Neuronal cell death in grasshopper embryos: variable patterns in different species, clutches, and clones

Transcription

1 J. Embryol. exp. Morph. 78, (1983) Printed in Great Britain The Company of Biologists Limited 1983 Neuronal cell death in grasshopper embryos: variable patterns in different species, clutches, and clones By CURTIS M. LOER 1-3, JOHN D. STEEVES 2 AND COREY S. GOODMAN 1 From the Department of Biological Sciences, Stanford University SUMMARY Previous studies showed that cell death plays an important role in adjusting the segmentspecific number of ganglionic neurones during grasshopper embryogenesis (Bate, Goodman & Spitzer, 1979; Goodman & Bate, 1981). In every segment, the single midline precursor 3 (MP3) divides once to produce two progeny. In some segments, one or both of these two progeny die; there is a general pattern of cell death of the MP3 progeny across the thoracic and abdominal segments. In the present study we examined the pattern of cell survival versus death of the MP3 progeny in 472 embryos from four different species, from the genetically related offspring within different clutches of the same species and from the genetically identical offspring within isogenic clones of the same species. We find variability in the pattern of cell survival versus death amongst embryos of the same species, clutch and clone, suggesting a significant epigenetic influence on this pattern. However, our results also show significant differences in the pattern of cell death between different genera and species, and between different clones and clutches within a single species, suggesting a genetic influence on this pattern as well. INTRODUCTION Each segmental ganglion in a grasshopper's metameric nervous system contains a highly specific pattern of neurones. From segment to segment, this pattern varies; each segment's complement of neurones is tailored to that segment's particular needs. One of the most striking differences between segments is the number of neurones; thoracic ganglia contain about 2000 neurones whereas abdominal ganglia contain about 500 neurones. Each different segmental pattern is produced, however, from a common segmentally repeated set of precursor cells (Bate, 1976; Bate & Grunewald, 1981; Goodman, Bate & Spitzer, 1981; Goodman & Bate, 1981). The differences in cell number arise in two ways: 1 Authors' address: Department of Biological Sciences, Stanford University, Stanford, CA 94305, U.S.A. 2 Author's address: Department of Zoology, University of British Columbia, Vancouver, BC V6T 2A9, Canada. 3 Author's present address: Biology Department B-022, University of California, San Diego, La Jolla, CA 92093, U.S.A.

2 170 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN differential production of cells by the neuronal precursors and differential death of the cells produced. Cell death appears to play the more important role in adjusting cell number; segmental differences are sculpted from a common block rather than constructed differently from the outset (Bate, Goodman & Spitzer, 1981; Goodman & Bate, 1981; Bate & Goodman, in preparation). This is illustrated by the progeny of one of the identified neuronal precursor cells, the median neuroblast (MNB). In the metathoracic (T3) segment, the MNB produces about 100 progeny, most of which survive. In the first abdominal segment, the MNB produces about 90 progeny, only 45 of which survive; thus cell death accounts for most of the segment-specific difference in cell number from this NB (Goodman & Bate, 1981). At specific stages of embryogenesis, there is massive cell death in the abdominal segments while there is relatively little death in the thoracic segments. Interestingly, many cells destined to die begin their morphological differentiation before they die (Bate & Goodman, in preparation). Whether cells in the oo ooooo ooooo ooooo ooooo oooo OOOO ( MNB NBS o oooo oooo ) OOO * OOOO <1 OOOO ooo Fig. 1. Schematic diagram of the grasshopper embryo at about 30 %, indicating the brain (B), the three suboesophageal segments (SI S3), the three thoracic segments (T1-T3), and the eleven abdominal segments (Al-All). On the right is a schematic diagram of the pattern and identification of neuronal precursor cells in each segment. There are two types of precursors: neuroblasts (NBs) and midline precursors (MPs) (Bate, 1976; Bate & Grunewald, 1981). Each segment contains two plates of 30 NBs each, a median neuroblast (MNB), and 7 MPs (MP1, MP2i, MP2 r, MP3, MP4, MP5, and MP6). In this paper we examine the two progeny of midline precursor 3 (MP3) whose position in the map is indicated by the arrow.

3 Neuronal cell death in grasshopper embryos 171 Fig. 2. Morphology of the 'H' cell and 'H cell sib' in the mesothoracic (T2) segment of a 60 % grasshopper embryo, based on a camera-lucida drawing of cellsfilledwith the fluorescent dye Lucifer yellow. Anterior is up. abdominal segments have a programmed commitment to die, or are directed to die by the segmental environment they find themselves in, remains an open question (e.g. Whitington et al 1982). Many neurones destined to die can be individually identified. For example, within the repeated pattern of precursor cells, every segment has a single midline precursor 3 (MP3) which divides once to give rise to two progeny (Fig. 1; Goodman, Bate & Spitzer, 1981). In the meso- (T2) and metathoracic (T3) segments, one of these two progeny differentiates into the distinctive 'H' cell; its sibling has a single anterior axon (Fig. 2). In some of the abdominal segments, one or both of these two cells dies (Bate, Goodman & Spitzer, 1981). These two neurones are ideal for studying cell death because they are highly accessible and easily recognized from birth simply on the basis of their cell body location. Previous

4 172 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN examination of the two MP3 progeny described segmental differences in their survival versus death (Bate, Goodman & Spitzer, 1981). Although there is a trend in the segment-specific differences of survival versus death, the precise pattern varies from embryo to embryo. In the present study we have been interested in three questions concerning the cell death of the MP3 progeny. First, how much variability in the pattern of cell death exists from embryo to embryo of the same species? Second, are there significant differences in the pattern of cell death between different species and genera? Third, to what degree is the variable pattern of cell death in animals within a species due to genetic versus non-genetic influences? To answer these questions, we counted the number of MP3 progeny present (either 0,1, or 2) in each segment after cell death occurs. We examined 472 embryos from four different species {Schistocerca americana, S. nitens, S. gregaria, and Melanoplus differ entialis). Furthermore, we examined the variable pattern of cell death in different clutches of the same species, and in different isogenic clones of the same species. Clutches are the offspring of single mated females; clones are the parthenogenetic offspring of single unmated females (Goodman, 1977, 1978; Steeves & Pearson, 1983). Thus, animals within a clutch are genetically related, whereas animals within an isogenic clone are genetically identical. In this paper we show significant differences in the pattern of cell death between different genera and species, and between different genetically related clutches and isogenic clones within a single species. Some of these results have previously been reported (Loer & Goodman, 1981). MATERIALS AND METHODS Four different species were used: Schistocerca americana, S. nitens, S. gregaria, and Melanoplus differential. S. americana and S. nitens were obtained from a laboratory colony at Stanford University, S. gregaria was obtained from University of British Columbia, and M. differential was obtained from Zoecon Corp. Although we do not know the precise number of generations that each colony has been bred in the laboratory, the S. nitens and S. gregaria colonies are quite old (well over 50 generations in the laboratory) whereas the S. americana colony is relatively young (about five generations in the laboratory at the time these experiments began). We report on 151 embryos of 5. americana, including 25 embryos each from four different clutches (S.a. 4, 5, 6, and 7); 125 embryos of S. nitens, including at least 20 embryos each from four different clutches (S.n. 1, 4, 5, and 6); 102 embryos of M. differential; and 94 embryos of S. gregaria from four different isogenic clones (7, 8,10, and 21). The clones were produced parthenogenetically at the University of British Columbia according to the method of Goodman (1977,1978). Embryos were examined at 55 %-65 % of development (Bentley, Keshishian, Shankland & Toroian-Raymond, 1979) after cell death of the MP3 progeny

5 Neuronal cell death in grasshopper embryos 173 I Fig. 3. Photograph of the dorsal surface of the mesothoracic (T2) segment in a 45 % grasshopper embryo showing the characteristic position and identification of the two MP3 progeny. Note that their cell bodies (arrows) are framed by the two anterior commissures (ac and be) and the posterior commissure (cc) and the longitudinal axonal pathways. The MP3 progeny have a distinctive size and appearance, and no other neuronal cell bodies lie near them on the dorsal surface within this frame of axonal pathways. These cell bodies are normally quite clearly visible and easy to identify in the living preparation with Nomarski optics. However, in the fixed preparation shown in thisfigure,we have enhanced their contrast for the photograph by application of a monoclonal antibody, 2C4, and an HRP-labelled second antibody (kindly provided by Kathryn Kotrla). Anterior is up. normally occurs. The cell bodies of the MP3 progeny were viewed in living embryos with Nomarski interference contrast optics (Goodman & Spitzer, 1979; Goodman, O'Shea, McCaman & Spitzer, 1979). Some cells were injected with

6 174 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN the fluorescent dye Lucifer Yellow to examine their morphology. The cell bodies of the two MP3 progeny are easily recognized after a minimal dissection. Their cell bodies are located on the dorsal surface of the developing ganglion and are framed by the anterior and posterior commissures and the longitudinal axonal pathways (Fig. 3). The MP3 progeny have a distinctive size and appearance, and no other neuronal cell bodies lie near them on the dorsal surface within this frame of axonal pathways. The MP3 progeny were examined in 14 contiguous segments: S3 (suboesophageal segment 3), T1-T3 (pro-, meso-, and metathoracic segments), and A1-A10 (abdominal segments 1-10). Altogether, the embryonic nervous system derives from 17 segmental ganglia (S1-S3, T1-T3, Al-All) plus a brain of unknown segmental origin. Each segment was scored as containing 0, 1, or 2 MP3 progeny. In the cases in which we see less than two MP3 progeny, we can rule out the possibility that MP3 has failed to divide or that the progeny have failed to migrate to the dorsal surface because, prior to the period of cell death, we always see two cells in this characteristic position in each segment (e.g. Fig. 3). Furthermore, after the period of cell death, we occasionally see the remains of one or two dead cells in the appropriate location. The S3, A8, A9, and A10 segments were often difficult to score because of the presence of other cells or tissues obscuring identification of the MP3 progeny. Although a segment was not scored unless we were quite certain about the presence or absence (and identification) of the MP3 progeny, the samples from S3, A8, A9, and A10 are nonetheless the least reliable. We have no doubt that our scoring of the other segments (T1-T3, A1-A7) is absolutely reliable. RESULTS Segmental pattern of MP3 progeny The typical pattern of survival of the MP3 progeny is one cell in S3, two cells in Tl-Al, one cell in A2-A7, and two cells in A8 (in A9 and A10 we are less confident about the 'typical' pattern). As shown in Fig. 4, the pattern is however quite variable. For example, in some embryos, only one cell survives in T2, or two cells survive in A2. Thus, although there is a general trend, the exact pattern of MP3 progeny survival is not constant from embryo to embryo. There is also a typical pattern of morphological differentiation of the MP3 progeny (Fig. 5), although this pattern too is not absolute. In T1-T3, the 'H' cell acquires its complete 'H' morphology (Figs 2, 5), while the 'H cell sib' acquires its characteristic morphology with a single axon extending anteriorly. In S3, the 'H' cell acquires a 'half-h' morphology with axons extending only anteriorly (Fig. 5). In Al and A8, the 'H' cell typically acquires a 'half-h' morphology, in this case, however, with axons extending only posteriorly. In A2-A7, the surviving cell has the 'H cell sib' morphology.

7 Neuronal cell death in grasshopper embryos 175 Schistocerca nitens not graphed n<5 0 No. of MP3 progeny surviving Melanoplus differentialis n = Schistocerca americana 100 n = n = Fig. 4. Segmental pattern of MP3 progeny survival versus death in three different species: 5. americana; Melanoplus differentialis; S. nitens. The 'n' indicates the number of embryos in which a given segment was scored. The key in the upper left applies to Figs 4, 6, 7, and 8 and indicates the number of MP3 progeny surviving: either 0,1, or 2. This pattern, like that of cell survival, is not absolute. For example, in segment A2, when two cells survive, the 'H' cell often (5/10) acquires the morphology typical for the 'H' cell in the Al segment. Only rarely does a second cell survive in A3 or A4 and acquire this morphology. Furthermore, the 'half-h' morphology typical of A8 is sometimes replaced by a complete 'H' morphology with axons extending anteriorly as well as posteriorly. Species patterns and differences The segmental patterns of cell survival (Fig. 4) and morphological differentiation of the MP3 progeny are quite similar in the three species examined: S. americana (n = 151), S. nitens (n = 125), and M. differentialis (n = 102). However, there are statistically significant differences in the death versus survival of the MP3 progeny amongst these three species (Table 1; each segment was compared using chi square analysis). The two species from the same genus (5. americana and S. nitens) are more similar than when compared to the unrelated species (M. differentialis).

8 176 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN S3 T1-T3 Al A2-A7 A8 Fig. 5. Schematic diagram showing the typical morphology of the 'H' cell and 'H cell sib' in 14 segments in the grasshopper embryo after the period of cell death and morphological differentiation. This pattern, like that of cell survival, is variable (see text). Anterior is up. Table 1. Statistically significant differences in the cell death of the MP3 progeny in different species of grasshopper segment Comparison S3 Tl T2 T3 Al A2 A3 A4 A5 A6 A7 A8 A9 A10 S.a. versus S.n. - ** - S.a. versus M.d. - S.n. versus M.d. - ** - - S.a. = Schistocerca americana. S.n. = S. nitens. M.d. = Melanoplus differentialis. * = P<O01. * = p< * * ** * - ** - - * * * ** * Few differences are detected in the thoracic segments, where the full 'H' morphology is expressed by the H cell. The significant differences tend to occur in the abdominal segments. In general, in these segments there was less survival in M. differentialis than in S. americana (Table 1). One of the more striking

9 Neuronal cell death in grasshopper embryos 111 differences between the species is seen in segment A2: in S. americana, 42 % of the embryos had two cells surviving, while in S. nitens only 7 % and in M. differentialis only 9 % of the embryos had two cells surviving. Table 2. Statistically significant differences in the cell death of the MP3 progeny in different clutches of the species S. americana Comparison clutches 4 versus 5 clutches 4 versus 6 clutches 4 versus 7 clutches 5 versus 6 clutches 5 versus 1 clutches 6 versus 7 S3 Tl segment T2 T3 Al A2 A3 A4 A5 A6 A7 A8 A9 A * * ** « * * * - - * = 005. S.a. clutch 4 S.a. clutch A S3 T A O " = S.a. clutch 6 S.a. clutch 7 n 37T S3 T1 n = A S3T1 2 3A Fig. 6. The pattern of survival versus cell death in four different clutches of 5. americana. Clutches are the offspring of single mated females. See legend of Fig. 4 for key to symbols.

10 178 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN S.n. clutch 4 n= n = n = n = Fig. 7. The pattern of survival versus cell death in four different clutches of S. nitens. Clutches are the offspring of single mated females. See legend of Fig. 4 for key to symbols. Table 3. Statistically significant differences in the cell death of the MP3 progeny in different clutches of the species S. nitens segment Comparison clutches 1 versus 4 clutches 1 versus 5 clutches 1 versus 6 clutches 4 versus 5 clutches 4 versus 6 clutches 5 versus 6 * = P<0-01. = />< S3 Tl T2 T3 Al A2 A3 A4 A5 A6 A7 A8 A9 A * _ _ * * _ * _ * _ ** _ Clutch patterns and differences We examined the pattern of cell survival of the MP3 progeny in four clutches each in S. americana (Fig. 6, Table 2) and 5. nitens (Fig. 7, Table 3). Clutches are the offspring from single mated females. Variable patterns of cell survival

11 Neuronal cell death in grasshopper embryos 179 were observed within the offspring of a single clutch. Furthermore, statistically significant differences were detected between clutches of the same species. One of the more striking examples of differences between clutches is seen in segment A2: in clutch 6 only 4 % of the embryos had two cells surviving, while in clutch 5 there were two cells surviving in 80 % of the embryos. It is interesting that more differences were detected between the clutches of S. americana than S. nitens (Tables 2 versus 3). One possible explanation is based on the age of these colonies: at the time of these assays, the S. americana colony had only been in the laboratory about five generations, whereas the S. nitens colony had been in the laboratory for over 50 generations. It seems quite possible that the inbreeding and/or laboratory selection resulted in greater homogeneity in the S. nitens colony and thus fewer differences between the S. nitens clutches. Clone patterns and differences We examined the pattern of cell survival of the MP3 progeny in four isogenic clones of S. gregaria (Fig. 8, Table 4). The isogenic clones are the parthenogenetic offspring of single unmated females. Despite the fact that the offspring within a clone are genetically identical, we observed variable patterns of cell S.g. clone 7 S.g. clone 8 n n = S 3 T A = : Fig. 8. The pattern of survival versus cell death in four different clones of 5. gregaria. Isogenic clones are the parthenogenetic offspring of single unmated females. See legend of Fig. 4 for key to symbols.

12 180 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN Table 4. Statistically significant differences in the cell death of the MP3 progeny in different isogenic clones of the species S. gregaria segment Comparison S3 Tl T2 T3 Al A2 A3 A4 A5 A6 A7 A8 A9 A10 clones 8 versus 10 _ * clones 8 versus 21 ** ** clones 8 versus 7 _ ** clones 10 versus 21 - ** clones 10 versus 7 _ _ - clones 7 versus 21 _ ** * = P<0-01. = />< survival within the clones. At the same time, however, we observed statistically significant differences between different clones (Table 4). For example, clone 7 had a high survival rate in the A2 segment, whereas clone 8 had an unusually high survival rate in A3 and A4. In regard to the small number of significant differences observed between the difference clones, it may be pertinent to note that the colony of S. gregaria had been inbred in the laboratory for over 80 generations prior to the production of these clones. The S. nitens colony, on the other hand, has been inbred for significantly fewer generations and, interestingly, shows a larger number of significant differences between different clutches. DISCUSSION Bate, Goodman & Spitzer (1981) showed a segmental pattern of cell survival versus death for the two MP3 progeny in segments T2-A6. The results presented in this paper expand the knowledge of this pattern to other segments (S3-A10), and in particular show that within this general pattern there is considerable variability from embryo to embryo. The statistically significant differences between different clutches and clones show that there is a genetic influence on the probability of survival versus death of these two cells. However, the striking variability within individual clutches and clones shows that there is also a significant epigenetic influence on the death of these neurones. A previous study (Goodman, 1977) using isogenic clones of grasshoppers showed that duplications and deletions of identified neurones can occur with a high degree of genetic control and specificity. The paper speculated: "The specificity of duplications and deletions could result from either selective cell division or selective cell death." The results from our present investigation show how genetic variability in the selective death of identified neurones can lead to such differences in cell number.

13 Neuronal cell death in grasshopper embryos 181 We have shown significant differences in cell death between different genera and species, and significant differences between the genetically related offspring of different clutches, and between the genetically identical offspring of different clones. Although variable patterns within clutches and clones indicate that the epigenetic influence is great, nevertheless we have demonstrated that certain differences in the pattern of survival versus death are heritable. It is interesting to speculate that these differences in the number of identified neurones between different clutches and clones may be the raw material on which natural selection acts to produce the different patterns observed in different populations and species. An interesting question for the future remains: what causes the segmentspecific death of the MP3 progeny? Are these segmental differences the result of (i) segment-specific differences in the intrinsic program of the cells (e.g. studies on the nematode Caenorhabditis elegans; reviewed by Horvitz, Ellis & Sternberg, 1982); (ii) segment-specific differences in the amount or response to some diffusing hormone or factor (e.g. studies on the moth Manduca sexta; reviewed by Truman & Schwartz, 1982); or (iii) segment-specific differences in the cellular environment contacted by the growth cones of these cells before the period of cell death; or (iv) segment-specific differences in the growth cones of other cells that contact these cells before the period of cell death? These last two possibilities can be tested in the grasshopper embryo by cell ablations and other manipulations in different segments prior to the period of cell death. Whatever the mechanism underlying the death of the MP3 progeny, its effectiveness is clearly variable from embryo to embryo and its variability is under both genetic and epigenetic control. This study was the undergraduate honors thesis of C.M.L. in the Department of Biological Sciences, Stanford University. We thank Kathryn Kotrla and Bill Kristan for criticism of the manuscript. The study was supported by grants from the N.S.F. and McKnight Foundation to C.S.G., and from N.S.E.R.C. to J.D.S. REFERENCES BATE, C. M. (1976). Embryogenesis of an insect nervous system: I. A map of the thoracic and abdominal neuroblasts in Locusta migratoria. J. Embryol. exp. Morph. 35, BATE, C. M. & GRUNEWALD, E. B. (1981). Embryogenesis of an insect nervous system: II. A second class of neuron precursor cells and the origin of the intersegmental connective$. J. Embryol. exp. Morph. 61, BATE, C. M., GOODMAN, C. S. &SPITZER, N. C. (1981). Embryonic development of identified neurons: segmental differences of the H cell homologues. /. Neurosci. 1, BENTLEY, D., KESHISHIAN, H., SHANKLAND, M. & TOROIAN-RAYMOND, A. (1979). Quantitative staging of embryonic development of the grasshopper, Schistocerca nitens. J. Embryol. exp. Morph. 54, GOODMAN, C. S. (1977). Neuron duplications and deletions in locust clones and clutches. Science 197, GOODMAN, C. S. (1978). Isogenic grasshoppers: genetic variability in the morphology of identified neurons. /. comp. Neurol. 182,

14 182 C. M. LOER, J. D. STEEVES AND C. S. GOODMAN GOODMAN, C. S. & BATE, M. (1981). Neuronal development in the grasshopper. Trends in Neuroscience. 4, GOODMAN, C. S. & SPITZER, N. C. (1979). Embryonic development of identified neurones: differentiation from neuroblast to neurone. Nature 280, GOODMAN, C. S., BATE, C. M. & SPITZER, N. C. (1981). Embryonic development of identified neurons: origins and transformation of the H cell. /. Neurosci. 1, GOODMAN, C. S., O'SHEA, M., MCCAMAN, R. E. & SPITZER, N. C. (1979). Embryonic development of identified neurons: temporal pattern of morphological and biochemical differentiation. Science 204, HORVITZ, H. R., ELLIS, H. M. & STERNBERG, P. W. (1982). Programmed cell death in nematode development. Neurosci. Commentaries 1, LOER, C. M. & GOODMAN, C. S. (1981). Variability in the cell death of identified neurons in grasshopper embryos. Soc. Neurosci. 7, 294. STEEVES, J. D. & PEARSON, K. G. (1983). Variability in the structure of an identified interneurone in isogenic clones of locusts. /. exp. Biol. 103, TRUMAN, J. W. & SCHWARTZ, L. M. (1982). Insect systems for the study of programmed neuronal death. Neurosci. Commentaries 1, WHITINGTON, P., BATE, M., SEIFERT, E., RIDGE, K. & GOODMAN, C. S. (1982). Survival and differentiation of identified embryonic neurons in the absence of their target muscles. Science 215, {Accepted 20 July 1983)