Hydrogel-Encapsulated Beads Enable Proximity-Driven Encoded Library Synthesis and Screening

Encoded combinatorial library technologies have dramatically expanded the chemical space for screening but are usually only analyzed by affinity selection binding. It would be highly advantageous to reformat selection outputs to ”one-bead-one-compound” solid-phase libraries, unlocking activity-based and cellular screening capabilities. Here, we describe hydrogel-encapsulated magnetic beads that enable such a transformation. Bulk emulsion polymerization of polyacrylamide hydrogel shells around magnetic microbeads yielded uniform particles (7 ± 2 μm diameter) that are compatible with diverse in-gel functionalization (amine, alkyne, oligonucleotides) and transformations associated with DNA-encoded library synthesis (acylation, enzymatic DNA ligation). In a case study of reformatting mRNA display libraries, transcription from DNA-templated magnetic beads encapsulated in gel particles colocalized both RNA synthesis via hybridization with copolymerized complementary DNA and translation via puromycin labeling. Two control epitope templates (V5, HA) were successfully enriched (50- and 99-fold, respectively) from an NNK5 library bead screen via FACS. Proximity-driven library synthesis in concert with magnetic sample manipulation provides a plausible means for reformatting encoded combinatorial libraries at scale.


Oligonucleotides
Oligonucleotides (Integrated DNA Technologies, Inc., Coralville, IA) were obtained as desalted lyophilates and used without additional purification.5' exonuclease assay probes were HPLC purified at the manufacturer.S3).
NGS library preparation and analysis.Bead aliquots (∼1000 beads) from each of the sorted populations were subjected to PCR amplification using primers P1 and P2 and gel purified to isolate the amplicons within the range of the templates (180-220 bp for all samples).Illumina sequencing libraries were constructed using the Bioo Scientific NEXTflex Rapid DNA-Seq kit and NEXTflex unique dual index DNA barcodes (Bioo Scientific Corporation, Austin TX).Amplicon (5 ng) was end-repaired, adenylated, and ligated to adapter.The adapter-ligated product was purified via solid-phase reversible immobilization (SPRI, Agencourt AMPure XP, Beckman Coulter, Inc. Brea, CA).Adapterligated and purified DNA was PCR-amplified (10 cycles) and purified by SPRI (AMPure XP beads) and quantified by Kapa qPCR (Kapa Biosystems, Inc. Wilmington, MA).The libraries were denatured, diluted (12 pM final concentration), and clustered (MiSeq, Illumina, Inc., San Diego, CA) using v2 Micro SR 300 cycles chemistry and dual indexing.
Analysis was performed by aligning each sequence to the control epitope or library se-S-12 quences and counting the proportion of (maximum of 3 mismatches) to each aligned sequence.Raw sequence data and processing scripts are located at (TBD).
Table S2.Translation and detection of non-canonical amino acid-containing gel particle library beads.NNT library particles (3 × 10 6 ) were subjected to an engineered IVT reaction that was recoded to install an azido-lysine (AzK) at CUG codons.Translated particles were washed and treated with an AF647-alkyne in a CuAAC reaction before analysis by flow cytometry.Particles were sorted to isolate the top 2% of the AF647 population and the sequences compared to the starting library.A 6-fold enrichment was observed after one round of screening.

Figure S2 .
Figure S2.Hydrogels prepared around 1 (left), 2.8 (middle) or 10 µm (right) magnetic beads are analyzed by flow cytometry (top).Particle populations are gated using the forward scatter height (FSC) and side scatter height (SSC-H) parameters.Particles (grey) are baseline resolved from particles hybridized to a FAM-labeled oligonucleotide complement (green).

Figure S3 .
Figure S3.The primary amine of the headpiece DNA (HDNA) was acylated with methacrylic acid to yield a methacrylamide-modified oligonucleotide for copolymerization into hydrogels.The MALDI-TOF mass spectrum displays the theoretical exact mass (cyan) and observed mass (black) for the [M+H] + .

Figure S4 .
Figure S4.HiBiT peptide standards (0.1-1000 nM HiBiT peptide) were combined with LgBiT/substrate mixture (5 µL, 2 × LgBiT/substrate in PBS) in a microtiter well plate (black, 384-well), and incubated (15 min, 37 °C).Luminescence was analyzed via plate reader and used to construct a standard curve.Unknown samples were quantitated by comparison to the standard curve.

Figure S5 .
Figure S5.NNT library particles (3 × 10 6 ) prepared by emPCR were subjected to an engineered IVT reaction that was recoded to install an azido-lysine (AzK) at CUG codons.Translated particles were washed and incubated with an AF647-alkyne in a CuAAC reaction before analysis by flow cytometry.Particles were sorted by gating the top 2% of the AF647 signal.

Figure S6 .
Figure S6.qPCR analysis of emPCR library preparation.(A) Aliquots of 100 beads are quantitated for to obtain average loading (blue traces) and limiting dilution of beads is sampled in 77 wells (green traces) to obtain single-bead quantitation of DNA templating.Traces for standards (gray) and negative template control (black) are shown.(B) Quantitation of 100-bead and DNA-templated single beads is shown in the box plots.Average particle loading for the 100-bead aliquot was 4,200 DNA molecules per bead.Average particle loading for single DNA-templated beads was 42,000 DNA molecules per bead.