Skip to main content
Premium Trial:

Request an Annual Quote

New York Genome Center Team Develops Single-Cell Method to Analyze Transcriptome, Surface Proteins


SAN FRANCISCO (GenomeWeb) – Researchers from the New York Genome Center have developed a method that allows them to analyze transcriptomes and cell surface proteins from single cells.

In a study published today in Nature Methods, the team demonstrated that the method can characterize single cells using either 10x Genomics' platform or the Drop-seq protocol on an in-house designed microfluidic device.

The method can enable researchers to "understand what cells are doing in a more detailed way," Marlon Stoeckius, research scientist at NYGC's Technology Innovation lab and lead author of the study, said. The single-cell field has been growing, but there has not yet been a method to "measure more than just the RNA," he added. With single-cell RNA sequencing, "you lose the knowledge of surface proteins that you get from flow cytometry," Stoeckius said. I thought it would be great to get these two worlds together."

The researchers have filed for a patent on the method and are interested in working with an industry partner to commercialize it.

The method, cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq), relies on oligonucleotide-labeled antibodies with a poly-A tail. Cells encapsulated into droplets are lysed and both the mRNA and antibody-derived oligos are indexed with a shared cellular barcode during reverse transcription. Stoeckius said that the poly-A tail on the antibody was the "main crux" to the method because it mimicked an mRNA molecule, enabling it to be reverse transcribed, sequenced, and read out.

The cDNAs and antibody-derived tags are separated and converted into sequencing libraries independently.

To test the method, the team first designed a proof-of-principle experiment in which they analyzed both human and mouse cells along with a highly expressed, species-specific marker CD29. Just over 97 percent of droplets that were identified as either human, mouse, or mixed by transcriptome data also had the same classification via antibody-derived tag count. In addition, they ran the experiment using both the Drop-seq protocol and 10x Genomics' platform and obtained comparable results.

Next, they wanted to see how the CITE-seq method for protein detection performed in comparison to flow cytometry. They performed CITE-seq and flow cytometry on a population of peripheral blood mononuclear cells and demonstrated that the distributions of cell populations based on marker protein expression were very similar, with both methods identifying subsets of T cells, B cells, plasmacytoids, myeloid dendritic cells, and monocytes.

After validating that the method works, the researchers tested it on just over 8,000 single cord blood mononuclear cells, analyzing the transcriptomes of the cells along with 10 cell surface proteins.

The team found that they were able to identify the different cell types based on both the transcriptome and cell surface markers. In addition, they were able to better distinguish between two similar subpopulations of natural killer cells. The method "lets us understand what the cell is doing in a more detailed way," Stoeckius said. He added that one application of the method would be to characterize immune cells from tissue biopsies — using both the surface markers and transcriptome data to identify the cell and understand gene expression.

Peter Smibert, co-manager of NYGC's Technology Innovation lab, said that the method could be used in conjunction with a number of different single-cell techniques. Aside from Drop-Seq and 10x Genomics' instrument, he said it could likely be used in conjunction with an in situ single-cell method that relies on combinatorial indexing and was developed at the University of Washington. It should also work on Becton Dickinson's Resolve instrument, the Illumina Bio-Rad Single-Cell Sequencing Solution, as well as with 10x Genomics' recently launched single-cell kit for analyzing immune cells.

Smibert declined to disclose whether the group was in discussion with any of these companies to commercialize the method, but said that one option he envisioned would be to partner with a company to develop CITE-seq as an "add-on reagent" to an existing system.

In the team's work with the Drop-seq method, Smibert said that the NYGC group has been using an in-house developed microfluidic device that it described earlier this year in a paper on the BioRxiv preprint server. In that paper, the team described a device that could be assembled from 3D printed parts for around $540. Smibert added that the NYGC team now performs all its Drop-seq experiments on that device and that it performs equivalently to the original lab set up.

Going forward, Smibert said the team plans to continue to use the method to better profile single cells, primarily from blood. "There are some cell types that are easily distinguishable by protein markers but not so easy to distinguish using transcriptomics," he said. "But, with multi-modal data, we can look at both."

Another advantage, he said, is that in theory there's not a limit to the number of markers that can be analyzed. In the Nature Methods study the team analyzed 10 protein markers, but Smibert said that could easily be scaled up. "You could theoretically make as many barcodes as there are antibodies," he said. That would enable both better characterization and understanding of cells as well as potentially the discovery of new cells. "I wouldn't be surprised if we find new subdivisions of cells or more defined cell types," he said.