Odd point mutations - (Aug/13/2009 )

Lately I've been doing a lot of cloning into pBluescript and some derivatives made in our lab. The plasmid is then going into XL1-Blue MRF' cells by electroporation. This has been a successful and efficient strategy in the past, but lately, sequencing of the plasmid constructs has shown a seemingly inordinate amount of point mutations in the sequence.

I originally asked a question regarding this a couple months ago, thinking it may have had something to do with using a ligation reaction product as template, and perhaps that had an effect on the fidelity of Phusion. However, sequencing the DNA that was used for ligation into pBluescript showed no mutations from amplification using Phusion, so the point mutations (there were 4 total in a 2 kb sequence) all occurred somewhere between ligation into pBluescript and a plasmid prep out of the E. coli strain. I have a hard time believing the mutations are being introduced by E. coli, but that seems to be where the mutations are occurring.

I tossed out the original construct described above and remade it from scratch. Upon insertion into pBluescript and sequencing, I again found 2 colonies with the insert and sequenced the plasmid. One of them was fine, the other again had some point mutations.

At this point, I figured maybe it was just the gene I was inserting, and maybe it was an isolated thing.

However, within the past couple weeks, I've been doing some more cloning and have seen a few more of these point mutations pop up, again in pBluescript in XL1-Blue MRF'. I have not yet gotten to sequencing the PCR products used for cloning, so it is possible that they are the source of the mutation, but based on my experience with the stuff above, I think it's probably not from the polymerase. I had 4 clones from a construct I'm making, and each one of them had a different point mutation.

Most of the mutations are G>T or C>A, but I noticed a couple today were C>T.

The really odd thing is that I'm doing different constructs simultaneously (basically I'm flag-tagging a bunch of bacterial proteins), and 3 of the other fusions I'm working on cloned fine, and their sequencing is perfect. That leads me to believe it's not the polymerase, and it's not any sort of mutagen contamination in any of the chemicals or media that I'm using, because they're all being made with the same reagents and grown in/on the same media.


Any ideas on what MIGHT be going on? I'm going to go back right now and compare the mutated to the non-mutated ones to see if any were gel-purified as opposed to PCR purified (UV damage). I really can't think of much else that it could be. Could the cells have lost their error-proofing while in the freezer? I don't know of any times where our stock cultures would have thawed, or if anyone would've let them thaw while using them.

fishdoc on Aug 14 2009, 06:35 AM said:

Hi,

Which polymerase are you using? I had 5 point mutations in my insert because i was using a non-cloning grade polymerase.. yet a fellow student of mine had perfect sequence using the same enzyme? Ever thought of using Blue-light (SYBR safe) instead of UV when gel purifying?

Cheers

Sequencing PCR products directly prior to cloning will never show the accuracy of the PCR. You must clone the fragments, and sequence individual clones. Cloning selects a single molecule, which may have errors. Sequencing from a PCR reaction provides the average sequence of a very large population of DNA fragments. The errors are happening in your PCR reaction; you just cannot see them by sequencing the result.

phage434 on Aug 16 2009, 06:07 AM said:

I agree with this. It is the most likely scenario. However, are you cloning from the original bacterial source or have they been propagated in the lab for too long such that point mutations might occur? This is a long shot, but something to think about when maintaining strains.

Teagan27 on Aug 15 2009, 11:31 PM said:

Phusion from NEB. I have thought it could be the UV, but if that were the case, wouldn't my sequencing of the gel purified product show the mutations?


edit: nevermind, I see where it could still be from the PCR product and the sequencing not picking it up. I will go back in my notebook and compare sequences that were gel purified as opposed to purified right from the PCR reaction to see if I can find any correlation to more mutations in the UV exposed samples.

phage434 on Aug 16 2009, 09:07 AM said:

I had considered that option, too, but wasn't sure it was a real possibility or not.

I went back in lab notebooks and compared clones using gel purified DNA to PCR purified DNA. Of 4 constructs made using DNA from gel purifications, 3 of the 4 had perfect sequences following cloning. The 4th had 3 isolates with an insert, two of which had point mutations. Of those 2, 1 of them had 4 mutations in a 2 kb sequence.

Two other clones using PCR purified DNA had 1 with a correct sequence, and another that had point mutations. Of the 4 isolates of the 2nd clone, each one of them had a single point mutation, all at different positions throughout the sequence. This insert was about 1.2 kb.

So in total, I've made 6 constructs recently, 4 using gel purified DNA and 2 using PCR purified DNA. Five of these originated as PCR products; gel purification was required for a couple because of some nonspecific amplification. Of the 6 constructs, two of them had multiple point mutations, but apparently it's not related to gel purification (UV), because it's happening with the PCR purifications not associated with UV exposure.

So I guess it's either in the bacteria after cloning, or it's a product of the PCR reaction. But as I said above, most of the cloning I've done that results in an insert originates from PCR reactions, and not all clones end up with any mutations. So I wonder why it would happen with some products, but not others. All reactions are done using the same conditions (other than annealing temps) with Phusion polymerase. I received the suggestion when I originally posted this problem that using DMSO (1.5% per reaction) could cause a problem, and maybe that's true. I haven't removed that from my reactions, but I think I will when I go back to amplify other DNA. In any case, however, it can't be a certainty that the DMSO is causing the mutations, because other products work just fine.

This is really odd, at least to me.


As for the question regarding the source of the template DNA, it's from a genomic DNA prep. The original source of that prep is a frozen stock. The genome of the bacterium has been sequenced. Had these mutations been consistently in one place in regards to the genetic sequence, I would think it possible that it may just be a change in the sequence from one strain to another or something like that... or too much time in the freezer. But seeing as the clones are having different mutations in different positions, I don't think it's due to the DNA template having mutations, but rather the polymerase is making mistakes (not likely, IMO) or the E. coli strain is making mistakes (even more not likely).

My next step would be to switch to another high fidelity polymerase and see if the situation improves. You could also try doing fewer cycles, or drop your extension temp a bit, or try another E. coli recipient strain...

But, a further question -- why do you need more than one mistake-free clone?

HomeBrew on Aug 16 2009, 08:19 PM said:

I don't. I was afraid I wouldn't be clear in describing my problems, and apparently I wasn't. If I get one mistake-free clone, I'm happy. But I'm doing 6 different constructs. Four of them are gene fusions to FLAG and the other two are mutagenesis projects. So when I'm talking about the different clones, it's the 6 different constructs (4 were gel purified for cloning, 2 were PCR purified) and so forth.


I've contemplated using another polymerase. We used to use Pfu Ultra II, and that always worked well, but then I saw Phusion one day, and it was about 1/2 the price for 2x the reactions. We tried it out, and it seemed to work great. We had been using it for a couple years before we (I) ran into the current issues. I don't think it was a bad stock of the polymerase, because we go through it pretty quickly, and I'm sure we've ordered new a few times since the first problem came up. I will try doing the current "bad" construct one more time with Phusion and if it gives poor results I will try Pfu Ultra II again.


As for the E. coli strain, I'm currently in a back and forth with Stratagene regarding the XL1 Blue cells to see if they have any ideas. I'm guessing they'll tell me that our freezer stock may just be old and to get a new one. I'm not sure when the freezer stock was put in there, if it's from an original or a subculture along the way, but I think we got that strain with the plasmid, and the plasmid has been in the freezer since 1999. Who knows.


Anyway, I appreciate all the thoughts and considerations. Maybe something will work, or I'll just get lucky and find a clone out of a few isolates without a mutation and just go from there.

OK, so here's where I am with this problem.


I have been in contact with NEB regarding their Phusion polymerase, and they've basically given me the "I have no friggin' clue" response. The DMSO I had been using was not a problem (according to them), my reaction conditions are fine, they had no history of complaints about that lot of Phusion, etc. They suggested maybe the dNTPs were out of whack, and asked if we made our own or bought commercially. We purchase from Bioline, and aliquot out to avoid to much freeze-thawing, and I've been through multiple aliquots with the same results (maybe the original stock from Bioline is bad?)

I then found some polymerase from about a year ago (Pfu Ultra II from Stratagene) and decided to use that side by side with the Phusion. Both amplified well, and I inserted a portion of each PCR product into pBluescript and screened with X-gal. For the Phusion, I got 15 white colonies and for Ultra II, only 5. I did colony PCR on each, and 3 of the 5 Ultra II clones had the correct sized product while only 1 of the 15 Phusion had the correct size. I sent the 4 good ones for sequencing. Of the 3 Ultra II clones, 2 of them had a single point mutation in them, but the third was fine. The Phusion sequence was very odd. The colony PCR using T3/T7 primers showed an insert, but the sequencing showed no insert. At this point, I really didn't care, because I had the one good one from Ultra II.

In the meantime, I was also constructing some complementation plasmids, and decided to use Ultra II. Today, I got back the sequencing for 2 of them, and in both cases, only 1 of 3 clones was correct sequence. Each of the others had 1 point mutation, all in different positions.

Thus, it seems it is not the polymerase that is causing these point mutations, because I am getting them with two (supposed) high fidelity enzymes.

Aside from one particular gene I'm trying to clone, I've ultimately gotten 1 good clone to use for further manipulation, but this level of mutation is very troubling.

The common reagents used in both the Phusion and Ultra II reactions are the dNTPs, the primers, and the template. The template is genomic DNA stored at 4 C, and it's been in there a couple years. The only other thing I could think of is that the 35 cycles is too much, thus causing both the Ultra II and Phusion to lose some fidelity. However, I'd guess you'd still expect a good majority of your amplicons to be of the correct sequence if that were the case.

I'm really at a loss for ideas on what could be causing this. We've got new Phusion and new Ultra II on the way, and I'm going to do some more side-by-side stuff to test them.

Any thoughts or suggestions?