Nature Biotechnology23, 463 - 468 (2005)
Published online: 13 March 2005; | doi:10.1038/nbt1076
Enrichment and analysis of peptide subsets using fluorous affinity tags and mass spectrometry
Scott M Brittain, Scott B Ficarro, Ansgar Brock
& Eric C Peters
Genomics Institute of the Novartis Research Foundation, 10675 John Jay Hopkins Drive, San Diego, California 92121, USA.
Correspondence should be addressed to Eric C Peters peters@gnf.org
Although mass spectrometry has become a powerful tool for the functional analysis of biological systems, complete proteome characterization cannot yet be achieved. Instead, the sheer complexity of living organisms demands fractionation of cellular extracts to enable more targeted analyses. Here, we introduce the concept of 'fluorous proteomics,' whereby specific peptide subsets from samples of biological origin are tagged with perfluorinated moieties and subsequently enriched by solid-phase extraction over a fluorous-functionalized stationary phase. This approach is extremely selective, yet can readily be tailored to enrich different subsets of peptides. Additionally, this methodology overcomes many of the limitations of traditional bioaffinity-based enrichment strategies, while enabling new affinity enrichment schemes impossible to implement with bioaffinity reagents. The potential of this methodology is demonstrated by the facile enrichment of peptides bearing particular side-chain functionalities or post-translational modifications from tryptic digests of individual proteins as well as whole cell lysates.
Functional proteome characterization involves the challenging task of identifying species of interest from among many thousands of proteins, each potentially altered by hundreds of possible post-translational modifications (PTMs). Additionally, living organisms often exhibit large dynamic ranges in protein expression levels, ranging from estimated values of 104 in yeast to 109−1012 in plasma. As a result of this extreme complexity, proteomics studies often use various fractionation methodologies to focus on only a subset of the overall protein complement. For example, numerous fractionation schemes based on the presence of a particular chemical moiety such as a native amino acid side-chain functionality1,
2 or a biologically important PTM have been described3,
4,
5,
6,
7,
8. Often, these affinity methods are specific for a particular functionality, such as immobilized metal affinity chromatography for the enrichment of phosphorylated peptides6 or various lectins for the enrichment of specific glycosylated species9. Although highly effective, these approaches require the discovery and development of an individual reagent for each specific functionality.
Alternatively, a more generalized approach based on a family of dual-functional reagents that possess different chemically reactive moieties coupled to a common affinity capture moiety has been described1,
10,
11. Typically, the classic biochemical affinity pair biotin-streptavidin has been used to isolate the labeled species. However, these custom reagents are relatively expensive, often difficult to fully elute from the capture resin12, and can complicate tandem mass spectrometry (MS/MS) spectral interpretation4. As such, an ideal enrichment method would readily be tunable with respect to its chemical selectivity, effect highly orthogonal but readily reversible isolations, be highly inert under MS/MS sequencing conditions and be economical enough to use with samples of varying levels of complexity.
Since the introduction of fluorous biphasic catalysis techniques13, the field of fluorous chemistry has expanded rapidly. The term 'fluorous' was coined to represent highly fluorinated (or perfluorinated) species in a way analogous to how 'aqueous' represents water-based systems. It has been demonstrated that appending a short perfluoroalkyl moiety to an organic compound enables the ready separation of these fluorous species from other mixture components by solid-phase extraction over fluorous-functionalized silica gel (or, fluorous solid-phase extraction (FSPE)) because of the selectivity arising from the nature of fluorine-fluorine interactions14,
15. This methodology enables the facile integration of synthesis and separation, and has been employed for the recycling and reuse of catalysts16, removal of reaction intermediates17 and the copurification of parallel synthesis products18. Recently, fluorous protecting groups have also aided in the purification of synthetic peptides19 and oligosaccharides20. However, to date, these methodologies have been used exclusively as processing aides for the targeted synthesis and purification of specific molecules in organic solvents.
Here, we demonstrate the use of fluorous affinity tags for the highly efficient enrichment and subsequent mass spectrometric characterization of various subsets of peptides from highly complex mixtures of biological origin. The versatility of this methodology is demonstrated by the isolation of different classes of peptides bearing various side-chain functionalities or specific PTMs. Additionally, the unique properties of these fluorous tags are utilized to implement fractionation schemes that would be impossible to implement using traditional bioaffinity tags.
Any viable proteomics fractionation methodology must enable both the efficient isolation of a desired subset from the remainder of the sample as well as the subsequent efficient recovery and analysis of that subset. Such a fluorous proteomics strategy is outlined in Figure 1a. Although this scheme depicts selective labeling occurring at the peptide level, it should be noted that the labeling of intact proteins followed by enzymatic digestion can also be performed, as discussed later in the text. The resulting peptide mixture is loaded onto a FSPE column, and nonlabeled peptides are removed from the column by isocratic washing with a solvent of suitable organic eluent concentration. Tagged peptides are then selectively eluted with a higher organic concentration and subjected to further analysis. Notably, all the various solutions used in this process are typically mixtures of methanol and water, enabling direct MS analysis of isolated peptides with the exact composition of the eluent solution depending on both the nature of the tag and the fluorous stationary phase used.
Figure 1. Fluorous proteomics strategy for the isolation of specific classes of peptides.
(a) Depiction of the overall method. A specific peptide class is selectively labeled with a fluorous tag ('F') bearing the appropriate functional group (for example, a thiol group to effect Michael addition after -elimination) and isolated using fluorous solid-phase extraction, followed by MALDI-MS or ESI-MS analysis. (b) MALDI-MS spectrum of -casein tryptic digest spiked with fluorous domain−containing peptides. (c) Nontagged peptides are observed in the MALDI-MS spectrum of the FSPE flow-through fraction. (d) Fluorous-tagged peptides are observed in the MALDI-MS spectrum of the FSPE elution fraction. q, pyroglutamic acid; sF17, dehydroalanine residue after reaction with 1H,1H,2H,2H-perfluorodecane-1-thiol.
To demonstrate FSPE enrichment of tagged peptides from a single-protein digest, two peptides bearing a C8F17 group were added to a tryptic digest of -casein (Fig. 1b). This mixture was loaded onto an FSPE column and washed with 60% methanol, resulting in the elution and collection of the nonlabeled peptides (Fig. 1c). Tagged species were then eluted from the column in 100% methanol. FSPE afforded a significant enrichment of the tagged peptides (Fig. 1d). Importantly, this enrichment was affected by a simple solid-phase extraction method utilizing commercially available inexpensive reagents. As such, this methodology could be used cost effectively with a technique such as two-dimensional gel electrophoresis that can separate protein mixtures into thousands of different samples. Based on these promising results, the ability of FSPE to enrich tagged peptides from far more complex mixtures was investigated. Thus, the same fluorous peptides were spiked into a tryptic digest of Jurkat T−cell total protein extract, and the resulting mixture was subjected to the same FSPE enrichment scheme. The tagged peptides were again easily isolated despite the far greater complexity of the mixture (Fig. 2).
Figure 2. Isolation of fluorous-tagged peptides from a highly complex peptide mixture.
MALDI-MS spectra of: (a) Jurkat cell total protein digest spiked with fluorous domain−containing peptides, (b) flow-through fraction after extensive washing with 50% acetonitrile and (c) FSPE elution fraction. q, pyroglutamic acid; sF17, dehydroalanine residue after reaction with 1H,1H,2H,2H-perfluorodecane-1-thiol.
To assess the relative sensitivity and percent recovery of this affinity methodology, the peptide VTQHFAK was synthesized and its free N-terminal amino group was selectively reacted with 2H,2H,3H,3H-perfluoroundecanoic acid while still attached to the resin using standard peptide coupling conditions. The resulting purified fluorous-labeled peptide served as a convenient quantifiable standard in repeat spiking studies, and at the same time enabled an initial assessment of whether the site of modification affects the efficiency of the isolation process. Thus, 20 pmol of the standard peptide was spiked into 100 g of a tryptic digest of Jurkat T−cell total protein extract, and this mixture was subjected to the same FSPE enrichment scheme. Matrix-assisted laser desorption ionization (MALDI) analysis of one-tenth of the FSPE eluent demonstrated a similar dramatic enrichment of the spiked peptide, whereas the percent recovery of the entire process measured between 50% and 55% (see Supplementary Fig. 1 online and Supplementary Methods online for details). Additionally, 500 fmol of the standard peptide was independently spiked into the same quantity of tryptic digest, and after subjecting this mixture to FSPE, the fluorous peptide was again seen to be the dominant species in the MALDI mass spectrum of one-fifth of the FSPE eluent. These preliminary results suggest that the fluorous affinity methodology, regardless of the site of modification, is indeed highly efficient.
We next sought to evaluate the compatibility of such fluorous tagged peptides with MS/MS sequencing strategies. Doubly charged ions of the fully tryptic peptides MPcF17TEDYLSLILNR (where cF17 represents a cysteine residue after reaction with N-[(3-perfluorooctyl)-propyl] iodoacetamide and MPcTEDYLSLILNR (where c represents a cysteine residue after reaction with iodoacetamide) were subjected to MS/MS on an quadrupole time-of-flight mass spectrometer. The two MS/MS spectra are nearly identical (Fig. 3a,b) when taking into account the clearly recognizable mass shift of fluorous-labeled fragment ions, with the exception that the spectrum of the fluorous-modified peptide (Fig. 3a) exhibits a characteristic immonium ion of monoisotopic mass 593.06 Da (labeled CI*) that is absent in the spectrum of the traditionally alkylated peptide. The presence of this unique marker ion can thus readily serve as confirmation that any given species is indeed labeled. Importantly, no ions corresponding to the loss or decomposition of the perfluoroalkyl group were observed. Thus, fluorous affinity tags have a distinct advantage compared to numerous biotin-based reagents, which often suffer fragmentation during MS/MS and thus complicate spectral interpretation4.
Figure 3. MS/MS spectra corresponding to (a) MPcF17TEDYLSLILNR (cF17 represents a cysteine residue after reaction with N-[(3-perfluorooctyl)-propyl] iodoacetamide), and (b) MPcTEDYLSLILNR (c represents a cysteine residue after reaction with iodoacetamide); CI*, immonium ion of the fluorous-labeled cysteine residue; *, loss of water/ammonia.
MALDI-MS spectra of (c) BSA tryptic digest after alkylation with N-[(3-perfluorooctyl)-propyl] iodoacetamide and (d) after FSPE elution. *, fluorous tagged cysteine-containing peptide.
Having shown the exquisite selectivity obtainable using fluorous affinity tags, we next sought to demonstrate that the labeling reactions could be performed directly on proteomics samples. Specifically, we investigated the enrichment of phosphorylated peptides from whole cell lysates utilizing established beta-elimination/Michael addition chemistry7,
21,
22. It should be noted that this reaction methodology suffers from several well-documented limitations, including the inability to distinguish whether an identified site was initially phosphorylated, modified with -N-acetylglucosamine or in some cases was originally an unmodified serine residue23. As such, other isolation methodologies may be more efficient for the global profiling of such modifications5,
24. Nevertheless, this study serves as a useful benchmark of our affinity approach as several similar studies using traditional bioaffinity-based reagents have already been reported.
Yeast total protein extract (400−500 g) was digested with trypsin, and the resulting peptides were treated with performic acid to oxidize cysteine residues to cysteic acid. The oxidized peptides were subjected to -elimination/Michael addition using 1H,1H,2H,2H-perfluorooctane-1-thiol as the Michael donor, and subsequently oxidized with H2O2 to generate a specific MS/MS marker for identification12. The labeled peptides were enriched by FSPE and analyzed by liquid chromatography (LC)/MS/MS. A total of 21 candidate phosphorylated peptides were identified (Supplementary Table 1 online), seven of which had previously been observed from yeast total protein extract using a carbodiimide-based isolation strategy3, indicating the validity of our approach. It should be noted that the C6F13-containing thiol was used in this case so that the resulting tagged peptides were directly compatible with reversed-phase LC columns typically used in proteomics studies. We observed that the C6F13 tag was slightly less selective than its C8F17 analog, as a small number of nonphosphorylated peptides were detected (data not shown). However, these unlabeled peptides were often easy to recognize as such because of the absence of the sulfenic acid neutral loss.
Similarly, Jurkat T−cell total protein extract (5 mg) was treated as described above, except that the tryptic peptides were first prefractionated on a C18 reversed-phase medium after performic acid oxidation (eluent steps of 5%, 15%, 25% and 40% acetonitrile in water). Each fraction was then individually subjected to -elimination/Michael addition and treated with H2O2. The resulting tagged peptides were enriched by FSPE and analyzed by LC/MS/MS. Nearly sixty candidate phosphopeptides were identified (as shown in Supplementary Table 2 online), 12% of which have been reported in the literature. Importantly, these two studies produced candidate phosphorylation sequences equal to or greater in number than those reported by other studies using chemically labeled affinity approaches3,
4 despite the far greater operational simplicity of the fluorous methodology.
The versatility of this enrichment strategy is greatly enhanced by the fact that it can readily be combined with a wide variety of highly selective chemical functionalities. For example, reagents combining a thiol-specific iodoacetamide moiety with a biotin affinity tag have been widely employed to simplify the complexity of a mixture before mass spectrometric analysis by effecting the selective isolation of cysteine-containing peptides1. By analogy, fluorous analogs of these reagents would be expected to effect a similar enrichment. To test this supposition, a tryptic digest of bovine serum albumin was reduced, treated with N-[(3-perfluorooctyl)-propyl] iodoacetamide and subjected to the same FSPE methodology described previously. As shown in Figure 3c,d, the tagged cysteine-containing tryptic peptides (marked '*') could clearly be isolated (Fig. 3d) from their nonlabeled counterparts. In fact, many of these cysteine-containing peptides were not detected in the MALDI-MS analysis of the unfractionated digest (Fig. 3c). Similarly, the relative quantification capability of some biotin-based reagents would similarly be expected to be affected by fluorous analogs having 13C substitutions in the fluorous domain, appropriate stable isotopic substitutions in the short linker region between the reactive moiety and the fluorous domain (required to attenuate the inductive effect of the fluorines on the reactive moiety), or some combination thereof.
Although the examples presented so far have involved chemical labeling of peptides, fluorous affinity reagents can also be used to label intact proteins. Chicken ovalbumin was reduced and reacted with N-[(3-perfluorohexyl)-propyl] iodoacetamide. The labeled protein was then subjected to all the operations typical of gel-based proteomics (that is, SDS-PAGE, in gel tryptic digestion and extraction of all resulting peptides from the gel whether labeled or not). The extracted digest was analyzed both before and after FSPE by LC/MS/MS, and the results were searched using MASCOT allowing for appropriate variable modification of the cysteine residues (Supplementary Fig. 2a,b online, respectively). Although the same three fluorous-modified peptides were identified in both analyses, numerous peptides identified in the sample before FSPE were no longer detected in the sample after FSPE despite the highly hydrophobic nature of some of these peptides. More importantly, MASCOT returned similar scores for the MS/MS spectra of iodoacetamide or N-[(3-perfluorohexyl)-propyl] iodoacetamide-modified tryptic peptides (Supplementary Fig. 2c online), and thus is able to unequivocally identify the protein from the entire NCBInr database using only the MS/MS spectra of fluorous-modified peptides (Supplementary Fig. 2d online), demonstrating that these affinity labels are completely compatible with commercially available protein identification software. Additional MS/MS spectra of fluorous-labeled peptides are shown in Supplementary Figures 3,4,5 online.
In addition to the production of an entire family of more easily used analogs of currently available labeling reagents, fluorous proteomic methodologies can also enable selective enrichment strategies that cannot be implemented using classical bioaffinity pairs. For example, numerous PTMs including disulfide bonds or ubiquitination lead to the production of branch sites within protein structures. After enzymatic digestion, these sites of modification can be particularly difficult to detect since often the only discriminating feature between a branched peptide and the far more numerous linear peptides is the presence of two N-terminal amino groups in the branched peptide. Given that it has been demonstrated that small molecules can be separated based primarily on the length of their fluorous tag25, we decided to investigate whether doubly tagged (branched) peptides could be separated from singly tagged (linear) peptides using a fluorous labeling strategy. Accordingly, a tryptic digest of polyubiquitin (Fig. 4a) was used as a model system. The effective implementation of the strategy requires the initial selective blocking of the -amino groups of lysine residues. However, this can readily be accomplished by guanidination (Fig. 4b)26,
27. The resulting mixture was reacted with the N-hydroxysuccinimide ester of 1H,1H,2H,2H-perfluorohexanoic acid, and the peptides were loaded onto an FSPE column. Isocratic washing with 50% acetonitrile resulted in the elution of singly labeled peptides, whereas the doubly labeled branched peptides were selectively eluted with pure methanol and subsequently analyzed (Fig. 4c). Figure 4d shows the MS/MS spectrum of the doubly labeled "Gly-Gly"28 branched peptide. These results suggest that this strategy could be a general approach for the study of these types of modifications that does not require the artificial transfection of affinity-tagged proteins into cell lines28, and is more selective than systems that rely solely on inducing changes in peptide hydrophobicity29.
Figure 4. Separation of peptides based on fluorine content.
MALDI-MS spectra of (a) tryptic digest of polyubiquitin, (b) polyubiquitin tryptic digest after guanidination of -amino groups with O-methylisourea, (c) FSPE elution fraction after derivatization of mixture with N-succinimidyl-3-perfluorobutyl propionate and (d) MS/MS spectrum of the doubly labeled "Gly-Gly"28 branched peptide.
We have described a methodology for the effective enrichment and subsequent characterization of peptide subsets from complex mixtures of biological origin using fluorous affinity tags. Specifically, perfluoroalkyl groups are selectively coupled to species of interest, and the resulting tagged peptides are readily isolated using a simple FSPE procedure, the conditions of which facilitate their subsequent direct characterization by mass spectrometry. The methodology is extremely selective and has been demonstrated for the enrichment of different classes of peptides bearing various native side-chain functionalities (thiol, amino) or specific PTMs (phosphorylation). Additionally, the methodology can readily be extended to the enrichment of other functional moieties owing to the facile synthesis of perfluoroalkyl chains bearing a wide array of reactive functional groups. Furthermore, the selectivity of the affinity enrichment can readily be tailored by adjusting the nature of the fluorinated moiety incorporated into the labeling reagent and/or the composition of the solid-phase extraction medium. Thus, this methodology avoids many of the problems associated with the use of classical bioaffinity strategies, including inefficient recovery of labeled substrates, fragmentation during MS/MS and the relatively high cost of such reagents.
This methodology also enables separation schemes not possible with conventional bioaffinity reagents. For example, we are unaware of any reports of peptides fractionated based on the number of biotin-based reagents with which they are labeled. By comparison, the ability to achieve enrichment based on differing fluorine content readily enables the fractionation of peptides bearing different numbers of a particular functionality or PTM. Alternatively, different PTMs (for example, phosphorylation and N-linked glycosylation) or functional moieties could be assayed in the same analysis by using affinity tags that differ in both their chemical reactivity and their fluorine content.
Future developments will undoubtedly extend the utility of this methodology. For example, in addition to its relative inertness under collision-activated dissociation conditions, fluorine exhibits a mass defect that could potentially be employed to enable the recognition of tagged species based on high-mass-accuracy measurements. Alternatively, the incorporation of cleavable linkers into the labeling reagents would enable highly selective enrichments to be followed by peptide characterization using classical high-performance liquid chromatography conditions. Finally, unlike classical bioaffinity reagents, this methodology could be directly applied to the selective isolation of different functional classes of small molecules in metabolomic studies.
Methods Proteolysis. Proteolysis with sequencing-grade modified trypsin (Promega) was carried out overnight at 37 °C using a substrate/enzyme ratio of 50:1 (wt/wt) in 100 mM ammonium bicarbonate. Bovine -casein, serum albumin and chicken ovalbumin were purchased from Sigma-Aldrich, and polyubiquitin was purchased from Affiniti Research Products.
Preparation of yeast and Jurkat whole cell protein fractions. Yeast cake (Saccharomyces cerevisiae) was purchased from a bakery supply store and subsequently pulverized under liquid N2 and stored at -80 °C. Jurkat (E6-1, ATCC) T cells were grown and harvested as described previously30. Cellular protein was isolated using Trizol reagent (Invitrogen), and subsequently oxidized by incubation overnight at 4 °C in 50 l oxidation solution (4.5 ml concentrated formic acid + 0.5 ml 30% H2O2). Finally, the oxidized protein fraction was dialyzed into 100 mM ammonium bicarbonate and digested with trypsin as described. Jurkat tryptic peptides were preparatively fractionated on a C18 reversed-phase cartridge (Haisil, Higgins Analytical) using step eluent solutions of 5%, 15%, 25% and 40% acetonitrile in 0.1% acetic acid.
Alkaline -elimination/fluorous Michael addition to O-phosphopeptides. Tryptic peptide digests (5 l) were combined with an equal volume of 3:1 dimethyl sulfoxide/ethanol (vol/vol), followed by the addition of 4.6 l saturated Ba(OH)2 and 1 l 500 mM NaOH. Finally, 0.7 l of 1H,1H,2H,2H-perfluorooctae-1-thiol (Fluorous Technologies) was added and the solution allowed to react at 37 °C for 1 h. The reactions were stopped by acidification of the solution with 5% trifluoroacetic acid (vol/vol), and subsequently oxidized by adjusting the solution to a final concentration of 3% H2O2 (vol/vol) and reaction for 30 min at room temperature (20°C).
Fluorous derivatization of cysteinyl peptides. Bovine serum albumin (BSA) tryptic digest was allowed to react with 50 mM N-[(3-perfluorooctyl)-propyl] iodoacetamide (Fluorous Technologies, prepared as a 500 mM stock solution in THF) in 25% ethanol (vol/vol) in 100 mM ammonium bicarbonate for 2 h in the dark. Excess alkylating reagent was removed by incubation with 2 mg of N-2-mercaptoethylaminomethyl polystyrene beads (Novabiochem) at room temperature for 2 h.
Fluorous derivatization and processing of cysteinyl-containing protein. Ovalbumin (40 M) was reduced with 10 mM TCEP in 6 M guanidinium hydrochloride, 20 mM Tris, pH 8.0 buffer for 10 minutes at room temperature, and reacted with 20 mM N-(1H,1H,2H,2H-perfluorohexyl)iodoacetamide for 1 h in the dark by addition of an equal volume of a THF solution of the fluorous iodoacetamide. Excess reagents were removed using a disposable gel filtration spin column packed with Biogel P6 beads (Micro Biospin P6, Bio-Rad). The fluorous-labeled protein fraction was combined with an equal volume of gel loading buffer (100 mM Tris, pH 6.8, 50 % glycerol, 0.1% bromophenol blue, 1% SDS). SDS-PAGE was performed using a 150 V constant voltage after loading several micrograms of derivatized protein into each well of a 12 well, 1 mm 8 cm 8 cm 10−20% Tris-Glycine polyacrylamide gel (Invitrogen). Following electrophoresis, the gel slab was stained with colloidal Coomassie blue (Invitrogen) for 4 h, followed by destaining in water overnight. Excised protein gel bands were cut into small cubes (2−3 mm) and subjected to several wash and dehydration cycles (ten min incubation with 50 l of 100 mM ammonium bicarbonate, removal of the liquid, ten min incubation with 25 l acetonitrile and removal of the liquid). The dehydrated gel cubes were vacuum dried for 5 min, rehydrated in 20 l of 50 mM ammonium bicarbonate solution containing 10 ng/l trypsin and incubated at 37 °C overnight. Tryptic peptides were recovered by repeated extraction (3 ) with 80% acetonitrile/0.2% TFA (vol/vol).
Guanidination and N-terminal fluorous derivatization. Peptide lysine -amino groups were first converted to their homoarginine analogs. Tryptic peptide digests were absorbed onto C18ZipTip (Millipore), and these tips were then repeatedly aspirated in 10 l of concentrated guanidination solution (2 l 0.5 M O-methylisourea hydrogensulfate dissolved in 8 l 0.25 M sodium carbonate, pH 11.7). The solution was then pipetted above the sorbent bed, and the solution-imbibed tips were incubated for 2 h at 37 °C. Following guanidination, peptides were thoroughly washed with 0.1% trifluoroacetic acid, eluted from the tips using 80% acetonitrile in 0.1% trifluoroacetic acid and dried under vacuum. N-terminal -amino groups were subsequently derivatized by addition of an equal volume of 0.25 M sodium bicarbonate buffer, pH 8.5, and freshly prepared 250 mM N-succinimidyl-3-perfluorobutyl propionate in THF (40 l total). The reactions proceeded for 2 h at room temperature, followed by the addition of 4 l of a 50% hydroxylamine solution to reverse unwanted reactions. This solution was allowed to stand for 10 min, followed by addition of 5 l of 5% trifluoroacetic acid.
Fluorous Solid Phase Extraction (FSPE). FSPE was performed in packed fused silica capillaries (360/200 m O.D/I.D 5−8 cm length) containing fluorous reversed-phase silica gel (FRPSG) (Fluoroflash Media; perfluorooctane bonded phase, 5 m particles, Fluorous Technologies). After equilibration with 60% methanol (vol/vol) in 10 mM ammonium formate, a labeled sample was loaded onto a column. The column was washed with multiple bed volumes of either wash solution A (60% methanol (vol/vol) in 10 mM ammonium formate) or wash solution B (50% acetonitrile (vol/vol)) depending on the fluorous tag used (wash solution A used with C6F13 tags, wash solution B with C8F17 or 2xC4F9 tags). Finally, fluorous tagged peptides were eluted using 50 column volumes of methanol.
Mass spectrometry and data analysis. MALDI-TOF analysis was performed on a Bruker Biflex III mass
spectrometer operated in the positive ion mode. Peptides were deposited onto MALDI targets using the dried droplet method by mixing
with 2,5-dihydroxybenzoic acid matrix solution (10 mg/ml in 50% acetonitrile with 0.2% trifluoroacetic acid). LC-ESI MS/MS was performed
using a Monitor C18 (Column Engineering) packed capillary column (3 m particles, 100 Å, 75 m I.D., 8 cm length, 5 m spray tip) interfaced to
a hybrid quadrupole time-of-flight (QqTof) mass spectrometer (Micromass Q-Tof-2) operating in
survey scan mode. Tandem MS data were processed and interrogated with the Mascot database
search algorithm (Matrix Science). Search parameters included the variable oxidation
of cysteine to cysteic acid and methionine to the methionine sulfone, as well as the C6F13 sulfoxide side-chain
modifications on former phosphoserine and phosphothreonine residues. The latter two modifications were identified on the basis of their unique
residual masses together with recognition of their neutral loss products, dehydroalanine (phosphoserine) or -methyldehydroalanine
(phosphothreonine) in the MS/MS patterns.
Natureproducts is an online service detailing information about specific
products used in this article, you can view the product descriptions, request
information and compare with other similar products. The products
used are listed in alphabetical order.
-elimination) and isolated using fluorous solid-phase extraction, followed by MALDI-MS or ESI-MS analysis. (b) MALDI-MS spectrum of 
-casein tryptic digest spiked with fluorous domain−containing peptides. (c) Nontagged peptides are observed in the MALDI-MS spectrum of the FSPE flow-through fraction. (d) Fluorous-tagged peptides are observed in the MALDI-MS spectrum of the FSPE elution fraction. q, pyroglutamic acid; sF17, dehydroalanine residue after reaction with 1H,1H,2H,2H-perfluorodecane-1-thiol.
g of a tryptic digest of Jurkat T−cell total protein extract, and this mixture was subjected to the same FSPE enrichment scheme. Matrix-assisted laser desorption ionization (MALDI) analysis of one-tenth of the FSPE eluent demonstrated a similar dramatic enrichment of the spiked peptide, whereas the percent recovery of the entire process measured between 50% and 55% (see 
-amino groups of lysine residues. However, this can readily be accomplished by guanidination (
-elimination/fluorous Michael addition to O-phosphopeptides.
40 
8 cm 
