We want to isolate the plasma membrane of cultured adipocytes for subsequent Western blotting, but we are finding some variety in the amount of cells necessary to achieve a sufficient sample amount (from just one T-75 flask to even as many as 50-100 million cells).
What would be the minimum area or number of cells required?
Also, has anybody experience with membrane protein extraction commercial kits? Are they more reliable or recommendable than the traditional sucrose gradient-ultracentrifuge method?
The key issue here is: adipocytes. I don't know what cells you use, but from my old experience with differentiated 3T3-L1 cells the critical step always was the removal of the fat. To make sure that you get nice plasma membrane fractions, do the following:
1. Break the cells using a syringe and a needle:
- scrape the cells in @1-2 ml buffer/T75 (I used PBS, but anything cole enough wil do)
- load them into a 1ml or 3 ml syringe
- pump them through a 24 gauge syringe needle
- repeat
2. Collect the stuff and place in some centrifuge tube (anything, from a 15ml table top to a 2ml eppendorf will do, depending on the volume).
3. Soft spin (3000 RPM for about 5 minutes).
4. Remove the BOTTOM layer (the fat floats!). You can substitute the soft spin by a much harder one (40000 in an ultracetrifuge) and then remove the top fat layer. If you trust your hands, just suck out the top fatty layer using vacuum. Whatever you do, the idea is to remove the fat, because it interferes with the plasma membrane extraction and whatever subsequent assay you need to perform.
5. Resuspend the (presumably) fat-free layer in fresh buffer.
6. Another soft spin (a little harder to remove nuclei; 5000 RPM is usually enough).
7. One final hard spin. Use a sucrose gradient if you want to get fancy. If you decide to use a sucrose gradient, add sufficient sucrose to your sample to make is 0.25 M. Then layer a centrifuge tube with 1.3 M sucose at the bottom, another layer of 1 M sucrose on top of this, and a final layer of 0.6 M sucrose on top. Put your sample on top of this and spin in one of those old-fashioned swinging bucket rotors. 25 K RPM in an old SW28 is what we used to do. Your PM fraction will be at the interface between the 1M and 0.6M sucrose layers. Our yield doing this was always of the order of 1mg membrane protein/10cm dish.
After this step, these plasma membranes are just like any others.
Most membrane extraction methods require detergent. In lipid-laden mature adipocytes, much of your detergent may be wasted in trying to solubilize triglycerides stored in these cells. Your protocol might work better if you try a hypotonic lysis step first to break open your cells, releasing the lipid droplets, and recover the membrane fraction by centrifugation. After this step, you can try to extract membrane proteins.
Actually the detergent step is just the last one, i.e. to solubilise the proteins out of the membrane lipid. It does not come into consideration when you are first trying to isolate the membranes. But hypotonic lysis and then centrifugation is the simple answer.
Do you really want purely plasma membrane , or is it sufficient to make a crude membrane prep, i.e. collect all membranes (including organelles like mitochndria)?. If a crude prep is sufficient, then follow the above suggestion. Just lyse the cells in enough buffer to dilute the cytoplasmic proteins and inhibit proteases etc, and spin down the membranes as a pellet.
you would just need enough cells to load on a blot. Depending on the detection levels of your protein , this is probably about 5,000 to 20,000 cells. Lyse them in a small volume, e.g. 50-100 microlitres , spin down at 16000-20000g for 30 min in an eppie, and then wash the resulting membrane pellet in the same buffer, spin again , and resuspend the pellet in SDS loading buffer.
Hi,
Isolation of plasma membrane proteins is a tricky thing. Biologically these proteins are so dynamic that they shift from plasma membrane to other endomembranes of the cell. So it is important to consider this cellular mechanism while quantifying the membrane protein (either it is plasma membrane or ER membrane or nuclear membrane). For isolation of total membrane proteins (other than mitochondrial and nuclear), Nature protocol by COX et al 2006 works very well and we have used it in our experiments for membrane protein profiling of brain tumors. commercial kits help to isolate only cell surface proteins or a fraction of plasma membrane proteins.
Another simple approach is to surface biotinylate and then capture with streptavidin or neutravidin. This could also be useful for tracking where your membrane fraction ends up during your fractionation, and would allow you to assess whether endocytic events are altering the distribution of your protein.
The key issue here is: adipocytes. I don't know what cells you use, but from my old experience with differentiated 3T3-L1 cells the critical step always was the removal of the fat. To make sure that you get nice plasma membrane fractions, do the following:
1. Break the cells using a syringe and a needle:
- scrape the cells in @1-2 ml buffer/T75 (I used PBS, but anything cole enough wil do)
- load them into a 1ml or 3 ml syringe
- pump them through a 24 gauge syringe needle
- repeat
2. Collect the stuff and place in some centrifuge tube (anything, from a 15ml table top to a 2ml eppendorf will do, depending on the volume).
3. Soft spin (3000 RPM for about 5 minutes).
4. Remove the BOTTOM layer (the fat floats!). You can substitute the soft spin by a much harder one (40000 in an ultracetrifuge) and then remove the top fat layer. If you trust your hands, just suck out the top fatty layer using vacuum. Whatever you do, the idea is to remove the fat, because it interferes with the plasma membrane extraction and whatever subsequent assay you need to perform.
5. Resuspend the (presumably) fat-free layer in fresh buffer.
6. Another soft spin (a little harder to remove nuclei; 5000 RPM is usually enough).
7. One final hard spin. Use a sucrose gradient if you want to get fancy. If you decide to use a sucrose gradient, add sufficient sucrose to your sample to make is 0.25 M. Then layer a centrifuge tube with 1.3 M sucose at the bottom, another layer of 1 M sucrose on top of this, and a final layer of 0.6 M sucrose on top. Put your sample on top of this and spin in one of those old-fashioned swinging bucket rotors. 25 K RPM in an old SW28 is what we used to do. Your PM fraction will be at the interface between the 1M and 0.6M sucrose layers. Our yield doing this was always of the order of 1mg membrane protein/10cm dish.
After this step, these plasma membranes are just like any others.
You don't need high speed ultracentrifugation, this is a misconception. If you keep the volumes small, you can do everything in 1.5 ml eppies in a normal microcentrifuge at 16000-20000g. See Abas & Luschnig (2010) Anal. Biochem 401.
Classical microsomal preps skew the results to the lighter membrane fractions, because the heavier domains are thrown away in the preclearance.. Plasma membranes will have light and heavy domains. You will throw away the heavier domains using classical protocols and collect only the lighter domains. This is a potentially bad bias that you should be aware of. See my explanation in Abas & Luschnig (2010) Anal. Biochem 401.
Hi Fatima,
t-75 flask is just fine.. Just make sure its confluent and use the pierce subcellular fractionation kit from pierce. I worked wonders for me. All the best
Membrane Protein extraction reagent from Pierce reliable and easy to work with
Just follow the protocol
I have used 2- 100 mm dishes seeded with 3.2 million cells each, so 6.4 million total. I can send you my protocol using a sucrose gradient and ultracentrifugation if you like... Based on using Caveolin as a plasma membrane marker, the fractionations seemed to work rather well.
It depend on the expression/abundance of membrane protein you are interested to work with. I suggest you to work with 2 - T 75 or 150 mm dishes will work fine. Please forward the protocol. Please let me know what membranes are you isolating plasma membrane or lipid rafts/DRMs from these cells. Also let me know your protocol.
Thanks everyone for all the replies.
Rajendra, our aim is to analyze GLUTs, and abundance will vary between the different types. Our protocol consists in collecting the cells in HES buffer, spining in a centrifuge (19,000 g), and then adding a sucrose gradient for the last ultracentrifuge steps (100,000 and 40,000 g).
Kristina, if your protocol is different, I would appreciate being able to take a look and compare with ours, thank you.
Lindy, I will have at look at yours as well. Many thanks!
Hi Fatima, Please find the protocol attached and let me know if you have any questions.
my protocol for preparing membranes from very small amounts of material using a benchtop centrifuge (no ultracentrifugation required ) is at
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3685806
lindy
Is your purpose looking at the surface membrane or internal membranes? For enriching surface membranes, after mechanical scraping to capture cells (do not trypsinize or other chemical release), you can gently sediment by mild tabletop centrifugation, decant, lyse very cold with distilled water (at least 4 volumes-- more may be necessary, solution should look clear) , then filter either through fine mesh or large-pore membrane filters; the filtrate is discarded (or used for membrane comparison, but remove fat as suggested by Guillermo), and the retentate held on the mesh should be a rich prep of plasma membrane and possible endoplasmic reticulum.
Although there is more loss, fine mesh works better than polycarbonate filters, which clog easily. The last step is a wash (at this point saline is okay, you already burst open the cells; may work better warm if released fat is a problem.). Now remove the mesh or filter and immerse in your membrane dissociation buffer (running PAGE of this prep versus the filtrate may prove to you that you succeeded-- look for an enriched surface protein).
If you want to look at nuclear or mitochondrial membranes only, do sucrose gradients first, then extract the membrane. You really need to separate the organelles first for meaningful information. My experience with looking at ultrastructure of isolated mitochondria was that calcium levels during isolation were critical for maintaining integrity. But... sucrose gradients with that small amount of material isn't a happy thought.
Some classical adipocyte membrane purification routines: (Shanahan, M. F., and Czech, M. P. (1977) J. Biol. Chem. 252, 6554-6561] and [Seals et al (1979) J Biol Chem 254,6991-6996.].
dear Eli
i have not tested the protocol with animal cells, but the underlying theory applies to any system. Since animal cells do not have a cell wall like plants, then the lysis step for cultured animal cells is usually much gentler (hypoosmosis bursting) compared to the harsh grinding needed to break open plant cells., So for animal cells, the organelles will be mostly intact and thus larger, and I would predict these would sediment even faster in any preclearance steps, i.e. there would be a tendency to discard more membranes in these early stages. Also, a big difference between plants and animals is that the lysosomes are one of the heavier membranes in animals, whereas plant vacuoles are one of the lightest, so one would probably lose more lysosomes than vacuoles in preclearance. My approach would be - just run all those traditional preclearance pellets along with your final microsomal fraction, and see what you would have thrown away as preclearance.
- Badges
- Science method