Introduction

The application of mass-spectrometry-based proteomics to profiling whole organs or tissues requires that the samples be fractionated to increase the detection limits of analysis^{1, 2, 3, 4}. We have developed a systematic protocol for fractionating whole tissues and organelles into subcellular components^{1, 5}. This protocol generates cytosol, nuclei, mitochondria and microsomes (a collection of endoplasmic recticulum (ER), Golgi, intracellular vesicles and plasma membrane) that can be extracted for both soluble and insoluble protein components^{1, 5}. Others have also published protocols for the isolation of multiple subcellular components by multiple-step gradient centrifugation⁶ and continuous gradients⁷. Continuous gradients are difficult to generate in a consistent fashion; in addition, there is then the need to collect fractions of the gradient and deconvolute the observed profiles⁷. Other fractionation protocols exist to specifically and highly purify subcellular compartments and organelles, but these generally require more starting material, time and special reagents^{2, 3, 4}. This protocol isolates these subcellular fractions from a single sample, reducing the quantity of starting material and generating an overall proteomics snapshot of the sample^{1, 5}. Although no generic fractionation procedure exists for all tissues, in two previous publications we have shown this method to be robust for a variety of organs and that the subcellular fractions obtained are of reasonably high purity^{1, 5}. Although the procedure described here works generally well for lung, kidney, placenta, brain and liver, changes may be required to: the volumes of buffer; homogenization; and density gradients used to optimize lysis and fractionation for other organs or tissues. Some detergents can be used, but may affect later steps of analysis, such as mass spectrometry. We have found this protocol to be readily applicable to cell culture; however, frozen tissue shows poor results as the freeze–thaw step generally lyses the cells and its subcellular components, resulting in higher cross-contamination of the fractions (B.C. and A.E., unpublished observations).

The homogenization is based on shearing with a Teflon-glass dounce homogenizer and an electric drill. We have found that glass–glass homogenizers and homogenization by hand also work (B.C. and A.E., unpublished observations). Rotary-blade homogenizers, such as a Polytron, will lyse tissues but will also tend to lyse nuclei and mitochondria and cause higher cross-contamination of the subcellular fractions (B.C. and A.E., unpublished observations). Other published protocols use digitonin for cell lysis; however, this is typically used for the specific isolation of mitochondria⁸. Digitonin can also lyse other intracellular membranes and, at higher concentrations, it will also lyse mitochondria⁹. The protocol below describes volumes for the fractionation of adult organ tissues; however, we have successfully applied the protocol to smaller samples, such as mouse lungs and hearts, at embryonic day 13.5. For these smaller samples, the initial centrifugation was carried out in 1.5-ml snap-cap microfuge tubes and the ultracentrifuge step was also scaled down to smaller tubes and rotors (B.C. and A.E., unpublished observations).

This protocol may not reveal the proteome of a specific subcellular fraction to the same depth as methods that are specifically optimized to generate highly purified subcellular components. However, our methodology does afford the advantage of comparison of multiple fractions so that different subcellular compartments from the same sample can be compared to judge the relative enrichment of a protein. The generation of a general microsomal fraction does combine many membranous fractions together, such as Golgi, ER and intracellular vesicles. In addition, we have noted that the ribosome is also highly enriched with this fraction, probably in association with the rough ER, as the lysis method is gentle and detergent-free⁵. However, the separation and isolation of each of these components of our general microsomal fractions would be more time-consuming and lead to a large increase in samples to be analyzed.

We have utilized Triton-X-100 for the extraction of proteins from the insoluble portion of many of the fractions. Detergents can interfere with the collection of high-quality mass spectra; however, Triton-X-100 is easily removed during sample clean-up. Although some methods for extraction of nuclear protein use DNAse, we have found that treatment with high salt and then detergent, combined with shearing through a syringe, fragments the DNA and releases the majority of proteins. In general, DNAse-based extraction is better for immunoprecipitation or pull-downs of tagged proteins, for which it is desirable to maintain intact protein complexes. We have found that Triton-X-100 readily extracts nuclear-pore components and histones⁵, but it may be suboptimal for the extraction of inner nuclear-membrane proteins; nevertheless, components such as inner nuclear-membrane protein Man1 and lamins are detected⁵. Users may wish to test different surfactants and detergents to alter the range of components extracted, bearing in mind that they need to be removed prior to mass-spectrometric analysis. Acetone tri-chloro-acetic acid precipitation seems to be an effective method of both concentrating the protein sample and removing detergents. For the microsomal fraction obtained after the 100,000g spin, we noted that the pellet readily dissolved in buffer with Triton-X-100. However, some membrane components may not be solubilized by this method and, therefore, users may wish to try different detergents to re-extract any material left behind. This comes with the obvious caveat that further fractionation will increase the number of fractions for analysis.

The protocol presented here is optimized to provide a survey of the cell or tissue proteome and may not be optimized for individual organelle or cellular subcomponents. Nevertheless, this provides an excellent starting point for global cellular protein analysis. We typically run the protocol from dissection to protein extraction in 5–6 h and freeze the samples for processing at a later date for analysis using Multidimensional Protein Identification Technology (MudPIT)⁵. By MudPIT analysis, we detect a slight loss of proteins with pI values in the neutral pH range^{1, 5}. Data collected on each fraction via MudPIT analysis has been shown to be applicable to quantification via the spectral counts method and amenable to comparison with microarray data^{1, 5}.

Materials

Reagents

• 1 M (1,000) dithiothreitol (DTT) (Sigma-Aldrich); aliquot and freeze at - 20 °C
• 1,000 phenylmethylsulphonylfluoride (PMSF) stock (Sigma-Aldrich), prepared as per the manufacturers instructions
• Phosphate-buffered saline (PBS) (Sigma-Aldrich)
• 25 mg ml^{- 1} (1,000) Spermine (Sigma-Aldrich); aliquot and freeze at - 20 °C
• 25 mg ml^{- 1} (1,000) Spermidine (Sigma-Aldrich); aliquot and freeze at - 20 °C
• 250-STM Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• 250-STMDPS Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• 2 M STM Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• 2 M STMDPS Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• NE Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• NET Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• HDP Buffer (Sigma-Aldrich) (see REAGENT SETUP)
• ME Buffer (Sigma-Aldrich) (see REAGENT SETUP)
  Critical Make all buffers with MilliQ grade water or better. Filter-sterilize all solutions and store as indicated.

Equipment

• Cheesecloth or gauze
• Electric drill capable of a minimum of 1000 r.p.m.
• Teflon-glass dounce homogenizer, range of sizes
• Fixed-angle centrifuge capable of 6,000g (most bands of table-top centrifuges and compatible rotors, such as Beckman C650)
• Ultracentrifuge with a swing-bucket rotor capable of 80,000g and 100,000g (SW 28 Ti and SW 41 Ti Beckman rotors or equivalent)
• Ultracentrifuge tubes
• Ice buckets/cold room
• 18-gauge needle
• Syringes: 1 ml and 5 ml
• Funnel
• Cannula (optional)
• 20–60 ml syringe (optional)
• Peristaltic pump (optional)

Reagent setup

• 250-STM Buffer Stock buffer for initial homogenization. Filter-sterilize and store at 4 °C. Consists of 250 mM sucrose, 50 mM Tris-HCl (pH 7.4), 5 mM MgCl₂.
• 250-STMDPS Buffer Working buffer for initial homogenization. Add DTT to 1 mM, Spermine to 25 g ml^{- 1}, Spermidine to 25 g ml^{- 1} and PMSF as per the manufacturer's specifications to a suitable volume of 250-STM Buffer. Other suitable protease inhibitors may also be used at the supplier's recommended concentration.
• 2 M STM Buffer Stock buffer for second homogenization and cushion. Filter-sterilize and store at 4 °C. Consists of 2 M sucrose, 50 mM Tris-HCl (pH 7.4), 5 mM MgCl₂, 1 mM DTT and 1 mM PMSF. Dissolve the sucrose by mixing at an elevated temperature.
• 2 M STMDPS Buffer Working buffer for second homogenization and cushion. Add DTT to 1 mM, Spermine to 25 g ml^{- 1}, Spermidine to 25 g ml^{- 1} and PMSF as per the manufacturer's specifications to a suitable volume of 2 M STM Buffer. Other suitable protease inhibitors may also be used at the supplier's recommended concentration.
• NE Buffer For extraction of nuclear proteins. Filter-sterilize and store at 4 °C. Consists of 20 mM HEPES (pH 7.9), 1.5 mM MgCl₂, 0.5 M NaCl, 0.2 mM EDTA and 20% glycerol.
• NET Buffer Make up to 1% Triton-X-100, 1 mM DTT, 1 mM PMSF in NE Buffer.
• HDP Hypotonic lysis buffer for extraction of soluble mitochondrial proteins. Filter-sterilize and store at 4 °C. Consists of 10 mM HEPES (pH 7.9), 1 mM DTT and 1 mM PMSF.
• ME Buffer Membrane extraction buffer for mitochondria and microsomes. Filter-sterilize and store at 4 °C. Consists of 20 mM Tris-HCl (pH 7.8), 0.4 M NaCl, 15% glycerol, 1 mM DTT, 1 mM PMSF and 1.5% Triton-X-100.
  Critical Make all buffers with MilliQ grade (10 M) water or better. Filter-sterilize all solutions and store as indicated.

Procedure

1. Points from here (point 1) up to and including point 13 are related to Initial homogenizationTiming: approximately 30 min, not including dissection time
   Critical step Time and temperature are critical: there are a few pause points, but typically all fractions are collected and extracted before freezing. Work quickly to remove and rinse tissues/organ of interest, as cell death will begin to occur only a few minutes after cardiopulmonary arrest, resulting in gene turnover and changes in the proteome. Keep all tissue and buffers ice-cold to prevent protein degradation. Pre-chill the glass homogenizer on ice and pre-cool the centrifuge and rotors. If multiple centrifuges are not available, the samples will keep on ice for the short centrifuge times. Kinase and phosphatase inhibitors could be added to buffers if desired, but have not been tested.Anesthetize and kill the animal.
   Critical step Consider the method of sacrifice carefully. Killing by CO₂ can affect the activity of many enzymes and kinases, thereby altering the phosphorylation state of proteins. Cervical dislocation as a means of killing may damage tissues of interest. In all cases, you should adhere to your institute's or facilitys animal care procedures.
   Caution If using primary human tissues, these must be handled in a biohazard containment hood or as per your institute's guidelines. Extra care should be taken as the tissues or cells may contain human pathogens. The user may also require immunizations to work with the tissues or cell lines.
2. If possible, clear blood from the organ by perfusing the organs of interest with cold PBS, as detailed in Box 1. If a suitable vein or artery is not available for the perfusion, or the tissue is not amenable to perfusion, proceed directly to Step 3.
   Critical step If perfusion is not carried out, the sample may contain large amounts of blood proteins that may need to be accounted for during data analysis.
3. Remove the organ and trim away any membranes or other adherent tissue that is not desired. For small samples, this can be performed under a dissecting microscope with the tissue being immersed in cold PBS.
4. Place the tissue in a Petri dish and mince the tissue with sharp scissors.
5. Rinse the minced tissue in cold PBS according to either option A or option B. If the tissue was perfused in Step 2, follow option A. If the tissue was not perfused, follow option B:
   1. If the tissue was perfused, rinse once in 5–10 volumes of cold PBS as follows:
      1. Place the tissue in a tube, preferably with a screw-cap that will fit into the rotor of the low-speed centrifuge.
         Caution Always balance the centrifuge sample to prevent rotor imbalance.
      2. Add 5 volumes or more of cold PBS.
      3. Gently spin (100g for 1 min or less) to pellet the minced tissue.
      4. Decant or remove liquid with a pipette.
   2. If the tissue was not perfused, and needs to be cleared of excess blood, rinse four times in cold PBS as follows:
      1. Place the tissue in a tube, preferably with a screw-cap that will fit into the rotor of the low-speed centrifuge.
         Caution Always balance the centrifuge sample to prevent rotor imbalance.
      2. Add 5 volumes or more of cold PBS.
      3. Gently spin (100g for 1 min or less) to pellet the minced tissue.
      4. Decant or remove liquid with a pipette.
      5. Repeat from Step (ii) three more times.
6. Rinse the tissue once in 5–10 volumes ice-cold 250-STMDPS Buffer.
7. Place the tissue into a pre-chilled dounce glass tube and add 4–8 volumes of ice-cold 250-STMDPS Buffer. Place into an ice bucket and secure the glass tube in place with a retort stand.
   Caution Wear safety glasses and a face shield, as the glass tube may shatter if too much pressure is applied or if the drill is moved at an angle other than vertical.
   Critical step The volume of buffer may need to be optimized for a particular tissue or organ; 4–8 ml of buffer per gram of tissue is a good start.
8. Homogenize for a minimum of 2 min using a tight-fitting Teflon pestle attached to a power drill (set to >1,000 rpm) by slowly stroking the pestle up and down, taking 15 s per stroke and 30 s to complete a down and up cycle.
9. Inspect the homogenate; if intact tissue is still apparent, re-homogenize for an additional 1 min.
10. Decant the homogenate into an appropriately-sized centrifuge tube and place into a pre-cooled rotor and centrifuge.
    Critical step For large samples, conical-bottomed 50- or 14-ml polypropylene tubes are handy, as they seal, are clear and have volume measurements.
    Caution Check that centrifuge tubes are free of flaws (cracks), are balanced and never fill tubes by more than half their volume when using a fixed-angle rotor; failure to do so may result in leakage of tube contents, resulting in possible damage to the rotor, centrifuge and/or persons.
11. Centrifuge at 800g for 15 min.Troubleshooting
12. Decant and save the supernatant on ice for isolation of mitochondria, cytosol and microsomes. Label as cyto-I for use at Step 25. Label and save the pellet on ice as nuclei-I.
13. If a cleaner preparation is required, an optional additional centrifuge step can be included. Re-spin the supernatant at 800–1,000g for 15 min to pellet any nuclei that have come free from the pellet during decanting. Decant and save the supernatant as cyto-I for isolation of mitochondria, cytosol and microsomes. The pellet may be discarded.
14. Points from here (point 14) up to and including point 24 are related to Isolation of nucleiTiming: 1 hAdd 5–10 volumes of 250-STMDPS Buffer to nuclei-I (from Step 12) and re-homogenize for 1 min in the dounce homogenizer. If desired, the homogenization can be checked by placing 10 l of homogenate on a microscope slide with a glass cover slip. Visualize by phase-contrast microscopy for cell lysis. Efficient lysis should have small nuclei floating freely in the homogenate, with few intact cells visible.
15. Decant the homogenate into an appropriately-sized centrifuge tube. For large samples, conical-bottomed 50 or 14 ml polypropylene tubes are handy, as they seal, are clear and have volume markings.
    Caution Check that centrifuge tubes are balanced and never fill tubes to more than half their volume when using a fixed-angle rotor; failure to do so may result in leakage of tube contents, resulting in possible damage to the rotor, centrifuge and/or people.
16. Centrifuge at 800g for 15 min.
17. Decant and save the supernatant as cyto-II (for use in Step 25) and save the pellet as nuclei-II.
18. Re-solubilize the nuclei-II pellet in 4 volumes of 2 M STMDPS Buffer by repeated pipetting, and homogenize with a single stroke of the dounce homogenizer.
19. Filter the re-suspension through several layers of cheesecloth (or gauze) using a funnel to remove debris.
20. Prepare a 4-ml cushion of 2 M STMDPS Buffer in a 14-ml ultracentrifuge tube.
21. Layer the suspended nuclei-II pellet onto the cushion by filling a pipette with the suspension and dispense slowly with the tip, touching the side wall of the tube.
    Critical step If there is less than 10 ml of sample, add more 2 M STMDPS Buffer to the sample and mix before pipetting onto the cushion.
22. Centrifuge in a swing-bucket ultracentrifuge at 80,000g for 35 min.
23. Slowly aspirate the supernatant with a pipette, and then remove the last 1 ml with a p1000 pipette to prevent disturbing the pellet, which contains pure nuclei.
24. To extract nuclear proteins from the pellet, proceed to Step 35. To continue with tissue fractionation of the supernatant (cyto-I), proceed to Step 25. These steps can be carried out in parallel, as outlined in Figure 1.

    Figure 1: Flow chart of the different centrifugation steps and the supernatant and pellets recovered from them.

    

    Along the bottom is shown an approximate time scale for the protocol. The indicated steps should be preformed within the appropriate time range to minimize the time required to complete the entire procedure. Sup, supernatant.

    Full size image (77 KB)
    
    Pause Point The pellet may be saved and frozen at - 70 °C as pure nuclei for the later extraction of proteins.
25. Points from here (point 25) up to and including point 31 are related to Isolation of mitochondriaTiming: 45 min to 1 hSpin cyto-I (from Steps 12 or 13) and cyto-II (from Step 17) supernatants in separate tubes at 6,000g for 15 min to pellet the mitochondria.
26. Decant the cyto-I supernatant and save as cyto-III for use in Step 32; the pellet is the mitochondria.
27. Decant the cyto-II supernatant and optionally combine with cyto-III for use in Step 32; the pellet is mitochondria.
28. Gently resuspend both mitochondria pellets and combine in 10 or more volumes of 250-STMDPS Buffer with a pipette.
29. Spin again at 6,000g for 15 min.
30. Decant the supernatant and save the pellet as the mitochondria.
31. To extract mitochondrial proteins from the pellet, proceed to Step 44. To continue with tissue fractionation of the supernatant, cyto-III (from Steps 26 and 27), proceed to Step 32. These steps can be carried out in parallel, as outlined in Figure 1.Pause Point The pellet may be frozen at - 70 °C for protein extraction at a later date.
32. Points from here (point 32) up to and including point 34 are related to Isolation of cytosol and microsomesTiming: 1 h 15 minPlace the cyto-III supernatant (from Steps 26 and 27) into two ultracentrifuge tubes of appropriate volume and balance.
    Caution Check that centrifuge tubes are balanced and never fill tubes by more than half their volume when using a fixed-angle rotor; failure to do so may result in leakage of tube contents, resulting in possible damage to the rotor, centrifuge and/or people.
33. Spin for 1 h at 100,000g in a swing-bucket ultracentrifuge.
34. Decant the supernatant and save as pure cytosol. Save the pellet as the microsomal fraction. To extract microsomal proteins from the pellet, proceed to Step 52.Pause Point The supernatant can be frozen at - 70 °C until required. The pellet may be frozen at - 70 °C and processed at a later date.
35. Points from here (point 35) up to and including point 43 are related to Extraction of nuclear proteinsTiming: 2 hResuspend the pure nuclei (from Step 23) in five volumes of NE buffer with a pipette. This can typically be performed in a 1.5-ml snap-cap microcentrifuge tube.
36. Incubate the nuclei for 30 min with gentle rocking at 4 °C.
37. Lyse the nuclei with 10 passages through an 18-gauge needle.Troubleshooting
38. Centrifuge the lysate at 9,000g for 30 min in a microcentrifuge.
39. Save the supernatant as Nuc-S. This will contain many of the soluble proteins and those not so tightly bound to DNA. Optionally, take a small aliquot to determine the concentration of protein; this will prevent having to thaw the whole sample at a later date.Pause Point The sample and aliquot can be frozen at - 70 °C until needed.
40. Resuspend the pellet in 5 volumes of NET buffer with a pipette and gently rock for 30 min at 4 °C.
41. Extract the nuclei debris by 10 passages through an 18-gauge needle.Troubleshooting
42. Centrifuge at 9,000g for 30 min in a microcentrifuge.
43. Save the supernatant as Nuc-T. This will contain nuclear membrane proteins and other proteins that are tightly bound to DNA, such as histones. Optionally, take a small aliquot to determine the concentration of protein; this will prevent having to thaw the whole sample at a later date.Pause Point The sample and aliquot can be frozen at - 70 °C until needed.
44. Points from here (point 44) up to and including point 51 are related to Extraction of mitochondrial proteinsTiming: 2.5 hResuspend the mitochondrial pellet (from Step 31) in 0.5 ml HDP buffer in a 1.5-ml snap-cap microfuge tube; if 0.5 ml is less than 10 volumes, split the pellet into multiple tubes.
45. Incubate on ice for 30 min.
46. Sonicate the suspension at a high setting to lyse the mitochondria. Keep the mitochondria on ice during sonication and use four 5–10-s bursts with 30-s pauses to prevent sample heating.
47. Microcentrifuge the lysate at 9,000g for 30 min.
48. Save the supernatant as Mito-S fraction, the soluble mitochondrial matrix proteins.
49. Resuspend the pellet in 0.5 ml of ME buffer and incubate for 30 min with gentle rocking.
50. Microcentrifuge at 9,000g for 30 min.
51. Save the supernatant as Mito-M, the mitochondrial membrane proteins.Pause Point Freeze the extracts at - 70 °C.
52. Points from here (point 52) up to and including point 54 are related to Extraction of microsome proteinsTiming: 1.5 hResuspend the microsome pellet (from Step 35) with 0.5 ml of ME buffer and incubate for 1 h with gentle rocking.
53. Microcentrifuge at 9,000g for 30 min.
54. Save the supernatant as Micro, the microsome proteins.Pause Point Freeze the extract at - 70 °C.

Timing

Although the entire protocol timing totals over 9 h, many of the steps can be carried out in parallel such that the entire procedure can be completed in less than 6 h. See Figure 1 for a flow chart and timing coordination.

Troubleshooting

Troubleshooting advice can be found in Table 1.

Table 1: Troubleshooting table

Full table

Anticipated results

After fractionation and extraction, typical protein yields are all in the several mg ml^{- 1} range, as determined by the Bradford assay¹⁰. As we have previously published, each fraction is highly enriched for proteins that are known to be localized to these fractions. Figure 2 shows a western blot of a preparation made from liver blotted for markers of cytosol, nuclei, microsomes (ER) and mitochondria. Also shown are the MudPIT profiles for the same proteins in six different organ preparations. As can be seen, the nuclei and cytosol show the highest degree of purity, with the mitochondria and microsomes being more variable depending on the tissue source. The depth of coverage is sufficient that numerous proteins specific to tissue types are readily identified in their correct subcellular location.

Figure 2: MudPIT profiles and western blot of markers for the four subcellular fractions: nuclei (Nuc), cytosol (Cyto), microsomes (Micro) and mitochondria (Mito).

Markers for cytosol (Transketolase), nuclei (RNA polymerase II large subunit), microsomes (Calreticulin) and mitochondria (F1 ATP synthase beta subunit) were blotted against extracts from liver subcellular fractions. Note the high degree of specificity. Quantification values obtained by densitometry are displayed below each lane. MudPIT profiles from brain, heart, kidney, liver, lung and placenta for these same proteins are shown alongside the western blots. Intensity of yellow corresponds to the relative quantity of protein detected. Note that the cytosol and nuclei show a high degree of purity for most tissues, whereas the microsomes and mitochondria specificity vary depending on the tissue source. B, brain; H, heart; K, kidney; Li, liver; Lu, lung; P, placenta. Reprinted and modified from ref. 5, with permission from Elsevier.

Full size image (30 KB)