This is another archival repost, originally from oct 2007.
Are you an addict? I bet you are (I saw you gambling addicts coming!). It might be coffee, chocolate, or tobacco, rather than alcohol, heroin, or self-harm, and you might not be wasting away because of it, but you’re probably addicted to something. There are a variety of reasons why we become addicted to a particular behaviour or chemical, but perhaps the most powerful is physical dependency. We become desensitised to drugs, take higher doses to achieve the same effects, and in many cases, experience physical pain when they are withdrawn. This is not just psychosomatic pain, and in “Drug-Induced Epigenetic Changes Produce Drug Tolerance“, Wang et al describe one of the mechanisms by which we become dependent on a drug.
The model of drug dependency used in this study was of benzyl alcohol, an organic solvent sedative (“sniffing glue”) in a species of fly (Drosophila melanogaster). The study looked at just one of the many genes that are involved in the process. The gene is called slo and it produces a calcium-activated potassium ion channel: a protein that sits on the surface of neurons, acting as a gatekeeper for the entry and exit of chemicals. Slo was chosen because it has previously been shown to have a variety of roles in regulating brain activity, including development of drug tolerance. Genes consist of coding regions, and non-coding regions, which contain metadata about gene expression, such as what events should trigger expression, and how much of the gene’s product should be made. The researchers were interested in one of these pieces of metadata, a “promoter” which switches on expression when a protein calledCREB is present. CREB is a signalling molecule — an intermediary between events occurring outside the cell, and regulation of gene expression within the cell. One of several roles thatCREB is known to perform, is to help mediate adaptation of neurons in response to experiences.
Genes are static: with only a few exceptions (random mutations, for example) they remain exactly the same in every cell of the body, day after day. How then, does a neuron adapt to a drug over long time periods, when the drug is only present for a short period? How, indeed, does a cell remember that it is supposed to be a neuron, and not a skin, bone or muscle cell? The answer is now thought to lie (at least partly) in epigenetics: the long term and heritable (at least over generations of cells) modification of gene expression. In particular, individual genes can be packaged up over long time frames in a protein called chromatin, which prevents the molecular machinery of expression from starting work. Under normal circumstances, slo tends to be packaged up in chromatin: not so tightly that no expression can occur at all, but enough that signals likeCREB are required to temporarily unpack the gene when expression is required. Wang et al gave some flies two rounds of solvent exposure and found that on the second exposure to the solvent, the flies recovered more quickly than on the first exposure, corresponding to greater slo expression. To support the hypothesis that this effect was caused by changes in the chromatin, they took flies that had not been exposed to the solvent before and exposed them first to a chemical which has a generalised (i.e. not specific to slo) effect of preventing the re-packaging of genes in chromatin, once they have been unpacked. Upon the exposure of these first-timers to the solvent, they recovered as quickly as those with the drug-tolerance. They also looked at a type of chemical modification, the addition of an acetyl group to chromatin proteins, which acts as a switch for unpacking the DNA, and found that following exposure to the drug, chromatin at the slo gene had indeed been acetylated. They also looked at the expression levels of slo, and found that preventing chromatin re-packing alone was enough to increase expression, even in the absence of the drug. To support the hypothesis that it was specifically slo and CREB that mediate the drug-tolerance effect, they knocked out CREB, and found that both the drug-tolerance effect, and the change in chromatin structure, disappeared with it.
To sum up, then: in the presence of the drug, the slo gene is expressed in large quantities for a very short period of time, producing a moderate quantity of a cell-surface protein which performs the task of bringing the neuron back to its normal state, thus aiding recovery from the drug. Over multiple drug exposures, the silencing of gene expression by chromatin is worn away, until expression of the gene lasts for much longer periods of time, and thus many more copies of the ion channel are produced. If the drug is then taken away, the ion channel continues to work away at bringing the neuron back to its normal state, but then overshoots, into a different abnormal state.
There are several potential applications of this knowledge. If we can prevent the drug tolerance, by preventing the wear down of chromatin regulation, can we prevent the addiction developing? (Or at least, reduce the probability or severity — multiple other molecular and psychological mechanisms are working alongside this one, after all.) Could we more carefully control expression of the relevant genes to prevent the overshoot, and subsequent withdrawal symptoms, thus making quitting easier? And how about legitimate pharmaceuticals? Tolerance becomes a problem when it’s applied to medicinal drugs; could that problem be reduced by careful control of chromatin structure?
- Krishnan HR, Ghezzi A, Yin JCP, Atkinson NS (2007) Drug-Induced Epigenetic Changes Produce Drug Tolerance. PLoS Biology 5(10): e265 doi.
- Ghezzi A, Al-Hasan YM, Larios LE, Bohm RA, Atkinson NS (2004) slo K(þ) channel gene regulation mediates rapid drug tolerance. Proc Natl Acad Sci USA 101: 17276–17281.
- Asyyed A, Storm D, Diamond I (2006) Ethanol activates cAMP response element-mediated gene expression in select regions of the mouse brain. Brain Res 1106: 63–71.
- Bird A (2007) Perceptions of epigenetics. Nature 447:396-398 doi
- Note that this is a much simplified model of what is actually happening.