ABSTRACT
Localized regions of Drosophila eggs were damaged at both nuclear multiplication and blastoderm stages, using the technique of microcautery. The resulting defects in embryos and adults were recorded in order to study the effects of microcautery on determination of larval and adult structures. Using 70 °C microcautery, few specific defects were found in either embryos or adults. Sensitivity maps relating embryonic mortality to sites of damage were prepared for both egg stages and compared. Microcautery at 75 °C was used to check experimentally the putative localization in the egg of adult disc cells, as indicated by genetic mosaic studies. Although some of the expected adult defects were found, there was also evidence for regulation. Localized embryonic defects were found using this higher temperature, such that anterior damage produced anterior defects and posterior damage led to posterior defects. Mid-region microcautery resulted in both anterior and mid-region defects, the latter being most common after damage to the blastoderm. Comparisons are made with results obtained using other experimental manipulations and with the development of some embryonic lethal mutants.
INTRODUCTION
One of the outstanding questions concerning the development of Drosophila is whether or not the egg is truly mosaic, i.e. if the fates of all parts of the embryo are fixed at fertilization such that localized damage will subsequently manifest itself as loss of parts. Early work suggested that this was the case, and ‘fate mapping’ of the egg using genetic mosaics might be similarly interpreted. However, the latter experiments do not prove this, and there is some evidence (see below) of regulation after damage to early egg stages. It is therefore possible that mosaicism is established only after the egg becomes cellular, or that different areas of the egg become determined at different times. Certainly if there is any period of lability it must be early and short. The experiments described here compare the consequences of treating two early stages, the acellular and cellular cortex, using microcautery of small areas to damage them.
Previous experiments of this sort have given conflicting results. Howland & Child (1935) removed cells from blastoderm stage eggs and noted defects in adults correlated with the site of damage, but they found very few defective flies, and many that might have regulated. Howland & Sonnenblick (1936), who pricked eggs at nuclear multiplication stages (pre-blastoderm), found no defective adults, and concluded that regulation was then possible, llmensee (1972), on the other hand, found adult defects roughly correlated with the site of damage after similar, but more extensive, pricking experiments. Nothiger & Strub (1972) found adult defects after irradiating early embryos with u.v., but these defects were not correlated with the sites of irradiation. All these experiments are concerned with the cells which will form the adult discs, and these are certainly determined shortly after blastoderm formation since dissociated cells from this stage can give rise to adult structures after culture in vivo (Chan & Gehring, 1971). But the problem-was the mosaic established before this ? — cannot be resolved by this transplantation technique.
Surprisingly, only Hathaway & Selman (1961) appear to have looked at the determination of larval structures. They found that u.v. damage to blastoderm stages produced a range of embryonic and larval defects, with some regularity in their pattern related to the site of initial injury. We have also looked at this aspect of determination.
MATERIALS AND METHODS
Origin and preparation of eggs
The eggs used for microcautery at 70 °C were Oregon-Samarkand hybrids, and the 75 °C microcautery eggs were Oregon K stock. Eggs were collected at 25 °C on agar plates coated with yeast paste. After a 1 h prelay, eggs were collected at 20 min intervals and used immediately, or incubated at 25 °C until they were of the required age. The eggs were washed from the agar with 0·9 % sodium chloride, then dechorionated with 3 % sodium hypochlorite for 5 min. They were then washed again with sodium chloride solution before treatment.
For the first set of microcautery experiments eggs were used from accurately timed collections. Nuclear multiplication (NM) stage eggs were 1 hr ± 15 min, and blastoderm (Bl) stage were 2 3/4 h ± 15 min. Eggs for the second set of experiments were selected using the binocular microscope. Stage NM was when the nuclei were dividing inside the yolk of the egg. This stage can be easily recognized from the lack of pole cells. Stage Bl was when the outer layer of blastoderm cells was clearly visible, but before any morphogenetic movements had begun.
Experimental technique
A few eggs were placed on a piece of lens tissue lying on black polythene. This prevents the eggs from drying out (which reduces their viability) and makes them clearly visible against the black background. The eggs were orientated as required under a dissecting microscope using a fine needle. A very fine tungsten needle was mounted on a micromanipulator and connected to a variable a.c. heat source. The needle was heated to 70 or 75 °C (for the 1st and 2nd set of experiments respectively), and calibrated by using paraffin waxes of known melting point. Each egg was touched at the relevant spot with the tip of the needle. Control eggs were treated in the same way, but not touched with the needle. For the first set of experiments one side of the egg was divided into reference points shown in Fig. 1 A. Samples of 10-20 eggs were cauterized at each of these points, at each of the two developmental stages (NM and Bl). For the second set of experiments the egg was treated at the points in Fig. 1B, which is an adaptation of Hotta & Benzer’s (1972) map showing the location of the presumptive adult disc cells on the surface of the egg.
(A) Grid reference of the egg, used for microcautery experiments at 70 °C and for describing regions of the egg with reference numbers. (B) An adaptation of the fate map produced by Hotta & Benzer (1972) showing the location of the presumptive adult disc cells. These points were damaged by microcautery at 75 °C.
(A) Grid reference of the egg, used for microcautery experiments at 70 °C and for describing regions of the egg with reference numbers. (B) An adaptation of the fate map produced by Hotta & Benzer (1972) showing the location of the presumptive adult disc cells. These points were damaged by microcautery at 75 °C.
There are a number of variable factors inherent in using this technique of thermocautery. The site damaged is judged by eye, and as the areas are small there is bound to be some inaccuracy in this. There will also be some variation in the actual temperature of the needle due to the amount of water left on the surface of the egg, which will conduct away heat; this will lead to variations in the degree of damage. The length of time the needle is in contact with the egg will also be subject to human error. By dissecting adult females and looking at the development of the ovary, it is possible to show that even with these variations the area damaged by this technique is fairly limited. The only areas of the egg leading to defective ovaries were in positions 10 and 11, both at the posterior of the egg; and flies were often found with an ovary one side of the body, but not on the other.
The treated eggs were picked up with a Pasteur pipette and put on an agar plate in a small amount of 0-9 % saline. The plate was covered with a Petri dish lid and incubated at 25 °C for 24 h in the dark.
Analysis of defective embryos
Those embryos which failed to hatch after 24 h were picked up in a saline solution, put on a slide and covered with a coverslip supported by broken pieces of coverslip. This arrangement prevents the eggs from bursting and also flattens them somewhat, making the internal organs more easily visible. The eggs were looked at under high magnification, described and photographed.
Analysis of adults
The hatched larvae from the agar plates were picked up with a paper spoon and put into vials containing yeasted Lewis medium, and then kept at 25 °C to continue development. These adults were observed at the time of emergence 10 days later, and scored for abnormalities in external morphology. Any pupae failing to hatch after some days were dissected and all abnormalities noted. Adults were kept for at least 5 days to see if any died early due to internal deficiencies. Preparations of relevant parts of the fly were made. Adults were left overnight in 10 % potassium hydroxide, which destroyed soft internal tissues, but not the chitinized parts of the body. The potassium hydroxide was rinsed away with water and the flies blotted dry. They were then dehydrated in glacial acetic acid for 5 min and blotted dry before placing them in clove oil to clear. Each fly was then put on a slide in a drop of oil and the relevant part dissected and arranged as required. The clove oil was then removed and the parts mounted in Canada balsam.
Statistical analysis of microcautery data
Chi-square tests at the 5 % significance level were used to compare the mortalities and abnormalities at different stages of development embryonic, larval, pupal and adult. The graphs of percentage embryonic abnormality against the position along the embryo (anterior-posterior) show the standard error of the percentage.
RESULTS AND DISCUSSION
Microcautery at 70 °C
General results
Microcautery at both NM and Bl increased the proportion of eggs failing to show any differentiation, and also the number of embryos developing abnormally (Table 1). Pupal death was unaffected, but more larvae died after treatment at Bl than at NM. The number of adult defects was increased, though only just significantly, at BL The NM treatment caused more embryonic death than the Bl treatment, with the greatest difference being at the poles (Fig. 2).
—, Bl; —, NM. x - x, Nôthiger & Strubb (1972) u.v. irradiation
(A) Graph of percentage embryonic mortality for the combined results of ventral regions ABC against the anterior-posterior position along the embryo 1-11. Standard errors of the percentages are shown by bars.
(B) Graph of percentage embryonic mortality for the combined results of dorsal regions DEF against the anterior-posterior position along the embryo 1-11. Standard errors of the percentages are shown by bars.
(C) Graph of the percentage embryonic mortality for the combined results CDE against the anterior-posterior position along the embryo 1-11. Also the percentage embryonic mortality after u.v. irradiation of an area equivalent to CDE (Nôthiger & Strub, 1972) plotted against their percentage values along the anterior-posterior axis of the egg. No results were given for the anterior 20 % of the egg. Standard errors of the percentage are shown by bars.
—, Bl; —, NM. x - x, Nôthiger & Strubb (1972) u.v. irradiation
(A) Graph of percentage embryonic mortality for the combined results of ventral regions ABC against the anterior-posterior position along the embryo 1-11. Standard errors of the percentages are shown by bars.
(B) Graph of percentage embryonic mortality for the combined results of dorsal regions DEF against the anterior-posterior position along the embryo 1-11. Standard errors of the percentages are shown by bars.
(C) Graph of the percentage embryonic mortality for the combined results CDE against the anterior-posterior position along the embryo 1-11. Also the percentage embryonic mortality after u.v. irradiation of an area equivalent to CDE (Nôthiger & Strub, 1972) plotted against their percentage values along the anterior-posterior axis of the egg. No results were given for the anterior 20 % of the egg. Standard errors of the percentage are shown by bars.
Embryonic mortality
The difference between the effects of microcautery at NM and Bl can best be seen in Fig. 2, which records the mortalities found when microcautery in dorsal (Fig. 2 A), in ventral (Fig. 2B) and in central (Fig. 2C) regions are plotted against position along the anterior axis of the egg. From the combination of these data, sensitivity maps for the nuclear multiplication and blastoderm stages were made to provide a visual summary of the results (Fig. 3). Clearly, the nuclear multiplication stage embryo is more easily damaged by microcautery, and the two stages show notably different sensitive areas.
In the ventral areas ABC (Fig. 2) the anterior is very sensitive at NM, but is considerably less so by Bl. Similarly region 10, and probably region 5, show this marked decrease in sensitivity between NM and Bl. The high mortality response to the treatment moves from the two poles of the egg at NM to the midline by Bl, essentially due to a decline in the sensitivity at the poles during this period. The dorsal regions DEF (Fig. 2B) are different and, generally, all regions are more sensitive at NM than Bl to about the same extent, the exceptions being areas around the posterior pole.
Nôthiger & Strub (1972) studied the effects of u.v. microbeam irradiation of NM eggs. Their irradiated area was approximately equivalent to the combined regions CDE. The relevant data is summarized in Fig. 2C, which shows virtually no difference between the two techniques of damaging the periplasm, except at the posterior pole, where thermocautery causes higher mortality than u.v. irradiation, llmensee (1972) tested for the most suitable region for injection of nuclei into eggs 17-20 min old (NM) by pricking the egg surface and subsequently scoring the sensitivity of the area by the number of eggs hatching. He found the least sensitive region was a lateral one, midway between the anterior and posterior poles ; a similar region can be seen in Fig. 2C. These similarities in the sensitivity of early eggs suggest that u.v. irradiation, pricking and thermocautery all damage the same ‘targets’, involved in embryonic development and located in the periplasm.
The sensitivity differences in the two maps at NM and Bl (Fig. 3) may be due to a distinct change when the nuclei migrate into the periplasm and cells form, or they may be two steps in a pattern which is constantly changing as the embryo develops. The fragmentation experiments of Herth & Sander (1973) with Protoformia and with Drosophila suggest that the latter hypothesis is more likely to be true. The segment pattern is gradually established from fertilization to blastoderm formation, so one might expect the sensitivity of different regions of the egg to change correspondingly.
Embryonic abnormalities and the formation of larvae
Several organs may be abnormally formed after microcautery at one particular region. There are two possible explanations for this kind of result.
(1) Normal morphogenesis may be impeded as a consequence of the primary defect, leading to a general deformity in which the primary cause may not readily be identified.
(2) Some cells, or groups of cells, may influence differentiation in other groups of cells, and damage may thus interfere with a sequential series of relationships.
It follows that only particular and specific abnormalities are useful in the context of mapping the egg for larval structures, although the frequent multiple effects of microcautery may tell us about some of the general mechanisms underlying the development of larval structures.
It was not possible to classify these embryonic defects in a sensible fashion; almost every egg was slightly different from the next. Many eggs (61·2 % at NM and 57·8 % at Bl) reached the stage of having some segmentation, some gut formation and often some tracheal formation. There were often small defects in several systems, and because of possibilities (1) or (2) above, these were not correlated to the initial site of damage. At NM, 18·3 % of eggs, and 13·9 % at Bl, died during early development before the formation of segments or tracheals. The gut was usually disarranged, a general defect found after microcautery at many sites. At Bl a few embryos formed specifically anterior or posterior defects. Those with the posterior well formed and the anterior undifferentiated or containing a contracting gut mass resulted from damage to regions 1-9, and embryos with anteriors well formed and posteriors containing a gut mass arose from damage to posterior regions 6-11. At NM this type of defect was rare, but was always the result of damage to sites causing the same defects at Bl. Such defects did not arise more frequently from particular regions on the dorso-ventral axis.
Abnormalities of the head and of mouthpart formation usually arose from damage in the anterior half of the egg at Bl; at NM damage at almost any point on the egg’s surface produced these defects, although with greater frequency from anterior microcautery.
Adult defects
Very few imaginai defects were found amongst the hatching adults and unemerged pupae. The kinds of abnormalities found were as follows:
(1) Defects of mouthparts: proboscis or labial palps absent or deformed.
(2) Wings absent, incomplete or deformed, usually accompanied by an abnormally shaped thorax and scutellum.
(3) Abnormalities of the legs: parts of the leg may be absent, joints crippled or too long; a leg may be completely absent, or fail to evert correctly.
(4) Abnormal abdomens: fusion of tergites of different segments, missing halves of tergites, distorted shape of abdomen as a whole.
No defects were found of eyes, antennae or halteres in these experiments.
When the treated embryos were less than 1-hr-old (NM) there were so few adult abnormalities that no conclusions can be drawn from them.
After microcautery at blastoderm formation, defects were on the expected side of the adult, except occasionally when both prothoracic legs were abnormal or the mouthparts were abnormal on both sides of the proboscis. Not enough defects were found to make an accurate map related to the surface of the egg, but defects found did show some correlation with the initial site of damage. Mouthparts were only abnormal after treatment of the anterior third of the embryo, and the abdomen was affected much more frequently with posterior than with anterior microcautery. The wing and leg abnormalities were spread over quite a large area of damage.
Microcautery at 75 °C
General results
As the eggs were selected to be in the correct developmental stage, the nuclear multiplication control sample showed a higher frequency of embryonic death than blastoderm controls. When corrections were made for this, a larger number of eggs failed to continue development after treatment at NM than at Bl, as shown in Table 2. More embryos developed abnormally, more larvae died, and a greater frequency of morphologically abnormal adults and pupae were found, after treatment at Bl. The only adult defects in the controls were abnormal abdomens.
Embryonic defects
Defective embryos arising from thermocautery at the sites shown in Fig. 1B were classified according to the scheme below, and the results were tabulated in relation to the area of the egg damaged (Table 3).
Class 1. Anterior defects
Embryos with abnormal head formation and with mouthparts abnormal or absent (Fig. 4); embryos where the posterior has some segmentation and the spiracles are present, the gut and yolk often extruded at the anterior of the embryo. This class includes all embryos from those with just a few abdominal segments partially or abnormally formed, to those with all segments, a well-formed tracheal system and some head formation (Fig. 5). In some embryos the anterior contains a large yolk patch and the posterior has a contracting mass of gut tissue; they sometimes also have a few abnormally arranged bristle rows on the surface, and one or two spiracles present (Fig. 6).
An embryo showing abnormal head formation and also some abnormal mouthpart (w) formation. (Class 1 in Table 3.) The anterior is at the top of the figure in this, and subsequent, photographs. The length of the egg is 0-42 mm.
An embryo showing abnormal head formation and also some abnormal mouthpart (w) formation. (Class 1 in Table 3.) The anterior is at the top of the figure in this, and subsequent, photographs. The length of the egg is 0-42 mm.
This embryo contains all the abdominal segments at the posterior, but the gut (–) is extruded at the anterior. (Class 1 in Table 3.)
This embryo contains all the abdominal segments at the posterior, but the gut (–) is extruded at the anterior. (Class 1 in Table 3.)
At the anterior is a yolk mass, and at the posterior a contracting gut mass (g) with signs of spiracles (s) and one or two abdominal segments. (Class 1 in Table 3.)
At the anterior is a yolk mass, and at the posterior a contracting gut mass (g) with signs of spiracles (s) and one or two abdominal segments. (Class 1 in Table 3.)
Class 2. Mid-region defects
Embryos where there is some formation of segments both at the anterior and posterior, but with an abnormality in the central region which does not usually extend right across the embryo (Fig. 7). In some embryos the abdominal segmentation is irregular, the bands running into one another as in ‘abnormal abdomen’ adults (Fig. 8).
Class 3. Posterior defects
Embryos where the head and thorax have formed abnormally, but there are signs of mouthpart formation. The posterior of the embryo contains a mass of yolk and gut (Fig. 9). In other embryos the head and thoracic segments are formed perfectly and the mouthparts are complete. There may be abdominal segments of varying number, either complete or abnormally formed behind the thorax; the gut is extruded at the posterior (Fig. 10). More damaged embryos may have a large yolk patch at the posterior with the anterior containing a gut mass (Fig. 11).
The abdominal segment boundaries are abnormal, two running into one as in ‘abnormal abdomen’ adults. (Class 2 in Table 3.)
The abdominal segment boundaries are abnormal, two running into one as in ‘abnormal abdomen’ adults. (Class 2 in Table 3.)
The posterior contains a yolk mass (y) and the anterior a contracting gut mass (g). (Class 3 in Table 3.)
The posterior contains a yolk mass (y) and the anterior a contracting gut mass (g). (Class 3 in Table 3.)
At the posterior there is a yolk mass (y), but at the anterior there are perfectly formed head and mouthparts and all eight abdominal segments, including one spiracle. This almost complete embryo occupies little more than half of the egg. (Class 3 in Table 3.)
At the posterior there is a yolk mass (y), but at the anterior there are perfectly formed head and mouthparts and all eight abdominal segments, including one spiracle. This almost complete embryo occupies little more than half of the egg. (Class 3 in Table 3.)
At the anterior the mouthparts (m) and a number of abdominal segments are perfectly formed. At the posterior the gut (g) is extruded. (Class 3 in Table 3.)
At the anterior the mouthparts (m) and a number of abdominal segments are perfectly formed. At the posterior the gut (g) is extruded. (Class 3 in Table 3.)
Class 4. Mouthpart formation amongst anterior defects
Embryos with a contracting gut mass, where there is often irregular segmentation but there has at some stage been abnormal mouthpart formation amongst the gut tissue. Some embryos have a well-formed posterior and gut extruded at the anterior, but also have a set of well-formed mouthparts facing into the embryo or to the side of it; there is no head formation or pharynx associated with this structure (Fig. 12 A, B).
(A) The posterior two-thirds of this embryo contains all the abdominal segments and part of some thoracic segments. A large mass of undifferentiated material is present at the anterior, yet amongst it is a set of perfectly formed mouthparts (w) facing into the embryo. (Class 4 in Table 3.) (B) High-power view of (A) to show detail of mouthparts.
(A) The posterior two-thirds of this embryo contains all the abdominal segments and part of some thoracic segments. A large mass of undifferentiated material is present at the anterior, yet amongst it is a set of perfectly formed mouthparts (w) facing into the embryo. (Class 4 in Table 3.) (B) High-power view of (A) to show detail of mouthparts.
Class 5. Non-specific defects
Embryos with a contracting mass of gut tissue (Fig. 13). Others are extremely transparent in appearance, with the gut formed, but the segmentation very faint. These embryos appear to have died before tracheal formation or mouthpart chitinization (Fig. 14), but not due to a defect in a specific region.
A contracting gut mass with no visible segmentation. (Class 5 in Table 3.)
Class 6. Others
Embryos which did not fit into any class and which were not found more than once.
In contrast to the previous set of microcautery data, the embryonic defects were easy to classify and closely correlated with the initial site of damage. In many cases they agreed well with the kinds of defects found after u.v. irradiation and pricking of similar areas. This is discussed further below, as each treated region is looked at in detail.
Treatment in head, proboscis and leg 1 regions
In all these regions treatment at both stages produced either anterior (Class 1) defects, or non-specific defects of contracting gut masses (Class 5), as with u.v. irradiation at Bl and NM (Bownes & Kalthoff, 1974). Class 4 embryos which were transparent in appearance (Fig. 14) were also found after microcautery in these regions at NM.
Treatment at leg 2 region
At NM, leg 2 region damage produced anterior defects. After treatment at Bl Classes 2 and 4 were found as well as anterior defects. These were not found after u.v. irradiation, possibly because the area damaged by microcautery is smaller, so that anterior morphogenesis is not always impaired, as may occur after u.v. irradiation of a central band of the egg. One embryo was found with anterior and posterior development and an abnormal mid-region after pricking the embryo at blastoderm formation, but usually anterior defects were found.
Treatment at leg 3 region
After treatment at NM mostly anterior and nonspecific defects were found, though examples of posterior and mid-region defects were also seen. Bl treatment produced anterior defects, non-specific posterior and mid-region defects, and also embryos with inverted mouthparts. The abnormal abdominal segmentation was found after treatment at Bl and NM. It should be noted that hatched larvae were not checked for defects, so it is possible that more defective embryos of this class were formed, which hatched and were thus omitted from the classification of abnormal embryo.
Treatment of wing and thorax regions
Damage at both Bl and NM produced mostly anterior defects and non-specific defects. After Bl treatment Classes 2 and 4 were also found, but these were generally absent after NM treatment. Embryos with inverted mouthparts occur most frequently after treatment of this region. It seems that mouthparts are able to form correctly amongst a mass of gut tissue when there is no head formation, and their formation does not seem to be dependent on induction from the pharynx, or on being in the correct position in the head ; yet when the head forms abnormally the mouthparts are always absent or defective. It is difficult to imagine a scheme of differentiation and determination of mouthpart material which can encompass both these types of defect.
Treatment of the abdominal region
sNon-specific and posterior defects, Classes 3 and 5, were most frequent after damage to this area at both stages. Some Class 2 defects were also found after damage at Bl. This is in agreement with the results of u.v. irradiation and pricking of the posterior egg regions (Bownes & Sang, 1974).
In general the areas located nearest to the anterior or posterior reacted exactly as in experiments using u.v. irradiation and pricking. Damage to more central regions sometimes produced anterior defects, as with u.v. irradiation and pricking, but occasionally produced central defects not found in previous experiments. These new defects arose much more frequently from treatment at blastoderm rather than at nuclear proliferation stages. This suggests that the axes for the development of the embryo are more completely established at blastoderm formation than they are at proliferation stages. This is in agreement with the findings of Herth & Sander (1973), who found that the segmentation pattern of Drosophila is gradually established with time, and is eventually complete at blastoderm formation. The fact that mid-region defects are occasionally found after treatment at NM might be due to the damaged area of cortex preventing this region reacting as part of a gradient system.
The higher temperature, producing more damage to the embryo, gives results rather different to the first set of microcautery experiments. Embryonic defects are well correlated with the area damaged, and are usually large defects.
By looking at particular embryonic abnormalities it is possible to deduce some facts concerning the interactions between certain areas of the embryo and the autonomy of development of some organ systems. It is quite possible for the anterior of the egg to develop normally when the other half is completely undifferentiated or contains a gut mass. It seems that the physical relationships between the anterior and posterior are not necessary for at least some morphogenetic movements to occur. It is also clear that the egg has some capacity to modify size. All abdominal segments and sometimes up to three thoracic segments and some head formation can be present after microcautery, but they occupy only half of the egg, the segments being much more closely crowded together than is usual.
Irregular segmentation has been observed several times around undifferentiated tissue. Segments occur on one side of the mass and not on the other in some cases, and in others may radiate out from a single point on the egg surface. Tracheal tubes are also able to form over a mass of tissue and are not dependent on normal segmentation or organization before they can form. Spiracles have been seen at the posterior of the egg without any segmentation being present, or any other recognizable organ formation in the embryo. We cannot deduce from these results that the egg is a mosaic of small areas determined to be specific structures. Although there is some ability for organs to develop autonomously, they are not always able to do this, suggesting that there are normally many interactions between cells and developing organ systems during embryogenesis.
Adult defects
The great majority of formed adults were normal (Table 4), but the higher temperature of microcautery did produce more specific defects in the adults than were found in the first experimental series. There is a greater frequency of adult defects with blastoderm damage than from damage to the periplasm (10 % of Bl formed adults and 5·7 % of NM formed adults). In both cases the defects found tend to be those expected within the limits of accuracy of the technique (see Materials and Methods). Where more than one structure is affected, the damage is to closely adjacent disc areas.
There could be four explanations of the high frequency of normal adults. Either the technique is so inaccurate that the necessary localization of the damage is not achieved, or some form of repair, regeneration, or regulation is possible. There is no certain way of eliminating the first possibility, but since damage to each area greatly reduces the chances of survival to the adult stage by causing malformations of the embryo or larva, it seems improbable that such a high proportion (21 % of all eggs) of normal adults would survive. It is possible that molecular repair processes may exist within the egg which can repair some of the microcautery damage, but as yet no information is available to decide if this unlikely event could happen. Several forms of cellular compensation are possible : either other cells take over the function of disc cells, or any surviving disc cells are capable of regenerating the whole (see Schubiger, 1971), or the undamaged disc cells become reorganized and regulate to form a whole disc.
Although defects correlated to the initial site of damage are produced after treatment at NM, this does not mean that there are localized regions in the periplasm which are committed to-a specific fate in development, since areas may be unable to respond to a later developmental signal to become committed regions, or nuclei may be unable to move into this damaged region and the disc cells would then not form.
llmensee (1972) also found imaginai defects correlated with the site of damage after pricking NM stage embryos, although Howland & Sonnenblick (1936) found no defective adults after similar experiments, and concluded that the Drosophila egg was capable of limited regulation at this stage. Nôthiger & Strub (1972) found adult defects, not correlated to the initial site of damage, after partial u.v. irradiation of nuclear multiplication eggs. This could be because they irradiated dorso-laterally instead of ventro-laterally where the presumptive disc cells are located; damage to the discs probably resulted from u.v. which had been scattered within the embryo.
The defective adults resulting from treatment at Bl do not necessarily show that the cells are determined to be adult disc cells at this time, because the damaged cells may remain in position and prevent the migration of nearby cells into this region so that no compensation occurs. Although possible, this seems unlikely since so many adults which hatch have normal morphology. The data agree on the other hand with the results of Chan and Gehring (1971) which show that cells are committed to become an adult disc after blastoderm formation. Chan and Gehring cut blastoderm eggs in half, disaggregated and reaggregated them. After culture in adults the fragments were injected into larvae and underwent metamorphosis. Head and thoracic structures formed from anterior fragments and thoracic and abdominal structures from posterior fragments, thus showing there was at least an anterior-posterior determination of adult disc cells at blastoderm formation. Combined with these results, microcautery suggests that the specification is for individual discs, i.e. the adult disc cells are determined to be a specific cuticular structure at blastoderm formation. The abnormalities in the adults were always large; structures were missing or grossly abnormal. There were never small defects such as a few bristles absent, or a small defect in the wings or legs. This suggests that there is no specificity within the group of cells determined to be a disc for specific parts of that disc. This ‘all or nothing’ type of response to microcautery suggests that if only a small number of disc cells are damaged they can be repaired or regenerated by the remaining cells, and that regulation will eventually produce a normal adult.
DISCUSSION
Microcautery at 75 °C gives a greater frequency of more precise defects than microcautery at 70 °C, probably because it produces a greater degree of damage at a similarly small site. Nonetheless there is a considerable variability in the effects produced by damage at any one site, and it is worth noting that this is not necessarily a consequence of the technique, but has also been found with mutants affecting embryogenesis. For instance, the dominant vestigial (VgD) mutant of Bull (1952) causes abnormal head formation during embryogenesis very similar to that described above (Fig. 4). Further study of the mutant (M. Bownes, unpublished observations) shows that the whole range of defects comprising Class 1 were produced by this small deletion of the second chromosome. This suggests that the large variety of defects found after experimental manipulations is a feature inherent in the system of a developing egg, and not totally due to small variations in the technique. Further, the variety of defects found amongst VgD homozygotes alters with the genetic background (A. L. Bull, personal communication; M. Bownes, unpublished observations) showing that the genetic background of the organism can also affect the type of aberrant development which results from this small deletion.
Shannon (1973) has described abnormal posterior development in eggs produced by the female sterile mutant almondex (amx) of Drosophila. Some embryos which are genetically heterozygous resemble embryos experimentally damaged at the posterior (Fig. 11). The occasional heterozygous females which hatch often show thoracic and abdominal defects, like those resulting from mid-region and posterior microcautery. It seems that both larval and adult structures may be affected by the same gene, the defect being dependent on the location in the posterior of the embryo of the presumptive cells of these structures. Similarly, microcautery produces adult and embryonic defects dependent on the site initially damaged.
The two stages of the embryo included in this study respond differently to microcautery not only with respect to survival but also in their patterns of larval and adult defects. In particular, mid-region defects and reversal of mouthparts were commonest, as were particular adult defects after microcautery of blastoderm eggs (Table 3), and we have the impression that the state of determination is more specific then than earlier. However, there is considerable variability between regions of the egg, and there is no clear evidence that the egg is a mosaic, sensu stricto, at either stage. Here we must note, however, that damage must also be affecting embryonic cell movement, and that our results may be confused by these secondary effects, which require further study. Since mapping with genetic mosaics (Hotta & Benzer, 1972) localizes adult structures very precisely, the implication is that mosaicism is established somewhat later than blastoderm formation. This certainly seems to be the case for larval cells removed from4-6 h embryos, some of which differentiate to their final state when cultured in vitro (Shields & Sang 1970, and unpublished observations). We conclude, therefore, that the distinction between regulative and mosaic eggs is one which depends on the time of determination, and is not absolute. Our evidence for regulation, particularly with respect to adult structures, is certainly open to the criticism that it may be a consequence of other mechanisms, particularly molecular repair, and this will be dealt with in a subsequent paper.