Safer Yeast Cell Viability Assays: Erythrosin B

A major focus of 2019 at Escarpment Labs was spent finding ways to improve our internal processes, especially as it related to things like health & safety, efficiency, and reducing waste. We reflected on the tools and supplies we use every day, so that we could gain a better understanding of our environmental impact and how we can make our workplace safer.  

One of the projects that we completed in 2019 was setting up improved systems for waste disposal. Lab waste is pretty specialized and requires the separation of different waste streams to ensure that toxic materials do not go to landfills. One challenge we encountered was the need for a separate waste stream for Trypan Blue, the cell viability stain we used to use at Escarpment Labs. 

That's right, we don't use it anymore! Enter Erythrosin B.

Erythrosin B (Red No. 3) stains dead yeast cells a vibrant pink colour and leaves live cells unstained. Foggy London Ale is pictured here (40x)  

The possibility recently arose of using Erythosin B (a red food dye; Red No. 3) as a vital stain for cell counts instead of Trypan Blue. Mike Tonsmeire (Mad Fermentationist, Sapwood Cellars) showed in a fortuitous tweet that this dye can stain yeast cells clearly and that tweet inspired us to do some in-house experimentation to see if we could get rid of Trypan Blue.  

Trypan Blue is commonly used as a viability stain for cell counts. It is used by yeast producers and brewery labs to ensure that yeast viability assays are accurate. Accurate viability numbers are critical to yeast management. Without being able to accurately assess whether a yeast culture is healthy or not, the brewer is flying blind and may pitch poor-quality yeast, potentially leading to sluggish fermentations and off-flavours. 

While methylene blue is also common in the brewing industry, this stain has been shown repeatedly to have poor reliability and appears to not be reliable for every yeast type. Methylene blue is also becoming more challenging for brewers to source due to its toxicity (although if you're not ingesting or injecting it, you're probably OK).  

Evidence in other cell types (mammalian cells and bacteria) suggests that Erythosin B can stain dead cells with as much accuracy as Trypan Blue, with the advantages of 1) being nontoxic (while Trypan Blue is a known teratogen), and 2) not staining
proteins, which is an awesome feature given that most yeast samples contain some degree of trub.

Twitter isn't all bad news and arguments, we happened upon this inspiring tweet from Michael Tonsmeire (The Mad Fermentationist/Sapwood Cellars).

The potential advantages of Erythrosin B can result in 1) decreased cost from waste stream management, 2) Increased workplace safety, and 3) potentially faster cell counts if protein aggregates and debris are not stained.

For us, the prospect of introducing a cell stain that is so safe it can go down the drain is exciting!

Additional advantages of Erythrosin B include:

  1. Reduced time of effectiveness since cell death (Trypan blue = 50 min, Erythrosin B = 1 min). 
  2. A less concentrated solution needed (reduced cost). 
  3. Non-toxic to cells for periods of over 3h, introducing the potential for preparing
  4. Samples ahead of time with no impact to viability.
  5. Can potentially serve as a vital stain for bacteria with an incubation time of 5 min.

The challenge we faced was that there wasn't much info on Erythrosin B for yeast cell staining, with exception to Tonsmeire's tweet. 

So Iz ran an experiment, where she compared Erythrosin B to Trypan Blue using some example yeast slurries. She picked old yeast slurry from two common yeast strains (Foggy London Ale and Czech Lager), since we store the occasional homebrew pouch for shelf-life studies as well as staff brewing.  

She did 20 cell counts of each strain and stain, to eliminate any potential issues with a smaller sample size and to ensure that we can apply statistical techniques to the data we get. The counts were performed as standard yeast cell counts, with a 1/200 dilution of the slurry and counting the standard "5 squares" to approximate the sample cell density. 

Then we plugged the data into some statistical software to figure out whether we can switch to Erythrosin B. 

Strain Dye Count Mean SD
Czech Lager Erythrosin B 20 0.225 0.0462
Czech Lager Trypan Blue 20 0.217 0.0351
Foggy London Ale Erythrosin B 20 0.921 0.0201
Foggy London Ale Trypan Blue 20 0.918 0.0201
Viability measurements for the comparison of Erythrosin B and Trypan Blue as yeast viability stains. Each slurry was counted 20 times for each strain. The two slurry samples were 6 months old, stored at refrigerator temperature.


A visual representation of the above data showing that there is no significant difference between the two dyes.

The result was favourable - there was no significant difference between the two strains, in terms of the percentage viability (~22% for Czech Lager and ~92% for Foggy London Ale, both with fairly low standard deviation in % viability). 

It was surprising to us that the Foggy London viability was so good - that's cool! We set our homebrew pouch shelf life at 4 months and pro pitch shelf life at 1 month so that we prioritize production and distribution of fresh yeast, but in some cases, our yeasts can survive a really long time. However, the Czech Lager data shows that some strains are more sensitive to storage. 

We also wanted to understand whether the distribution in cell count (not viability) was consistent, to help demonstrate that the work done diluting and counting the cells was performed consistently.


Distribution of viable cell counts for both strains. The cell count is on the x axis and the relative frequency is on the y axis. 

We see here that the distribution of the cell counts are quite similar between dyes, suggesting the dilutions were performed correctly. However, there is a wider spread in the data for Foggy London Ale. This is likely because it is a clumpier strain (English yeasts tend to aggregate) and that can result in greater variability in the cell count obtained. 

This raises an important point in all this discussion: A yeast cell count (even with 20 counts) should never be considered as absolute value! It is always an estimate since we are always diluting a yeast sample and counting a sample of that dilution. There is a lot of inherent error and noise in cell counting, an event with perfect technique. As a result, perhaps statistical techniques can help improve how yeast folks estimate and communicate cell counts in the future.  

This was a fun side project that resulted in meaningful change for us. We found a lower-cost option for cell counting, eliminated a waste stream, and validated Erythrosin B for yeast cell viability assays. 

We are aiming to offer ready-to-use Erythrosin B solutions soon, feel free to shoot us an email if you'd like to switch to a less toxic but still accurate cell counting option. In the interest of transparency, we also offer our protocol for preparation of Erythrosin B solution below. 

Statistical analysis (for our fellow nerds)

 Is there a significant difference between the viability % outcome of the two dyes? Shapiro-Wilk test of normality (% viability data): Foggy P = 0.39; Czech P = 0.33. P > 0.05 suggests data are normally distributed and a t test is appropriate. Unpaired two-sample t test results: Foggy P = 0.6117, Czech P = 0.5367. There is no significant difference in viability measurement between dyes for the two yeast samples.

F test between viable cell count of strains: P = 1.921e-05. Therefore there is a significant difference in the variances between the two strains. See plot above, Foggy London has a wider distribution, which may be due to a higher incidence of clumping for this strain. 

F test between viable cell count of dyes: P = 0.9745. Therefore there is no significant difference in the variances between the two dyes. 

Erythrosin B preparation procedure


  • Erythrosin B powder (FisherSci)

  • 1M (10x concentrated) Tris-HCl buffer, pH 7.5  


  1. To make 0.1M Tris-HCl buffer: In a 50 mL tube, dilute 1M Tris-HCl buffer 1:10 by adding 45 mL sterile distilled water and 5 mL of 1M Tris-HCl buffer.

  2. Weigh 0.8g EB powder directly into the tube.

  3. Top up to 40g with 0.1M Tris-HCl buffer

  4. Vortex until mixed.

  5. Store with parafilm around the cap.

  6. Do not attempt to make a higher concentration of stock solution; the EB will not dissolve.


  1. In a 15 mL centrifuge tube, add 9.5 mL sterile distilled water.

  2. Add 0.5 mL of 1M Tris-HCl buffer and vortex.

  3. Remove 0.5 mL of this solution to yield 9.5 mL in the tube. 

  4. Vortex EB stock solution until homogenous.

  5. Add 0.5 mL EB stock solution into 9.5 mL Tris-HCl buffer solution to yield 10 mL working solution.


  • Make a 1:1 mixture of cell dilution and EB working solution, for a final concentration of 0.05% EB.

  • We suggest investing in a P10 micropipette, so you only need to use 5-10µL of stain for each cell count performed.  


  • Erythrosin B is a non-toxic food dye, but ingesting large amounts leads to acute toxicity. 

  • The powder (like many powders) is flammable - see SDS

  • As a vital stain with very intense pigment, beware of unwanted staining on instruments and clothing.

References/Further Reading

Leave a comment

All comments are moderated before being published