Vol. 122, No. 1 Printed in U.S.A.
JOURNAL OF BACTERIOLOGY, Apr. 1975, p. 250-256 Copyright © 1975 American Society for Microbiology
Origin of the TEM Beta-Lactamase Gene Found
FRED HEFFRON,* R. SUBLETT,' R. W. HEDGES, ALAN JACOB, AND STANLEY FALKOW Department of Microbiology, University of Washington, Seattle, Washington 98195,* and the Department of Bacteriology, Royal Postgraduate Medical School, London, England Received for publication 4 November 1974
A sequence of deoxyribonucleic acid of 2.7 x 106 to 3.3 x 106 daltons which includes the TEM f,-lactamase gene is present on the small plasmid RSF1030 (R-Amp). This same sequence is present on plasmid derivatives that have received a translocation of deoxyribonucleic acid specifying the TEM 3-lactamase and is also present on naturally occurring plasmids of the F,, F,,, N, X, 0, I, C, and W incompatibility groups that specify the TEM f-lactamase. It is not found on plasmids of the F1, F, N, X, 0, I, C, and W incompatability groups that do not specify ampicillin resistance or specify 0-type f-lactamases.
Plasmid-mediated resistance to penicillins have shown translocation of TEM from RP4 to and cephalosporins is almost universally associ- the I compatibility plasmid R64. Richmond and ated with the elaboration of a f-lactamase. Two Sykes (25) later showed translocation of TEM general classes of drug resistance (R)-factor- from RP1 to the Escherichia coli chromosome. mediated f3-lactamase are now recognized (14). Recently, Hedges and Jacob (22) isolated a The so-called TEM-type f-lactamase (TEM) series of plasmid derivatives that received TEM has a high absolute activity of 60 to 1700 mU of resistance from RP4. These plasmids showed an f3-lactamase per 109 R+ bacteria gainst benzyl increase in molecular weight accompanying the penicillin. Relative to its activity on benzyl change to ampicillin resistance and, in turn, penicillin, TEM has high activity on ampicillin were able to translocate TEM resistance to and cephaloridine substrates but low activity other plasmids. The translocatable sequence(s) against isoazolyl penicillins such as oxacillin of DNA that was transmitted from RP4 to the and methacillin. TEM is the most common derivative plasmids and that must include the R factor-mediated ,B-lactamase and is found on structural information for the TEM fl-lactaa wide variety of naturally occurring plasmids mase was termed transposon A (Tn A) (22). In (21). Surprisingly, regardless of the species of the present study we show that a sequence of origin or the geographical source of the R factor, DNA, 2.7 x 106 to 3.3 x 106 daltons in size the TEM protein has been found to be similar (corresponding to Tn A), is common to natu(21). The other type of R factor-mediated f-lac- rally occurring plasmids of a variety of tamase, the so-called oxacillin-hydrolyzing en- compatability groups that specify TEMzyme (0), can be differentiated from TEM by mediated ampicillin resistance. its lower absolute activity against benzyl peniMATERIALS AND METHODS cillin and its ability to hydrolyze oxacillin at an appreciable rate (27). Although considerable Bacterial strains and plasmids. All bacteria used heterogeneity exists in the molar substrate pro- in this study were derivatives of E. coli K-12. They file of different 0-type ,B-lactamases isolated were: J53 (pro-, met), J62 (lac-, try-, leu-, thi ), from different R+ strains (21), these enzymes C600 (lac-, thr-, leu-, thi-), and 1485-1 (NalH, thy-, can be reasonably subdivided into two broad F-) (6). Many of the plasmids used were kindly to us by Naomi Datta. The molecular groups on the basis of their enzymatic activity supplied properties and origin of the plasmids employed in this against methicillin and their molecular weights study are shown in Table 1. (9).
Evidence has accumulated indicating that the TEM ,B-lactamase gene is sometimes situated on a sequence of deoxyribonucleic acid (DNA) that is capable of being translocated from one replicon to another. Datta et al. (13) 'Present address: Department of Microbiology, University of California, Los Angeles, Calif. 90024.
Hydroxylamine assays of ,B-lactamase activity. Hydroxylamine assays of ,B-lactamase activity were carried out by the method of Hedges et al. (21). Preparation of labeled plasmid DNA. 3H-labeled RSF1030 DNA was prepared from E. coli 1485. The bacteria were grown to about 109/ml in 121 media (17) with 0.8 gg of [3H Ithymine (New England Nuclear, 50 Ci/mmol) per 25 ml of culture and 1 gg of unlabeled thymine per ml. The cells were harvested by centrifu-
VOL. 122, 1975
TEM BETA-LACTAMASE GENE
TABLE 1. Molecular properties and origin of plasmids used in this study Resis-
G+Ca Size 106 (Mol (x dal%) tons)
388 R7K Sa Sa-1 906
W W W W P
62 62 62 62 ?
21 22 23 26 ?
Su, Tp Ap, Sm Cm, Km, Sm, Su Cm, Km, Sm, Su, Ap Ap, Cm, Su
RP4 R455 F Rl 222 R6K R16 R656a R648 R64 R66a R483 R64-1 R66a-1 R483-1
34 ? 70 62 68 26 48 ? ? 72 57 55 74 61
Ap, Tc, Km Ap, Sm, Tc, Cm, Su
I, Id I1 I,
59 ? 49 52 58 45 50 ? 50 50 50 ? 50 50 ?
Ap, Cm, Su, Gk Ap, Tc, Km, Su
Sm, Tc, Su
F, F, F, F, X 0
N 49 N ? Two plasmids (55 x 106 and 5.5 x 106) 48.5 RSF1030 Not N3
Origin of resistance
Escherichia coli Proteus rettgeri Shigella sp. Escherichia coli Bordetella bronchiseptica
Pseudomonas aeruginosa Proteus morganii E. coli IAp, Sm, Cm, Km, Su Salmonella paratvphi Sm, Su, Cm, Tc Shigella flexneri E. coli Ap, Sm E. coli Ap, Sm, Tc, Su Ap, Sm, Km, Su Salmonella typhimurium S. typhimurium Ap, Sm, Km S. typhimurium Tc, Sm Klebsiella pneumoniae Sm, Km E. coli Sm, Tp E. coli Ap, Tc, Sm E. coli Ap, Sm, Km E. coli Ap, Sm, Tp
U.K. Greece Japan
11 7 32 22 R. W. Hedges, A. E. Jacobs, and J. T. Smith, J. Gen. Microbiol., in press U.K. 13 S. Africa 20
Japan Greece Mexico S. Africa
U.K. U.K. U.S. U.K.
23 16 19 19 24 12 10 22 22 22
K. pneumoniae Providencia sp.
Japan S. Africa France
34 7 3
Ap, Sm, Tc, Cm, Su P. rettgeri S. panama Ap
By transformation from
Guanine plus cytosine. 'Su, Sulfonamide; Tp, trimethoprim; Ap, ampicillin; Sm, streptomycin; Cm. chloramphenicol; Km, kanamycin; Tc, tetracycline. a
gation and lysed by the Brij lysis procedure of Clewell and Helinski (5). Plasmid DNA was purified further by isopycnic centrifugation in a CsCl/ethidium bromide (EtBr) gradient (S. Falkow, P. Guerry, R. Hedges, and N. Datta, J. Gen. Microbiol., in press). The density was adjusted to 1.625 g/cm3 (refractive index, 1.3925), and the solution was centrifuged for 40 to 48 h at 15 C and 40,000 rpm (96,600 x g) in a Beckman type 50 fixed-angle rotor. The dense peak of covalently closed plasmid DNA was removed by fractionation of the gradient, the EtBr was extracted with CsCl-saturated isopropanol, and the DNA was dialyzed against 0.42 M NaCl. Extraction of unlabeled DNA for hybridization. Unlabeled whole-cell DNA from plasmid-containing strains was prepared by the method of So et al. (28). DNA-DNA duplex studies. Labeled and unlabeled DNA were sheared by sonic treatment prior to hybridization in a Branson model 140D sonic oscillator for 5 min at 4 C (output set at 3) and then dialyzed against 0.42 M NaCl. Hybridizations were carried out as described by Crosa et al. (8). Approximately 0.001 zg of 3H-labeled, sheared RSF1030 DNA (specific
activity, 1.5 x 106 counts/min per gg) was mixed in a glass vial with 150 ,gg of unlabeled whole-cell DNA in a total volume of 1 ml of 0.21 M NaCl. The DNA was denatured by placing it in a boiling-water bath for 10 min, and renaturation was immediately started by transfer of the vials to a 70 C bath. The DNA was allowed to reanneal for 50 min. Renaturation was stopped by placing the vials into a dry ice/ethanol bath. The S1 endonuclease reaction was carried out as described by Crosa et al. (8). The actual degree of duplex formation for the homologous reaction by the Si endonuclease assay was 81 to 86%, whereas the actual degree of duplex formation with the chromosomal control was 7 to 8%. Normalized values (recorded in the tables as percent homology) were obtained from raw data by subtracting the chromosomal control (7 to 8%) and dividing by the homologous reaction (81 to 86%). Analysis of heteroduplexes by electron microscopy. DNA-DNA heteroduplexes were prepared by the method of Davis et al. (15). Bacteria containing the appropriate plasmid were grown to stationary phase in 1 liter of antibiotic-supplemented
HEFFRON ET AL.
brain heart infusion broth. A Brij lysate was prepared and banded to equilibrium in CsCl/EtBr as described above, and the plasmid peak (about 50 Mg) was removed. This plasmid DNA was randomly nicked in a single strand by treating it with 5,000 to 6,000 rads of radiation (for a plasmid of 22 x 106 to 27 x 106 daltons) or 11,000 to 13,000 rads of radiation (for a plasmid of 5 x 106 to 6 x 106 daltons) while it was in
CsCl/EtBr containing 0.002 M histidine. The DNA rebanded in a second CsCl/EtBr gradient, and the nicked plasmid DNA was removed and used for heteroduplex analysis. Heteroduplex DNA molecules were obtained by mixing 0.1 to 0.2 ug of each X-ray-nicked plasmid DNA in 0.25 ml of a solution containing 0.1 N NaOH and 0.01 M ethylenediaminetetraacetate (pH 12.4) for 10 min and then reneutralizing to pH 8.5 with 2 M tris(hydroxymethyl)aminomethane (pH 7.1). The DNA was reannealed by adding 0.25 ml of formamide (Mallinckrodt) and allowing it to stand for 3 h at 25 C. Reannealed DNA (100 Ml) was spread from a solution of 40% formamide, 0.1 M tris(hydroxymethyl)aminomethane, 0.01 M ethylenediaminetetraacetic (pH 8.5), and 0.1 mg of cytochrome c per ml at the air-water interface on a glass slide onto a hypophase of a solution containing 10% formamide, 0.01 M tris(hydroxymethyl)aminomethane, 0.001 M ethylenediaminetetraacetic acid (pH 8.5) and was then picked up from the surface of the hypophase on a parlodion-coated electron microscope grid. The grid was dipped in freshly prepared 5 x 10-I M uranyl acetate in 90% ethanol for 30 s and in isopentane for 10 s. Grids were placed 0.75 cm below and 7.5 cm away from the bottom of a tungsten wire basket (and rotary shadowed with 2 cm of platinum palladium 23-gauge 80:20 wire an angle of about 7°). In some cases, OX 174 DNA was added as a single-strand length standard. Grids were examined with a JEOL model 100B electron was
DNA contour length. DNA contour length of plasmids was measured by spreading DNA according to the Kleinschmidt technique (15). Well-isolated molecules were photographed under an electron microscope, and the developed negatives were projected through a photographic enlarger onto paper and traced in pencil. The contour length of the tracing was determined by laying silk suture material on the tracings. True magnification was determined by calibration with a diffraction grating. RESULTS
The R-factor R6K is unusual in that it is self-transmissible (conjugative) and normally present within host bacterial cells in multiple copies. R6K has been placed in the X incompatibility group and has been the sole member of this class. Several naturally occurring R plasmids were examined at the Royal Postgraduate Medical School which, at first, appeared to be incompatible with R6K and possible candidates for inclusion in the X incompatibility class. R6K DNA was employed in DNA-DNA duplex
studies, and one of the putative X plasmids, Rlll (3), which conferred resistance to ampicillin was found to share some 27% sequences in common with R6K. When the strain harboring the Rlll plasmid was examined further, it was found that two distinct circular species were present-a transmissible plasmid of about 55 x 106 daltons and a smaller nonconjugative plasmid of about 5.5 x 106 daltons. Both species of Rlll plasmid DNA were employed to transform E. coli with selection for ampicillin resistance. Ampicillin-resistant transformants showed only the smaller plasmid species. This plasmid of 5.5 x 106 to 5.6 x 106 daltons, which we now designate RSF1030 (referred to hereafter as R-Amp for convenience), is found in host cells as a multicopy pool of about 30 copies per chromosome equivalent (J. Crosa, F. Heffron, and S. Falkow, manuscript in preparation), and specifies a f-lactamase with a molar substrate profile of a TEM ,B-lactamase. (The f-lactamase specified by R-Amp showed a higher absolute activity against benzyl penicillin than the TEM f-lactamase from R6K. It showed 102% of this activity against ampicillin and essentially no activity against methicillin.) Further DNA-DNA duplex studies revealed that all of the sequences in common detected between R6K and Rl1 were found on R-Amp (the 5.5 x 106-dalton derivative of Rlll). We initially entertained the idea that these shared sequences were concerned with plasmid maintenance; however, further genetic studies by Hedges and Datta (unpublished data) showed that R-Amp and R6K could be stably maintained within the same cell (i.e., they were compatible). Additionally, when labeled TABLE 2. DNA-DNA hvbridizations between RSF1030 and Plasmids carrying the transposed TEM sequence DNA Reit d Las Incomp Resist-Lac- bIt patiG±Ca tms fanct Mo tmas biity(Mo factr class
homolSize ofTEM inogy serted with
SA SA- 1 R483 R483-1 R66a R66a- 1 R64 R64-1
None TEM None TEM None TEM None TEM