pLenti-puro vector- effect of insert on plasmid stability?

Thumb 100 4420
Oct 05, 2017

I am trying to clone a 2.9 kb cDNA for a human long non-coding RNA (lncRNA) into Addgene vector pLenti-puro (7.1 kb). I am inserting directionally into NotI (sticky) and PmeI (blunt). The vector, which has lentiviral repeat sequences, has a tendency to be unstable. It tends to recombine after transformation even in recA-mutant host strains. I am using the recommended host strain Stbl3 which is supposed to reduce these recombinational events. The vector was dephosphoryated prior to ligation and the vector-only ligation transformation plate had no colonies. The colonies I got on the experimental (vector + insert) plate all produce what appear to be recombined plasmids. The desired construct would have been about 10 kb. These are varying in size from about 2.9 to 3.1 kb- just big enough to have the backbone ori and the AmpR gene, but not the lentiviral insert. I can cut them and get a single band with Asp700 (isoschizomer of XmnI; site located in the Amp gene, but there should be a second site in the lentiviral insert) but they do not cut with either NotI or PmeI. I also tried the same ligations in DH5alpha with the same results. I also tried incubating the transformation plates at 30°C, as recommended by some literature, to reduce recombination; this produced some late-appearing colonies, but again, they had spat out the lentiviral insert.

Yes, I know pLenti-puro has 2 NotI sites and I am using partial digests to get the vector cut only at the site located in the MCS. I have done partials before, and I have reasonable confidence in my prep of vector and insert, and my ligations. I am going to try testing the vector prep by cloning an innocuous 1.3 kb fragment of bacterial origin and see if I get anything from that ligation. But I know there are things you just can’t clone, or at least, they won’t go in in certain orientations, due either to the sequence being too “foreign” for E. coli, or too repetitive, or because it encodes something lethal to E. coli. And this lncRNA sequence has a lot of difficult regions: dinucleotide repeat runs, etc. I am beginning to suspect that the construct, if it does not recombine to spit out the insert, is lethal in E. coli. I suspect that the more I suppress recombination, the fewer clones I’m going to get.

The 2.9 kb lncRNA insert was originally cloned in pCR4-Blunt-TOPO (Thermo) by its originator, and I have been told that, despite the non-directionality of TOPO cloning, only one orientation of the insert was ever obtained, even though many clones were screened. The NotI and PmeI sites are provided by the pCR4 vector.

Meanwhile, if anyone has any experience with this vector, how large an insert have you successfully added? Has anyone else had trouble with inserts this size or larger in this vector? Or any other advice? It would be much appreciated.

Medium 100 4420


Former Senior Laboratory Technologist, A University in New York, USA

Cloning, Sanger Sequencing, PCR, and other Molecular Core functions, with an emphasis on Bacterial Resistance Plasmids


Thumb default avatar
mchlbrmn 10 months ago

I think you’ve already discussed this rather thoroughly yourself; I don’t know that I’m unlikely to add anything useful, so I’ll include more off the wall suggestions.
Use the tried and true method for this insert: clone it in backwards! : (
Ask the manufacturer of the vector about insert size; they should know.
They were able to clone it into pCR4, so it wasn’t completely toxic to the bacteria. Maybe a different vector would suit it better? Did you check for homology between the insert and vector?
If you’re going to such lengths as preparing a partial digest to clone the thing, shall we conclude that the insert cannot be PCRed? If it can be, then maybe a different cloning method, such as Gibson Assembly? Is it conceivable to PCR, and perhaps Gibson Assemble the entire construct, and transfect that?

Thumb 100 4420
OldCloner 10 months ago

Thanks, I appreciate the input. I am trying to troubleshoot a colleague’s project, and I would have planned it a bit differently if I’d been in at the start. The colleague wanted to avoid the extra expense of another round of amplification followed by the necessary re-sequencing of the insert, but that is what I am recommending at this point. This will allow the addition of more useful sticky R-sites and eliminate the need for the partial digest, which wastes so much vector DNA (I used up all they gave me). They are also looking at other vectors, with other viral sequences, like pBABE. But that has so few sites in its little MCS, amplification would be required, and they are dragging their feet on that one. I don’t think they’d go for Gibson, either, as I have no experience with it (beyond some reading) and there would be a learning curve.

Too bad backwards will not do! I can see several strategies for doing that! If the darn thing had gone into pCR4 in the other orientation I would have other options, too.

Thumb 100 4420
OldCloner 10 months ago

For any interested party, here’s the end of the story: I got the go-ahead to amplify the insert and added BamHI to the 3’ end and XhoI to the 5’ end. Also got permission to use a different vector, pBABE-puro. This vector has MoMuLV repeats instead of Lentiviral repeats, for infectivity. I have used pBABE vectors before and found them to be stable in regular E. coli strains like DH5-alpha.

I ligated the digested amplicon into pBABE that had been cut with BamHI and SalI (compatible with XhoI termini). But at first my few transformants were giving me the same results I got before with the pLenti vector: either empty vector or spontaneous recombinants that had spat out not only added insert but most of the viral sequences of the vector. One transformation I did had only one small colony and I just put the plate in the fridge until I could repeat the experiment and get more colonies to test. After another failure, and two weeks in the fridge for this plate, I decided to pick that colony-what the heck. But I noticed also some tiny pinpoint sized colonies on the plate and picked a few of those, too.

You can probably guess how this ends: those darn pinpoint colonies did indeed have the full size vector+insert that I was aiming to get. (The “one colony” was empty vector.) The desired clones grow very poorly and the plasmid yield is very low, but there they are. I think it is safe to say that the E. coli host does NOT like this insert at all, for whatever reason. Maybe they’ll only grow in the fridge?