 concentration to 55% of the pre-treatment level. On regular therapy, pituitary function- as,judged by basal thyrotrophin concentrations in serum and by the secretory response to thyrotrophin-releasing hormone- has been reported to be normal [3,10]. In addition, after discontinuation of therapy, the circulating concentrations of T4 and T3 return to normal [3]. These clinical results are consistent with fenclofenac acting as an inhibitor of the binding of thyroid hormones by serum proteins. . We have shown by equilibrium dialysis, with immunoassay of T4 and T3 in the dialysate, that fenclofenac, at concentrations similar to those found during therapy, displaces both thyronines from serum-binding proteins. Previous studies have used ultrafiltration of serum containing labelled T4 to show displacement of T4 and T3 by drugs [ 1,4]. Using isolated TBG we have shown competitive inhibition by fenclofenac of the binding of T4; it is of course also possible that fenclofenac has an

82

effect on thyroid hormone binding to albumin or pre~bu~n. With isolated TBG it has been possible to study inhibitors quantitatively; electrophoretic methods, which have been used to demonstrate drug-protein interactions are essentially qualitative in nature [1,4,9,1 I]. The hydroxylat~ derivatives of fenclofenac also show i~bitory activity, the monohydroxy derivative being more potent than the dihydroxy derivative. Monohydroxyfenclofenac has been identified as a major urinary metabolite of fenclofenac during therapy [15] but, as plasma concentrations are less than 5% of fenclofenac levels, it is unlikely that it plays a major role in the displacement of T4 and T3 from the circulation. Diclofenac is also a potent competitive inhibitor but during treatment the plasma concentrations of the drug are much lower (3 mg/l or less, according to the manufacturer’s data sheet) than are seen during fenclofenac treatment. ’ We have used two different in vitro methods to study the interaction of drugs with thyroid hormone binding proteins. Equilib~um dialysis provides a general method for such an investigation and has the advantage that a mixture of serum proteins can be studied under conditions that are near physiological. The solid-phase method, with antibody isolation of a single binding protein, is useful for studying a protein in isolation but under less physiological conditions. The latter type of system could be applied to the study of other binding proteins and their interaction with ligands and drugs.

S.J.C. is the recipient of a Science Research Council/CASE award. We are greatful to Dr. L. Wide (Uppsala) for the gift of ultrafine Sephadex and to Dr. 1. Flockhart (Reckitt and Colman Ltd.) for the assays of fenclofenac and monohydroxyfenclofenac. References 1 f&in, w. and S&u&r, G.C. (1968)Decreased serum free thyroxine concentrationin patientstreated with dinhenvlhvdantoin. J. Clin. Endocrinol. 28, 181- 186 r 2 Essigman, WK. (1980) Fenclofenac in osteoarthrosis. In: Fenclofenac, Royal Society of Medicine International Congress and Symposium Series, 28, pp. I- 7, Academic Press Inc., London 3 Humphrey, M.J., Capper, S.J. and Kurtz, A.B. (1980) Fenclofenac and thyroid hormone concentrations. Lancet 1,48?-488 4 Larsen, P.R. (1972) Salicylate-induced increases in free t~~o~yro~ne in human serum- evidence of inhibition of triiodothyronine binding to thyroxine-binding globulin and thyroxine-binding prealbumin. J. Clin Invest. 51, 1125- 1134 5 Larsen, P.R., Atkinson, A.J., Wellman, H.N. and Goldsmith, R.E. (1970) The effect of diphenylhydantoin on thyroxine metabo~sm in man. J. Clin. Invest. 49, t266- 1279 6 Liewendahl, K., Majuri, H. and Helenius, T. (1978) Thyroid function tests in patients on long-term treatment with various anti-convulsant drugs. Clin. Endocrinol. 8, 185- 191 I Oppenheimer, J. (1973) Interaction of drugs with thyroid hormone binding sites. Ann. N.Y. Acad. Sci. 226,333-340 J. (1968) Role of plasma proteins in the binding, distribution and metabo~sm of 8 Opp~~mer, thyroid hormones. New Engl. J. Med. 278, 1153- 1162 9 Oppenheimer, J. and Tavemetti, R. (1962) Studies on the thyroxine-diphenylhydantoin interaction: effect of 5,5’-DPH on the displacement of L-thyroxine from TBG. Endocrinology 71, 496-504 10 Ratcliffe, W.A., Hazleton, R.A. and Thompson, J.J. (1980) Effect of fenclofenac on thyroid function tests. Lancet I, 432 ,_I

83 1I Wolff, _I., Standaert, M.E. and Rall, J.E. (1961) Thyroxine displacement from serum proteins and depression of serum protein-bound iodine by certain drugs. J. Clin. Invest. 40 (2), 1373- 1379 12 Wide, L. (1969) Radioimmuno~says employing immunosorbents. In: Karolinska Symposia on Research Methods in Reproductive Endocrinology, (E., Dinfalusy, ed.,), Acta Endocrinol., Suppl. 142, 207-221 13 Ekins, R.P. and Ellis, SM. (1975) The radioimmunoassay of free thyroid hormones. In: Thyroid Research (Robbins, J. and Braverman, L.E., eds.). ICS 378, pp. 597-600, Excerpta Medica, Amsterdam 14 Dixon, M. and Webb, E. (1964) Enzymes, 2nd edn., pp. 315-335, Longmans, London I5 Greenslade, D., Havler, M.E., Humphrey, M.J., Jordan, B.J. and Rance, M.J. (1980) Species differences in the metabolism and excretion of fenclofenac, XenobioticalO, 753-760