Samuel Gershon: 7. The etiology of bipolar disorder: A biochemical puzzle


We pursued an answer to this puzzle for almost 15 years in Australia and America. I hope these pieces of the puzzle will induce new thinking about current beliefs concerning neurotransmitters and their effects on bipolar disorder beyond the unitary effect of any one transmitter alone.


Like so many scientific findings, serendipity played a role in our first step. The Chair of our Department in Melbourne was contacted by the Commonwealth Scientific and Industrial Research (CSIRO) concerning a problem. In one laboratory, many staff scientists were reporting sick, seen first by general physicians and then, in most cases, referred to a psychiatrist. Of these a total of 16 patients were diagnosed with either depression or schizophrenia. The following case study presents the clinical picture.

A man of 43 had been exposed to organo-phosphorous insecticides over a 10 year period. He complained from time to time of nausea, vomiting, excessive perspiration, giddiness and muscular incoordination. He became tired, had difficulty sleeping and worried about impaired mental abilities. In 1959, he consulted a psychiatrist with complaints of severe depression. He was restless, miserable, had nightmares and insomnia. He was irritable, cried occasionally and was unable to carry on at work or at home. He was given sedatives at night and treated with multiple antidepressants including an MAO inhibitor but with little effect. By December 1960, 8 months after toxic exposure ceased, he was no longer depressed, slept well, while memory and concentration returned to normal and he felt like his former self.

Of the 16 cases we saw, 7 were diagnosed as depression, 5 as schizophrenic and the remaining 4 were not clearly defined. But all had impaired memory and concentration. Gradual recovery was related not to treatment but to natural restoration of choline acetylase. We first drew attention to these effects in an article published in the Lancet (Gershon and Shaw, 1960).

We followed up this study with a field survey in two fruit growing regions in Victoria, where physicians reported that rates of depression were elevated after the spraying season. The results of our study and the survey were published the following year (Gershon and Shaw, 1961).

After publication, we received letters from investigators in other countries reporting similar observations in farming communities and several from Europe sent reprints of their work. It also had wider ramifications. A well-known murder mystery author quoted our paper and used insecticides as the murder weapon, while Rachel Carson’s popular book exposed the health risks associated with widespread spraying of foods by insecticides. As often happens, in preparing our findings for publication, we found earlier work of a similar nature conducted a decade earlier by researchers at the Maudsley Hospital in London (Rowntree, Nevin and Wilson, 1950).  They used the same irreversible cholinesterase as in our studies, di-isopropyl-fluoro-phosphate, (DUFLOS). This was given to 17 schizophrenic and 9 manic-depressive patients for from 7 to 37 days. In 6 schizophrenic patients, psychosis was activated and florid symptoms that characterized the onset of illness were reactivated. Consciousness was not impaired and this was not a delirium. In manic depressives, mania lessened while depression, insomnia and dreaming increased. These changes in mental state persisted for several months after drug administration.

Observations in 10 normal control subjects were also informative; they developed a characteristic picture of depression, irritability, lassitude and apathy. They looked dejected, miserable and unhappy.  These findings were essentially similar to our Australian cases and the EEG changes reported in London were similar to the one case of our own, where we had an EEG. In our sample, DUFLOS induced or reactivated both depressive and schizophrenic forms of illness and these effects lasted for from 6 to 12 months after exposure, despite any treatment interventions. This recovery time was dictated by the body manufacturing new acetyl cholinesterase.

In America, at NYU, we communicated and collaborated with Davis and Janowsky over their findings with manic-depressive patients administered physostigmine, a reversible acetyl cholinesterase inhibitor with effects lasting only up to 24 hours. Administered in the manic phase, this produced a marked diminution in all symptoms, including behavior, speech and thought often with a full reversion to baseline and even some symptoms of depression. My colleague, Burt Angrist, and I discussed these findings and decided they were related to our own work on induction of depression by insecticides. The patients we were seeing at Bellevue, admitted as emergencies, were probably more severe than those available to Davis and Janowsky in a State Hospital, so we invited them to join us in a collaborative study. Together, we interviewed selected patients, recorded their clinical picture and rated them on a mania scale. Our visitors injected physostigmine and the patients gradually became calmer, speaking in a more rational way and, after about an hour, were well behaved, talking sensibly and restored to a euthymic state. One patient began to show depressive symptoms with an increasing dose until administration was terminated.

This collaboration confirmed consistent findings with reversible and irreversible acetyl cholinesterase inhibitors and supported the role of acetylcholine in the biochemical etiology of both schizophrenia and bipolar disorder. It also demonstrated a level of scientific collegiality we currently miss (Shopsin, Janowsky, Davis and Gershon, 1975).


From the mid 1960’s on, biochemical speculation about etiology of the major psychiatric syndromes began to focus on the catecholamines. In 1965, Shildkraut proposed the “catecholamine hypothesis of affective disorders” (Shildkraut, 1965) and in 1970, Axelrod and his team won the Nobel Prize for their work on the uptake of catecholamines in humans.

This theory held that decreased levels of norepinephrine (NE) and its metabolites existed in the brain and cerebrospinal fluid in depression while elevated levels were present in mania. New antidepressants were introduced at a rapid rate and we had access to many of them at NYU for clinical trials and animal experiments. In our clinical trials, some of them were ineffective for depression but had fulfilled the animal requirements for this hypothesis. It was not long before experiments with synthesis inhibitors and treatment response to the SSRI antidepressants began to reconcile this anomaly and shift the focus to serotonin.

We now became involved at NYU with experiments in animals and humans using synthesis inhibitors of both amines, norepinephrine and serotonin. Patients with depression, successfully treated with imipramine and in remission, were treated with AMPT, a NE synthesis inhibitor or PCPA, a serotonin synthesis inhibitor. Only the serotonin inhibitor caused a return of depressive symptoms; its withdrawal led to return of remission. This clinical demonstration was preceded by numerous animal experiments to justify the logic and safety of conducting the experiment in patients (Shopsin, Gershon, Goldstein et al, 1975).

Closely related to our studies, was the pioneering work of our colleague Dr. Michael Stanley, showing reduced imipramine binding (serotonin transporters) in the brain of individuals who had committed suicide, published in Science (Stanley, Virgilo and Gershon, 1982).

This record of research conducted between 1960 and 1975 reveals the shifting sands of biochemical speculation about the etiology of bipolar disorders, from acetylcholine to norepinephrine to serotonin. As so often in neuroscience research, the brain produces more puzzles than definitive or singular answers. How do we reconcile our earlier findings with acetylcholine and more recent ones with serotonin? One could suggest that our experiments with these two sets of neurotransmitters are classical examples of translational neuroscience,  we hope may be followed up to clarify their roles.



Rowntree DW, Nevin S, Wilson A. The effect of disipropylphosphonate in psychosis. J.Neurol. Psychiat. 1950; 47-62.

Gershon S, Shaw FH. Psychiatric sequela of exposure to OP insecticides. Lancet 1960; 1:137- 4

Gershon S, Shaw FH. Psychiatric sequelae of chronic exposure to organophosphorus insecticides. Lancet: 1: 1371-1374, 1961.

Shildkraut JJ, The catecholamine hypothesis of affective disorders; a review of supporting evidence. Am.J.Psychiatry 1965; 122:509-22.

Shopsin B, Gershon S, Goldstein M, Friedman E. and Wilk S. Use of synthesis inhibitors in defining the role of biogenic amines during imipramine treatment in depressed patients. Psychopharmacol. Commun. 1 (2), 239-249, 1975.

Stanley M, Virgil, Gershon S. Tritiated imipramine binding sites are decreased in the frontal cortex of suicides. Science 1982; 216: 1337-9


Samuel Gershon

September 24, 2015