The neuronal building blocks of the navigational toolkit in the central complex of insects
Introduction
Spatial orientation is perhaps the most fundamental and vital task of the nervous system of all nonsessile animals. Any goal-directed behavior, such as food search, mate-finding, escaping, or homing, requires prior knowledge of the orientation of the animal’s body in space with respect to a defined external system of cues. Therefore, the nervous system has to hold internal representations of the axes of the body that is attached to it (Figure 1a). It also has to encode the heading direction: the spatial relationship between the sensory organs, which are often located on the head, and external stimuli that can be tracked. This heading direction does not always coincide with the traveling direction (Figure 1b). Therefore, it is important, especially for more sophisticated orientation behaviors, such as vector navigation, that the neuronal system also tracks the actual travel direction of the body. It is further useful (and for vector navigation indispensable) to be able to estimate travel velocity and integrate it over time to gauge distances. Many of the representations and mechanisms outlined above have specific neuronal correlates in the central complex (CX).
Volumetric movement (i.e. movement in 3D space) entails translation along or rotation around the three cardinal axes of the body (i.e. it has six degrees of freedom, Figure 1a). While the ability to perform a stable flight generally requires control over all six degrees of freedom, traveling between two points requires mainly control over movements in the horizontal plane. Although there is some evidence that neurons in the CX are not only tuned to azimuth (i.e. orientation in the horizontal plane), but also to elevation of visual stimuli 1, 2, 3, so far, most neuronal correlates of directional coding that have been identified in the CX, were limited to the horizontal plane 4, 5, 6, 7, 8••, 9••. One possible reason for this is that insects stabilize their gaze in the rollandpitch axes through compensatory head movements using visual and mechanosensory feedback, thereby reducing the input to the head-centered sensory systems to the horizontal plane 10, 11. A second possibility is that the restraining of insects to this plane in physiological setups has so far precluded the discovery of a three-dimensional system as described in mammals [12].
Electrophysiological work in crickets, monarch butterflies, dung beetles, bees, and particularly in locusts has elucidated many sensory and physiological properties of central-complex neurons in the past 20 years 13, 14, 15, 16, 17, 18, 19, 1, 3. In the past decade, the genetic toolkit available in Drosophila has advanced our understanding of the inner workings of the central- complex circuits at a breathtaking pace. In this review, I aim to provide a guided tour through the central complex, following the flow and processing of spatial information rather than providing a comprehensive overview of recent publications, which has been nicely done elsewhere 20, 21, 22.
Section snippets
Nomenclature
Parallel research in Drosophila and other species has led to different naming conventions for neuropils and neurons. Here, I will use the current Drosophila nomenclature [23], adding the alternate names, where appropriate. A lookup table for conversion between the systems can be found in [24].
General anatomy
The CX (Figure 1c) consists of the fan-shaped body (FB)/upper division of the central body (CBU), the ellipsoid body (EB)/lower division of the central body (CBL), the protocerebral bridge (PB), and the paired noduli (NO). The EB receives input mainly through the bulbs, which can consist of several substructures as in flies, butterflies, and beetles, or can be separated into medial and lateral bulbs as in bees and locusts 25, 21. FB, EB, and PB are divided into a variable number of vertical
The input stage: ER (TL) neurons
Sensory information enters the EB through tangential neurons that receive input in the bulbs or the lateral accessory lobes (LAL) and provide output to all columns of the EB (EB ring neurons, ER/tangential neurons of the CBL, TL, Figure 2) 26, 25. The majority of these neurons encode visual information, but a subset processes antennal mechanosensory stimuli 23••, 27•• or is involved in thermoresponsive motor behavior [28]. In many species, these neurons are sensitive to the plane of linearly
Heading direction: EPG (CL1a)
The next element downstream from ER neurons are ellipsoid body-protocerebral bridge-gall (EPG) neurons (Figure 2), which receive input in the EB and project to the PB and the gall 5, 36. These neurons are termed columnar neurons of the CBL type 1 (CL1a) in other species. EPG neurons are regarded as the head-direction neurons of the insectorientation-circuit 5, 37, 38, 39, 40. The activity of the population of these neurons manifests as a single ‘bump’ of activity in the EB, which, like the
Heading direction and rotational velocity: PEN (CL2)
When the animal turns, that is, it rotates around its yaw axis, the internal representation of head-direction in the EB needs to get updated. This requires a neuronal representation of rotation velocity, which is found in protocerebral bridge-ellipsoid body-noduli (PEN) neurons, termed columnar neurons of the CBL type 2 (CL2) in other species (Figure 2). PEN neurons receive input from EPG neurons in the EB and connect to the NO and the PB [23]. PEN and EPG neurons that branch in the same column
Reformatting and distribution of heading signal: Δ7 (TB)
Δ7 neurons are prominently characterized by having two output areas that are separated by seven columns that are either input areas or have no branches (4, 42, Figure 2). In locusts, the homologs of these neurons are called tangential neurons of the PB type 1 (TB1) were described to form an ordered representation of polarized light-orientation angles in the PB, which was the first description of a functional correlate of the columnar structure of the CX [4]. Δ7 Neurons are a central element of
Translational velocity: LNO (TN), SpsP
Processes described so far were only considering information about azimuth (heading) and changes thereof (rotational velocity). However, if an animal is moving, the orientation network also needs to incorporate information about the direction and velocity of translation. Translational-velocity information enters the CX through lateral accessory lobe-noduli (LNO) neurns, termed tangential neurons of the noduli (TN) in other species, that project from the LAL to specific compartments of the
Heading direction and translational velocity: PFN (CPU4)
Protocerebral bridge-fanshaped-body-noduli (PFN) neurons receive input in the PB and the NO and have outputs in the FB (42, 9••, 8••, Figure 2). These neurons are called columnar neurons of the CBU type 4 (CPU4) in other species. They are the most numerous columnar cell type in the CX (Drosophila: 437, bumblebee: 854) 23••, 24•. Many PFN subtypes respond to visual stimuli as well as airflow [43]. Two subtypes (PFNv and PFNd) have been shown to be conjunctively tuned to heading information and
Traveling direction: hΔB
The target of PFNv and PFNd neurons in the FB are neurons in the FB, termed hΔB 23••, 8••, 9••. Like other types of hΔ (or pontine) neurons, these connect two columns of the FB that are offset by half the width of the FB (Figure 3b, right). While EPG neurons encode the heading direction of the animal, that is, the direction that the head is facing with relation to an external frame of reference, hΔB neurons encode the actual direction of motion within that reference frame (Figure 3c). That
Conclusions
In the past years, huge advancements have been made in understanding the neuronal underpinnings of spatial orientation of insects at the neuronal level. Most of these studies were focused on circuits of the EB and PB. While increasingly more data from FB neurons are contributing to the understanding of the CX, the role of this neuropil is still poorly understood. Recent findings support the idea that state-dependency and context-dependency is incorporated into the system here, probably through
Conflict of interest statement
The author declares no conflict of interest.
Acknowledgements
I would like to thank Tanya Wolff, Stuart Berg, Hideo Otsuna, and Shin-ya Takemura, for providing the 3D model of the Drosophila central complex. Funding was obtained from the Deutsche Forschungsgemeinschaft (DFG, Germany), grant number: PF 714/5-1.
References and recommended reading (54)
- et al.
Receptive fields of locust brain neurons are matched to polarization patterns of the sky
Curr Biol
(2014) - et al.
Polarized skylight navigation in insects: model and electrophysiology of e-vector coding by neurons in the central complex
J Neurophysiol
(2008) - et al.
Neural coding underlying the cue preference for celestial orientation
Proc Natl Acad Sci
(2015) - et al.
Mechanisms underlying the neural computation of head direction
Annu Rev Neurosci
(2020) - et al.
A visual pathway for skylight polarization processing in Drosophila
eLife
(2021) - et al.
Network encoding heading and steering maneuvers in Drosophila
Neuron
(2020) - et al.
Mushroom bodies of the cockroach: their participation in place memory
J Comp Neurol
(1998) - et al.
Feature detection and orientation tuning in the Drosophila central complex
Nature
(2013) - et al.
Matched-filter coding of sky polarization results in an internal sun compass in the brain of the desert locust
Proc Natl Acad Sci USA
(2020) - et al.
Maplike representation of celestial E-vector orientations in the brain of an insect
Science
(2007)
Neural dynamics for landmark orientation and angular path integration
Nature
Angular velocity integration in a fly heading circuit
eLife
A neural circuit architecture for angular integration in Drosophila
Nature
Building an allocentric travelling direction signal via vector computation
Nature
Transforming representations of movement from body- to world-centric space
Nature
Vision and the organization of behaviour
Curr Biol
3-D maps and compasses in the brain
Annu Rev Neurosci
Sun compass integration of skylight cues in migratory monarch butterflies
Neuron
Sun compass neurons are tuned to migratory orientation in monarch butterflies
Proc Biol Sci
Neurons of the central complex of the locust Schistocerca gregaria are sensitive to polarized light
J Neurosci
Two compasses in the central complex of the locust brain
J Neurosci
Responses of compass neurons in the locust brain to visual motion and leg motor activity
J Exp Biol
Flexible navigational computations in the Drosophila central complex
Curr Opin Neurobiol
The insect central complex and the neural basis of navigational strategies
J Exp Biol
A connectome of the Drosophila central complex reveals network motifs suitable for flexible navigation and context-dependent action selection
eLife
A projectome of the bumblebee central complex
eLife
Cited by (9)
Variations on an ancient theme — the central complex across insects
2024, Current Opinion in Behavioral SciencesThe neurobiology of the Monarch butterfly compass
2023, Current Opinion in Insect ScienceEditorial overview: Insect neuroethology: More than behavior and neurons
2023, Current Opinion in Insect ScienceAnatomical organization of the cerebrum of the praying mantis Hierodula membranacea
2024, Journal of Comparative NeurologyNeural representation of goal direction in the monarch butterfly brain
2023, Nature CommunicationsNeuroarchitecture of the central complex in the Madeira cockroach Rhyparobia maderae: Pontine and columnar neuronal cell types
2023, Journal of Comparative Neurology