Hi Jing, I use PEI (1:1 ratio DNA to PEI).  For ~1-2 million HEK293s in a T-75 flask I use a range from 6 to 12ug of cDNA.  12ug is the upper limit of how much cDNA these cells can tolerate and they're not entirely happy after trypsinization and reseeding in a transwell.  

The assay I'm using is a transwell, specifically Corning's Fluroblok inserts.  I transfect cells with 6ug of my receptor, 6ug of GFP and seed them at the top.  After anywhere from 5-24H I image the green cells that pass through the membrane (the green cells in the top chamber are blocked from detection).

Also I realized I should note that I'm at the very beginning stages of assay development so transfection conditions and time points are a work in progress - but I was surprised to see no migration to a "standard positive control", especially when these same cells exhibited the ability to migrate towards serum free media and media with a chemokine.

Thank you Jing! Yes, I am doing control as my empty vector - so 6ug GFP, and 6ug of receptor or empty vector.  I'm working on a way to quantify the conditions so I'm not basing the results on my opinion of the wells/images.  Although in the case of the no migration to serum media - it's pretty apparent.  I'll look into supplements, I think spiking in EGF sounds like a good idea and it's cleaner.

Do you have a recommendation for how many images to take per well?  I tend to see more migration at the edges of the insert than in the center and haven't yet figured out a way to fairly represent the # of cells that migrated.  Alternatively I can just measure fluorescence with a plate reader.... 

I'm also looking into making a bicistronic vector with my receptor followed by an IRES GFP so that I only have to transfect 1 plasmid, and which hopefully mean less DNA.    

Bina

when using Transwell make sure to use the correct pore size. I do not know about your system specifications. We used the Costar system with either 5 or 8 micrometer pores. With pore sizes too big the cells just fall through and no chemoattractant effect can be detected; pore sizes too small cells do not make it even with strong chemottractants. We witnessed variations between insert batches. I do not know how you collect your migrated cells but very few cells will be detected in the bottom well. The cells pass the membrane pores and adhere on the filter on the lower side of the filter. We use trypsin to collect the adhered cells. It is definitely important that your 0% serum you do not have any cells collected whereas in the presence of chemoattractant (serum) you detect some cells.

Hi Thierry, I'm using an 8 micron pore size, people have published using 5 and 8 for HEK293s.  Since I can tag them with GFP and take images of the bottom of the well, I don't attempt to collect them.  This explains how the fluoroblok inserts work: http://www.ebiotrade.com/buyf/productsf/BD%20Biocoat/BD%20Falcon%20HTS%20FluoroBlok%2096-Multiwell%20Insert%20System.htm 

Do you transfect your cells prior to migration assay?  

Hi

I prefer to use cells stably expressing your GPCR or a cell line having endogenous receptor levels high enough to make them move. We witnessed that monocytic cell line THP1 do not do internalize well the CCR2 receptor after nucleofection with Amaxa kits. Be careful with the use of trypsin because some chemokine receptors are cleared off by trypsin.  What does your no serum medium contain? BSA?

While serum is a common positive control, some cells do not migrate toward it. For example, we find that pancreatic carcinomas migrate less toward serum or its major component, LPA, when compared to BSA. 10% serum can have very large amounts of LPA. At high levels, LPA and serum downregulate the main migratory LPA receptor, LPA1. At these levels of LPA, LPA2 receptor remains, but it does not always stimulate migration. I would recommend 100nM LPA as a good chemoattractant. But always remember to coat the transwells with matrix (e.g. fibronectin, collagen, laminin). Serum is handy because it supplies matrix (fibronectin and vitronectin) along with the chemoattractant. Personally, I didn't find the fluoroblock wells to work very well. My lab continues to count them manually.

Both your points have been incredibly helpful. I was inclined to think that the cells were overstimulated by 10% serum and actually performed a serial dilution – it seemed that 2% serum got some cells to migrate, but still the result is not robust (and needs to be repeated). It could also mean I was diluting out a chemorepellent. I really like the suggestions of adding the chemoattractant at a known concentration into serum-free media and will certainly look into LPA.

Regarding the matrix for HEK293 cells, I’ll need to look into the literature for what’s appropriate. But will this mean I don’t need to resuspend my cells in 1% FBS media? They seem very fragile after a ~20H transfection, followed by trypsinization, spin down and then resuspension in the final chemotaxis media (which is just 1% FBS DME). Should I coat the wells and give up the 1% serum?

Kathleen, thanks for your point about the fluoroblok wells – my problem with them is the migration seems uneven, I see more robust migration closer to the edge of the wells versus the center which makes it difficult to be fair when imaging.

Thank you for your advice!!

If you want good migration results, you should not recover your cells with serum after trypsinization. We use medium containing BSA (250ug/ml) to collect cells and then wash 3 times with that medium. If you stimulate the cells with serum upon harvest or resuspend your cells in serum, they will not migrate well to any chemoattractant because you will no longer have a good chemoattractant gradient. If you are interested, you can send me an e-mail at [email protected] and I'll send you our protocol.

I'm sending you an email right now!  Thank you!