- Email: [email protected]
Ultrastructural studies of attachment site formation in aortic smooth muscle cells cultured on collagen—hydroxyethylmethacrylate hydrogels
JOURNAL OF ULTRASTRUCTURERESEARCH 86, 2 5 2 - 2 6 1
(1984)
Ultrastructural Studies of Attachment Site Formation in Aortic Smooth Muscle Cells Cultured on Collagen-Hydroxyethylmethacrylate Hydrogels PAUL TOSELLI, BARBARA FARIS, PAMELA OLIVER, AND CARL FRANZBLAU
Department of Biochemistry, Boston University School of Medicine, Boston, Massachusetts 02118 Received January 13, 1984 Hydroxyethylmethacrylate (HEMA) hydrogels have a peculiar crater-like topography which renders them ideal for studying cell-to-substrate contact formation. The hydrogel is quite suitable for such studies since it is transparent, which allows for light microscopic observations, it is not toxic, and it will support cell growth only when an additional protein component such as collagen is integrated into its surface. Cultured rabbit aortic smooth muscle ceils (SMC) were grown on collagen-HEMA hydrogels, and the ultrastructure of developing cell attachment sites was studied. By 3 hr after cell seeding, both the rounded and the spreading SMC appeared anchored to the hydrogel via extracellular connective tissue-like material. The fully formed attachment site present at 5-8 days was characterized by large bundles of intraceUular myofilaments inserting onto areas of increased electron density along the plasmalemmal membrane; large amounts of extracellular connective tissue-like material also appeared attached to the areas of increased electron density. Fully formed cell substratum attachment sites were not observed when either elastin-HEMA hydrogels or hydrogels polymerized in the absence of protein were employed.
In vivo, most cell types are located on collagen-containing basal lamina or in collagen matrices. Because of this close proximity, experiments have been designed to probe cell-collagen interaction in a defined culture system. Since the observation of Ehrmann and Gey (1956) that cell growth in vitro is supported by collagen gels, there have been numerous studies reporting an enhancement of cell growth on collagen gels and on collagen- or gelatin-coated surfaces (Kleinman et al., 1980). More recently, we have described the use of protein-hydroxyethylmethacrylate (HEMA) hydrogels as a surface for cell growth (Civerchia-Perez et al., 1980; Fails et aL, 1983; Toselli et aL, 1983). The hydrogel surface is quite suitable for such studies because it is transparent, which allows for microscopic observations, it is not toxic, and it will support cell growth only when an additional protein component such as collagen is integrated into its surface. In this communication, cell-cell and cell-substratum interactions are evaluated when smooth muscle ceils are grown on collagen-HEMA hydrogels. 252 0022-5320/84 $3.00 Copyright © 1984 by Academic Press, Inc. All rights of reproduction in any form reserved.
Toselli et al. (1983) have described sample preparation necessary for improving the preservation of the ultrastructure of cells cultured on protein hydrogels. Thus, the unusual crater-like topography of the hydrogel can be utilized as an experimental aid in studying the ultrastructure of focal points &cell attachment and spreading on surfaces other than those which are uniformly fiat. Good preservation also increases the possibility for directly observing ultrastructural features of the smooth muscle cell (SMC) attachment-site formation, which is the main objective of this paper. MATERIALS AND METHODS
Preparation of HEMA Hydrogels Hydroxyethylmethacrylate (HEMA) hydrogels were prepared as described by Civerchia-Perez et aL (1980). Essentially, this consisted of mixing sequentially 1.0 ml of HEMA, 1.0 ml of ethylene glycol, 1.0 ml of a collagen solution in 0.05 M Tris-0.15 M sodium chloride buffer, pH 7.4 (0.42 mg pepsin-extracted calfskin collagen/ml), 0.1 ml of 6% a m m o n i u m persulfate, and 0.1 ml of 12% sodium metabisulfite. The HEMA hydrogels were then polymerized at 38°C between two glass microscope slides separated by two coverslips.
EM OF ATTACHMENT SITE FORMATION The polymerized hydrogels (which were approximately 0.5 m m thick), were removed from the glass slides and dialyzed exhaustively versus the Tris-sodium chloride buffer to remove residual monomer and ethylene glycol. After dialysis, the HEMA hydrogels were cut into 1.4-cm-diameter buttons with a size-10 cork borer. Each collagen-HEMA hydrogel button contained approximately 14 #g protein. The buttons were sterilized in Puck's Ca2+-Mg2+-free saline G containing 1000 units/ml penicillin, 50 #g/ml aehromycin, and 0.25 #g/ ml Fun#zone (Squibb) by placing them under ultraviolet light for 2 hr. They were subsequently stored at 4°C in the Puck's saline G containing 100 units/ml penicillin and 100 #g/ml streptomycin. Elastin-HEMA hydrogels were prepared in the same manner as the collagen-HEMA hydrogels. A solution of oxalic acid solubilized a-elastin (Partridge et aL, 1955) which contained 0.4 mg elastin/ml was used in these experiments.
Smooth Muscle Cell Cultures Rabbit smooth muscle cells were isolated and grown from the aortic arch of weanling rabbits as described previously (Fails et aL, 1976). Briefly, the aortic segments were cleaned of extraneous material and the medial layer was stripped and minced into 1-mm 2 pieces. These minced tissue pieces were allowed to attach to 75-cm 2 tissue culture flasks by standing the flasks upright in a 5% CO2/95% air humidified incubator at 37"C for 30 min. After tissue attachment, 10 ml medium was added to each flask and the flasks were left undisturbed for 1 week, at which time each flask was replenished with 20 ml medium on Days 7, 10, and 13. Dulbecco's modified Eagle's medium containing 3.7 g sodium bicarbonate per liter, 10% fetal bovine serum, 1 mMnonessential amino acids, 1 m M s o d i u m pyruvate, i00 units/ml penicillin B, and 100 ug/ml streptomycin was used in these experiments. First subcultivation was accomplished by obtaining a single-cell suspension with a 0.05% trypsin/EDTA solution and seeding at a density of 1.5 x 106 cells per 75-cm 2 flask. These cells were maintained for 1 week in 10 ml medium (which was changed twice) and then trypsinized and subeultivated again. Cells at this point (second subcultivation) were seeded onto the eollagen-HEMA hydrogels.
Cell Seeding Into each chamber ofa Costar cluster dish (24 chambers, 1.6 cm diameter) was placed an individual HEMA hydrogel button. Prior to cell plating, the buttons were incubated for 1 hr at 37°C in 1.0 ml of Dulbeeeo's modified Eagle's medium. This medium was then removed, and 0.5 ml of medium was added followed by seeding with 0.5 ml of medium containing 30000 smooth muscle cells in suspension. After incubation for 4 hr at 37"C, the HEMA hydrogel buttons were transferred into new chambers which contained 2.0 ml
253
of medium. In the case of the cells plated onto the chambers without HEMA hydrogels (also referred to as tissue-culture plastic), the medium was removed and 2.0 ml fresh medium was added to each chamber. The dishes were subsequently incubated at 37°C in a humidified atmosphere at 5% CO2/95°/0air and the medium was changed two times weekly. Phase-contrast photographs of the cells growing on the collagen-HEMA hydrogels and on the tissue-culture plastic were taken at specified times.
Electron Microscopy Cell preparation for electron microscopy was performed according to the method described by Toselli et al. (1983). The medium from each chamber was removed, and the cells were fixed for 1 hr with a 2% glutaraldehyde solution in Puck's saline G. The cells were then washed with Puck's saline G, and post-fixed for 1 hr in 1% osmium tetroxide in Puck's saline G. Dehydration was accomplished with the use of a graded series of ethanol solutions prepared with 15% polyethylene glycol (20 000) in Puck's saline G. Added to the samples sequentially for 20-rain periods were 50% ethanol, 70% ethanol (2x), 80% ethanol (2x), 95% ethanol (2 x ), and finally, 100% ethanol (2 x ). A 1:1 Araldire 502 (Ciba)--dodecenyl succinie anhydride epoxy mixture which was degased under vacuum for 10 to 20 min at room temperature was used in the embedment procedure. Infiltration with epoxy was accomplished by first incubating the cells with the epoxy mixture for 16 hr at 38°C. This was followed by two to three changes of epoxy during the next 24 hr at 38°C in a vacuum oven. The final embedding medium, which contained 2% (v/v) benzyldimethylamine catalyst, was degased as described above. The uncatalyzed epoxy was replaced with catalyzed epoxy, and the cells were placed under vacuum at 38°C for 10 to 20 min, followed by two to three 20-min changes of catalyzed epoxy. Most of the epoxy was then poured off so that a thin layer remained in each Costar chamber which was polymeilzed for 24 to 48 hr at 60°C. After cooling, the epoxy layer was cracked from the plastic chamber. Areas for sectioning were removed with a jeweler's saw and mounted on wood dowels with the HEMA hydrogel layer just below the upward surface. The block was mounted in a chuck and the bulk of the HEMA hydrogel was removed with a razor blade. Care was taken not to remove the hydrogel--celljunction. Thin sections were cut parallel to the plane o f the cell culture substratum. Serial sections were also cut through the hydrogel, the hydrogel---cell junction, and finally through the cells which pile up to form layers up to seven cells thick. Occasionally, sections were cut perpendicularly to the growing surface, and these sections were picked up with 300-mesh, Forrnvar-coated grids. All sections were cut on an LKB Ultratome V, stained with uranyl acetate followed by lead citrate (Reynolds, 1963), and examined with a Philips 300 electron microscope.
254
TOSELLI ET AL.
grow as well on collagen-hydroxyethylmethacrylate hydrogels (Fig. 1) as on tissueculture plastic. At the electron microscope level, 3-day cell cultures are recognized to have most o f the ultrastructural features (not shown) o f in vivo adult vascular SMC on both plastic and coUagen-HEMA. The cultured SMC is approximately 100-200 ~m long, 10-15 # m wide, and 1-3/zm thick and has a lobulated nucleus with p r o m i n e n t nucleoli. The p r e d o m i n a n t c o m p o n e n t s o f the SMC are the large bundles o f longitudinally aligned 7.5-nm thin myofilaments with periodically spaced dense bodies. Thick m y o filaments are not apparent. Microtubules, 10-nm filaments, and elongated m i t o c h o n dria are centrally located and aligned with their long axes parallel to the long axis o f the cell. N u m e r o u s free ribosomes, rough FIG. 1. Phase micrograph of rabbit aortic smooth endoplasmic reticulum, and m e m b r a n e muscle ceils cultured for 3 days on collagen-HEMA b o u n d lysosomes are also observed. A b u n hydrogels, x 130. dant plasmalemmal vesicles are present at the cell periphery. S m o o t h muscle cells are connected by d e s m o s o m e s (Toselli et al., Stages of Cell-Substratum Attachment For purposes of discussion, we have divided devel- 1981). Also, cultured SMC m a y contain poopment of smooth muscle cell attachment to collagen- lygonal networks o f thin myofilaments in the peripheral regions o f the cytoplasm (ToHEMA hydrogels in the followingmanner. Stage L Cells having a spherical profile, which, in selli et al., 1981) while the extracellular general, are observed lying at the bottom of the crater- spaces contain connective tissue proteins like hydrogel substratum. Stage II. Elongated or spreading cells attaching to such as elastin, collagen, and glycosamicrater walls. Stages I and II are observed shortly after noglycans (Fads et al., 1976; N a m i k i et al., seeding (3 hr). 1980). Stage III. Cells observed at this stage were obtained Ultrastructural data on the substratum from 3 to 8-day cultures. At this point, smooth muscle topography are essential for evaluating any ceils have a shape characteristic of in vivo adult smooth proposed m o d e l for the formation o f cellmuscle cells. substratum junctions. It is possible to deRESULTS lineate the configuration o f the collagenH E M A hydrogel surface by studying seCell Culture quential sections cut in a plane parallel to Light microscopic studies o f rabbit aortic the substratum (Toselli et al., 1983). I f the SMC cultured for 3 days indicate the cells sections always contain both cells and ....->
FIG. 2. Electron micrograph of rabbit aortic smooth muscle cell cultured for 3 hr on collagen-HEMA hydrogels. Note this portion of the cell is spherical in shape and is completely surrounded with the substratum. Sequential sections of the cell cut in a plane parallel to the plane of the hydrogel indicates the cell lies at the bottom of a hydrogel crater. Also, note extracellular connective tissue-like material (CT) is present at the cellsubstratum junction. Portions of the cell are not contacting the substratum (arrows) and extracellular fibrous material is not apparent. HEMA, eollagen-hydroxyethylmethacrylate;N, nucleus, x 9000.
.i1
p~
cr
4
HEMA
N"
255
2 56
TOSELLI ET AL.
HEMA, the substratum topography is craterlike. Therefore, the electron micrograph (Fig. 2) of the SMC (3-hr culture) showing a portion of the cell completely surrounded by HEMA represents a cell lying at the bottom of a crater. The electron micrograph of the SMC (3-hr culture) and adjacent substratum observed in Figs. 3 and 4 show the cell lying adjacent to the side of a crater. Cell-Substratum Attachment Site Formation Stages I and II. Three hours after seeding, the ultrastructural studies indicate the cells cultured on collagen-HEMA hydrogels vary in shape from ovoid (Fig. 2) to fusiform (Fig. 4). Sparse amounts of extracellular connective tissue-like fibrous material is present along a portion of the cell-substratum junction (Fig. 2, CT). This material is most abundant in areas of cell-substratum contact. Note portions of the cell are not contacting the substratum (Fig. 2, arrows) and extracellular fibrous material is not apparent in these areas. There are sparse amounts of intracellular 7.5-rim thin myo-
filaments observed in cells having a spherical profile, and the filaments are always observed in areas of cell-substratum contact. Figure 4 shows that only a small portion of the fusiform cell is anchored to the substratum (small arrows). These cells have numerous 7.5-nm thin myofilament bundles lying parallel to the long axis of the cell (Fig. 3, large arrow). The fusiform cell has a distinct cell body with filopodia extending from the cell body to focal pads (footpads) of attachment on the collagen-HEMA substratum. Sparse extracellular connective tissuelike fibrous material (Fig. 4, CT) is present at the cell-substratum junction. Stage 111. Electron microscopy of the cellhydrogel substratum junction at 3 days (Fig. 5) and 8 days (Fig. 6) reveals large bundles of longitudinally aligned thin myofilaments inserting onto areas of increased electron density near the plasma membrane. Longitudinally aligned extracellular connective tissue-like material also appears attached to areas of increased electron density near the plasma membrane (Fig. 7). In general, extracellular connective tissue-like material is
F~GS. 3 AND 4. Electron micrographs of rabbit aortic smooth muscle cells cultured for 3 hr on elastin-HEMA (Fig. 3) or collagen-HEMA (Fig. 4). The cells lie adjacent to the sides of craters. Figure 3 shows portions of two ceils forming a small aggregate. SMC cultured up to 18 days on elastin-HEMA show large cell aggregates with necrotic material when examined by electron microscopy (not shown), and extracellular connective tissue-like material is never observed at the cell-substratum junction. Figure 4 shows only a small portion of the cell is anchored to the substratum (small arrows). Sparse extracellular connective tissue-like fibrous material (CT) is present at the cell-substratum junction. Longitudinally aligned bundles of 7.5-nm thin myofilaments (large arrow) are present in the filopodia. HEMA, hydroxyethylmethacrylate; N, nucleus. Fig. 3, × 10 000; Fig. 4, x 5500. Flos. 5 AND 6. Electron micrographs of rabbit aortic smooth muscle cells cultured for 3 days (Fig. 5) and 8 days (Fig. 6) on eollagen-HEMA hydrogels showing longitudinally aligned thin myofilaments inserting onto areas of increased electron density (arrows) near the plasma membrane. Abundant longitudinally aligned extracellular connective tissue-like fibrous material (CT) also appears attached to these areas. This is shown at higher magnification in Fig. 7. HEMA, collagen-hydroxyethylmethacrylate. Fig. 5, x 9000; Fig. 6, x 12 000. FIG. 7. Electron micrograph of rabbit aortic smooth muscle cell cultured for 8 days on collagen-HEMA hydrogels showing a high magnification of the cell-substratum junction cut in a plane exactly parallel to the plane of the hydrogel. The cell lies at the uppermost portion of a crater. This fortuitous section depicts large areas of smooth muscle cell limiting plasma membrane (~) lying parallel in the section plane. Note that the intracellular 7.5-nm thin myofilaments (T) are found converging into areas of increased electron density (arrows) at the periphery of the cell. Extracellular connective tissue-like fibrous material (CT) also converges with areas of increased electron density. Examination of serial sections indicates the myofilament bundle shown here merges with a large bundle of thin myofilaments lying parallel to the long axis of the cell. HEMA, collagen-hydroxyethylmethacrylate, x 20 000.
EM OF ATTACHMENT SITE FORMAT/ON
L. g
257
258
TOSELLI ET AL.
259
EM OF ATTACHMENT SITE FORMATION
HEMA °
v" N
i
C T;~
;
!,S
ii¸~ r,, ::
ri.'
t~
"
~
2
~
¸
~,~ : :i~,~ ,:,~:
• ~ii~i
:
,:
~
:~i~-~:'
ii~
,+~ii:i ~
- •
•
7f~,~r~ ,-
260
TOSELLI ET AL.
most abundant in areas close to the ends of the elongated muscle cells, both at cell-substratum junctions and at cell--cell junctions. Here, the extracellular material most likely forms anchoring points for attaching thin myofilament bundles to the hydrogel or to neighboring cells. DISCUSSION Knox (1981) defined cell attachment as the initial phase of the interaction between a cell and the substratum. This interaction involves the contact and formation of an adhesive bond between the cell and the substratum. Cell attachment results in a spherical cell becoming attached to a substratum; and, once anchored, cells can spread and change from a sphere into a "flattened" form. By 3 hr after seeding of rabbit aortic SMC onto coUagen-HEMA hydrogels, both the "round" SMC (Fig. 2) and the "spreading" SMC (Fig. 4) appear anchored to the hydrogel via extracellular connective tissuelike material. Areas of increased electron density along the cell limiting plasmalemreal membrane are not apparent in Stage I and Stage II material. The fully developed attachment site present in Stage III material is characterized by large bundles of intracellular myofilaments inserting onto areas of increased electron density along the plasmalemmal membrane. In addition, large amounts of extracellular connective tissuelike material also appear attached to the areas of increased electron density. This suggests there may be an increase in electron density with a concomitant increase in myofilaments and connective tissue-like material during attachment site formation, resulting in a fully formed anchoring point for attaching the filaments to the hydrogel. Whereas highly developed attachment sites were observed in bovine aortic endothelial cells cultured for 2 days on coll a g e n - H E M A hydrogels (Toselli et al., 1983), such attachment sites were not apparent in endothelial cells cultured either on elastin-HEMA (Toselli et al., 1983) or on HEMA polymerized in the absence of added
protein (unpublished observations). These results are confirmed by the present study. Due to the close proximity of cells to collagen-containing basal lamina and collagen matrices in vivo, numerous in vitro studies have been described using collagen- or gelatin-coated surfaces (Kleinman et aL, 1981). Also, Civerchia-Perez et al. (1980) have described the use ofprotein-HEMA hydrogels as a surface for cell growth. It was shown that the number of fibroblasts growing on collagen-HEMA hydrogels is influenced by the number of collagen molecules polymerized in the hydrogel. The cells grow only when a protein such as collagen is integrated into the hydrogel surface. These hydrogels provide an ideal system in which to study cell-substrate interactions due to their transparency which allows for microscopic observation, their nontoxicity and their support of cell growth when a protein component is polymerized in the hydrogels. In addition, the pore size(s) of the hydrogel is such that cells do not penetrate into, but remain on the surface of the hydrogel. Also of great importance is the fact that the hydrogel surface is not completely fiat, but has a crater-like topography which aids in the ultrastructural examination of the focal points of cell attachment. Although most of this particular study incorporated only collagen-HEMA hydrogels, it should be emphasized that any protein can be incorporated into the HEMA hydrogel. HEMA hydrogels have been used to evaluate the ability of several complex substrates to support nerve fiber growth and neuronal differentiation (Carbonetto et al., 1982). Not only can one vary the individual protein being incorporated into the growth-supporting matrix, but one can also evaluate the effect of any protein with regard to cell growth and/or protein synthesis. We have shown that protein-HEMA hydrogels could be used to control cell morphology as well as the synthesis ofextracellular matrix components (Faris et al., 1983). Therefore the use of protein-HEMA hydrogels to facilitate the study of cell-surface interactions
EM OF ATTACHMENT SITE FORMATION a n d cell a t t a c h m e n t , as well as the m a n i p u l a t i o n o f the c y t o s k e l e t a l s t r u c t u r e o f cells i n c u l t u r e a d d s a n o t h e r d i m e n s i o n to the i n v i t r o a p p r o a c h for s t u d y i n g cells. We thank Professor A. Gedeon Matoltsy from the Department of Dermatology, Boston University School of Medicine, for his generosity with equipment and laboratory space. We also thank Valerie Verbitzki and Rosemarie Moscaritolo for their help in culturingcells. This study was supported by United States Public Health Service Grants HL19717 and HL13262.
REFERENCES CARBONETTO,S. T., GRUVER,M. M., ANDTURNER,D. C. (1982) Science 216, 897-899. CIVERCHIA-PEREZ,L., EARLS,B., LAPOINTE, B., BELDEKAS,J., LEIBOWITZ,H., ANDFRANZBLAU,C. (1980) Proc. NatL Acad. ScL USA 77, 2064-2068. EHRMANN, R. L., AND GEY, G. O. (1981) J. NatL Cancer Inst. 16, 1375-1402.
261
FARIS, B., SALCEDO, L. L., COOK, V., JOHNSON, L., FOSTER,J. A., ANDFRANZBLAU,C. (1976) Biochim. Biophys. Acta 418, 93-103. FARIS, B., MOZZICATO, P., MOGAYZEL, P. J., JR., FERRERA,R., GERSTENFELD,L. C., GLEMBOURTT,M., MAKARSVd,J. S., JR., HAUDENSCHILD,C. C., AND FRANZBLAU,C. (1983) Exp. CellRes. 143, 15-25. KNox, P. (1981) in KNox, P. (Ed.), Biochemistry of Cellular Regulation, Vol. IV, pp. 122-149, CRC Press, Boca Raton, Fla. KLEINMAN, H. K., KLEBE, R. J., AND MARTIN, G. R. (1981) J. CellBiol. 88, 473-485. NAMIKJ, O., FARIS,B., TSCHOPP,F., FUGLISTALLER,P., HOLLANDER, W., FRANZBLAU,C., AND SCHMID, K. (1980) Biochemistry 19, 1900-1904. PARTRIDGE, S. M., DAVIS, H. F., AND ADAIR, G. S. (1955) Biochem. J. 61, 11-20. REYNOLDS, E. S. (1963) J. CellBiol. 17, 208-212. TOSELLI, P., FARIS, B., OLIVER, P., WEDEL, N., AND FRANZBLAU,C. (1983) J. Ultrastruct. Res. 83, 220231. TOSELLI, P., OLIVER, P., COX, M., FARIS, B., AND FRANZBLAU,C. (1981) CellMotil. 1, 193-203.
(1984)
Ultrastructural Studies of Attachment Site Formation in Aortic Smooth Muscle Cells Cultured on Collagen-Hydroxyethylmethacrylate Hydrogels PAUL TOSELLI, BARBARA FARIS, PAMELA OLIVER, AND CARL FRANZBLAU
Department of Biochemistry, Boston University School of Medicine, Boston, Massachusetts 02118 Received January 13, 1984 Hydroxyethylmethacrylate (HEMA) hydrogels have a peculiar crater-like topography which renders them ideal for studying cell-to-substrate contact formation. The hydrogel is quite suitable for such studies since it is transparent, which allows for light microscopic observations, it is not toxic, and it will support cell growth only when an additional protein component such as collagen is integrated into its surface. Cultured rabbit aortic smooth muscle ceils (SMC) were grown on collagen-HEMA hydrogels, and the ultrastructure of developing cell attachment sites was studied. By 3 hr after cell seeding, both the rounded and the spreading SMC appeared anchored to the hydrogel via extracellular connective tissue-like material. The fully formed attachment site present at 5-8 days was characterized by large bundles of intraceUular myofilaments inserting onto areas of increased electron density along the plasmalemmal membrane; large amounts of extracellular connective tissue-like material also appeared attached to the areas of increased electron density. Fully formed cell substratum attachment sites were not observed when either elastin-HEMA hydrogels or hydrogels polymerized in the absence of protein were employed.
In vivo, most cell types are located on collagen-containing basal lamina or in collagen matrices. Because of this close proximity, experiments have been designed to probe cell-collagen interaction in a defined culture system. Since the observation of Ehrmann and Gey (1956) that cell growth in vitro is supported by collagen gels, there have been numerous studies reporting an enhancement of cell growth on collagen gels and on collagen- or gelatin-coated surfaces (Kleinman et al., 1980). More recently, we have described the use of protein-hydroxyethylmethacrylate (HEMA) hydrogels as a surface for cell growth (Civerchia-Perez et al., 1980; Fails et aL, 1983; Toselli et aL, 1983). The hydrogel surface is quite suitable for such studies because it is transparent, which allows for microscopic observations, it is not toxic, and it will support cell growth only when an additional protein component such as collagen is integrated into its surface. In this communication, cell-cell and cell-substratum interactions are evaluated when smooth muscle ceils are grown on collagen-HEMA hydrogels. 252 0022-5320/84 $3.00 Copyright © 1984 by Academic Press, Inc. All rights of reproduction in any form reserved.
Toselli et al. (1983) have described sample preparation necessary for improving the preservation of the ultrastructure of cells cultured on protein hydrogels. Thus, the unusual crater-like topography of the hydrogel can be utilized as an experimental aid in studying the ultrastructure of focal points &cell attachment and spreading on surfaces other than those which are uniformly fiat. Good preservation also increases the possibility for directly observing ultrastructural features of the smooth muscle cell (SMC) attachment-site formation, which is the main objective of this paper. MATERIALS AND METHODS
Preparation of HEMA Hydrogels Hydroxyethylmethacrylate (HEMA) hydrogels were prepared as described by Civerchia-Perez et aL (1980). Essentially, this consisted of mixing sequentially 1.0 ml of HEMA, 1.0 ml of ethylene glycol, 1.0 ml of a collagen solution in 0.05 M Tris-0.15 M sodium chloride buffer, pH 7.4 (0.42 mg pepsin-extracted calfskin collagen/ml), 0.1 ml of 6% a m m o n i u m persulfate, and 0.1 ml of 12% sodium metabisulfite. The HEMA hydrogels were then polymerized at 38°C between two glass microscope slides separated by two coverslips.
EM OF ATTACHMENT SITE FORMATION The polymerized hydrogels (which were approximately 0.5 m m thick), were removed from the glass slides and dialyzed exhaustively versus the Tris-sodium chloride buffer to remove residual monomer and ethylene glycol. After dialysis, the HEMA hydrogels were cut into 1.4-cm-diameter buttons with a size-10 cork borer. Each collagen-HEMA hydrogel button contained approximately 14 #g protein. The buttons were sterilized in Puck's Ca2+-Mg2+-free saline G containing 1000 units/ml penicillin, 50 #g/ml aehromycin, and 0.25 #g/ ml Fun#zone (Squibb) by placing them under ultraviolet light for 2 hr. They were subsequently stored at 4°C in the Puck's saline G containing 100 units/ml penicillin and 100 #g/ml streptomycin. Elastin-HEMA hydrogels were prepared in the same manner as the collagen-HEMA hydrogels. A solution of oxalic acid solubilized a-elastin (Partridge et aL, 1955) which contained 0.4 mg elastin/ml was used in these experiments.
Smooth Muscle Cell Cultures Rabbit smooth muscle cells were isolated and grown from the aortic arch of weanling rabbits as described previously (Fails et aL, 1976). Briefly, the aortic segments were cleaned of extraneous material and the medial layer was stripped and minced into 1-mm 2 pieces. These minced tissue pieces were allowed to attach to 75-cm 2 tissue culture flasks by standing the flasks upright in a 5% CO2/95% air humidified incubator at 37"C for 30 min. After tissue attachment, 10 ml medium was added to each flask and the flasks were left undisturbed for 1 week, at which time each flask was replenished with 20 ml medium on Days 7, 10, and 13. Dulbecco's modified Eagle's medium containing 3.7 g sodium bicarbonate per liter, 10% fetal bovine serum, 1 mMnonessential amino acids, 1 m M s o d i u m pyruvate, i00 units/ml penicillin B, and 100 ug/ml streptomycin was used in these experiments. First subcultivation was accomplished by obtaining a single-cell suspension with a 0.05% trypsin/EDTA solution and seeding at a density of 1.5 x 106 cells per 75-cm 2 flask. These cells were maintained for 1 week in 10 ml medium (which was changed twice) and then trypsinized and subeultivated again. Cells at this point (second subcultivation) were seeded onto the eollagen-HEMA hydrogels.
Cell Seeding Into each chamber ofa Costar cluster dish (24 chambers, 1.6 cm diameter) was placed an individual HEMA hydrogel button. Prior to cell plating, the buttons were incubated for 1 hr at 37°C in 1.0 ml of Dulbeeeo's modified Eagle's medium. This medium was then removed, and 0.5 ml of medium was added followed by seeding with 0.5 ml of medium containing 30000 smooth muscle cells in suspension. After incubation for 4 hr at 37"C, the HEMA hydrogel buttons were transferred into new chambers which contained 2.0 ml
253
of medium. In the case of the cells plated onto the chambers without HEMA hydrogels (also referred to as tissue-culture plastic), the medium was removed and 2.0 ml fresh medium was added to each chamber. The dishes were subsequently incubated at 37°C in a humidified atmosphere at 5% CO2/95°/0air and the medium was changed two times weekly. Phase-contrast photographs of the cells growing on the collagen-HEMA hydrogels and on the tissue-culture plastic were taken at specified times.
Electron Microscopy Cell preparation for electron microscopy was performed according to the method described by Toselli et al. (1983). The medium from each chamber was removed, and the cells were fixed for 1 hr with a 2% glutaraldehyde solution in Puck's saline G. The cells were then washed with Puck's saline G, and post-fixed for 1 hr in 1% osmium tetroxide in Puck's saline G. Dehydration was accomplished with the use of a graded series of ethanol solutions prepared with 15% polyethylene glycol (20 000) in Puck's saline G. Added to the samples sequentially for 20-rain periods were 50% ethanol, 70% ethanol (2x), 80% ethanol (2x), 95% ethanol (2 x ), and finally, 100% ethanol (2 x ). A 1:1 Araldire 502 (Ciba)--dodecenyl succinie anhydride epoxy mixture which was degased under vacuum for 10 to 20 min at room temperature was used in the embedment procedure. Infiltration with epoxy was accomplished by first incubating the cells with the epoxy mixture for 16 hr at 38°C. This was followed by two to three changes of epoxy during the next 24 hr at 38°C in a vacuum oven. The final embedding medium, which contained 2% (v/v) benzyldimethylamine catalyst, was degased as described above. The uncatalyzed epoxy was replaced with catalyzed epoxy, and the cells were placed under vacuum at 38°C for 10 to 20 min, followed by two to three 20-min changes of catalyzed epoxy. Most of the epoxy was then poured off so that a thin layer remained in each Costar chamber which was polymeilzed for 24 to 48 hr at 60°C. After cooling, the epoxy layer was cracked from the plastic chamber. Areas for sectioning were removed with a jeweler's saw and mounted on wood dowels with the HEMA hydrogel layer just below the upward surface. The block was mounted in a chuck and the bulk of the HEMA hydrogel was removed with a razor blade. Care was taken not to remove the hydrogel--celljunction. Thin sections were cut parallel to the plane o f the cell culture substratum. Serial sections were also cut through the hydrogel, the hydrogel---cell junction, and finally through the cells which pile up to form layers up to seven cells thick. Occasionally, sections were cut perpendicularly to the growing surface, and these sections were picked up with 300-mesh, Forrnvar-coated grids. All sections were cut on an LKB Ultratome V, stained with uranyl acetate followed by lead citrate (Reynolds, 1963), and examined with a Philips 300 electron microscope.
254
TOSELLI ET AL.
grow as well on collagen-hydroxyethylmethacrylate hydrogels (Fig. 1) as on tissueculture plastic. At the electron microscope level, 3-day cell cultures are recognized to have most o f the ultrastructural features (not shown) o f in vivo adult vascular SMC on both plastic and coUagen-HEMA. The cultured SMC is approximately 100-200 ~m long, 10-15 # m wide, and 1-3/zm thick and has a lobulated nucleus with p r o m i n e n t nucleoli. The p r e d o m i n a n t c o m p o n e n t s o f the SMC are the large bundles o f longitudinally aligned 7.5-nm thin myofilaments with periodically spaced dense bodies. Thick m y o filaments are not apparent. Microtubules, 10-nm filaments, and elongated m i t o c h o n dria are centrally located and aligned with their long axes parallel to the long axis o f the cell. N u m e r o u s free ribosomes, rough FIG. 1. Phase micrograph of rabbit aortic smooth endoplasmic reticulum, and m e m b r a n e muscle ceils cultured for 3 days on collagen-HEMA b o u n d lysosomes are also observed. A b u n hydrogels, x 130. dant plasmalemmal vesicles are present at the cell periphery. S m o o t h muscle cells are connected by d e s m o s o m e s (Toselli et al., Stages of Cell-Substratum Attachment For purposes of discussion, we have divided devel- 1981). Also, cultured SMC m a y contain poopment of smooth muscle cell attachment to collagen- lygonal networks o f thin myofilaments in the peripheral regions o f the cytoplasm (ToHEMA hydrogels in the followingmanner. Stage L Cells having a spherical profile, which, in selli et al., 1981) while the extracellular general, are observed lying at the bottom of the crater- spaces contain connective tissue proteins like hydrogel substratum. Stage II. Elongated or spreading cells attaching to such as elastin, collagen, and glycosamicrater walls. Stages I and II are observed shortly after noglycans (Fads et al., 1976; N a m i k i et al., seeding (3 hr). 1980). Stage III. Cells observed at this stage were obtained Ultrastructural data on the substratum from 3 to 8-day cultures. At this point, smooth muscle topography are essential for evaluating any ceils have a shape characteristic of in vivo adult smooth proposed m o d e l for the formation o f cellmuscle cells. substratum junctions. It is possible to deRESULTS lineate the configuration o f the collagenH E M A hydrogel surface by studying seCell Culture quential sections cut in a plane parallel to Light microscopic studies o f rabbit aortic the substratum (Toselli et al., 1983). I f the SMC cultured for 3 days indicate the cells sections always contain both cells and ....->
FIG. 2. Electron micrograph of rabbit aortic smooth muscle cell cultured for 3 hr on collagen-HEMA hydrogels. Note this portion of the cell is spherical in shape and is completely surrounded with the substratum. Sequential sections of the cell cut in a plane parallel to the plane of the hydrogel indicates the cell lies at the bottom of a hydrogel crater. Also, note extracellular connective tissue-like material (CT) is present at the cellsubstratum junction. Portions of the cell are not contacting the substratum (arrows) and extracellular fibrous material is not apparent. HEMA, eollagen-hydroxyethylmethacrylate;N, nucleus, x 9000.
.i1
p~
cr
4
HEMA
N"
255
2 56
TOSELLI ET AL.
HEMA, the substratum topography is craterlike. Therefore, the electron micrograph (Fig. 2) of the SMC (3-hr culture) showing a portion of the cell completely surrounded by HEMA represents a cell lying at the bottom of a crater. The electron micrograph of the SMC (3-hr culture) and adjacent substratum observed in Figs. 3 and 4 show the cell lying adjacent to the side of a crater. Cell-Substratum Attachment Site Formation Stages I and II. Three hours after seeding, the ultrastructural studies indicate the cells cultured on collagen-HEMA hydrogels vary in shape from ovoid (Fig. 2) to fusiform (Fig. 4). Sparse amounts of extracellular connective tissue-like fibrous material is present along a portion of the cell-substratum junction (Fig. 2, CT). This material is most abundant in areas of cell-substratum contact. Note portions of the cell are not contacting the substratum (Fig. 2, arrows) and extracellular fibrous material is not apparent in these areas. There are sparse amounts of intracellular 7.5-rim thin myo-
filaments observed in cells having a spherical profile, and the filaments are always observed in areas of cell-substratum contact. Figure 4 shows that only a small portion of the fusiform cell is anchored to the substratum (small arrows). These cells have numerous 7.5-nm thin myofilament bundles lying parallel to the long axis of the cell (Fig. 3, large arrow). The fusiform cell has a distinct cell body with filopodia extending from the cell body to focal pads (footpads) of attachment on the collagen-HEMA substratum. Sparse extracellular connective tissuelike fibrous material (Fig. 4, CT) is present at the cell-substratum junction. Stage 111. Electron microscopy of the cellhydrogel substratum junction at 3 days (Fig. 5) and 8 days (Fig. 6) reveals large bundles of longitudinally aligned thin myofilaments inserting onto areas of increased electron density near the plasma membrane. Longitudinally aligned extracellular connective tissue-like material also appears attached to areas of increased electron density near the plasma membrane (Fig. 7). In general, extracellular connective tissue-like material is
F~GS. 3 AND 4. Electron micrographs of rabbit aortic smooth muscle cells cultured for 3 hr on elastin-HEMA (Fig. 3) or collagen-HEMA (Fig. 4). The cells lie adjacent to the sides of craters. Figure 3 shows portions of two ceils forming a small aggregate. SMC cultured up to 18 days on elastin-HEMA show large cell aggregates with necrotic material when examined by electron microscopy (not shown), and extracellular connective tissue-like material is never observed at the cell-substratum junction. Figure 4 shows only a small portion of the cell is anchored to the substratum (small arrows). Sparse extracellular connective tissue-like fibrous material (CT) is present at the cell-substratum junction. Longitudinally aligned bundles of 7.5-nm thin myofilaments (large arrow) are present in the filopodia. HEMA, hydroxyethylmethacrylate; N, nucleus. Fig. 3, × 10 000; Fig. 4, x 5500. Flos. 5 AND 6. Electron micrographs of rabbit aortic smooth muscle cells cultured for 3 days (Fig. 5) and 8 days (Fig. 6) on eollagen-HEMA hydrogels showing longitudinally aligned thin myofilaments inserting onto areas of increased electron density (arrows) near the plasma membrane. Abundant longitudinally aligned extracellular connective tissue-like fibrous material (CT) also appears attached to these areas. This is shown at higher magnification in Fig. 7. HEMA, collagen-hydroxyethylmethacrylate. Fig. 5, x 9000; Fig. 6, x 12 000. FIG. 7. Electron micrograph of rabbit aortic smooth muscle cell cultured for 8 days on collagen-HEMA hydrogels showing a high magnification of the cell-substratum junction cut in a plane exactly parallel to the plane of the hydrogel. The cell lies at the uppermost portion of a crater. This fortuitous section depicts large areas of smooth muscle cell limiting plasma membrane (~) lying parallel in the section plane. Note that the intracellular 7.5-nm thin myofilaments (T) are found converging into areas of increased electron density (arrows) at the periphery of the cell. Extracellular connective tissue-like fibrous material (CT) also converges with areas of increased electron density. Examination of serial sections indicates the myofilament bundle shown here merges with a large bundle of thin myofilaments lying parallel to the long axis of the cell. HEMA, collagen-hydroxyethylmethacrylate, x 20 000.
EM OF ATTACHMENT SITE FORMAT/ON
L. g
257
258
TOSELLI ET AL.
259
EM OF ATTACHMENT SITE FORMATION
HEMA °
v" N
i
C T;~
;
!,S
ii¸~ r,, ::
ri.'
t~
"
~
2
~
¸
~,~ : :i~,~ ,:,~:
• ~ii~i
:
,:
~
:~i~-~:'
ii~
,+~ii:i ~
- •
•
7f~,~r~ ,-
260
TOSELLI ET AL.
most abundant in areas close to the ends of the elongated muscle cells, both at cell-substratum junctions and at cell--cell junctions. Here, the extracellular material most likely forms anchoring points for attaching thin myofilament bundles to the hydrogel or to neighboring cells. DISCUSSION Knox (1981) defined cell attachment as the initial phase of the interaction between a cell and the substratum. This interaction involves the contact and formation of an adhesive bond between the cell and the substratum. Cell attachment results in a spherical cell becoming attached to a substratum; and, once anchored, cells can spread and change from a sphere into a "flattened" form. By 3 hr after seeding of rabbit aortic SMC onto coUagen-HEMA hydrogels, both the "round" SMC (Fig. 2) and the "spreading" SMC (Fig. 4) appear anchored to the hydrogel via extracellular connective tissuelike material. Areas of increased electron density along the cell limiting plasmalemreal membrane are not apparent in Stage I and Stage II material. The fully developed attachment site present in Stage III material is characterized by large bundles of intracellular myofilaments inserting onto areas of increased electron density along the plasmalemmal membrane. In addition, large amounts of extracellular connective tissuelike material also appear attached to the areas of increased electron density. This suggests there may be an increase in electron density with a concomitant increase in myofilaments and connective tissue-like material during attachment site formation, resulting in a fully formed anchoring point for attaching the filaments to the hydrogel. Whereas highly developed attachment sites were observed in bovine aortic endothelial cells cultured for 2 days on coll a g e n - H E M A hydrogels (Toselli et al., 1983), such attachment sites were not apparent in endothelial cells cultured either on elastin-HEMA (Toselli et al., 1983) or on HEMA polymerized in the absence of added
protein (unpublished observations). These results are confirmed by the present study. Due to the close proximity of cells to collagen-containing basal lamina and collagen matrices in vivo, numerous in vitro studies have been described using collagen- or gelatin-coated surfaces (Kleinman et aL, 1981). Also, Civerchia-Perez et al. (1980) have described the use ofprotein-HEMA hydrogels as a surface for cell growth. It was shown that the number of fibroblasts growing on collagen-HEMA hydrogels is influenced by the number of collagen molecules polymerized in the hydrogel. The cells grow only when a protein such as collagen is integrated into the hydrogel surface. These hydrogels provide an ideal system in which to study cell-substrate interactions due to their transparency which allows for microscopic observation, their nontoxicity and their support of cell growth when a protein component is polymerized in the hydrogels. In addition, the pore size(s) of the hydrogel is such that cells do not penetrate into, but remain on the surface of the hydrogel. Also of great importance is the fact that the hydrogel surface is not completely fiat, but has a crater-like topography which aids in the ultrastructural examination of the focal points of cell attachment. Although most of this particular study incorporated only collagen-HEMA hydrogels, it should be emphasized that any protein can be incorporated into the HEMA hydrogel. HEMA hydrogels have been used to evaluate the ability of several complex substrates to support nerve fiber growth and neuronal differentiation (Carbonetto et al., 1982). Not only can one vary the individual protein being incorporated into the growth-supporting matrix, but one can also evaluate the effect of any protein with regard to cell growth and/or protein synthesis. We have shown that protein-HEMA hydrogels could be used to control cell morphology as well as the synthesis ofextracellular matrix components (Faris et al., 1983). Therefore the use of protein-HEMA hydrogels to facilitate the study of cell-surface interactions
EM OF ATTACHMENT SITE FORMATION a n d cell a t t a c h m e n t , as well as the m a n i p u l a t i o n o f the c y t o s k e l e t a l s t r u c t u r e o f cells i n c u l t u r e a d d s a n o t h e r d i m e n s i o n to the i n v i t r o a p p r o a c h for s t u d y i n g cells. We thank Professor A. Gedeon Matoltsy from the Department of Dermatology, Boston University School of Medicine, for his generosity with equipment and laboratory space. We also thank Valerie Verbitzki and Rosemarie Moscaritolo for their help in culturingcells. This study was supported by United States Public Health Service Grants HL19717 and HL13262.
REFERENCES CARBONETTO,S. T., GRUVER,M. M., ANDTURNER,D. C. (1982) Science 216, 897-899. CIVERCHIA-PEREZ,L., EARLS,B., LAPOINTE, B., BELDEKAS,J., LEIBOWITZ,H., ANDFRANZBLAU,C. (1980) Proc. NatL Acad. ScL USA 77, 2064-2068. EHRMANN, R. L., AND GEY, G. O. (1981) J. NatL Cancer Inst. 16, 1375-1402.
261
FARIS, B., SALCEDO, L. L., COOK, V., JOHNSON, L., FOSTER,J. A., ANDFRANZBLAU,C. (1976) Biochim. Biophys. Acta 418, 93-103. FARIS, B., MOZZICATO, P., MOGAYZEL, P. J., JR., FERRERA,R., GERSTENFELD,L. C., GLEMBOURTT,M., MAKARSVd,J. S., JR., HAUDENSCHILD,C. C., AND FRANZBLAU,C. (1983) Exp. CellRes. 143, 15-25. KNox, P. (1981) in KNox, P. (Ed.), Biochemistry of Cellular Regulation, Vol. IV, pp. 122-149, CRC Press, Boca Raton, Fla. KLEINMAN, H. K., KLEBE, R. J., AND MARTIN, G. R. (1981) J. CellBiol. 88, 473-485. NAMIKJ, O., FARIS,B., TSCHOPP,F., FUGLISTALLER,P., HOLLANDER, W., FRANZBLAU,C., AND SCHMID, K. (1980) Biochemistry 19, 1900-1904. PARTRIDGE, S. M., DAVIS, H. F., AND ADAIR, G. S. (1955) Biochem. J. 61, 11-20. REYNOLDS, E. S. (1963) J. CellBiol. 17, 208-212. TOSELLI, P., FARIS, B., OLIVER, P., WEDEL, N., AND FRANZBLAU,C. (1983) J. Ultrastruct. Res. 83, 220231. TOSELLI, P., OLIVER, P., COX, M., FARIS, B., AND FRANZBLAU,C. (1981) CellMotil. 1, 193-203.