LTRs are not functionally equivalent to poly-adenylation signals; please don't conflate the two. And blanket statements regarding endogenous RNAs and polyA signals are problematic. Small regulatory RNAs are not polyadenylated; and in fact, not all RNA Pol II mediated transcripts undergo polyadenylation (there is a whole literature on biases in cDNA libraries caused by oligo-dT tailing).

In terms of transgene delivery, the theoretical advantage of including a polyA signal is that it can improve mRNA stability and boost translational efficiency. Whether that is desirable depends on the level of transgene expression you wish to achieve, and certain intrinsic parameters of mRNA and protein stability for your gene of interest.

CMV-driven GFP does not need a polyA signal in lentiviruses, although efficient mRNA export from the nucleus can be enhanced by including the WPRE element from the Woodchuck hepatitis virus or the CTE element from Mason-Pfizer monkey virus. Many lentiviral constructs include the WPRE or CTE element 3' to the multiple cloning site (and upstream of the 3' LTR).

Some polyA sequences are tolerated in lentivirus when placed in the inverse orientation as Alok suggests. However, you cannot do this with CMV-driven transcription as you will have transcriptional interference between CMV and the 5' LTR that will screw up transcription of your virus during the production phase. So designing antisense gene cassettes using a polyA sequence is not completely trivial. Nonetheless, it has been done using weaker promoters and the BGHpA signal.

I've actually been dealing with this same issue. I'm using a lentiviral system that has two promoters, and I wanted to use a polyA after the first set of genes in addition to the polyA in the LTR after the second set. Here's what I've found:

I get expression without adding my own polyA. I think the expression is slightly better with a polyA. If I include a polyA in the sense direction, it reduced my titer about 3-fold (probably interferes with viral RNA production during packaging). How much the titer goes down probably depends on the strength of the polyA signal.

Check out HUMAN GENE THERAPY 19:840–850 (August 2008)

Bryce

Are you using a lentiviral vector???

@Bryce: that's very interesting, from all the accounts that I had heard before, adding more polyAs even in the same direction results in lower titer.

Unfortunately, I am not using a lenti or retro vector. I am basically wondering if I clone pgk-gfp into say, puc19 or some other minimal plasmid which does not contain a polyA signal, will I get detectable expression? Or is the stability and processing by polyA essential for expression?

Thanks!

You need polyA for mRNA export and translation termination.

In lentiviral vectors, the LTR acts as a polyA, so if you have LTR-Promoter-GFP-LTR, that should express GFP well at both transfection and virus level.

If you have LTR-Promoter-GFP-pA-LTR, you will poly-adenylate viral genomic transcripts and force them to not have the 3' LTR, reducing your titers.

The best way to put a polyA in a lenti is : LTR-RevCom(pA-GFP-Promoter)-LTR. Where the entire expression cassette is inserted in the reverse orientation.

LTRs are not functionally equivalent to poly-adenylation signals; please don't conflate the two. And blanket statements regarding endogenous RNAs and polyA signals are problematic. Small regulatory RNAs are not polyadenylated; and in fact, not all RNA Pol II mediated transcripts undergo polyadenylation (there is a whole literature on biases in cDNA libraries caused by oligo-dT tailing).

In terms of transgene delivery, the theoretical advantage of including a polyA signal is that it can improve mRNA stability and boost translational efficiency. Whether that is desirable depends on the level of transgene expression you wish to achieve, and certain intrinsic parameters of mRNA and protein stability for your gene of interest.

CMV-driven GFP does not need a polyA signal in lentiviruses, although efficient mRNA export from the nucleus can be enhanced by including the WPRE element from the Woodchuck hepatitis virus or the CTE element from Mason-Pfizer monkey virus. Many lentiviral constructs include the WPRE or CTE element 3' to the multiple cloning site (and upstream of the 3' LTR).

Some polyA sequences are tolerated in lentivirus when placed in the inverse orientation as Alok suggests. However, you cannot do this with CMV-driven transcription as you will have transcriptional interference between CMV and the 5' LTR that will screw up transcription of your virus during the production phase. So designing antisense gene cassettes using a polyA sequence is not completely trivial. Nonetheless, it has been done using weaker promoters and the BGHpA signal.

@Daniel: Thanks very much for the answer. So my understanding is that adding a polyA would definitely help a non-lentiviral plasmid express my protein of interest (eGFP) when it is under a pgk promoter. However, if I were to not include it, the amount of signal would be lowered by some indeterminate amount.

Do you have any personal experience with omitting the polyA for gfp?

The reason why I keep drilling at the question is because the addition of a polyA signal would require another fragment to be cloned into my construct, taking it to a total of 7 pieces which is quite high. But, to be safe I will probably just suck it up and clone it in since it seems to be quite unpredictable.

Thanks again!

Hi Brian. Yes, I've examined polyA manipulations in both plasmid and lentiviral contexts. If you're doing transient transfections in cell culture (you can add a ton of DNA and look at 24hrs), you don't need a polyA tail. If you need long-term expression or are looking in a context where it is hard to deliver your transgene, then I'd definitely recommend incorporating a polyA sequence into your gene cassette. Typically these are quite short (220 bp or so) and are present in a ton of commercial vectors. Perhaps you can assemble your 7 piece cassette into a plasmid that already has the polyA sequence. What is the design of your full gene cassette and does the level or duration of GFP expression matter for your downstream applications?

Hi everyone, 

I am kind of facing a similar issue and thought I would post my question here instead of opening a new thread. 

The construct I am working on is for transgenesis through recombination. So, basically I have a species-specific promoter which I know works. If I have a fluorescent gene of interest downstream to it, but no poly-A, once the cassette is recombined into the genome, do you think I would get a decent level of the fluorescent protein or should I make sure to include the poly-A site? 

Thanks!

I believe the consensus is that for long term sustained expression, a polyA is not necessary, but will help.

A bgh polyA signal is like 250 bp, so you should be able to add it without too much problem, as long as it's not part of a viral vector.

Including a polyA signal would yield higher levels of transgene expression (increased mRNA stability, and increased translational efficiency), and will promote proper termination of your mRNA. Without a termination signal, you may unintentionally generate a transcript with an extensive 3' UTR and/or interfere with downstream genes due to read-through transcription. I would recommend designing your construct to include a polyA signal.

Thank you very much!