Trypsin vs Accutase: Choosing a Cell Dissociation Reagent

Trypsin is a serine protease from the pancreas, supplied for cell culture either as an extract of porcine or bovine pancreas or as a recombinant protein. It hydrolyses peptide bonds on the carboxyl side of lysine and arginine residues, and it works well at 37 C and at slightly alkaline pH.

In dissociation it does two things at once. It digests the extracellular domains of the adhesion proteins that anchor cells to the culture surface and to each other, and it digests everything else on the cell surface that presents an accessible lysine or arginine. The second effect is the whole reason alternatives exist.

The "1:250" on a trypsin label is a potency designation, not a dilution and not a purity statement: it means one part of the preparation digests 250 parts of casein under the standardised assay. It says nothing about what else is in the bottle. Crude pancreatic preparations also carry chymotrypsin, elastase, carboxypeptidases and other pancreatic activities, and they vary between lots. This is worth knowing when a protocol suddenly behaves differently after a new bottle is opened. Recombinant trypsin, produced by fermentation, is a single defined component and does not carry that variability.

The standard formulations are straightforward. Gibco's 0.25% Trypsin-EDTA contains 2.5 g/L trypsin (1:250) with 0.38 g/L EDTA tetrasodium - about 1 mM - in a calcium- and magnesium-free balanced salt solution. The 0.05% version contains 0.5 g/L trypsin with 0.2 g/L EDTA tetrasodium, about 0.53 mM. The EDTA is not filler: it chelates the divalent cations that cadherins and integrins require, so enzyme and chelator attack the same adhesions by different routes.

Accutase is a cell detachment solution from Innovative Cell Technologies. Its active components are described as a mixture of proteolytic and collagenolytic enzymes derived from an invertebrate species. The exact enzyme identities and their concentrations are proprietary and unpublished, so no honest comparison can quote a unit activity for it - anything that does is guessing.

What is disclosed is the matrix. The enzymes are supplied in Dulbecco's phosphate-buffered saline without calcium and magnesium, containing 0.5 mM EDTA tetrasodium and 3 mg/L phenol red. The product contains no mammalian- or bacterially-derived components, which is the basis of its use where animal-origin material is a regulatory problem.

Three practical properties follow from the design. It is active at room temperature as well as at 37 C, so dissociation can be run cooler and slower where that helps viability. It does not require neutralisation with serum or a protease inhibitor - the standard workflow is to dilute in medium and pellet. And it is markedly gentler on surface proteins than trypsin, which is the property most people are actually paying for.

The cost is speed. Accutase generally needs longer than trypsin to release the same monolayer, and for firmly attached lines it may not release them completely at all.

The question is usually posed as trypsin versus Accutase, but two other reagents belong in the decision.

TrypLE is a recombinant trypsin-like protease produced by microbial fermentation and sold as an animal-origin-free replacement for porcine trypsin. Like Accutase, its enzyme concentration is proprietary. It sits between trypsin and Accutase in aggressiveness, is stable at room temperature for storage as well as for use, and is the straightforward substitution when a process needs to drop animal-derived material without redesigning the workflow. It is available with and without EDTA and with and without phenol red.

EDTA alone - sold as Versene at 0.48 mM EDTA in PBS, or as a 0.5 mM EDTA solution in calcium- and magnesium-free saline - contains no enzyme at all. It works purely by chelating Ca2+ and Mg2+, collapsing cadherin-mediated cell-cell junctions and integrin-mediated attachment without cleaving anything. Nothing on the cell surface is digested, so epitope preservation is complete.

The trade-off is that EDTA alone does not fully release cells from matrix in many systems, and it lifts cells in clumps rather than as single cells. For human pluripotent stem cells that is a feature rather than a defect: clump passaging with EDTA is the standard routine method precisely because hPSCs survive better as aggregates than as single cells.

If the cells you harvest are going straight back into a flask, dissociation chemistry barely matters - surface proteins are re-expressed within hours. If the cells you harvest are the measurement, it matters a great deal.

Trypsin cleaves the extracellular domains of many surface proteins used as identity and lineage markers. Epitopes commonly reported as trypsin-sensitive include CD4, CD8, CD14, CD19, CD31, CD62L (L-selectin), CXCR4 and several integrins. Sensitivity depends on the antibody clone as much as on the antigen, though: two antibodies raised against the same marker can behave completely differently after trypsinisation because they recognise different epitopes on it.

The consequence is a flow cytometry result that reports a phenotype the cells do not have. A population scored as marker-negative may simply have had the epitope removed at harvest. This failure is silent - the histogram looks like a genuine negative population, and nothing about the data flags it.

There are three ways out. Use EDTA alone or Accutase for the harvest. Or, where trypsin is unavoidable, let the cells recover in complete medium before staining so the surface proteome is re-expressed. Or validate directly, by staining the same sample harvested both ways and comparing - the only approach that actually proves the point for your particular panel.

A related point that is easy to miss: whichever reagent you choose, keep it constant across an experiment. A treated sample harvested with trypsin and a control harvested with Accutase is not a controlled comparison, whatever else is matched.

Both directions of error are expensive.

Over-digestion - too long, too concentrated, or too warm - damages the plasma membrane. The visible signs are blebbing, granularity, and cells that round up but never form a clean pellet. The measurable costs are reduced viability, reduced plating efficiency at the next passage, and a distorted marker profile. Trypsin does not stop working on its own, which is why the neutralisation step is not optional and why most over-digestion actually happens between the end of the incubation and the moment medium is added.

Under-digestion produces clumps. Clumps ruin haemocytometer and automated counts, which then produce uneven seeding across a plate. In droplet-based single-cell sequencing they cause multiplets and block microfluidics. In flow cytometry they are removed by doublet gating, so the cells that clumped are silently dropped from the analysis - and if clumping is not random with respect to cell type, that is a biased loss rather than merely a smaller sample.

The practical approach is to set incubation time by observation, not by protocol. Check the monolayer under the microscope from the earliest plausible time point and stop when roughly 80-90% of cells have rounded up, which is usually before they have fully detached; a firm tap on the side of the flask releases the rest. Add medium immediately at that point.

Temperature is a real lever. Trypsin at room temperature is much slower than at 37 C, which is occasionally useful for fragile cells. Accutase and TrypLE are designed to work at room temperature.

One further caution: trypsin solutions lose activity with repeated freeze-thaw. Aliquot on receipt rather than cycling a 500 mL bottle through the freezer, or a protocol that worked in January will not work in June and nobody will know why.

Robust adherent lines for routine passaging - HEK293, HeLa, CHO, most fibroblast and epithelial lines. Trypsin-EDTA at 0.25% or 0.05% is the sensible default: fastest, cheapest, and the epitope damage is irrelevant if the cells go back into culture. Use 0.05% for lines that lift readily and 0.25% for stubborn ones.

Human pluripotent stem cells (iPSC and ESC). EDTA alone for routine clump passaging; Accutase or a dedicated stem-cell dissociation reagent when single cells are genuinely required, as for clonal work, transfection or sorting. Single-cell dissociation of hPSCs triggers dissociation-induced apoptosis, and a ROCK inhibitor such as Y-27632 is standard practice to prevent it - Watanabe and colleagues reported cloning efficiency rising from roughly 1% to roughly 27% with it. Trypsin is a poor choice for these cells in either mode.

Primary neurons and other fragile primary cells. Accutase or a gentle enzyme-free reagent. Neurons are unusually sensitive both to over-digestion and to the mechanical trituration that follows it.

Cells destined for flow cytometry on surface markers. EDTA alone if it releases them, Accutase if it does not, and trypsin only with a recovery period and a validated panel.

Mesenchymal stromal cells and other cells for cell-therapy or GMP-adjacent processes. An animal-origin-free reagent - recombinant trypsin, TrypLE or Accutase - removes the porcine material and the adventitious-agent testing burden that comes with it.

Single-cell sequencing preparations. Whatever gives the cleanest single-cell suspension at the highest viability for that tissue, established empirically. Gentler is usually better, but an incompletely dissociated sample is worse than a slightly rough one.

Wash first, with a calcium- and magnesium-free solution. Serum contains alpha-1-antitrypsin and alpha-2-macroglobulin, which inhibit trypsin, so residual serum stalls the digestion. A Ca/Mg-free DPBS wash removes the serum and begins stripping divalent cations out of the junctions. Washing with a Ca/Mg-containing solution at this step works directly against you.

Use just enough volume to cover the monolayer. A thin film is sufficient - typically 1 mL for a T-25 and 2-3 mL for a T-75. Excess reagent costs money and improves nothing.

Neutralise trypsin properly. Complete medium containing serum is the usual method. For serum-free culture, use a defined soybean trypsin inhibitor rather than relying on dilution; dilution slows the enzyme but does not stop it.

Accutase does not need neutralisation, but it does need removing. Dilute in medium, pellet and resuspend rather than leaving cells sitting in it while you do something else.

Pellet gently. Around 200-300 x g for 5 minutes suits most mammalian lines. Spinning harder after an aggressive digest compounds the membrane damage rather than compensating for it.

Keep a written, line-specific protocol covering reagent, concentration, temperature, time and the visual endpoint. Dissociation is the step most likely to drift silently between operators, and it sits upstream of everything else you measure.