450 Energy Charge and Protein Synthesis Vol. 245, No.

Energy Charge and Protein Synthesis phosphate formed during the forward reaction inorganic pyro-
phosphatase is added in a quantity giving maximum stimulation
CONTROL OF AMINOACYL TRANSFER RIBONUCLEIC
ACID SYNTHETASES* of the reaction rate.
Cyclic adenosine 3’,5’-monophosphate has been found to in-
(Received for publication, November 10, 1969) fluence a number of regulatory properties of cells (7, 8). When
MICHAEL BRENNER,~ FRANCESCO DE LORENZO,~ AND tested with the histidyl-tRNA synthetase, cyclic 3’, 5’-AMP
BRUCE N. AMES (Sigma) at a concentration of 409 pM was found to have negli-
From the Department of Biochemistry, University of gible effect in the presence of ATP alone (100 PM), or together
California, Berkeley, California 9.JY20 with ATP (100 PM), AMP (380 PM), and ADP (500 PM).
The activity of the histidyl-tRNA synthetase as a function of
SUMMARY energy charge is shown in Fig. 2. Energy charge is seen to
The histidyl transfer RNA synthetase of Salmonella modulate significantly the eneymic activity, with an inhibition
fyphimurium LT-2 is inhibited by ADP and AMP, and conse- of 32% occurring at an energy charge as high as 0.8. Energy
quently is responsive to energy charge. The arginyl-, valyl-, charge would also presumably regulate the activities of the
and lysyl-tRNA synthetases from Salmonella are also in- arginyl-, valyl-, and lysyl-tRNA synthetases, with the severity
hibited by ADP and AMP. Reduced inhibition of a mutant of the effect probably paralleling the degree of inhibition by
histidyl-tRNA synthetase adds support to the idea that the ADP and AMP as shown in Table I.
greater inhibition of the wild type enzyme is not fortuitous, In conducting these experiments we have taken special care
but is maintained to serve a physiological function. to determine if the regulatory effects are realized under physio-
By affecting the activity of the various aminoacyl-tRNA logical conditions. The amounts of ATP, Mg2+, and histidine
synthetases, energy charge could regulate the size of the used were chosen to represent their estimated intracellular con-
charged tRNA pool and thus control the rate of protein syn- centrations.
thesis. Estimates of the ATP pool of rapidly growing Escherichia
coli range from 2 to 6 mM (9, 10). Under these favorable growth
Recently Atkinson and his collaborators have suggested that conditions, ATP probably accounts for a large portion of the
much of the metabolic machinery of a cell may be regulated by total adenylate pool, so we selected a concentration of 4 mM for
energy charge (energy charge = ([ATP] + )[ADP])/([ATP] + our experiments.
[ADP] + [AMP])) (1). It was proposed and documented (2-4) Choice of a Mg2+ concentration was somewhat more arbitrary.
that AMP and ADP would inhibit energy-utilizing systems, and Values generally ranging from 5 mrvr to 30 mM have been reported
that ATP would inhibit energy-generating systems. (11-14). However, the effective internal concentration is even
Since protein synthesis consumes a significant portion of the more uncertain, since a large portion of the ions may be tightly
energy supply of a cell, it may be subject to energy charge con- bound by other molecules, especially ribosomal RNA (12). We
trol. For every amino acid added to a growing polypeptide, chose a value of 8 mM for our studies.
three high energy bonds are hydrolyzed, two in the esterification Since adenylate kinase was not present in the reaction mixture,
of the amino acid to its corresponding tRNA, during which ATP the amounts of ATP, ADP, and AMP to be added for each level
is split to AMP and the pyrophosphate produced is hydrolyzed; of energy charge were calculated using 0.38 for the equilibrium
and one in the conversion of GTP to GDP as the ribosome is constant of the adenylate kinase reaction. This number was
ratcheted along the messenger RNA. One possible mechanism obtained from the equation of Rose (15), substituting the values
of energy charge control of protein synthesis is regulation of the for pH and Mgz+ concentration used here.
size of the pool of charged tRNA. In this paper w-e examine the The histidine concentration was selected to represent that of
possibility that such control is exerted by energy charge regula- Salmonella growing in minimal salts-glucose medium (16).
tion of aminoacyl-tRNA synthetases. We have found few references to ADP and AMP being tested
Since aminoacyl-tRNA synthetases are ATP-utilizing bio- as inhibitors of aminoacyl-tRNA synthetases. Both AMP and
synthetic enzymes, response to energy charge would be mediated ADP inhibit the glycyl-tRNA synthetase of chick embryo (17).
by inhibition of the enzymes by ADP or AMP or both. Data The glutamyl-tRNA synthetase of E. coli is inhibited by AMP,
showing such inhibition to occur are in Table I. The nature of but not ADP (18). The threonine enzyme from rat liver (19)
the inhibition of the histidine enzyme was probed by determining and the arginine enzyme from E. coli (20) are also inhibited by
the Km for ATP in the presence and absence of the inhibitors. AMP; studies with ADP were not reported. No regulatory
Fig. 1 shows double reciprocal plots of the data obtained. The significance of the inhibitions has previously been inferred.
similarity in Km for ATP in the presence and absence of ADP The finding that the activity of the histidyl-tRNA synthetase
and AMP suggests the inhibition is simple noncompetitive in responds to energy charge under conditions chosen to approxi-
nature. At higher histidine concentrations, however, preliminary mate those existing in the cell does not prove that such control
experiments suggest that ADP is competitive with ATP, while has physiological function. Should further studies of the histi-
AMP remains noncompetitive. dine enzyme reveal a regulatory binding site for ADP or AMP
Reversal of the aminoacylation reaction in the presence of which is distinct from that for ATP, the functioning of such
AMP does not occur in these experiments. To destroy pyro- control in the cell would be incontrovertible. But if ADP and
AMP bind at the ATP binding site, two interpretations would
* This research was supported in part by Grant AM 12092 from the
National Institutes of Health to Bruce N. Ames. be possible. On one hand, the structural requirements for an
$ Predoctoral Fellow studying under United States Public Health ATP-binding site may necessarily permit binding of ADP and
Service Training Grant 5TOlGM31.
$ Present address, Institute of Biochemistry, Medical School, Uni- AMP, with the resulting inhibition having no physiological
versity of Naples, Italy. meaning. On the other hand, as Atkinson has argued (l), the

This is an Open Access article under the CC BY license.

Issue of January 25, 1970 M. Brenner, F. De Lorenxo, and B. N. Ames 451

TABLE I
Inhibition of aminoacyl-tRNA synthetases by ADP and AMP
The tRNA and aminoacyl-tRNA synthetases were isolated from
Salmonella typhimurium LT-2. The tRNA was prepared by the method
of Silbert, Fink, and Ames (5). The histidyl-tRNA synthetase was
purified to homogeneity by fractionation on columns of hydroxylapatite,
DEAE-Sephadex, and phosphocellu1ose.i The synthetases for valine, E
= 60
arginine, and lysine were present in the breakthrough peak of the phos- E
phocellulose column, which was used as the last step in the puri6cation ‘Z
of the histidyl-tRNA synthetase. This breakthrough peak (Fraction p 40
PC) also contained high inorganic pyrophosphatase activity. Assays a?
were conducted at 37O in a final volume of 0.25 ml. The reaction mix- 20
ture contained 200 ELM ATP (Sigma), 0.1 M sodium cacodylate (pH 7.5),
8 mix MgCls, and 12.5 Am units of Salmonella tRNA giving concentra-
tions for tRNAn’* of 0.8 @r, tRNAArg of 1.2 ~C~RI,tRNAVa’ of 2.0 pM, 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 I .o
and tRNALya of 2.7 1~. When added, ADP and AMP (Schwars Bio- Energy Charge
Research) were at concentrations of 500 PM. The three nucleotides had
negligible cross contamination (6). Histidyl-tRNA synthetase was at a FIG. 2. Rate of the reaction catalyzed by histidyl-tRNA synthetase
final concentration of 8.8 ng per ml; Fraction PC was at 20 pg per ml as a function of energy charge. Except for variation in ATP, ADP,
for the assays of valine and arginine, and at 0.13 pg per ml for lysine. and AMP concentrations, the assays were performed as described in the
All synthetases were diluted in a buffer containing 50 rnrvr Tris-HCl of legend to Table I with histidine at a concentration of 10 PM. For the
pH 7.5, 20 mu NaCl, 0.5 m&r EDTA, 2 IIIM mercaptoethanol, and 1 mg lower cm-w (ATP + ADP + AMP) the total adenylate concentration
per ml of bovine serum albumin. was fixed at 4 mu, and the relative concentrations of the three nucleo-
Yeast inorganic pyrophosphatase (Nutritional Biochemicals) was tides to be added were calculated using 0.38 as the equilibrium constant
added at 0.6 unit/O.25 ml for assays of histidine and lysine activation. for the adenylate kinase reaction (see text). For the upper eerie (ATP
In the assays for arginine and valine activation the endogenous inorganic only) the same amount of ATP was added for each point as in the lower
pyrophosphatase activity in Fraction PC obviated the need for added CUTP~, but the ADP and AMP were omitted.
inorganic pyrophosphatase. Tritium labeled n-amino acids were used
for histidine, valine, and lysine (New England Nuclear), and i4C for
arginine (Tracerlab). Samples were previously incubated for 2 min,
and 20 ~1 of diluted enzyme were added to initiate the reactions (buffer
without enzyme added for blanks). At appropriate intervals 75-/.d
samples were withdrawn and deluged with about 4 ml of cold 10% tri- Mutant Synthetase
chloracetic acid. The precipitate was collected on glass fiber filters
(Gelman, type A, 1 inch in diameter), washed three times with cold 10%
trichloracetic acid, three times with 95% ethanol, and finally with ether.
The air-dried filters were counted in a toluene-2,5-diphenyloxaaole
(PPO)-1 ,4-bis[2-(5-phenyloxasolyl)]benzene (POPOP) mixture (Spec-
trafluor, Amersham/Searle) in a Nuclear-Chicago Mark I liquid scin-
tillation counter at an efficiency of 25% for sH and 90% for i4C. The
uninhibited reaction rates in picomoles of amino acid charged per ml per
min were 9.7 for histidine, 51 for arginine, 2.6 for valine, and 49 for II II II I II I
lysine. 0.1 0.2 0.3 0.4 0.5 0.6 07 0.8 0.9 I.0

Energy Charge
Percentage of
inhibition by 500 PY Fro. 3. Inhibition by ADP and AMP of the rate of the reaction
Amino acid ATP concentration catalyzed by mutant and wild type histidyl-tRNA synthetases as a func-
ADP AMP tion of energy charge. For each energy charge the rate of the reaction
-~ containing ATP -I- ADP + AMP is plotted relative to the rate obtained
with ATP alone. Data for the wild type enzyme are taken from Fig. 2.
PM % %
Except for the change in enzyme, the energy charge profile for the mutant
Histidine (10 PM). ............. 200 24 45 was done exactly as for the wild type (legend to Fig. 2). The mutant
Arginine (27 PM). .............. 200 47 58 enzyme was purified lo-fold on a hydroxylapatite co1umn.s It was used
Valine (21 PM). ................ 200 29 56 at a concentration of 34 pg per ml for energy charge points 0.1 to 0.4,
Lysine (28 PM). ............... 200 9 33 and at 5.8 pg per ml for energy charge points 0.5 to 1.0.

- 51 I I I I I I I I I ADP and AMP inhibitions may be maintained by natural selec-

.F
5 tion as part of an efficient control system for the cell.
2
i 4-
We have obtained an indication that the level of inhibition of
2 the wild type enzyme by ADP and AMP is not a necessary
consequence of the structural requirements for an ATP-binding
site. As part of an independent study on mutant histidyl-tRNA
synthetases,2 it was observed that the synthetase of one of these
mutants, strain HisSd280, has the same K, for histidine and
tRNAHis as the wild type enzyme, while its Km for ATP is 5-fold
higher. However, at physiological ATP concentrations (3 mu),
ATP does not limit the reaction rate. The mutant enzyme is
%. .;*0
considerably less inhibited by ADP and AMP than the enzyme
2 4 6 8 IO 12 14 16
of its wild type parent. Under the conditions given in Table I,
SW’ (S in mM ATP)
inhibition of the mutant enzyme by ADP is reduced to IS%,
FIG. 1. Rate of the reaction catalyzed by histidyl-tRNA synthetase while inhibition by AMP is cut to 6%. The attenuated inhibi-
as a function of ATP in the presence and absence of ADP and AMP.
Except for the variation in ATP concentrations, the assays were per- tion is reflected in the energy charge response. Fig. 3 compares
formed as described in the legend to Table I with 500 pM AMP or ADP the wild type and mutant enzymes with respect to the inhibition
when used, and histidine at 10 w.
* F. De Lorenzo, D. S. Straus, and B. N. Ames, manuscript in prep-
1 F. De Loreneo and B. N. Ames, J. Biol. Chm., in press. aration.
452 Energy Charge and Protein Synthesis Vol. 245, No. 2

produced by ADP and AMP at various energy charge levels. 3. SHEN, L. C., FALL, L., WALTON, G. M., AND ATKINSON, D. E.,
Biochemistry, 7, 4041 (1968).
Although the reduction in inhibition appears most dramatic at 4. BIGLER, W. N., AND ATKINSON, D. E., B&hem. Btiphys. Res.
low energy charge, even at an energy charge of 0.8 inhibition of Commun., 36, 381 (1969).
5. SILBERT, D. F., FINK, G. R., AND AMES, B. N., J. Mol. Biol., 22,335
the mutant enzyme is only half that of the wild type. The (1966).
properties of the mutant enzyme are consistent with the proposi- 6. RANDERATH, K., AND RANDERATE, E., J. Chromuiogr., 16,111 (1964).
7. PERLMAN, R. L., AND PASTAN, I., J. Bill. Chem., 243, 5420 (1968).
tion that the response of the wild type enzyme to energy charge 8. Kuo, J. F., AND GREENGARD, P., J. BX. Chem., 244, 3417 (1969).
may be maintained by natural selection. 9. COLE, H. A., WIMPENNY, J. W. T., AND HUGHES, D. E., B&him.
Biophys. Acta, 143, 445 (1967).
In conclusion, we propose that the inhibition by ADP and 10. BAGNARA, A. S., AND FINCH, L. R., Biochem. Biophye. Res. Commun.,
AMP of aminoacyl-tRNA synthetases is of physiological impor- 33, 15 (1968).
11. HURWITZ, C., AND ROSANO, C. L., J. Bbl. Chem., 242, 3719 (1967).
tance in the regulation of protein synthesis. Definitive proof of 12. LUBIN, M., AND ENNIS, H: L., Biochem. Biophys. Acta, 30,614 (1964).
this contention awaits further study of energy charge in general, 13. SILVER, S., Proc. Nat. Acad. Sci. U. S. A., 62, 764 (1969).
14. LUSK, J. E., AND KENNEDY, E. P., J. Biol. Chem., 244, 1653 (1969).
and aminoacyl-tRNA synthetases in particular. 15. ROSE, I. A., Proc. Nat. Acad. Sci. U. S. A., 61, 1079 (1968).
16. AMES, G. F., Arch. Biochem. Btiphys., 104,l (1964).
REFERENCES 17. BUBLITZ, C., B&him. Bbphys. Acta, 113, 158 (1966).
18. LAZZARINI, R. A., AND MERLER, A. H., Btichemistr~, 3, 1445 (1964).
1. ATKINSON, D. E., Biochemistry, 7, 4030 (1968). 19. ALLENDE, C. C., ALLENDE, J. E., GATICA, M., CELIS, J., MORA, G.,
2. KLUNGSOYR, L., HAGEMEN, J. H., FALL, L., AND ATKINSON, D. E., AND MATAIULA, M., J. Bill. Chem., 241, 2245 (1966).
Biochemistry, '7, 4035 (1968). 20. MITRA, S. K., AND MEHLER, A. H., J. Biol. Chem., 242, 5490 (1967).