Cloning - how to avoid religation and no insert? - (Nov/14/2005 )

Hi!

I've just taken over a "cloning-problem" from a collegue who has struggled with this for several months now...
We want to insert a 2 kb PCR-fragment into the pGL3 vector (4.8 kb) and transfect this to DH5alpha. Both the insert and the vector is cut with XmaI and BglII. The insert has restriction sites generated by the PCR primers and the number of adjacent bases should be OK according to the NEB catalgoue (2 bases for the BglII site, and 3 bases for XmaI). With regard to the vector, these restriction sites are located relatively close in the polylinker, but not so close that it should be a problem.
To make a long story short, the problem that my collegue has had was lack of insert in her clones, in other words religation of vector (transfection efficiency has been OK). She has tried cutting the vector with one restriction enzyme at the time (in its optimal buffer) as well as both enzymes together in a "compromise-buffer" (buffer 2 w/BSA) and with long incubation with excess enzyme. After purification by Qiagen's PCR purification kit, an aliquote of the vector has always looked nice after gel electrophoresis in the sence that there seems to be no intact vector present. However, I have changed that strategy now and have so far cut the vector with XmaI alone and will purify the linear vector from the gel (and, indeed, the cutting in this case was not complete but I used more DNA than my collegue so maybe her cuttings were more complete than mine). If I then cut with BglII and purify by the PCR extration kit, this vector should be ready to use, I guess? (another gel extraction will not make any sence since the BglII removes only a few bases so I won't be able to visualize if this cutting is incomplete)
Is there any point in dephosphorylating the vector??
With regard to the insert, this has been cut with both enzymes in the "compromise buffer" and purified with the PCR purification kit (again, no gel extraction due to loss of only a few bases during the cutting). And yes, both enzymes cut OK in this buffer if only the amount of DNA is not too high and it wasn't in this case.
So, to summarise I wonder if anyone out there has any comments on my cloning strategy, and if dephosphorylation would help here (it shouldn't be necessary since we use two different enzymes).

Thank's

---

I dont think you need to dephosphorylate, but I definately suggest cloning the insert into a t-vector ie: pGem T-easy first this way you can actually tell if your enzymes cut the insert b/c it drops out of the pGEM vector...

---

Hi,

I forgot to tell that we used the Pfu polymerase, which gives blunt ends without any extra deoxyadenosine. I guess this vector won't be of much help to us, then...(unless there is something I have misunderstood, which is quite possible).

Sonja

---

you can pass it over. Just add 1µl of Taq to you PCR reaction. And incubate 10' at 72°. You can then clone it into pGEM T or topo ta kits.

By doing a pre cloning, you can be sure that after a gel purification, 100% of your "inserts" are result of an effective digestion (and in PCR digested, not all fragments are digested) and almost 100% contains propper ends for cloning

---

A technique we have found useful is to prepare the "vector" by PCR rather than by plasmid purification. The idea is to PCR all the way around the vector, with a primer containing the correct cut sites. You then mix vector PCR and insert PCR reactions, cut with the two enzymes, ligate, and go. The background vector contamination is zero, because there is no (well, carryover PCR template amounts, which can be made negligible) of the original vector. You are not typically worried about PCR errors in the vector, since if it propagates in the cell with the correct resistances, that is really all that is required. The vector PCR can be done in bulk. Cutting and purifying the vector PCR prior to ligation can make this process slightly more efficient, and can be done ahead of time.

---

Wow, that sounds like a very painful approach for routine cloning, phage.

I might try that as a secondary or tertiary option if things aren't working well, but that would hardly be my go-to option unless I knew the recipient vector was not cutting properly. You should be able to determine if the sites are there on a gel with two single digests.

-Matt

---

Hi everybody

Thanks for all your input
No news from the lab yet, but I've started my "investigations" at my institute to hear if someone has the pGEM T easy vector. It sounds like a very elegant way to ensure an insert with the proper ends for ligation - brilliant
In the meantime I will try another ligation where my vector is purified from gel after XmaI cut (which I believe is the less efficient of my enzymes), then cut with BglII and purified by the PCR pur.kit (and yes, I know that both cutting sites are present in the vector based on single digests, but I think the previous cloning attempts have been corrupted by incomplete cutting and contamination by intact or only "single-cutted" vector that religated). I will measure the DNA concentrations properly (we have a nano drop instrument) and make sure the ratio of vector:insert is around 1:3. My insert may not be perfectly cut, but if my vector now is OK, then theoretically only inserts with the proper ends should be able to ligate (however, I'm aware that there may be a large distance between theory and practice...). Anyway, I'll give it a try.
Thanks again

---

I don't know why you think this is any harder than "normal" cloning. You prep the vector PCR product once, make a batch, and then it is no more difficult than any other cloning. Except you have no background, with no CIP/SAP digestion.

---

Hello again

Well, I've tried another ligation with purified and cut vector and my insert (cut and purified) but without much success. The control (uncut vector) was OK, so the method itself worked and the cut/purified vector gave no colonies which also is good since the previous "religation-of-vector-problem" seems to be gone. But the insert wouldn't ligate so I thought I should try your strategy now.
So, today I've purified my insert which is a blunt-ended PCR product (Pfu polymerase) and now I want to add the 3' A. However, I just found some information showing that the polymerases may be more or less efficient making such additions depending on the sequence of the template. If there's an A already at the blunt 3' end, then the addition of the yet another A isn't necessarily very easy (and, one of my primers has a 5'-T end... while the other is 5'-C which should be OK). This seems to be a problem for DyNaZyme II (from Finnzymes) while I couldn't find any information about this for the AmpliTaq DNA polymerase (Applied Biosystems). I could add that I regard enzymes needing hotstart as not very practical for my reaction since I don't want to denature my 2 kb insert - I'm not sure how easily it will get doublestranded again, which I guess it has to be before ligating it with the TOPO-vector.
Any idea of specific polymerases especially suitable for adding 3'-A no matter what the blunt ends are?
Sonja

---

This might be a bit painful to do, but do you have internal primers for your insert that could be used to sequence up to the ends of the PCR product? Sometimes oligo manufacturers make mistakes - are you sure that both restriction sites are there? There is no easy way to know whether your insert is really being digested since the size difference is negligible.

As a separate comment, I think that Phage's technique would be really useful for situations where the insert has to be in-frame with respect to particular restriction sites, for the purposes of cloning into one or more expression vectors later on. It may be a handy alternative to what I usually do, which is to ligate dsDNA oligo linkers, containing restriction sites, into a half-way house vector, and isolate clones containing the linker insert, and then clone my gene into the modified vector. Phage's technique would take one cloning step out of the whole process.

---