Warnings, solutions, and mysteries regarding extraction/PCR from large amounts of yeast stationary-phase genomes

This post is about some perplexing and difficult to get rid of contaminant that accumulates in yeast, especially stationary phase samples, that inhibits PCRs in a hard to predict way. Now I'm no biochemist, so I don't have a solution for you, but I do have a dirty hack that should let you do large batches of extracts that are good enough quality to load 1-2ug of genomes into a PCR to get good amplicon libraries out 👍 You know, this is really the kind of content that's really going to launch my career as an influencer 🤩 , but hey seriously maybe it might help someone out there who's doing yeast barseq - this ones for you.

Background I came across this issue when joining a lab that was doing amplicon-seq of lots of complex genomic-integrated barcode libraries (>1e6 barcodes). PCRs worked or they didn't, folks didn't seem to be too concerned that sample-to-sample variation could be affecting their PCR efficiencies dramatically with unknown effects on pcr-mediated recombination 😱😱😱 eek indeed. We wanted to fit as many genomes into the PCR as possible, because at 12.5 femtograms even 100ng is only 8x mean coverage of a 1 million barcode library. But 100ng or more extracted gDNA would often fail. Adding magnesium helped with efficiency, but it was hard to predict how much to add for consistent amplification

Because genotype effects on growth physiology confound measurements that use half-assed extraction techniques like enzymatic digestion (see https://doi.org/10.1074/jbc.M606973200 supplemental), we only use physical disruption to lyse ( see https://doi.org/10.1002/yea.2911 ). But if you do an ammonium precip protocol, or a phenol-chloroform extraction, you end up with this massive white chalky pellet. It doesn't seem to purify differently from the gDNA, absorb peaks in 260nm, recalcitrant to RNAseA treatment, precipitates in alcohol and PEG along with DNA, but doesn't react with qubit dsDNA dyes. Some ideas at the end about what it is

Methodz Anyways, the solution comes from the esteemed and wise Guillaume Diss - slilca beads. What worked was roughly:

1. Collect somewhere around 2.5e9 cells. Stationary phase works. Centrifuge, loosen the pellet before freezing it.
2. Resuspend the pellet in 400uL of 20mg/mL PVP, 20mg/mL CTAB, 82mg/mL NaCl, 100mM TrisHCL, 10mM EDTA. Use this to put it onto a bunch of glass beads in a lysis tube. Add about 20ug RNAseA.
3. Shake the bejezzus out of it. We use a Biospec 607 for 5 minutes.
4. Incubate at 65C for about 30 minutes.
5. Add 1 volume of 24:1 chloroform:isoamyl alcohol, put back at 65C for 30 minutes.
6. Spin and aspirate the aqueous layer. Wash it again with chloroform and spin.
7. Precipitate the aqueous layer in a new tube with 0.7x volume isopropyl alcohol, wash with 70% etOH, then dry the pellet. Observe the massive white chalky pellet.
8. Resuspend with 200uL TE. This takes a while. Then the fun part starts.
9. Get yourself a Qiaex2 kit (or explore homebrewing it! see below). Mix in 1 volume of their QX1 buffer. Watch as a significant ammount of that weird crap precipitates. Spin it and keep the supernatant!
10. In a new tube, add the Qiaex2 beads to the supernatant, and follow the kit for purifying with that. Elute at 50C with two washes of 30uL EB.

That should get you about 15ug of genomic DNA, and you should be able to load in 2ug of this to a PCR (with NEB OneTaq plus a little magnesium). But, of course, you should probably back off to 1ug for consistency!

Why The QX1 buffer is likely just sodium perchlorate to push the DNA onto the silica beads. It should be feasible to homebrew this with just sodium perchlorate and ... well silica or glass milk. There's some protocols for this, basically you just push DNA onto sand with a chaotropic salt. However, it seems that some non-DNA precipitate forms before addition of the silica beads, so there might be a way to just purify it away from the DNA with just the sodium perchlorate?

But what the hell is this crap? It acts like DNA in alcohols, phenol, chloroform, PEG, and there's still some left through the silica purification. There's more by 260nm absorbance that could be DNA. Could it be RNA made recalcitrant to RNAseA? What about polyphenols covalently linked to DNA in the extraction procedure? What about polyphenol polysaccharide chains that look like DNA (charge-wise)?

Anyways, I thought this stuff was weird. Holla if you've any thoughts about it. And if you need to extract an obscene amount of gDNA from yeast, maybe reach for the old glass milk.