Nitrogen Cavitation and Differential Centrifugation Allows for Monitoring the Distribution of Peripheral Membrane Proteins in Cultured Cells

Summary

Here we present protocols for detergent-free homogenization of cultured mammalian cells based on nitrogen cavitation and subsequent separation of cytosolic and membrane-bound proteins by ultracentrifugation. This method is ideal for monitoring the partitioning of peripheral membrane proteins between soluble and membrane fractions.

Abstract

Cultured cells are useful for studying the subcellular distribution of proteins, including peripheral membrane proteins. Genetically encoded fluorescently tagged proteins have revolutionized the study of subcellular protein distribution. However, it is difficult to quantify the distribution with fluorescent microscopy, especially when proteins are partially cytosolic. Moreover, it is often important to study endogenous proteins. Biochemical assays such as immunoblots remain the gold standard for quantification of protein distribution after subcellular fractionation. Although there are commercial kits that aim to isolate cytosolic or certain membrane fractions, most of these kits are based on extraction with detergents, which may be unsuitable for studying peripheral membrane proteins that are easily extracted from membranes. Here we present a detergent-free protocol for cellular homogenization by nitrogen cavitation and subsequent separation of cytosolic and membrane-bound proteins by ultracentrifugation. We confirm the separation of subcellular organelles in soluble and pellet fractions across different cell types, and compare protein extraction among several common non-detergent-based mechanical homogenization methods. Among several advantages of nitrogen cavitation is the superior efficiency of cellular disruption with minimal physical and chemical damage to delicate organelles. Combined with ultracentrifugation, nitrogen cavitation is an excellent method to examine the shift of peripheral membrane proteins between cytosolic and membrane fractions.

Introduction

Cellular proteins can be divided into two classes: those that are associated with membranes and those that are not. Non-membrane associated proteins are found in the cytosol, nucleoplasm and lumina of organelles such as the endoplasmic reticulum (ER). There are two classes of membrane-associated proteins, integral and peripheral. Integral membrane proteins are also known as transmembrane proteins because one or more segments of the polypeptide chain spans the membrane, typically as an α-helix composed of hydrophobic amino acids. Transmembrane proteins are co-translationally inserted into membranes in the course of their biosynthesis and remain so configured until they are catabolized. Peripheral membrane proteins are secondarily driven to membranes, usually as a consequence of post-translational modification with hydrophobic molecules such as lipids. In contrast to integral membrane proteins, the association of peripheral membrane proteins with cellular membranes is reversible and can be regulated. Many peripheral membrane proteins function in signaling pathways, and regulated association with membranes is one mechanism for activating or inhibiting a pathway. One example of a signaling molecule that is a peripheral membrane protein is the small GTPase, RAS. After a series of post-translational modifications that include modification with a farnesyl lipid, the modified C-terminus of a mature RAS protein inserts into the cytoplasmic leaflet of the cellular membrane. Specifically, the plasma membrane is where the RAS engages its downstream effector RAF1. To prevent constitutive activation of mitogen-activated protein kinase (MAPK) pathway, multiple levels of control of RAS are in place. Besides rendering RAS inactive by hydrolyzing GTP into GDP, active RAS also can be released from the plasma membrane by modifications or interactions with solubilizing factors to inhibit signaling. Although fluorescent live imaging affords cell biologists the opportunity to observe the subcellular localization of fluorescent protein-tagged peripheral membrane proteins¹, there remains a critical need to evaluate membrane association of endogenous proteins semi-quantitatively with simple biochemical approaches.

The proper biochemical evaluation of protein partitioning between membrane and soluble fractions is critically dependent on two factors: cellular homogenization and efficient separation of membrane and soluble fractions. Although some protocols, including the most widely used commercialized kits, depend on detergent-based cell homogenization, these methods can obfuscate analysis by extracting membrane proteins into the soluble phase². Accordingly, non-detergent based, mechanical methods of cell disruption provide cleaner results. There are several methods of mechanical disruption of cells grown in culture or harvested from blood or organs. These include Dounce homogenization, fine needle disruption, ball-bearing homogenization, sonication and nitrogen cavitation. Here we evaluate nitrogen cavitation and compare it to other methods. Nitrogen cavitation relies on nitrogen that is dissolved in the cytoplasm of the cells under high pressure. After equilibration, the cell suspension is abruptly exposed to atmospheric pressure such that nitrogen bubbles are formed in the cytoplasm that tear open the cell as a consequence of their effervescence. If the pressure is sufficiently high, nitrogen effervescence can disrupt the nucleus³ and membrane bound organelles like lysosomes⁴. However, if the pressure is kept low enough, the decompression will disrupt the plasma membrane and ER but not other organelles, thereby spilling both cytosol and intact cytoplasmic organelles into the homogenate that is designated the cavitate⁵. For this reason, nitrogen cavitation is the method of choice for isolating organelles like lysosomes and mitochondria.

However, it is also an excellent way of preparing a homogenate that can be easily separated into membrane and soluble fractions. The pressure vessel (henceforth called "the bomb") used during cavitation consists of a thick stainless steel casing that withstands high pressure, with an inlet for delivery of the nitrogen gas from a tank and an outlet port with an adjustable discharge valve.

Nitrogen cavitation has been used for cell homogenization since the 1960s⁶. In 1961, Hunter and Commerfold⁷ established nitrogen cavitation as a viable option for mammalian tissue disruption. Since then, researchers have adapted the technique to various cells and tissues with success, and nitrogen cavitation has become a staple in multiple applications, including membrane preparation^8,9, nuclei and organelle preparation^10,11, and labile biochemical extraction. Currently, cell biologists more often employ other methods of cell homogenization because the benefits of nitrogen homogenization have not been widely advertised, nitrogen bombs are expensive and there is a misconception that a relatively large number of cells is required. Protocols for nitrogen cavitation to achieve cell-free homogenates with intact nuclei have not been published, and in most published evaluations volumes of 20 mL of cell suspension were used. To adapt this classic technique to suit current requirements of working with small-scale samples, we present a modified protocol of nitrogen cavitation specifically designed for cultured cells. After nitrogen cavitation, the homogenate is separated into soluble (S) and membrane (P) fractions by differential centrifugation, first with a low-speed spin to remove nuclei and unbroken cells, and then with a high-speed spin (>100,000 x g) to separate membranes from the soluble fraction. We analyze the efficiency of the separation with immunoblots and compare nitrogen cavitation with other mechanical disruption techniques. We also investigate the osmotic effect of homogenization buffer during nitrogen cavitation.

Protocol

1. Buffer and Equipment Preparations

1. Chill 45 mL cell disruption bombs, 15 mL tubes, and ultracentrifugation tubes at 4 °C.
2. Prepare and chill 25 mL of homogenization buffer per 2 x 10⁷ cells at 4 °C. Add one protease inhibitor tablet just before use.
   NOTE: Homogenization buffers typically contain KCl rather than NaCl to better reflect intracellular salt composition. Homogenization buffer used in this protocol consists of 10 mM HEPES at pH 7.4, 10 mM KCl and 1.5 mM MgCl₂ (hereinafter referred to as hypotonic homogenization buffer). Most buffers can be adapted for nitrogen cavitation (see Discussion).
3. Prepare and chill 6 mL of 1x Phosphate-Buffered Saline (PBS) buffer per sample at 4 °C. Add protease inhibitor tablets fresh before use.
   NOTE: PBS buffer used in this protocol consists of 10 mM Na₂HPO₄ at pH 7.4, 1.8 mM KH₂PO₄, 137 mM NaCl and 2.7 mM KCl.
4. Prepare and chill 4 mL of solubilization buffer per sample at 4 °C. Add one protease inhibitor tablet just before use.
   NOTE: Solubilization buffer used in this protocol is 1x Radioimmunoprecipitation assay (RIPA) buffer, which consists of 25 mM Tris at pH 7.4, 150 mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, and 1% NP-40. See 4.6 NOTE.

2. Cell Harvesting

1. Grow 2 x 10⁷-10 x10⁹ tissue culture cells with the recommended culture media for the cell type. Typically, one 15-cm dish yields 2 x 10⁷ HEK-293 cells cultured in DMEM at 90% confluency (Figure 1).
2. Remove the growth medium by vacuum.
3. For adherent cells, place the culture dishes on ice, wash the cells directly on the culture dishes gently with chilled homogenization buffer twice (10 mL buffer per 15 cm dish per wash) and harvest the cells with a large cell scraper in an appropriate volume of homogenization buffer; For suspension cells, collect and wash the cell pellet in chilled homogenization buffer twice (10 mL buffer per 50-mL culture per wash) with 500 x g spin at 4 °C for 5 min, and resuspend the washed cell pellet in an appropriate volume of homogenization buffer on ice.
   NOTE: The volume should be based on the requirements for protein concentration in the intended experiments as well as the minimal/maximal volume allowed in the cell disruption bomb. A general guideline is 2-10 x 10⁷ cells/mL, or about 10 volumes of the cell pellet. This protocol is optimized for three 15-cm dishes of HEK-293 cells in 2 mL of homogenization buffer.

3. Nitrogen Cavitation

1. Transfer the cell suspension to a clean and chilled bomb in an ice bath on a stir plate.
   CAUTION: The bomb has high pressure, low temperature, nitrogen gas – wear appropriate personal protection.
2. Place a micro magnetic stir bar inside the bomb and turn on the stir plate to maintain suspension homogeneity.
3. Add one protease inhibitor tablet to the suspension and close the bomb per manufacturer's instruction.
4. Gradually pressurize the bomb with a nitrogen gas tank per manufacturer's instruction till the bomb pressure gauge reads 300 to 600 psi. Close all valves and disconnect the nitrogen tank.
   NOTE: The pressure required may vary with the cell type. Here we performed cavitation at 350 to 400 psi for HEK-293, NIH-3T3 and Jurkat cells.
5. Wait for 20 min to allow the nitrogen to dissolve and reach equilibrium within the cells.
6. Remove excess water around the discharge valve using a cloth towel. Open the discharge valve gently to achieve a dropwise release of homogenate and collect in a pre-chilled 15 mL tube.
   NOTE: Near the end of collection there will be a spurt of homogenate and gas will emerge with a hissing sound. Make sure that the gas does not cause previously collected cavitate to shoot out of the tube (hence the use of 15 mL tubes instead of 1.5 mL tubes). Once the spurt starts, close the discharge valve and open the nitrogen inlet valve abruptly to depressurize the bomb and achieve cavitation of the remaining cells in the bomb. Open the bomb for cavitate recovery and thorough cleaning.
   NOTE: The final cavitate should have a milky appearance with foam on top. Gently stir with a pipette tip to allow the foam to subside before centrifugation.
   NOTE: Examine the cavitate by phase-contrast microscopy to determine the homogenization efficiency. Add a 15 µL drop of cavitate to the surface of a microscopy slide and cover with a coverslip. Repeat step 3.4-3.6 only if too many unbroken cells are detected with a 20X objective.
   NOTE: If homogenization buffer does not contain EDTA or EGTA, add it to the collected cavitate at a final concentration of 1 mM within 5 min after discharge.

4. Separation of Cytosolic and Membrane Fractions

1. Centrifuge the cavitate at 500 x g for 10 min at 4 °C to remove unbroken cells and nuclei.
   NOTE: Repeat the centrifugation step until no visible pellet is produced and collect the Post Nuclear Supernatant (PNS) while avoiding the foam floating on top. Re-centrifuge the foam to further collect and combine PNS, if necessary (Figure 1).
2. Process the PNS as desired to obtain fractions of interest. For the purpose of separating cytosolic and membrane fractions, transfer the PNS to an ultracentrifuge tube and perform ultracentrifugation as desired. This protocol is optimized for a <3.5 mL sample in a polycarbonate ultracentrifuge tube and for ultracentrifugation at 350,000 x g for 1 h at 4 °C.
3. Collect the supernatant (the S fraction) using a 1 mL pipette.
4. Carefully rinse the pellet with 3 mL of cold PBS without disturbing it. Remove the PBS by vacuum.
   NOTE: If contamination of the membrane fraction by cytosolic proteins is a greater concern than the loss of sample, resuspend the pellet in 3 mL of cold PBS and re-ultracentrifuge as in step 4.2. Remove the PBS by vacuum.
5. Resuspend the pellet fully in an appropriate volume of detergent-containing solubilization buffer of choice. To achieve cell equivalence, use the same volume of solubilization buffer as the cytosolic fraction.
   NOTE: We suggest using 1x RIPA buffer as solubilization buffer for efficient membrane protein extraction. If no downstream assay is required for membrane fraction, use 1x laemmli sample buffer for maximal membrane protein extraction.
   NOTE: We suggest dislodging and transferring the pellet in solubilization buffer to a clean tube on a tube rotator at 4 °C for maximal membrane protein extraction.
   NOTE: If the pellet is too sticky (too many lipids) to be efficiently removed from the ultracentrifuge tube in solubilization buffer, we suggest snap freezing the pellet in liquid nitrogen and quickly dislodging the pellet from the ultracentrifuge tube with a mini metal spatula before the pellet thaws.
   NOTE: Alternatively, membrane pellets can be just resuspended and not solubilized in a non-detergent-containing buffer to produce a P fraction of membrane vesicle suspension, in which case the centrifugation step in 4.6 is not required. Such P fractions are useful when investigating functions such as enzyme activities that are dependent on membrane association.
6. Centrifuge the fully solubilized pellet suspension from step 4.5 at 20,000 x g in a tabletop centrifuge for 10 min at 4 °C. Collect the supernatant using a 1 mL pipette (the P fraction) and discard the pellet (insoluble lipids).
7. Perform desired assays such as western blotting with the cytosolic and/or membrane fractions, or save them at -80 °C for future use.

Results

Figure 2 shows the partitioning of cellular proteins from PNS into either the soluble cytosolic fraction (S) or membrane pellet fraction (P). We examined three representative cell lines from different cell types: HEK-293 (epithelial), NIH-3T3 (fibroblast), and Jurkat (lymphocyte). Rho Guanine Dissociation Inhibitor (RhoGDI) and cation-independent mannose-6-phosphate receptor (CIMPR) were used as positive controls for cytosolic and membrane fractions, respecti...

Discussion

The advantages of nitrogen cavitation over other methods of mechanical disruption are manifold. Perhaps the most significant benefit is its ability to gently yet efficiently homogenize specimens. The physical principles of decompression cools samples instead of generating local heating damage like ultrasonic and friction/shearing based techniques. Cavitation is also extremely efficient at disrupting the plasma membrane. Because nitrogen bubbles are generated within each individual cell upon decompression, the cavitation ...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Cell Disruption Vessel (45 mL)	Parr Instrument	4639	Nitrogen cavitation Bomb
Dounce homogenizer (2 mL)	Kontes	885300-0002	Dounce pestle and tube
U-100 Insulin Syringe 28G½	Becton Dickinson	329461	Needle
Atg12 antibody	Santa Cruz	271688	Mouse antibody, use at 1:1000 dilution
β-actin antibody	Santa Cruz	47778	Mouse antibody, use at 1:1000 dilution
β-tubulin antibody	DSHB	E7-s	Mouse antibody, use at 1:5000 dilution
Calnexin antibody	Santa Cruz	23954	Mouse antibody, use at 1:1000 dilution
Calregulin antibody	Santa Cruz	373863	Mouse antibody, use at 1:1000 dilution
Catalase antibody	Santa Cruz	271803	Mouse antibody, use at 1:1000 dilution
CIMPR antibody	Abcam	124767	Rabbit antibody, use at 1:1000 dilution
EEA1 antibody	Santa Cruz	137130	Mouse antibody, use at 1:1000 dilution
EGFR antibody	Santa Cruz	373746	Mouse antibody, use at 1:1000 dilution
F0-ATPase antibody	Santa Cruz	514419	Mouse antibody, use at 1:1000 dilution
F1-ATPase antibody	Santa Cruz	55597	Mouse antibody, use at 1:1000 dilution
Fibrillarin antibody	Santa Cruz	374022	Mouse antibody, use at 1:200 dilution
Golgin 97 antibody	Santa Cruz	59820	Mouse antibody, use at 1:1000 dilution
HDAC1 antibody	Santa Cruz	81598	Mouse antibody, use at 1:1000 dilution
Hexokinase 1 antibody	Cell Signaling Technology	2024S	Rabbit antibody, use at 1:1000 dilution
Lamin A/C antibody	Santa Cruz	376248	Mouse antibody, use at 1:1000 dilution
LAMP1 antibody	DSHB	H4A3-c	Mouse antibody, use at 1:1000 dilution
Na+/K+ ATPase antibody	Santa Cruz	48345	Mouse antibody, use at 1:1000 dilution
Rab7 antibody	Abcam	137029	Rabbit antibody, use at 1:1000 dilution
Rab9 antibody	Thermo	MA3-067	Mouse antibody, use at 1:1000 dilution
RCAS1 antibody	Santa Cruz	398052	Mouse antibody, use at 1:1000 dilution
RhoGDI antibody	Santa Cruz	360	Rabbit antibody, use at 1:3000 dilution
Ribosomal protein S6 antibody	Santa Cruz	74459	Mouse antibody, use at 1:1000 dilution
Sec61a antibody	Santa Cruz	12322	Goat antibody, use at 1:1000 dilution
Thickwall Polycarbonate ultracentrifuge tube	Beckman Coulter	349622	Sample tube for ultracentrifugation
TLK-100.3 rotor	Beckman Coulter	349481	rotor for ultracentrifugation
Optima MAX High-Capacity Personal Ultracentrifuge	Beckman Coulter	364300	ultracentrifuge
cOmplete protease inhibitor cocktail tablets	Roche	11697498001	protease inhibitors
Cell Scrapers with 25cm Handle and 3.0cm Blade	Corning	353089	large cell scraper
Magnetic Stir Bar	Fisher Scientific	14-513-57SIX	micro stir bar
Ceramic-Top Magnetic Stirrer	Fisher Scientific	S504501AS	magnetic stirrer