NEW YORK – A team of scientists at Brown University and their collaborators at PerkinElmer have developed a method for the label-free enrichment of trophoblast cells from cervical samples that may be useful in future noninvasive prenatal tests that rely on rare fetal cells.
Current NIPTs for fetal trisomies and other genomic aberrations analyze cell-free fetal DNA from maternal blood but this is always mixed with large amounts of maternal DNA. Having noninvasive access to fetal cells would be advantageous because the genetic material is pure and contains the entire fetal genome. Cells of fetal origin are known to occur in very small numbers in the blood and cervix of pregnant women and several academic and commercial research teams have been developing methods in recent years for enriching and genetically profiling them.
The Brown University group, which published its results in Scientific Reports last month, focused on the first part, the isolation of extravillous trophoblast cells — which originate from the placenta and invade the endometrium — from a large background of maternal cervical cells. "There is no easy technology available right now which is affordable to hospitals," said Anubhav Tripathi, a professor at the Center for Biomedical Engineering at Brown and the senior author of the study. "They are all either very manual or [involve] counting under a microscope or those kinds of things" he said.
Tripathi said his lab focuses on technologies related to reproductive and maternal health and has collaborated with PerkinElmer, a leader in newborn screening that is also getting into the NIPT space with its Vanadis platform, for several research projects in the past. For this study, PerkinElmer provided his group with cervical samples from standard Pap tests of pregnant women, he said.
His graduate student Christina Bailey-Hytholt, lead author of the study, came up with the hypothesis that intrinsic physical differences between cervical cells and trophoblasts might make them settle at different rates in solution.
As described in the paper, she added cervical samples containing about 72,000 cervical cells that were spiked with about 1,000 cells from a trophoblast cell line, JEG-3, to polystyrene wells of a 24-well plate, allowed the cells to settle for different amounts of time before removing the supernatant, and analyzed the cells at the bottom of the well.
The reason she added JEG-3 cells to the samples, which presumably already contained trophoblast cells since they came from pregnant women, was in order to normalize the number of trophoblasts, which can vary between samples, she said.
She found that the fetal cells appeared to settle to the bottom at a faster rate than cervical cells. After 4 minutes, about 70 percent of the trophoblast cells were attached to the well surface while more than 90 percent of cervical cells were removed with the supernatant, resulting in a 700 percent enrichment of fetal cells.
In a separate proof-of-concept experiment, the researchers transferred cells from the well plate to a slide, stained them for HLA-G, a trophoblast-specific protein, and used CyteFinder and CytePicker technology from RareCyte, to which they had access through PerkinElmer, to pick individual trophoblast cells, including both JEG-3 cells and "real" trophoblasts originating from the pregnancy. This was followed by whole-genome amplification with PicoPLEX and a PCR reaction for two Y-chromosome markers.
Tripathi said his group is now working on automating the process using a microfluidic device, including fetal cell enrichment, cell picking, cell lysis, and PCR or sequencing. It plans another publication on this aspect of the work, which also involves Anita Shukla, a professor of engineering at Brown who is an expert in biomaterials. "To go into the clinic, it's very important that it's automated, that somebody can press a button and things happen," Tripathi said.
The automation work also involves blood samples, in addition to cervical samples, he said, and the goal is be able to process eight patient samples per day in parallel.
There is no intellectual property associated with the enrichment technique. "It's a simple method which we have published and which any industry can use," Tripathi said. "We hope that people can adopt this in their workflow."
However, his group does not plan to develop the method into a clinical test. "For that, we need help from companies like PerkinElmer," he said.
PerkinElmer has already taken in interest in trophoblast-based noninvasive prenatal testing. In 2016, a group led by Randall Armant and Sascha Drewlo at Wayne State University School of Medicine published a different method for the isolation of trophoblast cells from cervical samples, followed by genome profiling. PerkinElmer took an exclusive license to that method at the time but has not commercialized it yet.
In contrast to the Brown team, the Wayne State researchers used magnetic beads and antibody labeling to capture trophoblast cells from cervical samples. "Our method is label-free – we don't use antibodies to cause the enrichment," Bailey-Hytholt said, which is easier.
In the meantime, Armant and Drewlo have cofounded a startup called Cradle Genomics that plans to launch a noninvasive prenatal test. The assay relies on highly pure fetal DNA, presumably from cells of fetal origin. In July, Cradle Genomics raised $17.1 million in Series A funding that was led by Illumina Ventures and Section 32. Drewlo and Cradle Genomics did not respond to a request for comment.
In any case, work to develop noninvasive trophoblast analysis into a clinical test will hinge on the availability of real samples rather than cell lines. "The bottleneck is having access to women's samples," Tripathi said, adding that his group is actively seeking collaborations with hospitals who could provide access to patient samples.