This approach revealed previously unseen, drastic cell identity transitions, namely “identity jumps” in the Waddington landscape, during neurodifferentiation. Following threads will elaborate on HisTrac-seq and and our findings.

While single-cell sequencing such as scRNA-seq are powerful, they inherently capture only static snapshots, and trajectory inference can only approximate dynamic processes. This has created a long-standing need for sc observations across multiple time points—history tracing.

We used Dam (DNA adenine methyltransferase) to label DNA with m6A, a mark absent from our genome. Because previously shown that m6A can persist for several hours until the next DNA replication, we expected it to serve as a long-term ”bookmark” in neurons, potentially for months.

By transiently activating Dam fused to POIs (proteins of interest), HisTrac-seq can record whole-genome multiomic profiles, including transcriptome, open chromatin, LAD-heterochromatin, Trithorax, Polycomb, and even bivalency.

We first demonstrated bulk HisTrac-seq in vitro, showing that transient Dam–POI activation in immature neurons beautifully reproduced their profiles after maturation! In addition, we extended the validation to the mesodermal cardiomyocyte lineage.

We next validated HisTrac-seq in vivo, pulse-labeling the neonatal brain and harvesting in adulthood (8 weeks). This captured neonatal epigenome and transcriptome in the adult brain, demonstrating that HisTrac-seq remains effective for at least two months in vivo!

To fully leverage history tracing, we developed a single-cell m6A detection method on the 10x Genomics platform. By targeting Tn5 transposase via the m6A-tracer, we captured in vivo sc profiles with 5–10K unique reads—among the highest quality achieved for 10x-based assays.

We also achieved sample multiplexing within the 10x Genomics microfluidics, dramatically reducing single-cell library preparation costs to just $0.04 per cell.

So far, so good—perhaps too good. For the final form, we developed scHisTrac-seq, enabling simultaneous profiling of past and present multiomic states from the same single cells via a nanobody-mediated dual scCUT&Tag protocol (past = m6A, present = H3K27ac, transcription marker).

Time for biology: Using scHisTrac-seq to profile transcriptome, chromatin accessibility, and bivalency during neurodifferentiation, we analyzed over 90,000 single cells and uncovered drastic cell identity transitions—namely “identity jumps.

Such abrupt identity jumps were previously unobservable with current single-cell technologies, as trajectory inference relies on continuity between single-cell snapshots.

We next explored the mechanisms underlying this hidden layer of developmental plasticity and found that identity jumps are associated with precocious activation of maturation programs, altered cell–cell interactions and downstream signaling, and resolution of enhancer bivalency.

In summary, we established whole-genome, multimodal single-cell history tracing, applicable to a wide range of experimental systems. Our lab is now applying HisTrac to study id specification and plasticity (e.g., memory) and we look forward to sharing exciting results soon!