Proc. Nati. Acad. Sci. USA
Vol. 87, pp. 3665-3669, May 1990 Biochemistry
Existence of two forms of rat liver arginyl-tRNA synthetase suggests channeling of aminoacyl-tRNA for protein synthesis (ubiquitin-dependent proteolysis/arginyl-tRNA protein transferase)
PILLARISETTI SIVARAM AND MURRAY P. DEUTSCHER Department of Biochemistry, University of Connecticut Health Center, Farmington, CT 06032
Communicated by Mary J. Osborn, March 1, 1990
Arginyl-tRNA synthetase (arginine-tRNA liABSTRACT gase, EC 6.1.1.19) is found in extracts of mammalian cells both as a free protein (Mr = 60,000) and as a component (Mr 72,000) of the high molecular weight aminoacyl-tRNA synthetase complex (Mr > 106). Several pieces of evidence indicate that the low molecular weight free form is not a proteolytic degradation product of the complex-bound enzyme but that it preexists in vivo: (i) the endogenous free form differs in size from the active proteolytic fragment generated in vitro, (ii) conditions expected to increase or decrease the amount of proteolysis do not alter the ratio of the two forms of the enzyme, and (iu) the free form contains an NH2-terminal methionine residue. A model is presented that provides a rationale for the existence of two forms of arginyl-tRNA synthetase in cells. In this model the complexed enzyme supplies arginyl-tRNA for protein synthesis, whereas the free enzyme provides arginyltRNA for the N112-terminal arginine modification of proteins by arginyl-tRNA: protein arginyltransferase. This latter process targets certain proteins for removal by the ubiquitindependent protein degradation pathway. The necessity for an additional pool of arginyl-tRNA for the modifcation reaction leads to the conclusion that the arginyl-tRNA destined for protein synthesis (and/or protein modification) is channeled and unavailable for other processes. Other evidence supporting channeling in protein synthesis is discussed. It is becoming increasingly clear that many, if not all, of the macromolecular components of cells are organized into multienzyme complexes or associated with subcellular structures (see ref. 1 for a review). A major consequence of such organization is that it can lead to compartmentalization or channeling of metabolic pathways (i.e., the transfer of metabolites from one enzyme to another without equilibration with the total fluid of the cell). Channeled pathways presumably would be more efficient than free ones by increasing the local substrate concentrations for a given amount of substrate, by minimizing destruction of unstable intermediates or loss due to side reactions, and by allowing for combined regulation of multiple activities. The protein biosynthetic machinery is among the most complex in the cell; it consists of a large number of protein and nucleic acid components. Organization of these various components has been studied most extensively in the case of the aminoacyl-tRNA synthetases. In extracts of higher eukaryotic cells, these enzymes are generally found in high molecular weight multienzyme complexes, often containing other components as well (2, 3). The number of synthetases found in these complexes can vary depending on the method of isolation, but as many as eight or nine are routinely found associated with each other (4-7), and even higher numbers have been observed (8, 9). It is now known that hydrophobic The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
3665
interactions play an important role in holding aminoacyltRNA synthetases in the complex (10-12) and that terminal extensions on the individual proteins, not required for catalysis, participate in these associations (13). For a few aminoacyl-tRNA synthetases (namely, those for arginine and aspartic acid), both high and low molecular weight forms of the enzyme coexist in the same extract (14, 15). Arginyl-tRNA synthetase (arginine-tRNA ligase, EC 6.1.1.19) is present in extracts of rat liver as a free protein (Mr 60,000) as well as a component (Mr 72,000) of the high molecular weight complex (Mr > 106) (5, 16). Comparison of the catalytic and immunological properties of the free and complexed forms of this enzyme initially suggested that they are closely related proteins (17), and peptide mapping indicated that they are probably identical except for a Mr = 12,000 NH2-terminal extension present on the complexed form (13). On the basis of a comparison of the endogenous free form of arginyl-tRNA synthetase with a proteolytically derived fragment released from the complex in vitro, it was thought that the free form might have arisen by a limited cleavage of the complex-bound protein by some endogenous protease (13, 17). However, several pieces of evidence did not fit with this idea (17), and the question of whether proteolysis occurs during isolation or whether the free form preexists in the cell was also left open. We have now examined this issue in more detail. Our resUlts indicate that the endogenous free form differs from the in vitro proteolytically derived enzyme, that the free form probably preexists in the cell, and that most likely it is a distinct translation product. The existence of two forms of arginyl-tRNA synthetase in mammalian cells, coupled with the recently discovered role of arginyl-tRNA in ubiquitindependent protein degradation (18), leads us to propose a model that has important consequences with regard to the channeling of aminoacyl-tRNAs for protein biosynthesis. =
MATERIALS AND METHODS Materials. Sephacryl S-300 was obtained from PharmaciaLKB Biotechnology. The protease inhibitors leupeptin, antipain, pepstatin A, and phenylmethylsulfonyl fluoride were purchased from Sigma. a2-Macroglobulin was from Boehringer Mannheim. [14C]Arginine was obtained from ICN. Rabbit livers (from young, fasted animals) and rat livers (from adult, fasted animals) were purchased from Pel-Freez Biologicals. Sprague-Dawley female rats (-200 g) were obtained from Charles River Breeding Laboratories. Normal rat liver cells (clone 9) were obtained from the American Type Culture Collection. They were maintained as monolayers in Ham's nutrient mixture F-12 with 10o (vol/vol) fetal bovine serum. Rabbit liver tRNA was prepared as described (19). Reagents for immunoblotting were products of Promega. All other materials were as reported (20). Preparation of Arginyl-tRNA Synthetases. Rat liver extracts for gel filtration were prepared from fresh liver by homoge-
3666
Biochemistry: Sivaram and Deutscher
nization and high-speed centrifugation (see the legend to Fig. 2). The low molecular weight free form of arginyl-tRNA synthetase was purified from frozen livers as described by Deutscher and Ni (16). The proteolyzed form of arginyltRNA synthetase was prepared from the high molecular weight complex purified through the Controlled-Pore glass (Electro-Nucleonics) column step followed by treatment with papain (17). Enzyme Assays. Aminoacyl-tRNA synthetase activity was determined by measurement of the incorporation of 14Clabeled amino acid into liver tRNA as described (20). One unit of arginyl-tRNA synthetase represents the incorporation of 1 nmol of amino acid per min under standard conditions. Electrophoresis and Immunoblotting. SDS/PAGE was carried out with 8.5% acrylamide gels as reported (20). Protein blotting and immunodetection were performed according to the instructions in the Promega kit using IgG prepared against the free form of arginyl-tRNA synthetase (13, 20).
Proc. Natl. Acad. Sci. USA 87 (1990)
.I ) .: __r--
-.
-.-,
1.
k
)il
1\ P.;.
RESULTS AND DISCUSSION The Endogenous Free Form of Arginyl-tRNA Synthetase Differs from the Proteolyticafly Derived Enzyme. ArginyltRNA synthetase differs from most other synthetases in that it is routinely found in both free and complexed forms in extracts of eukaryotic cells (5, 14-16). In rat liver, the ratio of high to low molecular weight enzyme is about 2:1 (5, 16, 17). It was originally thought that the free form might be derived from the complexed enzyme by proteolytic cleavage. First of all, the two proteins were closely related structurally, and second, in vitro treatment of the complex with various proteases released an active fragment of arginyl-tRNA synthetase that was the same size as the free form on gel filtration (17). In addition, since active, lower molecular weight forms of lysyl- and methionyl-tRNA synthetases could also be released from the complex by in vitro limited proteolysis (21, 22), it appeared that these enzymes might contain a specific protease-sensitive site between their catalytic domain and a domain required for complex formation. However, more detailed analysis of the proteolytically derived arginyl-tRNA synthetase has revealed that it actually differs in size from the endogenous free form. SDS/PAGE and immunoblotting of the two proteins showed that the Mr of the free form is -2000-3000 less than that of the proteolytically derived fragment; both forms are substantially smaller than the complexed enzyme (Fig. 1). The size difference between the two smaller forms may also account for the somewhat greater hydrophobicity previously observed for the proteolyzed protein compared to the endogenous free form (17). Thus, although a lower molecular weight arginyltRNA synthetase can be released from the complex by in vitro proteolysis, it is clear from these data that a single highly protease-sensitive site cleaved by both exogenous and endogenous proteases is not responsible for the production of both the in vivo and in vitro low molecular weight enzymes. As a consequence, the previously presumed identity of exogenous and endogenous cleavage sites no longer supports the idea that endogenous proteolysis is the origin of the free form of the protein. Extraction Conditions Do Not Alter the Ratio of Complexed to Low Molecular Weight Arginyl-tRNA Synthetase. Additional evidence that argues against artifactual proteolysis for the origin of the endogenous free enzyme is that extraction of rat liver by using a variety of conditions does not affect the ratio of the complexed form to the free form. In earlier work, conditions as extreme as homogenizing livers in the presence of seven protease inhibitors or incubating the homogenate for 1 hr at room temperature in the absence of protease inhibitors resulted in no change in the ratio of the two forms of the enzyme (5, 17).
FIG. 1. Electrophoresis and immunoblotting of different forms of arginyl-tRNA synthetase. The high molecular weight complex (25 /dl, 2 mg/ml) purified through Controlled-Pore glass (17) was incubated with or without 15 jul of papain (0.25 mg/ml) at 37°C for 25 min in a final volume of 75 ,u. The reaction was stopped by the addition of 10 ,.l of leupeptin (10 mg/ml). Untreated or treated complex (40 1I) was used for the analysis. Lane 1, untreated complex; lane 2, papaintreated complex; lane 3, endogenous free form of arginyl-tRNA synthetase (20 ,g) (16). Samples were electrophoresed and immu-
noblotted as described in Materials and Methods.
As a further test to eliminate extraneous proteolysis of the complex as the source of the free form of arginyl-tRNA synthetase, an extract of the supernatant resulting from centrifugation at 100,000 x g (S-100) was prepared from a liver that was immediately placed in liquid N2 to stop possible degradative reactions. Partially frozen liver was homogenized in a medium containing a mixture of protease inhibitors. After centrifugation, gel infiltration of this material gave the usual pattern of arginyl-tRNA synthetase activity with a ratio of complexed form to free form of about 2:1 (Fig. 2). In another experiment, cultured liver cells were preincubated for 90 min with a mixture of leupeptin (0.2 ,uM), pepstatin A (0.1 AM), and phenylmethylsulfonyl fluoride (500 ,uM). Again, the ratio of low to high molecular weight forms was unaffected by this treatment, despite the fact that these protease inhibitors can act intracellularly to inhibit proteolysis that might occur even prior to opening the cells (data not shown). The finding that these extreme procedures also do not change the proportion of free arginyl-tRNA synthetase lends strong support to the conclusion that this form of the enzyme preexists in the cell and is not generated by adventitious proteolysis subsequent to cell disruption. Low Molecular Weight Arginyl-tRNA Synthetase Is Probably a Distinct Translation Product. NH2-terminal sequence analysis of the endogenous low molecular weight arginyltRNA synthetase provides further evidence for the existence of this form of the enzyme in the cell and, in addition, suggests that it is a separate translation product. The free form was purified as reported (16) and also by SDS/PAGE and then subjected to 20 cycles of gas-phase sequencing (Table 1). Interestingly, in contrast to many eukaryotic proteins, the enzyme was found to be unblocked and, in addition, to begin with an NH2-terminal methionine residue. This finding strongly suggests that the low molecular weight form of the enzyme arises by translation. Since methionine residues represent only 3% of the amino acids in the complex-
Biochemistry: Sivaram and Deutscher
Proc. Natl. Acad. Sci. USA 87 (1990)
3667
Table 1. NH2-terminal sequence analysis of the low molecular weight free form of arginyl-tRNA synthetase Amino acid Cycle Yield, pmol
2
I2s
-
m 1-
4._
D
0
20 40 Fraction Number
60
FIG. 2. Gel filtration of a rat liver S-100 fraction prepared under conditions to minimize proteolysis. The liver was removed from an anesthetized rat, immediately frozen in liquid N2, and left for 30 min. The frozen liver was crumbled with a mortar and pestle, and the small pieces were allowed to partially thaw in an ice-cold medium containing 50 mM Hepes-KOH (pH 7.4), 165 mM potassium acetate (pH 7.0), 3 mM MgCI2, 3 mM glutathione, 0.2 mM EDTA, 0.4 mM phenylmethylsulfonyl fluoride, leupeptin (0.05 ,ug/ml), a2-macroglobulin (0.05 ,g/ml), and 7% (vol/vol) glycerol. The sample was homogenized by hand and centrifuged, and 10 ml of the S-100 fraction was chromatographed on a column (495 ml) of Sephacryl S-300 (17). Fractions of 6 ml were collected, and 15-Al portions were assayed for arginyl-tRNA synthetase activity for 4 min at 37°C.
bound enzyme (13), a proteolytic cleavage that generates an NH2-terminal methionine would be extremely unlikely. Additional evidence suggesting that the free arginyl-tRNA synthetase arises by translation comes from a comparison of its NH2-terminal sequence with the NH2 terminus of the Escherichia coli enzyme (23). As shown in Fig. 3, there is a high degree of homology between these two sequences. Seven of 19 residues (37%) within the sequence are identical, and an additional 7 are related by a single nucleotide change in their respective codons. Such a high level of sequence conservation between these widely separated organisms suggests that this portion of the protein has been maintained relatively intact during evolution and supports the idea that the sequence from the liver enzyme is actually the NH2 terminus. If the suggestion of a distinct translation event for the low molecular weight enzyme is correct, it is likely that both forms of the enzyme, one of which contains over 100 extra NH2-terminal residues but is otherwise identical to the free enzyme (13), arise from a single gene. This could be accomplished either by synthesis of two mRNAs from different transcription initiation sites or by a single mRNA with two alternate initiation codons. Several examples are now known in which two forms of an aminoacyl-tRNA synthetase are produced from a single gene (24, 25). In these cases, which were observed in yeast, transcription is initiated from two sites, leading to two mRNAs: the longer one encodes the mitochondrial form of the enzyme, and the 5' shortened one encodes the cytoplasmic protein. The extra NH2-terminal domains on the longer proteins are thought to target these forms to the mitochondria. Alternatively, there are also some situations in which two proteins arise from a single mRNA by Rat liver
E.coli
Met 1 97 2 88 lie 3 22 Asn 4 Ile 36 5 14 Asn 6 Ser 18 7 Xaa 17 8 Leu 16 9 Gln 10 14 Glu 11 Leu 17 12 Phe 19 13 17 Gly 14 Xaa 15 16 Ala 16 Ile 13 17 10 Lys 18 10 Ala 19 13 Ala 20 16 Tyr Arginyl-tRNA synthetase was purified as described (16) followed by further purification by SDS/PAGE on a 1o acrylamide gel. The sample was transferred to an Immobilon membrane (Millipore); the band corresponding to arginyl-tRNA synthetase was identified by immunoblotting and cut out for sequencing. Gas-phase sequencing was carried out on an Applied Biosystems model 470A sequencer equipped with a model 120A phenylthiohydantoin analyzer for 20 cycles. The equivalent "180 pmol of pure arginyl-tRNA synthetase protein was loaded on the gel. Xaa, amino acid could not be identified.
alternate translation initiation events from two different in-frame start codons (26, 27). Taken together, all the data presented here provide substantial support for the conclusion that mammalian cells contain two forms of arginyl-tRNA synthetase, one associated with other aminoacyl-tRNA synthetases and one free, and that the latter form arises from a distinct translation event. Although further work at the gene and mRNA levels will be necessary to conclusively prove these points, these conclusions fit well with the model presented below, which provides a consistent explanation for this unusual situation. Rationale for the Existence of Two Forms of Arginyl-tRNA Synthetase. For almost all other aminoacyl-tRNA synthetases, only a single form of each enzyme is found in cytoplasmic extracts of mammalian cells. What then might explain the presence of two arginyl-tRNA synthetases in these cells? One explanation, which reconciles a number of disparate observations about arginyl-tRNA metabolism, has important consequences for our understanding of protein biosynthesis. It has been known for years that all eukaryotic cells contain an arginyl-tRNA: protein transferase that catalyzes the posttranslational addition of arginine residues to the NH2 termini of certain acceptor proteins (28, 29). The enzyme utilizes all isoacceptors of arginyl-tRNA as the donor, but only proteins with NH2 terminal glutamic, aspartic, or cysteine residues can serve as acceptors (29). Until recently, the function of
Met Ile Asn Ile Asn Ser X Leu Gln Glu Leu Phe Gly X Ala Ile Lys Ala Ala Tyr
I1 1I I 1
1 I
1 11 11 11
Met Asn Ile Gln Ala Leu Leu Ser Glu Lys Val Arg Gln Ala Met Ile Ala Ala Gly
FIG. 3. Comparison of the NH2-terminal sequences of the rat liver low molecular weight synthetase and theE. coli arginyl-tRNA synthetase. The first 20 amino acids of the liver enzyme are aligned with the NH2-terminal 19 amino acids of the E. coli enzyme. Identities are denoted by double lines, and residues related by a single nucleotide change in the respective codons are denoted by a single line.
3668
Biochemistry: Sivaram and Deutscher
this unusual enzyme was unknown. However, several studies have now implicated this enzyme in the ubiquitin-mediated proteolytic degradation of certain proteins (18, 30, 31). Proteins destined for degradation by this pathway are modified by conjugation with ubiquitin on E-NH2 groups of lysine (32). An important determinant for reaction with ubiquitin is the identity of the targeted protein's NH2terminal amino acid (31). For proteins with an acidic NH2terminal residue, the presence of tRNA is required for ubiquitin conjugation (33), and this requirement is a consequence of the addition of an arginine residue to the NH2 termini of these proteins by arginyl-tRNA:protein transferase (18). Proteins with NH2-terminal arginine residues are among those most rapidly degraded by the ubiquitindependent system, and the arginine modification reaction serves as a signal for the rapid addition of ubiquitin (18, 31). It is now clear that in reticulocyte extracts the arginine modification reaction targets for degradation those proteins containing NH2-terminal glutamic, aspartic, cysteine, glutamine, or asparagine residues (31). Furthermore, the requirement for tRNA in this system can be accounted for completely by the arginyltransferase reaction (30, 31). On the basis of these considerations, arginyl-tRNA may be unique among mammalian aminoacyl-tRNAs because it is needed for a process other than protein biosynthesis. The existence of two forms of arginyl-tRNA synthetase in mammalian cells may be directly related to this requirement for a supply of arginyl-tRNA for two distinct processes. We propose that two arginyl-tRNA synthetases are necessary because there are two separate pools of arginyl-tRNA, one for protein synthesis and one for the arginine modification reaction. If one or both of these pools were sequestered and unavailable for other processes, the cell would require separate synthetases to keep each of the pools supplied with arginyl-tRNA. One conclusion that is suggested by such a model is that arginyl-tRNA, and presumably other aminoacyl-tRNAs, are channeled for protein synthesis, transferring directly from the aminoacyl-tRNA synthetase to the
Proc. Natl. Acad Sci. USA 87 (1990)
elongation factor, to the ribosome, and back to the synthetase without mixing with the bulk cytoplasm (Fig. 4). If correct, this would have important consequences for our understanding of protein synthesis in higher organisms. Channeling for Protein Synthesis. Despite extensive study of protein biosynthesis, relatively little attention has been given to an examination of channeling in this process. Nonetheless, several pieces of evidence support the conclusion that channeling does occur. Of particular relevance are the numerous observations that components of the proteinsynthesizing apparatus are organized. Such organization would provide a structural basis for channeling. These organized structures include the complexes of aminoacyl-tRNA synthetases described above, the association of aminoacyltRNA synthetases with ribosomes (34-36) and elongation factor (37), and the binding of many protein synthesis components to the cytoskeletal framework of the cell (38-41). In addition, there is evidence for a functional interaction between aminoacyl-tRNA synthetases and initiation factors (42). A number of other studies have also shown that exogenously supplied amino acids can enter the pathway of protein synthesis without mixing with the free amino acid pool of the cell (43-46), implying that channeling may begin early in the process. Two other studies also suggest that exogenously supplied tRNAs are not used for protein synthesis in intact systems. In one, tRNAPhe injected into Xenopus oocytes was a very poor substrate for aminoacylation by the endogenous aminoacyl-tRNA synthetase despite the fact that the same tRNA was an active substrate for this enzyme in extracts of these cells (47). The conclusion drawn from this experiment was that the endogenous aminoacyl-tRNA synthetase is compartmentalized and not available to exogenous tRNA. In a second study, a gently prepared protein-synthesizing system from rabbit reticulocytes was shown to incorporate radioactive amino acid with a linear time course, whereas incorporation from labeled aminoacyl-tRNA proceeded only after a sub-
Arginine
AA-tRNA Synthetase Complex (72 kDa Arg-tRNA synthetase) (bound tRNA)
[ Arg-tRNA ] EFi
Ribosome
Newly-synthesized protein
Arginine
C y T 0 p L A S M
tRNA 9
$
Arginyl-tRNA Synthetase ( 60 kDa free form )
Arg-tRNA
Arg-tRNA-Protein transferase
N-terminal-Arg- modified protein
Ubiquitin-dependent degradation FIG. 4. Model to explain the existence of two forms of arginyl-tRNA synthetase. Arginine that enters the high molecular weight aminoacyl-tRNA synthetase complex is added to endogenous tRNA by the 72-kDa form of arginyl-tRNA synthetase. The arginyl-tRNA is directly transferred to elongation factor 1 (EF1) and the ribosome and incorporated into protein without mixing with the bulk cytoplasm. Arginine interacting with the 60-kDa free form of arginyl-tRNA synthetase is added to free tRNAArg to generate arginyl-tRNA that is used by arginyl-tRNA-protein transferase to modify the NH2 terminus of certain proteins destined for ubiquitin-dependent protein degradation.
Biochemistry: Sivaram and Deutscher stantial lag (48). The conclusion from this work was that the aminoacyl-tRNA was first deacylated and that the free amino acid was then incorporated, bypassing use of the exogenous tRNA. Both of these studies are consistent with the idea that endogenous aminoacyl-tRNAs and synthetases are sequestered and that exogenous macromolecules do not enter the protein-synthesizing apparatus. Obviously, this is not true for the many disrupted systems studied over the years in which exogenously supplied aminoacyl-tRNAs can function in protein synthesis. The data and model presented here raise many interesting questions about protein synthesis in vivo and in vitro. In particular, what is the nature of the organization of the protein-synthesizing machinery and what is its location within the cell, and second, do in vitro protein-synthesizing systems accurately reflect the in vivo situation when the organization has been disrupted and when large excesses of exogenous tRNA or aminoacyl-tRNA are added to the system? A detailed understanding of protein synthesis will need to consider how organization of the various components and how channeling of the intermediates might influence the process in vivo. We thank Dr. Gary Vellekamp for some initial studies on this problem and George Korza and Dr. Juris Ozols for the amino acid sequence analyses. The technical assistance of Martin Worrall is greatly appreciated. This work was supported by Grant GM16317 from the National Institutes of Health. 1. Srere, P. A. (1987) Annu. Rev. Biochem. 56, 89-124. 2. Deutscher, M. P. (1984) J. Cell Biol. 99, 373-377. 3. Yang, D. C. H., Garcia, J. V., Johnson, Y. D. & Wahab, S. (1985) Curr. Top. Cell. Regul. 26, 325-335. 4. Mirande, M., LeCorre, D. & Waller, J.-P. (1985) Eur. J. Biochem. 147, 281-289. 5. Cirakaglu, B. & Waller, J.-P. (1985) Biochim. Biophys. Acta 829, 173-179. 6. Godar, D. E., Godar, D. E., Garcia, V., Jacabo, A., Aebi, U. & Yang, D. C. H. (1988) Biochemistry 27, 6921-6928. 7. Norcum, M. T. (1989) J. Biol. Chem. 264, 15043-15051. 8. Bandyopadhyay, A. K. & Deutscher, M. P. (1971) J. Mol. Biol. 60, 113-122. 9. Walker, E. J., Trecy, G. B. & Jeffrey, P. D. (1983) Biochemistry 22, 1934-1941. 10. Dang, C. V. & Yang, D. C. H. (1979) J. Biol. Chem. 254, 5350-5356. 11. Sihag, R. K. & Deutscher, M. P. (1983) J. Biol. Chem. 258, 11846-11850. 12. Cirakoglu, B., Mirande, M. & Waller, J.-P. (1985) FEBS Lett. 183, 185-190. 13. Vellekamp, G. & Deutscher, M. P. (1987) J. Biol. Chem. 262, 9927-9930. 14. Som, K. & Hardesty, B. (1975) Arch. Biochem. Biophys. 166, 507-517. 15. Ritter, P. O., Enger, M. D. & Hampel, A. E. (1979) Biochim. Biophys. Acta 562, 377-385.
Proc. Natl. Acad. Sci. USA 87 (1990)
3669
16. Deutscher, M. P. & Ni, R. C. (1982) J. Biol. Chem. 257, 6003-6006. 17. Vellekamp, G., Sihag, R. K. & Deutscher, M. P. (1985) J. Biol. Chem. 260, 9843-9847. 18. Ferber, S. & Ciechanover, A. (1987) Nature (London) 326, 808-811. 19. Deutscher, M. P. (1972) J. Biol. Chem. 247, 459-468. 20. Sivaram, P., Vellekamp, G. & Deutscher, M. P. (1988) J. Biol. Chem. 263, 18891-188%. 21. Kellerman, O., Viel, C. & Waller, J.-P. (1978) Eur. J. Biochem. 88, 197-204. 22. Mirande, M., Cirakoglu, B. & Waller, J.-P. (1983) Eur. J. Biochem. 131, 163-170. 23. Eriani, G., Dirheimer, G. & Gangloff, J. (1989) Nucleic Acids Res. 17, 5725-5736. 24. Chatton, B., Walter, P., Ebel, J.-P., Lacroute, F. & Fasiolo, F.
(1988) J. Biol. Chem. 263, 52-57.
25. Natsoulis, G., Hilger, F. & Fink, G. R. (1986) Cell 46, 235-243. 26. Conneely, 0. M., Kettelberger, D. M., Tsai, M.-J., Schrader, W. T. & O'Malley, B. W. (1989) J. Biol. Chem. 264, 1406214064. 27. Spence, A. M., Sheppard, P. C., Davie, J. R., Matuo, Y., Nishi, N., McKeehan, W. L., Dodd, J. G. & Matusik, R. J. (1989) Proc. Natl. Acad. Sci. USA 86, 7843-7847. 28. Kaji, H., Novelli, G. D. & Kaji, A. (1963) Biochim. Biophys. Acta 76, 474-477. 29. Deutch, C. E., Scarpulla, R. C. & Soffer, R. L. (1978) Curr. Top. Cell. Regul. 13, 1-28. 30. Ciechanover, A., Ferber, S., Ganoth, D., Elias, S., Hershko, A. & Arfin, S. (1988) J. Biol. Chem. 263, 11155-11167. 31. Gonda, D. K., Bachmair, A., Wunning, I., Tobias, J. W., Lane, W. S. & Varshavsky, A. (1989) J. Biol. Chem. 264, 16700-16712. 32. Rechsteiner, M. (1987) Annu. Rev. Cell Biol. 3, 1-30. 33. Ferber, S. & Ciechanover, A. (1986) J. Biol. Chem. 261, 3128-3134. 34. Irvin, J. D. & Hardesty, B. (1972) Biochemistry 11, 1915-1920. 35. Moline, G., Hampel, A. & Enger, M. D. (1974) Biochem. J. 143, 191-195. 36. Graf, H. (1976) Biochim. Biophys. Acta 425, 175-184. 37. Bec, G., Kerjan, P., Zha, X. D. & Waller, J.-P. (1989) J. Biol. Chem. 264, 21131-21137. 38. Nielson, P., Goelz, S. & Trachsel, H. (1983) CellBiol. Int. Rep. 7, 245-254. 39. Howe, J. G. & Hershey, J. W. B. (1984) Cell 37, 85-93. 40. Bonneau, A.-M., Darveau, A. & Sonenberg, N. (1985) J. Cell Biol. 100, 1209-1218. 41. Orrelles, D. A., Fey, E. G. & Penman, S. (1986) Mol. Cell. Biol. 6, 1650-1662. 42. Pollard, J. W., Galpine, A. R. & Clemens, M. J. (1989) Eur. J. Biochem. 182, 1-9. 43. vanVenrooij, W. J., Moonen, H. & vanLoon-Klaassen, L. (1974) Eur. J. Biochem. 50, 297-304. 44. Airhart, J., Vidrieh, A. & Khairallah, E. A. (1974) Biochem. J. 140, 539-548. 45. Hod, Y. & Hershko, A. (1976) J. Biol. Chem. 251, 4458-4467. 46. Gehrke, L. & Ilan, J. (1983) Proc. Natl. Acad. Sci. USA 80, 3274-3278. 47. Gatica, M., Allende, C. C. & Allende, J. E. (1980) Arch. Biochem. Biophys. 202, 653-656. 48. Hradec, J. & Dusek, Z. (1978) Biochem. J. 172, 1-7.
Vol. 87, pp. 3665-3669, May 1990 Biochemistry
Existence of two forms of rat liver arginyl-tRNA synthetase suggests channeling of aminoacyl-tRNA for protein synthesis (ubiquitin-dependent proteolysis/arginyl-tRNA protein transferase)
PILLARISETTI SIVARAM AND MURRAY P. DEUTSCHER Department of Biochemistry, University of Connecticut Health Center, Farmington, CT 06032
Communicated by Mary J. Osborn, March 1, 1990
Arginyl-tRNA synthetase (arginine-tRNA liABSTRACT gase, EC 6.1.1.19) is found in extracts of mammalian cells both as a free protein (Mr = 60,000) and as a component (Mr 72,000) of the high molecular weight aminoacyl-tRNA synthetase complex (Mr > 106). Several pieces of evidence indicate that the low molecular weight free form is not a proteolytic degradation product of the complex-bound enzyme but that it preexists in vivo: (i) the endogenous free form differs in size from the active proteolytic fragment generated in vitro, (ii) conditions expected to increase or decrease the amount of proteolysis do not alter the ratio of the two forms of the enzyme, and (iu) the free form contains an NH2-terminal methionine residue. A model is presented that provides a rationale for the existence of two forms of arginyl-tRNA synthetase in cells. In this model the complexed enzyme supplies arginyl-tRNA for protein synthesis, whereas the free enzyme provides arginyltRNA for the N112-terminal arginine modification of proteins by arginyl-tRNA: protein arginyltransferase. This latter process targets certain proteins for removal by the ubiquitindependent protein degradation pathway. The necessity for an additional pool of arginyl-tRNA for the modifcation reaction leads to the conclusion that the arginyl-tRNA destined for protein synthesis (and/or protein modification) is channeled and unavailable for other processes. Other evidence supporting channeling in protein synthesis is discussed. It is becoming increasingly clear that many, if not all, of the macromolecular components of cells are organized into multienzyme complexes or associated with subcellular structures (see ref. 1 for a review). A major consequence of such organization is that it can lead to compartmentalization or channeling of metabolic pathways (i.e., the transfer of metabolites from one enzyme to another without equilibration with the total fluid of the cell). Channeled pathways presumably would be more efficient than free ones by increasing the local substrate concentrations for a given amount of substrate, by minimizing destruction of unstable intermediates or loss due to side reactions, and by allowing for combined regulation of multiple activities. The protein biosynthetic machinery is among the most complex in the cell; it consists of a large number of protein and nucleic acid components. Organization of these various components has been studied most extensively in the case of the aminoacyl-tRNA synthetases. In extracts of higher eukaryotic cells, these enzymes are generally found in high molecular weight multienzyme complexes, often containing other components as well (2, 3). The number of synthetases found in these complexes can vary depending on the method of isolation, but as many as eight or nine are routinely found associated with each other (4-7), and even higher numbers have been observed (8, 9). It is now known that hydrophobic The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
3665
interactions play an important role in holding aminoacyltRNA synthetases in the complex (10-12) and that terminal extensions on the individual proteins, not required for catalysis, participate in these associations (13). For a few aminoacyl-tRNA synthetases (namely, those for arginine and aspartic acid), both high and low molecular weight forms of the enzyme coexist in the same extract (14, 15). Arginyl-tRNA synthetase (arginine-tRNA ligase, EC 6.1.1.19) is present in extracts of rat liver as a free protein (Mr 60,000) as well as a component (Mr 72,000) of the high molecular weight complex (Mr > 106) (5, 16). Comparison of the catalytic and immunological properties of the free and complexed forms of this enzyme initially suggested that they are closely related proteins (17), and peptide mapping indicated that they are probably identical except for a Mr = 12,000 NH2-terminal extension present on the complexed form (13). On the basis of a comparison of the endogenous free form of arginyl-tRNA synthetase with a proteolytically derived fragment released from the complex in vitro, it was thought that the free form might have arisen by a limited cleavage of the complex-bound protein by some endogenous protease (13, 17). However, several pieces of evidence did not fit with this idea (17), and the question of whether proteolysis occurs during isolation or whether the free form preexists in the cell was also left open. We have now examined this issue in more detail. Our resUlts indicate that the endogenous free form differs from the in vitro proteolytically derived enzyme, that the free form probably preexists in the cell, and that most likely it is a distinct translation product. The existence of two forms of arginyl-tRNA synthetase in mammalian cells, coupled with the recently discovered role of arginyl-tRNA in ubiquitindependent protein degradation (18), leads us to propose a model that has important consequences with regard to the channeling of aminoacyl-tRNAs for protein biosynthesis. =
MATERIALS AND METHODS Materials. Sephacryl S-300 was obtained from PharmaciaLKB Biotechnology. The protease inhibitors leupeptin, antipain, pepstatin A, and phenylmethylsulfonyl fluoride were purchased from Sigma. a2-Macroglobulin was from Boehringer Mannheim. [14C]Arginine was obtained from ICN. Rabbit livers (from young, fasted animals) and rat livers (from adult, fasted animals) were purchased from Pel-Freez Biologicals. Sprague-Dawley female rats (-200 g) were obtained from Charles River Breeding Laboratories. Normal rat liver cells (clone 9) were obtained from the American Type Culture Collection. They were maintained as monolayers in Ham's nutrient mixture F-12 with 10o (vol/vol) fetal bovine serum. Rabbit liver tRNA was prepared as described (19). Reagents for immunoblotting were products of Promega. All other materials were as reported (20). Preparation of Arginyl-tRNA Synthetases. Rat liver extracts for gel filtration were prepared from fresh liver by homoge-
3666
Biochemistry: Sivaram and Deutscher
nization and high-speed centrifugation (see the legend to Fig. 2). The low molecular weight free form of arginyl-tRNA synthetase was purified from frozen livers as described by Deutscher and Ni (16). The proteolyzed form of arginyltRNA synthetase was prepared from the high molecular weight complex purified through the Controlled-Pore glass (Electro-Nucleonics) column step followed by treatment with papain (17). Enzyme Assays. Aminoacyl-tRNA synthetase activity was determined by measurement of the incorporation of 14Clabeled amino acid into liver tRNA as described (20). One unit of arginyl-tRNA synthetase represents the incorporation of 1 nmol of amino acid per min under standard conditions. Electrophoresis and Immunoblotting. SDS/PAGE was carried out with 8.5% acrylamide gels as reported (20). Protein blotting and immunodetection were performed according to the instructions in the Promega kit using IgG prepared against the free form of arginyl-tRNA synthetase (13, 20).
Proc. Natl. Acad. Sci. USA 87 (1990)
.I ) .: __r--
-.
-.-,
1.
k
)il
1\ P.;.
RESULTS AND DISCUSSION The Endogenous Free Form of Arginyl-tRNA Synthetase Differs from the Proteolyticafly Derived Enzyme. ArginyltRNA synthetase differs from most other synthetases in that it is routinely found in both free and complexed forms in extracts of eukaryotic cells (5, 14-16). In rat liver, the ratio of high to low molecular weight enzyme is about 2:1 (5, 16, 17). It was originally thought that the free form might be derived from the complexed enzyme by proteolytic cleavage. First of all, the two proteins were closely related structurally, and second, in vitro treatment of the complex with various proteases released an active fragment of arginyl-tRNA synthetase that was the same size as the free form on gel filtration (17). In addition, since active, lower molecular weight forms of lysyl- and methionyl-tRNA synthetases could also be released from the complex by in vitro limited proteolysis (21, 22), it appeared that these enzymes might contain a specific protease-sensitive site between their catalytic domain and a domain required for complex formation. However, more detailed analysis of the proteolytically derived arginyl-tRNA synthetase has revealed that it actually differs in size from the endogenous free form. SDS/PAGE and immunoblotting of the two proteins showed that the Mr of the free form is -2000-3000 less than that of the proteolytically derived fragment; both forms are substantially smaller than the complexed enzyme (Fig. 1). The size difference between the two smaller forms may also account for the somewhat greater hydrophobicity previously observed for the proteolyzed protein compared to the endogenous free form (17). Thus, although a lower molecular weight arginyltRNA synthetase can be released from the complex by in vitro proteolysis, it is clear from these data that a single highly protease-sensitive site cleaved by both exogenous and endogenous proteases is not responsible for the production of both the in vivo and in vitro low molecular weight enzymes. As a consequence, the previously presumed identity of exogenous and endogenous cleavage sites no longer supports the idea that endogenous proteolysis is the origin of the free form of the protein. Extraction Conditions Do Not Alter the Ratio of Complexed to Low Molecular Weight Arginyl-tRNA Synthetase. Additional evidence that argues against artifactual proteolysis for the origin of the endogenous free enzyme is that extraction of rat liver by using a variety of conditions does not affect the ratio of the complexed form to the free form. In earlier work, conditions as extreme as homogenizing livers in the presence of seven protease inhibitors or incubating the homogenate for 1 hr at room temperature in the absence of protease inhibitors resulted in no change in the ratio of the two forms of the enzyme (5, 17).
FIG. 1. Electrophoresis and immunoblotting of different forms of arginyl-tRNA synthetase. The high molecular weight complex (25 /dl, 2 mg/ml) purified through Controlled-Pore glass (17) was incubated with or without 15 jul of papain (0.25 mg/ml) at 37°C for 25 min in a final volume of 75 ,u. The reaction was stopped by the addition of 10 ,.l of leupeptin (10 mg/ml). Untreated or treated complex (40 1I) was used for the analysis. Lane 1, untreated complex; lane 2, papaintreated complex; lane 3, endogenous free form of arginyl-tRNA synthetase (20 ,g) (16). Samples were electrophoresed and immu-
noblotted as described in Materials and Methods.
As a further test to eliminate extraneous proteolysis of the complex as the source of the free form of arginyl-tRNA synthetase, an extract of the supernatant resulting from centrifugation at 100,000 x g (S-100) was prepared from a liver that was immediately placed in liquid N2 to stop possible degradative reactions. Partially frozen liver was homogenized in a medium containing a mixture of protease inhibitors. After centrifugation, gel infiltration of this material gave the usual pattern of arginyl-tRNA synthetase activity with a ratio of complexed form to free form of about 2:1 (Fig. 2). In another experiment, cultured liver cells were preincubated for 90 min with a mixture of leupeptin (0.2 ,uM), pepstatin A (0.1 AM), and phenylmethylsulfonyl fluoride (500 ,uM). Again, the ratio of low to high molecular weight forms was unaffected by this treatment, despite the fact that these protease inhibitors can act intracellularly to inhibit proteolysis that might occur even prior to opening the cells (data not shown). The finding that these extreme procedures also do not change the proportion of free arginyl-tRNA synthetase lends strong support to the conclusion that this form of the enzyme preexists in the cell and is not generated by adventitious proteolysis subsequent to cell disruption. Low Molecular Weight Arginyl-tRNA Synthetase Is Probably a Distinct Translation Product. NH2-terminal sequence analysis of the endogenous low molecular weight arginyltRNA synthetase provides further evidence for the existence of this form of the enzyme in the cell and, in addition, suggests that it is a separate translation product. The free form was purified as reported (16) and also by SDS/PAGE and then subjected to 20 cycles of gas-phase sequencing (Table 1). Interestingly, in contrast to many eukaryotic proteins, the enzyme was found to be unblocked and, in addition, to begin with an NH2-terminal methionine residue. This finding strongly suggests that the low molecular weight form of the enzyme arises by translation. Since methionine residues represent only 3% of the amino acids in the complex-
Biochemistry: Sivaram and Deutscher
Proc. Natl. Acad. Sci. USA 87 (1990)
3667
Table 1. NH2-terminal sequence analysis of the low molecular weight free form of arginyl-tRNA synthetase Amino acid Cycle Yield, pmol
2
I2s
-
m 1-
4._
D
0
20 40 Fraction Number
60
FIG. 2. Gel filtration of a rat liver S-100 fraction prepared under conditions to minimize proteolysis. The liver was removed from an anesthetized rat, immediately frozen in liquid N2, and left for 30 min. The frozen liver was crumbled with a mortar and pestle, and the small pieces were allowed to partially thaw in an ice-cold medium containing 50 mM Hepes-KOH (pH 7.4), 165 mM potassium acetate (pH 7.0), 3 mM MgCI2, 3 mM glutathione, 0.2 mM EDTA, 0.4 mM phenylmethylsulfonyl fluoride, leupeptin (0.05 ,ug/ml), a2-macroglobulin (0.05 ,g/ml), and 7% (vol/vol) glycerol. The sample was homogenized by hand and centrifuged, and 10 ml of the S-100 fraction was chromatographed on a column (495 ml) of Sephacryl S-300 (17). Fractions of 6 ml were collected, and 15-Al portions were assayed for arginyl-tRNA synthetase activity for 4 min at 37°C.
bound enzyme (13), a proteolytic cleavage that generates an NH2-terminal methionine would be extremely unlikely. Additional evidence suggesting that the free arginyl-tRNA synthetase arises by translation comes from a comparison of its NH2-terminal sequence with the NH2 terminus of the Escherichia coli enzyme (23). As shown in Fig. 3, there is a high degree of homology between these two sequences. Seven of 19 residues (37%) within the sequence are identical, and an additional 7 are related by a single nucleotide change in their respective codons. Such a high level of sequence conservation between these widely separated organisms suggests that this portion of the protein has been maintained relatively intact during evolution and supports the idea that the sequence from the liver enzyme is actually the NH2 terminus. If the suggestion of a distinct translation event for the low molecular weight enzyme is correct, it is likely that both forms of the enzyme, one of which contains over 100 extra NH2-terminal residues but is otherwise identical to the free enzyme (13), arise from a single gene. This could be accomplished either by synthesis of two mRNAs from different transcription initiation sites or by a single mRNA with two alternate initiation codons. Several examples are now known in which two forms of an aminoacyl-tRNA synthetase are produced from a single gene (24, 25). In these cases, which were observed in yeast, transcription is initiated from two sites, leading to two mRNAs: the longer one encodes the mitochondrial form of the enzyme, and the 5' shortened one encodes the cytoplasmic protein. The extra NH2-terminal domains on the longer proteins are thought to target these forms to the mitochondria. Alternatively, there are also some situations in which two proteins arise from a single mRNA by Rat liver
E.coli
Met 1 97 2 88 lie 3 22 Asn 4 Ile 36 5 14 Asn 6 Ser 18 7 Xaa 17 8 Leu 16 9 Gln 10 14 Glu 11 Leu 17 12 Phe 19 13 17 Gly 14 Xaa 15 16 Ala 16 Ile 13 17 10 Lys 18 10 Ala 19 13 Ala 20 16 Tyr Arginyl-tRNA synthetase was purified as described (16) followed by further purification by SDS/PAGE on a 1o acrylamide gel. The sample was transferred to an Immobilon membrane (Millipore); the band corresponding to arginyl-tRNA synthetase was identified by immunoblotting and cut out for sequencing. Gas-phase sequencing was carried out on an Applied Biosystems model 470A sequencer equipped with a model 120A phenylthiohydantoin analyzer for 20 cycles. The equivalent "180 pmol of pure arginyl-tRNA synthetase protein was loaded on the gel. Xaa, amino acid could not be identified.
alternate translation initiation events from two different in-frame start codons (26, 27). Taken together, all the data presented here provide substantial support for the conclusion that mammalian cells contain two forms of arginyl-tRNA synthetase, one associated with other aminoacyl-tRNA synthetases and one free, and that the latter form arises from a distinct translation event. Although further work at the gene and mRNA levels will be necessary to conclusively prove these points, these conclusions fit well with the model presented below, which provides a consistent explanation for this unusual situation. Rationale for the Existence of Two Forms of Arginyl-tRNA Synthetase. For almost all other aminoacyl-tRNA synthetases, only a single form of each enzyme is found in cytoplasmic extracts of mammalian cells. What then might explain the presence of two arginyl-tRNA synthetases in these cells? One explanation, which reconciles a number of disparate observations about arginyl-tRNA metabolism, has important consequences for our understanding of protein biosynthesis. It has been known for years that all eukaryotic cells contain an arginyl-tRNA: protein transferase that catalyzes the posttranslational addition of arginine residues to the NH2 termini of certain acceptor proteins (28, 29). The enzyme utilizes all isoacceptors of arginyl-tRNA as the donor, but only proteins with NH2 terminal glutamic, aspartic, or cysteine residues can serve as acceptors (29). Until recently, the function of
Met Ile Asn Ile Asn Ser X Leu Gln Glu Leu Phe Gly X Ala Ile Lys Ala Ala Tyr
I1 1I I 1
1 I
1 11 11 11
Met Asn Ile Gln Ala Leu Leu Ser Glu Lys Val Arg Gln Ala Met Ile Ala Ala Gly
FIG. 3. Comparison of the NH2-terminal sequences of the rat liver low molecular weight synthetase and theE. coli arginyl-tRNA synthetase. The first 20 amino acids of the liver enzyme are aligned with the NH2-terminal 19 amino acids of the E. coli enzyme. Identities are denoted by double lines, and residues related by a single nucleotide change in the respective codons are denoted by a single line.
3668
Biochemistry: Sivaram and Deutscher
this unusual enzyme was unknown. However, several studies have now implicated this enzyme in the ubiquitin-mediated proteolytic degradation of certain proteins (18, 30, 31). Proteins destined for degradation by this pathway are modified by conjugation with ubiquitin on E-NH2 groups of lysine (32). An important determinant for reaction with ubiquitin is the identity of the targeted protein's NH2terminal amino acid (31). For proteins with an acidic NH2terminal residue, the presence of tRNA is required for ubiquitin conjugation (33), and this requirement is a consequence of the addition of an arginine residue to the NH2 termini of these proteins by arginyl-tRNA:protein transferase (18). Proteins with NH2-terminal arginine residues are among those most rapidly degraded by the ubiquitindependent system, and the arginine modification reaction serves as a signal for the rapid addition of ubiquitin (18, 31). It is now clear that in reticulocyte extracts the arginine modification reaction targets for degradation those proteins containing NH2-terminal glutamic, aspartic, cysteine, glutamine, or asparagine residues (31). Furthermore, the requirement for tRNA in this system can be accounted for completely by the arginyltransferase reaction (30, 31). On the basis of these considerations, arginyl-tRNA may be unique among mammalian aminoacyl-tRNAs because it is needed for a process other than protein biosynthesis. The existence of two forms of arginyl-tRNA synthetase in mammalian cells may be directly related to this requirement for a supply of arginyl-tRNA for two distinct processes. We propose that two arginyl-tRNA synthetases are necessary because there are two separate pools of arginyl-tRNA, one for protein synthesis and one for the arginine modification reaction. If one or both of these pools were sequestered and unavailable for other processes, the cell would require separate synthetases to keep each of the pools supplied with arginyl-tRNA. One conclusion that is suggested by such a model is that arginyl-tRNA, and presumably other aminoacyl-tRNAs, are channeled for protein synthesis, transferring directly from the aminoacyl-tRNA synthetase to the
Proc. Natl. Acad Sci. USA 87 (1990)
elongation factor, to the ribosome, and back to the synthetase without mixing with the bulk cytoplasm (Fig. 4). If correct, this would have important consequences for our understanding of protein synthesis in higher organisms. Channeling for Protein Synthesis. Despite extensive study of protein biosynthesis, relatively little attention has been given to an examination of channeling in this process. Nonetheless, several pieces of evidence support the conclusion that channeling does occur. Of particular relevance are the numerous observations that components of the proteinsynthesizing apparatus are organized. Such organization would provide a structural basis for channeling. These organized structures include the complexes of aminoacyl-tRNA synthetases described above, the association of aminoacyltRNA synthetases with ribosomes (34-36) and elongation factor (37), and the binding of many protein synthesis components to the cytoskeletal framework of the cell (38-41). In addition, there is evidence for a functional interaction between aminoacyl-tRNA synthetases and initiation factors (42). A number of other studies have also shown that exogenously supplied amino acids can enter the pathway of protein synthesis without mixing with the free amino acid pool of the cell (43-46), implying that channeling may begin early in the process. Two other studies also suggest that exogenously supplied tRNAs are not used for protein synthesis in intact systems. In one, tRNAPhe injected into Xenopus oocytes was a very poor substrate for aminoacylation by the endogenous aminoacyl-tRNA synthetase despite the fact that the same tRNA was an active substrate for this enzyme in extracts of these cells (47). The conclusion drawn from this experiment was that the endogenous aminoacyl-tRNA synthetase is compartmentalized and not available to exogenous tRNA. In a second study, a gently prepared protein-synthesizing system from rabbit reticulocytes was shown to incorporate radioactive amino acid with a linear time course, whereas incorporation from labeled aminoacyl-tRNA proceeded only after a sub-
Arginine
AA-tRNA Synthetase Complex (72 kDa Arg-tRNA synthetase) (bound tRNA)
[ Arg-tRNA ] EFi
Ribosome
Newly-synthesized protein
Arginine
C y T 0 p L A S M
tRNA 9
$
Arginyl-tRNA Synthetase ( 60 kDa free form )
Arg-tRNA
Arg-tRNA-Protein transferase
N-terminal-Arg- modified protein
Ubiquitin-dependent degradation FIG. 4. Model to explain the existence of two forms of arginyl-tRNA synthetase. Arginine that enters the high molecular weight aminoacyl-tRNA synthetase complex is added to endogenous tRNA by the 72-kDa form of arginyl-tRNA synthetase. The arginyl-tRNA is directly transferred to elongation factor 1 (EF1) and the ribosome and incorporated into protein without mixing with the bulk cytoplasm. Arginine interacting with the 60-kDa free form of arginyl-tRNA synthetase is added to free tRNAArg to generate arginyl-tRNA that is used by arginyl-tRNA-protein transferase to modify the NH2 terminus of certain proteins destined for ubiquitin-dependent protein degradation.
Biochemistry: Sivaram and Deutscher stantial lag (48). The conclusion from this work was that the aminoacyl-tRNA was first deacylated and that the free amino acid was then incorporated, bypassing use of the exogenous tRNA. Both of these studies are consistent with the idea that endogenous aminoacyl-tRNAs and synthetases are sequestered and that exogenous macromolecules do not enter the protein-synthesizing apparatus. Obviously, this is not true for the many disrupted systems studied over the years in which exogenously supplied aminoacyl-tRNAs can function in protein synthesis. The data and model presented here raise many interesting questions about protein synthesis in vivo and in vitro. In particular, what is the nature of the organization of the protein-synthesizing machinery and what is its location within the cell, and second, do in vitro protein-synthesizing systems accurately reflect the in vivo situation when the organization has been disrupted and when large excesses of exogenous tRNA or aminoacyl-tRNA are added to the system? A detailed understanding of protein synthesis will need to consider how organization of the various components and how channeling of the intermediates might influence the process in vivo. We thank Dr. Gary Vellekamp for some initial studies on this problem and George Korza and Dr. Juris Ozols for the amino acid sequence analyses. The technical assistance of Martin Worrall is greatly appreciated. This work was supported by Grant GM16317 from the National Institutes of Health. 1. Srere, P. A. (1987) Annu. Rev. Biochem. 56, 89-124. 2. Deutscher, M. P. (1984) J. Cell Biol. 99, 373-377. 3. Yang, D. C. H., Garcia, J. V., Johnson, Y. D. & Wahab, S. (1985) Curr. Top. Cell. Regul. 26, 325-335. 4. Mirande, M., LeCorre, D. & Waller, J.-P. (1985) Eur. J. Biochem. 147, 281-289. 5. Cirakaglu, B. & Waller, J.-P. (1985) Biochim. Biophys. Acta 829, 173-179. 6. Godar, D. E., Godar, D. E., Garcia, V., Jacabo, A., Aebi, U. & Yang, D. C. H. (1988) Biochemistry 27, 6921-6928. 7. Norcum, M. T. (1989) J. Biol. Chem. 264, 15043-15051. 8. Bandyopadhyay, A. K. & Deutscher, M. P. (1971) J. Mol. Biol. 60, 113-122. 9. Walker, E. J., Trecy, G. B. & Jeffrey, P. D. (1983) Biochemistry 22, 1934-1941. 10. Dang, C. V. & Yang, D. C. H. (1979) J. Biol. Chem. 254, 5350-5356. 11. Sihag, R. K. & Deutscher, M. P. (1983) J. Biol. Chem. 258, 11846-11850. 12. Cirakoglu, B., Mirande, M. & Waller, J.-P. (1985) FEBS Lett. 183, 185-190. 13. Vellekamp, G. & Deutscher, M. P. (1987) J. Biol. Chem. 262, 9927-9930. 14. Som, K. & Hardesty, B. (1975) Arch. Biochem. Biophys. 166, 507-517. 15. Ritter, P. O., Enger, M. D. & Hampel, A. E. (1979) Biochim. Biophys. Acta 562, 377-385.
Proc. Natl. Acad. Sci. USA 87 (1990)
3669
16. Deutscher, M. P. & Ni, R. C. (1982) J. Biol. Chem. 257, 6003-6006. 17. Vellekamp, G., Sihag, R. K. & Deutscher, M. P. (1985) J. Biol. Chem. 260, 9843-9847. 18. Ferber, S. & Ciechanover, A. (1987) Nature (London) 326, 808-811. 19. Deutscher, M. P. (1972) J. Biol. Chem. 247, 459-468. 20. Sivaram, P., Vellekamp, G. & Deutscher, M. P. (1988) J. Biol. Chem. 263, 18891-188%. 21. Kellerman, O., Viel, C. & Waller, J.-P. (1978) Eur. J. Biochem. 88, 197-204. 22. Mirande, M., Cirakoglu, B. & Waller, J.-P. (1983) Eur. J. Biochem. 131, 163-170. 23. Eriani, G., Dirheimer, G. & Gangloff, J. (1989) Nucleic Acids Res. 17, 5725-5736. 24. Chatton, B., Walter, P., Ebel, J.-P., Lacroute, F. & Fasiolo, F.
(1988) J. Biol. Chem. 263, 52-57.
25. Natsoulis, G., Hilger, F. & Fink, G. R. (1986) Cell 46, 235-243. 26. Conneely, 0. M., Kettelberger, D. M., Tsai, M.-J., Schrader, W. T. & O'Malley, B. W. (1989) J. Biol. Chem. 264, 1406214064. 27. Spence, A. M., Sheppard, P. C., Davie, J. R., Matuo, Y., Nishi, N., McKeehan, W. L., Dodd, J. G. & Matusik, R. J. (1989) Proc. Natl. Acad. Sci. USA 86, 7843-7847. 28. Kaji, H., Novelli, G. D. & Kaji, A. (1963) Biochim. Biophys. Acta 76, 474-477. 29. Deutch, C. E., Scarpulla, R. C. & Soffer, R. L. (1978) Curr. Top. Cell. Regul. 13, 1-28. 30. Ciechanover, A., Ferber, S., Ganoth, D., Elias, S., Hershko, A. & Arfin, S. (1988) J. Biol. Chem. 263, 11155-11167. 31. Gonda, D. K., Bachmair, A., Wunning, I., Tobias, J. W., Lane, W. S. & Varshavsky, A. (1989) J. Biol. Chem. 264, 16700-16712. 32. Rechsteiner, M. (1987) Annu. Rev. Cell Biol. 3, 1-30. 33. Ferber, S. & Ciechanover, A. (1986) J. Biol. Chem. 261, 3128-3134. 34. Irvin, J. D. & Hardesty, B. (1972) Biochemistry 11, 1915-1920. 35. Moline, G., Hampel, A. & Enger, M. D. (1974) Biochem. J. 143, 191-195. 36. Graf, H. (1976) Biochim. Biophys. Acta 425, 175-184. 37. Bec, G., Kerjan, P., Zha, X. D. & Waller, J.-P. (1989) J. Biol. Chem. 264, 21131-21137. 38. Nielson, P., Goelz, S. & Trachsel, H. (1983) CellBiol. Int. Rep. 7, 245-254. 39. Howe, J. G. & Hershey, J. W. B. (1984) Cell 37, 85-93. 40. Bonneau, A.-M., Darveau, A. & Sonenberg, N. (1985) J. Cell Biol. 100, 1209-1218. 41. Orrelles, D. A., Fey, E. G. & Penman, S. (1986) Mol. Cell. Biol. 6, 1650-1662. 42. Pollard, J. W., Galpine, A. R. & Clemens, M. J. (1989) Eur. J. Biochem. 182, 1-9. 43. vanVenrooij, W. J., Moonen, H. & vanLoon-Klaassen, L. (1974) Eur. J. Biochem. 50, 297-304. 44. Airhart, J., Vidrieh, A. & Khairallah, E. A. (1974) Biochem. J. 140, 539-548. 45. Hod, Y. & Hershko, A. (1976) J. Biol. Chem. 251, 4458-4467. 46. Gehrke, L. & Ilan, J. (1983) Proc. Natl. Acad. Sci. USA 80, 3274-3278. 47. Gatica, M., Allende, C. C. & Allende, J. E. (1980) Arch. Biochem. Biophys. 202, 653-656. 48. Hradec, J. & Dusek, Z. (1978) Biochem. J. 172, 1-7.
