Multidimensional proteomic analysis of proteolytic pathways involved in cell cycle control

 

Michael W. Schmidt1,2, Aruna Jain2, Dieter A. Wolf1,2

 

1Department of Cancer Cell Biology, 2Kresge Center Proteomics Facility, Harvard School of Public Health, Boston, MA 02115

 

 

 

 

1. Introduction

 

Many cell cycle transitions are controlled by ubiquitin-mediated proteolysis of key cell cycle regulatorr (1). The ubiquitin system targets substrates to the proteasome by attaching a polyubiquitin chain (2). The traditional ubiquitin transfer reaction involves a minimum of three enzymes: E1, which mediates the ATP-dependent activation of ubiquitin, and the E2 ubiquitin conjugating enzyme (UBC), which, together with an E3 ubiquitin ligase, transfers ubiquitin to the target protein. Despite the successful identification of many ubiquitin ligases only few of their substrates are known. This is because ubiquitin ligases share conserved motifs, while substrates seem to have little in common other than critical lysine residues.

Whereas systematic approaches to identifying components of ubiquitin ligases have been fruitful (3; 4; 5), to our knowledge, no systematic approaches have been employed to revealing their substrates. Here we describe a protocol for multidimensional proteomic analysis to systematically exploit a conserved biochemical rather than structural feature of ubiquitin ligase substrates: their delayed degradation and hence accumulation in fission yeast ubiquitin ligase mutants.

In a simplified description, protein lysates prepared from wild-type cells and from proteolysis mutants are separated by two-dimensional gel electrophoresis (2DGE), digital gel images are obtained, and proteins of higher abundance in proteolysis mutants are identified by image analysis and liquid chromatography coupled with tandem mass spectrometry (LC/MS/MS). The identified proteins can then be further validated biochemically with in vitro ubiquitination assays employing purified ubiquitin ligases.

Our initial studies have shown that the dynamic range of protein abundance is far too wide, even in a simple organism such as fission yeast, to enable a comprehensive proteome analysis by comparing total cell lysates by 2DGE (see for example Fig. 2C). We have therefore established a highly reproducible chromatographic prefractionation scheme that is outlined in Fig. 1. Total cell lysate is prepared from S. pombe cells by bead lysis and chromatographed on an anion exchange column. Three pooled fractions resulting from elution with a linear salt gradient and containing equal amounts of protein (~ 3.5 mg) are collected. Fractions are then precipitated with 10% trichloroacetic acid in acetone, resuspended, and analyzed by 2DGE. Gels are stained and analyzed with specialized 2D imaging software to reveal differentially accumulating proteins.

Fig. 2 shows results for a sample taken through the entire fractionation (Fig. 2A) and 2D analysis (Fig. 2C). A largely non-overlapping protein pattern can be revealed (Fig. 2B and C). In addition, many low abundance proteins not visible in the total cell lysate (Fig. 2C) are revealed after sample prefractionation. Comparative analysis of averaged gels from independent duplicate preparations of wild-type and mutant cell lysates enables the reliable identification of differentially accumulating proteins (Fig. 3).

 

Although 2DGE has continuously improved over the past several years, there are numerous frequently cited limitations many of which have been successfully circumvented in the described protocol:

(1.) The problem of the wide dynamic range of protein abundance that notoriously obscures low abundance proteins is addressed by chromatographic prefractionation. This procedure currently allows us to detect a minimum of 1800 independent features on our gels that are spread over a wide range of molecular weights and pIs (see Fig. 2C). This number represents approximately 40% of the theoretical fission yeast proteome. Although no other techniques currently afford a greater coverage, penetrance of the fission yeast proteome can be further increased by using narrow range pH gradient strips or further chromatographic fractionation of the flow-through fraction.

(2.) Protein losses during sample preparation can result form inefficient extraction, precipitation, and resolubilization. While the standard extraction buffer, containing non-ionic detergent, is difficult to improve, we have optimized the conditions for protein precipitation and resolubilization by testing numerous different agents and methods. While losses are impossible to avoid completely, we estimate that protein recovery with the described protocol is greater than 90% (data not shown).

(3.) The problem of reproducibility. We have found in numerous experiments that due to the use of an automated FPLC system, the reproducibility of the chromatographic steps is extremely high. Using the same lot of immobilized pH gradient strips for IF, consistent protein patterns have been obtained when running samples in duplicates (Fig. 3). In addition, averaging several gels from independently prepared samples further reduces gel-dependent variability.


2. Materials:

2.1: Lysate preparation

1.Yeast YE media (5 g/l Bacto Yeast Extract, 30 g/l Dextrose). Autoclave.

2.7.5 g/L adenine (50 x stock solution)

3.7.5 g/L uracil (100 x Stock solution)

4.Incubator

5.Yeast lysis buffer (prepare fresh: 25mM Tris-HCl pH 8.8, 0.5 % Triton x-100, 10 mg/ml leupeptin, 10 mg/ml pepstatin, 15 mg/ml aprotinin, 1 mM PMSF)

6.Bead beater and cup set (Biospec Products)

7.0.5 mm silica beads (Biospec Products)

8.Centrifuge RC 5B Plus (Sorvall)

9.Fixed angle rotor SS-34 (Sorvall)

10.1 L centrifuge buckets (Sorvall)

11.50 ml oak ridge screw cap polypropylene centrifuge tubes (Nalgene)

12.Dc-protein assay kit (Bio-Rad)

2.2: Prefractionation

1.Ätka FPLC system (Amersham Biosciences)

2.HiTrap Q HP 1 ml column (Amersham Biosciences)

3.Buffer A (25 mM Tris-HCl pH 8.8, 1 mM DTT)

4.Buffer B (25 mM Tris-HCl ph 8.8, 1 M NaCl, 1 mM DTT)

5.1.5 ml micro centrifuge tubes (VWR)

6.Glass centrifuge tubes 30 ml (Corex)

7.10 x DNAse/RNAse mix (1 mg/ml DNAse 1, 0.25 mg/ml RNAse A, 50 mM MgCl2). Freeze in aliquots.

8.Acetone / 13.3 % TCA / 0.093 % b-mercaptoethanol

9.Sample loading buffer (6):

a.7M urea (Genomic Solutions)

b.2 M thiourea (Sigma-Aldrich)

c.2 % CHAPS (EM Science)

d.2 % ampholytes pH 3-10 (Genomic Solutions)

e.65 mM DTT (Research Products Int.)

f.Serdolit MB-1 (Serva/Crescent Chemical Co.)

g.0.01 % bromophenol blue (Sigma-Aldrich)

10.10 % SDS mini gel

11.5 x SDS sample buffer

a.2M tris-HCl pH 6.8

b.25 % b-mercaptoethanol

c.10 % SDS

d.50 % glycerol

e.10 % H2O

f.5 %  bromophenol blue (0.1 % sol.)

12.SDS gel running buffer (tris-glycine pH 8.3, 10 % SDS)

2.3. 2D Analysis

1.Immobilized dry strips NL, pH 3-10, 18 cm (Amersham Biosciences)

2.Whatman filter paper 20 cm x 20 cm

3.30 cm x 30 cm plastic bag (Zip-Lock)

4.Investigator pHaser isoelectric focusing system (Genomic Solutions)

5.Investigator 5000 programmable power supply (Genomic Solutions)

6.Non-conducting oil for IPG (Genomic Solutions)

7.pHaser electrode wicks (Genomic Solutions)

8.Equilibration trays (Genomic Solutions)

9.Equilibration buffer 1:

a.6 M urea (Genomic Solutions)

b.375 mM tris-HCl pH 7.4 (ICN Biomedicals)

c.2 % SDS (Sigma-Aldrich)

d.2 % glycerol (ICN Biomedicals)

e.2 % DTT (Research Products International)

10.Equilibration buffer 2:

a.6 M urea (Genomic Solutions)

b.375 mM tris-HCl pH 7.4 (ICN Biomedicals)

c.2 % SDS (Sigma-Aldrich)

d.2 % glycerol (ICN Biomedicals)

e.2.5 % iodoacetamide (Sigma-Aldrich)

11.Upper (cathode) gel running buffer

a.200 mM tricine (Sigma-Aldrich)

b.200 mM tris-Base (Sigma-Aldrich)

c.0.4 % SDS (Sigma-Aldrich)

12.Lower (anode) gel running buffer:

25 mM tris-acetate pH 8.3 (Sigma-Aldrich/ EM Science)

13.Investigator 2-D running system (Genomic Solutions)

14.Pre-cast tricine gels (Genomic Solutions)

15.Gel gaskets (Genomic Solutions)

16.SDS-PAGE standard broad range markers (Bio-Rad)

17.Gel fixing solution

a.40 % methanol (EM Science)

b.10 % glacial acetic acid (EM Science)

18.Staining solution

a.10 % coomassie brilliant blue G-250 (Baker)

b.50 % methanol (EM Science)

c.10 % glacial acetic acid (EM Science)

19.Destaining solution

a.25 % methanol (EM Science)

b.5 % glacial Acetic Acid (EM Science)

20.Glass container

21.Rocking device

22.Flat bed scanner (Hewlett Packard)

23.Software (Phoretix2D Professional, Nonlinear Dynamics)

 

3. Methods:

 

3.1. Sample Preparation

3.1.1 Yeast Culture

1.Autoclave 1 L YES medium in a 2 L glass flask and add adenine and uracil when medium is cooled down.

2.Inoculate a single colony from an agarose plate (using a sterile loop) in 10 ml YES medium (YE medium plus supplements) and incubate for 24 h at 30 °C under agitation to produce a pre-culture.

3.Inoculate 2.5 ml of the pre-culture in 1 L YES and grow at 30 °C under agitation until OD595 1.5 is reached.

4.Spin culture for 15 min at 2000g in 1 L buckets and discard supernatant.

5.Resuspend pellet in 25 ml tris pH 7.5 and transfer to 40 ml tubes.

6.Spin 5 min at 2000g and remove supernatant.

7.Freeze pellet immediately at –80 ° C

3.1.2. Lysate Preparation (see Note1)

1.Thaw pellet of 1 L culture quickly (use warm water bath).

2.Add 25 ml chilled yeast lysis buffer (section 2.1.3.)

3.Transfer to bead beater medium cup and grind  ~ 6 x 1 min (see Note 2).

4.Transfer bead-cell homogenate in chilled glass beaker and separate cell homogenate from beads by aspiration with a 5 ml pipette.

5.Spin lysate in SS-34 rotor in prechilled Sorvall centrifuge at 20,000 rpm (48,000 g) for 40 min (see Note 3)

6.Determine protein concentration immediately using the Dc-protein assay.

3.1.3. Prefractionation (see Note 4)

1.Equilibrate the HiTrap Q 1ml column with 5 column volumes (CV) buffer A.

2.Apply 30 mg of cell lysate to column and collect flow through in 1 ml fractions.

3.Wash out unbound sample with 5 CV buffer A.

4.Elute bound sample with a linear gradient (0 to 100 % Buffer B) over 20 CV and collect 1 ml fractions (see Fig. 2A).

5.Clean column with 5 CV buffer B.

6.Reequilibrate with 5 CV buffer A.

7.Determine protein concentration in each 1 ml fraction and pool in larger fractions containing 3.5 mg each into 30 ml Corex tubes.

8.DNAse/RNAse treatment: add 0.1 volumes of the 10 x DNAse/RNAse mix and incubate for 10 min at 4° C.

3.1.4. Precipitation in Trichloracetic Acid in Acetone (see Note 5).

1.Add 3 volumes of the chilled (-20° C) stock solution (see section 2.2.8.) to the sample (final concentration is 10 % TCA and 0.07% b-mercaptoethanol in acetone) and incubate for 1.5 hours or over night at –20 ° C.

2.Spin samples 15 min at 2,500g (Sorvall SS-34) at -20° C.

3.Discard as much of the supernatant as possible (to avoid TCA-contamination) and wash pellet in chilled (-20° C) acetone containing 0.07 % b-mercaptoethanol.

4.Spin samples 15 min at 2,500g (Sorvall SS-34) at -20° C.

5.Remove all acetone and dry pellets at room temperature (~1.5 h).

6.Resuspend pellets in 1 ml sample loading buffer (see section 2.2.9.). (see Note 6).

7.Store sample at –80 °C.

 

3.2. Two-dimensional Gel Electrophoresis

For the 2D analysis, it is very convenient to process several samples in parallel.

3.2.1. Isoelectric Focusing

1.Thaw the samples to 30 °C, vortex, then briefly spin down any insoluble particles

2.Apply 400 ml (~1.5 mg) sample to a focusing tray and place an IPG-strip with the gel side facing down on the sample film (see Note 7).

3.Place a moist Whatman filter paper under the focusing tray and put it in a zip-lock plastic bag and incubate until the sample is completely absorbed by the strip (usually over night).

4.Focus the strip until 100,000 Vh are reached (see Note 7). Running parameters: run-time 24 h, max. Voltage 5,000 V, holding voltage 125 V, current per gel 80 mA.

3.2.2. Second Dimension

Before starting the equilibration it is convenient to set up the gel running tanks, the gels, and the gaskets as described in the product instruction manual. In order to speed up the process, the upper and the lower buffers can be prepared and cooled to 4° C in advance.

1.Equilibrate the IPG strip (gel side down) in 10 ml equilibration buffer 1 (see section 2.3.9.) at room temperature for 10 min under gentle agitation. Remove all buffer 1 before continuing.

2.Equilibrate the IPG strip (gel side down) in 10 ml equilibration buffer 2 (see section 2.3.10.) at room temperature for 10 min under gentle agitation. Remove all buffer 2.

3.Place the strips, plastic side facing backwards glass of the gel without trapping any air between the gel surface and the IPG strip.

4.Running parameters: run-time 8 h, max. voltage 500 V, max power/gel 20,000 mW.

5.Carefully take the gel out and place it in fixing solution for 2 h under agitation.

6.Remove the gel from fixing solution and place it in staining solution for 2 h under agitation (see Note 8).

7.Place the gel in destaining solution until gel is completely transparent.

3.2.3. Image Acquisition

In order to analyze the protein spots in a quantitative manner, it is necessary to obtain digital images. This is done in a regular flat bed scanner, which generates tiff that can then be analyzed using spot detection and quantitation software (Phoretix 2D).

1.Carefully remove the gel from the destaining solution, and place it between two layers of transparent plastic sheet protectors. Make sure not to trap any air bubbles between the gel and the plastic foil. The use of a light box can be helpful.

2.Scan the gel and make sure that no liquid gets on the scanner surface or between the gel and the scanner lid because this may lead to artifacts.

3.Save the image as a tiff file.

 

3.3 Data analysis

For increased accuracy, the sample preparation and the 2D gel electrophoresis should be repeated two to three times each time starting with an independent colony. Using specialized imaging software (2D Phoretix from Nonlinear Dynamics), independent replicate gels can be averaged, resulting in a virtual 2D gel (Fig. 3). Averaged gels from wild-type and mutants strains can then be compared for expression differences (Fig. 3).

 

 

4. Notes

1. Lysate preparation. In order to minimize protein degradation, it is important to chill the sample at all steps prior to denaturation in sample loading buffer. In case the sample warms up (e.g. during bead lysis), stop and chill before continuing. Try to do the sample preparation in one continuous workflow without any interruptions. In our experience, consistent expeditious sample preparation during all steps before isoelectric focussing is key to ensuring reproducibility of the procedure. Repeated 2D gel analyses of the same sample never revealed major variations, whereas comparative analysis of independently prepared samples showed variability, mainly due to differences in sample handling before isoelectric focussing.

2. Bead lysis. Check state of lysis by phase contrast microscopy during a chilling cycle. Broken cells loose their typical bright halo and appear dark (simulateously compare to a sample of unlysed cells, if in doubt). Lysis is complete, when ~60 % to 80 % of cells are broken.

3. Protein quantitation. It is convenient to perform the protein quantitation during the centrifugation step, in order to ensure quick processing.

4. Anion exchange chromatography. The flow rate for the entire run is 1 ml/min. In order to be able to exactly compare several individually prepared samples, always normalize the protein concentration and the sample volume before application onto the column. After start of the elution, make sure the fraction collector is set to collect all fractions to the end of the entire run. The void volume of tubing may lead to a shift of eluting proteins towards later fractions relative to the UV trace of protein elution.

5. TCA/acetone precipitation. Add b-mercaptoethanol to the chilled (-20° C) solution fresh. Do not use any plastic devices (tubes, pipettes) in each of the following steps, since this may result in background peaks during subsequent mass-spectrometric protein identification.

6. 2D sample loading buffer. It is important to resuspend the protein pellet completely. Any insoluble protein causes problems and low reproducibility. In order to achieve complete resuspension, place the samples in a 30 °C incubator under agitation. Do not heat the sample above 37° C as urea can cause protein modifications. At this point, it is recommended to analyze 5 ml of the samples on a regular SDS-mini-gel before continuing (see Fig. 2B). This step helps to monitor the quality and quantity of the sample prior to 2D analysis.

7. Isoelectric focusing. Avoid air bubbles under the IPG gel strip. Work expeditiously, since evaporating buffer can lead to urea or protein precipitation. The isoelectric focusing is a very critical step. Poor focusing is a major reason of "bad" gels and can be caused by the presence of salt in the sample. Make sure 100,000 Vh are reached before continuing the second dimension.

8. Protein quantities in this protocol are adjusted such that proteins can be visualized efficiently with Coomassie Brillian Blue. Silver staining is not recommended for this protocol, because of its nonquantitative characteristics. Sypro Ruby can be used instead of Coomassie Brillian Blue, but requires a laser imager and a robotic spot picker to excise protein spots form the gels.

References

 

1.         King, R. W., Deshaies, R. J., Peters, J. M. & Kirschner, M. W. (1996). How proteolysis drives the cell cycle. Science 274, 1652-9.

2.         Hershko, A. & Ciechanover, A. (1998). The ubiquitin system. Annu Rev Biochem 67, 425-79.

3.         Cenciarelli, C., Chiaur, D. S., Guardavaccaro, D., Parks, W., Vidal, M. & Pagano, M. (1999). Identification of a family of human F-box proteins. Curr Biol 9, 1177-9.

4.         Winston, J. T., Koepp, D. M., Zhu, C., Elledge, S. J. & Harper, J. W. (1999). A family of mammalian F-box proteins. Curr Biol 9, 1180-2.

5.         Regan-Reimann, J. D., Duong, Q. V. & Jackson, P. K. (1999). Identification of novel F-box proteins in Xenopus laevis. Curr Biol 9, R762-3.

6.         Gorg, A., Obermaier, C., Boguth, G., Harder, A., Scheibe, B., Wildgruber, R. & Weiss, W. (2000). The current state of two-dimensional electrophoresis with immobilized pH gradients. Electrophoresis 21, 1037-53.

 

 

 

Figure Legends

 

Fig. 1 Flow chart of multidimensional proteome analysis of fission yeast proteolysis mutant. Crude cell lysate from wild-type and mutant cells is fractionated by ion exchange chromatography, and the lueate is pooled into three fractions. The three fractions are analyzed by two-dimensional gel electrophoresis and corresponding gels derived from wild-type and mutant fractions are compared by image analysis. Differentially accumulating proteins are identified by tandem mass spectrometry.

 

Fig. 2 Chromatographic fractionation. (A) Total S. pombe lysate was fractionated by anion exchange chromatography. Fractions eluting with a linear salt gradient were pooled into three fractions of equal protein content. (B) Fractions were precipitated with TCA in acetone, resuspended in 2D sample buffer, and analyzed by one-dimensional gel electrophoresis to document sample quantity and quality. Total cell lysate is shown as a comparison. (C) The three fractions prepared in (B) were analyzed by 2DGE on nonlinear pH 3-10 strips. Gels were stained with Coomassie Brilliant Blue and scanned.

 

Fig. 3 Multidimensional proteome analysis. Two independently prepared fractions #2 from wild-type and mutant S. pombe were analyzed by 2DGE. Replicate gels were scanned, averaged with imaging software, and compared by spot matching and spot densitometry. Proteins accumulating to at least 3-fold higher levels in either wild-type or the mutant yeast are indicated.

 

Fig. 1
 
 

Fig. 2
 

Fig. 3