A somite pre-pattern is established shortly before visible segmentation. The pre-pattern results from the interaction of two components: a wave of cell behavioural change that passes along the axis, and, an underlying co-ordination of the cells that is the basis for their association into large somite-sized groupings. The evidence is derived from studies of the zones of abnormal segmentation that follow temperature shocks delivered between the neurula and tail-bud stages (Pearson & Elsdale, 1979).

Temperature shock given earlier at the mid-gastrula stage is however ineffective in inducing abnormalities in somitogenesis. Shocks given before the mid-gastrula stage reveal a prior period of sensitivity stretching back into the blastula. Thus early and late sensitive periods can be defined separated by a short refactory period. Quite different patterns in the distribution of somite abnormalities characterize the results of shock during the two sensitive periods, suggesting different aetiologies.

It is concluded that the wave of rapid cell change is set up early in embryogenesis during the blastula stage, and each cell of the prospective paraxial mesoderm carries a determination to change after a specific length of time, i.e. a countdown is set in each cell. As a result of the movements of gastrulation, the prospective paraxial mesoderm cells become laid out along the axis of the neurula in the order (antero-posterior sequence) in which they will change. The achievement of the correct redistribution of the cells depends crucially on the conservation of the sequence in the blastula by the maintenance of topological integrity throughout gastrulation. It is suggested that early shock disturbs gastrulation movements, causing some mixing up of the cells resulting in incoherence of the wavefront.

Whereas early shocks are thus assumed to affect the wave, the evidence suggests that late shock undoes co-ordination. It is concluded therefore that co-ordination is established later, after the refractory period, around the late gastrula stage.

Somitogenesis is governed by two time intervals: the long time taken for the somites to form from first to last is partitioned into equal short intervals between the formation of one somite and the next.

From experiments employing temperature shocks delivered to embryos about to form and forming their somites, and the analysis of the location, extent and anatomy of the characteristic abnormal zones developed in the somite files following this treatment, the following scheme has emerged. The longer interval is defined by the time taken to traverse the axis by a wave of rapid cell change moving in advance of somite segmentation. A second component, a co-ordination of the cells, contributes to the definition of the shorter intervals; a momentary coupling of these two components as the wavefront moves along the axis, establishes a somite pre-pattern which manifests as visible segmentation after a fixed time delay (Pearson & Elsdale, 1979).

This scheme has basic similarities with a theoretical model of somitogenesis which shows how a wave of change passing through a field of co-ordinated cells can function like an escapement mechanism in a clock to provide for a repetitive series of discontinuous events (Cooke & Zeeman, 1976).

In order to shed light on the origins in early embryogenesis of the two components that interact to specify the segmental pattern, we have studied embryos temperature-shocked during gastrulation.

Three experiments have been performed. First, a series of shocks to different gastrula stages shows that somite abnormalities are induced after shock to the younger gastrula but not the older. Furthermore, the distribution of abnormal somites in the former is different from what is observed in embryos shocked at the neurula and post-neurula stages. Second, the refractory period to temperature shock during later gastrulation indicated by the previous experiment, is investigated with longer duration shocks. We show in our first paper of this series how shocks to neurula and post-neurula stages induce somite abnormalities and in addition an enduring hidden effect revealed by a second shock; in a third experiment it is shown that a shock during the refractory period having no visible effect, has no hidden effect either. It is argued that a genuine refractory period separating earlier and later periods of sensitivity to temperature shock indicates disturbances to different components in segmental specification.

Rana temporaria embryos collected in season from natural habitats have three advantages over Xenopus: (1) each spawning provides a large population of synchronously developing embryos, (2) virtually all the embryos develop normally and (3) Rana embryos develop more slowly than Xenopus. The fact that Rana embryos are nearly twice the size of Xenopus is an added convenience.

Clutches of early cleavage stage Rana were collected in Perthshire and kept at 6°C in a cooled incubator in water collected with embryos. The day before each experiment, embryos were brought to room temperature to allow a period ofadaptation to the higher temperature of 19 ± 1°C. During most experiments the embryos were maintained in pond water, some use was made of tap water that had stood for a week under continous aeration. Embryos denuded of their jelly were re-examined immediately before shock in order that visibly damaged embryos could be discarded. To give a heat shock, embryos were transferred with minimal carry over of water to glass bottles containing 125 ml of pond water standing in a water bath at 37 ±0·2°C. After the desired time at 37°C, embryos were transferred to a larger body of water at room temperature, and reared until ready for scoring. In this way a standard 9-min duration shock was repeatable within an accuracy of 10 sec (2%).

Somite files were revealed for scoring by stripping the skin from the embryos. Embryos were transferred to a solution containing 0·5% potassium dichromate and 2·5% glacial acetic acid (Smith’s fixative without formaldehyde). After a few minutes in this solution the skin becomes strippable and remains so for some 10–15 min before turning brittle. Files were scored and photographed under a stereo microscope using incident oblique illumination.

Scanning electron microscopy and histological preparation employed the same techniques used by Pearson & Elsdale (1979).

At room temperature 19 ± 1°C it takes 22 h for Rana embryos to develop from the initial gastrula to the neural plate stage. This period is conveniently divided into 22 substages marking the progress of gastrulation hour by hour. The external appearance of the blastopore at representative substages is illustrated in Fig. 1 which gives a synoptic view of the time course of gastrulation. Three of these substages are reference substages, defined independently of the series on external criteria alone, these are: SS 1 the initial gastrula, SS 12 coincident with the completion of the circular blastopore, and SS 18 at which the yolk button disappears from view, a convenient marker for the end of gastrulation. Fig. 1 shows that the lateral extension of the blastopore to form the blastoporal circle is a relatively slow process, whereas the diminution of the yolk plug proceeds rapidly thereafter. Following common usage, SS 12 is referred to as the midgastrula stage although it is actually the stage two thirds of the time through gastrulation

Fig. 1

The course of gastrulation in Rana temporaria reared at 19°C. Each sketch shows the progress made by the blastopore viewed externally at one of the substages at which a 9-min temperature shock was delivered. A selection of these substages is illustrated. How these substages accord with Shumway’s (1940) staging of gastrulation in Rana pipiens is indicated.

Fig. 1

The course of gastrulation in Rana temporaria reared at 19°C. Each sketch shows the progress made by the blastopore viewed externally at one of the substages at which a 9-min temperature shock was delivered. A selection of these substages is illustrated. How these substages accord with Shumway’s (1940) staging of gastrulation in Rana pipiens is indicated.

1 9-min temperature shocks during gastrulation

Embryos developing synchronously at room temperature were divided into lots, each receiving a shock at one of the substages of gastrulation. The experiment comprised two series of shocks, one ovulation was used throughout the first series, and another throughout the second (Table 1).

Table 1

Scheme of temperature shocks to gastrula stages

Scheme of temperature shocks to gastrula stages
Scheme of temperature shocks to gastrula stages

Series 1

The first series comprised a control, and 16 lots receiving temperature shock. Fifteen of the later provided a series of shocks from SS 1 to SS 15. A late shock at SS 22 provided an overlapping comparison with previous work. Embryos were scored for abnormal gastrulation and abnormal somitogenesis. When possible each somite was scored in both files of the same embryo at around the 25-somite stage; these scores are the basis for Fig. 3.

For descriptive purposes the series is divided into four subseries. After noting the salient features in each, the subseries are described in detail.

Subseries 1: SS 1–SS 6. Gastrulation and somitogenesis grossly abnormal.

Subseries 2: SS 7–SS 11. Gastrulation and somitogenesis improving. Posterior relegation of abnormal somites.

Subseries 3: SS 12–SS 15. Normal gastrulation. Trivial somite irregularities.

Subseries 4: SS 22. Zone of abnormal segmentation in all embryos extending over first several somites. Characteristic effect of shock to neurula stage.

Shocks to SS 1-SS 6

Only in this subseries did a significant number of embryos die without developing a primary axis, and most of this mortality was associated with shocks to the first three substages. This result already points to the trend characterizing the series as a whole, namely, the later the 9-min shock is given, the less the effect.

The overall result, although variable, is a severe disturbance to both gastrulation and somitogenesis, without however a close correlation between the severity of these two effects within the same embryo. Although most surviving embryos were stunted, a minority showed near normal axial extension although their somite files were grossly abnormal.

(a) Gastrulation

The subseries shows that the earlier a temperature shock is given during gastrulation the greater the likelihood that the external course of gastrulation is arrested at the circular blastopore stage. The great majority of embryos so arrested however subsequently developed neural plates in the presence of a large persistent yolk plug and various degrees of spina bifida.

Persistence of the yolk plug was correlated with changes in the superficial pigmented ectoderm observed within a day of shock, a pronounced crinkling and folding of the ectoderm was observed that remained a feature of this tissue after its translocation to the ventro-lateral surface of the later embryo.

(b) Somitogenesis

Somitogenesis in this subseries was so grossly abnormal that it was often impossible to score files somite by somite. In lots 1 and 2, a great many files were abnormal throughout. For comparison with later subseries, scores are given for SS 3 and SS 6 only (Fig. 3).

The appearance of abnormalities can be described under the following heads, with the qualification that the appearances described vary widely in severity between affected embryos.

  1. Inhibition. A total inhibition of segmentation was invariably associated with severe stunting and an arrest of development after neurulation resulting in disintegration. Total inhibition was however rarely encountered. Much more common was a premature cessation of somitogenesis after the formation of a variable number of abnormal somites (Fig. 2a).

  2. Faint somite boundaries, suggesting attenuation of the segmental impulse, was commonly seen. There was often a gradient along the axis, with posterior somites less clearly delineated than anterior.

  3. Local inhibition. Local failures of segmentation, bounded by regions of more normal segmentation, were common. Such local areas were generally followed by gross abnormalities.

  4. Chaotic segmentation. The impression here is of fragmentation and disorganization of the segmenting mesodermal units (Fig. 2 b). The excess fissures between the small irregularly segmented groups of cells often ran in all directions.

  5. Frayed edges, a frequent concomitant of other abnormalities, where the ventral border of somitic mesoderm was imperfectly demarcated from the lateral plate (Fig. 2 b).

  6. ′Mistakes′. Trivial inconsistencies, to be described later.

Fig. 2

Each of these stripped embryos from Series 1 received a 9-min temperature shock early in gastrulation, (a) Shock delivered to SS 1. Inhibition of segmentation. Somites are not distinguished in this grossly abnormal specimen. Some faintly demarcated furrows are alone visible, (b) Shock delivered to SS 3. Abnormal segmentation. There is a partial inhibition of segmentation anteriorly, chaotic segmentation (Ch) posteriorly. Frayed ventral edges (f) indicate indistinct demarcation of the somite mesoderm from the lateral plate, (c) Shock delivered to SS 6. Posterior restriction of abnormal somites is evident in this embryo. The anterior somites are spared, behind them somitogenesis is abnormal.

Fig. 2

Each of these stripped embryos from Series 1 received a 9-min temperature shock early in gastrulation, (a) Shock delivered to SS 1. Inhibition of segmentation. Somites are not distinguished in this grossly abnormal specimen. Some faintly demarcated furrows are alone visible, (b) Shock delivered to SS 3. Abnormal segmentation. There is a partial inhibition of segmentation anteriorly, chaotic segmentation (Ch) posteriorly. Frayed ventral edges (f) indicate indistinct demarcation of the somite mesoderm from the lateral plate, (c) Shock delivered to SS 6. Posterior restriction of abnormal somites is evident in this embryo. The anterior somites are spared, behind them somitogenesis is abnormal.

Shocks to SS7–SS 11

There is no discontinuity between this subseries and the first. Most of the embryos gastrulated normally and the pigmented ectoderm remained smooth. The somite scores in Fig. 3 show the abnormalities in retreat; not only are there fewer abnormal somites, there is also a clear trend towards the sparing of anterior somites and a confinement of abnormalities to more posterior levels, (Fig. 2c). A further trend not brought out in Fig. 3 is towards a decline in the severity of abnormalities.

Fig. 3

9-min temperature shocks during gastrulation, Series 1. Results. Embryos were stripped and scored between the 20 and 25 somite stages. Each somite in each embryo was numbered 1, 2, 3…according to its location in the sequence of somites, and scored as either normal or abnormal. The degree of abnormality was not taken into account. These scores provide the primary data. Reference to Table 1 shows that a significant number of embryos shocked at SS 3 and SS 6 died before their batches were scored. The data from these substages are therefore biased. The remainder of the data is not biased. For each numbered somite, the scores, from all the embryos shocked at the same substage, were summed. The proportion of somites abnormal within each of these summations was converted to a percentage. These percentages are displayed in histograms. The trend towards fewer abnormalities and the sparing of anterior somites on passing from SS 6 to SS 11 is clear. From SS 12 to SS 15 abnormal scores are no higher than controls, and like the control scores reflect merely trivial irregularities of the type termed ‘mistakes’, see page 258 and Fig. 7. The figure shows in striking manner the sensitivity of the earlier gastrula to disturbance by temperature shock, and how as development proceeds the gastrula becomes more and more resistant, and finally refractory around SS 12–15. By SS 22 however the embryo is again easily disturbed, but now the distribution of abnormalities is quite different from before, and they are concentrated into an anterior abnormal zone.

Fig. 3

9-min temperature shocks during gastrulation, Series 1. Results. Embryos were stripped and scored between the 20 and 25 somite stages. Each somite in each embryo was numbered 1, 2, 3…according to its location in the sequence of somites, and scored as either normal or abnormal. The degree of abnormality was not taken into account. These scores provide the primary data. Reference to Table 1 shows that a significant number of embryos shocked at SS 3 and SS 6 died before their batches were scored. The data from these substages are therefore biased. The remainder of the data is not biased. For each numbered somite, the scores, from all the embryos shocked at the same substage, were summed. The proportion of somites abnormal within each of these summations was converted to a percentage. These percentages are displayed in histograms. The trend towards fewer abnormalities and the sparing of anterior somites on passing from SS 6 to SS 11 is clear. From SS 12 to SS 15 abnormal scores are no higher than controls, and like the control scores reflect merely trivial irregularities of the type termed ‘mistakes’, see page 258 and Fig. 7. The figure shows in striking manner the sensitivity of the earlier gastrula to disturbance by temperature shock, and how as development proceeds the gastrula becomes more and more resistant, and finally refractory around SS 12–15. By SS 22 however the embryo is again easily disturbed, but now the distribution of abnormalities is quite different from before, and they are concentrated into an anterior abnormal zone.

Shocks to SS 12–SS 15

The subseries is characterized by normal somite files; such anomalies as were observed were trivial irregularities classed as mistakes, to be described later. This subseries demonstrates the refractory period to temperature shock.

Shock to SS 22

The effect of a shock to embryos about to form and forming their somites is much more predictable than the effect of a shock to earlier stages; abnormal segmentation is confined to a limited zone the length of which is proportional to the duration of shock (Elsdale, Pearson & Whitehead, 1976). The location of the abnormal zone is precisely correlated with the progress of segmentation; a shock delivered shortly before somitogenesis commences, as in the present case, induces an abnormal zone including the very first somites; following a shock to embryos in the course of forming their somites, the zone invariably commences three to four somites posterior to the last somite formed at the time of shock. The abnormal zone is characterized by chaotic segmentation and an excess of fissures anteriorly, and a gradual amelioration and return to normal segmentation posteriorly.

1B Series 2

The first series gives a detailed picture of the transition from an early gastrula extremely sensitive to disturbance by temperature shock to a ‘mid-gastrula’ that is refractory. The gap between SS 15 and SS 22 however leaves undocumented the transition from the refractory stage to the sensitive early neurula.

The series comprised a control and six lots shocked at hourly intervals from SS 14 to SS 19. The results are presented in Fig. 4 and Table 2. Abnormalities among the first formed somites are significant in embryos shocked at SS 16–17, and typical abnormal zones are present in embryos shocked at SS 19.

Table 2

Results of second series of shocks to gastrula substage also including the data from Series 1, substage 22

Results of second series of shocks to gastrula substage also including the data from Series 1, substage 22
Results of second series of shocks to gastrula substage also including the data from Series 1, substage 22
Fig. 4

9-min temperature shocks during gastrulation, Series 2. Results, see in conjunction with Table 2. The format is the same as Fig. 3. Embryos were stripped and scored between the 10- and 15-somite stages, and for this reason the data are not strictly comparable with the data from the first series. By the 20- to 25-somite stages at which the first series embryos were scored the first three somites are becoming partially obscured by developments in the ear region. As shocks around the conclusion of the refractory period were expected to induce abnormalities in these first formed somites, the second series was scored at an earlier stage at which these somites could be viewed without dissection after stripping. The drawback to early scoring is that the segmental boundaries have not completely settled down and their appearance often presents small irregularities absent from older embryos. For this reason we have a rather high control count amounting to about one abnormal somite per file. The first series results revealed a refractory period to the standard temperature shock commencing around SS 11 and continuing to at least SS 15. The second series results show that in fact an SS 15 embryo is already approaching the end of the refractory period, and a shock at SS 17 reliably induces abnormalities among the first two or three somites. As embryos develop towards the neural plate stage, their susceptibility to shock further increases, and the abnormal zones induced by the 9-min shock lengthen.

Fig. 4

9-min temperature shocks during gastrulation, Series 2. Results, see in conjunction with Table 2. The format is the same as Fig. 3. Embryos were stripped and scored between the 10- and 15-somite stages, and for this reason the data are not strictly comparable with the data from the first series. By the 20- to 25-somite stages at which the first series embryos were scored the first three somites are becoming partially obscured by developments in the ear region. As shocks around the conclusion of the refractory period were expected to induce abnormalities in these first formed somites, the second series was scored at an earlier stage at which these somites could be viewed without dissection after stripping. The drawback to early scoring is that the segmental boundaries have not completely settled down and their appearance often presents small irregularities absent from older embryos. For this reason we have a rather high control count amounting to about one abnormal somite per file. The first series results revealed a refractory period to the standard temperature shock commencing around SS 11 and continuing to at least SS 15. The second series results show that in fact an SS 15 embryo is already approaching the end of the refractory period, and a shock at SS 17 reliably induces abnormalities among the first two or three somites. As embryos develop towards the neural plate stage, their susceptibility to shock further increases, and the abnormal zones induced by the 9-min shock lengthen.

1C Summary of gastrula shocks

A refractory period within the latter half of gastrulation during which temperature shock does not induce somite abnormalities is flanked by preceding and succeeding periods of greater reactivity.

Shocks delivered at the time the blastopore is initiated induce abnormal gastrulation and persistence of the yolk plug, and our data show that the first 25 somites are about equally at risk to severe disturbance. The trend exhibited in embryos shocked closer to the mid-gastrula stage is towards normal gastrulation, a progressive sparing of anterior somites and a confinement of abnormalities to more posterior levels concomitant with a declining incidence and severity. This trend culminates in the absence of abnormalities from embryos shocked at the beginning of the refractory period.

The trend exhibited, following a succession of shocks delivered as the late gastrula develops towards the neural plate stage, is towards increasing abnormalities among the first formed somites culminating in the typical abnormal zones characteristic of later shocks.

The two periods of reactivity to temperature shock separated by the refrac- tory period are distinguished by quite different patterns in the distribution of segmental anomalies.

For practical purposes the refractory period is preferably delimited in terms of our reference substages based on unequivocal external criteria. It needs to be kept in mind however that embryos neither enter nor leave the refractory period abruptly; in Fig. 5 the data are used to present the refractory period as a basin with sloping sides. Reference substage 12 the ‘mid-gastrula’ stage is a satisfactory marker for the beginning of the RP, reference substage 18 however is too advanced to represent the end of the period which corresponds with SS 16. In Rana embryos reared at 19°C the refractory period can be said to last about 4 h.

Fig. 5

The refractory period. Data from both series are employed to present the refractory period as a basin with sloping sides. Solid circles, Series 1: For each substage the total abnormal somites scored has been divided by the number of files scored, to give a figure for the mean number of abnormal somites per file. The appropriate axis is on the left. The base line represents the control for the series. Open circles, Series 2: The data along the bottom row of Table 2 are used. The appropriately scaled axis is on the right. The dashed horizontal line represents the control for the series. The figure shows that the refractory period is centred around SS 13.

Fig. 5

The refractory period. Data from both series are employed to present the refractory period as a basin with sloping sides. Solid circles, Series 1: For each substage the total abnormal somites scored has been divided by the number of files scored, to give a figure for the mean number of abnormal somites per file. The appropriate axis is on the left. The base line represents the control for the series. Open circles, Series 2: The data along the bottom row of Table 2 are used. The appropriately scaled axis is on the right. The dashed horizontal line represents the control for the series. The figure shows that the refractory period is centred around SS 13.

2 20-min shock during the refractory period (RP)

Shocks of long duration were employed in order to discover whether embryos passing through the RP were genuinely incapable of reacting to shock by abnormal segmentation, or whether they were merely more resistant to shock.

Early gastrulae are killed by a 20-min shock. Delivered after the RP a shock of this duration is not lethal, but induces very long abnormal zones; indeed only when the abnormal zone commences within the first 20 somites is there an eventual return to normal segmentation caudally.

Delivered at SS 13 (RP), around a quarter of the embryos are killed by a 20-min shock, showing it is near the maximum that can be used in a meaningful experiment employing embryos at this stage. The immediate effect of temperature shock is an arrest of development the duration of which is proportional to the length of the shock. Subsequently, embryos resume development at the normal rate. Survivors of a 20-min shock at SS 13 resume development after a protracted arrest lasting nearly three days; they nevertheless continue to grow normally forming an extended body axis and tail, apart from a kink in the tail that possibly indicates a local effect on notochordal development. Segmentation appears normal in these embryos and is nowhere seriously disturbed (Fig. 6). The sole irregularities are those classed as mistakes. These latter are increased to an average of 7·7 per file, a 40-fold increase over controls (Table 3).

Table 3

The frequency of mistakes counted at the 35-somite stage following temperature shocks delivered at SS 13 during the refractory period

The frequency of mistakes counted at the 35-somite stage following temperature shocks delivered at SS 13 during the refractory period
The frequency of mistakes counted at the 35-somite stage following temperature shocks delivered at SS 13 during the refractory period
Fig. 6

An embryo stripped at the 35-somite stage following a 20-min temperature shock during the refractory period (SS 13). (a) Anterior trunk; (b) part of tail. These pictures could pass for illustrations of a normal untreated embryo.

Fig. 6

An embryo stripped at the 35-somite stage following a 20-min temperature shock during the refractory period (SS 13). (a) Anterior trunk; (b) part of tail. These pictures could pass for illustrations of a normal untreated embryo.

Mistakes (Fig. 7) refer to trivial irregularities reflecting nothing more than a failure in the correct registration of the dorsal and ventral half somites. The commonest appearances are a Y formation when two adjacent dorsal half somites are confluent with a single ventral half somite, and the reverse situation giving an inverted Y. Occasionally a floating half somite may occur; three half somites joined to one has been observed but is unusual.

Fig. 7

Mistakes. Short lengths from the tails of two embryos stripped between the 25- and 30-somite stages. Left is anterior, top is dorsal, (a) The third complete segment counting from the left is a Y formation, two dorsal half somites are fused to one ventral half somite. This is counted as one mistake. Adjacent to the right is an inverted Y formation. The two mistakes compensate, for together they comprise three dorsal half somites and three ventral, and the pattern following is not put out. (b) Two similar mistakes to those in (a) but here separated by one whole somite. The imbalance due to the extra dorsal half somite created by the first mistake persists through an extra somite before being corrected by the second mistake. Mistakes do not always occur in pairs like this.

Fig. 7

Mistakes. Short lengths from the tails of two embryos stripped between the 25- and 30-somite stages. Left is anterior, top is dorsal, (a) The third complete segment counting from the left is a Y formation, two dorsal half somites are fused to one ventral half somite. This is counted as one mistake. Adjacent to the right is an inverted Y formation. The two mistakes compensate, for together they comprise three dorsal half somites and three ventral, and the pattern following is not put out. (b) Two similar mistakes to those in (a) but here separated by one whole somite. The imbalance due to the extra dorsal half somite created by the first mistake persists through an extra somite before being corrected by the second mistake. Mistakes do not always occur in pairs like this.

In the case of late shocks (delivered after the RP) inducing a limited abnormal zone, the length of this zone is proportional to the duration of shock; in response to minimal shock the abnormal zone reduces to a mistake. However, mistakes occur in other situations where they cannot be considered as reduced abnormalities. Mistakes are a natural occurrence, routinely observed both in controls reared in the laboratory and in embryos reared in their natural habitats. Among the latter, around one file in six bears one or more mistakes and it is interesting that these are not observed in the first 13 somites that alone persist after meta-morphosis. Thus mistakes can hardly be considered abnormal in themselves.

The incidence of mistakes can be increased by procedures other than heat shock that do not induce somite abnormalities. Mistakes are always increased following temperature shock; thus, for example, there is an increased incidence of mistakes posterior to the abnormal zones, within the remainder of the files otherwise segmenting normally following a late shock. In neither of these two cases can extra mistakes be considered reduced abnormalities.

It appears that mistakes reflect an aspect of somitogenesis that has remained unstable under the influence of natural selection, and is therefore easily disturbed. For this reason we do not consider extra mistakes in the absence of abnormalities as an effect of a severe temperature shock continuous with the production of gross abnormalities.

The significance of the experiment is that it clearly demonstrates the extra-ordinary refractoriness on the part of embryos at the mid-gastrula stage to the induction of somite abnormalities, and thereby incidentally contrives an unobscured window onto the raised background of mistakes that invariably accompanies temperature shock.

3 Two temperature shocks to the same embryo

By giving two temperature shocks to the same embryo, we have obtained two abnormal zones along the somite files separated by normal somites (Pearson & Elsdale, 1979). The interesting fact emerged from these experiments that the second abnormal zone was invariably only half as long as expected, that is to say half as long as the zones in control embryos that had not received the earlier of the two shocks. This result demonstrated an enduring effect of temperature shock persisting after segmentation had returned to normal, conferring a partial protection against a subsequent shock. Does a shock during the refractory period confer a similar protection?

We compared the repercussions of a shock delivered during the RP with the repercussions of a shock delivered at the first somite stage. Two experiments were performed. In the first, the test shocks were given to both groups at the 8-somite stage and the interval between the first and second shocks was different for the two groups. In the second experiment the interval between the two shocks was the same for both groups and the embryos received their second shocks at different stages. The results are presented in Table 4 and Fig. 8. In contrast to a shock at the first somite stage, a shock during the RP has no influence on the effect of a subsequent shock.

Table 4

Double temperature shock experiment

Double temperature shock experiment
Double temperature shock experiment
Fig. 8

Double temperature shock experiment. The primary data consist of estimates of the number of somites within abnormal zones induced by 8-min temperature shocks. Estimates were made according to Pearson & Elsdale (1979). In the case of embryos with two abnormal zones as a result of two shocks, the first zone was ignored as only the length of the second was relevant to the experiment. A few embryos were discarded in which the two somites, immediately preceding the second abnormal zone, were not unequivocally normal. In order to compare the lengths of the abnormal zones in two classes of embryos it is necessary to take into account the variation within each class. The figure therefore displays distributions showing the proportion of files in each class having abnormal zones of length 0,1,2,… somites. The proportions are expressed as percentages. Each graph compares two such distributions. The distribution bounded by the solid line, in each case refers to abnormal zones induced by a second shock. The accompanying distribution bounded by the dashed line refers to the corresponding control embryos that did not receive the earlier of the two shocks. Of significance is the extent to which the two distributions overlap; the overlaps are therefore cross hatched. The protective effect of a temperature shock is demonstrated by a leftward shift of the distribution bounded by the solid line in relation to the distribution bounded by the dashed line. Such an effect is demonstrated in the two graphs on the left (1a and 2a) referring to embryos receiving a first shock at the 1-somite stage. In contrast the distributions on the right (1b and 2b) overlap over most of their ranges indicating that a shock during the refractory period has no protective effect.

Fig. 8

Double temperature shock experiment. The primary data consist of estimates of the number of somites within abnormal zones induced by 8-min temperature shocks. Estimates were made according to Pearson & Elsdale (1979). In the case of embryos with two abnormal zones as a result of two shocks, the first zone was ignored as only the length of the second was relevant to the experiment. A few embryos were discarded in which the two somites, immediately preceding the second abnormal zone, were not unequivocally normal. In order to compare the lengths of the abnormal zones in two classes of embryos it is necessary to take into account the variation within each class. The figure therefore displays distributions showing the proportion of files in each class having abnormal zones of length 0,1,2,… somites. The proportions are expressed as percentages. Each graph compares two such distributions. The distribution bounded by the solid line, in each case refers to abnormal zones induced by a second shock. The accompanying distribution bounded by the dashed line refers to the corresponding control embryos that did not receive the earlier of the two shocks. Of significance is the extent to which the two distributions overlap; the overlaps are therefore cross hatched. The protective effect of a temperature shock is demonstrated by a leftward shift of the distribution bounded by the solid line in relation to the distribution bounded by the dashed line. Such an effect is demonstrated in the two graphs on the left (1a and 2a) referring to embryos receiving a first shock at the 1-somite stage. In contrast the distributions on the right (1b and 2b) overlap over most of their ranges indicating that a shock during the refractory period has no protective effect.

It is concluded therefore that a shock during the RP neither induces abnormalities along the somite file, nor has the enduring protective effect of a shock delivered to later stages.

The first paper in this series (Pearson & Elsdale, 1979) dealt in detail with the effect of a late temperature shock (delivered after the RP). Following a section on the significance of the RP, the bulk of this discussion is devoted to the effect of early temperature shock (delivered before the RP).

The refractory period

We have suggested on the basis of previous experiment a mechanism by which late shocks cause abnormalities in segmentation. It was assumed that late shocks disturb all of the paraxial mesoderm not yet committed to segmentation, by undoing a co-ordination of the cells essential for normal patterning. The disturbed tissue slowly recovers. Crucial is the time available for recovery before the arrival of the wave of rapid cell change. The somite pre-pattern established with the passing of the wave reflects the state of co-ordination of the cells, and disturbed co-ordination at this time will leave its indelible record in the somite file (Fig. 15, Pearson & Elsdale, 1979).

On the basis of this scheme it might be supposed that shocks during the RP produced no somite abnormalities because recovery is invariably complete before the pre-pattern of the first somites is established. There is no time for such recovery. The first somite forms as the neural folds close; utilizing the data displayed in fig. 7 of Pearson & Elsdale (1979), it can be estimated that the determination of the first somite follows within a few hours of the RP.

An alternative possibility is that temperature shock during the RP is not registered by the paraxial mesoderm. Two independent tests can be applied. First, one can look for somite abnormalities induced by shock. Second, the enduring protective effect of a shock can be investigated by a double shock experiment. Negative results from both tests indicate the unresponsiveness of the prospective paraxial tissue to a temperature shock during the RP.

Having previously concluded (Pearson & Elsdale, 1979) that late shock dis- turbs the co-ordination of the cells essential for normal patterning, while leaving the wave untouched, we infer that co-ordination is not established until the end of the RP. It follows that co-ordination cannot be the target of early shock; this target must be absent from the beginning of the RP. Moreover, separate aetiologies are strongly suggested by the characteristically different distributions of abnormalities along the somite files associated with early and late shocks.

The RP is interpreted therefore as an interval between two periods of differing reaction to temperature shock, in which the occasion for the earlier has passed and the conditions for the later have yet to mature.

The effect of early temperature shock

1 No recovery after early shock

We next enquire whether the data from early shock provide the same sort of evidence for a recovery process as do the data from late shocks.

A recovery process was postulated to account for the invariable finding that the longer the interval between a late shock and the formation of the somite pre-pattern coincident with the advance of the wave of rapid cell change, the less severe the induced abnormalities in segmentation, exemplified by the gradual return to normal segmentation behind the anterior of the abnormal zone (Elsdale et al. 1976). We enquire therefore whether a similar rule characterizes the result of early shock.

The data show that the shock of the same duration induces more numerous and more severe somite abnormalities delivered to the early gastrula than delivered to progressively later gastrula stages, although the time available for recovery is longest in the case of the former. Further, through SS 1 to SS 12 we observe a progressive sparing of anterior somites and a relegation of abnormalities to the more posterior of the first 20–25 somites, although anterior cells have less time to recover than posterior cells. These results do not merely offer no support for a recovery process, they provide a compelling contra-indication and justify the rather categorical conclusion that there is no recovery after early shock. Using minimal shocks it is likely that Cooke obtained a less convincing contra-indication than we did using more severe shocks, and he is not deterred on the basis of his results on Xenopus from envisaging a slow recovery after early shock (Cooke, 1978).

2 Delimiting the target of early shock

Unpublished observations reveal that the blastula is even more sensitive to heat shock than the early gastrula. Observations over a 16 h period show that the sensitivity of the late blastula declines steadily during the approach to gastru-lation. Seen from this perspective our results indicate that as gastrulation gets under way sensitivity to heat shock dies away sharply. In our Series 1 we are observing the tail end of a sensitive period not yet adequately explored, extending back into the blastula.

The target in the earliest gastrula comprises the precursors of no more than the first 30 somites (Cooke, 1978). As gastrulation proceeds the target area (in terms of prospective somites) contracts as the precursors of the anterior somites fall out of the target. We can picture a change from target to non-target overtaking the prospective somite mesoderm in the direction of the future anterior posterior axis. At the same time as the target is thus contracting, the number and severity of the abnormalities inducible within the residual target area is declining. We can picture therefore a concomitant attenuation or dilution of targets. By the mid-gastrula stage marking entry into the R.P., the target it seems has been eliminated as a result of the twin processes of contraction and attenuation.

We next seek to locate this target in the early embryo. In the absence of detailed knowledge of gastrulation in Rana we are forced to make assumptions on the basis of knowledge of other anurans, expecially Xenopus. A tentative correlation of gastrulation movements with the contraction and attenuation of the target of early shock and the timing of the R.P. suggests that the prospective somite mesoderm is especially sensitive prior to involution, and perhaps still somewhat sensitive during involution, whereas the tissue joining the mesodermal mantle following involution is no longer sensitive. However, the decline in sensitivity to temperature shock during the first two-thirds of gastrulation is much too rapid to correlate with the course of involution of the prospective somite tissue. It is Raymond Keller’s educated guess (personal communication) that perhaps no more than the first five prospective somites have involuted by the mid-gastrula stage marking the beginning of R.P. Even allowing considerable latitude to Keller’s estimate, we must envisage a sharp decline in sensitivity to temperature shock of the prospective somite tissue prior to its involution, a decline that may indeed already be under way in the late blastula.

These considerations raise the question whether it is the cells themselves that are disturbed by heat shock independently of the movements of gastrulation, or whether it is these movements that are disturbed in which case it is the consequent irregularities in the spatial distribution of the cells that are eventually translated into abnormalities in segmentation.

3 Abnormal gastrulation after early shock

Although the direct effect of heat shock on the prospective somite mesoderm is not observable externally, observation of the striking and characteristic effect of early shock on the superficial tissues of the embryo is of interest because it indicates that the same early shocks that lead to a disturbance of somitogenesis also interfere with the morphogenetic movements of gastrulation.

Epiboly, whereby the yolk is enclosed by a ventral streaming of the dorsal ectoderm, occurs during the second half of gastrulation following the formation of the circular blastopore. The inhibition of epiboly after shock to the early gastrula leading to persistent yolk plug and spina bifida has been described. It appears that early shock induces a premature and unco-ordinated expansion of the dorsal ectoderm manifested by folding and crinkling of the surface, with the further result that the co-ordinated ventral streaming movements by which the yolk is enclosed are incapacitated. It is interesting that the gastrulation advances towards the R.P. as does the somite-forming system.

The disturbance of ectodermal epiboly following early shock may point to a general disturbance of the gastrulation movements. However, somite abnormalities may be the only abnormal indications in embryos otherwise comparable to controls after allowing for the retardation which invariably follows early shock. In such embryos there can have been no gross disturbance of gastrulation; at most, local irregularities in the redistribution of the cells. If early shock were to induce such local irregularities these could be significant and have later repercussions only if in normal development the local ordering of the cells was conserved. Proof that the local ordering of the cells is conserved through gastrulation, is perhaps the most remarkable, if least emphasized result to emerge from vital dyeing experiments whose history stretches back over half a century (Vogt, 1929; Lovtrup, 1966,1975; Keller, 1975,1976). They show that dye marks made prior to gastrulation become plastically deformed but not fragmented within any one layer of the neurula. This indicates that topological integrity is maintained throughout gastrulation; the tissue redistribution of gastrulation occurs without tearing such that cells originally close together within the same germ layer of the blastula will be found (or their descendants will be found) still close together in the neurula.

Current thinking on the performance of stress-free topological redistribution during early development envisages the germ layers behaving as elasticoviscous liquids in which the cells are individually deformable and free to slip on one another (Phillips, Steinberg & Lipton, 1977; Phillips & Steinberg, 1978). If all the cells slip a little in the course of plastic deformation of the tissue layers, stresses are minimized and local neighbour relations minimally changed. Impairment of liquid behaviour would lead to the build up of stresses during tissue redistribution released by tearing and the fault-like movement of one line of cells on another parting old neighbours and making new neighbours of cells previously at a distance.

It is necessary to consider therefore whether local alterations in neighbour relations could be responsible for abnormalities in the segmental pattern.

4 The kinematic nature of the waves associated with segmentation

A wave that depends upon the propagation of a signal will be stopped at a cut across its path; a wave not so stopped that appears to jump across the cut is termed a kinematic wave (Zeeman, 1975). A simple transection experiment will therefore determine the nature of a wave of change crossing the embryo.

Embryos cut transversely in two after gastrulation and before any somites have formed, and the two halves reared separately, segment the same number of somites as intact embryos (Deuchar & Burgess, 1967). Pearson & Elsdale (1979) have confirmed this result and determined that the somites are segmented in the correct order, somite formation commencing in the posterior fragment immediately after it is completed in the anterior piece. Using temperature shocks they also proved that in their experiments the prior wave had not reached the posterior halves at the time of their separation. A length of unsegmented paraxial mesoderm cut out and replaced after rotation through 180° segments caudocranially Pearson (unpublished); the same result has been obtained in the chick (Christ, Jacob & Jacob, 1974; Menkes & Sandor, 1977).

These results rule out a stimulus propagated from cell to cell along the axis in a cranio-caudal direction responsible for the wave of rapid cell change, and imply that the cells of the paraxial mesoderm behave autonomously according to endogenous countdowns, the progress of the wave reflecting the layout of the cells along the axis in the order in which they are pre-set to change. The transection experiments show that the countdowns are set not later than the end of gastrulation. However, they are probably set much earlier for we know of no evidence for regulation of the temporal course of somitogenesis after the blastula stage. Indeed Zeeman (1974) has suggested that the countdowns are set in ordered sequence with the passage of a primary wave identified with mesodermal induction in the blastula (Nieuwkoop, 1969). The maintenance of topological integrity would be essential to ensure that the layout of the cells in the blastula was correctly carried over into the neurula under these circumstances.

5 An hypothesis to account for abnormal somitogenesis following early shock

Each cell of the prospective paraxial mesoderm in the early gastrula is set to undergo an abrupt change in behaviour at a later time. This change is one of the components in the establishment of the segmental pre-pattern. The pre-somitic cells in the early gastrula are arranged in a graded series with respect to the times they are set to perform their changes. Under normal circumstances following the redistribution of the tissue during gastrulation, starting in the late neurula a sharp coherent frontier between the changed and unchanged cells sweeps slowly, tailwards down the axis. This result depends on the maintenance of topological integrity throughout gastrulation to ensure that the layout of presomitic cells along the axis in the neurula is a continuous transformation of the layout of their precursors in the blastula. Temperature shock alters the prop-erties of the non-involuted prospective mesoderm in a way that makes gastrulation difficult, with the result that local breakdowns of topological integrity occur with consequent re-ordering of local neighbour relations among the cells. Across a region of tissue within which the order of the cells has been mixed up, the wavefront of abrupt change fragments and a confused pre-pattern results.

In a concurrent study of the effects of early temperature shocks carried out on Xenopus, Cooke (1978) proposed that early shocks might disturb development of the molecular machinery of hypothetical cellular oscillators, the entrained activities of which conferred co-ordination. This was an attempt to account for the seemingly random distribution of abnormalities along the somite files following shocks at any time during the early sensitive period. Because Rana embryos develop more slowly than Xenopus, and large numbers all at the same stage are obtainable it has been possible to investigate the different stages of the early sensitive period with a greater temporal resolution than was possible using Xenopus. This has resulted in the recognition of the trends described in this paper showing that the distribution of abnormalities is not in fact random but correlated with the precise stage of gastrulation shocked. Similar trends have since been confirmed in Xenopus (Cooke, personal communication). The data presently available thus permit the more concrete hypothesis here presented.

Our view of the significant events leading up to somite formation and the reactions to temperature shock are summarized in Table 5.

Table 5

Events leading up to somite formation and reaction to temperature shock

Events leading up to somite formation and reaction to temperature shock
Events leading up to somite formation and reaction to temperature shock

The authors wish to thank the Medical Research Council for occasional support to M. J. Pearson during a prolonged period of unemployment, Jonathan Bard, Jonathan Cooke, Duncan Davidson and Christopher Zeeman for their interest and helpful discussion, Allyson Ross for technical assistance, Sandy Bruce for art work and photography and Elaine Smith for manuscript preparation.

Christ
,
B.
,
Jacob
,
H. I.
, &
Jacob
,
M.
, (
1972
).
Experimentelle Untersuchungen zur Somitenentwicklung beim Hühnerembryo
.
Z. Anat. EntwGesch
.
138
,
82
97
.
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 modelf for control of the number of repeated structures during animal morphogenesis
.
J. theor. Biol
58
,
455
476
.
Deuchar
,
E. M.
&
Burgess
,
A. M. C.
(
1967
).
Somite segmentation in amphibian embryos: is there a transmitted control mechanism?
J. Embryol. exp. Morph
.
17
,
349
358
.
Elsdale
,
T.
,
Pearson
,
M.
&
Whitehead
,
M.
(
1976
).
Abnormalities in somite segmentation following heat shock to Xenopus embryos
.
J. Embryol. exp. Morph
.
35
,
625
635
.
Keller
,
R. E.
(
1975
).
Vital dye mapping of the gastrula and neurula of Xenopus laevis. I
.
Devi Biol
.
42
,
222
241
.
Keller
,
R. E.
(
1976
).
Vital dye mapping of the gastrula and neurula of Xenopus laevis. II
Devi Biol
.
51
,
118
137
.
Lovtrup
,
S.
(
1966
).
Morphogenesis in the amphibian embryo. Cell type distribution, germ layers and fate maps
.
Acta zool. Stockh
.
47
,
209
276
.
Lovtrup
,
S.
(
1975
).
Fate maps and gastrulation in Amphibia-a. critique of current view
.
Can. J. Zool
.
53
,
473
479
.
Menkes
,
B.
&
Sandor
,
S.
(
1977
).
Somitogenesis: regulation potencies, sequence determination, and primordial interactions
.
Vertebrate Limb and Somite Morphogenesis
Ed.
D. A.
Ede
,
T. R.
Hinchliffe
&
M.
Balls
, pp.
405
420
.
Nieuwkoop
,
P. D.
(
1969
).
The formation of the mesoderm in Urodelan amphibians. 1. Induction by the endoderm
.
Wilhelm Roux Arch. EntwMech. Org
.
162
,
341
373
.
Nieuwkoop
,
P. D.
&
Faber
,
J.
(
1956
).
Normal Table of Xenopus laevis (Daudin)
.
Amsterdam
:
North Holland
.
Pearson
,
M. J.
&
Elsdale
,
T.
(
1979
).
Somitogenesis in Amphibia. I. Experimental evidence for an interaction between two temporal factors in the specification of the osmite pattern
.
J. Embryol. exp. Morph
.
51
,
27
50
.
Phillips
,
H. M.
,
Steinberg
,
M. S.
&
Lipton
,
B. H.
(
1977
).
Embryonic tissues as elasticoviscous liquids. II. Direct evidence for cell slippage in centrifuged aggregates
.
Devi Biol
.
59
,
129
134
.
Phillips
,
H. M.
&
Steinberg
,
M. S.
(
1978
).
Embryonic tissues as elasticoviscous liquid: I. Rapid and slow shape changes in centrifuged cell aggregates
.
J. Cell Sci
.
30
,
1
20
.
Shumway
,
W.
(
1940
).
Stages in the normal development of Rana pipiens. I. External form
.
Anat. Rec
.
78
,
139
.
Vogt
,
W.
(
1929
).
Gestaltunganalyse am Amphibienkeim mit ortlicher Vitalfarbung. 11. Teil. Gastrulation und Mesodermbildung bei Urodelen und Anuren
.
Wilhelm Roux Arch. EntwMech. org
.
120
,
384
406
,
Zeeman
,
E. C.
(
1975
).
Primary and secondary waves in developmental biology. Lectures on Maths in the life Sciences, 7. Amer
.
Math. Soc., Rhode Island, U.S.A
, pp.
69
161
.