Intended for healthcare professionals

Clinical Review

Brain drugs of the future

BMJ 1998; 317 doi: https://doi.org/10.1136/bmj.317.7174.1698 (Published 19 December 1998) Cite this as: BMJ 1998;317:1698
  1. Susan Greenfield, professor
  1. Department of Pharmacology, University of Oxford, Oxford OX1 3QT

    People have been using drugs to alter brain states since the dawn of time. But the use of specific drugs to combat specific brain problems is a hallmark of this century. In previous eras doctors reached for laudanum to combat hysteria, oblivious to the underlying neurochemistry. However, in Paris the famous neurologist Charcot purposely gave belladonna to patients with Parkinson's disease to combat the constant salivation that accompanies loss of motor control because women of the time who used the drug to dilate their pupils complained of a dry mouth. Surprisingly, this anticholinergic drug proved effective in combating not just the dribbling but the motor symptoms of Parkinson's disease.

    Figure1

    Magnetic resonance scan of coronal section of brain in Parkinson's disease

    Today we owe much to the work of Henry Dale and Otto Loewi, who established that chemicals act as transmitters to relay a signal from one neurone to the next.1 The guiding principle of modern neuropharmacology is to mimic, block, amplify, or reduce the availability of certain transmitters, believed to be pivotal to the disease to be treated. Yet herein lies the problem.

    Summary points

    One transmitter may be linked to many disorders, and one disorder to many transmitters

    Classic transmitters have non-classic modulatory functions too

    Substances such as nitric oxide and acetylcholinesterase have unexpected signalling properties

    Neurodegeneration might be an aberrant form of development, so drugs promoting neuronal regeneration should be approached with caution

    Drugs in the future could be used as a Rosetta stone for linking brain and mind

    The promiscuous transmitter

    There is no one to one matching of a single chemical system to a disease. Consider, for example, the well known transmitter dopamine. In Parkinson's disease there is a deficit of dopamine in the substantia nigra, hence prompting administration of the precursor and eventual mimicry of the effects of dopamine with the use of an agent such as bromocriptine to stimulate dopamine receptors directly. But such treatment risks psychotic side effects such as visual hallucinations and thus a distortion of reality reminiscent of schizophrenia—a condition associated with a functional excess of dopamine and thus treated with dopamine receptor blockers such as chlorpromazine.

    Dopamine is far from being the transmitter for movement. Rather, its role in both the generation of movement and indeed the aberrations seen in schizophrenia is in conjunction with other transmitters. The reason that Charcot's chance treatment was effective not just in quelling dribbling but in inspiring a treatment for Parkinson's disease, is that the cholinergic and dopaminergic systems work as a chemical see-saw in the basal ganglia, such that reducing the brain concentrations of acetylcholine to a value commensurate with the reduced dopamine is effective enough to have been the preferred drug treatment for half a century. Now we know that dopamine interacts with other key systems such as glutamate,2 hence the rationale for pallidotomies, which will diminish the net, overexcitatory input of glutamate on to the dwindling dopamine neurones. Perhaps once these multiway see-saws are exhaustively understood they will be programmed and processed accurately. A polypharmacy could then be developed that circumvents side effects because it exploits the specific interactions and balances that constitute a neurochemical signature for any one particular brain region.

    The multichemical basis of brain disorders

    Just as one transmitter such as dopamine can play a part in a variety of different diseases, so any one disorder, such as Parkinson's disease, can be regarded as the product of an interaction of a number of diverse transmitters. Alzheimer's disease is another example in which the problem is unlikely to be attributable to a single deficient system. Despite the popularity of the cholinergic hypothesis of Alzheimer's disease, and the appeal of cholinergic promoting drugs such as donepezil hydrochloride,3 acetylcholine is unlikely to be the transmitter for holding dementia at bay. In the brain certain populations of cholinergic neurones remain intact in Alzheimer's disease, while other non-cholinergic cell populations are lost.4

    There must therefore be more to Alzheimer's disease than a simple and generic malfunction of cholinergic neurones. And just as a simple drug targeting a single transmitter system is unlikely to cure a disease of failing memory, so it is similarly unlikely to enhance memory specifically in non-degenerating brains. Despite the recent excitement over such cognitive enhancers,3 the rationale is again to work simply at the level of disembodied transmitter systems—if not of acetylcholine then of an amino acid such as glutamate—or on the general mechanisms involved in plasticity of neuronal connections.3 But since there is no transmitter for a complex net function such as memory, then the manipulation of such basic and ubiquitous components of the brain will be non-specific and side effects inevitable. The same problem at an even more reductionist level would also apply to targeting a gene thought to be for some final, sophisticated brain function or malfunction.

    The subtlety of non-classic neurochemical systems

    The drugs of the future will exploit mechanisms in the brain other than the classic processes of synaptic transmission. Over recent years, we have seen enormous advances in our understanding of such non-classic mechanisms that play crucial parts in neurone operations. For example, dopamine is secreted from dendrites of key populations of neurones in a fashion that is far more diffuse and modulatory than at the tightly regulated synaptic cleft.5 The concept of modulation itself is a relative newcomer: the basic idea is that a compound may not in itself produce a response but rather bias neurones for a limited sphere of time and space (fig 1). Drugs might be developed that put neurones on red alert and are thus contingent on specific, physiological signals for the effects to be realised. Such drugs would hold the promise of acting less like a sledgehammer sincetheir action would occur only among neurones in which some subsequent, independent eventtook place within a given time frame.

    FIG 1
    FIG 1

    Two ways in which the excitability of cells can be modulated by one chemical. In both cases the neurones were stimulated electrically (dot in upper trace, step in lower trace) to generate action potentials. In the thalamic neurone it makes the potential difference more negative (centre), which is a prerequisite for activation of calcium entry into the neurone and in turn causes the neurone to generate more action potentials. In the hippocampic neurone, however, acetylcholine stops positively charged potassium ions from leaving the cell. Because potassium does not leave the cell the potential difference remains more positive than when acetylcholine is absent (see record on left) so that more action potentials are generated. When acetylcholine is washed off (right), both neurones revert to their original responses. Adapted from Greenfield12

    In addition, to familiar neurochemicals working in unfamiliar, more sophisticated contexts, whole new classes of bioactive substances are being discovered that do not behave in the same way as the more familiar transmitters. Fifteen years ago who would have thought that a gas, nitric oxide, might be psychoactive?6 And at the other end of the scale in terms of size, there are peptides. Peptides were the centrepiece of many grant applications in the 1970s because they are often stored with the classic transmitters (fig 2). So consistent yet surprising was this observation that peptides were hailed as modulators, a separate class of substance distinct from a transmitter.

    Fig 2
    Fig 2

    Top: Comparison of classic transmitters and peptides. When a neurone generates only a modest number of action potentials (middle trace), transmitter alone is released. As the firing rate increases so both transmitter and peptide are released. When the firing rate becomes high, peptide release dominates. Bottom: Peptide is stored with transmitter in large vesicles (triangle plus dot) and is released outside the synaptic cleft unlike transmitter (dots alone). Reproduced from Hokfelt7 with permission

    Such a distinction should be much more blurred. Transmitters, such as dopamine, serotonin, and acetylcholine, can all act in a more modulatory way, whereas in certain cases—such as in pain—peptides such as substance P seem to have a more defined and specific role. That said, the peptides as molecules do have certain properties not displayed by their more familiar counterparts. They tend to be released only when a neurone is very active, as opposed to being a faithful reflection of any degree of neuronal activity7; moreover, the release is often outside of the synapse itself, suggesting a diffuse and promiscuous action. In addition, many peptides can be bilingual, functioning as hormones and thus with different time scales and possible targets beyond the brain into the rest of the body. In a recent book Candace Pert mused that peptides might be responsible for emotions—indeed, that there might be a peptide for each emotion.8 Once again the trap opens up, of reductionism of a complex brain state to molecular structure. On the other hand, the role of peptides in more generalised body functions, and indeed the interaction they might have with the immune system, present tantalising possibilities for the development of future drugs.

    Another promising molecule, dear to my own heart and even bigger than the peptide family, is the enzyme acetylcholinesterase, which has a well known role in cholinergic transmission. Evidence is accruing that this familiar chemical might have a completely different function—modulation of non-cholinergic neurones.4 One such non-cholinergic role seems to be in development of the brain since diverse reports now suggest that acetylcholinesterase can enhance neurite outgrowth in certain parts of the brain (fig 3). Furthermore, acetylcholinesterase makes a transient appearance in certain brain regions in development, vanishes in maturity, but reappears after insults.4 Clearly, acetylcholinesterase itself might hold promise for exploitation once its non-enzymatic actions are better understood. But, in addition, the reappearance of what seems to be a developmental marker after injury adds weight to a fascinating idea: that neurodegeneration might be an aberrant form of development.9

    Fig 3
    Fig 3

    Non-classic action of acetycholinesterase. Both sections show slides from organotypic tissue cultures of substantia nigra in which the cells have been stained for the synthetic enzyme dopamine tyrosine hydroxylase. Top: Cells were incubated with echothiophate, which blocks only the catalytic site of acetycholinesterase. Bottom: Comparable cells were treated with an agent BW284C51, which also blocks non-enzymatic sites of acetycholinesterase. Only the non-classical blocker has had a dramatic influence on cell survival. Reproduced from Jones et al13 with permission

    The future

    This hypothesis, if true, would have enormous implications for another currently attractive idea for future treatments for neurodegeneration—mimicking trophic agents, a class of compounds which, like acetylcholinesterase, can enhance neurone survival in culture and encourage rapid growth.10 The problem with these trophic factors, however, is that they are too large to cross the blood-brain barrier and thus would need some kind of implant. But beyond the resource and delivery problems involved, I think that caution should be exercised in assuming that agents which promote the survival of young neurones will automatically have the effect of arresting death in the aged brain. Even within a week the tolerance of growing neurones to large influxes of calcium (a common trigger in development) drops by a third. We now know that neurons can die by activation of an internal self destruct programme (apoptosis) as opposed to by necrosis, in which factors exterior to the neurone play a part that affects a cell population more globally. So, it may well be one thing to shut off an apoptotic mechanism in tissue culture and quite another to prevent necrotic cell death downstream, necrosis resulting from direct and indirect factors beyond the machinations within any one neurone. At the very least, such indirect factors, above and beyond apoptosis, contribute to the net effect of cell death, say, in Parkinson's disease.2

    Figure5

    Transmission electron micrograph of a synapse

    The key principle for guiding future drug design must surely be to remember that the chemicals in question function in a context. They are a nested hierarchy of circuits that are changeable over time and constitute brain regions, which in turn are not the centres for something but rather make complementary contributions to brain function and dysfunction, like instruments in an orchestra.

    Drugs offer a powerful bridge between linking our knowledge of what goes on in our brains with how we feel. But the what goes on must be seen in the context not of isolated transmitter systems or single synapses but in the site specificity of a three dimensional brain. Brain imaging will provide a valuable resource to place the actions of the drugs in their true physiological context.11 In the future, brain drugs may be given not only as treatment against a wider backdrop of brain organisation but to provide insights into the basic nature of consciousness itself.

    References

    View Abstract

    Log in

    Log in through your institution

    Subscribe

    * For online subscription