A NONHEBEDITABY, HOST-INDUCED VARIATION OF BACTERIAL VIRUSES' 8. E. Lmw AND WY L. HUMAN' Lbpartmant of Baccbriology, Uniwrity of Illinois, Urbana, Illinois .Baceived for publication April 16, 1952 One of virology's moat generally valid rules is that the properties of virus particlea are unaffected by the host in which they grow. Host adaptation and tissue adaptation, the apparent exceptions, ares explained today by selective reproduction of virus mutanta in new ho& or in new tissues. In analyzing the relation between certain phaga and certain mutants of their bacterial ho&s, we have encountered a novel &ration: the genotype of the host in which a virus reproduces affects the phenotype of the new virus. The phenotypic change sup presee.~ the ability of the virus to reproduce in certain hosta but not in others. It ie a transient change, in the sense that one cycle of 5owth in a &table host re- turns the virus to its original form. Both the 5owth ability of the moditkl vinu~ and, in some M, ita production from normal virus are controlled in part by the physiological &ate of the host cell. The pmnt paper describea these findinga and discussea their general implications. Other instancee of host-controlled pheno- typic changee in bacteriophages have recently been discovered (Bertani and Weigle, 1952). MA- AND MElmoD The phagear employed belong to the TLT7 @em active on Ewherkhia adi, strain B, and on Shiplla &mmt&ue, atrain Sh. Standard methods and media were employed (Adams, 1950). Microscopia obaervationa of phage-infected bac- teria were done at 39 C or 37 C on nutrient agar platea observed under a dry objective of N.A. 0.66. In the following sectiona, the expression "young cells" ia used to designate cella in the exponential growth phase in aerated nutrient broth; these generally were used when the viable count was about 101 per ml. "Old cells" refers to stationary phase cells in cultures continuously aerated for at least 18 hours (viable count about 4 X 1CP per ml). Ba&$ul mufuntu. About 199 independently arisen mutant straine B/4 were isolated at various times from aa many separate culturea of B grown from amall inocula and plated with an exceea of phage T4. After several restreakings, none 1 Thie work WM supported by granta-in-aid from the American Cancer Society (upon recommendation by the Committee on Growth, National Research Council) and from the Univemity Beeeerch Board of the Univereity of Illinois. The authors gratefully acknowl- edge the cooperation of Mica Martha R. Sheek in carrying out part of the experimenta. * Preeent addrurs: 3766 Honeyeuokle Avenue, Fort Worth, Texae. 657 558 8. E. LURIA AND MARY L. HUMAN [VOL. 04 of them was lysogenic for T4. As previously reported (Lurk, 1946), these B/4 mutants,8 when tested as hosts for T2 and T6, fall into three true breeding groups. Some strains (10 out of 94) are ss sensitive to T2 and T6 as strain B. Others (59 out of 94) give a somewhat low plate count with T2 and T6 (10 to 50 per cent of the count on B). The third group (25 out of 94) is the one whose relation with T2 and T6 is described in this paper. The strains of this group will be divided into two sub- groups, strain B/C and strain B/b; both groups give rise on nutrient agar to characteristic colonies, distinctly rougher than those of E. wli, strain B, and its other strain B/4 mutants. They are resistant to T3 and T7 and fully sensitive to Tl and T5. A rare, nonaerogenic, small colony mutant type (strain B/4,3,7, 1, tryptophan dependent) behaves like strain B/4,, with regard to T2 and T6. Some B/40 and B/k strains, after many subcultures, revert gradually either to the B/4 or to the B phenotype. This reversion has not yet been investigated. The behavior of strain B/40 will be described 6rst; then the distinguishing properties of strain B/h will be listed. The relatth of strain B/k with TB and T6. Phages T2 and T6 give no plaques when plated in moderate amounts with strain B/40. Plates seeded with at least 10) particles of T2 or T6 and up to 10 cells of strain B/40 give a partially lysed layer of growth or an almost complete lysis (with occasional resistant col- onies of strain B/40/2 or strain B/&/6). These observations suggested lysis of the initially plated bacteria without phage production. Tests in liquid medium revealed the following facts: 1. Strain B/40 cells adsorb T2 and T6 about as well as strain B cells. The adsorption rate constants per cell are 1 to 2 X Wn min+ for young cells, 1 to 5 X lO+ min-1 for old cells. 2. Lysis, as shown by microscopic observation of young cells of strain B/40 multiple-infected with T2 or T6, occurs at similar times as with strain B and is equally complete. In one experiment (no. 8-4), for example, 72 out of 30 bacteria lysed in 95 minutes. This observation explains the suppression of bacterial growth on strain B/h plates with an excess of phage T2 or T6. 3. Phage liberation by strain B/40 remained uncertain as long as plaques were sought using strain B or its mutants as indicators; the situation was clarified by the discovery that strain B/b liberates phage that forms plaques when plated with S. d~8e?&iae, strain Sh, while behaving abnormally toward E. wli, strain B. Normal T2 and T6 give equal counts on strain B and on strain Sh. Results from one step experiments of T2 and T6 on straiu B/4o with platings on strain Sh and on strain B (young cells) are given in table 1. In these experi- ments the unadsorbsd phage was removed by addition of antiphage serum before diluting the adsorption mixture; the only active phage was what came out of the infected cells of strain B/40. The results show that most and possibly all the young cslls of strain B/40 liberate phags that forms plaques on plates seeded with strain Sh but not on those with young cells of strain B. a T4-reeirtant mutsntr sre generally B/4,3,7, occeeionally B/4,3 (Demerea and Fano, 1945). TABLE 1 Tlb p~duciim of TY and TV in bacteria B/& mui B/& One rtep growth experimenb of T2 or Tfi on B/4, and B/4,,. Young or old oelle were mixed with phage. After 6 minute+ the mixturee were diluted into antiphage serum to stop adsorption. Four minuke later, when over 99 per cent of the phage that had remained free wae neu- tralised, further dilutions were made and eamplee were plated with young oelle of strain B and with rtrain Sh. All counts are given in plaquea per ml, referred to the adeorption mixturee. I 139 Baaterial strain.. . . . . . . . . . . . . . . . . . . . B/4,, atrain 21 Age of bacteria.. . . . . . . . . . . . . . . . . . . . Young Pblbge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T2 Phage input.. . . . . . . . . . . . . . . . . . . . . . . . 1 x 10' Number of infected bacteria that produce plaques on Sh.. . . . . . . . . . . 9.0 x 10' Number of infected baoteris that produce plaque0 on young cells Of B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 x10' Titer after lyeis, on strain Sh.. . . . . . 7.0 x 10' Titer after lyeir, on young aelle of B 2.5 X 10' Avemge yield per cell, T*. . . . . . . . . . . 8 Averageyieldpercell,T2 . . . . . . . . . . . - 139 S/4,, &rain 21 Old T2 1 x 10' 5.7 x 10' 1 x10' 5.0 x 10.t a x 10' 1t b/4,,, &rain 4 Young T2 5 x 10' 2.6 x 10' 2.5 X 10` 4.0 x 101 1.7 x 10` 16 )/4~, &rain 4 Old T2 6 x 10' 2.7 X 10' 2.3 X 10' 5.4 x 10' 6.1 x 10' - 2 138 I/4,, strain 21 Youllg T6 1.6 x 10' 1.1 x 10' 5 xw 1.6 x l(r 5 x 10' 14 - 139 b/m, rtrain 4 Young T6 3.0 x 10' 2.4 X 10' 2.5 X 10` 2.7 X 10' 1.3 x 10` 11 - t Phage liberation wae probably not completed at the time of plating. 560 8. R. Luau AND MABY L. EUKAN [VOL. 64 When old strain B/4( cells are infected with T2 or T6, only about half the cells produce some phage which is active on strain Sh but not on young cella of strain B. The average yield per young cell of strain B/4, measured on strain Sh, is 8 to 39 for T2 and T6; for old cells the yield is 1 to 4. Temperature and nature of the medium in which strain B/40 cells have been 5own do not affect the nature of the phage they produce. The typs of experiment shown in table 1 defines the modified form T+ of the phages T2 and T6 (T*2 and T*6, respectively) as one produced in strain B/40 and forming plaques on strain Sh but not on young cells of strain B. The proper- tiea of T* will be dsscribed later. The plague counts obtained in platings of infected strain B/40 bacteria on young cells of B (table l), although low, are significantly higher than can be ac- counted for by any serum-surviving free phage from the input. This might mean that a few strain B/40 cells (less than 1 in 1,966) liberate some normal phage particle. More probably it means that one out of many thousands T* particles plated hss an opportunity of producing a plaque on the B plates, possibly because on the plate it only becomes adsorbed late when some of the strain B cells have aged and become competent to support growth of T* (sss below). The ml&on of strains B/&J with T# and TB. Young celb of B/& strains infected with `I2 and T6 behave somewhat similarly to young oells of strain B/40 (table 1). About 97 to 99 per cent of the cells liberate only T* phage, active onstrainShbutnotonstrainB.About1to3percentoftheinfectedbacteria form plaques on young cells of B; the yield of phage that can form plaques on youngcellsofBisoftheorderofoneortwopercell. Old cells of strain B/400 infected with T2 or T6 liberats normal T2 or T6 in small amounta, similar to those produced by old cells of B (table 1). In mast experiments there is evidence for the production of some T* phage along with the T phage, as evidenced by higher counts on strain Sh than on young strain B cells (tables 1 and 4). The amounts of T* produced are variable, from less than 29 pss cent of the yield to about one-third of it. Thus, old cultures of strain B/~,w appear to contain some orgamsms resembling the young cells in their ability to produce T* from T phage; young culturss of B/4,,,,, on the other hand, contain a few cells that can liberate some T phage as such. The liberation of T* or T form of phags from B/400 cells in the phase of transition from old age to log phase and vice versa has not yet been analysed. Prop&&s of T*8 and TV. Qualitatively, T*2 and T*6 behave in similar ways. Since most experiments were done with T*2-the critical ones being con- firmed with T*6-we shall use the T2-T+2 system as a model. Phage T*2 is neutral&d by anti-T2 serum at the same rate as T2. It is as stable in sitro as T2. It is not converted to T2 in v&s by treatment with distilled water, which raised the titer of fresh T2 lysates by removing a phage inhibitor of bacterial origin (Bertani and Sagik, to be published). T*2 lysates can be pre- pared by infecting strain B/h with T2, removing most unadsorbed T2 by cen- trifugation, washing the infected cells before lysis, and allowing them to lyse in nutrient broth without added NaCl; readsorption of the liberated phage is 19521 HOST-INDUCED VABIATION OF BACTEBIAL VIRUSBS 561 thereby prevented. The best lysates gave counts about l,OW-fold higher on strain Sh than on young strain B plates. The count on Sh cells may be due partly to some reeidual T2 particles, partly to some T*2 particles that happen to infect bacteria late when the cells have become old on the plate (see item 4, below). 1. T*2 is adsorbed by strain Sh and lysea it, giving an apparently pure yield of normal T2. The adsorption of T*2 by strain Sh is similar to that of T2; the yield of T2 is the same whichever phage form is the infecting one (table 2). The liberation of T2 by bacteria infected with T*2 will be called "activation" of T*2. 2. T*2 is adsorbed by cells of strains B, B/4, B/6 (not by strain B/2). One particle per cell is sufkient to prevent colony formation by strain B cells. T*6 ia adsorbed by strain B/2, not strain B/6. TABLE 2 One stsp growth ezpurimenk with TV and Td on rtmin Sh Ezperimnt so. ISd. Young cella of strain Sh (10' celle per ml) were mixed with phage. After adsorption the free phage wae eliminated with antieerum and after dilution platin@ were made at intervale before and after lyeie. AU counte are given in plaquar per ml of the adeorption mixtures. Phage input I 4 x 10' I 1 x 10' Plaque count before lyeie, on young celle of B 2.2 x 10' i S.6 x 10' Plaque count beforelysie, on cellr of Sh 2.6 X 10' 6.9 x 10' Plaque count after lyeia, on young cells of B 2.8 X 10' 6.0 x 10' Plaque count after lysie, on celle of Sh 2.6 x 10' 6.0 x 10' Average yield of T2 per cell 116 105 3. If young cells of strain B that have adsorbed T*2 are plated with au excees of either strain B or strain Sh, about 99.9 per ceut of them produce no plaquea (no activation nor reproduction of T*2 as such; see table 3). Microscopic ob- servation at 39 C shows that the infected cella do not divide; many of them elongate, reaching lengths of 29 ~1 or more, then slowly disintegrate. 4. If old cella of strain B are infected with T*2 at 37 C, about 1 per cent of them liberatea T2 (equal cotmta on young strain B and on strain Sh plates; see table 3). The yield appears to consist of T2 phage only, without admixed T*2. The fraction of old strain B cells that activates T*2 is not greatly different, whether the cella have been infected in their old culture medium, or after washing and reauspending in buffer, or after washing in buffer and reauspending in fresh broth containiig T*, provided T* adsorption occurred promptly after contact with fresh medium. The remainin g cells liberate nothing. The fraction of old cells of strain B that activates T*2, whether in buffer or in the exhausted medium, is 562 8. E. LURIA AND MAItY L. HUMAN [VOL. ei4 higher when infection talus place at lower temperaturee-for example, 3 per cent at 25 C, 1 per cent at 37 C, 0.5 per cent at 43 C (experiment no. 123). 5. If old etrain B cella are resuspended in fresh aerated broth and preincubated before adding T*2, the fraction of infected cells that activates T*2 depend8 on the time and temperature of preincubation. A set of representative resulte ia shown in figure 1, a. Upon preincubation at 13 C little change occura in 2 hours. At 25 C the fraction of &rain B cells that activates T*2 increases at first, then slowly decreases. At 37 C there is no increase and the decrease ia more rapid. At 43 C the cells lose their activating ability in a few minutea and afterward be- have like young cells of strain B. +eincubation of old cella of &rain B in bufEer TABLE 3 Tad for phage liberalion by various Lypsr of calls inf6c&l with TY Ezpcw+neva~ no. f&3. A lyaate of TV wan mixed at 37 C with cells of variom atraiue, either young or old. After allowing 4.6 minutea for adeorption, the mixtures were treated with antiphage eerum, then diluted and plated. The concentration of old cells wee 6 X 10' per ml, that of young cell8 2 X 101 per ml. The phage lyeate had a plaque titer of 1.2 X 10' on strain Sh and 4.3 X 10' on young cells of &rain B. M DlmT . . . . . . . . . . . . . . . . . 4.3 x 1* 0.030 -~ @acterial atrain 8h 9.0 x 10' 7s Young B 7.2 x 10' 0.06+* Old B 6 x 10' 0.4 Young B/40 <4 x 10' - Old B/40 a x 10' - Young B/400 1.4 x 101 0.001 Old B/400 1.9 X 10' 1.6 o * These ratios are not dgnificautly different fr oells of B and on cella of 8h in the input phage (0.1 1.1 x 19 im I 7.6 x 10' 63 T2 7.2 X 10' 0.06*+ - 6 x 10` 0.6 T2 4 x 10' 0.03*+ - 6 x 10' 0.4 T.2 1.9 x 10` 0.16 T+2 2.6 X 10' 2.2 T2, T*2 m the ratio between counte on young 36). at varioue temperatures for up to 90 minutea doea not reduce their ability to activate T*2. 6. When young cella of etrain B/~w are infected with T02, over 99 per cent of the cells yield nothing, just aa young celle of &rain B (eee table 3). Specially de- signed experiment8 showed that of the few cells that liberate eomething, almost all yield only T*2, but a few yield Borne T2 just as if they had been infected with T2 (see table 4). If old cella of strain B/~oo are infected with T*2 at 37 C, about 1 to 3 per cent activate T02, aa old cells of &rain B would (table 3). The dependence of the frac- tion of cells of B/400 that activatea T*2 on the time and Wperature of prein- cubation in fresh medium ia qualitatively eimilar tc that obeerved with old atrain B (see figure 1, b). The old &rain B/400 cell often liberate Mime phage J I I I I I /o 20 Jo 40 50 min Figure 1. The proportion of old eella that liberata phage aftw infection at 25 C with Tq, aa a fun&ion of the time and temperature of incubation in fresh aerated nutrient broth before infection. Strain B (diagram a) liberater T2. Strain B/4@, (diagram b) liberated T2 with rome T+2. Strain B/4* (diagram c) &era- T??. TABLE 4 Comparboa offh6 proddh of TI and TY by Bj.#ro bactda, rtrain .J, infected with T# or T'# -Ilo.................................. Aon0~ucmu,mAuB/4Il................... mms........................................... Phage input (a) Plating8 before I*, on young cello of B (b) P1at~beforelyG,oncellaofSh Ratio W/W (c) Platings after lyair, on young cwlloofB (d) Plating &er lymia, on o&r of Sh Ratio (a)/(d) 14s Yocml n T1 1.4 x 10' 1.2 X 10' 1.0 X 1V 1.1 X l(r 8.0 X 100 9 X 10' 0.012 0.012 564 8. E. LTJMA AND MARY L. HUMAN [VOL. 64 T*2 along with the m:juet ae they do when infected with T2. The relative amouta of T*2 and T2 liberated by a given culture of strain B/400 are the same whether the cdla are infect& with T2 or T*2 (see table 4). Altogether, old B/400 cells resemble old B cells with regard tc ability to produce T2 from T2 and T2 from T*2. However, they aleo produce some T*2 from either T2 or T*2. 7. When young oella of strain B/40 are infected with T*2, at least 99.9 per cent yield nothing (see table 3). When old celle of strain B/40 are infected with T?2 a small fraction of them liberatea T3, the others liberate nothing (table 3). Age aud temperature relationa in the reproduction of T* by old &rain B/40 broadly resemble those iu the activation of T*2 by old strain B or strain B/4m (figure 1, cl- In eummary, T*2, produced in &rain B/40 or in strain B/400, ia &orbed by and kills the bacterial kaina susceptible to T2. It grows in every cell of &rain T2 + Shiga - T2 T*2+ Shiga -T 2 B young - T2 B young + - 0 old - T2 B old +T2 B/4. you ng - T"2 1 B/l0 old - T'2 BfiOb young v T'2 B&young + -. BM, old + T"2 BAO,, young + TX2 BA,, old + TZ+R I BAm old + T2+T*2 FQur~ d. Results of infeation of various bsctarial straina with phsges T2 and T# Sh and in a fraction of old B cells, giving a yield of normal T2. It growe in a fraction of old cells of strain B/40, giving a yield of T*2. It grows in a f&ion.of old B/400 cells, giving a yield that consista mainly of T2, but contains some T%. It grows in a very small fraction of young ce& of strain B/400, giving rk to production of Tq. The properties of T*6 are analogous, mutatia mutcmdizk, tc those of T*2. The r mutants behave like their parent phages. These facts are eum- marked in the scheme of figure 2. It is apparent that T*2 is distiied from T2 by the fact that it requires for growth a physiologically "competent" ho& Competence may involve ability to produce T2 from T% (m old strain B, in old &rain B/b, and in strain Sh) or to produce T02 from T+2 (in old strain B/h). It is important to note that whenever T*2 initiates phsge production in a given type of bacteria, the compositioe of the yield-obtained, if any, L exactly the same aa if the infecting phage had been T2. There ia no evidence of any influence of the 19521 HOS'CINDUtDD VABIATION OF BA- VmusBs 565 T or T* quality of the infecting phage on the quality of the progeny phage, which appeara to be exclusively host controlled. Interad~ of TY in mixed inject&m. 1. Genetic mombinution. Mixed infection of young celh of strain B with equal amounta of T*2 and T2hr gives a yield con- taining T2, T2r, T2h, and T2hr (table 5). The recombinant typea are more numerous than the parental type resembling the T*2 parent. The number of bac- teria that liberates phage with genetic chara&ristica of the T* parent ia sig- nificantly higher than the number of bacteria infected with any residual T2 particlea in the T*2 lyeate, but ia leea than one-tenth the number of bacteria in- fected with both T+2 and T2hr. Thus, the T* parent behavea in mixed infection TABLE 6 &wtic recombination bet-n T+I and T#hr Expurimmt no. 108. Equal amounta of T*2 and T2hr ruflicient to give multiple infeotion (g phagee of each type per oell) weremixed reparately or together with young cello of strain B in nutrient broth. After 6 minuter for adsorption and 4 minuter of antiserum treatment, the mixtures were diluted and assayed at intervals for plaque counts of T2, T2h, T2r, T2hr, either on young etrain B or on a mixture of young oeUr of B and young celle of B/2. Phage Platings before lyeis; plaque aount on &rain B Plating6 after lyeis, dilution 1:40; plaque oount on rtraine B + B/2 1 T+2 1.6 rt cIILI[N NO. 1 T2llr 709 r 546hr 3 T*2 + T%r 521 r 40 mottled (mixed r and rt) 495 lu 22 htr 16 hrt 2 Wrt The count in it&a wan obtained by plating a more concentrated aample and correoting for the difference in aoncentration. as a minority parent, aa though only a fraction of the T+ par&lee could contrib- ute to recombination in mixed infected bacteria. 2. Mutual ezclh. Mixed infection of young &rain B with T*2 and T6, with T'%? preceding T6 by 5 minutes, reducea the number of cells that liberatea T0, indicating that T* can to come extent exclude a related heterologoua phage. The reduction is leea than expecti if every adsorbed particle of T*2 could exclude T6. The partid an&vi& of strain B/&O to Td. Most B/4 &rains, including thoee of the strain B/6 group, are fully re&tant to T4, which they fail to adsorb, at lea& in the irreversible etep (Garen and Puck, 1951). Straina B/b, however, show an age dependence in their Astance to T4 (two etraine tested). Most old cells, plated on nutrient agar coated with an exceaa of T4, fail to form colonies; about 2 to 20 per cent 88 many colonies are formed on phaged platea 88 on non- phaged ones. The old cells adsorb T4 in liquid, but liberate none. Microscopic 566 8. E. LUBLI AND MABY L. HUMAN [VOL. 04 observation shows that most cells fail to increase in size and to divide; many of them slowly fade away. If old strain B/b cells are incubated in fresh aerated broth at 37 C and samples are plated at intervals on plates with and without T4 (see Qure 3), the proportion of colonies formed on T4 increasea slowly. A com- parison of figures 3 and l,b shows that the process of "rejuvenation" that sup- presses the susceptibility of strain B/$0 to T4 is much slower than the one that suppresses the ability to activate T*2. Preliminary experiments indicate, on the other hand, that the ability of old strain B/h to produce T2 from T2 is lost dur- ing rejuvenation at about the same rate as the susceptibility to killing by T4. The few colonies that appear after plating old strain B/b cells on T4 probably do not stem from genetically different cells; in fact, subcultures from these colo- nies give the same ratio of colony counts on plates with and without T4 as the FQure 3. Change-e in the proportion of strain B/408 celle killed by T4 aa a function of the time of incubation in freeh aerated nutrient broth at 37 C before plating for aolony oount. Platb with T4 were done by mixing the bacterial samplee with T4 (2 X 10" particle6 per ml) and plating 0.1 ml aliquota. c 100 original platings. The colonies formed on T4 plates may represent oells that, by chance, fail to meet T4 until they are rejuvenated. Attempta to act&x& free T*. Although the findings listed before indicate that activation of T* results from adsorption and growth in a competent bacterium, a variety of tests wsa done in an effort to eliminate the possibility that activation of T* in a mixture with bacteria be due, not to adsorption, but to the action of some soluble product of bacterial cells. All tests were negative. Filtrates and supernatants of strain Sh or of old strain B cultures were ine8ective under all conditions and temperatures tested. Suspensions of washed cells of strain Sh or old strain cultures of B, resuspended in buffer, were about as effective in activat- ing T* as the whole culture from which they came. Activation only occurred if T+ was mixed with cells under conditions where adsorption was possible. Also, the fact that activation fails with resistant bacteria (strain B/2 for T+2; strain B/6 for T%) would require unwarranted assumptior~ to be reconciled with the idea of extracellular activation. 19523 HOST-INDUCED VARIATION OF BACTERIAL VIBUSES 507 DIBCUEWON Product&m of the T* form of phage. The T* form of phages T2 and T6 is pro- duced only in some bacterial mutants (strains B/40 and B/4& resistant to T4, T3, and T7. The relation between the resistance to these phages and the modify- ing influence on T2 and T6 is unknown, as is also the relation between acquisition of resistance to certain phagea and changes in the nutritional requirements (Anderson, 1946; Wohman, 1947) or in the growth rate of bacteria (Luria, 1946). A physiological connection between msistance to T4 and modifying action on T2 and T6 is suggested by the fact that in the B/~oo strains the old cells are both capable of reproducing some T2 without modifying it and susceptible to killing by T4. These two properties of old cells disappear at similar rates during their rejuvenation in fresh medium. The remarkable fact is the role played by the host's heredity and by its physio- logical state in dete rmining the properties of the virus it liberates. This evidences a hitherto unsuspected plasticity of virus properties at the phenotypic level; it suggests that virus physiology, se expressed in the various phases of interaction with the host cell, may be controlled in part by nongenetic, "cytoplasmic" com- ponents of the virus particles. The existence of physiologically active, nongenetic components in some bacteriophages has become increasingly probable because of a number of recent 6ndings (Benser, 1952; Dulbecco, 1952; Luria, 1952; Hershey and Chase, 1952). Control by the genetic and developmental make-up of the host cells over the future reproductive ability of virus particles liberated from them could play a tremendous role in determining the course of virus infections, both at the epi- demiological and the pathological level. We may mention that some of the findings on the system T2-T2 provide an analogy for the cases of "masking" in some virus diseases of animals, for example, in rabbit papilloma where infectious virus with capacity for indefinite tr ansmisGon cannot be isolated from the tumors of the domestic rabbit, but only from the cottontail rabbit (Shope, 1950). The T+2 form produced in strains B/40 or B/400 cannot be detected with actively metabolising cells of strains B or B/4; only the availability of strain Sh per- mitted its identification. The role of the physiological state of strain B/h in determining the alternative T*2 or T2 phenotypea suggests by analogy the interesting possibility that a de- velopmentally specialized animal tissue may so modify a virus particle as to ren- der it incapable of reproduction in that tissue itself or in some other tissue, thus providing for an active rather than a passive determination of virus tropisms by the host. Speculative as these analogies are, we may gain from an awareness of their implications. The possible occurrence of host-induced restrictions in the host range of viruses should be kept in mind, for example, in considering the role of viruses in the etiology of "nonvirus" tumors. Growth phase control of the tirus-pr~d~ ability of bacteria. The T* form of phage, when infecting E. eoZi host cells, fails to multiply in young, actively metabolising cells but multiplies in an appreciable fraction of old, starved bac- teria. Considering the scheme of figure 2 we see that in "competent" cells, where it can reproduce, T+2 gives rise to a progeny identical to that which T2 would 568 8. E. LUBIA AND MABY L. HUMAN [VOL. 64 give in the same host. The quality, T or T*, of the phage that is produced appears to depend only on the host, not on the quality of the infecting phage. These phage changes clearly are not heritable since there is no cell type in which the composition of the yield is determined by the infecting phage. Guhlin (1948) described the lysis of several E. wli strains by phage C 16 with- out production of phage; C 16 reproduced normally in several dysentery organ- isms. It is possible that C 16, a phage related to T2 and T6 (Adams, 1952), gave riss in E. wli to a progeny analogous to T*, but not revealed by any of the host organisms tested in G&n's work. In view of recent evidence for a stagewise process of phage maturation (Luria, 1952), it is tempting to visualise T+ aa a somewhat imperfect form of the phage, resulting possibly from an inadequate supply of some essential component in the strain B/~o host. The defect would manifest iteelf as an inability to perform suc- cessfully some step needed for production or liberation of active phage in young cells of E. co& strain B, and of its mutants. In the competent cells of strain Sh and of old strain B cultures, either the defective phage activity may not be needed or its performance may only be required at an efilciency level inadequate for success in the young cella of strain B. Although we do not know at what state the process of phage production is stopped in young cells of E. wli infected with T*, the low recombination and exclusion e6iciencies of T+2 suggest that the block does not concern simply the procws of lysis and phage release. The data presented in this paper on the influence of the physiological state of strain B cells on the growth of T* do not permit a choice among a number of pos- sible hypotheses as to what phases of host metabolism are involved in the transi- tion of strain B cells from competence to incompetence and vice versa. The pro- portion of competent cells is affected by time and temperature of preincubation of old cells in fresh nutrient medium, but not in buffer. Evidently an active metabolism is required for the transition from competence to incompetence as for the concomitant transition from resting phase to active growth phsse. The complex effects of the time and temperature of preincubation on the proportion of competent cells suggest an interplay of various reactions with d&rent tem- perature dependencee. S-Y Several B/4 mutants of Eacherkh wZi, strain B, when infected with phages T2 or T6, liberate these phrrgea in a form de&rated as T*, which does not multi- ply in young cells of strain B or of its mutants. T* can multiply in a small propor- tion of old, starved cells of strain B, giving rise to a yield of the corresponding normalTphage.T*canbetransmittedseriallyassuchinasmallfractionof the old cells of some B/4 mutants. In r9higel24 dysataicrc, strain Sh, the T* partioles behave a~ normal T particles; infection of a strain Sh cell with a T* particles gives afullyieldofnormalTphage.~~findinge~~snssyetuneuapec~eue- ceptibility of viruses to transitory physiological changes induced by the host in which they have grown. 19521 HOST-INDUCED VAIUATION OF BACMCBIAL VIRUSES 569 REFERENCES ADAMB, M. H. 1950 Methode of rtudy of bacterial vinuae, pp. l-73. In &f&h& in mdicoZ rucurch. Vol. II. Year Book Puhliehere, Chicago, Ill. ADAUB, M. H. 1962 Claeeification of bacterial viruee~: characterietiw of the TS speciw and of the T2, Cl0 npeciea. J. Bact., U, 381-390. hDBBSON, E. H. lQ43 Growth requirements of virue-reeiatant mutanb of Emherichia coli &rain "B". Proo. Natl. Acad. Sci., 32,120-123. BINZEP, 8. 1962 Reeirtrrnae to ultraviolet light se an index to the reproduction of bacte- riophage. J. Bact., 0,3Q-72. BEBTANI, G., AND WEIOLI, J. 1962 Host-controlled variation in bacterial virueee. J. Bact., in prrrr. DzM~, M., AND FANO, U. lQ46 Baoteriophage-reel&ant mutanta in Ederichio coZi. Genetiw, a, IlQ-130. DULBEC