Analysis of sulfated peptides from the skin secretion of the Pachymedusa dacnicolor frog using IMAC-Ga enrichment and high-resolution mass spectrometry

Research Article

Received: 5 November 2010 Revised: 18 January 2011 Accepted: 20 January 2011 Published online in Wiley Online Library

Rapid Commun. Mass Spectrom. 2011, 25, 1017–1027

Analysis of sulfated peptides from the skin secretion of thePachymedusa dacnicolor frog using IMAC‐Ga enrichment andhigh‐resolution mass spectrometry

Griselda Demesa Balderrama1, Erika P. Meneses1, Lorena Hernández Orihuela1,Oscar Villa Hernández1, Ruben Castro Franco2, Victoria Pando Robles3 andCesar Vicente Ferreira Batista1*1Unidad de Proteómica‐Instituto de Biotecnología, Universidad Nacional Autónoma de México (UNAM). Av. Universidad, 2001,Col. Chamilpa, CP 62210, Cuernavaca, Mor., México

2Lab. Herpetología, Centro de Investigaciones Biológicas, UAEM, Av. Universidad 1001, Col. Chamilpa, CP 62209,Cuernavaca, Mor., Mexico

3Centro de Investigación Sobre Enfermedades Infecciosas – Instituto Nacional de Salud Pública. Av. Universidad No. 655. Col.Santa María Ahuacatitlán, Cerrada Los Pinos y Caminera, C.P. 62100, Cuernavaca, Mor., Mexico

Immobilizedmetal ion affinity chromatography (IMAC)has beenwidely used for the enrichment of phosphopeptides,whereas no report exists describing the use of IMAC columns for the enrichment of sulfopeptides. In this study, weused IMAC‐Ga microcolumns for the enrichment of sulfopeptides from a complex mixture of peptides, extractedfrom skin secretions of the Pachymedusa dacnicolor frog. The enriched fraction obtained by IMAC‐Ga was analyzedby liquid chromatograpy/electrospray ionization mass spectrometry (LC/ESI‐MS) in an Orbitrap XL and by matrix‐assisted laser desorption/ionization time‐of‐flight time‐of‐flight (MALDI‐TOF/TOF) in an ABI 4800 instrument.From this fraction, different sulfated and non‐sulfated peptides belonging to the caerulin and bradykinin familieswere structurally characterized. Other interesting negatively charged groups, such as phosphate adducts ofdermaseptins and pyridoxal phosphate attached to a protease inhibitor, were also characterized. Unexpectedly, somedermaseptin antimicrobial peptides were also enriched by IMAC‐Ga and a Sauvatine‐like peptide was also fullysequenced. Furthermore, neutral loss of sulfated peptides and their fragmentation patterns in the gas phase werealso compared using collision‐induced dissociation (CID) and high‐energy collision dissociation (HCD). Ourpresent study provides evidence that IMAC‐Ga enrichment is a fast, useful and promising method for high‐throughput analysis of sulfated‐peptides, since high‐resolution mass spectrometers can be used for this purpose.Copyright © 2011 John Wiley & Sons, Ltd.

(wileyonlinelibrary.com) DOI: 10.1002/rcm.4950

Sulfate, phosphate and glycosides are chemical groupscommonly attached to proteins.[1,2] They are considered toconstitute labile post‐translational modifications (PTMs) inmass spectrometry (MS) due to their fast elimination in thegas phase, especially when submitted to certain types ofdissociation events.[3,4] This chemical process is known asneutral loss and constitutes an analytical challenge in MS‐based‐proteomics, especially when complex mixtures ofpeptides are analyzed in a high‐throughput fashion. Revers-ible phosphorylation is a well‐studied PTM because itconstitutes part of important biochemical processes, such asregulation of cellular proteins, subcellular localization, degra-dation, and complex formation, and it also plays a key role incell signaling.[5,6] However, sulfated proteins remain poorlystudied, even though they represent one of the most

* Correspondence to: C. V. Ferreira Batista, Unidad deProteómica, Instituto de Biotecnología‐UNAM, Av.Universidad, 2001 Col. Chamilpa, CP 62210, Cuernavaca,Mor., México.E‐mail: [email protected]

Rapid Commun. Mass Spectrom. 2011, 25, 1017–1027

101

commonly PTMs found in proteins.[7] Tyrosine O‐sulfonationoccurs mainly on secreted and trans‐membrane spanningproteins. Although the biological function of tyrosine sulfate isstill not well understood, some reports have described therole it plays in the modulation of extracellular protein–protein interactions, hemostasis regulation and leukocytetrafficking.[8,9]

Interestingly, sulfated peptides are commonly found in theskin secretions of a number of amphibian species,mainly thosebelonging to the genus Litoria, Pachymedusa and Phyllomedusa.Previous works have reported the presence of caerulein andbradykinin peptides containing tyrosine O‐sulfonation, aswell as a varying number of hydroxyprolines in thesemolecules.[10–12]

Immobilized metal affinity chromatography (IMAC) hasbeen used extensively for the enrichment of phospho-peptides[13,14] and has also been used for the identificationof novel phosphorylation sites in Aurora A from the Xenopuslaevis frog.[15] However, the use of IMAC columns for theenrichment of sulfopeptides in sulfoproteomics screening isnot a common analytical procedure.[16] The characterizationof sulfopeptides by mass spectrometry can easily be

Copyright © 2011 John Wiley & Sons, Ltd.

7

G. D. Balderrama et al.

1018

misidentified as phosphopeptides, since the sulfate moiety isonly 9.5mDa lighter than phosphate.[17] However, the smallmolecular mass difference between both chemical groups canbe distinguished by the new generation of high‐resolutionmass spectrometers. Likewise, phospho‐ and sulfopeptidespossess similar chemical properties and, therefore, protocolsusing metal affinity chromatography could theoretically beused for selective enrichment of both types of modifiedpeptides. Furthermore, phosphate and sulfate groups presenta distinct neutral loss pattern when submitted to CID. Loss of98Da is frequently detected in the CID spectrum of peptidescontaining serine and threonine phosphorylated amino acidresidues and correspondingly neutral loss of 80Da iscommonly observed when CID is used to dissociate sulfatedpeptides.In this study, we have achieved the characterization of a

number of post‐translationally modified peptides from thecrude secretions of the Pachymedusa dacnicolor frog, byapplying the IMAC‐Ga enrichment protocol. We character-ized sulfated bradykinins, sulfated caeruleins, a pyridoxalphosphate‐modified protease inhibitor and a dermaseptinphosphate adduct taken from the enriched peptide fraction.In addition, a novel sauvatine‐like peptide was fully se-quenced. Finally, collision‐induced dissociation (CID) andhigh‐energy collision dissociation (HCD) methods wereevaluated, in order to verify the efficiency and the qualityof the data used for the structural characterization of sul-fated peptides. Our present study provides evidence thatIMAC‐Ga enrichment is a quick, useful and promisingmethod for the high‐throughput analysis of sulfated pep-tides, when high‐resolution mass spectrometers are employedfor this purpose.

EXPERIMENTAL

Collection of skin secretions

Adults from the P. dacnicolor frog species were collected inthe state of Morelos, Mexico, and extraction was performedin situ by applying mild electro‐stimulation as describedpreviously.[10] Frogs were briefly washed beforehand withtetra‐distilled water, electro‐stimulated and, subsequently, amore thorough wash was performed, in order to remove skinexudates. Following extraction, all specimens were deliveredto their native environment in a healthy state. Secretions fromeleven specimens were collected and subsequently pooledand then acidified with trifluoroacetic acid (TFA), beforebeing centrifuged at 15 000 rpm for 10min. The supernatantwas frozen at −20°C, until processing.

Enrichment of sulfopeptides by IMAC‐Ga

Sulfopeptides were isolated using tips with Ga‐IDA (gallium‐iminodiacetic acid) from Thermo Scientific‐Pierce (San Jose,CA, USA), in accordance with the manufacturer’s instruc-tions. Approximately 5μg of crude secretions was adjusted topH 3.0 by adding 50μL 0.1% acetic acid. The sample was firstloaded onto the column and then washed with 75μL of 0.1%acetic acid, followed by 75μL of 0.1% acetic acid in 10%acetonitrile and finally with 75μL water. The sample wassubsequently eluted with 20μL of 100mM ammonium

wileyonlinelibrary.com/journal/rcm Copyright © 2011 John Wile

bicarbonate. No phosphate or sulfate salts were used in thisprotocol.

HPLC fractionation

Two total fractions obtained from IMAC‐Ga enrichment wereindependently fractionated in an HPLC Agilent‐1200 system(Agilent, Santa Clara, CA, USA) comprised of quaternarypumps, a UV detector, an automatic degasser and a manualinjector connected to a RP C18 analytical column by metallines. Approximately 200μg of a Ga‐enriched fraction wasloaded onto an analytical Vydac C18 RP column (Hisperia,CA, USA) with a linear gradient ranging from 0% buffer A(0.1% TFA/water) to 60% buffer B (0.1% TFA/acetonitrile)and then run for 60min, at a flow rate of 1.0mL/min.Elution of the peptides was monitored by UV detection at230 nm. All fractions were collected manually and evapo-rated, using a Speed Vac Savant from ThermoFisher Co. (SanJose, CA, USA).

Mass spectrometric analysis

MALDI‐TOF

A sample obtained from the crude extract taken from the skinsecretion of the P. dacnicolor frog and a gallium‐enrichedfraction taken from the same extract were desalted with a ZipTip C18 (Millipore Co., Billerica, MA, USA) and spotted onto aMaldi plate, using α‐cyano‐4‐hydroxycinnamic acid (CHCA)matrix (10mg/mL in 50:50 ACN/0.1% TFA in water). MALDIspectra were recorded in positive‐ion mode using a 4800MALDI TOF/TOF analyzer (Applied Biosystems Inc.,Framingham, MA, USA), equipped with a Nd:YAG 355‐nmlaser. An accelerating voltage of 20 kV was applied and thereflectron voltage was 25 kV. Each spectrum represents thecumulative average of 400 profiles from each well.

LC/MS/MS

Samples were previously desalted using a ZipTip C18(Millipore Co., Billerica, MA, USA) and then analyzed byapplying LC/MS/MS. Chromatography was performedusing an Accela HPLC system (Thermo Fisher Scientific, SanJose, CA, USA) at a flow rate of 400 nL/min (splitter 19:1). Thecolumn was a PicoFrit Proteopep 2 C18 (75μm i.d. × 50mm;New Objective Inc., Woburn, MA, USA). Solvent Awas madeup of water/0.1% acetic acid, and solvent B consisted ofacetonitrile/0.1% acetic acid; peptides were eluted on agradient ranging from 5% to 70% solvent B for 120min.Eluted peptides were electrosprayedwith a nano‐electrosprayat a voltage of 2.0 kV into an LTQ‐Orbitrap XL massspectrometer (Thermo Fisher Scientific, San Jose, CA, USA),which is able to perform CID and HCD analysis. MS dataacquisition was performed automatically, using a methodspecifically devised for the purpose of sulfopeptide char-acterization: (i) all data were acquired with 30 000 resolutionin the FT‐Orbitrap analyzer in the positive ion mode; (ii) onlysingly, doubly and triply charged ions were selected fordissociation; (iii) dynamic exclusion of 60min was enabledand a pre‐exclusion list of 30 s was used; (iv) a data‐dependent scan performing HCD and CID on the three mostintense peaks, followed by MS3 events dependent on theneutral loss of 79.956, 39.978 and 26,652 from the previous

y & Sons, Ltd. Rapid Commun. Mass Spectrom. 2011, 25, 1017–1027

Sulfated peptides from the secretion of P. dacnicolor

CID spectra, was applied; (v) normalized collision energywas 35 (arbitrary units), with an activation Q of 0.250 andan activation time of 30ms for both CID and HCD. ManualMS acquisition was also performed by Tune Plus windowwhen peptides previously isolated by HPLC off‐line weredirectly infused into the instrument. In this way, dissocia-tion energies of 35, 45 and 55 were applied individually todissociate sulfopeptides by HCD. Likewise, all scans werealso performed in positive ion mode and CID and HCDfragmentation methods were used to dissociate all observedcharge state ions.

Data analysis

LC/MS analysis containing CID and HCD fragmentationdata were searched for using the program Sequest fromDiscoverer 1.0 (Thermo Fisher Scientific, San Jose, CA, USA)against a non‐redundant protein database. Raw files werealso converted into .dta format and searched for using theProtein Prospector (University of California in San Francisco)and Peaks Online (Bioinformatics Solutions, Canada) searchengines against SwissProt and nr.fasta databases. In the caseof CID data, a precursor mass error tolerance of 5 ppm and afragment mass tolerance of 0.1Da were permitted. Cysteineresidues were assumed to be carbamidomethylated. Othermodifications were taken into consideration including

Figure 1. MALDI‐TOF analysis of the selective enrichment of sulthe P. dacnicolor amphibian before enrichment with IMAC‐Ga (A).Asterisk (*) indicates the main sulfopeptides enriched. 1350.73peptides, phyllokinin and sauvatide‐like, respectively.

Copyright © 2011Rapid Commun. Mass Spectrom. 2011, 25, 1017–1027

peptide N‐terminal glutamine conversion into pyrogluta-mate, methionine oxidation, tyrosine, serine and threoninephosphorylation and sulfonation. Finally, all matched andunmatched spectra were analyzed manually.

RESULTS AND DISCUSSION

The IMAC‐Ga enrichment protocol was used for theselective isolation of sulfopeptides present in the total skinsecretions of the P. dacnicolor frog. The efficiency of theIMAC‐Ga enrichment protocol was monitored by LC/MSand MALDI‐TOF/TOF. Total peptide contents from the skinsecretion of the P. dacnicolor frog and the peptides obtainedafter enrichment with IMAC‐Ga were compared usingMALDI‐TOF/TOF (Figs. 1(A) and 1(B)). The enrichment ofsulfated peptides is clearly observed when the intensity ofthe ion at m/z 1446.69, corresponding to a sulfated peptide,was compared for both spectra. The insert on the upperright in Fig. 1(A) shows the great diversity of the peptidespresent in this amphibian secretion and the ionic suppres-sion effect along the molecular mass range (from 1220 to1540m/z), where the previously mentioned sulfated peptideis located. The efficiency of the IMAC‐Ga enrichment wasalso analyzed by LC/MS/MS. Figure 2(A) shows the total

fated peptides by IMAC‐Ga. Spectrum of the total secretion ofMALDI‐TOF spectrum of the IMAC‐Ga enriched fraction (B).and °1554.99 are the molecular weights of the non‐sulfated

wileyonlinelibrary.com/journal/rcmJohn Wiley & Sons, Ltd.

1019

Figure 2. Total ion current (TIC) chromatograms obtained by LC/MS/MS of the skin secretions of the P. dacnicolor frog. TICchromatogram of the total secretion presenting 234 different components as previously reported (A).[10] TIC chromatogram ofthe fraction enriched by an IMAC‐Ga microcolumn presenting 14 main peaks, which were labeled from ‘a’ to ‘n’ (B). Both TICchromatograms were obtained using a flow rate of 400 nL/min, a gradient system from 5 to 70% AcCN in 120min, and thedetection was carried out in a FT‐Orbitrap XL mass spectrometer.


1020
ion current (TIC) chromatogram for the secretions of theP. dacnicolor frog. The molecular contents of this secretion hadbeen studied previously and 194 different components weredetermined.[10] The TIC chromatogram of the IMAC‐Ga‐enriched fraction is shown in Fig. 2(B), where 14 main peakswere detected and labeled from ‘a’ to ‘n’. The decrease in thesample complexity can be clearly observed when both TICchromatograms are compared. The successful enrichment ofsulfated peptides using a protocol for phosphopeptideswas to be expected; since both groups share four equivalentoxygen atoms equally, in a tetrahedral arrangement. Theyalso manifest a negative charge and have similar molecularweights.One of the most abundant sulfopeptides present in the
skin secretion from P. dacnicolor is phyllokinin (RPPOH


GFTPFRIYSulf), which was previously detected in thesecretions of the Phyllomedusa sauvagei frog.[18] Distinctisoforms without hydroxyproline (Hyp) and/or sulfatedtyrosine were also found. The phyllokinin isoforms weredetected in the TIC chromatogram (Fig. 2(B)) at retentiontimes 22.58 and 23.68min, and are labeled with theletters ‘c’ and ‘d’, respectively. Phyllokinins containingHyp at position 3 and a sulfated tyrosine at position 11were detected by the doubly charged ion at m/z 723.847 inthe full scan MS event (Fig. 3(A)). The deconvolution of thedoubly charged signal permitted us to detect an experi-mental monoisotopic molecular mass of [M+H]+ 1446.694.When CID of the precursor ion at m/z 723.847 wasperformed, a strong signal was observed at m/z 683.869,corresponding to a sulfate neutral loss of 79.956 (Fig. 3(B)).


Figure 3. Structural analysis of the post‐translationally modified phyllokinin‐like peptide by MS/MS. Full scan spectrum of thedoubly charged ion at m/z 723.85 (A). MS2 spectrum of the precursor ion at m/z 723.85 showing an intense peak at m/z 683.87,corresponding to neutral loss of 39.9872+ Da (B). MS3 spectrum of the product ion 683.87 showing extensive fragmentation (C).HCD spectrum of the precursor ion doubly charged at m/z 723.85 (D). HCD spectrum from the precursor triply charged ion atm/z 656.69 of the sulfopeptide ‘e’ (E) and MS3 of the de‐sulfated triply charged ion at m/z 630.37 showing intense fragment of they ions series (F).


102

In order to obtain structural information referring to thepeptide, MS3 experiments of the ion at m/z 683.869 werealso carried out, indicating an extensive fragmentation ofthe peptide backbone corresponding mainly to the y ionseries (Fig. 3(C)). Unexpectedly, when the precursor ionwith m/z 723.847 was submitted to dissociation by HCD, noneutral loss of sulfate was observed and contrastingly they and b ion series were evident (Fig. 3(D)). The differentfragmentation pattern is due to the occurrence of multiplecollisions in the HCD octapole cell, where the amide bondcleavage also competes with neutral loss of sulfate,improving the quality of the data when comparedwith low‐energy CID performed in linear ion traps. Zhanet al.[19] have reported similar results when CID and HCDwere used in phosphopeptide studies. They found thatHCD is a highly efficient way of dissociating shortphosphopeptides containing 1–2 basic residues within the


sequence producing extensive fragmentation with nodominance of peaks corresponding to neutral loss. Ourresults concerning the HCD of sulfopeptides presented hereonly partially concurred with other previous results forphosphopeptides, as other sulfated peptides studied in thiswork did not manifest the same HCD fragmentation pattern.For example, the doubly (m/z 984.53) and triply charged(m/z 656.69) ion signals corresponding to a sulfated peptide(labeled as ‘e’ in Fig. 2) were submitted to HCD at differentconcentrations and different dissociation energies; however,no structural data was obtained (Fig. 3(E)). In order toaccess the primary structure of the peptide, CID‐MS3 wasperformed on the de‐sulfated triply charged ion at m/z 630.37,permitting a confident assignment of 9 amino acid residues(−YAKSFLAKW‐) to the sulfated peptide (Fig. 3(F)). Theresulting partial sequence was submitted for a search againstdifferent data banks but no significant matches were found.


1


1022

Phyllokinin is a short peptide, which possesses two basicresidues in its sequence and produces extensive fragmentationwhen submitted to HCD activation. However, this unknownpeptide fragment also possesses the structural featurescharacteristic of phyllokinin, but no elucidative HCD datawere obtained. It is clear that efficiency in terms of dissociatingsulfopeptides by HCD not only depends on the concentrationand the dissociation energy, but also on peptide sequence andpeptide length.A sulfated peptide previously isolated from the HPLC

fraction and eluted at a retention time of 30.72min was fullycharacterized by CID and HCD (Fig. 4). It presents apyroglutamic acid at the N‐terminus, possesses a sulfatedtyrosine at position four and the C‐terminal is amidated. Afterde novo sequencing and searching against protein data banks itwas identified as phyllocaerulein (pEEYsTGWMDF‐NH2).Phyllocaerulein is a well‐studied sulfated peptide which wascharacterized early on by Anastasi et al.[20] It is also found inthe skin secretions of P. dacnicolor and other frog speciesbelonging to the Phyllomedusinae genus. This componentwasdetected in the TIC chromatogram (Fig. 2(B)) at retention timeof 55.35min and labeled with the letter ‘n’.In the IMAC‐Ga‐enriched fraction, a dermaseptin phos-

phate adduct was also detected when submitted to CIDactivation (Figs. 5(A) and 5(C)), indicated by the fact thata neutral loss of 97.976Da was observed (Fig. 5(B)).Phosphate and sulfate adducts are commonly found in theMS spectra of proteins and peptides containing no protonatedstrong basic residues,when they are exposed to sulfuric acid orphosphoric acid. It is known that sulfate and/or phosphateimpurities present in the samples can produce smallamounts of corresponding acids during the ESI process.[21]

However, no phosphate salts were used in any of the sample‐processing steps performed in this work. We hypothesizedthat the identified adduct could have been formed as a result of

Figure 4. Collision‐induced dissociation (CID) of the singly charthe corresponding amino acid sequence of the de‐sulfated phyll


the presence of phosphate salts in the crude secretions, whichwere enriched by IMAC‐Ga and detected by MS. Thedermaseptin phosphate adduct was detected at a retentiontime of 32.77min and is labeledwith the letter ‘f’ in the LC/MSspectrum of the IMAC‐Ga‐enriched fraction (Fig. 2(B)).

Protease inhibitors have only been identified from the frogskin secretions of Phyllomedusa sauvagei and were namedPSKP‐1 and PSKP‐2 possessing molecular masses of 6.7 and6.6kDa, respectively.[22] Surprisingly, a very hydrophilic com-ponent that eluted at a retention time of 15.26min in theTIC chromatogram shown in Fig. 2(B) (letter ‘a’) was alsoenriched by the IMAC‐Ga protocol. This component hadbeen partially characterized previously by our [10] showinga high percentage of sequence identity with the previouslyreported protease inhibitors isolated from P. sauvagei andpossessing a molecular mass of 5799.59Da. However, CID ofthe signal at m/z 1161.17 [M+5H]5+, from the sample isolatedby HPLC, showed strong neutral loss signals of −98, −133, −213and −231Da obtained by deconvolution of the quadruplycharged ions (Fig. 6). The molecular mass differencesfound can be attributed to the presence of a complex PTMcorresponding to multiple pyridoxal phosphates (PPL), prob-ably forming a Schiff‐base linkage with the ε‐amino group oflysine. The CID spectrum shown in Fig. 6 demonstrates theneutral losses of phosphate by the difference between thequadruply charged ion atm/z [1450.957+4H]4+ and the fragmentat m/z [1426.188+4H]4+, a pyridoxal from [1426.188+4H]4+ to[1393.67+4H]4+, a PLP from [1450.957+4H]4+ to [1393.67+4H]4+,a second pyridoxal from [1397.918+4H]4+ to [1365.406+4H]4+, adehydro‐PLP from [1365.406+4H]4+ to [1312.384+4H]4+, and asecond PLP from [1365.406+4H]4+to [1308.135+5H]5+. Addi-tional signals at m/z [1141.548+5H]5+ and [1115.135+5H]5+

corresponding to neutral losses of phosphate and PLP,respectively, were also observed (data not shown). PLP(a modified form of the vitamin B6) is the unique PTM

ged ion at m/z 1158.47. The main b and y ion series values andocaerulein isoform are shown in the insert of the figure.


Figure 5. Partial sequence and phosphate‐adduct characterization from a dermaseptin‐like antimicrobial peptide. Full MSscan of triply and quadruply charged ions of the dermaseptin‐like peptide and the corresponding m/z signals from itsphosphate adduct (A). CID spectrum of the doubly charged ion at m/z 1070.597 showing an intense signal corresponding toneutral loss of 97.976Da (B). CID fragmentation of the triply charged ion at m/z 1037.938 and the partial sequence of thedermaseptin‐like (C).


102

involved in a number of catalyzing reactions such asdecarboxylation, amino group transfers, and amino acidisomerization.[23,24] CID spectra of PLP‐containing peptidesare similar to those presenting phosphate modification;including impaired backbone fragmentation.[25] This is the firstreport which shows a PLP attached to a protease inhibitorisolated from frog skin secretions. In future studies, it may bepossible to enrich protease inhibitors with IMAC‐Ga fromother amphibian species to determine the presence of PLP.


Additionally, extensive pharmacological and biochemicalscreening might be carried out in order to establish whetheramphibian PLP peptides are involved only in the inhibition ofproteases or whether they are able to catalyze an assortment ofreactions involved in a wide range of biological functions.

A sauvatine‐like peptide was unexpectedly detected inthe IMAC‐Ga‐enriched fraction, which is neither sulfated,nor phosphorylated, and nor is it an acid peptide. Thiscomponent possess a [M+H]+ monoisotopic molecular mass


3

Figure 6. CID spectrum of a protease inhibitor isolated from the skin secretion of the P. dacnicolor frog and the detection ofpyridoxal‐phosphate PTM. Dissociation of the precursor ion at m/z [1450.957+4H]4+ showing complex and multiple neutrallosses. The differences between the m/z values reveal neutral losses corresponding to phosphate (−Phos), pyridoxal (−Pyxl),pyridoxal phosphate (−PLP) and dehydro‐pyridoxal phosphate (Dehydro‐PLP).


1024

of 1554.995Da obtained by deconvolution of the signals atm/z [778.009+2H]2+ and [519.008+3H]3+. During the full MSscan event of this component, a phosphate adduct were alsoobserved by additional signals at m/z [826.997+2H]2+ and[551.666+3H]3+ (Fig. 7(A)). The signal at m/z 826.997 wassubmitted to fragmentation by CID showing a very intensesignal at m/z [778.009+2H]2+, whose difference correspondsto the neutral loss of phosphate (Fig. 7(B)). CID and HCDwere used to fragment this peptide and its full sequence isreported here for the first time (Fig. 7(C)). This peptide istwo amino acid residues longer than its previously de-scribed counterparts and matches the sequence of eightamino acids (2–10) present in the sauvatide peptide isolatedfrom P. sauvagei. However, it possesses four additionalamino acids at the C‐terminal region and the first aminoacid residue is replaced by phenylalanine. Table 1 presentsmultiple sequence alignment (ClustalW2) and sequenceidentity (Blast‐NCBI) for this sauvatine‐like peptide withthe most correlated peptides known.The sequences of sulfopeptides and non‐sulfated peptides

characterized in this work are presented in Table 2.Sulfonated amino acids and hydroxyprolines are presentedunder captions S and OH, respectively, following the aminoacids. Molecular weights are reported in Da as monoisotopicmasses [M+H]+.Seven dermaseptin‐like peptides were non‐specifically

enriched, using the IMAC‐Ga protocol. This fact may beexplained by their high concentration in the total secretion,or by their high ionization capacity or by the need to submitthem to a second enrichment process using the sameprotocol, or even an alternative one applying differentmetals such as Fe, Zr or TiO2. One trend in the futureconcerning large‐scale PTM analysis from a complex mixtureof proteins expressed at very low concentration will be the

Figure 7. MS characterization of the sauvatide‐like peptide. Detcorresponding to the peptide and its phosphate adduct (A). CIDsignal with m/z value of 708.008 corresponding to the neutral loss ofcharged ion at m/z 778.008 and the derived sequence (C).


development and optimization of protocols which are able toselect peptides containing specific chemical groups. In thecase of sulfoproteomics, a great effort must be made tooptimize a specific methodology in order to answer manyimportant questions concerning cellular behavior.

CONCLUSIONS

The IMAC‐Ga sulfopeptide enrichment protocol was suc-cessfully applied in order to isolate the main sulfopeptidesfrom the complex mixture of peptides present in the skinsecretion of the P. dacnicolor frog. However, the resultsindicate a clear lack of specificity. In spite of the lowspecificity of the protocol used, this seems to be a promisingmethod for high‐throughput studies involving proteinsulfonation, where well‐calibrated high‐resolution massspectrometers are applied. Neutral loss of 79.9568Da fromthe precursor sulfated ions generated by CID and the highaccuracy mass determination in the Orbitrap mass analyzerwere enough to distinguish between phosphate and sulfategroups. HCD experiments were used to fragment sulfopep-tides, which presented a doubtful performance, as in somecases these produced elucidative MS data and in others,no fragments were observed. This indicated that shortsulfopeptides containing one or two basic residues withintheir sequences are not the only requirements for obtainingMS/MS data with structural information. Another interest-ing observation was the dermaseptin and sauvatidephosphate adducts found in the fraction enriched withIMAC‐Ga, as absolutely no phosphate salts or phosphaterelated acids were used during the enrichment protocol.Finally, four sulfopeptides, two novel non‐sulfated peptides

ection of doubly and triply charged ions in the full MS scanof the double charged ion at m/z 826.997 producing an intense97.976Da (B). HCD (left) and CID (right) spectra of the doubly

►



wileyonlinelibrary.com/journal/rcmCopyright © 2011 John Wiley & Sons, Ltd.Rapid Commun. Mass Spectrom. 2011, 25, 1017–1027

1025

Table 1. CLUSTAL 2.0.12 multiple sequence alignment of Sauvatide correlated peptides

Peptide Sequence # AA Species % Identity

Sauvatide LR‐PAILVRTK‐‐‐‐ 10 (P. sauvagei) 90PDa‐1 ‐RFPAILVRTKGKGL 14 (P. dacnicolor) 100AMP‐1 LR‐PAVIVRTKAL‐‐ 12 (P. hypochondrialis) 58Hyposin‐3 LR‐PAVIVRTKGK‐‐ 12 (P. azurea) 75

* **::****

#AA: Number of amino acid residues; (:): conserved; *: equal

Table 2. Peptides characterized from the skin secretion of the Pachymedusa dacnicolor frog by IMAC‐Ga enrichmentand LC/MS analysis

PeakRT

(min) MW* (Da) Sequence Peptide

a 15.26 5,797.816 VIEPDCVKYHGSCKKSPNYICGTDGKTYY… Protease Inhibitor‐PLPb 19.47 1,554.995 RFPAXXVRTKGKGX‐NH2 Sauvatide‐like (novel)c 22.58 1,350.731;

1,367.740RPPGFTPFRIY; RPPOHGFTPFRIY Phyllokinin; POH‐

Phyllokinind 23.68 1,446.696;

1,430.696RPPOHGFTPFRIYs ; RPPGFTPFRIYs POH‐Phyllokinin‐Sulfated;

Phyllokinin‐Sulfatede 29.41 1,967.070 YAKSFXAKW… Unknown Sulfopeptidef 32.77 3,110.820 GMWKLKSVQLAALAAKAAKAA… Dermaseptin (Phos‐adduct)g 34.38 3,168.825 ‐ ‐h 36.69 2,871.632 GMWGKIKSTAKEAAKAAGKAALNAVSEAL‐

NH2

Dermaseptin‐like

i 38.40 2,870.632 GMWGKIKSTAKEAAKAAGKAALNAVSEAL‐NH2

Dermaseptin‐like

j 41.48 3,859.228 ‐ ‐k 42.43 4,300.480 ‐ ‐l 48.41 2,692.531 GVWGIAKIAGKVLGNILPHVFSSNQS Dermaseptin‐likem 50.61 3,205.875 ‐ ‐n 55.35 1,238.426;

1,158.469PyrEYsTGWMDF‐NH2; PyrEYTGWMDF‐NH2 Phyllocaerulin‐sulfated;

Phyllocaerulin

X: I/L; MW*: monoisotopic masses [M+H]+; Pyr: Pyroglutamic acid; PLP: Pyridoxal‐phosphate


1026

and a PPL‐protease inhibitor were structurally character-ized for the first time in the skin secretion of the P. dacnicolorfrog.

AcknowledgmentWe acknowledge Programa de Apoyo a Proyectos deInvestigación e Innovación Tecnológica (PAPIIT) of theDirección General de Asuntos de Personal Académico(DGAPA‐UNAM) for the partial support of this work bythe grant IN 72010 to CVFB.

REFERENCES

[1] O. N. Jensen. Nat. Rev. Mol. Cell Biol. 2006, 7, 391.[2] M. Mann, C. Choudhary. Nat. Rev. Mol. Cell Biol. 2010, 11,

427.[3] J. Villén, S. A. Beausoleil, S. P. Gygi. Proteomics 2008, 8, 4444.


[4] P. J. Boersema, S. Mohammed, A. J. Heck. J. Mass Spectrom.2009, 44, 861.

[5] T. Pawson, J. D. Scott. Trends Biochem. Sci. 2005, 30, 286.[6] J. Schlessinger. Cell 2000, 103, 211.[7] W. B. Huttner. Nature 1982, 299, 273.[8] K. L. Moore. J. Biol. Chem. 2003, 277, 23781.[9] J. R. Bundgaard, J. Vuust, J. F. Rehfeld. EMBO 1995, 14, 3073.

[10] E. P. Meneses, O. Villa‐Hernández, L. Hernández‐Orihuela,R. Castro‐Franco, V. Pando, C. V. Batista. Amino Acids 2010;Epub ahead of print.

[11] G. D. Brand, F. C. Krause, L. P. Silva, J. R. Leite, J. A. Melo,M. V. Prates, J. B. Pesquero, E. L. Santos, C. R. Nakaie,C. M. Costa‐Neto, C. Bloch Jr. Peptides 2006, 27, 2137.

[12] P. A. Wabnitz, J. H. Bowie, M. J. Tyler. Rapid Commun. MassSpectrom. 1999, 13, 2498.

[13] J. Ye, X. Zhang, C. Young, X. Zhao, Q. Hao, L. Cheng,O. N. Jensen. J. Proteome Res. 2010; Epub ahead of print.

[14] J. Villén, S. P. Gygi. Nat. Protocols. 2008, 3, 1630.[15] C. E. Haydon, P. A. Eyers, L. D. Aveline‐Wolf, K. A. Resing,

J. L. Maller, N. G. Ahn. Mol. Cell. Proteomics 2003, 2, 1055.[16] C. Seibert, T. P. Sakmar Biopolymers 2007, 90, 459.



[17] K. F. Medzihradszky, S. Guan, D. A. Maltby, A. L.Burlingame. J. Am. Soc. Mass Spectrom. 2007, 18, 1617.

[18] T. Chen, C. Shaw. Peptides 2003, 24, 1123.[19] Y. Zhan, S. B. Ficarro, S. Li, J. A. Marto J. Am. Soc. Mass

Spectrom. 2009, 20, 1425.[20] A. Anastasi, G. Bertaccini, J. M. Cei, G. De Caro, V. Erspamer,

M. Impicciatore. Br. J. Pharmacol. 1969, 37, 198.[21] K. Swapan, V. K. Chowdhury, R. C. Beavis, B. T. Chait. J.

Am. Soc. Mass Spectrom. 1990, 1, 382.


[22] L. G. Gebhard, F. U. Carrizo, A. L. Stern, N. I. Burgardt,J. Faivovich, E. Lavilla, M. R. Ermácora. Eur. J. Biochem.2004, 271, 2117.

[23] K. A. Denessiouk, A. I. Denesyuk, J. V. Lehtonen, T.Korpela, M. S. Johnson. Proteins: Struct. Funct. Genet. 1999,35, 250.

[24] M. D. Toney. Arch. Biochem. Biophys. 2005, 433, 279.[25] E. S. Simon, J. Allison. Rapid Commun. Mass Spectrom. 2009,

23, 3401.


1027

Documents

Analysis of sulfated peptides from the skin secretion of the Pachymedusa dacnicolor frog using IMAC-Ga enrichment and high-resolution mass spectrometry