I am interested in how changes in gene expression produce changes in behavior. As a post-doc, this led me to study the Drosophila slo gene. The slo gene intrigued me because mutant alleles caused subtle defects in behavior. My cloning of slo provided the first example of a gene encoding a BK-type Ca2+-activated K+ channel. In my own laboratory, I turned my attention to the transcriptional regulation of slo. Even though transcription regulation was an unusual topic for an ion channel lab, I felt confident that channel genes would be regulated in response to experience and that changes in channel expression would alter behavior. Furthermore, the fact that transcription regulation was not a focal point of the field served to convince me that it was here that I could make a unique contribution. To lay the groundwork for behavioral studies, we described the tissue- and development-specific regulation of the gene. We discovered that slo had a 7 kb transcriptional control region that included two neural promoters, a promoter responsible for muscle and tracheal cell expression, and a promoter that drives expression in a handful of epithelial cells. Control elements that dictated the expression pattern were numerous and scattered throughout the region. We studied the role of slo in flight muscles and in the so-called sticky feet response, and we examined the contribution of alternative splicing to channel diversity.

This groundwork enabled the pursuit of the topic that most intrigued me. The lab often discussed the idea of homeostatic regulation of ion channel gene expression and its behavioral consequences. I imagined that, when challenged with any agent that produced an abnormal change in neural activity, the nervous system would act to counter the agent in such a way as to restore normal signaling properties. I thought that slo was a likely target of such regulation. The very large conductance of the BK channel (10X to 20X as great as the conductance of any other K+ channels) meant that even small changes in expression could have large effects on membrane excitability. Moreover, slo channels respond to the two key modulators of synaptic release: membrane potential and Ca2+ levels. In addition, the proximity of slo channels to voltage-gated Ca2+ channels was expected to enable slo channels to affect synaptic signaling. These features make slo channels well situated for the task of modulating synaptic activity.

My lab is currently known for our work on functional tolerance to ethanol and to abused solvent inhalants. Tolerance is defined as a reduction in the response to a drug caused by prior drug exposure. However, it was an attempt to detect homeostatic regulation of slo that brought us into the drug abuse field. At the time, we were using sublethal doses of neural-specific insecticides as neural excitants. We asked whether the expression of ion channel genes, including slo, were altered by treatment with these insecticides. However, in one experiment a vehicle-only control showed very interesting results. The vehicle was the organic solvent benzyl alcohol, an occasionally used anesthetic in humans. Flies sedated with benzyl alcohol showed an up-regulation in the expression of slo. After substantial effort, we were able to determine that slo was up-regulated in response to sedation itself. Ethanol sedation produced similar induction. On the other hand, we observed that stimulants caused a decline in slo expression (Ghezzi et al., 2004). These observations suggested that slo gene expression is linked by some mechanism to neural activity.

Because of my interest in behavior, we examined the animals for any obvious behavioral changes resulting from treatment with the solvents. One particularly fortunate experiment was the measurement of the duration of sedation. We observed that prior sedation shortens the period of benzyl alcohol sedation. This shortening of the period of sedation is called tolerance.

To my great surprise, our work demonstrated that mutations in slo that block only neural expression absolutely eliminate the capacity to acquire rapid functional tolerance to sedation, indicating that this type of tolerance is dependent on slo expression in the nervous system.  In the lexicon of the field of drug and alcohol abuse, "functional" indicates that the tolerance arises because of a change in neural properties, and "rapid" means that the tolerance appears shortly after a single exposure. Using inducible transgenes, we went on to show that slo induction, alone, can phenocopy the tolerance phenotype (Ghezzi et al., 2004). Thus, induction of the slo gene appears to be necessary and sufficient to produce this form of behavioral tolerance. During this time, we also showed that slo was essential for the manifestation of rapid functional tolerance to ethanol sedation (Cowmeadow et al., 2005; Cowmeadow et al., 2006). While we have observed that sedation alters the expression of other K+ channel genes, the mutations in these genes did not block the tolerance phenotype.

These experiments changed the conceptual underpinnings of my lab. The central theme of the lab is now the molecular mechanisms that underlie tolerance to abused solvents and to ethanol. The acquisition of tolerance is an important component of the addictive process. In humans, tolerance is an insidious response to the consumption of addictive drugs because tolerance leads to increased drug consumption, which speeds the path to addiction.

We have been asking how the slo gene "senses" sedation and why the induction of slo ameliorates the effects of the drug. In addition, we have been searching for other components that participate in producing tolerance. To answer these questions and accomplish these goals, we have had to be very flexible and follow wherever the data lead. Doing so has involved much re-education and the adoption of a number of new techniques, which has been very exciting. We have used the chromatin immunoprecipitation assay to physically describe the changes in transcription factor binding and epigenetic modifications caused by sedation. We observed that shortly after sedation flies exhibit an increase in the binding of the CREB transcription factor at three positions within the slo transcriptional control region. After CREB binding there were changes in histone acetylation across the slo promoter region. These changes occur with a specific temporal and spatial time course and reflect modifications that are part of the induction of slo in response to sedation.  We observed that if CREB activity was inhibited (with a CREB mutant or an inducible dominant-negative CREB transgene), slo was not induced and tolerance did not occur (Ghezzi et al., 2004). We continue to describe the mechanics of slo activation by drug sedation. I now believe that we have identified the transcription factor that precedes CREB activation.

Our work leads one to the counterintuitive conclusion that the increased slo channel gene expression acts as a neural excitant that counters the sedating effects of drug sedation. The dogmatic view is that increased K+ channel activity is always correlated with the depression of neural activity. However, the literature is rich with examples of increased Ca2+-activated K+ channel activity reducing the neural refractory period and enhancing the firing rate of the neuron. We predicted that increased slo expression was having this enhancing effect. Using the giant fiber CNS preparation we then demonstrated that sedation-mediated slo induction was associated with a reduced refractory period and an increase in the maximum possible firing rate. Mutations in slo that eliminated neural expression blocked this increase, while the increase in firing rate could be phenocopied by inducing slo expression from a transgene. In the Drosophila giant fiber, prior sedation reduces the refractory period and enhances the capacity for repetitive neural activity. This result demonstrates a way that sedation-induced expression of slo can act as a neural excitant in Drosophila. These experiments are our first foray into examining the physiological basis of slo-dependent rapid functional tolerance and has been submitted for publication (Ghezzi et al., 2008).

While we were studying the slo gene and becoming more interested in the mechanics of drug tolerance, others in the drug/alcohol addiction field were also converging upon the slo gene. Unbiased genetic screens in C. elegans for alcohol resistant mutants identified mutations in the slo homolog (Davies et al., 2003). In mammals, substantial evidence implicates BK channels, encoded by slo, in the development of functional tolerance in mammals (Reviewed in Roberto et al (2006)). It is highly significant that the same ion channel gene has been independently shown to be involved in ethanol responses in nematodes, insects, and mammals. The functional conservation of slo in ethanol responses makes it likely that work in Drosophila is relevant for understanding the molecular mechanics of tolerance in humans.

Future work will have a strong focus on understanding, in fine detail, how the slo promoter region senses the changed neural activity associated with sedation. However, while studying the role of slo in producing tolerance we have developed methods and skills that are very useful for identifying other genes involved in the process of acquiring tolerance. For example, we developed a highly scalable computer-automated system for scoring tolerance in individual and groups of flies (Ramazani and Atkinson, 2006). We are using these new behavioral assays to perform high-throughput assays for ethanol-induced tolerance. Using such assays, we have identified about six additional genes that are required for the production of ethanol tolerance. Some of these seem to be directly involved in the regulation of the slo gene (transcription factors) or the post-transcriptional modification of the channel protein (a kinase). Two of these genes have been previously connected to drug responses in mammals. Perhaps even more interesting are the two genes that have never previously been connected to drug responses.

I believe that the Drosophila model system, with its powerful genetics, has much to offer to the understanding of how addictive drugs alter the nervous system. I am very excited about the direction that my lab has taken, and I look forward to the many future discoveries that are now within our reach.

References

Cowmeadow, RB, Krishnan, HR, Atkinson, NS (2005) The slowpoke gene underlies rapid ethanol tolerance in Drosophila. Alcoholism: Clinical and Experimental Research, 29:1777–1786.

Cowmeadow, RB, Krishnan, HR, Ghezzi, A, Al'hasan, Y, Wang, YZ, Atkinson, NS (2006) Ethanol Tolerance Caused by slowpoke Induction in Drosophila. Alcohol Clin Exp Res, 30:745–753.

Davies, AG, Pierce-Shimomura, JT, Kim, H, VanHoven, MK, Thiele, TR, Bonci, A, Bargmann, CI, McIntire, SL (2003) A central role of the BK potassium channel in behavioral responses to ethanol in C. elegans. Cell, 115:655–666.

Ghezzi, A, Al-Hasan, YM, Larios, LE, Bohm, RA, Atkinson, NS (2004) slo K+ channel gene regulation mediates rapid drug tolerance. Proc Natl Acad Sci U S A, 101:17276–17281.

Ghezzi, A, Wang, Y, Atkinson, NS (2008) Drug induced slo expression acts as a new neural excitant. Submitted.

Ramazani, RB, Atkinson, NS (2006) Computer automated quantification of movement in Drosophila melanogaster. Journal of Neuroscience Methods, 162:171–179.

Roberto, M, Treistman, SN, Pietrzykowski, AZ, Weiner, J, Galindo, R, Mameli, M, Valenzuela, F, Zhu, PJ, Lovinger, D, Zhang, TA, Hendricson, AH, Morrisett, R, Siggins, GR (2006) Actions of acute and chronic ethanol on presynaptic terminals. Alcohol Clin Exp Res, 30:222–232.

Statement of Research Interests

©2010. Nigel S. Atkinson. All Rights Reserved