Regenerative medicine aims to replace or regenerate tissues or organs to re-establish their normal function. In 1975, Rheinwald and Green developed a technique to isolate and amplify epidermal stem cells. Their discovery led to the development of cultured epidermal autografts (CEA), the first regenerative therapy using cultured cells. Adult stem cells are the working force behind tissue homeostasis and repair. Through constant division and specialization, they produce enough daughter cells to maintain tissue architecture and function. This process is orchestrated by an elegant cross-talk between the stem cells and their microenvironment. By using irradiated feeder cells (3T3-J2 cells), Rheinwald and Green were able to artificially instruct epidermal cells to grow in vitro. Later on, they found that these cells could regenerate a functional epidermis. Moreover, Barrandon and Green demonstrated that clonogenic keratinocytes lose progressively their growth potential in vitro. This process is called clonal conversion. The 3T3-J2 cells are mouse embryonic fibroblasts. The molecules that they produce are necessary to promote self-renewal of keratinocyte stem cells in vitro. If the quality of the culture system is not monitored, clonal conversion can occur rapidly and the therapeutic potential is lost. Although the system is now used in the clinics for the treatment of large burns and cornea injuries, the regulatory affairs express genuine concerns towards the animal origin of the feeder cells. Ultimately, we would like to replace the current 3T3-J2 culture system by a fully characterized system, devoid of animal products, for the production of CEA. In this thesis, we developed two strategies to work towards this goal. First, we developed a large scale RNAi strategy to investigate the cellular cross-talk between feeder cells and human keratinocytes. We have identified several putative “feeder genes”. One of these genes is Furin, a serine protease. Furin is expressed by the 3T3-J2 cells and is required to sustain the proliferation of human keratinocytes. Second, we investigated the impact of ROCK inhibition on the procurement and culture of human keratinocytes. We found that it promoted the adhesion and proliferation of freshly isolated human keratinocytes. In opposition to what was described previously, we did not observe evidences of cellular immortalization or reprogramming when keratinocytes where treated with Y-27632 (ROCK inhibitor). Together, the results of our two approaches provide new leads for the further development of a new culture system for human keratinocyte stem cells.