CUT&RUN and CUT&Tag Sample Prep from 3D Cultures

Team Dynamic Matrices | 15 August 2026

ChIP, CUT&RUN and CUT&Tag were developed on monolayer cells and dissected tissue, where you can get intact cells or nuclei with no polymer in the way. Organoids and hydrogel-embedded cultures add a matrix, a dissociation step and a smaller cell number, and each of those changes what the assay needs. We get asked about this often enough that here is our full answer.

Which assay, and what each needs from the sample

ChIP-seq crosslinks protein to DNA with formaldehyde, shears chromatin, immunoprecipitates. It needs fixed cells, tolerates rough handling after fixation, and needs millions of cells (Landt et al., Genome Research, 2012, ENCODE guidelines).

CUT&RUN (Skene and Henikoff, eLife, 2017) binds an antibody to native chromatin in permeabilised cells or nuclei, then a protein A-MNase fusion cuts around the bound site and the fragments diffuse out. No crosslinking, no shearing, works from 100,000 cells down to a few thousand.

CUT&Tag (Kaya-Okur et al., Nature Communications, 2019) replaces MNase with a protein A-Tn5 that inserts adapters directly. Even lower input, down to single cells, with a library ready without ligation.

Both CUT methods depend on intact nuclei with native chromatin. That is the constraint 3D cultures push against.

For ChIP: crosslink before or after dissociation?

Crosslinking intact organoids preserves native chromatin contacts better than crosslinking after dissociation, because dissociation itself perturbs transcription (see our dissociation stress post). But formaldehyde has to diffuse through the matrix and into the organoid, which takes longer than through a monolayer. Under-fixation of the core gives systematic bias toward the periphery; over-fixation reduces shearing efficiency and epitope accessibility.

Our approach: remove the matrix first if it dissolves gently (nuclease for DNA-crosslinked gels; cold depolymerisation for basement membrane extract), then crosslink the intact organoids, 1 percent formaldehyde, and titrate time (10, 15, 20 minutes) against shearing efficiency and a positive-control ChIP. The optimal time depends on organoid size and has to be found empirically. Do not assume the monolayer's 10 minutes.

For CUT&RUN and CUT&Tag: dissociation is the hard step

No crosslinking penetration problem, but a stricter requirement: nuclei must survive dissociation intact with native chromatin. Mechanical shear that is fine for flow cytometry can damage nuclear envelopes enough to lose chromatin and inflate background.

What we do:

1. Dissolve the matrix without touching the cells. For DyNAtrix, a 10 to 15 minute nuclease step at 37 °C; for BME, cold depolymerisation with washes.
2. Gentle enzymatic dissociation to single cells, TrypLE or a cold-active protease, with minimal trituration (10 passes through a wide-bore tip, no more).
3. Check nuclear integrity immediately by DAPI or trypan blue on nuclei: round, intact, no leakage. If more than 10 percent are damaged, back off the dissociation.
4. Bind to concanavalin A beads for CUT&RUN or CUT&Tag as usual. Organoid cells bind well once free of matrix; residual matrix protein interferes with binding, so clear it.

Alternatively, isolate nuclei directly from organoids with a gentle detergent lysis (0.1 percent NP-40 on ice) after matrix removal, skipping single-cell dissociation. For stress-sensitive lineages this is often better.

Chromatin yield runs low

Organoid-derived chromatin gives less than the same nominal cell number from a monolayer, because of incomplete recovery from the matrix, losses at each step, and lower per-cell accessibility from dense packing. Pilot the yield: count cells or nuclei going into the assay and quantify library output, and compare to your monolayer baseline. Discovering insufficient input after library prep wastes the antibody and the organoids.

Adopt low-input protocol variants from the start: adjusted bead ratios, longer antibody incubations, tuned PCR cycles for CUT&Tag (Kaya-Okur et al., Nature Protocols, 2020). They exist for sorted rare populations and they suit organoids.

QC checkpoints to add

• Nuclear integrity after dissociation (before binding to beads)
• Crosslinking efficiency for ChIP, by shearing profile on a Bioanalyzer
• Input cell or nucleus count recorded per reaction
• Library fragment distribution compared to a monolayer baseline for the same antibody
• A positive-control antibody (H3K27me3 or H3K4me3) alongside every experimental one, to separate sample failures from antibody failures