| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Molecular Biology and
Lymphocyte Biology Sections, Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892;
Metabolism Branch and
§
Laboratory of Biochemistry, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892; and
¶
Howard Hughes Medical Institute, Washington University School of Medicine, St. Louis, MO 63110
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Crystallographic structures of several TCR/MHC-I complexes indicate a
common orientation of the 
ß TCR in interacting with amino acid
side chains of both the MHC-I molecule and the bound peptide
(22, 23, 24, 25, 26). In contrast, relatively little is known about
the structure of Ly49 receptors. Ly49A is a member of a family of
related proteins encoded by closely linked genes and by amino acid
sequence is distantly related to members of the C-type lectin family
(14, 27). Ab blocking studies and transfection of target
cells with MHC-I-encoding genes indicate that Ly49A is an inhibitory
receptor for the H-2Dd and
H-2Dk molecules (15, 28) and that
this inhibitory effect for H-2Dd results from
interactions with the
12 domains (11, 15). Other studies indicate that Ly49A binds
H-2Dd directly, although the precise location of
the Ly49A interaction with the MHC-I molecule is unclear
(29, 30, 31).
We previously described the expression and function in transfected cells of single-chain (Sc) forms of the MHC-I molecules, H-2Dd and HLA-A2, which consist of the MHC-I L chain, ß2-microglobulin (ß2m), covalently linked to the MHC-I H chain through a peptide spacer (32, 33, 34). Others have reported similar Sc constructs of H-2Kd (35) and HLA-A2 (36). Sc H-2Dd molecules were expressed on the surface of transfected cells as detected by flow cytometry and could stimulate T hybridoma cells (B4.2.3) specific for H-2Dd complexed with P18-I10, a decapeptide from the HIV-1 envelope glycoprotein V3 loop (37, 38) in vitro when loaded with peptide. Analysis of the fine specificity of Ag presentation by the Sc molecules revealed only subtle quantitative differences when a panel of synthetic substituted peptide Ags was employed, suggesting that there are no major differences in the way the Sc molecules bind peptide or in the conformation of the resulting MHC/peptide complex (39). These data indicated that the overall conformation of the Sc H-2Dd molecule was adequate for binding peptides and for stimulating H-2Dd-restricted T cells. Because these earlier experiments addressed issues of mature T cell recognition of MHC/peptide complexes, we expected that transgenic expression of the Sc MHC-I molecules would allow questions of T cell and NK cell development and education to be explored. Here, we examine the expression, both in B6 and in B6 ß2m-/- mice, of a Sc H-2Dd (ScDd) molecule and examine its recognition in both education and effector phases by T and NK cells. These results lead to conclusions about the nature of the site of TCR and NKR interaction with H-2Dd.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To obtain a chimeric gene encoding a Sc
(ß2m-spacer-MHC-I H chain)
H-2Dd, in a form suitable for expression as a
transgene, three different fragments from two plasmids were used: 1) a
5' fragment (XbaI-BamHI) from the
pDd-1 plasmid (40, 41) containing
promoter sequences from 
400 bp upstream of the initiation ATG codon,
modified in its BamHI site with Klenow polymerase and dGTP
and dATP; 2) a SalI-FspI fragment derived from a
cDNA for a Sc H-2Dd construct that codes for
ß2m, a peptide linker, and the mature
H-2Dd H chain protein through to the end of the
2 domain (encoded by exon 3), which had been
modified at the SalI site with Klenow polymerase and dTTP
and dCTP (32); and 3) an FspI-EcoRI
fragment (which extends from the end of exon 3 to the 3' untranslated
region) from pDd-1. These were ligated in two
sequential steps into the pBluescript-IIKS -(+) vector.
The DNA construct encoding the Sc H-2Dd molecule was injected into fertilized eggs of homozygous C57BL/6NCr mice to generate transgenic mice. Two transgenic founder mice were backcrossed to B6 mice that contained an induced defect in the expression of ß2m (C57BL/6GphTac-{Ko}B2m N5, ß2m knockout (ß2m-/-) mice (3), maintained and bred at Taconic, Germantown, NY) to generate transgenic mice expressing cell-surface Sc H-2Dd in the absence of other ß2m-dependent molecules. (The formal names for the transgenic strains are C57BL/6NCr tScß2mDd-1 and -2, and we will refer to them here as ScDd-1 and ScDd-2, respectively. When bred onto the ß2m-/- strain, we refer to these as ScDd-1 ß2m-/- and ScDd-2 ß2m-/- as well.) Offspring were screened for H-2 expression with Abs specific for H-2Dd (34-2-12), H-2Db (28-14-8), and H-2Kb (AF6-88.5) by flow cytometry. Most of the experiments reported here have been performed with ScDd-1, and some have been performed with ScDd-2. Surface expression of the transgene-encoded molecule and functional behavior were indistinguishable for the two strains.
Abs and mice
The following Abs, purchased from PharMingen, San Diego, CA,
were used: anti-CD16/CD32 (2.4G2); FITC-conjugated
anti-H-2Dd (34-5-8,
12 domain specific),
(34-2-12, which binds the 3 domain),
(34-4-20), (3-25.4); anti-H-2Kb (AF6-88.5);
anti-H-2Db (KH95); anti-CD3
(145-2C11); anti-CD8 (53-6.7); anti-Ly49A (A1); PE-conjugated
CD4 (H129.19); anti-TCR (TCR) V2 (B20.1); anti-V8
(B21.14); anti-V11 (RR8-1); anti-Vß2 (B20.6);
anti-Vß3 (KJ25); anti-Vß4 (KT4); anti-Vß5 (MR9-4);
anti-Vß6 (RR4-7); anti-Vß7 (TR310); anti-Vß8 (MR5-2);
anti-Vß9 (MR10-2); anti-Vß10 (B21.5); anti-Vß11
(RR3-15); anti-Vß13 (MR12-3); biotinylated anti-NK1.1
(PK136); anti-TCR V3.2 (RR3-16); anti-Vß12 (MR11-1);
anti-Vß14 (14-2); anti-Vß17 (KJ23); cychrome-conjugated
anti-TCR Cß (H57-597); and streptavidin-cychrome. FITC-conjugated
F(ab')2 goat anti-mouse IgG mAb was purchased
from Jackson ImmunoResearch (West Grove, PA). A1 (anti-Ly-49A),
SW5E6 (anti-Ly49C/I), and 4D11 (anti-Ly49G2) were also used.
34-5-8S (anti-12
of H-2Dd) and 34-4-20S
(anti-H-2Dd) were obtained from American Type
Culture Collection (Manassas, VA). Ab was purified from cell-culture
supernatant by protein A or protein G affinity chromatography and
conjugated to FITC or biotin by standard protocols.
C57BL/6 (B6), BALB/c, ß2m-/-, and D8 (H-2Dd transgenic B6 (42)) mice were obtained from the National Institute of Allergy and Infectious Diseases production facility at Taconic (Germantown, NY). B6.C-H2bm1/By and C57BL6/J-H2bm3/Eg, which carry the H-2Kbm1 and H-2Kbm3 mutations of H-2Kb mice, were purchased from The Jackson Laboratory (Bar Harbor, ME) and are referred to here as B6.bm1 and B6.bm3, respectively.
FACS analysis
The expression of the ScDd transgene and lack of normal level of H-2Db and H-2Kb were documented by direct immunofluorescence analyses using FITC-conjugated 34-2-12, 3-25.4 (anti-H-2Dd), KH95 (anti-H-2Db), and AF6-88.5 (anti-H-2Kb) mAbs and indirect staining using biotinylated 28-14-8 (anti-H-2Db) followed by PE-conjugated streptavidin. To block Ab binding to Fc receptors, all samples were pretreated with anti-CD16/CD32 mAb. Percentages of CD4+ and CD8+ T cells from lymph nodes and spleen were determined among the TCR-positive cells. To study expression of NKR on NK cells, spleen cells were treated with ACK lysis buffer (Biofluids, Rockville, MD), passed over nylon wool columns to remove T cells, and stained with PE-conjugated anti-NK1.1 and FITC-conjugated anti-Ly49A, C/I, or G2 mAbs.
Cell-surface biotinylation and immunoprecipitation
Cells were surface-biotinylated using NHS-LC-biotin (Pierce, Rockfiord, IL) as previously described (43). The cells were then solubilized with 1% Nonidet P-40 in 10 mM Tris-HCl, pH 7.2, 140 mM NaCl, 1 mM PMSF, 5 mM iodoacetamide, 1 mM sodium orthovanadate, and Complete protease inhibitor mixture (Boehringer Mannheim, Indianapolis, IN). Nuclei were removed by centrifugation, and lysates from 5 x 107 cells were precleared with protein G-Sepharose and then subjected to immunoprecipitation with 10 µg of 34-2-12 and 150 µl of a 10% slurry of protein G-Sepharose beads (Pharmacia Biotech, Uppsala, Sweeden). Beads were washed and boiled in 1% SDS and 10 mM iodoacetamide for 5 min, and eluted proteins were then separated on a 14% SDS-polyacrylamide gel and transferred to an Immobilon P membrane (Millipore, Bedford, MA). Biotinylated proteins were visualized with streptavidin-HRP (Zymed, South San Francisco, CA) and enhanced chemiluminescence (Amersham, Chicago, IL).
Immunization with vaccinia virus and CTL Assay
ScDd ß2m-/- and D8 mice were immunized by i.p. injection of 2 x 107 pfu of vaccinia virus expressing HIV-1 envelope glycoprotein gp160IIIB (vPE16, the gift of P. Earl and B. Moss) (44). Three weeks later, splenocytes were cultured at 5 x 106/ml in 24-well culture plates in complete T cell medium (RPMI 1640 containing 10% FBS, 2 mM L-glutamine, penicillin (100 U/ml), streptomycin (100 mg/ml), and 50 µM 2-ME). Three days later, the cultures were supplemented with one-tenth volume of T-Stim (Collaborative Biomedical Products, Bedford, MA) as a source of IL-2. Spleen cells were stimulated in vitro for 7 days with 1 µM P18IIIB-I10 (RGPGRAFVTI) peptide together with 4 x 106 irradiated (3300 rad) syngeneic spleen cells as APC. Cytolytic activity of CTL lines was measured by a standard 4-h 51Cr-release assay (45). SEs of the mean of triplicate cultures were all <5% of the mean.
Generation of MLR and NK cell effector population
Primary mixed lymphocyte cultures were established essentially as described previously (46). Briefly, 2.5 x 107 responder splenocytes and 2.5 x 107 irradiated stimulator cells (3000 rad from a 137Cs source) were cultured together in 20 ml of complete T media with 5% FCS in upright T-25 flasks (Corning Glass Works, Corning, NY) for 5 days. Fresh NK effector cells were depleted of RBC, and nylon wool nonadherent splenocytes were taken from groups of mice that had been treated 20–24 h previously with 150 µg poly I:C (Sigma, St. Louis, MO), an NK stimulator. Four-day lymphokine-activated killer (LAK) NK effector cells were prepared by a procedure based upon that of Chadwick and Miller (47). Briefly, splenocytes were depleted of RBC by hypotonic lysis and passed over nylon wool. Nylon wool nonadherent cells were cultured for 4 days in RPMI 1640 plus 10% FCS, supplements (including 50 µM 2-ME), and 400 ng/ml recombinant human IL-2 (Chiron, Emeryville, CA). Adherent LAK (A-LAK) cells were prepared as described (15).
Lysis assays for NK cells and allogeneic CTL (MLR)
For target cell preparation, splenocytes were cultured in
complete T medium containing 5% FCS for 24–30 h in 24-well plates
(Falcon Plastics, Lincoln Park, NJ) at 2 x
106 cells/ml with 2 µg/ml Con A (Sigma). On the
day of assay, one-tenth volume of 1 M
methyl--D-mannopyranoside (Sigma) in RPMI 1640 or PBS
was added to target cell cultures to block Con A sites before labeling
for 1–2 h in 100 µl of 10 mCi/ml
[51Cr]Na2CrO4
(Amersham, Arlington Heights, IL) in PBS. All points were determined in
triplicate using 5 x 103 to 1 x
104 target cells per well.
Bone marrow transplantation
Bone marrow cells were obtained by flushing the lumina of the tibiae and femora of donor mice with HBSS under aseptic conditions. Suspensions of 3 x 106 cells were injected i.v. (subocular under anesthesia or by tail vein) into groups of six (or occasionally five) irradiated (900 rad from a 137Cs source) hosts. Five days after inoculation with bone marrow cells, mice were injected i.p. with 3 mCi of [125I] iododeoxyuridine (125IUdR; Amersham). Then, 20–24 h after injection of radiolabel, the mice were sacrificed and incorporation of radioactivity into the spleens was determined by gamma spectroscopy.
Biotinylation of Ly49A protein
The extracellular portion of Ly49A (amino acids 67 to 262) was expressed in bacteria as inclusion bodies, solubilized, refolded, and purified as described in detail elsewhere.5 The purified protein was chemically biotinylated with NHS-LC-biotin (Pierce) and further purified. Lymph node cells were stained for flow cytometry analysis using the biotinylated Ly49A and PE-streptavidin.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To explore the expression of the ScDd and
other MHC-I molecules in
ß2m-/- transgenic mice,
lymph node cells from ScDd
ß2m-/-,
ß2m-/-, BALB/c, and B6
mice were stained with mAb against H-2Dd,
H-2Db, and H-2Kb. As
expected, lymph node cells of the
ß2m-/- transgenic mice
expressed no dectectable H-2Kb or
H-2Db above the level observed in
ß2m-/- nontransgenic
cells (Fig. 1
A). In contrast,
mice homozygous for the ScDd transgene, expressed
in the ß2m-/-
background, revealed apparently normal levels of
H-2Dd epitopes as indicated by reactivity with
Abs 34-2-12 (3 domain), and 3-25.4
(12 domain) (Fig. 1A) as well as with 34-5-8
(12 domain) and
34-4-20 (data not shown).
|
B,
we detected low levels of H-2Db in both
ß2m-/- and in
ScDd
ß2m-/- mice expressing
the ScDd transgene. Presence of the
ScDd transgene did not perturb the expression of
H-2Db in
ß2m-/- mice (Fig. 1B).
With biochemical techniques we examined the Sc molecules made in these
transgenics. Although we expected that the Sc
H-2Dd molecules expressed at the cell surface
were the basis of the serological reactivity, it was possible that the
Sc molecules were proteolytically cleaved in the spacer region, leading
to cell surface H-2Dd/ß2m
heterodimers. H-2Dd molecules expressed on the
surface of cells of the transgenic mice were exclusively detected as
molecules with a molecular mass of 60–65 kDa as determined by
immunoprecipitation from splenocytes using 34-2-12 mAb
(anti-3 domain of
H-2Dd) (Fig. 1C). In contrast, the
normal MHC-I H-2Dd H chain on the spleen cells
from BALB/c was detected as a 50-kDa protein. Thus, the H chain of
transgenic ScDd is expressed as a molecular
species that is covalently linked with ß2m on
the cell surface, and there is no evidence for proteolytic cleavage of
the Sc molecule. (Because the 34-2-12 Ab used for the
immunoprecipitation is 3 domain specific and
reacts with the isolated H-2Dd
3 domain (51), this Ab would have
detected membrane molecules truncated in the
ß2m, spacer, 1, or
2 regions.)
ScDd educates a diverse repertoire of CD8+ T cells, similar to that induced by native H-2Dd
It is well known that normal expression of MHC-I molecules in the
thymus is necessary for the normal maturation of
CD8+ T cells, that is, the progression of
immature CD4+CD8+ cells to
mature single positive cells. Animals defective in
ß2m expression, and as a result lacking normal
cell-surface MHC-I expression, show a profound decrease in the number
of CD8+ (single positive) cells in the thymus as
well as in lymph node and spleen (3, 4). In addition,
animals defective in TAP expression, and thus deficient in the delivery
of self-peptides to MHC-I, show similarly abnormal
CD8+ T cell development (52). To
assess the selection of CD8+ T cells by
ScDd in the absence of normal expression of other
MHC-I molecules, we analyzed thymocytes, lymph node cells, and
splenocytes from B6, B6
ß2m-/-, and
ScDd
ß2m-/- mice for the
presence of mature CD8+ T cells by flow cytometry
(Fig. 2). Unlike nontransgenic
ß2m-/- thymocytes,
those from ScDd
ß2m-/- mice contained
mature CD8+ cells in numbers similar to those
seen in B6 thymuses (Fig. 2). As reported by others, normal B6 animals
showed a significant proportion of CD4 and CD8 single positive cells in
the thymus (10.65 and 1.70%, respectively), lymph node (62.39 and
36.32% of TCR Cß+ cells), and spleen (65.39
and 32.29% of TCR Cß+). In contrast,
thymocytes from ß2m-/-
animals showed a profound decrease in the proportion of CD8 single
positive cells in each tissue (0.01, 0.21, and 0.14% in thymus, lymph
node, and spleen, respectively) and a compensatory increase in the
proportion of CD4 single positive cells in the peripheral lymph nodes
and spleen. There was no apparent change in total number of
CD4+ cells. Despite the lack of proper expression
of MHC-I molecules other than the ScDd in the
transgenic ß2m-/-
animals, substantial numbers of CD8+ T cells were
detected in the thymus, lymph nodes, and spleen of
ScDd
ß2m-/- mice (1.70,
33.69, and 31.63% of total, respectively), indicating their normal
migration to peripheral lymphoid organs (Fig. 2).
|
and
Vß TCR repertoires of CD8+ T cells using
available mAbs. CD8+ T cells in lymph nodes from
ScDd
ß2m-/- exhibited a
diverse repertoire of TCR, indicating that these
CD8+ T cells do not represent expansion of a
small, oligoclonal population (Fig. 3A). Quantitative differences
in the level of expression of different TCR V region genes might reveal
subtle differences in the TCR repertoire selected by the
ScDd on the
ß2m-/- background as
compared with those repertoires selected by either B6 or B10.D2.
|
B). T
cells that develop in an environment in which their TCR are selected by
the ScDd molecule but retain discriminating
activity against the native two-chain H-2Dd
molecule would be expected to be present and be stimulated in the
ScDd anti-D8 MLR and should be capable of
killing D8 targets. Similarly, T cells expressing TCR that are selected
on H-2Dd yet remain reactive against unique
epitopes of ScDd would be expected to be
stimulated in the D8 anti-ScDd B6 MLR and
show cytolysis against ScDd B6 targets. A low
level of lysis of target cells of the stimulator mouse strain indicates
that few such cells are present and/or they are not particularly
reactive. CTL stimulated in both the D8
anti-ScDd B6 MLR and the
ScDd B6 anti-D8 MLR exhibited very low
cytolytic activity against target cells of the stimulator type (Fig. 3B). This was not due to a general inability of these mice
to respond to any MHC-I difference, as B6.bm3 stimulators elicited
substantial activity from the same responders against
H-2bm3 target cells (Fig. 3B).
H-2bm1 stimulators also elicited substantial
activity from the same responders (data not shown). This also was not
due to the inability of H-2Dd or
ScDd proteins to stimulate responses generally,
as B6 cells transgenic for either the native two chain or the
ScDd elicited substantial CTL activity from
nontransgenic B6 splenocytes (Fig. 3C). Furthermore, the
anti-H-2Dd alloreactive CTL raised against
either the Sc or the native molecule do not effectively distinguish
native H-2Dd from the Sc. Taken together, the
serological and functional data indicate that
ScDd participates in positive selection of
CD8+ T cells with a broad repertoire and that
this repertoire is functionally similar to that induced by the native
H-2Dd. ScH-2Dd mice can induce a virus-specific H-2Dd-restricted CD8+ CTL response
The generation of a diverse repertoire of
CD8+ T cells indicated that the
ScDd molecule was effective in the presentation
of self-Ag (presumably as peptides) to developing thymic cells. In
addition, the ability of cells expressing ScDd to
elicit allospecific CTL from B6 T cells indicated that the molecule is
effective in Ag presentation to mature T cells. We also wished to
evaluate the ability of the ScDd molecules to
effectively participate in an immune response by presenting foreign
Ags, such as those generated by a viral infection, to specific
CD8+ T cells. ScDd
ß2m-/- and D8 mice were
infected with recombinant vaccinia virus directing the expression of
the HIV-1 gp160 envelope glycoprotein (vPE16). Three weeks following
infection, spleen cells were restimulated in vitro by syngeneic spleen
cells loaded with gp160-derived H-2Dd-restricted
peptide, P18-I10 (37), and then were tested for effector
function in a 51Cr-release assay. CTL induced in
ScDd
ß2m-/- mice could kill
P18-I10-loaded P815 cells expressing wild-type
H-2Dd on the cell surface, whereas these cells
could not lyse P815 target cells that had not received the sensitizing
peptide (Fig. 3D). ScDd
ß2m-/- mice developed a
P18-I10 specific, H-2Dd-restricted
CD8+ CTL response that was comparable to that of
D8 mice in extent of cytolysis. Furthermore, these cells were able to
recognize peptide-loaded native H-2Dd
efficiently. Such CD8+ CTL had been generated in
vivo with the peptide produced endogenously from the full-length gp160,
and these T cells had not been exposed to native
H-2Dd-expressing APC before the cytolysis assay.
These findings clearly demonstrate that CD8+ T
cells selected by peptide in ScDd
ß2m-/- mice are capable
of mounting an effective peptide-specific anti-viral
cytotoxic response. Most importantly, CTL induced by
ScDd expressed on APC of the transgenic mice
interact effectively with native H-2Dd/P18-I10
complexes, that is, they are unable to distinguish native
H-2Dd from ScDd.
ScDd is defective in the education of NK cells
Native H-2Dd interacts with some of the Ly49 receptors on NK cells, in particular Ly49A (15) and Ly49G2 (16), and plays a critical role in the function and development of NK cells as well as CD8+ T cells (14). To explore the effects of the ScDd transgene in the development of NK cells, we tested whether the presence of ScDd could alter NK cell development in transgenic mice. Expression of native H-2Dd is sufficient to alter the NK cell specificity: H-2Dd, when expressed as a transgene in B6 mice, confers upon the NK cells in these mice the ability to reject B6 bone marrow grafts in vivo and the ability to lyse B6 target cells in vitro (53, 54). Additionally, expression of MHC-I molecules in normal mice is sufficient to permit development of NK cells capable of rejecting ß2m-/- bone marrow in vivo and killing ß2m-/- target cells in vitro (2, 55).
Surprisingly, we found that expression of the
ScDd transgene in
ß2m-/- mice did not
confer any similar function upon the NK cells in these mice. Poly
I:C-stimulated NK cells from ScDd
ß2m-/- mice could not
kill ß2m-/- Con A
blasts as measured in cytotoxicity assays (Fig. 4A), and
ScDd
ß2m-/- mice accepted
ß2m-/--derived bone
marrow donor cells (Fig. 4B). NK cells raised in short-term
in vitro cultures supplemented with IL-2 show similar cytotoxicity
patterns (data not shown). Thus, NK cells of ScDd
ß2m-/- mice are
defective in either their development and/or their effector function.
To determine whether NK cells of ScDd
ß2m-/- mice were
properly educated by ScDd in vivo, we studied
expression both of the level and the percentage of Ly49 receptors on NK
cells in ScDd
ß2m-/- mice. The
proportion of NK cells expressing receptors of the Ly49 family would be
expected to be reduced in the presence of an MHC-I ligand, and each NK
cell would also be expected to express a lower level of the given
receptor (18, 19, 21, 56). The level of expression and the
percentage of NK cells expressing Ly49A, Ly49C/I, and Ly49G2 on NK
cells from ScDd
ß2m-/- mice were
indistinguishable from those of NK cells from
ß2m-/- mice (Table I). This contrasts with the surface
expression of NKR of D8 mice that have significantly fewer Ly49A- and
Ly49G2-expressing NK cells. These cells also exhibit a lower cell
surface density of Ly49A when compared with B6 animals. Similarly,
ScDd B6 mice are not significantly different from
the B6 parental line. These results further support the conclusion that
NK cells of ScDd
ß2m-/- and of
ScDd B6 mice could not be properly educated by
ScDd in vivo during their development.
|
|
A). Furthermore,
ScDd B6 hosts could not reject B6 bone marrow
grafts (Fig. 4C). This result is particularly striking
because animals expressing a native H-2Dd
transgene are capable of rejecting B6 bone marrow grafts. Thus, the
ScDd transgene is unable to function like native
H-2Dd in the education of NK cells.
|
The MHC-I-dependent resistance to NK lysis is mediated by
inhibitory surface NKR that engage target cell MHC-I (57).
The native H-2Dd molecule interacts with at least
Ly49A and Ly49G2 receptors on NK cells and thus is capable of
delivering inhibitory signals to NK cells, preventing or reducing lysis
(15, 16, 28, 29). We wished to understand whether the
expression of the ScDd transgene-encoded
molecules on the target cells resulted in the delivery of inhibitory
signals to NK effector cells. In in vitro cytotoxicity assays,
short-term cultured NK cells derived from B10.D2 (data not shown) and
D8 mice killed ScDd B6 and nontransgenic B6
target cells equivalently (Fig. 5A) and also lysed
ScDd
ß2m-/- target cells and
ß2m-/- target cells
equivalently. Expression of ScDd on donor bone
marrow cells used in grafts was insufficient to prevent rejection of
either ß2m-/- (data not
shown) or B6 bone marrow by B10.D2 hosts (Fig. 5B). Although
it is unclear whether expression of only native
H-2Dd on
ß2m-/- bone marrow
grafts would be sufficient to prevent rejection by B10.D2 mice,
expression of native H-2Dd as a transgene was
sufficient to prevent rejection of B6 marrow by B10.D2 mice (Fig. 5B), as previously shown by Öhlén and colleagues
(53). Taken together, these data indicate that the
ScDd transgene, in contrast to the
H-2Dd transgene of D8 animals, has little or no
function with respect to NK cell inhibition.
The rejection of D8 bone marrow by B6 is dependent on an
NK1.1+ cell population, and is believed to be due
to engagement of H-2Dd-specific activation
receptors on B6 NK cells (53). Although the mechanism of
this phenomenon is not completely clear, it may involve recognition by
stimulatory receptors that lack cytoplasmic immunoreceptor
tyrosine-based inhibitory motifs, such as Ly49D and Ly49H (58, 59). Here we show that B6 mice are unable to reject
ScDd B6 bone marrow grafts and confirm the
ability of B6 mice to reject D8 grafts (Fig. 5C). (Because
the rejection of bone marrow grafts in this experimental system is due
to NK and not CTL activity, allospecific rejection is not observed.) If
ScDd B6 bone marrow were capable of eliciting a
weak rejection response, one would expect pretreatment of the B6 host
with Ab to NK1.1+ cells to reveal a higher level
of incorporation of [125I]iododeoxyuridine into
the grafted spleens. This was not detected (Fig. 5C), and
the results indicate that the ScDd molecule fails
to elicit any stimulatory response from NK cells that may be produced
by native H-2Dd.
Sc H-2Dd are defective in binding Ly49 receptors on NK cells
To explore the mechanism for the functional and developmental
deficiency of ScDd-expressing NK cells, we
investigated the ability of ScDd to interact
directly with Ly49 receptors known to bind native
H-2Dd. A soluble biotinylated recombinant form of
the Ly49A extracellular domain was employed to probe for binding to
lymph node cells expressing various types of MHC-I, and binding was
detected by flow cytometry (Fig. 6).
Biotinylated Ly49A stained lymph node T cells expressing native
H-2Dd (BALB/c, B10.D2, and D8) but did not stain
lymph node T cells from B6 or
ß2m-/- and
ScDd
ß2m-/- mice. This
reagent also failed to stain cells from ScDd
transgenic mice on either the B6 or
ß2m-/- background. Thus
in a direct assay based on the physical interaction of the recombinant
Ly49A with H-2Dd, the ScDd
molecule fails to bind Ly49A.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A recent report described the expression of a similar Sc human HLA-A2.1
at normal levels on cells in a transgenic mouse (36). In
those experiments, Sc constructs based on either murine or human
ß2m were expressed in
ß2m-/- or double
ß2m-/-
H-2Db-/- animals. This transgene also restored
a sizable T cell population of functional CD8+
cells when expressed in such MHC-I-deficient mice. As shown in our
experiments, transgenic ScDd restored a
significant number of CD8+ T cells expressing a
broad TCR repertoire in the absence of proper expression of other
MHC-Ia and MHC-Ib molecules. Furthermore, these
CD8+ T cells could mount a response against
specific endogenously processed peptide/MHC complexes in the periphery
and efficiently recognized specific peptide presented by native
H-2Dd. Although the proportion of
CD8+ T cells restored by
ScDd and Sc HLA-A2.1 molecules were quite
different from each other, the strategy using transgenic Sc molecules
was very effective for education of CD8+ T cells
in both cases. These results confirm that both the peptide binding site
and the region of the MHC-I needed for interaction with TCR are
conserved even after the structural modification generating Sc
molecules. The similarity of TCR repertoires generated by native and
ScDd is demonstrated by the virtual lack of
alloreactivity of T cells from D8 for stimulators from
ScDd B6 and of T cells stimulated in the reverse
MLR (Fig. 3).
It is valuable to think about our experiments along with those using transgenic mice that express MHC-II molecules covalently linked to a single peptide (60, 61) and those that, as a result of the MHC-II processing defect caused by H-2M deficiency, express MHC-II molecules predominantly in complex with the class II-associated invariant chain peptide (62, 63, 64). These experiments suggest that the normal diversity of self-peptides is not critical for positive selection of a broad TCR repertoire. However, the H-2M-/- animals, which express normal levels of MHC-II, mount a response against cells bearing normal MHC-II, indicative of the differences between their repertoires. The single peptide MHC-II transgenic mice show a similar response to parental MHC-II, but disparities in the level of expression make these experiments more difficult to interpret. Thus, it appears that in the MHC-II-restricted examples discussed above, there are significant differences in the full MHC-II/peptide repertoire of the mutant animals as compared with the parental strains, and the resulting TCR repertoires are quite distinct. In our studies, the ScDd functions well to positively select a broad TCR repertoire, a repertoire that shows little reactivity against native H-2Dd. Thus, the ScDd animals select a repertoire very similar to that selected by the native molecule. We conclude that the ScDd functions properly with respect to presentation of most self-peptides presented by native H-2Dd, presumably because the conformation of the peptide groove of ScDd is well conserved and ScDd present a broad array and an appropriate distribution of MHC/peptide complexes to TCR of developing thymocytes. The ability of ScDd to present peptides derived from the endogeneous processing pathway is consistent with our earlier studies examining the presentation of synthetic peptides by transfected cells expressing ScDd (33). In addition, we showed that the sequence motif of peptides eluted from ScDd molecules was the same as that of peptides derived from native H-2Dd (65).
In contrast to its ability to serve both in the education of and the
target cell recognition by CD8+ T cells, the
transgenic encoded ScDd did not function with
respect to the education of NK cells or inhibition of NK-mediated
cytoxicity in a variety of assays. ScDd
ß2m-/--derived NK cells
did not reject ß2m-/-
bone marrow grafts in vivo, nor did they kill
ß2m-/- target cells in
vitro (Figs. 4 and 5), indicating that the transgenic
ScDd protein was unable to alter the MHC-I
reactivity of NK cells (i.e., "educate" the NK cells). Consistent
with the failure of ScDd to educate NK cells, the
pattern of expression of Ly49 receptors on NK cells in
ScDd
ß2m-/- mice was
indistinguishable from that of
ß2m-/- mice (Table I).
Furthermore, expression of the ScDd protein on
either B6 or ß2m-/-
bone marrow grafts was unable to prevent their rejection from B10.D2
(H-2d) mice and unable to reverse their
sensitivity to cytolysis in vitro. This contrasts strikingly with the
function of the native H-2Dd expressed as a
transgene in B6 mice, which rescues the graft from bone marrow
rejection.
Recently, a ligand on NK cells for nonclassical MHC-I molecules has been identified. A cell-surface heterodimer consisting of NKG2 and CD94 recognizes HLA-E in humans (66, 67) and Qa-1 in mice (68). Furthermore, a significant proportion of murine NK cells express a receptor that binds to soluble Qa-1 tetramers (69). Because ScDd ß2m-/- mice cannot properly express MHC-Ib molecules, we examined whether ß2m-dependent MHC-Ib molecules may be critical in education of NK cells in vivo. It was also possible that the expression of only one MHC-Ia molecule might not be sufficient to educate functional NK cells. Thus, additional expression of other classical MHC-I molecules may be required to restore NK cell function from the state found in ß2m-/- mice. To explore these possibilities further, we performed bone marrow rejection assays and in vitro cytolysis assays using ScDd B6 mice expressing ß2m normally on the cell surface. However, ScDd could not restore the education and function of NK cells even in a ß2m+ environment. This conclusion is strengthened by the observation that expression of native H-2Dd as a transgene (in D8) is sufficient to induce all of the NK activities that we assayed and failed to find in ScDd B6 mice or cells. These results are all consistent with the view that ScDd molecules cannot interact properly with any H-2Dd-specific inhibitory NKR and thus that the expression of this molecule is simply not sensed effectively by NK cells.
The failure of biotinylated Ly49A protein to stain cells from the
ScDd transgenic animals (Fig. 6) as well as the
failure of ScDd on targets to inhibit killing by
sorted-Ly49A+G2- NK cells
(data not shown), indicates that structural alterations of the
ScDd protein prevent it from interacting
effectively with NK inhibitory receptors. Using chimeric
H-2Kd/H-2Dd molecules,
Matsumoto et al. showed that residues 53–65 of the
1 domain and 90–107 in the N-terminal part of
the 2 domain of H-2Dd
contributed to Ly49A recognition (70). It is possible that
these differences indirectly affect NKR binding by influencing the
conformation or selection of bound peptides. The specificity of Ly49A
for different MHC-I molecules may result from polymorphic residues
between reactive and nonreactive MHC-I alleles and/or nonpolymorphic
residues having different side chain conformations in different MHC-I
molecules (14, 71). Other mutagenesis studies using
cultured cell lines transfected with in vitro mutated
H-2Dd molecules suggest a role of specific
residues in the 1 and
2 domains in Ly49A recognition
(72).
Unlike the Ly49A/H-2Dd interaction, very little is known about the interaction of H-2Dd with Ly49G2. The expression level of Ly49G2 on NK cells is not perturbed by expression of the ScDd transgene. This result stands in contrast to the observed changes in the expression of the Ly49G2 receptor among NK cells when expressed in the presence of native H-2Dd (19, 21). Although this is not a direct indication, the fact that we fail to observe a change in Ly49G2 expression in the ScDd transgenic mouse suggests that Ly49G2 also fails to effectively interact with the ScDd molecule. Furthermore, Ly49A-G2+ A-LAK cells were not inhibited in their cytolysis of ScDd expressing target cells. Finally, Ly49G2 is expressed on approximately half of NK cells in various mouse strains, and an effective inhibitory interaction of the ScDd protein with Ly49G2 could be expected to alter NK function in bulk NK populations, which was not observed. Thus, we conclude that Ly49G2 also cannot interact effectively with the ScDd protein.
Although the mechanism of the failure in physical interaction between
Ly49A or Ly49G2 and ScDd molecules is unclear,
this may result from: 1) direct blocking of Ly49A binding by the
covalent peptide spacer linking the C terminus of
ß2m and the N terminus of the
H-2Dd H chain in this construct; 2) a
conformational change of the Ly49A binding site induced by this
covalent link; or 3) the formation of an obligate
ScDd/ScDd noncovalent dimer
on the cell surface due to "domain swapping" of the tethered
ß2m whereby ß2m
covalently linked to one molecule binds to the other, resulting in the
sequestration of the Ly49A binding site in the interface between the
two heterodimers. (Sc Ab Fv are known to form either dimers
("diabodies") (73, 74) or trimers depending on the
length of the peptide spacer that joins VH and
VL (75).) Because our data show no
evidence of function with respect to the entire population of NK cells
of the mouse, we conclude that this lack of interaction must apply to
most or all of the H-2Dd-specific inhibitory NKR,
including Ly49A and Ly49G2. In addition, we provide evidence that the
lack of interaction extends to H-2Dd-specific
stimulatory receptors as well (see Fig. 5C). We thus
conclude that some structure that is necessary for interaction with
most or all H-2Dd-specific NK activating and
inhibitory receptors is not present in the ScDd
protein, while the structures required for effective interaction with a
broad range of H-2Dd-specific TCR remain intact.
Because the site of interaction of MHC-I/peptide complexes with TCR has
been shown to consist of the 1 and
2 helices of the MHC and of exposed side
chains of the bound peptide, (22, 23, 24, 25) it seems likely that
this structural surface is not significantly distorted in the
ScDd molecule. Clearly, the general lack of
peptide specificity of Ly49A interaction with
H-2Dd as well as the profound difference in the
reactivity of ScDd with Ly49A both functionally
and in a direct binding assay using the recombinant Ly49A indicate that
the TCR and the NKR interact with distinct sites on
H-2Dd. In addition, these results support the
view that several different activating and inhibitory receptors of the
same structural family employ a structurally conserved surface of the
molecule to interact with H-2Dd.
In conclusion, transgenic Sc H-2Dd can induce development and function of a large number of mature functional CD8+ T cells expressing a broad TCR repertoire highly similar to the repertoire selected by native H-2Dd but cannot function in the education of NK cells or interact with H-2Dd-specific inhibitory receptors expressed by NK cells. Thus, we demonstrate that TCRs bind a part of the MHC/peptide complex distinct from the site where Ly49A and Ly49G2 bind.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: National Institute of Standards and Technology, Gaithersburg, MD 20899.
3 Address correspondence and reprint requests to Dr. David H. Margulies, Molecular Biology Section, Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, Building 10, Room 11N311, National Institutes of Health, Bethesda, MD 20892-1892. E-mail address:
4 Abbreviations used in this paper: MHC-I, MHC class I; NKR, NK receptor; Sc, single chain; ß2m, ß2-microglobulin; LAK, lyphokine-activated killer; A-LAK, adherent LAK.
5 Natarajan et al. Submitted for publication.
Received for publication May 4, 1999. Accepted for publication July 21, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß T cell receptor structure at 2.5 Å and its orientation in the TCR-MHC complex. Science 274:209.2 helix of an MHC class I molecule by a valine to proline mutation in position 165 does not prevent peptide-specific antigen presentation. J. Immunol. 159:4408.[Abstract]
host
combinations as described in Materials and Methods
(B and C). ScDd
ß2m-/- mice do not reject bone marrow cells
from ß2m-/- (B), nor do
ScDd B6 animals reject grafts from B6 (C).
Each group of recipients contains five or six mice in the bone marrow
transplantation. The results in B and C
each show two independent bone marrow transplantation
experiments.
