Uptake of indomethacin by the endometrium of rat uterus

PROSTAGLANDINS

UPTAKE OF INDOMETHACIN

BY THE ENDOMETRIUM OF RAT UTERUS

Philip V. Peplow and W. Howard Evans National Institute for Medical Research, Mill Hill, London NW7, U.K. ABSTRACT The uptake of indomethacin into rat endometrium was studied using biochemical subcellular fractionation techniques. Endometrial cell homogenates were separated by centrifugation in Percoll gradients, and the distribution of subcellular membranes and organelles in the fractions ascertained using biochemical and morphological markers. Two hours after intraluminal perfusion of uteri with [14C]-indomethatin, the drug showed a bimodal distribution. It was present in fractions containing soluble components and plasma membranes and/or their endocytic derivatives. The results show that indomethacin is taken up by rat endometrium, but it does not appear to associate with the Golgi-endoplasmic reticulum-lysosomal system. INTRODUCTION Indomethacin,when released locally from a silastic intrauterine device, causes metabolic changes in the rat uterus. These alterations include a decrease in uterine hypertrophy and in the basal levels of prostaglandins F&and E (I) and a reduction in the embryotoxicity of the uterine environment (2). The changes induced by indomethacin, taken with its effects on the morphology of endometrial cells (3, 4) make it pertinent to determine whether the drug is taken up by the tissue, and if so, its subcellular fate. The present study describes a rapid method for separating and identifying subcellular components of rat endometrium, and shows that intraluminally administered radioactiveindomethacin becomes tightly associated with subcellular membranes. MATERIALS AND METHODS Chemicals Indomethacin was obtained from Sigma. [14C] Indomethacin was obtained as a gift from Merck, Sharp and Dohme, Rahway, New Jersey, USA and purchased from the Radiochemical Centre, Amersham, U.K. Percoll was obtained from Pharmacia. Subcellular fractionation Preliminary experiments indicated that a satisfactory separation of endometrial homogenates could be achieved in Percoll density gradients. The experiments were carried out using pre-formed linear gradients, for they gave more reproducible separations than self-generating density gradients. Three groups of uterine samples were examined. Correspondence P.V.P., Dept.Anatomy, Univ.Otago, PO Box 913 Dunedin NZ

SEPTEMBER

1983 VOL. 26 NO. 3

365

PROSTAGLANDINS

Uteri from groups of 10 adult Wistar rats (body weight ZOO-250 g) at various stages of the estrous cycle were used in each experiment. The method for isolating and homogenising endometrial cells is described below. Percoll gradients were constructed in 18 ml cellulose nitrate tubes by mixing 8 ml 90% Percoll in 0.25 M sucrose - 3 mM Hepes buffer, pH7.4with8 ml 0.25 M sucrose - 3mM Hepes buffer, pH7.4 in a perspex gradient maker. Homogenates of endometrium were layered onto the gradients and centrifuged at 65,000 gav for 30 min at 4" (Beckman SW28 rotor). Gradients were unloaded by pumping out approximately 30 fractions from the bottom of the tubes. The densities of the fractions were obtained by weighing aliquots of known volume, or by using calibrated density beads (Pharmacia). Measurement of marker enzymes The following marker enzymes : alkaline phosphodiesterase, adenosine triphosphatase, succinic dehydrogenase, glucose-6-phosphatase, acid phosphatase and galactosyl transferase were measured using standard methods (5). Electron microscopy Fractions separated on Percoll gradients were ?e?&rmgTmmOOO g for 2 h at 4" and the pellets fixed in 3% glutaraldehyde in 50 mM sodium cacodylate buffer pH7.5 and post-fixed in 1% osmium tetroxide. Samples were stained --en bloc with IX uranyl acetate, dehydrated, embedded in Araldite and sections cut. Sections were examined in a Phillips 300 electron microscope after further staining with lead citrate and uranyl acetate. Indomethacin administration to uterine horns The rats were anaesthetized by intrapet-itoneal-amstration of sodium pentobarbitone, and with 6 animals within each group. a short mid-line incision was made in the abdominal wall and the uterine horns exposed. Indomethatin (320 ug, 8 uCi dissolved in 80 11 of 0.5% NaHC03, pH8.3) was injected slowly over 10 set into the lumen of each horn at the cranial end. The uterine horns were returned into the abdominal cavity and the cut edges of the abdominal wall brought together using metal clips or secured with a gauze square. Animals were kept warm for 2 h after indomethacin administration. The remaining 4 animals that did not receive the drug were included to provide additional endometrial tissue for fractionation. Uterine horns of the rats were perfused with cold 0.15 M NaCl via the abdominal aorta before they were removed, trimmed of mesometrium and fat, opened longitudinally, and the mucosal surface scraped lightly with a scalpel blade to remove the endometrium. Endometrial cells were collected in cold 0.25 M sucrose-3 mM Hepes buffer pH7.4, pelleted by centrifugation Cells at 1000 g for IO min at 4" and then washed in the same buffer. suspended in 1.5 ml 0.25 M sucrose-3 mM Hepes buffer, pH7.4. were hombgenised using 10 strokes of a tight-fitting Dounce'qlass homogeniser (Blaessiq, Rochester, USA). The homoqenate (6mq protein)was centrifuqed (600 g for 5 mTn at 4")to remove cell debris, and the-pellet rehomogenised using 1 stroke and the centrifugation step repeated. Combined supernatants were fractionated in Percoll gradients (see above).

366

SEPTEMBER

1983 VOL. 26 NO. 3

PROSTAGLANDINS Radioactivity in the fractions was measured by transferring 0.3 ml aliquots to 11 ml of a toluene-based TritonX-100 containing scintillant (6). Two groups of animals were used in separate experiments to study the uptake of ['4Cl-indomethacin. In experiments to measure the displacement of ['4C1-indomethacin, the subcellular fractions were incubated in a IOO-fold excess of unlabelled indomethacin for 30 min at 4", centrifuged at 100,000 g for 2 h at 4" and radioactivity in the supernatants measured. RESULTS Fractionation of endometrial homogenates in Percoll gradients The separationsol?fZiiZdon the pre-formed Percoll gradients were reproducible, with the density gradients retaining their linearity during the centrifugation. Marker enzymes for subcellular membranes and organelles were located in the density range 1.040 - 1.090 g/ml. Fig. 1 shows that the distribution of mitochondria was broad. The lysosomal marker used suggested the existence of heterogeneous populations of lysosomes, with the greatest contribution found at higher densities (Table I). Markers normally used for the plasma membrane were located in a narrow density range (1.040 - 1.055 g/ml). Markers for endoplasmic reticulum (glucose-6-phosphatase) and the Golgi apparatus (galactosyl transferase), were located within a narrow density range (1.055 - 1.070 g/ml). Morphological examination (Fig. 2) of the separated fractions confirmed the biochemical results, and provided additional information on the nature of the components separated in the various density positions. Thus the position where plasma membrane markers were located was characterised by the presence of smooth surfaced vesicles and free ribosomes (Table 1). The wide distribution of mitochondria was very apparent, but examination of sections from the various parts of themembranepellets showed that fractions of density 1.055 I.070 g/ml also contained vesicular membranes with increasing numbers of attached ribosomes becoming apparent as the density increased. The localisation of lysosomes at the denser region of the gradient (1.070 - 1.090 g/ml) was confirmed by examination of the morphological features of the fractions. Subcellular distribution of [14Cl-indomethacin Fig. 1 shows that at two hours after intraluminal Perfusion of rat uteri with T'4Cl-indomethacin, the subcellular fractions separated on Percoll gradients showed a bimodal distribution of radioactivity corresponding to the soluble phase and fractions where peaks of alkaline phosphodiesterase and ATP-ase activities were localised. There was only a small and varying degree of overlap between fractions containing radioactivity and those containing glucose-6-phosphatase, galactosyl transferase and acid phosphatase activities indicating that the drug had not penetrated appreciably into the Golgi-endoplasmic reticulum -

SEPTEMBER

1983 VOL. 26 NO. 3

367

PROSTAGLANDINS lysosomal networks of the cells. Assuming that the indomethacin was not degraded or modified, it was calculated that 38 ng was associated with particulate components compared to 28 ng recovered in the position of this gradient where soluble components were located.

cLiEi+

W-INKMETHACIN

I

PHOSPHATASE

30 F

GALACTOSlL TRANSFERASE

PHOSPHATASE

DENSITY

OF PERCOLL

FIG. 1 Distribution of ['4Cl-indomethacin and six marker enzymes in the fractions separated on Percoll gradients. Analytical subcellular fractionation was carried out as described in Materials and methods and frequency distribution of the components determined as in (17).

368

SEPTEMBER

1983 VOL. 26 NO. 3

PROSTAGLANDINS

Table 1

Features of subcellular fractions of uterine endometrium separated on Percoll density gradients Subcellular markers

Density range

Morphological features Soluble phase

1.030-l.O40g/ml

Free ribosomes small vesicles

1.040-l.O50g/ml

alkaline phosphodiesterase

1.050-l.O55g/ml

alkaline phosphodiesterase ATP-ase

+ Smooth vesicles Few membrane strips

1.055-l.O70g/ml

glucose-6-phosphatase galactosyl transferase

+ Smooth and ribosome coated vesicles

1.050-1.080g/ml

succinic dehydrogenase

mitochondria

1.070-l.O90g/ml

acid phosphatase

lysosomes

+

Membrane vesicles and mitochondria were located predominantly at the top and bottom respectively of the pellets examined

Further experiments were carried out to examine the disposition of the [14C]-radioactivity in pooled fractions of density 1.045-l.O60g/ml. Incubation of the fractions with IOO-fold excess unlabelled indomethatin displaced only 6% of the radioactivity. Approx 25% of the radioactivity was released into supernatants after addition of 0.1% Triton x-100. A major proportion of the indomethacin associated with the particulate fraction appears, therefore, to be tightly bound to subcellular structures. DISCUSSION The intrauterine silastic system (1,7) devised for the delivery of indomethacin offers the possibility of reducing prostaglandin-mediated side effects of the intrauterine device, such as metorrhagia and dysIt seems likely that the endometrium is the principal menorrhea. site of action of indomethacin released from an intrauterine depot. This tissue, however, is composed of a multiplicity of cell types including epithelia (luminal and glandular), cells of the connective tissue, and the endothelium. Cells of connective and vascular tissues synthesise prostaglandins (8-11) but the ability of epithelial cells to produce prostaglandins has been less well investigated although cyclooxygenase has been shown to be present in the glandular and luminal epithelium of the human uterus by an immunohistochemical method (12). To understand at the cellular level the mode of action of indomethacin on the uterus, it is necessary to know which cell types are involved and whether these are capable of interiorizing the drug. The present experiments show that indomethacin is taken up by the endometrial tissue, this being considered a prelude to a study of the extent to which individual cell types in the endometrium participateintheuptake.

SEPTEMBER

1983 VOL. 26 NO. 3

PROSTAGLANDINS The use of pre-formed Percoll gradients in this study has provided a subcellular fractionation method for endometrial homogenates which is rapid and reproducible. The subcellular components separated were identified by biochemical measurements of selected marker enzymes, and these results were confirmed by electron microscopy of pelleted membrane fractions. The subcellular fractionation results for the endometrial homogenates showed that enzymic markers for plasma membranes, mitochondria, endoplasmic reticulum, Golgi apparatus and lysosomes, although overlapping in some cases, showed characteristic distributions in the Percoll density gradient and were comparable to those obtained with cultured cells fractionated on sucrose gradients (13). Thus, markers for plasma membrane and Golgi-endoplasmic reticulum were well separated, although mitochondrial and lysosomal markers were more widely distributed. The densities of the endometrial subcellular components on Percoll gradients were similar to those reported for ileum homogenates (14). Indomethacin was administered intraluminally at a concentration known to have pronounced effects on uterine fun The ,s~E~l~~~ioa~~~v9~~~~' in reducing the implantation rate (15). located in soluble fractions and in fractions containing vesicles and free ribosomes. The radioactivity recovered in the soluble phase may correspond to free drug remaining in the uterine horns, or drug released from the soluble phase of the cell interior. However, in contrast to studies showing that rat embryo fibroblasts take up indomethacin into a soluble phase (13), the present work shows that this drug becomes tightly associated with particulate components of Although the fractions of density 1.040 - 1.065 endometrial cells. g/ml contained vesicles and ribosomes, it is likely that the drug or its metabolites were associated with vesicles which derive from the plasma membrane and its endocytic derivatives. The low degree of displacement of the particulate-associated radioactivity indicates that the drug was either located inside these vesicles, or that it was dissolved in the membranesthemselves, a possibility, on account of its high hydrophobicity. The association of the drug with endocytic vesicles may suggest that the drug was transported from the plasma membrane to intracellular membrane networks and organelles. However, the subcellular frac ionation results (Fig. 1) show that only a minor proportion of the [iiCl-radioactivity was associated with endoplasmic reticulum-Golgi-lysosome components. It is noteworthy that rat uterine tissue has receptors for melatonin (16), a compound with a similar structure to indomethacin.

In conclusion, the present results describe for the first time the subcellular fractionation of rat endometrial homogenates on Percoll density gradients and show that indomethacin administered intraluminally becomes associated with particulate components of the cells. Although the ultimate fate of indomethacin in the uterus is unknown, the results show that at 2 h after administration, the drug is associated with a vesicular membrane fraction with enzymic characteristics

370

SEPTEMBER

1983 VOL. 26 NO. 3

PROSTAGLANDINS

Electron microscopy of components in the subFig 2 nlar fractions separated on Percoll gradients. A, density 1.040 - l.D45 g/ml. 6, 1.050 - 1.055 g/ml. C, 1.065 - 1.070 g/ml. 0, 1.085 - 1.090 g/ml. Magnifications; A, 6 x 51,800; C, D x 18,750. See Table 1 for identification of components.

SEPTEMBER

1983 VOL. 26 NO. 3

371

PROSTAGLANDINS of the plasma membrane and/or its endocytic derivatives, and that only a small proportion is found in fractions containing components with enzymic characteristics of the lysosomes, endoplasmic reticulum and Golgi apparatus. Further experiments will investigate the subcellular localisation of indomethacin when other routes of administration, e.g. via the uterine vein, are used to attempt to increase the amount of drug taken up by the uterus. ACKNOWLEDGEMENTS We thank Mr. N. Flint for carrying out the galactosyl transferase assays and Miss Kate Sullivan for the electron microscopy. Part of the work was supported by the New Zealand Medical Research Council. REFERENCES Peplow, P.V. and Hurst, P.R. Influence of local release of indomethacin from a silastic device on uterine prostaglandins F2q and E. Prostaglandins leukotrienes Med. -9:437, 1982. Hurst, P.R., Peplow, P.V. and Gamble, V.M. Intrauterine delivery of indomethacin from a silastic device : effect on the preimplantation embryo and implantation. Contraception -26:97, 1982. Lundkvist, 0. and Nilsson, B.O. Ultrastructural changes of the trophoblast-epithelial complex in mice subjected to implantation blocking treatment with indomethacin. Biol. Reprod. -22:719, 1980. Peleg, S. and Lindner, H.R. The effect of prostaglandins on progestin receptor translocation and on the decidual cell reaction in vivo and in vitro. Endocrinol. -110:1647, 1982. _Evans, W.H. Preparation and characterisation of mammalian plasma membranes. Elsevier/North Holland, Amsterdam 1978. Evans, W.H. and GuSd, J.W. Biosynthesis of liver membranes. Incorporation of [ HI-leucine into proteins and of Ci4C]-glucosamine into proteins and lipids of liver microsomal and plasma membrane fractions. Biochem. J. -125:615, 1971. Peplow, P.V. and Hurst, P.R. An intrauterine silastic system for the sustained release of indomethacin. Prostaglandins and Medicine 7:563, 1981. Myatt, L., Bray, M.A., Gordon, D and Morley, J. Macrophages on intrauterine contraceptive devices produce prostaglandins. Nature -257~227, 1975. Bockman, R.S. Prostaglandin production by humanbloodmonocytes in vitro and mouse peritoneal macrophages : synthesis dependant on -culture conditions. Prostaglandins -21:9, 1981.

372

SEPTEMBER

1983 VOL. 26 NO. 3

PROSTAGLANDINS 10

Goldsmith, J.C. and Needleman, S.W. A comparative study of thromboxane and prostacyclin release from ex vivo and cultured bovine vascular endothelium. ProstaglandiiiF2iF-l-73, 1982. -

11

Moncada, S., Herman, A-G., Higgs, E.A. and Vane, J.R. Differential formation of prostacyclin (PGX or PG12) by layers of the arterial wall. An explanation for the anti-thrombotic properties of vascular endothelium. Thromb. Res. -11:323, 1977

12

Rees, M.C.P., Perry, D.M., Anderson, A.B.M. and Turnbull, A.C. Immunohistochemical localization of cyclooxygenase in the human uterus. Prostaglandins -23:207, 1982.

13

Blanchard, J.C., Schneider, Y.J., Tulkens, P. and Trouet, A. Cellular uptake and subcellular distribution of two non-steroidal anti-inflammatory drugs : indomethacin and ketoprofen. Arch. internat. Pharmacodynamie Ther. 240:4, 1979.

14

Jenkins, W.J. Empson, R., Jewell, D.P. and Taylor, K.B. Subcellular localization of vitamin B during absorption in the guinea-pig ileum. Gut 22:617, 1981.

15

Gupta, U., Malhotra, N., Varma, S.K. and Chaudhury, R.R. Effect of intrauterine administration of antiprostaglandin drugs on implantation in the rat. Contraception -24:283, 1981.

16

Cohen, M., Roselle, D., Chabner, B., Schmidt, T.J. and Lippman, M. Evidence for a cytoplasmic melatonin receptor. Nature 274:894, 1978.

17

i33aufay,H. and Amar-Costesec, A. Cell fractionation techniques : Methods in Membrane Biology (ed E.D. Korn) Plenum, New York, P.l, 1976. Editor: Harold R. Behrman Received: 2-21-83 Accepted: 5-24-83

SEPTEMBER

1983 VOL. 26 NO. 3

373