Background

The SLIT–ROBO pathway is a conserved signaling axis that regulates cytoskeletal remodeling, axon guidance, and epithelial cell migration [1–4]. Aberrant activation of this pathway has been implicated in cancer progression, angiogenesis, and immune modulation [5–8]. Despite its therapeutic relevance, no small-molecule modulators targeting SLIT2 have been reported until recently, largely due to the extracellular and glycosylated nature of SLIT2 that complicates biophysical screening [7].

To overcome these limitations, we established an orthogonal, immobilization-free screening platform combining temperature-related intensity change (TRIC) for primary binder discovery and time-resolved Förster resonance energy transfer (TR-FRET) for functional inhibition validation of the SLIT2–ROBO1 interaction. In this pipeline, the TRIC screen is followed by concentration-response affinity determination (Kd) for selected hits and TR-FRET concentration-response testing to report relative IC₅₀ values [9,12–14].

This combination leverages two complementary principles: First, TRIC (NanoTemper Dianthus) measures micro-environmental fluorescence changes induced by molecular binding under infrared (IR)-mediated thermal gradients. Because it requires no immobilization and is compatible with crude or glycosylated proteins, TRIC is well-suited for detecting weak or transient interactions typical of extracellular PPIs [11–14]. Second, TR-FRET [homogeneous time-resolved fluorescence (HTRF) format] detects the disruption of a specific protein–protein interaction using donor-acceptor energy transfer between terbium (Tb3+) and d2 fluorophores. A time delay (integration delay) is applied after excitation so that short-lived background signals decay before detection, while the long-lived terbium emission is retained. The assay is read ratiometrically as a 665/620 nm emission ratio, which improves robustness to well-to-well intensity variation and helps reduce sensitivity to nonspecific intensity fluctuations [9].

Dimethyl sulfoxide (DMSO) content is an important practical variable in both thermophoretic and fluorescence-based assays. In this workflow, DMSO was evaluated across a 0%–10% range, and intermediate levels (approximately 2%–5%) provided the most consistent signal-to-noise under the conditions described here. Lower than 2% or higher than 5% DMSO introduced increased noise in our hands (data not shown). For screening, we therefore fix the in-well DMSO at 2% across all wells and controls to minimize viscosity and matrix-driven variability.

Proteins were sourced in formats compatible with both TRIC labeling and TR-FRET detection. SLIT2 is used as a His-tagged recombinant protein to enable selective TRIC labeling using an NTA-based dye (RED-tris-NTA 2nd Gen) and recognition by the anti-6His d2 acceptor antibody in the TR-FRET assay. ROBO1 is used as an Fc-tagged recombinant protein so that the complex can be detected by the anti-human IgG-Tb donor antibody. A specific consideration of the ROBO1-Fc format is Fc-mediated dimerization, which can increase apparent avidity; however, under the low nanomolar protein concentrations used in this protocol, we did not observe obvious assay artifacts attributable to Fc valency. Reported potencies should be interpreted in the context of the constructs and assay geometry used here. If a monomeric ROBO1 construct is available, it can be substituted, but the TR-FRET signal window and protein ratios should be re-optimized because tag position and spacing influence FRET efficiency.

The workflow (as shown in the Graphical overview) proceeds as follows:

• Step 1: TRIC single-point screening of a focused small molecule library (TargetMol Lipid Metabolism Library) against fluorescently labeled SLIT2 in a 384-well format using NanoTemper Dianthus (Figure 1).

• Step 2: Removal of assay-interfering compounds by filtering wells flagged for autofluorescence, quenching, or aggregation using instrument quality control outputs, yielding preliminary hits.

• Step 3: Confirmation of binding and determination of binding affinity (Kd) by multi-point TRIC/microscale thermophoresis (MST) dose response measurements using NanoTemper Monolith X.

• Step 4: Orthogonal TR-FRET assay to evaluate the effect of confirmed binders on SLIT2–ROBO1 association, generating concentration response curves and reporting relative IC₅₀ values (Figure 2).

This platform enables quantitative and reproducible triage of early hits while minimizing false positives due to fluorescence artifacts, aggregation, or thermal instability. As a representative example, we previously identified bexarotene as a reproducible SLIT2 binder (Kd = 2.62 μM) that partially inhibits SLIT2–ROBO1 binding (IC₅₀ ≈ 77 μM). This example underlies the optimization steps detailed below and serves as a benchmark for troubleshooting and expected performance metrics [1,2].

Figure 1. Temperature-related intensity change (TRIC) screening steps overview. Schematic overview of the four-step screening pipeline used to identify SLIT2 binders by TRIC and confirm binding by dose response. (1) Compound library selection and preparation of intermediate dilution plates and final assay plates. (2) Primary single-point TRIC screen on the Dianthus platform using a 5 s laser-on time per well to measure changes in fluorescence signal (Fnorm) relative to vehicle controls. (3) Hit filtering based on fluorescence behavior and magnitude of signal change, retaining compounds that exceed a predefined threshold (for example, ±3× the standard deviation of the reference control distribution) while removing wells flagged for autofluorescence, quenching, or aggregation. (4) Concentration response testing of shortlisted hits using TRIC and microscale thermophoresis (MST) on the Monolith X to generate binding curves and quantify affinity from fitted dose response data.

Figure 2. TR-FRET screening overview. (1) Test compounds are dispensed into assay wells (2 μL; 20% DMSO stock) to achieve the desired final compound concentration and DMSO content after all additions. (2) Tagged proteins are added (4 μL of each) to form the complex, using the second LRR domain of hSLIT2 carrying a His tag (25 nM) and the first Ig domain of hROBO1 carrying an Fc tag (25 nM). (3) Detection reagents are added (5 μL of each): anti-His d2 (acceptor) and anti-human IgG terbium (donor) to enable time-resolved FRET between the two proteins when in proximity. (4) After incubation (1 h), the TR-FRET signal is measured as the 665 nm to 620 nm emission ratio, where inhibitor-mediated complex disruption reduces the ratiometric signal relative to vehicle controls.