European Commission logo
English English
CORDIS - EU research results

Generation of Novel Cardiotropic AAV-Based Serotypes to Evade Human Humoral Immunity

Final Report Summary - IMEVAAV (Generation of Novel Cardiotropic AAV-Based Serotypes to Evade Human Humoral Immunity)

Cardiovascular diseases remain the leading cause of morbidity and mortality in the modern world. Current therapeutic approaches, such as pharmacological approaches, attenuate disease progression but they do not cure. Cardiac gene therapy has emerged as a promising alternative to the traditional therapeutic approaches. Adeno-associated viral (AAV) vectors offer several advantages over other vectors, non-viral and viral. They display tropism for specific tissues, have low immunogenicity, lack pathogenicity and can sustain long-term expression. Indicative of their potential is their use in several clinical trials[1] and their establishment in the first approved gene therapy, Glybera[2]. Despite the growing success of AAV based gene therapy, the field is hampered by the increased prevalence of pre-existing immunity against different serotypes, which precludes almost 50% of the patients from receiving AAVs carrying therapeutic genes[1]. Several approaches have been undertaken to evade humoral immunity against AAVs, such as inclusion of only negative patients to trials, plasmapheresis to remove antibodies from the bloodstream, administration of high vector doses, pre-injection with empty capsids to absorb antibodies, and capsid engineering. These approaches suffer drawbacks, such as they exclude patients, they cannot be implemented to diseased patients, they evoke cellular immune responses (high dose and empty capsid), and they alter the properties of the capsid. Generation of novel serotypes however would affect the desirable properties. AAV serotypes 6 and 9 are cardiotropic, their tropism has been characterized extensively and have been used in small and large animal models[1]. The receptor for AAV6, N-sialic acid, and its binding motif on AAV6 has already been identified. Galactose has been identified as the AAV9 receptor, as well as its binding motive on the capsid, thus providing us with valuable information regarding the amino acids necessary for receptor binding and ultimately tropism. Studies in poliovirus, rhinovirus and most importantly AAV (in particular AAV2), have shown that the neutralizing antibodies and the receptor can bind to different amino acids.
Our goal has been to generate novel serotypes based on AAVs 6 and 9, which will retain the cardiac tropism, but will be able to evade humoral immunity. Based on the aforementioned studies, we proposed a directed mutagenesis approach, in which single mutations of specific amino acids would attenuate specifically the neutralization by antibodies.
Towards this goal, we have performed alanine scanning on the variable regions (VR) of the AAV capsid. The variable regions are located on the surface of the capsid and therefore are of interest, because are the only regions that can interact with antibodies, as they are exposed. The variable regions were also chosen based on several data in the literature that suggest they are involved in immune responses against AAVs. The approach that was undertaken was generation of single mutations, AAV production and validation of the immune evading properties using the neutralizing antibody assay.
Since the beginning of this project, 76 mutations have been generated on the AAV6 capsid and 104 on the AAV8. These mutations were generated on variable regions III, V, VI, VII, VIII and IX for both serotypes and IV for AAV9. The mutagenesis was performed from the original amino acid to alanine. When the original amino acid was alanine, it was changed to glycine or serine. The mutagenesis was verified by sequencing. As a first step toward the validation of the mutants, AAV were generated using the 3-plasmid system, transfection to 293T cells and collection of cells 3 days post transfection. To validate the packaging capacity of the generated mutants, the DNase Resistant Particles (DRP) were measured in crude lysates. As a control a mutation, Y388A, known to affect packaging was used. However, this approach proved not to be sufficiently sensitive to assess packaging. As an alternative, the viral lysates were purified from cell debris (non-packaged protein and DNA) using iodixanol gradients and an optimized protocol. Very few mutations were found to affect the packaging capacity in both serotypes, such as Y388A in VRIII. We then sought to test transduction efficiency in 293T cells and human induced pluripotent stem cell derived cardiac myocytes (hiPSC-CM). Some mutations affected the transduction capacity of the AAVs In particular mutations in VRIII of AAV9 overall suppressed transduction. The same effect, but to a lesser extent was observed with VRIX. Mutations in variable regions V, VI, VII and VIII generally did not depress transduction.
The major point of objective 1 of this project was the validation of the neutralizing antibody evading properties of the AAV mutants. Sera from normal donors were initially screened for neutralization of AAV6 and 9. Of these, the ones with seropositivity against AAV6 were chosen for further experiments with the mutants. We hypothesized that sera from individuals will contain antibodies against different epitopes on the AAV capsid. Therefore, the different mutants should display differing immune evading properties against different sera. Indeed, the neutralization pattern of each serum was not identical. Although some mutations had no effect on neutralization by all or some sera, some mutations alleviated neutralization by one serum but not by others. In general, although mutations in one variable region depressed transduction, they showed the highest alleviation of neutralization by most sera. Some mutations, either in single or in adjacent positions alleviated neutralization, whereas others increased neutralization by sera, which would be expected, as an amino acid substitution does not guarantee the desired properties. Finally, single mutations have been identified and will be used for subsequent in vivo experiments (objective 2 of this project).
With the completion of this project, the characterization of the AAV6 and AAV9 variable regions has been completed. To complete objective 1 of this project, double or multiple mutants were proposed to be generated, if necessary. The initial screen revealed few mutations that alone could alleviate neutralization by sera, thereby rendering the use of multiple mutations unnecessary. However, some double mutations will be further pursued. It is also noteworthy, in accordance to our hypothesis, that different mutations alleviate neutralization of different sera, which makes the combination of mutations against one serum challenging. The second objective of this project was to test the neutralizing evading properties of specific mutants against specific sera in vivo. For the in vivo experiments, mice were injected with serum and either the parent vector in a pilot study to establish the passive immunity model and the dosage of AAV per mouse to achieve sufficient expression levels. The viral genomes of AAV per mouse have been determined. However, the passive immunity mouse model needs further optimizations. In a second set of experiments, the parent vector and one mutant will be injected in passive immunity mice and transduction efficiency will be monitored. The parent vector is expected to be neutralized by the serum as opposed to the mutant AAV. In parallel the properties of the mutant, such as tropism, compared to the parent vector will also be monitored. The goal of this project is to identify amino acid mutations that are either specific or generally involved in humoral immune responses. This knowledge will be of substantial value to the patients that are currently excluded from clinical trials because of their pre-existing humoral immunity. Ultimately, these mutants, if they retain the favorable properties of their parent vectors, could be used for those patients.


1. Hammoudi N, Ishikawa K, Hajjar RJ (2015) Adeno-associated virus-mediated gene therapy in cardiovascular disease. Current opinion in cardiology 30 (3):228-234. doi:10.1097/HCO.0000000000000159
2. Watanabe N, Yano K, Tsuyuki K, Okano T, Yamato M (2015) Re-examination of regulatory opinions in Europe: possible contribution for the approval of the first gene therapy product Glybera. Molecular therapy Methods & clinical development 2:14066. doi:10.1038/mtm.2014.66