The mechanism of evagination of imaginal discs ofDrosophila melanogaster

Wilhem Roux' Archly 178, 123--138 (1975) �9 by Springer-Verlag 1975

The Mechanism of Evagination of Imaginal Discs of Drosophila melanogaster

I I . Studies on Tryps in-Acce le ra ted E v a g i n a t i o n

Eva Fekete, Dianne Fris t rom *, I s tvan Kiss, and James W. Fr is t rom *

Institute of Genetics, Biological Research Center, Hungarian Academy of Sciences, Szeged, Hungary

l%eceived December 16, 1974 Accepted in revised form May 21, 1975

Summary. The effects of trypsin treatment on the in vitro evagination of imaginal discs under different conditions are investigated. It is found that trypsin accelerates the eva- gination of discs which have previously been exposed to ecdysone in vivo or in vitro. Sub- stances which inhibit ecdysone-induced (unaccelerated) evagination, such as CytochMasin B, Concanavalin A and Mycostatin, also inhibit trypsin-accelerated evagination.

On the cellular level, evagination is associated with the flattening of the disc cells. How- ever, immature discs (i.e., those which have not been exposed to ecdysone) and discs pretreated with Cytochalasin B do not evaginate in response to trypsin even though pronounced cell flattening occurs. Cell flattening is an energy requiring process since it does not occur in response to trypsin treatment in the presence of oligomycin or nitrogen. We conclude that cell flattening is an active process that takes place during evagination but which does not itself produce evagination. An alternative mechanism for evagination may involve cell rearrange- ment. Trypsinization could facilitate both cell flattening and cell rearrangement by reducing intercellular adhesiveness.

Imaginal leg and wing discs occur in the hemocoel of the mature larva as f lat tened sac-like structures consisting of a folded, single cell thick columnar epithelium, the disc proper, which is continuous with a squamous epithelium, the peripodial membrane. I t is the disc proper t ha t during metamorphosis unfolds and extends (evaginates or everts) to form an adult appendage. The process of evaginat ion provides a convenient system for the s tudy of morphogenesis, since isolated larval discs undergo evagination i~ vitro when exposed to the molting hormone fl-ecdysone (Fristrom et al., 1970; Mandaron, 1973; Milner and Sang, 1974).

There is a striking change in the shape of the cells of the disc proper from columnar to cuboidal during evagination, which led Fris trom et al. (1970) and Poodry and Schnciderman (1970, 1971) to propose tha t it is the cause of eva- gination. Support ing this possibility was the impor tan t discovery by Poodry and Schneiderman (1971) tha t mild trypsinization of mature larval discs produces evagination similar to tha t produced by ecdysones. They noted tha t trypsini- zation causes cell f lattening presumably as a result of reduction in intercellular adhesivity, and concluded tha t evagination is a result of passive cell flattening. However, observations on disc evagination by Fris t rom and Fr is t rom (1975)

* Current address: Department of Genetics, University of California, Berkeley, California 94720.

124 E. Fekete et al.

indicate t ha t the process is more complex t han this. These authors showed tha t in the presence of Cytochalasin B evaginat ion was inhibited, bu t cell f la t tening still occurred under some conditions. Fur thermore, they noted t ha t it is no t s t ructura l ly feasible to derive a narrow, elongated structure such as a leg from a concentrically folded disc by cell f la t tening alone. They therefore postulated t h a t cell rear rangement might be involved in evaginat ion.

The current work was under taken to s tudy in more detail the process of evaginat ion in response to t rypsin. We have tested the effects of a n u m b e r of drugs known to inhibi t normal evaginat ion including Cytoehalasin B (Fristrom, i972; Mandaron and Sengel, 1973) ; Concanaval in A (Mandaron, 1974) ; inhibi tors

of RNA and protein synthesis (Fristrom et al., 1973) and inhibi tors of energy metabolism. We conclude t ha t the process of evaginat ion in the presence of t ryps in is similar to, bu t much faster than, tha t which occurs in its absence and,

like normal evaginat ion, requires an ini t ia l exposure to eedysone. The results also show tha t cell f la t tening by itself does no t result in evaginat ion and tha t both cell f la t tening and evaginat ion are energy dependent processes.

Material and Methods

Art Oregon R strain of Drosophila melanogaster was used throughout these experiments. Larvae were grown at room temperature in a standard medium. For most experiments discs from late third instar larvae and white prepupae were dissected in Robb's medium and transferred promptIy to the experimental conditions. To study the effects of trypsinization on cultured discs, late third instar mass-isolated (Fristrom, 1972) or dissected discs were in- cubated in Robb's medium (Robb, 1969) with 0.1 E~g/ml fl-ecdysone at 25~ for approximately 10 or 20 hours, which produces partial or complete evagination, respectively. Conditions for trypsinization were similar to those of Poodry and Schneiderman (1971). Discs were exposed to 0.1% trypsin (2 • recrystallized, Sigma Chemical Co.) in Robb's medium for 10 minutes at room temperature except where otherwise indicated. Discs were viewed and photographed before and after trypsinization using a Leitz photomicroscope. Inhibitors were obtained from commercial sources and were applied to discs in Robb's medium immediately prior to trypsin treatment. Exposure of discs to a nitrogen atmosphere was accomplished by placing the discs in a drop of culture medium on a microscope slide in a sealed box through which a constant stream of nitrogen was passed.

Microscopy. Discs were rinsed briefly in two changes of 0.1 M sodium eacodylate buffer (pI-I 7.4) and transferred to fixative, l~ixation was carried out for 1-2 hours on ice in a com- bination of 1% glutaraldehyde and 1% osmium in 0.075 M sodium cacodylate buffer, pH 7.4. Discs were post fixed in 1% osmium in 0.5 N buffer. Dehydration was in ethanol and embedding in Araldite.

Estimates of cell height were made by light microscope examination of 1 iz thick sections stained with 1% toluidine blue in 1% sodium borate. Since the disc epithelium consists of a single cell layer, cell height can be determined by measming epithelial thickness in sections cut perpendicular to the plane of the epithelium. Cell height estimates were based on at least 30 measurements from each of at least three wing discs taking care to measure equivalent areas in experimental and control discs.

Sections for electron microscopy were cut with glass knives on a Porter Blum MT1 ultra- microtome, stained with uranyI acetate and lead citrate and examined in a Jeol electron microscope.

Results

The Quantitation o/Evagination Occurring in Trypsin

The evaginat ion produced by incubat ing competent leg discs in t ryps in for 10 minu tes is depicted in Fig. 1. I n the studies described below, i t was necessary

Trypsin-Accelerated Evagination 125

Fig. 1 a and b. A leg disc from a white prepupa (a) before and (b) after treatment with 0.1% trypsin (w/v) in Robb's medium for 10 minutes. • 700

to have a means for evaluating the degree of evagination produced by trypsini- zation. We have, therefore, adopted the use of the Evagination Index (Chihara et al., 1972) in which the degree of evagination is scored integrally from 0, no evagination, to 10, complete evagination. In Fig. 1 the Evagination Indices of the discs before and after trypsinization are respectively 3 and 8.

The Role o/the Physiological Age o/the Discs in Producing Evagination in Trypsin

Poocb'y and Schneiderman (1971) indicated that trypsinization produced greater evagination of mature discs than of immature discs. Since mature discs have probably been exposed to ecdysone in vivo, this raised the possibility tha t successful trypsin-produced evagination (Evagination Index greater than 6) requires prior exposure to ecdysone. In vitro, the degree of evagination achieved by a disc is a function of the time of exposure to ecdysone (Borst et al., 1974). We assume that the degree of evagination of dissected discs is an indication of the duration of exposure to ecdysone in vivo. We have, therefore, determined the

10

8

i

x tlJ a Z

Z 6 O l--

Z

w 4 - - J

Z iT

126 E. Fekete et al.

10

tD LU (D

6--

z

Z

LJ ~9

2Z LLJ rr Ls tl- LI- r~

O + I I I I 0

0 1 2 3 4

INITIAL EVAGINATION INDEX

Fig. 2. The relationship between the initial Evaginatioa Index and the final Evagination Index obtained by incubating discs from larvae or white prepupae in 0.1% trypsin (w/v) in l~obb's medium for 10 minutes. The brackets represent standard deviations. Difference in

Evagination Indices --~--, Evagination Index - - . - -

amount of evagination produced by trypsin as a function of the initial degree of evagination. The results, depicted in Fig. 2, indicate tha t substantial evagination is only produced when the initial Evaginat ion Index is 2 or more (equivalent to an in vitro exposure to fl-ecdysone of 5 hours) and tha t virtually no evagination occurs when the Evaginat ion Index is 0 (equivalent to less than 2 hours in vitro

exposure to fl-ecdysone). We therefore conclude tha t prior exposure to ecdysone is required for evagination to occur in trypsin.

In a second experiment, paired discs dissected from white prepupae were divided into two groups. The first group was trypsinized and the second group incubated in Robb 's medium in the absence of t rypsin or exogenous hormone. The results (Table 1) show tha t these discs which are capable of evgginating in t rypsin in 10 minutes, can also evaginate in Robb ' s medium without additives in 2-3 hours (but not in Ringers or Ringers plus glucose).

Trypsin-Accelerated Evagination 127

Table 1. Capacity of discs capable of trypsin-accelerated evagination to evaginate in the absence of trypsin

Incubation conditions Evangination Index for given times of incubation

Initial 10 min 3 hrs 6 hrs

1. Robb's + Trypsin 2.8=[_0.57 7.7=t=1.17 (0.1%, wlv)

2. Robb's 2.8 ~ 0.76 - -

3. Ringer's 2.44-0.50 - -

4. P~inger's + Glucose 2.3:t:0.47 - - (2 mg/ml)

6.8• 8.8~0.96

3.6• 3.6•

2.7~0.94 2.7~0.94

Pairs of discs were dissected from white prepupae and one was exposed to trypsin and the other placed in Robb's medium, Ringer's plus glucose or Ringer's. At least 5 pairs of discs were scored in each experiment. I t should be noted that trypsin-accelerated cvagina- tion occurs in Ringer's or Ringer's plus glucose. Standard deviations are given.

In the light of the above results we adopt the view that trypsin greatly ac- celerates the normal process of evagination induced by/~-ecdysone in vivo. Thus, in the remainder of the paper we will refer to evagination in trypsin as trypsin- accelerated evagination, and that which occurs in vitro without trypsin as unac- celerated evagination.

Effect of Trypsinization of Immature Larval Discs Following Incubation in vitro with fl-eedysone

The ability of trypsin to accelerate evagination in mass-isolated, third instar leg discs which had been incubated in vitro with fl-ecdysone was also examined. The results are depicted in Fig. 3 where the Evagination Indices before and after trypsin t rea tment are again compared. In this ease, trypsin only has a dramatic accelerating effect if the initial Evagination Index is greater than 5, i.e., those discs in which the peripodial membrane has been ruptured by the evaginating appendage. This result is presented pictorially in Fig. 4 where the leg which has already broken through the peripodial membrane (leg " b ") undergoes substantial elongation while leg ' : a " is confined to the disc lumen by the peripodial mem- brane. During in vitro culture in Robb's medium, the opening in the stalk where the disc was originally attached to the epidermis heals over and encloses the part ly everted appendage in a sealed cavity. When discs are exposed to trypsin immediately Mter dissection the channel through the stalk is still open, allotting unimpeded elongation of the appendage. Hence, the difference in response to trypsin of discs exposed to fl-ecdysone in vivo and in vitro appears to involve the presence of a physical impediment in the latter.

Effects of [nhibitors of Evagination on Trypsin-accelerated Evagination

The conditions required for evagination in vitro in response to fi-ecdysone have been extensively studied. I t is an active process requiring the continued synthesis of RIgA and proteins (Fristrom et al., 1973) and is inhibited by a number of drugs

128 E. Fekete et al.

I0

9

X W r,, 8 Z

Z o 7

Z

~' 6

W

_J 5

Z ,T

INITIAL EVAGINATION INDEX

1/3 W

C3 Z

5 - Z O

4 Z

(D ~> 3 w

Z

W 2 c )

Z W r r w u_ u_

Fig. 3. The relationship between initial and final Evagination Indices for third instar larval discs incubated for five hours in i~obb's Medium containing 0.1 Fgm/mi fl-ecdysone and then

trypsinized as in Fig. 2. The brackets represent standard deviations

including CytochMasin B (CB) (Fristrom, 1972; Mandaron and Sengel, 1973), Concanavalin A (Con A) (Mandaron, 1974) Mycostatin (lZaikow and Fristrom, 1971). I f trypsin accelerates evagination, then drugs which affect the mechanism of evagination per se (rather than, for example, maintenance of the disc) should be equally effective against both trypsin-accelerated and unaccelerated evagination.

I. E/leers o/Inhibitor$ o / R N A and Protein Synthesis. Inhibitors of RNA and protein synthesis (Actinomycin D, puromycin and cycloheximide) were tested against trypsin-accelerated evagination in concentrations known to inhibit normal evagination. As can be seen from Table 2, inhibition of I%NA or protein synthesis has no immediate effect on trypsin-accelerated evagination. These results arc not surprising since trypsin-accelerated evagination occurs too rapidly for new macro- molecular synthesis to be a significant part of the process.

Juvenile hormone also does not affect trypsin-accelerated evagination but inhibits normal evagination. Juvenile hormone inhibits protein synthesis in discs (Fristrom, unpublished observations) and thus its inhibition of unaccelerated evagination can probably be attr ibuted to inhibition of protein synthesis.

II . Elleet o[ Cytochalasin B. Cytoehalasin B reversibly inhibits disc evagination in Drosophila (Fristrom, 1972; Mandaron and Sengel, 1973) and in Galleria (Hasskarl et al., 1973). This is also true of trypsin-accelerated evagination. Fig. 5 shows tha t the dose response curves for the inhibition of ev~gination by CB are similar for trypsin-accelerated and unaccelerated evagination. In both eases the inhibition is reversible within about 30 minutes after the removal of CB.

Trypsin-Accelerated Evagingt ion 129

Fig. 4a - - e . Effect of trypsinization on larval leg discs after five hours in culture (as in Fig. 3). (a) Before trypsin t rea tment . Note t ha t disc " a " is still surrounded by the peripodial mem- brane, pro. (b) The same two discs 10 minutes and (c) 60 minutes after addit ion of 0.1% trypsin. Only disc " b " shows substantial increase in length and reduction in width (note

reduction in diameter of the basitarsus +->). • 700

I30 E. Fekete et al.

Table 2. Effects of inhibitors that block unaeeelerated evagination on trypsin-accelerated evagir~tion

Inhibitor Concentration Evagination Index

Initial Final

Aetinomyein D 1 ~g/ml a, b 3.0 4- 0.89 7.4 4-1.5 20 rain before trypsin 10 vg/ml 2.8 • 0.98 7.t • 1.5

Puromyein 0.1 ~xg/ml 3.4 • 8.8 =t= 1.6 20 min before trypsin 1 ~g/ml a 3.04-1.0 7.5•

10 ~xg/ml 3.2 4-1.6 7.2 4- 1.6

Cycloheximide 0.1 ~xg/ml a 3.0 ~= 0.89 8.0 4- 2.5 20 rain before trypsin 1 ~g/ml b 2.5• 7.54-1.7

Juvenile hormone c 100 ag/ml a 3.04-0.89 7.24-0.98 10 min before trypsin

a Minimal concentration tested necessary to inhibit evagination (Fristrom et al., 1973). b Minimal concentration tested necessary to inhibit 95 % of incorporation of precursors into I~NA and protein (Fristrom et al., 1973), Standard deviations are given. Trypsin treatment (0.1% in gobb's) was for 10 min. e The juvenile hormone used was the Manduca C16 juvenile hormone.

Ultrastruetural observations (Fristrom and Fristrom, 1975) have shown tha t there is an increase in CB sensitive microfilaments associated with the basal cell surface of disc cells in response to fi-eedysone in vitro. We also find basal micro- filaments in partially evaginated discs dissected from white prepupae. However, preliminary observations of trypsin-accelerated discs have failed to reveal sig- nificant numbers of microfilaments. This is probably due to the inferior preser- vat ion obtained after trypsinization, especially of the basal cell surface where the filaments are predominant ly located. However, discs t reated with 0.2 ~zgms/ml CB followed by trypsinizat ion showed characteristic surface bulges containing filamentous material (Fig. 6). Similar fi lamentous bulges occur in high concen- trat ions of CB alone (Fristrom and Fristrom, 1975). These are commonly inter- preted as aggregates of nonfunctional filaments. Thus, CB seems to have similar effects at the ul t rastructural level on normally evaginated and trypsin-accelerated discs. Whether these ul t rastructural changees are directly related to inhibition of evagination remains to be determined.

I I I . Ef[ects of Con A and Mycostatin. Con A is a plant lectin which binds specifically to e-D manno- or ~-D glucopyranoside-like sites at the cell surface (Goldstein et al., 1965). Trea tment of discs with Con A irreversibly inhibits eva- gination (Mandaron, 1974). I t also inhibits trypsin-accelerated evagination (Table 3) al though in this case the effect is partially reversible after one hour, presumably due to partial hydrolysis of the bound Con A by trypsin. Trehalose, a nonreducing ~, e-disaeeharide prevents Con A inhibition of trypsin-accelerated evagination (Table 3) indicating tha t the binding of Con A specifically involves ~-glucosides or mannosides. Trehalose also prevents Con A inhibition of unac- eelerated evagination (Fristrom, unpublished observation). Agents which bind to microtubule protein (colehieine, colcemid and vinblas t ine)have been shown to

Trypsin-Accelerated Evagination 131

X LIJ 1:3 Z

z 0 I - , . .< z

X ILl

l:3 u.I N , - J

~E r 0 Z

lO

2

...... I,,,,

~t / i i i

0 o01 0.050.1

\

&

0.5 1 ~J 10 C O N C E N T R A T I O N CB ( /ug /ml )

Fig. 5. The effects of different concentrations of CB on trypsin-accelerated (--,--) evagination In the experiments on trypsin-accelerated evagination, the CB was added 20 minutes prior to trypsinization and the discs scored 10 minutes a~fter addition of 0.1% trypsin in Robb's medium. In the experiments on unaccelerated evagination (--,--), the CB was added at the start of the incubation of mass-isolated larval discs in Robb's medium with ~-eedysone (0.1 Fgm/ml) and the discs scored after 20 hours of culture. Direct comparison of inhibition by CB of trypsin-aceelerated and unaceelera.ted evagination was accomplished by "normaliz- ing" the trypsin results. This was accomplished by expanding and revaluing the Evagination Index scale change of 3 to 8 (typical for trypsin-accelerated evagination; see Fig. 2) to 0 to

l0 as in commonly used for unaeeelerated evagination.

reverse the effects of Con A (Yin et al., 1972; Edelman et al., 1973). Pre t rea tment of discs with colcemid eliminates the inhibitory effect of Con A on evagination (Table 3).

The antibiotic Mycostat in also interacts with the cell surface of sensitive organisms (see Kinsky, 1966 for review). Myeostat in inhibits normal evagination (l~aikow and Fristrom, 1971) and trypsin-accelerated ewgina t ion (Table 3). Thus, both these agents which i n t e r a c t specifically with the cell surface are effective inhibitors of evagination.

l~equirement of Energy for Trypsin-Accelerated Evaginat ion

Energy is clearly required for the normal evagination of discs in response to /~-ecdysone since continued R N A and protein synthesis are necessary. I n the ease

3 Wilhe lm l~oux's Archives

132 E. Fekete et al.

Fig. 6. Electron micrograph of a white prepupal leg disc treated for 20 minutes in 0.2 ,~gm/ml CB followed by 10 minutes in 0.1% trypsin. Note the surface bulge eontMning filamentous

material, f. •

Table 3. Effects of Concanavalin A and Mycostatin on trypsin-induced ewgination

Condition of incubation Evagination Index

Initial Final

gobb's 2.2 • 2.67 =1-0.94

Robb's -- Trypsin 2.4 =t-0.85 7.11~1.65

l~obb's @ Con A (10 ~g/ml, I0 rain) 2.55=~0.83 3.1 ~1.67 then Trypsin

Robb's + Con A (as above) 2.55=~ 0.83 4.7 • 1.3 then Trypsin (60 rain)

Robb's + Trehalose (3.5 ~xg/ml) and 2.8 =c0.91 6.8 i1 .12 Con A (10 gg/ml 10 rain) then Trypsin

l~obb's -k Colcemid (10 lzg/ml, 10 rain) 2.3 • 0.88 6.8 :[:0.9 then Con A (10 btg/ml, 10 min) then Trypsin

Robb's q- Mycostatin (10 rain) 2.5 ~-0.81 3.2 ~-1.25 then Trypsin

At least 6 discs were used in each experiment. Incubations with trypsin (0.1%) were for 10 min except as noted. Standard deviations are given.

of trypsin-accelerated evagination, no macromolecular synthesis is required and so it is possible to determine whether the cellular processes involved in the me- chanics of evagination are also energy requiring. Leenders st al. (1974) demon- strated that oligomycin and incubation in a nitrogen atmosphere were the agents most effective in reducing ATP levels in cul tured Drosophila sal ivary glands.

Trypsin-Accelerated Evagination ~[33

50 u') z 0 r~

.~ 40 7

I--- n- o 30 i l l n- - . J ...J tlJ U 20 . . J

- . l

w -r

10 n w

I N=2

._J r r

Z n " C

T

I ~=7 o~

n

0_ OJ r l - 0 -

Z

D- 2~ O - - i

O

Z O o

Fig. 7. A comparison of the cell height of discs following different treatments (described in Fig. 3 and Tables 3 and 4). Estimates of cell height were made as described in Materials a.nd Methods. All measurements were made on discs dissected from white prepupae except where otherwise indicated, and all discs except the controls were trypsinized as previously de-

scribed. Note that only the trypsinized prepupat discs (column 3) evaginated

The results presented in Table 4 shout that both oligomycin and exposure to a nitrogen atmosphere inhibit t rypsin-accelerated evagination. I n the case of nitrogen exposure the inhibition is reversible after 60 minutes in air, showing tha t no permanent damage was done to the discs. These results demonstra te tha t trypsin-accelerated evagination is not simply a passive process resulting from decreased intercellular adhesiveness.

Studies on Cell Flat tening

As noted in the Introduct ion, the most conspicuous change at the cellular level during evagination is cell f lattening (Fig. 7). The importance of cell f lattening for evagination was determined by measuring the change in epithelial cell height (see Material and Methods) after trypsinization under a var ie ty of conditions. I n agreement with Poodry and Sehneiderman (197t) we find tha t cell f lat tening occurs during trypsin-accelerated evagination (Figs. 8 and 9). The degree of f lat tening is much more pronounced in wing discs than in leg discs and so the cell height data presented here (see below) are restricted to ~dng discs, al though leg discs show similar but less extreme changes.

There are two instances in which cell f lattening occurs wi thout evagination. First, larval discs (Evaginat ion Index of 0-2) which were not able to respond to

134 E. Fekete et al.

ad

7

I

Fig. 8. A longitudinal section of a white prepupal wing disc; wp, wing pouch; ad, adepithelia cells. • 1800

Fig. 9. A longitudinal section of a white prepupal wing disc after 10 minutes in 0.1% trypsin. Note the decrease in the height of the epithelium el. Fig. 8. There are a number of morpholog- ical abnormalit ies characteristic of t rypsin evaginated wings; e.g., the wing pouch, wp has folded back towards the anterior end (identified by the adepithelial cells, ap). The dorsal and ventra l surfaces of the wing pouch fail to come together as in normal evagination (see

Fr is t rom and Fristrom, 1975 for comparison). • 1800

Fig. 10. A section of a white prepupal wing disc after t r ea tment with CB and t rypsin as in Fig. 6. Note the convoluted, f lat tened epithelium. • 1800

Trypsin-Accelerated Evagination 135

Table 4. Effects of inhibitors of ATP production on trypsin-induced evagination

Incubation condition Evagination index

Initial Final a

Robb's

P~obb's -~ Trypsin (0.1% )

Robb's 5- Oligomycin (10 txg/ml 20 rain) then Trypsin (0.1%)

Robb's ~- Nitrogen (20 min) then Trypsin (0.1% )

gobb's 5- Nitrogen (as above), incubation in air (60 min) then Trypsin (0.1%)

2.4~0.67 2.7~0.94

2.85-0.98 6.9•

2.7~1.05 3.15-0.74

2.85-0.87 3.05-1.0

2.5• 6.65-1.8

At least 9 discs were scored in each experiment. a Incubation with trypsin was for 10 rain. Standard deviations are given.

trypsin by evaginating frequently showed substantial cell flattening when compared to untreated controls. In these eases the increased surface area was accompanied by an increase in the number and/or depth of the folds of the disc. Secondly, when discs are treated with CB before trypsinization, cell flattening not only occurs, but is more extreme than that produced by trypsin alone (Fig. 7). This extreme flattening rather than resulting in evagination produces a highly convoluted epidermal structure (Fig. 10).

Con A inhibits both cell flattening and evagination (Fig. 7). Also, somewhat unexpectedly, energy deprivation by oligomycin treatment or incubation in a nitrogen atmosphere prevents cell flattening. This indicates that cell flattening itself is an energy dependent process.

Discussion

Our first conclusion from the observations presented here is that evagination produced by trypsinization of discs is an acceleration of the normal process already induced by fl-ecdysone. Only discs capable of evaginating in the absence of trypsin undergo rapid evagination in response to trypsin. The mechanisms involved in trypsin-accelerated evagination are presumably similar to those oc- curring during unaccelerated evagination. This is supported by the fact that inhibitors of unaccelerated evagination such as CB, Con A and Mycostatin also inhibit trypsin-accelerated evagination. Thus, trypsin treatment provides a tooi for studying the mechanics of evagination.

We can draw two main conclusions concerning the mechanism of evagination from the data presented here. First, cell flattening alone is not sufficient to produce evagination since some larval discs and CB treated discs undergo pronounced cell flattening in response to trypsin without evaginating. Second, cell flattening and evagination are energy dependent processes since neither occurs in the presence

136 E. Fekete et al.

of inhibitors of ATP production (oligomycin and nitrogen). We now consider evagination as involving at least two activities : cell flattening which provides the increased surface area of the evaginated disc and cell rearrangement, which we propose is responsible for the actual evagination. The occurrence of cell rearrange- ment was previously deduced from a study of the morphological changes which occur during evagination (Fristrom and Fristrom, 1975). There is an increase in the length of the appendage and a decrease in the diameter which cannot be explained by cell flattening alone. This is illustrated here in the trypsinization of a partially evaginated leg disc where the appendage becomes much longer and narrower within minutes (Fig. 4).

Mild trypsinization results in a reduction in intercellular adhesiveness (Poodry and Schneiderman, 1971). A reduction in adhesiveness between disc cells could hypothetically facilitate both cell flattening and cell rearrangement since in both cases a change in the area and/or position of intercellular contacts is required. Indeed, it is interesting to speculate as to whether some kind of proteolytic act ivi ty is a normal component of the evagination process in vivo. An increase in proteolytic activity of the hemolymph associated with the breakdown of larval tissues following pupariation seems reasonable and could explain the difference in the in vivo and in vitro kinetics of evagination (the late phases of evagination occur about twice as fast in vivo as unaccelerated evagination in vitro). Thus, partial loosening of intercellular at tachments by proteolysis might be a natural component of the evagination process.

The effects of some inhibitors of evagination tested here, CB, Con A and Mycostatin, are similar for trypsin-accelerated and unaccelerated evagination indicating tha t the mechanism of evagination is similar in both casts. The me- chanism of action of CB on normally evaginated discs has been discussed at length (Fristrom and Fristrom, 1975) and similar considerations apply here. The mode of action of Con A on cells deserves further comment. This plant lectin binds to the outer membrane of animal cells and can produce agglutination in certain types of cells; e.g., transformed cells (Inbar and Sachs, 1969) and embryonic cells (Moscona, 1971). Pretreatment of cells with agents which bind microtubule protein (colchicine, colcemid and vinblastine) can prevent agglutination by Con A (Yin et al., 1972). Burger and Noonan (1970) demonstrated tha t partial t ryptic digestion of Con A also resulted in loss of its agglutination properties, although it did not interfere with the binding of the lectin to the cell surface. In accordance with these observations, we find tha t t reatment of discs with colcemid prevents the Con A inhibition of evagination and that complete Con A inhibition is par t ly released after one hour presumably due to partial trypsinization of the bound Con A. Both of these observations are consistent with the view tha t the Con A in- hibition of evagination involves the agglutination of disc cells by the lectin thus preventing the cell dissociation we have proposed as necessary for cell flattening and cell rearrangement. However, it is also conceivable that Con A acts on discs by a more subtle effect on the cell surface. Edelman et al. (1973) have shown that Con A inhibits the mobility of surface receptors including Con A receptors on the surface of lymphocytes. This effect is also reversed by colcemid. I t is possible that the movement of some specific surface receptors play a role in evagination, especially with regard to establishing the final position of the cells in the disc.

Trypsin-Accelerated Evagination 137

W e can as ye t only speculate as to the cellular processes which occur in con- junc t ion with the reduct ion in in tercel lular adhesiveness to effect evaginat ion. However , two pieces of evidence suggest t h a t an in t race l lu lar contract i le mecha- nism m a y be involved. F i r s t ly , cell f la t ten ing and evagina t ion are A T P requir ing processes and secondly, microfi l~ments having some of the proper t ies of act in- l ike f i laments have been observed in evagina t ing discs (Fr i s t rom and Fr i s t rom, 1975). I n view of the widespread occurrence of ac tomyos in med ia t ed contract i le processes in nonmuscle cells (see Po l la rd and Weihing, 1974 for review), the opera t ion of some contract i le process in evagina t ion would no t be surprising. However , the existence and precise na tu re of contract i le processes in discs remains to be de- mons t ra ted . F ina l ly , the inhibi t ion of evag ina t ion by agents such as Con A and Mycosta t in , which in te rac t specif ical ly wi th the cell surface, indica tes the im- por tance of the p lasma membrane in this process. This aspect of evagina t ion has been emphasized by Mandaron (1974). The quest ion of whether changes a t the cell surface involve only quan t i t a t i ve changes in adhesiveness, or whether more specific surface in terac t ions are also involved, identif ies ano ther a rea for in- ves t iga t ion into the mechanism of evaginat ion.

We are indebted to Dr. Ferenc Joo for the use of his laboratory facilities. This work was supported in part by a grant from the United States Public Health Service (GM-19937) and through the auspices of UNDP Project HUN/71/506 from UNESCO.

References

Borst, D.W., Bollenbacher, W.E., O'Connor, J.D., King, D.S., Fristrom, J.W. : Ecdysone levels during metamorphosis of Drosophila melanogaster. Develop. Biol. 39,308-316 (1974)

Burger, M.M., Noonan, K.D.: Restoration of normal growth by covering of agglutinin sites on tumour cell surface. Nature (Lond.) 228, 512-515 (1970)

Chihara, C.J., Petri, W.It . , Fristrom, J.W., King, D.S.: The assay of ecdysones and ju- venile hormones on Drosophila imaginal disks in vitro. J. Insect Physiol. 18, 1115-1123 (1972)

Edelman, G.M., Yahara, I., Wang, J .L.: Receptor mobility and receptor-cytoplasmic interactions in lymphocytes. Proc. Nat. Acad. Sci. (Wash.) 70, 1442-1446 (1973)

Fristrom, D., Fristrom, J. W. : The mechanism of evagination of imaginal discs of Drosophila melanogaster. 1. General considerations. Develop. Biol. 48, 1-23 (1975)

Fristrom, J.W. : The biochemistry of imaginal disk development. In: Results and problems in cell differentiation: The biology of imaginal disks (H. Ursprung and R. NSthinger, eds.), vol. 5, p. 109-154. Berlin-Heidelberg-New York: Springer 1972

Fristrom, J.W., Logan, W.R., Murphy, C. : The synthetic and minimal culture requirements for evagination of imaginal discs of Drosophila melanogaster in vitro. Develop. Biol. 88, 441 456 (1973)

Fristrom, J.W., gaikow, g . , Petri, W., Stewart, D. : In vitro evagination and P~NA synthesis in imaginal discs of Drosophila melanogaster. In: RNA in development. (E. W. Hanly, ed.), Park City Symposium on Problems in Biology. p. 381 401. Salt Lake City, Utah: Uni- versity of Utah Press 1970

Goldstein, I .J . , Hollerman, C.E., Smith, E. E. : Protein-carbohych'ate interaction. II. In- hibition studies Oll the interaction of Concanavalin A with polysaceharides. Biochemistry (Wash.) 4, 876-883 (1965)

Hasskarl, E., 0berlander, H., Stephens, P~.E. : Microtubules and tracheole migration in wing disks of Galleria mellonella. Develop. Biol. 83, 334-343 (1973)

Inbar, )s Sachs, L. : Interaction of the carbohydrate-binding protein Concanavalin A with normal and transformed cells. Proc. Nat. Acad. Sci. (Wash.) 63, 1418-1421 (1969)

Kinsky, S.C., Luse, S.A., van Deenen, L.L.M.: Interaction of polyene antibiotics with natural and artificial membrane systems. Fed. Proc. 25, 1503-1510 (1966)

138 E. Feke~ et al.

Leenders, H.J., Kemp, A., Koninkx, J .F.J .G. , Rosing, J.- Changes in cellular ATP, ADP and AMP levels following treatments affecting cellular respiration and the activity of certain nuclear genes in Drosophila salivary glands. Exp. Cell Res. 86, 25-30 (1974)

Mandaron, P." Effects of ~-ecdysone, fl-ecdysone and inokosterone on the in vitro evagination of Drosophila leg discs and the subsequent differentiation of imaginal integumentary structures. Develop. Biol. 81, 101-113 (1973)

Mandaron, P. : Sur le m6canisme de l'@vagination des disques imaginaux de Drosophile cultiv@s in vitro: Effects de diverses substances affectant la membrane eellulaire. Wilhelm Roux' Archiv 175, 49-63 (1974)

Mandaron, P., Sengel, P. : Effects of cytochalasin B on the evagination in vitro of leg imaginal discs. Develop. Biol. 3~, 201-207 (1973)

Milner, M.J., Sang, J.H. : Relative activities of ~-ecdysone and fl-ecdysone for the differen- tiation in vitro of Drosophila melanogaster imaginal discs. Cell 3, 141-143 (1974)

Moscona, A.A.: Embryonic and neoplastic cell surfaces: Availability of receptors for Con- canavalin A and wheat germ agglutinin. Science 171, 905-906 (1971)

Pollard, T.D., Weihing, R.R.: Actin and myosin and cell movement. Critical Reviews of Biochem. 2, 1-65 (1974)

Poodry, C.A., Schneiderman, H.A.: The ultrastructure of the developing leg of Drosophila melanogaster. Wilhelm Roux' Archly 166, 1-44 (1970)

Poodry, C.A., Schneiderman, H.A.: Intercellular adhesivity and pupal morphogenesis in Drosophila melanogaster. Wilhelm Roux' Archiv 168, 1-9 (1971)

Raikow, R., Fristrom, J.W. : Effects of fl-ecdysone on R•A metabolism of imaginal discs of Drosophila melanogaster. J. Insect Physiol. 17, 1599-1614 (1971)

Robb, J.A. : Maintenance of imaginal discs of Drosophila melanogaster in chemically defined media. J. Cell Biol. 41, 876-884 (1969)

Yin, H.tt . , Ukena, T.E., Berlin, R.D. : Effect of eolchicine, colcemid, and vinblastine on the agglutination, by Concanavalin A, of transformed cells. Science 178, 867-868 (1972)

Eva Fekete Department of Genetics University of California Berkeley, California 94720 USA