Limitations of Semi-Automated Immunomagnetic Separation of HLA-G-Positive Trophoblasts from Papanicolaou Smears for Prenatal Genetic Diagnostics.

Background: In prenatal genetic diagnostics, the detection of single-gene defects relies on chorionic villus sampling (CVS) and amniocentesis, which carry a miscarriage risk of 0.2-0.3%. To mitigate this risk, fetal trophoblasts have been isolated from a Papanicolaou smear using Trophoblast Retrieval and Isolation from the Cervix (TRIC). However, this method is labor-intensive and has been shown to be challenging to implement in clinical practice. Here, we describe our experiences in using semi-automated immunomagnetic cell sorting for isolating trophoblasts from clinically obtained Papanicolaou smears during ongoing pregnancies. Methods: Using HLA-G-positive Jeg-3 and HLA-G-negative HeLa cell lines in 10%, 1%, and 0.1% dilutions, we tested and optimized the isolation of HLA-G-positive cells using FACS and semi-automated immunomagnetic cell sorting. We used the latter technique for isolation of HLA-G-positive cells from Papanicolaou smears collected from 26 pregnant women, gestational age between 6 and 20 weeks, who underwent CVS. Results: In four independent dilution series, the mean percentages of Jeg-3 cells went from 7.1% to 53.5%, 0.9% to 32.6%, and 0.4% to 2.6% (7.5, 36, and 6.5-fold enrichment, respectively) using immunomagnetic cell sorting. After sorting of the Papanicolaou smears, HLA-G-positive cells were moderately increased in the positive (14.61 vs. 11.63%) and decreased in the negative fraction (7.87 vs. 11.63%) compared to baseline pre-sorting. However, we could not identify fetal cells using XY-chromosomal FISH in a male sample. Conclusions: Our study supports previous findings that careful sampling of fetal cells from Papanicolaou smears in a clinical context poses significant challenges to cell retrieval.