Download Staining characteristics and antiviral activity of

Staining Characteristics and Antiviral Activity of
Sulforhodamine B and Lissamine Green B
James Chodosh,* Richard D. Dix,* R. Clark Howell* William G. Stroop,\
and Scheffer C. G. Tseng*
Purpose. Fluorescein and rose bengal are dyes used routinely in the examination of the ocular
surface. As part of an ongoing search for a superior ophthalmic dye with optimal specificity
and sensitivity and a lack of interference with subsequent viral cultures, and as part of studies
that use chemical dyes to understand better the pathophysiology of ocular surface disorders,
the staining characteristics and antiviral activity of sulforhodamine B and lissamine green B
were investigated.
Methods. Staining of rabbit corneal epithelial cell cultures by sulforhodamine B and lissamine
green B was compared to that of fluorescein and rose bengal. Diffusion of each dye through a
collagen gel was measured. Uptake of lissamine green B by herpes simplex virus type 1 (HSVl)-infected Vero cell cultures was compared at several times postinfection. The effect of sulforhodamine B and lissamine green B on HSV-1 plaque formation in Vero cells was determined.
The cellular toxicity of sulforhodamine B and lissamine green B in vitro was examined by a
quantitative 14C-amino acid uptake assay and by a qualitative cell viability assay. Finally, the
effect of sulforhodamine B and lissamine green B on viral replication was compared in vivo
with that of rose bengal in a rabbit model of herpetic epithelial keratitis.
Results. Rose bengal vividly stained cell monolayers of explant cultures of rabbit corneal epithelium. By light microscopy, sulforhodamine B and lissamine green B, like fluorescein, did not
stain the epithelial cells, but did stain the corneal explant stroma. Pretreatment of epithelial
cells with 0.25% trypsin for 5 minutes failed to induce dye uptake; however, pretreatment with
0.5% Triton X-100 for 5 minutes resulted in nuclear staining by lissamine green B, but not
sulforhodamine B. When added to a collagen gel, the relative diffusion rate was fluorescein >
lissamine green B > sulforhodamine B > rose bengal. By spectrophotometric analysis, HSV-1infected and uninfected Vero cells bound equivalent amounts of lissamine green B until late in
infection, when infected cells took up more dye (P < 0.001). A direct neutralization assay
showed that 0.06% lissamine green B or 0.5% sulforhodamine B reduced HSV-1 plaque formation in Vero cells by greater than 50%, when present at the time of viral adsorption. By a
quantitative 14C-amino acid uptake assay, lissamine green B was toxic to Vero cells in a dose-dependent manner, whereas sulforhodamine B was relatively nontoxic at the concentrations
tested. By a cell viability assay, however, neither dye showed significant cellular toxicity. In a
rabbit model of herpetic epithelial keratitis, rose bengal significantly reduced viral replication
and recovery, whereas sulforhodamine B and lissamine green B had no effect.
Conclusions. Neither sulforhodamine B nor lissamine green B stain healthy, normal cells. Lissamine green B stains membrane-damaged epithelial cells, but sulforhodamine B does not. Both
sulforhodamine B and lissamine green B stain corneal stroma. Lissamine green B inhibits
HSV-1 plaque formation at low concentrations of dye in vitro, which correlates with suppression of cellular metabolism as demonstrated by a 14C-amino acid uptake assay, but does not
From the *Bascom Palmer Eye Institute, Department of Ophthalmology, University
of Miami School of Medicine, Miami, Florida, and the -\Cullen Eye Institute,
Department of Ophthalmology, and Division of Molecular Virology, Baylor College
of Medicine, and Ophthalmology Research Laboratory, Houston Veterans Affairs
Medical Center, Houston Texas.
Presented in part at the Association for Research in Vision and Ophthalmology
Annual Meeting, Sarasota, Florida, May 2-7, 1993.
Supported by National Institutes of Health grarits NINCDS 1-PO-NS25569
(RDD), DC 01706 (WGS), NEIEY06819 (SCGT), and EY02I80 (Bascom Palmer
1046
Eye Institute); and the Department of Veterans' Affairs (WGS). WGS holds a Jules
and Doris Stein Research to Prevent Blindness Professorship. JC was supported by
a fellowship from the Heed Ophthalmic Foundation for 1992-1993.
Submitted for publication June 8, 1993; revised September 22, 1993; accepted
October 7, 1993.
Proprietary interest category: N.
Reprint requests: Scheffer C. G. Tseng, Bascom Palmer Eye Institute, 900 N.W. 17
Street, Miami, FL 33136.
Investigative Ophthalmology & Visual Science, March 1994, Vol. 35, No. 3
Copyright © Association for Research in Vision and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
1047
Sulforhodamine B and Lissamine Green B
affect cell viability. Neither sulforhodamine B nor lissamine green B inhibit viral replication or
recovery in vivo. Invest Ophthalmol Vis Sci. 1994; 35:1046-1058.
£ luorescein and its halide derivative, rose bengal, are
invaluable in the evaluation of ocular surface diseases.
Research has focused on the staining properties and
specificity of these two dyes,1'2 and on the antiviral
nature of rose bengal.3 Studies have shown that staining of the cornea by fluorescein occurs when surface
epithelial cell-to-cell junctions are disrupted.1 In contrast, rose bengal stains all healthy, normal cells in vitro, but does not stain the normal, healthy corneal or
conjunctival epithelium in vivo, because the normal
preocular tear film blocks access to the cells.12
Although fluorescein and rose bengal are the two
ophthalmic dyes used routinely in clinical practice in
the United States today, their clinical superiority to
other available dyes remains unproven. For example,
because of the intrinsic antiviral effect of rose bengal,
its application in herpetic epithelial keratitis before
viral culture has been discouraged.3'4 One dye thought
to be similar in its staining properties to rose bengal is
lissamine green B.5 It is not known, however, whether
lissamine green B shares the antiviral capacity of rose
bengal. Another dye, sulforhodamine B, has been reported to be superior to fluorescein for the visualization of both preocular tear film and conjunctival epithelial lesions.6 Although sulforhodamine B, a
member of the aminoxanthene dye family, structurally
resembles fluorescein and rose bengal, which are hydroxyxanthene dyes, the staining properties of sulforhodamine B have not been determined.
As part of an ongoing search for a superior ophthalmic dye with optimal specificity and sensitivity and
lack of antiviral activity, and as part of studies using
chemical dyes to understand better the pathophysiology of ocular surface disorders, we investigated the
staining characteristics and antiviral activity of both
sulforhodamine B and lissamine green B. As a first
step in the evaluation of sulforhodamine B and lissamine green B, we compared their staining with that of
fluorescein and rose bengal in rabbit corneal epithelial
(RCE) cell explant cultures. Because lissamine green B
has been reported to stain herpetic epithelial keratitis,5 we compared the binding to herpes simplex virus
type 1 (HSV-l)-infected and uninfected Vero cells in
vitro. We then investigated the effect of sulforhodamine B and lissamine green B on HSV-1 plaque formation in vitro. We examined the cellular toxicity of
each dye by a quantitative 14C-amino acid uptake assay
and by a qualitative cell viability assay. Finally, we compared the antiviral effect of sulforhodamine B, lissamine green B, and rose bengal in a rabbit model of
herpetic epithelial keratitis.
MATERIALS AND METHODS
Cell Cultures
Primary RCE cell cultures were obtained, as previously reported,2 using an explant culture of an approximately 1 X 1 X 1 mm3 piece of endothelium-free but
epithelium-containing rabbit cornea. Explants were
cultured in 100-mm2 chambers (Nunc, Inc., Naperville, IL), containing Dulbecco's modified Eagle's medium (DMEM/F-12), 2.5 mg/ml amphotericin B, 5%
fetal bovine serum (GIBCO, Inc., Grand Island, NY),
0.5% dimethyl sulfoxide, 2 ng/ml epidermal growth
factor, 33 ng/ml cholera toxin, 1 mg/ml insulin, and
50 mg/ml gentamicin sulfate (Sigma Chemical, Inc.,
St. Louis, MO). After 9 to 12 days of culture, epithelial
cells had migrated from the explant and formed a confluent cell layer with areas of stratification.
Vero cells, a continuous line of African green
monkey kidney cells, were cultured in 25-cm2 flasks
containing DMEM with 10% fetal bovine serum, 2.5
mg/ml amphotericin B, and 50 mg/ml gentamicin. All
cultures were maintained at 37°C in 5% CO2 and 95%
humidity.
All animals used in this study were handled in accordance with the National Institutes of Health
"Guide for the Care and Use of Laboratory Animals"
and the ARVO Resolution on the Use of Animals in
Research. Studies adhered to the tenets of the Declaration of Helsinki.
Virus
The HI 29 strain of HSV-1, used throughout these experiments, was originally isolated from a fatal case of
herpes encephalitis by Klassen and colleagues. Dix et
al have shown that HI29 is virulent in mice after peripheral infection.7'8 Using thymidine kinase assays, we
have shown HI29 to be thymidine kinase positive (data
not shown). In a rabbit model of acute and immunosuppression-induced reactivated infection, HI 29
causes severe epithelial keratitis and focal temporal
lobe necrotizing encephalitis.9"11
Comparison of Dye Uptake by Rabbit Corneal
Epithelial Cells
Sulforhodamine B, lissamine green B (Sigma), rose
bengal (Aldrich, Milwaukee, WI), and fluorescein (Alcon, Fort Worth, TX) were diluted with calcium-free,
phosphate-buffered saline (PBS). RCE cell explant
cultures were washed once with PBS. Serial dilutions
of each dye were applied to each culture well for 5
minutes, followed by three washes with PBS. The cul-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
1048
Investigative Ophthalmology & Visual Science, March 1994, Vol. 35, No. 3
ture chamber slides were examined grossly and by
light microscopy, and photographed. Because sulforhodamine B and lissamine green B failed to stain RCE
cell monolayers (see Results), additional monolayers
were pretreated for 5 minutes with 0.25% trypsin in
PBS with 0.02% ethylenediamine tetraacetic acid
(Sigma), followed by the addition of fetal bovine serum
to stop the enzymatic action of the trypsin, or with
0.5% Triton X-100 for 5 minutes before dye application.12
Diffusion of Dyes in Collagen Gel
As shown in Results, sulforhodamine B and lissamine
green B appeared to stain rabbit corneal explant
stroma, and thus the relative ability of each dye to
diffuse through a collagen gel was tested.1 Rat tail tendons were obtained and sterilized in 70% ethanol at
4°C for 24 hours. Collagen was extracted for 48 hours
at 4°C in 0.1% acetic acid solution (150 ml/g tendon),
then filtered through sterile gauze, and centrifuged at
16,000g for 1 hour at 4°C. To form the gel, the collagen supernatant was mixed simultaneously with 0.34
M NaOH and XI0 concentrated DMEM to a final ratio of 20:1:2 (v/v). The mixture was dispensed in test
tubes and incubated at 37°C for 1 hour to allow it to
become gelatinous. One milliliter of dye was added to
each tube and allowed to sit for 30 minutes before
three washes of the surface of the gel with PBS. Photographs of the gel-filled tubes were taken at several
times postapplication of dye.
Spectrophotometric Analysis of Lissamine
Green B Binding
Flask cultures (25 cm2) of Vero cells were washed once
with PBS, and then infected with HSV-1 at an approximate multiplicity of infection of ten plaque-forming
units (PFU)/cell. Control cell cultures were mock-infected with PBS. After an adsorption period of 1 hour,
cultures were washed with PBS, and incubated. At 6
and 12 hours postinfection, triplicate cultures of infected and uninfected cells were stained with 1% lissamine green B, rinsed three times with PBS, carefully
scraped from the flasks, and resuspended in 5 ml of
PBS. Each sample's cells were counted by hemocytometry to ensure equal numbers of cells in all flasks. The
cells were then centrifuged at 500g for 5 minutes,
lysed in 2 ml of 50% ethanol in PBS, vortexed, and
centrifuged again. The absorbance at 630 nm was
measured on 1 ml of the supernatant from the lysed
cells. For the 24 hours postinfection samples, cells
from the infected cultures were gently knocked loose
from the flasks, centrifuged, resuspended in PBS,
counted, centrifuged, and stained with 1% lissamine
green B for 5 minutes. Uninfected control cell cultures were trypsinized, centrifuged, resuspended in
PBS, counted, centrifuged, and stained with 1% lissamine green B. Both infected and uninfected sample
were thrice washed gently with PBS followed by centrifugation, before lysis in 50% ethanol in PBS, centrifugation, and quantitative spectrophotometry. The
data were analyzed by a two-factor analysis of variance
in which infection was one variable and day of infection was a blocking variable.
HSV-1 Plaque Reduction
Vero cells were cultured in six-well plates to confluence. Serial dilutions of sulforhodamine B or lissamine green B in PBS were mixed 1:1 with 100 PFU of
HSV-1 in PBS and allowed to adsorb onto the cells for
1 hour. Each culture was then washed three times with
PBS, overlayed with methylcellulose, and incubated
for 3 days, at which time plaques were counted.
Cellular Toxicity
Because, as shown in Results, both lissamine green B
and sulforhodamine B reduced HSV-1 plaque formation, we wished to determine whether the apparent
antiviral effect was the result of cellular toxicity of the
dyes. Uptake of 14C-amino acids into acid-precipitable
protein was used to assess the effect of lissamine green
B and sulforhodamine B on Vero cell metabolism, as
described previously.13 Individual monolayers of Vero
cells were treated in duplicate with serial dilutions of
each dye for 1 hour, washed four times with PBS, and
then incubated for 23 hours in DMEM containing
one-tenth the normal concentration of amino acids,
and supplemented with 3 to 4 mCi/ml of 14C-amino
acid mixture (Amersham, Inc., Arlington Heights, IL).
Cells were harvested by scraping, then pelleted, and
washed twice in PBS. The final cell pellet was resuspended in 0.2 ml of PBS, frozen at -70°C, and then
thawed. Next, 0.01 ml of each sample was spotted in
duplicate on GF/C Whatman filter paper, dried for 15
minutes at 150°C, and washed twice at 4°C with trichloroacetic acid for a total of 30 minutes. This was
followed by three 10-minute washes at 4°C with 95%
ethanol. After drying at 150°C, 2.5-cm squares of
filter paper containing acid-unsoluble material were
placed in vials containing 7.0 ml of scintillation fluid.
Radioactivity was counted on a Beckman (Columbia,
MD) LS5801 scintillation counter.
Staining Characteristics and Antiviral Activity
in an Animal Model of Herpetic Epithelial
Keratitis
Four- to 5-pound, specific pathogen-free, New Zealand White rabbits were obtained from Myrtle's Rabbitry (Thompson Station, TN). Nine rabbits were bilaterally inoculated with the HI29 strain of HSV-1 by
placing 2 X 106 PFU of virus in 100 ml onto the corneas, and gently massaging the eyelids. These nine
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
1049
Sulforhodamine B and Lissamine Green B
rabbits were divided into three groups. Beginning at 1
day postinfection (DPI), the right eyes of the rabbits in
each group of three rabbits received 50 ml of 1% lissamine green B, 1% sulforhodamine B, or 1% rose bengal dye, respectively. Treatments were given daily
through 11 DPI, except for the 6th DPI. To check for
obvious toxicity of daily dye treatment, the right eyes
of an additional three rabbits, nonvirus infected, were
treated with the dyes in parallel (one rabbit per dye).
The left eyes of these uninfected rabbits served as untouched controls.
After instillation of the dyes into the rabbits' right
eyes on DPI 1 to 4 and 7 to 11, the left and right eyes
of the virus-infected animals were swabbed with sterile
cotton applicators. Each swab was separately placed
into 1 ml of tissue culture media, which was subsequently titered for viral infectivity on Vero cells by a
tissue culture infectious dose assay. External eye photography was performed on each rabbit in this study at
5, 7, 9, and 11 DPI, immediately after the swabbing
procedure.
RESULTS
Lissamine Green B, but Not Sulforhodamine
B, Stains Membrane-Damaged Cells
Because of evidence that rose bengal stains normal,
healthy cells in vitro, whereas fluorescein staining of
normal cells is minimal and recordable only with a fluorescence microscope, we first sought to characterize
the propensity of sulforhodamine B and lissamine
green B to stain healthy cells in vitro. Serial dilutions
of each dye were added to RCE explant cultures (Fig.
1). Although rose bengal vividly stained the cell monolayers at concentrations greater than or equal to 5 X
10~4 M, sulforhodamine B and lissamine green B, like
fluorescein, did not stain the epithelial cells by gross
inspection or light microscopy at any concentration
tested. Because others have suggested that both sulforhodamine B and lissamine green B are taken up by
damaged cells,56 RCE cell monolayers were artificially
damaged to induce staining (Fig. 2). Monolayers were
treated before dye application with trypsin to break
cell-matrix and cell-to-cell adhesions (Fig. 2B), or with
the detergent, Triton X-100, to damage cell membranes (Fig. 2C). Neither treatment induced cellular
uptake of sulforhodamine B (data not shown). Treatment with trypsin failed to facilitate subsequent uptake of lissamine green B (Fig. 2B), but Triton X-100
pretreatment resulted in visible cellular uptake (Fig.
2C). The dye appeared to bind predominantly to the
epithelial cell nuclei.
Because rose bengal staining can be blocked by
mucin,1 we wished to determine if lissamine green B
CO
8
ffl
<D
C
1
#
E
CO
o
3
"5
FIGURE i. Staining of RCE cell explant cultures with serial
dilutions of rose bengal, lissamine green B, sulforhodamine
B, or fluorescein. Concentrations of dye are as follows: 1 =
10"a M, 2 = 5 X 1(T3 M, 3 = 10"3 M, 4 = 5 X 1(T4 M, 5 =
1(T4 M, 6 = 5 X 1CT5 M, 7 = 10"5, 8 = phosphate buffered
saline control. Note that only rose bengal visibly stains the
cell monolayer.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
1050
Investigative Ophthalmology & Visual Science, March 1994, Vol. 35, No. 3
FIGURE 2. Alteration of RCE cell monolayers to facilitate lissamine green B uptake. (A) Untreated monolayer after exposure to lissamine green B. (B) Monolayer pretreated with 0.25%
trypsin, followed by lissamine green B. (C) Monolayer pretreated with 0.5% Triton X-100,
followed by lissamine green B; or with (D) 0.5% Triton X-100, followed by coating with
porcine stomach mucin, and then lissamine green B. Lissamine green B uptake occurred only
in Triton X-100-pretreated cells, and was not blocked by coating the cells with mucin before
dye application. Note that Triton X-100 pretreatment before lissamine green B results in
nuclear staining of epithelial cells (C,D). The left lower corner of each micrograph is the area
from adjacent to the explant and appears defocused.
uptake by membrane-damaged cells could be blocked
in a similar fashion. Triton X-100-damaged cells were
coated with 1% porcine stomach mucin before application of lissamine green B. As shown in Figure 2D,
mucin failed to block lissamine green B uptake by
membrane-damaged cells. The treatment of cells with
porcine stomach mucin also failed to affect subsequent attempts to stain them with sulforhodamine B
(data not shown). Neither sulforhodamine B nor lissamine green B bound to porcine stomach mucin when
mixed together with the mucin and passed through a
G-75 gel filtration column (data not shown).
Lissamine Green B and Sulforhodamine B
Diffusion in Collagen Gels Is Intermediate
Between Rose Bengal and Fluorescein
Although sulforhodamine B and lissamine green B,
like fluorescein, did not stain healthy RCE cells, all
three dyes stained the corneal explant stroma. In contrast, rose bengal did not appreciably stain the stroma
(Fig. 1). To test the relative diffusion of each dye
through a collagenous stroma, rat tail collagen was
extracted and used to form a collagen gel. Each dye
was exposed to the surface of a gel-filled tube for 30
minutes before removal of the excess dye by washing
with PBS. Photographs of the gel-filled tubes, taken at
several times after dye application, revealed a relative
diffusion rate of fluorescein > lissamine green B >
sulforhodamine B > rose bengal (Fig. 3).
Lissamine Green B Binds to HSV-1-Infected
Vero Cells Late in the Infectious Cycle
Because Norn5 observed that lissamine green B stains
dendritic herpetic epithelial keratitis in a manner similar to rose bengal, and because it has been shown that
lissamine green B binds preferentially to the measlesinfected conjunctival epithelium in measles keratoconjunctivitis,14 we wished to examine if lissamine
green B uptake was enhanced in virus-infected epithelium. Flask cultures of Vero cells were infected with
HSV-1 or mock-infected with PBS (Fig. 4A). By spectrophotometric analysis, HSV-1-infected and uninfected Vero cells bound equivalent amounts of lissamine green B at 6 and 12 hours postinfection (Fig.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
Sulforhodamine B and Lissamine Green B
CQ
(0
O>
Ire
a)
o
CO
•o
CQ
c
Si
5
4)
o)
o
1051
4B). However, because we had already shown that cell
membrane damage was necessary to induce lissamine
green B uptake by RCE cells (Fig. 2C), we postulated
that uptake of lissamine green B by HSV-1-infected
Vero cells might occur maximally at a time late in the
viral replicative cycle, when virus-induced cell membrane alterations are most prominent.15 We tested the
uptake of lissamine green B at 24 hours postinfecdon,
just before cell lysis (Fig. 4A), and found that infected
cells took up more dye than uninfected cells (P <
0.001; Fig. 4B).
Lissamine Green B and Sulforhodamine B
Inhibit HSV-1 Plaque Formation In Vitro
Rose bengal possesses an intrinsic antiviral capacity,
and may reduce the predictive value of a negative ocular surface culture for HSV-1, if the dye is applied just
before culture.34 This fact limits the clinical usefulness
of rose bengal in ambiguous cases of epithelial keratitis where a culture is desirable. Therefore, we tested
lissamine green B, a dye thought to be specific for
epithelial keratitis,5 for its antiviral activity by a direct
neutralization assay.4 We found that 0.06% lissamine
green B, when present at the time of viral adsorption,
reduced HSV-1 plaque formation in Vero cells by
greater than 50% (Fig. 5A). In contrast, at least 0.5%
sulforhodamine B was necessary to show any significant decrease in plaque formation (Fig. 5B).
FIGURE 3. Diffusion of dyes in collagen gel. The times shown
reflect the number of minutes after application of dye to the
gel. Arrowhead indicates the level of {fluorescein) dye within
the gel. The relative diffusion rate isfluorescein> lissamine
green B > sulforhodamine B > rose bengal. PBS = phosphate-buffered saline as a control.
Lissamine Green B Has Intrinsic Cellular
Toxicity
In our plaque reduction experiment, Vero cells were
incubated with a dye—virus mixture for 1 hour. Hence,
it was of interest to determine whether the reduction
in HSV plaque formation was due to cellular toxicity
of the dyes. We therefore explored the effect of each
dye on the incorporation of 14C-amino acids into acidprecipitable protein in Vero cells as a measure of dye
effect on cell metabolism. Cells were treated with
various concentrations of lissamine green B or sulforhodamine B for 1 hour, and then labeled for 23 hours
with 14C-amino acids. By measuring incorporation of
14
C-amino acids into cellular acid-precipitable protein,
we showed that lissamine green B was toxic to Vero
cells in a dose-dependent manner (Fig. 6A), whereas
sulforhodamine B was relatively nontoxic at all concentrations tested (Fig, 6B). These results indicate that
the mild plaque reduction effect of sulforhodamine B
was not due to cellular toxicity by the dye, but does not
exclude this possibility for lissamine green B. We also
performed a qualitative cell viability assay that has
been used previously to demonstrate the toxicity of
rose bengal.2 In this assay, RCE cells in culture were
exposed to sulforhodamine B or lissamine green B for
only 5 minutes before washing the cells with PBS; this
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
Investigative Ophthalmology & Visual Science, March 1994, Vol. 35, No. 3
B
Hours Post Infection
FIGURE 4. (A) Phase contrast appearance of Vero cells at 0, 6, 12, and 24 hours after infection
with HSV-1. (B) Spectrophotometric measurement of binding of lissamine green B to infected versus uninfected Vero cells at 6, 12, and 24 hours after infection with HSV-1. Binding is expressed as the percent of binding in (infected cells/uninfected cells) X 100. By
quantitative spectrophotometry, a significant difference in dye uptake was found between
infected and uninfected cultures only at 24 hours postinfection, when infected cells bound
more dye (P < 0.001). Error bars = standard error of the mean.
more closely correlates with the human patient, in
which tears dilute the dye immediately after dye application. By this assay, we found no effect on cell viability for either dye, even up to the 1% concentration
(data not shown).
Lissamine Green B and Sulforhodamine B Do
Not Inhibit HSV-1 Replication In Vivo
The staining characteristics and antiviral properties of
lissamine green B and sulforhodamine B on HSV-1-infected corneas were studied and compared to rose
bengal in rabbits experimentally infected with the
HI 29 strain of HSV-1. Strain HI 29 typically produces
large epithelial defects as well as dendritic keratitis in
New Zealand White rabbits10 (Stroop, personal observations). For this experiment, both eyes of each rabbit
were infected, but the dyes were applied only to the
right eyes, leaving the left eyes to serve as virus-infected, non-dye-treated controls for the virus titration
studies. Three animals per group were treated daily
from 1 to 5 and 7 to 10 DPI by instillation of 50 ml of
1% solutions of each dye. Swabs from both the dyetreated right and non-dye-treated left eyes were titered separately for viral infectivity. To check for evidence of toxicity, the right eye of one uninfected rabbit per dye was treated in parallel.
Despite its apparent antiviral properties in vitro,
lissamine green B did not appreciably affect the replication of HSV-1 in rabbit corneas in vivo (Fig. 7A). No
statistical difference in the titers recovered from the
three dye-treated right eyes or the three non—dyetreated left eyes were found at any DPI. The dyetreated and non-dye-treated eyes all ceased shedding
measurable amounts of virus by 10 DPI (Fig. 7A).
Virus-induced, large, geographic corneal epithelial
defects were stained easily by lissamine green B (Fig.
8), suggesting that this dye stained the exposed
stroma. Lissamine green B also stained punctate and
dendritic lesions that often were seen at the periphery
of the large geographic lesions (Fig. 8). The geographic lesions became larger from 5 to 9 DPI, but had
shrunk considerably by 11 DPI. Corneal healing appeared to correlate with the rapid decrease in viral
titers observed in these animals from 9 to 11 DPI (Fig.
7A). The uninfected control animal that was treated
with lissamine green B showed no obvious signs of ocular surface toxicity through 11 DPI.
Like lissamine green B, sulforhodamine B exhibited no antiviral effect in vivo (Fig. 7B). Interestingly,
between 7 and 8 DPI, there appeared to be a resurgence in viral replication in the corneas of the dyetreated eyes. This increase in titer, however, was not
statistically significant, and was due to the reappear-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
Sulforhodamine B and Lissamine Green B
1053
bengal was profoundly antiviral, it is not surprising
that the corneal lesions stained by this dye were less
conspicuous than those seen in lissamine green Btreated eyes. No toxicity was observed in the eye of the
uninfected control rabbit treated with rose bengal.
120-
0.000
0.01S
0.030
0.060
0.125
0.250
Lissamine Green B (%)
O
80
o
O
60
0.060
B
0
0.125
0.0039
Sulforhodamine B (%)
0 . 0 1 5 0.062
0.25
Lissamine Green B (%)
FIGURE 5. HSV-1 plaque reduction caused by (A) lissamine
green B and (B) sulforhodamine B. Note that 0.06% lissamine green B, or 0.5% sulforhodamine B, when present at
the time of viral adsorption, reduced HSV-1 plaque formation in Vero cells by greater than 50%. Error bars indicate
the standard errors of the means of numbers of plaques
counted in triplicate cultures at each concentration of dye.
ance of virus in the right eye secretions of one of the
three rabbits studied. By analogy to other studies, the
reappearance of virus in this single eye may have been
due to the length of time required for HSV-1 to spread
to the ganglion (2 to 3 days), replicate within it (1 to 2
days), and return to the eye (2 to 3 days).9"11 Sulforhodamine B did not stain the eyes of infected rabbits as
well as lissamine green B (data not shown). No signs of
toxicity were observed in the eye of the uninfected
control animal treated with sulforhodamine B.
As expected from our previous studies with rose
bengal,3 this dye exhibited significant antiviral effect
in vivo (Fig. 7C). Virus was recovered from the three
rose bengal-treated eyes only on DPI 2 and 3; at all
other DPI, virus could not be recovered from the
swabbing medium. At 5 DPI, rose bengal stained the
edges, but not the centers, of the large virus-induced
corneal epithelial defects. This suggests that rose bengal stained the epithelium at the edge of the lesion, but
did not stain the exposed corneal stroma. At 7 and 9
DPI, rose bengal staining revealed only very small,
punctate and dendritic epithelial defects. Because rose
o
o
B
0.0039
0.015
0.062
0.25
Sulforhodamine B (%)
6. Cellular toxicity of (A) lissamine green B and (B)
sulforhodamine B as measured by l4C-amino acid incorporation into cellular acid-precipitable protein in Vero cells.
Whereas lissamine green B was toxic to cells in a dose-dependent fashion, sulforhodamine B toxicity was relatively
minor at clinically relevant concentrations. Error bars indicate standard errors of the means of acid-precipitable
counts from triplicate cultures for each concentration of
each dye.
FIGURE
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
Investigative Ophthalmology 8c Visual Science, March 1994, Vol. 35, No. 3
1054
DISCUSSION
10*
10*
10'
10'
10
10
:
1
I
s
r
0/\r
5
r/ -
4
r1
3
1
1
1
1
1
1
1
1 :
Ai
r
6
1
'
^ ^ ^ * < L
T
I
I
I
1
r
2
r
1
r
0
T ? ]
I
0
Q
LJ
1
1 1 |
10'
7
HIM
10
I
I
1 2 3 4 5
I
I
I
I
I
6 7 8 9
I
I
I
I
I
I
10 11
I
\
V
10'
O
Z
10'
CD
GO
10'
10
10'
10'
a:
LJ
a:
LJ
10
101
I
I
I
I
I
I
I
I
1 2 3 4 5 6 7 8 9
I
I
I
10 11
10
1 2 3 4 5 6 7 8 9
10 11
DAYS POST INFECTION
'
In the evaluation of ocular surface disease, fluorescein
is applied to improve visualization of corneal and conjunctival epithelial defects, to stain preocular tear film
in evaluation of tear breakup time,16 and to evaluate
tear clearance.17 Sulforhodamine B has been reported
to be superior to fluorescein for the visualization of
both preocular tear film and conjunctival epithelial
lesions.6 In contrast, rose bengal is applied to visualize
better the lesions of corneal epithelial keratitis, to assess the interpalpebral ocular surface in keratoconjunctivitis sicca, and to delineate the extent of intraepithelial dysplasia or squamous cell carcinoma at the
corneal limbus.18 Lissamine green B has been touted
as more sensitive than rose bengal for the screening of
xerophthalmia,19"22 but this was limited by a lack of
specificity for early xerosis.23'24 Lissamine green B also
has been used to highlight the lesions of herpetic epithelial keratitis,5 to quantify the ocular surface
changes in keratoconjunctivitis sicca,25"28 to assess the
function of meibomian gland orifices,29 to test the effects of air currents on corneal epithelium,30 and to
assess injury to corneal epithelium during cataract
surgery.31
We have summarized the staining characteristics,
antiviral activity, and cellular toxicity of these four
dyes in Table 1, by combining our present studies with
those published previously.1"3 Our recent work1 is
consistent with the long-held hypothesis that staining
of the ocular surface by fluorescein occurs when surface epithelial cell-to-cell junctions are disrupted. Fluorescein passage between epithelial cells or into corneal stroma occurs when the surface epithelium is absent, damaged, or not fully differentiated, and hence
without tight junctions. Our in vitro data suggest that
sulforhodamine B may act in much the same way as
fluorescein. We could not induce visible staining of
cells by sulforhodamine B in vitro, but did observe
staining of corneal stroma. In vivo, sulforhodamine B
stained HSV-1 -induced corneal lesions poorly. Our in
FIGURE 7. Recovery of HSV-1, strain HI29 from the ocular
secretions of New Zealand White rabbits treated with and
without (A) lissamine green B, (B) sulforhodamine B, and (C)
rose bengal. Animals were bilaterally infected at 0 DPI, and
50 ml of 1% solutions of the dyes were added to the right
eyes of three rabbits per dye on DPI 1 to 5 and 7 to 10. The
left eyes were not treated with dye. The figures show the titer
of virus recovered from swabs of the dye-treated right (•)
and non-dye-treated left eyes (O) (n = 3 per time point).
Although lissamine green B and sulforhodamine B had no
effect on viral recovery, daily application of rose bengal significantly reduced viral titers. The error bars indicate the
standard error of the mean. Downward arrows indicate that
the titer was below the lower limit of the assay (< 10 tissue
culture infectious dose/ml).
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
Sulforhodamine B and Lissamine Green B
1055
FIGURE 8. An HSV-1 strain H129-infected rabbit eye stained
with lissamine green B at 7 DPI. Note the large geographic
epithelial lesions in the upper left quadrant of the cornea
and the dendritic lesions in the lower right quadrant. Lissamine green B stained the epithelial edges of each ulcer, and
also stained the deepithelialized cornea stroma.
TABLE l.
vitro results are in accordance with the work of Araie
and Maurice,32 who showed that epithelial permeability by sulforhodamine B is similar to that of fluorescein, and that both dyes enter subepithelial stroma
primarily by passing between epithelial cells rather
than directly through them. In the same study, the
authors also showed that fluorescein diffused through
corneal stroma more quickly than did sulforhodamine
B. This is also in agreement with our results, which
showed that sulforhodamine B diffused into collagen
gel at an intermediate rate between that of fluorescein
(maximal) and rose bengal (minimal), and stained the
corneal stroma of HSV-1-induced corneal epithelial
defects more distinctly than rose bengal, but not
nearly as well as lissamine green B. Lissamine gTeen B
diffused into collagen gel at a rate faster than sulforhodamine B, but less than fluorescein, and stained corneal stroma well both in vitro and in vivo.
In contrast to the concept put forth by Norn,5 that
rose bengal stains only dead and devitalized ocular
surface epithelium, our previous data1"3 suggest that
rose bengal stains both healthy and "devitalized" cells
in vitro. Therefore, if a vital dye is defined as any dye
that can distinguish between normal or healthy and
abnormal or unhealthy cells, then fluorescein may be a
vital dye, but rose bengal is not. Rose bengal does not
Staining Characteristics and Antiviral Activity of Ophthalmic Dyes
Experimental data
Stains healthy cells*
Stains dead or
degenerated cells
Staining blocked by
mucin
Intrinsic toxicity
Photo toxic ity
Relative speed of
diffusion through
collagenous stroma
Clinical extrapolation
Staining promoted by
Antiviral activity
Inhibits HSV-1
plaque formation
in vitro
Inhibits HSV-1
replication in vivo
Rose Bengal
Lissamine Green
B
Sulforhodamine
Fluorescein
No
No
Yes
Yes
No
Yes
No
No
Yes
No
NAJ
No
Yes
Yes
Yes§
No
ND
No
ND
Fastest
Slowest
Fast
Flow
Disruption of
cell-cell
junctions
Insufficient
protection by
preocular tear film
Cell death or
degeneration;
and disruption
of cell-cell
junctions
Disruption of
cell-cell
junctions
No
Yes
Yes
Weakly
No
Yes
No
ND = noi done.
* Staining is defined by clinical means of detection, i.e., by the naked eye or cobalt blue filtered microscopy,
f Possibly "Yes" if appropriate excitation/barrier filter is used for observations/'
% Not applicable since no staining was observed without mucin precoating.
§ Metabolic suppression as shown by 14C-amino acid uptake, but no effect on cell viability.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
B
Nof
1056
Investigative Ophthalmology 8c Visual Science, March 1994, Vol. 35, No. 3
discriminate between normal and abnormal cells in vitro, yet often acts like a vital dye in vivo, because of a
complex and dynamic interaction between the surface
epithelial cells and the preocular tear film. For instance, we have postulated previously that rose bengal
highlights the edge of a herpetic dendritic or geographic ulceration, because the swollen, infected epithelial cells at the ulcer's edge are unable to bind or
produce a component of the preocular tear film that
normally would block penetration by rose bengal.3 We
noted that the critical tear film component that blocks
rose bengal staining of the normal ocular surface may
be a corneal epithelial-cell derived mucin.33 Based on
these findings, one might speculate that rose bengal
staining in keratoconjunctivitis sicca could result from
squamous metaplasia,34 which would produce a deficiency of this protective mucin layer. In the setting of
limbal dysplasia or frank carcinoma,18 rose bengal
staining could be explained by the same mechanism.
This explanation, however, cannot apply to staining of
the ocular surface by lissamine green B. As shown in
this report, the dye is not blocked by mucin, but does
stain those cells with detergent-induced membrane
damage. Therefore, lissamine green B is a vital dye.
Our report is not the first to demonstrate that
lissamine green is a vital dye. Goldacre and Sylven35
reported the selective capacity of lissamine green to
stain necrotic tumor cells. Holmberg36 showed that
healthy cells were impermeable to lissamine green B,
but cells damaged by freezing took up the dye in a
dose—response fashion where repeated freeze—thaw
cycles resulted in a greater percentage of dye-positive
cells. Jans and Hassard37 demonstrated that lissamine
green B was an effective "supravital" stain for determination of corneal endothelial cell viability, and suggested it might be useful in the pretransplantation assessment of donor corneal tissue. More recently, the
dye has been used as a marker for in vitro cell-killing
effects of antitumor drugs.38
We noticed that lissamine green B uptake in membrane-damaged cells appeared maximal in the nucleus. Rose bengal also binds to the cell nucleus,2'3 but
we found that although rose bengal binding to cells
was not reversible by repeated washes of the cells, the
binding of lissamine green B to membrane-damaged
cells was at least partially reversible by repeated washings (data not shown). Lissamine green has been documented previously to bind to serum proteins, in vitro,39 and we postulate that, like rose bengal,40 lissamine green B might bind to nuclear histones, although
more reversibly than rose bengal.
We also found that lissamine green B stains cells in
the late stages of cytopathic viral infection in vitro, and
highlights the epithelial edge of HSV dendritic and
geographic corneal ulceration in vivo. Our data are
consistent with previous reports that lissamine green B
stains viral-infected epithelium of HSV epithelial keratitis,5 and that the dye binds preferentially to measlesinfected conjunctival epithelium in measles keratoconjunctivitis.14 In the latter study, however, measlesinfected Vero cells did not stain with lissamine green.
Possibly, the dye was applied to the measles-infected
Vero cells too early in the infectious cycle to show
staining. We found, in HSV-1-infected Vero cells, that
significant lissamine green B uptake occurred only late
in infection, when virus-induced cell membrane alterations are prominent.15
Prior studies have demonstrated that rose bengal
possesses antiviral activity, and that rose bengal application before viral culture could reduce the predictive
value of a negative culture for HSV.3'4 The current
study, in which the daily application of rose bengal to
the ocular surface of HSV-1-infected rabbits reduced
the intensity and duration of infection, confirms the
antiviral capacity of rose bengal (Fig. 7C). Because the
antiviral activity of sulforhodamine B and lissamine
green B was unknown, we tested each dye for antiviral
effect both in vitro and in vivo. Using a direct neutralization assay,4 lissamine green B, but not sulforhodamine B, caused a significant reduction in HSV plaque
formation at very low doses in vitro. In vivo, however,
neither lissamine green B nor sulforhodamine B significantly reduced viral replication. Possible explanations
for the apparent antiviral effect of lissamine green B in
vitro include direct inactivation of the virus via chemical or photoinactivation, prevention of viral adsorption and penetration into the cell by dye bound to the
virion membrane or virus receptor site on the surface
of the cell, or direct dye-induced cellular toxicity
whereby the cells are unable to support viral replication. Our in vivo data suggest that direct inactivation
of the virus or prevention of viral adsorption are unlikely mechanisms for the plaque reduction found in
vitro. By a quantitative assay of cellular metabolism,13
in which cells were exposed to dye for 1 hour so as to
parallel the plaque reduction assay, we found that cellular toxicity of lissamine green B and sulforhodamine
B roughly paralleled HSV-1 plaque reduction. In contrast, by a cell viability assay,2 in which cells were exposed to lissamine green B or sulforhodamine B for
only 5 minutes before washing the cells with PBS, we
found no effect on viability. The 5-minute exposure to
dye more closely mimics the situation in vivo, in which
tears begin to dilute and clear the dye immediately
after instillation, and any effect of dye on cellular metabolism is minimized. The mechanism of HSV-1
plaque reduction by lissamine green B in vitro may be
a result of a prolonged exposure of the cells to the dye.
Yet, we hesitate to interpret our data as proof of a
cause-and-effect relationship between cell toxicity and
plaque reduction. Feenstra and Tseng2 have shown
previously that rose bengal is toxic to cells by the viabil-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
Sulforhodamine B and Lissamine Green B
ity assay. Chodosh and Stroop have shown that rose
bengal-pretreated human corneal epithelial cells, although unable to support HSV-1 replication, did support the replication of adenovirus types 5 and 8 in
vitro (unpublished observations). Together, these data
suggest that rose bengal, although toxic to cells, mediates the suppression of subsequent herpes viral replication, not by cellular toxicity, but rather by a direct
inactivation of the virus.
In summary, we have demonstrated that, in contrast to rose bengal, neither sulforhodamine B nor lissamine green B stain healthy, normal cells in vitro.
Lissamine green B, but not sulforhodamine B, selectively stains damaged epithelial cells, and is therefore a
vital dye. Like fluorescein, both sulforhodamine B and
lissamine green B stain corneal and collagenous
stroma, and could be used to demonstrate corneal and
conjunctival epithelial defects. The capacity of lissamine green B to stain membrane-damaged epithelial
cells, combined with its ability to stain denuded corneal stroma, could make the dye a useful alternative to
rose bengal and fluorescein in the diagnosis of ocular
surface diseases. Although rose bengal remains a
unique dye to detect the protective status of the preocular tear film,12 it should be reemphasized 3 that the
predictive value of a negative ocular surface culture
for herpes simplex may be significantly lowered by the
application of rose bengal before culture. For demonstration of the morphology of corneal epithelial keratitis in the human or animal research subject with HSV
epithelial keratitis, lissamine green B would be preferable to rose bengal, except at the end point of the study.
As we continue to search for the ideal vital dye, other
properties unique to sulforhodamine B and lissamine
green B are still under investigation, and may help us
in the future to understand better the pathophysiology of ocular surface diseases.
1057
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Key Words
herpes simplex virus type 1, lissamine green dye, sulforhodamine B, vital dye
16.
Acknowledgments
17.
The authors thank William Feuer for his help with statistical
analyses, and Careene Banks for technical assistance.
18.
References
1. Feenstra RPG, Tseng SCG. Comparison of fluorescein and rose bengal staining. Ophthalmology.
1992;99:6O5-617.
2. Feenstra RPG, Tseng SCG. What is actually stained by
rose bengal? Arch Ophthalmol. 1992; 110:984-993.
3. Chodosh J, Banks MC, Stroop WG. Rose bengal inhibits herpes simplex virus replication in Vero and human corneal epithelial cells in vitro. Invest Ophthalmol
VisSci. 1992; 33:2520-2527.
4. Roat MI, Romanowski E, Araulio-Cruz T, Gordon J.
19.
20.
21.
22.
23.
The antiviral effects of rose bengal and fluorescein.
Arch Ophthalmol. 1987; 105:1415- 1417.
Norn MS. Lissamine green: vital staining of cornea
and conjunctiva. Acta Ophthalmol. 1973; 51:483-491.
Eliason JA, Maurice DM. Staining of the conjunctiva
and conjunctival tear film. Br J Ophthalmol.
1990;74:519-522.
Dix RD, McKendall RR, Baringer JR. Comparative
neurovirulence of herpes simplex virus type 1 strains
after peripheral or intracerebral inoculation of
BALB/c mice. Infect Immun. 1983;40:103-112.
Hudson SJ, Dix RD, Streilein JW. Induction of encephalitis in SJL mice by intranasal infection with
herpes simplex virus type 1: a possible model of
herpes simplex encephalitis in humans. / Infect Dis.
1991; 163:720-727.
Stroop WG, Schaefer DC. Production of encephalitis
restricted to the temporal lobes by experimental reactivation of herpes simplex virus, f Infect Dis.
1986;153:721-731.
Stroop WG, Schaefer DC. Severity of experimentally
reactivated herpetic eye disease is related to the neurovirulence of the latent virus. Invest Ophthalmol Vis
5d.l987;28:229-237.
Stroop WG, McKendall RR, Battles EJ, Schaefer DC,
Jones B. Spread of herpes simplex virus type 1 in the
central nervous system during experimentally reactivated encephalitis. Microb Pathog. 1990; 8:119-134.
Aeschbacher M, Reinhardt CA, Zbinden C. A rapid
cell membrane permeability test using fluorescent
dyes and flow cytometry. Cell Biol Toxicol. 1986; 2:
247-255.
Dix RD, Courtney RJ. Effects of cytochalasin B on
herpes simplex virus type 1 replication. Virology.
1976;70:127-135.
Nommensen FE, Dekkers NWHM. Detection of
measles antigen in conjunctival epithelial lesions staining by lissamine green during measles virus infection.
JMed Virol. 1981; 7:157-162.
Spear PG. Glycoproteins specified by herpes simplex
virus. In: Roizman, ed. The Herpes Viruses. New York:
Plenum Press; 1985:313-356.
Norn MS. Dessication of the precorneal tear film: 1.
corneal wetting-time. Acta Ophthalmol. 1969;47:865880.
Mishima S, Gasset A, Klyce SD, Baum JL. Determination of tear volume and tear flow. Invest Ophthalmol.
1966;5:264-276.
Wilson FM. Rose bengal staining of epibulbar squamous neoplasms. Ophthalmic Surg. 1976;7:21-23.
Sauter JJM. Diagnosis of xerophthalmia by vital staining. Tropical Doctor. 1976;6:91-93.
Sauter JJM. Xerophthalmia and measles in Kenya. Doc
Ophthalmol. 1976; 42:1-235.
Kusin JA, Soewondo W, Sinaga HSRP. Rose bengal
and lissamine green vital stains: useful diagnostic aids
for early stages of xerophthalmia? Am f Clin Nutr.
1979;32:1559-1569.
Norn MS. Vitamin A-responsive punctate keratopathy
in xerophthalmia. AmJ Ophthalmol. 1979;88:955-6.
Emran N, Sommer A. Lissamine green staining in the
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017
1058
Investigative Ophthalmology 8c Visual Science, March 1994, Vol. 35, No. 3
clinical diagnosis of xerophthalmia. Arch Ophthalmol.
1979;97:2333-2335.
24. Sommer A. Conjunctival xerosis. Am J Clin Nutr.
1980;33:1313.
25. Franck C, Skov P. Foam at inner canthus in office
workers, compared with an average Danish population as control group. Ada Ophthalmol. 1989; 67:6168.
26. Franck C. Fatty layer of the precorneal film in the
"office
eye
syndrome."
Ada Ophthalmol.
1991;69:737-743.
27. Khurana AK, Chaudhary R, Ahluwalia BK, Gupta S.
Tear film profile in dry eye. Ada Ophthalmol.
1991;69:79-86.
28. Khurana AK, Sunder S, Ahluwalia BK, Malhotra KC.
Tearfilmprofile in Graves' ophthalmopathy. Ada Ophthalmol. 1992:70:346-349.
29. Norn M. Meibomian orifices and Marx's line studied
by triple vital staining. Ada Ophthalmol. 1985; 63:698700.
30. Wyon NM, Wyon DP. Measurement of acute response
to draught in the eye. Ada Ophthalmol. 1987; 65:385392.
31. Norn MS. Preoperative protection of cornea and conjunctiva. Ada Ophthalmol. 1981; 59:587-594.
32. Araie M, Maurice D. The rate of diffusion of fluoro-
33.
34.
35.
36.
37.
38.
39.
40.
phores through the corneal epithelium and stroma.
ExpEyeRes. 1987;44:73-87.
Mui MM, Tseng SCG. Characterization of monoclonal
antibodies against mucosal epithelium membrane-associated mucin-like protein (MEM). Invest Ophthalmol
VisSci. 1992; 33(suppl): 1176.
Pflugfelder SC, Huang AJW, Feuer W, et al. Conjunctival cytologic features of primary Sjogren's syndrome. Ophthalmology. 1990; 97:985-991.
Goldacre RJ, Sylven B. A rapid method for studying
tumour blood supply using systemic dyes. Nature.
1959;184:63-64.
Holmberg B. On the permeability to lissamine green
and other dyes in the course of cell injury and cell
death. Exp Cell Res. 1961; 22:406- 414.
Jans RG, Hassard DTR. Lissamine green: a supravital
stain for determination of corneal endothelial viability. Can J Ophthalmol. 1967;2:297-302.
Kopf-Maier P, Wagner W, Kopf H. In vitro cell
growth inhibition by metallocene dichlorides. Cancer
Chemother Pharmacol. 1981;5:237-241.
Brackenridge CJ. Factors affecting the uptake of lissamine green by serum proteins. / Clin Pathol.
1960;13:149-155.
Zhang SH, Tseng SCG. Interactions between rose
bengal (RB) and tear components. Invest Ophthalmol
VisSci. 1992;33(suppl):1286.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933180/ on 05/14/2017