Acta psychiat. scand. (1978) 57,447453 Department of Psychiatry (Head: Prof. C.N. Stefanis), Athens University Medical School, Eginition Hospital, Athens, Greece
Plasma cyclic AMP in manic-depressive illness E.LYJCOUIUS, E. VARSOU, E. GARELIS,C. N. STEFANIS AND D. MALLIAR~S The postulated disturbance of cyclic AMP (CAMP) in manic-depressive illness was investigated by using plasma as the biological material. Cyclic AMP was measured by a protein-binding assay, which was found very satisfactory for the purpose of this study. In the drug-free state, depressed patients (n = 28) had significantly lower and manic patients (n = 9) significantly higher plasma concentrations of CAMP than controls. Unrnedicated manic-depressive subjects had normal cAMP levels during normothymic phases (n = 7). Cyclic AMP was reduced by neuroleptics in mania and elevated by tricyclics in depression. Lithium exerted a normalizing effect on cAMP in both phases of the illness. It is concluded that manic-depressive illness is associated with a disturbance in the cAMP system. The use of plasma rather than urine for the investigation of the state of cAMP in psychiatric disorders is advocated.
Key w o r k : Cyclic AMP
- plasma - manic-deprssive illness.
In addition to its well-known role as a “second messenger” of many hormones (Robison et al. (1971)), adenosine 3: 5’-cyclic monophosphate (cyclic AMP, CAMP) seems to be involved in the function of the central nervous system (CNS) in several ways, affecting both presynaptic and postsynaptic events in neurotransmission (Bloom (1975), Duly (1975), Nuthanson (1977)). Psychotropic drugs interfere with the formation or breakdown of cAMP in many tissues, including the brain (Uzunov & Weiss (1972)), and this action has been proposed to be related to the mode of action of neuroleptics (Zversen (1975), Greengurd (1976)) and minor tranquilizers (Beer et al. (1972)). Since 1970, a series of reports have associated this nucleotide with manicdepressive illness (Abdulla (1970), PauZ et aZ. (1970, 1971), Naylor et ul. (1974), Sinanan et al. (1975)). Generally, in these studies CAMP excretion in urine was found low in depression and high in mania. On the assumption that changes in urinary CAMP might reflect similar changes in the brain, it was suggested that the cAMP system is disturbed in affective disorders. Such an assumption, however, is at variance with the existing evidence that much of the urinary CAMP is of renal origin (Broadus et al. (1970)). This contribution varies greatly among normal individuals, but may well exceed 50 % of the total cAMP concentration (Broadus et al. (1970), Kaminsky et al. (1970)). Moreover, only 1520 % of plasma CAMP is filtered into the urine, the rest being disposed of by
448
other mechanisms (Broadus et al. (1970)). It is thus clear that even a significant contribution to plasma cAMP from the brain would be largely obscured in urine. This has actually been directly demonstrated in the case of cerebral infarction, where increases of CAMP in plasma and CSF were high, whereas a negligible elevation was found in urine (Welch et al. (1975)). Similarly, changes of cAMP levels in plasma and urine are not parallel after treatment with hormones which increase extracellular levels of the nucleotide (Ball et al. (1972), Hamet et al. (1975)). Measurements in urine are subject to additional disadvantages. The amount of cAMP excreted may depend on the amount of creatinine and some processing of the samples is generally necessary (Murud (1973)). These factors indicate that urine is not a very appropriate material for the investigation of the cAMP system in psychiatric disorders and may explain the inability of some authors to replicate the reported changes in urinary cAMP of manic-depressive patients (Brown et al. (1972), Jenner et al. (1972)). Unlike urine, plasma represents a pool of CAMP deriving from many organs (Wehmann et al. (1974)), including the CNS (Broadus et al. (1970)). Concentrations in plasma are very stable in normal persons (Lykouras et al. (1978a)) and currently available methods can accurately measure cAMP in unprocessed plasma (Tovey et al. (1974)). Because of the above considerations, plasma rather than urine was used in the present study on the possible association of cAMP with manic-depressive illness. PATIENTS AND METHODS Forty-four m e d i c a t e d manic-depressive patients from the University Psychiatric Clinic at Eginition Hospital, or its outpatient department, were included in this study. The criteria of selection were as follows: 1. Agreement on the diagnosis of manic-depressive illness by two staff psychiatrists. Both bipolar and unipolar cases were included, but patients with involutional, neurotic or “atypical” depression were excluded from the study. 2. Age below 70 years. 3. Absence of clinical or laboratory evidence of physical disease. 4. Absence of medication for at least 3 weeks prior to the study (drug-free group). The effect of drugs on plasma CAMP was studied in 46 patients under treatment with tricyclic antidepressants (n = 22), lithium (n = 15) or neuroleptics (n = 9). The control group consisted of 76 healthy volunteers, age-matched to the patients (range 20-66 years). Blood was obtained from all subjects between 8.30 and 10 a.m. after overnight fasting. Samples were immediately transferred into tubes containing 0.5 M EDTA, 1 % v/v of blood. Plasma was then separated by centrifugation at 2800 r e v h i n for 10 min, and was kept frozen (-20°C) until analyzed. In our experience cAMP remains stable under these conditions for at least 1 year. De-
449
Table 1. Drugs not uflecting 3H-cAMP binding Dmg Diazepam Chlorpromazine Haloperidol Desipramine Li,CO,
Concentrations 7.10-M 1.4 X 1 P M 6.6 X l P M 5.3 X 1WM 0.1 mEqA
7.1WM 1.4 x 1 0 - 5 ~ 6.6 X 1WM 5.3 X l P M 1 mEqA
7.lPM 1.4 x 1 0 - 4 ~ 6.6 X 10dM 5.3 x 1 0 - 5 ~ 10 mEqA
terminations of cyclic AMP were made by the protein-binding method of T o v e y et ul. (1974) (“cyclic AMP assay kit”, Amersham, England). All samples were assayed in duplicate. Reproducibility of the method was very satisfactory. The coefficient of variation for eight measurements of six samples was 2-5.3 %. Recovery of 0.5 or 1 pmol of pure cAMP added to plasma was 95-97 % and therefore values were not corrected for recovery. Several factors which might influence results by this method were checked before the main study. Hemolysis, which has been considered such a factor (Murud (1973)), had no effect; recovery of cAMP from plasma was not affected by the presence of hemolysate of washed red blood cells (diluted 1:2 to 1:lOO). Endogenous CAMP in the hemolysate was previously destroyed by endogenous phosphodiesterase after incubation at 37OC for 24 h. Because blood contains a highly active phosphodiesterase (PDE), it is possible that some of the cAMP is destroyed by this enzyme in the interval between venipuncture and inhibition of PDE by the addition of EDTA (Brown et ul. (1972)). However, we found no difference in cAMP content in serial samples of eight subjects, mixed with EDTA 30, 60, 90 and 120 sec after venipuncture. All samples for the present study were obtained within this time interval. No interference with the binding of 3H-cAMP to the specific binding protein of the assay was found in the presence of several representative psychotropic agents, in a wide range of concentrations (Table 1). RESULTS
No difference in cAMP concentrations between males and females was found. The correlation of CAMP with age (control group) failed to reach significance (r = 0.092, NS). Mean ages did not differ significantly among the groups of our subjects. Table 2 shows concentrations of cAMP in the drug-free patients. Cyclic AMP levels are much higher in mania and much lower in depression, in comparison with the control group (P < 0.001 in both cases). Normothymic patients had normal concentrations. No difference in cAMP concentrations was found between unipolar and bipolar depressed patients. Psychoactive medication generally tended to normalize cAMP concentrations in all groups (Table 3).
450 Table 2. Plasma CAMP (pmoUml f s.d.) in drug-free manic-depressive patients Diagnosis
n
Depression Mania Normothymia Controls
28 9 7 76
Cyclic AMP
Difference from controls*
14.9 f 2.7 28.1 k 3.2 21.8 f 0.7 21.1 f 4.1
P < 0.001 P P
< 0.001 < 0.1
* Two-tailed t-test. Table 3. Effect of drugs
on plasma CAMP
CAMP @moI/ml f s.d.)
n
Drugs
Tricyclics Neuroleptics Li,CO, Controls
21.3 f 2.4 22.1 k 3.4 20.1 k 3.1 21.1 k 4.1
22 9
15 76
Table 4. Normalizing effect of lithium on plasma CAMP (pmoNm1) in manic-depressive patients
Cases
Before lithium Phase of illness
1
2 3 4
Mania Mania Depression Depression
During lithium cAMP
Phase of illness
cAMP
32 28.8 17.8 12.5
Normothymia Normothymia Normothymia Normothymia
24 21.8 21 16.5
None of the difTerences in Table 3 is significant. Most of the depressed patients being treated with tricyclic antidepressants were in normothymia when specimens were obtained. In nine of these patients, blood was taken before and after tricyclic drugs were given. Cyclic AMP rose signiiicantly after treatment (from 14.3 k 2.4 to 20.9 k 1.8 pmol, P < 0.001). The group receiving neuroleptics (Table 3) consisted of manic patients who were being treated with phenothiazines or butyrophenones and had improved to some extent at the time of blood sampling. Comparison of cAMP levels of this group to the drug-free manic subjects (Table 3) showed a significant fall in cAMP after treatment (P < 0.01). Lithium-treated patients were generally normothymic and had normal levels of cAMP (Table 3). In a few cases, however, specimens were also available before administration of lithium. Results in these cases indicate that lithium “normalizes” cAMP in plasma, i.e. decreases levels in mania and increases them in depression (Table 4).
451 DISCUSSION The finding of the present study that cAMP in plasma is elevated in mania and decreased in depression parallels similar observations in the urine of manicdepressive patients. However, several non-specific factors have to be accounted for before an association between manic-depressive illness and disturbance of the cyclic AMP system can be concluded. Some of these factors, related to the procedure of the assay, have been checked and found not to interfere with results (see Methods). Another possibility is that changes in CAMP may reflect changes in the psychomotor activity of the patients, as suggested by findings in urine (Eccleston et al. (1970)). Investigating this possibility in a parallel study, we found that psychomotor activity can affect concentrations of cAMP in plasma, but it is not the only factor responsible for the changes reported here (Lykouras et al. (1978b)). It can thus be concluded that abnormal affect is associated with changes in plasma CAMP. It has been adequately substantiated that changes in plasma cAMP reflect intracellular levels of the nucleotide (Broadus et al. (1970), Strange et al. (1974), Wehrnann et al. (1974)), perhaps through increased membrane permeability (Gillet al. (1975)). It is unlikely that these changes are due to disturbed excretion mechanisms, because clearance rates of cAMP from plasma are constant within a wide range of concentrations (Zssekutz (1975)). The tissues where plasma cAMP originates have not been identified with certainty. No single peripheral organ, including the liver, the kidney, the lungs, the adrenals and formed blood elements, is a major source of plasma CAMP (Wehrnann et at. (1974)). It has been suggested that the central nervous system, which is one of the richest sources of cAMP (Robison et al. (1971)), may be an important site of origin of plasma cAMP (Broadus et al. (1970)); direct evidence of a contribution from the brain, however, has only been provided in the case of acute cerebral infarction (Welch et al. (1975)). It is thus premature to conclude that elevation or decrease of plasma CAMP in manic-depressive patients reflects similar changes in the brain of these subjects. Whether changes in cAMP levels are due to a primary disturbance in production or breakdown of the nucleotide, or they are secondary to altered hormonal stimulation of adenyl cyclase, is an open question. It should be mentioned, in this context, that injection of catecholamines increases plasma cAMP (Murud (1973)), and endogenous elevation of catecholamines in cardial infarction runs parallel to the increase of cAMP in plasma (Strange et al. (1974)). Such a connection would be consistent with the long-standing catecholamine theory on affective disorders. The observation that patients under tricyclic antidepressants have normal plasma cAMP confirms similar findings in urine (Paul et al. (1971), Sinanan et al. (1975)). As most of these patients had recovered from depression when specimens were taken, it is not clear whether it is the action of the drugs or remission of the symptoms that normalizes CAMP in depression.
452
On the other hand neuroleptics decreased cAMP in manic patients, even before recovery; a similar observation was made in schizophrenic patients receiving PimozideB (Stefunis et al. (1977)). This effect can readily be explained by the inhibitory action of neuroleptics on adenyl cyclase (Zversen (1975)). The effect of lithium is somewhat more complex. It seems to act in both directions, i.e. it increases cAMP in depression and lowers it in mania. This stabilizing action is analogous to the drug’s prophylactic effect in both phases of manic-depressive illness and its ability to prevent drug-induced supersensitivity of receptors in experimental models (Kluwuns et ul. (1977)). In conclusion, this study provides evidence in support of the existence of a relationship between affective disorders and cyclic AMP. It also shows that plasma is a more suitable biological material than urine for the investigation of the role of cAMP in psychiatric disorders. REFERENCES Abdulla, Y . H., & K . Hamadah (1970): 3’, 5’-cyclic adenosine monophosphate in depression and mania. Lancet i, 378-381. Ball, 1. H., N . 1. Kaminsky, J . G . Hardman, A . E. Broadus, E . W . Sutherland & G . W . Liddle (1972): Effects of catecholamines and adrenergic-blocking agents on plasma and urinary nucleotides in man. J. clin. Invest. 51, 2124-2129. Beer, B., M . Chasin, D.E. Clody, J . R . Vogel & 2.P. Horovitz (1972): Cyclic adenosine monophosphate phosphodiesterase in brain effect on anxiety. Science 176, 428-430. Bloom, F.E. (1975): The role of cyclic nucleotides in central synaptic function. Rev. Physiol. biochem. Pharmacol. 74,1-104. Broadus, A . E., N . I . Kaminsky, J . G . Hardman, E. W . Sutherlund & G . W . Liddle (1970): Kinetic parameters and renal clearances of plasma adenosine 3’, 5’ monophosphate in man. J. clin. Invest. 49, 2222-2236. Brown, B. L., J . G . Salway, J . D.M . Albano, R . P. Hullin & R. P. Ekins (1972): Urinary excretion of cyclic AMP and manic-depressive psychosis. Brit. J. Psychiat. 120, 405-408. Duly, J . (1975): The role of cyclic nucleotides in the nervous system. In Zversen, L. L., S. D . lversen & S. H. Snyder (eds.): Handbook of psychopharmacology, Vol. I. Plenum, New York, pp. 47-130. Eccleston, D., R . Loose, I . A . Pullur & R . F. Sugden (1970): Exercise and urinary excretion of cyclic AMP. Lancet ii, 612-613. Gill, G . V., K . Prudhoe, D.B. Cook & A . L. Latner (1975): Effect of surgical trauma on plasma concentrations of cyclic AMP and cortisol. Brit. J. Surg. 62, 441-443. Greengard, P. (1976): Possible role for cyclic nucleotides and phosphorylated membrane proteins in postsynaptic actions of neurotransmitters. Nature (Lond.) 260, 101-108. Hamet, P., S. C. Loweder, I . G . Hardman & G . W. Liddle (1975): Effect of hypoglycemia on extracellular levels of cyclic AMP in man. Metabolism 24, 1139-1144. Zssekutz, T . B. (1975): Estimation of cyclic AMP turnover in normal and methylprednisolone treated dogs: effect of catecholamines. Amer. J. Physiol. 229, 291297. Iversen, L. L. (1975): Dopamine receptors in the brain. Science 188, 1084-1089. Jenner, F. A., G . A . Sampon, G . A . Thompson, A . R. Somerville, N . A . Beard & A . A . Smith (1972): Manic-depressive psychosis and urinary excretion of cyclic AMP. Brit. J. Psychiat. 121, 236-237. Kaminsky, N.Z., A . E. Broadus, J . G . Hardman, D . 1. Jones, 1. H . Ball, E . W . Sutherland & G . W . Liddle (1970): Effects of parathyroid hormone on plasma and urinary adenosine 3’, 5‘-monophosphate in man. J. clin. Invest. 49, 2387-2395.
453 Klawans, H. L., W. 1. Weiner & P. A. Nausieda (1977): The effect of lithium on an animal model of tardive dyskinesia. Progr. Neuro-Psychopharmacol. 1, 53-60. Lykouras, E., E. Garelis & C. N.Stefanis (1978a): Diurnal and longitudinal variation of cyclic AMP in plasma. (Submitted for publication). Lykouras, E., E. Garelis, E. Varsou & C.N . Stefanis (1978b): Physical activity and plasma cyclic AMP in manic-depressive patients and healthy adults. Amer. J. Psychiat. (in press). Murad, E. (1973): Clinical studies and applications of cyclic nucleotides. In Greengard, P., & G. A. Robison (eds.): Advances in cyclic nucleotide research, Vol. 3. Raven Press, New York. Nathanson, J. A. (1977): Cyclic nucleotides and nervous system function. Physiol. Rev. 57, 157-256. Naylor, G. L., S.A. Stansfield, S. F. Whyte & F. Hutchinson (1974): Urinary excretion of adenosine 3‘, 5’ monophosphate in depressive illness. Brit. J. Psychiat. 125, 275279. Paul, M . I., H. Cramer & F. K. Goodwin (1971): Urinary cyclic AMP excretion in depression and mania. Arch. gen. Psychiat. 24, 327-333. Paul, M . I., B. R. Ditzion, G. L. Pauk & D. S. Janowsky (1970): Urinary adenosine 3‘, 5’ monophosphate excretion in affective disorders. h e r . J. Psychiat. 126, 14931497. Robison, G. A., R. W. Butcher & E. W . Sutherland (1971): Cyclic AMP. Academic Press, New York. Sinanan, K., A. M . B. Keatinge, P. G. S. Beckett & W . Clayton-Love (1975): Urinary cyclic AMP in “endogenous” and “neurotic” depression. Brit. J. Psychiat. 126, 49-55. Stefanis, C . N., E. Lykouras, E. Garelis & E. Varsou (1977): The effect of Pimaid& on plasma cyclic AMP in chronic schizophrenics. Prog. Neuropsychopharmacol. 1, 323-327. Strange, R . C., N. Vetter, M . J . Rowe & M . F. Oliver (1974): Plasma cyclic AMP and total catecholamines during acute myocardial infarction in man. Eurap. J. clin. Invest. 4, 115-119. Tovey, K. C., K.G. Oldham & J . A . M . Whelan (1974): A simple direct sample using an improved competitive binding technique. Clin. chim. Acta 56, 221-234. Uzunov, P., & B . Weiss (1972): Psychopharmacological agents and the cyclic AMP system of rat brain. In Greengard, P.,& G. A. Robison (eds.): Advances in cyclic nucleotide research, Vol. 1. Raven Press, New York, pp. 435453. Wehmann, R . E., L. Blonde & A . L. Steiner (1974): Sources of cyclic nuclwtides in plasma. 3. clin. Invest. 53, 173-179. Welch, K. M. A., J . S. Mayer & A. N . C.Chee (1975): Evidence f o r disordered cyclic AMP metabolism in patients with cerebral infarction. Europ. Neurol. 13, 144-154.
Received October 26, 1977
E. Lykouras, M.D. Department of Psychiatry Eginition Hospital 74, Vasilissis Sophias Avenue Athens Greece
Plasma cyclic AMP in manic-depressive illness E.LYJCOUIUS, E. VARSOU, E. GARELIS,C. N. STEFANIS AND D. MALLIAR~S The postulated disturbance of cyclic AMP (CAMP) in manic-depressive illness was investigated by using plasma as the biological material. Cyclic AMP was measured by a protein-binding assay, which was found very satisfactory for the purpose of this study. In the drug-free state, depressed patients (n = 28) had significantly lower and manic patients (n = 9) significantly higher plasma concentrations of CAMP than controls. Unrnedicated manic-depressive subjects had normal cAMP levels during normothymic phases (n = 7). Cyclic AMP was reduced by neuroleptics in mania and elevated by tricyclics in depression. Lithium exerted a normalizing effect on cAMP in both phases of the illness. It is concluded that manic-depressive illness is associated with a disturbance in the cAMP system. The use of plasma rather than urine for the investigation of the state of cAMP in psychiatric disorders is advocated.
Key w o r k : Cyclic AMP
- plasma - manic-deprssive illness.
In addition to its well-known role as a “second messenger” of many hormones (Robison et al. (1971)), adenosine 3: 5’-cyclic monophosphate (cyclic AMP, CAMP) seems to be involved in the function of the central nervous system (CNS) in several ways, affecting both presynaptic and postsynaptic events in neurotransmission (Bloom (1975), Duly (1975), Nuthanson (1977)). Psychotropic drugs interfere with the formation or breakdown of cAMP in many tissues, including the brain (Uzunov & Weiss (1972)), and this action has been proposed to be related to the mode of action of neuroleptics (Zversen (1975), Greengurd (1976)) and minor tranquilizers (Beer et al. (1972)). Since 1970, a series of reports have associated this nucleotide with manicdepressive illness (Abdulla (1970), PauZ et aZ. (1970, 1971), Naylor et ul. (1974), Sinanan et al. (1975)). Generally, in these studies CAMP excretion in urine was found low in depression and high in mania. On the assumption that changes in urinary CAMP might reflect similar changes in the brain, it was suggested that the cAMP system is disturbed in affective disorders. Such an assumption, however, is at variance with the existing evidence that much of the urinary CAMP is of renal origin (Broadus et al. (1970)). This contribution varies greatly among normal individuals, but may well exceed 50 % of the total cAMP concentration (Broadus et al. (1970), Kaminsky et al. (1970)). Moreover, only 1520 % of plasma CAMP is filtered into the urine, the rest being disposed of by
448
other mechanisms (Broadus et al. (1970)). It is thus clear that even a significant contribution to plasma cAMP from the brain would be largely obscured in urine. This has actually been directly demonstrated in the case of cerebral infarction, where increases of CAMP in plasma and CSF were high, whereas a negligible elevation was found in urine (Welch et al. (1975)). Similarly, changes of cAMP levels in plasma and urine are not parallel after treatment with hormones which increase extracellular levels of the nucleotide (Ball et al. (1972), Hamet et al. (1975)). Measurements in urine are subject to additional disadvantages. The amount of cAMP excreted may depend on the amount of creatinine and some processing of the samples is generally necessary (Murud (1973)). These factors indicate that urine is not a very appropriate material for the investigation of the cAMP system in psychiatric disorders and may explain the inability of some authors to replicate the reported changes in urinary cAMP of manic-depressive patients (Brown et al. (1972), Jenner et al. (1972)). Unlike urine, plasma represents a pool of CAMP deriving from many organs (Wehmann et al. (1974)), including the CNS (Broadus et al. (1970)). Concentrations in plasma are very stable in normal persons (Lykouras et al. (1978a)) and currently available methods can accurately measure cAMP in unprocessed plasma (Tovey et al. (1974)). Because of the above considerations, plasma rather than urine was used in the present study on the possible association of cAMP with manic-depressive illness. PATIENTS AND METHODS Forty-four m e d i c a t e d manic-depressive patients from the University Psychiatric Clinic at Eginition Hospital, or its outpatient department, were included in this study. The criteria of selection were as follows: 1. Agreement on the diagnosis of manic-depressive illness by two staff psychiatrists. Both bipolar and unipolar cases were included, but patients with involutional, neurotic or “atypical” depression were excluded from the study. 2. Age below 70 years. 3. Absence of clinical or laboratory evidence of physical disease. 4. Absence of medication for at least 3 weeks prior to the study (drug-free group). The effect of drugs on plasma CAMP was studied in 46 patients under treatment with tricyclic antidepressants (n = 22), lithium (n = 15) or neuroleptics (n = 9). The control group consisted of 76 healthy volunteers, age-matched to the patients (range 20-66 years). Blood was obtained from all subjects between 8.30 and 10 a.m. after overnight fasting. Samples were immediately transferred into tubes containing 0.5 M EDTA, 1 % v/v of blood. Plasma was then separated by centrifugation at 2800 r e v h i n for 10 min, and was kept frozen (-20°C) until analyzed. In our experience cAMP remains stable under these conditions for at least 1 year. De-
449
Table 1. Drugs not uflecting 3H-cAMP binding Dmg Diazepam Chlorpromazine Haloperidol Desipramine Li,CO,
Concentrations 7.10-M 1.4 X 1 P M 6.6 X l P M 5.3 X 1WM 0.1 mEqA
7.1WM 1.4 x 1 0 - 5 ~ 6.6 X 1WM 5.3 X l P M 1 mEqA
7.lPM 1.4 x 1 0 - 4 ~ 6.6 X 10dM 5.3 x 1 0 - 5 ~ 10 mEqA
terminations of cyclic AMP were made by the protein-binding method of T o v e y et ul. (1974) (“cyclic AMP assay kit”, Amersham, England). All samples were assayed in duplicate. Reproducibility of the method was very satisfactory. The coefficient of variation for eight measurements of six samples was 2-5.3 %. Recovery of 0.5 or 1 pmol of pure cAMP added to plasma was 95-97 % and therefore values were not corrected for recovery. Several factors which might influence results by this method were checked before the main study. Hemolysis, which has been considered such a factor (Murud (1973)), had no effect; recovery of cAMP from plasma was not affected by the presence of hemolysate of washed red blood cells (diluted 1:2 to 1:lOO). Endogenous CAMP in the hemolysate was previously destroyed by endogenous phosphodiesterase after incubation at 37OC for 24 h. Because blood contains a highly active phosphodiesterase (PDE), it is possible that some of the cAMP is destroyed by this enzyme in the interval between venipuncture and inhibition of PDE by the addition of EDTA (Brown et ul. (1972)). However, we found no difference in cAMP content in serial samples of eight subjects, mixed with EDTA 30, 60, 90 and 120 sec after venipuncture. All samples for the present study were obtained within this time interval. No interference with the binding of 3H-cAMP to the specific binding protein of the assay was found in the presence of several representative psychotropic agents, in a wide range of concentrations (Table 1). RESULTS
No difference in cAMP concentrations between males and females was found. The correlation of CAMP with age (control group) failed to reach significance (r = 0.092, NS). Mean ages did not differ significantly among the groups of our subjects. Table 2 shows concentrations of cAMP in the drug-free patients. Cyclic AMP levels are much higher in mania and much lower in depression, in comparison with the control group (P < 0.001 in both cases). Normothymic patients had normal concentrations. No difference in cAMP concentrations was found between unipolar and bipolar depressed patients. Psychoactive medication generally tended to normalize cAMP concentrations in all groups (Table 3).
450 Table 2. Plasma CAMP (pmoUml f s.d.) in drug-free manic-depressive patients Diagnosis
n
Depression Mania Normothymia Controls
28 9 7 76
Cyclic AMP
Difference from controls*
14.9 f 2.7 28.1 k 3.2 21.8 f 0.7 21.1 f 4.1
P < 0.001 P P
< 0.001 < 0.1
* Two-tailed t-test. Table 3. Effect of drugs
on plasma CAMP
CAMP @moI/ml f s.d.)
n
Drugs
Tricyclics Neuroleptics Li,CO, Controls
21.3 f 2.4 22.1 k 3.4 20.1 k 3.1 21.1 k 4.1
22 9
15 76
Table 4. Normalizing effect of lithium on plasma CAMP (pmoNm1) in manic-depressive patients
Cases
Before lithium Phase of illness
1
2 3 4
Mania Mania Depression Depression
During lithium cAMP
Phase of illness
cAMP
32 28.8 17.8 12.5
Normothymia Normothymia Normothymia Normothymia
24 21.8 21 16.5
None of the difTerences in Table 3 is significant. Most of the depressed patients being treated with tricyclic antidepressants were in normothymia when specimens were obtained. In nine of these patients, blood was taken before and after tricyclic drugs were given. Cyclic AMP rose signiiicantly after treatment (from 14.3 k 2.4 to 20.9 k 1.8 pmol, P < 0.001). The group receiving neuroleptics (Table 3) consisted of manic patients who were being treated with phenothiazines or butyrophenones and had improved to some extent at the time of blood sampling. Comparison of cAMP levels of this group to the drug-free manic subjects (Table 3) showed a significant fall in cAMP after treatment (P < 0.01). Lithium-treated patients were generally normothymic and had normal levels of cAMP (Table 3). In a few cases, however, specimens were also available before administration of lithium. Results in these cases indicate that lithium “normalizes” cAMP in plasma, i.e. decreases levels in mania and increases them in depression (Table 4).
451 DISCUSSION The finding of the present study that cAMP in plasma is elevated in mania and decreased in depression parallels similar observations in the urine of manicdepressive patients. However, several non-specific factors have to be accounted for before an association between manic-depressive illness and disturbance of the cyclic AMP system can be concluded. Some of these factors, related to the procedure of the assay, have been checked and found not to interfere with results (see Methods). Another possibility is that changes in CAMP may reflect changes in the psychomotor activity of the patients, as suggested by findings in urine (Eccleston et al. (1970)). Investigating this possibility in a parallel study, we found that psychomotor activity can affect concentrations of cAMP in plasma, but it is not the only factor responsible for the changes reported here (Lykouras et al. (1978b)). It can thus be concluded that abnormal affect is associated with changes in plasma CAMP. It has been adequately substantiated that changes in plasma cAMP reflect intracellular levels of the nucleotide (Broadus et al. (1970), Strange et al. (1974), Wehrnann et al. (1974)), perhaps through increased membrane permeability (Gillet al. (1975)). It is unlikely that these changes are due to disturbed excretion mechanisms, because clearance rates of cAMP from plasma are constant within a wide range of concentrations (Zssekutz (1975)). The tissues where plasma cAMP originates have not been identified with certainty. No single peripheral organ, including the liver, the kidney, the lungs, the adrenals and formed blood elements, is a major source of plasma CAMP (Wehrnann et at. (1974)). It has been suggested that the central nervous system, which is one of the richest sources of cAMP (Robison et al. (1971)), may be an important site of origin of plasma cAMP (Broadus et al. (1970)); direct evidence of a contribution from the brain, however, has only been provided in the case of acute cerebral infarction (Welch et al. (1975)). It is thus premature to conclude that elevation or decrease of plasma CAMP in manic-depressive patients reflects similar changes in the brain of these subjects. Whether changes in cAMP levels are due to a primary disturbance in production or breakdown of the nucleotide, or they are secondary to altered hormonal stimulation of adenyl cyclase, is an open question. It should be mentioned, in this context, that injection of catecholamines increases plasma cAMP (Murud (1973)), and endogenous elevation of catecholamines in cardial infarction runs parallel to the increase of cAMP in plasma (Strange et al. (1974)). Such a connection would be consistent with the long-standing catecholamine theory on affective disorders. The observation that patients under tricyclic antidepressants have normal plasma cAMP confirms similar findings in urine (Paul et al. (1971), Sinanan et al. (1975)). As most of these patients had recovered from depression when specimens were taken, it is not clear whether it is the action of the drugs or remission of the symptoms that normalizes CAMP in depression.
452
On the other hand neuroleptics decreased cAMP in manic patients, even before recovery; a similar observation was made in schizophrenic patients receiving PimozideB (Stefunis et al. (1977)). This effect can readily be explained by the inhibitory action of neuroleptics on adenyl cyclase (Zversen (1975)). The effect of lithium is somewhat more complex. It seems to act in both directions, i.e. it increases cAMP in depression and lowers it in mania. This stabilizing action is analogous to the drug’s prophylactic effect in both phases of manic-depressive illness and its ability to prevent drug-induced supersensitivity of receptors in experimental models (Kluwuns et ul. (1977)). In conclusion, this study provides evidence in support of the existence of a relationship between affective disorders and cyclic AMP. It also shows that plasma is a more suitable biological material than urine for the investigation of the role of cAMP in psychiatric disorders. REFERENCES Abdulla, Y . H., & K . Hamadah (1970): 3’, 5’-cyclic adenosine monophosphate in depression and mania. Lancet i, 378-381. Ball, 1. H., N . 1. Kaminsky, J . G . Hardman, A . E. Broadus, E . W . Sutherland & G . W . Liddle (1972): Effects of catecholamines and adrenergic-blocking agents on plasma and urinary nucleotides in man. J. clin. Invest. 51, 2124-2129. Beer, B., M . Chasin, D.E. Clody, J . R . Vogel & 2.P. Horovitz (1972): Cyclic adenosine monophosphate phosphodiesterase in brain effect on anxiety. Science 176, 428-430. Bloom, F.E. (1975): The role of cyclic nucleotides in central synaptic function. Rev. Physiol. biochem. Pharmacol. 74,1-104. Broadus, A . E., N . I . Kaminsky, J . G . Hardman, E. W . Sutherlund & G . W . Liddle (1970): Kinetic parameters and renal clearances of plasma adenosine 3’, 5’ monophosphate in man. J. clin. Invest. 49, 2222-2236. Brown, B. L., J . G . Salway, J . D.M . Albano, R . P. Hullin & R. P. Ekins (1972): Urinary excretion of cyclic AMP and manic-depressive psychosis. Brit. J. Psychiat. 120, 405-408. Duly, J . (1975): The role of cyclic nucleotides in the nervous system. In Zversen, L. L., S. D . lversen & S. H. Snyder (eds.): Handbook of psychopharmacology, Vol. I. Plenum, New York, pp. 47-130. Eccleston, D., R . Loose, I . A . Pullur & R . F. Sugden (1970): Exercise and urinary excretion of cyclic AMP. Lancet ii, 612-613. Gill, G . V., K . Prudhoe, D.B. Cook & A . L. Latner (1975): Effect of surgical trauma on plasma concentrations of cyclic AMP and cortisol. Brit. J. Surg. 62, 441-443. Greengard, P. (1976): Possible role for cyclic nucleotides and phosphorylated membrane proteins in postsynaptic actions of neurotransmitters. Nature (Lond.) 260, 101-108. Hamet, P., S. C. Loweder, I . G . Hardman & G . W. Liddle (1975): Effect of hypoglycemia on extracellular levels of cyclic AMP in man. Metabolism 24, 1139-1144. Zssekutz, T . B. (1975): Estimation of cyclic AMP turnover in normal and methylprednisolone treated dogs: effect of catecholamines. Amer. J. Physiol. 229, 291297. Iversen, L. L. (1975): Dopamine receptors in the brain. Science 188, 1084-1089. Jenner, F. A., G . A . Sampon, G . A . Thompson, A . R. Somerville, N . A . Beard & A . A . Smith (1972): Manic-depressive psychosis and urinary excretion of cyclic AMP. Brit. J. Psychiat. 121, 236-237. Kaminsky, N.Z., A . E. Broadus, J . G . Hardman, D . 1. Jones, 1. H . Ball, E . W . Sutherland & G . W . Liddle (1970): Effects of parathyroid hormone on plasma and urinary adenosine 3’, 5‘-monophosphate in man. J. clin. Invest. 49, 2387-2395.
453 Klawans, H. L., W. 1. Weiner & P. A. Nausieda (1977): The effect of lithium on an animal model of tardive dyskinesia. Progr. Neuro-Psychopharmacol. 1, 53-60. Lykouras, E., E. Garelis & C. N.Stefanis (1978a): Diurnal and longitudinal variation of cyclic AMP in plasma. (Submitted for publication). Lykouras, E., E. Garelis, E. Varsou & C.N . Stefanis (1978b): Physical activity and plasma cyclic AMP in manic-depressive patients and healthy adults. Amer. J. Psychiat. (in press). Murad, E. (1973): Clinical studies and applications of cyclic nucleotides. In Greengard, P., & G. A. Robison (eds.): Advances in cyclic nucleotide research, Vol. 3. Raven Press, New York. Nathanson, J. A. (1977): Cyclic nucleotides and nervous system function. Physiol. Rev. 57, 157-256. Naylor, G. L., S.A. Stansfield, S. F. Whyte & F. Hutchinson (1974): Urinary excretion of adenosine 3‘, 5’ monophosphate in depressive illness. Brit. J. Psychiat. 125, 275279. Paul, M . I., H. Cramer & F. K. Goodwin (1971): Urinary cyclic AMP excretion in depression and mania. Arch. gen. Psychiat. 24, 327-333. Paul, M . I., B. R. Ditzion, G. L. Pauk & D. S. Janowsky (1970): Urinary adenosine 3‘, 5’ monophosphate excretion in affective disorders. h e r . J. Psychiat. 126, 14931497. Robison, G. A., R. W. Butcher & E. W . Sutherland (1971): Cyclic AMP. Academic Press, New York. Sinanan, K., A. M . B. Keatinge, P. G. S. Beckett & W . Clayton-Love (1975): Urinary cyclic AMP in “endogenous” and “neurotic” depression. Brit. J. Psychiat. 126, 49-55. Stefanis, C . N., E. Lykouras, E. Garelis & E. Varsou (1977): The effect of Pimaid& on plasma cyclic AMP in chronic schizophrenics. Prog. Neuropsychopharmacol. 1, 323-327. Strange, R . C., N. Vetter, M . J . Rowe & M . F. Oliver (1974): Plasma cyclic AMP and total catecholamines during acute myocardial infarction in man. Eurap. J. clin. Invest. 4, 115-119. Tovey, K. C., K.G. Oldham & J . A . M . Whelan (1974): A simple direct sample using an improved competitive binding technique. Clin. chim. Acta 56, 221-234. Uzunov, P., & B . Weiss (1972): Psychopharmacological agents and the cyclic AMP system of rat brain. In Greengard, P.,& G. A. Robison (eds.): Advances in cyclic nucleotide research, Vol. 1. Raven Press, New York, pp. 435453. Wehmann, R . E., L. Blonde & A . L. Steiner (1974): Sources of cyclic nuclwtides in plasma. 3. clin. Invest. 53, 173-179. Welch, K. M. A., J . S. Mayer & A. N . C.Chee (1975): Evidence f o r disordered cyclic AMP metabolism in patients with cerebral infarction. Europ. Neurol. 13, 144-154.
Received October 26, 1977
E. Lykouras, M.D. Department of Psychiatry Eginition Hospital 74, Vasilissis Sophias Avenue Athens Greece
