293T cells are barely adherent.  Any type of mechanical agitation, washing, etc. can cause cells to come off the dish.  If you were transfecting other types of cells, such as fibroblasts, they wouldn't come off, but 293T are still better for a lot of experiments because they transfect so efficiently.  Though you should certainly keep being gentle, your experiments still may work out fine.  Some cells will reattach during the transfection.  Also, you can still make lysates from the floating cells.  When I make western blot or IP samples, I detach all the cells into the media by pipetting, then spin down the cell suspension in a microfuge tube (thus collecting the ones that were floating to begin with).  The pellet can then be washed in PBS and a lysate made.  The only caution is that if you are going to do a western blot, you will need to be careful that equal cell equivalents are loaded in each lane.  As always you will need a loading control such as actin, but you may need to check the protein concentration of your lysates first so you can load the same micrograms/lane (which people often do anyways).

If you are going to do microscopy and the detachment is a huge problem, but you are more concerned with expression per cell and not per well, then I would try making the cells less confluent.  They will be less prone to come off in big sheets in this case.  Good luck!

293T cells are barely adherent.  Any type of mechanical agitation, washing, etc. can cause cells to come off the dish.  If you were transfecting other types of cells, such as fibroblasts, they wouldn't come off, but 293T are still better for a lot of experiments because they transfect so efficiently.  Though you should certainly keep being gentle, your experiments still may work out fine.  Some cells will reattach during the transfection.  Also, you can still make lysates from the floating cells.  When I make western blot or IP samples, I detach all the cells into the media by pipetting, then spin down the cell suspension in a microfuge tube (thus collecting the ones that were floating to begin with).  The pellet can then be washed in PBS and a lysate made.  The only caution is that if you are going to do a western blot, you will need to be careful that equal cell equivalents are loaded in each lane.  As always you will need a loading control such as actin, but you may need to check the protein concentration of your lysates first so you can load the same micrograms/lane (which people often do anyways).

If you are going to do microscopy and the detachment is a huge problem, but you are more concerned with expression per cell and not per well, then I would try making the cells less confluent.  They will be less prone to come off in big sheets in this case.  Good luck!

Thank you, Michael. The thing is I need to do microscopy. I am worried that the detached cells will affect the healthy or attached cells after 24-48hr incubation. But I will keep updating this experiment. Hopefully, the attached cells still work. 

Coat your wells with Poly-L-Lysine prior to plating the HEKs. That usually helps.

I had another thought after looking at your question again.  In the original protocol for lipofectamine, the plasmid and reagent were combined in a relatively small volume.  This was then added dropwise to a larger volume of media already on top of the cells; this helps disperse the reagent without the need for aggressive mixing.  However, if you are adding the entire 1 ml dropwise to the center of a well lacking media, it definitely would cause the cells to come of as a big "hole" in the center.  I always aspirate from the edge of a well and pipet fresh media down the side.  If the transfection cocktail is already mixed, you might try pipetting down the side instead of the dropwise approach.  Coating with poly-lysine as Lars suggested may also help and, since you are doing microscopy,  you might consider plating on glass chamber slides or coverslips.

Hi, Michael, thanks a lot.The good news is that the transfection worked.  I did add 1mL mixture to the cells without media. I think you are right, this would detach the cells. I will be very careful with this next time.

Hi Lars, thanks a lot. The plate says tissue culture treated. Does it mean that it has been pretreated with Poly-L-Lysine?

No, it’s just the normal TC treatment. You can buy plates that are treated with Poly-L-Lysin, or you can coat them yourself. Just ask around if anybody in your neighbor labs has some. It’s often sold as 0.01% solution. Most labs that produce virus with HEKs will have it. If not, ask for Gelatin, that should also work for you.

You can also try nucleofection (like the amaxa) or reverse transfection if nobody around have the nucleofector. In reverse transfection, you add the cells on the plate containing your tranfection reagent and vector mix. You won't have that detachment problem then. Several companies sell their own product, check it out. Here is the one from promega

https://www.promega.com/resources/pubhub/reverse-transfection-using-fugene-6-and-fugene-hd/

As far as I remember you can also use the invitrogen product to contransfect siRNA with vector as you can see at the end of this link.

https://www.thermofisher.com/us/en/home/references/protocols/cell-culture/transfection-protocol/rnaimax-reverse-transfections-lipofectamine.html

Cotransfecting DNA and RNA using Lipofectamine® RNAiMAX

For cotransfections of plasmid DNA and Stealth™ RNAi or siRNA into mammalian cells, we recommend using Lipofectamine® 2000 (Catalog no. 11668-027), which is superior for plasmid transfections. If you want to use Lipofectamine® RNAiMAX for your cotransfections, perform a reverse transfection  with the following modifications:

1: Add 20 ng (for 24-well format) of plasmid DNA to the diluted RNAi duplex.

2: Add cells such that they will be 80-100% confluent 24 hours after plating

Is there a specific reason why are you transfecting the cells at such high confluency?  

So you know, I always transfect around the 60-70% confluency when the cells are in exponential growth.  I found that by transfecting at lower confluency, you will obtain higher efficiency and the HEK293T's will be much less likely to lift away from the plate.

Also, you should be getting about 90% efficiency with HEK293T's, if not then try XFect.

http://www.clontech.com/US/Products/Transfection_and_Cell_Culture/Plasmid_Transfection/Xfect_Single_Shots?sitex=10020:22372:US

Hi Sonia, thank you. I just follow the suggested 80-90% confluency  by invitrogen. I find people transfect at a lower confluency if they use calcium phosphate method. So you use XFect?

Hi Lars, I see. Thanks a lot. I will try that.

Hi Daniel, thanks a lot. I will try the reverse transfection. It seems to be interesting.

You may also use collagen for plates coating, it works as well as gelatin. But as Michael proposed, i think, the point is the volume of the mixture added. I usually have about 10% detached Heks with every transfection (i`m doing transfections with Metafectene Pro by Biontex), but 10% is not such a problem for me. Considering the seeding density, Heks are usually detaching more easily when confluent, so transfecting at lower dencity (50-70%) could be preferable and also will give you higher transfection efficiency. Hope that helps.