1. Neuroscience
Download icon

A locomotor neural circuit persists and functions similarly in larvae and adult Drosophila

  1. Kristen Lee
  2. Chris Q Doe  Is a corresponding author
  1. Institute of Neuroscience, Howard Hughes Medical Institute, University of Oregon, United States
Research Advance
  • Cited 0
  • Views 749
  • Annotations
Cite this article as: eLife 2021;10:e69767 doi: 10.7554/eLife.69767

Abstract

Individual neurons can undergo drastic structural changes, known as neuronal remodeling or structural plasticity. One example of this is in response to hormones, such as during puberty in mammals or metamorphosis in insects. However, in each of these examples, it remains unclear whether the remodeled neuron resumes prior patterns of connectivity, and if so, whether the persistent circuits drive similar behaviors. Here, we utilize a well-characterized neural circuit in the Drosophila larva: the moonwalker descending neuron (MDN) circuit. We previously showed that larval MDN induces backward crawling, and synapses onto the Pair1 interneuron to inhibit forward crawling (Carreira-Rosario et al., 2018). MDN is remodeled during metamorphosis and regulates backward walking in the adult fly. We investigated whether Pair1 is remodeled during metamorphosis and functions within the MDN circuit during adulthood. We assayed morphology and molecular markers to demonstrate that Pair1 is remodeled during metamorphosis and persists in the adult fly. MDN-Pair1 connectivity is lost during early pupal stages, when both neurons are severely pruned back, but connectivity is re-established at mid-pupal stages and persist into the adult. In the adult, optogenetic activation of Pair1 resulted in arrest of forward locomotion, similar to what is observed in larvae. Thus, the MDN-Pair1 neurons are an interneuronal circuit – a pair of synaptically connected interneurons – that is re-established during metamorphosis, yet generates similar locomotor behavior at both larval and adult stages.

Introduction

Large-scale changes in neuronal morphology and function occur during mammalian puberty (Barendse et al., 2018; Mills et al., 2016; Sisk and Zehr, 2005), as well as several neurobiological disorders including depression (Patel et al., 2019), or chronic pain (Kuner and Flor, 2017). Similarly, major changes in neuronal numbers and type occur as a result of insect metamorphosis (Kanamori et al., 2015; Truman and Reiss, 1976; Yaniv and Schuldiner, 2016). Despite these changes, there are documented cases of individual insect neurons persisting from larval to adult stages. In Drosophila, individual motor and sensory neurons have been shown to persist throughout metamorphosis and undergo dramatic remodeling (Consoulas et al., 2002; Consoulas et al., 2000; Yaniv and Schuldiner, 2016; Yu and Schuldiner, 2014). Similar findings have been reported for the insect mushroom body, where Kenyon cells partners (projection neurons, DANs) exist at both larval and adult stages (Li et al., 2020; Marin et al., 2005). Yet, it remains unclear whether the remodeled neurons re-establish connectivity with the identical neurons in the larva and adult.

During Drosophila metamorphosis, the animal changes from a crawling limbless larva to a walking six-legged adult (Riddiford, 1980; Riddiford et al., 2003). Despite the obvious differences, some behaviors are similar: both larvae and adults undergo forward locomotion in search of food, backward locomotion in response to noxious stimuli, and pausing in between antagonistic behaviors (Carreira-Rosario et al., 2018). We and others identified a neuron that, when activated, can trigger backward locomotion in both larvae and adults (Bidaye et al., 2014; Carreira-Rosario et al., 2018; Sen et al., 2017), despite the obvious differences in limbless and six-legged locomotion. This neuron, named mooncrawler/moonwalker descending neuron (MDN), is present as a bilateral neuronal pair in each brain lobe, with all four MDNs having similar synaptic partners, and all four MDNs capable of eliciting backward larval locomotion in larvae (Carreira-Rosario et al., 2018). Larval MDNs function within a neural circuit that induces backward locomotion and coordinately arrests forward locomotion. Halting forward locomotion is achieved via activation of the Pair1 descending interneuron, which inhibits the A27h premotor neuron. Given that the A27h interneuron is required for forward locomotion, its inhibition via MDN-induced Pair1 activation prevents forward locomotion (Carreira-Rosario et al., 2018). Activating backward locomotion is likely to be due, in part, to MDN activation of the A18b premotor neuron, which is specifically active during backward locomotion (Carreira-Rosario et al., 2018). Thus, MDN-Pair1 are synaptically coupled members of a locomotor circuit in the Drosophila larva.

Here, we follow our previous work showing that MDN is remodeled during metamorphosis and persists into the adult (Carreira-Rosario et al., 2018) by asking: Is the MDN partner neuron Pair1 also maintained in the adult? Does the adult Pair1 induce an inhibition in forward locomotion, similar to its role in larvae? And, are the adult Pair1 and MDN synaptically coupled? We find that all of these questions are answered in the affirmative, showing that the core MDN-Pair1 interneuron circuit (a pair of synaptically connected interneurons) is re-established during metamorphosis despite profound neurite remodeling, and that this circuit coordinates forward/backward locomotion in both larvae and adults.

Results

The Pair1 neuron persists from larval to adult stages

To determine if Pair1 neurons were present in the adult, we mapped expression of a Pair1-Gal4 line (R75C02-Gal4) from early larval to adult stages. We identified the larval Pair1 neurons based on their characteristic cell body position in the medial subesophageal zone (SEZ), dense local ipsilateral dendritic arborizations (defined as dendritic based on enrichment for post-synapses in the TEM reconstruction of the larval Pair1 neuron; Figure 1—figure supplement 1), and contralateral axons descending into the ventral nerve cord (VNC) in an extremely lateral axon tract (Carreira-Rosario et al., 2018). Using the Pair1-Gal4 line, we could identify Pair1 neurons with this morphology at 28 and 96 hr after larval hatching (ALH; Figure 1A,B). The Pair1 neuron cell bodies and proximal neurites could still be observed at 24 hr after pupal formation (APF), but virtually all of the dendridic processes and descending axonal process are pruned (Figure 1C, only one neuron labeled). This is expected, given that many or all neurons undergo axon/dendrite remodeling during metamorphosis (Kanamori et al., 2015; Truman and Reiss, 1976; Yaniv and Schuldiner, 2016). At 48 hr APF, Pair1 neurons exhibited dendritic branching in the SEZ and a descending axon into the VNC, regaining morphological features similar to that of larval Pair1 neurons (Figure 1D). The axon innervated the T1 (prothoracic) neuropil and descended further down the VNC. These morphological features were maintained into the adult fly, where we could trace the Pair1 axon to primarily innervate the T1 neuropil (Figure 1E), with less extensive innervation of the mesothoracic (T2) and metathoracic (T3) neuropils.

Figure 1 with 1 supplement see all
The Pair1 neuron persists from larval to adult stages.

(A–B) Pair1 neurons (cell body: yellow asterisk; neurites: yellow arrowhead) in the larval CNS (gray outline) at 28 hr after larval hatching (ALH) (A) and 96 hr ALH (B). Here and in subsequent panels are maximum intensity projections of confocal sections containing the Pair1 neurons; anterior, up; dorsal view. Significant ‘off-target’ expression marked with white arrowheads. Scale bar, 50 μm. (A’–B’) Enlargement of the brain regions boxed in A,B. Scale bar, 20 μm. (A”–B”) Tracing to show Pair1 neuron morphology. Genotype: +; UAS-myr::GFP; R75C02-Gal4. (C–D) Pair1 neurons (cell body: yellow asterisk; neurites: yellow arrowhead) in the pupal CNS (gray outline) at 24 hr after pupal formation (APF) (C) and 96 hr APF (D). Significant ‘off-target’ expression marked with white arrowheads. Scale bar, 50 μm. (C’–D’) Enlargement of the brain regions boxed in C, D; cell body: yellow asterisk, neurites: yellow arrowhead. Scale bar, 10 μm. (C”) Tracing to show Pair1 neuron morphology. (D”) Focal plane showing Pair1 cell bodies (region boxed in D’, cell body marked with yellow asterisks). Scale bar, 10 μm. (D”’) Tracing to show Pair1 neuron morphology. Note that Pair1 can be followed to T1 in the 3D confocal stack but is difficult to represent here due to fasciculation of Pair1 with off-target neurons. Genotype: +; UAS-myr::GFP; R75C02-Gal4. (E) Pair1 neurons (cell body: yellow asterisk; neurites: yellow arrowhead) in the 4-day adult CNS (gray outline) Significant ‘off-target’ expression marked with white arrowheads. Scale bar, 50 μm. (E’) Enlargement of the brain region boxed in E. Scale bar, 10 μm. (E”) Focal plane showing Pair1 cell bodies (region boxed in E’, cell body marked with yellow asterisks). Scale bar, 10 μm. (E”’) Tracing to show Pair1 neuron morphology. Genotype: +; UAS-myr::GFP; R75C02-Gal4. (F) Pair1 neurons (cell body: yellow asterisk; neurites: yellow arrowhead) permanently labeled at 96 hr ALH and visualized in the 4-day old adult. See Materials and methods for details. Significant ‘off-target’ expression marked with white arrowheads. Scale bar, 50 μm. (F’) Enlargement of the brain region boxed in F; Pair1 cell body: yellow asterisk; Pair1 neurites: yellow arrowhead. Scale bar, 10 μm. (F”) Focal plane showing Pair1 cell bodies (region boxed in F’, cell body marked with yellow asterisks). Scale bar, 10 μm. (F”’) Tracing to show Pair1 neuron morphology. Genotype: Hs-KD,3xUAS-FLP; 13xLexAop(KDRT.Stop)myr:smGdP-Flag/+; 13xLexAop(KDRT.Stop)myr:smGdP-V5, 13xLexAop(KDRT.Stop)myr:smGdP-HA, nSyb(FRT.Stop)LexA::p65.

Although we can use the Pair1-Gal4 line to track neurons with Pair1 morphological features from larva to adult, it remains possible that the Gal4 line switches off in Pair1 and switches on in a similar descending neuron at a stage in between those we assayed. To conclusively demonstrate that the larval Pair1 neuron survives into adulthood, we used a genetic technique to permanently label or ‘immortalize’ the larval Pair1 neurons and assay for their presence in the adult brain. Briefly, the method achieves spatial specificity by using Pair1-Gal4 to drive UAS-FLP which removes a stop cassette from nSyb-FRT-stop-FRT-LexA resulting in permanent LexA expression in Pair1-Gal4 neurons; it achieves temporal specificity (e.g. labeling only larval Pair1-Gal4+ neurons) by using a heat inducible KD recombinase to ‘open’ the lexAop-KDRTstopKDRT-HA reporter (see Materials and methods for additional details). Thus, a heat shock will permanently label all Pair1-Gal4+ neurons at the time of heat shock. We immortalized Pair1 neurons in the larva, and assayed expression in the adult, and observed the two bilateral Pair1 neurons, based on characteristic medial SEZ cell body position, local ipsilateral arbors, and contralateral descending axons that preferentially innervate the prothoracic neuropil (Figure 1F). Pair1 innervation is clearer in neurons immortalized during larval stages, which reduces the off-target neuron expression in the adult VNC, and reveals an greatly enriched level of innervation in the T1 neuropil (Figure 1F).

The Pair1-Gal4 line is expressed in several off-target neurons in addition to Pair1. One of these, a sensory neuron from the proboscis can be removed from the adult Pair1 pattern by cutting off the proboscis a day prior to analysis (see Materials and methods) but is present at the 48 hr APF timepoint (Figure 1D,E). In addition, there are off-target neurons that innervate all three thoracic neuropils (T1-T3), obscuring Pair1 innervation (Figure 1E). We took advantage of the sparse labeling of the immortalization genetics and found brains that maintained preferential targeting of Pair1 to the prothoracic neuropil but lacked T1-T3 off-target innervation, confirming that they are indeed off-target neurons (Figure 1F–F''').

Pair1 neurons maintain the same molecular profile from larval to adult stages

If Pair1 neurons persist from larva to adult, they may express the same transcription factor (TF) profile at both stages. We screened a small collection of TF markers for expression in the larval and adult Pair1 neurons, and in all cases we found identical expression (Figure 2A–N). Larval and adult Pair1 neurons expressed Hunchback (Hb), Sex combs reduced (Scr), and Bicoid (Bcd); but did not express Visual system homeobox (Vsx1) or Nab (Figure 2A–N). However, there were many other Scr/Bcd/Hb triple-positive neurons in the SEZ (Figure 2O,P), showing that additional factors would be necessary to uniquely specify Pair1 identity. These results support the conclusion that Pair1 persists from larva to adult, maintaining both molecular and morphological features, and raises the interesting possibility that the three TFs (Hb, Bcd, Scr) may be part of a molecular code that directs both larval and adult Pair1 morphology and/or connectivity.

The Pair1 neuron expresses the same molecular markers at larval and adult stages.

(A) Schematic of the larval brain showing region of Pair1 neurons (red box) enlarged in panels below. Anterior up, dorsal view. (B–G) Larval Pair1 neurons (left column), indicated markers (middle column), and merge (right column) at 28 hr after larval hatching (ALH). In some cases the second Pair1 neuron is out of the focal plane, but both Pair1 neurons always have the same gene expression profile. Markers detect the following transcription factors: Hb, Hunchback; Scr, Sex combs reduced; Bcd, Bicoid; Vsx1, Visual system homeobox 1; and Nab. Scale bar, 5 μm. (G) Summary: marker expression matches that in adults. Genotype: +; UAS-myr::GFP; R75C02-Gal4. (H) Schematic of the adult brain showing region of Pair1 neurons (red box) enlarged in panels below. Anterior up, dorsal view. (I–N) Adult Pair1 neurons (left column), indicated markers (middle column), and merge (right column) in 4-day old adult. Scale bar, 5 μm. (N) Summary: marker expression matches that in larvae. Genotype: +; UAS-myr::GFP; R75C02-Gal4. (O–P) The number of cells expressing Scr (first column), Scr/Hb (second column), Scr/Bcd (third column), and Scr/Hb/Bcd (fourth column) in larvae (O) and adults (P). n = 5–6 whole brains.

Pair1 activation arrests forward locomotion in adults

We previously showed that larval MDN persists in adults and can induce backward locomotion at both stages despite the obvious difference in motor output – limbless crawling versus six-legged walking (Carreira-Rosario et al., 2018). This raised the question of whether the adult Pair1 neuron also maintains its larval function, that is, to pause forward locomotion. To test this hypothesis, we used Pair1-Gal4 to express the red light-gated cation channel CsChrimson (Chrimson) to activate Pair1 neurons in the adult. Experimental flies were fed all-trans retinal (ATR; required for Chrimson function) whereas control flies were fed vehicle only.

Control flies exposed to red light did not pause or arrest forward locomotion, did not show an increased probability of pausing, and did not have a decrease in distance traveled during the stimulus interval. In contrast, experimental flies expressing Chrimson in Pair1 neurons showed a near-complete arrest of forward locomotion, an increased probability of pausing, and a reduced distance traveled during the stimulus interval (Figure 3A–C; Figure 3—figure supplement 1). These effects were reversed after turning off the red light, with the exception of a slightly reduced distance traveled, likely due to a lingering physiological effect of the 30 s Pair1 activation (Figure 3A,C). Pair1 activation resulted in an increase in immobile flies (Figure 3E) and a corresponding decrease in whole body translocation (Figure 3F, defined as ‘large movements’). Importantly, Pair1 activation did not prevent small body part movements such as those involved in grooming (Figure 3G, defined as ‘small movements’). Note that Pair1-Gal4 off-target expression is common but variable from fly to fly, whereas its expression in Pair1 neurons is fully penetrant; because the Chrimson-induced behavior is also fully penetrant, we conclude that the arrest in forward locomotion is due to Chrimson activation of the Pair1 neurons. We conclude that Pair1 activation prevents a single behavior – forward locomotion – but does not produce general paralysis or interfere with non-translocating limb movements.

Figure 3 with 1 supplement see all
Pair1 activation for 30 s arrests forward locomotion but does not cause paralysis in adults.

(A) Speed (mm/s) of adult flies expressing Chrimson in Pair1 neurons following neuronal activation (+ATR [all-trans retinal] , blue) or no activation (vehicle control, black) in a closed loop arena. Speed was recorded for the 30 s prior to activation, the 30 s light-induced activation (red stipple), and 30 s after activation. Mean ± SEM, n = 10. Genotype for this and all subsequent panels: UAS-CsChrimson::mVenus; +; R75C02-Gal4. (B) Probability of forward locomotion pausing upon light-induced Pair1 activation (ATR treatment, blue) compared to vehicle control (black). Statistics: t-test, p < 0.001; n = 10. (C) Total distance traveled pre-light stimulus (‘pre’), during the light stimulus (‘light’) and post-light stimulus (‘post’) (terminology used here and in subsequent panels) of flies fed ATR (Pair1 activation, blue) compared to controls (fed vehicle, no Pair1 activation, black). Statistics: two-way ANOVA: drug treatment, F(1, 18) = 111.3, p < 0.0001; time, F(1.867, 33.61) = 47.03, p < 0.0001; interaction F(2, 26) = 38.24, p < 0.001; Bonferroni’s multiple comparisons between drug treatments within each timepoint: pre, p > 0.9999; light, p < 0.0001; post, p = 0.0001; n = 10. (D) Percent time doing large movements (whole body translocation, light gray), small movements (body part movement but no translocation, dark gray) or no movements (immobile, black) of flies fed vehicle (left side) or ATR (right side) during each time phase (pre, light, post). (E) Normalized duration of time spent immobile during each timepoint (pre, light, post) for flies fed ATR (Pair1 activation, blue) compared to controls fed vehicle (black). Statistics: two-way ANOVA: drug treatment, F(1, 18) = 112.8, p < 0.0001; time, F(1.930, 34.74) = 25.55, p < 0.0001; interaction, F(2, 36) = 27.81, p < 0.0001; Bonferroni’s multiple comparisons between drug treatments within each timepoint: pre, p > 0.9999; light, p < 0.0001; post, p = 0.0022; n = 10. (F) Normalized duration of time spent doing small movements during each timepoint (pre, light, post) for flies fed ATR (Pair1 activation, blue) compared to controls fed vehicle (black). Statistics: two-way ANOVA: drug treatment, F(1, 18) = 5.111, p = 0.036; time, F(1.923, 34.62) = 10.82, p = 0.0003; interaction, F(2, 36) = 4.225, p = 0.0225; Bonferroni’s multiple comparisons between drug treatments within each timepoint: pre, p > 0.9999; light, p = 0.0022; post, p > 0.9999; n = 10. (G) Normalized duration of time spent doing large movements during each time phase (pre, light, post) for flies fed ATR (Pair1 activation, blue) compared to controls fed vehicle (black). Statistics: two-way ANOVA: drug treatment, F(1, 18) = 53.56, p < 0.0001; time, F(1.869, 33.64) = 53.44, p < 0.0001; interaction, F(2, 36) = 52.20, p < 0.0001; Bonferroni’s multiple comparisons between drug treatments within each timepoint: pre, p > 0.9999; light, p < 0.0001; post, p = 0.0074; n = 10.

MDN and Pair1 are synaptic partners during adulthood

Given that MDN and Pair1 are synaptic partners in the larvae (Figure 1—figure supplement 1), MDN and Pair1 persist into adulthood (Figures 1 and 2), and MDN and Pair1 both regulate the same behavior in larvae and adults (Figure 3; Carreira-Rosario et al., 2018), we hypothesized that MDN and Pair1 may also be synaptic partners during adulthood. To test this hypothesis, we used the MDN-LexA and Pair1-Gal4 to label MDN and Pair1 neurons individually in the same animal (Figure 4A,B). We observed MDN and Pair1 neurites in close proximity to each other (Figure 4C–E).

Moonwalker descending neurons (MDN) and Pair1 are synaptic partners in adults.

(A–E) MDN and Pair1 show close membrane apposition. (A, B) MDN neurons (A) and Pair1 neurons (B) in the adult central brain. Neurons are in white, nc82 counterstain in magenta for whole brain orientation; cell bodies marked by yellow arrowheads. Here and in subsequent panels shows maximum intensity projection of volume; anterior, up; dorsal view. Scale bar, 50 μm. Genotype: UAS-mCD8::RFP, LexAop-mCD8::GFP; VT044845-LexA; R75C02-Gal4. (C–E) MDN neurites (C), Pair1 neurites (D), and merge (E) in the left subesophageal ganglion (red box in schematic). Scale bar, 10 μm. Genotype: UAS-mCD8::RFP, LexAop-mCD8::GFP; VT044845-LexA; R75C02-Gal4. (F–K) t-GRASP (targeted GFP reconstitution across synaptic partners) between MDN and Pair1. In all panels: Scale bar, 10 μm. Genotype:;; LexAop-pre-t-GRASP, UAS-post-t-GRASP/R75C02-Gal4. (F–G) Pupal t-GRASP at 24 hr after pupal formation (APF). (F) No detectable t-GRASP signal was observed in the subesophageal ganglion without expression of the pre-t-GRASP fragment in MDN. (G) t-GRASP signals between MDN and Pair1 were lacking in the subesophageal ganglion. (H–I) Pupal t-GRASP at 48 hr APF. (H) No detectable t-GRASP signal was observed in the subesophageal ganglion without expression of the pre-t-GRASP fragment in MDN. (I) t-GRASP signals between MDN and Pair1 were observed in the subesophageal ganglion. (J–K) Adult t-GRASP. (J) No detectable t-GRASP signal was observed in the subesophageal ganglion without expression of the pre-t-GRASP fragment in MDN. (K) t-GRASP signals between MDN and Pair1 were observed in the subesophageal ganglion.

To determine if MDN and Pair1 are synaptic partners in this region of neuropil, we utilized t-GRASP (targeted GFP reconstitution across synaptic partners), an activity-independent method to label synaptic contact sites (Shearin et al., 2018). MDN-Pair1 connectivity was absent at 24 hr APF (Figure 4F,G) when neurite pruning was maximal (Figure 1C), but was re-established at 48 hr APF (Figure 4H,I) when neurite regrowth was occurring (Figure 1D) and maintained into the adult (Figure 4J,K).

Across each timepoint, control flies only expressing pre-t-GRASP in MDN did not have detectable t-GRASP signal (Figure 4F,H and J). Thus, MDN-Pair1 larval connectivity is established twice during development: initially established in the late embryo, and then re-established in the pupae.

Discussion

Together with our earlier work (Carreira-Rosario et al., 2018), our results here show that a core interneuron circuit is preserved from larval stages into the adult. This neuronal circuit contains MDN and its monosynaptically coupled Pair1 neuron, allowing the fly to switch between antagonistic behaviors: forward versus backward locomotion. Our work raises several interesting questions: Do many other larval neural circuits persist and have similar function in adults? Do these results represent a rare occurrence or a common one? Are the cues that establish MDN-Pair1 connectivity in the larvae also used to re-establish MDN-Pair1 connectivity in the adult?

How much of the larval MDN-Pair1 circuit is maintained into the adult? The larval circuit contains the MDN partners Pair1, ThDN, and A18b, and the Pair1 partner A27h (Figure 5; Carreira-Rosario et al., 2018). In addition to MDN, we show here that Pair1 is maintained. There are no Gal4 lines or markers for the ThDN neuron. Although two independent A18b Gal4 lines have extensive off-target expression, permanently labeling the Gal4 expression during larval stages consistently labels an adult neuron in the abdominal ganglion (data not shown); since MDN axons do not extend to the abdominal ganglion, it is unlikely that MDN and A18b are synaptic partners during adulthood. Additionally, the A27h interneuron, which regulates forward crawling in the larvae, undergoes apoptosis during pupal stages (data not shown) and thus cannot regulate forward walking in the adult. This is not surprising as the A27h neurons are located in the abdominal segments, which do not have a role in adult walking.

Model describing the moonwalker descending neuron (MDN)-Pair1 circuit in larval and adult stages.

In the larvae, MDN and Pair1 neurons are located in the central brain. MDN and Pair1 neurons are synaptic partners. MDN neurons also extend axons into the abdominal region of the ventral nerve cord (VNC) and synapse onto A18b. A18b subsequently regulates backward crawling by synapsing onto motor neurons. Pair1 synapses onto and inhibits the pre-motor neuron A27h, which generates forward locomotion when activated. During metamorphosis, MDN and Pair1 neurons remain in the central brain, drastically prune their neurites, and survive. The A18b neuron and processes remain in the abdomen of the VNC and survives in the adult. The A27h neuron undergoes apoptosis. In the adult, synapses between MDN and Pair1 are re-established. MDN neurons also extend axons into the thoracic region of the VNC and synapse onto LBL40 and LUL130. Activation of LBL40 and LUL130 generates backward walking. MDN does not extend axons into the abdominal region of the VNC and is no longer synaptic partners with A18b. Pair1 neurons extend axons into the thoracic region of the VNC and synapse onto unknown neurons. Activation of Pair1 generates a pausing behavior, likely through the inhibition of neurons generating forward locomotion. In both larvae and adult, MDN and Pair1 neurons (blue) persist and function as a core circuit to regulate locomotion.

How much of the adult MDN-Pair1 circuit is present in the larvae? Recent work mapping the adult MDN circuit has identified over 30 VNC neurons downstream of MDN, including the LBL40 and LUL130 neurons required for hindleg backward stepping (Figure 5; Feng et al., 2020). Recent work has also identified adult neurons important for forward walking (Bidaye et al., 2020), but their relationship to adult MDN is unknown. In the future, it will be interesting to see if any of these adult neurons are present in the larvae, particularly those regulating forward and backward walking, and determine if they are also MDN or Pair1 target neurons.

Elegant recent work has shown that initiation of forward walking requires the forelegs, innervated by motor neurons in the prothoracic segment, whereas initiation of backward walking requires the hindlegs, innervated by motor neuron in the metathoracic segment (Feng et al., 2020). Similarly, adult MDN synaptic partners primarily innervate the metathoracic (T3) neuromere (Feng et al., 2020), a good location for inducing hindleg stepping and initiation of backward walking. A similar spatial segregation is likely to occur in the larva, where forward crawling is induced by A27h in posterior segments, and backward crawling is induced in anterior segments (Fushiki et al., 2016; Tastekin et al., 2018).

Is maintenance of neuronal circuits from larval to adult stages rare or common? Individual neurons that have similar functions in larva and adults have been identified, including select motor neurons, sensory neurons (Consoulas et al., 2002; Consoulas et al., 2000; Levine, 1984; Truman, 1992; Weeks, 2003), and Kenyon cells of the mushroom body (Eichler et al., 2017; Li et al., 2020). However, it remains unknown whether any of their individual synaptic partners also persist and retain the same pattern of connectivity. To date, the MDN-Pair1 circuit is the first pair of interneurons whose connectivity is established in the larva and then re-established after remodeling during metamorphosis, all while maintaining a similar function. It is currently unclear how common it is to maintain interneuron connectivity from larva to adult. If maintenance is rare, it raises the question of what is special about the MDN-Pair1 circuit? If a common occurrence, the MDN-Pair1 circuit provides an excellent model system to study a widely used mechanism.

Our work is the first, to our knowledge, to show that a pair of synaptically connected interneurons can persist from larva to adult and perform similar functions at both stages. Remarkably, both MDN and Pair1 undergo dramatic pruning and regeneration events during metamorphosis, only to re-form synapses with each other following neuronal remodeling. This suggests that synapse specificity cues are maintained from the late embryo, where MDN-Pair1 connectivity is first established, into pupal stages, where MDN-Pair1 connectivity is re-established. The importance of the MDN-Pair1 interneuron circuit is highlighted by its persistence from embryo to adult, despite adapting to different sensory input and motor output at each stage. Perhaps other descending or ascending interneurons will also persist into adults, switching inputs and outputs as needed. Indeed, the idea that an interneuron circuit that is stable across developmental stages is supported by recent elegant TEM reconstruction of neural circuits at different stages of Caenorhabditis elegans development (Witvliet et al., 2020). Here, the authors conclude that ‘Across maturation, the decision-making (interneuron) circuitry is maintained whereas sensory and motor pathways are substantially remodeled.’ These results, together with ours, raise the possibility that preservation of interneuron circuit motifs may be functional modules that can be used adaptively with different sensorimotor inputs and outputs. The presence of this circuit motif in both flies and worms suggests that it may be an ancient evolutionary mechanism for assembling sensorimotor circuits.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Genetic reagent (Drosophila melanogaster)R75C02-Gal4BDSCRRID:BDSC_39886Short genotype:
Pair1-Gal4
Genetic reagent (Drosophila melanogaster)VT044845-lexAGift from B Dickson, JRCShort genotype:
MDN-LexA
Genetic reagent (Drosophila melanogaster)UAS-myr::GFPBDSCRRID:BDSC_32198Gal4 reporter
Genetic reagent (Drosophila melanogaster)UAS-mChrimson::mVenusGift from Vivek Jayaraman, JRCWas used to excite/depolarize neurons of interest
Genetic reagent (Drosophila melanogaster)UAS-mCD8::RFP, LexAop-mCD8::GFPBDSCRRID:BDSC_32229Gal4 and LexA reporters
Genetic reagent (Drosophila melanogaster)LexAop-pre-t-GRASP, UAS-post-t-GRASPBDSC
(Shearin et al., 2018)
RRID:BDSC_79039t-GRASP
Genetic reagent (Drosophila melanogaster)Hs-KD,3xUAS-FLP; 13xLexAop(KDRT.Stop)myr:smGdP-Flag/ CyO-YFP; 13xLexAop(KDRT.Stop)myr:smGdP-V5, 13xLexAop(KDRT.Stop)myr:smGdP-HA, nSyb-(FRT.Stop)-LexA::p65/R75C02-Gal4This workUsed to permanently label Gal4 pattern
Antibody, polyclonalRabbit polyclonal anti-GFP A-11122Thermo Fisher Scientific, Waltham, MARRID:AB_221569(1:500)
Antibody, polyclonalChicken polyclonal anti-GFPAbcam, Eugene, ORRRID:BDSC_13970(1:1500)
Antibody, monoclonalRabbit polyclonal anti-GFP (G10362)Thermo Fisher Scientific, Waltham, MARRID:AB_2536526(1:300); used for t-GRASP
Antibody, monoclonalRat monoclonal anti-HA (3F10)Sigma, St. Louis, MOSKU: 11867423001(1:100)
Antibody, monoclonalMouse monoclonal anti-ScrDSHB (Iowa City, IA)RRID:AB_528462(1:10)
AntibodyRat polyclonal anti-BcdGift from John Reinitz, University of Chicago, IL(1:100)
AntibodyGuinea pig polyclonal anti-Vsx1Gift from Claude Desplan, NYU, New York, NY(1:500)
AntibodyGuinea pig polyclonal anti-NabGift from Stefan Thor, University of Queensland, Brisbane, Australia(1:500)
AntibodySecondary antibodiesJackson ImmunoResearch, West Grove, PA(1:400); all Donkey

Fly husbandry

Request a detailed protocol

All flies were reared in a 25°C room at 50% relative humidity with a 12 hr light/dark cycle. All comparisons between groups were based on studies with flies grown, handled, and tested together.

Fly stocks

Request a detailed protocol
  1. R75C02-Gal4 (Pair1 line; BDSC #39886).

  2. UAS-myr::GFP (BDSC #32198).

  3. UAS-CsChrimson::mVenus (Vivek Jayaraman, Janelia Research Campus).

  4. VT044845-lexA (adult MDN line; a gift from B Dickson, Janelia Research Campus).

  5. UAS-mCD8::RFP, LexAop-mCD8::GFP;; (BDSC #32229).

  6. LexAop-pre-t-GRASP, UAS-post-t-GRASP (BDSC #79039).

  7. Hs-KD,3xUAS-FLP; 13xLexAop(KDRT.Stop)myr:smGdP-Flag/ CyO-YFP; 13xLexAop(KDRT.Stop)myr:smGdP-V5, 13xLexAop(KDRT.Stop)myr:smGdP-HA, nSyb-(FRT.Stop)-LexA::p65/R75C02-Gal4 (line to permanently label Pair1; Doe Lab; modified from Ren et al., 2016).

Gal4 driver ‘immortalization’

Request a detailed protocol

Immortalization flies (see genotype #7, above) were allowed to lay eggs for 4 hr. Newly hatched larvae were placed in a food vial, and at 96 hr ALH the food vial was partially submerged in a 37°C water bath for 5 min, allowing the hs-KD to act as a recombinase to remove the KDRT Stop cassette, resulting in nSyb-LexA driving HA expression permanently in the neurons expressing Pair1-Gal4 at the time of heat shock (96 hr ALH). After the heat shock, larvae in the food vial recovered at 18°C for 5 min, and then grown to adulthood at 25°C.

Immunostaining and imaging

Request a detailed protocol

Standard confocal microscopy and immunocytochemistry methods were performed as previously described (Carreira-Rosario et al., 2018). Primary antibodies used recognize: GFP (rabbit, 1:500, Thermo Fisher Scientific, Waltham, MA; chicken, 1:1500, Abcam12970, Eugene, OR), HA (rat, 1:100, Sigma, St. Louis, MO), Hb (mouse, 1:400, AbcamF18-1G10.2), Scr (mouse, 1:10, Developmental Studies Hybridoma Bank, Iowa City, IA), Bicoid (rat, 1:100, John Reinitz, University of Chicago, IL), Vsx1 (guinea pig, 1:500, Claude Desplan, NYU, New York, NY), Nab (guinea pig, 1:500, Stefan Thor, University of Queensland, Brisbane, Australia), and t-GRASP signal (rabbit GFP G10362, 1:300, Invitrogen). Secondary antibodies were from Jackson ImmunoResearch (Donkey, 1:400, West Grove, PA). Confocal image stacks were acquired on a Zeiss 800 microscope. All images were processed in Fiji (https://imagej.new/Fiji) and Adobe Illustrator (Adobe, San Jose, CA). Images were processed as described previously (Carreira-Rosario et al., 2018). The primary neurites of Pair1 were traced using the Simple Neurite Tracer in Fiji.

Cell counts

Request a detailed protocol

Cell counting was done manually using the ‘Cell Counter’ plugin in Fiji (https://imagej.new/Fiji). Only cells expressing Scr were counted.

Adult behavioral experiment

Request a detailed protocol

Adult behavior was assayed using two arenas, a closed loop arena (Figure 3) and an open field arena (Figure 3—figure supplement 1). For the closed loop arena, adult female flies 1 day after eclosion were transferred to standard cornmeal fly food supplemented with 100 mL 0.5 mM ATR or 100% ethanol for 4 days (changed every 2 days). Animals, with intact wings, were starved for 4 hr and then placed in arenas and their behavior was recorded as described previously (Carreira-Rosario et al., 2018). Flies were exposed to low transmitted light, red light, and low transmitted light again for 30 s each. This was done three times for each animal. To calculate different parameters, the recorded videos were tracked and analyzed using the CalTech Fly Tracker (Fontaine et al., 2009) and JABA (Kabra et al., 2013). The speed, distance, and behavior reported were specific to the first trial. The reported speeds are the average speed of each second. The pausing probability was calculated as previously described (Carreira-Rosario et al., 2018). ‘Pre’ defines the 30 s prior to red light exposure, ‘light’ defines the 30 s of red light exposure, and ‘post’ defines the 30 s after red light exposure. Immobile movements were defined as the fly not translocating and not moving other body parts. Small movements were defined as the fly not translocating but moving body parts (i.e. grooming, moving wings). Large movements were defined as the fly translocating its body. All behavior measures were normalized by dividing them by the group average ‘pre’ values.

For the open field arena, adult flies were fed ATR and vehicle as described above. Three animals were placed in a circular arena with a diameter of 14.5 cm and height of 0.5 cm. After 5 min for environmental acclimation, animal behavior was recorded at 25 FPS using a Basler acA2040-25gm GigE camera under infrared light for 4 s followed by 4 s under red light and another 4 s under infrared light, as described previously (Risse et al., 2013). The was repeated three times, and tracked and analyzed as described above.

t-GRASP

Request a detailed protocol

Synapse establishment was investigated via t-GRASP. Flies with MDN-LexA and Pair1-Gal4 driving expression of t-GRASP (see genotype #6 above) were reared at 25°C and dissected at 24 hr APF, 48 hr APF, and 4 days post-eclosion. Control flies lacked MDN-LexA.

Statistics

All statistical analysis (t-test, one-way and two-way ANOVA with Bonferroni’s multiple comparison tests) were performed with Prism 9 (GraphPad Software, San Diego, CA). Numerical data in graphs show individual measurements (animals), means (represented by red bars) or means ± SEM (dashed lined), when appropriate. The number of replicates (n) is indicated for each data set in the corresponding legend.

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files.

References

  1. Book
    1. Kanamori T
    2. Togashi K
    3. Koizumi H
    4. Emoto K
    (2015)
    Chapter One - Dendritic Remodeling: Lessons from Invertebrate Model Systems
    In: Jeon K. W, editors. International Review of Cell and Molecular Biology. Academic Press. pp. 1–25.
    1. Yaniv SP
    2. Schuldiner O
    (2016) A fly's view of neuronal remodeling
    Wiley Interdisciplinary Reviews: Developmental Biology 5:618–635.
    https://doi.org/10.1002/wdev.241

Decision letter

  1. Ronald L Calabrese
    Senior and Reviewing Editor; Emory University, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This paper will be of interest to scientists interested in comparative neural circuits and how elements of circuit control are conserved across changes in development (or presumably evolution). The authors use compelling genetic experiments to show that Pair1 neurons, a synaptic partner of moonwalker descending neurons (MDN), are conserved between larval and adult stages, retracting during pupation and re-forming connections and much of their behavioral function in adulthood. This adds to our growing understanding of circuit conservation across insect brains.

Decision letter after peer review:

Thank you for submitting your article "A locomotor neural circuit persists and functions similarly in larvae and adult Drosophila" for consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by Ronald Calabrese as the Senior and Reviewing Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

This is a solid piece of work but the advance as presented is limited.

1. A primary concern is that there is no broader consideration of other connections of MDN in the larval transformation to adult. The focus on one specific connection being preserved may not represent the overall scale of remodeling during metamorphosis. There are a variety of experiments proposed that could provide information on this issue and give stronger insight into what types of connections are preserved/reconnected. For example, the trans-tango experiments proposed, even if they don't allow the authors to identify every connected cell-type, could provide a sense of scale. Part of this could be fixed by text revisions if specific experiments are not feasible.

2. A second concern is the breadth of TF staining suggesting that the specific TFs observed may not be controlling connectivity.

3. A third concern is the framing of the MDN-Pair1 connection as a decision-making circuit. This concern is best addressed by text revision. The detailed comments of the reviewers will provide specific reasoning/recommendations to the authors.

Reviewer #1:

This study from Lee and Doe examines the conservation of morphology, function, and connectivity in a Drosophila neural circuit that promotes backward locomotion. The study builds on a previous eLife paper from this group, Carreira-Rosario et al. 2018, which showed that a set of descending neurons- the Moonwalker Descending Neurons (MDN)- were conserved between larval and adult stages and performed a similar function in initiating backward locomotion, despite the very different locomotor strategies of crawling larvae and walking adults. Here, the authors examine a second neuron type- Pair1- that is downstream of MDN and inhibits forward locomotion. They first use a genetic immortalization strategy to show that the morphology of this neuron is similar between larvae and adults, although many of its processes are pruned during pupation. They next show that a set of transcription factor markers in these neurons are conserved between larval and adult stages. Finally these show that optogenetic activation of Pair1 in adults generates a similar stopping phenotype as in larvae, and that MDN and Pair 1 and synaptically connected (via GRASP) in adults as well. The authors conclude that these neurons form a core "decision-making" circuit that is conserved between developmental stages, and speculate that the continued expression of the same transcription factors may allow these neurons to re-form synapses after pupation. Overall this is a clear study that provides new data on which parts of the fly locomotor circuit are maintained across developmental stages. Together with on-going work in other parts of the fly brain, and across insect species, I expect this work to contribute to our understanding of how neural circuits become adapted through genetic and developmental mechanisms to different organismal niches.

The data are generally strong and convincing. One concern I have is that the transcription factor stains in Figure 2 show that most of the surrounding cell bodies share the same pattern of expression for the 5 TFs examined, with Hb, Scr, and Bcd highly expressed and Vsx1 and Nab absent. This might impact the authors' model that TF expression patterns in Pair1 allow it to make correct synaptic connections with MDN. The authors should address whether they expect these surrounding neuron to make similar synaptic connections, or whether connectivity is determined by a larger group of TFs other than those shown here. Experiments manipulating TF expression and looking at effects on connectivity would strength this part of the authors' model.

An additional concern is that the MDN-Pair1 neurons are referred to multiple time in the Discussion as a "decision-making circuit." However this term is poorly defined. The authors should specify what they mean by this term and how they would distinguish a "decision-making circuit" from other central circuits. More broadly, it would be helpful to be specific about what central circuit features they would expect to be conserved between larvae and adults or to state clearly if there too little data yet to make broad predictions of this type.

Reviewer #2:

This paper asks whether a neural connection that exists in the brain of the Drosophila larva also forms in the adult brain and serves a similar function. Past work (Carreira-Rosario et al., 2018) showed that in larvae, the mooncrawler descending neuron (MDN) drives the transition from forward to backward crawling in part through its connection onto the bilateral pair of Pair1 neurons. Here the authors find that Pair1 Gal4 lines label a pair of adult descending neurons with cell bodies in the gnathal ganglia which express the same developmental markers as the larval Pair1 neurons. Positive GRASP staining using MDN and Pair1 Gal4 lines suggests the adult neurons are synaptically connected. Furthermore, optogenetic activation of Pair1 neurons causes adult flies to stop walking or moving, similar to the effect found for Pair1 activation in larvae. They show that Pair1 neurons undergo dramatic remodeling of their axonal arbors in the ventral nerve cord (VNC).

This paper extends the work of Carreira-Rosario et al., (2018), by following the Pair1 neurons through development. As a conceptual advance, it is an interesting finding that Pair 1 neurons play similar functional roles in adults as in larvae, even though the downstream targets must be entirely different. Pair1 neurons were shown to form inhibitory synapses onto premotor neurons A27h in posterior abdominal segments, which themselves drive muscle contraction and initiate forward crawling. A27h undergo apoptosis during pupal development, abdominal segments lose their role in locomotion, and Pair1 develops extensive arbors in the adult T1 where locomotion is likely initiated.

The work addresses important and interesting questions, but its scope is limited. While the authors identify some similarities in connectivity and function of Pair1 neurons across metamorphosis, the characterization of the Pair1 neurons is far from complete.

I feel that two major revisions would improve the paper. First, there are some simple experiments that would help broaden the scope and impact of the study. The authors could survey the downstream partners of Pair1 in the T1 thoracic neuromere using Trans-tango. They could also ask whether silencing Pair1 (e.g. driving Kir2.1 expression) causes behavioral deficits, to test whether pair 1 neurons are necessary for the switch between forward and backward walking.

The paper would also benefit from a deeper exploration of the degree to which the preservation of MDN to Pair1 connectivity is representative or extraordinary. For example, are there connectivity motifs from the larvae that the authors did not find in the adult? Including these negative results would help provide context. Many neurons (e.g. thoracic motor neurons) gain a role in adulthood, others lose a role (e.g. abdominal motor neurons no longer control "locomotion"). What would one expect if it were possible to compare all the inputs and outputs of Pair1, or other fly neurons?

Given that the paper builds on a previous study, it could also be improved by editing for concision. There are some sections that distract from the key take-home message. For example, the authors state that this circuit plays a role in making decisions. I understand the point, that switching from forward to backward locomotion indicates that a decision has been computed, but it begs questions that are not addressed, such as what exactly constitutes a decision? Where is the decision being made? What mechanism, like a threshold, or competition between neurons, underlie when a decision takes place? Can adult flies stop (Pair1 activation by MDN) without moving backwards (LBL40, LUL130)? If the paper focuses on the developmental questions, provides background on what other interesting motifs are seen as circuits remodel, and position the results within those broader themes, I feel I would have a clearer sense of the novelty of these findings.

Issues:

Pair1 optogenetic activation causes larvae to pause briefly, whereas adults stop. The authors speculate it's due to circuitry. Could temporal differences in Channelrhodopsin kinetics and light protocols explain this?

Some edits:

This neuron, named Moonwalker/Mooncrawler Descending Neuron

(MDN) is present in two bilateral pairs per brain lobe, – Awkward phrasing.

"Halting forward locomotion is done via activation of the Pair1 descending

interneuron, which inhibits the A27h premotor neuron, to prevent it from inducing forward locomotion (Carreira-Rosario et al., 2018)." – Unclear.

"We find that all of these questions are answered in the affirmative, showing that the core MDN-Pair1 decision-making circuit (a pair of synaptically-connected interneurons) persists from larva to adult, despite profound remodeling during metamorphosis, and that this circuit coordinates forward/backward locomotion in both larvae and adults." – see point above.

"induce backward locomotion via the coordinate arrest of forward locomotion" – unclear.

77 – "but virtually all of the dendridic processes and descending axonal process had been pruned (Figure 1C, only one neuron labeled)."

107 – "We conclude that the Pair1 neurons are present from larval to adult stages, and that the Pair1 neurons are enriched for postsynaptic partners in the T1

neuromere." – "enriched" is a strange word in this context.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "A locomotor neural circuit persists and functions similarly in larvae and adult Drosophila" for consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Ronald Calabrese as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

Please shorten the Discussion and remove the Trans-Tango data as suggested by Reviewer #2.

Reviewer #1:

The authors have addressed my concerns.

Reviewer #2:

The authors have revised their manuscript, A locomotor neural circuit persists and functions similarly in larvae and adult Drosophila, which details the persistence of a circuit motif across metamorphosis, undergoing remodeling while maintaining a similar function. In my opinion, the writing has been significantly improved by raising and supporting the notion that remodeling on this scale has not been observed, both in the introduction and the discussion. The authors also found a positive, provocative way to suggest MDN-Pair1 participate in decision-making by citing recent results in C. elegans. The discussion is currently long and slightly repetitive with the intro and text; I would recommend another edit to shorten. For example, a paragraph dedicated to the identification of transcription factors in the discussion is unnecessary. Finally, the paper reads more clearly as an extension of Carreira-Rosario et al., 2018, which makes it more appropriate for an eLife Advance.

I appreciate the addition of the Trans-Tango data, but they are of limited utility since downstream neurons are not clearly labelled or identifiable, and what labelling does exist could indicate downstream partners of the sensory neurons labeled by the Pair1 Gal4 line. I noticed in the results that the authors dissected the adult CNS 4 days post-eclosion. However, it generally takes much longer for the trans-tango signal to appear. The original paper (Talay et al., 2017) used 15-20 day old flies. I recommend that the authors' either repeat the experiments with a positive control to ensure that trans-tango is working, or remove them from the manuscript entirely.

https://doi.org/10.7554/eLife.69767.sa1

Author response

Essential revisions:

This is a solid piece of work but the advance as presented is limited.

1. A primary concern is that there is no broader consideration of other connections of MDN in the larval transformation to adult. The focus on one specific connection being preserved may not represent the overall scale of remodeling during metamorphosis. There are a variety of experiments proposed that could provide information on this issue and give stronger insight into what types of connections are preserved/reconnected. For example, the trans-tango experiments proposed, even if they don't allow the authors to identify every connected cell-type, could provide a sense of scale. Part of this could be fixed by text revisions if specific experiments are not feasible.

We add several findings to address this interesting question. First, we document Pair1 trans-tango expression, showing Pair1 output neurons for both larva and adults. Pair1 has relatively few downstream partners at either larvae and adult stages (new Figure 1 – Supplement 2). This is substantiated by TEM analysis in larvae, where we find only two Pair1 output neurons that are bilateral with >1 synapse: the A27h and DN_mx (in addition to a lot of unknown neuron fragments). In the adult, none of the Pair1 output neurons are in similar anatomical locations as larval Pair1 output neurons. Thus, it is unlikely for there to many additional Pair1 output neurons that persist from larva to adult. Second, at least one of these Pair1 downstream partners, A27h, is not found in the adult (updated Figure 5). These data suggests that an interneuron-interneuron connection persisting from larvae to adulthood may be unusual, rather than common.

2. A second concern is the breadth of TF staining suggesting that the specific TFs observed may not be controlling connectivity.

We agree that Pair1 does not uniquely express Hb/Bcd/Scr. Quantification has been added to show that within the SEZ approximately 10-40% of cells are triple-positive neurons (updated Figure 2). Interestingly, only the SEZ has triple-positive neurons.

3. A third concern is the framing of the MDN-Pair1 connection as a decision-making circuit. This concern is best addressed by text revision. The detailed comments of the reviewers will provide specific reasoning/recommendations to the authors.

We have removed all mentions of the a “decision-making” circuit from the text.

Reviewer #1:

This study from Lee and Doe examines the conservation of morphology, function, and connectivity in a Drosophila neural circuit that promotes backward locomotion. The study builds on a previous eLife paper from this group, Carreira-Rosario et al. 2018, which showed that a set of descending neurons- the Moonwalker Descending Neurons (MDN)- were conserved between larval and adult stages and performed a similar function in initiating backward locomotion, despite the very different locomotor strategies of crawling larvae and walking adults. Here, the authors examine a second neuron type- Pair1- that is downstream of MDN and inhibits forward locomotion. They first use a genetic immortalization strategy to show that the morphology of this neuron is similar between larvae and adults, although many of its processes are pruned during pupation. They next show that a set of transcription factor markers in these neurons are conserved between larval and adult stages. Finally these show that optogenetic activation of Pair1 in adults generates a similar stopping phenotype as in larvae, and that MDN and Pair 1 and synaptically connected (via GRASP) in adults as well. The authors conclude that these neurons form a core "decision-making" circuit that is conserved between developmental stages, and speculate that the continued expression of the same transcription factors may allow these neurons to re-form synapses after pupation.

Overall this is a clear study that provides new data on which parts of the fly locomotor circuit are maintained across developmental stages. Together with on-going work in other parts of the fly brain, and across insect species, I expect this work to contribute to our understanding of how neural circuits become adapted through genetic and developmental mechanisms to different organismal niches.

The data are generally strong and convincing. One concern I have is that the transcription factor stains in Figure 2 show that most of the surrounding cell bodies share the same pattern of expression for the 5 TFs examined, with Hb, Scr, and Bcd highly expressed and Vsx1 and Nab absent. This might impact the authors' model that TF expression patterns in Pair1 allow it to make correct synaptic connections with MDN. The authors should address whether they expect these surrounding neuron to make similar synaptic connections, or whether connectivity is determined by a larger group of TFs other than those shown here. Experiments manipulating TF expression and looking at effects on connectivity would strength this part of the authors' model.

See general comment 2, above. We agree that Pair1 does not uniquely express Hb/Bcd/Scr. Quantification has been added to show that within the SEZ approximately 10-40% of cells are triple-positive neurons (new Figure 2).

An additional concern is that the MDN-Pair1 neurons are referred to multiple time in the Discussion as a "decision-making circuit." However this term is poorly defined. The authors should specify what they mean by this term and how they would distinguish a "decision-making circuit" from other central circuits. More broadly, it would be helpful to be specific about what central circuit features they would expect to be conserved between larvae and adults or to state clearly if there too little data yet to make broad predictions of this type.

We agree with this concern and have updated the text to exclude the phrase “decision-making circuit”.

Reviewer #2:

This paper asks whether a neural connection that exists in the brain of the Drosophila larva also forms in the adult brain and serves a similar function. Past work (Carreira-Rosario et al., 2018) showed that in larvae, the mooncrawler descending neuron (MDN) drives the transition from forward to backward crawling in part through its connection onto the bilateral pair of Pair1 neurons. Here the authors find that Pair1 Gal4 lines label a pair of adult descending neurons with cell bodies in the gnathal ganglia which express the same developmental markers as the larval Pair1 neurons. Positive GRASP staining using MDN and Pair1 Gal4 lines suggests the adult neurons are synaptically connected. Furthermore, optogenetic activation of Pair1 neurons causes adult flies to stop walking or moving, similar to the effect found for Pair1 activation in larvae. They show that Pair1 neurons undergo dramatic remodeling of their axonal arbors in the ventral nerve cord (VNC).

This paper extends the work of Carreira-Rosario et al., (2018), by following the Pair1 neurons through development. As a conceptual advance, it is an interesting finding that Pair 1 neurons play similar functional roles in adults as in larvae, even though the downstream targets must be entirely different. Pair1 neurons were shown to form inhibitory synapses onto premotor neurons A27h in posterior abdominal segments, which themselves drive muscle contraction and initiate forward crawling. A27h undergo apoptosis during pupal development, abdominal segments lose their role in locomotion, and Pair1 develops extensive arbors in the adult T1 where locomotion is likely initiated.

The work addresses important and interesting questions, but its scope is limited. While the authors identify some similarities in connectivity and function of Pair1 neurons across metamorphosis, the characterization of the Pair1 neurons is far from complete.

I feel that two major revisions would improve the paper. First, there are some simple experiments that would help broaden the scope and impact of the study. The authors could survey the downstream partners of Pair1 in the T1 thoracic neuromere using Trans-tango. They could also ask whether silencing Pair1 (e.g. driving Kir2.1 expression) causes behavioral deficits, to test whether pair 1 neurons are necessary for the switch between forward and backward walking.

We thank the reviewer for suggesting these experiments. We have included data showing a small number of Pair1 downstream partners labeled by Trans-tango in the adult (Figure 1 – Supplement 2). Given that silencing Pair1 neurons in larvae did not lead to any behavioral differences (Carreira-Rosario et al., 2018), we did not test this in the adult.

The paper would also benefit from a deeper exploration of the degree to which the preservation of MDN to Pair1 connectivity is representative or extraordinary. For example, are there connectivity motifs from the larvae that the authors did not find in the adult? Including these negative results would help provide context. Many neurons (e.g. thoracic motor neurons) gain a role in adulthood, others lose a role (e.g. abdominal motor neurons no longer control "locomotion"). What would one expect if it were possible to compare all the inputs and outputs of Pair1, or other fly neurons?

We tested multiple neurons from the larval MDN circuit for a role in the adult MDN circuit, and have added that data to the manuscript (updated Figure 5). From the larval MDN circuit, the A27h neurons do not persist into adulthood, and the A18b neurons are present in the adult but are not part of the MDN circuit. Yet it remains unclear whether maintaining anatomical and behavioral similarity from larva to adult is a rare or common phenomena. Either way is interesting: if rare, it leads to the question of how does persistent connectivity survive metamorphosis in the MDN-Pair1 circuit but not other circuits; if common, it provides an excellent model system to study a widely used mechanism. This is an important question for future study!

Given that the paper builds on a previous study, it could also be improved by editing for concision. There are some sections that distract from the key take-home message. For example, the authors state that this circuit plays a role in making decisions. I understand the point, that switching from forward to backward locomotion indicates that a decision has been computed, but it begs questions that are not addressed, such as what exactly constitutes a decision? Where is the decision being made? What mechanism, like a threshold, or competition between neurons, underlie when a decision takes place? Can adult flies stop (Pair1 activation by MDN) without moving backwards (LBL40, LUL130)? If the paper focuses on the developmental questions, provides background on what other interesting motifs are seen as circuits remodel, and position the results within those broader themes, I feel I would have a clearer sense of the novelty of these findings.

Thank you, we agree with these comments. We have removed all mention of decision-making circuits, because (as the reviewer points out) we have not provided sufficient context for this claim. Regarding novelty, we emphasize that this is the first interneuronal circuit known to persist from larva to adult while maintaining a similar function. We mention a few sensory-motor reflex circuits that are maintained, and we discuss whether larval Kenyon cells maintain the same inputs (DANs) and outputs (MBONs). Although all three classes of interneurons are maintained from larva to adult, it remains unknown whether the same neuronal partners are maintained, e.g. whether a specific pair of DAN-KC neurons in the larva will reform the same pairing in the adult.

Issues:

Pair1 optogenetic activation causes larvae to pause briefly, whereas adults stop. The authors speculate it's due to circuitry. Could temporal differences in Channelrhodopsin kinetics and light protocols explain this?

We thank the reviewer for this comment. We have added some discussion on the changes in the behavior between larvae and adult. Many differences do exist between the larvae and adult experiments – in addition to the differences in Channel rhodopsin kinetics and the light protocol, the adult flies also have a cuticle that makes light more difficult to pass through, whereas larvae are transparent. Even with the complicated nature of these experiments, the result that the adult flies pause for the entire duration of red-light exposure is very striking, making the differences in circuitry between larvae and adults our primary hypothesis.

Some edits:

This neuron, named Moonwalker/Mooncrawler Descending Neuron

(MDN) is present in two bilateral pairs per brain lobe, – Awkward phrasing.

Thank you, corrected on page 2, lines 50-51. We now say “This neuron, named Mooncrawler/Moonwalker Descending Neuron (MDN) is present as a bilateral neuronal pairs in each brain lobe".

"Halting forward locomotion is done via activation of the Pair1 descending

interneuron, which inhibits the A27h premotor neuron, to prevent it from inducing forward locomotion (Carreira-Rosario et al., 2018)." – Unclear.

Thank you, corrected on page 2, lines 54-57. We now say “Halting forward locomotion is achieved by via activation of the Pair1 descending interneuron, which inhibits the A27h premotor neuron. Given that the A27h interneuron is required for forward locomotion, its inhibition via MDN-induced Pair1 activation prevents forward locomotion".

"We find that all of these questions are answered in the affirmative, showing that the core MDN-Pair1 decision-making circuit (a pair of synaptically-connected interneurons) persists from larva to adult, despite profound remodeling during metamorphosis, and that this circuit coordinates forward/backward locomotion in both larvae and adults." – see point above.

"induce backward locomotion via the coordinate arrest of forward locomotion" – unclear.

Thank you, corrected on page 2, lines 53-54. We now say “Larval MDNs function within a neural circuit that induces backward locomotion and coordinately arrests forward locomotion".

77 – "but virtually all of the dendridic processes and descending axonal process had been pruned (Figure 1C, only one neuron labeled)."

Thank you, corrected on page 3, line 79. We now say “but virtually all of the dendridic processes and descending axonal process are pruned".

107 – "We conclude that the Pair1 neurons are present from larval to adult stages, and that the Pair1 neurons are enriched for postsynaptic partners in the T1

neuromere." – "enriched" is a strange word in this context.

Thank you, corrected on Page 3, line 109. We now say "Pair1 neurons have many postsynaptic partners in the T1 neuromere."

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

Please shorten the Discussion and remove the Trans-Tango data as suggested by Reviewer #2.

Reviewer #2:

The authors have revised their manuscript, A locomotor neural circuit persists and functions similarly in larvae and adult Drosophila, which details the persistence of a circuit motif across metamorphosis, undergoing remodeling while maintaining a similar function. In my opinion, the writing has been significantly improved by raising and supporting the notion that remodeling on this scale has not been observed, both in the introduction and the discussion. The authors also found a positive, provocative way to suggest MDN-Pair1 participate in decision-making by citing recent results in C. elegans. The discussion is currently long and slightly repetitive with the intro and text; I would recommend another edit to shorten. For example, a paragraph dedicated to the identification of transcription factors in the discussion is unnecessary. Finally, the paper reads more clearly as an extension of Carreira-Rosario et al., 2018, which makes it more appropriate for an eLife Advance.

I appreciate the addition of the Trans-Tango data, but they are of limited utility since downstream neurons are not clearly labelled or identifiable, and what labelling does exist could indicate downstream partners of the sensory neurons labeled by the Pair1 Gal4 line. I noticed in the results that the authors dissected the adult CNS 4 days post-eclosion. However, it generally takes much longer for the trans-tango signal to appear. The original paper (Talay et al., 2017) used 15-20 day old flies. I recommend that the authors' either repeat the experiments with a positive control to ensure that trans-tango is working, or remove them from the manuscript entirely.

We have precisely followed the reviewer’s suggestions in the text: (1) we remove the trans-tango data; (2) we shorten the discussion, including removing reference to trans-tango and deleting two entire paragraphs including the one reviewer suggested (on molecular markers).

https://doi.org/10.7554/eLife.69767.sa2

Article and author information

Author details

  1. Kristen Lee

    Institute of Neuroscience, Howard Hughes Medical Institute, University of Oregon, Eugene, United States
    Contribution
    Conceptualization, Resources, Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-4527-6468
  2. Chris Q Doe

    Institute of Neuroscience, Howard Hughes Medical Institute, University of Oregon, Eugene, United States
    Contribution
    Conceptualization, Data curation, Supervision, Funding acquisition, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    cdoe@uoregon.edu
    Competing interests
    Reviewing editor, eLife
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5980-8029

Funding

Howard Hughes Medical Institute

  • Chris Q Doe

National Institutes of Health (HD27056)

  • Kristen Lee

NIH (P40OD018537)

  • Chris Q Doe

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank John Reinitz, Claude Desplan, and Stefan Thor for antibodies; Barry Dickson, Matthieu Loius, and Vivek Jayaraman for fly stocks. Transgenic lines were generated by BestGene (Chino Hills, CA) or Genetivision (Houston, TX). Stocks obtained from the Bloomington Drosophila Stock Center (NIH P40OD018537) were used in this study. We thank Dr Sen-Lin Lai for the immortalization fly stock, and Sen-Lin Lai, Emily Heckman, and Arnaldo Carreira-Rosario for comments on the manuscript. Funding was provided by HHMI (CQD, KML).

Senior and Reviewing Editor

  1. Ronald L Calabrese, Emory University, United States

Publication history

  1. Preprint posted: April 27, 2021 (view preprint)
  2. Received: April 27, 2021
  3. Accepted: July 13, 2021
  4. Accepted Manuscript published: July 14, 2021 (version 1)
  5. Accepted Manuscript updated: July 15, 2021 (version 2)
  6. Version of Record published: July 22, 2021 (version 3)

Copyright

© 2021, Lee and Doe

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 749
    Page views
  • 108
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Neuroscience
    Domino K Schlegel et al.
    Research Article

    The RLBP1 gene encodes the 36 kDa cellular retinaldehyde binding protein, CRALBP, a soluble retinoid carrier, in the visual cycle of the eyes. Mutations in RLBP1 are associated with recessively inherited clinical phenotypes, including Bothnia dystrophy, retinitis pigmentosa, retinitis punctata albescens, fundus albipunctatus, and Newfoundland rod-cone dystrophy. However, the etiology of these retinal disorders is not well understood. Here, we generated homologous zebrafish models to bridge this knowledge gap. Duplication of the rlbp1 gene in zebrafish and cell-specific expression of the paralogs rlbp1a in the retinal pigment epithelium and rlbp1b in Müller glial cells allowed us to create intrinsically cell type-specific knockout fish lines. Using rlbp1a and rlbp1b single and double mutants, we investigated the pathological effects on visual function. Our analyses revealed that rlbp1a was essential for cone photoreceptor function and chromophore metabolism in the fish eyes. rlbp1a mutant fish displayed reduced chromophore levels and attenuated cone photoreceptor responses to light stimuli. They accumulated 11-cis and all-trans-retinyl esters which displayed as enlarged lipid droplets in the RPE reminiscent of the subretinal yellow-white lesions in patients with RLBP1 mutations. During aging, these fish developed retinal thinning and cone and rod photoreceptor dystrophy. In contrast, rlbp1b mutants did not display impaired vision. The double mutant essentially replicated the phenotype of the rlbp1a single mutant. Together, our study showed that the rlbp1a zebrafish mutant recapitulated many features of human blinding diseases caused by RLBP1 mutations and provided novel insights into the pathways for chromophore regeneration of cone photoreceptors.

    1. Neuroscience
    Kai Hwang et al.
    Research Article Updated

    Hubs in the human brain support behaviors that arise from brain network interactions. Previous studies have identified hub regions in the human thalamus that are connected with multiple functional networks. However, the behavioral significance of thalamic hubs has yet to be established. Our framework predicts that thalamic subregions with strong hub properties are broadly involved in functions across multiple cognitive domains. To test this prediction, we studied human patients with focal thalamic lesions in conjunction with network analyses of the human thalamocortical functional connectome. In support of our prediction, lesions to thalamic subregions with stronger hub properties were associated with widespread deficits in executive, language, and memory functions, whereas lesions to thalamic subregions with weaker hub properties were associated with more limited deficits. These results highlight how a large-scale network model can broaden our understanding of thalamic function for human cognition.