Nitric oxide synthase induction and cytotoxic nitrogen-related oxidant formation in conjunctival epithelium of dry eye (Sjögren's syndrome).

Abstract

Until now, the expression and possible role of nitric oxide and nitrogen related oxidants in the human dry eye have not been investigated. Therefore, we examined immunohistochemically nitric oxide synthase isomers (NOS), enzymes generated nitric oxide, nitrotyrosine, a cytotoxic byproduct of nitric oxide and malondialdehyde, a byproduct of lipid peroxidation, in conjunctival epithelium of patients with dry eye, Sjögren's syndrome (SS). Moreover, in conjunctival epithelium of patients with dry eye (SS) the immunohistochemical staining of some pro-inflammatory cytokines was demonstrated: mature interleukin-1 beta (IL-1beta), interleukin 6 (IL-6), interleukin 8 (IL-8) and tumor necrosis factor alpha (TNF-alpha). Conjunctival epithelial cells were obtained by the method of impression cytology. Normal eyes served as controls. In contrast to the normal eyes where endothelial nitric oxide synthase (NOS3) as well as inducible nitric oxide synthase (NOS2) were only slightly expressed in conjunctival epithelium, in dry eye both NOS (mainly NOS2) were gradually expressed along the severity of dry eye symptoms which was in accord with pro-inflammatory cytokine immunodetection (IL-1beta, IL-6, IL-8, TNF-alpha) in dry eye conjunctival cytology samples. This was in contrast to normal eyes where the staining of pro-inflammatory cytokines was weak or completely absent. Peroxynitrite formation (demonstrated by nitrotyrosine residues) and lipid peroxidation (evaluated by increased malondialdehyde staining) were also found in conjunctival epithelium of dry eye with highly pronounced symptoms of dryness. In conclusion, results point to the suggestion that reactive nitrogen species are involved in the pathogenesis or self-propagation of autoimmune dry eye (SS).

Full text

Page 1

Nitric oxide synthase induction and cytotoxic nitrogen-related
oxidant formation in conjunctival epithelium
of dry eye (Sjo ¨gren‘s syndrome)
J. Cˇejkova ´a,*, T. Ardana, Z. Sˇimonova ´a, Cˇ. Cˇejkaa,c, J. Malecc, K. Jirsova ´b,
M. Filipecb, D. Dotr ˇelova ´c, B. Bru ˚nova ´c
aLaboratory of Eye Histochemistry and Pharmacology, Institute of Experimental Medicine, Academy of Sciences of the Czech Republic, Vı ´den ˇska ´ 1083,
CR-14220 Prague 4, Czech Republic
bLaboratory of the Biology and Pathology of the Eye, Institute of Inherited Metabolic Disorders, General Teaching Hospital and Charles University,
CR-12800 Prague 2, Czech Republic
cDepartment of Ophthalmology for Children and Adults, Motol Hospital, Charles University, 2nd Medical School, CR-15000 Prague 5, Czech Republic
Received 9 February 2007; revised 24 April 2007
Available online 22 May 2007
Abstract
Until now, the expression and possible role of nitric oxide and nitrogen related oxidants in the human dry eye have not been inves-
tigated. Therefore, we examined immunohistochemically nitric oxide synthase isomers (NOS), enzymes generated nitric oxide, nitroty-
rosine, a cytotoxic byproduct of nitric oxide and malondialdehyde, a byproduct of lipid peroxidation, in conjunctival epithelium of
patients with dry eye, Sjo ¨gren‘s syndrome (SS). Moreover, in conjunctival epithelium of patients with dry eye (SS) the immunohistochem-
ical staining of some pro-inflammatory cytokines was demonstrated: mature interleukin-1 beta (IL-1beta), interleukin 6 (IL-6), interleu-
kin 8 (IL-8) and tumor necrosis factor alpha (TNF-alpha). Conjunctival epithelial cells were obtained by the method of impression
cytology. Normal eyes served as controls. In contrast to the normal eyes where endothelial nitric oxide synthase (NOS3) as well as induc-
ible nitric oxide synthase (NOS2) were only slightly expressed in conjunctival epithelium, in dry eye both NOS (mainly NOS2) were grad-
ually expressed along the severity of dry eye symptoms which was in accord with pro-inflammatory cytokine immunodetection (IL-1beta,
IL-6, IL-8, TNF-alpha) in dry eye conjunctival cytology samples. This was in contrast to normal eyes where the staining of pro-inflam-
matory cytokines was weak or completely absent. Peroxynitrite formation (demonstrated by nitrotyrosine residues) and lipid peroxida-
tion (evaluated by increased malondialdehyde staining) were also found in conjunctival epithelium of dry eye with highly pronounced
symptoms of dryness. In conclusion, results point to the suggestion that reactive nitrogen species are involved in the pathogenesis or
self-propagation of autoimmune dry eye (SS).
? 2007 Elsevier Inc. All rights reserved.
Keywords: Nitric oxide; Cytotoxic nitrogen-related oxidants; Dry eye; Ocular surface
Dry eye is a chronic disease with a deficiency of aqueous
tears, either in quantity or quality which may lead to ocular
surface dysfunction. The factors leading to the aqueous
tear deficiency are complex involving autoimmune disease
(SS), loss of hormonal support, and glandular inflamma-
tion [1]. Primary SS is characterized by diffuse lymphoid
cell infiltration in the salivary and lacrimal glands, resulting
in symptoms of dry eye and mouth due to insufficient secre-
tion [2]. Lacrimal glands from patients with SS produce
greatly increased levels of pro-inflammatory cytokines,
such as IL-1, interferon c and TNF alpha [3,4], which are
secreted into the tear fluid. In the tear fluid of patients with
SS increased pro-inflammatory cytokine levels were found
by Pflugfelder et al. [5] and Solomon et al. [6]. Solomon
1089-8603/$ - see front matter ? 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.niox.2007.04.006
*Corresponding author. Fax: +420 2 241062692.
E-mail address: cejkova@biomed.cas.cz (J. Cˇejkova ´).
www.elsevier.com/locate/yniox
Nitric Oxide 17 (2007) 10–17

Page 2

et al. [6] found that dry eye disease is accompanied by an
increase in the pro-inflammatory forms of IL-1 (IL-1 alpha
and mature IL-1 beta) and a decrease in the biologically
inactive precursor IL-1 beta in the tear fluid. Based on
the cytokine staining (expression of IL-1 family) in con-
junctival epithelial impression cytology specimens, these
authors suggested that the conjunctival epithelium is one
source of the increased concentration of IL-1 in the tear
fluid of patients with dry eye disease (SS). Beauregard
et al. [7] demonstrated in in vitro experiments that NOS2
expression can be induced in cultured rabbit gland acinar
cells by addition of the pro-inflammatory cytokine, IL-1
beta. The induction of NOS2 in these cells resulted in a sig-
nificant increase in nitric oxide production.
Nitric oxide (NO) is widely recognized to be quite an
important intercellular messenger in the cardiovascular
and nervous systems or immunological reactions, including
thatintheeye.ThismoleculeformedbyconstitutiveNOsyn-
thase (NOS), endothelial (eNOS, NOS3) and neuronal
(nNOS, NOS1), contributes to physiologically regulate ocu-
lar hemodynamics and cell viability and protects vascular
endothelial cells and nerve cells or fibers against pathogenic
factors associated with glaucoma, ischemia, and diabetes
mellitus. Ocular blood flow is regulated by NO derived from
the endothelium and efferent nitrergic neurons. Endothelial
dysfunction impairs ocular hemodynamics by reducing the
bioavailability of NO and increasing the production of
reactive oxygen species. On the other hand, NO formed by
inducible NOS (NOS2) expressed under influences of
inflammatory mediators, leading to serious ocular inflam-
matory reactions or diseases [8]. Ko et al. [9] described the
damaging role of nitric oxide in experimental allergic con-
junctivitis, Jeng et al. [10] and Meisler et al. [11] the effect
of increased NO production in corneas waiting for trans-
plantation. Cˇejkova ´ et al. [12] pointed out the danger from
nitric oxide overproduction in corneas irradiated with
UVB rays. Kao et al. [13] described that increased levels of
NO might be a risk factor in cataract formation. Evidence
of oxidative ocular damage from the cytotoxic byproduct
peroxynitrite, generated by the rapid reaction of NO with
the superoxide, has been demonstrated in endotoxin-
induced uveitis [14,15] and in the cornea irradiated with
UVBrays[12].Wuetal.[16]foundthatperoxynitritecaused
oxidativedamage oftheretinathroughlipidperoxidation of
photoreceptors in experimental autoimmune uveitis. Lipid
peroxidation is an important biological consequence of oxi-
dativedamageof cell membranes andthe formation of cyto-
toxic aldehydes. Increased levels of malondialdehyde, the
toxic aldehyde byproduct of lipid peroxidation, were
described in the cornea after the intrastromal injection of
liposaccharide[17].Buddietal.[18]investigatedthepresence
of malondialdehyde in human diseased corneas, and Park
and Tseng [19] examined distinct immunohistochemical
staining for malondiladehyde in the corneal stroma after
photorefractive keratectomy. The increased malondialde-
hyde stainingwas alsofound in the cornea repeatedlyirradi-
ated with UVB rays [12].
Although in vitro experiments Beauregard et al. [7]
showed that NOS2 expression can be induced in cultured
rabbitglandacinarcellsbyadditionofthepro-inflammatory
cytokine,thepossibleinductionofnitricoxideproductionin
human dry eye has not been examined in literature to date.
Therefore, we decided to study this problem to fill this gap.
Conjunctival epithelial cells (obtained by the method of
impressioncytology)of patientssufferingfrom autoimmune
dryeye (SS)were employed.Normaleyesserved as controls.
Using immunohistochemical methods, the expression of
NOSisomersandsomepro-inflammatorycytokines(matute
IL-1beta, IL-6, IL-8, TNF-alpha) were detected. Moreover,
the possible lipid peroxidation (evaluated by malondialde-
hyde staining) as well as peroxynitrite formation (demon-
strated by nitrotyrosine residues), the reaction product
between NO and superoxide anion were investigated in the
conjunctival epithelium. Superoxide anion can occur as a
consequence of the activities of several enzymes, including
mitochondrial oxidases, xanthine oxidoreductase, and vari-
ous NADPH oxidases. We found previously that xanthine
oxidoreductase/xanthine oxidase, enzymes that generate
reactive oxygen species, were much more expressed in the
conjunctival epithelium of dry eye patients with SS than in
theconjunctivalepitheliumofnormaleyes[36].Thepossible
increase in nitric oxide generation on the ocular surface
together with elevated levels of reactive oxygen species may
be dangerous to the eye from the increased toxic peroxyni-
trite formation.
Experimental procedures
Human subjects
Both eyes of eleven healthy controls (mean age 48.5 ± 5.2 years; two
men, nine women) and both eyes of eleven subjects with dry eye and ver-
ified SS (mean age 49.2 ± 5.4 years; two men, nine women) were examined
in this study. The study followed the tenets of the Declaration of Helsinki.
Informed consent was obtained from all subjects.
The clinical state of the ocular surface was evaluated, using the fluores-
cein tear break-up time, the degree of corneal staining with fluorescein, the
the vital dye staining, and tear production with the Schirmer test. Control
healthy volunteers did not wear contact lenses, had no symptoms of ocular
irritation, had normal tear production as confirmed by a Schirmer test,
and had no signs of ocular surface disease. Dry eye (in the group of
patients with Sjo ¨gren‘s syndrome examined in this paper) was diagnosed
according to Copenhagen criteria, i.e. at least two of the following tests
were abnormal: Schirmer test <10 mm/5 min, tear film break-up time
(BUT) <10 s, rose bengal staining >4 [20,21]. In our controls (normal sub-
jects) and patients with dry eye (SS) following mean values were found: In
healthy subjects the mean Schirmer-1 value was 13.20 ± 1.80 mm, the
mean value of BUT was 14.30 ± 2.80 s, and the mean staining value of
the rose bengal was 1.97 ± 0.70, whereas in dry eye patients these values
were 1. 36 ± 0.40 mm (the mean Schirmer-1 value), 5.40 ± 1.20 s (the
mean value of BUT) and 6.61 ± 0.70 (the mean staining value of the rose
bengal).
The dry eyes of patients examined in our study were classified as fol-
lows: Two men and six women, moderate dry eyes and three women mod-
erate to severe dry eyes, see classification of Murube et al. [22], for details).
The group of patients with moderate dry eye had less evident symptoms of
dryness and three women with moderate to severe dry eye had very much
pronounced symptoms of dryness. The first group of patients was classi-
fied as grade 2 and the second group as grade 2 to grade 3.
J. Cˇejkova ´ et al. / Nitric Oxide 17 (2007) 10–17
11

Page 3

Conjunctival impression cytology, sample collection
Conjunctival epithelial cells were obtained using the method of impres-
sion cytology with Millicell membranes (Millicell-CM, hydrophylic PTFE,
Millipore Corporation, Billerica, MA 01821, USA). A single drop of 0.4%
oxybuprocaine hydrochloride was first instilled to the eye. Impression
cytology was carried out bilaterally on the upper (at the 11–12 and 12–13
o ´clock positions) conjunctiva, 2 mm posterior from the limbus. Strips of
Millicell membrane were gently pressed onto the conjunctival surface to
remove superficial epithelial cells. The membrane was gently removed
and the specimens (conjunctival cells on the Millicells) were stored at
?80? C until they were processed (immunohistochemical examination).
Immunohistochemical examination
The Millicell membranes with conjunctival epithelial cells were fixed in
acetone for 5 min, 4? C, released from plastic holder, rinsed with PBS
(Phosphate buffered saline tablets, Sigma), placed cell side up on round
12 mm coverslips and then (after rinsing with PBS) permeabilised with
0.2% triton (Triton X 100, Sigma) in PBS.
For the immunohistochemical localization of NOS and nitrotyrosine,
the following primary antibodies were used: monoclonal mouse anti-
human NOS3 antibody, monoclonal mouse anti-human NOS1 antibody,
monoclonal mouse anti-human NOS2 antibody (BD Biosciences, San
Jose, CA, USA) and monoclonal mouse anti-nitrotyrosine (Abcam,
Cambridge, UK). Subsequently, an anti-mouse HRP/DAB Ultravision
Detection System (Lab Vision, Fremont, CA, USA) was employed as rec-
ommended by the manufacturer: hydrogen peroxide block (20 min), ultra
V block (5 min), primary antibody incubation (60 min), biotinylated goat
anti-mouse antibody incubation (10 min), and streptavidin peroxidase
incubation (10 min). Visualization was performed using a freshly prepared
DAB substrate/chromogen solution.
For the immunohistochemical localization of malondialdehyde goat
anti-malondialdehyde antibody (purched from United States Biological,
Swampscott, Massachusetts, USA) and Vectastain Elite ABC Kit (Vector
Laboratories, Burlingame, CA, USA) were employed. Staining procedure
was equivalent to Vectastain Elite ABC protocol: quenching of endoge-
nous peroxidase activity (0.3% H2O2, 20 min), normal blocking serum
incubation (20 min), primary antibody incubation (30 min), biotinylated
secondary antibody incubation (30 min), Vectastain Elite ABC reagent
incubation (30 min), DAB incubation (2 min).
The cytokines were localized immunohistochemically using following
antibodies: rabbit polyclonal to IL-1beta (primarily directed against
mature 17 kDa human IL-1beta), rabbit polyclonal to IL-6, rabbit poly-
clonal to IL-8 and rabbit polyclonal to TNF alpha (Abcam, Cambridge,
UK). The anti-rabbit HRP/DAB Ultravision Detection System (Lab
Vision, Fremont, CA, USA) was employed in accordance with manufac-
turer recommendations: hydrogen peroxide block (12 min), ultra V block
(5 min), primary antibody incubation (90 min), biotinylated goat anti-rab-
bit antibody incubation (10 min), streptavidin peroxidase incubation
(10 min), and DAB substrate/chromogen incubation (2 min).
Some samples were counterstained with Mayer ´s hematoxylin (Sigma).
Negative controls included the omission of the primary antibody. To con-
firm the specifity of the staining for nitrotyrosine, the antibody was incu-
bated with 10 mM nitrotyrosine [23].
After the rinsing in tap water, mounted in diluted Aquatex (Merck,
Darmstadt, Germany) and immediately examined using an Orthoplan
Leitz light microscope equipped with a Leica DC 500 digital camera.
Image analysis of NOS expression
Image analysis was performed on stored images taken on Orthoplan
Leitz light microscope equipped with a Leica DC 500 digital camera using
Leica Image Manager (Version 4.0). The luminance of all pixels in the
region of interest (ROI) was measured using Neurolucida image analysis
(MBF Bioscience, Williston, USA), and the optical density calculated as
OD (optical density) = 255 (maximal luminance value) ? actual lumi-
nance of the ROI. In NOS3 OD measurements, each experimental group
contained 12 clusters of cells; the clusters of cells were analyzed in groups
of 6–7, in total 76–80 cells were measured. In NOS2 OD measurements,
image analysis was done in 50 individual cells from each staining group.
Statistical analysis
For statistical analysis a one-way ANOVA with Tukey–Kramer multi-
ple comparison tests was employed, using GraphPad InStat software
(GraphPad Software, San Diego, CA, USA).
Results
Immunohistochemistry
In the conjunctival epithelium of dry eye NOS3 was
gradually expressed along the severity of dry eye symptoms
(Fig. 1b, dry eye with less evident symptoms of dryness,
grade 2; Fig. 1c, dry eye with pronounced symptoms of
dryness, grade 3). This is in contrast to the conjunctival epi-
thelium of normal eye (Fig. 1a), where NOS3 is only
slightly expressed. In the control sample for NOS3 (pri-
mary antibody omitted) no staining appears (Fig. 1d). Very
similar results were obtained with NOS2. Also the staining
of NOS2 in the conjunctival epithelium of dry eye increases
together with the more profound expression of dryness
symptoms (Fig. 1f, grade 2; Fig. 1g, grade 3). In the normal
eye (Fig. 1e) NOS2 is only slightly expressed in the con-
junctival epithelium. Fig. 1h—control (primary antibody
omitted) specimen for NOS2; no staining appears.
The staining of NOS1 is weak in conjunctival epithelium
of normal eye and does not significantly change in dry eye.
Peroxynitrite formation (demonstrated by nitrotyrosine
residues)andlipidperoxidation
increased malondialdehyde staining) can be seen only in
conjunctival epithelium of dry eye (SS, grade 3) (Figs. 2b
and 3a).
Normal conjunctival cytology samples do not reveal
cytokine staining or the staining is very weak (80–85% con-
junctival cells were negative, 15–20% slightly positive)
(IL1-beta, Fig. 4a; IL-6, Fig. 4d; IL-8, Fig. 4g; TNF alpha,
Fig. 4j). Only nuclei are stained with haematoxylin. How-
ever, all cytokine studied were already expressed in the con-
junctival epithelium of dry eye (SS) grade 2 (IL1-beta,
Fig. 4b; IL-6, Fig. 4e; IL-8, Fig. 4h; TNF-alpha Fig. 4k)
and the expression increased in the conjunctival epithelium
of dry eye (SS)—grade 3, highly pronounced symptoms of
dryness (IL-1 beta Fig. 4c; IL-6, Fig. 4f; IL-8, Fig. 4i; TNF-
alpha, Fig. 4l). (The cytokine expression studied was pres-
ent in 90–95% of conjunctival cells; dry eye samples, grade
2, grade 3). Arrows point to cytokine expression. In con-
trols (primary antibody omitted) no staining appears (not
shown).
(demonstratedby
Image analysis
The results of optical density measurements are shown
in Figs. 5 and 6.
12
J. Cˇejkova ´ et al. / Nitric Oxide 17 (2007) 10–17

Page 4

Fig. 2. Malondialdehyde staining in the conjunctival epithelium of normal eye and dry eye (SS, grade 3); nuclei were counterstained with haematoxylin.
(a) In the normal eye malondialdehyde staining is only slightly pronounced (arrows). (b) In dry eye (SS, grade 3) the expression of malondialdehyde is
highly pronounced (arrows).
Fig. 3. Peroxynitrite formation in conjunctival epithelium of dry eye (SS, grade 3) demonstrated by nitrotyrosine residues; nuclei were counterstained with
haematoxylin. (a) In dry eye (SS, grade 3) nitrotyrosine is highly expressed (arrows). (b) in control for nitrotyrosine where primary antibody was incubated
with 10 mM nitrotyrosine, no staining appears; only nuclei are stained.
Fig. 1. NOS3 and NOS2 expression in the conjunctival epithelium of normal eye and dry eye (SS). Scale bar: 10 lM. In the normal eye NOS3 is only
slightly expressed and the expression of NOS2 is nearly absent (a, NOS3; e, NOS2). In dry eye NOS3 as well as NOS2 are gradually expressed along the
severity of dry eye symptoms. b, NOS3 and f, NOS2 in dry eye with less pronounced symptoms of dryness—grade 2); c, NOS3 and g, NOS2 in dry eye with
highly pronounced symptoms of dryness—grade 3). In controls for NOS3, d and control for NOS2, h where primary antibodies were omitted, no staining
appears.
J. Cˇejkova ´ et al. / Nitric Oxide 17 (2007) 10–17
13

Page 5

NOS3: In normal eyes (Normal) only 1-3 conjunctival
cells from each cluster were positive for NOS3. In
patients with dry eye (Dry) (grade 2) and also in patients
with extreme dry eye (Extra Dry) (grade 3), all conjunc-
tival cells from each cluster were positively stained for
NOS3.
NOS2: Very similar results as described for NOS3
were also obtained for NOS2. In normal eyes (Normal)
NOS2 was very weak, NOS2 expression was increased
inpatientswith dryeye
highly increased in patients with dry eye (Extra Dry)
(grade 3).
(Dry) (grade 2)and
Discussion
Nitric oxide (NO), produced by the oxidative deamina-
tion of L-arginine by NOS, is an important mediator of
homeostatic processes in the eye, such as the regulation
of aqueous humor dynamics, retinal neurotransmission
and phototransduction [24,25]. In the cornea, NO sponta-
neously produced in the corneal endothelium is highly
involved in the maintainance of corneal thickness [26].
However, changes in NO generation or action have been
foundto contributeto various
[12,15,18,24,27,28–32].
pathological states
Fig. 4. The cytokine demonstration (IL-1 beta, IL-6, IL-8, TNF-alpha) in conjunctival epithelium of dry eye (SS, grade 2 and grade 3) and conjunctival
epithelium of normal eyes. Nuclei were counterstained with haematoxylin. Scale bar: 10 lM. In normal eyes the stainings of IL-1 beta (a), IL-6 (d), IL-8
(g), TNF-alpha (j) cytokines is nearly absent. Only nuclei are stained with haematoxylin. In dry eye conjunctival samples the stainings of IL-1 beta
cytokine (b, grade 2; c, grade 3), IL-6 cytokine (e, grade 2; f, grade 3), TNF-alpha (k, grade 2; l, grade 3) are gradually expressed along the severity of dry
eye symptoms; from grade 2 (less pronounced symptoms of dryness) to grade 3 (much more pronounced symptoms of dryness). Arrows point to cytokine
expression.
14
J. Cˇejkova ´ et al. / Nitric Oxide 17 (2007) 10–17