Protocol Online logo
Top : New Forum Archives (2009-): : Hematology

how to make Percoll solutions?! - (Oct/26/2010 )

hi all,

the percoll manual gives me the creeps really. i dont wanna study physics before being able to generate a percoll solution...

so is there anybody out there who could give me a "for dummies" procedure of how to generate this isotonic stock solution and then from that to get a 60% percoll solution?

many thanks!


PS. i work with mouse bone marrow.


I work with mouse bone marrow too! I've also been playing with Percoll and from studying the manual I've come up with the following:

You first need to make a 1.5M solution of NaCl. Then to 9 parts Percoll (from the bottle) you add 1 part 1.5M NaCl. This is to make Percoll isotonic, and from then on the manual refers to this solution as Stock Isotonic Percoll (SIP) as 100% Percoll (confusing, I know!).

To make further dilutions of Percoll, you dilute with 0.15M NaCl.

To get a 60% Percoll solution you want to have a final solution that is 60% SIP : 40% 0.15M NaCl. So to make 10mls of 60% Percoll, mix 6ml SIP with 4ml 0.15M NaCl.

Hope that helps!


We used hank's balanced salt solution, but same principle:

Mix 10 ml HBSS in 90 ml 100% Percoll (now 90%). This is the isotonic Percoll, like Piersgb said.
Dilute 60 ml isotonic Percoll with 40 ml 1X HBSS (plain).

The nomenclature in the manual confuses everyone at first, so I just gave these directions to the students to make their lives easier. The actual % percoll is around 54%.

-lab rat-

many thanks guys!!!
and sorry for the late answer... was busy with other stuff (the percoll thing is more of a sidekick project of mine).

what you say is reassuring, really. and i mean in two ways.
first, i dont seem to be the only one who feels a bit stupid when reading this horrible manual. :blink:
second, i allways did it as Piersgb described, but then sometimes the seperation didnt work, which means it wasnt due to the percoll but probably the sample itself. so i gotta work on handling the cells more gentle i guess. :o

anyways, i recently tried again and it worked out. btw, i ran ficoll in parallel and it worked as well but gave less sharp bands and less recovery than percoll.




Glad to hear it Bert. I thought of one more thing: it helps to make sure the cells are suspended in a solution of the same osmolarity as the Percoll. If the is lower in your suspension (e.g. PBS) and the cells come into contact with the higher Percoll during the layering, the cells will flatten and slide through the colloid.

-lab rat-

Hi all,


i am not sure if this topic is still active, but i also have a question.

I have almost established a culture of endothelial cells with some fibroblast contamination and i want to separate those two cell types.

Now i know that it can be done using Percoll, but the manual is a bit confusing and i get perticularly confused with the density specifications.

The endothelial cells should be separated, according to my reference, at 1.07 g/ml and the fibroblasts at 1.05-1.055.

Can anybody expain me how i can make those two specific densities?


Thank you all,




you can use this on-line calculator from ge lifesciences.


Hi mdfenko,


thanks a lot for your response.

I am aware of the on-line calculator, but i wil be performing this kind of separation for the first time.

So, if i understand correctly i have to make the solutions for specific densities. But what about the osmolarity, will the solution with my prefered density be ok for my cells? will they survive? besides that, i only want to collect just one type of cells, so i  guess if i make just one density percoll, then i will only get the cells that i want layered. But how does this look like in real time? I will be pouring the DMEM where i have my cells in on top of this Percoll solution, then i will spin it and then what happens?

Do you have any experience with Percoll?


I appreciate you responce a lot.


Thanks again.




sorry, i haven't used percoll myself.


here is the instruction booklet from ge lifesciences (pdf download)


osmolarity should be controlled by the buffer you use.