The propensity to form smooth, round "aggregates" of GFPs varies greatly among the GFP variants--and it is even more severe with all DsRed derivatives we've used.  In our experience, the "original" EGFP coded by the Clontech vectors is the least prone to forming these structures.  This may be because of a slightly longer N- or C-terminal stretches of amino acids flanking the EGFP sequence, perhaps due to polylinkers in the vector.

We recently compared EGFP to "monster" GFP (also called hmGFP) which is a very bright GFP isolated from sea anemone (expressed in some Promega-derived RNAi vectors, also sold by Superarray).  This GFP variant was far more prone to forming the smooth, round structures that looked like lipid droplets.  These killed our cells if expressed longer than 3 days.  To solve this problem, we had to cut out the hmGFP sequence and replace it with Clontech-derived EGFP.  

Bottom line:  not all EGFPs are the same.  Which one is coded in the adenovirus vector?  Is its sequence truncated relative to the Clontech version of EGFP?  Is is a EGFP from another species besides jellyfish?

Mike

Michael J. Schell, Ph.D., CIV, USUHS
Assist. Professor
Dept. of Pharmacology
Uniformed Services University
4301 Jones Bridge Rd.
Bethesda, MD  20814-3220
tel:  (301) 295-3249
[log in to unmask]
>>> David Burk <[log in to unmask]> 11/20/08 4:24 PM >>>
1. Q: Hi David - are you certain those are really lipid droplets and not something like endocytic vesicles? 

   A: We are pretty convinced they are lipid.  They are counterstained with Bodipy 558/568 and we have localized several "lipid-associated proteins" with them - like TIP47 and ADRP.


2. Q: I have the same question, and, further, are you sure that the GFP isn't simply aggregated? If aggregates form they could expose hydrophobic surfaces that favorably bind to the droplets.

   A = Q:  How would I determine if this were the case?

3. Q:  Are the "droplets" smooth?  I have seen many different aggregations, some smooth, some look like particles, also what size, roughly are they?

   A:  Yep, smooth and round - like nice lipid droplets :)  I don't have the size handy at the moment but can get back to you on that.

Thanks for your input!  I'll be happy to answer more questions if it helps us answer our initial question - quoted below.

David

__________________

GFP Experts:

We have a group transfecting cells with adenovirus.  This virus contains an EGFP sequence to let you know that, yes, the cell was transfected.
We have noticed in some of the transfected cells that the GFP signal appears to be strongly associated with lipid droplets.  This is new to me as I wouldn't expect WT-GFP to strongly associate with any particular subcellular organelle and is confounding their work since they are primarily interested in - of course - lipid droplets.

Do any of you know of a reason we would see wild-type GFP with no targeting sequence or modifications labeling lipid droplets?  We are beginning some additional control experiments with a large titration range of virus to determine if the lipid labeling is related to vector concentration (~GFP expression level).

Thanks for your help!

David H. Burk