sign in sign up
<<
Home , Cytokeratin

Investigative Ophthalmology & Visual Science. Vol. 31. No. 9. September 1990 Cops right £• Association lor Research in Vision and Ophthalmology

In Situ Localization of Cytoskeletal Elements in the Human Trabecular Meshwork and

Robert N. Weinreb* and Mark I. Ryderf

The authors compared cytoskeletal elements of the in situ human trabccular-mcshwork with in situ human corneal cells using indirect immunofluorcsccncc staining for and intermediate filaments (, , and ) and NBD-phallacidin staining for f- using both fixed frozen and unfixed frozen sections from postmortem eyes. Both f-actin and tubulin were found throughout the cell body of trabecular-meshwork cells, keratocytes, corneal endothelium, and corneal . The f-actin staining pattern was concentrated at the cell periphery of these four cell types. Vimentin stain was intensely localized in focal areas of the trabecular-meshwork cell, keratocytes, and throughout the corneal cndothelium. A general anticytokeratin antibody was intensely localized in corneal epithelium and endothelium. However, PKK-1 anticytokeratin antibody was seen only in superficial layers of corneal epithelium and not in corneal endothelium. The 4.62 anticytokeratin antibody was not observed in either corneal epithelium or endothelium. None of these three cytokera- tin antibodies were seen in trabccular-mcshwork cells or keratocytes. Desmin stain was not noted in any of these cell types. In general, cytoskeletal staining of unfixed frozen sections showed a similar staining pattern for f-actin and tubulin but a more uniform and intense staining pattern for vimentin and cytokcratin compared with fixed frozen material. The authors conclude that these cytoskclctal stains can differentiate human Irabeciilar-meshwork cells from cells of the cornea in situ. Invest Ophthalmol Vis Sci 31:1839-1847, 1990

Human trabecular-meshwork cells perform many subfragment used to label f-actin)-labeled, of the activities which have been hypothesized to con- critical-point dried cells. Similar techniques have tribute to the normal function of the trabecular been used to localize the major cytoskeletal elements meshwork.1"3 They maintain a flat appearance, re- in cultured bovine trabecular-meshwork cells.15 main attached to trabecular-meshwork beams, and However, the overall cell shape and cytoskeletal orga- possess the ability to spread to cover beams that be- nization of any cell is dependent on its external envi- come denuded. In addition, these cells have phago- ronment. Thus, the nature and organization of cyto- cytic abilities and can regulate the deposition and skeletal elements of cells propagated in serial culture degradation of the in the trabecu- may be different than the same cells in situ. lar meshwork.2-4"" As in other cells, these structural In the current study, we used fixed frozen and un- and motile functions arc effected through the three- fixed frozen tissue to obtain a comprehensive over- dimensional network of actin filaments (microfila- view of the in situ localization of actin filaments, ments). , and intermediate filaments tubulin in microtubules. and three different interme- known collectively as the . diate filament (vimentin. cytokeratin. and Recently, we examined the in vitro organization of desmin) in human trabecular-meshwork cells using actin . microtubules. and vimentin fil- fluorescent antibody or NBD-phallacidin labeling (a aments in cynomolgus monkey and human trabecu- fluoresccin-like stain specific for f-actin) of these ele- lar-meshwork cells12"14 with both fluorescent labeling ments. In addition, we compared the in situ localiza- of these cytoskeletal elements and with transmission tion of cytoskeletal elements of trabecular-meshwork electron-microscopic observations of extracted. S-1 (a cells with other cell types from the adjacent cornea. These studies show striking differences in the in situ cytoskeletal labeling patterns among these cell types. From the "Department of Ophthalmology. University of Califor- nia, San Diego, and the ["Department of Oral Biology. University of Materials and Methods California. San Francisco. California. Supported in part by NIH grant EY05990 (R.N.W.). Reprint requests: Robert N. Weinrcb. MD. University of Califor- Postmortem eyes were obtained by enucleation nia. San Diego. Department of Ophthalmology (T-014), La Jolla, within 2 hr after death. A corneoscleral button was CA 92093. excised 2-mm posterior to the limbus to include the

1839

Downloaded from iovs.arvojournals.org on 10/02/2021 1840 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Seprember 1990 Vol. 31

trabecular meshwork. In preliminary experiments, antialpha tubulin (Amersham, Arlington Heights, we found that direct fixation of the corneoscleral IL), (2) mouse monoclonal antivimentin (Amer- buttons resulted in a marked autofluorescence of the sham). (3) mouse monoclonal antidesmin (Amer- sectioned tissue. This autofluorescence was elimi- sham). (4) a general mouse monoclonal anticytoker- nated by storing the buttons in McCarey-Kaufman atin (Amersham) with unknown specific reactivity to media for 6-24 hr before fixation and freezing or particular cytokeratin species. (5) mouse monoclonal before freezing alone. Nine nonglaucomatous eyes PKK-1 anticytokeratin antibody (Labsystems. Hel- from nine patients (aged 23-78 yr) were processed for sinki. Finland) which crossreacts with the 44-kilodal- cytoskeletal labeling. ton (kD). 46-kD, 52-kD, and 54-kD of In the fixation and freezing procedure, the corneo- HcLa cells, and (6) mouse monoclonal anticytokera- scleral tissue from eight eyes were fixed in 2.0% para- tin 4.62 (Miles Laboratories. Naperville, IL) which formaldehyde in 0.1 M phosphate-buffered saline reacts with the 40-kD (no. 19) cytokeratin seen in 16 (PBS) with 10 mM of sodium azide at 4°C for 4-6 hr. differentiated simple epithelium. Each monoclonal washed three times with PBS. and stored in PBS with antibody was diluted in PBS with 0.1% Triton X-100. sodium azide for 3-14 days. A segment of tissue con- 1.0% bovine scrum albumin, and 10 mM sodium taining cornea, sclera, and trabecular meshwork was azide at the following concentrations: antialpha tu- excised from each corneoscleral button, mounted in bulin, 1:50: antivimentin, general anticytokeratin, OCT compound (Miles Laboratories, Naperville. IL) and antidesmin. 1:7; and PKK-1 and 4.62 anticyto- and frozen with dry ice and liquid nitrogen. In the . 1:20. freezing-alone procedure, the corneosclcral tissues All primary incubations were done for 60 min at from five eyes were directly mounted in OCT com- 20°C. The slides were then washed three times with pound and frozen without prior fixation. The frozen PBS with sodium azidc and incubated for 40 min at blocks of tissue were then transferred to a Slec HR 20°C with 50 ml of a 1:50 dilution of rhodamine- cryostat (London, England). For each piece of tissue. conjugated goat anti-mouse antibody (Cappel, Mal- 5-nm thick sections were cut and then collected on vern. PA) in the same buffer as used for the primary Chrome-Alum-treated slides (Becton-Dickinson. monoclonal incubation. Sunnyvale, CA). They were allowed to air dry for 1-2 Control incubations included incubating several hr. One slide from each tissue block was stained with sections with the primary monoclonal antibodies Mayer's hematoxylin and eosin (H & E) (Roboz Sur- alone and the secondary rhodamine-conjugated anti- gical, Washington, DC) to identify historically the body alone. After three more washes in PBS with different tissues. The remaining slides from each sodium azide, the sections were incubated with 50 JUL block were immersed in acetone for 20 min at of 1.5 mg/ml NBD-phallacidin (Molecular Probes, -20°C, air dried, and labeled for tubulin. vimentin. Junction City. OR) in PBS with sodium azide for 30 cytokeratin, ordesmin, using an indirect rhodamine- min at room temperature, washed in PBS with so- 12 conjugated antibody technique. Some sections were dium azidc, and mounted in a 1:1 mixture of PBS 12 then labeled for f-actin with NBD-phallacidin. and glycerol. In this labeling technique, the sections were first The H & E- and fluorescent-stained slides were incubated with one of the following antibodies to se- examined and photographed with an Olympus BH lected cytoskeletal elements: (1) mouse monoclonal microscope (Tokyo, Japan) using a 490-nm excita-

Figs. 1-8. Fig. I. Low power view of a fixed frozen section of the trabecular meshwork (TM) stained with Mayer's hemotoxylin and eosin (H & E). Corneal cndothclium (CN) and a portion of the ciliary body (CB) are shown, also. X 100. Fig. 2. Higher power H & E fixed frozen section of the trabecular meshwork region. Numerous dark staining nuclei of trabecular meshwork cells can be distinguished (arrows) covering the meshwork of beams. X1000. Fig. 3. High power fixed frozen section of the trabccular meshwork fluoresccntly stained for filamentous act in using NBD phallacidin. The actin localizes both around the nucleus (N) of the trabecular meshwork cell in and within the fine cell processes (CP) which cover the collagen beams of the meshwork. X900. Fig. 4. The same area of trabecular meshwork as in Figure 3 stained for tubulin using indirect rhodaminc conjugated antibody technique. The distribution of tubulin staining is similar to that of actin: it is found around the nucleus of the trabecular meshwork cell and within the fine processes (CP) which cover the collagen beams of the meshwork. X900. Fig. 5. High power fixed frozen section of the trabccular meshwork fluoresccntly stained for vimentin using an indirect rhodaminc conjugated antibody technique. The vimentin stain localizes to a few focal areas of the trabecular meshwork cells (arrows). A slight background fluorescence is noted in the collagen beams of the meshwork. x9()0. Fig. 6. High power unfixed frozen section of the trabccular meshwork stained for vimentin. Note a more uniform stain of vimentin (arrow) throughout the trabccular meshwork cells when compared to the fixed frozen section in Figure 5. X1000. Fig. 7. High power unfixed frozen section of the junction between the trabecular meshwork and ciliary body stained desmin. Note the marked staining of the ciliary muscle (CB) and the lack of specific stain in the trabccular meshwork (TM). X450. Fig. 8. High power fixed frozen section of the trabecular meshwork incubated with the secondary rhodamine conjugated antibody alone. A slight background stain is noted in the collagen beams but no staining is seen in the trabecular meshwork cells. X900.

Downloaded from iovs.arvojournals.org on 10/02/2021 No. 9 IN 5ITU HUMAN CYTOSKELETON / Wemreb ond Ryder 1841

CB CN TM 4fe>

Downloaded from iovs.arvojournals.org on 10/02/2021 1842 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / September 1990 Vol 31

tion filter and a 530-nm barrier filter for NBD and a rhodamine-conjugated antibody alone revealed a 545-nm excitation filter and a 610-nm barrier filter faint nonspecific staining in the collagen beams with- for rhodamine. Observations on the intensity of fluo- out specific staining in the trabccular-meshwork cells rescent staining were made with the 100X objective (Fig. 8). and graded on a scale of 0 (no discernible fluores- Corneal cndothclium. Descemefs membrane, and cence) to +++ (intense fluorescence) for each cell keratocytcs could be distinguished readily (Fig. 9). No type. specific fluorescent staining was noted with either antidesmin or with the secondary rhodamine-conju- Results gated antibody alone (Fig. 10). Anti-f-actin staining The H & E-stained sections of representative tissue was found diffusely throughout the corneal endothe- blocks from each eye enabled each tissue which was lium (with a somewhat more intense staining at the fluorescently stained for cytoskelctal elements to be cell periphery) and keratocytes but not in Descemet's identified (Fig. 1). From these preliminary observa- membrane (Fig. 1 1). Tubulin staining was seen tions, no major structural differences in tissue integ- throughout the conical endothelium (Fig. 12). In rity were noted between fixed frozen and unfixed fro- fixed frozen sections of keratocytes. the tubulin stain zen sections. The trabecular meshwork could be dis- was either seen throughout the cell or just around the tinguished clearly from the adjacent corneal nuclear region (Fig. 12), whereas in unfixed frozen endothelium, ciliary body, and overlying sclera. In sections a more uniform tubulin stain was observed the trabecular meshwork. cells covered the collagen throughout the cells. Antivimentin (Fig. 13) stain was beams (Fig. 2). Cytoskeletal staining patterns of these localized throughout the corneal-endothelial cells. An trabecular-meshwork cells, and of corneal endothe- intense stain was noted in the corneal endothelium lium, corneal epithelium, and keratocytcs, are sum- with the general anticytokcratin antibody (Fig. 14). marized in Table 1. For f-actin, tubulin. and desmin, However, no distinct stain pattern in the corneal en- the staining patterns were similar between fixed fro- dothelium was noted with the PKK-1 (Fig. 15) or zen and unfixed frozen tissue. However, with vimen- 4.62 (Fig. 16) anticytokeratin antibodies. tin and the various cytokeratins, a more uniform and Corneal epithelium was distinguished readily from intense pattern was observed on unfixed frozen mate- the underlying stroma (Fig. 17). Filamentous actin rial. was seen to localize in all cell layers (Fig. 18) with the In the trabecular meshwork. both anti-f-actin (Fig. most intense staining at the periphery of the cell. In 3) and antitubulin fluorescent staining (Fig. 4) were some sections, the actin stain was slightly more in- found around the trabecular-meshwork cell nucleus tense in the basal layer of cells (Fig. 18). Antitubulin and within the fine cell processes which cover the stain was diffusely distributed throughout these cells. collagen beams. Antivimentin staining localized in Fluorescent stain for antivimentin was very faint and only a few focal areas of the trabecular-meshwork nonspecific in the epithelium (Figs. 19, 20). In fixed cells in unfixed frozen sections (Fig. 5). but was seen frozen sections of underlying keratocytes, the antivi- throughout the trabecular cells in unfixed frozen sec- mentin stain varied from an intense stain of some tions (Fig. 6). Neither cytokeratin staining with any of cells to a general faint stain. By contrast, in unfixed the three tested antibodies (Fig. 7) nor desmin stain- frozen sections the vimentin stain of kcratocytes was ing was evident in the trabecular-mcshwork cells. more uniform and intense (Fig. 20). In fixed frozen Desmin staining, however, was noted in the adjacent sections, the general anticytokeratin staining was seen ciliary muscle (Fig. 7). Incubation with the secondary as a diffuse stain in the basal layers of corneal epithe-

Tablc I. Fluorescence staining of trabecular meshwork and corneal cells"

Trabecular meshwork Conical Corneal cells encloihelium Keraiocylcs epithelium

NBD (f-actin) +++/+++ + ++/-M- + + + -I-/+ + + ++/+ + Tubulin +++/+ + + + ++/+++ + -I-/I- + + +/+ Vimentin 0. ++/++ +-1--I-/+++ +/+ + + 0/0 Desmin 0/0 0/0 0/0 0/0 Cvtokcratin (Amcrsham) 0/0 -I-+/++ 0/0 -I-. + + +/+++ PKK-1 0/0 0/0. + 0/0 +. +++/+++ 4.62 0/0 0/0 0/0 0/0. + Secondary antibody alone 0/0 0. -1-/0. -i- 0/0 0. +/0. +

* Values represent fixcd/unlixed staining patterns. Fluoresccnc inicnsit\ is staining: (++ + ) intense staining. A comma between intensity gradings indi- graded as follows: (0) no disccrnable stain: (•(-) faint staining; (++) moderate cates intercellular variation.

Downloaded from iovs.arvojournals.org on 10/02/2021 No. 9 IN SITU HUMAN CYTOSKELETON / Wemreb ond Ryder 1843

Figs. 9-16. Fig. 9. H & E stained fixed frozen section of the area just anterior to the trabecular meshwork. A single flat layer of corneal endothelial cells (CN) covers Descemefs membrane (DM). Beneath the membrane, keratocytes (F) can be seen, X900. Fig. 10. High power fixed frozen section of the corneal endothelium region incubated with the secondary rhodamine conjugated antibody alone. A faint background stain is noted in the corneal endothelial cells and underlying stroma. X900, Fig. 11. High power fixed frozen section of the corneal endothclium region stained for filamentous actin with NBD phallacidin. Actin is seen to localize both in the corneal endothelium (CN), especially at the periphery, and in the keratocytes (F) beneath Descemet's membrane (DM). X800. Fig. 12. The same area as in Figure 11 fluorescenlly stained for microtubules. The microtubules staining is found throughout the corneal endothelium (CN) and in focal areas of the keratocytes (F) (DM—Descemet's membrane). X800. Fig. 13. Fixed frozen corneal endothelium fluorescently stained for vimentin. The vimentin stain is localized throughout the corneal endothelium. Fig. 14. Fixed frozen corneal endothelium fluorescently stained with the general anticytokeratin antibody. As with vimentin, the stain is seen throughout the corneal endothelium (DM—Descemet's membrane). X900. Fig. 15. Fixed frozen corneal endothelium flourescently stained with the PKK-1 anti-cytokeratin antibody. There is no discernable stain in the corneal endothelium (DM—Descemet's membrane). X900. Fig. 16. Fixed frozen corneal endothelium stained with the 4.62 anti-cytokeratin antibody. Again, there is no discernable stain in the corneal endothelium (DM—Descemet's membrane). X900.

Downloaded from iovs.arvojournals.org on 10/02/2021 1844 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Seprember 1990 Vol. 31

Figs. 17-22. Fig. 17. H & E stained fixed frozen section of the corneal epithelial region. The corneal epithelium appears in a stratified layer of cells over the connective tissue stroma. X560. Fig. 18. High power fixed frozen section of the corneal epithelial region stained for filamentous actin using NBD phallacidin. The actin stain is most intense at the periphery of each cell. The staining is slightly more intense in the basal layer of cells overlying Bowman's membrane (BM) than in the more superficial layer of cells (arrow). X900. Fig. 19. High power section of the corneal epithelium iluorescently stained for vimentin. A faint nonspecific stain is seen in the epithelial cells (BM—Bowman's membrane). X900. Fig. 20. Unfixed frozen section of the nasal region of the corneal epithelium and underlying stroma stained for vimentin. A faint, nonspecific stain is seen in the corneal epithelium, while an intense stain is observed within the underlying keratocytes (arrows). X900. Fig. 21. High power fixed frozen section of corneal epithelium stained with the general anti-cytokeratin antibody (Amersham). A diffuse stain is seen throughout the of the epithelial cells. This stain is markedly more intense at the superficial layer of epithelial cells (arrow) (BM—Bowman's membrane). X900. Fig. 22. Unfixed frozen section of corneal epithelium stained with the general anti-cytokeratin antibody (Amersham). Note a more uniform intense staining throughout all layers of epithelium when compared to Figure 21. X900.

Hum and a more intense stain in the most superficial ing of the epithelium or underlying connective tissue layers of epithelium (Fig. 21). By contrast, in unfixed was noted with either antidesmin staining or with frozen sections the anticytokeratin stain was more incubation with the secondary rhodamine-conju- uniformly intense throughout all cell layers (Fig. 22). gated antibody alone. A similar staining pattern between fixed and unfixed tissues was noted with the PKK-1 antibody, but the Discussion 4.62 anticytokeratin revealed no specific staining In the current study, the staining pattern for tubu- pattern in the corneal epithelium. No specific stain- lin and f-actin in human trabecular-meshwork cells

Downloaded from iovs.arvojournals.org on 10/02/2021 No. 9 IN SITU HUMAN CYTOSKELETON / Wemreb and Ryder 1845

in situ was somewhat different from that which we the present study. In the trabecular-meshwork cells found previously in cultured human and monkey tra- and keratocytes in situ, the vimentin stain was seen becular-meshwork cells.12"14 Cultured human cells only in a few focal areas of the cell processes on fixed and cynomolgus monkey cells had a well-defined frozen tissue, but it was seen more uniformly stress-fiber network of f-actin and microtubules ra- throughout the cells in unfixed frozen tissue. Such diating from the nuclear region. Such fine resolution observations suggest the possibility of a structural of structures was not observed in human trabecular- and/or functional heterogeneity of the trabecular- meshwork cells in situ. This is probably due to the meshwork cell population. However, this focal stain- fact that our observations of cultured cells were made ing pattern of trabecular-meshwork cells and kerato- on broad flat cells approximately 0.25-0.5 /urn in cytcs also may be due to the processing method used thickness, and observations of trabecular-meshwork in this study. Although the fixation step we used may cells in situ were made on 5-//m thick sections. In help to maintain the structural integrity of the tissue, these latter observations, only the nuclear profile and it may also alter the antigenicity of certain cytoskele- a cross-sectional area of thin cell processes enveloping tal proteins. Such a mechanism is supported by our the collagen beams were normally seen. Nevertheless, observations on keratocytes and trabecular-mesh- both f-actin and tubulin were demonstrated around work cells in unfixed frozen tissue and in a recent the nucleus and extending into the fine cell processes study using unfixed frozen corneal tissues which surrounding the collagen beams. Previous electron showed vimentin staining in keratocytes.25 Using microscopic observations on S-l labeled, thin-sec- fixed tissue, we may have detected only higher con- tioned in situ trabecular-meshwork cells noted bun- centrations of the vimentin in the trabecular- dles of actin filaments.17 As in other cell types,1819 meshwork cells and keratocytes. By comparison, in actin in these trabecular-meshwork cell processes corneal endothelium an intense vimentin stain was may play a role in the attachment of the cell to the seen throughout each cell. Although the biologic pro- collagen beam, maintaining the flat cell profile cesses mediated by vimenlin are not understood against the beam, and providing the motile force for clearly, it has been implicated as playing a role in trabecular-meshwork cell process migration over the maintaining cell shape and internal organization23 beams. Tubulin, which localized in the fine cell pro- and mediating cell-substrate and cell-cell attach- cesses, may enhance the structural integrity of the cell ment.26 Hence, it is likely that vimentin has similar and be involved in phagocytic and secretory func- functions in situ in trabecular-meshwork cells, cor- tions. ncal endothelium. and keratocytes. Tubulin and f-actin also were found throughout Equally striking was the distribution of the cyto- the corneal endolhclium and corneal epithelium. The intermediate-filament proteins as revealed by f-actin stain was especially prominent in the cortical three different anticytokeratin antibodies. The gen- regions of these two cell types. A similar concentra- eral anticytokeratin stain was prominent in corneal tion of f-actin in the cortical region of other cell types epithelium. As with the vimentin staining, the cyto- has been reported.2021 This cortical actin may play a kcratin staining was variable in the corncal epithe- role in maintaining the shape of the corneal endothe- lium from fixed frozen sections and more uniform in lium and epithelium and may form part of the at- unfixed frozen sections. It is possible that in fixed tachment apparatus for cell-cell or cell-substrate in- material the cytokeratins are antigenically altered teractions. The cortical staining pattern for actin was with less labeling affinity for the antibody. Further- particularly intense in the basal cells of the corneal more, this general anticytokeratin staining pattern epithelium. Similar observations have been made was seen also in the corneal endothelium. Although in migrating corneal-cpithelial cells during wound the presence of cytokeratin in endothelial cells is not healing.22 encountered normally, these observations of general One striking finding of this study was the differen- cytokeratin concur with those of a recent report in tial staining patterns seen with desmin. vimentin. and which keratin staining in normal human corneal-en- cytokeratin intermediate-filament proteins. Staining dothelial cells was shown on fixed sections using a for desmin intermediate-filament protein, found general cytokeratin derived from epidermal cells.27 In normally in muscle cells,21 was not seen in the exam- that study, the corneal endothelium also lacked sev- ined cell types, but was observed in the adjacent cili- eral histochemical and ultrastructural markers char- ary body. By contrast, in a recent in vitro study of acteristic of endothelium in other areas. From those cultured trabecular-meshwork cells, desmin staining observations, it was concluded that the corneal endo- was observed.24 This may be due to a different ex- thelium may not be a true endothelium and may pression of intermediate filaments in these cultured have several epithelial-cell characteristics. However, cells or to an in situ masking of the desmin protein in transmission electron-microscopic studies of corneal

Downloaded from iovs.arvojournals.org on 10/02/2021 1846 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / Seprember 1990 Vol 31

endothelium have not demonstrated keratin tonofi- the manuscript. The eyes were obtained from the San laments or .28 It is possible that the cyto- Diego Eye Bank. keratins in these cells exist in a more globular form. In fact, it is possible that morphologic features of References corneal endothelium may be similar to simple epithe- 1. Polansky JR. VVcinrcb R, and Alvarado JA: Studies on human lial cells in other tissues such as Bowman's capsule in trabccular meshwork cells propagated in vitro. Vision Res the kidney29 and the rete testis.™ Both vimentin and 21:155. 1981. cytokeratins arc seen routinely in the flat monolayer 2. Polansky JR. Wood IS. Maglio MT. and Alvarado JA: Trabcc- ular meshwork cell culture in glaucoma research. Ophthalmol- of simple epithelial cells which lines fluid chambers in ogy 91:580. 1984. these tissues. As in these other tissues, both cytokera- 3. Polansky JR. Bloom E. Konami D. Wcinrcb RN. and Alva- tin and vimentin may contribute to the regulation of rado JA: Cultured human trabecular meshwork cells: Evalua- the shape and mechanical resistance of the corneal tion of hormonal and pharmacological responses in vitro. /// endothelium. However, no marked cytokeratin- Recent Advances in Glaucoma. Ticho V and David R. editors. staining pattern in corneal-endothelial cells was Amsterdam. Exccrpta Mcdica. 1984. pp. 201-206. 4. Bill A: The drainage of aqueous humor. Invest Ophthalmol noted with either the PKK-I or the 4.62 anticytoker- 14:1. 1975. atin antibodies. The absence of stain with the 4.62 5. Rohen J\V and van der Zypcn E: The phagocytic activity of the antibody is of interest in that it specifically stains for trabecular meshwork endothelium: An electron-microscopic the 40-kD cytokeratin species (no. 19) seen in differ- study of the vcr\et (Ccrcopilhecus aethiops). Gracfcs Arch Clin entiated simple epithelium.16 Furthermore, the ExpOphihalmol 175:143. 1968. 6. Rohen JW and Lutjen-Drecoll E: Biology of the trabccular PKK-1 antibody has been shown to be reactive to the meshwork. //; Basic Aspects of Glaucoma Research. Lutjcn- 46-kD, 52-kD, and 54-kD cytokeratins in HeLa cells Drccoll E. editor. Stuttgart. Schattaucr. 1982, p. 141. which may be similar to the 46-kD, 52-kD. and 7. Francois J: The importance of the mucopolysaccharides in 54-kD cytokeratins seen in most simple cpithelia.16 It intraocular pressure regulation. Invest Ophthalmol 14:173. is possible that the general anticytokeratin used in 1975. this study and in the previous study27 may stain for 8. Schachtschabel DO. Bigalke B. and Rohen JW: Production of 1611 glycosjiminoglycans by cell cultures of the trabecular mesh- one of the 19 cytokeratin species not detected by work of the primate eye. Exp Eye Res 24:71. 1977. the PKK-I or 4.62 antibodies. 9. Polansky J. Gospodarowicz D. Weinrcb R. and Alvarado J: Human trabecular meshwork cell culture and glycosaminogly- Various fluorescent cytoskeletal stains to actin. tu- can synthesis. ARVO Abstracts. Invest Ophthalmol Vis Sci bulin, and especially intermediate filaments are now l7(Suppl):207. 1978. used widely to type both normal and neoplastic cells 10. Schachlschabel DO. Rohen JW. Wevcr J. and Sames K: Syn- in various tissues.2332 In our study, the use of unfixed thesis and composition of glycosaminoglycans by cultured frozen material did not appear to alter structural in- human trabecular meshwork cells. Gracfcs Arch Clin Exp tegrity markedly compared with fixed material. How- Ophthalmol 218:113. 1982. 1 1. Acott TS. Westcott M. Passo MS. and Van Buskirk EM: Tra- ever unfixed frozen tissues had a more uniform becular meshwork glycosaminoglycans in human and cyno- staining pattern. Therefore, it appears that using un- molgus monkey eyes. Invest Ophthalmol Vis Sci 26:1324. fixed frozen material is preferable to fixed material 1985. when studying in situ cytoskeletal localization on this 12. Ryder MI and VVcinrcb RN: The cytoskelelon of the cyno- type of tissue. Also, we demonstrated that trabecular- molgus monkey trabecular meshwork cell: I. General consider- ations. Invest Ophthalmol Vis Sci 27:1305. 1986. meshwork cells display a staining pattern to f-aclin. 13. VVcinrcb RN. Ryder MI, and Polansky J: The cytoskclcton of microtubules, and especially vimentin and cytokera- the cynomolgus monkey trabecular meshwork cell: II. Influ- tin intermediate-filament proteins which is clearly ence of cytoskelcton-activc drugs. Invest Ophthalmol Vis Sci distinct from neighboring corneal endothelium, cor- 27:1312. 1986. 14. Ryder Ml. Wcinrcb RN. Alvarado J. and Polansky J: The neal epithelium, and keratocytes. Such labeling tech- cytoskclcton of cultured human trabecular meshwork cells: I. niques may be valuable adjuncts in monitoring Characterization and drug responses. Invest Ophthalmol Vis phenotypic cell expression in glaucoma and other oc- Sci 29:251. 1988. ular diseases. 15. Grierson I. Millar L. Yong JD. Day J. McKechnie NM. Hitchins C. and Boulton M: Investigations of cytoskclctal ele- ments in cultured bovine meshwork cells. Invest Ophthalmol Key words: cornea, trabccular-mcshwork cell, glaucoma, Vis Sci 27:1318. 1986. 16. Sun T-T. Tseng SCG. Huang AJW, Cooper D. Schermcr A. actin, intermediate filaments Lynch MH. Weiss R. and Eichner R: Monoclonal antibody studies of mammalian epithelial keratins: A review. Ann N Y Acknowledgments Acad Sci 455:307, 1985. 17. Gipson IK and Anderson RA: Actin filaments in cells of The authors thank Yvonne DeSousa and Eileen Wong human trabccular meshwork and Schlemm's canal. Invest for their technical assistance. They also acknowledge the Ophthalmol Vis Sci 18:547. 1979. assistance of Michelle Elig and Carol Fiuren in preparing 18. Abercrombie MT. Heaysman EM, and Pegrum SM: The loco-

Downloaded from iovs.arvojournals.org on 10/02/2021 No 9 IN SITU HUMAN CYTOSKELETON / Wemreb ond Ryder 1847

motion of fibroblasts in culture: IV. Electron microscopy of the 26. Czernobilsky B. Moll R, Lcppicn G. Schweikhart G, and leading lamella. Exp Cell Res 67:359, 1971. Frankc WW: Dcsmosomal plaque-associated vimentin fila- 19. Opas M and Kalnius V: Light microscopical analysis of focal ments in human ovarian granulosa cell tumors of various his- adhesions of retinal pigmented epithelial cells. Invest Ophthal- tologic patterns. Am J Pathol 126:476, 1987. mol VisSci 27:1622. 1986. 27. Shamsuddin AKM. Nivankari VS. Purncll DM. and Chang 20. Higbce RG and Hazlclt 1_D: Actin filament localization and SH: Is the corneal posterior cell layer truly cndothclial? Oph- distribution in the young adult mouse cornea: A correlative thalmology 93:1298. 1986. immunolluorcscent and cytochcmical study. Exp Eye Res 28. Hogan MJ. Alvarado JA. and Weddcll JE: In Histology of the 36:171.1983. Human Eye. Philadelphia. WB Saunders, 1971, pp. 102-109. 21. Soong HK and Fairlcy JA: Actin in human conical epithelium. 29. Holthofer H. Micltincn A, Lento VP. Lchtancn E, and Vir- Arch Ophthalmol 103:565. 1985. tancn I: Expression of vimentin and cytokcratin types of inter- 22. Nakagawa S. Nishida T. and Manabe R: Actin organization in mediate filament proteins in developing and adult human kid- migrating conical epithelium of rabbits in situ. Exp Eye Res neys. Lab Invest 50:552. 1984. 41:335. 1985. 30. Achtstatter T. Mill R. Moore B. and Frankc WW: Cytokcratin 23. Osborn M. Geisler N. Shaw G. Sharp G. and Weber K: Inter- polypeptidc of different cpithclia of the human male urogenital mediate filaments. Cold Spring Harb Symp Quant Biol 46:413, tract: Immunoiluorcscence and gel electrophorctic studies. J 1982. Histochem Cytochcm 33:415. 1985. 24. Iwamoto Y and Tamura M: Immunocytochemical study of 31. Moll R. Franke WW. Schiller DL. Gcigcr B. and Krcplcr R: intermediate filaments in cultured human trabecular cells. In- The catalog of human cytokcratins: Patterns of expression in vest Ophthalmol Vis Sci 29:244. 1988. normal epithclia. tumors, and cultured cells. Cell 31:11. 1982. 25. Risen LA. Binder PS. and Nayak SK: Intermediate filaments 32. Osborn M and Weber K: Tumor diagnosis by intermediate and their organization in human corneal cndothelium. Invest filament typing: A novel tool for surgical . Lab Invest Ophthalmol VisSci 28:1933. 1987. 48:372. 1983.

Downloaded from iovs.arvojournals.org on 10/02/2021