The control of corneal transparency depends on the integrity of the mono-stratified corneal endothelium, which is considered devoid of regenerative capacity after birth in humans. In pathological conditions leading to blindness by irreversible corneal edema, the significant losses of endothelial cells (ECs) are not replaced efficiently, indicating that neither new ECs derived from stem cells (SC) nor the division of ECs neighboring the lesions can contribute to a form of regeneration. However, several works of the last 25 years demonstrated that ECs possess residual capacity of proliferation ex vivo, and more recently, two teams suggested the existence of SC or endothelial progenitors located in the corneal periphery. In this thesis work, we have firstly optimized an immunostaining technique specially adapted to intact endothelium of flat mounted whole corneas. Consequently, we now have simple protocols of fixation at the right temperature, and of antigen retrieval that allow detecting multiple proteins with a clear subcellular localization. Using important series of non-stored and of organ cultured corneas, and thanks to this technique, we investigated the cell cycle status of ECs and the location of potential SC in human corneal endothelium. Our results indicate that ECs homogeneously express positive regulators (PCNA, MCM2, cyclin D1, cyclin E and cyclin A) as well as negative regulators of the cell cycle (P16, P27); several original descriptions have been made: diffuse cytoplasmic location of MCM2, paranuclear location of cyclin D1, absence of P21. The expression patterns suggest that ECS could be arrested after the restriction point of the G1 phase and that numerous mechanism of DNA repair are stimulated in ECs exposed to an important oxidative stress throughout live. We identified a novel organization of the micro-anatomy of the endothelial periphery and extreme-periphery, where cells accumulate in multiple small multilayered clusters connected radial centripetal cell rows of nearly 1 mm of length. Associated with the observation of a lesser differentiation and an increased proliferation capacity in this area, these elements suggest a novel model of endothelial homeostasis in humans: during life, SC continuously and extremely slowly sustain the endothelial periphery with new ECs that migrate toward centre forming cells rows. These cells ensure the stability of the center, which optical fundamental properties require a perfect stability, as indicated by an annual cell loss of only 0.6%. Contrary to the corneal epithelium, this system is incapable of accelerations in case of important cell loss. Further studies are necessary to understand this limitation. Our works offer several perspectives: the next step is to isolate the SC or the progenitors from the extreme periphery and to cultivate them in an adapted microenvironment. The possibility to cultivate endothelial cells directly from SC or progenitors and not, as previously tried in the past, from senescent EC from the whole endothelium open the way of a true endothelial cell therapy. The increase of endothelial cell density during corneal storage by eye banks could be a first step before developing bioengineered endothelium on a specific carrier that could be implanted in the recipient eye. Finally, the identification of SC and of its microenvironment would allow developing the basis of an in vivo cell therapy able to treat early stages of endothelial dysfunctions