An efficient method for enrichment of histidyl-tRNA synthetase (Jo-1 antigen) from HeLa cells

Journal of Immunological Methods, 98 (1987) 235-241

235

Elsevier JIM 04296

An efficient method for enrichment of histidyl-tRNA synthetase (Jo-1 antigen) from HeLa cells Tapas Biswas, Frederick W. Miller, Sharon A. Twitty and Paul H. Plotz Arthritis and Rheumatism Branch, National Institute of Arthritis, Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD 20892, U.S.A.

(Received 22 September 1986, revised received 15 December 1986, accepted 17 December 1986)

A rapid method has been developed for enrichment of Jo-1 antigen (histidyl-tRNA synthetase) from HeLa cells. The enzyme has been prepared from post-ribosomal supernatant by successive chromatography with Blue Sepharose and Poly-U-Sepharose, followed by DEAE-high performance liquid chromatography (HPLC). By this method, enzyme could be obtained within 4 days of HeLa cell harvesting, with 40% recovery of the enzymatic activity. The apparent native molecular size of the enzyme as determined by HPLC-size exclusion column chromatography was approximately 120 kDa. Under denaturing conditions using SDS-polyacrylamide gel electrophoresis the enzyme subunit size was approximately 55 kDa. The antigen preparation, although not homogeneous, was found to react only with anti-Jo-1 positive antisera when tested by immunoblotting with many patient sera of defined autoantibody specificities, making the preparation useful for immunologic studies of anti-Jo-1 antibodies. Key words." HeLa cell; Histidyl-tRNA synthetase; Aminoacyl-tRNA synthetase; Myositis; Jo-1 antigen; La antigen; Chromatogra-

phy, high performance liquid; Immunoblotting

Introduction

In some patients with idiopathic inflammatory myopathies, antibodies have been described which react with an extractable cellular antigen, originally named Jo-1 (Nishikai and Reichlin, 1980; Arnett et al., 1981; Yoshida et al., 1983). The antigen has been identified as the enzyme histidyl-tRNA synthetase (EC 6.1.1.21), which Correspondence to: T. Biswas, Bldg. 10, Rm. 9N240, National Institutes of Health, Bethesda, MD 20892, U.S.A. Abbreviations: DEAE, diethylaminoethyl; MEM, minimal essential medium; PMSF, phenylmethylsulfonyl fluoride; 5'ATP, 5'-adenosine triphosphate; DTT, dithiothreitol; TCA, trichloroacetic acid; SDS, sodium dodecyl sulfate; PBS, phosphate-buffered saline (pH 7.4); DMSO, dimethyl sulfoxide; tRNA, transfer RNA; BSA, bovine serum albumin; HPLC, high performance liquid chromatography.

catalyzes the esterification of histidine to its cognate tRNA (Mathews and Bernstein, 1983). This enzyme, one of the aminoacyl-tRNA synthetases, has previously been biochemically purified from several sources including mammalian liver (Walker et al., 1983; Yang et al., 1984) and reticulocytes (Kane et al., 1978) and can be purified with the aid of immunoaffinity chromatography (Yang et al., 1984). The biochemical methods for purification, however, are long and tedious and yield little enzyme, and immunoaffinity chromatography purification depends upon the availability of monospecific antisera, which are rare. We have developed, therefore, a rapid biochemical purification procedure for the enzyme with high yield. In order to pursue biochemical and immunologic studies of this clinically important antigen from a human source we have used HeLa cells. Although

236 the enzyme has not been purified to homogeneity, upon immunoblotting with a variety of antisera of defined specificity from patients with rheumatic diseases, the preparation reacted only with those sera containing anti-Jo-1 antibodies.

HeLa cell harvesting Suspension cultures of HeLa cells were centrifuged in a refrigerated centrifuge with a swinging bucket rotor at 655 × g for 10 rain. Cell pellets were pooled and washed twice with PBS before lysis.

Materials and methods

HeLa cell lysate Packed cells were resuspended in 2.5 vol. of cold 10 m M Tris-HC1, 1.5 mM MgC12, 15 mM KCI, 5 mM D T T and 1 mM PMSF (pH 7.4). After 10 rain on ice, cells were disrupted with 5-6 strokes of a Dounce homogenizer and the NaCl concentration was immediately adjusted to 0.14 M with addition of 5 M NaC1.

Eagle's MEM with Earle's salts, penicillin/ streptomycin, L-glutamine, horse serum and goat serum were from Gibco Laboratories, Grand Island, NY. PMSF, 5'-ATP, D T T and calf liver tRNA were from Boehringer Mannheim Biochemicals, Indianapolis, IN. BSA was from Sigma Chemical Co., St. Louis, MO. RNasin was from Promega Biotec, Madison, WI. L-[3H]histidine (specific activity 54.4 C i / m m o l ) was from New England Nuclear, Boston, MA. Blue Sepharose CL-6B and Poly-U-Sepharose 4B were from Pharmacia, Piscataway, NJ. Peroxidase-conjugated affinity-purified goat IgG anti-human IgG and IgM were from Jackson ImmunoResearch Lab., Avondale, PA. All other reagents were the best analytical grade available. For ion exchange chromatography, an HPLC Spherogel-TSK DEAE 5 PW column (7.5 mm × 7.5 cm) from Altex Scientific Operations, Beckman Instruments, Berkeley, CA, was used, and for HPLC size exclusion chromatography the Dupont Zorbax GF-250 column (4 mm x 25 cm) from Du Pont Instrument Systems, Wilmington, DE, was used on a Beckman model 421 HPLC system. Analytical paper discs (0.5 in) were from Schleicher and Schuell, Keene, NH, and 24-well microtiter plates were from Costar, Cambridge, MA. Sera of defined specificity were kind gifts from Dr. M. Reichlin and Dr. I. Targoff of the Oklahoma Medical Research Foundation and from the Centers for Disease Control.

HeLa spinner culture HeLa cells were grown and maintained at 2.5-4.0 x 105 cells/ml in suspension cultures at 37 ° C in Eagles MEM with Earle's salts containing 5% h o r s e s e r u m , 1% ( v / v ) stock penicillin/streptomycin and 586 rag/1 L-glutamine. Cells were grown to 20 liters at 6-7 × 105 cells/ml before harvesting. All further steps were performed at 4 ° C.

Post-mitcrchondrial supernatant The cell lysate was centrifuged at 8000 x g for 10 rain and the supernatants obtained were pooled for further purification. Post-ribosomal supernatant The post-mitochondrial supernatant was centrifuged at 100000 x g for 90 min and the supernatant obtained was either immediately processed further or frozen at - 80 o C. Affinity chromatography on Blue Sepharose CL-6B The post-ribosomal supernatant, usually about 200 ml, was passed through 2 vols. of Blue Sepharose CL-6B, which had been swollen, washed and equilibrated with 50 mM Tris-HC1, 5 mM MgCI 2, 5 m M D T T and 1 mM PMSF, pH 7.5 (buffer A). The effluent was recycled twice through the column and the gel was washed with 10 bed volumes of buffer A. The enzyme preparation was then eluted with 50 m M 5'-ATP in buffer A using a volume equal to the volume of post-ribosomal supernatant originally applied. Affinity chromatography on Poly-U-Sepharose 4B The Blue Sepharose eluent was dialyzed for 12 h with three changes of 2 liters of 25 mM Tris-HC1, 3 mM MgC12, 0.1 mM EDTA, 0.5 mM D T T and 0.1 mM PMSF, pH 8.0. After dialysis, DTT to a final concentration of 5 mM and RNasin to a concentration of 0.06 U//~I were added to the eluent which was then applied to a Poly-U-Sepharose 4B column (0.25 v / v post-ribosomal super-

237 natant), which had been swollen, washed and equilibrated with buffer A. The effluent was recycled once over the column, and the column washed with six bed volumes of buffer A, prior to elution with two bed volumes of 200 mM MgC12 in buffer A at 0.9 ml/min. The Poly-U-Sepharose eluent was then dialysed overnight with three changes of 3 liters of 50 mM Tris-acetate, 5 mM Mg acetate, 5 m M DTT, 1 mM PMSF, p H 7.5 (buffer B).

Ion exchange DEAE-high performance liquid chromatography After dialysis the eluent obtained from the Poly-U-Sepharose 4B column was applied to the Spherogel T S K - D E A E 5PW HPLC column equilibrated with buffer B. After washing the c o l u m n extensively with buffer B, a 0-300 mM linear K H 2 P O 4 gradient in buffer B was applied and fractions were collected and assayed for enzymatic activity.

Assay of histidyl-tRNA synthetase The enzyme was assayed using a modification of the method of Kalousek and Konigsberg (1974). All samples for determining yield and purification were dialyzed against buffer A before assay. The assay was performed in a total volume of 25 #1 at 3 7 ° C for 10 min in 12 mM Tris-HC1, 10 mM MgC12, 5 mM 5'-ATP, 0.6 mM DTT, p H 7.5, with 125 /xg tRNA, 5 /11 of c-[3H]histidine and an appropriate amount of enzyme to be assayed (0.1 /~g-0.2/~g protein). After incubation for 10 min at 37 ° C the reaction was stopped by adding 5/~1 of ice cold 0.25 N HC1. A 25 /~1 aliquot from the assay was then put onto an analytical paper disc and dried, and then each pad was separately incubated in a well of the 24-well microtiter plates with 10% TCA at 4 ° C for 20 rain. The pads were washed with triple-distilled water and dried at 60°C. TCA-precipitable 3H-counts were determined by liquid scintillation counting. Other aminoacyl-tRNA synthetase activities were assayed under identical conditions with the substitution of an appropriate 3H-labeUed L-amino acid. A unit of activity was defined as one mmol of g-[3H]histidine incorporated into TCA-precipitable material as g-[SH]histidyl-tRNA under the assay conditions. The D E A E - H P L C fractions with

enzyme activity were pooled and stored at - 80 ° C. Protein was estimated with the Bio-Rad colorimetric protein assay using BSA as the standard (Bradford, 1976). SDS-polyacrylamide gel electrophoresis was performed according to Laemmli (1970). After electrophoresis, gels were either Coomassie brilliant blue or silver stained (Merril et al., 1979) or transblotted by a modification of standard methods (Renart et al., 1979; Towbin et al., 1979). The 10% polyacrylamide gels were equilibrated for 15 min in transfer buffer (12.5 mM Tris, 96 mM glycine and 20% (v/v) methanol, p H 8.2), and proteins electrophoretically transferred onto nitrocellulose at 70 V overnight with cooling. The nitrocellulose with transferred proteins was washed with PBS, air-dried, cut vertically into 1 cm strips and blocked overnight at 4 ° C with 10% goat serum in borate-buffered saline (pH 8.0). The strips were then incubated in sealed plastic bags with serum diluted 1 : 1 0 in 10% goat serum in borate-buffered saline (pH 8.0) and incubated for 2 h at room temperature with constant agitation. The strips were washed with five changes of PBS0.5% Tween 20 and incubated at room temperature for 1 h with a 1/500 dilution of peroxidaseconjugated affinity-purified goat IgG anti-human IgG and IgM. After incubation the strips were washed in PBS-0.5% Tween 20 and developed with a solution containing one part of 3 m g / m l 4chloro-l-naphthol in methanol and five parts of 50 mM Tris-saline, pH 7.5. H202 was then added to the developing solution at a final concentration of 0.0375%. The reaction was stopped by immersing the nitrocellulose strips in water, and strips were then blotted dry and photographed.

Results

Histidyl-tRNA synthetase was enriched over 5000-fold with a recovery of 40% from HeLa cells using this rapid method (Table I). Employing this purification scheme the enzyme preparation was obtained within 4 days of HeLa cell harvesting. The first chromatographic step, on Blue Sepharose CL-6B, is common to several published methods for purifying aminoacyl-tRNA synthetases (Moe and Piszldewicz, 1976; Drocourt et al., 1980; Putney et al., 1981), since enzymes re-

238

i

O

300

20 1,5-

%

[50

x (.~ I-

l0

II

Z

~ 5

,

. . . . . . . 10

J [5

. . . .

i . . . . 20

0

>l.-

25

1.0-

~RACTION (,.D

Fig. 1. D E A E - H P L C ion exchange chromatography of histidyl-tRNA synthetase. The enzymatic activity (O O) was eluted from the column by a linear K H z P O 4 gradient from 0 to 300 mM. 2 ml fractions were collected at 1.2 m l / m i n and assayed for enzymatic activity as described in the materials and methods section. The ion exchange D E A E - H P L C column (Spherogel TSK-5PW) was stored in 10% methyl alcohol and washed before use with 8 M guanidine-HC1 and 10% DMSO, regenerated with 0.1 N N a O H and then washed with 0 - 3 0 0 m M linear K H 2 P O 4 gradient in buffer B and equilibrated finally with buffer B prior to sample application.

quiring adenylyl-containing substances bind to this gel. The apparent high recovery of enzymatic activity at this stage is not understood, but might be due in part to the removal of inhibitors. The calculation of recovery of enzymatic activity compared to the amount present after Blue Sepharose

N Z

~ 0.5~

o--__o--o4, , 1

, ko.o.~_~.-.__.~,

5 I0 15 FRACTION N U M B E R

20

Fig. 2. HPLC size exclusion chromatography of purified histidyl-tRNA synthetase. The Dupont Zorbax GF-250 column was calibrated with molecular weight standard proteins. Then 1 ml fractions at 1 m l / m i n were collected after application of the purified sample and the enzymatic activity ( O - - - - - - O) eluted was compared with molecular weight (HPLC) standard protein kit (glutamate dehydrogenase, 290 kDa; lactate dehydrogenase, 140 kDa; enolase, 67 kDa; adenylate kinase. 32 k D a and cytochrome C, 12.4 kDa).

TABLE I P U R I F I C A T I O N OF Jo-1 A N T I G E N ( H I S T I D Y L - t R N A SYNTHETASE) F R O M HeLa CELLS Unit of enzyme activity as defined in the materials and methods section. Fraction

HeLa cell lysate Post-mitochondrial supernatant Post-ribosomal supernatant Blue Sepharose eluent Poly-U-Sepharose eluent HPLC-DEAE eluent

Protein concentration (mg/ml)

Total protein (mg)

28

5 544

Specific activity (units × 1 0 9 / m g protein)

Total activity (units x 109)

(%)

0.4305

2 386

100

1

Recovery

Purification

7.2

914.4

1.739

1 590

67

4

6.6

735

2.15

1 580

66

5

0.44

86.77

96

8 329

350

223

0.152

6.38

430

2 743

115

998

0.0465

0.465

2 200

1 023

43

5110

239

chromatography yields a value of 12.3%. In early experiments, the protein called La, which is very similar in size and charge to Jo-1 (Hoch and Billings, 1984; Francoeur et al., 1985) and to which autoantibodies are commonly found in the rheumatic diseases, contaminated all preparations of Jo-1. Therefore we introduced a chromatographic step with Poly-U-Sepharose since La binds nucleic acids with 3'-oligo-U sequences. This property had been used in the purification of La (Stefano, 1984). This step removed all La protein, as judged by immunoblotting (Fig. 3). At the last purification step, histidyl-tRNA

synthetase was eluted as a single peak of activity from the ion exchange DEAE-HPLC column at about 150 mM KH2PO 4 (Fig. 1). The preparation obtained from the ion exchange DEAE-HPLC chromatography contained only histidyl-tRNA synthetase activity when all the aminoacyl-tRNA synthetases were assayed. The apparent native molecular weight of the HeLa histidyl-tRNA synthetase as determined by HPLC size exclusion column chromatography was approximately 120 kDa (Fig. 2). When this antigen preparation was tested by immunoblotting after SDS-polyacrylamide gel electrophoresis under denaturing conditions, however, the only band recognized by anti-Jo-1 antibody-containing sera was at about 55 kDa (Fig. 3, lanes 5 and 6). Since all anti-Jo-1 positive sera inhibit the activity of histidyl-tRNA synthetase (Biswas et al., 1987), and since the enzyme activity and a protein of that size can be eluted from an affinity column made with an anti-Jo-1 positive serum IgG (unpublished

~Jiiii!ii:::!i~

92.5 66.2 55 45 31

otRo

o~Sm aLa

aRNP

aJo-1 aJo-1 NHS

Fig. 3. Immunoblotting of Jo-1 antigen (histidyl-tRNA synthetase) preparation with autoantibodies of defined specificities. The nitrocellulose strips with transblotted DEAE-HPLC purified histidyl-tRNA synthetase were blocked with 10% goat serum in borate-buffered saline (pH 8.0) and probed with a 1:10 dilution of patient sera with autoantibodies of known specificity or a 1 : 10 dilution of normal human serum (NHS). A 1:500 dilution of peroxidase-conjugated affinity-purified goat IgG anti-human IgG and IgM was used as the second antibody, and bands were developed as described in the materials and methods section.

Fig. 4. SDS-PAGE of the DEAE-HPLC histidyl-tRNA synthetase enriched fraction (75 ~g protein). Bio-Rad high molecular weight and low molecular weight standards (phosphorylase B 92.5 kDa, BSA 66.2 kDa, ovalbumin 45 kDa and carbonic anhydrase 31 kDa) were used to estimate the molecular size of the protein bands after Coomassie brilliant blue staining (arrows at right). The location of the estimated fnolecular size of Jo-1 antigen (55 kDa) is indicated by the arrow at the left.

240

results), it is reasonable to conclude that this band is histidyl-tRNA synthetase. Coomassie brilliant blue staining of a similar denaturing gel of this antigen preparation revealed a number of other protein bands (Fig. 4). Since histidyl-tRNA synthetase comprises less than 0.01% of the cellular protein of rabbit reticulocytes (Kane et al., 1978), the presence of other proteins in a 5000-fold enriched preparation is expected. In order to establish the usefulness of this enriched material for immunologic studies, it was necessary to show that other antigens to which autoantibodies are commonly directed are not present. As Fig. 3 shows, anti-Sm, anti-RNP, anti-Ro and anti-La antibodies did not detect any proteins. In the accompanying paper (Biswas et al., 1987) further studies document the absence of other antigens of importance in autoimmune rheumatologic diseases.

Discussion Although methods for the purification of histidyl-tRNA synthetase from several sources have been described, each has drawbacks. The biochemical methods, devised before the immunological interest in the enzyme as an autoantigen in human disease, are long and yield very little enzyme. The recently described immunoaffinity method depends upon the availability of relatively rare monospecific sera from patients with inflammatory myopathy. The method described here allows a high yield of the enzyme sufficiently purified for further immunologic studies from an easily available human source, HeLa cells. The molecular size for the undenatured enzyme has been reported as 122 kDa in rabbit reticulocytes by sucrose density gradient centrifugation (Kane et al., 1978), as 100-150 kDa in rat liver by gel filtration (Yang et al., 1984) and as 153 kDa in beef liver by gel filtration (Walker et al., 1983). We have determined an approximate molecular size for the enzyme from HeLa cells of 120 kDa by HPLC size exclusion chromatography. The subunit size has been reported as 64 kDa in rabbit reticulocytes by SDS-PAGE, consistent with a native homodimer; the rat liver enzyme was similar. The enzyme subunit size from HeLa cells

has been reported as 50 kDa (Mathews and Bernstein, 1983). Our finding of a subunit size of 55 kDa from HeLa cells is consistent with a dimer, although we cannot rule out the unlikely possibility that it is a heterodimer of which only one component is antigenic. These differences in molecular size estimates may reflect true size differences, differences in the methods of analysis, a n d / o r differences in proteolytic degradation because of differences in preparation. The final step of purification yields material containing protein bands other than the 55 kDa Jo-1 by silver or Coomassie brilliant blue staining of SDS-polyacrylamide gels, but it is free of other aminoacyl-tRNA synthetase activities and free of other cellular components recognized by immunoblotting with a variety of antisera of defined specificity from patients with rheumatic diseases. It has, therefore, been sufficiently purified to develop a sensitive and specific assay for measuring levels of anti-Jo-1 antibodies in patients with inflammatory myositis (Biswas et al., 1987).

Acknowledgements The authors thank Dr. Mary E. Cronin for helpful suggestions and Ms. Toni Vaughn and Ms. Dawn Smith for excellent secretarial help.

References Arnett, F.C., Hirsch, T.J., Bias, W.B., Nishikai, M. and Reichlin, M. (1981) J. Rheumatol. 8, 925. Biswas, T., Miller, F.W., Takagaki, Y. and Plotz, P.H. (1987) J. Immunol. Methods 98, 243. Bradford, M.M. (1976) Anal. Biochem. 72, 248. Drocourt, J.L., Gangloff, J., Dirheimer, G. and Thang, M.N. (1980) Biochem. Biophys, Res. Commun. 97, 787. Francoeur, A.M., Chan, E.K.L, Garrels, J.I. and Mathews, M.B. (1985) Mol. Cell. Biol. 5, 586. Hoeh, S.O. and Billings, P.B, (1984) J. Immunol. 133, 1397. Kalousek, F. and Konigsberg, W.H. (1974) Biochemistry 13, 999. Kane, S.M., Vugrincic, C., Finbloom, D.S. and Smith, D.W.E. (1978) Biochemistry 17, 1509. Laemmli, U.K. (1970) Nature 227, 680. Mathews, M.B. and Bernstein, R.M. (1983) Nature 304, 177. Merril, C.R., Switzer, R.C. and Van Keuren, M,L. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 4335. Moe, J.G. and Piszldewicz, D. (1976) FEBS Lett. 72, 147.

241 Nishikai, M. and Reichlin, M. (1980) Arthritis Rheum. 23, 881. Putney, S.D., Sauer, R.T. and Schimmel, P.R. (1981) J. Biol. Chem. 256, 198. Renart, J., Reiser, J. and Stark, G.R. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 3116. Stefano, J,E. (1984) Cell 36, 145. Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 4350.

Walker, E.J., Treacy, G.B. and Jeffrey, P.D. (1983) Biochemistry 22, 1934. Yang, D.C.H., Dang, C.V. and Arnett, F.C. (1984) Biochem. Biophys. Res. Commun. 120, 15. Yoshida, S., Akizuki, M., Mimori, T., Yamagata, H., Inada, S. and Homma, M. (1983) Arthritis Rheum. 26, 604.