This fractionation protocol will allow researchers to isolate cytoplasmic, nuclear, mitochondrial, and membrane proteins from mammalian cells. The latter two subcellular fractions are further purified via isopycnic density gradient.
This protocol describes a method to obtain subcellular protein fractions from mammalian cells using a combination of detergents, mechanical lysis, and isopycnic density gradient centrifugation. The major advantage of this procedure is that it does not rely on the sole use of solubilizing detergents to obtain subcellular fractions. This makes it possible to separate the plasma membrane from other membrane-bound organelles of the cell. This procedure will facilitate the determination of protein localization in cells with a reproducible, scalable, and selective method. This method has been successfully used to isolate cytosolic, nuclear, mitochondrial, and plasma membrane proteins from the human monocyte cell line, U937. Although optimized for this cell line, this procedure may serve as a suitable starting point for the subcellular fractionation of other cell lines. Potential pitfalls of the procedure and how to avoid them are discussed as are alterations that may need to be considered for other cell lines.
Subcellular fractionation is a procedure in which cells are lysed and separated into their constituent components through several methods. This technique can be used by researchers to determine protein localization in mammalian cells or for enrichment of low-abundance proteins that would otherwise be undetectable. While methods for subcellular fractionation currently exist, as do commercial kits that can be purchased, they suffer from several limitations that this procedure attempts to overcome. Most cell fractionation methods are exclusively detergent-based1,2, relying on the use of buffers containing increasing amounts of detergent to solubilize different cellular components. While this method is rapid and convenient, it results in impure fractions. These are designed to allow researchers to easily isolate one or two components of the cell, but are not complex enough to isolate multiple subcellular fractions from a sample at the same time. Relying solely on detergents usually results in membrane-enclosed organelles and the plasma membrane being indiscriminately solubilized, making separation of these components difficult. An additional complication from the use of these kits is the inability of researchers to alter/optimize them for specific applications, as most of the components are proprietary formulations. Finally, these kits can be prohibitively expensive, with limitations in the number of uses that make them less than ideal for larger samples.
Despite the availability of kits for the isolation of mitochondria that do not rely on detergents, they are not designed to isolate the plasma membrane and yield significantly lower amounts of sample than standard isolation protocols3,4. While differential centrifugation methods are more time-consuming, they often result in distinct fractions that cannot be obtained with exclusively detergent-based kits1. Separation without the sole use of solubilizing detergents also allows further purification using ultracentrifugation and isopycnic density gradients, resulting in less cross-contamination. This fractionation protocol demonstrates the isolation of subcellular fractions from U937 monocytes using a combination of detergent- and high-speed centrifugation-based approaches. This method will facilitate the isolation of the nuclear, cytoplasmic, mitochondrial, and plasma membrane components of a mammalian cell with minimal contamination between the fractions.
1. Prepare buffers and reagents
2. Cytosolic protein isolation
NOTE: The following steps will allow for the growth and expansion of U937 cells followed by extraction of cytosolic proteins. At the concentration used, digitonin will permeabilize the plasma membrane without disrupting it, allowing for the release of cytosolic proteins and retention of other cellular proteins.
3. Cell homogenization
NOTE: The following steps will allow for the mechanical homogenization of digitonin-treated cells (from step 2.9), which is necessary for the isolation of the mitochondrial and membrane protein fractions.
4. Debris removal and isolation of crude mitochondrial and membrane fractions
NOTE: The following steps will allow for the removal of cellular debris by centrifuging the homogenate at increasing speeds. This is followed by differential centrifugation for the isolation of crude mitochondrial and membrane fractions.
5. Isopycnic density gradient purification
NOTE: The following steps utilize isopycnic density gradient centrifugation to purify the crude mitochondrial and membrane fractions.
6. Nuclear protein isolation
NOTE: Using ionic and non-ionic detergents as well as techniques such as sonication and centrifugation, the following steps will solubilize all cellular membranes and allow for the isolation of nuclear proteins.
7. Protein quantification and western blot analysis
NOTE: The following steps will quantify total protein in each fraction and confirm the purity of the subcellular fractions.
A schematic flow chart of this procedure (Figure 1) visually summarizes the steps that were taken to successfully fractionate U9375 cells grown in suspension. Fractions collected from the top of the isopycnic density gradient in equal volumes (1 mL) show the purification of the mitochondrial and membrane fractions (Figure 2). Utilizing an antibody against VDAC, a protein localized to the outer mitochondrial membrane6, shows that the mitochondrial fraction migrated to the 25% and 30% iodixanol (v/v) fractions (Figure 2A). Using an antibody against the Na,K+-ATPase α1 subunit, part of an integral membrane heterodimer found primarily in the plasma membrane7, shows the separation of membrane contamination from the pure mitochondrial fraction (Figure 2A). The pure membrane fraction migrated to the least dense fractions, 10% and 15% iodixanol (v/v) (Figure 2B). Mitochondrial contamination of the membrane fraction was separated by the gradient.
A western blot8 performed with the additional localization markers (referenced in step 7.4) shows the purity of the cytosolic and nuclear fractions, while additionally verifying that the mitochondrial and membrane samples are free from contamination by proteins from other parts of the cell (Figure 3). Using an antibody against GAPDH, normally localized to the cytoplasm of the cell9, shows that this protein is only found in the cytosolic fraction (Figure 3A, Lane 1, first panel), and that no contamination is observed in the extracted nuclear proteins, the density-purified mitochondria, or membrane fractions (Figure 3A; Lanes 2, 3, and 4; first panel). Probing for histone H3, a protein found in the nucleus and involved in chromatin structure10, shows a successful nuclear extraction (Figure 3A, Lane 2, second panel), with some minimal detection in the cytoplasmic fraction and no cross contamination in the mitochondrial or membrane fractions.
Probing for VDAC in all fractions shows the presence of this protein in the pure mitochondrial fraction (Figure 3A, Lane 3, third panel), and that no cross contamination exists in the other fractions (Figure 3A; Lanes 1, 2, and 4; third panel). Probing for the Na/K-ATPase α1 subunit similarly shows that this protein is located only in the pure membrane fraction (Figure 3A, Lane 4, fourth panel). These fractions were analyzed by densitometry to confirm reproducibility and statistical significance (Figure 3B–E). In contrast to the results from the successful fractionation (Figure 3), improper execution of this method (or failure to adhere to all recommended steps) can result in cross-contamination of cellular components (Figure 4). A high concentration of histone H3 in the cytosolic fraction (Figure 4, Lane 1, second panel) can result from a failure to properly clarify the cytosolic fraction (referenced in step 2.6). This can occur if not enough clarification centrifuge spins are performed, or if the cytosolic fraction is not clarified quickly. If the cytosolic fraction is not clarified quickly enough, it may result in lysis of cell fragments, leading to contamination of the cytosolic fraction.
Failure to perform the isopycnic density purification step will result in contamination of the membrane fraction (Figure 4, Lane 4, all panels), depending on how heterogenous the sample is prior to density purification. Proper adherence to all steps of the protocol is critical to obtaining the desired separation of the subcellular fractions. When quantified using a Bradford assay, protein yield for each fraction can be determined. Expected protein yield per fraction is reported in Table 1. It is useful to perform a protein quantification assay prior to performing western blots for several reasons. First, it confirms that fractions indeed contain protein; second, it allows loading of SDS-PAGE gels based on protein quantity; and finally, assuming protein yields are similar to those expected (Table 1), it confirms proper execution of the procedure.
Figure 1: Diagram of the cell fractionation procedure. An overview of the cell fractionation protocol represented as a flow chart. Abbreviations: PBS = phosphate-buffered saline; CS = cell solubilization; NL = nuclear lysis. Please click here to view a larger version of this figure.
Figure 2: Isopycnic density purification of crude mitochondrial and membrane fractions. (A) Representative western blot of all fractions collected after density gradient purification of the crude mitochondrial fraction. (B) Representative western blot of all fractions collected after density gradient purification of the crude membrane fraction. Both density gradient purifications show migration of the mitochondrial marker, VDAC, for the 25% and 30% iodixanol (v/v) fractions and migration of the membrane marker, Na,K+ ATPase, for the 10% and 15% iodixanol (v/v) fractions. Abbreviation: VDAC = voltage-dependent anion channel. Please click here to view a larger version of this figure.
Figure 3: Successful isolation of U937 cytosolic, nuclear, mitochondrial, and membrane fractions. (A) Representative western blots of cell fractions isolated from a U937 cell culture with this technique and probed for markers of cytoplasm (GAPDH, first panel), nucleus (Histone H3, second panel), mitochondria (VDAC, third panel), and membrane (Na,K+ ATPase α1, fourth panel). (B–E) Densitometry of western blots of cell fractions isolated from U937 cell cultures with this technique. Results are from 3 independent experiments. Error bars represent standard deviation. One-way ANOVA. ***p<0.001. Abbreviations: GAPDH = glyceraldehyde-3-phosphate dehydrogenase; VDAC = voltage-dependent anion channel; cyto = cytosolic; nuc = nuclear; mito = mitochondrial; mem = membrane. Please click here to view a larger version of this figure.
Figure 4: Incomplete fractionation of U937 cell components. Representative western blots of cell fractions isolated from a U937 cell culture showing contamination of the cytosolic fraction with histone H3 (Lane 1, second panel) due to improper clarification of this fraction and a crude membrane fraction that was not subjected to isopycnic density gradient purification (Lane 4, all panels). Abbreviations: cyto = cytosolic; nuc = nuclear; mito = mitochondrial; mem = membrane. Please click here to view a larger version of this figure.
Fraction | Yield (µg/mL) | Standard Deviation | Approximate Volume |
Cytoplasm | 1500 | 146 | 5 mL |
Nucleus | 1200 | 172 | 500 µL |
Mitochondria | 400 | 66 | 500 µL |
Membrane | 200 | 23 | 500 µL |
Table 1: Protein yield for each subcellular fraction.
This method is a modified version of a previously published approach to subcellular fractionation without the use of high-speed centrifugation11. This modified method requires more specialized equipment to achieve the best results, but is more comprehensive and consistently reproducible.
The development of the initial protocol was necessary due to an inability to separate mitochondrial and membrane samples for the analysis of protein localization during necroptosis12. Attempts to use the exclusively detergent-based methods found in most commercially available kits resulted in a homogeneous mixture containing the plasma membrane and all membrane-enclosed organelles in the cell. Other limitations of these kits include an inability to make alterations to the procedure, cost per sample, volume restrictions, and number of samples that can be processed. The procedure presented here may be altered to any scale, changed to isolate fewer fractions, and can be performed without the use of expensive reagents. Fraction yields can be increased by utilizing more cells; steps can be tailored to the research being performed; and execution of the method is flexible. For example, if researchers are not examining a particular subcellular fraction, they do not need to isolate that sample in the course of the procedure. Likewise, the addition of particular inhibitors or reagents can be omitted if the researcher does not plan on studying the phosphorylation state of proteins (sodium orthovanadate) or is not concerned about denaturing proteins (hexylene glycol).
The use of iodixanol as the density gradient solution is optional; however, this reagent does not interfere with subsequent examination of samples via western blotting. It is also possible to remove the iodaxanol from samples by dilution and centrifugation to recover the mitochondria or membrane, although this will affect the final yield. Other alternative density gradient solutions can be used, including sucrose. To obtain optimal results and pure fractions, there are several factors to consider, and particular critical steps in the protocol that need careful attention. This protocol is optimized for the fractionation of U937 cells, and the concentration of cells recommended at particular points in the protocol are specific to this cell line. These values were determined empirically and will most likely need to be adjusted for different types of cells, particularly, if suboptimal results are obtained when executing the protocol.
Clarification of the cytoplasmic sample should occur as soon as possible to remove unbroken cells and debris that might result in cross-contamination by proteins from other subcellular fractions (Figure 4, Lane 1, second panel). Homogenization can be accomplished with any form of mechanical lysis, although results presented here were obtained using a bead-based method and blender device. Alternative manual forms of homogenization (a Dounce homogenizer or passage through a small gauge needle) can also be utilized, but may result in reproducibility issues due to variability of technique by the individual performing the procedure. This group's interest is primarily in circulating leukocytes, which is why U937 cells are used in this procedure. However, this procedure may be applied to other suspension cell lines with likely little need for alteration. Portions that may need to be adjusted to accommodate another suspension cell line include cell concentrations used throughout the procedure as well as the concentration of digitonin used to extract the cytosolic fraction.
While not optimized for adherent cell lines, this procedure may serve as a starting point to which adjustments can be made to accommodate adherent cells. These adjustments include cell concentration, concentration of digitonin, and homogenization time. In addition, adherent cells are limited by surface area while suspension cells are limited by volume; this makes scaling up of suspension cells simpler. To scale up adherent cells, tissue culture plates with a large surface area (>500 cm2) must be utilized. This procedure is a cost-effective, reproducible subcellular fractionation with the ability to separate cytosolic, nuclear, mitochondrial, and membrane fractions with great purity. One of the biggest advantages of this procedure is the separation of mitochondrial from the membrane fraction. This is not possible in exclusively detergent-based procedures. Although detergents are used in this procedure, they are used to permeabilize, but not disrupt the plasma membrane (digitonin) and to obtain the nuclear fraction after the isolation of the mitochondrial and membrane fractions.
The authors have nothing to disclose.
This work was supported by NIH R15-HL135675-01 and NIH 2 R15-HL135675-02 to T.J.L.
Benzonase Nuclease | Sigma-Aldrich | E1014 | |
Bullet Blender Tissue Homogenizer | Next Advance | 61-BB50-DX | |
digitonin | Sigma | D141 | |
end-over-end rotator | ThermoFisher | ||
Ethylenediaminetetraacetic acid (EDTA) | Sigma | E9884 | |
ethylene glycol-bis(β-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA) | Sigma | E3889 | |
GAPDH (14C10) | Cell Signalling Technologies | 2118 | |
HEPES | VWR | 97064-360 | |
Hexylene glycol | Sigma | 68340 | |
Igepal | Sigma | I7771 | Non-ionic, non-denaturing detergent |
KCl | Sigma | P9333 | |
Mannitol | Sigma | M9647 | |
MgCl2 | Sigma | M8266 | |
NaCl | Sigma | S9888 | |
Na, K-ATPase a1 (D4Y7E) | Cell Signalling Technologies | 23565 | |
Open-Top Polyclear Tubes, 16 x 52 mm | Seton Scientific | 7048 | |
OptiPrep (Iodixanol) Density Gradient Medium | Sigma | D1556-250ML | |
phenylmethanesulfonyl fluoride (PMSF) | Sigma | P7626 | |
Protease Inhibitor Cocktail, General Use | VWR | M221-1ML | |
refrigerated centrifuge | ThermoFisher | ||
S50-ST Swinging Bucket Rotor | Eppendorf | ||
Sodium dodecyl sulfate (SDS) | Sigma | 436143 | |
Sodium deoxycholate | Sigma | D6750 | |
sodium orthovanadate (SOV) | Sigma | 567540 | |
sonicator | ThermoFisher | ||
Sorvall MX120 Plus Micro-Ultracentrifuge | ThermoFisher | ||
Stainless Steel Beads 3.2 mm | Next Advance | SSB32 | |
Sucrose | Sigma | S0389 | |
Tris-buffered Saline (TBS) | VWR | 97062-370 | |
Tween 20 | non-ionic detergent in western blotting buffers | ||
VDAC (D73D12) | Cell Signalling Technologies | 4661 |