Content uploaded by John H Chang
Author content
All content in this area was uploaded by John H Chang
Content may be subject to copyright.
Expression of Toll-like Receptor 4 and Its Associated
Lipopolysaccharide Receptor Complex by Resident
Antigen-Presenting Cells in the Human Uvea
John H. Chang, Peter McCluskey, and Denis Wakefield
PURPOSE.To investigate the in vivo expression of toll-like re-
ceptor 4 (TLR4) and its associated lipopolysaccharide (LPS)
receptor complex in the human eye.
METHODS.Normal human ocular tissues were evaluated for in
vivo TLR4, MD-2, and CD14 mRNA and protein expression by
RT-PCR and immunohistochemistry, respectively. The distribu-
tion patterns and phenotypes of the cells expressing these
proteins were further characterized by confocal microscopy
and double-label immunofluorescence studies.
RESULTS.Normal human uvea, retina, sclera, and conjunctiva
constitutively expressed TLR4, MD-2, and CD14 mRNA. The
protein expression of these molecules was restricted, how-
ever, to resident antigen-presenting cells (APCs) in the normal
human uvea, consisting mainly of HLA-DR
⫹
dendritic cells
(DCs). These APCs endowed with the complete LPS receptor
complex appeared to be strategically positioned in perivascu-
lar and subepithelial locations for surveying blood-borne or
intraocular LPS. In contrast, other cell types of the normal
human cornea, conjunctiva, retina, and sclera did not express
TLR4/MD-2 protein in vivo as detectable by immunohistochem-
istry.
CONCLUSIONS.The present study demonstrates for the first time
that resident APCs in the normal human uvea express TLR4 and
its associated LPS receptor complex. This has significant impli-
cations for the understanding of normal ocular immunity as
well as unraveling the potential role of Gram-negative bacteria
in the pathogenesis of acute anterior uveitis (AAU). (Invest
Ophthalmol Vis Sci. 2004;45:1871–1878) DOI:10.1167/
iovs.03-1113
Uveitis is a relatively common form of potentially sight-
threatening intraocular inflammatory disease that predom-
inantly affects the iris, ciliary body, and choroid. Most forms of
uveitis are presumed to be of autoimmune etiology resulting
from a breakdown in the normal state of ocular immune priv-
ilege, although the exact pathogenic mechanisms for this or
the precise nature of the initiating cause(s) are unclear. Ante-
rior uveitis is the most common form of uveitis in most regions
of the world.
1
There is substantial clinical and experimental
evidence in human studies implicating Gram-negative bacteria
as triggers in the pathogenesis of acute anterior uveitis (AAU),
particularly those that are associated with HLA-B27. Among the
implicated infective triggers are the Gram-negative enterobac-
teria such as certain species of Klebsiella,Salmonella,Shi-
gella, and Yersinia.
2
Furthermore, in endotoxin-induced uve-
itis (EIU), a well-established animal model of AAU, the
administration of lipopolysaccharide (LPS) of Gram-negative
bacteria to certain susceptible strains of rodents, via various
routes remote from the eye, induces an acute and preferential
inflammation of the iris and ciliary body that closely resembles
AAU in humans.
3
Toll-like receptors (TLRs) are a recently discovered family of
type I transmembrane, pattern-recognition receptors (PRRs)
that are essential in the recognition of the highly conserved
pathogen-associated molecular patterns (PAMPs) that are
unique to microbes, such as LPS of Gram-negative bacteria,
mannans of yeast cell wall, and viral double-stranded RNA. To
date, 10 TLRs (TLR1 to -10) have been described, each recog-
nizing a specific class or classes of PAMP.
4,5
Stimulation of the
TLR, by its respective and specific PAMP, results in the activa-
tion of an immunostimulatory and immunomodulatory cell-
signaling pathway that is essential for innate immunity and for
the activation of the adaptive arm of the immune response.
4,6
It is now well established that TLR4 is the primary signaling
receptor for LPS-specific recognition and cellular activation.
5,7
CD14 is a glycosyl phosphatidylinositol-anchored cell surface
protein that functions as a coreceptor for LPS, as it, unlike
TLR4, is unable to activate cellular signal transduction due to
the absence of an intracellular signaling domain.
8
MD-2 is an
accessory molecule that associates with the extracellular do-
main of TLR4 conferring on it LPS responsiveness and is an
absolute requirement in TLR4-dependent LPS responses in
vivo.
9,10
TLR expression has been demonstrated on peripheral blood
monocytes,
11
dendritic cells (DCs),
12
B cells,
13
and dermal
vascular endothelial cells,
14
as well as in various human tissues
including lymphoid tissues,
15
intestinal epithelial cells,
16
and
skin.
17
The TLRs have also been implicated in the pathogenesis
of a variety of inflammatory or autoimmune human diseases
such as rheumatoid arthritis,
18
inflammatory bowel disease,
16
and psoriasis.
17
Song et al.
19
have recently reported the ex-
pression of functional TLR4 and CD14 in cultured human
corneal epithelial cells with implications for a role in the
pathogenesis of Gram-negative bacterial keratitis.
19
Although
CD14 expression was demonstrated in fresh whole human
corneas by immunohistochemistry, the in vivo corneal expres-
sion of TLR4 was not reported in that study. To date, no
investigations have been conducted, to the best of our knowl-
edge, to evaluate the in vivo expression of TLR4 in the normal
human eye, including the uvea. Given the particular predilec-
tion for the involvement of the uveal tract in intraocular in-
flammatory disease and the implicated role of Gram-negative
bacteria in their pathogenesis, the purpose of the present study
From the Laboratory of Ocular Immunology, Inflammatory Dis-
eases Research Unit, School of Medical Sciences, University of New
South Wales, Sydney, Australia.
Supported by National Health and Medical Research Council
(NH&MRC). JHC is a recipient of a NH&MRC Medical Postgraduate
Research Scholarship (Grant 222928).
Submitted for publication October 7, 2003; revised January 15,
2004; accepted February 12, 2004.
Disclosure: J.H. Chang, None; P. McCluskey, None; D. Wake-
field, None
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be marked “advertise-
ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Corresponding author: Denis Wakefield, School of Medical Sci-
ences, University of New South Wales, Sydney, NSW 2052, Australia;
d.wakefield@unsw.edu.au.
Investigative Ophthalmology & Visual Science, June 2004, Vol. 45, No. 6
Copyright © Association for Research in Vision and Ophthalmology 1871
was to determine the in vivo expression patterns of TLR4,
MD-2, and CD14, collectively constituting the LPS-signaling
receptor complex, in normal human ocular tissues. The results
of this study demonstrate, for the first time, that resident
antigen-presenting cells (APCs), including a significant subset
of the DCs in the normal human uvea, particularly within the
ciliary body and iris root, express all the components of the
LPS-signaling receptor complex. This finding has important
implications for our understanding of innate immunity and
ocular immune privilege in the eye. The identification of TLR4
and its associated LPS receptor complex in the human uvea
may also be of fundamental significance in unraveling the
potential role of Gram-negative bacterial triggers of AAU.
METHODS
Human Ocular Tissues
Twelve human eyes were obtained from six donors (Lions NSW Eye
Bank, Sydney, Australia) within 24 hours of death, in accordance with
institutional review board–approved protocol and the provisions of the
Declaration of Helsinki for research involving human tissue. The mean
age of the donors was 67.7 years with a range of 45 to 85 years (four
males and two females). No donors were known to have had any
ocular or systemic autoimmune or inflammatory disease and none
were on immunosuppressive therapy.
Eyes (n⫽2) used for total RNA isolation were enucleated within 4
hours of death, the corneas removed for grafting, and the remaining
ocular tissue transported in RNA stabilization solution (RNAlater; Am-
bion Inc., Austin, TX) at 4°C until further processing in the laboratory.
Iris, ciliary body, choroid, retina, cornea, sclera, and bulbar conjunctiva
were dissected and snap frozen in liquid nitrogen (for RNA isolation) or
were embedded in OCT medium (Sakura Finetek, Torrance, CA), snap
frozen, and sectioned at 4- to 20-
m thickness (for immunohistochem-
ical studies). These ocular tissues were stored at ⫺80°C until used.
Cryostat tissue sections were fixed for 10 minutes in acetone before
performing immunohistochemical studies.
RNA Isolation and RT-PCR Analysis
Total RNA was isolated from the iris, ciliary body, choroid, retina,
bulbar conjunctiva, and sclera (RNeasy Micro Kit; Qiagen Inc., Valen-
cia, CA) according to the manufacturer’s instructions and treated with
RNase-free DNase I. The RNA was reverse transcribed to single-
stranded cDNA using random hexamer-primed reverse transcriptase
(Superscript II RNase H
⫺
; Invitrogen, Carlsbad, CA). PCR amplification
was performed with Taq DNA polymerase (Platinum Taq; Invitrogen)
using the following gene-specific primers: TLR4
20
sense, 5⬘-TGGAT-
ACGTTTCCTTATAAG-3⬘and antisense, 5⬘-GAAATGGAGGCACCC-
CTTC-3⬘; MD-2
21
sense, 5⬘-GAAGCTCAGAAGCAGTATTGGGTC-3⬘and
antisense, 5⬘-GGTTGGTGTAGGATGACAAACTCC-3⬘; CD14
20
sense, 5⬘-
CCATGGAGCGCGCGTCCTGC-3⬘and antisense, 5⬘-GTCTTGGATCT-
TAGGCAAAGC-3⬘; and glyceraldehyde-3-phosphate dehydrogenase
(GAPDH)
22
sense, 5⬘-ACCACAGTCCATGCCATCAC-3⬘and antisense,
5⬘-TCCACCACCCTGTTGCTGTA-3⬘. Optimized PCR conditions were as
follows: TLR4 (32 cycles of 95°C for 30 seconds, 56°C for 30 seconds,
and 72°C for 45 seconds); MD-2 (28–32 cycles, depending on the
tissue source, of 95°C for 30 seconds, 60°C for 30 seconds, and 72°C
for 25 seconds); CD14 (35 cycles of 95°C for 30 seconds, 62°C for 2
minutes, and 72°C for 3 minutes); and GAPDH (28 cycles of 95°C for
30 seconds, 58°C for 30 seconds, and 72°C for 45 seconds). RT-PCR on
RNA extracted from human peripheral blood monocytes that are
known to express TLR4, MD-2, and CD14 mRNA served as positive
controls for the gene-specific primers.
11
Negative control reactions
included performing the PCR under identical conditions except for the
omission of the reverse transcriptase, the template cDNA, or the
primers. PCR products were analyzed by electrophoresis on 2% aga-
rose gels.
Single-Labeling Immunohistochemistry
Endogenous peroxidase activity in the cryostat tissue sections was
blocked by 0.3% H
2
O
2
and sodium azide for 5 minutes at room tem-
perature. After three washes with tris-buffered saline (TBS) of 5 min-
utes each, sections were incubated with 20% normal goat or rabbit
serum diluted in 2% bovine serum albumin (BSA)/TBS for 30 minutes
at room temperature to block nonspecific binding. Sections were then
incubated with the primary antibodies appropriately diluted in 2%
BSA-TBS in a humidified chamber overnight at 4°C. The following
primary monoclonal antibodies against the indicated specificity in
human tissues were used: anti-TLR4/MD-2 complex, 1
g/mL (clone
HTA125, mouse IgG2a; Santa Cruz Biotechnology, Santa Cruz, CA);
anti-CD14, 2.5
g/mL (clone B365.1, mouse IgG1,
; Biomeda Corp.,
Foster City, CA); anti-HLA-DR, 1
g/mL (clone YD1/63.4.10, rat IgG2a;
Cedarlane, Ontario, Canada); anti-CD68, 12.5
g/mL (clone PG-M1,
mouse IgG3,
; Dako, Glostrup, Denmark); and anti-von Willebrand
Factor, 6.3
g/mL (clone F8/86, mouse IgG1,
; Dako). After three
washes with TBS, the sections were incubated with biotinylated rabbit
anti-rat (3.3
g/mL, Dako) or goat anti-mouse (7.5
g/mL; Vector
Laboratories, Burlingame, CA) secondary antibodies for 30 minutes at
room temperature. After three further washes in TBS, the sections
were incubated with horseradish peroxidase-conjugated streptavidin
(Vector Laboratories) for 1 hour at room temperature. Immunolocal-
ization was performed with the addition of the substrate 3-amino-9-
ethylcarbazole (Sigma-Aldrich, St. Louis, MO) and the chromogen de-
velopment monitored by light microscopy. The reaction was stopped
once suitable color had developed or after a maximum of 10 minutes.
Sections were counterstained with hematoxylin and mounted in crys-
tal mounting medium (Biomeda Corp.).
As a positive control, human lymphoid tissues were stained for
TLR4/MD-2, CD14, HLA-DR, and CD68.
5,11–13
Negative controls in-
cluded the replacement of the primary antibody with species- and
isotype-matched monoclonal antibodies at the same concentrations or
the omission of the primary antibody.
Double-Labeling Immunofluorescence
Microscopy
Double immunofluorescence was performed by serially incubating the
cryostat tissue sections with two primary antibodies of different spe-
cies, using the same antibodies at the concentrations used for the
immunoperoxidase staining. After blocking nonspecific binding as
previously described with 20% normal rabbit serum, the sections were
incubated with the first primary antibody, rat anti-human HLA-DR
monoclonal antibody, for overnight at 4°C. After they were washed
with TBS, sections were incubated with biotinylated rabbit anti-rat
antibody for 30 minutes. After further washes with TBS, the sections
were then incubated with Alexa 488–conjugated streptavidin (2
g/
mL, excitation-emission maxima of 495/519 nm; Molecular Probes,
Eugene, OR) for 1 hour in the dark at room temperature. All subse-
quent steps were performed at room temperature with the sections
protected from light.
The sections were washed with TBS and blocked with 20% normal
goat serum for 30 minutes and then incubated for 3 hours with the
second primary antibody, which was a mouse anti-human monoclonal
antibody (anti-TLR4/MD-2, anti-CD14, or anti-CD68). After they were
washed with TBS, sections were incubated with Alexa 568–conju-
gated goat anti-mouse antibody (10
g/mL, excitation-emission max-
ima of 578/603 nm; Molecular Probes) for 1 hour. After further wash-
ing, slides were mounted in anti-fade medium (Vectashield; Vector
Laboratories). Negative controls included the replacement of the first
or second primary antibody or of both antibodies with the species- and
isotype-matched irrelevant antibodies.
1872 Chang et al. IOVS, June 2004, Vol. 45, No. 6
Slides were examined by microscope (BX60; Olympus, Tokyo,
Japan) equipped with a 100-W mercury burner for epifluorescence
illumination and wide-band interference filters for blue excitation-
green emission (460–490-nm band-pass, 515-nm long-pass) as well as
green excitation-red emission (520–550-nm band-pass, 590-nm long-
pass). Images were captured with a digital camera (Spot Cooled Color
Digital Camera; Diagnostic Instruments, Sterling Heights, MI) and the
pairs of images were superimposed for colocalization analysis using
image-management software (Photoshop ver. 7; Adobe Systems, Moun-
tain View, CA).
Confocal Microscopy
Double immunofluorescence of tissue sections was also examined by
inverted confocal laser scanning microscope (TCS; Leica Microsystems,
Mannheim, Germany) fitted with helium and argon lasers (Leica).
FIGURE 1. RT-PCR analysis demonstrated (A) TLR4, (B) MD-2, and (C) CD14 mRNA expression in fresh normal human ocular tissues. Single distinct
PCR product bands of the expected size were detected for TLR4 (506 bp), MD-2 (422 bp), and CD14 (1140 bp) in the iris/ciliary body (lane 1),
choroid (lane 2), retina (lane 3), sclera (lane 4), and conjunctiva (lane 5) by 2% agarose gel electrophoresis. cDNA from peripheral blood
mononuclear cells (PBMC) served as the positive control (lane 6), as these are known express TLR4, MD-2, and CD14. No products were generated
in the control reactions without the reverse transcriptase (RT⫺,lane 7), no-primer control (lane 8), and no-template control reactions (lane 9).
GAPDH was coamplified in all RT-PCR reactions (D).
IOVS, June 2004, Vol. 45, No. 6 Expression of TLR4 and Its LPS Receptor Complex by APCs in the Uvea 1873
1874 Chang et al. IOVS, June 2004, Vol. 45, No. 6
RESULTS
In Vivo Expression of TLR4, MD-2, and CD14
mRNA in Normal Human Ocular Tissue
We first performed RT-PCR analysis of normal fresh human
ocular tissues to investigate the in vivo expression of the LPS
receptor complex at the mRNA level. The normal human iris,
ciliary body, choroid, retina, sclera, and conjunctiva express
the mRNA transcripts for TLR4, MD-2, and CD14 (Fig. 1).
Agarose gel electrophoresis analysis demonstrated a single dis-
tinct band of the expected size for TLR4 (506 bp), MD-2 (422
bp), and CD14 (1140 bp) from the respective gene-specific
PCR amplification of the mRNA from normal human iris, ciliary
body, choroid, retina, sclera, and conjunctiva. No band was
observed in the control reactions with template and primers,
under identical PCR conditions, except for the omission of the
reverse transcriptase, indicating that the products were gener-
ated from mRNA and not from any contaminating genomic
DNA. RT-PCR with primers for GAPDH also showed a single
distinct product of the expected size (Fig. 1D), establishing the
integrity and relative abundance of the RNA isolated from the
various ocular tissues. RT-PCR analysis was unable to be per-
formed for the cornea, due to the unavailability of a suitable
normal human corneal tissue without significant RNA degrada-
tion.
Immunolocalization of TLR4, MD-2, and CD14 in
the Human Uvea
We next investigated the in vivo protein expression of TLR4,
MD-2, and CD14 in normal human ocular tissues by immuno-
histochemistry, which also allowed the examination of the
phenotype and pattern of distribution of the cells that ex-
pressed these proteins. The expression of these molecules was
restricted to a subpopulation of resident stromal cells of the
ciliary body, iris, and choroid (Fig. 2). These positive-staining
stromal cells demonstrated dendritiform morphology and a
pattern of distribution that was also consistent with the regular
network of resident APCs within the normal human uvea as
previously reported by others.
23–25
Immunoperoxidase stain-
ing for HLA-DR confirmed that these dendritiform stromal cells
were HLA-DR
⫹
and most were negative for CD68, a macro-
phage marker (data not shown), a pattern consistent with
previous reports for the human uvea
24
and thus further sug-
gesting that most of these cells were indeed resident uveal
DCs. The distribution of TLR4/MD-2
⫹
cells within the stroma
of ciliary body, iris, and choroid correlated with the distribu-
tion pattern and morphology of CD14
⫹
and HLA-DR
⫹
cells in
these tissues (Figs. 2A, 2B, 2H–K, and data not shown), sug-
gesting that TLR4/MD-2 complex and CD14 were coexpressed
on HLA-DR
⫹
resident APCs within the uveal tract. Therefore,
we next performed double-labeling immunohistochemistry
studies to confirm these findings.
Coexpression of TLR4/MD-2 and CD14 by
HLA-DR
ⴙ
APCs in the Human Uvea
To definitively determine the phenotype of TLR4/MD-2
⫹
cells
in uveal tissue, a series of double-immunofluorescence studies
were performed. First, uveal tissues were double-labeled for
HLA-DR with a green-emitting fluorochrome (Alexa-488, Fig.
2C) and for TLR4/MD-2 with a red-emitting fluorochrome
(Alexa-568, Fig. 2D). Confocal microscopy confirmed the den-
dritiform morphology of these HLA-DR
⫹
(Fig. 2C) and TLR4/
MD-2
⫹
cells (Fig. 2D) that were arranged in a regular network
within uveal tissues. These dendritiform stromal cells were
confirmed to coexpress TLR4/MD-2 complex and HLA-DR (Fig.
2E). Double-immunofluorescence studies for HLA-DR and
CD68 of serial sections demonstrated that most of of these
HLA-DR
⫹
cells were CD68
⫺
(data not shown). Therefore, these
HLA-DR
⫹
TLR4/MD-2
⫹
stromal cells fulfilled the currently ac-
cepted immunomorphologic criteria for DCs.
23,26
Although
most of the HLA-DR
⫹
DCs coexpressed TLR4/MD-2, there was
also a subpopulation of HLA-DR
⫹
TLR4/MD-2
⫺
DCs (Fig. 2F).
The expression pattern of the various TLRs in human DCs has
been reported to be different among the DC subsets, with the
myeloid DCs expressing TLR4 whereas the plasmacytoid DCs
do not.
5,27
Thus, the results of our double-immunofluores-
cence experiments are consistent with these reports and most
probably reflect the heterogeneous nature of the resident DC
population within the uvea. The subpopulation of HLA-DR
⫹
TLR4/MD2
⫹
DCs observed in our study most probably repre-
sents the myeloid DC subset and the HLA-DR
⫹
TLR4/MD-2
⫺
DCs representing plasmacytoid DC.
Double immunohistochemistry for TLR4/MD-2 and CD68
could not be performed, because these primary antibodies
Š
FIGURE 2. Immunohistochemical studies for TLR4/MD-2, CD14, and HLA-DR in human uvea. (A) Network of resident stromal cells at the junction
of the iris root (to the right in the photo) and ciliary body (to the left and superiorly) staining positively, as visualized with a red reaction product,
for TLR4/MD-2 complex. A subset of these TLR4/MD-2
⫹
stromal cells in the iris root and ciliary body were located adjacent to blood vessels (arrow)
that were confirmed to be vascular structures by their positive staining for von Willebrand factor, a marker for vascular endothelium (inset). Note
that the uveal vascular endothelium did not stain positively for TLR4/MD-2. (B) Higher power view of a similar region of the iris root and ciliary
body demonstrating a similar distribution pattern and morphology of positive-staining stromal cells for CD14, including their perivascular location
(arrowhead). No staining was seen with an isotype-matched irrelevant antibody (inset). (C–E) Double-labeling immunofluorescence microscopy
for colocalization studies in normal human ocular tissues using a (C)green-emitting (Alexa-488; Molecular Probes) and a (D)red-emitting
(Alexa-568; Molecular Probes) fluorochrome. Images shown are from the iris root and ciliary body, however similar staining patterns were seen
in other uveal tissues. (C) Confocal microscopy confirmed the dendritiform morphology of HLA-DR
⫹
(green fluorescence) resident stromal cells
in the iris root and ciliary body. Most of these dendritiform HLA-DR
⫹
cells were CD68
⫺
(data not shown). (D) These dendritic cells also expressed
TLR4/MD-2 (red fluorescence). Negative controls with the replacement of the first and/or second primary antibody with an isotype-matched
control antibody showed no staining (inset). (E) Colocalization of HLA-DR and TLR4/MD-2 (yellow). The inset demonstrates a high-resolution
confocal microscopy view of a DC coexpressing HLA-DR (green) and TLR4/MD-2 (red). (F) A membranous pattern of staining (arrowhead) was
seen for HLA-DR (green) and TLR4/MD-2 (red). A subset of these HLA-DR
⫹
DCs was TLR4/MD-2
⫺
(unmarked). (G) Double immunofluorescence
demonstrating coexpression (yellow) of HLA-DR (green) and CD14 (red) by resident stromal cells in the iris and ciliary body. High power views
of the ciliary process demonstrating positive-staining dendritiform stromal cells for TLR4/MD-2 (H) and CD14 (I), and no staining in the negative
controls (inset). Note the similar pattern of distribution of these TLR4/MD-2
⫹
and CD14
⫹
cells, including their subepithelial locations. The ciliary
epithelium and vascular endothelium did not stain positively for TLR4/MD-2 or CD14. (J) TLR4/MD-2
⫹
stromal cells (arrowhead) in the choroid
demonstrated a perivascular distribution (arrow: a choroidal blood vessel). Inset: von Willebrand factor–positive blood vessels in a serial section.
There was no positive staining for TLR4/MD-2 in the sclera or retina. (K) Perivascular distribution of CD14
⫹
stromal cells in the choroid. (L)
Negative control with an isotype-matched antibody. These images are representative of independent experiments performed on eyes from five
different donors. Original magnifications: (A,J)⫻200; (B,I,K,L)⫻400; (C–E)⫻600; (F–H)⫻1000.
IOVS, June 2004, Vol. 45, No. 6 Expression of TLR4 and Its LPS Receptor Complex by APCs in the Uvea 1875
were from the same species and thus indistinguishable at the
level of the secondary antibody. Therefore, the expression of
TLR4/MD-2 by resident uveal macrophages could not be di-
rectly examined in this study. Previous studies have shown that
monocytes/macrophages have high levels of expression of
TLR4
5,12,27
and thus it is likely that a minor subset of the
HLA-DR
⫹
TLR4/MD-2
⫹
cells demonstrated in this study were
resident uveal macrophages, although this has not been defin-
itively shown in the current experiments. Similarly, direct
colocalization studies for TLR4/MD-2 and CD14 were unable to
be performed as these antibodies also originated from the same
species. However, double immunofluorescence for HLA-DR
and CD14 demonstrated that HLA-DR
⫹
APCs coexpress CD14
(Fig. 2G). Thus, these results demonstrate that HLA-DR
⫹
APCs
coexpress both TLR4/MD-2 and CD14 that together constitute
the complete LPS-signaling receptor complex.
Pattern of Expression of the LPS Receptor
Complex in Normal Human Ocular Tissues
Within the normal human uvea, there was a relatively rich
network of TLR4/MD-2
⫹
CD14
⫹
APCs in the iris root and
stroma of the ciliary body, compared to the iris or choroid in
which only occasional positive-staining cells were found (Figs.
2 and data not shown). Some of these TLR4/MD-2
⫹
CD14
⫹
resident APCs within the uvea displayed a perivascular (Figs.
2A, 2B, 2J, 2K) or subepithelial location (Figs. 2H, 2I). No other
cell types in the normal human uvea expressed TLR4/MD-2 or
CD14 protein. Notably, in contrast to other tissues such as the
skin,
17,28
TLR4/MD-2 complex was not expressed on the epi-
thelial cells of the uvea, such as the ciliary or iris epithelium,
nor on the vascular endothelium (Figs. 2). Possible masking of
positive chromogenic staining by the melanin granules in the
pigmented epithelium of the iris and ciliary body was excluded
by immunofluorescence microscopy (data not shown).
We were unable to demonstrate in vivo protein expression
of TLR4/MD-2 complex in the normal human cornea, conjunc-
tiva (Fig. 3), retina, or sclera (Fig. 2J) as detectable by immu-
nohistochemistry.
DISCUSSION
In this study we demonstrate, for the first time, the in vivo
expression of the LPS receptor complex, namely TLR4, MD-2,
and CD14, by resident APCs, mostly HLA-DR
⫹
DCs, within the
normal human uvea. This finding has significant implications
for our understanding of the innate and adaptive immunity of
the eye as well as of the pathogenesis of ocular inflammatory
diseases, such as uveitis and Gram-negative bacterial endo-
phthalmitis.
Activation of TLR4 by its principal agonist, LPS, results in an
immunostimulatory intracellular signaling pathway leading to
the activation of the transcriptional factor, nuclear factor-
B
(NF-
B), and consequently the induction of various proin-
flammatory cytokines, chemokines, and antimicrobial activi-
ties.
4–6,29,30
The present study has shown that TLR4
⫹
MD-2
⫹
CD14
⫹
APCs appear to be strategically placed in perivascular
and subepithelial locations within the normal uvea and sug-
gests that such uveal APCs endowed with the complete LPS-
signaling receptor complex are optimally positioned to survey
and respond to either blood-borne or intraocular LPS of Gram-
negative bacteria. Thus, these innate immune cells in the uvea
would be expected to be able to respond rapidly to LPS of
Gram-negative bacteria, in contrast to the primary adaptive
immune response that requires several days to become effec-
tive as this involves clonal expansion and maturation of naive
effector cells. By becoming activated through its pattern rec-
ognition of the PAMP, LPS receptor–positive macrophages
would acquire enhanced effector functions such as phagocytic
activity,
5
thus facilitating the rapid containment and eradica-
tion of the microbe with minimal tissue damage. In addition to
their critical role in the innate immune response, TLRs have
been recognized to be important in the efficient priming and
FIGURE 3. (A) Positive staining for
TLR4/MD-2 by a subpopulation of
macrophage-like cells in the thymus
(positive tissue control). Inset:
higher power view of TLR4/MD-2
⫹
cells. (B) No staining was seen in the
thymus when using identical experi-
mental conditions but with the re-
placement of the primary antibody
with an isotype-matched irrelevant
antibody at the same concentration
(negative control). (C) Normal hu-
man cornea did not constitutively ex-
press TLR4/MD-2 complex. Upper
left inset: higher power view of the
corneal epithelium (epith.) and up-
per stroma. Lower right inset: higher
power view of the corneal endothe-
lium (endoth.) and lower stroma. (D)
Normal human conjunctiva did not
express TLR4/MD-2 complex in vivo.
Original magnifications: (A–C)⫻200;
(D)⫻400.
1876 Chang et al. IOVS, June 2004, Vol. 45, No. 6
initiation of the adaptive immune response by its activation of
APCs.
5,6,29
We propose that TLR4-dependent stimulation of
resident DCs by LPS in the human uvea induces DC maturation
with the induction of costimulatory molecules such as CD80
and CD86, the upregulation of major histocompatibility com-
plex class II molecules, and the secretion of proinflammatory
cytokines and chemokines that would recruit naive T cells and
other inflammatory cells to the uvea for activation as has been
shown in other settings.
4–7,27,29,31,32
The perivascular location
of TLR4
⫹
DCs in the uvea places them in an unique position to
activate the expression of vascular adhesion molecules and the
initiation of the multistep process of leukocyte recruitment
33
to the uvea through rolling, adhesion, and transendothelial
migration toward the chemokines secreted by these LPS-acti-
vated perivascular DCs. Furthermore, LPS activation of DCs via
TLR4 has been associated with a Th1 polarized immune re-
sponse by inducing cytokines such as IL-12 and thus may have
particular relevance to the pathogenesis of uveitis.
29
The initiating factors in AAU are unknown despite intensive
investigation. In addition to genetic factors, the most important
of which is the strong association with HLA-B27, there is
substantial evidence suggesting the role of environmental fac-
tors in its pathogenesis, particularly that of multiple Gram-
negative bacteria.
2
The development of AAU has been reported
in patients after acute Yersinia or Salmonella infections.
34,35
There have also been numerous studies demonstrating sero-
logic evidence of infection with these implicated Gram-nega-
tive bacteria in patients with AAU.
2,36
The inappropriate
and/or exaggerated TLR4-mediated activation of the innate and
adaptive immune responses within the uvea by LPS of the
implicated Gram-negative bacteria may be a major contributing
factor in the initiating mechanisms of AAU. LPS may also act as
an adjuvant by activating APC maturation in the presence of
the putative uveitogenic self-antigen(s) and thus mediate the
breakdown of peripheral tolerance resulting in the induction
of an autoimmune response. Current hypothesis on peripheral
tolerance suggests that APCs that capture self-antigens present
them to autoreactive T cells and induce T-cell tolerance by
deletion or anergy, as these APCs are relatively immature.
37
It
is recognized that TLRs can convert tolerogenic signals to
activating signals by promoting APC maturation.
4,5,37
Thus,
TLR4-mediated activation of resident uveal APCs may initiate
breakdown in ocular immune privilege and the development
of uveitis by various hypothesized autoimmune mechanisms
such as that of molecular mimicry.
2
In addition to the activa-
tion of TLR4 by LPS, its exogenous agonist, it has been more
recently discovered that these receptors may also be stimu-
lated by various endogenous agonists that are released at sites
of inflammation and tissue injury, such as heat shock proteins
and fibrinogen.
5,38
This may be a mechanism for the perpetu-
ation of uveitis and the progression to chronic inflammation as
the products of tissue inflammation further stimulate, via TLR4,
the immune cells that have infiltrated the eye.
Our finding that the epithelial and endothelial cells of the
normal human iris and ciliary body do not express TLR4 and
MD-2 proteins, in contrast to the corresponding cell types of
other tissues such as the skin,
14,17
may reflect the immunolog-
ically privileged nature of the eye. The strategic localization of
the LPS receptor complex within the uvea may be designed to
respond only to LPS of invasive organisms that have breached
the blood–ocular barrier or the iris-ciliary epithelium, and thus
minimizing the possibility of inappropriate and potentially
sight-threatening ocular inflammatory responses to LPS.
TLR4/MD-2 protein was not detected at the normal ocular
surface and this may again reflect the unique immune privi-
leged state of the eye, particularly with respect to the cornea.
Unique patterns of TLR expression appear to exist at different
host–environment tissue interfaces. For example, TLR4 is
strongly expressed by epithelial cells in the skin,
17
whereas in
the normal gastrointestinal tract, intestinal epithelial cells ex-
press minimal levels of TLR4, presumably to prevent inappro-
priate cellular activation in response to the constant exposure
to LPS of commensal organisms in the gut lumen.
16
We report
herein that the normal human conjunctiva and cornea do not
constitutively express the TLR4/MD-2 protein complex, and
this may be important in the maintenance of clarity of the
visual axis and prevention of inappropriate ocular surface in-
flammation in response to nonpathogenic LPS. Song et al.
19
have previously demonstrated in vitro, cell surface expression
of TLR4/MD-2 on cultured human corneal epithelial cells and
have shown that this expression was found to increase after
LPS treatment. The particular environment of cell culture me-
diums, including the presence of serum and/or LPS may mod-
ulate the in vitro expression of TLR4/MD-2. It is therefore
possible that upregulation of these LPS receptor complexes
from their low or undetectable constitutive levels may occur in
response to the appropriate stimuli in various ocular surface
inflammatory conditions, although these possibilities were not
examined in this study.
In summary, the results of the present study demonstrate
for the first time, the in vivo expression of TLR4 and its
associated LPS-receptor complex in the normal human eye and
their implications for ocular immunity in health and disease
have been discussed. The preferential expression of these
receptors on APCs within the uvea suggests a novel mechanism
for the initiating factors and immunopathogenesis of uveitis,
particularly HLA-B27–associated AAU, that have a particular
predilection for affecting this middle vascular layer of the eye.
Further studies, including functional studies, are indicated to
investigate further the role of these receptors in the context of
normal and pathologic ocular immunity.
Acknowledgments
The authors thank Paul Halasz (Confocal Microscopy Unit, University
of New South Wales) for expert technical assistance.
References
1. Chang JH, Wakefield D. Uveitis: a global perspective. Ocul Immu-
nol Inflamm. 2002;10:263–279.
2. Wakefield D, Montanaro A, McCluskey P. Acute anterior uveitis
and HLA-B27. Surv Ophthalmol. 1991;36:223–232.
3. Rosenbaum JT, McDevitt HO, Guss RB, Egbert PR. Endotoxin-
induced uveitis in rats as a model for human disease. Nature.
1980;286:611–613.
4. Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. A human homo-
logue of the Drosophila Toll protein signals activation of adaptive
immunity. Nature. 1997;388:394–397.
5. Takeda K, Kaisho T, Akira S. Toll-like receptors. Annu Rev Immu-
nol. 2003;21:335–376.
6. Schnare M, Barton GM, Holt AC, Takeda K, Akira S, Medzhitov R.
Toll-like receptors control activation of adaptive immune re-
sponses. Nat Immunol. 2001;2:947–950.
7. Poltorak A, He X, Smirnova I, et al. Defective LPS signaling in
C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science.
1998;282:2085–2088.
8. Frey EA, Miller DS, Jahr TG, et al. Soluble CD14 participates in the
response of cells to lipopolysaccharide. J Exp Med. 1992;176:
1665–1671.
9. Nagai Y, Akashi S, Nagafuku M, et al. Essential role of MD-2 in LPS
responsiveness and TLR4 distribution. Nat Immunol. 2002;3:667–
672.
10. Shimazu R, Akashi S, Ogata H, et al. MD-2, a molecule that confers
lipopolysaccharide responsiveness on Toll-like receptor 4. J Exp
Med. 1999;189:1777–1782.
IOVS, June 2004, Vol. 45, No. 6 Expression of TLR4 and Its LPS Receptor Complex by APCs in the Uvea 1877
11. Zarember KA, Godowski PJ. Tissue expression of human toll-like
receptors and differential regulation of toll-like receptor mRNAs in
leukocytes in response to microbes, their products, and cytokines.
J Immunol. 2002;168:554–561.
12. Muzio M, Bosisio D, Polentarutti N, et al. Differential expression
and regulation of toll-like receptors (TLR) in human leukocytes:
selective expression of TLR3 in dendritic cells. J Immunol. 2000;
164:5998–6004.
13. Mita Y, Dobashi K, Endou K, et al. Toll-like receptor 4 surface
expression on human monocytes and B cells is modulated by IL-2
and IL-4. Immunol Lett. 2002;81:71–75.
14. Faure E, Equils O, Sieling PA, et al. Bacterial lipopolysaccharide
activates NF-kappaB through toll-like receptor 4 (TLR-4) in cul-
tured human dermal endothelial cells: differential expression of
TLR-4 and TLR-2 in endothelial cells. J Biol Chem. 2000;275:
11058–11063.
15. Ochoa MT, Legaspi AJ, Hatziris Z, Godowski PJ, Modlin RL, Sieling
PA. Distribution of Toll-like receptor 1 and Toll-like receptor 2 in
human lymphoid tissue. Immunology. 2003;108:10–15.
16. Cario E, Podolsky DK. Differential alteration in intestinal epithelial
cell expression of toll-like receptor 3 (TLR3) and TLR4 in inflam-
matory bowel disease. Infect Immun. 2000;68:7010–7017.
17. Curry JL, Qin JZ, Bonish B, et al. Innate immune-related receptors
in normal and psoriatic skin. Arch Pathol Lab Med. 2003;127:178–
186.
18. Seibl R, Birchler T, Loeliger S, et al. Expression and regulation of
Toll-like receptor 2 in rheumatoid arthritis synovium. Am J Pathol.
2003;162:1221–1227.
19. Song PI, Abraham TA, Park Y, et al. The expression of functional
LPS receptor proteins CD14 and toll-like receptor 4 in human
corneal cells. Invest Ophthalmol Vis Sci. 2001;42:2867–2877.
20. Smith PD, Smythies LE, Mosteller-Barnum M, et al. Intestinal mac-
rophages lack CD14 and CD89 and consequently are down-regu-
lated for LPS- and IgA-mediated activities. J Immunol. 2001;167:
2651–2656.
21. Suzuki M, Hisamatsu T, Podolsky DK. Gamma interferon augments
the intracellular pathway for lipopolysaccharide (LPS) recognition
in human intestinal epithelial cells through coordinated up-regu-
lation of LPS uptake and expression of the intracellular Toll-like
receptor 4-MD-2 complex. Infect Immun. 2003;71:3503–3511.
22. Tso JY, Sun XH, Kao TH, Reece KS, Wu R. Isolation and charac-
terization of rat and human glyceraldehyde-3-phosphate dehydro-
genase cDNAs: genomic complexity and molecular evolution of
the gene. Nucleic Acids Res. 1985;13:2485–2502.
23. McMenamin PG, Crewe J, Morrison S, Holt PG. Immunomorpho-
logic studies of macrophages and MHC class II-positive dendritic
cells in the iris and ciliary body of the rat, mouse, and human eye.
Invest Ophthalmol Vis Sci. 1994;35:3234–3250.
24. Flugel C, Kinne RW, Streilein JW, Lu¨tjen-Drecoll E. Distinctive
distribution of HLA class II presenting and bone marrow derived
cells in the anterior segment of human eyes. Curr Eye Res. 1992;
11:1173–1183.
25. Forrester JV, McMenamin PG, Holthouse I, Lumsden L, Liversidge
J. Localization and characterization of major histocompatibility
complex class II-positive cells in the posterior segment of the eye:
implications for induction of autoimmune uveoretinitis. Invest
Ophthalmol Vis Sci. 1994;35:64–77.
26. Steinman RM. The dendritic cell system and its role in immunoge-
nicity. Annu Rev Immunol. 1991;9:271–296.
27. Kadowaki N, Ho S, Antonenko S, et al. Subsets of human dendritic
cell precursors express different toll-like receptors and respond to
different microbial antigens. J Exp Med. 2001;194:863–869.
28. Zhang FX, Kirschning CJ, Mancinelli R, et al. Bacterial lipopolysac-
charide activates nuclear factor-kappaB through interleukin-1 sig-
naling mediators in cultured human dermal endothelial cells and
mononuclear phagocytes. J Biol Chem. 1999;274:7611–7614.
29. Dabbagh K, Lewis DB. Toll-like receptors and T-helper-1/T-helper-2
responses. Curr Opin Infect Dis. 2003;16:199–204.
30. Smiley ST, King JA, Hancock WW. Fibrinogen stimulates macro-
phage chemokine secretion through toll-like receptor 4. J Immu-
nol. 2001;167:2887–2894.
31. Means TK, Hayashi F, Smith KD, Aderem A, Luster AD. The toll-like
receptor 5 stimulus bacterial flagellin induces maturation and che-
mokine production in human dendritic cells. J Immunol. 2003;
170:5165–5175.
32. Qureshi ST, Lariviere L, Leveque G, et al. Endotoxin-tolerant mice
have mutations in toll-like receptor 4 (Tlr4). J Exp Med. 1999;189:
615–625.
33. Springer TA. Traffic signals for lymphocyte recirculation and leu-
kocyte emigration: the multistep paradigm. Cell. 1994;76:301–314.
34. Saari KM, Laitinen O, Leirisalo M, Saari R. Ocular inflammation
associated with Yersinia infection. Am J Ophthalmol. 1980;89:84–
95.
35. Saari KM, Vilppula A, Lassus A, Leirisalo M, Saari R. Ocular inflam-
mation in Reiter’s disease after Salmonella enteritis. Am J Ophthal-
mol. 1980;90:63–68.
36. Wakefield D, Stahlberg TH, Toivanen A, Granfors K, Tennant C.
Serologic evidence of Yersinia infection in patients with anterior
uveitis. Arch Ophthalmol. 1990;108:219–221.
37. Ohashi PS, DeFranco AL. Making and breaking tolerance. Curr
Opin Immunol. 2002;14:744–759.
38. Flohe SB, Bruggemann J, Lendemans S, et al. Human heat shock
protein 60 induces maturation of dendritic cells versus a Th1-
promoting phenotype. J Immunol. 2003;170:2340–2348.
1878 Chang et al. IOVS, June 2004, Vol. 45, No. 6
