I received a lentiviral transfer plasmid from a colleague. The plasmid looks like:

LTR-Psi-RRE-cPPT-[EcoRI]-PromoterA-Gene-2A-mCherry-[NotI]-WPRE-LTR

I digested this plasmid with EcoRI/NotI and cloned in my expression cassette, so that the resulting plasmid looks like:

LTR-Psi-RRE-cPPT-[EcoRI]-PromoterB-Gene-GFP-IRES-mCherry-[NotI]-WPRE-LTR

My plasmid differs from the original in a few ways:

1) different promoter

2) different gene of interest

3) IRES rather than 2A peptide between the two coding sequences

Otherwise, the backbones are identical and the expression cassettes are roughly the same size. I transfected both of these plasmids into 293T cells for packaging, and I see expression of the appropriate fluorescent proteins in the packaging cells. When I harvest the supernatant, precipitate the virus, and transduce a different cell line, I see strong expression of the original construct from my colleague, but not of my construct.

I have sequenced the entire plasmid; everything is in frame (as also evidenced by the successful expression of GFP/mCherry in transfected 293 cells) and all the viral packaging elements are intact.

My initial thought was that "PromoterB" must not be active in my transduced cells, but it apparently works in 293 cells, so I then used my viral supernatants to transduce 293 cells. Once again, I see expression of the original construct but not mine. I've also given my construct to a different colleague to attempt packaging and we see the same result. All of this suggests that the virus I produced does not actually contain my expression construct. My question is: what could cause this to happen? All the evidence suggests that something in my expression cassette (PromoterB-Gene-GFP-IRES-mCherry) is inhibiting viral packaging.

I am planning a few tests to determine what is going on:

1) RT-qPCR on the virus to determine if it actually contains viral genome/my cassette

2) PCR from Genomic DNA of transduced cells to determine if the cassette is integrating.

3) Nucleofect the plasmids into the other cell line to test whether the construct can be expressed in that cell line when delivered by a different method - this will tell us if the problem is the construct itself or the virus.

Any other suggestions for troubleshooting or insight into why my construct is not being packaged would be greatly appreciated.