Yeasts Aint Yeasts

Lately I’ve been reading about how some recent discoveries have shown that some yeasts we thought were hybrids aren’t and some we thought we’re aren’t.

Now I’m just a brewing guy who studied biochemistry a while ago so I am going to probably stuff some of this up compared to people more well-versed in this topic than I am, but here goes…

What I mean by hybrids is that brewing yeasts in the Saccharomyces species fall into two main categories – “Lager” or pastorianus yeast, or “Ale”, cerevisiae yeast. Pastorianus is actually a hybrid of Cerevisiae with another yeast, Eubayanus, that gives it is low temperature and other characteristics that set it apart from the normal Ale type yeasts.

With recent advances in biotechnology including the practicality of PCR and gene sequencing it’s been shown that at least two yeasts and probably more used for a long time as lager / ale brewing yeasts aren’t actually what they seem. WLP800, which is a single strain isolate purportedly from Pilsner Urquell has been shown to be an ale yeast, and WLP051 California V Ale yeast has been shown to be a pastorianus yeast.

There’s still some debate out there about WLP800 which led me to search out for some extra data, as well as working out how else we could get some additional evidence. I jumped onto the NCBI SRA site and searched for WLP800, hoping to at least find some extra data that I could point someone at, and I must have got lucky as there was a huge collection of SRA data uploaded around 31st August to the NCBI site, from a whole number of commercial yeasts.

Turns out it’s for an upcoming study by Hittinger Lab, one of the world’s foremost yeast genomics labs into hybridisation of brewing yeasts, the paper is yet to be released but they have done a huge favour to everyone by releasing this data. From the descriptions of the short reads I could see a hybrid species identification had already been done – showing WLP800 was plain Cerevisiae, WLP029 was Cerevisiae x Eubayanus hybrid, as well as a couple of other interesting categorisations that go against common knowledge. E.g. WY1187 Ringwood being a lager, WLP838 S.German lager being an ale!

Now I don’t really have that much background in these matters so I reached out to a couple of people who I thought might help. One was Suregork, aka Kristoffer Krogerus, yeast researcher at VTT Finland, who’s got a heap of knowledge in the area for sure, even knowing how to make hybrids and study them, and the other was the owner of the data, Quinn Langdon.

That led to two things… Kristoffer agreed with me about how interesting it was, told me a little about how the identification would work and then wrote up this article where he found the WLP351 sequencing matched the sequencing they did on the Muri Kveik.

The owner of the data then got back to me and explained a couple of things… one was that the paper I thought the data was from wasn’t that, and this was for an upcoming paper, and the other thing was that the identifications came from a tool they developed called sppIDer run over the short sequences.

So this tool they released last year known as sppIDer is for analysing genomic content of short read gene sequence data. There’s a paper over here explaining the program: https://www.ncbi.nlm.nih.gov/pubmed/30184140

It takes short read data, and analyses its fit to a reference set of genomic data. You can then set up different genomic data to refence with, and the typical example with this would be S cerevisiae, S uvarum and S eubayanus to determine fit to the different hybrid genomes you’d find in a saccharomyces’ gene sequence. It can then also determine some level of ploidy by interpreting the data to see how much material maps to the different genomes, so this would be useful for identifying Frohberg (ie German) lager strains vs Saaz (ie Bohemian / Danish) lager strains.

I managed to the sppIDer program running. Being Docker makes it easy for me as I work with Docker containers constantly as part of my job at work, the program is based on R, Python and a number of bioinformatics programs such as bwa to process the short read data.

I don’t want to say too much about the results so as not to tarnish the full result release contained in that upcoming paper, but here’s an example of WLP029 short read data being fed into it, the analysis shows there’s plenty of Eubayanus DNA in there.


Once the paper’s released I will do some more analysis, and also Suregork is working on an updated yeast family tree. It’s going to be very interesting seeing where WLP838, 800 and some of these other strains end up in the tree!

Kölsch yeast experiment

So one of my favourite beer styles is Kölsch. While I’ve never been to Cologne where the style originates, I’ve had a couple of them and brewed them to the best of my ability, doing reasonably well in competitions with them. The trick is to mash for dryness, either a step mash or a 62-63C single infusion, with a long mash of 75-90 mins, and get the yeast right.

This is where this experiment comes in. I’m trying to find the best yeast choice for both my homebrewed efforts and for when I eventually make the leap to going commercial, and there’s a multitude of options out there now.

The yeasts of choice for this experiment are the new Lalbrew Köln dry yeast, Gigayeast’s GY021 and White Labs WLP029. Before I’d started this I have also used Wyeast 2565, rumoured to be the Weihenstephan 177 strain, and the White Labs European Ale strain, as well as WLP036 Alt. 2565 and European Ale are good flavour wise but take forever to clear to brightness, and WLP036 isn’t a kölsch yeast, it makes alright beer but I left it out.

Yeasts… The WLP029 was probably 15mL of thick slurry. Gigayeast was split up into 4x vials so approximately 50+ billion cells pitched.

Recipe was simple – 94% Pils , 4.5% Wheat, 1.5% Acidulated. Mash and sparge water adjusted with Phosphoric acid to hit a theoretical 5.2pH in Bru’n water along with about 4g of CaCl2 and 1g CaSO4 in the 15L batch. OG 1044, and around 27 IBU’s with Perle for bittering and Perle + Huell Melon for flameout. Looking back over my notes the Pils was actually a 60/40 split of 60% Barrett Burston Pale and 40% Weyermann Bohemian Pilsner.

Mash temperature

3x 5L demijohns were used for the fermentation, splitting the cooled wort into these, aerating each for around 1-2 mins with an airstone + bubbler then pitching. WLP029 was fresh slurry from a starter, GY021 was split from a fresh pack, and Koln was approx 5g rehydrated from a pack.

The demijohns.

Fermentation temperature was around 17C for all 3, and fermentation continued for 2 weeks – around 4-6 days active ferment and then conditioning / maturation at the same temperature – no cold crashing. Each demijohn got a small amount of Gelatin, and 2mL of White Labs Clarity Ferm was added during maturation to each, for no real reason other than testing that out.

Lag Time

WLP029 fired up first. GY021 was second and Koln took longer than both to get started and show signs of fermentation.

The WLP029 on the left is fermenting, but nothing’s happening over in the Koln!

Krausen

WLP029 had a decent krausen, GY021’s krausen was highest and Koln did eventually get a decent krausen, probably in the middle.

Sulfur & Diacetyl during fermentation

With only a small amount of wort no real sulfur was noticed near my fermentation, but pulling a couple of samples with a pipette did show a little more in the GY021 compared to the others. None of it stuck around though by the time it came to bottling.

I didn’t have enough of a sample to detect much diacetyl or run a forced diacetyl test during ferment but nothing seemed to be overly buttery after active ferment.

Attenuation & pH

The Koln took a fair amount longer in active fermentation than the other two. WLP029 was mostly over in 3-4 days, GY021 a little longer and then Koln probably bubbled away for 6 days. All of them seemed to still generate CO2 for a couple of days afterwards very slowly.

The final beers ended up a little different in attenuation, Koln 1.007, WLP029 1.008 and GY021 1.006. pH was around 4.1 for the Gigayeast, 4.2 for both the WLP029 and Koln.

Final Results & Conclusion

All 3 produced great, crisp and bright beer that’s the hallmark of a kolsch.

I’d have no hesitation in using any of these yeasts again, especially the Gigayeast 021 which may have now become my new favourite kolsch yeast, but the dry version is also good to my palate.

EDIT: I’ve tasted all the beers further now they’re carbonated and the WLP029 is no slouch, it’s crisp and clean. GY021 has a little more roundness of flavour, but there’s maybe a little harshness in the finish, and the dry yeast has a little bit of extra fruitiness/esters on the palate maybe a little like an English or American ale yeast, it’s not an overpowering flavour but noticeable. With such a subtle style any flavour sticks out, so all of these still made good beer, but really it’s a matter of preference as to how much flavour you want or prefer.

Over the next little while I’ll try out the beer with some other people and see what they think, while I don’t have enough beer to go full Brulosophy style with 20 tasters, it’ll be interesting to see what happens. And i’ve also entered the Gigayeast and Dry versions into a local homebrewing competition, so we’ll see what happens there (Edit: see the final section below for how it went)

I’ll take a couple of pictures when I can to show you how bright it is…

A note on price though… here in Australia at least there’s little difference in price between a vial of White Labs or Gigayeast and the 2 packets of the dry Koln yeast required to obtain the correct pitch rate recommended by Lallemand. I can see that a few microbreweries may like to use the bricks of this yeast to make Kolsch, clean ales or pseudo-lagers.

Other Thoughts

After I started this experiment, I also found some new gene sequence evidence that WLP029 may actually be a lager yeast… stay tuned for that one. It’s the same data from the Hittinger lab that Suregork was able to identify the Muri Kveik as being identical to WLP351 – http://beer.suregork.com/?p=4094 . I guess that explains some of the clean-ness and its use for “pseudo lagers” or quickly brewed lager-tasting beer in a lot of breweries…

Also in the same set of sequencing it turns out that WLP800 most probably is and WLP838 may be actually ale/cerevisiae type yeasts – these could also be interesting targets for an upcoming experiment.

Followup

I did pretty well with the GY021 and Lalbrew Koln versions at the recent NSW State homebrewing competition. The Koln-fermented version scored 84 points and the GY021 scored 78 points ( a couple of points lost from sulfur as it was only 2 weeks in bottle before tasting), however at the mini-BOS round in the category the GY021 ended up taking 1st place in the Pale Ale category and the Koln, 2nd place in the category. I guess maybe the sulfur disappeared a little between the scoring round and the placing round.