The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chiang, E. Y.
Right arrow Articles by Stroynowski, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chiang, E. Y.
Right arrow Articles by Stroynowski, I.
The Journal of Immunology, 2006, 177: 2123-2130.
Copyright © 2006 by The American Association of Immunologists

The Role of Structurally Conserved Class I MHC in Tumor Rejection: Contribution of the Q8 Locus1

Eugene Y. Chiang and Iwona Stroynowski2

Center for Immunology, Department of Microbiology and Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX 75390


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
The mouse multimember family of Qa-2 oligomorphic class I MHC genes is continuously undergoing duplications and deletions that alter the number of the two "prototype" Qa-2 sequences, Q8 and Q9. The frequent recombination events within the Q region lead to strain-specific modulation of the cumulative Qa-2 expression levels. Q9 protects C57BL/6 hosts from multiple disparate tumors and functions as a major CTL restriction element for shared tumor-associated Ags. We have now analyzed functional and structural properties of Q8, a class I MHC that differs significantly from Q9 in the peptide-binding, CTL-interacting {alpha}1 and {alpha}2 regions. Unexpectedly, we find that the extracellular domains of Q8 and Q9 act similarly during primary and secondary rejection of tumors, are recognized by cross-reactive antitumor CTL, have overlapping peptide-binding motifs, and are both assembled via the transporter associated with the Ag processing pathway. These findings suggest that shared Ag-presenting functions of the "odd" and "even" Qa-2 loci may contribute to the selective pressures shaping the haplotype-dependent quantitative variation of Qa-2 protein expression.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Recent studies in our laboratory demonstrated that Q9, a member of the Qa-2 family of Q region-encoded nonpolymorphic class Ib MHC Ags, orchestrates strong antitumor immune responses against several syngeneic tumors, including the B16 melanoma variant B78H1, 3LL Lewis lung carcinoma, and RMA T cell lymphoma (1, 2, 3). Transcription of the Q9 gene is generally silenced or significantly down-regulated in tumors arising in vivo in Qa-2-positive H2b or H2d haplotype mice (4). When surface expression of Q9 is restored by transfection, the tumors are rejected. Furthermore, mice surviving the initial tumor immunization acquire immunologic memory and are protected from secondary challenges with less immunogenic variants of the same tumor or other Q9-bearing malignancies. This "cross-protection" is dependent on CD8+ T cells. Accordingly, we proposed that Q9 functions as a CTL restriction element for shared tumor Ags during rejection of multiple disparate tumors. This case is, to the best of our knowledge, the only one reported in which nonclassical class Ib MHC promotes antitumor CTL in a syngeneic, rather than allogeneic (5), mouse model.

In addition to Q9, the Qa-2 family is comprised of several other members (6, 7, 8). Each mouse haplotype examined thus far encodes a different constellation of Qa-2 loci, suggesting that this segment of MHC is continuously being reshaped in response to selective pressures and/or inherent susceptibility to chromosomal rearrangements. In C57BL/6 and C57BL/10 mice, the Qa-2 subregion is composed of a set of "odd" Q loci, comprised of Q5, Q7, and Q9 genes, and a set of "even" Q loci, comprised of Q4, Q6, and Q8 genes (6). The odd/even Qa-2 gene clusters were proposed to have evolved by sequential duplication of a primordial Q gene pair consisting of one odd and one even locus. Accordingly, the odd genes are more similar to each other than they are similar to even genes and vice versa. The Q6-Q7 and Q8-Q9 gene pairs are nearly perfect duplicates of each other, whereas the Q4-Q5 pair is more divergent and will not be considered in this study. The translated protein products of Q7 and Q9 (referred to as Q9) differ from products of Q6 and Q8 (referred to as Q8) substantially, with over 20 of the predicted amino acid substitutions localizing to the {alpha}1 and {alpha}2 domains where they might affect peptide binding and/or T cell recognition. Earlier studies demonstrated that allogeneic CTL against odd and even Qa-2 products are selectively specific for Q8 and Q9 proteins and show no mutual cross-reactivity (9). We wondered how this protein divergence impacts on the function of the Qa-2 family and whether Q8 class Ib molecules contribute to antitumor immune responses. We have previously shown that Q8 is ubiquitously expressed in normal mouse tissues, like Q9. Interestingly, transcription of Q8 genes is frequently reduced in tumors and this down-regulation may occur independently of the transcriptional silencing of Q9 genes (4).

We demonstrate that extracellular Q8 peptide-binding domains restrict CTL-mediated recognition and rejection of tumors. Surprisingly, we find that Q8 and Q9 CTL responses are cross-reactive and that Q8 and Q9 have overlapping peptide-binding motifs. These results are discussed in the context of the unusual properties of the Qa-2 genes and their protein products in the murine immune system.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Mice

C57BL/6J or C57BL/6NCr (referred to as B6) mice were either bred and maintained in the microbiology animal colony at the University of Texas Southwestern Medical Center (Dallas, TX) or purchased from The Jackson Laboratory or from the National Cancer Institute Animal Production Program (Frederick, MD). SCID Beige mice were purchased from Charles River Breeding Laboratories. Adult mice (older than 8 wk) were used for all experiments. All experiments involving animals were performed according to institutional review board guidelines.

Cell lines and cell culture

B78H1 transfectants expressing cell surface Q9 (clone H1Q9. A7, designated in this study as H1Q9), Q9 in conjunction with TAP2 restoration (clone H1Q9TAP.11, designated as H1Q9TAP) and empty vector (vector), and GM-CSF-transduced B78H1 expressing both Q9 and TAP2 (clone GMQ9TAP.13, designated as GMQ9TAP) or TAP2 alone (GMTAP) have been previously described (1, 10), as have Q9-expressing 3LLA9F1 Lewis lung carcinoma (3LL-Q9) and RMA T cell lymphoma (RMA-Q9) transfectants (3). Q8-expressing B78H1 transfectants were derived by similar methods. Briefly, parental B78H1 cells were cotransfected with plasmid pcDNA3.1 (Invitrogen Life Technologies) and plasmid pcDNA1 carrying a Q8/Q9 fusion construct using the FuGENE 6 transfection reagent according to the manufacturer’s recommendations (Boehringer Mannheim). The Q8/Q9 fusion construct was made by restriction enzyme digest of Q8 cDNA with HindIII and BamHI to yield an excised fragment containing the {alpha}1 and {alpha}2 domains of Q8 and similar digest of pcDNA1 plasmid carrying full-length Q9 cDNA to remove the Q9 {alpha}1 and {alpha}2 domain-encoding region, followed by ligation. The resulting pcDNA1 plasmid carried the {alpha}1 and {alpha}2 domains of Q8 and the {alpha}3, transmembrane and cytosolic domains of Q9. B78H1 expressing Q8 alone (clone H1Q8.9, designated as H1Q8) and Q8 along with TAP2 (clone H1Q8TAP.11, designated as H1Q8TAP) and GM-CSF-transduced B78H1 expressing both Q8 and TAP2 (clone GMQ8TAP.10, designated as GMQ8TAP) clones were generated by single-cell sort of bulk transfections after selection using medium supplemented with 800 µg/ml G418 (Invitrogen Life Technologies). Culture medium consisted of 50% DMEM/50% RPMI 1640 supplemented with 10% FBS (Atlanta Biologicals), 1 mM sodium pyruvate (Invitrogen Life Technologies), 0.1 mM nonessential amino acids (Invitrogen Life Technologies), and 10 U/ml penicillin/10 µg/ml streptomycin (Sigma-Aldrich). All B78H1 transfectants were maintained in medium supplemented with 400 µg/ml G418 and grown at 37°C and 5% CO2. To maintain Mycoplasma-free conditions, all cell lines were treated with Mycoplasma Removal Agent (ICN Biochemicals) and periodically cultured in media supplemented with 10 µg/ml CellGro ciprofloxacin HCl (Mediatech).

Antibodies

Anti-Qa-2 mAb 1-1-2 was purchased from BD Pharmingen. Anti-Qa-2 mAbs M46 (11), 20-8-4 (12), and 34-1-2 (13) were purified from ascites as previously described (12). FITC-conjugated goat anti-mouse IgG (Cappel) was used as a secondary Ab.

Flow cytometry

Up to 1 x 106 cells were washed in staining buffer (PBS with 1% FBS and 0.1% sodium azide) and pelleted. A saturating amount of primary Ab was added in a volume of 100 µl, vortexed, and incubated for 30 min at 4°C. Excess unbound Ab was removed by washing with staining buffer. Saturating dilution of secondary Ab was added and incubated at 4°C for 20 min. Cells were washed twice, resuspended in 300 µl of staining buffer, and filtered through 35-µm pore size nylon mesh. A total of 1 x 104 events gated on live cells was collected on FACScan (BD Biosciences). Data were analyzed using CellQuest version 3.1f software (BD Biosciences)

Peptide-induced MHC stabilization assays

Peptide-binding assays were performed as previously described (1). Briefly, cells were cultured in the presence or absence of peptide present at a concentration of 50 µg/ml. Cultures were incubated for 16–20 h at 26°C, followed by 1 h of incubation at 37°C or an additional 1 h at 26°C. Cells were then collected and analyzed by flow cytometry. Peptides used were PEP7 (ELLSCSHLF derived from lymphocytic choriomeningitis virus L protein) (14), L19 (ILMEHIHKL) (15), PEP2 (NQLVNLHDL derived from rabies virus glycoprotein) (14), PEP9 (TLITVRHKI derived from vaccinia virus RNA polymerase) (14), and OVA (SIINFEKL) (16). Additionally, Ala scanning was performed using peptides synthesized to contain single Ala substitutions at each position of the L19 backbone.

Tumor challenge experiments

Tumor challenge experiments were performed as previously described (1, 2, 3). Briefly, 1 x 105 live tumor cells were injected s.c. in a volume of 200 µl of HBSS into the right rear flank. Mice were monitored daily and were considered tumor bearing when the tumor was palpable and measured at least 3 x 3 mm. Animals were sacrificed to avoid pain and suffering when the tumor burden showed physical signs of being excessive. For rechallenge experiments, mice that had completely rejected the initial tumor challenge were injected in the opposite flank with 1 x 105 live tumor cells or received at least three rounds of boosting immunization with the same tumor used in the initial challenge, each immunization given at least 2 wk apart, before rechallenge. Number of mice used in each experiment is indicated in the corresponding figures. Data are represented as Kaplan-Meier plots.

In vitro cytotoxicity assays

51Cr release assays were performed as previously described (2, 3, 10). Briefly, effector cells at various E:T ratios in a volume of 100 µl were added to the wells of a U-bottom 96-well plate. Tumor-reactive CTL were generated by harvesting splenocytes from animals that had rejected an initial tumor challenge and had received a minimum of three boosting immunizations given at least 2 wk apart and restimulating them for 5–7 days in vitro with mitomycin C-treated stimulator tumor cells in the presence of 10 U/ml murine rIL-2. Lymphokine-activated killer (LAK)3 cells were prepared by incubating splenocytes harvested from naive mice in the presence of 500 U/ml IL-2 for 5 days. Target cells were labeled by incubating 2 x 106 cells in a volume of 200 µl with 150–200 µCi of Na51CrO4 (Amersham Biosciences) for 1 h at 37°C and 5% CO2. For in vitro Ab blocking of Qa-2, target cells were incubated with saturating amounts of mAb M46 during the labeling step. Some 1000–5000 51Cr-labeled target cells in 100 µl of media were added to each well, and the plates were incubated at 37°C and 10% CO2 for 4 h. After incubation for 4 h, 100 µl of the supernatant was removed from each well and transferred to Skatron macrowell tubes. Radioactivity was counted in a Micromedic Gamma Counter (ICN Biomedicals). Data are expressed as the percentage of specific release calculated as: (experimental release – spontaneous release)/(maximum release – spontaneous release) x 100. Maximum release was determined by incubating target cells with 100 µl of 1% SDS. All experiments were performed in triplicate.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
TAP-dependent expression of Q8 on the surface of transfected melanoma cells

The amino acid differences between Q9 and Q8 proteins are confined to the {alpha}1 and {alpha}2 peptide-binding domains (discussed later in detail) and to the C-terminal portions of the precursor Qa-2 proteins. Exon 5 in Q9 genes specifies signals for the GPI attachment and phosphatidylinositol-phospholipase C-sensitive cell surface display (17, 18). The C-terminal region of Q8 is a few amino acids shorter and has a different composition than Q9 (7, 19). These changes result in the loss of GPI-processing, phosphatidylinositol-phospholipase C-insensitive surface display and drastically reduced surface expression of the "even" Qa-2 proteins relative to the "odd" Qa-2 (20, 21). It has been previously shown that higher expression levels of Q8 can be restored by swapping their 3' gene regions with the corresponding exons of the classical class I MHC genes (22). In this model, to enhance the surface display and allow for serological monitoring of the even Qa-2 products in transfected tumors, we substituted exons 4 and 5 of Q8 with exons 4 and 5 of Q9. The recombinant Qa-2 cDNA, driven by the CMV promoter, is designated as "Q8" to indicate that its product is identical with the Q8 protein from C57BL/6 in the three extracellular domains (i.e., {alpha}1, {alpha}2, and {alpha}3).

To study immunologic functions of Q8 proteins, we expressed them in B78H1 melanoma cells derived from B16 primary melanoma (1). This cell line is devoid of cell surface-expressed class I MHC due to the absence of H2-Kb, H2-Db, Qa-2, and other MHC class Ib H chain transcripts (10). In addition, B78H1 cells are deficient in transcription of Ag processing components, such as TAP2 and LMP7, but express high levels of beta2-microglobulin chains required for assembly of mature class I MHC. Surface display of transfected Q9 in B78H1 cells was previously shown to depend on cotransfection of TAP2 (10).

Q8 DNA was introduced into weakly immunogenic parental B78H1 (H1) and highly immunogenic GM-CSF-transduced B78H1 melanoma (GM) in TAP2-positive and TAP2-negative backgrounds as described previously (2). Surface expression was monitored with mAb reactive with Qa-2 (23) by flow cytometry. Fig. 1A shows that TAP2-positive Q8 transfectants (H1Q8TAP and GMQ8TAP) were strongly stained with {alpha}3 domain-reactive anti-Qa-2 mAb M46, whereas TAP2-negative variant (H1Q8) did not differ significantly from isotype control (Fig. 1A, upper left panel) or vector-transfected control (data not shown). This pattern of Q8 surface expression is nearly identical to the results seen with ectopically transfected Q9 (Fig. 1A, right panels), suggesting that both Qa-2 proteins bind peptides and share similar dependency on TAP.


Figure 1
View larger version (34K):
[in this window]
[in a new window]
  		FIGURE 1. Cell surface expression of Qa-2 family members Q8 and Q9 on transfected B78H1 tumors. A, Flow cytometry analysis was performed on B78H1 cells transfected with Q8 (left panels) or Q9 (right panels) in the absence or presence of TAP2 gene cotransfection. Pan-Qa-2 mAb M46 (thick line histogram) was used to quantify surface expression levels of Q8 and Q9 on B78H1 transfectants. Background staining (gray shaded histogram) with FITC-conjugated secondary Ab alone. B, Q8 hybrid molecule retains the wild-type Q8 conformation. GMQ8TAP and GMQ9TAP transfectants were stained with a panel of mAbs having defined specificities for Q8 and/or Q9 {alpha}1 and {alpha}2 domain-specific mAbs (anti-Q9 mAb 1-1-2 (thin line histogram); anti-Q9 mAb 20-8-4 (thick line histogram); anti-Q8, Q9 mAb 34-1-2 (filled histogram). Background staining with secondary Ab alone is shown by gray shaded histogram.

 
To determine whether B78H1 TAP-positive transfectants support expression of Q8 with previously defined serological properties (22, 23), transfectants were stained with additional mAbs. Fig. 1B demonstrates that Q9 reacts with {alpha}1- or {alpha}2-specific mAbs 20-8-4, 1-1-2, and 34-1-2, whereas Q8-positive cells stain only with mAb 34-1-2. This pattern of reactivities is consistent with the reported specificities of each mAb. The staining profiles indicate that the conformational epitopes of Q8 and Q9 imposed by their respective {alpha}1, {alpha}2, and {alpha}3 domains are maintained in the transfected melanoma cells.

Q8 mediates rejection of B78H1 melanoma and cross-protects against Q9 in vivo

We have previously defined experimental conditions for studying the Q9-mediated rejection of syngeneic tumors in B6 mice (1, 2, 3). The results of these studies established that ectopic expression of canonical, membrane-bound Q9 on a highly immunogenic variant of TAP-positive B78H1 melanoma (GMQ9TAP) protects ~50% of mice from tumor outgrowth following s.c. injection of melanoma. The primary immunization with 1 x 105 live GMQ9TAP tumor cells, a burden 10-fold higher than the minimal lethal dose (2), led to the generation of CTL immunological memory against subsequent challenges with equally high doses of poorly immunogenic (non-GM-CSF transduced) Q9-bearing melanoma (H1Q9TAP).

We used identical experimental approaches to test whether Q8 has antitumor effects in syngeneic B6 mice. Animals were injected s.c. with 1 x 105 live, highly immunogenic, GMQ8TAP melanoma cells and, as a control, with class I-deficient GMTAP parental cells. Tumor-challenged animals were monitored for tumor outgrowth and were scored as tumor-bearing using previously established criteria (1, 2, 3). Strikingly, all mice injected with Q8-expressing GMQ8TAP melanoma (10 of 10) remained tumor-free, whereas all mice receiving Q8-negative GMTAP had palpable tumor masses within 30 days postinjection (Fig. 2). To demonstrate that absence of tumors in the experimental group was not due to an inherent inability of the Q8 transfectants to grow in vivo, we showed that GMQ8TAP tumors were able to grow out in immune-deficient SCID Beige mice with kinetics comparable to that of class I-negative GMTAP (Fig. 2).


Figure 2
View larger version (18K):
[in this window]
[in a new window]
  		FIGURE 2. B78H1 melanoma cells expressing Q8 are rejected by syngeneic mice. B6 mice (n = 10 for each tumor) were challenged s.c. with 1 x 105 live class I-negative GMTAP or Q8-positive GMQ8TAP cells. As a control, GMQ8TAP tumor outgrowth was monitored in SCID Beige mice (GMQ8TAP/SCID Beige group, n = 5). Tumor growth was monitored for >100 days and the percentage of tumor-bearing mice was recorded as Kaplan-Meier plots.

 
Because B6 mice were able to reject GMQ8TAP tumors, we sought to determine whether these animals were protected by adaptive immune responses to a subsequent challenge with Q9-expressing tumor. The survivors from the initial GMQ8TAP immunization (mice tumor-free for at least 10 wk) were challenged in the opposite flank with 1 x 105 H1Q9TAP cells. As seen in Fig. 3A, in an unimmunized setting, all naive mice succumb to the poorly immunogenic H1Q9TAP variant within 7 wk. Mice previously immunized with GMQ8TAP were protected against subsequent H1Q9TAP challenge, with delayed outgrowth observed in most animals and rejection of H1Q9TAP in some animals. When a similar protocol was performed in GMQ8TAP surviving animals that had received multiple boosting immunizations with GMQ8TAP cells, the anti-H1Q9TAP effects were even more potent, with all animals remaining tumor-free.


Figure 3
View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 3. Cross-protection against Q8- and Q9-expressing melanoma in animals previously immunized against the other Qa-2 family member. A, Mice that had survived an initial tumor challenge of GMQ8TAP cells (immunized, n = 10) or that had survived and received a regimen of boosting immunizations (boosted, n = 10) were subsequently given a s.c. injection of 1 x 105 live H1Q9TAP cells in the opposite flank and H1Q9TAP tumor take was monitored. B, Similarly, GMQ9TAP-immunized (n = 10) or boosted (n = 9) mice were challenged in the opposite flank with H1Q8TAP cells, and H1Q8TAP outgrowth was recorded. As a control, tumor take was also monitored in naive animals (naive, n = 10).

 
Similar experiments were performed in mice previously immunized/boosted with GMQ9TAP tumor cells to determine whether these animals were protected against H1Q8TAP tumors (Fig. 3B). The presence of Q8 on weakly immunogenic B78H1 melanoma (H1Q8TAP) delayed tumor outgrowth in a majority of the naive animals and full protection was observed in only one animal, as compared with full protection (10 of 10) against highly immunogenic GMQ8TAP (Fig. 2). GMQ9TAP-immunized animals were able to mount a protective response that further delayed outgrowth of the subsequent H1Q8TAP challenge, whereas animals that had received multiple boosting rounds of immunization were more effectively able to eliminate the subsequent challenge. However, in this combination, full protection was not observed, suggesting that anti-Q8 immunity is more protective against Q9 than anti-Q9 immunity is against Q8. Nonetheless, in both cases, rejection of tumors expressing one member of the Qa-2 family establishes cross-protection against challenge with tumors expressing the other member.

It is important to point out that Q8 was overexpressed in our experimental tumor model. Therefore, it is possible that the CTL response against this MHC in vivo is less potent than that observed in this study and may depend on coreceptor interactions between T cells and their Q8-positive targets. In the case of costimulatory molecules B7-1 and B7-2, which are known to influence tumor rejection (24), we showed that their presence on GMQ9TAP melanoma has only a modest potentiating effect on tumor rejection (data not shown).

CTL raised in response to Q8-expressing tumor cells cross-react on tumors expressing Q9, and vice versa

In vivo tumor challenge experiments suggest that immunological memory is generated in mice that have rejected either highly immunogenic Q8- or Q9-positive melanoma. Importantly, this secondary response effectively eliminates tumor cells expressing Qa-2 family members to which the animal was not previously exposed. We have previously demonstrated that antitumor CTL are raised in response to challenge with GMQ9TAP melanoma (2). Thus, it is likely that CTL also have an important role in the observed cross-protection. This point is elaborated in detail below.

To assess whether antitumor CTL generated in response to GMQ8TAP or GMQ9TAP tumor challenge (hence restricted by Q8 or Q9, respectively) have fine specificity for their cognate restricting element or exhibit cross-reactivity on other Qa-2 family members, in vitro CTL cytotoxicity assays were performed (Fig. 4). Splenocytes harvested from mice that had survived an initial immunization with melanoma transfectants, and had received multiple boosts with either GMQ8TAP or GMQ9TAP derivatives, were restimulated in vitro with either mitomycin C-treated GMQ8TAP or GMQ9TAP cells to generate CTL. CTL responded equally well against Q9-bearing (Fig. 4A) or Q8-bearing (Fig. 4B) melanoma targets, but did not lyse Qa-2-negative GMTAP targets, regardless of the type of cells used for immunization or in vitro restimulation protocol (mean percentage of specific lysis for GMTAP targets was <6%, data not shown). In Fig. 4C, we show that CTL recognition of GMQ8TAP and GMQ9TAP was abrogated if target cells were preincubated with Qa-2-specific mAb M46, which is reactive with the {alpha}3 domains of Q8 and Q9. M46 effectively disrupts CTL recognition of Q9 targets on tumor cells (2), presumably by blocking potential interaction between CD8 and the class I molecule. These results indicate that at least some of the antitumor CTL raised in response to in vivo challenge with melanoma expressing Q9 or Q8 are not specifically restricted by the immunizing Qa-2 family member, but instead, cross-react against the odd as well as the even Qa-2 family members.


Figure 4
View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 4. CTL raised against melanoma expressing one Qa-2 family member react against disparate tumor targets bearing the other. CTL were generated by harvesting splenocytes from animals previously immunized and boosted with either GMQ9TAP ({blacksquare} and •) or GMQ8TAP ({square} and {circ}) and restimulating them in vitro with either mitomycin C-treated GMQ9TAP (• and {circ}) or GMQ8TAP ({blacksquare} and {square}). CTL were used as effectors against GMQ9TAP (A) or GMQ8TAP (B) target cells. Lysis of Qa-2-negative GMTAP targets by CTL generated against GMQ9TAP is shown as a negative control (x). Error bars are not shown for clarity; SE of triplicate measurements was <6% at each data point. One representative experiment of five is shown. Statistical analysis of mean percentage of specific lysis at the 200:1 E:T ratio for five experiments revealed no statistically significant differences between killing of GMQ9TAP and GMQ8TAP targets by Q9- and Q8-restricted CTL in A and B. Mean percentage of specific lysis (mean ± SE) of GMQ9TAP targets by CTL derived from GMQ9TAP-immunized mice restimulated with GMQ9TAP was 30.0 ± 2.0%; derived from GMQ9TAP-immunized mice restimulated with GMQ8TAP was 32.2 ± 3.2%; derived from GMQ8TAP-immunized mice restimulated with GMQ9TAP was 24.3 ± 1.8%; derived from GMQ8TAP-immunized mice restimulated with GMQ8TAP was 26.5 ± 1.8%. Against GMQ8TAP targets, mean percentage of specific lysis by CTL derived from GMQ9TAP-immunized mice restimulated with GMQ9TAP 29.6 ± 3.6%; derived from GMQ9TAP-immunized mice restimulated with GMQ8TAP was 30.6 ± 1.5%; derived from GMQ8TAP-immunized mice restimulated with GMQ9TAP was 27.3 ± 1.6%; derived from GMQ8TAP-immunized mice restimulated with GMQ8TAP was 28.8 ± 3.2%. C, Blocking with anti-Qa-2 mAb inhibits Q8-restricted CTL-mediated lysis of disparate tumors expressing Q9. Target cells were preincubated in the absence ({blacksquare}) or presence ({square}) of Qa-2 {alpha}3 domain-specific mAb M46 before mixing with CTL generated from splenocytes harvested from GMQ8TAP-immunized animals and cultured with GMQ8TAP stimulator cells. Data were collected at an E:T ratio of 200:1 and are shown as mean ± SE of triplicate measurements. A second experiment yielded similar results.

 
To test whether anti-Q8 CTL generated against melanoma transfectants recognize other Qa-2-positive tumors, CTL from GMQ8TAP-immunized mice were assayed against Q9-positive 3LLA9F1 Lewis lung carcinoma and RMA T cell lymphoma targets. Results in Fig. 4C indicate that CTL raised against Q8 expressed on melanoma cells recognize Q9-expressing 3LL and RMA targets with efficiency similar to that exhibited against melanoma. The killing of the lung carcinoma and lymphoma targets is inhibited by Abs against Qa-2 on tumor cells, confirming the specificity of anti-Q8 CTL for Qa-2 class I. Accordingly, the data suggest that anti-Q8 CTL, which recognize melanoma tumor Ags in the context of Qa-2 restriction elements, respond to similar/shared tumor-associated Ags expressed by other tumors.

Q8 binds peptides with the Q9 peptide-binding motif

The cross-reactivity of the antitumor Qa-2 CTL on Q8 and Q9 suggests that odd and even Qa-2 members have similar Ag-presenting properties. Earlier studies showed that Q9 has a unique peptide-binding motif among all class I Ags (14, 15, 25) and binds unusual peptides that form internal, intrapeptide contacts (26). The two dominant anchors include His residue at position 7 and a hydrophobic amino acid (Leu, Ile, or Phe) at position 9. In addition, one or two auxiliary anchors at variable positions are required for binding of nonameric peptides in the Q9 groove. Because the peptide motifs are defined by the architecture of the class I groove, we compared the homologous {alpha}1 and {alpha}2 regions of the two Qa-2 proteins. Primary Q8 and Q9 sequences from C57BL/6 differ from each other by 22 residues, 10 residing in {alpha}1 (Gln2, Gln6, Arg14, Gly18, Trp21, Ser24, Ile66, Gln72, Gly76, Ser83 in Q9, Pro2, Arg6, Trp14, Val18, Arg21, Ile24, Lys66, Glu72, Val76, Arg83 in Q8) and 12 in {alpha}2 (Met103, Gly107, Val136, Arg144, Arg145, Lys155, Gln157, Thr163, Met165, Glu173, Gly175, Arg181 in Q9, Val103, Glu107, Ala136, Leu144, His145, Arg155, Arg157, Ala163, Val165, Gln173, Arg175, Cys181 in Q8) (7, 19). Using a crystal structure of Q9 stabilized with the self-peptide L19 (26) as a template, Q8 substitutions were mapped onto the Q9 {alpha}1 and {alpha}2 ribbon model (data not shown). Remarkably, the vast majority of Q9 ->Q8 amino acid differences localize outside of the modeled Ag-binding site. Only one substitution, Ile66 ->Lys66, affects a position implicated in the interactions between Q9 and L19 peptide. Similarly, the known polymorphic residues of the odd genes (Gln173 in Q7, Glu173 in Q9) and even genes (Asn30, Lys31 in Q6, Asp30, Thr31 in Q8) are predicted to lie outside of the grooves.

To directly address whether Q8 binds peptides similar to Q9-binding ligands, two approaches were taken. First, we tested the ability of peptides known to stabilize Q9 to various degrees to reconstitute Q8 surface expression on TAP2-deficient H1Q8 melanoma cells (Fig. 5A). B78H1 melanoma is TAP2-deficient and surface expression of transfected "empty" class I molecules can be up-regulated by incubation at low temperatures or by stabilization with exogenous peptides interacting with class I H chains on the cell surface (1). The H1Q8 and H1Q9 B78H1 derivatives were incubated overnight at 26°C in the presence of saturating amounts of different peptides. The peptides L19 and PEP2 were previously shown to bind strongly to Q9; PEP9 has an intermediate Q9 binding phenotype; PEP7 is a weak Q9 binder; and OVA is a peptide presented by Kb and not bound by Q9 (14). As seen in Fig. 5A, peptides that could be exogenously loaded onto Q9 could also bind and stabilize Q8. The overall level of stabilization was partially reduced for Q8 compared with Q9. Expression levels of Q8-peptide complexes remained high after transfer to 37°C, showing that most of these complexes were heat stable. As with Q9, Q8 was stabilized more effectively with L19 and PEP2 than with the intermediate binder PEP9 or the weak binder PEP7.


Figure 5
View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 5. Q8 surface expression is elevated on TAP2-deficient B78H1 cells following exogenous stabilization with Q9-binding peptides. A, Peptide-induced stabilization of Q9 (left) and Q8 (right) was measured by incubating H1Q9 and H1Q8 cells, respectively, overnight at 26°C in the absence of peptide or presence of 50 µg/ml of the indicated peptide, followed by an additional 1-h incubation at either 26°C ({blacksquare}) or 37°C ({square}). Cell surface expression levels, recorded as mean fluorescence intensity of cells stained with mAb M46 and analyzed by flow cytometry, are shown as a percentage of expression levels on cells after overnight incubation at 37°C in the absence of peptides. Data shown are mean ± SE from four experiments. B, Peptides containing Ala substitutions on the L19 peptide backbone were used in peptide stabilization assays. Expression level of Q8 and Q9 on H1Q8 and H1Q9 transfectants, respectively, was determined after overnight incubation at 26°C followed by 1-h incubation at 37°C as described in A. The experiment was repeated with similar results. Ala-substituted L19 peptides exhibited a similar pattern of Q8 and Q9 stabilization after continuous incubation at 26°C (data not shown).

 
Second, we tested variants of Q9-binding peptides to determine whether the anchor requirements of Q8 were similar to that of Q9 (Fig. 5B). Single Ala substitutions on the L19 backbone were made, and these Ala-substituted peptides were exogenously loaded onto H1Q8 cells. Ala substitutions at the Q9-dominant anchor positions P7 and P9 completely abrogated the ability of L19 to stabilize the Q8 complex, and Ala substitutions at the auxiliary anchor positions, P2 and P3, also had an adverse effect on L19 binding.

These experiments demonstrate that Q8 and Q9 share a similar peptide-binding motif consisting of dominant anchors at positions P7 and P9 and auxiliary anchors at P2 and P3. The reduction in overall stabilization effect seen with Q8 (Fig. 5A) may indicate that the two Qa-2 grooves differ slightly in the affinity with which they bind the same peptides, or in the overall conformation/stability of complexes.

Q8 expression on melanoma targets inhibits lysis by LAK cells

It has been noted earlier by Levitsky et al. (27) that NK cells participate in rejection of B78H1 tumors in syngeneic C57BL/6 mice. We confirmed this finding and demonstrated that NK cells are needed during the initial stages of the antitumor immune response but are not required for tumor rejection once the CTL memory against Q9 has been established (2). NK cells may participate in tumor elimination by direct cytotoxicity leading to reduction of the tumor burden as well as by production of cytokines that promote an adaptive antitumor CTL response (28). Paradoxically, Q9 is known to inhibit NK cell killing, as deduced from in vitro cytotoxicity experiments (10). Importantly, the Q9-mediated inhibition is only partial and reduces NK cell cytotoxicity by <50%. Thus, the majority of NK cells (presumably those that do not express Q9-specific inhibitory receptors) are active against B78H1 variants and, depending on the size of the initial tumor burden, participate to varying degrees in the immune response against Qa-2-expressing melanoma cells.

Because Q8 proteins may influence innate antitumor responses as wells, we asked whether they inhibit or activate NK cells. Susceptibility of Q8- and Q9-positive tumor targets to NK cell killing was examined using LAK assays as previously described (10). The results in Fig. 6 demonstrate that H1Q8TAP and GMQ8TAP, like their Q9 counterparts, are less efficiently killed by LAK cells than the class I-negative, vector-transfected B78H1 targets. The protective effect is partial because the presence of Q8 on the tumor surface reduced killing by ~30–40%. Although these experiments do not address whether Q8 interacts with the same inhibitory receptor(s) as Q9, they do suggest that Q8 and Q9 have similar effects on NK cell function during the antitumor response.


Figure 6
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 6. Q8 expression on B78H1 targets protects against LAK cell-mediated cytotoxicity. Day 5 IL-2-cultured LAK cells were used as effectors in 4-h 51Cr release assays against a panel of B78H1 target cells expressing Q8 or Q9. Class I-negative vector-transfected B78H1 cells were used as a positive control. Results are shown as the mean of triplicate measurements. Error bars have been excluded for clarity (SD < 10% for all data points). One representative experiment of five is shown.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
The locus-specific relationships and the selective pressures shaping the evolution of the mouse MHC class I family members have been notoriously difficult to trace. Unlike human class Ia HLA-A, HLA-B, and HLA-C genes, which are tightly linked to each other and maintain sequence patterns characteristic of their allelic associations, the mouse class Ia are encoded by two distinct regions within the MHC complex, H2-K and H2-D, H2-L and lack locus specificity (6, 29). No orthologous relationships between mouse and human class I MHC genes can be delineated (30). Comparison of the class I genes’ flanking regions led Weiss et al. (6) and others (8) to conclude that the murine H2-K subregion, which contains class Ia H2-K gene and a pseudogene designated K1, arose by translocation of an ancestral pair of Qa-2-like genes, with one Q-odd and one Q-even sequence. Accordingly, the translocated Q6/Q8-like gene evolved into the polymorphic class Ia H2-K, while the Q7/Q9-like gene became the K1 pseudogene. This scenario is supported by serologic cross-reactivity of anti-Qa-2 Abs with H2-K proteins (23).

The physiological functions of the present-day Qa-2 genes are still under investigation. We have shown that Q9 protects mice from tumor outgrowth and, in C57BL/6 animals, induces rejection of melanoma to a greater extent than similarly expressed and analyzed Kb (1). The pathway underlying this immune response is most consistent with Q9 acting as a restriction element for presentation of conserved tumor-associated Ags to syngeneic antitumor CTL. Previous studies reported that Qa-2 proteins and genes show little evidence of class Ia-like selective pressures leading to the diversification of their peptide-binding grooves (8, 31, 32). Instead, the genetic polymorphism affects quantitative expression of Qa-2 products with multiple mechanisms controlling display of Q-odd and Q-even proteins on the cell surface (32). The characterized inbred and wild mice were classified as Qa-2high, Qa-2medium, Qa-2low, and Qa-2null. The gene dosage effect plays an important role in these regulatory phenomena resulting in a plastic Qa-2 locus occupied by a variable number of Q-odd and Q-even genes in different haplotypes. The variations in the Qa-2 gene copy number occur by duplications, deletions, unequal crossing over, or gene inactivation via frameshift mutations or transposition of external DNA sequences (6, 8, 33, 34). In addition, the surface Qa-2 expression levels are maintained at low levels due to a relatively low transcriptional activity in most cell types and due to alternative splicing, which interferes with production of canonical, beta2-microglobulin-associated surface Ags (4, 18). Expression at the cell surface is also down-regulated by alterations at the C terminus, which produce Qa-2 proteins with impaired ability to associate stably with the plasma membrane or structural features that enhance susceptibility to cleavage/shedding following their surface expression (35, 36). The intricate control of surface display suggests that strong selection operates on this locus to maintain Qa-2 surface display within a narrow permissible window.

The Q6/Q8 and Q7/Q9 genes in C57BL/6 (Qa-2high) mice encode two prototypical Qa-2 canonical products. The Q7/Q9 is GPI-linked and sensitive to cleavage with phospholipases (37). The Q6/Q8 proteins have phospholipase-resistant C termini, which are truncated due to a premature stop codon in exon 5. This latter alteration is responsible for low surface expression of Q6/Q8 in transfected, tissue-cultured cells (20, 35). Both types of Qa-2 genes have a wide tissue distribution, but some data suggest that their transcription is not always concordant. For example, in liver Q-odd transcripts, but not Q-even transcripts, are detected (4). In tumors, selective silencing of one set of the odd or even loci was observed in rare cases (4). Interestingly, some subpopulations of splenocytes display a large fraction of phospholipase-resistant Qa-2, indicating that high Q6/Q8 expression may occur in a cell-specific fashion (21). The low-expressing Q6 and Q8 cells are highly immunogenic and induce allogeneic CTL that do not cross-react on Q7 and Q9 proteins (9).

In this study, we examined whether overexpressed Q8 plays a role in tumor rejection in experimental models in which Q9 protects mice from tumor outgrowth. We find that high doses of melanoma cells expressing Q8 ectodomains on their cell surface were rejected efficiently by syngeneic C57BL/6 mice, but not by immunodeficient SCID Beige mice. The survivors immunized with Q8-positive tumors were protected against poorly immunogenic Q8+ as well as Q9+ variants of the parental tumor, thus showing that adaptive immune response induced against overexpressed Q8 is also effective against Q9-expressing cells. Consistent with this observation was the finding that CTL raised against either Q8 or Q9 tumors recognized both Qa-2 targets on disparate tumors in vitro, but failed to kill Qa-2-negative tumor cells regardless of immunization or in vitro restimulation protocol. The CTL-mediated killing was class I-specific as demonstrated by blocking experiments with anti-Q8/Q9 mAb. Thus, Q8, just as previously shown for Q9, appears to act as a restriction element for Q8 and Q9 cross-reactive CTL recognizing shared tumor-associated Ags from melanoma, Lewis lung carcinoma and T cell lymphoma.

These results were not anticipated because the predicted Q8 and Q9 proteins differ from each other by 22 residues in the {alpha}1 and {alpha}2 domains. Several of these substituted positions (66, 72, 76, 155, 157) are on the outer face of {alpha} helices and may represent potential TCR contacts, as defined by structural studies of class Ia MHC (38, 39, 40). In addition, an earlier study reported that allogeneic anti-Qa-2 CTL clones exclusively recognize either Q-odd or Q-even products (9). The distinct patterns of T cell reactivity on Q-odd and Q-even products by allogeneic and syngeneic CTL appear, at first glance, puzzling. They can be, however, reconciled when considered in the context of earlier observations that suggest that T cells educated on nonself vs self-MHC molecules have qualitatively different recognition modes.

First, the precursor frequency of alloresponse is orders of magnitude higher than the syngeneic response, which involves a relatively small number of epitope-specific precursors responding to a few immunodominant peptides (41, 42). This higher polyclonality of alloreactive T cells translates into higher diversity of TCR and diminished probability that they will have overlapping specificities. Second, the structural basis of MHC complex recognition may be different in the two cases. Although the great majority of alloreactive T cells see MHC in a peptide-dependent manner and their TCR assume similar diagonal orientations over the peptide groove/MHC helices as the syngeneic TCR (40, 41), it has been argued that the two types of TCR differ in their "bias" toward contacting MHC and peptide. Huseby et al. (43) suggested that negative thymic selection serves to filter TCR away from their MHC helical reactivity and toward peptide specificity. Accordingly, an alloresponse that has not been preceded by thymic deletion of T cells strongly reactive with target MHC will be enriched in TCR making multiple MHC-specific contacts. In contrast, syngeneic TCR repertoire is more likely to focus on peptides bound in the groove and has less interactions with polymorphic residues on MHC helices.

Because common Ag-binding properties may help to explain the cross-reactive mode during the recognition by syngeneic CTL, we asked whether Q8 and Q9 share similar peptide-binding motifs. Peptide stabilization assays indeed confirmed that Q-odd and Q-even proteins associate with peptides bearing the same anchors, His7 and Leu9, though Q8 binds Q9-specific ligands with diminished efficiency, as suggested by our semiquantitative assay. Analysis of the Q8 and Q9 sequences supports the experimental data because all but one of the Q9->Q8 substitutions (Ile66->Lys66) fall outside of the predicted peptide binding site. The slightly altered efficiency with which peptides stabilize Q8 and Q9 may be indicative of subtle differences between Q8 and Q9 grooves. These small distortions, in addition to the polymorphisms on the Q8/Q9 helices, might contribute to the epitopes seen by allogeneic or syngeneic CTL. Indeed, even though at the bulk CTL level we observe cross-reactivity of antitumor anti-Q8 and anti-Q9 responses, we cannot preclude the possibility that rare clones of CTL in the mixed populations express TCR with exclusive specificity for either Q8 or Q9.

One other set of observations may be relevant to our results. Recently it has been shown that CTL with TCR specific for an insulin-derived peptide presented by the class Ib molecule Qa-1 can cross-react with a class Ia molecule (44). The authors went on to propose that this reaction may be a common property of T cells selected by class Ib molecules. Indeed, CTL restricted on other class Ib molecules, such as H2-M3 and CD1b, has also been shown to be promiscuous for multiple ligands that share a similar molecular pattern (45, 46, 47, 48).

Taken together, we hypothesize that multiple factors influencing TCR selection as well as Ag presentation by Qa-2 may contribute to distinct recognition patterns of alloreactive and syngeneic anti-Qa-2 CTL.

An important, yet still unidentified, component of the targeted recognition of antitumor CTL is the putative peptide/ligand bound by the Qa-2 grooves in tumor cells. At this point it is unclear whether the tumor-expressed Qa-2 ligand is a nonameric peptide with unusual structural properties or perhaps another compound that has high affinity for the Qa-2 groove. The crystal structure of Q9 bound to one of the most abundant self-peptides associating with this class I molecule, L19, showed an unusual mode of interaction (26). The endogenous peptide makes few specific contacts and exhibits extremely poor shape complementarity to the MHC groove, raising a possibility that this ligand occupies the groove by default and may be easily displaced. This led us to speculate (3) that in tumors, or in diseased cells, Q9 associates preferentially with a set of conserved, stress-induced/tumor-associated Ags that have a better fit with the Q9 groove than L19 and most other self-derived peptides from normal cells. The sources and the nature of such putative tumor Ags are presently unknown and may differ from a family of prototypical tumor-associated Ags presented by class Ia MHC. Because Q8 appears to have similar structural properties to Q9, it is likely that Q-odd and Q-even members present similar ligands in healthy and nonhealthy cells and are recognized by CTL expressing identical or an overlapping set of TCR. This scenario could be responsible for frequent reshaping of the Qa-2 region and reshuffling of Q-odd and Q-even loci, under common selective pressures that attempt to balance beneficial vs detrimental (autoimmune?) effects of Q-odd and Q-even recognition. Whether this explanation proves correct and what is the nature and the physiological significance of the tumor-associated Ags presented by Q8 and Q9 are key issues that still need to be resolved.


   Acknowledgments
 
We are grateful to Maile Henson, Kathleen Robertson, Jill Mooney, and Dr. Piotr Tabaczewski for contributions in generating B78H1 transfectants.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by Grants 2-R01 AI 19624 and 1-F32 CA111041-01 from the National Institutes of Health. Back

2 Address correspondence and reprint requests to Dr. Iwona Stroynowski, Center for Immunology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-9093. E-mail address: iwona.stroynowski{at}utsouthwestern.edu Back

3 Abbreviation used in this paper: LAK, Lymphokine-activated killer. Back

Received for publication February 9, 2006. Accepted for publication May 24, 2006.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Chiang, E. Y., M. Henson, I. Stroynowski. 2003. Correction of defects responsible for impaired Qa-2 class Ib MHC expression on melanoma cells protects mice from tumor growth. J. Immunol. 170: 4515-4523. [Abstract/Free Full Text]
  2. Chiang, E. Y., I. Stroynowski. 2004. A nonclassical MHC class I molecule restricts CTL-mediated rejection of a syngeneic melanoma tumor. J. Immunol. 173: 4394-4401. [Abstract/Free Full Text]
  3. Chiang, E. Y., I. Stroynowski. 2005. Protective immunity against disparate tumors is mediated by a nonpolymorphic MHC class I molecule. J. Immunol. 174: 5367-5374. [Abstract/Free Full Text]
  4. Ungchusri, T., E. Y. Chiang, G. Brown, M. Chen, P. Tabaczewski, L. Timares, I. Stroynowski. 2001. Widespread expression of the nonclassical class I Qa-2 antigens in hemopoietic and nonhemopoietic cells. Immunogenetics 53: 455-467. [Medline]
  5. Griffiths, E., H. Ong, M. J. Soloski, M. F. Bachmann, P. S. Ohashi, D. E. Speiser. 1998. Tumor defense by murine cytotoxic T cells specific for peptide bound to nonclassical MHC class I. Cancer Res. 58: 4682-4687. [Abstract/Free Full Text]
  6. Weiss, E. H., L. Golden, K. Fahrner, A. L. Mellor, J. J. Devlin, H. Bullman, H. Tiddens, H. Bud, R. A. Flavell. 1984. Organization and evolution of the class I gene family in the major histocompatibility complex of the C57BL/10 mouse. Nature 310: 650-655. [Medline]
  7. Devlin, J. J., E. H. Weiss, M. Paulson, R. Flavell. 1985. Duplicated gene pairs and alleles of class I genes in the Qa-2 region of the murine major histocompatibility complex: a comparison. EMBO J. 4: 3203-3207. [Medline]
  8. Kumánovics, A., A. Madan, S. Qin, L. Rowen, L. Hood, K. Fischer Lindahl. 2002. Quod erat faciendum: sequence analysis of the H2-D and H2-Q regions of 129/SvJ mice. Immunogenetics 54: 479-489. [Medline]
  9. Lowen, L. C., C. J. Aldrich, J. Forman. 1993. Analysis of T cell receptors specific for recognition of class Ib antigens. J. Immunol. 151: 6155-6165. [Abstract]
  10. Chiang, E. Y., M. Henson, I. Stroynowski. 2002. The nonclassical major histocompatibility complex molecule Qa-2 protects tumor cells from natural killer cell- and lymphokine-activated killer cell-mediated cytolysis. J. Immunol. 168: 2200-2211. [Abstract/Free Full Text]
  11. Hasenkrug, K. J., J. M. Cory, J. H. Stimpfling. 1987. Monoclonal antibodies defining mouse tissue antigens encoded by the H-2 region. Immunogenetics 25: 136-139. [Medline]
  12. Ozato, K., D. H. Sachs. 1981. Monoclonal antibodies to mouse MHC antigens. III. Hybridoma antibodies reacting to antigens of the H-2b haplotype reveal genetic control of isotype expression. J. Immunol. 126: 317-321. [Abstract]
  13. Ozato, K., N. M. Mayer, D. H. Sachs. 1982. Monoclonal antibodies to mouse major histocompatibility complex antigens: IV. A series of hybridoma clones producing anti-H-2d antibodies and an examination of expression of H-2d antigens on the surface of these cells. Transplantation 34: 113-120. [Medline]
  14. Tabaczewski, P., E. Chiang, M. Henson, I. Stroynowski. 1997. Alternative peptide binding motifs of Qa-2 class Ib molecules define rules for binding of self and nonself peptides. J. Immunol. 159: 2771-2781. [Abstract]
  15. Joyce, S., P. Tabaczewski, R. H. Angeletti, S. G. Nathenson, I. Stroynowski. 1994. A nonpolymorphic major histocompatibility complex class Ib molecule binds a large array of diverse self-peptides. J. Exp. Med. 179: 579-588. [Abstract/Free Full Text]
  16. Rötzschke, O., K. Falk, S. Stevanovic, G. Jung, P. Walden, H. G. Rammensee. 1991. Exact prediction of a natural T cell epitope. Eur. J. Immunol. 21: 2891-2894. [Medline]
  17. Waneck, G. L., M. E. Stein, R. A. Flavell. 1988. Conversion of a PI-anchored protein to an integral membrane protein by a single amino acid mutation. Science 241: 697-699. [Abstract/Free Full Text]
  18. Ulker, N., L. E. Hood, I. Stroynowski. 1990. Molecular signals for phosphatidylinositol modification of the Qa-2 antigen. J. Immunol. 145: 2214-2219. [Abstract]
  19. Cai, W., W. Cao, L. Wu, G. E. Exley, G. L. Waneck, B. L. Karger, C. M. Warner. 1996. Sequence and transcription of Qa-2-encoding genes in mouse lymphocytes and blastocysts. Immunogenetics 45: 97-107. [Medline]
  20. Mann, D., J. Forman. 1988. Recognition by cytotoxic T lymphocytes of Qa-2 antigens: sensitivity of Qa-2 molecules to phosphatidylinositol-specific phospholipase C. J. Immunol. 141: 1813-1818. [Abstract]
  21. Stiernberg, J., M. G. Low, L. Flaherty, P. W. Kincade. 1987. Removal of lymphocyte surface molecules with phosphatidylinositol-specific phospholipase C: effects on mitogen responses and evidence that ThB and certain Qa antigens are membrane-anchored via phosphatidylinositol. J. Immunol. 138: 3877-3884. [Abstract]
  22. Straus, D. S., I. Stroynowski, S. G. Schiffer, L. Hood. 1985. Expression of hybrid class I genes of the major histocompatibility complex in mouse L cells. Proc. Natl. Acad. Sci. USA 82: 6245-6249. [Abstract/Free Full Text]
  23. Sharrow, S. O., J. S. Arn, I. Stroynowski, L. Hood, D. H. Sachs. 1989. Epitope clusters of Qa-2 antigens defined by a panel of new monoclonal antibodies. J. Immunol. 142: 3495-3502. [Abstract]
  24. Wu, T. C., A. Y. Hwang, E. M. Jaffee, H. I. Levitsky, D. M. Pardoll. 1995. A reassessment of the role of B7-1 expression in tumor rejection. J. Exp. Med. 182: 1415-1421. [Abstract/Free Full Text]
  25. Rötzschke, O., K. Falk, S. Stevanovic, B. Grahovac, M. J. Soloski, G. Jung, H. G. Rammensee. 1993. Qa-2 molecules are peptide receptors of higher stringency than ordinary class I molecules. Nature 361: 642-644. [Medline]
  26. He, X.-L., P. Tabaczewski, J. Ho, I. Stroynowski, K. C. Garcia. 2001. Promiscuous antigen presentation by the nonclassical MHC Ib Qa-2 is enabled by a shallow, hydrophobic groove and self-stabilized peptide conformation. Structure 9: 1213-1224. [Medline]
  27. Levitsky, H. I., A. Lazenby, R. J. Hayashi, D. M. Pardoll. 1994. In vivo priming of two distinct antitumor effector populations: the role of MHC class I expression. J. Exp. Med. 179: 1215-1224. [Abstract/Free Full Text]
  28. Biron, C. A., K. B. Nguyen, G. C. Pien, L. P. Cousens, T. P. Salazar-Mather. 1999. Natural killer cells in antiviral defense: function and regulation by innate cytokines. Annu. Rev. Immunol. 17: 189-220. [Medline]
  29. Pullen, J. K., R. M. Horton, Z. L. Cai, L. R. Pease. 1992. Structural diversity of the classical H-2 genes: K, D, and L. J. Immunol. 148: 953-967. [Abstract]
  30. Kumánovics, A., T. Takada, K. F. Lindahl. 2003. Genomic organization of the mammalian MHC. Annu. Rev. Immunol. 21: 629-657. [Medline]
  31. Michaelson, J., L. Flaherty, B. Hutchinson, H. Yudkowitz. 1982. Qa-2 does not display structural genetic polymorphism detectable on isoelectric-focusing gels. Immunogenetics 16: 363-366. [Medline]
  32. Tine, J. A., A. Walsh, D. Rathbun, L. Leonard, E. K. Wakeland, R. Dilwith, L. Flaherty. 1990. Genetic polymorphisms of Q region genes from wild-derived mice: implications for Q region evolution. Immunogenetics 31: 315-325. [Medline]
  33. Matsuura, A., R. Schloss, F. W. Shen, J. S. Tung, S. W. Hunt, III, D. A. Fisher, L. E. Hood, E. A. Boyse. 1989. Expression of the Q8/9d gene by T cells of the mouse. Immunogenetics 30: 156-161. [Medline]
  34. Nakayama, K., S. Tokito, C. Pannetier, H. Nakauchi, G. Gachelin. 1991. MHC gene Q8/9d of the BALB/cJ mouse strain cannot encode a Qa-2,3 class I antigen. Immunogenetics 33: 225-234. [Medline]
  35. Waters, J., L. Flaherty. 1991. Expression and regulation of Q8b in a transfected cell line. Immunogenetics 34: 179-184. [Medline]
  36. Tabaczewski, P., H. Shirwan, K. Lewis, I. Stroynowski. 1994. Alternative splicing of class Ib major histocompatibility complex transcripts in vivo leads to the expression of soluble Qa-2 molecules in murine blood. Proc. Natl. Acad. Sci. USA 91: 1883-1887. [Abstract/Free Full Text]
  37. Stroynowski, I., M. Soloski, M. G. Low, L. Hood. 1987. A single gene encodes soluble and membrane-bound forms of the major histocompatibility Qa-2 antigen: anchoring of the product by a phospholipid tail. Cell 50: 759-768. [Medline]
  38. Bjorkman, P. J., M. A. Saper, B. Samraoui, W. S. Bennett, J. L. Strominger, D. C. Wiley. 1987. The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature 329: 512-518. [Medline]
  39. Rudolph, M. G., J. G. Luz, I. A. Wilson. 2002. Structural and thermodynamic correlates of T cell signaling. Annu. Rev. Biophys. Biomol. Struct. 31: 121-149. [Medline]
  40. Garcia, K. C., E. J. Adams. 2005. How the T cell receptor sees antigen: a structural view. Cell 122: 333-334. [Medline]
  41. Whitelegg, A., L. D. Barber. 2004. The structural basis of T-cell allorecognition. Tissue Antigens 63: 101-108. [Medline]
  42. Blattman, J. N., R. Antia, D. J. Sourdive, X. Wang, S. M. Kaech, K. Murali-Krishna, J. D. Altman, R. Ahmed. 2002. Estimating the precursor frequency of naïve antigen-specific CD8 T cells. J. Exp. Med. 195: 657-664. [Abstract/Free Full Text]
  43. Huseby, E. S., J. White, F. Crawford, T. Vass, D. Becker, C. Pinilla, P. Marrack, J. W. Kappler. 2005. How the T cell receptoire becomes peptide and MHC specific. Cell 122: 247-260. [Medline]
  44. Reed-Loisel, L. M., B. A. Sullivan, O. Laur, P. E. Jensen. 2005. An MHC class Ib-restricted TCR that cross-reacts with an MHC class Ia molecule. J. Immunol. 174: 7746-7752. [Abstract/Free Full Text]
  45. Ploss, A., G. Lauvau, B. Contos, K. M. Kerksiek, P. D. Guirnalda, I. Leiner, L. L. Lenz, M. J. Bevan, E. G. Pamer. 2003. Promiscuity of MHC class Ib-restricted T cell responses. J. Immunol. 171: 5948-5955. [Abstract/Free Full Text]
  46. Lo, W. F., A. S. Woods, A. DeCloux, R. J. Cotter, E. S. Metcalf, M. J. Soloski. 2000. Molecular mimicry mediated by MHC class Ib molecules after infection with Gram-negative pathogens. Nat. Med. 6: 215-218. [Medline]
  47. Shamshiev, A., A. Donda, T. I. Prigozy, L. Mori, V. Chigorno, C. A. Benedict, L. Kappos, S. Sonnino, M. Kronenberg, G. De Libero. 2000. The {alpha}beta T cell response to self-glycolipids shows a novel mechanism of CD1b loading and a requirement for complex oligosaccharides. Immunity 13: 255-264. [Medline]
  48. Prendergast, M. M., A. J. Lastovica, A. P. Moran. 1998. Lipopolysaccharides from Campylobacter jejuni O:41 strains associated with Guillain-Barre syndrome exhibit mimicry of GM1 ganglioside. Infect. Immun. 66: 3649-3655. [Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 JEMHome page
P. A. Swanson II, C. D. Pack, A. Hadley, C.-R. Wang, I. Stroynowski, P. E. Jensen, and A. E. Lukacher
An MHC class Ib-restricted CD8 T cell response confers antiviral immunity
J. Exp. Med., July 7, 2008; 205(7): 1647 - 1657.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
D. C. Jay, L. M. Reed-Loisel, and P. E. Jensen
Polyclonal MHC Ib-Restricted CD8+ T Cells Undergo Homeostatic Expansion in the Absence of Conventional MHC-Restricted T Cells
J. Immunol., March 1, 2008; 180(5): 2805 - 2814.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chiang, E. Y.
Right arrow Articles by Stroynowski, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chiang, E. Y.
Right arrow Articles by Stroynowski, I.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS