Thymosin fraction 5 (TF5) stimulates secretion of adrenocorticotropic hormone (ACTH) from cultured rat pituitaries

Life Sciences, Vol. 42, pp. 2259-2268 Printed in the U.S.A.

Pergsmon Press

TETMOSIN FRACTION 5 (TF5) STIMULATES SECRETION OF ADRENOCORTICOTROPIC HORMONE (ACTB) FROM CDLTDRED RAT PITDITARIES Joseph P. McGillis, Nicholas R. Hall, and Allan L. Goldstein Department of Biochemistry George Washington University Washington, DC 20037 (Received in final form March 29, 1988) Summary In vivo administration of a partially purified thymic hormone-containing extract of the thymus gland, TF5, causes an increase in serum glucocorticoids. The lack of a direct effect of TF5 on adrenal corticosterone secretion suggests that it is mediated at the level of the pituitary. Cultured rat pituitary monolayers were used to determine if the effect is mediated by stimulation of ACTB secretion from the pituitary. Two lots of TF5, BPPlOO and C114080-01, caused a dose dependent secretion of ACTB from cultured pituitary monolayers. There was a synergistic effect when the cells were treated with both TF5 and corticotropin-releasing factor (CRF). Immunoneutralization studies were done in which the cells were treated with TF5 or CRF and an antibody to CRF. The antibody completely blocked CRF induced ACTB release, but had no effect on TF5 stimlated ACTB release, suggesting that the activity is not due to a CRF-like peptide in TF5. A number of peptides isolated from TF5, and certain other peptides produced by the immune system were evaluated for their ability to stimulate ACTB secretion. These included thymosin (TSN) al. all, and 8 , prothymosin a (PT a, thymopoeitin 5 (TP5), factuer thymique serique (Fe S), interferon o (INF a), INF y, interleukin 1 (IL-l), and interleukin 2 (IL-2). None of these factors had any effect on pituitary ACTB secretion. These results demonstrate that some peptide component of TF5 causes an increase in serum corticosteroids by stimulating pituitary ACTB release. Stimulation of adrenal function by products af the thymus gland and the immune system, which has recently been demonstrated experimentally (l-4), was first suggested by Sajous over 80 years ago (5). This early speculation was based on observations made in the late 19th century that injection of thymic extracts caused an increase in pulse rate and blood pressure. The existance of bidirectional communication between the thymus gland and the putitary-adrenal system is also supported by earlier observations that bilateral adrenalectomy causes an increase in thymic weight and that thymectomy causes an increase in adrenal weight (6-8). More recently, several investigators have suggested the existence of a thymicjimmune system-adrenal endocrine axis involved in the dynamic homeostatic regulation of both systems (g-11). A recurring theme in these hypothesis is that during an inflamatory response, there are physiological changes which include an alteration in circulating hormones and in neural activity. Besedovsky and co-workers have suggested that an elevation in glucocorticoids which parallels antibody production in an immune response acts as a negative feedback inhibitor on immunological processes (9). One of the major tasks at present is to identify the molecular components involved in 0024-3205188 $3.00 + .OO Copyright (c) 1988 Pergamon Press plc

2260

TF5 Stimulates ACTR Secretion

Vol. 42, No. 22, 1988

central nervous system-immune system communication, and to define their physiologic roles and sites of action. The demostration that lymphocytes can produce pro-opiomelanocortin and pro-enkephalin gene products (12, 14), peptides originally described with respect to their roles in neuroendocrine systems, and the identification of opiate binding sites on lymphocytes (15), suggests that the pathways involved are complex and consist of multiple neuroendocrine and immune system foci. TF5 is a thymus extract consisting of low molecular weight acid stable peptides (rev. in 16). Certain components of TF5 are involved in various stages of T-cell differentiation. Peptides that have been isolated from TF5 and sequenced include TSN QI, all, B,, B , f? , g,,, and PT a (17-22). PT a, a precursor to TSN al and all, is a ll& am!no acid peptide isolated from thymus. Another peptide that has recently been isolated, parathymosin a, shares about 50% homology with PT CL (23). Recently, it has been shown that injection of TF5 causes an increase in serum glucocorticoids in rodents (1). This increase is dose and time dependent with a maximal response at 1 hour in mice and at 2 hours in rats. In similar studies using monkeys it was observed that ACTR, 8 endorphin, and cortisol were elevated following administration of TF5 (2). The elevation in ACTR suggested that the stimulatory effect on the adrenal gland might be mediated directly through pituitary ACTR release. This site of action is also suggested by an earlier study in which it was shown that TF5 had no effect on in vitro adrenal corticosteroid release or adrenal CAMP production (24). However, it is not possible to conclude whether the stimulatory effects are mediated directly at the pituitary gland, or whether they are mediated by modulation of other factors, such as hypothalamic CRF. In order to answer this question TF5 and a number of other products of the immune systme were evaluated for their ability to stimulate secretion of pituitary ACTR in vitro. In these studies it was found that TF5 could stimulate pituitary ACTR secretion in vitro, whereas none of the TSN peptides, interleukins, or interferons had any effect on pituitary ACTR release. Methods

, and B, were gifts from Hoffmann-Laroche Materials: TF5' synthetic TSN1 a * a&als, Inc. (Washington, DC). Native rat NJ) and Alpha Bfome Inc., (Nutley, PT a was a gift from Dr. Bernie Horecker (Roche Institute for Molecular Biology, Nutly, NJ). Recombinant IL-2, IFN CL, INF y, and purified natural IL-l were a gift from Dr. Joost Cppenheim (National Cancer Institute, Frederick, MD). Concanavalin A (Con A) stimulated rat spleen cell supernatant fluids were prepared as described previously (16). Synthetic ovine and rat CRF were purchased from Peninsula Labs (San Carlos, CA), and Bachem (Torrance, CA). The reagents for the ACTR, lutienizing hormone (LH), prolactin (PRL), and growth hormone radioimmunoassays (RIA) were gifts from the National Pituitary Agency, Baltimore, MD. Antibodies to oCRF and rCRF were obtained from Peninsula labs and Immunonuclear Corp (Stillwater, MN). Arginine vasopressin (AVP) RIAs were done with an RIA kit purchased from Immunonuclear Corp. Sheep anti rabbit antiserum was obtained from Endocrine Sciences (Torrence, CA). DMRM, MPH non essential amino acids, penstrep, L-glutamine, horse serum and fetal calf serum were obtained from Gibco (Grand Island, NY). Pituitary Monolayer Cultures: Pituitary monolayer cultures were prepared by collanenase digestion of rat anterior pituitaries as described by Vale et al. (25,26). Fifteen to thirty male Wistar rats were used for eacQ preparation. After enzymatic dispersion the cells were diluted to 3.33 x 10 viable cells per ml in DMEM supplemented with 1 X MRM non essential amino acids, 100 U/oil100 ug/ml penstrep, 10 mM L-glutamine, 10 % horse serum, and 5 X fetal calf serum, and were plated at 0.5 ml/well in 48 well costar plates. The cells were

vol. 42, No. 22, 1988

TF5 StimulatesACTH Secretion

2261

placed in a humidified 37'C, 7 % CO incubatorand allowed to attach to the plates for at least 3 days prior to2experimentaltreatments. Prior to treatment, each well was washed 3 times with DMPM without serum. Followinga 60 minute preincubationthe media was replaced with 0.45 ml fresh DMRM without serum. The substancesto be tested were diluted to 10 times the final concentration and 50 ul were added to triplicateor quadruplicatewells. After a 4 hour incubaiionperiod the supernatantfluid was carefully removed and stored at -20 C. Radioimmunoassays: Peptides were iodinatedwith ch{qfamineT (27) as described for TSN a, (28) using concentrationsof peptide, Na I, chloramine-T, sodium metabislufiteand potassium iodide suggestedby the National Pituitary Agency for LH, GH, and PRL, by Orth (29) for ACTH, and by Moldow and Fishman (30) for oCRF and rCRF. The RIAs were done using the protocol described tor the TSN a RIA.(28) with the exception that the SPF,ABbuffer (50 mM Na PO , 25 mM KDTA, 100 mM NaCl, 0.1 % Na aside, 0.1 % BSA, and 0.1 % triton X2106, pH 7.3) described by Vale et al. (26) was used in the CRF RIAs. Dilution curves for the immunoreactivehormones in the culture supernatantfluids and in TF5 were paralled with the standards in all of the pituitaryhormone RIAs. The standardsused for the pituitary hormonea were NIADDK ACTH-RP 1, NIAMD LH-RP-3, NIADDK PRL-RP-3, or NIADDK rGH-RP-1. Statisticalanalysis of the results was done using either the Students t-Test or one way analysis of variance. Results ImmunoreactiveMaterials Present in TFS: The amount of immunoreactiveACTH (IrACTH)in TF5 was determined in order to insure that an increase in ACTH in the culture supematants was not due to background irACTH in TFS. The levels of other pituitaryhormones present in TP5 and the levels of certain peptides which are known to cause a release of ACTH were also determined. The results are summarizedin Table I. ImmunoreactiveACTH, LH PRL, and GH were all present in TFS. The source of these materials in TF5 could be from either blood and extracellularfluid, or could be immunoreactivematerials produced in the thymus gland. This latter possibilityis supportedby reports demonatrating that lymphocytescan produce immunoreactiveneuropeptidesand their mRNAs (12-14, 31). In each experimentincubationcontrolswere included in which TF5 was added to wells without cells and incubated for an identicalamount of time. The levels of immunoreactiveACTH in the incubationcontrolswere subtractedfrom the test groups. In the pituitary culture experimentsirACTH from TF5 was detected only at doses above 250 ug/ml. At the highest dose tested, 1 mg/ml, the irACTH in TF5 accounted for no more that 15% of the total ACTH measured. The levels of immunoreactiveAVP and CRP in TF5 were also determined. No immunoreactiveCRF could be detected using RIAs for either oCRF or rCRF. However, there were detectablelevels of AVP. In studies using the pituitary culture system descrived above it has been reported that the lowest doses of AVP which stimulatedACTH secretionwere in the low nanpylar range (32). The amount of immunoreactiveAVP present in TFS, 9.90 x 10 moles/mg, was insufficientto cause release of ACTH at the doses tested. Stimulationof ACTH release from cultured pituitarymonolayers: To validate the bioassay the pituitary cultureswere first treated with CRF to asses the response to-a known stimuli. -13 these_Ttudiesthe cells were treated with to 10 M. The sensffivityoclshe pituitary doses of rCRF ranging from 10 M CRF, peaked cultu_rgsto CRF induced ACTH release was between 1211 and 10 at 10 M CRF, and had an KDso of approximately10 M. ACTH productionby

2262

TF5

Vol. 42, No. 22, 1988

Stimulates ACTH Secretion

TABLE I Immunoreactive Materials In Thymosin Fraction 5

Material

Lot TF5

Content (w/w)

BPPlOO

1.04 ng/mg

C118040-01

5.69 nglmg

AVP

C118040-01

9.86 pglmg

rCRF

BPPlOO

NDa

C118040-01

ND

BPPlOO

ND

C118040-01

ND

ACTH

oCRF

BPPlOO

2.76 ng/mg

C118040-01

92 Pdmg

rLH

BPPlOO

1.14 ng/mg

rPRL

BPPlOO

220 Pdmp

C118040-01

122 Pdmg

BPPlOO

5.1

rGH

TSN "1

Kdmp

(M/mg)

2.29

x

lo-l3

1.25 x 10-12 9.09 x 10-15

1.31 x 10-13 4.38 x lo-l3 3.8 x 10-14 9.36 x lo-l5 5.19 x 10-15 1.64 x lo-'

a. ND; not detectable these cultures following CRF treatment was almost identical to the results reported by Vale et al. (26, 33). Stimulation of ACTH release by TF5 was evaluated for 2 separate lots of TF5. Fig. 1 shows a typical dose response curve for lot C114080-01. A significant increase in ACTH secretion was seen at concentrations above 5 pg/ml. Another lot of TF5, BPPlOO, was found to be 5 fold less potent in inducing ACTH release than lot C114080-01. Experiments were also done to determine if TF5 might act synergistically with CRF in inducing ACTH release. In these studies the cells were treated with suboptimal doses of CRF together with TF5 (lot BPPlOO). At the higher doses of TF5 the effects were merely additive, as is shown in Fig. 2, while at lower doses of TF5 there was a significant potentiation of the response to CRF. This enhancement occurred at doses of TF5 which had no effect by themselves. Immunoneutralization studies were done to rule out the possibility that CRF-like material in TF5 caused the ACTH release. In these experiments an antibody to CRF was added to the medium in order to block the corticotropic effects of TF5. As is shown in Fig. 3 anti-CRF at dilutions of 1:4000, 1:40,000, and 1:400,000 had no effect on TF5 stimulated ACTH release whereas it completely inhibited CRF activity. To determine whether any of the thymic peptides present in TF5 could cause a

Vol. 42, No. 22, 1988

TF5 StimulatesACTEISecretion

0.8

4

20

100

2263

500

DOSE TF5(rrg/ml)

FIG 1: TF5 stimulatedACTH secretion from cultured pituitary monolayers. Cultred rat pituitarymDnolayerswere incubatedwith TF5 (lot C118040-01) at doses ranging from 0.8 to 500 pg/ml for 4 hours. ACT8 in the supernatant fluid was measured by RIA. Basal release is shown by the shaded region. Each point representsthe mean + sem (n = 4). There was a significant increase in ACTR release at doses of TF5 of 20 pg/ml and greater (P < 0.05). release of ACTR, pituitary mono'?I':," ;'& treated with syntheticTSN al, all, M, or native PT a at oases ranging ~~mBfo"~2f~~o'~n~ing from 10 . None of these peptides had any effect on ACTR secretion. The range of concentrationsof TSN aI, and B4 tested in these studies bracket the concentrationsof these peptides that would have been present in the doses of TF5 which caused a release of ACTR. Two additionalthymic hormones, FTS and TP5, which are also present in TF5, had no effect on ACTR release. Other products of the immune system that were tested for their ability to cause a release of ACTR from pituitary monolayers includednative IL-l, recombinantIL-2, IFN a and IFN y, and con A sitmulatedspleen cell supernatants. While none of these materials are known to be in TF5, it has been reported that crude IL-2 containingpreparationsand recombinantIL-l can stimulate corticosteronesecretion in vivo (3, 4). The only experiment in which any effect on ACTR release was observed with any of these factors was in one experimentin shich the cells were treated with IL-2. The increase,although statisticallysignificant,was minimal and was seen only at doses of 10 and 100 U/nil(data not shown). These results could not be repeated in 4 subsequentexperiments.

2264

TF5 Stimulates ACTR Secretion

0

Un- CRF tmaI.dlO-“M

Vol. 42, No. 22, 1988

Fr.3 2Ojl2

TREATMENT

FIG 2: TF5 enhances CRF stimulated release of ACTS from cultured pityftary monolayers. Cultured pituitary monolayers were treated with 10 M CRF and various doses of TF5 or TF5 alone for 4 hours. ACTR was measured by RIA. Control gourps were untreated or received only CRF. The bars represent the mean 2 sem for triplicate samples. At the higher doses of TF5 the combined effects of CRF and TF5 were merely additive, whereas at lower doses of TF5, which had no effect by themselves, there was a significant enhancement of CRF induced ACTR release.

Discussion The data presented here demonstrate that the in vivo steroidogenic effect of TF5 is mediated directly at the level of the pituitary by causing a release of ACTR. A direct effect at the level of the pituitary gland is also supported by studies showing that in vivo administration of TF5 in monkeys results in an increase in ACTR and in B endorphin (2), and by studies showing that TF5 causes a release of ACTR from the AtT-20 pituitary tumor line (34). An effect at the level of the pituitary is also supported indirectly by the observation that TF5 has no direct effect on adrenal corticosterone release (24). In addition to an effect at the level of the pituitary, it is also possible that the active factor in TF5 acts at areas within the brain which regulate pituitary ACTR release. Further studies will be required to evaluate its effects at sites within the brain which regulate pituitary ACTR secretion. Although the ACTR stimulating activity is not associated with any of thymic peptides that have already been isolated and sequenced, it is probably not due to the presence in TF5 of other factors known to cause a release of ACTR, specifically CRF or AVP. No immunoreactive CRF-like material could be detected in TF5. and an antibody to CRF which completely inhibits CRF induced ACTR release had no effect on TF5 stimulated release of ACTR. Only a very small amount of immunoreactive AVP could be detected in TF5. The amount of

Vol. 42, No. 22, 1988

TF5 StimulatesACTH Secretion

2265

q 1OOrgTF5 p

3.0

Y !!I ii =

2.0

@

1.0

0

NO ANTI-CRF

1:4*oDo

1:40.000

1:400.000

ADDED DILUTION OF ANTI-CRF

FIG 3: Anit-CRF antibody does not inhibit TF5 stimulatedsecretionof ACTH from cultured piibitary monolayers. Cultured pituitary monolayer8 were treated with 10 M CRF or 100 us/ml TF5 and various dilutions of rabbit anti-CRF antibody. One group received no antibody and an untreated group was includedwith each antibody dilution. The bars represent the mean f sem for triplicatesamples. Anti-CRF completely inhibitedCRF but had no effect on TF5 stimulatedsecretion.

innrmnoreactive AVP that would have been present in the highest doses of TF5 was only in the order of 4 PM. These levels of AVP are not sufficeintto cause release of ACTH, although the stimulatoryfactor in TF5 is similar to AVP in that it can act synergisticallywith CRF in inducing ACTH release. A possible set of candidatesfor the stiaulatoryactivity are fragmentsof one of the peptidee in TF5 that has been characterized. A precursor for TSN a and all, PT a, has recently been isolated and sequenced (24). TSN al and 8. correspondto residues l-28 and 1-35 of PT a. While neither of these peptid' es nor PT a was active, there may be other fragmentsof PT a capable of stimulating ACTS secretion. PT a contains three pairs of basic amino acids, sites at which neuropeptideprecursorsare proteolyticallyprocessed to yield biologicallyactive peptides. These are located at residues 19-20, 104-105, and 107-108. The evaluationof syntheticpeptides based on these sequences in this and other systeam, may account for the ACTH stimulatoryactivity and other activitiesattributedto thymic extracts. A number of other peptides produced by component tissues of the irrmunesystem were evaluated for their ability to cause a release of ACTH in vitro. In these studies none of the materials tested caused a release of ACTH. f3neof the factors which had no effect on pituitaryACTS release in these studies, IL-1 has been reported to cause a release of ACTH in vitro from the AtT-20 cell line (35), and in vivo following injection into mice (4). The AtT-20

2266

TF5 Stimulates ACTH Secretion

Vol. 42, No. 22, 1988

line is derived from a mouse pituitary tumor and has been used as a model system for studying the cellular events associated with stimulation of ACTH secretion (36). Since the AtT-20 cell line is a tumor line, it is possible that the stimulation of ACTH secretion by IL-l is an artifact peculiar to that particular cell line, and has no physiological effect on primary pituitary corticotrophs. Other possible reasons for the discrepancy between studies with primary rat pituitary cultures and the mouse tumor line are: 1) species differences, and 2) that the IL-l tested in these studies was native IL-1 whereas the material reported to be active on the AtT-20 line was recombinant IL-l. Since recombinant eukaryotic proteins produced in prokaryotes may not be processed in the same manner as they are in eukaryotes, i.e., lack of glycosylation, it is possible that there are significant differences in the biolgical activities of native and recombinant IL-l with respect to stimulation of pituitary ACTH release. The lack of an effect directly on the pituitary suggests that this effect is mediated at some site other than the pitutiary in intact animals. One possibility is that the increase in corticosterone and ACTH is a result of the pyrogenic activity of IL-l, which is probably mediated by hypothalamic mechanisms. Another factor which has been reported to have a steroidogenic effect in vivo is an IL-2 containing lymphocyte supernatant fluid (3). In these studies both conA stimulated supernatant fluids and recombinant IL-2 were found to have no effect on ACTH release in vitro. In an earlier study it was also found that these supernatant fluids had no effect on corticosterone release in vitro (24). Since these materials appear to have no direct effect on adrenal corticosterone or pituitary ACTH release, there must be a third site which mediates this effect in vivo. While the molecular mediators involved are not clear, there is ample evidence supporting the existence of regulatory networks involved in bidirectional interactions between the immune system and the pituitary-adrenal system. The existence of 'tissue CRF's has been reviewed extensively (37), however their relevance in overall physiological pathways, and the conditions under which they in turn are regulated is unclear. Two potential roles for a thymic factor in influencing pituitary-adrenal function are: 1) regulation of growth and differentiation in that thymectomy ultimately can affect adrenal growth, and 2) homeostatic regulation of pituitary-adrenal function during an inflammatory process. In order to support such a role for the factor in TF5 it will be important to isolate and characterize the active factor. The pituitary monolayer culture system , which was used as a purification assay for CRF, should provide a sensitive bioassay for the isolation of the factor in TF5 which stimulates secretion of pituitary ACTH. Once this factor is isolated it will be possible to develop antibodies and other tools in order to study the functions and distribution of this factor, and define its roles in pituitary-adrenal function and immunoregulation. Acknowledgements The authors would like to thank Dr. Joost Oppenheim for the generous gift of interferons and lymphokines, and Ms. Janelle Oliver for the lymphokine containing supernatant fluids. This work was supported by grants and gifts from the National Institutes of Health (CA24974, NS21210, and MHOO648), Alpha 1 Biomedicals, Inc. and Hoffman-LaRoche, Inc. JPM was the recipient of an NSF predoctoral Fellowship.

vol. 42, No. 22, 1988

TF5 StimulatesACTH Secretion

2267

References

1

J.P. MCGILLIS, N.R. HALL, G.V. VAHOUNY, and A.L. GOLDSTEIN,J. Immunol. 134, 3592-3955 (1985). 2 D.L. l-W&Y, G.D. HODGEN, H.M. SCHULTE, G.P. CHROUSOS, S. LORIAUX, N.R. HALL, and A.L. GOLDSTEIN, Science222, 1353-1355 (1983). H-0. BESEDOVSKY,A. DEL REY, and E. SORKIN, J. Imunol. 126, 385-387(1981) 3 H.O. BESEDOVSKY,A.DEL REY, E. SORKIN, and C.A. DINARFLLO, Science 233, 4 652-654 (1986). C.E.M. SAJOUS, In: The Internal Secretionsand the Principlesof 5 Medicine, Vol. 1, pp. 171-185,F.A. Davis, Co., Philadelphia(1903). H.L. JAFFE, J. Exp. Med. 50, 325-342 (1942). D. MARINE, O.T. MANLEY, and E.J. BAUMEN, J. Exp. Med. 60, 439-443 (1924). J. KOMSA, Nature 179, 872-873 (1957). H.O. BESEDOVSKY,and E. SORKIN, In: Psychoneuroimnology, B. ADER, (ed.), pp. 545-574, Academic Press, New York, (1981). 10 E.M. SMITH, D. HARBOUR MCMENAMIN, and J.E. BLALOCK, J. Inununol. -’135 7796-782s (1985). 11 N.R. HALL, J.P. MCGILLIS, B.L. SPANGELO,and A.L. GOLDSTEIN,J. Immunol. 135, 806s-811s (1985). 12 E.M. SMITH, and J.E. BLALOCK, Proc. Natl. Acad. Sci. USA 79, 7530-7534 (1981). 13 H.J. WFSTLY, A.J. KLEISS, K.W. KELLY, P.K.Y. WONG, and P.H. YUEN, J. Exp. Med. l&, 1589-1594 (1986). 14 G. ZURAWSKI,M. BENEDIK, B.J. KAMS, J.S. ABRAMS, S.M. SURAWSKI,and F.D. LEE, Science 232, 772-775 (1986). 15 E. HAZUM, J. CHANG, and P. CUATRECASAS,Science 205, 1033-1035 (1979). 16 K.K. OATES, and A.L. GOLDSTEIN,Trends Pharm. Sci., 5, 347-352 (1984). 17 T.L.K. LOW, and A.L. GOLDSTEIN,J. Biol. Chem. 254, 987-993 (1979). 18 J. CALDARELLA,G.J. GOODAL, A.M. FELIX, E.P. HEIMER, S.B. SALVIN, and B.L. HQRECKER, Proc. Natl. Acad. Sci. USA&, 7424-7427 (1983). 19 '1. T.L.K..a,LQW, and A.L. GOLDSTEIN,J. Biol Chem. 254, 1000-1003 (1982). 20 r E. HANNAPPEL, S. DAVOUST, and B.L. HORECKER, Proc. Natl. Acad. Sci. USA 79, 1708-1711 (1982). S. ERICKSON-VIITANEN,S. RUGGIERI, P. NATALINI, and B.L. HORECKER, Arch. 21 Biochem. Biophys.225, 407-410 (1983). 22 A.A. HARITOS, R. BLACHER, S. STEIN, J. CALDARELLA,and B.L. HORECKER, Proc. Natl. Acad. Sci. USA 82, 343-346 (1985). 23 A.A. HARITOS, S.B. SALVIN, R. BLACHER, S. STEIN, and B.L. HORECKER, Proc. Natl. Acad. Sci. USA, 82, 1050-1053 (1985) G.V. VAHOUNY, E. KYEYUNE-NYOMBI,J.P. MCGILLIS, N.S. TARE, K.Y. HUANG, 24 R. TOMBES, A.L. GOLDSTEIN,and N.R. HALL, J. Immunol.130, 791-794 (1983). 25 W. VALE, G.F. GRANT, F. AMOSS, R. BLACKWELL,and R. GUILLEMIN,Endo. -91, 562-572 (1972). 26 W. VALE, J. VAUGHAN, G. YAMAMOTO, T BRUGH, C DOUGLAS, D. DALTON, C. RIVIER, and J. RIVIER, Meth. Enz. 103, 565-577 (1983). P.C. GREENWOOD,W.M. HUNTER, and Jx GLOVER, Biochem J. g, 114-123 27 (1963). J.P. MCGILLIS, N.R. HALL, and A.L. GOLDSTEIN,J. Immunol. 131, 148-151 28 (1983). 2nd Ed., B.M. JAFFE, 29 D. ORTH, In: Methods of Hormone Radioinrmunoassay, and J.R. BEHRMAN (eds.),pp. 245-284 (1979). 30 R.L. MOLDOW, and A.J. FISCHMAN, Peptides,2, 37-39 (1982). E.M. SMITH, M. PHAN, T.E. KRUGER, D.H. COPPENHAVER,and J.E. BLALOCK, 31 Proc. Natl. Acad. Sci. USA so, 6010-6013 (1983). 32 w. VALE, J. VAUGHAN, M. SMITH, G. YAMAMOTO, J. RIVIER, and C. RIVIER, Endo. 113, 1121-1131 (1983).

2268

33 34 35 36 37

TF5 StimulatesACTR Secretion

Vol. 42, No. 22, 1988

W. VALE, J. SPIESS, C. RIVIER, and J. RIVIER, Science 213, 1394-1397 (1981). J.M. FARAR, J.F. BISHOP, J.P. MCGILLIS, N.R. HALL, A.L. GOLDSTEIN, and T.L. O'DONAHUE, Society for NeuroscienceAbs., p. 886 (1985). B.M.R.N.J.WOLOSKI, E.M. SMITH, W.J. MEYER, G.M. FULLER, and J.E. BLALOCK, Science 230, 1035-1037 (1985). J. AKELROD, and T.D. REISINE, Science 224, 452-459 (1984). A. BRODISH, Ann. N.Y. Aced. Sci. 297, 420-435 (1977).