`
`© by Springer-Verlag 1979
`
`Amplification of Chloramphenicol Resistance Transposons Carried
`by Phage PlCm in Escherichia coli
`
`Jiirg Meyer and Shigeru lida
`Dept. of Microbiology, Biozentrum der Universitéit, CH—4056 Basel, Switzerland
`
`Summary. We have characterized a number of PlCm
`phages which contain the resistance genes to chloram-
`phenicol and fusidic acid as IS1-flanked Cm transpo-
`sons. Restriction cleavage and electron microscopic
`analysis showed that
`these Cm transposons were
`carried as monomers (M) or tandem dimers (D). Ly-
`sogens of PlCm (D) are more resistant to chloram-
`phenicol than those of its PlCm (M) presumably as
`a result of an increased gene dosage. Amplification
`of the Cm transposons to tandem multimers was fre-
`quently observed in PlCm (D) lysogens grown in
`the presence of high concentrations of chlorampheni-
`col or fusidic acid and was also detected in PlCm
`
`(M) lysogens. The degree of amplification varied in
`different clones which suggests that cells containing
`spontaneously
`amplified Cm transposons were
`selected by high doses of the antibiotics. The dimeric
`as well as the amplified Cm transposons carried in
`P1C1n lysogens grown in the absence of chloramphe-
`nicol displayed considerable stability. Mechanisms for
`the amplification of the ISl-flanked transposons are
`discussed.
`
`Introduction
`
`Gene amplification of plasmid-born functions, partic-
`ularly drug resistance, has repeatedly been observed
`in Proteus mirabilis (Rownd and Mickel, 1971; Hashi-
`moto and Rownd, 1975; Tanaka et a1., 1976) and
`it is also described in Streptococcus faecalis (Clewell
`et al., 1975; Clewell and Yagi. 1977). In Esclzeric/zia
`colt", similar phenomena have recently been discovered
`and are described in this and other papers (Mattes
`et al., 1979; Chandler et al., 1979).
`We have isolated several hybrid Pl phages (Mise
`and Arber. 1976; lida and Arber, 1977) which have
`acquired drug resistance genes from the R plasmid
`
`NR1 (also called R100) and which still carry all essen-
`tial genes for phage replication and lysogenization.
`Restriction enzyme cleavage and clcctron microscope
`analysis of their DNA revealed the presence of ISl-
`flanked transposons containing all or part of the r-dc-
`terminant (r-det) of NR1 inserted at the unique 151
`site of P1 (lida et al., 1978a) or at various other loca-
`tions in the Pl genome (Arber et al., 1978). These
`hybrid phages may have evolved either by a transposi-
`tion event or by cointegration of the P1 and NR1
`plasmids by recombination between their 1S1 ele-
`ments and subsequent deletion formation (Iida et al.,
`1978 h). Only a few of the resulting Cml transposons
`have the same size as the Tn9 which is also an 1S1-
`
`flanked Cm transposon (MacHattie and Jackowski,
`1977).
`Here we will present evidence that some of the
`Cm transposons are integrated in P1 DNA as tandem
`dimers of the structure IS1 — Cm — ISl — Cm — 1S1
`
`and that further amplification to higher oligomeric
`repeats occurs and can be detected after growth of
`E. coli lysogenic for PlCm in the presence of high
`concentrations of chloramphenicol or fusidic acid.
`Part of these results have been reported by Meyer
`et al. (1978).
`
`Materials and Methods
`
`Media, Bacteria and Bacteriophage
`
`The media used were as described by Iida and Arber (1977). Bacte-
`riophage P1 and its derivatives are listed in Table 1. Phage Plr—det
`is PlCmSmSuHg8lr]ts225 described by Arber et al. (1978). The
`isolation and the characterization of Cm’ Lransducing P1 phages
`
`Abbreviations. Cm:chloran1phenicol; Fasfusidic acid; bp=
`‘
`base pairs; kbskilobase pairs; PlCm (M)=Pl phage carrying
`a monomer Cm transposon; PlCm (D)=Pl phage carrying a tan-
`dem dimer Cm Lransposon; PlCm (A)=P1 phage carrying a tan-
`dem multimer (amplified) Cm transposon
`
`0026-8925/79/O176/0209/$02.20
`
`SANOFI v. GENENTECH
`IPR2015-01624
`EXHIBIT 2087
`
`
`
`210
`
`Table 1. Phage Pl strains
`
`Strain
`
`Reference
`
`J. Meyer and S. Iida: Gene Amplification in E. coli
`
`Cm transposon carried
`
`Insertion site Monomer size“
`on P1 DNA
`
`(map units)
`
`Number of
`Cm’ genes
`
`Designation
`used in the
`text
`
`PlcIts225
`PlCm0
`
`P1Cm0cZts225
`P1Cm0cItslOO
`PlCml3cIts225
`P1Cm89cIts225
`P1Cm88cIts225
`PlCm92cIts225
`PlCm80c]ts225
`PlCm248cIts225
`
`—
`Iida et al. (1978 a)
`new designation for P1CM isolated 20
`by Kondo and Mitsuhashi (1964)
`This study
`Rosner (1972)
`Iida et al. (l978a)
`Iida et al. (1978a)
`This study
`This study
`Arber et al. (1978)
`Arber et al. (1978)
`
`47
`20
`
`20
`
`2.7 kb =Tn9
`
`2.7 kb
`2.7 kb
`2.6 kb
`2.6 kb
`2.6 kb
`2.6 kb
`2.9 kb
`48 kb
`
`[\)>—-[\,)|\)»—->—|\,)>—-
`
`P1Cm0 (M)
`PlCm0 (D)
`PlCm13 (M)
`P1Cm89 (M)
`PlCm88 (D)
`P1Cm92 (D)
`PlCm80 (M)
`P1Cm248 (D)
`
`All PlCm phages listed are plaque forming and belong to type A as defined by Iida and Arber (1977). PlCml3 and PlCm89 DNA
`gave the same restriction cleavage patterns and so did P1Cm88 and P1Cm92 DNA
`a
`
`The monomer size gives the length of the Cm resistance determinant plus the two flanking 1S1 elements. This corresponds to
`the size of the Cm insertion at map units 4 and 47. At map unit 20, however, the insertion is about 800 bp smaller because there
`is a resident 1S1 at this location
`
`1977). E. coli WA921
`has been described (Iida and Arber,
`(thr leu met lac hsdSk) was used as a host for phage P1. The fusidic
`acid sensitive strain DB10 was obtained from Dr. J. Davies (Datta
`etal., 1974).
`
`Selection for High Levels of Drug Resistance
`
`Selection of lysogens resistant to high doses of chloramphenicol
`followed either of two protocols. Method]: A single colony of
`cells lysogenic for PlCm was suspended in LB containing 25 pg/ml
`Cm and grown at 30° C to saturation. After checking the structure
`of PlCm DNA by restriction cleavage analysis (see next section)
`the culture was diluted 50-fold in LB containing 300 pg/ml Cm
`and grown at 30° C. It usually took 1-3 days to grow the culture
`to saturation. Method 2: A single colony of cells lysogenic for
`PlCm was suspended in LB containing 25 pg/ml Cm and grown
`at 30° C to saturation. After checking the structure of PlCm DNA
`the culture was diluted 100-fold in LB containing twice the concen-
`tration of Cm. The procedure was repeated until the final Cm
`concentration had reached 1 mg/ml. In both methods the cultures
`were then diluted 50-fold in LBMg, grown to 2 x 108 cells/ml and
`phage was induced. The selection of bacteria resistant to high
`doses of Fa followed Method 1.
`
`Isolation and Restriction Cleavage Analysis of Phage DNA
`
`8.5 was added to a final concentration of 50 mM. After 5 min
`incubation at 37° C the mixture was extracted with phenol, DNA
`was concentrated by ethanol precipitation and redissolved in 200 pl
`of 10 mM Tris pH 7.4, 10 mM NaCl,
`1 mM EDTA. Samples of
`10 to 20 pl were digested with restriction enzymes and subjected
`to gel electrophoresis as described (Bachi and Arber, 1977).
`
`Electron Microscopy
`
`Heteroduplex molecules were prepared by denaturation and renatu-
`ration of a mixture of DNA fragments and mounted for electron
`microscopy by the formamide technique of Davis et al.
`(1971).
`Intrastrand reannealing of inverted repeats resulting from rapid
`renaturation was visualized as snap back structures. PM2 DNA
`(Espejo et al., 1969) and fd DNA (Beck et al., 1978) served as
`internal length standards for double-stranded and single-stranded
`DNA, respectively.
`
`Results
`
`1. Monomer and Tandem Dimer Forms
`
`of Diflerently Sized Cm Transposons are Found
`in Independent Isolates
`
`Phage was induced from 200 ml of cells lysogenic for PlCm and
`concentrated with polyethylene glycol (Iida and Arber, 1977). After
`treatment with 0.5 pg/ml each of pancreatic DNase and RNase
`A, phage particles were purified by a CsCl step gradient. Phage
`DNA was extracted with phenol.
`In order to analyse rapidly a large number of samples, 15 ml
`of phage lysate were prepared by heat induction. After removing
`the cell debris by low speed centrifugation, the lysate was treated
`with DNase and RNase and phage particles were precipitated with
`polyethylene glycol. They were resuspended in 1 ml of phage buffer
`(10 mM Tris-HCl pH 7.5, 10 mM MgSO4) to which EDTA pH
`
`A summary of studies on the location and size of
`the lS1—ilanked Cm insertions carried in PlCm phages
`is given by Arber et al. (1978). Detailed experimental
`data are presented here for the PlCm derivatives
`listed in Table 1. Several of the Cm transposons
`studied locate at P1 map unit 4, others at P1 map
`unit 20. Both these locations are within restriction
`
`fragment Bglll-2 of P1 DNA. None of the Cm inser-
`tions is cleaved by Bglll restriction endonuclease.
`
`
`
`J. Meyer and S. Iida: Gene Amplification in E. coli
`
`211
`
`a monomer and PlCm92 (D) a dimer of the resistance
`determinant. In PlCm92 (D) the insertion represents
`a tandem dimer of the structure ISl — Cm — ISl
`
`— Cm — ISl, with three directly repeated ISl elements
`(Fig. 3). This structure was confirmed by the observa-
`tion of single-stranded fragments of PlCm92 (D)
`DNA which formed snap back structures between
`sequences about 800 bp long, as have been described
`for PlCm89 (M) DNA (Iida et al., 1978a). The
`looped molecules fell into three classes which had
`the double—stranded part and one single—stranded end
`in common, but differed in the size of the loop and
`the other single—stranded end (Fig. 2, Table 2). The
`intramolecular reannealing occurred between the ISl
`element at map unit 20 of the P1 genome and either
`one of three ISl elements of opposite orientation pre-
`sent at both ends and in the centre of the insertion.
`
`Comparison of the double—stranded length of the
`BglII-2 DNA fragments of a second pair of PlCm
`derivatives PlCm0 (M) and PlCm0 (D) with the wild
`type fragment indicated additional sequences of 1.92
`kb and 3.95 kb, respectively. The sizes and structures
`of the Cm insertions were also determined by hetero-
`duplex mapping (Table 3). In heteroduplex molecules
`the variable location of the single—stranded loop
`within the ISI element at map unit 20 or within the
`monomer Cm transposon is direct evidence for the
`presence of ISl—flanked Cm insertions at this site and
`for the dimer structure of the Cm transposon on
`PlCm0 (D). Therefore, a similar monomer—dimer
`relation was demonstrated for PlCm0 (M) and
`PlCm0 (D) as for PlCm89 (M) and PlCm92 (D)
`(Fig. 3).
`
`b) Restriction Cleavage Studies. Since the Cm transpo-
`sons have no BglII site, as mentioned above, the Bglll
`fragment of PlCm DNA carrying the Cm insertion
`has a larger size. EcoRI cleaves the Cm transposons
`in question once (Arber et al., 1978), thus producing
`a fragment corresponding in size to one IS1 plus the
`Cm resistance determinant (“unit length” fragment)
`from tandem dimers, but not from monomers. PstI
`cuts once within the ISI element (Grindley, 1977;
`Othsubo and Ohtsubo, 1978; Iida et al., 1978 a) creat-
`ing one fragment from monomers and two fragments
`of the same length from dimers (Fig. 4C). EcoRI and
`Pstl restriction cleavage thus provides a rapid means
`for finding PlCm derivatives carrying a tandem dimer
`of an IS1—flanked Cm transposon.
`This situation is exemplified in Fig. 4: The Bgfll-2
`DNA fragment of PlCm13 (M) (Fig. 4A, slot b) is
`larger
`than the corresponding fragment of P1
`(Fig. 4A, slot a) and the one derived from PlCm88
`(D) is even larger (Fig. 4A, slot C). Psll cleavage of
`
`Fig. 1. Heteroduplex molecules between the Bg[II—2 fragments of
`PlCm89 (M) and PlCm92 (D) DNA. Only the relevant part is
`shown. The two electron micrographs give evidence for a tandem
`duplication by the presence of a single—stranded loop (arrows) of
`fixed size at variable positions between the boundaries of the mono-
`mer unit (principle outlined by Davidson and Szybalski, 1971).
`In the diagram the two extreme positions are shown. In this and
`all the following diagrams the straight lines represent Pl DNA,
`the black boxes identify IS1 elements and the Cm‘ determinant
`is indicated by the wavy lines. The bars on electron micrographs
`represent 1 kb length of DNA
`
`Electron microscope and restriction cleavage data re-
`veal that some of these Cm transposons occur as
`monomers, others as tandem dimers.
`
`a) Electron Microscopic Studies. Tandem repeats were
`detected in heteroduplex molecules between DNA
`fragments carrying the monomer and the dimer Cm
`transposon by the variable location of the single-
`stranded loop emanating from positions within the
`boundaries of the monomer unit. Figure 1 shows het-
`eroduplex molecules between the Bg[II-2 fragments
`of PlCm89 (M) and PlCm92 (D) DNA. A single-
`stranded loop of 1.66 kb originates from positions
`varying from 0.53 kb to 3.08 kb distance from one
`end (27 molecules). By EM criteria the Cm transpo-
`sons of PlCm92 (D) and PlCm89 (M) are inserted
`at the same site on Pl DNA and in the same orienta-
`
`tion. The only difference is that PlCm89 (M) carries
`
`
`
`J. Meyer and S. Iida: Gene Amplification in E. coli
`
`Fig. 2. Electron micrographs of snapback structures observed in single—stranded Bglll-2 fragments of PlCm92 (D) DNA. The diagram
`gives our interpretation of the structures: the intrastrand reannealing occurs between the IS1 at map unit 20 of the P1 genome and
`either of the three IS1 of the dimer Cm transposon. I, H and III refers to the class of molecules defined by the measurements given
`in Table 2
`
`Table 2. Length measurements of single-stranded snap back BglII—2
`fragments of PlCm92 (D) DNA
`
`Table 3. Analysis of heteroduplex molecules between the Bglll-2
`DNA fragments
`
`Class
`
`Number of Single-
`molecules
`stranded
`
`Single-stranded
`ends
`
`Double-
`stranded
`
`Heteroduplex
`
`measured
`
`loop T part
`Constant Variable
`
`Number Size of
`of
`loop
`mole-
`cules
`
`Location from
`right—hand end
`
`12.74
`:01“
`14.20
`--0.3
`16.14
`$0.2
`
`1.35
`i0.08
`1.33
`i0.16
`1.30
`i0.08
`
`3.80
`i0.l4
`2.04
`:0.11
`0.50
`j0.07
`
`a
`
`All measurements are given in kb
`
`DNA of P1Cm13 (M) and PlCm88 (D) (Fig. 4B,
`slots a and b) produces the “unit length” fragment
`once and twice, respectively. The same size DNA
`fragment is present in an Ec0RI digest of P1Cm88
`(D) (Fig. 4B, slot d), but is missing in that of P1Cm13
`(M) (Fig. 4B, slot c). It should be noted here that
`P1Cml3 (M) and P1Cm88 (D) are indistinguishable
`from P1Cm89 (M) and P1Cm92 (D), respectively,
`used for the electron microscopic characterization.
`The analogous situation is seen with PlCm0 (M)
`and PlCm0 (D) DNA. The BglI1—2 fragments are
`larger (Fig. 4A, slots e and t). The “unit length”
`fragment appearing in Pstl digests (Fig. 4B, slots f
`
`PlCm0 (M):P1
`
`PlCm0 (D) :P1
`
`15
`
`20
`
`P1Cm0(M):P1CmO (D) 17
`
`2
`
`1.96i0.16“ 1.2-2.2 kb
`(withinIS1)
`
`4.o1:0.24b 1.242.3kb
`(within IS1)
`1.3—4.4 kb
`(within monomer
`unit IS1-Cm-IS1)
`
`2.091013
`
`All measurements are given in kb
`Two molecules had two loops of 2 kb instead, resulting from
`1’
`reannealing of the central IS1 element of the dimer Cm insertion
`of PlCm0 (D) with the naturally occurring IS1 in the P1 DNA
`fragment
`
`_|_$l Cm|_$l Cm I51
`_..,..,.,.—.,.,.....‘;.
`~_
`4.2
`,.
`\;,..,.,.;«'
`*
`'
`
`
`‘~. 2 .6 ,'» z
`0.5
`
`P1Cm92(D)
`
`P1CmB9(M)
`
`<<
`IS1
`0.8 1.3
`
`L5; Cm IS1 Cm
`‘~.
`.
`I
`‘\ <§
`I
`P1CmO(D)
`3-
`‘*1;
`i IS1
`‘~W P1Cl'I'IO(M)
`\ :
`0.3
`1.3
`
`_
`
`135
`
`Fig. 3. Location, size and structure of monomer and dimer Cm
`transposons within the Bglll-2 fragment of P1 DNA. The figures
`give the size in kb
`
`
`
`J. Meyer and S. Iida: Gene Amplification in E. coli
`
`P1Cm(M)
`
`*
`|j.__J
`
`Pstl
`
`P1Cm(D)
`
`EcoRI_
`'
`__ '
`|_j__L___j_4 PSN
`
`Fig. 4A—C. Monomer, tandem dimer and oligomer structure of Cm transposons revealed by agarose gel electrophoresis. See text for
`detailed explanation. A Bglll digests of Pl and PlCm DNA: (a) P1 DNA. The fragments are numbered to the left according to
`Béichi and Arber (1977); (b) P1Cml3 (M) DNA; (c) P1Cm88 (D) DNA; (d) PlCm88 (A) DNA; (e) PlCm0 (M) DNA; (f) PlCm0
`(D) DNA; (g) PlCm0 (A) DNA; (h) PlCm80 (M) DNA; (i) PlCm80 (A) DNA. B P511 and EcoRl digests of P1 and PlCm DNA.
`(a) and (b) Pstl digests of PlCml3 (M) and PlCm88 (D) DNA.
`(c, d and e) Ec0RI digests of P1Cm13 (M), PlCm88 (D), and
`P1Cm88 (A) DNA. (f and g) Pstl digests of PlCm0 (M) and PlCm0 (D) DNA. (h,
`i and k) EcoRl digests of PlCm0 (M), PlCm0
`(D), and PlCm0 (A) DNA.
`(1) P511 digest of PlCm80 (M) DNA.
`(in to o) EcoRI digests of PlCm80 (M), PlCm80 (A) and P1
`DNA. The numbers to the right identify the Ec0Rl fragments of P1 DNA according to Bachi and Arber (1977). The lines to the
`left give the positions of the “unit length” fragments (corresponding to the length of the Cm’ determinant plus one IS1, see Fig. 4C)
`of the respective PlCm derivatives. C Schematic representation of the EcoRI and Pstl restriction cleavage of monomeric and dimeric
`Cm transposons on DNA ofPlCm13 (M), PlC1n88 (D), PlCm0 (M), and PlCm0 (D)
`
`and g) is about 100 bp larger than that of P1Cml3
`(M). Note that here the “unit length” fragment of
`PlCm0 (D) almost comigrates in the gel with frag-
`ment EcoRI—l4.
`
`2. Level of Resistance to C/iloramphenicol
`of PI Cm Lysogens Containing a Monomer
`or a Dimer Cm Transposon
`
`C) Cm Transposon Tn9 in PlCm0. The Cm transposon
`Tn9 was physically characterized in }tcam phages as
`an IS1-flanked Cm‘ determinant
`(MacHattie and
`Jackowski, 1977) which was derived from phage
`PlCm0 (Gottesman and Rosner, 1975). Phage PlCm0
`originated from growth of phage P1 in cells carrying
`an R plasmid now called pSMl4 (Kondo and Mitsu-
`hashi, 1964). We have determined that the Cm trans-
`poson carried by PlCm0 is integrated into the ISI
`site of P1 DNA (Iida et al., 1978a; Iida and Arber,
`1979; this study) and that by Pstl restriction cleavage
`analysis it has the same size as Tn9 in /lcam (data
`not shown).
`
`PlCm lysogens were grown in the absence of chloram—
`phenicol and then plated on agar containing increas-
`ing concentrations of the drug. The number of co-
`lonies formed was scored after incubation at 30° C
`
`for 48 h and the efficiency of colony formation was
`plotted as a function of the Cm concentration (Fig. 5).
`PlCm lysogens carrying a dimer Cm transposon
`displayed a slightly higher degree of resistance to chlo-
`ramphenicol than those with a monomer Cm transpo-
`son of the same size and inserted at the same location.
`
`Presumably, the increase in gene dosage allows faster
`inactivation of the drug. However, the location and/or
`size of the Cm transposon seem to be additional pa-
`rameters influencing the level of resistance.
`
`
`
`J. Meyer and S. lida: Gene Amplification in E. coli
`
`Table 4. Amplification of Cm transposons in P1Cmlysogens grown
`in 300 ug/ml Cm
`
`Number of phage
`populations
`with amplified Cm
`transposon
`
`13
`6
`1b
`3
`
`Prophage
`
`Number of
`experiments
`
`P1Cml3 (M)
`PlCm88 (D)
`PlCm0 (M)
`PlCm0 (D)
`
`Aberrant (see Discussion)
`Dimer
`
`a b
`
`DNA (Fig. 4C, Fig. 4B, slots a, d, f, i and 1) is generated
`several times from P1Cm(A) DNA. The correspond-
`ing band thus appears much more intense (Fig. 4B,
`slots e, k and n). In BglII digests of PlCm (A) DNA,
`the fragment carrying the multimeric Cm transposon
`is further enlarged. In PlCm88 (A) there are several
`new Bglll-2 bands, some even larger than the Bglll-1
`band (Fig. 4A, slot (1) and the BglII—2 band correspond-
`ing to that of the original PlCm88 (D) has faded.
`In PlCm0 (A) there is also a faint BglII—2 band at
`the position of that of PlCm0 (D) and a larger new
`Bglll-2 band (Fig. 4A, slot g). In P1Cm80 the Cm
`transposon is inserted in the BglII—3 fragment which
`in PlCm80 (M) results in an increase of its size to
`just below that of the BglII-2 fragment (Fig. 4A, slot
`h). In PlCm8O (A) this fragment is further enlarged
`to almost the size of Bg[II—l (Fig. 4A, slot i). Tandem
`oligomerization of the Cm transposons on P1 DNA
`was also confirmed by digestion with BCIWIHI and
`double digestion with BamHI : Pstl and BglII : Pstl re-
`striction endonucleases (data not shown).
`Furthermore, electron microscope analysis of 75
`single—stranded BglII—2 DNA fragments derived from
`a population of PlCm88 (A) carrying amplified Cm
`transposons revealed snapback structures analogous
`to those shown in Fig.2. The length of the loops
`fell into distinct size classes, beginning with classes
`I — III and further increasing by steps of about 1.8 kb
`up to a size of 23 kb (indicating the presence of a
`tandem hexamer, 3 molecules). One molecule had an
`exceptionally large loop of 32 kb (suggesting a tandem
`undecameric Cm transposon). The degree of oligo-
`merisation is higher than indicated by the size of
`the loops in these snapback molecules, and it can
`be determined from the total length of the Bg[II—2
`fragments. The majority of the BgllI—2 fragments had
`a size corresponding to the presence of a trimer, tet—
`ramer and pentamer Cm transposon.
`Thus restriction cleavage and electron microscope
`analysis of PlCm (A) DNA reveal mixtures of oligo-
`
`
`
`EFFICIENCYOFCOLONYFORMATION
`
`2S
`
`50
`
`70
`
`100
`
`140 200
`
`280 400
`
`Cm DOSE [ug/ml]
`Fig. 5. Level of resistance to chloramphenicol of two pairs of PlCm
`lysogens carrying monomer or dimer Cm transposons, compared
`to P1r—detlysogens. The efficiency of colony formation is plotted
`as a function of the chloramphenicol concentration in the plate.
`Lysogens — 1:1 — P1Cml3 (M), — I A P1Cm88 (D), — 0 ~ PlCm0
`(M), — 0 — PlCm0 (D), 4 A 4 Plr—det
`
`An analysis of the Cm transposons in some of
`the PlCm lysogens surviving 200 ug/ml Cm indicated
`that dimer Cm transposons were further amplified.
`Since the amplification of the r—determinant of R plas-
`mids has been observed in Proteus mirabilis but not
`
`the phenomenon of duplication of Cm
`in E. call",
`transposons carried on phage P1 DNA in the E. coli
`host was further studied.
`
`3. Higher Oligomeric Forms of Cm Transposonsk
`
`a) Populations of Increased Resistance to Chloramphe-
`nicol. PlCm lysogens were grown to saturation in
`the presence of 300 ug/ml Cm. Subcultures were
`grown in the absence of the drug, phage P1 was
`induced and the phage DNA analysed by Bglll and
`EcoRI restriction cleavage. Amplification of the Cm
`transposons was frequently observed when PlCm ly-
`sogens with tandem dimer Cm transposons were
`grown in medium containing high concentrations of
`chloramphenicol (Table 4). The phenomenon could
`also be detected in lysogens of PlCm (M), but
`in
`most experiments resistant populations produced Pl
`still carrying the original monomeric Cm transposon.
`The mechanism by which these cells may have ac-
`quired the higher resistance is discussed later.
`Tandem oligomerization of the Cm transposons
`was revealed in restriction cleavage analysis by the
`following observations documentedin Fig. 4. The “unit
`length” fragment produced once by Pstl cleavage of
`PlCm(M) DNA or by Ec0RI cleavage of PlCm(D)
`
`
`
`J. Meyer and S. Iida: Gene Amplification in E. coli
`
`mers of Cm transposons with a variable number of
`units. In many preparations the PlCm (D) carrying
`the original dimer Cm transposon was still observed.
`The plaque forming capacity of the PlCm (A)
`phage particles was tested for a few clones only. The
`results confirm earlier findings (Iida and Arber, 1977;
`Arber et al., 1978; Iida and Arber, 1979) that inser-
`tions in the Pl DNA which are larger than the termi~
`nal redundancy (about 10 kb) decrease the efficiency
`of plaque—formation. However, due to the circular
`permutation of the phage DNA, an oversized genome
`can be reconstituted by recombination after infection
`of a cell by several phage particles, each carrying
`less than a complete genome.
`
`19) Populations of Increased Resistance to Fusidic Acid.
`The R plasmid NR1 carries the Fa resistance determi-
`nant closely linked to and in the same operon as
`the Cm resistance gene (Lane and Chandler, 1977;
`Miki et al., 1978), possibly as overlapping genes (Mar-
`coli, Iida and Bickle, in preparation). All of our PlCm
`studied here also confer resistance to fusidic acid.
`
`In order to determine whether amplification of Cm
`transposons on P1 could be selected for by growing
`lysogens in the presence of high doses of fusidic acid,
`the Fas strain E. coli DB10 (Datta et al.,
`l974) was
`lysogenized with P1Cm88 (D). The resulting strain
`DBl0 (PlCm88 (D)) was grown in LB medium con-
`taining 300 ttg/ml Fa or on LA plates containing
`100 ug/ml Fa. Phages produced from these cultures
`also carried an oligomeric Cm transposon. The result
`was the same as for Cm selected populations. The
`degree of oligomerization of Cm transposons on
`PlCm phage DNA was higher in phages obtained
`from colonies formed on plates with 100 ug/ml Fa
`than in those derived from cells grown in LB medium
`with 300 ug/ml Fa. Plates provide a stronger selection
`than liquid cultures containing equal concentrations
`of Fa (and we obtained similar results with Cm),
`presumably due to a slower diffusion of the drugs.
`
`c) Amplification is Not Induced by the Selective Agent.
`The degree of amplification and the proportion of
`PlCm genomes showing amplification differed from
`experiment to experiment. A typical example is given
`in Fig. 6 (also compare to Fig. 4A, slot d). Three
`single colonies of PlCm88 (D) lysogens were grown
`in the presence of 25 ug/ml Cm and subcultures were
`made in medium containing 300 ug/ml Cm. Phage
`was induced and its DNA analysed as described.
`Judged from the mobility of the Bg[II—2 fragments
`containing the oligomeric Cm transposons, they con-
`tain predominantly trimeric (Fig. 6, slot c), tetrameric
`(slot b) or more than octameric (slot e) Cm transpo-
`sons. However, when several subcultures derived from
`
`Fig. 6. Degree of amplification of the Cm transposon of PlCm88
`(A) shown by agarose gel electrophoresis of Bglll digests.
`(a)
`PlCm88 (D) DNA; (b to e) PlCm88 (A) phage DNA derived
`from different cultures grown in the presence of 300 pg/ml Cm.
`See text for experimental details. Arrows give the positions of
`the Bglll-2 band of P1 (arrow 1) and of PlCm88 (D) (arrow
`2) DNA
`
`the same culture in medium containing 25 ug/ml Cm
`(originating from a single colony) were grown in the
`presence of 300 ug/ml Cm,
`they resulted in PlCm
`with amplified Cm transposons which gave indis-
`tinguishable restriction cleavage patterns
`(typical
`example shown in Fig. 6, slots c and d). These obser-
`vations suggest that the amplification occurs spon-
`taneously and the lysogens carrying amplified Cm
`transposons outgrow those containing non-amplified
`Cm transposons in the selective medium.
`This view is supported by the following experi-
`ment: E. coli DBl0(PlCm88 (D)) grown in 15 pg/ml
`Cm was separately grown in 300 ug/ml Cm and
`300 pg/ml Fa. The lysogens grown in the two different
`selective media produced PlCm with the same degree
`of amplification (data not shown).
`
`d) Stepwise Increase of Cm Concentration. The pro-
`portion of P1 with an amplified Cm transposon exist-
`
`
`
`B
`P1Cm248(0)
`
`'
`
`J
`
`l
`
`l
`
`J
`
`17a
`4
`4
`4
`Fig. 7. A Stability of the multimer Cm transposon on PlCm248
`(A) DNA shown by agarose gel electrophoresis of Bglll digests.
`(a) P1 DNA; (b) PlCm248 (D) DNA (c) PlCm248 (A) DNA
`derived from a population of cells adapted to 1 mg/ml Cm. (d)
`PlCm248 (A) DNA derived from a population of cells grown
`in the absence of Cm for more than 40 generations.
`(e to g)
`PlCm248 (A) DNA derived from cells grown from three individual
`colonies picked from the population of cells analysed in (d). Detec-
`tion of the tandem multimer Cm transposon on PlCm248 (A)
`DNA by agarose gel electrophoresis of EcoRI digests. (h) PlCm248
`(D) DNA. (i) PlCm248 (A) DNA. The intensified bands J and
`17a are labeled according to Arber et al. (1978). B Schematic repre-
`sentation of the EcoRI and Pstl restriction cleavage of PlCm248
`(D) DNA
`
`‘
`
`EcoRl
`Pstl
`
`J. Meyer and S. Iida: Gene Amplification in E. coli
`
`ments (Fig. 7B; Fig. 7A, slots c, and i). We did not
`observe an amplification of a larger monomeric Cm
`transposon on PlCmSm8l-l or of the monomeric
`r-determinant on Plr-det (Arber et al., 1978), even
`after growing lysogens in medium with gradually in-
`creasing doses of Cm or streptomycin up to l or
`2 mg/ml, respectively.
`
`e) Stability 0f0lig0mer Cm Tnmsposons on P1 DNA.
`Single colonies of PlCm0 (D) lysogens formed on
`plates without chloramphenicol were inoculated and
`grown in LB broth. All 25 colonies produced PlCm
`still carrying the original dimer Cm transposon as
`determined by restriction cleavage analysis. Even after
`storage of lysogens in airtight stab bottles containing
`nutrient agar for three years, 8 out of 8 individual
`subclones derived from 4 different stabs produced
`PlCm with the dimeric Cm transposon.
`The stability of amplified Cm transposons was
`examined in lysogens of PlCm248 (A) grown in the
`absence of Cm. Analysis of phage DNA revealed that
`the majority of induced PlCm carried the original
`oligomeric Cm transposon even after growth in the
`absence of Cm for as many as 40 generations
`(Fig. 7A, slot
`(1, compared to slot c). This finding
`was extended by the analysis of single clones derived
`from this population: 7 out of 10 colonies formed
`on LA plates produced PlCm phage with the Cm
`transposon still amplified to the original
`level
`(Fig. 7A, slots e and g), one carried the Cm transpo-
`son amplified to a lower degree (Fig. 7A, slot f), and
`the remaining two colonies did not produce Pl phage
`(and were not further examined). The procedure was
`repeated with independently obtained lysogens of
`PlCm248 (A): 3 out of 10 single colonies produced
`PlCm carrying the unchanged amplified Cm transpo-
`son and the other 7 colonies did not produce Pl
`phage. We conclude therefore that amplified Cm trans-
`posons carried by PlCm in rec+ cells show consider-
`able stability.
`
`Discussion
`
`PlCm derivatives have been isolated which have ac-
`
`quired the r-determinant together with the two flank-
`ing ISl from the R plasmid NR1. By ISl—mediated
`deletion formation smaller transposons evolved com-
`prising the part of the r-determinant carrying genes
`for resistance to chloramphenicol and fusidic acid
`between the two 1S1 elements. Most of the plaque
`forming PlCm phages carry the Cm transposon in
`a monomeric form, some, however, carry a tandem
`dimer of the structure 1S1 — Cm — IS1 — Cm — ISl.
`
`The cells harbouring a PlCm plasmid with a dimer
`
`ing prior to the treatment with high doses of Cm
`is too small to be detected by restriction cleavage
`or electron microscope analysis. It may even be so
`small that one culture step in the presence of 300 ug/
`ml Cm is not sufficient to enrich them to a detectable
`
`level, but cultures with stepwise increasing doses of
`Cm could allow their accumulation. Indeed, we did
`not observe P1Cm8O (M) derivatives with an amplified
`Cm transposon after treatment with 300 pg/ml Cm,
`but we did obtain such phages after gradual increase
`of the Cm concentration to lmg/ml (Fig. 4B, slot 11).
`Phage PlCm248 (D), a plaque forming deletion
`derivative of PlCmSmSuHg8l, carries a large dimeric
`Cm transposon with two EcoRI but no Bglll sites
`per monomer unit (Arber et al., 1978). Amplified
`transposons were detected on PlCm248 (A) obtained
`from cells adapted to increasing concentrations of
`chloramphenicol. Restriction cleavage analysis re-
`vealed a longer BgZII—2 fragment containing the multi-
`meric Cm transposon and two intensified EcoRI frag-
`
`
`
`J. Meyer and S. lida: Gene Amplification in E. coll‘
`
`217
`
`to
`Cm transposon were isolated as those resistant
`the commonly used concentration of 25 pg/ml of chlo-
`ramphenicol (as were those with a monomer), suggest-
`ing that the duplication occurred spontaneously.
`Once dimers of Cm transposons are carried on
`PlCm DNA further amplification to tandem multi-
`mers is readily observed in lysogens grown in the
`presence of high doses of chloramphenicol or fusidic
`acid. Our data bearing on the degree of oligomeriza—
`tion support the concept put forward by Hashimoto
`and Rownd (1975) that a small fraction of cells har-
`bouring P1Cm with a spontaneously amplified Cm
`transposon have a growth advantage and therefore
`accumulate under highly selective conditions. Since
`chloramphenicol is bacteriostatic,
`lysogens carrying
`a nonamplified Cm transposon may also replicate
`after the drug has been inactivated. This results then
`in the mixed populations observed.
`Spontaneous amplification occurs more frequently
`in PlCm carrying a tandem dimer than in those with
`a monomeric Cm transposon. All 4 tandem dimer
`Cm transposons tested were amplifiable to higher oli-
`gomers. The size of the monomer unit varied between
`2.6 kb (PlCm88 (D)) and 4.8 kb (P1Cm248 (D)) and
`they were carried at two different loci of the Pl ge-
`nome (map units 4 and 20).
`On several occasions we detected the duplication
`and amplification of monomeric Cm transposons on
`phage P1 DNA (Fig. 4, Table 4). Amplification of
`the Cm transposon Tn9 carried by plasmids was also
`observed by Chandler et al. (1979) and by R. Mat-
`tes and R. Schmitt (personal communication). Some
`limitations in detecting oligomerisation of Cm trans-
`posons could be due to the fact
`that we analyzed
`phage DNA. All amplifications of Cm transposons
`on PlCm plasmid DNA which abolished lytic phage
`functions would have remained unnoticed in our ex-
`
`perimental procedu