Chapter 8. Fun and Fights with Fungi, Part 2

Derek Jacoby and Vince Geisler

When last we left the stories of our fungal adventures, we had just sent the first two successful PCR-amplified mushroom samples off for sequencing. I was so sure that this was to be a triumphant presentation of dozens of barcoded samples. But we’re still in the midst of it. Today we’ll follow through on the bioinformatics of a successful barcoding sample, but most of our mushrooms are still unknown. We’ve just received the next level in the chemical arsenal to break usable DNA out of the mushroom, but we can report on our attempts so far. Outside of this last effort (which uses decidedly DIY-unfriendly chemicals such as chloroform) we’ve had interesting failures with sonicated samples (evenly sized DNA fragments [!] that are too short for our barcoding), with alkaline cell lysis buffers and detergent-based cell lysis buffers.

But let’s start with a description of the main problem. We have a mushroom sample we found at one of the mushroom shows, we may or may not know what mushroom it comes from, or more interestingly, we may think we know but are wrong. This mushroom has a genetic region that is not under evolutionary pressure, so it randomly accumulates mutations over evolutionary timescales. It turns out that due to this rate of genetic drift, different species of mushroom tend to have different DNA mutations in this region, causing it to be known as a barcoding region. There are a number of steps we have to follow in order to get our mushroom sample’s genetic sequence so that we can look it up in a database of known species. First, we have to first get the DNA out of the cells, then amplify our barcoding region using the polymerase chain reaction (PCR), then finally send it away for sequencing.

This is what we’ve been working on. Mostly the first step, so far. Last time, we talked about our troubles in disrupting the cells of the mushroom to release DNA for us to work with. One of our interesting attempts involved trying to mechanically break apart the cells using a sonification bath. This is just a Westinghouse parts washer bath, the kind where you’d normally throw your dirty, gummed-up car part in some solvent. You turn on the switch and it breaks the dirt and gunk off using transducers that produce a buzzing mostly above our range of hearing. So we thought maybe it would break our cells apart!

It was really successful. Too successful. The DNA came out sheared into about 500 base pair (bp) lengths. In the image at the end of this article, you can see the raw, purified DNA after it had been extracted from the cells but before PCR. It’s the band in the second row next to the ladder. This is great, except that the piece we want amplified is about 800 base pairs long. If all our chunks are only 500 bp, then no template DNA will exist to allow the PCR reaction to complete.

This is an interesting failure, though. We have no easy way to adjust the frequency or power in this parts washer, but with some work put into understanding the circuit and modifying it, we could adjust these variables. It did break the DNA out well, even if not into chunks we could use.

Which brings us back to the main problem: an inability to get DNA out of the cells. We’ve discovered we can’t break it apart too much or we get small fragments like in the sonicated samples.

Let’s look at the other samples. Our sole success, detailed below, was a jelly fungus. The others either produced no DNA or produced multiple bands after PCR amplification indicated an unusable sample. The prevailing theory is that polysaccharides in the fungal cell wall are binding with DNA, so we’re not getting any DNA from our purification protocol. We are using two different strengths of detergent for cell lysis, with and without disruptive sonification, so no release of DNA is unlikely. We have run successful barcoding reactions on insects with the same reagents to ensure that other aspects of our setup and technique are correct. If we are breaking the DNA out, but not recovering it, it seems likely that something in the mix is holding onto it. Enter the CTAB extraction. Essentially we break the cell with a strong detergent and then take the cell debris and cell wall component out of solution by dissolving them in chloroform. Interestingly, this leaves the DNA in the aqueous fraction so that we can work with it.

So in a way it’s a failure, because we needed to step back to chloroform and CTAB and the harsh chemicals that we hadn’t wanted to use. However, we are getting some DNA with the gentler processes, so let’s look at one of those sequences in more detail.

Sample 22 in last issue’s article was identified by the experts at the mushroom show as Pseudohydnum gelatinosum, a jelly fungus. It was the one with a single clean band in the gel that accompanied that article. When we got the sequence results back, we had two files, one from the forward direction and one from the reverse direction. This is important in longer sequences because of the limits of Sanger sequencing. Since we have an 800 base pair barcoding region, and the sequencing provider we use guarantees only 650 bases of good sequence read, we decided that reading from both directions was an important thing to do. To download the files in this section, go to http://reddwarf.ca/ and click Biocoder at the top. Our results (2_M13F.txt and 2_M13R.txt) arrived in what is called FASTA format. This is simply a text file format of sequence data. The first file is the forward sequence. We can now copy just the sequence information from that file (2_M13F.txt) and then paste it into the identification program at http://bit.ly/1o0edZy.

In this case, it returned our expected Pseudohydnum identification with about 99% sequence similarity, which was confirmed with the reverse direction read. The exact nucleotide match to our sequence is not found, and in fact, the returned results indicate a match with an uncultured Pseudohydnum species rather than the Pseudohydnum gelatinosum identified by our mushroom expert. So it’s time to enter it into the database!

As an aside, if you have never looked at sequencing results, please feel free to have a look at these files from http://reddwarf.ca/: positive.control.ab1, 2_M13F.ab1, and 2_M13R.ab1. Aside from the text file, the .ab1 files are a view of the actual chromatograph for the sequencing run. In Sanger sequencing, the sample is amplified by PCR; but instead of just using normal nucleotides, a small fraction are fluorescently labeled. When a labeled nucleotide is incorporated into the PCR product, amplification is not able to continue beyond that point. Since incorporation of the labeled bases is random, you end up with an assortment of partial products each with a labeled base at the end. These products are pulled through a capillary electrophoresis machine to sort them by size, pulled up by electric potential past a reader that reads the fluorescently labeled bases. The .ab1 file is a recording of those readings for each base, usually a clear single color, but sometimes there will be doubt at some of the bases. There are a number of programs that can read these files, but we used a trial of CodonCode Aligner.

At this point, we used up most of our samples in trying different extraction protocols. Lessons for this year’s mushroom show season include collect more material and get our protocols down early (and maybe use grocery store mushrooms to tune the procedures rather than our precious samples!).

Although the mushroom project itself hasn’t been as successful as we’d hoped, it’s been a lot of fun, and the side projects on the way have resulted in better PCR skills, a much-improved gel documentation station, and the completion of our fume hood (to work with chloroform). It was definitely worth doing, even if only for the side projects! Thanks for following along.