Despite having a limited number of senses, animals can perceive a huge range of sensations. One possible explanation is that the brain combines several stimuli to make each specific sensation.

The olfactory learning system in the fruit fly Drosophila melanogaster is in a part of the brain called the mushroom body. It allows fruit flies to associate a specific smell with a reward (e.g. food) or a punishment (e.g. poison) and behave accordingly. Two groups of neurons process stimuli from sensory receptors in the mushroom body: olfactory projection neurons carry information from the receptors and pass it on to neurons called Kenyon cells. The system relies on Kenyon cells receiving the combined input of multiple olfactory projection neurons, and therefore information from multiple receptors. The number of inputs each Kenyon cell receives is thought to determine the number of sensations that can be told apart, and thus, the number of signals that can be used for learning.

While many mechanisms dictating the complexity of a neuron’s shape have been described, the logic behind how two populations of neurons become connected to combine several inputs into a single sensation has not been addressed. A better understanding of how these connections are established during development can help explain how the brain processes information, and the D. melanogaster mushroom body is a good system to address these questions.

Elkahlah, Rogow et al. manipulated the number of olfactory projection neurons and Kenyon cells in the mushroom body of fruit flies during development. They found that despite there being a varying number of cells, the number of connections into a post-synaptic cell remained the same. This indicates that the logic behind the combinations of inputs required for a sensation depends on the Kenyon cell, while olfactory projection neurons can adapt during their development to suit these input demands. Thus, if there are fewer Kenyon cells, the olfactory projection neurons will each provide connections to fewer cells to compensate, and if there are fewer olfactory projection neurons, each of them will input into more Kenyon cells. To show that the developing mushroom body could indeed adapt to different numbers of olfactory projection neurons and Kenyon cells, the modified flies were tested for olfactory perception: their responses to odor were largely normal.

These results underline the robustness of neuronal circuits. During development, the mushroom body can compensate for missing or extra neurons by modifying the numbers of connections between two groups of neurons, thus allowing the olfactory system to work normally. This robustness may also predispose the system to evolutionary change, since it allows the system to continue working as it changes. These findings are relevant to any area of the brain where neurons rely on combined input from many sources.

The environmental stimuli animals encounter on a day-to-day basis are extraordinarily numerous. Olfactory systems have evolved to cope with this diversity by maximizing the chemicals that can be detected, through the amplification of chemosensory receptor gene families, and through combinatorial coding, which expands representation capacity from the number of receptors in the genome to the number of combinations among them. The arthropod mushroom body is a cerebellum-like associative learning structure with a well-understood role in representing sensory stimuli and associating sensory and contextual cues (Farris, 2011; Hige, 2018; Kennedy, 2015). While mushroom bodies of different insect species process information from a variety of sensory modalities, 90% of Kenyon cell inputs in Drosophila melanogaster are olfactory (Zheng et al., 2018). The mushroom body of each hemisphere has ~2000 Kenyon cells (KCs), which are two synapses from the sensory periphery. Each olfactory receptor neuron in the antennae of adult flies expresses one or two of 75 olfactory receptor genes encoded in the genome. The axons of neurons expressing the same receptor converge on one of 54 glomeruli in the antennal lobe. ~150 uniglomerular projection neurons (PNs) have dendrites in one of the 54 glomeruli and carry signals about distinct receptor channels to two regions of the protocerebrum, the lateral horn and the mushroom body calyx (Figure 1A). PN inputs to the lateral horn are thought to underlie innate behaviors, while inputs to the mushroom body allow flexible learned association of odor stimuli with behavioral context (Chin et al., 2018; de Belle and Heisenberg, 1994; Fişek and Wilson, 2014; Jefferis et al., 2007; Ruta et al., 2010).

Figure 1 with 3 supplements see all

Download asset Open asset

In the mushroom body calyx, the presynaptic sites of individual olfactory PNs cluster into multi-synaptic boutons, with PNs of different types (innervating different glomeruli) producing consistent, characteristic bouton numbers (Caron et al., 2013; Zheng et al., 2018). Each PN makes 1–20 boutons, and each bouton is wrapped by claws of ~10 KCs, such that each PN sends output to between 10 and 200 of the 2000 KCs (Leiss et al., 2009). KCs in turn have 3–10 (average of five) claws, which innervate boutons of various PNs (Caron et al., 2013; Gruntman and Turner, 2013; Zheng et al., 2018). Each KC therefore receives innervation from only a minority of the 54 incoming sensory channels, and different individual KCs receive different and relatively unstructured combinations of inputs (Caron et al., 2013; Eichler et al., 2017; Honegger et al., 2011; Murthy et al., 2008; Zheng et al., 2018). The sets of inputs to individual cells vary across hemispheres and likely across individuals (Caron et al., 2013; Eichler et al., 2017; Honegger et al., 2011; Murthy et al., 2008). Associative learning mechanisms operate at KC output synapses, in the mushroom body axonal lobes, to re-weight KC outputs depending on experience and shift animal behavior (Cohn et al., 2015; Handler et al., 2019; Hige et al., 2015; Owald and Waddell, 2015).

The mushroom body is a simplified and experimentally tractable example of an expansion layer, in which a set of sensory inputs is mapped combinatorially onto a much larger set of postsynaptic cells, increasing the dimensionality of sensory representations. Like the diversification of antibodies by V(D)J recombination, the diversification of sensory input combinations across KCs is thought to allow them to represent arbitrary odors, regardless of evolutionary experience. Neurons of many other expansion layers receive similarly few, or sparse, sensory inputs. These include the cerebellum proper, the electric organ of mormyrid fish, the dorsal cochlear nucleus, and the hippocampus (Bell et al., 2008; Keene and Waddell, 2007; Mugnaini et al., 1980). Cerebellar granule cells have an average of four large, claw-shaped dendrites that are innervated by clustered mossy fiber presynaptic sites in mossy fiber rosettes. The similar convergence ratios in the cerebellum and mushroom body (4 or 5 sensory inputs, respectively, per expansion layer cell) are thought to maximize dimensionality of sensory representations by optimizing the tradeoff between stimulus representation, which is maximized when expansion layer neurons receive large combinations of inputs, and stimulus separation, which is maximized when expansion layer neurons receive few inputs (Albus, 1971; Cayco-Gajic et al., 2017; Litwin-Kumar et al., 2017; Marr, 1969). The number of sensory inputs received by expansion layer neurons is thus a crucial parameter in sensory coding. How the density of inputs to expansion layer neurons is developmentally programmed is not understood in any system.

Innervation complexity more generally has been studied in the peripheral nervous system and in the developing mammalian cortex. In peripheral sensory neurons, most prominently those of the Drosophila larval body wall, cell-autonomous mechanisms profoundly influence dendritic complexity (Corty et al., 2016; Jan and Jan, 2010; Ziegler et al., 2017). However, sensory neurons do not need to coordinate their innervation with presynaptic partners. In the vertebrate peripheral nervous system, including the rabbit ciliary ganglion and vertebrate neuromuscular junction, postsynaptic neurons or muscles are thought to dictate wiring complexity (Gibbins et al., 2003; Hume and Purves, 1983; Hume and Purves, 1981; Purves and Hume, 1981; Turney and Lichtman, 2012). In contrast, in the developing cortex, extracellular signals including BDNF play a strong role in influencing dendritic complexity, suggesting that presynaptic cells and glia also influence connectivity density (Kohara et al., 2003; McAllister et al., 1997; McAllister et al., 1995; Purves, 1986). Therefore, while mechanisms acting in both pre- and post-synaptic cells can influence innervation complexity, there is a need to directly compare how pre- and post-synaptic cells influence one another.

We sought to ask how convergence ratio is set in the mushroom body calyx. By bidirectionally varying the populations of pre- and post-synaptic cells, we were able to make many different mushroom body aberrations. Across these conditions, we found a consistent pattern of compensations: the number of claws per KC remained largely constant, while the number of presynaptic boutons per olfactory PN varied bidirectionally and in response to changes of both the PN and KC populations. We therefore conclude that in this circuit, connectivity density is set by aspects of KC specification and is accomplished by flexible innervation of the calyx by PNs.

In order to ask whether presynaptic olfactory PNs or postsynaptic KCs (or both) dictate sparse wiring density in the mushroom body calyx, we used a variety of methods to manipulate the ratio between pre- and post-synaptic cells during embryonic or larval stages, well before adult connections form in the pupa. Through a combination of genetic and pharmacological interventions, we were able to vary both cell populations up and down. In these four conditions, we then examined gross calyx anatomy in adult animals and, when possible, the processes of individual PNs and KCs. These experiments allowed us to ask which population of neurons would adjust its production of processes in response to the perturbation. As we will describe, in all four cases, PNs adjusted their repertoire of presynaptic sites while KC claw number remained similar to that of unmanipulated animals.

Projection neurons reduce bouton number when Kenyon cell number is reduced

First, we sought to reduce the KC population, to ask whether remaining cells would increase their claw number to fully innervate incoming PN boutons, or whether PNs would scale down their boutons to the KC repertoire. To do this, we took advantage of existing pharmacological techniques for KC neuroblast ablation (de Belle and Heisenberg, 1994; Sweeney et al., 2012). In the fly, four KC neuroblasts in each hemisphere produce ~500 KCs each (Ito et al., 1997). While most neuroblasts pause their divisions for the first 8 hr after larval hatching (ALH), the KC neuroblasts continue to divide; if larvae are fed the mitotic poison hydroxyurea (HU) during this time, KC neuroblasts can be specifically ablated (de Belle and Heisenberg, 1994). In these animals, lacking all KCs that receive olfactory inputs, olfactory learning is abrogated while innate olfactory behaviors are spared (de Belle and Heisenberg, 1994). The uniglomerular, excitatory olfactory PNs that innervate the MB calyx are produced by two neuroblasts, called lNB/BAlc and adNB/BAmv3 . Besides the KC neuroblasts, lNB/BAlc is the only other neuroblast in the fly central nervous system dividing from 0 to 8 hr ALH and is thus also susceptible to HU ablation (Das et al., 2013; Ito and Hotta, 1992; Lovick et al., 2016; Lovick and Hartenstein, 2015; Stocker et al., 1997; Sweeney et al., 2012).

While ablation of all eight KC neuroblasts (4/hemisphere) was used to demonstrate the role of KCs in olfactory learning (de Belle and Heisenberg, 1994), we sought to perform more subtle manipulations of the mushroom body cell populations. We generated animals in which PNs and KCs were fluorescently labeled to facilitate scoring, and reduced the concentration and duration of HU application, producing mushroom bodies with a variety of Kenyon cell complements, likely due to sporadic loss of different neuroblast complements in each hemisphere (Figure 1B).

Together, these neuroblast losses could potentially produce 1–2 PN neuroblasts and mushroom bodies derived from 0 to 4 KC neuroblasts. Using our reporters, we scored the presence or absence of PNs ventrolateral to the antennal lobe, derived from lNB/BAlc, and counted the number of KCs (Figure 1C). In practice, some of the neuroblast states did not occur and some were overrepresented (Figure 2—figure supplement 1). Hemispheres from the same brain did not generally show the same neuroblast state as one another, but were often similar in severity of neuroblast losses (Figure 2—figure supplement 1).

KCs of different types are produced in a characteristic sequence in development: γ KCs are born in the embryo and early larva, α’β’ KCs are born in late larvae, and αβ KCs are born in pupae. We identified samples in which all 4 KC neuroblasts had been ablated by looking for samples lacking mushroom body vertical lobes, which are composed of axons of α’β’ and αβ KCs and thus do not form if all KC neuroblasts are killed by HU just after hatching. We confirmed previous findings that in animals with all four KC neuroblasts ablated, the PNs fail to target the calyx and project only to the lateral horn (Stocker et al., 1997). In these animals, ~100 KCs remained (Figure 1B). These represent the KC population that is born prior to larval hatching, and which is thus unaffected by KC neuroblast ablation after hatching. Recently, these earliest-born cells, called γd KCs, have been shown to receive visual and not olfactory input in adult flies (Vogt et al., 2016; Yagi et al., 2016). These observations suggest that γd KCs are molecularly distinct from later-born cells during pupal calyx wiring, such that even when all the olfactory KCs are absent, olfactory PNs cannot innervate visual KCs. Remarkably, these embryonic-born γd KCs receive olfactory inputs in the early larva and are sufficient for olfactory learning at that time (Eichler et al., 2017; Pauls et al., 2010); they must therefore switch their partner preferences across developmental stages.

We then examined calyx anatomy in animals with intermediate populations of KCs, likely derived from 1 to 3 remaining neuroblasts. We found a progressive decrease in the size of the mushroom body calyx in these animals as measured by the maximum cross-sectional area of the calyx (Figure 1B,C) or by calyx volume (Figure 1—figure supplement 1, Figure 1—video 1). To ask if this corresponded to a decline in bouton number, we counted the number of PN boutons in each hemisphere (Figure 1E, Figure 1—figure supplement 2). This revealed a linear relationship between KC number and PN bouton number, suggesting that PNs reduce their population of boutons to match the KC population. The lateral horn appeared normal in these animals and its size did not correlate with KC number, suggesting that PN projections to these two brain areas develop independent of one another (Figure 1B,D). The presence or absence of ventrolateral PNs did not obviously predict bouton number, suggesting individual PNs may be able both to tailor their bouton production to the number of KCs and to adjust to the number of other PNs. However, we were not able to obtain enough samples to draw firm conclusions about whether PN cell number affected PN bouton number in calyces derived from matched KC neuroblast complements. To test this, we developed alternative methods to ablate PNs, described below.

To score KC clones directly, we developed a strategy to label only the latest-born, ‘αβ core’ KCs. The somata and principle neurites of these late-born cells remain clustered in the adult, allowing us to count KC clones by counting groups of labeled soma or axon tracts as they leave the calyx (Figure 2A,B). Again, we found that calyx size tracked KC clone number, and that the presence or absence of the PN progeny of lNB/BAlc did not predict calyx size (Figure 2C). KC loss could reduce PN bouton number because PNs die or fail to be born when lacking KC contacts or trophic support. We thus counted the number of anterodorsal PNs (derived from adNB/BAmv3) and ventrolateral PNs (derived from lNB/BAlc) in this cohort. The number of anterodorsal PNs was not reduced in animals with some or all KC neuroblasts ablated, and the number of ventrolateral PNs was binary rather than graded, suggesting that vlPN number is determined by whether lNB/BAlc succumbed to HU, rather than by the KC complement (Figure 2D,E). These results suggest that PN neurogenesis and survival do not require KC signals. Like γd KCs, a few PNs derived from lNB/BAlc remained in adults; these likely represent PNs born embryonically from lNB/BAlc.

Figure 2 with 3 supplements see all

Download asset Open asset

Reduction in the number of PN boutons in the calyx could occur because individual PNs quantitatively reduce their bouton production. To ask whether individual PNs alter their production of presynaptic boutons as the KC repertoire changes, we identified two methods to label single PNs. We focused on anterodorsal PNs, as these are not susceptible to ablation of lNB/BAlc. First, we used 71D09-Gal4 to label the anterodorsal VM6 PN (Ward et al., 2015) and then subjected labeled animals to HU ablation. Because we found a linear relationship between KC number and calyx area (Figures 1C, 2C), we used calyx area as a proxy for KC number. While VM6 usually produces ~10 boutons, we found only ~5 boutons in animals with reduced calyx size, and 0–1 bouton in animals with severe calyx reductions, likely lacking all olfactory KCs (Figure 2F–H). These results suggest that individual PNs do reduce their bouton production as the KC complement is reduced, and that this reduction is graded rather than binary (i.e. in animals with some but not all KC neuroblasts ablated, bouton number is in between 0 and the wild type number). To characterize additional PNs, we expressed photoactivatable GFP under control of R42D01-Gal4, which labels ~10 anterodorsal PNs innervating the VM3 and VM4 glomeruli. We photoconverted individual somata using two photon microscopy (Figure 2—figure supplement 2). Though bouton counts were somewhat more variable than for VM6 as this strategy labels different types of PNs with different characteristic bouton numbers, we again found a correlation between KC loss and bouton reduction (Figure 2—figure supplement 2).

The graded reductions in bouton number we observed for different PN types would preserve the relative representation of these odor channels in the calyx. We also note that the range of bouton numbers we observed are consistent with adjustments to compensate for the presence or absence of ventrolateral PNs. For example, most VM6 neurons innervating a calyx derived from 1 KC neuroblast produced between 3 and 7 boutons, as compared to the median of 10 in unperturbed animals. If KCs are reduced by ¾ without loss of lNB/BAlc, perfect compensation by PNs would reduce boutons from 10 to 2.5. If lNB/BAlc was also lost, perfect compensation would reduce boutons from 10 to 5. These predicted values are similar to the bouton range we observed experimentally.

Finally, we sought to ask if KC claw number changes as KC number changes. We used GFP photoactivation to label individual KCs in animals subjected to HU ablation. The different KC classes have different median claw numbers: γ KCs have ~7 claws, αβ cells have ~5 claws, and α’β’ KCs have ~3 claws (Caron et al., 2013); we therefore used two different strategies to label individual KCs in animals subjected to HU ablation, both of which allow us to assign KC type. First, we expressed PA-GFP broadly in Kenyon cells using MB247-Gal4, and assigned KC type by imaging the axonal lobe innervation of photoactivated cells (Figure 2—figure supplement 3). Second, we expressed PA-GFP only in αβ or γ KCs using 58F02-Gal4 (Figure 2I) and 89B01-Gal4, respectively. Combining data from these two methods, we found no change in claw number on individual αβ KCs (Figure 2J–K) or γ KCs (Figure 2—figure supplement 3) as the KC population was reduced. From these experiments, we conclude that PNs reduce their bouton production on a cell-by-cell basis as the KC population shrinks, while KC claw number is unaffected by the size of the KC population (model, Figure 2L).

Olfactory projection neurons increase bouton repertoire as Kenyon cell number increases

In order to assess whether PNs can vary their presynaptic bouton number bidirectionally, we next created a method to specifically amplify the KC population. Recent work has suggested that ectopic or supernumerary populations of neurons can integrate into a variety of fruit fly neural circuits (Meng et al., 2019; Pop et al., 2019; Prieto-Godino et al., 2019; Seroka and Doe, 2019; Shaw et al., 2018), and previous studies identified a mutant, called mushroom body defect (mud) in which the mushroom body and other brain areas are dramatically expanded (Guan et al., 2000; Prokop and Technau, 1994). Mud/NuMA is a spindle orientation protein that ensures neuroblasts divide asymmetrically to produce one neuroblast and one differentiating neuron or neural progenitor (Siller et al., 2006). In mud mutants, neuroblasts occasionally divide symmetrically to produce two neuroblasts, amplifying the progeny of that neuroblast. To restrict amplification to KCs, we drove UAS-mud-RNAi (BL 35044) using OK107-Gal4, which is expressed in KC neuroblasts as well as in KCs (Liu et al., 2015). Importantly, in our RNAseq data from 45 hr APF and adult animals, we observe no expression of mud in PNs, KCs, or other brain cells, suggesting mud is not expressed in differentiating neurons (EJ Clowney, unpublished).

As OK107 labels mature KCs as well as KC neuroblasts, we included UAS-CD8GFP in these animals to count KCs. We observed potent amplification of KCs that was variable across animals, sometimes more than doubling KC number (Figure 3A,B). We measured calyx size, which expanded dramatically, and used GH146-driven fluorescence to count PN bouton number, which doubled in mushroom bodies with the largest KC expansions (Figure 3C, Figure 3—video 1). These results suggest that just as PNs scale down their boutons if the KC population is reduced, PNs scale up their boutons when the KC population is expanded.

To test whether PNs increased boutons on a cell-by-cell basis in these animals, we labeled VM6 using 71D09-LexA, and PNs innervating DA1, VA1d, and DC3 using MZ19-QF. First, we observed no change in VM6 or MZ19⁺ cell numbers in brains with KC expansions (Figure 3E,H), suggesting that PN numbers do not increase as KC number increases. We observed no change in VM6 bouton numbers, perhaps because the VM6 PN already has a relatively large number of boutons (Figure 3H). In contrast, MZ19 cells increased their number of presynaptic boutons as the KC population expanded, from an average of 3.5 boutons per cells in the control genotype to an average of 4.75 boutons per cell in the KC-expansion genotype. In brains with measurably enlarged calyces, indicative of KC population expansion, MZ19⁺ boutons doubled (Figure 3D,F,G).

Figure 3 with 2 supplements see all

Download asset Open asset

To rule out reductions in the number of claws per KC as the KC population expands, we expressed photoactivatable GFP under the control of OK107 in animals with UAS-mud-RNAi or control and subjected individual or small groups of KCs to photoactivation. We scored the number of claws per labeled KC in each calyx. As above, calyx size predicted KC number in these animals, so we used calyx size as a proxy for KC number. Rather than a decrease in claw number per KC in animals with expanded KC complements, we found a slight increase in claw number, from a median of 5 to a median of 6 claws. As different types of KCs produce different characteristic numbers of claws, this could represent a change in claw production, a change in the distribution of KC types in these animals, or bias in which KC types we labeled in expanded versus control brains. While we were unable to identify LexA lines to specifically label αβ or γ KCs in OK107>mudRNAi brains, we found a LexA line, R41C07, that labeled α’β’ KCs, and targeted these specifically for photoactivation (Figure 3—figure supplement 1). In these KCs of matched type, we found no change in KC claw number per cell, though we cannot rule out subtle differences. Together, these experiments suggest that in animals with an amplified KC complement, PNs dramatically increase their bouton production to match the KC population, while KCs may moderately increase, and do not reduce, their individual claw production.

Individual projection neurons decrease bouton number as the projection neuron population expands

We have shown that olfactory PNs bidirectionally vary their presynaptic boutons to suit populations of KCs derived from 0 to 8 KC neuroblasts. To ask whether PNs can also vary their bouton production in response to changes in their own numbers, or in response to changes in the ratio between PNs and KCs that do not affect KC clone number, we identified methods to vary the PN population. We found that just as we could expand the KC population by knocking down mud in KC neuroblasts, we could expand the PN population by knocking down mud in PN neuroblasts. This strategy produced GH146-labeled PN complements of up to 600 cells, as compared to ~120 GH146-labeled PNs we counted in controls. While the PN neuroblast Gal4 we used, 44F03, can label both lNB/BAlc and adNB/BAmv3 progeny (Awasaki et al., 2014), we found that ventrolateral PNs were much more susceptible to expansion than anterodorsal PNs.

Supernumerary PNs innervated the antennal lobe, causing a massive expansion of antennal lobe area (Figure 4A,C). The normal glomerular map in the AL appeared distorted, but glomerular divisions were still present. These PNs also projected to the calyx, observed by thickening of the PN axon tract crossing the calyx (red, Figure 4D), and increased the density of PN innervation of the lateral horn (Figure 4D), though not lateral horn area (Figure 4—figure supplement 1). However, there was no change in calyx area or total bouton number in these animals compared to wild type (Figure 4E–H). At a population level, the average number of boutons per PN was half that in controls (Figure 4I).

Figure 4 with 1 supplement see all

Download asset Open asset

To ask how small groups of PNs adjusted to PN expansion, we again used MZ19-QF, which labels 15–20 PNs in controls (Figure 4J,K). We counted MZ19⁺ cells and boutons in control versus animals of the expansion genotype (Figure 4K). As the expansion was highly variable and we rarely observed more than 25 MZ19⁺ cells in controls, we divided the expanded genotype into hemispheres with <25 PNs or ≥25 PNs. There was a modest but insignificant increase in MZ19⁺ boutons in the ≥25 PNs groups (Figure 4L), attributable to a strong and significant reduction of average number of boutons per PN (Figure 4M). Thus PNs scale down their individual bouton repertoire to compensate for increases in the size of the PN population. Like MZ19 PNs, the VM6 PN also maintained a stable number of total boutons as the PN population expanded (Figure 4—figure supplement 1). This was possible because the VM6 cell(s) had similar numbers of total boutons as the VM6 population expanded.

Projection neuron bouton expansion partially compensates for severe reductions in the projection neuron population

In order to reduce the PN repertoire independent of KC number, we identified two Gal4 lines, VT033006 and VT033008, that label large complements of uniglomerular PNs in the adult (Figure 5A). We drove diphtheria toxin A (Berdnik et al., 2006) and fluorescent reporters under control of these lines in animals where additional PNs were labeled by GH146-LexA (Figure 5B,G). In each case, DTA eliminated all Gal4⁺ cells as well as additional GH146⁺ PNs, suggesting that these lines drive DTA expression in broader PN populations at earlier stages of development. Driving DTA under control of VT033008 reduced the total PNs labeled by VT033008 and/or GH146 by nearly half (Figure 5C), while VT033006>DTA reduced PNs labeled by VT03006 and/or GH146 by >90% (Figure 5H). PN labeling by VT033006 and VT033008, and cell loss when used to drive DTA, were already evident in third instar larvae, suggesting PNs are lost well before pupal calyx wiring (Figure 5—figure supplement 1). While PN dendrites have been shown to occupy glomerular territories in pupal stages before OSNs innervate the antennal lobe (Jefferis et al., 2004), we found that antennal lobes lacking the vast majority of uniglomerular PNs retained glomerular divisions. This suggests that these PNs may not be required for the formation of glomerular divisions in the adult antennal lobe.

Figure 5 with 2 supplements see all

Download asset Open asset

We measured maximum calyx cross-sectional area and bouton number across these conditions and found that in each case, the effect on the calyx was measurable but much less severe than the reduction in PN number. For VT033008>DTA, loss of nearly half the PNs yielded no change in calyx area and a 20% reduction in overall bouton number (Figure 5D–F). Remarkably, while we estimate that only ~10 PNs remain in VT033006>DTA animals (~7% the normal complement, Figure 5H), the number of boutons in the calyx was reduced by only half (Figure 5H,L). This suggests that individual PNs could have expanded their bouton production by as much as 6-fold. While remaining PNs sometimes appeared to make exuberant collections of boutons (e.g. Figure 6D), distortion of the antennal lobes in these animals prevented us from assigning identities to these spared PNs and comparing bouton numbers to matched individual PNs in controls.

Figure 6

Download asset Open asset

This severe loss of inputs could lead to KC death. We therefore labeled and counted KCs in VT033006>DTA animals with severe PN loss. While we cannot rule out a small reduction in KC complement, at least 80% of KCs remained in these brains. We also subjected control and VT033006>DTA brains to flow cytometry and found that PN ablation brains had at least 85% as many KCs as did control brains (Figure 5—figure supplement 2).

Developmental plasticity preserves sparse odor coding despite perturbations to cell populations

The disparate and numerous strategies we used above to vary the ratio between PNs and KCs all produced the same result, that PNs adjusted to changes in PN::KC ratio by altering their presynaptic boutons, while the distribution of KC claw numbers changed little even when perturbations to PNs were so severe as to prevent production of the normal bouton complement. This finding suggests that developmental mechanisms prioritize the sparseness of olfactory inputs to KCs, and might therefore preserve sparse odor coding and high-dimensional odor representations despite stark changes to the underlying circuit constituents. We therefore sought to ask how KC odor responses were affected by perturbations to the PN::KC ratio.

We used OK107 to drive GCaMP6s expression in all KCs, and subjected animals to HU just after larval hatching. We then imaged KC somatic odor responses in vivo in 24 adult HU-treated and 14 adult sham-treated animals. We observed robust odor responses in both conditions, with 8/28 sham-treated and 4/46 HU-treated calyces completely unresponsive. To identify HU-treated hemispheres with reduced KC complement, following functional imaging we imaged the anatomy of the calyx by two-photon or by immunostaining and confocal. Among ablated animals, we identified 14 of 46 imaged hemispheres in which calyces were clearly still present but of reduced size, which we designate as ‘reduced-KC’ calyces. We subjected these datasets and 17 responsive controls to motion correction using Suite2p (Pachitariu et al., 2017). After motion correction, seven sham and seven reduced KC calyces were sufficiently still to allow us to define ROIs for individual somata. For each cell, we calculated peak change in fluorescence following odor stimulus (Figure 7—source data 1). Example responses for sham-treated and reduced-KC calyces are shown in Figure 7A, C, Figure 7—figure supplement 1, and in Figure 7—video 1, 2.

Figure 7 with 3 supplements see all

Download asset Open asset

Figure 7—source data 1

Peak response data for all animals in Figure 7.

https://cdn.elifesciences.org/articles/52278/elife-52278-fig7-data1-v2.xlsx

Download elife-52278-fig7-data1-v2.xlsx

In order to compare response magnitudes across conditions, we first pooled all cells from each condition and compared overall Δf/f in response to each odor (Figure 7D). The distributions were statistically indistinguishable between conditions for mineral oil (mechanosensory control), ethyl acetate, and isobutyl acetate, and significantly different for benzaldehyde and methylcyclohexanol. Using these aggregate responses, we set a threshold for ‘responsive’ cells of 20% increase in fluorescence over baseline, as we observed bifurcation of the cellwise response distribution at this cutoff. Using these criteria, a median of ~50% of cells from sham calyces responded to ethyl acetate and isobutyl acetate, and ~20–30% of cells to benzaldehyde and methylcyclohexanol (Figure 7E). These response rates are higher than has been reported with GCaMP3 and GCaMP1 (Honegger et al., 2011; Wang et al., 2004), but correspond with spiking response rates observed electrophysiologically (Murthy et al., 2008). We observed variation in odor responses in both control and reduced-KC calyces; interindividual variation is expected given the stochastic innervation of KCs by PNs.

We next compared the proportions of KCs per calyx responding to each odor between sham and reduced-KC calyces. These distributions were statistically indistinguishable (Figure 7E). Finally, we asked how many odors each cell responded to (Figure 7F). In each condition, ~60% of cells responded to 0 or one odors. Together, these findings suggest that the developmental plasticity mechanisms operating to set the density of olfactory inputs to KCs preserve qualitatively sparse KC odor responses when the KC population is experimentally reduced. Further experiments will be required to assess quantitative effects of KC reduction on responses to odors that typically stimulate fewer cells, like benzaldehyde and methylcyclohexanol.

Cerebellum-like architectures are found repeatedly in phylogenetically distinct organisms, suggesting that they have broad evolutionary utility. Yet the developmental production of sparseness, a key wiring feature allowing high-dimensional sensory representations, is not understood. Here, we have begun to investigate the development of sparse connectivity in the cerebellum-like associative learning center of insects, the mushroom body. By varying the ratio between presynaptic olfactory PNs and postsynaptic KCs, we find that connectivity density is set by KCs: KC claw number changes little across a wide range of PN::KC ratios, KC number predicts PN bouton number, and PNs exhibit wide variety in bouton number to satisfy KC demands. As described below, we expect that this strategy for generating connectivity density would preserve sparseness in KC inputs across developmental time and upon evolutionary changes in cell repertoires and thus maintain the olfactory coding function of the mushroom body over a range of conditions.

Analyzed data points generated during the study are included in the figures. Source data for Figure 7 is presented in a source data file.

1. 1. Akerboom J
   2. Chen TW
   3. Wardill TJ
   4. Tian L
   5. Marvin JS
   6. Mutlu S
   7. Calderón NC
   8. Esposti F
   9. Borghuis BG
   10. Sun XR
   11. Gordus A
   12. Orger MB
   13. Portugues R
   14. Engert F
   15. Macklin JJ
   16. Filosa A
   17. Aggarwal A
   18. Kerr RA
   19. Takagi R
   20. Kracun S
   21. Shigetomi E
   22. Khakh BS
   23. Baier H
   24. Lagnado L
   25. Wang SS
   26. Bargmann CI
   27. Kimmel BE
   28. Jayaraman V
   29. Svoboda K
   30. Kim DS
   31. Schreiter ER
   32. Looger LL

   (2012) Optimization of a GCaMP calcium Indicator for neural activity imaging

   Journal of Neuroscience 32:13819–13840.

   https://doi.org/10.1523/JNEUROSCI.2601-12.2012

   • PubMed
   • Google Scholar
2. 1. Albus JS

   (1971) A theory of cerebellar function

   Mathematical Biosciences 10:25–61.

   https://doi.org/10.1016/0025-5564(71)90051-4

   • Google Scholar
3. 1. Aso Y
   2. Hattori D
   3. Yu Y
   4. Johnston RM
   5. Iyer NA
   6. Ngo TT
   7. Dionne H
   8. Abbott LF
   9. Axel R
   10. Tanimoto H
   11. Rubin GM

   (2014) The neuronal architecture of the mushroom body provides a logic for associative learning

   eLife 3:e04577.

   https://doi.org/10.7554/eLife.04577

   • PubMed
   • Google Scholar
4. 1. Awasaki T
   2. Kao CF
   3. Lee YJ
   4. Yang CP
   5. Huang Y
   6. Pfeiffer BD
   7. Luan H
   8. Jing X
   9. Huang YF
   10. He Y
   11. Schroeder MD
   12. Kuzin A
   13. Brody T
   14. Zugates CT
   15. Odenwald WF
   16. Lee T

   (2014) Making Drosophila lineage-restricted drivers via patterned recombination in neuroblasts

   Nature Neuroscience 17:631–637.

   https://doi.org/10.1038/nn.3654

   • PubMed
   • Google Scholar
5. 1. Bell CC
   2. Han V
   3. Sawtell NB

   (2008) Cerebellum-like structures and their implications for cerebellar function

   Annual Review of Neuroscience 31:1–24.

   https://doi.org/10.1146/annurev.neuro.30.051606.094225

   • PubMed
   • Google Scholar
6. 1. Berdnik D
   2. Chihara T
   3. Couto A
   4. Luo L

   (2006) Wiring stability of the adult Drosophila olfactory circuit after lesion

   Journal of Neuroscience 26:3367–3376.

   https://doi.org/10.1523/JNEUROSCI.4941-05.2006

   • PubMed
   • Google Scholar
7. 1. Caron SJ
   2. Ruta V
   3. Abbott LF
   4. Axel R

   (2013) Random convergence of olfactory inputs in the Drosophila mushroom body

   Nature 497:113–117.

   https://doi.org/10.1038/nature12063

   • PubMed
   • Google Scholar
8. 1. Cayco-Gajic NA
   2. Clopath C
   3. Silver RA

   (2017) Sparse synaptic connectivity is required for decorrelation and pattern separation in feedforward networks

   Nature Communications 8:1116.

   https://doi.org/10.1038/s41467-017-01109-y

   • PubMed
   • Google Scholar
9. 1. Chen Y
   2. Akin O
   3. Nern A
   4. Tsui CY
   5. Pecot MY
   6. Zipursky SL

   (2014) Cell-type-specific labeling of synapses in vivo through synaptic tagging with recombination

   Neuron 81:280–293.

   https://doi.org/10.1016/j.neuron.2013.12.021

   • PubMed
   • Google Scholar
10. 1. Chin SG
    2. Maguire SE
    3. Huoviala P
    4. Jefferis G
    5. Potter CJ

    (2018) Olfactory neurons and brain centers directing oviposition decisions in Drosophila

    Cell Reports 24:1667–1678.

    https://doi.org/10.1016/j.celrep.2018.07.018

    • PubMed
    • Google Scholar
11. 1. Clowney EJ
    2. Iguchi S
    3. Bussell JJ
    4. Scheer E
    5. Ruta V

    (2015) Multimodal chemosensory circuits controlling male courtship in Drosophila

    Neuron 87:1036–1049.

    https://doi.org/10.1016/j.neuron.2015.07.025

    • PubMed
    • Google Scholar
12. 1. Cohn R
    2. Morantte I
    3. Ruta V

    (2015) Coordinated and compartmentalized neuromodulation shapes sensory processing in Drosophila

    Cell 163:1742–1755.

    https://doi.org/10.1016/j.cell.2015.11.019

    • PubMed
    • Google Scholar
13. 1. Connolly JB
    2. Roberts IJ
    3. Armstrong JD
    4. Kaiser K
    5. Forte M
    6. Tully T
    7. O'Kane CJ

    (1996) Associative learning disrupted by impaired gs signaling in Drosophila mushroom bodies

    Science 274:2104–2107.

    https://doi.org/10.1126/science.274.5295.2104

    • PubMed
    • Google Scholar
14. 1. Corty MM
    2. Tam J
    3. Grueber WB

    (2016) Dendritic diversification through transcription factor-mediated suppression of alternative morphologies

    Development 143:1351–1362.

    https://doi.org/10.1242/dev.130906

    • PubMed
    • Google Scholar
15. 1. Das A
    2. Gupta T
    3. Davla S
    4. Prieto-Godino LL
    5. Diegelmann S
    6. Reddy OV
    7. Raghavan KV
    8. Reichert H
    9. Lovick J
    10. Hartenstein V

    (2013) Neuroblast lineage-specific origin of the neurons of the Drosophila larval olfactory system

    Developmental Biology 373:322–337.

    https://doi.org/10.1016/j.ydbio.2012.11.003

    • PubMed
    • Google Scholar
16. 1. de Belle JS
    2. Heisenberg M

    (1994) Associative odor learning in Drosophila abolished by chemical ablation of mushroom bodies

    Science 263:692–695.

    https://doi.org/10.1126/science.8303280

    • PubMed
    • Google Scholar
17. 1. Eichler K
    2. Li F
    3. Litwin-Kumar A
    4. Park Y
    5. Andrade I
    6. Schneider-Mizell CM
    7. Saumweber T
    8. Huser A
    9. Eschbach C
    10. Gerber B
    11. Fetter RD
    12. Truman JW
    13. Priebe CE
    14. Abbott LF
    15. Thum AS
    16. Zlatic M
    17. Cardona A

    (2017) The complete connectome of a learning and memory centre in an insect brain

    Nature 548:175–182.

    https://doi.org/10.1038/nature23455

    • PubMed
    • Google Scholar
18. 1. Fan H
    2. Favero M
    3. Vogel MW

    (2001) Elimination of bax expression in mice increases cerebellar purkinje cell numbers but not the number of granule cells

    The Journal of Comparative Neurology 436:82–91.

    https://doi.org/10.1002/cne.1055

    • PubMed
    • Google Scholar
19. 1. Farris SM

    (2011) Are mushroom bodies cerebellum-like structures?

    Arthropod Structure & Development 40:368–379.

    https://doi.org/10.1016/j.asd.2011.02.004

    • PubMed
    • Google Scholar
20. 1. Fişek M
    2. Wilson RI

    (2014) Stereotyped connectivity and computations in higher-order olfactory neurons

    Nature Neuroscience 17:280–288.

    https://doi.org/10.1038/nn.3613

    • PubMed
    • Google Scholar
21. 1. Frank DD
    2. Jouandet GC
    3. Kearney PJ
    4. Macpherson LJ
    5. Gallio M

    (2015) Temperature representation in the Drosophila brain

    Nature 519:358–361.

    https://doi.org/10.1038/nature14284

    • PubMed
    • Google Scholar
22. 1. Gibbins IL
    2. Teo EH
    3. Jobling P
    4. Morris JL

    (2003) Synaptic density, convergence, and dendritic complexity of prevertebral sympathetic neurons

    The Journal of Comparative Neurology 455:285–298.

    https://doi.org/10.1002/cne.10404

    • PubMed
    • Google Scholar
23. 1. Goedhart J

    (2018a) Dispense with redundant P values

    Nature 554:31.

    https://doi.org/10.1038/d41586-018-01314-9

    • PubMed
    • Google Scholar
24. Website

    1. Goedhart J

    (2018b) Prevent p-value parroting

    The Node. Accessed September 20, 2019.

    https://thenode.biologists.com/p-value-parroting/research/
25. 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
26. 1. Gruntman E
    2. Turner GC

    (2013) Integration of the olfactory code across dendritic claws of single mushroom body neurons

    Nature Neuroscience 16:1821–1829.

    https://doi.org/10.1038/nn.3547

    • PubMed
    • Google Scholar
27. 1. Guan Z
    2. Prado A
    3. Melzig J
    4. Heisenberg M
    5. Nash HA
    6. Raabe T

    (2000) Mushroom body defect, a gene involved in the control of neuroblast proliferation in Drosophila, encodes a coiled-coil protein

    PNAS 97:8122–8127.

    https://doi.org/10.1073/pnas.97.14.8122

    • PubMed
    • Google Scholar
28. 1. Han DD
    2. Stein D
    3. Stevens LM

    (2000)

    Investigating the function of follicular subpopulations during Drosophila oogenesis through hormone-dependent enhancer-targeted cell ablation

    Development 127:573–583.

    • PubMed
    • Google Scholar
29. 1. Handler A
    2. Graham TGW
    3. Cohn R
    4. Morantte I
    5. Siliciano AF
    6. Zeng J
    7. Li Y
    8. Ruta V

    (2019) Distinct dopamine receptor pathways underlie the temporal sensitivity of associative learning

    Cell 178:60–75.

    https://doi.org/10.1016/j.cell.2019.05.040

    • PubMed
    • Google Scholar
30. 1. Hige T
    2. Aso Y
    3. Rubin GM
    4. Turner GC

    (2015) Plasticity-driven individualization of olfactory coding in mushroom body output neurons

    Nature 526:258–262.

    https://doi.org/10.1038/nature15396

    • PubMed
    • Google Scholar
31. 1. Hige T

    (2018) What can tiny mushrooms in fruit flies tell Us about learning and memory?

    Neuroscience Research 129:8–16.

    https://doi.org/10.1016/j.neures.2017.05.002

    • PubMed
    • Google Scholar
32. 1. Ho J
    2. Tumkaya T
    3. Aryal S
    4. Choi H
    5. Claridge-Chang A

    (2019) Moving beyond P values: data analysis with estimation graphics

    Nature Methods 16:565–566.

    https://doi.org/10.1038/s41592-019-0470-3

    • PubMed
    • Google Scholar
33. 1. Honegger KS
    2. Campbell RA
    3. Turner GC

    (2011) Cellular-resolution population imaging reveals robust sparse coding in the Drosophila mushroom body

    Journal of Neuroscience 31:11772–11785.

    https://doi.org/10.1523/JNEUROSCI.1099-11.2011

    • PubMed
    • Google Scholar
34. 1. Hong W
    2. Mosca TJ
    3. Luo L

    (2012) Teneurins instruct synaptic partner matching in an olfactory map

    Nature 484:201–207.

    https://doi.org/10.1038/nature10926

    • PubMed
    • Google Scholar
35. 1. Hume RI
    2. Purves D

    (1981) Geometry of neonatal neurones and the regulation of synapse elimination

    Nature 293:469–471.

    https://doi.org/10.1038/293469a0

    • PubMed
    • Google Scholar
36. 1. Hume RI
    2. Purves D

    (1983) Apportionment of the terminals from single preganglionic axons to target neurones in the rabbit ciliary ganglion

    The Journal of Physiology 338:259–275.

    https://doi.org/10.1113/jphysiol.1983.sp014672

    • PubMed
    • Google Scholar
37. 1. Ito K
    2. Awano W
    3. Suzuki K
    4. Hiromi Y
    5. Yamamoto D

    (1997)

    The Drosophila mushroom body is a quadruple structure of clonal units each of which contains a virtually identical set of neurones and glial cells

    Development 124:761–771.

    • PubMed
    • Google Scholar
38. 1. Ito K
    2. Hotta Y

    (1992) Proliferation pattern of postembryonic neuroblasts in the brain of Drosophila Melanogaster

    Developmental Biology 149:134–148.

    https://doi.org/10.1016/0012-1606(92)90270-Q

    • PubMed
    • Google Scholar
39. 1. Jan YN
    2. Jan LY

    (2010) Branching out: mechanisms of dendritic arborization

    Nature Reviews Neuroscience 11:316–328.

    https://doi.org/10.1038/nrn2836

    • PubMed
    • Google Scholar
40. 1. Jefferis GS
    2. Vyas RM
    3. Berdnik D
    4. Ramaekers A
    5. Stocker RF
    6. Tanaka NK
    7. Ito K
    8. Luo L

    (2004) Developmental origin of wiring specificity in the olfactory system of Drosophila

    Development 131:117–130.

    https://doi.org/10.1242/dev.00896

    • PubMed
    • Google Scholar
41. 1. Jefferis GS
    2. Potter CJ
    3. Chan AM
    4. Marin EC
    5. Rohlfing T
    6. Maurer CR
    7. Luo L

    (2007) Comprehensive maps of Drosophila higher olfactory centers: spatially segregated fruit and pheromone representation

    Cell 128:1187–1203.

    https://doi.org/10.1016/j.cell.2007.01.040

    • PubMed
    • Google Scholar
42. 1. Jenett A
    2. Rubin GM
    3. Ngo TT
    4. Shepherd D
    5. Murphy C
    6. Dionne H
    7. Pfeiffer BD
    8. Cavallaro A
    9. Hall D
    10. Jeter J
    11. Iyer N
    12. Fetter D
    13. Hausenfluck JH
    14. Peng H
    15. Trautman ET
    16. Svirskas RR
    17. Myers EW
    18. Iwinski ZR
    19. Aso Y
    20. DePasquale GM
    21. Enos A
    22. Hulamm P
    23. Lam SC
    24. Li HH
    25. Laverty TR
    26. Long F
    27. Qu L
    28. Murphy SD
    29. Rokicki K
    30. Safford T
    31. Shaw K
    32. Simpson JH
    33. Sowell A
    34. Tae S
    35. Yu Y
    36. Zugates CT

    (2012) A GAL4-driver line resource for Drosophila neurobiology

    Cell Reports 2:991–1001.

    https://doi.org/10.1016/j.celrep.2012.09.011

    • PubMed
    • Google Scholar
43. 1. Keene AC
    2. Waddell S

    (2007) Drosophila olfactory memory: single genes to complex neural circuits

    Nature Reviews Neuroscience 8:341–354.

    https://doi.org/10.1038/nrn2098

    • PubMed
    • Google Scholar
44. Book

    1. Kennedy A

    (2015)

    Representation and Learning in Cerebellum-Like Structures

    Columbia University.

    • Google Scholar
45. Preprint

    1. Kennedy A

    (2019) Learning with naturalistic odor representations in a dynamic model of the Drosophila olfactory system

    bioRxiv.

    https://doi.org/10.1101/783191

    • Google Scholar
46. 1. Kirkhart C
    2. Scott K

    (2015) Gustatory learning and processing in the Drosophila mushroom bodies

    Journal of Neuroscience 35:5950–5958.

    https://doi.org/10.1523/JNEUROSCI.3930-14.2015

    • PubMed
    • Google Scholar
47. 1. Kohara K
    2. Kitamura A
    3. Adachi N
    4. Nishida M
    5. Itami C
    6. Nakamura S
    7. Tsumoto T

    (2003) Inhibitory but not excitatory cortical neurons require presynaptic brain-derived neurotrophic factor for dendritic development, as revealed by chimera cell culture

    The Journal of Neuroscience 23:6123–6131.

    https://doi.org/10.1523/JNEUROSCI.23-14-06123.2003

    • PubMed
    • Google Scholar
48. 1. Kvon EZ
    2. Kazmar T
    3. Stampfel G
    4. Yáñez-Cuna JO
    5. Pagani M
    6. Schernhuber K
    7. Dickson BJ
    8. Stark A

    (2014) Genome-scale functional characterization of Drosophila developmental enhancers in vivo

    Nature 512:91–95.

    https://doi.org/10.1038/nature13395

    • PubMed
    • Google Scholar
49. 1. Lai SL
    2. Awasaki T
    3. Ito K
    4. Lee T

    (2008) Clonal analysis of Drosophila antennal lobe neurons: diverse neuronal architectures in the lateral neuroblast lineage

    Development 135:2883–2893.

    https://doi.org/10.1242/dev.024380

    • PubMed
    • Google Scholar
50. 1. Lai SL
    2. Lee T

    (2006) Genetic mosaic with dual binary transcriptional systems in Drosophila

    Nature Neuroscience 9:703–709.

    https://doi.org/10.1038/nn1681

    • PubMed
    • Google Scholar
51. 1. Lee T
    2. Lee A
    3. Luo L

    (1999)

    Development of the Drosophila mushroom bodies: sequential generation of three distinct types of neurons from a neuroblast

    Development 126:4065–4076.

    • PubMed
    • Google Scholar
52. 1. Lee T
    2. Luo L

    (1999) Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis

    Neuron 22:451–461.

    https://doi.org/10.1016/S0896-6273(00)80701-1

    • PubMed
    • Google Scholar
53. 1. Leiss F
    2. Groh C
    3. Butcher NJ
    4. Meinertzhagen IA
    5. Tavosanis G

    (2009) Synaptic organization in the adult Drosophila mushroom body Calyx

    The Journal of Comparative Neurology 517:808–824.

    https://doi.org/10.1002/cne.22184

    • PubMed
    • Google Scholar
54. 1. Lin AC
    2. Bygrave AM
    3. de Calignon A
    4. Lee T
    5. Miesenböck G

    (2014) Sparse, decorrelated odor coding in the mushroom body enhances learned odor discrimination

    Nature Neuroscience 17:559–568.

    https://doi.org/10.1038/nn.3660

    • PubMed
    • Google Scholar
55. 1. Litwin-Kumar A
    2. Harris KD
    3. Axel R
    4. Sompolinsky H
    5. Abbott LF

    (2017) Optimal degrees of synaptic connectivity

    Neuron 93:1153–1164.

    https://doi.org/10.1016/j.neuron.2017.01.030

    • PubMed
    • Google Scholar
56. 1. Liu Z
    2. Yang CP
    3. Sugino K
    4. Fu CC
    5. Liu LY
    6. Yao X
    7. Lee LP
    8. Lee T

    (2015) Opposing intrinsic temporal gradients guide neural stem cell production of varied neuronal fates

    Science 350:317–320.

    https://doi.org/10.1126/science.aad1886

    • PubMed
    • Google Scholar
57. 1. Lovick JK
    2. Kong A
    3. Omoto JJ
    4. Ngo KT
    5. Younossi-Hartenstein A
    6. Hartenstein V

    (2016) Patterns of growth and tract formation during the early development of secondary lineages in the Drosophila larval brain

    Developmental Neurobiology 76:434–451.

    https://doi.org/10.1002/dneu.22325

    • PubMed
    • Google Scholar
58. 1. Lovick JK
    2. Hartenstein V

    (2015) Hydroxyurea-mediated neuroblast ablation establishes birth dates of secondary lineages and addresses neuronal interactions in the developing Drosophila brain

    Developmental Biology 402:32–47.

    https://doi.org/10.1016/j.ydbio.2015.03.005

    • PubMed
    • Google Scholar
59. 1. Marin EC
    2. Watts RJ
    3. Tanaka NK
    4. Ito K
    5. Luo L

    (2005) Developmentally programmed remodeling of the Drosophila olfactory circuit

    Development 132:725–737.

    https://doi.org/10.1242/dev.01614

    • PubMed
    • Google Scholar
60. 1. Marr D

    (1969) A theory of cerebellar cortex

    The Journal of Physiology 202:437–470.

    https://doi.org/10.1113/jphysiol.1969.sp008820

    • PubMed
    • Google Scholar
61. 1. Masuda-Nakagawa LM
    2. Tanaka NK
    3. O'Kane CJ

    (2005) Stereotypic and random patterns of connectivity in the larval mushroom body Calyx of Drosophila

    PNAS 102:19027–19032.

    https://doi.org/10.1073/pnas.0509643102

    • PubMed
    • Google Scholar
62. 1. McAllister AK
    2. Lo DC
    3. Katz LC

    (1995) Neurotrophins regulate dendritic growth in developing visual cortex

    Neuron 15:791–803.

    https://doi.org/10.1016/0896-6273(95)90171-X

    • PubMed
    • Google Scholar
63. 1. McAllister AK
    2. Katz LC
    3. Lo DC

    (1997) Opposing roles for endogenous BDNF and NT-3 in regulating cortical dendritic growth

    Neuron 18:767–778.

    https://doi.org/10.1016/S0896-6273(00)80316-5

    • PubMed
    • Google Scholar
64. 1. Meng JL
    2. Marshall ZD
    3. Lobb-Rabe M
    4. Heckscher ES

    (2019) How prolonged expression of hunchback, a temporal transcription factor, re-wires locomotor circuits

    eLife 8:e46089.

    https://doi.org/10.7554/eLife.46089

    • PubMed
    • Google Scholar
65. 1. Mugnaini E
    2. Osen KK
    3. Dahl AL
    4. Friedrich VL
    5. Korte G

    (1980) Fine structure of granule cells and related interneurons (termed golgi cells) in the cochlear nuclear complex of cat, rat and mouse

    Journal of Neurocytology 9:537–570.

    https://doi.org/10.1007/BF01204841

    • PubMed
    • Google Scholar
66. 1. Murthy M
    2. Fiete I
    3. Laurent G

    (2008) Testing odor response stereotypy in the Drosophila mushroom body

    Neuron 59:1009–1023.

    https://doi.org/10.1016/j.neuron.2008.07.040

    • PubMed
    • Google Scholar
67. 1. Muthukumar AK
    2. Stork T
    3. Freeman MR

    (2014) Activity-dependent regulation of astrocyte GAT levels during synaptogenesis

    Nature Neuroscience 17:1340–1350.

    https://doi.org/10.1038/nn.3791

    • PubMed
    • Google Scholar
68. 1. Owald D
    2. Waddell S

    (2015) Olfactory learning skews mushroom body output pathways to steer behavioral choice in Drosophila

    Current Opinion in Neurobiology 35:178–184.

    https://doi.org/10.1016/j.conb.2015.10.002

    • PubMed
    • Google Scholar
69. Preprint

    1. Pachitariu M
    2. Stringer C
    3. Dipoppa M
    4. Schröder S RLF
    5. Dalgleish H
    6. Carandini M
    7. Harris KD

    (2017) Suite2p: beyond 10,000 neurons with standard two-photon microscopy

    bioRxiv.

    https://doi.org/10.1101/061507

    • Google Scholar
70. 1. Pauls D
    2. Selcho M
    3. Gendre N
    4. Stocker RF
    5. Thum AS

    (2010) Drosophila larvae establish appetitive olfactory memories via mushroom body neurons of embryonic origin

    Journal of Neuroscience 30:10655–10666.

    https://doi.org/10.1523/JNEUROSCI.1281-10.2010

    • PubMed
    • Google Scholar
71. 1. Perkins LA
    2. Holderbaum L
    3. Tao R
    4. Hu Y
    5. Sopko R
    6. McCall K
    7. Yang-Zhou D
    8. Flockhart I
    9. Binari R
    10. Shim HS
    11. Miller A
    12. Housden A
    13. Foos M
    14. Randkelv S
    15. Kelley C
    16. Namgyal P
    17. Villalta C
    18. Liu LP
    19. Jiang X
    20. Huan-Huan Q
    21. Wang X
    22. Fujiyama A
    23. Toyoda A
    24. Ayers K
    25. Blum A
    26. Czech B
    27. Neumuller R
    28. Yan D
    29. Cavallaro A
    30. Hibbard K
    31. Hall D
    32. Cooley L
    33. Hannon GJ
    34. Lehmann R
    35. Parks A
    36. Mohr SE
    37. Ueda R
    38. Kondo S
    39. Ni JQ
    40. Perrimon N

    (2015) The transgenic RNAi project at Harvard medical school: resources and validation

    Genetics 201:843–852.

    https://doi.org/10.1534/genetics.115.180208

    • PubMed
    • Google Scholar
72. 1. Pfeiffer BD
    2. Truman JW
    3. Rubin GM

    (2012) Using translational enhancers to increase transgene expression in Drosophila

    PNAS 109:6626–6631.

    https://doi.org/10.1073/pnas.1204520109

    • PubMed
    • Google Scholar
73. 1. Pitman JL
    2. Huetteroth W
    3. Burke CJ
    4. Krashes MJ
    5. Lai SL
    6. Lee T
    7. Waddell S

    (2011) A pair of inhibitory neurons are required to sustain labile memory in the Drosophila mushroom body

    Current Biology 21:855–861.

    https://doi.org/10.1016/j.cub.2011.03.069

    • PubMed
    • Google Scholar
74. Preprint

    1. Pop S
    2. Chen C-L
    3. Sproston CJ
    4. Kondo S
    5. Ramdya P
    6. Williams DW

    (2019) Patterned cell death sculpts functional and adaptive neural networks in flies

    bioRxiv.

    https://doi.org/10.1101/626465

    • Google Scholar
75. 1. Potter CJ
    2. Tasic B
    3. Russler EV
    4. Liang L
    5. Luo L

    (2010) The Q system: a repressible binary system for transgene expression, lineage tracing, and mosaic analysis

    Cell 141:536–548.

    https://doi.org/10.1016/j.cell.2010.02.025

    • PubMed
    • Google Scholar
76. Preprint

    1. Prieto-Godino LL
    2. Silbering AF
    3. Khallaf MA
    4. Cruchet S
    5. Bojkowska K
    6. Pradervand S
    7. Hansson BS
    8. Knaden M
    9. Benton R

    (2019) Functional integration of “undead” neurons in the olfactory system

    bioRxiv.

    https://doi.org/10.1101/623488

    • Google Scholar
77. 1. Prokop A
    2. Technau GM

    (1994) Normal function of the mushroom body defect gene of Drosophila is required for the regulation of the number and proliferation of neuroblasts

    Developmental Biology 161:321–337.

    https://doi.org/10.1006/dbio.1994.1034

    • PubMed
    • Google Scholar
78. 1. Purves D

    (1986) The trophic theory of neural concentrations

    Trends in Neurosciences 9:486–489.

    https://doi.org/10.1016/0166-2236(86)90155-4

    • Google Scholar
79. 1. Purves D
    2. Hume RI

    (1981) The relation of postsynaptic geometry to the number of presynaptic axons that innervate autonomic ganglion cells

    The Journal of Neuroscience 1:441–452.

    https://doi.org/10.1523/JNEUROSCI.01-05-00441.1981

    • PubMed
    • Google Scholar
80. 1. Ramaekers A
    2. Magnenat E
    3. Marin EC
    4. Gendre N
    5. Jefferis GS
    6. Luo L
    7. Stocker RF

    (2005) Glomerular maps without cellular redundancy at successive levels of the Drosophila larval olfactory circuit

    Current Biology 15:982–992.

    https://doi.org/10.1016/j.cub.2005.04.032

    • PubMed
    • Google Scholar
81. 1. Riemensperger T
    2. Völler T
    3. Stock P
    4. Buchner E
    5. Fiala A

    (2005) Punishment prediction by dopaminergic neurons in Drosophila

    Current Biology 15:1953–1960.

    https://doi.org/10.1016/j.cub.2005.09.042

    • PubMed
    • Google Scholar
82. 1. Ruta V
    2. Datta SR
    3. Vasconcelos ML
    4. Freeland J
    5. Looger LL
    6. Axel R

    (2010) A dimorphic pheromone circuit in Drosophila from sensory input to descending output

    Nature 468:686–690.

    https://doi.org/10.1038/nature09554

    • PubMed
    • Google Scholar
83. 1. Seroka AQ
    2. Doe CQ

    (2019) The hunchback temporal transcription factor determines motor neuron axon and dendrite targeting in Drosophila

    Development 146:dev175570.

    https://doi.org/10.1242/dev.175570

    • PubMed
    • Google Scholar
84. 1. Shaw RE
    2. Kottler B
    3. Ludlow ZN
    4. Buhl E
    5. Kim D
    6. Morais da Silva S
    7. Miedzik A
    8. Coum A
    9. Hodge JJL
    10. Hirth F
    11. Sousa‐Nunes R

    (2018) In vivo expansion of functionally integrated GABA ergic interneurons by targeted increase in neural progenitors

    The EMBO Journal 37:201798163.

    https://doi.org/10.15252/embj.201798163

    • Google Scholar
85. 1. Siller KH
    2. Cabernard C
    3. Doe CQ

    (2006) The NuMA-related mud protein binds pins and regulates spindle orientation in Drosophila neuroblasts

    Nature Cell Biology 8:594–600.

    https://doi.org/10.1038/ncb1412

    • PubMed
    • Google Scholar
86. 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
87. 1. Stopfer M
    2. Laurent G

    (1999) Short-term memory in olfactory network dynamics

    Nature 402:664–668.

    https://doi.org/10.1038/45244

    • PubMed
    • Google Scholar
88. 1. Strausfeld NJ
    2. Sinakevitch I
    3. Brown SM
    4. Farris SM

    (2009) Ground plan of the insect mushroom body: functional and evolutionary implications

    The Journal of Comparative Neurology 513:265–291.

    https://doi.org/10.1002/cne.21948

    • PubMed
    • Google Scholar
89. 1. Sweeney ST
    2. Hidalgo A
    3. de Belle JS
    4. Keshishian H

    (2012) Hydroxyurea ablation of mushroom bodies in Drosophila

    Cold Spring Harbor Protocols 2012:pdb.prot067777.

    https://doi.org/10.1101/pdb.prot067777

    • PubMed
    • Google Scholar
90. 1. Tenedini FM
    2. Sáez González M
    3. Hu C
    4. Pedersen LH
    5. Petruzzi MM
    6. Spitzweck B
    7. Wang D
    8. Richter M
    9. Petersen M
    10. Szpotowicz E
    11. Schweizer M
    12. Sigrist SJ
    13. Calderon de Anda F
    14. Soba P

    (2019) Maintenance of cell type-specific connectivity and circuit function requires tao kinase

    Nature Communications 10:1–16.

    https://doi.org/10.1038/s41467-019-11408-1

    • PubMed
    • Google Scholar
91. 1. Turner GC
    2. Bazhenov M
    3. Laurent G

    (2008) Olfactory representations by Drosophila mushroom body neurons

    Journal of Neurophysiology 99:734–746.

    https://doi.org/10.1152/jn.01283.2007

    • PubMed
    • Google Scholar
92. 1. Turney SG
    2. Lichtman JW

    (2012) Reversing the outcome of synapse elimination at developing neuromuscular junctions in vivo: evidence for synaptic competition and its mechanism

    PLOS Biology 10:e1001352.

    https://doi.org/10.1371/journal.pbio.1001352

    • PubMed
    • Google Scholar
93. 1. Vogt K
    2. Aso Y
    3. Hige T
    4. Knapek S
    5. Ichinose T
    6. Friedrich AB
    7. Turner GC
    8. Rubin GM
    9. Tanimoto H

    (2016) Direct neural pathways convey distinct visual information to Drosophila mushroom bodies

    eLife 5:e14009.

    https://doi.org/10.7554/eLife.14009

    • PubMed
    • Google Scholar
94. 1. Wang Y
    2. Guo HF
    3. Pologruto TA
    4. Hannan F
    5. Hakker I
    6. Svoboda K
    7. Zhong Y

    (2004) Stereotyped odor-evoked activity in the mushroom body of Drosophila revealed by green fluorescent protein-based Ca2+ imaging

    Journal of Neuroscience 24:6507–6514.

    https://doi.org/10.1523/JNEUROSCI.3727-03.2004

    • PubMed
    • Google Scholar
95. 1. Ward A
    2. Hong W
    3. Favaloro V
    4. Luo L

    (2015) Toll receptors instruct axon and dendrite targeting and participate in synaptic partner matching in a Drosophila olfactory circuit

    Neuron 85:1013–1028.

    https://doi.org/10.1016/j.neuron.2015.02.003

    • PubMed
    • Google Scholar
96. 1. Yagi R
    2. Mabuchi Y
    3. Mizunami M
    4. Tanaka NK

    (2016) Convergence of multimodal sensory pathways to the mushroom body Calyx in Drosophila Melanogaster

    Scientific Reports 6:29481.

    https://doi.org/10.1038/srep29481

    • PubMed
    • Google Scholar
97. 1. Yu HH
    2. Kao CF
    3. He Y
    4. Ding P
    5. Kao JC
    6. Lee T

    (2010) A complete developmental sequence of a Drosophila neuronal lineage as revealed by twin-spot MARCM

    PLOS Biology 8:e1000461.

    https://doi.org/10.1371/journal.pbio.1000461

    • PubMed
    • Google Scholar
98. 1. Zheng Z
    2. Lauritzen JS
    3. Perlman E
    4. Robinson CG
    5. Nichols M
    6. Milkie D
    7. Torrens O
    8. Price J
    9. Fisher CB
    10. Sharifi N
    11. Calle-Schuler SA
    12. Kmecova L
    13. Ali IJ
    14. Karsh B
    15. Trautman ET
    16. Bogovic JA
    17. Hanslovsky P
    18. Jefferis G
    19. Kazhdan M
    20. Khairy K
    21. Saalfeld S
    22. Fetter RD
    23. Bock DD

    (2018) A complete electron microscopy volume of the brain of adult Drosophila Melanogaster

    Cell 174:730–743.

    https://doi.org/10.1016/j.cell.2018.06.019

    • PubMed
    • Google Scholar
99. 1. Ziegler AB
    2. Thiele C
    3. Tenedini F
    4. Richard M
    5. Leyendecker P
    6. Hoermann A
    7. Soba P
    8. Tavosanis G

    (2017) Cell-Autonomous control of neuronal dendrite expansion via the fatty acid synthesis regulator SREBP

    Cell Reports 21:3346–3353.

    https://doi.org/10.1016/j.celrep.2017.11.069

    • PubMed
    • Google Scholar

Author details

1. Najia A Elkahlah

   Department of Molecular, Cellular and Developmental Biology, The University of Michigan, Ann Arbor, United States

   Contribution

   Formal analysis, Investigation, Visualization, Methodology, Writing - review and editing

   Contributed equally with

   Jackson A Rogow

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9657-1777

2. Jackson A Rogow

   Laboratory of Neurophysiology and Behavior, The Rockefeller University, New York, United States

   Contribution

   Formal analysis, Investigation, Visualization, Methodology, Writing - review and editing

   Contributed equally with

   Najia A Elkahlah

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5060-6770

3. Maria Ahmed

   Department of Molecular, Cellular and Developmental Biology, The University of Michigan, Ann Arbor, United States

   Contribution

   Formal analysis, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0054-3315

4. E Josephine Clowney

   Department of Molecular, Cellular and Developmental Biology, The University of Michigan, Ann Arbor, United States

   Contribution

   Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing

   For correspondence

   jclowney@umich.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9150-9464

• 3,839

  views

• 583

  downloads

• 13

  citations

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