Person:

Harris, Deshea L.

Purpose: To test the feasibility of altering the phenotype of umbilical cord blood mesenchymal stem cells (UCB MSCs) toward that of human corneal endothelial cells (HCEC) and to determine whether UCB MSCs can “home” to sites of corneal endothelial cell injury using an ex vivo corneal wound model. Methods: RNA was isolated and purified from UCB MSCs and HCECs. Baseline information regarding the relative gene expression of UCB MSCs and HCEC was obtained by microarray analysis. Quantitative real-time PCR (q-PCR) verified the microarray findings for a subset of genes. The ability of different culture media to direct UCB MSCs toward a more HCEC-like phenotype was tested in both tissue culture and ex vivo corneal endothelial wound models using three different media: MSC basal medium (MSCBM), a basal medium used to culture lens epithelial cells (LECBM), or lens epithelial cell-conditioned medium (LECCM). Morphology of the MSCs was observed by phase-contrast microscopy or by light microscopic observation of crystal violet-stained cells. Immunolocalization of the junction-associated proteins, zonula occludins-1 (ZO1) and N-cadherin, was visualized by fluorescence confocal microscopy. Formation of cell-cell junctions was tested by treatment with the calcium chelator, EGTA. A second microarray analysis compared gene expression between UCB MSCs grown in LECBM and LECCM to identify changes induced by the lens epithelial cell-conditioned culture medium. The ability of UCB MSCs to “home” to areas of endothelial injury was determined using ZO1 immunolocalization patterns in ex vivo corneal endothelial wounds. Results: Baseline microarray analysis provided information regarding relative gene expression in UCB MSCs and HCECs. MSCs attached to damaged, but not intact, corneal endothelium in ex vivo corneal wounds. The morphology of MSCs was consistently altered when cells were grown in the presence of LECCM. In tissue culture and in ex vivo corneal wounds, UCB MSC treated with LECCM were elongated and formed parallel sheets of closely apposed cells. In both tissue culture and ex vivo corneal endothelial wounds, ZO1 and N-cadherin localized mainly to the cytoplasm of UCB MSCs in the presence of MSCBM. However, both proteins localized to cell borders when UCB MSCs were grown in either LECBM or LECCM. This localization was lost when extracellular calcium levels were reduced by treatment with EGTA. A second microarray analysis showed that, when UCB MSCs were grown in LECCM instead of LECBM, the relative expression of a subset of genes markedly differed, suggestive of a more HCEC-like phenotype. Conclusions: Results indicate that UCB MSCs are able to “home” to areas of injured corneal endothelium and that the phenotype of UCB MSCs can be altered toward that of HCEC-like cells. Further study is needed to identify the specific microenvironmental conditions that would permit tissue engineering of UCB MSCs to replace damaged or diseased corneal endothelium.

To study bilateral nerve changes in a newly developed novel mouse model for neurotrophic keratopathy by approaching the trigeminal nerve from the lateral fornix. Surgical axotomy of the ciliary nerve of the trigeminal nerve was performed in adult BALB/c mice at the posterior sclera. Axotomized, contralateral, and sham-treated corneas were excised on post-operative days 1, 3, 5, 7 and 14 and immunofluorescence histochemistry was performed with anti-β-tubulin antibody to evaluate corneal nerve density. Blink reflex was evaluated using a nylon thread. The survival rate was 100% with minimal bleeding during axotomy and a surgical time of 8±0.5 minutes. The blink reflex was diminished at day 1 after axotomy, but remained intact in the contralateral eyes in all mice. The central and peripheral subbasal nerves were not detectable in the axotomized cornea at day 1 (p<0.001), compared to normal eyes (101.3±14.8 and 69.7±12.0 mm/mm2 centrally and peripherally). Interestingly, the subbasal nerve density in the contralateral non-surgical eyes also decreased significantly to 62.4±2.8 mm/mm2 in the center from day 1 (p<0.001), but did not change in the periphery (77.3±11.7 mm/mm2, P = 0.819). Our novel trigeminal axotomy mouse model is highly effective, less invasive, rapid, and has a high survival rate, demonstrating immediate loss of subbasal nerves in axotomized eyes and decreased subbasal nerves in contralateral eyes after unilateral axotomy. This model will allow investigating the effects of corneal nerve damage and serves as a new model for neurotrophic keratopathy.

Purpose: Human corneal endothelial cells (HCEC), particularly from older donors, only proliferate weakly in response to EGF. The protein tyrosine phosphatase, PTP1B, is known to negatively regulate EGF-induced signaling in several cell types by dephosphorylating the epidermal growth factor receptor (EGFR). The current studies were conducted to determine whether PTP1B plays a role in regulating cell cycle entry in HCEC in response to EGF stimulation. Methods: Donor corneas were obtained from the National Disease Research Interchange and accepted for study based on established exclusion criteria. PTP1B was localized in the endothelium of ex vivo corneas and in cultured cells by immunocytochemistry. Western blot analysis verified PTP1B protein expression in HCEC and then compared the relative expression of EGFR and PTP1B in HCEC from young (<3 years old) and older donors (>60 years old). The effect of inhibiting the activity of PTP1B on S-phase entry was tested by comparing time-dependent BrdU incorporation in subconfluent HCEC incubated in the presence or absence of the PTP1B inhibitor, CinnGEL 2Me, before EGF stimulation. Results: PTP1B was localized in a punctate pattern mainly within the cytoplasm of HCEC in ex vivo corneas and cultured cells. Western blots revealed the presence of three PTP1B-positive bands in HCEC and the control. Further western blot analysis showed no significant age-related difference in expression of EGFR (p=0.444>0.05); however, PTP1B expression was significantly higher in HCEC from older donors (p=0.024<0.05). Pre-incubation of HCEC with the PTP1B inhibitor significantly increased (p=0.019<0.05) the number of BrdU positive cells by 48 h after EGF stimulation. Conclusions: Both immunolocalization and western blot studies confirmed that PTP1B is expressed in HCEC. Staining patterns strongly suggest that at least a subset of PTP1B is localized to the cytoplasm and most likely to the endoplasmic reticulum, the known site of EGFR/PTP1B interaction following EGF stimulation. PTP1B expression, but not EGFR expression, was elevated in HCEC from older donors, suggesting that the reduced proliferative activity of these cells in response to EGF is due, at least in part, to increased PTP1B activity. The fact that inhibition of PTP1B increased the relative number of cells entering S-phase strongly suggests that PTP1B helps negatively regulate EGF-stimulated cell cycle entry in HCEC. These results also suggest that it may be possible to increase the proliferative activity of HCEC, particularly in cells from older donors, by inhibiting the activity of this important protein tyrosine phosphatase.

The cornea is the shield to the foreign world and thus, a primary site for peripheral infections. However, transparency and vision are incompatible with inflammation and scarring that may result from infections. Thus, the cornea is required to perform a delicate balance between fighting infections and preserving vision. To date, little is known about the specific role of antigen-presenting cells in viral keratitis. In this study, utilizing an established murine model of primary acute herpes simplex virus (HSV)-1 keratitis, we demonstrate that primary HSV keratitis results in increased conventional dendritic cells (cDCs) and macrophages within 24 hours after infection. Local depletion of cDCs in CD11c-DTR mice by subconjuntival diphtheria toxin injections, led to increased viral proliferation, and influx of inflammatory cells, resulting in increased scarring and clinical keratitis. In addition, while HSV infection resulted in significant corneal nerve destruction, local depletion of cDCs resulted in a much more severe loss of corneal nerves. Further, local cDC depletion resulted in decreased corneal nerve infection, and subsequently decreased and delayed systemic viral transmission in the trigeminal ganglion and draining lymph node, resulting in decreased mortality of mice. In contrast, sham depletion or depletion of macrophages through local injection of clodronate liposomes had neither a significant impact on the cornea, nor an effect on systemic viral transmission. In conclusion, we demonstrate that corneal cDCs may play a primary role in local corneal defense during viral keratitis and preserve vision, at the cost of inducing systemic viral dissemination, leading to increased mortality.