J. Cell Sci. 50, 225-243 (1981) 225Printed in Great Britain © Company of Biologists Limited 1081

THE SECRETION OF THE EGGSHELL OFSCHISTOCERCA GREGARIA. ANALYSIS OFTHE KINETICS OF SECRETION IN VITROBY LIGHT AND ELECTRON MICROSCOPEAUTORADIOGRAPHY

S. J. KIMBER»Department of Zoology, University of Cambridge

SUMMARYThe secretion of the 2 main layers (endochorion and exochorion) of the eggshell of the desert

locust Sckistocerca gregaria was investigated using light and electron microscope autoradio-graphy. Follicles undergoing endochorion secretion were labelled using a 3 min ' pulse' of["HJleucine in vitro followed by a o—115 min non-radioactive 'chase'. Immediately after thepulse the silver grains were distributed over the cytoplasm and organelles including roughendoplasmic reticulum, while by 2 and 5 min Golgi bodies contained radioactivity. By 12 minfrom the beginning of the chase the cell apex containing small secretory vesicles was labelled.By 20 man most of the silver grains were over the endochorion. The half-transport time (tM)was 14-15 min (from mid pulse), the lag time was 9-10 min and the percentage transport ratewas 14-15 % per min. When a 3 min pulse of pHJgalactose was used to label exochorion pre-cursors, the shorter tK (11 min) and the clumped grain distribution in light microscope auto-radiographs after o-min chase suggested that galactose was incorporated in Golgi bodies. Thesecretion of exochorion precursors appears to occur at a similar rate to that of endochorionprecursors (approximately 15 % per min). The results indicate that the follicle cells are amongthe fastest secreting cells.

INTRODUCTION

In a previous paper (Kimber, 1980) the ultrastructure of eggshell secretion in thedesert locust Schistocerca gregaria was described. In this report the kinetics of secretionof the endochorion and the exochorion, the 2 major layers of the eggshell, are in-vestigated using light microscope (LM) and electron microscope (EM) autoradio-graphy. Chorion proteins make up the major proportion of the proteins synthesizedby the cells at this time and they are secreted by the synchronized epithelium over arigidly defined time-scale (Paul, Goldsmith, Hunsley & Kafatos, 1972; Kimber, 1980).In addition, the secretory product is deposited directly adjacent to the cell from whichit is derived, and the programme of eggshell secretion .continues autonomously withthe same time-scale in vitro as in vivo (Paul & Kafatos, 1975; Kafatos et al. 1977;Margaritis, Kafatos & Petri, 1980; Mazur, Regier & Kafatos, 1980; Nadel & Kafatos,1980; Regier, Mazur & Kafatos, 1980). These criteria make this system ideal for thestudy of the dynamic aspects of the secretory process, using a precursor-product model.

• Present address: A.R.C. Institute of Animal Physiology, Animal Research Station,307 Huntingdon Road, Cambridge CB3 oJQ, England.

226 S.J.Kimber

In this paper quantitative light microscope autoradiography has been used to assessthe kinetics of the secretory process. Similar studies have been carried out in only afew other insect cells (Kafatos & Kiortsis, 1971; Lai-Fook & Gupta, 1976; Couble,Prudhomme & Daillie, 1977; Blau & Kafatos, 1978, 1979; Gupta, unpublished). Thepresent study reveals the extreme rapidity of the secretion of the locust eggshell.

MATERIALS AND METHODS

Preparation for pulse-chase experiments with radioactive precursors

Tissue culture medium 199 (Tc 199, Flow Laboratories, Irvine, Scotland) was made upusing Hepes buffer, gassed with 95 % air, 5 % CO! and the pH corrected to give a pH of 6-8after sterilization through a o-22-fim Millipore filter. Ovarioles with terminal follicles under-going chorion secretion were dissected from the ovaries of 12- to 14-day-old female locusts ona cold stage. They were stored under Tc 199 on ice while additional ovaries were dissected.

A single ovariole from each ovary was removed to 2 % formaldehyde in sodium cacodylatebuffer for the assessment of background radiation. The remaining ovarioles were incubated inlocust Ringer solution: 0-105 M-NaCl, o-oi M-KCI, io~* M-CaC^, 2 x io~s M-MgCU, 0-015 M-NaHgPO«, 0-0175 M-NajHPO^ (Weis-Fogh & Gupta, unpublished, modified from Weis-Fogh,1956) for 5-12 min on ice followed by 5-6 min at 35 °C. They were then subjected to a 3-8 minradioactive 'pulse' at 35 °C in locust Ringer solution containing 50 /*Ci/ml [*H]leucine (sp.act. 56 Ci/mmol) 0r2oo/*Ci/ml [8H]galactose (sp. act. 12 Ci/mmol). After three 0-5-1 min rinsesin Tc 199 they were incubated in Tc 199 (which contains 0-46 mM-leucine) for the' cold chase',also at 35 °C. When fHJgalactose was the precursor 4 mM-galactose was added to the Tc 199(which lacks galactose) used for the cold chase. After 0-115 min cold chase ovarioles were fixedin 2 % formaldehyde at room temperature (2x1 min + 28 min) to minimize non-specific bindingof PH]leucine and possibly [3H]galactose (Peters & Ashley, 1967; Williams, 1973; Meyrick &Reid, 1975). This was followed by 2-5-4 h fixation in 2-5 % glutaraldehyde in sodium cacodylatebuffer, and the ovarioles were processed further for electron microscopy as previously described(Kimber, 1980).

In describing the results the following terminology has been used: a 3 min radioactive pulseincubation in pi-Tjleucine, 1-5 min rinse and 12 min non-radioactive chase incubation has beenreferred to as 3L*+ 1-5 +12 min. When [*H]galactose was used as the radioactive precursorthe letter G replaces L.

Preparation for light and electron microscopy

All glassware was cleaned as recommended by Rogers (1973). About 30 o-$-fim or i-o-/*msections from each follicle were cut on a Cambridge ultramicrotome and distributed between3 slides for LM autoradiography.

For electron microscope autoradiography pale-gold 100-nm-thick sections were cut on aReichert ultramicrotome. The sections were deposited on celloidin-coated slides (Salpeter &Bachmann, 1972) and stained when completely dry in uranyl acetate and lead citrate. Theywere coated with an approx. 50-nm layer of carbon from a vacuum coating unit (Salpeter &Bachmann, 1964).

Slides were coated with Ilford L4 emulsion (average crystal diameter 0-14 /*m) melted at45 °C and diluted 2:3 for LM autoradiography or between 1:3 and1! 12-5 for EM autoradio-graphy with 3 x distilled water containing 0-005 % sodium lauryl sulphate (Moses, 1964). Theslides were coated using an automatic coating device (Kopriwa, 1966; Gupta, Moreton &Cooper, 1973) at 31 °C for LM autoradiographs and at 32 °C for EM autoradiographs. Thedilution used for EM autoradiographs gave an emulsion layer > 150 nm thick, with purple/blue interference colours, which was observed in the EM to be an overlapping monolayer ofsilver halide crystals.

Slides were dried vertically for 1 h at 18-20 °C, 60-70% relative humidity and transferredto light-tight boxes containing dried silica gel, which were sealed and stored at 4 °C.

Kinetics of secretion of locust eggshell 227

Exposure for LM autoradiographs was for 5-7 days after [*H]leucine-labelling and 10-14days after ["HJgalactose-labelling. Exposure for EM autoradiographs was for 5-6 months.

LM slides were developed in Kodak Dio,b developer for 3-5 min and fixed in 25 % sodiumthiosulphate. The sections were viewed under phase contrast and then stained in o-i % meth-ylene blue in 1 % borax and mounted under 80% glycerol (Caro, 1964). Staining caused nodiminution or movement of silver grains.

EM autoradiographs were developed using 2 or 5 min gold latensification (James, 1948;Wisse & Tates, 1968; Salpeter & Bachmann, 1972; Gupta et al. 1973) followed by Elon ascorbicacid with sodium sulphite (EAS) for 4-6 min (Salpeter & Szabo, 1972). Fixation was in 25 %sodium thiosulphate for 2-3 min. Slides were soaked for 1-12 h in 3 x distilled water at 4 °Cto aid floating off the celloidin film (Salpeter & Bachmann, 1964). Acetone-cleaned grids weredeposited over the sections and the films picked up by immersion using a clean wire gauze.The EM autoradiographs were examined in a Philips EM 200 electron microscope operatedat 60 kV.

Quantitative analysis of LM autoradiographs

Silver grains were counted and areas measured for each source compartment on light micro-graphs at a magnification of approximately x 8000. The source compartments considered were:(1) nucleus, (2) total cytoplasm, (3) basal cytoplasm, (4) apical cytoplasm, (5) endochorion and(6) exochorion (where present).

Because of the variability in the incorporation of [*H]leucine and PHJgalactose (see Tables 1,2) it was necessary to normalize the grain distribution by expressing the number of grains inany compartment as a percentage of the total grains in each field. Each field, comprising a cellplus that portion of the chorion adjacent to it, was assumed to be a closed system. Possibledistortions of the data introduced by considering percentages (Salpeter, 1973) were checkedfor by plotting the data as relative grain density (percentage grains divided by percentagearea) against chase time (Kimber, 1979). This gave curves with similar shapes to those obtainedwhen percentage grains were plotted against chase time. The grain density over the folliclecells in unlabelled control autoradiographs was only 0-25 grains/100/tm1.

Considering the system on a precursor-product model, radioactive material lost from thecytoplasm is transferred to the chorion with first-order kinetics:

-dA/dt = dB/dt,

where A is the radioactive protein in the cytoplasm, B is that in the chorion and t is time.Strictly, the radioactivity per volume relative to the concentration of chorion protein in cyto-plasm and chorion should be considered (Droz, 1976). However, assuming that the specificradioactivity (grain density divided by precursor macromolecule density) remains constant,valuable information can be obtained from a precursor-product analysis using percentage data.

RESULTS

Preliminary experiments indicated that prechorionated follicles would survive forat least 5 h in Tc 199 without noticeable deterioration (observed by light and electronmicroscopy).

A few follicles undergoing vitellogenesis were incubated in pHjleucine-containingmedium followed by non-radioactive chase. The radioactive precursor was taken upand incorporated by the follicle cells but in no case did radioactive molecules pass intothe extracellular space or the yolk.

Qualitative analysis of endochorion secretion

Figs. 1-6 show light microscope autoradiographs from follicles subjected to 3L* +1-5 + 0 — 55 min. In follicles fixed immediately after a 1-5 min rinse silver grains were

228 S.J.Kimber

Kinetics of secretion of locust eggshell 229

distributed throughout the cytoplasm (Fig. 1). In EM autoradiographs silver grainswere found to be scattered throughout the cell, over rough endoplasmic reticulum(RER), Golgi bodies and cytoplasm (Fig. 7). In LM autoradiographs clumping ofgrains was observed by 3L*+ 1-5 + 2 min (Fig. 2) and was significant by 3L*+ 1-5 +5 min (Fig. 3). EM autoradiography revealed that the silver grains were predomi-nantly associated with the convex face of Golgi bodies adjacent to RER by 3L* + 1 -5 +2 min (Fig. 8) and were localized towards the concave face after 3!,*+1-5+ 5 min(Fig. 9). After 12 min chase (Fig. 4) a large number of grains were localized over theapical cytoplasm and the hah0 of the endochorion adjacent to the follicle cells. In EMautoradiographs silver grains were found over small electron-dense vesicles at theconcave face of Golgi bodies as well as over the cell apex (Fig. 10). Quantitative ultra-structural analysis (Kimber, 1979; and unpublished) indicated that the grains overthe apex originate from radioactivity in small secretion vesicles. Most of the radio-activity had entered the chorion by 3L*+ 1-5 + 20 min chase (Figs. 5, n ) . Even after55 min (Fig. 6) and 115 min chase (not shown) a few silver grains remained over thecytoplasm.

Quantitative analysis of endochorion secretion

The average incorporation of [3HJleucine per follicle cell varied amongst folliclesfrom a single ovary (Table 1). However, since there was no progressive increase inthe number of grains or in the average grain density with chase time after [3H]leucine-labelling, the experimental regime seemed to provide an effective pulse of radio-activity.

Figs. 1-6 show light microscope autoradiographs of cells from follicles pulse-labelledfor 3 min in fHJleucine. en, endochorion; fc, follicle cell. Ilford L4 emulsion, 0-5 fimEpon sections stained with methylene blue, x 3000.

Fig. 1. A follicle cell and adjacent endochorion after 3L# +1-5+0 min. Silvergrains are localized over the follicle-cell cytoplasm and there are few over either thenucleus or the endochorion.

Fig. 2. A follicle cell and adjacent endochorion after 3L* + 1-5 + 2 min. The silvergrains are over the follicle cell cytoplasm and have a clumped distribution. Thenucleus and endochorion are unlabelled.

Fig. 3. A follicle cell and adjacent endochorion after 3L# +1 -5 + 5 min. The clump-ing of silver grains over the follicle-cell cytoplasm is very prominent. The endochorionis still unlabelled.

Fig. 4. A follicle cell and adjacent endochorion after 3L* + 1-5 + 12 min. Silvergrains are concentrated over the apical cytoplasm of the follicle cell and over the endo-chorion adjacent to the follicle cell. However, considerable numbers of silver grainsremain over the rest of the cytoplasm.

Fig. 5. A follicle cell and adjacent endochorion after 3L* + i s + 2O min. Themajority of the silver grains are now over the endochorion. The silver grains are distri-buted over the entire width of the endochorion, but have a greater density adjacentto the follicle cell.

Fig. 6. A follicle cell and adjacent endochorion after 3L* + 1 -5 + 55 min. Silver grainsare distributed throughout the endochorion but are more concentrated adjacent tothe follicle cell. The silver grains over the cytoplasm probably represent PHJleucineincorporated into cytoplasmic protein.

230 SJ. Kimber

Kinetics of secretion of locust eggshell 231

Table 1. Total areas, number of grains and overall grain density for follicle cells andchorion after a 3 min pulse of [3H]leucine and 0-55 min chase

Chase time(min)

02

S122 0

55

Total gainscounted

325826331807272331533267

Av. total grainsper cell plus

chorion (±S.E.)

407 ±19220 ± 19

201 ± io272 ±19350 ±40327 ±29

Av. total areaper cell plus

chorion(jim* ± s.E.)

312 ±20250 ± 14237 ±i°240! 10

269 ±11350 ±29

Av. overall graindensity (cells +

chorion)(grains//im')

i-3o±o-O3O-92±O-OIo-86±o-oz1-13 ±0-02I"3O±O-O2O'94±O'Oi

Background 0-25 grains per 100 fim1 over cells in unlabelled follicle; 9—11 cells analysed foreach chase time. Silver grains were counted and areas measured on micrographs at x 8000magnification using a Perspex overlay. The results were processed using a Quantimet 720.Average overall grain density was obtained by dividing average total grains per cell pluschorion by average total area per cell plus chorion.

The percentages of grains in different compartments for a representative follicleare plotted in Fig. 13. After 3L* + 1-5 + 0 min only 2% of the grains were over thechorion while 94% were over the cytoplasm. After 2 min and 5 min chases there waslittle change in grain distribution. In contrast, by 3L*+ 1-54-12 min almost 40% ofthe grains were over the chorion and 57 % over the cytoplasm. By3L*+i-5 + 55 minthe endochorion contained 59% of the radioactivity and the cytoplasm 36%. A similarpercentage of radioactivity was retained in the cytoplasm after 55 and 115 min chases(Kimber, 1979) and this was considered to be due to the incorporation of ^HJleucineinto cytoplasmic proteins. Therefore, the percentages of grains representing radio-active protein destined for the chorion are plotted in a precursor-product analysis inFig. 14 on the assumption that all radioactive chorion precursor protein had left thecell by 55 min chase.

Data from 6 ovaries indicate that chorion protein synthesis accounts for 60-80%of the total protein synthesized during chorion secretion. This assumes that newlysynthesized protein is not broken down during the course of the experiment, theleucine contents of cytoplasmic and chorion proteins are similar and the same

Figs. 7-11 show EM autoradiographs from follicles pulse-labelled for 3 min in [*H]-leucine; 100 nm Epon sections stained with uranyl acetate and lead citrate, coatedwith Ilford L4 emulsion and developed in EAS. x 19000.

Fig. 7. EM autoradiograph showing a region of a follicle cell after 3L# + 1-5 +0 min.The silver grains are distributed throughout the cytoplasm and some are in thevicinity of RER.

Fig. 8. EM autoradiograph showing a region of a follicle cell after 3L# + 1-5 + 2 min.The silver grains are mainly over the periphery of the convex face (arrows) of theGolgi bodies (g) rather than over the electron dense material at the concave face.

S.J. Kimber

Kinetics of secretion of locust eggshell 233

cytoplasmic pool of amino acids is used for the synthesis of cytoplasmic and chorionproteins.

Three factors are considered in interpreting the curves in Fig. 14 (Kafatos &Kiortsis, 1971): (1) the half-transport time (tM), which is the time measured from themiddle of the radioactive-pulse incubation within which 50% of the radioactiveproduct has been secreted into the chorion; (2) the lag phase, which is the minimumtime taken from the middle of the radioactive pulse for the radioactive product toappear in the chorion; (3) the percentage transport rate measured from the maximumslope of the precursor-product curve. From Fig. 14 and Table 1, and from data on3 other ovaries it appears that during mid to late endochorion formation the percentagetransport rate is 14-15% per min, the tM is 14-15 min and the lag time 9-10 min.

Qualitative analysis of exochorion secretion

When follicles were incubated in [3H]leucine during exochorion formation radio-activity was found in both exochorion and endochorion after 25 and 55 min chase(Fig. 12). Histochemical studies (Kimber, 1979) suggested the presence of poly-saccharide in the exochorion but not endochorion. Therefore, [3H]galactose was usedin order to investigate exochorion secretion without interference from endochorionsecretion, which continues in parallel.

Figs. 15-20 show cells from follicles incubated for 3G*+ 1-5 + 0 — 55 min. Therewas a clear increase in the number of grains over the pale-staining exochorion materialduring this period. After o, 2 and 5 min non-radioactive chase (Figs. 15, 16, 17) mostof the incorporated radioactivity was still in the cell. Even after o min chase the silvergrains had a clumped distribution suggesting that the precursor was incorporated inGolgi bodies. After 3G#+ 1-5 + 12 and 3 C + 1-5 + 20 min (Figs. 18, 19) a significantnumber of grains were associated with the interface between cell and exochorion.However, after 3G* +1-5 + 55 m m (Fig- 20) the grains were mainly over the exo-chorion, where they tended to be deposited in a layer.

Quantitative analysis of exochorion secretion

Table 2 shows data derived from a representative ovary subjected to a 3 min pulseincubation in [sHJgalactose-containing medium followed by 0-55 min chase in non-radioactive medium. The average overall grain density varied between differentfollicles from the same ovary but there was no consistent increase with chase time.

Fig. 21 shows the percentage of grains over different compartments plotted against

Fig. 9. EM autoradiograph showing a region of a follicle cell after 3L* + 1 -5 + 5 min.The silver grains are now over Golgi saccules, which contain electron-dense material.Fig. 10. EM autoradiograph showing part of a follicle cell + chorion after 3L* +1-5 + 12 min. The silver grains are mainly over the extreme cell apex and over theendochorion (en) adjacent to the follicle cell. Note that the large mosaic vesicles (v)do not appear to be labelled. Inset: the Golgi region still contains radioactive materialafter 12 min chase in some cases. Silver grains occur over electron-dense spherulesbudded from Golgi saccules.

234 S.J.Kimber

J

Kinetics of secretion of locust eggshell 235

Table 2. Total areas, number of grains and overall grain density for follicle cells andchorion after a 3 min pulse of \?H]galactose and 0-55 min non-radioactive chase

Chase time(min)

Total grainscounted

Av. total grainsper cell plus

chorion (± s.E.)

Av. total areaper cell plus

chorion(/»m' ± s.E.)

Av. overall graindensity (cells +

chorion)(grains//im*)

o 873 58 ±6 328114 018 ±0-0042 657 55 ±3 4°2±i5 0-1410-0025 671 6717 334±H 0-2010-005

12 1587 I34±I5 362121 0-3710-00720 1269 H5±7 377±i8 o-3o±o-oo455 1068 i°7±4 339±i9 o-30±o-oo2

Background 0-24 grains per 100 fim* over cells in unlabelled follicle; 10-15 cells analysedfor each chase time. Silver grains were counted by eye. Areas were calculated using a Quantimet720, which was also used to process the results. Average overall grain density was obtained bydividing average total grains over cell plus chorion by the average total area per cell pluschorion. Note the low average grain density compared to that after [*H]leucine-labelling.

time. After 3G* + 1-5 + 0 and 3 C + 1-5 + 2 min over 90% of the silver grains wereover the cytoplasm. The percentage of grains over the exochorion begins to rise after5 min chase and increases dramatically between 3G*+i-5 + 5 min and 3G* + i-5 +12 min. By 12 min chase 58% and by 55 min chase 84% of the silver grains wereover the chorion. In contrast after pHJleucine incubation only 25-37% of the radio-activity had passed to the endochorion after 12 min chase. The possibility of error inallocation of grains at the cell/exochorion border (especially after 12 min chase) isreduced because, in most cases, there is a small gap between the cell and exochorion.After pHjgalactose-labelling the silver grains over the endochorion never exceeded6% of the total.

The shape of the precursor-product curve is also different after pHJgalactoseincubation (compare Fig. 14 and Fig. 22). After pHJgalactose-labelling there was aslow linear release of a further 20 % radioactive product into the exochorion between3G*+ 1-5 + 20 min and 3G#+ 1-5 + 55 min (Fig. 22). This might indicate that 2 com-ponents with different secretory kinetics are produced.

Fig. 11. EM autoradiograph of the apical region of a follicle cell and adjacent chorionafter 3L*+ 1-5 + 20 min. Almost all the silver grains are localized over the endo-chorion (en), mainly in the half adjacent to the follicle cells. The cytoplasm is almostfree of silver grains in the field shown. Some do occur, however, over the microvillibetween small secretion vesicles in the process of being released. The mosaic vesicles(v) in the cytoplasm are unlabelled. Thin endochorion tubules (tu) are. visible adjacentto the follicle cell microvilli, while thicker ones are found further from the cell.

Fig. 12. Light microscope autoradiograph of a follicle cell during exochorion forma-tion after 7L* + 5 + 25 min. Silver grains are present over the follicle cells (Jc) exo-chorion (ex) and endochorion (en). Section stained with methylene blue,. Ilford K2emulsion, x 4000.

236 S.J.Kimber

100

90

80

70c

a> 60

ior 50o

40

30

20

10

0_L _l

0 2 5 12 20 30 40 50 55Chase time (min)

Fig. 13. The percentage of the total silver grains over different regions of the cell andthe chorion as a function of chase time after incubation with ["HJleucine. The barrepresents the pulse incubation period in PH]leucine (3 min). The first ovariole wasfixed after a 1-5 min rinse (o min chase point). # — # , total cytoplasm; • — • , apicalcytoplasm; A—A, endochorion; • — • , basal cytoplasm; x — x , nucleus.

Less than 10% of the silver grains were found over the cytoplasm after 55 minchase. Therefore, it was assumed that this residual radioactivity represented incorpora-tion into cytoplasmic and membrane proteins. The real value is probably between0% (Fig. 21) and 10% (Fig. 22). Both Figs. 21 and 22 indicate a lag time of approxi-mately 6 min and a t60 of 11 min. The percentage transport rate in the rapid phase ofsecretion is between 13% and 16% per min, similar to that of pHJleucine-labelledendochorion precursors.

DISCUSSION

The most striking feature of the secretion of chorion protein precursors in S. gre-garia is its rapidity. tM is probably the best measure of secretion speed. It is influencedboth by the time required for the first radioactive chorion precursors to reach thechorion, the lag phase, and by the percentage transport rate. The lag phase includesthe minimum time required for processing polypeptide chains in the Golgi apparatus,for the transport of precursor to the apical membrane and for exocytosis. The per-centage transport rate is influenced by the speed of all transport steps, by the different

Kinetics of secretion of locust eggshell 237

3CD

co

a

Chase time (min)

Fig. 14. The percentage of the silver grains representing chorion material in the cyto-plasm = (the percentage of the total grains in the cytoplasm at a particular time— the percentage of total grains in cytoplasm after 55 min chase) multiplied by 100,divided by the percentage of grains representing chorion material in cytoplasm andchorion. The bar represents the pulse incubation period in PHJleucine (3 min). <M>the half-transport time = 14 min; lag time = 9 min. O—O, total cytoplasm; A—A,endochorion.

distances precursor must move from Golgi bodies to the apical membrane, and alsoby the length of the pulse (Kafatos & Kiortsis, 1971).

It is interesting that insect cells such as the ovarian follicle cells in silkmothssecreting eggshell proteins with a t60 of 20-25 In^n a* 25 °C (Blau & Kafatos, 1978),the epidermal cells of S.gregaria secreting resilin with a tw of 12-14 min at 37 °C(Lai-Fook & Gupta, 1976; Gupta, unpublished data) and the ovarian follicle cells ofS.gregaria (the present study) are capable of secreting proteins faster than manysecretory systems in vertebrates. Estimates of /M from quantitative and semi-quantitative data (see Case, 1978) and those values of tso that are available in theliterature suggest that 30 min is about the minimum tM value for secretion by verte-brate cells. However, a faster secretion rate is not a phylogenetic characteristic ofinsects since silkmoth colleterial glands show a t^ of 3-4 h (Grayson & Berry, 1974)and silkmoth galea, secreting cocoonase zymogen, have a tM of 38-74 min (Kafatos &Kiortsis, 1971) while in Bombyx silk gland, silk fibroin is secreted with a t6Q of approxi-mately 45-50 min (from data of Couble et al. 1977).

238 S.J. Kimber

Kinetics of secretion of locust eggshell 239

The similar rates of secretion of exochorion glycoprotein and endochorion pre-cursor protein, seen in the present study, suggest that during the secretion of theexochorion both endochorion precursors and exochorion presursors are handled inparallel. Similarly, in the silkmoth secreted proteins are released with the samekinetics during the last three-quarters of the period of chorion formation in spite ofthe rapidly changing nature of the molecular species produced (Blau & Kafatos, 1978).It is assumed that the tso and lag time calculated for protein synthesis during exo-chorion secretion are the same as those during endochorion secretion, since thepercentage transport rates for both processes are approximately the same. The 3-4min shorter tM (11 min) and lag time (6 min) after using pHJgalactose from theseparameters after using [sH]leucine could be interpreted to mean that most galactoseis incorporated in the Golgi apparatus, as appears to occur in other cell types (e.g.see Leblond & Bennett, 1977). This is supported by the clumped distribution ofautoradiographic grains after [3H]galactose-labelling, even after o min chase. Assumingthat the protein is synthesized on ribosomes of the RER then 3-4 min will representthe time required for the synthesis of an average chorion protein and its transfer to theGolgi apparatus. This is supported by quantitative EM autoradiographic analysis(Kimber, 1979 and unpublished) and is of the same order of magnitude as the timeestimated in other cells (Jamieson, 1972; Case, 1978; Blau & Kafatos, 1978).

Figs. 15-20 show LM autoradiographs of cells from follicles pulse-labelled for 3 minin PH]galactose during exochorion formation, en, endochorion; ex, exochorion; fc,follicle cell. Ilford L4 emulsion, i-o-fim sections stained with methylene blue, x 3000.

Fig. 1 s. A follicle cell and adjacent chorion after 3G* + 1-5+0 min. The silver grainsare localized over the follicle-cell cytoplasm. They appear to be clumped (especiallyover the portion of cytoplasm to the left of the figure). The endochorion and exo-chorion are unlabelled.

Fig. 16. A follicle cell and adjacent chorion after 3G* + 1 '5 + 2 min. The silver grainsare over the cytoplasm of the follicle cells. Note the fragmented appearance of thetops of the endochorion ridges.

Fig. 17. A follicle cell and adjacent chorion after 3G# + i'5 + s min. The graindistribution is similar to that in Fig. 16. The endochorion ridge has been displacedfrom the interfollicle cell junction. This is a common occurrence as the follicle cellstake on a squamous shape during exochorion formation.

Fig. 18. A follicle cell and adjacent chorion after 3G* +1-5 + 12 min. The silvergrains are localized over the junction between the apical follicle-cell membrane andthe exochorion. Where these 2 structures have separated (right) an approx. equalnumber of silver grains is associated with each structure.

Fig. 19. A follicle cell and adjacent chorion after 3G* + 1-5 + 20 min. Most of thesilver grains are localized over the exochorion, mainly over the outer edge but a feware over the exochorion further from the cell. Some silver grains remain associatedwith the apical membrane of the follicle cells but the basal and lateral membranesare not labelled.

Fig. 20. A follicle cell and adjacent chorion after 3G* + i-5 + 55 min. Almost allthe silver grains are over the exochorion and few appear to remain associated withthe apical follicle cell membrane.

240 S.J.Kimber

Chase time (min)

Fig. 21. The percentage of the total grains over different regions of the cell and thechorion as a function of chase time after frTlgalactose incubation. The bar representsthe pulse incubation period in [*H]galactose (3 min). The first ovariole was fixedafter a 1-5 min rinse (o min chase point). 0 — # , total cytoplasm; A—A, exochorion;O—O, apical cytoplasm; • — Q , basal cytoplasm; x — x, nucleus; O—©,endochorion.

After [3H]leucine-labelling the disappearance of incorporated radioactivity frombasal follicle-cell cytoplasm was only slightly faster than from the apical cytoplasmindicating that exocytosis is a less time-consuming factor than Golgi processing. Thisis supported by results from EM autoradiography (Figs. 8, 9; Kimber, 1979 andunpublished). However, there was a much more rapid decrease in radioactivityincorporated in the basal cytoplasm compared to the apical cytoplasm after [8HJ-galactose-labelling. Since Golgi elements are prevalent in the basal cytoplasm (seefig. 9 of Kimber, 1980) as well as in the apical cytoplasm, this may indicate that thesecreted product is transported to the apex of the cell almost immediately after theaddition of galactose to exochorion molecules in Golgi bodies.

Although the first third of the endochorion thickness is secreted en masse (Kimber,1980), autoradiography indicates that further molecules are intercalated through thethickness of the shell during the subsequent phases of secretion (Figs. 5, 6). Duringexochorion formation, there may be greater localization of precursor macromoleculesin the endochorion nearer the follicle cells (Fig. 12). pHJleucine-labelled macro-molecules are added through the thickness of the exochorion but [3H]galactose-

Kinetics of secretion of locust eggshell 241

0 2 5 12 20 30

Chase time (min)

40 50 55

Fig. 22. The percentage of silver grains representing chorion material as a functionof chase time after fHJgalactose incubation. The percentage of silver grains repre-senting chorion material = (the percentage of total grains in cytoplasm at a particularchase time - the percentage of total grains in the cytoplasm after 55 min chase) multi-plied by 100, divided by the percentage grains representing chorion material in cyto-plasm + chorion. ?w> the half-transport time =11 min, approximate lag time =6 min. The bar represents the pulse incubation period in ["Hjgalactose (3 min). A—A,exochorion + endochorion; A—A, exochorion; • — 0 , total cytoplasm.

labelled macromolecules are predominantly localized adjacent to the follicle cells(Figs. 19, 20). Similarly, in the silkmoth follicles there are stage-specific patterns ofmacromolecular addition to the partially completed chorion (Blau & Kafatos, 1979).

This investigation of the overall kinetics of secretion of chorion proteins by thefollicle cells of S. gregaria has revealed the extreme rapidity of the process in commonwith secretion in certain other insect cells. The rapid secretion of certain proteins byinsect cells is probably necessary for the rapid growth rates that insects sustain. Thereproductive cycle is correspondingly short: oocyte growth and maturation in S.gregaria 19 accomplished in approximately 14 days. Thus, it is not surprising that thefinal phase in the maturation of the oocyte should take place in 30-36 h (Kimber,1980). In order for the 20-40 /im thick eggshell to be secreted in such a short spaceof time it is necessary for the proteins to be secreted at rates that are probably nearthe upper limit for secretory cells.

242 5. J. Kimber

The author wishes to thank Dr B. L. Gupta for much invaluable advice, Dr D. W. Parryfor providing facilities in the Zoology Department, University of Cambridge, and Mr M. J.Day, Mr N. Maskell, Mr J. Rodford and Mr G. Runnalls for photographic assistance. Thiswork was carried out in partial fulfilment of the Ph.D. degree under a Research Studentshipfrom the Science Research Council.

REFERENCES

BLAU, H. M. & KAFATOS, F. C. (1978). Secretory kinetics of the follicular cells of silkmothsduring eggshell formation. J. Cell Biol. 78, 131-151.

BLAU, H. M. & KAFATOS, F. C. (1979). Morphogenesis of the silkmoth chorion. Patterns ofdistribution and insolubilization of the structural proteins. Devi Biol. 73, 211-225.

CARO, L. G. (1964). High-resolution autoradiography. In Methods in Cell Physiology, vol. 1,(ed. D. M. Prescott), pp. 327-363. New York, London: Academic Press.

CASE, R. M. (1978). Synthesis, intracellular transport and discharge of exportable proteins inthe pancreatic acinar cell and other cells. Biol. Rev. 53, 211-354.

COUBLE, P., PRUDHOMME, J.-C. & DAILLIE, J. (1977). The biosynthesis of fibroin. III. Involve-ment of the Golgi apparatus in transport and secretion. An electron microscopic autoradio-graphic study. Expl Cell Res. 109, 139-150.

DROZ, B. (1976). Kinetic analysis of macromolecular movements in cells by means of radio-autography. J. Microsc. Biol. Cell. 27, 191-200.

GRAYSON, S. & BERRY, S. J. (1974). Synthesis and intracellular transport of protein by thecolleterial gland of the Cecropia silkmoth. Devi Biol. 38, 150-156.

GUPTA, B. L., MORETON, R. B. & COOPER, N. C. (1973). Reconsideration of resolution in EMautoradiography using a biological line source. J. Microsc. 99, 1-25.

JAMES, T. H. (1948). The site of reaction in direct photographic development. Kinetics ofdevelopment initiated by gold nuclei. J. Colloid Set. 3, 447-455.

JAMIESON, J. D. (1972). Transport and discharge of exportable proteins in pancreatic exocrinecells: in vitro studies. Current topics in membranes and transport, vol. 3, (ed. Bonner, F. &Kleinzeller, A.), pp. 273-338. New York: Academic Press.

KAFATOS, F. C. & KIORTSIS, V. (1971). The packaging of a secretory protein: kinetics ofcocoonase zymogen transport into a storage vacuole. J. Cell Biol. 48, 426-431.

KAFATOS, F. C, REGIER, J. C, MAZUR, G. D., NADEL, M. R., BLAU, H. M., PETRI, W. H.,WYMAN, A. R., GELINAS, R. E., MOORE, P. B., PAUL, M., EFSTRATIADIS, A., VOURNAKIS,J. N., GOLDSMITH, M. R., HUNSLEY, J. R., BAKER, B., NARDI, J. & KOEHLER, M. (1977).The eggshell of insects: differentiation specific proteins and the control of their synthesisin development. In Biochemical Differentiation of Insect Glands (ed.W. Beermann), pp. 45-145.Results and problems in cell differentiation, vol. 8. Berlin, Heidelberg: Springer-Verlag.

KIMBER, S. J. (1979). Follicle cell differentiation and the secretion of the eggshell in Schistocercagregaria (Forsk) the desert locust: a cytological study. Ph.D. thesis, University of Cambridge.

KIMBER, S. J. (1980). The secretion of the eggshell of Schistocerca gregaria. The ultrastructureof the follicle cells during the termination of vitellogenesis and eggshell secretion. J. Cell Set.46, 455-477-

KOPRIWA, B. M. (1966). A semiautomatic instrument for the radioautographic coating tech-nique. J. Histochem. Cytochem. 14, 923-928.

LAI-FOOK, J. & GUPTA, B. L. (1976). Synthesis intracellular transport and release of resilinin Schistocerca. J. Cell Biol. 70, 215 a.

LEBLOND, C. P. & BENNETT, G. (1977). Role of the Golgi apparatus in terminal glycosylation.In International Cell Biology, 1976—1977. First International Congress of Cell Biology,Boston, Mass., 1976 (ed. Brinkley, B. R. & Porter, K. R.), pp. 326-336. New York: TheRockefeller University Press.

MARGARITIS, L. H., KAFATOS, F. C. & PETRI, W. H. (1980). The eggshell of Drosophila melano-gaster. I. Fine structure of the layers and regions of the wild-type eggshell. J. Cell Sci. 43,I-3S-

MAZUR, G. D., REGIER, J. C. & KAFATOS, F. C. (1980). The silkmoth chorion: morphogenesisof surface structures and its relation to synthesis of specific proteins. Devi Biol. 76, 305-321.

Kinetics of secretion of locust eggshell 243

MEYRICK, B. & REID, L. (1975). In vitro incorporation of sH-threonine and 'H-glucose by themucous and serous cells of the human bronchial submucosal gland. A quantitative electronmicroscopical study. J. Cell Biol. 67, 320-344.

MOSES, M. J. (1964). Application of autoradiography to electron microscopy. J. Histochem.Cytochem. 12, 115-130.

NADEL, M. R. & KAFATOS, F. C. (1980). Specific protein synthesis in cellular differentiation.IV. The chorion proteins of Botnbyx mori and their program of synthesis. Devi Biol. 75,26-40.

PAUL, M., GOLDSMITH, M. R., HUNSLEY, J. R. & KAFATOS, F. C. (1972). Specific proteinsynthesis in cellular differentiation. Production of egg-shell proteins by silkmoth folliclecells. J. Cell Biol. 55, 653-680.

PAUL, M. & KAFATOS, F. C. (1975). Specific protein synthesis in cellular differentiation. II.The program of protein synthetic changes during chorion formation by silkmoth folliclesand its implementation in organ culture. Devi Biol. 42, 141-159.

PETERS, T. Jr & ASHLEY, C. A. (1967). An artefact in radioautography due to binding of freeamino acid to tissue by fixatives. J. Cell Biol. 33, 53-60.

REGIER, J. C, MAZUR, G. D. & KAFATOS, F. C. (1980). The silkmoth chorion: morphologicaland biochemical charcterization of four surface regions. Devi Biol. 76, 286-304.

ROGERS, A. W. (1973). Techniques of Autoradiograpky, 2nd edn. Amsterdam: Elsevier.SALPETER, M. M. (1973). Electron microscope autoradiography: a personal assessment. In

Electron Microscopy and Cytochemistry (ed. E. Wisse, W.Th. Daems, I. Molenaar &P. N. van Duijn), pp. 315-326. Amsterdam: North Holland.

SALPETER, M. M. & BACHMANN, L. (1964). Autoradiography with the electron microscope.J. Cell Biol. 22, 469-475.

SALPETER, M. M. & BACHMANN, L. (1972). Autoradiography. In Princples and Techniques ofElectron Microscopy. Biological Applications, vol. 2 (ed. M. A. Hay at), pp. 221-278. NewYork: Van Nostrand Reinhold.

SALPETER, M. M. & SZABO, M. (1972). Sensitivity in electron microscope autoradiography.I. The effect of radiation dose. J. Histochem. Cytochem. 20, 425-434.

WEIS-FOGH, T. (1956). Tetanic force and shortening in locust flight muscle. J. exp. Biol. 33,668-684.

WISSE, E. & TATES, A. D. (1968). A gold latensification-elon ascorbic acid developer for IlfordL4 emulsion. In Electron Microscopy, vol. 2 (ed. D. S. Bocciarelli), pp. 465-466. FourthEuropean regional conference on electron microscopy, Rome. Rome: Tipografia PoliglottaVaticana.

WILLIAMS, M. A. (1973). Electron microscopic autoradiography. Its application to proteinbiosynthesis. In Techniques in Protein Biosynthesis (ed. P. N. Campbell & J. R. Sargent),pp. 125-190. New York: Academic Press.

(Received 19 September 1980)