Case Study: From Sample to Lead In Just 3 Weeks

The Challenge
An effective response to a pathogenic threat requires comprehensive deep screening and characterization of a human antibody response at the maximum speed possible, so that pathogen-specific therapeutics can be quickly identified, developed and deployed. Ideally, samples from index patients (patients who are the first to have been confirmed infected with the pandemic virus) who recovered would be made available to deploy our pandemic response platform.

Our Approach

We have been working with DARPA since March 2018 as part of the P3 program to optimize our technology stack to find effective and field-ready therapeutics against pathogenic threats in record time.

We developed rapid antibody screening, expression, purification and characterization pipelines to deeply mine human antibody responses. As part of the P3 program, and prior to COVID-19, we pressure tested our technology stack twice in simulated pandemic responses:

  • MERS-CoV: In late 2018, we demonstrated rapid isolation of hundreds of Middle Eastern Respiratory Syndrome Coronavirus, or MERS-CoV, heavy chain only antibodies, or HcAbs, from infected camelids (a natural host for MERS-CoV) in less than 96 hours from sample receipt. Many of these HcAbs were more potent neutralizers than benchmark antibodies.
  • Pandemic Influenza (H1N1): In early 2019, together with our partners, we discovered influenza-neutralizing antibodies from a single sample from a human donor, and demonstrated that we could deploy our platform from sample receipt to successful testing in animals in 55 working days. Our seven lead antibodies were all 100% protective against a 20-times lethal dose of the 2009 pandemic H1N1 strain of influenza virus in rodents.

The Result

We rapidly deployed our pandemic response platform to find a therapeutic antibody against COVID-19 in the spring of 2020 starting from a blood sample obtained from a U.S. patient. We screened approximately 5.8 million single cells to identify over 500 unique anti-SARS-CoV-2 antibodies. Each of these antibodies was evaluated computationally and experimentally to identify approximately 500 different properties per antibody which yielded 220,000 data points, which allowed us to filter down to a smaller group of lead candidates.

Within 23 days of receiving the sample, we and our partners identified 24 lead antibodies for further development and clinical testing. One antibody drug candidate was selected by our partner Lilly, and the first patients were dosed in the first-ever COVID-19 monoclonal antibody clinical trial in North America. This was only 90 days from when we received the sample. The antibody, LY-CoV555, has undergone or is undergoing a Phase 1 clinical trial, three Phase 2 clinical trials and two Phase 3 clinical trials.

From sample to lead in just 3 weeks: real-time response to COVID-19

On October 7, 2020 Lilly submitted a request for LY-CoV555 to the U.S. Food and Drug Administration (FDA) for emergency use authorization (EUA) as a monotherapy treatment in higher-risk patients recently diagnosed with COVID-19. Interim Phase 2 data of the monotherapy shows a reduced rate of hospitalization for patients treated with LY-CoV555.

For more information on how AbCellera’s technology can deliver large panels of antibody candidates for your therapeutic programs, please contact our team.

Case Study: Amplifying Antibody Diversity

Probing and searching natural immune diversity for therapeutic antibodies is a daunting task. Nature produces an immense amount of diversity in response to any given antigen, and from an antibody discovery perspective, the nature of the response determines the overall quality and diversity of the antibodies that can be identified for therapeutic applications.

In recent years, humanized transgenic mice have become the preferred starting source of natural diversity given their ability to generate fully human antibodies in a timely manner and under constraints of in vivo selection. However, to fully utilize the potential of these animals, deep screening of the immune response is required. Frequently employed outdated technologies, such as hybridoma and display-based approaches, are unable to optimally mine the natural diversity in these animals.

Next-generation antibody discovery platforms, such as AbCellera’s high-throughput single-cell screening systems, enable an enhanced level of sensitivity to search and identify rare and valuable antibodies. This is in large part due to unbiased deep screening of millions of antibody-producing cells in any given day, as opposed to only a few tens of thousands of cells with legacy systems, which capture only a small fraction of the natural diversity in any immune system. When combined with humanized mice, this deep screening approach generates large panels of paired heavy/light chain antibody sequences that bypass the need for additional humanization.

To mine the immune response even deeper, the process of immune repertoire sequencing (RepSeq) has recently emerged. RepSeq uses high-throughput sequencing to perform near-comprehensive profiling of the repertoire of heavy and light chain antibody genes present in a pooled or single sample. In its highest-throughput implementation, a single RepSeq sequencing run generates approximately 800 million antibody sequences. This vast diversity is often used to provide a high level profile of an immune response but lacks (1) contextual information about the binding properties of the corresponding antibodies and, (2) is not as sensitive to rare antibodies that might be present in the initial sample.

To combine the best of both worlds, we have integrated high throughput single-cell screening with RepSeq. With this approach, we can generate very large panels of antibodies with known properties and expand diversity around single cell data with RepSeq.

The Challenge: A discovery program to identify an antibody against a validated human target, with 100% homology across mice, rats and humans. Accessing a diverse set of antibodies was a must. In addition, the human target had two related protein homologs, against which antibodies had to be carefully screened.

Our Approach: In this campaign we deployed our optimized immunization protocols to generate a robust immune response in humanized rodents. Next, we developed a high throughput single-cell screening strategy that used multiplexed fluorescence detection to find antibodies specific to the target, but also with binding profiles that would cross-react with the closely related targets.

The Result: We screened four million single cells from immunized rodents and identified more than 1,900 target-specific antibodies, a hit-frequency that demonstrates a robust immune response from breaking tolerance against a 100% homologous target. Of these hits, we identified 428 unique antibody leads with a degree of somatic hypermutation that indicates a mature and directed immune response against a difficult target. We also performed immune repertoire sequencing in parallel and identified clonal lineages in Celium™.

Case Study: Single Chain Antibodies

HcAbs are single chain antibodies lacking light chains, found naturally in camelids such as llamas and alpacas (along with conventional paired heavy and light chain IgG antibodies, at approximately 50% frequency). HcAbs are valuable for various antibody-based therapeutics, such as bispecific antibodies, antibody-drug conjugates and CAR-Ts. In addition, the decreased complexity of having only single chains comprising the antibody molecules means they are more straightforward to produce and manufacture.

Find small & sleek single-chain antibodies for any modality (VHH)

The Challenge: A partner needed to identify HcAbs from immunized llamas against transforming growth factor beta-3 (TGF-ß3) that could also discriminate between closely-related homologs, transforming growth factor beta-1 (TGF-ß1) and transforming growth factor beta-2 (TGF-ß2). The partner did not have the screening technology to identify antibodies with such restricted specificity.

Our Approach: We designed custom reagents to distinguish HcAbs from conventional IgGs in llamas and also designed custom assays to identify HcAbs specific to TGF-ß3 that could differentiate between TGF-ß1 and TGF-ß2. To find these rare antibodies, we deployed our deep screening platform and screened approximately 20 million single cells over four days to deeply mine the immune response.

The Result: We identified 67 unique HcAbs, which is a <0.001% antigen specific HcAb hit frequency, indicating an exceedingly rare antibody. This illustrates the flexibility of our platform to discover non-conventional antibody modalities, from alternative species, and the depth required to find rare antibodies with very specific properties.

Find out how our single domain discovery technology can give you the flexibility to take on tough challenges by contacting our team.