Skip to main content
Premium Trial:

Request an Annual Quote

New Droplet-Based Single-Cell RNA-Seq Method Enables Transcriptome Profiling of Individual Synapses


NEW YORK – Researchers from Baylor College of Medicine and their collaborators have developed a single-cell total RNA sequencing approach that is able to profile the transcriptome of individual synapses.

Named Multiple-Annealing-and-Tailing-based Quantitative scRNA-seq in Droplets (MATQ-Drop), the method, described in Nature Biotechnology last month, enables high-throughput transcriptome profiling of the synaptic region of a neuron, namely the synaptosome.

“The synapse is one of the most interesting subcellular structures in our system,” said Chenghang Zong, a professor of molecular and human genetics at Baylor and the corresponding author of the paper. Because synapses are crucial for signal transmission between neurons, Zong added, it is important to study their heterogeneity and characterize their individual transcriptome profiles.

Previously, the majority of methods for studying synaptosome transcriptomes were based on bulk assays. The field still lacked an effective way to interrogate the transcriptome of single synaptosomes due to several technical barriers, according to Muchun Niu, a graduate student in Zong’s lab and the first author of this study.

For one, he said, individual synaptosomes are typically small in volume and thus contain small quantities of RNA for analysis. “Normally for a single cell, the diameter is around 10 microns,” Niu explained, “but for single synapses, it is on the order of several 100 nanometers.” Therefore, an assay needs to be highly sensitive in order to capture and study the sparse RNA content within a single synaptosome.

In addition, after individual synaptosomes are isolated, the cellular content needs to be fixed immediately to prevent RNA molecules from leaking out, Niu said. As a result, an RNA-seq chemistry that is compatible with fixed samples is also required. Lastly, the assay needs to be able to simultaneously detect mature and nascent RNA in order to characterize locally spliced transcripts in the synapses.

To address these challenges, the Baylor researchers teamed up with Harvard University physicist David Weitz and developed a specialized microfluidic platform based on a previously developed single-cell total-RNA sequencing method dubbed MATQ-seq.

Initially described in a Nature Methods paper in 2017, MATQ-seq, which was also invented by Zong’s group, deploys primers based on multiple annealing and looping-based amplification cycles (MALBAC). This method enables the detection of total RNA, including noncoding and nonpolyadenylated RNA. Meanwhile, the approach aims to reduce PCR amplification bias during second-strand synthesis by introducing unique molecular identifiers.

In MATQ-Drop, in situ reverse transcription and poly(A) tailing are carried out on fixed nuclei before they are encapsulated in droplets containing barcoded hydrogel beads. Once inside the droplet, barcoded primers are enzymatically released from the beads to bind to the poly(A) tail of the cDNA to perform second-strand synthesis. After that, the material in the droplet is released for PCR amplification and sequencing.

For their study, the researchers applied MATQ-Drop to profile the transcriptome of individual synaptosomes in mouse and frozen human brain samples. They were able to identify different subtypes of synaptosomes and neuron-glia junctions. Specifically, they observed presynaptic and postsynaptic clusters, as well as a special subcluster associated with synapses that are in the process of assembly and maturation.

Compared with mature mRNA-based single-cell platforms, such as the 10x Genomics Chromium, Zong said, one technical advantage of MATQ-Drop is that it can effectively identify nascent RNAs using the reads mapped to intronic regions. In the study, his team characterized the landscape of intron retention for various clusters of synapses.

The researchers also performed a benchmark comparison between MATQ-Drop and the 10x Chromium using single-nucleus transcriptome data of the mouse brain. MATQ-Drop achieved up to 135 percent improvement in gene detection sensitivity across different cell types in comparison to the 10x Chromium platform, they found.

Furthermore, Zong’s team applied MATQ-Drop to help explore the transcriptome of single synaptosomes in an Alzheimer’s disease mouse model. They discovered AD-associated synaptic gene expression changes that were not detected using previous single-cell methods.

With MATQ-Drop, “we are going from bulk tissue to single cell and now to subcellular whole-transcriptome RNA sequencing, which has huge implications,” said Luke Child Dabin, a postdoctoral researcher at Indiana University School of Medicine whose research focuses on neurodegeneration and scRNA-seq. “Potentially, this is the start of a series of developments that could lead us to an even greater subcellular resolution [for transcriptome profiling].”

According to Dabin, who was not involved in the development of the method, one of the strengths of MATQ-Drop is that the method is able to capture the intronic sequences of mRNAs, affording the technology “much more sensitivity” compared with the mature mRNA-based methods.

Additionally, Dabin applauded the method's ability to effectively isolate and analyze individual synapses in situ, which has previously been very difficult to achieve. However, whether these are single or multiple synapses was not demonstrated in the paper, Dabin said, suggesting that a species-mixing experiment may be one way to demonstrate this.

However, despite the method’s promises, Dabin remains unsure whether MATQ-seq can be easily replicated in other labs due to its specialized setup and design. “I think that this method could be quite intimidating to newcomers,” Dabin said. “[The protocol] is a lot more custom — I think that will limit the number of laboratories that can do it.”

“I honestly think if they start a company, and if they start selling kits, that will make it easier for other labs to do this experiment,” Dabin added.

Additionally, given that the experiment can potentially be expensive to perform, Dabin noted that it remains unclear from the paper what the appropriate sequencing depth of the method is, which makes the technology’s sequencing cost “an unanswered question.”

Commenting on the method’s cost, Zong said MATQ-Drop “should be pretty manageable,” and if the procedures are carried out in-house, the method should be cheaper than 10x experiments.

At this point, MATQ-Drop does require labs to build a bespoke microfluidic platform to perform the droplet-based protocols, though. Zong noted that for some labs, microfluidics has already become a commonly used technology, and those who are not equipped to build a platform could potentially seek a collaboration with other labs.

Zong also said that his team, including Niu, is interested in starting a company to commercialize MATQ-Drop, and the group has filed a patent application for the method.

Moving forward, Zong said the team will work to continue to increase the throughput of MATQ-seq to help with large-scale cell atlas construction. Right now, the method can process thousands of synapses, he said, and the goal is to boost that to millions. Aside from neurons, his group is also planning to apply the method to frozen tumor samples, offering an in situ solution to help profile the transcriptomes of individual tumor cells.

“Sensitivity does matter, and we are doing an in situ assay,” he said. “I believe those are all the directions this method can help [with].”