I designed a synthetic construct on paper and got it synthesized from a company. The objective was to make a vector which can be used to study both transcriptional and post-transcriptional regulation by different cis-elements (2-3 different cis-elements can be studied simultaneously).

The construct is like this:



Kozak consensus sequence = GCCACCATGG

GFP, RFP and YFP are fast turnover variants with MODC pest sequence. I took the sequences for these proteins from Evrogen website. I took the Sv40 polyA signal sequence from one of their vectors in order to avoid any event of incompatibility. Kozak sequence was also taken from those vectors only. The total size it ~4.8kb. For CMV promoter I took the standard sequence used in most vectors.

With this vector I could clone any post-transcriptional regulatory sequence between stop codon and SV40PA, and could change promoters as well.

With the CMV promoter and no 3'UTRs I see a very faint expression of GFP, but could not see any RFP expression. I used pMaxGFP and dsRed as a control for filter sensitivity.

I transfected the plasmid in Hela and Neuro2a and at various concentrations (1-5μg) using lipofectamine. The controls express well (transfected 1μg each) but I see very faint GFP expression and dont see any RFP expression.

I checked the sequencing QC report given by the company and there were no indels or other mismatches. I am clueless about why the construct fails to express.

Is the linker a problem; is 100bp not sufficient ?

  • $\begingroup$ Did the company synthesize the sequence as above or did they codon optimize it for a species? You can really mess up folding if you're not careful with translation gradients. Also where did you put the Kozak sequence? I couldn't tell from the question. Need to make sure we have everything in the right order, gene start-stop, etc. $\endgroup$
    – Atl LED
    Commented Oct 7, 2013 at 18:17
  • $\begingroup$ i just gave them the sequence.. it is codon-optimized for mammalian expression. Kozak sequence was obtained from the ORF source i.e Evrogen vector.. Kozak sequence is overlapping with the start codon and I just took the entire sequence from start of kozak sequence to stop codon $\endgroup$
    Commented Oct 7, 2013 at 18:47
  • $\begingroup$ I have used an GFP+RLuc plasmid before that had linker just shy of 100BP, so I don't think that's the issue. Have you done any westerns or northerns to see if you're getting some kind of read-through? $\endgroup$
    – Atl LED
    Commented Oct 7, 2013 at 19:34
  • $\begingroup$ no.. I dont have the antibody or the primers/probes yet.. I have also tried electroporation.. no great difference.. I'll try it once more and see.. Point of using GFP was to have a direct readthrough $\endgroup$
    Commented Oct 7, 2013 at 20:10
  • $\begingroup$ GFP antibodies are not too hard to run across. I might try and see what your base level of expression is. Otherwise the only way I know to go above CMV in expression is to use T7. And then you have to worry about your T7 source. $\endgroup$
    – Atl LED
    Commented Oct 7, 2013 at 20:12

1 Answer 1


I'm afraid there's not much of answer but back to the bench with you! (That's why it's called re-search).

There are several factors that could be playing a role here. You could have transcriptional or translational read through. You should check your sequence and the cells you are going into to make sure you aren't getting Amber (UAG) suppression.

Before you do anything, you should check your DNA of all your groups. Run it uncut on a gel and make sure the majority of the DNA is super coil. The next easiest thing to check is translation problems. Get an GFP antibody, possibly a few that recognize different epitopes, and see if your GFP is coming out the right size. It may not be processed correctly or folded correctly.

You can run into problems by optimizing the sequence. First, synonymous mutations can cause folding variants (1). Second, you might be causing a hair pen or other disadvantageous secondary or tertiary structure. If you can't detect any problems in your Western, you might consider doing a Northern to trouble shoot the transcription.

If you rule everything else out, and you are just getting poor expression, then the only way I know to boost over CMV is to use T7. The problem with T7 is always going to be what use for a source. If you can use BSR/T7 cells and you can get your hands on them, they work great. Otherwise you're looking at a T7 plasmid or MVA-T7 as common sources. The plasmids never lead to a great deal of expression in anyone's hands I've ever known. MVA-T7 works, but to say that it confounds things would be the under statement of the year.

  • 1
    $\begingroup$ Re-search That's perfect. $\endgroup$
    – Amory
    Commented Oct 7, 2013 at 22:05
  • $\begingroup$ Can't take credit for that one. Heard it back in the 90's from an advisor, and I'd bet he didn't come up with it ether. $\endgroup$
    – Atl LED
    Commented Oct 7, 2013 at 22:08
  • $\begingroup$ It never had the time or resources to modify the vector and eventually gave up this project. I am accepting this answer as it summarizes all issues and would be useful for others who seek to work on synthetic biology. $\endgroup$
    Commented Jul 24, 2014 at 6:27

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .