Detection and Characterization of Tn2501, a Transposon Included

JOURNAL OF BACTERIOLOGY, June 1984, p. 866-871

Vol. 158, No. 3

0021-9193/84/060866-06$02.00/0 Copyright © 1984, American Society for Microbiology

Detection and Characterization of Tn2501, a Transposon Included Within the Lactose Transposon Tn951 THOMAS MICHIELS AND GUY CORNELIS* Unite' de Microbiologie, Universite de Louvain, UCL 30.58, B1200 Brussels, Belgium Received 5 December 1983/Accepted 5 March 1984

The DNA sequence spanning coordinates 9.9 to 16.4 kilobases of the lactose transposon Tn951 (Cornelis et al., Mol. Gen. Genet. 160:215-224, 1978) constitutes a transposable element by itself. Unlike Tn951 (Cornelis et al., Mol. Gen. Genet. 184:241-248, 1981), this element, called Tn2501, transposes in the absence of any other transposon. Transposition of Tn2501 proceeds through transient cointegration and duplicates 5 base pairs of host DNA. Tn2501 is flanked by nearly perfect inverted repeats (44 of 48), related to the inverted repeats of Tn2J (Zheng et al., Nucleic Acids Res. 9:6265-6278, 1982). Unlike Tn21, Tn2501 does not confer mercury resistance.

Tn951 is a 16.6-kilobase (kb) transposon encoding lactose fermentation (8). Its lactose operon only includes the lacI, lacZ, and lacY genes aqd is homologous to the lac operon of Escherichia coli K-12 (8), as well as to the lac genes of several plasmids from various origins (6). The homology between Tn951 and the E. coli chromosome is restricted to the 5.7 kb of the lactose operon itself (8). Tn951 also contains a single copy of the IS1 sequence, which was shown to generate deletions and inversions within the transposon (10). The terminal 41 base pairs (bp) of Tn951 constitute a perfect inverted repeat (IR), and the outside 38 bp are identical to the ends of Tn3. Tn951 contains 100 bp corresponding to the C terminus of the transposase gene of Tn3 with two mismatches (11), but the 6.5 kb of DNA between this homologous region and the end of lac Y is not homologous to Tn3. In agreement with these observations, Tn951 does not transpose by itself at a detectable frequency, but it is complemented by the tnpA gene of Tn3 (11). Tn951 was first detected on pGC1, a lactose plasmid that originated in Yersinia enterocolitica 842 (7). Electron microscopy analysis of pGC1 DNA showed that it contained two different transposon-like structures bordered by short IRs. One was 16.6 kb long (Tn951), and the other was about 6 kb long (8). Plasmid pGC1 forms with RP1 transient cointegrates that do not resolve into RP1::Tn951 derivatives (8). This suggests that Tn951 is not the element involved in the formation of these cointegrates (For a review of the transposons related to Tn3, see reference 21.) Moreover, it is unlikely that these cointegrates were generated by the Tnl element present on RP1 since this transposon possesses a very efficient resolution system (30). Thus, there is structural as well as functional evidence that pGC1 contains, besides Tn951, a second transposable element, devoid of a known phenotype. The aim of the present work was to identify and to characterize this second element present on pGC1. We show that it is a transposon that is 6.3 + 0.2 kb long and flanked by IRs related to those of Tn2l. Interestingly, it is contained within Tn951, where it spans the unassigned region located between the tnpA vestige and the lactose operon.

MATERIALS AND METHODS Bacterial strains and plasmids. The E. coli strains used were: JC3272 (his lys trp AlacX74 rpsL) (40), JC6310 (his lys trp AlacX74 recAS6 rpsL) (40), W1177 (thr leu thi lac Y rpsL) (27), 84Nx (met cys trp lacY gyrA) (from P. Fredericq), and KL131 (leu arg met aro recAl lacY) (28) (strain CGSC4332 of the E. coli Genetic Stock Center; received from B. Bachmann). Plasmids are listed in Table 1. Growth of bacteria and genetic transfer. These methods are quoted in Cornelis'et al. (7). Selective agents were tetracycline (10 ,ug/ml), sulfathiazole (250 p.gIml), chloramphenicol (20 ,ugIml), kanamycin (25 ,uglml), nalidixic acid (35 ,ug/ml), and trimethoprim (50 ,ug/ml). Trimethoprim was used in minimal agar plates (31). The other antibiotics were used in MacConkey agar. Preparation and manipulation of DNA. The method of Kado and Liu (24) was used to visualize plasmids and to estimate their size. For rapid restriction analysis, plasmid DNA was prepared as described by Birnboim and Doly (2). For other purposes, plasmid DNA was prepared essentially according to Sharp et al. (35). Gel electrophoresis, nick translation, and Southern blotting were done according to the methods quoted in Cornelis and Saedler (10). DNA-DNA hybridizations were carried out for 48 h in 2x SSC (lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate, pH 7.0)-10x Denhardt solution (Denhardt solution is 0.02% Ficoll, 0.02% bovine serum albumin, and 0.02% polyvinylpyrrolidone)-40% formamide at 40°C. DNA was sequenced by the method of Maxam and Gilbert (29). RESULTS Analysis of the cointegrates formed between pGCl and RP1. Transposition of the Tn3-like elements takes place via a cointegrated intermediate that is subsequently resolved by the action of the tnpR gene product. In spite of their instability, transient cointegrates can be exchanged during conjugation. The Tn3-like transposons thus promote the mobilization of nonconjugative plasmids, and the efficiency of mobilization is an inverse function of the resolution efficiency (see reference 21 for a review). The mobilization of plasmids lacking mobilization genetic determinants thus allows selection for transposition of elements devoid of a selectable phenotype. This strategy was

* Corresponding author. 866

VOL. 158, 1984

Tn250/ CONTAINED WITHIN Tn951

according to the element that was transposed before or during the first cross (Fig. 1). In the first conjugation, Lac' Kmr recombinants occurred at a frequency of 10' per Kmr transconjugant. Seventy-four such Lac' Kmr recombinants were crossed with the Nalr strain 84Nx, and Kmr Nalr recombinants were selected. Of the 74 clones, 11 gave 100% Lac' colonies among the Kmr Nalr transconjugants, suggesting that they contained an

TABLE 1. Plasmids used Plasmid

Relevant markersa

Origin (reference)

RP1 Ap Km Tc Tra+ pGC1 Lac' Tra+(Ts) R388 Tp Su Tra+ pACYC184 Cm Tc MobpBR322 Ap Tc MobRSF1050 Ap pJOE105 Ap Tc pUB781 Hg pGC704 Sm Su Lac' pGC9114 Ap Km Tc Lac' Tra+ pGC823 Tc Lac' Rep(Ts) a Abbreviations are those recommended

(18) (7) (12) (5) (3) pMB8::Tn3 (22) pJOE100::Tnl721 (34) ColE1::TnS01 (1) RSF1010-4::Tn951 (11) RP1::Tn951 (8) pPM103::Tn951 (11) by Novick et al. (32).

RP1::Tn951 derivative. The other 63 clones gave less than 1.5% Lac' colonies among their transconjugants. These clones presumably arose by transfer of transient cointegrates formed between pGC1 and RP1 and mediated by TnJ or by the 6-kb transposon. Among these 63 clones, 48 gave less than 0.1% Lac' colonies among their Kmr Nalr transconjugants, and 15 clones gave between 0.1% and 1.5% Lac' colonies among their transconjugants. Since resolution of Tnl/Tn3-mediated cointegrates is highly efficient (21), we decided to analyze clones among this latter group of 15 to isolate RP1 derivatives carrying the 6-kb sequence. For the sake of clarity, we shall call their plasmid RP1*. Plasmid DNA of 5 clones out of the 15 was analyzed by a rapid agarose gel electrophoresis procedure (data not shown). The size of RP1* was slightly larger than that of RP1 itself, suggesting that RP1* gained a DNA sequence, as was expected. Plasmid DNA from two different RP1* derivatives was purified and analyzed by restriction (data not shown). In both cases, the inserted sequence was about 6 kb long and contained three identical internal SmaI fragments (0.25, 0.7 and 2.6 kb). The insertion sites were different in both cases. We called the transposed sequence Tn2501 (26). Tn2501 encodes its transposition functions. The mobilization strategy was also used to transpose Tn2501 onto pACYC184: strain JC6310(RP1: :Tn2501)(pACYC184) was mated with strain 84Nx, and selection was applied for

first introduced by Guyer (19) to show the transposable character of gamma-delta. We used the same strategy to analyze the presumptive 6kb transposon present on pGC1. E. coli W1177(RP1)(pGC1) was grown and mated at 37°C (nonpermissive temperature for pGC1 transfer) with strain JC3272. Selection was applied for lactose utilization and kanamycin resistance. This was shown to select for transposition of Tn951 onto RP1 (8), in spite of a possible immunity barrier (21) due to the presence of Tnl on the recipient replicon. This could also select for cointegration mediated by Tnl, Tn951, or the presumptive 6kb transposon sequence. Transferred cointegrates are expected to resolve in the transconjugant of this first mating. To sort out the transpositional event that occurred in this first step, the various JC3272 transconjugants were purified and mated with E. coli 84Nx. Selection was only applied for kanamycin and nalidixic acid resistance. The lactose phenotype of the transconjugant of the second cross will differ Donor cell

I

m

867

Lac' Km R transcon jugant -

E

D

KmR transconjugant

Structure -

Phenotype

-

E

Lac+

Lac

,II1

4

.11m

* RESOLUTION

f-I F""111NI ll

m

-

XIII

Ev~

---

_ ~~~~~Ir It ZI, )

C jIID ,

[m

Lac'

-1

Lac'

I

| RESOLUTION

ImID

Lac'

Lac'

LIImI

Lac-

FIG. 1. Strategy used to detect the presumptive 6-kb transposable element present on pGC1. The circles represent RP1, and the squares represent pGC1. The hatched lines represent Tn951, and the dotted lines represent the presumptive transposable sequence. Three different types of events are presented: transposition of Tn951 (I), cointegration mediated by Tn951 (II), and cointegration mediated by the 6-kb sequence (III). Horizontal arrows indicate transfer by conjugation, whereas vertical arrows indicate the resolution of the structure within the same host strain. Thick arrows represent the most probable event, and dashed arrows represent rare events. The lactose phenotype of the second transconjugant is indicated in the last column.

MICHIELS AND CORNELIS

868

J. BACTERIOL.

TA BLE 2. Tn2501-mediated mobilization of pACYC184a

Frequency of mobilization of Donor strain pACYC184 KL131(1pACYC184)(RP1) .1 X 10-6 KL131(1pACYC184::Tn2501)(RP1) ............... 2 x 10-4 KL131(1pACYC184)(R388) ......................

(U

i

l

"\

/

'

/

0

c

WITHIN

10.6

/* t III~~~~

I>

,,

u-

FIG. 4. Restriction pBR322::Tn2501 derivative.

map

of

Tn2501

and

Digestions with

Hinfl,

SalI-PvuIl

sequencing. HindlIl; Pvu, PvuII; of

The

thick

Sal,

line

Sall;

used

AluI,

Tn2501

the

HpaII indicated that the insertion occurred between 430-bp

coordinates

2449

and

2642

Hinfl-Hinfl

3' (*). DNA, and the black boxes represents the termini. Bam,

the

I

to

and

and

and on

represents

Sma,

I

C

strategy

of the pBR322 sequence (37). The 600-bp 5' end by using the polynucleotide kinase (0) tion

r

II C

a.

end

BamHI;

Eco,

EcoRI;

Hind,

SmaI.

Lac-

12th clone were Kms. Plasmid DNA different transformants was subjected to SmaI (data not shown). In all 12 cases, pACYC184 three fragments (0.25, 0.7, and 2.6 and Tn951. This restriction, as well as BamHI, digests (data not shown), thus confirms

from

(

the

(left

side) and

a

TACTA

digestion

acquired

kb)

Tn2501

common

DISCUSSIC

BstEII,

PvuII

that

pG

mobiliza-

tion of pACYC184 by pGC9114 was the transposition of Tn2501 (1 clone) or of Tn951 clones). is thus entirely included within Tn951. is related to Tn2l. The termini of Tn250O sequenced. The map of a derivative for DNA sequencing and the sequencing strategy in Fig. 4. For the left terminus, both strands quenced. For the right terminus, only sequenced because this sequence was already earlier while sequencing the termini of Tn95J sequence B of the termini of (Fig. 5) with four mismatches. Interestingly, the these IRs are clearly related to the termini of 38 bp). The sequence analysis showed that ed into pBR322 between nucleotides 2498 and insertion did not cause any deletion and duplicated nucleotides 2494 through 2498 of pBR322. result (11

Tn2501 Tn2501

pBR322::Tn2501

JC1, the original vector of

37°C,

at in a process that n951. They also observed i of about 6 kb flanked by observatioiins are explained by the ion of pGC1 involves the Tn250O, 0.2 kb flanked elemer nt of 6.3

Tn951,

were

±

are

contair n resistance genes, but its b)y the mobilization that it

were

one

strand

for mobilization, only which suggests that the o represents an intermetransposition products. C 184 onto R388 and from no resident transposfar on these plasmids, we Tn25( encodes its own trans-

rn observe(d,

determined

Tn2501

(11).

form

ted

48-bp

external

Tn2501

Tn2J

2499

Tn2501 Tn951

Relation between and Tn25O1 region of Tn9SJ spanning coordinates 9.9 to right side of Tn25OJ starts at nucleotide which corresponds exactly to the point where

Tn951

of Tn9SJ diverges from the

Tn2501 EcoRI-SalI

sequence of

Tn3

ince 01

(37).

occupies

16.4

int(

(41)

Tn25O1

nearly perfect IRs (44 of related to the IR flanking (21). In agreement with dupliccates 5 bp of target DNA iat Tn25O1 transposition mobiilization of plasmids sugI transient cointethrough ahso functionally related to

48-bp-long,

kb.

137

the

(11).

We sequenced the first 100 bp of the between Tn95J and For this sequence used the 1.5-kb fragment spanning 8.5 to 10.0 kb of Tn95J (11). This sequence homologous neither to the sequence of Tn3 sequence of lacY (4). Within Tn951, Tn25O1

other

junction

analysis,

coordinates

(not

(23)

Tn250

1

5'

of

5'

L

........T.A.........

......

sequence

Tn2501

of the IRs of been

replaced

T....C

GT.TCA .A..AAG.........

G.AC.TCA

R

FIG. 5. Nucleotide of the IRs of Tn25OI (left) have

10-4,

GGGGTCCGCCTGGAAAACGGAAATTATCCCACGCTAAGACTGTTTTTT R

Tn2l

is in the Tn2501-promo tednatelymobilization the transposition

nor

appears

......

Tn25O1

Tn2l

and comparison

44/ 4 8

.........

35/ 3 8

A..AAG.

29

/3 8

Tn21

those

MICHIELS AND CORNELIS

870

J . BACTERIOL .

Tn 951

A

tr... _0

.. .

_

.E

.

LA

C

LO

Li

C%J

()

.

am

_

m I- CL-. 4n

co

)

oC) "OCD "LActo c wen c ,

C 0Cv)D0

cooo-

V)

LACD

I

COLD

_

._

I

.

0

mn

C

Isi

I

Lac

Z

Tn 2501

y

IR

B

Tn951

(LEFT)

5

-

R

Tn2501

......CGGCATAAGCCTGCTTIAAAAAACAGGCTTAACGTGGGATA

TTTTCCGTTTTCCAAGCGGACCCCTATCAATATGCTCGGCCATTATTCCTT -ACCAAAGGAC

I

Iso c

-

TERMINUS

1 30

14 0

15

tnp A

AT CT GAG A CC AT T AA AA GAG 710 80

1'20

110o9go I R -R Tn95 1

(100 bp)

G C G T CA GA 60

G T AG A A A AC GF T G CT T A AC G T GAG T TT

50

410

310

T T CC ACT G AG C G T C AG A C C C C

210

110

1

FIG. 6. Structure of Tn951. (A) Genetic and restriction map. This map shows the position of the BamHI (Bam), BstEII (Bst), EcoRI (Eco), HindlIl (Hind), Pstl (Pst), Sall (Sal), and SmaI (Sma) sites within Tn951. Only Pvull (Pvu) sites present on Tn2501 are shown (sites outside Tn2501 are not given). The position of ISI (10) and the position and orientation of the lac operon (6) are also given. The localization of Tn2501 is discussed in the present paper. The coordinates of the restriction cuts are Tn951 coordinates (kb) (9). (B) Sequence of the last 200 nucleotides of Tn951 (right side) showing the insertion of Tn2501 into the tnpA gene of Tn951. This sequence was presented in reference 11. Coordinates are in bp, starting from the right end of Tn951. These coordinates correspond to those of Tn3 (starting from the left end) up to nucleotide 136 (23). Tn2501 starts at nucleotide 137. The IRs of the two elements are boxed.

frequency of the Tn3-like transposons. It was not feasible to measure the transposition frequency directly since Tn2501 lacks a selectable marker. However, the high mobilization frequency suggests that the ratio between mobilization and transposition frequencies must be quite high. This feature is more reminiscent of TnS01 (36) and Tn2l (13) than of Tn3. The sequence of the IR is also closer (29 of 38) to the IR of Tn2l (41) than to the IR of Tn3 (22 of 38) (33). Tn2501 is probably a new representative of the TnSO/TnJ721/Tn2J subgroup (15, 17). However, Tn2501 does not hybridize with TnS01 and TnJ721 under stringent conditions, and it does not confer mercury resistance. Complementation experiments, presently being carried out, should shed light on the relations between Tn2501 and this subgroup of elements. It is already clear from previous observations on Tn951 that Tn2501 does not complement the transposition of Tn3 (11). The fact that Tn2501 is contained within Tn95J is consistent with previous structural and functional observations concerning the latter transposon. In particular, the sequence of Tn2501 in Tn951 starts exactly at the point where the sequences of Tn951 and Tn3 have been shown to diverge (11). This clearly suggests that the tnpA gene of Tn9SJ was inactivated by the insertion of Tn2501. The fact that the copy of Tn2501 present in Tn951 is not flanked by a 5-bp direct repeat indicates that Tn2501 generated a deletion within Tn951, in the direction of the lactose operon (38). This deletion presumably removed all of the transposition region of the original lactose transposon since the sequence immediately flanking Tn2501 at the left-hand side is not homoloTn3. The discovery of Tn2501 explains

gous to

a

paradox: Tn9OM does

not contain DNA homologous to the tnpR gene of Tn3, and it does not repress expression of the tnpA gene of Tn3 (11). Moreover, it does not hybridize (G. Cornelis, unpublished observation) with DNA of the PvuII-BamHI fragment containing the resolution site of Tn3 (16). Thus, Tn951 seems to be completely devoid of a Tn3-like resolution system. However, complementation of Tn951 by a tnpR mutant of Tn3 did not result in unresolved cointegrates (M. Van Bouchaute and G. Cornelis, unpublished observation). The presence of Tn2501 explains this paradox, since Tn2501 certainly encodes a functional resolution system and thus provides such a system to Tn951. Thus, Tn951 is a transposon harboring another transposon. This is not unprecedented in the Tn3 family. In particular, it is the case with Tn4, a transposon consisting of Tn3 inserted into the mercury resistance gene of Tn2J (25). Similarly, Grinsted and Brown (SGM Meet., 98th, Leeds, England, abstr. no. 1400, p. M6, 1983) recently showed that TnS01 contains a sequence identical to the IR of Tn2J, about 50 bp from its own left IR, which leads to the speculation that TnS01 arose by the transposition of a Tn2J-like structure encoding mercury resistance, within a Tnl721-like element. Transposons of the Tn3 family thus appear to evolve frequently by successive insertions of related elements. ACKNOWLEDGMENTS We are indebted to M. Van Bouchaute for expert technical assistance. This work was supported by a Credit aux Chercheurs of the Belgian Fonds National de la Recherche Scientifique to G.C.

VOL. 158, 1984

Tn2501 CONTAINED WITHIN Tn951 LITERATURE CITED

1. Bennett, P. M., J. Grinsted, C. L. Choi, and M. H. Richmond. 1978. Characterization of TnS01, a transposon determining resistance to mercuric ions. Mol. Gen. Genet. 159:101-106. 2. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523. 3. Bolivar, F., R. L. Rodriguez, P. J. Greene, M. C. Betlach, H. L. Heineker, and H. W. Boyer. 1977. Construction and characterization of amplifiable multicopy DNA cloning vehicles. II. A multipurpose cloning system. Gene 2:95-113. 4. Buchel, D. E., B. Groneborn, and B. Muller-Hill. 1980. Sequence of the lactose permease gene. Nature (London) 283:541545. 5. Chang, A. C. Y., and S. N. Cohen. 1978. Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J. Bacteriol. 134:1141-1156. 6. Cornelis, G. 1981. Sequence relationships between plasmids carrying genes for lactose utilisation. J. Gen. Microbiol. 124:9197. 7. Cornelis, G., P. M. Bennett, and J. Grinsted. 1976. Properties of pGC1, a lac plasmid originating in Yersinia enterocolitica 842. J. Bacteriol. 127:1058-1062. 8. Cornelis, G., D. Ghosal, and H. Saedler. 1978. Tn9OM: a new transposon carrying a lactose operon. Mol. Gen. Genet. 160:215-224. 9. Cornelis, G., D. Ghosal, and H. Saedler. 1979. Multiple integration sites for the lactose transposon Tn951 on plasmid RP1 and establishment of a coordinate system for Tn9OM. Mol. Gen. Genet. 168:61-67. 10. Cornelis, G., and H. Saedler. 1980. Deletions and an inversion induced by a resident IS1 of the lactose transposon Tn9OM. Mol. Gen. Genet. 178:367-374. 11. Cornelis, G., H. Sommer, and H. Saedler. 1981. Transposon Tn951 (TnLac) is defective and related to Tn3. Mol. Gen. Genet. 184:241-248. 12. Datta, N., and R. W. Hedges. 1972. Trimethoprim resistance confered by plasmids in Enterobacteriaceae. J. Gen. Microbiol. 72:349-356. 13. de la Cruz, F., and J. Grinsted. 1982. Genetic and molecular characterization of Tn2l, a multiple resistance transposon from R100.1. J. Bacteriol. 151:222-228. 14. Depicker, A., M. De Block, D. Inze, M. Van Montagu, and J. Schell. 1980. IS-like element IS8 in RP4 plasmid and its involvement in cointegration. Gene 10:329-338. 15. Diver, W. P., J. Grinsted, D. C. Fritzinger, N. L. Brown, J. Altenbuchner, P. Rogowsky, and R. Schmitt. 1983. DNA sequences of and complementation by the tnpR genes of Tn2l, TnS01 and Tn1721. Mol. Gen. Genet. 191:189-193. 16. Grindley, N. D. F., M. R. Lauth, R. G. Wells, R. J. Wityk, J. J. Salvo, and R. R. Reed. 1982. Transposon-mediated site-specific recombination: identification of three binding sites for resolvase at the res sites of -y& and Tn3. Cell 30:19-27. 17. Grinsted, J., F. de la Cruz, J. Altenbuchner, and R. Schmitt. 1982. Complementation of transposition of tnpA mutants of Tn3, Tn2l, Tn501 and Tnl721. Plasmid 8:276-286. 18. Grinsted, J., J. R. Saunders, L. C. Ingram, R. B. Sykes, and M. H. Richmond. 1972. Properties of an R factor which originated in Pseudomonas aeruginosa 1822. J. Bacteriol. 110:529-537. 19. Guyer, M. S. 1978. The -yb sequence of F is an insertion sequence. J. Mol. Biol. 126:347-365. 20. Hedges, R. W., and A. E. Jacob. 1974. Transposition of ampicil-

21. 22. 23. 24. 25.

26. 27.

28.

29. 30.

31. 32. 33.

34.

35.

36. 37. 38.

39. 40.

41.

871

lin resistance from RP4 to other replicons. Mol. Gen. Genet. 132:31-40. Heffron, F. 1983. Tn3 and its relatives, p. 223-260 In J. A. Shapiro (ed.), Mobile genetic elements. Academic Press, Inc., New York. Heffron, F., P. Bedinger, J. J. Champoux, and S. Falkow. 1977. Deletions affecting the transposition of an antibiotic resistance gene. Proc. Natl. Acad. Sci. U.S.A. 74:702-706. Heffron, F., B. McCarty, H. Ohtsubo, and E. Ohtsubo. 1979. DNA sequence analysis of the transposon Tn3: three genes and three sites involved in transposition of Tn3. Cell 18:1153-1163. Kado, C. I., and S.-T. Liu. 1981. Rapid procedure for detection and isolation of large and small plasmids. J. Bacteriol. 145:13651373. Kopecko, D. J., J. Brevet, and S. N. Cohen. 1976. Involvement of multiple translocating DNA segments and recombinational hotspots in the structural evolution of bacterial plasmids. J. Mol. Biol. 108:333-360. Lederberg, E. 1981. Plasmid reference center registry of transposon Tn allocation through July 1981. Gene 16:59-61. Lederberg, J. 1951. Prevalence of Escherichia coli strains exhibiting genetic recombination. Science 114:68-69. Low, B. 1973. Rapid mapping of conditional and auxotrophic mutations in Escherichia coli K-12. J. Bacteriol. 113:798-812. Maxam, A. M., and W. Gilbert. 1980. Sequencing end-labelled DNA with base-specific chemical cleavages. Methods Enzymol. 65:499-559. McCormick, M., W. Wishart, H. Ohtsubo, F. Heffron, and E. Ohtsubo. 1981. Plasmid cointegrates and their resolution mediated by transposon Tn3 mutants. Gene 15:103-118. Miller, J. H. 1972. Experiments in molecular genetics, p. 432. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Novick, R. P., R. C. Clowes, S. N. Cohen, R. Curtiss III, N. Datta, and S. Falkow. 1976. Uniform nomenclature for bacterial plasmids: a proposal. Bacteriol. Rev. 40:168-189. Ohtsubo, H., H. Ohmori, and E. Ohtsubo. 1979. Nucleotide sequence analysis of the ampicillin resistance transposon Tn3: implications for insertion and deletion. Cold Spring Harbor Symp. Quant. Biol. 43:1269-1278. Schoffl, F., W. Arnold, A. Pikhler, J. Altenbuchner, and R. Schmitt. 1981. The tetracycline resistance transposons Tnl721 and Tnl771 have three 38-base-pair repeats and generate fivebase-pair direct repeats. Mol. Gen. Genet. 181:87-94. Sharp, P. A., M. T. Hsu, E. Ohtsubo, and N. Davidson. 1972. Electron microscope heteroduplex studies of sequence relations among plasmids of Escherichia coli. I. Structure of F-prime factors. J. Mol. Biol. 71:471-497. Sherratt, D., A. Arthur, and M. Burke. 1981. Transposonspecified site-specific recombination systems. Cold Spring Harbor'Symp. Quant. Biol. 45:275-281. Sutcliffe, J. G. 1979. Complete nucleotide sequence of the Escherichia coli plasmid pBR322. Cold Spring Harbor Symp. Quant. Biol. 43:77-90. Tu, C.-P., and S. N. Cohen. 1980. Effect of DNA sequences adjacent to the termini of Tn3 on sequential translocation. Mol. Gen. Genet. 177:597-601. Warren, G., and D. Sherratt. 1977. Complementation of transfer deficient ColEl mutants. Mol. Gen. Genet. 151:197-201. Willetts, N. S. 1975. Recombination and the Escherichia coli K12 sex factor F. J. Bacteriol. 121:36-43. Zheng, Z. X., M. Chandler, R. Hipskind, M. Clerget, and L. Caro. 1982. Dissection of the r-determinant of plasmid R100-1: the sequence of the extremities of Tn2l. Nucleic Acids Res. 9:6265-6278.