The treatment of a patient affected by an incurable genetic skin disease demonstrates the efficacy, feasibility and safety of replacing almost the whole skin using genetically corrected stem cells. See Letter p.327
Stem-cell and gene therapies are often considered to be the future of medicine, but there have been many barriers to putting these approaches into practice. Indeed, there are few examples of truly useful human stem-cell therapies1. On page 327, Hirsch et al.2 describe a success in this area — the use of gene therapy to correct the cells of a child who had a devastating genetic disease associated with skin blistering.
The skin is composed of the epidermis, which acts as a barrier against the external environment, and the underlying dermis, in which the epidermis is firmly anchored, conferring elasticity and mechanical resistance3. In the disease epidermolysis bullosa, genetic mutations prevent normal epidermal resistance or anchoring4, making the skin fragile. Mechanical stress and minor trauma provoke epidermal fragmentation or detachment from the dermis, causing skin blistering and ulcers. This produces chronic, painful and untreatable wounds, and ultimately leads to skin cancers, infection and sometimes death4. There is currently no cure.
The group that performed the current study previously used gene therapy to treat a mild form of epidermolysis bullosa caused by mutations in the gene laminin β3 (LAMB3), which encodes an epidermal anchoring protein5. In that study, the authors isolated a small piece of epidermis from a patient and added a normal version of LAMB3 to the isolated epidermal cells, using a retroviral vector to carry the gene into the cells' nuclei. The vector integrated into each cell's genome, enabling normal LAMB3 expression. The group grew the genetically corrected cells in vitro to form a larger piece of epidermis that they transplanted onto the patient's leg, where it engrafted.
Hirsch et al. have taken this strategy much further. A seven-year-old child who had an extremely severe form of epidermolysis bullosa caused by LAMB3 mutations was admitted to hospital in a life-threatening condition, having lost almost his entire skin. The authors took a 4-square-centimetre biopsy from an unaffected part of his skin, genetically corrected cells using a retroviral vector carrying LAMB3, and grew the corrected cell population to obtain 0.85 m2 of transgenic epidermal grafts. They replaced 80% of the patient's skin with the grafts in three separate operations (Fig. 1). After 21 months' follow-up, the child seemed to have made a full recovery, with no blistering. His skin was resistant to stress and healed normally.
One possible complication of gene therapy is that, because the vector integrates into the host genome at random sites, it might disrupt essential genes or trigger overexpression of genes that control tumour development. To investigate this possibility, Hirsch et al. sequenced DNA from the patient's genetically corrected skin. Sequencing revealed that most integrations occurred in non-protein-coding sequences. The genes that contained integrated retroviral vectors are not known to be directly involved in cancer, demonstrating the safety of the approach.
Next, the authors compared integration patterns in the corrected in vitro cultures with patterns in the regenerated epidermis in vivo, to determine whether particular patterns (for example, integration into promoter sequences that drive gene expression) conferred a survival advantage that might predispose to cancer in the future. They found similar patterns in both conditions, indicating that neither the culture protocol nor natural skin-cell turnover led to preferential survival and expansion of particular cell subsets. Finally, the researchers found no sign of autoantibody production against the transgene, which would lead to the rejection of the graft, further demonstrating the safety of the approach.
The epidermis is completely renewed about once a month3. Whether renewal is ensured by stem cells at the top of a cellular hierarchy, or whether all proliferative cells behave as equipotent progenitors, choosing randomly between proliferation and differentiation, is a matter of debate6. Culturing epidermal cells in vitro produces three types of cell colony — holoclones, paraclones and meroclones. Holoclones are proliferative colonies composed of undifferentiated cells that have self-renewal capacity; paraclones are differentiated cells that have little renewal capacity; and meroclones are intermediate between the two7. Although it has been hypothesized that holoclone colonies contain epidermal stem cells, this relationship has not been formally demonstrated.
Hirsch et al. mapped the positions of viral integration sites in the genomes of holoclones in vitro, and compared them with the integration sites retained in cells from the child's skin four and eight months after grafting. There were many fewer different integration sites in the biopsied cells taken at four months than in the initial cultures. By contrast, many insertion sites were retained between the four- and eight-month samples. These data suggest that most cells present in the initial culture are lost over the first four months, with only a few stem cells contributing to long-term epidermal maintenance.
Moreover, there was a massive increase in the frequency at which the in vitro holoclone integration sites appeared in the rejuvenated skin over time (this skin contained not only holoclones, but also newly formed paraclones and meroclones). Thus, holoclone colonies contain stem cells that repopulate the regenerated skin. After eight months, almost the entire epidermis was derived from holoclones. Clearly, a few long-lived stem cells sustain the human epidermis.
Hirsch and colleagues' study demonstrates the feasibility and safety of replacing the entire epidermis using combined stem-cell and gene therapy. In addition, the work provides insights into the cellular hierarchy that governs epidermal maintenance in humans. But there are several considerations to be addressed before rolling the treatment out to other patients.
Epidermolysis bullosa can be caused by mutations in different genes, not all of which will be easy to correct. Strategies such as the use of CRISPR–Cas9 gene-editing technology will be needed to correct some mutations. It will also be necessary to adapt the procedure to different sites in the body, in particular in people who have less-severe skin conditions. The treatment might be more effective in children, whose stem cells have higher renewal potential and who have less total skin to replace, than in adults.
Finally, longer-term follow-up of the child in the current study and other patients will be needed, to ensure that there are no adverse consequences — for example, the development of skin cancers or changes that lead to the loss of transgene expression in some cells, which could result in blistered zones. Nonetheless, the authors' work marks a major step forward in the quest to use stem-cell therapies to treat disease.