NEW YORK – Researchers from Yale University and Karolinska Institutet and their collaborators have developed two new spatial biology methods that enable simultaneous profiling of the epigenome and transcriptome in mammalian tissues.
Described in a Nature study published earlier this month, the methods, named spatial ATAC–RNA-seq and spatial CUT&Tag–RNA-seq, open the door for researchers to study genome-wide epigenetic modifications and gene expression side by side in order to gain a better understanding of gene regulation events within complex tissues.
"Once you know the epigenetic modifications, you also want to know whether or not they actually control the [gene] expression," said Rong Fan, a biomedical engineering professor at Yale and a corresponding author of the study. "I really think this is a home-run tool to understand the biology of gene expression."
Both spatial ATAC–RNA-seq and spatial CUT&Tag–RNA-seq are built upon previous tools developed by Fan’s lab, using the deterministic barcoding in tissue sequencing (DBiT-seq) technology.
For spatial ATAC–RNA-seq, the method combines spatial-ATAC-seq (spatially resolved chromatin accessibility profiling of tissue sections using next-generation sequencing) with spatial transcriptomics analysis, resulting in a workflow that can jointly profile chromatin accessibility and mRNA expression. Similarly, spatial CUT&Tag–RNA-seq is a cross between spatial-CUT&Tag, which was developed for genome-wide profiling of histone modifications, and spatial RNA-seq.
To integrate spatial-ATAC-seq or spatial-CUT&Tag with RNA sequencing, the researchers designed a co-barcoding scheme where they sequentially place two microfluidic channel array chips with flows perpendicular to each other onto a tissue section. The microchannels each release a spatial barcode onto the tissue, generating a two-dimensional grid of spatially barcoded tissue pixels, each labeled with a unique combination of barcodes.
"I want to emphasize how versatile this in-tissue barcoding approach is," Fan said. "As long as you can get initial molecular information tagged … we're using the same microfluidics to add a spatial address code to those tags for epigenome and transcriptome."
Fan noted that DBiT-seq can also achieve joined protein, epigenome, and transcriptome analysis, for which he has already started generating data.
Because spatial epigenome and transcriptome tools are not naturally compatible with each other, Fan said, his team has done a lot of protocol development to come up with the new integrated workflows.
"It’s not trivial," he said. "There's a lot of optimization we have to do to find a balanced condition that works for epigenome and transcriptome."
For instance, while spatial transcriptomics analysis requires the tissues to be fixed, the epigenomics workflow needs the samples to remain fresh. As a result, the researchers deployed a much lower paraformaldehyde concentration for fixing, Fan said.
In addition, to enhance resolution for the tissue area covered, the researchers developed a microfluidic barcoding chip that is able to achieve a 100 × 100 barcode scheme, dividing a tissue section into 10,000 pixels, each measuring 20 μm. As such, the platform can cover an area of about 16 mm2 at nearly single-cell resolution, the authors noted.
"We didn't compromise," Fan said. "Actually, we did it even better in terms of throughput and the content of the data."
The researchers applied the two new techniques to embryonic and juvenile mouse brains and adult human brain hippocampus in order to investigate the causal relationships between epigenetic state and gene expression, and their implication for cell regulation.
By co-profiling chromatin accessibility and RNA, for instance, the authors mapped more than 21,000 statistically significant correlations between epigenetic regulatory elements and target genes in the juvenile mouse brain.
Moreover, the researchers are now able to perform pseudo-time analyses to gain insights into the cell differentiation process. "Now with two different omics, you can see which cells are already epigenetically primed. That's something people hypothesized but there was no way you could prove [this previously]," Fan noted.
Even more surprisingly, the team found that some genes characteristic of oligodendrocyte precursor cells remained accessible through open chromatin even though their expression started to diminish. "That's a total surprise," Fan said. "I think this is a very cool finding. The only way you can see this is from this multiomics data."
"A lot of times we look at epigenomic modifications, it's like, ‘What are the consequences?'" said Claudia Lalancette, managing director of the epigenomics core at the University of Michigan Medical School who was not involved with the study. "Having the transcriptome and epigenome layers [together] allows you to answer that within the tissue context."
Lalancette said that by enabling spatial epigenome and transcriptome co-profiling, the new tools can help researchers gain insight into various biological questions, such as gene expression mechanisms in the cancer environment.
Despite their promise, Lalancette said for these new methods to be adopted in her lab, the microfluidics platform used to deliver the in-tissue barcodes needs to be commercialized first, given that it can be challenging for her team to build one in house.
Once the platform is available, she said, her team will be "very excited" to try the new tools. "I'd love to be able to provide this to my investigators," she said. "I have already told my institution that I want this [technology]."
The study authors have already filed a patent related to the new methods. In addition, Fan serves as the scientific founder and adviser of AtlasXomics, a Yale spinout that is offering commercial solutions for spatial-ATAC-seq and spatial-CUT&Tag.
While Fan did not disclose whether AtlasXomics has any plans to commercialize the new methods, he said it is "not too difficult" for the company to offer spatial epigenome-transcriptome co-profiling commercially.
In terms of downstream analysis, because joint profiling of the epigenome and transcriptome generates somewhat of a new data type, there is still a need to develop specific data pipelines. "I would welcome people to jump into our data and come up with better approaches to do this type of data analysis in a much more standardized manner," Fan said.
Moving forward, he said his team will continue to use the new methods in areas such as epigenetic mechanism of senescence and aging.
"I just think this is a better tool to study tissue biology," Fan said. "My next step is to utilize these technologies to better understand some interesting biological questions."