Perspectives in Diabetes - American Diabetes Association

Perspectives in Diabetes
The Role of Cell Adhesion Molecules in the
Development of IDDM
Implications for Pathogenesis and Therapy
Xiao-dong Yang, S a r a A. Michie, Reina E. Mebius, Roland Tisch, Irving Weissman,
and Hugh 0. McDevitt
IDDM is a chronic inflammatory disease in which there is
autoimmune-mediated organ-specific dtstmction of the
insulin-producing p-cells in the pancreatic islets of Langerhans. 'lhe migration of autoreactive lymphocytes and
other leukocytes from the bloodstream into the target
organ is of clear importance in the etiology of many
organ-specific autoimmunelinflanunatoqy disorders, including IllDM. In IDDM, this migration results in lymphocytic invasion of the islets (formation of insulitis) and
subsequent destruction of p-cells. Migration of lymphocytes from the bloodstream into tissues is a complex
process involving sequential adhesioe and activation
events. This migration is controlled in part by selective
expression and functional regulation of cell adhesion
molecules (CAMs) on the surface of lymphocytes and
vascular endothelial cells or in the extracellular matrix.
Understanding the mechanisms that regalate lymphocyte
migration to the pancreatic islets will lead to further
understanding of the pathogenesis of IaDM. In this article, we summaxize the recent advances regarding the
function of CAMs in the development of IDDM in animal
models and in humans and discuss the potential for
developing CAM-based therapies for IDDM. Diabetes 45:
705-710, 1996
ost lymphocytes recirculate throughout the
body, migrating from blood through organized
lymphoid tissues such as lymph nodes (LNs)
and Peyer's patches (PPs) or extranodal tissues
such as the pancreas, then to lymph and back to blood. This
migration is important in establishing effective immune surveillance ill that it allows effector lymphocytes to migrate to
almost any site of antigen deposition. However, the migraFrom the Department of Immunology (X.-D.Y.), Cell Genesys, Foster City; the
Departments of Microbiology and Immunology (R.T., H.O.M.) and Pathology
(S.A.M., R.E.M.,
I.W.), Stanford Universiw School of Medicine; and the Department
of Veterans Affairs (S.A.M.), Center for Molecular Biology, Palo Alto, California.
Address cor~espondenceand reprint requests to Dr. Xiao-dong Yang, Department of lmrnunology, Cell Genesys Inc., 324 Lakeside Dr., Foster City, CA 94404.
Received for publication 16 August 1995 and accepted in revised form 11
January 1996.
CAM, celI adhesion molecule; EC, endothelial cell; ICAM, intercelIular CAM; LN,
lymph node; LFA, lymphocyte function-associated antigen; mAb, monoclonal
antibody; MAtlCAM-1, mucosal addressin CAM-I; PNAd, peripheral node addressin; PP, Peyer's patches; STZ, streptozotocin; TNF, tumor necrosis factor;
VCAM-1, vascular CAM-1; VLA, very late antigen.
tion can also cause localized chronic inflammation that leads
to damage and destruction of host tissue. IDDM is one
example of such a tissue-specific chronic inflammatory disorder (1-3).
Lymphocyte migration from blood into tissue is a highly
complex process involving a cascade of adhesion and activation events (4-7). As illustrated in Fig. 1, many of the
adhesion events involve interaction of lymphocyte cell adhesion molecules (CAMs) with their endothelial cell (EC)
ligands. Some CAMs, such as lymphocyte function-associated
antigen-1 (LFA-1) and EC vascular cell adhesion molecule-1
(VCAM-I),seem to be involved in lynlphocyte migration to a
wide variety of tissues (8,9). In contrast, other CAMs are
involved in lymphocyte migration to one or a few tissues.
The lymphocyte CAMs associated with tissue-selective migration are known as homing receptors (10-13); their endothelial ligands, expressed in a tissue-selective manner either
constitutively in uninflamed tissues or induced in inflamed
tissues, are called vascular addressins (14-16).
CAMS can be classified into at least three major groups based
on their molecular structure: selectins, integrins, and the
immunoglobulin (Ig) superfamily (17). We will focus on
lymphocyte and EC CAMs that appear to be involved in the
pathogenesis of IDDM (Table 1). Several excellent reviews
are available that give more detailed information on the
other leukocyte and EC CAMs (9,17,18).
L-selectinRNAd. L-selectin is a member of the selectin
family of CAMs characterized by an NH,-terminal carbohydrate-binding lectin domain that binds to carbohydrate determinants on several EC glycoproteins (10,19). These
glycoproteins are collectively known as peripheral node
addressin (PNAd), many of which react with monoclonal
antibody (mAb) MECA-79 (Table 1) (14). L-selectin is expressed on most mononuclear cells except for some activated andlor menlory lymphocytes and functions to mediate
an initial and transient attachment, or rolling, of leukocytes
on vascular ECs (9,17,20). L-selectin is involved in lymphocyte traficking to some extranodal sites of inflammation,
including the thymus, skin, and pancreas (21-26). PNAd is
highly expressed by ECs in LNs but not in PPs, implying that
the degree to which L-selectin participates in tissue-selective
Altachmentl Rolling
Triggering 1 Activation
Firm Adhesion
Blood flow
Selective targets
for anti-adhesion
molecule mAbs
KBA, 2E6
YN 111.7
MK 2.6
KBA, 2E6
YN 1H.7
lymphocyte migration may be partially determined by the
expression pattern of its ligands.
a4P7 integrin/BIAdCAM-1. The integrins are a large family
of noncovalently-linked heterodimers that mediate a variety
of cell-cell and cell-extracellular matrix interactions (17,27).
Integrin molecules are structurally characterized by a unique
combination of 1 of 15 known a-chains with 1 of 8 P-chains.
Three integrins found on lymphocytes, cw4P1, a4P7, and
aLP2, have well-defined roles in lymphocyte migration and in
immune responses via their interactions with ligands on ECs,
extracellular matrix, or antigen-presenting cells.
a4P7 integrin is a lymphocyte homing receptor that mediates lymphocyte migration to mucosal sites such as PPs via
binding to the mucosal addressin CAM-1 (MAdCAM-1)
(11,12,28). MAdCAM-1, a glycoprotein with several Ig-like
domains and a single mucin domain, is constitutively expressed by vessels in mucosal lymphoid organs (PPs and
mesenteric LNs), intestinal lamina propria, and exocrine
pancreas (15,19,25). a4P7 also binds to VCAM-1 and fibronectin, both of which are ligands for a4pl integrin (30).
a4pl integrin/VCAM-Yfibronectin. The pl-integrh (CD29)
can pair with several a-chain integrins to form a unique aXP1
FIG. 1. Sequential multistep model of
lymphocyte/endothelial adhesion. Extravasculation of
lymphocytes from blood into the tissue is mediated by a
cascade of adhesive interactions between lymphocytes
and endothelial cells. Such interactions involve a
number of diverse CAMS in a sequential and
multiple-step process. This model implies that
lymphocyte migration can be controlled and regulated
at multiple points. It also provides a working model for
designing CAM-based therapeutic intervention
heterodimer (previously termed very late antigen [VLA])
(31). a 4 p l integrin (VLA-4, CD49dlCD29) has at least two
distinct ligands: endothelial VCAM-1 and the extracellular
matrix protein fibronectin (Table 1). VCAM-1, a member of
the Ig family of adhesion molecules, is strongly expressed in
vivo by ECs in a wide variety of inflammatory sites (32,33).
aLP2 integrin/ICAMs. The P2-integrin (CD18) can pair
with various cw-chain integrins (CD11) to form functionally
distinct CAMs. For instance, P2 pairing with aL forms cwLP2,
known as LFA-1 (CDlla/CD18), which is expressed on all
leukocytes, while Mac-1 (CDllbICDlS), formed by pairing of
P2 with cwM, is predominantly expressed by macrophages
and neutrophils (Table 1). Lymphocyte LFA-1 is thought to
participate in the binding of lymphocytes to vascular endothelium in many tissues, including LNs, PPs, and multiple
extranodal sites of inflammation (6,8). The endothelial ligands for LFA-1 include ICAM-1 and ICAM-2 (5).
Multistep model of lymphocyte/endothelial adhesion.
Studies on the adhesive interactions between leukocytes and
vascular endothelium under flow or shear force reveal that
such interactions involve a number of diverse CAMs in a
multiple-step process (Fig. 1) (4,6,17). An initial transient
Cell adhesion molecules: nomenclature and expression
Lymphocyte CAMS
selecti in
(Mel-14 Ag, CD62L)
Leukocytes, including most lymphocyte
Peripheral LNs, activated endothelium
Mucosal lymphoid tissues, activated endothelium
Activated endothelium, dendritic cells
Extracellular matrix
(LFA-1, CDllaICDl8)
(Mac-1, CDllb/CD18)
Monocytes and neutrophils
Factor X
Endothelium, activated lymphocyte, dendritic cells
Activated endothelium, dendritic cells
Extracellular matrix
Epithelial cells
Extracellular matrix
Extracellular matrix
Extracellular matrix
(WAS, CD49cICD29)
Endothelium, activated lymphocyte, dendritic cells
attachment, or rolling, is mediated by the interaction of
selectins such as L-selectin with endothelial carbohydrate
ligands or by the interaction of lymphocyte a4p7 with
MAdCAhl-1 (step 1). If the rolling leukocytes encounter
appropriate activating, chemotactic, or chemokinetic factors
in the local microenvironment, they may then undergo an
activation step that leads to strong adhesion, usually mediated by activated integrins (step 2). These microenvironmental factors may include cytokies, such as tumor necrosis
factor (TNF)-a, which can activate both attached lymphocytes and vascular ECs, and chemokines such as RANTES
(regulated upon activation, normally T-cell expressed and
secreted), which may play an important role in the recruitment of different lymphocyte subsets. The strong adhesion
can be followed by lymphocyte transendothelial migration
(diapedesis) into tissue (step 3). Depending on the differentiation ancVor activation stages of the cells and the nature of
the stimu1;atory signals they receive, different types of leukocytes and different subsets of lymphocytes may undergo this
process by slightly different steps (6,7). Nevertheless, this
model of sequential cellular adhesive interactions implies
that lymphocyte migration can be regulated at multiple
points. It also provides an important working model for
designing m vivo therapeutic intervention protocols (Fig. 1).
Obviously, CAMs are not only involved in lymphocyte
migration but also play critical roles in many other aspects of
the immune response. However, in this article we will focus
our discussion on the involvement of CAMs in lymphocyte
migration into the pancreatic islets.
To investigate lymphocyte migration in a chronic inflammatory disease such as IDDM, a series of studies covering
different time points during the disease process is necessary.
For this purpose, animal models have been used to delineate
the lymphocyte/endothelial adhesion pathways involved in
the diabetogenic process. The NOD mouse offers an ideal
model for these studies because 1) it spontaneously develops IDDM that closely resembles human IDDM; 2 ) it develops lymphocytic infiltrates in several wgans, including
pancreas, thyroid, and salivary glands, allowing comparison
of different tissue-selective adhesion pathways in the same
animal; and 3 ) mAbs are available against a wide variety of
murine lymphocyte and endothelial CAMs.
The developmental biology of vascular CAM expression in
NOD mouse pancreas has been investigated using immunohistochemistry. MAdCAM-1, but no PNAd, is found on many
vessels in exocrine pancreas of both NOD and control
C57BLKA mice during the first 3 weeks after birth. In NOD
mice, T-cells and macrophages are first detected in the islets
at -18-21 days of age. Macrophages (Mac-1') appear to
enter the islets a few days earlier than CD3+ T-cells. Vessels
with the morphology of high endothelial venules, which are
the vessels involved in most lymphocyte migration from
blood into LNs and PPs, develop aaacent to the islets at the
time when T-cells enter the islets (X.-D.Y., S.A.M., R.E.M.,
I.W., H.O.M., unpublished observations). These observations
suggest that monocytes or macrophages may be the initial
islet-infiltrating cells and that the formation of high endothelial venules may be crucial for the early lymphocytic accumulation.
From 3 weeks of age, increasing numbers of mononuclear
CAM expression on infiltrating lymphocytes and vascular ECs
within the endocrine pancreas of NOD mice
Lymphocyte CAMS
Endothelial CAMS
L-selectin a4P7 LFA-1 PNAd MAdCAM-1 VCAM-1
For expression of lymphocyte CAMS on the islet infiltrating cells:
-I+, <lo%; +, 10-3004 ++, 30-60%; +++, >60%of the infiltrating
cells expressing CAMs. For expression of EC C.4Ms in the infiltrated
islets: -/+, <5%; +, 510%) ++, 10-40%; +++, >40% of the
infiltrated islets expressing CAMs.
cells, including T- and p-cells, accumulate around the islets
@eri-insulitis) and gradually invade the islets (intra-insulitis). In parallel with the development of insulitis, expression
of some CAMs is upregulated, while that of others remains
unchanged (Table 2) (25,261.
MAdCAM-1 seems to be the predominant addressin expressed on ECs in and acijacent to the islets at early stages of
insulitis (5-7 weeks). In contrast, very little PNAd expression
is seen until there is significant insulitis (-8 weeks of age).
As insulitis progresses, the number of vessels expressing
MAdCAM-1 and PNAd increases (Table 2). Both of these
addressins are expressed mainly by peri-islet high endothelial venules (25).
While few L-selectin expressing lymphocytes are found in
the islets during the early stages of insulitis, most infiltrating
cells express high levels of the a4p7 integrin throughout the
disease process. The expression of a4p7 by lymphocytes
correlates with high expression of MAdCAM-1 by ECs,
further suggesting a predominant role for the mucosal lymphocyte homing pathway in the development of insulitis
(25,341. Moreover, most islet-infiltrating T-cells express low
levels of L-selectin and high levels of LFA-1, u4p7, VLA-4, and
CD44, a phenotype for activatedlmemory T-cells (26).
The tissue nonspecific CAMs, such as LFA-1 and ICAM-1,
are expressed by almost all islet-infiltrating lymphocytes
during the disease process. ICAM-1 is expressed by the
endothelium of most islet vessels in both uninflamed and
inflamed islets (25,261. It remains uncertain whether islet
endocrine cells, particularly the p-cells, express ICAM-1. If
ICAM-1 is expressed, an LFA-l/ICAM-1 interaction between
lymphocytes and target p-cells could have at least two
potential functions: 1) delivery of costimulatory signals for
activation of T-cells andlor 2 ) mediation of direct killing of
p-cells by cytotoxic T-cells. The failure to detect expression
of ICAM-1 on islet endocrine cells may simply be due to the
low sensitivity of current immunohistochemical approaches.
VCAM-1 is expressed by vessels in exocrine and endocrine
pancreas of diabetes-prone NOD mice as well as nondiabetes-prone BALB/c mice (25,26,35). Only a few VCAM-Iexpressing vessels are found within the islets (uninflamed or
inflamed), whereas many are found in the exocrine-pancreas
(Table 2). There is also VCAM-1 expression on some dendritic cells infiltrating the NOD islets (26). These findings
suggest that these tissue nonspecific CAMs may function as
costimulatory molecules in the local immune response.
Anti-CAM-based imnmnotherapy in IDDM in mice
Inhibition of various forms of IDDM
Cyclophosphamideor STZ-induced
24, 34, 36
24, 34, 35, 36, 38
Unpublished data
Unpublished data
41, 43
35, 36, 41, 43
41, 42, 43
41, 42, 43
-, no effect; -I+, 0-500h inhibition of diabetes incidence; +, 50-60% inhibition; + +, 60-80% inhibition; + + +, >80%inhibition; ND, not
Alternatively, they may play accessory roles in mediating
lymphocytic infiltration into the pancreas.
In addition to inflammation of the pancreatic islets, NOD
mice develop lymphocytic infiltration of other organs such
as salivary gland and thyroid. This makes NOD mice an
invaluable model to study the mechanisms involved in tissueselective lymphoc-yte trafficking to sites of chronic infiammation. In contrast to insulitis, which develops in the first
few weeks of life, overt lymphocytic infiltration of the
salivary gland (sialadenitis) is not seen in our colony until
around 6 months of age. The inflamed salivary glands show
marked vascular expression of PNAd and VCAM-1 but no
detectable expression of MAdCAM-1 (26,34). ICAM-1 is expressed by ECs as well as dendritic cells in the infiltrated
areas. The lymphocytes infiltrating the salivary glands show
CAM expression similar to that of LN lymphocytes: LFA-I+,
CD44+, some L-selectin+,and many a4+ (26).
Overall, these findings suggest that an interaction of mucosal homing receptor [email protected] and its vascular addressin,
MAdCAM-1, is a major adhesion pathway responsible for
selective lymphocytic migration to the islets, while VL4-41
VCAM-1 and probably L-selectin1PNAd interactions may primarily mediate lymphocyte migration into salivary glands
and possibly other organs (34). More importantly, this type of
tissue-selective homing in a chronic inflammatory disease
appears t o be determined by the selective expression of
vascular addressins on ECs (34).
In vivo studies using mAbs that recognize lymphocyte or
EC CAMS have been carried out to determine which adhesion pathway or pathways have a prominent role during the
diabetogenic process (Table 3). Treatment of NOD mice with
mAb Mel-14 (anti-L-selectin) leads to efficient protection
against the spontaneous occurrence of IDDM if given early in
the disease process, e.g., from birth to 4 weeks of age (34,36).
However, Mel-14 is unable to inhibit an ongoing diabetogenic
process if administered after the onset of insulitis, for
instance at 10 weeks of age (34). This may simply be due to
the fact that most T-cells in young animals have a naive
phenotype se select in^'^^), whereas more T-cells with an
activatedmemory phenotype (L-selectin- Or 'Ow ) are found in
older mice. Recently, it has been shown that only the
L-selectin- T-cell population from diabetic NOD mice is able
t o transfer the disease (37). These data imply that L-selectin
may only play a role in the initial phase of insulitis development in young animals (<4 weeks) and that in the late phase,
integrins such as [email protected] function to mediate both lymphocyte rolling and strong adhesion on ECs in the mucosal
lymphoid tissues (6,7). Strikingly, when the a4-integrin receptor is blocked by the Rl-2 mAb, NOD mice are signscantly protected from both spontaneous diabetes and
adoptive-transfer disease (24,34,35,38). This treatment also
prevents the development of insulitis, suggesting that protection from IDDM by R1-2 mAb may result from direct
inhibition of lymphocyte migration into the islets (24,34).
Moreover, our preliminary data demonstrate that antiLp7integrin and anti-MAdCAM-1 mAbs can also inhibit spontaneous development of IDDM in NOD mice (Table 3) (X.-D.Y.,
S.A.M., H.K. Sytwu, H.O.M.). Anti-VCAM-1 antibody marginally delays but fails to prevent disease in an adoptive-transfer
model (35). Unlike anti-LFA-1 and anti-ICAM-1 treatment,
which induces tolerance (39), blockade of L-selectin or
a4-integrin receptor appears to affect neither autoimmune
responses to @-cellsnor immune responses to foreign antigens, suggesting that the IDDM-protective effect induced by
blocking L-selectin or a4-integrin receptors is not a result of
immune suppression (24,34,40). Interestingly, the inhibition
seems to be tissue-selective, because sialadenitis is unaffected by either anti-L-selectin or anti-a4integrin treatment
Blocking LFA-1 andlor ICAM-1 receptors in vivo by specific mAbs has been shown to inhibit the spontaneous
development of IDDM in NOD mice and streptozotocin
(STZ)-induced insulitis and diabetes in mice (Table 3) (4143). Nevertheless, it remains controversial whether the LFAIACAM-1 lymphocytelendothelial adhesion pathway is
involved in the diabetogenic process, since administration of
both anti-LFA-1 and anti-ICAM-1mAbs is required to prevent
STZ-induced disease and treatment with either anti-LFA-1 or
anti-ICAM-1 fails to inhibit disease in adoptive-transfer
model in NOD mice (42,43). This raises the question of how
LFA-1 and ICAM-1 function in the development of autoimmunity and diabetes. On one hand, it is conceivable that
LFA-1ACAM-1 may be involved in strengthening the interaction between T-cells and target p-cells and facilitating p-cell
killing andor provide costimulatory signals for T-cell activation (5). On the other hand, there does not appear to be
increased LFA-1 and ICAM-1 expression in inflamed islets
during disease progression (Table 2). Therefore, LF'A-MCAM-1
interactions may not be a s important as the a407hLAdCAM-1
pathway in the control of tissue-selective lymphocyte homDIABETES, VOL. .ZZ, JUNE 1996
ing to the islets. One potential problem in using anti-LFA-1
and anti-ICAM-1 treatment as a therapy is the risk of inducing
immune suppression and/or exacerbation of the disease.
Treatment of mice with anti-LFA-1 or anti-ICAM-1 enhances
the severity of experimental allergic encephalomyelitis, an
animal model for multiple sclerosis (44).
Blocking ccMP2-integrin (Mac-1) with ahti-Mac-1 antibody
also prevents dlabetes in an adoptive-transfer model in NOD
mice, possibly via inhibiting recruitment of monocytes, macrophages, and other inflammatory cells in the islets (45).
During the evolution of diabetes, expression of certain
CAMs such as MAdCAM-1 in the islets may be developmentally programmed and may have a crucial role in initial
tissue-selective lymphocyte migration and in the priming of
autoreactive lymphocytes. After activation, lymphocytes and
inflammatory cells release a variety of biologically active
materials, such as cytokines and chemokines, which in turn
activate and stimulate expression of CAMs on ECs and
lymphocytes, resulting in more lyrnphocytic influx (46,47).
Among these cytokines, TNF-a, y-interferon, and interleukin10, are known to be involved in stimulation of CAM expression and are essential for the formation of insulitis (48-53).
Thus, expression of CAMs and their role in the diabetogenic
process may in turn be regulated by cytokine production and
The function of CAMs in human IDDM has recently been
investigated. However, such studies are hampered by the
limited availability of human pancreatic tissues, particularly
from prediabetic individuals. Much of the published work
has compared expression of certain CAMs in the pancreases
of diabetlc or newly diagnosed diabetic patients with the
pancreases from individuals without diabetes (54-57). By
immunostaining, ICAM-1 has been found on the vessels of
inflamed islets but not on vessels from uninflamed pancreases. It is still unclear whether ICAM-1 i s expressed on
endocrine cells found within the islets (54,57,58). LFA-3, an
EC ligand for CD2, is found on ECs of pancreases taken from
people with IDDM. In two IDDM patients, VCAM-1 expression was found on some dendritic cells scattered throughout
the pancreas but not on vascular endothelium (65,57).
The significance and implication of these results in human
IDDM remains uncertain, especially with regard to the function of these molecules in the control of lymphocyte trafficking. Furthermore, since the majority of studies have used
pancreatic material from overtly diabetic patients, the end
stage of the diabetogenic process, it is difficult to assess the
potential function of CAMs during the early development of
One interesting aspect of such studies in humans is the
observation that the serum levels of the extracellular domain
of CAMs such as soluble ICAM-1 (SICAM-1) appear to be
increased in some IDDM patients and their first-degree
relatives (59,60). In addition, SICAM-1 appears to inhibit
specific anti-p-cell T-cell responses (59). Whether these
observed phenomena are specific to the diabetogenic process
and what role soluble CAMs may play in the autoimmune
response anti diabetes process need further investigation.
DIABETES, VOl,. 45, JUNE 1996
Lymphocyte-endothelial adhesion followed by transendothelial migration is a key event in the development of organspecific autoimmunity. Selective interactions of certain cellsurface molecules regulate lymphocyte migration under
normal as well as pathological inflammatory conditions.
NOD mice are an ideal model for investigation of the role of
CAMs in regulation of lymphocyte migration t o the target
organs in chronic inflammatory diseases such as IDDM. Both
immunohistological and in vivo therapeutic studies in NOD
mice strongly suggest that there are tissue-selective lymphocytic homing pathways for insulitis development; the mucosal (a4P7/MAdCAM-1) adhesion system appears to be one of
the predominant ones. The fact that blocking these pathways
leads to successful intervention in the diabetogenic process
suggest that CAMs provide a potential therapeutic target for
human IDDM (36,61). Further studies on IDDM patients will
prove helpful for understanding IDDM pathogenesis, in addition to providing a basis for designing CAM-based therapeutic approaches.
This work was supported by grants from the National
Institutes of Health and the Department of Veterans Affairs.
X.-D.Y., R.E.M., and R.T. were recipients of Juvenile Diabetes
Foundation International, Arthritis Foundation, and National
Institute of Health fellowships, respectively.
1. Castano L, Eisenbarth GS: Type-I diabetes: a chronic autoimmune
disease of human, mouse, and rat. Annu Re?) Immunol 8:647-679, 1990
2. Hossini AA, Greiner DL, Friednlan HP, Mordes JP: Immunopathogenesis
of diabetes mellitus. Diabetes Rev 1:43-75, 1993
3. Bach J-F: Insulin-dependent diabetes mellitus as an autoimmune disease.
Endocrine Re11 15:516-542, 1994
4 . Butcher EC: Leukocyte-endothelial cell recognition: three (or more)
steps to specificity and diversity. Cell 6$:10'33-1036, 1991
5. Springer TA: Adhesion receptors of the immunc system. iliituw 34k42.5434, 1990
6. Bargatze RF, Jutila MA, Butcher EC: Distinct roles of L s e l ~ c t i nand
integrin alpha 4 beta 7 and LFA-I in lymphocyte honiing to Peyer's
patch-HEV in situ: the multistep model confirmed and refined. I~rrrrttcnity
399-108, 1995
7. Jones DA, McIntire LV, Smith CW, Picker LJ:A two-step adhesion
cascade for T cclVendothelial cell interactions under flow conditions. d
Clin Invest 94:2443-2450, 1994
8. Hamann A, Jahlonski-Westrich D, Duijvestijn A, Butcher EC. Baisch H,
Harder R, Thiele H-G: Evidence for an accessory role of LFA-1 in
lymphocyte-high endothelium interaction during homing. J Infmc~nol
140593-699, 1988
9. Carlos TM, Harlan JM: Leukocyte-endothelial adhesion molecules. B/ood
84:2068-2101, 1994
10. Gallatin WM, Weissman lL, Butcher EC: A cell-surface molecule involved
in organ-spccific homing of lynryl~ocyles.Nnturc 30430-34, 1983
11. Holzmann B, Mclntyre BW, Weissman IL: Identification of a murine
Peyer's patch specific lynlphocyte homing receptor as an integrin molecule with an a chain homologous to human VL4-4a. CcLL ,5637-46, 1989
12. Hu MCT, Croue DT. Weissman IL, Holzmann R: Cloning and expression
of mouse integrin Pp(p7): a functional role in Peyer's patch-specific
lymphocyte homing. Proc Nntl Acad Sci USA 89:8254-8258, 1992
13. Picker W, Michie SA, Rott LS, Butcher EC: A unique phenotype of
skin-associated lymphocytes in hun~an:preferential expression of the
HECA-452 cpitope by benign and malignant T cells at cutaneous sites.
Am J Pnthol 136:10.53-1068, 1990
14. Streeter PR, Rouse RT, Butcher EC: Immunohistologic and functional
characterization of a vascular addressin involved in lynxphocyte homing
into peripheral lymph nodes. J Cell Blol 107:1853-1862, 1988
15. Streeter PR, Berg EL, Rouse BTW, Bargatze RK, Butcher EC: A tissue
specific endothelial cell molecuIe involved in lymphocyte homing. Natrcw
331:41-46, 1988
16. Picker LT,Kishimoto TK, Smith CW, Wernack RA, Butcher EC: ELAM-I
is an adhesion molccule for skin-homing T cells. Natclw 349:796-799,
17. Springer TA: Traffic signals for lymphocyte recirculation and leukocyte
emigration: the multistep paradigm. Cell 76:301-314, 1994
18. Bevilacqua MP: Endothelial-leukocyte adhesion molecules. Annu Reu
Immunol 11:767-804, 1993
19. Lasky LA, Singer MS, Yednock TA, Doebenko D, Femie C, Rodriguez 8,
Nguyen T, Stachel S, Rosen SD: Cloning of a lymphocyte homing receptor
reveals a lectin dommn. Cell 56:104&1055, 1989
20. Jung TM, Gallatin WM, Weissman IL, Dailey MO: Down-regulation of
homing receptors after T cell activation. J I m m u l ~ o 141:4110-4117,
21. Michie SA, Rouse RV: Tr&c of peripheral B and T lymphocytes to
hyperplastic, preneoplastic thymuses of AKR mice. Am J P a l h 138:10151025, 1991
22. Michie SA, Streeter PR, Butcher EC, Rouse RV: Lselectin and a4p7
integrin hon~ingreceptor pathways mediate peripheral lyn~phocytetraffic
to AKR mouse hyperplastic thymus. .4m JPothol 147:412-421, 1995
23. Dawson J, Sedgwick AD, Edwards JCW, Lees P: The n~onoclonal
antibody MEL14 can block lymphocyte migration into a site of chronic
inflammation. Eur J Immunol 221647-1650, 1992
24. Yang X-D, Karin N, Tisch R, Steinman L, McDevitl HO: Inhibition of
insulitis and prevention of diabetes in NOD mice by blocking L-selectin
and VLA4 adhesion receptors. Proc Natl Acad Sci USA 90:10494-10498,
25. Hannincn A, Taylor C, Streeter PR, Stark I S , Sarte JM, Shizunl JA, Simell
0 , Michie SA: Vascular addressins are induced on islet vessels during
insnlitis in nonobese diabetic mice and are involved in lymphoid cell
binding to islet endothelium. J Clin Invest 92:3509-2515, 1993
26. Faveeuw C, Gagnerault MC, Lepault F: Expression of homing and
adhesion niolecules in infiltrated islets of Langerhans and salivary glands
of nonobese diabetic mice. J Irnw~unol1525969-5978, 1994
27. Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69: 11-25, 1992
28. Berlin C, Rrrg EL, Briskin MJ. Andrew DP, Kilshaw PJ, Holzmann B,
Weissman IL, Hamann A, Butcher EC: a4P7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAh-I. Cell 74: 185195, 1993
29. Briskin MJ, McEvoy Lhl, Butcher EC: MAdCAM-I has homology to
immunoglobulin and mncin-like adhesion receptors and to 1gA 1. Nnt71re
363:461-464, 1993
30. Ruegg C, Postigo k4, Sikorski EE, Butcher EC, Pytela R, Erle DJ: Role of
integrin alpha 4/beta P in lymphocyte adherence to fibronectin and
VCAM-I and in homotypic cell clustering. J Cell Biol 117:179-189, 1992
:31. Hemler ME: VLA proteins in the integrin family: structures, funclions,
and their role on leukocytes. Annu Reu Itr~w~unol
8:365-400, 1990
:i2. Usborn L: Leukocyte adhesion to endothelium in inflammation. Cell
623-6. 1990
33. Lobb RR, Hcrnler ME: The pathophysiologiral role of alpha4 integrin in
vivo. J Clin Invest 94:1722-1728, 1994
34. Yang X-D, Michie SA, Karin N, Tisch R, Steinman L, McDeritt H O A
predominant role of integrin a4 in the spontaneous development of
autoimmune diabetes in nonobese diabetic mice. Plac Na,tlAcad Sci USA
91:12604-12608, 1994
35. Baron JL, Reich EP, Visintin I, Janeway CW: The pathogenesis of
adoptive murine autoimmune diabetes requires an interaction between
alpha Cintegrins and vascular cell adhesion molecule-1. Clin Inuest
93: 1700-1708, 1994
36. Yang X-D,Michie SA, Tisch R, Karin N, Steinman L, McDevitt HO: Cell
adhesion molecnles: a selective therapeutic target for allerlation of
IDDM. J AutoCmmunity 7:859-864, 1994
37. Lepault F, Gagnerault MC, Faveeuw C. Bazin H, Boitard C: Lack of
I.-selectir~expression by cells transferring diabetes in NOD mice: ins~ghts
into the mechanisms involved in diabetes prevention by Mel-14 antibody
treatment. Eur J Iw~munol25 1502-1507, 1995
38. Burkly LC, Jakubowski A, Hattori M: Protection against adoptive transfer
of autoimmune diahetes mediated through very late antigen-4 integnn.
Diobetes 43:529-534, 1994
39. Isobe M, Yagita H, Okunlura K, Ihara A. Specific acceptance of cardiac
allografl after treatment with antibodies to ICAM-I and LFA-1. Science
255:112&1127, 1992
40. Tisch R, Yang X-D, Singer SM, Liblau RS, Fugger L, McDevitt HO: Immune
response to glutamic acid decarboxylase correlates with insulitis in
non-obese diabetic mice. Nature 366:72-75, 1993
41. Hasegawa Y, Yokono K, Taki T, Ammo K, Tominaga Y, Yoneda R, Yagi N,
Maeda S, Yagita H, Okumua K, Kasuga M: Prevention of autoimmune
insulin-dependent diabetcs in non-obese diabetic mice by anti-LFA-1 and
anti-ICAM-l mAb. Int Immunol 6:831-838, 1994
42. Hayashi T, Hashimoto S, Kameyama Y: Reduced streptozotocin-induced
insulitis in CD-1 mice by treatment with anti-intercellular adhesion
molecule-1 and anti-lymphocyte function associated antigen-1 monoclonal antibodies together with lactic dehydrogenase v i ~ u sinfection. Int J
Exp Path 75:117-121, 1994
43. Herold KC, Vrzys V, Gage A, Montag AG: Prevention of autoimmune
diabetes by treatment with anti-LFA-l and anti-ICAM-1 monoclonal
antibodies. Cell Immn~nol157:48R-500, 1994
44. Cannella B, Cross AH, Raine CS: Anti-adhesion molecule therapy in
experimental autoimmune encephalomyelitis. J Nenoroirnmn~nol46:4356, 1993
45. Ilutchings P, Kosen H, O'Reilly L, Simpson E, Gordon S, Cooke A:
Transfer of diabetes in mice prevented by blockade of adhesion-promoting receptor on macrophages. Nature 348:639-642, 1990
46. Doukas J , Pober JS: IFN-y enhances endothelial activation induced by
tumor necrosis factor but not 1L-1. J I m n ~ u n o 1451727-1733,
47. Doukas J, Mordes JP: T lymphocytes capable of activating endothelial
cells in vitro are present in rats with autoimmune diabetes. .I I?nmunol
150:1036-1046, 1993
48. Yang X-D, Tisch R, Singer SM, Cao ZA, Liblau R, Schreiber RD, McDcbitt
HO: Effect of tumor necrosis factor CY on insulin-dependent diabetes
~uellitusin NOD mice. I. The early development of autoimmunity and
diabetogenic process. J E q Med 180:995-1004, 1994
49. Yang X-D, McDevitt HO: Role of TNF-a in the development of autoimmunity and the pathogenesis of insulin-dependent diabetes mellitus in
&OD mice. Circ Shock 43:198-201, 1994
50. Picarella DE, Kralz A, Li C , Huddle NH, Flavell RA: Transgenic tumor
necrosis factor (TNF)-a production in pancreatic islets leads to insulitis,
not diabetes. J Irnmunol 150:4136-4150, 1993
51. Higuchi Y, EIerrera P, Muniesa PJH, Belin D, Ohashi P, Aichele P, Orci L,
Vassalli J-D, Vassalli P: Expression of a tumor necrosis factor a trar~sgene
in murine pancreatic p cells results in severe and permanent insulitis
without evolution towards diabetes. J Exp Med 176:1719-1731, 1992
52. Lce MS, Sarvetnick N: Induction of vascular addressins and adhesion
molecules in the pancreas of IFN-gamma transgenic mice. J Irnmunol
1524597-4603, 1894
53. Wogensen L, Huang X, Sawetnick N: Leukocyte extravasation into the
pancreatic tissue in transgenic mice rxprrssing interleukin 10 in the
islets of Langerhans. J E x p hfed 178.17&185, 1993
54. Vives M, Soldevila G, Alcalde L, Lorenzo C, Son~ozaN, wol-Borrell R:
Adhesion molecoles in human islet p-cells: de novo induction of ICAM-1
but not LFA-3. Diabetes 40:1382-1390, 1991
65. Hanninen A, Jalkanen S, Salnii M, Toikkanerr S, Nikolakaros G, Simell 0 :
hlacrophages, T cell receptor usage, and endothelial cell activation in the
pancreas at the onset of insulin-dependent diabetes mellitus. J Clin
Invest 90: 1901-1910, 1992
56. Itoh N. Hanafi~saT, Miyazaki A, Miyagawa J-1, Yamagata K, Yanlarnoto K,
Waguri M, I~nagawaA, Tamura S, Inada M, Kawata S, Tarui S, Kono N,
Matsr~zawaY: Mononulear cell infiltration and its relation to the expression of major histocompatibility complex antigens a r ~ dadhesion molecules in pancreas biopsy specimens from newly diagnosed insulindependent diabeles mellitus patients. J C1i.r~Invest 922313-2322, 1993
57. Somoza N, Vargas F, Roura-hlir C, Vives-Pi M, Teresa M, Jaraquemada D,
Pujol-Borrell R: Pancreas in rrrent onset insulin-dependent diabetes
mellitus: changes in HLA, adhesion molecules and autoantigens, restricted T cell receptor VB usage, and cytokine profile. J Imrnunol
153:1360-1377, 1994
58. Campbell IL, Cutri A, Wilkinson D, Boyd AW, Ilanison LC: Intercellular
adhes~onmolecr~le1 is induced on isolated endocrine islet cells cytokines but not by reovinls infection. Pror Natl Acad Sci USA 8642824286, 1989
59. Roep BO, Heidenthal E, de Vries RR, Kolb H, Martin S: Soluble fonns of
interrellr~laradhesion molecule-1 in insulin-dependent diabetes mellitus.
Lancet 343:1590-1593, 1994
60. Lampeter ER, Kishimoto TK, Rothlein R, Mainolfi EA, Bertrarns J , Kolb H,
Martin S: Elevated levels of circulaling adhesion molecules in IDDM
patients and in subjects at risk for IDDM. Diob~fes41:1668-1671, 199'2
61. Yang X-D, Tisch R, McDcvitt HO: Selective targets for imn~unotherapyin
autoimmune disease. Chem Immur~ol60:20-31, 1994