A Combination of Immobilised pH Gradients Improves Membrane Proteomics J. M. Chick,† P. A. Haynes,†,‡ B. Bjellqvist,§ and M. S. Baker*,†,‡ Department of Chemistry & Biomolecular Sciences, Macquarie University, NSW, 2109, Australia, Australian Proteome Analysis Facility, Macquarie University, NSW, 2109, Australia, and GE Healthcare, Biosciences AB, Bjo¨rkgatan 30, SE-751 84 Uppsala, Sweden Received May 9, 2008
Membrane protein analyses have been notoriously difficult due to hydrophobicity and the general low abundance of these proteins compared to their soluble cytosolic counterparts. Shotgun proteomics has become the preferred method for analyses of membrane proteins, in particular the recent development of peptide immobilized pH gradient isoelectric focusing (IPG-IEF) as the first dimension of two-dimensional shotgun proteomics. Recently, peptide IPG-IEF has been shown to be a valuable shotgun proteomics technique through the use of acidic narrow range IPG strips, which demonstrated that small acidic pI increments are rich in peptides. In this study, we assess the utility of both broad range (BR) (pI 3-10) and narrow range (NR) (pI 3.4-4.9) IPG strips for rat liver membrane protein analyses. Furthermore, the use of these IPG strips was evaluated using label-free quantitation to demonstrate that the identification of a subset of proteins can be improved using NR IPG strips. NR IPG strips provided 2603 protein assignments on average (with 826 integral membrane proteins (IMPs)) compared to BR IPG strips, which provided 2021 protein assignments on average (with 712 IMPs). Nonredundant protein analysis demonstrated that in total from all experiments, 4195 proteins (with 1301 IMPs) could be identified with 1428 of these proteins unique to NR IPG strips with only 636 from BR IPG strips. With the use of label-free quantitation methods, 1659 proteins were used for quantitative comparison of which 319 demonstrated statistically significant increases in normalized spectral abundance factors (NSAF) in NR IPG strips compared to 364 in BR IPG strips. In particular, a selection of six highly hydrophobic transmembrane proteins was observed to increase in NSAF using NR IPG strips. These results provide evidence for the use of alternative pH gradients in combination to improve the shotgun proteomic analysis of the membrane proteome. Keywords: Membrane proteins • Shotgun proteomics • Isoelectric focusing • Protein identification • Mass Spectrometry • Label-Free Quantitation
Introduction Comprehensive membrane proteomics has become achievable because of the use of various shotgun proteomic methodologies to separate and analyze proteolytic peptides derived from membrane protein enriched samples. Recently, methods for peptide and protein fractionation have been used for the large-scale identification of membrane proteins, including offline strong cation exchange (SCX),1-6 online SCX,7 capillary isoelectric focusing (cIEF),8,9 immobilized pH gradient (IPG)IEF,10 and 1D SDS-PAGE.11 These methods have provided the identification of hundreds to over a thousand annotated integral membrane proteins (IMPs), explaining why shotgun proteomics has now become the method of choice for analyz* To whom correspondence should be addressed. Professor Mark S. Baker, Australian Proteome Analysis Facility, Building F7B4, Research Park Drive, Macquarie University 2109, Australia. Phone: +61 2 9850 6209; fax, +61 2 9850 6200; e-mail,
[email protected]. † Department of Chemistry & Biomolecular Sciences, Macquarie University. ‡ Australian Proteome Analysis Facility, Macquarie University. § GE Healthcare.
4974 Journal of Proteome Research 2008, 7, 4974–4981 Published on Web 10/07/2008
ing membrane proteins. With the knowledge that approximately 70% of all protein based drug targets are active against IMPs,12 these methods will be used more frequently to determine the role that IMPs play in various biological environments. Like most large-scale proteomics experiments, membrane proteomics has improved with the use of methods that employ high sample loading, increased peptide and protein fractionation and extended liquid chromatography (LC) separation prior to the application of tandem mass spectrometry (MS). Wang et al. have achieved approximately 1400 IMP identifications using a combination of high sample loading and 100 min LC gradient prior to tandem MS analysis.6 Similar extended LC gradients have provided comprehensive analysis of IMPs for studies in human epithelial ovarian carcinoma8 and yeast.9 Alternatively, Blonder et al. have identified 876 IMPs simply by increasing the number of SCX fractions (96 fractions) prior to the use of a 120 min LC separation before tandem MS analysis of rat natural killer cells.1 Recently, we have demonstrated the use of peptide IPG-IEF as an alternative method for shotgun proteomics of membrane 10.1021/pr800349f CCC: $40.75
2008 American Chemical Society
research articles
IPG Strip Proteomic Analysis for Membrane Protein Identifications proteins with the identification of 690 IMPs from rat liver membranes.10 IPG-IEF has been shown to be advantageous as a high resolution shotgun proteomics method primarily because of high sample loading capacity and the ability to use the theoretical pI of peptides as an additional means to reduce false discovery rates.10,13-15 Peptide IPG-IEF has also been demonstrated to provide more high confidence protein identifications compared to 1D-SDS-PAGE based shotgun proteomics.15 IMP digestion has been shown to be improved with the use of 60% methanol,10 similar to earlier studies on IMP’s.1,3 In addition, unlike 2D gel based protein separation,16 peptide IPG-IEF was shown to be amenable to analysis of hydrophobic IMPs since peptides do not have the same solubility constraints that hydrophobic IMPs otherwise have. The amenability of hydrophobic IMPs to IEF based shotgun proteomics methods has also been demonstrated with cIEF.8,9 Alternatively, studies using narrow range (NR) IPG strips (pI 3.5-4.5) on whole cell lysates from rat testis samples have demonstrated the separation of over 7000 peptides across this small acidic region of the pI spectrum.14 These studies using NR IPG strips demonstrated that small acidic pI increments were rich in peptides14 that may otherwise go undetected with broad range (BR) IPG strips (pI 3-10). In addition, with high loading capacity of IPG strips, NR IPG-IEF should improve IMP analysis when applied to the separation of peptides derived from membrane protein enriched samples. Furthermore, the use of NR IPG strips compared to SCX fractionation was shown to provide similar total number of peptide identifications, even though the IEF was conducted in a very narrow region of the pH spectrum.14 The use of NR IPG-IEF as the first-dimension fractionation technique for shotgun proteomics of membrane proteins has yet to be evaluated and may provide an alternative to running extended LC gradients prior to tandem MS. One approach for label-free quantitation that has been developed for shotgun proteomics for comparing biological samples uses a protein’s calculated normalized spectral abundance factor (NSAF).17 These quantitative methods were used originally for determining differences in yeast cells grown in two different media and demonstrated the utility of this method for distinguishing differently abundant proteins.17 Label-free quantitation has now been applied to many other biological questions,18,19 suggestive of its valuable role in comparative proteomics studies. Label-free quantitation in essence is not restricted to use in evaluating biological questions but may also be applicable to comparing technical studies such as comparing different shotgun proteomics methods. The aim of this study was to quantify the differences between BR IPG-IEF (pI 3-10) and NR IPG-IEF (pI 3.4-4.9) shotgun proteomics using label-free quatitation for the analysis of rat liver membrane proteins. Here we demonstrate the first comparison of BR and NR IPG strips for the first-dimension separation of peptides in two-dimensional shotgun proteomics. The results demonstrate that NR acidic IPG strips provide a 2-fold increase in peptide identifications for the identical pI region from BR IPG strips. Furthermore, NR IPG strips provide a significant increase in total protein identifications and IMPs. Label-free quantitation was also valuable in showing a significant enrichment of acidic peptides in NR IPG strips through both raw spectral abundance counts and NSAF values. NR IPGIEF enriched for a subset of proteins that would otherwise go undetected from BR IPG-IEF. Our results demonstrate that NR IPG-IEF in combination with BR IPG-IEF is a valuable firstdimension separation method for shotgun proteomics of
membrane enriched samples and provides the first use of labelfree quantitation for assessing technical comparisons.
Materials and Methods Membrane Protein Isolation and Preparation for Peptide Immobilised pH Gradient Isoelectric Focusing (IPG-IEF). Rat livers obtained from 8 week old Dark Agouti rats (Save Sight Institute, Sydney Eye Hospital, Australia) were perfused with 0.9% (w/v) phosphate buffered saline. Rat liver tissue (1.5 g) was then homogenized in 10 mM HEPES buffer (pH 7.8) supplemented with 2 mM NaCl, 10 mM NaOH, 500 mM EDTA and protease inhibitors (3 mg of antipain-dihydrochloride, 0.5 mg of aprotinin, 0.5 mg of bestatin, 1 mg of chymostatin, 3 mg of E-64, 10 mg of EDTA-Na2, 0.5 mg of leupeptin, 20 mg of Pefabloc SC, 0.5 mg of pepstatin, 3 mg of phosphoramidon) (Roche Diagnostics, Switzerland) using an Omni TH homogenizer (Omni International, Inc., VA). The homogenized tissue was then centrifuged at 13,000g for 15 min at room temperature (RT). Membrane proteins were isolated using a modified sodium carbonate stripping method.20,21 Briefly, the supernatant was collected and diluted to a final volume of 40 mL in 0.1 M sodium carbonate (pH 11), then incubated for 1 h rotating at 4 °C. The carbonate-treated membranes were sedimented by ultracentrifugation at 120,000g for 1 h at 4 °C. The supernatant was removed and the membrane pellet was washed once with 0.1 M sodium carbonate (pH 11) and resuspended in 10 mM ammonium bicarbonate (NH4HCO3, pH 7.8). Sample was then transferred to a 20 mL glass scintillation vial and pulsesonicated using a Branson 450 Sonifier (Branson, Danbury, CT) using 2 s bursts for 15 intervals on ice. The sonicated sample was then reduced with 10 mM dithiothreitol for 1 h at 37 °C and subsequently alkylated with 55 mM iodoacetamide at RT for 30 min. Protein quantification was conducted by Bradford Assay (Sigma-Aldrich, MO, USA). Trypsin digestion was carried out in the presence of 60% (v/v) methanol in 10 mM ammonium bicarbonate (pH 7.8) for approximately 8 h at 37 °C. All digested samples were evaporated to dryness in a vacuum centrifuge and resuspended in 250 µL of 8 M urea supplemented with a trace of bromophenol blue. Peptide IPG-IEF. Digested proteins (1 mg) in 8 M urea were used to passively rehydrate either a linear pH 3-10 (in which the actual pH range of the IPG strip is 3.85-9.36, information available at www.gehealthcare.com) 18 cm IPG strip (broad range) or pH 3.4-4.9 (with an actual pH range of 3.35-5) 22 cm IPG strip (narrow range) for 6 h at room temperature in triplicate. Isoelectric focusing was conducted on an IPGphorII (GE Healthcare) with a current limit of 50 µA per strip for the BR IPG strip and 200 µA limit for NR IPG strip at 20 °C with the following focusing program: 300 V for 1 h, a gradient to 1000 V for 1 h, a gradient to 4000 V for 3 h, a gradient to 8000 V for 3 h and 8000 V until 100 kVh was reached. The IPG strips were treated differently for their respective µA limits so that both IPG strips would have 8000 V as their maximum volt step. The strips were then cut (with plastic backing still in place) with a scalpel blade into 24 equal length pieces. Peptides were extracted from each fraction by incubation in 100 µL of 0.1% (v/v) formic acid for 1 h at RT. The extraction was repeated twice and subsequently combined with the initial fractions. Combined peptide extracts were desalted using C18 tips (Omix, Journal of Proteome Research • Vol. 7, No. 11, 2008 4975
research articles Varian, Inc., CA) and the eluate was dried using a vacuum centrifuge followed by resuspension in 0.1% (v/v) formic acid. Nanoflow Liquid Chromatography-Tandem Mass Spectrometry (NanoLC-MS/MS). Each of the 24 fractions from the triplicate BR and NR IPG-IEF experiments were analyzed by nanoLC-MS/MS using a LTQ-XL ion-trap mass spectrometer (Thermo, CA) according to Breci et al.22 Reversed phase columns were packed in-house to approximately 7 cm (100 µm i.d.) using 100 Å, 5 mM Zorbax C18 resin (Agilent Technologies, CA) in a fused silica capillary with an integrated electrospray tip. A 1.8 kV electrospray voltage was applied via a liquid junction upstream of the C18 column. Samples were injected onto the C18 column using a Surveyor autosampler (Thermo, CA). Each sample was loaded onto the C18 column followed by initial wash step with buffer A (5% (v/v) ACN, 0.1% (v/v) formic acid) for 10 min at 1 µL min-1. Peptides were subsequently eluted from the C18 column with 0-50% Buffer B (95% (v/v) ACN, 0.1% (v/v) formic acid) over 58 min at 500 nL min-1 followed by 50-95% Buffer B over 5 min at 500 nL min-1. The column eluate was directed into a nanospray ionization source of the mass spectrometer. Spectra were scanned over the range 400-1500 amu. Automated peak recognition, dynamic exclusion, and tandem MS of the top six most intense precursor ions at 35% normalized collision energy were performed using the Xcalibur software (Thermo, CA). Protein and Peptide Identification. Raw files were converted to mzXML format and processed through the global proteome machine (GPM) software using version 2.1.1 of the X!Tandem algorithm, freely available from www.thegpm.org.23,24 For each experiment, the 24 fractions were processed sequentially with output files for each individual fraction and a merged, nonredundant output file was generated for protein identifications with Log(e) values less than -1. Peptide identification was determined using a 0.4 Da fragment ion tolerance. Carbamidomethyl was considered as a complete modification and partial modifications were also considered, which included oxidation of methionine and threonine, and deamidation of asparagine and glutamine. MS/MS spectra were searched against the Rattus norvegicus database (Database derived from Swiss-Prot, Ensemble and NCBI) and reverse database searching was used for estimating false discovery rates.25 Protein identifications were validated using peptide pI filtering as described in our previous report.10 Briefly, pI was calculated through an open source pI calculator, Compute pI from ExPASy (http://au.expasy.org/tools/pi_tool.html), which calculates peptide pI from amino acid values described by Bjellqvist et al.26 Peptide pI filtering was conducted on the data obtained from each of the fractions by removal of statistical outliers that were defined as peptides with pI values that fall outside the pI boundaries, determined using (0.5 of the standard deviation (calculated over the entire fraction) from the median pI value of the fraction. For analysis of total protein identifications, proteins were further validated using a 1% false discovery rate by searching the MS spectra against a reverse database and applying the following formula: FDR ) reverse/ (reverse + forward) × 100. Transmembrane segment annotation of the identified proteins was determined using the TransMembrane Hidden Markov Model (TMHMM, http://www.cbs.dtu.dk/services/ TMHMM-2.0/).27-29 Calculation of Normalized Spectral Abundance Factors. Normalized spectral abundance factors were calculated according to Zybailov et al.,17 using the following formula: NSAF 4976
Journal of Proteome Research • Vol. 7, No. 11, 2008
Chick et al. ) (Spc/L)/∑(Spc/L), where Spc refers to spectral count (number of nonredundant peptide identifications for a given protein), L is the length of the protein. Protein identifications were only included in NSAF data analysis if a given protein was identified in each of the triplicate experiments from either BR or NR IPG strips. The reduced protein list was then adjusted with a fraction of a spectral count to allow the incorporation of proteins with zero spectral counts for statistical analysis.17 An optimal adjustment factor was determined through fitting the natural log of the NSAF values to a Gaussian curve, assessed by computing both R2 values and applying a Shapiro-Wilk test. To satisfy the Shapiro-Wilk test, a W value greater than 0.05 was required. Statistical Analysis. Statistical analysis was conducted with Microsoft Excel using the Student t-test with equal sample sizes and P < 0.05 for assignment of statistical confidence. Origin8 software package was used for assessing Gaussian distribution of data for label-free quantitation by computing both R2 values and a Shapiro-Wilk test.
Results Peptide Separation and Analysis. To provide an understanding of why NR IPG strips are valuable for protein and peptide identifications, we first sought to determine what would be the theoretical distribution of rat proteolytic peptides corresponding to annotated spectrum libraries (ftp://ftp. thegpm.org/proteotypic_peptide_profiles/eukaryotes/) for 63 964 peptides according to Craig et al.30 Of all the proteolytic peptides, 46% (29 623) lie within the pI range of 3-5, which suggests that NR IPG-IEF analysis could be a valuable tool for protein identifications. To demonstrate the reproducible separation of tryptic peptides by IEF using BR and NR IPG strips, the number of peptides per fraction for each strip was compiled. Analysis of the number of peptides per fraction demonstrates that NR IPG strips provide a well-separated sample with an even distribution of peptides with 642 ( 33 peptides on average per fractions after pI filtering (Figure 1A). BR IPG-IEF also demonstrated a well-separated sample with three distinct pI regions in which peptides were concentrated. The average reduction in the number of peptides per fraction after pI filtering for NR IPG strips was 21%, suggesting a minority of peptides are false positive identifications. This is similar to results from BR pH gradients with a 28% reduction. The high reproducibility of peptide IPG-IEF for both types of IPG strips was indicated by low standard deviation in the number of peptides per fraction between experimental replicates. To demonstrate that the separation of peptides on the BR and NR IPG strips is due to isoelectric point, the average peptide pI per fraction was calculated (Figure 1B). The average peptide pI per fraction after pI filtering starts at approximately 4 and reaches as high as 5.89 ( 0.005 pI for the NR IPG strips and as high as 8.65 ( 0.01 for the BR IPG strips. In comparison to the actual pH of the IPG strip, the average pI per fraction for the NR IPG strips is slightly higher than the actual pH of the IPG strip between fractions 1-16. However, the opposite is seen for fractions 19-23 with lower pI values than the pH of the IPG strip. Furthermore, fraction 24 shows a large increase in the average pI per fraction where this is attributed to the IPG strip retaining peptides with higher pI values greater than a pI of 5. The BR IPG strip has more variability in average pI per fraction in comparison to the actual pH of the IPG strip. In addition, the average pI per fraction was reproducible as
IPG Strip Proteomic Analysis for Membrane Protein Identifications
research articles
Figure 1. IPG-IEF separation of peptides using BR (pI 3-10) and NR (pI 3.4-4.9) IPG strips. (A) Number of peptides per fraction after pI filtering. (B) Average peptide pI per fraction for BR and NR IPG strips after pI filtering and their respective pH gradients.
shown by the low standard deviation in individual fractions between experimental replicates. These results clearly demonstrate that peptides derived from trypsin digestion of rat liver membrane proteins were reproducibly separated by isoelectric point, and therefore, comparisons can be made at the total peptide and protein levels. In comparing the differences between broad range (pI 3-10) and narrow range (pI 3.4-4.9) IPG-IEF, peptides were binned into 0.5 pI increments across the pI intervals of 3-10 (Figure 2). The NR IPG strip demonstrated an expected higher number of peptide identifications than the BR IPG strip within the pI ranges of 3.5-4, 4-4.5 and 4.5-5 (n ) 3, p < 0.05). The NR IPG strips within pI increment of 3.5-4 provided 1134 ( 125 peptides compared to 316 ( 12 from the BR IPG strips. The pI increment between 4-4.5 demonstrated the highest difference in peptide numbers with 6292 ( 576 from NR IPG strips and 2197 ( 215 from BR IPG strips. Furthermore, the pI increment of 4.5-5 contained 2968 ( 279 peptides for the NR IPG strips and 1790 ( 123 peptides from the BR IPG strips. The number of peptides within the pI range of 3-5 was 2-fold greater for the number of peptides from the NR IPG strips over the BR IPG strips. This fold change represents the identification of almost 5500 more peptide identifications on average from the NR IPG strips compared to the same acidic region from BR IPG strips. These results demonstrate compelling evidence that
narrow range IPG strips can enrich for more peptides in a discrete acidic pI region when compared to the same region in BR IPG strips. Comparison of BR and NR IPG Strips for Total Peptide and Protein Identifications. After demonstrating that NR IPG strips provide more peptide assignments compared to the same pI region from BR IPG strips, it was of interest to compare total peptide, protein and IMP assignments between the two sets of IPG strips (Figure 3). Comparisons between BR and NR IPG strips for total peptide identifications demonstrates no significant increase in peptide assignments with 10 752 ( 1002 and 11 173 ( 1583.9, respectively. At the protein level, NR IPG strips provided an increase in protein identifications with 2603.7 ( 355 compared to BR IPG strips with 2021 ( 138.6; however, these differences are not statistically significant. Similarly, the analysis of IMPs demonstrated that NR IPG strips provided more IMPs with 826.3 ( 65.1 compared to BR with 712 ( 36.2, which is also not a statistically significant difference. These results demonstrate the added value of using NR IPG strips within the pI region of 3.4-4.9 for providing an increase in total protein and IMP assignments. To determine the differences in the nonredundant protein identifications between BR and NR IPG strips, nonredundant proteins for each set of IPG strips were compiled and compared (Figure 4). Of the total 4195 protein identifications found in all Journal of Proteome Research • Vol. 7, No. 11, 2008 4977
research articles
Chick et al.
Figure 2. Peptides were binned into 0.5 pI increments and presented in a pI histogram representing a pH gradient from 3-10. Broad range IPG strips (black) provided peptide identifications throughout the pH spectrum. NR IPG strips (gray) provided more peptide identifications over the acidic pI regions of 3-5 (* n ) 3, p < 0.05).
Figure 5. Number of IMP identifications from triplicate BR and NR IPG strips based on the proteins number of transmembrane segments. Slight increases in protein numbers were observed for transmembrane proteins with single and double transmembrane segment containing proteins. Figure 3. Number of peptide, protein and IMP identifications from BR and NR IPG strip triplicates. Peptide identifications (black) between IPG strips had no observable differences. However, both protein (dark gray) and IMP (light gray) identifications were significantly increased using NR IPG strips.
Figure 4. Venn diagram representing the number of nonredundant protein and IMP identifications compiled from the triplicate IPG experiments from BR and NR IPG strips.
three replicates, 2131 proteins (51%) were common to both sets of IPG strips. However, 1428 (34%) of the protein identifications were unique to NR IPG strips, which was more than twice the percentage unique to BR IPG strips (15%, 636). The proteins that were unique to the BR IPG strips included a subset which were identified from entirely, or a majority of, peptides that lay outside the pI ranges of the NR IPG strip, principally at the basic end (data not shown). Furthermore, comparing the similarities and differences for IMP identifications demon4978
Journal of Proteome Research • Vol. 7, No. 11, 2008
strated that 57.8% of IMPs were common to NR and BR IPG strips. NR IPG strips provided an increase in the percentage of total IMP’s with 26.6% compared to BR IPG strips with 15.6%. These data demonstrate that in combination, BR and NR IPG strips provide a comprehensive characterization of the liver membrane proteome. After demonstrating that the percentage of unique IMPs was higher for NR IPG strips, an analysis of the types of IMPs based on transmembrane segments from BR and NR IPG strips was conducted (Figure 5). Single transmembrane proteins represented the most common IMP identifications with a slight increase in protein identifications for NR IPG strips but this increase was found not to be statistically significant. In contrast, IMPs with two transmembrane segments showed a significant increase (n ) 3, P < 0.05) in protein identifications from NR IPG strips (106 ( 7) compared to BR IPG strips (91 ( 2.7). IMPs with greater than two transmembrane segments did not show an increase in protein identifications, however, both types of IPG strips were amenable to the analysis of multitransmembrane segment containing IMPs. In total, 1301 nonredundant IMPs were identified from the BR and NR IPG strips suggesting that together BR and NR IPG-IEF are highly valuable techniques for the identification and characterization of IMPs. Analysis of the Relative Abundance of Peptides between BR and NR Range IPG Strips Using NSAF. To demonstrate that NR IPG strips provided both a unique set and an increase
IPG Strip Proteomic Analysis for Membrane Protein Identifications
research articles
Table 1. Comparison of BR and NR IPG Strips Using Label-Free Quantitation on 1659 Proteins That Were Identified in All Three Experiments from Either IPG Stripa
Total Protein Identifications Unique Proteins Total Proteins with Increased NSAF IMPs with Increased NSAF
broad range
narrow range
1235 (74%) 287 (17%) 364 (22%) 123 (7%)
1365 (82%) 406 (24%) 319 (19%) 130 (9%)
a The presented percentages are in relation to the total protein identifications.
in peptide identifications for a subset of proteins, label-free quantification was used to distinguish differences with statistical confidence. From each of the pI filtered data sets, proteins that were identified from each of the replicate experiments in either NR or BR IPG strips were only used for further analysis. In addition, proteins with single spectral counts were removed from the analysis to prevent error in assigning differential quantities of proteins (i.e., proteins found to have statistically significant changes in abundance between IPG strips but which actually contain the same spectral count resulting in false assessment). To allow the incorporation of proteins with a zero spectral count in one data set, 0.5 of a spectral count was added to each data point according to Zybailov et al.17 Normalized spectral abundance factors were calculated for each protein and a Student t-test was calculated for comparison between the two IPG strips. To confirm that the data set was indeed Gaussian and therefore valid for analysis by a Student t-test both R2 values and a Shapiro-Wilk test was performed. Analysis of the BR and NR IPG strips demonstrated a Gaussian fit with R2 values >0.98 and Shapiro-Wilk test of 0.061 and 0.05, respectively. The results from this analysis demonstrated that 1659 proteins with a 100 spectral counts (n ) 3, p < 0.05) (Figure 6). These data demonstrate that NR IPG strips are valuable for increasing the spectral counts for a subset of proteins compared to BR range IPG strips and improve upon the characterization of some IMPs identified from the BR IPG strips. After demonstrating that NR IPG-IEF could improve the characterization of a subset of proteins including IMPs, it was of interest to investigate if large multitransmembrane segment containing proteins had increased NSAF values (Figure 7). In total, 6 IMPs containing greater than 10 transmembrane segments had significant increases (n ) 3, p < 0.05) in NSAF values in the NR IPG strips compared to the BR IPG strips. For example, multidrug resistance-associated protein 6 (Abc C6) had a 2.79-fold change in NSAF for NR compared to BR IPG strips (p < 0.05). Furthermore, multidrug resistance protein 2
Figure 6. Comparison of proteins grouped by spectral counts, which demonstrated significant increases in proteins with spectral counts of 1-10 and >100 (* n ) 3, p < 0.05).
Figure 7. A subset of highly hydrophobic, multitransmembrane segment (>10) containing IMPs with increased NSAF values using NR IPG strips compared to broad range IPG strips (n ) 3, p < 0.05). Slc9a1 ) solute carrier family 9 member 1, AbcC2 ) ATP-binding cassette family C member 2, Abc a8b ) ATP-binding cassette family a member 8b, Abc b1a ) ATP-binding cassette family b member 1a, Abc c6 ) ATP-binding cassette family C member 6, Slc 29(1) ) solute carrier family 29 member 1.
(Abc C2) had a 1.79-fold change in NSAF values for NR compared to BR IPG strips (p < 0.05). These 6 proteins are largely hydrophobic IMPs that would not be amenable for analysis by 2D gels,16 hence, demonstrating that the use peptide IPG-IEF, in particular NR IPG-IEF, is a valuable technique for analysis of IMPs.
Discussion The aim of this study was to compare the use of BR and NR IPG strips for the first-dimension shotgun proteomics separation of membrane enriched samples from rat liver. Recently, improved membrane protein analysis using shotgun proteomics has been achieved mainly through high sample loading, increased first-dimension fractionation or through extended LC gradients prior to tandem MS analysis.3,6,8 NR IPG strips have been shown to provide an increase in peptide identifications over acidic pI regions.14 These studies demonstrated that acidic pH regions of the pH spectrum are quite rich in peptides derived from proteolytic digestion of rat testis proteins.14 When proteotypic peptides derived from membrane enriched samples were applied to NR IPG strips, it was hypothesized that NR IPG strips would improve the characterization of some proteins in particular a subset of IMP’s. Our analysis of rat proteolytic peptides from annotated spectrum libraries demonstrated that a large proportion of experimentally attained peptide sequences have pI values that Journal of Proteome Research • Vol. 7, No. 11, 2008 4979
research articles fall with the pI region of 3-5. This analysis confirmed that peptide IPG-IEF using NR IPG strips within the acidic pI region of 3-5 would be a valuable alternative method to BR IPG strips. In this study, both BR and NR IPG strips provided a wellseparated sample over 24 fractions. BR IPG strips demonstrated a tripartite distribution of peptides based on the number of peptides per fraction, which is in agreement with previously published data.10,13,15 The number of peptides per fraction within the NR IPG strip demonstrated a more even distribution across all 24 fractions, again similar to previously reported data.14 The separation of peptides based on isoelectric point was confirmed by the average pI per fraction, which followed a linear gradient similar to the actual pH of the IPG strip. However, some discrepancies between the average pI of some fractions and the actual pI in each strip do exist. In particular, fraction 24 from the NR IPG strip had quite a high average pI and further analysis of this fraction demonstrates that peptides with much higher pI values are still being retained in this fraction. In comparing the number of peptides identified from the NR IPG strip with the corresponding region on the BR IPG strip, approximately 5500 more peptides, representing a 2-fold increase in peptide identifications, were observed. This is the first study to demonstrate an increase in peptide identifications as a direct comparison between the BR and NR IPG strips. The total unique peptide identifications from BR and NR IPG strips demonstrated that approximately 10 000-12 000 high confidence peptide identifications could be assigned for each of the IPG strips after peptide pI filtering and 1% FDR. At the protein level, the number of protein identifications increased by approximately 600 protein identifications on average for the NR IPG strip with ∼2600 as compared to the BR IPG strip with ∼2000. The number of protein identifications on average from the NR IPG strip was similar to previously reports from rat testis samples using a similar sample load.14 Importantly, NR IPG strips allowed 114 more IMP identifications then BR IPG strips. On the basis of each NR IPG strip experimental replicates, an average of 826 IMPs was identified in this study, which is remarkably similar to the data from Blonder et al.1 What should be noted in this comparison is that we are using less firstdimension fractions and shorter LC gradients but using a much higher sample load. The total number of nonredundant protein identifications across all three BR and all three NR IPG strips was 4195 high confidence protein assignments of which 1301 proteins were annotated as IMPs. The total redundant IMP identifications is similar to previous reports of 1400 IMPs from human epithelial ovarian carcinoma cells6 and 1190 from mouse liver.31 Analysis of the similarities and differences between the nonredundant protein identifications between BR and NR IPG strips demonstrated that 2131 proteins were common to both types of IPG strips. However, NR IPG strips identified 1428 unique protein assignments, while BR IPG strips provided 636 unique protein identifications. Similarly, NR IPG strips provided more unique IMP identifications. These results together demonstrate that BR and NR IPG strips provide a comprehensive inventory of protein identifications and in particular IMP identifications. The distribution of IMPs based on the number of transmembrane segments is similar to past reports.1,6,10,32 The number of total protein and IMP identifications from NR IPG strips demonstrates the utility of NR IEF for improved proteome coverage, in particular the membrane proteome. The number of integral membrane proteins in this study and in previous studies1,6,10,32 demonstrates that single transmembrane pro4980
Journal of Proteome Research • Vol. 7, No. 11, 2008
Chick et al. teins are the most common IMPs. Furthermore, the use of shotgun proteomics has also allowed the identification of highly hydrophobic, multitransmembrane spanning IMPs. Our data establishes for the first time the use of label-free quantitation for use in technique comparisons as well as biological studies. In this study, 1659 proteins were used for label-free quantitation as they contain at least two spectral counts from all three replicates from either the BR or NR IPG strips. NR IPG strips contained 319 proteins that had a significant increase in NSAF values of which 142 were IMPs. It is also important to note that 364 proteins, in particular 109 IMPs, decreased in NSAF values. However, these results provide evidence that NR IPG strips can improve the analysis of a subset of IMPs over the use of BR IPG strips. NR IPG strips were also shown to increase the recovery of proteins containing between 1-10 and >100 spectral counts. The higher number of spectral counts for proteins between 1-10 spectral counts in NR IPG strips is expected when the number of protein identifications also increases. However, the increase in the number of proteins with >100 spectral counts is unexpected since the total number of peptide identifications is similar. Multitransmembrane segment containing IMPs are a notoriously difficult class of proteins to analyze since they are poorly soluble due to their largely hydrophobic nature. Our results demonstrate that a subset of IMPs increased in NSAF values when using NR IPG strips. In particular, multitransmembrane segment containing IMPs were better represented. As an example, proteins such as the ABC transporters known to contain >10 transmembrane segments increased in spectral counts due to the use of NR IPG strips. These proteins, due to their largely hydrophobic nature, are refractory to analysis by 2D gels because of poor solubility.16 Here we present 6 IMPs, with greater than 10 transmembrane segments, for which NSAF values have significantly improved through the use of NR IPG strips.
Conclusion In summary, this study demonstrates the utility of NR IPG strips as the first dimension separation of peptides for comparative two-dimensional shotgun proteomics of a membrane protein enriched sample. NR IPG strips were shown to be advantageous with results suggesting that NR IPG strips could increase protein and IMP identifications. Furthermore, NR IPG strips also provided a set of proteins and specifically IMPs that currently go undetected in BR IPG strips. In addition, labelfree quantitation was used to demonstrate the differences, with statistical significance, between BR and NR IPG strips based on each of the proteins measured spectral abundance. These results demonstrate that the combination of BR and NR IPG strips as the first-dimension separation of shotgun proteomics provides a comprehensive profile of the rat liver membrane proteome. Abbreviations: IMP, integral membrane proteins; NSAF, normalized spectral abundance factors; IEF, isoelectric focusing; IPG, immobilized pH gradient; 7, immobilized pH gradients isoelectric focusing; BR, broad range; NR, narrow range; SCX, strong cation exchange; MS, mass spectrometry; MS/MS, tandem mass spectrometry; LC, liquid chromatography; LCMS/MS, liquid chromatography tandem mass spectrometry; FDR, false discovery rate.
Acknowledgment. The research was supported by the Australian Research Council and GE Healthcare. The authors
research articles
IPG Strip Proteomic Analysis for Membrane Protein Identifications would like to thank Susanna Lindman, Lena Hornsten, Kristena Uhlen and Brian Hood for technical and product support. This research was facilitated by access to the Australian Proteome Analysis Facility Ltd., established under the Australian Commonwealth Government Major National Research Facilities Schemes and NCRIS.
Supporting Information Available: Supplementary table contains the nonredundant protein identifications used for the quantitative comparison between broad range and narrow range IPG strips. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Blonder, J.; Rodriguez-Galan, M. C.; Chan, K. C.; Lucas, D. A.; Yu, L. R.; Conrads, T. P.; Issaq, H. J.; Young, H. A.; Veenstra, T. D. Analysis of murine natural killer cell microsomal proteins using two-dimensional liquid chromatography coupled to tandem electrospray ionization mass spectrometry. J. Proteome Res. 2004, 3 (4), 862–70. (2) Blonder, J.; Chan, K. C.; Issaq, H. J.; Veenstra, T. D. Identification of membrane proteins from mammalian cell/tissue using methanolfacilitated solubilization and tryptic digestion coupled with 2DLC-MS/MS. Nat. Protoc. 2006, 1 (6), 2784–90. (3) Blonder, J.; Goshe, M. B.; Moore, R. J.; Pasa-Tolic, L.; Masselon, C. D.; Lipton, M. S.; Smith, R. D. Enrichment of integral membrane proteins for proteomic analysis using liquid chromatography-tandem mass spectrometry. J. Proteome Res. 2002, 1 (4), 351–60. (4) Blonder, J.; Goshe, M. B.; Xiao, W.; Camp, D. G., II; Wingerd, M.; Davis, R. W.; Smith, R. D. Global analysis of the membrane subproteome of Pseudomonas aeruginosa using liquid chromatography-tandem mass spectrometry. J. Proteome Res. 2004, 3 (3), 434–44. (5) Blonder, J.; Terunuma, A.; Conrads, T. P.; Chan, K. C.; Yee, C.; Lucas, D. A.; Schaefer, C. F.; Yu, L. R.; Issaq, H. J.; Veenstra, T. D.; Vogel, J. C. A proteomic characterization of the plasma membrane of human epidermis by high-throughput mass spectrometry. J. Invest. Dermatol. 2004, 123 (4), 691–9. (6) Wang, H.; Qian, W. J.; Chin, M. H.; Petyuk, V. A.; Barry, R. C.; Liu, T.; Gritsenko, M. A.; Mottaz, H. M.; Moore, R. J.; Camp Ii, D. G.; Khan, A. H.; Smith, D. J.; Smith, R. D. Characterization of the mouse brain proteome using global proteomic analysis complemented with cysteinyl-peptide enrichment. J. Proteome Res. 2006, 5 (2), 361–9. (7) Wu, C. C.; MacCoss, M. J.; Howell, K. E.; Yates, J. R., III. A method for the comprehensive proteomic analysis of membrane proteins. Nat. Biotechnol. 2003, 21 (5), 532–8. (8) Wang, W.; Guo, T.; Rudnick, P. A.; Song, T.; Li, J.; Zhuang, Z.; Zheng, W.; Devoe, D. L.; Lee, C. S.; Balgley, B. M. Membrane proteome analysis of microdissected ovarian tumor tissues using capillary isoelectric focusing/reversed-phase liquid chromatography-tandem MS. Anal. Chem. 2007, 79 (3), 1002–9. (9) Wang, W.; Guo, T.; Song, T.; Lee, C. S.; Balgley, B. M. Comprehensive yeast proteome analysis using a capillary isoelectric focusing-based multidimensional separation platform coupled with ESI-MS/MS. Proteomics 2007, 7 (8), 1178–87. (10) Chick, J. M.; Haynes, P. A.; Molloy, M. P.; Bjellqvist, B.; Baker, M. S.; Len, A. C. Characterization of the rat liver membrane proteome using peptide immobilized pH gradient isoelectric focusing. J. Proteome Res. 2008, 7 (3), 1036–45. (11) Andon, N. L.; Hollingworth, S.; Koller, A.; Greenland, A. J.; Yates, J. R., III.; Haynes, P. A. Proteomic characterization of wheat amyloplasts using identification of proteins by tandem mass spectrometry. Proteomics 2002, 2 (9), 1156–68. (12) Hopkins, A. L.; Groom, C. R. The druggable genome. Nat. Rev. Drug Discovery 2002, 1 (9), 727–30. (13) Cargile, B. J.; Talley, D. L.; Stephenson, J. L., Jr. Immobilized pH gradients as a first dimension in shotgun proteomics and analysis
(14)
(15) (16)
(17)
(18)
(19)
(20)
(21)
(22)
(23) (24) (25)
(26)
(27)
(28) (29) (30) (31) (32)
of the accuracy of pI predictability of peptides. Electrophoresis 2004, 25 (6), 936–45. Essader, A. S.; Cargile, B. J.; Bundy, J. L.; Stephenson, J. L., Jr. A comparison of immobilized pH gradient isoelectric focusing and strong-cation-exchange chromatography as a first dimension in shotgun proteomics. Proteomics 2005, 5 (1), 24–34. Krijgsveld, J.; Gauci, S.; Dormeyer, W.; Heck, A. J. In-gel isoelectric focusing of peptides as a tool for improved protein identification. J. Proteome Res. 2006, 5 (7), 1721–30. Molloy, M. P.; Herbert, B. R.; Walsh, B. J.; Tyler, M. I.; Traini, M.; Sanchez, J. C.; Hochstrasser, D. F.; Williams, K. L.; Gooley, A. A. Extraction of membrane proteins by differential solubilization for separation using two-dimensional gel electrophoresis. Electrophoresis 1998, 19 (5), 837–44. Zybailov, B.; Mosley, A. L.; Sardiu, M. E.; Coleman, M. K.; Florens, L.; Washburn, M. P. Statistical analysis of membrane proteome expression changes in Saccharomyces cerevisiae. J. Proteome Res. 2006, 5 (9), 2339–47. Paoletti, A. C.; Parmely, T. J.; Tomomori-Sato, C.; Sato, S.; Zhu, D.; Conaway, R. C.; Conaway, J. W.; Florens, L.; Washburn, M. P. Quantitative proteomic analysis of distinct mammalian Mediator complexes using normalized spectral abundance factors. Proc. Natl. Acad. Sci. U.S.A. 2006, 103 (50), 18928–33. Sardiu, M. E.; Cai, Y.; Jin, J.; Swanson, S. K.; Conaway, R. C.; Conaway, J. W.; Florens, L.; Washburn, M. P. Probabilistic assembly of human protein interaction networks from label-free quantitative proteomics. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (5), 1454–9. Fujiki, Y.; Hubbard, A. L.; Fowler, S.; Lazarow, P. B. Isolation of intracellular membranes by means of sodium carbonate treatment: application to endoplasmic reticulum. J. Cell Biol. 1982, 93 (1), 97–102. Molloy, M. P.; Herbert, B. R.; Slade, M. B.; Rabilloud, T.; Nouwens, A. S.; Williams, K. L.; Gooley, A. A. Proteomic analysis of the Escherichia coli outer membrane. Eur. J. Biochem. 2000, 267 (10), 2871–81. Breci, L.; Hattrup, E.; Keeler, M.; Letarte, J.; Johnson, R.; Haynes, P. A. Comprehensive proteomics in yeast using chromatographic fractionation, gas phase fractionation, protein gel electrophoresis, and isoelectric focusing. Proteomics 2005, 5 (8), 2018–28. Craig, R.; Beavis, R. C. A method for reducing the time required to match protein sequences with tandem mass spectra. Rapid Commun. Mass Spectrom. 2003, 17 (20), 2310–6. Craig, R.; Beavis, R. C. TANDEM: matching proteins with tandem mass spectra. Bioinformatics 2004, 20 (9), 1466–7. Peng, J.; Elias, J. E.; Thoreen, C. C.; Licklider, L. J.; Gygi, S. P. Evaluation of multidimensional chromatography coupled with tandem mass spectrometry (LC/LC-MS/MS) for large-scale protein analysis: the yeast proteome. J. Proteome Res. 2003, 2 (1), 43–50. Bjellqvist, B.; Hughes, G. J.; Pasquali, C.; Paquet, N.; Ravier, F.; Sanchez, J. C.; Frutiger, S.; Hochstrasser, D. The focusing positions of polypeptides in immobilized pH gradients can be predicted from their amino acid sequences. Electrophoresis 1993, 14 (10), 1023–31. Krogh, A.; Larsson, B.; von Heijne, G.; Sonnhammer, E. L. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 2001, 305 (3), 567– 80. Sonnhammer, E. L.; von Heijne, G.; Krogh, A. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc. Int. Conf. Intell. Syst. Mol. Biol. 1998, 6, 175–82. Speers, A. E.; Wu, C. C. Proteomics of integral membrane proteinstheory and application. Chem. Rev. 2007, 107 (8), 3687–714. Craig, R.; Cortens, J. C.; Fenyo, D.; Beavis, R. C. Using annotated peptide mass spectrum libraries for protein identification. J Proteome Res 2006, 5 (8), 1843–9. Lai, K. K.; Kolippakkam, D.; Beretta, L. Comprehensive and quantitative proteome profiling of the mouse liver and plasma. Hepatology 2008, 47 (3), 1043–51. Wu, C. C.; MacCoss, M. J. Shotgun proteomics: tools for the analysis of complex biological systems. Curr Opin Mol Ther 2002, 4 (3), 242–50.
PR800349F
Journal of Proteome Research • Vol. 7, No. 11, 2008 4981