Treatment of rats with thyrotropin (TSH) reduces the adrenoceptor sensitivity of adenylate cyclase from cerebral cortex

Neurochem. Int. Vol. 10, No. 2, pp. 173-178, 1987 Printed in Great Britain. All rights reserved

0197-0186/87 $3.00+ 0.00 ~) 1987PergamonJournals Ltd

TREATMENT OF RATS WITH THYROTROPIN (TSH) REDUCES THE ADRENOCEPTOR SENSITIVITY OF ADENYLATE CYCLASE FROM CEREBRAL CORTEX JOACHIM E. SCHULTZ* and BERNARD H. SCHMIDT Pharmazeutisches Institut, Universit~itTiibingen, Morgenstelle 8, 7400 Tiibingen, F.R.G. (Received 9 July 1986; accepted 20 August 1986)

Abstract--Prolonged treatment of rats with TSH attenuates the noradrenaline (NA) sensitivity of the cerebral cortical cAMP generating system. This effect can be measured in slices from brain cortex and with a membraneous, cellfree adenylate cyclase preparation from the same brain region. The development of this downregulation is time (7 d treatment required) and dose dependent (EC50= 3.5 mU/kg for 9 d). After discontinuation of treatment, about 12 d were required for reversal of this effect. Density of fl-adrenoceptors in cerebral cortex as measured by (-)[3H]dihydroalprenolol (DHA) binding was reduced by TSH treatment (22% reduction). The catalytic activity of adenylate cyclase as tested with guanosinemonophosphate phosphoimidophosphate (GMPPNP) was unaffected by TSH treatment. It is suggested that TSH-mediated feedback mechanisms on TRH levels and receptors in brain are responsible for the changes of the central adrenoceptor system. The data are discussed with respect to the blunted TSH-response to TRH observed in certain psychiatric illnesses.

Modulation of adrenergic receptor systems in animals after in vivo treatment with neurotropic drugs has been observed for more than a decade. Several treatment regimes believed to relieve manicdepressive disorders in man lead to a decrease in DHA binding sites in cortical membranes and a reduction of NA elicited cAMP formation in rat cerebral cortex (for review see Sulser et al., 1978; Kopanski et al., 1983). These findings resulted in an extension of the catecholamine hypothesis of affective disorders which emphasized a correlation between increased central adrenergic activity and the depressive state (Schildkraut, 1965; Sulser et al., 1978). The mechanism of action of many antidepressant drugs is now discussed in terms of a specific downregulation of the fl-adrenoceptor system (Sulser et al., 1978; O'Donnell and Frazer, 1985). Modulation of adrenoceptor activity in vivo appears also to occur within certain limits without overtly pathologic consequences, since changes in the endocrinological balance seem to affect the functional level of adrenergic neurotransmission. One such endocrine system is the thyroid (Whybrow et al., 1969). An association between thyroid gland function, the concurrent mental state and basic adrenergic mechanisms has been acknowledged since the earliest descriptions of thyroid dysfunction (Libow and *To whom correspondence should be addressed.

Durell, 1965). Therefore, the thyroid hormone system, comprising hypothalamic thyrotropin-releasing hormone (TRH), pituitary TSH, and thyroidal triodothyronine (T3) and thyroxine (T4) has been investigated in clinical studies of various psychiatric syndromes (Kirkegaard et al., 1978; Langer et al., 1984; Whybrow and Prange, 1981). Concomitant treatment of patients with imipramine and T3 or TRH has been proposed for some cases of depression (Prange et al., 1970; 1972; Furlong et al., 1976). The TSH response pattern to i.v. TRH stimulation was suggested as a diagnostic indication of recovery during antidepressant treatment (Kirkegaard et al., 1978; Langer et al., 1983; 1984). We reported that long-term treatment of rats with T4 downregulates the noradrenergic cAMP generating system in cerebral cortex (Schmidt and Schultz, 1985). Here, we evaluated the effect of a prolonged TSH treatment on the same parameters, i.e. fl-adrenoceptor number and NA stimulated cAMP formation. We also established that cellfree particulate adenylate cyclase from cerebral cortex can be used instead of brain slices to study the effects of antidepressant treatment on the adrenoceptorregulated adenylate cyclase. EXPERIMENTAL PROCEDURES

Pretreatment of animals. Male Sprague-Dawley rats, initial weight approximately 150 g, (from Jautz,. Kisslegg, 173

174

JOACHIM E. SCHULTZ a n d BERNARD H. SCHMIDT

F.R.G.) were treated i.m. once daily between 8 and 10 a.m. with TSH dissolved in 0.9% NaC1 solution. No toxic side effects were observed during the treatment. Controls received vehicle only. Animals were kept on a 12 h dark-light cycle and had free access to tap water and a standard laboratory chow. Body weight was monitored daily. Experiments with brain slices. 24 h after the last TSH administration animals were sacrificed by decapitation and brain cortical slices (2601tm) were prepared with a Mcllwain tissue slicer within 5 min. Slices were incubated and washed in Krebs-Ringer bicarbonate buffer as described earlier (Schultz and Daly 1973a). The final incubations contained in 2mL 100#M 1-NA, 600/~M ascorbic acid, and 1 mM of the phosphodiesterase inhibitor 1-methyl-3-isobutylxanthine. Controls were with ascorbic acid and inhibitor only. After a 10 min incubation at 37'C 1 mL ice-cold 0.4 N perchloric acid was added. Purification and determination of cAMP were carried out as described (Schultz and Daly, 1973a). The values were corrected for yield determined with an internal radioactive cAMP standard. Each brain was analyzed in duplicate for basal levels and in triplicate for the repsonse to 100,uM NA. Determination o f fl-adrenoceptor density. Binding of the fl-adrenoceptor antagonist [3H]DHA was determined at five concentrations ranging from 1 to 10 nM as descirbed (Alexander et al., 1975). Unspecific binding was measured in the presence of 10/~M (_)-propranolol. It was not affected by up to 100#M (+)-propranolol. Adenylate cTclase in cerebral cortical membranes. Cerebral cortices were quickly dissected and homogenized with 48mM Tris-HCl, pH 7.4, 12mM MgC12, and 0.1 mM EGTA in a Dounce homogenizer by 15 strokes with a loose fitting pestle and 15 strokes with a tight fitting one. The homogenate was centrifuged for 15min at 12,000g. The pellet was dispersed in 5 mL 1 mM KHCO 3 and centrifuged again for 10min at 12,000g. The pellet was suspended in 2.5 mL 1 mM KHCO 3 for each cerebral cortex and was used immediately. The adenylate cyclase assay carried out as described earlier (T/irck et al., 1980) contained 40 mM Tris-HC1, pH 7.4, 1 mM MgCI 2, 0.I1 mM CaC12, 0.1 mM EGTA, 1 mM methylisobutylxanthine, 14 mM kreatinphosphate, 70 EU kreatin kinase and 80 to I00/tg protein. The reactions of 250/~ L were started by addition of ct-[32p]ATP (0.5/iCi, 250FLM) and proceeded for 10min at 37'C. Purification of cAMP was carried out as described (Salomon et al., 1974) using [3H]cAMP to monitor yield. Data were calculated as pmol cAMP formed/min/mg protein. Total and free plasma concentrations of T3 and T4 were kindly determined by Dr Wahl, Laboratory of Clinical Chemistry, University of Tiibingen, using a radioimmunoassay obtained from Corning (IMA, Giessen, F.R.G.). ED~j values were estimated graphically from respective plots. Protein was assayed by the Lowry method with bovine serum albumin as a standard. Statistical methods'. Data were analyzed with Student's t-test. Controls, usually 4 rats, were always run in parallel with the experiments. Since there were no significant differences in the NA responsivity of adenylate cyclase from these control groups, they were normalized and pooled for data analysis. Drugs and chemicals. TSH (-)-isoproterenol and ( - )-NA bitartrates were from Sigma, Munich. (+)-Propranolol was a gift from ICI-Pharmacia, Planckstadt, F.R.G. All radiochemicals were purchased from Amersham International.

RESULTS

A 9 d treatment o f rats with TSH ( 1 0 0 m U / k g ) significantly reduced the gain in body weight o f 8 g/d by a b o u t 25% c o m p a r e d to placebo treated controls. This effect o f TSH was dependent on the daily dose administered. After discontinuation o f TSH application the normal growth rate o f the animals resumed. In this context, we also measured blood plasma levels o f b o u n d and free T3 and T4 from rats treated for 9 d with 100 m U / k g TSH. No differences in T3 and T4 plasma concentrations between TSH-treated and control animals were found (n = 11). In almost all reports on hormonally regulated c A M P formation in brain tissue from young adult animals brain slices were used as experimental system. With this system the n u m b e r of individual incubations per animal brain is restricted to about five because o f the limited a m o u n t of properly dissected tissue and the minimal tissue requirements for appropriate handling during the preincubation and washing procedure (for details see Schultz and Daly, 1973a). This makes that type o f experiment very laborious and expensive since a considerable number o f experimental animals has to be used to arrive at a reasonably sized statistical sampling. Therefore, wc c o m p a r e d the N A sensitivity o f the c A M P synthesizing system in the brain slice and in a cellfree m e m b r a n e o u s adenylate cyclase from the same region after treatment with different drugs (Table 1). The N A responsiveness o f the cellfree adenylate cyclase was consistently much lower to that o f the brain slice.

Table 1. Comparison of the effect of pretreatment of rats with various drugs on the NA sensitivity of the cerebral cortical cAMP generating system in tissue slices and celt membranes NA-stimulated cAMP formation (% of unstimulated control incubations) Kind of treatment Sham Imipramine (30 mg/kg) Rolipram (10 mg/kg) Tiflucarbine ( 15 mg/kg) TSH (100 mU/kg)

Cerebral cortical slices

Celffree adenylate cyclase in cerebral cortical membranes

265 + 5.3 (27) 169 + 13.4" (4)

152 + 9.6 (25) 126 + 2* (6)

201 + 8.1" 16)

108 -. 3.6* (5)

188 + 19.2" (6)

115 + 2.4* (4)

207 ± 7.8* (12)

130 + 1.5" (12)

cAMP levels in unstimulated control incubations of cerebral cortical slices were 20.5 _+2 pmol/mg. Basal adenylate cyclase activity in cerebral cortical membranes was 167 _+ 19pmol/min/mg protein. Data with imipramine are from Tfirck et al. (1981), with rolipram from Schultz and Schmidt (1986), with triflucarbine from Schmidt and Schultz (1986). The daily dose administered for 9 consecutive days is indicated in brackets. *2P < 0.01 compared to corresponding sham treated group.

175

TSH effect on adrenergic adenylate cyclase in brain 220

A t~n

E

x.

E

"~ 200 -S

~ 18o O

~ 16o - "

[I

0 0.1

[,

I

1

I

4

L

10

40

I

100

Norodrenoline (#M)

Fig. l. Dose-response curve for NA-stimulated adenylate cyclase in membranes from cerebral cortex of either control (I--]--[]) or TSH-treated (@--0) rats. TSH was administered once daily i.m. at a dose of 100 mU/kg for 9 d. Values represent X _ SEM of 8 independent experiments. *2P < 0.01; **2P < 0.05. Yet, probably because of the more controllable incubation conditions of the cellfree system, the reproducibility of parallel incubations of a membraneous adenylate cyclase and of individual preparations from different animals was considerably greater than that of brain slices. The reduced NA sensitivity after antidepressant treatment was retained in the cellfree adenylate cyclase (Table 1). Concerning the TSH treatment, we found that 100 mU/kg for 9 d reduced the NA responsiveness of the cAMP generating system significantly in both, cerebral coritical slices and cellfree adenylate cyclase from brain cortex (Table 1). Because of the better practicability of the membraneous adenylate cycalse, e.g. complete dose-response curves for NA can be carried out starting with a single rat brain cortex, we continued our studies with this type of preparation. The downregulation induced by long-term TSH treatment was characterized by a reduced NA sensitivity of the adenylate cyclase system (Fig. 1). As indicated by the identical EDs0 values for NA in vehicle and TSH-treated animals (4.5 # M) the affinity of the adrenoceptors for the catecholamine was not affected (see also Fig. 4). The response to NA was due to stimulation of fl-adrenoceptors since it could be mimicked by the agonist isoprenaline and was completely blocked by 50#M of the antagonist

(_+)-propranolol, In brain slices a small (about 15% of the total), yet very reproducible ct-adrenoceptor mediated stimulation of cAMP formation is observed (Perkins and Moore, 1973; Schultz and Daly, 1973b). Such an effect was not observed using the cellfree adenylate cyclase from cerebral cortex (data not shown). The possibility exists that TSH reduced the amount of the enzymatic entity of adenylate cyclase or that it interfered in the coupling of the receptor with the adenylate cyclase by the GTP-binding regulatory proteins. We therefore measured adenylate cyclase activity from TSH- and sham-treated animals with 100#M of the nonhydrolyzable GTP analog GMPPNP and with 5 mM NaF. No differences between both groups were detected. GMPPNP stimulated adenylate cyclase activities were 310 _+ 11.3 vs 309.8 _+ 11.5 pmol cAMP/min/mg protein, and NaF activated activities were 707 +_ 33 vs 697 +_ 57 pmol cAMP/min/mg protein for control and TSH-treated groups, respectively (n = 4). This is evidence against a direct effect of TSH treatment on receptorindependent adenylate cyclase. Also, addition of TSH to the adenylate cyclase assays (10 mU/mL) had no effect on basal or stimulated cAMP formation. Figure 2 demonstrates the slow onset of the development of adrenoceptor subsensitivity after TSH administration. A significant decrease of NA re-

..E 100

= o

80

/ / /

,~

/

'

.

,¢,

/

lh

5

5

7

9

14

9+5

9+12

Days of TSH treatment

Fig. 2. Effect of various periods of daily administration of TSH on the NA-stimulated adenylate cyclase in rat cerebral cortical membranes. TSH (100 mU/kg) was given i.m. for 1 h and for 1 to 14d. With the exception of the 1 h pretreatment all experiments were carried out 24 h after the last dose. In the experiments depicted by the shaded bars TSH treatment was discontinued after day 9 and rats were sacrificed after day 5 and 12, respectively. Values are percents of NA-stimulated adenylate cyclase (+SEM) of saline treated controls (number of animals analyzed = 4-8 per group). *2P < 0.01.

176

JOACHIM E. SCHULTZ

60

~_ o

50

o)

i! -£

and

BERNARD H, SCHMIDT

5 0 0 --

--

400

--

300

,1, E

o

{3c o

Jo
E•

o

[] o

200

°°

30 1 O0

Z

o

o

i 20 [~L. 0

[ 0.5

I _ _ 1 1 TSH

I 10

I

J 100

(mU/kg)

Fig. 3. Dose response curve for TSH. All animals were treated for 9 d i.m. with the TSH doses indicated at the abscissa. Experiments were carried out 24 h after the last drug application, adenylate cyclase in cerebral cortical membranes was stimulated by 100#M NA. Unstimulated adenylate cyclase activity of 218 + 3.3 pmol/min/mg is subtracted in this graph. Each point reflects X _+ SEM of four animals analyzed in triplicates. *2P < 0.01 and **2P < 0.05 compared to the sham-treated control group.

sponsiveness was evident only after a 5 d treatment period, but not after an acute dose 60 rain prior to sacrifice and cerebral cortical membrane preparation. Complete recovery was observed 12 d after discontinuation of treatment (Fig. 2). The recovery phase required a longer period of time compared to that observed with antidepressants or T4 (Sulser et al., 1978; Schmidt and Schultz, 1985, 1986; Schultz and Schmidt, 1986). Initially carried out with brain slices, we found that the downregulation of NA-stimulated c A M P formation occurred already after a dose of 20 m U / k g for 9 d. A more detailed dose-response relationship for TSH pretreatment was established using the cellfree adenylate cyclase (Fig. 3). A maximal effect was obtained already with 1 0 m U TSH/kg, the EDs0 as determined graphically from Fig. 3, was 3.5 mU/kg. Downregulation of NA-stimulated c A M P formation is often accompanied by a concomitant decrease in/~-adrenoceptor density, although an absolutely stringent correlation does not exist (Mishra et al., 1980; Schmidt and Schultz, 1985). Therefore, we determined binding of the/~-adrenoceptor antagonist [3H]DHA in membranes from cerebral cortex of rats treated with 1 0 0 m U T S H / k g for 9d. A significant

I 0

20 DHA

40

60

80

bound (fmol/free(nM)/mg)

Fig. 4. Scatchard analysis of specific [3H]DHA-binding to membranes from cerebral cortex prepared from rats treated with 100mU TSH/kg ( O - - O , n = 5) or saline ( O - - O , n = 5) for 9 d. The values are corrected for unspecific binding determined in the presence of 10pM (_.+)-propranolol. The lines drawn were fitted by a computer (all correlation coefficients were >0.91). The Bmax was 472 + 15 fmol bound DHA/mg protein for saline treated controls and 364 + 12 fmol/rng for TS-treated animals, respectively (2P<0.001). The dissociation constants were 5.8 + 0.6 and 5.6 _+0.7 nM, respectively.

reduction of Bma x w a s found compared to placebotreated controls (364 + 12 vs 472 + 15 fmol D H A bound/mg protein, 2P < 0.001; n = 6 for each group). The almost parallel lines in the Scatchard analysis (Fig. 4) reveal that the dissociation constant was unaltered by the TSH treatment (5.8 + 0.6 nM of drug-treated vs 5 . 6 + 0 . 7 n M of sham treated animals). D H A binding was not affected by addition of TSH (10 m U / m L ) or T R H (100/~M) to the incubation. DISCUSSION

There seems to be no doubt that neuroendocrine factors play an important role in psychobiology, Clinically, a blunted TSH response to an exogenous T R H challenge is a firmly established diagnostic marker of psychiatric illnesses such as depressions (Kastin et al., 1972; Prange et al., 1972; Kirkegaard et al., 1978; Langer et al., 1984; Loosen, 1985). Normalization of this T R H response ( " d i s b l u n t i n g ' ) in patients treated with psychotropic drugs is suggested as indicator for lasting recovery (Langer et al.,

TSH effect on adrenergic adenylate cyclase in brain 1983, 1984). A general consent exists that the physiological manifestations of altered thyroid states may be mediated through modulation of peripheral and central adrenoceptor systems (Greene, 1976; Sitaghal et al., 1975; Gross et al., 1980). However, we do not know in sufficient detail the biochemical sites where and how thyroid hormones and catecholamines intersect in the central nervous system. Our data may point toward that direction. TSH treatment of rats downregulates the fl-adrenoceptor regulated cAMP generating system in cerebral cortex and diminishes fl-adrenoceptor density much like several proven antidepressant drugs and T3 and T4 (Kopanski et al., 1983; Schmidt and Schultz, 1985). THS does so without a significant and lasting alteration of T3 and T4 plasma levels. Probably, TSH has several effects in brain in addition to those on the thyroid gland. Of particular interest may be its feedback influence on hypothalamic TRH which it should share with T3 and T4. While it is unclear whether all brain TRH originates from the hypothalamus, it is increasingly evident that TRH, a tripeptide which produces endocrine effects, exerts direct behavioral effects and may modulate the action of other brain neurotransmitter substances (Rastogi et al., 1981; Garbutt and Loosen, 1984). Specific and regionally differentiated binding sites have been identified in the brain (Morley, 1979; Simasko and Horita, 1984). Thus, it seems conceivable that in the adult rat the actions of TSH and of T4 treatment (Schmidt and Schultz, 1985) are mediated via TRH sensitive pathways and do not constitute direct effects on central adrenoceptors. The blunted TSH response in certain psychiatric illnesses can be considered as an indicator of either a dysfunction of central TRH receptors or of an improper secretory mechanism for TSH from its pituitary storage sites. The first alternative seems to be the more attractive one. The normalization of the blunted TSH-response in patients during psychotropic drug treatment would, although of diagnostic value, have no direct involvement in the etiology of psychiatric illnesses. Conversely, malfunctioning TRH receptors in brain regions other than the pituitary gland which are coupled in a modulatory manner with different neurotransmitter receptors, may constitute a link to pathopsychological states. Indeed, TRH administration has been demonstrated to reduce rat cerebral fl-adrenoceptor density, to potentiate the effects of imipramine on rat brain serotonergic systems, and to act synergistically with psychotropic drugs in patients (Rastogi et al., 1981; Schmidt and Schultz, 1985; Garbutt and Loosen,

177

1984). Further, TRH levels increase in brain of rats treated chronically with antidepressants (Lighton et al., 1985). One could then speculate that the actions of catecholamines, antidepressants, and the thyroid hormone system in brain interphase at a common biochemical level: the phosphatidylinositolphosphate system. TRH and lithium have been shown to strongly affect the latter system (Berridge et al., 1982; Drummond et al., 1984). It is apparent that such an extended interpretation of our TSH data spotlights several questions which can be experimentally investigated in the future. Acknowledgement--This work was supported by the Deutsche Forschungsgemeinschaft. REFERENCES

Alexander R. W., Davis J. N. and Lefkowitz R. J. (1975) Direct identification and characterization of fl-adrenergic receptors in rat brain. Nature, Lond. 258, 437-440. Berridge M. J., Downes C. P. and Hanley M. R. (1982) Lithium amplifies agonist-dependent phosphatidylinositol responses in brain and salivary glands. Biochem. J. 2116, 587-595. Drummond A. H., BushfieldM. and Macphee C. H. (1984) Thryrotropin-releasing hormone-stimulated [3H]inositol metabolism in GH 3 pituitary tumor cells. Studies with lithium. Molec. Pharmac. 25, 201-208. Furlong F. W., Brown G. M. and Beeching M. F. (1976) TRH: differential antidepressant and endocrinological effects. Am. J. Psychiat. 133, 1187-1190. Garbutt J. C. and Loosen P. T. (1984) A dramatic behavioral response to TRH following low-dose neuroleptics. Psychoneuroendocrinology 9, 311-314. Greene R. (1976) The thyroid gland: its relationship to neurology. In: Handbook of Clinical Neurology, (Vinken P. J. and Bruyn G. W., eds), Vol. 27, pp. 255 277. North Holland, Amsterdam-New York. Gross G., Brodde O.-E. and Schiimann H.-J. (1980) Effects of thyroid hormone deficiency on pre- and postsynaptic noradrenergic mechanisms in the rat cerebal cortex. Arch. Int. Pharmacodyn. 244, 219-230. Kastin A. J., Ehrensing R. H., Schalch D. S. and Anderson M. S. (1972) Improvement in mental depression with decreased thyrotropin-response after administration of thyrotropin-releasing hormone. Lancet 740-742. Kirkegaard C., Bjorum N., Cohn D. and Lauridsen U. B. (1978) Thyrotropin-releasing hormone (TRH) stimulation test in manic-depressiveillness.Archs gen. Psychiat. 35, 1017-1021. Kopanski C., Tiirck M. and Schultz J. E. (1983) Effects of long-term treatment of rats with antidepressants on adrenergic receptor sensitivity in cerebral cortex: structure activity study. Neurochern. Int. 5, 649~59. Langer G., Aschauer H., Koinig G., Resch F. and Sch6nbeck G. (1983) The TSH-response to TRH: a possible predictor of outcome to antidepressant and neuroleptic treatment. Prog. Neuropsychopharmac. Biol. Psychiat. 7, 335-342. Langer G., Resch F., Aschauer H., Keshaven M. S., Koinig

178

JOACHIM E. SCHULTZ and BERNARD H. SCHM1DT

G., Sch6nbeck G. and Dittrich R. (1984) TSH-response patterns to TRH stimulation may indicate therapeutic mechanisms of antidepressant and neuroleptic drugs. Neuropsychology 11, 213-218. Libow L. S. and Durell J. (1965) Clinical studies on the relationship between psychosis and the regulation of thyroid gland activity. Psychosom. Med. 23, 369--376. Lighton C., Bennet G. W. and Marsden C. A. (1985) Increase in levels and ex vivo relase of TRH in specific regions of the CNS of the rat by chronic treatment with antidepressants. Neuropharmacology 24, 401~406. Loosen P. T. (1985) The TRH-induced TSH response in psychiatric patients: a possible neuroendocrine marker. Psychoneuroendocrinology 10, 237 260. Mishra R., Janowsky A. and Sulser F. (1980) Action of mianserin and zimelidine on the norepinephrine receptorcoupled adenylate cyclase system in brain: subsensitivity without reduction in fl-adrenergic receptor binding. Neuropharmacology 19, 983-987. Morley J. E. (1979) Extrahypothalamic TRH--its distribution and its function. Life Sci. 25, 1461 1465. O'Donnell J. M. and Frazer A. (1985) Effects of clenbuterol and antidepressant drugs on fl-adrenergic receptor/N-coupling in the cerebral cortex of the rat. J. Pharmac. exp. Ther. 234, 3~36. Perkins J. P. and Moore M. M. (1973) Characterization of the adrenergic receptors mediating a rise in cyclic 3',5'-adenosine monophosphate in rat cerebral cortex. J. Pharmac. exp. Ther. 185, 371-378. Prange A. J., Wilson I. C., Knox A., McClane T. K. and Lipton M. A. (1970) Enhancement of imipramine by thyroid stimulating hormone: clinical and theoretical implications. Am. J. Psychiat. 127, 191 199. Prange A. J., Wilson I. C., Lara P. P., Alltop L. G. and Breese G. R. (1972) Effects of thyrotropin-releasing hormone in depression. Lancet 999 1002. Rastogi R. B., Singhal R. L. and Lapierre Y. D. (1981) Thyrotropin releasing hormone potentiates the effects of imipramine on brain serotonergic system. Brain Res. Bull. 7, 449~151. Salomon Y., Londos C. and Rodbell M. (1974) A highly sensitive adenylate cyclase assay. Anal. Biochem. 58, 541 548. Schildkraut J. J. (1965) The catecholamine hypothesis of

effective disorders: A review of supporting evidence. Am. J. Psychiat. 122, 509-522. Schmidt B. H. and Schultz J. E. (1985) Chronic thyroxine treatment of rats downregulates the noradrenergic cyclic AMP generating system in cerebral cortex. J. Pharmac. exp. Ther. 233, 466~72. Schmidt B. H. and Schultz J. E. (1986) The potential antidepressant tiflucarbine downregulates fl-adrenoceptors in rat brain. Eur. J. Pharmac. 130, 27 36. Schultz J. E. and Daly J. W. (1973a) Cyclic adenosine 3',5'-monophosphate in guinea pig cerebral cortical slices. I. Formation of cyclic adenosine 3',5'-monophosphate from endogenous and from radioactive adenosine triphosphate formed during a prior incubation with radioactive adenine. J. biol. Chem. 248, 843 852. Schultz J. E. and Daly J. W. (1973b) Adenosine Y,5'-monophosphate in cerebral cortical slices from rat and mouse: stimulatory effect of alpha- and betaadrenergic agents and adenosine. J. Neurochem. 21, 573-579. Schultz J. E. and Schmidt B. (1986) Rolipram, a stereospecific inhibitor of calmodulin-independent phosphodiesterase, causes fl-adrenoceptor subsensitivity in rat cerebral cortex. Naunyn-Schmiedebergs Arch. Pharmac. 333, 23 30. Simasko S. M. and Horita A. (1984) Localization of TRH receptors in the septal nucleus of the rat brain. Brain Res. 296, 393-395. Singhal R. L., Rastogi R. B. and Hrdina P. D. (1975) Brain biogenic amines and altered thyroid function. Life Sci. 17, 1617--1626. Sulser F., Vetulani J. and Mobley P. L. (1978) Mode of action of antidepressant drugs. Biochem. Pharmac. 27, 257 261. T~rck M., Schocker W. F., Fathy A. M. and Schultz J. E. (1980) Adrenergic subsensitivity of a cell-free adenylate cyclase from rat brain after chronic imipramine treatment. Archs Pharmacy 313, 768 773. Whybrow P. C. and Prange A. J. (1981) A hypothesis of thyroid-catecholamine-receptor interaction. Archs gen. Psychiat. 38, 10¢~113. Whybrow P. C., Prange A. J. and Treadway C. R. (1969) Mental changes accompanying thyroid gland dysfunction. Archs gen. Psychiat. 20, 48 63.