| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
ICOS Corporation, Bothell, WA 98021; and
Burnham Institute, La Jolla, CA 92037
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Integrins are heterodimers consisting of 
and 
subunits. Within
the amino-terminal region of the CD11a (L) subunit of LFA-1 there is
an A or I domain that contains a critical binding site for ICAM-1. The
I domain consists of approximately 200 residues and possesses a
structure with a central open twisted sheet surrounded by
helixes (15). This fold, designated dinucleotide binding
domain or Rossman fold, belongs to a subfamily of the / domain
superfamily. Purified recombinant I domain binds ICAM-1
(16), and mutational studies indicate that the ICAM-1
binding interface is located on the upper face of the LFA-1 I domain
around a metal coordination site (17, 18, 19). This site
consists of a conserved motif, the metal ion-dependent adhesion site
(MIDAS),2 present in
all integrin I domains. Metal was shown to be critical to LFA-1 binding
to ICAM-1, and an essential glutamate (E34) in ICAM-1 was proposed to
contribute to metal coordination (20). Recently, it was
reported that Mg2+ in the MIDAS of
21 forms a direct
contact with the carboxylate oxygen of a glutamate in its ligand
collagen, forming a strong electrostatic bond between integrin and
ligand (21).
Two different I domain conformations, open and closed, have been
reported for Mac-1 and 2 I domain crystal
structures (21, 22). The open, high affinity conformation
was observed in crystals that had Mg at the MIDAS bound to a glutamate
from either its ligand collagen (2) or an
adjacent I domain (Mac-1). The open form differed from the closed,
ligand-free, low affinity structure in metal coordination at the MIDAS
as well as a large 10A shift in the position of the C-terminal
7 helix. Flexibility of the
7 helix of LFA-1 I domain is suggested by a
crystal structure in which the helix appears to be displaced through a
crystal packing interaction and by nuclear magnetic resonance solution
structure determination (23, 24).
We have reported recently, using nuclear magnetic resonance, that
residues around the LFA-1 MIDAS are affected by ICAM-1 binding. ICAM-1
binding was also found to affect a second site distal to the MIDAS in
the region around the C-terminal 7 helix and
proximal sheet (25). COS cell transfectants expressing
mutant LFA-1 with alanine substitutions in the core of this second site
demonstrated constitutively active binding. These mutations were
predicted to destabilize the closed or inactive conformation visualized
in crystal structures and/or stabilize or induce the active or open
conformation. In addition, alanine substitution of hydrophilic residues
around the I domain allosteric site (IDAS) core decreases binding to
ICAM-1, but does not appear to contribute to the ICAM-1 binding
interface. The corresponding wild-type (WT) residues may stabilize the
active or open conformation of the I domain. In addition, a small
molecule antagonist, Lovastatin, that binds close to this site may
function by stabilizing the inactive conformation (26).
This second site, the IDAS, thus regulates binding at MIDAS.
We demonstrate here that regulation at LFA-1 IDAS controls adhesion to ICAM-1 activated through inside-out signaling in lymphocytes. In addition, modulation of adhesion by IDAS mutations correlated to a conformational change detected with a novel activation epitope CD18 mAb.3 Furthermore, using surface plasmon resonance with soluble LFA-1, we report that the mechanism of this regulation by the IDAS appears to be mediated by alteration of the off rate and thereby the affinity state of LFA-1.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hybridoma cells producing anti-human CD11a mAb (TS1/22, IgG1), or anti-human CD18 mAb (TS1/18, IgG1) were obtained from American Type Culture Collection (ATCC; Manassas, VA). The stimulatory anti-human CD18 mAb 240Q (IgG1), the activation epitope mAb 327C (IgG1), and control mAb 1B7 (IgG1) were generated at ICOS (Bothell, WA). All cell culture media and antibiotics were obtained from Life Technologies (Gaithersburg, MD). All other reagents were obtained from Sigma (St. Louis, MO).
Generation of CD11a double mutants
To generate activation/inactivation double mutations of CD11a, a fraction of G1115A cytoplasmic tail activation mutation was shuttled into each of the knockout mutant CD11a backgrounds. Specifically, D137A, K305A, Y307A, E310A, and WT CD11a in pDC1 were digested with EcoRI, internal site in CD11a, and XbaI, 3' vector site, and the 7.4-kb vector/5' portion of CD11a was isolated and purified. Similarly, G1115A CD11a in pDC1 was digested with EcoRI and XbaI, and the 2.3-kb fragment containing the 3' portion of CD11a was isolated and purified. The 5' CD11a fragments containing the IDAS or MIDAS mutations were individually ligated to the 3' portion of CD11a containing the G1115A activation mutation. Resulting clones were sequenced to verify the integrity of the junctions and the presence of both activating and inactivating mutations.
COS cell transfections and COS cell adhesion assay
COS cell transfections and adhesion assays were conducted as previously described (25). Every mutation was tested in triplicate samples in at least three independent transfections and assays, and representative assays are displayed. The percentage of cell binding (percentage of capture) was determined using the mean values for each triplicate in a given assay and the formula: [(A570 (binding to ICAM-1) - A570 (binding to BSA))/A570 (binding to CD11a + CD18 mAb)] x 100.
Generation J2.7 clones
To evaluate the effects of key IDAS mutations of CD11a in a lymphoid setting, three inhibitory mutations, D137A, Y307A, and E310A, and two activating mutations, I235A and G1115A, as well as WT CD11a were subcloned into pCDNA3. Briefly, the pDC1 CD11a constructs were digested with HinDIII and XbaI. The CD11a-containing fragments were isolated, purified, and ligated into similarly digested pCDNA3. Resulting clones were sequenced to verify both the integrity of the junctions and the presence of the mutations.
J2.7 cells (provided by L. Klickstein, Brigham and Women’s
Hospital, Boston, MA) were then electroporated with 30 µg mutant or
WT CD11a pCDNA3 DNA/13.2 x 106 cells in
RPMI, 10% FBS, 1.25% DMSO at a density of 40 x
106 cells/ml, using 0.4-cm path-length cuvettes,
250 V, and 250 µF. Immediately following electroporation cells were
diluted into 30 ml of the same medium and cultured overnight at 37°C.
After 24 h, the cells were spun down and recultured in RPMI, 10%
FBS, 1 mg/ml G418.
Following selection, the J2.7 transfectants were stained with
anti-CD11a Ab (TS1/22) and sorted by FACS. The top 1% of each
population was plated out for cloning by limited dilution in 96-well
plates. Resulting clones were reanalyzed by FACS with TS1/22 to assess
the level of LFA-1 expression. Clones with comparable levels of LFA-1
were used in all studies. The level of LFA-1 expression was confirmed
before all experiments.
J2.7 adhesion assay
The adhesion assay was performed as described previously (25) with modifications. Briefly, Costar EasyWash 96-well plates (Cambridge, MA) were coated overnight at 4°C with 50 µl/well ICAM-1 (4 µg/ml for assays with inactivation mutations or 2 µg/ml for assays with activation mutations), capture mAb (TS1/18 and TS1/22 at 5 µg/ml each), or buffer alone overnight at 4°C. The next morning, plates were washed with 200 µl/well Dulbecco’s PBS. Plates were blocked at room temperature for 1.5 h with 100 µl/well adhesion buffer (AB; RPMI/5% heat-inactivated FBS). Cells were spun down and resuspended in AB at 1 x 106 cells/ml. Serial dilutions of PMA (final concentrations in assay, 1.56–100 ng/ml) and of the activating mAb 240Q (final concentrations in assay, 0.156–10 µg/ml) were prepared at a 3x final concentration in AB. Blocking and control Abs (final concentration in assay, 20 µg/ml) were also prepared as 3x stocks in AB. Plates were emptied and blotted dry, and 100 µl/well AB was added. Then, 100 µl/well AB with or without stimulatory agent or Ab was added, and the plates were incubated for 10 min at 37°C. Next, 100 µl/well cell suspension was added, and the plates were spun down at 1000 x g for 1 min to promote cell contact. The plate was then incubated for 15 min at 37°C. Adherent cells were fixed, stained, and analyzed as in the COS cell adhesion assay described above.
FACs staining
In a 96-well plate, 1–5 x 105 cells of each transfectant were stained with an Ab to CD18 (TS1/18; ATCC), an Ab to CD11A (TS1/22; ATCC), and an activating Ab to CD18 (mAb 240Q, ICOS). Sheep anti-mouse Ig -FITC (F-2883, Sigma, St. Louis, MO) at a 1/200 dilution was added to each sample. Washed cells were fixed in 1% formaldehyde and analyzed on the same day. Controls included unstained cells, cells stained with secondary Ab only, and cells stained with an isotype-matched control Ab (1B7).
Production and purification of sLFA-1 proteins
The soluble forms of recombinant WT LFA-1 (sLFA-1/WT) and I235A mutant (sLFA-1/I235A) contain deletions of the transmembrane and cytoplasmic domains of CD11a and CD18, and substitution of these regions for acidic and basic leucine zipper cassettes, respectively, which promote and stabilize specific heterodimerization, were generated as described for the production of soluble TCR (27).
Both WT and mutant I235A CD11a were truncated after position Q1063 in the mature polypeptide, and the 47-aa acidic leucine zipper cassette (LZA, TRSSADLVPRGSTTAPSAQLEKELQALEKENAQLEWELQALEKELAQ) was added in-frame, using standard methods. CD18 was truncated after position N678 in the mature polypeptide, and the 47-aa basic leucine zipper cassette (LZB, TRSSADLVPRGSTTAPSAQLKKKLQALKKKNAQLKWKLQALKKKLAQ) was added in-frame. Both sLFA-1/WT and sLFA-1/I235A were expressed in CHO cells and purified from the supernatants using separate immunoaffinity columns created by coupling 2 mg of an anti-CD18 mAb (23I11B, ICOS)/ml activated cyanogen bromide-Sepharose (Amersham Pharmacia, Uppsala, Sweden), according to the manufacturer’s suggested protocol, and eluted using a 20 mM Tris (pH 7.5), 2.5 M MgCl2 buffer. Purity and approximate m.w. were verified by SDS-PAGE and Coomassie blue staining. Soluble LFA-1/WT and sLFA-1/I235A were then purified twice by gel filtration chromatography over a HiLD SuperDex 200 column (Amersham Pharmacia) in Dulbecco’s PBS buffer using standard methods to remove any single chain, aggregated, and/or degraded material. The peak fractions corresponding to homogenous suspensions of purified heterodimers were pooled and concentrated using Ultrafree-4 Centrifugal Filter Units with Biomax-30 membranes (Millipore, Bedford, MA), then dialyzed in HBS buffer (10 mM HEPES (pH 7.4), 150 mM NaCl, and 2 mM MgCl2) at 4°C, and quantitated using a Bio-Rad protein assay (Bio-Rad, Philadelphia, PA) and the manufacturer’s protocol.
Activation epitope mAb 327C isolation and binding to J2.7 clones
Ab 327C was generated by immunizing mice with recombinant sLFA-1/WT using standard methods and screening for Abs that stained LFA-1-expressing cells with increased intensity in the presence of 1 mM Mn2+. The Ab detects a conformational change in CD18 that occurs when LFA-1 function in primary human lymphocytes is stimulated by PMA or TCR cross-linking, and the appearance of this neoepitope does not require the presence of ICAM ligands. This conformational neoepitope is also up-regulated by high concentrations of ICAM-1 in the absence of cellular stimulation. Cellular stimulation of 327C binding with PMA or TCR cross-linking is more modest in Jurkat than in primary isolates of human lymphocytes; however, stimulation of 327C in Jurkat is robust upon ICAM-1 or 240Q stimulation (see Footnote 3).
For analysis of J2.7 clones, 327C was first directly conjugated with
Alexa-488 (Molecular Probes, Eugene, OR) following the manufacturer’s
protocol. Then, 2.5 x 105 cells were
stained in 75 µl RPMI/10% heat-inactivated serum with 10 µg/ml
327C-Alexa-488 or isotype-matched control-Alexa-488 conjugate. Stimuli
included medium alone, 20 ng/ml PMA, or 150 µg/ml ICAM-1/Fc for 20
min at 37°C. Cells were then washed, fixed, and subjected to analysis
on the FACS. The mean fluorescence intensity of the sample was
corrected for background staining by subtracting the staining of
isotype-matched negative control Ab and was normalized for LFA-1
expression level by staining with TS1/18.
BIAcore analysis
Binding analysis of sLFA-1/WT and sLFA-1/I235A was performed on
a BIAcore 2000 biosensor (Pharmacia Biosensor). All experiments were
performed at 25°C. All proteins for injection were diluted with HBS
buffer. To immobilize ICAM-1/Fc to the Biosensor, polyclonal rabbit
anti-human IgG Fc (Pierce, Rockford, IL) was coupled to the sensor
chip (
12,000 relative units (RU)) using the amine coupling
kit (Pharmacia Biosensor) as previously described (28),
except that the Ab was injected at 50 µg/ml in 10 mM sodium acetate
(pH 4.5). For each binding reaction ICAM-1/Fc was injected at 10
µg/ml until approximately 200 RU were bound, then sLFA-1/WT or
sLFA-1/I235A was injected at a flow rate of 10 µl/min. The sensor
surface was regenerated before each new reaction with 0.1 N
HCl.
Analysis of the binding data in BIAcore
Analysis of the kinetic data for sLFA-1/WT and sLFA-1/I235A binding to immobilized ICAM-1/Fc was performed using standard kinetic equations (28) and nonlinear curve fitting within the BIA Evaluation 2.0 program (Pharmacia Biosensor). The portions of the sensorgrams corresponding to the dissociation of sLFA-1/WT and sLFA-1/I235A from immobilized ICAM-1/Fc were first analyzed to obtain the Kd. The Ka was then determined by nonlinear curve fitting to the association phase data using the model of one site. Kd was calculated from the ratio Kd:Ka.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
By analyzing 24 independent alanine substitution mutations within
the CD11a I domain and five within the linker region, we previously
demonstrated that different mutations can increase or decrease the
binding of LFA-1 to ICAM-1 relative to WT when expressed in COS cells
(summarized in Table I
). These mutations
aided in defining an I domain allosteric site of regulation we termed
the IDAS. We previously determined that substitution of certain
residues within the IDAS hydrophobic pocket, typified by I235A,
resulted in constitutively active binding to ICAM-1 (25)
(Fig. 1A), and these mutations
behaved very similarly to an activating mutation within the cytoplasmic
GFFKR region, G1115A. However, substitution at other residues proximal
to the IDAS hydrophobic pocket, typified by K305A, decreased binding to
ICAM-1. The binding of these mutants to ICAM-1 could be recovered by
stimulation with the CD18 mAb, 240Q (inducible mutants). In addition,
certain substitutions in the carboxy-terminal I domain linker, such as
Y307A and E310A, resulted in an inactive phenotype that did not bind
ICAM-1 in the presence or the absence of 240Q stimulation. These
mutations behaved similarly to an inactivating mutation at the MIDAS,
D137A. All mutations were expressed at equivalent levels, as determined
by FACS analysis with both CD18 and CD11a mAbs (25) (data
not shown). Therefore, these data suggested that the IDAS may regulate
the activation state of the integrin. However, these experiments did
not determine whether activating signals from within the cell
(inside-out signaling) were also controlled at the IDAS.
|
|
B). Binding of
the double mutant, K305A+G1115A, could be recovered by stimulating the
ectodomain of LFA-1 directly with 240Q. As expected, the MIDAS
substitution mutant, D137A, also blocked the increased binding of
G1115A, and binding was not recovered with 240Q, similar to E310A.
Interestingly, the G1115A mutation partially compensated the Y307A
inactivating mutation such that 240Q stimulation resulted in
approximately 20% WT levels of binding. Thus, these data suggest that
inside out signaling mediated through the GFFKR region can be regulated
by the IDAS. The IDAS regulates one determinant of adhesion
Although the data reported above were suggestive of a role for the
IDAS in regulating inside-out signaling, limitations of the COS-7
system prevented further characterization. Therefore, to further
address IDAS involvement in inside-out signaling, representative single
mutants and WT CD11a were expressed in a CD11a-negative lymphoid cell
line J2.7. Lymphoid cell clones expressing inactivating mutants were
chosen that had equivalent or greater levels of expression relative to
their matched WT clones (Fig. 2A). In contrast, clones
expressing the activating mutants were chosen that expressed equivalent
or lower levels of expression relative to their matched WT clones (Fig. 3A). This insured that any
effects observed were not due to differences in expression levels of
the mutants relative to the WT LFA-1.
|
|
2.7 cells PMA
stimulated adhesion of WT expressing clones to ICAM-1 in a
dose-dependent fashion (Figs. 2B and 3B).
PMA-stimulated adhesion of WT-expressing clones reached a maximum of
about 60% binding of cells. In contrast, PMA-stimulated adhesion to
ICAM-1 was significantly inhibited by the IDAS-proximal linker
mutations Y307A and E310A (Fig. 2B). PMA-stimulated adhesion
of Y307A- and E310A-expressing clones reached a maximum of about 10%
binding of cells. Similarly, 240Q-stimulated adhesion of WT-expressing
clones resulted in maximum adhesion of 100% of binding cells, whereas
neither Y307A nor E310A could respond 240Q stimulation (Fig. 2C). Thus, consistent with COS cell transfectants, the Y307A
and E310A mutations interfere with inside-out signaling in lymphoid
cells and block mAb-stimulated adhesion mediated through
CD18.
Interestingly, in contrast to COS cell transfectants, I235A did not
demonstrate constitutive binding when expressed in lymphoid cells (Fig. 3B). However, I235A binding was inducible, with increased
binding at lower PMA concentrations relative to WT. PMA concentrations
that resulted in 100% binding of cells expressing I235A (50 ng/ml)
stimulated only 40–50% binding of cells expressing WT (Fig. 3B). In contrast to both WT and I235A, when G1115A was
expressed in lymphoid cells it bound constitutively to ICAM-1 as it did
in COS transfectants (Fig. 3B). In addition, binding was
further increased with PMA stimulation and reached a maximum much
earlier (10 ng/ml PMA) than either WT or I235A. Thus, I235A binding
can be negatively regulated by a cytoplasmic mechanism that can no
longer function with the cytoplasmic mutation G1115A. Therefore, the
IDAS regulates only one determinant of binding, and an additional
determinant exists that is under cytoplasmic control. In addition,
these data strongly suggest that I235A does not spontaneously induce
the activated state of LFA-1, but, rather, may either stabilize the
activated state once it forms or destabilize the inactive state and
thereby lower the threshold necessary for activation. In COS-7 cells
integrins become partially activated perhaps due to differences within
the regulation of the plasma membrane or heterodimer formation between
COS-7 and Jurkat cells. Therefore, in our previous report in COS-7
cells, I235A appeared to be constitutively active.
Stabilization of the active state of LFA-1 through I235A mutation would
result in a slower dissociation rate of the receptor resulting in the
net production of more high avidity LFA-1/ICAM-1 complexes per cell and
thereby increase the number of cells strongly adherent to ICAM-1
compared with WT. Similarly, the decreased ability of G1115A to be
negatively regulated and thereby shut off would result in a similar
effect. These effects would result in a greater number of I235A- or
G1115A-expressing cells becoming adherent than WT-expressing cells, as
observed in Fig. 3B. If correct, this hypothesis would
predict that WT-expressing clones should achieve the same level of
maximum adhesion as the I235A and G1115A mutant-expressing clones if
stimulated by the 240Q activating Ab. Indeed, this is precisely what
was observed, as 240Q stimulation results in all clones reaching 100%
of binding (Fig. 3C).
Ligand-induced activation-epitope expression are regulated by the IDAS
At least two determinants, which are not mutually exclusive, may contribute to formation of the activated state of LFA-1. An overall increase in LFA-1 binding to ligand may be achieved by either 1) an increase in affinity of LFA-1 for ICAM-1 (the strength of binding between the two molecules) or 2) an increase in overall avidity/clustering (the affinity multiplied by the number of interactions occurring at one time). Much debate in the literature has surrounded the relative contributions of these two determinants in the generation of integrin activation states (7, 9, 10, 11, 12, 13, 14). The ability of I235A to functionally increase the active state of LFA-1 allowed us to probe this question of avidity vs affinity.
The binding of certain conformation-dependent mAbs has previously been
shown to strongly correlate with the activated state of LFA-1 induced
upon ligand binding (29). These mAbs are dependent on
LFA-1 binding to its ligand and therefore have been used as indirect
detection reagents of ligand binding affinity. We have developed our
own activation-dependent mAb (327C) that responds to ICAM-1 binding
(see footnote 3). We tested the ability of this Ab to bind to LFA-1 on
the different J2.7 clones constitutively and in response to PMA or
ICAM-1/Fc binding (Fig. 4). Inactivating
mutants Y307A and E310A and the control MIDAS mutant D137A demonstrated
decreases in both constitutive binding of mAb 327C and binding upon
stimulation with PMA or ICAM-1/Fc. However, the epitope can be induced
upon Y307A and E310A with 240Q stimulation, indicating that the epitope
itself is not eliminated directly by the mutation. Although 240Q
stimulation could not bring the level of epitope expression on Y307A or
E310A up to the levels observed with WT, it did significantly increase
their expression levels of the epitope. The lower level of constitutive
binding to the inactivating mutants may reflect a lack of LFA-1/ICAM-1
binding events between cells that occurs at low levels in WT-expressing
clones. Both I235A and G1115A mutants demonstrated modest increases in
constitutive and notable increases in ligand-induced 327C binding,
which suggested that these mutants have higher affinities for ICAM-1.
Interestingly, again a significant difference between I235A and G1115A
was observed in PMA-stimulated 327C binding. Like the difference
between these two mutants in ICAM-1 binding in response to PMA (Fig. 3B), these data on activation epitope expression in response
to PMA demonstrate that at least two determinants affect the LFA-1
activation state. Interestingly, the response to 240Q was similar
between WT, I235A, and G1115A, and similar data were obtained with
330E, another CD18 neoepitope Ab that binds to a distinct epitope (see
footnote 3).
|
Although the results presented above suggest that I235A, and therefore the IDAS itself, influences the affinity state of LFA-1, since the previous assays rely on cell surface expression of the integrin, secondary effects due to cross-linking (avidity) could not be completely ruled out. Therefore, to accurately compare the relative affinity for ICAM-1 of WT LFA-1 and the IDAS mutant I235A, we generated recombinant soluble forms, sLFA-1/WT and sLFA-1/I235A (see Materials and Methods). Both were produced in secreted form in CHO cells and purified from the cell culture supernatants over an anti-CD18 immunoaffinity column, then repurified by gel filtration chromatography twice over a Pharmacia HiLD SuperDex 200 column in PBS buffer to remove any aggregated and/or degraded material. The resulting suspensions of purified heterodimers were concentrated, then dialyzed in HBS buffer before analysis.
The affinity of sLFA-1/WT and sLFA-1/I235A was then measured by surface
plasmon resonance using a BIAcore sensor chip CM5 to which an
anti-human Fc Ab (Pierce) was covalently linked. Following the
method of Tominaga et al. (30), a subsaturating amount of
ICAM-1/Fc was first captured to the chip via an anti-human Fc Ab
(Fig. 5A), then sLFA-1/WT or
sLFA-1/I235A was allowed to bind at different concentrations in HBS
buffer, and the surface plasmon resonance was recorded (Fig. 5, B and C). After each concentration of LFA-1 or
LFA-1/I235A was allowed to bind and dissociate, the chip was stripped
of ICAM-1/LFA-1 complexes with 0.1 N HCl and regenerated with fresh
ICAM-1/Fc before the next concentration of LFA-1 or LFA-1/I235A was
analyzed (Fig. 5A).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). These substitutions block the CD18
neoepitope induced by PMA stimulation or ICAM-1 binding, but not
neoepitope expression induced by 240Q stimulation. We reported
previously that the conformational or quaternary change in LFA-1 I
domain, presumably induced by the activating CD18 mAb, 240Q, is blocked
by these I domain linker substitutions (25). Thus, the
ability of structural changes in CD18 to induce activated ligand
binding also appears to be restrained or regulated by CD11a I domain
C-terminal linker; however, restriction at the CD11a C-terminal linker
may not directly control all conformational changes in CD18. When distal cytoplasmic restraints are removed through GFFKR substitution or cellular activation, the activating IDAS substitution I235A, in the presence of ICAM-1, may either stabilize the active or open I domain conformation or destabilize the inactive or closed I domain conformation. In COS cells, where WT LFA-1 demonstrates a low level of constitutive binding, the I235A substitution resulted in increased constitutive binding, and therefore appeared to stabilize or induce the open/active conformation. However, in lymphoid cells where LFA-1 may be under negative cytoplasmic regulation, cellular stimulation was required to demonstrate the enhanced adhesion and neoepitope expression mediated by I235A relative to WT. Therefore, I235A appeared to destabilize the closed/inactive conformation, thereby lowering the energy barrier to convert to the open conformation.
Activation of LFA-1-dependent adhesion may involve an increase in
avidity due to LFA-1 clustering or affinity resulting from a
conformational change. We demonstrate here that affinity modulation is
indeed a mechanism by which the IDAS can regulate LFA-1-dependent
adhesion. Different affinities for LFA-1 binding to ICAM-1 have been
reported. The binding affinity of monomeric soluble ICAM-1 to
immobilized LFA-1 was determined to be 130 nM (32). Two
other studies have used soluble receptor and ligand and surface plasmon
resonance to determine affinity. In the first study by Labidia et al.
(33) LFA-1 was covalently linked to the BIAcore chip, and
affinity was measured by probing with soluble ICAM-1. Using this method
a Kd of 133 nM was reported. In the second
study by Tominaga et al. (30) ICAM-1 was noncovalently
linked to the BIAcore chip, and affinity was measured using soluble
LFA-1. Using this method a Kd of 500 nM was
reported. Importantly, Tominaga et al. also demonstrated that accurate
measurement of the affinity of LFA-1 was dependent on purifying soluble
heterodimers away from spontaneously aggregated material. This
secondary purification by size-exclusion chromatography was not
performed in the study by Labidia et al. and may explain the
differences in affinity reported when determined in a similar manner by
surface plasmon resonance. Without the purification of heterodimers the
assay may be more prone to multivalent interactions or deposition of
aggregated LFA-1 on the chip. In our study we followed the method of
Tominaga et al. and noncovalently linked ICAM-1 to the BIAcore chip,
followed by probing with sLFA-1 that had been purified twice by
size-exclusion chromatography to insure that aggregated material was
removed. The Kd of WT sLFA-1 in our study (547
nM) is nearly identical with the 500 nM reported by Tominaga et al.
(30). The Kd of sLFA-1/I235A, which
was purified under identical conditions and analyzed in the same
experiment, was 86 nM. This affinity is strikingly similar to the 80 nM
Kd of PMA-activated V3 binding to a ligand
mimetic (34). In a cellular system data from a competition
binding assay between sICAM-1 and a blocking LFA-1 mAb were used to
obtain evidence for a subpopulation of LFA-1 that converts to a higher
affinity state (35). However, the high affinity state
reported in this system (400 nM) is approximately 4-fold lower than the
affinity we determined for I235A using surface plasmon resonance. One
reason for this difference in affinities measured in a cell or
cell-free system may be the presence of leukocyte glycocalyx
sialogylcoproteins such as CD43, which can interfere with LFA-1
binding to ICAM-1 (36).
The I domain structures for CD11a, CD11b, CD49a, and CD49b have been described (22, 24, 37, 38, 39). Except for CD11b, all these structures were of the closed form, which led to the suggestion that the open form was a crystal artifact (40) However, recent studies have shown that the closed and open forms are functionally relevant and may correspond to the low affinity (inactive) and high affinity (active) states of the corresponding integrin (8, 14, 21, 38).
Although no structural data exist for the open form of the LFA-1 I
domain, we have modeled its structure based on CD11b (Fig. 6, right) and compared it to
the known structure of the closed form (Fig. 6, left).
Interestingly, in the closed form the strand residue, I235,
interacts closely with side chains from five other residues (L132,
F153, L302, I259, and I306); however, in the predicted open form I235
loses interactions with two C-terminal 7 helix
residues (L302, I306). The substitution of a methyl group for the
isopropyl group in the I235A mutant would be predicted to also cause a
loss of these interactions in the closed form and therefore may promote
the transition to the open form. This provides a likely biochemical
explanation for the increased affinity. Thus, loss of these
interactions may decrease the energy barrier to switch from a closed to
an open conformation. The data for LFA-1 strongly suggest that
regulation of LFA-1 activation occurs at least in part through an
allosteric switch at the IDAS and are consistent with the functional
conformational changes in
21 and CD11b
(8, 21, 41). We have observed regulation of the ligand
binding activity of CD11b (M. Lupher, unpublished observation), similar
to that which we previously described for CD11a. In this regard, it is
very interesting that while this article was in preparation, Xiong et
al. (14) recently described a similar correlation between
mutation at the IDAS region of CD11b, affinity of the isolated I
domain, and the open conformation of the CD11b I domain.
|
). The substitution I235A
increases ICAM-1-induced activation. Thus, high affinity binding may be
the product of a cell activation-dependent release of a restraint at
the IDAS that favors the closed conformation and an ICAM-1-induced fit.
We propose here that the IDAS regulates a conformational switch within
the I domain that controls the formation of the high affinity state of
LFA-1 and thereby regulates adhesion of LFA-1 to ICAM-1.
|
Footnotes |
|---|
2 Abbreviations used in this paper: MIDAS, metal ion-dependent adhesion site; IDAS, I domain allosteric site; PKC, protein kinase C; AB, adhesion buffer; WT, wild type; sLFA, soluble form of LFA; RU, relative units.
3 C. R. Beals, A. C. Edwards, R. J. Gottschalk, T. W. Kuijpers, and D. E. Staunton. CD18 activation epitopes induced by leukocyte activation. Submitted for publication.
Received for publication March 5, 2001. Accepted for publication May 21, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2, 3, and 7 integrin activity. J. Immunol. 160:5622.L2) integrin. Proc. Natl. Acad. Sci. USA 92:10277.2 integrin lymphocyte function-associated antigen-1 contains a binding site for ligand intercellular adhesion molecule-1. J. Biol. Chem. 269:12395.21. Cell 101:47.[Medline]
subunit of integrin CR3 (CD11b/CD18). Cell 80:631.[Medline]
and T-cell receptor extracellular segments. Proc. Natl. Acad. Sci. USA 91:11408.V3 detected with a novel patch-engineered monovalent ligand. J. Biol. Chem. 274:21609.21. J. Biol. Chem. 272:28512.1, 2 and 7 integrins differentially regulate LFA-1 function in K562 cells. Mol. Biol. Cell. 8:719.[Abstract]
M2 on monocytic cells. J. Biol. Chem. 269:3159.This article has been cited by other articles:
|
|
 |
|
  M. R. Sarantos, H. Zhang, U. Y. Schaff, N. Dixit, H. N. Hayenga, C. A. Lowell, and S. I. Simon Transmigration of Neutrophils across Inflamed Endothelium Is Signaled through LFA-1 and Src Family Kinase J. Immunol., December 15, 2008; 181(12): 8660 - 8669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Green, U. Y. Schaff, M. R. Sarantos, A. F. H. Lum, D. E. Staunton, and S. I. Simon Dynamic shifts in LFA-1 affinity regulate neutrophil rolling, arrest, and transmigration on inflamed endothelium Blood, March 1, 2006; 107(5): 2101 - 2111. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Sarantos, S. Raychaudhuri, A. F. H. Lum, D. E. Staunton, and S. I. Simon Leukocyte Function-associated Antigen 1-mediated Adhesion Stability Is Dynamically Regulated through Affinity and Valency during Bond Formation with Intercellular Adhesion Molecule-1 J. Biol. Chem., August 5, 2005; 280(31): 28290 - 28298. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kim, C. V. Carman, W. Yang, A. Salas, and T. A. Springer The primacy of affinity over clustering in regulation of adhesiveness of the integrin {alpha}L{beta}2 J. Cell Biol., December 20, 2004; 167(6): 1241 - 1253. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Fraemohs, R. R. Koenen, G. Ostermann, B. Heinemann, and C. Weber The Functional Interaction of the {beta}2 Integrin Lymphocyte Function-Associated Antigen-1 with Junctional Adhesion Molecule-A Is Mediated by the I Domain J. Immunol., November 15, 2004; 173(10): 6259 - 6264. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Vitte, A. Pierres, A.-M. Benoliel, and P. Bongrand Direct quantification of the modulation of interaction between cell- or surface-bound LFA-1 and ICAM-1 J. Leukoc. Biol., September 1, 2004; 76(3): 594 - 602. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. E. Green, D. N. Pearson, R. T. Camphausen, D. E. Staunton, and S. I. Simon Shear-Dependent Capping of L-Selectin and P-Selectin Glycoprotein Ligand 1 by E-Selectin Signals Activation of High-Avidity {beta}2-Integrin on Neutrophils J. Immunol., June 15, 2004; 172(12): 7780 - 7790. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Wigelsworth, B. A. Krantz, K. A. Christensen, D. B. Lacy, S. J. Juris, and R. J. Collier Binding Stoichiometry and Kinetics of the Interaction of a Human Anthrax Toxin Receptor, CMG2, with Protective Antigen J. Biol. Chem., May 28, 2004; 279(22): 23349 - 23356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Yang, M. Shimaoka, A. Salas, J. Takagi, and T. A. Springer Intersubunit signal transmission in integrins by a receptor-like interaction with a pull spring PNAS, March 2, 2004; 101(9): 2906 - 2911. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Dunne, R. G. Collins, A. L. Beaudet, C. M. Ballantyne, and K. Ley Mac-1, but Not LFA-1, Uses Intercellular Adhesion Molecule-1 to Mediate Slow Leukocyte Rolling in TNF-{alpha}-Induced Inflammation J. Immunol., December 1, 2003; 171(11): 6105 - 6111. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. P. Wojcikiewicz, X. Zhang, A. Chen, and V. T. Moy Contributions of molecular binding events and cellular compliance to the modulation of leukocyte adhesion J. Cell Sci., June 15, 2003; 116(12): 2531 - 2539. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R.C. Liddington and M.H. Ginsberg Integrin activation takes shape J. Cell Biol., September 3, 2002; 158(5): 833 - 839. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. F. H. Lum, C. E. Green, G. R. Lee, D. E. Staunton, and S. I. Simon Dynamic Regulation of LFA-1 Activation and Neutrophil Arrest on Intercellular Adhesion Molecule 1 (ICAM-1) in Shear Flow J. Biol. Chem., May 31, 2002; 277(23): 20660 - 20670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Welzenbach, U. Hommel, and G. Weitz-Schmidt Small Molecule Inhibitors Induce Conformational Changes in the I Domain and the I-like Domain of Lymphocyte Function-associated Antigen-1. MOLECULAR INSIGHTS INTO INTEGRIN INHIBITION J. Biol. Chem., March 15, 2002; 277(12): 10590 - 10598. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Q. Ma, M. Shimaoka, C. Lu, H. Jing, C. V. Carman, and T. A. Springer Activation-induced Conformational Changes in the I Domain Region of Lymphocyte Function-associated Antigen 1 J. Biol. Chem., March 15, 2002; 277(12): 10638 - 10641. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
