Loss of protein expression following double transformation - Protein expression post transformation (Feb/12/2011 )

Hello!

I am doing double transformations of E.coli using the bacterial expression vector pTRC (Invitrogen) and a mammalian expression vector (Vical). The vectors are selectable using Ampicillin and Kanamycin respectively.
The plasmids for which transformations were done have been sequenced.
If I do a single transformation using my bacteria and the pTRC plasmid I get great GFP expression (This is the inserted protein). However when the bacteria has both plasmids the percentage of bacteria expressing the protein drops from nearly 100% to between 5 and 60% depending on the colony. I need to get it back to 100%.

What I have done so far....

pTRC is inducible using IPTG. I have tried a range of concentrations in double transformed colonies without significant variations in proportion of expressing bacteria.
I have tried various concentration of AMP (50-100 micrograms/mL) and Kan (10-50 micrograms/mL), again there is limited variations detected.
I have done transformations using 1 or 2 steps. When using the 2 step method I selected the optimal pTRC GFP expressing colony and made it competent. Nothing!

Does anyone have a solution for this problem?

Regards
Dave @ QIMR Australia

the success of your described strategy will be very much dependent on the origin of replication that your two plasmids will have!

Common plasmids used everyday in the lab have a so called ColE1-origin of replication (or sometimes referred to as pUC, pMB1, pBR322 ...but at the end they are ColE1-like origins). It is know that plasmids containing the same origin of replication are incompatible. Here is a link to a historical paper.
When using no selection pressure plasmid loss will be very rapid ...since you mentioned that you are using both anitibiotics i would assume that plasmid loss will be less pronounced ...but will happen as well ...leading to loss of one of the two plasmids and consequently to growth cessation.

Do you know what origins of replication your plasmids encode?
Why do you need two co-transform your E. coli cells with two plasmids ...one of them a mammalian expression vector? ...that does not make really sense to me? ...maybe you can explain your strategy in detail?

Regards,
p

WormsPhD on Sun Feb 13 05:41:25 2011 said:

Ampicillin resistance is generally rather sloppy. Beta-lactamase is excreted, so a culture can become resistant without the individual cells being resistant. I would suggest changing your antibiotic if possible (chloramphenicol or tetracycline, perhaps), or at least switching to carbenicillin instead of ampicillin as the antibiotic -- it is a little more stable.

Thanks for your input!

I don't believe that the problem is specific to ampicillin, because my positive control is a single transformed colony which produces high amounts of GFP even after continuous subculturing for at least 3 days. But I am prepared to give anything a shot, so I will track down some carbenicillin and give that a go.

I had read about the requirement for different origins of replication and was happy when I saw that one was pUC and the other pBR322. Very disappointed to be informed that they are the same creature under a different name. In fact a nucleotide blast confirms their identity. I assumed that because there was still resistance to both of the selective antibiotics, that the plasmids must still be present in the bacteria.

Is it possible that the daughter cells are being produced without the ampicillin resistant gene/plasmid are still viable at the time of analysis but the lifespan would be short as it would be susceptible to ampicillin? Remembering that the GFP plasmid is low copy number (pBR322) and the other plasmid is a high copy number (pUC). Working on this theory I will grow the bacteria at 37 degrees for the starter culture but when I add IPTG to induce protein expression I will lower the temperature allowing protein production but limiting growth/division of the bacteria. This will also allow the ampicillin to have its effect. I will try a group where I refresh the ampicillin at the same time that IPTG is added.

I will read up on the appropriate temperatures for incubation. If anyone has looked at this issue previously then please reply with results.

Thanks again
Dave

It would be best to spin down the cells and replace the medium rather than simply replacing the ampicillin, since the beta-lactamase is excreted and is enzymatically active.

What E. coli host strain do you use? ...if there is homology of your two plasmids and you use an rec+ strain, like BL21(DE3), you will end up with a recombination variant of your two plasmmids ...that maybe contains both resistance genes but has one of the expression units destroyed.

Regards,
p

I am using 2 strains. DS410 and P678-54. None of these are listed as being rec+. There is definately regions of homology between the 2 plasmids, particularly at the origin of replication.

Dave

interesting! ...are you doing secretion of proteins in E. coli?

there is also recA-independet recombination in E.coli (with a much lower frequency) ...do you ever tried to recover your plasmids after cultivation to check if they are still intact or if one is really lost? ...this would be an interesting.

Regards,
p

WormsPhD on Mon Feb 14 08:14:42 2011 said:

No I want the proteins to stay within the bacteria. These strains produce bacterial minicells which will be used as a vaccine delivery parcels with protein and plasmid DNA within. This has been previously done with some success but it is very experimental.

In fact today I did a miniprep and digest of the plasmids in a double transformed culture and both plasmids have remained intact. This was done from a clone that has been cultured over 3 days.

Dave

okay, than recombination seems not be an issue

i'm not used to the work with minicells ...but i can imagine that the fact that you use ori's of the same type somehow biases the segregation of your plasmids. If i were in your situation i would switch on of the plasmid backbones to another ori (like p15A, SC101 or some other origin) and another resistance than Ampicillin, like phage434 suggested. Maybe this helps.
But there is also the possibility that the phenomenon that you investigate is some intrinsic problems of minicells? ...do you ever tried co-expression in another "normal" E. coli strain?

Anyway, very interesting approach using minicells as a protein and DNA vaccine at the same time (never heard of that) ...elegant!

Regards,
p

WormsPhD on Mon Feb 14 08:45:39 2011 said: