an investigation of mistranslation in vivo induced by streptomycin by an examination of the...

Eur. J . Biochetn. 74, 285-292 (1977)

An Investigation of Mistranslation in vivo Induced by Streptomycin by an Examination of the Susceptibility of Abnormal Proteins to Degradation

Jane HEWITT and Margot KOGUT

Biochemistry Department, King’s College, University of London

(Received October 19, 1976)

Proteolysis rates in vivo were measured in Escherichia coli cultures during treatment with dihydrostreptomycin and under various other conditions. Dihydrostreptomycin treatment caused an increase in the proteolysis rate, compared to untreated controls. The proteolytic system in vivo responsible for the elevated proteolysis in the early stages of dihydrostreptomycin treatment, or that during canavanine and puromycin treatment, were not inhibited by addition of phenylniethanesulphonyl fluoride. This agent did inhibit proteolysis rates in cultures whose growth was inhibited by starvation, or had been completely stopped by dihydrostreptomycin. It seems, therefore, that the extremely high proteolysis rates in cultures at this stage of dihydrostreptomycin treatment were due to the action of two protease systems: the one concerned with the breakdown of abnormal proteins, and the other concerned with normal protein turnover and active during a non-specific decline of growth.

The proteolytic rate at complete growth inhibition brought about by dihydrostreptomycin was intermediate between those induced by canavanine and puromycin at the same stage of treatment. This indicated a similar hierarchy in the extent and nature of abnormality in the proteins synthesised under these conditions. The relationship between the abnormality of proteins induced by dihydrostreptomycin and the importance of this in the antibiotic mechanism is discussed.

Streptomycin was demonstrated to cause mistranslation during protein synthesis in vitro with synthetic [ l ] and natural messengers [2-41. Since in vitro, the misreading effect of streptomycin is effective at the level of codon-anticodon interaction [ 5 ] and is limited to certain bases [6], one can envisage that streptomycin-induced misreading could occur to cause both missense and nonsense. The misreading effect was used to account for the phenotypic suppression in vivo of both nonsense and missense mutations in streptomycin-resistant and streptomycin- sensitive Escherichia coli strains treated with streptomycin 17-12]. Misreading has been proposed as effective in the growth inhibitory effect of streptomycin, but this hypothesis has been contested [S]. On the other hand, indications of misreading effects of streptomycin being related to the growth inhibitory effects are supported by the demonstration of growth of mutants by phenotypic suppression at low streptomycin concentrations, whereas at higher streptomycin concentrations there is non-growth of such mutants [ l l - 121. Further, a correlation has been found between the extent of misreading in vitro caused by streptomycin, paramomycin or ethanol [ 1,2,13]

and the response of strains sensitive, resistant or dependent on streptomycin to these agents in vivo [ 141. This is particularly pertinent for streptomycin- dependent strains which are resistant to growth inhibition in the presence of streptomycin or paramomycin, but sensitive to their combined presence. Also, in a protein-synthesising system in vitro, the poly- peptides produced were smaller in streptomycin- treated preparations than untreated ones [15].

Other hypotheses based on observations in vitro propose streptomycin interference at some stage soon after initiation [16], termination [17], or distortion at the P site [18,19]. However their importance in the situation in vivo is in doubt, since predictions of their respective effects on the ribosome profile are at variance with those observed in streptomycin-treated bacterial cultures [20 - 241. Furthermore, because the effects observed on the ribosome profile were late with respect to growth inhibition, it was deduced that streptomycin-affected ribosomes could initiate, under- go some translational movement and were then released from mRNA and could repeat the cycle [24]. Further, as chloramphenicol prevented loss of viability during streptomycin’s action [25] it is a likely con-

286 The Susceptibility of Streptomycin-Induced Proteins to Proteolysis

clusion that the antibiotic was only effective against ribosomes actively involved in protein synthesis.

Therefore, in order to further investigate the relationship between streptomycin-induced misreading and growth inhibition, an indirect assay was chosen, based on the evidence that abnormal proteins are subject to greater rates of proteolysis [26]. The increase in rate of proteolysis depends mainly on the magnitude of the abnormality with perhaps con- formation playing a part [26]. It is slight with single missense or nonsense mutations [27,28], e.g. the suppressor activity of a tRNA 1291, but larger with those resulting from puromycin treatment, which brings about premature peptide release [26,29], or with amino acid analogues [26]. As proteolysis is a first-order reaction, the extent of protein breakdown will depend on both the concentration of abnormal proteins as well as the extent of the abnormality. Examination of the relationship between streptomycin-induced growth inhibition and effects on proteolysis rates in vivo, as well as the time sequence of events, permits evaluation of the misreading effect as the antibiotic's mechanism of action. (Others [26,30, 311 have briefly investigated the effect of streptomycin on proteolysis rates.)

Since dihydrostreptomycin can be more readily obtained in a pure soluble form than streptomycin, and the antibiotic action of the two components is essentially the same [2], dihydrostreptomycin was used in these studies.

MATERIALS AND METHODS

Ma terials

['4C]Leucine 342 (Cijmol) was obtained from the Radiochemical Centre (Amersham), dihydrostreptomycin sulphate (potency 787 iu/mg) from Glaxo Labs, puromycin dihydrochloride from Nutritional Bio- chemical Corporation and canavanine sulphate from the Sigma Chemical Company.

Bacterial Strains and Growth Conditions Employed

E. coli B 3 63 (streptomycin-sensitive, requires leucine, histidine and methionine) was obtained from Dr M. Cannon (King's College, London). This was grown in basal salts medium with 0.4 % (w/v) glucose and NH4CI at 32 "C [32,33] and supplemented by 50 pgjml of the required amino acids. When cultures were in the logarithmic phase of growth, they were treated with dihydrostreptomycin (30 pg/ml) or puromycin (1 00 pg/ml).

E. coli K 12 AB 11 57 (streptomycin-resistant, requires thiamin, leucine, proline, histidine, arginine and threonine) was obtained from Dr R. Buxton

(National Institute for Medical Research) and was grown on a similar medium with the required supple- ments [34]. For treatment, cells were washed and subsequently grown in an arginine-free medium which contained an analogue of arginine, canavanine

Cultures were depleted of glucose by both washing and re-suspending the cells in glucose-free medium.

Growth of cultures was measured as absorbance at 500 nm and plotted as log to the base 2, and the extent of growth inhibition at particular intervals during drug treatment was measured after removal of extracellular antibiotic as previously described [32,35].

(100 pg/ml).

Determination of Protein Degradation in vivo

Volumes, estimated to contain 11 mg dry weight of cells [32] were removed from cultures growing logarithmically or at various stages during treatments, and the cells recovered by filtration. After washing, the cells were re-suspended in 50 ml of the same medium but with [14C]leucine at a lower concentration (3 pg/ml) (0.3 Ci/mol). After 5-min incubation, the cells were washed in antibiotic-free growth medium supplemented with excess unlabelled leucine (75 pg/ml), and re-suspended in 40-50 ml of this same medium. Duplicate samples were taken im- mediately and at intervals during subsequent incubation.

Perchloric acid (5 2,) and bovine serum albumin as co-precipitant were added to these samples, and after standing at room temperature for 30 min, the precipitates were collected by filtration. 1 .O-ml aliquots of the filtrates were added to 10 ml scintillant (750 ml toluene, 250 ml Triton X-100,2 g PPO, 0.05 g POPOP per 1) and the radioactivity counted in a Packard Tri- Carb model 3003. In addition, 1-ml samples were taken from each culture incubation and the total radioactivity was counted. By difference between the radioactivity in this sample and the radioactivity soluble in perchloric acid at zero time, an estimate of the radioactivity in protein at the beginning of the assay was obtained. The rate of breakdown was expressed as the increase of acid-soluble radioactivity as a percentage of that originally in protein.

In some experiments, phenylmethanesulphonyl fluoride (1 mg/ml in 2.5 vjv ethanol unless otherwise stated) was added to the medium after the 5-min [1-14C]leucine pulse, and ethanol added to control cultures.

RESULTS

The Relationship between Proteolysis Rates and Dihydrostreptomycin Treatment

At various stages throughout the treatment of a growing culture with dihydrostreptomycin, the pro-

J . Hewitt and M. Kogut

c

287

9.5

9.0

7.5

I I

0 40 80 120 160 2 0 0 Time after dihydrostreptornycin addition (min) 24r 2 0

E

E 8 c

I

0 0 4 8 12 16

Tlrne (minl

Fig. 1. The effect of dihydrostreptornycin treatment on protein degradation. (A) Growth curves: (0-0) control cells; (A-A) dihydrostreptomycin-treated (30 bg/ml) cells. Samples (F 1 - 5) were removed and pulse-labelled with [l-’4C]leucine. (B) Protein degradation: (A-A) F I ; (A-A) F2; (.A) F3; (0-0) F 4 ; (Y -v) F5. The extent of growth inhibition was respectively 0 ”/, (F l), 37 % (F2), 86% (F3), 100 % (F4 and F5)

teolysis rates of the most recently synthesised proteins were measured as illustrated in Fig. 1 A. The shape of the curves (Fig.lB), where the increment in the amount of acid-soluble material released becomes less as time passes, may possibly indicate a first-order reaction dependent on the concentration of prelabelled proteins susceptible to proteolysis. The proteolysis rates increased with increased time of dihydrostreptomycin treatment, to reach maximum rates when growth was completely inhibited (Fig. 1 B). The combined data of several experiments (Fig. 2) show an exponential relationship between proteolysis rates and the growth inhibitory effect exerted by dihydrostreptomycin, at least up until complete growth in-

I L” 5 5 1.2 - V

V 0 8 1

c . / c

0 4 1

Growth inhibition (%) - P, 0 4 -0.4 -

Fig. 2. Proteolysis rates as a function of growth inhibition by dihydrostreptomycin

hibition, at which stage there could be an additional non-specific stimulation of proteolysis [26]. The relationship between loglo of proteolysis rates and percentage growth inhibition was expected to be linear. This is because although the quantity of dihydrostreptomycin bound to 30-S ribosomes increases linearly as growth inhibition increases [15], such affected ribosomes function catalytically so that the production of abnormal proteins as substrates for proteolysis can be expected to be exponential. (Of course the plotted percentage growth inhibition is computed from the residual growth rates which are themselves exponential.) It is important to notice that the onset and progressive increase of both proteolysis rates and growth inhibition were found to occur with the same time sequence. Other workers [30,31] have similarly shown, although in less detail, that proteolysis rates progressively increase with dihydrostreptomycin or streptomycin treatment.

Before these observations could be further inter- preted, it was necessary to establish whether the increased proteolysis was in response to the nature of the proteins synthesised or due to a general stimulation of proteolysis by dihydrostreptomycin. Fig. 3 shows that proteins prelabelled in the absence of dihydrostreptomycin were degraded at approximately the same low rate after exposure of the culture to dihydrostreptomycin as the untreated controls. There- fore, the proteins synthesised during dihydrostreptomycin treatment must be responsible for the elevated proteolysis rates, and hence they are deduced to be abnormal.

It is relevant here to emphasise that this technique is an indirect assay of misreading. Hence, the direct relationship between an exponential increase in proteolysis rates and the growth inhibitory effect by dihydrostreptomycin is difficult to interpret in terms of quantitative changes in the extent or nature of misreading. Further, this technique does not distin-

288

- c 6 - 0 +.. m

The Susceptibility of Streptomycin-Induced Proteins to Proteolysis

A

A

A

" 0 U

0 v 0 20 40 60 80 100 120 140

Time [min)

Fig, 3. The ejfect qf dihydrostreptoncin treatment on the proteolysis qf'proteins synthesized prior to lreatment. (0-0) Control; (A-A) dihydrostreptomycin-treated (30 pg/ml)

guish between any misreading as missense and/or nonsense produced by dihydrostreptomycin treatment. It is, however, useful for a qualitative comparison of the proteolysis rates consequent to dihydrostreptomycin treatment with model systems for the faulty incorporation of an amino acid and premature termination. For these, treatment of E. coli cultures with canavanine, an analogue of arginine, or with puromycin were used respectively.

The Relationship between Proteolysis Rates and Puromycin and Canavanine Treatment

Proteins were prelabelled when growth was nearly completely inhibited by either dihydrostreptomycin, canavanine or puromycin, transferred subsequently to media free of these agents and the degradation of the labelled proteins determined and compared as shown in Fig. 4. Under these conditions, the cultures treated with puromycin gradually recovered to a normal growth rate (data not shown) and the culture treated with canavanine continued exponential growth but at 10 "/, of the rate of untreated controls (data not shown). There was no recovery of growth for dihydrostreptomycin-treated cultures. In the puromycin-treated cultures, the rate of proteolysis in the first 30 min was 7 times greater than in untreated controls. For the canavanine-treated cultures, the increase was 2-3-fold. This result was considered to be low, not only because reported values were 6 - 8-fold those of controls [26,29,34], but because these were reported for conditions where canavanine had little effect on growth. The proteolysis rate in the dihydrostreptomycin-treated culture was intermediate in effect between that observed here for puromycin and canavanine; puromycin causing the greatest stimulation

2 20 0-

4

0 0 30 60 90 120 150 180

Time (min)

Fig. 4. The eSfects of puromycin, cunavanine or dil~ydrostreptomycin treatment on protein degradation in cultures whose grotcth is corn- pletely inhibited. Untreated controls: (&4) E. coli B163, (O--O) E. coli K12 AB1157; (A--A) canavanine (100 pg/ml); (m-m) puromycin (100 pg/ml) ; (A--A) dihydrostreptomycin (30 lg/ml)

I0

- 8

c

2 6 D m

m a' - 4

0 0 30 60 90 120 150 180

Time (win) Fig. 5. The eJJect of dgJerent concentrations of plienylnic~thylsulpho- nyl .fluoride on proteolysis induced in carbon-starved cultures. (0-0) Logarithmically growing control; culture starved of glucose (A-A), glucose-starved culture treated with (It--.) 0.5 mg/rnl or (A-A) 1 mg,'ml phenylmethylsulphony1 fluoride

of proteolysis. Assuming that this hierarchy reflects the extent of production of abnormal proteins and the relative susceptibility of such proteins to proteolysis, then it may be concluded that the extent of abnormality induced by streptomycin was intermediate between those caused by puromycin and canavanine.

Proteases

The proteolysis rates for those cultures whose growth had been compIeteIy inhibited by dihydrostreptomycin for some time were extremely high (Fig. 1, 2). It was questioned whether this was entirely due to dihydrostreptomycin or if the effects of the drug were supplemented by non-specific stimulation of proteolysis. Goldberg [29] found that proteases activated by starvation of nitrogen or carbon in E. coli

J . Hewitt and M. Kogut 289

Table 1 The effect of yhenylmethylsulphonylSluoride on dihydrostreptornycin-induced proteolysis

Extent of culture's growth inhibition Addition of phenylmethylsulphonyl fluoride

Piotein degration at

5 15 30 60 120 180 rnin _________________ ______ _ _

50 x Just reached 100

-

+

35 min after reaching 300 % -

+

4.0 8.3 10.5 18.0 17.8 21.5 2.8 7.5 10.0 12.3 16.8 18.8

12.5 24.3 26.0 26.8 32.0 34.0 5.0 12.5 14.3 21.5 24.0 28.8

Table 2. The ejj"ct of phenylmethylsulphonyl fluoride on canavanine-induced or purom~cin-inducedproteolysis in cultures whose growth is 90- 100 % inhibited

Culture treatment Addition of phenylmethylsulphonyl fluoride

Protein degradation at

5 15 30 _______---~

60 130 180 min

%

Canavanine -

+ 0.8 1.7 2.8 5.2 7.0 9.0 0.8 1.5 2.6 5.0 10.0 12.5

Puromycin - 0.5 3.5 14.0 20.3 24.3 26.5 + 0.3 7.5 14.3 19.3 32.0 36.0

cultures were inhibited by sulphonyl fluorides such as phenylmethylsulphonyl fluoride. However, such com- pounds did not inhibit the proteolysis of proteins containing amino acid analogues [37], i. e. abnormal proteins. Therefore, the sensitivity of proteolysis to phenylmethylsulphonyl fluoride was examined in dihydrostreptomycin-treated cultures.

First, it was confirmed (Fig. 5 ) that the stimulation of proteolysis induced by carbon starvation was a 2 - 3-fold increase over that of logarithmically growing cultures [26,38 - 401. The same extent of stimulation occurred whether the proteins were labelled under starvation or under normal growth conditions. This is consistent with the growth condition acting as stimulus for proteolysis rather than the nature of the proteins synthesised. The effect of phenylmethylsulphonyl fluoride in lowering such proteolysis rates to those found in logarithmically growing cultures depended on the concentration of the inhibitor [29] (Fig. 5), 1 mg/ml being completely effective. Under these conditions of phenylmethylsulphonyl fluoride treatment, it was found (Table 1) that the proteolysis rates of dihydrostreptomycin-treated cultures were unaffected when growth was 50% inhibited or had just reached 100 %inhibition. However, phenylmethylsulphonyl fluoride caused a partial lowering of the proteolysis rates in dihydrostreptomycin-treated cul-

tures where growth had ceased for some time (Table 2). Also, it was found that the elevated proteolysis rates in cultures from the later stages of canavanine or puromycin treatments (Table 2) were insensitive to phenylmethylsulphonyl fluoride. Thus, the proteolysis system which recognises abnormal proteins, whether with altered amino acid residues, or altered length, are insensitive to phenylmethylsulphonyl fluoride. This implies that in dihydrostreptomycin-treated cultures which have ceased growth for some time, there are two protease systems, namely one insensitive to phenylmethylsulphonyl fluoride and recognising abnormal proteins and the other sensitive to phenylmethylsulphonyl fluoride and activated by the non- growing condition.

DISCUSSION

The higher proteolysis rates found in dihydrostreptomycin-treated cultures compared to untreated ones appear to be specifically due to the nature of the proteins synthesised. This a priori requires that dihydrostreptomycin-affected ribosomes are functional in translation. The proteolysis rates, which depend on the nature and extent of abnormality of the proteins, progressively increased as the inhibition of growth became more severe; the loglo of proteolysis rates

290 The Susceptibility of Streptomycin-Induced Proteins to Proteolysis

varied directly with the extent of growth inhibition. Further, effects on proteolysis rates and growth inhibition were first detected at the same time. This relationship indicates that dihydrostreptomycin-induced misreading is a primary or contributory mechanism for the antibiotic effect of dihydrostreptomycin, i. e. irreversible inhibition of growth, and not second- ary.

In order to evaluate the importance of misreading for the antibiotic effect, one must consider the size of the misreading effect found or to be predicted. Im- mediately there is a difficulty, because whilst the extent of growth inhibition at any time during dihydrostreptomycin treatment could be quantified, values for the extent of misreading at such times could not. The reason for this is that, although proteolysis rates are first order and can therefore be expected to depend directly on the concentration of susceptible proteins and on the nature of their abnormality, the factors in such a relationship could not be ascertained. These are the ones relating measured proteolysis rates to the actual peptide bonds cleaved (since a large protein requires multiple hydrolysis of peptide bonds before it becomes soluble) and to the extent of abnormality in an amino acid sequence which is required to render a protein susceptible to proteolysis. There- fore, one is limited to a qualitative analysis of the data. For instance, a comparison can be made between proteolysis rates and the nature and extent of abnormality of proteins induced by various agents with known effects on the proteins synthesised. Thus, puromycin acts as an analogue of the 3' end of an aminoacyl-tRNA. It binds weakly to ribosomes in the A site and reacts with peptidyl-tRNA to form peptidyl-puromycin which is then released from the ribosomes [41- 431. It seems that after degradation of the abnormal protein, the puromycin can recycle in the process [29]. The frequency with which puromycin can react with peptidyl-tRNA, and hence the size of the released product, depends on the concentration of puromycin, at least in mammalian systems [44]. Therefore, unless the attached nucleoside per se renders the puromycin-peptide susceptible to proteolysis, this is a model for premature release of peptides.

Canavanine is an analogue of arginine, which occurs as about 8 mo1/100 mol of all amino acids. It should therefore be incorporated to that extent by the arginine-requiring strain of E. coli growing in the absence of arginine but presence of canavanine. In view of the structural similarity between arginine and canavanine, this is a model system for completely efficient misreading, which is however conservative and restricted to the translation of the codons for one amino acid.

By contrast, as a consequence of the dihydrostreptomycin-ribosome interaction misreading could occur to cause nonsense. This would result in pre-

mature termination, as caused by puromycin, but without the physical combination of streptomycin with the nascent chain and would give proteins of a length depending on the specificity and frequency of such mistranslations. Misreading resulting in missense could be compared with the effect of canavanine treatment. However, whereas canavanine can be incorporated into proteins synthesised by the total cellular ribosome population and in response to all arginine-specifying codons, dihydrostreptomycin would only cause missense translation by a proportion of cellular ribosomes, i. e. those combined with the drug. It also would probably not cause the mistranslation of all possible codons to result in missense. This would result in only a portion of cellular proteins being abnormal and containing only some missenses. On the other hand, depending on the specificity of misreading, a greater variety of amino acids could be replaced non-conservatively. Hence, although the concentration of abnormal proteins in dihydrostreptomycin-treated cultures would be less than in canavanine-treated ones, the nature of the abnormality would be greater and could render these proteins more susceptible to proteolysis.

When the proteolysis rates were compared in cultures which had just reached complete growth inhibition by either puromycin, dihydrostreptomycin or canavanine, they were found to rank in that order, with puromycin causing the highest proteolysis rate. Thus this hierarchy correlates with the accepted or postulated effects these drugs have on the synthesis of proteins. However, the relative extent to which dihydrostreptomycin induced misreading or nonsense cannot be evaluated quantitatively by these methods. In addition, a mechanism by which dihydrostreptomycin would cause the premature release of peptides by affecting the stability of the peptidyl-tRNA . ribosome complex cannot be excluded.

Theoretical estimates of the extent and nature of misreading in dihydrostreptomycin-treated cultures at defined stages of growth inhibition are also limited. The relevant parameters can be identified but their quantitative contribution is not known in all cases. These parameters are as follows.

a) The proportion of ribosomes combined with dihydrostreptomycin : minimum estimates range from 15 - 27 % to 45 - 80 % at growth inhibitions from 10 % to complete inhibition, respectively [36].

b) The selectivity of misreading which determines the incidence of those codons which can be misread. From studies in vitro this was shown to be only one base in a codon, and this either guanine or cytosine in the 5' position; or any base in an internal position misread as adenine or uracil. This was further restricted with certain base neighbours [6,45]. This selectivity may not necessarily apply to the situation in vivo, although a necessarily limited study of the strepto-

J . Hewitt and M. Kogut 291

mycin-induced misreading in vivo of all three nonsense codons by a nonsense tRNA suppressor indicated it to be the same [46].

c) Due to the degeneracy of the code [47], not all misreadings result in missense.

d) The efficiency of misreading of possible codons. It can be calculated from the data of Davies et al. above [6], that between 8.0 and 0.6% of all possible codons were misread to result in missense, remember- ing again that this applies to the situation in vitro. Thus, although no definite values may be given, it seems that the predicted extent of misreading is small.

The next question to be considered then, is whether misreading could theoretically cause complete inhibition of growth. With respect to this, situations are considered where faulty translation occurs, and the extent of this and its toleration or not by the organism, is described. Logarithmically growing wild- type cells have a low background level of ambiguity in translation. This was thought to be of the order of 0.1 % by Gorini [48] or rather more, namely 1.8 % based on the extent of translational leakiness causing suppression of a nonsense mutation [46]. Strains carrying tRNA suppressors or the ram mutation [49] have additional mistranslation effects. These can be estimated from studies of their efficiency in suppression of a nonsense or missense mutation. It can be assumed that the tRNA suppressor will recognise and ‘misread’ with about the same efficiency as it causes suppression, the specific codon wherever else it appears. The extent of misreading as a percentage of all codons then becomes the product of the percentage suppression frequency and the occurrence of the suppressed codon. For missense, the occurrence of a codon was determined from the molar proportion of the encoded amino acid, corrected for by the degeneracy of the code. Terminator codons were estimated to be 0.21 of all translated codons, on the basis of the average number of amino acids per protein [50], each protein being equivalent to one terminator codon. Thus, two strains with missense suppressors [51] caused the mistranslation of 0.24% and 0.34% of codons, and nonsense suppressors [52] of 0.1 -0.006% of codons. ram mutations cause less selective mistranslation and so the measured percentage suppression of one codon could be extrapolated to be at least the same order for all codons. In one case, the suppression was found to be 3.3% [46]. All strains with suppressors or ram mutations, or even strains carrying both ram and missense suppressors, can grow, though sometimes at lower rates (20% less or severely diminished, respectively) when compared with the wild-type parent strain [52]. Thus the above extent of ‘misreading’, namely at least up to 3.3

Further information concerning the relationship between the extent of misreading and growth inhibition comes from the following observations on

can be tolerated.

streptomycin treatment. The growth of ram mutants is inhibited by lower concentrations of streptomycin than the ram’ parent [49]. The same is true for another strain which additionally carries a missense suppressor. This is more sensitive than its parent strain carrying the ram mutation and a restrictive allele (strA 1) which restricts translational ambiguity. These results are explained by the streptomycin concentration dependence for the extent of misreading as demonstrated for conditionally streptomycin- dependent mutants [53] and by the cumulative misreading effects of streptomycin, ram and tRNA suppressors.

Information on the extent of misreading which cannot be tolerated can be obtained from studies where cultures of mutants are grown in the presence of analogues of amino acids for which they are auxotrophic. This is equivalent to misreading to the extent of that amino acid’s occurrence in proteins. Incorporation of canavanine, azatryptophan or 0- methyl threonine which are analogues of arginine, tryptophan and threonine respectively and which occur in proteins to 8, 1 or 7.5% respectively lead to irreversible growth inhibition of cultures 154,551. Thus, ‘misreading’ of 8 % or less of codons can cause irreversible growth inhibition.

In summary, it seems that since the onset of misreading, as indicated by increased proteolysis, and of growth inhibition, occurred at the same time in dihydrostreptoniycin-treated cultures, the misreading could have a primary, or contributory role in causing the antibiotic effect of dihydrostreptomycin in streptomycin-sensitive strains. Further, although the extent of misreading could not be directly quantified (and was thought to be small), its overall effect could be evaluated from a survey of the level and degree of misreading in E. coli strains which could either be tolerated, or cause growth inhibition. This again suggested that misreading, followed by proteolysis of abnormal proteins, could have at least a contributary role in causing irreversible growth inhibition. Of course, the converse may also be true; namely that the increasing scarcity of normal proteius in dihydrostreptomycin- treated cultures contributes to growth inhibition.

J . H. thanks the Medical Research Council for a research training award.

REFERENCES 1. Davies, J., Gilbert, W. & Gorini, L. (1964) Proc. Nut/ Aead.

2. Davies, J., Gorini, L. & Davies, B. D. (1965) Mol. Pharmacol,

3 . Van Knippenberg, P. H., van Ravensway Claasen, J. C., Grijm- Vos, M., Veldstra, H. & Posch, L. (1965) Biochim. Biophys. Acta, 95, 461 -473.

4. Anderson, W. E., Gorini, L. & Breckenbridge, L. (1965) Proc. Natl Acud. Sci. U.S.A. 54, 1076-1083.

Sci. U.S.A. 51, 883-890.

I , 93 - 106.

292 J. Hewitt and M. Kogut: The Susceptibility of Streptomycin-Induced Proteins to Proteolysis

5 . Nirenberg, M. W. & Leder, P. (1964) Science (Wash. D.C.)

6. Davies, J., Jones, D. S. & Khorana, H. G. (1966) J . Mol. Bid.

7. Valentine, R. C. & Zinder, N. D. (1964) Science (Wash. D.C.)

8. Gorini, L. & Kataja, E. (1964) Proc. Natl Acad. Sci. U.S.A. 51,

9. Gorini, L. & Kataja, E. (1965) Biochem. Biophys. Res. Commun.

10. Orias, E. & Gartner, T. K. (1966) J . Bacteriol. 91, 2210-2215. 11. Whitfield, H., Martin, R. & Ames, B. (1966) J . Mol. Biol. 21,

12. Reale-Scafati, A. (1967) Virology, 32, 543-552. 13. So, A. G. & Davie, E. W. (1964) Biochemistry, 3, 1165- 1169. 14. Zimmerman, R. A., Rosset, R. & Gorini, L. (1971) J . Mol.

15. Wallace, B. J., Tai, P. C., Herzog, E. L. & Davis, B. D. (1973)

16. Modolell, J. & Davis, B. D. (1970) Proc. N a t l Acad. Sci.

17. Scolnick, E., Tompkins, R., Caskey, C. T. & Nirenberg, M.

18. Modolell, J . & Davis, €3. D. (1968) Proc. Nut1 Acad. Sci.

19. Modolell, J. & Davis, B. D. (1969) Nature j lond.) 224, 345-

20. Kogut, M. (1969) J . Gen. Microbiol. 9, 42-49. 21. Luzzatto, L., Apirion, D. & Schlessinger, D. (1969) J . Mol.

22. Luzzatto, L., Apirion, D . & Schlessinger, D. (1969) J. Bacteriol.

23. Kogut, M. & Prizant, E. P. (1975) Antimicrob. Agents Chemo-

24. Kogut, M. & Prizant, E. P. (1970) FEES Lett. 12, 17- 20. 25. Plotz, P. H. &Davis, B. D. (1962) J . Bacteriol. 83, 802-805. 26. Pine, M. J . (1967) J . Bacteriol. 93, 1527-1533. 27. Lin, S. & Zabin, I . (1972) J . Bid. Chem. 247, 2205-2211. 28. Goldschmidt, R. (1970) Nature (Lond.) 228, 1151 - 1154. 29. Goldberg, A. L. (1971) Nut. New Biol. 234, 51 -52. 30. Hipkiss, A. R. & Kogut, M. (1973) Biochem. SOC. Trans. 1,

145, 1399- 1407.

18,48- 57.

144, 1458- 1459.

487-493,995-1001.

18,656- 663.

335- 355.

Biol. 57,403 - 422.

Proc. Natl Acad. Sci. U.S.A. 70, 1234- 1237.

U.S.A. 67, 1148-1155.

(1968) Proc. Nut1 Acad. Sci. U.S.A. 61, 168-7774.

U.S.A. 67, 1148-1155.

348.

Biol. 42, 315-335.

99,206 - 209.

ther. 7, 341 - 348.

595 - 596.

31. Pinkett, M. 0. & Brownstein, B. L. (1974) J . Bacteriol. 119,

32. Kogut, M., Lightbown, J . W. & Isaacson, P. (1965) J . G m .

33. Kogut, M. & Harris, M. (1969) Eur. J . Biochern. Y,42-49. 34. Kempshead, J . T. & Hipkiss, A. R. (1974) Eur. J . Biochem. 45,

35. Kogut, M., Lightbown, J. W. & Isaacson, P. (1966) J . Cen.

36. Prizant, E. P. & Kogut, M. (1975) FEES Lett. 50, 37-42. 37. Prouty, W. F. & Goldberg, A. L. (1972) J. Biol. Cliem. 247,

38. Pirie, M. J. (1965) Biochim. Biophys. Acta, 104,439-456. 39. Willetts, N. S. (1965) Biochem. Biophys. Res. Commun. 20,

40. Brunschede, H. & Bremer, H. (1971) J . Mot. Biol. 57, 35-57. 41. Allen, D. W. & Zamecnik, P. C. (1962) Biochim. Biophys. Acta,

42. Nathans, D. (1964) Proc. Nut1 Acad. Sci. U.S.A. 51, 585-592. 43. Smith, J. D., Traut, R. R., Blackburn, G. M. & Monro, R. E.

44. Williamson, A. R . & Schweet, R. (1965) J . Mol. Biol. 11,

45. Strigini, P. & Gorini, L. (1970) J . Mol. Biol. 47, 517-530. 46. Strigini, P. & Brickman, E. (1973) J . Mol. Biol. 75, 659-672. 47. Crick, F . H. C. (1966) J . Mol. Bid. 19, 548-555. 48. Gorini, L. (1974) Ribosomes (Nomura, M., Tissieres, A. &

Lengyel, P., eds) pp. 791 -803, Cold Spring Harbor Labor- atory, New York.

345- 350.

Microbiol. 39, 155 - 164.

535 - 540.

Microbiol. 42, 333- 346.

3341 - 3352.

692- 696.

55,86- 874.

(1965) J. Mol. B i ~ l . 13, 617-628.

358 - 372.

49. Rosset, R. & Gorini, L. (1969) J . Mol. Biol. 39, 95-112. 50. McQuillen, K. (1965) in Function and Structure in Micro-

organisms (Pollock, M. R. & Richmond, M. H., eds) p. 134, Cambridge University Press.

51. Biswas, D. K. & Gorini, L. (1972)J. Mol. Biol. 64, 119- 134. 52. Gorini, L. (1971) Nat. New Bid . 234, 261 - 264. 53. Gorini, L. & Kataja, E. (1965) Biochem. Biophys. Res. Commun.

54. Schachtele, C. F. & Rogers, P. (1965) J . Mol. Biol. 14, 474-

55. Rabinovitz, M., Finkelman, R. L., Reagen, R. L. & Brutman,

18,656- 663.

489.

T. R. (1969) J. Bucteriol. 9Y, 336- 338.

J. Hewitt, Institut fur Genetik der Universitiit zu Koln, Weyertal 121, D-5000 Koln-Lindenthal, Federal Republic of Germany

M. Kogut, Department of Biochemistry, King’s College London, Strand, London, Great Britain, WC2R 2LS

an investigation of mistranslation in vivo induced by streptomycin by an examination of the...

Documents