This focus of this project is the genetics of the response to general anesthetics. General anesthesia is not only a vital tool for the practice of surgery but also a potentially powerful resource for the exploration of the nervous system. Anesthetized patients are sedated, have no motor or autonomic responses to painful stimuli, and do not form memory traces of ongoing events. Nevertheless, a variety of monitoring techniques demonstrate that both spontaneous and evoked neural activity persist during anesthesia; it is the higher order functions that are lost. Thus, general anesthetics could point the way toward identifying the molecular and cellular substrates that underlie this level of communication. In seeking to harness the scientific insights offered by anesthesia, genetics should be a valuable complement to physiological and biochemical approaches. For a model genetic organism, I chose Drosophila because it seemed to have the proper combination of simple husbandry and a complex nervous system. Our early work established that Drosophila and vertebrates have similar sensitivities to a broad array of volatile anesthetics, suggesting conservation of the molecular targets. We also showed that at least one neural circuit continued to function quite well while anesthetics had rendered the fly immobile and unresponsive. Thus, just as in vertebrates, the Drosphila nervous system appears to have elements with differential sensitivity to anesthetics. We concluded that the fruit fly was a proper model organism for this study of and established two lines of research. In one, we use behavioral tests to identify genetic loci that are important for anesthesia sensitivity. During the past year we explored inaF, a gene that regulates expression and function of Trp channels. In the other line of research we set up electrophysiological recording techniques and used them to monitor the response of Drosophila to volatile anesthetics. The driving force behind this effort is the desire to find ways to explore how specific genes influence neural function in the absence and presence of anesthetics. We use two systems?a nerve-muscle preparation from Drosophila larvae and the retina/outer optic lobe of the adult fly. Our principal tool for examining anesthetic effects on visual processing is the electroretinogram (ERG), a record of extracellular currents measured at the surface of the eye in response to a pulse of light. Although the gross shape of the ERG is unaffected by anesthetics, we noted that slowing of the recovery phase at the end of a light pulse. This effect on the ?return to baseline? (RTB) was more marked with ERGs that were elicited by long pulses of light than by short-pulses. To our knowledge, the only descriptions of a slow RTB used mutations in photoreceptor (PR) components needed for the termination of the photoresponse. However,the RTB of typical termination mutants is equally defective after both long and short pulses of light. Further experiments suggested that anesthetics might interfere with a previously unrecognized feedback loop between eye and brain. We examined the ERG phenotype of mutations that block synaptic transmission from PRs to second order neurons and found a slow RTB. Just as with halothane, the effect is most evident after a long light pulse. Intracellular recording from PRs showed that changes in the RTB reflect altered ionic fluxes in PRs and not merely currents that arise in other cell types. It appears that PRs do not function well when their synaptic partners in the brain can not receive a signal from the retina. The implication is that higher order neurons feed back onto PRs and influence their termination kinetics. To test this idea we manipulated cholinergic elements, since acetylcholine is the principal neurotransmitter in the brain but is not synthesized in the retina. Indeed, pharmacological and genetic manipulation of cholinergic transmission caused a slowing of the RTB. While efferent control of primary sensory neurons occurs in other systems, our investigation of anesthetic effects has uncovered a previously unsuspected mode, one that operates in an important model organism and one that serves to influence not the sensitivity but the temporal responsiveness of the system. In earlier work, we found that the principal effect of isoflurane on the larval nerve-muscle preparation is a reduction in excitability of the motorneuron. Ion channels are attractive candidates for the molecules that mediate this effect. To explore the complex relationship between channel function, conduction velocity, calcium influx and neurotransmitter release, we turned to computer simulation and constructed a model of an axon and its terminal bouton. By changing parameters in this model, we could calculate the effect of progressive alteration of ionic conductances (representing the putative anesthetic effects) on the conduction of a simulated action potential and on the resulting Ca2+ currents in the terminal. Neither measure alone was adequate to distinguish among anesthetic effects on different conductances. However, channel modulation did produce a distinct pattern in the relationship between conduction velocity and Ca2+ influx. When a specific channel blocker (tetrodotoxin) was tested experimentally, the pattern of synaptic delay (a measure of conduction velocity) as a function of EJC (which reflects Ca2+ entry into boutons) perfectly followed the pattern predicted by the simulation. Moreover, of the conductances that were modeled, only modulation of a non-voltage-gated (?leak?) K+ conductance produced the pattern determined experimentally for isoflurane. Mutations in the inaF gene make flies more resistant to halothane. The gene was known to be needed for proper expression and function of Trp channels, the principal carrier of light-induced currents in the retina, but nothing was known about the nature of its involvement. Insight came from an in silico reanalysis of the transcript produced by gene. The mRNA had been annotated as generating a modest-sized (241 aa) polypeptide, but database searches with this ORF detected no orthologs. We noticed a smaller ORF (81 aa), located closer to the 5?-end of the message, and found that Drosophila encoded 4 related ORFs, all within a few kb of each other. Importantly, this new family had orthologs in organisms ranging from nematodes to mice and humans. In no case had these ORFs been clearly recognized and annotated; we have apparently uncovered a new family of proteins. In the past year, we have tested our model of this locus. We used RT-PCR to show that the four Drosophila paralogs are alternative exons, each of which is spliced to a common downstream exon. We also assembled a cosmid bearing a 14 k genomic segment that covers all the exons and showed that a single copy restores Trp channel levels. Derivatives of this cosmid with either the 241 or 81 aa ORF disrupted by a frameshift mutation proved that the former was not needed for inaF function. Finally, we modified the cosmid to put a hemagglutinin tag at the carboxy-terminus of the putative 81aa polypeptide. Western blots of head extracts from transgenic flies with anti-HA indeed revealed a band of the predicted size. Moreover, immunohistochemistry showed that the HA tag colocalizes with Trp channels. There has been enormous interest in these channels, but the way they are targeted to the membrane and the way they are opened in response to light remains contention. Our work opens the way for examining the way these channels are influenced by an important regulator and for exploring its contribution to anesthesia sensitivity.
Showing the most recent 10 out of 17 publications
