covalent attachment of 4-hydroxynonenal to glyceraldehyde-3

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 268, No. 9, Iseue of March 25, pp. 6388-6393,1993 Printed in U S A .

Covalent Attachment of 4-Hydroxynonenal to Glyceraldehyde-3- Phosphate Dehydrogenase A POSSIBLE INVOLVEMENT OF INTRA- AND INTERMOLECULAR CROSS-LINKING REACTION*

(Received for publication, August 13, 1992)

Koji UchidaS and Earl R. Stadtmant From the Laboraton, of Biochemistrv. National Heart. Lung, and Blood Institute, National Institutes of Health, Bethesda, Mary.!& 20892

_ .

Cytotoxic action of membrane lipid peroxidation product 4-hydroxynonenal (HNE) is due mainly to its facile reactivity with proteins (Esterbauer, H., Schaur, R. J., and Zollner, H. (1991) Free Radical Biol. Med. 11, 77-80). In the present study, the detailed mechanism of HNE modification of a key enzyme in inter- mediary metabolism, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), is studied mainly focusing on the formation of HNE-amino acid adducts in the enzyme. When GAPDH (1 mglml) was treated with 0-2 mM HNE in sodium phosphate buffer (pH 7.2) for 2 h at 37 “C, the enzyme was inactivated by HNE in a concentration-dependent manner. The loss of enzyme activity was associated with the loss of free sulfhydryl groups. Following its reduction with NaEiH4, amino acid analysis of the HNE-modified enzyme demonstrated that histidine and lysine residues were also modified. At concentrations lower than 0.5 mM, HNE reacts preferentially with cysteine and lysine residues. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the HNE-modified enzyme suggested the formation of intra- and intermolecular cross-links of the enzyme subunit. The HNE-dependent loss of amino acid residues was accompanied by the generation of protein-linked carbonyl derivatives as assessed by reduction with NaBraH]H4 and reaction with 2,4-dinitrophenylhydrazine. Thus, the conjugation of all the amino acids appears to involve Michael addition type reactions in which the carbonyl function of HNE would be preserved. The modified histidine residues were quantitatively recovered as the HNE-histidine adduct. However, only 28% of the missing lysine could be accounted for as the HNE-lysine derivative, and only 15.6% of the modified cysteine could be accounted for as the HNE-cysteine thioether derivative. It is pro- posed that the carbonyl groups of the HNE-derived Michael addition products may undergo secondary reactions with the amino acid groups of lysine residues to yield inter- and intrasubunit cross-links.

The cytotoxicity of lipid peroxidation products has been attributed to the formation of reactive aldehydes, such as

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

3 Present address: Dept. of Food Science and Technology, School of Agriculture, Nagoya University; Nagoya 464-01, Japan.

$ To whom correspondence should be addressed: National Insti- tutes of Health, 9000 Rockville Pike, Bldg. 3, Rm. 222, Bethesda, MD 20892. Tel.: 301-496-4096 Fax: 301-496-0599.

alkenals, 2-alkenals, and 4-hydroxyalkenals (1, 2). These aldehydes are considerably more stable than free radicals and thus are likely to diffuse into the cellular medium. Among these aldehydes, 4-hydroxynonenal (HNE)’ is one of the major products of membrane peroxidation and shows many biological effects (2-6) such as high toxicity to cells, the lysis of erythrocytes, and inhibition of the synthesis of DNA and protein. It is believed that HNE may be responsible for some of the tissue damage that occurs in vivo under conditions of oxidative stress. It was shown earlier (7) that HNE reacts readily with sulfhydryl groups of proteins to form thioether adducts which further undergo cyclization to form hemiace- tals.

Based on the fact that .Raney nickel catalyzes cleavage of the thioether bond, a procedure was developed for the detection and quantification of thioether derivatives produced by the interaction of HNE and other 2-alkenals with sulfhydryl groups of model compounds (8). Using this procedure, it was shown that, in addition to the formation of simple thioether adducts, the reaction of HNE with sulfhydryl groups of proteins leads to other as yet unidentified products (8).

It is generally accepted that the aldehyde moiety of HNE and other 2-alkenals reacts with e-amino groups of lysine residues in proteins to form a,@-unsaturated aldimines (2). In fact, the modification of low density lipoprotein by HNE is associated with a significant loss of lysine residues (9). It was established also that HNE reacts with histidine residues of low density lipoprotein (10) and several other proteins (11). It is therefore evident that HNE can modify a number of amino acid residues in proteins. It is thus important to char- acterize the HNE-modified protein at the molecular level in order to understand the mechanism of a large number of biological effects induced by HNE.

To this end, the present investigation was undertaken using GAPDH as a convenient model protein, which is known to be highly sensitive to the inactivation by lipid peroxidation products in vitro (12-14). In the present paper, we studied HNE modification of GAPDH mainly focusing on the formation of HNE adducts in the enzyme.

EXPERIMENTAL PROCEDURES

Materials-Stock solution of trans-4-hydroxy-2-nonenal was prepared by the acid treatment (1 mM HCl) of 4-hydroxynonenal diethy- lacetal which was generously provided by Dr. H. Esterbauer (Univer- sity of Graz). The concentration of the HNE stock solution was determined by the measurement of UV absorbance at 224 nm (15). Sodium borotetratritide (5-15 Ci/mmol) was obtained from Du Pont-

The abbreviations used are: HNE, 4-hydroxynonenal; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; 2,4-DNP, 2,4-dinitrophenylhydrazine; NaBH4, sodiumborotetrahydride; OPA, o-phthaldehyde; HPLC, high performance liquid chromatography.

6388

HNE Modification of GAPDH 6389

New England Nuclear and trypsin treated with ~-1-tosylamide-2- phenylethyl chloromethyl ketone, Raney nickel-activated catalyst, NAD, ~~-glyceraldehyde-3-phosphate, N"-acetyl-L-lysine, and Nu- acetyl-L-histidine were from Sigma. Rabbit muscle GAPDH was obtained from Calbiochem, 2,4-dinitrophenylhydrazine (2,4-DNP) was from Fluka, and 5,5'-dithiobis-(2-nitrobenzoic acid) was from Eastman Kodak Company. Raney nickel was rinsed thoroughly with water and ethanol prior to use.

Inactivation of GlyceraMehyde-3-phsposphate Dehydrogenase by 4- Hydroxynonenal-The reaction mixture (1 ml) containing 1 mg ofGAPDH and 0-2 mM HNE in 50 mM sodium phosphate buffer (pH 7.2) was incubated for 2 h at 37 "C (hereafter referred to as "mixture P).

Assay of Enzyme Activity-A 10-p1 aliquot of mixture A was assayed in 3 ml of 15 mM sodium pyrophosphate, 30 mM sodium arsenate buffer (pH 8.5). Reaction of GAPDH was initiated by the addition of 100 pl of 7.5 mM NAD, 100 p1 of 0.1 M dithiothreitol, and 100 pl of 15 mM ~~-glyceraldehyde-3-phosphate. The mixture was incubated at room temperature for 3 min and the absorbance at 340 nm was measured.

Sulfhydryl Assay-An aliquot (400 pl) of mixture A was treated with 10% trichloroacetic acid. The precipitated protein was dissolved with 0.9 ml of 8 M guanidine hydrochloride, 13 mM EDTA, 133 mM Tris solution (pH 8.0) containing 1 mM EDTA and then 0.1 ml of 10 mM 5,5'-dithiobis-(2-nitrobenzoic acid) was added. After 5 min, the absorbance at 412 nm was measured. For quantitation, the reaction with cysteine served as a reference standard.

Amino Acid Composition-An aliquot (400 pl) of mixture A was treated with 40 pl of 10 mM EDTA, 40 pl of 1.0 N NaOH, and 40 pl of 100 mM NaBH, in 0.1 N NaOH. After 1 h at 37 'C, the mixture was treated with 10% trichloroacetic acid, and the precipitated protein was then hydrolyzed with 6 N HC1 (200 pl) for 20 h at 110 "C under nitrogen atmosphere. The hydrolyzed sample was evaporated to dryness and redissolved in 0.1 ml of 50 mM sodium phosphate buffer (pH 8.0) containing 1 mM EDTA. An aliquot (10 p l ) was labeled with o- phthaldehyde (OPA) for determination of amino acid composition by HPLC (16). Reverse-phase HPLC was performed on a Hewlett- Packard model 1090 chromatography equipped with a Hewlett-Pack- ard model 1046A programmable fluorescence detector.

Determination of Protein Carbonyl Content by Reduction with NaBH4-An aliquot (400 p1) of mixture A was treated with 10% trichloroacetic acid (w/v, final concentration); the precipitate was dissolved with 400 pl of 8 M guanidine hydrochloride, 13 mM EDTA,

EDTA (40 pl) and 1 N NaOH (40 pl) in a 1.5-ml Sarstedt tube fitted 133 mM Tris solution (pH 7.2). The solution was mixed with 0.1 M

with O-ring and a cap. Then, 40 pl of 0.1 M NaB[3H]H4 in 0.1 N NaOH was added, and the mixture was incubated at 37 "C for 1 h. The reaction was terminated by the addition of 100 p1 of 1 N HC1. The mixture was then applied to PD-10 column (Sephadex G-25), equilibrated in 6 M guanidine hydrochloride, in order to separate 3H- labeled protein from other radioactive contaminants. The protein fraction is hereafter referred to as "mixture B." Recovery of protein was determined by the measurement of UV absorbance at 278 nm ( 6

= 10.2). The protein carbonyl content was calculated from radioactivity measurement of the suitable aliquota.

Determination of Protein Carbonyl Content by 2,4-Dinitropheny- lhydrazine-An aliquot (400 pl) of mixture A was treated with an equal volume of 0.1% (w/v) 2,4-DNP in 2 N HC1 and incubated for 1 h at room temperature. This mixture was treated with 400 p1 of 20% trichloroacetic acid (w/v, final concentration), and after centrifugation, the precipitate was extracted three times with ethanol/ethyl acetate (l:l, v/v). Then protein sample was dissolved with 1 ml of 8 M guanidine hydrochloride, 13 mM EDTA, 133 mM Tris solution (pH 7.2) and measured UV absorbance at 365 nm. The results were expressed as moles of 2,4-DNP incorporated/mole of protein subunit based on an average absorptivity of 21.0 mM" cm" for most aliphatic hydrazones (17).

4-Hydroxynonenal-Histidine Adduct-The standard sample of HNE-histidine adduct was prepared by reaction of HNE with N- acetylhistidine as reported previously (11). N-Acetylhistidine (50 mg) was treated with 8 mM HNE in 2 ml of 50 mM sodium phosphate buffer (pH 7.2) for 20 h at 37 "C. Formation of products was determined by HPLC. A linear gradient of 0.05% trifluoroacetic acid in water (solvent A)-acetonitrile (solvent B) (time = 0, 100% A; 20 min 0% A), a t a flow rate of 1 ml/min, was used with an Apex Octadecyl 5U column (0.46 X 15 cm) (Jones Chromatography). After isolation of the adduct, an aliquot (0.1 ml) was taken in a hydrolysis vial and treated with 10 pl of 10 mM EDTA, 10 p1 of 1 N NaOH, and 10 p1 of

0.1 M NaBH, in 0.1 N NaOH for 1 h at 37 "C. After incubation, the mixture was treated with 1 N HCl (30 pl) and then hydrolyzed with 6 N HCl (0.2 ml) for 20 h at 110 "C under nitrogen atmosphere. After incubation, the mixture was evaporated to dryness, and then 0.2 ml of 50 mM sodium phosphate buffer (pH 8.0) containing 1 mM EDTA was added. This solution was used for amino acid analysis as the standard sample of HNE-histidine adduct. An aliquot (10 p1) of this solution was prelabeled with OPA and used for determination of HNE-histidine adduct by amino acid analysis.

4-Hydroxynonenal-Cysteine Adduct-For quantitation of HNE- cysteine adduct in the HNE-modified protein, an aliquot (100 pl) of mixture B was mixed with 300 pl of 8 M guanidine hydrochloride, 13 mM EDTA, 133 mM Tris solution (pH 7.2) and 400 mg of Raney nickel (8). The mixture was incubated for 15 h at 55 "C. Control samples received no catalyst. After incubation, the reaction mixture was extracted twice with 500 pl of chloroform/methanol (9:l). The extract was dried with sodium sulfate, and 100 pl was taken for radioactivity measurement.

4-Hydroxynonenal-Lysine Adduct-Nu-Acetyl lysine (50 mg) was incubated with 8 mM HNE in 2 ml of 0.1 M HEPES buffer (pH 7.2) for 20 h at 37 "C. Formation of product was determined by HPLC under the same HPLC chromatographic conditions as those of HNE- histidine adduct. The major component was collected and character- ized by fast atom bombardment-mass spectrometry on a JEOL JMS- S X 102 mass spectrometer. The mass spectral analysis of the purified compound revealed a quasi-molecular (M+H) of 335, which is consistent with the formation of Michael addition type HNE-lysine adduct. The isolated HNE-lysine adduct was reduced with NaBHd and then hydrolyzed under the same procedures of HNE-histidine adduct. The hydrolyzed sample was used for amino acid analysis as the standard sample of HNE-lysine adduct. The OPA derivative at HNE-lysine adduct was eluted right after the lysine peak on the HPLC chromatogram.

Sodium Dodecyl Sulfate-Polyacrylumide Gel Electrophoresis-Mix- ture A (100 pl) was treated with 10 p1 of 10 mM EDTA, 10 pl of 1 N NaOH, and 10 pl of 0.1 M NaBH, in 0.1 N NaOH for 1 h at 37 "C. The sample was then subjected to SDS-polyacrylamide gel electrophoresis (10% acrylamide) according to Laemmli (18). Protein bands were stained with Coomassie Brilliant Blue R-250.

Trypsin Digestion-GAPDH (0.5 mg) was incubated in the presence and absence of 0.2 mM HNE in 0.5 ml of 50 mM sodium phosphate buffer (pH 7.2) for 2 h at 37 "C. Reaction mixtures of native and HNE-modified protein were then treated with 50 pl of 10 mM EDTA, 50 p l of 1 N NaOH, and 50 p1 of 0.1 M NaBH, in 0.1 N NaOH. After 1 h at 37 "C, mixtures were treated with 10% trichloroacetic acid. The precipitate was collected by centrifugation and was suspended in 0.5 ml of 1 M Tris buffer (pH 8.0) and digested with 20 p1 of trypsin for 20 h at 37 "C. Peptides were analyzed by HPLC on a Vydac C18 peptides and proteins column (0.21 X 25 cm) which had been previously equilibrated with 0.05% trifluoroacetic acid in water and developed using a linear gradient of acetonitrile (l%/min) at a flow rate of I ml/min. A decrease in three peptides was confirmed by comparing the peptide map from native and HNE-modified proteins. These peptides were isolated form the trypsin digest of native protein and subjected to amino acid analysis. Sequences of the peptides were assessed by comparing the amino acid analysis data with the known sequence of GAPDH (19).

RESULTS

Modification of Glyceraldehyde-3-phosphate Dehydrogenase by Reaction with 4-Hydroxynonenal-To test the effect of HNE on the enzyme activity, GAPDH was treated with various concentrations of HNE (0-2 mM) for 2 h at 37 "C. As shown in Fig. lA, activity of the enzyme declined rapidly to only 20% of the initial value as the concentration of HNE was varied from 0 to 0.5 mM. The loss of activity was accompanied by the loss also of cysteine (sulfhydryl groups), lysine, and histidine residues (Fig. 1B). It is evident from the data in Fig. 1, A and B that the loss of catalytic activity is correlated with the loss of about two sulfhydryl groups, 3 lysine residues, and possibly 1 histidine residuelsubunit. The relationship between histidine modification and the loss of activity is not clearly evident. At a concentration of 0.5 mM HNE, where 80% of the catalytic activity is lost, approximately 1 residue

6390 HNE Modification of GAPDH

I 0.0 1 .o 2.0

4-Hydroxynonenal (mM)

6 Bl

0.0 1 .o 2.0 4-Hydroxynonenal (mM)

FIG. 1. HNE-provoked loss of enzyme activity (A) and loss of amino acid residues ( B ) in GAPDH. The enzyme (1 mg/ml) was incubated with 0-2 mM HNE in 50 mM sodium phosphate buffer (pH 7.2) for 2 h at 37 "C. Symbols: 0, HNE-cysteine; A, HNE-lysine; 0, HNE-histidine.

of histidine/subunit is lost. However, the modification of histidine residues continues to increase with further increases in the concentration of HNE, even beyond the concentration (1.0 mM) at which the loss of activity is nearly complete. At the highest level of HNE tested (2.0 mM), about 5 histidine residues/subunit were modified.

Effect of Substrates on the Modification of Glyceraldehyde- 3-phosphate-Dehydrogenase by HNE-Trypsin digestion of the HNE (0.2 mM)-modified enzyme followed by peptide mapping demonstrated the selective loss of three peptides (Ile'G3-Lys'6@, Val'@"Lys'm, and Va1'32-Arg2'6) (data not shown). Among them, both Ile1G3-Lys'6@ and Val'@"Lys'm are known to locate at the active center of GAPDH. To determine if amino acid residues at the catalytic site of GAPDH are among those modified by HNE, the effects of NAD+ and glyceraldehyde-3-P on the modification reactions was exam- ined. As shown in Table I, 1 less cysteine and 2 fewer histidine residues were modified when NAD+ was present in the reaction mixture, but NAD+ had no effect on the modification of lysine residues. In contrast, glyceraldehyde-3-P had no effect on the number of cysteine and histidine residues that were modified, but it appears to promote the modification of 1 additional lysine residue by HNE.

The protective effect of NAD+ on the modification of cysteine and histidine by HNE is understandable in light of the fact that one cysteine (Cys'") and 1 histidine (His"*) residues are situated at the NAD+-binding site of the enzyme (19, 20). A second histidine (His=) is partially protected from photo- oxidation modification by NAD+, but its presumed implica- tion in the catalytic mechanism is suspect since it is located some distance from the catalytic site and its replacement by a glutamate residue in the lobster enzyme does not compro- mise catalytic function (19, 20). The observation that NAD+

TABLE I Effect of the substrates on the loss of amino acids in glyceraldehyde-

3-phosphate dehydrogenase by reaction with 4-hydrorynonenal Glyceraldehyde-3-phosphate dehydrogenase (1 mg/ml) was incu-

bated with 2 mM HNE in the presence and absence of substrates (1.5 mM glyceraldehyde-3-phosphate or 75 pM NAD) in 50 mM sodium phosphate buffer (pH 7.2) for 12 h at 36 "C. Cysteine content was determined by 5,5'-dithiobis-(2-nitrobenzoic acid), and lysine and histidine contents were determined by HPLC according to the meth- ods under "Experimental Procedures." Abbreviation: GAP, DL-glyceraldehyde-3-phosphate.

Amino acid content Systems (mol/mol subunit)

Cysteine Lysine Histidine __ GAPDH 3.38 27.64 11.24 GAPDH/HNE 0.89 23.47 5.61 GAPDH/HNE/GAP 0.83 22.54 5.89 GAPDH/HNE/NAD 1.64 23.28 7.56 __

had no effect on the number of lysine residues modified is surprising since 1 lysine residue (Lys'=) and possibly others as well have been shown to influence the binding of NAD+ (20) to the enzyme. Perhaps conformational changes of protein structure elicited by the binding of NAD' leads to exposure and HNE-dependent modification of several other lysine residues and thereby masks the effect of HNE on Lys" (see "Discussion").

Qmntitution of the 4-Hydroxynonenul Protein Adducts-It is generally accepted that the aldehyde moiety of HNE reacts with t-amino group of lysine residues in proteins to form Schiff base conjugates (2). However, in recent studies with N - acetyl lysine and glucose-6-phosphate dehydrogenase,' we showed that the preferred reaction with HNE involves addition of the amino group to the a,@-double bond to form a secondary amine, with retention of the aldehyde function. As reported previously (ll), the imidazole group of histidyl residues in proteins also reacts with the double bond of HNE. The HNE adducts of histidine and lysine residues are both stabilized by reduction of their respective aldehyde moieties with NaBH4, and upon acid hydrolysis, they yield new amino acid derivatives. The OPA derivatives of these new amino acids are readily separated from all normal amino acids by HPLC. By this procedure, it was demonstrated that at least three different HNE-histidine derivatives (perhaps isomers) are formed, all of which elute from the HPLC column between the OPA derivatives of leucine and lysine ( l l ) , whereas the OPA derivatives of the HNE-lysine adduct elute just after the OPA derivative of unmodified lysine.' As shown in Fig. 2, the OPA derivatives of all three HNE-histidine adducts and the HNE-lysine adduct were detected in the HNE-modified GAPD. Moreover, the amounts of each increased as the level of HNE in the incubation mixture was increased from 1 to 2 mM. Interestingly, the total amount of HNE-histidine present in the acid hydrolysate of HNE-treated GAPH, as determined by the HPLC procedure, accounted for about 90% of the histidine that was lost upon HNE treatment (Fig. 3). In contrast, the amount of HNE-lysine adduct that was recovered accounted for only about 28% of the lysine that disappeared during the HNE treatment (Fig. 3). To determine the amount of HNE that becomes conjugated in thioether linkage to cysteinyl residues in proteins, we developed a procedure which takes advantage of the fact that after reduction with NaB['H]H4 the labeled HNE moiety can be cleaved from the protein by Raney nickel desulfurization, and after extraction into an organic solvent, the 3H-labeled product can

- ~~

Szweda, L. I., Uchida, K., Tsai, L., and Stadtman, E. R. (1993) J. Bwl. Chem. 268,3342-3347.


Leu LYS

19.0 19.5 20.0 20.5 Time (min.)

FIG. 2. Separation of the 4-hydroxynonenal adducts by HPLC. The enzyme (1 mg/ml) was incubated with 0

buffer (pH 7.2) for 2 h at 37 “C. After reduction with NaBH,, reaction (. . . . ), 1 (- - -), and 2 mM (-) HNE in 50 mM sodium phosphate

mixtures were hydrolyzed and the amino acid composition was analyzed by HPLC following derivatization with OPA. The arrows in A indicate the peaks corresponding to the HNE-histidine adducts. The arrow in B indicates the HNE-lysine peak. The peaks marked Leu and Lys correspond to the OPA derivatives of leucine and lysine, respectively.

1

His Lys Cys FIG. 3. Comparison of the amounts of amino acid residues

lost and the yields of HNE-adducts. GADPH (1 mg/ml) was incubated with 2 mM HNE in 50 mM sodium phosphate buffer (pH 7.2) at 37 ”C. The amount of HNE-cysteine present as a simple thioester was determined by the Raney nickel procedure (8) as described under “Experimental Procedures.” For quantitation of HNE- lysine and HNE-histidine adducts, the reaction mixture was incubated with NaBH,, then hydrolyzed with 6 N HC1 for 20 h at 110 “C, and the amino acids were separated by HPLC after labeling with OPA as described under “Experimental Procedures.” open bars indicate the number of amino acid residues lost per subunit. Closed bars indicate the number of amino acid residues recovered as recognizable HNE-adducts.

be quantitated by radioactive measurements (8). Using this procedure, it was established that only 15.6% of the cysteine residues in HNE-modified GAPDH which disappeared could be present as simple thioether adducts (Fig. 3).

Introduction of Carbonyl Groups into Protein by Reaction with HNE-In earlier studies, it was demonstrated that the oxidative inactivation of enzymes by metal ion-catalyzed reactions or by exposure to y-irradiation leads to the conversion of some amino acid residues to carbonyl derivatives (21-24). Highly sensitive procedures for the detection and quantitation of protein carbonyl groups were developed based on the reactions with NaB[3H]H4 to form 3H-labeled hydroxy derivatives or with 2,4-DNP to yield protein hydrazones. It is therefore noteworthy that the reaction of HNE with func- tional groups of proteins leads to products which will react with both kinds of carbonyl reagents.

As shown in Fig. 4, HNE can react with proteins to form three kinds of derivatives. (i) The aldehyde group of HNE

can react directly with the €-amino group of lysine residues to form Schiff base derivatives (reaction a ) . (ii) The cu,P-double bond of HNE can react with sulfhydryl, histidyl, or amino groups to produce Michael addition type adducts (reaction 4 . (iii) The aldehyde moiety of the Michael addition products can undergo secondary reactions with e-amino groups of residues to form intra- or interprotein Schiff base cross-links (reaction f ). The products produced by all three mechanisms will react with either NaEi[3H]H, (reactions b, e, and h) or 2,4-DNP (reactions c, i , and g) as illustrated in Fig. 4. There- fore, the carbonyl assays now in use, cannot distinguish between carbonyl groups introduced into proteins by reactions d , or d + f (Fig. 4) from those generated by direct oxidation of amino acid side chains (21-24). However, HNE-derived Schiff base adducts produced by reaction a (Fig. 4) can be distinguished from born fide protein carbonyl groups and from carbonyl groups introduced by Michael addition type mechanism (reaction d ) by a comparison of their reactions with 2,4- DNP and NaB[3H]H4. Reduction of the Schiff bases by NaB[3H]H4 (reactions b, e, and h) will in each case lead to the generation of a 3H-labeled protein derivative. However, reaction of the Schiff base produced by reaction a with 2,4- DNP will lead to cleavage of the Schiff base, and concomi- tantly to regeneration of the protein amino group, and to release of the HNE moiety as the free hydrazone derivative (reaction c). This hydrazone would be separated from the protein by the procedure used. Thus, products formed by reaction a will give a positive test for protein carbonyl groups when assayed by the 3H-labeling technique but will not yield a protein-associated 2,4-DNP derivative. In light of these considerations, it is significant that the amount of protein carbonyl groups produced upon treatment of GADPH with HNE is the same, whether determined by the 3H-labeling technique or the 2,4-DNP procedure (Fig. 5). This precludes a significant contribution of simple Schiff base formation (reaction a, Fig. 4) in the modification of GADPH by HNE, but leaves open the possibility that Schiff base formation could occur as a secondary reaction after formation of a Michael addition-type reaction (reaction d , Fig. 4).

Relationship between the Number of Amino Acid Residues Modified and the Number of Products Identified-From the data in Fig. 3, it is evident that about 90% of the His residues which are modified by HNE followed by NaBH4 reduction can be accounted for as a mixture of new amino acids which, after derivatization with o-phthaldehyde, can be separated from all other amino acid derivatives (see “Experimental Procedures”). However, only 28% of the lysine residues which were modified could be accounted for as the HNE-lysine adduct, and only 15.6% of the modified cysteine residues could be identified as the thioether adduct. Interestingly, the sum of the amounts of Cys-, Lys-, and His-HNE adducts (5.88 mol/mol subunit) which were recovered as recognizable products is exactly equal to the amount of protein carbonyl groups (5.78 mol/mol subunit) detected. Although possibly fortui- tous, this relationship invites speculation that the protein carbonyl groups are associated with the products formed by the Michael addition mechanism (reaction d, Fig. 4) and leads to the proposition that the amounts of amino acid residues not yet accounted for are due to secondary reactions in which the primary Michael addition products formed with Cys and His residues react further with Lys residues of the same or different subunits (reaction f , Fig. 4) to form intra- or interprotein cross-linkages, respectively. These linkages would be stabilized by the treatment with NaBH4 (reactions e and h, Fig. 4). This interpretation is consistent with the observation in Fig. 3 that the sum of the amounts of Cys and His residues

6392 HNE Modification of GAPDH OH

OH

nu . . . NaB'' '

FIG. 4. Plausible mechanisms for the modification of GAPDH by HNE. Consequences of reactions with NaBH4 (reactions 6, e, and h) or with 2,4-DNP (reactions c, i, and g). HJV refers to the c-amino group of a lysine residue in the protein. H x refers to the side chain of either His, Lys, or Cys residues. The wavy lines indicate the polypeptide chains.

0 0 "

0 1 2 4-Hydroxynonenal (mM)

FIG. 5. Introduction of carbonyl groups into GAPDH by reaction with HNE. The enzyme (1 mg/ml) was incubated with 0- 2 mM HNE in 50 mM sodium phosphate buffer (pH 7.2) for 2 h at 37 "C. The carbonyl content was determined by the procedures using 2,4-DNP (0) and [3H]NaBH4 (0).

(2.67 mol/mol subunit) that cannot be accounted for as primary Michael addition products is equal to the amount of Lys that cannot be accounted for as the simple HNE Michael addition products (2.52 mol/mol subunit).

Inter- and Intraprotein Cross-linking-If the above interpretation of our data is correct, then the treatment with HNE should lead to the production of intra- or interprotein cross- linkages. This was experimentally verified by the results summarized in Fig. 6, showing that upon SDS-polyacrylamide gel electrophoresis, the native GAPDH migrates as a single protein band of 36 kDa (lanes 1 and 7), whereas the HNE- modified enzyme migrated as slightly broad bands (lanes 2- 6). After reaction with NaBH4, the electrophoresis pattern is more complex. The HNE-modified enzyme migrated as two distinct bands, one of 36 kDa and one of slightly lower mobility (lanes 8-12), suggesting that the band of lower

M 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 KDa ~ __ . . . . ~ " - -. - " . -.

200-

116-

66-

45-

36- -7 "" - "" .... -

FIG. 6. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of GAPDH modified by HNE. The enzyme (1 mg/ml) was incubated with 0-2 mM HNE in 50 mM sodium phosphate buffer (pH 7.2) for 2 h at 37 "C. Prior to denaturation of proteins, samples in lanes 7-12 were treated with 10 pl of 10 mM EDTA, 10 pl of 1 N NaOH, and 10 pl of 0.1 M NaBH, in 0.1 N NaOH for 1 h at 37 "C. Protein bands were stained with Coomassie Brilliant Blue R-250. Lane M, marker proteins; lanes 1 and 7, 0 mM HNE; lanes 2 and 8, 0.1 mM HNE; lanes 3 and 9, 0.2 mM HNE; lanes 4 and 10, 0.4 mM HNE; lanes 5 and 11,l mM HNE lanes 6 and 12,2 mM HNE.

mobility represents a more globular form as could be produced by intramolecular cross-linkage reactions. In addition to the two protein bands of approximately 36 kDa, a less dense band of approximately 72 kDa is evident. The results suggest that without NaBH4 treatment, the Schiff base conjugation which links two different subunits together is disrupted during prep-


aration of the samples for electrophoresis and that the treatment with NaBH4 stabilized the linkage.

DISCUSSION

The high sensitivity of GAPDH to oxidative inactivation (12-14) is attributable to the oxidation of cysteine 149 which is at the catalytic site of the enzyme. This cysteine must be present in the sulfhydryl form because it participates in the formation of an S-acyl thiolester intermediate which in the catalytic mechanism undergoes phosphorolysis to form 1,3- diphosphoglyceric acid (20). In addition, it is known that histidine residue 176 and lysine residue 183 are located at the NAD+-binding site (19). Moreover, NAD+ protects Hism from photo-oxidation even though it is not directly involved in NAD+ binding or in the catalytic mechanism (25-27). It is therefore not surprising that the loss of activity associated with the binding of HNE to the protein is accompanied by the loss of His, Cys, and Lys residues (Fig. 1) and that NAD' protects 1 Cys and 2 His residues from HNE modification (Table I). However, the inability of NAD' to protect some Lys residues against the HNE modification is surprising in view of the fact that Lys'= is situated at the NAD+-binding site, and N-acetylation of Lysl= leads to a loss of catalytic activity. Moreover, Lys"' and Lys''' are both modified by pyridoxal phosphate in the presence of NAD+, and this leads to complete loss of activity. However, in the absence of NAD+, only LysIg1 is modified by pyridoxal phosphate (28). Thus, NAD+ promotes the modification of Lys'l'. Perhaps in a similar manner the apparent inability of NAD+ to protect Lys residues (Lysis) from HNE modification reflects an NAD+ provoked change in protein conformation, which leads to the exposure of an additional Lys residues (e.g. Lys''*) to HNE modification, and thus obscures its effect on Lyslm modification.

The fact that the number of protein carbonyl groups in HNE modified protein is the same whether measured by the 2,4-DNP or the 3H-labeling technique (Fig. 5) precludes reactions in which only the aldehyde moiety of HNE reacts with protein amino groups to form Schiff base adducts. There- fore, the generation of protein carbonyl groups must reflect the attachment of His, Cys, or Lys residues to the a,&double bond of HNE. This is consistent with the fact that the number of protein carbonyl groups is approximately equal to the sum of the number of His, Lys, and Cys residues that could be identified as born f ide HNE derivatives (Fig. 3). Nevertheless, a considerable fraction of the amino acid residues which were modified by HNE could not be accounted for by the procedures used. We suggest that the amino acid residues which were not recovered as identifiable products were involved in secondary reactions in which the aldehyde moieties of the primary HNE derivatives undergo secondary reactions with proximal lysine residues or with lysine residues of a different subunit to form intra- or intersubunit Schiff base cross-links (reaction f, Fig. 4). This possibility is consistent with the observations that: (a) the number of Cys and His residues that could not be accounted for as recognizable HNE adducts is equal to the number of unaccounted for Lys residues (Fig. 3); (6) both intra- and intermolecular cross-linkages are formed (Fig. 6). It is noteworthy that the sulfhydryl group of CYS"~~S readily acetylated by reaction with acetyl-P, and that the acetyl group of the thiolester derivative undergoes spon- taneous transfer to the e-amino group of LyslW to form the

N-acetyl derivative (29). Thus, in the tertiary structure, the amino group of LydW and the sulfhydryl group of Cys14' are close enough to permit direct interactions between them. This invites speculation that an intramolecular cross-link may be formed between the HNE thioether adduct of Cys14' and the amino group of Lys'=.

Many key enzymes in metabolism accumulate as catalyti- cally inactive or less active forms during animal aging (21, 23). Among other modifications, the conversion of some amino acid residues to carbonyl derivatives has been observed (30-32). There is evidence that protein carbonyl groups can be generated by metal-catalyzed oxidative reactions (21-23), or by reactions with oxygen-free radicals derived from radiol- ytic processes (24), and by oxygen-free radical-mediated gly- cation reactions (33). The data presented here and the results of earlier studies (1-7) serve notice that protein carbonyl groups may arise also as the result of interactions with lipid peroxidation products.

Acknowkdgrnent-We thank Dr. Hermann Esterbauer of the Uni- versity of Graz for the gift of HNE.

REFERENCES 1. Schauenstein, E., and Esterbauer, H. (1979) in Submolecular Biology of

Cancer, pp. 225-244, Ciba Foundation Series 67, Excerpta Medica/ Elsevier, Amsterdam

2. Esterbauer, H., Schaur, R. J., and Zollner, H. (1991) Free Rudieal Biol. Med. 11,81-128

3. Benedett, A., Comporti, M., and Esterbauer, H. (1980) Biochim. Biophys.

4. Winkler, P., Schaur, R. J., and Schauenstein, E. (1984) Biochim. Biophys.

5. Winkler, P., Lhder, W., Esterbauer, H., Schauenatein, E., and Schaur, R

6. Esterbauer, H., Zollner, H., and Lang, J. (1985) Biochem. J. 228,363-373 7. Esterbauer, H., Zollner, H., and Scholz, N. (1975) Z . Nuturforsch 30c, 466-

8. Uchida, K,, and Stadtman, E. R. (1992) Proc. Nutl. Acud. Sci. U. S. A. 8 9 ,

9. Jurgens, G., Lang, J., and Esterbauer, H. (1986) Biochim. Biophys. Acta

Acta 620, 281-296

Acta 796,226-231

J. (1984) Biochim. Biophys. Acta 796,232-237

473

5611-5615

876,103-114 10. Uchida, K., and Stadtman, E. R. (1992) FASEB J. 6,2133 (abstr.) 11. Uchida, K., and Stadtman, E. R. (1992) Proc. Natl. Acud. Sci. U. S . A. 8 9 ,

12. Brodie, A. E., and Reed, D. J. (1987) Biochem. Biophys. Res. Commun.

13. B e > , A. E., and Reed, D. J. (1990) Arch. Biochem. Biophys. 276, 212-

45444548

148 , 120-125

ala 14. Kanmawa, K., and Ashida, H. (1991) Arch. Biochem. Biophys. 288.71-78 15. Schauenstein, E., Taufer, M., Esterbauer, H., Kylianek, A., and Seelich, T. 16. Church, F. C., Porter, 0. H., Catignani, G . L., and Swaisgmd, G. L. (1985)

(1971) Manatch Chem. 102,517-529

A d . Biochem. 146.343-348 17. Jones, L. A., Holmes, J. C., &d Seligman, R. B. (1956) Anal. Biochem. 28 ,

191-198 18. 19.

20.

21. 22.

24. 23.

Laemmli, U. K. (1970) Nature 227,680-685 Buehner, M., Ford, G. C., Moras, D., Olsen, K. W., and Rossman, M. G.

Harris, J. J., and Waters, M. (1976) in The Enzymes (Boyer, P. D., ed) Vol. (1974) J. Mol. Biol. 90,25-49

Stadtman, E. R. (1991) Biochemistry 29,6323-6331 XIII, 3rd ed., pp. 1-49, Academic Press, New York

Stadtman, E. R. (1990) Free Radical BioL Med. 9,315-525 Stadtman, E. R., and Oliver, C. N. (1991) J. Biol. Chem. 266,2005-2008 Garrison, W. M., Jayko, M. E., and Bennett, W. (1962) Rudiot. Res. 16,

m3-m 25. Bond, J. C., Francis, S. H., and Park, J. H. (1970) J. Biol. Chem. 245,

26. Francis, s. H., Meriwether, B. P., and Park, J. H. (1973) Biochemistry 12,

27. Francis, S. H., Meriwether, B. P., and Park, J. H. (1973) Biochemistry 12,

28. Forc,ina, B. G., Ferri, G., Zapponi, M. G., and Ronchi, S. (1971) Eur~ J.

29. Mathew, E., Meriwether, B. P., and Park, J. H. (1967) J. Bwl. Chem. 2 4 2 ,

30. Farber, J. M., and Levine, R. L. (1986) J. Biol. Chem. 261 4574-4578 31. Amici, A., Levine, R. L., Tsai, L., and Stadtman, E. R. (1986) J . Biol. Chem

32. Climent, I., Tsai, L., and Levine, R. L. (1989) A d Biochem. 182, 226-

_" "_ 1041-1053

346-355

355-359

Blochem. 20,535-540

5024-5033

264,3341-3346

2.12 33. DuY; J. A. Ahmed M. V., Murtiashaw, M. H. Richardson J. M. Wdla

M. D., Tiorpe, S.'R., and Baynes. J. W. (1996) Biochemistry ZS,'109&d 10970

covalent attachment of 4-hydroxynonenal to glyceraldehyde-3

Documents