Temperate Coliphages: Classification and Correlation with

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 1980, p /80/ /08$02.00/0 Vol. 39, No. 5 Temperate Coliphages: Classification and Correlation with Habitats ELVERA K. S. DHILLON,' T. S. DHILLON,2* Y. Y. LAM,2 AND ALFRED H. C. TSANG' Department of Biology, The Chinese University of Hong Kong,1 and Department of Botany, University of Hong Kong,2 Hong Kong Temperate coliphages were recovered from sewage, mammalian feces, and lysogenic strains of Escherichia coli. A total of 32 phages of independent origin were divided into six groups by applying the criteria of host range, antigenic homology, and the ultraviolet inducibility of the prophage. The demonstration of genetic interactions in some cases has confirmed the classification scheme. Nine phages were assigned to the P2 family and 19 to the lambda family. The remaining four isolates may represent some novel phylogenetic types. Phages recovered from the lysogenic strains of E. coli were all found to be P2 related, whereas a majority of the phages recovered as cell-free plaque-forming units were assignable to the lambda family. It is proposed that the biological attributes of the phages belonging to the two principal families are reflected in the distribution patterns observed. The virions of phage HK256 show multiple tail fibers and may thus represent a "new" virion form among the temperate coliphages. We have been interested in various aspects of the natural history of coliphages in general and the temperate coliphages in particular. In earlier communications it was shown that the temperate coliphages comprise a very small proportion of the total plaque-forming units (PFU) found in natural habitats (5, 6, 8). Over a period of time, we have accumulated a sizable number of temperate phages isolated from diverse habitats. In this communication we have attempted a classification of these phages by the criteria of host range and antigenic homology. The validity of the taxonomic groups has been established in some instances by genetic interactions between members assigned to a group. In conclusion, an attempt has been made to explain the relative abundance of certain types of phages in particular habitats. MATERIALS AND METHODS Media. Tryptone broth agar was used for phage assays and for the isolation of phage-resistant mutants. Bacterial dilutions were made in phosphate-saline, and bacteriophages were diluted either in tryptone broth or in phosphate-saline containing 10% tryptone broth. Composition of these media has been described previously (5, 7, 8). Bacterial and bacteriophage strains. Some of the more important and basic bacterial strains are described in Table 1. Derivatives of these strains harboring the prophages, or some that were selected as phage-resistant mutants, will be mentioned in the text and are not listed in Table 1. Table 1 also lists the standard coliphages, such as Ti, T5, A, P2, P4, and 080, and their mutants. The locally isolated temperate phages will be identified by 1046 the prefix HK. The origins of the HK series of phages are given in Table 2. Methods. Methods routinely employed for phage assay, preparations of indicator cultures, "spot tests" for study of host range, and the preparation of antiphage sera have been described previously (6, 8). Methods adopted for the isolation of clear-plaque mutants of phages, isolation of single and double lysogens, and the ultraviolet (UV) induction of lysogens have been described elsewhere (7). A prophage was classified as UV inducible if the irradiated culture showed a 10-fold or higher titer than the control culture. The rationale and the methods for determining prophage immunity types and for the study of superinfection immunity have been described previously (5, 8). All tests were carried out using Escherichia coli C strain Cla and its derivatives, unless specified otherwise. lambda-resistant bacterial mutants were selected by challenge with phage HK243r-1 (8) on MacConkey medium containing maltose. The tona mutants were selected by challenge with phage Ti or T5 or HK022 vir, and the tonb mutants were selected for simultaneous resistance to phage TI and colicin B (10). Identification of P4 helper phage. Phage P4 can grow only on bacteria lysogenic for P2 or a related phage (4; E. Six, Bacteriol. Proc., p. 138, 1963). Lysates of P4 were, therefore, prepared on bacterial strain CIa (P2). The frequency of P2 PFU in such lysates was l0-5 per P4 PFU. Samples of P4 lysates containing 102 to 104 PFU were plated on lysogens harboring one or the other of the locally isolated phages. Lysogens that registered the P4 plaques under these conditions were concluded to harbor a P4 helper prophage. RESULTS Comments on classification. Initially, all the local isolates were classified by their host

2 VOL. 39, 1980 TEMPERATE COLIPHAGE HABITATS TABLE 1. Bacterial and bacteriophage strains Strain designation Relevant properties Source (reference) Bacterial strainsa Bb Prototrophic; "B" in the text A. D. Hershey Clab Prototrophic; "C" in the text G. Bertani (2) oogb Su', prototrophic, K-12; "K" in the text R. Herriott (6) FL104 Su- derivative of 009 (7) ST107 E. coli C; does not adsorb phage P2; strain C-2082 G. Bertani (personal of the Bertani collection communication) ST51 Col B+ strain; K-12 S. Kondo (10) Bacteriophages Ti and T5 Members of the T-series A. D. Hershey Apapa Wild-type lambda phage H. Echols A ci857 Temperature-sensitive mutant of the repressor pro- H. Echols tein gene; lysogens do not grow at 41 C A vir Virulent mutant; can form plaques on bacteria lyso- H. Echols genic for lambda prophage P2 Wild-type P2 phage G. Bertani (3) P2 vir Can form plaques on P2 lysogens P4 Helper-dependent phage M. G. Sunshine' 4)80 Wild-type )80 A. Matsushiro (1) 480 vir Virulent mutant of 480; can form plaques on 480 A. Matsushiro (perlysogens sonal communication) 4)80 sus sus23 (gene 10, tail formation), sus37 (gene 14, A. Matsushiro (12) early function), and sus622 (gene 1, head formation) HK243 and Adsorbs to A-receptor sites of E. coli (8; unpublished data) HK243r-1 a All are E. coli species. bthese strains will be referred to as the "primary strains" in the text. 'Source: Six, Bacteriol. Proc., p. 138, range. As the next step, antisera were prepared against a number of phages (the immunizing phage), and the inactivation rate constant (K) values of the sera were measured using the respective immunizing phages. These antisera were then tested against phage isolates that had not been used for immunization. Whenever a tested phage yielded a K value of l/io or more of the K value of the immunizing phage, the two were concluded to be antigenically homologous. The antigenic homologs were given a common serotype designation, usually the designation of the immunizing phage. In some instances partial antigenic homology was ignored on the strength of other criteria, and the phages in question were accorded different serotype designations. Such exceptional cases will be dealt with in the subsequent text. By applying the criteria of host range and antigenic homology, 32 local isolates were classified into six groups. Table 3 shows the more important characteristics of these six groups. The three standard temperate coliphages, namely, A, P2, and 480, are also listed in Table 3 to indicate the affinities of the local isolates TABLE 2. Sources of locally isolated temperate bacteriophagesa Source Bacteriophages Sewage HK022, HK104, HK105, HK106, HK542, HK543, HK544 Duck pond HK238 Cow dung... HK239, HK240 Pig dung HK244, HK245, HK246, HK247 Human feces HK241, HK248, HK249, HK250, HK251, HK252, HK253, HK254, HK255, HK256, HK257 Lysogenic E. coli strainsb HK107, HK108, HK109, HKl1O, HK111, HK113, HK114 areferences 5 and 8 and earlier publications referred to in reference 8 describe methods of isolation. 'The E. coli strains were isolated from clinical specimens. Of 110 E. coli strains tested on E. coli C, 20 were found to be lysogenic. The seven selected isolates were those that could be easily amplified on the bacterial strain Cla; the other 13 showed extremely poor propagation on E. coli C, B, or K and have, therefore, not been studied. with these three. Additional characteristics of the six groups of phages and the genetic tests indicating phylogenetic homology between members of the various groups are presented below.

3 1048 DHILLON ET AL. APPL. ENVIRON. MICROBIOL. TABLE 3. Classification of temperate bacteriophages of local origin Properties' Group Sensitive Archetype Prophage immunity types" primary Resistant mutants Serotype strains I C Not obtainable Unidentified but HKllO [HKllO] unique II C Unidentified Unique (?) HK253 [HK250] [HK253] III C, B, K tona Unique HK256 [HK256] IV C, B, K' P2 resistant P2 P2 [P2, HK240, HK241] [HKI09, HKI13, HK239] [HKI07, HKI08, HK111, HK114] V C, B. K. tona and tonb 480/HK [480, HK252] [HK254] VI C, B. K tona HK022 HK022 [X, HK106, HK244, HK542, HK544] [HK105, HK245, HK255] [HKI04, HK248] [HK238, HK247] [HK022] [HK246] [HK249] [HK251] [HK257] [HK543] "All phages except those of group IV were UV inducible. b Homoimmune isolates are bracketed together. Isolates appearing in separate brackets are heteroimmune. 'See the text for variability of this parameter. Group I. No bacterial mutants resistant to phage HK110 were obtainable, and this phage can grow on all the phage-resistant mutants of our collection. Thus, it must possess a unique host range. The antigenic uniqueness of this phage is indicated by its insensitivity to the following antisera: anti-lambda, anti-hk022, anti-hk239 (a P2-related phage; see below), anti-hk253, anti-mul, and anti-hk256. Group II. The two phages of this group are differentiable from one another by their heteroimmune nature and by their plaque morphology; plaques of phage HK250 are approximately 3 mm in diameter, and those of HK253 are less than 1 mm in diameter. Bacterial mutants selected for resistance to the clear-plaque mutants of HK253 are also resistant to HK250. However, such mutants remain sensitive to all the other phages listed in Tables 1 and 2. Both these phages are insensitive to the following antisera: anti-lambda, anti-hk022, anti-hk239, anti- MU1, and anti-hk256. However, the reciprocal tests indicate that HK253 may have some antigen(s) in common with phages of group IV. Anti- HK253 serum when tested with phage HK253 gave a K value of 80. The same antiserum inactivated group IV phages, giving K values for HK109 (K = 26), HK239 (K = 25), P2 (K = 20), and HK111 (K = 7). Inactivation of phage Pl vir was also observed, and a K value of 7 was obtained. Phage P4 does not form plaques on the lysogens harboring either prophage HK250 or HK253. This property and their inducibility by UV are the principal reasons for not assigning them to group IV. UV-induced lysates and the lytic lysates of clear-plaque mutants of these two phages failed to transduce either trp or his mutants. Group Ill. Phage HK256, which is the sole member of this group, resembles group VI phages in its host range but is unique in other respects. When plated at different temperatures, it gave an efficiency of plating (EOP) of 1.0 at 41 C, 0.6 at 37 C, and 0.1 at 32 C. In addition to its partially thermophilic nature, this phage also shows some other anomalous properties. The plaques of this phage show only slight bacterial growth. However, the cells recovered from the plaques are invariably lysogenic. Five successive single-colony isolations of the lysogens failed to yield any nonlysogenic derivatives. The lysogenic state, therefore, appears to be moderately stable. However, spontaneous curing of the lysogens has been observed in stab-cultures stored at room temperature for about 2 years. This observation suggested that the apparent instability of the prophage might be due to its existence as a plasmid (13). An attempt was therefore made to see whether the lysogens can be cured of the prophage by acridine orange treatment. No prophage curing was observed when lysogenic cells were grown for at least eight generations in ph 7.6 broth containing 35,ug of acridine orange per ml. Release of cell-free PFU by cultures of lysogenic bacteria seems to depend upon the host cell genetic background. Spontaneous release of cell-free PFU of triplicate cultures of B, C, and K lysogens was determined. E. coli C lysogens were found to contain 107 to 5 x 107 cell-free PFU per ml, whereas the density of such PFU in B and K lysogens was invariably found to be less than 2 x 102 per ml. Both the induced lysates and the lytic lysates of the wild-type phage and of a clear-plaque mutant failed to transduce either the trp or the his genetic markers.

4 VOL. 39, 1980 Figure 1 shows electron micrographs of HK256. A lamboid morphology may be concluded from isometric heads and flexible tails. However, most tails can be seen to terminate in a tangled mass of more than one fibrous appendage, and in this respect, HK256 represents a novel morphological type among the temperate coliphages. TEMPERATE COLIPHAGE HABITATS 1049 Group IV. The nine local isolates of this group can be unambiguously assigned to the P2 family of temperate coliphages. Besides the properties listed in Table 3, all phages of this group resemble P2 in their ability to act as P4 helpers. The homoimmune nature of P2, HK240, and HK241 was confirmed by the ability of a virulent A FIG. 1. Electron micrographs ofphage HK256 after staining with 1% uranyl acetate. Bars, 100 nm.

5 1050 DHILLON ET AL. mutant of phage P2 to form plaques on single lysogens harboring prophage P2, HK240, or HK241. The EOP of P2 vir on the single lysogens varied from 0.2 to 0.6. Virulent mutants of phages HK109 and HK239 were able to form plaques on single lysogens harboring either prophage HK109, HK113, or HK239, thus confirming their homoimmune relationship. The remaining four local isolates seem to be homoimmune, but this inference has not been confirmable due to the inavailability of the appropriate virulent mutants. Existence of further intragroup variability is indicated by the inability of some of the group IV phages to grow on one or more of the primary bacterial strains. Phages HK109, HK240, and HK241 have not been seen to form plaques on either the K or the B strain of E. coli (EOP less than 10-7). The basis of this inability to grow on K and B strains has not been determined. Group V. A very close kinship between the lambdoid phage 480 and the local isolates HK252 and HK254 is evident from the data presented in Table 3 and the following additional observations. Phage 480 is naturally temperature sensitive (1). We have found HK252 and HK254 to be also temperature sensitive (EOP at 41 C less than 10-8 of that at 34 C). Anti-HK252 serum gave a K value of 50 when the immunizing phage was tested. The same antiserum inactivated 480, giving a K value of 20, and HK254, giving a K value of 30. No detectable inactivation of phage A or HK022 (group VI, see below) by the use of anti-hk252 serum was observed. The homoimmune relationship of 480 and HK252 was confirmed by the use of a virulent mutant of 480. The 480 vir phage plated equally efficiently on the nonlysogenic bacteria and on the single lysogens harboring either prophage 480 or HK252. In spite of the close phylogenetic kinship indicated by their homoimmune relationship, two criteria have permitted differentiation between 480 and HK252. On strain B, 480 produces pinpoint plaques less than 1.0 mm in diameter, whereas plaques of HK252 are 3 to 4 mm in diameter. Prophage 480 appears to exclude phage HK022; the EOP of HK022 on Cla and Cla(4)80) is 1.0 and 10-7, respectively. In contrast, phage HK022 plates equally well on Cla and Cla(HK252). A close phylogenetic kinship between 480 and the heteroimmune phage HK254 was demonstrable by the use of three sus mutants of 480. All three sus mutants were found to plate at comparable efficiencies on the Su' and Su-(HK254) strains, whereas the EOP on the Su- nonlysogenic strain varied from i0-5 to APPL. ENVIRON. MICROBIOL. These results suggest either the occurrence of genetic recombination between the two phages or the transactivation of sus+ prophage alleles by the superinfecting sus particles (17). In either case, a close phylogenetic affinity between the two phages is proved. Group VI. Of 32 local isolates, 17 were assignable to this group. By the criteria employed, phage lambda is not a member of this group; its inclusion in the last column of Table 3 is explained later in the text. Anti-HK022 serum gave a K value of 68 for the immunizing phage. The same antiserum when tested by using group V phages gave K values of 21 for 480, 8 for HK254, and only 2 for HK252. Thus phage HK022 appears to bear some antigenic homology to phages of group V although, as mentioned earlier, in the reciprocal test anti-hk252 serum did not show any detectable inactivation of phage HK022. All local isolates of group VI when tested with anti-hk022 serum gave K values of 31 or more. Phage lambda showed no detectable inactivation by anti-hk022 serum. Group VI phages plate equally well at temperatures ranging from 32 to 41 C and are thus differentiable from group V phages in being temperature insensitive. The 17 phages of group VI encompass at least 10 different types of prophage immunity specificities (Table 3). Even though phages of this group bear no resemblance to phage lambda either in host range or in antigenic homology, four of the local isolates were found to be homoimmune to lambda. The availability of well-characterized mutants of phage lambda has permitted us to confirm the homoimmune relationship of A and the four local isolates. The virulent mutant of lambda was plated on single lysogens harboring either of the prophages A, HK106, HK244, HK542, or HK544. The A vir phage plated at comparable EOPs on the five single lysogens and an isogenic nonlysogen. In another series of tests, double lysogens harboring the ci857 mutant of lambda and one or the other of the local isolates were sought at 41 C. Since single lysogens of A ci857 show the lethal phenotype at 41 C, the emergence of double lysogens harboring A ci857 and a c+ phage would show that the latter phage can furnish repressor protein molecules for the maintenance of prophage A ci857. For unknown reasons, double lysogens harboring A ci857 and HK542 or HK544 have not been obtainable. Double lysogens harboring ci857 and HK106 or HK244 were obtained, and some of their properties are described below. Prophage HK106 is able to furnish the func-

6 VOL. 39, 1980 tion required for the maintenance of prophage A ci857 at 410C (Table 4), thus confirmning the homoimmune relationship between these two phages. The double lysogen harboring prophages A ci857 and HK244 was investigated in an identical manner. Thirty colonies for each of the four different types of growth temperatures given in Table 4 were tested. No loss of either prophage was observed. UV-induced lysates of the doubly lysogenic strains produced both parental and recombinant genotype progeny. For example, in the lysate obtained from strain A ci857(hk106), 7% of the progeny produced turbid plaques on the tona indicator (genotype c+h'), and 2% ofthe progeny produced clear plaques on the lambda-resistant indicator (genotype cih'"). Phage A does not form plaques on lysogenic bacteria harboring prophage HK022 (unpublished data). To determine whether the A-homoimmune local isolates were also excluded by prophage HK022, lysates of X and the four homoimmune local phages were plated on strain Cla(HK022). The EOPs of A, HK244, HK542, and HK544 were found to be less than 106, whereas that of HK106 was found to be 1.0. Lysates of 12 of the group VI phages were prepared on B, C, and K strains, and their EOPs were determined on strains B, C, and K. All 12 phages were found to be restricted by strain B (EOP on B relative to C varied from 6 x 10-7 to 2 x 10-4), and 9 of them were also restricted by strain K (EOP on K relative to C varied from 10-4 to 8 x 10-3). Phage HK244 was found to be insensitive to the K type of restriction and could thus be distinguished from other lambda-ho- TABLE 4. Incubation temp (IC) ofa: Stability of the doubly lysogenic bacteria at 33 and 41 C No. of colonies with prophage: Broth Agar ci857b HK106 Both culture plate a Nutrient broth was inoculated with fewer than 105 colony-forming units per ml, and duplicate cultures in parallel were static incubated overnight at 330C and at 410C. The overnight cultures were plated on nutrient agar plates, and duplicate plates were incubated at 33 and 410C. b The singly lysogenic derivatives listed in this column were tested, and all four were found to be incapable of growth at 410C. TEMPERATE COLIPHAGE HABITATS 1051 moimmune phages. Similarly, HK247 could be distinguished from its homoimmune relative HK238 by virtue of its insensitivity to the K- type of host cell restriction. Absence of Rex' phenotype in group V and VI phages. Five different T4rII deletion mutants were tested, and all were found to plate efficiently on the K-12 lysogens harboring prophages of the local isolates assigned to groups V and VI. Thus, none of these locally isolated phages, including those that are homoimmune to lambda, showed the Rex' phenotype, which is a distinctive trait of prophage lambda. Incidence of different phages in diverse habitats. Table 5 shows the incidence of phage groups in diverse habitats. It is clear that phages recovered from the naturally occurring lysogenic strains of E. coli are predominantly members of the P2 family (group IV). As cell-free PFU, P2- related phages have been recovered from cow dung, but they appear to be rare in human feces and in pig dung. A total of 25 of the phage isolates were obtained after chloroform treatment of the habitat samples. Of these, 17 (68%) were assignable to group VI. In the present collection, all the phages recovered from sewage and from pig dung were found to belong to group VI. Human fecal matter has yielded the greatest diversity of phages; 11 isolates of fetal origin contained at least one member assignable to five of the six groups. DISCUSSION On the basis of the experimental evidence presented, the two local isolates of group V have been referred to the lambda family. The four isolates of group VI that are homoimmune to lambda are clearly closely related to lambda and may logically be referred to the lambda family. Thus, six isolates of our collection are unambiguously lambdoid in nature. Phage HK022, which has been chosen as the archetype TABLE 5. Incidence ofphage groups in various habitats Incidence in habitats: Phage group Pro- Sewa Cow Pig Human phage dung' dung' feces' I 1 II 2 III 1 IV V 2 VIb a Samples of these habitats were treated with chloroform before plating. 'One isolate of this group (HK238) was recovered from a sample of duck pond.

7 1052 DHILLON ET AL. of group VI, does not recombine either with X or with )80. However, when crossed with imm080 ha recombinant phage, it did yield some recombinant progeny (unpublished data), and therefore HK022 can also be referred to the lambda family. Direct genetic evidence for the phylogenetic affinities of the remaining 12 phages of group VI is not available. However, considering the very close serological and host range similarities between these phages and HK022, we are inclined to refer them to the lambda family as well. Consequently, we find that 19 out of 32 local isolates (59%) are lambdoid phages. In contrast, only 28% are P2 related, and the remaining 13% may represent some novel phylogenetic types. In an unpublished study, G. Bertani and his associates analyzed temperate phages recovered from lysogenic bacteria obtained from clinical samples at the Los Angeles County Hospital. Of 28 phages, 16 were found to be P2 related, and 12 were lambda related (personal communication). Their results agree with ours with respect to an excess of P2-related phages. However, they are drastically different from ours with respect to our failure to recover lambda-related prophages. Conceivably, the conflicting results are in some way related to different ecological forces operative in the two geographical localities. Assuming that our study has not been seriously biased by the operation of selectional artifacts, we conclude that in their natural habitats, P2- related phages exist principally as prophages whereas the lambdoid phages exist mostly as cell-free PFU. In most studies, including ours, that have involved the isolation and characterization of temperate coliphages from natural habitats, a significant majority of the isolates have been unambiguously assignable either to the lambda family or to the P2 family. Disregarding the less frequently encountered phages, we propose to consider the biological properties of the lambdoid and P2 types of phage, which may be of ecological relevance. Members of the two families, in general, show contrasting attributes with regard to UV inducibility, overall rate of genetic recombination, number of attachment sites on the host chromosome, stability of virions under laboratory conditions and perhaps even in the natural habitats, possession of genes controlling general recombination, and the conferral of selectively advantageous phenotype on the lysogens. From the distribution pattern of the lambdoid phages it would seem that under natural conditions these phages are preferentially propagated by successive cycles of lytic infection of sensitive cells. In contrast, the distribution pattern of the P2-type phages suggests that these APPL. ENVIRON. MICROBIOL. viruses may be primarily propagating as the prophages. Known biological characteristics of the members of the two phage families seem to be in accord with the "habitual" modes of propagation suggested by their distribution patterns. For example, a lytic mode of propagation is expected to be favored by the stability of the virions and by enhanced genetic recombination. The large size of group VI of our collection would imply stable virions and is confirmed by stable titers of such phages under laboratory conditions. Comparatively high incidence of genetic recombination and its catalysis by phage-coded proteins in lambda are well documented (14, 15), and it is not illogical to conclude that, if examined, this property would also be demonstrable for phages that bear close phylogenetic relationship to lambda. Prophage P2 and related phages can convert the lysogens, rendering them ineffectual hosts for the propagation of unrelated phages (5, 11, 16). This characteristic and the ability of P2 (and possibly other related pages) to integrate at multiple attachment sites (3) may be looked upon as adaptations to the prophage as the principal mode of propagation. We note one characteristic of phage lambda which renders it a unique rather than a typical member of the lambda family. Over 40 lambdoid phages have been studied by us and none of them has been found to exclude the rii mutants of T4. Since the exclusion of T4rII mutants is shown by some P2-related phages (5, 9), it is possible that P2 or a related phage may have contributed the rex gene to lambda. A similar proposal has previously been made to explain the antigenic homology of two very dissimilar phages, namely, P2 and P1 (3). ACKNOWLEDGMENTS We are grateful to G. Bertani for providing unpublished data and for help in preparation of the manuscript. Part of this work was done in the laboratory of Stuart Linn, and we thank him for the hospitality. Alfred N. C. Lai provided valuable technical assistance, and we thank Alice Taylor for taking the electron micrographs. This investigation was financially supported from the Research Funds of our universities and by Public Health Service grant GM-9020 from the National Institutes of Health to Stuart Linn. LITERATURE CITED 1. Aizawa, S., and A. Matsushiro Studies on temperature sensitive growth of phage 080. I. Prophage excision. Virology 67: Bertani, L. E Abortive induction of bacteriophage P2. Virology 36: Bertani, L. E., and G. Bertani Genetics of P2 and related phages. Adv. Genet. 16: Calendar, R., J. Geisselsoder, M. G. Sunshine, E. W. Six, and B. H. Lindqvist The P2-P4 transactivation system, p In H. Fraenkel-Conrat and R. R. Wagner (ed.), Comprehensive virology, vol. 8. Plenum Publishing Corp., New York.

8 VOL. 39, Dhillon, E. K. S., and T. S. Dhillon HK239: a P2 related temperate phage which excludes rii mutants of T4. Virology 55: Dhillon, T. S., and E. K. S. Dhillon Studies on bacteriophage distribution. II. Isolation and host range based classification of phages active on three species of Enterobacteriaceae. Jpn. J. Microbiol. 16: Dhillon, T. S., and E. K. S. Dhillon Temperate coliphage HK022: clear plaque mutants and preliminary vegetative map. Jpn. J. Microbiol. 20: Dhillon, T. S., E. K. S. Dhillon, H. C. Chau, W. K. Li, and A. H. C. Tsang Studies on bacteriophage distribution: virulent and temperate bacteriophage content of mammalian feces. Appl. Environ. Microbiol. 32: Dove, W. F The extent of rii deletions in phage T4. Genet. Res. (Cambridge) 11: Ishii, X., and S. Kondo Spontaneous and radiation-induced mutations in Escherichia coli strains with different repair capacities. Mutat. Res. 16: TEMPERATE COLIPHAGE HABITATS Rutberg, L., and B. Rutberg On the expression of the rii mutation of T-even bacteriophages in Escherichia coli strain B. Virology 22: Sato, K., Y. Nishimune, M. Sato, R. Numich, and A. Matsushiro Suppressor-sensitive mutants of coliphage 480. Virology 34: Signer, E. R Plasmid formation: a new mode of lysogeny in phage A. Nature (London) 223: Signer, E. R General recombination, p In A. D. Hershey (ed.), The bacteriophage lambda. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 15. Signer, E. R., and J. Weil Recombination in bacteriophage A. I. Mutants deficient in general recombination. J. Mol. Biol. 34: Smith, H. S., L. I. Pizer, L. Pylkas, and S. Lederberg Abortive infection of Shigella dysenteriae P2 by T2 bacteriophage. J. Virol. 4: Thomas, R Control circuits, p In A. D. Hershey (ed.), The bacteriophage lambda. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.