Trypan Blue and the Dye Exclusion Viability Assay

Trypan blue (CAS 72-57-1, C34H24N6Na4O14S4, molecular weight 872.88 g/mol) is a bis-azo dye originally developed as a stain for trypanosomes, which is where the name comes from. In cell culture it is used for one job: distinguishing cells with an intact plasma membrane from cells without one.

The dye carries four sulfonate groups and is strongly negatively charged at physiological pH. That charge is the entire basis of the assay. A healthy plasma membrane is essentially impermeable to a large, tetra-anionic molecule of this size, so the dye is excluded and the cell interior stays unstained. When membrane integrity is lost, the dye enters freely and binds to intracellular proteins, giving the cell a distinct blue cast under brightfield illumination.

This is why the technique is properly called a dye exclusion assay rather than a viability assay. What is measured is exclusion of a dye by a membrane. Viability is an inference drawn from that measurement, and the gap between the two is where most trypan blue errors live.

Under the microscope, an unstained viable cell appears bright, round and refractile with a clear halo. A dead cell appears flat, dark blue, often slightly swollen, and loses its refractile edge. The contrast is high enough to score by eye at 100x magnification, which is a large part of why the method has survived essentially unchanged since the 1920s.

Trypan blue for cell counting is supplied as a 0.4% (w/v) solution -- 4 mg/mL -- almost always in isotonic buffer (0.81% sodium chloride and 0.06% potassium phosphate is a common formulation) and sterile-filtered through a 0.22 um membrane.

The 0.4% figure is a working compromise rather than a physical constant:

• Strong enough to score quickly. At 0.4%, dead cells reach full colour saturation within seconds of mixing, so the operator is not waiting for the stain to develop.
• Dilute enough to see through. Trypan blue is intensely coloured. Much above 0.4% the background of the counting chamber darkens to the point where unstained cells lose contrast and small cells become hard to find.
• Compatible with the standard 1:1 dilution. Because the protocol mixes equal volumes of sample and dye, the cells actually sit in 0.2% dye. Supplying the stock at 0.4% means the dilution arithmetic is a clean factor of 2.

Some protocols use 0.2% or 0.1% trypan blue for very small cells or for automated counters with sensitive optics. If you change the dye concentration, the dilution factor in your calculation changes with it, and comparisons against historical counts stop being valid.

Store the solution at room temperature or 2-8 C, protected from light. Trypan blue solutions form fine precipitates over time, particularly after refrigeration; filter through a 0.2 um syringe filter before use if you see particulates, since dye crystals are easily miscounted as small dead cells or debris.

This is the standard manual protocol using a Neubauer-improved hemocytometer. Total time is about five minutes per sample once you are practised.

Before you start. Clean the hemocytometer and coverslip with 70% ethanol and lint-free tissue. Press the coverslip onto the chamber until Newton's rings -- faint rainbow interference bands -- appear at the contact edges. Their presence confirms the coverslip is seated at the correct 0.1 mm height, which is what makes the volume calculation valid.

1. Prepare a single-cell suspension. For adherent cultures, detach with trypsin-EDTA, neutralise, and pipette gently to break clumps. Clumps are the single largest source of manual counting error, because a clump of eight cells is scored as one object.
2. Take a representative sample. Mix the suspension thoroughly but without foaming, then immediately remove an aliquot. Cells settle within a minute or two; a sample drawn from a standing tube will not reflect the culture.
3. Mix 1:1 with dye. Combine 10 uL of cell suspension with 10 uL of 0.4% trypan blue. This gives a dilution factor of 2. Mix by gentle pipetting.
4. Wait, but not long. Let the mixture stand about 1 minute. Do not exceed 3-5 minutes before counting. Trypan blue is cytotoxic, and viable cells progressively take it up on standing, so viability falls the longer the sample sits. A sample counted at 15 minutes can read several percentage points lower than the same sample at 2 minutes.
5. Load the chamber. Pipette 10 uL into the notch at the edge of the coverslip and let capillary action fill the chamber in one continuous movement. Do not overfill into the moat and do not introduce bubbles. If the chamber floods, clean and reload.
6. Let the cells settle for about 30 seconds, then examine at 100x total magnification.
7. Count the four large corner squares. Each corner square is 1 mm x 1 mm with the coverslip 0.1 mm above it, so each encloses 0.1 mm3 = 1 x 10^-4 mL. Count stained and unstained cells separately.
8. Apply a consistent boundary rule. Count cells touching the top and left boundary lines; do not count cells touching the bottom and right. Any rule works as long as it is used consistently, because its purpose is to avoid counting the same cell in two adjacent squares.
9. Check the count is in range. Aim for 20-50 cells per corner square. Below about 20 the Poisson counting error becomes large; above about 100 cells overlap and are undercounted. Dilute or concentrate the sample and repeat if you are outside this window.

Calculate:

``` Cells/mL = (average count per corner square) x (dilution factor) x 10^4

% Viability = (unstained cells / total cells) x 100 ```

The 10^4 term converts the 1 x 10^-4 mL chamber volume to a per-millilitre figure. It is a fixed property of the hemocytometer geometry, not something to adjust.

Worked example. Corner square counts of 42, 38, 45 and 39 unstained cells, plus 3, 2, 4 and 3 stained cells.

• Average total per square = (45 + 40 + 49 + 42) / 4 = 44
• Average unstained per square = 41
• Total cells/mL = 44 x 2 x 10^4 = 8.8 x 10^5 cells/mL
• Viable cells/mL = 41 x 2 x 10^4 = 8.2 x 10^5 cells/mL
• Viability = 164/176 x 100 = 93.2%

Count both chamber sides and average them if you need better precision. The counting error on a single 100-cell total is roughly 10% by Poisson statistics alone, before any operator or sampling error.

Automated trypan blue counters run the identical chemistry with image analysis substituted for the operator's eye. The sample is mixed 1:1 with 0.4% dye exactly as above, loaded onto a disposable slide or a reusable cassette, and a brightfield image is segmented to classify objects by size, circularity and optical density.

What automation genuinely fixes is operator-to-operator variance and the tedium of counting a plate of samples. Two analysts scoring the same slide routinely differ by several percent on viability, and that difference is systematic -- some people are simply more willing to call a pale cell dead.

What automation does not fix is anything upstream. It cannot separate a clump, and it cannot make an unrepresentative sample representative. It also inherits a specific failure mode: debris, dye crystals and cell fragments in the same size range as small cells get classified as cells. A culture in poor condition, full of fragments, can produce an automated count that is materially too high and a viability that is too low.

Practical guidance:

• Set and record the size gate for your cell type, and keep it fixed across an experiment. Changing the gate changes the answer.
• Verify a new instrument or a new cell line against manual counts on the same samples before trusting it.
• Treat viability below roughly 70-80% as a regime where automated classification degrades, because dying cells drift into intermediate sizes and staining intensities.
• Fluorescence-based counters using acridine orange with propidium iodide are considerably more accurate than brightfield trypan blue for samples with heavy debris, primary cells, or blood-derived samples, since the nuclear stains ignore anything without DNA.

Trypan blue exclusion is fast, cheap and adequate for routine passaging. It is genuinely unreliable as a measure of culture health, and the reasons are worth understanding because they determine when you need a different assay.

It detects late death only. Loss of membrane integrity is one of the last events in cell death. A cell that has committed to apoptosis -- caspases active, DNA fragmenting, mitochondria permeabilised -- excludes trypan blue perfectly well for hours. A culture reading 95% viable can contain a substantial apoptotic fraction that will be dead by tomorrow. This is the single most consequential limitation. If you are measuring a cytotoxic treatment, trypan blue will underestimate the effect and will underestimate it most at early timepoints.

It is not a proliferation or function assay. Cells that are viable by dye exclusion may be growth-arrested, senescent, or unable to produce your protein of interest. Membrane integrity says nothing about whether a cell will do useful work.

Dead cells disappear. Cells that have fully lysed or fragmented are not counted at all. In a culture that has been declining for days, the worst-affected cells have already disintegrated, so the surviving population reads deceptively healthy. Viability on a badly crashed culture is a survivorship-biased number.

The dye itself kills. Prolonged contact reduces the measured viability of a real sample. This is why the 3-5 minute window matters and why a batch of samples should be stained one at a time rather than all at once.

Protein binds the dye. Trypan blue binds serum proteins, which raises background staining and reduces contrast. Samples in high-serum medium, and especially protein-rich samples such as ascites or lysates, give poorer discrimination. Washing into PBS before staining improves the read.

Scoring is subjective. Partially stained, pale-blue cells are common in a stressed culture and different operators score them differently. Decide in advance whether pale counts as dead, write it down, and apply it consistently.

Small and low-viability samples are poorly served. Manual counting needs roughly 1 x 10^5 cells/mL to land in the useful counting window, and precision degrades sharply at the extremes.

Match the assay to the question. Trypan blue answers "how many cells do I have and how many are already dead". It does not answer "is this compound killing my cells" or "is this culture healthy".

• Routine passaging and seeding density. Trypan blue. Fast, cheap, gives a concentration and a viability in one operation.
• Cytotoxicity and dose-response. A metabolic assay (resazurin, MTT/XTT) or an ATP luminescence assay. These detect functional decline well before membrane rupture and give a continuous readout suitable for IC50 fitting.
• Distinguishing apoptosis from necrosis. Flow cytometry with annexin V and propidium iodide. Annexin V detects phosphatidylserine exposure on the outer leaflet, which happens early; PI enters only after membrane failure. The two together separate early apoptotic, late apoptotic and necrotic populations, which no single-dye method can do.
• Primary cells, blood products, debris-heavy samples. Fluorescent nuclear staining, typically acridine orange with propidium iodide, on an automated counter. DNA-specific dyes ignore platelets, red cells and debris that confound brightfield counting.
• Release testing of a cell therapy or a bioprocess harvest. Usually more than one orthogonal method, because membrane integrity alone is not accepted as sufficient evidence of a functional product.

In a bioreactor or process-development setting, trypan blue viability tracked over a run remains useful as a trend. The absolute number matters less than the shape of the curve, and a viability that starts falling is a real signal even if its exact value is optimistic.

Viability is much lower than expected. Check the time between staining and counting first -- this is the most common cause. Then check trypsinisation: over-digestion damages membranes directly. Check that the medium and PBS were warmed; cold shock and osmotic stress from a mismatched buffer both show up as reduced exclusion. Finally, check the dye is not past its useful life or heavily precipitated.

Viability is implausibly high on a culture that looks sick. Suspect the survivorship effect described above -- fully lysed cells are not counted. Look at the culture under the microscope before detaching: floating debris and granular, vacuolated adherent cells tell you more than the number does. Consider a metabolic assay.

Counts are irreproducible between replicates. Almost always clumping or sampling. Pipette the suspension to a genuine single-cell state, mix immediately before each aliquot, and check whether the count is drifting downward across replicates, which indicates cells settling in the tube.

Cells look grey rather than clearly blue or clear. Usually excess protein binding the dye. Wash cells into PBS and restain. Also check the coverslip is properly seated and the condenser is set for good brightfield contrast.

Chamber fills unevenly or cells pile at one edge. The coverslip is not seated or the chamber was overfilled. Clean, dry, reseat until Newton's rings appear, and reload with a single smooth capillary fill.

Automated counter disagrees with manual count. Check the size gate and the dilution entered into the instrument. Image a field and inspect the classification overlay directly -- most instruments will show you what they called a cell, and the answer is frequently debris.

Trypan blue is a laboratory reagent, not a therapeutic material, and should be handled as a chemical with recognised hazards. It is classified in several jurisdictions as a suspected carcinogen and reproductive toxicant, reflecting its structural relationship to benzidine-derived azo dyes. Handle in gloves and a lab coat, work in a biosafety cabinet or with adequate ventilation when preparing solutions from powder, and dispose of stained material as chemical waste in accordance with local rules rather than down the drain.

For the solution itself:

• Store at room temperature or 2-8 C, protected from light.
• Filter through 0.2 um before use if precipitate is visible.
• Sterile-filtered product is used to keep counting samples from contaminating a chamber that is reused, not because the assay itself requires sterility -- stained cells are discarded.
• Do not return unused stained sample to the stock bottle.

A 100 mL bottle at 20 uL consumed per count supports several thousand counts, so shelf life rather than volume is usually what determines replacement.