Protein Scission by Metal Ion–Ascorbate System

Protein Scission by Metal Ion–Ascorbate System

Jolanta Sereikaite,1,3

Jelena Jachno,2Rasa Santockyte,

2Piotr Chmielevski,

1Vladas-Algirdas

Bumelis,1 and Gervydas Dienys2

About 14 proteins were tested for specific oxidative scission catalyzed by metal ions in thepresence of ascorbate and oxidizing agents (O2 or hydrogen peroxide). Only four of them weredegraded by Fe3+/Fe2+- ascorbate, twelve – by Cu2+/Cu+-ascorbate and two proteins (a-and b-caseins) were degraded by Pd2+ ions. The rate and the intensity of degradation are very

different for various proteins. For the most of tested proteins only a small fraction ofmolecules was degraded. None of them was degraded completely. Two possible reasons ofprotein stability against oxidative degradation may be proposed as follows: either there is no

metal binding site in a protein molecule, or metal binding ligands of protein undergo a rapidoxidative modification and the metal ion is released from the binding site. Human growthhormone was cut specifically at two sites by Cu2+/Cu+-ascorbate system. At least one of

amino acid residues of this protein was modified by formation of reactive carbonyl.

KEY WORDS: Ascorbate; human growth hormone; metal binding site; metal ions; protein scission.

1. INTRODUCTION

Specific degradation of proteins (catalase, humanserum albumin, transferin) into smaller fragments inthe presence of ascorbate and Cu2+ ions underunusually mild conditions (pH � 7.0, room temper-ature) was discovered accidentally in 1983 by Chiouduring the study of anticancer activity of ascorbateand copper peptide complex (Chiou, 1983). He alsotested several other transition metals and found thatonly copper and iron ions have the protein – scis-sion activity in the presence of ascorbate. Laterother authors described similar observations of scis-sion by Fe/asc and/or Cu/asc systems of such pro-teins as glutamine synthase (Kim et al., 1985), malicenzyme (Chou et al., 1995; Wei et al., 1994), phos-phoenolpyruvate carboxykinase (Hlavaty and No-wak, 1997) and Na/K-ATPase (Goldsheleger and

Karlish, 1997). It was found that ascorbate as areducing agent may be replaced by thiol compounds(dithiothreitol, 2-mercaptoethanol). An oxidizingagent (H2O2 or molecular oxygen) was also re-quired. If the reaction was run under an oxygenfree atmosphere the protein remained intact (Kimet al., 1985).

Attempts were made to elucidate the mecha-nism of protein scission. Rana and Meares reportedthat in the presence of ascorbate and H2O2 (or O2),an iron chelate attached to a cysteine residue in car-bonic anhydrase cleaved the protein into two dis-crete fragments. Unmodified N- and C-ends wereproduced at the cleavage point. The transfer of an

1 Department of Chemistry and Bioengineering, Vilnius Gediminas

Technical University, Sauletekio al. 11, LT-2040, Vilnius,

Lithuania.2 Institute of Biotechnology, Vilnius, Lithuania.3 To whom correspondence should be addressed. E-mail: sjolanta@

fm.vtu.lt

Abbreviations: BSA, bovine serum albumin; CAPS, 3-(cyclo-

hexylamino) propanesulfonic acid; Cu/asc, Cu2+/Cu+-ascorbate;

DNPH, 2,4-dinitrophenylhydrazine; DTT, dithiothreitol; Fe/asc,

Fe3+/Fe2+-ascorbate; GCSF, recombinant human granulocyte

colony-stimulating factor; GuHCl, guanidinium hydrochloride;

hGH, recombinant human growth hormone; SDS-PAGE, sodium

dodecyl sulfate-polyacrylamide gel electrophoresis; TCA, trichlo-

roacetic acid; TFA, trifluoroacetic acid; TNF, recombinant human

tumor necrosis factor-alpha; Tricine, N-[2-hydroxy-1,1-bis(hy-

droxymethyl)ethyl]glycine

3691572-3887/06/0900-0369/0 � 2006 Springer Science+Business Media, Inc.

The Protein Journal, Vol. 25, No. 6, September 2006 (� 2006)DOI: 10.1007/s10930-006-9014-7Published Online: October 6, 2006

18O atom from [18O] H2O2 or [18O] O2 to the newlyformed carboxyl group was demonstrated by massspectrometry. The carboxyl group did not incorpo-rate significant amounts of 18O when the reactionwas carried out in the presence of [18O] H2O Theseresults may be explained by generation of a highlynucleophilic oxygen species such as peroxide coordi-nated to the iron chelate that attacks a carbonyl ofa peptide bond nearby (Rana and Meares, 1991).

The oxidative mechanism which exists along-side the hydrolytic mechanism and in some cases isthe dominant one was also proposed (Hoyer et al.,1990; Platis et al., 1993). The reaction is initiated bya-H abstraction, presumably by hydroxyl radical,leading to Ca radical. The oxidation at Ca and thecleavage of NH–Ca bond then follows. The newlyformed C-ends as well as in some cases N-ends aremodified, therefore sequencing of the fragments byEdman and carboxypeptidase methods is not possi-ble, contrary to the case of the hydrolyticmechanism.

In 1990, the idea was raised that it could bepossible to develop a new method for specific pep-tide bond splitting at a predetermined site using thisreaction (Hoyer et al., 1990; Rana and Meares,1990; Schepantz and Cuenoud, 1990), and a numberof publications followed. Certain positive resultswere obtained, but a general method was not devel-oped, because its specificity was not high enough. Acomplicating factor is the oxidative modification ofsome amino acid residues going in parallel with apeptide bond splitting (Berlett and Stadman, 1997;Stadman, 1993). Above-mentioned oxidative modifi-cation is also specific, like the peptide bond splittingcatalysed by metal ions. Only some amino acid resi-dues are modified, presumably those taking part inmetal ion binding or situated nearby.

Peptide bonds splitting catalyzed by metal ionswas successfully used for solving certain structuralproblems: for identificating metal binding sites(Gobson et al., 2000; Mustaev et al., 1997; Soundarand Colman, 1993), providing the information oninteracting regions of domains and subunits (Ghaimet al., 1995; Goldshleger and Karlish, 1999; Heydukand Heyduk, 1994; Miyake et al., 1998), studyingthe proximity relationships in membrane proteins(Shimon et al., 1998; Wu et al., 1995) and evaluat-ing the changes in protein conformation (Ermacoraet al., 1992, 1994).

Transition metal–ligand complexes that canperform hydrogen peroxide decomposition with theresulting production of reactive oxygen species are

considered to be widespread oxidative systems.They may be used by fungi for oxidative degrada-tion of wood components. Synthetic dyes are suc-cessfully decolorized by such systems; succinic acidas a ligand and Cu2+ or Co2+ were found to be themost effective combinations (Nerud et al., 2004).

The aim of this work was a systematic study ofoxidative protein degradation in the presence ofascorbate and metal ions. The influence of both thereactants and reaction conditions on result of reac-tion was investigated.

2. MATERIALS AND METHODS

2.1. Materials

Reagents were obtained from the following sour-ces: a-casein, b-casein, bovine serum albumin (BSA),EDTA disodium salt, FeCl3 � H2O, lysozyme, myo-globin from equine skeletal muscle, ovalbumin, pep-sin, sodium ascorbate, Sigma; 3-(cyclohexylamino)propanesulfonic acid (CAPS), CuCl2 � H2O, dith-iothreitol (DTT), guanidinium hydrochloride (Gu-HCl), HgCl2, PdCl2, MnCl2, sodium dodecyl sulfate(SDS), trichloroacetic acid (TCA), trifluoroacetic acid(TFA), Fluka; CaCl2 � 6H2O, CoCl2 � 6H2O, 2,4-dinitrophenylhydrazine (DNPH), FeSO4 � 7H2O,MgCl2 � 6H2O, NiCl2 � 6H2O, Pb(NO3)2, ZnCl2,Reachim (Russia); RuCl3 � xH2O, Aldrich; E. colib-galactosidase, E. coli alkaline phosphatase, E. coliDNA polimerase I, MBI Fermentas (Lithuania); pro-tein UK-114, recombinant granulocyte colony stimu-lating factor (GCSF), recombinant human growthhormone (hGH), recombinant human tumor necrosisfactor alpha (TNF) were kindly provided by SicorBiotech UAB (Lithuania). Distilled water furtherpurified with a Millipore MilliQ system was usedthroughout the present work.

2.2. Protein Cleavage Reaction

The reaction mixture containing 0.02–0.03 mMof protein, 20 mM of ascorbate and 1 mM of metalion in 0.1 M Tris–HCl buffer, pH 7.2, was incu-bated at 30�C for 120 min. Samples were withdrawnand the reaction was stopped by adding 5-fold con-centrated gel sample buffer (313 mM Tris–HClpH 6.8, 50% glycerol, 10% SDS, 500 mM DTT,0.05% bromophenol blue) (Goldshleger and Kar-lish, 1997). On purpose to detect peptide bondcleavage, samples of the reaction mixture were

370 Sereikaite, Jachno, Santockyte, Chmielevski, Bumelis, and Dienys

analyzed by sodium dodecyl sulfate-polyacrylamidegel electrophoresis (SDS-PAGE).

2.3. SDS-PAGE

SDS-PAGE was carried out on the MightySmall electrophoresis unit (Hoefer Scientific Instru-ments) according to the Laemmli method (Laemmli,1970). The total acrylamide concentrations used inthe separating gel were 10, 12 and 15%, dependingon molecular weight of proteins. A wide range pro-tein standard from MBI Fermentas (Lithuania) wasused as molecular weight markers. The gels werestained with Coomassie Brilliant Blue R-250 andscanned at 633 nm using the Dual Wavelength TLCscanner (CS 930, Shimadzu).

2.4. Tricine-SDS-PAGE

Tricine-SDS-PAGE was performed accordingto the method of Schagger and von Jagow (Schag-ger and von Jagow, 1987) using in the separatinggel a 16.5% total concentration of acrylamide. Theconcentration of cross-linker was equal to 3%. Pro-teins were visualized by silver staining. Low molecu-lar weight markers from Sigma were used.

2.5. Blotting and N-terminal Amino Acid Sequencing

Membrane Immobilon from Millipore wasused. The process of blotting was carried out atroom temperature and current density 0.5 mA/cm2

for 1 h, using a transfer buffer containing 10 mMCAPS, and 10% ethanol, pH 11.0. The amino acidsequence of the isolated fragments was directlydetermined by Edman degradation for 20–25 cycles.Automated Edman degradation of blotted hGHsamples was performed with an Applied BiosystemsProcise 491 protein sequencer.

2.6. Determination of Reactive Carbonyls in Proteins

Reactive carbonyl groups were analyzed asdescribed by Adams (Adams et al., 2001). After thereaction according to section 2.2., the protein wasprecipitated with ice-cold solution of trichloroaceticacid (final concentration 10% v/v). After a 10-minincubation at 4�C, samples were centrifuged at11,000 g for 3 min. The protein pellet was resus-pended in 0.5 ml of 10 mM solution of DNPH in

2 M HCl. The samples were incubated at roomtemperature for 1 h. The modified protein was pre-cipitated with 0.5 ml of 20% TCA, centrifuged at11,000 g for 3 min. The pellet was washed with1 ml of ethanol–ethyl acetate (1:1; v/v) to removefree DNPH reagent, then centrifuged for 5 min at11,000 g and the supernatant was discarded. Thewashing procedure was repeated three times. Theresulting protein pellet was resuspended in 1.0 ml of2 mM potassium phosphate solution (pH 2.3, ad-justed with TFA) containing 6 M GuHCl. To dis-solve protein, the samples were incubated at 37�Cfor 15–30 min. or at increased temperature up to50�C in the case of low protein solubility. Then allsamples were centrifuged to remove any insolublematerial remaining in suspension. The concentrationof DNPH derivatives was determined at its maxi-mum wavelength (360 nm) and the molar absorp-tion coefficient of 22 000 M)1 cm)1 was used toquantify the levels of protein carbonyls. Sampleswere spectrophotometrically analyzed against ablank of guanidine solution (6 M GuHCl with2 mM potassium phosphate). Protein concentrationwas determined spectrophotometrically at 280 nm.

3. RESULTS

3.1. Experimental Conditions of Protein Degradation

Protein splitting by metal ion–ascorbate systemis usually performed at mild conditions (pH � 7.0–7.5, 25–37�C). The concentration of metal ions var-ies from 5lM to 2 mM (Chou et al., 1995; Goldsh-leger and Karlish, 1997; Rana and Meares, 1991;Wei et al., 1994). In order to understand betterpeculiarities of the reaction, we studied the influenceof experimental factors (pH, temperature, time, con-centration of metal ions and ascorbate) on the rateand intensity of protein degradation. Products ofprotein cleavage were analyzed by SDS-PAGE. Thepercentage of residual uncleaved protein was esti-mated by the gel scanning. An original proteinamount was presumed as 100%. The peptide bondcleavage of a-casein by Fe/asc system and hGH byCu/asc system were examined in detail, and stan-dard conditions used for cleavage experiments werechosen.

pH dependence of a-casein and hGH degrada-tion is given on the Fig. 1. Protein scission isstrongly pH dependent in the case of both iron andcopper ions with narrow maximum at pH 7.0–7.6.

371Protein Scission by Metal Ion

The reaction is very slow below pH 6.0 and overpH 8.5.

Effects of reaction temperature and time arequite different for degradation of a-casein and hGH(Figs. 2 and 3). a-casein in Fe/asc system isdegraded more intensively at elevated temperatures(40–50�C). The temperature has little effect on deg-radation of hGH by Cu/asc system. Low maximumof degradation is located at 40�C. The a-casein scis-sion is a time-dependent process. After 120 min ofincubation approximately 50% of a-casein iscleaved. Degradation of hGH is rapid during initial30 min. Later it stops and the addition of moreascorbate and/or Cu2+ ions gives no effect. A pro-longation of the reaction time doesn’t change theextent of the cleavage, too. Control without addi-tion of metal ions and ascorbate incubated for 5 hat 30�C does not undergo any changes (Fig. 7).

The effect of reactant concentrations (metalions and ascorbate) is shown on Figs. 4 and 5. Theincreasing of metal ions concentrations results inmore intensive cleavage. A saturation of ascorbateis reached at the 20 mM concentration in the caseof both Fe3+ and Cu2+ ions.

Using the information presented on theFigs. 1–5 we selected a standard procedure for pro-tein degradation experiments: pH 7.2, 30�C, concen-tration of metal ions and ascorbate equal to 1 and20 mM, respectively, and incubation time 120 min.

In order to characterize the accuracy of thedensitometry method used in our experiments, thesame reaction mixture from one experiment (degra-dation of a-casein by Fe/asc system) was analyzedin 20 parallel lanes of electrophoregram Standarddeviation was equal to 6% of the mean value of theuncleaved protein.

3.2. Relative Intensity of Protein Degradation

We examined 14 proteins for the peptide bondcleavage by metal/ascorbate system under standardconditions. Fe3+ and Cu2+ ions were tested for theexhibition of this activity. The results are presentedin the Table 1. Only four of the examined proteinswere degraded by Fe/asc system. Cu/asc system wasfound to be more effective than the Fe/asc one. The

Fig. 1. Effect of pH on efficiency of a-casein and hGH cleavage

for 120 min at 30�C. Concentration of metal ions, 1 mM; ascor-

bate, 20 mM. Buffers used: pH 4.8, pH 5.8–6.6, pH 6.8–9.8 – ace-

tate, phosphate and Tris–HCl buffer, respectively. —n—

degradation of a-casein by Fe/asc system; � � �n� � � degradation of

hGH by Cu/asc system.

Fig. 2. Effect of temperature on efficiency of a-casein and hGH

cleavage for 120 min and pH 7.2. Concentration of metal ions,

1 mM; ascorbate, 20 mM. —n— degradation of a-casein by Fe/

asc system; � � �n� � � degradation of hGH by Cu/asc system.

Fig. 3. Time dependence of a-casein and hGH cleavage at 30�Cand pH 7.2. Concentration of metal ions, 1 mM; ascorbate,

20 mM. —n— degradation of a-casein by Fe/asc system; � � �n� � �degradation of hGH by Cu/asc system.


cleavage was registered for twelve proteins. Notonly the cleavage, but also the formation of SDS-resistant oligomers that were not separated intomonomers with DTT was observed in the presenceof metal/ascorbate in some cases (Figs. 6 and 7).Differences in the action of Fe/asc system and Cu/asc system were observed, especially in the case ofBSA. Different patterns of scission fragments wereobtained.

The degradation of the proteins is a ratherindividual process. The rate and intensity of degra-dation are very different for various proteins. Mosttested proteins were degraded to a low extent. Ourexperiments showed that BSA, ovalbumin, a-caseinand b -casein were also degraded by traditionalFenton chemistry (Fe2+/H2O2), EDTA reduced the

degradation of all these proteins in both case ofFe2+/H2O2 and Fe/asc systems.

Lysozyme and GCSF behaved in the differentway than other proteins examined for the scissionby Cu/asc system. The cleavage of lysozyme wasnegligible and similar after 1 and 6 h of the incuba-tion. GCSF underwent absolutely no cleavage.However, Cu2+/H2O2 system caused an appearanceof smaller molecular weight bands in the electr-

Fig. 5. Effect of ascorbate concentration on efficiency of a-caseinand hGH cleavage for 120 min at 30�C and pH 7.2. Concentra-

tion of metal ions, 1 mM. —n— degradation of a-casein by Fe/

asc system; � � �n� � � degradation of hGH by Cu/asc system.

Fig. 4. Effect of metal ions concentration on efficiency of a-case-in and hGH cleavage for 120 min at 30�C and pH 7.2. Concen-

tration of ascorbate, 20 mM. —n— degradation of a-casein by

Fe/asc system; � � �n� � � degradation of hGH by Cu/asc system.

Lane 1

116 66.2

45

35

25

18.4

14.4

7 2 3 4 5 6

Fig. 6 SDS-PAGE (15%) of time course TNF cleavage (0.1 M

Tris–HCl buffer pH 7.2). Concentration of Cu2+, 1 mM; ascor-

bate, 20 mM. Lane 1, protein markers; lane 2, control incubated

for 6 h at 30�C; lane 3–7, TNF plus Cu/asc incubated for 1, 2, 3,

4 and 6 h, respectively, at 30�C.

Table 1. Protein Cleavage by Metal–Ascorbate Systema

Protein Metal ion

Fe3+ Cu2+

BSA + +

Ovalbumin + +

a) casein + +

b) casein + +

hGH ) +

TNF ) +

E. coli b-galactosidase ) +

Lysozyme ) +

Pepsin ) )E. coli DNA polimerase I ) +

UK-114 ) +

GCSF ) )Myoglobin ) +

E. coli alkaline phosphatase ) +

a 30�C, 120 min, concentration of metal ions, 1 mM; ascorbate,

20 mM; 0.1 M Tris–HCl buffer, pH 7.2, reaction was stopped by

adding 5-fold concentrated gel sample buffer.

+ protein was cleaved.

) no cleavage.


ophoregram of both lysozyme and GCSF. Severalauthors have also shown the ability of Cu2+/H2O2

to degrade proteins (Cooper et al., 1985; Hawkinsand Davies, 1997). Co2+/H2O2 system also causedthe specific cleavage of these proteins resulting inthe appearance of smaller molecular weight band.

In the Table 2 we present the results of experi-ments aimed to reach the highest possible intensityof degradation (50–60�C, incubation time 5 h, high-er concentration of reactants in comparison withstandard scission procedure). Along with the prod-ucts of specific scission, lower molecular mass pep-tides were apparent after 5 h of BSA and hGHscission reaction. Moreover, in the lane of b-caseinthe band of initial protein diminished considerably,but no products of degradation were observed.Maybe small peptides were formed which escaped

from the gel during electrophoresis. Evidently, non-specific degradation prevails under harsh experi-mental conditions. The specificity of splitting maybe diminished by conformational changes of proteinat elevated temperature (Ermacora et al., 1992,1994). BSA was cleaved more effectively by Cu/ascthan by Fe/asc system. In the case of Cu/asc systemthe depth of cleavage was reached up to 80%. Noneof the tested four proteins was degraded completely.We could not reach more than 20% of hGH degra-dation. An addition of more splitting reagents (me-tal ions and ascorbate), higher temperature andlonger reaction time did not result in higher degreeof degradation (Table 2).

3.3. Catalytic Activity of Metal Ions

The screening of metal ions for their ability tocatalyze oxidative protein scission was described byseveral authors (Cr2+, Mn2+, Fe2+, Fe3+, Co2+,Ni2+, Cu2+, Zn2+ – Chiou, 1983; Mg2+, Ca2+,Mn2+, Fe2+, Co2+, Ni2+, Zn2+ – Wei et al., 1994;Cr2+, Mn2+, Fe2+, Fe3+, Ni2+, Cu2+, Zn2+,Mo2+ – Goldshleger and Karlish, 1997). Only ironand copper ions showed catalytic activity of all theions tested.

We tested a number of metal ions such asMg2+, Ca2+, Mn2+, Fe2+, Fe3+, Co2+, Ni2+,Cu2+, Zn2+, Ru3+, Pd2+, Hg2+, Pb2+ for degra-dation of a-casein, b-casein, hGH, and BSA; Co2+

and Pd2+ were additionally tested on hGH, UK-114, GCSF, TNF, E. coli b – galactosidase andlysozyme. Fe2+ and Fe3+ ions were equally effec-tive for protein scission in the presence of ascor-bate. Besides iron and copper ions protein scissionwas registered in the presence of Pd2+ and ascor-bate (Fig. 8). Pd2+-ascorbate system degraded onlytwo of the tested proteins – a-casein and b-casein.This degradation was inhibited by EDTA. The scis-sion fragments of a-casein were very similar in deg-radation by both Pd2+-ascorbate and Fe/ascsystem. Purity of PdCl2 used in scission experimentswas verified by the flame photometry method. Onlytrace quantity of iron was found.

It was described that addition of other metalions may inhibit the degradation of proteins by Fe/asc or Cu/asc systems (Goldshleger and Karlish,1997; Hlavaty and Nowak, 1997; Wei et al., 1994).We tested influence of nine metal ions (Mg2+,Ca2+, Mn2+, Co2+, Ni2+, Zn2+, Ru3+, Hg2+,Pb2+) on degradation of a-casein by Fe/asc system.

Lane

1 7

62 3 4 5

Fig. 7. SDS-PAGE (15%) of time course hGH cleavage (0.1 M

Tris–HCl buffer pH 7.2). Concentration of Cu2+, 1 mM; ascor-

bate, 20 mM; Lane 1, control of protein; lane 2, control incu-

bated for 5 h at 30�C; lane 3–7, hGH plus Cu/asc incubated at

30�C for 1, 2, 3, 4 and 5 h, respectively.

Table 2. Determination of Protein Degradation Degreea

Protein Uncleaved protein (%)

Fe/asc system Cu/asc system

a ) casein 20 30

b ) casein 50 50

BSA 70 20

HGH 100 81

a 50–60�C, 5 h, concentration of metal ions, 3 mM; ascorbate,

60 mM; 0.1 M Tris–HCl buffer, pH 7.2.


No significant influence on degradation rate wasnoticed using the concentrations of added metalions up to 30 times higher than one of Fe3+.

3.4. Degradation of hGH

hGH has a well characterized metal-bindingsite which shows high affinity to bivalent metal ionssuch as Co2+ and Zn2+ (Cunningham et al., 1991).This site consists of His18 and His21 from N-termi-

nal helix, and Glu174 from another helix laying inparallel with the first one.

The picture of hGH degradation by Cu/asc sys-tem given by SDS-PAGE is presented in Fig. 7.Only small fraction of the protein is degraded andsome cross-linked oligomers are formed The scis-sion products appears as two bands of lower molec-ular weight (Mw� 18.9 and 13.2 kDa, Mw of theintact protein being equal to 22 kDa). After separa-tion of the cleaved protein fragments by SDS-PAGE, they were transblotted electrophoreticallyonto polyvinylidene difluoride membrane and N-terminal amino acid sequence was analyzed with anautomatic protein sequenator. Both fragments hadthe same N-terminal sequences as intact hGH (10residues were identified). It implies that the mole-cule of hGH was cut specifically at two sites. Twosmaller peptides should have been formed, too.They were not seen on electrophoresis plate(Fig. 7). These smaller peptides were identified usingTricine-SDS-PAGE (Fig. 9), their molecular weightwas estimated to be 8.7 and 3 kDa. Attempts todetermine their N-end amino acid sequences wereunsuccessful. Perhaps their N-terminal amino acidresidues were modified during peptide bond scis-sion. It may happen if the reaction proceeds by theradical mechanism (Berlett and Stadman, 1997;Platis et al., 1993).

Besides peptide bond scission metal/ascorbatesystem also catalyze oxidative modification of aminoacid residues (Berlett and Stadman, 1997; Stadman,1993). Recently the modification of His18 and His21

residues of hGH by Cu/asc system was described.Both histidine residues were transformed into 2-oxo-histidines (Zhao et al., 1997). We used the methodfor determination of reactive carbonyl groups,

Lane

116

66.2

45

35

25

18.4

14.4

1 2 3 4

Fig. 8. SDS-PAGE (15%) analysis of 0.8 mg/ml solution of

a-casein incubated at 30�C and pH 7.2 for 120 min. Lane 1, pro-

tein markers; lane 2, control of protein; lane 3, degradation of

a-casein by Fe/asc system; lane 4, degradation of a-casein by

Pd2+ – ascorbate system.

Lane 10

17.2 14.6

8.24 6.38

2.56

2 3 4 5 6 7 8 91

Fig. 9. Tricine-SDS-PAGE analysis of 0.88 mg/ml solution of hGH at pH 7.2. Lane 1, protein markers; lane 3, control incubated at 30�Cfor 2 h; lanes 4–5, hGH plus Cu/asc incubated at 30�C for 2 h, concentration of Cu2+, 1 mM, ascorbate, 20 mM; lane 6, hGH plus Cu/

asc incubated at 30�C for 2 h, concentration of Cu2+, 2 mM, ascorbate, 20 mM; lane 7, control incubated at 40�C for 2 h; lanes 8–9,

hGH plus Cu/asc incubated at 40�C for 2 h, concentration of Cu2+, 1 mM, ascorbate, 20 mM; lane 10, hGH plus Cu/asc incubated at

40�C for 2 h, concentration of Cu2+, 2 mM, ascorbate, 20 mM.


described by Adams et al. (2001), in order to find ifmodification of other amino acid residues also takesplace. This method is based on the reaction of car-bonyl groups with DNPH. We tested the methodusing two proteins BSA and a-casein prior modifiedby metal-catalyzed oxidation (see section 2.2.). Thenwe applied this method to carbonyl group determina-tion of hGH. We found the significant increase incarbonyl levels in hGH modified by Cu/asc. At leastone more amino acid residue of hGH is modifiedbesides His18 and His21 by action of Cu/asc system.

4. DISCUSSION

Most publications about protein scission in thepresence of metal ions and ascorbate describe trans-formations of only one protein which for some rea-sons is important to the authors. After readingthem, one may be under the impression that all ornearly all proteins are degraded by these ox-red sys-tems. In this work we tested fourteen proteins andfound that ten of them are not degraded by Fe/ascsystem (Table 1) and two – by Cu/asc system. Pro-teins which are degraded also differ between them-selves in the rate and intensity of scission. Anotherpeculiarity of the reaction is that protein scissiondoes not reach completion. It is clear from theresults, presented in the Table 2 and from picturesof electrophoresis plates in this work (Figs. 6 and 7)and in other publications (Chiou, 1983; Goldshlegerand Karlish, 1997). An addition of more ox-redreagents to the reaction mixture does not move theprocess any further, as if some of the protein mole-cules are resistant to scission. Therefore, the oxida-tive degradation catalyzed by metal ions can notpretend to become a general method for preparativespecific splitting of proteins, except rare favourablecases. However, it may be useful for solving cer-tain structural problems of protein chemistry (seeIntroduction).

The mechanism of cleavage reaction was inves-tigated by several authors. Rana and Meares foundthat if an iron EDTA chelate was attached to Cys-212 of human carbonic anhydrase I, the modifiedenzyme in the presence of ascorbate and H2O2 (orO2) was rapidly cleaved between the residues Leu-189 and Asp-190 to produce two discrete fragments.Both fragments were not modified suggestinghydrolytic cleavage of the peptide bond. If thecleavage of the enzyme was performed using labeledoxidizing agents [18O]H2O2 or [18O]O2, an

18O atom

was incorporated into the newly formed carboxylicgroup of Leu-189. These results may be explainedby generation of a highly nucleophilic oxygen spe-cies, such as peroxide coordinated to the iron che-late that attacks a carbonyl carbon nearby (Fig. 10)The proximity and orientation of the reactinggroups are crucial to the observed selectivity of thereaction (Rana and Meares, 1991).

Analyzing reaction mixtures of peptides andproteins cleavage catalyzed by FeÆEDTA a numberof products were discovered that could not beformed by the hydrolytic mechanism (Hoyer, et al.,1990; Platis et al., 1993; Ermacora et al., 1994).Formation of these products was explained by radi-cal mechanism of peptide bond cleavage. Reactiveoxygen species are formed by Fe/asc and Cu/ascsystems in the presence of O2 or H2O2. The mostactive of them are hydroxyl-radicals. Radicals HOÆmay abstract Ca–H hydrogen atoms. Then the car-bon-centered radical formed reacts rapidly with O2

to form an alkyl peroxyl radical intermediate trans-formed later to alkylperoxide and alkoxyl radical(Fig. 11). The generation of alkoxyl radicals sets thestage for cleavage of the peptide bond by either thea-amidation or diamide pathways (Berlett andStadman, 1997; Platis et al., 1993). Peptide bondcleavage can also occur as a result of radical oxida-tion of prolyl, glutamyl and aspartyl side chains

Fig. 10. Hydrolytic mechanism of peptide bond scission (Rana

and Meares, 1991; Platis et al., 1993).


(Berlett and Stadman, 1997). Thus a number ofpolypeptide cleavage mechanisms must coexist.

The cleavage reaction seems to be independenton the nature of amino acid side chain. The detailedstudy of staphylococcal nuclease scission mediatedby attached FeÆEDTA showed that in 26 identifiedcleavage sites hydrophobic and hydrophilic, chargedand uncharged amino acid chains are represented,with no clear preference to any of them. Proximity tothe FeÆEDTA and accessibility of peptide bond wereproposed as the main factors determining the cleav-age sites (Ermacora et al., 1994). Partly a-helical pep-tide of 16 residues modified by covalent attachmentof FeÆEDTA was cleaved almost uniformly over itslength with no clear sequence selectivity among theresidues examined. The uniform cleavage may be dueto the partially unfolded state of the peptide and theshort lifetime of an a-helix, allowing an access to allpeptide units in each molecule during the lifetime ofthe active oxygen species (Platis et al., 1993).

Metal binding to the protein seems to be a nec-essary condition for protein splitting. A proximity

of peptide bond to the metal binding site and afavorable steric orientation are probably the mostimportant parameters in determining the cleavagesites. Only in this case active nucleophilic or radicalagents formed in the metal coordination sphere maytake part in the splitting reaction (Hoyer, et al.,1990; Rana and Meares, 1990). Enzymes, whichhave metal ions in their active centers are amongdegradable proteins (Chou et al., 1995; Hlavatyet al., 2000; Wei et al., 1994).

hGH is an interesting example of metal ionsbinding to protein. It is known that in blood hGHexists and is biologically active in monomeric form.In the presence of Zn2+ ions it associates into a di-meric quaternary structure which may be a storageform of this protein in pituitary gland (Cunning-ham, 1991). It was shown in our previous work (Di-enys et al., 2000) that some other ions also causedimerization of hGH, among them there is Cu2+,but not Fe2+ or Fe3+. Probably iron ions are notbinded in a hGH metal-binding site. As it is shownin the Table 1, hGH is degraded by Cu/asc system,

Fig. 11. Radical mechanism of peptide bond splitting: a – a-amidation pathway, b – diamide pathaway (Platis et al., 1993; Berlett and

Stadman, 1997).


but not by Fe/asc. It supports the assumption thatbinding of copper ions is the necessary conditionfor hGH degradation.

In parallel with peptide bonds scission, an oxi-dative modification of protein functional groupstakes place. After binding to a protein, metal ionssuch as copper and iron catalyze the transformationof side chains of amino acid residues situated incoordination sphere. Carbonyl containing groupsand other oxidation products are formed (Berlettand Stadman, 1997; Stadman and Oliver, 1991;Stadman, 1993). Probably it is the reason why pro-tein scission does not go to completion. If metalcoordinating amino acid residues becomes modified,metal atom is released from the coordination centreand the scission reaction stops. Therefore if proteinis not degraded by metal/ascorbate systems, it maybe that there is no metal binding center in the pro-tein molecule or amino acid residues in the bindingcenter are rapidly modified. Modification of aminoacid residues is usually more rapid process thanpeptide bond splitting (Stadman, 1993). For exam-ple, two of hGH histidine residues were rapidlymodified using the low concentration of Cu2+ ions(10 lM), at 25�C. Scission of the protein could notbe registered in such conditions (Zhao et al., 1997).As it is shown in our work, at least one other ami-no acid residues is oxidized to carbonyl derivatives.Oxidative modification of metal binding ligandsmay be the reason of limited degradation of hGHby Cu/asc system.

It is believed that oxidative degradation andmodification of proteins take place in vivo duringaging and in many pathological processes and mayhave biological significance. Acumulation of oxi-dized proteins is associated with a number of dis-eases (Adams et al., 2001; Berlett and Stadman,1997; Stadman, 1993).

REFERENCES

Adams, S., Green, P., Claxton, R., Simcox, S., Williams, M. V.,Walsh, K., and Leeuwenburgh, C. (2001). Frontiers inBioscience 6: 17–24.

Berlett, B. S., and Stadman, E. R. (1997). J. Biol. Chem. 272:20313–20316.

Chiou, S.-H. (1983) J. Biochem. 94: 1259–1267.Chou, W. Y., Tsai, W. P., Lin, C. C., and Chang, G.-G. (1995).

J. Biol. Chem. 270: 25935–25941.

Cooper, B., Creeth, J. M., and Donald, S. R. (1985). Biochem. J.228: 615–626.

Cunningham, B. C., Mulkerrin, M. G., and Wells, J. A. (1991).Science 253: 545–548.

Dienys, G., Sereikaite, J., Luksa, V., Jarutiene, O., Mistiniene,E., and Bumelis, V. A. (2000). Bioconjugate Chem. 11: 646–651.

Ermakora, M. R., Delfino, J. M., Cuenoud, B., Schepartz, A.,and Fox, R. O. (1992). Proc. Natl. Acad. Sci. 89: 6383–6387.

Ermakora, M. R., Ledman, D. W., Hellinga, H. W., Hsu, G. W.,and Fox, R. O. (1994). Biochemistry 33: 13625–13641.

Ghaim, J. B., Greiner, D. P., Meares, C. F., and Gennis, R. B.(1995). Biochemistry 34: 11311–11315.

Godson, G. N., Schoenich, J., Sun, W., and Mustaev, A. A.(2000). Biochemistry 39: 332–339.

Goldshleger, R., and Karlish, J. D. (1997). Proc. Natl. Acad. Sci.94: 9596–9601.

Goldshleger, R., and Karlish, J. D. (1999). J. Biol. Chem. 274:16213–16221.

Hawkins, C. L., and Davies, M. J. (1997). Biochimica etBiophysica Acta 1360: 84–96.

Heyduk, E., and Heyduk, T. (1994). Biochemistry 33: 9643–9650.Hlavaty, J. J., Benner, J. S., Hornstra, L. J., and Schildkraut, I.

(2000). Biochemistry 39: 3097–3105.Hlavaty, J., and Nowak, T. (1997). Biochemistry 36: 15514–15525.Hoyer, D., Cho, H., and Schultz, P. G. (1990). J. Am. Chem. Soc.

112: 3249–3250.Kim, K., Rhee, S. G., and Stadtman, E. R. (1985). J. Biol. Chem.

260: 15394–15397.Laemmli, U. K. (1970) Nature 227: 680–685.Miyake, R., Murakami, K., Owens, J. T., Greiner, D. P.,

Ozoline, O. N., Ishihama, A., and Meares, C. F. (1998).Biochemistry 37: 1344–1349.

Mustaev, A., Kozlov, V., Markovtsov, V., Zaychikov, E.,Denissova, L., and Goldfarb, A. (1997). Proc. Natl. Acad. Sci.94: 6641–6645.

Nerud, F., Baldrian, P., Eichlerova, I., Merhautova, V., Gabriel,J., and Homolka, L. (2004). Biocatal. Biotransform. 22: 325–330.

Platis, J. E., Ermakora, M. R., and Fox, R. O. (1993).Biochemistry 32: 12761–12767.

Rana, T. M., and Meares, C. F. (1990). J. Am. Chem. Soc. 112:2457–2458.

Rana, T. M., and Meares, C. F. (1991). Proc. Natl. Acad. Sci. 88:10578–10582.

Schagger, H., and Jagow, G.von (1987). Anal. Biochem. 166: 368–379.

Schepartz, A., and Cuenoud, B. (1990). J. Am. Chem. Soc. 112:3247–3250.

Shimon, M. B., Goldshleger, R., and Karlish, S. J. (1998). J. BiolChem. 273: 34190–34195.

Soundar, S., and Colman, R. F. (1993). J. Biol. Chem. 268: 5264–5271.

Stadman, E. R. (1993) Annu. Rev. Biochem. 62: 797–821.Stadman, E. R., and Oliver, C. N. (1991). J. Biol. Chem. 266:

2005–2008.Wei, C. H., Chou, W. Y., Huang, S. M., Lin, C. C., and Chang,

G.-G. (1994). Biochemistry 33: 7931–7936.Wu, J., Perrin, D. M., Sigman, D. S., and Kaback, H. R. (1995).

Proc. Natl. Acad. Sci. 92: 9186–9190.Zhao, F. Z., Schoneich, E. G., Aced, G. I., Hong, J., Milby, T.,

and Schoneich, C. (1997). J. Biol. Chem. 272: 9019–9029.


Protein Scission by Metal Ion–Ascorbate System

Documents

Transcript of Protein Scission by Metal Ion–Ascorbate System