Phosphoproteome Profiling of Human Skin Fibroblast Cells in

First, this spectrum is not of particularly high quality (evidenced by the low intensity and signal/noise ratio). Second, there is an unexplained inte...
0 downloads 0 Views 289KB Size
Download PDF
Phosphoproteome Profiling of Human Skin Fibroblast Cells in Response to Low- and High-Dose Irradiation Feng Yang,† David L. Stenoien,† Eric F. Strittmatter,† Junhua Wang,‡ Lianghao Ding,‡ Mary S. Lipton,† Matthew E. Monroe,† Carrie D. Nicora,† Marina A. Gristenko,† Keqi Tang,† Ruihua Fang,† Joshua N. Adkins,† David G. Camp, II,† David J. Chen,‡ and Richard D. Smith*,† Biological Sciences Division and Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, P.O. Box 999, Richland, Washington 99352, and Division of Molecular Radiation Biology, Department of Radiation Oncology, UT Southwestern Medical Center, Dallas, Texas 75390 Received January 26, 2006

A hallmark of the response to high-dose radiation is the up-regulation and phosphorylation of proteins involved in cell cycle checkpoint control, DNA damage signaling, DNA repair, and apoptosis. Exposure of cells to low doses of radiation has well documented biological effects, but the underlying regulatory mechanisms are still poorly understood. The objective of this study is to provide an initial profile of the normal human skin fibroblast (HSF) phosphoproteome and explore potential differences between low- and high-dose irradiation responses at the protein phosphorylation level. Several techniques including Trizol extraction of proteins, methylation of tryptic peptides, enrichment of phosphopeptides with immobilized metal affinity chromatography (IMAC), nanoflow reversed-phase HPLC (nano-LC)/ electrospray ionization, and tandem mass spectrometry were combined for analysis of the HSF cell phosphoproteome. Among 494 unique phosphopeptides, 232 were singly phosphorylated, while 262 peptides had multiple phosphorylation sites indicating the overall effectiveness of the IMAC technique to enrich both singly and multiply phosphorylated peptides. We observed ∼1.9-fold and ∼3.6-fold increases in the number of identified phosphopeptides in low-dose and high-dose samples respectively, suggesting both radiation levels stimulate cell signaling pathways. A 6-fold increase in the phosphorylation of cyclin dependent kinase (cdk) motifs was observed after low- dose irradiation, while highdose irradiation stimulated phosphorylation of 3-phosphoinositide-dependent protein kinase-1 (PDK1) and AKT/RSK motifs 8.5- and 5.5-fold, respectively. High- dose radiation resulted in the increased phosphorylation of proteins involved in cell signaling pathways as well as apoptosis while low-dose and control phosphoproteins were broadly distributed among biological processes. Keywords: phosphoproteome profiling • mass spectrometry • normal human skin fibroblast phosphoproteome • low-dose irradiation • high-dose irradiation

Introduction The general population is constantly being exposed to low levels of background radiation, e.g., from cosmic radiation and from naturally occurring radioactive materials in the earth. Currently, the health risks associated with exposure to lowdose ionizing radiation are estimated by extrapolating empirical linear fits of data that correlate with humans exposed to relatively high doses. This procedure is controversial because the cancer risks from low-dose exposure are suggested to be overestimated.1 However, several studies have shown that transcriptional changes and protein modifications occur in response to very low doses of ionizing irradiation.2-5 In addition, some genes have been reported to be changed only at low doses of radiation, but not at high doses.6 Recent cDNA * To whom correspondence should be addressed. Fax: (509) 376-7722. E-mail: [email protected]. † Pacific Northwest National Laboratory. ‡ UT Southwestern Medical Center.

1252

Journal of Proteome Research 2006, 5, 1252-1260

Published on Web 04/14/2006

microarray studies on normal human skin fibroblast (HSF) cells suggest that although a quantitative dose response is predominant, quantitative and qualitative differences in cellular responses to low- and high-dose ionizing radiation coexist.7 While previous studies have aimed to provide a global view of how cells react in response to very low-dose (2 rad) radiation at the gene expression level,7 the relationship between mRNA and protein levels is not necessarily linear, and it is the activity of the protein molecules that are directly responsible for maintaining correct cellular function.8 Therefore, examinations of mRNA and protein levels alone cannot provide critical details of the cellular events involved in DNA repair, DNA damage signaling, and cell cycle checkpoint control, all of which are hallmarks of the radiation response.9,10 Phosphorylation events are likely to play important roles in the prompt cellular response to radiation-induced stress occurring prior to upregulation of protein synthesis. 10.1021/pr060028v CCC: $33.50

 2006 American Chemical Society

Phosphoproteome Profiling of Human Skin Fibroblast Cells

The primary goals of our study are to profile the phosphoproteome of normal human fibroblast cells and determine how irradiation affects the phosphoproteome profiles. Among the challenges that make phosophoproteome analysis difficult are the often low-abundance of phosphorylated proteins, lowstoichiometry of phosphorylation, and the inefficient backbone fragmentation of phosphopeptide ions (e.g., in commercially available ion trap mass spectrometers) due to the generally labile nature of phosphorylation.11 A large number of phosphorylation sites have been identified from different biological samples using various methods. To date, the most extensive data for global phosphoproteome characterization was obtained by pre-fractionation of a whole cell lysate using gel electrophoresis, followed by the strong cation exchange (SCX) liquid chromatography to enrich for phosphopeptides based on their charge differences at pH 2.7.12 In this work, more than 2000 phosphorylation sites were identified through the analysis of SCX fractions by reversed-phase LC and MS/MS. However, with this approach, the majority of the peptides identified were singly phosphorylated; multiply phosphorylated peptides and phosphopeptides with internal basic residues may have been missed in the analysis. In addition, a large sample size (8 mg) and extensive analysis time were required to obtain these impressive results. Another popular method for enriching phosphopeptides utilizes immobilized metal affinity chromatography (IMAC). IMAC with Fe(III) as the ligand has proved useful for enriching phosphopeptides in both simplified and complex proteomic mixtures. With a careful sample handling and methylation protocol,13 IMAC followed by liquid chromatography (LC) coupled to MS can provide high throughput (e.g., detect >1000 phosphopeptide candidates in a single analysis) analyses using small sample amounts (100 µg or less)14 and with a high specificity for phosphopeptides.15 Recently, Kim et al. utilized IMAC to enrich phosphopeptides from HT-29 human colon adenocarcinoma and identified 238 phosphorylation sites from 116 proteins using ∼200 µg protein.16 However, to provide a more complete catalog of phosphorylation events, and to interrogate the potential role of these events in signal transduction, phosphoproteome profiles under different biological conditions are needed. Here, we utilize an Fe(III)-IMAC method to examine the HSF cell phosphoproteome response to both low and high doses of irradiation. Sample preparation was optimized to improve the phosphopeptide binding to the IMAC. To our knowledge, this is the first global characterization of the phosphoproteome of HSF cells and under different biological conditions. Importantly, even though this is a qualitative study, our results suggest potentially significant differences between low- and high-dose irradiation responses exist at the protein phosphorylation level that can provide a basis for future comparative quantitative proteomic work and/or biology studies.

Experimental Procedure Cell Culture. HSF42 cells were maintained in alpha MEM medium (Invitrogen Inc., Carlsbad, CA) supplemented with 10% fetal bovine serum (HyClone, Logan, UT) and antibiotics. Cells nearing 100% confluence in T75 flasks were irradiated at different doses (i.e., 0, 2 rad, and 4 Gy) using a Pantak X-ray generator operated at 320 kV/10 mA. After 2 h, the cells were digested with 0.25% trypin-EDTA, washed twice with PBS, and stored at -80 °C. Protein Sample Preparation Optimization, Digestion, and Methylation. Cells were lysed using Trizol reagent (Invitrogen,

research articles Inc., Carlsbad, CA) and proteins were subsequently isolated according to the manufacturer’s instructions, but with the following optimizations. Since high nucleic acid content in the protein sample competes for binding sites in the IMAC column, protein sample quality following Trizol extraction was carefully monitored using the A260/A280 (absorbance at 260 nm/ absorbance at 280 nm) assay. Protein samples used in our analyses had absorbance ratios of A260/A280 < 1.0, indicating more than ∼95% of the sample was protein and that the majority of the nucleic acid was successfully removed.17 In addition, we found that doubling the Trizol volume (∼2 mL Trizol reagent/5 × 106 cells) and resolubilizing the protein pellet with guandine HCl (6 M) instead of SDS (1%) increased the protein yield up to 2-fold. Protein pellets were re-suspended in 6 M guanidine HCl, then diluted 10-fold with 50 mM NH4HCO3 (pH 7.4). Modified trypsin was added at a trypsin-toprotein ratio of 1:20, and the mixture was incubated overnight at 37 °C. The tryptic digests were desalted using a C18 RP peptide MacroTrap cartridge (Michrom BioResources, Inc., Auburn, CA), and methylated according to the published protocol14 with the exception that the 1 h methylation reaction was repeated one time to ensure complete methylation. IMAC and Capillary HPLC-MS/MS Analysis. Methylated peptides from each sample were resuspended in 100 µL of 0.1% acetic acid (HOAc). Aliquots (∼100 µg) were loaded onto a reversed phase desalting column (360 µm O. D. × 200 µm I. D. fused silica; PolymicroTechnologies) packed with 10 cm of YMC 5-15 µm C18 resin (Waters Inc., Milford, MA), washed with 50 µL of 0.1% HOAc, and then eluted to the IMAC (360 µm O. D. × 200 µm I. D. fused silica; PolymicroTechnologies) column with 20 µL of 40% acetonitrile (MeCN) in 0.1% HOAc. The IMAC column was disconnected from the desalting column, washed with 20 µL of 25/74/1 ACN/water/HOAc containing 100 mM NaCl, 40 µL 0.1% acetic acid, and eluted with 20 µL of 20 mM ascorbic acid18 to a precolumn (360 µm O. D. × 100 µm I. D. fused silica; PolymicroTechnologies) packed with 5 cm of YMC 5-15 µm C18 resin. IMAC replicates were performed for both the low-dose (one replicate) and high-dose samples (2 replicates) and showed high reproducibility with 99%). This result indicates that the IMAC technique efficiently prevents nonspecific bindings by methylation and provides specific enrichment of phosphopeptides. A typical LC-MS analysis using ∼100 µg of starting protein will detect ∼700-1400 phosphopeptide candidates when the neutral loss program is used to scan for the most likely phosphopeptide MS/MS spectrum indicating extensive neutral loss of phosphate (>90% of the most intense fragmentation peak in a given spectrum).14 The number of phosphopeptides detected depends on both the quality and the nature of the sample. We found effective nucleic acid removal from the samples increased the number of phosphopeptides detected. In our analyses, the number of phosphopeptide candidates was observed to increase in low-dose and high-dose treated samples, suggesting that phosphorylation events increase with irradiation stimulation. In addition, the quality of LC separation was observed to have a significant bearing on the detection of phosphopeptides. In our current analysis, a relatively short analytical column (13 cm long with 5 µm C18 particles) was used. Identification of Phosphopeptides. To identify phosphopeptides from HSF cells, all MS/MS spectra passing the neutral loss program were searched against the human database; tryptic peptides were restricted to increase the confidence in identifications. Since phosphopeptides are well-known for their inefficient fragmentation in ion trap mass spectrometers, these peptides may not be effectively identified by the SEQUEST software, i.e., often resulting in low Xcorr and dCn12,14 values. Thus, commonly accepted MS/MS data filtering criteria are not suitable for data analysis with post-translational modifications such as phosphorylation. On the basis of our experience,18 a spectrum with a very low SEQUEST Xcorr value could be correct (as confirmed by biological experiments); however, only spectra with Xcorr g 1.5 were manually examined in the present work because of the large number of spectra that were generated. These results are given in Supporting Information Table 1. Figure 2 shows two representative phosphopeptide spectra. Using the criteria described above, we confidently confirmed the SEQUEST assignment to the MS/MS spectrum in Figure 2A for the phosphopeptide R.RRT*PT*PPPR.R, even though the Xcorr value was well below that acceptable for an unmodified peptide with a 3+ charge state. However, in the case of the spectrum shown in Figure 2B, several issues became apparent

Phosphoproteome Profiling of Human Skin Fibroblast Cells

research articles

Figure 1. Chromatograms (a-d) obtained in the LC-ESI-MS analysis of the IMAC enriched methylated tryptic digest of ∼100 µg proteins from low-dose irradiation-treated cells. (a) Base peak chromatogram; (b) Base peak chromatogram for precursor ions showing the neutral loss of phosphate with m/z 98 in MS/MS; (c) Base peak chromatogram for precursor ions showing the neutral loss of phosphate with m/z 49 in MS/MS; (d) Base peak chromatogram for precursor ions showing the neutral loss of phosphate with m/z 32.7 in MS/MS.

during evaluation of the SEQUEST assignment. First, this spectrum is not of particularly high quality (evidenced by the low intensity and signal/noise ratio). Second, there is an unexplained intense peak at 673.0 m/z. This peak shows a 49 Da m/z difference from the most abundant peak at 722.4 m/z, which is probably due to the first neutral loss from the precursor ion, suggesting a double neutral loss of the phosphate group from the 2+ charge state precursor ion. If this double neutral loss ion assignment is correct, then the sequence assigned to this spectrum should contain at least two phosphorylation sites. On the basis of this reasoning, we ultimately excluded the SEQUEST assignment of K.GFGFGQGAGALVHS*E.- to this spectrum, even though with an Xcorr of 3.44 and dCn of 0.39 for a 2+ peptide, it would have passed the commonly used MS/MS data filtering criteria. SEQUEST assignments similar to the assignment in Figure 2B were not included in our results. More than 95% of the peptides identified after IMAC enrichment were determined to be phosphorylated, consistent with the work of Stover and colleagues15 using a similar IMAC protocol. The number of unique phosphopeptides confirmed from the control, low-dose, and high-dose irradiation-treated samples is 89, 169, and 323 respectively, corresponding to 59, 106 and 268 distinctive proteins and 131, 218, 484 unique phosphorylation sites. A total of 494 unique phosphopeptides (692 phosphorylation sites) and 347 distinctive proteins were identified from the three samples (Supporting Information Table 1). Among the 494 unique phosphopeptides, 232 are singly phosphorylated, 220 are doubly phosphorylated, 41 are triply phosphorylated and 1 peptide has four sequential phospho-

rylations. These numbers demonstrate the overall effectiveness of the IMAC technique for enriching both singly and multiply phosphorylated peptides. Some of the triply and quadruply phosphorylated peptides may be excluded due primarily to the difficulty in confidently assigning a sequence to the associated spectrum. Fragmentation was hindered for multiply phosphorylated peptides because of the multiple neutral losses of phosphate groups. Gygi and co-workers12 used MS3 to further fragment the dephosphorylated species from mostly singly phosphorylated peptides enriched in early fractions by strong cation exchange chromatography. However, this MS3 method may not be as useful for multiply phosphorylated peptide identification with an ion trap mass spectrometer. Due to the nature of multiply phosphorylated peptides, concurrent fragmentation steps using MS3 may not result in sufficiently intense nonphosphorylated fragment ion peaks for effective identification. Identification of these multiply phosphorylated peptides is very important since our data show that ∼50% of the phosphopeptides contain multiple phosphorylations. Since phosphoproteome studies are often qualitative in nature and do not provide a complete delineation of all phosphorylation sites due to false negatives, it is sometimes difficult to make a direct comparison between different conditions based on the presence or absence of a specific phosphorylation site. The lack of a particular phosphorylated peptide could be due to it being missed for a variety of reasons including low abundance, sample complexity, and “under sampling” issues often associated with MS/MS analyses. To identify a potential link between phosphorylation events and radiation treatment, we first performed an analysis of proteins common to all three samples (Table 1). Many of these common Journal of Proteome Research • Vol. 5, No. 5, 2006 1255

research articles

Yang et al.

Table 1. Proteins Found Phosphorylated in All Samples and Their Phosphorylation Sites Identified in Control, Low-Dose, and High-Dose Samples protein

description

control

low-dose

high-dose

IPI00013297.1 28 Kda Heat- And AcidStable Phosphoprotein. IPI00013743.1 Hypothetical Protein.

S57, S60, S63

S57, S60, S63

S57, S60, S63

T174, S175

S197, S201

IPI00017297.1 Matrin 3. IPI00020984.1 Calnexin Precursor. IPI00021405.1 Splice Isoform Lamin A Of P02545 Lamin A/C.

S598, S604 S583 S390, S392, T394, S404, S406, S407

IPI00023640.1 Programmed Cell Death Protein 5. IPI00025512.2 Heat Shock 27 Kda Protein. IPI00027152.1 Centromere Protein A. IPI00029601.2 Src Substrate Cortactin.

S119

T123, S127, T135, S139, T159, S163, S236, S240 S598, S604 S583 S390, S392, S395, S403, S404, S406, S407, T409, T416 S119

S598, S604 S583 S390, S392, T394, S395, S404, S406, S407, T409 S119

S78, S83

S78, S82, S83

S15

S17, S19 T407, S411

S17, S19 T405, T407, S411, T417, S424 S139

S17, S19, T21 T407, S411

S295, S297, S2407, S2409, S2581 S167, S169

IPI00069693.2 Similar To Ribosomal Protein L14. IPI00099730.1 Rna Binding Protein.

S139

IPI00176903.1 Leucine-Zipper Protein Fksg13. IPI00179700.1 High Mobility Group AtHook 1 Isoform A. IPI00216047.1 Similar To Swi/Snf Related, Matrix Associated, Actin Dependent Regulator Of Chromatin, Subfamily C, Member 2. IPI00216312.1 Vimentin.

S169

S295, S297, S2702, S2706 S169

S99, S102, S103

S99, S102, S103

S99, S102, S103

S302, S306

S302, S304

S302, S304, S306

IPI00216919.1 Similar To Yes-Associated Protein, 65 Kda. IPI00219160.1 Ribosomal Protein L34. IPI00295392.3 60s Ribosomal Protein L3. IPI00328293.1 Similar To Serine/Arginine Repetitive Matrix 1.

S295, S300

S71, S72

S38, S55, S71, S72, S418, S419 S103, S105, S109, T110 S61, S105, T110, S127, S128, S138, T141 S12, Y13 S12 S12 S12 S402, T406, S549, S551, S560, S562, S402, T406, S429, T572, T574, S605, S607, S614, S616, S431, S450, T572, S626, S628, S636, S645, S646, S715, T574, S626, S628, S718, S738, S743 S738, S743, T872, S874 S83 S83

IPI00328526.1 Hypothetical Protein Dkfzp451f1219.CampRegulated Phosphoprotein 19. IPI00383404.1 Hypothetical Protein. T206

proteins are highly abundant making it less likely that phosphorylation sites were missed due to the reasons stated above. Some phosphorylation sites are found in all treatments and may represent basal phosphorylation sites or sites regulated by signaling pathways unaffected by radiation. Proteins such as vimentin, Yes-Associated protein, RNA binding protein, cortactin and Lamin A are found in all three samples, but show differences in phosphorylation sites in each treatment. In the case of vimentin, we found seven phosphorylation sites, two are conserved in all samples (S71, S72), one is present in both low- and high-dose samples (S419), two are present in lowdose only (S38, S55) and two are present in high-dose only (S41, S82). The differential phosphorylation of vimentin on S55 in low-dose and S82 in high-dose is interesting because this has recently been shown to occur in a two-step reaction.36 Vimentin is first phosphorylated by Cdk1 on Ser55 creating a binding site for Polo-like kinase 1 and a subsequent phosphorylation on Ser82 to regulate intermediate filament dynamics. Only a few proteins are phosphorylated exclusively in the low- and high-dose samples and in most cases the phosphorylation sites on these proteins differ with the dose of radiation (See Supporting Information Table 1A, B, and C). These results suggest that differential phosphorylation might be occurring in these common proteins and in the phosphoproteome at large with different doses of radiation. As a second approach for determining if distinct signaling pathways are activated by low- and high-dose irradiation, we 1256

Journal of Proteome Research • Vol. 5, No. 5, 2006

T206, T207

S139

S41, S71, S72, S82, S419 S61, T63 S12 S12 S402, T406, S429, S431, S450, S560, S562, T572, T574, S597, S605, S607, T614, S616, S713, T718, S901 S83 T206, T207

analyzed our dataset for the presence of 12 different phosphorylation motifs (Figure 3).23 Proline-directed, casein kinase, and basophilic kinase 1, 2, and 3 were the predominant phosphorylation motifs identified in all three samples. For most motifs, we observed an ∼2- to 3-fold increase in the number of sites identified in low- and high-dose samples, respectively. Phosphorylation of cyclin dependent kinase (CDK) motifs was elevated in the low-dose sample 6-fold (4 peptides identified in control vs 24 peptides in low-dose). PDK1 (8.5-fold), AKT/ RSK (5.5-fold) and CDK (5-fold) motifs showed the greatest fold increase in the high-dose sample. PDK1 is a key mediator of the 3′-phosphoinositide kinase signaling pathway that regulates a number of cellular processes including cell cycle progression, differentiation, gene expression and cell survival. Interestingly, PDK1 is a known activator of AKT suggesting a possible link between these two kinases in the high-dose response. To examine the functional differences of protein activations at different doses of irradiation, we extracted unique gene process categories that are activated at either high- or lowdose treatment (See Table 2 for category breakdown and Supporting Information Table 2 for complete listing). Many of the unique phosphorylated proteins in the high-dose irradiated samples are involved in signal transduction pathways. These include a number of proteins involved in wnt signaling such as catenins as well as GTP regulated signal transduction. Also, two members of the Gab family of scaffolding proteins that play a critical role in the regulation of receptor tyrosine kinase

research articles

Phosphoproteome Profiling of Human Skin Fibroblast Cells

Figure 2. Manual validation of two phosphopeptide MS/MS spectra assigned by SEQUEST analysis.

signaling27 were phosphorylated in the high-dose sample. DNA damage is a consequence of high-dose radiation and several of the phosphoproteins identified have direct roles in the DNAdamage response. These include SMC5 which is phosphorylated on Thr771/Thr774 and plays a major role in multiple DNA repair and DNA damage response pathways.29 Five proteins shown to have a direct role in the regulation of apoptosis are phosphorylated in the high-dose samples compared to only one protein in low-dose. The increased phosphorylation observed following high-dose radiation treatment is consistent with the activation of cell signaling pathways. Increased kinase activity of ATM, ATR, Chk1/2, JNK and other kinases occurs in the presence of DNA damage. While we did not detect phosphorylation of these kinases in our study, several kinases were found to be phosphorylated in the high-dose compared to one in low-dose (Cdk1) and none in control samples, suggesting a further role of kinase phosphorylation in regulating the response to radiation treatment. The phosphorylated kinases found in the highdose sample include Protein Kinase C Delta (PKCDelta), MAP kinase kinase kinase 4 (MAPKKK4), LIM domain kinase 1, cdc2-

like 5, and the rho-associated kinase, P160 ROCK. PKCDelta is required to mount an apoptotic response to DNA damage through a poorly understood pathway that may involve Chk1 and P53.24 MAPKKK4 phosphorylation is consistent with the observed activation of multiple p38 MAP kinase signaling pathways in response to radiation and genotoxic stress.25 LIM domain kinase 1 and P160Rock are regulators of the actin cytoskeleton components of which are phosphorylated following radiation treatment. Moreover, P160 ROCK is activated during apoptosis and is necessary for membrane blebbing, redistribution of fragmented DNA into blebs, and formation of apoptotic bodies.26 In contrast to the high-dose treatment, low-dose irradiation resulted in no dominant gene process categories (Table 2). However, analysis of the individual proteins phosphorylated in the low-dose sample revealed some potentially interesting phosphorylation events. The p53 tumor suppressor protein is a key regulator of cell cycle progression, DNA damage response and apoptosis and is phosphorylated on multiple sites following a variety of cell stresses including radiation treatment.37 In the present study, we identified a quadruply phosphorylated pepJournal of Proteome Research • Vol. 5, No. 5, 2006 1257

research articles

Yang et al.

Figure 3. Survey of phosphorylation motifs. Primary sequences surrounding the identified phosphorylation sites in each phosphopeptide were queried against distinct kinase motifs and revealed potential kinase activities. Table 2. Numbers of Phosphorylated Proteins by Biological Process Category Uniquely Found in Control, Low-Dose, and High-Dose Samples category

metabolism signal transduction regulation of biological process response to stimulus protein modification transport cell organization and biogenesis cell communication development other biological process biological process unknown cellular process regulation of progression through cell cycle small GTPase mediated signal transduction apoptosis chromosome organization and biogenesis negative regulation of biological process positive regulation of biological process G-protein coupled receptor protein signaling pathway negative regulation of protein metabolism protein kinase cascade total

control

4 5 2 1 2 1 1 1 3

low-dose

high-dose

10 2 6 3 1 4 1 1 3 1 1 3 2

39 25 22 17 15 12 11 10 9 8 7 7 7 6

1

1 1

5 5

2

5 5

1

1

3 2

22

43

2 224

tide from p53 that includes Ser315 in the low-dose sample, a site known to be phosphorylated in the presence of DNA damage. Cdk1, one of the key regulators of progression through mitosis is found to be phosphorylated on Thr14/Tyr15 during low-dose treatment which is associated with inhibition of its 1258

Journal of Proteome Research • Vol. 5, No. 5, 2006

activity35 and this could explain the growth arrest associated with DNA damage. The Yes-Associated Protein is phosphorylated in vivo by AKT on Ser127 which recruits 14-3-3 proteins and suppresses the induction of pro-apoptotic genes in the presence of DNA damage.34 We identified phosphorylation of Ser127 in the low-dose, but not high-dose or control samples suggesting a possible mechanism for the attenuation of apoptosis observed in low- versus high-dose treatments. We also identified a potential link to translational control of protein expression following low-dose treatment. It is well understood that cells respond to stress stimuli by altering protein levels through the activation of transcription pathways. Regulation of protein expression at the protein translation level is another important control mechanism that is less understood.30 A recent study demonstrated that a series of ordered phosphorylation events involving mTOR and ribosomal S6 kinase, S6K1, regulate eukaryotic translation initiation factors.31 This study found that phosphorylation of eukaryotic translation initiation factor 4B (eIF4B) on Ser422 results in the efficient translation of a subset of target proteins. A separate study found that translation up-regulation of tousled-like kinase by this pathway plays a key role in an adaptive response to radiation.32 In our low-dose sample we found 4 eIF4B peptides containing a phosphorylated Ser422 either alone or doubly phosphorylated on surrounding amino acids as well as an additional eIF4B phosphopeptide (Ser498/Ser504). Another component of this translation regulatory pathway, eukaryotic translation initiation factor 4 gamma 1, was found exclusively in the low-dose sample suggesting that regulation of translation may play a key role in the low-dose cellular response. The identification of multiple phosphopeptides for eIF4B in the low-dose compared to none in control and high-dose samples increases our confidence that eIF4B phosphorylation is indeed elevated by low-dose irradiation.

research articles

Phosphoproteome Profiling of Human Skin Fibroblast Cells

Differential phosphorylation of cytoskeletal proteins and their regulators are observed in both the low-dose and highdose samples and could reflect changes in cellular architecture when cells enter into mitosis or apoptosis. During the final stages of apoptosis, cell contraction, and membrane blebbing occur that are associated with altered regulation of the cytoskeleton. Phosphorylation of several actin associated proteins including Cortactin, Drebrin, Myosin, Coronin (phosphorylated in control, but not irradiated samples) is seen suggesting alterations in actin cytoskeletal function may occur following radiation treatment. In addition to vimentin, the intermediate filament protein, nestin is phosphorylated in both low- and high-dose samples (Ser345/Ser351). Nestin plays a role in regulating intermediate filament protein dynamics during mitosis through interactions with phosphorylated vimentin.33 A number of microtubule-associated proteins including Map1b, Map4, Cytoplasmic Dynein heavy chain 2, axonemal dynein heavy chain, Kinesin-like protein KIF14, Septin 2 are also differentially phosphorylated. Stathmin, a regulator of microtubule dynamics during mitosis, is phosphorylated on Ser25 (found in both low-dose and control samples) via Apoptosis signal-regulating kinase 1 (Ask1).

Conclusions and Future Perspective The present study has shown that Fe (III)-IMAC with LCMS is an efficient method for phosphoproteome analysis. A total of 692 phosphorylation sites were assigned to 347 proteins. In some cases, a clear cellular function has been associated with the phosphorylation sites we identify here, providing possible insights into the role of phosphorylation following radiation treatment. However, many of the phosphorylation sites identified in this study have not been previously characterized. Therefore, the identification of these novel phosphorylation events may shed light on the regulatory mechanisms involved in cell cycle regulation, DNA damage response, apoptosis, cancer progression as well as other important cellular functions. Importantly, the phosphorylation profile for proteins identified from the low-dose (2 rad) irradiated HSF cells was shown to be different from the profile obtained for proteins irradiated at the high-dose (4 Gy) level. This observation is significant since it suggests that both low- and highdoses of radiation might affect distinct signaling pathways. These dose-dependent radiation-responses may form a basis for identifying relevant molecular markers that can be used in the future to better evaluate human health risks at low doses of irradiation and to develop low-dose radiation counter measurements. This powerful proteomic approach provides an initial view of the phosphorylation events in response to low- and highdose radiation and a basis for determining potential functions for site-specific protein phosphorylation. However, given the difficulties in performing phosphoproteomic analyses, the results presented here (and elsewhere) also represent an incomplete view of the total phosphoproteome. The potential differences suggested by the present study need to be further validated by future more quantitative proteomic studies (e.g., using stable isotope labeling) in conjunction with statistically guided biology studies. Additionally, improved methods for detecting low abundance phosphorylation events and quantifying the relative phosphorylation status of each protein under different conditions are clearly desirable. Moreover, validation of phosphoproteomic analyses will require, e.g., the development of immunological reagents specific for important phos-

phorylation sites. Clearly, improved high throughput approaches for phosphorylation validation and site-specific antibody production will be beneficial for both allowing the desired depth of proteome coverage and the analysis of significant numbers of biological replicates. Further improvements to the present measurements are anticipated. For example, electron capture dissociation (ECD)38 and electron-transfer dissociation (ETD)39 are two fragmentation techniques that have been used to successfully sequence phosphopeptides without neutral loss of phosphate on most fragmentation ions, with the latter offering greater efficiency and utility with ion trap mass spectrometers. These methods can provide the identification of presented missed phosphopeptides that can then be subsequently identified in higher throughput analyses exploiting accurate mass measurements. In addition to qualitative analysis, quantitative changes need to be monitored to better understand the difference between low- and high-dose irradiation effects. Additionally, the implementation of a significantly improved separations (e.g., using long packed or monolithic columns with integrated electrospray emitter tips and having smaller inner diameters)40 are expected to provide benefits that include an improved dynamic range, phosphoproteome coverage, and sensitivity for quantitative phosphopeptide analyses. Finally, once a large set of phosphopeptides has been confidently identified (and their separation times determined), the application of the accurate mass and tag approach42 will further increase both the throughput and sensitivity of measurements, largely eliminate the under-sampling of MS/MS approaches, and will likely improve quantitation.

Acknowledgment. This research was supported by the U.S. Department of Energy (DOE) Office of Biological and Environmental Research (BER) Low Dose Radiation program at the Pacific Northwest National Laboratory (PNNL), a multiprogram national laboratory operated by Battelle for the DOE under Contract DE-AC05-76RL01830. Experiments were performed in the Environmental Molecular Sciences Laboratory, a DOE national scientific user facility at PNNL, sponsored by BER. We would like to acknowledge Mr. Mark L. McCleland at the University of Virginia and Prof. Sandra Rossie at Purdue Univeristy for providing useful information and assistance with sample preparation. Supporting Information Available: The complete list of unique phosphopeptides and proteins identified from control, low-dose, and high-dose samples and literature search12,41 for the phosphorylation sites reported are provided as Supporting Information Table 1A, B, and C, respectively. The complete list of gene categories for control (C), Low-dose (L) and High-dose (H) proteins according to the Gene Ontology (GO) database obtained from the European Bioinformatics Institute is provided in Supporting Information Table 2. This material is available free of charge via the Internet at http:// pubs.acs.org. References (1) Brooks, A. L. Developing a scientific basis for radiation risk estimates: goal of the DOE Low Dose Research Program. Health Phys. 2003, 85, 85-93. (2) Diano, K.; Ichimura, S.; Nenoi, M. Early induction of CDKN1A (p21) and GADD45 mRNA by a low dose of ionizing radiation is due to their dosedependent posttranscriptional regulation. Radiat Res. 2002, 157, 478-482.

Journal of Proteome Research • Vol. 5, No. 5, 2006 1259

research articles (3) Fornace, A. J.; Amundson, S. A.; Do, K. T.; Meltzer, P.; Trent, J.; Bittner, M. Stress-gene induction by low-dose gamma irradiation. Mil. Med. 2002, 167, 13-15. (4) Mothersill, C.; Lyng, F.; Mulford, A.; Seymour, C.; Cottell, D.; Lyons, M.; Austin, B. Effect of low doses of ionizing radiation on cells cultured from the hematopoietic tissue of the Dublin Bay prawn, Nephrops norvegicus. Radiat Res. 2001, 156, 241-250. (5) Sutherland, B. M.; Bennett, P. V.; Cintron-Torres, N.; Hada, M.; Trunk, J.; Monteleone, D.; Sutherland, J. C.; Laval, J.; Stanislaus, M.; Gewirtz, A. Clustered DNA damages induced in human hematopoietic cells by low doses of ionizing radiation. . Radiat Res. (Tokyo) 2002, 43 Suppl., S149-152. (6) Suzuki, K.; Kodama, S.; Watanabe, M. Extremely low-dose ionizing radiation causes activation of mitogen-activated protein kinase pathway and enhances proliferation of normal human diploid cells. Cancer Res. 2001, 61, 5396-5401. (7) Ding, L. H.; Shingyoji, M.; Chen, F. Q.; Hwang, J.-J.; Burma, S.; Cheng, J.-F.; Chen, D. J. Gene Expression Profiles of Normal Human Fibroblasts After Ionizing Radiation: A Comparative Study of Low and High Doses. Radiation Research 2005, 164 (1), 17-26. (8) Gygi, S. P.; Rochon, Y.; Franza, B. R.; Aebersold, R. Correlation between Protein and mRNA Abundance in Yeast. Mol. Cell. Biol. 1999, 19, 1720-1730. (9) Park, S. J.; Ciccone, S. L.; Freie, B.; Kurimasa, A.; Chen, D. J.; Li, G. C.; Clapp, D. W.; Lee, S. H. A positive role for the Ku complex in DNA replication following strand break damage in mammals. J. Biol. Chem. 2003, 279, 6046-6055. (10) Gerion, D.; Chen, F.; Kannan, B.; Fu, A.; Parak, W. J.; Chen, D. J.; Majumdar, A.; Alivisatos, A. P. Room-temperature single-nucleotide polymorphism and multiallele DNA detection using fluorescent nanocrystals and microarrays. Anal. Chem. 2003, 75, 4766-4772. (11) DeGnore, J. P.; Qin, J. Fragmentation of phosphopeptides in an ion trap mass spectrometer. J. Am. Soc. Mass Spec. 1998, 9, (11), 1175-1188. (12) Beausoleil, S. A.; Jedrychowski, M.; Schwartz, D.; Elias, J. E.; Villen, J.; Li, J.; Cohn, M. A.; Cantley, L. C.; Gygi, S. P. Large-scale characterization of HeLa cell nuclear phosphoproteins. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12130-12135. (13) Brill, L. M.; Salomon, A. R.; Ficarro, S. B.; Mukherji, M.; StettlerGill, M.; Peters, E. C. Robust Phosphoproteomic Profiling of Tyrosine Phosphorylation Sites from Human T Cells Using Immobilized Metal Affinity Chromatography and Tandem Mass Spectrometry Anal. Chem. 2004, 76 (10), 2763-2772. (14) Ficarro, S. B.; McCleland, M. L.; Stukenberg, P. T.; Burke, D. J.; Ross, M. M.; Shabanowitz, J.; Hunt, D. F.; White, F. M. Phosphoproteome Analysis by Mass Spectrometry and its Application to Saccharomyces cerevisiae. Nat. Biotechnol. 2002, 20, 301-305. (15) Stover, D. R.; Caldwell, J.; Marto, J.; Root, R.; Mestan, J.; Stumm, M.; Ornatsky, O.; Orsi, C.; Radosevic, N.; Liao, L.; Fabbro, D.; Moran, M. F. Differential phosphoprofiles of EGF and EGFR kinase inhibitor-treated human tumor cells and mouse xenografts. Clin. Proteomics 2004, 1, 069-080. (16) Kim, J.-E.; Tannenbaum, S. R.; White, F. M. Global Phosphoproteome of HT-29 Human Colon Adenocarcinoma Cells J. Proteome Res. 2005, 4 (4), 1339-1346. (17) Bollag, D. M.; Rozycki, M. D.; Edelstein, S. J. Protein Methods; Wiley: 1996; p 135. (18) Sweeney, F. D.; Yang, F.; Chi, A.; Shabanowitz, J.; Hunt, D. F.; Durocher, D. Saccharomyces cerevisiae Rad9 Acts as a Mec1 Adaptor to Allow Rad53 Activation. Curr. Biol. 2005, 15, 13641375. (19) Martin, S. E.; Shabanowitz, J.; Hunt, D. F.; Marto, J. A. Subfemtomole MS and MS/MS Peptide Sequence Analysis Using NanoHPLC Micro-EST Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. Anal. Chem. 2000, 72, 4266-4274. (20) Corbin, R. W.; Paliy, O.; Yang, F.; Shabanowitz, J.; Platt, M.; Lyons, C. E. J.; Root, K.; McAuliffe, J.; Jordan, M. I.; Kustu, S.; Soupene, E.; Hunt, D. F. Toward a protein profile of Escherichia coli: Comparison to its transcription profile. PNAS 2003, 100, 92329237. (21) Eng, J. K.; McCormack, A. L.; Yates, J. R., An Approach to Correlate Tandem Mass Spectral Data of Peptides with Amino Acid Sequences in a Protein Database. J. Am. Soc. Mass Spectrom. 1994, 5 (11), 976-989.

1260

Journal of Proteome Research • Vol. 5, No. 5, 2006

Yang et al. (22) Yates, J. R.; Eng, J. K.; McCormack, A. L.; Schieltz, D. Method to Correlate Tandem Mass Spectra of Modified Peptides to Amino Acid Sequences in the Protein Database. Anal. Chem. 1995, 67 (8), 1426-1436. (23) Ballif, B. A.; Villen, J.; Beausoleil, S. A.; Schwartz, D.; Gygi, S. P. Phosphoproteomic analysis of the developing mouse brain. Mol. Cell. Proteomics 2004, 3, 1093-1101. (24) Santiago-Walker, A. E.; Fikaris, A. J.; Kao, G. D.; Brown, E. J.; Kazanietz, M. G.; Meinkoth, J. L. Protein kinase C delta stimulates apoptosis by initiating G1 phase cell cycle progression and S phase arrest. J. Biol. Chem. 2005, 280 (37), 32107-32114. (25) Dent, P. Y. A.; Fisher, P. B.; Hagan, M. P.; Grant, S. MAPK pathways in radiation responses. Oncogene 2003, 22 (37), 58855896. (26) Coleman, M. L.; Sahai, E. A.; Yeo, M.; Bosch, M.; Dewar, A.; Olson, M. F. Membrane blebbing during apoptosis results from caspasemediated activation of ROCK I. Nat. Cell Biol. 2001, 3 (4), 339345. (27) Gu, H.; Neel, B. G. The “Gab” in signal transduction. Trends Cell Biol. 2003, 13 (3), 122-130. (28) Szumiel, I. Monitoring and signaling of radiation-induced damage in mammalian cells. Rad. Res. 1998, 150, S92-101. (29) Lehmann, A. R. The role of SMC proteins in the responses to DNA damage. DNA Repair (Amst). 2005, 4 (3), 309-314. (30) Holcik, M.; Sonenberg, N. Translational control in stress and apoptosis. Nat. Rev. Mol. Cell Biol. 2005, 6 (4), 318-327. (31) Holz, M. K.; Ballif, B. A.; Gygi, S. P.; Blenis, J. mTOR and S6K1 mediate assembly of the translation preinitiation complex through dynamic protein interchange and ordered phosphorylation events. Cell. 2005, 123 (4), 569-580. (32) Sunavala-Dossabhoy, G.; Fowler, M.; De Benedetti, A. Translation of the radioresistance kinase TLK1B is induced by gammairradiation through activation of mTOR and phosphorylation of 4E-BP1. BMC Mol. Biol. 2004, 5, 1. (33) Chou, Y. H.; Khuon, S.; Herrmann, H.; Goldman, R. D. Nestin promotes the phosphorylation-dependent disassembly of vimentin intermediate filaments during mitosis. Mol. Biol. Cell. 2003, 14 (4), 1468-1478. (34) Basu, S.; Totty, N. F.; Irwin, M. S.; Sudol, M.; Downward, J. Akt phosphorylates the Yes-associated protein, YAP, to induce interaction with 14-3-3 and attenuation of p73-mediated apoptosis. Mol Cell. 2003, 11 (1), 11-23. (35) Gould, K. L.; Nurse, P. Tyrosine phosphorylation of the fission yeast cdc2+ protein kinase regulates entry into mitosis Nature 1989, 342, 39-45. (36) Yamaguchi, T.; Goto, H.; Yokoyama, T.; Sillje, H.; Hanisch, A.; Uldschmid, A.; Takai, Y.; Oguri, T.; Nigg, E. A.; Inagaki, M. Phosphorylation by Cdk1 induces Plk1-mediated vimentin phosphorylation during mitosis J. Cell Biol. 2005, 171 (3), 431-436. (37) Bode, A. M.; Dong, Z. Posttranslational modification of p53 in tumorigenesis. Nat. Rev. Cancer 2004, 4 (10), 793-805. (38) Shi, S. D.; Hemling, M. E.; Carr, S. A.; Horn, D. M.; Lindh, I.; McLafferty, F. W. Phosphopeptide/phosphoprotein mapping by electron capture dissociation mass spectrometry. Anal. Chem. 2001, 73, 19-22. (39) Syka, J. E.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt, D. F. Peptide and protein sequence analysis by electron-transfer dissociation mass spectrometry. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528-9533. (40) Luo, Q.; Shen, Y.; Hixson, K. K.; Zhao, R.; Yang, F.; Moore, R. J.; Mottaz, H. M.; Smith, R. D. Preparation of 20-m-i.d. Silica-Based Monolithic Columns and Their Performance for Proteomics Analyses Anal. Chem. 2005, 77 (15), 5028-5035. (41) Diella, F.; Cameron, S.; Gemu ¨ nd, C.; Linding, R.; Via, A.; Kuster, B.; Sicheritz-Ponte´n, T.; Blom, N.; Gibson, T. J. Phospho. ELM: A database of experimentally verified phosphorylation sites in eukaryotic proteins. BMC Bioinform. 2004, 5 (1), 79. (42) Zimmer, J. S.; Monroe, M. E.; Qian, W.-J.; Smith, R. D. Advances in proteomics data analysis and display using an accurate mass and time tag approach. Mass Spectrom. Rev. 2006, 25 (3), 450-482.

PR060028V