Communication THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 265, No. 31, Issue of November 5, pp. 18753-18756,199O

0 1990 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S. A.

Progesterone Photoaffinity Labels P-glycoprotein in Multidrug- resistant Human Leukemic Lymphoblasts*

(Received for publication, April 2, 1990)

Xiao-dong Qian$ and William T. Beck@7

From the jDepartment of Biochemical and Clinical Pharmacology, St. Jude Children’s Research Hospital, Memphis, Tennessee 38101 and the SDepartment of Pharmacology, College of Medicine, University of Tennessee, Memphis, Tennessee 38163

We show for the first time that [3H]progesterone ([3H] PRG) can directly photoaffinity label membrane pro- teins prepared from a multidrug-resistant human leu- kemic lymphoblastic cell line CEM/VLBS~. A 170-kDa protein in CEM/VLBS~ cell membranes was specifically labeled by [3H]PRG, which we identified as P-glyco- protein (Pgp) by immunoprecipitation with monoclonal antibody C219. The anticancer drug vinblastine and multidrug resistance reversing agent verapamil as well as several steroidal hormones were examined for their ability to interfere with [3H]PRG binding to Pgp. We found that 200-fold molar excess of vinblastine strongly inhibited (93%) the binding of [3HlPRG to Pgp compared with verapamil (SO%), progesterone (78%), testosterone (46%), dexamethasone (25%), and aldos- terone (56%). The results of this study provide direct evidence that progesterone can bind to Pgp and support the hypothesis that under physiological conditions Pgp may play a role in the excretion of progesterone from certain cells. Importantly, our results show that under our conditions vinblastine and verapamil are better able to compete with [3H]PRG for binding to Pgp than are other steroids, including testosterone, corticoster- oids, and mineralocorticoids.

P-glycoprotein (Pgp)’ is a 170-kDa protein originally found to be overexpressed in plasma membranes of MDR cancer cells (1,2). A cDNA has been cloned from human and mouse MDR cell lines (3, 4), and it was found to encode Pgp, a protein with 12 transmembrane domains and two nucleotide binding sites having homology with ATP binding sites in

*This work was supported by Research Grant CA 40570 and Cancer Center Support (CORE) Grant CA 21756 from the National Cancer Institute, Department of Health and Human Services, and in part by American Lebanese Syrian Associated Charities. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

7 To whom correspondence should be addressed: Dept. of Biochem- ical and Clinical Pharmacology, St. Jude Children’s Research Hos- pital, 332 N. Lauderdale, Memphis, TN 38101.

’ The abbreviations used are: Pgp, P-glycoprotein; MDR, multi- drug-resistant; ALD, aldosterone, DEX, dexamethasone; PRG, pro- gesterone; TST, testosterone; EP, p-estradiol; VLB, vinblastine; VRP, verapamil; SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

other proteins. A variety of photoactive analogs of anticancer drugs (VLB (5) and colchicine (6)) and MDR reversing agents (VRP (7-9) and azidopine (10)) has been shown to bind specifically to this protein. These results led to the hypothesis that a function of Pgp is to bind intracellular anticancer drugs and extrude them to the extracellular space, resulting in a decreased cytotoxic drug concentration in the cancer cells. The energy required for this active efflux function has been postulated to come from hydrolysis of ATP, since a photoac- tive analog of ATP can bind to Pgp (11) and purified Pgp displayed M$+-dependent ATPase activity (12).

High levels of Pgp have also been found in the membranes of some normal human and mouse tissues such as liver, pancreas, kidney, intestine, adrenal cortex (13), placenta (14), uterine epithelium (15), and capillary endothelial cells of the blood-brain barrier (16). Although its physiological function is not clear, the existence of Pgp in these normal tissues has led to the speculation that this protein may play an important role in the secretory functions of these tissues. Yang et al. (17) and Naito et al. (18) demonstrated that a steroidal hormone, PRG, enhanced vinblastine accumulation and cy- totoxicity in MDR cells. Based on their demonstrations that PRG inhibited the binding of [3H]azidopine to Pgp, these investigators postulated that PRG reversed multidrug resist- ance by interacting with Pgp in these cells. However, no direct evidence was presented to support this hypothesis.

Earlier studies showed that the a&unsaturated ketone structure in PRG is inherently photoactive (19). Synthetic progestin R5020 has been used to photolabel progestin recep- tors in chicken oviduct (20) and human breast cancer cells (21). In the present study, we found that [3H]PRG, at a concentration comparable with that of plasma levels in later term pregnancy (22), will also photolabel Pgp, thus providing direct physical evidence that PRG interacts with Pgp. Our results suggest that Pgp may have physiological importance in the transport of PRG across the uterine epithelium during pregnancy. Our data also indicate that PRG can be used as a new reagent to study the normal physiological function of &P.

EXPERIMENTAL PROCEDURES

Materials-[3H]PRG, specific activity 144 Ci/mmol, was purchased from Amersham Corp. [‘H]PRG with lower specific activity (24 Ci/ mmol) was prepared by adding an appropriate amount of nonradioac- tive PRG to radioactive PRG solution. VRP, PRG, TST, DEX, E2, and ALD were purchased from Sigma. VLB was provided by Dr. Homer Pearce, Lilly. Monoclonal antibody C-219 was obtained from Centocor, Inc. (Malvern, PA). Other chemicals and supplies were of the highest commercial grade available.

Cell Culture-CCRF-CEM human leukemic lymphoblasts (CEM) and the MDR subline, CEM/VLBsx, were grown as described earlier (23). CEM/VLB5~ was selected for 2636.fold resistance to VLB, as described (9).

Plasma Membrane Protein Preparation and Photoaffinity Label- ing-preparation of membrane proteins from CEM and CEM/VLBsK cells was done as described previously (9). Membrane proteins (200 pg) were incubated in buffer containing 250 mM sucrose and 10 mM Tris-HCl, pH 7.4, at 25 “C in the dark for 30 min with 1 pM [‘HI PRG (specific activity 144 Ci/mmol, unless otherwise specified) in the absence or presence of various competitors in a total volume of 150 ~1. The incubation mixture was then irradiated with a germicidal UV light (General Electric Germicidal lights, G30T8, 30 watts), commonly used in laminar flow hoods, for 30 min at a distance of 10 cm.

18753

by guest on April 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

18754 Progesterone Photoaffinity Labeling of P-glycoprotein

Immunoprecipitation, Gel Electrophoresis, and Fluorography-Im- munoprecipitation was performed as described earlier (9). Photola- beled and/or immunoprecipitated proteins were separated by one- dimensional 515% SDS-PAGE, and the radioactivities associated with proteins were detected by fluorography as described previously (9).

RESULTS AND DISCUSSION

Photoaffinity Labeling Pgp with PRG--To study the inter- action between Pgp and PRG, we photoaffinity labeled mem- brane proteins prepared from the drug-resistant human leu- kemic cell line, CEM/VLBeK, with decreasing concentrations of [ ‘H]PRG, ranging from 20 fiM to 80 nM. The photoaffinity- labeled membrane proteins were analyzed by SDS-PAGE and fluorography. The fluorograph in Fig. 1 shows that [“HIPRG photoaffinity labels a 170-kDa membrane protein prepared from CEM/VLB5k cells. Several proteins of smaller molecular weight were also found to be labeled by [“HIPRG. The 170- kDa protein was found to be present only in the CEM/VLBsk membranes and not in the CEM cell membranes (data not shown). At [“HIPRG concentrations >8 pM, nonspecific label- ing of other membrane proteins was seen. However, relative to these other proteins, it appeared that specific labeling of Pgp with [“H]PRG at concentrations higher than 8 FM was

not significantly increased, suggesting that saturation of Pgp binding sites occurred around this concentration. The amount of the 170-kDa protein labeled by PRG (or detected by fluo- rography) decreased markedly as the concentration of [“HI PRG was decreased.

We incubated membranes with 4 pM [‘H]PRG for different times between 20 min and 4 h, exposed them to UV light as before, and then analyzed for the amount of labeled 170-kDa protein. We found that the reaction was essentially complete by 20 min, and no further labeling of the 170-kDa protein occurred. We conclude, therefore, that under the 30-min incubation conditions of our experiments, we are measuring binding of [“HIPRG to the 170-kDa protein under equilibrium conditions.

It has been shown that plasma PRG levels during late pregnancy can reach 0.3-0.5 WM (22). Since the major syn- thetic site of PRG is the placenta (22), its local concentration in the uterus during pregnancy is likely to be higher than in the plasma. We found that the labeling of Pgp by PRG can be detected at 0.1 pM by fluorography when [“HIPRG with higher specific activity (144 Ci/mmol) was used. These data

200 -

10 116-

b 97-

42-

1 RIP

16 8 4 0.8 0.4 0.08 I.IM

CPRGI

FIG. 1. Photoaffinity labeling of plasma membrane proteins from CEM/VLBI-,~ cell lines with 13H]PRG. Cell membrane frac- tions were prepared according to methods previously described (9). CEM/VLB,K membrane uroteins were nhotoaffinitv-labeled with de- creasing concentration (20 FM to 80 nM) of [“H]PRG (specific activity 24 Ci/mmol), separated by SDS-PAGE, and detected by fluorography. See “Experimental Procedures” for experimental details.

strongly suggest that PRG, at concentrations present in the body under normal physiological conditions, is capable of binding to Pgp.

Recently, several photoaffinity labeling agents have been identified that specifically bind to Pgp and have been used to study its function (5-9). PRG has the advantage among other photoaffmity labeling agents in that it is an inherently pho- toactive agent, and structural modification, such as attach- ment of azido (9) or azidobenzoyl(5-8) groups, is not required for its covalent binding to protein under UV light.

Immunological Identification of PRG-labeled 170-kDa Pro- tein as Pgp-To identify the 170-kDa protein labeled by [3H] PRG, a monoclonal antibody C-219, which specifically rec- ognizes Pgp (24), was used to immunoprecipitate the CEM/ VLBsk membrane proteins labeled by [“HIPRG. Fig. 2 dem- onstrates that the [“HIPRG-labeled 170-kDa protein in CEM/ VLB,k was specifically immunoprecipitated by antibody C- 219. This protein was not detected in the immunoprecipitates when nonimmune mouse serum was used. No radioactivity was detected on the fluorograph at the 170-kDa region in the CEM membrane proteins labeled by [3H]PRG, although sev- eral other proteins with lower molecular weight were labeled in a pattern that is similar to that of CEM/VLB5x membrane preparations. These results indicate clearly that the 170-kDa protein labeled by [“HIPRG is Pgp.

Effect of Nonradioactive PRG, VLB, and VRP on the Bind- ing of [“HIPRG to Pgp-To demonstrate that PRG specifi- cally binds to Pgp, we used nonradioactive PRG to inhibit its binding. We also examined whether a cytotoxic drug, VLB, and an MDR reversing agent, VRP, could compete with [3H] PRG for labeling of Pgp. Fig. 3 shows that nonradioactive PRG, VLB, and VRP all strongly inhibit the binding of [3H] PRG to Pgp, while their effect on the labeling of other proteins by [“HIPRG is negligible. These results suggest that the binding of PRG to Pgp is specific and that the binding site(s) for PRG on Pgp is saturable. Of considerable interest, this experiment demonstrates that although VLB and VRP ap- parently share little or no structural similarity with PRG, the inhibitory effects of VRP (80%) and VLB (93%) are the same or greater than that of PRG (78%). We cannot determine

C219 N.S.

- 200

(I)- - 116 -97 2

-66 z -. b

-43

L.-l- CEM CEM/VLBJK

FIG. 2. Immunoprecipitation of [‘H]PRG-labeled Pgp with C-219 monoclonal antibody. Membrane proteins (200 ~gj nre- pared from CEM and CEM/VLB5s cells were photoaffinity-iabileded with 1 pM [“HIPRG, solubilized in sodium deoxycholate buffer, and then incubated with 50 pg of normal mouse serum (N.S.) or C-219 monoclonal antibody (C219). Immunocomplexes were precipitated with protein A-Sepharose. Proteins were separated by SDS-PAGE and detected by fluorography. Solubilized membrane samples from CEM and CEM/VLB5~ were electrophoresed without antibody treat- ment as control. See “Experimental Procedures” for experimental details.

by guest on April 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Progesterone Photoaffinity Labeling of P-glycoprotein

-200

-I16 2 -97 x

-66 6 L

CEM , I

CEM/VLB5K

FIG. 3. Inhibitory effect of progesterone, vinblastine, and verapamil on photoaffinity labeling Pgp by [3H]PRG. Mem- brane proteins prepared from CEM or CEM/VLBsK cells were photo- affinity-labeled by 1 pM [“H]PRG in the presence of 200 pM nonra- dioactive PRG, \iLB, or -VKP. The labeled proteins were separated bv SDS-PAGE and detected bv fluoroaranhv. The extent of inhibition (‘r, Inhib.) of [“H]PRG labeling of P& b; nonradioactive PRG, VLB, and VRP was determined by scanning the film with a densitometer. See “Experimental Procedures” for details.

from these results whether the inhibition of [3H]PRG binding to Pgp by VLB and VRP is due to either an allosteric effect or a direct competition. Further experiments are needed to clarify the competition patterns among the various agents shown to bind to Pgp. Nevertheless, the results of this exper- iment correlate well with the data from the earlier studies by Yang et al. (17), showing that 1000-fold molar excess of PRG (100 pM) was required to inhibit 69% of the binding of VLB (100 nM) to membrane vesicles from MDR cells (17).

The high efficacy of VLB in inhibiting the binding of other photoaffinity labeling agents to Pgp has also been demon- strated in several recent studies. Photoaffinity labeling of Pgp by azidopine (10) and azido derivatives of colchicine (6) and verapamil (7-9) is specifically and more strongly inhibited by VLB than by other competing agents, although these photo- affinity compounds do not have apparent structural similari- ties.

Effects of Steroidal Hormones on the Binding of PRG to Pgp-To begin to study the structural requirements for PRG binding to Pgp, we examined the effects of several steroidal hormones on the labeling of Pgp by [“HIPRG (Fig. 4). The inhibitory effects of a 200-fold molar excess of the steroid hormones on the binding of PRG to Pgp were: PRG, 82%; TST, 46%; DEX, 25%; and ALD, 56%. It is of considerable interest that the steroids used in this experiment, TST, DEX, and ALD, had less of an effect on the binding of [3H]PRG to Pgp than that of VLB and VRP (Fig. 3), although these steroidal compounds are more closely related to the structure of PRG than is VLB.

We attempted to determine whether E2 also inhibited [3H] PRG binding to Pgp. We found that it did, but in contrast to the other competitors, we could not completely solubilize EP. This led us to ask whether physical filtering of incident UV light caused an apparent inhibition of [3H]PRG labeling of Pgp. To test this, we exposed [“H]PRG to UV light in the presence of 200-fold molar excesses of PRG, VLB, TST, DEX, and ALD but in the absence of membrane proteins. The photolyzed [“H]PRG was then chromatographed on a silica gel TLC plate, developed in hexane/ethyl acetate (3:1), and the reaction products were analyzed by autoradiography. The autoradiograms of the plate showed that [3H]PRG was com- pletely photolyzed whether or not 200-fold molar excesses of

-200

/ II I

CEM CEM/VL&,

FIG. 4. Effect of steroidal hormones on photoaffinity label- ing Pgp by [“HlPRG. Membrane proteins prepared from CEM/ VLBsK-cells were photoaffinity-labeled by ~-PM [“H]PRG in the nresence of 200 uM nonradioactive PRG. TST. DEX. or ALD. The labeled proteins’ were separated by Sk-PAGE and detected by fluorography. Membrane proteins from CEM/VLB,,,, labeled with [ ‘H]PRG in the absence of competitors, were used as controls. The extent of inhibition (% Inhib ) of nonradioactive PRG, TST, DEX, and ALD on Pgp labeling by [ ‘H]PRG was determined by scanning the film with a densitometer. See “Experimental Procedures” for details.

nonradioactive competitors were present. These results indi- cated that the inhibitions of [“H]PRG labeling of Pgp by these compounds were true competitions and not apparent inhibi- tions due to physical filtering of the UV light.

We also asked whether another steroid with the a,/?-unsat- urated ketone group can also photoaffinity label Pgp. Accord- ingly, we attempted to photolabel Pgp in MDR cell mem- branes with [3H]hydrocortisone. We found that, under the same experimental conditions and at the same concentration range as those used for PRG, [“Hlhydrocortisone did not label Pgp (data not shown). This result indicates that the cu,fl- unsaturated ketone functionality at the 3-position of PRG is essential but not sufficient for its binding to Pgp. It also indicates that the interaction between Pgp and PRG may be of a hydrophobic nature, since the three hydroxy groups at the ll& 17a, and 21 position of hydrocortisone essentially abolished its binding to Pgp. Clearly, although Pgp can rec- ognize a variety of structurally unrelated compounds, it still requires certain structural features for these compounds to bind to Pgp (25, 26).

Summary and Conclusions-The 170-kDa Pgp has been shown to be expressed at an elevated level in MDR cancer cells and in some normal tissues, including the gravid uterus. Although it has been demonstrated convincingly that Pgp plays a pivotal role in MDR cancer cells for their resistance to a variety of cytotoxic agents (l-4), its function in normal tissues and under normal physiological conditions is not clear. In the present study, we have shown that a steroidal hormone, PRG, can specifically bind to Pgp. This result provides ex- perimental support for the hypothesis that Pgp may be im- portant in normal cell physiology. Further, we showed here that the binding of PRG to Pgp can be inhibited specifically by VLB, a cytotoxic agent, as well as by an MDR reversing agent, VRP, although they appear to share little or no struc- tural similarity with PRG. (We have not yet compared their three-dimensional structures, however.) These results are consistent with the “broad specificity” of Pgp for a variety of cytotoxic agents. These results also suggest a possible basis for the observations that VRP retards fetal growth during early gestation and is also embryocidal (27).

by guest on April 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

18756 Progesterone Photoaffinity

Finally, it is of considerable interest that the structurally more related steroidal hormones, such as TST and DEX, were less able to inhibit the binding of PRG to Pgp than were VLB and VRP. The diminished effects of these steroids could be a function of lower binding affinities for Pgp, binding to differ- ent site(s) on Pgp, or largely nonspecific binding. This enigma clearly warrants further extensive and systematic study. To the best of our knowledge, the binding of so many structurally unrelated chemicals to Pgp has not been shown for other receptors or membrane transporters. We do note, however, that the cytochrome P-450 system, which is also involved in drug and xenobiotic detoxification, also recognizes many di- verse chemical structures. Indeed, it has been proposed that both Pgp and cytochrome P-450 systems subserve similar complementary functions in the body (28, 29).

Achnowledgmenb-We are grateful to Danette Harris and Bobbi Wysong in the Biochemical Communications Department for excel- lent preparation of the artwork. We thank Vicki Gray for her excellent word processing efforts in preparing this manuscript.

REFERENCES

1. 2.

3.

4. 5.

6.

Beck, W. T. (1987) Biochem. Pharmacol. 36, 2879-2887 Juranka, P. F., Zastawany, R. L., and Ling, V. (1989) FASEB J.

3,2583-2592 Chen, C.-j., Chin, J. E., Ueda, K., Clark, D. P., Pastan, I.,

Gottesman, M. M., and Roninson, I. B. (1986) Cell 47, 381- 389

Gros, P., Croop, J., and Housman, D. (1986) Cell 47, 371-389 Safa, A. R., Glover, C. J., Meyers, M. B., Biedler, J. L., and

Felsted, R. L. (1986) J. Biol. Chem. 261, 6137-6140 Safa, S. R., Mehta, N. D., and Agresti, M. (1989) Biochem.

Bioph.ys. Res. Commun. 162, 1402-1408 7. Safa,A.-R. (1988) PFOC. Natl. Acad. Sci. U. S. A. 85, 7178-7191 8. Yusa. K.. and Tsuruo. T. (1989) Cancer Res. 49,5002-5006 9. Qian; X-d., and Beck,‘W. T. (1990) Cancer Res. 50, 1132-1137

10. Safa, A. R., Glover, C. J., Sewell, J. L., Meyers, M. B., Biedler, J. L., and Felsted, R. L. (1987) J. Biol. Chem. 262, 7884-7888

Labeling of P-glycoprotein

11.

12. 13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Cornwell, M. M., Tsuruo, T., Gottesman, M. M., and Pastan, I. (1987) FASEB J. 1, 51-54

Hamada, H., and Tsuruo, T. (1988) Cancer Res. 48,4926-4932 Thiebaut, F., Tsuruo, T., Hamada, H., Gottesman, M. M., Pastan,

I.. and Willingham. M. C. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,7735-7738

Sugawara, I., Kataoka, I., Morishita, Y., Hamada, H., Tsuruo, T., Itovama, S., and Mori, S. (1988) Cancer Res. 48. 1926-1929

Arceci, R. J.,. Croop, J. M., Horwitz, S. B., and.Housman, D. (1988) Proc. Nutl. Acad. Sci. U. S. A. 85, 4350-4354

Cordon-Cardo, C., O’Brian, J. P., Casals, D., Rittman-Grauer, Biedler, J. L., Melamed, M. R., and Bertino, J. R. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 695-698

Yang, C. H., DePinho, S. G., Greenberger, L. M., Arceci, R. J., and Horwitz, S. B. (1989) J. Biol. Chem. 264, 782-788

Naito, M., Yusa, K., and Tsuruo, T. (1989) Biochem. Biophys. Res. Commun. 158,1066-1071

Martyr, R. J., and Bernisek, W. T. (1973) Biochemistry 12,2172- 2178

Dure, L. S., IV, Schrader, W. T., and O’Malley, B. W. (1980) Nature 283, 784-786

Horwitz, K. B., and Alexander, P. S. (1983) Endocrinology 113, 2195-2201

Casey, M. L., MacDonald, P. C., and Simpson, E. R. (1985) in Williams Textbook of Endocrinological Changes of Pregnancy (Wilson, J. D., and Foster, D. W., eds) 7th Ed., pp. 422-437, W. B. Saunders Co., Philadelphia

Beck, W. T., Muller, T. J., and Tanzer, L. R. (1979) Cancer Res. 39,2070-2076

Kartner, N., Evernden-Porelle, D., Bradley, G., and Ling, V. (1985)Nature 316,820-823

Pearce, H. L., Safa, A. R., Bach, N. J., Winter, M. A., Cirtain, M. C., and Beck, W. T. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,5128-5132

Zamora, J. M., Pearce, H. L., and Beck, W. T. (1988) Mol. Pharmacol. 33,454-462

Spatz, R. J., Pillalamari, E. D., Diab, G. M., Abadir, A. R., and Gintantas, J. (1986) Proc. West. Pharmacol. Sot. 29, 319-320

Gottesman. M. M.. and Pastan. I. (1988) J. Biol. Chem. 263. 12163-12166

Moscow, J. A., and Cowan, K. H. (1988) J. Natl. Cancer Inst. 80, 14-20

by guest on April 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

X D Qian and W T Beckleukemic lymphoblasts.

Progesterone photoaffinity labels P-glycoprotein in multidrug-resistant human

1990, 265:18753-18756.J. Biol. Chem. 

  http://www.jbc.org/content/265/31/18753Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/265/31/18753.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on April 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from