Tuesday, 13 August 2013

Another Wild Yeast Identification Test

Micrograph of the Wild Yeast (left) versus
conventional ale yeast (right)
Over a series of posts, listed below, I've covered my work designing a method to identify what species I'm purifying in my attempts to isolate wild yeasts.  This technique has concentrated on the sequencing ribosomal DNA; stretches of DNA which are frequently sequenced for the purpose of species identification.  My first attempt at this did not work well, but I had a higher degree of success upon re-targeting my sequencing attempts to a different region of the ribosome.

In parallel with these experiments I have been purifying a range of wild strains (edit: completed here) which will soon be tested for their fermentation characteristics.  Once good brewing strains are identified, I will further characterize them - including identifying the species via ribosomal sequencing.

Before I go down that road I want to test the sequencing method on a bona fide wild strain.  For this test I am using a strain kindly provided by fellow wild brewer Doc_Drive.  This yeast was featured in an earlier post of mine as an example of suing old yeast-identification methods.  In this post I tentatively identified this yeast as Kloeckera apiculata based on its morphology and described fermentation characteristics.

I have now positively identified this yeast as Kloeckera, based on ribosomal sequencing; the full story can be found below the fold...yes, I was right!

Something isn't right...
Left: Ladder
Middle: ITS PCR
Right: 26s PCR
As per usual, this trial starts with a DNA isolation - in this case I tried the simplest method available; a small amount of yeast was suspended in 100ul of water, and boiled for 15 minutes.  The solution was then centrifuged to pellet any cell debris.  I then PCR amplified both sections of the ribosome I have primers for (e.g. the 'old' (ITS) and 'new' (26s) primers), using a 49C annealing temperature and 45 amplification cycles.  When run on a gel, I got the picture on the left.

The resulting bands were then purified, and sent for sequencing.  Due to the small amounts of DNA present, I didn't get a good sequence off of the NS1 primers.  The ITS primers gave the following sequences on NCBI Blast:

PCR Sequence Identification
Escherichia coli P12b

Well shit, not what I was hoping for.  Looks like somewhere in the mailing/reculturing chain some nasty bacteria got in - guess that's what I get for not using an antibiotic containing plate.

So I streaked out my stored 'yeast' on an antibiotic plate, repeated the sequencing and got the following:
PCR Sequence Identification
ITS aaaccaactg ggattacctt agtaacggcg agtgaagcgg taaaagctca aatttgaaat ctggtacttt cagtgcccga gttgtaattt gtagaatttg tctttgatta ggtccttgtc tatgttcctt ggaacaggac gtcatagagg gtgagaatcc cgtttggcga ggataccttt tctctgtaag actttttcga agagtcgagt tgtttgggaa tgcagctcaa agtgggtggt aaattccatc taaagctaaa tattggcgag agaccgatag cgaacaagta cagtgatgga aagatgaaaa gaactttgaa aagagagtga aaaagtacgt gaaattgttg aaagggaagg gcatttgatc agacatggtg ttttttgcat gcactcgcct ctcgtgggct tgggcctctc aaaaatttca ctgggccaac atcaattctg gcagcaggat aaatcatt Hanseniaspora uvarum

It worked!  Not only did it work, but my first tentative identification of the yeast - based on morpholog & a bit of a description of where the yeast was found - ended up being right!  It is hanseniasporia - AKA the Kloeckera I had tentatively identified it as before.


  1. Nicely done. I have been toying with ITS PCR and RFLP for differentiating Brett and Sacc. I like your approach and may start to send off for sequencing.

    1. The method does seem to be working. And with the low cost of sequencing and PCR, is becoming more affordable by the day!

  2. Congrats on your Kloeckera results. Did you use the following primers for this identification: NS1 (GCATATCAATAAGCGGAGGAAAAG), NS4 (GGTCCGTGTTTCAAGACGG)?

    Never did a PCR with 45 cycles before. I would expect a lot of product after so many cycles. How much DNA could you isolate from the gel?

    1. I used the NS primers outlined in my older post:


      45 cycles is about the maximum number most PCR enzymes can handle, and with smaller products like the one sequenced here, can greatly improve yield. Especially when starting with the trace amounts of DNA that I have using the boil-method to recover yeast DNA.

      I do not remember how much DNA I recovered off hand, but it was enough that I had to dilute it 1:5 or so to reduce the concentration to what the sequencing lab prefers (10-20ng/ml).

  3. Are you using a master mix or making your own? What Mg+ concentration?

    1. I am using home-made PFU, but with a commercial buffer. I am not adding additional Mg+. It contains:
      200 mM Tris-HCl (pH 8.8), 100 mM (NH4)2SO4, 100 M KCl, 1% (v/v) Triton X-100, 1 mg/mL BSA.

  4. This is great information, thanks’ for share!

  5. I'm not sure if you will see this comment since this post is so old, but I felt it best to go here! If I don't see a response, I will repost on your latest entry.

    I am trying to identify a yeast cultured from a commercial source. After streaking out the yeast and letting it grow for a couple days at 30C, I then used the PCR method you outlined in your blog (ITS1/4 primers).

    I gel extracted the only band (side note: do you include DMSO in your PCR reaction? I find that helps reduce laddering, among other issues, quite a bit!), which was a bit larger than 800 bp. I sent this for DNA sequencing. When I BLAST the resulting sequence (chromatogram looks good), I get S. cerevisae hits (only about 89% identical though over the 800 bp).

    My real question is, how can I figure out what type of Sacc this is? Or most closely related to? I am going to do some test fermentations with this isolate, but I was perhaps naively hoping to at least figure out what type of yeast this was at the genetic level first...

    1. It sounds like your PCR is working fine - ITS should be between 800 and 900 bp in size. The level of homology is lower than I would expect - I typically see 95% or higher - are you trimming the sequence to the best quality sequence?

      In terms of DMSO, I've not found it necessary, but different enzymes may benefit from it under standard ITS conditions.

      You cannot get strain-specific information from ITS sequences; only the species/genus gets identified. There are other PCR fingerprinting methods for strain-typing saccharomyces, but I've done minimal experimentation with those and do not know how well they work.

    2. I am going to try Interdelta PCR (http://onlinelibrary.wiley.com/doi/10.1016/S0378-1097(03)00205-2/full) and see if that works better. At a minimum I should be at least able to compare my isolated strain with strains in my library.

    3. Inter-delta works well for sacch. I'm trying other methods that may work for sacch and other yeasts (e.g. brett). Nothing to report yet on the alternate methods.