Both 293T and Fibroblasts should be able to express your protein (assuming all of the cellular machinery is in place) and all that requires is getting higher transfection efficiency. For 293T (https://altogen.com/product/293t17-transfection-reagent-emb-kidney-crl-11268/) and for Fibroblasts (https://altogen.com/product/fibroblast-transfection-reagent-fibroblast-cells/) lipofection can provide reliable results with little off-target effects as one would expect from transduction.

It sounds like your problem is with the Kozak sequence / start codon.  If you can detect the protein in principle ("wt fibroblasts indicates expression" -- I assume this means they express it naturally, not transfected?), and there is message there, but not protein, then that means the cell can't figure out what you want it to do with the message.  

I think the most likely explanation at this point would be a design error, or an error during cloning, that ruined the ribosome-binding site on your message.  Can you post the gene sequence around your start codon? 

If it's indeed a cell-type specific issue, that would be interesting, and I'd love to learn about it. 

Is your protein tagged? I would also make sure your antibody can detect the protein if it's tagged. I've had that issue before and could only pick up the protein in 293T's with a specific antibody.

Thanks for your answer. I don't know if it was entirely clear, but I've been working with 293T-cells. Since I couldn't detect my protein in wt or transfected cells, I've suspected problems with my antibody. However, when I ran normal fibroblast cells on Western, I suddenly got a clear band.

But even if the 293-cells have no detectable constitutional expression of the protein, it should be possible to overexpress it, right?

I fear you might be right about errors in the vector. I have sequenced these previously and didn't notice anything strange, but I'm pretty new at cloning. I will post the sequence as soon as I have access to it.

Elyse - my protein is not tagged. If I have to redesign my plasmid I'll bring that up - my vector is very simple.

I agree with the other answers that there is probably a problem with the vector.  Fibroblasts can be tricky to transfect.  We routinely use Amaxa on fibroblast-like synoviocytes and can get reasonable expression levels (although not as good as 293 cells of course). 

I am not transfecting fibroblasts! I am considering trying to transfect them since they seem to have a constitutional expression of my protein - and 293T cells do not. ;)

Jonathan, have you used Lipofectamine with fibroblasts? Do you think it's plausible? :)

It depends on the type of fibroblast.  You definitely would not get expression in every cell, but if you are just looking to see if you are getting expression of your transgene, even expression in 10-20% of the cells might be enough.  

I am a bit confused, if fibroblasts make the protein that you are expressing from the plasmid, how would you tell your expressed protein from the endogenous protein?  I have never heard of a protein that would express fine in fibroblasts, but not in 293 cells.  Is your protein large and have a complex structure?  Maybe the chaperone proteins are different between 293s and fibroblasts?  Does the protein need a co-factor to stabilize?

I'm sorry to be confusing.

I have been working with 293T-cells. I am trying to overexpress a ~500AA one-exon relatively simple protein. I can not get bands in the western for transfected nor untransfected 293T-cells.

However, when I run fibroblasts, I get a clear band. That's what made me think that trying to transfect fibroblasts would be a good idea. I have not transfected any fibroblasts - yet. :)

Also, it's quite easy to tell overexpression from endogenous expression because of the intensity of the bands in Western. I am doing this with an NDRG1 overexpression construct as a positive control for the transfection, and the difference in signal strength is huge.

If you want to use lipofectamine, I would be hesitant to assume that you will be able to tell if your protein is expressed just by looking at the intensity of the western blot.  Fibroblasts do not express as much protein as 293s, and with lipofectamine most of your cells will not be expressing at all.  Even if your transfected cells are expressing 10 fold higher than endogenous levels, if you are only expressing in 10% of the cells, this will only look like a 2 fold increased intensity on WB.  I for one would not trust this as an indication that my plasmid worked, especially as transfections are inherently toxic to cells so you could easily have 2 fold differences in protein levels just from variations due to the transfection toxicity. 

Long story short, if the protein isn't expressing in 293's, it is a long shot that it will suddenly start expressing in fibroblasts.  If you want me to take a look at your sequence and see if there is anything obviously wrong, shoot me an email ([email protected]).

Good luck!

Can You Describe your Vector?? because the problem is the vector. I produce a lot of proteins in HEK 293 T and I don´t have any problem.

I apologize for the late answer, and I appreciate all of your replies.

Jonathan: I know that the wild type expression of my protein in 293T cells is low. Do you still think I wouldn't be able to see overexpression in a western blot if my transfection was successful?

I am not very confident in my antibodies and have suspected that they weren't working, which is why I was very surprised to see a band in the fibroblasts.

I don't have my qPCR data at hand at the moment, but the overexpression in the transfected 293T-cells were many thousand fold compared to cells transfected with psirpg (GFP-carrying plasmid as positive control for transfection).

From the psirpg-transfected cells I have estimated a 60-70% transfection efficiency.

Should I assume that I have overexpression of transcript - and that my protein is either not overexpressed OR my antibodies are simply faulty?

A map of the vector is attached below. It has been constructed in our lab (before my time!) and has been used successfully many times (with different inserts, of course). The gene is ~1500bp long (~500AA protein) and is inserted by Gateway Cloning.

This is the region around the start site. 95 nucleotides of UTR precedes the start codon.

CCGGACTCTAGAGGATCCCTACCGGTGATATCCTCGAGCCCATCAACAAGTTTGTACAAAAAAGCAGGCTCAACCAGCATATAGAAGTGACATATTGTTATATTTTCACCATAG_95bpUTR_ATG

I am using a NDRG1-construct (similar plasmid backbone as above) as a positive control for overexpression, and it works very well. The overexpression is strongly evident on Western.

Nadia, we still can't tell if there's a functional Kozak (translation start site) there... We need to see a couple of BP *around* your intended "ATG" start codon.  Before your ATG, there's a "95bpUTR" that doesn't reveal the sequence, and we can't see behind your ATG at all.  You can use online software too, which will reliably tell you if the start site is good.  It should read something like NCCNCCATGG (but many variations are possible)

Hi Nadia,

I think that you should be able to see overexpression of your protein in 293 cells by western blot (if it is being expressed). 

Thank you all for your input. I am designing a new Gateway clone now, with an edited Kozak sequence.

Both 293T and Fibroblasts should be able to express your protein (assuming all of the cellular machinery is in place) and all that requires is getting higher transfection efficiency. For 293T (https://altogen.com/product/293t17-transfection-reagent-emb-kidney-crl-11268/) and for Fibroblasts (https://altogen.com/product/fibroblast-transfection-reagent-fibroblast-cells/) lipofection can provide reliable results with little off-target effects as one would expect from transduction.