- Email: [email protected]
β Core Neurons in Long-Term Memory Consolidation in Drosophila
Current Biology 22, 1981–1989, November 6, 2012 ª2012 Elsevier Ltd All rights reserved
http://dx.doi.org/10.1016/j.cub.2012.08.048
Article A Permissive Role of Mushroom Body a/b Core Neurons in Long-Term Memory Consolidation in Drosophila Cheng Huang,1,4 Xingguo Zheng,1,4,5 Hong Zhao,1,4 Min Li,2 Pengzhi Wang,1,6 Zhiyong Xie,1 Lei Wang,1 and Yi Zhong1,3,* 1School of Life Sciences, Tsinghua University, Beijing 100084, PR China 2Department of Neurology, the First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, PR China 3Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724, USA
Summary Background: Memories are not created equally strong or persistent for different experiences. In Drosophila, induction of long-term memory (LTM) for aversive olfactory conditioning requires ten spaced repetitive training trials, whereas a single trial is sufficient for LTM generation in appetitive olfactory conditioning. Although, with the ease of genetic manipulation, many genes and brain structures have been related to LTM formation, it is still an important task to identify new components and reveal the mechanisms underlying LTM regulation. Results: Here we show that single-trial induction of LTM can also be achieved for aversive olfactory conditioning through inhibition of highwire (hiw)-encoded E3 ubiquitin ligase activity or activation of its targeted proteins in a cluster of neurons, localized within the a/b core region of the mushroom body. Moreover, the synaptic output of these neurons is critical within a limited posttraining interval for permitting consolidation of both aversive and appetitive LTM. Conclusions: We propose that these a/b core neurons serve as a ‘‘gate’’ to keep LTM from being formed, whereas any experience capable of ‘‘opening’’ the gate is given permit to be consolidated into LTM.
olfactory conditioning, single-session training, associating an odor with electric shock, only produces short-term, intermediate-term, and weak anesthesia-resistant memories (ARMs) that last less than 1 day, whereas spaced training (ten repetitive trials with 15 min resting intervals in between) induces protein synthesis-dependent LTM that lasts for at least 7 days [2, 12, 13]. However, single trial of appetitive olfactory conditioning, associating an odor with sucrose, is capable of inducing LTM that lasts for days [7, 8]. In the last 20 years, extensive investigations revealed molecular and neural anatomic components underlying aversive and appetitive olfactory LTM, including involvement of cyclic AMP response element-binding protein (CREB)-regulated gene expression [7, 14], the neuronal activities of mushroom body (MB) [7, 15], dorsal-paired medial neurons [16], and dorsal-anteriorlateral neurons [17]. However, the mechanisms of LTM regulation are still far from clear. The current study of the highwire (hiw) gene shed light on this problem. Our analysis of behaviorally identified LTM mutants led to the finding that the hiw gene [18] and its downstream proteins, wallenda (wnd) [19], a mitogen-activated protein kinase kinase kinase (MAPKKK), and basket (bsk) [19], a c-Jun N-terminal kinase (JNK), play an interesting role in affecting number of training trials needed for inducing aversive olfactory LTM. The hiw gene encodes an E3 ubiquitin ligase, which is important in regulation of synaptic growth at the neuromuscular junction [18, 20] and axon guidance in fly central nervous system [21]. We then identified a subset of Kenyon cells, located within the a/b core region of the MB, with intriguing features. The nature of this cluster of neurons led us to propose that they serve as a gate to keep LTM from being formed. Whether an experience is allowed to be consolidated into LTM depends on its ability in ‘‘opening’’ the gate, which is easier for experiences important for survival.
Introduction
Results
One of the common features of memory induction is that the repetitive event induces the long-term memory (LTM), which requires de novo protein synthesis and a postacquisition consolidation process, much more easily than single experience [1, 2]. The repetitive event is viewed to be capable of activating intracellular signaling pathways much stronger than single experience, thereby more likely triggering mechanisms that regulate LTM induction-relevant gene expression [3, 4]. Yet there are well-documented examples of LTM induction through single experience [5–9]. Current work suggests a gating mechanism involved in giving permission of whether an experience should be consolidated into LTM. In Drosophila, repetition is commonly required for LTM induction in many training paradigms [2, 10, 11]. In aversive
We previously constructed and screened 2,021 homozygous adult-viable P{lacW} transposants for lines with defected aversive LTM after spaced training. One identified mutant, named benP1, has a P element insertion within an intron of the hiw gene (insertion site at X:14,918,446). Within this intron also resides another small gene, ben [22] (Figure 1A). Further analysis revealed that the impaired LTM phenotype in the benP1 mutant resulted from altered expression of ben rather than hiw. On the contrary, memory was enhanced in hiw mutants.
4These
authors contributed equally to this work address: Department of Biology, Washington University, St. Louis, St. Louis, MO 63130, USA 6Present address: The School of Molecular and Cellular Biology, University of Illinois, Urbana-Champaign, Urbana, Illinois 61801, USA *Correspondence: [email protected] 5Present
Inhibition of Hiw Activity Facilitates LTM Formation As shown in Figure 1B, 24 hr memory induced by singlesession or two-session spaced aversive training was significantly increased in two hiw mutants, hiwND9 and hiwDN (see Figure 1A) [20], whereas LTM remained unaltered in hiw mutants subjected to regular spaced training. This phenotype did not result from changes in sensorimotor abilities that are essential for the task (see Table S1 available online). To determine the nature of the enhanced memory, we compared the retention curves at various time points after onesession training. hiwND9 flies had a normal memory in the first
Current Biology Vol 22 No 21 1982
Figure 1. Facilitated LTM Formation in hiw Mutants (A) The gene structure of hiw. Gray boxes indicate exons. hiwND9 has a nonsense mutation at amino acid 2,001, and hiwDN has a deletion at the N-terminal of the Hiw protein. (B) hiw mutants displayed significantly enhanced aversive 24 hr memories after one-session or two-session spaced training (p < 0.01, n = 8–10, Student’s t test) but showed normal LTM after ten-session spaced training (p > 0.79, n = 8, Student’s t test). (C) Retention curves were generated by testing conditioned odor avoidance at various time points after one-session training. hiwND9 flies displayed a normal memory performance before 8 hr (p = 0.31, 0.51, 0.17 compared to WT for 0, 3, and 8 hr, respectively, n = 6, Student’s t test) but enhanced memory thereafter (p = 0.001 for 12 hr, Student’s t test, and p = 0.02 for 24 hr, n = 6, Mann-Whitney test). (D) The enhanced 24 hr memory in hiwND9 was blocked in the CXM feeding group (CXM+) (p > 0.5, n = 8, Mann-Whitney test) but was not affected in the vehicle group (CXM2) (p < 0.01, n = 8, Mann-Whitney test). Average data are presented as mean 6 SEM; **p < 0.01, ***p < 0.001. See Table S1 for task-relevant sensorimotor abilities. See also Figure S1.
3 hr after training and a slightly but not significantly elevated memory at 8 hr, whereas they exhibited significantly enhanced memory at the time points of 12 and 24 hr (Figure 1C). The unaltered performance in the first 3 hr suggested that the observed memory enhancement is not due to a strengthened acquisition of the initial memory. To exclude the possibility of a ceiling effect, we determined the acquisition scores of hiwND9 flies at different electric shock intensities (20V, 40V, and 60V) and numbers of shock pulses (4, 8, and 12) (Figure S1A). Consistently, no statistically significant difference was found. Taken together, the results suggested that hiw mutants have normal short-term and intermediate-term memories but increased memory after 12 hr. There are two types of well-described consolidated memory forms that can last more than 1 day: LTM and ARM [2, 23]. Protein synthesis is critical for induction of LTM, but not for ARM. Our results indicated that the enhanced memory was protein synthesis dependent because feeding hiw mutant flies with a protein synthesis inhibitor, CXM (cycloheximide), eliminated the 24 hr memory enhancement (Figure 1D), whereas feeding with vehicle had no effects. Thus, LTM formation was facilitated in hiw mutants such that only a single training session was needed for its induction as compared to the ten-spaced sessions normally required in wild-type (WT) flies. Reported effects of Hiw regulating synaptic growth through E3 ligase activity [20] raised the concern that the observed memory phenotype resulted from abnormal development of the nervous system, instead of acute involvement of the Hiw physiological functions. To address this concern, we used the RU486-inducible GAL4 Gene-Switch (GS) system [24] to acutely manipulate hiw transgene expression in adult flies. Expression of transgenes of either UAS-hiw (the WT hiw+ transgene) or UAS-hiwDRING (the dominant-negative hiw transgene with the abolished ligase activity) [20] was effectively induced in response to 2 days of RU486 feeding after eclosion (Figure S2). Acutely, expression of hiw+ transgene in
the adult nervous system was sufficient to rescue the mutant phenotype (Figure 2A). Additionally, the facilitated LTM phenotype was presented (Figure 2B) in flies with acutely attenuated Hiw ligase activity through pan-neuronal expression of the hiwDRING. Expression of hiwDRING in hiwND9 mutant did not lead to further enhancement in 24 hr memory (Figure 2C), reflecting the specific dominant-negative effect in inhibiting Hiw function. Taken together, these results supported the idea that inhibition of Hiw E3 ligase activity, likely leading to the accumulation of target proteins, acutely caused facilitation of LTM formation. Hiw Functions in the MB a/b Core Region Given Hiw’s role in facilitating LTM formation, we sought to determine the expression pattern of this gene in the adult fly brain. Because no suitable antibody was available for immunocytochemistry staining in adult fly brain [20], we visualized the GAL4 expression pattern of hiwNP4132 (termed hiw-GAL4 below) with a P{GawB} element insertion in an intron of the hiw gene with the same transcription direction (insertion site at X:14,918,437) (Figure 1A). Confocal imaging of the green fluorescent protein (GFP) signal in hiw-GAL4/+;UAS-mCD8:: GFP/+ flies revealed a preferential expression in four major compartments of the adult central brain (Figure 3A; Movie S1), including fan-shaped body (FB), ellipsoid body (EB), glial cells (GCs), and the MB. The other two hiw enhancer-trap lines (NP1041 and NP2284) showed weaker GFP signals but very similar expression patterns to hiw-GAL4 (Figure S3A). A close examination of confocal images mapped expression of hiwGAL4 within the MB to the core region of a/b lobes (Figure 3B). Such images, obtained through double labeling, indicated that hiw-GAL4 labels intrinsic Kenyon cells at the calyx region, the inner part of lobes, and peduncle of MB outlined through MB247-dsRed (Figure 3B, red signal) [25]. The expression pattern of hiw-GAL4 in MB is reminiscent of the MB core neurons by comparing to other MB GAL4 (Figure S3B).
Mushroom Body Subregion Gates Memory Consolidation 1983
Figure 2. Hiw E3 Ligase Activity Acutely Regulates LTM Formation (A) Facilitated LTM in hiwND9 can be rescued by acutely inducing hiw+ expression. In the vehicle group (RU4862), hiwND9/Y;UAS-hiw/+;elav-GS flies displayed the enhanced 24 hr memory after one-session training (p < 0.01, n = 8–10, ANOVA), which was restored to normal levels in the RU486 feeding group (RU486+) (p > 0.43, n = 8–10, ANOVA). (B) Acute expression of E3 ligase activitydead HiwDRING protein in WT background facilitated LTM formation (p < 0.01, n = 8, ANOVA). (C) Acute expression of HiwDRING protein in hiw mutant adults cannot further enhance 24 hr memory (p > 0.9, n = 8, ANOVA). Average data are presented as mean 6 SEM; **p < 0.01. n.s., nonsignificance. See also Figure S2 for the efficiency of acute induction of the elav-GS system.
Furthermore, the MB expression could be eliminated by introducing additional expression of MB-GAL80 [16] targeted to the MB, while leaving expression in all other compartments unchanged (see the effect of removing hiw-GAL4-induced GFP expression specifically within the MB by comparing the two columns in Figure 3B; Figure S3C). Subsequent behavioral assays demonstrated that Hiw regulates LTM formation in MB, specifically within the core region of the a/b lobes. Inhibition of Hiw functions in hiw-GAL4 neurons through expression of UAS-hiwDRING led to the enhanced 24 hr memory after one-session training (Figure 3C), which reliably phenocopied the hiw mutants. Consistently, the learning was unchanged (Figure S3D), and the enhanced 24 hr memory was also protein synthesis dependent (Figure S3E). However, suppression of the MB expression in hiw-GAL4 failed to induce the LTM facilitation. The hiwDRING-induced memory enhancement could also be observed in other GAL4 drivers that include the a/b core region (NP7175, NP6024, and c739) [26, 27], but not in the GAL4 driver that labels a/b posterior and surface neurons (NP5286) [26] or GC (repoGAL4) [28] (Figure 3C). Similar enhancement was also achieved by expressing UAS-hiwRNAi in hiw-GAL4, NP7175, and NP6024 (Figure S3F). The morphological structure of MB was not obviously affected by hiwDRING expression (Figure S4G). In contrast, overexpression of hiw+ transgene in hiw-GAL4 or NP7175 severely impaired LTM formation induced by spaced training (Figure 3D), whereas presence of MB-GAL80 eliminated the LTM impairment. Meanwhile, the learning was unchanged in hiw-GAL4/+;UAS-hiw/+ flies (Figure S3H). Task-relevant sensorimotor responses were not significantly altered in most lines, except NP7175/+;;UAShiwDRING/+ and c739/+;UAS-hiwDRING/+, which performed significantly worse than the UAS-hiwDRING control in odor acuity but indistinguishably from their GAL4 controls (Table S2). Taken together, we showed that inhibition of Hiw activity within a/b core neurons could facilitate LTM formation, whereas excessive Hiw blocks LTM formation. Involvement of Hiw Downstream Targets in the Facilitation of LTM Formation To verify the functional role of Hiw and gain more mechanism insights, we tested the genetic regulation of Hiw downstream signals. Wallenda (Wnd), a MAPKKK homologous to the vertebrate dual-leucine zipper-bearing kinases (DLKs and LZKs, respectively), plays a crucial role in Hiw-mediated synaptic overgrowth [19]. In the established model for Hiw function (Figure 4A), Hiw constrains the protein level of Wnd, which is thought to function as an upstream regulator of the JNK/ MAPK, encoded by bsk in Drosophila [19].
As shown in Figure 4B, reduction of Wnd and Bsk by expressing wnd RNAi or Bsk dominant-negative protein (BskDN) [29] blocked the enhancement of LTM formation induced by expressing UAS-hiwDRING in hiw-GAL4 while leaving learning unaffected (Figure S4A). To further explore whether the accumulation of Hiw downstream protein is sufficient to facilitate LTM formation, we used temperature-sensitive GAL80 system [30] to acutely overexpress Wnd protein. After 4 days’ induction, overexpression of Wnd in hiw-GAL4 and NP7175 both resulted in enhanced LTM formation, which mimicked the phenotype in hiwDRING group (Figure 4C) or hiw RNAi group after induction (Figure S4B). The data also showed that the kinase activity of Wnd was critical for facilitating LTM because expression of kinase-dead Wnd protein failed to induce such phenotype (Figure 4C). In all groups, learning was unaffected (Figure S4C). Western blotting revealed the efficiency of the overexpression of Wnd and, more importantly, showed that expressing hiwDRING did increase the Wnd protein level (Figures 4D and 4E). Moreover, after acute upregulation of Wnd, immunostaining with anti-FASII revealed that the morphology of MB was unaffected (Figure S4D). The results above suggested the involvement of Wnd/JNK pathway in Hiw-regulated LTM formation. MB a/b Core Neurons Permit Aversive LTM Consolidation The facilitated LTM phenotype achieved through manipulation of Hiw activity within MB a/b core neurons posed the immediate question of whether such manipulation acts to affect acquisition, consolidation, or retrieval. To address this question, we used the temperature-sensitive shibire (shi) mutant transgene [31], UAS-shits1, to manipulate synaptic output reversibly. The shits1 gene encodes a mutated Dynamin that allows normal neurotransmitter release at permissive temperatures (below 23 C) but blocks synaptic transmission at restrictive temperatures (above 30 C) [31]. At the permissive temperature (18 C), expression of UASshits1 in all drivers tested had no effects on LTM induced by regular ten-session spaced training (Figure 5A). LTM was also not affected by blockade of synaptic transmission in either hiw-GAL4 or NP7175 during acquisition (Figure 5B) or during retrieval (Figure 5C). However, LTM was impaired when synaptic transmission was blocked during the whole consolidation period in hiw-GAL4 group (Figure S5A) or in the interval 6 hr after completion of spaced training for a period of 12 hr in hiw-GAL4, c739, NP6024, or NP7175 groups (Figure 5D). This memory impairment, observed in the drivers containing the a/b core region, was not shown in repo-GAL4, NP5286, and hiw-GAL4;MB-GAL80 group (Figure 5D).
Current Biology Vol 22 No 21 1984
Figure 3. Hiw Regulates LTM Formation in MB Core Neurons (A) Confocal imaging of GFP expression in the whole-mount adult central brain and thoracic ganglion of hiw-GAL4-driven mCD8::GFP (green) flies revealed preferential expression in four compartments: FB F5 layer neurons, EB, and a subset of the MBs and GCs. Scale bars represent 50 mm. (B) Expression pattern of hiw-GAL4, hiw-GAL4;MB-GAL80 in MB lobes, peduncle, and Kenyon cell region. The whole MB structure was indicated by MB247dsRed-labeled region (red signal), and the GAL4 expression pattern was displayed by mCD8::GFP-labeled region (green signal). Top view shows the expression in the Kenyon cell region. Middle view is the expression in the peduncle. Bottom view demonstrates the expression in MB lobes. In hiw-GAL4 and hiw-GAL4;MB-GAL80 images, the cell bodies of the FB are outlined by white lines. In the hiw-GAL4;MB-GAL80 peduncle and lobe region, white dashed lines outline the a/b core subdivision; yellow arrow indicates some extrinsic innervations onto MB lobes. Scale bars represent 20 mm. (C) Attenuating Hiw activity in hiw-GAL4-labeled neurons was sufficient to facilitate LTM formation. UAS-hiwDRING flies were crossed with indicated GAL4 drivers. Facilitated LTM formation can be seen in hiw-GAL4, NP7175, NP6024, and c739 groups (p < 0.05, n = 10–12, ANOVA), but not in hiw-GAL4; MB-GAL80, NP7175;MB-GAL80, or NP5286 (p > 0.58, n = 10–12, ANOVA and Kruskal-Wallis ANOVA). (D) Overexpression of the WT Hiw protein impaired LTM. hiw-GAL4/+;UAS-hiw/+ or NP7175/+;UAS-hiw/+ flies displayed significantly impaired LTM after ten-session spaced training compared with control flies (p < 0.01, n = 8, ANOVA), where LTM was normal in hiw-GAL4/+;MB-GAL80/UAS-hiw or NP7175/+;MB-GAL80/UAS-hiw flies (p > 0.9, n = 8, ANOVA and Kruskal-Wallis ANOVA). For the X chromosome-located hiw-GAL4, hiw-GAL4;MB-GAL80, NP7175, NP7175;MB-GAL80, and NP6024, only female results are shown. Average data are presented as mean 6 SEM; *p < 0.05, **p < 0.01, ***p < 0.001. See Table S2 for task-relevant sensorimotor abilities. See also Figure S3.
The memory impairment was specific for LTM because blocking synaptic transmission in a/b core neurons within the whole consolidation period or indicated 12 hr time window had no effects on ARM induced by ten-session massed training in tested flies (Figure S5B; Figure 5E), and blocking synaptic output from hiw-GAL4 and NP7175 neurons did not affect learning (Figure S5C). In most cases, task-relevant sensorimotor responses were not significantly altered, except NP7175/+;;UAS-shits1/+ and c739/+;UAS-shits1/+, which performed significantly worse than the UAS-shits1 control in odor acuity but indistinguishably from their GAL4 controls at the permissive temperature (Tables S3 and S4). As shown in Figure 5F, blocking the transmission from hiwGAL4 during consolidation period could also eliminate the
hiwDRING-induced LTM facilitation. Moreover, the similar requirement of core neurons in LTM consolidation was also observed by using another odor pair (OCT/BEN) as shown in Figures S5D–S5F. Thus, synaptic output from the identified a/b core neurons plays a critical role during the period of LTM consolidation in a very specific manner that permits aversive olfactory LTM that is induced by either single-session training or regular spaced repetitive training. MB a/b Core Neurons Also Permit Appetitive LTM Consolidation The suggested permissive function of identified MB a/b core neurons made us curious about its general role in LTM induction. In particular, appetitive LTM is induced with only
Mushroom Body Subregion Gates Memory Consolidation 1985
Figure 4. Involvement of Hiw Downstream Signals in the LTM Facilitation (A) The model of Hiw function. Hiw negatively regulates downstream Wnd/DLK, which is an activator of JNK/MAPK. (B) Reducing Wnd and Bsk suppressed the Hiw-mediated LTM facilitation. Expression of hiwDRING in hiw-GAL4 promoted the LTM formation, whereas coexpression of wnd RNAi or BskDN with hiwDRING failed to induce such LTM facilitation (p < 0.01, n = 8–12, ANOVA). (C) LTM formation was significantly enhanced by acutely overexpressing WT Wnd (UAS-wnd) protein or Hiw dominant-negative protein (UAShiwDRING) in MB core region, labeled by hiwGAL4 and NP7175 (p < 0.05, n = 8–12, ANOVA), whereas expression of kinase-dead Wnd protein (UAS-wndKD) had no such effect (p > 0.54, n = 8, ANOVA and Kruskal-Wallis ANOVA). (D) Western blots showed acutely induced expression of hiwDRING and Wnd in hiw-GAL4elevated Wnd protein level (p < 0.05, n = 6, ANOVA). Left panel is representative of western blots for uninduced and induced group. Right panel shows quantified results. (E) Western blots showed acutely induced expression of hiwDRING and Wnd in hiw-GAL4elevated Wnd protein level (p < 0.05, n = 6, ANOVA). Left panel is representative of western blots for uninduced and induced group. Right panel shows quantified results. Average data are presented as mean 6 SEM; *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S4.
single-session training, rather than ten-session spaced training as in induction of aversive LTM. In the appetitive-learning task, we measured LTM 48 hr later after training. Because flies cannot form or retrieve appetitive memory unless they are hungry [7], flies were food deprived for 36 hr before training, reintroduced to food for 24 hr after completing training (to avoid starvation and heat shock caused death), and then food deprived for 24 hr before testing. Control experiments showed that at the permissive temperature (23 C), one session-induced appetitive LTM was normal in all genotypes tested (Figure 6A). We then demonstrated that disrupting synaptic output from hiw-GAL4-labeled neurons or from a/b lobe core regions (c739 and NP6024) at the restrictive temperature (30 C) in the posttraining period impaired appetitive LTM (Figure 6B), whereas the task-relevant sensorimotor responses were not significantly altered (Tables S5 and S6). However, appetitive LTM was normal in flies when synaptic transmission was blocked in those hiw-GAL4-labeled brain regions outside of the MB (hiw-GAL4;MB-GAL80 group), in the a/b posterior and surface region (NP5286 group), or in GCs (repo-GAL4 group) (Figure 6B). It is interesting that blockade of output from the NP7175-labeled inner part of core neurons alone did not impair appetitive LTM consolidation. This observation can either result from a further functional division within the core region or a complementary effect from NP7175-unlabeled core neurons. We further tested the idea that hiw-GAL4-labeled MB a/b core neurons are specifically engaged in consolidation by
introducing two time windows during which synaptic transmission was blocked. Appetitive LTM was normal when the blockade of synaptic transmission (at 30 C) occurred 12 hr after completing training (Figure 6C) but was defective if synaptic transmission was disrupted immediately after training for 12 hr (Figure 6D). This result showed that disruption of synaptic transmission from hiw-GAL4labeled MB a/b core neurons affected LTM only during the phase of consolidation and thus supported the idea that hiw-GAL4-labeled MB a/b core neurons play a permissive role in gating consolidation of not only aversive but also appetitive LTM. Discussion The current study began with the finding that 24 hr memory resulting from single session was enhanced in two hiw mutant alleles. This enhanced memory component was identified as facilitated LTM, given that it was sensitive to protein synthesis inhibition. The behavioral effect of hiwDRING and the presence of a hiw-GAL4 line allowed us to map the neural circuitry to a cluster of MB a/b core neurons, within which Hiw and its downstream targets regulate LTM. Furthermore, we showed that the MB a/b core neurons were involved in the consolidation of both aversive and appetitive LTM. In conclusion, the observations that the MB a/b core neurons were capable of both facilitating and limiting LTM suggested a working model in which the ability to form LTM is gated through these neurons. The significance of results presented is further elaborated below.
Current Biology Vol 22 No 21 1986
Figure 5. MB a/b Core Neurons Permit Aversive LTM Consolidation Female flies of indicated GAL4 drivers were crossed to WT male flies (+/+) or to UAS-shits1 male flies, and all the progenies were raised at 18 C to minimize any potential ‘‘leaky’’ effect of UAS-shits1 on development. Temperature-shift protocols are shown above each graph. For the X chromosome-located hiwGAL4, hiw-GAL4;MB-GAL80, NP6024, and NP7175, only female results are shown. (A) The permissive temperature (18 C) did not affect LTM after ten-session spaced training of all genotypes used in this study (p > 0.2, n = 8–10, ANOVA and Kruskal-Wallis ANOVA). See also Table S3 for task-relevant sensorimotor abilities. (B) Blocking synaptic output from hiw-GAL4 and NP7175 neurons during the training period did not affect LTM performance (p > 0.9, n = 8, ANOVA). (C) Blocking synaptic output from hiw-GAL4 and NP7175 neurons during the testing period did not affect LTM performance (p > 0.9, n = 8, ANOVA). (D) Blocking synaptic output from hiw-GAL4, c739, NP6024, and NP7175 neurons during the LTM consolidation period (6–18 hr after the end of training) disrupted LTM performance (p < 0.001, n = 8–10, ANOVA and Kruskal-Wallis ANOVA), whereas LTM was not affected in hiw-GAL4;MB-GAL80, NP5286, and repo-GAL4 flies. See also Table S4 for task-relevant sensorimotor abilities. (E) Blocking synaptic output during the memory consolidation period (6–18 hr after the end of training) did not impair ARM after ten-session massed training (p > 0.48, n = 8, ANOVA and Kruskal-Wallis ANOVA). (F) Blocking synaptic output from hiw-GAL4 during the consolidation period eliminated the facilitated LTM caused by hiwDRING expression (p < 0.01, n = 8, ANOVA). Average data are presented as mean 6 SEM; **p < 0.01, ***p < 0.001. See also Figure S5.
Involvement of Hiw and Its Downstream DLK/JNK Pathway in LTM Regulation Not only synthesis but also degradation of proteins plays a critical role in the remodeling of synapses, learning, and memory [32]. Altered memory formation in ubiquitin ligase mutants has
been reported in mice and Drosophila [33, 34]. Here, we reported that Hiw, an evolutionarily conserved E3 ubiquitin ligase, negatively regulated LTM formation through restraining its downstream target Wnd [19]. Our results indicated Hiw function as an inhibitory constraint, as the memory suppressor
Mushroom Body Subregion Gates Memory Consolidation 1987
gene [35], on LTM formation. Removal of this suppressor or direct activation of its downstream signals could lead to the facilitated LTM induction without the repetitive training that is normally required. So far, the physiological consequence of Hiw or Wnd in core neurons still remains an open question. Because of the extensively shared components with hiw’s function in synaptic growth and transmission [18, 19], the attenuated Hiw activity may elevate Wnd level and then lead to the excessive synaptogenesis or abnormal synaptic activity in core neurons. It will also be interesting to test whether the hiw-mediated LTM facilitation shares some common molecular components with the corkscrew-regulated spacing effect in LTM induction [36], in which the MB a/b lobes also play an important role. In a recent report, Hiw regulates the axon guidance in MB [21]. We also observed the morphological defect in MB a/b lobes in a portion of hiw mutant flies (Figure S1B). It is striking that hiw mutants with MB defect can form LTM even more efficiently, given the observation of LTM impairment in another MB structural mutant, ala [15]. However, comparing to the total loss of vertical lobes (including a and a0 ) in ala mutant, most hiw mutants had the abnormal thickness of a/b lobes caused by the unequal distribution of the MB axonal projections between the a and b lobes, and about 39% of hiwDN mutant had the shortened a lobe [21]. One of the possible explanations is that the remaining function of a/b lobe in hiw mutants is sufficient to support the LTM. Moreover, expression of Hiw dominant-negative protein or acutely increasing Wnd protein level in MB was sufficient to promote LTM but did not give rise to any observable gross morphological change in MB (Figures S3G and S4D). Thus, we suggest that Hiw mediates memory phenotype through a different mechanism from the one that led to the structure change in the MB.
Figure 6. MB a/b Core Neurons Permit Appetitive LTM Consolidation Female flies of indicated GAL4 drivers were crossed to WT male flies (+/+) or to UAS-shits1 male flies, and all the progenies were raised at 18 C to minimize any potential ‘‘leaky’’ effect of UAS-shits1 on development. Temperature-shift and food-deprivation protocols are shown above each graph. For the X chromosome-located hiw-GAL4, hiw-GAL4;MB-GAL80, NP6024, and NP7175, only female results are shown. (A) The permissive temperature (23 C) did not affect appetitive LTM after one-session appetitive training (AP training) for all genotypes used in this study (p > 0.22, n = 8-10, ANOVA and Kruskal-Wallis ANOVA). See also Table S5 for task-relevant sensorimotor abilities. (B) Disrupting synaptic output from hiw-GAL4, NP6024, or c739 neurons during the LTM consolidation period could impair appetitive LTM performance (p < 0.01, n = 8–10, ANOVA), but LTM was not affected in hiw-GAL4;MB-GAL80, NP5286, repo-GAL4 flies. See also Table S6 for task-relevant sensorimotor abilities. (C) Disrupting synaptic output from hiw-GAL4 neurons during 12–24 hr after training did not affect appetitive LTM (p > 0.7, n = 8–10, ANOVA). (D) Disrupting synaptic output from hiw-GAL4 neurons during the first 12 hr after training was sufficient to impair appetitive LTM. (p < 0.01, n = 8–9, ANOVA). Average data are presented as mean 6 SEM; **p < 0.01, ***p < 0.001.
MB a/b Core Region in Giving the Permission to Consolidation of LTM The involvement of MB in the hiw-mediated LTM facilitation led us to further probe the function of this structure in LTM regulation. It has been well documented that the MB, a bilateral brain structure that consists of approximately 2,500 neurons in each hemisphere [37], plays the central role in olfactory memories, both aversive and appetitive [7, 16, 38–40]. Intrinsic MB neurons are organized into physically distinct a, b, a0 , b0 , and g lobes [41, 42]. All three lobes exhibited different functions in memory processing, such that the output of a/b lobes was required for retrieval of memory [7, 16, 43], a0 /b0 lobes were transiently required to stabilize memory [16] or to retrieve immediate memory [44], and g lobe mediated rutabagadependent mechanism and dopaminergic signal to support short-term memory (STM) and LTM formation [45, 46]. Moreover, memory traces mapped to different lobes exhibited different temporal features [44, 47–49]. Through gene expression patterns and enhancer trap lines, each lobe of MB can be classified into more specific subgroups such as the posterior, surface, and core regions in the a/b lobes [26, 50]. The current work shows that a/b core neurons play a distinct role in LTM induction. The synaptic outputs of these neurons are critical during consolidation of LTM for both aversive and appetitive conditioning, but these neurons are not involved in LTM cellular consolidation per se because a landmark of LTM cellular consolidation, CREB-mediated protein synthesis, occurs in non-MB neurons [17]. Thus, one of the roles for this cluster of neurons can be viewed as simply providing connections to channel learning
Current Biology Vol 22 No 21 1988
information to the downstream neurons in which LTM is formed. However, the remarkable feature of enabling singletrial induction of aversive LTM through targeted genetic manipulation of this cluster of neurons suggests that they play a unique permissive role in determining whether an experience should be consolidated. Gating LTM This newly identified function for permitting an experience to be consolidated leads us to propose a gating theory. This theory proposes that the a/b core neurons serve as a ‘‘gate,’’ and activation of this gating mechanism functions as a checkpoint that keeps LTM from being formed for general experiences, whereas only specific experiences, capable of ‘‘opening’’ this gate, can and are bound to trigger LTM consolidation and to form LTM ultimately. There is little survival advantage in committing never-to-be-repeated episodes to memory, particularly because the very act of LTM formation may be deleterious to the fly [11]. In contrast, repetitively occurring experience, such as spaced repetitive aversive conditioning, and events critical for survival, such as finding food or single-trial appetitive conditioning, would be able to ‘‘open’’ the gate, and therefore, LTM is formed for such experiences. Experimental Procedures Behavioral Assays Aversive and appetitive Pavlovian olfactory-conditioning procedures were performed as described previously [2, 7, 12]. For details of behavioral procedures, see Supplemental Experimental Procedures. Statistical Analysis Shapiro-Wilk test was used for the normality test. Normally distributed data were analyzed by two-tailed Student’s t test or one-way ANOVA following Bonferroni test (Origin version 8; OriginLab). For the data not normally distributed, Mann-Whitney test and Kruskal-Wallis ANOVA were used. Statistical results are presented as mean 6 SEM. Asterisks indicate critical values (*p < 0.05, **p < 0.01 and ***p < 0.001). ‘‘n.s.’’ indicates no significant difference (p > 0.05). Supplemental Information Supplemental Information includes five figures, six tables, Supplemental Experimental Procedures, and one movie and can be found with this article online at http://dx.doi.org/10.1016/j.cub.2012.08.048. Acknowledgments We thank Aaron DiAntonio, Catherine A. Collins, Chunlai Wu, Ronald L. Davis, Scott Waddell, Kei Ito, and Andre´ Fiala for providing flies and reagents. We are grateful for helpful comments on the manuscript from Jennifer Beshel, Yichun Shuai, and Lisha Shao. This work was supported by grants from the National Basic Research Project (973 program) of the Ministry of Science and Technology of China (2006CB500806 and 2009CB941301) and the Tsinghua University Initiative Scientific Research Program (20111080956, all to Y.Z.). Received: May 25, 2012 Revised: July 18, 2012 Accepted: August 21, 2012 Published online: October 11, 2012 References 1. Carew, T.J., Pinsker, H.M., and Kandel, E.R. (1972). Long-term habituation of a defensive withdrawal reflex in aplysia. Science 175, 451–454. 2. Tully, T., Preat, T., Boynton, S.C., and Del Vecchio, M. (1994). Genetic dissection of consolidated memory in Drosophila. Cell 79, 35–47.
3. Kandel, E.R. (2001). The molecular biology of memory storage: a dialogue between genes and synapses. Science 294, 1030–1038. 4. Hawkins, R.D., Kandel, E.R., and Bailey, C.H. (2006). Molecular mechanisms of memory storage in Aplysia. Biol. Bull. 210, 174–191. 5. Fulton, D., Kemenes, I., Andrew, R.J., and Benjamin, P.R. (2005). A single time-window for protein synthesis-dependent long-term memory formation after one-trial appetitive conditioning. Eur. J. Neurosci. 21, 1347–1358. 6. Sugai, R., Azami, S., Shiga, H., Watanabe, T., Sadamoto, H., Kobayashi, S., Hatakeyama, D., Fujito, Y., Lukowiak, K., and Ito, E. (2007). One-trial conditioned taste aversion in Lymnaea: good and poor performers in long-term memory acquisition. J. Exp. Biol. 210, 1225–1237. 7. 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. 8. 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. 9. Garcia, J., Kimeldorf, D.J., and Koelling, R.A. (1955). Conditioned aversion to saccharin resulting from exposure to gamma radiation. Science 122, 157–158. 10. McBride, S.M., Giuliani, G., Choi, C., Krause, P., Correale, D., Watson, K., Baker, G., and Siwicki, K.K. (1999). Mushroom body ablation impairs short-term memory and long-term memory of courtship conditioning in Drosophila melanogaster. Neuron 24, 967–977. 11. Mery, F., and Kawecki, T.J. (2005). A cost of long-term memory in Drosophila. Science 308, 1148. 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. Margulies, C., Tully, T., and Dubnau, J. (2005). Deconstructing memory in Drosophila. Curr. Biol. 15, R700–R713. 14. Yin, J.C.P., Wallach, J.S., Del Vecchio, M., Wilder, E.L., Zhou, H., Quinn, W.G., and Tully, T. (1994). Induction of a dominant negative CREB transgene specifically blocks long-term memory in Drosophila. Cell 79, 49–58. 15. Pascual, A., and Pre´at, T. (2001). Localization of long-term memory within the Drosophila mushroom body. Science 294, 1115–1117. 16. 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. 17. Chen, C.C., Wu, J.K., Lin, H.W., Pai, T.P., Fu, T.F., Wu, C.L., Tully, T., and Chiang, A.S. (2012). Visualizing long-term memory formation in two neurons of the Drosophila brain. Science 335, 678–685. 18. Wan, H.I., DiAntonio, A., Fetter, R.D., Bergstrom, K., Strauss, R., and Goodman, C.S. (2000). Highwire regulates synaptic growth in Drosophila. Neuron 26, 313–329. 19. Collins, C.A., Wairkar, Y.P., Johnson, S.L., and DiAntonio, A. (2006). Highwire restrains synaptic growth by attenuating a MAP kinase signal. Neuron 51, 57–69. 20. Wu, C., Wairkar, Y.P., Collins, C.A., and DiAntonio, A. (2005). Highwire function at the Drosophila neuromuscular junction: spatial, structural, and temporal requirements. J. Neurosci. 25, 9557–9566. 21. Shin, J.E., and DiAntonio, A. (2011). Highwire regulates guidance of sister axons in the Drosophila mushroom body. J. Neurosci. 31, 17689–17700. 22. Zhao, H., Zheng, X., Yuan, X., Wang, L., Wang, X., Zhong, Y., Xie, Z., and Tully, T. (2009). ben Functions with scamp during synaptic transmission and long-term memory formation in Drosophila. J. Neurosci. 29, 414–424. 23. Quinn, W.G., and Dudai, Y. (1976). Memory phases in Drosophila. Nature 262, 576–577. 24. Roman, G., Endo, K., Zong, L., and Davis, R.L. (2001). P[Switch], a system for spatial and temporal control of gene expression in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 98, 12602–12607. 25. Riemensperger, T., Vo¨ller, T., Stock, P., Buchner, E., and Fiala, A. (2005). Punishment prediction by dopaminergic neurons in Drosophila. Curr. Biol. 15, 1953–1960. 26. Tanaka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila mushroom body. J. Comp. Neurol. 508, 711–755. 27. O’Dell, K.M.C., Armstrong, J.D., Yang, M.Y., and Kaiser, K. (1995). Functional dissection of the Drosophila mushroom bodies by selective
Mushroom Body Subregion Gates Memory Consolidation 1989
28. 29.
30.
31.
32. 33.
34.
35.
36.
37. 38.
39.
40.
41.
42.
43. 44.
45.
46.
47.
48.
49. 50.
feminization of genetically defined subcompartments. Neuron 15, 55–61. Sepp, K.J., Schulte, J., and Auld, V.J. (2001). Peripheral glia direct axon guidance across the CNS/PNS transition zone. Dev. Biol. 238, 47–63. Adachi-Yamada, T., Nakamura, M., Irie, K., Tomoyasu, Y., Sano, Y., Mori, E., Goto, S., Ueno, N., Nishida, Y., and Matsumoto, K. (1999). p38 mitogen-activated protein kinase can be involved in transforming growth factor beta superfamily signal transduction in Drosophila wing morphogenesis. Mol. Cell. Biol. 19, 2322–2329. McGuire, S.E., Le, P.T., Osborn, A.J., Matsumoto, K., and Davis, R.L. (2003). Spatiotemporal rescue of memory dysfunction in Drosophila. Science 302, 1765–1768. 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. Bingol, B., and Sheng, M. (2011). Deconstruction for reconstruction: the role of proteolysis in neural plasticity and disease. Neuron 69, 22–32. Jiang, Y.H., Armstrong, D., Albrecht, U., Atkins, C.M., Noebels, J.L., Eichele, G., Sweatt, J.D., and Beaudet, A.L. (1998). Mutation of the Angelman ubiquitin ligase in mice causes increased cytoplasmic p53 and deficits of contextual learning and long-term potentiation. Neuron 21, 799–811. Pavlopoulos, E., Anezaki, M., and Skoulakis, E.M.C. (2008). Neuralized is expressed in the a/b lobes of adult Drosophila mushroom bodies and facilitates olfactory long-term memory formation. Proc. Natl. Acad. Sci. USA 105, 14674–14679. Abel, T., Martin, K.C., Bartsch, D., and Kandel, E.R. (1998). Memory suppressor genes: inhibitory constraints on the storage of long-term memory. Science 279, 338–341. Pagani, M.R., Oishi, K., Gelb, B.D., and Zhong, Y. (2009). The phosphatase SHP2 regulates the spacing effect for long-term memory induction. Cell 139, 186–198. Heisenberg, M. (2003). Mushroom body memoir: from maps to models. Nat. Rev. Neurosci. 4, 266–275. de Belle, J.S., and Heisenberg, M. (1994). Associative odor learning in Drosophila abolished by chemical ablation of mushroom bodies. Science 263, 692–695. 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. 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. Lee, T., Lee, A., and Luo, L. (1999). Development of the Drosophila mushroom bodies: sequential generation of three distinct types of neurons from a neuroblast. Development 126, 4065–4076. 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. Isabel, G., Pascual, A., and Preat, T. (2004). Exclusive consolidated memory phases in Drosophila. Science 304, 1024–1027. Wang, Y., Mamiya, A., Chiang, A.S., and Zhong, Y. (2008). Imaging of an early memory trace in the Drosophila mushroom body. J. Neurosci. 28, 4368–4376. Blum, A.L., Li, W., Cressy, M., and Dubnau, J. (2009). Short- and longterm memory in Drosophila require cAMP signaling in distinct neuron types. Curr. Biol. 19, 1341–1350. Qin, H., Cressy, M., Li, W., Coravos, J.S., Izzi, S.A., and Dubnau, J. (2012). Gamma neurons mediate dopaminergic input during aversive olfactory memory formation in Drosophila. Curr. Biol. 22, 608–614. Yu, D., Akalal, D.B.G., and Davis, R.L. (2006). Drosophila a/b mushroom body neurons form a branch-specific, long-term cellular memory trace after spaced olfactory conditioning. Neuron 52, 845–855. Akalal, D.B., Yu, D., and Davis, R.L. (2010). A late-phase, long-term memory trace forms in the g neurons of Drosophila mushroom bodies after olfactory classical conditioning. J. Neurosci. 30, 16699–16708. Davis, R.L. (2011). Traces of Drosophila memory. Neuron 70, 8–19. Aso, Y., Gru¨bel, K., Busch, S., Friedrich, A.B., Siwanowicz, I., and Tanimoto, H. (2009). The mushroom body of adult Drosophila characterized by GAL4 drivers. J. Neurogenet. 23, 156–172.
http://dx.doi.org/10.1016/j.cub.2012.08.048
Article A Permissive Role of Mushroom Body a/b Core Neurons in Long-Term Memory Consolidation in Drosophila Cheng Huang,1,4 Xingguo Zheng,1,4,5 Hong Zhao,1,4 Min Li,2 Pengzhi Wang,1,6 Zhiyong Xie,1 Lei Wang,1 and Yi Zhong1,3,* 1School of Life Sciences, Tsinghua University, Beijing 100084, PR China 2Department of Neurology, the First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, PR China 3Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724, USA
Summary Background: Memories are not created equally strong or persistent for different experiences. In Drosophila, induction of long-term memory (LTM) for aversive olfactory conditioning requires ten spaced repetitive training trials, whereas a single trial is sufficient for LTM generation in appetitive olfactory conditioning. Although, with the ease of genetic manipulation, many genes and brain structures have been related to LTM formation, it is still an important task to identify new components and reveal the mechanisms underlying LTM regulation. Results: Here we show that single-trial induction of LTM can also be achieved for aversive olfactory conditioning through inhibition of highwire (hiw)-encoded E3 ubiquitin ligase activity or activation of its targeted proteins in a cluster of neurons, localized within the a/b core region of the mushroom body. Moreover, the synaptic output of these neurons is critical within a limited posttraining interval for permitting consolidation of both aversive and appetitive LTM. Conclusions: We propose that these a/b core neurons serve as a ‘‘gate’’ to keep LTM from being formed, whereas any experience capable of ‘‘opening’’ the gate is given permit to be consolidated into LTM.
olfactory conditioning, single-session training, associating an odor with electric shock, only produces short-term, intermediate-term, and weak anesthesia-resistant memories (ARMs) that last less than 1 day, whereas spaced training (ten repetitive trials with 15 min resting intervals in between) induces protein synthesis-dependent LTM that lasts for at least 7 days [2, 12, 13]. However, single trial of appetitive olfactory conditioning, associating an odor with sucrose, is capable of inducing LTM that lasts for days [7, 8]. In the last 20 years, extensive investigations revealed molecular and neural anatomic components underlying aversive and appetitive olfactory LTM, including involvement of cyclic AMP response element-binding protein (CREB)-regulated gene expression [7, 14], the neuronal activities of mushroom body (MB) [7, 15], dorsal-paired medial neurons [16], and dorsal-anteriorlateral neurons [17]. However, the mechanisms of LTM regulation are still far from clear. The current study of the highwire (hiw) gene shed light on this problem. Our analysis of behaviorally identified LTM mutants led to the finding that the hiw gene [18] and its downstream proteins, wallenda (wnd) [19], a mitogen-activated protein kinase kinase kinase (MAPKKK), and basket (bsk) [19], a c-Jun N-terminal kinase (JNK), play an interesting role in affecting number of training trials needed for inducing aversive olfactory LTM. The hiw gene encodes an E3 ubiquitin ligase, which is important in regulation of synaptic growth at the neuromuscular junction [18, 20] and axon guidance in fly central nervous system [21]. We then identified a subset of Kenyon cells, located within the a/b core region of the MB, with intriguing features. The nature of this cluster of neurons led us to propose that they serve as a gate to keep LTM from being formed. Whether an experience is allowed to be consolidated into LTM depends on its ability in ‘‘opening’’ the gate, which is easier for experiences important for survival.
Introduction
Results
One of the common features of memory induction is that the repetitive event induces the long-term memory (LTM), which requires de novo protein synthesis and a postacquisition consolidation process, much more easily than single experience [1, 2]. The repetitive event is viewed to be capable of activating intracellular signaling pathways much stronger than single experience, thereby more likely triggering mechanisms that regulate LTM induction-relevant gene expression [3, 4]. Yet there are well-documented examples of LTM induction through single experience [5–9]. Current work suggests a gating mechanism involved in giving permission of whether an experience should be consolidated into LTM. In Drosophila, repetition is commonly required for LTM induction in many training paradigms [2, 10, 11]. In aversive
We previously constructed and screened 2,021 homozygous adult-viable P{lacW} transposants for lines with defected aversive LTM after spaced training. One identified mutant, named benP1, has a P element insertion within an intron of the hiw gene (insertion site at X:14,918,446). Within this intron also resides another small gene, ben [22] (Figure 1A). Further analysis revealed that the impaired LTM phenotype in the benP1 mutant resulted from altered expression of ben rather than hiw. On the contrary, memory was enhanced in hiw mutants.
4These
authors contributed equally to this work address: Department of Biology, Washington University, St. Louis, St. Louis, MO 63130, USA 6Present address: The School of Molecular and Cellular Biology, University of Illinois, Urbana-Champaign, Urbana, Illinois 61801, USA *Correspondence: [email protected] 5Present
Inhibition of Hiw Activity Facilitates LTM Formation As shown in Figure 1B, 24 hr memory induced by singlesession or two-session spaced aversive training was significantly increased in two hiw mutants, hiwND9 and hiwDN (see Figure 1A) [20], whereas LTM remained unaltered in hiw mutants subjected to regular spaced training. This phenotype did not result from changes in sensorimotor abilities that are essential for the task (see Table S1 available online). To determine the nature of the enhanced memory, we compared the retention curves at various time points after onesession training. hiwND9 flies had a normal memory in the first
Current Biology Vol 22 No 21 1982
Figure 1. Facilitated LTM Formation in hiw Mutants (A) The gene structure of hiw. Gray boxes indicate exons. hiwND9 has a nonsense mutation at amino acid 2,001, and hiwDN has a deletion at the N-terminal of the Hiw protein. (B) hiw mutants displayed significantly enhanced aversive 24 hr memories after one-session or two-session spaced training (p < 0.01, n = 8–10, Student’s t test) but showed normal LTM after ten-session spaced training (p > 0.79, n = 8, Student’s t test). (C) Retention curves were generated by testing conditioned odor avoidance at various time points after one-session training. hiwND9 flies displayed a normal memory performance before 8 hr (p = 0.31, 0.51, 0.17 compared to WT for 0, 3, and 8 hr, respectively, n = 6, Student’s t test) but enhanced memory thereafter (p = 0.001 for 12 hr, Student’s t test, and p = 0.02 for 24 hr, n = 6, Mann-Whitney test). (D) The enhanced 24 hr memory in hiwND9 was blocked in the CXM feeding group (CXM+) (p > 0.5, n = 8, Mann-Whitney test) but was not affected in the vehicle group (CXM2) (p < 0.01, n = 8, Mann-Whitney test). Average data are presented as mean 6 SEM; **p < 0.01, ***p < 0.001. See Table S1 for task-relevant sensorimotor abilities. See also Figure S1.
3 hr after training and a slightly but not significantly elevated memory at 8 hr, whereas they exhibited significantly enhanced memory at the time points of 12 and 24 hr (Figure 1C). The unaltered performance in the first 3 hr suggested that the observed memory enhancement is not due to a strengthened acquisition of the initial memory. To exclude the possibility of a ceiling effect, we determined the acquisition scores of hiwND9 flies at different electric shock intensities (20V, 40V, and 60V) and numbers of shock pulses (4, 8, and 12) (Figure S1A). Consistently, no statistically significant difference was found. Taken together, the results suggested that hiw mutants have normal short-term and intermediate-term memories but increased memory after 12 hr. There are two types of well-described consolidated memory forms that can last more than 1 day: LTM and ARM [2, 23]. Protein synthesis is critical for induction of LTM, but not for ARM. Our results indicated that the enhanced memory was protein synthesis dependent because feeding hiw mutant flies with a protein synthesis inhibitor, CXM (cycloheximide), eliminated the 24 hr memory enhancement (Figure 1D), whereas feeding with vehicle had no effects. Thus, LTM formation was facilitated in hiw mutants such that only a single training session was needed for its induction as compared to the ten-spaced sessions normally required in wild-type (WT) flies. Reported effects of Hiw regulating synaptic growth through E3 ligase activity [20] raised the concern that the observed memory phenotype resulted from abnormal development of the nervous system, instead of acute involvement of the Hiw physiological functions. To address this concern, we used the RU486-inducible GAL4 Gene-Switch (GS) system [24] to acutely manipulate hiw transgene expression in adult flies. Expression of transgenes of either UAS-hiw (the WT hiw+ transgene) or UAS-hiwDRING (the dominant-negative hiw transgene with the abolished ligase activity) [20] was effectively induced in response to 2 days of RU486 feeding after eclosion (Figure S2). Acutely, expression of hiw+ transgene in
the adult nervous system was sufficient to rescue the mutant phenotype (Figure 2A). Additionally, the facilitated LTM phenotype was presented (Figure 2B) in flies with acutely attenuated Hiw ligase activity through pan-neuronal expression of the hiwDRING. Expression of hiwDRING in hiwND9 mutant did not lead to further enhancement in 24 hr memory (Figure 2C), reflecting the specific dominant-negative effect in inhibiting Hiw function. Taken together, these results supported the idea that inhibition of Hiw E3 ligase activity, likely leading to the accumulation of target proteins, acutely caused facilitation of LTM formation. Hiw Functions in the MB a/b Core Region Given Hiw’s role in facilitating LTM formation, we sought to determine the expression pattern of this gene in the adult fly brain. Because no suitable antibody was available for immunocytochemistry staining in adult fly brain [20], we visualized the GAL4 expression pattern of hiwNP4132 (termed hiw-GAL4 below) with a P{GawB} element insertion in an intron of the hiw gene with the same transcription direction (insertion site at X:14,918,437) (Figure 1A). Confocal imaging of the green fluorescent protein (GFP) signal in hiw-GAL4/+;UAS-mCD8:: GFP/+ flies revealed a preferential expression in four major compartments of the adult central brain (Figure 3A; Movie S1), including fan-shaped body (FB), ellipsoid body (EB), glial cells (GCs), and the MB. The other two hiw enhancer-trap lines (NP1041 and NP2284) showed weaker GFP signals but very similar expression patterns to hiw-GAL4 (Figure S3A). A close examination of confocal images mapped expression of hiwGAL4 within the MB to the core region of a/b lobes (Figure 3B). Such images, obtained through double labeling, indicated that hiw-GAL4 labels intrinsic Kenyon cells at the calyx region, the inner part of lobes, and peduncle of MB outlined through MB247-dsRed (Figure 3B, red signal) [25]. The expression pattern of hiw-GAL4 in MB is reminiscent of the MB core neurons by comparing to other MB GAL4 (Figure S3B).
Mushroom Body Subregion Gates Memory Consolidation 1983
Figure 2. Hiw E3 Ligase Activity Acutely Regulates LTM Formation (A) Facilitated LTM in hiwND9 can be rescued by acutely inducing hiw+ expression. In the vehicle group (RU4862), hiwND9/Y;UAS-hiw/+;elav-GS flies displayed the enhanced 24 hr memory after one-session training (p < 0.01, n = 8–10, ANOVA), which was restored to normal levels in the RU486 feeding group (RU486+) (p > 0.43, n = 8–10, ANOVA). (B) Acute expression of E3 ligase activitydead HiwDRING protein in WT background facilitated LTM formation (p < 0.01, n = 8, ANOVA). (C) Acute expression of HiwDRING protein in hiw mutant adults cannot further enhance 24 hr memory (p > 0.9, n = 8, ANOVA). Average data are presented as mean 6 SEM; **p < 0.01. n.s., nonsignificance. See also Figure S2 for the efficiency of acute induction of the elav-GS system.
Furthermore, the MB expression could be eliminated by introducing additional expression of MB-GAL80 [16] targeted to the MB, while leaving expression in all other compartments unchanged (see the effect of removing hiw-GAL4-induced GFP expression specifically within the MB by comparing the two columns in Figure 3B; Figure S3C). Subsequent behavioral assays demonstrated that Hiw regulates LTM formation in MB, specifically within the core region of the a/b lobes. Inhibition of Hiw functions in hiw-GAL4 neurons through expression of UAS-hiwDRING led to the enhanced 24 hr memory after one-session training (Figure 3C), which reliably phenocopied the hiw mutants. Consistently, the learning was unchanged (Figure S3D), and the enhanced 24 hr memory was also protein synthesis dependent (Figure S3E). However, suppression of the MB expression in hiw-GAL4 failed to induce the LTM facilitation. The hiwDRING-induced memory enhancement could also be observed in other GAL4 drivers that include the a/b core region (NP7175, NP6024, and c739) [26, 27], but not in the GAL4 driver that labels a/b posterior and surface neurons (NP5286) [26] or GC (repoGAL4) [28] (Figure 3C). Similar enhancement was also achieved by expressing UAS-hiwRNAi in hiw-GAL4, NP7175, and NP6024 (Figure S3F). The morphological structure of MB was not obviously affected by hiwDRING expression (Figure S4G). In contrast, overexpression of hiw+ transgene in hiw-GAL4 or NP7175 severely impaired LTM formation induced by spaced training (Figure 3D), whereas presence of MB-GAL80 eliminated the LTM impairment. Meanwhile, the learning was unchanged in hiw-GAL4/+;UAS-hiw/+ flies (Figure S3H). Task-relevant sensorimotor responses were not significantly altered in most lines, except NP7175/+;;UAShiwDRING/+ and c739/+;UAS-hiwDRING/+, which performed significantly worse than the UAS-hiwDRING control in odor acuity but indistinguishably from their GAL4 controls (Table S2). Taken together, we showed that inhibition of Hiw activity within a/b core neurons could facilitate LTM formation, whereas excessive Hiw blocks LTM formation. Involvement of Hiw Downstream Targets in the Facilitation of LTM Formation To verify the functional role of Hiw and gain more mechanism insights, we tested the genetic regulation of Hiw downstream signals. Wallenda (Wnd), a MAPKKK homologous to the vertebrate dual-leucine zipper-bearing kinases (DLKs and LZKs, respectively), plays a crucial role in Hiw-mediated synaptic overgrowth [19]. In the established model for Hiw function (Figure 4A), Hiw constrains the protein level of Wnd, which is thought to function as an upstream regulator of the JNK/ MAPK, encoded by bsk in Drosophila [19].
As shown in Figure 4B, reduction of Wnd and Bsk by expressing wnd RNAi or Bsk dominant-negative protein (BskDN) [29] blocked the enhancement of LTM formation induced by expressing UAS-hiwDRING in hiw-GAL4 while leaving learning unaffected (Figure S4A). To further explore whether the accumulation of Hiw downstream protein is sufficient to facilitate LTM formation, we used temperature-sensitive GAL80 system [30] to acutely overexpress Wnd protein. After 4 days’ induction, overexpression of Wnd in hiw-GAL4 and NP7175 both resulted in enhanced LTM formation, which mimicked the phenotype in hiwDRING group (Figure 4C) or hiw RNAi group after induction (Figure S4B). The data also showed that the kinase activity of Wnd was critical for facilitating LTM because expression of kinase-dead Wnd protein failed to induce such phenotype (Figure 4C). In all groups, learning was unaffected (Figure S4C). Western blotting revealed the efficiency of the overexpression of Wnd and, more importantly, showed that expressing hiwDRING did increase the Wnd protein level (Figures 4D and 4E). Moreover, after acute upregulation of Wnd, immunostaining with anti-FASII revealed that the morphology of MB was unaffected (Figure S4D). The results above suggested the involvement of Wnd/JNK pathway in Hiw-regulated LTM formation. MB a/b Core Neurons Permit Aversive LTM Consolidation The facilitated LTM phenotype achieved through manipulation of Hiw activity within MB a/b core neurons posed the immediate question of whether such manipulation acts to affect acquisition, consolidation, or retrieval. To address this question, we used the temperature-sensitive shibire (shi) mutant transgene [31], UAS-shits1, to manipulate synaptic output reversibly. The shits1 gene encodes a mutated Dynamin that allows normal neurotransmitter release at permissive temperatures (below 23 C) but blocks synaptic transmission at restrictive temperatures (above 30 C) [31]. At the permissive temperature (18 C), expression of UASshits1 in all drivers tested had no effects on LTM induced by regular ten-session spaced training (Figure 5A). LTM was also not affected by blockade of synaptic transmission in either hiw-GAL4 or NP7175 during acquisition (Figure 5B) or during retrieval (Figure 5C). However, LTM was impaired when synaptic transmission was blocked during the whole consolidation period in hiw-GAL4 group (Figure S5A) or in the interval 6 hr after completion of spaced training for a period of 12 hr in hiw-GAL4, c739, NP6024, or NP7175 groups (Figure 5D). This memory impairment, observed in the drivers containing the a/b core region, was not shown in repo-GAL4, NP5286, and hiw-GAL4;MB-GAL80 group (Figure 5D).
Current Biology Vol 22 No 21 1984
Figure 3. Hiw Regulates LTM Formation in MB Core Neurons (A) Confocal imaging of GFP expression in the whole-mount adult central brain and thoracic ganglion of hiw-GAL4-driven mCD8::GFP (green) flies revealed preferential expression in four compartments: FB F5 layer neurons, EB, and a subset of the MBs and GCs. Scale bars represent 50 mm. (B) Expression pattern of hiw-GAL4, hiw-GAL4;MB-GAL80 in MB lobes, peduncle, and Kenyon cell region. The whole MB structure was indicated by MB247dsRed-labeled region (red signal), and the GAL4 expression pattern was displayed by mCD8::GFP-labeled region (green signal). Top view shows the expression in the Kenyon cell region. Middle view is the expression in the peduncle. Bottom view demonstrates the expression in MB lobes. In hiw-GAL4 and hiw-GAL4;MB-GAL80 images, the cell bodies of the FB are outlined by white lines. In the hiw-GAL4;MB-GAL80 peduncle and lobe region, white dashed lines outline the a/b core subdivision; yellow arrow indicates some extrinsic innervations onto MB lobes. Scale bars represent 20 mm. (C) Attenuating Hiw activity in hiw-GAL4-labeled neurons was sufficient to facilitate LTM formation. UAS-hiwDRING flies were crossed with indicated GAL4 drivers. Facilitated LTM formation can be seen in hiw-GAL4, NP7175, NP6024, and c739 groups (p < 0.05, n = 10–12, ANOVA), but not in hiw-GAL4; MB-GAL80, NP7175;MB-GAL80, or NP5286 (p > 0.58, n = 10–12, ANOVA and Kruskal-Wallis ANOVA). (D) Overexpression of the WT Hiw protein impaired LTM. hiw-GAL4/+;UAS-hiw/+ or NP7175/+;UAS-hiw/+ flies displayed significantly impaired LTM after ten-session spaced training compared with control flies (p < 0.01, n = 8, ANOVA), where LTM was normal in hiw-GAL4/+;MB-GAL80/UAS-hiw or NP7175/+;MB-GAL80/UAS-hiw flies (p > 0.9, n = 8, ANOVA and Kruskal-Wallis ANOVA). For the X chromosome-located hiw-GAL4, hiw-GAL4;MB-GAL80, NP7175, NP7175;MB-GAL80, and NP6024, only female results are shown. Average data are presented as mean 6 SEM; *p < 0.05, **p < 0.01, ***p < 0.001. See Table S2 for task-relevant sensorimotor abilities. See also Figure S3.
The memory impairment was specific for LTM because blocking synaptic transmission in a/b core neurons within the whole consolidation period or indicated 12 hr time window had no effects on ARM induced by ten-session massed training in tested flies (Figure S5B; Figure 5E), and blocking synaptic output from hiw-GAL4 and NP7175 neurons did not affect learning (Figure S5C). In most cases, task-relevant sensorimotor responses were not significantly altered, except NP7175/+;;UAS-shits1/+ and c739/+;UAS-shits1/+, which performed significantly worse than the UAS-shits1 control in odor acuity but indistinguishably from their GAL4 controls at the permissive temperature (Tables S3 and S4). As shown in Figure 5F, blocking the transmission from hiwGAL4 during consolidation period could also eliminate the
hiwDRING-induced LTM facilitation. Moreover, the similar requirement of core neurons in LTM consolidation was also observed by using another odor pair (OCT/BEN) as shown in Figures S5D–S5F. Thus, synaptic output from the identified a/b core neurons plays a critical role during the period of LTM consolidation in a very specific manner that permits aversive olfactory LTM that is induced by either single-session training or regular spaced repetitive training. MB a/b Core Neurons Also Permit Appetitive LTM Consolidation The suggested permissive function of identified MB a/b core neurons made us curious about its general role in LTM induction. In particular, appetitive LTM is induced with only
Mushroom Body Subregion Gates Memory Consolidation 1985
Figure 4. Involvement of Hiw Downstream Signals in the LTM Facilitation (A) The model of Hiw function. Hiw negatively regulates downstream Wnd/DLK, which is an activator of JNK/MAPK. (B) Reducing Wnd and Bsk suppressed the Hiw-mediated LTM facilitation. Expression of hiwDRING in hiw-GAL4 promoted the LTM formation, whereas coexpression of wnd RNAi or BskDN with hiwDRING failed to induce such LTM facilitation (p < 0.01, n = 8–12, ANOVA). (C) LTM formation was significantly enhanced by acutely overexpressing WT Wnd (UAS-wnd) protein or Hiw dominant-negative protein (UAShiwDRING) in MB core region, labeled by hiwGAL4 and NP7175 (p < 0.05, n = 8–12, ANOVA), whereas expression of kinase-dead Wnd protein (UAS-wndKD) had no such effect (p > 0.54, n = 8, ANOVA and Kruskal-Wallis ANOVA). (D) Western blots showed acutely induced expression of hiwDRING and Wnd in hiw-GAL4elevated Wnd protein level (p < 0.05, n = 6, ANOVA). Left panel is representative of western blots for uninduced and induced group. Right panel shows quantified results. (E) Western blots showed acutely induced expression of hiwDRING and Wnd in hiw-GAL4elevated Wnd protein level (p < 0.05, n = 6, ANOVA). Left panel is representative of western blots for uninduced and induced group. Right panel shows quantified results. Average data are presented as mean 6 SEM; *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S4.
single-session training, rather than ten-session spaced training as in induction of aversive LTM. In the appetitive-learning task, we measured LTM 48 hr later after training. Because flies cannot form or retrieve appetitive memory unless they are hungry [7], flies were food deprived for 36 hr before training, reintroduced to food for 24 hr after completing training (to avoid starvation and heat shock caused death), and then food deprived for 24 hr before testing. Control experiments showed that at the permissive temperature (23 C), one session-induced appetitive LTM was normal in all genotypes tested (Figure 6A). We then demonstrated that disrupting synaptic output from hiw-GAL4-labeled neurons or from a/b lobe core regions (c739 and NP6024) at the restrictive temperature (30 C) in the posttraining period impaired appetitive LTM (Figure 6B), whereas the task-relevant sensorimotor responses were not significantly altered (Tables S5 and S6). However, appetitive LTM was normal in flies when synaptic transmission was blocked in those hiw-GAL4-labeled brain regions outside of the MB (hiw-GAL4;MB-GAL80 group), in the a/b posterior and surface region (NP5286 group), or in GCs (repo-GAL4 group) (Figure 6B). It is interesting that blockade of output from the NP7175-labeled inner part of core neurons alone did not impair appetitive LTM consolidation. This observation can either result from a further functional division within the core region or a complementary effect from NP7175-unlabeled core neurons. We further tested the idea that hiw-GAL4-labeled MB a/b core neurons are specifically engaged in consolidation by
introducing two time windows during which synaptic transmission was blocked. Appetitive LTM was normal when the blockade of synaptic transmission (at 30 C) occurred 12 hr after completing training (Figure 6C) but was defective if synaptic transmission was disrupted immediately after training for 12 hr (Figure 6D). This result showed that disruption of synaptic transmission from hiw-GAL4labeled MB a/b core neurons affected LTM only during the phase of consolidation and thus supported the idea that hiw-GAL4-labeled MB a/b core neurons play a permissive role in gating consolidation of not only aversive but also appetitive LTM. Discussion The current study began with the finding that 24 hr memory resulting from single session was enhanced in two hiw mutant alleles. This enhanced memory component was identified as facilitated LTM, given that it was sensitive to protein synthesis inhibition. The behavioral effect of hiwDRING and the presence of a hiw-GAL4 line allowed us to map the neural circuitry to a cluster of MB a/b core neurons, within which Hiw and its downstream targets regulate LTM. Furthermore, we showed that the MB a/b core neurons were involved in the consolidation of both aversive and appetitive LTM. In conclusion, the observations that the MB a/b core neurons were capable of both facilitating and limiting LTM suggested a working model in which the ability to form LTM is gated through these neurons. The significance of results presented is further elaborated below.
Current Biology Vol 22 No 21 1986
Figure 5. MB a/b Core Neurons Permit Aversive LTM Consolidation Female flies of indicated GAL4 drivers were crossed to WT male flies (+/+) or to UAS-shits1 male flies, and all the progenies were raised at 18 C to minimize any potential ‘‘leaky’’ effect of UAS-shits1 on development. Temperature-shift protocols are shown above each graph. For the X chromosome-located hiwGAL4, hiw-GAL4;MB-GAL80, NP6024, and NP7175, only female results are shown. (A) The permissive temperature (18 C) did not affect LTM after ten-session spaced training of all genotypes used in this study (p > 0.2, n = 8–10, ANOVA and Kruskal-Wallis ANOVA). See also Table S3 for task-relevant sensorimotor abilities. (B) Blocking synaptic output from hiw-GAL4 and NP7175 neurons during the training period did not affect LTM performance (p > 0.9, n = 8, ANOVA). (C) Blocking synaptic output from hiw-GAL4 and NP7175 neurons during the testing period did not affect LTM performance (p > 0.9, n = 8, ANOVA). (D) Blocking synaptic output from hiw-GAL4, c739, NP6024, and NP7175 neurons during the LTM consolidation period (6–18 hr after the end of training) disrupted LTM performance (p < 0.001, n = 8–10, ANOVA and Kruskal-Wallis ANOVA), whereas LTM was not affected in hiw-GAL4;MB-GAL80, NP5286, and repo-GAL4 flies. See also Table S4 for task-relevant sensorimotor abilities. (E) Blocking synaptic output during the memory consolidation period (6–18 hr after the end of training) did not impair ARM after ten-session massed training (p > 0.48, n = 8, ANOVA and Kruskal-Wallis ANOVA). (F) Blocking synaptic output from hiw-GAL4 during the consolidation period eliminated the facilitated LTM caused by hiwDRING expression (p < 0.01, n = 8, ANOVA). Average data are presented as mean 6 SEM; **p < 0.01, ***p < 0.001. See also Figure S5.
Involvement of Hiw and Its Downstream DLK/JNK Pathway in LTM Regulation Not only synthesis but also degradation of proteins plays a critical role in the remodeling of synapses, learning, and memory [32]. Altered memory formation in ubiquitin ligase mutants has
been reported in mice and Drosophila [33, 34]. Here, we reported that Hiw, an evolutionarily conserved E3 ubiquitin ligase, negatively regulated LTM formation through restraining its downstream target Wnd [19]. Our results indicated Hiw function as an inhibitory constraint, as the memory suppressor
Mushroom Body Subregion Gates Memory Consolidation 1987
gene [35], on LTM formation. Removal of this suppressor or direct activation of its downstream signals could lead to the facilitated LTM induction without the repetitive training that is normally required. So far, the physiological consequence of Hiw or Wnd in core neurons still remains an open question. Because of the extensively shared components with hiw’s function in synaptic growth and transmission [18, 19], the attenuated Hiw activity may elevate Wnd level and then lead to the excessive synaptogenesis or abnormal synaptic activity in core neurons. It will also be interesting to test whether the hiw-mediated LTM facilitation shares some common molecular components with the corkscrew-regulated spacing effect in LTM induction [36], in which the MB a/b lobes also play an important role. In a recent report, Hiw regulates the axon guidance in MB [21]. We also observed the morphological defect in MB a/b lobes in a portion of hiw mutant flies (Figure S1B). It is striking that hiw mutants with MB defect can form LTM even more efficiently, given the observation of LTM impairment in another MB structural mutant, ala [15]. However, comparing to the total loss of vertical lobes (including a and a0 ) in ala mutant, most hiw mutants had the abnormal thickness of a/b lobes caused by the unequal distribution of the MB axonal projections between the a and b lobes, and about 39% of hiwDN mutant had the shortened a lobe [21]. One of the possible explanations is that the remaining function of a/b lobe in hiw mutants is sufficient to support the LTM. Moreover, expression of Hiw dominant-negative protein or acutely increasing Wnd protein level in MB was sufficient to promote LTM but did not give rise to any observable gross morphological change in MB (Figures S3G and S4D). Thus, we suggest that Hiw mediates memory phenotype through a different mechanism from the one that led to the structure change in the MB.
Figure 6. MB a/b Core Neurons Permit Appetitive LTM Consolidation Female flies of indicated GAL4 drivers were crossed to WT male flies (+/+) or to UAS-shits1 male flies, and all the progenies were raised at 18 C to minimize any potential ‘‘leaky’’ effect of UAS-shits1 on development. Temperature-shift and food-deprivation protocols are shown above each graph. For the X chromosome-located hiw-GAL4, hiw-GAL4;MB-GAL80, NP6024, and NP7175, only female results are shown. (A) The permissive temperature (23 C) did not affect appetitive LTM after one-session appetitive training (AP training) for all genotypes used in this study (p > 0.22, n = 8-10, ANOVA and Kruskal-Wallis ANOVA). See also Table S5 for task-relevant sensorimotor abilities. (B) Disrupting synaptic output from hiw-GAL4, NP6024, or c739 neurons during the LTM consolidation period could impair appetitive LTM performance (p < 0.01, n = 8–10, ANOVA), but LTM was not affected in hiw-GAL4;MB-GAL80, NP5286, repo-GAL4 flies. See also Table S6 for task-relevant sensorimotor abilities. (C) Disrupting synaptic output from hiw-GAL4 neurons during 12–24 hr after training did not affect appetitive LTM (p > 0.7, n = 8–10, ANOVA). (D) Disrupting synaptic output from hiw-GAL4 neurons during the first 12 hr after training was sufficient to impair appetitive LTM. (p < 0.01, n = 8–9, ANOVA). Average data are presented as mean 6 SEM; **p < 0.01, ***p < 0.001.
MB a/b Core Region in Giving the Permission to Consolidation of LTM The involvement of MB in the hiw-mediated LTM facilitation led us to further probe the function of this structure in LTM regulation. It has been well documented that the MB, a bilateral brain structure that consists of approximately 2,500 neurons in each hemisphere [37], plays the central role in olfactory memories, both aversive and appetitive [7, 16, 38–40]. Intrinsic MB neurons are organized into physically distinct a, b, a0 , b0 , and g lobes [41, 42]. All three lobes exhibited different functions in memory processing, such that the output of a/b lobes was required for retrieval of memory [7, 16, 43], a0 /b0 lobes were transiently required to stabilize memory [16] or to retrieve immediate memory [44], and g lobe mediated rutabagadependent mechanism and dopaminergic signal to support short-term memory (STM) and LTM formation [45, 46]. Moreover, memory traces mapped to different lobes exhibited different temporal features [44, 47–49]. Through gene expression patterns and enhancer trap lines, each lobe of MB can be classified into more specific subgroups such as the posterior, surface, and core regions in the a/b lobes [26, 50]. The current work shows that a/b core neurons play a distinct role in LTM induction. The synaptic outputs of these neurons are critical during consolidation of LTM for both aversive and appetitive conditioning, but these neurons are not involved in LTM cellular consolidation per se because a landmark of LTM cellular consolidation, CREB-mediated protein synthesis, occurs in non-MB neurons [17]. Thus, one of the roles for this cluster of neurons can be viewed as simply providing connections to channel learning
Current Biology Vol 22 No 21 1988
information to the downstream neurons in which LTM is formed. However, the remarkable feature of enabling singletrial induction of aversive LTM through targeted genetic manipulation of this cluster of neurons suggests that they play a unique permissive role in determining whether an experience should be consolidated. Gating LTM This newly identified function for permitting an experience to be consolidated leads us to propose a gating theory. This theory proposes that the a/b core neurons serve as a ‘‘gate,’’ and activation of this gating mechanism functions as a checkpoint that keeps LTM from being formed for general experiences, whereas only specific experiences, capable of ‘‘opening’’ this gate, can and are bound to trigger LTM consolidation and to form LTM ultimately. There is little survival advantage in committing never-to-be-repeated episodes to memory, particularly because the very act of LTM formation may be deleterious to the fly [11]. In contrast, repetitively occurring experience, such as spaced repetitive aversive conditioning, and events critical for survival, such as finding food or single-trial appetitive conditioning, would be able to ‘‘open’’ the gate, and therefore, LTM is formed for such experiences. Experimental Procedures Behavioral Assays Aversive and appetitive Pavlovian olfactory-conditioning procedures were performed as described previously [2, 7, 12]. For details of behavioral procedures, see Supplemental Experimental Procedures. Statistical Analysis Shapiro-Wilk test was used for the normality test. Normally distributed data were analyzed by two-tailed Student’s t test or one-way ANOVA following Bonferroni test (Origin version 8; OriginLab). For the data not normally distributed, Mann-Whitney test and Kruskal-Wallis ANOVA were used. Statistical results are presented as mean 6 SEM. Asterisks indicate critical values (*p < 0.05, **p < 0.01 and ***p < 0.001). ‘‘n.s.’’ indicates no significant difference (p > 0.05). Supplemental Information Supplemental Information includes five figures, six tables, Supplemental Experimental Procedures, and one movie and can be found with this article online at http://dx.doi.org/10.1016/j.cub.2012.08.048. Acknowledgments We thank Aaron DiAntonio, Catherine A. Collins, Chunlai Wu, Ronald L. Davis, Scott Waddell, Kei Ito, and Andre´ Fiala for providing flies and reagents. We are grateful for helpful comments on the manuscript from Jennifer Beshel, Yichun Shuai, and Lisha Shao. This work was supported by grants from the National Basic Research Project (973 program) of the Ministry of Science and Technology of China (2006CB500806 and 2009CB941301) and the Tsinghua University Initiative Scientific Research Program (20111080956, all to Y.Z.). Received: May 25, 2012 Revised: July 18, 2012 Accepted: August 21, 2012 Published online: October 11, 2012 References 1. Carew, T.J., Pinsker, H.M., and Kandel, E.R. (1972). Long-term habituation of a defensive withdrawal reflex in aplysia. Science 175, 451–454. 2. Tully, T., Preat, T., Boynton, S.C., and Del Vecchio, M. (1994). Genetic dissection of consolidated memory in Drosophila. Cell 79, 35–47.
3. Kandel, E.R. (2001). The molecular biology of memory storage: a dialogue between genes and synapses. Science 294, 1030–1038. 4. Hawkins, R.D., Kandel, E.R., and Bailey, C.H. (2006). Molecular mechanisms of memory storage in Aplysia. Biol. Bull. 210, 174–191. 5. Fulton, D., Kemenes, I., Andrew, R.J., and Benjamin, P.R. (2005). A single time-window for protein synthesis-dependent long-term memory formation after one-trial appetitive conditioning. Eur. J. Neurosci. 21, 1347–1358. 6. Sugai, R., Azami, S., Shiga, H., Watanabe, T., Sadamoto, H., Kobayashi, S., Hatakeyama, D., Fujito, Y., Lukowiak, K., and Ito, E. (2007). One-trial conditioned taste aversion in Lymnaea: good and poor performers in long-term memory acquisition. J. Exp. Biol. 210, 1225–1237. 7. 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. 8. 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. 9. Garcia, J., Kimeldorf, D.J., and Koelling, R.A. (1955). Conditioned aversion to saccharin resulting from exposure to gamma radiation. Science 122, 157–158. 10. McBride, S.M., Giuliani, G., Choi, C., Krause, P., Correale, D., Watson, K., Baker, G., and Siwicki, K.K. (1999). Mushroom body ablation impairs short-term memory and long-term memory of courtship conditioning in Drosophila melanogaster. Neuron 24, 967–977. 11. Mery, F., and Kawecki, T.J. (2005). A cost of long-term memory in Drosophila. Science 308, 1148. 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. Margulies, C., Tully, T., and Dubnau, J. (2005). Deconstructing memory in Drosophila. Curr. Biol. 15, R700–R713. 14. Yin, J.C.P., Wallach, J.S., Del Vecchio, M., Wilder, E.L., Zhou, H., Quinn, W.G., and Tully, T. (1994). Induction of a dominant negative CREB transgene specifically blocks long-term memory in Drosophila. Cell 79, 49–58. 15. Pascual, A., and Pre´at, T. (2001). Localization of long-term memory within the Drosophila mushroom body. Science 294, 1115–1117. 16. 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. 17. Chen, C.C., Wu, J.K., Lin, H.W., Pai, T.P., Fu, T.F., Wu, C.L., Tully, T., and Chiang, A.S. (2012). Visualizing long-term memory formation in two neurons of the Drosophila brain. Science 335, 678–685. 18. Wan, H.I., DiAntonio, A., Fetter, R.D., Bergstrom, K., Strauss, R., and Goodman, C.S. (2000). Highwire regulates synaptic growth in Drosophila. Neuron 26, 313–329. 19. Collins, C.A., Wairkar, Y.P., Johnson, S.L., and DiAntonio, A. (2006). Highwire restrains synaptic growth by attenuating a MAP kinase signal. Neuron 51, 57–69. 20. Wu, C., Wairkar, Y.P., Collins, C.A., and DiAntonio, A. (2005). Highwire function at the Drosophila neuromuscular junction: spatial, structural, and temporal requirements. J. Neurosci. 25, 9557–9566. 21. Shin, J.E., and DiAntonio, A. (2011). Highwire regulates guidance of sister axons in the Drosophila mushroom body. J. Neurosci. 31, 17689–17700. 22. Zhao, H., Zheng, X., Yuan, X., Wang, L., Wang, X., Zhong, Y., Xie, Z., and Tully, T. (2009). ben Functions with scamp during synaptic transmission and long-term memory formation in Drosophila. J. Neurosci. 29, 414–424. 23. Quinn, W.G., and Dudai, Y. (1976). Memory phases in Drosophila. Nature 262, 576–577. 24. Roman, G., Endo, K., Zong, L., and Davis, R.L. (2001). P[Switch], a system for spatial and temporal control of gene expression in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 98, 12602–12607. 25. Riemensperger, T., Vo¨ller, T., Stock, P., Buchner, E., and Fiala, A. (2005). Punishment prediction by dopaminergic neurons in Drosophila. Curr. Biol. 15, 1953–1960. 26. Tanaka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila mushroom body. J. Comp. Neurol. 508, 711–755. 27. O’Dell, K.M.C., Armstrong, J.D., Yang, M.Y., and Kaiser, K. (1995). Functional dissection of the Drosophila mushroom bodies by selective
Mushroom Body Subregion Gates Memory Consolidation 1989
28. 29.
30.
31.
32. 33.
34.
35.
36.
37. 38.
39.
40.
41.
42.
43. 44.
45.
46.
47.
48.
49. 50.
feminization of genetically defined subcompartments. Neuron 15, 55–61. Sepp, K.J., Schulte, J., and Auld, V.J. (2001). Peripheral glia direct axon guidance across the CNS/PNS transition zone. Dev. Biol. 238, 47–63. Adachi-Yamada, T., Nakamura, M., Irie, K., Tomoyasu, Y., Sano, Y., Mori, E., Goto, S., Ueno, N., Nishida, Y., and Matsumoto, K. (1999). p38 mitogen-activated protein kinase can be involved in transforming growth factor beta superfamily signal transduction in Drosophila wing morphogenesis. Mol. Cell. Biol. 19, 2322–2329. McGuire, S.E., Le, P.T., Osborn, A.J., Matsumoto, K., and Davis, R.L. (2003). Spatiotemporal rescue of memory dysfunction in Drosophila. Science 302, 1765–1768. 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. Bingol, B., and Sheng, M. (2011). Deconstruction for reconstruction: the role of proteolysis in neural plasticity and disease. Neuron 69, 22–32. Jiang, Y.H., Armstrong, D., Albrecht, U., Atkins, C.M., Noebels, J.L., Eichele, G., Sweatt, J.D., and Beaudet, A.L. (1998). Mutation of the Angelman ubiquitin ligase in mice causes increased cytoplasmic p53 and deficits of contextual learning and long-term potentiation. Neuron 21, 799–811. Pavlopoulos, E., Anezaki, M., and Skoulakis, E.M.C. (2008). Neuralized is expressed in the a/b lobes of adult Drosophila mushroom bodies and facilitates olfactory long-term memory formation. Proc. Natl. Acad. Sci. USA 105, 14674–14679. Abel, T., Martin, K.C., Bartsch, D., and Kandel, E.R. (1998). Memory suppressor genes: inhibitory constraints on the storage of long-term memory. Science 279, 338–341. Pagani, M.R., Oishi, K., Gelb, B.D., and Zhong, Y. (2009). The phosphatase SHP2 regulates the spacing effect for long-term memory induction. Cell 139, 186–198. Heisenberg, M. (2003). Mushroom body memoir: from maps to models. Nat. Rev. Neurosci. 4, 266–275. de Belle, J.S., and Heisenberg, M. (1994). Associative odor learning in Drosophila abolished by chemical ablation of mushroom bodies. Science 263, 692–695. 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. 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. Lee, T., Lee, A., and Luo, L. (1999). Development of the Drosophila mushroom bodies: sequential generation of three distinct types of neurons from a neuroblast. Development 126, 4065–4076. 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. Isabel, G., Pascual, A., and Preat, T. (2004). Exclusive consolidated memory phases in Drosophila. Science 304, 1024–1027. Wang, Y., Mamiya, A., Chiang, A.S., and Zhong, Y. (2008). Imaging of an early memory trace in the Drosophila mushroom body. J. Neurosci. 28, 4368–4376. Blum, A.L., Li, W., Cressy, M., and Dubnau, J. (2009). Short- and longterm memory in Drosophila require cAMP signaling in distinct neuron types. Curr. Biol. 19, 1341–1350. Qin, H., Cressy, M., Li, W., Coravos, J.S., Izzi, S.A., and Dubnau, J. (2012). Gamma neurons mediate dopaminergic input during aversive olfactory memory formation in Drosophila. Curr. Biol. 22, 608–614. Yu, D., Akalal, D.B.G., and Davis, R.L. (2006). Drosophila a/b mushroom body neurons form a branch-specific, long-term cellular memory trace after spaced olfactory conditioning. Neuron 52, 845–855. Akalal, D.B., Yu, D., and Davis, R.L. (2010). A late-phase, long-term memory trace forms in the g neurons of Drosophila mushroom bodies after olfactory classical conditioning. J. Neurosci. 30, 16699–16708. Davis, R.L. (2011). Traces of Drosophila memory. Neuron 70, 8–19. Aso, Y., Gru¨bel, K., Busch, S., Friedrich, A.B., Siwanowicz, I., and Tanimoto, H. (2009). The mushroom body of adult Drosophila characterized by GAL4 drivers. J. Neurogenet. 23, 156–172.