The effect of heat shock (15 min at 48 °C) on segmentation has been investigated in the short germ embryo of the locust (Schistocerca gregaria). Prior to formation of the germ anlage and at the disc stage heat shock considerably reduced the survival of eggs but appeared to have little effect upon segmentation. At later stages heat shock had no effect on survival but resulted in disruptions of the segmental pattern. The location of abnormal segments depended upon the stage at heat shock and the number affected depended on its severity. A constant number of normal segments developed between the last segment visible at the time of heat shock and the first abnormal segment. These results are similar to the disruptions observed in amphibian somites following heat shock. However, different parts of the segment pattern varied in their response; the head segments were very rarely affected, and disrupted regions rarely started in the middle abdomen (segments A5 and A6).

The results are discussed in relation to two models (the clock and wavefront and progress zone models) that have been proposed as an explanation for the specification of the somite pattern in amphibians.

The body of an insect consists of a constant number of similar, but not identical, segmental units arranged in an anterior-to-posterior sequence. The mechanism by which this segmented pattern is specified during early embryonic development is not known.

At a descriptive level, insect embryos seem to form segments in two different ways (reviewed Sander, 1976, 1981). In short germ insects (e.g. the locust, Schistocerca gregaria) the embryo (or ‘germ anlage’) forms from a small region of the blastoderm cell layer and segments become visible in a reliable sequence during the gradual elongation of the posterior tip of the germ anlage. In contrast, the germ anlage in long germ insects (e.g. Drosophila, Smittia) forms from a large area of the blastoderm which is subdivided into segments. There are also intermediate germ insects (e.g. Eucelis, Acheta) in which anterior segments form as in a long germ embryo but the more posterior segments are added gradually as in a short germ embryo.

In long germ insects, the results of a wide variety of experiments (such as fragmentation and localized u.v.-irradiation) have suggested that the embryonic segments form with reference to levels of a gradient (or gradients) established between the anterior and posterior poles of the egg, prior to the formation of the germ anlage (see Herth & Sander, 1973; Schubiger & Wood, 1977; Kalthoff, 1983). In short germ insects the response to experimental manipulation at similar stages tends to be all-or-nothing; the embryo either develops normally or not at all (Miya & Kobayashi, 1974). This suggests that the mechanism of segmentation may differ from that in long germ embryos. The way in which segments appear during development suggests that cell proliferation at the posterior tip of the embryo may play an important role in the formation of the segment pattern (reviewed Sander, 1976).

Segmentation of insects appears to be rather similar to somitogenesis in anuran amphibian embryos. Somites appear in anterior-to-posterior sequence, the posterior somites developing as the tail bud extends. The normal pattern of somites is disrupted in particular ways following heat shock at specific stages during development. A short heat shock early in development causes small disruptions throughout the somite pattern. Heat shock given later (just prior to and during visible somitogenesis) results in a small disruption at a predictable position, posterior to the somites visible at the time of heat shock (Elsdale, Pearson & Whitehead, 1976; Cooke, 1978; Pearson & Elsdale, 1979; Elsdale & Pearson, 1979). Thus, as heat shock is given at progressively later stages, the disrupted area occurs in an increasingly posterior position, with a constant number of normal somites forming between the most posterior somite visible at the time of heat shock and the most anterior abnormal somite.

In the present experiments short germ locust embryos were given a short heat shock at different times prior to and during visible segmentation and the location and nature of the resulting segmental abnormalities were analysed (see Mee & French, 1986).

Populations of the locust, Schistocerca gregaria, were maintained at 31 ± 0·5 °C, with a 12h/12h light/dark cycle, on a diet of sprouted barley and bran. Mature adults were provided with honey pots filled with damp sand for oviposition. Eggs are laid in a cluster (a ‘pod’) of between 25 and 90 eggs deposited over a period of 1·5 to 3h. When eggs of a known age were required, ovipositing females were observed and pots removed when the female withdrew her abdomen from the sand. At this time the age of all eggs in the pod was designated Oh. Pods of eggs were removed from the sand, the eggs carefully separated, washed and kept on moist cotton wool in an incubator at 30 ± 0·5°C until required.

Embryos were examined after fixation. Eggs were punctured (in a region remote from the embryo), fixed in acetic acid: formalin: ethanol: water (1:6:16:30) for 3 h at 60°C (see Lawrence, 1973), washed and then stored in 70% ethanol. The chorion was removed by dissection, using tungsten needles and watchmakers’ forceps, or by treatment with 3 % sodium hypochlorite solution. In older eggs the serosal cuticle, serosa and amnion were removed also.

The period of development from germ anlage to germ band was staged on the basis of the morphological appearance of embryos. To determine the mean age at which each stage occurred, 22 pods of varying ages were sampled at 2 h intervals over a period of 12 or 24 h. The size of the sample taken from each pod (two to seven eggs) was dictated by the number of eggs per pod and the number of samples. The stage of each of the embryos within the sample (of known age) was then determined.

Eggs were heat shocked at intervals between egg deposition and the completion of visible segmentation. The eggs from one pod were treated together, as they develop fairly synchronously (Mee, 1984) while different pods develop at rather different rates (Tyrer, 1970; Bentley, Keshishian, Shankland & Toroian-Raymond, 1979). Prior to the formation of the germ anlage, eggs were heat shocked at a known age, subsequently at a recognizable morphological stage. A pod was staged by examining a sample of 10 or 15 eggs fixed immediately before experimentation, and eggs from the pod were assumed to be at the median stage observed in the sample. Eggs were heat shocked in water at 48°C for 15 min and then returned to a 30°C incubator until hatching. The consequences of altering the temperature or duration of heat shock were investigated and the effect of handling was examined by giving a sham ‘shock’ of 30 °C.

The effect of heat shock on the segmental pattern was determined in first instar hoppers, fixed shortly after hatching, usually after the development of pigmentation. Eggs that failed to hatch were fixed, the embryos dissected out and scored.

(A) Stages of development

The development of the germ anlage into the fully segmented germ band can be described by 14 morphological stages (see Fig. 1). During this period the segments of the gnathos (3), thorax (3) and abdomen (11) become visible as the posterior tip of the embryo elongates. The first four stages are the disc, heart shape (HS), elongating protocorm (EP) and segmented thorax (sT) (see also Shulov & Pener, 1963; Bentley et al. 1979) and the remaining stages are based on the sequential appearance of the eleven segments of the abdomen. These stages, named according to the most posterior visible segment, are the segmented abdominal segment 1 (sAl) stage, the segmented abdominal segment 2 (sA2) stage and so on to the fully segmented stage (sA10/11).

Fig. 1.

Camera lucida drawings of embryonic stages of S. gregaria. The embryo lies at the posterior pole of the egg. Egg and embryo are viewed ventrally to show the first four stages (disc, HS, EP and sT), two of the stages during segmentation of the abdomen (sA1and sA2) and the fully segmented stage. Segments (of the gnathos and thorax) first become visible in anterior regions of the embryo at the sT stage, and subsequently the eleven segments of the abdomen appear in anterior-to-posterior sequence. G1, G2 and G3 are the mandibular, maxillary and labial segments; T1, T2 and T3 the pro-, meso- and metathoracic segments and A1, A2 and A11 the 1st, 2nd and 11th abdominal segments.

Fig. 1.

Camera lucida drawings of embryonic stages of S. gregaria. The embryo lies at the posterior pole of the egg. Egg and embryo are viewed ventrally to show the first four stages (disc, HS, EP and sT), two of the stages during segmentation of the abdomen (sA1and sA2) and the fully segmented stage. Segments (of the gnathos and thorax) first become visible in anterior regions of the embryo at the sT stage, and subsequently the eleven segments of the abdomen appear in anterior-to-posterior sequence. G1, G2 and G3 are the mandibular, maxillary and labial segments; T1, T2 and T3 the pro-, meso- and metathoracic segments and A1, A2 and A11 the 1st, 2nd and 11th abdominal segments.

The fully segmented germ band is the presumptive ventral surface and appendages. The dorsal surface is derived from the flanks of the germ band which later extend dorsally and finally fuse at the dorsal midline (Sander, 1976).

The mean age at which each stage of development was reached is shown in Fig. 2. The disc stage germ anlage forms approximately 36 h after oviposition and segmentation is completed at approximately 73 h. Comparison of the interval between the mean ages for each stage shows that the duration of the first three stages (disc, heart shape and elongating protocorm) is longer than that of subsequent stages.

Fig. 2.

Mean age (with standard deviation) at each morphological stage (for abbreviations see text). The total number of embryos observed at each stage is given.

Fig. 2.

Mean age (with standard deviation) at each morphological stage (for abbreviations see text). The total number of embryos observed at each stage is given.

(B) The effects of heat shock

Following heat shock three categories of result were observed: (i) normal animals; (ii) animals with segmental defects; (iii) eggs that failed to develop (most were dead and discoloured, or had burst and extruded yolk).

The response, in terms of the proportion of animals falling into each of these categories varied with the stage of development at which heat shock was given and between different pods of eggs at a given stage.

The location of defects in the head was based on the appearance of the eyes and the appendages (labrum, antennae and mouthparts) as individual head segments are not visibly delineated. In the abdomen and thorax defects were located in the segments and their appendages. Abdominal segments A2 to A7 (and A8 in males) are similar and therefore the location of heat shock defects was determined by reference to unambiguously recognizable segments. This was a problem only when one or more of the similar segments were absent. If there was an additional defect it was assumed that the loss and the disruption had occurred in adjacent segments (e.g. the absence of one segment and an abnormality in the 2nd and 3rd abdominal segments present would be scored as disruption of abdominal segments 2,3 and 4). Abdominal segments A10 and All are fused in the hopper and were treated as a single segment (A10/11). Segments were scored without reference to the location of the disruption within the segment circumference or its appendages.

In an affected animal, heat shock usually resulted in the disruption of two or three consecutive segments. Two areas of disruption separated by one or more normal segments occurred rarely (in only 15 % of animals with abnormal segmentation resulting from heat shock prior to germ anlage formation, and 6 % of those resulting from heat shock given during the subsequent period).

(C) The response to heat shock before germ anlage formation

When comparing the results from heat- and sham-shocked eggs the variability between eggs from different pods was taken into account by using an Analysis of χ2 Test. The results from eggs, heat and sham shocked at 3h, showed that heat shock resulted in a significant increase in the proportion of eggs failing to develop, but had no significant effect on the frequency of surviving animals with segmental abnormalities.

The frequency of eggs failing to develop and of eggs developing into normal animals fluctuated with age at heat shock (Table 1) and the proportion of animals with abnormal segmentation was consistently less than 10 %. After heat shock at all stages, disruptions were observed in the head, thorax and abdomen, and their location within the segment file was apparently not related to age (Fig. 3).

Table 1.

The response of eggs following early heat shock Control

The response of eggs following early heat shock Control
The response of eggs following early heat shock Control
Fig. 3.

The location of segmental abnormalities following heat shock prior to the formation of the germ anlage, and following a sham shock (at 3h 5min). The kite diagrams show the frequency with which each segment was abnormal, calculated as a proportion (%) of the total number of abnormal segments in that age class. A frequency of 10 % is shown by the length of the scale bar. Segments are labelled, anterior to posterior, gnathos, G1 to G3; thorax, T1 to T3; abdomen, Al toA10/11. The other pattern elements scored for defects were the labrum (lb), eyes and antennae (ant).

Fig. 3.

The location of segmental abnormalities following heat shock prior to the formation of the germ anlage, and following a sham shock (at 3h 5min). The kite diagrams show the frequency with which each segment was abnormal, calculated as a proportion (%) of the total number of abnormal segments in that age class. A frequency of 10 % is shown by the length of the scale bar. Segments are labelled, anterior to posterior, gnathos, G1 to G3; thorax, T1 to T3; abdomen, Al toA10/11. The other pattern elements scored for defects were the labrum (lb), eyes and antennae (ant).

(D) The response to heat shock after germ anlage formation

(1) Sensitivity to heat shock

At all stages after germ anlage formation (except disc and sA7) heat shock resulted in an increase in the frequency of animals with segmental abnormalities (with respect to that of the sham-shocked controls) with a peak in the frequency of animals affected between the sT and sA2 stages. At disc and sA7 stages the frequency of segmental defects was low and similar to that of the control (see Table 2). The frequency of eggs failing to develop fluctuated following a sham or a heat shock between heart shape and sA7 stages (see Table 2), but at any given stage (apart from sA2 and sA3) values for sham and heat shock were not significantly different. At the earlier disc stage, there was a significant increase in the frequency of eggs failing to develop after a heat shock. At the disc stage, therefore, the response to heat shock was similar to that observed following early heat shock (failure to develop, no effect on segmentation) while at later stages it differed (little effect on viability, disrupted segmentation).

Table 2.

The results of (i) heat shock and (ii) control sham shock of eggs at stages between disc and sA7

The results of (i) heat shock and (ii) control sham shock of eggs at stages between disc and sA7
The results of (i) heat shock and (ii) control sham shock of eggs at stages between disc and sA7

(2) The segments affected by heat shock

At the disc stage the distribution of defects within the segment file resembled that observed following early heat shock, and all segments (apart from T3) were affected at similar low frequencies (Fig. 4).

Fig. 4.

The location of segments affected by heat shock at stages from disc to sA7. The data are presented as in Fig. 3. A frequency of 10 % is shown by the length of the scale bar. Frequencies of less than 1 % are represented by a Une. (Defects in which the only abnormality was the apparent loss of one (or more) of the middle abdominal segments could not be included -see Materials and Methods.)

Fig. 4.

The location of segments affected by heat shock at stages from disc to sA7. The data are presented as in Fig. 3. A frequency of 10 % is shown by the length of the scale bar. Frequencies of less than 1 % are represented by a Une. (Defects in which the only abnormality was the apparent loss of one (or more) of the middle abdominal segments could not be included -see Materials and Methods.)

In contrast, at the heart shape and subsequent stages heat shock induced defects that were located in the thorax and abdomen at relatively high frequencies, while the procephalon and gnathos were rarely disrupted. The location of the few segmental defects observed following sham shocks at heart shape and later stages (171 affected segments in 44 animals) showed a similar distribution.

The location of segmental defects depended upon the stage at which heat shock was given (Fig. 4); heat shock at progressively later stages affected increasingly posterior segments (whereas the location of the few defects occurring after a sham shock was not stage specific). A latency was apparent between the heat shock and the register of its effect upon the pattern. The first (i.e. most anterior) abnormal segment was the third or fourth to become visible (Fig. 5); two or three normal segments developing posterior to the last segment visible at the time of heat shock (Table 3). When measured as the time between age at heat shock and age at which the most anterior affected segment became visible (see Fig. 2), the latency was 12–9 h at early stages and 7 or 5h during appearance of the abdomen.

Table 3.

The latency in the response to heat shock in terms of the number of normal segments developed between the last segment visible at the time of heat shock and the first (most anterior) abdominal segment affected (see Fig. 5), and the interval of time between heat shock and the appearance of the first abnormal segment (see Fig. 2)

The latency in the response to heat shock in terms of the number of normal segments developed between the last segment visible at the time of heat shock and the first (most anterior) abdominal segment affected (see Fig. 5), and the interval of time between heat shock and the appearance of the first abnormal segment (see Fig. 2)
The latency in the response to heat shock in terms of the number of normal segments developed between the last segment visible at the time of heat shock and the first (most anterior) abdominal segment affected (see Fig. 5), and the interval of time between heat shock and the appearance of the first abnormal segment (see Fig. 2)
Fig. 5.

Identity of the first (most anterior) segment affected by heat shock at stages from heart shape to sA6. Kite diagrams show the frequency with which a given segment was the most anterior in a series of affected segments as a proportion (%) of the total number of animals with segmental defects (defects in which the only abnormality was the loss of one or two abdominal segments were not included -the number of such defects is given in brackets). The mean number of segments/disruption (with standard deviation) is also shown (defects in which the only abnormality was the loss of one or two abdominal segments were scored as affecting 1 and 2 segments respectively). Data were not plotted for disc and sA7 stages as at the former the pattern of response was unlike that observed for the heart shape and subsequent stages (see text) and at the latter the frequency of animals with abnormal segments was low, suggesting that the majority of eggs from pods staged sA7 had reached a stage at which heat shock no longer had any effect.

Fig. 5.

Identity of the first (most anterior) segment affected by heat shock at stages from heart shape to sA6. Kite diagrams show the frequency with which a given segment was the most anterior in a series of affected segments as a proportion (%) of the total number of animals with segmental defects (defects in which the only abnormality was the loss of one or two abdominal segments were not included -the number of such defects is given in brackets). The mean number of segments/disruption (with standard deviation) is also shown (defects in which the only abnormality was the loss of one or two abdominal segments were scored as affecting 1 and 2 segments respectively). Data were not plotted for disc and sA7 stages as at the former the pattern of response was unlike that observed for the heart shape and subsequent stages (see text) and at the latter the frequency of animals with abnormal segments was low, suggesting that the majority of eggs from pods staged sA7 had reached a stage at which heat shock no longer had any effect.

A discontinuity was observed in the distribution of defects within the abdomen. Although segments A5 and A6 were often affected by heat shock, they were rarely found as first abnormal segment in a disrupted sequence (Fig. 6). Since large numbers of eggs were heat shocked at appropriate stages (i.e. sAl, sA2 and sA3), the absence of disruptions beginning with segments A5 and A6 was not due to inadequate sampling. The effect of the discontinuity is clearly seen in the distribution of defects following heat shock at sA2 (Fig. 5).

Fig. 6.

The total number of times (i) each segment was affected by heat shock (total number of affected segments was 5373); (ii) each segment occurred as the most anterior in an affected region (total number of defects was 1998). Data taken from the heart shape to sA6 stage. Again disruptions in which the only abnormality was the loss of a complete segment(s) cannot be plotted.

Fig. 6.

The total number of times (i) each segment was affected by heat shock (total number of affected segments was 5373); (ii) each segment occurred as the most anterior in an affected region (total number of defects was 1998). Data taken from the heart shape to sA6 stage. Again disruptions in which the only abnormality was the loss of a complete segment(s) cannot be plotted.

(3) The extent of defects

In any given animal the extent of a heat shock defect was usually two or three segments (Fig. 5). However, at any given stage the range of segments affected at relatively high frequencies was from two or three (at stages sA3, sA4 and sA5, for example) to four, five or more (heart shape, sAl and sA2 stages -see Fig. 4). The interpretation of these results is difficult as the overall extent of disruptions was restricted; posteriorly, because A10/11 is the last segment of the animal; anteriorly, because gnathal segments were rarely affected. The absence of defects beginning with segments A5 and A6 also influenced the distribution.

At heart shape and sA2 stages it seems that the larger range of segments affected was due, at least in part, to variability in the location of the disrupted region within the segment file, as indicated by the identity of the most anterior abnormal segment (see Fig. 5). A greater number of segments would be affected by heat shock within a stage if that stage was of longer duration than the others (assuming the rate of segmentation is constant), as is the case at the heart shape and other early stages (but not the sA2 stage -see Fig. 2).

At the sA1 stage the location of the abnormal area was not particularly variable but the mean number of segments per disruption was slightly higher than at other stages (Fig. 5).

(4) The response to heat shock of different severities

Heat shocks of a different duration or temperature were delivered to embryos at the elongating protocorm stage. As the heat shock became more extreme (i.e. the temperature or duration of the heat shock increased) the frequency of normal animals declined and the proportion failing to develop increased (Table 4). All heat shocks raised the frequency of animals with abnormal segments above that of the sham-shocked control (Table 2). The heat shock utilized in the main experimental series (i.e. 48°C, 15min) gave the maximum frequency of animals with segmental abnormalities.

Table 4.

The effect of varying the severity of heat shock at the elongating protocorm stage

The effect of varying the severity of heat shock at the elongating protocorm stage
The effect of varying the severity of heat shock at the elongating protocorm stage

The severity of the heat shock had little effect on the identity of the most anterior segment affected (Fig. 7). The median position of the first abnormal segment was Al or A2, regardless of the temperature or duration of heat shock. However, as the heat shock became more extreme the mean number of segments per defect increased as successively more posterior segments were affected.

Fig. 7.

Identity of the first (most anterior) segment affected following heat shocks of varying severity delivered at the elongating protocorm stage of development. Data plotted as in Fig. 6. Each kite diagram shows the result following a heat shock of temperature and duration indicated. The mean number of segments per disruption is also shown.

Fig. 7.

Identity of the first (most anterior) segment affected following heat shocks of varying severity delivered at the elongating protocorm stage of development. Data plotted as in Fig. 6. Each kite diagram shows the result following a heat shock of temperature and duration indicated. The mean number of segments per disruption is also shown.

The effect of heat shock early in development

Prior to germ anlage formation and at the disc stage heat shock has a considerable effect on the survival of locust eggs. Eggs heat shocked at any time up to and including the disc stage show a mortality of 40–60 %, compared to around 20 % in control sham-shocked eggs and in eggs heat shocked at the later stages (Tables 1, 2). A similar result was observed in Drosophila following heat shock. Bergh & Arking (1984) found that only a small proportion (6 %) of eggs survived a heat shock given before blastoderm, but at gastrulation the frequency of eggs continuing to develop (85 %) was similar to that of control eggs (see also Graziosi et al. 1983).

Early heat shock, however, had no effect on the segmentation of locust embryos; the frequency of animals with segmental defects was low and not significantly different from that of the sham-shocked controls; the location of defects within the segment file was not influenced by age at heat shock (see Figs 3, 4, for 3h and disc stages, respectively). However, the segment pattern of non-developing embryos was not investigated and some of these may have had segmental abnormalities. Berg & Arking (1984) observed morphological abnormalities, following heat shock, in nonhatching Drosophila embryos but not in hatching larvae.

These results are rather different to those described for anuran amphibians following early heat shock (at stages during gastrulation, prior to somitogenesis). In amphibians, mortality and the degree to which somitogenesis was disrupted declined as heat shock was given at progressively later stages. The location of the disruptions was also found to depend upon stage; heat shock at earlier stages induced widespread somite abnormalities while at later stages they were restricted to more posterior regions (Elsdale & Pearson, 1979).

The effect of heat shock late in development

From the heart shape stage onwards, the standard heat shock of 15 min at 48°C had little effect on the survival of locust eggs but resulted in a high frequency of animals with segmental abnormalities. The location of segmental defects was stage specific and shifted posteriorly from thoracic to posterior abdominal segments as the heat shock was delivered at progressively later stages (Fig. 4). A fairly constant number of normal segments developed between the most posterior segment visible at the time of heat shock and the most anterior segment affected, regardless of the stage at which heat shock was delivered (Table 3) or the severity of the heat shock (Fig. 7). The first abnormal segment was the third or fourth to form after heat shock. A cellular sensitivity to heat shock therefore seems to precede visible segmentation. The number of segments affected ranged from one to four or more and was influenced by the severity of the heat shock; segments posterior to the disrupted region were normal.

The effects of heat shock on the locust during this period of development are similar in many ways to those described for amphibians heat shocked immediately prior to and during visible somitogenesis (Elsdale et al. 1976; Pearson & Elsdale, 1979).

Several aspects of the results show, however, that all parts of the locust segment pattern do not respond equally to heat shock.

  • (1) The frequency of heat-shock-induced abnormalities varied with stage, reaching high levels at stages between sT and sA2 (Fig. 4). Low frequencies at relatively early (e.g. heart shape) or late (e.g. sA6, sA7) stages may indicate that a proportion of the eggs in these pods was outside the sensitive period. This is not a complete explanation, however, since the frequency was also reduced at stages (e.g. EP and sA3) when the location of abnormalities was not at the anterior or posterior end of the segment pattern.

  • (2) The range of segments affected by heat shock was greater at some stages than at others and, as discussed above, this seems to reflect variability in location and in one case, an increase in the size of defects.

  • (3) Defects involving gnathal segments were rare (Fig. 6). It seems unlikely that this was due either to inadequate sampling (large numbers of eggs were shocked at disc, HS and EP stages) or to the fact that only the appendages of the gnathos were scored (since 84 % of thoracic disruptions involved the appendages). In short germ embryos the development of gnathal segments is believed to be similar to that of the thoracic and abdominal segments (for example, see Anderson, 1972; Sander, 1976), but the different response to heat shock suggests that this may not be the case.

  • (4) It is also clear that, although segments A5 and A6 are frequently affected by heat shock, areas of disruption do not often begin with these segments (Fig. 6), and this suggests that the abdominal segments may not all be formed in the same way.

Segmental abnormalities have also been described in the intermediate germ cricket, Acheta, following irradiation with X-rays (Heinig, 1967, reviewed Sander, 1976). Defects induced at the time of germ anlage formation affected thoracic and gnathal segments. At subsequent stages more anterior head segments were affected and finally the segments of the abdomen. Within the abdomen, which appears to form segments as in a short germ embryo, the location of disruptions followed a different pattern to that observed in the locust. Initially defects occurred throughout the abdomen but later they were restricted to increasingly posterior segments. This difference in the location of defects in the cricket and the locust may reflect different effects of heat shock and X-irradiation or a difference in the process of segmentation in the intermediate and short germ insects.

Segmental disruptions have been induced by a severe heat shock (4 h at 35 °C) to long germ Drosophila embryos but, in contrast to those in the locust, they were distributed throughout the segment file irrespective of the stage at which heat shock was given (Maas, 1949). The maximum frequencies of abnormalities occurred following heat shock at about the blastoderm stage (when the segment pattern is being formed) and then later, at about the time of dorsal closure.

Heat shock and models of segmentation

Pearson & Elsdale (1979) interpreted the effects of heat shock on amphibian development in terms of the ′Clock and Wavefront′ model (Cooke & Zeeman, 1976) in which somites are generated by the interaction of two components. These are a gradient, set up early in development across the anterior-posterior length of presumptive somite tissue, which specifies the time at which cells become competent to undergo segmentation (the wavefront), and a cellular oscillation with respect to which all cells are normally synchronized (the clock). During one short phase of oscillation all presumptive somite cells that have become competent will co-operate and eventually form a visible somite.

In amphibians heat shock results in the chaotic subdivision of a region of the presomite tissue. The disruption is most severe anteriorly, the pattern gradually becoming normal posteriorly. It is proposed that heat shock disturbs the synchrony of cellular clocks (Pearson & Elsdale, 1979). The boundary between anterior normal somites and the abnormal region would therefore locate the position of the wavefront of competence at the time of heat shock. The increasingly normal appearance of posterior somites was interpreted in terms of a recovery of synchrony within the unsegmented tissue; the severity of the disruption at any position reflecting the time available for recovery before competence was achieved (Pearson & Elsdale, 1979).

The clock and wavefront model was proposed for the formation of the somites of the trunk, which appear to form in a similar way to segments in long germ insects. However, the model can also be applied to somitogenesis in the elongating tail bud, where the disruptions are similar to those of the trunk.

Variations in the timing and severity of heat shock had similar effects on the location and extent of defects in anuran amphibians and locusts, but there are problems in interpreting the locust results in terms of the clock and wavefront model. The abnormality is not an irregular subdivision of the embryo (as in amphibians), but a deletion of parts of the segmental pattern (see Mee & French, 1986). The disrupted patterns may not be directly comparable since amphibian somite fissures were studied shortly after heat shock, while locust defects were examined in the cuticular pattern formed after the further development of the segments during embryogenesis. However, it is not clear how deletions could arise in the clock and wavefront model and, although a period of sensitivity to heat shock clearly precedes visible segmentation, the model cannot readily account for the results observed in the locust.

The ‘Progress Zone’ model for sequential pattern formation during growth was proposed for the specification of pattern elements in the developing chick limb bud (Summerbell, Lewis & Wolpert, 1973), and has since been suggested for somitogenesis in the amphibian tail (Cooke, 1975) and segmentation in the short germ insect embryo (Sander, 1976). It is postulated that a ‘progress zone’ of constant size is located at the tip of the chick limb bud and the positional value of the cells within the zone is labile, becoming increasingly distal with time. As cell division occurs within the progress zone, cells emerge from it proximally, retaining their current positional value, so that the pattern is laid down in a proximal-to-distal sequence. Following the X-irradiation of early limb buds proximal structures were missing, and more distal structures were lost as irradiation was carried out at successively later stages (Wolpert, Tickle & Sampford, 1979). It was suggested that X-irradiation caused cell death and so depleted the cell population of the progress zone. Consequently surviving cells remained within the zone for longer than normal and thereby acquired more distal positional values. This would result in the absence of cells possessing a particular range of positional values, according to the time of irradiation, and would lead to the formation of a limb lacking corresponding structures.

In a short germ insect embryo, a progress zone may occur at the posterior tip of the germ anlage, and cells emerging early in development would form thoracic segments, while those leaving late would form posterior abdominal segments. Heat shock could produce a stage-dependent loss of segmental structures if it resulted in cell death or merely delayed cell division, cells remaining in the progress zone and consequently acquiring a more distal positional value before they emerged.

The applicability of the progress zone model to segmentation in short germ insects will be reexamined in the light of results from fragmentation experiments (Mee, 1986).

We thank David Wright and Neil Toussaint for stimulating discussion and comments, Linda Partridge for help with the statistics and Bert Stewart for background noise. This work was funded by an SERC studentship to JEM.

Anderson
,
D. T.
(
1972
).
The development of Hemimetabolous Insects
.
In Developmental Systems: Insects
, vol.
1
(ed.
S. J.
Counce
&
C. H.
Waddington
), pp.
95
163
.
London and New York
:
Academic Press
.
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
,
47
74
.
Bergh
,
S.
&
Arking
,
R.
(
1984
).
Developmental profile of the heat shock response in early embryos of Drosophila
.
J. exp. Zool
.
231
,
379
391
.
Cooke
,
J.
(
1975
).
Control of somite number during development of a vertebrate, Xenopus laevis
.
Nature, LoncL
254
,
196
199
.
Cooke
,
J.
(
1978
).
Somite abnormalities caused by short heat shocks to pre-neurula stages of Xenopus laevis
.
J. Embryol. exp. Morph
.
45
,
283
294
.
Cooke
,
J.
&
Zeeman
,
E. C.
(
1976
).
A clock and wavefront model for control of the number of repeated structures during animal morphogenesis
.
J. theor. Biol
.
58
,
445
476
.
Elsdale
,
T.
,
Pearson
,
M.
&
Whitehead
,
M.
(
1976
).
Abnormalities in somite segmentation following heat shock to Xenopus embryos
.
J. Embryol. exp. Morph
.
35
,
625
635
.
Elsdale
,
T.
&
Pearson
,
M.
(
1979
).
Somitogenesis in amphibia. II. Origins in early embryogenesis of two factors involved in somite specification
.
J. Embryol. exp. Morph
.
53
,
245
267
.
Graziosi
,
G.
,
De Cristini
,
F.
,
Di Marcotullio
,
A.
,
Marzari
,
R.
,
Micali
,
F.
&
Savoini
,
A.
(
1983
).
Morphological and molecular modifications induced by heat shock in Drosophila melanogaster embryos
.
J. Embryol. exp. Morph
.
77
,
167
182
.
Herth
,
W.
&
Sander
,
K.
(
1973
).
Mode and timing of body pattern formation (regionalization) in the early embryonic development of cyclorrhapic dipterans (Protophormia, Drosophila)
.
Wilhelm Roux’ Archiv. EntwMech. Org
.
172
,
1
27
.
Kalthoff
,
K.
(
1983
).
Cytoplasmic determinants in dipteran eggs
.
In Time, Space and Pattern in Embryonic Development
(ed.
W. R.
Jeffery
&
R.
Raff
), pp.
313
348
.
New York
:
Alan R. Liss
.
Lawrence
,
P. A.
(
1973
).
A clonal analysis of segment development in Oncopeltus (Hemiptera)
.
J. Embryol. exp. Morph
.
30
,
681
699
.
Maas
,
H.
(
1949
).
Über die Auslôsbarkeit von Temperaturmodifikationen wâhrend der Embryonalentwicklung von Drosophila melanogaster
.
Wilhelm Roux’ Archiv. EntwMech. Org
.
143
,
515
572
.
Miya
,
K.
&
Kobayashi
,
Y.
(
1974
).
The embryonic development of Atrachya Menetriesi Faldermann (Coleóptera, Chrysomelidae). II. Analyses of early development by ligation and low temperature treatment
.
J. Fac. Agrie. Iwate Univ
.
12
,
39
55
.
Mee
,
J. E.
(
1984
).
The development of segments in locust embryos
.
Ph.D. thesis: University of Edinburgh
.
Mee
,
J. E.
(
1986
).
Pattern formation in fragmented eggs of the short germ insect, Schistocerca gregaria
.
Wilhelm Roux’ Archiv. devl. Biol, (submitted)
.
Mee
,
J. E.
&
French
,
V.
(
1986
).
Disruption of segmentation in a short germ insect embryo. II. The structure of segmental abnormalities induced by heat shock
.
J. Embryol. exp. Morph
.
96
,
00
00
.
Pearson
,
M.
&
Elsdale
,
T.
(
1979
).
Somitogenesis in amphibian embryos. I. Experimental evidence for an interaction between two temporal factors in the specification of somite pattern
.
J. Embryol. exp. Morph
.
32
,
533
555
.
Sander
,
K.
(
1976
).
Specification of the basic body pattern in insect embryogenesis
.
Adv. Insect Physiol
.
12
,
125
238
.
Sander
,
K.
(
1981
).
Pattern generation and pattern conservation in insect ontogenesis - problems, data and models. In Progress in Developmental Biology
(ed.
H. W.
Sauer
).
Fortschr. Zool
.
26
,
107
119
.
Schubiger
,
G.
&
Wood
,
W. J.
(
1977
).
Determination during early embryogenesis in Drosophila melanogaster
.
Am. Zool
.
17
,
565
576
.
Shulov
,
A.
&
Pener
,
M. P.
(
1963
).
Studies on the development of eggs of the desert locust Schistocerca gregaria and its interruption under particular conditions of humidity
.
Anti-Locust Bull
.
41
,
1
59
.
Summerbell
,
D.
,
Lewis
,
J. H.
&
Wolpert
,
L.
(
1973
).
Positional information in chick limb morphogenesis
.
Nature, Lond
.
224
,
492
496
.
Tyrer
,
N. M.
(
1970
).
Quantitative estimation of the stage of embryonic development in the locust, Schistocerca gregaria
.
J. Embryol. exp. Morph
.
23
,
705
718
.
Wolpert
,
L.
,
Tickle
,
C.
&
Sampford
,
M.
(
1979
).
The effect of cell killing by X-irradiation on pattern formation in the chick limb
.
J. Embryol. exp. Morph
.
50
,
175
198
.