Journal: bioRxiv

Article Title: Sequence and structural determinants of efficacious de novo chimeric antigen receptors

doi: 10.64898/2025.12.12.694033

Figure Lengend Snippet: (a) Schematic of CD22 target antigen, with the arrow highlighting the preferred binding pocket as a hydrophobic patch. (b) Summary of YSD screening of de novo designed proteins from two campaigns against CD22. (c) Identification of four hits from the BindCraft campaign via sequencing. (d) CAR co-culture of four de novo CD22 binders (D1-D4) compared to m971 (clinical CAR). Shown is the %CD69 + Jurkats among GFP + cells. (e) Validation of CD22 expression of cell lines in the CAR co-culture. Arrow highlights the absence of CD22 expression in RPMI 8226. (f) Cocultures of three CARs with variable effector to target (E:T) ratios. Statistical test: Wald test of linear regression comparing D1 de novo binder to m971 clinical CAR, adjusting for E:T ratio. (g) Diversifying CD22 binder sequences given a single binder (D1). (h) Triplicate CAR Jurkat co-cultures with variable CAR binders. Statistical test: Two-sided Student’s t test. (i) Summary of diversified CD22 sequences in CAR co-culture. “X” highlights off-target activation from parental binder, D1. (j) Activation scores from scRNA-seq profiles of five CAR binders cultured against two different cell lines. Statistical test: two-sided Mann-Whitney U test. (k) Primary CAR T killing curves against RPMI 8226 (CD22 - ) showing off-target-specific killing in the de novo D1 binder. Statistical test: Wald test of linear regression interaction term between D1.N0 binder and time compared to D1, adjusting for time and binder.

Article Snippet: The following day, cells were washed once with 1× PBS-B (0.25% BSA) and incubated with varying concentrations of biotinylated recombinant antigen BCMA (Sino Biological, Cat. 10620-H40H-B), CD22 (Sino Biological, Cat. 11958-H49H-B), or CD19 (Sino Biological, Cat. 11880-H49H-B) for 1 hour at room temperature.

Techniques: Binding Assay, Sequencing, Co-Culture Assay, Biomarker Discovery, Expressing, Activation Assay, Cell Culture, MANN-WHITNEY

Journal: bioRxiv

Article Title: Sequence and structural determinants of efficacious de novo chimeric antigen receptors

doi: 10.64898/2025.12.12.694033

Figure Lengend Snippet: (a) Summary of mutations introduced to each of the CARPNN diversified CD22 D1 binder. Red residue index denotes interface residues while blue index denotes non-interface residues. (b) Comparison of CAR activation of the evolved CD22 D1 binders in CD22 - RPMI 8226 cell lines and CD22-overexpressing K562 cell lines. (c) Summary of diversified sequences from antigen CAR flow (top) and co-cultures with variable cell lines (bottom). (d) Representative Incucyte killing assays showing the cytolytic activity of CAR T cells expressing either CD22-specific minibinder- or scFv-based receptors. Time-course plots showing normalized red calibrated unit (%RCU) intensity relative to time 0h for each construct. (e) Cytokine productions from CD22-specific CAR T cells in co-cultures with CD22 + and CD22 - target cell lines. Heatmap shows mean cytokine levels across triplicates, revealing elevated cytokine release specifically in response to CD22-expressing targets, consistent with antigen-specific activation and killing. (f) Representative images at 0h and 72h for NB and at 72h for each binder condition to illustrate target-cell killing. Green fluorescence denotes CAR T cells, and red fluorescence denotes the corresponding target cell line. (g) Characterization of CAR antigen binding at variable CD22 concentrations. (h) Identification of plausible candidates of D1 off-target interaction via subsetting HPA surfaceome and GTEx overlap. (i) Comparison of average cofolding ipSAE score between parental D1 to all plausible off-target genes and the evolved D1.N0 binder to plausible off-target genes. (j) Predicted binding site of parental D1 binder towards CXCR4 aligned to a solved structure of CXCR4 (PDB: 8U4R).

Article Snippet: The following day, cells were washed once with 1× PBS-B (0.25% BSA) and incubated with varying concentrations of biotinylated recombinant antigen BCMA (Sino Biological, Cat. 10620-H40H-B), CD22 (Sino Biological, Cat. 11958-H49H-B), or CD19 (Sino Biological, Cat. 11880-H49H-B) for 1 hour at room temperature.

Techniques: Residue, Comparison, Activation Assay, Activity Assay, Expressing, Construct, Fluorescence, Binding Assay

Journal: JACS Au

Article Title: Molecular Insights into the Engagement of High-Affinity Sialylated Ligands to Human CD22

doi: 10.1021/jacsau.5c01013

Figure Lengend Snippet: Targeting CD22 with sialic acid-based ligands. (A) (Left) CD22 proximity to the B cell receptor (BCR) on the surface of B cells is required for its ability to antagonize BCR signaling. (Middle) Stimulation of B cells with antigen and sialic acid-based CD22 ligands can induce CD22–BCR proximity, leading to dampening of BCR signaling. (Right) Nanoparticles or liposomes coated with modified sialosides can deliver therapeutic payloads to targeted B cells. Image created in Biorender. (B) Overview of sialic acid-based ligands decorated with diverse chemical fragments targeting human CD22. Sialic acid derivatives with hydrophobic 9- N -biphenylcarboxamide (BPC)­56 or 9- N - m -phenoxybenzamide (MPB), , substituents at C9 exhibit high affinity and selectivity for human CD22.

Article Snippet: Human CD22 recombinant protein containing the most N-terminal Ig domains (CD22 d1‐d3 ) (UniprotKB P20273 , residues 20–330), alone or fused to human IgG1 Fc region (UniprotKB P01857 , residues 99–330), and with a C-terminal 6x His tag, was subcloned into pcDNA 3.4 (Invitrogen) and codon optimized for expression in human cells.

Techniques: Liposomes, Modification

Journal: JACS Au

Article Title: Molecular Insights into the Engagement of High-Affinity Sialylated Ligands to Human CD22

doi: 10.1021/jacsau.5c01013

Figure Lengend Snippet: Ligand screening and binding affinity measurement. (A) Workflow to identify high-affinity ligands for CD22. LiFlow cytometry was used for detection of the fluorescence emitted from PE antihuman IgG targeting CD22-Fc. (B) Schematic representation of magnetic bead binding assay. The streptavidin-coupled magnetic beads were coated with glycan-biotin probe, which competes with the test compounds. The glycan-ligand (polyacrylamide derivative with the biotin tag) consists of Neu5Gcα2–6Galβ1–4GlcNacβ-SpNH-PAA-1MD (6′SLN­(Gc)-PAA). (C) Titration curves of a competitive bead binding assay for compounds 1B , 7–012 , 12a-b and 16 .

Article Snippet: Human CD22 recombinant protein containing the most N-terminal Ig domains (CD22 d1‐d3 ) (UniprotKB P20273 , residues 20–330), alone or fused to human IgG1 Fc region (UniprotKB P01857 , residues 99–330), and with a C-terminal 6x His tag, was subcloned into pcDNA 3.4 (Invitrogen) and codon optimized for expression in human cells.

Techniques: Binding Assay, Cytometry, Fluorescence, Magnetic Beads, Glycoproteomics, Titration

Journal: JACS Au

Article Title: Molecular Insights into the Engagement of High-Affinity Sialylated Ligands to Human CD22

doi: 10.1021/jacsau.5c01013

Figure Lengend Snippet: Affinity of Synthesized Human CD22 Ligands

Article Snippet: Human CD22 recombinant protein containing the most N-terminal Ig domains (CD22 d1‐d3 ) (UniprotKB P20273 , residues 20–330), alone or fused to human IgG1 Fc region (UniprotKB P01857 , residues 99–330), and with a C-terminal 6x His tag, was subcloned into pcDNA 3.4 (Invitrogen) and codon optimized for expression in human cells.

Techniques: Synthesized

Journal: JACS Au

Article Title: Molecular Insights into the Engagement of High-Affinity Sialylated Ligands to Human CD22

doi: 10.1021/jacsau.5c01013

Figure Lengend Snippet: Thermodynamic parameters of the interaction between 1B , 7–012 , 12a , 12b , 16 , or 17 and CD22. (A) Representative raw (top) and processed (bottom) ITC data for 1B , 7–012 , 12a , 12b , 16 , or 17 binding to CD22. The calculated mean K D ± SEM for at least two independent experiments is indicated. (B) Bar graph comparing the measured binding thermodynamic parameters (Δ G , Δ H , and − T Δ S ) for the interactions between 1B , 7–012 , 12a , 12b , 16 , or 17 with CD22. Mean ± SEM for at least two independent experiments is reported.

Article Snippet: Human CD22 recombinant protein containing the most N-terminal Ig domains (CD22 d1‐d3 ) (UniprotKB P20273 , residues 20–330), alone or fused to human IgG1 Fc region (UniprotKB P01857 , residues 99–330), and with a C-terminal 6x His tag, was subcloned into pcDNA 3.4 (Invitrogen) and codon optimized for expression in human cells.

Techniques: Binding Assay

Journal: JACS Au

Article Title: Molecular Insights into the Engagement of High-Affinity Sialylated Ligands to Human CD22

doi: 10.1021/jacsau.5c01013

Figure Lengend Snippet: Structural details of the interaction of 7–012 , 17 ,and 1B ligands and CD22. (A) Crystal structures of the ligand binding pocket of CD22 (in gray) in complex with 7–012 (left panel), 17 (middle panel), and 1B (right panel). The ligands are depicted with sticks: Neu5Ac sugar moiety colored in magenta, Gal in yellow, GlcNAc in blue; the unnatural C2-R in wheat and the C9-R in green. H-bonds are represented with black dashed-lines. (B) Two dimensional schematic representation of the CD22- 7–012 , CD22- 17 , and CD22- 1B interaction network generated with Discovery Studio 4.0. The legend represents the type of interaction between the ligand atoms and amino acid residues of the protein (see also for further details).

Article Snippet: Human CD22 recombinant protein containing the most N-terminal Ig domains (CD22 d1‐d3 ) (UniprotKB P20273 , residues 20–330), alone or fused to human IgG1 Fc region (UniprotKB P01857 , residues 99–330), and with a C-terminal 6x His tag, was subcloned into pcDNA 3.4 (Invitrogen) and codon optimized for expression in human cells.

Techniques: Ligand Binding Assay, Generated

Journal: JACS Au

Article Title: Molecular Insights into the Engagement of High-Affinity Sialylated Ligands to Human CD22

doi: 10.1021/jacsau.5c01013

Figure Lengend Snippet: Computed binding energies ( E BSSE , kcal mol –1 ) for the most relevant noncovalent contacts between the substituents and C2 and C9 and the amino acid side chains of CD22 found in the crystal structures of CD22 with 7–012 (A), 1B (B), and 17 (C). For 7–012 (panel A), cation−π interactions take place between the BPC aromatic rings with R131, while CH−π forces are deduced between the C9 methylene with W128 and the nitroimidazole fragment with Y64. For 1B (panel B), the interaction between the guanidinium of R131 with the BPC fragment is mostly of CH−π character, while that between the dichlorophenyl moiety with Y64 is mostly π–π. The CH−π interaction between the C9 methylene with W128 is kept. Finally, for trisaccharide 17 , the cation−π interactions between the MPB aromatic moiety with R131 is somehow off-axis, while CH-p interactions are deduced between the acetamide at C5 with W128 (weak) and the Gal moiety with Y64 (strong).

Article Snippet: Human CD22 recombinant protein containing the most N-terminal Ig domains (CD22 d1‐d3 ) (UniprotKB P20273 , residues 20–330), alone or fused to human IgG1 Fc region (UniprotKB P01857 , residues 99–330), and with a C-terminal 6x His tag, was subcloned into pcDNA 3.4 (Invitrogen) and codon optimized for expression in human cells.

Techniques: Binding Assay