Design strategy and computational workflow for miniprotein binder design against CCL25.

(A) Chemokine interacts with chemokine receptors primarily through CRS1 (Pink, receptor N-terminus with chemokine body) and CRS2 (Purple, chemokine N-terminus with receptor TM domain). The receptor:chemokine complex is generated using AF3 with sequences of CCR9 and CCL25. (B) Conceptual framework illustrating the strategy for inhibiting canonical chemokine to receptor binding by directly targeting chemokine with designer miniprotein binders. (C) Computational workflow used to generate chemokine-binding miniproteins. (D) Molecular interaction map summarizing contacts between all miniprotein designs and CCL25. The most frequently targeted structural elements, the α-helix, the 310-turn, and the 40s-loop, are highlighted in red, green, and blue, respectively.

VUPs (VU miniproteins) are confirmed to directly bind CCL25.

(A) Structural models of the generated designs for CCL25. CCL25 is shown in blue with a surface representation, and the binders are displayed as cartoon in various colors. (B) Cartoon schematic for the fluorescence polarization assay. When the miniprotein binds CCL25_AZ488, the size increases, as does the polarization of the emitted light. (C) The polarized fluorescent emission was measured for 15 nM CCL25_AZ488 in the presence of increasing concentrations of VUP binders. Binding of the miniproteins leads to an increase in polarization. Data points represent the average ± SD of three independent experiments and internally normalized to the minimum and maximum polarization of each binder.

VUPs interfere with CCL25 binding to CCR9 and ACKR4 through clashes with CRS1.

(A) Design of binding assay using TR-FRET between CCL25_ΔCT-AZ647 and Terbium (Tb) labelled SNAP-CCR9 and SNAP-ACKR4. TR-FRET binding results for CCR9 (B) and ACKR4 (C) using 20 nM CCL25_ΔCT-AZ647 with increasing concentrations of the miniproteins, which shows a decreasing TR-FRET signal indicating an inability of the chemokine:VUP complex from binding the target receptor. Values represent the mean ± SD of three independent experiments performed in triplicate, presented as fold over vehicle condition. (D, E) Predicted structural models of VUP25101 and VUP25111 superimposed on the AF3-predicted CCR9:CCL25 (D) and ACKR4:CCL25 (E) complexes with zoom-in panels highlighting where the miniproteins are predicted to sterically clash with receptor:chemokine interactions.

VUPs selectively inhibit CCL25-induced arrestin recruitment towards CCR9 and ACKR4.

(A) Schematic illustration of BRET sensor pair CCR9/ACKR4-RlucII and GFP10-β-arrestin2. The structure of β-arrestin2 is generated using AF3. β-Arrestin2 recruitment towards CCR9 (B) or ACKR4 (C) following stimulation across a titration of CCL25 concentrations (black) or a fixed concentration of CCL25 (20 nM) preincubated with increasing concentration of miniprotein. Values represent the mean ± SD of three independent experiments performed in duplicate and normalized to min and max CCL25 response.

Most VUPs, but not VUP25111, effectively inhibit CCL25-induced Gi protein activation.

(A) Schematic illustration of BRET pair sensor CCR9 and Venus-mGi. The complex representing mGi is taken from PDB 9KO4[45]. (B) Concentration-dependent mGi recruitment to CCR9 following stimulation with a titration of CCL25 (black) or with 20 nM CCL25 preincubated with increasing concentrations of miniproteins. (C) Suppression of cAMP signaling monitored by the BRET-based cAMP sensor CAMYEL. The structure of CAMYEL is generated using AF3. (D) CCR9 induced cAMP suppression following stimulation with a titration of CCL25 (black) or a fixed concentration of CCL25 (20 nM) preincubated with a titration of miniproteins. Values represent mean ± SD of three independent experiments performed in duplicate, normalized to min and max CCL25 response.

CCL25-mediated GRK3 recruitment to CCR9 is suppressed by VUP25101, VUP25107, and VUP25112, but not VUP25111.

(A) Schematic presentation of the GRK3 recruitment BRET setup to detect GRK3-Nluc translocation to activated CCR9-mV. (B) Recruitment of GRK3 to CCR9 measured by BRET across a concentration range of either CCL25 (black) or VUP miniproteins with a constant 20 nM CCL25. Values represent mean ± SD of three independent experiments performed in duplicate, normalized to min and max CCL25 response.

All VUPs inhibit CCL25-promoted internalization of CCR9.

(A) Schematic representation of the BRET assay to track CCR9 internalization by monitoring the resonance energy transfer between CCR9-RlucII and rGFP-CAAX. The BRET between the probes decreases as the receptor trafficks away from the membrane through internalization. (B) Internalization response measured by the BRET setup presented in (A) across a range of concentrations of CCL25 (black) or VUP miniproteins with 20 nM CCL25. Values represent the mean ± SD of three independent experiments performed in duplicate, normalized to min and max CCL25 response.

MOLT-4 chemotactic response towards CCL25 is suppressed by VUP binding.

(A) Percentage of migrated cells towards CCL25 across a titration of chemokine concentrations. (B) Percentage of migrated cells towards a fixed concentration of CCL25 (260 nM, corresponding to the highest migration observed in (A)) preincubated with miniproteins (10 µM). Values represent the mean ± SD of three independent experiments measured in technical triplicates by flow cytometry. Significance was determined using Welch ANOVA followed by a Dunnet’s T3 multiple comparisons test comparing to no VUP. *p<0.05.

Mechanisms by which VUPs (VU miniproteins) modulate CCL25 signaling at CCR9 and ACKR4.

(A) Canonical engagement of CCL25 with CCR9 produces a balanced signaling response, activating both G protein-and β-arrestin2-dependent pathways. Signal suppression occurs when miniproteins (VUP25101, VUP25107, and VUP25112) sterically block CCL25:CCR9 interactions, leading to reduced activation of both G protein and β-m arrestin2 signaling. Signal modification can also arise from co-binding of the CCL25:miniproteins complex to the receptor (VUP25111), resulting in biased signaling characterized by selective attenuation of the β-arrestin2 pathway while preserving G protein-mediated signaling for CCR9. (B) Canonical engagement of CCL25 with ACKR4 results in β-arrestin2 signaling. Signal suppression occurs when miniproteins (VUP25101, VUP25107, and VUP25112) sterically block CCL25:ACKR4 interactions, leading to reduced β-arrestin2 signaling. Miniprotein VUP25111 does not inhibit activation of ACKR4 by CCL25. The complex representing G protein is taken from PDB 9KO4 [45], the structure of β-arrestin2 was generated using AF3.