qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation

5
JOURNAL OF MASS SPECTROMETRY J. Mass Spectrom. 2002; 37: 755–759 Published online 3 July 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.341 Qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation Carlos Afonso, Yetrib Hathout and Catherine Fenselau Chemistry and Biochemistry Department, University of Maryland, College Park, Maryland 20742, USA Received 11 February 2002; Accepted 29 March 2002 Nanospray and collisionally induced dissociation (CID) on a quadrupole/time-of-flight mass spectrometer were used to examine the complexes formed between the zinc ion binding protein metallothionein and a series of peptides related to glutathione. The objective of the study was to determine if CID could be used to distinguish complexes that are stabilized by co-chelation of a zinc ion from non-covalent complexes that were formed in some other way. Differences in the collision energy required for dissociation and, more importantly, differences in the distribution of zinc ions between the pairs of dissociation products suggest that mass spectrometry can provide qualitative information about the bimolecular chelation of metal ions. The potential application to zinc chelates is particularly important, since biological chelates do not provide signals directly detectable by NMR, M ¨ ossbauer or other spectroscopies. The observations reported here also allowed a molecular mechanism to be proposed to explain the differences observed by others in the physiological interactions of reduced and oxidized glutathione with metallothionein. Copyright 2002 John Wiley & Sons, Ltd. KEYWORDS: zinc chelates; metallothionein; glutathione; nanospray; collisionally induced dissociation INTRODUCTION Studies through the last decade have shown that qualitative and quantitative information can be obtained by study- ing the gas-phase decomposition of singly 1 and multiply charged 2 bimolecular complexes bound by protons or metal ions. 3 Collisionally induced dissociation (CID) has also been employed to study the relative strengths of interactions in complexes comprising multiple proteins. 4,5 In this study, CID was applied to the qualitative characterization of pep- tide/protein complexes, in which both partners contribute ligands to the chelation of a shared zinc ion. The role of such complexes is under intense scrutiny by biochemists, in efforts to understand both intracellular zinc ion homeostasis 6 and also mechanisms of gene regulation. 7 These zinc ion chelates are silent to most physicochemical probes, 8 so it is impor- tant to find new techniques to study the role of zinc in biology. Mammalian metallothionein (MT) is a small protein that chelates seven zinc ions across two domains and occurs in every mammalian cell. It has been shown that MT provides zinc ions to other proteins and polypeptides, enzymes such as carbonic anhydrase 9,10 and also non- catalytic zinc fingers. 11,12 It is generally accepted that Zn 2C is transferred in transient bimolecular complexes, presumably via a series of intermediate co-chelates. 13,14 Considerable Ł Correspondence to: Catherine Fenselau, Chemistry and Biochemistry Department, University of Maryland, College Park, Maryland 20742, USA. E-mail: [email protected] Contract/grant sponsor: National Institutes of Health; Contract/grant number: (GM-21248). work remains to characterize these transfer complexes, 15 both structurally and kinetically. Oxidized glutathione (GSSG) (Table 1) has been reported to enhance the release of zinc ions from metallothionein to other proteins, 16,17 while reduced glutathione (GSH) (Table 1) is reported to inhibit Zn 2C transfer from MT to other proteins. 16 The important hypothesis has been put forward for testing that the GSSG/GSH couple links zinc transfer to the cellular redox state. 18 CID was used here to examine the complexes formed between metallothionein and reduced glutathione, oxidized glutathione and a series of glutathione analogs. In all cases adducts can be observed in nanoelectrospray mass spectra. However, differences are seen in the distribution of the putative shared zinc ions into the dissociation products, and also in the energy required to dissociate the complexes. EXPERIMENTAL Materials Rabbit liver metallothionein 2a, glutathione, glutathione disulfide, oxidized Cys-Gly, Glu-Cys, oxidized Glu- Cys-Glu, zinc atomic absorption standard solution and ammonium hydrogencarbonate were obtain from Sigma (St. Louis, MO, USA). Deionized water (18.2 M) was obtained from a MilliQ apparatus (Millipore, Bedford, MA, USA). Acetyl chloride was obtained from Aldrich (Milwaukee, WI, USA). HPLC-grade methanol and acetonitrile were obtained from Fisher Scientific (Pittsburgh, PA, USA). Copyright 2002 John Wiley & Sons, Ltd.

Upload: carlos-afonso

Post on 15-Jun-2016

215 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation

JOURNAL OF MASS SPECTROMETRYJ. Mass Spectrom. 2002; 37: 755–759Published online 3 July 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.341

Qualitative characterization of biomolecular zinccomplexes by collisionally induced dissociation

Carlos Afonso, Yetrib Hathout and Catherine Fenselau∗

Chemistry and Biochemistry Department, University of Maryland, College Park, Maryland 20742, USA

Received 11 February 2002; Accepted 29 March 2002

Nanospray and collisionally induced dissociation (CID) on a quadrupole/time-of-flight mass spectrometerwere used to examine the complexes formed between the zinc ion binding protein metallothionein and aseries of peptides related to glutathione. The objective of the study was to determine if CID could be usedto distinguish complexes that are stabilized by co-chelation of a zinc ion from non-covalent complexes thatwere formed in some other way. Differences in the collision energy required for dissociation and, moreimportantly, differences in the distribution of zinc ions between the pairs of dissociation products suggestthat mass spectrometry can provide qualitative information about the bimolecular chelation of metal ions.The potential application to zinc chelates is particularly important, since biological chelates do not providesignals directly detectable by NMR, Mossbauer or other spectroscopies. The observations reported herealso allowed a molecular mechanism to be proposed to explain the differences observed by others in thephysiological interactions of reduced and oxidized glutathione with metallothionein. Copyright 2002John Wiley & Sons, Ltd.

KEYWORDS: zinc chelates; metallothionein; glutathione; nanospray; collisionally induced dissociation

INTRODUCTION

Studies through the last decade have shown that qualitativeand quantitative information can be obtained by study-ing the gas-phase decomposition of singly1 and multiplycharged2 bimolecular complexes bound by protons or metalions.3 Collisionally induced dissociation (CID) has also beenemployed to study the relative strengths of interactions incomplexes comprising multiple proteins.4,5 In this study,CID was applied to the qualitative characterization of pep-tide/protein complexes, in which both partners contributeligands to the chelation of a shared zinc ion. The role of suchcomplexes is under intense scrutiny by biochemists, in effortsto understand both intracellular zinc ion homeostasis6 andalso mechanisms of gene regulation.7 These zinc ion chelatesare silent to most physicochemical probes,8 so it is impor-tant to find new techniques to study the role of zinc inbiology.

Mammalian metallothionein (MT) is a small protein thatchelates seven zinc ions across two domains and occursin every mammalian cell. It has been shown that MTprovides zinc ions to other proteins and polypeptides,enzymes such as carbonic anhydrase9,10 and also non-catalytic zinc fingers.11,12 It is generally accepted that Zn2C istransferred in transient bimolecular complexes, presumablyvia a series of intermediate co-chelates.13,14 Considerable

ŁCorrespondence to: Catherine Fenselau, Chemistry andBiochemistry Department, University of Maryland, College Park,Maryland 20742, USA. E-mail: [email protected]/grant sponsor: National Institutes of Health;Contract/grant number: (GM-21248).

work remains to characterize these transfer complexes,15 bothstructurally and kinetically. Oxidized glutathione (GSSG)(Table 1) has been reported to enhance the release ofzinc ions from metallothionein to other proteins,16,17 whilereduced glutathione (GSH) (Table 1) is reported to inhibitZn2C transfer from MT to other proteins.16 The importanthypothesis has been put forward for testing that theGSSG/GSH couple links zinc transfer to the cellular redoxstate.18

CID was used here to examine the complexes formedbetween metallothionein and reduced glutathione, oxidizedglutathione and a series of glutathione analogs. In allcases adducts can be observed in nanoelectrospray massspectra. However, differences are seen in the distributionof the putative shared zinc ions into the dissociationproducts, and also in the energy required to dissociate thecomplexes.

EXPERIMENTAL

MaterialsRabbit liver metallothionein 2a, glutathione, glutathionedisulfide, oxidized Cys-Gly, �Glu-Cys, oxidized �Glu-Cys-Glu, zinc atomic absorption standard solution andammonium hydrogencarbonate were obtain from Sigma (St.Louis, MO, USA). Deionized water (18.2 M�) was obtainedfrom a MilliQ apparatus (Millipore, Bedford, MA, USA).Acetyl chloride was obtained from Aldrich (Milwaukee, WI,USA). HPLC-grade methanol and acetonitrile were obtainedfrom Fisher Scientific (Pittsburgh, PA, USA).

Copyright 2002 John Wiley & Sons, Ltd.

Page 2: Qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation

756 C. Afonso, Y. Hathout and C. Fenselau

Table 1. Peptides studied

Compound Molecular formula Structure M (Da)

�ECG (GSH) reduced C10H17 N3O6S1HN

NH2

O

O

CH2

SH

NH

COOHHO

O

307.1a

307.3b

�ECG (GSSG) oxidized C20H32N6O12S2

CH2

CH2

S

HN

O

COOH

HN

NH2

O

S

NH

NH

O

O

COOHHO

O

HO

O

NH2

612.1a

612.6b

�EC oxidized C16H26N4O10S2

CH2O

CH2O

O

NH

S

NH2

HN

NH2

S

NH2

O

OH

O

NH2

O

OH

498.1a

498.5b

CG oxidized C10H18N4O6S2

CH2

S

HN

O

COOH

H2N

H2N

CH2

S

NH

O

COOH

354.1a

354.4b

�ECE oxidized C26H40N6O16S2

NH

S

HN

O

HN

NH2

CH2

CH2

S

NH

O

OH

O

O

O

COOH OH

COOH

O

OH

O

NH2

O

OH

756.2a

756.8b

�ECG (tetramethyl ester) oxidized C24H40N6O12S2

CH2

CH2

S

HN

O

COOCH3

HN

NH2

O

S

NH

NH

O

O

COOCH3

H3CO

O

H3CO

O

NH2

668.2a

668.8b

a Average mass.b Monoisotopic mass.

Copyright 2002 John Wiley & Sons, Ltd. J. Mass Spectrom. 2002; 37: 755–759

Page 3: Qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation

Biomolecular zinc complexes 757

Sample preparationRabbit liver metallothionein was purified by reversed-phase(RP) HPLC and reconstituted with zinc following a publishedprocedure.11 Excess of zinc ions was eliminated using5 kDa cutoff centrifugal ultrafiltration cartridges (Ultrafree,Millipore).

Esterification of glutathione was performed using 5 ml ofmethanol acidified by addition of 100 µl of acetyl chloride.19

The compound was purified by semi-preparative RP-HPLC.The commercial peptides were used without further

purification and were diluted to concentrations of 10–20 µM

with 10 mM aqueous ammonium hydrogencarbonate solu-tion. About 10 equiv. of each peptide were mixed with theZn7MT solution and allowed to equilibrate for 15 min beforeinterrogation by nanoelectrospray mass spectrometry.

InstrumentationAn Applied Biosystems (Foster City, CA, USA) Q-Star hybridquadrupole/time-of-flight mass spectrometer equipped witha Protana (Odense, Denmark) nanoelectrospray source wasused for all the experiments. About 2 µl of sample wereloaded into a Protana metallized tip using a 1–10 µlGELoader (Eppendorf, Westbury, NY, USA). The instrumentwas set to analyze positive ions with an ionization voltage of900 V. The declustering potential was set at 100 V to limit thecontribution of non-specific interactions. For tandem massspectrometry, argon was used as the collision gas and thecollision energy was set between 10 and 45 V. Reportedcollision energies correspond to the voltage value needed todissociate 50% of the complex5 (Table 1).

External calibration was performed using polypropyleneglycol 3000 standard (Applied Biosystems). The quadrupolemass filter was tuned to transmit ions in the m/z 200–2500range. Simulation of isotopic distribution and calculationof monoisotopic and average mass were performed withthe ICR-2LS v2.66.79 program (EMSL, Richland, WA, USA).Determinations of atomic formulae were performed usingthe GPMAW 4.12 program (Lighthouse Data, Odense,Denmark). The formula M C nMemC � nmH was used tocalculate the average mass of the metallothionein, were Mis the mass of the apoprotein calculated in its reduced (SH)form, n, m and Me correspond to the number, the charge andthe mass of the metal ions, respectively, and H is the mass ofthe proton.20,21

RESULTS

The structures of the six glutathione analogs studied areshown in Table 1. Nanospray spectra of mixtures of eachof these incubated with metallothionein were found tocontain ions corresponding to bimolecular complexes. Forexample, in Fig. 1 the GSSG–MT complex is detected with1 : 1 stoichiometry, at charge states of 5C and 6C. InFig. 2, a complex formed between GSH and MT with 1 : 1stoichiometry and a 5C charge state is detected. Table 2summarizes the m/z values and stoichiometry observed foreach complex.

These complexes were then subjected to CID experi-ments. The same parent ion charge state (C5) was used in all

500 1000 1500 2000 25000

20

40

60

80

100

[GS

SG

+H

]+

Rel

ativ

e ab

unda

nce

m/z

[Zn 7

MT

+6H

]6+

[Zn 7

MT

+4H

]4+

[Zn 7

MT

+5H

]5+[Z

n 7M

T+

GS

SG

+6H

]6+

[Zn 7

MT

+G

SS

G+

5H]5+

Figure 1. Mass spectrum of a mixture of GSSG and Zn7MTrecorded with a declustering potential of 100 V and displayingmultiply charged Zn7MT–GSSG adducts.

500 1000 1500 20000

20

40

60

80

100

Rel

ativ

e ab

unda

nce

m/z

[GS

H+

H]+

[GS

H+

Na]

+

[Zn 7

MT

+6H

]6+

[Zn 7

MT

+3H

]3+

[Zn 7

MT

+4H

]4+

[Zn 7

MT

+5H

]5+[Z

n 7M

T+

GS

H+

5H]5+

[Zn 7

MT

+G

SH

+4H

]4+

Figure 2. Mass spectrum of a mixture of GSH and Zn7MTrecorded with a declustering potential of 100 V and displayingmultiply charged Zn7MT–GSH adducts.

Table 2. Characteristics of complexes of metallothionein withthe peptides in Table 1

Compound m/zCollision

energya (V)Zinc ions

transferred

GSH 1376.3 15 0GSSG 1437.3 35 1�ECox 1414.5 25 1CGox 1385.7 <10 0�ECEox 1466.2 40 1,2Me4�ECGox 1448.6 <10 0

a Collision energy for the decomposition of ¾50% of thecomplexes.

Copyright 2002 John Wiley & Sons, Ltd. J. Mass Spectrom. 2002; 37: 755–759

Page 4: Qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation

758 C. Afonso, Y. Hathout and C. Fenselau

cases in order to obtain consistent data. Each complex couldbe disassociated into the MT moiety and the correspond-ing glutathione analog. Interestingly, the collision energyrequired for these decompositions (see Experimental) variedacross the set. These values are summarized in Table 2. It canbe seen that complexes formed with GSSG, �ECox and �ECEox

require more energy for dissociation than complexes formedbetween MT and GSH, CGox and tetramethylated GSSG.

Another intriguing difference is found among thecomplexes in the distribution of zinc ions among thedecomposition products. In the CID spectrum shown inFig. 3(b), the products detected from the [GSH–MT]5C

complex (m/z 1376.3) are [GSH]C and [ZN7MT]4C. Incontrast, the CID spectrum in Fig. 3(a) reveals that the[GSSG–MT]5C complex decomposes in part to form [GSSG �H C Zn]C (m/z 675.1) and [Zn6MT]4C (m/z 1643.3). A zincion is transferred from MT to GSSG in the bimolecularcomplex. Metallothionein itself was found to be stable underthe conditions used for these CID studies.

This phenomenon was further investigated by examiningthe CID spectra of the other bimolecular complexes (Table 1).Product ion scans are shown in Figs 4 and 5 for complexesof MT with CGox, �ECox and �ECEox tetramethylated GSSG.The decomposition of the MT–CGox complex led to theformation only of the protonated CGox moiety (Fig. 4(a)).Similarly, CID of the Me4GSSG–MT complex only producedprotonated Me4GSSG (Fig. 4(b)). In both CID spectra of the

100

0

100

0

Rel

ativ

e ab

unda

nce

Rel

ativ

e ab

unda

nce

200 400 600 800 1000 1200 1400 1600 1800 2000m/z

[GSH+H]+

[GSSG+H]+

[GSSG-H+Zn]+

[Zn7MT+4H]4+

[Zn6MT+6H]4+

[Zn7MT+4H]4+

[Zn7MT+GSH+5H]5+

[Zn7MT+GSSG+5H]5+ (a)

(b)

Figure 3. Product ion scan of (a) [Zn7MT C GSSG C 5H]5C,m/z 1437.3, recorded with a collision energy of 35 V and (b)[Zn7MT C GSH C 5H]5C, m/z 1376.3, recorded with a collisionenergy of 15 V.

100

0

100

0

Rel

ativ

e ab

unda

nce

Rel

ativ

e ab

unda

nce [CGox+H]+

[Zn7MT+CGox+5H]5+

[Zn7MT+Me4GSSG+5H]5+

[Me4GSSG+H]+

[Zn7MT+4H]4+

[Zn7MT+4H]4+

(a)

(b)

400 600 800 1000 1200 1400 1600 1800 2000

m/z

Figure 4. Product ion scan of (a) [Zn7MT C CGox C 5H]5C, m/z1385.7, recorded with a collision energy of 10 V and (b)[Zn7MT C Me4GSSG C 5H]5C, m/z 1448.6, recorded with acollision energy of 10 V.

complexes that contain �Glu residues, pairs of dissociationproducts are observed that contain complementary numbersof zinc ions, as well as complementary numbers of charges(Fig. 5(a) and (b)). The presence of four glutamate residuesallows the transfer of one and two zinc ions (Fig. 5(b)), assummarized in Table 2.

DISCUSSION

Several explanations can be advanced to explain why highercollision energies are required to decompose MT complexeswith peptides that contain more than one glutamic acid. Thetwo side-chain carboxylate groups could form salt bridgeswith some of the eight protonated lysine side-chains thatextend out from the surface22 of folded metallothionein.Alternatively, these carboxylate groups may participate, withsulfhydryl groups from MT, in mixed-ligand co-chelationof one or more zinc ions. The redistribution of zinc ionsinto both dissociation products in three of the complexesreported here appears to be consistent only with the latterexplanation.

Based on the experimental evidence, it is proposedthat complexes formed with peptides containing multipleglutamate residues are stabilized by zinc sharing and thatthe other complexes are not stabilized in this way. Thisinterpretation also allows a mechanistic explanation to beadvanced for the difference observed by others16,17 in the

Copyright 2002 John Wiley & Sons, Ltd. J. Mass Spectrom. 2002; 37: 755–759

Page 5: Qualitative characterization of biomolecular zinc complexes by collisionally induced dissociation

Biomolecular zinc complexes 759

[Zn7MT+γECox+5H]5+

[Zn7MT+γECEox+5H]5+

[Zn6MT+6H]4+

[Zn5MT+8H]4+

[Zn6MT+6H]4+

[Zn7MT+4H]4+

[Zn7MT+4H]4+

[γECox+H]+

[γECEox+H]+

[γECox+Zn-H]+

[γECEox+Zn-H]+

[γECEox+2Zn-3H]+

400 600 800 1000 1200 1400 1600 1800 20000

100

100

0

Rel

ativ

e ab

unda

nce

Rel

ativ

e ab

unda

nce

(a)

(b)

m/z

Figure 5. Product ion scan of (a) [Zn7MT C �ECox C 5H]5C,m/z 1414.5, recorded with a collision energy of 25 V and (b)[Zn7MT C �ECEox C 5H]5C, m/z 1466.2, recorded with acollision energy of 40 V.

intracellular effects of reduced and oxidized glutathioneon zinc transfer from metallothionein to other proteins.In this hypothesis, oxidized glutathione co-chelates zincions in the metallothionein cluster, thus loosening orpartially removing them, while reduced glutathione iscomplexed by interaction with the protein in such a wayas to provide steric hindrance to other potential zinc ionrecipients.

It is not clear whether the qualitative experimentsreported here can be extended to provide meaningfulquantitative values for zinc ion affinities of MT and peptideco-chelators. Metallothionein binds seven zinc ions in twofluid clusters23 and it is not known if Zn2C is consistentlytransferred from the same position in the clusters oreven whether the same Zn2C is transferred to differentpeptides.

Work is under way to extend this qualitative approachto studies of biomolecular ligand fields involving othermetal ions.

AcknowledgementsThe authors thank Dr X. Yao and K. Reynolds for helpful discussions.This work was financially supported by the National Institutes ofHealth (GM-21248).

REFERENCES1. Cooks RG, Patrick JS, Kotiaho T, Mcluckey SA. Mass Spectrom.

Rev. 1994; 13: 287.2. Kaltashov IA, Fenselau C. Rapid Commun. Mass Spectrom. 1996;

10: 857.3. Nemirovskiy OV, Gross ML. J. Am. Soc. Mass Spectrom. 2000; 11:

770.4. Tito MA, Miller J, Walker N, Griffin KF, Williamson ED,

Despeyroux-Hill D, Titball RW, Robinson CV. Biophys. J. 2001;81: 3503.

5. Rogniaux H, Van Dorsselaer A, Barth P, Biellmann JF,Barbanton J, van Zandt M, Chevrier B, Howard E, Mitschler A,Potier N, Urzhumtseva L, Moras D, Podjarny A. J. Am. Soc. MassSpectrom. 1999; 10: 635.

6. Suhy DA, Simon KD, Linzer DIH, O’Halloran TV. J. Bio. Chem.1999; 274: 9183.

7. Kagi JHR. J. Trace Elem. Exp. Med. 2001; 14: 171.8. Lippard SJ, Berg JM. Principles of Bioinorganic Chemistry.

University Science Books: Mill Valley, CA, 1994.9. Udom AO, Rady FO. Biochem. J. 1980; 187: 329.

10. Ejnik J, Munoz A, Gan T, Shaw CF, Petering DH. J. Biol. Inorg.Chem. 1999; 4: 784.

11. Hathout Y, Fabris D, Fenselau C. Int. J. Mass Spectrom. 2001; 204:1.

12. Zaia J, Fabris D, Wei D, Karpel RL, Fenselau C. Protein Sci. 1998;7: 2398.

13. Li TY, Kraker AJ, Shaw CF, Petering DH. Proc. Natl. Acad. Sci.USA 1980; 77: 1838.

14. Otvos JD, Liu X, Li H, Shen G, Basti M. In Metallothionein III,Suzuki KT, Imura N, Kimura M (eds), Birkhauser Verlag: Basel,Switzerland, 1993; 58–74.

15. Brouwer M, Hoexum-Brouwer T, Cashon RE. Biochem. J. 1993;294: 219.

16. Maret W. Proc. Natl. Acad. Sci. USA 1994; 91: 237.17. Jiang L-J, Maret W, Vallee BL. Proc. Natl. Acad. Sci. USA 1998; 95:

3483.18. Jacob C, Maret W, Vallee BL. Proc. Natl. Acad. Sci. USA 1998; 95:

3489.19. Haroldsen PE. Rapid Commun. Mass Spectrom. 1994; 8: 371.20. Lei QP, Cui X, Kurtz DM, Amster J, Chernushevich IV,

Standing KG. Anal. Chem. 1998; 70: 1838.21. Fabris D, Joseph Z, Hathout Y, Fenselau C. J. Am. Chem. Soc.

1996; 118: 12 242.22. Robbins AH, McRee DE, Williamson M, Collett SA, Xuong NH,

Furey WF, Wang BC, Stout CD. J. Mol. Biochem. 1991; 221: 1269.23. Robbins AH, Stout CD. Methods Enzymol. 1991; 205: 485.

Copyright 2002 John Wiley & Sons, Ltd. J. Mass Spectrom. 2002; 37: 755–759