Engineering of triketone-insensitive human 4-hydroxyphenylpyruvate dioxygenase enzymes as basis to develop e new gene therapy for hereditary tyrosinemia type 1

Hereditary tyrosinemia type 1 (HT1) is a rare, life-threatening disorder of the tyrosine degradation pathway that mainly affects the liver cells. The overall incidence of this disease is estimated at 1 in 100.000 newborns worldwide. HT1 is characterized by a dysfunctional fumarylacetoacetate hydrolase (FAH) enzyme that leads to the progressive loss of hepatocytes accompanied by an increased risk for hepatocellular carcinoma (HCC), renal tubular dysfunction with rickets and neurological crises. This is caused by the accumulation of the toxic upstream metabolites fumarylacetoacetate (FAA), maleylacetoacetate (MAA), and succinylacetone (SA). Without proper treatment, most HT1 patients die in their infancy.
In the past thirty years, HT1 patients have been successfully treated with the drug NTBC [2-(2-nitro-4trifluoromethylbenzoyl)-1,3-cyclohexandione; nitisinone], a potent inhibitor of the 4hydroxyphenylpyruvate dioxygenase (HPD) enzyme, in combination with a tyrosine- and phenylalaninerestricted diet. NTBC partly corrects the biochemical deregulation and rescues patients from severe illness or even death. However, it does not cure HT1 or restore the enzymatic defect, but it inflicts a new disease state comparable to HT3.
HT1 patients may also be at risk to develop HCC, even when continuously treated with NTBC but this is related to multifactorial reasons. This was supported by the observation that the expression of a number of genes related to liver disease and HCC development was increased in HT1 mouse livers under continuous NTBC treatment but with residual SA levels. Especially, in the context of HCC development, a significantly increased expression of some markers, of which a number were further modulated upon NTBC therapy discontinuation, was detected. This indicates that NTBC therapy may not completely resolve HT1-driven liver disease nor completely abolishes the risk to develop HCC. This study highlights the importance of treatment adherence to NTBC and the need for other therapies to be developed, as NTBC does not present as the perfect solution. Whether NTBC will prevent HCC development forever, is not known yet.
The main aim of this dissertation was to engineer a human HPD enzyme variant with decreased sensitivity to NTBC (HPD∆NTBC) by directed protein evolution as a basis for the development of a novel gene therapy for the treatment of HT1 that is functional under NTBC therapy. In combination with the WT version of FAH, this HPD∆NTBC variant will allow to fully restore the tyrosine catabolic pathway in transduced hepatocytes and will work without NTBC discontinuation. The engineering of potent variants of the human HPD enzyme that exhibit significantly decreased sensitivity to the inhibitor NTBC and other triketones was accomplished by applying a knowledge-gaining directed protein evolution approach. The metabolic competence of the HPD variants was measured by evaluating their ability to metabolize 4hydroxyphenylpyruvate to homogentisic acid in the presence and absence of NTBC with a new in-house developed robust, colorimetric high-throughput screening system. This led to the selection of one lead HPDΔNTBC variant that harbors 8 beneficial amino acid substitutions, all situated at the outer surface of the enzyme. The selected lead HPDΔNTBC variant exhibits in vitro 100 % metabolic activity at a 40-fold higher NTBC concentration compared to the WT enzyme.
Next, HPDΔNTBC was integrated into a bicistronic adeno-associated viral (AAV) vector with liver tropism (AAV-2/8) in combination with WT FAH, and the therapeutic efficacy was evaluated in a preclinical Fahdeficient mouse model of HT1 under NTBC treatment. The ability of the AAV-FAH-HPDΔNTBC vector to 2 correct the metabolic defect and prevent hypertyrosinemia under NTBC treatment was defined by closely monitoring clinical parameters such as plasma SA, tyrosine, phenylalanine, and NTBC. The success of the transgenic expressed FAH enzyme was demonstrated by its ability to lower the SA plasma levels, however, the functionality and ability of HPDΔNTBC to lower the tyrosine concentrations could not be demonstrated. As a consequence, troubleshooting was performed to optimize the expression and functionality of the transgenic HPD, i.e. use of single vectors and injecting a mixture of AAV-FAH and AAV-HPD, use of an immune deficient HT1 mouse model to rule out immunogenicity, insertion of a stabilizing T382A mutation and use of another TIMD mouse model, alkaptonuria mouse model, where HPDΔNTBC is applicable. Further research is needed to elucidate the problems associated with human HPD functionality in mice, which must be performed foremost in an in vitro setting, thereafter in vivo experiments could be reconsidered.