by Susan Greenfield Royal Institution London With Alzheimer’s disease we are faced not so much with a technical problem but with an ethical problem. Is it always appropriate to tell someone they have an incurable disease?

Genetic research holds exciting promise for anticipating, diagnosing and treating some of our most intractable brain disorders, such as Alzheimer's disease and serious strokes.

We have a large number of genes, nigh on 30,000, in our bodies. Amazingly enough, as many as half of them will play some part in brain function. But daunting though this figure may sound, it pales by comparison with the number of connections between our brain cells: one trillion (15 noughts).

So even if every single gene in our body were devoted to a connection, we would still have a shortfall of 100 billion genes. We can't really speak, therefore, of a specific gene ‘for’ a certain mental function or dysfunction.

Instead, a good way of visualising how genes might relate to brain function is to think of them as spark plugs of the engine of a car: necessary, but not sufficient technology. Genes will express – trigger the manufacture of – proteins. Those proteins in turn, will play an important part in the biochemical machinery that enables one brain cell to communicate with another. This simple building block is then incorporated into circuits. Circuits are clustered into assemblies of neurons and, via a layered hierarchy of organisation, neurons are grouped into gross regions that together constitute the brain.

So we cannot really jump from a gene to a final function. We do not have such characteristics as homosexuality, criminality, good housekeeping, or bad cooking, trapped inside a strand of DNA.

Even when, in rare cases, a brain disorder can be traced to a single gene, it is still not the case that one is automatically destined to have that affliction. This remarkable truth came to light recently in a study on the gene for Huntington's chorea, a severe disorder of movement, characterised by a wild and involuntary flailing of the limbs.

A single aberrant gene is at the root of this disorder. We can now tweak this rogue gene in mice so that they will inevitably develop a movement disorder providing us with an animal model of the disease for further study. Yet even in a simpler brain of a mouse, and even when the genetic destiny seems to have doomed them to eventually develop the murine equivalent of Huntington's chorea , the outcome is still not straightforward. Incredibly, enrichment of their environment can delay the onset of the impairment. Nurture can trump Nature.

Brain genomics
So how will the study of genes – genomics – be used to best advantage in the future for brain diseases? One obvious application is for diagnosis: to identify individuals that might be at risk with strongly genetic-based diseases, such as early-onset familial Alzheimer's disease.

Here we are faced not so much with a technical problem but with an ethical problem. Is it always appropriate to tell someone they have an incurable disease? Identifying the genes for diseases also raises a further ethical issue with the unborn. If one can screen a foetus for aberrant genes, how should that affect our choice of abortion or not? Where do we draw the Rubicon between when one can have a useful life, even though one is physically and mentally imperfect, and when not? Huntington's chorea, for example, does not become apparent until the middle years and, even then, one is still the individual one always was. Should that person never have lived?

Another way in which diagnostic screening, developed through the study of genes, might help in the clinic is by the process of pharmacogenomics. Frequently, the same symptoms will be apparent but will be due to different genes. Treatment today may therefore be far less than optimal and entail side-effects. An obvious analogy is an off-the-peg suit where one size purports to fit all, compared to the desirability of custom-made tailoring. Pharmacogenomics could lead to tailor-made medication, with minimal side-effects.

This general strategy of taking a snap-shot of one's genes in health and disease is known as ‘expression profiling’ . If we knew what sets of genes are different between one condition (such as Alzheimer's disease) and another, we would have a good clue to new drug targets.

But unless we know what proteins the genes are actually making, it is hard to devise ways of stopping them acting out their biological fate. Hence the rise of the science of proteomics, where one dispenses with genes altogether, and looks instead at the rogue proteins resulting from discrepancies in the sets of genes.

So it might seem that we are now home and dry in identifying new brain mechanisms. In fact, we need to revisit the same question: what are the proteins this time doing?

‘Real brain’
Perhaps, instead, proteomics will lead us, not into rushing to develop a drug, but rather to appreciate how proteins interact in the context of real brain. We might then see how the complete chemical landscape, not just the isolated proteins, might be modified.

In the longer term, we are already hearing much of cloning. It is important to remember, however, that genetic factors are necessary but not sufficient for brain function. There is also an important contribution from the environment. Even if you are a clone (that is, an identical twin), no two people will ever be the same person. Moreover, with reproductive cloning, one would be separated by a whole generation, complete with differences in culture, diet and medication.

A more realistic scenario will be to clone stem cells, the starter cells from which all the different cells of the body are made. If a stem cell is introduced into a particular micro-environment in the body or brain, it will become that type of cell, whether it is a blood cell, a bone cell, a heart cell or, indeed, a brain cell. Through the possibility of introducing cloned stem cells into the brain we might by-pass the ethical problems of donations from, say, foetal brain tissue, and even offer the prospects of a treatment where none currently exists – for example, in large-scale strokes.

In summary then, the notion of a simple bridge between genes and brain function is perhaps a bridge too far. But if the study of genes prompts new types of treatments, it should certainly offer a bridge into the future.