Download Developmental organization of the mushroom bodies of Thermobia

EVOLUTION & DEVELOPMENT
7:2, 150 –159 (2005)
Developmental organization of the mushroom bodies of Thermobia
domestica (Zygentoma, Lepismatidae): insights into mushroom body
evolution from a basal insect
Sarah M. Farris
Department of Biology, West Virginia University, Morgantown, WV 26506, USA
Correspondence (email: [email protected])
SUMMARY The mushroom bodies of the insect brain are
sensory integration centers best studied for their role in learning
and memory. Studies of mushroom body structure and
development in neopteran insects have revealed conserved
morphogenetic mechanisms. The sequential production of
morphologically distinct intrinsic neuron (Kenyon cell)
subpopulations by mushroom body neuroblasts and the
integration of newborn neurons via a discrete ingrowth tract
results in an age-based organization of modular subunits in the
primary output neuropil of the mushroom bodies, the lobes.
To determine whether these may represent ancestral
characteristics, the present account assesses mushroom
body organization and development in the basal wingless
insect Thermobia domestica. In this insect, a single calyx
supplied by the progeny of two neuroblast clusters, and three
perpendicularly oriented lobes are readily identifiable. The lobes
are subdivided into 15 globular subdivisions (Trauben). Lifelong
neurogenesis is observed, with axons of newborn Kenyon cells
entering the lobes via an ingrowth core. The Trauben do not
appear progressively during development, indicating that they
do not represent the ramifications of sequentially produced
subpopulations of Kenyon cells. Instead, a single Kenyon cell
population produces highly branched axons that supply all lobe
subdivisions. This suggests that although the ground plan for
neopteran mushroom bodies existed in early insects, the
organization of modular subunits composed of separate
Kenyon cell subpopulations is a later innovation. Similarities
between the calyx of Thermobia and the highly derived fruit fly
Drosophila melanogaster also suggest a correlation between
calyx morphology and Kenyon cell number.
INTRODUCTION
ulation that allows the wings to be folded (Borror et al. 1989).
The phylogenetically older Paleoptera (dragonflies and mayflies) lack this articulation, whereas the basal Apterygota
(bristletails and firebrats) are primitively wingless.
Beyond the basic groundplan described above, mushroom
bodies display extensive morphological diversity across different insect lineages (reviewed in Strausfeld et al. 1998).
Dedicated input regions (calyces) may be present or absent,
single or double. Mixed input/output structures (lobes) vary
in size, shape, and number. There is strong evidence that some
aspects of this diversity are correlated with the number and
type of sensory inputs to the mushroom bodies, which are in
turn associated with functional demands imposed by a species’ behavioral ecology. The social Hymenoptera display a
particularly robust interconnection between behavior, afferent
supply, and mushroom body morphology. Species that depend on visual navigation during foraging, such as the honey
bee (Apis mellifera), have extensive input from optic neuropils
to a specialized zone in the calyx called the collar (Mobbs
1984; Gronenberg 2001; Ehmer and Gronenberg 2002). In
The mushroom bodies are distinctive lobed neuropils in the
brains of an array of arthropod and invertebrate groups
(Holmgren 1916; Hanström 1940; Strausfeld et al. 1995),
where they are defined by the presence of minute intrinsic
neurons that form a dense protocerebral neuropil with hundreds or thousands of parallel-projecting axon-like processes
(Strausfeld et al. 1995, 1998; Fig. 1). Insect mushroom bodies
are the best studied, and have been ascribed a number of
higher processing functions. These include context-dependent
multimodal sensory integration (Schildberger, 1984; Li and
Strausfeld 1997, 1999), prediction and monitoring of motor
behavior (Mizunami et al. 1998a; Okada et al. 1999), place
memory (Vowles 1964; Mizunami et al. 1998b), context generalization (Liu et al. 1999), and olfactory learning and memory (reviewed by Waddell and Quinn 2001; Heisenberg 2003).
These studies were performed in neopteran species, which
represent the greatest and most recent radiation of the insects
and are characterized in part by the acquisition of an artic150
& BLACKWELL PUBLISHING, INC.
Farris
Fig. 1. Schematic diagram of Kenyon cell morphology in representative neopteran and apterygote brains. (A) Sagittal representation of the mushroom body of the cockroach, Periplaneta
americana, illustrating general neopteran mushroom body morphology and the developmental organization of four distinct cell
types (Class I–III, and newborn) by age in the calyx (Ca), pedunculus (P), and lobes (V, vertical; M, medial). (B) Sagittal representation of the mushroom body of Thermobia domestica,
showing the invariant arrangement of Trauben (numbered) on
the medial, vertical, and horizontal (H) lobes. Thermobia Kenyon
cells (purple) do not segregate into discrete lobe subdivisions, but
appear to provide processes to all of the Trauben via a single,
highly branched axon. MbNb, mushroom body neuroblasts; Core,
ingrowth core composed of axons of newborn Kenyon cells; Collaterals, transient axon branches produced by core Kenyon cells; g
layer, glia-delineated lobe subdivision supplied by axons of Class II
Kenyon cells; Acc Ca, accessory calyx of Class III Kenyon cells;
Lobelet, vertical lobe tract of Class III Kenyon cells. (A) was reprinted from Arthropod Structure and Development, 32, S. M.
Farris and I. Sinakevitch, Development and evolution of the insect
mushroom bodies: towards the understanding of conserved developmental mechanisms in a higher brain center, Copyright (2003),
with permission from Elsevier.
ants, in which foraging is more dependent on olfactory cues,
visual input is diminished whereas olfactory input is increased.
Subsequently, the collar region of the mushroom body calyx
is greatly reduced, whereas the lip, which receives olfactory
input, is expanded (Gronenberg and Hölldobler 1999;
Gronenberg 2001).
To determine the interplay between cellular, developmental, and functional factors during mushroom body evolution,
homologous aspects of mushroom body organization must be
identified across taxa, and the ancestral or derived status of
these traits determined. For example, all neopteran mushroom bodies studied to date are characterized by a particular
pattern of morphogenesis, suggesting that this might characterize a universal feature of development and organization
(Farris and Sinakevitch 2003). In each species, three or more
Kenyon cell subpopulations are produced in an invariant
Mushroom bodies of Thermobia
151
sequence during mushroom body development and occupy
discrete areas of the lobes and calyces (Schürmann 1973;
Mobbs 1982; Ito et al. 1997; Farris et al. 1999, 2004; Lee et al.
1999; Strausfeld and Li 1999b; Farris and Strausfeld 2001;
Kurusu et al. 2002; Malaterre et al. 2002; Strausfeld 2002;
Farris and Strausfeld 2003; Strausfeld et al. 2003; Zhu et al.
2003). For example, the three major lobe systems of the
mushroom bodies in the fruit fly Drosophila melanogaster are
made up of three Kenyon cell populations produced in the
embryo/early larva, late larva, and pupal stage, respectively
(Lee et al. 1999). As a result, the adult mushroom bodies are
composed of age-based layers consisting of morphological
and functional modules of intrinsic neurons.
Another shared feature is the mushroom body calyx, composed of Kenyon cell dendrites and receiving input from primary sensory neuropils, particularly from the antennal lobe
via the inner antennocerebral tract (iACT; Mobbs 1982; Ito
et al. 1998; Li and Strausfeld 1999; Gronenberg 2001; Ehmer
and Gronenberg 2002). Evolutionary relationships among the
many calyx morphologies, however, are not clear. For example, calyx doubling is observed in both basal and derived
species representing widely divergent lineages (such as
the cockroach Periplaneta americana and the honey bee
A. mellifera; for a review, see Strausfeld et al. 1998). The
quadripartite organization of the single calyx in D. melanogaster has led to the suggestion that it arose from the secondary
fusion of two ancestral calyces, each additionally composed of
two hemicalyces (Yang et al. 1995). Based on comparisons of
cellular morphology between the single primary calyx of Orthoptera and the double calyces of Periplaneta, however,
Weiss (1981) has proposed that double calyces are the derived
condition, at least in these closely related insect lineages.
Further insight into the homology of mushroom body
components may be gleaned by comparisons of neopterans
with phylogenetically basal non-neopterans. Little is currently
known about the mushroom bodies of basal paleopteran or
apterygote insects. The apterygotes, which are further divided
into the Orders Archeognatha (the bristletails, also called
Microcoryphia) and Zygentoma (firebrats and silverfish), are
considered to be among the most basal of insect lineages,
having likely arisen in the early to late Devonian, respectively
(Grimaldi 2001; Mendes, 2002; Engel and Grimaldi 2004).
Interestingly, the Archeognatha completely lack mushroom
bodies, whereas in the Zygentoma, mushroom bodies are well
developed (Hanström 1940; Strausfeld et al. 1995, 1998). The
Zygentoma thus represent the most basal living example of
insect mushroom bodies. The goal of the present account is to
assess the developmental organization and cellular composition of primitive mushroom bodies in a representative zygentoman species, the firebrat Thermobia domestica (Family
Lepismatidae). Comparisons with mushroom bodies of neopteran insects are interpreted in terms of potentially ancestral
and derived traits of these complex neuropils.
152
EVOLUTION & DEVELOPMENT
Vol. 7, No. 2, March^April 2005
MATERIALS AND METHODS
Insects
Breeding colonies of T. domestica were maintained in ventilated
cages in an incubator at 351C, on a 12:12 light:dark cycle. Pans of
tap water were placed in the incubator to increase ambient humidity. Meow Mix brand cat food (Meow Mix Co., Seacaucus, NJ,
USA) and Cricket Quencher (Fluker Farms, Port Allen, LA, USA)
were provided as food and water sources, respectively. Cotton balls
were placed in the cages to serve as egg-laying substrates. First
instar nymphs were collected by placing eggs in a separate container and collecting nymphs within 24 h of hatching. The largest
insects (8–10 mm) in the mature breeding colony were identified as
adults.
Histology
Cason’s staining was performed according to the protocol of Kiernan (1990); additional details are provided by Farris and Strausfeld
(2001). Insects were anesthetized with cold, whole heads or brains
were removed under physiological saline (O’Shea and Adams
1981), and fixed for 1–2 h in Carnoy’s fixative followed by paraffin
embedding and sectioning at 8 mm on a rotary microtome. Golgi
impregnations utilized a combined Colonnier/Rapid Golgi protocol described by Li and Strausfeld (1997) and Farris and Strausfeld
(2001). Histological preparations were viewed on a Zeiss AxioSkop
2 brightfield microscope, and images were captured with a Zeiss
AxioCam MRC5 camera and Axiovision 4 software (Carl Zeiss
AG, Oberkochen, Germany). Reconstructions of Golgi-impregnated neurons from 20 mm sections were generated with Z-stack
and Extended Focus modules for the Axiovision software.
50 -Bromo-2-deoxyuridine (BrdU) labeling and detection
Proliferating neuroblasts were visualized by incorporation of BrdU
and subsequent immunodetection according to Farris and
Strausfeld (2001). Twenty-five mg/ml BrdU in O’Shea–Adams saline was injected into anesthetized adults using a tuberculin syringe.
Small nymphs were anesthetized and a droplet of solution was
placed on the mouthparts for approximately 5 min. After BrdU
treatment, insects were returned to the incubator in separate cups
with food and water. Tissue was collected the following day for
paraffin embedding and sectioning prior to immunodetection of
BrdU incorporation. Mouse anti-BrdU primary antibody (ICN
Biomedicals Inc., Aurora, OH, USA) was used at a concentration
of 1:50 and detected by horseradish peroxidase-conjugated goat
anti-rabbit secondary (Jackson Immunochemicals, West Grove,
PA, USA) and the chromogenic substrate 3,30 -diaminobenzidine
(Sigma-Aldrich, St. Louis, MO, USA). Sections were counterstained with a quick incubation (1 min) in Cason’s stain prior to
dehydrating, clearing, and coverslipping in xylene-based mounting
medium.
Anti-DC0 and phalloidin staining
Anti-DC0 polyclonal antibody (a generous gift of Dr. Daniel
Kalderon) is directed against the catalytic subunit of D. melanogaster protein kinase A, which has been shown to be highly
expressed in the fly mushroom bodies (Skoulakis et al. 1993). AntiDC0 reliably labels all Kenyon cell subpopulations in the devel-
oping and adult mushroom bodies of neopteran insects (Farris and
Sinakevitch 2003; Farris and Strausfeld 2003; Farris et al. 2004).
The anti-DC0 primary was used at a 1:1000 concentration and
visualized using a Texas Red-conjugated goat anti-rabbit secondary (Molecular Probes, Eugene, OR, USA). Alexa 488-conjugated
phalloidin applied at a 1:500 concentration (Molecular Probes) labels filamentous actin and has a strong affinity for extending axons
of newborn Kenyon cells in the insect mushroom bodies (Kurusu
et al. 2002; Farris et al. 2004). Fluorescence double staining was
performed as described by Farris et al. (2004) and visualized with a
Zeiss LSM 510 confocal microscope.
DiI fills
Small crystals of 1,10 -dioctadecyl-3,3,30 ,30 -tetramethylindocarbocyanine perchlorate (DiI; Molecular Probes) were applied to the
mushroom bodies of paraformaldehyde-fixed brains using a pulled
glass pipette in order to label intrinsic mushroom body tracts. The
protocol described by Farris et al. (2004) was followed.
Image processing
Additional processing of images for contrast, brightness, and color
balance was performed as needed using Adobe Photoshop 7.0 for
Macintosh (Adobe Systems Incorporated, San Jose, CA, USA).
Schematic diagrams were created in Adobe Illustrator 10.0.
RESULTS
Overview of neopteran mushroom body anatomy
and development
The insect mushroom bodies are distinctive paired neuropils
located in the dorsal protocerebrum of the brain. Mushroom
body intrinsic neurons, called Kenyon cells, are characterized
by small, tightly packed somata located in the dorsoposterior
brain (Fig. 1A). Each Kenyon cell body gives rise to a single
neurite that projects anteroventrally, producing dendritic arborizations that form one or more cup- or knob-shaped calyces. Further distal to the calyx, the neurite forms axon-like
processes in the pedunculus that typically bifurcate to form a
medial and a vertical lobe. Further branching by each axon
within a lobe may occur, but extensive collaterals are typically
observed only on immature axons (Farris and Strausfeld
2001; Farris et al. 2004). In the adult mushroom bodies, distinct Kenyon cell subpopulations can be identified in a given
species based on process morphology, gene expression, and
immunostaining affinities, and afferent and efferent connectivity (reviewed in Farris and Sinakevitch 2003).
Mushroom body development begins with the appearance
of dedicated Kenyon cell progenitors (mushroom body neuroblasts (MbNbs)) in the dorsoposterior protocerebrum of the
late embryo (Noveen et al. 2000; Malaterre et al. 2002). The
MbNbs may form one or more clusters as in the honey bee
(Malun 1998; Farris et al. 1999), or remain separate as in the
Farris
Mushroom bodies of Thermobia
153
fruit fly (Ito and Hotta 1992). Production of Kenyon cells
commonly begins in the embryo and continues until late in
postembryonic development, in some species continuing
throughout adulthood (Cayre et al. 1996). Distinct Kenyon
cell subpopulations as described above are produced sequentially during development (Panov 1957; Farris et al. 1999; Lee
et al. 1999; Cayre et al. 2000; Farris and Strausfeld 2001),
resulting in an age-based layering of Kenyon cell axons according to subpopulation identity in the lobes (Farris and
Strausfeld 2001; Kurusu et al. 2002; Farris et al. 2004). Lobe
substructures such as laminae and lobe pairs represent the
ramifications of distinct subpopulations of Kenyon cells and
thus also appear sequentially during development (Lee et al.
1999; Farris and Strausfeld 2001).
General anatomy of Thermobia mushroom bodies
The overall appearance of the mushroom bodies of T. domestica is readily compared with that of neopteran insects. The
mushroom bodies consist of small cell somata located in the
dorsoposterior brain that supply with their processes a single
calyx, a pedunculus, and lobes (Figs. 1B and 2). The lobes
form three major branches oriented perpendicular to one another and here termed the vertical, medial, and horizontal
lobes. Each lobe is further subdivided into bulbous subdivisions, the ‘‘Trauben’’ (grapes) described by Böttger (1910) and
Hanström (1940) in the closely related lepismatid Lepisma
saccharina. Because of the invariant number and position of
Trauben across individuals, each can be uniquely and reliably
identified (Figs. 1B and 2A).
The adult calyx is roughly globular, with a small dorsal
indentation formed by an apodeme (muscle attachment point)
of the head endoskeleton that passes in close proximity to the
brain (Fig. 2B). In the first instar nymph (Fig. 2B, inset), the
relatively poorly developed calyx wraps around the ventral
and lateral surfaces of the apodeme, forming an asymmetrical
U shape. The majority of Kenyon cell bodies are positioned
medial to this apodeme, with neurites projecting into the calyx
via a central fiber tract (Fig. 2C). The outer surface of the
calyx is composed of glomerular structures. Although they
appeared to be unusually large in Thermobia, microglomerular organization is characteristic of neopteran calyces
(Yusuyama et al. 2002; Frambach et al. 2004). Homology of
the Thermobia calyx with that of neopterans is further supported by the presence of a small tract originating in the
deutocerebrum and extending to the calyx. The tract was
labeled in the adult brain both by phalloidin (Fig. 2D) and by
DiI fills to the deutocerebrum (data not shown), and has a
trajectory identical to that of the iACT that supplies antennal
lobe afferents to the neopteran calyx (for example, see Ito et
al. 1998). Although the methods used did not resolve the exact
origin of iACT neurons in the Thermobia deutocerebrum,
Cason’s staining revealed a small region composed of
Fig. 2. Morphology of Thermobia mushroom bodies. (A) Anterior
frontal section of the medial (M) and vertical (V) lobes of the
mushroom bodies in a Thermobia adult. The horizontal (H) lobe is
not visible, but sprouts from the pedunculus (Pe) prior to the medial–vertical branch point. Arrows indicate Trauben of the main
lobe branches. Dotted line indicates the brain midline. (B) Posterior
frontal section of the mushroom body calyces (Ca) in the same
brain as in (A). Each mushroom body (one per hemisphere) is
equipped with a single roughly spherical calyx. An arrow indicates
apodeme passing through the calyx. (B, inset) Calyx of a first instar
nymph, showing distorted U-shape (brackets) and apodeme passing through the calyx center (arrow). (C) High-magnification frontal view of the adult calyx with glomerular subdivisions in the outer
rind (arrows). Kenyon cell bodies (Kcb) having an average diameter of 3.38 mm (SD 5 0.20 mm, n 5 8) lie dorsal to the calyx and
provide neurites to the calyx core. (D) Anti-DC0/phalloidin double
staining of a frontal section of the calyx and pedunculus. The putative inner antennal cerebral tract (iACT) approaches the calyx
from the ventral brain (arrows). Mus, muscles of the head; P, pedunculus. Scale bars: A, B 5 50 mm; B, inset, C, D 5 20 mm.
glomeruli reminiscent of those in the neopteran antennal lobe
(data not shown). The putative iACT was the only calycal
input tract showing strong phalloidin labeling in the adult.
The affinity of phalloidin for newly extending axons (see Ma-
154
EVOLUTION & DEVELOPMENT
Vol. 7, No. 2, March^April 2005
mushroom bodies (Ito et al. 1997), and are therefore termed
the MbNbs. In adult Thermobia, the BrdU-labeled MbNb
clusters were positioned immediately dorsal to the calyx (Fig.
3B). Further examination of serial sections revealed two
MbNb clusters per hemisphere, arranged anterodorsally (Fig.
3, C and D). The progeny of each MbNb cluster provided
neurites to the single calyx via a separate fiber tract (Fig. 3E).
Morphology and development of Thermobia
Kenyon cells
Fig. 3. Mushroom body progenitor cells. (A) 50 -Bromo-2-deoxyuridine (BrdU) incorporation reveals small clusters of mushroom body neuroblasts (arrows) in the dorsoposterior
protocerebrum of the first instar nymph. The neuroblasts are surrounded by small Kenyon cell bodies (Kcb) that are situated
dorsoposterior to the cell bodies of other protocerebral neurons
(Pcb). (B) BrdU-labeled neuroblasts (arrow) reside directly dorsal
to the calyx (Ca) in the adult. (C and D) Serial oblique sections of
BrdU-labeled neuroblast clusters in a nymph, showing the presence
of two separate proliferative centers (arrows). (E) Anti-DC0 and
phalloidin-stained sagittal section of the adult calyx illustrating two
neuronal tracts (arrows) supplying the single calyx. Each tract
arises from cell bodies (Kcb) produced by a separate neuroblast
cluster (Nb). (F) BrdU incorporation in the mushroom body neuroblasts (arrows) of the pterygote insect Periplaneta americana,
illustrating the more typical ‘‘one proliferative center per calyx’’
organization. LCa, lateral calyx; MCa, medial calyx; Pe, pedunculus. Scale bars 5 20 mm.
terials and Methods) suggests that in contrast to the Neoptera
(Frambach et al. 2004), deutocerebral inputs are added to the
Thermobia mushroom bodies throughout life.
Postembryonic neurogenesis in the Thermobia
mushroom bodies
Lifelong cell proliferation was observed in the mushroom
bodies of Thermobia. BrdU treatment of first instar nymphs
labeled small cell clusters in the dorsoposterior protocerebrum
(Fig. 3A). In Drosophila, progenitors in the dorsoposterior
protocerebrum contribute neurons and glia solely to the
At all stages examined, Kenyon cell axons in the Thermobia
mushroom bodies were highly branched and supplied all
subdivisions of the lobes. DiI crystals applied to the dorsoposterior cell body region filled small ensembles of Kenyon
cells whose processes could be traced into the lobes (Fig. 4A,
inset). Kenyon cell axons labeled in this way invariably provided branches to all three lobes and their Trauben. DiI-filled
axon bundles were tightly packed in the main axis of the
lobes, but fanned out at their tips into the bulbous Trauben.
Additional experiments confirmed the highly branched
nature of Thermobia Kenyon cell axons. Phalloidin staining
revealed a tract of Kenyon cell processes emanating from
presumably newborn neurons located next to each MbNb
cluster and entering the pedunculus and lobes at all postembryonic stages, indicating that the MbNbs continue to produce neurons even in the adult insect. This ‘‘ingrowth core’’
supplied all of the lobes and their Trauben (Fig. 4B). In addition, stochastic labeling of neurons using the Golgi impregnation method repeatedly demonstrated a Kenyon cell
morphology identical to that revealed by DiI and phalloidin
(Fig. 4C). Axon tracts arising from progeny of each of the
MbNb clusters fused in the distal pedunculus to form a single
tract that supplied all parts of the lobes (Fig. 4D). In all cases
where Kenyon cell axons were impregnated, branches entering each Trauben were present.
In the Thermobia calyx, Kenyon cell dendrites were oriented roughly perpendicular to the centrally located main
neurite and extended into the outer glomerular layer (Fig.
4E). With the staining methods used, it was not possible to
determine whether an individual neurite gave rise to dendrites
that projected radially into all regions of the calyx, or that
were restricted to a single radial zone.
Serial sections of anti-DC0/phalloidin-stained brains allowed individual Trauben to be uniquely identified (Fig. 1B).
The branching pattern of the phalloidin-labeled ingrowth core
into anti-DC0-labeled Trauben was invariant throughout
postembryonic development (Fig. 5). Fifteen individual Trauben were observed both in newly hatched first instar nymphs
(Fig. 5, A–C) and in adults (Fig. 5, D–F). Although the lobes
increased in size between these two developmental time
points, the number and orientation of individual Trauben did
not change. This stability further supports the conclusion that
the mushroom bodies of Thermobia are made up of a single,
Farris
Mushroom bodies of Thermobia
155
Fig. 5. The number of lobe subdivisions is invariant throughout
postembryonic development in Thermobia. Five vertical (V), two
horizontal (H), and eight medial lobe (M) Trauben are present in
the first instar nymph (A–C). Adult Trauben show increased size
but the same number and arrangement, indicating that these subdivisions do not represent sequentially generated Kenyon cell populations. Anti-DC0- and phalloidin-stained sagittal sections. Ca,
calyx; Mus, muscle of the head; Pe, pedunculus. Scale bar5 20 mm.
Fig. 4. Morphology of Thermobia Kenyon cells. (A) DiI crystals
applied to the Kenyon cell body region of the adult label small
ensembles of neurons that innervate all of the Trauben. Kenyon
cell axons are tightly packed in the central axis of each lobe (arrows) and fan out into the Trauben at their tips (arrowheads). (A,
inset) DiI fill in a different brain further illustrating the branching
of a small cohort of Kenyon cells into all of the Trauben. (B) AntiDC0 and phalloidin double staining localizes the ingrowth core
(arrow) containing newly extending axons. Axons within the core
also supply all of the Trauben. (C) Golgi impregnation of a small
cohort of Kenyon cell axons supplying all of the Trauben. Fine
fibers project from the main lobe axis into the Trauben (arrows).
(D) Golgi impregnation of Kenyon cell axons in the pedunculus,
horizontal (H), and medial (M) lobes showing two separate fiber
tracts in the pedunculus (arrows) fusing into a single tract prior to
the lobe branchpoint (arrowhead). (E) Phalloidin staining of the
adult calyx reveals fine branches (arrows) emanating from the
central tract to suffuse the glomerli of the outer calyx rind (arrowheads). H, horizontal lobe; M, medial lobe; Pe, pedunculus; V,
vertical lobe. All panels frontal sections with scale bars 5 20 mm
except for D (horizontal section, scale bar 5 50 mm).
highly branched Kenyon cell population rather than multiple
subpopulations that contribute to different parts of the lobes.
The lobe subdivisions in Thermobia mushroom bodies cannot
represent the ramifications of Kenyon cell subpopulations
produced sequentially during development.
Phalloidin also labeled fine Kenyon cell processes associated with the ingrowth core. As observed with DiI fills and
Golgi impregnations, phalloidin-labeled axons in Thermobia
were tightly packed in the main axes of the lobes (Fig. 6A). A
branch of the ingrowth core penetrated the center of each of
the Trauben, at which point the core fibers appeared to defasciculate into a network of collaterals that radiated throughout the body of each Trauben (Fig. 6A). At Trauben tips,
these collaterals formed whorls encircling the axis of the ingrowth core (Fig. 6B). Similar phalloidin-staining collaterals
emanating from the ingrowth core are observed in neopteran
insects such as Periplaneta (Fig. 6C).
DISCUSSION
Cellular morphology of Thermobia mushroom
bodies
The mushroom bodies of T. domestica, a basal apterygote
insect, are clearly homologous to those of neopteran species.
156
EVOLUTION & DEVELOPMENT
Vol. 7, No. 2, March^April 2005
Fig. 6. The ingrowth core of the Thermobia mushroom body lobes.
(A) Anti-DC0 (purple) and phalloidin (green) double staining
shows the location and fine structures of axons composing the
ingrowth core. (A) The ingrowth core (large arrow) in Thermobia
produces a fine spray of collaterals that suffuse the Trauben (small
arrows). (B) At lobe tips, fibers of the ingrowth core may also take
on a whorled morphology (arrows). (C) In Periplaneta, similar fine
collaterals (small arrows) emanate from the ingrowth core (large
arrows) and infiltrate the remainder of the medial and vertical
lobes. Ca, calyx; H, horizontal lobe; M, medial lobe; Pe, pedunculus; V, vertical lobe. All frontal sections except C (sagittal).
Scale bars: A, B 5 20 mm; C 5 50 mm.
Thermobia mushroom bodies are composed of minute intrinsic neurons (Kenyon cells) whose dendritic and axonal processes are partitioned into a calyx and lobes, respectively. The
calyx is supplied at least in part by afferents from the deutocerebrum via a tract resembling the iACT of Neoptera. In
both neopterans and apterygotes, MbNb clusters in the
dorsoposterior brain produce Kenyon cells throughout most
or all of postembryonic development (Ito and Hotta 1992;
Cayre et al. 1996; Farris et al. 1999; Farris and Strausfeld
2001), and phalloidin labels axons of newborn Kenyon cells
comprising a discrete core tract in the lobes (Kurusu et al.
2002; Farris et al. 2004). Core fibers also produce a network
of fine collaterals with a similar high affinity for phalloidin
(Farris et al. 2004). The D. melanogaster anti-DC0 antibody,
demonstrated to show a high affinity for Kenyon cells in
neopteran species (Skoulakis et al. 1993; Farris and Sinakevitch, 2003; Farris and Strausfeld, 2003; Farris et al., 2004),
also labels Kenyon cell processes in the Thermobia mushroom
bodies.
The organization of Thermobia mushroom bodies departs
somewhat from that of the Neoptera at the cellular level.
Lobe subdivisions in the Neoptera, which take the form of
separate layers or lobe systems, are composed of the axons of
distinct subpopulations of Kenyon cells (Fig. 1A; Farris and
Sinakevitch 2003). Kenyon cell subpopulations are produced
sequentially during development as are the lobe subdivisions
they comprise. In Thermobia, 15 lobe subdivisions (Trauben)
arranged along three lobe axes (medial, vertical, and horizontal) are present from nymphal hatching to adulthood. No
new subdivisions are added; rather, the 15 Trauben increase in
size only. This evidence, combined with DiI, Golgi, and phalloidin staining results, strongly suggests that the Thermobia
mushroom bodies are made up of a single Kenyon cell population in which individual neurons produce a highly
branched axon that supplies all subdivisions of the lobes.
Another possible explanation for the observed pattern of
development is that the MbNbs of Thermobia produce several
subpopulations of Kenyon cells simultaneously rather than
sequentially. The processes of these different subpopulations
might enter the lobes via the ingrowth core and then partition
into different arrays of Trauben. This study was unable to
demonstrate the axon morphology of single Kenyon cells;
however, this latter scenario is still highly unlikely in light of
the results of DiI fills and Golgi impregnations that randomly
stain small ensembles of Kenyon cells. These methods repeatedly revealed an invariant axon morphology in which highly
branched processes supplied all of the Trauben (Fig. 4, A, C,
and D). Furthermore, phalloidin staining of the ingrowth core
does not support the continuous partitioning of axons into
separate Trauben. Were this the case, attenuation of phalloidin labeling would be expected at the lobe tips, as the only
axons left would be those projecting into the distalmost
Trauben (somewhat like the DiI fills in Fig. 4A). Such attenuation of phalloidin labeling was not observed; instead, the
ingrowth core appeared uniform throughout the lobes (Figs.
4B and 5).
Fibers in the Thermobia ingrowth core were associated
with fine axon collaterals with high affinity for phalloidin that
fanned out across the body of each of the Trauben (Fig. 6, A
and B). Kenyon cell collateral networks associated with the
Farris
ingrowth core have been observed in widely divergent neopteran species (Fig. 6C; Farris and Strausfeld, 2001; Sinakevitch et al., 2001; Farris et al. 2004), indicating that immature
axon morphology has been conserved across the insects. In
Neoptera such as the cockroach (Fig. 6C), collaterals are a
transient aspect of Kenyon cell morphology and are lost as
the main axons are displaced from the ingrowth lamina (Farris and Strausfeld 2001; Farris et al. 2004). DiI fills and Golgi
impregnations, which are not selective for newborn Kenyon
cells, repeatedly revealed collateral-like processes in the Trauben of Thermobia. Based on this evidence, it is possible that all
Kenyon cells in Thermobia, regardless of age, produce these
processes. Once established, however, they lose their affinity
for phalloidin, and it is these processes that make up the body
of the globular Trauben.
Thermobia and the evolution of insect mushroom
bodies
The current evidence suggests that Thermobia mushroom
bodies are composed of a single highly branched Kenyon cell
population. The basal phylogenetic position of Thermobia
suggests that the acquisition of additional Kenyon cell subpopulations was likely a neopteran innovation. What factors
could have driven this increase in mushroom body complexity? In the Neoptera, different Kenyon cell subpopulations
connect separately or in varying combinations with afferent
and efferent neurons (Mauelshagen 1993; Rybak and Menzel
1993; Li and Strausfeld 1997; Strausfeld 2002). Kenyon
cell subpopulations may therefore be regarded as differential
processing modules that produce a variety of combinatorial
outputs (Strausfeld et al. 1998; Strausfeld 2002). Additional Kenyon cell subpopulations could increase the functional plasticity of the mushroom bodies by acquiring new
afferent sources and/or generating new combinatorial outputs. An increase in processing power in a learning and
memory and sensory integration neuropil such as the mushroom bodies is likely to be important for the evolution of
complex behaviors; indeed, the mushroom bodies of the social
Hymenoptera are not only characterized by their large size,
but by a bewildering variety of Kenyon cell types and the
concomitant precise segregation of afferents and efferents
among them (Mobbs 1982; Ehmer and Gronenberg 2002;
Strausfeld 2002). At present, little is known about the extrinsic
connections of the Thermobia mushroom bodies, although the
lack of Kenyon cell diversity may reflect a similar lack of
diversity in extrinsic neuron circuitry. One intriguing possibility is that Kenyon cell processes in the Thermobia lobes
behave like amacrine processes rather than axons, forming
local circuits with afferent and efferent neurons in each of the
Trauben. Such circuitry has already been observed in the
cockroach mushroom bodies (Li and Strausfeld 1997, 1999),
and has been proposed to represent an ancestral mode of
Mushroom bodies of Thermobia
157
mushroom body organization (Strausfeld and Li 1999a;
Strausfeld 2001).
The morphology of the Thermobia calyx also provides
clues about the evolution of this substructure. Each mushroom body in the Thermobia brain possesses a single calyx
that is supplied by two fiber tracts arising from the progeny of
two MbNb clusters. This may be interpreted to represent an
ancestral mode of calyx organization based on the basal
phylogenetic position of Thermobia, but comparisons with
neopterans cast doubt on this conclusion. The Thermobia
calyx is reminiscent of that of the highly derived neopteran
D. melanogaster, in which the Kenyon cells supply a single
calyx neuropil. Based on the symmetrical subdivisions of
Kenyon cell populations revealed by enhancer trap expression, Yang et al. (1995) suggest that Drosophila calyx organization results from the secondary fusion of two ancestral
calyces. Double calyces are also observed in the mushroom
bodies of divergent species, such as the cockroach P. americana and the honey bee A. mellifera.
Evidence from these and other species shows that calyx
morphology is flexible and not solidly correlated with basal or
derived species. Instead, the overall mushroom body size may
be a better determinant of calyx morphology. The single-calyx
mushroom body of Drosophila contains only about 2500
Kenyon cells per hemisphere (Hinke 1961). Periplaneta and
Apis, both of which possess double calyces, have among the
largest mushroom bodies (170,000 Kenyon cells per hemisphere in the honey bee [Witthöft 1967], 175,000 per hemisphere in Periplaneta [Neder 1959]. It is therefore possible that
double calyces arose independently in several lineages in
which a trend toward increased mushroom body size occurred. Like the gyri and sulci of the mammalian cerebral
cortex to which the mushroom body calyces were originally
compared (Dujardin 1850), the extra surface area provided by
calyx duplication can accommodate increasing numbers of
Kenyon cell dendrites and their afferent connections. In highly derived lineages such as the brachyceran Diptera, which
possess small mushroom bodies, this trend could be reversed
such that two ancestral calyces would secondarily fuse into a
single structure. Further comparative studies will be needed to
definitively establish the possibility of this simple connection
between calyx morphology and Kenyon cell number.
The basic groundplan for insect mushroom body organization was established before the Zygentoma diverged from
the insect lineage, approximately 400 Myr (Engel and Grimaldi 2004). The presence of well-developed mushroom bodies
in Thermobia and other lepismatids (Böttger 1910; Hanström
1940) raises questions as to when these neuropils first appeared. Members of the even older apterygote order Archeognatha do not possess mushroom bodies (Hanström 1940;
Strausfeld 1998), indicating that the insect mushroom bodies
arose between the time of appearance of the two apterygote
orders, in the early Devonian (Grimaldi 2001; Engel and
158
EVOLUTION & DEVELOPMENT
Vol. 7, No. 2, March^April 2005
Grimaldi 2004). Among the basal non-insect Hexapoda, the
Collembola and Protura also lack mushroom bodies (Hanström 1940; Bullock and Horridge 1965), but Holmgren
(1916) and Hanström (1940) describe in the Diplura dense,
multi-lobed protocerebral neuropils associated with a dorsoposterior cluster of minute intrinsic neurons. This suggests
either that mushroom bodies arose even earlier in hexapod
evolution and were subsequently lost in the lineage leading to
the Archeognatha, or that mushroom body–like structures
have evolved twice independently in the hexapods. This question restates the century-old debate as to the homology of
mushroom body–like structures in invertebrates such as polychaete annelids, onychophorans, chelicerates, and myriapods (Holmgren 1916; Hanström 1940; Strausfeld et al. 1995;
Strausfeld 1998). Future studies of mushroom body development, cellular composition, and gene expression, combined
with the existing body of morphological evidence, will provide further insight into the evolutionary history and the
functional necessity of these complex sensory integration
neuropils.
Acknowledgments
The author would like to thank Mr. Jonathan Benincosa and Ms.
Deborah Hardee for their assistance with the Golgi protocol and
tissue processing, respectively. The author would also like to thank
Dr. Daniel Kalderon for providing the anti-DC0 antibody, and Dr.
Paul Liu and Dr. Thomas Kaufmann for providing Thermobia
domestica. Dr. Ronald Bayline, Dr. Ashok Bidwai, and Dr. Nicholas
Strausfeld provided helpful comments on the manuscript. This work
was supported by West Virginia University.
REFERENCES
Borror, D. J., Triplehorn, C. A., and Johnson, N. F. 1989. An Introduction
to the Study of Insects. Saunders College Publishing, Philadelphia.
Böttger, O. 1910. Das Gehirn eines niederen Insektes (Lepisma saccharina
L.). Jenaische Zeitschrift für Naturwissenschaft 46: 801–844.
Bullock, T. H., and Horridge, G. A. 1965. Structure and Function in the
Nervous Systems of Invertebrates. W. H. Freeman and Company, San
Francisco.
Cayre, M., Malaterre, J., Charpin, P., Strambi, C., and Strambi, A. 2000.
Fate of neuroblast progeny during postembryonic development of mushroom bodies in the house cricket, Acheta domesticus. J. Insect. Physiol.
46: 313–319.
Cayre, M., et al. 1996. Neurogenesis in adult insect mushroom bodies. J.
Comp. Neurol. 371: 300–310.
Dujardin, F. 1850. Mémoire sur le système nerveux des insectes. Ann. Sci.
Nat. Zool. 14: 195–206.
Ehmer, B., and Gronenberg, W. 2002. Segregation of visual input to the
mushroom bodies in the honeybee (Apis mellifera). J. Comp. Neurol. 451:
362–373.
Engel, M., and Grimaldi, D. 2004. New light shed on the oldest insect.
Nature 427: 627–630.
Farris, S. M., Abrams, A. I., and Strausfeld, N. J. 2004. Development and
morphology of Class II Kenyon cells in the mushroom bodies of the
honey bee, Apis mellifera. J. Comp. Neurol. 474: 325–339.
Farris, S. M., Robinson, G. E., Davis, R. L., and Fahrbach, S. E. 1999.
Larval and pupal development of the mushroom bodies in the honey bee,
Apis mellifera. J. Comp. Neurol. 414: 97–113.
Farris, S. M., and Sinakevitch, I. 2003. Development and evolution of the
insect mushroom bodies: towards the understanding of conserved developmental mechanisms in a higher brain center. Arthropod Struct. Dev.
32: 79–101.
Farris, S. M., and Strausfeld, N. J. 2001. Development of laminar organization in the mushroom bodies of the cockroach: Kenyon cell proliferation, outgrowth, and maturation. J. Comp. Neurol. 439: 331–351.
Farris, S. M., and Strausfeld, N. J. 2003. A unique mushroom body substructure common to both basal cockroaches and to termites. J. Comp.
Neurol. 456: 305–320.
Frambach, I., Rössler, W., Winkler, M., and Schürmann, F.-W. 2004.
F-Actin at identified synapses in the mushroom body neuropil of the
insect brain. J. Comp. Neurol. 475: 303–314.
Grimaldi, D. 2001. Insect evolutionary history from Handlirsch to Hennig,
and beyond. J. Paleontol. 75: 1152–1160.
Gronenberg, W. 2001. Subdivisions of hymenopteran mushroom body
calyces by their afferent supply. J. Comp. Neurol. 436: 474–489.
Gronenberg, W., and Hölldobler, B. 1999. Morphologic representation of
visual and antennal information in the ant brain. J. Comp. Neurol. 412:
229–240.
Hanström, B. 1940. Inkretorische Organe, Sinnesorgane und Nervensystem
des Kopfes einiger niederer Insektenordnungen. K. Sven. Vetenskapsakad. Handl. 18: 1–266.
Heisenberg, M. 2003. Mushroom body memoirs: from maps to models.
Nat. Neurosci. Rev. 4: 266–275.
Hinke, W. 1961. Das relative postembryonale Wachstum der Hirnteile von
Culex pipiens, Drosophila melanogaster und Drosphila mutanten.
Z. Morphol. Okol. Tiere. 50: 81–118.
Holmgren, N. 1916. Zur Vergleichenden anatomie des gehirns von Polychaeten, Onychophoren, Xiphosuren, Arachniden, Crustaceen, Myriapoden und Insekten. K. Sven. Vetenskapsakad. Handl. 56: 1–303.
Ito, K., Awano, W., Suzuki, K., Hiromi, Y., and 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.
Ito, K., and Hotta, Y. 1992. Proliferation pattern of postembryonic neuroblasts in the brain of Drosophila melanogaster. Dev. Biol. 149: 134–148.
Ito, K., Suzuki, K., Estes, P., Ramaswami, M., Yamamoto, D., and
Strausfeld, N. J. 1998. The organization of extrinsic neurons and their
implications in the functional roles of the mushroom bodies in Drosophila melanogaster Meigen. Learn. Memory 5: 52–77.
Kiernan, J. 1990. Histological and Histochemical Methods: Theory and
Practice. Pergamon Press, Oxford.
Kurusu, M., Awasaki, T., Masuda-Nakagawa, L. M., Kawauchi, H., Ito,
K., and Furukubo-Tokunaga, K. 2002. Embryonic and larval development of the Drosophila mushroom bodies: concentric layer subdivisions
and the role of fasciclin II. Development 129: 409–419.
Lee, T., Lee, A., and Luo, L. 1999. Development of the Drosophila mushroom bodies: sequential generation of three distinct types of neurons
from a neuroblast. Development 126: 4065–4076.
Li, Y.-S., and Strausfeld, N. J. 1997. Morphology and sensory modality of
mushroom body extrinsic neurons in the brain of the cockroach, Periplaneta americana. J. Comp. Neurol. 387: 631–650.
Li, Y.-S., and Strausfeld, N. J. 1999. Multimodal efferent and recurrent
neurons in the medial lobes of cockroach mushroom bodies. J. Comp.
Neurol. 409: 647–663.
Liu, L., Wolf, R., Ernst, R., and Heisenberg, M. 1999. Context generalization in Drosophila visual learning requires the mushroom bodies.
Nature 400: 753–756.
Malaterre, J., Strambi, C., Chiang, A.-S., Aouane, A., Strambi, A., and
Cayre, M. 2002. Development of cricket mushroom bodies. J. Comp.
Neurol. 452: 215–227.
Malun, D. 1998. Early development of mushroom bodies in the brain of the
honeybee Apis mellifera as revealed by BrdU incorporation and ablation
experiments. Learn. Memory 5: 90–101.
Mauelshagen, J. 1993. Neural correlates of olfactory learning in an identified neuron in the honey bee brain. J. Neurophysiol. 69: 609–625.
Mendes, L. 2002. Taxonomy of Zygentoma and Microcoryphia: historical
overview, present status and goals for the new millenium. Pedobiologia
46: 225–233.
Farris
Mizunami, M., Okada, R., Li, Y.-S., and Strausfeld, N. J. 1998a. Mushroom bodies of the cockroach: activity and identities of neurons recorded
in freely moving animals. J. Comp. Neurol. 402: 501–519.
Mizunami, M., Weibrecht, J., and Strausfeld, N. J. 1998b. Mushroom
bodies of the cockroach: their participation in place memory. J. Comp.
Neurol. 402: 520–537.
Mobbs, P. G. 1982. The brain of the honeybee Apis mellifera L. The connections and spatial organization of the mushroom bodies. Phil. Trans.
R. Soc. Lond. B 298: 309–354.
Mobbs, P. G. 1984. Neural networks in the mushroom bodies of the honeybee. J. Insect Physiol. 30: 43–58.
Neder, R. 1959. Allometrisches Wachstum von Hirnteilen bei drei verschieden grossen Schabenarten. Zool. Jahrb. Anatomie 77: 411–464.
Noveen, A., Daniel, A., and Hartenstein, V. 2000. Early development of the
Drosophila mushroom body: the roles of eyeless and dachshund. Development 127: 3475–3488.
Okada, R., Ikeda, J., and Mizunami, M. 1999. Sensory responses and
movement-related activities in extrinsic neurons of the cockroach mushroom bodies. J. Comp. Physiol. A 185: 115–129.
O’Shea, M., and Adams, M. 1981. Pentapeptide (proctolin) associated with
an identified neuron. Science 213: 567–569.
Panov, A. A. 1957. The structure of the brain in insects in successive stages
of postembryonic development. Rev. Entom. USSR 36: 269–284.
Rybak, J., and Menzel, R. 1993. Anatomy of the mushroom bodies in the
honey bee brain: the neuronal connections of the alpha-lobe. J. Comp.
Neurol. 334: 444–465.
Schildberger, K. 1984. Multimodal interneurons in the cricket brain: properties
of identified extrinsic mushroom body cells. J. Comp. Physiol. 154: 71–79.
Schürmann, F. 1973. Über die Struktur der Pilzkörper des Insektengehirns.
III Die Anatomie der Nervenfasern in den Corpora pedunculata bei
Acheta domesticus L. (Orthopt.): eine Golgi-Studie. Z. Zellforsch.
Mikrosk. Anat. 145: 247–285.
Sinakevitch, I., Farris, S. M., and Strausfeld, N. J. 2001. Taurine-, aspartate- and glutamate-like immunoreactivity identifies chemically distinct subdivisions of Kenyon cells in the cockroach mushroom body. J.
Comp. Neurol. 439: 352–367.
Skoulakis, E. M. C., Kalderon, D., and Davis, R. L. 1993. Preferential
expression in mushroom bodies of the catalytic subunit of protein kinase
A and its role in learning and memory. Neuron 11: 197–208.
Strausfeld, N. J., Buschbeck, E. K., and Gomez, R. S. 1995. The arthropod
mushroom body: its functional roles, evolutionary enigmas and mistaken
Mushroom bodies of Thermobia
159
identities. In O. Briedbach and W. Kutsch (eds.). The Nervous Systems of
Invertebrates: An Evolutionary and Comparative Approach. Birkhäuser
Verlag, Basel, Switzerland, pp. 349–381.
Strausfeld, N. J. 1998. Crustacean–insect relationships: the use of brain
characters to drive phylogeny amongst segmented invertebrates. Brain
Behav. Evol. 52: 186–206.
Strausfeld, N. J., Hansen, L., Li, Y., Gomez, R. S., and Ito, K. 1998.
Evolution, discovery, and interpretations of arthropod mushroom bodies. Learn. Memory 5: 11–37.
Strausfeld, N. J., and Li, Y.-S. 1999a. Organization of olfactory and multimodal afferent neurons supplying the calyx and pedunculus of the
cockroach mushroom bodies. J. Comp. Neurol. 409: 603–625.
Strausfeld, N. J., and Li, Y.-S. 1999b. Representation of the calyces in the
medial and vertical lobes of cockroach mushroom bodies. J. Comp.
Neurol. 409: 626–646.
Strausfeld, N. J. 2001. Insect brain. In G. Roth and M. Wulliman (eds.).
Brain, Evolution and Cognition. John Wiley and Sons, New York.
Strausfeld, N. J. 2002. Organization of the honey bee mushroom body:
representation of the calyx within the vertical and gamma lobes.
J. Comp. Neurol. 450: 4–33.
Strausfeld, N. J., Sinakevitch, I., and Vilinsky, I. 2003. The mushroom
bodies of Drosophila melanogaster: an immunocytological and Golgi
study of Kenyon cell organization in the calyces and lobes. Microsc. Res.
Technol. 62: 151–169.
Vowles, D. 1964. Olfactory learning and brain lesions in the wood ant
(Formica rufa). J. Comp. Physiol. Psychol. 58: 105–111.
Waddell, S., and Quinn, W. 2001. Flies, genes and learning. Ann. Rev.
Neurosci. 24: 1283–1309.
Weiss, M. J. 1981. Structural patterns in the corpora pedunculata of
Orthoptera: a reduced silver analysis. J. Comp. Neurol. 203: 515–553.
Witthöft, W. 1967. Absolute Anzahl und Verteilung der Zellen im Hirn der
Honigbiene. Z. Morphol. Tiere. 61: 160–164.
Yang, M. Y., Armstrong, J. D., Vilinski, I., Strausfeld, N. J., and Kaiser, K.
1995. Subdivision of the Drosophila mushroom bodies by enhancer-trap
expression patterns. Neuron 15: 45–54.
Yusuyama, K., Meinertzhagen, I., and Schürmann, F.-W. 2002. Synaptic
organization of the mushroom body calyx in Drosophila melanogaster.
J. Comp. Neurol. 445: 211–226.
Zhu, S., Chiang, A.-S., and Lee, T. 2003. Development of the Drosophila
mushroom bodies: elaboration, remodeling and spatial organization of
dendrites in the calyx. Development 130: 2603–2610.