Serial Isoelectric Focusing as an Effective and Economic Way to

Oct 8, 2005 - Here, we introduce a novel, effective, and economic method designed ... Sample consumption, time required, and associated costs are redu...
0 downloads 0 Views 531KB Size
Serial Isoelectric Focusing as an Effective and Economic Way to Obtain Maximal Resolution and High-Throughput in 2D-Based Comparative Proteomics of Scarce Samples: Proof-of-Principle Murtada H. Farhoud*,,† Hans J. C. T. Wessels,† Ron A. Wevers,‡ Baziel G. van Engelen,§ Lambert P. van den Heuvel,†,‡ and Jan A. Smeitink† Nijmegen Center for Mitochondrial Disorders at the Department of Pediatrics, Laboratory of Pediatrics and Neurology, and Neuromuscular Center Nijmegen, Department of Neurology, Radboud University Nijmegen Medical Centre, PO Box 9101, 6500HB Nijmegen, The Netherlands Received July 24, 2005

Abstract: In 2D-based comparative proteomics of scarce samples, such as limited patient material, established methods for prefractionation and subsequent use of different narrow range IPG strips to increase overall resolution are difficult to apply. Also, a high number of samples, a prerequisite for drawing meaningful conclusions when pathological and control samples are considered, will increase the associated amount of work almost exponentially. Here, we introduce a novel, effective, and economic method designed to obtain maximum 2D resolution while maintaining the high throughput necessary to perform large-scale comparative proteomics studies. The method is based on connecting different IPG strips serially head-to-tail so that a complete line of different IPG strips with sequential pH regions can be focused in the same experiment. We show that when 3 IPG strips (covering together the pH range of 3-11) are connected head-to-tail an optimal resolution is achieved along the whole pH range. Sample consumption, time required, and associated costs are reduced by almost 70%, and the workload is reduced significantly.

zoom-in immobilized pH gradient strips (IPG) has improved the resolution of 2D significantly.1,2 Using these gel systems, it was possible to resolve more than 10 000 proteins from higher eukaryotic total cell homogenates.3 To further expand the protein profiling capacity of the technique to allow comprehensive analysis of very complex proteomes such as the proteomes of mammalian cells, tissues, and biological fluids, very efficient microscale pre-fractionation methods have been established.4,5 These improvements however are paralleled by an increased workload associated with the need to run several overlapping first dimensions and, when no pre-fractionation is performed, the use of more samples in proportion to the number of used IPG strips. The latter is often a major problem (due to a limited patient material for example). Another limiting factor of this approach is the extra costs associated with the extra material used, in particular when special labeling methods are used for the purpose of quantification (e.g., CyDye fluors). In the present work, we show that serial isoelectric focusing (SIEF), which is based on connecting different IPG strips headto-tail, significantly reduces workload, sample consumption, and costs in 2D-based comparative proteomics.

Keywords: serial isoelectric focusing(SIEF) • two-dimensional gel electrophoresis • DIGE • prefractionation • scarce samples • mitochondria

Materials and Reagents. The serial IEF unit (SIEF unit) used in this work was made in-house and consisted of a Lexan electrophoresis tray resting on a lexan-cooling element (Figure 2). Lexan is a tough, dimensionally stable thermoplastic that has excellent electrical properties (e.g., dielectric strength 1567 kV/mm). The anode, cathode, and connection bars were all made from Lexan. The anode and cathode bars contained further platina electrodes. The electrophoresis tray can accommodate up to 12 serial IPG strips each consisting of up to 3 × 18 cm IPG strips. The anode bar can only be placed at the extreme left side while the cathode bar is movable as well as the connection bars. In this way, (serial) IPG strips of lengths between 10 and 56 cm can be focused in this unit. The cooling element ensures constant temperature in the electrophoresis unit by active cooling. An external thermostatic circulator (not shown) is used to deliver the cooled medium (MilliQ). A removable lid made from Plexiglas (not shown) protects the unit. The plastic unit base contains a safety mechanism that allows a closed circuit to be formed only when the lid covers the unit.

Introduction Comparative proteomics has provided unique opportunities to study genotype-phenotype relations and stimulated the hunt for new disease markers. In 2D-based proteomics, the protein content of a given sample is first separated by twodimensional gel electrophoresis (2D). When coupled to highthroughput mass spectrometry, the separated proteins could be identified with sufficient rates to establish proteome-wide reference maps that could be used in subsequent work to identify differentially expressed proteins. The introduction of * To whom correspondence should be addressed. Tel: +31-(0)24-3614454. Fax: +31-(0)24-3668532. E-mail: [email protected]. † Nijmegen Center for Mitochondrial Disorders, Department of Pediatrics. ‡ Laboratory of Pediatrics and Neurology. § Neuromuscular Center Nijmegen, Department of Neurology.

2364

Journal of Proteome Research 2005, 4, 2364-2368

Published on Web 10/08/2005

Experimental Procedures

10.1021/pr050231a CCC: $30.25

 2005 American Chemical Society

technical notes

Farhoud et al.

Figure 1. Principal of Serial Isoelectric focusing (SIEF). IPG strips of sequential, minimally overlapping, pH gradients are connected head-to-tail to construct a “serial IPG strip”. The connections between the different strips are facilitated by filter papers wetted with rehydration buffer and kept during the whole SIEF session under light pressure delivered by the connection bars. The sample can be added to the rehydration buffer of any (or all) strips but can also be applied via conventional cup-loading.

Figure 2. Serial IEF unit (SIEF unit): The in-house made SIEF unit can accommodate up to 12 “serial IPG strips” each consisting of 3 × 18 cm IPG strips. The IPG strips are placed in the grooves on the electrophoresis tray (1). The anode bar (5) can only by placed at the extreme left side while the cathode bar (3) is movable as well as the connection bars (4). In this way (serial) IPG strips of lengths between 10 and 56 cm can be focused in this unit. The cooling element (2) ensures constant temperature in the electrophoresis unit by active cooling. An external thermostatic circulator (not shown) is used to deliver the cooled medium (milliQ). The unit is protected by a removable lid (not shown). The unit base (6) contains a safety mechanism that allows a closed circuit to be formed only when the lid covers the unit.

Human heart mitochondrial fraction was kindly provided by Dr. Leo Nijtmans (NCMD, Radboud University Nijmegen Medical Centre). Thiourea and Iodoacetamide were from ICN Biomedicals, the cross-linker Piperazine Diacrylamide was from

Biorad, and G-250 Serva Blue was from Serva. With the exception of the in-house made electrophoresis unit, all other chemicals, materials and equipment used for sample preparation and two-dimensional gel electrophoresis were from GE HealthCare. Sample Preparation. A 200-µL human heart mitochondrial fraction (20 mg/mL) was mixed with 600 µL solubilization buffer (8 M urea, 2 M Thiourea, 4% CHAPS, 50 mM DTT, 1% (v/v) nuclease mix), vortexed thoroughly, and incubated at room temperature for 30 min. The sample was cleared by centrifugation at 12 000 × g and 4 °C for 30 min. The supernatant was divided in portions of about 50 µL per portion and frozen in liquid nitrogen prior to storing at -80 °C. Protein concentration was determined with 2D-Quant protein assay using BSA as standard (GE HealthCare). Preparing the IPG Strips. All IPG strips used in this work were 18 cm strips. IPG strips of pH 5.0-6.0, 6.2-7.5, and 3-10 NL were used in conventional IEF, while IPG strips of pH 3-5.6, 5.3-6.5, and 6-11 were used in serial IEF. In serial IEF, the rehydration buffer of the anodic strip (pH 3-5.6) contained 8 M urea, 2 M thiourea, 2% CHAPS, 5% glycerol, 35 mM DTT and 0,4% IPG Buffer pH 3-5.5. A 300-µg portion of human heart mitochondrial fraction was added to the rehydration buffer of the middle strip (pH 5.3-6.5), which contained 8 M urea, 2 M thiourea, 2% CHAPS, 5% glycerol, 35 mM DTT, and 0,5% IPG Buffer pH 5.5-6.7 (rehydration loading). The rehydration buffer of the cathodic strip (pH 6-11) contained 8 M urea, 2 M thiourea, 2% CHAPS, 5% glycerol, 3.5% (v/v) DeStreak Agent, and 0,5% IPG buffer pH 6-11. 1. In conventional IEF, the rehydration buffer contained always 300 µg of human heart mitochondrial fraction in 8 M urea, 2 M thiourea, 2% CHAPS, 5% glycerol, and 0,5% IPG Buffer. The rehydration buffer for the IPG strips of pH 5.0-6.0 and 3-10 NL contained also 35 mM DTT while that of the IPG strips of pH 6.2-7.5 contained 3.5% DeStreak Agent instead. IPG strips were incubated overnight in the corresponding rehydration buffer. Serial IEF (SIEF). The rehydrated strips were prepared for serial connection by cutting with a razor blade the extra packing material from one or both ends of the strip depending on its location: the anodic (left-side) strip is cut at the right end, the middle strip is cut at both sides, and the cathodic (right-side) Journal of Proteome Research • Vol. 4, No. 6, 2005 2365

technical notes

Serial Isoelectric Focusing (SIEF)

Figure 3. Optimal resolution is achieved along the whole pH range of 3-11 in the “serial IPG strip”. In the lower panel, 300 microgram of human heart mitochondrial fraction was separated along 3 serially connected IPG strips covering the pH region of 3-11 as described in experimental procedures. The 18 cm IPG strips used here were from left to right of pH 3-5.6, 5.3-6.5, and 6-11. The sample was added to the rehydration buffer of the middle strip. 11% SDS-PAGE was used in the second dimension. The 2D gels, which were stained with colloidal coomassie, are shown in false colors. For comparison a standard 2D of 300 microgram of human heart mitochondrial fraction separated along an 18 cm 3-10 NL IPG strip is shown in the upper panel. The false colors used for the single 2D in the upper panel correspond to those used for the different 2D’s in the lower panel.

strip is cut at the left side. This is done to remove the extra packing material, which usually protrudes at both sides of the IPG strip, and to expose a sticky-end of the gel material. To ensure that the exposed ends of the gel strips will stick to each other and form a stable connection, it is helpful to make the cuts at a constant angle (preferably 45°) relative to either the anode- or the cathode-side of the construction. To prevent drying of the strips during this preparative step, the strips were overlaid with a few milliliters of DryStrip Cover fluid (GE HealthCare). The IPG strips were positioned in the grooves always with the gel side up and the anodic side (lowest pHside) to the left. Starting with the anodic IPG strips, the prepared strips were positioned in the following order: anodic, middle, and cathodic. Electrode pads (GE HealthCare) wetted with rehydration buffer were placed on top of gel at the anodic and cathodic extremes of the assembled strips (contact with electrodes) and on top of each connection between two neighboring strips. This was done to facilitate movement of proteins across the connections. To ensure proper contact between the connected strips a light pressure was maintained at each connection using the connection bars. When the IPG strips were properly positioned and the electrode and connection bars were installed, 500 mL DryStrip cover fluid was used to cover all IPG strips completely. SIEF was performed at a constant temperature of 18 °C and a maximum current of 2366

Journal of Proteome Research • Vol. 4, No. 6, 2005

50 µA per used groove. SIEF was started with a step of constant voltage at 300 V for 3 h followed by a linear gradient from 300 V to 14 000 V in 12 h and finally a step of constant voltage at 14 000 V for 10 h (ca. 200 kVhr at finish). Conventional IEF. Conventional IEF was performed essentially as described elsewhere2 but using the SIEF unit. Separate IEF sessions were performed using IPG strips of pH 5.0-6.0, 6.2-7.5, and 3-10 NL. The rehydrated IPG strips were always positioned with the gel side up. IEF was started with a step of constant voltage at 300 V for 3 h followed by a linear gradient from 300 to 8000 V in 6 h and finally a step of constant voltage at 8000 V for 6 h (ca. 70 kVhr at finish). Second Dimension. The second dimension was performed as described elsewhere2 and modified as in ref 6: the 11% homogeneous polyacrylamide gels, which were cross-linked by piperazine diacrylamide, contained no SDS. The electrophoresis buffer contained 0.2% SDS to maintain the necessary negative charge on proteins. The 2D gels were stained with colloidal Coomassie blue as described elsewhere.7

Results and Discussion To increase the resolution power of 2D, we tried to connect multiple IPG strips head-to-tail so that a given sample could be separated over all the different strips simultaneously (Figure 1). In this way a complete line of different IPG strips with

technical notes

Farhoud et al.

Figure 4. No protein loss is observed at the junctions between neighboring strips. A 300-µg portion of human heart mitochondrial fraction was separated along zoom-in single IPG strips of pH 5.0-6.0 (A) and pH 6.2-7.5 (C). Corresponding regions on the “serial IPG strip”, which contain the junctions (indicated by solid white lines), are shown in B and D, respectively. 11% SDS-PAGE was used in the second dimension. The 2D gels were stained with colloidal coomassie (shown as negatives). Arrowheads indicate matching spots on corresponding images.

sequential pH regions could be used in the same experiment. Besides decreasing the workload associated with running the first dimension, the same original amount of sample usually used with a single strip could be separated over all connected strips. Precious (labeled) samples could be used more efficiently in this way, which would enable very limited amounts of sample to be analyzed thoroughly with 2D. To test this idea, we developed an electrophoresis unit capable of accommodating up to 3 × 18 cm IPG strips connected serially to each other. In this serial IEF (SIEF) unit (Figure 2) up to 12 assemblies (each consisting of up to 3 × 18 cm IPG strips) can be run simultaneously. As proof-of-principle, we separated 300 µg of human heart mitochondrial fraction on 3 serially connected IPG strips covering together the pH region of 3-11. The 18 cm IPG strips used here were, from left to right, of pH 3-5.6, 5.3-6.5, and 6-11, which ensures a minimal overlap between two neighboring strips. The heart-

muscle sample was applied to the middle strip (pH 5.3-6.5) by rehydration loading. As can be seen in the lower panel of Figure 3, optimal resolution was achieved along the whole pH region of 3-11. For comparison the upper panel of Figure 3 shows a typical 2D pattern obtained for this sample when an 18 cm 3-10 NL IPG strip is used in the first dimension. The overall increased resolution is quite evident (more than 2000 spots were detected in the 3 lower images compared to less than 900 spots in the upper image). The duplicate pH regions between 2 neighboring strips, which are minimal in this work, seem not to interfere with the movement of proteins. Similar results were obtained when the duplicate regions were trimmed (data not shown). Trimming duplicate regions, however, might be necessary when the overlap between the neighboring strips is much greater than in the combination used in this work. To investigate whether there is any protein loss at the junctions between two neighboring strips in the serial IPG strip, Journal of Proteome Research • Vol. 4, No. 6, 2005 2367

technical notes

Serial Isoelectric Focusing (SIEF)

also 300 µg of human heart mitochondrial fraction was separated along zoom-in IPG strips of pH 5.0-6.0 and 6.2-7.5 using conventional IEF. The former pH region is comparable with that around the junction between the IPG strips of pH 3-5.6 and pH 5.3-6.5, while the latter one is comparable with that around the junction between the IPG strips of pH 5.3-6.5 and pH 6-11. Figure 4 shows that separation of the same sample along the strips pH 5.0-6.0 and pH 6.2-7.5 has comparable spot patterns to the corresponding regions of the “serial IPG strip”, which excludes any protein loss at the junctions. In this study, we show that the chosen combination (pH 3-5.6, 5.3-6.5, and 6-11) results in optimal resolution of the protein content of human heart mitochondrial fraction when the sample is applied to the middle strip. Other combinations are still possible and are currently under investigation but in general we should expect that the suitability of a given combination will depend on type of sample. Also, the application of sample either by rehydration loading or by cup loading is under further investigation to determine the optimal conditions and, more importantly, the loading capacity of the system. The linearity of the whole “serial IPG strip” will depend not only on the linearity within the individual IPG strips but also on that of the used strips with respect to each other. The “serial IPG strip” described in the present work is nonlinear at both the acidic and the alkaline sides: on the acidic side due to the nonlinear 3-5.6 IPG strip and on the alkalaine side due to the narrow-range 6-11 strip compared to the zoom-in 5.3-6.5 strip (5 pH units/18 cm for the former compared to 1.2 pH units/18 cm for the latter). Since more users will try this approach in the future, as we believe, users could request new types of IPG strips that suit their optimal serial IEF conditions. Optimal resolution and coverage are important criteria in comparative proteomics. The unique opportunity provided by this method to prepare the different regions of the “serial IPG strip” separately allows ultimate fine-tuning of the buffering conditions used per pH region. In this way, different concentrations of IPG Buffer can be applied to the different pH regions as needed. In the same way, the reducing agent DTT, which is only active at pH < 7, could be applied to the acidic pH region while organic disulfides such as DeStreak Agent (GE HealthCare), which are active only at pH > 7, could be used in the basic pH region. The concomitant use of DTT and DeStreak agent, which seems to work, is under further investigation at this moment to define optimal conditions and drawbacks if any. An important advantage of this novel method is that the sample is fractionated over the connected strips during IEF, which could render prefractionation prior to IEF unnecessary. In 2D-comparative proteomics of scarce samples and especially when large numbers of samples are to be analyzed SIEF will increase the throughput substantially. It will also avoid any experimental variations that could be introduced during prefractionation of the individual samples. Another important

2368

Journal of Proteome Research • Vol. 4, No. 6, 2005

advantage of this method is that it is very economic. In 2Dbased comparative proteomics, DIGE is the method of choice since an internal standard is included in the experimental setting, which is used to correct for possible experimental variations. DIGE however remains very expensive (almost $300 per gel) and is still far away from being a routine application. To our knowledge, so far DIGE has only been used with either broad range IPG strips or single zoom-in gels. In both cases, most of the costs invested in the label are simply wasted because only part of the sample is being separated (optimally). SIEF provides the solution for this problem because the labeled sample can be separated simultaneously over more than one strip and is used more efficiently in this way. According to the experimental setting described here, the average label cost per gel is reduced to one-third. In ongoing research and using the same experimental setting, we were able to use less than 1.6 nmol each of Cy2, Cy3, and Cy5 to label less than 300 µg each of patient and control samples to perform a complete DIGE experiment along the entire pH range of 3-11 utilizing a total number of 12 gels. In summary, we deliver here the proof-of-principle for serial isoelectric focusing (SIEF). By applying serial isoelectric focusing, the resolution and throughput of 2D are increased significantly while sample consumption, workload, experimental variations, and costs are kept at minimum. In this way significant numbers of samples can easily be included in 2Dbased comparative proteomics. This will enhance the possibility of drawing meaningful conclusions, especially when a large number of (scarce) samples from pathological and control conditions are to be compared.

Acknowledgment. Part of this work was supported by the Neuromuscular Centre Nijmegen. Also, part of this work was supported by the Sixth Framework Program Priority 1, project titled “Rational treatment strategies combating mitochondrial oxidative phosphorylation (OXPHOS) disorders (Eumitocombat), Contract No 503116. We would like to thank Ernest Prins, Stef Mientki, and Bernard Janssen from ID-SPON, Radboud University Nijmegen, where the SIEF unit was made. References (1) Hoving, S.; Voshol, H.; van Oostrum, J. Electrophoresis 2000, 21 (13), 2617-2621. (2) Gorg, A.; Obermaier, C.; Boguth, G.; Harder, A.; Scheibe, B.; Wildgruber, R.; Weiss, W. Electrophoresis 2000, 21 (6), 1037-1053. (3) Fey, S. J.; Larsen, P. M. Curr. Opin. Chem. Biol. 2001, 5 (1), 2633. (4) Zuo, X.; Speicher, D. W. Proteomics 2002, 2 (1), 58-68. (5) Herbert, B.; Righetti, P. G. Electrophoresis 2000, 21 (17), 36393648. (6) Hochstrasser, D. F.; Patchornik, A.; Merril, C. R. Anal. Biochem. 1988, 173 (2), 412-423. (7) Candiano, G.; Bruschi, M.; Musante, L.; Santucci, L.; Ghiggeri, G. M.; Carnemolla, B.; Orecchia, P.; Zardi, L.; Righetti, P. G. Electrophoresis 2004, 25 (9), 1327-1333.

PR050231A