The Clinical Time-Course of Experimental Autoimmune
Uveoretinitis Using Topical Endoscopic Fundal Imaging
with Histologic and Cellular Infiltrate Correlation
David A. Copland,1Michael S. Wertheim,1,2W. John Armitage,1Lindsay B. Nicholson,1,3
Ben J. E. Raveney,*,3,4and Andrew D. Dick*,1,2,3
PURPOSE. EAU is an established preclinical model for assess-
ment of immunotherapeutic efficacy toward translation of ther-
apy for posterior uveitis. Reliable screening of clinical features
that correlate with underlying retinal changes and damage has
not been possible to date. This study was undertaken to de-
scribe, validate, and correlate topical endoscopic fundus imag-
ing (TEFI) with histologic features of murine experimental
autoimmune uveoretinitis (EAU), with the intent of generating
a rapid noninvasive panretinal assessment of ocular inflamma-
METHODS. EAU was induced in B10.RIII mice by immunization
with the peptide RBP-3161-180. The clinical disease course (days
0–63) was monitored and documented using TEFI. Disease
severity and pathology were confirmed at various time points
by histologic assessment. The composition of the cell infiltrate
was also examined and enumerated by flow cytometry.
RESULTS. TEFI demonstrated the hallmark features of EAU, par-
alleling many of the clinical features of human uveitis, and
closely aligned with underlying histologic changes, the severity
of which correlated significantly with the number of infiltrating
retinal leukocytes. Leukocytic infiltration occurred before man-
ifestation of clinical disease and clinically fulminant disease, as
well as cell infiltrate, resolved faster than histologic scores.
During the resolution phase, neither the clinical appearance
nor number of infiltrating retinal leukocytes returned to pre-
CONCLUSIONS. In EAU, there is a strong correlation between
histologic severity and the number of infiltrating leukocytes
into the retina. TEFI enhances the monitoring of clinical dis-
ease in a rapid and noninvasive fashion. Full assessment of
preclinical immunotherapeutic efficacy requires the use of all
three parameters: TEFI, histologic assessment, and flow cyto-
metric analysis of retinal infiltrate. (Invest Ophthalmol Vis Sci.
human uveitides and, as a result, is a successful preclinical
model for translation of immunotherapies.1,2Furthermore, the
model serves to dissect immunopathogenic mechanisms relat-
ing to immune-mediated tissue damage, which in turn highlight
avenues for future immunotherapies.3–6Murine EAU is gener-
ated after systemic activation of ocular-specific CD4?T cells
that are frequently located within or around photoreceptor
segments.7–11In particular, EAU can be induced via adminis-
tration of dominant peptides from retinoid binding protein
(RBP)-3 (previously called interphotoreceptor retinoid binding
protein [IRBP]) in an appropriate adjuvant.12Disease occurs
subsequent to T cell infiltration into the target organ that
recruits and activates macrophages into the eye, generating
structural damage via mechanisms including secretion of nitric
To quantify the extent and severity of disease, which is
clearly essential for validating the efficacy of preclinical thera-
pies, two approaches have been used to date: nonvalidated
clinical scoring and semiquantitative histologic scoring and
grading. Clinical EAU assessment involves in vivo examination
of the eye using indirect slit lamp biomicroscopy and scoring
the features of retinal, anterior chamber, and pupil appearance
during disease.11In this regard, fundus photography has until
now been limited by technical difficulties and the poor reso-
lution of existing techniques for disease assessment.14Immu-
nohistochemical assessment of retinal sections, with grading
according to the degree of inflammatory infiltrate and struc-
tural damage, has been used for assessment of disease severi-
ty,15but this technique has inherent limitations, such as the
fact that only a small proportion of the whole retina can be
examined. Therefore, a new easy-to-use imaging system that
facilitates rapid, reproducible, live clinical assessment of the
whole fundus, closely correlating with histologic changes is
required as an approach to monitor progression of retinal
disease in experimental models, including EAU.
Topical endoscopic fundus imaging (TEFI) is a recently
described compact system that allows high-resolution in vivo
color photography of the retina in rodents and was developed
in normal eyes of mice.16TEFI is based on the use of an
endoscope with parallel, lateral, crescent-shaped illumination
connected to a digital camera. This technique facilitates rapid
assessment and capture of high-quality images of the whole
fundus, including the peripheral retina and ciliary body, with-
out distress to the mouse or the requirement for general anes-
The objectives of this study were to validate a platform by
using the TEFI system for assessment of the clinical disease
time course of RBP-3161-180–induced EAU in B10.RIII mice, and
correlate clinical features to both matched published histologic
xperimental autoimmune uveoretinitis (EAU) is a suitable
correlate to the spectrum of clinicopathologic features of
From the1Academic Unit of Ophthalmology, Department of Clin-
ical Sciences at South Bristol, and the3Department of Cellular and
Molecular Medicine, University of Bristol, Bristol, United Kingdom;
2Bristol Eye Hospital, Lower Maudlin Street, Bristol, United Kingdom;
science NCNP, Tokyo, Japan.
Supported by a National Eye Research Centre grant.
Submitted for publication May 28, 2008; revised July 17 and
August 13, 2008; accepted October 24, 2008.
Disclosure: D.A. Copland, None; M.S. Wertheim, None; W.J.
Armitage, None; L.B. Nicholson, None; B.J.E. Raveney, None; A.D.
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.
*Each of the following is a corresponding author: Andrew D. Dick,
Academic Unit of Ophthalmology, Department of Clinical Sciences at
South Bristol, University of Bristol, Bristol BS8 1TD, UK;
Ben J. E. Raveney, Department of Immunology, National Institute of
Neuroscience NCNP, Tokyo, Japan; email@example.com.
4Department of Immunology, National Institute of Neuro-
Investigative Ophthalmology & Visual Science, December 2008, Vol. 49, No. 12
Copyright © Association for Research in Vision and Ophthalmology
severity scores and the extent of inflammatory retinal cell
infiltrate determined by flow cytometric analysis.
MATERIALS AND METHODS
B10.RIII mice were originally obtained from Harlan UK, Ltd. (Oxford,
UK) and a breeding colony established within the Animal Services Unit
at Bristol University (Bristol, UK). All mice were housed in specific
pathogen-free conditions with continuously available food and water.
Female mice, immunized for disease induction, were aged between 6
and 8 weeks. Treatment of animals conformed to UK legislation and to
the ARVO Statement for the Use of Animals in Ophthalmic and Vision
The peptide RBP-3161-180(SGIPYIISYLHPGNTILHVD) was synthesized
by Sigma-Genosys Ltd. (Poole, UK). Peptide purity was ?95% as deter-
mined by HPLC.
EAU Induction and Scoring
B10.RIII mice were immunized SC in one flank with 50 ?g/mouse
RBP-3161-180peptide in PBS (2% DMSO), in Complete Freud’s Adjuvant
(CFA; 1 mg/mL; 1:1 vol/vol) supplemented with 1.5 mg/mL Mycobac-
terium tuberculosis complete H37 Ra (BD Biosciences, Oxford, UK),
and 1.5 ?g Bordetella pertussis toxin (Sigma-Aldrich, Poole, UK) was
given intraperitoneally. At various time points after immunization, the
eyes were enucleated, oriented in optimal cutting temperature (OCT)
compound (R. Lamb Ltd., East Sussex, UK), and carefully snap frozen.
Serial 12-?m sections were cut and stored at ?80°, before thawing at
room temperature and fixation in acetone for 10 minutes. Sections
were stained with rat anti-mouse CD45 monoclonal antibody (Serotec,
Oxford, UK), counterstained with hematoxylin (ThermoShandon,
Pittsburgh, PA), and then scored for inflammatory infiltrate (presence
of CD45-positive cells) and structural disease (disruption of morphol-
ogy). Cellular infiltrate was scored within the ciliary body, vitreous,
vessels, rod outer segments, and choroid, whereas structural disease
was scored within the rod outer segments, neuronal layers, and retinal
morphology. Both scores were added together to calculate a final
disease total (Table 1).
Topical Endoscope Fundus Imaging
Using a method adapted from Paques et al.,16we connected an endo-
scope with a 5-cm-long tele-otoscope with a 3-mm outer diameter
(1218AA; Karl Storz, Tuttlingen, Germany) a digital camera (D80 with
a 10-million-pixel charge-coupled device [CCD] image sensor and AF
85/F1.8 D objective (Nikkor; all from Nikon, Tokyo, Japan), with a
additional ?4.00-D magnifying lens. The settings of the camera were as
follows: large and superfine image, manual focus; operating mode S
(shutter speed priority), shutter set at 1/100 s, and white balance set at
fluorescent. A xenon lamp (201315-20; Karl Storz) connected through
a flexible optic fiber to the endoscope was used as the light source.
The pupils of the mice were dilated with topical tropicamide 1%
and phenylephrine 2.5% (Minims; Chauvin Pharmaceuticals, Romford,
UK), then topical oxybuprocaine 0.4% (Minims) and eye gel (Novartis
Pharmaceuticals, Camberley, UK), were applied for corneal anesthesia
and endoscope contact, respectively. For imaging, the camera with
endoscope was attached to a bench-clamp, and the mouse was slowly
moved toward the tip of the endoscope. Once contact with the gel
covering the cornea was obtained, focus and illumination were ad-
justed by using the camera, and the fundus was examined and the
image was captured. Images were transferred to computer for process-
ing (Photoshop; Adobe Systems, Mountain View, CA). Images were
cropped to a size of 6 ? 4.85 in. The blue curves tool was used to
render the image a natural color. We did not use RAW imaging, as no
image manipulation (other than color adjustment) was required. We
found that the superfine setting was more than adequate for our
purposes and each image was around 3 MB in size. After numerous
trials, we found that using the fluorescent light white balance setting
generated the best image detail after further blue curve adjustment in
the image analysis software.
Isolation of Retinal Infiltrating Cells
Infiltrating retinal cells were isolated by using a previously described
method.17In brief, the eyes were enucleated and the retinas (including
the ciliary body) of each animal were dissected microscopically and
washed in wash media (complete RPMI supplemented with 10% [vol/
vol] FCS and 1 mM HEPES; all from Invitrogen, Paisley, UK). Retinas
were then cut into small pieces and digested in 1 mL wash medium,
supplemented with 0.5 mg/mL collagenase D (Roche, Welwyn Garden
City, UK) and 750 U/mL DNase I (Sigma-Aldrich) for 20 minutes at
37°C. An additional 0.5 mg/mL collagenase D and DNase 750 U/mL
was added before incubation for a further 10 minutes at 37°C. Cell
suspensions were forced through a 40-?m cell strainer (BD-Falcon,
Cambridge, UK), with a syringe plunger, and the cell suspensions were
stained for flow cytometric analysis.
TABLE 1. Summary of EAU Disease Scoring
Ciliary body Cell infiltrate ? 5 cells
Cells ? 5
Cells ? 100
?10% vessels involved
Cells in or around wall
Mild perivascular cuffing
Granulomas ? 5
Vasculitis (mural or
Rod outer segments
Rod outer segments Cell infiltrate
Folds ? 10%
Folds ? 50%
* Maximum possible total score: infiltration, 30; structural, 12.
IOVS, December 2008, Vol. 49, No. 12
Clinical Time Course of EAU5459
The cell suspensions were incubated with 24G2 cell supernatant for
5 minutes at 4°C. For cell counting, retinal cell suspensions were
stained with PE-Cy5-conjugated anti-mouse CD4 monoclonal antibody
(mAb), APC-Cy7-conjugated anti-mouse CD11b mAb, and PE-Cy7-
conjugated anti-mouse CD45 mAb (all BD Pharmingen, Oxford, UK),
at 4°C for 20 minutes. Cell suspensions were acquired with a flow
cytometer (LSR-II; BD Cytometry Systems, Oxford, UK). Analysis
was then performed (FlowJo software; TreeStar, San Carlos, CA).
The number of cells counted was calculated by reference to a
Partial correlation was performed (SPSS Inc, ver. 14; SPSS, Chicago, IL)
and used to explore the relationship between the number of CD45?
cells (after square root transformation) and histologic score, while
controlling for time (days) after immunization.
TEFI Imaging of the Retina during EAU
TEFI is a system that allows high resolution in vivo color
photography of the retina in rodents and was developed in the
normal eyes of C57BL/6 and BALB/c mice.16We found that
high-quality panretinal images, with clear visualization of the
peripheral retina could also be obtained in the B10.RIII mouse
strain, when the pupil was suitably dilated. The optimal pupil
dilation was achieved by using a drop combination of phenyl-
ephrine 2.5% and tropicamide 1%, instilled at least 5 minutes
before initiating TEFI.
Using this adapted TEFI method, we sought to monitor the
clinical changes in the retina that occur during disease progres-
sion in mice immunized for EAU. We used the highly suscep-
tible B10.RIII mouse strain, in which the immunizing regimen
generates reliable disease induction and consistent moderate
disease severity in our hands, thus ensuring that any clinical
changes to the retina would be clearly evident.
Mice were immunized SC with 50 ?g RBP-3161-180emulsi-
fied in CFA, and pertussis toxin was coadministered intraperi-
toneally. In the initial experiment, 10 mice were immunized
and the disease progression was monitored from days 0 to 63.
The TEFI method enabled us to capture a variety of clinical
images (Fig. 1). Clinical features of EAU were clearly observed,
including vasculitis and optic nerve swelling (Fig. 1A); exuda-
tive retinal detachment (Fig. 1B); retinal folds, observed as
retinal flecks (Fig. 1E); and choroidal lesions, analogous to
chorioretinal lesions in uveitis in humans (Fig. 1F). The periph-
ery of the retina could also be visualized, demonstrating the
anatomy of the ciliary body and the drainage angle (Figs. 1C,
1D). Figure 1D demonstrates how we were able to increase the
magnification of views of the ciliary body and drainage angle
by virtue of imaging through the mouse lens.
The time-course of EAU in the right eye of a representative
individual mouse demonstrates the significant changes that
occurred during disease progression (Fig. 2). The retina and
vasculature remained normal in appearance, with no clinical
evidence of disease from day 0 to 10 postimmunization (pi).
However, by day 13 pi, we recorded swelling of the optic
nerve that increased in severity to include the central retinal
vasculature (analogous to retinal vasculitis in humans). With
time (days 14–18 pi), vitritis (cellular infiltrate within the
vitreous gel) made the fundus increasingly indistinct (vitreous
haze). Despite this vitreous haze, large exudative retinal de-
tachments were documented from day 17 pi onward and
resolved after day 21 pi. From day 15 pi onward, white retinal
flecks were observed uniformly throughout the retina, which
are presumed to be clinical evidence of small retinal folds
(described later). Accompanying resolution of the exudative
detachments was clinical resolution of features of retinal vessel
involvement (vasculitis) and optic nerve swelling. Over the
time period examined, the retina did not regain its normal
appearance, as clinical images from day 28 pi onward demon-
strated persistence of retinal flecks throughout the resolution
phase and up to and including day 63 pi. Examination of the
contralateral eye in selected mice verified that there was sim-
ilar clinical appearance between eyes at all time points (data
Comparison of TEFI, Histologic Features, and
Composite of Cellular Infiltrate during EAU
Given the relative ease and reproducibility of TEFI when used
to monitor disease in immunized mice, we wanted to deter-
mine whether clinical features would correlate with histologic
changes and the kinetics of cellular infiltrate.
We immunized 40 B10.RIII mice, and on days 12, 13, 14, 15,
18, 19, 21, 28, 35, 42, and 63 pi, TEFI images of the right eye
were obtained from four mice at each time point (three mice
on days 42 and 63) before death. The right eyes were enucle-
ated and sections prepared for immunohistochemical staining
with anti-CD45 antibody. Three sections per retina per time
point were scored for inflammatory infiltrate and structural
damage, as described previously (Table 1).
Figure 3 shows our findings as a representative comparison
of TEFI and histology images taken from the same eye. Obser-
vations from days 0 to 12 pi demonstrated a normal retinal
appearance by TEFI, which was confirmed histologically in
sections that displayed normal morphology and no inflamma-
tory infiltrate. By days 13 and 14 pi, clinical changes that
included a raised appearance of the optic nerve were observed
in 75% of the mice, although at this stage there was no clinical
or histologic evidence of altered retinal morphology. The in-
crease in histologic disease score was secondary to infiltrate
that arose at the ciliary body and scleral–choroidal interface in
that area (Fig. 3, inset).
From days 15 to 19 pi, exudative retinal detachments and
signs of cellular infiltrate (white lesions) and perivascular
sheathing (vasculitis) were evident in all (100%) animals exam-
ined at these times. Where severe vitritis in the mice prevented
clear visualization of the retina, histologic assessment con-
firmed the characteristics of clinical disease. This result corre-
lates with extensive retinal disruption and folding, vasculitis,
and perivascular infiltrate associated with increased CD45?
infiltrate observed by histology. However, by day 21 pi, retinal
detachments were reduced clinically, whereas perivascular in-
filtrate persisted and by histology, both infiltrate and retinal
morphologic disruption remained clearly evident. The overall
clinical appearance improved from days 28 to 63 pi in all
(100%) animals, with reduced inflammation of the optic nerve
and retina (as observed as reduced optic nerve head swelling,
reduced perivascular infiltrate and reduced creamy chorioreti-
nal deep presumed infiltrative lesion), although during the
resolution phase (postpeak disease), white, worm-like retinal
flecks persisted. Histologic assessment of the eyes, corrobo-
rated such findings and demonstrated markedly reduced infil-
trate, but persistent small retinal folds, likely to represent flecks
observed by TEFI, were still apparent. Such folds are similar to
those previously documented in other models.18
Analysis of the inflammatory infiltrate and structural scores
throughout the time course by histology exhibited the classic
monophasic disease course of EAU in B10.RIII mice (Fig. 4A).
From days 12 to 14 pi, increased levels of CD45?cell infiltrate
were detected, while little or no structural damage was ob-
served within the retina. Disease progressed from day 15 pi
5460Copland et al.
IOVS, December 2008, Vol. 49, No. 12
in a representative cohort of B10.RIII mice immunized for EAU using RBP-3161-180in CFA. Images show raised and swollen optic nerve, with typical
perivascular cuffing and caliber changes to vessels (arrowhead, A), and an inferior exudative retinal detachment (B), at day 20 pi. Images including
the ciliary body demonstrate peripheral chorioretinal inflammation and inflammatory vascular changes of the marginal vein (C, D). Scattered flecks
which correlate to histologic features of retinal folds, are typically observed after day 15 pi (E). Multiple choroidal lesions (arrowhead) associated
with inflammatory vascular changes (perivascular cuffing) and swollen optic nerve persisted at day 28 pi (F).
Clinical observations of EAU in B10.RIII mice using topical endoscopic fundus imaging. Shown are examples of clinical disease observed
right eyes were dilated and photographed.
Clinical features during the study time course in a B10.RIII mouse immunized for EAU. At each time point from days 10 to 63 pi, the
IOVS, December 2008, Vol. 49, No. 12
Clinical Time Course of EAU5461
onward, with a peak of disease at day 19 pi, as reflected by high
scores for inflammatory infiltrate and associated structural dam-
age. The period from day 21 pi onward is often termed the
resolution phase, and although disease scores are reduced,
morphologic changes (structural damage) and CD45?cellular
infiltration persists through to day 63 pi. Although this infers a
level of regulation and repair, neither the number of CD45?
cells nor retinal morphology returned to normal predisease
Inflammatory Cell Infiltrate
Considering the clinical pictures obtained using TEFI and the
close relationship we observed to the underlying histologic
changes, we wished to determine whether the clinical features
also related to the kinetics and levels of inflammatory cell
infiltrate present in the eye during the course of EAU. The
isolation and analysis of retinal infiltrate using flow cytometric
methods has been used to determine the normal immune
status of the eye,19to quantify and monitor the kinetics of
inflammatory infiltrate in the retina, and also evaluate the
effects and efficacy of potential new immunomodulatory
agents in EAU.17Therefore, at the same time points described
earlier, the left eye was also enucleated (as we noted synchro-
nous bilateralism of clinical features during EAU), dissected,
and single cell suspensions were prepared from the retina and
the ciliary body. Cells were then stained with fluorochrome-
conjugated monoclonal antibodies against CD11b (macro-
phages), CD4 (T cells), and CD45 (leukocytes) surface markers
and analyzed by flow cytometry to enable quantification of
total cell number and phenotype (Fig. 4B).
We observed elevated levels of leukocytes compared with
normal retina from day 12 pi forward. The number of CD11b?
and CD4?cells increased steadily from day 13 pi onward, with
the main expansion of both cell types occurring after day 15 pi,
and at a maximum on day 18 pi. Furthermore, during this time
CD4?cells were present at lower levels, with a predominance
of CD11b?cells. Of note was the fact that an increased num-
ber of cells was detected before any evidence of clinical (TEFI
images) or histologic disease. From day 19 to 21 pi, the level of
inflammatory cell infiltrate reduced and CD45?cell numbers
remained throughout (to day 63 pi) at levels equivalent to
those on day 13 pi. The number of cells never returned to
normal predisease levels, indicating that CD45?infiltrate per-
sists and may contribute to the clinical changes observed
during the resolution phase. The ratio of CD11b?to CD4?
cells during this phase is also reduced with both cell types
present in equal amounts at the later time points.
Correlation between CD45 Infiltrate
Figure 5A shows the change in the number of CD45?cells
compared with the change in histologic score with time after
immunization. The data suggest an association between these
variables with both staying low at days 12 to 13 pi, increasing
between days 15 and 20 pi and then reducing again thereafter.
Although the number of cells fell to levels similar to those at
days 12 to 13 pi, the histologic score remained somewhat
elevated, albeit lower than the peak scores. Partial correlation
was used to explore the relationship between histologic score
and numbers of CD45?cells while controlling for time (days)
after immunization. This confirmed that there was a strong,
positive partial correlation (r ? 0.78, df ? 32, P ? 0.001), with
high histologic scores being associated with high cell counts
(Fig. 5B). The zero order correlation (r ? 0.73) suggests that
time has little influence on the strength of the association
between these two variables.
RBP-3161-180in CFA. At the days after immunization indicated, clinical images were obtained before the mice were killed. The right eyes were
enucleated, sectioned, and stained for CD45?infiltrate. A 12-?m retinal section from the right eye with total disease score is shown next to the
corresponding clinical image from the right eye at each day, representative of each group (n ? 4). Histology images from days 13 and 14 include
insets showing the ciliary body–ciliary marginal zone and surrounding CD45?perivascular infiltrate.
Comparison of clinical and histologic images during the EAU time course. Forty female B10.RIII mice were immunized for EAU using
5462 Copland et al.
IOVS, December 2008, Vol. 49, No. 12
TEFI has been an effective technique that permits high resolu-
tion in vivo imaging of the clinical changes that occur during
EAU disease progression in mice. Previous techniques were
limited, and TEFI now offers an improved, rapid, no-anesthesia
approach to generating detailed panfundal images in mice. It
also facilitates the reduction, replacement, and refinement
goals now favored by ethics committees in animal experimen-
tation. With TEFI, when comparisons and correlation are made
with histologic scoring and flow cytometric assessment of
retinal infiltrate, several important unrecognized features of
this model become apparent. First, significant retinal cell infil-
trate is observed whereas clinically the retina appears largely
unaffected; second, the clinical resolution of peak disease is
much faster than resolution of histologic disease, and finally, in
the resolution phase of EAU, neither the clinical appearance or
the extent and composition of CD45?cells within the retina
return to predisease levels (up to 63 days pi). Although we
noted a correlation between histologic scoring and flow cyto-
metric assessment of infiltrating leukocyte numbers, cell
counts resolved faster than severity of histologic changes; and,
with the use of TEFI, our results emphasize that significant
changes may occur that are not always clinically manifest.
The objectives of this study were to validate a platform that
uses the TEFI system for assessment of clinical time course of
RBP-3161-180–induced EAU in B10.RIII mice and correlate clin-
icopathologic features to both histologic severity and the ex-
tent of inflammatory cell infiltration of the retina. The clinical
images obtained using the TEFI approach overall closely asso-
ciate to the pathologic features of disease observed by histol-
ogy. Histologic assessment demonstrated that the disease in
this model of EAU followed the classic monophasic profile,
with peak disease severity observed at day 19. Analysis of the
dynamics and kinetics of retinal infiltrate demonstrated that an
expansion of CD45?cells, including CD11b?and CD4?pop-
ulations, was present before this peak. Although the number of
infiltrating cells was reduced during the later resolution phase,
histologic disease scores and infiltrate levels never returned to
normal; this finding has not been appreciated in this model of
autoimmune destruction of the retina. Statistical correlation
analyses demonstrated a positive association between the num-
Changes in the number of CD45?cells and histologic score with time
after immunization. (B) Scatterplot of total histologic score against the
square root of total number of CD45?cells. Partial correlation (r ?
0.78, df ? 32, P ? 0.001) shows a strong, positive association between
these variables while controlling for time after immunization.
Correlation of infiltrate and histologic disease score. (A)
trate during the EAU time course. (A) Right eyes were enucleated at
the post immunization day indicated. Eyes were snap frozen before
cryosectioning and staining for immunohistochemical analysis of
CD45?infiltrate. Average disease score ? SD of inflammatory infiltrate
and structural disease is shown (n ? 4/time point). Histology demon-
strated that the EAU was a monophasic disease that peaked at day 19
pi, but did not fully resolve or return to normal levels. (B) Left eyes
were simultaneously enucleated at each time point, and the retina and
ciliary body excised and digested with collagenase. The number of
immune cells per eye was measured by flow cytometry. The total
number of immune cells is detailed as follows: CD45?CD11b?
(CD11b), CD45?CD4?(CD4), and CD45?CD11b?CD4?(CD45) (n ?
4/time point). Elevated levels of retinal infiltrate were observed from
day 12 pi, with an expansion of CD45?cells including macrophages
and T cells seen from day 15 pi, peaking at day 18 pi. From days 19 to
21 pi, the level of infiltrate was reduced but the cells persisted through-
out the resolution phase.
Comparison of histologic scores and retinal cellular infil-
IOVS, December 2008, Vol. 49, No. 12
Clinical Time Course of EAU 5463
ber of infiltrating CD45?cells and the resulting histologic
By using TEFI, it is now possible to record and monitor the
dramatic clinical changes that occur in B10.RIII mice during
the normal disease course of RBP-3161-180–induced EAU. From
the time of immunization until day 12 pi, the retina and vas-
culature appeared normal and healthy, followed by a series of
clinical changes from days 13 to 18, including raised optic
nerve, perivascular infiltration developing that manifests the
perivascular infiltrate and vitritis normally appreciated as hall-
marks of this model. The development of large exudative
retinal detachments from day 17, which resolve, along with the
other clinical features of perivascular infiltrate and vitritis, can
also be observed. The emergence of retinal flecks, uniformly
distributed across the retina is demonstrated from day 15 pi.
The retinal flecks correspond to the retinal folds we observed
histologically. Clinically and histologically, retinal integrity
never normalizes to the predisease state during the EAU time
course. We also noted with TEFI that clinical features of EAU
were constant between contralateral eyes.
EAU serves as a model for the spectrum of human posterior
uveitis including sympathetic ophthalmia and Vogt-Koyanagi-
Harada syndrome (VKH; particularly in relation to exudative
retinal detachments), multifocal choroiditis, ocular sarcoidosis,
and other forms of idiopathic disease.1,20For example, the
clinical features seen in this study correlate well with clinical
features of VKH, in which resolving exudative retinal detach-
ments are observed. After resolution of acute VKH, the classic
clinical features of sunset-glow retina with its appreciated de-
generative features are seen, again correlating with our TEFI
images from day 28 onward.
Flow cytometric analysis of cells isolated from the retina
demonstrated that the elevated levels of inflammatory infiltrate
observed from day 12 onward during the time course of EAU,
consisted of macrophages, T cells, and other CD45?leuko-
cytes. Infiltration of cells at this time has been examined by
histology, which demonstrates the perivascular accumulation
of CD45?cells in the retina,21although this static analysis
cannot fully assess the dynamics of infiltration. The infiltration
kinetics revealed that the main expansion of cells occurred
after day 15 pi, culminating in a peak at day 18, and during this
time, the proportion of CD4?cells present was reduced com-
pared to the number of CD11b?cells. After the infiltrative
peak, total CD45?cells were greatly reduced over the remain-
der of the time course, but never returned to predisease levels.
During this resolution phase, both the main CD11b?and
CD4?populations were present at equal levels. Persistence of
elevated levels of infiltrate in the eye would suggest that reso-
lution and recovery do not equate to normal leukocyte counts,
and may further suggest that certain regulatory mechanisms
are maintained in the eye after inflammation.22,23Similarly
assessment of immunotherapeutic agents, given our current
findings of temporal disparity between clinical appearance and
cell infiltrate in the earlier stages of disease, and together with
previous observations of maintained cellular infiltrate despite
reduced histologic scores,24shows that it is plausible that
changes in constituents and number of infiltrating cells are not
appreciated in the face of normal clinical phenotype and may
conversely not always indicate preservation of function.
Nevertheless, TEFI is a method that allows confirmation of
disease status and severity. It will aid in the design of experi-
mental protocols according to clinical observations. TEFI will
also greatly assist with current approaches to preclinical test-
ing of experimental eye models, as it allows direct observation
and assessment of therapeutic efficacy of new potential ocular
therapy. It will also provide a rapid assessment to determine
potential adverse effects incurred due to invasive procedures
including intravitreous or subretinal injections.
Although, unlike experimental autoimmune encephalomy-
elitis (EAE),25in which we are unable to ascribe directly func-
tional deficit (paralysis) to histologic change or with the more
technically demanding imaging of cellular infiltrate in the
CNS,26we are now able in EAU to directly correlate and assess
clinical changes with histologic and flow cytometric analysis of
cellular infiltrate. In both models, we now understand that
significant cellular infiltrate occurs before the onset of clinical
signs in the fundus of EAU and clinically in EAE.
Furthermore, the current published clinical grading of dis-
ease17,27,28in both B10.RIII and/or C57BL/6 mouse models
have been developed without incorporating evolution of clin-
ical phenotype and comparison of such temporal characteris-
tics with respect to the extent and timing of leukocytic infil-
tration (e.g., by flow cytometry analysis) and contemporaneous
histopathologic appearances throughout the course of EAU.
Although these scores may still be used, and indeed clinical
features we show can mirror underlying histologic change,
ascribing scoring of clinical severity or damage in light of this
new data necessitates further investigation of EAU progression
with larger groups of mice and in other strains (C57BL/6) to
generate and then validate such a proposed grading system.
The most recent report29in C57BL/6 model of TEFI grading of
clinical changes in chronic EAU supports our findings in this
model of EAU. The advantage of adapting TEFI is therefore
highlighted in both models and serves to assess more repro-
ducibly the signs of inflammatory disease and correlate with
underlying histologic and flow cytometric data.
Arguably, to fully assess preclinical immunotherapeutic ef-
ficacy requires the use of all three parameters: TEFI, histologic
assessment, and flow cytometric analysis of retinal infiltrate.
Combined TEFI and histologic methods enable the observation
of clinical features and severity of disease, but information
regarding the dynamics, phenotype, function and quantity of
cellular traffic through the eye is only provided through de-
tailed analysis of cell populations present in the eye at various
stages of disease progression.
1. Forrester JV, Liversidge J, Dua HS, Towler H, McMenamin PG:
Comparison of clinical and experimental uveitis. Curr Eye Res.
2. Dick AD. Experimental approaches to specific immunotherapies in
autoimmune disease: future treatment of endogenous posterior
uveitis? Br J Ophthalmol. 1995;79:81–88.
3. Caspi RR. Regulation, counter-regulation, and immunotherapy of
autoimmune responses to immunologically privileged retinal anti-
gens. Immunol Res. 2003;27:149–160.
4. Dick AD, Forrester JV, Liversidge J, Cope AP. The role of tumour
necrosis factor (TNF-alpha) in experimental autoimmune uveoreti-
nitis (EAU). Prog Retin Eye Res. 2004;23:617–637.
5. Copland DA, Calder CJ, Raveney BJ, et al. Monoclonal antibody-
mediated CD200 receptor signaling suppresses macrophage acti-
vation and tissue damage in experimental autoimmune uveoretini-
tis. Am J Pathol. 2007;171:580–588.
6. de Smet MD, Chan CC. Regulation of ocular inflammation: what
experimental and human studies have taught us. Prog Retin Eye
7. Wacker WB. Retinal autoimmunity: two decades of research. Jpn J
8. Wacker WB. Proctor Lecture. Experimental allergic uveitis: inves-
tigations of retinal autoimmunity and the immunopathologic re-
sponses evoked. Invest Ophthalmol Vis Sci. 1991;32:3119–3128.
9. Atalla L, Linker-Israeli M, Steinman L, Rao NA. Inhibition of auto-
immune uveitis by anti-CD4 antibody. Invest Ophthalmol Vis Sci.
10. Caspi RR. Immunogenetic aspects of clinical and experimental
uveitis. Reg Immunol. 1992;4:321–330.
5464Copland et al.
IOVS, December 2008, Vol. 49, No. 12
11. Thurau SR, Chan CC, Nussenblatt RB, Caspi RR. Oral tolerance in Download full-text
a murine model of relapsing experimental autoimmune uveoreti-
nitis (EAU): induction of protective tolerance in primed animals.
Clin Exp Immunol. 1997;109:370–376.
12. Caspi RR, Chan CC, Wiggert B, Chader GJ. The mouse as a model
of experimental autoimmune uveoretinitis (EAU). Curr Eye Res.
13. Hoey S, Grabowski PS, Ralston SH, Forrester JV, Liversidge J. Nitric
oxide accelerates the onset and increases the severity of experi-
mental autoimmune uveoretinitis through an IFN-gamma-depen-
dent mechanism. J Immunol. 1997;159:5132–5142.
14. Hawes NL, Smith RS, Chang B, Davisson M, Heckenlively JR, John
SW. Mouse fundus photography and angiography: a catalogue of
normal and mutant phenotypes. Mol Vis. 1999;5:22.
15. Dick AD, Cheng YF, Liversidge J, Forrester JV. Immunomodulation
of experimental autoimmune uveoretinitis: a model of tolerance
induction with retinal antigens. Eye. 1994;8:52–59.
16. Paques M, Guyomard JL, Simonutti M, et al. Panretinal, high-
resolution color photography of the mouse fundus, Invest Oph-
thalmol Vis Sci. 2007;48:2769–2774.
17. Raveney BJ, Richards CM, Aknin ML, et al. The B subunit of
Escherichia coli heat-labile enterotoxin inhibits Th1 but not Th17
cell responses in established autoimmune uveoretinitis. Invest
Ophthalmol Vis Sci. 2008;49:4008–4017.
18. Akhmedov NB, Piriev NI, Chang B, et al. A deletion in a photore-
ceptor-specific nuclear receptor mRNA causes retinal degeneration
in the rd7 mouse. Proc Natl Acad Sci U S A. 2000;97:5551–5556.
19. Dick AD, Ford AL, Forrester JV, Sedgwick JD. Flow cytometric
identification of a minority population of MHC class II positive
cells in the normal rat retina distinct from CD45lowCD11b/
c?CD4low parenchymal microglia. Br J Ophthalmol. 1995;79:
20. Singh VK, Biswas S, Anand R, Agarwal SS. Experimental autoim-
mune uveitis as animal model for human posterior uveitis. Indian
J Med Res. 1998;107:53–67.
21. Jiang HR, Lumsden L, Forrester JV. Macrophages and dendritic
cells in IRBP-induced experimental autoimmune uveoretinitis in
B10RIII mice. Invest Ophthalmol Vis Sci. 1999;40:3177–3185.
22. Robertson MJ, Erwig LP, Liversidge J, Forrester JV, Rees AJ, Dick
AD. Retinal microenvironment controls resident and infiltrating
macrophage function during uveoretinitis. Invest Ophthalmol Vis
23. Kerr EC, Raveney BJ, Copland DA, Dick AD, Nicholson LB. Analysis
of retinal cellular infitrate in experimental autoimmune uveoreti-
nitis reveals multiple regulatory cell populations. J Autoimmun. In
24. Dick AD, McMenamin PG, Korner H, et al. Inhibition of tumor
necrosis factor activity minimizes target organ damage in experi-
mental autoimmune uveoretinitis despite quantitatively normal
activated T cell traffic to the retina. Eur J Immunol. 1996;26:
25. Wekerle H, Kojima K, Lannes-Vieira J, Lassmann H, Linington C.
Animal models. Ann Neurol. 1994;36(suppl):S47–S53.
26. Kawakami N, Nagerl UV, Odoardi F, Bonhoeffer T, Wekerle H,
Flugel A. Live imaging of effector cell trafficking and autoantigen
recognition within the unfolding autoimmune encephalomyelitis
lesion. J Exp Med. 2005;201:1805–1814.
27. Taylor AW, Yee DG, Nishida T, Namba K. Neuropeptide regulation
of immunity; the immunosuppressive activity of alpha-melanocyte-
stimulating hormone (alpha-MSH). Ann N Y Acad Sci. 2000;917:
28. Shao H, Liao T, Ke Y, Shi H, Kaplan HJ, Sun D. Severe chronic
experimental autoimmune uveitis (EAU) of the C57BL/6 mouse
induced by adoptive transfer of IRBP1–20-specific T cells. Exp Eye
29. Xu H, Koch P, Chen M, Lau A, Reid DM, Forrester JV. A clinical
grading system for retinal inflammation in the chronic model of
experimental autoimmune uveoretinitis using digital fundus im-
ages. Exp Eye Res. 2008;87(4):319–326.
IOVS, December 2008, Vol. 49, No. 12
Clinical Time Course of EAU5465