A workflow to reduce red blood cell autofluorescence for imaging IL-4/IL-4R interactions in the inflamed lung

Autofluorescence is a common challenge in fluorescence microscopy of inflamed tissue, particularly where red blood cell (RBC) autofluorescence can mimic true fluorescence signals. This complicates image interpretation, reduces counting accuracy, and undermines reproducibility. Reliable methods are needed to distinguish true signals from autofluorescence, while preserving tissue architecture. In this study, we address this challenge by developing an optimized workflow that integrates antigen retrieval, selective fluorophore choice, and spectral separation.

Autofluorescence hinders the detection of protein interactions in inflamed tissues by generating signals indistinguishable from true positives, leading to misinterpretation and errors. This study aims to establish and validate a reproducible workflow that enables discriminating between RBC autofluorescence and true proximity ligation assay (PLA) signals in paraffin-embedded inflamed lung tissue. Our approach combines antigen retrieval, selective fluorophore excitation, spectral unmixing, and post-processing to minimize autofluorescence while maintaining tissue integrity. We demonstrate the workflow's efficacy using the PLA, a fluorescence-based technique in which antibody-coupled DNA probes generate a detectable signal only when two proteins are located within nanometer-scale proximity. Using this approach, we detected PLA signals suggesting spatial proximity between interleukin-4 (IL-4) and its receptor (IL-4R) on RBCs within inflamed lung tissue, a context in which RBC autofluorescence is particularly problematic.