3. F. R. Carbone and Y. Paterson, J. Immunol. 135, 2609 (1985). 4. R. R. Porter, Biochem. J. 46, 479 (1950); R. R. Porter, ibid. 73, 119 (1959); A. Nisonoff, F. C. Wissler, L. N. Lipman, D. L. Woemnley, Arch. Biochem. Biophys. 89, 230 (1960). 5. E. Margoliash, E. L. Smith, G. Kreil, H. Tuppy, Nature (London) 192, 1125 (1961). 6. The monoclonal antibodies (3 mg), affinity-purified on the respective antigen, were incubated with 5% trypsin (w/w of total protein) overnight at room temperature and then separated from trypsin and any released eptidesby chromatography on Sepha- dex G-75. TFhe cytochrome c-specific antibodies were then incubated with a saturatsng amount of the antigen (1 mg) for 1 hour at 4°C. Unbound cyto- chrome c was separated from the IgG-bound antigen by chromatography on Sephadex G-75 in 50 mM ammonium bicarbonate, pH 8.3. Between 1.4 and 2.0 moles of cvtochrome c remained bound to each mole of trypsinized IgG1. An equimolar amount of cytochrome c that just saturated the binding sites on the specific antibodies was added to anti-DNP IgG1 in control experiments. 7. R. E. Dickerson, Sci. Am. 226, 58 (April 1972); T. Takano and R. E. Dickerson, J. Mol. Bid. 153, 95 (1981). 8. Further evidence that residue 60 is immunodomi- nant for E8 is based on site-specific chemical modifi- cation of Trp59 by N-formylation. [I. Aviram and A. Schejter, Biochim. Biophys. Acta 229, 113 (1971)]. Cytochromec that contained N-formyl Trp" did not bind E8 but did bind C3 in solid-phase radioim- munoassay. However, our unpublished results indi- cate that the modification procedure also appears to cause changes in the heme-peptide structure. 9. The following abbreviations were used for amino acids: A, alanine; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; Q, glutamine; T, threonine; W, tryptophan; Y, tyrosine. 10. R. B. Merrifield,J. Am. Chem. Soc. 86, 304 (1964); J. M. Stewart and J. D. Young, Solid Phase Peptide Synthesis (Pierce Chemical, Rockford, IL, ed. 2, 1984). 11. The a-amino group of 21-28-KGG was acetylated while the peptide was still protected and attached to the benzhydrylamine resin. Then residues 41-47 were synthesized on the side chain N-E-amino group of the lysine following residue 28 (which does not occur in the natural sequence). In the synthesis of 21-28-KGG, the lysine following residue 28 was side chain-protected with the 9-fluoroenylmeth- yloxycarbonyl group. This allowed deprotection of its s-amino group under basic conditions without affecting other side chain protecting groups (10). The COOH-terminal glycine residues were used to separate the growing end of the peptide chain from the resin to increase the efficiency of subsequent coupling. 12. R. emmnerson, P. R. Morrow, N. R. Klinman, Y. Paterson, Proc. Natl. Acad. Sci. U.S.A. 82, 1508 (1985). 13. H. Towbin, T. Stachlin, J. Gordon, ibid. 76, 4350 (1979). 14. D. L. Brautigan, S. Ferguson-Miller, E. Margoliash, Methods Enzymol. 53, 128 (1978). 15. H. L. Spiegelberg and W. 0. Weigle, J. Exp. Mcd. 121, 323 (1965). 16. T. J. O'Donnel and A. J. Olson, Comput. Graphics 15, 133 (1981); M. L. Connolly and A. J. Olson, Comput. Chem. 9, 1 (1985). 17. Supported by a grant from Johnson and Johnson Inc. and by NIH grants GM 31841, A121486, and AI 19499. The authors are gratefuil to A. Everson of Johnson and Johnson Biotechnology Center, Inc., for determining the amino acid compositions of the peptides, and to N. Klinman, Scripps Clinic and Research Foundation, for providing the antibodies to DNP. 21 October 1985; accepted 10 March 1986 Steroid Hormone Metabolites Are Barbiturate-Like Modulators of the GABA Receptor MARIA DOROTA MAJEWSKA,* NEIL L. HARRISON, ROCHELLE D. SCHWARTZ, JEFFERY L. BARKER, STEVEN M. PAUL Two metabolites of the steroid hormones progesterone and deoxycorticosterone, 3a- hydroxy-5a-dihydroprogesterone and 3a,5a-tetrahydrodeoxycorticosterone, are po- tent barbiturate-like ligands of the -y-aminobutyric acid (GABA) receptor-chloride ion channel complex. At concentrations between 10-7 and 10-5M both steroids inhibited binding of the convulsant t-butylbicyclophosphorothionate to the GABA-receptor complex and increased the binding of the benzodiazepine flunitrazepam; they also stimulated chloride uptake (as measured by uptake of 36C1-) into isolated brain vesicles, and potentiated the inhibitory actions of GABA in cultured rat hippocampal and spinal cord neurons. These data may explain the ability of certain steroid hormones to rapidly alter neuronal excitability and may provide a mechanism for the anesthetic and hypnotic actions of naturally occurring and synthetic anesthetic steroids. STEROID HORMONES ACT ON THE central nervous system (CNS) to pro- duce diverse neuroendocrine and be- havioral effects (1). Both adrenal and gonad- al steroids interact with intracellular recep- tors in the CNS and trigger genomically directed alterations in protein synthesis, which occur in minutes to hours (2). In addition, many steroids produce more rapid alterations in CNS excitability (1). Over 40 years ago Selye (3) described the rapid and reversible CNS depressant actions of various steroids in the rat. The gonadal steroid progesterone, and the mineralocorticoid de- oxycorticosterone, as well as several of their metabolites, were the most potent among a series of steroids in inducing sedation and anesthesia (4). On the basis of these obser- vations a class of steroidal anesthetics was developed and has been used clinically (5). The mechanisms responsible for the rapid effects of steroids on neuronal excitability are poorly understood, although the short latency (seconds to minutes) of the effects makes it unlikely that they are mediated by "classical" intracellular receptors. The anes- thetic and hypnotic actions of certain drugs, including the benzodiazepines (6), barbitu- rates (7), and the anesthetic steroid 3a- hydroxy-5a-pregnane-11,20-dione (alphax- alone) (8) may be due in part to their enhancement of the inhibitory action of the neurotransmitter y-aminobutyric acid (GABA). The potentiation of GABA-medi- ated synaptic inhibition by these drugs oc- curs by a direct interaction with GABAA receptors, which are coupled to chloride (CF-) ion channels (6-9). The GABAA re- ceptor is an oligomeric receptor complex consisting of several subunits with indepen- dent but interacting binding sites for GABA, benzodiazepines, and barbiturates (9). Because the GABAA-receptor complex is an important site of anesthetic and hyp- notic drug action, we examined a series of naturally occurring steroids for their ability to interact with one or more sites on this receptor complex. We now report that the ring A reduced metabolites of progesterone and deoxycorticosterone-namely, 3a-hy- droxy-5a-dihydroprogesterone (3a-OH- DHP) and 3a,5a-tetrahydrodeoxycortico- sterone (3a-THDOC), respectively-are potent modulators of the GABA-receptor complex and interact at a site close to or identical with that for barbiturates. A series of steroids was tested for their ability to inhibit the specific binding of the 35S-labeled convulsant t-butylbicyclophos- phorothionate ([35S]TBPS), a ligand that labels a site close to or on the GABA- operated Cl- channel (10). The specific binding of [35S]TBPS to brain membranes is inhibited both by barbiturates and by GABA antagonists such as picrotoxin, and there is a good correlation between the pharmacological potencies of these com- pounds and their ability to displace [35S]TBPS binding (10). Both 3a-OH- DHP and 3a-THDOC were relatively po- tent inhibitors of [35S]TBPS binding to the GABAA receptor-Cl- channel complex in crude synaptosomal membranes from rat M. D. Majewska, R. D. Schwartz, S. M. Paul, Sections on Molecular Pharmacology and Preclinical Studies, Clinical Neuroscience Branch, National Institute of Mental Health, Bethesda, MD 20892. N. L. Harrison and J. L. Barker, Laboratory of Neuro- physiology, National Institute of Neurological and Com- municative Disorders and Stroke, National Institutes of Health, Bethesda, MD 20892. *To whom correspondence should be addressed. SCIENCE, VOL. 232 1004. on June 12, 2015 www.sciencemag.org Downloaded from on June 12, 2015 www.sciencemag.org Downloaded from on June 12, 2015 www.sciencemag.org Downloaded from on June 12, 2015 www.sciencemag.org Downloaded from