Recessive dystrophic epidermolysis bullosa (RDEB) is an incurable, devastating, inherited skin disease for which there is only supportive care. RDEB is due to mutations in COL7A1 gene that encodes for type VII collagen (C7), the major component of anchoring fibrils (AFs) mediating epidermal-dermal adherence. Approximately 20% of COL7A1 mutations are nonsense mutations leading to premature stop codons and a truncated C7 with diminished function. The investigators demonstrated that aminoglycosides such as gentamicin readily induce premature termination codon (PTC) "read through" and produce biologically functional C7 in 22 reported COL7A1 nonsense mutations. Importantly, aminoglycoside-induced C7 reversed the abnormal RDEB cell phenotype and incorporated into the dermal-epidermal junction. Herein, the investigators propose the first clinical trial of gentamicin (topical and intradermal) in RDEB patients with nonsense mutations that the investigators have fully characterized. The milestones include increased C7 and AFs at the patients' dermal-epidermal junction and absence of significant gentamicin side effects.

Evaluate C7 and AF expression with topical and intradermal gentamicin treatment Topical application of gentamicin to skin lesion: We will apply commercially available gentamicin 0.1% ointment three times a day under Tegaderm occlusion for two weeks to a target RDEB skin erosion. Likewise, we will apply the ointment control vehicle three times a day to a similarly sized RDEB erosion. This dosage schedule is based on the successful clinical trial in CF patients that resulted in clinical efficacy without side effects.19 In our in vitro cell culture studies, we showed that a single dose of gentamicin (400 ug/ml) induced C7 expression at a level of 20-40% of that seen in normal cells (Figure 1A and 1B).23 In our clinical trial, the experimental ointment will contain 1 milligram of gentamicin per milliliter of ointment vehicle.

Injection of gentamicin into the high dermis of RDEB patient skin: In addition, we will identify target 2.0 cm x 2.0 cm areas of unwounded intact skin within areas prone to blister formation and intradermally inject commercially available sterile gentamicin solution (40 mg/ml in saline). A similar control area will be injected with equal volumes of saline. We will inject 200 ul (8 milligrams) into each site once on day 0 and once again on day 1. A single injection will be administered to the site on each of the two days. The total dose will be 16 milligrams injected into the upper dermis where it will contact the patient's dermal fibroblasts and basal keratinocytes. In our published in vitro read-through study, each 1 cm2 of cultured cells was exposed to 400 ugs/ml of gentamicin that showed read-through activity with no cytoxocity. For the proposed intradermal study, we will be using a total dose approximately 5-fold greater than the in vitro dose.

Initial and follow-up Parameters: Prior to any treatment, the RDEB patients will be subjected to a 9 mm shave biopsy of intact skin that will be divided into 3 parts and evaluated for H&E histology, transmission electron microscopy and direct immunofluorescence for C7 expression. At 1 and 3 months after gentamicin treatment of RDEB erosions or intact skin in blister prone areas, we will biopsy the treated sites and repeat the histological, ultrastructural and C7 expression evaluations. For the assessment of C7 expression at the DEJ by immunofluorescence (IF), 5 micron cryosections will be probed with anti-C7 polyclonal antibodies to the NC1 and NC2 domains of C7. The increased expression of the NC2 domain of C7 at the DEJ of gentamicin-treated skin or erosions will serve as one major "milestone" in this study since it would indicate PTC read-through and restoration of a full-length C7. The third part of the biopsy will be evaluated ultrastructurally, and AFs will be enumerated by computer-assisted morphometry. These studies will assess if there is restoration of normal AFs. Skin sections from normal human subjects will serve as positive controls, while skin sections from the vehicle control site will serve as negative controls.

Patient clinical assessment: Skin Erosion Sites: Patients will be blinded to the gentamicin and vehicle treatments of the Experimental Site and Control Site. Each week, the patients will assess the sites and grade their healing as follows: - 1 = enlargement of the erosion compared to its initial size; 0 = no change in the size of the erosion; +1 = partial healing and a smaller erosion than its initial size, and +2 = complete closure of the wound.

Secondly, baseline photographs of the erosions will be generated and the area of the erosions calculated by computer-assisted planimetry. Identical assessments will be made at 1 and 3 months post treatment. Therefore, a second "milestone" for this study will be decreased surface areas of gentamicin-treated erosions compared with vehicle control-treated erosions.

Evaluation of RDEB intact skin treated by intradermal injections of gentamicin: Patients and Investigators will be blinded to the treatments of the Experimental Area and Control Area. Each week, the patient will evaluate the areas and grade them as follows: - 1 = new blister or erosion formation in the site, and +1 = no new blisters or erosions. At USC visits at 1 and 3 months, biopsies will be obtained from the sites and evaluated as above for the expression of C7 at the DEJ and enumeration of AFs.

Evaluation of Patients' Safety: Patients will have baseline histories, review of systems (ROS), vital signs (including weight) and physical examinations on Day 0 before treatment, and at Day 1 (one day after treatment) and then at 1 and 3 months during their visits to USC. At these same time points, blood tests will be performed and include a complete blood count, electrolytes, liver enzymes, erythrocyte sedimentation rate, creatinine, and BUN. Creatinine clearances will be also calculated, and the patient's treatment sites will be evaluated for erythema, edema, blistering and erosions. At Day 0, 1 month and 3 months after treatment, audiometry evaluations will be done. Patients will complete a ROS questionnaire daily at home, and be telephoned weekly by a USC study member inquiring about any new signs, symptoms, or ROS changes.

E. Characterization of Immune Responses to Gentamicin-Induced C7: Our study patients all express lower levels of C7 including the NC1 domain, which is the most antigenic domain of C7. Therefore, with the exception of patient B, we doubt that reading through the patient's PTC will induce a protein that is viewed as antigenic by the patient's immune system. Nevertheless, we hope that gentamicin will generate a functional, rather than non-functional, species of C7. To evaluate if this change triggers an immune response and generates anti-C7 antibodies, patient serum will be obtained at baseline, 1 month, and 3 months for evaluation of anti-C7 antibodies by salt-split IIF and ELISA. If a patient develops antibodies to C7, we will then examine their skin for C7 antibody deposits by DIF.

Inclusion Criteria:

(i) RDEB patients with a nonsense mutation in COL7A1 in either one or two alleles (ii) An absence or decrease in C7 expression at their DEJ when compared to that of normal human skin.

Exclusion Criteria:

(i) Pre-existing renal or auditory impairment (ii) Allergies to aminoglycosides or sulfate compounds (iii) Pregnancy (iv) Exposure to gentamicin within the past 6 weeks.

5 RDEB patient with bonafide nonsense mutations. One patient declined to participate in the intradermal injection portion of the study.

3 months

Ointments were formulated by an independent pharmacy and masked before receipt by investigators and delivered to participants as such. Intradermal injections were blinded by clinic staff (unrelated to the study) before injection.