Neuroscience Letters 495 (2011) 115–120
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/neulet
Characterization of epithelial V-like antigen in human choroid plexus epithelial
cells: Potential role in CNS immune surveillance
Elise Wojcika, Lisette M. Carrithersa, Michael D. Carrithersa,b,∗
aDepartment of Neurology and Program in Cellular and Molecular Pathology, University of Wisconsin School of Medicine and Public Health, Madison, WI 53706, USA
bNeurology Service, William Middleton VA Hospital, Madison, WI 53705, USA
a r t i c l ei n f o
Received 11 February 2011
Received in revised form 16 March 2011
Accepted 17 March 2011
Blood-cerebrospinal fluid barrier
CD4 T lymphocyte
a b s t r a c t
Prior work demonstrated that immune surveillance of the brain occurs primarily through the blood-
cerebrospinal (CSF) fluid barrier rather than the blood-brain barrier endothelium. Recently, we identified
epithelial V-like antigen (EVA), an immunoglobulin-like adhesion molecule, as a regulator of blood-CSF
plexus epithelial cells and analyzed its role in CD4 T lymphocyte adhesion. In human choroid plexus
epithelial cells and a subset of CD4 T lymphocytes, EVA is expressed at high levels. Epithelial adhesion
of T lymphocytes is inhibited by a blocking monoclonal antibody that recognizes EVA. T cell adhesion
elicits calcium flux in choroid plexus epithelial cells that also can be blocked by an EVA-specific antibody.
cytoskeletal epithelial morphology. These results demonstrate that EVA is expressed in human choroid
plexus epithelial cells and CD4 T lymphocytes and regulates CD4+ T lymphocyte adhesion to human
choroid plexus epithelial cells in vitro. These data suggest a novel mechanism to regulate CNS immune
© 2011 Elsevier Ireland Ltd. All rights reserved.
The neuroprotective barriers include the blood brain barrier (BBB),
which is formed by tight junctions between brain endothelial cells,
and the blood-cerebrospinal fluid (CSF) barrier, which is regu-
lated by epithelial cells within the choroid plexus. T lymphocyte
trafficking across these barriers occurs in health in addition to
inflammatory disease states . Our prior work in mouse in vivo
models of immune cell trafficking demonstrated that immune
the blood-CSF barrier  and is partly mediated by the adhesion
molecule P selectin expressed on the choroid plexus epithelium
. Human memory T lymphocytes likely utilize the same path-
way to enter the cerebrospinal fluid . This pathway protects
the individual from developing infections of the central nervous
system (CNS) but also may lead to the initiation of new inflam-
matory lesions in patients with multiple sclerosis as suggested by
work in animal models [8,14].
Recently, our laboratory demonstrated that reduced lym-
phocyte immune surveillance in lymphocyte-deficient mice is
associated with increased blood-CSF barrier permeability. Based
on that observation, we analyzed CSF-barrier function and global
gene expression in immune-competent and lymphocyte-deficient
∗Corresponding author at: 1300 University Avenue, Room 2679, Madison, WI
53706, USA. Tel.: +1 608 265 8596; fax: +1 608 265 3170.
E-mail address: email@example.com (M.D. Carrithers).
for T lymphocyte CNS entry . These data revealed differential
expression of epithelial V-like antigen (EVA), a member of the
immunoglobulin superfamily of proteins , by choroid plexus
epithelial cells. Decreased expression of EVA in the choroid plexus
of lymphocyte-deficient mice is associated with increased per-
meability of the blood-CSF barrier, reduced cadherin expression
at epithelial cell contacts, and a less ordered morphology of the
Based on those results, we hypothesized that EVA regulates
permeability of the choroid plexus. Additional work from another
may mediate T lymphocyte adhesion to mouse thymic epithelial
cells . Here our goals were to characterize EVA expression in
human choroid plexus epithelial cells and analyze its role in CD4 T
lymphocyte adhesion. Our results suggest that EVA is expressed on
both human choroid plexus epithelial cells and a unique subset of
the two cell types.
Human choroid plexus epithelial cells were obtained from Sci-
enCell and grown according to the supplier’s recommendations in
Epithelial Cell Media (ECM) supplemented with 2% FBS, epidermal
cells at ScienCell and used from cell passages 1–4 in our labora-
0304-3940/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.
E. Wojcik et al. / Neuroscience Letters 495 (2011) 115–120
tory. Human CD4 T lymphocytes were obtained from Lonza. These
cells, which express the adhesion molecules VLA-4 and LFA-1,
were obtained from healthy, adult donors and purified by negative
selection. They were expanded in vitro for 3–5 days in the pres-
ence of ionomycin (2?g/ml) and phorbol myristate acetate (PMA)
C57BL6/J mice (6–8 weeks of age) were obtained from Jackson
Laboratory. All mice were maintained in sterile and pathogen-free
conditions. All animal studies were reviewed and approved by the
University of Wisconsin-Madison Institutional Animal Care and
Review Committee (IACUC).
Mice were anesthetized with ketamine/xylazine, and intrac-
ardiac perfusion was performed with ice cold PBS followed
by 4% paraformaldehyde. Isolated brains were post-fixed in
paraformaldehyde, sucrose infused, embedded in OCT compound,
and sectioned (7?m sections) on a cryostat.
Following column purification of RNA and reverse transcription
using Superscript III (Gibco), qPCR was performed using TaqMan
primers (FAM-labeled; Applied Biosystems). Samples were run
on a SmartCycler (Cepheid) . The following Taq-Man primers
were used: Hs00170684 m1 (EVA; mpzl2), Hs00174914 m1
(Transthyretin; ttr), Hs01023894 m1 (E-cadherin; cdh1), and
Hs00170684 m1 (GAPDH). Samples were normalized to GAPDH Ct
values for each experiment.
Cells and tissue sections were stained according to standard
protocols . The following primary antibodies were used: mouse
anti-pan-cytokeratin (PKC-26, 25?g/ml) and cadherin (CH-19,
81?g/ml) (Abcam), rabbit anti-EVA (rabbit polyclonal, 1:100 dilu-
tion) (Proteintech), mouse anti-EVA (27–136, 2.5?g/ml) (Abnova),
mouse anti-CD4 (RIV-6, 1?g/ml) (Pierce-Thermo Bioscience), rab-
bit anti-CD3 (rabbit monoclonal, 04–460, 5?g/ml) (Millipore) and
mouse anti-?-tubulin (236–10501, 1?g/ml) (Invitrogen). Sections
were washed and incubated with fluorescent goat or donkey sec-
were counterstained with DAPI or no additional stain.
Activated (ionomycin+PMA as described above) CD4 T lym-
phocytes were stained with mouse anti-EVA primary antibody
followed by Alexa 488 labeled donkey anti-mouse secondary anti-
body (Invitrogen). For intracellular cytokine staining, fixed and
permeabilized cells were stained for EVA, IL-17 (APC-conjugated),
and by FlowJo software.
Stained cells were imaged on a Zeiss Axiovert 200 fluorescence
microscope using a 63× (Zeiss Plan Apochromat, 1.4 oil) objec-
tive and an Axiocam-MRm CCD camera. Images were acquired
using Zeiss Axiovision 4.6.3 software. Some of these images were
acquired in multiple Z-planes and then deconvoluted using the
Monolayer cultures of human choroid plexus epithelial cells
were imaged on a Zeiss Axiovert 200 fluorescent microscope using
the 63× objective and Axiocam CCD camera as described above
and a DIC filter in an environmentally controlled chamber. Acti-
vated CD4 T lymphocytes (1×106cells/ml) were added at t=0, and
time lapse images were acquired at 1min intervals for 60min in
multiple Z-planes to image both diffusing T cells and the epithe-
lial monolayer. Adherence was defined as motion arrest of the T
cells on the epithelial monolayer for multiple time-lapse images
within the microscopic field (100?m×140?m), and the num-
ber of adherent cells per time point was averaged across all time
points for each 60min experiment (average number of adherent
of a blocking monoclonal antibody (2.5?g/ml final concentration,
concentration; mouse IgG2a kappa; BD Bioscience), and in the
mRNA expression in primary human choroid plexus epithelial cells.
GeneCt (threshold cycle)±SEM
Real time PCR analysis was performed. Ct (threshold cycle) values were normalized
using GAPDH as a control for each experiment.
absence of antibody treatment. Data was analyzed using Kaleida-
graph software using a Student’s t test.
Choroid plexus epithelial cells were labeled with the cal-
cium indicator dye Fluo-4 (1?M) (Invitrogen) for 30min and
then washed. T cells were added as described above. Live
cell imaging was performed in an environmental chamber,
and time lapse images were acquired with a QuantEM:512SC
EMCCD camera using the same microscopy set-up and soft-
ware as for fixed samples. Time-lapse sequences were analyzed
using the Time Series Analyzer plugin (Balaji) for ImageJ soft-
ware (http://rsb.info.nih.gov/ij/plugins/time-series.html). For each
experiment the maximal response of the Fluo-4 labeled choroid
plexus epithelial cells to an adherent lymphocyte was calculated,
and the pooled data from multiple experiments was analyzed by
Kaleidagraph software using the Student’s t test.
Human choroid plexus epithelial cells were grown in vitro
and characterized for expression of epithelial and choroid
plexus markers by immunofluorescence and quantitative PCR.
Immunofluorescence staining demonstrated expression of the
epithelial marker, cytokeratin, in these cells and the presence of
extensive cellular processes within monolayer cultures (Fig. 1A).
Quantitative PCR analysis demonstrated high expression of the
choroid plexus epithelial marker transthyretin (Ct=25.24+0.06,
n=3; Ct=threshold cycle normalized to GAPDH expression with
GAPDH expression normalized to Ct=17.00) with lower levels of
E-cadherin mRNA detected (Ct=35.99+0.59, n=6) (Table 1).
To demonstrate expression of EVA in human choroid plexus
epithelial cell function, we first analyzed the expression of EVA
mRNA and its protein product. Expression of EVA mRNA was mea-
sured using quantitative PCR (Table 1). EVA mRNA was expressed
at high levels (Ct=19.25±0.18, n=3; Ct=threshold cycle normal-
ized to GAPDH expression with GAPDH expression normalized to
Ct=17.00). Immunofluorescence staining was performed to char-
acterize subcellular expression of EVA protein in choroid plexus
epithelial monolayers. Positive staining for EVA was observed in a
locations (Fig. 1B).
cells, mouse brain sections were stained with anti-EVA antibody
and analyzed by fluorescent microscopy (Fig. 1C). In most epithelial
cells, EVA staining demonstrated expression throughout the cell
on both the basal and apical (ventricular) surfaces. Staining was
not observed in brain endothelial cells. These results suggest that
the surface expression of EVA is not polarized in vivo during health
and that adhesion events could occur on either side of the barrier.
Because EVA can form intercellular homophilic adhesions
between thymic epithelial cell and developing T lymphocytes ,
we analyzed expression of EVA in human CD4 T lymphocytes. CD4
T lymphocytes were activated in vitro and then analyzed for EVA
expression by flow cytometry. Approximately 30% of cells were
EVA positive, and, within that population, approximately 6% of the
total expressed very high levels of EVA (Fig. 2A). Consistent with
their potential role in homing to choroid plexus epithelium, these
EVA-high expressing cells demonstrated increased expression of
the inflammatory cytokine interleukin 17 (IL-17) as compared to
E. Wojcik et al. / Neuroscience Letters 495 (2011) 115–120
Fig. 1. Characterization of human choroid plexus epithelial cells and EVA expression. (A) Staining for the epithelial marker cytokeratin in monolayer cultures demonstrated
expression throughout the cell and included filamentous cellular processes. Scale bar, 10?m. (B) EVA staining (green; mouse anti-EVA) was present in a golgi, peri-nuclear
pattern and in a filamentous pattern. Scale bar, 10?m. (C) In brain tissue from C57/BL6/J mice, EVA staining (rabbit anti-EVA) was observed in the choroid plexus epithelium
in a uniform pattern on both apical (ventricular) and basal surfaces without evidence of polarized expression. Scale bar, 20?m.
Fig. 2. EVA expression in primary human CD4 T lymphocytes. Primary human CD4 T lymphocytes were analyzed by flow cytometry. (A) Approximately 30% of cells were
EVA+ and approximately 6% of cells demonstrated high expression. (B and C) Intracellular cytokine staining was performed, and separate gates for EVA-negative cells and
EVA-high expressing cells were analyzed. As compared to EVA-negative cells, the high expressing EVA+ cells demonstrated increased expression of IL-17 and very high
expression of IL-22.
E. Wojcik et al. / Neuroscience Letters 495 (2011) 115–120
Fig. 3. EVA-dependent adhesion and calcium signaling between choroid plexus epithelial cells and CD4 T lymphocytes is inhibited by an anti-EVA antibody. (A) Quantitative
analysis of anti-EVA treatment revealed a reduction in CD4 T cell adhesion to epithelial monolayers as compared to the untreated and antibody isotype control conditions.
This analysis demonstrated 13.1±1.2 adherent cells/14,000?m2/min for the isotype antibody control condition (n=10), 10.5±0.5 adherent cells/14,000?m2/min for the
untreated condition (n=12), and 4.7±0.4 adherent cells/14,000?m2/min for the anti-EVA condition (n=9; P<0.01 as compared to each control condition). (B and C) For
detection of EVA-dependent calcium signaling, choroid plexus epithelial cells were labeled with the calcium indicator dye Fluo-4 and then imaged in the presence of CD4 T
lymphocytes for 30min. Only cells that demonstrated contact with a CD4 cell were analyzed. Epithelial cells treated with isotype control antibody demonstrated a rapid and
persistent increase in cytosolic calcium that was significantly decreased in the presence of anti-EVA antibody (mouse monoclonal, 2.5?g/ml). The peak calcium response in
isotype control cells was 151.8±23.8 ?Relative Fluorescent Units (n=5) and in the anti-EVA condition, 3.6±1.6 ?Relative Fluorescent Units (n=5; P<0.01).
EVA-negative cells (Fig. 2B) and very high expression of IL-22, a
regulator of epithelial cells (Fig. 2C) . These results suggest that
EVA could mediate adhesion to the choroid plexus epithelium with
a subset of CD4 T lymphocytes through homophilic adhesions.
To quantify EVA-dependent adhesion, time-lapse images of
co-cultures of human choroid plexus epithelial cells and CD4 T
lymphocytes were analyzed. Within the choroid plexus, immune
cells can exit the circulation through fenestrated capillaries
and then interact with choroid plexus epithelial cells. To mea-
sure these interactions in vitro, monolayer cultures of human
choroid plexus epithelial cells were imaged by DIC microscopy
and then activated CD4 T lymphocytes were added in the pres-
ence or absence of antibodies that block EVA interactions and
those that do not (isotype control). Treatment with a blocking
monoclonal antibody inhibited adherence as defined as motion
arrest on the epithelial monolayer. Quantitative analysis of adhe-
sion revealed 13.1±1.2 adherent cells/14,000?m2/min for the
isotype antibody control condition (n=10), 10.5±0.5 adherent
cells/14,000?m2/min for the untreated condition (n=12), and
(n=9; P<0.01 as compared to each control condition) (Fig. 3A).
To demonstrate that EVA and lymphocyte adhesion could
elicit cellular signaling in choroid plexus epithelial cells, calcium
signaling was analyzed in epithelial cells in response to lympho-
cyte interactions. Prior studies demonstrated that cross-linking of
epithelial cells were labeled with the calcium indicator dye Fluo-4
and then challenged with CD4 T lymphocytes as described for the
adherence assays. In cells treated with an isotype control antibody,
cytosolic calcium in choroid plexus epithelial cells. Treatment with
an anti-EVA antibody significantly reduced this response (Fig. 3B).
?Relative Fluorescent Units (n=5) and in the anti-EVA condition,
3.6±1.6 ?Relative Fluorescent Units (n=5; P<0.01) (Fig. 3C).
immunofluorescent staining of co-cultures of human choroid
plexus epithelial cells and CD4 T lymphocytes was performed.
Activated CD4 T lymphocytes were added to monolayer cultures
of human choroid plexus epithelial cells for 30min, washed to
remove non-adherent cells, fixed, and then analyzed for EVA,
E. Wojcik et al. / Neuroscience Letters 495 (2011) 115–120
Fig. 4. EVA localization in a co-culture of human choroid plexus epithelial cells and CD4 T lymphocytes. Human CD4 T lymphocytes were added to epithelial monolayers
for 30min, and the monolayer was washed to remove non-adherent lymphocytes prior to fixation. A. Two EVA+, CD4+ cells are shown (red) that have contacted a choroid
plexus epithelial cell (larger cell in green). EVA staining is most intense at points of contact between the two cell types. Scale bar, 10?m. B. EVA-positive cellular processes
(green) from the epithelial cells contact and entangle an EVA+, CD3 T lymphocyte (red). Scale bar, 10?m.
CD4 or CD3 expression. As shown in Fig. 4, all adherent CD4 T
lymphocytes that were analyzed demonstrated a high level of
expression of EVA. EVA was expressed at particularly high levels
at regions of cell–cell contact between CD4 T lymphocytes and
choroid plexus epithelial cells (Fig. 4A and Fig. S1).
At some contacts between T lymphocytes and choroid plexus
epithelial cells, EVA-positive cytoskeletal processes extend from
the epithelial cells and entangle the lymphocyte (Figs. 4B and
Fig. S2). This observation is consistent with the proposed asso-
ciation of EVA with the cytoskeleton  and suggests that T
lymphocytes can become entangled within the epithelium.
We here demonstrate that EVA is expressed in human choroid
with prior mouse studies in this laboratory  and others [7,10]
that showed a role for EVA in regulation of choroid plexus barrier
permeability and thymic lymphocyte development. To our knowl-
edge, the present study is the first to demonstrate a direct role for
epithelium. These results suggest a novel, EVA-dependent mecha-
nism to regulate CD4 T lymphocyte immune surveillance of the
EVA expression by a subset of activated human CD4 T lympho-
cytes was confirmed by flow cytometry (Fig. 2). Those cells that
demonstrated the highest levels of EVA (Fig. 2) express IL-17 and
very high levels of IL-22, a cytokine that regulates epithelial barrier
function . Localized secretion of IL-22 may provide a mecha-
nism for this T cell subset to repair and remodel the choroid plexus
Treatment with a monoclonal antibody to EVA blocked CD4
T lymphocyte adhesion to choroid plexus epithelium by greater
than 50%. These results suggest that, although lymphocyte adhe-
sion to choroid plexus epithelium can occur in the absence of EVA,
its expression increases the number of adherent cells and may
be necessary for the development of tighter and more prolonger
interactions between the two cell types. These EVA-dependent
plexus and facilitate localized immune cell interactions and host
response. Since EVA can mediate homophilic interactions, similar
to CD31 (PECAM-1) and cadherins, adhesion to the choroid plexus
epithelial cell may occur through EVA expressed by the lympho-
cyte. Alternatively, based on work from another laboratory, EVA
may bind to CD3 expressed on the T lymphocyte surface. Addi-
tional work is required to determine the relative importance of
lymphocyte CD3 and EVA in binding to epithelial EVA.
From a clinical standpoint, these findings may be relevant to
the development of CNS opportunistic infections in some patients
treated with immune modulatory agents that reduce lympho-
cyte trafficking or other immune suppressive medications. These
patients include those treated with chemotherapeutic agents for
neoplasms and autoimmune disorders as well as those with mul-
tiple sclerosis who require treatment with medications such as
natalizumab or fingolimod [1,6,12]. We speculate that targeting
of EVA pharmacologically could repair barrier function in individ-
uals treated with immune therapy and potentially decrease the
incidence of adverse events. Additional work is required to deter-
mine the effects of EVA activation and inhibition during health and
inflammatory disease states.
Affairs or the United States Government.
The authors declare no competing interests.