| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
,,,
*
The Chemo-Sero-Therapeutic Research Institute, Kumamoto, Japan;
Department of Immunology, Juntendo University School of Medicine, Tokyo, Japan; and
Core Research for Evolutional Science and Technology, Japan Science and Technology Corporation, Tokyo, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
To date, at least 18 molecules, including TNF-
, lymohotoxin
(LT)-/TNF-
, FasL, LT-, CD27L, CD30L, CD40L, 4-1BBL, OX40L,
TNF-related apoptosis-inducing ligand (TRAIL)/APO2L, TNF-like weak
inducer of apoptosis/APO3L, and B cell activation factor from the TNF
family/NH2-terminal kinase/B lymphocyte
stimulator, have been reported to be the TNF family members
(8, 9). All these members, except for LT-, exist as
type II cell surface proteins and exhibit pleiotropic biological
activities such as cell death, differentiation, and proliferation via
their binding to cognate receptors belonging to the TNFR family
that have several repeats of characteristic cysteine-rich domain (CRD)
in their extracellular region (8, 9, 10). Like other members
of the TNF family of molecules, the biologically active FasL molecule
exists as a homotrimerized complex (11). A FasL trimer
recruits three Fas molecules, resulting in oligomerization of a
cytoplasmic death domain, which recruits and activates procaspase-8 via
an adaptor molecule, Fas-associated death domain. The activated
caspase-8 in turn activates downstream caspases that lead to apoptotic
cell death characterized by structural changes such as DNA
fragmentation and chromatin condensation (1).
Among the TNF family members, the crystal structures of TNF-,
LT-, CD40L, and TRAIL have been revealed (12, 13, 14, 15, 16, 17, 18).
Despite relatively low amino acid sequence similarity between these
molecules, their three-dimensional structures are highly
conserved. Each monomer consists of sandwich (jellyroll) topology
composed of several strands and loops bridging the strands.
Further studies of the crystal structures of LT-/TNFRI and
TRAIL/TRAIL-R2 complexes revealed that one individual receptor molecule
interacts with two ligand molecules in the trimer complex via two
different contact sites, one between the outer tip-forming D-E loop on
the ligands and the loop motif on the second CRD in the receptors and
another between the residues relatively near the top of the ligands and
the loop on the second/third CRD of the receptors (18, 19).
Previously, we established murine neutralizing anti-human FasL mAbs, NOK1, NOK2, and NOK3, and characterized the biological nature of FasL (20). Furthermore, we indicated the therapeutic usefulness of neutralizing anti-FasL mAb in murine models of fulminant hepatitis and lethal acute GVHD (6, 7). In the present study, we generated a humanized version of NOK2, which might be useful for clinical application to human diseases. By extensive alanine-scanning mutagenesis and computed molecular modeling, we determined amino acid residues on human FasL that constitute the conformational epitopes recognized by murine NOK1, -2, and -3 and humanized NOK2. We also determined which human FasL residues are critical for Fas binding and cytotoxic activity. Based on these results, a molecular model of FasL/Fas interaction is proposed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
Chinese hamster ovary (CHO) cells and COS cells were obtained from American Type Culture Collection (Manassas, VA) and cultured in RPMI 1640 medium containing 10% FCS, 100 µg/ml streptomycin and penicillin, and 2 mM glutamine (culture medium). WR19L cells and human Fas (hFas) cDNA transfectants (hFas/WR19L) were kindly provided by Dr. S. Yonehara (Kyoto University, Kyoto, Japan) and cultured in the culture medium.
Reagents
Murine anti-human FasL mAbs, NOK1 (IgG1-
), NOK2
(IgG2a-), and NOK3 (IgM-), were purified from culture
supernatants as previously described (20). An
anti-FLAG mAb M2 and disuccnimidyl suberate (DSS) were purchased
from Eastman Kodak (New Haven, CT) and Pierce (Rockford, IL),
respectively.
Humanization of NOK2
The cDNAs encoding murine Ig VH and
VL regions were isolated by RT-PCR from NOK1-,
NOK2-, or NOK3-producing hybridoma. The VH region
was amplified using AAG CTT GCC GCC ACC ATG GAA TGG AGC TGG GTC TTT as
the 5' primer and GGA TCC ACT CAC CTG AGG AGA CGG TGA as the 3' primer.
The VL region was amplified using AAG CTT CGC CAC
CAT GAA GTT GCC TGT TAG GCT G as the 5' primer and GGA TCC ACT TAC GTT
TTA TTT CCA GCT T as the 3' primer. Each 5' and 3' primer was tagged
with the HindIII and BamHI sites, respectively.
After subcloning of the cDNA into pBluescript II SK(+) TA cloning
vector (Stratagene, La Jolla, CA), the nucleotide sequence was
determined by using an automated sequencer (Applied Biosystems, Foster
City, CA) and the fluoresceinated dye terminator cycle sequencing
method. Humanization of NOK2 was performed essentially as described
previously (21, 22). Human VH cDNA
(SGI, kindly provided by Dr. M. M. Bendig, Medical Research
Council Collaborative Center, London, U.K.) and
VL cDNA from a human PBMC cDNA library were used
as the templates. In practice, most parts of
complementarity-determining regions (CDRs) of human
VH and VL regions were
genetically replaced with the amino acid residues corresponding NOK2
CDRs by using a site-directed mutagenesis kit (Stratagene) according to
the manufacturer’s instruction. Some additional residues in human
framework regions (FRs) were also were genetically substituted by
corresponding amino acid residues in the NOK2 FRs on the basis of our
computed molecular models and the knowledge from previous
Ab-humanization studies. The HindIII-BamHI
fragments of humanized NOK2 VH and
VL cDNAs were then transferred to
HindIII-BamHI sites of pCAG-
1 and pCAG-,
respectively. pCAG-1 is a mammalian expression vector modified from
HCMV-VH0.5-1 (23) in which the CMV promoter was
replaced with the chicken -actin promoter and rabbit -globin
splicing acceptor. pCAG- was generated by replacing the human 1
constant region in pCAG-1 with the constant region. CHO cells
were transfected with both pCAG-1 and pCAG- carrying humanized
NOK2 VH and VL cDNA,
respectively, by electroporation. After selection with neomycin and
methotrexate (both from Sigma-Aldrich, St. Louis, MO), humanized and
reshaped NOK2 mAbs were purified by using a protein G column from the
culture supernatants. Among several versions of humanized NOK2 mAbs,
one mAb, designated RNOK203, was selected for its strong ability to
inhibit FasL cytotoxicity against hFas/WR19L.
Construction and expression of FasL mutants
The cDNA encoding the extracellular domain of FasL was amplified
by RT-PCR from pMKITNeo (kindly provided by Dr. K. Maruyama, Tokyo
Medical and Dental University, Tokyo, Japan) containing full-length
human FasL (20) by using TCT GGT ACC TGT GGG CAG CTC GAC
TAC AAG GAC GAC GAT GAC AAG CAC CTA CAG AAG GAG CTA GCA GAA CTC CGA GAG
TCT as the 5' primer and GCC AAG CTT GGA TCC TTA GAG CTT ATA TAA GCC
GAA as the 3' primer. The KpnI site plus the FLAG tag
sequence and the BamHI site were added to the 5' primer and
the 3' primer, respectively. After digestion with KpnI and
BamHI, the PCR product of 577 bp was subcloned into the
KpnI and BamHI sites of pCAGn-C25VL, which has
been constructed to express the V region of
anti-HIV mAb C25 under the control of chicken -actin promoter,
rabbit -globin splicing acceptor, and human CMV enhancer (details
will be described elsewhere), resulting in in-frame fusion of
the human Ig signal sequence, the FLAG tag, and the extracellular
domain of human FasL (native hFasL/pCAGn). For alanine-scanning
substitutions, nucleotide mutations with alanine conversions were
introduced into native hFasL/pCAGn using a site-directed mutagenesis
kit (Stratagene). At amino acid residue A240, a glycine was used for
the substitution. Transient expression in COS cells was performed by
using lipofectACE (Life Technologies, Rockville, MD) according to the
manufacturer’s instruction. After 24 h of incubation, the culture
medium was changed to ASF serum-free medium (Ajinomoto, Tokyo,
Japan). The culture supernatant was collected after 72 additional
hours, and the concentration of recombinant soluble FasL was evaluated
by ELISA using anti-FLAG mAb M2 as described below.
ELISA for FasL/anti-FasL mAb binding and for FasL/Fas binding
To evaluate the concentration of FLAG-tagged native or mutant FasL, an Immulon 2 plate (Dynatech Laboratories, Chantilly, VA) was coated with serially diluted native or mutant FasL for 16 h at 4°C. After washing with 0.05% Tween 20-PBS, the wells were blocked with 1.0% BSA-PBS for 2 h at 37°C, 50 µl of anti-FLAG mAb (2 µg/ml) was added to each well, and wells were incubated for 1 h at 37°C. Then, 50 µl of x5000 diluted HRP-conjugated goat anti-mouse Ig Ab (American Qualex, San Clemente, CA) was added and incubated for 1 h at 37°C. The wells were developed with 100 µl 1 mg/ml orthophenylenediamine in 50 mM citrate-phosphate buffer (pH 5.0) containing 0.03% H2O2 and stopped with 50 µl of 2 N H2SO4. OD (A450–490) was measured on an automated ELISA reader (Molecular Devices, Sunnyvale, CA). Serial dilutions of native FasL, which was affinity purified by a NOK2 column from the supernatant of native hFasL/pCAGn cDNA-transfected COS cells, was used as the standard.
For evaluating binding of anti-FasL mAbs to native or mutant FasL, the Immulon 2 plate was coated with native or mutant FasL (5 µg/ml) for 16 h at 4°C. After blocking with 1.0% BSA-PBS, serially diluted NOK1, -2, and -3 or humanized NOK2 (RNOK203) was added and incubated for 1 h at 37°C. Then HRP-conjugated goat anti-mouse Ig Ab and anti-human Ig Ab (American Qualex) were used to detect the bound NOK1, -2, and -3 and RNOK203, respectively.
For evaluating binding to Fas, the Immulon 2 plate was coated with 1 µg/ml human Fas-human IgG1 Fc chimeric protein (Fas-Ig; Alexis, San Diego, CA) for 16 h at 4°C. After blocking with 1.0% BSA-PBS, 500 ng/ml of native or mutant FasL was added to each well and incubated for 1 h at 37°C. The bound FasL was evaluated by subsequent incubations with anti-FLAG mAb and HRP-labeled anti-mouse Ig Ab as described above.
Chemical cross-linking and Western blot analysis
The culture supernatants containing native or mutant FasL were
concentrated 
40-fold by using a Centriprep-10 (Amicon, Beverly, MA).
Ten microliters of aliquot was diluted with 30 µl PBS and then
incubated with or without 0.15 mM DSS (Pierce) for 10 min at 4°C.
Then the samples were subjected to SDS-PAGE under reducing conditions,
followed by Western blot analysis using anti-FLAG mAb, goat
HRP-labeled anti-mouse Ig Ab, and ECL (Amersham Pharmacia Biotech,
Little Chalfont, U.K.).
Cytotoxicity assay
Cytotoxic activity of native or mutant FasL was tested against hFas/WR19L cells by the Alamar blue method according to the manufacturer’s instruction (Alamar Biosciences, Sacramento, CA). Briefly, hFas/WR19L (1 x 104) cells were incubated with serially diluted native or mutant FasL in a total volume of 100 µl. After a 16-h incubation, 10 µl Alamar blue was added to the culture, and the culture was further incubated for 4 h. Fluorescence of the reduced Alamar blue was detected on a Fluoroscan (Titertec Fluoroscan II; Labosystems, Tokyo Japan) at 590 nm by an excitation at 544 nm.
Computer modeling of human FasL and Fas
A three-dimensional molecular model of human FasL monomer was
built by MODELER (Accelrys, San Diego, CA), a program that
implements comparative modeling by maximal satisfaction of spatial
restraints. The crystallographic structures of both TNF- (12, 13) and LT- (14) (Brookhaven Protein Data Bank
entries 1TNF and TNR) were used as the templates. After molecular
modeling of FasL monomer, three copies of each monomer were
superimposed on the structure of TNF trimer (1TNF). The trimer
structure was optimized by 100 steps of conjugate gradient minimization
using the CHARMm program (Accelrys). In a similar way, molecular models
of the FasL/Fas complex were predicted on the basis of crystal
structure of the human LT--human TNFRI x-ray complex
(19).
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
To perform humanization, we first determined the amino acid
sequences of VH and VL of
murine anti-human FasL mAbs, NOK1, NOK2, and NOK3, by sequencing
RT-PCR-generated cDNAs (Fig. 1
A). Although the NOK1, -2,
and -3 mAbs were derived from the same mouse, their amino acid
sequences in the VH and VL
CDR3 regions, which correspond to V(D)J junctions, were significantly
different. In addition, V and J gene usages in NOK1, -2, and -3 were
also distinct as follows: NOK1,
VH1/JH1 and
V10/J1;
NOK2, VH1/JH4 and
V1/J2; and NOK3,
VH1/JH1 and
V19/J1. These
differences suggest that NOK1, -2, and -3 were derived from distinct B
cell clones.
|
B). A recombinant humanized NOK2,
designated RNOK203, was purified from culture supernatants of CHO cells
transfected with both the humanized VH (1)
cDNA and VL () cDNA. As represented in Fig. 1C, RNOK203 neutralized FasL cytotoxicity more efficiently
than the original NOK2.
Some humanized or human-mouse chimera mAbs that can neutralize
proinflammatory cytokines such as TNF- have already been
successfully used in clinical trials (24, 25, 26). We
previously demonstrated the therapeutic effects of neutralizing
anti-FasL mAb in murine model of hepatitis and GVHD (6, 7). The humanized anti-FasL mAb (RNOK203) we generated here
may be useful for clinical application to human diseases.
Epitope mapping of anti-human FasL mAbs
We previously demonstrated that NOK1, -2, and -3 mAbs inhibited
the binding of soluble Fas-Ig to FasL transfectants (20).
Our preliminary epitope analysis by peptide mapping suggested that
these mAbs recognize amino acid residues present in the area from Y196
to A240 in human FasL (data not shown). Thus, to identify the amino
acid residues critically involved in shaping the NOK1, -2, and -3
epitopes, each amino acid residue from Y196 to A240 was subjected to
alanine-scanning mutagenesis. The mutant FasL molecules were
constructed as a FLAG-tagged soluble form. Forty-five expression
plasmids carrying native or mutant FasL cDNA were transiently
transfected into COS cells. Concentration of the secreted FasL proteins
was determined by ELISA using anti-FLAG mAb. Although most of the
mutant constructs as well as the native one led to the production of
>1 µg/ml of soluble FasL in the culture supernatant, the
alanine-substitutions at Y196, F197, C202, N204, P206, L207, H209,
V221, M224, M225, S231, Y232, C233, G236, W239, and A240 mostly
abolished the FasL production in the supernatant (Table I). It has previously been reported that
some amino acid substitutions in human FasL residues resulted in a
significant reduction of FasL protein production. For example, amino
acid replacements at N-glycosylation sites such as D260
significantly reduced FasL protein production (27).
Although we presently do not know precise reasons for the dramatic
reduction of FasL production by the alanine substitutions performed in
this study, these substitutions may affect the intracellular
trafficking or posttranscriptional modification of FasL protein.
Further studies are needed to address these possibilities.
|
, an alanine
substitution at Q220A, led to an almost complete loss of NOK2
reactivity. In addition to the alanine substitution at Q220A, the loss
of >75% reactivity with NOK2 was also found by alanine substitution
of R198, G199, L205, D221, and M238, whereas that of Y218, L222, G227,
M230, or Q237 led to 35–40% loss of NOK2 reactivity (Fig. 2 and
Table I). In a similar way, we also examined the binding of NOK1 and
NOK3 (Table I). As was the case with NOK2, widespread distributions of
amino acid residues with the loss of reactivity were also noted with
NOK1 and NOK3, suggesting that NOK1, -2, and -3 recognize
conformational epitopes. Comparison of each epitope for NOK1, -2, and
-3 revealed a largely overlapped distribution of the amino acid
residues with remarkable loss of reactivity. However, relative
importance of each residue in shaping the epitopes appeared to vary
among NOK1, -2, and -3. For example, the alanine-substitution of N203
and K228 significantly reduced FasL binding to NOK2 but not to NOK1 and
NOK3 (Table I). These results indicated that the epitopes for NOK1, -2,
and -3 are similar but distinct. Consistent with this notion, our
previous study showed that NOK1, -2, and -3 mAbs only partially
competed with each other for their binding to FasL
(20).
|
). This
difference may be responsible for the higher neutralizing activity of
RNOK203 than NOK2 observed in Fig. 1C. Amino acid residues important for FasL cytotoxicity
A previous study by Schneider et al. (27) using some
human FasL mutants demonstrated that the P206 and T218 residues in
human FasL were critically involved in the interaction with Fas
receptor. However, more details about the FasL/Fas interaction remain
to be determined. Thus, we next tested the cytotoxic activity of our
panel of FasL mutants. A mouse T lymphoma cell line-derived human Fas
cDNA transfectant, hFas/WR19L, which is highly sensitive to human
FasL-induced apoptosis, was used to evaluate the cytotoxic activity of
the mutants. As reported previously (27), we also found
that the alanine substitution at T218 led to an almost complete loss of
the cytotoxicity (Table I). In addition, the alanine substitution at
R198, G199, K210, Y212, N215, Y218, Q220, D221, L222, K228, M230, and
M238 also mostly abolished FasL cytotoxicity against hFas/WR19L
(Table I).
It has been known that the trimeric structure of TNF family ligands is
required for execution of biological activity. Indeed, a substitution
of F275 to L275 in human FasL, which mimics the loss-of-function
mutation found in gld mice, has been reported to lead to an
inadequate oligomerization of FasL molecule with significant loss
of cytotoxic activity (27, 28). Thus, we next examined
whether the alanine substitutions affected the homotrimeric nature of
FasL by Western blot analysis using anti-FLAG mAb for detection.
With a chemical cross-linker, DSS, which stabilizes the FasL trimer
complex, 90- and 60-kDa bands representing FasL trimer and dimer,
respectively, were observed with native FasL (Fig. 3). Similar levels of trimerization were
also observed with all the FasL mutants tested, as represented by
L222A, V223A, E226A, G227A, and K228A. Consistent with a previous
report (11), a 30-kDa band representing FasL monomer was
dominant without the DSS cross-linking (Fig. 3). In a similar way, we
also tested all the other FasL mutants for their trimer formation and
found no difference among native and mutant FasL (data not shown).
|
, all the alanine
substitutions with significant loss of cytotoxicity were accompanied
with >65% loss of Fas binding as compared with native FasL. These
results suggest that the reduced cytotoxicity of the FasL mutants was
primarily due to the reduced binding to Fas receptor rather than
inadequate trimer formation.
|
). However, some other residues such as K210 and N215 were
necessary for cytotoxicity but not for the NOK1, -2, and -3 epitopes.
As compared with the NOK1 epitope, the NOK2 and NOK3 epitopes exhibited
relatively higher similarity with the distribution of amino acid
residues crucial for the cytotoxicity (Table I). Molecular modeling of conformational epitopes for anti-FasL mAbs
As shown in Fig. 5A, the
amino acid residues subjected to alanine scan were predicted to locate
on the region from the middle part of the -strand C to the end of
the E-F loop in the FasL sequence. Although the TNF family members
exhibit only low sequence similarity, it is well known that their
three-dimensional structures are well conserved (12, 13, 14, 15, 16, 17, 18).
As represented in Fig. 5B, like other members of the TNF
family including TNF- (12, 13), LT-
(14), CD40L (15), and TRAIL
(16, 17, 18), FasL monomer is deduced to conform a strand
jellyroll topology composed of two flat sheets. In the ribbon
diagram, the region subjected to alanine scan was extended from the top
to bottom at one side of the FasL molecule (Fig. 5B). Then
we deduced conformational positions of these amino acid residues on
FasL by using a space-filling comparative molecular model of FasL in
trimer form. As represented in Fig. 6A, the residues subjected to
alanine scan are mainly located along a long groove at the FasL
monomer-monomer interface. However, some hydrophobic amino acid
residues such as Y196 and F197 were deduced not to exist on the surface
due to the conformational folding of sheets (Table I). As described
above, we revealed the wide-spread distribution of amino acid residues
critical for NOK1, -2, and -3 binding (Fig. 2 and Table I), and we next
predicted the conformational epitopes for NOK1, -2, and -3 on the
three-dimensional FasL model. As illustrated in Fig. 6, B–D, each conformational epitope for NOK1, -2, and -3 on
the FasL trimer model could be classified into two clusters; one
locates around N205 near the top, and another locates around Q220 and
D221 corresponding to the edge of -strand D (Fig. 5A).
This suggests that NOK1, -2 and -3 recognize a largely overlapped, huge
three-dimensional epitope composed of at least two spatially distant
clusters of amino acids. Consistent with this notion, a computed
molecular model of NOK2 V regions revealed that the Ag-recognition site
of NOK2 has the almost same size as a circle with a 17-Å radius that
can cover all of the amino acid residues on the upper and lower
clusters on the FasL trimer (data not shown). However, the reactivities
of these mAbs to these clusters, especially to the upper cluster,
appeared to be slightly different. For example, R198 and N203 in the
upper cluster were not required for NOK1 and NOK2 binding, respectively
(Table I and Fig. 6, B–D). This suggests that the upper
cluster might be mainly responsible for the different specificities of
these mAbs. As described above, neutralizing activity of RNOK203 was
significantly increased as compared with original NOK2, possibly due to
the increased reactivity to the amino acid residues at 212, 222, 228,
and 230 (Fig. 1C and Table I). The conformational positions
of these amino acid residues were deduced to be located in the upper
and lower clusters (Fig. 6E), suggesting that the
humanization of NOK2 increased the affinity for both the upper and
lower epitopes.
|
|
We further deduced conformational locations of the amino acid
residues important for FasL cytotoxicity. As illustrated in the space
full-filling molecular model of FasL (Fig. 6F), the amino
acid residues critical for FasL cytotoxicity are mostly located on the
groove at FasL/FasL interface. Like the epitope for NOK1, -2, and -3
and RNOK203, these residues could be classified into two clusters, one
around D221 near the bottom corresponding to the
D-E loop and strand E (Fig. 5, A
and B) and another around K238 near the top. This suggests
that the interaction with Fas takes place at two distinct sites on
FasL, the outer tip-forming D-E loop and the
region near the top of the FasL molecule. At these contact sites, one
FasL molecule binds two Fas molecules, which leads to the aggregation
of cytoplasmic death domains resulting in Fas-associated death domain
recruitment. NOK1, -2, and -3 and RNOK203 seem to interrupt the
FasL/Fas interaction by competing with Fas at both of these FasL/Fas
contact sites on FasL. However, the lower cluster around Q220 important
for FasL cytotoxicity was relatively lager than those for NOK1, -2, and
-3 and RNOK203 epitopes (Fig. 6, B–F). Although the upper
cluster crucial for cytotoxicity was relatively well overlapped with
those for NOK2 and NOK1 epitopes, the upper cluster for NOK1 epitope
appeared to be shifted more closely to the top of FasL. These results
suggest that NOK1, -2, and -3 and RNOK203 might be so bulky that these
mAbs can inhibit the binding of Fas by directly or indirectly masking
the Fas-binding sites on FasL.
Molecular modeling of FasL/Fas contact sites
Previous crystallographic studies of LT-/TNF-RI and
TRAIL/TRAILR2 complexes revealed that these ligand-receptor pairs
interact with each other at two sites (18, 19). Also in
our three-dimensional model of the FasL/Fas complex, the trimerization
of FasL monomer forms three clefts at the interfaces between FasL
monomers (Fig. 7, A and
B). It seems that three Fas molecules bind diagonally along
these clefts on the FasL trimer. Furthermore, one Fas molecule binds
two FasL molecules via two distinct contact sites on FasL monomer at
the outer tip-forming loop near the bottom and the side of cleft near
the top (Fig. 7, A and B).
|
C),
R198 on FasL -strand C (Fig. 5A) was predicted to form a
hydrogen bond with D92 on the second CRD of Fas. Furthermore, FasL G199
was expected to interact with Fas E93 and G94 via van der Waals force.
In the lower contact region, Y218 on the FasL D-E
loop that is conserved among death-inducing TNF family members (Fig. 5A) has been reported to be important for LT- and TNF-
binding to TNFRI (19, 29, 30). In addition, R86 on the
second CRD of Fas has been shown to be a critical residue for direct
interaction with FasL Y218 (31). In agreement with these
notions, FasL Y218 was predicted to directly bind to Fas R86 by forming
a hydrogen bond in our molecular model (Fig. 7D).
Furthermore, our model suggested that Fas R86 also interacts with FasL
Q220 and D221 on -strand E, forming a network of hydrogen bonds and
van der Waals contact. In conclusion, we revealed that the anti-human FasL mAbs inhibit the cytotoxic activity of FasL by masking both of two distinct Fas binding sites on FasL. Our molecular model of the FasL/Fas interaction would facilitate the drug design for interrupting the FasL/Fas interaction. In addition, the humanized NOK2 we generated in this study would be useful for clinical application to human diseases in which the FasL/Fas system plays a key role in the pathogenesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Hideo Yagita, Department of Immunology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan. E-mail address: hyagita{at}med.juntendo.ac.jp
3 Abbreviations used in this paper used in this paper: L, ligand; GVHD, graft-vs-host disease; LT, lymphotoxin; TRAIL, TNF-related apoptosis-inducing ligand; CRD, cysteine-rich domain; CDR, complementarity-determining region; FR, framework region; DSS, disuccnimidyl suberate; hFas, human Fas; CHO, Chinese hamster ovary.
Received for publication May 16, 2001. Accepted for publication July 5, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionReferences |
|---|
antibodies on acute graft-versus-host disease pathologies. Blood 91:4051. at 2.6-Å resolution: implications for receptor binding. J. Biol. Chem. 264:17595.) at 1.9-Å resolution. J. Biol. Chem. 267:2119. complex: implications for TNF receptor activation. Cell 73:431.[Medline]
. Arthritis Rheum. 36:1681.[Medline]
(cA2) versus placebo in rheumatoid arthritis. Lancet 344:1105.[Medline]
This article has been cited by other articles:
|
|
 |
|
  B. Poonia, M. S. Salvato, H. Yagita, T. Maeda, K. Okumura, and C. D. Pauza Treatment with anti-FasL antibody preserves memory lymphocytes and virus-specific cellular immunity in macaques challenged with simian immunodeficiency virus Blood, August 6, 2009; 114(6): 1196 - 1204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. O'Hara, A. J. Small, J. B. Nelson, A. D. Badley, X.-M. Chen, G. J. Gores, and N. F. LaRusso The Human Immunodeficiency Virus Type 1 Tat Protein Enhances Cryptosporidium parvum-Induced Apoptosis in Cholangiocytes via a Fas Ligand-Dependent Mechanism Infect. Immun., February 1, 2007; 75(2): 684 - 696. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Yao, W. Chen, H. Luo, Q. Jiang, Z. Xia, L. Zang, J. Zuo, X. Wei, Z. Chen, X. Shen, et al. Identification of core functional region of murine IL-4 using peptide phage display and molecular modeling Int. Immunol., January 1, 2006; 18(1): 19 - 29. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nikiforow, K. Bottomly, G. Miller, and C. Munz Cytolytic CD4+-T-Cell Clones Reactive to EBNA1 Inhibit Epstein-Barr Virus-Induced B-Cell Proliferation J. Virol., November 15, 2003; 77(22): 12088 - 12104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Dan, T. Kanai, T. Totsuka, R. Iiyama, M. Yamazaki, T. Sawada, T. Miyata, H. Yagita, K. Okumura, and M. Watanabe Ameliorating effect of anti-Fas ligand MAb on wasting disease in murine model of chronic colitis Am J Physiol Gastrointest Liver Physiol, October 1, 2003; 285(4): G754 - G760. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, 
, 
, and 
) FasL was
evaluated by ELISA. Culture medium was used as a negative control
(
). Data are shown as the means of triplicated wells. SDs were <5%
(data not shown). Similar results were obtained from three independent
experiments.
