Eureka, we are back to science. Today, I would like to start with a series of posts covering various spoilage yeasts. The yeast of today is widely used in food production such as bread and bakery products but has a connection to beer as well. The yeast I am talking about is called Torulaspora delbrueckii. I stumbled upon T. delbrueckii a while ago as this yeast is apparently used in the production in Bavarian Wheat beers. However, I could not find any scientific reference discussing the use of Torulaspora in beer. The only published cases of T. delbrueckii in beer cover T. delbrueckii as spoilage organism.
Where do I work?
In general, all non-Saccharomyces yeasts are considered as spoilage yeasts associated with negative traits such as introducing off-flavors, impact on clarity and different sugar preferences leading to different attenuation levels (degree of fermentation). This is now changing and lots of efforts and research is put into examining the effects of different “spoilage” yeasts in either single inoculation or in mixed fermentations along with Saccharomyces cerevisiae, the working horse of most of the beer brewers, wine makers, spirit producers and lets not forget the bakers. One other “spoilage yeast” which gets a lot of attention lately is Brettanomyces for example.
The first positive effects of Torulaspora in mixed fermentations has been initially studied in wine where the use of Torulaspora increases the complexity of the final wines [Tataridis P, van Breda V, 2013]. And yeast products with this yeast are already available.
What about beer?
The first published evidence that Torulaspora has positive effects in beer was published by Tataridis et al in 2013. The authors fermented 3.5 L of malt extract wort (OG 1.044) each with WB-06 and a strain of Torulaspora delbrueckii and compared the beers. They first noticed that T. delbrueckii was capable of metabolizing maltose the most abundant sugar in wort. However, the fermentation using T. delbrueckii took a while longer to reach terminal gravity compared to the WB-06 fermentation (157 h vs 204 h). The beer fermented with T. delbrueckii was more hazy and had a higher terminal gravity (1.012 vs 1.009). Despite the higher terminal gravity and a slower fermentation activity, the most interesting differences could be observed in the final beers. T. delbrueckii showed higher ester notes (mainly banana, rose and bubblegum) and a decreased phenolic character than WB-06. Demonstrating that T. delbrueckii might have a potential positive role in the production of wheat beers.
Now that we covered some basics about the possible advantages of the yeast, lets look at the taxonomy and biochemistry.
A quick taxonomy journey
Questions to be addressed in this chapter are:
- What is the closest relative yeast of T. delbrueckii?
- How closely related are Saccharomyces cerevisiae and Dekkera bruxellensis (aka Brettanomyces bruxellensis) to T. delbrueckii?
To address these questions, one can look at certain DNA sequences of the different yeasts and compare them in terms of how similar they are. I will try to make this very simple here. Think of a mother yeast cell from which all existing yeasts originate and evolved during time. Kind of the ur-mother-yeast-cell. Lets assign the letter A to the mother yeast cell and B to be a yeast daughter cell of A. Let me walk you through some possible examples of B and its impact on the DNA compared to A. Please notice that this is a simplified version and I am fully aware that biology is a bit more complicated than depicted in this example.
- B is a direct ancestor of A. B is a daughter cell of A and originates from a budding/fission event of A directly creating B. In this example, the DNA of both cells are the same (I intentionally leave mutations etc aside here)
- B is an ancestor of A but not a direct one and evolved during time thereby changing the DNA sequences in B compared to A. B is still in the lineage of A but not very close any more due to evolutionary events. There can be several billion, billion, billion daughter cells between A and B. In general, more similar DNA sequences are more likely to be closer related
- If B is very distant of A (in terms of DNA similarities), B is classified as separate species than A. In this case, B and A cannot interbreed any more because they are too distant of each others. S. cerevisiae and Dekkera for example would be daughter cells of A but very distant and form their own species
Lets take another example, horses. Zebra, horses and donkeys look very alike but are different species. (I intentionally leave mules aside here as this these animals are hybrids of horse and donkeys). It is very likely that all these species originate from some kind of ur-horse but individually adapted to new environments forming three different, new animals. To investigate which animal is closer related to which one, one could isolate DNA from the three animals and compare them.
To address what the relationship between T. delbrueckii and S. cerevisiae and Dekkera is, one can look at the large subunit of the ribosome (LSU rRNA). The ribosome is a complex of various subunits and is responsible for the protein synthesis in the cell. Because the ribosome is a very important machinery in a cell, the changes over time on the DNA level which encode parts of the ribosome are rather low. And can therefore be used to assess relationships among different species and strains. For T. delbrueckii, the relationship between some other yeasts is depicted in Fig 1 as a phylogeny tree. The tree begins with a common ancestor and the branches represent different fates.
I included additional members from the Torulaspora genus to have some close relatives of T. delbrueckii in the tree. And Saccharomyces and Dekkera to see how they end up in the phylogeny tree. One can observe that S. cerevisiae seems to be closer to Torulaspora than Dekkera.
Addressing the closest yeast relative of Torulaspora delbrueckii is a bit more complicated. First of all, it all depends on the data one uses to construct the phylogeny trees. If you for example do not include the true closest yeast relative in the dataset you will not pick it up anyway. Looking through some published phylogeny trees makes it hard to give a final answer. In one example published by Kurtzman et al (2011), the closest relative of T. delbrueckii is S. cerevisiae (like shown in Fig 1). On the other hand, another phylogeny tree published by Kurtzman et al (2011) ends up grouping Zygotorulaspora and Zygosaccharomyces closer to Torulaspora than the Saccharomyces group. In the latter case, Zygosaccharomyces mrakii and Z. rouxii end up being the closest relatives. It is therefore not possible to give a final answer here based on my investigations. However, what is obvious from the phylogeny tree shown in Fig 1, Dekkera is farther apart from Torulaspora than Saccharomyces.
Torulaspora delbrueckii has a very long list of synonyms which include a lot of different genera like Saccharomyces, Debaryomyces, Zygosaccharomyces and Torulaspora. In 1970, Kurtzman et al assigned Torulaspora and Zygosaccharomyces to Saccharomyces leaving Debaryomyces as a separate species. Five years later, van der Walt and Johannsen recreated the genus Torulaspora and incorporated all Debaryomyces species to it as well. Nine years later, Kurtzman et al accepted all four species again. This is actually not very uncommon in yeast taxonomy which is why yeast taxonomy can be very confusing and undergo lots of changes.
One reason why Saccharomyces, Debaryomyces, Zygosaccharomyces and Torulaspora make the lives of taxonomists so hard is their biochemical and phenotypical similarities and behaviour. Thus making it hard to differentiate the species. In addition, different yeasts were initially assigned to species based on morphology and biochemical properties. Nowadays, yeasts are assigned to species based on DNA. Which can lead to a lot of taxonomical changes and re-assignments of various yeasts. That’s how it is.
Beside T. delbrueckii, five other Torulaspora species exist being T. globosa, T. franciscae, T. globosa, T. maleeae, T. microellipsoides and T. pretoriensis. All other species with the exception of T. microellipsoides and T. delbrueckii are not associated with beverages. T. microellipsoides could be isolated from apple juice, tea-beer and lemonade and is a contaminant of soft drinks [Kutzman et al, 2011].
Where can you find me?
Most of the Torulaspora species and strains are isolated from soil, fermenting grapes (wine), berries, agave juice, tea-beer, apple juice, leaf of mangrove tree, moss, lemonade and tree barks [Kutzman et al, 2011]. With a bit of luck, you may find yourself some Torulaspora or you may go with the available Torulaspora delbrueckii yeast products.
Some say that Wyeast’s WY3068 Weihenstephan is a Torulaspora delbrueckii strain or contains Torulaspora delbrueckii. At least based on micrographs, its hard to tell whether WY3068 Weihenstephan is different from a typical Ale yeast (such as WY1056 American Ale) (Fig 2, 3). If anyone has rRNA seqs from WY3068, please let me know.
Some biochemical stats about me for yeast ranchers
Below a summary of the biochemical properties of T. delbrueckii. Data is summarized from Kutzman et al (2011). One way of differentiating between S. cerevisiae and T. delbrueckii can be performed using RFLP using HinfI on amplified ITS1-5.8S-ITS2 amplicons [van Breda, 2013]. Or obviously by sequencing the ITS1-5.8S-ITS2 amplicons.
|Systematic name:||Torulaspora delbrueckii|
|Synonyms:||There are a lots of accepted synonyms for this yeasts. Just some examples: Saccharomyces delbrueckii, S. rosei, S. fermentati, S. torulosus, S. chevalieri, S. vafer, S. saitoanus, S. florenzani
|Growth in malt extract:||Cell morphology:||Spherical, ellipsoidal, 2-6 µm x 6.6 µm
|Clustering:||Occurring as single cells or in pairs
|Growth in malt extract:||Colony morphology:||After 3 days: Butyrous, dull to glistening, and tannish-white in color
That’s all about Torulaspora delbrueckii so far. I hope this was in a way informative to you. At least keep in mind that spoilage yeasts do not inevitably have to be bad. If one can use their potential for our advantage, we can make something really unique. Have fun playing around with Torulaspora delbrueckii.
- Kurtzman CP, Fell JW, Boekhout T (2011) The Yeasts, a Taxonomic Study. Volume 1. Fifth edition. Elsevier (Link to sciencedirect)
- Tataridis P, Kanelis A, Logotetis S, Nerancis E (2013) Use of non-saccharomyces Torulaspora delbrueckii yeast strains in winemaking and brewing. Zbornik Matice srpske za prirodne nauke, Vol 124, 415-426
- van Breda V., Jolly N, van Wyk J (2013) Characterisation of commercial and natural Torulaspora delbrueckii wine yeast strains. International Journal of Food Microbiology, 163, 80-88
I definitely agree with your conclusion that “spoilage yeasts do not inevitably have to be bad.” This was way informative. Thanks!
Cheers, some other posts will follow covering Candida, Pichia, Debaryomyces and some Saccharomyces species as well. And I have some other yeasts on my wish-list. Thanks for reading
Huh, if Debaromyces has some of the same phenotypic properties as Torulaspora (if both genus are the same) it might be possible to differentiate using a cycloheximide containing media (D. hansenii is resistant to cycloheximide). I’ve used a 10%w/w and 5%w/w glucose broth to differentiate between D. hansenii and other yeasts before as well. D. hansenii is very tolerant to high salinity.
Fantastic blog! I have isolated some T. delbrueckii and Pichia sp. and have used them in Cider and beer. I think the non-sac species are truly underappreciated in brewing. I especially think this is the case because when you look at highly rated wild-fermented beers, wine, and ciders they are composed of multiple species that change over time, like in succession. This flavor complexity is not as well matched with single strain. Furthermore S. cerevisiae has been selected for many years where as theses non-sacc species are under utilized and/or studied. We now know that different strains of S. cerevisiae behave very differently and can contribute undesirable flavors and smells dependent on temperature regimen, ingredients used, pitching rate, etc. Very little of this has been examined for these non-sac yeast. But from my own experience wild strains vary in their characteristics much more widely than cultivated strains. Once we understand more about these strains then we will better know how to predict use them in producing beer, wine, and ciders. I have used Pichia fermentans in beer and it will soon be used in a commercialized brewery (Haw River). If you have patients and time to experiment with these strains they can be put to exception use and lead to a novel and distinguished product. I’ve been acquiring some novel yeast strains and can say that for sure Laffont alpha is T. delbrueckii if you want to do a side-by-side comparison. I’ll try to purchase and sequence WY3068 for comparison also.
Thanks Deb, indeed lots of non-Saccharomyces yeasts are under appreciated and maybe unknown territory to lots of brewers. Hence my efforts to briefly discuss them. And I completely agree that mixed fermentation is more powerful than single inoculations.
I could send you some WY3068 as well if you like. The sequencing would indeed be quite interesting. Some say that the Wyeast 3056 Bavarian Wheat blend contains Torula but I cannot see any different morphologies on agar plates nor on micrographs. However, this does not mean that there is no Torula in there. I simply cannot detect it with the methods applied.
What does P. fermentans taste/smell like in your experience? The only previous mention I know is from JasperYeast, who found his various isolates revolting, like Limburger cheese. I imagine if you and a commercial brewery are happily using it they must be rather different strains!
Thanks for this, I have been super curious about T. delbrueckii ever since I saw that suggestion about WY3068.
Sounds like a good series–I’m really looking forward to the rest. Dr. Roy Ventullo (who reads and posts on a lot of yeast blogs) and his students did PCR ID on my various strains (DCY01, 02 and 03) to the species level by comparing with known strains. They found 01 to be Saccharomyces, 02 to be Brettanomyces but 03 was something else they did not have. I’ll be curious to see if it matches up in description to anything else you are going to write about.
You are very welcome. UFO-strains such as your DCY03 are anyway more interesting than Bretts. Maybe its the next big deal 🙂
Roy Ventullo & students run PCR on my 20 strain from the BBA/EBY experiment so far. The preliminary results I got look quite promising and all the strains looked like Brett in the first run. We will have to see how the next experiments look like.
I guess the WY3068 question might be answered soon…
I routinely do plating and strain isolations along with PCR so I would be happy to add in an additional strain. For testing fermentation/taste I use Apple juice or cider. I buy it in bulk at Costco and the plastic 1 gal bottles make an easy, cheap and reproducible test system. I did a three-way comparison with S. cerevisiae, T. delbrueckii and P. fermentans. They all obtained an ABV around 6-6.5% after 8-14 days. I personally liked the flavor and aroma of the non-sac yeast in that trial. The non-sac were a tad sweeter and not as dry but they were both very aromatic. Furthermore We have brewed beer head-to-head between S. cerevisiae and P. fermentans. They are very similar but again the Picha had a sweeter aroma and flavor profile. We did two taste test with friends (1) telling them one was a wild yeast strain and (2) the other was completely blinded. We found that whenever you inform tasters that one is a wild strain they automatically say one taste “wild” or “yeasty” or even “bad” even if they wrongly guess which one is the wild-strain. In our fully blinded test those descriptors never really entered the discussion. But the results in both taste test were the same. About 50/50 on preference which seemed to depend on if the person liked a sweeter or a dryer profile. I find Torulaspora gives a very fruity aromatic profile and even an apricot like or fruity flavor so I think it would be a great strain for a wheat or belgian beer. One important difference I noted is that the non-sac strains typically ferment a little bit slower. Furthermore the strains I’ve isolated appear to ferment best at warmer temperature (70’s) ranges and can be quickly overgrown by S. cerevisiae contamination.
Thanks for your explanations. I wonder how such strains would work in met as I am not quite fond of cider. Although this may change soon because I came across some nice cider examples lately.
Very nice article, as usual.
Keep up the educational work 🙂
Cheers, the next one will follow soon. So many interesting wild yeasts out there 🙂
Pingback: Reporte Ciencia UANL » For Yeast’s Sake: The Benefits of Eating Cheese, Chocolate, and Wine
Pingback: The Weihenstephan Strain (WLP300/WY3068) – Saccharomyces or Torulaspora? Part 1 | ode to zymase
I love your blog and these posts in particular. I’ve spent a few months now gathering a small library of spoilage yeasts and am planning on fermenting wort with pure cultures of them. I have 6 to start, the first one will be Kluyveromyces lactis followed by this species mentioned in this article. Thanks for the background information. I really appreciate it.
Hi Matt, thx for your feedback. I for myself got me some pure non-sacc cultures but started with some other varieties in the meantime. Lets see what potential these yeasts have.
I just found your site and I thank you for sharing your knowledge with all of us. My son who likes to brew beer recently expressed interest in trying some non-sacc fermentations. Can’t wait to tell him about your site!
Thanks a lot. More non-sacc stuff to come soon.