to BioTechniques free email alert service to receive content updates.
Gateway Cloning
Kristie Nybo, PhD
BioTechniques, Vol. 56, No. 4, April 2014, pp. 170–171
Full Text (PDF)

This month's question from the Molecular Biology Forums (online at comes from the “DNA and General PCR Methods” section. Entries have been edited for concision and clarity. Mentions of specific products and manufacturers have been retained from the original posts, but do not represent endorsements by, or the opinions of, BioTechniques.

Molecular Biology Techniques Q&A

With Gateway cloning, why do I find empty destination vectors? (Thread 33177)

Q I'm performing LR reactions using Gateway cloning, but I repeatedly end up with empty destination vectors. The Gateway cassette with the ccdB and chloramphenicol gene is cut out, but my gene of interest is not ligated into the vector. Is this a common problem?

A In my experience, Gateway cloning is more reliable than ligation. I just checked my recent experiments, and three clones each of seven inserts all worked. My only concern with Gateway cloning is that in my lab, these constructs seem to have a lower success rate for expression than clones made by ligation.

Did you sequence the clones?

Q I haven't sequenced yet, but I found the empty vectors by digestion.

I tried a reaction with only the destination vector and LR Clonase and transformed into E. coli. The LB plates with the selection antibiotic are full of small colonies; this should not happen. Even though this plasmid contains the antibiotic resistance gene, it should not become circular.

A What host are you using? If you use DH5, the cells can't grow if the ccdB gene is still in the plasmid.

A It could grow if it kicked out the ccdB insert. I have seen that before, but infrequently. Some vectors may have more trouble with this than others because of the potential for recombination at the inverted att sequences.

Q I use TOP10 cells, so I don't think this should be a problem. Isn't it strange that this reaction is happening without all the att sites? I thought the LR Clonase needed both the attL and attR sites to work.

A Are you sure TOP10 cells work with ccdB?

Cloning bacteria have some recombination genes deleted, but strange things can happen in the presence of strong selection, which might be outnumbered by productive rearrangements in the presence of a good entry vector. Maybe something is wrong with the entry vector or the bacteria don't tolerate the insert?

A The empty vector might be a minority contaminant with the destination vector. Have you tried transforming from just your destination vector plasmid prep (without Clonase)?

If this is a high-copy destination vector (or even if it isn't), it is possible that the colony you originally picked to prep the destination vector was double-transformed with the Dest-ccdB-selection vector and the empty parent vector, and managed to maintain both plasmids. Then you would have purified both plasmids. So after the Clonase reactions, you're transforming both the Gateway vector and a minute quantity of empty parent vector that is not visible on a gel. A supercoiled empty parent vector would be smaller and more abundant than successfully recombined Gateway final vectors in your reaction mix, so the majority of your colonies would carry the parent vector.

You might need to go back and reconstruct the destination vector, or at least re-transform and isolate a colony that is guaranteed to have only the destination vector. This would all be easier if the parent vector could be digested with an enzyme that doesn't cut the destination vector; you could look at the multiple cloning site of the parent vector to see if this is possible.

Q Yes, they are sensitive to the ccdB toxin. The plasmid is in the TOP10 cells, so if the toxin is active, the cells should die. With the LR Clonase, the toxin might get removed, but this seems unlikely without the attR sites. The selection was done using a normal concentration of kanamycin.

A Are you sure the att sites you chose for the destination vector cannot recombine with each other? Is it possible that there is a small PCR fragment, or some other ligation fragment, that contains both of the attL/R sites needed to recombine with the destination vector sites but with no ORF between them?

Q How could I check for those things? The att sites are just the ones that came with the Gateway conversion kit. The attR sites are based on the example PCR mentioned in the user manuals.

I don't think there could be a small PCR fragment carrying those sites. I received the plasmids with the ORF between attL sites in a “donor plasmid,” and I only have the destination vector with the Gateway cassette in it from the kit.

A Without Clonase, do you see any growth? If not, there is the possibility that the entry vector itself is contaminated with some non-ligated parent entry vector. This vector might preferentially recombine. This wouldn't explain why you get results with the destination vector and Clonase though.

Have you tried sequencing across the insertion site on some of the negative clones? That might tell you what is going on.

Alternately, have you run the entry vector on a gel? I would recommend a restriction digest with an enzyme that would leave about a 0.5 kb band for an AttL-MCS-AttL vector, and a 2.5 kb band (or whatever size) for the expected AttL-ORF-AttL vector. That should be sufficient to determine if you are using a mixed-colony plasmid prep. Are there any sequences in your parent destination vector that resemble an attR site? How large is the entry vector and what is its antibiotic selection? Have you tried transforming the Clonase itself? Perhaps it was contaminated.

Q I have an entry plasmid with the ORF, and all plasmids were sequenced by the people who sent them to me, so they should be correct.

I don't think there are any sequences in the parent destination vector that resemble an attR site, but I will need to check it again.

The donor vector is about 8–9 kb, and the destination vector is about 10 kb. I am using gentamicin for the entry vector and kanamycin for the destination vector.

I will try the suggested experiments to look for any contamination in the reagents.