Directed transposition of the arabinose operon: A technique for the isolation of specialized transducing bacteriophages for any Escherichia coli gene

J. Mol. Biol. (1969) 44, 117-127

Directed Transposition of the Arabinose Operon: A Technique for the Isolation of Specialized Transducing Bacteriophages for any Escherichia coli Genei SUSAN GOTTESMAN AND JONATHAN R. BECEWITH Department of Bacteriology and Immunology Haruard i5fedical School Boston, Mam. 02115, U.S.A. (Received 8 April 1969, and in rev&d form 22 Nay 1969) An F-thr-araf episome has been inserted into the Emherichiu coli chromosome at the Tl locus near the attachment site for bacteriophage 480. From a strain carrying such an insertion, a 480dara transducing phage was isolated. The techniques used should be applicable for the isolation of .specialized transducing phages for many E. co.5 genes.

1. Introduction The analysis of regulatory systems in Escherichia coli has been facilitated by the existence of two useful genetic tools. First, F’ factors carrying bacterial genes can be used to examine the behavior of regulatory mutants in strains diploid for the genes under study. Such analyses have been essential in determining the two major types of gene control in E. c&-negative control by gene repression (Jacob Q Monod, 1961) and positive control by gene activation (Sheppard t Englesberg, 1967). The recently developed technique of Low (1968) for isolating F’ factors makes the isolation of an F’ factor for any gene in E. co.3relatively simple. A second genetic tool for studies of regulation is the specialized transducing phage. Such transducing phages have been particularly useful in studies on the luc, trp and gal operons. The isolation of a @Odluc transducing phage has made possible the following studies on the regulation of kzc operon expression: (1) in strains lysogenic for the @Odluc, a thorough deletion analysis of the lac operon can be done. The characterization of these deletions has led to much new information including the determination of the location of the lac promoter site (Miller, Ippen, Scaife & Beckwith, 1968; Miller, Beckwith & Miiller-Hill, 1968; Reznikoff & Beckwith, 1969; Reenikoff, Miller, Scaife & Beckwith, 1969); (2) induction of strains lysogenic for certain derivatives of #80dlac allow an enormous increase in the number of lac gene copies per cell, t when the reprint supply of this paper is exhausted, no replies to reprint requests will be made. $ Abbreviations used: lac, ara, gal, utilization of lactose, arabinose, gslactose; hb, p?lr, PUT, arg, thr, LX, pro, trp requirement for histidine, pyrimidine, purine, srginine, threonine, leuoine, proline, tryptophan; Tl’, Tl,V, resistance to pbage Tl, phagee Tl and Tb; attE0, chromosome attachment site for 480; Smn, Sm’, sensitive and resistant to streptomycin; 480dara, 480dZac, defective 480 transducing phage carrying the ara region, the Zao region; FTs, F factor with mutation rendering its replication temperature sensitive; &Oh, 480 host range; 48Ov, 480 virulent; h h480, h-480 hybrid having A immunity and 480 host range and attachment spec3loity. 117

118

S. GOTTESMAN

AND

J. R. BECKWITH

thus increasing the concentration of luc repressor and making its isolation much simpler (Miiller-Hill, Crapo & Gilbert 1968) ; (3) the isolation of DNA from @OdZuc high frequency transducing lysates provides DNA highly enriched for the lac genes. Such preparations have been necessary in studying the interaction of lac repressor with Zac operator DNA (Gilbert $ Miiller-Hill, 1967; Riggs, Bourgeois, Newby & Cohn, 1968) and in studying the expression of the lac genes in vitro (Zubay, Lederman & DeVries, 1967). Until recently, the usefulness of specialized transducing phages such as X and $80 was limited to the transduction of those genes which are located close to the sites of insertion of the phage DNA in the bacterial chromosome. However, we have shown that it is possible to transpose the luc genes from their normal site on the E. coli chromosome to sites close enough to both dth and att80 so that Xdluc and 480dlac phages can be obtained. The isolation of such phages is not possible in wild-type strains. In the case of att80, this was done by selecting for insertion of an F-lac factor into the nearby locus (Fig. 1) determining sensitivity to bacteriophage Tl (Beckwith & Signer, 1966). This insertion renders the locus inactive and makes the bacteria resistant to Tl. In the case of atth, this was done by selecting for insertion of the F-luc into the nearby gal operon, (Beckwith, unpublished results) using the technique for selection of gal mutants described by Soffer (1961). We call these derivatives “directed transposition” strains (Beckwith, Signer & Epstein, 1966). Three conditions of the F&c directed transposition experiments seemed particularly critical. First, strains in which the transpositions were selected carried extensive deletions of the luc region, so that there was no possibility for insertion of the F-Zuc at the normal kzclocation. Second, the F&c episomes used were temperature-sensitive in their replication (F&UC), so that at 42°C they were not maintained in a bacterial population (Cuzin & Jacob, 1967). Strains in which the F,,lac has been integrated into the bacterial chromosome no longer exhibit a temperature-sensitive character and can be detected as temperature-stable lac + derivatives. Finally, even when strains are isolated in which insertions of the lac genes close to a phage attachment site have occurred, there must not be too much of the episomal material between the attachment site and the lac genes, or the isolation of the transducing phage may be impossible. The isolation of hdlac and 480dlac transducing phages raises the possibility that any gene or group of genes in E. coli might be incorporated into a specialized transducing phage. The arabinose operon of E. coli appears to be a particularly important group of genes to analyze. This operon represents the best studied example of a system which, according to the genetic evidence, is regulated by positive control (Sheppard & Englesberg, 1967). In this paper, we describe the extension of the directed transposition technique to the ara genes of E. coli. From one of the directed transposition strains, a 480dara transducing phage has been isolated. The findings presented here indicate that the technique should be generalizable to any gene in E. coli. Furthermore, we have found that not all of the conditions used in the isolation of the $80dlac are essential for the isolation of the +80dara.

2. Materials and Methods (a) Bacterial &-aim The F’ara episome, derived from E. coli B, is described in Sheppard & Englesberg (1967). It carries thr, ara and possibly part of the Zeulocus. We found that transferring the episome

ISOLATION

OF +30dara

TRANSDUCING

PHAUE

119

into some leucine negative strains (all presumably carrying the same Ieu mutation) rendered the strains &u+, while other strains remained &u-. This suggests that only part of the lcu locus is carried by the episome. The R,,,, locus (Cohen & Jacob, 1969) which maps on the opposite side of thr from am is not carried by the episome. The deletion ara 714 covers genes C, B, and A of the arabinose locus. The F’am-/ara A& B,, C, strains used to test Tlr-aradeletions are E. wli B strains. These strains were provided by Dr Englesberg. The ara- point mutant was u.v.-induced by J. Miller. me %114 Zac episome and the (Za~-pro)x~~~ deletion have been described previously (Chin 8s Jacob, 1964,1987). The pyC-, &-, pyrF-, trp- markers and the strain Xl78 which carries the colicinogenie factors are described in Signer, Beckwith & Brenner (1966). The arg- mutant, which lies in the arg-BCEH cluster, waz obtained from F. Jacob. The purE marker was obtained from Dr P. G. deHaan. #SO, #3oV and +SOiLwere obtained from E. Signer. (b) Media The following media were used: M63 (Pardee, Jacob BE Monod, 1969) minimal agar with 0.2% carbon source, 1~ vitamin Bl/ml., 114 cl~ streptomycin/ml., 20% amino acid/ ml.; M63 salts and buffer for dilution tubes; B broth (Signer, 1966); LB broth and F top 1968) ; arabinose tetrazolium indicator agar agar (Miller, Ippen, Scaife & Beckwith, (Ohlsson, Strigini & Beckwith, 1968) containing 1% arabinose (Calbiochem); Zac MacConkey (Difco) with 1% arabinose added or MacConkey base, also with 1% arabinose; H plates (10 g Bacto-Tryptone, 8 g NaCl, 10 g Bacto-agar/litre water) and H top agar (8 g Bacto-egar instead of 10/l.), and TYE plates (16 g Bacto-agar, 8 g NeCl, 10 g BactoTryptone, and 6 g yeast extract/litre water). (c) Mating pocedurea Matings were done as described in Miller, Ippen, Scaife L Beckwith

(1968).

(d) Stock prepara&on (i) 980 viruht and #SO hoat vange Plate lysates were made of #SO virulent and 480h (Adams, 1969). In order to be effective in Tlr selections, a +8Ov titer of at least 10” phage/ml. is necessary. (ii)

lyea& Strain Xl78 (lysogenic for colicins V and B) was grown up overnight in B broth and spread on Zuc-tetrazolium agar. The plate was grown overnight at 30°C. From the lawn a loopful of bacteria was inoculated into 10 ml. of B broth. These cultures were grown with aeration to 3 x 10s bacteria/ml., centrifuged and the pellet resuspended in 6 ml. of O-1 wM~SO,. Each 6 ml. culture was irradiated for 60 set with a General Electric 16 w germicidal lamp, G16T8, at a distance of 69 cm. The irradiated culture was added to a t-liter flask covered with aluminum foil and holding 6 ml. of B broth for each 6 ml. to be added. A total volume of 100 ml, per flask was grown with vigorous aeration for 4 hr. Five ml. of chloroform was added, the bacterial debris pelleted, and the supernatant pooled and used for Tlr selections. Chloroform inactivates colicin during storage, so an effort was made to exclude the chloroform in collecting the supernatant after centrifugation. COZioin

(e) Selection of TIT straina A 6.ml. overnight culture of the strain is centrifuged to concentrate the cells. The supernatant is discarded, and the pellet is resuspended in the few drops remaining by vigorous agitation. Three drops of the suspension are removed to a second tube and 0.6 ml. of colicin and 0.6 ml. of 980~ lysate added. The adsorption mixture is incubated for 20 min at 37”C, and then 3 to 6 drops are spread on arabinose MacConkey agar and incubated at 42°C. Tlr cells grow more slowly than Tl’ cells both in broth and on solid media. On tetrazolium agar, Tlr colonies give a poor color reaction. MacConkey plates are more sensitive and the color is quite clear within 20 hr, even for TIC colonies. This improvement of the

120 technique dent Hfr

S. GOTTESMAN

AND

J. R. BECKWITH

was developed after the selections had begun, and the fist of the three indepenstrains discussed below was detected on tetrazolium plates. (f) $80 transduction

procedures

A stationary phase culture of the recipent was concentrated and resuspended in the original volume of 0.1 M-MgSO,. This was aerated at 37°C for 20 min. Two-tenths of a ml. of undiluted phage lysate was adsorbed to 0.2 ml. of this starved cell culture for 20 min at 37’C, and then plated in 2.5 ml. F top agar on ara,-minimal agar containing streptomycin. The multiplicity of infection is about 10 phage/bacterium. When concentrated phage lysates were used, the recipient culture was also concentrated lo-fold and starved for 1 hr instead of 20 min. (g) Formation

of defective

lysogen

A culture of a strain carrying ara-deletion 714 was transduced using a number of dilutions of high frequency transducing lysate. After adsorption at 37”C, the bacteria were spread on pre-warmed ara-minimal agar at 42°C. Since 480 does not carry out successful infections at 42”C, single lysogens could not be reinfected by active phage released on the plate. Single colonies from the highest dilution plates were purified and grown up in culture. These cultures were tested for $80 immunity and for release of active phage. Defective lysogens were also expected to require helper to give transducing lysates. (h) Selection

of TP-ara-

deletions

in the defective

lysogen

Independent 5 ml. cultures of the defective lysogen were grown overnight and treated with colicin and +8Ov as described above. A few drops of each treated culture were spread on arabinose MacConkey-base agar. Tlr-amcolonies were purified and tested for 480 immunity by spotting a drop of a @Oh lysate on a lawn of bacteria on TYE agar. Cultures having immunity do not show lysis at the site of the $80 spot. Spot mating tests were carried out with these Tl’-aradeletions and F’ara-/arudonors homogenotic for either araA-, araBor araCpoint mutants. A drop of a broth culture of each is mixed on arabinose minimal agar containing tryptophan (20 pg/ml.) and sodium citrate (0.004 M). The Tlr cells require citrate for growth on minimal agar, and tryptophan is added in case the Tl deletion also deleted part of the nearby tqn locus. A positive test usually was obor as at least 20 recombinant ura+ served as a confluent spot of ara+ complementation colonies.

3. Results It should be remembered that all of these experiments involve the use of an episome from E. coli B (Sheppard & Englesberg, 1967) introduced into E. coli K12 strains. This episome carries the thr locus and possibly a portion of the leu operon in addition to the ara genes. It was isolated from an Hfr strain which transfers in a counterclockwise direction and the episome itself mobilizes the chromosome in that direction, indicating that its structure is as shown in Figure 1A. (a) Conversion

of an F-ara to an F,,-ara

In order to make the F-ara temperature-sensitive in its replication, we have recombined into it the FTSllC character (Cuzin & Jacob, 1967) from the F,,Zac used in previous studies. Normally, male cells of E. coli cannot act as recipients of genetic material from another male. However, it is possible to make what are known as female phenocopies of male strains by, for instance, allowing them to be aerated in stationary phase for long periods of time (Brinton, Gemski & Carnahan, 1964). We have made a female phenocopy of a male strain of genotype FTsll&c+/ ~(ZUCara- by growing it for 24 hours in LB broth at 30°C. It was crossed for 30 Po)-x111,

ISOLATION

OF

@Odara

TRANSDUCING

121

PHAGE

thr

A

:; I 1

I

I

-

Tl

B

aft80 I

I

Tt

5

thr I

DZC Illll

b

Tl 3 I

tJ-P

FIG. 1. The F-ara episome and its insertion into the Tl loous. The dotted lines in part A indicate a single reciprocal crossover event leading to insertion episome.

of the

minutes with an exponential phase culture of a donor of genotype F-um+/ara-trp~TF-his-. Ara+tq+pyrF+his+ recombinants, appearing at a frequency of l/lo3 recipient cells, are selected on am-minimal agar containing proline at 3O”C, repurified on minimal agar with proline and then tested for the temperature sensitivity of the ara character. From each of 94 stabs that grew up well, cells were streaked out to single colonies on arabinose-tetrazolium agar at 30 and 42°C. All streaks were predominantly white (am’) at 30°C. A clone was judged temperature sensitive for the am character and tested further if the 42°C streak showed red @a-) sectoring of many colonies. Of these 94 ara+ recombinanta, three were ara+ at 3O”C, but lost the F-ara episome at 42°C. These strains were la.c- at both temperatures. We presume that these strains arose from recombination events between the two episomes F,,lac and F-ara, in cells which temporarily harboured them both. This episome is thus probably F,rs,,,aru. (b) Integration of Frzara into the Tl locus In isolating directed transpositions of Eat, a strain carrying the F,&e and an extensive chromosomal deletion of the lac region was used. As far as we know, there is no deletion in E. coli which covers the entire region aontained in the F-ara episome, thr-ara-leu. Therefore, we sought to determine whether directed transposition could be found even in strains which contained chromosomal regions with extensive homology for the episome. Specifically, we sought integration of the episome in the Tls locus of E. wli K12. This locus maps between att80 and trp and probably specifies a portion of the receptor sites for Tl, $30 and colicins V snd B (Gratis, 1964). Details of selection of Tl’ cells are given in Materials and Methods. The diploid strain, EX2, of genotype FTsara+/ara- V (Zuc-~ro),,,,Sma with an unchars&erized ara- point mutant was used. This strain, when spread at 42°C on ara-MacConkey agar indicator medium gave approximately 3% am+ colonies. Any stable am+ colonies at 42°C might be due to : (1) recombination in the ara region so that the am+ allele is located on the chromosome ; or (2) integration of the episome

122

S. GOTTESMAN

AND

J. R.

BECKWITH

into the chromosome in the ara region or at some other site. This, then, is the background of temperature-stable ara+ colonies we must expect in seeking the specific directed transpositions. Eighty-two broth cultures from independent colonies of this strain were treated with agents necessary to select for Tl’ mutants and spread on ara-MacConkey agar at 42°C. Ordinarily, about lo0 cells from each culture were treated and between 50 and 300 TIP mutants found. Nearly every culture gave rise to a small proportion (2 to 3%) of Tl’ colonies which were ara+ at 42°C. Most of these are presumably the background we have discussed. The colonies grown up after one day were examined under low magnification (15 x ) for non-sectoring ara+ colonies. These colonies appear uniformly red on the MacConkey plates, while colonies in which the episome is not stably integrated range from totally white to red with fine light sectors. The stable ara + colonies were further purified and tested for the site of integration of the episome. (c) A caution: nature of the P,,&ac The particular temperature-sensitive FTsl14 lac used is not completely inhibited in its replication at 42°C (Cuzin & Jacob, 1967). As a result, there is some growth of the strain of genotype F,slac+/ (lac deletion) on lacminimal agar at 42”C, since a small proportion of the bacteria still carry the episome. In addition, on the indicator media lactose-EMB agar and lactose-MacConkey agar, colonies of this strain have considerable red color at 42°C. However, on these indicator media close examination of the colonies of this strain show that the color is due to sector-like shapes in the colony rather than to a uniform red color. Thus, a colony which is stable luc + (carries an integrated episome) at 42°C can be easily distinguished from one which is losing the episome. On lac-tetrazolium indicator medium, it is much easier to distinguish temperature-resistant EaC+ colonies from the temperature-sensitive ones. The properties are emphasized here because it may be more difficult to distingush a strain with an integrated episome from one which is losing its episome on a selective medium. We have not examined this problem in detail. Furthermore, in the crosses carried out to convert the F-ara into an F,sara, an indicator medium is again used to test the temperature-sensitive character. Therefore, it is possible that for certain markers it may be necessary to isolate a mutant of the F-factor which is more temperature-sensitive. Of the E”,slac strains described by Cuzin & Jacob (1967), none seem more suitable than Fr,,,,Zac. Although certain of them are more temperature sensitive in their replication, they are also much more unstable and are transferred from cell to cell at very low frequency. (d) Testing for the site of integration of P,,ara Since the sex factor, F, is inserted in the chromosome along with the ara genes, any strains in which the F,,ara is inserted in the Tl locus are likely to be Hfr strains which donate with their origin in the Tl region between pyrC and trp. From each selection plate a number of ara+ colonies were picked and tested to determine if the F,,ara were in the Tl region. The strains were mated with a strain of genotype F-pyrCtrp-SmP. The mating was interrupted at 30 minutes, and O-1 ml. of a l/100 dilution of the mating mixture plated on glucose-minimal agar containing streptomycin and either tryptophan or uracil. A large number of colonies on one plate of each set (> 1000) and < 10 colonies on the corresponding plate was considered indication of integration in or near the Tl locus. Temperature-resistant ara+ derivatives from 3 of

ISOLATION

OF

+80daro

TRANSDUCING

PHAGE

123

the 82 cultures tested were of the type we were seeking; all three transferred the WC marker very early and the trp marker late. Two of these three cultures derived from plates with “jackpots” (30% and 60%) of Tl’ara+ colonies. These were the only two ara+ jackpots found in the 82 cultures. One representative from each of the three cultures (EX3, EX6 and EX7) was chosen for further study. Some strains which were temperature-stable but did not show evidence of episomal integration in the Tl locus by these crosses were mated for short times (1 hr) with strains of genotype F-purE-Smr, (selecting for pur+Smr) and F-arg-Smr (selecting for arg+Smr). These two markers are on opposite sides of the ura locus; a high frequency of arg+ and a low frequency of purE + recombinants suggests episomal integration in the ara region. All strains which transferred any markers transferred arg early and purE late, supporting the hypothesis that there is a background of ara+ colonies due to integration of the episome in the chromosomal ara locus. The directed transposition Hfr strains were characterized to determine whether the episome had inserted in such a way that the Hfr transferred the thr and/or the ara markers early or late. Interrupted mating experiments showed that both markers are transferred early by all three strains. Since the original F’-ara has the orientation origin+ra (C-BAD)-thr, we believe these strains to have the structure indicated in Figure 1B. All three Hfr strains were somewhat unstable in that the Fzsara came out of the chromosome at an appreciable frequency. In each experiment, the strains were checked for their temperature-resistant, ara + character, and for their ability to donate the ~~yrrCmarker, but not the trp marker, early.

(e) Isolation of +80&a

tramtducing bacteriap?taga

The thr and ara markers are normally cotransduced by bacteriophage Pl at a frequency of 2% (Lennox, 1955). This indicates a sizeeble distance between the two markers, probably representing a DNA length of more than 30 x IO6 daltons, since the molecular weight of the DNA of 8 Pl transducing particle is 60x 10” dsltons (Ikede & Tomizawa, 1965). $80 DNA has a molecular weight of approximately 30 x lo6 (Yamagishi, Yoshizako & S&o, 1966). Therefore, it is difficult to see how a #3Odara transducing phage particle could be formed which would include all the material between a.tt80 and ara. However, we would expect that certain types of double events could give rise to a #@dura transducing phagge(Fig. 2A and B). A strain which carried a deletion mutation removing the thr genes and much of the material between the phage genome and ara could bring ara close enough to the phage genes so that a +Oduru could be isolated. Alternatively, an inversion of the ar&hr region could bring the ara genes closer to a$t80. Thus, if we look hard enough, i.e. at enough phsge, from an induced lysate of $80 lysogens of the directed transposition strrtins, we might expect to find a rare 48Odura transducing particle, due to a double event. Since Tl’ strains are also +80-resistant, we have made 480 lysogens of these strsins with a $80 host range mutant (48Oh).Lysates were induced with ultraviolet light and test4 for transduction of an ara- strsin. A recipient strain carrying an ara- (CBA) deletion was used to avoid contamination with revertants. Since we expected arabinose transductants to be rare, transductions were also done with a tenfold concentrated phage lysate. Colonies growing up on the transduction plates were purified on arabinose-tetrezolium egar and used to make u.v.-induced lysates. A drop of each of these lysates

124

S. GOTTESMAN Early

AND

I

imm 80

wz

R. BECKWITH

Late

off 80 imm 80 I I

atf80

J.

h I

fhr I I

aft80 I

Deletion

Inversion

ara DABC

off 80 I

‘!‘I’

1

imm 80 I 2

I

aft80

h

I

80

off 80 I

Lysogenization aft80 I

fhr I

I Y

aft80 ara 1 immaODABC, I I Ill.

+ ara CBAD I I

Transducing phage formation

imm

aft80

ara DABC I ! I I I

Tl I

atf 80 -

imm80

ara Y C,B,AD

off80 I

(II

Tl ’ -----

C

FIG. 2. The structure of the 480 lysogen of EX3, the am transducing phages derived from it, end the defeotive 48Odara lysogens. The sequence of events in part A represent the formation of +80dara by a deletion and subsequent transduoiug phage formation and in part B by an inversion and subsequent transducing phage formation. Part C represents the classes of Tl’-mm- deletions obtained from the 480dara defective lysogen. Although specific classes of deletions and inversions are presented here to account for the ~SOdam obtained, clearly other such events could be equally suitable. For instance, instead of inversion of the h-am&f region as the Grst step, the entire phage along with the t/w locus could have hem inverted and given rise to the same transduoing phage.

ISOLATION

OF

~8Odara

TRANSDUCING

PHAQE

125

was spotted on a lawn of ara- bacteria on arabinose-minimal agar; high frequency transducing lysates cause growth in the spot. The transducing titer of high frequency transducing lysates was consistently lob3 times lower than the active phage titer. However, we have evidence from density gradient centrifugation that the transducing particles are present in nearly equal number to the active phage particles (Gottesman & Beckwith, unpublished results). Several independent lysates were made from lysogens of EX3 and EX7. All lysates gave rise to ara+ transductants at frequencies around lo-l1 per active phage. Nine of these transductants derived from one lysate of EX3 were themselves induced with ultraviolet light and seven gave rise to high frequency ara transducing lysates. The transducing particles were shown to be 48Oara particles by the following tests. (1) Strains resistant to #8Oh (T1,5’) cannot be transduced by these lysates to ara+. (2) The ara+ transductants can be cured by superinfection with M80, indicating that the +SOdara is at the +SO site (Signer & Beokwith, 1966). The superinfecting phage lysate was spread on arabinose-tetrazolium agar, allowed to dry, and the transductant to be tested was streaked over this background. Curing is seen as extensive ara- sectoring of colonies. (3) Properties of defective lysogen (see below). As expected from the discussion above, none of the +SOdara high frequency transducing lysates could transduce LSthr- strain to thr + . It appears then that there has been a rearrangement of the genetic material between the $80 insertion site and the ara genes as part of or prior to the formation of the +8Odara. An indication of the type of rearrangement involved comes from a study of 8 defective lysogen of one of these 48Odara strains. Ordinarily, a 480 transducing phage carrying ohromosomal genes from the trp side of a.#80 would be expected to be missing certain early genes but contain all the late genes (Franklin, Dove & Yanofsky, 1965; Signer & Beckwith, 1966; Kayajanian, 1968). A 48Odara defective lysogen was obtained (see Materials and Methods) and shown to still carry the phage immunity region. This 48Odara is inserted at the att80 site as shown by its ability to be cured by superinfection with hh80. We can test for the presence of a trp-proximal late gene of 480, since the phage used in the isolation of the #80dara and subsequent superinfection with wild-type $80 (h+) gave rise to a lysate which was titrated on a Tl’ (+80r, +80hs) strain at a low dilution and on a T18 (+80s, +80hs) strain at higher dilutions. No $SOh recombinants (
120

S. GOTTESMAN

AND

J. R. BECKWITH

4. Discussion We have described a technique for isolating a @Odara transducing phage. The procedure used should be applicable to any gene in E. coli (and closely related organisms) subject to technical problems mentioned in Materials and Methods. Although the principles involved in the isolation of the @Odara are the same as those for @Odlac and Xdluc, three new technical points come out of these studies. First, a wild-type temperature-insensitive episome was easily converted to a temperature-sensitive episome by the type of cross described. The results indicate relatively frequent recombination between sex factors. Second, the strains in which directed transpositions are being sought need not contain chromosomal deletions corresponding to the regions carried by the F,s ep isome. The generality of this point may depend on the size of the episome used. F’-factors carrying larger regions of the chromosome may integrate more frequently at their homologous regions, resulting in an intolerable background of temperature-resistant derivatives. Ideally, one would use strains oarrying the chromosomal deletions where feasible. However, as seen here, this is not an essential component of the procedure. Finally, it seems likely that once a directed transposition strain has been isolated in a site near attA or dt80, it should be possible to isolate a transducing phage for the desired marker no matter how far that marker is from the phage attachment site. In the strain EX3, the distance between the ara genes and att80 is too great to allow formation of a @Odara by a single event. Either a deletion or an inversion of material in this region could give rise to strains in which q%Odara formation by the normal mechanism is possible. Double events are thought to be involved in the formation of certain Xpbio transducing phages (Manley, Signer & Radding, 1969) and ~80peu+,,, (R. L. Russell, J. N. Abelson, A. Landy, S. Brenner, M. L. Gefter & J. D. Smith, manuscript in preparation). Two properties of the @Odara analyzed here indicate that it is the product of a double event. First, the ara genes have replaced the late gene region of 480 contrary to the expectations of a Campbell model type single event. This finding can be explained by either a deletion or an inversion of material in this region. Secondly, the ara genes in the 480dara have become inverted from their original orientation. This structure could not have arisen solely by a deletion and subsequent transducing particle formation. The origin of the q580dara is most easily explained then by postulating an inversion and subsequent Campbell (1962) model type event in the 480 lysogen of EX3. (The U.V. irradiation may have been responsible for the inversion.) These findings raise the possibility that even genes very far away from a%80 can be transduced by +80 at very low frequency. One potentially useful tinding is that in two of the three cultures in which directed transpositions of the ara genes had occurred, the selection plate showed a high proportion of stable ara + strains. These “jackpots” are presumably due to the accumulation of integrations occurring relatively early in the history of the particular clones. The failure to find “jackpots” for strains with the F-ara integrated in the ara region may, in part, be explained by the expected instability of such integrations. Thus, it may be necessary only to screen TIP selection plates for such “jackpots” in order to detect the directed transposition. The @Odara transducing phages should be extremely useful in characterizing the control mechanisms involved in regulation of the ara genes. As pointed out in the Introduction, the isolation of the G gene product becomes much simpler with such

ISOLATION

OF 480dara

TRANSDUCING

PHAGE

127

a @age. Subsequent studies on interaction of the araC product with ara operon DNA will use DNA isolated from @Odara transducing lysates. This work was supported by grant no. GM-13017 and by a Career Development Award both from the National Institutes of Health to one of us (J.R.B.). We wish to thank Dr Ellis Englesberg for generously providing the episome and most of the ara- mutations. We also acknowledge the excellent technical assistance of Miss Veronica MacGillivray. REFERENCES Adams, M. H. (1959). Bacteriophagea, p. 454. New York: Interscience. Beckwith, J. R. & Signer, E. R. (1966). J. Mol. BioZ. 19, 254. Beckwith, J. R., Signer, E. R. & Epstein, W. (1966). Cold Spr. Harb. Symp. Quant. Biol. 31, 393. Brinton, C. C. Jr., Gemski, P. & Carnahan, J. (1964). Proc. Nat. Acad. Sci., Wmh. 52, 776. Campbell, A. (1962). Adwanc. Genetica, 11, 101. Cohen, G. N. & Jacob, F. (1959). C.R. Acud. Sci. Park, 248, 3490. Cuzin, F. & Jacob, F. (1964). C.R. Acud. Sci. Paris, 258, 1350. Cuzin, F. & Jacob, F. (1967). Ann. Imt. Pmteur, 112, 397. Franklin, N. C., Dove, W. F. t Yanofsky, C. (1965). Biochem. Biophys. Res. Con-m. 18,910. Gilbert, W. t Miiller-Hill, B. (1967). Proc. Nat. Acad. Sci., Wash. 58, 2415. Gratia, J. P. (1964). Ann. Imt. Pasteur, 107, 132. Ikeda, K. & Tomizawa, J. (1965). J. Mol. Biol. 14, 85. Jacob, F. t Monod, J. (1961). J. Mol. BioZ. 3, 318. Kayajanian, G. (1968). Virology, 36, 30. Lennox, E. S. (1955). v&roZogy, 1, 190. Low, B. (1968). PTOC. Nat. Acud. Sci., Wash. 60, 160. Manley, K. F., Signer, E. R. t Radding, C. M. (1969). vi’irology, 37, 177. Miller, J. H., Beckwith, J. R. & Mi.iller-Hill, B. (1968). Nature, 220, 1287. Miller, J. H., Ippen, K., Scaife, J. G. I% Beckwith, J. R. (1968). J. Mol. BioZ. 38, 413. Miiller-Hill, B., Crapo, L. & Gilbert, W. (1968). Proc. Nat. Acud. SC&, Wash. 59, 1259. Ohlsson, B. M., Strigini, P. F. & Beckwith, J. R. (1968). J. Mol. BioZ. 36, 209. Pardee, A. B., Jacob, F. & Monod, J. (1959). J. Mol. Biol. 1, 165. Reznikoff, W. S. & Beckwith, J. R. (1969). J. Mol. Biol. 43, 215. Reznikoff, W. S., Miller, J. H., Scaife, J. G. & Beckwith, J. R. (1969). J. Mol. BioZ. 43, 201. Riggs, A. D., Bourgeois, S., Newby, R. F. & Cohn, M. (1968). J. Mol. BioZ. 34, 365. Sheppard, D. E. & Englesberg, E. (1967). J. Mol. BioZ. 25, 443. Signer, E. R. (1966). J. Mol. Biol. 15, 243. Signer, E. R. BEBeckwith, J. R. (1966). J. Mol. BioZ. 22, 33. Signer, E. R., Beckwith, J. R. & Brenner, S. (1965). J. Mol. BioZ. 14, 153. Soffer, R. L. (1961). J. Butt. 82, 471. Yamagishi, H., Yoshizako, F. & Sato, K. (1966). ViTology, 30, 29. Zubay, G., Lederman, M. & DeVries, J. K. (1967). Proc. Nat. Acad. Sci., Wash. 58, 1669.