Glutamic Acid As a Precursor to N-Terminal Pyroglutamic Acid In

Proc. Nat. Acad. Sci. USA Vol. 69, No. 1, pp. 274-277, January 1972

Glutamic Acid as a Precursor to N-Terminal Pyroglutamic Acid in Mouse Plasmacytoma Protein (protein synthesis/initiation/immunoglobulins/pyrrolidone carboxylic acid) DANIEL R. TWARDZIK AND ALAN PETERKOFSKY Laboratory of Biochemical Genetics, National Heart and Lung Institute, National Institutes of Health, Bethesda, Maryland 20014 Communicated by Marshall Nirenberg, November 8, 1971

ABSTRACT Cell suspensions derived from a mouse (w/v = 1/5) was filtered through a Mylar screen (mesh size plasmacytoma (RPC-20) that secretes an immunoglobulin centrifugation then light chain containing N-terminal pyroglutamic acid can 10.5 mm), sedimented by low-speed synthesize protein in vitro. Chromatographic examination washed twice in chilled Dulbecco's phosphate-buffered of an enzymatic digest of protein labeled with glutamic saline (pH 7.2). Cells were resuspended in Eagle's minimum acid shows only labeled glutamic acid and pyroglutamic essential medium (MEM) (14), without glutamine (cells acid; hydrolysis of protein from cells labeled with gluta- from 1 g of tumor/10 ml of media) and the suspension was mine, however, yields substantial amounts of glutamic acid in addition to glutamine and pyroglutamic acid. The equilibrated with 5% C02-95% air. Incubation at 37"C was absence of glutamine synthetase and presence of glutami- initiated by the addition of radioactive amino acids. nase in plasmacytoma homogenates is consistent with these findings. These data indicate that N-terminal pyro- Enzyme Assays. Preparation of homogenates and reaction glutamic acid can be derived from glutamic acid without mixtures for the radiochemical assay of glutamine synthetase prior conversion of glutamic acid to glutamine. Since free (EC 6.3.1.2) were as described by Lund (15). Glutaminase or bound forms of glutamine cyclize nonezymatically to (EC 3.5.1.2) activity was measured by a modification of the pyroglutamate with ease, while glutamic acid does not, as the data suggest that N-terminal pyroglutamic acid for- glutamine synthetase assay, with radioactive glutamine mation from glutamic acid is enzymatic rather than spon- the substrate. The loss of radioactivity in the effluent from a taneous. Dowex-1 column was a measure of the enzyme activity. Pyroglutamic acid (pyrrolidone carboxylic acid; pGlu) is RESULTS found as the N-terminal amino acid in various proteins and Amino acid incorporation by cell suspensions of polypeptides. It has been demonstrated in both immuno- mouse plasmacytoma globulin light and heavy chains (1), fibrinogen (2), and Cell suspensions of mouse plasmacytoma, when incubated in collagen (3), as well as in kinins and venoms (4). Although Eagle's minimum essential medium, actively incorporate pGlu has a blocked amino group and a structure about its radioactive amino acids into protein; maximum incorporation a-carbon atom similar to that of N-formylmethionine, there is reached at about 30 min (Fig. 1). Substantial incorporation is no convincing evidence implicating it in the initiation was observed for glutamine, glutamic acid, leucine, histidine, process. Since pGlu cannot be acylated directly to transfer and methionine. The great difference in the extent of in- RNA (5, 6), attempts have been made to demonstrate a corporation of glutamine compared to leucine and glutamic mechanism for pGlu formation in proteins similar to the acid, as shown on the semilogarithmic plot, probably is due to formylation of methionine on fMet-tRNA (7, 8). Preliminary differences in uptake rates or pool sizes in the cells. The data studies demonstrating that pGlu is formed from glutaminyl- of Fig. 2 support this idea. As shown in the bar graph, the tRNA (9) or from N-terminal glutamine (10) are available, addition of unlabeled glutamine to cell suspensions stimulated but the spontaneous cyclization of glutamine to pGlu (11) the incorporation of glutamic acid and leucine. This stimula- makes the interpretation of these data difficult. This tendency tion was not observed for the other amino acids tested. This of glutamine and N-terminal glutaminyl peptides to form is in keeping with the greater requirement for glutamine than pGlu under mild conditions has introduced some reservations for other amino acids for the growth of L cells and HeLa as to whether pGlu is present in native protein or is derived cells in tissue culture (16). from the spontaneous cyclization of glutamine (11). We report here experiments that identify the precursor of Identification of the precursor of pGlu in N-terminal pGlu in plasmacytoma protein. A preliminary plasmacytoma protein report of this work has been published (12). A cell suspension was prepared in minimal essential medium minus glutamine. One 10-ml incubation contained 54 nmol METHODS (100 ,Ci) of [3H]glutamic acid, in the presence of an equimolar Preparation of Plasmacytoma Cell Suspensions. Tumors amount of cold glutamine; another incubation contained (RPC-20) were transplanted subcutaneously in female, 54 nmol of ['4C]glutamine (10 ,uCi), in the presence of an inbred BALB/c mice. Freshly excised tumors were washed equimolar amount of cold glutamic acid. After 90 min of and minced in a solution of 5% Puck's D1 (13) saline con- incubation at 37°C, the cell suspensions were frozen and taining 2% sucrose and 0.1% glucose. The cell suspension thawed to break the cells, concentrated by lyophilization, 274 Downloaded by guest on October 6, 2021 Proc. Nat. Acad. Sci. USA 69 (1972) Glutamic Acid as N-Terminal Pyroglutamate Precursor 275

dialyzed extensively against H20, and again lyophilized. Glu + - - - -- The relyophilized material was then suspended in 2 ml of Gi- I+ 0.05 M Tris- HC1 (pH 8.0) and digested with Pronase (Cal- + - -- biochem, 100 mg, 4500 proteolytic units) for 24 hr at 370C. Any remaining insoluble material was removed by centrifugation. A preliminary analysis of protein hydrolyzates from each

z cell incubation was made on Dowex-50 columns. pGlu, which 0 has a blocked amino group, is not retained on the negatively-

0 charged resin and is found in the effluent. About 10% of the a-ac: radioactivity derived from protein labeled with either gluta- 0 mine or glutamic acid was found in the effluent (Table 1). z The mobility on high-voltage electrophoresis [0.05 M potassium formate (pH 4.5)] of the radioactive material in column effluents was identical with that of standard pGlu. The protein hydrolyzates described above were also examined by ascending paper chromatography in phenol- FIG. 2. The bar graph schematizes, from left to right, the water 88:12 to identify the amino acids labeled by ['H]glutam- incorporation of glutamic acid, glutamine, and leucine into hot- acid-precipitable material. Experimental details were as in Fig. 1. The 100% value for incorporation represents the maximum 1000 ,- value of incorporation achieved in 90 min for each radioactive GMn amino acid in a complete incubation medium as described in Fig. 1. Where indicated by a minus sign, the incubation medium was deficient in that amino acid. Samples were assayed for incorporation of radioactivity into protein as in Fig. 1.

ic acid or ['4C]glutamine. Chromatography of the hydrolyzate from cells labeled with glutamic acid (Fig. 3, upper panel) demonstrated that only glutamic acid and pGlu were labeled. There was no evidence of conversion of glutamic acid!to glutamine. In contrast, an examination of the protein hydroly- 100 zate from cells labeled with glutamine (Fig. 3, lower panel) showed substantial amounts of labeled glutamic acid and glutamine, in addition to pGlu. We suggest therefore that the 0 pGlu in protein from glutamate-labeled cells is derived from glutamic acid without the intervention of a glutamine 00 intermediate. Enzyme activities in plasmacytoma and liver homogenates The above hypothesis is supported by a survey of glutamine synthetase and glutaminase activity in plasmacytoma cell glutamin anLecneepeprd aho heeu0m n extracts. The results of Fig. 4 (upper panel) indicate that glutamine synthetase activity was not detectable in plasmacytoma homogenates, whereas substantial activity was seen in mouse liver. On the other hand, a very active glutaminase activity (Fig. 4, lower panel) was observed in plasmacytoma

TABLE 1. Analysis of protein digests on Dowex-50 columns

cpm In % In Label Applied effluent effluent cubation mixtures were supplemented with 54 nmol of glutamic acid, glutamine, and leucine. One incubation contained [.14C]- [14C]Glutamine 6770 735 10.8 glutamine (10,uCi), another [14C]glutamic acid (13.7 ACi), and ['HIGlutamic acid 4855 413 8.5 the third ['H]leucine (100,uCi). The mixtures were incubated at 37°C in screw-capped tubes. 1-ml samples were removed at the Protein labeled with ['4C~glutamine or ['HJglutamic acid was time indicated, precipitated by the addition of 1 ml of 10% hydrolyzed with Pronase. 0.5-ml samples of the hydrolyzate were CL6CCOOH, and then heated in a boiling-water bath for 5 min. applied to columns (0.5 X 4 cm) of Dowex-50W X 2 (hydrogen The precipitates were trapped on glass-fiber filters and washed form, 200-400 mesh). The columns were washed with 1.0 ml of with 5% CL6CCOOH. Radioactivity on the filters was deter- HO. The combined effluent and washes were counted with 10 ml minedMA byver scnilto c with 1 of Triton-toluene-Liquifluor (New England Nuclear Co.) count- Liquifluor 6:12: 1. ing fluid. Downloaded by guest on October 6, 2021 276 Biochemistry:Twardzik and Peterkofsky Proc. Nat. Acad. Sci. USA 69 (1972)

*Z 14 ~~~~~~~~Liver 12

E cL 0. 4 2 Plasmacytoma

@ 14 :° : Plasmacytoma .i2- 0

10 Inches migrated 8 _ FIG. 3. Ascending paper chromatography in 88% phenol- E 6 12% H20 of radioactive protein hydrolyzate from cells labeled with [3H]glutamic acid (upper panel) or [14C]glutamine (lower 4

panel). The relative mobilities of glutamic acid, glutamine, and ~~~~~~~Liver pGlu in this solvent are indicated in the upper panel. 5 10 20 40 MINUTES extracts, while liver extracts showed undetectable activities FIG. 4. Homogenates from liver and plasmacytomas were pre- of this enzyme. pared in 4 volumes of 0.05 M Tris- HCl, (pH 7.5)-5 mM EDTA- DISCUSSION 0.9% NaCl. Cellular debris was removed by centrifugation at 10,000 X g, and the supernatant was used for both the glutamine This study has focused attention on two interrelated ques- synthetase and glutaminase assay. Incubation mixtures for the tions: whether glutamic acid or glutamine is the more direct glutamine synthetase assay (top) were as described by Lund precursor of pGlu in protein, and whether N-terminal pGlu (15); glutaminase was assayed (bottom) in 0.05 M Tris-HCl (pH formation in immunoglobulins is a biological or artifactual 8.5). Reaction mixtures from both assays were equilibrated to process. A metabolic distinction between glutamic acid and 370C and reactions were started by addition of the crude homo- glutamine is difficult in many tissues because of the enzymatic genate. Samples were withdrawn at various times and pipetted interconversion of the two amino acids catalyzed by glutamine into 0.25 ml of cold 95% ethanol; the precipitate was removed by synthetase or glutaminase. Thus, while this study might centrifugation and the supernatants were assayed on Dowex-50 have been complicated in other tissues, the unique properties columns. of the plasmacytoma extracts allowed us to obtain interpret- able results. The absence of glutamine synthetase (Fig. 4) different possible interpretations of our precursor studies. If prevents glutamine formation from glutamic acid (Fig. 3, we had found that glutamine was the precursor to pGlu, we upper panel), although the glutaminase present does not would be uncertain whether the isolated pGlu had been suppress glutamic acid formation from glutamine. These formed by an enzymatic or nonenzymatic process. Since the data are consistent with previous observations that mam- data suggested that pGlu was derived from glutamic acid, malian cells in culture have a glutamine requirement that the interpretation was greatly simplified. We can, therefore, cannot be replaced by glutamic acid (16). assume that an enzymatic process is necessary for the cycliza- The artifactual formation of pGlu occurs primarily with tion of glutamic acid to pGlu. glutamine, and not with glutamic acid. Table 2 emphasizes Fig. 5 outlines two possible models for the conversion of this distinction between glutamic acid and glutamine. glutamic acid to pGlu. In the first mechanism, analogous to Neither glutamic acid nor glutamyl-tRNA formed significant the formation of formylmethionyl-tRNA in bacterial systems, amounts of pGlu after incubation at pH 8.0 and 370C. On glutamyl-tRNA is cyclized to pyroglutamyl-tRNA. An the other hand, about 8% of the added glutamine was ATP-dependent phosphorylation of the y-carboxyl of the recovered as pGlu after 18 hr under these conditions. The nonenzymatic cyclization of glutamine probably explains Glutamic Acid the finding (Table 1) that pGlu was recovered from hydrolyzates of glutamine-labeled protein (10% after 24 hr at pH 8.0). When the a-carboxyl group of glutamine was esterified .RNA Pyroglutomyl- (glutaminyl-tRNA), the nonenzymatic cyclization was I (i)Protein accelerated about 4-fold. This distinction in the chemical Glutamyl-Protein * stabilities of glutamic acid and glutamine led us to anticipate FIG. 5. Mechanisms for pGlu formation from glutamic acid. Downloaded by guest on October 6, 2021 Proc. Nat. Acad. Sci. USA 69 (1972) Glutamic Acid as N-Terminal Pyroglutamate Precursor 277

TABLE 2. Test of stability of amino acids and aminoacyl-tRNA reactions (17). Pyroglutamyl-tRNA, as an initiator tRNA, might respond to specific codons and initiation factors. The Incubation second suggested mechanism involves the post-translational Labeled compound time (hr) % pGlu conversion of protein-bound N-terminal glutamic acid to [3H]Glutamic acid 0 0.1 N-terminal pGlu. This mechanism does not dictate that the P3H]Glu1amic acid 1 0.5 protein be initiated by glutamic acid. pGlu might be formed [3H] Glutamic acid 3 0.1 in conjunction with the proteolytic cleavage of an adjacent [3H]Glutamic acid 6 0.4 amino acid or peptide. Further experiments designed to [3H]Glutamic acid 18 0.2 discriminate between these alternative mechanisms are in [3H]Glutamyl-tRNA 1 0.1 progress. [14C]Glutamine 0 0.5 [14C] Glutamine 1 3.1 Dr. Ettore Appella kindly provided an initial transplant of the [14C]Glutamine 3 4.0 mouse plasmacytoma. The conditions for preparing dispersed cell ['4C]Glutamine 6 suspensions for amino acid-incorporation studies were developed 8.2 in this laboratory by Dr. George Jones. Mrs. Helmi Carpenter [14C]Glutamine 18 8.5 assisted in the preparation of the manuscript. [14C]Glutaminyl-tRNA 1 12.7 1. Dayhoff, M. 0. (1969) in Atlas of Protein Sequence and The indicated compounds were incubated at Structure (National Biomedical Research Foundation Silver 370C in 0.05 M Spring, Md.), Vol. 4, pp. D73-D1O. Tris.HCl, pH 8.0, containing 5 mg/ml of Pronase, for the indi- 2. Dayhoff, M. 0. (1969) in Atlas of Protein Sequence and cated times. 0.05-ml Aliquots of the incubation contained: Structure (National Biomedical Research Foundation, [3H]glutamic acid, 1.3 X 106 cpm; ['H]glutamyl-tRNA, 1.08 X Silver Spring, Md.), Vol. 4, p. D68. 10' cpm; ['4C]glutamine, 5.5 X 106 cpm; [14C]glutaminyl-tRNA, 3. Dayhoff, M. 0. (1969) in Atlas of Protein Sequence and 8.6 X 104 cpm. 0.05-ml Aliquots of the incubations were diluted Structure (National Biomedical Research Foundation, to 1.0 ml with H20 and passed through Dowex-50 columns as in Silver Spring, Md.), Vol. 4, p. D180. Table 1. 4. Dayhoff, M. 0. (1969) in Atlas of Protein Sequence and Aminoacyl-tRNA preparations were made by incubation of Structure (National Biomedical Research Foundation, Escherichia coli B tRNA with the designated amino acids and E. Silver Spring, Md.), Vol. 4, pp. D-166-D172. 5. Moav, G. & Harris, T. N. (1967) Biochem. Biophys. Res. coli B aminoacyl-tRNA synthetase preparations. Incubation Commun. 29, 773-776. mixtures for aminoacylation contained: 50 mM potassium caco- 6. Rush, E. & Starr, J. L. (1970) Biochim. Biophys. Acta 199, dylate buffer (pH 7.2); 10 mM MgCl2; 5 mM ATP; 2 mM phos- 41-55. phenolpyruvate; 50 Axo/ml E. coli B tRNA (Schwarz Bio- 7. Adams, J. M. & Capecchi, M. R. (1966) Proc. Nat. Acad. Research); 0.1 mg (43 units)/ml pyruvate kinase (Sigma); about Sci. USA 55, 147-155. 0.15 mg/ml aminoacyl-tRNA synthetase preparation, sufficient 8. Webster, R. E., Engelhardt, D. L. & Zinder, N. D. (1966) to completely acylate the tRNA with glutamine or glutamic acid Proc. Nat. Acad. Sci. USA 55, 155-161. in 15 min. For acylation with glutamine, the mixture included 9. Bernfield, M. R. & Nestor, L. (1969) Fed. Proc. 28, 837. 10. Baglioni, C. (1970) Biochem. Biophys. Res. Commun. 38, [14C]glutamine (187 Ci/mol), 1.25 MCi/ml. For acylation with 212-219. glutamic acid, the incubations contained ['H]glutamic acid (15 11. Blomback, B. (1967) in Methods in Enzymology, ed. Hirs, Ci/,umol), 5,Ci/ml. After incubation for 15 min, the mixture was C. H. W. (Academic Press, New York and London), Vol. applied to a column (1.3 X 4 cm) of DEAE-cellulose, and the XI, pp. 398-411. column was washed with about 25 ml of 0.3 M NaCl containing 12. Twardzik, D. R., Appella, E. & Peterkofsky, A. (1971) Fed. 0.1 M sodium acetate (pH 4.5)-10 mM MgCl,-1 mM EDTA. Proc. 30, 1216. The bound tRNA was then eluted with 10 ml of 1 M NaCl con- 13. Ham, R. G. & Puck, T. T. (1962) in Methods in Enzymology, taining the same additions as the 0.3 M NaCl buffer. tRNA was ed. Colowick, S. P. & Kaplan, N. 0. (Academic Press, New precipitated by the addition of 3 volumes of cold ethanol. The York and London), Vol. V, pp. 90-119. 14. Eagle, H. (1959) Science 130, 432-437. precipitate was dissolved in 0.05 M Tris-HCl (pH 8.0). 15. Lund, P. (1970) Biochem. J. 118, 35-39. 16. Eagle, H. (1955) Science 122, 501-504. 17. Meister, A. (1962) in The Enzymes, ed. Boyer, P. D., Lardy, esterified glutamic acid, followed by either a spontaneous or H. & Myrback, K. (Academic Press, New York and Lon- enzymatic cyclization, would havc a precedent in similar don), Vol. 6, 2nd Ed., p. 247. Downloaded by guest on October 6, 2021