Gln-Gly cleavage: Correlation between collision-induced dissociation and biological degradation

Tryptic digestion of the 150-residue human acidic salivary proline-rich protein 1 (PRP-1) generated eight peptides, two of which corresponded to the N-terminal 30-residue segment. In each of the other six tryptic peptides, a consensus repeat with the structure PQGPPQQGG was present. A facile Gln-Gly cleavage between the second and the third residues of the repeat was observed during collision-induced dissociation experiments. We postulate possible mechanisms to account for this reactivity, involving attack on the peptidyl carbonyl group by the Gln sidechain. Significantly, the Gln-Gly cleavage has been shown to be biologically important in the bacterial degradation of PRPs in saliva, generating bacteria-binding Pro-Gln C-termini. We suggest a link between the gas-phase chemistry and the biochemical degradation of these molecules.     

Gas-phase cleavage of PTC-derivatized electrosprayed tryptic peptides in an FT-ICR trapped-ion cell: Mass-based protein identification without liquid chromatographic separation

Condensed phase protein sequencing typically relies on N-terminal labeling with phenylisothiocyanate ("Edman" reagent), followed by cleavage of the N-terminal amino acid. Similar Edman degradation has been observed in the gas phase by collision-activated dissociation of the N-terminal phenyl thiocarbamoyl protonated peptide [1] to yield complementary b1 and y(n-1) fragments, identifying the N-terminal amino acid. By use of infrared multiphoton (rather than collisional) activation, and Fourier transform ion cyclotron resonance (rather than quadrupole) mass analysis, we extend the method to direct analysis of a mixture of tryptic peptides. We validate the approach with bradykinin as a test peptide, and go on to analyze a mixture of 25 peptides produced by tryptic digestion of apomyoglobin. A b1+ ion is observed for three of the Edman-derivatized peptides, thereby identifying their N-terminal amino-acids. Search of the SWISS-PROT database gave a single hit (myoglobin, from the correct biological species), based on accurate-mass FT-ICR MS for as few as one Edman-derivatized tryptic peptide. The method is robust-it succeeds even with partial tryptic digestion, partial Edman derivatization, and partial MS/MS IRMPD cleavage. Improved efficiency and automation should be straightforward.     

Characterization of γ-carboxylated tryptic peptides by collision-induced dissociation and electron transfer dissociation mass spectrometry

Vitamin K-dependent carboxylation of glutamic acid (Glu) residues into γ-carboxyglutamic acid (Gla) is a post-translational modification essential for normal protein activity of, for example, proteins involved in the blood coagulation system. These proteins may contain as many as 12 sites for γ-carboxylation within a protein sequence of 45 amino acid residues. In the biopharmaceutical industry, powerful analytical techniques are required for identification and localization of modified sites. We here present comparatively easy and rapid methods for studies of Gla-containing proteins using recent technology. The performances of two mass spectrometric fragmentation techniques, collision-induced dissociation (CID) and electron transfer dissociation (ETD), were evaluated with respect to γ-carboxylated peptides, applying on-line LC-ion trap MS. ETD MS has so far not been reported for Gla-containing peptides and the applicability of CID for heavily γ-carboxylated proteins has not been evaluated. The anticoagulant protein, protein C, containing nine Gla-sites, was chosen as a model protein. After tryptic digestion, three peptides containing Gla-residues were detected by MS; a 1.2 kDa fragment containing two Gla-residues, a 4.5 kDa peptide containing seven residues and also the 5.6 kDa tryptic peptides containing all nine Gla-residues. Regarding the shortest peptide, both CID and ETD provided extensive peptide sequencing. For the larger peptides, fragmentation by CID resulted in loss of the 44 Da CO(2)-group, while little additional fragmentation of the peptide chain was observed. In contrast, ETD resulted in comprehensive fragmentation of the peptide backbone. The study demonstrates that the combination of both techniques would be beneficial and complementary for investigation of γ-carboxylated proteins and peptides.     

On-line capillary liquid chromatography tandem mass spectrometry on an ion trap/ reflectron time-of-flight mass spectrometer using the sequence tag database search approach for peptide sequencing and protein identification

Capillary high-performance liquid chromatography has been coupled on-line with an ion trap storage/reflectron time-of-flight mass spectrometer to perform tandem mass spectrometry for tryptic peptides. Selection and fragmentation of the precursor ions were performed in a three-dimensional ion trap, and the resulting fragment ions were pulsed out of the trap into a reflectron time-of-flight mass spectrometer for mass analysis. The stored waveform inverse Fourier transform waveform was applied to perform ion selection and an improved tickle voltage optimization scheme was used to generate collision-induced dissociation. Tandem mass spectra of various doubly charged tryptic peptides were investigated where a conspicuous y ion series over a certain mass range defined a partial amino acid sequence. The partial sequence was used to determine the identity of the peptide or even the protein by database search using the sequence tag approach. Several peptides from tryptic digests of horse heart myoglobin and bovine cytochrome c were selected for tandem mass spectrometry (MS/MS) where it was demonstrated that the proteins could be identified based on sequence tags derived from MS/MS spectra. This approach was also utilized to identify protein spots from a two-dimensional gel separation of a human esophageal adenocarcinoma cell line.     

Analysis of photo-oxidized amino acids in tryptic peptides of calf lens gamma-II crystallin.

Insoluble and crosslinked proteins and increased pigmentation in the eye lens are features of aging and cataracts. Determining the amino acids which are involved in insolubilization, crosslinking and visible light scattering will shed light on the mechanisms by which cataracts form. Calf lens gamma-II crystallin was irradiated at 295 nm, digested and separated into tryptic peptides. Additional tryptic peptides were found in the digest of irradiated gamma-II which were not present in the dark control digest. These peptides were identified by amino acid sequencing and shown to correspond to expected tryptic fragments of the protein, indicating more facile digestion in the UV-irradiated protein than in dark controls. Amino acid analysis of the irradiated protein and peptides showed losses of histidine, methionine and cysteine residues as compared to control samples. Tryptophan, which is not detected by amino acid analysis, was also found to be reactive since losses in its fluorescence intensity were observed after irradiation. Some of the photochemically active amino acids had lower than expected responses in amino acid sequencing experiments. This suggested specific sites of photochemical activity in the various peptides. The evidence for peptide crosslinks is also discussed.     

ANALYSIS OF PHOTO-OXIDIZED AMINO ACIDS IN TRYPTIC PEPTIDES OF CALF LENS γ-II CRYSTALLIN

Insoluble and crosslinked proteins and increased pigmentation in the eye lens are features of aging and cataracts. Determining the amino acids which are involved in insolubilization, crosslinking and visible light scattering will shed light on the mechanisms by which cataracts form. Calf lens γ-II crystallin was irradiated at 295 nm, digested and separated into tryptic peptides. Additional tryptic peptides were found in the digest of irradiated γ-II which were not present in the dark control digest. These peptides were identified by amino acid sequencing and shown to correspond to expected tryptic fragments of the protein, indicating more facile digestion in the UV-irradiated protein than in dark controls. Amino acid analysis of the irradiated protein and peptides showed losses of histidine. methionine and cysteine residues as compared to control samples. Tryptophan, which is not detected by amino acid analysis, was also found to be reactive since losses in its fluorescence intensity were observed after irradiation. Some of the photochemically active amino acids had lower than expected responses in amino acid sequencing experiments. This suggested specific sites of photochemical activity in the various peptides. The evidence for peptide crosslinks is also discussed.     

'De novo' sequencing of Atlantic cod vitellogenin tryptic peptides by matrix-assisted laser desorption/ionization quadrupole time-of-flight tandem mass spectrometry: similarities with haddock vitellogenin

Vitellogenin is a protein produced by the liver of oviparous animals in response to circulating estrogens. The amino acid sequence of vitellogenin from Atlantic cod (Gadus morhua) has not yet been determined. In this study Atlantic cod vitellogenin was characterized using a 'bottom-up' mass spectrometric approach. Vitellogenin synthesis was induced 'in vivo' with beta-estradiol, and subjected to trypsin digestion for characterization by matrix-assisted laser desorption/ionization quadrupole time-of-flight tandem mass spectrometry. A peptide mass fingerprint was obtained and 'de novo' sequencing of the most abundant tryptic peptides was performed by low-energy collision-induced dissociation tandem mass spectrometry. As a result of these experiments, the sequences of various tryptic peptides have been elucidated. The database search has shown that Atlantic cod vitellogenin shares a series of common peptides with the two different known vitellogenin sequences of haddock, a closely related species. These findings allow us to propose that Atlantic cod might also co-express at least two distinct forms of vitellogenin.     

A combination of chemical derivatisation and improved bioinformatic tools optimises protein identification for proteomics

The identification of individual protein species within an organism's proteome has been optimised by increasing the information produced from mass spectral analysis through the chemical derivatisation of tryptic peptides and the development of new software tools. Peptide fragments are subjected to two forms of derivatisation. First, lysine residues are converted to homoarginine moieties by guanidination. This procedure has two advantages, first, it usually identifies the C-terminal amino acid of the tryptic peptide and also greatly increases the total information content of the mass spectrum by improving the signal response of C-terminal lysine fragments. Second, an Edman-type phenylthiocarbamoyl (PTC) modification is carried out on the N-terminal amino acid. The renders the first peptide bond highly susceptible to cleavage during mass spectrometry (MS) analysis and consequently allows the ready identification of the N-terminal residue. The utility of the procedure has been demonstrated by developing novel bioinformatic tools to exploit the additional mass spectral data in the identification of proteome proteins from the yeast Saccharomyces cerevisiae. With this combination of novel chemistry and bioinformatics, it should be possible to identify unambiguously any yeast protein spot or band from either two-dimensional or one-dimensional electropheretograms.     

The effect of phosphorylation on the electron capture dissociation of peptide ions

The effect of site and frequency of phosphorylation on the electron capture dissociation of peptide ions has been investigated. The ECD of a suite of synthetic peptides (APLSFRGSLPKSYVK; one unmodified, three singly-phosphorylated, three-doubly phosphorylated, and one triply-phosphorylated); two tryptic phosphopeptides (YKVPQLEIVPN(p)SAEER, alpha-casein and FQ(p)SEEQQQTEDELQDK, beta-casein) and their unmodified counterparts, were determined over a range of ECD cathode potentials. The results show that, for doubly-charged precursor ions, the presence of phosphorylation has a deleterious effect on ECD sequence coverage. The fragmentation patterns observed suggest that for peptides with multiple basic residues, the phospho-groups exist in their deprotonated form and form salt-bridges with protonated amino acid side chains. The fragmentation observed for the acidic tryptic peptides suggested the presence of noncovalent interactions, which were perturbed on phosphorylation. Increasing the ECD electron energy significantly improves sequence coverage. Alternatively, improved sequence coverage can be achieved by performing ECD on triply-charged precursor ions. The findings are important for the understanding of gas-phase fragmentation of phosphopeptides.     

Sequencing Lys-N Proteolytic Peptides by ESI and MALDI Tandem Mass Spectrometry

In this study, we explored the MS/MS behavior of various synthetic peptides that possess a lysine residue at the N-terminal position. These peptides were designed to mimic peptides produced upon proteolysis by the Lys-N enzyme, a metalloendopeptidase issued from a Japanese fungus Grifola frondosa that was recently investigated in proteomic studies as an alternative to trypsin digestion, as a specific cleavage at the amide X-Lys chain is obtained that provides N-terminal lysine peptide fragments. In contrast to tryptic peptides exhibiting a lysine or arginine residue solely at the C-terminal position, and are thus devoid of such basic amino acids within the sequence, these Lys-N proteolytic peptides can contain the highly basic arginine residue anywhere within the peptide chain. The fragmentation patterns of such sequences with the ESI-QqTOF and MALDI-TOF/TOF mass spectrometers commonly used in proteomic bottom-up experiments were investigated.     

Sequence Elucidation of an Unknown Cyclic Peptide of High Doping Potential by ETD and CID Tandem Mass Spectrometry

Identification of an unknown substance without any information remains a daunting challenge despite advances in chemistry and mass spectrometry. However, an unknown cyclic peptide in a sample with very limited volume seized at a Pennsylvania racetrack has been successfully identified. The unknown sample was determined by accurate mass measurements to contain a small unknown peptide as the major component. Collision-induced dissociation (CID) of the unknown peptide revealed the presence of Lys (not Gln, by accurate mass), Phe, and Arg residues, and absence of any y-type product ion. The latter, together with the tryptic digestion results of the unusual deamidation and absence of any tryptic cleavage, suggests a cyclic structure for the peptide. Electron-transfer dissociation (ETD) of the unknown peptide indicated the presence of Gln (not Lys, by the unusual deamidation), Phe, and Arg residues and their connectivity. After all the results were pieced together, a cyclic tetrapeptide, cyclo[Arg-Lys-N(C₆H₉)Gln-Phe], is proposed for the unknown peptide. Observations of different amino acid residues from CID and ETD experiments for the peptide were interpreted by a fragmentation pathway proposed, as was preferential CID loss of a Lys residue from the peptide. ETD was used for the first time in sequencing of a cyclic peptide; product ions resulting from ETD of the peptide identified were categorized into two types and named pseudo-b and pseudo-z ions that are important for sequencing of cyclic peptides. The ETD product ions were interpreted by fragmentation pathways proposed. Additionally, multi-stage CID mass spectrometry cannot provide complete sequence information for cyclic peptides containing adjacent Arg and Lys residues. The identified cyclic peptide has not been documented in the literature, its pharmacological effects are unknown, but it might be a “designer” drug with athletic performance-enhancing effects.     

Periodic sequence distribution of product ion abundances in electron capture dissociation of amphipathic peptides and proteins

The rules for product ion formation in electron capture dissociation (ECD) mass spectrometry of peptides and proteins remain unclear. Random backbone cleavage probability and the nonspecific nature of ECD toward amino acid sequence have been reported, contrary to preferential channels of fragmentation in slow heating-based tandem mass spectrometry. Here we demonstrate that for amphipathic peptides and proteins, modulation of ECD product ion abundance (PIA) along the sequence is pronounced. Moreover, because of the specific primary (and presumably secondary) structure of amphipathic peptides, PIA in ECD demonstrates a clear and reproducible periodic sequence distribution. On the one hand, the period of ECD PIA corresponds to periodic distribution of spatially separated hydrophobic and hydrophilic domains within the peptide primary sequence. On the other hand, the same period correlates with secondary structure units, such as α-helical turns, known for solution-phase structure. Based on a number of examples, we formulate a set of characteristic features for ECD of amphipathic peptides and proteins: (1) periodic distribution of PIA is observed and is reproducible in a wide range of ECD parameters and on different experimental platforms; (2) local maxima of PIA are not necessarily located near the charged site; (3) ion activation before ECD not only extends product ion sequence coverage but also preserves ion yield modulation; (4) the most efficient cleavage (e.g. global maximum of ECD PIA distribution) can be remote from the charged site; (5) the number and location of PIA maxima correlate with amino acid hydrophobicity maxima generally to within a single amino acid displacement; and (6) preferential cleavage sites follow a selected hydrogen spine in an α-helical peptide segment. Presently proposed novel insights into ECD behavior are important for advancing understanding of the ECD mechanism, particularly the role of peptide sequence on PIA. An improved ECD model could facilitate protein sequencing and improve identification of unknown proteins in proteomics technologies. In structural biology, the periodic/preferential product ion yield in ECD of α-helical structures potentially opens the way toward de novo site-specific secondary structure determination of peptides and proteins in the gas phase and its correlation with solution-phase structure.     

A database of 660 peptide ion cross sections: Use of intrinsic size parameters for bona fide predictions of cross sections

An ion trap/ion mobility/time-of-flight mass spectrometry technique has been used to measure collision cross sections for 660 peptide ions generated by tryptic digestion of 34 common proteins. Measured cross sections have been compiled into a database that contains peptide molecular weight and sequence information. The database is used to generate average intrinsic contributions to cross section (size parameters) for different amino acid residues by solving systems of equations that relate the unknown contributions of individual residues to the sequences and cross sections of database peptides. Size parameters are combined with information about amino acid composition to calculate cross sections for database peptides. Bona fide cross section predictions (made prior to measurement) for peptides observed in tryptic digests of sperm whale myoglobin and yeast enolase are made. Eight of 10 predicted cross sections are within 2% of the experimental values and all 10 are within 3.2%. The utility of size parameters for cross section prediction is explored and discussed.     

Backbone cleavages of [M - H](-) anions of peptides. Cyclisation of citropin 1 peptides involving reactions between the C-terminal [CONH](-) residue and backbone amide carbonyl groups. A new type of beta cleavage: a joint experimental and theoretical study

This paper reports the study of backbone cleavages in the collision-induced negative-ion mass spectra of the [M - H](-) anions of some synthetic modifications of the bioactive amphibian peptide citropin 1 (GLFDVIKKVASVIGGL-NH(2)). The peptides chosen for study contain no amino acid residues which could effect facile side-chain cleavage, i.e. Ser (-CH(2)O, side-chain cleavage) and Asp (-H(2)O) are replaced by Ala or Lys. We expected that such peptides should exhibit standard and pronounced peaks due to alpha cleavage ions (and to a lesser extent beta cleavage ions) in their collision-induced negative-ion spectra. This expectation was realised, but the spectra also contained peaks formed by a new series of cleavage anions. These are produced following cyclisation of the C-terminal CONH(-) moiety at carbonyl functions of amide groups along the peptide backbone; effectively transferring the NH of the C-terminal CONH(-) group to other amino acid residues. We have called the product anions of these processes beta' ions, in order to distinguish them from standard beta ions. Some beta' ions also fragment directly to some other beta' ions of smaller mass. The reaction coordinates of alpha,beta and beta' backbone processes have been calculated at the HF/6-31G*//AM1 level theory for simple model systems. The initial cyclisation step of the beta' sequence is barrierless and exothermic. Subsequent steps have a maximum barrier of +40 kcal mol(-1), with the overall reaction being endothermic by some 30 kcal mol(-1) at the level of theory used. These calculations take no account of the complexity of the conformationally flexible peptide system, and it is surprising that each of the two reacting centres can 'find' each other in such a large system.     

Improved sequencing of oxidized cysteine and methionine containing peptides using electron transfer dissociation

Oxidative modifications to the side chains of sulfur-containing amino acids often limit the number of product ions formed during collision-induced dissociation (CID) and thus make it difficult to obtain sequence information for oxidized peptides. In this work, we demonstrate that electron-transfer dissociation (ETD) can be used to improve the sequence information obtained from peptides with oxidized cysteine and methionine residues. In contrast to CID, ETD is found to be much less sensitive to the side-chain chemistry, enabling extensive sequence information to be obtained in cases where CID fails to provide this information. These results indicate that ETD is a valuable technique for studying oxidatively modified peptides and proteins. In addition, we report a unique and very abundant product ion that is formed in the CID spectra of peptides having N-terminal cysteine sulfinic acid residues. The mechanism for this unique dissociation pathway involves a six-membered cyclic intermediate and leads to the facile loss of NH(3) and SO(2), which corresponds to a mass loss of 81 Da. While the facile nature of this dissociation pathway limits the sequence information present in CID spectra of peptides with N-terminal cysteine sulfinic acid residues, extensive sequence information for these peptides can be obtained with ETD.     

Capillary electrophoresis-atmospheric pressure ionization mass spectrometry for the characterization of peptides. Instrumental considerations for mass spectrometric detection.

On-line capillary electrophoresis (CE) separations are shown for a synthetic peptide mixture and a tryptic digest of human hemoglobin in an uncoated fused-silica capillary with detection using atmospheric pressure ionization mass spectrometry (API-MS). The CE system utilized a 1-m capillary column of either 75- or 100-microns I.D. These somewhat larger inside diameters allow higher sample capacities for MS detection and the 1-m length facilitates connecting the CE column to the liquid junction-ion spray interface and MS system. Low volatile buffer concentrations (15-20 mM) of ammonium acetate or ammonium formate, and high organic modifier content (5-50%) of methanol or acetonitrile facilitates ionization under electrospray conditions. This study shows that peptides separated by CE may be transferred to the API-MS system through a liquid junction coupling to the pneumatically assisted electrospray (ion spray) interface at low buffer pH when the electroosmotic flow is low (0-0.04 microliter/min). CE-MS as described herein is facilitated by features in modern CE instrumentation including robotic cleaning and pressurization of the capillary inlet. The latter is particularly useful for repetitive rinsing and conditioning of the capillary column between analyses in addition to continuous 'infusion' of sample to the mass spectrometer for tuning purposes. In addition to facile molecular weight determination, amino acid sequence information for peptides may be obtained by utilizing on-line tandem MS. After the tryptic digest sample components enter the API-MS system, the molecular ion species of individual peptides may be focussed and transmitted into the collision cell of the tandem triple quadrupole mass spectrometer. Collision-induced dissociation of protonated peptide molecules yielded structural information for their characterization following injection of 10 pmol of a tryptic digest from human hemoglobin.     

Effect of N-terminal glutamic acid and glutamine on fragmentation of peptide ions

A prominent dissociation path for electrospray generated tryptic peptide ions is the dissociation of the peptide bond linking the second and third residues from the ammo-terminus. The formation of the resulting b₂ and y _n−2 fragments has been rationalized by specific facile mechanisms. An examination of spectral libraries shows that this path predominates in diprotonated peptides composed of 12 or fewer residues, with the notable exception of peptides containing glutamine or glutamic acid at the N-terminus. To elucidate the mechanism by which these amino acids affect peptide fragmentation, we synthesized peptides of varying size and composition and examined their MS/MS spectra as a function of collision voltage in a triple quadrupole mass spectrometer. Loss of water from N-terminal glutamic acid and glutamine is observed at a lower voltage than any other fragmentation, leading to cyclization of the terminal residue. This cyclization results in the conversion of the terminal amine group to an imide, which has a lower proton affinity. As a result, the second proton is not localized at the N-terminus but is readily transferred to other sites, leading to fragmentation near the center of the peptide. Further confirmation was obtained by examining peptides with N-terminal pyroglutamic acid and N-acetyl peptides. Peptides with N-terminal proline maintain the trend of forming b₂ and y _n−2 because their ring contains an imine rather than imide and has sufficient proton affinity to retain the proton at the N-terminus.     

A computer program (COMPOST) for predicting mass spectrometric information from known amino acid sequences

A computer program (COMPOST) is described that carries out predictive computations on known amino acid sequences. The program is designed to be of use to mass spectrometrists with an interest in protein and peptide sequencing. Mass values (monoisotopic and average) for protonated peptide and protein molecules and elemental compositions are calculated. COMPOST also calculates mass to charge ratio values for protonated peptides expected from specified digests, locates specified amino acid subsequences or peptides of a specifIed molecular weight within a longer sequence, and predicts mass to charge ratio values for fragment ions from high-energy collision-induced dissociation of protonated peptides.     

CHASE, a charge-assisted sequencing algorithm for automated homology-based protein identifications with matrix-assisted laser desorption/ionization time-of-flight post-source decay fragmentation data

We describe CHASE, a novel algorithm for automated de novo sequencing based on the mass spectrometric (MS) fragmentation analysis of tryptic peptides. This algorithm is used for protein identification from sequence similarity criteria and consists of four steps: (1) derivatization of tryptic peptides at the N-terminus with a negatively charged reagent; (2) post-source decay (PSD) fragmentation analysis of peptides; (3) interpretation of the mass peaks with the CHASE algorithm and reconstruction of the amino acid sequence; (4) transfer of these data to software for protein identifications based on sequence homology (Basic Local Alignment Search Tool, BLAST). This procedure deduced the correct amino acid sequence of tryptic peptide samples and also was able to deduce the correct sequence from difficult mass patterns and identify the amino acid sequence. This allows complete automation of the process starting from MS fragmentation of complex peptide mixtures at low concentration (e.g. from silver-stained gel bands) to identification of the protein. We also show that if PSD data are collected in a single spectrum (instead of the segmented mode offered by conventional matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) instrumentation), the complete workflow from MS-PSD data acquisition to similarity-based identification can be completely automated. This strategy may be applied to proteomic studies for protein identification based on automated de novo sequencing instead of MS or tandem MS patterns. We describe the Charge Assisted Sequencing Engine (CHASE) algorithm, the working protocol, the performance of the algorithm on spectra from MALDI-TOFMS and the data comparison between a TOF and a TOF-TOF instrument.     

Charge remote fragmentation in electron capture and electron transfer dissociation

Secondary fragmentations of three synthetic peptides (human αA crystallin peptide 1-11, the deamidated form of human βB2 crystallin peptide 4-14, and amyloid β peptide 25-35) were studied in both electron capture dissociation (ECD) and electron-transfer dissociation (ETD) mode. In ECD, in addition to c and z· ion formations, charge remote fragmentations (CRF) of z· ions were abundant, resulting in internal fragment formation or partial/entire side-chain losses from amino acids, sometimes several residues away from the backbone cleavage site, and to some extent multiple side-chain losses. The internal fragments were observed in peptides with basic residues located in the middle of the sequences, which was different from most tryptic peptides with basic residues located at the C-terminus. These secondary cleavages were initiated by hydrogen abstraction at the α-, β-, or γ-position of the amino acid side chain. In comparison, ETD generates fewer CRF fragments than ECD. This secondary cleavage study will facilitate ECD/ETD spectra interpretation, and help de novo sequencing and database searching.