- Email: [email protected]
Steroid Hormones: Membrane transporters of steroid hormones
STEROID HORMONES STEROID HORMONES
E. BRAD THOMPSON
Membrane transporters of steroid hormones The discovery of a plasma membrane ABC protein that exports steroids in yeast highlights the possibility that similar membrane sorting systems in mammalian cells may modulate the access of steroids to their receptors. The discovery of intracellular receptors for steroid hormones provided a powerful model that has driven thought and experimentation in its area of modern biology for the last 25-30 years. The basic principle of steroid receptors' function is that they are ligand-activated transcription factors which bind to specific DNA regulatory sites and to certain other transcription factors, thereby modulating the transcription of specific gene sets. Steroid hormones are generally thought to diffuse into cells and there encounter, bind to, and activate, their receptors which are already in the nucleus or enter it on activation. During the 1970s and '80s, as steroid binding by the glucocorticoid receptor was studied in mammalian cell and tissue systems, many reports documented a correlation between the functional potency of various glucocorticoids and their relative binding affinities for the receptor [1]. These correlations were strong enough that they could generally be carried over into in vivo applications, although pharmacokinetic issues also have to be taken into account in vivo. However, even as the general concept was developed that glucocorticoids' potency correlated with their binding affinities for the receptor, exceptions were noted. These were usually seen as quantitative discrepancies between binding and functional potency, unique to one cell or tissue, or varying between differing cells and tissues [1,2]. No known mechanism completely explains these discrepancies, which are often dismissed as being due to trivial methodological causes. It seems unlikely, however, that all such variations can be set aside on these grounds. One early suggestion was that there might be multiple glucocorticoid receptors with different binding affinities and different behaviors on activation, providing the grounds for the differences between binding and potency. Indeed, mutations in the ligand-binding domains of several steroid receptors are known to alter responses to ligand, even causing steroids that normally act as antagonists to become agonists. Thus, each individual steroid-receptor interaction must cause a change in receptor structure that affects interactions between the receptor and a complex set of transcription factors. Molecular cloning of the glucocorticoid receptor, however, has shown it to be the product of a single gene, and although some mutants of the glucocorticoid and other steroid receptors are known to occur in natural mammalian populations, there does not seem to be an array of cell-type-specific receptors in normal tissues sufficient to explain the frequent discrepancies between the potency and binding of particular 730
steroid hormones. Similarly, the known post-translational modifications of the glucocorticoid receptor have been unable to account for the data. Another suggested explanation for the discrepancy between binding and potency invokes modulatory DNA motifs. A regulatory DNA sequence named GME (glucocorticoid modulatory element), which confers enhanced responsiveness to glucocorticoids without itself binding receptor (which has specific binding motifs elsewhere in the promoter) has been identified in the promoter of the tyrosine aminotransferase gene [3]. When present, this control sequence shifts the dose-response curve for glucocorticoid induction of the gene to the left, increasing the apparent potency of several glucocorticoids. The GME sequence can even cause the anti-glucocorticoid dexamethasone mesylate to become an agonist [3]. One interpretation of these results is that interactions between the glucocorticoid receptor-ligand complex and other transcription factors are affected by the presence of the GME, which may have its own unique binding proteins. If such interactions can change the function of a ligand antagonist to that of an agonist, then it is possible that the pattern of relative potencies of other ligands could shift as well, depending on subtle modulations in receptor structure and protein-protein interactions. The distribution of the GME sequence in the genome is unknown; nor do we have a detailed knowledge of proteins that bind to it. Yet another way that cells might alter the effectiveness of specific steroids is by controlling access of steroids to their receptors. For example, metabolic processes are sometimes used to convert inactive steroids to active forms, and vice versa. An example of the former is the conversion of cortisone, which has little or no binding activity, to cortisol, by conversion of the 11-one to 11-ol. A well known example of steroid inactivation is removal of the 11 -hydroxyl group of cortisol and other glucocorticoids by 113p-hydroxysteroid dehydrogenase in the kidney; occupancy and inappropriate activation of the mineralocorticoid receptor by glucocorticoids is thus prevented, leaving the way free for the action of aldosterone, which circulates at lower concentrations than glucocorticoids. The cell's plasma membrane is a natural filtering and sorting device for many compounds, and it may be used to sort steroids, too. In the laboratory practice of using non-mammalian eukaryotes to study steroid actions, this can be an annoyance. Recently, several labs have carried out studies of steroid receptor function in yeast. The
© Current Biology 1995, Vol 5 No 7
DISPATCH Fig. 1. How the yeast plasma membrane ABC transporter LEM1 alters the biological potency of glucocorticoids from that predicted by their binding affinities for the glucocorticoid receptor. Three glucocorticoids (G1, G2 and G3) are shown, with affinities for binding to the receptor in descending order, G3>G2>G1. When applied to yeast cells, however, their effective potencies are changed as a result of the action of LEM1, which exports much of G2 and G3 but allows free entry of G1.
powerful genetic manipulations that can be performed in these simple organisms makes them very attractive for analysis of expressed mammalian genes. Furthermore, yeast lack glucocorticoid receptors, providing a 'zero background' against which to study the actions of expressed receptors. However, several popular, potent synthetic glucocorticoids, used widely to activate their receptors in mammalian systems, have proven to be only weak agonists in yeast. On the other hand, the relatively weak agonist deoxycorticosterone functions well in yeast cells. Kralli et al. [4] have made a virtue of this problem in recently published work. Correctly suspecting that there was a membrane barrier involved in the failure of glucocorticoid receptors expressed in yeast cells to respond to low concentrations of some agonists, they used the power of yeast genetics to identify a mutation that allowed activation of the glucocorticoid receptor by concentrations of triamcinolone and dexamethasone far below those usually required, while not affecting the deoxycorticosterone dose-response curve. By complementation analysis, the gene responsible for this effect was obtained and its protein product identified. The wild-type protein, named LEM1 by the authors, was found to be a member of the ABC (ATP-binding cassette) superfamily of transporter proteins. Members of this family have a structure in which two ATP-binding domains alternate with two hydrophobic motifs, each of which spans the membrane six times. They have ATPase and chloride channel activity and function to transport a variety of compounds of more or less hydrophobic character; they have been the subject of several recent reviews [5,6]. LEM1 selectively pumps dexamethasone and triamcinolone, but not deoxycorticosterone, from yeast cells. Thus its inactivation in the mutant allows the synthetic ligands to enter freely, activating the glucocorticoid receptor according to their binding affinities (see Fig. 1).
Interestingly, based on entirely different selection criteria, the LEM1 gene was independently cloned and identified as encoding an ABC transporter protein by two other groups - in one case as PDR5, cloned for resistance to cycloheximide [7], and in the other as STS1, conferring resistance to sporidesmin [8]. If this sort of behavior - a single membrane protein providing resistance to several unrelated compounds - sounds familiar, it should. A widely studied member of the ABC superfamily in mammalian cells is the P glycoprotein product of the multiple drug resistance (MDR1) gene [6]. This protein confers resistance to a number of structurally dissimilar compounds. It is often expressed at high levels in cancer cells, but is also found in some normal tissues, particularly in the membranes of steroid-secreting cells and cells of transporting epithelia. The P glycoprotein can bind several steroids and can export some of them [6,9-12]; it has been speculated that its naturally high expression in the adrenal gland permits adrenal cells to handle their highsteroid environment. Years ago, selection for mouse lymphomas produced a 'membrane mutant' which excluded glucocorticoids, and this effect turned out to be due to high expression of MDR 1 [9]. Kralli et al. [4] suggest that membrane sorting of steroids in mammalian cells may be more widespread than we think. ABC proteins are found across vast taxonomic and phylogenetic spreads, from the ascomycete yeasts in the Kingdom Fungi, to the Animal Kingdom, where they have been widely studied in two disparate phyla insects and vertebrates. Nature has a way of using the same structural motifs in proteins for differing functions across phylogenetic distances. This could mean that the functions of LEM1 may be special for yeast. Nature also has a way, however, of preserving vital functions lodged in particular proteins. Whether the steroid barrier that LEM1 provides for yeast will prove to be reflected as an analogous major natural function for the P glycoprotein
731
732
Current Biology 1995, Vol 5 No 7 and other yet-to-be-found members of the ABC group of proteins in mammals remains to be seen. At present we can only speculate about the natural functions of both LEM1 and P glycoprotein. Perhaps subtle sorting of steroids will prove to be one duty of proteins of the ABC family. References 1. Munck A, Leung K: Glucocorticoid receptors and mechanisms of action. In Receptors and Mechanism of Actions of Steroid Hormones. Edited by Pasqualini JR. New York: Marchel Dekker; 1977: 311-397. 2. Harmon JM, Schmidt TJ,Thompson EB: Deacylcortivazol acts through glucocorticoid receptors. J Steroid Biochem 1981, 14:273-279. 3. Oshima H, Simons SS Jr: Modulation of transcription factor activity by a distant steroid modulatory element. Mol Endocrinol 1992, 6:416 428, 1992. 4. Kralli A, Bohen SP, Yamamoto KR: LEM1, an ATP-binding-cassette transporter, selectively modulates the biological potency of steroid hormones. Proc Natl Acad Sci USA 1995, 92:4701-4705. 5. Higgins CF: ABC transporters: from microorganisms to man. Annu Rev Cell Biol 1992, 8:67-113. 6. Gottesman MM, Pastan : Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu Rev Biochem 1993, 62:385-427.
7. Balzi E, Wang M, Leterme S, Van Dyck L, Goffeau A: PDRS, a novel yeast multidrug resistance conferring transporter controlled by the transcription regulator PDR1. J Biol Chem 1994, 269: 2206-2214. 8. Bissinger PH, Kuchler K: Molecular cloning and expression of the Saccharomyces cerevisiae STS1 gene product. J Biol Chem 1994, 269:4180-4186. 9. Bourgeois S, Gruol DJ, Newby RF, Rajah FM: Expression of an mdr gene is associated with a new form of resistance to dexamethasoneinduced apoptosis. Mol Endocrinol 1993, 7:840-851. 10. Ueda K, Okamura N, Hirai M, Tanigawara Y, Saeki T, Kioka N, Komano T, Hori R: Human p glycoprotein transports cortisol, Biol aldosterone, and dexamethasone, but not progesterone. Chem 1992, 267:24248-24252. 11. Wolf DC, Horwitz SB: P glycoprotein transports corticosterone and is photoaffinity-labeled by the steroid. Int Cancer 1992, 52: 141-146. 12. van Kalken CK, Broxterman HJ, Pinedo HM, Feller N, Dekker H, Lankelma J, Giaccone G: Cortisol is transported by the multidrug resistance gene product p glycoprotein. Br Cancer 1993, 67: 284-289.
E. Brad Thompson, Department of Human Biological Chemistry and Genetics, 603 Basic Science Building 0645, The University of Texas Medical Branch, Galveston, Texas 77555-0645, USA.
THE AUGUST 1995 ISSUE (VOL. 7, NO. 4) OF CURRENT OPINION IN CELL BIOLOGY will include the following reviews, edited by Jean Gruenberg and Thomas Kreis, on Membranes and Sorting: Quality control in the secretory pathway by Craig Hammond and Ari Helenius Sorting and retrieval between the endoplasmic reticulum and the Golgi apparatus by Hugh R.B. Pelham The role of coat proteins in the biosynthesis of secretory proteins by Nina R. Salama and Randy W. Schekman Protein transport to the yeast vacuole by Bruce E Horazdovsky, Daryll B. DeWald and Scott D. Emr Membrane transport in the endocytic pathway by Frederick Maxfield and Jean Gruenberg Antigen processing/presentation by Ira Mellman The emergence of clathrin-independent pinocytic pathways by Christophe Lamaze and Sandra L. Schmid SNAREs and the specificity of transport vesicle targeting by Mark K. Bennett Molecular motors, membrane movements and physiology: emerging roles for myosins by Tama Hasson and Mark S. Mooseker The same issue will also include the following reviews on Membrane Permeability, edited by Qais Al-Awqati: Chloride channels of intracellular organelles by Qais Al-Awqati Cell volume regulated transporters of compatible osmolytes by H. Moo Kwon andJoseph S. Handler The extracellular calcium receptor by Steven C. Hebert and Edward M. Brown Aquaporin water channels: unanswered questions and unresolved controversies by Peter Agre, Dennis Brown and Soren Neilson Molecular characterization of the Na-K-Cl cotransporter isoforms by John Payne and Bliss Forbush Structure and function of fusion pores in exocytosis and ectoplasmic membrane fusion by Manfred Lindau and Wolfhard Almers
E. BRAD THOMPSON
Membrane transporters of steroid hormones The discovery of a plasma membrane ABC protein that exports steroids in yeast highlights the possibility that similar membrane sorting systems in mammalian cells may modulate the access of steroids to their receptors. The discovery of intracellular receptors for steroid hormones provided a powerful model that has driven thought and experimentation in its area of modern biology for the last 25-30 years. The basic principle of steroid receptors' function is that they are ligand-activated transcription factors which bind to specific DNA regulatory sites and to certain other transcription factors, thereby modulating the transcription of specific gene sets. Steroid hormones are generally thought to diffuse into cells and there encounter, bind to, and activate, their receptors which are already in the nucleus or enter it on activation. During the 1970s and '80s, as steroid binding by the glucocorticoid receptor was studied in mammalian cell and tissue systems, many reports documented a correlation between the functional potency of various glucocorticoids and their relative binding affinities for the receptor [1]. These correlations were strong enough that they could generally be carried over into in vivo applications, although pharmacokinetic issues also have to be taken into account in vivo. However, even as the general concept was developed that glucocorticoids' potency correlated with their binding affinities for the receptor, exceptions were noted. These were usually seen as quantitative discrepancies between binding and functional potency, unique to one cell or tissue, or varying between differing cells and tissues [1,2]. No known mechanism completely explains these discrepancies, which are often dismissed as being due to trivial methodological causes. It seems unlikely, however, that all such variations can be set aside on these grounds. One early suggestion was that there might be multiple glucocorticoid receptors with different binding affinities and different behaviors on activation, providing the grounds for the differences between binding and potency. Indeed, mutations in the ligand-binding domains of several steroid receptors are known to alter responses to ligand, even causing steroids that normally act as antagonists to become agonists. Thus, each individual steroid-receptor interaction must cause a change in receptor structure that affects interactions between the receptor and a complex set of transcription factors. Molecular cloning of the glucocorticoid receptor, however, has shown it to be the product of a single gene, and although some mutants of the glucocorticoid and other steroid receptors are known to occur in natural mammalian populations, there does not seem to be an array of cell-type-specific receptors in normal tissues sufficient to explain the frequent discrepancies between the potency and binding of particular 730
steroid hormones. Similarly, the known post-translational modifications of the glucocorticoid receptor have been unable to account for the data. Another suggested explanation for the discrepancy between binding and potency invokes modulatory DNA motifs. A regulatory DNA sequence named GME (glucocorticoid modulatory element), which confers enhanced responsiveness to glucocorticoids without itself binding receptor (which has specific binding motifs elsewhere in the promoter) has been identified in the promoter of the tyrosine aminotransferase gene [3]. When present, this control sequence shifts the dose-response curve for glucocorticoid induction of the gene to the left, increasing the apparent potency of several glucocorticoids. The GME sequence can even cause the anti-glucocorticoid dexamethasone mesylate to become an agonist [3]. One interpretation of these results is that interactions between the glucocorticoid receptor-ligand complex and other transcription factors are affected by the presence of the GME, which may have its own unique binding proteins. If such interactions can change the function of a ligand antagonist to that of an agonist, then it is possible that the pattern of relative potencies of other ligands could shift as well, depending on subtle modulations in receptor structure and protein-protein interactions. The distribution of the GME sequence in the genome is unknown; nor do we have a detailed knowledge of proteins that bind to it. Yet another way that cells might alter the effectiveness of specific steroids is by controlling access of steroids to their receptors. For example, metabolic processes are sometimes used to convert inactive steroids to active forms, and vice versa. An example of the former is the conversion of cortisone, which has little or no binding activity, to cortisol, by conversion of the 11-one to 11-ol. A well known example of steroid inactivation is removal of the 11 -hydroxyl group of cortisol and other glucocorticoids by 113p-hydroxysteroid dehydrogenase in the kidney; occupancy and inappropriate activation of the mineralocorticoid receptor by glucocorticoids is thus prevented, leaving the way free for the action of aldosterone, which circulates at lower concentrations than glucocorticoids. The cell's plasma membrane is a natural filtering and sorting device for many compounds, and it may be used to sort steroids, too. In the laboratory practice of using non-mammalian eukaryotes to study steroid actions, this can be an annoyance. Recently, several labs have carried out studies of steroid receptor function in yeast. The
© Current Biology 1995, Vol 5 No 7
DISPATCH Fig. 1. How the yeast plasma membrane ABC transporter LEM1 alters the biological potency of glucocorticoids from that predicted by their binding affinities for the glucocorticoid receptor. Three glucocorticoids (G1, G2 and G3) are shown, with affinities for binding to the receptor in descending order, G3>G2>G1. When applied to yeast cells, however, their effective potencies are changed as a result of the action of LEM1, which exports much of G2 and G3 but allows free entry of G1.
powerful genetic manipulations that can be performed in these simple organisms makes them very attractive for analysis of expressed mammalian genes. Furthermore, yeast lack glucocorticoid receptors, providing a 'zero background' against which to study the actions of expressed receptors. However, several popular, potent synthetic glucocorticoids, used widely to activate their receptors in mammalian systems, have proven to be only weak agonists in yeast. On the other hand, the relatively weak agonist deoxycorticosterone functions well in yeast cells. Kralli et al. [4] have made a virtue of this problem in recently published work. Correctly suspecting that there was a membrane barrier involved in the failure of glucocorticoid receptors expressed in yeast cells to respond to low concentrations of some agonists, they used the power of yeast genetics to identify a mutation that allowed activation of the glucocorticoid receptor by concentrations of triamcinolone and dexamethasone far below those usually required, while not affecting the deoxycorticosterone dose-response curve. By complementation analysis, the gene responsible for this effect was obtained and its protein product identified. The wild-type protein, named LEM1 by the authors, was found to be a member of the ABC (ATP-binding cassette) superfamily of transporter proteins. Members of this family have a structure in which two ATP-binding domains alternate with two hydrophobic motifs, each of which spans the membrane six times. They have ATPase and chloride channel activity and function to transport a variety of compounds of more or less hydrophobic character; they have been the subject of several recent reviews [5,6]. LEM1 selectively pumps dexamethasone and triamcinolone, but not deoxycorticosterone, from yeast cells. Thus its inactivation in the mutant allows the synthetic ligands to enter freely, activating the glucocorticoid receptor according to their binding affinities (see Fig. 1).
Interestingly, based on entirely different selection criteria, the LEM1 gene was independently cloned and identified as encoding an ABC transporter protein by two other groups - in one case as PDR5, cloned for resistance to cycloheximide [7], and in the other as STS1, conferring resistance to sporidesmin [8]. If this sort of behavior - a single membrane protein providing resistance to several unrelated compounds - sounds familiar, it should. A widely studied member of the ABC superfamily in mammalian cells is the P glycoprotein product of the multiple drug resistance (MDR1) gene [6]. This protein confers resistance to a number of structurally dissimilar compounds. It is often expressed at high levels in cancer cells, but is also found in some normal tissues, particularly in the membranes of steroid-secreting cells and cells of transporting epithelia. The P glycoprotein can bind several steroids and can export some of them [6,9-12]; it has been speculated that its naturally high expression in the adrenal gland permits adrenal cells to handle their highsteroid environment. Years ago, selection for mouse lymphomas produced a 'membrane mutant' which excluded glucocorticoids, and this effect turned out to be due to high expression of MDR 1 [9]. Kralli et al. [4] suggest that membrane sorting of steroids in mammalian cells may be more widespread than we think. ABC proteins are found across vast taxonomic and phylogenetic spreads, from the ascomycete yeasts in the Kingdom Fungi, to the Animal Kingdom, where they have been widely studied in two disparate phyla insects and vertebrates. Nature has a way of using the same structural motifs in proteins for differing functions across phylogenetic distances. This could mean that the functions of LEM1 may be special for yeast. Nature also has a way, however, of preserving vital functions lodged in particular proteins. Whether the steroid barrier that LEM1 provides for yeast will prove to be reflected as an analogous major natural function for the P glycoprotein
731
732
Current Biology 1995, Vol 5 No 7 and other yet-to-be-found members of the ABC group of proteins in mammals remains to be seen. At present we can only speculate about the natural functions of both LEM1 and P glycoprotein. Perhaps subtle sorting of steroids will prove to be one duty of proteins of the ABC family. References 1. Munck A, Leung K: Glucocorticoid receptors and mechanisms of action. In Receptors and Mechanism of Actions of Steroid Hormones. Edited by Pasqualini JR. New York: Marchel Dekker; 1977: 311-397. 2. Harmon JM, Schmidt TJ,Thompson EB: Deacylcortivazol acts through glucocorticoid receptors. J Steroid Biochem 1981, 14:273-279. 3. Oshima H, Simons SS Jr: Modulation of transcription factor activity by a distant steroid modulatory element. Mol Endocrinol 1992, 6:416 428, 1992. 4. Kralli A, Bohen SP, Yamamoto KR: LEM1, an ATP-binding-cassette transporter, selectively modulates the biological potency of steroid hormones. Proc Natl Acad Sci USA 1995, 92:4701-4705. 5. Higgins CF: ABC transporters: from microorganisms to man. Annu Rev Cell Biol 1992, 8:67-113. 6. Gottesman MM, Pastan : Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu Rev Biochem 1993, 62:385-427.
7. Balzi E, Wang M, Leterme S, Van Dyck L, Goffeau A: PDRS, a novel yeast multidrug resistance conferring transporter controlled by the transcription regulator PDR1. J Biol Chem 1994, 269: 2206-2214. 8. Bissinger PH, Kuchler K: Molecular cloning and expression of the Saccharomyces cerevisiae STS1 gene product. J Biol Chem 1994, 269:4180-4186. 9. Bourgeois S, Gruol DJ, Newby RF, Rajah FM: Expression of an mdr gene is associated with a new form of resistance to dexamethasoneinduced apoptosis. Mol Endocrinol 1993, 7:840-851. 10. Ueda K, Okamura N, Hirai M, Tanigawara Y, Saeki T, Kioka N, Komano T, Hori R: Human p glycoprotein transports cortisol, Biol aldosterone, and dexamethasone, but not progesterone. Chem 1992, 267:24248-24252. 11. Wolf DC, Horwitz SB: P glycoprotein transports corticosterone and is photoaffinity-labeled by the steroid. Int Cancer 1992, 52: 141-146. 12. van Kalken CK, Broxterman HJ, Pinedo HM, Feller N, Dekker H, Lankelma J, Giaccone G: Cortisol is transported by the multidrug resistance gene product p glycoprotein. Br Cancer 1993, 67: 284-289.
E. Brad Thompson, Department of Human Biological Chemistry and Genetics, 603 Basic Science Building 0645, The University of Texas Medical Branch, Galveston, Texas 77555-0645, USA.
THE AUGUST 1995 ISSUE (VOL. 7, NO. 4) OF CURRENT OPINION IN CELL BIOLOGY will include the following reviews, edited by Jean Gruenberg and Thomas Kreis, on Membranes and Sorting: Quality control in the secretory pathway by Craig Hammond and Ari Helenius Sorting and retrieval between the endoplasmic reticulum and the Golgi apparatus by Hugh R.B. Pelham The role of coat proteins in the biosynthesis of secretory proteins by Nina R. Salama and Randy W. Schekman Protein transport to the yeast vacuole by Bruce E Horazdovsky, Daryll B. DeWald and Scott D. Emr Membrane transport in the endocytic pathway by Frederick Maxfield and Jean Gruenberg Antigen processing/presentation by Ira Mellman The emergence of clathrin-independent pinocytic pathways by Christophe Lamaze and Sandra L. Schmid SNAREs and the specificity of transport vesicle targeting by Mark K. Bennett Molecular motors, membrane movements and physiology: emerging roles for myosins by Tama Hasson and Mark S. Mooseker The same issue will also include the following reviews on Membrane Permeability, edited by Qais Al-Awqati: Chloride channels of intracellular organelles by Qais Al-Awqati Cell volume regulated transporters of compatible osmolytes by H. Moo Kwon andJoseph S. Handler The extracellular calcium receptor by Steven C. Hebert and Edward M. Brown Aquaporin water channels: unanswered questions and unresolved controversies by Peter Agre, Dennis Brown and Soren Neilson Molecular characterization of the Na-K-Cl cotransporter isoforms by John Payne and Bliss Forbush Structure and function of fusion pores in exocytosis and ectoplasmic membrane fusion by Manfred Lindau and Wolfhard Almers