Proc. Natl Acad. Sci. USA Vol. 79, pp. 2460-2464, April 1982 Biochemistry Methylation at D-aspartyl residues in erythrocytes: Possible step in the repair of aged membrane proteins (protein methylation/amino acid racemization/aging/erythrocytes) PHILIP N. MCFADDEN AND STEVEN CLARKE Department of Chemistry and Biochemistry and the Molecular Biology Institute, University of California, Los Angeles, California 90024 Communicated by Paul D. Boyer, December 23, 1981 ABSTRACT Reversibly methylated aspartyl residues in hu- rate of 0.14% per yr (11). It seemed conceivable to us that sim- man erythrocyte membrane proteins are shown to be in the "un- ilar amino acid racemization might be occurring in aging eryth- natural" D configuration. This is demonstrated by treatment of rocytes, providing the substrate for the methylation ofaspartyl proteolytically derived aspartic acid fi-[3H]methyl ester with L- residues in erythrocyte proteins. and D-amino-acid oxidases and by the resolution ofdiastereomeric We report here that all of the methylated aspartic acid that L-leucyl dipeptides containing either L- or D-aspartic acid j3- we have isolated from erythrocyte membrane and cytoskeletal methyl ester by ion-exchange chromatography. Based on this ob- proteins has the uncommon D configuration. We propose that servation, we propose a novel role for eukaryotic protein carboxyl acid residues can be methyltransferases (protein 0-methyltransferase; S-adenosyl-L- D-aspartic enzymatically recognized and methionine:protein O-methyltransferase, EC 2.1.1.24). We sug- modified in a step leading to the repair ofaging proteins. This gest that these widely distributed enzymes function to recognize process may help maintain functional proteins at a metabolic aspartyl residues that have racemized spontaneously for a sub- cost that is low compared to that ofreplacing damaged proteins sequent repair reaction. This repair function is postulated to cou- by de novo translation. ple ester hydrolysis with the restoration of the original L config- We suspect that this postulated protein repair mechanism is uration of the aspartyl residue. It is possible that similar types of widely distributed because protein carboxyl methylation occurs racemization repair processes may occur by reversible covalent in all mammalian tissues examined (1, 12). Other varieties of modifications at other residues. Other possible functions for D- reversible protein modification reactions, many ofwhich have aspartic acid fl-methyl ester residues in proteins are considered. not been assigned a function, may also be involved in the repair of racemized or otherwise altered proteins. A widespread reaction in nature is the posttranslational modi- fication of protein carboxyl groups by methyl ester formation MATERIALS AND METHODS (1). At least two classes of the associated enzyme, protein car- Materials. L-Aspartic acid ,B-methyl ester hydrochloride was boxyl methyltransferase (protein O-methyltransferase; S-ad- purchased from Vega Biochemicals (Tucson, AZ). D-Aspartic enosyl-L-methionine:protein 0-methyltransferase, EC 2.1.1.24), acid 3methyl ester hydrochloride was synthesized by the are known to exist (2). The first of these is a bacterial enzyme method ofde Groot and Lichtenstein (13) from 5 g ofD-aspartic that catalyzes the methylation of chemoreceptors at glutamyl acid (37.6 mmol; Sigma) dissolved in a mixture of 38 ml of an- residues in an adaptive response to sensory stimuli (2, 3). The hydrous methanol (940 mmol) and 5.4 ml ofacetyl chloride (75 second class ofenzyme activity, found in both prokaryotes and mmol). After recrystallization from diethylether/methanol, 2:1 eukaryotes, shows a much broader substrate specificity and (vol/vol), the product [1.98 g (10.8 mmol); 29% yield] was char- catalyzes the formation ofextremely labile methyl esters, pre- acterized by its melting point (186.5-189.5°C), titrimetric be- sumably at aspartyl residues (4). The function ofthis latter pro- havior (1 equivalent of a group with a pKa of 2.2 and 1 equiva- cess is not clear. lent ofa group with a pKa of7.6), hydrolysis rate [tl/2 = 89 min We are interested in determining the function ofmammalian at pH 10.5 and 37°C; the literature value for L-aspartic acid /3- protein carboxyl methylation and are studying this reaction in methyl ester is 82 min (6)], and migration in thin-layer chro- the human erythrocyte. We have shown that specific cytoskel- matography (cf. ref. 4). Amino acid analysis of the product on etal and membrane proteins are reversibly methylated at as- a Beckman model 120C analyzer revealed up to 10% impurity partyl residues (4-6). The carboxyl methylation of these pro- of aspartic acid. A ninhydrin color constant of 0.20 relative to teins is substoichiometric (less than 0.02 methyl groups per aspartic acid was calculated for the ester. polypeptide chain) in all cases (5, 6), but we have determined Both the L- and D-aspartic acid /8-methyl esters were >99% that this level is consistently higher in older populations of optically pure as determined by the diastereomer method of erythrocytes (7). Similarly, experiments with the purified car- Manning and Moore (14); the contaminating aspartic acid had boxyl methyltransferase from both erythrocytes and other mam- the same stereoconfiguration as the parent ester. malian tissues indicate that in vitro substrates are also substoi- L-Proline was from Sigma. Liquid scintillation cocktail chiometrically methylated (8, 9). To understand why aspartyl (Aquamix) was from West Chem (San Diego, CA), and 7-10 residues at a given position in a sequence may be only partially volumes were used per 1 volume of aqueous sample. methylated, we have investigated the nature ofthe aspartic acid Isolation ofAspartic Acid fl-[3H]Methyl Ester from Human residues that are methylated in the red cell. Erythrocyte Membranes. Membranes containing [3H]methyl It has been shown that racemization ofaspartic acid residues groups (referred to hereafter as *membranes) were prepared occurs in aging mammalian proteins (for a review, see ref. 10). from fresh erythrocytes incubated with L-[methyl-3H]methionine For example, D-aspartic acid accumulates in lens proteins at a (70-90 Ci/mmol; 1 Ci = 3.7 X 1010 becquerels) as described by Freitag and Clarke (5). These *membranes (5 mg ofprotein The publication costs ofthis article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertise- Abbreviations: LeuCA, L-leucine N-carboxyanhydride; *membranes, ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. membranes containing tritiated methyl groups. 2460 Downloaded by guest on September 27, 2021 Biochemistry: McFadden and Clarke Proc. Natl. Acad. Sci. USA 79 (1982) 2461 per ml; 160 pmol of [3H]methyl groups per ml) were digested of erythrocyte *membrane proteins (4). The radioactive yield for 16-25 hr at 370C in 2-3 volumes of a Sigma preparation of of aspartic acid 83-[3H]methyl ester from the *membranes in bakers' yeast carboxypeptidase Y (2 mg/ml; 100 units/mg of these experiments was 5-10%; much of the remaining radio- protein) containing 8 mg ofcitrate buffer (pH 5) per ml. In some activity chromatographed as [3H]methanol and may have been experiments, radiolabeled aspartic acid /B-methyl ester was iso- formed by ester hydrolysis during the digestion step. lated by ion-exchange chromatography, desalted by Sephadex Aspartic acid /3-[3H]methyl ester in unfractionated digests G-15 gel filtration in 0.1 M acetic acid, and concentrated by was treated with L-amino-acid oxidase. The radioactive methyl lyophilization as described by Janson and Clarke (4). ester was unaffected by the oxidase, even though a standard of L-Amino-Acid Oxidase Treatment. Carboxypeptidase Y di- L-aspartic acid ,3methyl ester added to the mixture was com- gestion of*membranes (0.2 mg of*membrane protein) was ter- pletely degraded (Fig. 1). L-Aspartic acid, which also was pres- minated after 16 hr by the addition of0.6 mg ofphenylmethyl- ent in the mixture, was a poor substrate for the L-oxidase (17) sulfonyl fluoride (Sigma). The digestion products were mixed and provided an internal standard for the ninhydrin analysis. with 4 gnmol of either L- or D-aspartic acid /3-methyl ester and This experiment also was performed with isolated aspartic acid 100 ,Amol of Tris HCl (pH 7.5) in a final volume of 0.34 ml. -P[3H]methyl ester mixed with an internal standard of D-as- Aliquots (150 ,tl) were incubated with either 20 1.l of water or partic acid /3-methyl ester. In this case, the L-amino-acid oxi- 20 A.l of solution containing 1.6 units of L-amino-acid oxidase dase was inactive toward both the radioactive aspartic acid /3- (from Crotalus adamanteus venom, Sigma type IV; 8.1 units/ methyl ester and the standard D-aspartic acid /3-methyl ester mg of protein). These samples were incubated at 370C for 3.5 (data not shown). Overall, the results indicate that L-amino-acid hr and then were quenched with 10 Al of 1 M HC1 and 220 IlI oxidase specifically oxidizes standards of L- and not D-aspartic of sodium citrate analyzer sample buffer (pH 2.2; Pierce). Por- acid /3-methyl ester and that this enzyme does not detectably tions (350 ,l) were chromatographed on a 0.9 cm X 30 cm col- oxidize erythrocyte-derived aspartic acid /3-[3H]methyl ester. umn of sulfonated polystyrene analyzer resin (Dionex DC-6A) Erythrocyte-Derived Aspartic Acid f3-[3H]Methyl Ester Is at 50°C in citrate (pH 3.25; 0.2 M in Na+). The column was a Substrate for D-Amino-Acid Oxidase. Radioactive aspartic eluted at a flow rate of 70 ml/hr with this buffer, followed by acid P-methyl ester isolated from membrane digests by ion-ex- elution with 0.2 M NaOH.