- Email: [email protected]
The morphology of the cervical giant fiber neuron ofDrosophila
Brain Research, 221 (1981) 213-217
213
Elsevier/North-Holland Biomedical Press
Research Reports THE MORPHOLOGY
OF T H E C E R V I C A L G I A N T F I B E R N E U R O N OF
DR OSOPHILA
MARY KOTO, MARK A. TANOUYE, ALBERTO FERRUS, JOHN B. THOMAS and ROBERT J. WYMAN ( M.K., J.B.T. and R.J. W. ) Yale University, P.O. Box 6666, New Haven, Conn. 06511 and ( M.A.T. and A.F.) Division of Biology, California Institute of Technology, Pasadena, Calif. 91125 (U.S.A.)
(Accepted February 19th 1981) Key words: Drosophila melanogaster - - giant fiber - - Lucifer yellow
SUMMARY The morphology of the cervical giant fiber ( C G F ) neuron of Drosophila melanogaster was studied by intracellular injection of Lucifer yellow dye. The C G F neuron is the c o m m a n d cell in a motor circuit causing visually driven escape behavior: a single action potential in a C G F axon produces patterned activity in j u m p and flight muscles. The present study identified the C G F cell body, a large soma located in the posterior part of the lower ipsilateral protocerebrum. The main process runs anteriorly from the cell body, extends three branches, and turns posteromedially while descending through the brain. The C G F axon courses through the cervical connective and ends within the mesothoracic neuromere of the thoracic ganglion. Thus, the C G F neuron is an interneuron, not a motoneuron as previously believed. We have been isolating mutants that affect C G F neuron-mediated behavior. Comparison of C G F neuron morphology in wildtype strains with that in these mutants will allow identification of genes that affect the development, structure, and connections of the C G F neuron.
INTRODUCTION The cervical giant fiber ( C G F ) neurons in Drosophila melanogaster are a bilateral pair of nerve cells that extend from the brain to the thoracic ganglion. The C G F axons were first described by Power in 1948 as giant fibers 7. Though only 5 # m in diameter, the C G F axons are several times larger than any other axon in the cervical connective1, 7. The C G F neuron is a c o m m a n d cell in a fast-conducting motor pathway. A single C G F action potential elicited by intracellular stimulation drives the
214 jump and the flight muscles with a stereotyped pattern of spikes 9. A light-off stimulus elicits a spike in the CGF neuron, producing the same pattern of muscle activity 10. It thus appears that a visually induced escape response is mediated by the C G F neurons. In the present study, the CGF neuron was stained by intracellular injection of the fluorescent dye, Lucifer yellow CH 8. This method allowed visualization of the entire neuron, from its cell body in the brain to its terminus in the thoracic ganglion. The morphology of the CGF neuron is notable among invertebrates for its simplicity. Our results, together with King and Wyman's EM study 4 of the thoracic ganglion, show that the C G F neuron is an interneuron, not a motoneuron as previously believeda, 7. An abstract describing preliminary results appeared earlierL MATERIALS AND METHODS Experiments were performed on adult female Drosophila melanogaster. Flies were either a wildtype strain (Canton-Special), or white-eyed flies of the white, brown, scarlet, or cinnabar brown genotype. No difference in CGF neuron morphology within the thoracic ganglion was found between the wildtype (n = 5) and the 3 white-eyed strains (n = 6). The brain tissue of wildtype flies has a natural fluorescence that obscures visualization of the stained C G F neuron in brain wholemounts. Removal of the eyes before fixation reduced, but did not eliminate the fluorescence. Because whiteeyed flies had a minimal amount of fluorescence, they were used to view the brain in wholemount. Preliminary results comparing brain sections of wildtype (n = 2) and white-eyed (n = 2) flies showed no significant differences in the structure of the C G F neuron. Dissection, recording techniques, and verification of CGF axon impalement in the cervical connective are described by Tanouye and Wyman 9. Micropipettes, in which tips had been filled with 3 ~ Lucifer yellow, were backfilled with 0.05 ~,, LiCI. Lucifer yellow was iontophoretically injected into a CGF axon for 5-30 min with hyperpolarizing pulses or DC current of4-15 nA. Immediately after dye injection, the flies were fixed in buffered formalin for at least 30 min before dissecting the brain and ganglion out of the body. Wholemounts were prepared by dehydrating tissue in an ascending alcohol series and clearing in methyl benzoate. Sections were prepared by embedding brains in Spurr's medium and sectioning at 4/~m. All preparations were viewed under a Zeiss microscope with fluorescence optics as described by Stewart s. RESULTS AND DISCUSSION The cell body of the C G F neuron is very large, about 20/zm in diameter (n = 9). Ipsilateral to its axon, the cell body is located close to the posterior border of the brain adjacent to the neuropile of the protocerebrum about 30 #m dorsal to the oesophageal canal. (Fig. 1A and B). A neurite about 60/~m long extends anteriorly, connecting the cell body with the main process. The main process then descends posteromedially, curves around the oesophageal canal, and exits in the cervical connective. Prior to its posteromedial descent, the main process has 3 branches - - two ipsilateral and one
215
Fig. 1. A: wholemount of the brain. The anterior part of the brain is at the top of the picture. The CGF cell body is in the posterior part of the brain. The large ipsilateral branch that ends in the antennal glomerulus runs anteriorly. The contralateral branch and an unidentified contralateral cell body are faintly stained. The CGF axon turns medially, descends ventrally, and exits the brain in the cervical connective. B : horizontal section of the brain with CGF cell bodies. One CGF cell body is stained (right arrow) while its contralateral homologue (left arrow) is not. The small ipsilateral branch is in the central part of the brain. The section is 36 /~m dorsal to the oesophageal canal. C: wholemount of the thoracic ganglion. The tuft region is 57 #m from the axon terminus. In this preparation, two separate CGF axon penetrations were made. One lasted only 2 rain and is most likely responsible for the faint filling of the contralateral homologue. All space bars = 100 #m. contralateral. The larger ipsilateral branch, stained in 8 preparations, runs ventroanteriorly and ends in the protocerebral lobe close to the anterior edge o f the brain. This b r a n c h sends out n u m e r o u s short, spiny processes, especially near its terminus. In 4 preparations, a second ipsilateral branch was visible. Shorter and thinner than the first, this b r a n c h extends dorsomedially f r o m the main axon. The contralateral branch, faintly stained in the 7 preparations where it was seen, crosses the midline above the oesophageal canal and ends in the posterior part o f the protocerebrum. The contralateral b r a n c h in one preparation ran to the cell b o d y o f the contralateral C G F neuron which also stained. The staining o f the C G F neuron's contralateral h o m o l o g u e p r o b a b l y results f r o m dye-coupling between the two neurons. Since dye-coupling occurs between cells that are electrically coupled a, the C G F neurons m a y be connected by an electrical synapse. In 3 other preparations, the contralateral b r a n c h ended in small dots, processes, or an unidentified cell b o d y in the posterior part o f the brain. In all ganglion preparations (n = l l ) , the structure o f the thoracic C G F processes was simple and unvarying (Fig. 1C). Entering the ganglion along the dorsal midline, the C G F axon gradually descends ventrally. The C G F axon turns ventrolaterally within the mesothoracic neuromere. Where the axon begins to turn, it sends out a dorsal spray o f short, fine processes that resembles a tuft. The C G F axon ends
216 about 80/~m from the lateral edge of the ganglion without branching. Only one of the pair of C G F axons stained in the thoracic ganglion when a single penetration of a C G F axon was made. Thus, the C G F axons are not dye-coupled within the thoracic ganglion. In the present study, the C G F cell body was located close to the neuropile in the posterior part of the brain just above the oesophageal canal. In a [3H]thymidine labeling experiment with postembryonic DrosophilamelanogasterWhite and Kanke111 found that the neurons with cell bodies closest to the neuropile had the earliest birthdays. Therefore, the close proximity of the cell body to the neuropile probably indicates that the C G F neuron underwent its final division early in development. The C G F cell bodies, conspicuous for their large size, were described earlier by Power, but he did not associate them with the giant fibers~. Our results conflict with those of Levine and Tracey 5. Looking at wholemounts prepared by backfilling the entire cervical connective with cobalt dye, Levine and Tracey believed that the CGF cell body was in the anterior portion of the brain. The structure that they term the C G F cell body could be that of any cervical fiber. The branches of the CGF neuron in the brain send out short, fine processes in several regions, including the protocerebral lobe, the central commissure, and the central body. It is likely that the C G F neuron makes synaptic contacts via the fine processes in these areas. Fibers in these areas include interneurons from the eyes and antennae as well as from association centers in the brain 6. In the thoracic ganglion, our results are consistent with King and Wyman's finding that the CGF neuron is an interneuron 4. In their serial section micrographs, the C G F axon ended within the mesothoracic neuromere; in the present study, the large, brightly stained CGF axon always ended abruptly within the mesothoracic neuromere. In previous reports the C G F neuron was described with two extra processes in the thoracic ganglionS, 7. Processes other than the C G F axon stained faintly in a number of our thoracic ganglion preparations (n = 5). Since all processes had cell bodies associated with them, it is most likely that they are separate neurons dyecoupled to the CGF neuron. When the antennal nerve was backfilled with cobalt dye, Strausfeld and his associates also found cells in the thoracic ganglion dye-coupled to the CGF neuron (personal communication). In the present study, two cells were frequently dye-coupled to the CGF neuron. One runs from the terminus of the C G F neuron to the lateral edge of the ganglion near the ipsilateral posterior dorsal mesothoracic nerve. King and Wyman 4 found that the large motoneuron to the jump muscle (the tergotrochanteral muscle) contacts the terminus of the C G F neuron and exits the ipsilateral posterior dorsal mesothoracic nerve (PDM nerve). The other cell found in the present study crosses from the tuft region of the CGF neuron to the contralateral side of the ganglion, where it exits in the PDM nerve. King and Wyman 4 also found an interneuron (the peripherally synapsing interneuron) that contacts the CGF neuron near the midline, exits the ganglion in the contralateral PDM nerve, and synapses in the peripheral nerve with the 5 motoneurons innervating the dorsal longitudinal flight muscle. It is likely that the cells stained in the present study - - the
217 cell that abuts the C G F neuron and the contralateral cell - - are, respectively, the tergotrochanter motoneuron and the PSI found by King and Wyman 4. Both the dyecoupling of these cells to the C G F neuron and physiological evidence 9 suggest that the C G F neuron makes electrical synapses with the tergotrochanter motoneuron and the PSI. Our results can be applied to developmental studies of the C G F neuron and its connections. The large size of the C G F cell body readily allows identification of the neuron, even in unstained preparations (Fig. 1B). I f the C G F cell body attains its large size early in development, its cell body can be injected with Lucifer yellow and outgrowth of its processes traced a. Mutations which affect the structure and function of the C G F pathway could be used to analyze the events in C G F netiron development. As a first step in such an analysis, comparison of wildtype and mutant flies would reveal defects in C G F neuron morphology. ACKNOWLEDGEMENTS Development of the Lucifer yellow injection technique in Drosophila melanogaster and the initial results were obtained in the laboratory of Dr. Seymour Benzer. The authors wish to thank him for the use of his facilities and his support. The authors also thank Walter Stewart for his gifts of Lucifer yellow. Research was supported by NSF PCM-7911771 to Seymour Benzer, N S F predoctoral fellowship to M.K., Muscular Dystrophy Association postdoctoral fellowship to M.A.T., Gosney Fund fellowhip to A.F., N I H training grant to J.B.T., and N I H NS-07314 and NS14887 to R.J.W.
REFERENCES 1 Coggshall, J. C., Boschek, C. B. and Buchner, S. M., Preliminary investigations on a pair of giant fibers in the central nervous system of Dipteran flies, Z. Naturforsch., 28C (1973) 783-784. 2 Ferrus, A., Koto, M. L. and Tanouye, M. A., Profiles of the giant fiber of Drosophila as revealed by lucifer yellow injection, Cal. Tech. BioL ann. Rep., (1980) 225. 3 Goodman, C. S., Bate. M. and Spitzer, N. C., Embryonic development of identified neurons: origin and transformation of the H cell, J. Neurosci., 1 (1981) 94-102. 4 King, D. G. and Wyman, R. J., Anatomy of the giant fiber pathway in Drosophila. I. Three thoracic components of the pathway, J. Neurocytol., 9 (1980) 753-770. 5 Levine, J. and Tracey, D., Structure and function of the giant motoneuron of D. melanogaster, J. comp. Physiol., 87 (1973) 213-235. 6 Power, M. E., The brain of Drosophila melanogaster, J. Morphol., 72 (1943) 517-559. 7 Power, M. E., The thoracico-abdominal nervous system of an adult insect, D. melanogaster, J. comp. Neurol., 88 (1948) 347-409. 8 Stewart, W. W., Functional connections between cells as revealed by dyecoupling with a highly fluorescent naphthalimide tracer, Cell, 14 (1978) 741-759. 9 Tanouye, M. A. and Wyman, R. J., Motor outputs of giant nerve fiber in Drosophila, J. Neurophysiol., 44 (1980) 405-421. 10 Thomas, J. B., Mutations affecting the giant fiber system of Drosophila, Neurosci. Abstr., 6 (1980) 742. 11 White, K. and Kankel, D. R., Patterns of cell division and cell movement in the formation of the imaginal nervous system in Drosophila melanogaster, Develop. Biol., 65 (1978) 296-321.
213
Elsevier/North-Holland Biomedical Press
Research Reports THE MORPHOLOGY
OF T H E C E R V I C A L G I A N T F I B E R N E U R O N OF
DR OSOPHILA
MARY KOTO, MARK A. TANOUYE, ALBERTO FERRUS, JOHN B. THOMAS and ROBERT J. WYMAN ( M.K., J.B.T. and R.J. W. ) Yale University, P.O. Box 6666, New Haven, Conn. 06511 and ( M.A.T. and A.F.) Division of Biology, California Institute of Technology, Pasadena, Calif. 91125 (U.S.A.)
(Accepted February 19th 1981) Key words: Drosophila melanogaster - - giant fiber - - Lucifer yellow
SUMMARY The morphology of the cervical giant fiber ( C G F ) neuron of Drosophila melanogaster was studied by intracellular injection of Lucifer yellow dye. The C G F neuron is the c o m m a n d cell in a motor circuit causing visually driven escape behavior: a single action potential in a C G F axon produces patterned activity in j u m p and flight muscles. The present study identified the C G F cell body, a large soma located in the posterior part of the lower ipsilateral protocerebrum. The main process runs anteriorly from the cell body, extends three branches, and turns posteromedially while descending through the brain. The C G F axon courses through the cervical connective and ends within the mesothoracic neuromere of the thoracic ganglion. Thus, the C G F neuron is an interneuron, not a motoneuron as previously believed. We have been isolating mutants that affect C G F neuron-mediated behavior. Comparison of C G F neuron morphology in wildtype strains with that in these mutants will allow identification of genes that affect the development, structure, and connections of the C G F neuron.
INTRODUCTION The cervical giant fiber ( C G F ) neurons in Drosophila melanogaster are a bilateral pair of nerve cells that extend from the brain to the thoracic ganglion. The C G F axons were first described by Power in 1948 as giant fibers 7. Though only 5 # m in diameter, the C G F axons are several times larger than any other axon in the cervical connective1, 7. The C G F neuron is a c o m m a n d cell in a fast-conducting motor pathway. A single C G F action potential elicited by intracellular stimulation drives the
214 jump and the flight muscles with a stereotyped pattern of spikes 9. A light-off stimulus elicits a spike in the CGF neuron, producing the same pattern of muscle activity 10. It thus appears that a visually induced escape response is mediated by the C G F neurons. In the present study, the CGF neuron was stained by intracellular injection of the fluorescent dye, Lucifer yellow CH 8. This method allowed visualization of the entire neuron, from its cell body in the brain to its terminus in the thoracic ganglion. The morphology of the CGF neuron is notable among invertebrates for its simplicity. Our results, together with King and Wyman's EM study 4 of the thoracic ganglion, show that the C G F neuron is an interneuron, not a motoneuron as previously believeda, 7. An abstract describing preliminary results appeared earlierL MATERIALS AND METHODS Experiments were performed on adult female Drosophila melanogaster. Flies were either a wildtype strain (Canton-Special), or white-eyed flies of the white, brown, scarlet, or cinnabar brown genotype. No difference in CGF neuron morphology within the thoracic ganglion was found between the wildtype (n = 5) and the 3 white-eyed strains (n = 6). The brain tissue of wildtype flies has a natural fluorescence that obscures visualization of the stained C G F neuron in brain wholemounts. Removal of the eyes before fixation reduced, but did not eliminate the fluorescence. Because whiteeyed flies had a minimal amount of fluorescence, they were used to view the brain in wholemount. Preliminary results comparing brain sections of wildtype (n = 2) and white-eyed (n = 2) flies showed no significant differences in the structure of the C G F neuron. Dissection, recording techniques, and verification of CGF axon impalement in the cervical connective are described by Tanouye and Wyman 9. Micropipettes, in which tips had been filled with 3 ~ Lucifer yellow, were backfilled with 0.05 ~,, LiCI. Lucifer yellow was iontophoretically injected into a CGF axon for 5-30 min with hyperpolarizing pulses or DC current of4-15 nA. Immediately after dye injection, the flies were fixed in buffered formalin for at least 30 min before dissecting the brain and ganglion out of the body. Wholemounts were prepared by dehydrating tissue in an ascending alcohol series and clearing in methyl benzoate. Sections were prepared by embedding brains in Spurr's medium and sectioning at 4/~m. All preparations were viewed under a Zeiss microscope with fluorescence optics as described by Stewart s. RESULTS AND DISCUSSION The cell body of the C G F neuron is very large, about 20/zm in diameter (n = 9). Ipsilateral to its axon, the cell body is located close to the posterior border of the brain adjacent to the neuropile of the protocerebrum about 30 #m dorsal to the oesophageal canal. (Fig. 1A and B). A neurite about 60/~m long extends anteriorly, connecting the cell body with the main process. The main process then descends posteromedially, curves around the oesophageal canal, and exits in the cervical connective. Prior to its posteromedial descent, the main process has 3 branches - - two ipsilateral and one
215
Fig. 1. A: wholemount of the brain. The anterior part of the brain is at the top of the picture. The CGF cell body is in the posterior part of the brain. The large ipsilateral branch that ends in the antennal glomerulus runs anteriorly. The contralateral branch and an unidentified contralateral cell body are faintly stained. The CGF axon turns medially, descends ventrally, and exits the brain in the cervical connective. B : horizontal section of the brain with CGF cell bodies. One CGF cell body is stained (right arrow) while its contralateral homologue (left arrow) is not. The small ipsilateral branch is in the central part of the brain. The section is 36 /~m dorsal to the oesophageal canal. C: wholemount of the thoracic ganglion. The tuft region is 57 #m from the axon terminus. In this preparation, two separate CGF axon penetrations were made. One lasted only 2 rain and is most likely responsible for the faint filling of the contralateral homologue. All space bars = 100 #m. contralateral. The larger ipsilateral branch, stained in 8 preparations, runs ventroanteriorly and ends in the protocerebral lobe close to the anterior edge o f the brain. This b r a n c h sends out n u m e r o u s short, spiny processes, especially near its terminus. In 4 preparations, a second ipsilateral branch was visible. Shorter and thinner than the first, this b r a n c h extends dorsomedially f r o m the main axon. The contralateral branch, faintly stained in the 7 preparations where it was seen, crosses the midline above the oesophageal canal and ends in the posterior part o f the protocerebrum. The contralateral b r a n c h in one preparation ran to the cell b o d y o f the contralateral C G F neuron which also stained. The staining o f the C G F neuron's contralateral h o m o l o g u e p r o b a b l y results f r o m dye-coupling between the two neurons. Since dye-coupling occurs between cells that are electrically coupled a, the C G F neurons m a y be connected by an electrical synapse. In 3 other preparations, the contralateral b r a n c h ended in small dots, processes, or an unidentified cell b o d y in the posterior part o f the brain. In all ganglion preparations (n = l l ) , the structure o f the thoracic C G F processes was simple and unvarying (Fig. 1C). Entering the ganglion along the dorsal midline, the C G F axon gradually descends ventrally. The C G F axon turns ventrolaterally within the mesothoracic neuromere. Where the axon begins to turn, it sends out a dorsal spray o f short, fine processes that resembles a tuft. The C G F axon ends
216 about 80/~m from the lateral edge of the ganglion without branching. Only one of the pair of C G F axons stained in the thoracic ganglion when a single penetration of a C G F axon was made. Thus, the C G F axons are not dye-coupled within the thoracic ganglion. In the present study, the C G F cell body was located close to the neuropile in the posterior part of the brain just above the oesophageal canal. In a [3H]thymidine labeling experiment with postembryonic DrosophilamelanogasterWhite and Kanke111 found that the neurons with cell bodies closest to the neuropile had the earliest birthdays. Therefore, the close proximity of the cell body to the neuropile probably indicates that the C G F neuron underwent its final division early in development. The C G F cell bodies, conspicuous for their large size, were described earlier by Power, but he did not associate them with the giant fibers~. Our results conflict with those of Levine and Tracey 5. Looking at wholemounts prepared by backfilling the entire cervical connective with cobalt dye, Levine and Tracey believed that the CGF cell body was in the anterior portion of the brain. The structure that they term the C G F cell body could be that of any cervical fiber. The branches of the CGF neuron in the brain send out short, fine processes in several regions, including the protocerebral lobe, the central commissure, and the central body. It is likely that the C G F neuron makes synaptic contacts via the fine processes in these areas. Fibers in these areas include interneurons from the eyes and antennae as well as from association centers in the brain 6. In the thoracic ganglion, our results are consistent with King and Wyman's finding that the CGF neuron is an interneuron 4. In their serial section micrographs, the C G F axon ended within the mesothoracic neuromere; in the present study, the large, brightly stained CGF axon always ended abruptly within the mesothoracic neuromere. In previous reports the C G F neuron was described with two extra processes in the thoracic ganglionS, 7. Processes other than the C G F axon stained faintly in a number of our thoracic ganglion preparations (n = 5). Since all processes had cell bodies associated with them, it is most likely that they are separate neurons dyecoupled to the CGF neuron. When the antennal nerve was backfilled with cobalt dye, Strausfeld and his associates also found cells in the thoracic ganglion dye-coupled to the CGF neuron (personal communication). In the present study, two cells were frequently dye-coupled to the CGF neuron. One runs from the terminus of the C G F neuron to the lateral edge of the ganglion near the ipsilateral posterior dorsal mesothoracic nerve. King and Wyman 4 found that the large motoneuron to the jump muscle (the tergotrochanteral muscle) contacts the terminus of the C G F neuron and exits the ipsilateral posterior dorsal mesothoracic nerve (PDM nerve). The other cell found in the present study crosses from the tuft region of the CGF neuron to the contralateral side of the ganglion, where it exits in the PDM nerve. King and Wyman 4 also found an interneuron (the peripherally synapsing interneuron) that contacts the CGF neuron near the midline, exits the ganglion in the contralateral PDM nerve, and synapses in the peripheral nerve with the 5 motoneurons innervating the dorsal longitudinal flight muscle. It is likely that the cells stained in the present study - - the
217 cell that abuts the C G F neuron and the contralateral cell - - are, respectively, the tergotrochanter motoneuron and the PSI found by King and Wyman 4. Both the dyecoupling of these cells to the C G F neuron and physiological evidence 9 suggest that the C G F neuron makes electrical synapses with the tergotrochanter motoneuron and the PSI. Our results can be applied to developmental studies of the C G F neuron and its connections. The large size of the C G F cell body readily allows identification of the neuron, even in unstained preparations (Fig. 1B). I f the C G F cell body attains its large size early in development, its cell body can be injected with Lucifer yellow and outgrowth of its processes traced a. Mutations which affect the structure and function of the C G F pathway could be used to analyze the events in C G F netiron development. As a first step in such an analysis, comparison of wildtype and mutant flies would reveal defects in C G F neuron morphology. ACKNOWLEDGEMENTS Development of the Lucifer yellow injection technique in Drosophila melanogaster and the initial results were obtained in the laboratory of Dr. Seymour Benzer. The authors wish to thank him for the use of his facilities and his support. The authors also thank Walter Stewart for his gifts of Lucifer yellow. Research was supported by NSF PCM-7911771 to Seymour Benzer, N S F predoctoral fellowship to M.K., Muscular Dystrophy Association postdoctoral fellowship to M.A.T., Gosney Fund fellowhip to A.F., N I H training grant to J.B.T., and N I H NS-07314 and NS14887 to R.J.W.
REFERENCES 1 Coggshall, J. C., Boschek, C. B. and Buchner, S. M., Preliminary investigations on a pair of giant fibers in the central nervous system of Dipteran flies, Z. Naturforsch., 28C (1973) 783-784. 2 Ferrus, A., Koto, M. L. and Tanouye, M. A., Profiles of the giant fiber of Drosophila as revealed by lucifer yellow injection, Cal. Tech. BioL ann. Rep., (1980) 225. 3 Goodman, C. S., Bate. M. and Spitzer, N. C., Embryonic development of identified neurons: origin and transformation of the H cell, J. Neurosci., 1 (1981) 94-102. 4 King, D. G. and Wyman, R. J., Anatomy of the giant fiber pathway in Drosophila. I. Three thoracic components of the pathway, J. Neurocytol., 9 (1980) 753-770. 5 Levine, J. and Tracey, D., Structure and function of the giant motoneuron of D. melanogaster, J. comp. Physiol., 87 (1973) 213-235. 6 Power, M. E., The brain of Drosophila melanogaster, J. Morphol., 72 (1943) 517-559. 7 Power, M. E., The thoracico-abdominal nervous system of an adult insect, D. melanogaster, J. comp. Neurol., 88 (1948) 347-409. 8 Stewart, W. W., Functional connections between cells as revealed by dyecoupling with a highly fluorescent naphthalimide tracer, Cell, 14 (1978) 741-759. 9 Tanouye, M. A. and Wyman, R. J., Motor outputs of giant nerve fiber in Drosophila, J. Neurophysiol., 44 (1980) 405-421. 10 Thomas, J. B., Mutations affecting the giant fiber system of Drosophila, Neurosci. Abstr., 6 (1980) 742. 11 White, K. and Kankel, D. R., Patterns of cell division and cell movement in the formation of the imaginal nervous system in Drosophila melanogaster, Develop. Biol., 65 (1978) 296-321.