| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Gastroenterology Division, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA 02115;
Medical Research Council, Molecular Haematology Unit, John Radcliffe Hospital, Oxford, United Kingdom;
Institut Fuer Medizinische Immunologie, Charité-Humboldt-Universität zu Berlin, Berlin, Germany;
Imperial Cancer Research Fund-Medical Oncology Unit, St. Bartholomew’s Hospital Medical College, London, United Kingdom;
¶
Immunopathology Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114; and
||
Combined Program in Pediatric Gastroenterology and Nutrition, Massachusetts General Hospital and Harvard Medical School, and
#
Department of Adult Oncology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA 02115
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
+ TCR-+ but
CD28- T lymphocytes that may function in intestinal
epithelial cell immunosurveillance. In an attempt to define novel cell
surface molecules involved in iIEL function, we raised several mAbs
against activated iIELs derived from the small intestine that
recognized an Ag on activated, but not resting, iIELs. Using expression
cloning and binding studies with Fc fusion proteins and transfectants,
the cognate Ag of these mAbs was identified as the N domain of biliary
glycoprotein (CD66a), a carcinoembryonic Ag-related molecule that
contains an immune receptor tyrosine-based inhibitory motif.
Functionally, these mAbs inhibited the anti-CD3-directed and
lymphokine-activated killer activity of the P815 cell line by iIELs
derived from the human small intestine. These studies indicate that the
expression of biliary glycoprotein on activated human iIELs and,
potentially, other mucosal T lymphocytes is involved in the
down-regulation of cytolytic function. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A significant fraction of human iIELs of both the small and large
intestine are CD8 + and
CD45RO+ T cells that express a limited array of
and, to a lesser extent, 
TCRs (2, 3, 4, 5). These
phenotypic properties indicate that most iIELs are memory cells that
localize to the basolateral surface of IECs for the recognition of a
limited number of Ags in the context of MHC class I or class I-like
molecules on the IEC. However, the majority of iIELs in mouse and human
are CD28-, suggesting that other costimulatory
molecules for TCR/CD3 complex-mediated activation may be important in
providing necessary secondary signals for iIEL activation
(6, 7, 8, 9, 10, 11). Candidate costimulatory molecules for human iIELs
include CD2 (10), CD101 (8), BY-55
(9), and the
E7 integrin
(11), which are expressed by the majority of iIELs.
It has also become increasingly evident that in addition to activating
costimulatory molecules, T cells can express a variety of molecules
that deliver an inhibitory signal such that either the initial
activation of the T cell is prevented or the activated state is
down-regulated. The former type includes the killer inhibitory
receptors (KIR), which are expressed on a subset of T cells and bind
specific types of MHC class I molecules on the target cells
(12). The latter type includes CTLA-4 (CD152) which, when
expressed after T cell activation, binds either CD80 (B7.1) or CD86
(B7.2) on APC (13, 14). KIRs characteristically contain
Ig-like extracellular domains and one or more immune receptor
tyrosine-based inhibitory motifs (ITIM) in their cytoplasmic tails with
a consensus sequence of I/L/VxYxxL/V (15). In the case of
CTLA-4, the cytoplasmic tail contains the sequence GxYxxM, which is
highly similar to, but not identical with, the ITIM of KIRs.
ITIM-containing receptors function in the recruitment of either the Src
homology domain-containing protein tyrosine phosphatases, SHP-1 and
SHP-2, or the SH2 domain-containing inositol polyphosphate
5-phosphatase, SHIP (16). These phosphatases function in
the dephosphorylation of signaling molecules recruited by immune
receptor tyrosine-based activation motif (ITAM)-bearing receptors such
as those contained in the CD3-, -, -
, and -
chains that
associate with the TCR. As such, ITIM-bearing receptors on T cells are
predicted to down-regulate activation events elicited by ITAM-bearing
receptors if both are ligated in close proximity to one another.
Importantly, neither KIRs and CTLA-4 nor CD80/CD86 have been observed
on human iIELs or IECs of the intestine, respectively.
In this report we provide evidence in support of a novel role for biliary glycoprotein (BGP; CD66a), a member of the carcinoembryonic Ag family (CEA), as an inhibitory receptor for activated T cells contained within the human intestinal epithelium. These studies also suggest that, in a regional microenvironment that is predominantly CD28/CTLA-4-CD80/CD86 negative, other receptor-ligand interactions may provide the necessary down-regulatory signals to limit T cell activation and immunopathology.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 34B1, 26H7, and 5F4 mAbs were produced by immunizing BALB/c
mice with the activated human iIEL line, 191E, obtained from a subject
with celiac disease as previously described (8). Hybridoma
supernatants were screened by indirect immunoperoxidase staining of
frozen intestinal and tonsillar tissue sections to identify and
characterize the mAbs used in this report as previously described
(17). The isotypes of 34B1 (IgG1), 26H7 (IgG1), and 5F4
(IgG1) were determined by ELISA using murine isotype-specific mAb
(Boehringer Mannheim, Indianapolis, IN). OKT3 (IgG2a) is a mouse
anti-human CD3 mAb (provided by Dr. Robert Finberg, Dana-Farber
Cancer Institute, Boston, MA). TS 2/18 (provided by Dr. Lloyd
Klickstein, Brigham and Women’s Hospital) is an anti-CD2 mAb
(mouse IgG2a). OKT4 (mouse IgG2b) and OKT8 (mouse IgG2a) are mAbs
specific for human CD4 and CD8, respectively (obtained from American
Type Culture Collection, Manassas, VA). MA22 (CD66abce; clone YG-C94G7;
IgG1), MA26 (CD66ae; clone 4.3.17; IgG1), MA27 (CD66e; clone 26/5/1;
IgG2a), MA28 (CD66e; clone 26/3/13; IgG1), MA30 (CD66c; clone 9A6;
IgG1), MA41 (CD66b; clone BIRMA17c; IgG1), MA61 (CD66b; clone 80H3;
IgG1), MA76 (CD66ae; clone 12-140-4; IgG1), MA79 (CD66b; clone B13.9;
IgG1), MA81 (CD66b; clone G10F5; IgG1), MA83 (CD66e; clone b7.8.5;
IgG1), MA84 (CD66de; clone COL-1; IgG2a), MA86 (CD66acde; clone B6.2;
IgG1), and MA91 (CD66e; cloneT84.66; IgG1) are mouse mAbs that were
obtained from the Sixth Leukocyte Typing Workshop, Osaka, Japan. The
isotype-matched mouse IgG1 negative control mAb was purchased from
Cappel (West Chester, PA). mAbs were affinity purified with protein A
or Sepharose G columns by standard methods.
Cells and cell lines
Human iIELs were obtained, and cell lines EEI-10 (small intestine), EEI-5 (small intestine), KJ-3 (small intestine), and CLI (large intestine) were generated from donors as previously described (18) and maintained by stimulation with 1 µg/ml PHA-P (Murex, Dartford, U.K.) in RPMI 1640 containing 10% human serum (type AB; Sigma, St. Louis, MO), 5 U/ml rIL-4 (Genzyme, Cambridge, MA), 2 nM rIL-2 (a gift from Ajinomoto, Japan), and irradiated PBMC as feeders. iIEL cell lines EEI-10, EEI-5, and KJ-3 were >90% CD8+, whereas the CLI cell line was 40% CD8+, 30% CD4+, and 30% double negative. HT29 is a human IEC line obtained from the American Type Culture Collection. COS (monkey kidney fibroblast), CHO (Chinese hamster ovary), HeLa (human cervical epithelium), and HT29 cell lines were maintained in RPMI 1640 containing 10% heat-inactivated FCS (Life Technologies, Gaithersburg, MD), penicillin and streptomycin, nonessential amino acids, and 10 mM HEPES (complete medium) at 37°C in 5% CO2.
Immunohistology
Tissue samples, obtained under the auspices of human studies approval from the Massachusetts General Hospital and Brigham and Women’s Hospital, were mounted in OCT compound (Ames, Elkart, IN), frozen in liquid nitrogen or in a cryostat, and stored at -70°C. Frozen tissue sections 4 µm thick were fixed in acetone for 5 min, air-dried, and stained by an indirect immunoperoxidase method using avidin-biotin-peroxidase complex (Vector Laboratories, Burlingame, CA) and 3-amino-9-ethylcarbazole (Aldrich, Milwaukee, WI) as the chromogen as previously described (17).
Two-color immunohistochemical analysis was performed as previously
described (19). Four-micron-thick specimens were fixed in
cold acetone for 10 min, air-dried for 20 min, and incubated with
normal horse serum (Vector Laboratories) for 30 min. The specimens were
then incubated with the 5F4 mAb (10 µg/ml) for 1 h at room
temperature. For detection, 5 µg/ml biotinylated horse anti-mouse
Ig (Vector Laboratories) was used, followed by incubation with
peroxidase-labeled avidin (Dako, Carpinteria, CA). These specimens were
developed with a solution of 3-amino-9-ethylcarbazole (Aldrich). The
reaction was stopped by dipping the specimens in distilled water for 10
min and washing with PBS for 10 min. The specimens were incubated with
purified anti-CD3 mAb, Leu4 (10 µg/ml; Becton Dickinson, San
Jose, CA), for 1 h. For detection, biotinylated horse
anti-mouse Ig (Vector Laboratories) was used, followed by
incubation with alkaline phosphatase-labeled avidin reagent (Vector
Laboratories) for 30 min. After development using the alkaline
phosphatase substrate kit III (Vector Laboratories) for 15 min, the
specimens were postfixed with 2% paraformaldehyde and mounted with
Glycerogel (Dako). Each step was followed by three washes with PBS.
Incubation with 0.3% H2O2
in PBS was used to block endogenous peroxidase activity, and sequential
incubations with avidin and biotin (Vector Laboratories) were used to
block endogenous biotin.
Transfectants
The BGPx' molecule was constructed as follows. The N-terminal
domain and the transmembrane/cytoplasmic domains of human
BGPc were each amplified separately by PCR with the primer
pairs BGPAMP-S (CATCATCATCATAAGCTTATGGGGCACCTC)/NTM-AS
(GCCATTTTCTTGGGGCAGCTCCGGGTATAC) and
NTM-S/(GTATACCCGGAGCTGCCCCAAGAAAATGGC)/BGP
TRANS-CYT-AS(CTACTACTACTAAGACTATGAAGTTGGTTG), respectively, where
the NTM primers were hybrids of the 3' end of the N-terminal domain
and the 5' end of the transmembrane domain. Each PCR
consisted of 50 µl of 1 mM Tris (pH 8.3), 5 mM KCl, 0.01% gelatin,
0.09 mM MgCl2, 0.03 mM of each dNTP, 1 µM of
each primer, 1 U of Taq polymerase, and 1 µg of
cDNA. The PCR was conducted at 94°C for 10 min, followed by 25 cycles
of 94°C for 1 min, 55°C for 1 min, and 72°C for 2 min, plus a
final extension of 10 min at 72°C. After passing the PCR products
through S-300 columns, 5 µl of each PCR product was used in a second
PCR. After the PCR products had annealed, the BGPAMP-S and BGP
TRANS-CYT-AS primers were added to the reaction mix, and the PCR
reaction was conducted as described above. The resulting PCR product
was cloned into the pAMP 1 vector using the CloneAMP system as detailed
by the manufacturer (Life Technologies), transformed into DH5
competent bacteria and positive transformants selected by PCR. The
resulting BGPx' cDNA was extracted and sequenced by standard methods.
The BGPx' cDNA was digested with EcoRI and NotI
restriction enzymes and subcloned into the pcDNA1/Amp vector
(Invitrogen, San Diego, CA). The BGPx' cDNA in this vector and the
pSV2neo plasmid (Clontech, Palo Alto, CA) were linearized with
XhoI and BamHI, respectively, and electroporated
into CHO cells at a ratio of 15:1, which was selected in G418 and on
the FACS cell sorter to create a stable CHO-BGPx' cell line as
described previously (20). CHO cells stably transfected
with BGPx', neomycin, BGPc (21), and BGPa
(22) and HeLa cells stably transfected with CEA, CGM1,
NCA, and CGM6 have been previously described (20, 21, 22).
Flow cytometry
Flow cytometry was performed as previously described (2, 3). Staining was performed with 1 µg of the primary Ab followed by incubation with 1 µg of a goat anti-mouse FITC secondary Ab (Zymed, San Francisco, CA) with analysis on a FACSCalibur (Becton Dickinson, Sunnyvale, CA) flow cytometer.
COS cell expression cloning
A cDNA library was constructed in the pCDM8 vector using poly(A)+ RNA from resting and activated human peripheral blood T cells (PBT) in the vector pAEXF (23). For the first round of selection, COS cells were transfected via the DEAE-dextran procedure (24) with 0.2 µg of library DNA/100-mm dish. After 48 h, cells were harvested, incubated with the 34B1 mAb (1/500 dilution of ascites), washed, and panned on anti-IgG1-coated plates as previously described (23, 24, 25). Episomal DNA was purified from adherent cells, reintroduced into Escherichia coli, and transfected into COS cells by polyethylene glycol-mediated fusion of spheroplasts (24), and the panning with 34B1 mAb was repeated. Individual plasmid DNAs were transfected into COS cells via the DEAE-dextran procedure and analyzed after 72 h for cell surface expression by indirect immunofluorescence and flow cytometry.
Radiolabeling, immunoprecipitation, and electrophoresis
COS cells, 96 h after transient transfection, were removed nonenzymatically from plastic petri dishes, and iIELs were labeled with Na-[125I] by the lactoperoxidase-catalyzed method as previously described (26). Immunoprecipitations, digestion with N-glycanase, and SDS-PAGE were performed as previously described (26).
Production of soluble recombinant proteins and analysis of Ab binding
Details of the pIG vector (R&D Systems Europe, Abingdon, U.K.) containing the Fc genomic fragment of human IgG1 and incorporating the hinge (H), CH2, and CH3 domains and of the construction and purification of the CD66a-Fc soluble proteins containing the N, NA1B1, and NA1B1A2 extracellular domains, Muc-18-Fc (R&D Systems) and NCAM-Fc, have been described previously (20, 27, 28). Ab binding was quantified by ELISA with detection by alkaline phosphatase-conjugated goat anti-mouse Ig (Boehringer Mannheim, Indianapolis, IN) and paranitrophenyl phosphate (Sigma) as substrate as previously described (29).
Cytotoxicity assays
Cytotoxicity was evaluated as previously described (30). Briefly, the P815 mouse mastocytoma cell line was labeled with 100 µCi of 51Cr (New England Nuclear, Boston, MA) at 37°C for 30 min. The radiolabeled cells (2 x 103), in 100 µl of complete medium, were added to 100 µl of varying concentrations of effector T cells in 100 µl of complete medium in triplicate in a 96-well V-bottom plate. Before addition of target cells, the effector cells were incubated for 20 min at room temperature with medium, the OKT3 mAb (100 ng/ml of purified Ab), and/or varying concentrations of the 34B1 mAb, the 26H7 mAb, the 5F4 mAb as purified mAbs, or purified IgG1 Ab as a control (Sigma). Lymphokine-activated killer activity was assessed by examining cytotoxicity in the absence of OKT3 mAb. After 5 h, 100 µl of supernatant was removed for analysis in a gamma counter (LKB Wallac Clini Gamma 1272, Turku, Finland). Spontaneous and maximal release were measured by incubating target cells with medium or 1% Nonidet P-40, respectively. The percent cytotoxicity was calculated using the formula [(experimental release - spontaneous release) x 100/(maximal release - spontaneous release)].
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During the development of iIEL-specific mAbs, obtained by
immunizing mice with an iIEL T cell line from human small intestine
propagated in vitro, it was observed that several of the mAbs stained
IECs, as shown by immunohistology of normal human small and large
intestines. Staining of human intestinal tissue sections showed that
these three mAbs (34B1, 26H7, and 5F4) only stained IECs, not iIELs
(Fig. 1
, A–C). The in vivo
tissue staining with these Abs appeared to be on the cell surface, as
confirmed by flow cytometric analysis of a normal human IEC line, HT29
(data not shown). Because these three Abs did not stain iIELs in situ,
as determined by immunohistochemistry (Fig. 1, A–C), or
immediately after isolation as determined by flow cytometry (Fig. 1D), it was predicted that iIELs, activated during the
process of in vitro cultivation, expressed neoantigens that were
constitutively expressed by IECs. Indeed, after maintenance in vitro as
continuous cell lines with PHA-P activation in the presence of
allogeneic feeder cells, the majority of iIELs expressed the Ag
recognized by these three mAbs. Staining of an iIEL T cell line, EEI-5,
established from the small intestine, that was 90%
CD8+ and 10% CD4+
indicates that all iIELs expressed the Ag recognized by the three mAbs
after this type of in vitro activation (Fig. 1E). Similar
observations were made with an iIEL T cell line prepared from the large
intestine, CLI, which was 40% CD8+, 30%
CD4+, and 30% double negative
(CD4-CD8-) at the time of
staining consistent with the in vivo phenotype of iIELs in this tissue
site (31) (data not shown). The expression of this Ag was observed
within 7 days of in vitro activation of freshly isolated normal human
iIELs, indicating that the observations were not an artifact of in
vitro cultivation (Fig. 1F).
|
, numerous
CD3+ 5F4- cells, which
stained blue, consistent with T cells, were observed throughout the
lamina propria and epithelia (open arrow). CD3-
5F4+ cells, which stained brown, included cells
with the morphology of granulocytes (small arrow) and intestinal
epithelial cells (large arrow). Rare double-positive cells (arrowheads)
that were both blue and brown, consistent with 5F4-staining,
CD3+ cells, were observed in both the
lamina propria and epithelium. These results suggest that the
expression of this activation Ag is likely to be relevant to in vivo
situations, because low numbers of mucosal T cells could be stained
with the 5F4 mAb in active celiac disease (Fig. 2). These phenotypic
studies with iIELs from multiple donors and tissue sites (small
intestine and colon) indicate that induction of this activation Ag is
likely to be a common feature of human iIELs.
|
To define the nature of this activation Ag, the 34B1 mAb was used
to clone the cDNA that coded for the cognate Ag of the 34B1 mAb by COS
cell expression cloning. Because the three mAbs (34B1, 26H7, and 5F4)
were also noted to stain activated T cells from peripheral blood (data
not shown), COS cells were transfected with a mixture of three cDNA
libraries from resting and activated human PBTs. Transiently
transfected COS cells were subjected to three rounds of panning with
the 34B1 mAb. After the third round of panning, 17 of 50 randomly
selected E. coli transformants contained plasmids with a
3.3-kb insert. The inserts in these plasmids were similar by
restriction digest analysis. COS cells transfected with these plasmids
were stained specifically with the 34B1 mAb. One of these clones,
pPAN3.1, was selected for further characterization. This plasmid
directed the translation, when transfected into COS cells, of a 120-kDa
glycoprotein that was specifically recognized by the 34B1 and 5F4 mAbs
and that resolved as major band of 70 kDa and several minor bands of
lower molecular mass after digestion with N-glycanase (Fig. 3). A similar glycoprotein was
immunoprecipitated from radiolabeled cell surface iIEL proteins by all
three mAbs (Fig. 3). Complete DNA sequencing of both strands of this
cDNA revealed a sequence that was 97% identical with the b splice
variant of BGP or CD66a (GenBank accession no. X14831), with all the
differences occurring outside the coding region. Because the cDNA
predicted a polypeptide backbone of 58 kDa, the data in Fig. 3 suggest
that several of the carbohydrate modifications were relatively
resistant to N-glycanase digestion.
|
A). As shown in Fig. 4B, these studies confirmed the recognition of BGP (CD66a)
by the three mAbs and showed that all three mAbs reacted with the N
domain.
|
). Except for CGM1 (CD66b), the 34B1 mAb
stained all the CD66 family members tested, including all the CD66a
splice variants. The 26H7 and 5F4 mAbs, however, were quite interesting
in that they only stained the CD66a splice variants, suggesting that
they were probably specific for the N domain of this molecule. Notably,
there is only one other previous report of a CD66a-specific
single-chain Ab fragment and no previously described CD66a-specific
mAbs (33). These results clearly identify the N domain of
BGP as the cognate Ag of the 34B1, 26H7, and 5F4 mAbs and show that the
34B1 mAb recognizes other CD66 forms. In addition, the 34B1, 26H7, and
5F4 mAbs are probably specific for a determinant contained within the
polypeptide chain and not a carbohydrate side-chain modification,
because these mAbs recognized the protein from tunicamycin-treated cell
lines and specific mutations in the N domain-abrogated Ab binding to N
domain mutants (S. Watt and R. Blumberg, unpublished observations).
|
Our observation that BGP was expressed on activated iIELs, as
defined by staining with the BGP-specific mAbs, 34B1, 26H7, and 5F4,
was novel, because BGP has previously been primarily viewed as a
molecule expressed on epithelial cells and granulocytes and involved in
cell-cell adhesion and regulation of epithelial cell growth. In
addition, BGP is the only CD66 isoform expressed by activated human
iIELs. Fig. 6 shows the staining of an
activated human iIEL cell line from the small intestine, EEI-10, with a
panel of mAbs specific for CD66a–e. As shown, mAbs MA76 (CD66ae), MA86
(CD66acde), 34B1 (CD66acde), and 5F4 (CD66a), which are capable of
recognizing CD66a-specific mAbs, but not mAbs specific for CD66b
(MA41), CD66c (MA30), CD66e (MA27), or CD66de (MA84), exhibited
significant staining. Similarly, the mAbs MA28 (CD66e), MA61 (CD66b),
MA79 (CD66b), MA81 (CD66b), MA83 (CD66e), and MA91 (CD66e) did not
stain the activated human iIEL cell line, EEI-10 (data not shown).
|
shows a
dose-finding study with the three CD66-specific mAbs described here at
a range of concentrations. As predicted, all three Abs specifically
inhibited the anti-CD3-directed cytolysis and did not directly
activate the cytolytic function of the iIEL cell lines. Inhibition of
cytolysis required 100 µg/ml for the two highly specific mAbs (5F4
and 26H7), but only 10 µg/ml or less for the broadly reactive mAb
(34B1). To confirm these results, the two highly specific Abs were
examined at a concentration of 100.0 µg/ml at varying E:T cell ratios
in comparison with an isotype-matched Ab with the KJ-3 cell line. As
shown in Fig. 8A, neither the
5F4 nor the 26H7 mAbs directly stimulated the cytolytic activity of the
iIELs. In addition, both the 5F4 and 26H7 mAbs, but not the control Ab,
significantly inhibited the anti-CD3-directed lysis. The 5F4 mAb
inhibited the lysis by 22, 35, and 38% at E:T cell ratios of 100:1,
50:1, and 25:1, respectively. Similarly, the 26H7 mAb inhibited the
lysis by 21, 39, and 46% at E:T cell ratios of 100:1, 50:1, and 25:1,
respectively. On the other hand, the inhibition by the control Ab at
similar E:T cell ratios was 4, 5, and 5%. Moreover, the
anti-CD3-directed cytolysis was not inhibited by a CD2-specific
mAb, TS 2/18, which would be expected to inhibit CD58-like interactions
with the P815 cell line, suggesting that the inhibition by the
anti-CD66a mAbs was probably not due simply to an effect on
adhesion (Fig. 8B).
|
|
C, this type
of cytolysis was extremely sensitive to the inhibitory effects of all
three Abs (34B1, 26H7, and 5F4) in comparison with an IgG1 control Ab,
with up to 50% inhibition at a dose of 2 µg/ml of mAb. When
cytolysis was assessed with a pool of the three Abs (2 µg/ml each),
the lymphokine-activated killer activity-related cytolysis was
inhibited by 70%. This finding suggested synergistic inhibition by the
three mAbs and is consistent with recent findings that the three mAbs
recognize distinct epitopes within the N domain, based upon an analysis
of binding to N domain mutants (S. Watt and R. Blumberg, unpublished
observations). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human BGP is a member of the CEA family of glycoproteins, part of the Ig supergene family, and encoded in a large cluster on chromosome 19 (20, 22, 28, 29, 32). The CEA cluster is highly related to the genetically linked, pregnancy-specific gene cluster (32, 35). The CEA subgroup of this family is serologically defined as CD66a (BGP or C-CAM), CD66b (CGM6), CD66c (NCA), CD66d (CGM1), and CD66e (CEA). These structurally related glycoproteins consist of a highly homologous membrane distal amino-terminal IgV-like N domain and variable numbers of membrane distal IgC2-like domains in the case of BGP, NCA, CGM6, and CEA. In contrast to human CEA, CGM6, and NCA, which are linked to the membrane by a glycosyl phosphatidylinositol anchor, CGM7, CGM1, and BGP are type 1 transmembrane glycoproteins. The latter exist as isoforms containing short or long cytoplasmic tails.
BGP and its mouse and rat homologues C-CAM (35, 39, 40) have been regarded mainly as cell-cell adhesion and signaling molecules that are expressed primarily by epithelial cells of the gastrointestinal tract and biliary tree, neutrophils, and, more recently, B cells and human PBTs (37, 38, 41). Consistent with this we have observed that the mAbs described here stain epithelial cells in a number of human tissues (including intestine, tonsil, biliary tract, thymus, and kidney), tonsillar B cells, and granulocytes as determined by immunohistology (data not shown). BGP also serves as a receptor for mouse hepatitis virus (42) and for Opa proteins of the Neisseria species of bacteria (43). It is of interest that ligation of BGP on epithelial cells may deliver a negative growth signal, which may be decreased during tumor formation due to diminished expression of BGP (44). BGP also exhibits a high degree of alternate transcriptional processing, resulting in at least eight potential alternate transcripts. Two of these transcripts, BGPa and BGPb, encode a long cytoplasmic tail of 73 aa containing two ITIM motifs, which suggests a role as inhibitory receptors (35). Indeed, this cytoplasmic tail, when tyrosine phosphorylated, is capable of binding SHP-1 in a mouse colon carcinoma cell line (34). Such interactions may account for the inhibitory growth effect of this molecule on epithelial cells.
The studies contained in this report show that whereas BGP is constitutively expressed by IECs, it is an activation molecule on T cells adjacent to the epithelium, similar to the findings of two previous reports with PBTs (37, 38). However, in contrast to these earlier reports, which noted low levels of CD66a on PBTs and a subset of NK cells that were increased after in vitro activation (37, 38), we did not observe CD66a expression on resting iIELs, suggesting that CD66a expression may be actively suppressed in the epithelium under normal conditions. More importantly, the function of CD66a on iIELs and T cells in general is unknown.
In this regard, taking advantage of several newly generated mAbs with unique specificity for BGP, we have been able to determine that BGP regulates CD3-directed and lymphokine-activated killer activity of activated human iIELs. In preliminary studies the Abs described here also inhibit the activation of human PBTs, suggesting that the results contained in this report may be extensible to T cells in general (data not shown). The mechanism of the inhibition of cytolysis is unknown. However, given the observations that the cytoplasmic tail of the BGPa and BGPb splice variants contain the ITIM motif (34) and that the BGP homologue in mouse binds SHP-1 (33), it is possible that BGP on activated human iIELs interacts with intracellular phosphatases that down-regulate the function of ITAM-containing receptors such as CD3. Alternatively, BGP may function as an adhesion molecule that stabilizes effector cell interactions with the target such that blockade leads to diminished cytolysis. These hypotheses will be examined in future studies. Interestingly, the BGP gene maps to human chromosome 19q13.3 adjacent to the KIR locus on chromosome 19q13.4 (32, 45).
Although the ligand for BGP on the IEC is unknown, a good candidate is BGP itself or another CD66 family member in view of the known homophilic and heterophilic interactions among the CD66 group members (20, 21, 29, 35). These studies also suggest that BGP might provide inhibitory signals to iIELs in the absence of conventional inhibitory receptors such as KIRs and CTLA-4, which are notably absent from human CD8+ iIELs. It is also possible that the epithelial cell may actively regulate CD66a function and the capacity of the cell to mediate cytotoxicity, which would be highly relevant to epithelial cell infections, epithelial cell cancers, and chronic inflammatory diseases of the intestine. Indeed, we have observed CD66a expression on low numbers of mucosal T cells in celiac disease, as reported here.
The possibility that BGP might inhibit cytolytic T cell function during epithelial cell-T cell interactions extends the function of BGP to immunoregulation, making this the second example of a CD66 family member potentially involved in epithelial cell-T cell interactions. Mayer and colleagues have recently provided strong evidence for a role of a novel CD66e-related molecule, gp180, in directly ligating CD8 and activating p56lck on T cells (46). In conclusion, our studies strongly suggest a much larger role for CD66 family members in regulating T cell activation and deactivation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Richard S. Blumberg, Gastroenterology Division, Brigham and Women’s Hospital, Harvard Medical School, 75 Francis Street, Boston, MA 02115. E-mail address:
3 Abbreviations used in this paper: iIEL, intestinal intraepithelial lymphocyte; IEC, intestinal epithelial cells; KIR, killer inhibitory receptors; ITIM, immune receptor tyrosine-based inhibitory motif; BGP, biliary glycoprotein; PBT, peripheral blood T cell; CEA, carcinoembryonic Ag.
Received for publication October 16, 1998. Accepted for publication May 12, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cell receptor genes. J. Immunol. 150:5144.[Abstract]
1 T cell repertoire in human small intestine and colon. J. Exp. Med. 180:183.7 integrins on human mucosal lymphocytes. Proc. Natl. Acad. Sci. USA 89:1924. mutant mice with inflammatory bowel disease. J. Exp. Med. 183:847.2-Microglobulin-independent MHC class Ib molecule expression by human intestinal epithelium. Science 265:259.v3 as a heterotypic ligand for CD31/PECAM-1. J. Cell Sci. 109:437.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  M. Fujita, T. Otsuka, M. Mizuno, C. Tomi, T. Yamamura, and S. Miyake Carcinoembryonic Antigen-Related Cell Adhesion Molecule 1 Modulates Experimental Autoimmune Encephalomyelitis via an iNKT Cell-Dependent Mechanism Am. J. Pathol., September 1, 2009; 175(3): 1116 - 1123. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. MARKEL, T. COHEN-SINAI, M. J. BESSER, K. OVED, O. ITZHAKI, R. SEIDMAN, A. J. TREVES, E. FRIDMAN, Y. KEISARI, Z. DOTAN, et al. Preclinical Evaluation of Adoptive Cell Therapy for Patients with Metastatic Renal Cell Carcinoma Anticancer Res, January 1, 2009; 29(1): 145 - 154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Chen, L. Chen, S.-W. Qiao, T. Nagaishi, and R. S. Blumberg Carcinoembryonic Antigen-Related Cell Adhesion Molecule 1 Inhibits Proximal TCR Signaling by Targeting ZAP-70 J. Immunol., May 1, 2008; 180(9): 6085 - 6093. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Stern-Ginossar, S. Nedvetzki, G. Markel, R. Gazit, G. Betser-Cohen, H. Achdout, M. Aker, R. S. Blumberg, D. M. Davis, B. Appelmelk, et al. Intercellular Transfer of Carcinoembryonic Antigen from Tumor Cells to NK Cells J. Immunol., October 1, 2007; 179(7): 4424 - 4434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Yu, E. M. C. Chow, H. Wong, J. Gu, O. Mandelboim, S. D. Gray-Owen, and M. A. Ostrowski CEACAM1 (CD66a) Promotes Human Monocyte Survival via a Phosphatidylinositol 3-Kinase- and AKT-dependent Pathway J. Biol. Chem., December 22, 2006; 281(51): 39179 - 39193. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Markel, R. Seidman, N. Stern, T. Cohen-Sinai, O. Izhaki, G. Katz, M. Besser, A. J. Treves, R. S. Blumberg, R. Loewenthal, et al. Inhibition of Human Tumor-Infiltrating Lymphocyte Effector Functions by the Homophilic Carcinoembryonic Cell Adhesion Molecule 1 Interactions J. Immunol., November 1, 2006; 177(9): 6062 - 6071. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Azuz-Lieberman, G. Markel, S. Mizrahi, R. Gazit, J. Hanna, H. Achdout, R. Gruda, G. Katz, T. I. Arnon, S. Battat, et al. The involvement of NK cells in ankylosing spondylitis Int. Immunol., July 1, 2005; 17(7): 837 - 845. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. L. Servin Pathogenesis of Afa/Dr Diffusely Adhering Escherichia coli Clin. Microbiol. Rev., April 1, 2005; 18(2): 264 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Hemmila, C. Turbide, M. Olson, S. Jothy, K. V. Holmes, and N. Beauchemin Ceacam1a-/- Mice Are Completely Resistant to Infection by Murine Coronavirus Mouse Hepatitis Virus A59 J. Virol., September 15, 2004; 78(18): 10156 - 10165. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Markel, R. Gruda, H. Achdout, G. Katz, M. Nechama, R. S. Blumberg, R. Kammerer, W. Zimmermann, and O. Mandelboim The Critical Role of Residues 43R and 44Q of Carcinoembryonic Antigen Cell Adhesion Molecules-1 in the Protection from Killing by Human NK Cells J. Immunol., September 15, 2004; 173(6): 3732 - 3739. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Chen, H. Iijima, T. Nagaishi, A. Nakajima, S. Russell, R. Raychowdhury, V. Morales, C. E. Rudd, N. Utku, and R. S. Blumberg Carcinoembryonic Antigen-Related Cellular Adhesion Molecule 1 Isoforms Alternatively Inhibit and Costimulate Human T Cell Function J. Immunol., March 15, 2004; 172(6): 3535 - 3543. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-J. Chen and J. E. Shively The Cell-Cell Adhesion Molecule Carcinoembryonic Antigen-Related Cellular Adhesion Molecule 1 Inhibits IL-2 Production and Proliferation in Human T Cells by Association with Src Homology Protein-1 and Down-Regulates IL-2 Receptor J. Immunol., March 15, 2004; 172(6): 3544 - 3552. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Sundberg, N. Beauchemin, and B. Obrink The cytoplasmic domain of CEACAM1-L controls its lateral localization and the organization of desmosomes in polarized epithelial cells J. Cell Sci., March 1, 2004; 117(7): 1091 - 1104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Markel, H. Mussaffi, K.-L. Ling, M. Salio, S. Gadola, G. Steuer, H. Blau, H. Achdout, M. de Miguel, T. Gonen-Gross, et al. The mechanisms controlling NK cell autoreactivity in TAP2-deficient patients Blood, March 1, 2004; 103(5): 1770 - 1778. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Iijima, M. F. Neurath, T. Nagaishi, J. N. Glickman, E. E. Nieuwenhuis, A. Nakajima, D. Chen, I. J. Fuss, N. Utku, D. N. Lewicki, et al. Specific Regulation of T Helper Cell 1-mediated Murine Colitis by CEACAM1 J. Exp. Med., February 17, 2004; 199(4): 471 - 482. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Fournes, J. Farrah, M. Olson, N. Lamarche-Vane, and N. Beauchemin Distinct Rho GTPase Activities Regulate Epithelial Cell Localization of the Adhesion Molecule CEACAM1: Involvement of the CEACAM1 Transmembrane Domain Mol. Cell. Biol., October 15, 2003; 23(20): 7291 - 7304. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Greicius, E. Severinson, N. Beauchemin, B. Obrink, and B. B. Singer CEACAM1 is a potent regulator of B cell receptor complex-induced activation J. Leukoc. Biol., July 1, 2003; 74(1): 126 - 134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Finkenzeller, B. Fischer, S. Lutz, H. Schrewe, T. Shimizu, and W. Zimmermann Carcinoembryonic Antigen-Related Cell Adhesion Molecule 10 Expressed Specifically Early in Pregnancy in the Decidua Is Dispensable for Normal Murine Development Mol. Cell. Biol., January 1, 2003; 23(1): 272 - 279. [Abstract] [Full Text]
|
|
|
|
 |
|
 T. A. Plunkett and P. A. Ellis CEACAM1: A Marker With a Difference or More of the Same? J. Clin. Oncol., November 1, 2002; 20(21): 4273 - 4275. [Full Text] [PDF]
|
|
|
|
|
|
B. B. Singer, I. Scheffrahn, R. Heymann, K. Sigmundsson, R. Kammerer, and B. Obrink Carcinoembryonic Antigen-Related Cell Adhesion Molecule 1 Expression and Signaling in Human, Mouse, and Rat Leukocytes: Evidence for Replacement of the Short Cytoplasmic Domain Isoform by Glycosylphosphatidylinositol-Linked Proteins in Human Leukocytes J. Immunol., May 15, 2002; 168(10): 5139 - 5146. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Markel, N. Lieberman, G. Katz, T. I. Arnon, M. Lotem, O. Drize, R. S. Blumberg, E. Bar-Haim, R. Mader, L. Eisenbach, et al. CD66a Interactions Between Human Melanoma and NK Cells: A Novel Class I MHC-Independent Inhibitory Mechanism of Cytotoxicity J. Immunol., March 15, 2002; 168(6): 2803 - 2810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Klein, S. Hermann, C. F. Dietrich, D. Hoelzer, H. Martin ;, and D. Bunjes Transplantation-related toxicity and acute intestinal graft-versus-host disease after conditioning regimens intensified with Rhenium 188-labeled anti-CD66 monoclonal antibodies Blood, March 15, 2002; 99(6): 2270 - 2271. [Full Text] [PDF]
|
|
|
|
|
|
A. Nakajima, H. Iijima, M. F. Neurath, T. Nagaishi, E. E. S. Nieuwenhuis, R. Raychowdhury, J. Glickman, D. M. Blau, S. Russell, K. V. Holmes, et al. Activation-Induced Expression of Carcinoembryonic Antigen-Cell Adhesion Molecule 1 Regulates Mouse T Lymphocyte Function J. Immunol., February 1, 2002; 168(3): 1028 - 1035. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. Blau, C. Turbide, M. Tremblay, M. Olson, S. Letourneau, E. Michaliszyn, S. Jothy, K. V. Holmes, and N. Beauchemin Targeted Disruption of the Ceacam1 (MHVR) Gene Leads to Reduced Susceptibility of Mice to Mouse Hepatitis Virus Infection J. Virol., September 1, 2001; 75(17): 8173 - 8186. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Watt, A. M. Teixeira, G.-Q. Zhou, R. Doyonnas, Y. Zhang, F. Grunert, R. S. Blumberg, M. Kuroki, K. M. Skubitz, and P. A. Bates Homophilic adhesion of human CEACAM1 involves N-terminal domain interactions: structural analysis of the binding site Blood, September 1, 2001; 98(5): 1469 - 1479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Kammerer, D. Stober, B. B. Singer, B. Obrink, and J. Reimann Carcinoembryonic Antigen-Related Cell Adhesion Molecule 1 on Murine Dendritic Cells Is a Potent Regulator of T Cell Stimulation J. Immunol., June 1, 2001; 166(11): 6537 - 6544. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
