A Pair of Inhibitory Neurons Are Required to Sustain Labile Memory in the Drosophila Mushroom Body

Current Biology 21, 855–861, May 24, 2011 ª2011 Elsevier Ltd All rights reserved

DOI 10.1016/j.cub.2011.03.069

Report A Pair of Inhibitory Neurons Are Required to Sustain Labile Memory in the Drosophila Mushroom Body Jena L. Pitman,1,2 Wolf Huetteroth,1,2 Christopher J. Burke,1 Michael J. Krashes,1,3 Sen-Lin Lai,1,4 Tzumin Lee,1,5 and Scott Waddell1,* 1Department of Neurobiology, University of Massachusetts Medical School, 364 Plantation Street, Worcester, MA 01605, USA

Summary Labile memory is thought to be held in the brain as persistent neural network activity [1–4]. However, it is not known how biologically relevant memory circuits are organized and operate. Labile and persistent appetitive memory in Drosophila requires output after training from the a0 b0 subset of mushroom body (MB) neurons and from a pair of modulatory dorsal paired medial (DPM) neurons [5–9]. DPM neurons innervate the entire MB lobe region and appear to be pre- and postsynaptic to the MB [7, 8], consistent with a recurrent network model. Here we identify a role after training for synaptic output from the GABAergic anterior paired lateral (APL) neurons [10, 11]. Blocking synaptic output from APL neurons after training disrupts labile memory but does not affect long-term memory. APL neurons contact DPM neurons most densely in the a0 b0 lobes, although their processes are intertwined and contact throughout all of the lobes. Furthermore, APL contacts MB neurons in the a0 lobe but makes little direct contact with those in the distal a lobe. We propose that APL neurons provide widespread inhibition to stabilize and maintain synaptic specificity of a labile memory trace in a recurrent DPM and MB a0 b0 neuron circuit. Results and Discussion Fruit flies form robust aversive or appetitive olfactory memory following a training session pairing odorant exposure with electric shock punishment or sucrose reward, respectively [12, 13]. Olfactory memories are believed to be stored in the output synapses of third-order olfactory system neurons in the mushroom body (MB) [5, 14–16], a symmetrical structure comprised of roughly 2500 neurons on each side of the brain that can be structurally and functionally dissected into ab, a0 b0 , and g neuron systems [5, 17]. Similar to aversive memory, appetitive memory measured 3 hr after training is referred to as middle-term memory and is comprised of a labile anesthesia-sensitive memory (ASM) and an anesthesia-resistant memory (ARM) component [18–20]. Both of these phases and later long-term memory

2These

authors contributed equally to this work address: Beth Israel Deaconess Medical Center, Harvard Medical School, Center for Life Sciences, 3 Blackfan Circle, Boston, MA 02215, USA 4Present address: Howard Hughes Medical Institute, University of Oregon, Eugene, OR 97403, USA 5Present address: Janelia Farm Research Campus, Howard Hughes Medical Institute, Ashburn, VA 20147, USA *Correspondence: [email protected] 3Present

(LTM) require the action of the dorsal paired medial (DPM) neurons [5–9, 21, 22]. DPM neurons exclusively innervate the lobes and base of the peduncle regions of the MB [9, 21], where functional imaging suggests they are pre- and postsynaptic to MB neurons [7]. DPM neuron projections to the a0 b0 MB neuron subdivision appear to be of particular importance, and blocking output from a0 b0 neurons themselves during a similar time period after training phenocopies a DPM neuron block [5, 8, 9]. These data led us to propose that reverberant activity in a recurrent MB a0 b0 -to-DPM neuron circuit is required to hold labile memory and for consolidation to LTM within ab neurons [5, 9, 23, 24], where output is critical for retrieval of LTM [9, 23, 24]. As part of a screen for additional neurons contributing to appetitive memory processing after training, we tested for a role of the second-order olfactory projection neurons (PNs). We expressed a uas-shibirets1 transgene [25] with the two most frequently utilized PN GAL4 drivers, GH146 and NP225 [26]. The uas-shits1 transgene allows one to temporarily block synaptic transmission from specific neurons by shifting the flies from the permissive temperature of <25 C to the restrictive temperature of >29 C. We tested appetitive olfactory memory in GH146;uas-shits1 and NP225;uas-shits1 flies in parallel with control flies harboring the GAL4 drivers or the uas-shits1 transgene alone. We also tested c316/uas-shits1 flies, in which the DPM neurons were blocked for comparison. No defects were apparent when the flies were trained and tested at the permissive temperature (Figure 1A). To test for a role after training, we trained all flies at 23 C and immediately after training shifted them to 31 C for 2 hr to disrupt neurotransmission from PN or DPM neurons. All flies were then returned to 23 C and tested for 3 hr memory. Memory was significantly impaired by GH146;uas-shits1 and c316/uasshits1 manipulation, but not by NP225;uas-shits1. The performance of GH146;uas-shits1 and c316/uas-shits1 flies was significantly different from their respective control flies. In contrast, the performance of NP225;uas-shits1 flies was not significantly different from control flies (Figure 1B). GH146 and NP225 label a large number of largely overlapping PNs [26]. However, because NP225;uas-shits1 flies did not exhibit a memory defect (n = 24), we concluded that other neurons labeled by GH146 that are downstream of PNs could be responsible for the observed memory defect. GH146 most obviously differs from NP225 by also expressing in two anterior paired lateral (APL) neurons that innervate the MB [10, 11, 26]. Each APL ramifies throughout the entire ipsilateral MB [10, 11]. This anatomy is similar to the DPM neurons, which project ipsilaterally throughout the MB lobes and base of the peduncle [10, 21] (Figure 1C). We therefore further investigated whether the APL neurons were required for memory processing after training. The NP5288 and NP2631 GAL4 lines have also been reported to label the APL neurons [10]. NP5288 is expressed in a subset of PNs similar to that of NP225 [27], as well as a few other distributed neurons in the brain. NP2631 does not label PNs but labels many other neurons in the brain including those in the median bundle, protocerebral bridge, and subesophageal ganglion. We tested the consequence on memory of

Current Biology Vol 21 No 10 856

Figure 1. Blocking APL or DPM Neurons after Training Disrupts Memory Temperature-shift protocols are shown pictographically above each graph. (A) The permissive temperature of 23 C does not affect 3 hr appetitive odor memory in any of the lines used in this study. All genotypes were trained and tested for 3 hr memory at 23 C. Data in (A) and (B) are mean 6 standard error of the mean (SEM). (B) Blocking APL or DPM neurons after training impairs 3 hr appetitive odor memory. Flies were trained at 23 C and immediately after training were shifted to 31 C for 2 hr. Flies were then returned to 23 C and tested for 3 hr odor memory. *p < 0.05 versus relevant controls. (C) Three-dimensional reconstruction of a fly brain showing a single APL (red) and DPM (green) neuron to illustrate the location of their soma and their ipsilateral projections in the mushroom body (MB).

blocking synaptic output after training from the neurons labeled in these additional APL-expressing lines. As before, no apparent defects were observed when the flies were trained and tested at the permissive temperature (Figure 1A). However, flies trained at 23 C, shifted to 31 C for 2 hr after training, and tested for 3 hr memory at 23 C revealed defective memory. Memory performance of NP5288;uas-shits1 and NP2631;uas-shits1 flies was statistically different from the performance of their genetic control groups (Figure 1B). These data are consistent with a role for APL neurons in memory processing after training. Others have reported that combining a ChaGAL80 transgene with GH146 inhibits expression in the APL neurons but leaves expression in PNs relatively intact [11]. We utilized this approach to further test the requirement of uas-shits1 expression in APL neurons for our observed memory defects. We combined the ChaGAL80 transgene with the GH146, NP5288, and NP2631 GAL4 drivers and uas-mCD8::GFP to visualize the extent of GAL4 inhibition by ChaGAL80 in these flies. As described for GH146;ChaGAL80 flies [11], confocal imaging of the GFP-labeled brains revealed that the ChaGAL80 transgene efficiently suppressed APL expression. The APL neurons were evident in all flies lacking ChaGAL80 but were not labeled in any of the three genotypes containing ChaGAL80 (Figures 2A–2F). ChaGAL80 affected the expression in other neurons labeled by each GAL4 line to varying degrees. Our analysis revealed a strong inhibition in GFP expression in the PNs labeled by GH146 (Figure 2D) and

NP5288 (Figure 2E), although more PNs retained expression in NP5288 than in GH146, consistent with these two GAL4 drivers labeling partially nonoverlapping PN populations. ChaGAL80 inhibited APL expression in NP2631 and also removed expression from several other neurons (Figure 2F). Expression was lost in some neurons innervating the subesophageal ganglion, whereas robust expression remained in the median bundle and protocerebral bridge of the central complex. Unfortunately, several intersectional approaches to create more specific control of APL neurons were unsuccessful (see Supplemental Results available online). We next combined ChaGAL80 with each APL-expressing GAL4 driver and the uas-shits1 transgene to test whether APL expression was necessary for the observed memory phenotypes when GH146, NP5288, and NP2631 neurons were blocked after training (Figure 2G). We assayed memory performance of GH146, NP5288, and NP2631 flies expressing uas-shits1 with or without the ChaGAL80 transgene along with GAL4;ChaGAL80 and uas-shits1 control flies for comparison. We again trained flies at 23 C, shifted them to 31 C for 2 hr after training, and tested 3 hr appetitive memory at 23 C. This manipulation significantly impaired memory performance in all flies without the ChaGAL80 transgene but not in flies with the ChaGAL80 transgene. Memory performance of GH146;uas-shits1, NP5288;uas-shits1, and NP2631;uas-shits1 flies was significantly different from uas-shits1 and GAL4; ChaGAL80 flies. In contrast, memory performance of all flies also harboring the ChaGAL80 transgene was not significantly different from the performance of the genetic control flies (Figure 2G). These data suggest that expression in APL neurons is critical to disrupt 3 hr memory when blocking neurotransmission after training. The memory experiments described did not disrupt synaptic transmission during training or testing. Nevertheless, to control for possible confounding effects, we tested the olfactory acuity and motivation to seek sucrose in naive flies following

Neural Circuitry for Labile Memory in Drosophila 857

Figure 2. Expression in APL Neurons Is Required for the Memory Defect (A–F) Projection view of a brain from a GH146;uasmCD8::GFP (A), NP5288;uas-mCD8::GFP (B), NP2631; uas-mCD8::GFP (C), GH146;ChaGAL80/ uas-mCD8::GFP (D), NP5288;ChaGAL80/uasmCD8::GFP (E), and NP2631; ChaGAL80/uasmCD8::GFP (F) fly. Insets in (C) and (F) represent single optical sections at the depth of the horizontal lobes or the calyx of the MB. Whereas NP2631 flies (C) show clear APL innervation in the MB calyx and horizontal lobes, this expression is missing in NP2631;ChaGAL80 flies (F). Scale bar represents 50mm. (G) Removing expression from APL neurons with ChaGAL80 reverses the observed memory defects. *p < 0.05 versus relevant controls by analysis of variance (ANOVA). Data are mean 6 SEM.

a 2 hr disruption of synaptic transmission and 1 hr recovery as employed in the memory experiments (Table S1). No olfactory acuity defects were observed in GH146;uas-shits1 or NP5288;uas-shits1 flies. However, NP2631;uas-shits1 flies exhibited a pronounced defect, which questions the validity of the memory experiments with this line (Table S1). We therefore rely on the GH146;uas-shits1 and NP5288;uas-shits1 flies and the comparison to NP255;uas-shits1 flies to draw our conclusions. GH146;uas-shits1 and NP5288;uas-shits1 flies also exhibited sucrose acuity that was statistically indistinguishable from uas-shits1 controls. NP5288;uas-shits1 flies performed better than NP5288, which had an apparent defect (Table S1). These data suggest that 3 hr appetitive memory requires synaptic output from the APL neurons after training, similar to the requirement for output from DPM [8, 9] and MB a0 b0 neurons [5]. DPM neuron output is also required after training for appetitive LTM [9]. We therefore used GH146 and NP5288 to test whether APL block disrupted LTM. We blocked APL output for 2 hr after training and tested 24 hr memory. Surprisingly, performance of GH146;uas-shits1 and NP5288;uas-shits1 flies was not significantly different from uas-shits1 or GAL4 flies (Figure 3A), suggesting that APL output is specifically required for

an earlier memory phase. Appetitive memory at 3 hr has been shown to be sensitive to cold-shock anesthesia delivered 2 hr after training [20]. We therefore tested whether APL block affected this labile component by performing experiments with cold shock (Figure 3B). We trained wild-type flies, subjected half of them 2 hr afterward to a 2 min cold shock, allowed them to recover at room temperature, and tested 3 hr memory. Performance of these flies was significantly different from those not receiving a cold shock, consistent with previous literature [20]. Interestingly, the performance of GH146;uas-shits1 flies in which APL neurons were blocked for 2 hr after training was statistically indistinguishable from cold-shocked wild-type flies. To further test whether APL-blocked flies were missing the cold-shock-sensitive memory component, we combined the shits1 block and cold-shock treatments. We trained GH146;uas-shits1 flies, blocked APL after training by shifting flies to 31 C for 105 min, returned them to 25 C for 15 min, gave them a 2 min cold shock, and tested 3 hr memory (Figure 3B). The performance of these flies was statistically different from GH146;uas-shits1 flies that received all treatment except the cold shock, suggesting that some ASM was present in GH146;uas-shits1-blocked flies. Importantly, memory performance was not totally abolished. Because significant memory remained following the uas-shits1 block and the cold shock, we conclude that APL neuron block largely affects the labile anesthesia-sensitive appetitive memory. However, it is worth noting that APL block and cold shock cannot be considered to be operationally equivalent because the 2 min cold shock at 2 hr reduced memory observed at 24 hr (Figure S1), whereas blocking APL for 2 hr did not impact 24 hr memory (Figure 3A). Therefore, blocking GH146 and NP5288 neurons appears to be more specific to labile appetitive memory than cold-shock treatment at this time. The APL neurons are known to be GABAergic [11], and the finding that they are critical for labile memory is consistent with inhibitory input stabilizing the putative MB-DPM recurrent

Current Biology Vol 21 No 10 858

Figure 3. Blocking APL Neurons after Training Does Not Impair Long-Term Memory Temperature-shift protocols are shown pictographically above each graph. Data are mean 6 SEM. (A) Flies were trained at 23 C, shifted to 31 C immediately after training for 2 hr, returned to 23 C, and tested for 24 hr odor memory. *p < 0.05 versus relevant controls by ANOVA. (B) Cold-shock treatment at 2 hr impairs 3 hr appetitive memory. Wild-type flies were trained and tested at the permissive temperature or received cold shock at 2 hr and were tested for 3 hr odor memory. GH146;uas-shits1 flies were trained at 23 C, shifted immediately after training to 31 C for 105 min, returned to 23 C, received a 2 min cold shock 15 min later, and were tested for 3 hr odor memory. *p < 0.05 between indicated groups by ANOVA. GH146;uas-shits1 + cold shock flies were significantly different from zero (p < 0.001, Mann-Whitney U). See also Figure S1.

network [28]. We therefore examined the anatomy of APL neurons in relation to DPM and MB neurons. Prior work has shown that APL neurons ramify throughout the calyx, peduncle, and all lobes of the MB [10, 11], whereas DPM neuron projections are confined to the lower peduncle and MB lobes [10, 21]. We simultaneously imaged DPM and APL neuron projections onto the MB by expressing uasmCD8::mCherry in DPM neurons with the c316-GAL4 driver and QUAS-mCD8::GFP in APL neurons with an APL-expressing GH146-QF driver [29] (Figures 4A–4C). This analysis revealed that APL neurons have a more reticular structure throughout the lobes (Figure 4A), whereas DPM processes are punctate and are most dense in the a0 b0 lobes (Figure 4C). APL processes are interspersed with those of the DPM in regions of overlapping innervation (Figure 4B). To determine possible sites of cell-cell contact, we used GFP reconstitution across synaptic partners (GRASP) [30, 31]. GRASP is detectable when neurons expressing complementary parts of an extracellular split-GFP are close enough that functional GFP is reconstituted [30]. We constructed flies that express lexAop-mCD4::spGFP11 [31] in MB with 247-LexA and uas-mCD4::spGFP1-10 [31] in APL or DPM with NP5288 or c316-GAL4. This analysis revealed distinct innervation of the MB by DPM and APL. DPM-MB GRASP was very dense and punctate throughout the MB lobes and peduncle and generally resembled the mCD8::GFP pattern covering all the major MB lobe regions (Figure 4D; Movie S1). APL-MB GRASP was most notable for structure that is absent (Figure 4F; Movie S3). The regular net-like appearance of APL seen with mCD8::GFP (Figure 4A) was not apparent, and label mostly decorated fibers running in parallel with MB neurons in the lobes. APL-MB GRASP in the vertical lobes was particularly revealing. Whereas the APL mCD8::GFP network extended throughout the vertical lobes (Figure 4A and for APL mCD8::mCherry in Figure 4I), APL-MB GRASP labeled processes extending in the a0 lobe but very little in the a lobe (Figure 4F). Because GRASP is most reliably an indicator of proximity rather than connectivity, these data indicate that

much of the APL network is distant to the MB neurons in the a lobe. We also used GRASP to visualize contact between APL and DPM neurons using NP5288-GAL4 for APL and L0111-LexA for DPM (Figure 4E; Figure S2; Movie S2). APL-DPM GRASP revealed punctate labeling throughout the MB lobes that was most dense in the a0 b0 lobes and base of the peduncle region (Figure 4E). We conclude that APL contacts DPM and MB neurons preferentially in the a0 lobe. In the horizontal lobes, APL contacts DPM throughout and makes dense contact with proximal portions of the MB b, b0 , and g neurons. The density of contact decreases toward the distal end of each horizontal lobe. It seems plausible that APL contacts other unidentified neurons, especially in the areas where they are apparently avoiding MB neurons. We also labeled presynaptic active zones in APL and DPM neurons by expressing a uas-Bruchpilot::GFP [32] with a mCD8::mCherry [33] transgene that should label the entire cell surface. Brp::GFP driven in APL with GH146 revealed presynaptic zones throughout the MB lobes with elevated levels in the a0 b0 lobes (Figures 4G–4I). In contrast, Brp::GFP driven in DPM neurons with c316 revealed presynaptic zones throughout the lobes but very pronounced labeling in the ab lobes (Figures 4J–4L). The anatomical data are consistent with a model of a recurrent MB a0 b0 -DPM-APL circuit and flow of activity from the a0 b0 lobes through the DPM neurons to the ab lobes (Figure S3). Importantly, GRASP suggests that APL and DPM contact is most dense within the a0 b0 lobes, and Brp::GFP indicates strongest APL neurotransmitter release in a0 b0 . APL-MB GRASP indicates that APL preferentially contacts a0 b0 MB neurons (most apparent in the vertical lobes). Interestingly, others have found that APL and DPM neurons are electrically coupled via heterotypic gap junctions ([34]; this issue of Current Biology). It will therefore be important to determine whether APL-DPM contact in a0 b0 is exclusively electrical or a mixture of electrical and chemical. In conclusion, we have identified a role after training for synaptic output from the GABAergic APL neurons [10, 11]. APL neurons appear to be specifically required for labile memory, and not for consolidation of long-term memory. APL and DPM neurons are functionally connected, yet outside

Neural Circuitry for Labile Memory in Drosophila 859

A

B

C

Figure 4. APL and DPM Processes in the Mushroom Body Are Morphologically Different

(A–C) Single confocal sections through a GH146QF/QUAS-mCD8::GFP;c316/uas-mCherry fly brain at the level of the horizontal lobes of the MB. (A) APL-driven mCD8::GFP (green) reveals an even reticular network across all MB subdomains and denser labeling in the heel region. (B) APL processes intermingle with mCherryD E F marked DPM processes (red). (C) DPM processes are most dense in the a0 b0 lobes and fairly sparse in the ab core. (D) Voltex projection of the lobe region of DPM-MB GRASP (MBGAL80/lexAop-mCD4::spGFP11;c316, 247-LexA/uas-mCD4::spGFP1-10). The GRASP signal reveals dense contact between DPM and MB neurons throughout the DPM innervation pattern. (E) A voltex projection of DPM-APL GRASP G H I (NP5288,L0111-LexA/lexAop-mCD4::spGFP11;uasmCD4::spGFP1-10). Sparse contacts are present in all of the lobes and particularly evident in the a0 b0 lobes (outlined) and the base of the peduncle. (F) Voltex projection of the lobe region of APL-MB GRASP (NP5288/lexAop-mCD4::spGFP11;247LexA/uas-mCD4::spGFP1-10). Contacts between APL and MB neurons are most obvious in the heel and the region of each lobe most proximal to the junction. Unlike the evenly distributed net-like APL innervation pattern, APL-MB J K L contacts diminish toward the tips of the lobes, most obviously in the distal a lobe where no APL-MB proximity is apparent. (G–I) Single confocal sections at the level of the MB lobes from a GH146/uas-Brp::GFP,uas-mCD8:: mCherry fly brain. (G and H) Brp::GFP labels putative synaptic active zones of APL throughout the MB lobes, with denser labeling in the a0 b0 lobes. (I) Coexpressed mCherry reveals normal APL morphology. (J–L) Single confocal sections at the level of MB lobes from a uas-Brp::GFP,c316/uas-mCD8::mCherry fly brain. (J and K) Brp::GFP labels putative synaptic active zones of DPM throughout the MB lobes, with densest labeling in the ab lobes. (L) Coexpressed mCherry shows normal DPM morphology. Scale bar in (K) represents 25 mm and applies to all panels. See also Figures S2–S4.

of labile memory described here, disrupting either neuron can have different consequences. First, reducing GABA synthesis in APL neurons enhances learning [11], whereas DPM neurons are not required during acquisition [6, 8]. Functional imaging data suggests that learning specifically increases DPM neuron activity but reduces APL activity driven by the conditioned odor [7, 11]. We suspect that these differences relate to APL also having processes in the MB calyx, where GRASP suggests that APL directly contacts MB neurons (Figure S4). Second, APL neurons are only required for earlier labile memory, whereas DPM neurons are required for labile and consolidated memory. We suspect that this reflects the mode and function of their respective transmitters. We propose that APL provides broad nonselective cross-inhibition to maintain synaptic specificity in the recurrent DPMMB-APL circuit that was originally set by the conditioned odor at acquisition. DPM in contrast might return activity to MB a0 b0 neurons and supply consolidating signals to MB ab neurons (see model in Figure S3). We expect that additional neurons contribute to the network and await identification. It will also be important to gain exclusive control of APL neurons. Active memory storage is thought of mostly on a secondsto-minute timescale in mammals [1, 2]. ASM in Drosophila

suggests a prolonged-duration active memory system. It will be important to determine the physiological property that is ‘‘held’’ in the putative recurrent network. A step change in membrane potential accompanies periods of persistent activity in the oculomotor neural integrator of the goldfish [35]. Such a change in the MB neurons coding olfactory memory would render them more easily excited by the conditioned odorant. Physiology will be needed for us to definitively add ASM in Drosophila to goldfish gaze stabilization [35] and head direction [36] and prefrontal cortical circuits in mammals [1, 4] as models to understand how memory is stored as persistent activity in recurrent neural networks. Nevertheless, the architecture and prolonged requirement for neurotransmission within the MB-DPM-APL neural circuit are suggestive. In addition, a recent gene profiling study of developing vertebrate cortex and annelid MB indicates a common evolutionary origin [37]. Experimental Procedures Fly Strains Fly stocks were raised on standard cornmeal food at 25 C and 60% relative humidity. The wild-type strain was Canton-S. The c316, GH146, NP225,

Current Biology Vol 21 No 10 860

NP2631, and NP5288 GAL4 and uas-mCD8::GFP strains were as described in [10, 21, 26, 38]. The uas-shits1 flies [25] carried an insertion on the third chromosome. The ChaGAL80 strain was as described in [11] and was provided by Ronald Davis (Scripps Florida); GH146QF, QUAS-mCD8::GFP, and uas-mCD8::mCherry were as described in [29, 33]. The uas-Brp::GFP transgene was as described in [31], and flies harboring uas-Brp::GFP and uas-mCD8::mCherry were provided by Motojiro Yoshihara (University of Massachusetts Medical School). 247-LexA::VP16 flies were generated by screening hundreds of P element transformation lines carrying the 247 bp D-Mef2 MB enhancer (a gift from Ronald Davis) upstream of LexA::VP16 [33]. L0111 flies expressing LexA in DPM neurons were a generous gift from Ann-Shyn Chiang (National Tsing Hua University, Taiwan). The GRASP reporters lexAop-mCD4::spGFP11 and uas-mCD4::spGFP1-10 were as described in [31] and were provided by Kristen Scott (University of California, Berkeley). Behavioral Analysis For behavior experiments, wild-type and uas-shits1 female flies were crossed to male flies harboring GH146, NP225, c316, NP2631, or NP5288 or to GH146;ChaGAL80, NP2631;ChaGAL80, or NP5288;ChaGAL80 flies. Mixed-sex populations were tested together in all behavior experiments. Flies were food deprived for 18–20 hr before training in milk bottles containing a damp filter paper. The olfactory appetitive paradigm was performed as described in [22]. Odors were 3-octanol (7 ml in 8 ml mineral oil) or 4-methylcyclohexanol (15 ml in 8 ml mineral oil). Following training, flies were transferred into prewarmed vials containing damp filter paper and stored in a temperature-controlled room or incubator at 31 C for 2 hr. For 3 hr memory experiments, vials were returned to 23 C for 1 hr before testing. For permissive-temperature experiments, flies were kept at 23 C at all times. For permissive-temperature experiments incorporating cold shock, flies were trained and kept at the permissive temperature for 2 hr, cold shocked for 2 min, and tested at 3 hr or 24 hr. For restrictive-temperature experiments, flies were trained at 23 C, shifted to the restrictive temperature for 105 min, returned to permissive temperature, and cold shocked 15 min later. The performance index (PI) was calculated as the number of flies in the arm containing the conditioned odor minus the number of flies in that with the unconditioned odor, divided by the total number of flies in the experiment. A single PI value is the average score from flies of the identical genotype tested with each odor. Olfactory acuity was assessed according to [8]. The odor concentrations used for conditioning are not strongly aversive to naive wild-type flies. Therefore, we increased concentrations 5-fold for acuity experiments. Flies were stored in vials with damp filter paper for 2 hr at 31 C and shifted to 23 C for 1 hr prior to testing (Table S1). Sugar-seeking motivation was performed using a protocol based on [39]. Flies were stored in vials with damp filter paper for 2 hr at 31 C and shifted to 23 C for 1 hr prior to testing. Flies were transferred to a T maze and allowed 2 min to choose between an arm containing dry filter paper and an arm containing dry filter paper soaked with a 3 M sucrose solution. Odorless air was pulled through the T maze. Scores were calculated as for a performance index. Statistical analyses were performed using GraphPad Prism. Overall analyses of variance (ANOVAs) were followed by planned pairwise comparisons between the relevant groups with a Tukey’s honestly significant difference post hoc test. In Figure 3B, a Mann-Whitney U test was used to evaluate whether the GH146GAL4;uas-shits1 + cold score was statistically different from zero. Unless otherwise stated, n R 8 trials per genotype. Imaging To visualize native GFP or mCherry, adult female flies were collected 4–9 days after eclosion (1 day for GRASP flies) and brains were dissected in ice-cold 4% paraformaldehyde solution in phosphate-buffered saline (PBS) (1.86 mM NaH2PO4, 8.41 mM Na2HPO4, 175 mM NaCl) and fixed for an additional 60–120 min at room temperature under vacuum. Samples were washed for 10 min three times with PBS containing 0.1% Triton X-100 (PBT) and twice in PBS before mounting in Vectashield (Vector Labs). Imaging was performed on a Zeiss LSM 510 Pascal confocal microscope, and images were processed in Amira 5.2 (Mercury Systems). In some cases, debris on the brain surface and/or antennal and gustatory nerves was manually deleted from the relevant confocal sections to permit construction of a clear projection view of the Z stack. Reconstruction of neurons was performed in Amira 5.2, including an add-on described previously [40]. Brain neuropil surfaces were slightly modified from [41].

Supplemental Information Supplemental Information includes Supplemental Results, one table, four figures, and three movies and can be found with this article online at doi:10.1016/j.cub.2011.03.069. Acknowledgments We thank Monika Chitre for assistance with data collection and Motojiro Yoshihara, Chris Potter, Ron Davis, Kristin Scott, and the Bloomington Drosophila Stock Center for reagents. We are especially grateful to AnnShyn Chiang and Glenn Turner for sharing results and reagents prior to publication. This work was supported by National Institutes of Health grants MH09883 to S.W. and MH80739 and MH83607 to T.L. Received: January 14, 2011 Revised: March 7, 2011 Accepted: March 29, 2011 Published online: April 28, 2011 References 1. Goldman-Rakic, P.S. (1995). Cellular basis of working memory. Neuron 14, 477–485. 2. Wang, X.J. (2001). Synaptic reverberation underlying mnemonic persistent activity. Trends Neurosci. 24, 455–463. 3. Compte, A. (2006). Computational and in vitro studies of persistent activity: Edging towards cellular and synaptic mechanisms of working memory. Neuroscience 139, 135–151. 4. Douglas, R.J., and Martin, K.A. (2007). Recurrent neuronal circuits in the neocortex. Curr. Biol. 17, R496–R500. 5. Krashes, M.J., Keene, A.C., Leung, B., Armstrong, J.D., and Waddell, S. (2007). Sequential use of mushroom body neuron subsets during drosophila odor memory processing. Neuron 53, 103–115. 6. Keene, A.C., Stratmann, M., Keller, A., Perrat, P.N., Vosshall, L.B., and Waddell, S. (2004). Diverse odor-conditioned memories require uniquely timed dorsal paired medial neuron output. Neuron 44, 521–533. 7. Yu, D., Keene, A.C., Srivatsan, A., Waddell, S., and Davis, R.L. (2005). Drosophila DPM neurons form a delayed and branch-specific memory trace after olfactory classical conditioning. Cell 123, 945–957. 8. Keene, A.C., Krashes, M.J., Leung, B., Bernard, J.A., and Waddell, S. (2006). Drosophila dorsal paired medial neurons provide a general mechanism for memory consolidation. Curr. Biol. 16, 1524–1530. 9. Krashes, M.J., and Waddell, S. (2008). Rapid consolidation to a radish and protein synthesis-dependent long-term memory after singlesession appetitive olfactory conditioning in Drosophila. J. Neurosci. 28, 3103–3113. 10. Tanaka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila mushroom body. J. Comp. Neurol. 508, 711–755. 11. Liu, X., and Davis, R.L. (2009). The GABAergic anterior paired lateral neuron suppresses and is suppressed by olfactory learning. Nat. Neurosci. 12, 53–59. 12. Tully, T., and Quinn, W.G. (1985). Classical conditioning and retention in normal and mutant Drosophila melanogaster. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 157, 263–277. 13. Tempel, B.L., Bonini, N., Dawson, D.R., and Quinn, W.G. (1983). Reward learning in normal and mutant Drosophila. Proc. Natl. Acad. Sci. USA 80, 1482–1486. 14. Dubnau, J., Grady, L., Kitamoto, T., and Tully, T. (2001). Disruption of neurotransmission in Drosophila mushroom body blocks retrieval but not acquisition of memory. Nature 411, 476–480. 15. McGuire, S.E., Le, P.T., and Davis, R.L. (2001). The role of Drosophila mushroom body signaling in olfactory memory. Science 293, 1330– 1333. 16. Heisenberg, M. (2003). Mushroom body memoir: From maps to models. Nat. Rev. Neurosci. 4, 266–275. 17. Crittenden, J.R., Skoulakis, E.M., Han, K.A., Kalderon, D., and Davis, R.L. (1998). Tripartite mushroom body architecture revealed by antigenic markers. Learn. Mem. 5, 38–51. 18. Folkers, E., Drain, P., and Quinn, W.G. (1993). Radish, a Drosophila mutant deficient in consolidated memory. Proc. Natl. Acad. Sci. USA 90, 8123–8127. 19. Tully, T., Preat, T., Boynton, S.C., and Del Vecchio, M. (1994). Genetic dissection of consolidated memory in Drosophila. Cell 79, 35–47.

Neural Circuitry for Labile Memory in Drosophila 861

20. Schwaerzel, M., Jaeckel, A., and Mueller, U. (2007). Signaling at A-kinase anchoring proteins organizes anesthesia-sensitive memory in Drosophila. J. Neurosci. 27, 1229–1233. 21. Waddell, S., Armstrong, J.D., Kitamoto, T., Kaiser, K., and Quinn, W.G. (2000). The amnesiac gene product is expressed in two neurons in the Drosophila brain that are critical for memory. Cell 103, 805–813. 22. Tamura, T., Chiang, A.S., Ito, N., Liu, H.P., Horiuchi, J., Tully, T., and Saitoe, M. (2003). Aging specifically impairs amnesiac-dependent memory in Drosophila. Neuron 40, 1003–1011. 23. Keene, A.C., and Waddell, S. (2007). Drosophila olfactory memory: Single genes to complex neural circuits. Nat. Rev. Neurosci. 8, 341–354. 24. Yu, D., Akalal, D.B., and Davis, R.L. (2006). Drosophila alpha/beta mushroom body neurons form a branch-specific, long-term cellular memory trace after spaced olfactory conditioning. Neuron 52, 845–855. 25. Kitamoto, T. (2001). Conditional modification of behavior in Drosophila by targeted expression of a temperature-sensitive shibire allele in defined neurons. J. Neurobiol. 47, 81–92. 26. Thum, A.S., Jenett, A., Ito, K., Heisenberg, M., and Tanimoto, H. (2007). Multiple memory traces for olfactory reward learning in Drosophila. J. Neurosci. 27, 11132–11138. 27. Okada, R., Awasaki, T., and Ito, K. (2009). Gamma-aminobutyric acid (GABA)-mediated neural connections in the Drosophila antennal lobe. J. Comp. Neurol. 514, 74–91. 28. Chance, F.S., and Abbott, L.F. (2000). Divisive inhibition in recurrent networks. Network 11, 119–129. 29. Potter, C.J., Tasic, B., Russler, E.V., Liang, L., and Luo, L. (2010). The Q system: A repressible binary system for transgene expression, lineage tracing, and mosaic analysis. Cell 141, 536–548. 30. Feinberg, E.H., Vanhoven, M.K., Bendesky, A., Wang, G., Fetter, R.D., Shen, K., and Bargmann, C.I. (2008). GFP Reconstitution Across Synaptic Partners (GRASP) defines cell contacts and synapses in living nervous systems. Neuron 57, 353–363. 31. Gordon, M.D., and Scott, K. (2009). Motor control in a Drosophila taste circuit. Neuron 61, 373–384. 32. Wagh, D.A., Rasse, T.M., Asan, E., Hofbauer, A., Schwenkert, I., Du¨rrbeck, H., Buchner, S., Dabauvalle, M.C., Schmidt, M., Qin, G., et al. (2006). Bruchpilot, a protein with homology to ELKS/CAST, is required for structural integrity and function of synaptic active zones in Drosophila. Neuron 49, 833–844. 33. Lai, S.L., and Lee, T. (2006). Genetic mosaic with dual binary transcriptional systems in Drosophila. Nat. Neurosci. 9, 703–709. 34. Wu, C.-L., Shih, M.F.-M., Lai, J.S.-Y., Yang, H.-T., Turner, G.C., Chen, L., and Chiang, A.-S. (2011). Heterotypic gap junctions between two neurons in the Drosophila brain are critical for memory. Curr. Biol. 21, this issue, 848–854. 35. Aksay, E., Gamkrelidze, G., Seung, H.S., Baker, R., and Tank, D.W. (2001). In vivo intracellular recording and perturbation of persistent activity in a neural integrator. Nat. Neurosci. 4, 184–193. 36. Taube, J.S., and Bassett, J.P. (2003). Persistent neural activity in head direction cells. Cereb. Cortex 13, 1162–1172. 37. Tomer, R., Denes, A.S., Tessmar-Raible, K., and Arendt, D. (2010). Profiling by image registration reveals common origin of annelid mushroom bodies and vertebrate pallium. Cell 142, 800–809. 38. Lee, T., and Luo, L. (1999). Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis. Neuron 22, 451–461. 39. Colomb, J., Kaiser, L., Chabaud, M.A., and Preat, T. (2009). Parametric and genetic analysis of Drosophila appetitive long-term memory and sugar motivation. Genes Brain Behav. 8, 407–415. 40. Schmitt, S., Evers, J.F., Duch, C., Scholz, M., and Obermayer, K. (2004). New methods for the computer-assisted 3-D reconstruction of neurons from confocal image stacks. Neuroimage 23, 1283–1298. 41. Rein, K., Zo¨ckler, M., Mader, M.T., Gru¨bel, C., and Heisenberg, M. (2002). The Drosophila standard brain. Curr. Biol. 12, 227–231.