Thomas A.Ban, editor. Lithium in Psychiatry in Historical Perspective.


Robert Haim Belmaker’s comment

A Saga of Lithium Research 1975-2020




         The Jerusalem Mental Health Center, where I was Director of Research (1974-1984), was a hospital of about 300 beds with a catchment area of Northern Jerusalem. The first clinical trial I planned was with a young resident in psychiatry named Joseph Biederman. I gave him the task of seeing whether there was a cut-off at a particular place along the dimension between bipolar disorder and schizoaffective disorder where lithium ceased to work. The design was a double-blind “add-on” of lithium or placebo to haloperidol-treated, acutely psychotic patients. Some of the patients were bipolar manics; others were schizophrenic patients with “excited psychosis.” Most patients were excited schizoaffectives. The findings of that first study were published in the Archives of General Psychiatry in 1979 (Biederman, Lerner and Belmaker 1979). There were about 35 patients included in it. We found highly significant benefits for lithium without any difference between the bipolar manics and other patients. It was infrequently quoted, perhaps because DSM-3 diagnoses were introduced soon after and became an inclusion criterion for review articles. Another possible reason for this study’s obscurity is that it was an “add-on” study and at the time “add-on” studies were considered uninformative about basic psychopharmacological questions.   Lithium added benefit whenever there was an affective component; it really didn’t matter whether the patients met DSM3- criteria for bipolar disorder or schizoaffective disorder or schizophrenia. That conclusion might have been hard to accept for  all those of us who hoped that lithium would allow a biochemical dissection of bipolar disorder. 

         We did another “add-on” study to test indirectly the specificity of the lithium effect which was also published in 1984 in the Archives of General Psychiatry (Klein, Bental,  Lerer and Belmaker 1984). It was “add on” of carbamazepine or placebo in haloperidol-treated patients with “excited psychosis” done by Ehud Klein who was a young psychiatry resident then. Again, we found benefit of carbamazepine without any relationship to strict bipolar diagnosis.



         The laboratory at the Jerusalem Mental Health Center was headed by Richard Ebstein, PhD.  He set up an assay for measuring cyclic AMP in human plasma, based on a paper by Sutherland (Ball, Kaminsky, Hardman et al. 1972). Sutherland demonstrated that  if one gives a dose of epinephrine, in a dose  equivalent to that given in those days in an asthma attack subcutaneously, and takes blood samples every 10 minutes from an indwelling venous  catheter, one  sees a rise in cyclic AMP, exactly the same as occurs inside cells in response to stimulation of the adrenergic receptor. We were able to show that in euthymic patients on lithium at therapeutic doses the cyclic AMP response to epinephrine was completely blocked, compared to controls (Ebstein, Belmaker, Grunhaus and Rimon 1976). This was an early  translational research study where a finding in basic science that lithium affects second messengers was replicated in humans. A finding that might have been put aside as occurring at non-therapeutic concentrations was shown to work in patients.

         Vetulani and Sulser (1975) published findings that turned the old catecholamine hypothesis on its head, saying antidepressants induce postsynaptic sub-sensitivity rather than an increase in synaptic monoamines. Salbutamol had just been released for clinical use in the treatment of asthma. We hypothesized that  salbutamol should increase plasma cyclic AMP just like other adrenergic agonists. Bernard Lerer, a young resident at the time, tested depressed patients at baseline for cyclic AMP response to salbutamol, then again after a month of treatment and then treated for depression. We were then able to demonstrate in vivo in humans the sub-sensitivity of the ß-receptor second messenger response (Lerer, Ebstein and Belmaker 1981). This seemed to connect the effect of lithium on the second messenger cyclic AMP system to that of unipolar acting antidepressants (Belmaker 1981).  With a new post-doctoral fellow, Michael Newman, we tried to examine the mechanism of lithium’s effects on the cyclic AMP generating system using tools such as forskolin that bypass the adrenergic receptor. We found definite effects of lithium distal to the receptor, consistent with developing concepts in basic science revealing that the receptor is G-protein coupled (Newman and Belmaker 1987)

         In 1985 I moved to the Beersheva Mental Health Center of Ben Gurion University and began working with Gabriel Schreiber MD, a psychiatry resident, and Sophia Avissar PhD, a post-doctoral student. They developed a receptor binding assay for G-protein activation and we found that lithium could block the receptor activated increase in G-protein binding in tissue from rat brain (Avissar, Schreiber, Danon and Belmaker 1988). The effect has since been found to be dependent on magnesium concentration in the experimental medium and not clearly relevant to human therapeutic conditions. As a result, I became interested in looking at lithium effects on a newly emerging second messenger system, phosphatidyl inositol (PI), and inspired by the work of Michael Berridge (Berridge, Downes and Hanley 1989) who received the Wolf Prize in 1994 for his work in the Israeli Knesset.  Our first efforts were behavioral, working with a young post-doctoral fellow at the time in our lab, Ora Kofman PhD, and a young resident in psychiatry at the time, Yuly Bersudsky.  The work was based on the 1980 serendipitous finding by Hallcher and Sherman that lithium at therapeutic concentrations inhibits inositol monophosphatase, the final step in the dephosphorylation of a key product in the PI cycle.   Sherman  and colleagues Honchar and Olney (1983) found that lithium in rats hugely potentiated  the effects of cholinergic agonists to cause a peculiar limbic seizure syndrome and connected this convincingly to inositol depletion. While lithium was not thought to have cholinergic mechanisms, this effect was the most powerful behavioral effect of lithium I had seen before or since.  I spent the summer of 1989 in Sherman’s lab at Washington University in St. Louis to acquaint myself with this phenomenon. Sherman was a great scientist and teacher, an organic chemist, as he said it, “for an organic psychiatry department.” Ora set up the lithium-pilocarpine model in our lab in Beersheva and we soon proved that the lithium effect was stereospecifically reversed by the natural isomer of inositol (Kofman and Belmaker, 1993) but Yuly found that inositol, unfortunately, was not an antidote to lithium toxicity in general (Bersudsky, Vinnitsky, Ghelber et al. 1993).  When Haim Einat that joined us as a doctoral student we found that, as with the cyclic AMP effects of lithium described above, the effects of lithium on PI could be joined to the effects of inositol and the PI cycle in depression models and treatment more generally (Einat and Belmaker 2001). This theme of lack of specificity of lithium effects in humans to the specific narrow diagnosis of bipolar disorder and the lack of specificity of lithium mechanisms to any unique system in biochemistry has run throughout my years of research, much against my hopes and inclinations in the direction of Occam’s razor. Galila Agam joined the team as a biochemist in the 1990s and some of our exciting new  approaches to lithium and the PI cycle  involved knockout mice (Agam, Bersudsky, Berry et al. 2009). The inositol monophosphatase-1 knockout mice have pilocarpine sensitivity and behave on the Porsolt forced swimming test as if the animal is taking lithium. 

         Most recently, Nisha Singh, a doctoral student from Grant Churchill’s lab at Oxford in the UK, came to our lab in Beersheva to study ebselen compared to lithium in the pilocarpine potentiation model. Ebselen is a compound developed for treatment of inflammation found to be an inositol monophosphatase inhibitor and to lower brain inositol, both biochemical effects of lithium. It’s possible clinical effectiveness will be the ultimate test of the inositol depletion hypothesis, first proposed by Berridge, Downes and Hanley (1989) as discussed above. Unfortunately for the lithium-pilocarpine model, Nisha and coworkers in our lab found that ebselen did not potentiate pilocarpine  seizures (Singh, Serres, Toker et al. 2020).  This may not be decisive evidence against the inositol depletion hypothesis of lithium action but it is evidence against the generalizability of lithium pilocarpine seizures as a useful behavioral  model of lithium-like biochemical effects.

         In a 2004 article in the New England Journal of Medicine I developed the concept  that lithium research has found effects of lithium in every dominant paradigm in each generation of psychopharmacology. When monoamine metabolites were being investigated, lithium had effects on monoamine metabolites; when receptors were the frontier, lithium had effects on receptor hypersensitivity; when second messengers were at the forefront, lithium was found to have effects on second messengers; and when gene expression was of interest to all, lithium was found to have effects on gene expression. This generational shifting has continued but my strong impression of my own 45 years of work is that we still do not know the mechanism of lithium action and are not even close in lithium research to a concept parallel to the central theme of dopamine blockade in neuroleptic action and monoamine reuptake or MAO inhibition in antidepressants.



         I still use lithium in the clinic and find it very beneficial in many patients with bipolar disorder, some with recurrent unipolar disorder and many with schizoaffective disorders; atypical antipsychotics are often but not always equally useful and are very convenient (Osher, Bersudsky and Belmaker 2010). Given the long period of lithium usage in many of my patients, kidney effects including progression to renal dialysis and transplant have become a very real issue (Azab, Schnaider, Osher et al. 2015). Basic research on the mechanism of lithium action has not led to application  of Occam’s razor; but clinical usage surely justifies the adage that there is no free lunch.



Agam G, Bersudsky Y, Berry GT, Moechars D, Lavi-Avnon Y, Belmaker RH. Knockout mice in understanding the mechanism of action of lithium. Biochem Soc Trans. 2009; 37(Pt 5):1121-5.  

Avissar S, Schreiber G, Danon A, Belmaker RH. Lithium inhibits adrenergic and cholinergic increases in GTP binding in rat cortex. Nature. 1988; 331(6155):440-2. 

Azab A, Schnaider A, Osher Y, Wang D, Bersudsky Y, Belmaker R. Lithium Nephrotoxicity. Int J Bipolar Disord. 2015; 3:13. 

Ball JH, Kaminsky NI, Hardman JG, Broadus AE, Sutherland EW, Liddle GW. Effects of catecholamines and adrenergic-blocking agents on plasma and urinary cyclic nucleotides in man. J Clin Inves.1972; 51(8):2124-9. 

Belmaker R. Bipolar Disorder. N Engl J Med. 2004; 351(5):476-86.

Belmaker R. Receptors, adenylate cyclase, depression and lithium. Biol Psychiatry. 1981; 16(4):333-50.  

Berridge MJ, Downes CP, Hanley MR. Neural and developmental actions of lithium: a unifying hypothesis. Cell. 1989; 59(3):411-9.  

Bersudsky Y, Vinnitsky I, Ghelber D, Kofman O, Kaplan Z, Belmaker R. Mechanism of lithium lethality in rats. Journal of Psychiatric Research. 1993; 27(4):415-22. 

Biederman J, Lerner Y, Belmaker R. Combination of lithium carbonate and haloperidol in schizo-affective disorder:  a controlled study. Arch Gen Psychiatry. 1979; 36(3):327-33.  

Ebstein R, Belmaker R, Grunhaus L, Rimon R. Lithium inhibition of adrenaline sensitive adenylate cyclase in humans. Nature. 1976; 259:411-13.  

Einat H, Belmaker R. The effects of inositol treatment in animal models of psychiatric  disorders. J Affect Disord. 2001; 62(1-2):113-21. 

Hallcher L, Sherman W. The effect of lithium ion and other agents on the activity of myo-inositol-1-phosphatase from bovine brain. J Biol Chem 1980; 255(22):10896-901. 

Honchar MP, Olney JW, Sherman WR. Systemic cholinergic agents induce seizures and brain damage in lithium-treated rats. Science. 1983; 220(4594):323-5. 

Klein E, Bental E, Lerer B, Belmaker RH. Carbamazepine and haloperidol v placebo and haloperidol in excited psychoses:  A controlled study. Arch Gen Psychiatry. 1984; 41(2):165-70. 

Kofman O, Belmaker RH. Biochemical, behavioral and clinical studies of the role of inositol in lithium treatment and depression. Biol Psychiatry. 1993; 34(12):839-52. 

Lerer B, Ebstein RP, Belmaker RH. Subsensitivity of human-beta adrenergic adenylate cyclase after salbutamol treatment of depression. Psychopharmacology (Berl). 1981; 75(2):169-72.  

Newman ME, Belmaker RH.  Effect of lithium in vitro and ex vivo on components of the adenylate cyclase system in rat cerebral cortex membrane. Neuropharmacology. 1987; 26:211-7. 

Osher Y, Bersudsky Y, Belmaker RH. The new lithium clinic. Neuropsychobiology 2010; 62:17-26  

Singh N, Serres F, Toker L, Sade Y, Blackburn V, Batra SA, Saiardi A, Agam G., Belmaker RH, Sharp T, Vasudevan SR, Churchill G. Effects of the putative lithium mimetic ebselen on pilocarpine-induced neural activity. European J of Pharmacology. 2020; 883:173377. 

Vetulani J, Sulser F. Action of various antidepressant treatments reduces re-activity of noradrenergic cyclic AMP generating system in limbic forebrain. Nature. 1975; 257: 495-6. 


September 24, 2020