A Bernard

3D map of the human corneal endothelial cell.

Corneal endothelial cells (CECs) are terminally differentiated cells, specialized in regulating corneal hydration and transparency. They are highly polarized flat cells that separate the cornea from the aqueous humor. Their apical surface, in contact with aqueous humor is hexagonal, whereas their basal surface is irregular. We characterized the structure of human CECs in 3D using confocal microscopy of immunostained whole corneas in which cells and their interrelationships remain intact. Hexagonality of the apical surface was maintained by the interaction between tight junctions and a submembraneous network of actomyosin, braced like a drum. Lateral membranes, which support enzymatic pumps, presented complex expansions resembling interdigitated foot processes at the basal surface. Using computer-aided design and drafting software, we obtained a first simplified 3D model of CECs. By comparing their expression with those in epithelial, stromal and trabecular corneal cells, we selected 9 structural or functional proteins for which 3D patterns were specific to CECs. This first 3D map aids our understanding of the morphologic and functional specificity of CECs and could be used as a reference for characterizing future cell therapy products destined to treat endothelial dysfunctions.

Comparison of endothelial cell density of organ cultured corneas with cornea donor study.

Purpose: Determination of the endothelial cell density (ECD) by eye banks is paramount in donor cornea qualification. Unbiased measurement avoids wastage and grafts with an increased risk of premature failure. Internal calibration of the counting method is essential, but external validation would add an extra stage in the assessment of reliability. In this respect, data published by the multicenter Cornea Donor Study (CDS) in 2005 is a reference. The aim of the study was to compare ECD determined within a single eye bank, which uses calibrated image analysis software designed for transmitted light microscopy images of organ cultured corneas, with the CDS data determined on specular microscopy images of corneas stored at 4°C. Methods: ECD of consecutive corneas retrieved between 2005 and 2013 was determined after exposure to 0.9% NaCl. More than 300 ECs were counted on 3 fields of the central 8 mm. Endothelial cell boundaries were automatically drawn and verified by a skilled technician who performed all necessary corrections. Results: Three thousand fifty-two corneas were analyzed, of which 48.5% donors were >75 years (CDS upper age limit). Between 10 and 75 years, the ECD varied according to donor age exactly in the same manner as in the CDS, but were consistently higher of 100 ± 25 cells per square millimeter (P < 0.001). Conclusions: ECD determined by a computer-aided method from transmitted light microscopy images compares favorably with the American CDS reference series. The slight systematic difference on either side of the Atlantic Ocean could be due to (1) differences in counting principles and/or (2) higher shrinkage of the cornea caused by stromal edema in organ culture.

Delivery of molecules into human corneal endothelial cells by carbon nanoparticles activated by femtosecond laser

Corneal endothelial cells (CECs) form a monolayer at the innermost face of the cornea and are the engine of corneal transparency. Nevertheless, they are a vulnerable population incapable of regeneration in humans, and their diseases are responsible for one third of corneal grafts performed worldwide. Donor corneas are stored in eye banks for security and quality controls, then delivered to surgeons. This period could allow specific interventions to modify the characteristics of CECs in order to increase their proliferative capacity, increase their resistance to apoptosis, or release immunosuppressive molecules. Delivery of molecules specifically into CECs during storage would therefore open up new therapeutic perspectives. For clinical applications, physical methods have a more favorable individual and general benefit/risk ratio than most biological vectors, but are often less efficient. The delivery of molecules into cells by carbon nanoparticles activated by femtosecond laser pulses is a promising recent technique developed on non-adherent cells. The nanoparticles are partly consummated by the reaction releasing CO and H2 gas bubbles responsible for the shockwave at the origin of cell transient permeation. Our aim was to develop an experimental setting to deliver a small molecule (calcein) into the monolayer of adherent CECs. We confirmed that increased laser fluence and time exposure increased uptake efficiency while keeping cell mortality below 5%. We optimized the area covered by the laser beam by using a motorized stage allowing homogeneous scanning of the cell culture surface using a spiral path. Calcein uptake reached median efficiency of 54.5% (range 50.3–57.3) of CECs with low mortality (0.5%, range (0.55–1.0)). After sorting by flow cytometry, CECs having uptaken calcein remained viable and presented normal morphological characteristics. Delivery of molecules into CECs by carbon nanoparticles activated by femtosecond laser could prove useful for future cell or tissue therapy.

Delivery of macromolecules into the endothelium of whole ex vivo human cornea by femtosecond laser-activated carbon nanoparticles

Background The targeted delivery of drugs or genes into corneal endothelial cells (ECs) during eye banking could help improve graft quality and quantity. Physical methods raising less safety concerns than viral ones, we previously adapted, for in vitro ECs, a recent innovative technique of drug delivery based on the activation of carbon nanoparticles (CNPs) by a femtosecond laser (fsL). The aim of the present pilot study was to adapt this method to enable molecule delivery into the intact endothelium of ex vivo human corneas. Methods ECs from 40 organ-cultured corneas were perforated by photoacoustic reaction induced by irradiation of CNPs by a fsL. This enabled intracellular delivery of Alexa Fluor 488 dextran, a 4000 Da fluorescent macromolecule. The influence of increasing laser fluences (15, 20, 30 and 40 mJ/cm2) and of protective additives (ROCK inhibitor and poloxamer 407) on delivery and mortality rates was quantified using ImageJ. Results No dextran was delivered with a fluence lower than 20 mJ/cm2. Dextran was delivered into 3% (range 0%–7%) of cells at 20 mJ/cm2, 7% (range 2%–12%) at 30 mJ/cm2 and reaching a median 13% (range 3%–24%) for 40 mJ/cm2, showing that dextran uptake by ECs increased significantly with fluence. Induced mortality varied from 0% to 53% irrespective of fluence, but likely to be related with the endothelial status (EC density and morphometry, donor age, storage duration and presence of Descemets folds). ROCK inhibitor slightly increased uptake efficiency, unlike poloxamer. However, none of them decreased the mortality induced by laser. Conclusions This study shows that a macromolecule can be delivered specifically into ECs of a whole organ-cultured human cornea, using fsL-activated CNPs. The delivery rate was relatively high for a non-viral method. Further optimisation is required to understand and reduce variability in cell mortality.

Femtosecond Laser Cutting of Multiple Thin Corneal Stromal Lamellae for Endothelial Bioengineering

Purpose: To assess the feasibility of cutting multiple thin stromal lamellae in human donor corneas using a commercial femtosecond laser (FSL) to provide cell carriers for future endothelial graft bioengineering. Method: Eight edematous organ-cultured corneas not suitable for grafting for endothelial reasons were mounted on a Ziemer anterior chamber and cut with a Z6 FSL with 6 successive parallel cuts, from depth to surface. Target thickness of each lamella ranged from 100 to 150 &mgr;m depending on initial corneal thickness. Thickness was measured using anterior segment optical coherence tomography before and after cutting on mounted corneas, and on each stromal lamella after detachment. Scanning electron microscopy observation was performed on 4 lamellae and histological cross sections on 1 cornea before detachment. Results: A median of 5 (minimum 3, maximum 7) lamellae was obtained per cornea. All lamellae still attached were the most posterior ones, suggesting that FSL was less efficient because of light scattering by edematous stroma. Cut precision and postdetachment swelling were correlated with anterior–posterior position within the cornea. Median lamella thickness was 127 &mgr;m (56–222 &mgr;m) before detachment and 196 &mgr;m (80–304 &mgr;m) after detachment. Surface state was consistent with previously reported FSL lamellar cuts during Descemet stripping automated endothelial keratoplasty. Conclusions: Up to 7 thin lamellae can be cut in stored corneas with an FSL. This method, once optimized primarily by using deswelled, more transparent corneas, could prove effective for recycling unsuitable donor corneas in corneal bioengineering processes.

Femtosecond Laser Cutting of Endothelial Grafts: Comparison of Endothelial and Epithelial Applanation

Purpose: Stromal surface quality of endothelial lamellae cut for endothelial keratoplasty with a femtosecond laser (FSL) with epithelial applanation remains disappointing. Applanation of the endothelial side of the cornea, mounted inverted on an artificial chamber, has therefore been proposed to improve cut quality. We compared lamellar quality after FSL cutting using epithelial versus endothelial applanation. Method: Lamellae were cut with an FSL from organ-cultured corneas. After randomization, 7 were cut with epithelial applanation and 7 with endothelial applanation. Lamellae of 50-, 75-, and 100-&mgr;m thickness were targeted. Thickness was measured by optical coherence tomography before and immediately after cutting. Viable endothelial cell density was quantified immediately after cutting using triple labeling with Hoechst/ethidium/calcein–AM coupled with image analysis with ImageJ. The stromal surface was evaluated by 9 masked observers using semiquantitative scoring of scanning electronic microscopy images. Histology of 2 samples was also analyzed before lamellar detachment. Results: Precision (difference in target/actual thickness) and thickness regularity [coefficient of variation (CV) of 10 measurements] were significantly better with endothelial applanation (precision: 18 &mgr;m; range, 10–30; CV: 11%; range, 8–12) than with epithelial applanation (precision: 84 &mgr;m; range, 54–107; P = 0.002; CV: 24%; range, 13–47; P = 0.001). Endothelial applanation provided thinner lamellae. However, viable endothelial cell density was significantly lower after endothelial applanation (1183 cells/mm2; range, 787–1725 versus 1688 cells/mm2; range, 1288–2025; P = 0.018). Conclusions: FSL cutting of endothelial lamellae using endothelial applanation provides thinner more regular grafts with more predictable thickness than with conventional epithelial applanation but strongly reduces the pool of viable endothelial cells.

Cutting and Decellularization of Multiple Corneal Stromal Lamellae for the Bioengineering of Endothelial Grafts

Purpose Engineered corneal endothelial grafts able to provide numerous functional endothelial cells for the restoration of corneal transparency would be a worthwhile way of replacing donor tissue, which is extremely scarce. The grafts are simply constructed: a biocompatible thin and transparent carrier colonized by a monolayer of cultured endothelial cells (ECs). Here we describe a process able to obtain appropriate carriers by recycling human corneas unsuitable for graft in their original state, but liable to provide multiple thin lamellae when cut with a femtosecond laser as used in refractive surgery. Methods We selected a robust method of stromal decellularization. To demonstrate that neither this process nor long-term storage hindered cell adherence, lamellae were endothelialized with an EC line. Results The constructs achieved up to very high EC density (the main quality criterion for regular donor corneas) while remaining transparent and thin. We verified that they could be inserted in the anterior chamber of a human eye, like a conventional endothelial graft. Human decellularized cornea will likely be directly compatible with the recipient cornea and comply with the requirements of health regulatory authorities. Conclusions This study demonstrates that thin human corneal lamellae could have high potential as carriers in next-generation therapy for endothelial dysfunctions.