Laboratory Water Quality Affects Protein Separation by 2D Gel Electrophoresis 2D gel electrophoresis remains a dominant technique in proteomics. Obtaining high quality gels requires careful and tedious work, and the use of high quality reagents including the water used in every step of the experiment is critical. Introduction Effective protein separation by 2D gel electrophoresis is characterized by well-distributed, dark, non-overlapping spots that have well-defined borders. The gel should be free of streaks and background staining. 1 Well-resolved spots are important, not only for isolating proteins for downstream analysis, but also for accurate imaging, in order to detect differences in specific proteins between two conditions. For example, 2D gels should accurately reflect protein changes when comparing samples from healthy and diseased individuals. 2D gel electrophoresis is labor-intensive, expensive, and time-consuming. Furthermore, the slightest variations in sample preparation, electrophoresis procedure, staining, and data acquisition can affect 2D separation results. 2 One of the simplest steps to ensure reproducible, high quality 2D gels is to use only the highest purity reagents, including water. Bottled water and water purification systems are two common sources of high purity water in laboratories. A well-designed purification system delivers ultrapure water that contains very few ions (resistivity 18.2 MΩ cm) and organic compounds (less than 5 ppb of total oxidizable carbon, TOC). This is achieved by combining reverse osmosis (RO), electrodeionization (EDI), ion exchange resins, activated carbon, and UV photooxidation. Point-of-use (POU) cartridges placed at the end of the purification chain customize ultrapure water to specific laboratory needs. Examples of POU cartridges are ultrafilters and 0.22 µm membrane filters. This study has two parts. In the first part, gels prepared with freshly delivered ultrapure water from a purification system equipped with an ultrafilter were compared with gels prepared using bottled water. In the second part, a water purification system was equipped with two different POU cartridges, and we used water from each source to determine whether the choice of POU purifier affects 2D electrophoretic separation. 1
Fresh ultrapure water versus bottled water for 2D gel electrophoresis In the first step, we compared sterile bottled water (Baxter), and freshly delivered ultrapure water. Ultrapure water was delivered from a Milli-Q Synthesis system (Merck Millipore) with an internal ultrafilter. This system delivers ultrapure water with resistivity of 18.2 MΩ cm and TOC of less than 10 ppb. The system was fed with pure water from an Elix system (Merck Millipore), which produced pure water from tap water using RO and EDI. To compare the performance of the two water types for 2D gel electrophoresis, we prepared parallel 2D gels of E coli extracts (see Figure 1A for experimental schematic). For the first dimension, a strip with a linear ph range of 3-10 was used. Separation in the second dimension was carried out in a 12% SDS-polyacrylamide gel. Proteins were visualized using Coomassie blue or silver stain solutions. Coomassie blue is the most common staining procedure for gels. Silver staining is at least 100 times more sensitive, and therefore more sensitive to the effects of contaminants. The two different sources of water were used in preparing rehydration and equilibration buffers for the first dimension, the 12% SDS-polyacrylamide gel and separation buffer for the second dimension, and in preparing the staining solutions. Figure 1. (A) Schematic of procedure used to compare 2D gels of protein extracts from E. coli prepared using bottled sterile water (S) and ultrapure water (U); (B) Coomassie stained gels; (C) Three-dimensional densitogram of gels shows more spots detected on gels made with ultrapure water (bottom) than with bottled water (top). As shown in Figure 1B, streaking was more extensive on the gels prepared and processed with bottled water. Horizontal and vertical streaking are the most common 2D gel problems. There are several possible sources of this problem, such as incomplete or excessive focusing, overloaded protein, poor sample preparation, and poor protein solubilization. Serious contamination of the water used to prepare the sample, buffers, and gels could also contribute to streaking. For example, salts and charged organics result in poor focusing in the first dimension, causing horizontal streaks. Water contamination can also lead to 2
uneven or incomplete polymerization of the second dimension gel, which may cause poor separation during SDS-PAGE, causing vertical streaking. Horizontal streaks are caused by problems with isoelectric focusing, while vertical streaks are caused by poor separation during SDS-PAGE. In addition to decreased streaking, spots were more numerous and better distributed throughout the gel that was prepared and processed with ultrapure water (Figure 1B, and also in silver stained gels that are not shown here). Using Progenesis software (Nonlinear Dynamics Ltd), we detected 407 protein spots in the gel processed with ultrapure water, while in the gel processed with bottled water, only 206 spots were detected (Figure 1C). The gels were prepared in parallel, following the same protocol and reagents except the water used to prepare the samples, buffers, and gels. This suggests that the quality of water influenced the 2D gel electrophoretic separation of proteins. 0.22 µm membrane point-of-use final filter recommended for 2D gel electrophoresis Next, we compared ultrapure water dispensed from two points of delivery equipped with different POU purifiers, one with an ultrafilter, the other with a 0.22 µm membrane filter. Ultrapure water with resistivity of 18.2 MΩ cm and TOC 5 ppb was obtained from a Milli-Q Advantage A10 system fed with pure water from an Elix 3 UV system. The Milli-Q system was equipped with two POU purifiers, an ultrafilter (BioPak, Merck Millipore) and 0.22 µm filter (Millipak, Merck Millipore). The ultrafilter is used in many life science laboratories because it effectively removes nucleases from ultrapure water. However, 0.22 µm membrane POU filters may be more suited for proteomic analyses, because they exhibit very low extractables. To compare the performance of the two water types for proteomic separations, we prepared 2D gels of U937 cell extracts (see Figure 2A for experimental schematic). For the first dimension, a strip with a linear ph range of 3-10 was used. Separation in the second dimension was carried out in a 12% SDS-polyacrylamide gel. Proteins were visualized using Coomassie blue or silver stain solutions. Ultrapure water from an ultrafilter or 0.22 µm membrane was used to prepare: (1) the rehydration and equilibration buffers for the first dimension, (2) the 12% SDS-polyacrylamide gel and separation buffer for the second dimension, and (3) the staining solution. Gels were scanned using a Bio-Rad scanner and images were analyzed using ImageMaster 2D Platinum software from GE Healthcare Life Sciences. 3
Figure 2. (A) Schematic of procedure used to compare 2D gels of protein extracts from U937 cells. (B) Coomassie blue-stained 2D gels prepared using ultrapure water delivered from a 0.22 µm membrane (M) and ultrafilter (U). (C) Densitometry shows that water processed with a 0.22 µm membrane (M) yielded a gel with better resolution and well-defined spots compared to an ultrafilter (U). Coomassie blue staining showed streaking in the high MW and basic ph region of the gel that was processed with ultrapure water from an ultrafilter. (Figure 2B, upper right portion of gels.) Spots were better distributed throughout the gel when ultrapure water from a 0.22 µm membrane POU cartridge was used (Figure 2B). Densitometry confirmed that adjacent spots were better resolved and the spots were well-defined (Figure 2C). To measure background signal, portions of the gels without spots were analyzed (images not shown). The gel processed with ultrapure water after ultrafiltration had more background peaks compared to the gel processed with ultrapure water after 0.22 µm membrane filtration. To assess reproducibility, triplicate 2D gels were prepared using water from each POU cartridge. Both types of water yielded highly reproducible gels (95% homology for each triplicate set). However, the homology between a gel processed with 0.22 µm membrane filtered water and ultrafiltered water was only 84%. Ten spots common to both gels were excised and digested with trypsin. The tryptic digests were then spotted onto a MALDI plate with 2-cyano-4-hydroxycyanuric acid matrix and analyzed by MALDI-ToF MS on an Applied Biosystems 4700 Proteomics Analyzer. Spectra were of good quality, with sufficiently intense peaks corresponding to peptides that allowed identification using Mascot (data not shown). 4
Organic contamination when using ultrafilter To compare levels of contamination in ultrapure water filtered through an ultrafilter with water filtered through a 0.22 µm membrane, we performed high performance liquid chromatography - mass spectrometry (LC-MS). Water was first pre-concentrated on a C18 XTerra column (3 µm, 4.6 x 30 mm) for 60 min at 1 ml/min with the flow directed to waste. After pre-concentration, a C18 Atlantis column (3 µm, 2.1 x 150 mm) was connected to the XTerra column and the following gradient elution profile was used to elute any organic molecules that were adsorbed on the XTerra column: 100% to 0% water for 30 mins, hold at 0% water for 10 mins. Flow was 0.25 ml/min. LC-MS was conducted on a Waters Alliance 2695 HPLC system equipped with a 2996 photodiode array detector (PDA) and a ZQ 2000 mass spectrometer. In the pre-concentration step, only water passed through the C18 pre-concentration column for 60 mins. During this time, organic contaminants in the water collected at the head of the column. During gradient elution, the amount of organic solvent passing through the column increased over time, causing the organic contaminants to elute off the column. These contaminants were detected by the PDA and by the mass spectrometer. Figure 3 shows many more peaks in the UV and mass chromatograms of water dispensed from the POU ultrafilter than from the POU membrane filter, suggesting that the POU ultrafilter was the source of organic impurities contaminating the water. These impurities could affect the quality of the gel, ultimately resulting to poor protein separation. Figure 3. HPLC chromatograms of pre-concentrated water dispensed from either a POU ultrafilter (green traces) or a POU 0.22 µm membrane filter (blue traces). (A) UV 210 detection (B) electrospray ionization mass spectrometry positive mode (ESI+). 5
Summary The data presented indicate that the source of water used in 2D gel electrophoresis can affect separation of proteins. Serious ionic contamination by salts and charged organics could alter the ionic strength of buffers and affect migration of proteins. Such ionic contamination may have caused the streaking seen in the gels prepared with bottled water. Extractables from storage containers or certain purification cartridges can also lead to poor separation as observed in the gels prepared using ultrapure water dispensed from an ultrafilter. Among the options presented in this study, ultrapure water, freshly delivered from a system with a 0.22 µm membrane POU filter, provides the best protein separation by 2D gel electrophoresis. References 1. Lopez, J.L. J. Chromtogr. B 2007, 849, 190. 2. Schroder, S.; et. al. J. Proteome Res. 2008, 7, 1226. Authors: Maricar Tarun, Stéphane Mabic Merck Millipore Lab Water business field, St.-Quentin-en-Yvelines, France 6