[I haven’t exactly mastered this and have merely assembled this post while learning things for myself. Please submit any tips or tricks or great videos and educational material. My immediate goals are making uncontaminated pure culture Pombe yeast starters with sufficient cell concentration for scale up. Even if you use a counting app, double check your math. I was making a mistake for a while that dramatically under counted my yeast.]
The hemacytometer is a specialty slide for a microscope originally designed to count blood cells hence the name, but it can even count yeast and sometimes bacteria. It works on the concept of a fixed volume and precisely etched grid for counting cells. Its use is a requisite skill set of the brewer and wine maker so it should also be requisite for any distiller aiming to make fine spirits. There is incredible educational resources available to get familiar with the tool which I aim to collect here. I’m also soliciting tips and tricks for those that understand it in the distillery context.
Many new American distillers claim to be too busy to tackle analysis, so my recommendation on some of these things is to simply own the supplies because they are cheap. When free time comes, or you hire an employee that already has experience, it will all be there, ready for service. I encourage the same for manual acid titration.
Cell counting can help properly pitch a ferment so that it finishes on time without surprises thus minimizing risk. Many people still pitch yeast by starter volume while they should be pitching by yeast concentration. Cell density numbers often range from 5-10 million/mL. and Arroyo gives us a little bit of data on the reduction in time it takes to complete a ferment as cell density increases which we will revisit in a bit. We can extrapolate from these numbers to get a sense of how much we need to dilute our samples for effective counting.
The hemacytometer can also reveal how viable cells are. Observation can note the percentage of cells that are alive or dead. Recent concepts also emphasize the amount of cells that are actively dividing (for most people this will be budding). Yeasts have a chronological life span which is there entire life but also a reproductive life span which is the percentage of their life they can viably reproduce. A lower cell density with a high percentage of division may be much more valuable than a high cell density with a lower percentage undergoing division. Noting the percentage of dead cells is a completely standard practice, but what we need to learn is how easy it is to also observe cell division for both budding and fission yeast. We also need to collect all the numbers we can to help interpret our results.
K.B. Smith from Jack Daniels in his spectacular chapter: Yeast practices in the production of American whiskies in Lallemand’s Alcohol Textbook 6th edition notes that eventually data can be collected to correlate brix attenuation of starters to yeast concentration so decisions can be made without a long format cell count. For example, in his data set, cell concentration reaches the maximum level at an attentuation of 60% of the original brix.
Smith then introduces the concept which suggests that pitching based on maximum budding rate may be superior to maximum cell count. This means pitching could be done much earlier (80-90% of original brix), but the trade off is having to more aggressively control and manage yeasting. I’m still not sure how easy it is to fully observe reproduction via hemocytometer. New distillers may not be able to fully take advantage of this concept but should be aware of it.
I bought a cheap $15 hemacytometer, but good ones can run $100. There are also mid quality ones in between. Common advice is to own a nice one and a cheap one as a backup. Quality differences are due to clarity, lack of imperfections like chips, and precision of the grinding and cell spacing. My cheap one is functional, but it has chips and I swear particular squares in the grid that should be the same size are not. It works, but I understand why nicer ones exist.
It is really helpful to understand the anatomy of the device to understand where precision lies and basically how it works.
The cover glass forms a bridge and what that rests on is precisely ground to form an exact height above the grid. Hemacytometer cover glasses are different than typical microscope slides. They are both larger and thicker (and flatter). If you buy cover glass, you will be looking for something specialized. In some videos you will see an operator putting small drops of water on the bridge supports. When perfectly mated, refraction rings become visible on the cover glass illustrating optical flatness. Underneath, the bridge is still dry. Each side of the hemocytometer has a little ramp. Eventually a sample will be placed on the ramp and capillary action will suck the sample up, filling the cavity where the grid lies. Only enough sample should be dispensed via micro pipet to fill the grid but ideally not enter the over flow zone which is called the “moat”. If sample enters the moat, it may raise the cover glass distorting the volume measured within the grid. Samples should be observed and measured within a fairly short time frame or evaporation may also distort the measure.
From what I can tell, you should never observe samples on a hemocytometer with oil immersion because the lens may flex the cover glass as you focus. Yeast cells will be counted at 400x.
The first thing you do when you get any new equipment is learn to clean it. A hemacytometer and their covers can be cleaned with soap and water, followed by a rinse, followed by alcohol, followed by a distilled water rinse, followed by wiping with lint free “Kim wipes“. The final containers that samples are prepared in should probably be disposable because of the stains.
From the Hausser website (finest maker of these devices):
To clean the counting chamber: After completing the count, remove the cover glass and clean the counting chamber with water or a mild cleaning solution (10% solution of bleach). Dry the counting chamber with a soft cloth or wipe, or rinse with acetone.
I’m attracted the acetone rinse and no doubt will start that.
The vials that samples are mixed and diluted in are often called either centrifuge tubes or “Eppendorf vials”. I have not figured out what is best, but it is looking like these should be 5.0 ml to as small as 2.0 ml. Samples often have an initial dilutions with water followed by a final 1:1 dilution with stain. Samples are mixed in these small volumes by repeatedly withdrawing into the pipet and then dispensing.
A key to successful cell counting is dilution. A typical rule of thumb is to use a dilution level that allow counting of 100-200 cells in 5 of the 25 central squares so only 20-40 yeasts per square. Sometimes that can be done at 1:1 (2x) while other times it can be done at 1:9 (10x) or even more for slurries. Often the last two parts of dilution come from adding the stain so for a 1:9 (10x) dilution, a 1.0 ml of sample will but diluted into 5.0 ml (adding 4.0) and then 0.5 or 1.0 ml of that will be paired with an equal volume of stain. This may require two vials, one for diluting the yeast sample, and then another for final dilution and mixing with the stain. There are even protocols for dense hard to dispense slurries that work by weight and are measured on an analytic balance.
Finding dilutions and the best vessels to do it in has been my biggest initial hurdle (I’ve slowly settled on 1:4 followed by 1:1 with stain). A starter will need one thing and then a ferment another as the cell count exponentially increases. I have had to try something, dilute more, and start again, but you can quickly start to understand the lay of the land.
Arroyo gives us some data that will inform dilution:
This is a loose example and we don’t even know the sugar concentration, but what we can say is if each ferment is given this initial seeding concentration and each starter is 5% of the total volume, then the starter has 20X the concentration. This will give us an idea of what to expect and strive for in our scale ups.
According to various literature, distillery ferments are typically seeded at a range of 5-10 million/mL, implying that starters need to have a concentration of 100-200 million/mL. No authors like to give exact recommendation but rather insist you know how to count as well as conduct progressive experiments to discover an optimal for a given yeast.
If 20 individual yeasts are counted on average in 5 squares for a total of 100, a dilution factor of 2 will give a density of 2.5 million/mL.
At the high end, 200 million/mL with a dilution factor of only 2 would require counting 400 yeast on each of 5 squares for a total of 2000! To bring this back to 20, the dilution factor would have to increase to 40 which means an initial dilution of 1:19 before being diluted again 1:1 by the stain.
Staining comes for the most part in two options: Trypan blue and Methylene blue. These also come at various concentrations which we will need to sort out (0.4% Trypan for 1:1 and likely 0.1% Methylene blue for 1:1 final dilution, but some people appear to use 1%). Trypan blue is more specific to seeing only dead cells because of the way it in excluded from living cells. Methylene Blue can work, but it is a more heavy handed general stain and is influenced by pH in a way that may impact some distillery work.
[Another stain idea is Erythrosine B which gets a glowing review from hemocytometer.org, but I have not seen it anywhere else. [I have finally acquired some.]
Here is another page that promotes Erythrosine B E127 over Trypan though it appears to be a new idea with not a lot of traction. Note the scaling 5μg/mL (0.000005) which I have a hard time believing at the moment. Yet another brief article.]
This paper mentions a 0.1% solution for Erythrosine B but also notes some was not soluble and excess was removed by filtration. They used it 1:1 as the final dilution for the stain.
From the Illustrated Guide to Home Biology:
For some staining protocols you can substitute erythrosine B also known as FD&C Red #3. Results are generally inferior to those obtained with eosin Y, but are still quite usable. The red food coloring dye sold in grocery stores is usually a mixture of erythrosine B and Allura Red AC (FD&C Red #40). We’ve used this mixture in diluted form with reasonable good results.
Maybe this will help someone who already has some red food coloring in the kitchen. Eosin Y does not appear to be relevant for yeast counting.
We find great writing from BKYeast and are introduced to an author I intend to read more of:
Methylene Blue – a blue dye that will stain the dead cells blue. Living cells also take in the dye, but the active enzymes within them will process (reduce) the dye and make it colorless. Problems with this stain include that some dead cells still have enough active enzymes left in them to reduce the dye, the dye may block some intracellular processes and result in failure to reduce it by living cells, below pH 4.6 the uptake of the dye increases so that many living cells also become blue, also it may stain stressed live cells with fractured membranes. In my search for more info I found some articles about using various microscopic techniques to tell between the false positive and false negative cells, all of which are beyond the homebrewing level. Despite all its shortcomings, methylene blue remains the standard viability stain most likely due to its availability.
Availability: easily available and inexpensive.
Trypan Blue – also a blue dye that stains dead cells blue. Unlike methylene blue, this stain’s action is based on cell membrane integrity. It is a negatively charged compound that can’t cross the membrane unless it has been damaged. Dead cells have ruptured membranes, allowing the dye to enter and stain them while living cells are very selective in respect to what goes in or out and trypan blue is not allowed to enter. This stain is not dependent on chemical reactions, gradients, reduction states etc., which makes it the gold standard for non-fluorescent vital dyes. It should be noted that it is toxic and cells exposed to it for too long will die and stain blue. Other two stains in this study will be compared to it to measure their effectiveness.
Availability: not easily available to homebrewers and expensive.
It appears that we should get Methylene blue because it is easy to find and cheap, while we should also strive for Trypan which may work better, but is harder to find and pricier. It is cheap enough to have both.
BKYeast gives us a note on clumping which I’ve encountered a lot of:
I found it especially curious how trypan blue and safranin can differentially stain cells in clumps while methylene blue has a bit more trouble differentiating between live and dead ones.
Here is another BKYeast note that will no doubt help the sour mash crowd:
Another cool thing is that unlike methylene blue, trypan blue and safranin don’t stain living bacteria, which means it can be used to assess the viability of your lactic acid bacteria cultures. This is something that’ll have to be followed up on.
White Labs notes a strategy of using 0.5% H2SO4 (sulfuric acid) dilution water as a strategy if cells are clumping [If the goal here is a chelating agent to unclump the cells, distilled vinegar may also work and should be easy to try). They also give us a note on counting actively budding cells:
Yeast cell buds emerging from mother cells are counted as a separate cell if the bud is at least one-half the size of the mother cell.
Here are a lot more strategies for clumping yeast samples.
A killer protocol on counting can be found here. Credit to braukaiser.com :
To keep track of counts many people use a mechanical clicker like a night club bouncer, but I really enjoy the ($1.99 HemocyTap app for iPhone from the hemocytometer.org team which makes counting and calculation easy.
This is a great video that shows 0.1 mL sample being diluted with 0.1 mL Trypan blue and put into a small centrifuge tube. She uses the larger squares and does not only go into the center 5×5 square and count count those. These may not be yeast cells.
***This video covers the anatomy of the device that others ignore and notes that the cover glass are thicker and optically flat which implies they are more precise than other cover glass. Optical flatness can even be observed because of rings it creates. The bridge supports the glass and creates a fixed volume underneath because of how it is precisely ground. This bridge system is probably why you cannot use oil immersion on a hemacytometer because you distort the glass over the bridge. Note, the cover glass goes on dry after cleaning with alcohol. You do not want the fluid to flow into the moat.
This video dilutes a very dense slurry 1:100 with water. Recommends 1:1 dilution with stain for a total dilution of 1:200. Their intense slurry was 1.4 billion/mL. They use capillary action to suck the sample under the cover glass.
This video shows yeast dilution by weight. This is aimed at commercial work with heavy slurries that may be hard to dispense by volume. 1:60 dilution by weight with an analytic balance. He uses a methylene violet stain.
This video again shows 100 μl (0.1 ml) mixed with 100 μl 0.4% Trypan blue. Here the cover slip is moistened and slid back and forth until refraction rings appear proving optical flatness.
This video shows that the moistening water is only placed on the bridge to get the cover glass to adhere. No water should be placed on the grid. The micro pipet is also used to mix the cell suspension after it has sat for a while.
Red blood cells are countered in the center square just like yeast.
This very clever video breaks down the math and makes it very intuitive (I’ve personally re-referenced this many times to get a handle on the math).
This video from White Labs uses 15 ml conical tubes as well as “citration Methylene blue stain or Alkaline Methylene Violet, 0.1%”. Yeast cells are counted at 400x. Notes that budding cells may be stained as dead, but they shouldn’t be counted.
What I haven’t resolved is how clearly we can differentiate budding and fission yeasts at 400x. Ashby’s century old photographs appear to be only be magnified 350x:
[I have started successfully observing and differentiating both fission and budding yeasts using 400x. You become a lot more confident when you observe cell division. For budding yeasts its the bud which is the most easily apparent while in fission yeasts its seeing them elongated with a sort of belt or band. I have not exactly differentiated top and bottom type, but mine look more elongated like the top type. I’m still not sure if I can count a reproduction rate with any confidence.]
Hemocytometer.org has some amazing educational resources.
Something to note is that there are specialized cell counting slides for bacteria which are smaller than yeast. This Hausser model (20 micron depth) is designed to work with oil immersion and requires special cover glass. Note that these specialized designs come with depth options. Who knows if this will be helpful for counting bacteria starters for rum some day.
This is a wonderful article if you are wondering about the different grid sizes. Yeast use a particular part of the grid because of their size.
Erythrosine B is promoted as a stain over even Trypan blue which I have not seen anywhere else.