Flexible Xxx–Asp/Asn and Gly–Xxx Residues of Equine Cytochrome c in Matrix-Assisted Laser Desorption/Ionization In-Source Decay Mass Spectrometry

Abstract

The backbone flexibility of a protein has been studied from the standpoint of the susceptibility of amino acid residues to in-source decay (ISD) in matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS). Residues more susceptible to MALDI-ISD, namely Xxx–Asp/Asn and Gly–Xxx, were identified from the discontinuous intense peak of c′-ions originating from specific cleavage at N–C_α bonds of the backbone of equine cytochrome c. The identity of the residues susceptible to ISD was consistent with the known flexible backbone amides as estimated by hydrogen/deuterium exchange (HDX) experiments. The identity of these flexible amino acid residues (Asp, Asn, and Gly) is consistent with the fact that these residues are preferred in flexible secondary structure free from intramolecular hydrogen-bonded structures such as α-helix and β-sheet. The MALDI-ISD spectrum of equine cytochrome c gave not only intense N-terminal side c′-ions originating from N–C_α bond cleavage at Xxx–Asp/Asn and Gly–Xxx residues, but also C-terminal side complement z′-ions originating from the same cleavage sites. The present study implies that MALDI-ISD can give information about backbone flexibility of proteins, comparable with the protection factors estimated by HDX.

Introduction

The flexibility of proteins is associated with “looseness” and “rigidness” of the backbone of polypeptides, and is also of importance from the standpoints of protein evolution and design,^1,2) and intramolecular self-organization such as in secondary and tertiary structure formation. The backbone flexibility of peptide and protein is often studied using solution-based NMR spectroscopy coupled with hydrogen/deuterium exchange (HDX).³⁾ Although mass spectrometry (MS) coupled with soft ionization methods such as matrix-assisted laser desorption/ionization (MALDI)^4–6) and electrospray ionization (ESI)^7,8) are widely used for identification of protein due to its sensitivity and rapidity, it is in general difficult for MS to obtain detailed information about the backbone flexibility of protein. Recent top-down MS strategies have addressed the analysis of secondary structures,⁹⁾ folding processes and identification of protein,^10–13) by using the specific backbone cleavage methods of electron capture dissociation and/or electron transfer dissociation coupled to ESI MS and by using in-source decay (ISD)^14,15) coupled with MALDI. MALDI-ISD has provided unique strategies for molecular imaging,^16,17) collision-induced dissociation combined with ISD,^16,18,19) and conformation analysis of protein.^20,21)

MALDI-ISD is a rapid fragmentation process which occurs within several tens of nanoseconds in the ion source, leading to specific cleavage at the N–C_α bond of the peptide backbone. The N–C_α bond cleavage occurs due to hydrogen radicals released from the hydrogen-donating matrix.^22,23) Following formation of transient hypervalent radical species mainly results in amino (N)-terminal side c′-ions and a radical species z^· (Scheme 1), while carboxyl (C)-terminal side z′-ion²⁴⁾ and collision-induced dissociation (CID)-like b- and/or y-ions originating from thermal pathways²⁵⁾ are observed. Early MALDI-ISD studies of backbone flexibility of protein showed that the N–C_α bond between Gly35 and His36 residues lying in a turn region between α-helix segments 21–35 and 37–40 of myoglobins was more susceptible to ISD than the rest of the protein,²⁶⁾ while the N–C_α bond between Xxx–Gly, Xxx–Val and Xxx–Ile residues was less susceptible to ISD than between other amino acid residues.^27,28) We have recently reported that in ISD spectra of equine myoglobin, equine cytochrome c and bovine serum albumin the N–C_α bonds between Xxx–Asp, Xxx–Asn and Gly–Xxx residues, which are preferred in flexible secondary structures such as turn and bend than in intra-molecular hydrogen-bonded structures such as α-helix, are more susceptible to MALDI-ISD than other residues.²⁹⁾ It is therefore of interest to examine the relationships between the amino acids susceptible to MALDI-ISD and the rate of HDX of the backbone amides of protein, because amino acid residues free from intra-molecular hydrogen-bonded helix and sheet would be expected to have enhanced hydrogen radical attachment and HDX. With respect to this, Rand et al. have reported that MALDI-ISD experiments coupled with HDX provide practical information about protein conformation which gives detailed mapping of dynamic structure particularly with regard to the backbone flexibility of a protein in solution.²¹⁾

Scheme 1. The formation mechanism for c′-ions and radical z^·-ion in MALDI-ISD.

Here I describe that Xxx–Asp/Asn and Gly–Xxx residues are more susceptible to MALDI-ISD in equine cytochrome c with the susceptibility comparable to the protection parameters calculated by HDX experiments in equine cytochrome c.³⁰⁾ It will be further reported that the MALDI-ISD spectrum of equine cytochrome c obtained with the newly introduced matrix 5-amino-1-naphthol (5,1-ANL)²⁹⁾ gives z′-ions with Asp/Asn residues at the N-terminus. The resulting z′-ions have complemented the c′-ions originating from the N–C_α bond cleavage at the Xxx–Asp/Asn residues of the protein. The ISD spectrum showed the covalent complex ions [z-ANL+H]⁺ generated from radical z^·-ions and matrix radical 5,1-ANL.

Experimental

Materials

The MALDI matrix 5-amino-1-naphthol (5,1-ANL) was purchased from Tokyo Chemical Industry (Tokyo, Japan). Acetonitrile was purchased from Wako Pure Chemicals (Osaka, Japan). Water used in all experiments was purified using a MilliQ water purification system from Millipore (Billerica, MA, USA). Equine cytochrome c (M_r 12360.4) was purchased from Sigma (Milwaukee, WI, USA). All reagents were used without further purification.

Sample preparation

Analyte was dissolved in water at a concentration of 20 pmol/µL. The matrix material was dissolved in water–acetonitrile (3 : 7, v/v) without any acid additives. A sample solution was prepared by mixing a volume of 10 µL of analyte solution with a volume of 10 µL of matrix solution. A volume of 0.5 µL of the sample solution was deposited onto a stainless-steel MALDI plate and the solvents were removed by allowing evaporation in air at room temperature.

MALDI MS

MALDI-ISD spectrum was acquired on a time-of-flight mass spectrometer AXIMA-CFR (Shimadzu, Kyoto, Japan) equipped with a nitrogen laser (337 nm wavelength) operating at a pulse rate of 10 Hz. The pulse width of the laser was 4 ns. The laser spot size on the target substrate was about 100 µm in diameter. The ions generated by MALDI were accelerated using 20 kV with delayed extraction. The analyzer was operated in linear mode and the ions were detected using a microchannel plate detector. A total of 500 shots were accumulated for each mass spectrum acquisition. The reproducibility of the ISD spectrum was confirmed.

Results and Discussion

Identifying flexible amino acid residues of equine cytochrome c in MALDI-ISD

Positive ion MALDI-ISD spectrum of equine cytochrome c obtained with 5,1-ANL as matrix, which resulted in steady generation of ISD fragment ions with high reproducibility and without sweetspots,²⁹⁾ is shown in Fig. 1. Discontinuous intense peaks of c′-ions were observed as c₂₄, c₃₀, c₃₄, c₃₇, c₄₅, c₄₉, c₅₁, c₅₃, c₆₉ and c₇₉ ions originating from the N–C_α bond cleavage at Gly24–Lys25, Pro30–Asn31, Gly34–Leu35, Gly37–Arg38, Gly45–Phe46, Thr49–Asp50, Ala51–Asn52, Lys53–Asn54, Glu69–Asn70, and Lys79–Met80, respectively. To estimate the susceptibility of amino acid residues to ISD, the intensity of each c′-ion was defined by the ratio of the intensity of n-th c′-ion to the average intensity of adjacent to both sides of the peaks as follows:

Fig. 1. MALDI-ISD spectrum of equine cytochrome c (M_r 12360) with 5-amino-1-naphthol (5,1-ANL) matrix. The c₂₄, c₃₀, c₃₄, c₃₇, c₄₅, c₄₉, c₅₁, c₅₃, c₆₉ and c₇₉ ions represent the outstanding higher peaks of c′-ions than adjacent to both sides of the peaks.

It may be expected that amino acid residues possessing the values beyond unity are more susceptible to MALDI-ISD than other residues. By this criterion, several amino acid residues susceptible to ISD could be taken from the red line in Fig. 2. Relatively intense c′-ions and values beyond unity correspond to c₂₄ (1.31), c₃₀ (1.73), c₃₄ (1.22), c₃₇ (1.21), c₄₅ (1.48), c₄₉ (1.88), c₅₁ (1.41), c₅₃ (1.64), c₆₉ (5.65) and c₇₉ (2.34) which in turn correspond to N–C_α bond cleavage of Gly24–Xxx, Xxx–Asn31, Gly34–Xxx, Gly37–Xxx, Gly45–Xxx, Xxx–Asp50, Xxx–Asn52, Xxx–Asn54, Xxx–Asn70 and Xxx–Met80, respectively. As described above, the amino acid residues with higher values of the c′-ion intensity beyond unity indicate greater susceptibility to ISD than the rest. The values described above therefore suggest that the Xxx–Asp, Xxx–Asn and Gly–Xxx residues are more susceptible to ISD than the other residues. The intensity value of zero which corresponds to cleavage at the N–C_α bond of Xxx–Pro is due to incomplete cleavage at the N-terminal side of the Pro residue. The c₆₉ ion has the highest intensity having a value of 5.65 and furthermore was observed by an accompanying complement peak corresponding to a C-terminal side ion of z₃₅ (Fig. 3). With regard to the complement z′-ions, it is known that the radical z^·-ions bind to matrix radicals to form covalent complex species (z-ANL in Fig. 3 and Scheme 2).³¹⁾ The other z′-ions which complement the c′-ions were observed as the pairs c₄₉/z₅₅ (Xxx–Asp50), c₃₇/z₆₇ (Gly37–Arg38) and c₃₀/z₇₄ (Xxx–Asn31, data not shown), and the corresponding complex ions z₅₅-ANL, z₆₇-ANL and z₇₄-ANL (data not shown) were also observed (Fig. 3). The observation of both the intense N-terminal side c′-ions and the C-terminal side complement z′-ions is very useful for identification of protein via a top-down proteomics strategy.

Fig. 2. Stacked relationship between the intensity of c′-ions, Int(c′-ion), estimated by MALDI-ISD (red line) and the protection factor, log P, estimated by HDX-NMR (blue line) in equine cytochrome c. The protection factors were obtained from the report by Milne et al.³⁰⁾ The amino acids more susceptible to MALDI-ISD, Pro30–Asn31, Thr49–Asp50, Ala51–Asn52, Lys53–Asn54, Lys53–Asn54 and Lys79–Met80, are consistent with the HDX data which identified the flexible amide hydrogens as Asn31, Thr49, Asn52, Lys53, Asn54 and Met80, respectively. The amino acid sequence and secondary structure of equine cytochrome c was obtained from X-ray crystallography data obtained from the PDB database (PDB ID: 1HRC).

Fig. 3. Partial MALDI-ISD spectra of equine cytochrome c (M_r 12360) with 5-amino-1-naphthol (5,1-ANL) matrix. The z₃₅, z₅₅ and z₆₇ ions represent C-terminal side species that complement N-terminal side c-ions c₆₉, c₄₉ and c₃₇, respectively. The z₃₅-ANL, z₅₅-ANL and z₆₇-ANL represent covalent complex between z^·-ions and 5,1-ANL radical species.

Scheme 2. The production of the covalent complex ion z-ANL formed from the radical z^·-ion species and 5,1-ANL matrix radical.

The susceptibility of amino acids to MALDI-ISD with respect to the formation of c′-ions and the covalent complex ion z-B can be rationalized by considering the following processes:

(1)

(2)

(3)

(4)

where B, [B–H]^· and H^· represent matrix molecule, matrix radical and hydrogen radical, respectively. The species M>C=O represents carbonyl oxygen on the backbone of a protein molecule M. The prompt cleavage of the resulting radical species M–C^·–OH occurs leading to the formation of c′- and z^·-ions. The resulting radical z^·-ions bind to matrix radical [B–H]^·, and follows the formation of the covalent z-B complex ions, a process which may occur simultaneously with ISD, through the interactions between the backbone amides and matrix molecules in the plume. It may be assumed therefore that the rate-determining step for the MALDI-ISD is the hydrogen attachment reaction (2). For the preferential cleavage at the N–C_α bond of Xxx–Asp/Asn and Gly–Xxx residues to generate c′- and z′-ions, the hydrogen radicals have to bind to the carbonyl oxygen between Xxx and Asp/Asn residues and between Gly and Xxx residues, respectively. This indicates that carbonyl oxygens in the Xxx–Asp/Asn and Gly–Xxx residues are in the flexible regions free from intra-molecular hydrogen-bonded secondary structures such as helix and sheet, and are exposed to matrix molecules. The formation of the complex ions z-B originating from cleavage at the Xxx–Asp/Asn and Gly–Xxx residues suggests that the backbone amide regions, –C_α–CO–NH–C_α–, in those residues susceptible to MALDI-ISD are able to interact with matrix molecules. The kinetics of the MALDI-ISD processes may be affected by the backbone flexibility in protein, because the intra-molecular hydrogen bonding in the helix and sheet protects the backbone carbonyl oxygen from the attachment of hydrogen radicals. It should be noted that there are criteria for which amino acids are preferred in the secondary structures of protein,^32,33) i.e., Asp, Asn, Gly, Pro, and Ser residues are preferred in the flexible secondary structure of turn rather than in helix and sheet. The amino acid residues Glu, Ala, Leu, Met, Gln, and Lys are preferred in helix, while Pro, Gly, Ser, Cys, and Tyr residues tend to destroy the formation of helix. Amino acids more susceptible to MALDI-ISD, i.e., Xxx–Asp/Asn, and Gly–Xxx residues, as ascertained here are consistent with the criteria described above. This indicates that carbonyl oxygens of the Asp, Asn and Gly residues are exposed to matrix molecules and are able to bind to hydrogen radicals to form the transient hypervalent radical species (Scheme 1).

Flexible Asp/Asn residues are comparable with those in protection factors estimated by hydrogen/deuterium exchange with NMR

Milne et al. reported the protection factors log P for a large number amide hydrogens in equine cytochrome c by measuring the rate of HDX with NMR.³⁰⁾ Relatively small or large values of the protection factor for an amino acid residue mean that the corresponding backbone amides are either in a flexible or rigid state, respectively. According to the HDX study, the relatively flexible amino acid residues (log P values) were Asn31 (3.1), Thr40 (3.6), Gln42 (3.6), Ala43 (3.2), Thr49 (2.7), Asn52 (3.7), Lys53 (3.3), Asn54 (3.5), Ile57 (3.1), Lys73 (3.4), Thr78 (3), and Met80 (2.8). Here we have plotted the values of the protection factor for each amino acid residue of equine cytochrome c (blue line) on a graph together with the susceptibility of amino acids to MALDI-ISD (red line), as shown in Fig. 2. Relatively small values (smaller than 4.0) of the protection factor mean that the rate of amide hydrogen exchange of the flexible amino acid residues is faster than that in other residues. This indicates that amide hydrogen in the flexible residues, –NH–C_α–, is exposed to solvent molecules. Interestingly, among the amino acid residues having small protection factors described above, Asn31, Thr49, Asn52, Lys53, Asn54, and Met80 are consistent with the backbone regions of amino acid residues identified as being susceptible to MALDI-ISD, namely Pro30–Asn31, Thr49–Asp50, Ala51–Asn52, Lys53–Asn54, Lys53–Asn54, and Lys79–Met80, respectively. However, the site of amide hydrogen, –NH–C_α–, in the residues with the small values of the protection factor was not necessarily compatible with the site of carbonyl oxygen, –C_α–CO–, in the susceptible residues to ISD. In the case of Asn31 and Pro30–Asn31 (c₃₀), for instance, the flexible amide hydrogen in HDX is in the N-terminal side of Asn31 and susceptible carbonyl carbon to ISD is in the C-terminal side of Pro30 (Scheme 3a), while in the case of Thr49 and Thr49–Asp50 (c₄₉), the flexible amide hydrogen and susceptible carbonyl carbon are in the same Thr49 residue (Scheme 3b). Further, it should be mentioned that the Glu69–Asn70 residues which give the highest intensity value of c₆₉ ion did not have a small value of the protection factor. The incompatibilities in the sites of flexible or susceptible residues between HDX and ISD described above may be due to that MALDI-ISD experiment is performed in crystal phase, while the HDX is in solution phase. And also the presence of matrix in MALDI-ISD and solvent in HDX with NMR may affect protein conformation such as secondary and tertiary structures.

Scheme 3. The sites of amide hydrogen, –NH–C_α–, in the residues with the small values of the protection factor and the sites of carbonyl oxygen, –C_α–CO–, in the susceptible residues to MALDI-ISD.

In spite of several incompatibilities described above, it is noteworthy that amino acid residues susceptible to MALDI-ISD, especially Xxx–Asp/Asn, are comparable with the flexible residues by means of HDX with NMR. Although MALDI-ISD has been reported as a tool in a top-down proteomics strategy, it would seem to be useful for studying the flexibility or susceptibility of the backbone amides and amino acid residues in proteins.

Conclusion

In this paper, I have demonstrated that in MALDI-ISD experiments with equine cytochrome c especially Xxx–Asp/Asn residues identified as being susceptible to MALDI-ISD are consistent with the residues having flexible backbone amides as identified by having faster HDX than the rest, regardless of the HDX differs from ISD in the sites of interaction with hydrogens. The N–C_α bond of Xxx–Asp/Asn and Gly–Xxx residues susceptible to MALDI-ISD gave not only intense N-terminal side c′-ions, but also C-terminal side z′-ions. The observation of the complement pair of c′-/z′-ions may be useful for identifying proteins and could become a top-down strategy. Furthermore, z′-ions were observed by accompanying covalent complex z-ANL ions with 5,1-ANL matrix radical species, which can give C-terminal side information. The present study implies that MALDI-ISD can give information about backbone flexibility of proteins, comparable with the protection factors estimated by HDX, as well as N-/C-terminal information.

Dedicated to the late Dr. Hisashi Matsuda of Prof. Emeritus Osaka University