Contractile proteins in pericytes. I. Immunoperoxidase localization of tropomyosin.

In these studies we have compared the relative amounts and isoforms of tropomyosin in capillary and postcapillary venule pericytes, endothelial cells, and vascular smooth muscle cells in four rat microvascular beds: heart, diaphragm, pancreas, and the intestinal mucosa. The results, obtained by in situ immunoperoxidase localization, indicate that (a) tropomyosin is present in capillary and postcapillary venule pericytes in relatively high concentration; (b) the tropomyosin content of pericytes appears to be somewhat lower than in vascular smooth muscle cells but higher than in endothelia and other vessel-associated cells; and (c) pericytes, unlike endothelia and other nonmuscle cells, contain detectable levels of tropomyosin immunologically related to the smooth muscle isoform. These results and our previous findings concerning the presence of a cyclic GMP-dependent protein kinase (Joyce, N., P. DeCamilli, and J. Boyles, 1984, Microvasc. Res. 28:206-219) in pericytes demonstrate that these cells contain significant amounts of at least two proteins important for contraction regulation. Taken together, the evidence suggests that pericytes are contractile elements related to vascular smooth muscle cells, possibly involved, as are the latter, in the regulation of blood flow through the microvasculature.

Contractile proteins in pericytes. II. Immunocytochemical evidence for the presence of two isomyosins in graded concentrations.

This paper describes the localization of isomyosins in the pericytes of four rat microvascular beds: heart, diaphragm, pancreas, and the intestinal mucosa, by use of immunoperoxidase techniques and IgGs specific for either nonmuscle or smooth muscle isoforms. Based on the semiquantitative nature of the peroxidatic reaction, we concluded that the amount and distribution of these isoforms vary with the microvascular bed and also with vascular segments within the same bed. In the pericytes of small capillaries, nonmuscle isomyosin is the predominant form, whereas the smooth muscle isomyosin is present in very low concentration. A reversed relationship is found in the pericytes associated with larger capillaries and postcapillary venules. These results, taken together with previous findings on actin (Herman, I., and P. A. D'Amore, 1983, J. Cell Biol. 97:278a), tropomyosin (Joyce, N. C., M. F. Haire, and G. E. Palade, 1985, J. Cell Biol. 100:1379-1386), and cyclic GMP-dependent protein kinase (Joyce, N., P. DeCamilli, and J. Boyles, 1984, Microvasc. Res. 28:206-219), indicate that pericytes contain proteins essential for contraction in higher concentration than any other cells associated with the microvasculature, except smooth muscle cells. Pericytes appear to be, therefore, cells differentiated for a contractile function within the microvasculature.

Transforming growth factor-beta suppresses proliferation of rabbit corneal endothelial cells in vitro.

Corneal endothelial cells in vivo appear to be inhibited in G1 phase of the cell cycle. Studies were carried out to determine whether cultured rabbit corneal endothelium expresses transforming growth factor-beta (TGF-beta) receptor types I, II, and III, suggesting they would be sensitive to a TGF-beta-induced signal. In addition, we explored if TGF-beta might mediate this G1 phase inhibition by implementing flow cytometry and 5-bromo-2'-deoxyuridine (BrdU) immunofluorescence. Reverse transcription-polymerase chain reaction (RT-PCR) products of the expected size were obtained for all three TGF-beta receptor types. Flow cytometry revealed a dose-dependent suppression in the percentage of S phase cells in cultures treated with TGF-beta1 or TGF-beta2. The lowest percentage of S phase cells was found for 10 ng/ml TGF-beta1 and 0.1 ng/ml TGF-beta2. BrdU, an S phase marker, was immunolocalized, and semiquantitative analysis of stained cells showed a maximum suppression of S phase entry at 18 h for 10 ng/ml of TGF-beta11 and 24 h for 10 ng/ml of TGF-beta2. In rabbit, the corneal endothelium expresses TGF-beta receptor types I, II, and III, permitting a TGF-beta signal to be transduced. Flow cytometry reveals a dose-dependent response to both TGF-beta1 and TGF-beta2, and the cells are more sensitive to TGF-beta2. At optimal TGF-beta concentrations, the percentage of S phase cells is comparable to that of a non-proliferating culture, suggesting TGF-beta prevents the cells from proceeding through the G1/S phase transition. This suppression was also seen with BrdU labeling. Together, these results indicate that TGF-beta could be one of the pathways that leads to G1 phase arrest in corneal endothelial cells.

Cell cycle kinetics in corneal endothelium from old and young donors.

PURPOSE: To compare cell cycle kinetics in corneal endothelial cells from young and old donors. METHODS: Human corneas were obtained from the eye bank and separated into two groups: young (19 corneas, <30 years of age) and old (40 corneas, >50 years of age). Corneas were cut in quarters, and the endothelium was released from contact inhibition by producing a 2-mm scrape wound. Unwounded endothelium acted as a negative control. Corneal pieces were exposed for 24, 36, 48, 60, 72, and 84 hours to medium containing 10% fetal bovine serum, 20 ng/ml fibroblast growth factor, and 50 mg/ml gentamicin or the same medium supplemented with 10 ng/ml epidermal growth factor (EGF). Tissue was fixed, immunostained for Ki67 (a marker for the late G1-through M-phase) or for 5-bromo-2'-deoxyuridine (BrdU; a marker for the S-phase), and mounted in medium containing propidium iodide (PI) to visualize all nuclei. Confocal images were evaluated using an image analysis program to count Ki67-positive and PI-stained cells and to evaluate cell cycle position. Cells were counted in 15x100 microm2 areas randomly selected from each wound, and the mean was used for subsequent calculations. RESULTS: Human corneal endothelial cells could be reliably scored for their position within the cell cycle using Ki67 staining patterns. In both age groups, cells repopulating the wound area stained positively for Ki67, whereas no Ki67 staining was observed in unwounded areas under any condition tested. Cells from old donors treated with fetal bovine serum and FGF stained positively for Ki67, indicating that these cells were actively cycling. Compared with cells from young donors, old cells entered the cell cycle more slowly (48 versus 36 hours), the peak of Ki67 staining occurred later (72 versus 60 hours), and fewer cells proliferated (23% versus 47%) or exhibited mitotic figures (4% versus 7%). Addition of EGF to the culture medium increased Ki67 staining in both groups, but the effect on old cells was more dramatic. More cells from old donors entered the cell cycle by 36 hours after wounding, the number of proliferating cells increased 1.6-fold, and the relative number of mitotic figures increased 2.5-fold over cells treated in the absence of EGF. CONCLUSIONS: Regardless of donor age, corneal endothelial cells can enter and complete the cell cycle. In the presence of fetal bovine serum and FGF, cells from old donors can proliferate but respond more slowly and to a lesser extent than cells from young donors. EGF added to the medium stimulates cells from old donors to enter the cell cycle faster, increases the relative number of actively cycling cells, and increases the number of cells exhibiting mitotic figures. The resultant hypothesis is that it is possible to stimulate a significant proliferative response in corneal endothelial cells from old individuals. Administration of an optimal combination of stimulatory growth factors is required under conditions in which cells have been transiently released from contact inhibition.

Transforming growth factor-beta receptor expression in human cornea.

PURPOSE: Limbal basal cells and corneal endothelial cells appear to be inhibited in the G1 phase of the cell cycle. As a preliminary to determining whether transforming growth factor-beta (TGF-beta) might mediate this inhibition, investigation was made to determine whether human corneal and limbal cells express TGF-beta receptor types I (RI), II (RII), and III (RIII). METHODS: Corneas from eight human donors, aged stillborn to 85 years, were fresh frozen, cryostat sectioned, and prepared for indirect immunofluorescence localization of RI, RII, and RIII, using an established protocol. Corneas from donors 50 years of age or older were used to prepare RNA from the epithelium and endothelium. Reverse transcription-polymerase chain reaction was conducted using primers specific for each TGF-beta receptor type. RESULTS: Immunolocalization patterns for RI, RII, and RIII were similar, regardless of donor age. Binding of RI and RII antibodies was barely detectable in central corneal epithelium; however, most limbal basal cells stained positively for RI and RII. All layers of central corneal epithelium and the suprabasal layers of the limbus stained positively for RIII, whereas staining for this receptor was markedly decreased in limbal basal cells. Corneal endothelium bound the antibody for all three TGF-beta receptor types. In the same tissue sections, antibody staining for the RIII protein was more intense in corneal endothelial cells than in limbal basal cells. Polymerase chain reaction product for RI, RII, and RIII was detected in the epithelium and in the endothelium. CONCLUSIONS: Limbal basal cells and corneal endothelial cells expressed mRNA and protein for TGF-beta receptor types I, II, and III, suggesting that both cell types can transmit a TGF-beta-induced signal. These two cell types may differ in their relative response to those TGF-beta isoforms that require binding to RIII for signal transduction, in that staining intensity for RIII was relatively low in limbal basal cells compared with that in the endothelium. That limbal basal and corneal endothelial cells express receptors for TGF-beta suggests that this cytokine could mediate G1 phase arrest in these two cell types.

Protein kinase C activation during corneal endothelial wound repair.

In previous studies, the authors have shown that the two forms of cell translocation that occur during corneal endothelial monolayer wound repair can be pharmacologically separated. Epidermal growth factor (EGF) enhanced the breaking of cell-cell contacts and movement of individual cells from the wound edge, while indomethacin, an inhibitor of PGE2 synthesis, promoted cell enlargement and spreading of the confluent monolayer sheet into the wound defect. From these findings, the authors hypothesized that the two forms of cell translocation were stimulated by different but coordinately regulated second messenger systems. The current studies used selected protein kinase C (PKC) stimulators and inhibitors, Rh-phalloidin staining of actin filaments, and immunofluorescent localization of PKC to show that: (1) PKC acts as a mediator of the EGF-induced enhancement of the migratory response; (2) the enhanced migratory response results, at least in part, from short-term EGF stimulation of PKC; (3) PKC is a mediator of the EGF-induced alterations in the actin cytoskeleton; and (4) PKC becomes activated in cells at the wound edge during normal, endogenously stimulated wound repair. The results of these studies provide suggestive evidence that wounding of the corneal endothelial monolayer must produce an endogenous, EGF-like stimulation of PKC activity in cells at the wound edge. One effect of PKC activation that must contribute to stimulation of individual cell migration is the induction of cytoplasmic changes that lead to alterations in actin filament organization.

In vitro pharmacologic separation of corneal endothelial migration and spreading responses.

Repair of corneal endothelial wounds involves two forms of cell translocation: (1) "migration," in which individual cells at the wound edge break contacts with neighboring cells and move as individuals into the wound defect, and (2) "spreading," in which cells within the confluent monolayer adjacent to the wound move as a group into the wound area. The authors combined morphometric analysis of Giemsa-stained cultures, phase-contrast video microscopy, and Rh-phalloidin staining of actin filaments to study the effects of epidermal growth factor (EGF) and indomethacin on the migratory and spreading responses to wounding using an in vitro wound-closure model which mimics the amitotic state and general behavior of human corneal endothelium. They found that EGF stimulated the migration of individual cells from the wound edge, induced cellular elongation, and promoted a diffuse distribution of actin filaments. Indomethacin promoted spreading of the confluent monolayer into the wound defect, induced enlargement and flattening of cells, and promoted the formation of long, thick actin stress fibers. These results provide evidence that the migration and spreading responses of corneal endothelial cells to wounding can be pharmacologically separated. The findings suggest that migration of individual cells during wound repair may result from an endogenous form of EGF-like stimulation and that the elongated shape associated with this form of translocation results, at least in part, from an EGF-like alteration in actin-filament organization. Spreading of the confluent monolayer to cover the wound defect may result from a decrease in cyclic adenosine monophosphate induced by a transient reduction in prostaglandin E2 synthesis. This form of translocation may result, in part, from enlargement and flattening of corneal endothelial cells secondary to an enhancement of actin stress-fiber formation.

EDTA: a promoter of proliferation in human corneal endothelium.

PURPOSE: To determine whether it is possible to induce proliferation in the endothelium of older donor corneas by treatment of the intact monolayer with EDTA. METHODS: Corneas from donors 52 to 75 years of age were obtained from an eye bank and were usually cut in quarters to increase sample size. The effect of EDTA dose (0.02-2.0 mg/ml) and incubation time (6, 30, and 60 minutes) on endothelial cell-cell contacts was evaluated by staining for ZO-1, a cell junction marker. Cell death was tested by a commercial live-dead assay. Corneal pieces were incubated for 0, 24, 48, or 60 hours in culture medium (M-199, 10% fetal bovine serum, 10 ng/ml epidermal growth factor, 20 ng/ml fibroblast growth factor) before EDTA treatment. After treatment, pieces were incubated in the same medium for 24, 48, 72, or 96 hours to permit cell cycle entry. Tissue was fixed, stained for Ki67 (a marker for late G1-phase through the M-phase), and mounted in medium containing propidium iodide to visualize all nuclei. Confocal images were evaluated by computer (Image software; NIH, Bethesda, MD) to count Ki67-positive and propidium iodide-stained cells. RESULTS: EDTA released corneal endothelial cell-cell contacts in a dose- and time-dependent manner. At doses and incubation times tested, EDTA did not induce significant cell death. Preincubation in culture medium for 24 hours was needed for endothelial cells to efficiently initiate proliferation in response to EDTA. The endothelium of corneas incubated in mitogen-containing medium for up to 108 hours without EDTA treatment did not stain for Ki67. EDTA at 2.0 mg/ml for 60 minutes appeared optimal and stimulated 16% to 18% of the cells to proliferate. Ki67-positive mitotic figures were visible 48 hours after exposure to EDTA. Formation of daughter cells was visible after double-staining for Ki67 and ZO-1. CONCLUSIONS: EDTA released cells from contact inhibition and promoted proliferation in corneal endothelium from older donors. The authors hypothesize that corneal endothelium from older individuals divide in situ when exposed to positive growth factors under conditions in which cells have been transiently released from contact inhibition.

Immune privilege and immunogenicity reside among different layers of the mouse cornea.

PURPOSE: To determine the extent to which each layer of the mouse cornea displays alloimmunogenicity or immune privilege. METHODS: Intact corneas or individual or combined layers of corneas from normal or cauterized eyes of BALB/c, C57BL/6, and CD95L-deficient B6-gld mice were grafted beneath the kidney capsule of normal BALB/c, B10.D2, BALB.B mice or of BALB/c mice presensitized to donor antigens. Graft fate was assessed clinically and histologically and acquisition of donor-specific delayed hypersensitivity (DH) was assessed at selected intervals after grafting. RESULTS: Full-thickness allogeneic corneas induced vigorous DH and were rejected acutely. Similar results were obtained with allografts of corneal epithelium alone (if supported by syngeneic viable stroma), allografts of epithelium from cauterized corneas (containing Langerhans' cells), and stromal allografts deprived of endothelium. Grafts comprised of stroma plus endothelium (without epithelium) were not rejected, nor did they induce DH unless the graft had no CD95L expression. If stroma-endothelium grafts had no CD95L expression, DH directed against major histocompatibility complex (MHC), but not minor histocompatibility, alloantigens was induced. Moreover, CD95L expressed on stroma-endothelium grafts protected endothelial cells, but not stromal cells, from rejection in presensitized recipients. CONCLUSIONS: When grafted to a heterotopic site, the alloimmunogenicity of the normal cornea resides within its epithelial and stromal layers, whereas immune privilege arises from the endothelium. In normal mice, CD95L-expressing endothelium can inhibit the stroma from inducing immunity directed at MHC alloantigens, but in presensitized mice the endothelium can protect itself only from immune rejection.

Cell cycle protein expression and proliferative status in human corneal cells.

PURPOSE: To determine the relative expression of cell cycle-association proteins in human corneal and limbal epithelium and corneal endothelium in situ, to correlate staining patterns of cell cycle-associated proteins with known proliferative status of corneal and limbal epithelial cells, and to determine the relative proliferative status of corneal endothelial cells in situ by comparing their staining patterns with those of corneal and limbal epithelial cells. METHODS: Corneas from donors 6 weeks and 17, 27, 37, 53, 66, and 67 years of age were preserved in Optisol, received on ice within 24 to 36 hours of death, and immediately fresh frozen. Transverse 6-micron corneal sections were prepared for indirect immunofluorescence localization using commercial antibodies that specifically recognize the following cell cycle-associated proteins: cyclins D, E, A, and B1; protein kinases p33cdk2 and p34cdc2; and Ki67, a marker of actively cycling cells. RESULTS: All cells of the corneal and limbal epithelium and corneal endothelium stained positively for protein kinases, p33cdk2 and p34cdc2, and for cyclin B1. Staining patterns for cyclins D, E, and A and for Ki67 differed depending on the relative proliferative status of the cells. Terminally differentiated, noncycling corneal epithelial subrabasal cells did not stain significantly for cyclins D, E, or A, or for Ki67. Some corneal epithelial basal cells showed nuclear staining, particularly for cyclin D and Ki67, indicating the presence of actively cycling cells in this regenerative cell layer. In peripheral corneal epithelium, the relative number of basal cells with positive cytoplasmic staining for cyclins D, E, and A increased with proximity to the limbus. Within this region, an occasional nucleus stained positively for Ki67. Limbal basal cells, which contain a slow-cycling stem cell population, stained positively for cyclins D, E, and A within the cytoplasm. Nuclear staining for cyclin D and Ki67 was observed in a few basal cells. Occasional cells within the suprabasal layers of the limbus stained positively for Ki67. The corneal endothelium, considered a nonrenewing population, exhibited staining patterns similar to those of limbal basal cells, except that in no specimen was Ki67 staining observed. CONCLUSIONS: All corneal and limbal epithelial and corneal endothelial cells express protein kinases, p33cdk2 and p34cdc2, and cyclin B1. Relative staining patterns of the cell cycle-dependent proteins, cyclins D, E, and A, and of Ki67 acted as markers to distinguish terminally differentiated epithelial suprabasal cells that have exited the cell cycle, actively cycling epithelial basal cells, and slowly-cycling limbal basal (stem) cells. Staining patterns of the corneal endothelium most closely corresponded to those of limbal basal cells, suggesting that endothelial cells are arrested in G1-phase and have not exited the cell cycle.

TGF-beta2 in aqueous humor suppresses S-phase entry in cultured corneal endothelial cells.

PURPOSE: Corneal endothelium in vivo is arrested in G1, the phase of the cell cycle that prepares cells for DNA synthesis. In many cell types, transforming factor (TGF)-beta inhibits proliferation by inducing G1-phase arrest. Evidence indicates that corneal endothelial cells synthesize mRNA for TGF-beta1 and are also bathed in aqueous humor that contains TGF-beta2 (mainly in a latent form). As such, this cytokine may maintain the corneal endothelium in a G1-phase-arrested state in vivo. The purpose of these studies was to determine the effect of exogenous TGF-beta2 and TGF-beta2 in aqueous humor on DNA synthesis in cultured corneal endothelial cells. METHODS: Rat corneal endothelial cells were grown in explant culture and identified by morphology and reverse transcription-polymerase chain reaction using primers for specific corneal cell markers. Subconfluent cells were synchronized in the G0 phase (quiescence) by serum starvation for 24 hours. Serum was then added to the cells in the presence or absence of exogenous TGF-beta2 or activated rat aqueous humor. [3H]Thymidine was added, and radioactivity was measured at various time points to detect DNA synthesis. Preincubation of exogenous TGF-beta2 or activated rat aqueous humor with neutralizing antibody was used to test for cytokine specificity. RESULTS: A linear increase in [3H]thymidine incorporation began approximately 16 hours after serum addition, and peak incorporation occurred at approximately 24 hours. Exposure of cells to serum plus TGF-beta2 suppressed [3H]thymidine incorporation in a dose-dependent manner at concentrations ranging from 5 pg/ml to 5 ng/ml. [3H]Thymidine incorporation was also suppressed in cells exposed to serum plus rat aqueous humor diluted 1:10. Neutralizing antibody reversed the effects of both exogenous TGF-beta2 and aqueous humor. CONCLUSIONS: Exogenous TGF-beta2 and TGF-beta2 in aqueous humor suppress S-phase entry of rat corneal endothelial cells. These results suggest that this cytokine in aqueous humor could help maintain the corneal endothelium in a G1-phase-arrested state in vivo.

Mitotic inhibition of corneal endothelium in neonatal rats.

PURPOSE: Corneal endothelium in humans does not divide to any significant extent after birth; therefore, with age there is a gradual loss of cells. When cell density is reduced to a critical level, the endothelium cannot function to maintain corneal clarity, and the cornea becomes permanently cloudy. Currently, the blindness that results can be treated only by corneal transplantation. The long-term goal is to find methods to stimulate corneal endothelial proliferation in a clinically relevant manner. The first step toward achieving this goal is to identify mechanisms responsible for the induction and maintenance of mitotic inhibition of the corneal endothelium in vivo. During corneal development, the endothelium is formed by migration and proliferation of mesenchymal cells from the ocular periphery. Soon after the monolayer is formed, proliferation ceases. In tissue culture, many cell types cease proliferating upon formation of stable cell-cell and cell-substrate attachments. The goal of the present studies was to determine whether establishment of stable contacts correlates with cessation of endothelial proliferation during corneal development in vivo. METHODS: Corneas from neonatal (days 1, 3, 7, 10, 13, 14, 17, 21, 28, and 42) and adult rats were used for immunolocalization of the following: bromodeoxyuridine (BrdU), an S-phase marker; p27kip1 and p21cip1, G1-phase inhibitors; connexin-43 and ZO-1, proteins associated with gap and tight junctions, respectively; Na+/K+-ATPase and beta3-integrin, markers of plasma membrane polarity; and fibronectin and collagen type IV, constituents of Descemet's membrane. Nuclei staining positively for BrdU were counted to determine the relative number of S-phase cells at various times after birth. Marker protein expression and localization were determined by conventional fluorescence microscopy and by confocal microscopy. RESULTS: The number of endothelial cells staining positively for BrdU gradually decreased between postnatal days 1 and 13. After postnatal day 13, positive BrdU staining was no longer detectable. During the first postnatal week, cells stained positively for the G1-phase inhibitor p27kip1 but not for p21cip1. Connexin-43 achieved its mature location by postnatal day 1. ZO-1, Na+/K+-ATPase, beta3-integrin, fibronectin, and collagen type IV achieved their mature localization patterns between postnatal days 14 and 21. CONCLUSIONS: In neonatal rat, corneal endothelial cells are still entering the cell cycle at birth, but cell cycle entry gradually decreases, so that by postnatal day 13 cells are no longer entering the S-phase. The G1-phase inhibitor p27kip1, but not p21cip1, may help mediate this inhibition. Stable cell-cell and cell-substrate contacts gradually form, and monolayer maturation is complete between postnatal days 14 and 21. The results lead to the hypothesis that, in developing rat cornea in vivo, the establishment of stable cell-cell and cell-substrate contacts initiates a cascade of events, mediated by p27kip1, which induces mitotic inhibition in the endothelial monolayer.

Expression of cell cycle-associated proteins in human and rabbit corneal endothelium in situ.

PURPOSE: It is unknown why human corneal endothelium exhibits limited capacity to divide while the endothelia of other species, such as rabbit, divide in vivo at wounding and in culture. A potentially valuable source of information concerning why human endothelium has such a limited proliferative capacity lies in elucidating any differences in the molecular events governing the cell cycle of these two species. A recent study of the relative expression of cell cycle-associated proteins in donor corneas suggests that human corneal endothelial cells in vivo have not exited the cell cycle but are arrested in G1-phase. The purpose of the current study was to identify differences in cell cycle protein expression in human and rabbit endothelium that would explain the difference in their relative proliferative capacities. Specifically, the authors first ascertained the relative proliferative status of rabbit corneal endothelial cells in vivo. The expression and intracellular distribution of G1-phase regulatory proteins was then determined in both species, and the results were compared. METHODS: Corneas from New Zealand white rabbits (weight range, 2 to 3 kg) and from human donors (age range, 6 months to 67 years) were fresh frozen, cryostat sectioned, and prepared for indirect immunofluorescence microscopy using an established protocol. The following monoclonal antibodies were localized in rabbit corneal endothelium only: cyclins D, E, A, and B1; protein kinase p34cdc2; and Ki67, a marker of actively cycling cells. Localization patterns for the following G1-phase regulatory proteins were compared in both human and rabbit corneal endothelia: the tumor suppressors, pRb, p53, and p16INK4, and the transcription factor, E2F. Reverse transcription-polymerase chain reaction studies were conducted to detect mRNA for Ki67 in human and rabbit corneal cells. RESULTS: Cyclins D, E, and A were localized in the cytoplasm of rabbit corneal endothelium, whereas cyclins B1 and p34cdc2 were detected in the nucleus. No Ki67 protein or mRNA expression was detected in the endothelium of either species. In human and rabbit endothelia, p53 and p16INK4 were localized to the cytoplasm, whereas pRb was detected in the nucleus. E2F exhibited a nuclear and a cytoplasmic localization in each species. CONCLUSIONS: The corneal endothelium of rabbits stained positively for cyclins D, E, and A and did not stain for Ki67, suggesting that, as in humans, rabbit corneal endothelium in vivo is arrested in G1-phase upstream from Ki67 synthesis. Cyclin E was located in the cytoplasm of rabbit cells, whereas it was found in the nucleus in human endothelium. The apparent difference in cellular distribution of cyclin E in these two species may be significant because this cyclin is active during the G1-/S-phase transition. It is possible that in situ human and rabbit corneal endothelial cells are arrested at different points within G1-phase and/or that the difference in relative proliferative capacity exhibited by the corneal endothelium in these two species may be caused by differences in their relative ability to overcome G1-phase arrest.

Effects of EGF and indomethacin on rabbit corneal endothelial wound closure in excised corneas.

Corneal endothelial cells of rabbit corneas stored in M-K medium at 37 degrees C were wounded by touching them lightly with a micropipet under video specular microscope observation. Three groups were studied: control, with EGF, and with EGF + indomethacin. The wound closure process (initial wound area about 8500 microns 2) was observed and recorded with time-lapse videography for 6 hr. The recorded video images were digitized and computer assisted morphometric analysis was performed. (1) Addition of either EGF (10 ng/ml) + indomethacin (1 microM), or EGF (10 ng/ml) alone to the M-K medium statistically significantly shortened the wound closure time as compared with the control group. (2) Both EGF + indomethacin and EGF alone resulted in a greater average percent relative change of the shape factor, more than three times greater with EGF + indomethacin and more than two times greater with EGF alone, than in the control group 150 min after wounding. (3) The maximum cell shape change occurred at about 150 min after wounding in the groups EGF + indomethacin and EGF alone, and at about 200 min in the control group. After this time in all three groups the cells began to approach a normal shape. (4) The cells near the wound boundary moved faster in the EGF + indomethacin and the EGF groups as compared with the control group. These results suggest that EGF and indomethacin may be of therapeutic value in promoting closure of traumatized human corneal endothelium.

Corneal endothelial wound closure in vitro. Effects of EGF and/or indomethacin.

In response to stress, the corneal endothelium must maintain or region its barrier function. To study cellular responses of the corneal endothelium, our laboratory has developed an in vitro model of rabbit corneal endothelial wound closure. When cells are free to divide, a 3 mm diameter wound closes within 4 days. 5-fluorouracil added to these cultures does not affect the cellular morphology or ultrastructure, but does inhibit cell division. In the presence of 5-fluorouracil, wounds close in approximately 7 days. These conditions mimic the amitotic state and general behavior of adult human corneal endothelium in vivo. Using this model, we studied the effects of epidermal growth factor (EGF) and/or indomethacin treatment on corneal endothelial wound closure in mitotically competent and inhibited cultures. EGF appeared to stimulate migration, whereas indomethacin appeared to enhance cell spreading in response to wounding, particularly in mitotically inhibited cultures. Treatment with the above agents at the time of wounding had little effect on wound closure rates, but did affect closure patterns. In contrast, pretreatment of cultures, particularly with indomethacin, significantly accelerated closure in mitotically inhibited cultures. In the presence of indomethacin, wounds closed in 3-4 days compared to 7-8 days for controls. These results indicate that the response of corneal endothelial cells to wounding can be pharmacologically manipulated, and perhaps accelerated, and suggest that the treatment of the endothelium with nonsteroidal anti-inflammatory drugs or EGF-like growth factors may be clinically useful.

Protein kinase C regulation of corneal endothelial cell proliferation and cell cycle.

PURPOSE: The purpose of this study was to determine the role of protein kinase C (PKC) in corneal endothelial cell proliferation. METHODS: Immunocytochemistry and Western blotting were used to define the PKC isoforms expressed in primary cultures of rat corneal endothelial cells. For proliferation studies, primary cultures of rat corneal endothelial cells were serum-starved for 48 hours and incubated for 2 hours with the PKC inhibitors staurosporine (10(-9) to 10(-7) M), chelerythrine (10(-9) to 5 x 10(-8) M), or calphostin C (10(-9) to 10(-7) M). Individual PKC isoforms were inhibited using PKCalpha antisense oligonucleotide transfection or exposure for 1 hour to myristoylated, pseudosubstrate-derived peptide inhibitors against PKCalpha, -alphassgamma, -epsilon, and -delta (10(-8) to 10(-6) M). Cells were then stimulated with 2.5% serum for 24 hours. Cell proliferation was measured with bromodeoxyuridine (BrDU) and Ki67 immunocytochemistry. Protein level of cyclin E was determined by Western blotting. RESULTS: PKCalpha, -ssII, -delta, -epsilon, -iota, -eta, -gamma, and -theta were detected in corneal endothelial cells. Maximum inhibition of PKC with staurosporine, calphostin C, and chelerythrine reduced cell proliferation to 7%, 31%, and 48% of control, respectively. Myristoylated peptide inhibition of PKCalpha and -epsilon reduced cell proliferation to 57% and 59% of control, respectively. PKCalpha antisense oligonucleotide reduced cell proliferation to 35% of control. Cyclin E protein level was decreased to 70%, 38%, 57%, and 43% of control in cells treated with calphostin C, staurosporine, chelerythrine, and PKCalpha antisense, respectively. CONCLUSIONS: PKC activity, in particular PKCalpha and -epsilon activity, is important in promoting corneal endothelial cell proliferation. Inhibition of PKC activity prohibits G1/S-phase progression and reduces cyclin E protein levels.

Synchronization of the G1/S transition in response to corneal debridement.

PURPOSE: This study's intention was to examine the progression of ocular surface epithelium through the G1/S transition of the cell cycle after corneal epithelial debridement. METHODS: Three-millimeter debridements were made in central rat cornea and allowed to heal 4 to 48 hours in vivo. Unwounded contralateral eyes served as controls. Two hours before the animals were killed, 5-bromo-2-deoxyuridine (BrdU) was injected to detect S-phase cells. Incorporated BrdU was visualized by indirect immunofluorescence microscopy, and expression of G1 cell-cycle markers cyclins D and E was examined by indirect immunofluorescence and immunoblotting. RESULTS: The number of BrdU-labeled cells in conjunctival, limbal, and peripheral epithelium peaked at 28 hours after wounding (3.9-, 4.5-, and 3.2-fold increases, respectively). In unwounded eyes, cyclin D showed diffuse cytoplasmic localization with occasional basal cells exhibiting a nuclear localization, while anti-cyclin E showed intense localization in limbal and conjunctival basal cells but only minimal labeling in corneal epithelium. Within 8 to 12 hours after wounding, the nuclei of most corneal basal cells outside the wound area were bound intensely by anti-cyclins D and E. Immunoblotting revealed that cyclin D and E protein levels increased 4.5- and 12.1-fold after wounding, respectively. Epithelium migrating into the wound area did not incorporate BrdU and did not exhibit nuclear localization of cyclins D and E. CONCLUSIONS: Corneal epithelial debridement stimulates basal cells outside the wound area to synchronously enter the cell cycle. However, cells migrating to cover the wound area do not progress through the cell cycle. These data suggest a compartmentalization of the proliferative and migratory phases of wound repair.

TGF-beta receptor types I and II are differentially expressed during corneal epithelial wound repair.

PURPOSE: It has been demonstrated that cells migrating to cover an epithelial débridement wound exit the cell cycle and that the cell-cycle inhibitor p15(INK4b) is upregulated in these cells. TGF-beta signaling has been implicated in both of these processes, and this study was conducted to determine whether the expression and localization of TGF-beta receptor (TbetaR)-I and -II are altered during corneal epithelial wound repair. METHODS: Three-millimeter superficial keratectomy wounds and 3-mm débridement wounds were made in central rat cornea and allowed to heal in vivo for 1 to 48 hours. Immunofluorescence microscopy and Western blot analysis were used to determine the localization and expression of TbetaR-I and -II. Unwounded rat corneas served as control samples. To determine the effect of epidermal growth factor (EGF) and TGF-beta1 on p15(INK4b) and TbetaR-I and -II expression, human corneal epithelial cells were grown in culture to 50% to 60% confluence, and EGF (5 ng/ml) and/or TGF-beta1 (2 ng/ml) were added for 6 hours. Cells were harvested and p15(INK4b) and TBR-I and -II levels were assayed by using Western blot analysis. RESULTS: In unwounded corneas, TbetaR-I and TbetaR-II were present at low levels across the cornea, with higher levels in limbal epithelium. Both TbetaR-I and -II were upregulated after wounding. However, levels of TbetaR-II appeared to increase in the epithelial cells that had migrated to cover the wound area, whereas TbetaR-I was upregulated in the entire corneal epithelium. Western blot analysis indicated that both TbetaR-I and -II were upregulated threefold after wounding. In cultured cells, EGF and TGF-beta1 stimulated TbetaR-II; however, neither one stimulated TbetaR-I expression. TGF-beta1 stimulated p15(INK4b) protein levels threefold. CONCLUSIONS: After wounding, TbetaR-I and TbetaR-II were both expressed at high levels in cells migrating to cover a corneal wound, suggesting that TGF-beta signaling is involved in blocking migrating cells from progressing through the cell cycle. This blockage, at least in part, involves the inhibitor p15(INK4b). In addition, although both TbetaR-I and TbetaR-II are upregulated during wound repair, they appear to be differentially regulated.

Expression patterns of retinoblastoma and E2F family proteins during corneal development.

PURPOSE: To determine the expression patterns of the retinoblastoma protein and the E2F transcription factor families in limbal and corneal epithelia and in corneal keratocytes in situ during corneal development and differentiation. METHODS: Retinoblastoma protein (pRb) and its family members p107 and p130; E2F-1, -2, and -4, members of the E2F family of transcription factors; and Ki67, a marker of actively cycling cells, were localized by indirect immunofluorescence microscopy, in corneas of neonatal, juvenile, and adult rats. Presence of mRNA for pRb, p107, p130, and E2F types 1 to 5 in adult corneal epithelium was determined by reverse transcription-polymerase chain reaction. RESULTS: mRNA for all members of pRb and E2F families was present in adult corneal epithelium. The greatest number of Ki67-positive corneal and limbal epithelial cells were present at days 13 to 19, and Ki67-positive stromal keratocytes at day 2. pRb and E2F-2 were localized to all cells in neonatal, juvenile, and adult corneas. With age, p130 localization became more intense and nuclear in stromal keratocytes and suprabasal cells of corneal and limbal epithelia; p107, initially nuclear in limbal and corneal epithelia, became increasingly cytoplasmic in corneal epithelium. E2F-1 was initially nuclear in keratocytes and diminished after day 10. E2F-1 was localized in the basal cell layer of limbal and corneal epithelia after day 10. E2F4 was always nuclear in limbal epithelium and cytoplasmic in corneal epithelium. CONCLUSIONS: Expression patterns of pRb and E2F family proteins vary with corneal cell differentiation, but are most apparent with p130 and p107. Nuclear localization of p130 appears to correlate with terminal differentiation in epithelium and entrance into a quiescent state by keratocytes. In contrast, p107 is nuclear in the undifferentiated limbal basal cells and is cytoplasmic in the remainder of the corneal epithelial cells.