The effects of ablating individual earthworm embryo ectoteloblasts were studied in terms of the morphology of the ventral nerve cord (VNC) and a discrete population of histochemically identified cells normally present in each segment of the worm. The activity of ectoderm cells in the vicinity of the ablated teloblast was assessed by observing the pattern of cells undergoing DNA synthesis for up to 72 h after an ablation. While examination of the morphology of the VNC in whole mounts suggests that a reconstitutive process is elicited by ectoteloblast ablation, this is not borne out in sections and at the level of individual, histochemically identified cells, which up to the hatchling stage appeared to be missing following ablation of particular ectoteloblasts. A proliferative response to ablation by the ectoderm cells in the vicinity of the ablated teloblast, however, suggests that some reconstitutive process(es) operate in this annelid, even if they do not result in a completely normal nerve cord at the hatchling stage.

In 1974 Devries published the results of an experiment in which he used ultraviolet (UV) light to irradiate the entire ectoteloblastic region of the embryo of the earthworm Eisenia foetida (unicolor), simultaneously destroying all eight ectoteloblasts. The ectoteloblasts are stem cells characteristic of annelid and some arthropod embryos (Anderson, 1973), which give rise to neural and epidermal tissues. 9 days after the operation, Devries observed the addition of apparently normal segments posterior to the irradiated zone. His drawings show numerous segments in which the VNC, nephridia and setae are present in their normal arrangement. According to Devries the new segments are generated by the proliferation and differentiation of the so-called banal ectoderm in the absence of all the ectoteloblasts.

This apparently regulative response to ectoteloblast ablation in the earthworm embryo is not a unique phenomenon among the oligochaetes. Penners (1934, 1937) has described a similar response in the Tubifex embryo. He found that ablation of the ectoteloblasts induced cells derived from the mesoteloblasts to proliferate and migrate ventrally, forming nerve cord and ectodermal structures. Further, there are oligochaetes, such as Stylaria (Naididae), which reportedly lack ectoteloblasts altogether (Dawydoff, 1942). Instead, the 2d cell, from which the ectoteloblasts are normally derived, gives rise to an apparently homogeneous sheet of ectoderm, which subsequently differentiates into neural and ectodermal structures, suggesting that a set of uniquely committed ectodermal precursor cells is not necessary for the generation of the annelid nerve cord or other ectodermal derivatives.

In contrast, experiments in another annelid embryo, the leech, have demonstrated that each ectoteloblast generates a specific population of cells and that while the loss of epidermal progeny is compensated for by the proliferation of the progeny of other ectoteloblasts, the neural progeny of the ectoteloblasts cannot be replaced (Blair, 1983; Blair and Weisblat, 1984). Thus, the ablation of the N teloblast, which contributes the majority of neurons in each segmental ganglion (Weisblat et al. 1980; Kramer and Weisblat, 1985), results in a dramatic depletion and disorganisation of the VNC (Blair and Weisblat, 1982). As noted in the accompanying paper, adult oligochaetes and leeches differ in other ways as well; while oligochaetes have a variable number of segments and are able to regenerate those that are lost (Morgan, 1901; Sawyer, 1986), leeches have a fixed segment number and cannot replace lost segments (Needham, 1952; Sawyer, 1986).

The experiments described here are an attempt to clarify Devries’s findings by examining the effects of ablating individual ectoteloblasts in the earthworm embryo. The effects are assessed in terms of their impact on the formation of the VNC, since in the leech it is nerve cells that are not replaced. Because of difficulties inherent in working with the earthworm embryo (see companion paper), the ablation experiments reported here were not performed in conjunction with lineage tracers. However, the normal pattern of contribution of the N teloblast to the developing nerve cord has been established and shown to be very similar to that made by the N teloblast in the leech embryo (companion paper). Comparisons with other equivalent teloblasts in the leech embryo (the O,P and Q teloblasts), help to provide a context in which to study the effects of their ablation in the earthworm embryo.

The effects of ablation on the morphology of the VNC were studied in whole mounts and in sections, and in terms of the presence or absence of a set of segmentally repeated and histologically identifiable neurons. The response of the ectoderm cells in the vicinity of an ablation was investigated by direct observation of cell movements and by observing the pattern of cells undergoing DNA synthesis in this region for up to 72h after the operation.

Staging

Experiments were carried out on embryos between 6 and 7 ·5 days old, which measure between 0 ·2 and 0 ·6 mm from the stomodeum to the most posterior point of the embryo (for details see companion paper). This stage corresponds to stage 7 in the leech embryo, Fernandez (1980).

Ablation technique

Embryos were released from cocoons into a dish of earth worm saline (see companion paper) in a small Petri dish. Each embryo was prepared for teloblast ablation by placing it in a pool of saline in a small depression on a slide coated with Sylgard resin and viewing it with a ×40 water-immersion lens on a fixed stage Zeiss microscope. A teloblast was ablated by introducing a glass micropipette into the cell and impaling and removing its nucleus. All four ectoteloblasts on either side of the embryo were identified by position (companion paper), before an ablation took place.

The effectiveness of an ablation was initially confirmed by direct observation. The nucleus occupies about two thirds of the volume of a teloblast and its removal on the end of the pipette leaves an easily identifiable ablation site, the position of which can then be checked with respect to its chain of daughter cells and the remaining teloblasts. The ablation site was observed at intervals up to 3 h after the ablation. In later experiments, a random sample of operated embryos was removed from culture and exposured to bromodeoxyuridine (BUdR) in saline. BUdR is an analogue of thymidine that is incorporated into cells during DNA replication and can be revealed immunocytochemically (for details see companion paper). Following exposure to BUdR, the teloblasts and their bandlets are labelled in a characteristic pattern (see companion paper), so the presence or absence of a teloblast and its bandlet in an operated embryo can be established.

Culture of embryos

Embryos were cultured to hatching as described previously (see companion paper). Between 30 and 40% of embryos placed in culture survived to hatching (20 –22 days in culture). This percentage was the same for animals with ablations as for controls cultured for the same period. Only those embryos that ingested large amounts of the medium and which remained virtually transparent until dorsal closure were included. Prior to staining, all animals were anaesthetised in 0 ·07 % chlorobutanol in saline and were then dissected open along the dorsal midline allowing the gut to be removed.

Staining techniques

Two staining methods were used to assess the consequences of ablation. The majority of animals were stained with toluidine blue (Altman and Bell, 1973) and were assessed for the effects of ablation on the gross morphology of the VNC. A smaller group of embryos was stained with glyoxylic acid to reveal the monoamine containing cells in the VNC, (de la Torre and Sturgeon, 1976; Stuart, 1981; Blair, 1983). Dissected embryos were immersed in the glyoxylic acid solution for 1 min, dried for 5 to 20 min under a stream of cold air, covered with a drop of immersion oil and baked for 3 min at 95 °C, before the addition of a drop of fluoromount (Gurr, BDH 36098), coverslipping and viewing with a fluorescence microscope.

8 μm wax sections were cut through posterior segments of hatchlings and stained with Mayer’s haemalum and eosin Y (Grimstone and Skaer, 1972).

Effects of ectoteloblast ablation on the morphology of the hatchling VNC

Ablations were performed on 6- to 7 ·5-day-old embryos which were subsequently cultured to hatching. The effects of unilateral and bilateral N teloblast ablation and the ablation of all four ectoteloblasts (NOPQ) on one side of the embryo were assessed and are summarised in Fig. 1 and shown in Fig. 2.

Fig. 1.

Symbolic representation of the effects of ectoteloblast ablation on the morphology of the VNC seen in whole mounts. Arrows indicate ablation sites (the effects of ablation were most extensive in the most anterior segments in which they appeared and the first of these segments are therefore referred to as the ablation site). Occasionally ganglion deficits were also seen on the nonablated side of the animal. This is indicated in the last two ganglia in the ablation region showing the effects of unilateral N teloblast ablation (B) and the last two showing the effects of unilateral ablation of N and Q teloblasts (D). This phenomenon is also indicated in the second ganglion illustrating the combined effects of unilateral ablation of all four ectoteloblasts. Anterior is up. (A) Control. The VNC of the hatchling consists of a continuous cord punctuated in each segment by a small swelling which corresponds to a single ganglion (also see Fig. 2A). Each ganglion has three pairs of bilaterally symmetrical nerves. Nerve 1 leaves the ganglion anteriorly. Nerves 2 and 3 constitute a double nerve and exit parallel and close together in the midposterior part of the ganglion. The majority of cell bodies are ventral while the dorsal surface (view seen here) is dominated by two thick fibre tracts (the Dorsal Fibre Tracts, DFTs), which run parallel either side of the midline. Note: in control animals occasional irregularly shaped ganglia were seen as a result of the normal morphogenetic movements of ganglion formation: dorsolateral cell bodies move to the ventral surface of the ganglion as segments grow (Klienenberg, 1879; K.G.S. unpublished). Irregular shapes in control animals could be easily distinguished from more dramatic irregularities seen in operated animals, see below. (B) Effects of unilateral N teloblast ablation. An ablation site was identified in five out of six animals between segments 15 and 20 and was marked by a fusion or distortion of the ganglia. In succeeding segments ganglia were often irregularly shaped, both sides of the cord could be affected and occasional aberrations in the pattern of nerves leaving the ganglia were observed. The DFTs were also sometimes distorted. However, in the most posterior segments four out of five animals had symmetrical, normallooking ganglia (see Fig. 2B). These animals also had the greatest number of segments (ranging 74 –62 compared with 46 segments in the animal with irregular posterior segments). (The sixth animal, in which an ablation site was not identified, had 71 segments). (C) Effects of bilateral N teloblast ablation. An ablation site was identified in five out of seven animals between segments 12 and 17. Again in succeeding segments ganglia were distorted and deformed. Ganglia appeared shrunken and the DFTs, normally parallel, ran together in a central groove. However, despite such deficits the most posterior segments in three out of five animals possessed symmetrical normal-looking ganglia (see Fig. 2C). Again these animals also had the most segments (ranging from 63 –60 as compared with 57 and 40 in the two animals with irregular posterior segments). (The sixth and seventh animals in this study had no conspicuous ablation site and possessed the most segments 82 and 105). (D) Effects of simultaneous ablation of all four ectoteloblasts on one side of the embryo. Two ablation sites were identified in six out of eight animals. The first in the same position as that seen following N teloblast ablation, between segments 12 and 20 and the second between segments 28 and 34. This second site marks the onset of the effects of O and P ablation (see below). Ganglia immediately succeeding the N ablation site were irregularly shaped in all animals. There appeared to be no difference between the effect of unilateral N teloblast ablation (B) and ablation of both the N and Q teloblasts. (The effects of ablating Q are predicted to occur in a similar position to those of N as they both contribute twice the number of blast cells as O and P (see below). Q is the most lateral teloblast and is unlikely to make a large contribution to the VNC). In three out of the six animals with a conspicous second ablation site, symmetrical, normal looking ganglia were observed in the most posterior segments (see Fig. 2 DI and D2). Again those animals with symmetrical posterior ganglia tended to have the most segments (ranging 69 –44 compared with 42 –47). (Two animals in which only the first ablation site was identifed had 79 and 68 segments respectively). The appearance of a second ablation site can be explained by the finding that N (and Q) contribute two blast cells for every one contributed by O and P (see companion paper). The position of the first effects of N teloblast ablation can be predicted by a simple calculation. N is bom on the third day of development (Devries, 1973) and like all the teloblasts produces approximately nine blast cells a day (see companion paper). After 6 to 7 ·5 days of development N has generated between 27 and 40/41 blast cells. The VNC does not start to form until the fourth segment (counting the prostomium and the first segment as jointly forming segment 1) and two ganglia appear to be fused in the fourth segment, constituting the suboesophageal ganglion. So an addition of 3 to account for the position of the VNC and subtraction of 1 to account for the presence of two ganglia in one segment is made. Thus, if two n blast cells contribute to each segment the effects of ablation on day 6 to 7 ·5 will be seen first between segments 14/15 and 21/22. Similarly, taking into account the morphology of the VNC, as O and P are bom on day 4 (Devries, 1973) and contribute one blast cell per segment the first effects of their ablation between day 6 and 7-5 will be expected between segments 21 and 34/35. These predictions compare well with the observed pattern of ablation sites.

Fig. 1.

Symbolic representation of the effects of ectoteloblast ablation on the morphology of the VNC seen in whole mounts. Arrows indicate ablation sites (the effects of ablation were most extensive in the most anterior segments in which they appeared and the first of these segments are therefore referred to as the ablation site). Occasionally ganglion deficits were also seen on the nonablated side of the animal. This is indicated in the last two ganglia in the ablation region showing the effects of unilateral N teloblast ablation (B) and the last two showing the effects of unilateral ablation of N and Q teloblasts (D). This phenomenon is also indicated in the second ganglion illustrating the combined effects of unilateral ablation of all four ectoteloblasts. Anterior is up. (A) Control. The VNC of the hatchling consists of a continuous cord punctuated in each segment by a small swelling which corresponds to a single ganglion (also see Fig. 2A). Each ganglion has three pairs of bilaterally symmetrical nerves. Nerve 1 leaves the ganglion anteriorly. Nerves 2 and 3 constitute a double nerve and exit parallel and close together in the midposterior part of the ganglion. The majority of cell bodies are ventral while the dorsal surface (view seen here) is dominated by two thick fibre tracts (the Dorsal Fibre Tracts, DFTs), which run parallel either side of the midline. Note: in control animals occasional irregularly shaped ganglia were seen as a result of the normal morphogenetic movements of ganglion formation: dorsolateral cell bodies move to the ventral surface of the ganglion as segments grow (Klienenberg, 1879; K.G.S. unpublished). Irregular shapes in control animals could be easily distinguished from more dramatic irregularities seen in operated animals, see below. (B) Effects of unilateral N teloblast ablation. An ablation site was identified in five out of six animals between segments 15 and 20 and was marked by a fusion or distortion of the ganglia. In succeeding segments ganglia were often irregularly shaped, both sides of the cord could be affected and occasional aberrations in the pattern of nerves leaving the ganglia were observed. The DFTs were also sometimes distorted. However, in the most posterior segments four out of five animals had symmetrical, normallooking ganglia (see Fig. 2B). These animals also had the greatest number of segments (ranging 74 –62 compared with 46 segments in the animal with irregular posterior segments). (The sixth animal, in which an ablation site was not identified, had 71 segments). (C) Effects of bilateral N teloblast ablation. An ablation site was identified in five out of seven animals between segments 12 and 17. Again in succeeding segments ganglia were distorted and deformed. Ganglia appeared shrunken and the DFTs, normally parallel, ran together in a central groove. However, despite such deficits the most posterior segments in three out of five animals possessed symmetrical normal-looking ganglia (see Fig. 2C). Again these animals also had the most segments (ranging from 63 –60 as compared with 57 and 40 in the two animals with irregular posterior segments). (The sixth and seventh animals in this study had no conspicuous ablation site and possessed the most segments 82 and 105). (D) Effects of simultaneous ablation of all four ectoteloblasts on one side of the embryo. Two ablation sites were identified in six out of eight animals. The first in the same position as that seen following N teloblast ablation, between segments 12 and 20 and the second between segments 28 and 34. This second site marks the onset of the effects of O and P ablation (see below). Ganglia immediately succeeding the N ablation site were irregularly shaped in all animals. There appeared to be no difference between the effect of unilateral N teloblast ablation (B) and ablation of both the N and Q teloblasts. (The effects of ablating Q are predicted to occur in a similar position to those of N as they both contribute twice the number of blast cells as O and P (see below). Q is the most lateral teloblast and is unlikely to make a large contribution to the VNC). In three out of the six animals with a conspicous second ablation site, symmetrical, normal looking ganglia were observed in the most posterior segments (see Fig. 2 DI and D2). Again those animals with symmetrical posterior ganglia tended to have the most segments (ranging 69 –44 compared with 42 –47). (Two animals in which only the first ablation site was identifed had 79 and 68 segments respectively). The appearance of a second ablation site can be explained by the finding that N (and Q) contribute two blast cells for every one contributed by O and P (see companion paper). The position of the first effects of N teloblast ablation can be predicted by a simple calculation. N is bom on the third day of development (Devries, 1973) and like all the teloblasts produces approximately nine blast cells a day (see companion paper). After 6 to 7 ·5 days of development N has generated between 27 and 40/41 blast cells. The VNC does not start to form until the fourth segment (counting the prostomium and the first segment as jointly forming segment 1) and two ganglia appear to be fused in the fourth segment, constituting the suboesophageal ganglion. So an addition of 3 to account for the position of the VNC and subtraction of 1 to account for the presence of two ganglia in one segment is made. Thus, if two n blast cells contribute to each segment the effects of ablation on day 6 to 7 ·5 will be seen first between segments 14/15 and 21/22. Similarly, taking into account the morphology of the VNC, as O and P are bom on day 4 (Devries, 1973) and contribute one blast cell per segment the first effects of their ablation between day 6 and 7-5 will be expected between segments 21 and 34/35. These predictions compare well with the observed pattern of ablation sites.

Fig. 2.

Effects of unilateral N teloblast ablation (B, UL N-T), bilateral N teloblast ablation (C, BL N-T), and unilateral ablation of all 4 ectoteloblasts (DI and D2, UL-ET), on the hatchling VNC compared with a control animal (A, control). Anterior segments including ablation sites (see Fig. 1) are shown together with the most posterior segments of each animal (except in 2 and 3 in which the last 10 –20 segments were removed for sectioning). Ablation sites are indicated with an arrowhead. Control (A) shows normal shape of hemiganglia in an animal cultured from 6 to 7 ·5 days to hatching. Unilateral N teloblast ablation (B) is marked by ganglion depletion and distortion which in this case switches sides, but is then confined to one side. The posterior segments of this animal, shown below, appear regular. Bilateral N teloblast ablation (C) results in depletion of both sides of the ganglia, although again, posterior segments appear regular. Two examples of the effects of ablating all four ectoteloblasts on one side of the embryo are shown. In DI two ablation sites are indicated, the first is due to ablation of the N (and possibly Q) teloblasts, while the second is due to ablation of the O and P teloblasts (see Fig. 1). In this animal, segments posterior to the second ablation site are all irregular. In D2 only the first ablation site is indicated. The second site in this animal took the form of a reduction and slight distortion of the ganglia and this effect can be seen in the first eight of the posterior segments shown here. In contrast, the remaining segments in this animal are regularly shaped. Scale bar = 200 μm.

Fig. 2.

Effects of unilateral N teloblast ablation (B, UL N-T), bilateral N teloblast ablation (C, BL N-T), and unilateral ablation of all 4 ectoteloblasts (DI and D2, UL-ET), on the hatchling VNC compared with a control animal (A, control). Anterior segments including ablation sites (see Fig. 1) are shown together with the most posterior segments of each animal (except in 2 and 3 in which the last 10 –20 segments were removed for sectioning). Ablation sites are indicated with an arrowhead. Control (A) shows normal shape of hemiganglia in an animal cultured from 6 to 7 ·5 days to hatching. Unilateral N teloblast ablation (B) is marked by ganglion depletion and distortion which in this case switches sides, but is then confined to one side. The posterior segments of this animal, shown below, appear regular. Bilateral N teloblast ablation (C) results in depletion of both sides of the ganglia, although again, posterior segments appear regular. Two examples of the effects of ablating all four ectoteloblasts on one side of the embryo are shown. In DI two ablation sites are indicated, the first is due to ablation of the N (and possibly Q) teloblasts, while the second is due to ablation of the O and P teloblasts (see Fig. 1). In this animal, segments posterior to the second ablation site are all irregular. In D2 only the first ablation site is indicated. The second site in this animal took the form of a reduction and slight distortion of the ganglia and this effect can be seen in the first eight of the posterior segments shown here. In contrast, the remaining segments in this animal are regularly shaped. Scale bar = 200 μm.

All three kinds of ablation experiment show that most animals possessed apparently normal ganglia in their posterior segments despite the effects of ablation seen more anteriorly. In many cases the posteriorly normal animals were those with the greatest number of segments (Fig. 1 legend). The success of all ablations monitored directly or by exposure to BUdR at intervals after ablation (see below), also support the view that these animals are able to reconstitute themselves following ectoteloblast ablation, as proposed by Devries (1974).

To test this conclusion in detail, the effects of ectoteloblast ablation were monitored (a) in a stereotypical set of histochemically identifiable cells which differentiates in the normal development of the VNC, (b) in the structure of the most posterior ganglia of experimental animals, and (c) in the cells immediately surrounding the ablated teloblast.

Normal pattern of monoamine-containing cells in hatchling ganglia

Glyoxylic acid staining of the hatchling VNC reveals a pair of bilaterally arranged cells just anterior to the exit of nerve 2 (see Fig. 1), and a set of bilaterally symmetrical fibres in the dorsal fibre tracts (DFTs), Fig. 3. The position of the two cells corresponds to the group C neurons identified by Rude (1969) in L. lerristris. To distinguish them from the c interneuron in the leech I have called them cg cells. Rude (1969) observed that the fibres stained in the DFTs were processes put out by sensory cells located in the body wall. The cg cells and fibres in the DFTs are the first monoamine-containing cells and fibres to stain in the developing cord, appearing in anterior ganglia between 20 and 22 days of development (hatchlings emerge on day 28).

Fig. 3.

Symbolic representation of the effects of ectoteloblast ablation on monoamine containing cells in the hatchling ganglion. Segment numbers are notional and indicate that two n blast cells are incorporated for every one o and p blast cell (see Fig. 1 legend). The normal staining pattern (two cg cells (black circles) in each ganglion, one on either side and heavy staining of the DFTs (blacked parallel lines)), is represented in the first three ganglia. The next three ganglia represent the N teloblast ablation site (again there was no difference between the effects of N ablation and ablation of N and Q). N teloblast ablation does not affect the pattern of cg cells or staining in the DFTs, although the irregular shape of ganglia at the N ablation site can distort the DFTs. The last four ganglia represent the effects of O and P ablation (see Fig. 1 legend). One cg cell in each ganglion is deleted. The remaining cg cell occasionally appeared on the ablated side of the animal. Staining in the DFTs is depleted on both sides of the animal. Depletion on the non ablated side could reflect the contralateral projection of fibres which originate on the ablated side.

Fig. 3.

Symbolic representation of the effects of ectoteloblast ablation on monoamine containing cells in the hatchling ganglion. Segment numbers are notional and indicate that two n blast cells are incorporated for every one o and p blast cell (see Fig. 1 legend). The normal staining pattern (two cg cells (black circles) in each ganglion, one on either side and heavy staining of the DFTs (blacked parallel lines)), is represented in the first three ganglia. The next three ganglia represent the N teloblast ablation site (again there was no difference between the effects of N ablation and ablation of N and Q). N teloblast ablation does not affect the pattern of cg cells or staining in the DFTs, although the irregular shape of ganglia at the N ablation site can distort the DFTs. The last four ganglia represent the effects of O and P ablation (see Fig. 1 legend). One cg cell in each ganglion is deleted. The remaining cg cell occasionally appeared on the ablated side of the animal. Staining in the DFTs is depleted on both sides of the animal. Depletion on the non ablated side could reflect the contralateral projection of fibres which originate on the ablated side.

In the last 6 to 10 segments of the hatchling VNC, the pattern of staining was weaker than in the anterior segments. This was observed in control and experimental animals and is probably due to the later differentiation of the more posterior segments. As the effects of ablation are predicted to occur in more anterior segmerits it was anticipated that weaker staining in the posterior segments would not interfere with the assessment of ablation effects.

The cg cells and the fibres stained in the DFTs have a bright blue-green fluorescence indicative of the presence of dopamine, while other groups in the adult and in the anterior segments of the hatchling have a bright yellow fluorescence characteristic of serotonin (de la Torre and Sturgeon, 1976; Stuart, 1981).

Unilateral N teloblast ablation

Ablation sites in the three animals in this study were found between segments 13 and 18. The DFTs were distorted in all animals and in one case a cg cell was missing at the ablation site. However, these effects were confined to one or two ganglia only (Fig. 3). The ablation site in all three animals was followed by ganglia in which both cg cells were present and in which staining in the DFTs was normal. In all, 258 such ganglia were examined.

These findings suggest either that N teloblast ablation is followed by almost complete regulation (only one cg cell was missing and in only one animal), or that the N teloblast does not give rise to the cg cell or the cells that project fibres into the DFTs, and that the distortion effects are caused simply by the absence of other cells in the ganglia. In the next experiment, all four ectoteloblasts on one side of the embryo were removed. As the effects of N teloblast ablation are seen in segments anterior to those affected by ablation of the O and P teloblasts (see above), it is possible to distinguish the effects of N and, O and P ablation in these animals.

Unilateral ablation of all four ectoteloblasts

The effects of this ablation are summarised in Fig. 3. The site of N teloblast ablation was identified in four out of the six animals. In the two remaining animals no disturbance in the pattern was seen in the first 27 and 30 segments, respectively.

Segments in the region posterior to the N teloblast ablation site contained both eg cells and normal staining in the DFTs. This region lasted for (on average) 14 segments. A second ablation site, between segments 27 and 30, was distinguished by the absence of a cg cell and disturbance and depletion of the fibres in the DFTs (Figs 3,4).

Fig. 4.

Example of the disturbance in the pattern of cg cells (arrowheads) and fibres stained in the DFTs seen at the second ablation site, in an animal with unilateral ablation (right) of all four ectoteloblasts (stained with glyoxylic acid). One eg cell is missing on the ablated side in the most anterior ganglion and this is accompanied by depletion of the DFTs on both sides. It is difficult to draw segment boundaries beyond this point as some segments are fused. Staining in the DFTs continues to be depleted on both sides and is erratic on the ablated side. Three cg cells appear on the ablated side and five on the non ablated side. No more cg cells appeared on the ablated side, while cg cells were present on the non ablated side to the end of the animal. Midline is indicated by a broken line, anterior is up and scale bar = 25 μm. This figure and Fig. 5 are negative images, in which the fluorescently stained structures appear black.

Fig. 4.

Example of the disturbance in the pattern of cg cells (arrowheads) and fibres stained in the DFTs seen at the second ablation site, in an animal with unilateral ablation (right) of all four ectoteloblasts (stained with glyoxylic acid). One eg cell is missing on the ablated side in the most anterior ganglion and this is accompanied by depletion of the DFTs on both sides. It is difficult to draw segment boundaries beyond this point as some segments are fused. Staining in the DFTs continues to be depleted on both sides and is erratic on the ablated side. Three cg cells appear on the ablated side and five on the non ablated side. No more cg cells appeared on the ablated side, while cg cells were present on the non ablated side to the end of the animal. Midline is indicated by a broken line, anterior is up and scale bar = 25 μm. This figure and Fig. 5 are negative images, in which the fluorescently stained structures appear black.

The sixth animal, in which the first ablation site was not observed until segment 22, showed a different pattern of staining to that observed in the other five animals and the segments in which ablation effects were first seen suggest that this embryo was older (aged 7 ·5 days) than the other embryos when the ablations were performed. The pattern of cg cell deletions in this animal are shown in Fig. 5 and summarised in Fig. 6. It is clear in Fig. 5 that the DFTs are strongly and uniformly stained while some cg cells are missing.

Fig. 5.

Glyoxylic acid staining in ganglia in segments 24, 25, 26 and the anterior portion of 27, in an animal in which all four ectoteloblasts on one side were removed (right), also see Fig. 6. In all other animals with this ablation, a disturbance in the pattern of cg cells and DFTs occurred simultaneously and was observed in more posterior segments. In this animal, the absence of one cg cell in five more anterior segments (three of which are shown here) may be due to o-bandlet damage (see text), cg cells are indicated with arrowheads and the midline by a broken line. Anterior is up and scale bar = 25 μm.

Fig. 5.

Glyoxylic acid staining in ganglia in segments 24, 25, 26 and the anterior portion of 27, in an animal in which all four ectoteloblasts on one side were removed (right), also see Fig. 6. In all other animals with this ablation, a disturbance in the pattern of cg cells and DFTs occurred simultaneously and was observed in more posterior segments. In this animal, the absence of one cg cell in five more anterior segments (three of which are shown here) may be due to o-bandlet damage (see text), cg cells are indicated with arrowheads and the midline by a broken line. Anterior is up and scale bar = 25 μm.

Fig. 6.

Symbolic representation of the pattern of glyoxylic acid staining in the vnc of one animal with all four ectoteloblasts removed on one side. The pattern in ganglia 22 –27 may be explained by local damage to the o bandlet during N teloblast ablation. Ganglia were normal (see above) until segment 21 after which one cg cell was missing in each ganglion (except segment 24) through to segment 27. The staining in the DFTs, however, appeared normal in these ganglia (see text). Ganglia in segments 28 to 37 were again normal and the second ablation site (corresponding to O and P teloblast ablation), began in segment 38.

Fig. 6.

Symbolic representation of the pattern of glyoxylic acid staining in the vnc of one animal with all four ectoteloblasts removed on one side. The pattern in ganglia 22 –27 may be explained by local damage to the o bandlet during N teloblast ablation. Ganglia were normal (see above) until segment 21 after which one cg cell was missing in each ganglion (except segment 24) through to segment 27. The staining in the DFTs, however, appeared normal in these ganglia (see text). Ganglia in segments 28 to 37 were again normal and the second ablation site (corresponding to O and P teloblast ablation), began in segment 38.

One explanation for the unique pattern of staining in this embryo springs from the observation that in one case one cg cell was missing at the N teloblast ablation site. As the o bandlet runs alongside the N teloblast, the ablation of the N teloblast may distort the o bandlet to such an extent that the effect of the loss of O teloblast progeny may be seen earlier than predicted. If this is the case, it suggests that the o lineage gives rise to the cg cell while the p lineage gives rise to the cells whose processes appear in the DFTs.

Comparison of the average number of segments of animals following the each type of ablation (Table 1) suggests that ablation of all four ectoteloblasts may result in the early termination of segment production. Further, in animals with unilateral and bilateral N teloblast ablation, segments furthest from the ablation site appeared the most regular. It could be argued, therefore, that the sustained deficits seen in the glyoxylic-acid-treated animals with unilateral ablation of all four ectoteloblasts are not recovered because reconstitutive processes are only manifest in segments after a sufficient developmental delay. The small number of segments between the second ablation site and the last segment in such animals may then account for the consistent deficits observed.

Table 1.
graphic
graphic

Does effective reconstitution occur in posterior segments: the structure of posterior ganglia in operated animals with many segments

Table 2 summarises observations made of transverse light microscope sections of the last 7 to 13 segments of four animals with unilateral N teloblast ablations which had between 65 and 105 segments. Observations made of transverse sections of the last 8 to 10 segments of three control animals cultured for the same period (to hatching) are also presented. Normal ganglia are bilaterally symmetrical. Ganglia were scored as irregular if they had marked asymmetries or were distorted, for example, by large gaps between the cells. Sections were scored for whether these irregularities occurred on the left, right or both sides of the ganglia.

Table 2.

Transverse sections through the posterior segments of animals with unilateral N teloblast ablation scored for the regularity of their ganglia

Transverse sections through the posterior segments of animals with unilateral N teloblast ablation scored for the regularity of their ganglia
Transverse sections through the posterior segments of animals with unilateral N teloblast ablation scored for the regularity of their ganglia

60 –70 % of sections scored in the experimental animals had irregularly shaped ganglia, as opposed to 10 –25 % in controls. The effect of ablation was largely confined to the ablated side in only one animal and in the other three ablation effects were seen consistently on both sides of the ganglia. All ganglia in experimental animals had at least one section in which an irregularity appeared.

All ganglia in experimental animals were consistently smaller than those of controls. As the extent to which embryos develop in culture can vary, the width of the ganglia was measured in four animals (two controls and two experimentáis) which had developed to approximately the same extent. To allow for any reduction in size due to the overall smaller size of an animal the width of the ganglion in each section was expressed as a proportion of the width of the whole segment at that point. The width of the ganglion (in sections that passed though approximately the same position in the ganglion) was measured in eighteen sections of control and experimental animals. It was found that the proportion of the width of the segment occupied by the ganglia ranged from 0 ·25 to 0 ·31 in experimental animals and 0 ·36 to 0 ·51 in controls. This amounts to a difference of, on average, 18 μm (approximately a quarter of the diameter of the VNC, Fig. 7).

Fig. 7.

Transverse sections through the posterior segments of hatchlings raised in culture from 6 to 7 ·5 days of development. Sections A and B were chosen as they pass through the same region in the ganglion. A is from a control animal, which was simply removed from its cocoon and cultured to hatching. B and C are from animals with unilateral N teloblast ablation. Ganglia in experimental animals were consistently smaller than in controls (B) (see text) and cells sometimes looked jumbled, as if supporting cells such as glia were missing (C). No dramatic depletions of cells (C) were seen in controls. Nerves 1, 2, and 3 and the three giant fibres in the DFTs were identified in all ganglia of experimental animals. Ventral nerve cord, vnc, gut,g, setae, s and nepridia, np. Ventral is down and scale bar = 40 μm.

Fig. 7.

Transverse sections through the posterior segments of hatchlings raised in culture from 6 to 7 ·5 days of development. Sections A and B were chosen as they pass through the same region in the ganglion. A is from a control animal, which was simply removed from its cocoon and cultured to hatching. B and C are from animals with unilateral N teloblast ablation. Ganglia in experimental animals were consistently smaller than in controls (B) (see text) and cells sometimes looked jumbled, as if supporting cells such as glia were missing (C). No dramatic depletions of cells (C) were seen in controls. Nerves 1, 2, and 3 and the three giant fibres in the DFTs were identified in all ganglia of experimental animals. Ventral nerve cord, vnc, gut,g, setae, s and nepridia, np. Ventral is down and scale bar = 40 μm.

These observations show that the apparently regular ganglia observed in whole-mount preparations of unilateral N teloblast ablated animals are in fact abnormal and these abnormalities persist even in the most posterior segments. The apparent regularity observed in whole mounts might to some extent be accounted for by the ‘sharing out’ of the effects of ablation between the two sides.

Effects of ablation on the cells neighbouring the ablated teloblast

The initial effects of ectoteloblast ablation were investigated by direct observation of the movement of cells close to the ablated teloblast and by examining the pattern of cells synthesising DNA (as revealed by BUdR incorporation, see Materials and methods), in this region.

N teloblast ablation was taken as an example of the effects of ectoteloblast ablation as N lies at the medial edge of the germ band and the cells surrounding it are more easily defined than those associated with O and P (Fig. 8). In addition, at the magnifications used, the N teloblast lies in the same field of view as its contralateral homologue, so facilitating comparison between them.

Fig. 8.

Sketches of cell positions before N teloblast ablation and 1 h and 3 h after ablation. Four kinds of cells are found in the region of the N teloblast at 6 to 7 ·5 days: medial ciliated cells, ectoderm cells which constitute a sheet in which the ectoteloblasts and their bandiets are embedded, the progeny of the N teloblast (n blast cells) and the progeny of the O teloblast, some of which have already undergone their first division. Scale bar = 10 μm.

Fig. 8.

Sketches of cell positions before N teloblast ablation and 1 h and 3 h after ablation. Four kinds of cells are found in the region of the N teloblast at 6 to 7 ·5 days: medial ciliated cells, ectoderm cells which constitute a sheet in which the ectoteloblasts and their bandiets are embedded, the progeny of the N teloblast (n blast cells) and the progeny of the O teloblast, some of which have already undergone their first division. Scale bar = 10 μm.

Direct observations

The ablation site of the N teloblast was observed in ten embryos for 3h following ablation. Embryos were monitored at intervals of 30 min throughout this period and the positions of cells were sketched (Fig. 8). In most embryos the debris of the ablated N teloblast was ejected after 2h.

30 min after ablation

After 30 min the hole left by the ablation of the N teloblast appeared as a dark sphere surrounded by ectoderm cells. In two cases, one of these cells appeared swollen and and may have died subsequently. As the teloblast was ablated by removing its nucleus (see Materials and methods) neighbouring cells should not have been directly affected. The subsequent death of a neighbouring cell might suggest that it is in close contact with the teloblast, perhaps playing a supportive role similar to the cap cells associated with grasshopper neuroblasts (Doe and Goodman, 1985). Ablation of the N teloblast may also disturb the cells in the o-bandlet (see above).

2 and 3 h after ablation

After 2h the cells between the ablation site and the ciliated cells had moved into the hole (see Fig. 8). By 3 h these cells had filled the hole and appeared to be in contact with the small ectoderm cells next to the o bandlet. This movement may be a passive filling in of available space by neighbouring cells, while its direction is perhaps dictated by the normal lateral spread of the ciliated cells over the n bandlets as the embryo develops.

DNA synthesis in cells in the region of the ablation

All embryos were exposed to BUdR in saline for 6h, either immediately after ablation (first group) or for the last 6h of a longer culture period. Exposure to BUdR for a given duration produces a characteristic pattern of labelling in the teloblasts and their immediate progeny (companion paper). This pattern was used to confirm the presence of only seven ectoteloblasts in each animal with unilateral N teloblast ablation up to 48 h after the operation (see below). The site of ablation in these embryos is defined as the region directly opposite the intact, contralateral N teloblast. It is assumed that the ectodermal sheet, in which the ectoteloblasts are set, proliferates alongside and is displaced backwards along with the ectoteloblasts as the embryo develops (although the O and P teloblasts and their bandlets are ‘displaced’ at a different rate relative to the N and Q teloblasts (companion paper). Thus, the ablation site of the N teloblast is also displaced backwards and appears opposite the intact N teloblast. The distance between this site and the end of the discontinued n bandlet increases with time and this region was also studied.

6 h after ablation

In five (out of eight) embryos pulsed for just 6h following ablation no labelled cells appeared in the region of the ablation. In the three others four to eight strongly labelled cells were observed at a level just posterior and contralateral to the intact N teloblast. However, in two embryos, a few labelled cells were also seen close to the intact N teloblast.

10 h after ablation

10 h after ablation the three embryos examined had between four and eight labelled cells in the ablation region. These cells appeared to be ectoderm cells which were normally observed surrounding the N-teloblast and at the fringes of the ectoteloblast bandlets. However, three or four labelled cells were also seen just posterior to the intact N teloblast. (It was also noted that the configuration of cells in the o bandlets was normal -as judged by the symmetry of the first division of the o blast cells (see companion paper) - despite the absence of the normally adjacent n bandlet).

24 h after ablation

Five embryos (including two with unilateral ablation of all four ectoteloblasts) were examined at 24 h. In each embryo, a group of labelled cells was found in the ablation region, opposite the intact, contralateral N teloblast. While the frequently observed three or four labelled cells were seen again in the vicinity of the intact N teloblast, between eight and ten such cells were seen on the ablated side of the embryo (see Fig. 9).

Fig. 9.

(A) The line of cells seen 24 h after unilateral N teloblast ablation. The intact N teloblast (N), its bandiet and the line of cells on the contralateral side of an embryo (exposed to BUdR for 6h) can be seen. The arrowheads indicate the gap between the n-bandlet (1) and the beginning of the line (2). A labelled division in the line is indicated by a small arrowhead. Anterior is up and scale bar = 20 μm. (B) Tracing of A highlighting salient features: the intact N teloblast and its bandiet, and the contralateral line of cells in which labelled divisions are taking place, the gap (between 1 and 2) filled in by ciliated cells, and the end of the n bandiet on the ablated side (arrowhead 1). Scale as in B.

Fig. 9.

(A) The line of cells seen 24 h after unilateral N teloblast ablation. The intact N teloblast (N), its bandiet and the line of cells on the contralateral side of an embryo (exposed to BUdR for 6h) can be seen. The arrowheads indicate the gap between the n-bandlet (1) and the beginning of the line (2). A labelled division in the line is indicated by a small arrowhead. Anterior is up and scale bar = 20 μm. (B) Tracing of A highlighting salient features: the intact N teloblast and its bandiet, and the contralateral line of cells in which labelled divisions are taking place, the gap (between 1 and 2) filled in by ciliated cells, and the end of the n bandiet on the ablated side (arrowhead 1). Scale as in B.

In two of these embryos, the cells in the region of the ablation were positioned in single file along the anterior posterior axis of the embryo and in one. two labelled mitotic figures were observed opposite the intact N teloblast. Fig. 9. This suggests that the line of cells is formed by the division of the ectoderm cells observed in this region. The linear arrangement of the new cells is not what would be expected if this was simply a response to local damage to the surrounding ectoderm cells. No new teloblasts were observed.

Anterior to the line of cells was an area in which no cells were labelled. This area was occupied by ciliated cells. Anterior to these cells was the posterior end of the discontinued bandlet. The cells in the bandlet in all five embryos showed similar patterns of labelling to the n blast cells in the contralateral bandlet at the equivalent level (Fig. 9). No additional proliferation was observed in the discontinued bandlet.

48 h after ablation

After 48 h a similar gap between the labelled ectoderm cells and the discontinued n bandlet was observed in all five embryos examined (two of which had undergone unilateral ablation of all four ectoteloblasts). In each of these embryos, the labelled cells formed a strip two or three cells wide, running along the anterior/posterior axis of the embryo. In embryos in which all four ectoteloblasts on one side had been ablated, the position of the gap was not the same in each bandlet: the o and p bandlets continued posteriorly, past the gap in the n bandlet and, when the o and p bandlets ceased (which they did together), a second gap was seen. This was abutted posteriorly by a group of strongly labelled ectoderm cells (Fig. 10). This observation is consistent with the more posterior position of the O and P teloblasts and with the observation that the N teloblast contributes two n blast cells for every one o and p blast cell (see companion paper).

Fig. 10.

(A) An embryo 48 h after ablation of all four ectoteloblasts on one side of the embryo (left) (exposed to BUdR for 6h). The arrowheads marked 1, indicate the gap between the end of the n-bandlet and a line of labelled cells which are first seen 24 h after N teloblast ablation. The arrowheads marked 2, indicate the gap between the end of the o and p bandlets and a band of strongly labelled ectoderm cells, which may generate a similar line of cells to that observed following N teloblast ablation. Anterior is up and scale bar = 40 μm. (B) Tracing of A highlighting salient features: the gap between the end of the n bandlet (hemiganglia primordia) and the next group of labelled cells in line with the n bandlet, on the ablated side of the embryo (between arrowheads 1), and a second gap, in the o and p bandlets between the ends of these bandlets and the next group of labelled cells, (between arrowheads 2). The outlining of groups of cells in this figure does not imply that cells in these groups are all clonally related. Scale as in A.

Fig. 10.

(A) An embryo 48 h after ablation of all four ectoteloblasts on one side of the embryo (left) (exposed to BUdR for 6h). The arrowheads marked 1, indicate the gap between the end of the n-bandlet and a line of labelled cells which are first seen 24 h after N teloblast ablation. The arrowheads marked 2, indicate the gap between the end of the o and p bandlets and a band of strongly labelled ectoderm cells, which may generate a similar line of cells to that observed following N teloblast ablation. Anterior is up and scale bar = 40 μm. (B) Tracing of A highlighting salient features: the gap between the end of the n bandlet (hemiganglia primordia) and the next group of labelled cells in line with the n bandlet, on the ablated side of the embryo (between arrowheads 1), and a second gap, in the o and p bandlets between the ends of these bandlets and the next group of labelled cells, (between arrowheads 2). The outlining of groups of cells in this figure does not imply that cells in these groups are all clonally related. Scale as in A.

72 h after ablation

After 72 h the ectoteloblasts could only be identified in one of seven embryos with unilateral N teloblast ablation. In the remaining embryos the ectoteloblasts and the posterior ends of their bandlets had sunk below a covering ectodermal sheet and the spreading band of ciliated cells. The strip of labelled cells observed earlier could not, therefore, be identifed by its unique position relative to the intact teloblasts and their bandlets.

However, in six embryos a gap was observed in the n bandlet (the most median strip of cells). As the anterior end of the n bandlet had formed into groups in a segmentally repeated fashion (as defined by the position of developing nephidia and underlying fibres, see companion paper) the position of this gap could be found in terms of the number segments from the stomodeum. This position was consistent with the position of ablation sites identified in the hatchling VNC.

The findings presented in this paper show that despite the apparent regularity of the ganglia in the posterior segments of experimental animals, sustained deficits are indeed produced by the ablation of the ectoteloblasts. This is demonstrated by deficits in the pattern of monoamine-containing cells and by a disorganisation and reduction in size of the most posterior segments of operated animals.

It could be argued that the deficits seen in the pattern of monoamine-containing cells in animals with unilateral ablation of all four ectoteloblasts, are sustained because there are only a small number of segments between the ablation site and the last segment of the animal; that is, that reconstitutive processes are only manifest after a ‘developmental delay’, in segments some distance from the ablation site. The disorganisation of the most posterior ganglia seen in sections of operated animals with many segments argues against this idea. Nonetheless, to exclude any form of regulation it would be necessary to raise operated animals to maturity before assessing the results of teloblast ablations. The work reported here shows that to the formation of the hatchling the effects of ectoteloblast ablation are similar to those observed in the glossipho-niid leech Helobdella triserialis (Blair, 1983), in that ablation of specific ectoteloblasts results in the loss of a specific population of nerve cells.

Clearly the cell divisions observed in the region of the ablated teloblast are not able to replace the lost teloblast lineages (at least up to the hatchling stage). On the other hand, their presence may mitigate the effects of ectoteloblast ablation: there is a striking difference between the dramatic depletion of the VNC seen in the leech after ablation of an N teloblast or of the OPQ precursor cell (Blair and Weisblat, 1982) and the effects described here.

Blair and Weisblat (1982) propose that the progeny of the N teloblast provide a focus or ‘organising centre’ for the progeny of the o,p and q blast cells which migrate ventromedially in order to contribute to the VNC (Weisblat et al. 1984; Torrence and Stuart, 1986). As shown by HRP injection of the N teloblast, (companion paper) and by these ablation studies, it is likely that other teloblasts also contribute to the VNC in the earthworm, thus it may be that the line of cells seen following N teloblast ablation provide a substitute focus for the migrating progeny of the o,p and q blast cells, facilitating contact between the remaining cells in the ganglion and hence their more normal differentiation.

A similar proliferation of the cells in the ectodermal sheet but in response to ablation of micromeres rather than teloblasts has been reported in the leech (Ho and Weisblat, 1987). Injection of HRP into the micromeres reveals that each contributes a specific ‘territory’ in the sheet of ectoderm that covers the ectoteloblast bandlets. Ablation of one of the micromeres results in the expansion of neighbouring territories by the proliferation of cells derived from other micromeres. Thus, it may be that ablation of the ectoteloblasts in the earthworm embryo elicits a similar response in the surrounding ectoderm cells. The large teloblasts of H. triserialis are not set into a sheet of ectoderm so the cells are simply not available to respond to teloblast ablation as they are in earthworm.

The sheet of ectoderm cells that surrounds the ectoteloblasts in the earthworm embryo is thought to be derived from small cells budded off from the N and OPQ cells (n ′,opq ′) which Devries (1973,a) has observed migrating with their parent cells from the dorsal to the ventral surface during gastrulation. In the glossi-phoniid leech, the opq ′ cells lie initially over the ectoteloblast bandlets and later contribute to the epithelium lying over the dorsal midline (Ho and Weisblat, 1987). The n ′ cells, on the other hand, have been shown to contribute a strip of epithelial cells overlapping the dorsal edge of the germ band and later these cells also contribute to the dorsal epithelium. The position of the progeny of the n ′ and opq ′ cells in leech and earthworm embryos therefore appears to be different. Indeed, while in the leech embryo, the ectoteloblast bandlets are covered over by cells derived from other cells in the micromere population, these cells do not appear to be interspersed between the bandlets, which lie in a tightly packed formation (Weisblat et al. 1980). A similar arrangement has also been described in the medical leech (Fernandez and Stent, 1982) and in this embryo there also appear (from the published Figures) to be no ectoderm cells associated with the ectoteloblast bandlets. If the presence of a set of ectoderm cells surrounding the ectoteloblasts were associated with albumenotrophic feeding, perhaps as part of a general expansion of the ectodermal sheet, which forms a provisional epithelium to contain ingested albumen, then the small ectoteloblasts in the albumenotrophic medical leech might be expected to be associated with a similar population of cells. The fact that they are not suggests that the position of this population of ectoderm cells may be special to the earthworm.

The proliferation of cells in the ectodermal sheet in the earthworm embryo in response to N teloblast ablation appears to correspond to the phenomenon of ‘dédoublement’ described by Devries (1973). In Devries’s (1974) experiments, UV irradiation was followed by the proliferation of ectoderm cells posterior to the irradiated zone. Devries describes the process of ‘dédoublement’ as the splitting of this sheet into two layers, an outer superficial layer confluent with the normal covering ectoderm, and an inner layer corresponding to that normally generated by the ectoteloblasts. The cells posterior to the irradiated region were observed to form a pair of ‘germ bands’ which, while lacking ectoteloblasts and bandlet-like structures, differentiated into tissues which Devries assumes to be derived from the ectoteloblasts: the VNC, circular muscles and epidermis.

While the findings presented here are in agreement with Devries’s observations of the initial response to ablation, the results differ on the fundamental question of whether the new cells acquire the specific fates of cells produced by the ectoteloblasts. The experiments described here show that up to the hatchling stage, discrete populations of nerve cells are lost when specific ectoteloblasts are ablated. However, the fate of the ectoderm cells produced in response to ablation has yet to be determined and it may be that the epidermal progeny of the teloblasts, with which the latter merge during the normal formation of the epidermis, are compensated for, while nerve cells are not replaced. This would be radically different from the leech embryo, in which the loss of the epidermal progeny of teloblasts is compensated for only by the proliferation of the progeny of remaining ectoteloblasts (Blair and Weisblat, 1984). If the micromere-derived ectoderm cells were shown (by injection of lineage tracers) to contribute to structures normally derived from the ectoteloblasts, the further study of this phenomenon might lead us to an understanding of how it is that annelids such as the earthworm are able to regenerate as adults, while others, such as the leech, do not have this capacity.

I thank my supervisor Michael Bate for his enthusiasm and support. I am also grateful to Helen Skaer and Jim Truman for valuable discussions and to Maggie Bray, Mick Day, Dennis Unwin and Bill Westley for excellent technical assistance. Joe Gentle master carpenter in the Zoology Department in Cambridge, made me a beautiful worm box, which is still in use today. This work was supported by a studentship from the Medical Research Council.

Altman
,
J. S.
&
Bell
,
E.
(
1973
).
A rapid method for demonstration of nerve cell bodies in invertebrate CNS’s
.
Brain Res
.
63
,
482
489
.
Anderson
,
D. T.
(
1973
).
Embryology and Phylogeny in Annelids and Arthropods
.
Pergamon Press, N.Y
.
Blajr
,
S. S.
(
1983
).
Blastomere ablation and the developmental origin of identified monoamine-containing neurons in the leech
.
Devi Biol
.
95
,
65
72
.
Blair
,
S. S.
&
Weisblat
,
D. A.
(
1982
).
Ectodermal interaction during neurogenesis in the glossiphoniid leech Helobdella triserialis
.
Devi Biol
.
91
,
64
72
.
Blair
,
S. S.
&
Weisblat
,
D. A.
(
1984
).
Cell interactions in the developing epidermis of the leech Helobdella triseriahs
.
Devi Biol
.
101
,
318
325
.
Dawydoff
,
C.
(
1942
).
Etude sur l’embryologie des Naididae indochinos
.
Arch. Zool. Exp. gen. Notes and revue
.
81
,
173
194
.
De La Torre
,
J. C.
&
Sturgeon
,
J. W.
(
1976
).
A methodological approach to rapid and sensitive monoamine histofluresence using a modified glyoxylic acid technique. The SPG method
.
Histochemistry
49
,
81
93
.
Devries
,
J.
(
1973
).
La formation et la destinee des feuillets embryonnaires chez le lombricien Eisenia foenda (Annelide oligochete)
.
Arch. Anat. Microsc. Morph, exp
.
62
,
15
38
.
Devries
,
J.
(
1974
).
Le mesoderme, feuillet directeur de l’embryogenese chez le lombricien Eisenia foetida. III. La determination des ectoteloblastes. Acta embryologiae experimentalis
.
Issue
2
,
181
190
.
Doe
,
C. Q.
&
Goodman
,
C. S.
(
1985
).
Early events in insect neurogenesis. I. Developmental and sequential differences in the pattern of neural precursor cells
.
Devi Biol
.
111
,
193
205
.
Fernandez
,
J.
(
1980
).
Embryonic development of the glossiphoniid leech Thermoyzon rude: characterisation of developmental stages
.
Devi Biol
.
76
,
245
262
.
Fernandez
,
J.
&
Stent
,
G.
(
1982
).
Embryonic development of the hirudinid leech Hirudo medicinalis: structure, development and segmentation of the germinal plate
.
J. Embryol. exp. Morph
.
72
,
71
96
.
Grimstone
,
A. V.
&
Skaer
,
R. J.
(
1972
).
A Guidebook to Microscopical Methods
.
Cambridge University Press
.
Ho
,
R.
&
Weisblat
,
D. A.
(
1987
).
A provisional epithelium in the leech embryo: cellular origins and influence on a develonmental equivalence group
.
Devi.Biol
.
120
,
520
534
.
Kleinenberg
,
N.
(
1879
).
The development of the earthworm Lumbricus trapeziodes
.
Q.J. microsc. Set
.
14
,
206
244
.
Kramer
,
A. P.
&
Weisblat
,
D. A.
(
1985
).
Developmental kinship groups in the leech
.
J. Neurosci
.
5
,
388
407
.
Needham
,
A. E.
(
1952
).
Regeneration and Wound Healing
.
London and New York
:
Meuthuen & Co. Ltd. John Wiley & Sons Ltd
.
Penners
,
A.
(
1934
).
Experimentelle untersuchlangen zum determinations problem am keim von Tubifex rivolorum Lam. III. Der teloblasten auf verschiedenen entwichlungs-stradien des keimstreifs. Z
.
wiss. Zool
.
145
,
220
260
.
Penners
,
A.
(
1937
).
Regulation am keim von Tubifex rivulorum lam. nach auschaltung des ektodermalen keimstriefs
.
Z. wiss. Zool
.
149
,
86
130
.
Rude
,
S.
(
1969
).
Monoamine-containing neurons in the nerve cord and body wall of Lumbricus terrestns
.
J. comp. Neurol
.
128
,
397
412
.
Sawyer
,
R. T.
(
1986
).
Leech Biology and Behaviour
, vol.
1
, p.
58
and 160-162
.
Stuart
,
D. K.
(
1981
).
Monoamine neurons in embryonic and adult leech
.
Neurosa. Abstr
.
7
,
43
.
Torrence
,
S.
&
Stuart
,
D.
(
1986
).
Gangliogenesis in leech embryos: migration of neural precursor cells
.
J. Neurosci
.
6
,
2736
2746
.
Weisblat
,
D. A.
,
Harper
,
G.
,
Stent
,
G. S.
&
Sawyer
,
R. T.
(
1980
).
Embryonic cell lineage in the nervous system of the glossiphoniid leech, Helobdella triserilais
.
Devi Biol
.
76
,
58
78
.
Weisblat
,
D. A.
,
Kim
,
S. Y.
&
Stent
,
G. S.
(
1984
).
Embryonic origins of cells in the leech Helobdella triserialis
.
Devi Biol
.
104
,
65
85
.