Reaching out to senior scientists can feel daunting, but it’s often less intimidating—and more rewarding—than you’d expect. In this @ecrLife article, I share my experience as an early-career researcher and encourage others to take that step too.

RT @eLifeCommunity: "Only diamonds are made under pressure; good scientists are made with love and support.".

I’m grateful for the unwavering support from my advisor @wchnicholas and everyone from the Wu Lab. It has been an amazing journey so far, and I’m excited for more to come! We envision a near future where antibody binding characterization is fast, simple, and cost-effective. (14/).

oPool+ display also holds promise for promoting future synergy between machine learning models and experimental systems by enabling rapid validation of prediction results at scale. We believe it represents a starting point for the future of high-throughput antibody biology. (13/)

oPool+ display can be generalized to any antigens of interest as long as they can be recombinantly purified. More importantly, oPool+ display also offers great flexibility in experimental design, as the synthesized library can be stored, reused, and expanded at any time. (12/).

This suggests that a diverse native pairing library of close to ~20,000 unique antibody sequences can be assembled in just one 96-well plate PCR, followed by concurrent specificity characterization against 10-20 antigens of choice. Such a scale would be truly unprecedented. (11/)

In the update, we tested the limit of our strategy as we increased the # of scFv to assemble in a single PCR. Our one-pot PCR strategy performed amazingly well, as it achieved high reproducibility and coverage even with the assembly of different 200 scFvs in a single tube. (10/)

We overcame this challenge by splitting a defined scFv sequence into fragments and assembling them back through PCR. To ensure precise assembly, we utilized the highly diverse CDR regions to computationally design overlaps that are unique at the nucleotide level. (9/)

The key innovation of oPool+ display lies within the assembly of native paring antibodies, as current technologies only permit massively parallel oligo synthesis <350nt, which is only 1/3 of a single-chain variable fragment (scFv), the shortest human antibody format. (8/)

We then selected 25 antibodies and thoroughly tested their binding against all 9 HAs by both BLI (using Fab) and ELISA (using IgG). Both validations indicated >70% of true positive rate and >95% of true negative rate, demonstrating the robust performance of oPool+ display. (7/)

Our screening showed that 114 of the 325 antibodies bound to at least one of the nine HA screened, 45 of which were further identified to target the more conserved stem domain through either HA stem screens or binding competition with CR9114, a known HA stem antibody. (6/n)

As a proof-of-concept, we applied oPool+ display to rapidly synthesize >300 uncharacterized and uncommon influenza hemagglutinin (HA) antibodies and measure their binding to 9 HA variants through 16 different screens. Over 5,000 binding tests were performed in < 5 days. (5/)

Here, we combined oligo pool synthesis with mRNA display to synthesize and screen antibodies with defined sequences at scale. Compared to the conventional approaches, our platform is significantly faster (3-5 days), costs 80-90% less, and only requires one person to perform. (4/)

Considering that hundreds to thousands of natively paired antibody sequences can be isolated in a clinical study, this process would take dedicated research teams weeks to months to complete and would cost up to hundreds of thousands of dollars. A bottleneck clearly exists. (3/)

Antibody discovery is crucial to developing therapeutics and understanding adaptive immunity. However, the characterization of monoclonal antibodies today remains low throughput, as it involves cloning, expressing, and characterizing each antibody one by one (2/):

In this latest version of the preprint, we are excited to (re)introduce oPool+ display, a high-throughput cell-free platform that enables rapid synthesis and specificity characterization of natively paired antibodies at an unprecedented scale (1/):

Thanks for following our journey! I’m grateful for the contribution of all co-authors, especially @huibinlv1992 , Ken (Wenkan), and my mentor @wchnicholas. It was truly awesome seeing this project come alive from a spark of idea to where we are now! (14/14).

oPool+ display can be generalized to any antigens as long as they can be recombinantly purified. It also has the potential to benefit ML models for antibody engineering, specificity prediction, and de novo design in the future through rapid experimental validation (13/14)

Compared to conventional methods, oPool+ display is significantly more cost-efficient (10%-20% cost) and faster (~48-72 hours). It also has a relatively simple protocol, as it only requires standard benchtop equipment commonly found in a regular molecular biology lab (12/14).

Despite the unique binding mode, 16.ND.92 remained broadly reactive, binding to multiple H1s and recent avian H5s. It also conferred in vivo protection against lethal influenza challenge. This finding offers new insights into antibody sequence features on HA stem (11/14)