Development of an innovative strategy based on engineered autologous erythrocytes as “enzyme replacement therapy” for the treatment of Phenylketonuria

INTRODUCTION Phenylketonuria (PKU) is a genetic disease caused by a mutated Phenylalanine Hydroxylase. The consequent accumulation of Phe to toxic levels leads to severe neurological disability if adequate treatment is not applied early. Current dietary therapies are often not complied with, hence alternative approaches are needed. Previous investigations have proven that administration of the enzyme Phenylalanine Ammonia Lyase (PAL) from Anabaena variabilis could be an effective strategy as an Enzyme Replacement Therapy (ERT) for PKU. METHODS Our aim was to investigate if RBCs loaded with PAL may function as a safe delivery system able to overcome bioavailability and immunity issues, in order to lower blood Phe. Murine RBCs were loaded with recombinant AvPAL and their ability to act as bioreactors was assessed in vivo. Three groups of BTBR-Pahenu2 mice received a single i.v. of rAvPAL-RBCs at three different doses to select the most appropriate one. Three i.v. of the selected dose were then performed every 18 days to assess the best time-lag between injections and, subsequently, we tested seven repeated i.v. of the same dose. Anti-rAvPAL IgG titers were also evaluated by ELISA. Then, the loading protocol for proteins has been optimized (with hexokinase as model protein) for human RBCs through the Red Cell Loader equipment and software 3.2.0. Three different loading conditions have been tested (loading step set at 3, 5 and 10 min) and the final products were challenged by a stress test after addition of different buffers (or varying concentrations of the same buffer) to verify the possibility to store loaded RBCs for periods longer than 30 min (current maximum acceptable period before reinfusion). The determination of supernatant free Hb and HK was employed to evaluate the alterations suffered by RBC membranes following the stress test, assuming that these proteins could leak out the cells through a damaged membrane. RESULTS 0.5 IU was the lowest dose able to keep Phe in a safe range for at least 8 days; the best inter-injections time span in which Phe stayed below the pre-treatment limit was 9-10 days. In both cases the repeated administrations had a persistent effect, not affected by neutralizing antibodies; the increasing anti-rAvPAL IgG titers actually did not correspond to a reduced efficacy. As concerns results obtained from RBC loading by means of the Red Cell Loader, we confirmed the procedure involving a 10-minute loading step as the one yielding the best characteristics of the final product. The best RBC membrane resistance to the stress test was found when phosphate buffer was added to the final product, with reduced levels of HK and Hb leakage. This results was enhanced when increasing concentrations of phosphate buffer were added to loaded RBCs. CONCLUSIONS Given the positive results obtained, this approach deserves further in vivo evaluation, either as an ERT for PKU or also as a model to develop similar treatments for other metabolopathies. We also managed to obtain protein-loaded human erythrocytes meeting international requirements for blood components by employing the EryDex system and the Red Cell Loader, which could facilitate the transition from the laboratory to the clinics. A clinical trial should be desirable and could be feasible through this device.