This article has been updated to correct the affiliation of the Castelo-Branco lab.
NEW YORK – Research teams in New York and Sweden have independently developed new methods for multimodal, single-cell epigenetic profiling of chromatin. The methods offer a simpler workflow than previous ones, such as CUT&Tag (Cleavage Under Targets & Tagmentation), potentially providing results for more cells from an initial sample.
One of the approaches, nanobody-tethered transposition followed by sequencing (NTT-seq), was developed by Ivan Raimondi of the New York Genome Center and Tim Stuart, a postdoc in Rahul Satija’s lab at New York University and the New York Genome Center. The other method, nanoCUT&Tag (nano-CT), comes out of Gonçalo Castelo-Branco's group at Sweden's Karolinska Institute. Both use nanobodies, antibodies made of a small single polypeptide chain, that are fused to Tn5 transposase. The fusion protein binds specifically to primary antibodies targeting chromatin marks of interest — say, repressive histone marks — and the transposase fragments and barcodes the DNA surrounding the mark to enable detection and mapping by next-generation sequencing.
The groups discovered in March that they were both working on the same idea and have coordinated the release of their preprints and the submission of manuscripts, which are still under review. In its preprint, the New York team used NTT-seq to analyze up to three features at a time, including multiple histone marks and protein-DNA binding sites in single cells. They also extended the method to analyze cell surface protein expression.
The SciLifeLab team used nano-CT to simultaneously profile chromatin accessibility and two histone marks in mouse brain tissue. They reported being able to conduct analyses with as few as 25,000 cells with a 16-fold increase in the number of fragments per cell.
"Everything is going multimodal," said Marek Bartosovic, a postdoc in Castelo-Branco's lab. "It was the obvious next step, to look at more than one histone modification."
With this new approach, "you can see how these different histone marks work together," he added. "Which one is first, which one comes later, how they interact with open chromatin."
Though they've coordinated their papers, it appears the groups haven't cooperated on commercialization. Both have filed for patents, and the New York group, under the umbrella of the Center for Integrated Cellular Analysis, is giving away boxed kits with the nanobody-Tn5 fusion protein already purified.
The kits could help these nanobody-based single-cell chromatin profiling methods — and NTT-seq in particular — take root in a competitive field. "Multiple papers in the past year are doing the same multi-factor profiling, just not with fusions," said Kami Ahmad, a researcher at the Fred Hutchinson Cancer Research Center who led development of CUT&Tag, a method that is largely replacing chromatin immunoprecipitation sequencing (ChIP-seq) and the basis for both new protocols. Some methods, such as antibody-guided chromatin tagmentation for two factors followed by sequencing (ACT2-seq) and MultiCUT&Tag, both published late last year, aggregate the antibodies and Tn5 enzyme before introducing the reagent to the sample. Another approach, exemplified by Multiple Target Identification by Tagmentation (MulTI-Tag), uses oligo-conjugated antibodies.
Ahmad's lab, which also developed CUT&Run, a similar method that used Mnase instead of Tn5, took a similar approach to the New York group by mailing aliquots of their fusion proteins. "We knew people would just do what they'd been doing," he said. "There's just inertia because it's hard to pick up something from a paper, but if someone is going to provide you the reagent, you can try it out, even on a small scale."
Ahmad is "very excited" by the nanobody methods and has requested an NTT-seq kit.
"Nanobodies are extremely well-defined reagents," he said. "We know how they work, and we know from prep to prep that they're going to do the same thing," while larger antibodies can vary from lot to lot.
While nanobodies can specifically bind the IgG antibody targeting the histone mark or cell surface protein of interest, the approach still requires those antibodies to be from separate species, Ahmad said, rabbit and mouse in the case of NTT-seq. "What if you have two rabbit [antibodies]?" Ahmad said.
But one of the biggest advantages is that the nanobody methods "reduce the number of steps dramatically," he said. "One of the hardest things about single-cell epigenomics is keeping cells around. CUT&Tag isn’t that long, but it is multiple steps, and if you’re losing cells at every step, you end up with nothing."
"Especially working with precious materials, it's very sad if you run through and there’s nothing left at the end," he said.
NTT-seq
Satija called NTT-seq the "flagship project" of the Center for Integrated Cellular Analysis, started with a National Human Genome Research Institute Center for Excellence grant awarded in 2020. The five-year grant provides between $2 million and $3 million a year, for a total of up to $15 million. CICA is comprised of 10 labs across six institutions: NYU, Columbia University, Weill Cornell Medicine, NYGC, Rockefeller University, and Memorial Sloan Kettering Cancer Center.
"The neat thing about the program is, we are encouraged to take new directions as they arrive," he said. Initially, they had focused on ATAC-seq (assay for transposase-accessible chromatin by sequencing), but as soon as CUT&Tag arrived in 2019, Raimondi and Stuart were interested in trying it.
"What we're finding when looking at CUT&Tag by itself is that it's hard to make sense of the data when you only have a repressive histone mark," Stuart said. "But when you have a combo, you can see how they change together."
The New York team declined to estimate the cost for each kit or for the method in general. "Some of the major costs would be in protein purification, which can vary depending on how much you produce," Satija said. CICA has already received about 150 orders and will process "as many as we can," he said.
If researchers can run single-cell ATAC-seq, they can also run NTT-seq, including on the 10x Chromium platform, Satija said. Sequencing requirements are "on par" with those existing workflows.
Satija said his group does not want to become long-term suppliers for labs and is interested in commercializing the method. "We have filed a provisional patent covering NTT-seq and some apps of it," said Silas Maniatis, a researcher in the Technology Innovation lab at NYGC. "We've had some conversations with commercial entities over possibilities of licensing it." Out-licensing would be "ideal," Satija said, suggesting that he would like to see the method commercialized in the same manner as CITE-seq, a single-cell transcriptomics assay that also includes cell surface protein data and was licensed to BioLegend.
"This is a fantastic initiative from them," Bartosovic said of the kit launch, though he said he won't be able to order a kit because they've limited their distribution to the US.
Nano-CT
Developing nano-CT was a natural outgrowth of work on a single-cell version of CUT&Tag, which the Castelo-Branco lab published in Nature Biotechnology in April 2021, Bartosovic said.
They came up with the nanobody-Tn5 fusion design in March of that year and got it working just a few months later. In their preprint, they said they were able to infer chromatin velocity between accessible chromatin and H3k27ac histone modification in oligodendrocytes, to deconvolute H3K27me3 repressive states, and to "infer two sequential waves of H3K27me3 repression at distinct gene modules during oligodendrocyte lineage progression."
The group has also submitted a patent application, Bartosovic said. "We don’t have any immediate commercialization plans, but we’ll keep this option for the future." Like the NTT-seq group, his team has submitted plasmids for the fusion protein to AddGene.
Next Steps
Aside from distributing kits, the NTT-seq team is looking to expand the number of markers it can assay at a time, Raimondi said. The researchers are considering using multiplexed antibodies, as well as different rounds of staining and tagmentation. Every additional step up in multiplexing adds the ability to integrate datasets. "You'd have a common feature to use in data analysis," he said.
How other groups may use these multimodal assays remains to be seen, but Ahmad suggested cell trajectories is an obvious one. "A lot of things come into play when you're making decisions. We're very good and have a lot of descriptions of cell types in the human body. We have them refined down to morphology, gene expression, etc. But those are the end states. For all those cell types, earlier on, [they] could have gone one way or another."