I am trying to clone 2 copies of the same mammalian gene into the pSF-CMV-Ub-Puro Ascl (contains HindIII, KpnI and NheI cut sites) plasmid.
This is what the final product is supposed to look like in the pSF plasmid:
[split GFP -- Protein A -- P2A sequence] -- [Protein A -- split GFP]
For the inserts, I PCR amplified the [split GFP -- Protein A -- P2A sequence] to add HindIII and KpnI restriction sites, and [Protein A -- split GFP] to add KpnI and NheI sites. Then, I cut the PCR products with their corresponding RE enzymes (NEB high fidelity enzymes, double simultaneous digest at 37c, overnight), and gel purified the insert.
For the vector, I cut the pSF plasmid with KpnI and NheI in a simultaneous double digest, gel purified it, then I ligated the [Protein A -- split GFP] fragment with KpnI and NheI sites into the cut pSF plasmid. This was successful and it has been confirmed by sanger sequencing to be correct.
This is where the problem starts. To insert the [split GFP -- Protein A -- P2A sequence] fragment, I cut the pSF plasmid containing the [Protein A -- split GFP] insert with HindIII and KpnI again, then ligated the [split GFP -- Protein A -- P2A sequence] fragment into the plasmid. I saw many tiny colonies. However, when I did colony PCR, they were negative. The truly strange part is when I picked some clones to sequence, the results show that only part of the split GFP and part of the P2A sequence was present. The entire protein A sequence was missing.
I am puzzled. How can an entire gene go missing after ligation? Is it some strange recombination event in DH5a? Is this particular insert toxic? Any ideas VERY appreciated.
Extra details:
- The final plasmid size is ~11 kb. The pSF plasmid is ~6 kb, and each insert is ~2.5 kb and ~2.8 kb.
- I have done single digests of the pSF plasmid containing the [Protein A -- split GFP] insert with HindIII and KpnI. I saw single bands corresponding to the size of the linearized plasmid, indicating that these enzymes are functional and cutting as they should.
- I have sequenced the [split GFP -- Protein A -- P2A sequence] insert. It is correct and all components are present. Sequencing results also indicate no extra HindIII or KpnI cut sites within the insert. I also have run the insert after RE digest on a gel to confirm this.
- I have added extra bases to the end of my primers that contain RE sites. I followed the NEB table showing optimal bases flanking restriction sites for maximum RE cleavage activity.
- I am using both the Promega and NEB T4 ligase kit. For ligation, I have tried 1:3, 1:5, 1:7 and 1:10 insert:vector ratios, and vector concentrations of 10 ng, 100 ng and 300 ng. I have tried ligating at room temp for 3h and overnight, as well as 16c overnight. I used the NEBioCalculator to calculate my ratios and concentrations.
- After ligation, I transformed the ligation mix into DH5a competent cells with heat shock. My protocol for heat shock is: add 10uL of ligation mix into 50uL of DH5a, ice for 20 min, 42c for 45 sec, ice for 5 min, add 700uL SOC, recover with shaking for 1h, concentrate cells and plate on kan (50ug/mL working concentration).
- I have run positive (pSF plasmid with single cut) and negative (pSF plasmid containing the [Protein A -- split GFP] insert after double cut) ligation controls. As expected, positive control show colonies, while negative controls do not.
- I tried to insert the [split GFP -- Protein A -- P2A sequence] fragment into an empty pSF plasmid. There were no colonies at all.
- I use UV to visualize all my gels.
Are the two copies of the gene literally identical?
If so, you'll have real problems with homologous recombination. Any slippage during replication, or strand breaks during normal cellular activity, and the chances are high that one copy will get spliced out.
Since a plasmid with only a single copy is much, much more stable, even if this is a rare event it will propagate.
It's also a pretty big plasmid, final size: bugs hate that, so that again promotes plasmid truncation.
I guess, really, the question is...why are you doing this?
The other thing to look for is the chi sequence, which GCTGGTGG, in your construct. These sites trigger recombination via the recBC recombination system, which DH5a, being recA-, likely has intact.
Dear Sy Teo
Based on the points made by John Hildyard & Gilmore and your logical observations, I could find some possible solutions.
1. If the bacterial host is imposing the pressure to reduce the size of the recombinant plasmids, you may go with 10-beta cells (NEB) to overcome this issue.
2. As the insert you are using is about three times smaller than the recombinant vector, better you use the insert 3 times lesser than the vector, using not more than 50ng vector in 10ul ligation mix.
3. As the excess amount of ligase enzymes have a negative effect on the transformation efficiency, the amount of ligated product you are using for transformation is too high for 50ul of comp cells. Optimum would be 5ul in 100ul comp cells.
4. Most of the conventional enzymes work better at 4*C for overnight incubation.
Best wishes
- Badges
- Science topic