Hi everyone,

I am working on brain inflammation and I try to separate the mouse brain vessels from the parenchyma after the removal of the olfactory bulbs and the meninges.

Currently I proceed to a manual homogenization  of one single perfused brain in 3 mL PBS/FBS-5% and then I centrifuge at 1100g/2650rpm -10 min - 4°C x 3 times.

The supernatant, which is supposed to contain the parenchymal fraction, is first discarded in another tube = P (parenchyma) and each new supernatant is added to the tube P then centrifuged once at the same speed.

The pellet is resuspended in 1 mL of PBS/FBS-5% + 2 mL are used to pour the pellet over a 100um mesh filter which is flipped over to collect  the filter-bound vessels in a new tube = V (vessels) using with 3 mL of PBS/FBS-5%.

Because some argue that the parenchymal fraction is still in the pellet obtained after the first centrifugation steps, therefore filtrate may contain some parenchymal components/cells (neurons, astrocytes, microglia), so I would call this fraction X. 

Unfortunately, when I use western blot or dot blot to assess the relative purity of my fractions, I can't get a consistent positive signal for CD31/PECAM-1 (endothelial cells) specific to the fraction V and a signal for GFAP (astrocytes) specific to fraction P. Both CD31 and GFAP signals obtained in fraction X are inconsistent, as well as for the homogenate.

I know some researchers use either Percoll, Ficoll, mostly Dextran and rarely sucrose to make  centrifugation gradient in order to separate fractions V and P, but here again the literature is really inconsistent on this point. As it is inconsistent too about where exactly the brain parenchyma remains .

Soon I will try to filtrate right after I homogenized one mouse brain (or after one round of centrifugation at 400g max - 10 min - 4°C to get rid of some debris), discard the supernatant, resuspend and filtrate the pellet over a 100um mesh to consider then that I will get some parenchymal components in filtrate and some vessels bound to the filter. To improve the purity of the vascular fraction I think of doing a second filtration step through a 100um mesh. And as for the parenchymal fraction I wonder what would be the best way to improve its purity (eg. filtration or centrifugation by density gradient?).

I believe it is easier to reduce the quantity of vessels from the bulk homogenate through filration steps than to separate the parenchymal fraction from the vascular fraction, especially if I intend to keep intact the vessel - but I might be all wrong...

I have been troubleshooting this technique for so long, I really would appreciate any kind of help on the topic.

Sincerely, thanks!