UKC

I have two human proteins A and B that I want to test the interaction between (by transient co-transfection in 293 cells):

Protein A I have as WT and multiple mutants that I want to test the interactions of with Protein B. I have these Protein A mutants tagged with HA, Flag and GFP tags.
GFP expresses, IPs and blots nicely. However, I've found that HA and Flag express well (as seen in lysates) but does not IP at all well, so the availability of these small tags appears to be low in the native state, whereas with the large GFP tag this isn't a problem

Protein B is tagged with mCherry - expresses, IPs and western blots nicely.

So the question is, can I combine mCherry tagged protein A and GFP tagged protein B? They are evolutionarily related, GFP from Aequorea victoria (Jellyfish) and mCherry a modification of RFP from Discosoma sp, but by eyeball, conservation is relatively low. So, the question is, do you think a GFP Ab will pull down mCherry or an RFP Ab pull down GFP? I'm hoping not!!

Comparison of GFP and RFP Sequences:

GFP MSKGEELFTGV VPILVELDGD VNGHKFSVSG EGEGDATYGK LTLKFICTTG -KLPVPWPT
RFP MRSSKNVIKEF MRFKVRMEGT VNGHEFEIEG EGEGRPYEGH NTVKLKVTKG GPLPFAWDI


GFP LVTTFSYGVQ CFSRYPDHMK QHDFFKSAMP EGYVQERTIF FKDDGNYKTR AEVKFEG
RFP LSPQFQYGSK VYVKHPADIP --DYKKLSFP EGFKWERVMN FEDGGVVTVT QDSSLQD


GFP DTLVNRIELK GIDFKEDGNI LGHK-LEYNY NSHNVYIMAD KQKNGIKVNF KIRHNI
RFP GCFIYKVKFI GVNFPSDGPV MQKKTMGWEA STERLYPRDG VLKGEIHKAL KLK---


GFP EDGSVQLADH YQQNTPIGDG PVLLPDNHYL STQSALSKDP NEKRDHMVLL EFVTAA
RFP -DGGHYLVEF --KSIYMAKK PVQLPGYYYV DSKLDITS-- -HNEDYTIVE QYERTE


GFP GITHGMDELYK
RFP GRHH----LFL

Thanks for any help!

In reply to Jimbo W:

I don't know about RFP, but cherry seems pretty close to GFP, at least from my sequences? I use ant-GFP for YPET. Antibodies recognize structures rather than sequences as well, I think you are out of luck. Or, your controls for cross binding will have to be excellent.

>mCherry
MVSKGEEDNMAIIKEFMRFK
VHMEGSVNGHEFEIEGEGEG
RPYEGTQTAKLKVTKGGPLP
FAWDILSPQFMYGSKAYVKH
PADIPDYLKLSFPEGFKWER
VMNFEDGGVVTVTQDSSLQD
GEFIYKVKLRGTNFPSDGPV
MQKKTMGWEASSERMYPEDG
ALKGEIKQRLKLKDGGHYDA
EVKTTYKAKKPVQLPGAYNV
NIKLDITSHNEDYTIVEQYE
RAEGRHSTGGMDELYK

In reply to wilding:


Using today's ring settings that translates to co-ordinates for a convoy in the North Atlantic.

In reply to Jimbo W:

I think that GFP and RFP can be used for the same thing as can YFP but I'm not sure about mCherry. I gave this article a quick skim and it might answer your questions

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2875877/

In reply to Jimbo W: As Wilding said, it's the epitope you need to be worried about. Think you will probably need a strong blocking solution.

You might want to try and test pulldowns at different temperatures for different lengths of time to see how good your secondary is and how much noise you have.

Interactions between your two proteins of interest could be tricky, one of the problems I came across when doing confocal microscopy was that GFP was practically engulfing my protein of interest (it was very small and GFP very large) so it couldn't localise to where it should have (the nuclear localisation signal was masked).

In reply to Jimbo W:

... And another consonant please Rachel.

In reply to Jimbo W:

Are you using monoclonals or polyclonal sera? (I'm sure you know this, but although polyclonals are messier they tend to be better for IP) If monoclonals you must know what epitopes they see, don't you? Where did you get the antibodies from?

If you can't find out, I guess it might save you a lot of time to try IPing some native GFP and RFP to check how much cross-reactivity there is, but I can't believe this information isn't already available.

BTW, you don't really mean Protein A do you? That could complicate things!

In reply to Dave Garnett:

Thanks Dave. Yes I should have said protein x and y! We make our own sheep Gfp polyclonal which is great in western blot and had been for ip as well. However for the reasons you suggest, I've been looking for alternatives. I also have rabbit and chicken polyclonals from abcam, which I use for ChIP and IF respectively, but I don't have a GFP monoclonal. However, the RFP ab I'm using is from chromatek, and it is a rat monoclonal:

http://chromotek.com/en/reagents/conventional-antibodies/rfp-antibody-5f8/

Which I don't know the epitope, but which they are explicit about being specific for RFP and not GFP.

We have some of this companies GFP trap beads in the unit, that people use for mass spec, but they're expensive, though for IP they claim they should be very specific. I also wonder whether I could pre incubate my Gfp polyclonal coupled to beads in a lysate from cells expressing RFP alone.

I'm setting up an expt. today to test this out. However, I've discovered another potential issue, which is that we use the eGFP, and this does not have the non dimerising mutations. While the mCherry should be monomeric, I'm now worried about the potential that these might dimerise together, Gfp and RFP. I can't find anything specific on the literature about this, but I think it is unlikely. However for my understanding, I wonder if you can tell me whether something that dimerises inefficiently like Gfp, uM kd I think, could this be rendered problematic by locally high concentration such as when using multiple Gfp derivatives on proteins known to complex, or worse could Gfp dimerisation enhance an interaction of Gfp on a proteon known to interact with itself. Does that make sense? If there is a potential for Gfp and mCherry to dimerise, can I get round this in Ips by using a large volume for my dynabead pull downs?

Sorry for the length! Need to make sure I get this right! Thanks for your help.. ..I've presumed before that you have worked in biotech, so appreciate the advice very much.

In reply to Jimbo W:

It's been a long time since I worked in a lab and although I did some recombinant expression work, I never myself made anything tagged with GFP.

That said, I don't recall GFP dimerisation being a big issue (as it was in some immunoglobulin constructs with spare disulphides, for instance). I suspect you're either going to have to suck it and see, or be supercautious and use a non-dimerising mutant (I've googled and there is one: A206K). There's more to dimerisation than just local concentration, and I'd have thought steric considerations would be more important. Unless the GFP tags can align correctly they won't dimerise (but also, of course, if you are GFP-tagging proteins you want to dimerise you need to be sure a relatively large tag like this won't get in the way). If it's largely hydrophobic dimerisaion you might be able to add a little bit of a mild detergent like nOG or something.

Anyway, I'm hypothesising, you'd be far better talking to someone in a lab with direct experience of this. You could try asking the technical people at Dynal about the pulldowns. As it happens I know the guys in the lab in Oslo well and I could put you in touch.


Good luck - I still miss the lab sometimes. Although then I think back to the nights spent in the coldroom with only the glow from the P-32 to keep me warm and I don't miss it quite so much!