- Email: [email protected]
Relationship of staphylococcal conjugative plasmids to large gentamicin-resistance nonconjugative plasmids in clinical isolates of Staphylococcus epidermidis
DIAG. MICROBIOL.INFECT. DIS. 1990;13:227-234
227
MYCOLOGY
Relationship of Staphylococcal Conjugative Plasmids to Large G e n tami cin- R es is tan c e Nonconjugative Plasmids in Clinical Isolates of Staphylococcus epidermidis Marvin Rogolsky and Rae Gobert
Self-mobilization of large plasmids was not observed in 15 of 15 strains of gentamicin-resistant (Gm r) Staphylococcus epidermidis clinical isolates that were taken from bacteremic newborns in a neonatal intensive care unit. Alternatively, nine Gm r Staphylococcus aureus clinical isolates, which were isolated along with the S. epidermidis strains, were all shown previously to contain Gm r conjugative plasmids. All of the nonconjugative strains had plasmids that were similar in size to the S. aureus conjugative plasmids. Transfer of seven of these nonconjugative plasmids by protoplast transformation indicated that they carried Gm ~ determinants. The rationale of these studies was to detect the presence of genes for conjuga-
tion (tra) on the large Gm r nonconjugative plasmids. It was thought that these plasmids might contain either defective or deleted tra gene sequences. To gain insight into these possibilities, a 6.3-kb probe, which contained a major tra gene region, was hybridized with EcoR/and XbaI digests of nonconjugative plasmid DNA. Hybridization occurred with only one of eight plasmids. It was concluded that seven of the large S. epidermidis plasmids were not self-transmissible because they lacked tra genes. However, pMH6502 contained an excess of tra gene regions compared to prototype conjugative plasmids and was still not self-transmissible.
INTRODUCTION
Staphylococcus aureus and S. epidermidis strains, genes for aminoglycoside resistance are commonly associated with high-molecular-weight, Gm r self-transmissible (conjugative) plasmids. These genes code for a bifunctional protein that modifies aminoglycosides by acetylation and phosphorylation (Lyon et al., 1987). Inter- and intraspecies transfer of Gm r plasmids by conjugation is well documented (Archer and Johnston, 1983; Forbes and Schaberg, 1983; Goering and Ruff, 1984; M c D o ~ ! l et al., 1983; Archer et al., 1985; Lyon et al., 1987; Evans and Dyke, 1988; Zorbas et al., 1988; Thomas and Archer, 1989). Aside from selecting for resistance plasmids in general, antibiotic chemotherapy has probably also selected for strains with conjugative plasmids, which has broadened the reservoir of these plasmids. Staphylococcus aureus and S. epidermidis conjugative plasmids from widely separated geographical areas are very similar in restriction digestion patterns and in DNA homology (Archer et al., 1985; Schaberg et
Staphylococcus epidermidis has emerged as a signifi-
cant pathogen among compromised patients, neonatal intensive care unit populations, and patients receiving catheters and prosthetic devices (Jaffe et al., 1980; Baumgart et al., 1984; Parisi, 1985; Hall et al., 1987). It also ranks among the top drug-resistant Gram-positive cocci, especially among nosocomial strains (Parisi, 1985). In gentamicin-resistant (Gmr), From the Division of Cell Biologyand Biophysics, School of Basic Life Sciences, University of Missouri, Kansas City, Missouri. Address reprint requests to: Marvin Rogolsky, School of Basic Life Sciences; Division of Cell Biologyand Biophysics, Biological Sciences Building, University of Missouri-Kansas City, Kansas City, MO 64110-2499. Received December 15, 1989; revised and accepted February 2, 1990. © 1990Elsevier Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/90/$3.50
228
al., 1985; Zorbas et al., 1988; Thomas and Archer, 1989). The genes for conjugation (tra) appear to occupy similar positions on these plasmids, and the same situation appears to be true with genes for aminoglycoside resistance. A major region for tra genes is positioned on a 6.3-kb EcoRI fragment of pGO2 (Gordon L. Archer, personal communication), pJEI (Evans and Dyke, 1988), and pCRG1600 (Asch et al., 1984). Cloning experiments have shown that all of the genetic information, which was necessary for conjugation, was included in a 14.5-kb BglII region of pGO1 (Thomas and Archer, 1989). The 6.3-kb EcoRI fragment of pGO1 and pGO2 is homologous and contains tra gene sequences that are located within the 14.5-kb BglII segment of pGO1 (Gordon L. Archer, personal communication). In 1985, our laboratory received 336 staphylococcal strains that were isolated from bacteremic newborns in the Neonatal Intensive Care Unit at Children's Mercy Hospital (NICU-CMH) in Kansas City, Missouri. The API Staph Ident System was used to identify 35 of these strains as S. aureus and the remainder as coagulase-negative staphylococci (Hall et al., 1987). Of the latter, 194 were S. epidermidis and 59% of these isolates were GmL Our laboratory previously observed that nine of nine S. aureus strains from these isolates carried conjugative Gm r plasmids that were 43.8-63 kb in size (Zorbas et al., 1988). However, nine of nine Gm r S. epidermidis strains from these isolates were nonconjugative but all contained plasmids that were similar in size to prototype Gm r conjugative plasmids (Zorbas et al., 1988). The NICUCMH S. aureus conjugative plasmids and conjugative Gm r plasmids from other geographical areas had a high degree of homology and shared common restriction digestion fragments. The main objective of this paper was to analyze the large plasmids from the Gm r S. epidermidis clinical isolates to determine w h y they were not self-transmissible.
MATERIALS A N D METHODS Bacterial Strains A total of 336 clinical isolates were obtained during a 6-month period during 1985 from 152 patients with bacteremia in the NICU-CMH in Kansas City, Missouri. Isolates were obtained from multiple sites, primarily blood, throat, skin, and nose (Hall et al., 1987). The API Staph Ident System was used to identify 35 of 336 isolates as S. aureus and the remainder as coagulase-negative staphylococci. Of the latter, 194 were S. epidermidis and 59% of these (or 115) were Gm r. Fifteen of these Gm r S. epidermidis strains, primarily from blood cultures, were selected for use in this work and are listed in Table 1. All of these strains were slime producers.
M. Rogolsky and R. Gobert
Conjugation Clinical isolates were tested for their ability to transfer plasmids to recipient strains by the filter-mating method of Forbes and Schaberg (1983), which was modified as described previously by this laboratory (Zorbas et al. 1988). The concentration of selective agents used was fusidic acid (25 ~g/ml), gentamicin (10 ~g/ml), and novobiocin (10 ~g/ml).
Protoplast Transformation Because the NICU-CMH S. epidermidis strains contained four or more plasmids, only the large plasmids from these strains were transformed into protoplasts of strain SAl13, a rifampin-resistant, restriction-deficient strain of S. aureus. A modification of the method of Lindberg (1981) was used for protoplast transformation. Modifications included resuspending the pellet in SMM media (100 mM Tris, 20 mmol MgSO4 . 7 H 2 0 , 0.8 M sucrose at pH 7.6) before the addition of lysostaphin. Also the HBM hypertronic broth, into which protoplasted cells were resuspended, was prepared with 1 M sodium succinate in place of sucrose. After regeneration of protoplasts on DM3 media, Gm r transformants were selected on heart infusion agar (Difco Laboratories, Detroit, MI), with 10 ~g/ml of gentamicin and 10 ~g/ml of rifampin. All Gm r transformants were screened for plasmid DNA after electrophoresis on agarose gels.
Isolation of Plasmid D N A When small quantities of DNA were required, "mini" cleared lysates of S. epidermidis strains were obtained after using a method described by Parisi and Hecht (1980). Larger quantities of plasmid DNA, which were required for preparation of probe and transforming DNA, were obtained from a cesium chloride (CsC1) density gradient by a modification of the method described by Jaffe et al. (1980). Modifications included growing cells at 37°C overnight on brainheart infusion (BHI) (Difco) agar, which was prepared with Noble agar (Difco). After the cells were lysed and the proteins denatured by heating the mixture for 15 min at 65°C, the lysate was centrifuged at 10,400 g for 10 min. The DNA in the supernatant was then added to 0.313 volume of 42% (wt/vol) polyethylene glycol 8000 (Sigma Chemical Company, St. Louis, MO) to precipitate the DNA. Donor large plasmid DNA for protoplast transformation was separated from smaller plasmids in the denser band taken from the CsC1 gradients after centrifugation through neutral sucrose velocity gradients, which separated the plasmids into distinct bands. Neutral sucrose velocity centrifugation was
229
Plasmids of Staphylococcus epidermidis
TABLE 1.
Staphylococcal Strains
Strain
Relevant Phenotype
Comments
Source
S. epidermidis
S.
MH1202 MH1419 MH1520 MH3902 MH4014 MH6502 MH8702 MHl182 MH2614 MH1282 MH1322 MH1445 MH1442 MH1452 MH5220 G2
Gm r GmrTcrEmr Gm r Gm r Gm r Gm r Gm r Gm r Gm r Gm r Gm ~ Gm r Gm r Gm r Gm r Gm r
UM899
GmrTcrEmr
131 1312
RifrFus ~ RifrFusrNov r
S18
SmrCdrRif ~
Clinical isolate; blood culture Clinical isolate; skin culture Clinical isolate; throat culture Clinical isolate; blood culture Clinical isolate; nose culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; nose culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; throat culture Contains pGO2, which was the source of the EcoRI 6.3-kb probe Contains conjugative plasmid pAM899-1 Plasmid-free recipient for conjugation Strain 131 made Novr; conjugation recipient Plasmid-free recipient for conjugation
NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; G.L. Archer
Restriction-deficient, plasmid-free recipient for conjugation Restriction-deficient, plasmid-free recipient for conjugation Restriction-deficient, plasmid-free recipient for conjugation Restriction-deficient, plasmid-free recipient for protoplast transformation Obtained after conjugation of pAM899-1 from strain UM899 to strain ISP781
R.V. Goering
this this this this this this this this this this this this this this this
work work work work work work work work work work work work work work work
D.R. Schaberg J.T. Parisi This work R.W. McDonnell
aureus ISP781
RifrNovr
RN2677
Nov r
879R4
RifrFus r
SAl13
Rifr
SA8991
Gm r
R.P. Novick D.R. Schaberg M. Lindberg
This work
Gm, gentamicin; Tc, tetracycline; Em, erythromycin; Rif, rifampin; Fus, fusidic acid; Nov, novobiocin; Sm; streptomycin; and Cd, cadmium.
p e r f o r m e d according to the p r o c e d u r e of W a r r e n et al. (1975), with the exception that e t h i d i u m b r o m i d e w a s a d d e d to the g r a d i e n t s a n d the large p l a s m i d s w e r e localized, isolated, a n d purified b y the s a m e m e t h o d s u s e d after CsC1 d e n s i t y g r a d i e n t centrifugation.
Agarose Gel Electrophoresis All lysates w e r e s c r e e n e d on horizontal slab gels of 0.6-0.8% t y p e II a g a r o s e (Sigma) b y u s i n g TBE buffer (0.89 M Tris, 0.089 M boric acid, 0.002 M EDTA at p h 8.0). Gels w e r e r u n in a Bio-Rad electrophoresis
cell (Bio-Rad Laboratories, R i c h m o n d , CA) for 1719 hr at either 50 V or at 10 m A . Gels w e r e stained with e t h i d i u m b r o m i d e (1 ~g/ml) for 30 rain a n d rinsed in distilled w a t e r for 15 min. Bands w e r e visualized with a transilluminator (Ultra-Violet Products, Inc., San Gabriel, CA). The gels w e r e p h o t o graphed on Kodak Tri-X film with a Pentax Spotmatic c a m e r a a n d a M a c r o - T a k u m a r f/1.4 lens.
Endonuclease Restriction Digestion Restriction digestion w a s p e r f o r m e d b y the m e t h o d of Maniatis et al. (1982). The digested D N A w a s
230
electrophoresed in 0.7% low melting temperature agarose (Bio-Rad) for 16 hr at 9 mA. After staining and visualization with ethidium bromide, the appropriate bands were cut out and prepared for nick translation by removing the agarose according to the methods of Chen and Thomas (1980), with some modification. Restriction fragments that were used to make probe DNA required additional concentration and purification before undergoing nick translation. EcoRI restriction fragments were first concentrated with ammonium acetate (NH4OAc) and added to two volumes of 95% ethanol. After overnight incubation at 0°C, precipitated DNA was centrifuged. The pellet was then air-dried and suspended into a Tris (1 mM)/EDTA (1 mM) buffer. An alternative method of removing the agarose from the probe DNA was with the use of the GeneClean Kit (Bio 101, Inc., La Jolla, CA) according to the manufacturer's directions.
Nick Translation The purified pGO2 fragment was biotinylated by incorporating biotin-ll-dUTP (Bethesda Research Laboratories, Gaithersburg, MD) via nick translation according to the procedure of Rigby et al. (1977) and using the Nick Translation Reagent Kit (Bethesda Research Laboratories) according to the instructions of the manufacturer. The unincorporated nucleotides were removed by ethanol precipitation in the presence of NH4OAc, or they were removed with the GeneClean Kit, as described previously.
Southern Blot Hybridization Southern blots consisted of either a gel containing whole plasmid DNA or a gel of a restriction digest of a single plasmid. The DNA was transferred from the gel to a nitrocellulose membrane by using the method described by Southern (1975) and the modification of Meinkoth and Wahl (1984). Hybridization and biotin detection procedures followed the manufacturer's instructions (BIuGENE Nonradioactive Nucleic Acid Detection System [Bethesda Research Laboratories]), according to the methods of Leary et al. (1983). After performing the initial hybridizations, the conditions of stringency were experimentally increased by raising the temperature at which the membranes were hybridized from 42 ° to 48°C. The concentration of saline sodium citrate in the washes was changed from 2 x, 0.2 x, and 0.16 x to 0 . 1 x , 0.16x and 0 . 1 x , respectively. Wash times were increased from 3 min, 3 min, and 20 min to 5 rain, 5 min, and 30 rain, respectively. After color development, the blot was photographed wet with a Polaroid MP3 Copy Camera and 4 x 5 Kodak Professional Copy film.
M. Rogolsky and R. Gobert
RESULTS Conjugation Neither inter- nor intraspecies transfer of gentamicin resistance from 15 NICU-CMH Gm r strains (Table 1) to three different S. aureus recipient strains (Table 1) and to three different S. epidermidis nonconjugative strains (Table 1) was observed. Suspensions of recipient cells consisted of - 1 0 9 cells/ml. This should therefore allow for detection of conjugation frequencies of at least 1 0 - 9 . Because the donor strains did not yield frequencies within the limits of detection, it was assumed that they were unable to transfer gentamicin resistance by conjugation. As a positive control, an S. epidermidis UM899 donor strain was able to transfer gentamicin resistance to S. aureus RN2677 and ISP781 strains and to S. epidermidis strains 131 and 312 at frequencies of 10-5 to 1 0 - 7 .
Transfer of Gm r Plasmids by Protoplast Transformation Large plasmids were isolated from 10 nonconjugative NICU-CMH S. epidermidis strains after CsC1 density gradient centrifugation, followed by neutral sucrose velocity centrifugation. Plasmids pMH1442, pMH1445, pMH1322, pMH8702, pMH6502, pMH3902, and pMH1452, but not pMH2614, pMH1182, or pMH1419, were observed to transfer gentamicin resistance to recipient SAl13 strains after protoplast transformation. Frequency of transfer ranged between 7 x 10~ and 1 x 1 0 - 7 . Agarose gel electrophoresis was used to show that Gm r transformants had acquired a specific large nonconjugative plasmid. Thus, gentamicin resistance could be transferred by 7 of 10 large plasmids from the NICU-CMH S. epidermidis strains by transformation but not by conjugation.
Plasmid Analyses Plasmid DNA from 7 of the 15 Gm r S. epidermidis nonconjugative strains was analyzed after agarose gel electrophoresis and compared on the same gel with plasmid DNA from four Gm r conjugative strains. The 11 strains used in these analyses are listed in Table 2. Whereas S. aureus Gm r conjugative strains from the NICU-CMH clinical isolates contained only one or two plasmids (Zorbas et al., 1988). S. epidermidis strains contained from four to eight different plasmids. Many of the plasmids from the nonconjugative strains were equal in molecular mass, but no strain had identical plasmid patterns. All large nonconjugative plasmids were similar in molecular mass to the conjugative plasmids against which they were compared on the same agarose gel (Table 2).
231
Plasmids of Staphylococcus epidermidis
TABLE 2. Comparative Sizes of Conjugative Plasmids and the Large Plasmids from Nonconjugative Gm r S. epidermidis Clinical Isolates Plasmid a
Host
Conjugative pGO2 pCRG1600 pAM899-1 pIZ7814
G2 CRG1600 SA8991c MH7814
Nonconjugative pMH3902 pMH6502 pMH8702 pMH1182 pMH2614 pMH1322 pMH1452
MH3902 MH6502 MH8702 MHl182 MH2614 MH1322 MH1452
Size (kb) b 50 52.9 50 48.3 51 51 50 43 43 51 50
Source G.L. Archer R.V. Goering D.R. Schaberg This laboratory This This This This This This This
work work work work work work work
aPlasmids were analyzed after electrophresison an agarose ~lasmid sizes werecalculatedby correlatingthem to the known sizes of the standard plasmid markers in Escherichiacoli V517 and using a double log plot of distance migrated versus size (in kb). cObtained after conjugation of pAM899-1 from strain UM899 (Forbes and Schaberg 1983) to strain ISP781.
All large nonconjugative plasmids, with the exception of pMH1182 and pMH2614 (Table 2), were found to express gentamicin resistance.
Restriction Digests Restriction digest bands from eight large nonconjugative plasmids showed a significant degree of variation after electrophoresis on agarose gels. The sizes (in kilobases) of most of these plasmids are given in Table 2. Digests of nonconjugative S. epidermidis plasmids with EcoRI (Fig. 1A) and XbaI (Fig. 2A) indicated that they had similar, but never identical, restriction patterns. This was even the case for plasmids that had identical molecular masses. Plasmids pMH1452 (50 kb), pMH1322 (51 kb), and pMH8702 (50 kb) (Fig. 2A, lanes 6, 7, and 9, respectively) show identical XbaI restriction digest patterns, with the exception that each had a band that was not present in the other two digests. Many plasmids shared common EcoRI (Fig. 1A) and XbaI (Fig. 2A) fragments, but no one fragment was common to all of the nonconjugative plasmids. In contrast, conjugative plasmids of similar molecular masses from the S. aureus NICU-CMH strains always had identical restriction patterns (Zorbas et al., 1988). Evidence has been provided to indicate that all NICUCMH plasmids used (Figs. 1 and 2) coded for gentamicin resistance, with the exception of pMH2614.
Plasmid DNA Hybridization Analyses An attempt was then made to identify tra gene sequences on the nonconjugative plasmids. A probe was made from the 6.3-kb EcoRI fragment of pGO2, which contains a major tra gene region (Gordon L. Archer, personal communication). This probe was then used to seek out tra genes on the large nonconjugative plasmids. However, w h e n the 6.3-kb probe was hybridized with an EcoRI digest of these plasmids, hybridization occurred with only one of seven plasmids (Fig. 1B). Gentamicin resistance was found to be expressed by all of these plasmids, with the exception of pMH2614. Note that a fragment of 6.3 kb can be seen in the EcoRI digests of several of the nonconjugative plasmids (Fig. 1A). For example, an EcoRI fragment of - 6.0 kb can be seen in pMH3902 (Fig. 1A, lane 8), yet it did not hybridize with the probe. This 6.0-kb fragment was originally thought to be analogous to the 6.3-kb EcoRI fragment of conjugative plasmids but with a 0.3-kb deletion (Zorbas et al. 1988). The data in Figure 1 show that this is obviously not the case. The exceptional nonconjugative plasmid was pMH6502. In the EcoRI digest seen in Figure 1, lanes 7, the probe not only hybridized very strongly with a 6.3-kb fragment but with two larger EcoRI fragments as well. It was also observed after agarose gel electrophoresis that a single band of near complete homology was quite evident between the 6.3-kb probe and the 6.3-kb EcoRI fragments of conjugative plasmids pAM899-1, pIZ7814, pCRG1600, and pGO2. Hybridization of the probe to pGO2 is shown in Figure 1B, lane 9. In contrast to that seen with pMH6502, homology is apparent with only the 6.3-kb EcoRI band of pGO2. The 6.3-kb probe was also hybridized to an XbaI restriction digest of conjugative and nonconjugative plasmids (Fig. 2). The probe hybridized most strongly with a large fragment of - 16.9 kb in size, which was present in all of the conjugative plasmids (Fig. 2, lanes 1-4). This XbaI fragment has already been shown to be a major tra region for pCRG1600 (Asch et al., 1984) and pJE1 (Evans and Dyke, 1988). The high degree of homology of the conjugative plasmids with the probe indicates the presence of homologous tra gene regions on the pGO2, pCRG1600, pIZ7814, and pAM899-1 plasmids (Fig. 2B, lanes 14, respectively). Alternatively, the probe did not hybridize with the nonconjugative plasmids (Fig. 2B, lanes 5-7 and 9), with the exception of pMH6502 (Fig. 2B, lane 8). All of these nonconjugative plasmids expressed gentamicin resistance.
DISCUSSION Self-mobilization of large plasmids was not observed in 15 of 15 strains of Gm r S. epidermidis isolates taken
M. Rogolsky and R. Gobert
232
A
1
2 3 4 5
6
789
B
"
~
"*"
=
=
"
_6.3
|
FIGURE 1. Agarose gel electrophoresis of EcoRI restriction endonuclease-digested DNA of the large plasmid from Gm r nonconjugative S. epidermidis clinical isolates (A). Filter hybridization of the EcoRI restriction digests with the 6.3-kb EcoRI fragment from the conjugative plasmid pGO2 as probe DNA (B). The probe DNA contains a major tra gene region. (Lanes 1) Phage lambda DNA, (lanes 2) pMH1442, (lanes 3) pMH1445, (lanes 4) pMH1322, (lanes 5) pMH2614, (lanes 6) pMH8702, (lanes 7) pMH6502, (lanes 8) pMH3902, and (lanes 9) pGO2. Molecular masses are in kilobases.
1
2
34
5 6 7 8
910
16. c 12.2 10.1
6.2 4..q
2.~
FIGURE 2. Agarose gel electrophoresis of XbaI restriction endonuclease-digested DNA of conjugative plasmids and the large plasmid from Gm r conjugative S. epidermidis clinical isolates (A). Filter hybridization of the XbaI restriction digests with the 6.3-kb EcoRI fragment from the conjugative plasmid pGO2 as probe DNA (B). The probe DNA contains a major tra gene region. (Lanes 1) pGO2, (lanes 2) pCRG1600, (lanes 3), pIZ7814, (lanes 4) pAM899-1, (lanes 5) pMH3902, (lanes 6) pMH1452, (lanes 7) pMH1322, (lanes 8) pMH6502, (lanes 9) pMH8702, and (lanes 10) phage lambda DNA. Molecular masses are in kilobases.
Plasmids of Staphylococcus epidermidis
from newborns in a neonatal intensive care unit. However self-transmissible plasmids among Gm" S. epidermidis clinical isolates were routinely observed by other laboratories (Archer and Johnston, 1983; Forbes and Schaberg, 1983; McDonnell et al., 1983; Archer et al., 1985). The fact that nine of nine NICUCMH Gm r S. aureus strains, which were all isolated along with the 15 nonconjugative strains, had selftransmissible plasmids (Zorbas et al. 1988) makes this situation even more puzzling. In a survey of 18 Gm r S. epidermidis isolates and 23 S. aureus Gm r isolates from a Virginia hospital, all of the S. aureus strains, but only 50% of the S. epidermidis strains, were found to be conjugative (Archer and Johnston, 1983). Thus, surveys from this and other investigations (Archer and Johnston, 1983; Archer et al., 1985; Zorbas et al., 1988) appear to indicate that Gm r strains of S. aureus clinical isolates are more likely to contain a self-transmissible plasmid that Gm r S. epidermidis clinical isolates. This is in spite of the fact that Gm ~ S. epidermidis strains are more prevalent than Gm r strains of S. aureus in a nosocomial environment (Weinstein et al., 1982; Hall et al., 1987). Transfer of seven large nonconjugative plasmids from the NICU-CMH strains by protoplast transformation demonstrated that they carried Gm r genes and were indeed Gm r plasmids that were similar in size to conjugative Gm ~ plasmids. It was then thought that a search for possible tra genes on the large plasmids might determine w h y these plasmids were not selftransmissible. Identical 6.3-kb EcoRI tra gene fragments appear to be present in all conjugative plasmids (Asch et al., 1984; Archer et al., 1986; Evans and Dyke, 1988; Zorbas et al., 1988). However, it was surprising to discover that this fragment only hybridized to one of eight of the large nonconjugative plasmids. It is therefore reasonable to conclude that seven of these large Gm ~ plasmids are not self-transmissible because they lack tra genes. Whereas these seven plasmids are devoid of tra genes, pMH6502 (Fig. 1B, lane 7) contains an excess of tra genes in comparison to prototype conjugative plasmids and still is not selftransmissible. The orientation of pMH6502 tra genes shows obvious variation from the consistent pattern that is present on prototype conjugative plasmids (Fig. 1B, lanes 7 and 9). This indicates that molecular rearrangements of pMH6502 have probably resulted in defective tra gene expression. Our laboratory is presently doing a molecular analysis of this plasmid to gain more insight into this possibility. Data from the hybridization studies indicated that complete or near complete 6.3-kb EcoRI segments from large nonconjugative Gm r plasmids are not necessarily
233
analogous to the common 6.3-kb EcoRI fragments of conjugative plasmids. Restriction digestion analyses showed that the nonconjugative plasmids did not share a high degree of relatedness with conjugative plasmids or, to a large extent, even with themselves. Alternatively, conjugative plasmids even from widely separated geographical areas share much physical relatedness (Schaberg et al., 1985; Zorbas et al., 1988; Thomas and Archer, 1989). In summary, it is concluded that the observations made with the NICU-CMH strains and those made by other investigators seem to indicate that the conjugative plasmids evolved from one or more common large ancestral Gm r plasmids and that S. epidermis was the natural environment for these plasmids. If the tra gene region was transposable (Thomas and Archer, 1989), this large ancestral Gm r plasmid might have served as the natural target replicon for this region. Transpositions and other inter- and intramolecular rearrangements of the ancestral plasmids over a long time period might have resulted in diversity of large Gm r plasmids in S. ep ~ idermidis strains. This molecular activity, especially in regard to transposition, could also have led to the depletion or duplication of any existing tra gene sequences. Such events might have molded the large plasmids in the NICU-CMH S. epidermidis strains. Alternatively, conjugative plasmids appear to belong to a family of highly related Gm r replicons. Because the Gm r plasmid is nearly always of the conjugative variety in the nosocomial S. aureus host, one might conclude that this plasmid was recently acquired by S. aureus from S. epidermidis donors. This highly conserved plasmid was most likely selected for transfer from the donor because it contained a functional set of tra genes, which were lacking on the other Gm r plasmids of the donor. This thinking is supported by the evidence that recent chemotherapy with aminoglycosides has broadened the nosocomial reservoir of Gm r plasmids in S. epidermidis (Weinstein et al., 1982). It can then be assumed that this was followed by dramatic increases in the transfer of specific Gm r plasmids to S. aureus after in vivo conjugation. Growth of the Gm r conjugative and nonconjugative plasmid populations is expected to continue to cause serious problems in the treatment of staphylococcal infections. Gratitude is extended to William von David, who helped us to modify the procedures for protoplast transformation and for recovery of large quantities of plasmid DNA. We are also grateful to G. L. Archer for helpful suggestions and discussion.
234
M. Rogolsky and R. Gobert
REFERENCES Archer GL, Johnston J (1983) Self-transmissible plasmids in staphylococci that encode resistance to aminoglycosides. Antimicrob Agents Chemother 24:70-77. Archer GL, Dietrick DR, Johnston J (1985) Molecular epidemiology of transmissible gentamicin resistance among coagulase-negative staphylococci in a cardiac surgery unit. J Infect Dis 151:24-251. Archer GL, Coughter JP, Johnston JL (1986) Plasmid encoded trimethoprim resistance in staphylococci. Antimicrob Agents Chemother 29:733-740. Asch DK, Goering RV, Ruff EA (1984) Isolation and preliminary characterization of a plasmid mutant derepressed for conjugal transfer in S. aureus. Plasmid 12:197202. Baumgart S, Hall SE, Campos JM, Polin RA (1984) Sepsis with coagulase-negative staphylococci in critically ill newborns. Am J Dis Child 137:461-467. Chen CW, Thomas Jr. CA (1980) Recovery of DNA fragments from agarose gels. Anal Biochem 101:339-341. Evans J, Dyke KGH (1988) Characterization of the conjugation system associated with the Staphyloccus aureus plasmid pJE1. J Gen Microbiol 134:1-8. Forbes BA, Schaberg DR (1983) Transfer of resistance plasmids from Staphylococcus epidermidis to Staphylococcus aureus: evidence for conjugative exchange of resistance. J Bacteriol 153:627-634. Goering RV, Ruff, EA (1983) Comparative analysis of conjugative plasmids mediating gentamicin resistance in Staphylococcus aureus. Antimicrob Agents Chemother 24:450452. Hall RT, Hall SL, Barnes WG, Izuegbu J, Rogolsky M, Zorbas I (1987) Characteristics of coagulase negative staphylococcus from infants with bacteremia. Pediatr Infect Dis J 6:377-383. Jaffe HW, Sweeney HM, Nathan C, Weinstein RA, Kabins SA, Cohen S (1980) Identity and interspecific transfer of gentamicin resistance plasmids in Staphylococcus aureus and Staphylococcus epidermidis. 141:738-746. Leary JJ, Brigati DJ, Ward DC (1983) Rapid and sensitive colorimetric method for visualizing biotin-labelled DNA probes hybridized to DNA or RNA immobilized on nitrocellulose: bio-blots. Proc Natl Acad Sci USA 80:40454044. Lindberg M (1981) Genetic studies in Staphylococcus aureus using protoplasts: cell fusion and transformation. In Staphylococci and Staphylococcal Infections. Ed., J Jeljaszewicz, Zbl Bakt Suppl 10. New York: Gustav Fisher Verlag, pp 535-541.
Lyon BR, Gillespie MT, Byrne ME, May JW, Skurray RA (1987) Plasmid-mediated resistance to gentamicin in Staphylococcus aureus: the involvement of a transposon. J Med Microbiol 23:101-110. Maniatis T, Fritsch EF, Sambrook J (1982) Molecular Cloning: A Laboratory Manual. New York: Cold Spring Harbor Laboratory. McDonnell RW, Sweeney HM, Cohen S (1983) Conjugational transfer of gentamicin resistance plasmids intra- and interspecifically in Staphylococcus aureus and Staphylococcus epidermidis. Antimicrob Agents Chemother 23:151-160. Meinkoth J, Wahl G (1984) Hybridization of nucleic acids immobilized on solid supports. Anal Biochem 138:267283. Parisi JT, Hecht DW (1980) Plasmid profiles in epidemiologic studies of infections by Staphylococcus epidermidis. J Infect Dis 141:637-643. Parisi JT (1985) Coagulase negative staphylococci and the epidemiological typing of Staphylococcus epidermidis. Microbiol Rev 49:126-139. Rigby PWJ, Dieckman M, Rhodes C, Berg P (1977) Labelling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J Mol Biol 113:237-252. Schaberg DR, Power G, Betzold G, Forbes BA (1985) Conjugative R plasmids in antimicrobial resistance of Staphylococcus aureus causing nosocomial infections. ] Infect Dis 152:43-49. Southern, E (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517. Thomas WD, Archer GL (1989) Identification and cloning of the conjugative transfer region of Staphylococcus aureus plasmid pGO1. J Bacteriol 171:684-691. Warren R, Rogolsky M, Wiley BB, Glasgow LA (1975) Isolation of extrachromosomal DNA for exfoliative toxin production from phage group II Staphylococcus aureus. J Bacteriol 122:99-105. Weinstein RA, Kabins SA, Nathan C, Sweeney HM, Jaffe HW, Cohen S (1982) Gentamicin resistant staphylococci as hospital flora: epidemiology and resistance plasmids. J Infect Dis 145:374-381. Zorbas I, Hall RT, Hall SL, Barnes WG, Rogolsky M (1988) Molecular analyses of conjugative, gentamicin-resistance plasmids from staphylococcal clinical isolates. Can J Microbiol 34:1050-1057.
227
MYCOLOGY
Relationship of Staphylococcal Conjugative Plasmids to Large G e n tami cin- R es is tan c e Nonconjugative Plasmids in Clinical Isolates of Staphylococcus epidermidis Marvin Rogolsky and Rae Gobert
Self-mobilization of large plasmids was not observed in 15 of 15 strains of gentamicin-resistant (Gm r) Staphylococcus epidermidis clinical isolates that were taken from bacteremic newborns in a neonatal intensive care unit. Alternatively, nine Gm r Staphylococcus aureus clinical isolates, which were isolated along with the S. epidermidis strains, were all shown previously to contain Gm r conjugative plasmids. All of the nonconjugative strains had plasmids that were similar in size to the S. aureus conjugative plasmids. Transfer of seven of these nonconjugative plasmids by protoplast transformation indicated that they carried Gm ~ determinants. The rationale of these studies was to detect the presence of genes for conjuga-
tion (tra) on the large Gm r nonconjugative plasmids. It was thought that these plasmids might contain either defective or deleted tra gene sequences. To gain insight into these possibilities, a 6.3-kb probe, which contained a major tra gene region, was hybridized with EcoR/and XbaI digests of nonconjugative plasmid DNA. Hybridization occurred with only one of eight plasmids. It was concluded that seven of the large S. epidermidis plasmids were not self-transmissible because they lacked tra genes. However, pMH6502 contained an excess of tra gene regions compared to prototype conjugative plasmids and was still not self-transmissible.
INTRODUCTION
Staphylococcus aureus and S. epidermidis strains, genes for aminoglycoside resistance are commonly associated with high-molecular-weight, Gm r self-transmissible (conjugative) plasmids. These genes code for a bifunctional protein that modifies aminoglycosides by acetylation and phosphorylation (Lyon et al., 1987). Inter- and intraspecies transfer of Gm r plasmids by conjugation is well documented (Archer and Johnston, 1983; Forbes and Schaberg, 1983; Goering and Ruff, 1984; M c D o ~ ! l et al., 1983; Archer et al., 1985; Lyon et al., 1987; Evans and Dyke, 1988; Zorbas et al., 1988; Thomas and Archer, 1989). Aside from selecting for resistance plasmids in general, antibiotic chemotherapy has probably also selected for strains with conjugative plasmids, which has broadened the reservoir of these plasmids. Staphylococcus aureus and S. epidermidis conjugative plasmids from widely separated geographical areas are very similar in restriction digestion patterns and in DNA homology (Archer et al., 1985; Schaberg et
Staphylococcus epidermidis has emerged as a signifi-
cant pathogen among compromised patients, neonatal intensive care unit populations, and patients receiving catheters and prosthetic devices (Jaffe et al., 1980; Baumgart et al., 1984; Parisi, 1985; Hall et al., 1987). It also ranks among the top drug-resistant Gram-positive cocci, especially among nosocomial strains (Parisi, 1985). In gentamicin-resistant (Gmr), From the Division of Cell Biologyand Biophysics, School of Basic Life Sciences, University of Missouri, Kansas City, Missouri. Address reprint requests to: Marvin Rogolsky, School of Basic Life Sciences; Division of Cell Biologyand Biophysics, Biological Sciences Building, University of Missouri-Kansas City, Kansas City, MO 64110-2499. Received December 15, 1989; revised and accepted February 2, 1990. © 1990Elsevier Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/90/$3.50
228
al., 1985; Zorbas et al., 1988; Thomas and Archer, 1989). The genes for conjugation (tra) appear to occupy similar positions on these plasmids, and the same situation appears to be true with genes for aminoglycoside resistance. A major region for tra genes is positioned on a 6.3-kb EcoRI fragment of pGO2 (Gordon L. Archer, personal communication), pJEI (Evans and Dyke, 1988), and pCRG1600 (Asch et al., 1984). Cloning experiments have shown that all of the genetic information, which was necessary for conjugation, was included in a 14.5-kb BglII region of pGO1 (Thomas and Archer, 1989). The 6.3-kb EcoRI fragment of pGO1 and pGO2 is homologous and contains tra gene sequences that are located within the 14.5-kb BglII segment of pGO1 (Gordon L. Archer, personal communication). In 1985, our laboratory received 336 staphylococcal strains that were isolated from bacteremic newborns in the Neonatal Intensive Care Unit at Children's Mercy Hospital (NICU-CMH) in Kansas City, Missouri. The API Staph Ident System was used to identify 35 of these strains as S. aureus and the remainder as coagulase-negative staphylococci (Hall et al., 1987). Of the latter, 194 were S. epidermidis and 59% of these isolates were GmL Our laboratory previously observed that nine of nine S. aureus strains from these isolates carried conjugative Gm r plasmids that were 43.8-63 kb in size (Zorbas et al., 1988). However, nine of nine Gm r S. epidermidis strains from these isolates were nonconjugative but all contained plasmids that were similar in size to prototype Gm r conjugative plasmids (Zorbas et al., 1988). The NICUCMH S. aureus conjugative plasmids and conjugative Gm r plasmids from other geographical areas had a high degree of homology and shared common restriction digestion fragments. The main objective of this paper was to analyze the large plasmids from the Gm r S. epidermidis clinical isolates to determine w h y they were not self-transmissible.
MATERIALS A N D METHODS Bacterial Strains A total of 336 clinical isolates were obtained during a 6-month period during 1985 from 152 patients with bacteremia in the NICU-CMH in Kansas City, Missouri. Isolates were obtained from multiple sites, primarily blood, throat, skin, and nose (Hall et al., 1987). The API Staph Ident System was used to identify 35 of 336 isolates as S. aureus and the remainder as coagulase-negative staphylococci. Of the latter, 194 were S. epidermidis and 59% of these (or 115) were Gm r. Fifteen of these Gm r S. epidermidis strains, primarily from blood cultures, were selected for use in this work and are listed in Table 1. All of these strains were slime producers.
M. Rogolsky and R. Gobert
Conjugation Clinical isolates were tested for their ability to transfer plasmids to recipient strains by the filter-mating method of Forbes and Schaberg (1983), which was modified as described previously by this laboratory (Zorbas et al. 1988). The concentration of selective agents used was fusidic acid (25 ~g/ml), gentamicin (10 ~g/ml), and novobiocin (10 ~g/ml).
Protoplast Transformation Because the NICU-CMH S. epidermidis strains contained four or more plasmids, only the large plasmids from these strains were transformed into protoplasts of strain SAl13, a rifampin-resistant, restriction-deficient strain of S. aureus. A modification of the method of Lindberg (1981) was used for protoplast transformation. Modifications included resuspending the pellet in SMM media (100 mM Tris, 20 mmol MgSO4 . 7 H 2 0 , 0.8 M sucrose at pH 7.6) before the addition of lysostaphin. Also the HBM hypertronic broth, into which protoplasted cells were resuspended, was prepared with 1 M sodium succinate in place of sucrose. After regeneration of protoplasts on DM3 media, Gm r transformants were selected on heart infusion agar (Difco Laboratories, Detroit, MI), with 10 ~g/ml of gentamicin and 10 ~g/ml of rifampin. All Gm r transformants were screened for plasmid DNA after electrophoresis on agarose gels.
Isolation of Plasmid D N A When small quantities of DNA were required, "mini" cleared lysates of S. epidermidis strains were obtained after using a method described by Parisi and Hecht (1980). Larger quantities of plasmid DNA, which were required for preparation of probe and transforming DNA, were obtained from a cesium chloride (CsC1) density gradient by a modification of the method described by Jaffe et al. (1980). Modifications included growing cells at 37°C overnight on brainheart infusion (BHI) (Difco) agar, which was prepared with Noble agar (Difco). After the cells were lysed and the proteins denatured by heating the mixture for 15 min at 65°C, the lysate was centrifuged at 10,400 g for 10 min. The DNA in the supernatant was then added to 0.313 volume of 42% (wt/vol) polyethylene glycol 8000 (Sigma Chemical Company, St. Louis, MO) to precipitate the DNA. Donor large plasmid DNA for protoplast transformation was separated from smaller plasmids in the denser band taken from the CsC1 gradients after centrifugation through neutral sucrose velocity gradients, which separated the plasmids into distinct bands. Neutral sucrose velocity centrifugation was
229
Plasmids of Staphylococcus epidermidis
TABLE 1.
Staphylococcal Strains
Strain
Relevant Phenotype
Comments
Source
S. epidermidis
S.
MH1202 MH1419 MH1520 MH3902 MH4014 MH6502 MH8702 MHl182 MH2614 MH1282 MH1322 MH1445 MH1442 MH1452 MH5220 G2
Gm r GmrTcrEmr Gm r Gm r Gm r Gm r Gm r Gm r Gm r Gm r Gm ~ Gm r Gm r Gm r Gm r Gm r
UM899
GmrTcrEmr
131 1312
RifrFus ~ RifrFusrNov r
S18
SmrCdrRif ~
Clinical isolate; blood culture Clinical isolate; skin culture Clinical isolate; throat culture Clinical isolate; blood culture Clinical isolate; nose culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; nose culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; blood culture Clinical isolate; throat culture Contains pGO2, which was the source of the EcoRI 6.3-kb probe Contains conjugative plasmid pAM899-1 Plasmid-free recipient for conjugation Strain 131 made Novr; conjugation recipient Plasmid-free recipient for conjugation
NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; NICU-CMH; G.L. Archer
Restriction-deficient, plasmid-free recipient for conjugation Restriction-deficient, plasmid-free recipient for conjugation Restriction-deficient, plasmid-free recipient for conjugation Restriction-deficient, plasmid-free recipient for protoplast transformation Obtained after conjugation of pAM899-1 from strain UM899 to strain ISP781
R.V. Goering
this this this this this this this this this this this this this this this
work work work work work work work work work work work work work work work
D.R. Schaberg J.T. Parisi This work R.W. McDonnell
aureus ISP781
RifrNovr
RN2677
Nov r
879R4
RifrFus r
SAl13
Rifr
SA8991
Gm r
R.P. Novick D.R. Schaberg M. Lindberg
This work
Gm, gentamicin; Tc, tetracycline; Em, erythromycin; Rif, rifampin; Fus, fusidic acid; Nov, novobiocin; Sm; streptomycin; and Cd, cadmium.
p e r f o r m e d according to the p r o c e d u r e of W a r r e n et al. (1975), with the exception that e t h i d i u m b r o m i d e w a s a d d e d to the g r a d i e n t s a n d the large p l a s m i d s w e r e localized, isolated, a n d purified b y the s a m e m e t h o d s u s e d after CsC1 d e n s i t y g r a d i e n t centrifugation.
Agarose Gel Electrophoresis All lysates w e r e s c r e e n e d on horizontal slab gels of 0.6-0.8% t y p e II a g a r o s e (Sigma) b y u s i n g TBE buffer (0.89 M Tris, 0.089 M boric acid, 0.002 M EDTA at p h 8.0). Gels w e r e r u n in a Bio-Rad electrophoresis
cell (Bio-Rad Laboratories, R i c h m o n d , CA) for 1719 hr at either 50 V or at 10 m A . Gels w e r e stained with e t h i d i u m b r o m i d e (1 ~g/ml) for 30 rain a n d rinsed in distilled w a t e r for 15 min. Bands w e r e visualized with a transilluminator (Ultra-Violet Products, Inc., San Gabriel, CA). The gels w e r e p h o t o graphed on Kodak Tri-X film with a Pentax Spotmatic c a m e r a a n d a M a c r o - T a k u m a r f/1.4 lens.
Endonuclease Restriction Digestion Restriction digestion w a s p e r f o r m e d b y the m e t h o d of Maniatis et al. (1982). The digested D N A w a s
230
electrophoresed in 0.7% low melting temperature agarose (Bio-Rad) for 16 hr at 9 mA. After staining and visualization with ethidium bromide, the appropriate bands were cut out and prepared for nick translation by removing the agarose according to the methods of Chen and Thomas (1980), with some modification. Restriction fragments that were used to make probe DNA required additional concentration and purification before undergoing nick translation. EcoRI restriction fragments were first concentrated with ammonium acetate (NH4OAc) and added to two volumes of 95% ethanol. After overnight incubation at 0°C, precipitated DNA was centrifuged. The pellet was then air-dried and suspended into a Tris (1 mM)/EDTA (1 mM) buffer. An alternative method of removing the agarose from the probe DNA was with the use of the GeneClean Kit (Bio 101, Inc., La Jolla, CA) according to the manufacturer's directions.
Nick Translation The purified pGO2 fragment was biotinylated by incorporating biotin-ll-dUTP (Bethesda Research Laboratories, Gaithersburg, MD) via nick translation according to the procedure of Rigby et al. (1977) and using the Nick Translation Reagent Kit (Bethesda Research Laboratories) according to the instructions of the manufacturer. The unincorporated nucleotides were removed by ethanol precipitation in the presence of NH4OAc, or they were removed with the GeneClean Kit, as described previously.
Southern Blot Hybridization Southern blots consisted of either a gel containing whole plasmid DNA or a gel of a restriction digest of a single plasmid. The DNA was transferred from the gel to a nitrocellulose membrane by using the method described by Southern (1975) and the modification of Meinkoth and Wahl (1984). Hybridization and biotin detection procedures followed the manufacturer's instructions (BIuGENE Nonradioactive Nucleic Acid Detection System [Bethesda Research Laboratories]), according to the methods of Leary et al. (1983). After performing the initial hybridizations, the conditions of stringency were experimentally increased by raising the temperature at which the membranes were hybridized from 42 ° to 48°C. The concentration of saline sodium citrate in the washes was changed from 2 x, 0.2 x, and 0.16 x to 0 . 1 x , 0.16x and 0 . 1 x , respectively. Wash times were increased from 3 min, 3 min, and 20 min to 5 rain, 5 min, and 30 rain, respectively. After color development, the blot was photographed wet with a Polaroid MP3 Copy Camera and 4 x 5 Kodak Professional Copy film.
M. Rogolsky and R. Gobert
RESULTS Conjugation Neither inter- nor intraspecies transfer of gentamicin resistance from 15 NICU-CMH Gm r strains (Table 1) to three different S. aureus recipient strains (Table 1) and to three different S. epidermidis nonconjugative strains (Table 1) was observed. Suspensions of recipient cells consisted of - 1 0 9 cells/ml. This should therefore allow for detection of conjugation frequencies of at least 1 0 - 9 . Because the donor strains did not yield frequencies within the limits of detection, it was assumed that they were unable to transfer gentamicin resistance by conjugation. As a positive control, an S. epidermidis UM899 donor strain was able to transfer gentamicin resistance to S. aureus RN2677 and ISP781 strains and to S. epidermidis strains 131 and 312 at frequencies of 10-5 to 1 0 - 7 .
Transfer of Gm r Plasmids by Protoplast Transformation Large plasmids were isolated from 10 nonconjugative NICU-CMH S. epidermidis strains after CsC1 density gradient centrifugation, followed by neutral sucrose velocity centrifugation. Plasmids pMH1442, pMH1445, pMH1322, pMH8702, pMH6502, pMH3902, and pMH1452, but not pMH2614, pMH1182, or pMH1419, were observed to transfer gentamicin resistance to recipient SAl13 strains after protoplast transformation. Frequency of transfer ranged between 7 x 10~ and 1 x 1 0 - 7 . Agarose gel electrophoresis was used to show that Gm r transformants had acquired a specific large nonconjugative plasmid. Thus, gentamicin resistance could be transferred by 7 of 10 large plasmids from the NICU-CMH S. epidermidis strains by transformation but not by conjugation.
Plasmid Analyses Plasmid DNA from 7 of the 15 Gm r S. epidermidis nonconjugative strains was analyzed after agarose gel electrophoresis and compared on the same gel with plasmid DNA from four Gm r conjugative strains. The 11 strains used in these analyses are listed in Table 2. Whereas S. aureus Gm r conjugative strains from the NICU-CMH clinical isolates contained only one or two plasmids (Zorbas et al., 1988). S. epidermidis strains contained from four to eight different plasmids. Many of the plasmids from the nonconjugative strains were equal in molecular mass, but no strain had identical plasmid patterns. All large nonconjugative plasmids were similar in molecular mass to the conjugative plasmids against which they were compared on the same agarose gel (Table 2).
231
Plasmids of Staphylococcus epidermidis
TABLE 2. Comparative Sizes of Conjugative Plasmids and the Large Plasmids from Nonconjugative Gm r S. epidermidis Clinical Isolates Plasmid a
Host
Conjugative pGO2 pCRG1600 pAM899-1 pIZ7814
G2 CRG1600 SA8991c MH7814
Nonconjugative pMH3902 pMH6502 pMH8702 pMH1182 pMH2614 pMH1322 pMH1452
MH3902 MH6502 MH8702 MHl182 MH2614 MH1322 MH1452
Size (kb) b 50 52.9 50 48.3 51 51 50 43 43 51 50
Source G.L. Archer R.V. Goering D.R. Schaberg This laboratory This This This This This This This
work work work work work work work
aPlasmids were analyzed after electrophresison an agarose ~lasmid sizes werecalculatedby correlatingthem to the known sizes of the standard plasmid markers in Escherichiacoli V517 and using a double log plot of distance migrated versus size (in kb). cObtained after conjugation of pAM899-1 from strain UM899 (Forbes and Schaberg 1983) to strain ISP781.
All large nonconjugative plasmids, with the exception of pMH1182 and pMH2614 (Table 2), were found to express gentamicin resistance.
Restriction Digests Restriction digest bands from eight large nonconjugative plasmids showed a significant degree of variation after electrophoresis on agarose gels. The sizes (in kilobases) of most of these plasmids are given in Table 2. Digests of nonconjugative S. epidermidis plasmids with EcoRI (Fig. 1A) and XbaI (Fig. 2A) indicated that they had similar, but never identical, restriction patterns. This was even the case for plasmids that had identical molecular masses. Plasmids pMH1452 (50 kb), pMH1322 (51 kb), and pMH8702 (50 kb) (Fig. 2A, lanes 6, 7, and 9, respectively) show identical XbaI restriction digest patterns, with the exception that each had a band that was not present in the other two digests. Many plasmids shared common EcoRI (Fig. 1A) and XbaI (Fig. 2A) fragments, but no one fragment was common to all of the nonconjugative plasmids. In contrast, conjugative plasmids of similar molecular masses from the S. aureus NICU-CMH strains always had identical restriction patterns (Zorbas et al., 1988). Evidence has been provided to indicate that all NICUCMH plasmids used (Figs. 1 and 2) coded for gentamicin resistance, with the exception of pMH2614.
Plasmid DNA Hybridization Analyses An attempt was then made to identify tra gene sequences on the nonconjugative plasmids. A probe was made from the 6.3-kb EcoRI fragment of pGO2, which contains a major tra gene region (Gordon L. Archer, personal communication). This probe was then used to seek out tra genes on the large nonconjugative plasmids. However, w h e n the 6.3-kb probe was hybridized with an EcoRI digest of these plasmids, hybridization occurred with only one of seven plasmids (Fig. 1B). Gentamicin resistance was found to be expressed by all of these plasmids, with the exception of pMH2614. Note that a fragment of 6.3 kb can be seen in the EcoRI digests of several of the nonconjugative plasmids (Fig. 1A). For example, an EcoRI fragment of - 6.0 kb can be seen in pMH3902 (Fig. 1A, lane 8), yet it did not hybridize with the probe. This 6.0-kb fragment was originally thought to be analogous to the 6.3-kb EcoRI fragment of conjugative plasmids but with a 0.3-kb deletion (Zorbas et al. 1988). The data in Figure 1 show that this is obviously not the case. The exceptional nonconjugative plasmid was pMH6502. In the EcoRI digest seen in Figure 1, lanes 7, the probe not only hybridized very strongly with a 6.3-kb fragment but with two larger EcoRI fragments as well. It was also observed after agarose gel electrophoresis that a single band of near complete homology was quite evident between the 6.3-kb probe and the 6.3-kb EcoRI fragments of conjugative plasmids pAM899-1, pIZ7814, pCRG1600, and pGO2. Hybridization of the probe to pGO2 is shown in Figure 1B, lane 9. In contrast to that seen with pMH6502, homology is apparent with only the 6.3-kb EcoRI band of pGO2. The 6.3-kb probe was also hybridized to an XbaI restriction digest of conjugative and nonconjugative plasmids (Fig. 2). The probe hybridized most strongly with a large fragment of - 16.9 kb in size, which was present in all of the conjugative plasmids (Fig. 2, lanes 1-4). This XbaI fragment has already been shown to be a major tra region for pCRG1600 (Asch et al., 1984) and pJE1 (Evans and Dyke, 1988). The high degree of homology of the conjugative plasmids with the probe indicates the presence of homologous tra gene regions on the pGO2, pCRG1600, pIZ7814, and pAM899-1 plasmids (Fig. 2B, lanes 14, respectively). Alternatively, the probe did not hybridize with the nonconjugative plasmids (Fig. 2B, lanes 5-7 and 9), with the exception of pMH6502 (Fig. 2B, lane 8). All of these nonconjugative plasmids expressed gentamicin resistance.
DISCUSSION Self-mobilization of large plasmids was not observed in 15 of 15 strains of Gm r S. epidermidis isolates taken
M. Rogolsky and R. Gobert
232
A
1
2 3 4 5
6
789
B
"
~
"*"
=
=
"
_6.3
|
FIGURE 1. Agarose gel electrophoresis of EcoRI restriction endonuclease-digested DNA of the large plasmid from Gm r nonconjugative S. epidermidis clinical isolates (A). Filter hybridization of the EcoRI restriction digests with the 6.3-kb EcoRI fragment from the conjugative plasmid pGO2 as probe DNA (B). The probe DNA contains a major tra gene region. (Lanes 1) Phage lambda DNA, (lanes 2) pMH1442, (lanes 3) pMH1445, (lanes 4) pMH1322, (lanes 5) pMH2614, (lanes 6) pMH8702, (lanes 7) pMH6502, (lanes 8) pMH3902, and (lanes 9) pGO2. Molecular masses are in kilobases.
1
2
34
5 6 7 8
910
16. c 12.2 10.1
6.2 4..q
2.~
FIGURE 2. Agarose gel electrophoresis of XbaI restriction endonuclease-digested DNA of conjugative plasmids and the large plasmid from Gm r conjugative S. epidermidis clinical isolates (A). Filter hybridization of the XbaI restriction digests with the 6.3-kb EcoRI fragment from the conjugative plasmid pGO2 as probe DNA (B). The probe DNA contains a major tra gene region. (Lanes 1) pGO2, (lanes 2) pCRG1600, (lanes 3), pIZ7814, (lanes 4) pAM899-1, (lanes 5) pMH3902, (lanes 6) pMH1452, (lanes 7) pMH1322, (lanes 8) pMH6502, (lanes 9) pMH8702, and (lanes 10) phage lambda DNA. Molecular masses are in kilobases.
Plasmids of Staphylococcus epidermidis
from newborns in a neonatal intensive care unit. However self-transmissible plasmids among Gm" S. epidermidis clinical isolates were routinely observed by other laboratories (Archer and Johnston, 1983; Forbes and Schaberg, 1983; McDonnell et al., 1983; Archer et al., 1985). The fact that nine of nine NICUCMH Gm r S. aureus strains, which were all isolated along with the 15 nonconjugative strains, had selftransmissible plasmids (Zorbas et al. 1988) makes this situation even more puzzling. In a survey of 18 Gm r S. epidermidis isolates and 23 S. aureus Gm r isolates from a Virginia hospital, all of the S. aureus strains, but only 50% of the S. epidermidis strains, were found to be conjugative (Archer and Johnston, 1983). Thus, surveys from this and other investigations (Archer and Johnston, 1983; Archer et al., 1985; Zorbas et al., 1988) appear to indicate that Gm r strains of S. aureus clinical isolates are more likely to contain a self-transmissible plasmid that Gm r S. epidermidis clinical isolates. This is in spite of the fact that Gm ~ S. epidermidis strains are more prevalent than Gm r strains of S. aureus in a nosocomial environment (Weinstein et al., 1982; Hall et al., 1987). Transfer of seven large nonconjugative plasmids from the NICU-CMH strains by protoplast transformation demonstrated that they carried Gm r genes and were indeed Gm r plasmids that were similar in size to conjugative Gm ~ plasmids. It was then thought that a search for possible tra genes on the large plasmids might determine w h y these plasmids were not selftransmissible. Identical 6.3-kb EcoRI tra gene fragments appear to be present in all conjugative plasmids (Asch et al., 1984; Archer et al., 1986; Evans and Dyke, 1988; Zorbas et al., 1988). However, it was surprising to discover that this fragment only hybridized to one of eight of the large nonconjugative plasmids. It is therefore reasonable to conclude that seven of these large Gm ~ plasmids are not self-transmissible because they lack tra genes. Whereas these seven plasmids are devoid of tra genes, pMH6502 (Fig. 1B, lane 7) contains an excess of tra genes in comparison to prototype conjugative plasmids and still is not selftransmissible. The orientation of pMH6502 tra genes shows obvious variation from the consistent pattern that is present on prototype conjugative plasmids (Fig. 1B, lanes 7 and 9). This indicates that molecular rearrangements of pMH6502 have probably resulted in defective tra gene expression. Our laboratory is presently doing a molecular analysis of this plasmid to gain more insight into this possibility. Data from the hybridization studies indicated that complete or near complete 6.3-kb EcoRI segments from large nonconjugative Gm r plasmids are not necessarily
233
analogous to the common 6.3-kb EcoRI fragments of conjugative plasmids. Restriction digestion analyses showed that the nonconjugative plasmids did not share a high degree of relatedness with conjugative plasmids or, to a large extent, even with themselves. Alternatively, conjugative plasmids even from widely separated geographical areas share much physical relatedness (Schaberg et al., 1985; Zorbas et al., 1988; Thomas and Archer, 1989). In summary, it is concluded that the observations made with the NICU-CMH strains and those made by other investigators seem to indicate that the conjugative plasmids evolved from one or more common large ancestral Gm r plasmids and that S. epidermis was the natural environment for these plasmids. If the tra gene region was transposable (Thomas and Archer, 1989), this large ancestral Gm r plasmid might have served as the natural target replicon for this region. Transpositions and other inter- and intramolecular rearrangements of the ancestral plasmids over a long time period might have resulted in diversity of large Gm r plasmids in S. ep ~ idermidis strains. This molecular activity, especially in regard to transposition, could also have led to the depletion or duplication of any existing tra gene sequences. Such events might have molded the large plasmids in the NICU-CMH S. epidermidis strains. Alternatively, conjugative plasmids appear to belong to a family of highly related Gm r replicons. Because the Gm r plasmid is nearly always of the conjugative variety in the nosocomial S. aureus host, one might conclude that this plasmid was recently acquired by S. aureus from S. epidermidis donors. This highly conserved plasmid was most likely selected for transfer from the donor because it contained a functional set of tra genes, which were lacking on the other Gm r plasmids of the donor. This thinking is supported by the evidence that recent chemotherapy with aminoglycosides has broadened the nosocomial reservoir of Gm r plasmids in S. epidermidis (Weinstein et al., 1982). It can then be assumed that this was followed by dramatic increases in the transfer of specific Gm r plasmids to S. aureus after in vivo conjugation. Growth of the Gm r conjugative and nonconjugative plasmid populations is expected to continue to cause serious problems in the treatment of staphylococcal infections. Gratitude is extended to William von David, who helped us to modify the procedures for protoplast transformation and for recovery of large quantities of plasmid DNA. We are also grateful to G. L. Archer for helpful suggestions and discussion.
234
M. Rogolsky and R. Gobert
REFERENCES Archer GL, Johnston J (1983) Self-transmissible plasmids in staphylococci that encode resistance to aminoglycosides. Antimicrob Agents Chemother 24:70-77. Archer GL, Dietrick DR, Johnston J (1985) Molecular epidemiology of transmissible gentamicin resistance among coagulase-negative staphylococci in a cardiac surgery unit. J Infect Dis 151:24-251. Archer GL, Coughter JP, Johnston JL (1986) Plasmid encoded trimethoprim resistance in staphylococci. Antimicrob Agents Chemother 29:733-740. Asch DK, Goering RV, Ruff EA (1984) Isolation and preliminary characterization of a plasmid mutant derepressed for conjugal transfer in S. aureus. Plasmid 12:197202. Baumgart S, Hall SE, Campos JM, Polin RA (1984) Sepsis with coagulase-negative staphylococci in critically ill newborns. Am J Dis Child 137:461-467. Chen CW, Thomas Jr. CA (1980) Recovery of DNA fragments from agarose gels. Anal Biochem 101:339-341. Evans J, Dyke KGH (1988) Characterization of the conjugation system associated with the Staphyloccus aureus plasmid pJE1. J Gen Microbiol 134:1-8. Forbes BA, Schaberg DR (1983) Transfer of resistance plasmids from Staphylococcus epidermidis to Staphylococcus aureus: evidence for conjugative exchange of resistance. J Bacteriol 153:627-634. Goering RV, Ruff, EA (1983) Comparative analysis of conjugative plasmids mediating gentamicin resistance in Staphylococcus aureus. Antimicrob Agents Chemother 24:450452. Hall RT, Hall SL, Barnes WG, Izuegbu J, Rogolsky M, Zorbas I (1987) Characteristics of coagulase negative staphylococcus from infants with bacteremia. Pediatr Infect Dis J 6:377-383. Jaffe HW, Sweeney HM, Nathan C, Weinstein RA, Kabins SA, Cohen S (1980) Identity and interspecific transfer of gentamicin resistance plasmids in Staphylococcus aureus and Staphylococcus epidermidis. 141:738-746. Leary JJ, Brigati DJ, Ward DC (1983) Rapid and sensitive colorimetric method for visualizing biotin-labelled DNA probes hybridized to DNA or RNA immobilized on nitrocellulose: bio-blots. Proc Natl Acad Sci USA 80:40454044. Lindberg M (1981) Genetic studies in Staphylococcus aureus using protoplasts: cell fusion and transformation. In Staphylococci and Staphylococcal Infections. Ed., J Jeljaszewicz, Zbl Bakt Suppl 10. New York: Gustav Fisher Verlag, pp 535-541.
Lyon BR, Gillespie MT, Byrne ME, May JW, Skurray RA (1987) Plasmid-mediated resistance to gentamicin in Staphylococcus aureus: the involvement of a transposon. J Med Microbiol 23:101-110. Maniatis T, Fritsch EF, Sambrook J (1982) Molecular Cloning: A Laboratory Manual. New York: Cold Spring Harbor Laboratory. McDonnell RW, Sweeney HM, Cohen S (1983) Conjugational transfer of gentamicin resistance plasmids intra- and interspecifically in Staphylococcus aureus and Staphylococcus epidermidis. Antimicrob Agents Chemother 23:151-160. Meinkoth J, Wahl G (1984) Hybridization of nucleic acids immobilized on solid supports. Anal Biochem 138:267283. Parisi JT, Hecht DW (1980) Plasmid profiles in epidemiologic studies of infections by Staphylococcus epidermidis. J Infect Dis 141:637-643. Parisi JT (1985) Coagulase negative staphylococci and the epidemiological typing of Staphylococcus epidermidis. Microbiol Rev 49:126-139. Rigby PWJ, Dieckman M, Rhodes C, Berg P (1977) Labelling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J Mol Biol 113:237-252. Schaberg DR, Power G, Betzold G, Forbes BA (1985) Conjugative R plasmids in antimicrobial resistance of Staphylococcus aureus causing nosocomial infections. ] Infect Dis 152:43-49. Southern, E (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517. Thomas WD, Archer GL (1989) Identification and cloning of the conjugative transfer region of Staphylococcus aureus plasmid pGO1. J Bacteriol 171:684-691. Warren R, Rogolsky M, Wiley BB, Glasgow LA (1975) Isolation of extrachromosomal DNA for exfoliative toxin production from phage group II Staphylococcus aureus. J Bacteriol 122:99-105. Weinstein RA, Kabins SA, Nathan C, Sweeney HM, Jaffe HW, Cohen S (1982) Gentamicin resistant staphylococci as hospital flora: epidemiology and resistance plasmids. J Infect Dis 145:374-381. Zorbas I, Hall RT, Hall SL, Barnes WG, Rogolsky M (1988) Molecular analyses of conjugative, gentamicin-resistance plasmids from staphylococcal clinical isolates. Can J Microbiol 34:1050-1057.