Relationship between Histidyl-tRNA Level and Protein Synthesis Rate in Wild-type and Mutant Chinese Hamster Ovary Cells DON J. LOFGREN ' AND LARRY H.THOMPSON Lawrence Livermore Laboratory, L-452, University of California, P. 0. Box 5507, Biomedical Sciences Diuision, Livermore, California 94550
ABSTRACT A preliminary investigation was carried out to determine how conditional lethal mutants affected in particular aminoacyl-tRNA synthetases may be used to study the role of tRNA charging levels in protein synthesis. The relationship between rate of protein synthesis and level of histidyl-tRNA in wild-type cultured Chinese hamster ovary cells was determined using the analogue histidinol to inhibit histidyl-tRNA synthetase activity. This response was compared with that obtained using a mutant strain with a defective histidyl-tRNA synthetase that phenotypically shows decreased rates of protein synthesis a t reduced concentrations of histidine in the growth medium. The approach used was based on measuring the histidyl-tRNA levels in live cells. The percentage charging was estimated by comparing [*4Clhistidineincorporated into alkali-labile material in paired samples, one of which was treated with cycloheximide, five minutes before terminating during the incubation, to produce maximal aminoacylation. Wild-type cells under histidinol inhibition exhibited a sensitive, sigmoidal relationship between the level of histidyl-tRNA and the rate of protein synthesis. A decrease in the relative percentage of acylated tRNAHisfrom 46%to 35% elicited a large reduction in the rate of protein synthesis from 90%to 30%relative to untreated cells. An unpredicted result was t h a t the relationship between protein synthesis and histidyl-tRNA in the mutant was essentially linear. High acylation values for tRNAHiawere associated with rates of protein synthesis t h a t were not nearly as high as in wild-type cells. These findings suggest that the charging levels of tRNAHisisoacceptors could play a regulatory role in determining the rate of protein synthesis under conditions of histidine starvation in normal cells. The mutant appears to be a potentially useful system for studying the pivotal role of tRNA charging in protein synthesis, assuming that the altered response in the mutant is caused by its altered synthetase. The molecular mechanisms governing translation in response to amino acid deficiency have yet to be clearly defined for mammalian cells. Amino acid starvation results in inhibiting protein synthesis and disassembling polysomes i n rat liver (Munro, '68; Fishman e t al., '69; Pronczuk et al., '70) and cultured cells (Vaughan et al., '71; Van Venrooij et al., '72; Stanners and Thompson, '74). It has been proposed that the molecular mechanism regulating protein synthesis under amino acid limiting conditions is mediated through t h e J. CELL. PHYSIOL. (1979) 99: 303-312.
amount of aminoacylation of tRNA (Allen et al., '69). Levels of aminoacyl-tRNAin the liver of starved rats were found to be the same as in the controls (Allen et al., '69; Shenoy and Rogers, '771, but rats force fed diets deficient in essential amino acids did show significant reductions in t h e respective aminoacyl-tRNA pools (Allen et al., '69). Vaughan and Hansen Received June 29, '78. Accepted Jan. 17, '79. ' Department of Biomedical and Environmental Sciences, School of Public Health, University of California, Berkeley. 2Correspondence:Dr. L. H. Thompson, L-452, Lawrence Livermore Laboratory, P. 0. Box 5507, Livermore, California 94550.
303
304
DON J. LOFGREN AND LARRY H. THOMPSON
('73) examined the relationships between the rate of protein synthesis and acylated tRNA1Ie and tRNAHisin cultured HeLa cells. They reported that both tRNA species were normally highly acylated and that the appearance of small amounts of deacylated tRNA of either species was accompanied by pronounced inhibition of protein synthesis and reduced initiation. According to the model of Vaughan and Hansen, regulating protein synthesis could occur directly via uncharged tRNA acting a t the initiation step. Therefore, in terms of the inhibition of protein synthesis it should not matter how deacylated tRNA is generated. As a test of the Vaughan and Hansen model, in the current study we used two independent methods to produce uncharged tRNAHl3 in Chinese hamster ovary (CHO) cells: (1)inhibiting the charging of histidine to tRNAH1"by competition with the analogue histidinol, as previously reported in HeLa cells (Vaughan and Hansen, '731, and (2) depressing the charging of histidine in a conditional mutant CHO strain that is defective for histidyltRNA synthetase (Thompson et al., '77, '78) and sensitive to reduced histidine in the growth medium. To our surprise, the response of the cells to the level of uncharged tRNAHiS differed in the two cases. These unpredicted results suggest that the level of acylation in wild-type cells could be a mechanism for regulating protein synthesis i n response to histidine availability and that the mutant may be defective in this regard. MATERIALS AND METHODS
Cell culture conditions Stock CHO cell suspension cultures were maintained a t 34°C as previously described (Thompson et al., '77). The wild-type cell line used here is the strain designated Gat-, while His-1 is a temperature-sensitive mutant with an increased histidine requirement for growth and was selected from Gat- (Thompson et al., '77; Adair et al., '78). The exponential doubling times for His-1 and wild-type were 31 to 35 hours and 19 hours, respectively. Chemicals and radioactive amino acids Cycloheximide, L-histidinol, and all Lamino acids were obtained from Calbiochem. Ribonuclease A and puromycin were from Sigma. L-[2,5-3H1Histidine (1 mCi/ml, 60 Ci/ mmol) and L-[U-'4Clleucine (50 pCi/ml, 330 mCi/mmol) in aqueous solution were purchased from Amersham Corporation.
Preparation of medium and cells for protein synthesis and histidyl-tRNA measurements The experimental medium was prepared immediately prior to the assay. Thawed frozen stock solutions of asparagine, glutamine, histidine, cysteine, and dialyzed fetal bovine serum (K.C. Biological, Inc.) a t 10% (V/V) were added to an otherwise fully supplemented a-MEM medium, as reported elsewhere (Thompson et al., '77). All amino acid concentrations were standard for a-MEM, except that for wild-type cells the medium contained 2 0 p M histidine (to facilitate histidinol inhibition), and 200 p M leucine and valine. The medium for His-1 cells was initially lacking histidine (see below), and leucine and valine were each at 400 p M . Wild-type cells growing exponentially at 34°C (2.5-3.0 X l o 5cells/ml) were collected by centrifugation and rinsed with the experimental medium. The cells were again centrifuged and resuspended in experimental medium a t a concentration of 1 x l o 7 cells per ml. L-Histidinol was added to 1-ml suspension cultures which were preincubated at 34°C for 25 minutes before being given labeled amino acid. Sample preparation for His-1 cells was similar to that for wild-type except that the experimental medium lacked histidine, and various amounts of histidine were added to the tubes. Protein synthesis rate measurements Each 1-ml sample received 7 5 p l of 114Clleucine (3.75pCi) or [3Hlhistidine(75pCi). Fiftymicroliter aliquots of 5 x lo5 cells were removed a t 0, 5 , 10, 15, and 20 minutes and placed into glass tubes, on ice, containing 1.5 ml of 0.1 M KOH/0.05 M KC1 to hydrolyze the aminoacyl-tRNA bond. The samples were precipitated with an equal volume of cold 10% TCA, collected on Whatman GF/C filters, and rinsed twice with 10 ml of 5% TCA two times. The dried filters were counted for radioactivity, with counting efficiencies of 78% for "C and 16%for 3H, in 1.5 ml of toluene based liquid scintillation fluid.
Whole cell histidyl-tRNA assay The labeling and extracting of aminoacyltRNA were similar to procedures previously published and are outlined here with pertinent modifications (Thompson et al., '77; Stanners et al., '78). Cell samples received 75
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE
pCi of 13Hlhistidine (zero time) plus in some cases, cycloheximide or puromycin after ten minutes of incubation. After 15 minutes a t 34"C, incubation was terminated by rapidly pouring the sample into 9 ml of 70% ethanol:20 mM Na acetate, pH 5.0, a t 34°C while continuously mixed on the Vortex. The fixed cells were chilled on ice and centrifuged a t 4°C. They were then resuspended a t room temperature in 2.4 ml of buffer containing 0.05 M Na acetate pH 5.0,0.15 M NaC1,O.Ol M EDTA. After adding two ml of phenol and mixing for one minute, each sample was centrifuged. The aqueous layer was removed and extracted with CHC13:isoamylalcohol (99:1, v/v). After phase separation, one half of the aqueous layer was precipitated with 10%TCA, and the other treated on ice with a n equal volume of 0.2 M NaOH for five minutes before precipitation with TCA. Precipitates were collected on 0.8 p m Millipore filters and rinsed three times with 5%TCA. Radioactivity in the dried filters was counted as described above with an estimated efficiency of 10%.
Calculation of histidyl-tRNA The procedure described above allowed us to measure the amount of tRNAH1" acylated when the rate of protein synthesis was constant (as in fig. 1).Evidence t h a t the radioactive amino acid was acylated to tRNA was furnished by the loss of TCA-precipitable radioactivity upon treatment with mild alkali (Gatica e t al., '66)or ribonuclease (results not shown). The alkali-treated fraction (see above) was used for background subtraction, and the alkali-labile radioactivity represents histidyl-tRNA. To calculate the fraction of tRNAHisthat was acylated we obtained a n estimate of the amount of tRNAHist h a t could be acylated in the presence of 200 Fg/ml of the inhibitor cycloheximide. Cycloheximide blocks the utilization of histidyl-tRNA and should allow residual histidyl-synthetase activity to acylate most, if not all, of the tRNAH1".Thus, this method was used to derive relative 100% acylation values since at 200 Fg/ml protein synthesis was inhibited by >98% within five minutes under normal conditions. The relative percentage of acylation was then calculated by dividing the alkalilabile radioactivity of each experimental sample by the radioactivity from a duplicate that was additionally treated with cycloheximide a t the end of the labeling interval. The validity of this approach requires t h a t the
305
brief cycloheximide or puromycin treatment alone not significantly alter the specific activity of the histidine pool. RESULTS
Protein synthesis rate versus level of acylated tRNAHLs under conditions of histidinol inhibition L-Histidinol was previously found to be a competitive inhibitor of L-histidine in the aminoacylation of tRNA in crude extracts of HeLa cells (Hansen e t al., '72). The analogue was further shown to inhibit protein synthesis in whole cells without affecting histidine transport (Hansen e t al., '72; Vaughan and Hansen, '73). In our study histidinol was likewise used to inhibit protein synthesis by limiting the activity of histidyl-tRNA synthetase in wild-type CHO cells. Both protein synthesis rates and histidyl-tRNA levels a t various concentrations of histidinol in the medium were determined under exactly the same experimental conditions. The histidine concentration in the medium was 20 p M or 1/10 t h a t of normal a-MEM medium. A comparison of ['4Clleucine incorporation a t 200 and 20 g M histidine (without histidinol) showed no difference, demonstrating t h a t 20 /.LMwas not limiting for protein synthesis (see fig. 4 below). Rates of protein synthesis were estimated separately by [14Clleucineor L3Hlhistidine incorporation into alkali-resistant, TCA precipitable material as shown in figure 1. Under the conditions employed a small amount of labeled amino acid is added while the system is a t equilibrium. The specific activity of the intracellular amino acid pool and the comparatively very small aminoacyl-tRNA pool rapidly reaches that of the medium as evidenced by the linear rates after a lag of one minute for leucine and less than one minute for histidine. Rates of protein synthesis obtained from the slopes of figure 1 and two other experiments, and expressed as percentages, are shown in figure 2. There was no difference between the relative rates of [ 14Clleucine and i3Hlhistidine incorporation a t various levels of histidinol inhibition. If ['Tlleucine incorporation can be taken as a valid measure of the rate of protein synthesis, then the specific activity of histidyl-tRNA is independent of histidinol concentration. This condition, however, is not necessarily required for validity of the percentage charging determination.
306
DON J. LOFGREN AND LARRY H. THOMPSON '
I
'
I
i
I
'
I
I
- A
0
0
20
15
I
I
I
I
I
I
e
120
10
I
5
10
15
20
Minutes
Minutes
Fig. 1 Incorporation of IITlleucine or f3Hlhistidine into wild-type cell protein a t various concentrations of histidinol. Samples of 1 X 10' cells in 1 ml of fresh medium were prepared and labeled as detailed in MATERIALS AND METHODS. The concentrations of histidinol were: zero (0); 4 0 p M ( A ) ; 1 5 0 p M ( O ) ; and 800 pM(X). (A) I'"C1leucine incorporation, final specific activity a t 18 Ci/mol; a zero time background value of 250 cpm was subtracted. (B) 13Hlhistidine incorporation, specific activity at 3.5 Ci/mol; background 900 cpm.
a?
.-&
100
v)
s 2 .-c c 0
2 a
r
50
c W
2
.-*>0 al r
0
100
200
400
800
Histidinol concentration, pM Fig. 2 Relative protein synthesis rates as a function of histidinol concentration for wild-type cells. All rates were obtained as exampled in figure 1, with the uninhibited sample in each experiment defining the 100%level. In one experiment {Whistidine (X)was used, and three others I'Clleucine ( 0 , A . O ) .
The histidyl-tRNA assay was used to determine the level of acylation a t histidinol concentrations in the range shown in figure 2. After the cell samples had preincubated for 25 minutes with the analogue, L3Hlhistidinewas added (time zero) and left in for 15 minutes before termination. Protein synthesis was constant for a t least a n additional five minutes (fig. 1) Cyclohemixide was added to duplicate cultures during the last five minutes of incubation. Samples were then processed for histidyl-tRNA as described in
MATERIALS AND METHODS. Table 1 shows the data from one of three experiments. The apparent amount of histidyl-tRNA (cpm without cycloheximide) decreased with increasing concentration of histidinol. Samples treated with cycloheximide during the last five minutes of labeling had similar cpm and were used to calculate the percentage of tRNAHisthat was acylated. Several points can be made from the consistent values among the cycloheximide or puromycin treated samples within each experiment. First, a t the various levels of histi-
307
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE TABLE 1
Percentage acylated tRNAHis for wild-typecells at uarious concentrations ofhistidinol Histidinol
PM
0 5 10 20 40 150 150 300 800
cpm without cycloheximide
cpm with cycloheximide
2,258 1,521 1,139 989 1,150 142
2,395
-
2,333 2,496 2,653
634 221
2,611
% hiatidyl-tRNA
90 61 46 40 46 30 -
25 9
Cells were prepared as detailed in MATERIALS AND METHODS and incubated for 25 minutes a t 34OC with histidinol. Each sample was then labeled with [%]histidine and the incubation terminated after 15 minutes. Duplicate samples were additionally treated with cycloheximide a t 200pgIml during the last five minutes of incubation. Cpm representing histidyl-tRNA were obtained as described in MATERIALS AND METHODS. The backgrounds, represented by the alkali-treated samples, were always less than 10%of the total cpm. The percentage histidyl-tRNA was calculated by dividing the cpm obtained without cycloheximide by the average cpm obtained in the presence of cycloheximide (2,498 % 137 cpm). In two other repeat experiments not shown the means and standard deviations of t h e cycloheximide treated samples were 2,550 ? 195 and 3,115 2 90 cpm. I cpm without cycloheximide divided by the average cpm with cycloheximide. ‘Sample treated with 550 pg/ml of puromycin instead of cycloheximide during the last five minutes of incubation. TABLE 2
Percentage acylated tRNAHis for mutant His-1 cells at various concentrations of histidine Histidine PM
200 200 100 75 50 35 20
’
cpm without cycloheximide
cpm with cycloheximide
308
484 469,446 824 924 1,418 1,684 2,808
-
535 582 666 596 595
% his-tRNA
64 -
65 63 47 35 21
His-1 cells were prepared as detailed in MATERIALS AND METHODS and incubated with various concentrations of histidine for 25 minutes a t 34°C. The histidyl-tRNA was then labeled and extracted a s described in table 1. The alkali-treated sample cpm (all S 16 cpm) were subtracted, and the alkali-labile cpm shown represent [ Whistidine acylated to tRNA. The percentage histidyl-tRNA was calculated by dividing the cpm obtained without cycloheximide by the cpm of the duplicate additionally treated with 200 pglrnl cycloheximide. The values in the “with cycloheximide” column are increasing because of the change in specific activity. Treated with 550 pg/ml puromycin instead of cycloheximide.
’
dinol tested, residual synthetase activity appeared to be sufficient to acylate the tRNAHiE pool to the same high level. Second, there is no indication t h a t the specific activity of the histidine pool was changed by the varying degree of histidinol inhibition. This is consistent with the data presented above where the incorporation r a t e s of L3Hlhistidine and [14Clleucine are compared under these conditions. Thus, a constant specific activity allows us to use the average of the cycloheximide treated samples for the 100%relative acylation. Third, conditions for producing a high
level of acylated tRNAHiswere not peculiar to cycloheximide since samples treated with puromycin gave similar values (tables 1, 2). The relationship between levels of histidyltRNA and relative rates of protein synthesis for various degrees of histidyl-tRNA synthetase inhibition, shown in figure 3, was determined by combining data from table 1 and figure 2 with similar data from other experiments. The resulting curve is unmistakably sigmoidal. At maximal rate of protein synthesis, the level of charging of tRNAHiais approximately 80%i n the presence of 20 p M
308
DON J. LOFGREN AND LARRY H. THOMPSON
Percentage of tRNAH'' acylated Fig. 3 Percentage acylated tRNAHis versus t h e relative rate of protein synthesis for wild-type cells. Three separate histidyl-tRNA experiments (O,A,U) were completed as in table 1, and t h e relative percentage histidyl-tRNA found for a given concentration of histidinol was plotted against the corresponding percentage protein synthesis as shown in figure 2. These experiments were conducted at 20 p M histidine. The percentage histidyl-tRNA a t 200 pM histidine without histidinol is identified by (X).
100
c
50
E
0
100
0 100
20
20c
I
Histidine concentration, p M Fig. 4 Rate of protein synthesis for His-1 and wild-type cells as a function of histidine concentration in the medium. Cell samples of 1 X lo7 cells in 1 ml of fresh medium were prepared as detailed in MATERIALS AND METHODS and incubated for 25 minutes a t 34°C. Samples were then treated with 75 ~1 of ["clleucine (3.75pCi, final specific activity 9.1 Cifmol) and rates of incorporation were taken from the slopes between 5 and 20 minutes. Three separate His-1 experiments ( O , A , O ) were performed and wild-type cells ( 0 ) were measured once in parallel. No significant difference in Lowry protein content (Lowry e t al., '51) per cell was detected between wild-type and His-1 cells for these experimental conditions. Standard concentration of histidine in a-MEM medium is 200pM.
histidine. As the level drops to 50%,protein synthesis decreases by only about 10%. A threshold value is reached a t 45% charging, and protein synthesis then responds dra-
matically to small changes in the level of acylated tRNAHis.At very low levels of acylation, protein synthesis becomes more refractory to inhibition.
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE
309
,/---
I
/'
71
.3
- P
I
" r Y A
10-
o
0
I
7
h I
I
10
I
I
I
I
50
I
I
100
Percentage of tRNAHisacylated Fig. 5 Percentage acylated tRNAHi8 versus the relative rate of wild-type protein synthesis for His-I cells. The data from table 2 ( A ) and two other similar experiments ( 0 , O ) were used to plot the histidyl-tRNA level for a given concentration of histidine as a function of the corresponding percentage of wild-type protein synthesis as derived from the relationship in figure 5; the line was drawn by eye. The 100%histidyl-tRNA values produced with cycloheximide were approximately the same for wild-type and His-1 cells (table 1 and text vs. table 2, 20 p his), indicating no significant differences between the two strains in the total concentration of tRNAHis per cell. The dotted line represents the relationship found for wild-type cells under histidinol inhibition (fig. 3).
Protein synthesis rate uersus acylated tRNAHis leuel in a histidyl-tRNA synthetase mutant An alternative method of altering the histidyl-tRNA pool utilized a mutant with a defective histidyl-tRNA synthetase. It was previously shown that protein synthesis and growth for the mutant His-1 were abnormally sensitive to reduced concentrations of histidine in the medium (Thompson et al., '77); histidyl-tRNA synthetase activity in vitro was reduced compared to the wild-type (Thompson et al., '78). The mutant's sensitivity to reduced histidine at 34°C was used in this study to affect levels of histidyl-tRNA and rates of protein synthesis. Experimental conditions were similar to those of the histidinol inhibition experiments with wild-type cells. Protein synthesis was estimated with [' Clleucine, and incorporation was linear for at least 20 minutes a t all concentrations of histidine tested (data not shown). The rate of protein synthesis for wild-type and mutant cells a t various concentrations of histidine is shown in figure 4.The mutant exhibits histidine dependence over a range t h a t does not affect the wild-type. At histidine concentrations up to 2 mM, the rate did not increase over that
shown for 200 p M . His-1 had 45% the protein synthesis rate of wild-type under the most optimum conditions tested, consistent with the mutant's exponential growth rate in culture being about one half of wild-type's. Representative data for a histidyl-tRNA assay a t various histidine concentrations are presented in table 2. A sample treated with cycloheximide was essential for each condition because of the varying specific activity of histidine in the medium. L3H1histidine incorporation into protein was linear with time at 20 p M and 200 p M histidine, indicating t h a t the label had equilibrated with the histidyltRNA pool (results not shown). The level of acylated tRNAHiSdecreased from 64% to 21% over a range of 200 to 20 p M histidine. In contrast, wild-type cells maintained a value of about 80% over this range (fig. 3). The relationship between histidyl-tRNA and protein synthesis for the mutant was compared with that for t h e wild-type on a per cell basis, as shown in figure 5. The difference between the two cell types is striking. At its maximal rate of protein synthesis (45% relative to wild-type), His-1 shows a n acylation value of approximately 65%.Protein synthe-
310
DON J. LOFGREN AND LARRY H. THOMPSON
sis in His-1 is, therefore, depressed despite the relatively high percentage of acylated tRNAHis.The dependence of protein synthesis on the level of histidyl-tRNA was essentially linear, contrasting with the sigmoidal response of wild-type cells treated with histidinol. By presenting the His-1 data relative to wild-type, we have made the tacit assumption that the reduced rate of protein synthesis of His-1 under the optimal conditions used is the result of the mutation affecting the histidyltRNA synthetase. We cannot exclude the possibility of other mutation(s1 affecting protein synthesis and rate of growth. DISCUSSION
We conducted a preliminary investigation to determine if a tRNA synthetase conditional mutant could be utilized, in a manner analogous to a chemical inhibitor of the s y n thetase, to study the involvement of tRNA aminoacylation in protein synthesis. We speculated, based on the model of Vaughan and Hansen ('73) concerning the inhibitory role of deacylated tRNAs in translational initiation, that i t might be possible to obtain the same empirical relationship between the rate of protein synthesis and percentage of acylation of tRNAHisusing the two independent methods of inhibiting histidyl-tRNA synthetase. This comparison, based on a method for measuring the relative percentage acylation of total tRNAHlSin vivo gave results (fig. 5) that were unexpected in two aspects. First, we found that the protein synthesis vs. charging curve observed for histidinol with wild-type cells was qualitatively different from that reported by Vaughan and Hansen ('73) for HeLa cells. In their study the relationships for both tRNAHisand tRNA1le were strongly curvilinear and resembled the lower half of our curve (fig. 3). They observed a 4fold reduction in protein synthesis with the appearance of a relative 20% deacylated tRNAHiswhereas our data indicate that a decrease in level of acylation to below 50% occurs before there is significant reduction in protein synthesis. This disparity could be a t tributed to a difference between t h e two cell types, the temperature (34' vs. 37'1, or assay procedures. Their method of stopping cellular metabolism a t cold neutral pH prior to nucleic acid extraction was similar to one that we believe disrupts certain aminoacyl-tRNA pools by favoring acylation (Atherly and Suchanek, '71; Thompson et al., '78). The possibility that
the conditions we used to terminate the incubation might slow the charging of tRNAHis faster than its utilization in protein synthesis seems quite unlikely in view of the multiple steps a t which protein synthesis will be affected. The second unpredicted result in our findings was the altered behavior of the mutant strain in that i t exhibited a linear relationship between protein synthetic rate and tRNAHis charging level when the extracellular concentration of histidine was varied. Compared with the response seen with wild-type cells, the mutant also had abnormally low rates of protein synthesis a t charging levels in the range of 45-70%. If these altered characteristics of the mutant cells are all the results of the mutation affecting the histidyl-tRNA synthetase, then the faulty enzyme might be defective specifically in charging one of several tRNAHiSisoacceptors that represents a minor, but essential, fraction of tRNAHis. More speculatively, the mutant's enzyme might be defective for some hypothetical regulatory involvement of the normal enzyme in controlling the rate of translation. It is interesting to note that the mutant cells contain a n altered profile of tRNAHisisoacceptors compared with the wild-type (Doctor I. Andrulis, personal communication). The withholding of essential amino acids from cells in culture has been shown to result i n inhibiting protein synthesis initiation (Vaughan e t al., '71; Van Venrooij et al., '72; Stanners and Thompson, '74). This effect on initiation is believed to be mediated by the aminoacyl-tRNA levels (Vaughan and Hansen, '73). A response t h a t reduces initiation of protein synthesis and sustains a relatively high level of acylation (fig. 3) might decrease the probability of translational errors (Parker et al., '78). A variety of other observations also suggest a pivotal role for tRNA charging. Deacylated tRNA was shown to affect the formation of the eukaryotic 8 0 s initiation complex in vitro (Kyner et al., '73). Pain and Henshaw ('75) observed that lysine deprivation caused decreased binding of Met-tRNAf to native 40s ribosomal subunits, presumably via deacylated tRNALys,and concomitant inhibition of initiation in Ehrlich ascites cells. Warrington et al. ('77) found that histidinol inhibited initiation of protein synthesis in mouse L-cells and suggested that deacylated tRNAHisaffected the initiation complex's ability to proceed
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE
with elongation. It has also been suggested that deacylated tRNA is responsible for negative pleiotypic responses associated with the inhibition of growth (Warrington et al., '77; Grummt and Grummt, '76). The whole-cell aminoacyl-tRNA assay utilized may have some advantages over the periodate procedure predominantly used for the measurement of aminoacyl-tRNA in cell culture (Tockman and Vold, '77; Andrulis and Arfin, '78). The periodate method requires more cells and a subsequent in vitro aminoacylation assay. Though applicable to most tRNAs the procedure excludes certain tRNA acceptors (Allen et al., '69; Shenoy and Rogers, '77) and has been shown t o change t h e chromatographic behavior of some isoacceptor tRNAs (Tockman and Vold, '77). The procedure described here, which entails less manipulation and is completed in about four hours, should prove useful in cell culture for measuring changes in aminoacyl-tRNA levels in general. In conclusion, our results suggest that acylation of tRNAHi" is under rather precise control when histidine is deficient. Under such conditions, protein synthesis appears to be limited by other aspects of aminoacylation besides simple availability of histidyl-tRNA. The further use of temperature-sensitive mutants in addition to amino acid analogues specifically inhibiting aminoacylation should help to reveal how cells respond to conditions of limited availability of amino acid. ACKNOWLEDGMENTS
We wish to express gratitude for the adroit cell culture assistance of Susan Fong, for the helpful comments of Doctor Cliff Stanners during t h e preparation of the manuscript, and for the editorial assistance of Doctor Leila Abrahamson. Work performed under t h e auspices of the U.S. Department of Energy by the Lawrence Livermore Laboratory under Contract W-7405-ENG-48. Reference to a company or product name does not imply approval or recommendation of the product by the University of California or the U.S. Department of Energy to the exclusion of others that may be suitable. LITERATURE CITED Adair, G. M., L. H. Thompson and P. A. Lindl 1978 Six complementation classes of conditionally lethal protein synthesis mutants of CHO cells selected by %-amino acid. Somatic Cell Genet., 4; 27-43. Allen, R. E., P. L. Raines and D. M. Regen 1969 Regulatory
311
significance of transfer RNA charging Levels I. Measurements of charging levels in livers of chow-fed rats, fasting rats, and rats fed balanced or inbalanced mixtures of amino acids. Biochim. Biophys. Acta., 190: 323-336. Andrulis, I. L., and S. M. Arfin 1978 Methods for determining the extent of tRNA aminoacylation in uiuo in cultured mammalian cells. In: Methods in Enzymology. Vol. LIX. Academic Press, pp. 268-271. Atherly, A. G., and M. C. Suchanek 1971 Characterization of mutants of Escherichia coli temperature-sensitive for ribonucleic acid regulation: an unusual phenotype associated with a phenylalanyl transfer ribonucleic acid synthetase mutant. J. Bacteriol., 108: 627-638. Fishman, B., R. J. Wurtman and H. N. Munro 1969 Daily rhythms in hepatic polysome profiles and tyrosine transaminase activity: role of dietary protein. Proc. Nat. Acad. Sci., 64: 677-682. Gatica, M., C. C. Allende, G. Moro, J. E. Allende and J. Medina 1966 Effect of pH on the stability of several aminoacyl-sRNA's, Biochim. Biophys. Acta., 129: 201-203. Grummt, F., and I. Grummt 1976 Studies on the role of uncharged tRNA in pleiotypic response of animal cells. Eur. J. Biochem., 64: 307-312. Hansen, B. S., M. H. Vaughan and L. Wang 1972 Reversible inhibition by histidinol of protein synthesis in human cells a t the activation of histidine. J. Biol. Chem., 247: 3854-3857. Kyner, D., P. Zabos and D. H. Levin 1973 Inhibition of protein chain initiation in eukaryotes by deacylated transfer RNA and its reversibility by spermine. Biochim. Biophys. Acta., 324: 386-396. Lowry, 0.H., N. J. Rosebrough, A. L. Farr a ndR. J. Randall 1951 Protein measurement with the folin phenol reagent. J. Biol. Chem., 193: 265-275. Munro, H. N. 1968 Role of amino acid supply in regulating ribosome function. Fed. Proc., 27: 1231-1237. Pain, V. M., and E. C. Henshaw 1975 Initiation of protein synthesis in Ehrlich ascites tumour cells. Evidence for physiological variation in the association of methionyltRNAf with native 40-5 ribosomal subunits in uiuo. Eur. J. Biochem., 57: 335-342. Parker, J., J. W. Pollard, J. D. Friesen and C. P. Stanners 1978 Stuttering: high level mistranslation in animal and bacterial cells. Proc. Nat. Acad. Sci., 75: 1091-1095. Pronczuk, A. W., Q.R. Rogers and H. N. Munro 1970 Liver plysome patterns of rats fed amino acid imbalanced diets. J. Nutr., 100: 1249-1258. Shenoy, S. T., and Q. R. Rogers 1977 Effect of starvation on the charging levels of transfer ribonucleic acid and total acceptor capacity in ra t liver. Biochim. Biophys. Acta., 476: 218-227. Stanners, C. P., and L. H. Thompson 1974 Studies on a mammalian cell mutant with a temperature-sensitive leucyl-tRNA synthetase. In: Control of Proliferation in Animal Cells, Cold Spring Harbor Conferences on Cell Proliferation. B. Clarkson and R. Baserga, eds. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 191-203. Stanners, C. P., T. M. Wightman and J. L. Harkins 1978 Effect of extreme amino acid starvation on the protein synthetic machinery of CHO cells. J. Cell. Physiol., 95: 125-138. Thompson, L. H., D. J. Lofgren and G . M. Adair 1977 CHO cell muta nts for arginyl-, asparagyl-, glutaminyl-, histidyl-, and methionyl-transfer RNA synthetases: Identification and initial characterization. Cell, 11: 157-168. Thompson, L. H., D. J. Lofgren and G. M. Adair 1978 Evidence for structural gene alterations affecting amino-
312
DON J. LOFGREN AND LARRY H. THOMPSON
acyl-tRNA synthetases in CHO cell mutants and revertants. Somatic Cell Genet., 4: 423-435. Tockman, J., and B. S. Vold 1977 In vivo aminoacylation of transfer ribonucleic acid in Bacillus subtilis and evidence for differential utilization of lysine-isoaccepting transfer ribonucleic acid species. J. Bacteriol., 130: 1091-1097. Van Venrooij, W. J. W., E. C. Henshaw and C. A. Hirsch 1972 Effects of deprival of glucose or individual amino acids on plyribosome distribution and rate of protein synthesis in cultured mammalian cells. Biochim. Biophys. Acta., 259: 127-137.
Vaughan, M. H., and B. S. Hansen 1973 Control of initiation of protein synthesis in human cells. J. Biol. Chem.. 248: 7087-7096. Vaughan, Jr., M. H., P. J. Pawlowski and J. Forchhammer 1971 Regulation of protein synthesis initiation in HeLa cells deprived of single essential amino acids. Proc. Nat. Acad. Sci., 68: 2057-2061. Warrington, R. C., N. Wratten and R. Hechtman 1977 L. Histidinol inhibits specifically and reversibly protein and ribosomal RNA synthesis in mouse L cells. J. Biol. Chem., 252: 5251-5257.
ABSTRACT A preliminary investigation was carried out to determine how conditional lethal mutants affected in particular aminoacyl-tRNA synthetases may be used to study the role of tRNA charging levels in protein synthesis. The relationship between rate of protein synthesis and level of histidyl-tRNA in wild-type cultured Chinese hamster ovary cells was determined using the analogue histidinol to inhibit histidyl-tRNA synthetase activity. This response was compared with that obtained using a mutant strain with a defective histidyl-tRNA synthetase that phenotypically shows decreased rates of protein synthesis a t reduced concentrations of histidine in the growth medium. The approach used was based on measuring the histidyl-tRNA levels in live cells. The percentage charging was estimated by comparing [*4Clhistidineincorporated into alkali-labile material in paired samples, one of which was treated with cycloheximide, five minutes before terminating during the incubation, to produce maximal aminoacylation. Wild-type cells under histidinol inhibition exhibited a sensitive, sigmoidal relationship between the level of histidyl-tRNA and the rate of protein synthesis. A decrease in the relative percentage of acylated tRNAHisfrom 46%to 35% elicited a large reduction in the rate of protein synthesis from 90%to 30%relative to untreated cells. An unpredicted result was t h a t the relationship between protein synthesis and histidyl-tRNA in the mutant was essentially linear. High acylation values for tRNAHiawere associated with rates of protein synthesis t h a t were not nearly as high as in wild-type cells. These findings suggest that the charging levels of tRNAHisisoacceptors could play a regulatory role in determining the rate of protein synthesis under conditions of histidine starvation in normal cells. The mutant appears to be a potentially useful system for studying the pivotal role of tRNA charging in protein synthesis, assuming that the altered response in the mutant is caused by its altered synthetase. The molecular mechanisms governing translation in response to amino acid deficiency have yet to be clearly defined for mammalian cells. Amino acid starvation results in inhibiting protein synthesis and disassembling polysomes i n rat liver (Munro, '68; Fishman e t al., '69; Pronczuk et al., '70) and cultured cells (Vaughan et al., '71; Van Venrooij et al., '72; Stanners and Thompson, '74). It has been proposed that the molecular mechanism regulating protein synthesis under amino acid limiting conditions is mediated through t h e J. CELL. PHYSIOL. (1979) 99: 303-312.
amount of aminoacylation of tRNA (Allen et al., '69). Levels of aminoacyl-tRNAin the liver of starved rats were found to be the same as in the controls (Allen et al., '69; Shenoy and Rogers, '771, but rats force fed diets deficient in essential amino acids did show significant reductions in t h e respective aminoacyl-tRNA pools (Allen et al., '69). Vaughan and Hansen Received June 29, '78. Accepted Jan. 17, '79. ' Department of Biomedical and Environmental Sciences, School of Public Health, University of California, Berkeley. 2Correspondence:Dr. L. H. Thompson, L-452, Lawrence Livermore Laboratory, P. 0. Box 5507, Livermore, California 94550.
303
304
DON J. LOFGREN AND LARRY H. THOMPSON
('73) examined the relationships between the rate of protein synthesis and acylated tRNA1Ie and tRNAHisin cultured HeLa cells. They reported that both tRNA species were normally highly acylated and that the appearance of small amounts of deacylated tRNA of either species was accompanied by pronounced inhibition of protein synthesis and reduced initiation. According to the model of Vaughan and Hansen, regulating protein synthesis could occur directly via uncharged tRNA acting a t the initiation step. Therefore, in terms of the inhibition of protein synthesis it should not matter how deacylated tRNA is generated. As a test of the Vaughan and Hansen model, in the current study we used two independent methods to produce uncharged tRNAHl3 in Chinese hamster ovary (CHO) cells: (1)inhibiting the charging of histidine to tRNAH1"by competition with the analogue histidinol, as previously reported in HeLa cells (Vaughan and Hansen, '731, and (2) depressing the charging of histidine in a conditional mutant CHO strain that is defective for histidyltRNA synthetase (Thompson et al., '77, '78) and sensitive to reduced histidine in the growth medium. To our surprise, the response of the cells to the level of uncharged tRNAHiS differed in the two cases. These unpredicted results suggest that the level of acylation in wild-type cells could be a mechanism for regulating protein synthesis i n response to histidine availability and that the mutant may be defective in this regard. MATERIALS AND METHODS
Cell culture conditions Stock CHO cell suspension cultures were maintained a t 34°C as previously described (Thompson et al., '77). The wild-type cell line used here is the strain designated Gat-, while His-1 is a temperature-sensitive mutant with an increased histidine requirement for growth and was selected from Gat- (Thompson et al., '77; Adair et al., '78). The exponential doubling times for His-1 and wild-type were 31 to 35 hours and 19 hours, respectively. Chemicals and radioactive amino acids Cycloheximide, L-histidinol, and all Lamino acids were obtained from Calbiochem. Ribonuclease A and puromycin were from Sigma. L-[2,5-3H1Histidine (1 mCi/ml, 60 Ci/ mmol) and L-[U-'4Clleucine (50 pCi/ml, 330 mCi/mmol) in aqueous solution were purchased from Amersham Corporation.
Preparation of medium and cells for protein synthesis and histidyl-tRNA measurements The experimental medium was prepared immediately prior to the assay. Thawed frozen stock solutions of asparagine, glutamine, histidine, cysteine, and dialyzed fetal bovine serum (K.C. Biological, Inc.) a t 10% (V/V) were added to an otherwise fully supplemented a-MEM medium, as reported elsewhere (Thompson et al., '77). All amino acid concentrations were standard for a-MEM, except that for wild-type cells the medium contained 2 0 p M histidine (to facilitate histidinol inhibition), and 200 p M leucine and valine. The medium for His-1 cells was initially lacking histidine (see below), and leucine and valine were each at 400 p M . Wild-type cells growing exponentially at 34°C (2.5-3.0 X l o 5cells/ml) were collected by centrifugation and rinsed with the experimental medium. The cells were again centrifuged and resuspended in experimental medium a t a concentration of 1 x l o 7 cells per ml. L-Histidinol was added to 1-ml suspension cultures which were preincubated at 34°C for 25 minutes before being given labeled amino acid. Sample preparation for His-1 cells was similar to that for wild-type except that the experimental medium lacked histidine, and various amounts of histidine were added to the tubes. Protein synthesis rate measurements Each 1-ml sample received 7 5 p l of 114Clleucine (3.75pCi) or [3Hlhistidine(75pCi). Fiftymicroliter aliquots of 5 x lo5 cells were removed a t 0, 5 , 10, 15, and 20 minutes and placed into glass tubes, on ice, containing 1.5 ml of 0.1 M KOH/0.05 M KC1 to hydrolyze the aminoacyl-tRNA bond. The samples were precipitated with an equal volume of cold 10% TCA, collected on Whatman GF/C filters, and rinsed twice with 10 ml of 5% TCA two times. The dried filters were counted for radioactivity, with counting efficiencies of 78% for "C and 16%for 3H, in 1.5 ml of toluene based liquid scintillation fluid.
Whole cell histidyl-tRNA assay The labeling and extracting of aminoacyltRNA were similar to procedures previously published and are outlined here with pertinent modifications (Thompson et al., '77; Stanners et al., '78). Cell samples received 75
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE
pCi of 13Hlhistidine (zero time) plus in some cases, cycloheximide or puromycin after ten minutes of incubation. After 15 minutes a t 34"C, incubation was terminated by rapidly pouring the sample into 9 ml of 70% ethanol:20 mM Na acetate, pH 5.0, a t 34°C while continuously mixed on the Vortex. The fixed cells were chilled on ice and centrifuged a t 4°C. They were then resuspended a t room temperature in 2.4 ml of buffer containing 0.05 M Na acetate pH 5.0,0.15 M NaC1,O.Ol M EDTA. After adding two ml of phenol and mixing for one minute, each sample was centrifuged. The aqueous layer was removed and extracted with CHC13:isoamylalcohol (99:1, v/v). After phase separation, one half of the aqueous layer was precipitated with 10%TCA, and the other treated on ice with a n equal volume of 0.2 M NaOH for five minutes before precipitation with TCA. Precipitates were collected on 0.8 p m Millipore filters and rinsed three times with 5%TCA. Radioactivity in the dried filters was counted as described above with an estimated efficiency of 10%.
Calculation of histidyl-tRNA The procedure described above allowed us to measure the amount of tRNAH1" acylated when the rate of protein synthesis was constant (as in fig. 1).Evidence t h a t the radioactive amino acid was acylated to tRNA was furnished by the loss of TCA-precipitable radioactivity upon treatment with mild alkali (Gatica e t al., '66)or ribonuclease (results not shown). The alkali-treated fraction (see above) was used for background subtraction, and the alkali-labile radioactivity represents histidyl-tRNA. To calculate the fraction of tRNAHisthat was acylated we obtained a n estimate of the amount of tRNAHist h a t could be acylated in the presence of 200 Fg/ml of the inhibitor cycloheximide. Cycloheximide blocks the utilization of histidyl-tRNA and should allow residual histidyl-synthetase activity to acylate most, if not all, of the tRNAH1".Thus, this method was used to derive relative 100% acylation values since at 200 Fg/ml protein synthesis was inhibited by >98% within five minutes under normal conditions. The relative percentage of acylation was then calculated by dividing the alkalilabile radioactivity of each experimental sample by the radioactivity from a duplicate that was additionally treated with cycloheximide a t the end of the labeling interval. The validity of this approach requires t h a t the
305
brief cycloheximide or puromycin treatment alone not significantly alter the specific activity of the histidine pool. RESULTS
Protein synthesis rate versus level of acylated tRNAHLs under conditions of histidinol inhibition L-Histidinol was previously found to be a competitive inhibitor of L-histidine in the aminoacylation of tRNA in crude extracts of HeLa cells (Hansen e t al., '72). The analogue was further shown to inhibit protein synthesis in whole cells without affecting histidine transport (Hansen e t al., '72; Vaughan and Hansen, '73). In our study histidinol was likewise used to inhibit protein synthesis by limiting the activity of histidyl-tRNA synthetase in wild-type CHO cells. Both protein synthesis rates and histidyl-tRNA levels a t various concentrations of histidinol in the medium were determined under exactly the same experimental conditions. The histidine concentration in the medium was 20 p M or 1/10 t h a t of normal a-MEM medium. A comparison of ['4Clleucine incorporation a t 200 and 20 g M histidine (without histidinol) showed no difference, demonstrating t h a t 20 /.LMwas not limiting for protein synthesis (see fig. 4 below). Rates of protein synthesis were estimated separately by [14Clleucineor L3Hlhistidine incorporation into alkali-resistant, TCA precipitable material as shown in figure 1. Under the conditions employed a small amount of labeled amino acid is added while the system is a t equilibrium. The specific activity of the intracellular amino acid pool and the comparatively very small aminoacyl-tRNA pool rapidly reaches that of the medium as evidenced by the linear rates after a lag of one minute for leucine and less than one minute for histidine. Rates of protein synthesis obtained from the slopes of figure 1 and two other experiments, and expressed as percentages, are shown in figure 2. There was no difference between the relative rates of [ 14Clleucine and i3Hlhistidine incorporation a t various levels of histidinol inhibition. If ['Tlleucine incorporation can be taken as a valid measure of the rate of protein synthesis, then the specific activity of histidyl-tRNA is independent of histidinol concentration. This condition, however, is not necessarily required for validity of the percentage charging determination.
306
DON J. LOFGREN AND LARRY H. THOMPSON '
I
'
I
i
I
'
I
I
- A
0
0
20
15
I
I
I
I
I
I
e
120
10
I
5
10
15
20
Minutes
Minutes
Fig. 1 Incorporation of IITlleucine or f3Hlhistidine into wild-type cell protein a t various concentrations of histidinol. Samples of 1 X 10' cells in 1 ml of fresh medium were prepared and labeled as detailed in MATERIALS AND METHODS. The concentrations of histidinol were: zero (0); 4 0 p M ( A ) ; 1 5 0 p M ( O ) ; and 800 pM(X). (A) I'"C1leucine incorporation, final specific activity a t 18 Ci/mol; a zero time background value of 250 cpm was subtracted. (B) 13Hlhistidine incorporation, specific activity at 3.5 Ci/mol; background 900 cpm.
a?
.-&
100
v)
s 2 .-c c 0
2 a
r
50
c W
2
.-*>0 al r
0
100
200
400
800
Histidinol concentration, pM Fig. 2 Relative protein synthesis rates as a function of histidinol concentration for wild-type cells. All rates were obtained as exampled in figure 1, with the uninhibited sample in each experiment defining the 100%level. In one experiment {Whistidine (X)was used, and three others I'Clleucine ( 0 , A . O ) .
The histidyl-tRNA assay was used to determine the level of acylation a t histidinol concentrations in the range shown in figure 2. After the cell samples had preincubated for 25 minutes with the analogue, L3Hlhistidinewas added (time zero) and left in for 15 minutes before termination. Protein synthesis was constant for a t least a n additional five minutes (fig. 1) Cyclohemixide was added to duplicate cultures during the last five minutes of incubation. Samples were then processed for histidyl-tRNA as described in
MATERIALS AND METHODS. Table 1 shows the data from one of three experiments. The apparent amount of histidyl-tRNA (cpm without cycloheximide) decreased with increasing concentration of histidinol. Samples treated with cycloheximide during the last five minutes of labeling had similar cpm and were used to calculate the percentage of tRNAHisthat was acylated. Several points can be made from the consistent values among the cycloheximide or puromycin treated samples within each experiment. First, a t the various levels of histi-
307
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE TABLE 1
Percentage acylated tRNAHis for wild-typecells at uarious concentrations ofhistidinol Histidinol
PM
0 5 10 20 40 150 150 300 800
cpm without cycloheximide
cpm with cycloheximide
2,258 1,521 1,139 989 1,150 142
2,395
-
2,333 2,496 2,653
634 221
2,611
% hiatidyl-tRNA
90 61 46 40 46 30 -
25 9
Cells were prepared as detailed in MATERIALS AND METHODS and incubated for 25 minutes a t 34OC with histidinol. Each sample was then labeled with [%]histidine and the incubation terminated after 15 minutes. Duplicate samples were additionally treated with cycloheximide a t 200pgIml during the last five minutes of incubation. Cpm representing histidyl-tRNA were obtained as described in MATERIALS AND METHODS. The backgrounds, represented by the alkali-treated samples, were always less than 10%of the total cpm. The percentage histidyl-tRNA was calculated by dividing the cpm obtained without cycloheximide by the average cpm obtained in the presence of cycloheximide (2,498 % 137 cpm). In two other repeat experiments not shown the means and standard deviations of t h e cycloheximide treated samples were 2,550 ? 195 and 3,115 2 90 cpm. I cpm without cycloheximide divided by the average cpm with cycloheximide. ‘Sample treated with 550 pg/ml of puromycin instead of cycloheximide during the last five minutes of incubation. TABLE 2
Percentage acylated tRNAHis for mutant His-1 cells at various concentrations of histidine Histidine PM
200 200 100 75 50 35 20
’
cpm without cycloheximide
cpm with cycloheximide
308
484 469,446 824 924 1,418 1,684 2,808
-
535 582 666 596 595
% his-tRNA
64 -
65 63 47 35 21
His-1 cells were prepared as detailed in MATERIALS AND METHODS and incubated with various concentrations of histidine for 25 minutes a t 34°C. The histidyl-tRNA was then labeled and extracted a s described in table 1. The alkali-treated sample cpm (all S 16 cpm) were subtracted, and the alkali-labile cpm shown represent [ Whistidine acylated to tRNA. The percentage histidyl-tRNA was calculated by dividing the cpm obtained without cycloheximide by the cpm of the duplicate additionally treated with 200 pglrnl cycloheximide. The values in the “with cycloheximide” column are increasing because of the change in specific activity. Treated with 550 pg/ml puromycin instead of cycloheximide.
’
dinol tested, residual synthetase activity appeared to be sufficient to acylate the tRNAHiE pool to the same high level. Second, there is no indication t h a t the specific activity of the histidine pool was changed by the varying degree of histidinol inhibition. This is consistent with the data presented above where the incorporation r a t e s of L3Hlhistidine and [14Clleucine are compared under these conditions. Thus, a constant specific activity allows us to use the average of the cycloheximide treated samples for the 100%relative acylation. Third, conditions for producing a high
level of acylated tRNAHiswere not peculiar to cycloheximide since samples treated with puromycin gave similar values (tables 1, 2). The relationship between levels of histidyltRNA and relative rates of protein synthesis for various degrees of histidyl-tRNA synthetase inhibition, shown in figure 3, was determined by combining data from table 1 and figure 2 with similar data from other experiments. The resulting curve is unmistakably sigmoidal. At maximal rate of protein synthesis, the level of charging of tRNAHiais approximately 80%i n the presence of 20 p M
308
DON J. LOFGREN AND LARRY H. THOMPSON
Percentage of tRNAH'' acylated Fig. 3 Percentage acylated tRNAHis versus t h e relative rate of protein synthesis for wild-type cells. Three separate histidyl-tRNA experiments (O,A,U) were completed as in table 1, and t h e relative percentage histidyl-tRNA found for a given concentration of histidinol was plotted against the corresponding percentage protein synthesis as shown in figure 2. These experiments were conducted at 20 p M histidine. The percentage histidyl-tRNA a t 200 pM histidine without histidinol is identified by (X).
100
c
50
E
0
100
0 100
20
20c
I
Histidine concentration, p M Fig. 4 Rate of protein synthesis for His-1 and wild-type cells as a function of histidine concentration in the medium. Cell samples of 1 X lo7 cells in 1 ml of fresh medium were prepared as detailed in MATERIALS AND METHODS and incubated for 25 minutes a t 34°C. Samples were then treated with 75 ~1 of ["clleucine (3.75pCi, final specific activity 9.1 Cifmol) and rates of incorporation were taken from the slopes between 5 and 20 minutes. Three separate His-1 experiments ( O , A , O ) were performed and wild-type cells ( 0 ) were measured once in parallel. No significant difference in Lowry protein content (Lowry e t al., '51) per cell was detected between wild-type and His-1 cells for these experimental conditions. Standard concentration of histidine in a-MEM medium is 200pM.
histidine. As the level drops to 50%,protein synthesis decreases by only about 10%. A threshold value is reached a t 45% charging, and protein synthesis then responds dra-
matically to small changes in the level of acylated tRNAHis.At very low levels of acylation, protein synthesis becomes more refractory to inhibition.
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE
309
,/---
I
/'
71
.3
- P
I
" r Y A
10-
o
0
I
7
h I
I
10
I
I
I
I
50
I
I
100
Percentage of tRNAHisacylated Fig. 5 Percentage acylated tRNAHi8 versus the relative rate of wild-type protein synthesis for His-I cells. The data from table 2 ( A ) and two other similar experiments ( 0 , O ) were used to plot the histidyl-tRNA level for a given concentration of histidine as a function of the corresponding percentage of wild-type protein synthesis as derived from the relationship in figure 5; the line was drawn by eye. The 100%histidyl-tRNA values produced with cycloheximide were approximately the same for wild-type and His-1 cells (table 1 and text vs. table 2, 20 p his), indicating no significant differences between the two strains in the total concentration of tRNAHis per cell. The dotted line represents the relationship found for wild-type cells under histidinol inhibition (fig. 3).
Protein synthesis rate uersus acylated tRNAHis leuel in a histidyl-tRNA synthetase mutant An alternative method of altering the histidyl-tRNA pool utilized a mutant with a defective histidyl-tRNA synthetase. It was previously shown that protein synthesis and growth for the mutant His-1 were abnormally sensitive to reduced concentrations of histidine in the medium (Thompson et al., '77); histidyl-tRNA synthetase activity in vitro was reduced compared to the wild-type (Thompson et al., '78). The mutant's sensitivity to reduced histidine at 34°C was used in this study to affect levels of histidyl-tRNA and rates of protein synthesis. Experimental conditions were similar to those of the histidinol inhibition experiments with wild-type cells. Protein synthesis was estimated with [' Clleucine, and incorporation was linear for at least 20 minutes a t all concentrations of histidine tested (data not shown). The rate of protein synthesis for wild-type and mutant cells a t various concentrations of histidine is shown in figure 4.The mutant exhibits histidine dependence over a range t h a t does not affect the wild-type. At histidine concentrations up to 2 mM, the rate did not increase over that
shown for 200 p M . His-1 had 45% the protein synthesis rate of wild-type under the most optimum conditions tested, consistent with the mutant's exponential growth rate in culture being about one half of wild-type's. Representative data for a histidyl-tRNA assay a t various histidine concentrations are presented in table 2. A sample treated with cycloheximide was essential for each condition because of the varying specific activity of histidine in the medium. L3H1histidine incorporation into protein was linear with time at 20 p M and 200 p M histidine, indicating t h a t the label had equilibrated with the histidyltRNA pool (results not shown). The level of acylated tRNAHiSdecreased from 64% to 21% over a range of 200 to 20 p M histidine. In contrast, wild-type cells maintained a value of about 80% over this range (fig. 3). The relationship between histidyl-tRNA and protein synthesis for the mutant was compared with that for t h e wild-type on a per cell basis, as shown in figure 5. The difference between the two cell types is striking. At its maximal rate of protein synthesis (45% relative to wild-type), His-1 shows a n acylation value of approximately 65%.Protein synthe-
310
DON J. LOFGREN AND LARRY H. THOMPSON
sis in His-1 is, therefore, depressed despite the relatively high percentage of acylated tRNAHis.The dependence of protein synthesis on the level of histidyl-tRNA was essentially linear, contrasting with the sigmoidal response of wild-type cells treated with histidinol. By presenting the His-1 data relative to wild-type, we have made the tacit assumption that the reduced rate of protein synthesis of His-1 under the optimal conditions used is the result of the mutation affecting the histidyltRNA synthetase. We cannot exclude the possibility of other mutation(s1 affecting protein synthesis and rate of growth. DISCUSSION
We conducted a preliminary investigation to determine if a tRNA synthetase conditional mutant could be utilized, in a manner analogous to a chemical inhibitor of the s y n thetase, to study the involvement of tRNA aminoacylation in protein synthesis. We speculated, based on the model of Vaughan and Hansen ('73) concerning the inhibitory role of deacylated tRNAs in translational initiation, that i t might be possible to obtain the same empirical relationship between the rate of protein synthesis and percentage of acylation of tRNAHisusing the two independent methods of inhibiting histidyl-tRNA synthetase. This comparison, based on a method for measuring the relative percentage acylation of total tRNAHlSin vivo gave results (fig. 5) that were unexpected in two aspects. First, we found that the protein synthesis vs. charging curve observed for histidinol with wild-type cells was qualitatively different from that reported by Vaughan and Hansen ('73) for HeLa cells. In their study the relationships for both tRNAHisand tRNA1le were strongly curvilinear and resembled the lower half of our curve (fig. 3). They observed a 4fold reduction in protein synthesis with the appearance of a relative 20% deacylated tRNAHiswhereas our data indicate that a decrease in level of acylation to below 50% occurs before there is significant reduction in protein synthesis. This disparity could be a t tributed to a difference between t h e two cell types, the temperature (34' vs. 37'1, or assay procedures. Their method of stopping cellular metabolism a t cold neutral pH prior to nucleic acid extraction was similar to one that we believe disrupts certain aminoacyl-tRNA pools by favoring acylation (Atherly and Suchanek, '71; Thompson et al., '78). The possibility that
the conditions we used to terminate the incubation might slow the charging of tRNAHis faster than its utilization in protein synthesis seems quite unlikely in view of the multiple steps a t which protein synthesis will be affected. The second unpredicted result in our findings was the altered behavior of the mutant strain in that i t exhibited a linear relationship between protein synthetic rate and tRNAHis charging level when the extracellular concentration of histidine was varied. Compared with the response seen with wild-type cells, the mutant also had abnormally low rates of protein synthesis a t charging levels in the range of 45-70%. If these altered characteristics of the mutant cells are all the results of the mutation affecting the histidyl-tRNA synthetase, then the faulty enzyme might be defective specifically in charging one of several tRNAHiSisoacceptors that represents a minor, but essential, fraction of tRNAHis. More speculatively, the mutant's enzyme might be defective for some hypothetical regulatory involvement of the normal enzyme in controlling the rate of translation. It is interesting to note that the mutant cells contain a n altered profile of tRNAHisisoacceptors compared with the wild-type (Doctor I. Andrulis, personal communication). The withholding of essential amino acids from cells in culture has been shown to result i n inhibiting protein synthesis initiation (Vaughan e t al., '71; Van Venrooij et al., '72; Stanners and Thompson, '74). This effect on initiation is believed to be mediated by the aminoacyl-tRNA levels (Vaughan and Hansen, '73). A response t h a t reduces initiation of protein synthesis and sustains a relatively high level of acylation (fig. 3) might decrease the probability of translational errors (Parker et al., '78). A variety of other observations also suggest a pivotal role for tRNA charging. Deacylated tRNA was shown to affect the formation of the eukaryotic 8 0 s initiation complex in vitro (Kyner et al., '73). Pain and Henshaw ('75) observed that lysine deprivation caused decreased binding of Met-tRNAf to native 40s ribosomal subunits, presumably via deacylated tRNALys,and concomitant inhibition of initiation in Ehrlich ascites cells. Warrington et al. ('77) found that histidinol inhibited initiation of protein synthesis in mouse L-cells and suggested that deacylated tRNAHisaffected the initiation complex's ability to proceed
HISTIDYL-tRNA LEVEL AND PROTEIN SYNTHESIS RATE
with elongation. It has also been suggested that deacylated tRNA is responsible for negative pleiotypic responses associated with the inhibition of growth (Warrington et al., '77; Grummt and Grummt, '76). The whole-cell aminoacyl-tRNA assay utilized may have some advantages over the periodate procedure predominantly used for the measurement of aminoacyl-tRNA in cell culture (Tockman and Vold, '77; Andrulis and Arfin, '78). The periodate method requires more cells and a subsequent in vitro aminoacylation assay. Though applicable to most tRNAs the procedure excludes certain tRNA acceptors (Allen et al., '69; Shenoy and Rogers, '77) and has been shown t o change t h e chromatographic behavior of some isoacceptor tRNAs (Tockman and Vold, '77). The procedure described here, which entails less manipulation and is completed in about four hours, should prove useful in cell culture for measuring changes in aminoacyl-tRNA levels in general. In conclusion, our results suggest that acylation of tRNAHi" is under rather precise control when histidine is deficient. Under such conditions, protein synthesis appears to be limited by other aspects of aminoacylation besides simple availability of histidyl-tRNA. The further use of temperature-sensitive mutants in addition to amino acid analogues specifically inhibiting aminoacylation should help to reveal how cells respond to conditions of limited availability of amino acid. ACKNOWLEDGMENTS
We wish to express gratitude for the adroit cell culture assistance of Susan Fong, for the helpful comments of Doctor Cliff Stanners during t h e preparation of the manuscript, and for the editorial assistance of Doctor Leila Abrahamson. Work performed under t h e auspices of the U.S. Department of Energy by the Lawrence Livermore Laboratory under Contract W-7405-ENG-48. Reference to a company or product name does not imply approval or recommendation of the product by the University of California or the U.S. Department of Energy to the exclusion of others that may be suitable. LITERATURE CITED Adair, G. M., L. H. Thompson and P. A. Lindl 1978 Six complementation classes of conditionally lethal protein synthesis mutants of CHO cells selected by %-amino acid. Somatic Cell Genet., 4; 27-43. Allen, R. E., P. L. Raines and D. M. Regen 1969 Regulatory
311
significance of transfer RNA charging Levels I. Measurements of charging levels in livers of chow-fed rats, fasting rats, and rats fed balanced or inbalanced mixtures of amino acids. Biochim. Biophys. Acta., 190: 323-336. Andrulis, I. L., and S. M. Arfin 1978 Methods for determining the extent of tRNA aminoacylation in uiuo in cultured mammalian cells. In: Methods in Enzymology. Vol. LIX. Academic Press, pp. 268-271. Atherly, A. G., and M. C. Suchanek 1971 Characterization of mutants of Escherichia coli temperature-sensitive for ribonucleic acid regulation: an unusual phenotype associated with a phenylalanyl transfer ribonucleic acid synthetase mutant. J. Bacteriol., 108: 627-638. Fishman, B., R. J. Wurtman and H. N. Munro 1969 Daily rhythms in hepatic polysome profiles and tyrosine transaminase activity: role of dietary protein. Proc. Nat. Acad. Sci., 64: 677-682. Gatica, M., C. C. Allende, G. Moro, J. E. Allende and J. Medina 1966 Effect of pH on the stability of several aminoacyl-sRNA's, Biochim. Biophys. Acta., 129: 201-203. Grummt, F., and I. Grummt 1976 Studies on the role of uncharged tRNA in pleiotypic response of animal cells. Eur. J. Biochem., 64: 307-312. Hansen, B. S., M. H. Vaughan and L. Wang 1972 Reversible inhibition by histidinol of protein synthesis in human cells a t the activation of histidine. J. Biol. Chem., 247: 3854-3857. Kyner, D., P. Zabos and D. H. Levin 1973 Inhibition of protein chain initiation in eukaryotes by deacylated transfer RNA and its reversibility by spermine. Biochim. Biophys. Acta., 324: 386-396. Lowry, 0.H., N. J. Rosebrough, A. L. Farr a ndR. J. Randall 1951 Protein measurement with the folin phenol reagent. J. Biol. Chem., 193: 265-275. Munro, H. N. 1968 Role of amino acid supply in regulating ribosome function. Fed. Proc., 27: 1231-1237. Pain, V. M., and E. C. Henshaw 1975 Initiation of protein synthesis in Ehrlich ascites tumour cells. Evidence for physiological variation in the association of methionyltRNAf with native 40-5 ribosomal subunits in uiuo. Eur. J. Biochem., 57: 335-342. Parker, J., J. W. Pollard, J. D. Friesen and C. P. Stanners 1978 Stuttering: high level mistranslation in animal and bacterial cells. Proc. Nat. Acad. Sci., 75: 1091-1095. Pronczuk, A. W., Q.R. Rogers and H. N. Munro 1970 Liver plysome patterns of rats fed amino acid imbalanced diets. J. Nutr., 100: 1249-1258. Shenoy, S. T., and Q. R. Rogers 1977 Effect of starvation on the charging levels of transfer ribonucleic acid and total acceptor capacity in ra t liver. Biochim. Biophys. Acta., 476: 218-227. Stanners, C. P., and L. H. Thompson 1974 Studies on a mammalian cell mutant with a temperature-sensitive leucyl-tRNA synthetase. In: Control of Proliferation in Animal Cells, Cold Spring Harbor Conferences on Cell Proliferation. B. Clarkson and R. Baserga, eds. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 191-203. Stanners, C. P., T. M. Wightman and J. L. Harkins 1978 Effect of extreme amino acid starvation on the protein synthetic machinery of CHO cells. J. Cell. Physiol., 95: 125-138. Thompson, L. H., D. J. Lofgren and G . M. Adair 1977 CHO cell muta nts for arginyl-, asparagyl-, glutaminyl-, histidyl-, and methionyl-transfer RNA synthetases: Identification and initial characterization. Cell, 11: 157-168. Thompson, L. H., D. J. Lofgren and G. M. Adair 1978 Evidence for structural gene alterations affecting amino-
312
DON J. LOFGREN AND LARRY H. THOMPSON
acyl-tRNA synthetases in CHO cell mutants and revertants. Somatic Cell Genet., 4: 423-435. Tockman, J., and B. S. Vold 1977 In vivo aminoacylation of transfer ribonucleic acid in Bacillus subtilis and evidence for differential utilization of lysine-isoaccepting transfer ribonucleic acid species. J. Bacteriol., 130: 1091-1097. Van Venrooij, W. J. W., E. C. Henshaw and C. A. Hirsch 1972 Effects of deprival of glucose or individual amino acids on plyribosome distribution and rate of protein synthesis in cultured mammalian cells. Biochim. Biophys. Acta., 259: 127-137.
Vaughan, M. H., and B. S. Hansen 1973 Control of initiation of protein synthesis in human cells. J. Biol. Chem.. 248: 7087-7096. Vaughan, Jr., M. H., P. J. Pawlowski and J. Forchhammer 1971 Regulation of protein synthesis initiation in HeLa cells deprived of single essential amino acids. Proc. Nat. Acad. Sci., 68: 2057-2061. Warrington, R. C., N. Wratten and R. Hechtman 1977 L. Histidinol inhibits specifically and reversibly protein and ribosomal RNA synthesis in mouse L cells. J. Biol. Chem., 252: 5251-5257.
