AAV COVID-19 vaccine aims for the sweet spot between antibody and T cell immunity

The rationale behind the only AAV-based vaccine in development for COVID-19

BioCentury is providing this content for free given the urgent need for information about the coronavirus crisis. Further analysis can be found in our COVID-19 Resource Center. For more, sign up for our daily email.

The academic team behind the only AAV vaccine candidate for COVID-19 thinks its vector can get the best of both worlds and induce stronger antibody responses than adenoviral vaccines, while still activating T cells. Moreover, the researchers contend it can avoid viral vectors' Achilles heel: pre-existing immunity.

The dominant focus for COVID-19 vaccine development has been on their ability to generate antibodies against SARS-CoV-2, the virus that causes the disease. Whereas generating antibodies is the domain of B cells, their counterpart role of T cells has received less attention.

Activating T cell immunity brings multiple advantages, including recruiting CD4+ T cells that help B cells develop more potent antibody responses, and CD8+ T cells that kill infected cells directly.

Vectors encoding viral antigens are thought to be among the most efficient at generating immune responses, in part because they stimulate both B and T cells (see "Guide to COVID-19 Vaccine Modalities in the Clinic").

But the immune profiles of different vector types can come with trade-offs. Adenoviruses, a leading modality among COVID-19 vaccines, “are great T cell vaccines, but actually quite meek B cell vaccines,” said Luk Vandenberghe, director of the Grousbeck Gene Therapy Center at Massachusetts Eye and Ear.

He believes that AAVs, a vector class more commonly used for gene therapies, can be leveraged to strongly activate both B and T cells against COVID-19.

"The vaccine platform we developed back then also had no pre-existing immunity, which is unique in the space."

Luk Vandenberghe, Mass. Eye & Ear

Vandenberghe is heading the AAVCOVID program, which aims to get at least one candidate targeting the SARS-CoV-2 spike protein into the clinic in 2H20. It has lined up a phalanx of manufacturing and research partners including University of Pennsylvania's James Wilson, a gene therapy pioneer, and the AveXis unit of Novartis AG (NYSE:NVS; SIX:NOVN), which created the first gene therapy for spinal muscular atrophy (SMA) (see “Partners to Deploy Gene Therapy Tools for COVID-19”).

The AAVCOVID program hinges on a vector dubbed AAVrh32.33, which Vandenberghe helped generate in Wilson's lab over 10 years ago.

He told BioCentury that AAVrh32.33 had properties that made it well-suited for vaccine development, beyond the well-known ability of AAVs to induce strong antibody responses to foreign -- but not self -- cargo.

“Not only did it have the inherent AAV class benefit of inducing antibodies, but this particular platform also induced quite potent and qualitatively high memory responses on a T cell level,” he said. “The vaccine platform we developed back then also had no pre-existing immunity, which is unique in the space.”

Pre-existing immunity has already cropped up as a red flag for the first viral vector vaccine candidate to enter the clinic for COVID-19, Ad5-nCoV from CanSino Biologics Inc. (HKEX:6185), which uses a human adenoviral vector (see “Deluge of Early COVID-19 Vaccine Data”).

AAVrh32.3 was derived from rhesus macaque viral isolates.The COVIDAAV program will be the first time it is tested in humans.

But Vandenberghe thinks the “tremendous amount of platform experience and industry buildup” around AAVs used for gene therapy gives the team a head start in understanding the program's safety profile, as well as accessing manufacturing capacity.

The team can get 10,000-100,000 vaccine doses from the same amount of “manufacturing lift” needed to produce a single dose of Novartis' SMA gene therapy Zolgensma onasemnogene abeparvovec-xioi, according to Vandenberghe.

“The dose benefit for a vaccine, compared to a gene therapy, is really an incredible multiplier,” he said.

The program is supported by philanthropic funds and guided by a strategic advisory team that includes former Editas Medicine Inc. (NASDAQ:EDIT) CEO Katrine Bosley and Partners Innovation Fund’s Roger Kitterman.

Flexing both arms

AAVrh32.33 is a hybrid of two AAVs isolated from the spleen of a healthy rhesus macaque. In a 2009 Journal of Virology study, Vandenberghe, Wilson, and their co-authors showed vaccination via a AAVrh32.33 vector induced high antibody titers and potent CD8+ T cell responses in mice and non-human primates.

A follow-up paper in the same journal pinpointed the functional domain on the capsid required for activating T cell immunity, while another study in Journal of Infectious Diseases showed minimal pre-existing antibody responses to AAVrh32.33 in 888 serum samples from healthy volunteers in 10 countries around the world.

The authors reported that across all countries and regions surveyed, the prevalence rates of neutralizing antibodies against AAVrh32.33 were never more than 2% in serum dilutions of >1:20, and neutralizing antibodies were never detected in dilutions of >1:80. In contrast, other AAV serotypes, including ones used for gene therapy, had double-digit pre-existing antibody prevalences at both dilutions.

“This is a capsid that is 40% divergent from most of the AAVs that we're currently bringing to the clinic in the gene therapy world,” said Vandenberghe.

“We do believe there is a certain window where re-administration in a vaccine setting will be feasible.”

Luk Vandenberghe, Mass. Eye & Ear

The team has mouse studies ongoing and is poised to begin non-human primate (NHP) studies, with plans to publish its data.

Two adenoviral vector COVID-19 vaccines have entered the clinic. The first, Ad5-nCoV from CanSino, uses a human adenovirus 5 (Ad5) vector; the second, AZD1222, licensed by AstraZeneca plc (LSE:AZN; NYSE:AZN) from University of Oxford, uses a chimpanzee adenovirus vector. At least seven adenovirus vector vaccines are in preclinical development for COVID-19.

At least one AAV vaccine has been tested in the clinic: an AAV2-based HIV vaccine from Targeted Genetics Corp. The compound induced modest B and T cell responses to HIV in a Phase II study, and was discontinued. Vandenberghe said the field has since recognized that AAV2 is among the most weakly immunogenic AAV vectors.

Vandenberghe thinks AAVCOVID will be potent enough to administer as a single low dose. But if a second administration were required, he thinks the vaccine's low dose intramuscular delivery has a better chance of avoiding the anti-vector immunity that prevents repeat dosing of systemic gene therapies.

“We do believe there is a certain window where re-administration in a vaccine setting will be feasible,” he said. “The question is, whether we can get this particular vaccine in that window.”

Stocking up

Because AAV manufacturing is predominantly geared to produce high doses of gene therapies, Vandenberghe thinks the program's partners could generate up to one million low doses of AAVrh32.33 vaccines in a single typical run.

He said that multiplier, plus partners' commitments to prioritize production of AAVCOVID, will enable the program to rapidly generate large numbers of doses despite the notorious bottlenecks that plague AAV manufacturing (see “Mapping the Path to Gene Therapy 2.0”).

On May 28, the AAVCOVID team announced manufacturing partnerships with AveXis as well as CDMOs Aldevron LLC, Catalent Inc. (NYSE:CTLT) and Viralgen Vector Core, a Spain-based JV between Asklepios Bio Pharmaceutical Inc. (AskBio) and Columbus Venture Partners.

Vandenberghe said the partnerships give AAVCOVID two supply lines, which speeds up production and mitigates risk.

In one supply line, DNA plasmids produced by Aldevron go to AveXis for insertion into the AAV vector, and finally to Catalent for QC. In the second, plasmids from Aldevron or Viralgen are inserted into AAV vectors manufactured by Viralgen; QC can then be performed by either Viralgen or Catalent. Wilson’s team at UPenn is conducting some IND-enabling studies and analytics on manufacturing.

Two AAVCOVID candidates with different dosing went into GMP production in late April; the team plans to bring at least one to the clinic this year.

At least 24 other COVID-19 vaccine developers have mapped out the number of doses they expect to produce. Among viral vector vaccine developers, AstraZeneca and Johnson & Johnson (NYSE:JNJ) have each projected they could reach the one billion dose mark for their adenoviral vaccines in 2021 (see “Adding Up Manufacturing Capacity for COVID-19 Vaccines”).


SARS-CoV-2 S - SARS-CoV-2 spike protein

How to gain access

Continue reading with a
two-week free trial.