Continuous neurogenesis in the olfactory brain of adult shore crabs, Carcinus maenas

Brain Research 762 Ž1997. 131–143

Continuous neurogenesis in the olfactory brain of adult shore crabs, Carcinus maenas Manfred Schmidt

)

Technische UniÕersitat ¨ Berlin, Institut fur ¨ Biologie, Franklinstr. 28 r 29, 10587 Berlin, Germany Accepted 18 March 1997

Abstract To scrutinize the common belief that the number of neurons in the CNS of adult decapod crustaceans stays constant, in spite of their dramatic postlarval increase in size, I counted olfactory projection neurons ŽOPNs. in the brains of differently-sized postlarval shore crabs, Carcinus maenas, and performed in vivo labeling of proliferating cells with 5-bromo-2X-deoxyuridine ŽBrdU. on brains of adults. The number of OPNs increases continuously throughout the postlarval life of shore crabs and approximately doubles from the very young to the oldest animals. Brain sections from adult crabs labeled with BrdU revealed ongoing proliferation of cells in the lateral soma cluster, which consists of OPN cell bodies, and in the cluster of somata of hemiellipsoid body local interneurons, which are the targets of the OPNs. Post-injection survival times from 5.5 to 120 h revealed a small but relatively constant number of labeled nuclei with neuronal morphology in both soma clusters of all specimens Ž31.3 " 9.5 S.D. nuclei per lateral cluster, n s 29; 20.1 " 4.5 S.D. nuclei per hemiellipsoid body cluster, n s 10.. The labeled nuclei were located in a distinct proliferative zone in each cluster. There were significantly more labeled nuclei in both soma clusters after a prolonged post-injection survival time of 1 month Ž71.3 " 7.8 S.D. nuclei per lateral cluster, n s 4; 38.2 " 7.1 nuclei per hemiellipsoid body cluster, n s 6.. In both soma clusters the labeled nuclei formed a compact group that was dislocated from the proliferation zone towards the outer edge of the cluster. In the proliferation zone of the lateral cluster histological stainings revealed cell bodies of typical neuronal shape that are slightly smaller and more intensely stained than the surrounding OPN somata. Some of these cell bodies were captured in various stages of mitosis. Collectively, these data indicate that continuous neurogenesis occurs in the central olfactory pathway of the brain of shore crabs throughout their entire adult life. This unexpected structural plasticity may enable long-lived decapod crustaceans to adapt to ever-changing olfactory environments. q 1997 Elsevier Science B.V. X

Keywords: Carcinus maenas; Crustacean; Adult neurogenesis; Brain; Olfactory pathway; 5-Bromo-2 -deoxyuridine; BrdU; Neuron number

1. Introduction Recent studies demonstrate that neurogenesis is common in the adult vertebrate brain. Besides other areas w5,20,34,56x, especially the olfactory bulbs acquire new neurons throughout adult life in diverse vertebrate species w1,2,8,25,26,36x. Among arthropods, however, evidence for neurogenesis in the adult brain is very limited. In some insect species Žthe fruit fly Drosophila melanogaster, several crickets and beetles. neurogenesis is restricted to the Kenyon cells, the intrinsic neurons of the mushroom bodies, which represent second order olfactory neuropils w3,6,7,53x. In other insect species Žthe honey bee Apis ) Corresponding author. Present address: Universitat ¨ Hamburg, Zoologisches Institut und Museum, Martin-Luther-King-Platz 3, 20146 Hamburg, Germany.

0006-8993r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 7 . 0 0 3 7 6 - 4

mellifera, the locust Locusta migratoria, the cockroach Periplaneta americana., however, neurogenesis is lacking throughout the adult brain w6,9x. Decapod crustaceans, which usually lack a terminal moult, may increase in size up to 100 000-fold after having obtained adult morphology, e.g. w12x. This size increase is associated with a substantial increase in the number of sensory receptor neurons w24,27,52x, whereas the number of motoneurons remains constant w11,12x. The latter finding led Purves Ž1988. w33x to assume the number of neurons in the CNS Žof lobster. was fixed throughout adult life. To date, however, this assumption has not been tested directly in any decapod crustacean. The significant, growth-related increase in the number of receptor neurons, raises the question, how do the central neurons involved in the processing of this sensory information adapt to the increasing input, in particular do they show neurogenesis. To address this question,

132

M. Schmidtr Brain Research 762 (1997) 131–143

which was first raised by Laverack w23x, I studied morphological changes in the central olfactory pathway of several size Žand hence age. classes of postlarval shore crabs, Carcinus maenas and performed in vivo BrdU labeling in juvenile and adult specimens to visualize proliferating cells. Part of these data have been presented earlier in abstract form w42,43x. The central olfactory pathway was chosen for this study because of several reasons. Firstly, it is known that the

receptor neurons providing input to this pathway increase in number during adult growth w27,40,52x. Secondly, the components of the central olfactory pathway especially its soma clusters are clearly separated from other sensory pathways and represent almost pure populations of specific neuron types, e.g. w45x. Finally, it was recently reported that neurogenesis persists into the early juvenile stage in the central olfactory pathway of a crab Ž Hyas araneus . w16x.

Fig. 1. Organization of the olfactory system in the shore crab, Carcinus maenas. a: olfactory sensilla are located on the antennule Žfirst antenna; arrow.. The two animals represent size-range of crabs analyzed in this study. b: the central olfactory pathway consists of two levels. The olfactory lobe ŽOL. in the central brain receives olfactory afferents Ž1. from the antennular nerve ŽA I Nv. and is comprised of local interneurons Ž2. and projection neurons Ž3. with their somata in the medial ŽMC. and lateral soma cluster ŽLC., respectively. The axons of the projection neurons form the olfactory globular tract ŽOGT. ascending to the hemiellipsoid body ŽHB. in the eyestalk ganglion, which represents the second stage. The somata of local HB interneurons form a cluster adjacent to the HB neuropil Žasterisk.. Medulla terminalis ŽMT.; oesophageal connective ŽOC.; optic lobe ŽOPL.. c: the lateral soma cluster ŽLC. adjacent to the neuropil of the olfactory lobe ŽOL. is comprised of small, globuli-type somata of olfactory projection neurons showing a size and staining gradient from the inside to the outside of the cluster. Light micrograph of toluidine blue stained semithin horizontal section; male crab of 39 mm cw.

M. Schmidtr Brain Research 762 (1997) 131–143

2. Materials and methods All experiments were performed with shore crabs, Carcinus maenas, ranging from 7.5 to 53 mm in carapace width Žcw. ŽFig. 1a.. The crabs were collected at the North Sea coast and kept in aquaria with circulating artificial seawater at a temperature of about 158C. Two independent measures were used to determine the number of OPNs in brains of differently sized postlarval shore crabs. In brains of seven specimens the total number of OPN somata was determined morphometrically by light microscopy; in three of these brains Žthe smallest, the largest and an intermediate-sized specimen. the axons of the OPNs were counted on cross sections through the olfactory globular tract. For light and electron microscopy brains were fixed in 5% glutaraldehyde Ž4 h., postfixed in 2% OsO4 Ž2 h., embedded in EPON, and cut into semithin Ž1.5 mm. or ultrathin Ž100 nm. sections according to methods described in detail elsewhere w10x. Semithin sections were dried onto glass slides, stained with a 1% aqueous toluidine blue solution containing 6% sodiumtetraborate for 2–3 min at 608C, and coverslipped with Permount ŽFisher.. On a microscope with bright-field optics ŽOlympus BH-2. sections were viewed, photographed, drawn with a camera lucida, or digitized and analyzed using an image analysis system ŽAnalySis, Software Imaging System.. Ultrathin sections were collected on slit-grits covered with pioloform, stained with lead citrate for 10 min and viewed and photographed in a transmission electron microscope ŽSiemens ELMI 1A.. Prerequisites for the morphometric determination of

133

OPN soma number were their exclusive occurrence and tight packing in the LC leaving less than 5% of the total cluster volume for glial cells and arterioles, as well as their relatively uniform size ŽFig. 1c.. This allowed the total number of OPN somata to be estimated by dividing the total LC volume by the mean soma volume. Both volumes were measured in series of horizontal semithin sections stained with toluidine blue. The volume of the LC was delineated by measuring the area it occupies on every 10th section and adding up the partial volumes obtained by multiplying each area with the section distance Ž15 mm.. The area measurements were performed with a 3-D reconstruction system ŽHVEM-3d. w55x. The mean soma volume was determined by measuring the area of 500 neuronal somata in each LC individually Ž100 somata in each of the 5 semithin sections from different regions of the LC. with the image analysis system ŽAnalySis. and calculating from each soma area the soma volume assuming a spherical shape of the somata. The axons of OPNs were counted on electron micrographs of cross sections through the OGT on the level of the protocerebral tract. This was possible because the axons in the OGT are considerably thinner than surrounding axons allowing precise identification ŽFig. 2a,b.. For in vivo labeling with 5-bromo-2X-deoxyuridine ŽBrdU., 5 mg BrdU per 100 g of body weight was injected into the hemolymph of crabs in a 0.5% BrdU solution in Carcinus saline w47x. After Žpost-injection. survival times of 2–720 h, brains were fixed with Bodian a 2 fixative Ž90 ml 80% ethanol, 5 ml formol, 5 ml glacial acidic acid., for 45 min, embedded in gelatine and cut on a vibratome

Fig. 2. Organization of the olfactory globular tract ŽOGT. of Carcinus maenas. a: cross-section through the protocerebral tract connecting central brain and eyestalk ganglion. The OGT, containing only extremely thin axons of olfactory projection neurons ŽOPNs., is clearly delimitated from other axons of larger diameter running in the PT. Light micrograph of toluidine blue stained semithin section; male crab of 33 mm cw. b: TEM micrograph of cross-section through a part of the OGT as used for counting axons of OPNs. Note the very uniform and small diameter Ž0.1–0.5 mm. of the OPN axons. Thin processes of glial cells ŽG. ensheathe groups of axons. Male crab 33 mm of cw.

134

M. Schmidtr Brain Research 762 (1997) 131–143

in 70 mm thick serial sections. After degelatinization on a warm water bath, incubation in 2 N HCl for 20 min, and rinsing Ž4 = 15 min. in 0.1 M Sorensen phosphate buffer ¨ ŽSPB., the free-floating sections were incubated over-night in anti-BrdU ŽCell Proliferation Kit; Boehringer, Mannheim.. Subsequently sections were rinsed in SPB Ž4 = 30 min., incubated for 4 h in goat anti-mouse CY3labeled secondary antibody ŽJackson Immunoresearch. diluted 1:100 in SPB, and coverslipped in glycerolrSPB after final rinsing in SPB Ž4 = 30 min.. Sections were viewed and photographed in a microscope with epifluorescence ŽOlympus BH-2.. Despite increased section thickness Ž70 mm compared to the 20-mm-thick cryostat sections normally used in BrdU immunostainings of mammalian tissue. antibody penetration was not a problem encountered. Intensely labeled nuclei were detectable throughout the entire thickness of the vibratome sections. In studies on larval neurogenesis in decapod crustaceans BrdU immunostaining has been successfully employed even on whole mount preparations of the CNS w15,16x.

3. Results As in other decapod species w27,37,44x and insects w4x, the central olfactory pathway of shore crabs consists of two levels ŽFig. 1b.. In the deutocerebrum, a pair of glomerular neuropils, the olfactory lobes ŽOL. similar to the antennal lobes of insects, receive the olfactory afferents and serve as the first synaptic relay. In the protocerebrum the paired hemiellipsoid bodies ŽHB., similar to the insect mushroom bodies, serve in second-order olfactory processing. The OLs are comprised of local interneurons, the somata of which constitute the medial soma clusters ŽMC., and of olfactory projection neurons ŽOPN., whose somata form the lateral soma clusters ŽLC.. The axons of the OPNs constitute the large olfactory globular tracts ŽOGT., which ascend to the HBs in the eyestalk ganglia via the interconnecting protocerebral tracts ŽPT. ŽFigs. 1 and 2a,b.. The HBs are comprised of axon terminals of OPNs and local interneurons, whose somata constitute specific soma clusters ŽHBC.. The morphometric analysis showed that the volume of the LC increases linearly with the size Žcarapace width s cw. of the shore crabs and quintuples from the smallest Žcw 7.5 mm; weight 105 mg. to the largest crab Žcw 53 mm; weight 39.1 g. included in the study ŽFig. 3a.. Over the same size range of crabs the mean soma volume increases only about 2.5-fold ŽFig. 3b.. Soma sizes are distributed almost normally across all size classes of crabs but the soma size distribution broadens considerably towards larger specimens. The soma number calculated from LC volume and mean soma volume increases linearly and approximately doubles over the size range of crabs under study ŽFig. 3c.. The counts of axons in the OGT ŽFig. 2b. show a similarly linear, approximately 2-fold increase in

axon number over the same size range of crabs ŽFig. 3c.. In each of the three analyzed animals the number of axons in the OGT is about 25% higher than the number of somata estimated for one LC. This apparent discrepancy can be attributed to the presence of a subpopulation of OPNs with axons that bifurcate in the chiasm of the OGT and project bilaterally w27,45x. In the shore crab backfilling one OGT with neurobiotin Ž n s 8. labeled axons in the contralateral OGT demonstrating the existence of bilateral projection neurons ŽSchmidt and Demuth, unpublished.. In summary, the morphometrical data show that over the size range under study and hence almost over the entire life-span of postlarval shore crabs, the number of olfactory projection neurons increases continuously and in total approximately doubles from about 23 000 to about 48 000 per brain. This substantial and continuous increase in neuron number requires neurogenesis in the brain of juvenile and adult shore crabs. In vivo labeling with the thymidine analogue 5-bromo2X-deoxyuridine ŽBrdU. was used to directly demonstrate proliferating cells in brains of postlarval shore crabs Ž n s 26. ranging from 20 to 56 mm in cw and 2.0 to 45.1 g in weight, including mostly large, sexually mature specimens Žcw ) 37 mm; weight) 10.4 g; n s 19.. A Post-injection survival time of 2 h did not reveal any labeled nuclei in the brain Ž n s 2.; a survival time of 4 h resulted in very lightly labeled nuclei, whose number could not be delineated Ž n s 4.. All brains Ž n s 16. with survival-times of 5.5 to 120 h after BrdU injection contained distinctly labeled nuclei in each LC Ž19 to 56; mean: 31.3 " 9.5 S.D.; n s 29 LCs., with the number of labeled nuclei being very similar Žmaximal difference: 25%. in each of the two LCs of an individual brain ŽFig. 4a.. The number of labeled nuclei did not vary systematically with the size ŽFig. 7a., sex or moult stage of the tested crabs, nor with post-injection survival time ŽFig. 7b.. Almost all somata in the LC whose nuclei had incorporated BrdU were arranged in a line, that was located in the ventral-most quarter of the cluster and radiated from the region of the LC attached to the OL towards its medial surface ŽFig. 4b.. Thus, a small area of the LC was demarcated as a ‘‘proliferative zone’’ by the somata with labeled nuclei. The labeled nuclei in the proliferative zone showed a gradient in size and staining intensity, with the smallest and most intensely labeled nuclei directly adjacent to the OL neuropil and larger, less intensely stained nuclei towards the outer surface of the LC ŽFig. 4c.. Rarely very small but intensely labeled nuclei were present close to the proliferative zone, allowing for the possibility that some of the proliferating cells in the LC undergo pyknotic or apoptotic cell death. In some crabs a few isolated nuclei were labeled in other areas of the LC. After a post-injection survival time of 1 month significantly more Ž P - 0.001; Student’s t-test. nuclei in the LC were labeled than after survival times of up to 120 h Ž60 to 78; mean: 71.3 " 7.8 S.D.; n s 4 LCs. ŽFig. 4d,e and Fig.

M. Schmidtr Brain Research 762 (1997) 131–143

7b.. As with shorter survival times the labeled nuclei were still linearly arranged but these cells were displaced from the proliferative zone towards the outer surface of the LC ŽFig. 4d.. The staining gradient was reversed with respect to shorter survival times, with the most intensely labeled nuclei located furthest from the OL ŽFig. 4e.. In many of the lightly labeled nuclei immunoreactivity was distributed in a spot-like fashion indicative of BrdU dilution through further mitoses ŽFig. 4e.. In the proliferative zone of the LC occasionally somata with BrdU-labeled nuclei were captured in various phases of mitosis Žmetaphase, telophase. showing directly that division and not only DNA synthesis occurred in these somata ŽFig. 5d.. This was further corroborated by the analysis of toluidine blue stained semithin sections through

135

the LC of untreated specimens that also revealed somata in different phases of mitosis in the proliferative zone of the LC ŽFig. 5a–c.. The mitotically active somata in the LC were spherically shaped ŽFig. 4c,e and Fig. 5a–d. and clearly resembled the morphology of mature neuronal cell bodies of OPNs and not of mature glial cell somata, which are characterized by very flat, elongated nuclei ŽFig. 5a inset.. The mitotically-active cells were among the smallest somata of neuronal shape in the LC ŽFig. 5a.. Histological analyses based on toluidine blue stained semithin sections demonstrated a clear morphological gradient within the LC with the smallest and most densely stained neuronal somata close to the OL in the proliferative zone and larger, less densely stained somata at the periphery ŽFig. 1c and Fig. 5a..

Fig. 3. Increase in volume and number of olfactory projection neurons ŽOPN. in growing postlarval shore crabs, Carcinus maenas. a: volume of the lateral soma cluster increases continuously and about 5-fold over the size range of studied crabs Ž7.5 to 53 mm cw; n s 7.; linear regression Ž r s 0.98.. b: mean volume of neuronal somata in the LC increases only about 2.5-fold over the size range of studied crabs; linear regression Ž r s 0.79.. c: the number of OPNs increases continuously and approximately doubles over the size range of crabs studied. From the data presented in Ža. and Žb. the number of OPN somata in each LC was calculated by dividing cluster volume by mean soma volume Žlower curve; linear regression r s 0.95.. As independent measure of OPN number, their axons were counted on TEM micrographs in brains of three crabs also used for the determination of soma number Župper curve; linear regression r s 0.99.. The consistently higher number of axons than somata is likely due to the presence of OPNs with bifurcating axons and bilateral projections.

136

M. Schmidtr Brain Research 762 (1997) 131–143

In vivo BrdU labeling showed another proliferative zone similar to that in the LC in the soma cluster of HB local interneurons ŽHBC. ŽFig. 6a–c. but not in the medial

soma cluster of the OL ŽFig. 4a.. After short post-injection survival times Ž17–120 h. a relatively constant number of nuclei Ž13 to 28; mean: 20.1 " 4.5 S.D.; n s 10 HBCs.

Fig. 4. Identification of proliferating cells in the lateral soma cluster ŽLC. of adult shore crabs, Carcinus maenas, with in vivo BrdU labeling. a–c: short post-injection survival times Ž17–40 h.. a: in the ventral aspect of both LCs a line-shaped group of nuclei is labeled Žarrowheads. demarcating a small region directly adjacent to the neuropil of the olfactory lobe ŽOL. as a proliferative zone. Some isolated nuclei are labeled in other areas of the central brain including the medial soma cluster Žarrows. of the OL. Horizontal, 70 mm thick vibratome section through central brain of a male shore crab measuring 44 mm in cw Ž17 h post-injection survival time.. b,c: higher magnification reveals tight clustering as well as neuronal size and shape Žspherical. of BrdU-labeled nuclei in the proliferative zone of the LC. Large neuroblasts or cells with glial morphology Žflat, elongated nuclei. are lacking. Note size and staining gradient of the labeled nuclei, with the smallest and most intensely labeled nuclei close to the OL Žtop. and larger, lightly labeled nuclei further away Žbottom.. Horizontal, 70 mm thick vibratome section through central brain of a female shore crab measuring 40 mm in cw Ž40 h post-injection survival time.. d,e: long post-injection survival time Ž1 month.. One month after BrdU injection labeled nuclei still forming a tightly clustered string-like group are present in the LC. Horizontal, 70 mm thick vibratome sections through central brain of a male shore crab measuring 45 mm in cw Ž720 h post-injection survival time.. d: the group of labeled nuclei has moved away from the proliferative zone at the OL towards the outside of the LC. e: higher magnification reveals neuronal shape Žspherical. and size of labeled nuclei. Many nuclei contain only spot-like label indicative of dilution of BrdU through further mitoses Žarrows.. Note that gradient of labeling intensity is reversed with respect to short post-injection survival times Žc. with the most intensely labeled nuclei furthest away Žbottom. from the OL Žtop..

M. Schmidtr Brain Research 762 (1997) 131–143

were labeled in the HBC ŽFig. 6a,b and Fig. 7c.. The number of labeled nuclei did not vary systematically with post-injection survival time ŽFig. 7c.. The labeled nuclei were of neuronal size and shape Žspherical. and formed a dense, clump- or string-shaped group at the inner edge of

137

the cluster adjacent to the HB neuropil ŽFig. 6b.. Among the labeled nuclei a size and staining gradient was present, with the smallest and most intensely labeled nuclei close to the HB neuropil and larger, less intensely stained nuclei towards the outer surface of the HBC. Occasionally la-

Fig. 5. Mitoses in somata located in the proliferative zone of the lateral soma cluster ŽLC. of Carcinus maenas. a: a neuronal soma captured in metaphase Žarrow. is located in the proliferative zone of the LC adjacent to the neuropil of the olfactory lobe ŽOL.. The proliferative zone contains somata of neuronal morphology that are slightly smaller and more densely stained than those in the surrounding; glial cell Žarrowhead.. Toluidine blue stained semithin section through central brain of a male crab measuring 14 mm in cw. Inset: morphological differentiation of mature neurons and glial cells. Somata of mature olfactory projection neurons in the LC have nuclei of spherical shape and loose karyoplasm Žasterisks.. In contrast, the nuclei of glial cells at the border between LC and OL are flat and elongated and have very dense karyoplasm Žarrowheads.. Toluidine blue stained semithin section through central brain of a male crab measuring 14 mm in cw. b,c: near the OL neuropil ŽOL. neuronal somata are captured in metaphase Žarrows. or late telophase Žarrowheads., respectively. Toluidine blue stained semithin section through central brain of a juvenile crab measuring 7.5 mm in cw. d: among BrdU-labeled nuclei in the proliferative zone of the LC one is captured in late telophase Žarrowheads.. Vibratome section through central brain of female crab measuring 37 mm in cw Žpost-injection survival time 22 h..

138

M. Schmidtr Brain Research 762 (1997) 131–143

beled somata were captured in different phases of mitosis. After a post-injection survival time of 1 month significantly more Ž P - 0.001; Student’s t-test. nuclei in the HBC were labeled than after survival times of up to 120 h Ž29 to 45; mean: 38.2 " 7.1 S.D.; n s 6 HBCs. ŽFig. 6c and Fig. 7c.. As with shorter survival times the labeled nuclei were still arranged in a tightly packed group, but

Fig. 7. Graphic representation of the quantitative results of in vivo BrdU labeling in the brain of post-larval shore crabs. a: number of BrdU-positive nuclei per LC in crabs of different size Žcarapace width. after a post-injection survival time of 17 h. Note that no clear correlation between the size of a crab and the number of BrdU-labeled nuclei exists. b,c: number of BrdU-positive nuclei per LC Žb. or HBC Žc. in crabs after different post-injection survival times. Note that from 5.5 to 120 h post-injection survival time the number of BrdU-labeled nuclei does not change systematically, that after a prolonged survival time of 1 month, however, the number of labeled nuclei approximately doubles.

Fig. 6. Identification of proliferating cells in the soma cluster containing the cell bodies of hemiellipsoid body local interneurons ŽHBC. with in vivo BrdU labeling. a,b: short post-injection survival time Ž120 h.. Vibratome section through eyestalk ganglion of a male crab measuring 38 mm in cw. a: in the edge of the HBC Žasterisk. attached to the neuropil of the hemiellipsoid body ŽHB. a line of nuclei is labeled Žarrowhead. demarcating the edge of the cluster as a proliferative zone. Some isolated somata are labeled in other areas of the eyestalk ganglion, especially the optic lobe ŽOPL.. b: higher magnification reveals dense clustering as well as neuronal size and shape Žspherical. of labeled nuclei. Note size and staining gradient of labeled nuclei, with the smallest and most intensely labeled nuclei close to the HB Žright. and larger, lightly labeled nuclei further away Žleft.. c: long post-injection survival time Ž1 month.. One month after BrdU injection labeled nuclei still forming a tightly clustered group are present in the HBC. Note that group of labeled nuclei has moved away from the proliferative zone at the HB towards the outside of the HBC. Vibratome section through eyestalk ganglion of a male crab measuring 45 mm in cw Žpost-injection survival time 720 h..

M. Schmidtr Brain Research 762 (1997) 131–143

this group was displaced from the proliferative zone into the center of the HBC ŽFig. 6c.. After one month the nuclei farthest from the HB neuropil were stained with the highest intensity. In addition to a relatively constant number of nuclei in the proliferative zones of the LC and the HBC, in vivo BrdU labeling stained a highly variable number of nuclei throughout the entire brain Žpost-injection survival times 5.5–120 h; Fig. 4a and Fig. 6a.. By their size, shape and location Žinside of neuropils. most of these additional nuclei were characterized as belonging to somata of either glial cells or cells of the connective tissue forming the blood vessels. Some of them, however, occurred in the other soma clusters and by size and shape Žspherical. could represent neurons or neuronal precursor cells ŽFig. 4a.. Usually very few labeled nuclei were scattered throughout the other soma clusters, only in the sheet-like soma clusters of the visual neuropils Žlamina, medulla externa, medulla interna. groups of small labeled nuclei with varying shapes occurred at their medial-most edges. The number of labeled nuclei in these regions resembling proliferative zones varied from very few to more than 50 between individuals, without obvious correlation to size, sex or moult stage Ža detailed report on the proliferation in the visual system will appear elsewhere..

4. Discussion The observations reported here provide strong accumulative evidence that neurogenesis occurs in the brain of juvenile and adult shore crabs. By itself, positive BrdUlabeling of nuclei with neuronal shape, size and localization after relatively short survival times Žas shown for the LC and the HBC, two soma clusters of the central olfactory pathway. does not provide conclusive evidence for neurogenesis. This is because it leaves the possibilities that Ž1. the proliferating somata differentiate to non-neuronal cells, that Ž2. a high percentage of them undergoes apoptosis shortly after mitosis, as has been observed among a population of constitutively proliferating cells in the subependymal layer of the adult mammalian forebrain w30x, or that Ž3. DNA is synthesized without subsequent mitosis. For various reasons, however, these alternatives are highly unlikely. The first two possibilities are contradictory to the result of the BrdU labeling experiments with a post-injection survival time of 1 month. Even after this prolonged period clearly labeled nuclei were present in both soma clusters showing that cell death can be the fate of only a minor fraction of the proliferating cells. Furthermore the labeled nuclei clearly retained neuronal morphology Žspherical. and as a group migrated further into the center of the respective soma clusters, where histological methods show tightly packed somata of mature neurons only rarely interspersed with non-neuronal cell types with distinctive

139

morphological characteristics Žglial cells, connective tissue forming the wall of arterioles.. Thus it is unlikely that the majority of proliferating cells in the LC and the HBC differentiate into glial cells or into connective tissue. The possibility of BrdU incorporation without subsequent mitosis is inconsistent with the observation of somata in the proliferative zone of the LC and the HBC with neuronal morphology captured in various phases of mitosis with histological methods andror in vivo BrdU labeling. Therefore the results of the in vivo BrdU labeling strongly indicate that continuous neurogenesis occurs in the central olfactory pathway throughout the entire juvenile and adult life of shore crabs. This interpretation is further corroborated by the linear increase in the number of OPNs throughout the post-larval life of shore crabs as shown by the independent counts of their somata and axons. Per se the increase in soma and axon number could also have alternative explanations not involving neurogenesis like migration of neurons from other areas into the LC or change in the percentage of bifurcating axons. Although these possibilities cannot be ruled out entirely, these alternative explanations appear unlikely considering the strong positive evidence for neurogenesis from the in vivo BrdU labeling. Still it remains that final proof for neurogenesis in the LC and the HBC of adult shore crabs is lacking. Such definitive proof could be obtained by double-labeling experiments with BrdU and markers of mature neurons. For the CNS of vertebrates a host of antibodies against specific neuronal antigens Že.g. neurofilaments, neuron specific enolase, neuN. are available and these antibodies have been utilized in combination with BrdU or other proliferation markers Žw 3 Hxthymidin; replication deficient retrovirus containing the reporter gene lacZ . to conclusively demonstrate neurogenesis in the olfactory bulb and the dentate gyrus of the adult mammalian brain, e.g. w8,22,32x. Unfortunately, specific neuronal markers for the arthropod CNS have not yet been identified. Therefore the lack of final proof for the neuronal fate of cells proliferating in the CNS of adult arthropods is a more general problem and also applies to the recently reported ‘‘neurogenesis’’ among the Kenyon cells in the adult brain of diverse insect species w6,7x. In these investigations the neuronal fate of the newly formed somata was concluded from their location Žin the cluster of Kenyon cell somata., their neuron-like size and shape, and lack of labeling by a glial cell marker. Two other approaches potentially could provide final evidence for the neuronal fate of the cells proliferating in the LC of adult shore crabs. One would be to utilize the neurochemical phenotype of the OPNs for double labeling with BrdU. Unfortunately, however, the neurotransmitters of the OPNs of any decapod crustacean have not been identified to date in spite of numerous immunocytochemical investigations, e.g. w39,41,46x. The other approach would be to use the unique property of the OPNs to possess an axon projecting into the eyestalk ganglia for double labeling experiments.

140

M. Schmidtr Brain Research 762 (1997) 131–143

Backfilling the OPN somata via their axons in the protocerebral tract has been achieved with biocytinrneurobiotin w45x and other neuronal tracers w38x. Currently, experiments combining neurobiotin backfills of OPN axons with in vivo BrdU labeling in adult shore crabs are underway and recently the first double-labeled somata have been identified in the LC ŽSchmidt and Demuth, unpublished.. Further corroboration of these initial results is required to provide conclusive direct evidence for neurogenesis in the LC of adult shore crabs. One important question arising from the results reported here concerns the mechanism by which the neurons in the central olfactory pathway pare generated in juvenile and adult shore crabs, especially whether this mode of neuron formation conforms to or differs from neurogenesis during embryonic and larval development. The general scheme for neurogenesis during embryonic and larval development is consistent throughout the CNS of insects and decapod crustaceans: neuroblasts differentiate from the ectoderm and through a series of unequal mitoses produce ganglion mother cells, which each divide into two neurons, e.g. w15,49,54x. Typically, neuroblasts and ganglion mother cells are considerably larger than terminally differentiated neurons and they are located at the outer edge of the soma clusters formed by their mitotic activity, e.g. w15,19,31x. The precursor cells that are mitotically active in the Kenyon cell cluster of adult insects w7x and in the LC of early postlarval crabs w16x share these criteria and therefore are regarded as neuroblasts or ganglion mother cells, respectively. In contrast, the nuclei that were BrdU-labeled in the LC and HBC of juvenile and adult shore crabs belong to the smallest somata of these cluster and they are located in the center of the soma cluster. These pronounced differences indicate that the precursor cells, that are mitotically active in the LC and HBC of the juvenile and adult shore crabs, are no typical neuroblasts or ganglion mother cells. Thus, I conclude that most likely neurogenesis in the olfactory pathway of the brain of juvenile and adult shore crabs occurs by a different mode than during embryonic and larval development. The experiments reported here do not allow to characterize this presumptively novel mode of neurogenesis in terms of the underlying proliferation kinetics of the putative precursor cells given by the total cell cycle time, duration of S-phase and total proliferating population. To determine these parameters multiple BrdU injections are required w30x, which have not yet been done. However, the data reported here allow some more general conclusions. Ž1. The finding that the total number of labeled nuclei in the LC and the HBC is independent of post-injection survival times ranging from 5.5 to 120 h indicates that BrdU is available for incorporation into DNA for not longer than 5.5 h, since otherwise an increasing number of labeled nuclei would be expected at longer survival times. This is consistent with BrdU incorporation times known from vertebrates; for arthropods, however, it has not yet

been determined how long BrdU is available for incorporation. Ž2. This finding also leads to the conclusion that after the injection of BrdU and its incorporation into DNA no more than one cell division occurs in the labeled cells during the first 120 h. Otherwise a dilution of the initially introduced label and an increase in the number of labeled nuclei would be expected during this time. Ž3. Based on the same line of argumentation, the significant increase in the number of labeled nuclei in both soma clusters after a prolonged survival time of 1 month that is accompanied by a clear dilution of the label in many of the stained nuclei, strongly suggests that in the period from 5 days to one month after BrdU injection further mitoses occur in some of the cells that were initially labeled. Ž4. The observation that after this prolonged survival time of 1 month the labeled nuclei were located further away from the proliferative zones of both clusters than after survival times up to 120 h but still formed a compact group indicates that likely they do not migrate actively but rather are pushed outward by newly arising cells in the proliferative zones. Ž5. From this observation it can further be concluded that likely an age-gradient of cells exists in the LC and HBC with the youngest neurons close to the neuropil and older ones closer towards the outer surface. This age-gradient may be reflected by the clear size and staining gradient that structures the LC in the same pattern. Ž6. The independence of the number of labeled nuclei from the size and hence age of the studied adult crabs indicates that neurogenesis in the LC Žand likely also the HBC. is a continuous process that does not slow down in older animals. This is comparable to the genesis of granule cells in the olfactory bulb of adult rats that are produced at a constant rate for the entire life span after an initial decline in the rate of production during the first year w21x. In contrast, the production of new granule cells in the dentate gyrus of adult rats decreases dramatically in older animals w22,51x. Ž7. If the assumption of a constant rate of neurogenesis is correct the minimal value of this rate can be estimated from the total life-span of Carcinus maenas Žapproximately 3 years w29x. and the total increase in the number of somata in the LC of post-larval crabs Ž11 500 to 24 000. as determined in this study. From this calculation it follows that at least 10 new OPNs per day have to be generated in each LC, provided that no OPNs die during adult life. If neuronal death occurs, which is indicated by the observation of some BrdU-positive pyknotic nuclei in the LC reported here, this would have to be compensated by a higher rate of neurogenesis. The mean number of 31.3 BrdU-positive nuclei per LC appears to be quite sufficient to support this estimated rate of neurogenesis in the LC. Neurogenesis in the brain of juvenile and adult shore crabs shows interesting parallels and differences to the few cases of neurogenesis in the brain of adult insects that have been reported to date w3,6,7,53x. In the brain of adult insects neurogenesis has exclusively been observed among Kenyon cells, which are likely equivalent to the HB local

M. Schmidtr Brain Research 762 (1997) 131–143

interneurons of decapod crustaceans w14x. Thus, the finding that neurogenesis occurs in the HBC of adult shore crabs directly conforms to the insect situation. For the even more prominent neurogenesis of OPNs in the shore crab, however, no parallel exist in insects. Interestingly, in decapod crustaceans, the OPNs represent globuli cells as do the HB local interneurons and insect Kenyon cells, e.g. w14x. Globuli cells are characterized mainly by a small and uniform size, a relatively large nucleus and their occurrence in large populations w14,27,45x. A much Ž10 to 100-fold. higher number of OPNs is one important feature in which the central olfactory pathway of decapod crustaceans differs from that of insects w28,45x and this higher neuron number has been interpreted as an evolutionary older state w45x. Consequently, one may assume that neurogenesis in the adult arthropod brain, since it appears to be primarily present among globuli cells, could also represent an evolutionary older principle. This interpretation would explain why in the more derived insects neurogenesis in the adult brain is limited to Kenyon cells, which are the only true globuli cells left, and why among the few insect species examined, several – the honeybee Apis mellifera, the locust Locusta migratoria, and the cockroach Periplaneta americana – do not even show it w6,9x. If it is true that neurogenesis is a basic principle of globuli cell organization, the lack of a proliferative zone in the medial soma cluster of the shore crab olfactory deutocerebrum is particularly surprising, since this cluster also contains predominantly globuli cells w27,45x. Analyses of other, phylogenetically more basic decapod species Žbrachyuran crabs are considered to be highly derived w50x. are needed to decide if neurogenesis really does not occur among all globuli cells of the decapod brain or if shore crabs are an exceptional case among decapod crustaceans. A common theme emerges when comparing neurogenesis in the brain of adult arthropods and vertebrates w1,2,5,8,20,25,26,34,36,56 x. Although neurogenesis in the brain of adult vertebrates is found in various regions like the hippocampus w5x, the cerebellum w56x and the visual pathway w20,34x, proliferation appears to be most prominent and species-independent among neurons constituting the olfactory bulb, e.g. w1,2,8,25,26,36x. Since adult neurogenesis is limited to or most prominent in the central olfactory pathway of insects w3,7,53x and decapod crustaceans Žthis study., respectively, there appears to be a general tendency for brain areas processing olfactory input to acquire new neurons throughout the entire life. Whether this apparent similarity is a mere coincidence or reflects a basic functional constrain of olfactory processing remains an open question. In this context, however, it is interesting that in mammals, olfaction is the only sensory pathway in which primary sensory neurons undergo a continuous turnover, e.g. w13x. A similar turnover of olfactory receptor cells, in addition to their continuous formation through successive moults, has recently been reported in a decapod crustacean w40x. Thus it may be speculated that neurogene-

141

sis in the olfactory pathway of the adult brain is linked to continuous turnover and formation of olfactory receptor neurons in the periphery. If these processes are indeed functionally linked, a life-long adaptive respecification of the olfactory system driven by changes in the olfactory environment could be envisioned. In decapod crustaceans also non-olfactory sensory input to the CNS is mediated by primary sensory neurons, for some of which an increase in number throughout the juvenile and adult life of the animals has been demonstrated w24,48x. Thus, the question remains if neurogenesis also occurs in the corresponding central sensory pathways. In the shore crab neurogenesis may indeed occur in other sensory pathways since soma clusters not belonging to the olfactory deutocerebrum, especially in the optic lobe, contained some nuclei of neuronal shape that were BrdUlabeled. However, conclusive evidence that the BrdUlabeled nuclei in these soma clusters are neurons or neuronal precursor is lacking. In this context it is interesting that the number of neuronal somata in the terminal ganglion of crayfish was reported to increase significantly from juvenile to adult animals w35x, but this brief report has never been confirmed by a thorough study. Clearly, basic morphological analyses in juvenile and adult decapod crustaceans are needed to determine the extent to which neurogenesis is associated with the growth of the CNS in general w17,18x and with the increasing number of sensory afferents in particular.

Acknowledgements I thank W. Roloff for the crab drawing and Dr. Mike Michel ŽUniversity of Utah. for help in preparing the manuscript. This research was supported by a grant from the Deutsche Forschungsgemeinschaft ŽSchm 738r4-1..

References w1x J.R. Alonso, J. Lara, E. Vecino, R. Covenas, ˜ J. Aijon, ´ Cell proliferation in the olfactory bulb of adult freshwater teleosts, J. Anat. 163 Ž1989. 155–163. w2x S.A. Bayer, 3 H-thymidine-radiographic studies of neurogenesis in the rat olfactory bulb, Exp. Brain Res. 50 Ž1983. 329–340. w3x M. Bieber, D. Fuldner, Brain growth during the adult stage of a holometabolous insect, Naturwissenschaften 66 Ž1979. 426. w4x J. Boeckh, K.-D. Ernst, Contribution of single unit analysis in insects to an understanding of olfactory function, J. Comp. Physiol. A 161 Ž1987. 549–565. w5x H.A. Cameron, C.S. Woolley, B.S. McEven, E. Gould, Differentiation of newly born neurons and glia in the dendate gyrus of the adult rat, Neuroscience 56 Ž1993. 337–344. w6x M. Cayre, C. Strambi, P. Charpin, R. Augier, M.R. Meyer, J.S. Edwards, A. Strambi, Neurogenesis in adult insect mushroom bodies, J. Comp. Neurol. 371 Ž1996. 300–310. w7x M. Cayre, C. Strambi, A. Strambi, Neurogenesis in an adult insect brain and its hormonal control, Nature 368 Ž1994. 57–59. w8x F.S. Corotto, J.A. Henegar, J.A. Maruniak, Neurogenesis persists in

142

w9x

w10x

w11x w12x w13x w14x w15x

w16x

w17x

w18x

w19x

w20x w21x

w22x

w23x w24x

w25x w26x

w27x

w28x

w29x

w30x

w31x

M. Schmidtr Brain Research 762 (1997) 131–143 the subendymal layer of the adult mouse brain, Neurosci. Lett. 149 Ž1993. 111–114. S.E. Fahrbach, J.L. Strande, G.E. Robinson, Neurogenesis is absent in the brains of adult honey bees and does not explain behavioral neuroplasticity, Neurosci. Lett. 197 Ž1995. 145–148. W. Gnatzy, M. Schmidt, J. Rombke, Are the funnel-canal organs the ¨ ‘‘campaniform sensilla’’ of the shore crab Carcinus maenas ŽCrustacea, Decapoda.? I. Topography, external structure and basic organization, Zoomorphology 104 Ž1984. 11–20. C.K. Govind, Development of asymmetry in the neuromuscular system of lobster claws, Biol. Bull. 167 Ž1984. 94–119. C.K. Govind, J. Pearce, Growth of inhibitory innervation in a lobster muscle, J. Morphol. 199 Ž1989. 197–205. P.P.C. Graziadei, Cell dynamics in the olfactory mucosa, Tissue Cell 5 Ž1973. 113–131. B. Hanstrom, ¨ The olfactory centers in crustaceans, J. Comp. Neurol. 38 Ž1925. 221–250. S. Harzsch, R.R. Dawirs, Neurogenesis in larval stages of the spider crab Hyas araneus ŽDecapoda, Brachyura.: proliferation of neuroblasts in the ventral nerve cord, Roux’s Arch. Dev. Biol. 204 Ž1994. 93–100. S. Harzsch, R.R. Dawirs, Neurogenesis in the developing crab brain: postembryonic generation of neurons persists beyond metamorphosis, J. Neurobiol. 29 Ž1996. 384–398. S.M. Helluy, J.L. Benton, K.A. Langworthy, M.L. Ruchhoeft, B.S. Beltz, Glomerular organization in developing olfactory and accessory lobes of American lobsters: stabilization of numbers and increase in size after metamorphosis, J. Neurobiol. 29 Ž1996. 459–472. S.M. Helluy, M.L. Ruchhoeft, B.S. Beltz, Development of the olfactory and accessory lobes in the American lobster: an allometric analysis and its implications for the deutocerebral structure of decapods, J. Comp. Neurol. 357 Ž1995. 433–445. K. Ito, Y. Hotta, Proliferation pattern of postembryonic neuroblasts in the brain of Drosophila melanogaster, Dev. Biol. 149 Ž1992. 134–148. M.S. Kaplan, Neurogenesis in the 3-month-old rat visual cortex, J. Comp. Neurol. 195 Ž1981. 323–338. M.S. Kaplan, N.A. McNelly, J.W. Hinds, Population dynamics of adult-formed granule neurons of the rat olfactory bulb, J. Comp. Neurol. 239 Ž1985. 117–125. H.G. Kuhn, H. Dickinson-Anson, F.H. Gage, Neurogenesis in the dentate gyrus of the adult rat: age-related decrease of neuronal progenitor proliferation, J. Neurosci. 16 Ž1996. 2027–2033. M.S. Laverack, The numbers of neurones in decapod crustacea, J. Crust. Biol. 8 Ž1988. 1–11. J.G. Letourneau, Addition of sensory structures and associated neurons to the crayfish telson during development, J. Comp. Physiol. 110 Ž1976. 13–23. C. Lois, A. Alvarez-Buylla, Long-distance neuronal migration in the adult mammalian brain, Science 264 Ž1994. 1145–1148. M.B. Luskin, M.S. Boone, Rate and pattern of migration of lineally-related olfactory bulb interneurons generated postnatally in the subventricular zone of the rat, Chem. Senses 19 Ž1994. 695–714. D. Mellon Jr., V. Alones, Cellular organization and growth-related plasticity of the crayfish olfactory midbrain, Microsc. Res. Tech. 24 Ž1993. 231–259. D. Mellon Jr., V. Alones, M.D. Lawrence, Anatomy and fine structure of neurons in the deutocerebral projection pathway of the crayfish olfactory system, J. Comp. Neurol. 321 Ž1992. 93–111. M. Moritz, H.-E. Gruner and W. Dunger, Lehrbuch der speziellen Zoologie. Band 1: Wirbellose Tiere. 4. Teil: Arthropoda Žohne Insecta., Gustav Fischer Verlag, Stuttgart, 1993. C.M. Morshead, D. Van der Kooy, Postmitotic death is the fate of constitutively proliferating cells in the subependymal layer of the adult mouse brain, J. Neurosci. 12 Ž1992. 249–256. R.H. Nordlander, J.S. Edwards, Postembryonic brain development in the monarch butterfly, Danaus plexippus plexippus L. III. Morpho-

w32x

w33x w34x

w35x w36x

w37x

w38x

w39x

w40x

w41x

w42x w43x

w44x

w45x

w46x

w47x

w48x

w49x

w50x

w51x w52x

w53x

genesis of centers other than the optic lobes, Wilhelm Roux’ Arch. 164 Ž1970. 247–260. H.J. Okano, D.W. Pfaff, R.B. Gibbs, RB and Cdc2 expression in brain: correlations with 3 H-thymidine incorporation and neurogenesis, J. Neurosci. 13 Ž1993. 2930–2938. D. Purves, Body and Brain. A Trophic Theory of Neuronal Connections, Harvard University Press, Cambridge, London, 1988. P.A. Raymond, S.S. Easter Jr., Postembryonic growth of the optic tectum in goldfish. I. Location of germinal cells and numbers of neurons produced, J. Neurosci. 3 Ž1983. 1077–1091. R.L. Roth, P. Suppes, Postembryonic addition of neurons to the abdominal nerve cord of crayfish, Anat. Rec. 175 Ž1973. 430. P. Rousselot, C. Lois, A. Alvarez-Buylla, Embryonic ŽPSA. N-CAM reveals chains of migrating neuroblasts between the lateral ventricle and the olfactory bulb of adult mice, J. Comp. Neurol. 351 Ž1995. 51–61. D. Sandeman, R. Sandeman, C. Derby, M. Schmidt, Morphology of the brain of crayfish, crabs, and spiny lobsters: a common nomenclature for homologous structures, Biol. Bull. 183 Ž1992. 304–326. D.C. Sandeman, R.E. Sandeman, Electrical responses and synaptic connections of giant serotonin-immunoreactive neurons in crayfish olfactory and accessory lobes, J. Comp. Neurol. 341 Ž1994. 130–144. D.C. Sandeman, R.E. Sandeman, A.R. Aitken, Atlas of serotonincontaining neurons in the optic lobes and the brain of the crayfish Cherax destructor, J. Comp. Neurol. 269 Ž1988. 465–478. R.E. Sandeman, D.C. Sandeman, Pre- and postembryonic development, growth and turnover of olfactory receptor neurons in crayfish antennules, J. Exp. Biol. 199 Ž1996. 2409–2418. R.E. Sandeman, D.C. Sandeman, A.H.D. Watson, Substance P antibody reveals homologous neurons with axon terminals among somata in the crayfish and crab brain, J. Comp. Neurol. 294 Ž1990. 569–582. M. Schmidt, Postlarval neurogenesis in the olfactory midbrain of a decapod crustacean, Soc. Neurosci. Abstr. 21 Ž1995. 1524. M. Schmidt, Neurogenesis in the central olfactory pathway of the shore crab brain continues throughout adult life, Verh. Dtsch. Zool. Ges. 89 Ž1996. 77. M. Schmidt, B.W. Ache, Antennular projections to the midbrain of the spiny lobster. II. Sensory innervation of the olfactory lobe, J. Comp. Neurol. 318 Ž1992. 291–303. M. Schmidt, B.W. Ache, Processing of antennular input in the brain of the spiny lobster, Panulirus argus. II. The olfactory pathway, J. Comp. Physiol. wAx 178 Ž1996. 605–628. M. Schmidt, B.W. Ache, Immunocytochemical analysis of glomerular regionalization and neuronal diversity in the olfactory deutocerebrum of the spiny lobster, Cell Tissue Res. 287 Ž1997. 541–563. M. Schmidt, W. Gnatzy, Specificity and response characteristics of gustatory sensilla Žfunnel-canal organs. on the dactyls of the shore crab, Carcinus maenas ŽCrustacea Decapoda., J. Comp. Physiol. A 166 Ž1989. 227–242. B. Schmitz, Postembryonic development of the crayfish Procambarus clarkii and its tailfan mechanosensory system, in: R.N. Singh ŽEd.., Nervous Systems. Principles of Design and Function, Wiley Eastern Limited, New Delhi, 1992, pp. 69–90. G. Scholtz, Cell lineage studies in the crayfish Cherax destructor ŽCrustacea, Decapoda.: germ band formation, segmentation, and early neurogenesis, Roux’s Arch. Dev. Biol. 202 Ž1992. 36–48. G. Scholtz, S. Richter, Phylogenetic systematics of the reptantian Decapoda ŽCrustacea Malacostraca., Zool. J. Linn. Soc. 113 Ž1995. 289–328. T. Seki, Y. Arai, Age-related production of new granule cells in the adult dentate gyrus, Neuroreport 6 Ž1995. 2479–2482. M. Spencer, K.A. Linberg, Ultrastructure of aesthetasc innervation and external morphology of the lateral antennule setae of the spiny lobster Panulirus interruptus ŽRandall., Cell Tissue Res. 245 Ž1986. 69–80. G.M. Technau, Fiber number in the mushroom bodies of adult

M. Schmidtr Brain Research 762 (1997) 131–143 Drosophila melanogaster depends on age, sex and experience, J. Neurogenet. 1 Ž1984. 113–126. w54x J.B. Thomas, M.J. Bastiani, M. Bate, C.S. Goodman, From grasshopper to Drosophila: a common plan for neuronal development, Nature 310 Ž1984. 203–207. w55x S.J. Young, S.M. Royer, P.M. Groves, J.C. Kinnamon, Three-di-

143

mensional reconstructions from serial micrographs using the IBM PC, J. Electron Microsc. Tech. 6 Ž1987. 207–217. w56x K.H. Zupanc, I. Horschke, Proliferation zones in the brain of adult gymnotiform fish: a quantitative mapping study, J. Comp. Neurol. 353 Ž1995. 213–233.