Botulinum toxin, or ‘botox’ as scientists know it, is usually associated with food poisoning, not medical treatment. But, ironically, the same qualities that make botox highly poisonous and sometimes fatal also make it an effective therapy.

Ten years ago, an American ophthalmologist in San Francisco hit on the idea that the muscle paralysis induced by the toxin might be put to clinical use in cases where patients suffered from excessive muscular activity. Alan Scott of the Smith-Kettlewell Eye Research Institute in San Francisco thought of using the toxin for strabismus, or misalignment of the eyes. In this condition, a person’s eyes do not look in the same direction, and they are cross-eyed.

In the past decade, scientists have accumulated enough evidence to persuade the US Food and Drug Administration to approve injection of botox for strabismus, blepharospasm (forcible closure of the eye-lids) and hemifacial spasm (muscle contractions on one side of the face). The NIH panel recommends that these uses be expanded to include other dystonias, which involve the neck, jaw, limbs, and vocal cords, as well as stuttering and rare, but quite troublesome, spastic closure of the anal or urinary sphincter. At present, the toxin is available from two sources: Smith-Kettlewell and the Centre for Applied Microbiological Research, Porton Down.

Botulinum toxin is a composite of two toxins – ‘binary’ toxin and neurotoxin – together with an auxiliary protein, which acts as a stabiliser. When botulinum neurotoxin is ingested in spoiled food, the neurotixon travels to the junctions of skeletal muscles and nerves, where it eventually blocks the release of the neurotransmitter acetylcholine, causing muscle weakness and, sometimes, paralysis.

According to Lance Simpson at the Jefferson Medical College in Philadelphia, Pennsylvania, no one knows exactly how the toxin reaches the acetylcholine receptor. However, researchers know that the neurotoxin consists of two linked chains of polypeptides, and that one chain binds to the muscle cell’s plasma membrane while the other blocks the acetylcholine receptor.

If the chains are separated, says Simpson, toxicity is lost. This could be important for future manipulation of the toxin, says Simpson. For example, it may be possible to decouple the chains and attach a different kind of toxin or other drug to the binding chain, or change the binding site of the receptor-blocking chain.

For the time being, however, clinicians and researchers are happy with the neurotoxin as it is. Virtually everyone who has used the toxin in clinical trials reported their results at the NIH meeting, and the consensus was that when properly administered, the toxin is safe, effective, and, in some cases, the treatment of choice over surgery and drug therapy – particularly for focal dystonias, which affect only one or a few muscles in the body.

The response to localised injections of botox is far more reliable than that of other drugs such as anticholinergics and other neurotransmitter inhibitors. If a person is going to respond to botox, he usually does at the beginning of treatment, and continues to respond through repeated injections. Furthermore, botox is less drastic, less disfiguring, and less expensive than surgery (hundreds of dollars per botox treatment).

Of course, botox is not entirely problem-free. The panel stressed that the toxin is not a cure – nerves that the toxin blocks eventually sprout new terminals – so the blocking must be maintained through repeated injections. Although given in much smaller amounts than would be needed to produce botulism, it can still cause side effects, such as difficulty in swallowing, which result from excessive weakening of the target and surrounding muscles. However, the panel stated that such effects are relatively uncommon, usually transient, and often avoidable if clinicians are properly trained and muscle activity is accurately measured before treatment.

Furthermore, there is so far no report of any patient or handler of the toxin developing botulism or of the toxin adversely affecting patients with compromised immune systems. Indeed, the main problem with patients’ response is that some form antibodies to the toxin. ÐÓ°ÉÔ­´´s speculate that such a reaction could be the result of too much toxin in too short a period, or perhaps a genetic predispostiion to immunity.

*NIH Consensus Development Conference on Clinical Use of Botulinum Toxin, 12-14 November, 1990, Bethesda, Maryland.

Despite this insight, however, researchers have not been able to pin
down where the system goes wrong or why. They know, however, that certain
drugs can alleviate the feelings of helplessness and hopelessness that accompany
clinical depression, as well as the intense restlessness and irritability
of mania. But no one knows how these drugs work. Some researchers argue
that if they can find out how, they will have the key to understanding the
basic mechanisms of these complex disorders.

Recently, researchers have focused on lithium carbonate, a drug which
can moderate both depression and mania. Lithium’s ability to work at both
ends of the mood spectrum seems at first paradoxical: the substance appears
to both lower and raise the levels of neurotransmitters in the brain. However,
the researchers argue, there is no paradox if the drug acts not on the transmitters
themselves, but on some other point during the process of transmitting neurochemicals.
If the process is disrupted at this point, an abnormal mood results. Lithium
could prevent that disruption.

Charles Glatt, a medical student at Johns Hopkins University in Baltimore,
Maryland, suggests that lithium may act at the fulcrum of a chemical ‘seesaw’,
keeping it balanced.

Neurotransmitters that influence mood do not communicate information
directly to neurons; instead, they activate other chemical pathways, or
second messengers, that ultimately cause a biochemical change in the receiving
neurons. Could some critical step in the system of second messengers represent
the fulcrum? Glatt’s work indicates that a particular protein in the cell
membrane of neurons may be the site at which lithium acts. This protein,
known as the guanine nucleotide binding protein, or G protein, links the
external signal – the neurotransmitter – to the second messenger through
a series of chemical reactions which limit themselves. The second messenger
then carries the message inside the cell. Physiologically, these reactions
require magnesium. But although the magnesium ion has two positive charges
and the lithum ion only one, their sizes are similar. This means that they
can compete in chemical reactions.

To try to find out where lithium acts, Glatt conducted an experiment
in which he added three substances to a culture of neurons. One was lithium,
the next was the G protein. The third was a drug that is an agonist for
a particular neurotransmitter – in other words, it acts in a similar way
to the neurotransmitter.

Glatt found that the agonist bound to its receptor in the membrane of
the neuron, but the G protein remained inactive, thus preventing activation
of the second messenger system. It appears, therefore, that lithium acts
not at the receptor but on the G protein. If it had acted on the receptor,
it would have prevented the agonist from binding.

In the experiment, it took lithium several minutes to block the G protein.
Yet lithium (and other antidepressant drugs) has its clinical effect only
after several weeks. This observation led Glatt to examine the possibility
that – in the body and not the test tube – lithium produces a long-term
change in the amounts of protein that genes produce in the body. Glatt showed
that animals that had been treated with lithium had reduced amounts of messenger
RNA for a specific G protein. In other words, they were making smaller quantities
of the protein. He concluded that lithium in the brain does not simply block
the G protein as it did in his laboratory experiment. Instead, it lowers
the amount of protein that the G protein gene makes.

Glatt suggests that the G protein is the fulcrum of the seesaw of moods,
and that lithium may remove some of the ‘oil’ from that fulcrum. It may
interfere with the G protein’s ability to link neurotransmitters with their
second messengers and thus keep the balance from tipping in either the manic
or the depressive direction.

Cavenee began with the two-step concept of how cells become malignant,
put forward by Peter Nowell in 1978. According to Nowell’s now widely accepted
scheme, the first step probably consists of some sort of genetic mutation,
which transforms a normal cell. This transformed cell, though no longer
normal, will not necessarily become malignant. It takes a second event to
push this cell into generating a clone of identically damaged cells that
eventually become the clinical disease known as cancer.

Indeed, say Cavenee and others, there are many more stages beyond these
initial two involved in producing human cancer. Yet each generation of cells
in the progression towards malignancy should show signs of the event that
marked a former stage. So, like the largest in a nest of Russian dolls,
the ultimate stage of given cancer should contain all of the events that
led to it.

At least some of these events are aberrations of the chromosomes, in
which part of a chromosome becomes deleted, leading to the loss of a normal
gene from one of a pair of matching (homologous) chromosomes. To test this
idea, Cavenee first looked at retinoblastoma, a rare human cancer of the
eye known to run in families. Geneticists have shown that the cells of people
with the familial form of this cancer all have the same defect. Part of
one of the pair of chromosomes numbered 13 is missing: there is a deletion
of one gene, or allele, from one chromosome of the pair. This means that
either the father or mother of that individual contributed a defective chromosome
13.

This is the first step in the progression towards this cancer. This
genetic defect is present in all tissues but, for some reason, tumours arise
in cells of the retina that experience the second step: the loss of the
normal allele of the gene on their other chromosome 13. Before this loss,
the second ‘normal’ gene can apparently compensate for the absence of its
fellow. This final loss leaves the cell without a gene that apparently plays
a role in the suppression of cell growth.

Returning to Nowell’s model, the early event in retinoblastoma would
create a genetic predisposition to this form of cancer caused by the presence
of one normal and one defective chromosome 13. The second event, the loss
of the normal allele, pushes the cell towards malignancy. According to Cavenee,
80 per cent of retinoblastomas show this loss of mixed ancestry, or ‘heterozygosity’,
by elimination of the normal allele. Moreover, many other solid human tumours
lose the initial heterozygosity of the chromosomes.

One such cancer is glioblastoma, a malignancy of those brain cells known
as the glial cells, which nourish and structurally support neurons in the
brain. Glioblastoma is the name for the ultimate stage of astrocytoma, a
malignancy of a subtype of glial cells, the astrocytes. All glioblastomas
show a loss of heterozygosity in chromosome 10. This is compelling evidence
that the loss marks a stage in the progression of astrocytoma to glioblastoma.

Not everyone agrees with Cavenee, however. Many pathologists argue that
astrocytoma and glioblastoma are not stages of the same disease but separate
cancers arising either from a common precursor or two different precursors.
Cavenee and his colleague Peter Collins, a neuropathologist at the Karolinska
Institute in Stockholm, have now marshalled enough genetic evidence to suggest
a model for the progression of astrocytoma to glioblastoma. Working backwards,
they examined chromosome 10 in both earlier and ultimate stages of astrocytic
malignancy. Although they found that late-stage tumours had uniformly lost
the normal allele, they found no such loss in the earlier ones.

Next, they examined the receptors for a chemical, known as epidermal
growth factor, on the tumour cells. They found that these receptors were
more plentiful than normal on both the ultimate and mid-stage tumours, but
not in the earliest ones. Now they were ready to seek the initial step,
and for this they turned to a disease made famous by John Merrick, otherwise
know as the Elephant Man.

Merrick suffered from neurofibromatosis, an inherited disease traceable
to a defect in chromosome 17. People with this disease are likely to develop
astrocytoma as well. With this last fact in place, Cavenee and Collins came
up with the following genetic model for the progression of astrocytic tumours.

The initial event is a mutation in chromosome 17. This mutation makes
the cell genetically predisposed to neurofibromatosis. Cavenee and his colleagues
have recently published work showing that loss of the normal allele on the
other chromosome 17 is the second event (Proceedings of the National Academy
of Sciences, vol 86, p 2858). The result of this loss is the array of grotesque
but nonmalignant tumours that characterise neurofibromatosis.

Proliferation of the receptors on the cells for epidermal growth factor
is the next stage, leaving traces in the next generation of tumours, which
now invade the brain. Finally, in the last stages of astrocytoma, in the
so-called glioblastoma, loss of the normal allele of chromosome 10 appears,
along with the amplification of the receptors for epidermal growth factor
and the loss of the normal allele of chromosome 17.

Cavenee is quick to point out that his model is descriptive rather than
predictive. He cannot say for certain whether these genetic markers are
cause or effect, although the correlation between their presence and the
accepted pathological and histological stages of the cancer is so high as
to suggest a causal relationship. Nevertheless, Cavenee says, he still does
not know precisely how, or even whether, these chromosomal changes are linked,
or how the loss of the normal chromosome 10 might lead to malignant progression.
That is the subject of future work.

Although Cavenee works with heritable forms of cancer, he believes that
the same initiating event – for example, the loss of a regulatory gene by
mutation – is responsible in both familial and sporadic types of human cancers.
The difference may be that in the forms that run in families, the mutation
occurs in the germ line (egg and sperm cells) while in the sporadic form,
the mutation arises in somatic cells.

It is a very long and largely unlit road, however, from initiation to
progression. Indeed, glioblastoma does not progress to the same degree as
other solid tumours, because the blood-brain barrier prevents the cancerous
cells from spreading. It is probable, say Cavenee, that the progression
of human malignancies proceeds by many routes and that cancers, although
they may have a common origin in a deranged genome, are many different diseases
requiring many different lines of attack.