On the influence of fixations on the normal hydration of rabbit cornea and the total amount of acid mucopolysaccharides in the stroma

Summary The authors studied the influence of fixations on the normal hydration of the rabbit cornea and the total amount of acid mucopolysaccharides (AMPS) in the stroma. The following fixatives were used: formol-calcium chloride at 19° C for 24 hours, formolcetylpyridinium chloride (CPC) at 19 and 28° C for 48 hours, Lillie's fixation at 19° C for 24 hours and Carnoy's fluid at 19° C for 30 minutes. The sections of the corneae were stained with Alcian blue, colloidal Fe³⁺ in the modification according to Rinehart and Abu'l Haj and with toluidine and methylene blue. The amount of AMPS was determined with the method of Rondle and Morgan and the total hydration of the stroma by weighing the corneae before and after using different fixative fluids and by calculation of obtained values on dry weight.The best results were obtained by using formol-CPC at 28° C. At the ordinary room temperature (±19° C) it was the poorest fixation, however, as the corneae in this solution became hydrated. Formol-calcium chloride was the second in the row and it was much better than Lillie's and Carnoy's fluid.The amount of AMPS in the stroma was not essentially changed by the effect of fixatives. Within 24–48 hours formol-CPC at 28° C retained the normal content and formol-calcium chloride caused the 11% decrease of AMPS maximally. The loss of AMPS after other fixatives was minimal.The intensity of staining with cationic dyes in paraffin sections was different after individual fixatives and after the kind of their application and was dependent chiefly on the state of hydration of the corneal stroma: It is impossible to interpret the results of staining reactions in terms of the quantity of AMPS as it was hitherto done.     

The effect of hypertonic saline on human corneal hydration.

Hydrophilic gel contact lenses, presoaked in various strength saline solutions, were held in intimate contact with the in vivo human cornea. A change in corneal water content was observed taking place against the induced osmotic gradient. Direct stimulation of some aspect of the mechanisms controlling corneal hydration is postulated, in a manner which may be similar to that previously reported for the in vitro rabbit cornea.     

Alkali burns of the rabbit cornea. II. A histochemical study of glycosaminoglycans.

In alkali burned rabbit cornea the stainability of glycosaminoglycans in cold microtome setions was investigated. Staining by Alcian blue in 3% acetic acid, Alcian blue in various MgCl2 concentration and toluidine blue (pH 4.5) was employed. From the 1st to the 4th experimental day the intensity of reactions was decreased. This is most probably due to an increased hydration of the corneal stroma. On the 7th day hydration was markedly suppressed and reached nearly the normal level. In this time interval a decreased stainability of glycosaminoglycans was seen accompanied by a complete loss of staining in the marginal zone. On the 14th day the stainability in the traumatized area began to restore and in the marginal zone appeared. On the 32nd day the staining intensity of both areas was normalised, however when lower concentrations of MgCl2 were used; in the presence of higher concentrations of MgCl2 the decreased staining intensity persisted and points to a lower sulfatation of glycosaminoglycans. This was particularly remarkable in the area bordering the injured zone. This decrease runs parallel to the increased activities of acid glycosidases (especially of acid beta-galactosidase) which were reported previously.     

Dipeptidyl peptidase IV (DPPIV) activity in the tear fluid as an indicator of the severity of corneal injury: a histochemical and biochemical study

Comparative histochemical and biochemical studies on the catalytically active protease Dipeptidyl peptidase IV (DPPIV), have been performed in the rabbit cornea and the tear fluid using a sensitive fluorogenic substrate, Gly-Pro-7-amino-4-Trifluoromethyl Coumarine (AFC). In both normal and experimentally injured corneas, DPPIV activity was detected histochemically and in the tear fluid biochemically. In contrast to the normal cornea where DPPIV activity was absent and in the tear fluid where it was low, during continuous wearing of contact lenses or repeated irradiation of the cornea with UVB rays, slight DPPIV activity appeared first in the superficial layers of the corneal epithelium, while later increased activity was present in the whole epithelium. This paralleled elevated DPPIV activity in the tear fluid. Moreover, during continuous contact lens wear, the increased DPPIV activity in the tear fluid was, in many cases, coincidental with the presence of capillaries in the limbal part of the corneal stroma. After severe alkali burns when corneal ulcers appeared, collagen fragments were active for DPPIV, which was associated with high DPPIV activity in the tear fluid. In conclusion, Gly-Pro-AFC was found to be useful for comparative histochemical and biochemical studies on DPPIV activity in the experimentally injured rabbit eye. Using the method of the tear film collection by a short touch of substrate punches to the respective site of the cornea or conjunctiva we can show that in experimental injuries (wearing of contact lenses, irradiation of the cornea with UVB rays), the damaged corneal cells were the main source for DPPIV activity in the tear fluid. It is suggested that the activity of DPPIV measured in the tear fluid might serve as an indicator of early corneal disorders, e.g. corneal vascularization related to contact lens wear.     

STUDIES ON THE CORNEA : I. The Fine Structure of the Rabbit Cornea and the Uptake and Transport of Colloidal Particles by the Cornea in Vivo

Physiological studies have demonstrated that ions, as well as large molecules such as hemoglobin or fluorescein, can diffuse across and within the cornea. Most of the substrates for corneal metabolism are obtained from aqueous humor filling the anterior chamber. In order to receive its nutrients and in order to maintain its normal conditions of hydration, the avascular cornea must transport relatively large amounts of solute and solvent across the cellular layers which cover this structure. It has been suggested in the past that there may be a morphological basis for the transport of large amounts of solvents and solutes by cells by the mechanism of pinocytosis. The use of electron-opaque markers to study fluid movements at the electron microscope magnification level was described by Wissig (29). The present study describes the fine structure of the normal rabbit cornea and the pathways of transport of colloidal particles by the cornea in vivo. Rabbit corneas were exposed in vivo to suspensions of saccharated iron oxide, thorium dioxide, or ferritin by injection of the material into the anterior chamber. In other experiments thorium dioxide or saccharated iron oxide was injected into the corneal stroma, producing a small bleb. Particles presented at the aqueous humor surface of the rabbit corneal endothelium are first attached to the cell surface and then pinocytosed. It appears that the particles are carried around the terminal bar by an intracellular pathway involving the pinocytosis of the particles and their subsequent transport in vesicles to the lateral cell margin basal to the terminal bar. Particles introduced at the basal surface of the endothelium (via blebs in the corneal stroma) are apparently carried through the endothelial cells in membrane-bounded vesicles without appearing in the intercellular space. There appears to be free diffusion of these particles through Descemet's membrane and the corneal stroma. The stromal cells take up large quantities of the particles when blebs are injected into the stroma.     

Survival and integration of tissue-engineered corneal stroma in a model of corneal ulcer

Tissue-engineered replacement of diseased or damaged tissue has become a reality for some types of tissue, such as skin and cartilage. Tissue-engineered corneal stroma represents a promising concept to overcome the limitations of cornea replacement with allograft. In this study, porcine cornea was decellularized by a series of extraction methods, and the in vivo biocompatibility of the scaffold was measured subcutaneously in rabbits (n = 8). These were not acutely rejected and no abscesses were observed by hematoxylin and eosin staining at the 8th week, indicating that the scaffolds had good biocompatibility. To investigate the potential value of clinical applications, rabbit stromal keratocytes were implanted onto decellularized scaffolds to fabricate tissue-engineered corneal stroma. Allograft, tissue-engineered corneal stroma, or scaffolds were implanted into a model of corneal ulcer. The survival and reconstruction of corneal transplantation were morphologically evaluated by light and electron microscopy until the 32nd week after implantation. Experiments involving transplantation indicated that the epithelial and stromal defect healed quickly, with improvement in corneal clarity. The integration of the graft was accompanied by neurite ingrowth from the host tissue. By 16 weeks after transplantation, the cornea had gradually regained an intact state similar to that of normal cornea. Our results demonstrate that the tissue-engineered corneal stroma with allogenetic cells is a promising therapeutic method for corneal injury. This study was supported by the Nature Science Foundation of China (project no. 30572046) and the Development of High and New Science and Technology (863 Project) of China (2002AA205041, 2005AA205241).     

Fibronectin promotes epithelial migration of cultured rabbit cornea in situ

We investigated the effect of fibronectin on epithelial migration onto the stroma in cultured rabbit cornea. Rabbit plasma fibronectin was purified by affinity chromatography using gelatin-Sepharose 4B, and its purity was confirmed by SDS polyacrylamide slab gel electrophoresis. Antibody against rabbit plasma fibronectin raised in guinea pigs formed a single precipitin line against rabbit plasma and purified rabbit plasma fibronectin by Ouchterlony double diffusion test. When rabbit cornea was cut into small blocks and cultured in TCM-199 medium alone, corneal epithelial cells began to migrate on the cut edge of the corneal stroma. The addition of purified rabbit plasma fibronectin to the culture medium significantly enhanced epithelial migration. The degree of enhancement depended on the amount of fibronectin added. When guinea pig IgG anti-rabbit plasma fibronectin was added, epithelial migration was significantly inhibited when compared with that in control cultured corneal blocks. The results demonstrate that fibronectin promotes epithelial migration in the cornea and thus plays an important role in corneal wound healing.     

Ionic basis for color changes in the iridescent cornea of the sand goby (Pomatoschistus minutus).

The iridescence from the cornea of the sand goby (Pomatoschistus minutus) occurs because of thin layer interference from the platelet-like cells in the stroma. It is suggested that ionic pumps across the epithelium control the water content in the stroma and thus the spectral reflection. A saline was perfused over goby eyes and simple ion manipulation was carried out to observe any changes in the iridescent characteristics. It was found that removal of Cl- and K+ ions reduced the peak reflected wavelength to the blue end of the spectrum, whereas Na+ had little effect. The removal of K+ also caused a dramatic change to the normal shift in reflected spectral intensity. The iridescence was also found to be sensitive to pH, and the buffer HEPES was detrimental to the cornea compared to controls. These results suggest similarities to amphibian and mammalian corneal hydration control.     

Genetics of corneal endothelial dystrophies

The corneal endothelium maintains the level of hydration in the cornea. Dysfunction of the endothelium results in excess accumulation of water in the corneal stroma, leading to swelling of the stroma and loss of transparency. There are four different corneal endothelial dystrophies that are hereditary, progressive, non-inflammatory disorders involving dysfunction of the corneal endothelium. Each of the endothelial dystrophies is genetically heterogeneous with different modes of transmission and/or different genes involved in each subtype. Genes responsible for disease have been identified for only a subset of corneal endothelial dystrophies. Knowledge of genes involved and their function in the corneal endothelium can aid understanding the pathogenesis of the disorder as well as reveal pathways that are important for normal functioning of the endothelium.     

Ultrastructural localization of sulfated and unsulfated keratan sulfate in normal and macular corneal dystrophy type I

Keratan sulfate (KS) proteoglycans are of importance for the maintenance of corneal transparency as evidenced in the condition macular corneal dystrophy type I (MCD I), a disorder involving the absence of KS sulfation, in which the cornea becomes opaque. In this transmission electron microscope study quantitative immuno- and histochemical methods have been used to examine a normal and MCD I cornea. The monoclonal antibody, 5-D-4, has been used to localize sulfated KS and the lectin Erythrina cristagalli agglutinin (ECA) to localize poly N -acetyllactosamine (unsulfated KS). In normal cornea high levels of sulfated KS were detected in the stroma, Bowman's layer, and Descemet's membrane and low levels in the keratocytes, epithelium and endothelium. Furthermore, in normal cornea, negligible levels of labeling were found for N -acetyllactosamine (unsulfated KS). In the MCD I cornea sulfated KS was not detected anywhere, but a specific distribution of N -acetyllactosamine (unsulfated KS) was evident: deposits found in the stroma, keratocytes, and endothelium labeled heavily as did the disrupted posterior region of Descemet's membrane. However, the actual cytoplasm of cells and the undisrupted regions of stroma revealed low levels of labeling. In conclusion, little or no unsulfated KS is present in normal cornea, but in MCD I cornea the abnormal unsulfated KS was localized in deposits and did not associate with the collagen fibrils of the corneal stroma. This study has also shown that ECA is an effective probe for unsulfated KS.     

Disturbances in the rabbit cornea after short-term and long-term wear of hydrogel contact lenses

Summary The influence of soft contact lenses (SCL) with low (37%, L) and high (65%, H) water content on rabbit corneas was investigated. The lenses were worn continuously for 1, 2, 4, 7, 10, 14, 21 or 28 days. The changes in corneal transparency, hydration and enzyme activities were studied. A slight change in corneal transparency due to higher hydration caused by a decreased activity of Na⁺–K⁺-dependent adenosine triphosphatase (Na⁺–K⁺-ATPase) in the corneal endothelium is followed by a decrease in the activity of -glutamyl transferase (GGT). Slight morphological disturbances appear within 4 days in animals wearing SCL (L). SCL (H) produce similar changes one week later. Subsequently, the corneal epithelium becomes thinner and changes in the size of corneal endothelial cells are obvious. Disturbances of enzyme activities in cells of all corneal layers are present. In the epithelium highly increased activities of acid glycosidases, acid phosphatase, and dipeptidyl peptidase I and II, in keratocytes decreased activities of alkaline phosphatase and GGT, and in the endothelium decreased activity of Na⁺–K⁺-ATPase and GGT were found. These changes are more severe after SCL (L). In this case, inflammatory cells displaying high activities of lysosomal hydrolases appear in the anterior part of the stroma during the 3rd and 4th weeks and local degradation of glycosaminoglycans and proteins takes place. In contrast, after SCL (H) a remarkable thinning of the corneas was observed during extended wear, accompanied by decreased stainability of stromal glycosaminoglycans and highly decreased enzyme activities in keratocytes. The histochemical methods proved very useful in the assessment of tesions caused by a continuous wear of SCL.     

Neutron diffraction studies of the corneal stroma

Low-angle neutron diffraction patterns have been obtained from demembranated beef corneal stroma using the D11 camera available at the Institut Laue-Langevin, Grenoble. These diffraction patterns show three peaks: the first peak corresponds to the interfibrillar distance in the stroma and the others are the third and fifth orders of the collagen spacing. The position of the interfibrillar peak as a function of the hydration of the tissue and the swelling properties of the stroma have been studied at different pH values. The results suggest that in swollen cornea there may be a system of mutually repelling cylinders that expands uniformly to fill the space made available by the water present, and there seems little possibility for fibril cross-linking.     

A triphasic analysis of corneal swelling and hydration control.

Physiological studies strongly support the view that hydration control in the cornea is dependent on active ion transport at the corneal endothelium. However, the mechanism by which endothelial ion transport regulates corneal thickness has not been elaborated in detail. In this study, the corneal stroma is modeled as a triphasic material under steady-state conditions. An ion flux boundary condition is developed to represent active transport at the endothelium. The equations are solved in cylindrical coordinates for confined compression and in spherical coordinates to represent an intact cornea. The model provides a mechanism by which active ion transport at the endothelium regulates corneal hydration and provides a basis for explaining the origin of the "imbibition pressure" and stromal "swelling pressure." The model encapsulates the Donnan view of corneal swelling as well as the "pump-leak hypothesis."     

Efficacious and safe tissue-selective controlled gene therapy approaches for the cornea

Untargeted and uncontrolled gene delivery is a major cause of gene therapy failure. This study aimed to define efficient and safe tissue-selective targeted gene therapy approaches for delivering genes into keratocytes of the cornea in vivo using a normal or diseased rabbit model. New Zealand White rabbits, adeno-associated virus serotype 5 (AAV5), and a minimally invasive hair-dryer based vector-delivery technique were used. Fifty microliters of AAV5 titer (6.5×10(12) vg/ml) expressing green fluorescent protein gene (GFP) was topically applied onto normal or diseased (fibrotic or neovascularized) rabbit corneas for 2-minutes with a custom vector-delivery technique. Corneal fibrosis and neovascularization in rabbit eyes were induced with photorefractive keratectomy using excimer laser and VEGF (630 ng) using micropocket assay, respectively. Slit-lamp biomicroscopy and immunocytochemistry were used to confirm fibrosis and neovascularization in rabbit corneas. The levels, location and duration of delivered-GFP gene expression in the rabbit stroma were measured with immunocytochemistry and/or western blotting. Slot-blot measured delivered-GFP gene copy number. Confocal microscopy performed in whole-mounts of cornea and thick corneal sections determined geometric and spatial localization of delivered-GFP in three-dimensional arrangement. AAV5 toxicity and safety were evaluated with clinical eye exam, stereomicroscopy, slit-lamp biomicroscopy, and H&E staining. A single 2-minute AAV5 topical application via custom delivery-technique efficiently and selectively transduced keratocytes in the anterior stroma of normal and diseased rabbit corneas as evident from immunocytochemistry and confocal microscopy. Transgene expression was first detected at day 3, peaked at day 7, and was maintained up to 16 weeks (longest tested time point). Clinical and slit-lamp eye examination in live rabbits and H&E staining did not reveal any significant changes between AAV5-treated and untreated control corneas. These findings suggest that defined gene therapy approaches are safe for delivering genes into keratocytes in vivo and has potential for treating corneal disorders in human patients.     

The mechanical properties of the rabbit and human cornea

The extensibility of rabbit and human corneas was measured by raising the pressure within the intact globe of the eye and measuring the displacements of two very small mercury drops on the corneal surface. The human cornea showed a negligible extensibility under low stresses. The rabbit tissue, however, underwent a 9% strain under low pressures with a curvilinear relationship between stress and strain. At higher pressures the relationship was linear, and the tissue showed some creep. The low pressure stress-strain relationship of the rabbit could not be explained on the basis that the collagen fibrils were being straightened out from an initial set in a sinusoidal wave. When the stroma was isolated from Descemet's membrane, it showed a negligible low pressure extensibility in rabbit and man. On the other hand, isolated Descemet's membrane was very extensible in both species. The difference between them in the behavior of the intact cornea seems to lie in the relative initial strain in the stroma and Descemet's membrane.     

Ocular distribution of keratan sulfates during pre- and postnatal development in rabbits

Twenty-six monoclonal antibodies (MAbs) developed against rabbit corneal proteokeratan sulfate (PKS), were used to evaluate immunohistochemically the ocular distribution of PKS during prenatal and early postnatal development in rabbits. These MAbs were directed against epitopes located in the keratan sulfate (KS) chains of the proteoglycan (SundarRaj et al., 1985). Staining of cryostat sections of the eyes was carried out using an indirect peroxidase-conjugated technique. Only one of the MAbs reacted with the presumptive corneal region at day 13 or 16 of fetal development. By day 20, more MAbs reacted with the corneal stroma. There were distinct differences, however, in the distribution of the epitopes recognized by the various MAbs. A few of them stained only the posterior region of the cornea, whereas others showed a decreasing staining gradient from the posterior to the anterior region. By day 24, all of the MAbs reacted with the corneal stroma, but some reacted also with the limbal region and with the conjunctival stromal matrix. One MAb also reacted with the conjunctival epithelial layer, but only at this stage of development. Conjunctival staining was more intense at day 28 of fetal development and at day 2 postnatally. KS was not detectable in the conjunctiva of adult rabbits with any of the MABs. These results suggest that although KS synthesis starts at very early stages of fetal development, there are progressive changes in its antigenic structure in specific regions of the cornea and conjunctiva during corneal development.     

Proteoglycan distribution in developing rabbit cornea

We used a staining procedure specific for sulfated glycosaminoglycans, cuprolinic blue dye (CBD), and immunohistochemical techniques to determine the histological distribution and ultrastructural organization of proteoglycans in developing rabbit cornea. We found several types of CBD-stained structures located throughout the corneal stroma, indicative of the distribution and perhaps the chemical heterogeneity of proteoglycans in this tissue. Keratan sulfate-specific immunohistochemical evidence supports our cytochemical findings. Our results suggest that low-sulfated keratan sulfate proteoglycans are found throughout most of the developing stroma, with the exception of the posterior margin of this tissue. Highly sulfated keratan sulfate proteoglycans in young fetal corneas, initially restricted to the subepithelial stroma, progressively extend to deeper portions of the stroma with development. Dermatan sulfate proteoglycans are located throughout the stroma, including the posterior margin. Invoking a recently published "oxygen-lack hypothesis" and correlating the tissue location of proteoglycans with the source of oxygen, we hypothesize that the distribution of proteoglycans in the developing rabbit cornea is related to the selective synthesis of keratan sulfate glycosaminoglycans under hypoxic conditions.     

Staining of rabbit eosinophil and pseudoeosinophil leukocytes.

Air-dried rabbit blood was stained by HE, PAS and a modification of the Undritz II method. Eosin stained granules red in the eosinophil leukocytes. PAS was negative and the modified Undritz method failed to give consistent results. Cells with eosinophilic granules appeared in the corneal stroma 1 h after removing the corneal epithelium. They were stained red consistently by both eosin and the modified Undritz II method. Electron micrographs failed to demonstrate crystalloids in the granules. Because of the staining characteristics and the lack of crystalloids in their granules these cells were classified as pseudoeosinophil leukocytes. The electron micrographs showed some glycogen 12 h after denuding the cornea, however, glycogen was not well stained by PAS until 18 h after denuding.     

Application of catastrophe theory to corneal swelling

Stromal swelling in human, cat, and rabbit cornea is biphasic, interpretable as an elementary cusp catastrophe proposed by Thom, with t* = log t and Q* = log Q (stromal charge Q, time t) as control parameters, and H0.5 (hydration H) as the state variable. A thermodynamic potential with two attractor regions, each with a local minimum, governs corneal stromal swelling. Transitions follow a 'saturation convention' whereby the second minimum is preferred upon availability. Corneal swelling is an example of a space-equivalent unfolding, where the transition plane moves in time. It is proposed that the transition plane coincides with the uncoupling of interfibrillary linkages or 'springs' in the corneal stroma, and is associated with a critical hydration of ca. 10 kg H2O per kilogram dry mass, and stromal charge ca. 1 x 10(-7) mol electrons.