Reprinted from the PROCEEDINGS oF THE NATIONAL ACADEMY OF SCLENCES Vol. 51, No. 4, pp. 678-682. April, 1964. INTERACTION OF STREPTOMYCIN AND A SUPPRESSOR FOR GALACTOSE FERMENTATION IN E. COLI K-12* By E. M. Leprersere, L. CAVALLI-SFORZA, AND J. LEDERBERG DEPARTMENT OF GENETICS, STANFORD UNIVERSITY, AND ISTITUTO DI GENETICA, UNIVERSITA, PAVIA, ITALY Communicated February 5, 1964 An observation was made some time ago on a suppressor of a galactose-negative mutation in #. coli. The suppressor action was found to be overcome by a muta- tion to streptomycin resistance. Furthermore, in some of the streptomycin-re- sistant stocks, the action of the suppressor was partially restored by the addition of streptomycin (Table 1). Although the situation was not analyzed in much detail, this brief report may be relevant to current interest in the effect of suppressor genes,!—3 and particularly in the light of Gorini’s suggestion‘ that streptomycin may act as a phenotypic suppressor altering the reading of the genetic code. The Suppressor Mutation —As summarized in Table 1, a Gals” mutant, blocked in the capacity to produce UDP galactose transferase, was isolated after ultraviolet irradiation of E. cold strain Y-87,!! a derivation of K-12. A reversion from it, selected on EMB galactose medium, was indicated to be the result of a suppressor mutation, as lambda phage prepared from it by UV induction transduced Gal+ to all mutants tested, except to Gal;-. The recovery of Gal- recombinants in crosses between the suppressed strain and a Ga/+ strain is also observed in Table 8, cross 2. On closer examination, the suppressed strain, W-1802, can actually be dis- tinguished from wild-type Gal+ by virtue of slower fermentation on EMB indicator Vow. 51, 1964 GENETICS: LEDERBERG ET AL. 679 TABLE 1 History oF THE STRAINS Galactose transferase Name Genotype phenotype Strain no. Wild-type Gait Normal (+) Y-87 Gal mutant Gal,- Absent (—) W-518* Revertant (suppressed) Gal, su(Gal)+ Partially restored W-1801 and W-1802 Streptomycin-resistant Gal,—su( Gal) Sm Streptomycin Present Absent + —ft W-4903 —t +t W-4904 * Intermediate mutations not relevant to the history of these strains! are omitted here as well ag other markers not directly concerned. he phenotype of this strain will be described here as Gait, for slow fermentation. : } The phenotype of these strains is usually negative on the first day of observation and may show some weak fermentation on the second day. It is therefore called here Gal” (very slow). TABLE 2 Cross 1* Sanaa Percentages of Male Alleles for Markers{——————————. Time (min) sut Xyl~ Mt- M+ Laca~ Ade~ 20 77 4 2 0 0 0 40 28 19 16 0 0 0 60 32 32 24 0 0 0 90 52 54 48 7 2 0 * Cross 1, between strains: Mali+ Xyl- Mtl- M+ Th~- Lacy (Gals*+@alo-) Ade~ (W-4884, also: su*) Mal~ Xyl* Mu+ M~- Th* Lacy* (Gala-Galo*) Ade*+ (W-4878, also: su(Gal)) t Selection for Mali * recombinants was carried out after interruption of the mating at the times indicated. Per- centages obtained from 40-50 recombinants each. media, as might be expected of a suppressed strain. A direct assay for transferase carried out by R. L. Soffer (unpublished) confirmed the re-establishment of trans- ferase activity in W-1802. The specificity of the suppressor, su(@al), with respect to other mutants in the same cistron, and other cistrons, has not been determined. Streptomycin-resistant Mutants and Streptomycin-dependent Fermentation (sdf).— When a streptomycin-resistant mutant was selected from the suppressed, galactose- fermenting strain W-1802, it was found that the capacity to ferment had been lost. again if the fermentation test was carried out in the absence of streptomycin, but was similar to that of the parental W-1802 strain in the presence of streptomycin. Ten independent streptomycin-resistant mutants were then selected from W-1802 to test for possible differences. All of them had lost the capacity to ferment galac- tose, at least by the EMB test, although three of them were still capable of ferment- ing at a very low rate. This behavior had not been encountered before in strepto- mycin-resistant mutants from normal, galactose-fermenting strains. When, how- ever, the fermentation test was carried out on EMB media supplemented with streptomycin in concentrations toxic to sensitive strains, it was found that four of the ten resistant strains were capable of fermenting at a rate similar to that of W-1802. In other words, some of the strains had become streptomycin-dependent for the fermentation of galactose (sdf), though not for growth. Genetic instability was a remarkable feature of most of the suppressor-carrying streptomycin-resistant mutants. Mapping the Suppressor Locus.—The strain carrying the su(Gal), gene is a female, methionine auxotroph (F-M~). A Mal,— marker was added by selecting for re- sistance to a virulent phage mutant of lambda,® thus obtaining strain W-4878. The latter was crossed to male W-4884 (Vfr, or very high frequency of recombina- 680 GENETICS: LEDERBERG ET AL. Proc. N. A. 8. tion’). This male injects the Mal locus at about 20 min. The order of entry of the other markers is presented in Table 2. Both male and female are streptomycin- sensitive so that the segregation of the suppressor, su(Gal), carried by the female, can thus be scored directly. The data in Table 2 show clearly that su+(Gal) enters earlier than any of the other loci tested. Therefore, it is closely linked to Mal and probably enters earlier than Mal. Crosses using a streptomycin-resistant marker were carried out with strepto- mycin-resistant Vjfr males (like W-4884 provided by E. A. Adelberg). W-4882 (AB-312) used in cross 2, Table 3, is believed to have the same order of entry as the male used in Table 2, while W-4883 (AB-313) used in cross 3, Table 3, has the TABLE 3 CrossEs 2 AND 3 No. Malt Sm —— Sm. Cross Time (min) recombinants Gals* Gal- Gals Galss* Gal - 2 20-60 314 46 0 0 261 7 3 20 56 55 0 0 0 1 30 42 28 0 0 14 0 40-90 136 71 0 5 90 0 * See second footnote, Table 1. Crosses 2 and 3 were interrupted at various intervals; only time intervals showing differences are reported sepa- rately. Only Mal* recombinants tested. reverse order of entry, with Mal entering at 15 min. The appearance of Gal- segregants in a Gal’ X Gal+ mating is shown by cross 2, Table 3. Because the segregations of Gai and of Sm are independent of time, both markers are believed to be located on the same side with respect to Mal to suggest the order: O...su...- Sm...Mal. Cross 3 shows that Sm enters after Mal with this male, but su+, ex- pected to enter after Sm on the basis of the order given above (reversed with this male as O...Mal...Sm...su...) does not seem to enter to any significant extent. The appearance of a few Gal’ and one Gal-Sm’ is not readily explained. Apart from this inconsistency, which may be due to chance, to the existence of modifiers in this male, or other peculiarities of the chromosomal region under investigation, it would seem that the suppressor may be mapped not far from Sm, away from Mal. Diseussion.—Our findings may be summarized as follows: in a particular strain of £. colt K-12, mutation to streptomycin resistance was found to affect the enzyme galactose-transferase, whose production then becomes streptomycin-dependent. In several other streptomycin-resistant mutants, however, there is an almost com- plete elimination of enzymatic action, both in the presence and absence of strepto- mycin. The strain showing this peculiar behavior was capable of producing enzyme at a subnormal rate, thanks to the presence of a gene suppressing the action of another mutation, which had in turn inhibited the formation of the enzyme. Streptomycin resistance made the action of the suppressor gene streptomycin-dependent. The physiology of the suppressor in question seems to merit further investigation. Genetic studies are incomplete, but a chromosomal or mapping location not far from the streptomycin gene seems reasonable on the basis of the data summarized above. According to present views, many suppressors act by perturbing the code of specifie amino acids, at least partially, in such a way that a mutant making an Vou. 51, 1964 GENETICS: LEDERBERG ET AL. 681 altered and inactive protein because of an amino acid change, under the action of the suppressor can make some normal protein.'~? On the other hand, strepto- mycin resistance is believed to be a property of the ribosomes. This may seem at first sight to conflict with the view that streptomycin resistance is recessive, at least in F. coli.2 In heterozygotes. both types of ribosomes, streptomycin-sensitive and resistant, should be produced; if, in the presence of streptomycin, only the former do not function, protein synthesis would be reduced to one half, presumably compatible with life, and making resistance dominant. The aggregation of ribo- somes into polysomes, however, coupled with the hypothesis that streptomycin may prevent the progression of the messenger by jamming the mechanism of ad- vancement, may explain the dominance of sensitivity. Under such a hypothesis, in fact, it would be enough if one ribosome in a polysome chain were of the sensitive type, to prevent the formation of protein by all the ribosomes of the chain. The residual activity would then be only 1 in 2”, if n is the number of elements in the polysome. The drastic reduction of the rate of protein synthesis thus determined might therefore explain the dominance of streptomycin sensitivity in cells heterozy- gous for resistance sensitivity. In its simplest form, the hypothesis would assert that streptomycin tends to displace messenger RNA from the ribosome thus perturbing its transcription and eventually jamming its passage. Some mutations, by altering ribosome structure, also disturb the messenger-ribosome complex, and because they perturb transcrip- tion, act as suppressors. According to this lemma, neighboring codons could in- fluence the extent of perturbations and allow some discrimination in the occurrence of transcription noise.? The mutation for streptomycin resistance (Sm’) modifies the ribosome in the opposite sense so as to increase its affinity for typical messengers to mitigate the perturbations resulting from the presence either of streptomycin or of certain suppressor mutations. The mitigation may, however, fail to cope with both disturbances simultaneously, and streptomycin-dependent suppression may result. Finally, without regard to suppressors, the altered ribosome may bind the messenger too tightly for normal function, in this case producing the phenotype of streptomycin dependence for growth. There is a close relationship between our studies and those of Gorini,* who found mutants for several amino acids, such that the requirements of one amino acid in a given mutant could be dispensed with by the addition of streptomycin. This ‘‘con- ditional streptomycin dependence” is quite analogous in that it seems that strepto- mycin restores the production of a specific enzyme, albeit a different one in each strain. Gorini suggests, on the basis of these results, that streptomycin acts by increasing the ambiguity of the code, thus leading to the synthesis of wrong protein. The argument of dominance of sensitivity mentioned above, which could now be tested directly in vitro, would specify that streptomycin jams the advancement of the messenger in the polysome chain, perhaps by linking it to ribosomal RNA. Summary.—An E. coli K-12 strain which has lost its capacity to produce galactose transferase carries a suppressor mutation (mapping not far from streptomycin re- sistance) which has partially restored the capacity to ferment. Some mutations to streptomycin resistance in this suppressed strain make galactose fermentation streptomycin-dependent. The implications of the similarity between suppressor genes and streptomycin drug action are discussed. It is possible that some sup- 682 GENETICS: LEDERBERG ET AL. Proc. N. A. 8. pressors act by altering the ribosome and that streptomycin acts by jamming the mechanism of advancement of the messenger, and it is suggested that this might explain the dominance of streptomycin sensitivity in heterozygotes. * This research was supported by grant (G-6411) from the National Science Foundation, and by grant (Al-5160-06) from the National Institutes of Health. ' Benzer, 8., and 8. P. Champe, these Proczeprves, 47, 1025 (1961). * Garen, A., and O. Siddiqi, these ProckEpines, 48, 1121 (1962). 3 Brody, 8., and C. Yanofsky, these Procrrpines, 50, 9 (1963). *Gorini, L., and E. Kataja, these Procerpinas, 51, 487 (1964). 5 Kurahashi, K., Science, 125, 114 (1957). ® Lederberg, E. M., Genetics, 40, 580 (1955). * Taylor, A. L., and E. A. Adelberg, Genetics, 45, 1233 (1960). § Spotts, C. R., and R. Y. Stanier, Nature, 183, 618 (1959); Speyer, J. F., P. Lengyel, and C. Basilio, these ProcEEpinas, 48, 684 (1962). ® Lederberg, J., J. Bacteriol., 61, 549 (1951). * Warner, J. R., P. F. Knopf, and A. Rich, these Procrepinas, 49, 122 (1963). "Lederberg, E. M., Genetics, 37, 469 (1952).