Michael Rosbash is interested in circadian rhythms, which reflect a circa 24-hour pacemaker that is nearly ubiquitous among higher organisms. The same mechanism and proteins function in mammals and flies, indicating deep conservation and making Drosophila an important model system. A myriad of tissues and systems—including much of animal biochemistry, endocrinology, and physiology—are under clock control. Circadian clocks also function within the brain to influence sleep.
MicroRNAs, Neurons, and Behavior
MicroRNAs (miRNAs) function in translational control, which has played only a minor role in our understanding of circadian rhythms. Nonetheless, we and others have begun to address the contribution of miRNAs to circadian rhythms, in flies and in mammals. To address a possible relationship between miRNAs and circadian rhythms, we subjected Drosophila miRNAs to high-throughput sequencing "around the clock," meaning at different circadian times, and found miRNAs that undergo substantial circadian oscillations in abundance. One cluster of miRNAs is transcribed in a circadian manner; its six mature miRNAs must therefore be rapidly degraded, which is a requirement for circadian regulation of RNA level. Synthesis appears to take place predominantly in the fat body, and the system appears to be related to the timing of feeding behavior. Interestingly, feeding also appears to regulate the temporal phase of synthesis.
More recently, we identified several different miRNAs that undergo a similar diurnal regulation in level but only in specific circadian neurons. These miRNAs appear to be under light as well as circadian clock control (different regulation for different miRNAs), which presumably reflects unusual features of translational control within these specific neurons. Literature is sparse on the topic of the relationship between neuronal function and miRNA levels, and our preliminary data suggest that at least some of these regulated miRNAs contribute to circadian timekeeping. We are keen to understand in detail the circadian and behavioral contributions of these miRNAs as well as how they are regulated. This project is multidimensional, extending from genetics and RNA biology to behavior.
Sleep and Aging: Cause or Effect?
It is generally agreed that sleep is essential in all complex organisms, essential meaning an animal fully deprived of sleep dies prematurely. Moreover, studies in people over ~50 years indicate that both short sleep and long sleep correlate with reduced life span, raising two possibilities. The first is that covert morbidity shortens or lengthens sleep time, namely, individuals are already in the initial stages of a disease that first manifests itself by causing either too little or too much sleep. The second is that the unusually short or long sleep time is the illness, which reduces life span.
From a more general perspective, work on Drosophila over the past decade has improved our understanding of the nature and regulation of sleep. Indeed, much progress has been made in the identification of genes, anatomical regions, and neurochemicals with strong effects on fly sleep. Among the most important wake-promoting chemicals is dopamine, and a couple of chemicals strongly promote sleep. The circadian system is also important for the normal timing of sleep and wake, that is, the daily regulation of sleepiness and alertness. What is more poorly understood is homeostatic regulation, or whatever keeps track of sleep need. For example, sleep deprivation leads to increased sleep or "sleep rebound" (and an increased depth of sleep in mammals), and this regulation is largely independent of circadian time, in flies as well as people.
Aging also affects sleep regulation; this effect is superficially similar in humans and flies. The strongest effect of aging is on sleep consolidation, or sleep fragmentation, which is, roughly speaking, an increase in the number of awakenings per night, whereas total sleep time is only modestly affected by aging. We also found that sleep rebound is compromised in old flies, suggesting a relationship between aging and homeostatic sleep regulation, which is explored in more detail below.
We would like to know whether modest and chronic sleep deprivation, or even an increase in sleep fragmentation, affects fly life span, because these levels of deprivation mimic the modern human lifestyle. (Note: Total sleep deprivation has been reported to be lethal in mammals.) The strategy of activating dopaminergic neurons with the temperature-sensitive dTrpA1 cation channel and mild temperature elevation is gentle and therefore amenable to long-term use.
We are also using biochemical methods to look for molecular signatures of sleep deprivation and sleep rebound, and we are looking for mutants that enhance or suppress the effects of the sleep manipulations in old versus young flies. We can perhaps use lethality as an assay if there is a life-span effect. Screening directly for modifiers in older flies and subscreening in young flies might reveal interesting suppressors and enhancers with milder or perhaps even no effect in younger flies. This strategy is designed to reveal vulnerabilities or idiosyncrasies of the aging nervous system, which might in turn reveal an interesting aspect of homeostatic sleep regulation, aging, or their interface.
Neuron-Specific Gene Expression and Circadian Locomotor Activity
The circadian network in Drosophila consists of ~75 pairs of neurons in the central brain. Rhythmic circadian locomotor activity of flies and many insects is crepuscular, namely, two activity peaks occur per day: One is centered on dawn and one on dusk. We and others have shown that these two activity peaks derive from two subgroups of the 75 circadian neurons: morning and evening cells. Recent experiments have shown that each of these activities can be further localized to just 4 morning neurons and 4 evening neurons. This statement comes in part from the fact that clock gene expression in only these 8 neurons is sufficient to rescue much if not all circadian behavior in an otherwise arrhythmic clock gene mutant animal. We are interested in the relevant features of these 8 neurons; how they differ; and how they communicate, not only with each other, but also with other relevant circadian neurons and output pathways. Part of this project involves the spatial and temporal profiling of differential gene expression by RNA-Seq. Spatial profiling is a comparison of different neuronal subtypes, primarily morning versus evening neurons, and temporal profiling is across the 24 hours of the day. The data may also indicate important RNA editing events, because of the spatial or temporal regulation of editing activity on specific genes. RNA editing is known to affect the properties of many important neuronal proteins. Candidate genes of interest are identified and then assayed for their effect(s) on behavior or on neuronal communication, by RNA interference for example.