SEM localization of laminin on the basement membrane of the chick corneal epithelium with immunolatex microspheres

DEVELOPMENTAL

BIOLOGY

106, 83-88 (1984)

SEM Localization of Laminin on the Basement Membrane of the Chick Cornea1 Epithelium with lmmunolatex Microspheres STEPHEN Department

of Zoology, Center

MEIER’

AND CHRISTOPHER

for Developmental

Received April

Biology,

University

5, 1984; accepted in revised form

DRAKE of Texas, Austin,

Texas 78712

June II, 1984

Embryonic chick cornea1 explants were soaked in mild detergent and the anterior cornea1 epithelium was peeled from its basement membrane, leaving the lamina lucida surface exposed and supported on the subjacent primary stroma. Explants were treated with rabbit anti-laminin IgG, followed by sheep anti-rabbit IgG linked microspheres, and processed for SEM. The lucida surface was heavily decorated with microspheres, whereas controls treated with preimmune rabbit IgG were essentially beadless. Laminin distribution was not regular, appearing denser in some regions than others. However, the connective tissue surface of the basement membrane was never laminin-positive, even after treatment with hyaluronidase. These results suggest the basal lamina of the cornea1 epithelium is 6 1984 Academic asymmetric, with preferential location of laminin to the lucida surface of the basement membrane. Press, Inc.

ment membrane of the &day-old cornea1 epithelium was laminin positive. Furthermore, the primary stroma underlying the epithelium serves as support for the basement membrane once the epithelial cells have been removed, facilitating study of the lucida surface. By using an immunolatex probe, we were able to localize laminin sites with the SEM. Moreover, we demonstrate that laminin is located asymmetrically in the basement membrane, in that the lamina lucida side, but not the connective tissue side, of the lamina densa is lamininpositive.

INTRODUCTION

Basement membranes are complex structures containing a variety of collagenous and noncollagenous glycoproteins and proteoglycans. Ultrastructurally, the basement membrane consists of an electron-lucent region, the lamina lucida, directly adjacent to the epithelial cell membrane, and an electron-dense region, the lamina densa, a felt-like mat of varible thickness located between the lamina lucida and the connective tissue. How the various molecular components interact to produce a structured basement membrane remains a matter of controversy. Although there is general agreement that the lamina densa contains most of the type IV collagen (Yaoita et al, 1978; Courtoy et al., 1982; Monaghan et aZ., 1983), glycoproteins such as laminin have been localized in both the lamina lucida and lamina densa (Foidart et ab, 1980; Courtoy et al., 1983; Laurie et aZ., 1982; Inoue et ab, 1983). Laminin, a large noncollagenous glycoprotein identified by Timpl and Rohde (Rohde et ah, 1979; Timpl et al, 1979), is particularly of interest because it is a component that serves in the attachment of epithelial cells to basement membrane collagen (Terranova et al, 1980). How laminin is distributed in the plane of the basement membrane is difficult to appreciate from sectioned material. Therefore, we have devised a way to expose large expanses of both the lamina lucida and connective tissue sides of the basement membrane (lamina densa). We chose to manipulate the cornea1 epithelium because preliminary immunofluorescent studies of sectioned material indicated that the base’ To whom all correspondence

MATERIALS

AND

METHODS

White leghorn chick eggs were incubated for 5 days and tissue explants containing the cornea were removed from the rest of the ocular tissue. To prepare a lucida surface, cornea1 explants were soaked in 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 5 min at room temperature and the anterior cornea1 epithelium was hooked with a needle and pulled as a sheet from the underlying stroma. Shorter periods of detergent treatment caused some of the epithelial cells to remain attached whereas longer detergent treatment cause the epithelium to lose its sheet-like integrity and disintegrate. Other methods for exposing the lucida surface were tried, including treatment of tissues with 10X Ca2+-Mga+-free Hanks’ (to crenate the epithelium), and prolonged (60 min or more) exposure to EDTA (Dodson and Hay, 1971). Although our results were the same with these procedures, neither of the latter techniques was as effective in preparing a clean lucida surface as detergent treatment. Explants containing the detergent exposed lamina lucida were rinsed and

should be sent. 83

OOlZ-1606/84 $3.00 Copyright All rights

0 1984 by Academic Press. Inc. of reproduction in any form reserved.

MEIER AND DRAKE

SEM

Visualizaticm

treated for 30 min with either (a) 140 pg/ml rabbit anti-mouse laminin IgG, (b) 200 pg/ml preimmune rabbit IgG, or (c) PBS. Tissues were rinsed in PBS and fixed in l/2 strength Karnovsky’s (1965) for 30 min. To prepare a connective tissue surface, cornea1 explants were soaked in 0.04% EDTA in Ca2+-Mgz+-free Hanks’ for 5-10 min and rinsed in PBS. The cornea1 epithelium was hooked with a needle and pulled with its basement membrane from the subjacent primary stroma. Epithelia were rinsed in PBS, treated with the primary antibodies, and fixed as for the lucida preparations. Although it is possible to obtain cornea1 epithelia with attached basement membrane by simple dissection, large expanses of intact epithelium were difficult to achieve this way, and so brief EDTA treatment was used to loosen the connection between the lamina densa and the connective tissue of the primary stroma. After fixation, specimens were exposed to immunolatex microspheres (900 A-diameter methacrylate polymers) conjugated to sheep anti-rabbit IgG (Meier and Drake, 1982), prepared by a modification of a procedure reported by Molday et al. (1975). Specimens were rinsed of excess microspheres with PBS, postfixed in osmium, dehydrated through graded alcohols, critical point dried using CO2 as the exchange fluid, and mounted and sputter coated with 6-7 nm of goldpalladium alloy as described previously (Meier and Drake, 1984). For identification of laminin by light microscopy, dissected ocular tissue was fixed in 95% ethanol/glacial acetic acid (99:l) (Icardo and Manasek, 1983). Tissues were dehydrated in 100% ethanol, cleared in two changes of cold xylene, embedded in paraplast, and sectioned. Sections were cleared of paraplast by xylene, rehydrated in cold ethanol, and finally placed in PBS. Sections were exposed to (a) rabbit anti-mouse laminin IgG (140 lg/ml), (b) 200 pug/ml preimmune rabbit serum, or (c) fresh PBS. Samples were rinsed in PBS, exposed to fluorescein for 30 min, rinsed, and mounted for microscopy. Sections were examined with a Zeiss microscope equipped with epifluorescence. In some cases, lucida and densa preparations were treated for 30 min at 37°C with 3500 U/ml testicular

of Laminin

on Basal

Lamina

85

hyaluronidase (Sigma, Type VI-S) (pH 5.6) in PBS prior to exposure to primary antibodies in order to remove large glycosaminoglycans that might obscure laminin sites in the basement membrane. We have previously reported (Meier and Drake, 1934) that hyaluronidase treatment enhances microsphere labeling of fibronectin sites on cranial basement membranes of chick embryos. Rabbit anti-mouse laminin IgG was a generous gift of Dr. Charles Little at the University of Virginia. Briefly, laminin antigen was purified from the EHS mouse sarcoma tumor according to Timpl et al. (1979). Antibodies were raised in rabbits utilizing a subcutaneous injection schedule. Polyclonal rabbit IgG to laminin was affinity purified and antibody specificity to embryonic mouse laminin was confirmed by radioimmune precipitation assay and by Western gel analysis. Furthermore, when the anti-mouse laminin IgG was exposed to [3H]leucine-labeled basement membrane proteins solubilized from chick chorioallantoic membrane, only two bands appeared on radiofluorograms, at MW of approximately 400,000 to 200,000. This strongly suggests that the anti-mouse laminin IgG cross-react with chicken laminin. RESULTS

In order to remove the cornea1 epithelium from its basement membrane, ocular tissue containing 5-dayold chick cornea was soaked briefly in mild detergent. The cornea1 and surrounding surface epithelium (now prospective conjunctival ectoderm) was peeled away, revealing a large expanse of basement membrane left behind on the explant (Fig. 1). Such explants were exposed first to rabbit anti-mouse laminin IgG and then to immunolatex spheres conjugated to sheep antirabbit IgG. Spheres will only attach to sites where the rabbit IgG is bound. As seen in Fig. 2, the lamina lucida side of the basement membrane is smooth and appears as a gently undulating mat over the underlying primary stroma. This surface of the basement membrane is laminin positive as it is highly decorated with microspheres. Microspheres are seen to be attached individually to the lamina. Labeling is uneven in that

FIG. 1. Overview of ocular tissue, with cornea1 and surrounding conjunctival epithelium removed by detergent treatment. Epithelial cells detach from the basement membrane and the cornea1 portion collapses on the corrugated primary stroma, whereas the basement membrane of the adjacent conjunctiva ectoderm (star) rests on periocular mesenchyme. X160. FIG. 2. Lamina lucida surface of cornea1 basement membrane from the region marked “cornea” in Fig. 1, treated with laminin probe. Microspheres are abundant but are not distributed uniformly. Similar heterogeneous patterns were seen on the lucida side of surface ectoderm as well (region of star in Fig. 1). ~10,200. FIG. 3. Control specimen prepared similarly to the one in Fig. 1, except that following detergent treatment, preimmune rabbit serum was used as the primary antibody. Although small bits of contaminants sometimes appear, the lucida surface is beadless. Similarly, specimens exposed only to microspheres were undecorated. ~20,600.

86

MEIER AND DRAKE

SEM Visualization

some areas appear less densely decorated than others, and some patchy clustering and alignment into short rows occurs as well. However, there was no repetitive or rigorous organization to the patterned distribution of microspheres on the lucida side of the basement membrane of either cornea1 or conjunctival epithelium. Examination of the undersurface of the cornea1 epithelium removed from the explant and exposed to laminin probe showed it to be extremely blebbed, but devoid of microspheres and therefore laminin-negative (not shown). The basement membranes of control specimens, treated with preimmune serum and microspheres, or treated with microspheres only, were beadless (Fig. 3). This indicates that little, if any, rabbit IgG sticks nonspecifically to our tissue preparations and that the spheres do not adhere without primary antibody. To obtain a view of the matrix side of the cornea1 basement membrane, ocular explants were soaked briefly in EDTA, a treatment that first loosens the connection between the primary stroma and the cornea1 basement membrane. Epithelia with their basement membrane adherent were peeled back from the primary stroma and treated with our probe to laminin (Fig. 4). SEM examination revealed obvious differences in the thickness of the basement membrane, in that the outlines of cornea1 epithelial cells are visible through the membrane, whereas cells of the conjunctival ectoderm are not as apparent. Regardless of its thickness, we were unable to label the matrix surface of basement membranes with our laminin probe (Fig. 5). Even after hyaluronidase treatment, to remove glycosaminoglycans on the basement membrane that might obscure laminin sites, the matrix side of the basement membrane remained undecorated after exposure to laminin probe. Further evidence strengthening the conclusion that laminin is distributed asymmetrically in the basement membrane comes from the careful examination of some lucida preparations. Often times a portion of the basement membrane is pulled back from the stroma and reflected onto the lucida surface (Fig. 6). Thus, in one picture, it is possible to show that the stromal or matrix side of the basement memnbrane is laminin negative, whereas the lucida side of the basement membrane of the same specimen is laminin positive.

of Luminin

on Basal Alumina

87

DISCUSSION

By exposing large expanses of the lamina lucida and applying our SEM probe, we obtain a cellular view of laminin distribution along the basement membrane to which the epithelium is attached. That little laminin is lost from this surface during tissue separation is supported by the fact that the undersurface of the epithelium removed was laminin negative. Likewise, there was no immunologically detectable laminin in the dissection medium. By using a polyclonal antibody as the first part of our probe, we maximized our opportunity for binding anti-laminin IgG because many parts of the laminin molecule could be recognized. Even though we feel the tissue was treated with excess quantities of rabbit anti-laminin IgG and microspheres, it is likely we are still underestimating the actual number of laminin sites. A sphere with a diameter of 900 A could easily cover several laminin molecules that might be in its immediate vicinity. However, if we assume the patterned distribution of laminin we see to be representative of epithelial attachment, then the epithelium is only attached in spots to the basement membrane. Moreover, that laminin is not blanketed in regular array on the basement membrane in all locations suggests there may be differences in the adhesiveness of basement membranes based on the variable distribution of laminin. However, other epithelia will have to be examined before any meaningful quantitations of adhesiveness can be made. It is not surprising that laminin is located in the 5-day-old cornea1 epithelium. Laminin has been localized to a wide variety of basement membranes and it seems to be a ubiquitous component (Rohde et ah, 1979; Foidart et al., 1980; Laurie et al., 1983). Sugrue and Hay (1982) have shown that g-day-old cornea1 epithelium, stripped of its basement membrane, blebs, and that the addition of laminin to the culture medium restores the smooth appearance of the basal cell surface and stimulates cornea1 matrix production. Furthermore, laminin-coated covaspheres have been shown to bind to the basal cell surface of cornea1 epithelia divested of their basement membrane (Sugrue and Hay, 1983). Our preliminary immunofluorescent studies indicated that the basement membrane of cornea1 epithelium was laminin positive, but even more positive was that

FIG. 4. Overview of ocular epithelium peeled from the subjacent stroma after treatment with EDTA. The basement membrane remains with the epithelium, thus exposing the connective tissue side of the basement membrane. There is a clear transition (arrow) in the appearance of the basement membrane. Basement membrane under the cornea1 epithelium (CE) is noticeably thinner than under the surface ectoderm (SE). X700. FIG. 5. Higher magnification of the connective tissue side of the basement membrane underlying cornea1 epithelium as shown in Fig. 5, exposed to laminin probe. This surface of the basement membrane is undecorated, and therefore is laminin negative. ~21,000. FIG. 6. High magnification of the basement membrane of a detergent treated specimen exposed to laminin probe. A portion of the basement membrane has been pulled from the stroma and turned back onto the lucida surface. Although the lucida side of the cornea1 epithelium is laminin positive, the connective tissue side of the reflected basement membrane (and the stroma) is laminin-negative. X23,750.

88

DEVELOPMENTAL BIOLOGY

of the adjacent surface ectoderm and nearby lens. However, we were never able to identify laminin on the connective tissue surface of the basement membranes of any of these tissues with our immunolatex probe. It has been shown previously that hyaluronidase treatment greatly enhances the labeling of fibronectin on the connective tissue surface of several embryonic chick basal laminae (Meier and Drake, 1984; Harrison et al., 1984), but hyaluronidase treatment did not reveal additional laminin on the basement membrane of the cornea1 epithelium. The asymmetric distribution of laminin in the cornea1 epithelium supports the findings for skin and kidney (Madri et al., 1980; Foidart et ak, 1980; Courtoy et ab, 1982), where laminin is located only in the lamina lucida. However, our results do not contradict the findings of others who locate laminin to the interstices of the lamina densa (Laurie et al,, 1982; Inoue et al., 1983; Amenta et al., 1983), since we cannot see these regions by our technique. However, the fact that laminin is not available on the connective tissue surface of the basement membrane does suggest that stromal fibroblasts, which migrate along the undersurface of the cornea1 epithelium, do not use laminin as an attachment molecule (Couchman et al., 1983). Instead, it seems more likely that the stromal fibroblasts attach to fibronectin deposited on the connective tissue surface of the basement membrane. We thank Dr. Charles Little of the University of Virginia for his kind gift of rabbit anti-mouse laminin IgG, and Betty Galloway for her excellent photographic assistance. This work was supported by NIH Grant DE 05616 and NSF Grant PCM 8203488 to S.M. REFERENCES AMENTA, P. S., CLARK, C. C., and MARTINEZ-HERNANDEZ, A. (1983). Deposition of fibronectin and laminin in the basement membrane of rat parietal yolk sac: Immunohistochemical and biosynthetic studies. J. Cell. Biol 96,104-111. COUCHMAN, J. R., Hijij~, M., REES, D. A., and TIMPL, R. (1983). Adhesion, growth, and matrix production by fibroblasts on laminin substrates. J. Cell. Biol. 96, 177-183. COURTOY, P. J., TIMPL, R., and FARQUHAR, M. G. (1982). Comparative distribution of laminin, type IV collagen, and fibronectin in rat glomerulus. J Histochem. Cytochem. 30, 874-886. DODSON, J. W., and HAY, E. D. (1971). Secretion of collagenous stroma by isolated epithelium grown in vitro. Exp. Eye Res. 65, 215-220. FOIDART, J. M., BERE, E. W., JR., YAAR, M., RENNARD, S. I., GULLINO, M., MARTIN, G. R., and KATZ, S. I. (1980). Distribution and immunoelectron microscopic localization of laminin, a non-collagenous basement membrane protein. Lab. Invest. 42, 336-342.

VOLUME 106, 1984

HARRISSON, F., VANROELEN, C., FOIDART, J.-M., and VAKAET, L. (1983). Expression of different regional patterns of fibronectin immunoreactivity during mesoblast formation in the chick blastoderm. Dev. Biol. 101, 373-381. ICARDO, J. M., and MANASEK, F. J. (1983). Fibronectin distribution during early chick embryo heart development. Den Bid 95, 19-30. INOUE, S., LEBLOND, C. P., and LAURIE, G. W. (1983). Ultrastructure of Reichert’s membrane, a multilayered basement membrane in the parietal wall of the rat yolk sac. J. Cell Biol 97, 1524-1537. KARNOVSKY, M. J. (1965). A formaldehyde-glutaraldehyde fixative of high osmolarity for use in electron microscopy. J. Cell Biol. 27, 137a. LAURIE, G. W., LEBLOND, C. P., and MARTIN, G. R. (1982). Localization of the type IV collagen, laminin, heparan sulfate proteoglycan, and fibronectin to the basal lamina of basement membranes. J. Cell Biol. 95, 340-344. LAURIE, G. W., BEBLOND, C. P., and MARTIN, G. R. (1983). Light microscopic immunolocalization of type IV collagen, laminin, heparan sulfate proteoglycan and fibronectin in basement membranes of a variety of rat organs. Amer. J Anat. 167, 71-82. MADRI, J. A., ROLL, F. J., FURTHMAYR, H., and FOIDART, J. (1980). Ultrastructural localization of fibronectin and laminin in the basement membrane of the murine kidney. J. Cell Biol. 86, 682687. MEIER, S., and DRAKE, C. (1982). Development of a latex-conjugated immunocytological marker for SEM analysis of quail-chick chimeras. J. Exp. Zool. 224, 26-37. MEIER, S., and DRAKE, C. (1984). SEM localization of cell-surfaceassociated fibronectin in the cranium of chick embryos utilizing immunolatex microspheres. J. Embryol. Exp. Morphol. 80,175-195. MOLDAY, R. S., DREYER, W. J., REMBAUM, A., and YEN, S. P. S. (1975). New immunolatex spheres: visual markers of antigens on lymphocytes for scanning electron microscopy. J. Cell Biol. 64, 75-88. MONAGHAN, P., WARBURTON, M. J., PERUSINGHE, N., and RUDLAND, P. S. (1983). Topographical arrangement of basement membrane proteins in lactating rat mammary gland: comparison of distribution of type IV collagen, laminin, fibronectin, and Thy-l at the ultrastructural level. Proc. N&l. Acad Sci. USA 80, 3344-3348. ROHDE, H., WICK, G., and TIMPL, R. (1979). Immunochemical characterization of the basement membrane glycoprotein laminin. Eur. .I Biochem. 102,195-201. SUGRUE, S. P., and HAY, E. D. (1982). Interaction of embryonic cornea1 epithelium with exogenous collagen, laminin, and fibronectin: Role of endogenous protein synthesis. Dev. Biol. 92, 97-106. SUGRUE, S. P., and HAY, E. H. (1982). Identification of extracellular matrix binding sites on the cell surface of isolated embryonic cornea1 epithelia. J. Cell Biol. 97, 319a. TERRANOVA, V. P., ROHRBACH, D. H., and MARTIN, G. R. (1980). Role of laminin in the attachment of PAM 212 (epithelial) cells to basement membrane collagen. Cell 22,719-726. TIMPL, R., ROHDE, H., ROBEY, P. G., RENNARD, S. I., FOIDART, J. M., and MARTIN, G. R. (1979). Laminin-a glycoprotein from basement membranes. J. Biol. Chem. 254,9933-9937. YAOITA, H., FOIDART, J. M., and KATZ, S. J. (1978). Localization of the collagenous component in skin basement membrane. J. Invest. Dermatol 70, 191-193.,