Low recovery from Ampure XP beads

[nicolerp]

Hello, Working on a library prep for Minion sequencing. We are working with high molecular weight DNA. Attempting to recover very large fragments (>100kb). Losing about 75% on beads (seems smaller fragments removed first). We are loading 2ug gDNA in 60ul with 60ul Ampure beads with 2x fresh 70% washes. Elution in water. Sometimes, we can get more with second elution. Curious about changing to 80% ethanol or longer elution. Any suggestions?

[Olaf Blue]

I typically use 80% ethanol and two washes @ 175 ul each. Allow to dry for 5 minutes - elute with Low EDTA TE or "RSB" (Resuspension buffer). Water should be fine, too. Make sure you mix the beads well after washing with your eluant to make sure recovery is as complete as possible. Also, I always make sure that the 80% ethanol is fresh (no more than two days old or so).

[Ola]

I typically use 80% ethanol and two washes @ 175 ul each. Allow to dry for 5 minutes - elute with Low EDTA TE or "RSB" (Resuspension buffer). Water should be fine, too. Make sure you mix the beads well after washing with your eluant to make sure recovery is as complete as possible. Also, I always make sure that the 80% ethanol is fresh (no more than two days old or so).

Ethanol concentration in washes is not a problem with HMW DNA. Long elutions at 37* may help, just like for dissolving a HMW pellet after ethanol precipitation. Sometimes it seems you really need to shear it by pipetting to get it off the beads (you can still get a nice >60 kb peak).

[JackieM]

Ethanol concentration in washes is not a problem with HMW DNA. Long elutions at 37* may help, just like for dissolving a HMW pellet after ethanol precipitation. Sometimes it seems you really need to shear it by pipetting to get it off the beads (you can still get a nice >60 kb peak).

I second this. I've found that the more you mix the beads with your DNA, you can actually see the beads get lodged into your sample. I would make sure not to mix the beads for too long, not to overdry when you're on the magnet, and elute at 37C for...a long time. Also, I would recommend lobind tips when handling beads and not pipette mixing your beads and sample. You can get the bead/sample lodged into your tip and it's really hard to get out.

Another thing to consider- how do you know that you're losing your gDNA? I've found that if I nanodrop gDNA and then qubit it, the discrepancy will usually be pretty large coming out of extraction (Phenol chloroform extraction followed by several EtOH precipitations), even after RNAse treatment. When I clean it up with XP beads, the nanodrop will drop to match the Qubit because I've gotten rid of the little random nucleotides and things that artificially increase the absorbance on the nanodrop.

I've also found that after gDNA extraction, my concentration will increase dramatically if I let it sit at 4C for a few days. There is usually a snot-like piece of gDNA that takes a while to dissolve at anything above a few hundred ng/ul. EtOH precipitation and the XP beads are similar enough that you may be having the same problem.

Good luck with this! I tried the MinION high MW sequencing too a while back, just as a fun little project, and had very little success. I would be interested to hear if you are able to sequence some large fragments!

[pmiguel]

And just so I can sound like a broken record on this topic, unless you checked the concentration of your genomic DNA using an agarose gel or fluorimetry, you may be fooling yourself as to how much DNA you actually started with.
That is UV spectrophotometry always overestimates the amount of DNA in a genomic prep unless heroic efforts were made to remove potential confounding molecules other than double-stranded DNA that absorb UV at 260nm.
See here for more details...

--
Phillip

[gringer]

And even fluorimetry can get it wrong. If there's a huge amount of RNA in the sample (e.g. 100x as much as DNA), or contamination with a fluorescent substance (e.g. coffee), it can affect the result.

At PoreCampAU this year, it was suggested that we should elute at higher temperatures (e.g. 37°C) if we were having trouble with getting enough DNA.

[JackieM]

All this talk of gDNA quantification gave me a thought- when you add your buffer to the beads and incubate it, is your concentration/molecular weight high enough that you can see the DNA in solution? If so, perhaps you can wait until you can see the pellet dissolve?

And even fluorimetry can get it wrong. If there's a huge amount of RNA in the sample (e.g. 100x as much as DNA), or contamination with a fluorescent substance (e.g. coffee), it can affect the result.

This is very true. And I'd like to add an additional problem for the fluorometer- the concentration you get is as good as your ability to pipette. High molecular weight gDNA does not come into solution well and it tends to pull itself out of your pipette tip. If you can see the gDNA (you will see a clear blob in the solution that catches bubbles) you need to incubate longer and/or add more buffer before you quantify.

If I care at all about the real concentration of my DNA I will use a fluorometer and a spectrophotometer. If there is 100x as much RNA, you will still get a massive difference between the two methods. (spectro being way higher.) Usually when I care about the concentration, I will also care about the quality, so I will also visualize it on an agarose gel- which, in my experience, has only disagreed with the fluorometer when I have freshly isolated, concentrated, high molecular weight gDNA that is still in the "snot in a tube phase" and really should have been incubated overnight with additional buffer.

I had never heard about the coffee thing though! That's very interesting. In my case, if there's coffee in my sample it's time to go home and start over, but now I'm interested to know if there are other similar substances that will throw off a fluoro reading...

[Carcharodon]

I find that over-drying the beads is incredibly easy to do, and will cause massive reductions in yield. For smaller volumes of bead, I don't recommend drying for more than two minutes.

Reducing drying time from five minutes to two to three minutes has had an enormously positive impact for me and some of my colleagues.

Good luck!

- Sean

[JackieM]

I find that over-drying the beads is incredibly easy to do, and will cause massive reductions in yield. For smaller volumes of bead, I don't recommend drying for more than two minutes.

Reducing drying time from five minutes to two to three minutes has had an enormously positive impact for me and some of my colleagues.

Good luck!

- Sean

I don't mean to spam the thread, but this is a great point and reminded me of something ONT tech support advised me: If you're not sure, it's better to add your RSB before the ethanol has dried as opposed to over-drying and the tube can be left open at 37C while eluting to evaporate the remaining ethanol. "Wet but not shiny" is a good rule.

[pmiguel]

And even fluorimetry can get it wrong. If there's a huge amount of RNA in the sample (e.g. 100x as much as DNA), or contamination with a fluorescent substance (e.g. coffee), it can affect the result.

I have seen "genomic DNA" samples that had 100X the amount of RNA in them as DNA. And yes, even a double-strand-specific fluorescent dye will not faithfully report the correct DNA concentration in such a situation. But it will get a couple of orders of magnitude closer than UV spectrophotometry will.

Being 50% off on the concentration of your input DNA is a different sort of error than being 10,000% off. In the case of the former you may well still get results. But in the case of the latter -- that is a catastrophic error that will not yield a usable library.

The deeper issue here is that all measurements are subject to systematic errors of one sort or another. So you want to be aware of the limitations of the assay you are performing, not treat them as infallible. Absorbance at 260nm of a solution is just that -- the amount a solution absorbs at 260nm, not necessarily the concentration of your DNA.

--
Phillip

[devking]

I'm having this same issue. nicolerp, were you able to find any solutions with the AMPure bead cleanup of HMW fragments? When you guys say to elute at 37C for a long time, are we talking minutes or hours?