Sickle cell disease (SCD) is a genetic blood disorder caused by a single-point mutation in the β-globin gene (c.20A>T; p.Glu6Val), leading to the production of hemoglobin S (HbS). In low-oxygen environments, HbS polymerizes and causes red blood cells (RBCs) to adopt a sickled shape, triggering vaso-occlusion, hemolysis, and systemic complications. Despite decades of research, the dynamic mechanisms driving sickling remain incompletely understood — partly due to the lack of functional tools capable of capturing these rapid and complex morphological changes in real time.

In this study, we present a robust, automated fluorescence imaging assay designed to monitor RBC sickling ex vivo under oxygen-restricted conditions.

RBCs (5 × 10⁴/well) obtained from SCD patients were plated on 96-well plates coated with PBS-BSA to promote adherence. A coverslip and silicone oil layer were applied to reduce oxygen diffusion. Hypoxia was chemically induced using 2% sodium metabisulfite. In parallel, we also tested an enzymatic system using glucose oxidase and catalase to evaluate a more physiological method of oxygen depletion. Images were captured every 10 seconds for 60 minutes using the ImageXpress Micro XLS system, configured with transmitted light and a DAPI filter (λ: 417–477 nm), optimizing contrast via Hb absorption at 420 nm. To analyze cell morphology, a customized pipeline in CellProfiler was combined with CellPose for AI-driven segmentation, eliminating the need for fluorescent staining.

Over 140 timepoints per cell were analyzed. A comprehensive panel of morphometric parameters—including eccentricity, solidity, compactness, and area—was extracted at each timepoint. Machine learning classification was performed using CellProfiler Analyst. RBCs were manually labeled as “Round” or “Sickled” to train a Random Forest model, which achieved an accuracy of 97.42%. Heatmaps revealed sharp morphometric transitions during the sickling process, while individual cell tracking uncovered diverse behavioral patterns—ranging from immediate sickling to complete resistance.The assay enabled quantification of the percentage of sickled cells over time and the generation of a sickling time distribution across the entire population. These findings highlight distinct subpopulations with varying degrees of resilience to hypoxia-induced deformation. The platform also opens opportunities to explore modulatory factors, such as fetal hemoglobin (HbF). Ongoing experiments with bimodal cHbF distributions aim to determine the threshold HbF content associated with delayed or suppressed sickling.

Altogether, this high-content imaging approach provides a powerful and scalable framework for dissecting the pathophysiology of SCD, offering valuable insights for personalized risk assessment and therapeutic screening. This study was financed, in part, by the São Paulo Research Foundation (FAPESP), Brazil (Process Number: #2022/12856-6).