BALTIMORE – Researchers from Tokyo University of Science and their collaborators have developed a new cDNA amplification and sequencing technique that promises to improve the efficiency of single-cell RNA sequencing compared with existing approaches.
The method, called terminator-assisted solid-phase cDNA amplification and sequencing (TAS-Seq), is built upon previous scRNA-seq methods that utilize a template-independent terminal transferase for solid-phase cDNA amplification using a nanowell bead-based platform. By leveraging dideoxynucleotide triphosphates (ddNTPs) to achieve stochastic terminations during cDNA synthesis, TAS-Seq can reportedly reduce the technical difficulties of the terminal transferase reaction, yielding scRNA-seq data with high sensitivity and accuracy.
"We aimed to create [a] more sensitive, precise, stable, easy-to-use, scalable scRNA-seq method," Shigeyuki Shichino, a researcher at Tokyo University of Science whose team described TAS-Seq in a study published in the journal Communications Biology last month, wrote in an email.
According to Shichino, TAS-Seq was originally conceived to help his team acquire more precise and deeper scRNA-seq data for a previous project that started in 2017 to construct a single-cell atlas of multiple inflamed tissues.
Currently, many widely used scRNA-seq methods, such as the 10x Genomics Chromium platform and Smart-Seq2, employ template-switching reactions for cDNA synthesis, where Moloney murine leukemia virus reverse transcriptase (MMLV-RT) and template switching oligonucleotides (TSO) are deployed to synthesize the second cDNA strand.
However, Shichino said the efficiency of template switching-based scRNA-seq methods can be hindered by the 5' structure of the RNA template. To account for that, scientists previously developed methods, including Quartz-Seq and Quartz-Seq2, that circumvent template-switching by using a template-independent polymerase called terminal transferase (TdT) to add homopolymer tails to the 3' end of the cDNA.
Still, due to the tricky nature of TdT, one disadvantage of the current TdT-based scRNA-seq methods is that they require stringent control of the TdT reaction conditions — including reaction time, temperature, primer amount, and enzyme activity — to suppress excessive byproduct generation derived from free RT primers, Shichino added.
To overcome this challenge, Shichino's team developed TAS-Seq, which performs solid-state cDNA synthesis on exonuclease I-treated BD Rhapsody magnetic beads. By spiking in ddNTP, specifically dideoxycytidine triphosphate (ddCTP), TAS-Seq can effectively suppress excessive elongation of the polyC tail by the TdT enzyme, eliminating the need for strict temperature and time control of the TdT reaction, Shichino noted.
To test the performance of TAS-Seq, the researchers applied the method to murine and human lung samples. The results showed that TAS-Seq yielded scRNA-seq data that were "highly correlated" with flow-cytometric data. In addition, TAS-Seq demonstrated "higher gene detection sensitivity" and "more robust detection" of important cell-cell interactions and growth factors compared with the commonly used scRNA-seq methods such as 10X Chromium and Smart-Seq2.
Shichino said the team also internally tested the batch effects of TAS-Seq using human peripheral blood mononuclear cells. According to him, the results showed that the batch effects between different experiments were "very limited," and no batch effect correction is necessary when analyzing the datasets.
Furthermore, the researchers showed that TAS-Seq is compatible with antibody-based cell hashing technology, where researchers mix cells from different samples that are labeled with oligo-tagged antibodies and analyze them on the same scRNA-seq platform, thus allowing sample multiplexing within the same scRNA-seq experiment.
TAS-Seq is "an exciting alternative" to the currently available scRNA-seq strategies, said Junyue Cao, head of the single-cell genomics and population dynamics laboratory at Rockefeller University.
Judging by the study, Cao said one "key improvement" of TAS-Seq is that it appears to have higher RNA detection efficiency. He explained that traditional scRNA-seq methods that depend on the template-switching mechanism typically require reverse transcription reactions to propagate all the way to the 5' end of the mRNA template during first-strand cDNA synthesis to enable cDNA amplification downstream. However, because not all reverse transcription reactions can successfully make it to the end of the mRNA molecules, some cDNA molecules may drop out during the experiment.
In contrast, by optimizing the extension length of the TdT reaction, which adds a homopolymer tail at the end of the first-strand cDNA independent of the RNA template, Cao said TAS-Seq upgrades the existing TdT-based scRNA-seq approaches while achieving higher efficiency for transcription detection.
Additionally, Cao considers TAS-Seq's compatibility with cell hashing "another great feature" of the technology.
Besides TAS-Seq's promises, Cao said one remaining question he has is cost, as he did not see a "very obvious" cost comparison between this method and other commonly used ones.
Shichino said most of the cost of TAS-Seq stems from the reagent cost for BD Rhapsody magnetic beads. According to him, his lab usually analyzes 20,000 cells per run, which costs JPY 210,000 (approximately $1,530) for the BD Rhapsody beads, or JPY 10.5 (about $.08) per cell. Compared with the cost of 10X Chromium, which is between JPY 13.1 to JPY 25 ($.10 to $.18) per cell, Shichino said, TAS-Seq is "slightly cheaper."
In terms of the sample types compatible with TAS-Seq, Shichino said because the method relies on a bead-based nanowell scRNA-seq platform for cell isolation, the input material is restricted to live cells, cell nuclei, or non-crosslinked fixed cells, such as methanol-fixed cells.
As for turnaround time, a typical TAS-Seq workflow usually takes three days, although he said it largely depends on the amount of RNA from the starting cells.
Moving forward, Shichino said his group is performing "extensive optimization" of the TAS-Seq workflow, which they had not done for the published study. While the team has already doubled the method's sensitivity compared to the original version in terms of detected gene number, he said further investigation is necessary to improve the method's reaction conditions — including buffer compositions, primer structures, and DNA polymerases — while broadening its applications.
Lastly, Shichino said the team is also trying to apply TAS-Seq to other solid-support single-cell platforms, including the 10X Genomics Visium. "We believe that expanding [the applications of] TAS-Seq might be helpful to acquire deeper single-cell multiomics data," he added.