Studying metal integration in native and recombinant copper proteins by hyphenated ICP-DRC-MS and ESI-TOF-MS capabilities and limitations of the complementary techniques

Abstract

The complementary use of LC-ESI-MS and LC-ICP-MS for characterization of native and recombinant copper proteins (molar mass range 10–20 kDa) was investigated. SEC and IC separation protocols were implemented for hyphenated ICP-MS analysis. The studies showed that validation of the methods addressing metal integration on a quantitative basis via metal to sulfur ratios demanded complementary determinations of molar mass. Reversed phase LC-ESI-TOF-MS measurements showed point mutation for the investigated recombinant apo-plastocyanin. Moreover, both recombinant proteins, i.e. plastocyanin and the Cu_A domain of cytochrome c oxidase from the cyanobacterium Synechocystis, lost their N-terminal amino acid upon expression. Since in both cases methionine formed the N-terminus the theoretical metal to sulfur ratio of the protein was changed. Excellent precision ranging at 3 ppm (N = 5) could be achieved for the determination of multiply charged ion patterns by LC-ESI-TOF-MS. The precision of the molar mass determination after deconvolution ranged at 5–15 ppm (N = 5). The intact metal containing copper proteins were measured by flow injection-ESI-TOF-MS under non-denaturing conditions (pH 5). Mass accuracy of ESI-TOF-MS allowed confirming not only stoichiometry of metal ligation but also the oxidation state of the metal center in plastocyanin and the Cu_A domain of cytochrome c oxidase. Moreover, IC-ICP-MS measurements on isotopically enriched Cu proteins were accomplished showing the potential of hyphenated ICP-MS analysis in future tracer studies.

---

Introduction

The potential of atomic spectroscopy in inorganic biochemistry is currently being propagated by researchers involved in speciation analysis.^1–4 Speciation is defined as the distribution of an element amongst defined chemical species in a system.⁵ In case of biological applications the ultimate goal of speciation analysis becomes comprehensive mapping of metal—biomolecule association.⁶ The key advance of inorganic mass spectrometry in the analysis of metal–biomolecule interaction relies in the fact that quantitative results can be obtained with species unspecific standards. The determination of metal to sulfur ratios can be exploited to assess stoichiometry of metal binding in biomolecules, since the sulfur containing amino acids exhibit a natural abundance of 4%. This approach was pioneered by Prange et al., investigating metallothioneins by CE-ICP-MS.^7–9 Quantification was carried out by on-line isotope dilution. Stoichiometry of metal binding could be assessed in commercially available metallothionein standards with known amino acid sequence. Application in real samples was complex due to the high number of unknown metallothionein sequences.

Inorganic biochemistry investigates the chemical reactivity of metal ions in biological environments.¹⁰ Proteins that must bind metal ions for their function constitute a significant share of the proteome of any living organism. A metalloprotein may need to incorporate a metal ion or a metal containing cofactor because it is involved in the catalytic mechanism and/or because it is important to determine the protein tertiary or quarternary structure.¹¹ Recently, Haraguchi and Matsura introduced the term metallomics to denote metal-assisted function biochemistry.¹² The metallomic information refers to identities of the individual metal species and their concentrations and can therefore be considered as a subset (referring to cellular biochemistry) of speciation analysis.¹³ As a matter of fact, proteins constitute the most prominent biological ligands of metal ions. Accordingly, it was attempted to transfer methodological approaches stemming from proteomics combining gel electrophoresis with laser ablation ICP-MS for detection of metals.^14–17 As a drawback data validation of these approaches is difficult. Alternative analytical strategies for investigation of metalloproteome were presented by Smith et al.¹⁸ In this study numerous components of the human hepatic copper proteome were identified using copper(II)-immobilized metal affinity chromatography, two-dimensional gel electrophoresis and subsequent identification by MALDI QqTOF-MS.

In contrast to the comprehensive metallomics approach, bioinorganic studies mostly are centered on the characterization of selected metalloproteins. In the postgenomic era research addresses structure-function relationships of the innumerable translation products (proteins, enzymes) found in various organisms. The differential effects of the transition metals and many non-metals through binding to proteins are studied. The analytical method spectrum includes mainly spectroscopic techniques (flow-flash and stopped flow kinetic spectroscopy,¹⁹ paramagnetic spectroscopy including MCD and EPR,^20,21 high field NMR^22–24 and X-ray crystallography^24,25) and organic mass spectrometry^1,26–28. Very few studies applied ICP-MS in this field.^29–35 Based on the fact that many spectroscopic techniques need protein concentrations in the micromolar to millimolar range, it is also often necessary to establish a method for the production of the corresponding recombinant proteins as well as reconstitution protocols for metal integration.

Recently, SEC-ICP-MS was introduced to study metal integration in biotechnologically produced metalloproteins supporting the development of metalloprotein expression procedures.³⁶ The fundamental experiments clearly indicated the complexity of measuring metal–protein association in biological samples. In the case of catalase peroxidase for example, SEC-ICP-MS could not separate unspecifically bound Fe (unspecific binding of hemin to protein surface was assumed).

These basic studies formed in our opinion an important starting point for future “metallomics” research. Accordingly, in the present work investigations will be extended to biotechnologically produced Cu proteins. The majority of copper proteins fulfil one of two functions, namely electron transfer or binding and activation of small molecules.³⁷ Four copper proteins were selected that differed both in copper ligation and protein size. The capabilities and limitations for the analysis of recombinant proteins by both techniques, hyphenated ICP-DRC-MS and ESI-TOF-MS, an already established technique²⁶ for metalloprotein analysis will be addressed. Moreover, fundamental experiments addressed the use of using isotopically enriched ⁶³Cu proteins for species specific isotope dilution exemplifying the key advantage of ICP-MS, i.e. the quantitative aspect of the method. Recently, it was reported for the first time that it was feasible to produce metalloproteins containing enriched metal isotopes.³⁸ The stability of the isotopically enriched copper containing rusticyanin under SEC-ICP-MS conditions was studied. In our study isotopically enriched plastocyanin will be investigated under IC conditions. The potential of isotope dilution for quantitative analysis of endogenous trace elemental species in biological systemsis is evident.³⁹ However, today only a few applications of species specific isotope dilution (IDMS)^40,41 have been published in the field of bio-inorganic speciation analysis, which can be regarded as the starting points of future developments.

Experimental

LC-ESI-TOF-MS

A cap-LC-MSD TOF system (Agilent Technologies, Palo Alto, CA, United States) was used for determination of the molecular weight of the proteins. The Poroshell SB C8 separation column (0.5 mm × 75 mm, Agilent Technologies, Palo Alto, CA, United States) was employed (column temperature 75 °C). The reversed phase separation was accomplished using 5% HCOOH in water (A) and 5% HCOOH in acetonitrile (B). Initial conditions were 10% B going after 1 min to 50% B within 5 min. After 1 min 50% B initial conditions were re-established. The eluent flow rate was 250 μL min⁻¹. A sample volume of 1 μL was injected.

ICP-MS

The ICP-MS used in this work was a quadrupole-based system equipped with a dynamic reaction cell (ELAN DRC-II, PE SCIEX, Ontario, Canada). Oxygen (purity 4.5, Linde Gas GmbH, Vienna, Austria) was used as reaction gas. The ICP-MS operation parameters are given in Table 1.

Table 1 (a) ICP-MS operation parameter. (b) ESI-TOF-MS operation parameter

	ELAN DRC II
(a)
Nebulizer	PFA
Spray chamber	Cyclon
Nebulizer gas flow	1.0 L min^−1
Auxiliary gas flow	1.275 L min^−1
Plasma gas flow	15 L min^−1
ICP RF power	1075 W
Ion lens voltage	6.35 V
O2 flow rate (as Ar)	0.6 mL min^−1
RPQ	0.3 (SO), 0.6 (Cu)
Axial field voltage	250 V
Measured isotopes	^32S, ^16O, ^65Cu
Scan mode	Peak hopping
Dwell time per isotope	50 ms
(b)
Ion polarity	Positive
Fragmentor voltage	230 V
Skimmer voltage	60 V
OCT RF voltage	250 V
Gas temperature	350 °C
Drying gas (nitrogen)	10.0 L min^−1
Nebulizer gas pressure	20 psi g
Capillary voltage	4000 V
Transients scan^−1	10 058
Min m/z range	500
Max m/z range	3500

---

SEC-ICP-MS/IC-ICP-MS

The metal-free chromatographic system consisted of an AS 50 autosampler (including a custom made temperature control device), a GP 40 gradient pump, an AD 20 UV-Vis detector and the Chromeleon Chromatography Management System (Version 6.40), all from Dionex (Sunnyvale, CA, USA). All injected samples were filtered in-line using a 0.45 μm PEEK filter located in front of the column.

A 350 × 2.0 mm PEEK column, which was packed with Sephadex G25 superfine (Sigma-Aldrich Chemie GmbH, Vienna, Austria) was used for size exclusion chromatography. The SEC eluent consisted of 150 mM NaCl and 20 mM Tris-HCl, pH 7.4. The chromatographic system was operated at a flow rate of 300 μL min⁻¹. The injection volume was 20 μL. The separation column was kept at 20.0 °C throughout the measurements.

The anion-exchange column was a TSK-DEAE-NPR column (particle diameter 2.5 μm, TOSOH Bioscience GmbH, Stuttgart, Germany) made of PEEK (30 × 2 mm, Grom Chromatography GmbH, Rottenburg-Hailfingen, Germany). The injection volume was 5 μL. The column temperature was 20 °C. The mobile phase was delivered by the gradient pump at a flow rate of 350 μL min⁻¹ and consisted of 20 mM Tris, at pH 8.5 (eluent A), and 1 M NaCl + 20 mM Tris, pH 8.5 (eluent B). Initial composition was 100% eluent A, after 1 min a linear gradient elution ran to 50% B within 10 min. The column effluents were directly coupled to the ICP-MS introduction system, consisting of a PFA-nebulizer and a cyclonic spray chamber.

Metalloprotein samples

Native superoxide dismutase (purity 95–100%, source: bovine) and azurin (purity 95–100%, from Pseudomonas aeruginosa) and were purchased at Sigma Aldrich GmbH, Vienna, Austria.

Purification and reconstitution of the soluble domain of subunit II from Synechocystis PCC 6803 (i.e. Cu_A, a binuclear copper protein) cytochrome c oxidase has been described elsewhere.³⁴ Shortly, the protein obtained from lysis of E. coli was loaded on a Q Sepharose Fast Flow column and fractions with the protein of interest were pooled and the Cu_A site was reconstituted by addition of Cu(His)₂. Further purification of the holoprotein included DEAE Sepharose Fast Flow and Superdex 75 HR FPLC. The theoretical mass of the apoprotein (which has lost the N-terminal methionine, 131 Da) is 20 984.8 Da. Adding copper (2 × 63.5 Da) gives a theoretical mass of the holoprotein of 21 111.8 Da.

Cloning, heterologous overexpression, purification and reconstitution of the type-1 copper protein plastocyanin from Synechocystis have been described recently.⁴² In the present study the variant Lys56Arg was investigated, which has a theoretical mass of the apoprotein of 10 280.5 Da. Adding copper (63.5) gives a theoretical mass of the holoprotein of 10 344 Da. Its purification procedure from E. coli included Q Sepharose Fast Flow, reconstitution of pooled fractions with Cu(His)₂ and further purification by Phenyl Sepharose Fast Flow chromatography. The production of isotopically enriched plastocyanin was accomplished by using a copper standard enriched in ⁶³Cu (99.9 ± 0.15%) purchased at Chemotrade, Düsseldorf, Germany.

Chemicals and standards

Sulfur and Cu ICP-MS standard solutions for preparation of calibration standards were purchased at VWR. The accuracy of the Cu and Zn calibrations was assessed by measurement of a standard reference material (TM-27.2, National Water Research Institute, Canada).

Eluents were prepared from sodium hydroxide (suprapure, Merck, Darmstadt, Germany), hydrochloric acid (p.a., Merck; further purified in a quartz sub-boiling system from Milestone-MLS GmbH, Leutkirch, Germany), tris(hydroxymethyl)aminomethane (p.a., Merck) and water purified in a water purification system (>18 MΩ cm resistance; HQ, USF, Vienna, Austria).

Uncertainty of measurement

In this work, reference is made to the terminology: accuracy of measurement, standard uncertainty, expanded uncertainty and coverage factor according to the Guide to the Expression of Uncertainty in Measurement.⁴³ All combined uncertainties were calculated according to the ISO/GUM guide using the uncertainty propagation procedure.⁴⁴ Dedicated software (GUM-workbench software, Metrodata GmbH, Grenzach-Wyhlen, Germany) was employed for the calculations based on the numerical method of differentiation.⁴⁵ Processing of chromatographic data was carried out using Chromeleon (Version 6.40., Dionex Corp., Sunnyvale, CA, USA).

Results and discussion

Four copper-containing metalloproteins differing in metal ligation were investigated in this study. Based on the fact that the majority of electron transfer copper centers are found around the type 1 or “blue” site, whose structure is derived from a trigonal pyramidal motif [Cu(Cys)(His)₂(L)]^0/+1 with L being methionine or a main chain carbonyl group, in the present work two monomeric type-1 copper proteins have been selected. We studied plastocyanin in the recombinant form from the cyanobacterium Synechocystis, which functions as electron carrier in cyanobacterial and plant oxygenic photosynthesis and azurin from Pseudomonas Aeruginosa, which is involved in electron transport of Gram-negative bacteria such as Pseudomonas, methylotrophs and Alcaligenes strains. Furthermore, we selected a dicopper transfer site that is closely related to the type 1 centers, termed Cu_A, which constitutes the electron entry domain of cystochrome c oxidase (i.e. the terminal electron acceptor in the mitochondrial respiratory chain). In detail, the recombinant Cu_A domain of cytochrome c oxidase from the cyanobacterium Synechocystis was investigated. The structure of Cu_A can be thought of as two fused type-1-like sites linked by bridging cysteinate groups, to yield a [{Cu(His)(L)(μ-Cys)}2]^0/+2 complex, with L being either a methionine side chain or a main chain carbonyl group. The fourth copper protein investigated here is a representative of type-2 copper centers, which generally contain mononuclear copper ions that do not contain thiolate ligation. A variety of structures fall into this classification and these copper sites take part in substrate binding and activation. A well known representative is Cu/Zn superoxide dismutase (SOD) that catalyzes the very rapid two-step dismutation of the toxic superoxide radical. In homodimeric Cu/Zn SOD the copper ion has four histidine ligands arranged in an irregular tetrahedral distortion from square planar geometry, whereas the zinc is tetrahedrally surrounded by histidines. One of the histidine bridges the two metals in an almost planar manner, giving a Cu–Zn separation of about 6.2 Å. In the present study native bovine Cu/Zn superoxide dismutase was used as a model type-2 protein.

Determination of molar mass by RP-LC-ESI-TOF-MS

Accurate determination of molar mass allows identification of amino acid exchanges in the protein of interest, since point mutations and loss of the N-terminus are often a consequence of heterologous expression. These exchanges could affect metal ligation and number of sulfur containing amino acids characterizing the protein. In a first step, molar mass of the apoforms of the two recombinant proteins, namely plastocyanin and Cu_A domain of cytochrome c oxidase was assessed by RP-LC-ESI-TOF-MS. Moreover, the RP-LC chromatograms obtained for the investigated Cu proteins proved the absence of other protein impurities. The ESI-TOF-MS instrument used in this study offered a mass accuracy of 3 ppm. The molar mass of the investigated plastocyanin from Synechocystis deviated from the theoretical value (see Table 2).⁴⁶ A molar mass of 10 280.38 was obtained after deconvolution (see Fig. 1c). An excellent precision of 4 ppm was achieved for N = 5 consecutive determinations using the reference mass option of the instrument. In combination with DNA sequencing the variant Lys56Arg having removed the N-terminal methionine was identified. Metal ligation was not affected by this mutation. Moreover, as can be readily observed in Table 2, for the recombinant Cu_A domain the experimentally obtained molar mass was consistent with the calculated mass again without N-terminal methionine. Hence, both proteins lost the N-terminal methionine upon expression changing the theoretical S/Cu ratio from 4 to 3, and 7 to 6.5 for plastocyain and CuA domain of cytochrome c oxidase, respectively.

Fig. 1 (a) Separation of plastocyanin by LC-ESI-MS (TIC, 10 pmol plastocyanin) (b) multiply charged ion pattern obtained for plastocyanin and (c) deconvoluted mass corresponding to the apo-protein. A relative molecular weight (MH⁺) of 10 281.38 was determined with a precision 4 ppm (n = 5).

Table 2 Molar mass determination of the two recombinant proteins by ESI-TOF-MS. (m/z determination of multiply charged ions with precision of 3 ppm)

	Predicted theoretical mass^46	Deconvoluted mass (M-H^+) LC-ESI-TOF-MS	Deconvoluted mass (M-H^+) FI-ESI-TOF-MS	Apparent mass difference metalloproteins − apoprotein	Metal ligation	Oxidation state of metal center
Plastocyanin recombinant from E. coli	10 383.60
Plastocyanin recombinant from E. coli	10 280.48	10 281.38	10 342.64	61.26	1 Cu	2
Variant Lys56Arg
Cytochrome c oxidase from Synechocystis	20 984.89	20 985.65	21 109.73	124.08 (corresponding to 62.04/Cu)	2 Cu	1.5

---

This characterization of the investigated protein set is a prerequisite for the validation of hyphenated ICP-MS analysis relying on the knowledge of theoretical sulfur/metal ratios.

Determination of stoichiometry of copper ligation/degree of metal integration by SEC-ICP-MS/IC-ICP-MS

Next, S/Cu ratios were assessed by SEC-ICP-MS. Fig. 2 shows the SEC-ICP-MS chromatogram of recombinant plastocyanin. The protein elutes at 2 minutes with the implemented experimental set-up. As can be seen separation is an absolute prerequisite since inorganic impurities obfuscate otherwise the accuracy of S/Cu determination. Table 3 summarizes the results obtained by species unspecific calibration. Correction factors considering the different sensitivity of Cu and S detection in order to assess the correct molar ratio were introduced via flow injection of inorganic Cu and S standards in the form of Cu²⁺ and SO₄²⁻. The accuracy of species unspecific calibration has been shown elsewhere.³⁶ The S/Cu intensity ratios were determined with standard uncertainties (precision of N = 4 determinations) ranging at 0.8–1.5%. Hence, for this experimental set-up the S/Cu ratio determination was associated with a total combined uncertainty ranging at 5% calculated according to Eurachem guidelines. In the case of native proteins azurin and Cu/Zn SOD the determined S/Cu agreed with the theoretical value within the total combined uncertainty. Both recombinant proteins were expressed in the apoform and metal addition occurred in an extra-reconstitution step. The Cu_A domain of cytochrome c oxidase showed a S/Cu ratio higher than the predicted theoretical value indicating unspecific copper binding. In this case a significant fraction of offered Cu²⁺ was bound not to the specific binding sites, but adsorbed unspecifically to the protein surface and could not be separated by SEC-ICP-MS.

Fig. 2 Rapid separation of isolated recombinant plastocyanin from impurities by (above) size exclusion chromatography and (below) ion chromatography (detection by ICP-DRC-MS; sulfur was measured as ³²S¹⁶O using oxygen as reaction gas).

Table 3 Determination of stoichiometric S/Cu ratios by SEC-ICP-MS and IC-ICP-MS. The ratios are given with standard uncertainties (precision of N = 4 determinations)

Protein	Purity	Sulfur containing amino acids per molecule	S/Cu ratio theoretical	S/Cu ratio theoretical heterologous expression	S/Cu ratio measured by SEC-ICP-MS	S/Cu ratio measured by IC-ICP-MS calibration by azurin
Cu Zn superoxide dismutase bovine	>99	4	4		3.99 ± 0.02	8.41 ± 0.16
Azurin from Pseudomonas aeruginosa	>95	9	9		9.20 ± 0.13
Plastocyanin recombinant from E. coli	>80	4	4	3	2.83 ± 0.02	2.87 ± 0.20
Cytochrome c oxidase from Synechocystis	>80	14	7	6.5	4.29 ± 0.01	6.20 ± 0.30

---

In the next step, IC-ICP-MS was evaluated for studying metal integration in these metalloproteins. Since all investigated metallo-proteins were acidic proteins,^34,42 anion exchange was selected (see Fig. 2b). A non-porous DEAE functionalized stationary phase with 2.1 μm particle diameter allowed fast protein separation at pH 8.5. Both recombinant proteins, i. e. plastocyanin and Cu_A domain were investigated using azurin as calibration standard. The uncertainty of S/Cu ratio determination was assessed: as the retention times of plastocyanin (4.9 min) and azurin (4.8 min) coincided, calibration was straightforward with a total combined uncertainty of 8%. In the case of the Cu_A domain (retention time 7 min), gradient elution affected the calibration of the S/Cu ratio. Hence, flow injection analysis of an inorganic standard containing 1000 ng g⁻¹ S and 100 ng g⁻¹ Cu was performed using two different concentrations of NaCl in the carrier flow corresponding to the eluent composition at 5 and 7 min. The experiment represented the conditions during elution of azurin, plastocyanin and Cu_A domain, respectively. The signal intensity ratios ⁴⁸SO/⁶⁵Cu obtained for the two different salt concentrations agreed within their uncertainty of 2%. Hence a correction factor of K = 1 was assumed and associated with a standard uncertainty of 2% for calibration of Cu_A domain via azurin. Finally, a total combined uncertainty of 11% (coverage factor: 2) was assessed for species unspecific Cu/S ratio quantification of the Cu_A domain. The standard uncertainties of the measured intensity ratios and the influence of the changing NaCl concentrations during gradient elution represented the major contribution to the total combined uncertainty. As can be seen in Table 3, IC-ICP-MS was successfully applied to determine the degree of specifically bound Cu in the two recombinant proteins. Both determined S/Cu ratios agreed with the theoretically predicted value. Fig. 2b shows the IC-ICP-MS analysis of plastocyanin. The broad Cu peak (1–4 min retention time) could be due to inorganic Cu impurities or unspecifically protein bound Cu as a result of Cu addition in an extra step in the heterologous expression procedure. At this point, it has to be stressed that this method was only applicable provided alternative analytical methods confirmed the stability of metal–protein binding at pH 8.5. Spectroscopic studies (UV-Vis and circular dichroism spectroscopy) clearly demonstrated that plastocyanin, azurin and the Cu_A protein are stable at pH 8.5 and do not loose the coordinated copper ion(s), whereas in case of Cu/Zn superoxide dismutase loss of copper was observed due to beginning of protein unfolding.⁴⁷ Accordingly, the S/Cu ratio found by IC-ICP-MS was > than the theoretical target value (8.4 versus 4).

Table 4 compares the limit of detection for both proteins obtained for IC and SEC-ICP-MS, respectively. The LODs were based on three times the standard deviation of the baseline signal quantified by peak height calibration.

Table 4 Limits of detection and for comparison absolute protein amount investigated by SEC/IC-ICP-MS and LC-ESI-TOF-MS. The LODs were based on three times the standard deviation of the baseline signal quantified by peak height calibration. A 5 nmol mL⁻¹ plastocyanin solution was used in SEC/IC-ICP-MS, 10 nmol mL⁻¹ plastocyanin was investigated in LC-ESI-MS. Here, the most prominent multiply charged ion (i.e.m/z 935.685 was extracted as peak (extraction range 1.5 amu) for the assessment of LOD

	Injection volume/μL	Absolute amount of protein injected/pmol	LOD S/μg L^−1	LOD Cu/μg L^−1	LOD plastocyanin/μg L^−1	LOD plastocyanin/nmol L^−1	LOD plastocyanin/fmol
a Quantified considering sulfur (i.e. limiting factor).
SEC-ICP-MS	20	100	4.6	0.8	400^a	40^a	800^a
IC-ICP-MS	5	25	16	4.8	1300^a	130^a	650^a
LC-ESI-MS	1	10			50	4.7	4.7

---

Determination of metal binding stoichiometry and oxidation state of copper by ESI-TOF-MS under non-denaturing conditions

Finally, the investigated Cu proteins were introduced into the ESI source by flow injection under non-denaturing conditions (pH = 6) in order to observe intact metal containing protein ions. For plastocyanin a molar mass of 10 342.64 (M-H⁺) was assessed (see Fig. 3b). This mass was calculated from the multiply charged ion pattern assuming that all ionizable sites in the amino acid sequence are neutral and that consequently all charge of an ion is due to excess protons. However, the metals in metalloproteins as plastocyanin occur in oxidation states other than zero. Hence, the apparent mass will differ from the actual mass by the number of protons equal to the charge present on the metal center.⁴⁸ As can be seen in Fig. 3a flow injection analysis of plastocyanin resulted in only 2 multiply charged ions (+6, +5), accordingly the precision of molar mass determination decreased to 15 ppm (N = 5). For Cu_A domain a molar mass of 21 109.73 was determined with a precision of 5 ppm. Hence, the known stoichiometry of 1 and 2 Cu/protein were obtained for plastocyanin and the Cu_A domain, respectively. However, quantification of metal integration achieved by the recombinant production was impeded due to the lack of standards. Moreover, as a drawback flow injection analysis relied on the purity of protein solution. Table 2 summarizes the ESI-TOF-MS measurements with the obtained molar mass difference between apo- and holoprotein corresponding to the molar mass of 1 Cu and 2 Cu for plastocyanin and Cu_A domain, respectively. Since the deconvolution defines uniform charge carrying species (protons), comparison of apparent mass (i.e. 61.3 Da for plastocyanin, 62.0 × 2 Da for Cu_A domain) and predicted theoretical mass (63.5) reveals the charge i.e. the oxidation state of Cu.⁴⁹ In plastocanin copper is present as Cu^II.⁴² The Cu_A center of the Cu_A domain in its oxidized state is composed of two electronically coupled, mixed-valence Cu^I/Cu^II copper ions.⁵⁰ Hence, for both proteins the known Cu oxidation states could be confirmed experimentally by ESI-TOF-MS. So far studies concerning the redox status of metal centers in large biomolecules were carried out exclusively by FT-ESI-MS.

Fig. 3 Flow injection ESI-TOF-MS analysis (pH 5.5, 20 mM ammonium acetate) of plastocyanin (above) obtained multiply charged ion pattern and (below) deconvoluted mass corresponding to the metalloproteins. A relative molecular weight 10 342.64 was determined with a precision of 15 ppm (n = 5). The mass difference between the apoprotein and the holoprotein corresponding to Cu is 61.3 (and not 63.5), which is due to limitations in the deconvolution algorithm. This algorithm is limited by the fact that the assumed charge carrying species is uniform for the multiply charged ion pattern and is calculated for H⁺. Since copper is present in the protein carrying two charges the resulting molecular weight differs by 2 from the theoretical target value.

Production of isotopically enriched Cu proteins

Although ID ICP-MS has been widely employed for trace element analysis in a variety of sample matrixes, its application to species specific determinations has been limited by the non-availability of commercial species-specific enriched spikes. Application of the species-specific isotope dilution analysis to large biomolecules is a challenging task.

In our work, isotopically enriched ⁶³Cu plastocyanin was produced according to the procedure described in the experimental section. The quality of the assay was controlled by IC-ICP-MS. Fig. 4 shows the signals obtained at m/z 63 and 65 for (a) the “natural” metalloprotein, (b) the enriched ⁶³Cu metalloprotein and (c) the blend, which was prepared by mixing gravimetrically determined amounts of the two.

Fig. 4 Determination of the ⁶⁵Cu/⁶³Cu ratio in plastocyanin by IC-ICP-MS. The natural ratio is 0.4458. (Top) In this study a ratio of 0.4480 (±1.7%, N = 3) was determined in plastocyanin by dividing the peak areas obtained for ⁶⁵Cu and ⁶³Cu, respectively. (Middle) For the isotopically enriched protein a ⁶⁵Cu/⁶³Cu a ratio of 0.0077 (±2.2%, N = 3) was determined. (Bottom) A protein blend was prepared showing the feasibility of species specific IDMS. The measured ⁶⁵Cu/⁶³Cu ratio in this blend was 0.3550. Excellent precisions of 1.2% (N = 3) could be achieved.

As a matter of fact the accuracy of species specific on-line IDMS results is mainly determined by the precision of measurement of the isotopic ratio of the blend because the contribution of gravimetric measurands to the total combined uncertainty is generally low. In our study the precision of isotopic measurement ranged at 1–2% (peak ratios) allowing quantification with a total combined uncertainty of 5% (coverage factor 2).

Conclusion

Both methods, LC-ICP-MS and LC-ESI-MS have shown to be valuable tools for quality control of biotechnologically produced metalloproteins, which are the subjects of bioinorganic studies. By LC-ESI-MS, micro-heterogeneities in the primary amino acid sequence, point mutations or side products often occurring during recombinant protein expression and consequently compromising enzymatic activity and/or metal binding could be excluded. The stoichiometry of metal binding and even oxidation state of the metal center was confirmed by FIA-ESI-MS. Moreover, the measurements showed in the case of the two studied recombinant proteins, the removal of the N-terminal methionine upon heterologous expression, resulting in an altered theoretical metal/sulfur ratio. Accordingly, determination of molar mass was a prerequisite for studying metal integration by LC-ICP-MS based on metal/sulfur ratios. The potential of ICP-MS relied in the fact that the yield of metal integration achieved in the process of heterologous protein expression could be quantified (knowing amino acid sequence and stoichiometries of metal binding). Again this parameter has a significant impact on the studied enzymatic activity of the metalloproteins and has to be taken into account in bioinorganic studies related to structure-function relationship. Moreover, preliminary studies using isotopically enriched metalloproteins standards showed the feasibility of species specific isotope dilution analysis of metallobiomolecules.

Acknowledgements

Andrea Zenker from Agilent Technologies Basel, Switzerland, Rainer Schumacher and Franz Berthiller (IFA-Tulln, BOKU-Vienna) are highly acknowledged for useful discussions.

References

1. R. Lobinski, D. Schaumloffel and J. Szpunar, Mass Spectrom. Rev., 2006, 25, 255.
2. N. Jakubowski, R. Lobinski and L. Moens, J. Anal. At. Spectrom., 2004, 19, 1.
3. A. Hagège, T. Baldinger, M. Martin-Jouet, F. Zal, M. Leroy, E. Leize and A. Van Dorsselaer, Rapid Commun. Mass Spectrom., 2004, 18, 735.
4. J. Szpunar and R. Lobinski, Anal. Bioanal. Chem., 2002, 373, 404.
5. A. D. Jenkins, in IUPAC Compendium of Chemical Terminology, IUPAC, Research Triangle Park, NC 27709, USA, 2004, vol. 76, p. 1070.
6. J. Szpunar, R. Lobinski and A. Prange, Appl. Spectrosc., 2003, 57, 102a.
7. D. Schaumloeffel, A. Prange, G. Marx, K. G. Heumann and P. Bratter, Anal. Bioanal. Chem., 2002, 372, 155–163.
8. A. Prange and D. Schaumloffel, Anal. Bioanal. Chem., 2002, 373, 441.
9. D. Profröck, P. Leonhard and A. Prange, Anal. Bioanal. Chem., 2003, 377, 132.
10. J. A. Cowan, Inorganic biochemistry—An introduction, VCH Publishers Inc., New York, 1993.
11. J. J. R. Frausto da Silva and R. J. P. Williams, The Biological Chemistry of the Elements, The Inorganic Chemistry of Life, Clarendon Press, Oxford, 1993.
12. H. Haraguchi and H. Matsura, in Bitrel, Wako (Saitama), ed. S. Enomoto, Fujiyoshida (Yamanashi), Japan, 2003.
13. J. Szpunar, Anal. Bioanal. Chem., 2003, 378, 54.
14. R. L. Ma, C. W. McLeod, K. Tomlinson and R. K. Poole, Electrophoresis, 2004, 25, 2469.
15. J. S. Becker, M. Zoriy, C. Pickhardt, M. Przybylski and J. S. Becker, Int. J. Mass Spectrom., 2005, 242, 135.
16. J. S. Becker, M. Zoriy, U. Krause-Buchholz, J. S. Becker, C. Pickhardt, M. Przybylski, W. Pompe and G. Rodel, J. Anal. At. Spectrom., 2004, 19, 1236.
17. S. Becker, M. Zoriy, J. S. Becker, C. Pickhardt and M. Przybylski, J. Anal. At. Spectrom., 2004, 19, 149.
18. S. D. Smith, Y. M. She, E. A. Roberts and B. Sarkar, J. Proteome Res., 2004, 3, 834.
19. A. L. De Lacey, A. Pardo, V. M. Fernandez, S. Dementin, G. Adryanczyk-Perrier and E. C. Hatchikian, M. Rousset. J. Biol. Inorg. Chem., 2004, 9, 636.
20. M. R. Cheesman, V. S. Oganesyan, N. J. Watmough, C. S. Butler and A. J. Thomson, J. Am. Chem. Soc., 2004, 126, 4157.
21. W. R. Hagen, Coord. Chem. Rev., 1999, 192, 209.
22. T. E. Machonkin, W. M. Westler and J. L. Markley, Inorg. Chem., 2005, 44, 779.
23. M. R. Jensen, D. F. Hansen, U. Ayna, R. Dagil, M. A. S. Hass, H. E. M. Christensen and J. J. Led, Magn. Reson. Chem., 2006, 44, 294.
24. L. Banci, I. Bertini and S. Mangani, J. Synchroton Radiat., 2005, 12, 94.
25. C. M. Wilmot and A. R. Pearson, Curr. Opin. Chem. Biol., 2002, 6, 202.
26. J. A. Loo, Mass Spectrom. Rev., 1997, 16, 1.
27. T. D. Veenstra, K. L. Johnson, A. J. Tomlinson, T. A. Craig, R. Kumar and S. Naylor, J. Am. Soc. Mass Spectrom., 1998, 9, 8.
28. P. F. Hu and J. A. Loo, J. Mass Spectrom., 1995, 30, 1076.
29. M. E. D. Busto, M. Montes-Bayon, E. Blanco-Gonzalez, J. Meija and A. Sanz-Medel, Anal. Chem., 2005, 77, 5615.
30. M. H. Nagaoka and T. Maitani, J. Bioorg. Biochem., 2005, 99(9), 1887.
31. M. H. Nagaoka, H. Akiyama and T. Maitani, Analyst, 2004, 129, 51.
32. M. H. Nagaoka and T. Maitani, Biochim. Biophys. Acta, 2001, 1526, 175.
33. M. H. Nagaoka and T. Maitani, Analyst, 2000, 125, 1962.
34. M. Paumann, B. Lubura, G. Regelsberger, M. Feichtinger, G. Köllensperger, C. Jakopitsch, P. G. Furtmüller, G. A. Peschek and C. Obinger, J. Biol. Chem., 2004, 279, 10293.
35. M. H. Nagaoka, T. Yamazaki and T. Maitani, Biochem. Biophys. Res. Commun., 2002, 296, 1207.
36. S. Hann, G. Koellensperger, Ch. Obinger, P. G. Furtmueller and G. Stingeder, J. Anal. At. Spectrom., 2004, 19, 74.
37. M. A. Halcrow, P. F. Knowles and S. E. V. Phillips, in Handbook of Metalloproteins, ed. A. Sigel, H. Sigel and I. Bertini, Marcel Dekker Inc., New York, 2001, p. 709.
38. C. F. Harrington, D. S. Vidler, M. J. Watts and J. F. Hall, Anal. Chem., 2005, 77, 4034.
39. D. Schaumlöffel and R. Lobinski, Int. J. Mass Spectrom., 2005, 242, 217.
40. J. R. Encinar, D. Schaumloffel, Y. Ogra and R. Lobinski, Anal. Chem., 2004, 76, 6635.
41. S. McSheeny, L. Yang, R. Sturgeon and Z. Mester, Anal. Chem., 2005, 77, 344.
42. M. Paumann, M. Bernroitner, B. Lunra, M. Peer, C. Jakopitsch, P. G. Furtmüller, G. A. Peschek and C. Obinger, FEMS Microbiol. Lett., 2004, 239, 301.
43. EURACHEM, Citac Guide CG4, Quantifying Uncertainty in Analytical Measurement, 2nd edn, ed. S. L. R. Ellison, M. Rosslein and A. Williams, London, 2000.
44. Guide to the Expression of Uncertainty in Measurement, International Organization for Standardization, ISO/GUM, Geneva, Switzerland, 1995, ISBN 92-67-10188-9.
45. J. Kragten, Analyst, 1994, 119, 2161.
46. E. Gasteiger, C. Hoogland, A. Gattiker, S. Duvaud, M. R. Wilkins, R. D. Appel and A. Bairoch, in The Proteomics Protocols Handbook, ed. J. M. Walker, Humana Press, Totowa, NJ, USA, 2005, p. 571.
47. K. M. Beem, W. E. Rich and K. V. Rajagoplan, J. Biol. Chem., 2002, 249, 7298.
48. K. A. Johnson, M. Verhagen, P. S. Brereton, M. W. W. Adams and J. I. Amster, Anal. Chem., 2000, 72, 1410.
49. K. A. Johnson, B. A. Shira, J. L. Anderson and I. J. Amster, Anal. Chem., 2001, 73, 803.
50. F. Malatesta, F. Nicoletti, V. Zickermann, B. Ludwig and M. Brunori, FEBS Lett., 1998, 434, 322.

---