2021-09-14_12h03_45

Custom Antibody Services

Antibody Lead Optimization Services

Create next-generation therapeutics. Our STage Enhanced Maturation (STEM™) platform has successfully improved antibody functions including affinity, pH sensitivity, expression, and developability. By leveraging our platform, you’ll maximize your chances of engineering the most optimal and developable therapeutic leads.

 

Antibody Lead Optimization Services

Overview

We offer our lead optimization services standalone or as part of our end-to-end antibody development services. R&D teams can get high-quality antibodies with our flexible, optimized, and validated approach that improves affinity, specificity, functionality, and developability.

Our success rates are high, and it all comes down to our extensive experience, fully optimized processes, and innovative (patent-pending) technology:

STEM™ (STage-Enhanced Maturation) , our multi-stage antibody affinity maturation platform, combines advanced computational design with comprehensive benchtop selection, thereby mimicking the natural immune system to engineer enhanced function. By iterating over multiple stages, we can explore a sequence landscape vastly larger compared to traditional approaches. And by using directed evolution to constrain diversity, instead of artificial parameters designed in silico, we ensure that each antibody possesses not only improved function but also passes our strict developability filters.

After each engineering stage, we exhaustively characterize lead candidates and discuss the project outcomes with you before proceeding. This allows us to quickly adapt our engineering approach and saves you time and resources by avoiding unnecessary steps.

 
 
The final library pairs pre-selected light and heavy chains

STage-Enhanced Maturation platform. All CDRs are rendered blue, in red are rendered CDRs targeted in the given stage for mutagenesis. Separate single CDR mutant libraries are made in Stage A1 and selected mildly against the target to collect all functional clones. Then, CDR pools are combined separately for light chain and heavy chain in Stage A2 and stringently selected to obtain high affinity binders. In Stage B, these light and heavy chain pools are combined as a single library and selected using the most stringent condition to obtain clones with highest affinity. Strict stability and polyreactivity filters are included at each stage.

 

To get started designing your libraries, all we need is the amino acid sequence of your antibody!

 

Benefits of Phage Technology for Lead Optimization

Phage selection is extremely versatile and can be tailored to improve specificity, affinity, thermostability, and pH sensitivity to get you high-quality antibodies as early as possible. Unique in vitro selections enable on-rate and/or off-rate targeting for kinetic optimization. We also have extensive experience phage panning on native or transfected cells.

Flexibility

High specificity & affinity

With decades of collective experience, Abwiz Bio has led dozens of successful affinity maturation campaigns, producing >10-fold affinity improvement for human, mouse, rabbit, and even feline antibodies with low nanomolar starting affinity.

No royalty payments

More diversity

Our approach covers more diversity and yields higher affinity clones that cannot be achieved by traditional methods such as CDR walking or AI design. Phage libraries are easily >1010 in size, compared to 108 maximum size of yeast libraries.

100_success

Ensured developability

Developability is integrated into all selections unlike with AI modeling or yeast display methods. Phage can be heated for stability selection and subtracted on polyspecificity reagents to remove  clones possessing poor developability profiles.

STEM™ used to engineer broadly-neutralizing COVID-19 mAbs

Our novel STEM™ affinity engineering method was used successfully to create a panel of therapeutic antibodies that broadly neutralize all major SARS-CoV-2 Omicron variants to date. Using our rabbit monoclonal antibody discovery platform Rabwiz, we developed a potent lead mAb that showed strong neutralization of the Wuhan-Hu-1 strain. We humanized this antibody and subjected it to our proprietary STEM™ platform, iteratively improving affinity and broadening neutralization potency on SARS-CoV-2 variants as they emerged. This resulted in the most potent, broadest neutralizing COVID-19 therapeutic mAbs to date. See our peer-reviewed work, detailing our engineering strategy and antibody characterization: Rapid engineering of SARS-CoV-2 therapeutic antibodies to increase breadth of neutralization including BQ.1.1, CA.3.1, CH.1.1, XBB.1.16, and XBB.1.5.

Even when the parent mAb showed no detectable binding to or neutralization of later Omicron subvariants, we successfully improved neutralization to low ng/mL EC50 for even the latest subvariant strains circulating as of January 2024 including EG.5, HK.3, and JN.1

COVID

Use the same technology to improve function for your lead mAb today!

 

AI design with benchtop selection combines the best of both worlds

We use the latest tools in antibody modeling and bioinformatics analysis to carefully design our libraries. Our goal is to maximize library diversity while maintaining developability. The end result is functional, scale up-ready antibodies with CDRs that have been further humanized to reflect the natural immune repertoire.

AI modeling and bioinformatics are used to identify important residues to target…

We design diversity based based on:

  • AI modeling – based on DeepMind’s AlphaFold, we run a local instance to help define CDRs and peripheral CDR anchor residues
  • Germline comparison – which residues are different?
  • Comparison to observed CDR diversity of published antibody structures sharing the same germline (bioinformatics analysis) to identify conserved residues important for structure
  • CDR diversity can reach above 1010 in size, orders of magnitude larger than other display technologies

...then the entire library is subjected to benchtop filters.

Traditional AI-based services rely on in silico design alone, which can miss complex interactions that are not well understood or modeled such as stability or polyreactivity. Instead, we subject our smart libraries to strict filters and allow directed evolution to remove problematic CDRs, while simultaneously improving the desired function.

 

Timeline

 
  Steps Time
Library Construction   4-6 weeks
  •Gene/oligo synthesis
•Library synthesis
•Pilot studies
2-3 weeks
2-3 weeks
1 week (concurrent)
Panning/Screening   2-3 weeks
  •Phage panning
•Fab screening
•Sequence analysis
1 week
3-5 days
2 days
Stage A1 Completion   6-9 weeks
Stage A2 Completion Informed by Stage A1 results 4-6 weeks (3-4 months total)
Stage B Completion Informed by Stage A2 results 4-6 weeks (4-5 months total)
 

Case Studies

 

Case Study 1: Affinity maturation

Increasing affinity is the most common request for antibody engineering that we receive, and we have successfully improved target binding affinity for dozens of human, mouse, rabbit, and feline lead candidates using peptide, recombinant protein, or cell surface displayed targets. Because our CDR libraries reflect natural diversity observed within the human repertoire, the final engineered clones possess a more human and thus less immunogenic CDRs, which is especially important for humanized lead mAbs. We have often been successful improving affinity into the pM KD range even when starting with relatively strong binding candidates showing low nM KD starting affinity.

 

In the below example, a lead candidate possessing 5 nM KD affinity was engineered from Stage A1 through Stage B, resulting in mutations across three CDRs. The final engineered candidates were tested as IgG by surface plasmon resonance (SPR) for binding to target analyte. The top candidate possessed >1300-fold increased binding to 4 pM KD and was patented for therapeutic use.

 

CaseStudy 1

 

Case Study 2: Improving expression and developability

Antibodies possessing good developability profiles can easily be scaled-up and manufactured for therapeutic use without problems. Issues that can cause poor developability include polyreactivity, poor expression or stability, high viscosity at increased concentration, or susceptibility to post-translational modifications. Many of these properties are poorly understood and can be difficult to predict even with AI models.

 

In the below example, a client antibody possessed good function but poor expression. Using only cells natively expressing the antigen, STEM was used to select for clones that retained similar antigen binding compared to the parent but better expression yield. Using a previously validated technique pioneered by Abwiz scientists (publication link), the phage library was first transiently heated and cooled to remove unstable clones. Then, the library was incubated on various reagents to subtract undesired polyreactive clones, before performing positive selection on antigen cells. Stability and polyreactivity filter stringency was increased throughout each stage. The engineered clones showed >10-fold increased expression level with similar function in cell assays; a therapeutic patent was filed on the final candidates.

 

Case Study 2

Case Study 3: Engineering pH sensitivity for increased potency

Engineering mild acid pH sensitivity is a useful tool for antibody therapeutics. The tumor microenvironment is mildly acidic, so clones possessing increased binding to antigen at pH ~6 compared and decreased binding at pH ~7.4 in the plasma can have fewer off-target effects. For traditional antigens, endosome recycling releases antigen from the antibody at mild pH ~6, allowing degradation of the antigen and releasing the free antibody to return to the bloodstream. Antibodies that possess weaker binding at pH ~6 compared to pH 7.4 possess more efficient endosome recycling and thus can have both higher potency and increased half-life.

 

In the example below, we created biobetter versions of infliximab (Remicade) and adalimumab (Humira). We performed all three stages of STEM and selected for high affinity clones that retained high affinity binding to tumor necrosis factor alpha (TNFα) under physiological conditions but quickly released TNFα at pH 5.8. The final top engineered candidates for each parent antibody were screened by Octet to measure binding at pH 7.4 and dissociation at either pH 7.4 or pH 5.8. While infliximab and adalimumab showed negligible pH sensitivity, the engineered clones were quickly released under mildly acidic conditions.

 

Case Study 3

Case Study 4: Tuning binding kinetics by avidity engineering

For some clones, such as recombinant single-chain T-cell receptors (scTCRs), chimeric antigen receptors (CARs), or for functional agonistic or antagonistic antibodies, extremely high affinity binding may not be desired. In these cases, kinetic tuning may be required to selectively increase or decrease the on or off rate. Our platform is uniquely capable of performing the specialized in vitro selections necessary to elicit these unique binding profiles. In the below example, we engineered a Fab to create mutants possessing both increased and decreased affinity, which were then further screened for improved function.

 

In Stage A1, single CDR variants were selected under mild conditions on target antigen to remove non-productive clones. Then the combined CDR Stage A2 library was subjected to a variety of selection conditions including all combinations of fast or slow on and off-rates. The final clones were screened by both (2) monovalent ELISA to identify high affinity candidates and (2) bivalent ELISA, where Fab is pre-complexed with secondary antibody to create pseudo-IgG complexes, to identify high avidity, weakly binding candidates. Engineered clones possessing a range of binding affinities were successfully obtained.

 

Case Study 4

FAQs

AI-based affinity maturation methods typically iterate through a large sequence space in silico to model and predict mutations that confer increased binding, then synthesize and screen IgGs to determine the accuracy of their predictions. These methods have limitations in that models typically have difficulty predicting success beyond ~5 or so mutations. Additionally, developability issues such as polyreactivity are inherently difficult to predict and thus can arise in the final engineered clones.

In STEM, we use the latest modeling techniques paired with bioinformatics analysis to build targeted libraries up to >1010 per CDR. Because CDR pools are combined in later stages, the effective sequence space explored in STEM is exceptionally large. At each stage, strict thermostability and polyreactivity filters are included to ensure that the final mutant clones can directly move to clinical studies. In this way, we combine the best of both AI and directed evolution by pairing smart libraries with rigorous benchtop selection.

While we can’t guarantee the outcome of a specific project because every antibody is different, we always guarantee the quality of our work. Each library undergoes QC check by gel electrophoresis and DNA sequencing, and we validate the mAbs via ELISA, flow cytometry, functional assays, and more upon request.

Because we use phage display, our library sizes are larger than competitors (1E10 compared to 1E8 for yeast), so you have a greater chance of getting the exact characteristics you need. Phage also allows us to select for developability (via thermoresistance and polyreactivity selection) so lead candidates can get to the clinic faster.

No. Our affinity maturation services are flexible and tailored to your needs. For our multi-stage STEM process, we can stop at the stage where you are satisfied with the improved antibodies. We work with you at every stage of the process and discuss what our plan is for each subsequent stage to get approval before moving forward.

After each stage of the STEM method, we can characterize the Fabs using multiple methods:

  • ELISA
  • Flow cytometry
  • Functional Assays
  • Full sequence analysis
  • Octet Bio-Layer Interferometry (BLI)

Our dual-purpose vector allows soluble Fab expression in 96-well format, so we can screen lead candidates in high-throughput using crude bacterial supernatant before proceeding to the expensive, time-consuming IgG stage.

Yes! Abwiz Bio offers mouse, rabbit, llama, alpaca, and other species antibody humanization services. When paired with our STEM platform, we can not only humanize frameworks but can deliver engineered clones possessing more human CDRs.

Learn more about our Antibody Humanization and Optimization services.

SCIENTIST 1@2x
ScienceExchange 1@2x

Get in touch today to discuss your project requirements- we’d love to help.