Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 702-708
October 31, 1991
THE NH2-TERMINAL EXTENSION OF RAT LIVER ARGINYL-tRNA SYNTHETASE IS RESPONSIBLE FOR ITS HYDROPHOBIC PROPERTIES Shouting Huang and Murray P. Deutscher Department of Biochemistry University of Connecticut Health Center Farmington, CT 06030
Received September 9, 1991
SUMMARY: Rat liver arginyl-tRNA synthetase is found in extracts either as a component (.~_r
= 72,000) of the multienzyme aminoacyl-tRNA synthetase complex or as a low molecular weight = 60,000) free protein. The two forms are thought to be identical except for an extra peptide extension at the NH2-terminus of the larger form which is required for its association with the complex, but is unessential for catalytic activity. It has been suggested that interactions among synthetases in the multienzyme complex are mediated by hydrophobic domains on these peptide extensions of the individual proteins. To test this model we have purified to homogeneity the larger form of arginyl-tRNA synthetase and compared its hydrophobicity to that of its low molecular weight counterpart. We show that whereas the smaller protein displays no hydrophobic character, the larger protein demonstrates a high degree of hydrophobicity. No lipid modification was found on the high molecular weight protein indicating that the amino acid sequence itself is responsible for its hydrophobic properties. These findings support the proposed model for synthetase association within the multienzyme complex. © 1991Academic Press,
Inc.
Studies from many laboratories have firmly established that aminoacyl-tRNA synthetases from higher eukaryotic cells can be isolated as organized complexes containing multiple synthetase activities (see Ref. 1 for a recent review). Although some uncertainty still exists as to the exact number of synthetases in these high molecular weight assemblies, the main focus of work in this area has shifted from proving the existence of these complexes to understanding their structure and function. Of particular interest are the observations that hydrophobic interactions are important for maintaining the structural integrity of the complex (2,3), and that synthetases in this structure contain peptide extensions that are unessential for catalytic activity, but whose removal by proteolysis leads to release of low molecular weight forms of the enzymes (4-7). Based on these findings, one of us proposed a model (8) suggesting that the peptide extensions on individual synthetases are hydrophobic domains that serve to organize the various proteins into the multienzyme complex, possibly in association with bound lipids. The attachment of synthetases to the complex in this model would be similar to the association of certain proteins with membranes, and the possibility that synthetases may associate with membranes in vivo has also been considered (8,9). Some support for this model has come from observations showing that isoleucyl-, leucyland lysyl-tRNA synthetases isolated from the high molecular weight complex are hydrophobic proteins (5,10), and that a low molecular weight form of lysyl-tRNA synthetase released from the complex by proteolysis loses its hydrophobic character (5). The sequence of human aspartyl0006-291X/91 $1.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
702
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tRNA synthetase revealed that this protein also has a potential NH2-terminal amphiphilic helix that could confer hydrophobic properties (11). Furthermore, it was found that the synthetase complex itself displays a high degree of hydrophobicity which results in its association with various lipids (6,12), and that removal of the bound lipids has profound effects on the structure and activity of synthetases within the complex (12,13). Additional evidence supporting the model has been obtained from studies of rat liver arginyl-tRNA synthetase. This enzyme has the unusual property that it is found in extracts both as a free form (Mr = 60,000) and as a component of the multienzyme complex (Mr = 72,000) (6,14). Comparison of the two purified proteins showed that they are identical except for an additional peptide extension at the NH2-terminus of the larger form (6,15). Since proteolysis of the complex releases a low molecular weight form similar to the endogenous free form, the NH2terminal extension on the larger protein apparently is responsible for its association with the complex (6). Recent evidence has also suggested that the two forms may co-exist in ceils and serve different functions (16). It is not known, however, whether the larger form is a hydrophobic protein and whether its NH2-terminal extension plays a role in this property. In this paper we compare the hydrophobicity of high molecular weight arginyl-tRNA synthetase isolated from the complex to that of the low molecular weight protein, and show directly that the additional - 12 kDa NH2-terminal extension on the larger protein makes this form much more hydrophobic. We also show that the increased hydrophobicity is not due to covalent modification by lipid, indicating that the NH2-terminal amino acid sequence itself is responsible for the hydrophobic properties of the longer protein.
EXPERIMENTAL PROCEDURES
Materials - Sodium dodecyl sulfate (SDS), diazomethane, polyethylene glycol 6000 (PEG), phenylmethylsulfonyl fluoride and Triton X-114 were obtained from Sigma. [14C] arginine was from ICN. Phenyl-Sepharose and Sepharose 4B were products of Pharmacia, and Bio-Gel A-5m was from Bio-Rad Laboratories. Fatty acid methyl ester standards were obtained from Nu Chek Prep (Elysian, MN). Reagents for immunoblotting were purchased from Promega. Frozen rat livers (from adult, fasted animals) were obtained from Pel-Freez Biologicals. Rabbit liver and Escherichia ¢oli tRNA were prepared as described previously (17,18). tRNA-Sepharose was synthesized according to Remy et al. (19). Assavs- Arginyl-tRNA synthetase activity was determined by incorporation of [14C] arginine into liver tRNA as described (14). Protein was measured according to Bradford (20). Protein blotting and immunodetection were carded out according to the instructions in the Promega kit using IgG prepared against the low molecular weight form of arginyl-tRNA synthetase (12,15). Prenaration of Rat Liver A m i n o a c v l - t R N A Svnthetase Comnlex - The synthetase complex was purified by a modification of the procedure of Kellerman et al. (21). Frozen rat liver (28 g) was homogenized in 2 volumes of 50 mM Tris-C1, pH 7.3, 5 mM MgCI2, 1 mM dithiothreitol, 1.2 mM phenylmethylsulfonyl fluoride, 10% (v/v) glycerol and 3.5% (v/v) isopropanol using five 45 sec treatments in a Sorvall Omni-mixer with cooling between treatments. The extract was then centrifuged at 15,000 rpm for 30 min to remove cell debris and larger organelles. The supematant material was fractionated at 4°C with PEG 6000 by slowly mixing with a 50% (w/v) solution and stirring for 30 rain. The material precipitating between 2% and 5% PEG was recovered and dissolved in 10 ml of column buffer (25 mM K phosphate, pH 7.5, 10 mM 2-mercaptoethanol, 10% glycerol). Insoluble material was removed by high speed centrifugation, and the clarified sample was loaded on a column (90 x 2.8 cm) of Bio-Gel A-5m equilibrated in the column buffer. Fractions containing the aminoacyl-tRNA synthetase complex 703
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
were identified by assaying arginyl-tRNA synthetase. The active fractions were combined and added to a column of tRNA-Sepharose (6 x 1.5 cm) prepared in the column buffer. The synthetase complex was eluted with a 360-ml gradient from 25 to 400 mM K phosphate. Fractions containing the synthetase complex eluted between 100 and 220 mM K phosphate. The active fractions were pooled, concentrated to 4 mg/ml with an Amicon YM-10 membrane, and stored at -20°C. Preparation of Ar~invl-tRNA Svnthetases - Native low molecular weight arginyl-tRNA synthetase was prepared as described (14). The enzyme was denatured by preparative SDSPAGE (9 x 12 x 0.4 cm) gel run according to Laemmli (22). The high molecular weight form of arginyl-tRNA synthetase was isolated and denatured in one step by SDS-PAGE of the tRNASepharose fraction of the aminoacyl-tRNA synthetase complex described above. The band corresponding to arginyl-tRNA synthetase was identified by immunoblotting. Protein bands were visualized by soaking the gels in 0.25 M KCI (23). Bands corresponding to the high and low molecular weight forms of arginyl-tRNA synthetase were sliced from their respective gels, and electroeluted from the gel slice according to Harrington (24). SDS was removed from the sample by electrodialysis against 50 mM NH4HCO3, pH 8.5 containing 0.001% SDS, and then against 50 mM NH4 HCO3, pH 8.5 (24). Hvdrovhobic Chromatoeravhv - The hydrophobicities of the high and low molecular weight arginyl-tRNA synthetases were compared by chromatography on phenyl-Sepharose. All operations were carried out at 4°C. Protein, 5 to 10 Ixg, was dialyzed against 10 mM Na phosphate, pH 7.5, 0.2 mM dithiothreitol, 0.2 mM EDTA, 0.2 mM phenylmethylsulfonyl fluoride, 10% glycerol. After dialysis the sample was diluted to 500 Ixl with additional buffer, and loaded on a phenyl-Sepharose column (50 ~tl) prepared in a pipet tip and equilibrated with the dialysis buffer. The column was eluted with this buffer alone or containing various concentrations of ethylene glycol or Triton X-114, as noted in the figure legends. Each fraction was dialyzed against water, lyophilized, and analyzed by SDS-PAGE and immunoblotting or staining. Analvsis for Covalentlv-bound Fattv Acids - To remove non-covalently-bound lipids, purified aminoacyl-tRNA synthetase complex or purified high molecular weight arginyl-tRNA synthetase were extracted with 10 ml of chloroform:methanol (2:1). The insoluble protein was separated from the solvent by centrifugation, and the precipitate was washed repeatedly (up to 10 times) until fatty acid was no longer detectable in the washes by gas chromatography. The extracted pellet was dried in a stream of N2 at room temperature, and suspended in 5 ml of 83% methanol-2 N HC1. The sample was hydrolyzed in vacuo at 95oc for 60 h. The hydrolysate was extracted three times with 3 ml of hexane each, the extracts combined, and dried in a stream of N2. Lipids were methylated with diazomethane (25), dried and dissolved in hexane. Portions of the lipid fraction were analyzed on a 10% Silar 10C column (183 x 0.2 cm) using a Varian 3700 gas chromatograph as described (26).
RESULTS AND DISCUSSION Purification of Ar~invl-tRNA Svnthetase - Considerable difficulty has been encountered in attempts to purify native arginyl-tRNA synthetase from the high molecular weight synthetase complex. Thus, upon dissociation of the synthetase complex, arginyl-tRNA synthetase activity is rapidly lost, the protein tends to aggregate and it also sticks tightly to glass (3,6). This behavior is quite different from that of the low molecular weight form of the enzyme, which is a typical soluble protein (14). The difference between the two forms of the enzyme led to the suggestion that the extra extension, which is thought to be responsible for its association with the complex, also confers hydrophobic character to the protein (6,8). This possibility is directly tested here. In order to minimize the effects of secondary and tertiary structure, proteins that had been denatured were used for the studies presented. Both the high and low molecular weight forms of 704
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
of ethylene glycol up to 30% concentration (lanes 2 to 4, Fig. 1B), and only a small amount was released at 50% (lane 5). The arginyl-tRNA synthetase was efficiently removed from the column only upon addition of 1% Triton X-114 to the column buffer (lane 6, Fig. 1B). These data demonstrate that the complex-derived form of arginyl-tRNA synthetase is a strongly hydrophobic protein, and that the ~ 12 kDa NH2-terminal extension is sufficient, by itself, to dramatically increase hydrophobicity implying that it is a very hydrophobic domain. Analvsis for Covalentlv-bound Linid - In earlier work (15) we presented amino acid
compositions for the high and low molecular weight forms of arginyl-tRNA synthetase, and by difference, calculated a presumed composition for the - 12 kDa NH2-terminal extension. Although there are pitfalls in this type of difference analysis, the apparent composition of the NH2-terminal extension did not include a large number of hydrophobic amino acid residues. So, even though the presence of a hydrophobic domain was not completely excluded, this analysis raised the question of whether some type of lipid modification, rather than a sequence of amino acids, might be responsible for the hydrophobic character of this domain. To examine this possibility in more detail, we have analyzed the purified arginyl-tRNA synthetase for covalently bound fatty acids using gas chromatography. Analysis was also carried out on the purified aminoacyl-tRNA synthetase complex for comparison. Both preparations were first exhaustively extracted with chloroform-methanol to remove free lipid prior to hydrolysis to release any covalently-bound residues that might be present. The data in Table I show that arginyl-tRNA synthetase is devoid of the various saturated and unsaturated fatty acids examined. In addition, no unidentified peaks were present in the fatty acid region of the chromatogram. Thus, these findings suggest that the hydrophobicity of the
Table I Analysis for eovalentlv-bound fatty acids in the aminoaevl-tRNA svnthetase comDlex and in nurified ar~invl-tRNA svnthetase
$toichiometry Fattv acid
Complex ArgRSa nmol fatty_acid/nmol protein < 0.01 < 0.02 O.11 < 0.02 0.06 < 0.02 < 0.01 < 0.02 < 0.01 < 0.02 < 0.01 < 0.02
Mydsfic (14:0) Palmitic (16:0) Stearic (18:0) Linoleic (18:2) Linolenic (18:3) Arachidoni¢ (22:6) aArgRS, arginyl-tRNA synthetase
Samples were extracted, hydrolyzed and analyzed as described in "Experimental Procedures". Synthetase complex (4 mg) and arginyl-tRNA synthetase (200 [tg) were used for the analysis. The amount of fatty acid in a peak was calculated by comparing its area to that of a known standard. Sensitivity was about 0.05 nmol. The data are presented as the ratio of moles of fatty acid to moles of protein assuming the average molecular weight of a protein in the complex is 100,000. 706
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
arginyl-tRNA synthetase remained inactive, even after removal of the SDS. Use of denatured protein allowed us to isolate homogeneous arginyl-tRNA synthetase from the purified synthetase complex using the denaturing conditions of SDS-PAGE.
Purification was followed by
immunoblotting rather than by enzyme activity. Arginyl-tRNA synthetase isolated from the complex in this manner migrated as a ~ 72 kDa protein on denaturing gels. Approximately 200 pg of purified protein could be obtained from 4 mg of synthetase complex by the procedure described. The aggregative properties of the larger arginyl-tRNA synthetase were retained by this procedure since removal of the SDS by acetone precipitation led to a precipitate that could not be re-dissolved. For this reason, we resorted to electro-dialysis to remove the detergent prior to the column studies. Chromatography
on PhenvI-SeDharose - The hydrophobicities of the high and low
molecular weight arginyl-tRNA synthetases were compared based on their retention on columns of phenyl-Sepharose (Fig. 1). The low molecular weight enzyme was not retained by phenylSepharose in 10 mM Na phosphate; all the recovered protein was observed in lanes 1 and 2 (Fig. 1A) which represent the flow-through and buffer wash fractions, respectively. These findings are in agreement with those previously observed for the native, fully active enzyme (6), and they suggest that denaturation does not expose any hydrophobic domains sufficient for binding the protein to phenyl-Sepharose. In contrast, the high molecular weight form of arginyl-tRNA synthetase displayed a high degree of hydrophobic character. No protein was eluted from phenyl-Sepharose in the presence
A
1
2
3
4
5
6
-94
B
1
2
3
4
5
6
--68 -94
-46
-68
-45
Fig. 1: Chromatography of arginyl-tRNA synthetases on phenyl-Sepharose. Chromatography was carried out as described in "Experimental Procedures". A. Low molecular weight enzyme (5 pg) was added to the column and eluted with the indicated solution. Lane 1, flow through; lane 2, buffer; lane 3, buffer + 30% ethylene glycol; lane 4, buffer + 50% ethylene glycol; lane 5, buffer + 1% Triton X-114; lane 6, 2 ktg low molecular weight enzyme. Arginyl-tRNA synthetase was detected by immunoblotting. B. High molecular weight enzyme (10 tag) was added to the column and eluted with the indicated solution. Lane 1, 2 I.tg purified enzyme control; lane 2, flow-through; lane 3, buffer; lane 4, buffer + 30% ethylene glycol; lane 5, buffer + 50% ethylene glycol; lane 6, buffer + 1% Triton X-114. Arginyl-tRNA synthetase was detected by silver staining. The molecular weight standards shown on the right are from top to bottom: phosphorylase b, bovine serum albumin, ovalbumin. 705
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
NH2-terminal domain of arginyl-tRNA synthetase is not due to covalently-bound lipid, and must be a property of the amino acid sequence itself. On the other hand, covalently-bound fatty acids (palmitic and stearic) were detectable in the synthetase complex at a stoichiometry of about onetenth to one-twentieth that of protein, raising the possibility that some proteins within the complex are modified by lipids. Further work will be necessary to identify which protein(s) within the complex might be affected. The data presented here demonstrate directly that the NH2-terminal extension on the high molecular weight form of arginyl-tRNA synthetase is responsible for the hydrophobic character of this protein, and that it is the amino acid sequence itself, rather than a lipid modification, that is involved. In the case of human aspartyl-tRNA synthetase, whose sequence was recently determined at the DNA level (11), the NH2-terminal sequence was suggested to take the form of a neutral amphiphilic helix which confers hydrophobic properties. We cannot rule out the presence of such an amphiphilic helix in the extra extension of arginyl-tRNA synthetase, although it would have had to re-form after removal of the SDS. It is more likely that the NH2-terminal domain remains denatured, and that the hydrophobicity of this protein is a direct consequence of a sequence of hydrophobic amino acid residues that were not apparent by the difference amino acid analysis (15). Inasmuch as the enzymatic and immunological properties of the high- and lowmolecular weight forms of arginyl-tRNA synthetase are so similar (6), and since denatured proteins were used for these studies, we think it is highly improbable that the presence of the NH2-terminal extension leads to a major conformational change that exposes a hydrophobic domain elsewhere on the protein. However, sequence information for the NH2-terminal region and detailed structural analysis of the protein will be necessary to completely eliminate such a possibility. The most likely possibility that the NH2-terminal domain of arginyl-tRNA synthetase is hydrophobic, and is responsible for association of this protein with the synthetase complex (6), supports the previous suggestion that hydrophobic interactions among the peptide extensions play an important role in maintaining the structure of this assembly (3,8). It is an open question, however, whether in vivo the hydrophobic extensions serve to maintain synthetases in a multienzyme complex similar to that isolated in vitro, or whether these extensions serve to associate synthetases with membranous structures. ACKNOWLEDGMENTS We would like to thank Dr. Sanoj Suneja and Dominick Cinti for assistance with the gas chromatographic analysis. This work was supported by Grant GM16317 from the National Institutes of Health. REFERENCES 1. Yang, D.C.H., and Jacobo-Molina, A. (1990) in Structural and Oreanizational Asnects of Metabolic Regulation (Srere, P.A., Jones, M.E., and Mathews, C.I(., eds.) pp. 199-214, Alan R. Liss, New York 2. Johnson, D.L., Dang, C.V., and Yang, D.C.H. (1980) J. Biol. Chem. 255, 4362-4366 3. Sihag, R.K., and Deutscher, M.P. (1983) J. Biol. Chem. 258, 11846-11850 4. Mirande, M., Cirakoglu, B., and Waller, J-P. (1983) Eur. J. Biochem. 1~1,163-170 707
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
5. Cirakoglu, B., and Waller, J-P. (1985) Eur. J. Biochem. 151,101-110 6. Vellekamp, G., Sihag, R.H., and Deutscher, M.P. (1985) J. Biol. Chem. 260, 98439847 7. Siddiqui, F.A., and Yang, D.C.H. (1985) Biochim. Biophys. Acta ~.2_8_,177-187 8. Deutscher, M.P. (1984) J. Cell Biol. 99, 373-377 9. Sivaram, P., and Deutscher, M.P. (1989) Life Science Advances - Biochemistrv 8, 7-12 10. Lazard, M., Mirande, M., and Waller, J-P. (1985) Biochemistry 2__44,5099-510-6 11. Jacobo-Molina, A., Peterson, R., and Yang, D.C.H. (1989) J. Biol. Chem. 264, 1660816612 12. Sivaram, P., Vellekamp, G., and Deutscher, M.P. (1988) ,I, Biol. Chem. 263, 1889118896 13. Sivaram, P., and Deutscher, M.P. (1990) ,I, Biol. Chem. 265, 5774-5779 14. Deutscher, M.P., and Ni, R.C. (1982) J. Biol. Chem. 257, 6003-6006 15. Vellekamp, G., and Deutscher, M.P. (1987) J. Biol. Chem. 262, 9927-9930 16. Sivaram, P., and Deutscher, M.P. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 3665-3669 17. Deutscher, M.P. (1972) J. Biol. Chem. 247, 459-468 18. Deutscher, M.P., and Ghosh, R.K. (1978) Nucleic Acids Res. 5, 3821-3830 19. Remy, P., Birmele, C., and Ebel, J.P. (1972) FEBS Lett. 27, 134-138 20. Bradford, M.M. (1976) Anal. Biochem. 72, 248-254 21. Kellermann, O., Brevet, A., Honetti, H., and Waller, J-P. (1979) Eur. ,I, Biochem. 99, 541-550 22. Laemmli, U.K. (1970) Natur~ 227, 680-685 23. Hager, D.A., and Burgess, R.R. (1980) Anal, Biochem. 109, 76-86 24. Harrington, M.G. (1990) Methods Enzvmol. 182, 488-495 25. Holloway, P.W. (1975) Methods Enzvmol. 35, 253-262 26. Nagi, M.N., Cook, L., Laguna, J.C., and Cinti, D.L. (1988) Arch. Biochem. Biophy~, 2~i7, 1-12
708
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 702-708
October 31, 1991
THE NH2-TERMINAL EXTENSION OF RAT LIVER ARGINYL-tRNA SYNTHETASE IS RESPONSIBLE FOR ITS HYDROPHOBIC PROPERTIES Shouting Huang and Murray P. Deutscher Department of Biochemistry University of Connecticut Health Center Farmington, CT 06030
Received September 9, 1991
SUMMARY: Rat liver arginyl-tRNA synthetase is found in extracts either as a component (.~_r
= 72,000) of the multienzyme aminoacyl-tRNA synthetase complex or as a low molecular weight = 60,000) free protein. The two forms are thought to be identical except for an extra peptide extension at the NH2-terminus of the larger form which is required for its association with the complex, but is unessential for catalytic activity. It has been suggested that interactions among synthetases in the multienzyme complex are mediated by hydrophobic domains on these peptide extensions of the individual proteins. To test this model we have purified to homogeneity the larger form of arginyl-tRNA synthetase and compared its hydrophobicity to that of its low molecular weight counterpart. We show that whereas the smaller protein displays no hydrophobic character, the larger protein demonstrates a high degree of hydrophobicity. No lipid modification was found on the high molecular weight protein indicating that the amino acid sequence itself is responsible for its hydrophobic properties. These findings support the proposed model for synthetase association within the multienzyme complex. © 1991Academic Press,
Inc.
Studies from many laboratories have firmly established that aminoacyl-tRNA synthetases from higher eukaryotic cells can be isolated as organized complexes containing multiple synthetase activities (see Ref. 1 for a recent review). Although some uncertainty still exists as to the exact number of synthetases in these high molecular weight assemblies, the main focus of work in this area has shifted from proving the existence of these complexes to understanding their structure and function. Of particular interest are the observations that hydrophobic interactions are important for maintaining the structural integrity of the complex (2,3), and that synthetases in this structure contain peptide extensions that are unessential for catalytic activity, but whose removal by proteolysis leads to release of low molecular weight forms of the enzymes (4-7). Based on these findings, one of us proposed a model (8) suggesting that the peptide extensions on individual synthetases are hydrophobic domains that serve to organize the various proteins into the multienzyme complex, possibly in association with bound lipids. The attachment of synthetases to the complex in this model would be similar to the association of certain proteins with membranes, and the possibility that synthetases may associate with membranes in vivo has also been considered (8,9). Some support for this model has come from observations showing that isoleucyl-, leucyland lysyl-tRNA synthetases isolated from the high molecular weight complex are hydrophobic proteins (5,10), and that a low molecular weight form of lysyl-tRNA synthetase released from the complex by proteolysis loses its hydrophobic character (5). The sequence of human aspartyl0006-291X/91 $1.50 Copyright © 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
702
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tRNA synthetase revealed that this protein also has a potential NH2-terminal amphiphilic helix that could confer hydrophobic properties (11). Furthermore, it was found that the synthetase complex itself displays a high degree of hydrophobicity which results in its association with various lipids (6,12), and that removal of the bound lipids has profound effects on the structure and activity of synthetases within the complex (12,13). Additional evidence supporting the model has been obtained from studies of rat liver arginyl-tRNA synthetase. This enzyme has the unusual property that it is found in extracts both as a free form (Mr = 60,000) and as a component of the multienzyme complex (Mr = 72,000) (6,14). Comparison of the two purified proteins showed that they are identical except for an additional peptide extension at the NH2-terminus of the larger form (6,15). Since proteolysis of the complex releases a low molecular weight form similar to the endogenous free form, the NH2terminal extension on the larger protein apparently is responsible for its association with the complex (6). Recent evidence has also suggested that the two forms may co-exist in ceils and serve different functions (16). It is not known, however, whether the larger form is a hydrophobic protein and whether its NH2-terminal extension plays a role in this property. In this paper we compare the hydrophobicity of high molecular weight arginyl-tRNA synthetase isolated from the complex to that of the low molecular weight protein, and show directly that the additional - 12 kDa NH2-terminal extension on the larger protein makes this form much more hydrophobic. We also show that the increased hydrophobicity is not due to covalent modification by lipid, indicating that the NH2-terminal amino acid sequence itself is responsible for the hydrophobic properties of the longer protein.
EXPERIMENTAL PROCEDURES
Materials - Sodium dodecyl sulfate (SDS), diazomethane, polyethylene glycol 6000 (PEG), phenylmethylsulfonyl fluoride and Triton X-114 were obtained from Sigma. [14C] arginine was from ICN. Phenyl-Sepharose and Sepharose 4B were products of Pharmacia, and Bio-Gel A-5m was from Bio-Rad Laboratories. Fatty acid methyl ester standards were obtained from Nu Chek Prep (Elysian, MN). Reagents for immunoblotting were purchased from Promega. Frozen rat livers (from adult, fasted animals) were obtained from Pel-Freez Biologicals. Rabbit liver and Escherichia ¢oli tRNA were prepared as described previously (17,18). tRNA-Sepharose was synthesized according to Remy et al. (19). Assavs- Arginyl-tRNA synthetase activity was determined by incorporation of [14C] arginine into liver tRNA as described (14). Protein was measured according to Bradford (20). Protein blotting and immunodetection were carded out according to the instructions in the Promega kit using IgG prepared against the low molecular weight form of arginyl-tRNA synthetase (12,15). Prenaration of Rat Liver A m i n o a c v l - t R N A Svnthetase Comnlex - The synthetase complex was purified by a modification of the procedure of Kellerman et al. (21). Frozen rat liver (28 g) was homogenized in 2 volumes of 50 mM Tris-C1, pH 7.3, 5 mM MgCI2, 1 mM dithiothreitol, 1.2 mM phenylmethylsulfonyl fluoride, 10% (v/v) glycerol and 3.5% (v/v) isopropanol using five 45 sec treatments in a Sorvall Omni-mixer with cooling between treatments. The extract was then centrifuged at 15,000 rpm for 30 min to remove cell debris and larger organelles. The supematant material was fractionated at 4°C with PEG 6000 by slowly mixing with a 50% (w/v) solution and stirring for 30 rain. The material precipitating between 2% and 5% PEG was recovered and dissolved in 10 ml of column buffer (25 mM K phosphate, pH 7.5, 10 mM 2-mercaptoethanol, 10% glycerol). Insoluble material was removed by high speed centrifugation, and the clarified sample was loaded on a column (90 x 2.8 cm) of Bio-Gel A-5m equilibrated in the column buffer. Fractions containing the aminoacyl-tRNA synthetase complex 703
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
were identified by assaying arginyl-tRNA synthetase. The active fractions were combined and added to a column of tRNA-Sepharose (6 x 1.5 cm) prepared in the column buffer. The synthetase complex was eluted with a 360-ml gradient from 25 to 400 mM K phosphate. Fractions containing the synthetase complex eluted between 100 and 220 mM K phosphate. The active fractions were pooled, concentrated to 4 mg/ml with an Amicon YM-10 membrane, and stored at -20°C. Preparation of Ar~invl-tRNA Svnthetases - Native low molecular weight arginyl-tRNA synthetase was prepared as described (14). The enzyme was denatured by preparative SDSPAGE (9 x 12 x 0.4 cm) gel run according to Laemmli (22). The high molecular weight form of arginyl-tRNA synthetase was isolated and denatured in one step by SDS-PAGE of the tRNASepharose fraction of the aminoacyl-tRNA synthetase complex described above. The band corresponding to arginyl-tRNA synthetase was identified by immunoblotting. Protein bands were visualized by soaking the gels in 0.25 M KCI (23). Bands corresponding to the high and low molecular weight forms of arginyl-tRNA synthetase were sliced from their respective gels, and electroeluted from the gel slice according to Harrington (24). SDS was removed from the sample by electrodialysis against 50 mM NH4HCO3, pH 8.5 containing 0.001% SDS, and then against 50 mM NH4 HCO3, pH 8.5 (24). Hvdrovhobic Chromatoeravhv - The hydrophobicities of the high and low molecular weight arginyl-tRNA synthetases were compared by chromatography on phenyl-Sepharose. All operations were carried out at 4°C. Protein, 5 to 10 Ixg, was dialyzed against 10 mM Na phosphate, pH 7.5, 0.2 mM dithiothreitol, 0.2 mM EDTA, 0.2 mM phenylmethylsulfonyl fluoride, 10% glycerol. After dialysis the sample was diluted to 500 Ixl with additional buffer, and loaded on a phenyl-Sepharose column (50 ~tl) prepared in a pipet tip and equilibrated with the dialysis buffer. The column was eluted with this buffer alone or containing various concentrations of ethylene glycol or Triton X-114, as noted in the figure legends. Each fraction was dialyzed against water, lyophilized, and analyzed by SDS-PAGE and immunoblotting or staining. Analvsis for Covalentlv-bound Fattv Acids - To remove non-covalently-bound lipids, purified aminoacyl-tRNA synthetase complex or purified high molecular weight arginyl-tRNA synthetase were extracted with 10 ml of chloroform:methanol (2:1). The insoluble protein was separated from the solvent by centrifugation, and the precipitate was washed repeatedly (up to 10 times) until fatty acid was no longer detectable in the washes by gas chromatography. The extracted pellet was dried in a stream of N2 at room temperature, and suspended in 5 ml of 83% methanol-2 N HC1. The sample was hydrolyzed in vacuo at 95oc for 60 h. The hydrolysate was extracted three times with 3 ml of hexane each, the extracts combined, and dried in a stream of N2. Lipids were methylated with diazomethane (25), dried and dissolved in hexane. Portions of the lipid fraction were analyzed on a 10% Silar 10C column (183 x 0.2 cm) using a Varian 3700 gas chromatograph as described (26).
RESULTS AND DISCUSSION Purification of Ar~invl-tRNA Svnthetase - Considerable difficulty has been encountered in attempts to purify native arginyl-tRNA synthetase from the high molecular weight synthetase complex. Thus, upon dissociation of the synthetase complex, arginyl-tRNA synthetase activity is rapidly lost, the protein tends to aggregate and it also sticks tightly to glass (3,6). This behavior is quite different from that of the low molecular weight form of the enzyme, which is a typical soluble protein (14). The difference between the two forms of the enzyme led to the suggestion that the extra extension, which is thought to be responsible for its association with the complex, also confers hydrophobic character to the protein (6,8). This possibility is directly tested here. In order to minimize the effects of secondary and tertiary structure, proteins that had been denatured were used for the studies presented. Both the high and low molecular weight forms of 704
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
of ethylene glycol up to 30% concentration (lanes 2 to 4, Fig. 1B), and only a small amount was released at 50% (lane 5). The arginyl-tRNA synthetase was efficiently removed from the column only upon addition of 1% Triton X-114 to the column buffer (lane 6, Fig. 1B). These data demonstrate that the complex-derived form of arginyl-tRNA synthetase is a strongly hydrophobic protein, and that the ~ 12 kDa NH2-terminal extension is sufficient, by itself, to dramatically increase hydrophobicity implying that it is a very hydrophobic domain. Analvsis for Covalentlv-bound Linid - In earlier work (15) we presented amino acid
compositions for the high and low molecular weight forms of arginyl-tRNA synthetase, and by difference, calculated a presumed composition for the - 12 kDa NH2-terminal extension. Although there are pitfalls in this type of difference analysis, the apparent composition of the NH2-terminal extension did not include a large number of hydrophobic amino acid residues. So, even though the presence of a hydrophobic domain was not completely excluded, this analysis raised the question of whether some type of lipid modification, rather than a sequence of amino acids, might be responsible for the hydrophobic character of this domain. To examine this possibility in more detail, we have analyzed the purified arginyl-tRNA synthetase for covalently bound fatty acids using gas chromatography. Analysis was also carried out on the purified aminoacyl-tRNA synthetase complex for comparison. Both preparations were first exhaustively extracted with chloroform-methanol to remove free lipid prior to hydrolysis to release any covalently-bound residues that might be present. The data in Table I show that arginyl-tRNA synthetase is devoid of the various saturated and unsaturated fatty acids examined. In addition, no unidentified peaks were present in the fatty acid region of the chromatogram. Thus, these findings suggest that the hydrophobicity of the
Table I Analysis for eovalentlv-bound fatty acids in the aminoaevl-tRNA svnthetase comDlex and in nurified ar~invl-tRNA svnthetase
$toichiometry Fattv acid
Complex ArgRSa nmol fatty_acid/nmol protein < 0.01 < 0.02 O.11 < 0.02 0.06 < 0.02 < 0.01 < 0.02 < 0.01 < 0.02 < 0.01 < 0.02
Mydsfic (14:0) Palmitic (16:0) Stearic (18:0) Linoleic (18:2) Linolenic (18:3) Arachidoni¢ (22:6) aArgRS, arginyl-tRNA synthetase
Samples were extracted, hydrolyzed and analyzed as described in "Experimental Procedures". Synthetase complex (4 mg) and arginyl-tRNA synthetase (200 [tg) were used for the analysis. The amount of fatty acid in a peak was calculated by comparing its area to that of a known standard. Sensitivity was about 0.05 nmol. The data are presented as the ratio of moles of fatty acid to moles of protein assuming the average molecular weight of a protein in the complex is 100,000. 706
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
arginyl-tRNA synthetase remained inactive, even after removal of the SDS. Use of denatured protein allowed us to isolate homogeneous arginyl-tRNA synthetase from the purified synthetase complex using the denaturing conditions of SDS-PAGE.
Purification was followed by
immunoblotting rather than by enzyme activity. Arginyl-tRNA synthetase isolated from the complex in this manner migrated as a ~ 72 kDa protein on denaturing gels. Approximately 200 pg of purified protein could be obtained from 4 mg of synthetase complex by the procedure described. The aggregative properties of the larger arginyl-tRNA synthetase were retained by this procedure since removal of the SDS by acetone precipitation led to a precipitate that could not be re-dissolved. For this reason, we resorted to electro-dialysis to remove the detergent prior to the column studies. Chromatography
on PhenvI-SeDharose - The hydrophobicities of the high and low
molecular weight arginyl-tRNA synthetases were compared based on their retention on columns of phenyl-Sepharose (Fig. 1). The low molecular weight enzyme was not retained by phenylSepharose in 10 mM Na phosphate; all the recovered protein was observed in lanes 1 and 2 (Fig. 1A) which represent the flow-through and buffer wash fractions, respectively. These findings are in agreement with those previously observed for the native, fully active enzyme (6), and they suggest that denaturation does not expose any hydrophobic domains sufficient for binding the protein to phenyl-Sepharose. In contrast, the high molecular weight form of arginyl-tRNA synthetase displayed a high degree of hydrophobic character. No protein was eluted from phenyl-Sepharose in the presence
A
1
2
3
4
5
6
-94
B
1
2
3
4
5
6
--68 -94
-46
-68
-45
Fig. 1: Chromatography of arginyl-tRNA synthetases on phenyl-Sepharose. Chromatography was carried out as described in "Experimental Procedures". A. Low molecular weight enzyme (5 pg) was added to the column and eluted with the indicated solution. Lane 1, flow through; lane 2, buffer; lane 3, buffer + 30% ethylene glycol; lane 4, buffer + 50% ethylene glycol; lane 5, buffer + 1% Triton X-114; lane 6, 2 ktg low molecular weight enzyme. Arginyl-tRNA synthetase was detected by immunoblotting. B. High molecular weight enzyme (10 tag) was added to the column and eluted with the indicated solution. Lane 1, 2 I.tg purified enzyme control; lane 2, flow-through; lane 3, buffer; lane 4, buffer + 30% ethylene glycol; lane 5, buffer + 50% ethylene glycol; lane 6, buffer + 1% Triton X-114. Arginyl-tRNA synthetase was detected by silver staining. The molecular weight standards shown on the right are from top to bottom: phosphorylase b, bovine serum albumin, ovalbumin. 705
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
NH2-terminal domain of arginyl-tRNA synthetase is not due to covalently-bound lipid, and must be a property of the amino acid sequence itself. On the other hand, covalently-bound fatty acids (palmitic and stearic) were detectable in the synthetase complex at a stoichiometry of about onetenth to one-twentieth that of protein, raising the possibility that some proteins within the complex are modified by lipids. Further work will be necessary to identify which protein(s) within the complex might be affected. The data presented here demonstrate directly that the NH2-terminal extension on the high molecular weight form of arginyl-tRNA synthetase is responsible for the hydrophobic character of this protein, and that it is the amino acid sequence itself, rather than a lipid modification, that is involved. In the case of human aspartyl-tRNA synthetase, whose sequence was recently determined at the DNA level (11), the NH2-terminal sequence was suggested to take the form of a neutral amphiphilic helix which confers hydrophobic properties. We cannot rule out the presence of such an amphiphilic helix in the extra extension of arginyl-tRNA synthetase, although it would have had to re-form after removal of the SDS. It is more likely that the NH2-terminal domain remains denatured, and that the hydrophobicity of this protein is a direct consequence of a sequence of hydrophobic amino acid residues that were not apparent by the difference amino acid analysis (15). Inasmuch as the enzymatic and immunological properties of the high- and lowmolecular weight forms of arginyl-tRNA synthetase are so similar (6), and since denatured proteins were used for these studies, we think it is highly improbable that the presence of the NH2-terminal extension leads to a major conformational change that exposes a hydrophobic domain elsewhere on the protein. However, sequence information for the NH2-terminal region and detailed structural analysis of the protein will be necessary to completely eliminate such a possibility. The most likely possibility that the NH2-terminal domain of arginyl-tRNA synthetase is hydrophobic, and is responsible for association of this protein with the synthetase complex (6), supports the previous suggestion that hydrophobic interactions among the peptide extensions play an important role in maintaining the structure of this assembly (3,8). It is an open question, however, whether in vivo the hydrophobic extensions serve to maintain synthetases in a multienzyme complex similar to that isolated in vitro, or whether these extensions serve to associate synthetases with membranous structures. ACKNOWLEDGMENTS We would like to thank Dr. Sanoj Suneja and Dominick Cinti for assistance with the gas chromatographic analysis. This work was supported by Grant GM16317 from the National Institutes of Health. REFERENCES 1. Yang, D.C.H., and Jacobo-Molina, A. (1990) in Structural and Oreanizational Asnects of Metabolic Regulation (Srere, P.A., Jones, M.E., and Mathews, C.I(., eds.) pp. 199-214, Alan R. Liss, New York 2. Johnson, D.L., Dang, C.V., and Yang, D.C.H. (1980) J. Biol. Chem. 255, 4362-4366 3. Sihag, R.K., and Deutscher, M.P. (1983) J. Biol. Chem. 258, 11846-11850 4. Mirande, M., Cirakoglu, B., and Waller, J-P. (1983) Eur. J. Biochem. 1~1,163-170 707
Vol. 180, No. 2, 1991
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
5. Cirakoglu, B., and Waller, J-P. (1985) Eur. J. Biochem. 151,101-110 6. Vellekamp, G., Sihag, R.H., and Deutscher, M.P. (1985) J. Biol. Chem. 260, 98439847 7. Siddiqui, F.A., and Yang, D.C.H. (1985) Biochim. Biophys. Acta ~.2_8_,177-187 8. Deutscher, M.P. (1984) J. Cell Biol. 99, 373-377 9. Sivaram, P., and Deutscher, M.P. (1989) Life Science Advances - Biochemistrv 8, 7-12 10. Lazard, M., Mirande, M., and Waller, J-P. (1985) Biochemistry 2__44,5099-510-6 11. Jacobo-Molina, A., Peterson, R., and Yang, D.C.H. (1989) J. Biol. Chem. 264, 1660816612 12. Sivaram, P., Vellekamp, G., and Deutscher, M.P. (1988) ,I, Biol. Chem. 263, 1889118896 13. Sivaram, P., and Deutscher, M.P. (1990) ,I, Biol. Chem. 265, 5774-5779 14. Deutscher, M.P., and Ni, R.C. (1982) J. Biol. Chem. 257, 6003-6006 15. Vellekamp, G., and Deutscher, M.P. (1987) J. Biol. Chem. 262, 9927-9930 16. Sivaram, P., and Deutscher, M.P. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 3665-3669 17. Deutscher, M.P. (1972) J. Biol. Chem. 247, 459-468 18. Deutscher, M.P., and Ghosh, R.K. (1978) Nucleic Acids Res. 5, 3821-3830 19. Remy, P., Birmele, C., and Ebel, J.P. (1972) FEBS Lett. 27, 134-138 20. Bradford, M.M. (1976) Anal. Biochem. 72, 248-254 21. Kellermann, O., Brevet, A., Honetti, H., and Waller, J-P. (1979) Eur. ,I, Biochem. 99, 541-550 22. Laemmli, U.K. (1970) Natur~ 227, 680-685 23. Hager, D.A., and Burgess, R.R. (1980) Anal, Biochem. 109, 76-86 24. Harrington, M.G. (1990) Methods Enzvmol. 182, 488-495 25. Holloway, P.W. (1975) Methods Enzvmol. 35, 253-262 26. Nagi, M.N., Cook, L., Laguna, J.C., and Cinti, D.L. (1988) Arch. Biochem. Biophy~, 2~i7, 1-12
708
