Sehgal Supplementary Figure legends
ms # 2005-07-08291B
Table S1. The GAL4 drivers used are listed alongside their published expression patterns.
Figure S1. Expression patterns of eight GAL4 drivers within the MB peduncle. (a) Expression of a nuclear-targeted GFP reporter was driven by each of eight MB GAL4 drivers. Confocal images of brain whole mounts were deconvoluted, and virtual cross-sections were reconstructed through the MB peduncle proximal and distal to Kenyon cell bodies. See (b) for a schematic representation of the structures labeled. Based upon the plane of the section and proximity to the Kenyon cell bodies, all four fiber tracts may not be visible in the proximal sections. Examples of alpha/beta expression are denoted by arrowheads in the MB-Switch and c309 panels. Note that the core region within each tract shows at best weak expression in MB-Switch and appears absent in c309. (b) Schematic of MB peduncle anatomy proximal and distal to Kenyon cell bodies1. Axons extend from Kenyon cell bodies through the peduncle, where they form four main tracts. Proximal to the Kenyon cell bodies these can be identified in transverse sections of the peduncle as: four bundles of fibers that extend to the core region of alpha/beta lobes, four groups of fibers surrounding each of the core bundles that also extend to alpha/beta lobes, a diffuse group of fibers surrounding the previous two groups that extend to the alpha’/beta’ lobes, and an outer ring of fibers that extends to the gamma lobes. Distal to Kenyon cell bodies the four tracts of core and alpha/beta fibers merge into two groups, and fibers destined for the alpha’/beta’ and gamma lobes become more distinguishable in cross-sections of the peduncle1. (c) Summary of the expression patterns of MB GAL4 drivers. Expression patterns are denoted with “++” for high levels of fluorescence, “+” for moderate levels of fluorescence, “+/-“ for low to barely detectable levels of fluorescence and “-“ for no detectable fluorescence. Since alpha/beta and alpha’/beta’ fibers could not always be distinguished, we have grouped them here for simplicity. Expression of 201Y in the alpha/beta core has previously been documented2. Lack of expression of c309 in the core region has also previously been described3. The panneuronal elavGeneSwitch driver (Fig 1) also expressed weakly in the core region and strongly in alpha/beta-alpha’/beta’ fibers (data not shown).
1.Strausfeld, N.J., Sinakevitch, I. & Vilinsky, I. The mushroom bodies of Drosophila melanogaster: an immunocytological and golgi study of Kenyon cell organization in the calyces and lobes. Microsc Res Tech. 62, 151-69 (2003).
2.Wang, J., Zugates, C.T., Liang, I.H., Lee, C.H. & Lee, T. Drosophila Dscam is required for divergent segregation of sister branches and suppresses ectopic bifurcation of axons. Neuron14, 559-71 (2002).
3.Connolly, J.B., Roberts, I.J., Armstrong, J.D., Kaiser, K., Forte, M., Tully, T. & O’Kane, C.J. Associative learning disrupted by impaired Gs signaling in Drosophila mushroom bodies. Science20, 2104-7 (1996).
Figure S2. Expression of constitutively active PKA with the 238Y or 30Y GAL4 drivers leads to suppression of sleep at night and enhancement of sleep during the day. (a,b) 238Y/UAS-mc*, 30Y/UAS-mc* and control driver/y w females were raised as in Fig 1. Sleep profiles are means of n=15-16 animals. Data from experimental groups and controls are represented by stippled and solid lines, respectively. (c,d) GAL4 expression profiles from 5-9 day old females carrying both the driver and UAS-GFPn transgene. Brains were prepared as in Fig 2.
Figure S3. Sleep architecture of animals expressing PKA under the control of c309 and 201Y drivers. Animals were prepared as in Fig 1 and analyzed as in Figs 3d,e. (a) Sleep bout distribution shifts leftward to reflect overall shorter sleep bouts in c309/UAS-mc* female flies compared to controls. Black bars represent w1118/y w;c309/UAS-mc*; white bars represent w1118/y w;c309/+; grey bars represent w1118/y w;UAS-mc*/+. (b) Average daily number of sleep bouts is similar between w1118/y w;c309/UAS-mc* (black bar), w1118/y w;c309/+ (white bar) and w1118/y w;UAS-mc*/+ (grey bar). (c) Sleep bout distribution in 201Y/UAS-mc* female flies compared to controls. Black bars represent w1118/y w;201Y/UAS-mc*; white bars represent w1118/y w;201Y/+; grey bars represent w1118/y w;UAS-mc*/+. (d) Average daily number of sleep bouts in w1118/y w;201Y/UAS-mc* (black bar) compared to w1118/y w;201Y/+ (white bar) and w1118/y w;UAS-mc*/+ (grey bar) controls. N = 30-61 female flies in each group. All calculations are plotted as averages + SEM. n.s. is not significant and ** p<.01 by 1-way ANOVA. In the case of the latter, posthoc comparison between w;201Y/UAS-mc* and w1118/y w; UAS-mc*/+ showed a significant difference while the difference between w;201Y/UAS-mc* and w1118/y w;201Y/+ was marginally insignificant.
Figure S4. MBSwitch/UAS-mc* female flies show homeostatic rebound sleep following 12 hours of mechanical sleep deprivation that is similar in magnitude to PKA-induced rebound. MBSwitch/UAS-mc* animals were prepared as in Fig 3. After 1.5 days of acclimation, activity and sleep profiles of uninduced animals were measured for 2.5 days in LD. Beginning at day 2.5 (lights off) and ending at day 3 (lights on), experimental animals were shaken for 2 seconds at randomized intervals with a mean inter-shake interval of 20 seconds. Control animals were undisturbed. (a) During the shaking period, marked in the figure as “dep”, activity was elevated in experimental animals (dotted line), but not in genotypically identical control animals (solid line). (b) Sleep in the experimental group was virtually abolished during the shaking period, while a normal sleep profile was maintained in control animals. For approximately 8 hours after the deprivation period, the experimental group slept substantially more than the control group. (c) Sleep in the control group was subtracted from sleep in the experimental group and integrated over time to keep a running tabulation of cumulative change in sleep. During the deprivation period, experimental animals lost ~500 minutes of sleep compared to controls. Following the deprivation period, the experimental animals recovered 137 + 26 min relative to controls before resuming a sleep profile with little net change in sleep relative to controls. Plotted values are derived from means of 21 experimental animals and 25 control animals.
Figure S5. Decreasing excitability in a subset of MB neurons using the EKO transgene increases sleep. Animals were prepared and assayed for sleep as in Fig 4. RU486 suppressed sleep slightly in female flies bearing either the MBSwitch driver or 3 copies of the transgene for UAS-EKO but significantly increased sleep in females carrying the driver/transgene combination. Black bars represent animals that were fed RU486; white bars are controls. From left to right: for w1118/y w;2xUAS-EKO/+;MBSwitch/1xUAS-EKO n = 45; for w1118/y w;;MBSwitch/+ n = 30-31; for w1118/y w;2xUAS-EKO/+;1xUAS-EKO/+ n = 29-49. Plotted values represent averages + SEM. ** p<.01; n.s. is not significant by unpaired t-test.
Figure S6. Ablation of mushroom bodies with hydroxyurea leads to increased activity and decreased rest. (a) Daily locomotor activity of HU-treated (black) vs untreated (white) animals, averaged over 3 days. (b) Average daily locomotor activity divided by waking time for HU-treated (black) vs control (white) animals. (c) Average daily sleep for HU-treated (black) vs control (white) animals. N=59 and 53 for drug and no drug, respectively. All panels depict averages + SEM. ** p<.001; * p<.05 by unpaired t-test.
Figure S7. Model of sleep-regulating circuitry involving MBs. We propose that within the MBs, sleep-promoting neurons (e.g. 201Y) are normally most active at night, and wake promoting/sleep-inhibiting neurons (e.g. c309/MBSwitch) are normally most active during the day (diurnal influences are indicated by upward and downward deflections in sinusoids). Antagonistic signals from these two sets of cells are integrated to generate sleep/wake activity cycles. Dashed arrows indicate possible influences of MB neurons on each other. Constitutive expression of PKA in a single set of relevant MB neurons leads to increased sleep or waking at temporally ectopic times of day. When PKA is constitutively expressed in both sets an increase in daytime sleep accompanies a decrease in nighttime sleep with little net change in total daily sleep (Fig S2). When MBs are ablated the excitatory influence of the integrator increases waking and locomotion.