- Email: [email protected]
Epithelial morphogenesis in developing Artemia: The role of cell replication, cell shape change, and the cytoskeleton
DEVELOPMENTAL
BIOLOGY
152, %‘9-2%? (1998)
Epithelial Morphogenesis in Developing Artemia: The Role of Cell Replication, Cell Shape Change, and the Cytoskeleton
The roles of cell replication and shape change as morphogenetic forces in epithclial inxagination were examined in instar II Arte~in. The epidcrmal cells underwent a fixed pattern of cell division during the first 5 hr of instar II. Greater cell replication in the thoracopod bud (ThB) than in the arthrodial membrane (AM) region resulted in a higher density of epidermal cells in the ThB region (differential cell density). The ratio of cell density (AM/ThB) declined from 1.0 to less than 0.80 by Hour 2 of instar II. Invagination of the AM occurred during Hour 1 when the AM/ThB reached 0.75. A 2-hr pulse with 5’fluorodeoxguridine (FudR) during instar I delayed completion of the cell replication pattern and devclopmerit of transverse cell files in the ThB region for a period equal to the length of the exposure. The delay in thv cell division program resulted in a ccl1 density ratio of 0.93 at Hour 4, a value normally observed in Hour 2 larvae, and cvagination of the epidermis did not occur at apolysis (Hour 4). The FudR treatment did not perturb the cytoskcteton or the initial steps in cell shape change and the larvae formed small segments during instar III. Cell shape change Lvithin the AM began during Hour 4 as this region became significantly thinner than the neighboring ThB region (thickness ratio, AM/ThB - 0.77). Before apolysis the AM cells became wedge shaped, a change which occurred when the basal region of the cell enlarged. The microtubules and microfilaments were reorganizcad from the apical cytoplasm to thr lateral border of apposing AM cells. Following apolysis (Late Hour 4) shape change was completed as the cells attained a thin spindle form, with microtubuleand microfilament-rich filopodial extensions which overlapped adjacent AM cells. As contact with ThB cells shifted from lateral to apicolateral, the AM cells formed the innermost edge of the inragination. Microtubules in the differentiating AM cells contained tyrosinated, detyrosinated, and acetylated o-tubulin isoforms. Treatment with nocodazole, colchicine, tasol, or cytochalasin B blocked AM ~11 shape change and inhibited segmentation, but did not affect the mitotic pattern or differential cell density. Re conclude that the specific pattern of cell division led to differential cell density which, along with AM cell shape change, established the conditions ntressar> ‘( 1992 Academic Press, Inc. to achieve epidermal evagination.
Smuts et r/l., 1978; Meier, 1978; Goldin and Wessels, 1979; Bard and Ross, 1982). Evidence supporting the contenMorphogenesis in many developing tissues begins by tion that one mechanism generates the primary force in evagination or invagination of an epithelial sheet of a certain tissue and that the other mechanisms are inactive has not been forthcoming. Indeed, it is likely that cells to form a tube or pouch (Ettensohn, 1985a; Hilfer and Searls, 1986; Kolega, 1986; Fristrom, 1988, for re- tissue development results from the simultaneous or seview). The cellular basis for these tissue changes may quential occurrence of several morphogenetic processes, include microtubule- and microfilament-mediated cell each contributing to the shape of the epithelium. shape change (Wessels et ul., 1971; Burnside, 1973; Although growth is usually seen as a simple increase Spooner, 1973; Ode11et al, 1981; Owaribe et ccl.,1981), cell in the mass of the developing tissue, cell replication and rearrangement or repacking (Fristrom, 1976; Mit- growth may become a morphogenetic force during early tenthal and Mazo, 1983; Ettensohn, 198513;Hardin and organogenesis if it contributes to lateral (mitotic) presCheng, 1986), disparity of cell adhesion (Nardi, 1981; sure (Ettensohn, IY%a, for review). For example, by inMittenthal and Mazo, 1983), and mitotic pressure (Pic- creasing cell density within a limited area of the epithetet et nl., 1972; Hilfer, 1973; Zwaan and Hendrix, 1973; lium, lateral force would be generated. The force may be restrained by contact with a rigid substratum, and thus be a potential force, or it may be immediately exerted on adjacent regions. Uneven distribution of force through’ Present address: Department of Internal Medicine, University of Alabama Hospital, 619 19th St. South, Birmingham, AL 3.i233. out the tissue would exert a greater effect in weak reINTRODIJCTION
279
280
DEVELOPMENTALBIOLOGY
VOLUME
152,1992
FIG. I. Brine shrimp larvae at instars I (A), II (B), and IV (C). The bracket in each micrograph indicates the region of morphogenesis of the first thorax segment. Arrowheads in B and C indicate the AM region. Anterior is to the top in each figure. The bar in A represents 100 pm, and the magnification is the same for all micrographs
gions, creating a situation in which buckling of the epithelium occurs. Mitotic pressure may work in concert with one or more other mechanisms, as demonstrated for neural development (Morriss-Kay and Tuckett, 1987; Schoenwolf and Alvarez, 1989) and salivary gland (Spooner, 1973). To examine further the interplay of these potential morphogenetic mechanisms, we investigated the role of cell replication, shape change, and rearrangement during invagination of the epithelium of the developing brine shrimp, Artemia (Fig. 1). The epidermal cells constitute a monolayer epithelium of polarized cells attached to the cuticle and organized in a pattern of longitudinal or transverse files (Fransenmeier, 1940; Weisz, 1947; Freeman, 1989a; MacRae et al., 1991). The limb primordium of the thoracopod (ThB) evaginates as the adjacent arthrodial membrane (AM) epidermis invaginates upon release of the epidermis from the cuticle (apolysis) (Freeman, 1989b). In this study, the contribution of each morphogenetic mechanism to epithelial invagination and segment development was investigated. METHODS
Brine shrimp cysts (Sanders Brine Shrimp, Salt Lake City, UT) were hatched and reared in artificial seawater (Instant Ocean, Mentor, OH) at room temperature (23°C). Experiments were conducted with larvae se-
lected for synchronous development at hatching thus at the same age and developmental stage.
Larvae prepared for cell density
analysis
and
were fixed
in toto in Carnoy’s fluid, rinsed, and stained with 1.0 pg/ml Hoechst 33258 in phosphate-buffered saline (PBS), pH 7.4. The larvae were mounted in 80% glycerol for viewing. For cell cycle analysis and determination of cell density, the video image (Dage-MTI67M video camera) was captured with a video frame grabber board (PCVISIONplus, Imaging Technology Inc., Woburn, MA) controlled by ImageMeasure 4100 software (MicroScience, Inc., Federal Way, WA). The cell cycle phase was determined using fluorescence intensity of the nucleus. The image was averaged 10 times and the background subtracted before measuring the fluorescence. The fluorescence intensity was compared to the area of the nucleus to confirm the cell cycle phase, as previously defined (Freeman and Chronister, 1989). The cell density was ascertained by digitally overlaying a 50 x 18-pm rectangle over the image of the epiderma1 cells and counting the nuclei within the box. The ratio of the number of cells in the presumptive AM versus the number of cells in the ThB was used to define differential cell density, reported here as the AM/ThB ratio. Analysis of variance was performed after arcsine transformation of the ratio.
FIG. 2. Spatial pattern of cells (Hoechst-stained nuclei) in the ventral epidermis of segments 1 and 2 during the first 5 hr of instar II. (A) Ikring the first hour of instar II (Hour 0) the cell pattern changes from seven longitudinal files to eight transverse files numbed from anterior (1) to posterior (8). The middle ThB cells of the first segment (file 1) are aligned transversely (large arrowhead) and mediolateral divisions have occurred in the first three ~11s. The cell indicated hy “2” is the fourth cell in the mediolatcral file and is ahout to divide. The cell to the right is in metaphasr and the division axis is mediolateral. Cells in the AM region (transverse files 3 and 5) are indicated hg small arrowheads. (8) Iiour 1. (C) Hour 2. (D) Hour 3. The increase in cell replication in the ThB cells of the first segment (transverse files l-3) has led to a differential cell density. (E) Hour 4. Cell shape change is beginning in the AM region (files 4 and 5). (F) Hour 5. Apolysis has occurred. Anterior is to the top in all figures and the ventral midline is indicated hy “m”. The rectangle in A has the dimensions of 18 X 50 pm and serves a:: a six> rc~fcrencc for the cell density determinations. Bar in A 2 20 pm, and the magnification is the same for all micrographs.
To examine the importance of differential cell density in segmentation and evagination, the cell cycle of the .epidermal cells in the ThB region was blocked by exposure to 5’-fluorodeoxyuridine (FudR) (10 pg/ml seawater) during Hours 13 and 14 of instar I. This time
period includes S phase for the cells in the anterior ventral region of the thorax (Freeman and Chronister, 1989). Larvae were then transferred to drug-free seawater after which the cell division pattern and the differential cell density at Hours O-5 of instar II were de-
282
D&EISWMENTAI.
1.00
P
BIOLOGY
-
0.80 : 0.90 0.70
-
0.60
-
~‘q~~~:;,:i 1
0.50 '
I 0
1
3
2
Hour
of
lnstar
4
5
II
FIG. 3. Change in the ratio (AM/ThB, ordinate) of cell density (open triangles) and cell thickness (black triangles) during early instar II. Each point represents the mean and one standard deviation of 20-24 larvae. The change in cell density occurs during the first 3 hr while the differential cell thickness does not change significantly until Hour 4.
VOLUME 152,1992
PBS) (Molecular Probes, Eugene, OR), and mounted in 80% glycerol containing 2% n-propyl gallate. Antitubulin antibodies used and their specificities were: DMlA (cu-tubulin, Sigma), KMX (/$tubulin, a gift from Dr. K. Gull), anti-Glu (detyrosinated a-tubulin, a gift from Dr. C. Bulinski), 6-llB-1 (acetylated cu-tubulin, a gift from Dr. G. Piperno), and YL1/2 (tyrosinated cu-tubulin, Dimension Laboratories Inc. Mississauga, Ontario). These and other specimens were observed and photographed with a Leitz Dialux epifluorescence microscope or a Sony video printer. To study the role of microtubules and microfilaments in AM cell shape change, larvae were exposed to colchitine or nocodazole (10 pg/ml) to disrupt microtubules, taxol(2 PM) to stabilize microtubules, or cytochalasin B (20 pg/ml) to disassemble microfilaments. The drugs were placed in artificial seawater at specified times, after which the larvae were observed for progress in segmentation and fixed for immunocytochemistry.
termined. These larvae were also fixed and stained for microtubules and microfilaments, as described below. Analysis
of AM Cell Shupe Ch,ange
The thickness of the epidermis in the AM and ThB regions was determined in living and fixed specimens by image analysis. After the image was captured as described above, measurements were made using the Line Distance subroutine from the ImageAction program (Imaging Technology). The measurements included (1) the thickness of the AM and ThB regions (distance from apical to basal surfaces), (2) the width of a group of three ThB cells in one file (cumulative cell diameter) for cell packing measurement, and (3) the width (diameter) of apical and basal regions of the AM cell at different times during instar II. For the cumulative cell diameter determinations, the cell boundary was considered to be a point equidistant between nuclei. All dimensions were measured digitally after the X and Y coordinates of the field were calibrated with a slide micrometer. Microfubu,les
and Microjilamexts
Microfilaments and microtubules were detected as previously described (MacRae et al., 1991). Briefly, larvae were fixed overnight in 4% paraformaldehyde in PBS. After treatment with PBS containing 0.5% albumin and 0.5% Triton X-100, the specimens were stained overnight with antitubulin antibodies, rinsed in several changes of PBS, stained with FITC-labeled goat antimouse IgG (1:40) (Sigma Chemical Co., St. Louis, MO), rinsed in PBS containing TRITC-phalloidin (20 PM in
RESIJLTS
Pattern of’ Replicution AM and ThB Cells
in
arid Cell Cycle qf’ Presumptive bxtar II
Epidermal cells involved in formation of segments 1 and 2 were identified by cell lineage studies (Freeman, 198913,unpublished). At ecdysis to instar II (Hour 0) the presumptive AM cells were arranged in seven longitudinal files in the ventral epidermis; each file extending from the cephalon to the tail. The orientation of the file
TABLE
1
MORPHOMETRICDETERMINATIONOFCELLDIMENSIONSOF AM CELLS AND THB CELLS ADJACENT TO THE AM REGION AT HOUR 2, HOUR 4 (BEFOREAPOLYSIS),ANDHOUR~(AFTERAPOLYSIS) Morphometric dimension AM cell height ThB cell height Diameter of apical AM cell surface Diameter of basal AM cell region Cumulative diameter of three ThB
Hour 4 after apolysis
Hour 2
Hour 4 before apolysis
9.0 (1.06) 9.6 (.61)
9.4 (Xl) 11.1 (.81)
7.5 (94) 9.3 (2.1)
6.6 (.54)
6.9 (1.3)
5.05 (.67)
5.6 (X4)
7.8 (.97)
6.5 (1.0)
23.1 (2.4) 22.9 (1.3)
18.6 (2.8) 17.7 (1.0)
20.45 (2.0) 21.95 (1.4)
WllS
CT) (S)
Not<>. Each value (in pm) is the mean (standard deviation) of lo-15 measurements of that dimension. The cumulative diameter of ThB cells was measured from transverse (T) and surface (S) views.
FREEMAN,CHESHIRE,ANDMACRAE
E~ithdicrl
12k,r.pl/o!/rt/c’sis in .A rtfjirc it/
283
FIG. 4. Recovery from treatment with Y-fluorodeoxyuridine. (A and B) Hoechst-stained nuclei in control and treated larvae, respectively, at Hour 2 instar II. The drug slowed the temporal pattern of cell replication but did not afl’ect cell position in the epithclium or the orientat ion 01 mitosis. ((1) Surface view of Hour 2, instar II FudR-treated larva showing no disruption of microtubules by the drug. (D) Same larva as in (’ showing no disruption of microfilaments. iE) Surface view of FudR-treated Late Hour 4, instar II larva showing microfilaments in AM cells. The >lM cells hare begun to change shape. (F and G) Appearance of segment 1 in control (F) and treated (G) larvae in instar III showing a smaller hut otherwise normal limb bud in recovering larvae. Cells in the mid-ThB region (file 1 in A, (’ and E, file 2 in B and D, are indicated by the large arrowhead. Cells in the AM region (files 4 and j) are indicated by two small arroa-heads. Antrrior is to the top in all ligures. m, the ventral midline. Bar in .4 wpresents 20 pm for A-E. Bar in F represents 20 pm for F and G.
is indicated by the pattern of nuclei rather than the cell borders. During the first hour of the instar the cells that were to form the posterior region of the ThB of segment I, the AM, and the anterior region of segment 2 underwent a change such that they appeared to be organized in eight transverse files (Fig. 2A). This transformation is a result of cytoplasmic changes rather than alteration of cell neighbors. Cells in files numbered 1 and 8 will form the apex of the first two segments, respectively. Transverse files 2 and 3 formed the posterior face of the first bud 1 and files 6 and 7 formed the anterior face of the second bud. Cells in transverse files 4 and 5 formed
the AM, the region between the ThB buds of the first two segments. The cell cycle behavior in the AM region differed from that in the ThB region and the tissue types were defined by the cell cycle pattern. The cells were in G2 phase at ecdysis to instar II. Those cells in the presumptive ThB region of the first segment underwent one rapid replication in the transverse direction, beginning at Hour 0 with the medial cells of file 1 (Fig. 2A). The same pattern was observed at progressively later times in files 2 and 3 (Figs. 2B and 2C). By Hour 2, all cells in files 1 and 2, but none in file 3, had divided (Fig. 2C). Divisions of
284
DEVELOPMENTAL
BIOLOGY
0.90
152, 1992
invagination of the AM region removed the nuclei from the plane of focus. During the period of replication, the ThB cells increased in height (Hour 2 to Hour 4, Table 1). In addition, the cumulative width (diameter) of three ThB cells in one file decreased from approximately 23 to 18 pm. These morphometric determinations suggest that the diameter of the cells decreased during division and the height increased as the postmitotic cells grew.
1 .oo
F
VOLUME
p!r
0.80 0.70 0.60 t 0.50 ’
I 0
1
2
Hour
4
3
of
lnstar
5
II
FIG. 5. ERect of a 2-hr pulse with FudR on cell density (triangles) and cell thickness (circles) in control (open symbols) and treated (solid symbols) larvae. Each point represents the mean and one standard deviation of lo-20 larvae.
the presumptive ThB cells of the first segment always preceded those in segment 2. Similar to the first segment, cell divisions in the second segment began with the file furthest from the AM region (file 8) and then occurred in files 6 and 7. By Hour 3 almost all cells in files 1-3 and 6-8 had divided (Fig. 2D). In contrast to the ThB cells, the AM cells (files 4 and 5) remained in G2 throughout the first 4 hr of the instar (Figs. 2A-2D) and at the time of cell shape change were in the G2 phase (Figs. 2E and 2F). After invagination, the AM cells were no longer in the plane of focus and were difficult to observe. It was possible that the AM cells divided later in the instar, but no mitotic figures were noticed at Hour 5.
Mitotic pressure may act as a morphogenetic force if the pattern of replication results in a significant difference in the density (number of cells per 50 X l&pm area) of cells in adjacent regions of the epithelium, hereafter referred to as differential cell density. Cell replication in the presumptive ThB regions of segments 1 and 2 increased the number of cells in the measurement region from approximately 12 cells at Hour 0 to 1’7 cells at Hour 4. These cells were still attached to the cuticle at their apical surfaces. During the same time, the number of AM cells did not change significantly. The ratio of cell density (AM/ThB) declined from 0.99 at Hour 0 of instar II to 0.75 at Hour 3 (Fig. 3), and the ratio did not change significantly (P > 0.05) for the next 2 hr, during which apolysis and evagination occurred. It was difficult to accurately measure the cell density after Hour 5 since
Diflerential
Cell Density Produces Mitotic
Pressure
To determine the role of changing cell density in segment morphogenesis it was necessary to separate the process of cell replication from other potential effecters, such as cell shape change, by altering the temporal pattern of developmental events during early instar II. To accomplish this, replication of the ThB cells was inhibited by treating larvae with a pulse of FudR at Hours 13-15 of instar I, when almost all of the epidermal cells in the presumptive ThB region of the first segment were traversing S phase (Freeman and Chronister, 1989). We reasoned that if, upon removal of the drug, the normal cell replication pattern resumed, division of the ThB cells compared to that of the AM cells would be delayed by 2 hr. The appearance of differential cell density between the ThB and the AM regions would also be delayed by 2 hr. Examination of larvae in control and treated groups at hourly intervals during early instar II showed that the spatial and temporal patterns of transverse cell replication were normal following drug treatment, but they were delayed by 2 hr (Figs. 4A and 4B). The cell density ratio at Hour 4 in treated larvae was 0.94, a value not significantly different from that of Hour 2 larva (Fig. 5). Thus, the delay in pattern generation and differential cell density was similar to the period of the FudR block and there was no indication that the drug-treated cells accelerated or otherwise altered the timing of replication. At apolysis (Hour 4), neither ThB evagination nor AM invagination had occurred in FudR-treated larvae. The results suggest that disparate replication patterns within the epithelium established differential cell density which in turn was necessary for segmentation. The FudR treatment did not affect the pattern of microtubules or microfilaments in the epidermal cells of the ThB and AM regions (Figs. 4C and 4D) ahd the AM cells demonstrated typical cell shape change at Hour 4 (Fig. 4E). Similarly, the AM/ThB thickness ratio (see below) for both the control and the treated larvae decreased at Hour 4 (Fig. 5). During the next instar the treated larvae continued segment formation, although
FREEMAN.~HESHIRE,AND
MACRAE
&pith
fditrl ~~~~~~~~l~o~~r~~~f~.si.s ii, .-l rfcii, itr
285
I;. Appearance of the AM and ThB regions of the epidermis at Hour 2 CA) and Latr Hour 4 (BJ. In B apolysis has occurred over the AM and nt ThB cells; the old cuticle is indicated hg “c”. Large arrowhead indicates the Th8 rexion; small arrowhead indicates the AM region The lines indicate the location of the lateral borders of one AM cell. They show the ne&,rc-shaped ccl1 in R. Bar represents 20 pm for both
the segments were smaller than those of control (Figs. 4F and 4G).
larvae
The AM cells Ivere the first of the epidermal cells in the thorax to differentiate during segmentation (Cheshire, 1987; Freeman, 1989a), undergoing a complex shape change before assuming a thin spindle form. The shape change of the AM cells began shortly before apolysis during Hour 4 and was completed after apolysis (Figs. 6 and 7). By Hour 4 the ThB cells had undergone enhanced replication and were columnar. The AM cells were still attached to the cuticle and their apical surfaces did not change in area (diameter of apical region, Table 1) while the basal region enlarged significantly (P < 0.05) (diameter of basal region, Table 1). The height of the AM cells did not change appreciably (Table 1). In some larvae the basolateral regions of these cells shifted slightly with respect to adjacent AM cells (Fig. 7, Hours 4 and 5).
Following apolysis over the AM and neighboring ThB cells, invagination began. Contact between the AM cells and their ThB neighbors changed from lateral to more apicolateral as the diameter of the apical region of the AM cell decreased (Figs. 6B and 7; Table 1). By Hour 5, apolysis was completed over all of the thorax epidermis and the AM cells had reached an innermost position in the invaginating AM region. The epithelium in the AM region became thinner than that in the ThB region as shape change progressed. It is expected that the thinner epithelium would deform (invaginate) before a thicker region in response to lateral pressure. To determine if the difference in cpidermal thickness was important in the invagination process, the thickness of the AM and ThB regions was compared at different times in instar II (AM/ThB ratio, Figs. 3 and 5). Over the first 3 hr of instar II the ratio declined slightly (P > 0.05). However, a decrease in the height of the AM epithelium combined with an increase in the height of the ThB cells during Hour 4 brought the ratio
286
DEVELOPMENTALBIOLOGYVOLUME~~~, 1992
Hr2 Hr4
Hr4
Hr4 APO
Hf5 Hr 5
FIG. 7. Diagrammatic view of the cell shape changes in the vcntrolateral epidermis. The transverse view of a 60-wrn region of the integument is shown on the left and the surface view of the AM region and neighboring ThB cells (two from segment 1 and two from segment 2) is shown on the right. The bar at the bottom of each panel represents 20 pm. (Left) At Hour 2 the cells are similar in size, still attached to the cuticle (C), and demonstrate a polarized cgtoskeleton with microtubules (MT) and microfilaments (MF) enriched in the apical region. The AM cells are indicated by arrows. By Hour 4 the ThB cells have undergone enhanced replication and are columnar. The AM cells are still attached to the cuticle and their basal regions have enlarged. The organization of the microtubules and microfilaments has shifted from primarily apical to enriched along the lateral membranes of apposinp AM cells. The basolateral regions of one AM cell may shift transversely with respect to the cell in the other AM file. Later in Hour 4 the izM cells and the neighboring ThB cells undergo apolysis and invagination begins. The neighboring ThB cells now decrease in height and are not as densely packed. As invagination proceeds, cell-cell contact of the AM cell with the ThB neighbors changes to a more apicolateral position. By Hour 5, apolysis is complete and the AM cell has invaginated to the greatest extent. The arrowhead indicates the position of the other AM cell. (Right) Surface view of the invaginating AM and neighboring ThB cells. Depth of inragination is indicated by stippling. Anterior is to the left in left panel and to the top in right panel.
to below 0.8, where it remained through the period of invagination (Figs. 3 and 5). Morphometric measurements before and after apolysis demonstrated that the packing of neighboring ThB cells increased prior to Hour 4 as the ThB cells repli-
cated (cumulative diameter of three ThB cells, Table 1). Measurement of ThB density showed that both density and height decreased after apolysis, suggesting that the ThB cells were spreading following release from the cutitle. This change in packing density may be sufficient
FIG. 8. Fluorescence staining of microtubules and microfilaments in the vcntrolatcral epidermis of instar II Artemicr. Large arrowhead indicates the ThB region of the first segment; small arrowhead indicates the AM region between the first two segments. (A) Surface view of microtubules in epidermis of Hour 3 larva. (B) Surface view of microfilaments in epidermis of Hour 3 larva. (C) Transverse view of microlilaments of Hour 3 larva showing apical staining of cells. (D) Surface view of Hour 4 larva stained for acetylated cu-tubulin. The AM cells are beginning to change shape (compare to A). (E) Surface view of Late Hour 4 larva showing tubulin in AM cells during invagination. (F) Microfilaments in Hour 4 larva during shape change (compare to B). (G and H) Surface view showing enrichment of microtubules (G) or microfilaments (H) in apposing regions of AM cells (files 4 and 5) at Hour 4. (I and J) Transverse view of Hour 4 larva showing mid-AM border with concentration of microfilaments (I) and microtubules (J) (small arrowhead) in the adjacent regions of apposing AM cells in files 4 and 5. The white dashed lines indicate the border of the AM cells with the neighboring ThB cells and demonstrate the wedge shape of the cell. (K) Hour 1 larva with plane of photograph at the level of transverse view of AM on the left side and surface view on the right side. The microfilaments are enriched in the lateral region of the wedge-shaped cell (small arrowhead, left side) and between apposing cells (arrowhead, right side). The white dashed lines elucidate the location of lateral horders of one AM cell and demonstrate the wedge shape of the cell. The specimens were stained with antitubulin antibodies KMX in A, DMlA in E and G, 6-llB-1 in D, YL1/2 in J, and phalloidin in B, C, F, II, I, and K. The bar represents 20 pm in all figures. n, nucleus.
288
DEVELOPMENTAL BIOLOGY
VOLUME 152, 1992
FIG. 9. View of the basal surface of epidermal cells in the invaginating AM region. The filopodia (arrowheads) extend transversely and overlap adjacent AM cells. (A and B) The same larva stained, respectively, for microtubules containing acetylated cu-tubulin and microfilaments. (C and DI Same larva stained, respectively, for tu-tubulin (DMlA) and microfilaments. Bar represents 20 pm for all figures. n, nucleus.
to produce lateral pressure which would cause deformation of the AM cells.
Reorgunixation of Microtubules during Cell Shape Change
and Microjilaments
Immunolocalization of microtuhules and microfilaments revealed that cells in the presegmentation thoracic epithelium contained a polarized cytoplasm with the apical population of microtubules located near the surface and approaching the lateral borders of the cell (Fig. 8A; see also MacRae et al., 1991). Medial and basal regions of the cell contained fewer microtubules. Microfilaments were distributed in a manner similar to that of the microtubules (Figs. 8B and 8C). The intracellular location of microtubules and microfilaments in the AM cells changed during Hour 4, instar II (Figs. 8D-8K). Surface views of larvae revealed that microtubules and microfilaments were colocalized as the cells changed shape and assumed a spindle form (Figs. 8D-8F, compare to A and B). Early in Hour 4 there was an enrichment of microtubules and microfilaments near the apposing membranes of AM cells located in files 4 and 5, the point of eventual invagination (Figs. 8F-8H). A transverse view of the AM region revealed that, as the cell became wedge shaped, there was a redistribution of the microtubules and microfilaments from the apical to the midlateral and basal regions of the cell (Figs. 81-8K). The change in location of microtubules and microfilaments was completed just prior to apolysis (Late Hour 4, instar II). After apolysis, shape change was completed as the AM cells became spindle shaped and extended lateral filopdia-like projections (Fig. 9). The extensions were always directed mediolaterally and overlapped basolatera1 regions (Figs. 9A and 9B) and filopodia (Figs. 9C
and 9D) of neighboring cells. Microfilaments and microtubules were concentrated in the extensions. Invagination was complete shortly after apolysis (Fig. 7, Hour 5). Specimens in which the dorsal thorax had been dissected away revealed that the leading, curved edge of the invagination was formed by one AM cell, either from file 4 or file 5 (Figs. lOA-1OC). The remaining AM file and adjacent ThB files formed the walls of the invagination (Fig. 10D). Studies with several cell types from different organisms have demonstrated that certain cell activities may involve a change in the composition of post-translationally modified tubulins (MacRae and Langdon, 1989, for review). To determine if post-translationally modified tubulins were important in AM cell shape change, larvae were stained with antibodies to tyrosinated, detyrosinated (not shown, MacRae et al., 1991), and acetylated cu-tubulins. Antibodies to acetylated (6-llB-1) and detyrosinated cu-tubulin isoforms stained the invaginating cells in a manner similar to DMlA (oc-tubulin) and KMX (/3-tubulin). All isotubulins occurred in the AM and ThB cells (Figs. 8A, 8D, 8G, 85; 9A). Microtubules and Microjilaments Cell Shape Change
Are Necessary for AM
To determine the role of cytoskeletal elements in cell shape change and thinning of the AM epithelium, the effects of microtubule and microfilament reactive drugs on Hour 4 larvae were examined (Table 2). In larvae exposed to cytochalasin B (20 pg/ml), taxol (5 PM), colchicine (10 pg/ml), or nocodazole (10 wg/ml), the AM/ ThB cell thickness ratio never dropped below 0.9 (Table 2), showing that the reduction in cell height associated with cell shape change involved both of the cytoskeletal elements. The cell pattern and the differential cell density achieved prior to Hour 3 were not affected by the
FREEMAN,CHESHIRE.ANDMACRAE
inhibitors (Table 2). These compounds did, however, block formation of spindle-shaped cells and filopodial extensions. Colocalization of microtubules and microfilaments during cell shape change suggests that the interaction of the two elements is essential to the process and that they are structurally interdependent. We observed, however, that taxol, nocodazole, and colchicine disrupted the organization of microtubules but did not affect microfilaments. Similarly, larvae treated with cytochalasin B showed loss of microfilaments but the microtubule staining pattern was unaffected.
To determine if cell rearrangement was active in limb bud development, the pattern of cells was observed during the 5hr period in which the limb bud was delineated. Based on the position of the nuclei, we could discern no evidence for movement of any cell in the ThB region other than that due to mitosis. Prior to apolysis in the AM, no shift was more than one-half of a cell diameter. We cannot rule out the possibility that some rearrangement may have occurred following apolysis (Hour 5), as suggested by the shift in basolateral regions and spindle cell formation. I)ISCIJSSION
FIG. IO. View of the l~sal surface of AWI invagination at different [)lancls of focus. (A and 8) Innermost e&y of invagination showing a single wll stained for microtubules (A) or microfilaments (B). Thtl tilopotlia arc intlicntcd hy arrowheads. CC:)Slightly more apical view ol’ ncighhoring cells in the e&!c of the invagination. (D) Apical-most vitj\v at the v&y of the AM inragination, stained for nricrotubulcs. Slwcimcans in A, (‘. and L) n-cre stained with 1)MlA while that in B was staincvl with I)halloidin. n. nucleus. The bar wpresents 20 pm fur all liguws
The results of this study support a model of epithelia l invagination in which cell replication plays an impor tant role as a morphogenetic force in segmentation. Th e temporal and spatial patterns of cell division are highl! regulated, and such control is required to attain the differential cell density in the AM and ThB. The AM and ThB cell populations are distinguished by their replication during the second instar with patterns similar to those found in the segmenting regions of other crustacean embryos (Scholtz, 1990). While the developmental step that regulates the pattern of division, and therefore defines the two cell populations, is not known, it clearly is initiated prior to instar II (Freeman, 19891)). Although mitotic pressure is involved in the processes of evagination (Zwaan and Hendrix, 1973; Smuts cjt trl, 19’78), accordion folding (Pictet ef ul., 1972; Bard and Ross, 1982), and budding of epithclia (Goldin and Wcssells, 19’79), it has been difficult to evaluate its importance with respect to other morphogenetic activities. In this study the role of cell replication was examined by experimentally dissecting morphogenesis into replication-dependent (FudR-sensitive) and replication-indcpendent processes. The FudR-induced delay in the increase in cell density did not perturb the shape change in the ,4M cells or the timing of apolysis. Without dif-
290
DEVELOPMENTALBIOLOGY
VOLUME 152,1992
TABLE 2 EFFECTOFMICROTUBULEANDMICROFILAMENTREACTIVEDRUGSONCHANGESINCELLDENSITYANDSHAPE Cell density Treatment Control Hour 0 Hour 2 Hour 3 Hour 4 Hour 5 Cytochalasin Taxol Colchicine Nocadazole
AM
B
11.9 13.2 12.3 13.7 14 13.2 12.8 13 11.4
(.7) (.6) (.63) (.5) (.4) (.6) (.5) (.5, (.7)
ThB
12.1 16.5 15.2 17.7 19.5 18.7 17.4 17.3 15.6
(.7) (.9) (.74) (.6) (.9) (.Y) (.7) (.6) (.6)
Cell thickness AM/ThB
.98 .8 .81 .77 .71 .71 .75 .75 .73
AM
10.4 9.8 11.1 9.2 8.2 11.7 11.4 11.5 11.6
(3) (1.0) (.9) (3) (.4) (.7) (.7) (.6) C.7)
ThB
10.6 10.4 11.8 11.97 11.8 12.1 12 12.3 12.3
(.9) (1.2) (.9) (.8) (.7) (.9) (3) (.5) (.6)
(pm) AM/ThB
.98 .94 .94 .77 .7 .96 .95 .93 .94
LVofr. Exposure to the drugs began at Hour 3 and the larvae were fixed at Hour 5. Each value in the cell density columns AM and ThB represents the mean (-tl standard deviation, w = 10-20) number of nuclei per 18 X 50-pm region of the epidermis. Each value in the AM and ThB columns under cell thickness represents the mean (+l standard deviation, 7~= 10-20) epidermal thickness (gm) in that region.
ferential cell density, the AM region did not invaginate at apolysis. Although we can not eliminate the possibility that the delay in cell replication simply resulted in an insufficient number of ThB cells for complete invagination, we favor the idea that differential density could provide the morphogenetic force needed for mitotic pressure after apolysis. Morphometric measurements show that, prior to apolysis, ThB cells increase in height as the cell number and density increase, suggesting that both replication and growth contribute to the increase in cell density and mitotic pressure. During this period the mitotic pressure may be stored as “potential” lateral force since expansion of the epithelium is inhibited by attachment to the l-pm-thick cuticle at the apical surface (Freeman, 1989a,b; Horst, 1989; Criel, 1991). Contact with the cuticle would also have inhibited cell rearrangement, as has been found in developing imaginal disc cells (Pino-Heiss and Schubiger, 1989; Birr et a,l., 1990; Condic et al., 1991). Immediately after release from the cuticle the free ThB cells decrease in height as the epithelium spreads slightly (Fig. 7), establishing a kinetic force. Since the mid-segment ThB cells are still attached to the cuticle (Freeman, 1989b), the lateral pressure generated by the expanding ThB region cannot be directed in any direction except toward the AM. The cellular mechanism by which the mitotic pressure could be stored and subsequently transformed into lateral pressure is not understood. Expansion of the ThB epithelium is associated with a decrease in cell height, suggesting a reorganization of ThB cell cytoplasm. The results do not permit us to discern whether this reorganization is passive, occurring in response to release from the cuticle, or is an active shape change
that provides the force for lateral expansion and AM invagination. We were unable to detect any alteration in the distribution of microtubules and microfilaments in the ThB cells during expansion, as would be expected if the cytoskeleton was active in cell shape change. For mitotic pressure to effect epithelial deformation there must be a region of cells that deforms before the rest of the epithelium. In brine shrimp this is the AM region where the epithelium becomes thinner as the cells undergo shape change. Shape change begins as the cuboidal cells become wedge shaped while still attached to the cuticle. This change occurred in a manner that appears to be different from apical constriction (pursestring effect) exhibited by neural epithelium (Burnside, 1973; Schoenwolf and Franks, 1984), thyroid (Hilfer, 1973), salivary gland (Spooner, 1973), and lens (Zwaan and Hendrix, 1973) tissues. The AM cell was attached to the cuticle and there was no reduction in the apical surface area prior to apolysis. Rather, the wedge shape was attained through an enlargement of the basal region, a process associated with a shift of the microtubules and microfilaments from the apical region to the lateral membrane of apposing AM cells and translocation of the nucleus. The second phase of the shape change occurs after apolysis. The cells become spindle shaped with lateral extensions aligned mediolaterally, overlapping adjacent cells. Microtubules and microfilaments were concentrated in the extensions and agents that blocked the assembly of either cytoskeletal element inhibited filopodia formation. The role of the lateral extensions may be to (1) increase the interaction of the two AM cell files, (2) provide directionality and strength to the invaginating cells, (3) facilitate the formation of the inner edge of
FREEMAN,~HESHIRE,AND
MACRAE
the invagination, or (4) rapidly extend the lateral borders of the cell to achieve the spindle shape. Similar extensions occur in insects (Locke and Huie, 1981; Nardi and Magee-Adams, 1986; Delhanty and Locke, 1989) and sea urchins (Hardin, 1989) where they are thought to carry out several developmental processes. Microtubules and microfilaments are colocalized throughout both phases of AM cell shape change, suggesting that there is a functional interaction between them. The microtubule and microfilament inhibitor studies demonstrate that, although both microtubules and microfilaments are required for the shape changes occurring before and after apolysis, they appear to be structurally independent of one another. Thus, microtubules and microfilaments may have complementary functions in AM morphogenesis. The microtubules contain tyrosinated, detyrosinated, and acetylated isoforms of cu-tubulin which were not restricted to certain regions of the cytoplasm, specific cell type, or discrete periods of development. The findings support the contention that the isotubulin diversity does not change during cell differentiation and morphogenesis in instar II (Langdon et al., 1991; MacRae et ul., 1991). Our findings demonstrate a strong role for cell replication in establishing the conditions necessary for epithelial invagination. Mitotic pressure functions in conjunction with cell shape change and, possibly, a change in cell adhesion. Segment morphogenesis in Artenliu then appears to be a result of the temporal and spatial orchestration of the different morphogenetic processes. This research was supported by National Institutes of Health Grant HD24219 to J.A.F., NATO Grant 0865187 to J.A.F. and T.H.M., and an NSERC Grant to T.H.M.
REFERENCES Bard, J. B. I,., and Ross, A. S. A. (1982). The morphogenesis of the ciliary body of the avian eye. II. Differential cnlargment causes an epithelium to form radial folds. De?,. Lzio/. 92, 87-96. Birr, C. A., Fristrom, D.. King, D. S., and Fristrom, J. W. (1990). Ecdysane-dependent proteolysis of an apical surface glycoprotein may play a role in imaginal disc morphogenesis in Drtwopl~ilo. I~ew/o~~~r/c?/t 110, 239-248. Burnside, B. (19731. Microtubules and microfilaments in amphibian neurulation. An,. Zoo/. 13, 989-1006. Cheshire, L. B. (1987). “Role of microfilaments, microtubules, and intermediate filaments in epithelial evagination in the brine shrimp, Artc~v~itr scr~inu.” MS Thesis, IJniversity of South Alabama. 76 pp. Condic, M. L., Fristrom, D., and Fristrom, J. W. (1991). Apical cell shape changes during f~,osophilrc imaginal leg disc elongation: A now1 morphogcnetic mechanism. Dwc/opnzc?~t 111, 23-33. Criel, G. R. J. (1991). Morphology of Arte~mia. 1~ “Artmnio Biology” (R. A. Broane, P. Sorgeloos, and C. N. A. Trotman, Eds.), pp. 119153. CRC Press, Boca Raton, FL. Delhanty, P.. and Locke, M. (1989). The development of epidermal feet
Epitlrdirf
I JI01~~~1~0~~~~~c.si.s iu A rtr~fir icl
291
in preparation for metamorphosis in an insect. Tiswr~ Wl 21, 891909. Ettensohn, C. A. (1985a). Mechanisms of epithelial imagination. Q,ctrrf.Rf?. Hid. 60, 289-307. Ettrnsohn, C. A. (1985b). Gastrulation in the sea urchin is accompanied by the rearrangement of invaginating epithelial cells. Uvi,. Bid. 112, x-390. Fransenmeier, L. (1940). Zur fraye der herkunft des metanauplialen mesoderms und die segmentbildung hei drtcrvitr so/it/o Leach. &if. ,f/tr MSs.sc~rscl/c~,flicl/Zool. 152, 43S472. Freeman, J. A. (1989a). The integument ofArtc>/tlirr during early dcvelopment. Ire “Biochemistry and Cvll Biology of Arfvwitr” (T. H. MacRae, J. C. Bagshaw, and A. H. Rarner, Eds.). pp. 233256. CRC’ Press, Boca Raton, FI,. Freeman, J. A. (IRXDbl. Segment morphoyenesis in drtc~wicc lar\-ac. 111 “Cell and Molecular Biology of Artemia Development” (A. H. Warner, T. H. MacRae. and J. C. Bagshaw, Eds.1. pp. 77-W. Plenum, New York. Freeman, J. A., and Chronister, R. B. (1989). Cell cycle kinetics of epidermal cells in dcvcloping A,?prr/itr: Ir/ sit/l analysis. .I. (:r/t.st. Bid. 9, 193-201, Fristrom, D. (1976). The mechanism of evagination of the imaginal discs of ~rwol,hiltr rt/cltr~co!/trstc~,: III. Evidence for cell rcarrang,rcment. UVI.. Biol. 54, 163171. Fristrom. D. (1988). The cellular hasis of epithelial morphogenesis: A review. Tis.s~~ (‘J/I 20, 6-l.5690. Goldin, G. I’., and \\~eswlls. N. K. (1979). Mammalian lung develol)ment: The possible rolr of cell proliferation in the formation of supernumerary tracheal buds and in branching morphogenesis. .J. Erp. Zool. 208, 337~316. Hardin, J. (1989). Local shifts in position and polarized motility drive cell rearrangement during sva urchin gastrulation. 11~~. Rio/. 156, J3@-l45. Hardin, J. D.. and Cheng (1986). The mechanisms and mechanics of archenteron elongation during sea urchin gastrulation. I)(,,,. 8iol. 115, 19OL!lOl. Hilfer, S. R. (19731. Extracellular and intracellular correlates of organ initiation in the cmhryonic chick thyroid. ..lnl. %oo/. 13, 1023~1038. Hilfer, S. R., and Searls, R. L. (19861. Cytoskclctal dynamics in animal morphogcnesis. 11, “Developmental Biology: A Comprehensive Synthesis” (L. IV. Browder. Ed.), Vol. 2, pp. 4-29 Plenum, Ntv York. Horst, M. N. (1989). Molecular and cellular aspects of chitin synthesis in larval .4r?c,witr. 1~ “Cell and Molecular Biology of Arfc~n/itr Drvrlopmcnt” (Warner. A. H., T. H. MacRac, and Bagshaw, J. (:. Eds.), pp. 59-76. Plenum, New York. Kolega, J. (19861. The cellular basis of epithelial morphogenesis. I,( “Developmental Biology: B Comprehensive Synthesis” (L. 15’. Browder, Ed. I, Vol. 2, pp. 103-143. Plenum, New York. Langdon, C. M., Freeman, J. A., and MacRae, T. H. (1991). Post-translationally modified tubulins in Avtc~wio: Prelarval development in the absence of drtyrosinated tuhulin. Ilcc Bid. 148, 147m155. Locke, M., and Huie. P. (1981). Epidermal feet in insect segment morphogenesis. .\‘tr/~w 293, 733735. MacRac, T. H., and Langdon, C. M. (19891. Tubulin synthesis, structure, and function: What arc’ the relationships? Ilioc//c~r)r. Co// nio/. 67, 770~790. MacRae, T. H., Landgon, C. M., and Freeman, J. -4. (1991 I. Spatial distribution of post-translationally modified tuhulins in polarized cells of developing A rfc>witr. Cc& Mdil. Cytoskc/. 18, 189-203. Meier, S. (19781. Dewlopment of the embryonic chick otic placode: Light microscopic analysis. Attut. Rrc 191, 44Tm458. Mittenthal. J. E., and Maze, R. M. (1983). A model for shape generation
292
DEVELOPMENTALBIOLOGY
by strain and cell-cell adhesion in the epithelium of an arthropod leg segment. J. Tl,con Rid. 100, 443-483. Morriss-Kay, G., and Tuckett, F. (1987). Fluidity of the neural epithelium during forebrain formation in rat embryos. .I. Cell Sci. Supljl. 8, 433-449. Nardi, J. B. (1981). Induction of invagination in insect epithelium: A paradigm for embryonic invagination. S~ienw 214, 564-566. Nardi, J. B., and Magee-Adams, S. M. (1986). Formation of scale spacing patterns in a moth wing. I. Epithelial feet may mediate cell rearrangement. Del%. Rid. 116,278-290. Odell, G. M., Oster, G., Alberch, P., and Burnside, B. (1981). The mechanical basis of morphogenesis. I. Epithelial folding and invagination. Dee Bid. 85, 446-462. Owaribe, K., Kodama, R., and Eguchi, G. (1981). Demonstration of contractility of circumferential actin bundles and its morphogenetic significance in pigmented epithelium it/ ,sitro and it/ viw ,1 (‘PI/ Rid SO, 507-514. Pictet, R. L., Clark, W. R., Williams, R. H., and Rutter, W. J. (1972). An ultrastructural analysis of the developing embryonic pancreas. Ijut*. Bid. 29, 436-467. Pino-Heiss, S., and Schubiger, G. (1989). Extracellular protease production by Drosophi/u imaginal discs. [jet.. Biol. 132, 282-291. Schoenwolf, G. C., and Alvarez, I. S. (1989). Roles of neuroepithelial
VOLUME158. 1992
cell rearrangement and division in shaping of the avian neural plate. I)c,IlPlopntr)lf 106, 427-439. Schoenwolf, G. C., and Franks, M. V. (1984). Quantitative analyses of changes in cell shapes during bending of the avian neural plate. Dw.
Rid. 10.5,257-272. Scholtz, (:. (1990). The formation, differentiation and segmentation of the post-naupliar germ band of the amphipod C;U~~UI.IIS /,/r/e.r L. (Crustacea, Malacostraca, Pericarida). Ptvc. R S’oc: Lordorc H 239, 163~211. Smuts, M. S., Hilfer, S. R., and Searls, R. 1~.(1978). Patterns of cellular proliferation during thyroid organogenesis. J. E>wtlrr!yol. Exp. ;2for-
/hd. 48, 269-286. Spooner, B. S. (1973). Microlilaments, cell shape changes, and morphogenesis of salivary epithelium. .+ln,. Zoo/. 13, 1007-1022. Weisz, P. IS. (1947). The histological pattern of metamcric development in :1 rtcarr/itr .scdit/tr. J. MwpAol. 81, 45-M. Wessels, N. Ii., Spooner, B. S., Ash, J. F., Bradley, M. O., Luduena, M. A., Taylor, E. L., Wrenn, J. T., andyamada, K. M. (1971). Microfiaments in cellular and developmental processes. Scirtlce 171, 1X-
143. Zwaan, J., and Hendrix, K. LV. (1973). Changes in cell and organ shape during early development of the ocular lens. 4,~. Zoo/. 13, 103910.19.
BIOLOGY
152, %‘9-2%? (1998)
Epithelial Morphogenesis in Developing Artemia: The Role of Cell Replication, Cell Shape Change, and the Cytoskeleton
The roles of cell replication and shape change as morphogenetic forces in epithclial inxagination were examined in instar II Arte~in. The epidcrmal cells underwent a fixed pattern of cell division during the first 5 hr of instar II. Greater cell replication in the thoracopod bud (ThB) than in the arthrodial membrane (AM) region resulted in a higher density of epidermal cells in the ThB region (differential cell density). The ratio of cell density (AM/ThB) declined from 1.0 to less than 0.80 by Hour 2 of instar II. Invagination of the AM occurred during Hour 1 when the AM/ThB reached 0.75. A 2-hr pulse with 5’fluorodeoxguridine (FudR) during instar I delayed completion of the cell replication pattern and devclopmerit of transverse cell files in the ThB region for a period equal to the length of the exposure. The delay in thv cell division program resulted in a ccl1 density ratio of 0.93 at Hour 4, a value normally observed in Hour 2 larvae, and cvagination of the epidermis did not occur at apolysis (Hour 4). The FudR treatment did not perturb the cytoskcteton or the initial steps in cell shape change and the larvae formed small segments during instar III. Cell shape change Lvithin the AM began during Hour 4 as this region became significantly thinner than the neighboring ThB region (thickness ratio, AM/ThB - 0.77). Before apolysis the AM cells became wedge shaped, a change which occurred when the basal region of the cell enlarged. The microtubules and microfilaments were reorganizcad from the apical cytoplasm to thr lateral border of apposing AM cells. Following apolysis (Late Hour 4) shape change was completed as the cells attained a thin spindle form, with microtubuleand microfilament-rich filopodial extensions which overlapped adjacent AM cells. As contact with ThB cells shifted from lateral to apicolateral, the AM cells formed the innermost edge of the inragination. Microtubules in the differentiating AM cells contained tyrosinated, detyrosinated, and acetylated o-tubulin isoforms. Treatment with nocodazole, colchicine, tasol, or cytochalasin B blocked AM ~11 shape change and inhibited segmentation, but did not affect the mitotic pattern or differential cell density. Re conclude that the specific pattern of cell division led to differential cell density which, along with AM cell shape change, established the conditions ntressar> ‘( 1992 Academic Press, Inc. to achieve epidermal evagination.
Smuts et r/l., 1978; Meier, 1978; Goldin and Wessels, 1979; Bard and Ross, 1982). Evidence supporting the contenMorphogenesis in many developing tissues begins by tion that one mechanism generates the primary force in evagination or invagination of an epithelial sheet of a certain tissue and that the other mechanisms are inactive has not been forthcoming. Indeed, it is likely that cells to form a tube or pouch (Ettensohn, 1985a; Hilfer and Searls, 1986; Kolega, 1986; Fristrom, 1988, for re- tissue development results from the simultaneous or seview). The cellular basis for these tissue changes may quential occurrence of several morphogenetic processes, include microtubule- and microfilament-mediated cell each contributing to the shape of the epithelium. shape change (Wessels et ul., 1971; Burnside, 1973; Although growth is usually seen as a simple increase Spooner, 1973; Ode11et al, 1981; Owaribe et ccl.,1981), cell in the mass of the developing tissue, cell replication and rearrangement or repacking (Fristrom, 1976; Mit- growth may become a morphogenetic force during early tenthal and Mazo, 1983; Ettensohn, 198513;Hardin and organogenesis if it contributes to lateral (mitotic) presCheng, 1986), disparity of cell adhesion (Nardi, 1981; sure (Ettensohn, IY%a, for review). For example, by inMittenthal and Mazo, 1983), and mitotic pressure (Pic- creasing cell density within a limited area of the epithetet et nl., 1972; Hilfer, 1973; Zwaan and Hendrix, 1973; lium, lateral force would be generated. The force may be restrained by contact with a rigid substratum, and thus be a potential force, or it may be immediately exerted on adjacent regions. Uneven distribution of force through’ Present address: Department of Internal Medicine, University of Alabama Hospital, 619 19th St. South, Birmingham, AL 3.i233. out the tissue would exert a greater effect in weak reINTRODIJCTION
279
280
DEVELOPMENTALBIOLOGY
VOLUME
152,1992
FIG. I. Brine shrimp larvae at instars I (A), II (B), and IV (C). The bracket in each micrograph indicates the region of morphogenesis of the first thorax segment. Arrowheads in B and C indicate the AM region. Anterior is to the top in each figure. The bar in A represents 100 pm, and the magnification is the same for all micrographs
gions, creating a situation in which buckling of the epithelium occurs. Mitotic pressure may work in concert with one or more other mechanisms, as demonstrated for neural development (Morriss-Kay and Tuckett, 1987; Schoenwolf and Alvarez, 1989) and salivary gland (Spooner, 1973). To examine further the interplay of these potential morphogenetic mechanisms, we investigated the role of cell replication, shape change, and rearrangement during invagination of the epithelium of the developing brine shrimp, Artemia (Fig. 1). The epidermal cells constitute a monolayer epithelium of polarized cells attached to the cuticle and organized in a pattern of longitudinal or transverse files (Fransenmeier, 1940; Weisz, 1947; Freeman, 1989a; MacRae et al., 1991). The limb primordium of the thoracopod (ThB) evaginates as the adjacent arthrodial membrane (AM) epidermis invaginates upon release of the epidermis from the cuticle (apolysis) (Freeman, 1989b). In this study, the contribution of each morphogenetic mechanism to epithelial invagination and segment development was investigated. METHODS
Brine shrimp cysts (Sanders Brine Shrimp, Salt Lake City, UT) were hatched and reared in artificial seawater (Instant Ocean, Mentor, OH) at room temperature (23°C). Experiments were conducted with larvae se-
lected for synchronous development at hatching thus at the same age and developmental stage.
Larvae prepared for cell density
analysis
and
were fixed
in toto in Carnoy’s fluid, rinsed, and stained with 1.0 pg/ml Hoechst 33258 in phosphate-buffered saline (PBS), pH 7.4. The larvae were mounted in 80% glycerol for viewing. For cell cycle analysis and determination of cell density, the video image (Dage-MTI67M video camera) was captured with a video frame grabber board (PCVISIONplus, Imaging Technology Inc., Woburn, MA) controlled by ImageMeasure 4100 software (MicroScience, Inc., Federal Way, WA). The cell cycle phase was determined using fluorescence intensity of the nucleus. The image was averaged 10 times and the background subtracted before measuring the fluorescence. The fluorescence intensity was compared to the area of the nucleus to confirm the cell cycle phase, as previously defined (Freeman and Chronister, 1989). The cell density was ascertained by digitally overlaying a 50 x 18-pm rectangle over the image of the epiderma1 cells and counting the nuclei within the box. The ratio of the number of cells in the presumptive AM versus the number of cells in the ThB was used to define differential cell density, reported here as the AM/ThB ratio. Analysis of variance was performed after arcsine transformation of the ratio.
FIG. 2. Spatial pattern of cells (Hoechst-stained nuclei) in the ventral epidermis of segments 1 and 2 during the first 5 hr of instar II. (A) Ikring the first hour of instar II (Hour 0) the cell pattern changes from seven longitudinal files to eight transverse files numbed from anterior (1) to posterior (8). The middle ThB cells of the first segment (file 1) are aligned transversely (large arrowhead) and mediolateral divisions have occurred in the first three ~11s. The cell indicated hy “2” is the fourth cell in the mediolatcral file and is ahout to divide. The cell to the right is in metaphasr and the division axis is mediolateral. Cells in the AM region (transverse files 3 and 5) are indicated hg small arrowheads. (8) Iiour 1. (C) Hour 2. (D) Hour 3. The increase in cell replication in the ThB cells of the first segment (transverse files l-3) has led to a differential cell density. (E) Hour 4. Cell shape change is beginning in the AM region (files 4 and 5). (F) Hour 5. Apolysis has occurred. Anterior is to the top in all figures and the ventral midline is indicated hy “m”. The rectangle in A has the dimensions of 18 X 50 pm and serves a:: a six> rc~fcrencc for the cell density determinations. Bar in A 2 20 pm, and the magnification is the same for all micrographs.
To examine the importance of differential cell density in segmentation and evagination, the cell cycle of the .epidermal cells in the ThB region was blocked by exposure to 5’-fluorodeoxyuridine (FudR) (10 pg/ml seawater) during Hours 13 and 14 of instar I. This time
period includes S phase for the cells in the anterior ventral region of the thorax (Freeman and Chronister, 1989). Larvae were then transferred to drug-free seawater after which the cell division pattern and the differential cell density at Hours O-5 of instar II were de-
282
D&EISWMENTAI.
1.00
P
BIOLOGY
-
0.80 : 0.90 0.70
-
0.60
-
~‘q~~~:;,:i 1
0.50 '
I 0
1
3
2
Hour
of
lnstar
4
5
II
FIG. 3. Change in the ratio (AM/ThB, ordinate) of cell density (open triangles) and cell thickness (black triangles) during early instar II. Each point represents the mean and one standard deviation of 20-24 larvae. The change in cell density occurs during the first 3 hr while the differential cell thickness does not change significantly until Hour 4.
VOLUME 152,1992
PBS) (Molecular Probes, Eugene, OR), and mounted in 80% glycerol containing 2% n-propyl gallate. Antitubulin antibodies used and their specificities were: DMlA (cu-tubulin, Sigma), KMX (/$tubulin, a gift from Dr. K. Gull), anti-Glu (detyrosinated a-tubulin, a gift from Dr. C. Bulinski), 6-llB-1 (acetylated cu-tubulin, a gift from Dr. G. Piperno), and YL1/2 (tyrosinated cu-tubulin, Dimension Laboratories Inc. Mississauga, Ontario). These and other specimens were observed and photographed with a Leitz Dialux epifluorescence microscope or a Sony video printer. To study the role of microtubules and microfilaments in AM cell shape change, larvae were exposed to colchitine or nocodazole (10 pg/ml) to disrupt microtubules, taxol(2 PM) to stabilize microtubules, or cytochalasin B (20 pg/ml) to disassemble microfilaments. The drugs were placed in artificial seawater at specified times, after which the larvae were observed for progress in segmentation and fixed for immunocytochemistry.
termined. These larvae were also fixed and stained for microtubules and microfilaments, as described below. Analysis
of AM Cell Shupe Ch,ange
The thickness of the epidermis in the AM and ThB regions was determined in living and fixed specimens by image analysis. After the image was captured as described above, measurements were made using the Line Distance subroutine from the ImageAction program (Imaging Technology). The measurements included (1) the thickness of the AM and ThB regions (distance from apical to basal surfaces), (2) the width of a group of three ThB cells in one file (cumulative cell diameter) for cell packing measurement, and (3) the width (diameter) of apical and basal regions of the AM cell at different times during instar II. For the cumulative cell diameter determinations, the cell boundary was considered to be a point equidistant between nuclei. All dimensions were measured digitally after the X and Y coordinates of the field were calibrated with a slide micrometer. Microfubu,les
and Microjilamexts
Microfilaments and microtubules were detected as previously described (MacRae et al., 1991). Briefly, larvae were fixed overnight in 4% paraformaldehyde in PBS. After treatment with PBS containing 0.5% albumin and 0.5% Triton X-100, the specimens were stained overnight with antitubulin antibodies, rinsed in several changes of PBS, stained with FITC-labeled goat antimouse IgG (1:40) (Sigma Chemical Co., St. Louis, MO), rinsed in PBS containing TRITC-phalloidin (20 PM in
RESIJLTS
Pattern of’ Replicution AM and ThB Cells
in
arid Cell Cycle qf’ Presumptive bxtar II
Epidermal cells involved in formation of segments 1 and 2 were identified by cell lineage studies (Freeman, 198913,unpublished). At ecdysis to instar II (Hour 0) the presumptive AM cells were arranged in seven longitudinal files in the ventral epidermis; each file extending from the cephalon to the tail. The orientation of the file
TABLE
1
MORPHOMETRICDETERMINATIONOFCELLDIMENSIONSOF AM CELLS AND THB CELLS ADJACENT TO THE AM REGION AT HOUR 2, HOUR 4 (BEFOREAPOLYSIS),ANDHOUR~(AFTERAPOLYSIS) Morphometric dimension AM cell height ThB cell height Diameter of apical AM cell surface Diameter of basal AM cell region Cumulative diameter of three ThB
Hour 4 after apolysis
Hour 2
Hour 4 before apolysis
9.0 (1.06) 9.6 (.61)
9.4 (Xl) 11.1 (.81)
7.5 (94) 9.3 (2.1)
6.6 (.54)
6.9 (1.3)
5.05 (.67)
5.6 (X4)
7.8 (.97)
6.5 (1.0)
23.1 (2.4) 22.9 (1.3)
18.6 (2.8) 17.7 (1.0)
20.45 (2.0) 21.95 (1.4)
WllS
CT) (S)
Not<>. Each value (in pm) is the mean (standard deviation) of lo-15 measurements of that dimension. The cumulative diameter of ThB cells was measured from transverse (T) and surface (S) views.
FREEMAN,CHESHIRE,ANDMACRAE
E~ithdicrl
12k,r.pl/o!/rt/c’sis in .A rtfjirc it/
283
FIG. 4. Recovery from treatment with Y-fluorodeoxyuridine. (A and B) Hoechst-stained nuclei in control and treated larvae, respectively, at Hour 2 instar II. The drug slowed the temporal pattern of cell replication but did not afl’ect cell position in the epithclium or the orientat ion 01 mitosis. ((1) Surface view of Hour 2, instar II FudR-treated larva showing no disruption of microtubules by the drug. (D) Same larva as in (’ showing no disruption of microfilaments. iE) Surface view of FudR-treated Late Hour 4, instar II larva showing microfilaments in AM cells. The >lM cells hare begun to change shape. (F and G) Appearance of segment 1 in control (F) and treated (G) larvae in instar III showing a smaller hut otherwise normal limb bud in recovering larvae. Cells in the mid-ThB region (file 1 in A, (’ and E, file 2 in B and D, are indicated by the large arrowhead. Cells in the AM region (files 4 and j) are indicated by two small arroa-heads. Antrrior is to the top in all ligures. m, the ventral midline. Bar in .4 wpresents 20 pm for A-E. Bar in F represents 20 pm for F and G.
is indicated by the pattern of nuclei rather than the cell borders. During the first hour of the instar the cells that were to form the posterior region of the ThB of segment I, the AM, and the anterior region of segment 2 underwent a change such that they appeared to be organized in eight transverse files (Fig. 2A). This transformation is a result of cytoplasmic changes rather than alteration of cell neighbors. Cells in files numbered 1 and 8 will form the apex of the first two segments, respectively. Transverse files 2 and 3 formed the posterior face of the first bud 1 and files 6 and 7 formed the anterior face of the second bud. Cells in transverse files 4 and 5 formed
the AM, the region between the ThB buds of the first two segments. The cell cycle behavior in the AM region differed from that in the ThB region and the tissue types were defined by the cell cycle pattern. The cells were in G2 phase at ecdysis to instar II. Those cells in the presumptive ThB region of the first segment underwent one rapid replication in the transverse direction, beginning at Hour 0 with the medial cells of file 1 (Fig. 2A). The same pattern was observed at progressively later times in files 2 and 3 (Figs. 2B and 2C). By Hour 2, all cells in files 1 and 2, but none in file 3, had divided (Fig. 2C). Divisions of
284
DEVELOPMENTAL
BIOLOGY
0.90
152, 1992
invagination of the AM region removed the nuclei from the plane of focus. During the period of replication, the ThB cells increased in height (Hour 2 to Hour 4, Table 1). In addition, the cumulative width (diameter) of three ThB cells in one file decreased from approximately 23 to 18 pm. These morphometric determinations suggest that the diameter of the cells decreased during division and the height increased as the postmitotic cells grew.
1 .oo
F
VOLUME
p!r
0.80 0.70 0.60 t 0.50 ’
I 0
1
2
Hour
4
3
of
lnstar
5
II
FIG. 5. ERect of a 2-hr pulse with FudR on cell density (triangles) and cell thickness (circles) in control (open symbols) and treated (solid symbols) larvae. Each point represents the mean and one standard deviation of lo-20 larvae.
the presumptive ThB cells of the first segment always preceded those in segment 2. Similar to the first segment, cell divisions in the second segment began with the file furthest from the AM region (file 8) and then occurred in files 6 and 7. By Hour 3 almost all cells in files 1-3 and 6-8 had divided (Fig. 2D). In contrast to the ThB cells, the AM cells (files 4 and 5) remained in G2 throughout the first 4 hr of the instar (Figs. 2A-2D) and at the time of cell shape change were in the G2 phase (Figs. 2E and 2F). After invagination, the AM cells were no longer in the plane of focus and were difficult to observe. It was possible that the AM cells divided later in the instar, but no mitotic figures were noticed at Hour 5.
Mitotic pressure may act as a morphogenetic force if the pattern of replication results in a significant difference in the density (number of cells per 50 X l&pm area) of cells in adjacent regions of the epithelium, hereafter referred to as differential cell density. Cell replication in the presumptive ThB regions of segments 1 and 2 increased the number of cells in the measurement region from approximately 12 cells at Hour 0 to 1’7 cells at Hour 4. These cells were still attached to the cuticle at their apical surfaces. During the same time, the number of AM cells did not change significantly. The ratio of cell density (AM/ThB) declined from 0.99 at Hour 0 of instar II to 0.75 at Hour 3 (Fig. 3), and the ratio did not change significantly (P > 0.05) for the next 2 hr, during which apolysis and evagination occurred. It was difficult to accurately measure the cell density after Hour 5 since
Diflerential
Cell Density Produces Mitotic
Pressure
To determine the role of changing cell density in segment morphogenesis it was necessary to separate the process of cell replication from other potential effecters, such as cell shape change, by altering the temporal pattern of developmental events during early instar II. To accomplish this, replication of the ThB cells was inhibited by treating larvae with a pulse of FudR at Hours 13-15 of instar I, when almost all of the epidermal cells in the presumptive ThB region of the first segment were traversing S phase (Freeman and Chronister, 1989). We reasoned that if, upon removal of the drug, the normal cell replication pattern resumed, division of the ThB cells compared to that of the AM cells would be delayed by 2 hr. The appearance of differential cell density between the ThB and the AM regions would also be delayed by 2 hr. Examination of larvae in control and treated groups at hourly intervals during early instar II showed that the spatial and temporal patterns of transverse cell replication were normal following drug treatment, but they were delayed by 2 hr (Figs. 4A and 4B). The cell density ratio at Hour 4 in treated larvae was 0.94, a value not significantly different from that of Hour 2 larva (Fig. 5). Thus, the delay in pattern generation and differential cell density was similar to the period of the FudR block and there was no indication that the drug-treated cells accelerated or otherwise altered the timing of replication. At apolysis (Hour 4), neither ThB evagination nor AM invagination had occurred in FudR-treated larvae. The results suggest that disparate replication patterns within the epithelium established differential cell density which in turn was necessary for segmentation. The FudR treatment did not affect the pattern of microtubules or microfilaments in the epidermal cells of the ThB and AM regions (Figs. 4C and 4D) ahd the AM cells demonstrated typical cell shape change at Hour 4 (Fig. 4E). Similarly, the AM/ThB thickness ratio (see below) for both the control and the treated larvae decreased at Hour 4 (Fig. 5). During the next instar the treated larvae continued segment formation, although
FREEMAN.~HESHIRE,AND
MACRAE
&pith
fditrl ~~~~~~~~l~o~~r~~~f~.si.s ii, .-l rfcii, itr
285
I;. Appearance of the AM and ThB regions of the epidermis at Hour 2 CA) and Latr Hour 4 (BJ. In B apolysis has occurred over the AM and nt ThB cells; the old cuticle is indicated hg “c”. Large arrowhead indicates the Th8 rexion; small arrowhead indicates the AM region The lines indicate the location of the lateral borders of one AM cell. They show the ne&,rc-shaped ccl1 in R. Bar represents 20 pm for both
the segments were smaller than those of control (Figs. 4F and 4G).
larvae
The AM cells Ivere the first of the epidermal cells in the thorax to differentiate during segmentation (Cheshire, 1987; Freeman, 1989a), undergoing a complex shape change before assuming a thin spindle form. The shape change of the AM cells began shortly before apolysis during Hour 4 and was completed after apolysis (Figs. 6 and 7). By Hour 4 the ThB cells had undergone enhanced replication and were columnar. The AM cells were still attached to the cuticle and their apical surfaces did not change in area (diameter of apical region, Table 1) while the basal region enlarged significantly (P < 0.05) (diameter of basal region, Table 1). The height of the AM cells did not change appreciably (Table 1). In some larvae the basolateral regions of these cells shifted slightly with respect to adjacent AM cells (Fig. 7, Hours 4 and 5).
Following apolysis over the AM and neighboring ThB cells, invagination began. Contact between the AM cells and their ThB neighbors changed from lateral to more apicolateral as the diameter of the apical region of the AM cell decreased (Figs. 6B and 7; Table 1). By Hour 5, apolysis was completed over all of the thorax epidermis and the AM cells had reached an innermost position in the invaginating AM region. The epithelium in the AM region became thinner than that in the ThB region as shape change progressed. It is expected that the thinner epithelium would deform (invaginate) before a thicker region in response to lateral pressure. To determine if the difference in cpidermal thickness was important in the invagination process, the thickness of the AM and ThB regions was compared at different times in instar II (AM/ThB ratio, Figs. 3 and 5). Over the first 3 hr of instar II the ratio declined slightly (P > 0.05). However, a decrease in the height of the AM epithelium combined with an increase in the height of the ThB cells during Hour 4 brought the ratio
286
DEVELOPMENTALBIOLOGYVOLUME~~~, 1992
Hr2 Hr4
Hr4
Hr4 APO
Hf5 Hr 5
FIG. 7. Diagrammatic view of the cell shape changes in the vcntrolateral epidermis. The transverse view of a 60-wrn region of the integument is shown on the left and the surface view of the AM region and neighboring ThB cells (two from segment 1 and two from segment 2) is shown on the right. The bar at the bottom of each panel represents 20 pm. (Left) At Hour 2 the cells are similar in size, still attached to the cuticle (C), and demonstrate a polarized cgtoskeleton with microtubules (MT) and microfilaments (MF) enriched in the apical region. The AM cells are indicated by arrows. By Hour 4 the ThB cells have undergone enhanced replication and are columnar. The AM cells are still attached to the cuticle and their basal regions have enlarged. The organization of the microtubules and microfilaments has shifted from primarily apical to enriched along the lateral membranes of apposinp AM cells. The basolateral regions of one AM cell may shift transversely with respect to the cell in the other AM file. Later in Hour 4 the izM cells and the neighboring ThB cells undergo apolysis and invagination begins. The neighboring ThB cells now decrease in height and are not as densely packed. As invagination proceeds, cell-cell contact of the AM cell with the ThB neighbors changes to a more apicolateral position. By Hour 5, apolysis is complete and the AM cell has invaginated to the greatest extent. The arrowhead indicates the position of the other AM cell. (Right) Surface view of the invaginating AM and neighboring ThB cells. Depth of inragination is indicated by stippling. Anterior is to the left in left panel and to the top in right panel.
to below 0.8, where it remained through the period of invagination (Figs. 3 and 5). Morphometric measurements before and after apolysis demonstrated that the packing of neighboring ThB cells increased prior to Hour 4 as the ThB cells repli-
cated (cumulative diameter of three ThB cells, Table 1). Measurement of ThB density showed that both density and height decreased after apolysis, suggesting that the ThB cells were spreading following release from the cutitle. This change in packing density may be sufficient
FIG. 8. Fluorescence staining of microtubules and microfilaments in the vcntrolatcral epidermis of instar II Artemicr. Large arrowhead indicates the ThB region of the first segment; small arrowhead indicates the AM region between the first two segments. (A) Surface view of microtubules in epidermis of Hour 3 larva. (B) Surface view of microfilaments in epidermis of Hour 3 larva. (C) Transverse view of microlilaments of Hour 3 larva showing apical staining of cells. (D) Surface view of Hour 4 larva stained for acetylated cu-tubulin. The AM cells are beginning to change shape (compare to A). (E) Surface view of Late Hour 4 larva showing tubulin in AM cells during invagination. (F) Microfilaments in Hour 4 larva during shape change (compare to B). (G and H) Surface view showing enrichment of microtubules (G) or microfilaments (H) in apposing regions of AM cells (files 4 and 5) at Hour 4. (I and J) Transverse view of Hour 4 larva showing mid-AM border with concentration of microfilaments (I) and microtubules (J) (small arrowhead) in the adjacent regions of apposing AM cells in files 4 and 5. The white dashed lines indicate the border of the AM cells with the neighboring ThB cells and demonstrate the wedge shape of the cell. (K) Hour 1 larva with plane of photograph at the level of transverse view of AM on the left side and surface view on the right side. The microfilaments are enriched in the lateral region of the wedge-shaped cell (small arrowhead, left side) and between apposing cells (arrowhead, right side). The white dashed lines elucidate the location of lateral horders of one AM cell and demonstrate the wedge shape of the cell. The specimens were stained with antitubulin antibodies KMX in A, DMlA in E and G, 6-llB-1 in D, YL1/2 in J, and phalloidin in B, C, F, II, I, and K. The bar represents 20 pm in all figures. n, nucleus.
288
DEVELOPMENTAL BIOLOGY
VOLUME 152, 1992
FIG. 9. View of the basal surface of epidermal cells in the invaginating AM region. The filopodia (arrowheads) extend transversely and overlap adjacent AM cells. (A and B) The same larva stained, respectively, for microtubules containing acetylated cu-tubulin and microfilaments. (C and DI Same larva stained, respectively, for tu-tubulin (DMlA) and microfilaments. Bar represents 20 pm for all figures. n, nucleus.
to produce lateral pressure which would cause deformation of the AM cells.
Reorgunixation of Microtubules during Cell Shape Change
and Microjilaments
Immunolocalization of microtuhules and microfilaments revealed that cells in the presegmentation thoracic epithelium contained a polarized cytoplasm with the apical population of microtubules located near the surface and approaching the lateral borders of the cell (Fig. 8A; see also MacRae et al., 1991). Medial and basal regions of the cell contained fewer microtubules. Microfilaments were distributed in a manner similar to that of the microtubules (Figs. 8B and 8C). The intracellular location of microtubules and microfilaments in the AM cells changed during Hour 4, instar II (Figs. 8D-8K). Surface views of larvae revealed that microtubules and microfilaments were colocalized as the cells changed shape and assumed a spindle form (Figs. 8D-8F, compare to A and B). Early in Hour 4 there was an enrichment of microtubules and microfilaments near the apposing membranes of AM cells located in files 4 and 5, the point of eventual invagination (Figs. 8F-8H). A transverse view of the AM region revealed that, as the cell became wedge shaped, there was a redistribution of the microtubules and microfilaments from the apical to the midlateral and basal regions of the cell (Figs. 81-8K). The change in location of microtubules and microfilaments was completed just prior to apolysis (Late Hour 4, instar II). After apolysis, shape change was completed as the AM cells became spindle shaped and extended lateral filopdia-like projections (Fig. 9). The extensions were always directed mediolaterally and overlapped basolatera1 regions (Figs. 9A and 9B) and filopodia (Figs. 9C
and 9D) of neighboring cells. Microfilaments and microtubules were concentrated in the extensions. Invagination was complete shortly after apolysis (Fig. 7, Hour 5). Specimens in which the dorsal thorax had been dissected away revealed that the leading, curved edge of the invagination was formed by one AM cell, either from file 4 or file 5 (Figs. lOA-1OC). The remaining AM file and adjacent ThB files formed the walls of the invagination (Fig. 10D). Studies with several cell types from different organisms have demonstrated that certain cell activities may involve a change in the composition of post-translationally modified tubulins (MacRae and Langdon, 1989, for review). To determine if post-translationally modified tubulins were important in AM cell shape change, larvae were stained with antibodies to tyrosinated, detyrosinated (not shown, MacRae et al., 1991), and acetylated cu-tubulins. Antibodies to acetylated (6-llB-1) and detyrosinated cu-tubulin isoforms stained the invaginating cells in a manner similar to DMlA (oc-tubulin) and KMX (/3-tubulin). All isotubulins occurred in the AM and ThB cells (Figs. 8A, 8D, 8G, 85; 9A). Microtubules and Microjilaments Cell Shape Change
Are Necessary for AM
To determine the role of cytoskeletal elements in cell shape change and thinning of the AM epithelium, the effects of microtubule and microfilament reactive drugs on Hour 4 larvae were examined (Table 2). In larvae exposed to cytochalasin B (20 pg/ml), taxol (5 PM), colchicine (10 pg/ml), or nocodazole (10 wg/ml), the AM/ ThB cell thickness ratio never dropped below 0.9 (Table 2), showing that the reduction in cell height associated with cell shape change involved both of the cytoskeletal elements. The cell pattern and the differential cell density achieved prior to Hour 3 were not affected by the
FREEMAN,CHESHIRE.ANDMACRAE
inhibitors (Table 2). These compounds did, however, block formation of spindle-shaped cells and filopodial extensions. Colocalization of microtubules and microfilaments during cell shape change suggests that the interaction of the two elements is essential to the process and that they are structurally interdependent. We observed, however, that taxol, nocodazole, and colchicine disrupted the organization of microtubules but did not affect microfilaments. Similarly, larvae treated with cytochalasin B showed loss of microfilaments but the microtubule staining pattern was unaffected.
To determine if cell rearrangement was active in limb bud development, the pattern of cells was observed during the 5hr period in which the limb bud was delineated. Based on the position of the nuclei, we could discern no evidence for movement of any cell in the ThB region other than that due to mitosis. Prior to apolysis in the AM, no shift was more than one-half of a cell diameter. We cannot rule out the possibility that some rearrangement may have occurred following apolysis (Hour 5), as suggested by the shift in basolateral regions and spindle cell formation. I)ISCIJSSION
FIG. IO. View of the l~sal surface of AWI invagination at different [)lancls of focus. (A and 8) Innermost e&y of invagination showing a single wll stained for microtubules (A) or microfilaments (B). Thtl tilopotlia arc intlicntcd hy arrowheads. CC:)Slightly more apical view ol’ ncighhoring cells in the e&!c of the invagination. (D) Apical-most vitj\v at the v&y of the AM inragination, stained for nricrotubulcs. Slwcimcans in A, (‘. and L) n-cre stained with 1)MlA while that in B was staincvl with I)halloidin. n. nucleus. The bar wpresents 20 pm fur all liguws
The results of this study support a model of epithelia l invagination in which cell replication plays an impor tant role as a morphogenetic force in segmentation. Th e temporal and spatial patterns of cell division are highl! regulated, and such control is required to attain the differential cell density in the AM and ThB. The AM and ThB cell populations are distinguished by their replication during the second instar with patterns similar to those found in the segmenting regions of other crustacean embryos (Scholtz, 1990). While the developmental step that regulates the pattern of division, and therefore defines the two cell populations, is not known, it clearly is initiated prior to instar II (Freeman, 19891)). Although mitotic pressure is involved in the processes of evagination (Zwaan and Hendrix, 1973; Smuts cjt trl, 19’78), accordion folding (Pictet ef ul., 1972; Bard and Ross, 1982), and budding of epithclia (Goldin and Wcssells, 19’79), it has been difficult to evaluate its importance with respect to other morphogenetic activities. In this study the role of cell replication was examined by experimentally dissecting morphogenesis into replication-dependent (FudR-sensitive) and replication-indcpendent processes. The FudR-induced delay in the increase in cell density did not perturb the shape change in the ,4M cells or the timing of apolysis. Without dif-
290
DEVELOPMENTALBIOLOGY
VOLUME 152,1992
TABLE 2 EFFECTOFMICROTUBULEANDMICROFILAMENTREACTIVEDRUGSONCHANGESINCELLDENSITYANDSHAPE Cell density Treatment Control Hour 0 Hour 2 Hour 3 Hour 4 Hour 5 Cytochalasin Taxol Colchicine Nocadazole
AM
B
11.9 13.2 12.3 13.7 14 13.2 12.8 13 11.4
(.7) (.6) (.63) (.5) (.4) (.6) (.5) (.5, (.7)
ThB
12.1 16.5 15.2 17.7 19.5 18.7 17.4 17.3 15.6
(.7) (.9) (.74) (.6) (.9) (.Y) (.7) (.6) (.6)
Cell thickness AM/ThB
.98 .8 .81 .77 .71 .71 .75 .75 .73
AM
10.4 9.8 11.1 9.2 8.2 11.7 11.4 11.5 11.6
(3) (1.0) (.9) (3) (.4) (.7) (.7) (.6) C.7)
ThB
10.6 10.4 11.8 11.97 11.8 12.1 12 12.3 12.3
(.9) (1.2) (.9) (.8) (.7) (.9) (3) (.5) (.6)
(pm) AM/ThB
.98 .94 .94 .77 .7 .96 .95 .93 .94
LVofr. Exposure to the drugs began at Hour 3 and the larvae were fixed at Hour 5. Each value in the cell density columns AM and ThB represents the mean (-tl standard deviation, w = 10-20) number of nuclei per 18 X 50-pm region of the epidermis. Each value in the AM and ThB columns under cell thickness represents the mean (+l standard deviation, 7~= 10-20) epidermal thickness (gm) in that region.
ferential cell density, the AM region did not invaginate at apolysis. Although we can not eliminate the possibility that the delay in cell replication simply resulted in an insufficient number of ThB cells for complete invagination, we favor the idea that differential density could provide the morphogenetic force needed for mitotic pressure after apolysis. Morphometric measurements show that, prior to apolysis, ThB cells increase in height as the cell number and density increase, suggesting that both replication and growth contribute to the increase in cell density and mitotic pressure. During this period the mitotic pressure may be stored as “potential” lateral force since expansion of the epithelium is inhibited by attachment to the l-pm-thick cuticle at the apical surface (Freeman, 1989a,b; Horst, 1989; Criel, 1991). Contact with the cuticle would also have inhibited cell rearrangement, as has been found in developing imaginal disc cells (Pino-Heiss and Schubiger, 1989; Birr et a,l., 1990; Condic et al., 1991). Immediately after release from the cuticle the free ThB cells decrease in height as the epithelium spreads slightly (Fig. 7), establishing a kinetic force. Since the mid-segment ThB cells are still attached to the cuticle (Freeman, 1989b), the lateral pressure generated by the expanding ThB region cannot be directed in any direction except toward the AM. The cellular mechanism by which the mitotic pressure could be stored and subsequently transformed into lateral pressure is not understood. Expansion of the ThB epithelium is associated with a decrease in cell height, suggesting a reorganization of ThB cell cytoplasm. The results do not permit us to discern whether this reorganization is passive, occurring in response to release from the cuticle, or is an active shape change
that provides the force for lateral expansion and AM invagination. We were unable to detect any alteration in the distribution of microtubules and microfilaments in the ThB cells during expansion, as would be expected if the cytoskeleton was active in cell shape change. For mitotic pressure to effect epithelial deformation there must be a region of cells that deforms before the rest of the epithelium. In brine shrimp this is the AM region where the epithelium becomes thinner as the cells undergo shape change. Shape change begins as the cuboidal cells become wedge shaped while still attached to the cuticle. This change occurred in a manner that appears to be different from apical constriction (pursestring effect) exhibited by neural epithelium (Burnside, 1973; Schoenwolf and Franks, 1984), thyroid (Hilfer, 1973), salivary gland (Spooner, 1973), and lens (Zwaan and Hendrix, 1973) tissues. The AM cell was attached to the cuticle and there was no reduction in the apical surface area prior to apolysis. Rather, the wedge shape was attained through an enlargement of the basal region, a process associated with a shift of the microtubules and microfilaments from the apical region to the lateral membrane of apposing AM cells and translocation of the nucleus. The second phase of the shape change occurs after apolysis. The cells become spindle shaped with lateral extensions aligned mediolaterally, overlapping adjacent cells. Microtubules and microfilaments were concentrated in the extensions and agents that blocked the assembly of either cytoskeletal element inhibited filopodia formation. The role of the lateral extensions may be to (1) increase the interaction of the two AM cell files, (2) provide directionality and strength to the invaginating cells, (3) facilitate the formation of the inner edge of
FREEMAN,~HESHIRE,AND
MACRAE
the invagination, or (4) rapidly extend the lateral borders of the cell to achieve the spindle shape. Similar extensions occur in insects (Locke and Huie, 1981; Nardi and Magee-Adams, 1986; Delhanty and Locke, 1989) and sea urchins (Hardin, 1989) where they are thought to carry out several developmental processes. Microtubules and microfilaments are colocalized throughout both phases of AM cell shape change, suggesting that there is a functional interaction between them. The microtubule and microfilament inhibitor studies demonstrate that, although both microtubules and microfilaments are required for the shape changes occurring before and after apolysis, they appear to be structurally independent of one another. Thus, microtubules and microfilaments may have complementary functions in AM morphogenesis. The microtubules contain tyrosinated, detyrosinated, and acetylated isoforms of cu-tubulin which were not restricted to certain regions of the cytoplasm, specific cell type, or discrete periods of development. The findings support the contention that the isotubulin diversity does not change during cell differentiation and morphogenesis in instar II (Langdon et al., 1991; MacRae et ul., 1991). Our findings demonstrate a strong role for cell replication in establishing the conditions necessary for epithelial invagination. Mitotic pressure functions in conjunction with cell shape change and, possibly, a change in cell adhesion. Segment morphogenesis in Artenliu then appears to be a result of the temporal and spatial orchestration of the different morphogenetic processes. This research was supported by National Institutes of Health Grant HD24219 to J.A.F., NATO Grant 0865187 to J.A.F. and T.H.M., and an NSERC Grant to T.H.M.
REFERENCES Bard, J. B. I,., and Ross, A. S. A. (1982). The morphogenesis of the ciliary body of the avian eye. II. Differential cnlargment causes an epithelium to form radial folds. De?,. Lzio/. 92, 87-96. Birr, C. A., Fristrom, D.. King, D. S., and Fristrom, J. W. (1990). Ecdysane-dependent proteolysis of an apical surface glycoprotein may play a role in imaginal disc morphogenesis in Drtwopl~ilo. I~ew/o~~~r/c?/t 110, 239-248. Burnside, B. (19731. Microtubules and microfilaments in amphibian neurulation. An,. Zoo/. 13, 989-1006. Cheshire, L. B. (1987). “Role of microfilaments, microtubules, and intermediate filaments in epithelial evagination in the brine shrimp, Artc~v~itr scr~inu.” MS Thesis, IJniversity of South Alabama. 76 pp. Condic, M. L., Fristrom, D., and Fristrom, J. W. (1991). Apical cell shape changes during f~,osophilrc imaginal leg disc elongation: A now1 morphogcnetic mechanism. Dwc/opnzc?~t 111, 23-33. Criel, G. R. J. (1991). Morphology of Arte~mia. 1~ “Artmnio Biology” (R. A. Broane, P. Sorgeloos, and C. N. A. Trotman, Eds.), pp. 119153. CRC Press, Boca Raton, FL. Delhanty, P.. and Locke, M. (1989). The development of epidermal feet
Epitlrdirf
I JI01~~~1~0~~~~~c.si.s iu A rtr~fir icl
291
in preparation for metamorphosis in an insect. Tiswr~ Wl 21, 891909. Ettensohn, C. A. (1985a). Mechanisms of epithelial imagination. Q,ctrrf.Rf?. Hid. 60, 289-307. Ettrnsohn, C. A. (1985b). Gastrulation in the sea urchin is accompanied by the rearrangement of invaginating epithelial cells. Uvi,. Bid. 112, x-390. Fransenmeier, L. (1940). Zur fraye der herkunft des metanauplialen mesoderms und die segmentbildung hei drtcrvitr so/it/o Leach. &if. ,f/tr MSs.sc~rscl/c~,flicl/Zool. 152, 43S472. Freeman, J. A. (1989a). The integument ofArtc>/tlirr during early dcvelopment. Ire “Biochemistry and Cvll Biology of Arfvwitr” (T. H. MacRae, J. C. Bagshaw, and A. H. Rarner, Eds.). pp. 233256. CRC’ Press, Boca Raton, FI,. Freeman, J. A. (IRXDbl. Segment morphoyenesis in drtc~wicc lar\-ac. 111 “Cell and Molecular Biology of Artemia Development” (A. H. Warner, T. H. MacRae. and J. C. Bagshaw, Eds.1. pp. 77-W. Plenum, New York. Freeman, J. A., and Chronister, R. B. (1989). Cell cycle kinetics of epidermal cells in dcvcloping A,?prr/itr: Ir/ sit/l analysis. .I. (:r/t.st. Bid. 9, 193-201, Fristrom, D. (1976). The mechanism of evagination of the imaginal discs of ~rwol,hiltr rt/cltr~co!/trstc~,: III. Evidence for cell rcarrang,rcment. UVI.. Biol. 54, 163171. Fristrom. D. (1988). The cellular hasis of epithelial morphogenesis: A review. Tis.s~~ (‘J/I 20, 6-l.5690. Goldin, G. I’., and \\~eswlls. N. K. (1979). Mammalian lung develol)ment: The possible rolr of cell proliferation in the formation of supernumerary tracheal buds and in branching morphogenesis. .J. Erp. Zool. 208, 337~316. Hardin, J. (1989). Local shifts in position and polarized motility drive cell rearrangement during sva urchin gastrulation. 11~~. Rio/. 156, J3@-l45. Hardin, J. D.. and Cheng (1986). The mechanisms and mechanics of archenteron elongation during sea urchin gastrulation. I)(,,,. 8iol. 115, 19OL!lOl. Hilfer, S. R. (19731. Extracellular and intracellular correlates of organ initiation in the cmhryonic chick thyroid. ..lnl. %oo/. 13, 1023~1038. Hilfer, S. R., and Searls, R. L. (19861. Cytoskclctal dynamics in animal morphogcnesis. 11, “Developmental Biology: A Comprehensive Synthesis” (L. IV. Browder. Ed.), Vol. 2, pp. 4-29 Plenum, Ntv York. Horst, M. N. (1989). Molecular and cellular aspects of chitin synthesis in larval .4r?c,witr. 1~ “Cell and Molecular Biology of Arfc~n/itr Drvrlopmcnt” (Warner. A. H., T. H. MacRac, and Bagshaw, J. (:. Eds.), pp. 59-76. Plenum, New York. Kolega, J. (19861. The cellular basis of epithelial morphogenesis. I,( “Developmental Biology: B Comprehensive Synthesis” (L. 15’. Browder, Ed. I, Vol. 2, pp. 103-143. Plenum, New York. Langdon, C. M., Freeman, J. A., and MacRae, T. H. (1991). Post-translationally modified tubulins in Avtc~wio: Prelarval development in the absence of drtyrosinated tuhulin. Ilcc Bid. 148, 147m155. Locke, M., and Huie. P. (1981). Epidermal feet in insect segment morphogenesis. .\‘tr/~w 293, 733735. MacRac, T. H., and Langdon, C. M. (19891. Tubulin synthesis, structure, and function: What arc’ the relationships? Ilioc//c~r)r. Co// nio/. 67, 770~790. MacRae, T. H., Landgon, C. M., and Freeman, J. -4. (1991 I. Spatial distribution of post-translationally modified tuhulins in polarized cells of developing A rfc>witr. Cc& Mdil. Cytoskc/. 18, 189-203. Meier, S. (19781. Dewlopment of the embryonic chick otic placode: Light microscopic analysis. Attut. Rrc 191, 44Tm458. Mittenthal. J. E., and Maze, R. M. (1983). A model for shape generation
292
DEVELOPMENTALBIOLOGY
by strain and cell-cell adhesion in the epithelium of an arthropod leg segment. J. Tl,con Rid. 100, 443-483. Morriss-Kay, G., and Tuckett, F. (1987). Fluidity of the neural epithelium during forebrain formation in rat embryos. .I. Cell Sci. Supljl. 8, 433-449. Nardi, J. B. (1981). Induction of invagination in insect epithelium: A paradigm for embryonic invagination. S~ienw 214, 564-566. Nardi, J. B., and Magee-Adams, S. M. (1986). Formation of scale spacing patterns in a moth wing. I. Epithelial feet may mediate cell rearrangement. Del%. Rid. 116,278-290. Odell, G. M., Oster, G., Alberch, P., and Burnside, B. (1981). The mechanical basis of morphogenesis. I. Epithelial folding and invagination. Dee Bid. 85, 446-462. Owaribe, K., Kodama, R., and Eguchi, G. (1981). Demonstration of contractility of circumferential actin bundles and its morphogenetic significance in pigmented epithelium it/ ,sitro and it/ viw ,1 (‘PI/ Rid SO, 507-514. Pictet, R. L., Clark, W. R., Williams, R. H., and Rutter, W. J. (1972). An ultrastructural analysis of the developing embryonic pancreas. Ijut*. Bid. 29, 436-467. Pino-Heiss, S., and Schubiger, G. (1989). Extracellular protease production by Drosophi/u imaginal discs. [jet.. Biol. 132, 282-291. Schoenwolf, G. C., and Alvarez, I. S. (1989). Roles of neuroepithelial
VOLUME158. 1992
cell rearrangement and division in shaping of the avian neural plate. I)c,IlPlopntr)lf 106, 427-439. Schoenwolf, G. C., and Franks, M. V. (1984). Quantitative analyses of changes in cell shapes during bending of the avian neural plate. Dw.
Rid. 10.5,257-272. Scholtz, (:. (1990). The formation, differentiation and segmentation of the post-naupliar germ band of the amphipod C;U~~UI.IIS /,/r/e.r L. (Crustacea, Malacostraca, Pericarida). Ptvc. R S’oc: Lordorc H 239, 163~211. Smuts, M. S., Hilfer, S. R., and Searls, R. 1~.(1978). Patterns of cellular proliferation during thyroid organogenesis. J. E>wtlrr!yol. Exp. ;2for-
/hd. 48, 269-286. Spooner, B. S. (1973). Microlilaments, cell shape changes, and morphogenesis of salivary epithelium. .+ln,. Zoo/. 13, 1007-1022. Weisz, P. IS. (1947). The histological pattern of metamcric development in :1 rtcarr/itr .scdit/tr. J. MwpAol. 81, 45-M. Wessels, N. Ii., Spooner, B. S., Ash, J. F., Bradley, M. O., Luduena, M. A., Taylor, E. L., Wrenn, J. T., andyamada, K. M. (1971). Microfiaments in cellular and developmental processes. Scirtlce 171, 1X-
143. Zwaan, J., and Hendrix, K. LV. (1973). Changes in cell and organ shape during early development of the ocular lens. 4,~. Zoo/. 13, 103910.19.