Download A hemolysin-encoding plasmid contributes to bacterial

Investigative Ophthalmology & Visual Science, Vol. 33, No. 5, April 1992
Copyright © Association for Research in Vision and Ophthalmology
A Hemolysin-Encoding Plasmid Contributes to Bacterial
Virulence in Experimental Enterococcus
Faecalis Endophthalmitis
5. X. Stevens,* H. G. Jensen,* D. D. Jerr,t and M. 5. Gilmoret
Beta-hemolysin production is a variable trait of the Lancefield group D streptococcus, Enterococcus
faecalis. The E. faecalis hemolysin is encoded by large transmissible plasmids. The variable nature of
this putative virulence factor provided an ideal system for testing its contribution in experimental
endophthalmitis. In this study, isogenic E. faecalis strains were compared to determine whether the
presence of the hemolysin-encoding plasmid affected the severity of disease in a rabbit endophthalmitis
model. Experimental infections (n = 6) with lO'-lO 4 E. faecalis organisms harboring the hemolysin-encoding plasmid resulted in a 98% loss of retinal function (by electroretinography [ERG]) and white
reflex by postoperative day 3. By contrast, infections of similar numbers of plasmid-free E. faecalis
organisms (n = 5) resulted in retention of some retinal function (23% per ERG) with a red reflex
demonstrated on postoperative day 3. Results of light microscopy, slit-lamp examination, ERG, and
indirect ophthalmoscopy indicated that infections with hemolysin-encoding plasmid-containing E. faecalis resulted in a more aggressive endophthalmitis compared with the endophthalmitis caused by
plasmid-free E. faecalis. This is the first endophthalmitis model to the authors' knowledge that specifically evaluates bacterial virulence using isogenic strains. Invest Ophthalmol Vis Sci 33:1650-1656,
1992
Exogenous bacterial endophthalmitis is a rare and
devastating complication of intraocular surgery. The
incidence of exogenous bacterial endophthalmitis
ranges from 0.056-1.3%.l~s The visual outcome is
thought to be inversely proportional to the virulence
of the offending organism.1"6 Since 1979, 197 cases of
culture-positive cases of bacterial endophthalmitis
have been treated at the Dean A. McGee Eye Institute. Streptococci have been isolated from 23% of
these cases. Typically, streptococcal endophthalmitis
is fulminant and often leaves the patient with no vision in the affected eye. The Lancefield group D streptococcus, Enterococcus faecalis, was isolated from
seven patients with endophthalmitis. The visual outcome in four of these cases was no light perception.
From the *Dean A. McGee Eye Institute and the f Department of
Microbiology and Immunology, University of Oklahoma Health
Sciences Center, Oklahoma City, Oklahoma.
Supported in part by grants from the United States Public Health
Service (EY08289), The Gustavus and Louis Pfeiffer Foundation,
Research to Prevent Blindness, Inc. (New York, NY), and the private philanthropy of the citizens of Oklahoma.
Submitted for publication: April 11, 1991; accepted October 25,
1991.
Reprint requests: M. S. Gilmore, PhD, Department of Microbiology and Immunology, University of Oklahoma Health Sciences
Center, PO Box 26901, Oklahoma City, OK 73190.
The three remaining patients maintained hand motions visual acuity. E. faecalis has been reported to be
the primary etiologic agent in late endophthalmitis
associated with filtering blebs,7'8 and its presence has
been associated with endophthalmitis after trauma
and intraocular surgery.9"16 Only 15% of the reported
cases resulted in a final visual acuity of 20/200 or
better.
E. faecalis strains frequently harbor large transmissible plasmids that encode a beta-hemolysin (cytolysin).17'18 This hemolysin possesses general membrane
lytic properties and lyses both bacterial and eukaryotic cells. 19~21 It recently was shown that hemolysin production by E. faecalis is correlated with strains causing human parenteral infections.18 Moreover, hemolysin production has been shown experimentally to
contribute to the toxicity of E. faecalis in a mouse
intraperitoneal infection model.22
Because of the severity of E. faecalis endophthalmitis and the correlation between hemolysin production
and parenteral infection, we believed it was important
to determine whether the presence of the hemolysinencoding plasmid contributes to the course or severity
of endophthalmitis. As a first step in analyzing the
virulence properties of E. faecalis in intraocular infection, isogenic strains of E. faecalis, which differed
only in the presence of a hemolysin-encoding plas-
1650
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017
No. 5
E. FAECALIS ENDOPHTHALMITIS / Stevens er ol
mid, were compared in a rabbit endophthalmitis
model.
Materials and Methods
The isogenic E. faecalis strains used in this study
were JH2SS and JH2SS(pAM714). E. faecalis JH2SS
was selected as a spontaneous streptomycin- and
spectinomycin-resistant mutant23 of the plasmid-free
clinical isolate JH2.24 Plasmid pAM714 is a derivative
of the well-characterized hemolysin-encoding E. faecalis plasmid pADl,25"27 into which the erythromycin-resistant transposon Tn917 was inserted at a site
that does not affect hemolysin expression.28 The chromosomal resistances to streptomycin and spectinomycin and the hemolysin plasmid-associated resistance to erythromycin facilitated strain maintenance,
identification, and differentiation of recovered organisms. Strains JH2SS and JH2SS(pAM714) were provided by D. B. Clewell, PhD (University of Michigan,
Ann Arbor, MI). E. faecalis strains were cultivated
routinely in brain heart infusion (BHI; Difco, Detroit,
MI) at 37°C without aeration. Before intravitreal inoculation, each E. faecalis strain was washed twice by
centrifugation at 5000 X g for 10 min, resuspended in
1 volume of balanced salt solution (BSS; Dey, Napa,
CA), and diluted to the appropriate concentration
(per 0.1 ml of BSS). After inoculation of rabbit eyes,
the actual concentration of organisms present in the
inoculum was confirmed by plating serial dilutions of
bacterial cultures onto BHI supplemented with agar
(1.5% (w/v); Difco). Bacterial colonies were counted
after an 18-hr incubation at 37°C, and the original
inoculum was calculated. Quantitation of bacteria in
vitreous recovered from animals postinfection was
done similarly.
New Zealand white rabbits weighing 3-5 kg were
selected for these experiments. The animals were
cared for in accordance with the ARVO Resolution
on the Use of Animals in Research. To analyze the
course of infection, eight animals were injected with
JH2SS(pAM714), and seven animals received plasmid-free JH2SS. In addition, six animals were used to
quantify in situ bacterial growth, and nine animals
were used for additional absolute and surgical control.
To establish base values before experimentation, the
rabbits underwent general anesthesia with intramuscular ketamine (35 mg/kg) and xylazine (5 mg/kg)
followed by topical ocular anesthesia with proparacaine hydrochloride 0.5%. The rabbits were examined
by slit-lamp biomicroscopy, indirect ophthalmoscopy, and electroretinography (ERG). Scotopic
bright-flash ERG was done after pharmacologic mydriasis and 30 min of dark adaptation.
1651
The animals were allowed to recover for 72 hr before intravitreal paracentesis with instillation of E.
faecalis. Three doses of topical tropicamide 1% and
phenylephrine hydrochloride 2.5% were instilled 5
min apart and 30 min before intravitreal paracentesis.
The animals underwent general anesthesia, and the
eye was draped in a sterilefield.The globe was anesthetized with topical proparacaine hydrochloride 0.5%.
A 25-gauge needle attached to a 1.0-ml disposable
sterile syringe was introduced through the pars plana.
The bevel of the needle was visualized and placed posteriorly in the anterior vitreous. Approximately 0.15
ml of vitreous humor was aspirated slowly. The syringe was detached from the needle, and a second 1.0ml syringe containing E. faecalis in suspension was
attached. The bevel of the needle was placed anteriorly, and 0.1 ml of the E. faecalis suspension was infused slowly. The eye was examined immediately postinjection by indirect ophthalmoscopy to detect intraoperative complications. Slit-lamp biomicroscopy,
indirect ophthalmoscopy, and ERG were done 24 and
72 hr postoperatively. Cell and flare were established
using the highest slit-lamp magnification and the
smallest spot size. Cell was graded on a scale of 0-4+,
where 0 was defined as occasional cells; 1 +, 1-5 cells;
2+, 6-12 cells; 3+, 13-18 cells; and 4+, > 18 cells
with hypopyon. Rare was graded on a scale of 0-4+,
with 0 defined as none to 4+ defined as severe.
A range of inocula were used to determine the level
of E. faecalis required to establish endophthalmitis
within a period of 4 days. Instillation of 103-104 organisms occurred in only one eye. The matched eye
served either as a surgical control (which received intravitreal paracentesis followed by instillation of sterile BSS) or as an unoperated (absolute) control. The
rabbits infected with lO'-lO2 E. faecalis organisms
received intravitreal paracentesis and organisms in
both eyes. One eye received JH2SS(pAM714) strain
organisms, and the matched eye received an equivalent number of plasmid-free JH2SS. The clinical investigator was blinded as to the type of organism injected into each eye.
The vitreous of the eyes used for bacterial quantitation was harvested after the animals were killed with
pentobarbital and phenytoin, and the number of E.
faecalis organisms present was determined as follows.
The eyes were enucleated and flushed with sterile saline. A sterile 15° super blade (Storz, Inc., St. Louis,
MO) was used to make a 2-mra incision in the pars
plana. The scleral wall was cut with Wescott scissors
(Storz), and the vitreous was shelled into a sterile Petri
dish. The lens was separated from the vitreous, and
the vitreous was placed into a sterile test tube, minced
by passage through a 10-ml glass syringefittedwith an
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017
1652
INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / April 1992
18-gauge needle, and diluted to 5 ml with sterile BSS.
The vitreous samples then were plated, and the concentration of bacteria was determined as described
previously.
Vol. 33
Table 2. Percent loss of B-wave amplitude in eyes
that served as absolute or surgical controls
Absolute controls
Loss
Results
Several parameters were used to evaluate the relative severity of experimental endophthalmitis induced by strains JH2SS and JH2SS(pAM714).
ERG
Scotopic ERG b-wave amplitude results are presented in Tables 1 and 2. The b-wave amplitude diminished secondary to bacterial endophthalmitis
caused by either E.faecalis strain as anticipated. However, infections caused by hemolytic E. faecalis
JH2SS(pAM714) showed a significantly more rapid
loss in b-wave amplitude (mean ± standard error of
the mean, 48.0% ± 4.7% at 24 hr and 98.3% ± 1.0% at
72 hr) than infections caused by plasmid-free JH2SS
(15.1% ± 9.7% at 24 hr and 76.6% ± 7.9% at 72 hr; P
< 0.01 at 24 hr; P < 0.01 at 72 hr). For both groups,
no significant difference in the rate of b-wave ampliTable 1. Percent loss of B-wave amplitude in eyes
infected with Eterococcus faecalis, with and without
the hemolysin plasmid pAM714
E.faecalis JH2SS (pAM714) [hly+]
Eye
89-81
89-80
89-104
89-96
89-252
89-253
89-254
89-95
OD
OD
OD
OS
OD
OD
OD
OD
No. of
organisms
POD no. 1
104
47%
103
53%
POD no. 3
95%
100%
100%
NA
100%
NA
100%
102
102
10
45%
62%
50%
38%
23%
2
102
102
101
X ± SEM = 48.0% ± 4.7%
98.3% ±
E. faecalis JH2SS [hly-]
Loss
Rabbit no.
Eye
No. of organisms
POD#I
POD #3
89-83
89-82
89-104
89-252
89-253
89-254
89-95
OD
OS
OS
OS
OS
OS
OS
104
103
102
102
102
102
10'
25%
10%
43%
-14%
4%
-13%
51%
93%
78%
87%
NA
47%
NA
78%
X ± S E M = 15.1% ± 9.7%
Eye
89-96
89-80
89-81
89-82
90-227
90-230
90-231
90-234
OD
OS
OS
OD
OD
OS
OD
OS
POD no. 1
76.6% ± 7.9%
pAM714 = pADl +Tn917 not affecting hemolysin expression; hly, hemolysin; POD, postoperative day; OD, right eye; OS, left eye; NA, data not
available-animal died during anesthesia; SEM, Standard error of mean.
POD no. 3
18%
3%
52%
-10%
36%
5%
-30%
8%
12%
-81%
12%
X ± S E M = 5.9%± 14.1%
9.5% ± 7.7'
Surgical controls
Loss
Rabbit no.
Eye
89-83
90-227
90-228
90-228
90-229
90-229
90-230
90-231
90-232
90-232
90-233
90-233
90-234
OS
OS
OD
OS
OD
OS
OD
OS
OD
OS
OD
OS
OD
POD no. 1
POD no. 3
32%
37%
47%
-7%
23%
21%
34%
14%
12%
12%
7%
31%
25%
19%
18%
19%
5%
25%
X ± SEM = 23.3% ± 4.3%
% Loss
Rabbit no.
Rabbit no.
12.1% ±4.2%
POD, postoperative day; OD, right eye; OS, left eye; SEM, standard error of
mean.
tude loss was observed for differing inoculum sizes
over the range of 102-104 organisms (Table 1). Eyes
infected with plasmid-free JH2SS did not have a significant reduction in b-wave amplitude 24 hr postinfection compared with surgical or absolute control eyes
(P > 0.3), although eyes infected with JH2SS(pAM714) did (P < 0.01). Eyes from both infection
groups had a significant reduction in b-wave amplitude at 72 hr over those of control eyes (P < 0.01).
Indirect Ophthalmoscopy
Indirect ophthalmoscopy was done preoperatively,
immediately postoperatively, and on postoperative
days 1 and 3. There was no evidence of retinal detachment or vitreous bleeding immediately after intravitreal injection. Postoperative day 1 results revealed an
average 2+ vitreous cell and flare for both groups. An
excellent red reflex was also present. At this time, it
was impossible to distinguish which eye had been infected with the hemolysin plasmid-containing E. fae-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017
No. 5
1653
E. FAECALIS ENDOPHTHALMITIS / Srevens er ol
calis strain. On postoperative day 3, a clear view of the
retina was not possible in any of the eyes as a result of
marked vitreous inflammation, anterior segment fibrin clot formation, and rarely, irregular small pupils
with posterior synechiae. On day 3, all three eyes that
received 102 or fewer plasmid-free JH2SS had a faint
red reflex on examination. By contrast, only one of
four eyes that received an equivalent number of
JH2SS(pAM714) had a faint red reflex on day 3 by
indirect ophthalmoscopy.
Slit-Lamp Examination
Injection of 102 or fewer organisms of either E. faecalis strain resulted in minimal anterior chamber cell
and flare 24 hr postinoculation. When 103 or 104 colony-forming units per 0.1 ml of JH2SS(pAM714)
were injected, an intense anterior chamber reaction
was observed at 24 hr, with a 4+ cell and flare and
hypopyon. The eye receiving 103 colony-forming
units per 0.1 ml of plasmid-free JH2SS had minimal
flare and no cells at 24 hr. However, the one receiving
104 organisms resulted in a more marked anterior
chamber reaction with a 4+ cell and flare. A red retinal reflex was observed in all animals 24 hr postinoculation.
Slit-lamp examination at 72 hr showed an intense
anterior chamber reaction with 4+ cell and flare and
hypopyon, regardless of the amount or strain of E.
faecalis injected. Although inflammation and a fibrin
clot were present in the anterior chamber of all eyes
infected, a more intense vitreous reaction was observed in those eyes injected with JH2SS(pAM714),
as defined by the red reflex. Generally, a red reflex
could be seen in eyes inoculated with JH2SS but not
those that received JH2SS(pAM714). The cornea was
edematous with numerous mutton-fat keratic precipitates in all eyes 72 hr postinoculation.
Light Microscopy
Light microscopic examination of the specimens
that received JH2SS(pAM714) revealed a more
marked vitritis, posterior retinitis, and optic neuritis
24, 48, and 72 hr postinoculation compared with
matched paired eyes receiving an equal number of
JH2SS (Fig. 1). The eyes that received JH2SS(pAM714) had more areas in which the posterior retina had undergone autolytic changes.
Eyes infected with JH2SS(pAM714) had intense
vitritis with early involvement of the posterior retina
at 24 hr and progressive involvement through 48 hr.
Numerous areas of the retinal architecture were disorganized and replaced with inflammatory cells by 72
hr. By contrast, eyes receiving the plasmid-free strain
JH2SS had a mild vitritis with no involvement of the
posterior retina at 24 hr. This progressed to a moderate vitritis with a few foci of inflammatory cells contained in the retina at 48 hr. At 72 hr, a severe vitritis
was observed in eyes infected with JH2SS, with inflammatory cells in the retina. However, the retinal
architecture was largely intact.
Bacterial Quantitation
Figure 2 depicts the relative rates of bacterial
growth in eyes infected with JH2SS(pAM714) and
plasmid-free JH2SS. Values at each time represent
the mean of at least two serial dilutions plated in duplicate. The data show similar growth kinetics for
both strains during the course of infection.
Discussion
All intraocular infections are sight threatening; however, severe cases generally are associated with more
virulent organisms. The severity of infection is related
to several variables: duration of infection, size of inoculum, previous therapy, toxic byproducts, associated
ocular damage, and the individual immune response.29 Endophthalmitis associated with organisms
such as Bacillus cereus, streptococcal species, E. faecalis, Staphylococcus aureus, Pseudomonas aeruginosa, and Serratia marcescens generally results in poor
final visual acuity even with multiple vitrectomies
and repeated instillation of intravitreal antibiotics.30"33 The severity of intraocular infection from
these organisms is thought to result largely from the
ability of these bacteria to establish an intraocular infection rapidly with the production of extracellular
cytotoxins.32'34 We tested this hypothesis using isogenic strains of E. faecalis differing only in the presence of a hemolysin-encoding plasmid. The ERG results revealed that eyes inoculated with JH2SS(pAM714) showed a significantly greater loss in ERG
b-wave amplitude after 24 hr than eyes inoculated
with an equivalent number of plasmid-free JH2SS. By
contrast, eyes inoculated with JH2SS showed a
change in b-wave amplitude at 24 hr that was not
significantly different than absolute and surgical control eyes. No residual retinal function was detected at
72 hr in all except two eyes that received JH2SS(pAM714). The two eyes with residual function
showed 95% reduction in response at 72 hr. Reduction in b-wave amplitude would be expected with
medial opacification, but the near total loss demonstrated in these eyes suggests direct retinal involvement in the infection, as was found histopathologically. All eyes that received plasmid-free JH2SS retained significantly greater residual retinal function
and a red reflex.
Indirect ophthalmoscopy, slit-lamp biomicroscopic
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017
1654
INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / April 1992
Vol. 33
Fig. 1. (A-Q Progressive changes in rabbit retinas at 24,48, and 72 hr from eyes injected intravitreally with 102 colony forming units (cfu's)
of the hemolytic strain EnterococcusfaecalisJH2SS (pAM714). After 72 hr, the retinal architecture was disorganized and largely replaced with
inflammatory cells. Hematoxylin-eosin; original magnification, X63. (D-F) Contrasting changes in rabbit retinas at 24, 48, and 72 hr from
eyes injected intravitreally with 102 cfu's of the plasmid-free strain E.faecalis JH2SS. The retinas remain largely intact throughout the time
period, with inflammatory cells beginning to infiltrate at 72 hr. Vitritis is apparent at 72 hr (F). Hematoxylin-eosin; original magnification
X63,
examination, and light microscopic pathologic results
correlated well with the ERG findings. Slit-lamp and
indirect ophthalmoscopic examinations demonstrated that eyes inoculated with JH2SS(pAM714)
were highly inflamed at 72 hr, with marked anterior
segment inflammation and a white reflex on examination. By contrast, eyes that received plasmid-free
JH2SS frequently showed severe anterior segment in-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017
E. FAECAUS ENDOPHTHALMITIS / Srevens er ol
No. 5
24
48
72
Time post-infection (hr)
Fig. 2. Quantitation of bacteria through course of experimental
Enterococcus faecalis endophthalmitis. • = hemolytic E. faecalis
JH2SS (pAM714); • = nonhemolytic, E. faecalis JH2SS. Concentrations of organims are mean of at least two 10-fold serial dilutions
performed in duplicate.
flammation by 72 hr; however, a reduced but easily
detected red reflex was still observed. In both groups,
the posterior chamber was full of inflammatory cells,
preventing a direct view of the retina at 72 hr postinoculation. Histopathologically, experimental infections with hemolytic JH2SS(pAM714) showed more
severe inflammation at 24,48, and 72 hr than did eyes
infected with plasmid-free JH2SS. A severe inflammatory response with virtually total retinal disintegration with few infiltrating inflammatory cells was observed in stained retinal sections from eyes infected
with the nonhemolytic strain. Taken together, these
results show that the endophthalmitis associated with
hemolysin-producing E. faecalis was a more aggressive infection, resulting in more rapid loss in all diagnostic measures of potential vision than endophthalmitis associated with the nonhemolytic E. faecalis.
Eyes infected with the nonhemolytic E. faecalis may
have salvageable vision 72 hr postinfection, based on
the observation of red reflex and residual ERG bwave values. They may recover useful vision if treated
aggressively with vitrectomy and intravitreal antibiotics.
Although we showed that the hemolysin-encoding
plasmid-containing E. faecalis strain produced a
quantitatively and qualitatively more severe infection, the mechanism by which the hemolysin may
contribute to retinal damage is uncertain. In addition
to direct cytotoxic effects, hemolysin production may
affect the retina indirectly by mediating a more intense inflammatory reaction, which, in turn, may aid
in the destruction of the retinal architecture. Increased inflammation is associated with a more virulent intraocular infection, and its occurrence during
1655
endophthalmitis is thought to play a large role in retinal destruction.113536 Intraocular inflammation has
been shown to increase even after the vitreous has
become sterile in experimentally induced infections,
suggesting that bacterial products initiate inflammation.37"40 Additionally, other plasmid-encoded functions may contribute to the course and severity of the
endophthalmitis because the plasmid we used encodes other functions, including those related to conjugal plasmid transfer. Experiments using isogenic E.
faecalis strains differing only in expression of hemolysin-encoding related genes will be required to relate
unambiguously the pathogenesis of fulminant E. faecalis endophthalmitis to hemolysin expression. Even
so, the observed differences in infection severity are
not likely to be the result of an intraocular growth
advantage for the hemolytic strain because cultures
taken at regular intervals revealed similar bacterial
growth rates for hemolytic and nonhemolytic strains
(Fig. 2).
The virulence of enterococci has been the subject of
debate. Enterococci generally are regarded as commensal flora; however, they are among the leading
causes of nosocomial infections in the United
States.41 Moreover, the incidence of plasmid-encoded
hemolysin production in clinical isolates of E. faecalis
is 40-70%, suggesting a role for hemolysin in their
pathogenesis.1842 In a mouse model, it was shown that
an approximately tenfold decrease occurred in the
lethal dose for 50% of animals that received hemolytic
E. faecalis compared with an isogenic nonhemolytic
strains.22 The results of this study support the role of
E. faecalis hemolysin as a virulence factor.
Key words: Enterococcus faecalis, endophthalmitis, cytolysin, beta-hemolysin
Acknowledgments
The authors thank Scottye Davis and Mark Dittmar for
technical assistance.
References
1. Rowsey JJ, Newsom DL, Sexton DJ, and Harms WK: Endophthalmitis: Current approaches. Ophthalmology 89:1055,
1982.
2. Mandelbaum S and Forster RK: Postoperative endophthalmitis. Int Ophthalmol Clin 27:95, 1987.
3. Klimek JJ, Ajemian E, Andrews L, Hryb K, and Hill DA: Outbreak of bacterial endophthalmitis after cataract surgery and
lens implantation: Lack of direct evidence for exogenous contributing factors. Am J Infect Control 14:184, 1986.
4. Weber DJ, Hoffman KL, Thoft RA, and Baker AS: Endophthalmitis following intraocular lens implantation: Report of 30
cases and review of the literature. Rev Infect Dis 8:12, 1986.
5. Carney M and Peyman GA: Vitrectomy in endophthalmitis.
Int Ophthalmol Clin 27:127, 1987.
6. Jacobs DR and Cohen HB: The inflammatory role of endo-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017
1656
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / April 1992
toxin in rabbit gram-negative bacterial endophthalmitis. Invest
Ophthalmol Vis Sci 25:1074, 1984.
Mandelbaum S, Forster RK, Gelender H, and Culbertson W:
Late onset endophthalmitis associated withfilteringblebs. Ophthalmology 92:964, 1985.
Katz LJ, Cantor LB, and Spaeth GL: Complications of surgery
in glaucoma. Ophthalmology 92:959, 1985.
Affeldt JC, Flynn HW, Forster RK, Mandelbaum S, Clarkson
JG, and Jarus GD: Microbial endophthalmitis resulting from
ocular trauma. Ophthalmology 94:407, 1987.
Driebe WT, Mandelbaum S, Forster RK, Schwartz LK, and
Culbertson WW: Pseudophakic endophthalmitis: Diagnosis
and management. Ophthalmology 93:442, 1986.
Mandelbaum C and Forster RK: Postoperative endophthalmitis. Ophthalmic Forum 3:70, 1985.
Eichenbaum DM, Jaffe NS, dayman HM, and Light DS: Pars
plana vitrectomy as a primary treatment for acute bacterial
endophthalmitis. Am J Ophthalmol 86:167, 1978.
Leveille AS, McMullan FD, and Cavanagh HD: Endophthalmitis following penetrating keratoplasty. Ophthalmology
90:38, 1983.
Croxatto JO, Galentine P, Cupples HP, Harper D, Reader A,
and Zimmerman LE: Sympathetic ophthalmia after pars plana
vitrectomy-lensectomy for endogenous bacterial endophthalmitis. Am J Ophthalmology 91:342, 1981.
Abel R Jr, Binder PS, and Bellows R: Postoperative bacterial
endophthalmitis: Section III. Ann Ophthalmol 8:1253, 1976.
Jones S, Cohen EJ, Arentsen JJ, and Laibson PR: Ocular streptococcal infections. Cornea 7:295, 1988.
Ike Y and Clewell DB: High incidence of hemolysin production by Streptococcus faecalis strains associated with human
parenteral infections: Structure of hemolysin plasmids. In
Streptococcal Genetics, Ferretti J and Curtiss R, editors. Washington DC, American Society for Microbiology, 1987,pp.159—
164.
Ike Y, Hashimoto H, and Clewell DB: High incidence of hemolysin production by Enterococcus (Streptococcus) faecalis
strains associated with human parenteral infection. J Clin Microbiol 25:1524, 1987.
Basinger SF and Jackson RW: Bacteriocin (hemolysin) of
Streptococcus zymogenes. J Bacteriol 96:1895, 1968.
Brock TD and Davie JM: Probable identity of group D hemolysin with bacteriocin. J Bacteriol 86:708, 1963.
Clewell DB: Plasmids, drug resistance, and gene transfer in the
genus Streptococcus. Microbiol Rev 45:409, 1981.
Ike Y, Hashimoto H, and Clewell DB: Hemolysin of Streptococcus faecalis subspecies zymogenes contributes to virulence
in mice. Infect Immun 45:528, 1984.
Tomich PK, An FY, and Clewell DB: Properties of erythromycin-inducible transposon Tn97 7 in Streptococcus faecalis. J
Bacteriol 141:1366, 1980.
Jacob AE and Hobbs S: Conjugal transfer of plasmid-borne
multiple antibiotic resistance in Streptococcus faecalis var. zymogenes. J Bacteriol 117:360, 1974.
Clewell DB, Tomich PK, Gawron-Burke MC, Franke AE, Yagi
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
Vol. 33
Y, and An FY: Mapping of Streptococcus faecalis plasmids
pADl and pAD2 and studies relating to transposition of
Tn917. J Bacteriol 152:1220, 1982.
Ehrenfeld EE and Clewell DB: Transfer function of Streptococcus faecalis plasmid PADl: Organization of plasmid DNA encoding response to sex pheromone. J Bacteriol 169:3473,1987.
Tomich PK, An FY, Damle SP, and Clewell DB: Plasmid-related transmissibility and multiple drug resistance in Streptococcus faecalis subsp. zymogenes strain DS16. Antimicrob
Agents Chemother 15:828, 1979.
Ike Y and Clewell DB: Genetic analysis of pADl pheromone
response in Streptococcus faecalis using transposon Tn977 as
an insertional mutagen. J Bacteriol 158:777, 1984.
Diamond JG: Changing concepts in treatment of endophthalmitis. Ophthalmic Forum 3:66, 1985.
Rowsey JJ: Diagnosis of endophthalmitis. Ophthalmic Forum
3:67, 1985.
Rowsey JJ, Jensen H, and Sexton DJ: Clinical diagnosis of
endophthalmitis. Int Ophthalmol Clin 27:82, 1987.
O'Day DM, Smith RS, Gregg CR, Turnbull PCB, Head WS,
Ives JA, and Ho PC: The problem of Bacillus species infection
with special emphasis on the virulence of Bacillus cereus. Ophthalmology 888:833, 1981.
Davey RT Jr and Tauber WB: Posttraumatic endophthalmitis:
The emerging role of Bacillus cereus infection. Rev Infect Dis
9:110, 1987.
Peyman GA, Vastine DW, and Raichand M: Symposium: Postoperative endophthalmitis, experimental aspects and their
clinical application. Ophthalmology 85:374, 1978.
Graham RO and Peyman GA: Intravitreal injection of flexamethasone: Treatment of experimentally induced endophthalmitis. Arch Ophthalmol 92:149, 1974.
Meredith TA, Aguilar EE, Miller MJ, Gardner SK, Trabelsi A,
and Wilson LA: Comparative treatment of experimental Staphylococcus epidermidis endophthalmitis. Arch Ophthalmol
108:857, 1990.
Meredith TA, Trabelsi A, Miller MJ, Aguilar E, and Wilson
LA: Spontaneous sterilization in experimental Staphylococcus
epidermidis endophthalmitis. Invest Ophthalmol Vis Sci
31:181, 1990.
Meyers-Elliot RH and Dethlefs BA: Experimental Klebsiellainduced endophthalmitis in the rabbit. Arch Ophthalmol
100:1959, 1982.
Kreger AS and Gray LD: Purification of Pseudomonas aeruginosa proteases and microscopic characterization of pseudomonal protease-induced rabbit corneal damage. Infect Immun
19:630, 1978.
Forrester JV, Worgul BV, and Merriam GR Jr: Endotoxin-induced uveitis in the rat. Graefes Arch Clin Exp Ophthalmol
213:221, 1980.
Centers for Disease Control: Infection surveillance summaries.
MMWR35(Suppl 1SS):17SS, 1984.
Huycke MM, Spiegel CA, and Gilmore MS: Bacteremia caused
by hemolytic, high-level gentamicin-resistant Enterococcus
faecalis. Antimicrob Agents Chemother 35:1626, 1991.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933167/ on 05/04/2017