One of the main obstacles to understanding how the brain works is that we know relatively little about how its nerve cells or neurons are connected to one another. These connections make up the brain’s wiring diagram. Current methods for revealing this wiring all have limitations. The most popular method – serial electron microscopy – can reveal the connections in a small region of the brain in great detail, but it cannot show connections between neurons that are far apart.

Huang et al. have now created a genetic system for visualizing these connections. For neurons to communicate, one neuron must produce a signal called a ligand. This ligand can then bind to and activate its partner neuron. Huang et al. modified the DNA of neurons so that every time those cells produced a specific ligand, they also produced a red fluorescent protein. Similar modifications ensured that every time the ligand activated a partner neuron, the activated neuron produced a green fluorescent protein. Viewing the red and green neurons under a microscope enabled Huang et al. to see which cells were communicating with which others.

While these experiments took place in fruit flies, the same approach should also work in other laboratory animals, including fish, mice and rats. Once we know the wiring diagram of the brain, the next step is to investigate the role of the various connections. To understand how a computer works, for example, we might change the connections between its circuit components and look at how this affects the computer’s output. With this new method, we can change how neurons communicate with one another in the brain, and then look at the effects on behavior. This should provide insights into the workings of the human brain, and clues to what goes wrong in disorders like schizophrenia and autism.

Comprehensively mapping the connectivity of diverse neural circuits across many brain regions and many organisms is a major goal of modern neuroscience (Denk et al., 2012; Bargmann and Marder, 2013; Swanson and Lichtman, 2016). In addition, recent research indicates that aberrant neuronal wiring may be the cause of several neurodevelopmental disorders (Peça and Feng, 2012; Rubinov and Bullmore, 2013; Mevel and Fransson, 2016), further emphasizing the importance of identifying the wiring diagrams of brain circuits. To address this issue, several new approaches have been developed. Each of these methods has strengths and limitations, which are discussed in detail in a recent review (Lee et al., 2017).

In a previous work, we demonstrated that a strategy based on ligand-induced intramembrane proteolysis can be used to monitor cell-cell interactions in the nervous system of transgenic Drosophila (Huang et al., 2016). In this strategy, cells expressing an artificial ligand (‘donor’ cells) bind to and activate an artificial receptor on their interacting partners (‘receiver’ cells) via interactions across the intercellular space. We previously used this method to investigate the interactions between neurons and glial cells in the Drosophila central nervous system (Huang et al., 2016). However, in its original implementation, the system was not useful to trace neuronal circuits because it revealed all forms of cell-cell contact, including non-synaptic contacts.

To allow this strategy to trace neuronal circuits it is necessary to ensure that the interaction of ligand and receptor occurs exclusively across synapses. To prevent activation of the system between neurons that had non-synaptic cell-cell contacts, we engineered the ligand so that it would be selectively located at the presynaptic cleft (Südhof, 2012; Van Vactor and Sigrist, 2017). To validate that the optimized ligands revealed neurons specifically connected by synapses, we tested the system on the Drosophila antennal lobe, an olfactory center with well-understood connectivity established by light and electron microscopy, and electrophysiological methods (Grabe et al., 2016; Stocker et al., 1990; Wilson, 2013; Rybak et al., 2016). We observed that adding domains from the synaptic proteins synaptobrevin (nSyb) or syndecan (sdc) was sufficient to achieve reliable synaptic tracing in the antennal lobe. In addition, we used TRACT to investigate the connectivity of the Drosophila circadian system, a compact brain circuit with a small number of neurons (Dubowy and Sehgal, 2017; Nitabach and Taghert, 2008), and discovered new candidate postsynaptic targets for the PDF neurons of the circadian system in the central brain, some of which that express the circadian-related gene, per. These results demonstrate that the TRACT method will be a useful addition to the arsenal of methods available to investigate the connectivity of brain circuits.

In our previously published work, we described an artificial ligand-receptor system based on intramembrane proteolysis to investigate connections between cells (Huang et al., 2016). In short, upon ligand-receptor interaction at sites of cell-cell contact the transmembrane domain of an engineered receptor is cleaved by intramembrane proteolysis and releases a protein fragment that regulates transcription in the interacting partners (Figure 1a, Gordon et al., 2015; Morsut et al., 2016; Huang et al., 2016). In its original implementation, we used a ligand called CD19mch that contains the extracellular and transmembrane domains (ECD and TMD) of the mouse lymphocyte antigen CD19 (Fujimoto et al., 1998) fused to the red fluorescent protein mCherry in its intracellular domain (ICD). The original artificial receptor is called SNTG4 and contains the following domains (in N-to-C terminal order): (i) the ECD from a single chain antibody (ID3) that recognizes mouse CD19 (Kochenderfer et al., 2009), (ii) the Notch regulatory region (NRR) and the TMD from Drosophila Notch (spanning from EGF repeat 36 until the ICD) (Kovall et al., 2017), and (iii) esn, a simplified version of the yeast transcriptional activator Gal4 (Figure 1a) (Sprinzak et al., 2010). In this configuration, expression of the ligand and receptor in neurons leads to the distribution of these two proteins uniformly throughout the membrane, including cell body, dendrite branches, and axons. To detect the activation of the receptor upon interaction with its ligand in vivo in Drosophila, we included a UAS-GFP reporter transgene, which expresses GFP in response to Gal4 activity (Brand and Perrimon, 1993).

Figure 1

Download asset Open asset

To test whether the expression of the ligand in donor neurons could reveal the subset of receiver neurons that receive synaptic input from them, we focused on the antennal lobe, the second-order olfactory processing area in the Drosophila brain that receives direct input from the primary olfactory receptor neurons (ORNs) (Figure 2a, Wilson, 2013; Laissue and Vosshall, 2008). Synapses in the antennal lobe are organized into discrete compartments, called glomeruli, and each glomerulus corresponds to a distinct ORN class (defined by the odorant receptor it expresses) (Berck et al., 2016). All ORNs of a specific class project their axons to a common glomerulus (Vosshall et al., 1999; Fishilevich and Vosshall, 2005; Couto et al., 2005). The principal second-order olfactory neuron is called a projection neuron (PN). The vast majority of PNs are uniglomerular PNs (uniPNs) (Figure 2b), and each of these uniPNs have dendrites that branch into single glomeruli where they receive direct synaptic input from a single class of ORNs. Axonal output from uniPNs projects to two major third-order olfactory areas, the mushroom body and the lateral horn, via the medial antennal lobe tract (mALT, formerly iACT) (Ito et al., 2014), (Figure 2a). In addition, a smaller subset of PNs, multiglomerular PNs (multiPNs) (Figure 2c) (Stocker et al., 1990; Parnas et al., 2014; Liang et al., 2013), have dendritic arbors which pool input across multiple glomeruli in the antennal lobe, and send their output to the lateral horn only via the mediolateral ALT (mlALT, formerly the mACT) (Figure 2a) (Tanaka et al., 2012). Finally, a large set of local neurons (LNs) send and receive synaptic input from both ORNs and PNs, densely interconnecting the glomeruli (Figure 2d) (Chou et al., 2010; Hong and Wilson, 2015; Seki et al., 2010).

Figure 2 with 7 supplements see all

Download asset Open asset

Expression of the original TRACT ligand and receptor (without any domains that would localize them to synaptic sites) in neurons failed to reveal any synaptic connections in the antennal lobe. For instance, when the CD19mch ligand is expressed in the majority of ORNs using the Orco-lexA driver (Stocker et al., 1997) and the SNTG4 receptor is expressed from the elav enhancer (Yao and White, 1994), we predicted labeling of both PNs and LNs, the major postsynaptic targets in the antennal lobe. However, in flies carrying Orco > CD19 mch, elav-SNTG4, UAS-GFP transgenes, no GFP induction was observed in the antennal lobe (data not shown), consistent with the findings from another publication (He et al., 2017), suggesting that the system needed adjustments to detect synaptic connections.

To generate a genetic system that could be used to reliably identify synaptically connected neurons we tested a number of different modifications with the ligand, receptor, the drivers used to direct their expression, and reporters (discussed in full detail in the supplementary methods). The following modifications proved to be effective to achieve specific transneuronal labeling:

• To achieve panneuronal expression of the SNTG4 receptor we used the enhancer from the synaptobrevin gene (nSybE), and confirmed that this driver directs expression of the receptor in the vast majority of neuronal types in the brain at detectable levels (Figure 2, Figure 2—figure supplement 1, Figure 2—figure supplement 2).

• To allow for the immunodetection of the receptor, we added the V5 tag to the ICD of SNTG4 (Figure 2, Figure 2—figure supplement 1, Figure 2—figure supplement 2).

• To allow for the immunodetection of the ligand we added the OLLAS tag to the ICD of mCD19 (Figure 2, Figure 2—figure supplement 3).

• To achieve selective localization of the ligand into presynaptic sites we added the ICDs of syndecan or synaptobrevin into CD19 (CD19sdc and nSybCD19, respectively), and confirmed that these ICDs targeted mCD19 into presynaptic sites that co-localized with or were adjacent to the presynaptic marker BRP ((Figure 2, Figure 2—figure supplement 4).

• To achieve selective localization of the receptor into postsynaptic sites we added the ICD of the Drosophila neuroligin gene into SNTG4 (nlgSNTG4) (Figure 1a), and confirmed that this ICD enriched the localization of the SNTG4 receptor in the neuropil regions, which contain neuronal synapses ((Figure 2, Figure 2—figure supplement 1, Figure 2—figure supplement 2).

• To enable visualization of the dendrites and axons of neurons we used the reporter UAS-CD4tdGFP, in which tdGFP is localized into the plasma membrane (Han et al., 2011).

To test the ability of TRACT to reveal the connectivity originating from ORNs, we generated flies carrying orco>nSybCD19 or orco>CD19 sdc, nSybE-nlgSNTG4, and UAS-CD4tdGFP. In these flies, we observed GFP expression in dozens of neurons in the antennal lobe. Roughly half the GFP+ neurons were PNs, which can be identified by their axons projecting into the mushroom body and lateral horn, consistent with the known connectivity between ORNs and PNs (Figure 2e, top panels). In addition, the cell bodies of these axon-bearing neurons were located in the anterodorsal, lateral, and ventral sectors of the antennal lobe, consistent with them being PNs. Finally, most of the putative PNs were immunopositive for CHAT, a known marker of PNs (Figure 2, Figure 2—figure supplement 5). Around half the GFP+ neurons in the antennal lobe did not have axons, were immunopositive for GABA, and their cell bodies were located dorsolateral and ventrolateral, all features consistent with them being LNs (Figure 2e, top panels and Figure 2, Figure 2—figure supplement 5) (Okada et al., 2009).

The experiments with the orco driver revealed that the TRACT system can be used to activate gene expression in neurons known to be connected by synapses, but it did not prove that the LNs and PNs that were highlighted by TRACT were strictly connected by synapses to the ORNs expressing ligand (Figure 1b). In principle, any type of cell-to-cell contact, including non-synaptic membrane-to-membrane contact, could be sufficient to activate the TRACT receptor in the receiver neurons. We took advantage of the anatomy of the Drosophila antennal lobe to investigate whether TRACT could be used to reveal neurons solely connected by synapses. In Drosophila, all ORNs expressing the same olfactory receptor project their axons into two bilaterally symmetric glomeruli in the antennal lobe (Grabe et al., 2016; Tanaka et al., 2012). In addition, the dendrites of individual uniPNs branch into single glomeruli where they make synapses with the axons of ORNs (Figure 2a and b). Due to this feature of the connectivity of uniPNs, it is possible to unambiguously confirm that a given uniPN is connected to a single type of ORN through synaptic contacts that will exclusively occur in an identified glomerulus. Thus, to test the synaptic specificity of TRACT, we focused on the connectivity between ORNs and uniPNs. To selectively express the receptor in PNs, we generated transgenic flies in which the nlgSNTG4 receptor is driven by the GH146 enhancer, which is expressed in the majority of antennal lobe PNs (Stocker et al., 1997) (Figure 2e bottom panels, 3 and 4). To investigate the connectivity between ORNs and PNs, we generated flies carrying orco>nSybCD19, GH146-nlgSNTG4, and UAS-CD4::tdGFP (Han et al., 2011). In these flies we observed dozens of PNs (identifiable by their axons projecting into the mushroom body and lateral horn) with GFP expression in the antennal lobe, consistent with the known connectivity between ORNs and PNs (Figure 2e, bottom panels).

To investigate the ability of TRACT to reveal neurons exclusively connected by synapses we crossed GH146-nlgSNTG4 flies with flies that expressed the nSyb::CD19 and CD19::sdc ligands in identified glomeruli. If TRACT exclusively revealed neurons connected by synaptic contacts, the uniPNs whose receptors were activated by interaction with the ligand would have GFP+ dendrites that would selectively branch in the glomeruli onto where the ORNs would converge their axons. We crossed the GH146-nlgSNTG4 flies with flies expressing the nSyb::CD19 and CD19::sdc ligand in glomeruli VC1 and DA1/VA6/VA1lm under the enhancers of R28H10 and R17H02 LexA drivers, respectively (http://www.virtualflybrain.org) (Figure 3). We observed that with the nSyb::CD19 and CD19::sdc ligands, the uniPNs labeled projected their dendrites selectively into the glomeruli where the ligand was expressed, consistent with previous work (Figure 3) (Grabe et al., 2016). We counted the number of PNs induced by the nSyb::CD19 or CD19::sdc ligands when expressed in identified glomeruli, and found that the observed numbers matched those from previously published work (Grabe et al., 2016) (Table 1a and b). For example, a study using photoactivatable GFP (paGFP) indicated that between 8 and 10 PNs receive input from the DA1 glomerulus (Grabe et al., 2016), and when we expressed the nSyb::CD19 ligand with the R17H02-LexA driver in the DA1 glomerulus, we observed ~11 labeled uniPNs (Figure 3a and Table 1a). Similarly, previous work indicated that between 1–3 and 1–2 PNs receive input from the VA6 and VC1 glomeruli, respectively, and with the nSyb::CD19 ligand, we observed that one uniPN expressed GFP in each of these glomeruli (Figure 3a and Table 1a) (Grabe et al., 2016). In the VA1lm glomerulus, we did not observe any induction in PNs with the nSyb::CD19, which may be due to the low level of the receptor expression in this glomerulus (Figure 3a). When using the CD19::sdc ligand in identified glomeruli, we observed one PN labeled in VA6 and VC1 and ~12 PNs labeled in DA1 (Figure 3b and Table 1b).

Figure 3

Download asset Open asset

A previous work demonstrated that, in addition to the uniPNs, the GH146 enhancer drives expression of transgenes into one multiPN (Marin et al., 2002). With the CD19::sdc and nSyb::CD19, in addition to the uniPNs described above, one multiPN was labeled with both R17H02 and R28H10 drivers, which has been identified previously (Figures 3b and 4 and Table 1a and b) (Marin et al., 2002). The dendrites of this neuron almost cover the whole antennal lobe, and its axon projects first to the lateral horn and then anterodorsally toward the midline. (Figure 3c and d).

Figure 4

Download asset Open asset

Although we did not observe any differences in the number or distribution of uniPNs when the ligands were CD19::Sdc or nSyb::CD19, we observed that the labeling of multiPNs had different requirements for sdc and nSyb. Whereas we could achieve labeling of multiPNs with a single copy of CD19::sdc (in heterozygote animals) (Figures 3b and 4b), we could only detect multiPNs when the nSyb::CD19 ligand was present in two copies (homozygote animals) (Figures 3a and 4a). This observation suggests first, that the amount of ligand may be a limiting factor for transneuronal labeling, and second, that in some cases, the CD19::sdc could be more effective at revealing some synaptic partners than nSyb::CD19. These results indicate that both the nSyb::CD19 and CD19::sdc ligands selectively reveal neurons connected by synaptic contacts between ORNs and PNs, match the available data from previously published works, and confirm that TRACT can be used to perform anterograde transneuronal tracing of brain circuits in the Drosophila brain.

To explore the ability of the TRACT method to trace neuronal circuits in other brain areas, we investigated the brain circuits involved in the control of Drosophila circadian behavior (Figure 5a) (Konopka and Benzer, 1971; Hall, 1998; Rosbash et al., 2003; Young, 1998; Nitabach and Taghert, 2008). There are two key advantages of the circadian circuit to test the usefulness of the TRACT system: (i) it consists of a relatively low number of neurons (around 150), and (ii) the function and anatomy of many of those neurons are known in some detail, although their connectivity is not fully understood (Beckwith and Ceriani, 2015; Peschel and Helfrich-Förster, 2011; Shafer et al., 2006).

Figure 5 with 1 supplement see all

Download asset Open asset

As a test case, we explored the ability of TRACT to reveal the connectivity of the PDF neurons, a well-characterized set of cells in the Drosophila brain that are critical regulators of circadian rhythm (Figures 5b and 6 and Figure 5, Figure 5—figure supplement 1; Renn et al., 1999; Lear et al., 2009). There are two types of PDF neurons in the Drosophila brain: (i) s-LNv neurons, which project their axons into the dorsal regions of the central brain, and (ii) l-LNv neurons, which have large dendrites that branch in the optic lobe, and an axon that projects to the contralateral optic lobe (Figure 5a). In addition to pdf, the period (per) gene is another critical regulator of circadian function in Drosophila (Konopka and Benzer, 1971; Bargiello and Young, 1984; Reddy et al., 1984). Previous works using GRASP suggested that s-LNv neurons were connected to the DN1p (Cavanaugh et al., 2014; Seluzicki et al., 2014) and DN2 neurons (Tang et al., 2017) in the adult fly. In addition, it has been reported that s-LNv neurons make weak connections with LNd neurons that vary throughout the day (Gorostiza et al., 2014). We tested whether we could use TRACT, first, to confirm the connection between s-LNv and DN1 and/or LNd neurons, and second, to identify new synaptic partners of s-LNvs in other brain areas. We used a pdf-lexA driver to express the nSyb::CD19 or CD19::sdc ligands, performed immunostaining against the OLLAS tag and confirmed that the ligands were selectively expressed in the axon terminals of the PDF neurons (Figures 5b and 6 and Figure 5—figure supplement 1).

Figure 6

Download asset Open asset

In the flies carrying pdf>nSyb::CD19, nSybE-nlgSNTG4 and UAS-CD4::tdGFP, we observed several neurons located in the central brain that expressed GFP (Figures 5b and 6 and Table 1c). Consistent with a previous work (Tang et al., 2017) indicating the s-LNvs make synaptic contacts with DN2 neurons, we performed immunocytochemistry using an antibody against the PER protein and observed GFP+, PER+ neurons in the DN2 cluster in 10 out of 18 hemispheres (Figures 5b and 6b, and Table 1c). There are two PER neurons in the DN2 cluster (Helfrich-Förster et al., 2007). In our experiments, in seven hemispheres we observed one GFP+, PER+ DN2 neuron, and in three hemispheres we observed two GFP+, PER+ DN2 neurons (Figures 5b and 6b, and Table 1c). In contrast to previous works, we did not observe any GFP+ cells in the DN1 cluster but we observed several GFP+ cells that were PER-, close to the DN1 cluster (Figures 5b and 6a) (Cavanaugh et al., 2014) (Seluzicki et al., 2014). In addition, we did not detect any induction in the LNd neurons (Figure 5b) (Gorostiza et al., 2014). Finally, we observed GFP induction in some of the DN3 neurons in all the brains we analyzed (Table 1c), and we confirmed that all the GFP+ cells in DN3s were also PER+ (Figures 5b and 6b).

We observed that whereas some of the sets of neurons that were identified as potential synaptic partners with CD19::sdc were similar to those detected with nSyb:CD19, there were also some differences (Figures 5b and 6, and Figure 5, Figure 5—figure supplement 1). Like nSyb::CD19, CD19::sdc also revealed potential synaptic connections with neurons in the DN2 and DN3 cluster (Figure 5, Figure 5—figure supplement 1). However, the numbers of neurons labeled with CD19::sdc were fewer than the ones with nSyb::CD19. This might be due to the fact that CD19::sdc driven by the pdf driver was expressed at very high levels in the cell body, but at much lower levels in the presynaptic terminals (Figure 5, Figure 5—figure supplement 1). In contrast, nSyb::CD19 was clearly detectable in the presynaptic terminals (Figure 5b). In addition, nSyb::CD19, but not CD19::sdc, detected some PER- neurons near the DN1 cluster, and, with CD19::sdc, but not with nSyb::CD19, we observed GFP induction in the l-LNvs and s-LNvs (the pdf neurons themselves) (Figures 5b and 6a and Figure 5, Figure 5—figure supplement 1). Finally, we observed that the patterns of GFP induction between animals in this circuit were more consistent with nSyb::CD19 than with CD19::sdc (Figure 5, Figure 5—figure supplement 1). These observations indicate that TRACT can be used to discover candidate synaptic partners in an unbiased manner. Whereas the overall pattern of transneuronal labeling observed is broadly comparable when using the sdc or nSyb ligands, in some cases there may be potential synaptic partners that may be specifically revealed by one of the ligands, and that further experiments will be required to validate these putative connections.

Our experiments demonstrate that it is possible to take advantage of the molecular mechanisms of ligand-induced intramembrane proteolysis to trace neuronal circuits. We have generated transgenic animals in which neurons expressing an artificial ligand (‘donor’ cells) activate a genetically modified receptor on their synaptic partners (‘receiver’ cells). Using this system, called TRACT (for TRansneuronal ACtivation of Transcription) we have shown that expressing the ligand in a set of donor neurons can activate transcription in synaptically connected neurons in the Drosophila brain, in an anterograde manner. Using TRACT, we have confirmed the connectivity between ORNs and PNs in the antennal lobe, and have discovered new potential connections between PDF and PER neurons in the circadian circuit.

There are several advantages of transneuronal tracing based on ligand-induced membrane proteolysis. (1) TRACT is fully genetically-encoded, it only requires three constructs (a ligand, a receptor, and a reporter) and can be used with high reproducibility in transgenic animals. Moreover, as we demonstrate here, the ligand and/or the receptor can be driven with promoters specific to selective neuronal populations to reveal circuits of synaptically connected neurons. (2) In principle, it can be used in any species amenable to transgenesis. This feature is particularly advantageous for Drosophila (Bellen et al., 2010; St Johnston, 2002), mice (Anderson and Ingham, 2003; Kile and Hilton, 2005) and zebrafish (Fetcho and Liu, 1998; Patton and Zon, 2001), three model organisms of great interest to neurobiologists, with an extensive arsenal of genetic tools. (3) The synaptically connected neurons can be studied in vivo with electrophysiological recordings, live imaging, and optical monitoring of activity, or in fixed tissue, combined with light or electron microscopy. (4) The system can be used in high-throughput experiments because, unlike electron microscopy (Mikula and Denk, 2015; Bock et al., 2011; Tapia et al., 2012), it is not labor intensive or time-consuming. This feature would be very useful to perform genetic screens seeking to identify mutations that affect neuronal connectivity, and to investigate how neuronal connections change in response to behavioral experience, diseases, or exposure to different environmental variables, such as drugs or toxic chemicals. (5) The system can be used to induce the expression of transgenes in the synaptically connected cells that allow for imaging of neuronal activity, such as genetically encoded Ca²⁺ sensors (Chen et al., 2013) or for experimental modification of neuronal activity, such as optogenetic tools (Kim et al., 2017) or ion channels (Lin et al., 2010). (6) TRACT could be used to control neuronal function by regulating endogenous genes indirectly through nuclear translocation of molecules such as Cre, Flp, LexA, QF2, or TetA (Riabinina et al., 2015; Sauer and Henderson, 1988; Dymecki, 1996; Lewandoski, 2001; Venken et al., 2011; del Valle Rodríguez et al., 2011) or directly by fusing endogenous transcription factors to the artificial receptor. (7) In the experiments presented here, we localized the ligand in presynaptic sites and observed anterograde tracing, in a presynaptic to postsynaptic neuron direction. In principle, it should be possible to design a retrograde tracing system by localizing the ligand selectively in the postsynaptic sites (Sheng and Kim, 2011), and/or the receptor in the presynaptic site (Südhof, 2012). (8) Finally, we demonstrate that directing expression of the ligand into a subset of donor neurons localized to a restricted area of the nervous system activates transcription in a very selective subset of neurons that receive synaptic contact from those donor neurons. For example, we were able to activate transcription selectively in specific subsets of antennal lobe PNs that receive synaptic input from individual glomeruli. This observation indicates that even if there are no specific promoters capable of directly driving expression of transgenes into certain neuronal types (such as antennal lobe uniPNs that receive synaptic input from glomerulus VC1 and DN3 neurons), this strategy makes it possible to genetically manipulate highly specific populations of neurons based not on the genes that they express, but on the cells from which they receive synaptic input.

In any strategy for mapping synaptic connectivity there are two key problems that are important to be aware of. First, there may be neurons connected by synapses but the system fails to detect that they are connected (false negatives). For example, it is probably easier to miss the connection between neurons that have very few synapses between them. Second, there may be neurons that are not connected by synapses, but the system indicates that they are connected (false positives). For example, it is possible that some of the connections revealed by the tracing system could be between neurons that are close to each other, but not connected by synapses. This is a problematic scenario, as designing models of computations by brain circuits will be completely inadequate if they include synaptic connections that do not exist. This indicates that with any of the currently available methods to study neuronal connectivity, it is advisable to confirm that any connections revealed are validated by other complementary methods.

Our results of the connectivity between ORNs and the antennal lobe with TRACT are consistent with the published literature. First, with the nSyb and sdc-targeted ligand, we never observed any uniPN whose dendrites projected outside of the glomeruli where the ligand was expressed, thus indicating that TRACT has a very low rate of false positives. Second, with these ligands we consistently observed a number of uniPNs consistent with the published works, suggesting that for the connectivity between ORNs and uniPNs, TRACT has a very low rate of false negatives. However, we could only detect the expected multiPN (Marin et al., 2002) (Figures 3b and 4 and Table 1) in approximately 50% of the cases. Several reasons could account for the variable labeling of multiPNs that we observed in our data. First, we do not know whether the GH146 enhancer drives expression of the nlgSNTG4 receptor into this multiPN in a consistent manner. If the expression of the receptor is variable between animals, this could explain why we detect the multiPN in some brains but not others. Second, whereas uniPNs have all their synapses concentrated in a single glomerulus, multiPNs have synapses distributed throughout multiple glomeruli. This means that the number of synapses between the multiPN and the ligands presented by ORN axons in an individual glomerulus is likely lower than between those of uniPNs and ORNs. This suggests that TRACT probably can detect contacts between ORNs and uniPNs with a higher sensitivity than it can between ORNs and multiPNs.

Our analysis of the connectivity of the PDF neurons in the circadian circuit using TRACT confirmed a recent report indicating a connection between PDF and DN2 neurons (Tang et al., 2017). In addition, using TRACT with the nSyb::CD19 and CD19::sdc ligand revealed that DN3 neurons are new potential postsynaptic targets for PDF neurons, an observation consistent with recent studies demonstrating that applying PDF causes a delay in the circadian phases of Ca²⁺ activity in DN3 (Liang et al., 2017). The role of DN3s in the control of circadian rhythm is still poorly understood, and it would be interesting to use TRACT to selectively manipulate these labeled DN3s to investigate their function in regulating circadian behavior. It is important to note that although the axons from the PDF neurons (l-LNvs and s-LNvs) project very widely through the brain, we identified a handful of putative synaptic targets in brain regions that are consistent with a circadian function. These observations suggest that TRACT probably has a low rate of false positives, consistent with our data on the connectivity of the antennal lobe. Finally, our experiments suggest that TRACT may have some false negatives as it failed to reveal the connections between s-LNvs and DN1s, and between s-LNvs and LNds that have been postulated by previous works using GRASP (Cavanaugh et al., 2014; Seluzicki et al., 2014; Gorostiza et al., 2014). Several reasons could account for this discrepancy. First, in the previous published works, neither of the two components of the GRASP system were localized to synaptic sites, so it is possible that the GRASP signal that was detected could be due to proximity between axons of PDF and DN1 or LNd neurons that were not connected by synapses. Resolving this issue will require validating the connectivity between PDF and DN1 neurons using complementary methods, including electrophysiological methods or electron microscopy. Second, although the nSybE driver is supposed to be pan-neuronally homogeneous, it is likely that it may express transgenes in certain neuronal populations at weaker levels. For example, If the expression of the receptor is weak in DN1 and LNd neurons, this could explain why we did not detect any connections between s-LNvs and DN1 and LNd neurons. Third, it is possible that in our current implementation of the system, TRACT can reveal neurons connected by certain synapses but not others. PDF neurons are characterized by the expression of the neuropeptide PDF, and it is possible that the secretion of neurotransmitters and neuropeptides could occur at different presynaptic sites. It is possible that in this current implementation, TRACT may be able to detect synaptic contacts mediated by neurotransmitter vesicles, but perhaps it may fail to detect neuropeptide release sites. Further tests of the system in multiple circuits would be required to fully assess the sensitivity of TRACT.

Our results indicate that the choice of the domain used to target the ligand into the presynaptic sites determines the specific potential synaptic partners identified by TRACT. In initial experiments (described in the supplementary materials) we observed that fusing the CD19 ligand to the intracellular domains of several molecules known to be localized in presynaptic sites (including syntaxin, dpr-10, dip- ɣ, and neurexin) was sufficient to enrich the localization of the TRACT ligand into presynaptic compartments. However, when using the stringent test afforded by the anatomical specificity of uniPNs (whose dendrites branch into a single glomerulus where they receive synapses from ORNs expressing a single olfactory receptor molecule), we observed that using these ligands we could detect GFP expression in uniPNs whose dendrites branched in glomeruli other than the one where the TRACT ligand was expressed. These results suggest that the domains that we used from syntaxin, dpr-10, dip- ɣ, and neurexin were not useful to restrict transneuronal labeling between neurons that were exclusively connected by synapses (as opposed to any other non-synaptic forms of cell-cell contact, including mere proximity between their respective plasma membranes). In contrast, we observed that fusing the CD19 ligand to the intracellular domains from sdc and nSyb enables reliable transneuronal tracing that was restricted to neurons connected by synapses, as assessed by the stringent test of the branching of uniPNs’ dendrites into single glomeruli. We did not observe any difference in the number or distribution of uniPNs labeled with nSyb::CD19 or CD19::sdc in the antennal lobe. However, we observed that whereas a single copy of CD19::sdc was sufficient to label multiPNs, two copies of nSyb::CD19 were necessary to produce any multiPN labeling. This observation indicates that both ligands are effective at detecting both uni- and multiPNs. However, in some cases the amount of the ligand could be a limiting factor to the intensity of the labeling.

In the circadian circuit we observed that whereas the overall pattern of transneuronal labeling revealed by TRACT was similar with nSyb::CD19 and CD19::sdc, there were some differences.

First, we observed that the patterns of GFP induction were more consistent between animals with nSyb::CD19 than with CD19::sdc (Figures 5b and 6 and Figure 5, Figure 5—figure supplement 5). Second, we observed that some neuronal populations were labeled with one of the ligands, but not the other. Most notably, nSyb::CD19 (but not CD19::sdc) labeled some PER- neurons near the DN1 cluster and CD19::sdc (but not nSyb::CD19) labeled l-LNvs, and s-LNVs (the pdf neurons themselves). Several reasons could account for these differences, including the different biological functions of nSyb and sdc. nSyb is a molecule predominantly localized into the synaptic vesicles, which is displayed on the plasma membrane after synaptic vesicle fusion. Sdc is a heparan sulfate proteoglycan that in neurons is enriched in presynaptic sites. Thus, given the different biological functions of sdc and nSyb, and their specific subcellular localizations, it is expected that their domains may target expression of the TRACT ligands into slightly different locations in the presynaptic site, or they may influence its abundance on the synaptic cleft. Finally, it is possible that creating a hybrid molecule that combines, for example, some of the domains from a presynaptic marker and CD19 may perturb its targeting to the intended synaptic compartment, its abundance, or stability.

The two methods currently available to unambiguously confirm neuronal connectivity are dual single-cell electrophysiological recordings and serial electron microscopy. However, because single-cell recordings have a low throughput (in a typical experiment, only a handful of connections can be tested every day), this method is well suited to confirm the connections suggested by a different method, but it is not feasible to identify neuronal connections in an exploratory manner. Serial electron microscopy is an extremely powerful method to identify neuronal connections, but currently it has two main limitations. First, because it has a low throughput it cannot be used to study the connectivity of brain circuits across multiple animals. Second, the area that can be acquired in a single imaging event is restricted to a maximum of around 1 mm². This makes it well suited to investigate local connectivity, but extremely challenging to follow the connections between neurons located far away from each other, because they cannot be observed in a single imaging acquisition. The results shown here demonstrate that TRACT allows investigators to discover potential synaptic targets for genetically identified neurons in an unbiased manner, across multiple animals, thus allowing for high-throughput approaches. Then, the candidate synaptic partners identified by TRACT can be validated using complementary low-throughput methods that allow for unambiguous confirmation of connectivity, including dual single-cell electrophysiological recordings and/or electron microscopy.

While the present work was under review another work describing a genetic strategy to identify synaptically connected neurons called trans-TANGO was published (Talay et al., 2017). Trans-TANGO is the implementation of the TANGO system with a membrane-bound ligand, so that it can be applied to the study of neuronal connectivity. Both TRACT and trans-TANGO depend of ligand-induced proteolysis and subsequent release of a membrane anchored transcription factor. TRACT is based on intramembrane proteolysis by the endogenous ɣ-secretase, which is ubiquitously present in the plasma membrane of metazoan cells. Trans-TANGO depends on the cleavage and release of a transcription factor by the reconstitution of the viral protease TEV, which is triggered by the interaction between arrestin and the activation of an exogenous, engineered GPCR upon interaction with its ligand. The initial report of trans-TANGO also studied the connectivity between ORNs and antennal lobe neurons in Drosophila to validate its specificity. Our results in the Drosophila antennal lobe with TRACT, when the ligand was expressed into individual glomeruli, revealed connections between ORNs and uniPNs in the antennal lobe that are consistent with the published literature, but we only detected multiPNs in around 50% of the brains tested. In the trans-TANGO report with ligand expression into individual glomeruli, there was no evidence of detection of multiPNs, and the number of postsynaptic uniPNs observed with transTANGO was higher than those reported in previously published works. The authors concluded that the higher numbers of PNs observed with trans-Tango could represent false-positive signals that might have resulted from inefficient synaptic localization of the ligand due to its overexpression. Further experiments will clarify the respective rate of false positives and false negatives by TRACT and transTANGO, and will allow for the optimization of these systems to enable the reliable application of these strategies for the investigation of neuronal connections in brain circuits.

In recent years, there has been a surge in interest for new methods to investigate synaptic connectivity in brain circuits (Meinertzhagen and Lee, 2012; Denk et al., 2012; Bargmann and Marder, 2013; Swanson and Lichtman, 2016). Identifying how neurons are connected is a valuable guide towards understanding how computations take place in the brain. In addition, recent research indicates that abnormal neuronal wiring might be the cause of several neurodevelopmental and psychiatric disorders, including autism and schizophrenia (Peça and Feng, 2012; Rubinov and Bullmore, 2013; Mevel and Fransson, 2016; Narr and Leaver, 2015). We anticipate that the advantages of TRACT will make it a useful addition to the arsenal of tools available to identify the wiring diagrams of brain circuits, and will open new avenues for enabling the genetic manipulation of neurons connected by synapses.

1. 1. Anderson KV
   2. Ingham PW

   (2003) The transformation of the model organism: a decade of developmental genetics

   Nature Genetics 33 Suppl:285–293.

   https://doi.org/10.1038/ng1105

   • PubMed
   • Google Scholar
2. 1. Bargiello TA
   2. Young MW

   (1984) Molecular genetics of a biological clock in Drosophila

   PNAS 81:2142–2146.

   https://doi.org/10.1073/pnas.81.7.2142

   • PubMed
   • Google Scholar
3. 1. Bargmann CI
   2. Marder E

   (2013) From the connectome to brain function

   Nature Methods 10:483–490.

   https://doi.org/10.1038/nmeth.2451

   • PubMed
   • Google Scholar
4. 1. Beckwith EJ
   2. Ceriani MF

   (2015) Communication between circadian clusters: The key to a plastic network

   FEBS Letters 589:3336–3342.

   https://doi.org/10.1016/j.febslet.2015.08.017

   • PubMed
   • Google Scholar
5. 1. Bellen HJ
   2. Tong C
   3. Tsuda H

   (2010) 100 years of Drosophila research and its impact on vertebrate neuroscience: a history lesson for the future

   Nature Reviews Neuroscience 11:514–522.

   https://doi.org/10.1038/nrn2839

   • PubMed
   • Google Scholar
6. 1. Berck ME
   2. Khandelwal A
   3. Claus L
   4. Hernandez-Nunez L
   5. Si G
   6. Tabone CJ
   7. Li F
   8. Truman JW
   9. Fetter RD
   10. Louis M
   11. Samuel AD
   12. Cardona A

   (2016) The wiring diagram of a glomerular olfactory system

   eLife 5:e14859.

   https://doi.org/10.7554/eLife.14859

   • PubMed
   • Google Scholar
7. 1. Bock DD
   2. Lee WC
   3. Kerlin AM
   4. Andermann ML
   5. Hood G
   6. Wetzel AW
   7. Yurgenson S
   8. Soucy ER
   9. Kim HS
   10. Reid RC

   (2011) Network anatomy and in vivo physiology of visual cortical neurons

   Nature 471:177–182.

   https://doi.org/10.1038/nature09802

   • PubMed
   • Google Scholar
8. 1. Brand AH
   2. Perrimon N

   (1993)

   Targeted gene expression as a means of altering cell fates and generating dominant phenotypes

   Development 118:401–415.

   • PubMed
   • Google Scholar
9. 1. Cavanaugh DJ
   2. Geratowski JD
   3. Wooltorton JR
   4. Spaethling JM
   5. Hector CE
   6. Zheng X
   7. Johnson EC
   8. Eberwine JH
   9. Sehgal A

   (2014) Identification of a circadian output circuit for rest:activity rhythms in Drosophila

   Cell 157:689–701.

   https://doi.org/10.1016/j.cell.2014.02.024

   • PubMed
   • Google Scholar
10. 1. Chen TW
    2. Wardill TJ
    3. Sun Y
    4. Pulver SR
    5. Renninger SL
    6. Baohan A
    7. Schreiter ER
    8. Kerr RA
    9. Orger MB
    10. Jayaraman V
    11. Looger LL
    12. Svoboda K
    13. Kim DS

    (2013) Ultrasensitive fluorescent proteins for imaging neuronal activity

    Nature 499:295–300.

    https://doi.org/10.1038/nature12354

    • PubMed
    • Google Scholar
11. 1. Chou YH
    2. Spletter ML
    3. Yaksi E
    4. Leong JC
    5. Wilson RI
    6. Luo L

    (2010) Diversity and wiring variability of olfactory local interneurons in the Drosophila antennal lobe

    Nature Neuroscience 13:439–449.

    https://doi.org/10.1038/nn.2489

    • PubMed
    • Google Scholar
12. 1. Couto A
    2. Alenius M
    3. Dickson BJ

    (2005) Molecular, anatomical, and functional organization of the Drosophila olfactory system

    Current Biology 15:1535–1547.

    https://doi.org/10.1016/j.cub.2005.07.034

    • PubMed
    • Google Scholar
13. 1. del Valle Rodríguez A
    2. Didiano D
    3. Desplan C

    (2011) Power tools for gene expression and clonal analysis in Drosophila

    Nature Methods 9:47–55.

    https://doi.org/10.1038/nmeth.1800

    • PubMed
    • Google Scholar
14. 1. Denk W
    2. Briggman KL
    3. Helmstaedter M

    (2012) Structural neurobiology: missing link to a mechanistic understanding of neural computation

    Nature Reviews Neuroscience 13:351–358.

    https://doi.org/10.1038/nrn3169

    • PubMed
    • Google Scholar
15. 1. Dubowy C
    2. Sehgal A

    (2017) Circadian Rhythms and Sleep in Drosophila melanogaster

    Genetics 205:1373–1397.

    https://doi.org/10.1534/genetics.115.185157

    • PubMed
    • Google Scholar
16. 1. Dymecki SM

    (1996) Flp recombinase promotes site-specific DNA recombination in embryonic stem cells and transgenic mice

    PNAS 93:6191–6196.

    https://doi.org/10.1073/pnas.93.12.6191

    • PubMed
    • Google Scholar
17. 1. Fetcho JR
    2. Liu KS

    (1998) Zebrafish as a model system for studying neuronal circuits and behavior

    Annals of the New York Academy of Sciences 860:333–345.

    https://doi.org/10.1111/j.1749-6632.1998.tb09060.x

    • PubMed
    • Google Scholar
18. 1. Fishilevich E
    2. Vosshall LB

    (2005) Genetic and functional subdivision of the Drosophila antennal lobe

    Current Biology 15:1548–1553.

    https://doi.org/10.1016/j.cub.2005.07.066

    • PubMed
    • Google Scholar
19. 1. Fujimoto M
    2. Poe JC
    3. Inaoki M
    4. Tedder TF

    (1998) CD19 regulates B lymphocyte responses to transmembrane signals

    Seminars in Immunology 10:267–277.

    https://doi.org/10.1006/smim.1998.9999

    • PubMed
    • Google Scholar
20. 1. Gordon WR
    2. Zimmerman B
    3. He L
    4. Miles LJ
    5. Huang J
    6. Tiyanont K
    7. McArthur DG
    8. Aster JC
    9. Perrimon N
    10. Loparo JJ
    11. Blacklow SC

    (2015) Mechanical Allostery: Evidence for a Force Requirement in the Proteolytic Activation of Notch

    Developmental Cell 33:729–736.

    https://doi.org/10.1016/j.devcel.2015.05.004

    • PubMed
    • Google Scholar
21. 1. Gorostiza EA
    2. Depetris-Chauvin A
    3. Frenkel L
    4. Pírez N
    5. Ceriani MF

    (2014) Circadian pacemaker neurons change synaptic contacts across the day

    Current Biology 24:2161–2167.

    https://doi.org/10.1016/j.cub.2014.07.063

    • PubMed
    • Google Scholar
22. 1. Grabe V
    2. Baschwitz A
    3. Dweck HKM
    4. Lavista-Llanos S
    5. Hansson BS
    6. Sachse S

    (2016) Elucidating the Neuronal Architecture of Olfactory Glomeruli in the Drosophila Antennal Lobe

    Cell Reports 16:3401–3413.

    https://doi.org/10.1016/j.celrep.2016.08.063

    • PubMed
    • Google Scholar
23. 1. Hall JC

    (1998)

    Genetics of biological rhythms in drosophila

    Advances in Genetics 38:135–184.

    • PubMed
    • Google Scholar
24. 1. Han C
    2. Jan LY
    3. Jan YN

    (2011) Enhancer-driven membrane markers for analysis of nonautonomous mechanisms reveal neuron-glia interactions in Drosophila

    PNAS 108:9673–9678.

    https://doi.org/10.1073/pnas.1106386108

    • PubMed
    • Google Scholar
25. 1. He L
    2. Huang J
    3. Perrimon N

    (2017) Development of an optimized synthetic Notch receptor as an in vivo cell-cell contact sensor

    PNAS 114:5467–5472.

    https://doi.org/10.1073/pnas.1703205114

    • PubMed
    • Google Scholar
26. 1. Helfrich-Förster C
    2. Shafer OT
    3. Wülbeck C
    4. Grieshaber E
    5. Rieger D
    6. Taghert P

    (2007) Development and morphology of the clock-gene-expressing lateral neurons of Drosophila melanogaster

    The Journal of Comparative Neurology 500:47–70.

    https://doi.org/10.1002/cne.21146

    • PubMed
    • Google Scholar
27. 1. Hill CA
    2. Fox AN
    3. Pitts RJ
    4. Kent LB
    5. Tan PL
    6. Chrystal MA
    7. Cravchik A
    8. Collins FH
    9. Robertson HM
    10. Zwiebel LJ

    (2002) G protein-coupled receptors in Anopheles gambiae

    Science 298:176–178.

    https://doi.org/10.1126/science.1076196

    • PubMed
    • Google Scholar
28. 1. Hong EJ
    2. Wilson RI

    (2015) Simultaneous encoding of odors by channels with diverse sensitivity to inhibition

    Neuron 85:573–589.

    https://doi.org/10.1016/j.neuron.2014.12.040

    • PubMed
    • Google Scholar
29. 1. Huang TH
    2. Velho T
    3. Lois C

    (2016) Monitoring cell-cell contacts in vivo in transgenic animals

    Development 143:4073–4084.

    https://doi.org/10.1242/dev.142406

    • PubMed
    • Google Scholar
30. 1. Ito K
    2. Shinomiya K
    3. Ito M
    4. Armstrong JD
    5. Boyan G
    6. Hartenstein V
    7. Harzsch S
    8. Heisenberg M
    9. Homberg U
    10. Jenett A
    11. Keshishian H
    12. Restifo LL
    13. Rössler W
    14. Simpson JH
    15. Strausfeld NJ
    16. Strauss R
    17. Vosshall LB
    18. Insect Brain Name Working Group

    (2014) A systematic nomenclature for the insect brain

    Neuron 81:755–765.

    https://doi.org/10.1016/j.neuron.2013.12.017

    • PubMed
    • Google Scholar
31. 1. Kile BT
    2. Hilton DJ

    (2005) The art and design of genetic screens: mouse

    Nature Reviews Genetics 6:557–567.

    https://doi.org/10.1038/nrg1636

    • PubMed
    • Google Scholar
32. 1. Kim CK
    2. Adhikari A
    3. Deisseroth K

    (2017) Integration of optogenetics with complementary methodologies in systems neuroscience

    Nature Reviews Neuroscience 18:222–235.

    https://doi.org/10.1038/nrn.2017.15

    • PubMed
    • Google Scholar
33. 1. Kochenderfer JN
    2. Feldman SA
    3. Zhao Y
    4. Xu H
    5. Black MA
    6. Morgan RA
    7. Wilson WH
    8. Rosenberg SA

    (2009) Construction and preclinical evaluation of an anti-CD19 chimeric antigen receptor

    Journal of Immunotherapy 32:689–702.

    https://doi.org/10.1097/CJI.0b013e3181ac6138

    • PubMed
    • Google Scholar
34. 1. Konopka RJ
    2. Benzer S

    (1971) Clock mutants of Drosophila melanogaster

    PNAS 68:2112–2116.

    https://doi.org/10.1073/pnas.68.9.2112

    • PubMed
    • Google Scholar
35. 1. Kovall RA
    2. Gebelein B
    3. Sprinzak D
    4. Kopan R

    (2017) The Canonical Notch Signaling Pathway: Structural and Biochemical Insights into Shape, Sugar, and Force

    Developmental Cell 41:228–241.

    https://doi.org/10.1016/j.devcel.2017.04.001

    • PubMed
    • Google Scholar
36. 1. Laissue PP
    2. Vosshall LB

    (2008) The olfactory sensory map in Drosophila

    Advances in Experimental Medicine and Biology 628:102–114.

    https://doi.org/10.1007/978-0-387-78261-4_7

    • PubMed
    • Google Scholar
37. 1. Lear BC
    2. Zhang L
    3. Allada R

    (2009) The neuropeptide PDF acts directly on evening pacemaker neurons to regulate multiple features of circadian behavior

    PLoS Biology 7:e1000154.

    https://doi.org/10.1371/journal.pbio.1000154

    • PubMed
    • Google Scholar
38. 1. Lee D
    2. Huang TH
    3. De La Cruz A
    4. Callejas A
    5. Lois C

    (2017) Methods to investigate the structure and connectivity of the nervous system

    Fly 11:224–238.

    https://doi.org/10.1080/19336934.2017.1295189

    • PubMed
    • Google Scholar
39. 1. Lewandoski M

    (2001) Conditional control of gene expression in the mouse

    Nature Reviews Genetics 2:743–755.

    https://doi.org/10.1038/35093537

    • PubMed
    • Google Scholar
40. 1. Liang L
    2. Li Y
    3. Potter CJ
    4. Yizhar O
    5. Deisseroth K
    6. Tsien RW
    7. Luo L

    (2013) GABAergic projection neurons route selective olfactory inputs to specific higher-order neurons

    Neuron 79:917–931.

    https://doi.org/10.1016/j.neuron.2013.06.014

    • PubMed
    • Google Scholar
41. 1. Liang X
    2. Holy TE
    3. Taghert PH

    (2017) A series of suppressive signals within the drosophila circadian neural circuit generates sequential daily outputs

    Neuron 94:1173–1189.

    https://doi.org/10.1016/j.neuron.2017.05.007

    • PubMed
    • Google Scholar
42. 1. Lin CW
    2. Sim S
    3. Ainsworth A
    4. Okada M
    5. Kelsch W
    6. Lois C

    (2010) Genetically increased cell-intrinsic excitability enhances neuronal integration into adult brain circuits

    Neuron 65:32–39.

    https://doi.org/10.1016/j.neuron.2009.12.001

    • PubMed
    • Google Scholar
43. 1. Marin EC
    2. Jefferis GS
    3. Komiyama T
    4. Zhu H
    5. Luo L

    (2002) Representation of the glomerular olfactory map in the Drosophila brain

    Cell 109:243–255.

    https://doi.org/10.1016/S0092-8674(02)00700-6

    • PubMed
    • Google Scholar
44. 1. Meinertzhagen IA
    2. Lee CH

    (2012) The genetic analysis of functional connectomics in Drosophila

    Advances in Genetics 80:99–151.

    https://doi.org/10.1016/B978-0-12-404742-6.00003-X

    • PubMed
    • Google Scholar
45. 1. Mevel K
    2. Fransson P

    (2016) The functional brain connectome of the child and autism spectrum disorders

    Acta Paediatrica 105:1024–1035.

    https://doi.org/10.1111/apa.13484

    • PubMed
    • Google Scholar
46. 1. Mikula S
    2. Denk W

    (2015) High-resolution whole-brain staining for electron microscopic circuit reconstruction

    Nature Methods 12:541–546.

    https://doi.org/10.1038/nmeth.3361

    • PubMed
    • Google Scholar
47. 1. Morsut L
    2. Roybal KT
    3. Xiong X
    4. Gordley RM
    5. Coyle SM
    6. Thomson M
    7. Lim WA

    (2016) Engineering customized cell sensing and response behaviors using synthetic notch receptors

    Cell 164:780–791.

    https://doi.org/10.1016/j.cell.2016.01.012

    • PubMed
    • Google Scholar
48. 1. Narr KL
    2. Leaver AM

    (2015) Connectome and schizophrenia

    Current Opinion in Psychiatry 28:229–235.

    https://doi.org/10.1097/YCO.0000000000000157

    • PubMed
    • Google Scholar
49. 1. Nitabach MN
    2. Taghert PH

    (2008) Organization of the Drosophila circadian control circuit

    Current Biology 18:R84–R93.

    https://doi.org/10.1016/j.cub.2007.11.061

    • PubMed
    • Google Scholar
50. 1. Okada R
    2. Awasaki T
    3. Ito K

    (2009) Gamma-aminobutyric acid (GABA)-mediated neural connections in the Drosophila antennal lobe

    The Journal of Comparative Neurology 514:74–91.

    https://doi.org/10.1002/cne.21971

    • PubMed
    • Google Scholar
51. 1. Parnas M
    2. Lin AC
    3. Huetteroth W
    4. Miesenböck G

    (2013) Odor discrimination in Drosophila: from neural population codes to behavior

    Neuron 79:932–944.

    https://doi.org/10.1016/j.neuron.2013.08.006

    • PubMed
    • Google Scholar
52. 1. Patton EE
    2. Zon LI

    (2001) The art and design of genetic screens: zebrafish

    Nature Reviews Genetics 2:956–966.

    https://doi.org/10.1038/35103567

    • PubMed
    • Google Scholar
53. 1. Peça J
    2. Feng G

    (2012) Cellular and synaptic network defects in autism

    Current Opinion in Neurobiology 22:866–872.

    https://doi.org/10.1016/j.conb.2012.02.015

    • PubMed
    • Google Scholar
54. 1. Peschel N
    2. Helfrich-Förster C

    (2011) Setting the clock--by nature: circadian rhythm in the fruitfly Drosophila melanogaster

    FEBS Letters 585:1435–1442.

    https://doi.org/10.1016/j.febslet.2011.02.028

    • PubMed
    • Google Scholar
55. 1. Reddy P
    2. Zehring WA
    3. Wheeler DA
    4. Pirrotta V
    5. Hadfield C
    6. Hall JC
    7. Rosbash M

    (1984) Molecular analysis of the period locus in Drosophila melanogaster and identification of a transcript involved in biological rhythms

    Cell 38:701–710.

    https://doi.org/10.1016/0092-8674(84)90265-4

    • PubMed
    • Google Scholar
56. 1. Renn SC
    2. Park JH
    3. Rosbash M
    4. Hall JC
    5. Taghert PH

    (1999) A pdf neuropeptide gene mutation and ablation of PDF neurons each cause severe abnormalities of behavioral circadian rhythms in Drosophila

    Cell 99:791–802.

    https://doi.org/10.1016/S0092-8674(00)81676-1

    • PubMed
    • Google Scholar
57. 1. Riabinina O
    2. Luginbuhl D
    3. Marr E
    4. Liu S
    5. Wu MN
    6. Luo L
    7. Potter CJ

    (2015) Improved and expanded Q-system reagents for genetic manipulations

    Nature Methods 12:219–222.

    https://doi.org/10.1038/nmeth.3250

    • PubMed
    • Google Scholar
58. 1. Rosbash M
    2. Allada R
    3. McDonald M
    4. Peng Y
    5. Zhao J

    (2003)

    Circadian rhythms in Drosophila

    Novartis Foundation Symposium 253:223–232.

    • PubMed
    • Google Scholar
59. 1. Rubinov M
    2. Bullmore E

    (2013) Fledgling pathoconnectomics of psychiatric disorders

    Trends in Cognitive Sciences 17:641–647.

    https://doi.org/10.1016/j.tics.2013.10.007

    • PubMed
    • Google Scholar
60. 1. Rybak J
    2. Talarico G
    3. Ruiz S
    4. Arnold C
    5. Cantera R
    6. Hansson BS

    (2016) Synaptic circuitry of identified neurons in the antennal lobe of Drosophila melanogaster

    Journal of Comparative Neurology 524:1920–1956.

    https://doi.org/10.1002/cne.23966

    • PubMed
    • Google Scholar
61. 1. Sauer B
    2. Henderson N

    (1988) Site-specific DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1

    PNAS 85:5166–5170.

    https://doi.org/10.1073/pnas.85.14.5166

    • PubMed
    • Google Scholar
62. 1. Seki Y
    2. Rybak J
    3. Wicher D
    4. Sachse S
    5. Hansson BS

    (2010) Physiological and morphological characterization of local interneurons in the Drosophila antennal lobe

    Journal of Neurophysiology 104:1007–1019.

    https://doi.org/10.1152/jn.00249.2010

    • PubMed
    • Google Scholar
63. 1. Seluzicki A
    2. Flourakis M
    3. Kula-Eversole E
    4. Zhang L
    5. Kilman V
    6. Allada R

    (2014) Dual PDF signaling pathways reset clocks via TIMELESS and acutely excite target neurons to control circadian behavior

    PLoS Biology 12:e1001810.

    https://doi.org/10.1371/journal.pbio.1001810

    • PubMed
    • Google Scholar
64. 1. Shafer OT
    2. Helfrich-Förster C
    3. Renn SC
    4. Taghert PH

    (2006) Reevaluation of Drosophila melanogaster's neuronal circadian pacemakers reveals new neuronal classes

    The Journal of Comparative Neurology 498:180–193.

    https://doi.org/10.1002/cne.21021

    • PubMed
    • Google Scholar
65. 1. Sheng M
    2. Kim E

    (2011) The postsynaptic organization of synapses

    Cold Spring Harbor Perspectives in Biology 3:a005678.

    https://doi.org/10.1101/cshperspect.a005678

    • PubMed
    • Google Scholar
66. 1. Sprinzak D
    2. Lakhanpal A
    3. Lebon L
    4. Santat LA
    5. Fontes ME
    6. Anderson GA
    7. Garcia-Ojalvo J
    8. Elowitz MB

    (2010) Cis-interactions between Notch and Delta generate mutually exclusive signalling states

    Nature 465:86–90.

    https://doi.org/10.1038/nature08959

    • PubMed
    • Google Scholar
67. 1. St Johnston D

    (2002) The art and design of genetic screens: Drosophila melanogaster

    Nature Reviews Genetics 3:176–188.

    https://doi.org/10.1038/nrg751

    • PubMed
    • Google Scholar
68. 1. Stocker RF
    2. Heimbeck G
    3. Gendre N
    4. de Belle JS

    (1997) Neuroblast ablation in Drosophila P[GAL4] lines reveals origins of olfactory interneurons

    Journal of Neurobiology 32:443–456.

    https://doi.org/10.1002/(SICI)1097-4695(199705)32:5<443::AID-NEU1>3.0.CO;2-5

    • PubMed
    • Google Scholar
69. 1. Stocker RF
    2. Lienhard MC
    3. Borst A
    4. Fischbach KF

    (1990) Neuronal architecture of the antennal lobe in Drosophila melanogaster

    Cell and Tissue Research 262:9–34.

    https://doi.org/10.1007/BF00327741

    • PubMed
    • Google Scholar
70. 1. Südhof TC

    (2012) The presynaptic active zone

    Neuron 75:11–25.

    https://doi.org/10.1016/j.neuron.2012.06.012

    • PubMed
    • Google Scholar
71. 1. Swanson LW
    2. Lichtman JW

    (2016) From Cajal to Connectome and Beyond

    Annual Review of Neuroscience 39:197–216.

    https://doi.org/10.1146/annurev-neuro-071714-033954

    • PubMed
    • Google Scholar
72. 1. Talay M
    2. Richman EB
    3. Snell NJ
    4. Hartmann GG
    5. Fisher JD
    6. Sorkaç A
    7. Santoyo JF
    8. Chou-Freed C
    9. Nair N
    10. Johnson M
    11. Szymanski JR
    12. Barnea G

    (2017) Transsynaptic mapping of second-order taste neurons in flies by trans-tango

    Neuron 96:783–795.

    https://doi.org/10.1016/j.neuron.2017.10.011

    • PubMed
    • Google Scholar
73. 1. Tanaka NK
    2. Endo K
    3. Ito K

    (2012) Organization of antennal lobe-associated neurons in adult Drosophila melanogaster brain

    The Journal of Comparative Neurology 520:4067–4130.

    https://doi.org/10.1002/cne.23142

    • PubMed
    • Google Scholar
74. 1. Tang X
    2. Roessingh S
    3. Hayley SE
    4. Chu ML
    5. Tanaka NK
    6. Wolfgang W
    7. Song S
    8. Stanewsky R
    9. Hamada FN

    (2017) The role of PDF neurons in setting the preferred temperature before dawn in Drosophila

    eLife 6:e23206.

    https://doi.org/10.7554/eLife.23206

    • PubMed
    • Google Scholar
75. 1. Tapia JC
    2. Kasthuri N
    3. Hayworth KJ
    4. Schalek R
    5. Lichtman JW
    6. Smith SJ
    7. Buchanan J

    (2012) High-contrast en bloc staining of neuronal tissue for field emission scanning electron microscopy

    Nature Protocols 7:193–206.

    https://doi.org/10.1038/nprot.2011.439

    • PubMed
    • Google Scholar
76. 1. Van Vactor D
    2. Sigrist SJ

    (2017) Presynaptic morphogenesis, active zone organization and structural plasticity in Drosophila

    Current Opinion in Neurobiology 43:119–129.

    https://doi.org/10.1016/j.conb.2017.03.003

    • PubMed
    • Google Scholar
77. 1. Venken KJ
    2. Simpson JH
    3. Bellen HJ

    (2011) Genetic manipulation of genes and cells in the nervous system of the fruit fly

    Neuron 72:202–230.

    https://doi.org/10.1016/j.neuron.2011.09.021

    • PubMed
    • Google Scholar
78. 1. Vosshall LB
    2. Amrein H
    3. Morozov PS
    4. Rzhetsky A
    5. Axel R

    (1999) A spatial map of olfactory receptor expression in the Drosophila antenna

    Cell 96:725–736.

    https://doi.org/10.1016/S0092-8674(00)80582-6

    • PubMed
    • Google Scholar
79. 1. Wilson RI

    (2013) Early olfactory processing in Drosophila: mechanisms and principles

    Annual Review of Neuroscience 36:217–241.

    https://doi.org/10.1146/annurev-neuro-062111-150533

    • PubMed
    • Google Scholar
80. 1. Yao KM
    2. White K

    (1994) Neural specificity of elav expression: defining a Drosophila promoter for directing expression to the nervous system

    Journal of Neurochemistry 63:41–51.

    https://doi.org/10.1046/j.1471-4159.1994.63010041.x

    • PubMed
    • Google Scholar
81. 1. Young MW

    (1998) The molecular control of circadian behavioral rhythms and their entrainment in Drosophila

    Annual Review of Biochemistry 67:135–152.

    https://doi.org/10.1146/annurev.biochem.67.1.135

    • PubMed
    • Google Scholar

Author details

1. Ting-hao Huang

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2546-3525

2. Peter Niesman

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Formal analysis, Validation, Investigation, Visualization, Writing—review and editing

   Competing interests

   No competing interests declared

3. Deepshika Arasu

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Validation, Investigation, Visualization

   Competing interests

   No competing interests declared

4. Donghyung Lee

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Validation, Investigation, Visualization

   Competing interests

   No competing interests declared

5. Aubrie L De La Cruz

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Data curation, Validation, Investigation, Visualization, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7351-0285

6. Antuca Callejas

   1. Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States
   2. Department of Cell Biology, School of Science, University of Extremadura, Badajoz, Spain

   Contribution

   Validation, Investigation, Visualization

   Competing interests

   No competing interests declared

7. Elizabeth J Hong

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Experimental design, Analysis and data interpretation, Funding

   Competing interests

   No competing interests declared

8. Carlos Lois

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   clois@caltech.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7305-2317

• 10,712

  views

• 1,907

  downloads

• 26

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.