Patterning during pupal commitment of the epidermis in the butterfly, Precis coenia: The role of intercellular communication

DEVELOPMENTAL

BIOLOGY

133,336-347

(1989)

Patterning during Pupal Commitment of the Epidermis in the Butterfly, Preck coenk: The Role of Intercellular Communication CLAIREKREMEN Department

of Zoology,

Duke Accepted

University,

Durham,

January

North

Carolina

27706

23, 1989

The commitment of cells to pupal development in the larvae of holometabolous insects can be prevented by treatment with juvenile hormone (JH) or a JH mimic during a critical period early in the last larval instar. By treating larvae of different ages with a JH mimic, pupal commitment of the epidermis of the butterfly, Precis coenia, was found to occur in a strict temporal and spatial progression, which was serially homologous and occurred independently in each segment. The mechanism underlying this sequence of pupal commitment was examined by cauterizing regions of the epidermis to observe the effects of local ablation on the pattern of pupal commitment revealed by treatment with the JH mimic. Cautery of the segmental site of origin of pupal commitment, the dorsal midline, suppressed pupal commitment in the rest of the operated segment, indicating that the midline has a special effect on commitment of the rest of the segment. Cautery off the midline produced asymmetries in the pattern of pupal commitment; when placed close to the midline, such cauteries prevented pupal commitment in the region “downstream” of the cautery, suggesting that a signal (diffusible or transducible) emanates from the midline. Finally, cautery of a circle around the midline inhibited pupal commitment only outside the circle, showing that cautery could act as a barrier to the passage of a signal coming from the midline. These results suggest that inductive as well as hormonal signals are involved in the regulation of pupal commitment in the epidermis of the lepidopteran, P. coenia. 0 1989 Academic Press. Inc.

vide a tool for visualizing the process of pupal commitment. Tissues are deemed committed to pupal development when ecdysteroid-induced metamorphosis at the end of the instar can no longer be prevented by juvenoid treatment earlier in the instar (Truman et al., 1974). The sequence of pupal commitment in different regions of the integument forms a complex temporal and spatial pattern (Fukuda, 1944; Sehnal & Schneiderman, 1973; Truman et ab, 1974; Yin and Chaw, 1984; Kremen, 1987). While this has been recognized for many years, few studies have examined the mechanism(s) producing this pattern, focusing instead on the hormonal control of pupal commitment. Indeed, the studies which have analyzed temporal and spatial patterns of pupal commitment (Sehnal, 1968; Truman et ab, 1974; Ohtaki et al., 1986) have generally interpreted these patterns as resulting from differences in the sensitivity of tissues to the hormones that control metamorphosis. For example, patterns of pupal commitment in the moth, Manduca sexta, have been attributed to differential ecdysteroid requirements of different tissues (Truman et al, 1974; Fain, 1975). In Manduca, a small, early increase in the ecdysteroid titer in the absence of JH has been shown to induce pupal commitment of the epidermis (Riddiford, 1976,1978). Thus, a simple explanation for the temporal and spatial sequence of pupal commitment is that tissues differ in the level or duration of ecdysteroid exposure required to induce com-

INTRODUCTION

The metamorphosis of insects involves complex morphogenetic changes resulting from altered patterns of gene expression; these changes are coordinated by hormones. In the larval-pupal metamorphosis of holometabolous insects, the anti-metamorphic or status quo hormone, juvenile hormone (JH), must be below a threshold level during a critical period early during the last larval instar in order for the epidermis and imaginal disks to attain a pupal commitment (Kurushima and Ohtaki, 1975; Riddiford, 1976, 19’78; Nijhout and Wheeler, 1982). However, the expression of pupal fate does not occur until the end of the instar, in response to the prepupal surge of the molting hormone, 20-hydroxyecdysone, which stimulates the evagination of the imaginal disks (Oberlander, 1969) and the secretion of pupal cuticle (Wigglesworth, 1934; Riddiford, 1982). Exogenous JH can prevent pupal commitment of larval cells if applied prior to or during the critical period (Oberlander & Silhacek, 1976; Riddiford, 1976, 1978; Fain and Riddiford, 1977; Nijhout and Wheeler, 1982). The degree to which exogenous JH prevents pupal commitment depends on the time of application; frequently, treatment with JH prevents pupal commitment in only a portion of the tissues, leading to the production of larval-pupal mosaics. Larval-pupal mosaics produced by JH treatment at different times can therefore pro0012-1606189 Copyright All rights

$3.00

0 1989 by Academic Press, Inc. of reproduction in any form reserved.

336

CLAIRE

KREMEN

Pupal

Commitment

mitment; consequently, as the ecdysteroid titer increases through time, it stimulates pupal commitment first in more and then in less sensitive tissues. While epidermal tissues in M. sexta last instar larvae do differ in the duration of ecdysteroid exposure required to induce molting (Truman et al., 1974; Mitsui and Riddiford, 1976), no comparable evidence exists to suggest that tissues also vary in the duration (or level) of ecdysteroid exposure required to induce pupal commitment. In addition, since certain tissues such as the imaginal disks achieve pupal commitment several days prior to the increase in ecdysteroid concentration that induces pupal commitment (Truman et al, 1974; Nijhout, 1976; Safranek and Williams, 1984; Kremen, unpublished), this mechanism clearly cannot explain the entire observed temporal pattern of pupal commitment. Alternatively, temporal and spatial patterns of pupal commitment could result from differential sensitivity of tissues to JH. In this view, tissues become committed to pupal development once the JH titer declines below a tissue-specific JH threshold. The decline of the JH titer early during the last larval instar (Granger and Bollenbacher, 1981) would thus naturally generate a temporal sequence of pupal commitment due to differential sensitivities of tissues. Neither of these hypotheses for the origin of patterns of pupal commitment has been tested. In this paper, evidence is presented which is not compatible with models of temporal patterns of pupal commitment that are based on differential hormonal sensitivity. Instead, the ontogeny of pupal commitment in the epidermis of P. coenia is ascribed to a pattern formation process whereby sources located in discrete regions of the epidermis induce pupal commitment in adjacent epidermis by intercellular communication. MATERIALS

Rearing and Staging

AND

METHODS

Animals

Animals were reared as described by Nijhout (1980) at 25°C under a 16-hr L:8-hr D photoperiod using an artificial diet modified from Yamamoto (1969), containing about 1% by weight of Plantago lanceolata, a larval food plant. Pharate last instar larvae were collected each morning at 12:00 (hrs after lights off), weighed at 17:00 (after the molt), and reweighed 24 hr later. These two weight measures were used to assessthe probability, calculated from a discriminant function (see below), that an individual larva would spend either 5 (Gate 1) or 6 (Gate 2) days in the last larval instar. Only larvae with a 70% or better probability of being Gate 2 were chosen for experiments: of these, individual probabilities of being Gate 2 ranged from 70 to 95%.

Patterns in Precis

337

To generate the discriminant function used to determine the probabilities of a larva being Gate 1 versus Gate 2, larvae reared individually in 1 oz cups were weighed each day at 17:00, from the day they entered the last larval instar (Day 0) until pupation. Knowing their weights on each day and the length of the instar, a discriminant function could be calculated based on weight data for any subset of days. This discriminant function could then be used to predict the probability that a larva would spend 6 days in the instar (Gate 2), based on its weight on particular days. Comparison of discriminant analyses using successively fewer days of weight data showed that the combination of weights from Days 0 and 1 was as good a predictor of gate as any more inclusive set of weight data (Kremen, 1987). Hence larvae were staged by weighing only on Days 0 and 1. Hormone Treatments Fifty micrograms of the JH mimic, S-hydroprene (Zoecon), in 1 ~1 of acetone (HPLC grade), was applied topically to the dorsal prothorax of each larva, at different time during the last larval instar. Specimens were examined under a dissecting microscope after molting and scored for larval versus pupal characteristics of the epidermis. Some of the treated individuals were either photographed using a Wild microscope equipped with a photoautomat and/or drawn using a camera lucida on a Zeiss dissecting microscope. Previous work (Kremen, 1987; Kremen and Nijhout, 1989) established the required dosage and timing of JH mimic treatments and showed that the effects of the JH mimic were identical to the effects of JH 1, a juvenoid found in lepidopteran hemolymph. These studies also showed that the effects of juvenoid treatment were not dependent upon the site of topical application (different regions of the dorsal integument were tested) and that the effects of topical applications were equivalent to those of injections of similar doses of hormone contained in mineral oil. Cauteries After larvae were anesthetized for 0.5 hr under COz, they were cauterized in specific locations by brief contact with the needle of a high-frequency monopolar electrocoagulation apparatus (Hyfrecator-Birtcher Corp.). After cauteries were performed (between 2:00 and 15:00 on Day 3), larvae were returned to food. Three to ten hours after cautery, larvae were treated with the JH mimic, S-hydroprene, as described above. Whole mounts of the integument of several larvae cauterized along the dorsal mesothoracic midline (but not treated with S-hydroprene) were prepared by fixing them in 10% buffered formalin containing 0.25 M su-

33%

DEVELOPMENTAL

BIOLOGY

crose, washing them in dHaO for 24 hr, stripping them of fat bodies, muscles, and tracheae, staining them with a fluorescent nuclear stain (Hoechst; 0.5 @g/ml), and mounting them in a gycerol-based mounting medium for microscopic examination at 400X. RESULTS

De&nitirm

AND

DISCUSSION

of Pupal Commitment

Pupal commitment of an epidermal region is the time after which JH application no longer prevents that region’s metamorphosis, as evidenced by the production of pupal cuticle in response to the natural surge of ecdysteroid at the end of the instar. Specimens were therefore scored for pupal commitment after the molt. Pupal Commitment

of the Epidermis

Previous studies showed that pupal commitment of the epidermis progressed roughly from anterior to posterior, beginning with portions of the head epidermis on Day 2 and continuing on Days 3-4 into the thorax and abdomen (Kremen, 1987; Kremen and Nijhout, 1989). The last regions of the epidermis to become committed to pupal development included those features whose fate at metamorphosis was cell death (setae, tubercles, prolegs, etc.) (Kremen, 1987; Kremen and Nijhout, 1989). Pupal commitment of the epidermis was completed approximately 29 hr before the ecdysteroid titer peaked at the end of the instar, inducing the molt. While the sequence of pupal commitment of the epidermis was consistent between individuals, the popula-

VOLUME

133, 1989

tion showed considerable variation with respect to the onset of the sequence within the 24-hr period of epiderma1 commitment. For example, a group of staged individuals treated at any one time produced an array of larval-pupal mosaics. As treatment occurred progressively later in the instar, a higher proportion of more advanced (more pupalj mosaics was produced. Utilizing a range of treatment times allowed reconstruction of the entire developmental progression of pupal commitment, as described below. In each thoracic and abdominal segment, pupal commitment began at specific regions located along the dorsal midline and spread in a stereotyped manner through the segment (shown for the thoracic segments in Fig. 1). In the thorax, pupal commitment began roughly simultaneously in each of the segments. In the abdomen, pupal commitment did not begin simultaneously in all segments, but occurred first in the most anterior and the penultimate segments, progressing anteriorly and posteriorly toward the medial abdominal segments. Pupal commitment originated in homologous locations in each segment and spread through each segment in a similar fashion (with slight variations reflecting differences in segment morphology and rate of spread). In each thoracic and abdominal segment, the site of origin of the sequence of pupal commitment was a small triangular region located along the dorsal midline, slightly posterior to the row of dorsal tubercles (or their homologs, the setae, in the prothorax) found in each segment (Fig. 1). Pupal commitment spread from this triangular region around the tubercles. The poste-

FIG. 1. Pupal commitment in the epidermis of the dorsal head and thorax of Precis coeka, as revealed by treatment with the juvenile hormone mimic, S-hydroprene. The schematic illustrations represent larval-pupal mosaics produced by treatmet with S-hydroprene at different times on Day 3 of the last larval instar and are arranged in a developmental series from least (left) to most (right) advanced in pupal commitment. The shaded portions indicate regions of pupal cuticle; unshaded is larval. The most advanced specimen still retains some larval features, such as the setae and tubercles. AN, antenna; H, head; ST, larval setae; Tl, prothorax; T2, mesothorax; T3, metathorax; TU, larval tubercle; WI, wing. For photograph, see Fig. 3A.

CLAIRE KREMEN

Pupal

Ccnnmitment

Pattemzs

in Precis

339

To determine the size of the region affected by caurior base of the triangle expanded posteriorly and laterally posterior to the tubercles, while the apex spread tery, Hoechst-stained whole mounts of cauterized infirst anteriorly and medially between the tubercles and tegument (not treated with S-hydroprene) were prethen anteriorly and laterally anterior to the tubercles. pared. Such whole mounts were found to have a band of Eventually, the anterior and posterior lateral arms of distorted nuclei approximately 500 pm wide in the repupally committed tissue encircled the tubercles and gion of the cautery-equivalent to a string of 125 epifused distally with one another (and with the wing base dermal cells. in the meso- and metathoracic segments). In the mesoBecause of individual differences in the timing of epidermal pupal commitment, it was necessary to dethoracic segment, fusion of the arms was also aided by velop a method for distinguishing between the effects of the independent origin of a region of pupal commitment the experiment and population variation due to devellocated near the wing base (see Fig. 4). The ventral opmental asynchrony of larvae. More specifically, a epidermis was the last region to achieve a pupal commethod was needed which could distinguish larvae mitment, followed by the loss of the tubercles. treated too early or too late relative to the onset of An Alternative to the Differential Hormonal pupal commitment from larvae treated at the approSensitivity Hypothesis priate time in development. The problem of individual differences in developmenBecause pupal commitment of the epidermis consistal timing was overcome by finding that within individtently spread from particular regions such as the dorsal uals the level of commitment obtained by certain segmidline to adjacent regions (Fig. l), it was hypothesized ments was highly correlated with that obtained by the that these sites of origin might act as sources of a signal others. For example, in S-hydroprene-treated larvae, which induced pupal commitment in adjacent epithe pattern of pupal commitment found in the prothodermis as it was transmitted from cell to cell. This hyrax was reliably correlated with that found in the mepothesis provided an alternative to the two differential sothorax (lOO%, N = 54). This correlation in the thohormonal sensitivity hypotheses and was tested by racic segments was maintained in larvae first cauterkilling small groups of epidermal cells. ized on the midline of the first abdominal segment and The differential hormonal sensitivity hypotheses then treated with S-hydroprene (lOO%, N = 44; Figs. 2A predict that killing a small group of cells would produce and 3A). Thus, under the experimental conditions, deonly a local defect in the pattern of pupal commitment, gree of pupal commitment of the prothorax could be due to the loss of cells with particular “prepatterned” taken to indicate the expected level of commitment in hormonal sensitivities. In contrast, a cell-cell commuthe mesothorax, and vice versa. Accordingly, cauteries nication hypothesis predicts that killing a group of cells were always located in either the prothorax or the meshould produce a nonlocal defect in the pattern of pupal sothorax, and the pattern found in the unoperated segcommitment, by influencing neighboring cells which ment (the internal control segment, as in Fig. 2) was formerly communicated with the cells that were killed. used to predict the expected pattern in the operated Furthermore, killing cells of the “source” region would segment. Specimens with no pupally committed epibe expected to abolish the entire pattern influenced by dermis in the thorax and abdomen were discarded from that source, while killing cells near the source should the analysis, since cautery and S-hydroprene treatment distort the pattern produced by the source “downapparently preceded the onset of pupal commitment in stream” of the wounded region. these animals. The use of the internal control segment to predict the Arrest of Pupal Commitment of the Epidermis expected pattern of pupal commitment in another segby Cautery ment required the assumption that segments acted as The dorsal thoracic midline was chosen as the first independent units with respect to pupal commitment. site for ablation since patterns of pupal commitment in This assumption seemed justified given that pupal the thoracic segments began in the regions located commitment appeared to originate independently along the dorsal midline (Fig. 1). The cells located on within each segment at serially homologous, nonadjathe dorsal midline epidermis were killed by cautery and cent sites along the midline. The issue of segment indelarvae were then treated at various times (3-10 hr later) pendence is discussed in more detail below. with the JH mimic, S-hydroprene, to observe the effect of cautery on pupal commitment of adjacent epidermis. The Midline Induces Pupal Commitment This range of times was used in order to visualize effects of cautery on the entire sequence of pupal comThe results of cautery of the prothoracic midline (N mitment. = 55) indicate that the dorsal midline epidermis played

DEVELOPMENTALBIOLOGY

VOLUME 133,1989

A. CONTROLS

B. PROTHORACIC

C. MESOTHORACIC

MIDLINE

MIDLINE

CAUTERY

CAUTERY

t

FIG. 2. Results of cauterizing different regions of the thoracic dorsal midline and then treating with the JH mimic, S-hydroprene, to reveal pupal commitment. Shaded is pupal; unshaded is larval. The arrow marks the segment containing the cautery, which is delineated by the double tracks. IC indicates the internal control segment, the segment which indicates the level of advancement of pupal commitment. (A) A developmental series of the head and thorax of control animals, cauterized on the dorsal midline of the first abdominal segment (not shown) and treated with S-hydroprene. (B) A developmental series of the head and thorax of experimental animals cauterized along the prothoracic midline. Comparing the prothoracic segment of experimentals with the control specimens directly above shows that cautery of the prothoracic dorsal midline inhibited pupal commitment in the prothorax. In the most advanced specimens (right) a remnant of the prothoracic pupal commitment pattern can be observed. Of 55 specimens, 51% showed complete inhibition (first three drawings in series), 44% showed partial inhibition (last drawing in series), and 5% showed no inhibition (not illustrated). For photograph see Fig. 3B. (C) Developmental series of

CLAIRE

KREMEN

Pupal

Commitment

Patterns

in Precis

341

FIG. 3. Photographs of the dorsal heads and thoraces of control and experimental specimens that were first cauterized and then treated with the JH mimic, S-hydroprene, to reveal pupal commitment. Arrows mark sites of cautery. Mottled, smooth, light colored cuticle is pupal; dark-colored cuticle is larval. (A) Control specimen, cauterized on the first abdominal segment. See also Fig. 2A. (B) Experimental specimen, cauterized on the dorsal prothoracic midline. See also Fig. 2B. (C) Experimental specimen, cauterized to the left of the mesothoracic dorsal midline. See also Fig. 5B. (D) Experimental specimen, cauterized on the mesothoracic dorsal midline. See also Fig. 2C. H, head; Tl, prothorax; T2, mesothorax; wi, wing; tu, larval tubercle.

a special role in establishing pupal commitment in the prothorax. Prothoracic midline cautery inhibited pupal commitment in the segment either completely (51%) or partially (44%; Figs. 2B and 3B). Specimens which were only partially inhibited by cautery consistently displayed a more advanced level of pupal commitment in the other thoracic segments. Specimens completely unaffected by prothoracic cautery (5%, N = 3) displayed the highest level of pupal commitment in the other thoracic segments. Thus, in developmentally more advanced larvae, cautery did not wholly abolish pupal commitment, probably because part or all of the pupal commitment pattern had already been determined. However, in less advanced larvae, cautery of the prothoracic midline did inhibit pupal commitment of the

prothorax, indicating that the dorsal midline epidermis influenced pupal commitment in the rest of the segment. Results from cautery of the mesothoracic midline (N = 25) likewise showed that the epidermis of the dorsal midline influenced pupal commitment in adjacent epidermis (Figs. 2C and 3D). As in the prothorax, cautery of the mesothoracic midline inhibited pupal commitment in the mesothorax either completely (40%), partially (32%), or not at all (28%), and the degree of inhibition was related to the level of pupal commitment attained in the rest of the thorax, as described above. In those specimens which were only partially inhibited by cautery of either the prothoracic or mesothoracic midline, the regions of pupal commitment which

animals cauterized on the mesothoracic midline. Comparison of the mesothorax in experimentals and controls shows that cautery of the dorsal mesothoracic midline inhibited pupal commitment in the rest of the mesothorax. Note again that a remnant of the pupal commitment pattern can be observed only in the most advanced experimental specimens. Of 25 specimens, 40% showed complete inhibition, 32% partial inhibition, and 28% no inhibition. For photograph, see Fig. 3D.

342

DEVELOPMENTAL

BIOLOGY

persisted occurred not randomly, but in homologous locations, lateral and posterior to the row of tubercles or setae. This location corresponded to the site of a discrete patch of pupal commitment also observed in the mesothorax of several control larvae (Fig. 4); it could therefore represent a late-acting source of origin of pupal commitment independent of the hypothesized dorsal midline source. Efects

of Wounding

To test whether wounding by cautery could produce the results observed above as artifacts, larvae were cauterized above the prothoracic spiracle instead of on the midline. These cauteries were of magnitudes equivalent to those located on the midline and should therefore have suppressed pupal commitment to the same degree if this inhibition were due to wounding alone. I found that these cauteries did not inhibit pupal commitment in the prothorax (N = 24), although they did produce slight asymmetries in the prothoracic patterns of pupal cuticle (discussed below, see also Fig. 5C). This experiment therefore rules out the possibility that a response to wounding, either generalized, such as a systemic suppression of determinative processes within a segment, or localized, such as a reversion to a larval commitment by cells regenerating in the presence of JH (Sehnal and Schneiderman, 1973), could produce the inhibition of pupal commitment observed after ablation of the dorsal midline. Cuuteries

Act as Barriers

to Communication

or as Sinks

The experiments described so far show that of groups of cells in the epidermis produced alterations in the patterns of commitment, which cannot be attributed to wounding and

ablation nonlocal a result which is

VOLUME

133,1989

not consistent with the hypotheses of differential hormonal sensitivities. These experiments further indicate that the midline was of special importance in establishing pupal commitment of adjacent epidermis; the midline might therefore function as the source of a diffusible or transducible inductive signal. This leads to the prediction that cautery in the path of the signal should inhibit pupal commitment in “downstream” epidermis, either by creating a barrier to diffusion (Kuhn and von Engelhardt, 1933), disrupting transduction of a signal from cell to cell (Sander and NtiblerJung, 1981), or creating a sink for the diffusible signal (Nijhout, 1985a,b). Indeed, tubercles appeared to form a natural barrier between the midline and the epidermis downstream of the tubercles, since this was one of the last regions of epidermis to become pupally committed (Fig. 1). To test this prediction, larvae were cauterized to the left of the mesothoracic midline, in the center of a white pigment spot anterior to the tubercles. Forty-one percent of these larvae (N = 56) subsequently had dramatically asymmetric distributions of pupal cuticle about the midline of the mesothoracic segment, with the epidermis to the left of the wound (downstream) making larval cuticle, while the epidermis to the right of the wound and across the midline made pupal cuticle (Figs. 3C and 5B). The remaining larvae (59%) had symmetric distributions of pupal cuticle about the midline; these larvae were also further advanced in their level of pupal commitment. These results further support the hypothesis that the midline acted as a source for a diffusible or transducible signal which travelled laterally away from the midline; accordingly, regions cauterized during the process of communication acted as sinks or barriers, inhibiting pupal commitment downstream, while cautery occurring after the passage of the signal had no effect on pupal commitment downstream.

FIG. 4. Additional intermediates in the developmental series of larval-pupal mosaics resulting from prene, on Day 3 of the last larval instar. Note the discrete sites of pupal commitment in the mesothorax to the remnants of the pupal commitment pattern observed after cautery of the mesothoracic midline

treatment with the JH mimic, S-hydro(marked by arrows) which correspond in advanced specimens (Fig. 2C).

CLAIRE KREMEN

Pupal

Commitment

Patterns

in Precis

A. CONTROLS

8. MESOTHORACIC

CPROTHORACIC

OFF-CENTER

OFF-CENTER

CAUTERY

CAUTERY

FIG. 5. Results of cauterizing regions off of the dorsal thoracic midline and treating with the JH mimic, S-hydroprene, to reveal pupal commitment. Arrows mark segments in which cautery (dashed line) was performed. IC indicates internal control segment. (A) Control series (as in Fig. 2A). (B) Developmental series of specimens cauterized on the left side of the mesothoracic midline. Comparison of the mesothorax of experimentals and controls shows that cautery inhibited pupal commitment “downstream” of the mesothoracic midline, except in the most advanced specimens. Of 56 specimens, 41% showed asymmetric patterns of pupal commitment (first three drawings in series) and 59?& were not affected (last drawing in series). For photograph, see Fig. 3C. (C) Developmental series of specimens cauterized above the left prothoracic spiracle. Comparison of the prothoraces of experimentals and controls reveals that cautery produced slight asymmetries in the resultant pattern of pupal commitment in the prothorax. Of 24 specimens, 83% showed asymmetric patterns of pupal commitment, and 17% were unaffected.

344

DEVELOPMENTAL

BIOLOGY

As noted previously, larvae which were cauterized above the left prothoracic spiracle often (83%, N = 24) showed slight asymmetries in the distribution of pupal cuticle (Fig. 5C). However, in this case, cauteries prevented pupal commitment in epidermis located upstream of the cautery (with respect to the midline). These cauteries may have affected another independent source located in that region, such as the source postulated to exist in a homologous location in the mesothoracic segment (Fig. 4). Alternatively, cauteries above the spiracle may have acted as local sinks for a diffusible signal coming from the midline. Nijhout (1985a), in explaining the induction of supernumerary eyespots that resulted from cautery of the hindwing of Precis pupae, suggested that the inductive signal might be destroyed at the site of cautery due to wounding or wound healing, resulting in a sink. Thus in general, cauteries may have acted as sinks in both the prothoracic and

VOLUME

133,1989

mesothoracic experiments; in accordance with the observations of this study, such sinks would have had their most dramatic effects when located close to the source of the signal, as in the case of the mesothoracic experiment (Fig. 5B). Further

Evidence

That the Midline

Is an Active Source

To further examine the action of cauteries as sinks or barriers, a final experiment was carried out in which a circular region around the mesothoracic midline and its two flanking tubercles was cauterized. From the results of the previous experiment, it was predicted that such a cautery would allow development of the pupal commitment pattern to spread from the midline out to the limit imposed by the circular cautery, but no further. As predicted, 15 of 21 specimens had normal patterns of pupal cuticle within or up to the boundaries of the

A. CONTROLS

8. MESOTHORACIC

CIRCULAR

CAUTERY

FIG. 6. Results of circular cautery about the mesothoracic midline followed by treatment with the JH mimic, S-hydroprene, to reveal the state of commitment. (A) Control series. (B) Experimental series. Comparison of the mesothorax of experimentals and controls shows that, in all but the most advanced specimens, cautery provided a barrier preventing the spread of pupal commitment outside the circle. Cautery had no effect on pupal commitment within the circle. Of 21 specimens, 71% had normal patterns of pupal cuticle within the circle, 0% had pupal cuticle outside the circle alone, and 29% were unaffected. Note particularly the third member of the experimental series, in which the internal control segment (IC), the prothorax, indicates that pupal commitment would normally have spread out beyond the limits imposed by circle cautery.

CLAIRE KREMEN

Pupal

Ciwnmitment

cauterized circle (Fig. 6). Of these, the critical specimens were those in which the prothorax (the internal control segment) indicated that pupal commitment in the mesothorax would normally have proceeded beyond the limit imposed by the cautery (see the third member of the developmental series in Fig. 6B). Of the 6 specimens that had pupal cuticle outside of the circle boundary, all were more advanced in pupal commitment of the prothorax, suggesting that cautery must have occurred after the passage of the commitment signal. Prevention of the passage of the commitment signal by appropriately timed cauteries further confirms the hypothesis that the dorsal midline actively induced pupal commitment in adjacent epidermis, via an inductive signal travelling through the epidermal sheet. It is of interest to note that specialized morphological and developmental characteristics of the dorsal thoracic midline have also been noted for the cockroach, Gromphadorhina portentosa (Shelton, 1979), suggesting that the dorsal thoracic midline may have important developmental properties in other insects as well.

Patterns

in Precis

345

ecdysteroid to become committed (Truman et al., 1974; Fain, 1975). In this hypothesis, differential hormonal sensitivity to ecdysteroid could account for the temporal and spatial complexity of pupal commitment patterns which have been observed in a number of Lepidoptera (Fukuda, 1944; Riddiford, 1972; Sehnal and Schneiderman, 1973; Truman et al., 1974; Yin and Chaw, 1984; Ohtaki et al, 1986). Alternatively, pupal commitment could result simply from the decline of the JH titer below a threshold (Ohtaki et al., 1986; Kremen and Nijhout, 1989); tissue-specific differences in threshold sensitivity could generate a temporal/spatial sequence of pupal commitment in response to a declining JH titer. The results of the present study imply that in P. coenia, such a fixed spatial pattern of ecdysteroid requirements or JH thresholds does not exist; rather, the ontogeny of pupal commitment is established dynamically by cell-cell communication. If the ontogeny of pupal commitment resulted instead from a predetermined pattern of differential hormonal sensitivities, then cauteries would have created local defects in the Segment Independence pattern, rather than affecting regions larger than the cauterized area. Instead, the size and distribution of The method of observing the effects of cautery on regions of larval tissue observed after cautery suggest pupal commitment required the assumption that segthat cell-cell communication is responsible for the obments acted independently of one another with respect served ontogeny of pupal commitment. to the ontogeny of pupal commitment. This assumption The role of a communication signal in the ontogeny of was supported by the observation that in cauterized pupal commitment might be to induce cells to destroy individuals, unoperated segments were correlated with JH receptors (O’Connor, 1983) and/or to synthesize cyone another in their levels of pupal commitment, in the toplasmic JH-specific esterases (Sparks, 1984), thereby same way as in uncauterized larvae. causing cells to lose sensitivity to JH, and ultimately Segment independence could result from reduced leading to pupal commitment. The process of cell-cell gap-junction permeability at segment boundaries, as communication might itself be sensitive to the levels of has been observed in other insects (Warner and JH (Kremen and Nijhout, 1989), possibly beginning only Lawrence, 1982; Blennerhassett and Caveney, 1984). In when the JH titer fell below a threshold and terminataddition, much evidence exists to support the current ing if the JH titer then rose again above threshold. In view of the insect segment as a communication comthis manner, the level of JH would ultimately control partment and developmental field, including studies of the onset and progress of the sequence of pupal comcell lineage (Lawrence, 1975), gene expression (Lewis, mitment. 1978; Kornberg et al., 1985), and pattern formation It is possible also that ecdysteroid influences the pro(Locke, 1960; Lawrence, 1973; Nibbler-Jung and Grau, gress of pupal commitment by increasing the conduc1987). Indeed, many models of pattern formation in the tance of junctional membranes in the epidermis, insect segment require the assumption of segment inthereby facilitating intercellular communication. In the dependence (Lawrence, 1973; Sander and Nibbler-Jung, mealworm beetle, Tenebrio molitor, the junctional con1981). ductance of abdominal epidermis was increased by 66% after exposure to 20-hydroxyecdysone in vitro, a reClassical Endocrine Models of Pupal Commitment sponse which was specific to 20-hydroxyecdysone (CaPrevious workers have hypothesized that pupal com- veney and Blennerhassett, 1980). Changes in junctional mitment of the epidermis occurs in response to a surge conductance during development were also observed in in the concentration of ecdysteroid in the absence of JH vivo (Caveney, 1978); these changes could have resulted (Riddiford, 1976, 1978) and that different regions of the from alterations in both pore size and number of gap epidermis require different durations of exposure to junctions in the lateral membranes of epidermal cells

346

DEVELOPMENTAL

BIOLOGY

(Caveney, 1978; Caveney and Blennerhassett, 1980). Thus increases in the ecdysteroid titer during development may facilitate intercellular communication leading to pattern formation (Caveney, 1978); accordingly, the role of 20-hydroxyecdysone in committing the epidermis to pupal development in the absence of JH (Riddiford, 1976,1978) could be reinterpreted as facilitating pupal commitment by enhancing cell to cell communication. In fact, JH and 20-hydroxyecdysone might act as antagonists to pupal commitment by having opposite effects on the conductance of junctional membranes, explaining both the ability of 20-hydroxyecdysone to promote pupal commitment (Riddiford, 1976,1978) and of JH to prevent the onset and progress of the sequence of pupal commitment. The author thanks Dr. H. Frederik Nijhout for his numerous insights and suggestions given during the course of this study, and for his critical readings of this manuscript. The comments of two anonymous reviewers improved the quality of this manuscript. Drs. N. A. Granger, M. M. Nijhout, P. Mabee, M. D. Rausher, and V. L. Roth commented on earlier drafts of this manuscript. S-Hydroprene was a gift of Dr. D. A. Schooley (Zoecon). This study was supported by NSF Grant DCB-851’7210 to H. F. Nijhout and by a NSF Graduate Fellowship, a James B. Duke Fellowship, and a Sigma Xi Grant-in-Aid to the author. REFERENCES BLENNERHASSETT, M. G., and CAVENEY, S. (1984). Separation of developmental compartments by a cell type with reduced junctional permeability. Nature &on&m) 309,361-364. CAVENEY, S. (1978). Intercellular communication in insect development is hormonally controlled. Science 199,192-195. CAVENEY, S., and BLENNERHASSETT, M. G. (1980). Elevation of ionic conductance between insect epidermal cells by P-ecdysone in vitro. J Insect Physiol. 26, 13-25. FAIN, M. J. (1975). “Endocrine Physiology of Larval Molting in the Tobacco Hornworm.” Ph.D. thesis, Harvard University, Cambridge, MA. FAIN, M. J., and RIDDIFORD, L. M. (1977). Requirements for molting of the crochet epidermis of the tobacco hornworm larva in viva and in vitro. Wilhelm Roux’s Arch. Dev. BioL 181,285-307. FUKUDA, S. (1944). The hormonal mechanism of larval molting and metamorphosis in the silkworm. J. Fuc. Sci. Imp. Univ. Tokyo Sect. .4, 6,477-532. GRANGER, N. A., and BOLLENBACHER, W. E. (1981). Hormonal control of insect metamorphosis. In “Metamorphosis” (L. I. Gilbert and E. Frieden, Eds.), pp. 105-138. Plenum, New York. KORNBERG, T., SIDEN, I., O’FARRELL, P., and SIMON, M. (1985). The engrailed locus of Drosophila: In situ localization of transcripts reveals compartment-specific expression. Cell 40,45-53. KREMEN, C. (1987). Metamorphosis of the butterfly, Precis coenia, (Nymphalidae): Commitment of the imaginal disks and epidermis to pupal development. Ph.D. thesis, Duke University, Durham, NC. KREMEN, C., and NIJHOUT, H. F. (1989). Juvenile hormone controls the onset of pupal commitment in the epidermis and imaginal disks of Precis coenia (Lepidoptera: Nymphalidae). J. Insect Physiol., in press. KUHN, A., and VON ENGELHARDT, M. (1933). Uber die Determination des Symmetriesystems auf dem Vorderfltigel von Ephestia kiihniellu. Wilhelm Roux’s Arch. 130, 660-703.

VOLUME

133.1989

KURUSHIMA, M., and OHTAKI, T. (1975). Relation between cell number and pupal development of wing disks in Bombyx mori. J. Insect Physiol. 21, 1705-1712. LAWRENCE, P. A. (1973). The development of spatial patterns in the integument of insects. In “Developmental Systems: Insects.” (S. J. Counce and C. H. Waddington, Eds.), pp. 157-211. Academic Press, London/New York. LAWRENCE, P. A. (1975). The structure and properties of a compartment border: The intersegmental boundary in Gncepeltus. In “Cell Patterning,” Ciba Foundation Symposium, pp. 3-16. Associated Scientific Pub., Amsterdam. LEWIS, E. B. (1978). A gene complex controlling segmentation in Drosophila. Nature (London) 276,565-570. LOCKE, M. (1960). The culticular pattern in an insect-The intersegmental membranes. J. Exp. Biol. 37,398-406. MITSUI, T., and RIDDIFORD, L. M. (1976). Pupal cuticle formation by Manducu se&a epidermis in vitro: patterns of ecdysone sensitivity. Dev. Biol. 54,172-186. NIJHOUT, H. F. (1976). The role of ecdysone in pupation of Manduca sexta. J. Insect Physiol. 22,453-463. NIJHOUT, H. F. (1980). Pattern formation on lepidopteran wings: Determination of an eyespot. Dev. Biol. 80,267-274. NIJHOUT, H. F. (1985a). Cautery-induced colour patterns in Precis coenia (Lepidoptera: Nymphalidae). J EmbryoL Exp. MorphoL 86, 191-203. NIJHOUT, H. F. (1985b). The developmental physiology of color patterns in Lepidoptera. In “Advances in Insect Physiology” (J. E. Treherne, M. J. Berridge, and V. B. Wigglesworth, Eds.), Vol. 18, pp. 181-247. Academic Press, London/New York. NIJHOUT, H. F., and WHEELER, D. E. (1982). Juvenile hormone and the physiological basis of insect polymorphisms. Q. Rev. Biol. 57, 109-133. NOBLER-JUNG, K., and GRAU, V. (1987). Pattern control in insect segments: Superimposed features of the pattern may be subject to different control mechanisms. Wilhelm Roux’s Arch. 196,290-294. O’CONNOR, J. D. (1983). Ecdysteroid and juvenile hormone receptors. In “Endocrinology of Insects” (R. G. H. Downer and H. Laufer, Eds.), pp. 559-565. A. R. Liss, New York. OBERLANDER, H. (1969). Effects of ecdysone, ecdysterone, and inokosterone on the in vitro initiation of metamorphosis of wing disks of Gall&a mellonella. J. Insect PhysioL 15, 297-304. OBERLANDER, H., and SILHACEK, A. L. (1976). Action of juvenile hormone on imaginal discs of the Indian meal moth. In “The Juvenile Hormones” (L. I. Gilbert, Ed.), pp. 220-232. Plenum, New York. OHTAKI, T., YAMANAKA, F., and SAKURAI, S. (1986). Differential timing of pupal commitment in various tissues of the silkworm, Bombyx mori. J. Insect Physiol. 32, 635-642. RIDDIFORD, L. M. (1972). Juvenile hormone in relation to the larvalpupal transformation of the cecropia silkworm. Biol. Bull. 142, 310-325. RIDDIFORD, L. M. (1976). Hormonal control of insect epidermal cell commitment in vitro. Nature (London) 259,115-117. RIDDIFORD, L. M. (1978). Ecdysone-induced change in cellular commitment of the epidermis of the tobacco hornworm, Manduca sexta, at the initiation of metamorphosis. Gen. Comp. Endocrinol. 34, 438-446. RIDDIFORD, L. M. (1982). Changes in translatable mRNAs during the larval-pupal transformation of the epidermis of the tobacco hornworm. Dev. BioL 92,330-342. SAFRANEK, L., and WILLIAMS, C. M. (1984). Critical weights for metamorphosis in the tobacco hornworm, Manduca sextu. Biol. Bull. 167, 555-567. SANDER, K., and NOBLER-JUNG, K. (1981). Polarity and gradients in

CLAIRE

KREMEN

Pupal

Commitment

insect development. In “International Cell Biology 1980/81” (H. G. Schweiger, Ed.), pp. 497506. Springer-Verlag, Berlin/Heidelberg. SEHNAL, F. (1968). Influence of the corpus allatum on the development of internal organs in Galleria mellonella L. J Insect Physiol. 14,73-85.

SEHNAL, F., and SCHNEIDERMAN, H. A. (1973). Action of the corpora allata and of juvenilizing substances on the larval-pupal transformation of Galleria mellonella L. (Lepidoptera). Acta Ent. Bohemoslnu. 70,289-302. SHELTON, P. M. J. (1979). Post-embryonic determination of the ecdysial line in the cockroach: Evidence for pattern regulation in the medio-lateral axis. J. EmbryoL Exp. Morphol. 49, 27-46. SPARKS, T. C. (1984). Effects of JH 1 and the anti-JH fluoromevalonate on development and JH esterase activity in postfeeding laststadium larvae of Trichoplusia ni (Hubner). J. Insect Phgsiol. 30, 225-234.

Patterns

347

in Precis

TRUMAN, J. W., RIDDIFORD, L. M., and SAFRANEK, L. (1974). Temporal patterns of response to ecdysone and juvenile hormone in the epidermis of the tobacco hornworm, Munduca se&a. Dev. Bid. 39, 247-262.

WARNER, junctions

A. E., and LAWRENCE, P. A. (1982). Permeability at the segmental border in insect epidermis.

of gap Cell 28,

243-252.

WIGGLESWORTH, V. B. (1934). The physiology prolixus (Hemiptera). II. Factors controlling morphosis.” Q. J. Microsc. Sci. 77, 191-222. YAMAMOTO, R. T. (1969). Mass Larval rearing and pupation.

of ecdysis moulting

in Rho&Gus and “meta-

rearing of the tobacco hornworm. J. Econ. EntomoL 62,1427-1431.

YIN, C-M., and CHAW, W-D. (1984). Haemolymph and epidermal cell commitment of the female borer larvae. J. Insect Ph ysiol. 30,257-263.

ecdysteroid southwestern

II. titre corn