748 This work was supported in part by grants from the National Science Foundation and the National Cancer Institute of National Institutes of Health. 4 Waksman, 8. A., and Woodruff, H. B., Proc. Soc. Exp. Biol. Med., 45, 609 (1940). 2 Brockmann, H., Forts, Chem. Org. Naturstoffe, 18, 1 (1960). > Reich, E., Acs, G., and Franklin, R. M. (in preparation). 4 Acs, G., and Reich, E. (unpublished observations). 5 Reich, E., Franklin, R. M., Shatkin, A Science, 184, 556 (1961); (1962). * Goldberg, I. H., and Rabinowitz, M., Science, 186, 315 (1962). ? Hurwitz, J., Furth, J. J., Malamy, M., and Alexander, M., Proc. U.S. Nat. Acad. Sci., 48, 1222 (1962). * Kirk, J. M., Biochim. Biophys. Acta, 42, 167 (1960). *Rauen, H, M., Kersten, H., and Kersten, W., Z. physiol. Chem. Hoppe-Seyler, 821, 139 (1960). 3» Kawamata, J., and Imaniski, H., Biken’s J., 4, 13 (1961). 11 Burchenal, J. H., Oettgen, H. F.. Reppert, J. A., and Coley, V., Ann. N.Y. Acad, Sci., 89, 399 (1960). .J., and Tatum, E. L., Proc. U.S. Nat. Acad, Sei., 48, 1238 NATURE 3/1/95 November. 24, 1962 as , A. J, Be B., Franklin, R. M., and Tat ' Shatkin, A. phe cian 6, 277 (962). um, By ry sar M., and Berg, P., Proc. U.S. Nat. Acad. Sci., By 4 Goldberg, I. H., and Rabinowitz, M., Biochim. Biophys. Ada, K 202 (1961). / 18 Goldberg, I. H., Biochim. Biophys. Acta, 51, 201 (1061). 16 Goldberg, I. H., and Rabinowitz, M., Biochem. Biophys. Res, Come, 6, 394° (1961). 17 Rabinowitz, M., and Goldberg, I. H., Fed. Proe., 21, 384 (1909, 18 Kahan, F. M., Fed. Proc., 21, 371 (1962). 1 Brockmann, H., and Manegold, J. H., Ber., 98, 2971 (1960), 20 Miller, W., Naturwiss., 40, 156 (1962). 1 Schachman, H. K., Adler, J., Radding, C. M., Lehman, I. R., gy Kornberg, A., J. Biol. Chem., 285, 3242 (1960). a Furth, J. J., Hurwitz, J., and Goldmann, M., Biochem. Biophy Res. Comm., 4,°431 (1961). : *3 Goldberg, I. H., Rabinowitz, M., and Reich, B., Proc. U.S! %, Acad. Sci. (in the press). % Sueoka, N., J. Mol. Biol., 8, 31 (1961). . % Kersten, W., Biochim. Biophys. Acta, 47, 610 (1961). VOL. 198 APPEARANCE OF STREPTOMYCIN RESISTANCE FOLLOWING THE UPTAKE OF TRANSFORMING DEOXYRIBONUCLEIC ACID IN PNEUMOCOCCUS By Dr. HARRIETT EPHRUSSI-TAYLOR* Laboratoire de Genetique Physiologique du ‘C.N.R.S., Gif-sur-Yvette, Seine-et-Oise N bacterial transformation, the experimenter is able to examine the physiological action of normal hereditary determinants, introduced into cells in the ‘form of DNA, by examining the events leading to the manifestation of the newly acquired hereditary trait. Thus far, such studies are limited in number??. In one’, the appearance of resistance to streptomycin was followed in populations of pneumococci which had been made to fix DNA from a streptomycin- resistant donor strain. The conclusion reached was that the resistant phenotype appears as a discrete change from sensitive to resistant: no stages of partial resistance could be recognized. It was found, furthermore, that the probability of a cell becoming resistant was normally distributed over a time- interval ranging from about 15 min to 90 min follow- ing penetration of the streptomycin-resistance gene, even though DNA fixation had been limited to a 5- min period. The most probable moment for a cell to become resistant was about 60 min following uptake of DNA. The interpretation of these results at the time of their publication was rendered difficult owing to the absence of information concerning the mechanism of streptomycin-resistance and the types of syntheses involved in its establishment. Later, ‘a brief description was made of experiments showing that the discrete event, described here, is not in fact the development of typical streptomycin- resistance, but, after all, only an intermediate stage in its development*. Experiments documenting this contention are presented here, in conjunction with a hypothesis concerning the mode of action of the streptomycin-resistance gene. The hypothesis is a development of the recently published theory of Spotts and Stanier‘ concerning the mode of action of strepto- mycin and the nature of streptomycin-resistance. Since it may open some interesting new approaches to the study of gene action, and, in particular, to the question of the relationship between genes and ribosomes, the publication of these experi- * Present address: Developmental Biology Center, Western Reserve University. Cleveland, Ohio. ments, and the accompanying hypothesis ond mode of action of the streptomycin-gene, § worth while. , (1) Thé experimental demonstration of the a of resistance. All investigations of the appearant antibiotic resistance following uptake of transfor DNA uso a single basic procedure. Followi period of DNA fixation which is sharply limited destruction of unabsorbed DNA with DNase, the’ are diluted into fresh medium and incubate 37° C. At intervals, samples are withdrawn’ plated in agar containing the antibiotic. The numg of cells able to give rise to a colony in the presen the antibiotic are thus scored. Fig. 1, curve a, sht how pneumococci transforming for streptomyg resistance develop this ability. This curve is typy of those obtained by Fox? and Schaeffer' as well ag me. The number of cells able to give rise to coi0g in streptomycin-agar rises rapidly from abou a 15th min following DNA fixation. A shouldeg observed at about 80-90 min, following which}& numbers increase again, but at a slower, exponeng rate characteristic of the overall population incr of the growing culture. Various experiments*” J shown that: (1) at 90 min, virtually every cell 1 fixed a transforming molecule is able to fo colony in the presence of streptomycin; (2) that increase observed after 90 min is due to the fo of genetically transformed daughter cells. In it is established that the transmission of an acqu! gene to both daughter cells may begin as early 959 second generation after DNA fixation’, that 5% about 45 min. It may, however, begin only’& third or fourth generation in some cells’. The transmission of an acquired gene is, however g reflected immediately in the numbers of str mycin-resistant colony-forming units observed, OWf to the tendency of sister cells to remain atts after cell division. Differences in the degree to W: shoulder is observed at 80-90 min are almost cert due to differences in the extent of chain formatt different media. : November 24, 1962 Ie. 4656 iT \ \ ewes oe ee Neo Ke e I \ ow 1 1 me wre Putri J J | { J | 60 120 180 Min after contact with DNA ‘1, Evolution of the numbers of transformed cells able to form nies in streptomycin agar. A, Transformation culture growing Wwabsence of streptomycin ;' B, streptomycin has been added 90 min after DNA uptake; C. streptomycin ag been added 135 min after DNA uptake CET TER eRe \ f, in the agar medium used to score the resistant “the concentration of streptomycin is insufficient arrest rapid growth of the recipient strain, Midual metabolism in the presence of streptomycin @l cause some transformants to complete the change mM sensitive to resistant on the agar plate?. Thus, transformed phenotype will seem to appear earlier ncentrations of streptomycin below a critical l, and the precocity of the appearance of the iotype will appear to be a function of streptomycin @icentration. Therefore, in order to avoid under- imation of the time required for resistance to de- wlop it is necessary to challenge the cells in agar Whtaining streptomycin at a concentration yielding # maxi ly selective effect. In the experiments eed here, it was found that the rate of appearance t-etreptomycin-resistant cells is the same at all Wncentrations of streptomycin equal to or greater than 200 pg/ml. The majority of experiments were Performed at 200 pg/ml., but in some the concentra- were higher. *(2) Limitations of the procedure for demonatrating Ntistance. The foregoing procedure for determining the appearance of resistance, adequate at first sight, ‘fact, leaves one parameter unexplored. The method reveals only when a transforming cell can 8 colony at a maximally selective concentration streptomycin. It does not tell us whether the formant does so at once, or whether its growth ‘ad division is temporarily suspended by the chal- lenge. The streptomycin-resistant donor strain is ‘mpletely indifferent to streptomycin at the con- Cmirations used: Is the newly resistant cell, ch yields a colony, also really indifferent to tomycin ? (8) Two steps in development of resistance. To test in question, instead of challenging the transformants Streptomycin agar, a small amount of strepto- eee 78 added to the cells in liquid medium, at time when resistance is generally presumed to be “mplete. Following the addition of streptomycin, NATURE 749 platings in agar were performed in order to determine the evolution of the numbers of resistant cells: Fig. 1 shows the results of one such experiment, in which cells which had fixed DNA for 2 min were diluted 100-fold into fresh medium and the culture divided into three portions: (a) no streptomycin is present in the liquid culture and platings are made directly into streptomycin agar; (b) 50 ug/ml. of streptomycin were added after 90 min of growth, and platings made into streptomycin agar (at 200 pg/ml.); and, {c) 50 ug/ml. of streptomycin were added after 135 min of growth, and platings are made in streptomycin agar. If, as is generally believed, all transformants have achieved the synthesis of the streptomycin- resistant phenotype by 90 min, there should be no difference in the numbers of streptomycin-resistant cells present in these three liquid cultures. This is, however, clearly not the case. The increase in the numbers of resistant cells present in the cultures receiving a small amount of streptomycin is almost immediately arrested by the antibiotic. Thus, the ‘immediate replication of the newly formed resistant cells is blocked by as little as 50 pg/ml. of streptomy- cin. Yet these cells are able to form colonies in agar containing 200 ug/ml. or more. Other experiments showed that, in fact, streptomycin transformants become completely indifferent to streptomycin only after some 150-180 min have elapsed following DNA fixation. , . Two explanations of these observations can be offered: (1) that resistance develops in two steps. First, the bacteria are altered so that streptomycin is no longer bacteriocidal, and secondly, they become completely indifferent to streptomycin. If this explana- tion is to be retained, it must be assumed also that cells can pass from the first state to the second in the presence of streptomycin. (2) That the cells which survive the streptomycin challenge are not genetically transformed. For example, the acquired factor can be supposed to be not yet a part of the linear array of genes of the bacterial chromosome, but transmitted via an extra-chromosomal mechanism. Streptomycin could then be supposed to block the extra-chromo- somal mechanism so that the majority of the daughter cells produced in its presence would be streptomycin-sensitive and therefore die. This would be analogous to the situation found in the induction of ‘petites’ by acriflavine acting on yeast®. The eventual formation of a colony in streptomycin-agar would reflect a shift from the extra-chromosomal state to a chromosomal state, achieved through recombina- tion at one of the numerous cell divisions which the mother cell could make. Results of a number of types of experiments invalid- ate the second hypothesis. One critical argument against it is the fact that when a cell acquires a DNA particle, recombination does ensue very shortly thereafter**, Further, it is reported that when a particle of transforming DNA is genetically marked at several points, so that it is able to give rise to several types of different, recognizable recombinants, one observes that a unique recombinant type is formed from a single absorbed particle, most if not all of the time’®. Were the acquired particle transmitted at the outset by an extra-chromosomal mechanism prior to recombination, this result could not be observed. The first hypothesis is, therefore, to be’ retained in considering why streptomycin arrests the multiplica- tion of resistant cells newly formed by transformation. Accordingly, in order to explain the results exempli- fied hy Fig. 1, we can assume that even though from 750 90 min on, every cell which acquired the streptomy- cin-resistance gene can form a colony at high con- centrations of streptomycin they do so only after a considerable period of arrested growth. In other words, at this stage of phenotypic transformation,,. streptomycin is a bacteriostatic substance from the effects of which the transforming cell can eventually escape. This characterizes what we shall call stage 1 in the development of resistance, while complete indifference charactérizes stage 2, the definitive state. Another type of experiment confirms this point of view, and, in addition, informs us of further characteristics of stage 1. Following a challenge of 500 pg/ml. of streptomycin for a 30-min period at 37° C, surviving transformed cells are washed on a membrane filter to eliminate unbound streptomycin, transferred to fresh medium by washing. them off the membrane, and- their growth followed by plating samples at intervals, in two different media: agar with and without streptomycin. In the experi- ment shown in Fig. 2, the resistant transformants were selected 60, 110 and 180 min after DNA fixation as described. One may note the following features of the curves in Fig. 2. (1) There is a small lag in the onset of replication of the resistant cells selected 180 min after DNA fixation which is not observed when strepto- mycin is simply added at this time to 4 transforming culture and left there. The lag observed in Fig. 2 is almost certainly caused by the vigorous aeration of the cells during washing on the membrane filter (Pneumococct are microaerophilic). (2) Growth of streptomycin-resistant cells selected 60 and 110 min after DNA fixation is severely retarded, even though NATURE ? 5,000 | / / / C / : J e i x / A # 1,000 = ‘ ¥ / f 500 Z er Og ff / 0 / e a / f 100 x Le if ' zg & 1 75 160 210 l | ! l | 1 2 3 4 5 6 Hours of incubation Fig 2. Onset of division of streptomycin-resistant transformants in the absence of extracellular streptomycin. The resistant transformants were selected: A, 60 min; B, 110 min; C, 180 min after DNA uptake by treating the transforming population with 500 yvg/ml. of streptomycin for 30 min. Survivors were collected on a ‘Millipore’.membrane, washed, resuspended in medium, and their growth followed: x, on plates containing no streptomycin and: @ on 200 ug/ml. of streptomycin ‘transforming DNA. Yet some of these cells November 24, 1962 the antibiotic has been removed. This means either the cells have fixed enough streptomycin that the extracellular concentration of the antibig: is no longer critical, or that the 30-min treatment hy inflicted a finite damage which is not subject , reversal by the removal of streptomycin. The la seems more likely since the amount of streptomyg bound to bacteria is very small". (3) No sensitive survive the selection at 500 pg/ml., and few ory sensitive progeny are formed by stage 1 resistay cells. By extrapolating the exponential slopes of ty curves of Fig. 2, one can calculate from curve ¢ th delay caused by aeration, and from curves a and } th delay caused by the combined factors of aeration ay streptomycin-inflicted damage. Correcting for ty delay caused by aeration, one finds that the cells ing required 135 min of incubation to resume exponentig increase, while the cells in b required 85 min. Thy difference between these two times is 50 min, whid is the same as the difference in the incubation timy of the two cultures prior to the streptomycin ch@ VOL. 198 ‘lenge. In other words, the time required for definitiy resistance to develop is constant, and independeg of the moment of application of the streptomy¢é challenge. Thus, cells which are at stage 1 in th development of resistance, and which may hg arrived at this stage at very different moments, 4 homogeneous population in so far as their attainmel of definitive resistance is concerned. With respect! definitive resistance, primary transformants 3% apparently no different from their second, thirdf even fourth generation daughters. ii In the experiment of Fig. 1, 50 pg/ml. of strepy mycin was added to the liquid culture, while in experiment of Fig. 2, 500 ug/ml. were added. concentrations arrested the multiplication off@ resistant transformants. Irrespective of whethetdj damage to the cells was inflicted by 50 or 500 pgfy and of whether the streptomycin was left in cong with the survivors, the moment of onset of inorg of the streptomycin-resistant cells was at aboutgy min. This again suggests that streptomycin infj finite damage on stage 1 transformants, and that. recovery is independent of the external concentragg of streptomycin. To show this more clearly@ experiment was performed in which fluctuationg the numbers of streptomycin-resistant colony-forng units was followed in a control and two streptomyg containing cultures. The latter received 50 and,§ ug/ml. of streptomycin, respectively, 60 min si DNA fixation. Fig. 3 shows the results of sucltg experiment. It can be seen that the time required stage 1 resistant transformants to resume divig after the addition of streptomycin is approximay the same, irrespective of the external streptomy concentration. Hence, the conversion of 9 stagy resistant transformant into a definitely resistant is essentially independent of streptomycin concen™ tion. Further, the damage inflicted on the stag cells must be finite and independent of streptom™ concentration, within the limits explored. The most striking feature of the way in definitive resistance develops in a transfort population is that cells destined to transform, © their immediate progeny, show this resistance ate same time, that is, about’ 180 min after fixatia original transformants and some are theif } November 24, 1962 yo: 4855 ——— wt x5 978 oes f { a th 50 100 150 200 Min of incubation Wt. Evolution of the numbers of streptomycin-resistant cells etransforming culture x in the absence of streptomycin; in the “mlence of @ 50 uve/ml.;O, 500 ug/ml. of streptomycin added 60 ‘pi‘after DNA uptake. Some of the irregularities in the curves 4 are probably due to synchrony of division. toc id or later generation progeny. Therefore, as maforming cells grow and divide, they must produce enters which are similar to themselves not only Motypically but also with respect to the degree to mn they have developed definitive phenotypic ince. Since genetic integration usually occurs of the first two or three divisions following DNA mpon, it is difficult to imagine that this phenotypic ormity of transformants and their progeny is hed by the process of genetic integration itself. other hand, DNA fixation has occurred during y short interval. Its penetration into the cell “very well be the event which initiates the of phenotypic transformation. is were the case, the following mechanism of seoced. On penetration of the DNA, the strepto- w20'r gene immediately induces the formation of system which confers resistance. Since the com- fey menotype is manifested only some 180 min after be ation, we can suppose that resistance results the synthesis of a very large number of specific WEomolocules. As cell division proceeds, both the . ting system initiated by the acquired gene and lertic macromolecules which it determines are lt uted more or less equally between sister cells. those cells where the resistance gene is fixed hently by genetic recombination, the resistance thes Will also be transmitted. It is only in these cells lag U2, 8enorating system will be stable enough for apt eence to be manifested. What. then, is l resistance? As shown by Fox, it is a discrete ee? which occurs on an average after about 60 ave elapsed following DNA-uptake, and which & normal but fairly wide distribution with to the moment it occurs!. As shown by the NATURE 751 foregoing experiments, it is a change which enables a cell to survive a challenge of maximally selective amounts of streptomycin, and to escape from a strong bacteriostatic effect of the antibiotic. Further, the rate at which a stage I resistant escapes is independ- ent of the external streptomycin concentration in the growth medium. A suitable explanation of stage 1 resistance was not evident so long as theories of the nature of strepto- mycin resistance were based on supposing the resist- ant cell impermeable to streptomycin. Even with the publication of a theory? to the effect that, in the presence of streptomycin, sensitive bacteria synthesize an abnormal membrane constituent which results in disruption of transport mechanisms, an explanation of stage] resistance did not seem possible. Supposing that, at the onset, the acquired resistance gene were to confer on the cell the capacity to form normal membrane substance in the presence of streptomycin, at early stages the cell membrane could be at best a mosaic, for the old membrane and membrane-forming system should still be present in the cell. It is hard to see how a mosaic membrane could confer on cells an immunity to the lethal effects of streptomycin. The recent hypothesis of Spotts and Stanier‘* provides, on the other hand, an explanation of the nature of stage 1 resistance. According to these authors, streptomycin attacks the ribosomes of sensitive cells, causing their disruption. Resistant cells, according to the theory, contain ribosomes which do not combine with streptomycin, and are, therefore, resistant to its action. There is, indeed, some direct evidence in favour of this view!*. In the light of this hypothesis, stage 1 resistance can be inter- preted as resulting from the synthesis of adequate numbers of streptomycin-resistant ribosomes so that at least one copy of each of the different messenger RNA’s of the cell which are necessary for the continua- tion of vital specific functions could be housed in streptomycin-resistant ribosomes. Bacteriostasis would ensue, however, at this stage owing to the destruction of residual streptomycin-sensitive ribo- somes, which could still represent the majority of the ribosomes of the cell. Stage 1 resistant cells would recover their ability to divide as soon as the streptomycin-resistant ribosome population were built up to a level compatible with normal growth and division. Recovery-rate would be independent of the amount of streptomycin in the system, for recovery would result from the function of surviving streptomycin-resistant ribosomes. The ribosome hypothesis is particularly satisfying because it explains why stage 1 resistance appears after an interval which is normally distributed over a fairly wide time-range. There are presumably many different messenger RNA’s determining vital functions which must be housed, and the probability that any one cell possesses one streptomycin-resistant ribosome-messenger RNA particle of each type would be expected to be dis- tributed in this way, assuming random association of the RNA with ribosomes. Further, the ribosome hypothesis of streptomycin action can explain why all transforming cells and their progeny show definitive resistance at about the same time: the existing ribo- somes would be shared at each division. However, it should be mentioned that there is one fact concerning the action of streptomycin which the Spotts and Stanier theory does not explain. This is the observation that if chloramphenicol and streptomycin are added simultaneously to sensitive 752 cells, streptomycin has no lethal action. It appears, thus, that the lethal effects of streptomycin are the consequence of protein synthesis. It has been proposed that streptomycin does not enter cells unless a special permease is synthesized", following contact of cells with streptomycin, and this could account for the protective effect of chloramphen- icol. Since so many of the biological actions of streptomycin can be explained by the theory of Spotts and Stanier, including the very particular way in which resistance develops following trans- formation, one is inclined to conclude that only a minor modification of it may be necessary in order to explain why chloramphenicol eliminates the bacterio- cidal effect of streptomycin. This work was supported by a grant from the Rockefeller Foundation to the C.N.RB.S. NATURE November 24, 1962 1 Fox, M. S., J. Gen. Phys., 42, 737 (1958). ? Lack, 8. F., and Hotchkiss, R. D., Biochim. Biophys. Acta, 4 (1960). 3 Bphrussi-Taylor, H., in Microbial Genetics, 132 (Camb. Univ. y « Spotts, C. R., and Stanier, R., Nature, 192, 633 (1961). * Schaeffer, P., Ann. Inst. Pasteur, 91, 323 (1956). * Hotchkiss, R. D., in Chemical Basis of Heredity, 321 (Johns Hoy Univ. Press, 1956). 1 Ephrussi-Taylor, H., in Recent- Progress in Microbiol, (Almavist and Wiksells, Stockholm, 1958). . * Ephrussi, B., and Hottinguer, H., Cold Spring Harbor Symp, Quant. Biol., 16, 75 (1951). * Ephrussi-Taylor, H., J. de Chimie-Physique, 58, 1090 (1961), ) ‘M. 8., Nature, 187, 1006 (1960). Voll, M. J., and Goodgal, 5 Proc. U.S. Nat. Acad. Sci., 47, 505 (1961). 1® Ravin, A. W., Hap. Cell Res., 7, 58 (1954). 41 Hancock, R., Biochem. J., 78, 7P (1961). as Anand, Nitya, Davis, B. D., and Armitage, A. K., Nature, lg 13 Speyer, J. F.. Lengyel, P., and Basilio, C., Proc. U.S. Nat. Acad.g 48, 684 (1962). . wo 1 Hurwitz, C., and Rosano, C., Biochim. Biophys. Acta, 41, 162 (1y VOL, ty