Department of Physics, Chemistry and Biology
Molecular Biotechnology Division
Final Thesis
Enrichment strategy development for
phosphoproteome analysis of saccharomyces
cerevisiae
Pontus Lundemo
LITH-IFM-A-EX--09/2040--SE
Department of Physics, Chemistry and Biology Linkoping University
Department of Physics, Chemistry and Biology
Enrichment strategy development for
phosphoproteome analysis of saccharomyces
cerevisiae
Pontus Lundemo
2009-03-02
Supervisor:
Timothy J Griffin
Examiner:
Maria Sunnerhagen
Upphovsrätt
Detta dokument hålls tillgängligt på Internet – eller dess framtida ersättare – under 25 år från publiceringsdatum under förutsättning att inga extraordinära omständigheter uppstår.
Tillgång till dokumentet innebär tillstånd för var och en att läsa, ladda ner, skriva ut enstaka kopior för enskilt bruk och att använda det oförändrat för ickekommersiell forskning och för undervisning. Överföring av upphovsrätten vid en senare tidpunkt kan inte upphäva detta tillstånd. All annan användning av dokumentet kräver upphovsmannens medgivande. För att garantera äktheten, säkerheten och tillgängligheten finns lösningar av teknisk och administrativ art.
Upphovsmannens ideella rätt innefattar rätt att bli nämnd som upphovsman i den
omfattning som god sed kräver vid användning av dokumentet på ovan beskrivna sätt samt skydd mot att dokumentet ändras eller presenteras i sådan form eller i sådant sammanhang som är kränkande för upphovsmannens litterära eller konstnärliga anseende eller egenart.
För ytterligare information om Linköping University Electronic Press se förlagets hemsida http://www.ep.liu.se/.
Copyright
The publishers will keep this document online on the Internet – or its possible replacement – for a period of 25 years starting from the date of publication barring exceptional
circumstances.
The online availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for his/hers own use and to use it unchanged for non-commercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility.
According to intellectual property law the author has the right to be mentioned when his/her work is accessed as described above and to be protected against infringement.
For additional information about the Linköping University Electronic Press and its procedures for publication and for assurance of document integrity, please refer to its www home page: http://www.ep.liu.se/.
A
BSTRACT
The reversible phosphorylation of proteins is central to regulating most aspects of cell function. Malfunction in this critical cellular process have been implicated to cause diseases such as diabetes, cancer, and Alzheimer’s. Recent advances in mass spectrometry have made it possible to study this important post translational modification on a proteome-wide scale. However, to be able to do so, enrichment of phosphorylated peptides is required. Pairwise comparison of individual steps in an enrichment procedure and simultaneous improvement of data analysis resulted in a protocol which allowed high confidence identification of 2,131 unique phosphorylated peptides from 1,026 proteins. Thereby not only establishing a working protocol for phosphopeptide enrichment in the Griffin Lab, but also generating the largest list of proteins phosphorylated under normal conditions in yeast to date.
2DE - 2-dimensional gel electrophoresis
ACN - Acetonitrile
AcOH - Acetic acid
AMU - Atomic mass unit
BCA - Bicinchoninic acid
CID - Collision induced dissociation
EDC - 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide
hydrochloride
EDTA - Ethylenediaminetetraacetic acid
ESI - Electrospray ionization
ETD - Electron transfer dissociation
FA - Formic acid
FDR - False discovery rate
FT - Fourier transformation
ICR - Ion cyclotron
IDA - Imminodiacetric acid
IMAC - Immobilized metal affinity chromatography
LTQ - Linear ion-trap quadrupole
m/z - Mass-to-charge ratio
MALDI - Matrix-assisted laser desorption/ionization
MS - Mass spectrometry
MS/MS - Tandem mass spectrometry
MSA - Multistage activation
NHS - H-hydroxysuccinimide
NL - Neutral loss
NTA - Nitrilotriacetic acid
OSCC - Oral squamous cell carcinoma
PBS - Phosphate buffered saline
pI - Isoelectric point
ppm - Parts per million
PTM - Post-translational modification
PTP - Protein tyrosine phosphatase
RF - Radio frequency
SCX - Strong cation exchange
SDS - Sodium Dodecyl sulfate
TCEP - tris(2-carboxyethyl)phosphine
TFA - Trifluoroacetic acid
TPP - Trans-proteome pipeline
Tris - tris(hydroxymethyl)aminomethane
1
C
ONTENTS
Abstract ... 3 2 Introduction ... 7 2.1 Background ... 7 2.2 Previous studies ... 7 2.3 Aim ... 8 3 Theory ... 9 3.1 Proteomics ... 9 3.2 Mass spectrometry ... 10 3.2.1 Instrument ... 10 3.2.2 Output ... 11 3.2.3 Database searching ... 12 3.3 BCA assay ... 13 3.4 Phosphorylation ... 13 3.5 Enrichment methods ... 14 3.5.1 ProQ enrichment ... 143.5.2 Strong cation exchange ... 14
3.5.3 Immobilized metal affinity chromatography ... 15
3.5.4 Methyl esterification ... 15
3.5.5 Other methods ... 16
3.6 Fractionation methods ... 16
3.6.1 Strong cation exchange ... 16
3.6.2 OFFGEL- Isoelectric focusing ... 16
3.6.3 Ethanolamide modification ... 16
4 Experimental details ... 18
4.1 Approach... 18
4.2 Yeast growth and lysis ... 18
4.3 ProQ enrichment ... 18
4.4 Trypsin digestion ... 19
4.5 Desalting ... 19
4.6 Strong cation exchange (SCX) ... 19
4.8 Ethanol amide modification ... 20
4.9 Esterification ... 20
4.10 IMAC ... 20
4.11 MS analysis ... 20
4.12 Database searching ... 21
5 Results and Discussion ... 22
5.1 Phosphoproteins identified ... 22 5.2 Enrichment method ... 24 5.3 Fractionation method ... 25 5.4 Chemical modification ... 25 5.5 Search algorithm ... 26 5.6 Sequest parameters ... 27 6 Conclusions ... 29 7 Future ... 30 7.1 Future directions ... 30 7.2 Future applications ... 30 8 Acknowledgement... 31 9 References ... 32
7
2
I
NTRODUCTION
2.1 B
ACKGROUNDThe term proteome was first used in an article 1995 (Wasinger, et al., 1995), revealing that proteomics is a relatively new field. It does not only include the identification and
quantification of proteins, but also the determination of their localization, modifications, interactions, activities and functions (Fields, 2001). This is indeed a very challenging task since the approximately 30,000 genes in humans can result in more than 100 times as many different proteins due to RNA processing and Post-translational modifications (PTMs) (O'Donovan, et al., 2001). One of the most abundant, and also most important,
modifications is phosphorylation. Phosphorylation/dephosphorylation reactions are one of the dynamic mechanisms through which cells modulate protein activity in response to environmental stimuli and it is estimated that up to 30 % of all cellular proteins are
phosphorylated at any given time. Furthermore, dysfunctional phosphorylation have been implied to cause diseases such as diabetes, cancer, Alzheimer’s and a large number of kinases, phosphatases and their substrates are potential drug targets (Cohen, 2002). Although phosphorylation is an intensively studied PTM, most studies have focused on its regulatory function on a single protein basis. Since phosphorylation often occurs in complex signaling networks linking many proteins and phosphorylation sites (Laviola, et al., 2007) (Nollau, et al., 2003) the study of phosphorylation through proteomics is an intriguing possibility.
To be able to study phosphorylation with proteomics several complications have to be addressed. The present method of choice for large scale proteomics is mass spectrometry, which puts several demands on the sample analyzed. First of all, the sample needs to be ionized. Phosphorylated peptides are harder to ionize than unmodified, due to their additional negative charges. Also, phosphorylated peptides are often in low stoichiometry ratios and are therefore overshadowed by their unmodified counterparts. Hence to detect phosphorylated peptides with mass spectrometry, they have to be enriched. Furthermore, the model organism used in this study, saccharomyces cerevisiae, expresses around 6,000 proteins (Goffeau, et al., 1996). Since a mass spectrometer can only analyze a limited amount of peptides at a time, the sample will need to be fractionated before introduced to the mass spectrometer.
2.2 P
REVIOUS STUDIESSaccharomyces cerevisiae, budding yeast, is used in this study mainly for two reasons. First
of all, it is a model organism for eukaryotic cells which is simple to grow and generate a lot of protein. Secondly, several phosphoproteomics analyses of budding yeast have been
conducted, allowing comparison of the methods success. In 2002 Ficarro et al. identified 383 phosphorylation sites by methyl esterification prior to immobilized metal affinity
8
used a combination of strong cation exchange (SCX) and IMAC to find 729 phosphorylation sites (Gruhler, et al., 2005). Chi et al. identified 1252 phosphorylation sites on 629 proteins with endo-Lys C as proteolytic enzyme, IMAC enrichment and an alternative fragmentation method for MS/MS (Chi, et al., 2007). The same year Gygi et al. identified 2288
phosphorylation sites from 985 proteins (Li, et al., 2007), which is the largest yeast phosphoprotein dataset described so far. When looking at an overview of these studies (Table I), it is apparent that the most successful methods are diverse and that no consensus about optimal strategy exists.
Table I, overview of previously reported studies of the yeast phosphoproteome, showing what enrichment strategy, mass spectrometer, fragmentation type and search algorithm was used.
2.3 A
IMThis thesis will investigate methods for enrichment of phosphoproteins, to be used prior to mass-spectrometric analysis. The model organism used will be saccharomyces cerevisiae and the goal is to get an enrichment strategy working in the Griffin lab at least as efficiently as described previously in any other lab. There are two separate parts that need to be addressed to reach this goal. The first is to find an experimental protocol that leads to as high enrichment and fractionation of the sample as possible. The second part is equally important, and consists of optimizing the data analysis to be able to identify most of the phosphopeptides introduced into the mass spectrometer. Hopefully the same enrichment protocol can be applied to more biologically relevant yeast samples as well as to other biological samples, saliva in particular. In the long run qualitative studies of changes in phosphorylation can complement quantitative studies in the search for biomarkers of oral cancer.
9
3
T
HEORY
3.1 P
ROTEOMICSThe early years of proteomics was shaped by 2-dimensional gel electrophoresis (2DE) (Klose, 1975). It gives a visual representation of the proteome through separation of proteins by isoelectric point (pI) and molecular weight. Essentially it remained a descriptive technique until matrix-assisted laser desorption/ionization (MALDI) (Tanaka, et al., 1988) and
Electrospray ionization (ESI) (Fenn, et al., 1989) were invented in the late 80’s. By then e.g. PTMs, detected as spot-trains on a 2DE (Figure 1) could be analyzed by MS to identify the corresponding peptide sequence. However, the gel based method faced a series of issues, including limited sample capacity, low detection sensitivity, difficulty to separate
transmembrane proteins and limited throughput. Furthermore the method is difficult and high labor and therefore has no potential for full automation (Gygi, et al., 2000). Even though some of these issues have been addressed by the use of fluorescent stains, zoom gels and automated robots, another approach is now more common, based on Mass spectrometry (MS).
Figure 1, 2DE, a technique frequently used in the early years of proteomics. Spot trains indicate isoforms of the same protein (Carrette, et al., 2006)
10
3.2 M
ASS SPECTROMETRY3.2.1
I
NSTRUMENTBy definition, a mass spectrometer (MS) consists of an ion source, a mass analyzer that separates the ions by mass-to-charge ratio (m/z), and a detector that registers the number of ions at each m/z value. For biological samples it is important to have a soft ionization method, otherwise the sample peptide is fragmented too much, into pieces that give no information about the original composition of the sample. Therefore electrospray ionization (ESI) is commonly used, which ionizes the peptides by pushing them out of a thin capillary, dissolved in a volatile solvent, to create an aerosol. The aerosol enters a strong electric field that accelerates the ions towards the mass analyzer.
For a mass analyzer there are several important properties such as sensitivity, resolution, mass accuracy and spectral recording speed. Sensitivity tells how small an amount of sample that can be detected and resolution is defined as the degree of separation between peaks. Resolution is important in proteomics because a big number of peptides are introduced to the mass spectrometer at the same time, and the mass analyzer has to be able to tell them apart. For the same reason, speed is an important property. When many peptides are introduced continuously a slow mass analyzer will not have time to look at every individual peptide. The last important property is mass accuracy, which is measured relative to the mass of the analyte in parts per million (ppm) or in actual atomic mass units (AMU). High mass accuracy increase confidence in peak assignment since the probability of ambiguous peak assignment in the database searching is decreased. (Patterson, et al., 2000)
Figure 2, experimental setup of the mass spectrometer used. a) Electrospray ionization (ESI) to ionize the peptides b) Linear ion trap (LTQ) used for MS/MS measurements c) C-trap required to get ions into the Orbitrap d) Orbitrap mass analyzer e) Fourier transformation (FT) to get data for every peptide. Modified
from (Scigelova, et al., 2006)
In this study a linear ion-trap quadrupole (LTQ) Orbitrap MS (Thermo) was used, seen schematically in Figure 2. It is a hybrid instrument with both an LTQ and an Orbitrap mass analyzer. The ion-trap gathers ionized peptides in an electric field for a short period of time and ejects the ions sequentially by changing radio frequency (RF) fields to a frequency were
a) b) c)
d) e)
11
they are unstable. The Orbitrap is the newest type of mass analyzer. Basically it is an ion-trap, but without RF- or magnetic-fields. Instead ions are trapped around an electrode, where the centrifugal force from their initial velocity compensates for the electrostatic interaction with the electrode. The ions will get an axial oscillation frequency that is
dependent on the ions mass. The frequency of the oscillation can be measured as an image current on the two halves of an electrode encapsulating the Orbitrap. Since more than one ion is trapped at a time, Fourier transformation is needed to extract data for individual ions. These measurements obtain very high resolution data, comparable to data from FT-ICR instruments. (Scigelova, et al., 2006)
3.2.2
O
UTPUTThe output from the mass spectrometer is called a mass-spectrum. For protein
identification, tandem mass spectrometry (MS/MS) is needed to get confident data. In MS/MS a peak from the MS spectrum is selectively fragmented and analyzed again, to get more information of the composition of the peptide. There are several methods for fragmentation but the most common is collision induced dissociation (CID) in which the selected peptide is allowed to collide with a neutral molecule and thereby fragment into smaller ions. Cleavage mainly occurs at the amide bond. If the charge from the peptide ends up on the N-terminal part the generated ion is called a b-ion, and if the charge ends up on the C-terminal part the ion is called a y-ion. When several connected b- and y-ions are identified an amino acid sequence is acquired, that can be used to identify the peptide and eventually what protein the peptide was derived from. An example of the output from every step can be seen in Figure 3. MS spectra are acquired using the Orbitrap and MS/MS spectra are acquired in the LTQ. The LTQ have lower mass accuracy than the Orbitrap, but makes up for it by being faster, allowing more ions to be fragmented.
12
Figure 3, a) Total Ion Chromatogram (TIC) from a 100minute Reverse phase (RP) microcapillary liquid chromatography (µLC) run b) one mass spectrum acquired at retention time 50.04 min c) MS/MS of the
peptide with m/z 669.99 Da d) Identified b- and y-ions and the corresponding amino acid sequence.
3.2.3
D
ATABASE SEARCHINGThe output files from the Orbitrap are called RAW files and they contain the masses for all peaks and their respective intensities. They also connect MS spectra to corresponding MS/MS spectra. This data is evaluated with one of many available search algorithms. The algorithms compare the acquired data to theoretical spectra generated from a selected protein sequence database.
To evaluate the quality of the dataset, a target-decoy database-searching strategy is useful (Elias, et al., 2007). For every protein sequence in the searched database, a corresponding reversed sequence is added. The probability of getting a false positive to the forward database is equally high as getting a hit to the reversed database. Thereby the false
a)
b)
c)
13 O- O -O O N H3 + O O -P
discovery rate (FDR) can be calculated by Equation 1. This allows filtering of data according to several different parameters to optimize the number of hits while minimizing false positives and allows assessment of the datasets confidence.
FDR = #reverse hits ∗ 2#total spectra
Equation 1, false discovery rate (FDR), used as a quality measurement of the dataset
3.3 BCA
ASSAYTo measure the amount of protein obtained bicinchoninic acid (BCA) assay (Thermo scientific) was used. The proteins reduce Cu2+ to Cu+
3.4 P
HOSPHORYLATIONso that it can form a complex with BCA, a complex which absorb at 562 nm. The concentration of an unknown sample is derived from a calibration curve made from know concentrations of bovine serum albumin.
The importance of phosphorylation is emphasized by the Nobel Prize awarded to Krebs and Fisher 1992 for discovering the regulatory function of the PTM (Nobel Web AB). It is present in both prokaryotes and eukaryotes (Chang, et al., 1998) and works by an equilibrium mediated by protein kinases and phosphatases with ATP as a frequently used phosphoryl donor, due to its ubiquitous presence (Figure 4). There are 4 different known types of phosphorylation; , N-, S- and acyl-phosphorylation. The most common is
O-phosphorylation, which occurs on serine, threonine or tyrosine residues (Reinders, et al., 2005).
Figure 4, a) phosphorylation and dephosphorylation of proteins with ATP as phosphoryl donor, mediated by protein kinases or phosphatases. b) Structure of phosphorylated serine (pSer)
When it comes to studying phosphorylated proteins with MS methods, there are some complications. Phosphorylated proteins are often in low stoichiometry ratio, and they are harder to ionize than unmodified peptides. Another problem with analyzing phosphorylated
14
proteins is that the phosphate group tends to absorb most of the collision energy, leaving the peptide intact apart from losing a phosphate group. This process is called neutral loss (NL) and the problem has been addressed by fragmenting the NL peak again, in a process called MS3, to increase the amount of b- and y-ions generated. Although a recent article by Villén et al. showed that MS3
Another complication is the sheer number of phosphorylation events as there are over 100,000 estimated phosphorylation sites in the human proteome and several thousand in yeast (Zhang, et al., 2002). This is not only a problem for phosphoproteomics but for proteomics in general. Not only are there too many peptides to analyze in a single run, but they are in varying abundance. This dynamic range problem has been approached with a variety of methods including depletion of high abundance proteins (Greenough, et al., 2004). The most utilized approach to deal with this problem however, is fractionation. By
separating complex peptide mixtures according to various physical and chemical properties the dynamic range, and amount of proteins introduced into the mass spectrometer is greatly reduced, thereby making it easier for the instrument to detect less abundant peptides.
is not worthwhile in proteomic studies when using a high mass accuracy instrument (Villén, et al., 2008). Removing most unmodified peptides however, solves all of these issues. If there are only phosphorylated peptides they will be ionized and, with nothing else in the sample, they won’t be overshadowed. An enriched sample also leads to a sufficient amount of phosphorylated peptides introduced into the spectrometer for b- and y- ions peaks to be intense enough for peptide identification.
3.5 E
NRICHMENT METHODS3.5.1
P
ROQ
ENRICHMENTPro-Q® Diamond Phosphoprotein Enrichment Kit (Molecular Probes) contains a commercial resin which has affinity for phosphorylated proteins. No information about how the resin works is given by the manufacturer, but it is capable of enriching both denatured and native phosphorylated proteins.
3.5.2
S
TRONG CATION EXCHANGESCX have been shown to be useful as a phosphopeptide enrichment strategy (Beausoleil, et al., 2004). Peptides bind to a negatively charged resin and are released into discrete
fractions by an elution gradient with increasing salt concentration. Since phosphorylated peptides have extra negative charges, they will elute earlier than their corresponding unmodified peptides and can therefore be enriched by collecting the first fractions that are eluted from the column, as shown in Figure 5.
15
Figure 5, a) theoretical charge state distribution of 5-40 amino acid long tryptic peptides generated from a human protein database. This shows that most peptides have a charge state of 2+. Any peptide in this category would shift to a 1+
3.5.3
I
MMOBILIZED METAL AFFINITY CHROMATOGRAPHYcharge state upon phosphorylation. b) strong cation exchange separation of HeLa cell lysate after trypsin digestion with some identified peptiedes shown. (Beausoleil, et al., 2004)
Immobilized metal affinity chromatography (IMAC), first introduced by Porath (Porath, et al., 1986), is one of the most popular methods to enrich phosphopeptides today. It is based on the strong affinity between certain metals, such as Fe3+, Ga2+, Zn2+, and the phosphate group. There are a number of different linkers to choose between for connecting the metals to beads but nitrilotriacetic acid (NTA) linkers have been shown to have superior affinity for Fe3+ compared to another common linker, imminodiacetric acid (IDA), due to its
tetradentate structure (Zhou, et al., 2000). Furthermore it has been shown that IDA incompletely loaded with Fe3+
3.5.4
M
ETHYL ESTERIFICATIONcan co-purify His- and Lys-containing peptides (Scanff, et al., 1991) which would lead to a more complex sample and less identified phosphopeptides.
A downside with IMAC is unspecific binding of peptides with negative charges. A proposed method of dealing with this problem is to block all carboxyl groups (Ficarro, et al., 2002). This is done through an esterification reaction, where thionyl chloride is used to couple methanol to the c-termini, aspartic- and glutamic- side chains, as seen in Figure 6. If water is present, the acid chloride can react with it to reform a carboxyl group, therefore the methanol must be anhydrous and the sample must be carefully dried.
Figure 6, mechanism for modification of carboxyl groups, to maintain neutral charge over a broader pH range and thereby reduce nonspecific binding to the IMAC column. Thionyl chloride reacts with the carboxyl
groups and forms acid chlorides. Acid chlorides have a very electrophilic carbon, which the oxygen in methanol can attack to form an ester.
16
3.5.5
O
THER METHODSApart from the methods used in this study, there are several other ways to enrich phosphopeptides. One relatively straight-forward approach is to use phosphotyrosine specific antibodies for affinity purification (Yeung, et al., 1998), another methods is replacing the phosphate group of serine and threonine with ethanedithiol by β-elimination followed by Michael addition and introduction of a biotin tag. This allows purification of the peptides with an avidin column. The major disadvantage with this method is that you need a large amount of sample, and that only phosphorylated serine and threonine is enriched (Oda, et al., 2001). Several methods similar to IMAC have also been developed, including the use of titanium (Pinkse, et al., 2004)and zirconium dioxide (Kweon, et al., 2006).
3.6 F
RACTIONATION METHODS3.6.1
S
TRONG CATION EXCHANGEAs described in section 3.5.2, peptides can be bound to negatively charged resin and eluted according to charged state with increasing salt concentration. If the gradient is gentle enough, even phosphorylated peptides will elute at different salt concentrations according to their charge distribution, allowing SCX to be used as a fractionation method.
3.6.2
OFFGEL-
I
SOELECTRIC FOCUSINGAn OFFGEL fractionator works in the same way as in gel isoelectric focusing, but has the advantage of recovering the peptides directly in solution. On top of a gradient gel, 12 or 24 enclosed compartments are filled with sample together with a buffer containing ampholytes. When applying a current through the gel, charged peptides will migrate towards the anode or cathode depending on if they are negatively or positively charged, until they reach the well were the pH equals their pI (Michel, et al., 2003). A potential advantage of using isoelectric focusing as fractionation method is that the pI of peptides can be calculated theoretically and used as an extra constraint to remove false positives (Cargile, et al., 2004).
3.6.3
E
THANOLAMIDE MODIFICATIONAn OFFGEL can also be used for phosphopeptide enrichment, by first blocking all negatively charged groups on a peptide, except the phosphorylation. Thereby the only peptides with a pI around or below neutral should be phosphorylated peptides. But the methyl group of esterified peptides is unstable at high pH, which can result in non-phosphorylated peptides in OFFGEL fractions that should only contain phosphorylated peptides (Xu, et al., 2007). A more stable alternative is 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride (EDC) mediated ethanol amine modification (Figure 7). This modification would block all negative charges except phosphate groups without hydrolyzing at high pH.
17
Figure 7, cross linking reaction for attaching ethanolamine to carboxyl groups. R-COOH represent all Asp, Glu and c-terminus and R2-NH2 is ethanolamine. EDC reacts with the carboxyl groups to form an amine reactive
18
4
E
XPERIMENTAL DETAILS
4.1 A
PPROACHThe general analysis process used is illustrated in Figure 8. Highlighted steps have been evaluated by pairwise comparisons while only minor modifications have been made to other steps. All chemicals used are purchased from Sigma Aldrich unless stated otherwise.
Figure 8, general analysis process showing the order in which the various methods were used. Bold text indicates steps which were thoroughly investigated.
4.2 Y
EAST GROWTH AND LYSISBudding yeast, Saccharomyces cerevisiae , strain BY4741, was grown for 3 days in room temperature on a Yeast Peptone Dextrose (YPD) plate before transferred to liquid YPD medium. Cells were harvested by centrifugation, 6000*g for 5 min, while in log phase (OD600
4.3 P
ROQ
ENRICHMENT< 1.5), to avoid hardening of cell wall, and lysed by vortexing in 8M Urea buffer with added glass beads for 10*20 s with a 120 s cool down period between cycles to avoid overheating.
Prior to loading the samples on a ProQ column (Invitrogen), they were precipitated with methanol and chloroform, and re-suspended in Focus Extraction Buffer –IV (Biosciences). After loading, the column was washed with provided wash buffer (Invitrogen). The flow-through from the loading and washing steps was pooled and saved for MS analysis. Enriched phosphoproteins were eluted with 50 mM tris(hydroxymethyl)aminomethane (Tris) pH 8, 2% Sodium Dodecyl Sulfate (SDS), 10 mM ethylenediaminetetraacetic acid (EDTA), 5 mM
tris(2-19
carboxyethyl)phosphine (TCEP) and concentrated to about 1mg/ml by centrifugal concentration, all according to the ProQ protocol (Invitrogen, 2005).
4.4 T
RYPSIN DIGESTIONAll protein mixtures were treated with TCEP for one hour to reduce disulfide bridges before trypsin (Promega) was added to the enzyme:substrate ratio of 1:200. The samples were incubated in 37°C over night and digestion was stopped by adding one third of sample volume 1% Trifluoroacetic acid (TFA).
4.5 D
ESALTINGDesalting was done with SepPack reverse phase (RP) C18 columns (Waters Corporation) or StageTips (Rappsilber, et al., 2007) with 2 disks of Empore 3M C18 material (Chrom Tech). Columns were prepared by washing with 80% Acetonitrile (ACN), 0.1% TFA and 0.1% TFA, loaded and washed with 1% TFA and eluted with 80% ACN, 0.1% TFA. All washing, loading and elution was performed under air pressure to speed up the procedure and give more uniform peptide loading. Samples were dried in a SpeedVac (Savant).
4.6 S
TRONG CATION EXCHANGE(SCX)
High pressure liquid chromatography (HPLC) was used with a quaternary Agilent 1100 pump at 3 ml min-1 flow rate and a photodiode array detector set up to read at 280 nm. The sample was dissolved in 5 mM K2HPO4
Figure 9
, 30% ACN, pH 2.65 before manually injected into a 20 ml loading loop, loaded on a semipreparative polysulfoethyl SCX column, 9.4 mm inner diameter, 100 mm length, 5 mm particle size, 200 Å pore size (PolyLC) and eluted with the gradient described in . Fractions of 4 min were collected, dried in SpeedVac to remove ACN and desalted with SepPack.
Figure 9, gradient used for SCX fractionation of trypsin digested peptides. 13 fractions of 4 minutes were collected from injection up to minute 52. The rest of the gradient is washing and regeneration of the column.
0% 20% 40% 60% 80% 100% 0 100 200 300 400 500 600 0 20 40 60 80 100 120 Con ce nt rat ion (m M ) Time (min) K2HPO4 (mM) KCl (mM) ACN (%)
20
4.7 OFFGEL
Isoelectric separation was to be performed on a 3100 OFFGEL fractionator according to User Manual (Agilent Technologies, 2006). A 24 well set-up was used and the samples were diluted to a final volume of 2.6 mL using OFFGEL peptide sample solution. An IPG gel strip with a 3-10 linear pH range and a preprogrammed electrofocusing gradient up to 50 kV was used for separation.
4.8 E
THANOL AMIDE MODIFICATIONThe peptides were dissolved in Phosphate buffered saline (PBS) and modified by adding 500 mM EDC in the presence of an abundance of ethanolamine (1 M) and 25 mM
H-hydroxysuccinimide (NHS) as catalyst. The samples were incubated for 2 hours in room temperature before adding 10% TFA and after another 30 minutes the samples were dried down in a SpeedVac.
4.9 E
STERIFICATION1 % thionyl chloride in anhydrous methanol was mixed and added directly to peptide
samples directly after SpeedVac. The samples were sonificated 10 min and incubated for 1 h at room temperature before being dried with a SpeedVac.
4.10 IMAC
To reduce the number of steps in the purification procedure, IMAC and subsequent desalting was made in a single step by loading the IMAC beads, PHOS-Select™ Iron Affinity Gel (Sigma), directly onto a desalting column, as described previously (Villén, et al., 2008). Prior to
loading the IMAC beads on SepPack or StageTips they were washed with binding buffer containing 40% ACN, 25 mM Formic Acid (FA) and incubated together with an equal amount of binding buffer and the peptide samples for 1 hour. Also the SepPack/StageTips were washed with MeOH, 50% ACN 1% Acetic Acid (AcOH) and twice with 1% FA. After loading the beads the column was washed with binding buffer to wash out the non-phosphorylated peptides, this flow-through was saved and subsequently analyzed on the MS to optimize loading amounts used. The C18 material was equilibrated with 1% FA and the peptides was eluted from the IMAC beads and transferred to the C18 material by addition of 500mM K2HPO4
4.11 MS
ANALYSIS. Then phosphate salts was washed away with another volume of 1% FA. Finally, phosphopeptides were eluted with 50% ACN 1% AcOH and dried in a Speed Vac.
The samples were redissolved in 5-20 µl 5% ACN, 0,1% FA, 2-5 µl, loaded onto a RP-µLC column and separated by a 60 min gradient of 5-35% ACN before injected and analyzed on an LTQ Orbitrap (Thermo). For each scan cycle, one full MS run (400-1800 m/z) at 60,000 resolution with 1,000,000 ions as target and 1,000 ms maximum ion accumulation time is followed by 8 in MS/MS mode with 100 ms maximum accumulation time and 30 ms activation at 35 % normalized collision energy. Dynamic exclusion was set up not to fragment ions that had already been selected for MS/MS the previous 35 s. Ions with a charge of +1 were also excluded, since they are rarely peptides. Lock-mass ion real time
21
calibration was used on masses 371.1012, 445.12 and 519.1388 to increase mass accuracy (Olsen, et al.).
4.12 D
ATABASE SEARCHINGMS/MS-data was searched against a Saccharomyces cerevisiae database containing both the forward and reverse sequences using several different algorithms. 2 versions of the Sequest algorithm (Eng, et al., 2004), Mascot (Perkins, et al., 1999) and X!Tandem (Craig, et al., 2004) were evaluated. For converting RAW files into MzXML prior to database searching, 2
different programs were tested. Mascot requires a different input type and therefore 2 different methods of conversing RAW data into mascot generic format (mgf) were evaluated as well. Furthermore, peptide hits were assigned probability scores (P score) by
PeptideProphet (Keller, et al., 2002), through 2 different software alternatives: Scaffold (Proteome Software) and with a web based software called the Trans-Proteome Pipeline (TPP) developed at the Institute for systems biology, Seattle (Institute for Systems Biology). Filtering of data was performed in Microsoft Office Excel (Microsoft Corporation) where several different parameters, including P score values, precursor mass error, number of enzymatic termini, Sequest Xcorr and Mascot ionscore values, was tested to minimize false discovery rate and maximize the number of phosphorylated peptides identified.
All searches were set up to detect the following variable modifications: oxidation of
methionine (+15.99491 AMU), phosphorylation of serine, threonine and tyrosine (+79.96633 AMU) and fixed modifications were set up on c-termini, aspartic- and glutamic acids in the methyl esterified (14.01565 AMU) and ethanol amine modified samples (+13.031634 AMU).
22
5
R
ESULTS AND
D
ISCUSSION
Progress towards an optimal identification of phosphorylated proteins was made in two parallel areas, as depicted in Figure 10. To find an optimal enrichment procedure for
phosphopeptides a series of pairwise comparisons were made, after an initial decision to use IMAC enrichment. IMAC was chosen as a starting point because all of the most successful previous studies have used it. The problem of nonspecific binding in IMAC was approached by evaluating methyl esterification. To further reduce the number of unmodified peptides, ProQ phosphoprotein enrichment was assessed and increasing the number of total
identified peptides was attempted by using different fractionation methods. However, the most crucial step surprisingly turned out to be optimization of data evaluation. A starting amount of 10 mg protein was used in every later test, as large starting amounts have been suggested to be a key factor for identifying large number of protein (Villén, et al., 2008). Moreover no negative effects have been observed from overloading IMAC resin (Ishihama, et al., 2007).
Figure 10, a) timeline showing major changes made to the method by marking when new techniques were added or removed b) table illustrating the major differences between the initial and final methods.
5.1 P
HOSPHOPROTEINS IDENTIFIEDWith the final enrichment procedure, 2,131 unique phosphorylated peptides from 1,026 proteins were identified with a FDR of 0.42%. Mostly singly phosphorylated peptides were identified as shown in Figure 11. It has previously been suggested this is because multiply
a)
23
phosphorylated peptides bind so tightly to the IMAC resin that they never elute (Thingholm, et al., 2008). Therefore an extra elution with ammonium water was tested, but did not lead to identification of any new peptides (data not shown).
Figure 11, a) distribution of single, double and multiply phosphorylated peptides identified out of total 1,899 unique phosphorylated peptides from the Sequest algorithm b) distribution of peptides phosphorylated on
serine, threonine and tyrosine residues from the same dataset
As expected, most phosphorylations were identified at serine residues but interestingly some on tyrosine were found as well, which will be discussed under future applications. In comparison with previous successful studies of the phosphoproteome of yeast under normal conditions this study was successful, as seen in Table I. The acquired dataset identified more phosphorylated proteins than any previous method. When all of these datasets are
combined with the dataset from this study, a list of 1,624 phosphoproteins is obtained (252 unique from this study). That number is comparable to the number of phosphorylated proteins estimated to be phosphorylated at any time in yeast (30% of 6,000).
80.9% (1537) 18.3% (347) 0.8% (15)
Number of phosphorylation sites
1 2 3+ 80% (1721) 18% (376) 2% (40) phosphorylated proteins pSer pThr pTyr a) b)
24
5.2 E
NRICHMENT METHODSCX can enrich phosphopeptides, lately however IMAC and TiO2
Figure 12
enrichment have been more popular, and since none of them have been shown to be superior to the other (Wolschin, et al., 2005) IMAC have been used in this study. A commercial resin exist that can enrich phosphorylated proteins and possibly increase enrichment further, ProQ. This method was evaluated as an extra enrichment step, applied prior to trypsin digestion. As seen in
, ProQ did not lead to any improvement in phosphopeptide enrichment. The relatively low overlap of identified proteins might suggest that technical doubles would increase the number of identified proteins for both methods, evening out their differences. However without indication of ProQ helping enrichment and considering that an extra method leads to loss of sample amount, ProQ was not used in later attempts.
Figure 12, a) phosphopeptide enrichment with and without ProQ enrichment. ME indicates methyl esterification prior to IMAC enrichment. b) Number of phosphorylated proteins identified with and without
ProQ protein enrichment, and the overlap between the 2 methods. No fractionation method was used.
0% 20% 40% 60% 80% Pho spho pe pt ide s / to ta l pe pt ide s a) b)
25
5.3 F
RACTIONATION METHODNo pairwise comparison between fractionation methods was possible, since no phosphopeptides were identified after OFFGEL fractionation. A large amount of
nonphosphorylated peptides were found in the flow-through from IMAC, but no peptides were found in the eluate. This indicates that the phosphate groups were lost during OFFGEL separation or that the IMAC step failed for one of several reasons thereby not enriching phosphorylated peptides. Possible reasons include IMAC buffers or resin being too old or the samples being stored for too long or without sufficient amounts of phosphatase inhibitors.
5.4 C
HEMICAL MODIFICATIONTo evaluate the effects of methyl esterification as directly as possible, every SCX fraction was split in two and one of them methyl esterified before loading on IMAC beads. The results are ambiguous but indicate that modification reduce nonspecific binding and thereby allow identification of additional phosphorylated peptides. As seen in Figure 13 a, methylation helps more for the later SCX fractions, were a bigger portion of nonphosphorylated peptides start to elute from the column.
Figure 13, evaluation of methyl esterification a) enrichment obtained represented as the fraction of phosphopeptides over all peptides identified, showing slightly better enrichment after methyl esterification
b) distribution of phosphopeptides over SCX fractions c) Phosphorylated proteins identified from modified versus unmodified samples. Only data from the Sequest algorithm is used for this comparison.
All of the methyl esterified samples were searched both with and without methyl
esterification as a constant modification to evaluate whether the modification is quantitative or not. Logically this should be possible to do by searching with methyl esterification as a
0% 20% 40% 60% 80% 100% 1 2 3 4 5 6 7 8 9 10 11 12 13 % pho spho pe pt ide s SCX Fraction unmodified methyl esterified 0 200 400 600 800 1000 1200 1 2 3 4 5 6 7 8 9 10 11 12 13 pho spho pe pt ide s SCX Fraction unmodified methyl esterified a) b) c)
26
variable modification, but as seen in Figure 14, from an initial search of part of the dataset, that was not the case. This may imply that the algorithm does not find every possible phosphorylated peptide, since phosphorylation is treated as variable modifications as well.
Figure 14, Venn diagram showing that Sequest searches with methylation as a variable modification does not find all hits that searches with or without methylation as a constant modification does. All searches are
performed on the same dataset, 3 SCX fractions.
Further supporting methyl esterification is the fact that it can be used for quantification without any additional steps, simply by using deuterated methanol on one sample, as shown in previous studies (Ficarro, et al., 2002).
5.5 S
EARCH ALGORITHMIt has been suggested that different search algorithms give complementing hits (Searle, et al., 2008). Therefore 3 of the most commonly used were evaluated by searching the same samples and looking at the overlap, as shown in Figure 15. It turns out Sequest is by far most successful at identifying phosphorylated peptides. This is surprising since Mascot is set up to handle neutral loss of phosphate groups (Matrix Science, 2007). Mascot searches seem to extend the number of identified proteins and is therefore still worth considering. X!Tandem however require further parameter optimization to be a useful addition for phosphopeptide identification.
27
Figure 15, number of unique phosphorylated peptides identified by each searching algorithm. Data consists of 3 SCX fractions, both the methyl esterified and unmodified fractions.
No differences were found between Sequest versions 27 and 28, neither between TPP and Scaffold for P score assignment or between scripts for generating mgf-files from RAW-files (data not shown).
5.6 S
EQUEST PARAMETERSSince the MS/MS data was collected using an LTQ, with limited mass accuracy the peptide mass tolerance was set to 0.8 AMU. Logically the parent mass tolerance should be set low, because the Orbitrap is a very accurate instrument. But empirically it was found that a too narrow parent mass tolerance window resulted in an increased number of reverse hits (Figure 16). The reason for this is unclear, but one hypothesis is that the scoring algorithm weighs hits in a narrow window so high that sequence coverage and number of b- and y- ions become less important, thereby giving false hits high probabilities score. It is obvious from Figure 16 that a higher mass tolerance in Sequest and a lower mass tolerance in Scaffold reduce false positive hits. This is in accordance with recent findings made by Villén and Gygi (Villén, et al., 2008).
28
Figure 16, false discovery rate for different combinations of Scaffold and Sequest parent mass tolerances. The data used is from sample 1 “IMAC”
After testing several different parameters to remove false positives, mass filtering (±6-16 ppm) together with setting minimum number of enzymatic termini to 2, was found most efficient. 0.0% 5.0% 10.0% 15.0% 20.0% 0.1 0.5 1 1.1 2 3 FD R
SEQUEST peptide mass tolerance
0.1 0.5 1 1.1 ∞ Scaffold peptide mass tolerance
29
6
C
ONCLUSIONS
With IMAC enrichment as a starting point, through a series of pairwise tests, an enrichment strategy capable of identifying more than 1,000 phosphorylated proteins was obtained. SCX fractionation in combination with IMAC enrichment worked better than any other method tested, and as good as any previously described strategy. The enrichment was further improved by methyl esterification prior to IMAC. Moreover no beneficial effect was shown after adding an extra enrichment step prior to trypsin digestion. Unexpectedly, data evaluation required much attention before deriving at high confidence dataset. After software modifications and parameter optimization high quality data was eventually obtained and a combination of Sequest and Mascot was found to improve the number of identified proteins. Finally, filtering data with parent mass error and number of tryptic termini as constrains was shown to be more efficient in increasing dataset confidence than the use of peptide and protein probability scores.
30
7
F
UTURE
7.1 F
UTURE DIRECTIONSIf I would have had more time on this project, there are several different paths I would pursue. The first would be different fragmentation techniques. Already some samples were fragmented with multistage activation (MSA) (Schroeder, et al., 2004) which showed promising results, at least for less complex samples. Another fragmentation method with promising results in phosphorylation studies is electron transfer dissociation (ETD) (Chi, et al., 2007), a technique that became available to the lab just before I left. Furthermore I would have liked to proceed with OFFGEL fractionation, which has some intriguing advantages such as an extra constraint to remove false positives and low fraction overlap (Hubner, et al., 2008). If I could get the EDC mediated ethanolamine modification to work as well it would be an interesting enrichment step. Lastly I would have liked to try alternatives on some of the other steps in the enrichment procedure, e.g. other proteolytic enzymes than trypsin or TiO2
7.2 F
UTURE APPLICATIONSinstead of IMAC enrichment.
Tyrosine phosphorylation is well studied in humans since dysregulation of this type of phosphorylation is a known causal factor for cancer and diabetes (Dubé, et al., 2005). However, until recently tyrosine phosphorylation was only thought to exist in higher
eukaryotes and is therefore not intensely studied in yeast. A 14 year old study indicates that tyrosine phosphorylation does exist in yeast, and that several substrates are
dephosphorylated by protein tyrosine phosphatase 1 (PTP1) (Wilson, et al., 1995). Only one of the substrates was identified at the time, but advances in mass spectrometry should allow identification of several other substrates. Therefore PTP1 and PTP2 deficient BY4741 strains are now studied with the final enrichment method. Potential substrates will hopefully be found by comparing identified phosphorylated proteins from tyrosine phosphatase deficient strains with wild type yeast.
Enriching phosphorylated peptides is also useful is in the study of Oral Squamous Cell
Carcinoma (OSCC). It is the 6th most common cancer worldwide and accounts for more than 30,000 cases of cancer and approximately 7,500 deaths each year in the United States alone (Jemal, et al., 2008). As with most types of cancer, early diagnosis of OSCC leads to a higher survival rate. The 5 year survival rate of patients if the cancer is detected in an early stage is 80% compared to 20-40% if detected at a later stage (Hu, et al., 2007). This combined with the fact that the survival rate has not improved much in 30 years (Shiboski, et al., 2000) underlines the need for new methods of detection. New diagnosis methods need to be faster, cheaper and simpler to account for a population growing older and older (UN, 2002). An example of a diagnostic fluid that potentially fulfills these criterions is saliva. It is easy to collect, fast, cost-efficient and non-invasive. By studying differences in phosphorylation of salivary proteins between subjects with and without cancer, potential biomarkers could be discovered.
31
8
A
CKNOWLEDGEMENT
I am very thankful to Dr. Maria Sunnerhagen and Prof. David Bernlohr who arranged for me to get the opportunity to visit the University of Minnesota. Both of them have been very supportive and helped me with various things throughout the fall. I would also like to thank Prof. Tim Griffin for letting me work in his laboratory, and his entire group for all their help. Special thanks to Dr. Sri Bandhakavi who has been an inspiring and helpful tutor and to Dr. Matt Stone for help with mass spectrometry analyses and whiskey advice.
Thanks to my family for refueling my energy over Christmas and thanks to my housemates on Oak St, especially Renee for helping me experience American culture and being a good friend. Final thanks to Marie, without whom the stay in Minneapolis wouldn’t have been the same.
32
9
R
EFERENCES
Agilent Technologies. 2006. Agilent technologies. Agilent Literature Library. [Online] October 13, 2006. [Cited: January 6, 2009.]
http://www.chem.agilent.com/Library/usermanuals/Public/G3100-90000_ebook.pdf. Beausoleil, Sean A, et al. 2004. Large-scale characterization of HeLa cell nuclear
phosphoproteins. Proceedings of the National Academy of Sciences of the United States of
America. 2004, Vol. 101, 33, pp. 12130–12135.
Cargile, Benjamin J, et al. 2004. Gel based isoelectric focusing of peptides and the utility of isoelectric point in protein identification. Journal of proteome research. 2004, Vol. 3, 1, pp. 112-119.
Carrette, Odile, et al. 2006. State-of-the-art two-dimensional gel electrophoresis: a key tool of proteomics research. Nature Protocols. 2006, Vol. 1, 2, pp. 812 -823.
Chang, Caren and Stewart, Richard C. 1998. The two-component system. Regulation of diverse signaling pathways in prokaryotes and eukaryotes. Plant Physiololgy. 1998, Vol. 117, pp. 723-31.
Chi, An, et al. 2007. Analysis of phosphorylation sites on proteins from Saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proceedings of the
National Academy of Sciences. 2007, Vol. 104, 7, pp. 2193–2198.
Cohen, Philip. 2002. Protein kinases--the major drug targets of the twenty-first century?
Nature reviews Drug Discovery. 2002, Vol. 1, 4, pp. 309-315.
Craig, R and Beavis, R C. 2004. TANDEM: matching proteins with tandem mass spectra.
Bioinformatics. 2004, Vol. 20, 9, pp. 1466-1467.
Dubé, Nadia and Tremblay, Michel L. 2005. Involvement of the small protein tyrosine phosphatases TC-PTP and PTP1B in signal transduction and siseases: from diabetes, obesity to cell cycle, and cancer. Biochimica et Biophysica Acta. 2005, Vol. 1754, pp. 108-117. Elias, Joshua E and Gygi, Steven P. 2007. Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature Methods. 2007, Vol. 4, 3, pp. 207-214.
Eng, Jimmy K, McCormack, Ashley L and Yates, John R. 2004. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database.
Journal of The American Society for Mass Spectrometry. 2004, Vol. 5, 11, pp. 976-989.
Fenn, John B, et al. 1989. Electrospray ionization for mass spectrometry of large biomolecules. Science. 1989, Vol. 246, pp. 64-71.
33
Ficarro, Scott B, et al. 2002. Phosphoproteome analysis by mass spectrometry and its application to Saccharomyces cerevisiae. Nature biotechnology. 2002, Vol. 20, 3, pp. 301-305.
Ficarro, Scott, et al. 2002. Phosphoproteome analysis of capacitated human sperm. The
Journal of Biological Chemistry. 2002, Vol. 278, 13, pp. 11579-11589.
Fields, Stanley. 2001. Proteomics in Genomeland. Science. 2001, Vol. 291, pp. 1221 - 1224. Goffeau, A, et al. 1996. Life with 6000 genes. Science. 1996, Vol. 274, 5287, pp. 563-567. Greenough, Carrie, et al. 2004. A method for the rapid depletion of albumin and
immunoglobulin from human plasma. Proteomics. 2004, Vol. 4, pp. 3107-3111. Gruhler, Albrecht, et al. 2005. Quantitative Phosphoproteomics Applied to the Yeast
Pheromone Signaling Pathway. Molecular & Cellular Proteomics. 2005, Vol. 4, 3, pp. 310-327. Gygi, Steven P, et al. 2000. Evaluation of two-dimensional gel electrophoresis-based
proteome analysis technology. Proceedings of the National Academy of Sciences of the
United States of America. 2000, Vol. 97, 17, pp. 9390-9395.
Hu, Shen, et al. 2007. Implications of salivary proteomics in drug discovery and
development: a focus on cancer drug discovery. Drug Discovery Today. 2007, Vol. 12, pp. 911-916.
Hubner, Nina C, Ren, Shubin and Mann, Matthias. 2008. Peptide separation with immobilized pI strips is an attractive alternative to in-gel protein digestion for proteome analysis. Proteomics. 2008, Vol. 8.
Institute for Systems Biology. Proteomics Tools. Seattle Proteome Center. [Online] [Cited: November 4, 2008.] http://tools.proteomecenter.org/software.php.
Invitrogen. 2005. Pro-Q® Diamond Phosphoprotein Enrichment Kit. Invitrogen Web site. [Online] June 28, 2005. [Cited: September 28, 2008.]
http://probes.invitrogen.com/media/pis/mp33358.pdf.
Ishihama, Yasushi, et al. 2007. Enhancement of the Efficiency of Phosphoproteomic Identification by Removing Phosphates after Phosphopeptide Enrichment. Journal of
Proteome Research. 2007, Vol. 6, pp. 1139-1144.
Jemal, Ahmedin, et al. 2008. Cancer Statistics, 2008. CA: A Cancer Journal for Clinicans. 2008, Vol. 58, pp. 71-96.
Keller, Andrew, et al. 2002. Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Analytical chemistry. 2002, Vol. 74, 20, pp. 5383 -5392 .
34
Klose, J. 1975. Protein mapping by combined isoelectric focusing and electrophoresis of mouse tissues. A novel approach to testing for induced point mutations in mammals.
Humangenetics. 1975, Vol. 26, pp. 231-43.
Kweon, Hye Kyong and Hakansson, Kristina. 2006. Selective Zirconium Dioxide-Based Enrichment of Phosphorylated Peptides for Mass Spectrometric Analysis. Analytical
Chemistry. 2006, Vol. 78, pp. 1743-1749.
Laviola, L, Giorgino, F and Natalicchio, A. 2007. The IGF-I signaling pathway. Current
Pharmaceuticall Design. 2007, Vol. 13, pp. 663-669.
Li, Xue, et al. 2007. Large-Scale Phosphorylation Analysis of α-Factor-Arrested
Saccharomyces cerevisiae. Journal of proteome research. 2007, Vol. 6, 3, pp. 1190-1197. Matrix Science. 2007. Help > Modifications. Matrix Science. [Online] Mars 09, 2007. [Cited: January 16, 2009.] http://www.matrixscience.com/help/pt_mods_help.html.
Michel, Philippe E, et al. 2003. Protein fractionation in a multicompartment device using Off-Gel isoelectric focusing. Electrophoresis. 2003, Vol. 24, pp. 3-11.
Nobel Web AB. nobelprize.org. [Online] [Cited: September 22, 2008.] http://nobelprize.org. Nollau, Peter, Machida, Kazuya and Mayer, Bruce J. 2003. Profiling the Global Tyrosine Phosphorylation. Molecular & Cellular Proteomics. 2003, Vol. 2, pp. 215-233.
Oda, Yoshiya, Nagasu, Takeshi and Chait, Brian T. 2001. Enrichment analysis of
phosphorylated proteins as a tool for probing the phosphoproteome. Nature Biotechnology. 2001, Vol. 19, 4, pp. 379-382.
O'Donovan, Claire, Apweiler, Rolf and Bairoch, Amos. 2001. The human proteomics initiative (HPI). TRENDS in Biotechnology. 2001, Vol. 19, 5.
Olsen, Jesper V, et al. Parts per Million Mass Accuracy on an Orbitrap Mass Spectrometer via Lock Mass Injection into a C-trap. Molecular & Cellular Proteomics. Vol. 4, 12, pp. 2010-2021. Patterson, Scott D, Aebersold, Reudi and Goodlett, David R. 2000. Mass spectrometry-based methods for protein identification and phosphorylation site analysis. [book auth.] S R Pennington and M J Dunn. Proteomics. s.l. : BIOS Scientific Publishers Ltd, 2000, p. 334. Perkins, DN, et al. 1999. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis. 1999, Vol. 20, 18, pp. 3551-3567. Pinkse, Martijn W H, et al. 2004. Selective isolation at the femtomole level of
phosphopeptides from proteolytic digests using 2D-NanoLC-ESI-MS/MS and titanium oxide precolumns. Analytical Chemistry. 2004, Vol. 76, 14, pp. 3935-3943.
35
Porath, J and Andersson, L. 1986. Isolation of phosphoproteins by immobilized metal (Fe3+) affinity chromatography. Analytical Chemistry. 1986, Vol. 154, pp. 250-254.
Rappsilber, Juri, Mann, Matthias and Ishihama, Yasushi. 2007. Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nature Protocols. 2007, Vol. 2, 8, pp. 1896-1906.
Reinders, Joerg and Sickmann, Albert. 2005. State-of-the-art in phosphoproteomics.
Proteomics. 2005, Vol. 5, pp. 4052-4061.
Scanff, P, Yvon, M and Pellissier, J P. 1991. Immobilized Fe3+ affinity chromatographic isolation of phosphopeptides. Journal of Chromtograph. 1991, Vol. 539, 2, pp. 425-432. Schroeder, Melanie J, et al. 2004. A neutral loss activation method for improved
phosphopeptide sequence analysis by quadrupole ion trap mass spectrometry. Analytical
chemistry. 2004, Vol. 76, 13, pp. 3590-3598.
Scigelova, Michaela and Makarov, Alexander. 2006. Orbitrap Mass Analyzer - Overview and Applications in Proteomics. Practical Proteomics. 2006, pp. 16-22.
Searle, Brian C, Turner, Mark and Nesvizhshii, Alexey I. 2008. Improving sensitivity by probabilistically combining results from multiple MS/MS search methodologies. The Journal
of Proteome Research. 2008, Vol. 7, pp. 245-253.
Shiboski, Caoline H, Shiboski, Stephen C and Silverman Jr., Sol. 2000. Trends in oral cancer rates in the United States, 1973-1996. Community Dentistry And Oral Epidemiology. 2000, Vol. 28, 4, pp. 249-256.
Tanaka, Koichi, et al. 1988. Protein and polymer analyses up to m/z 100 000 by laser ionization time-of-flight mass spectrometry. Rapid communications in mass spectrometry. 1988, Vol. 2, 8, pp. 151-153.
Thermo scientific. BCA Protein Assay. Thermo scientific Web site. [Online] Thermo scientific. [Cited: September 26, 2008.]
http://www.piercenet.com/Products/Browse.cfm?fldID=02020101.
—. 2007. Instruction books, Pierce Protein Reaserch Products. Thermo scientific. [Online] 2007. [Cited: January 14, 2008.] http://www.piercenet.com/files/0475as4.pdf.
Thingholm, Tina E, et al. 2008. SIMAC (Sequential Elution from IMAC), a Phosphoproteomics Strategy for the Rapid Separation of Monophosphorylated from Multiply Phosphorylated Peptides. Molecular & Cellular Proteomics. 2008, Vol. 7, 4, pp. 661-671.
UN. 2002. United Nations. Population Division, DESA. [Online] 2002. [Cited: January 09, 2009.]
36
Villén, Judit and Gygi, Steven P. 2008. The SCX/IMAC enrichment approach for global phosphorylation analysis by mass spectrometry. Nature protocaols. 2008, Vol. 3, 10, pp. 1630-1638.
Villén, Judit, Beausoleil, Sean A and Gygi, Steven P. 2008. Evaluation of the utlity of neutral-loss-dependent MS3 strategies in large-scale phosphorylation analysis. Proteomics. 2008, Vol. 8, pp. 4444-4452.
Wasinger, VC, et al. 1995. Progress with gene-product mapping of the Mollicutes: Mycoplasma genitalium. Electrophoresis. 1995, Vol. 16, pp. 1090-1094.
Wilson, Linda K, et al. 1995. The Yeast Immunophilin Fpr3 Is a Physiological Substrate of the Tyrosine-specific Phosphoprotein Phosphatase Ptp1. The Journal of Biological Chemistry. 1995, Vol. 270, 42, pp. 25185-25193.
Wolschin, F, Wienkoop, S and Weckwerth, W. 2005. Enrichment of phosphorylated proteins and peptides from complex mixtures using metal oxide/hydroxide affinity chromatography (MOAC). Proteomics. 2005, Vol. 5, 17, pp. 4389-4397.
Xu, Chong-Geng, et al. 2007. Selective Enrichment and Fractionation of Phosphopeptides from Peptide Mixtures by Isoelevtioc Focusing after Methyl Esterification. Analytical
Chemistry. 2007, Vol. 79, pp. 2007-2014.
Yeung, YG, et al. 1998. Colony-stimulationfactor-1 stimulates the formation of multimeric cytosolic complexes of signaling proteins and cytoskeletal components in macrophages.
Journal of Biological Chemistry. 1998, Vol. 273, 27, pp. 17128-17137.
Zhang, Hui, et al. 2002. Phosphoprotein analysis using antibodies broadly reactive against phosphorylated motifs. The Journal of biological chemistry. 2002, Vol. 277, 42, pp. 39379-39387.
Zhou, Wei, et al. 2000. Detection and sequencing of phosphopeptides affinity bound to immobilized metal ion beads by matrix-assisted laser desorption/ionization mass
spectrometry. Journal of The American Society for Mass Spectrometry. 2000, Vol. 11, 4, pp. 273-282.
Presentationsdatum
2009-01-22
Publiceringsdatum (elektronisk version)
2009-03-02
Institution och avdelning
Department of Physics, Chemistry and Biology
Linköping University
URL för elektronisk version
http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-16997
Publikationens title
Enrichment strategy development for phosphoproteome analysis of saccharomyces cerevisiae
Författare
Pontus Lundemo
Sammanfattning
The reversible phosphorylation of proteins is central to regulating most aspects of cell function. Malfunction in this critical cellular process have been implicated to cause diseases such as diabetes, cancer, and Alzheimer’s. Recent advances in mass spectrometry have made it possible to study this important post translational modification on a proteome-wide scale. However, to be able to do so, enrichment of phosphorylated peptides is required. Pairwise comparison of individual steps in an enrichment procedure and simultaneous improvement of data analysis resulted in a protocol which allowed high confidence identification of 2,131 unique phosphorylated peptides from 1,026 proteins. Thereby not only establishing a working protocol for phosphopeptide enrichment in the Griffin Lab, but also generating the largest list of proteins phosphorylated under normal conditions in yeast to date.
Nyckelord
Proteomics, mass spectrometry, phosphorylation, yeast, IMAC, SCX, Sequest
Språk
Svenska
X Annat (ange nedan) Engelska Antal sidor 36 Typ av publikation Licentiatavhandling X Examensarbete C-uppsats D-uppsats Rapport
Annat (ange nedan)
ISBN
ISRN: LITH-IFM-A-EX--09/2040--SE Serietitel
