Retinoid binding lipoprotein in neoplastic cells

41

Cancer Letters, 22 (1984) 41-47 Elsevier Scientific Publishers Ireland Ltd.

RETINOID

BINDING

LIPOPROTEIN

IN NEOPLASTIC

CELLS

D. SKLAN and R. LOTAN Faculty of Agriculture, Rehouot tute, Rehouot 76-l 00 (Israel)

76-100 and Department

of Biophysics,

Weizmann

Insti-

(Received 26 September 1983) (Accepted 1 December 1983)

SUMMARY

A lip:id-protein aggregate (LPA), binding both retinol and retinoic acid, was detected in the cytosols of several human and murine tumor cells by gel filtration after incubation of cell extracts with labelled retinoids. In addition to this high M, lipoprotein (M, 2 X 106) the cytosols also contained low M, retinoid binding proteins (M, 14,500). The amounts of retinoids bound to these 2 distinct components were similar and both bound some 40-200-fold more retinol than retinoic acid. Retinyl palmitate hydrolase, copper and zinc were found in association with LPA. The LPA may be involved in subcellular compartmentalization, transport and function of retinoids.

INTRODUCTION

The presence of the cytosolic lipoprotein containing retinol and retinyl esters as well as retinyl palmitate hydrolase has been recently described [1,2]. This LPA was found to contain about 50% lipid, and exhibited a hydrated density in the range 1.052-1.13, and in addition it co-eluted on gel filtration with zinc and copper [ 1,2]. It was suggested that LPA may possess a function in intracellular retinol transport [ 1,2]. An additional specific intracellular retinol binding protein (cRBP), with a Mr of 14,500, identified in many tissues [3] has been suggested to mediate biological effects of retinol [ 31. In liver, cRBP was found to be associated with the lipoprotein [ 1,2]. One of the major active metabolites of retinol is the acid -Address all correspondence Rehovot 76-100, Israel.

to: Dr. D. Sklan, Faculty of Agriculture, P.O. Box 12,

o 1984 Elsevier Scientific Publishers Ireland Ltd. 0304-3835/84/$03.00 Published and Printed in Ireland

42

form, retinoic acid [4], which is partially found in a complex with a specific ~tracellular binding prote~~RABP [ 3 ] . This binding protein is very similar to cRBP [3]. While most studies relating to intracellular binding and transport of retinoids have focused on the low IMrretinoid binding proteins, little attention has been devoted to the possible role of the LPA in vitamin A function. In recent years there has been increasing interest in the relation between vitamin A and cancer, due to various reports on the ability of vitamin A and analogs (retinoids) to inhibit carcinogenesis [ 5,6] and supress the growth of some tumor cells in vivo and in culture [6]. Although the mechanism of these effects has not yet been elucidated, it is thought that the intracellular retinoid binding proteins are intimately involved [3,6]. While it is conceivable that the LPA may also play a role in mediation of the anti-neoplastic effects of retinoids, this possibility has not been studied to date, and constituted the object of the present study. MATERIALS AND METHODS

Materials [ 15-14C]Retinol (16 pCi/pmol)

and [11,12-3H]retinoic nmol) were kind gifts of Hoffmann-LaRoche, Basel.

acid (2.5 ,uCi/

Cells

The human cervical carcinoma HeLaS3 [ 71, the human osteosarcoma I-Is791 IS], the murine melanoma S91 C-2, a retinoic acid sensitive clone [ 91 and its retinoic acid resistant mutant derivative S91 C-l 54 [lo] and the murine lymphosarcoma RAWl17P [ 111 were cultured as described previously. ~epara tion of cy tosol

Subconfluent cultures were harvested with 2 mM EDTA and the cells were washed twice with PBS and suspended in 50 mM Tris-HCl buffer (pH 8.0) at approximately lo8 cells/ml. Homogenisation was carried out at 4’C with 50 strokes of a motor-driven teflon-glass homogeniser (speed 6, Heidolph) or sonicated for 3 X 10 s at setting 3 in a model W-375 Sonicator Cell Disruptor (Heat Systems, Ultrasonic Inc., New York), followed by cent~~gation at 105,000 Xg for 60 min. The resulting clear supernatant is termed cytosol. Ligand binding

Incubation of cytosol samples with labelled or unlabelled retinoids was carried out at 4°C in the dark overnight. Analysis of retinoid binding proteins

Gel filtration was carried out on columns of Sepharose 4-B, 6-B and Sephadex G-50 (Pharmacia, Uppsala), eluting with 50 mM Tris-HCl con-

43

taming 25 mM mercaptoethanol. Columns were calibrated using protein standards [ 21. High performance liquid chromatography (HPLC) was performed using a TSK 3000 gel permeation column (7.5 X 300 mm) eluting with 0.1 M phosphate buffer containing 0.1 M NaCl, or with a C-18 reversed phase column (Partisil, ODS 2) eluting with 8% water in methanol. Ultracentrifugation was carried out in different sucrose densities for 24 h at 105,000 X g in the Ti 50 rotor of a Beckman L3-50 centrifuge. Chemical analyses of LPA Lipid.s were extracted and the lipid content determined as previously described [ 21. Retinyl palmitate hydrolase activity was determined as described by Prystowsky et al. [ 121. Concentrations of zinc and copper were determined by atomic absorption spectroscopy using a Perkin Elmer 2380. RESULTS

Chromatographic analysis of the cytosol of cells incubated with labelled retinol and retinoic acid is shown in Figs. 1 and 2. In all cells tested a significant peak of both 3H and 14Celuted close to the void volume of the Sepharose 6-B column, and an additional included peak of label was detected at a M, of approximately 13,000-15,000. The amounts of retinol and retinoic acid bound is variable between the peaks and within each peak. In

20

30

40

50

60

70

ELUTION VOLUME (ml)

60 ELUTION VOLUME (ml)

Fig. 1. Gel chromatography of Hs791 (A,B) and HeLa S3 (C,D) cytosols on Sepharose 6-B follo8wing incubation with labelled retinol (0) and retinoic acid (0). Cytosol (0.6 ml, 5-10 mg protein) was loaded onto the column (45 X 1.7 cm) and ‘H and “C were determined in the eluent (B and D). In addition copper (A) and zinc (A) were determined in each fraction (AC).

44

d

f25 c

20

30

40

50

60

70

ELUTION VOLUME(ml)

80

20

30

40 50 60 70 80 ELUTION VOLUME(ml)

Fig. 2. Gel chromatography of S91 C-2 (A,B) and RAWl17P (C,D) cytosols on Sepharose 6-B following incubation with labelled retinol (0) and retinoic acid (0). Other details as in Fig. 1.

the LPA, binding of retinol was greater than that of retinoic acid by a factor of between 40 and 200 for the different cell types. A similar trend was observed in the binding of retinoids to the low M, binding proteins. With the exception of the S91 C-2 cells the amounts of retinoids bound by both LPA and the low M, binding proteins were comparable. As previously described in other tissues [ 1,2] zinc and copper co-eluted with the LPA peak [ 21. The elution profile of a retinoic acid resistant mutant of S91 melanoma cells (C-154) was essentially identical with that of S91 C-2 (data not shown). In addition, growth of S91 cells in the presence of retinoic acid (10 FM) for 5 days prior to cytosol preparation did not alter the pattern of retinoid binding. Similar elution patterns were obtained when cytosols were prepared by homogenization or by sonication. The material eluting at 13,000-15,000 daltons was concentrated and re-chromatographed on Sephadex G-50 column (60 X 1.5 cm) before or after incubation with a lOOO-fold excess of unlabelled retinoid. This procedure resulted in a single peak of both ‘H and 14C at a M, of 14,000-15,000; this peak diminished specifically by some 80% following incubation with the respective unlabelled ligand. A peak with similar chromatographic and binding properties was dissociated from LPA after 6 M urea treatment. Both LPA and the low M, proteins bound material which cochromatographed with retinol on HPLC following extraction (data not shown). Rechromatography of the LPA isolated from the Sepharose 6-B column on a Sepharose 4-B column (55 X 2 cm) resulted in elution of the LPA well within the included volume and very close to the elution volume of blue dextran, this suggests a M, close to 2 X 106. Extraction of this LPA fraction showed the

45

CELL LINE

Fig. 3. Retinyl palmitate hydrolase activity in the LPA isolated from tumor cell cytosols. Activity was determined by measuring the rate of release of labelled palmitic acid from retinyl-[1-“Cl palmitate [12].

presence of 50.7% lipids. Ultracentrifugation revealed label in the fraction floating in the density range 1.06-l .ll. Retinyl palmitate hydrolase activity was detected in LPA from all cells (Fig. 3). The specific activities varied considerably between cell types. DISCUSSION

In the present report we have shown that several neoplastic cells of different histopathological origin, from human and murine tumors, contain a cytosolic lipid-protein aggregate capable of binding retinoids. With this LPA are associated the low M, retinoid binding proteins, retinyl palmitate hydrolase, zinc and copper. The LPA found in tumor cells was very similar both in these characteristics and in physicochemical properties to that found in normal tissues [ 1,2]. Little has been reported concerning the intracellular compartmentalization of retinoids before and after metabolism. As retinoids are lipophilic, they are found dissolved in cellular membranes [ 131. However, the site of action of retinoids is by no means confined to the membrane, in view of the many reports of retinoid effects on gene expression [ 14-171. The recent demonskations of translocation to the nuclei of various cells of retinoids bound to low Mr specific binding proteins [ 18-201 suggest that these binding proteins play a role in transport between intracellular organelles. Our demonstration of the co-existance of the LPA together with the low Mr binding proteins, and the finding that its retinoid binding capacity is similar to that of cRBP and cRABP raises the possibility that LPA is involved in sub-cellular retinoid compartmentalization and transport. The mechanism by which retinoids inhibit the proliferation and suppress the exprlession of the transformed phenotype of certain tumor cells has not yet been clarified. Since the modulations of these fundamental properties

46

of neoplastic cells may be the result of retinoid induced changes in gene expression, it has been proposed that cRBP and cRABP mediate these effects [3,6,21]. However, the presence of these low M, retinoid binding proteins could not be correlated with susceptibility or resistance to the growth inhibitory action of retinoids [10,22]. The tumor cells examined in this study, with the exception of S91 C-154, are susceptible to growth inhibitory action of retinoids [ 7-101. All these cells contain both LPA and the low M, binding proteins, and this has not been previously detected in the studies concerning the role of binding proteins. It thus appears important that the role of the LPA in mediation of retinoid action on neoplastic cells should be evaluated. REFERENCES 1 Sklan, D., Blaner, W., Adachi, N., Smith, J.E. and Goodman, D.S. (1982) Association of cellular retinol binding protein and several lipid hydrolase activities with a vitamin A containing high molecular weight lipid-protein aggregate from rat liver cytosol. Arch. Biochem. Biophys., 214, 35-44. 2 Sklan, D. and Donoghue, S. (1982) Association of acylglyceride and retinyl palmitate hydrolase activities with zinc and copper metalloproteins in a high molecular weight lipid-protein aggregate fraction from chick liver cytosol. Biochim. Biophys. Acta, 711, 532-538. 3 Chytil, F. and Ong, D.E. (1979) Cellular retinol- and retinoic acid binding proteins in vitamin A action. Fed. Proc., 38,2510-2514. 4 Roberts, A.B. and Frolik, C.A. (1979) Recent advances in the in vivo and in vitro metabolism of retinoic acid. Fed. Proc., 38,2524-2527. 5 Sporn, M.B. and Newton, D.L. (1979) Chemoprevention of cancer with retinoids. Fed. Proc., 38,2528-2534. 6 Lotan, R. (1980) Effects of vitamin A and its analogs (retinoids) on normal and neoplastic cells. Biochim. Biophys. Acta, 605,33+X. 7 Lotan, R., Kramer, R.H., Neumann, G., Lotan, D. and Nicolson, G.L. (1980) Retinoic acid induced modifications in the growth and cell surface components of a human carcinoma (HeLa) cell line. Exp. Cell Res., 130,401-414. 8 Thein, R. and Lotan, R. (1982) Sensitivity of cultured human osteosarcoma and chondrosarcoma cells to retinoic acid. Cancer Res., 42,4771-4775. 9 Lotan, R., Giotta, G., Nork, E. and Nicolson, G.L. (1978) Characterization of the inhibitory effects of retinoids on the in vitro growth of two malignant murine melanomas. J. Natl. Cancer Inst., 60,1035-1041. 10 Lotan, R., Stolarsky, T. and Lotan, D. (1983) Isolation and analysis of melanoma cell mutants resistant to the antiproliferative action of retinoic acid. Cancer Res., 43, 2868-2875. 11 Brunson, K.W. and Nicolson, G.L. (1978) Selection and biologic properties of malignant variants of a murine lymphosarcoma. J. Natl. Cancer Inst., 61,149s1504. 12 Prystowsky, J.H., Smith, J.E. and Goodman, D.S. (1981) Retinyl palmitate hydrolase activity in normal rat liver. J. Biol. Chem., 256,4498-4503. 13 Mack, J.P., Lui, N.S.J., Roels, 0-A. and Anderson, O.R. (1972) The occurrence of vitamin A in biological membranes. Biochim. Biophys. Acta, 288, 203-219. 14 Strickland, S., Smith, K.K. and Marotti, K.R. (1980) Hormonal induction of differentiation in teratocarcinoma stem cells: generation of parietal endoderm by retinoic acid and dibutyryl CAMP. Cell, 21,347-355.

47 15 Linder, S., Krondahl, U., Sennerstam, R. and Ringer& N.R. (1981) Retinoic acid induced differentiation of F9 embryonal carcinoma cells. Exp. Cell Res., 132,453460. 16 Fuchs, E. and Green, H. (1981) Regulation of terminal differentiation of cultured human keratinocytes by vitamin A. Cell, 25.617-625. 17 Omori, M. and Chytil, F. (1982) Mechanism of vitamin A action: gene expression in retinol deficient rats. J. Biol. Chem., 267.14370-14374. 18 Wiggert, B., Russell, P., Lewis, M. and Chader, G. (1977) Differential binding to soluble nuclear receptors and effects on cell viability of retinol and retinoic acid in cultured retinoblastoma cells. Biochem. Biophys. Res. Commun., 79, 218-225. 19 Jetten, A. and Jetten, M. (1979) Possible role of retinoic acid binding protein in retinoid stimulation of embryonal carcinoma cell differentiation. Nature, 278, 180-1182. 20 Takase, S., Ong, D.E. and Chytil, F. (1979) Cellular retinol binding protein allows specific interaction of retinol with the nucleus in vitro. Proc. Natl. Acad. Sci. U.S.A., 76,2204-2208. 21 Schindler, J., Matthaei, K.I. and Sherman, M.I. (1981) Isolation and characterisation of mouse mutant embryonal carcinoma cells which fail to differentiate in response to retinoic acid. Proc. Natl. Acad. Sci. U.S.A., 78,1077-1080. 22 Lotan, R., Ong, D.E. and Chytil, F. (1980) Comparison of the level of cellular retinoid binding proteins and susceptibility to retinoid induced growth inhibition of various neoplastic cell lines. J. Natl. Cancer Inst., 64, 1259-1262.