Company, Science, Services

Next Generation Sequencing-Enhanced Antibody Discovery

digital proteomics
March 3, 2020

Using Reptor for NGS-enhanced antibody discovery

Traditional antibody discovery methods, such as hybridoma and phage display, as well as newer single cell screening methods, capture only a small portion of the adaptive immune response, and as a result isolate a small number of highly similar antibodies For the antibody clones that emerge as hits, sequence liabilities such as N-glycan sites are red flags that cannot be overcome easily. In addition, natural variation of antibody sequences within a CDR3 family can guide engineering and design.

Bulk sequencing of the immune repertoire with ReptorTM provides a method for identifying relatives of hits with varying affinity and developability attributes. In this blog post we look at the opportunities for NGS-enhanced antibody discovery by analyzing antibody hits discovered by Alicanto® in the context of their CDR3 families.

Clonal lineages guide antibody engineering and design

In a recent Alicanto® case study [1], we identified a rabbit monoclonal antibody targeting a peptide. Briefly, we performed bulk antibody repertoire sequencing on multiple tissues and blood samples taken during the immunization. We identified antibody candidates from the repertoire that were present in serum and reactive to our target using affinity chromatography and mass spectrometry. Candidate sequences were synthesized, expressed, and validated via ELISA. We selected one candidate for analysis in this blog post.

We identified 70 heavy chain sequences in the repertoire with the identical CDR3. While the validated candidate was only observed in bone marrow, other members of the lineage were identified in spleen or peripheral blood mononuclear cells (PBMCs). Members of the lineage ranged in abundance from 2 reads to 148 reads (our validated candidate had abundance 7).

An anchored lineage constructed from of a subset of the 70 sequences is shown on the right, with the location of the validated heavy chain sequence starred (at the bottom of the lineage).

We selected the nearest clade to our hit in order to determine sequence logos for the CDRs and flanking sequences (shown at bottom of figure). Interestingly, within the clade we see greater variability in the CDR1 (three positions) versus only two positions in the CDR2. By analyzing the conserved and variable positions, it may be possible to identify target-binding residues [3] or design a rational variant library for screening [4].

Clonal lineage from NGS

Avoiding developability flags in candidate selection

In another example, Alicanto® was used to identify single chain antibodies (HCAbs) from a llama (we wrote about the project in a blog post) against both a peptide and the carrier, keyhole limpet hemocyanin (KLH). From that project, we selected the highest affinity candidate identified by Alicanto®. The selected candidate was a member of a CDR3 family containing 115 HCAbs with identical CDR3 amino acid sequences. The 51 family members with at least abundance 3 were used to construct an anchored lineage (below).

The lineage shows the hit HCAb (starred) among its relatives. However, selecting variants from this large pool of sequences can be tricky. Here we highlight sequence-based developability flags detected on several of the sequences.

  • Red squares indicate an un-paired cysteine in the variable region.
  • Red circles indicate a variable region charge greater than our hit, an attribute associated with faster non-specific clearance [2].

The lineage on the left is constructed on HCAbs with identical CDR3s. However, the CDR3 is a member of a collection of 143 highly similar CDR3s (CDR3 cluster). The sequence motif shows the positions that vary across the CDR3 cluster.

Developability overlayed on clonal lineage from NGS

Conclusion

As shown above, bulk antibody repertoire sequencing with NGS enables discovery of even more target-reactive candidates than single-cell and traditional approaches. The NGS-enhanced antibody discovery described here has been applied in a variety of species beyond rabbit and llama and can be further adapted to discover antibodies in transgenic animals and humans.

Our NGS-enhanced antibody discovery workflow can add value to your antibody discovery campaign. Read more about our unique Alicanto® platform for antibody discovery from serum, or our high-throughput hybridoma sequencing service with ReptorTM.

References

[1] Bonissone, Stefano R., et al. “Serum proteomics reveals high-affinity antibodies in immunized rabbits missed by B-cell repertoire sequencing.” bioRxiv (2019): 833871. (bioRxiv)

[2]Sharma, Vikas K., et al. “In silico selection of therapeutic antibodies for development: viscosity, clearance, and chemical stability.” Proceedings of the National Academy of Sciences 111.52 (2014): 18601-18606. (PubMed)

[3]Ramirez‐Benitez, Maria del Carmen, and Juan Carlos Almagro. “Analysis of antibodies of known structure suggests a lack of correspondence between the residues in contact with the antigen and those modified by somatic hypermutation.” Proteins: Structure, Function, and Bioinformatics 45.3 (2001): 199-206. (PubMed)

[4]Nimrod, Guy, et al. “Computational design of epitope-specific functional antibodies.” Cell reports 25.8 (2018): 2121-2131. (PubMed)