Download Peptide Repertoire Class I Molecule Q10 Binds a Classical The

The Murine Liver-Specific Nonclassical MHC
Class I Molecule Q10 Binds a Classical
Peptide Repertoire
This information is current as
of June 16, 2017.
Francesca Zappacosta, Piotr Tabaczewski, Kenneth C.
Parker, John E. Coligan and Iwona Stroynowski
J Immunol 2000; 164:1906-1915; ;
doi: 10.4049/jimmunol.164.4.1906
http://www.jimmunol.org/content/164/4/1906
Subscription
Permissions
Email Alerts
This article cites 65 articles, 28 of which you can access for free at:
http://www.jimmunol.org/content/164/4/1906.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2000 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
References
The Murine Liver-Specific Nonclassical MHC Class I Molecule
Q10 Binds a Classical Peptide Repertoire1
Francesca Zappacosta,2,3* Piotr Tabaczewski,2† Kenneth C. Parker,4* John E. Coligan,5* and
Iwona Stroynowski6†‡
M
olecular and biochemical analyses of class I MHC
molecules led to the identification of two major subgroups of these proteins. The classical class I (class Ia)
Ags are highly polymorphic, nearly ubiquitously expressed
polypeptides that associate with self- and nonself 8 –10 residuelong peptides (1, 2). They play key roles in T cell and NK cellmediated elimination of virally infected and/or malignantly transformed cells. The nonclassical class I (class Ib) Ags are a
heterogeneous group of ␤2-microglobulin (␤2m)7-associated proteins that display little polymorphism and frequently exhibit low
level expression and/or unique tissue distributions (3, 4). Further*Laboratory of Allergic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, MD 20852; and †Center for Immunology and ‡Departments of Microbiology and Internal Medicine, University of Texas
Southwestern Medical Center, Dallas, TX 75235-9093
Received for publication September 9, 1999. Accepted for publication December
1, 1999.
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 in part by grants from the National Institutes of Health
(AI19624 and AI37818).
2
F.Z. and P.T contributed equally to this work.
3
Current address: Department of Physical and Structural Chemistry, SmithKline
Beecham Pharmaceutical, King of Prussia, PA 19406.
4
Current address: PE Biosystems, 500 Old Connecticut Path, Framingham,
MA 01701.
5
Address correspondence and reprint requests to Dr. John E. Coligan, Laboratory of
Allergic Diseases, National Institute of Allergy and Infectious Diseases, National
Institutes of Health, Twinbrook II, Room 205, Rockville, MD 20852. E-mail address:
[email protected]
6
Address correspondence and reprint requests to Dr. Iwona Stroynowski, Center for
Immunology, University of Texas Southwestern Medical Center, Dallas, TX 752359093. E-mail address: Stroynow@UTSW. SWMED.edu
7
Abbreviations used in this paper: ␤2m, ␤2-microglobulin; GPI, glycosylphosphatidylinositol; CAD, collision-activated dissociation; MS, mass spectrometry; PI-PLC,
phosphatidylinositol-specific phospholipase C; ESI, electrospray ionization; m/z,
mass:charge ratio.
Copyright © 2000 by The American Association of Immunologists
more, many of the class Ib molecules exist in soluble forms that are
secreted into the serum and body fluids (5, 6).
Recent studies of rodent and human members of class Ib families revealed remarkable diversity of their ligands, Ag-presenting
capacities, and immune as well as nonimmune functions (7–12).
Some of the membrane-bound class Ib proteins are dedicated to
presentation of structurally unique forms of ligands. For example,
M3 Ag widely expressed on murine tissues, binds selectively Nformylated peptides of mostly prokaryotic origin (13). This property allows M3 to be recognized as a restriction element during
CD8⫹ T cell-mediated clearance of bacterial infections. Another
ubiquitously expressed murine Ag, Qa-1, as well as its proposed
human homologue HLA-E, associate preferentially with a limited
set of hydrophobic leader peptides from class I MHC Ags (14, 15).
The resulting class Ib complexes serve as targets for alloreactive
cytotoxic T cells, as shown for Qa-1 (16), and as recognition elements for NK receptors (17–19).
Not all of the known class Ib proteins bind structurally unique
ligands. Some, such as murine Qa-2 and human HLA-G, associate
with diverse repertoires of peptides reminiscent of class Ia peptides
(20, 21). The biological significance of these types of classes Ib
complexes is still poorly understood (22). Additionally, very little
is currently known about Ag-presenting properties or function(s)
of soluble class Ia or Ib molecules reported to exist in a wide range
of species, including mouse (23) and human (24, 25).
To address these issues, we performed analysis of ligands associated with the soluble Q10 class Ib protein. This 38- to 40-kDa
␤2m-associated molecule is detectable in serum as a multivalent
complex of 200 –300 kDa, at concentrations ranging from 20 to 60
␮g/ml, depending on the mouse strain (26, 27). The Q10 proteins
are encoded in the Q region of the H-2 complex, which also contains Qa-2 genes (3), and a cluster of several other class Ib sequences. In common with other Q region class Ib genes, Q10
shows ⬎80% homology with the classical H-2K, D, and L loci
(28). Structurally, the protein is truncated at the C terminus and
0022-1767/00/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
The biological properties of the nonclassical class I MHC molecules secreted into blood and tissue fluids are not currently
understood. To address this issue, we studied the murine Q10 molecule, one of the most abundant, soluble class Ib molecules. Mass
spectrometry analyses of hybrid Q10 polypeptides revealed that ␣1␣2 domains of Q10 associate with 8 –9 long peptides similar to
the classical class I MHC ligands. Several of the sequenced peptides matched intracellularly synthesized murine proteins. This
finding and the observation that the Q10 hybrid assembly is TAP2-dependent supports the notion that Q10 groove is loaded by
the classical class I Ag presentation pathway. Peptides eluted from Q10 displayed a binding motif typical of H-2K, D, and L
ligands. They carried conserved residues at P2 (Gly), P6 (Leu), and P␻ (Phe/Leu). The role of these residues as anchors/auxiliary
anchors was confirmed by Ala substitution experiments. The Q10 peptide repertoire was heterogeneous, with 75% of the groove
occupied by a multitude of diverse peptides; however, 25% of the molecules bound a single peptide identical to a region of a TCR
V ␤-chain. Since this peptide did not display enhanced binding affinity for Q10 nor does its origin and sequence suggest that it is
functionally significant, we propose that the nonclassical class I groove of Q10 resembles H-2K, D, and L grooves more than the
highly specialized clefts of nonclassical class I Ags such as Qa-1, HLA-E, and M3. The Journal of Immunology, 2000, 164:
1906 –1915.
The Journal of Immunology
1907
carries several substitutions in the hydrophobic region corresponding to the transmembrane segments of class Ia heavy chains. These
features account for the inability of Q10 to insert into the plasma
membrane and explain why Q10 is secreted (26, 28).
The Q10 locus exhibits two hallmarks of class Ib genes: it is
well conserved, with ⬎99.4% homology between different sequenced alleles (28), and it is expressed in tissue-specific fashion.
In adult mice, the protein is synthesized mainly by liver and, in
trace amounts, by kidney and stomach (26, 29). During early development, Q10 transcripts are detectable in major organs of fetal
hematopoiesis: visceral yolk sac and fetal liver (30). This expression pattern led to the speculation that Q10 participates in the
induction of T cell tolerance and/or regulation of embryonic hematopoiesis. We demonstrate here that the peptide-binding (␣1␣2)
domains of Q10 associate with eight and nine residue-long selfpeptides similar to the class Ia ligands and discuss this finding in
the context of potential T cell and NK cell recognition.
Cell lines and tissue culture
Materials and Methods
Flow cytometry
Cloning of Q10 cDNA and construction of hybrid genes
Cells from subconfluent cultures were stained by indirect immunofluorescence using FITC-conjugated goat-anti mouse IgG as the secondary Ab
(Cappel, Durham, NC). The acquisition was performed by FACScan (Becton Dickinson, Mountain View, CA). Data were analyzed with the Lysis
program (Becton Dickinson). Dead cells were excluded by a combination
of gates set on forward/side scatter and by exclusion of cells staining positive with propidium iodide dye.
Antibodies
The mAbs 46 (anti-␣3 of Qa-2) (34) and S19.8 (anti-mouse ␤2mb) (35) used
in Q10-affinity purification and ELISA were purified from mouse ascites fluid
or from ␥-globulin-free tissue culture supernatants with protein A-Sepharose
CL-4B (Pharmacia, Piscataway, NJ) using standard protocols (36). Secondary
mAbs used in ELISA were biotinylated with N-hydroxysuccinimidobiotin
(Sigma, St. Louis, MO) as described previously (36).
Radiolabeling and immunoprecipitation
Radiolabeling and immunoprecipitations were conducted by a modification
of a standard method described previously (5, 36). Briefly, RMA transfectants (107 cells) were harvested at the logarithmic phase of growth (8 ⫻
105/ml), washed twice in ice-cold PBS, and resuspended in labeling medium: 1 ml of methionine/cysteine-deficient RPMI 1640 medium (ICN
Pharmaceuticals, Costa Mesa, CA) supplemented with 10% dialyzed FBS
and 0.5 mCi of [35S]methionine and cysteine (Trans 35S-label; ICN Pharmaceuticals). For phosphatidylinositol-specific phospholipase C (PI-PLC)
treatment, tissue culture media were supplemented with 0.3 U of PI-PLC
(American Radiolabeled Chemicals, St. Louis, MO). After incubation for
4 h at 37°C, cell supernatants were harvested and precleared with 50 ␮l of
normal rabbit serum. Recombinant Q10 and control proteins were precipitated with saturated amounts of mAb 46 Ab. Ag-mAb 46 complexes were
bound to protein A-Sepharose CL-4B (Sigma), washed six times with PBS,
denatured, reduced, and analyzed by one-dimensional SDS-PAGE. Gels
were stained with Coomassie blue. Radioactively labeled proteins were
detected by autoradiography.
Measurement of expression levels and stability of Q10 molecules
by ELISA
RMA, RMA-S cells, and their transfectants were grown to a density of 8 ⫻
105 cells/ml. Caps of tissue culture flasks were tightened, and cells were
incubated overnight at room temperature. Cells were harvested and washed
three times with ice-cold PBS. Pellets were lysed with 0.5% nonionic detergent Nonidet P-40 (Sigma) in 0.2 M phosphate buffer (pH 7.05) and in
the presence of proteinase inhibitors: pepstatin A, 5 ␮g/ml; leupeptin, 2
␮g/ml; benzamidine, 2.5 mM; soybean trypsine inhibitor, 20 ␮g/ml;
PMSF, 100 ␮M; and EDTA, 4 mM. Cell nuclei were pelleted by centrifugation. The protein concentrations were measured using the bicinchoninic
acid protein assay (Pierce, Rockford, IL). Where appropriate, adjustments
were made to standardize protein concentrations of lysates. Supernatants
containing class I complexes were stored on ice until needed (no more than
16 h). To measure MHC levels, we used a modified semiquantitative
two-Ab sandwich ELISA assay (41). For MQ10 and SQ10 measurements,
mAb 46 (anti-␣3 of Qa-2) was used as primary Ab and biotinylated mAb
S19.8 (anti-␤2mb) as secondary Ab. The assay for Qa-2 was performed
using the same protocol on cell lysates of MQa-2 transfectants (41). The
assay for H-2Kb was performed similarly with mAb 20-8-4 as a primary Ab
and biotinylated mAb Y3 as a secondary Ab (41).
Isolation of endogenous peptides from Q10 complexes
MQ10 and SQ10 complexes and their ligands were purified using two
different methods. Endogenous peptides bound to MQ10 molecules were
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
The nonpolymorphic Q10 cDNA was isolated from the NOD/Lt (H-2g7)
cDNA liver library derived by Girgis et al. (31). The cDNA fragment
encoding the N-terminal portion of Q10 (exons 1–3) was amplified by
PCR, subcloned into pIC20H plasmid (American Type Culture Collection
(ATCC), Manassas, VA), and sequenced. It was found to be identical to the
genomic sequence of Q10 from C3H mouse (32) and cDNAs amplified
from C57BL/6 and C57BL/10 mice (data not shown).
We designed two hybrid Q10/Qa-2 molecules. The first, MQ10, encodes the N-terminal portion of Q10 (leader peptide, ␣1 and ␣2) and the
C-terminal portion of Qa-2 (␣3 and the glycosylphosphatidylinositol (GPI)
moiety linking Qa-2 product to the cell surface). MQ10 is membrane
bound. The second hybrid molecule, SQ10, consists of the same N-terminal
domains of Q10 linked to the ␣3 domain of the soluble form of Qa-2,
followed by six additional histidines (6xHis-tag), and is secreted. The Cterminal domains of MQ10 and SQ10 were derived from different isoforms
of Qa-2 genes, Q9m and Q7s, respectively (5).
The following pairs of primers were used to amplify parts of H-2 molecules: for the ␣1␣2 region of Q10, (P1) 5⬘-AAACCCGTCGACGATC
CCAGATGGGGGCGATGGCG-3⬘ (signal peptide sequence in bold,
SalI site underlined) and (P2) 5⬘-AAACCCAGATCTGTGCGCAG
CAGCGTCT-3⬘ (C-terminal part of ␣2 domain in bold, BglII site underlined); for the ␣3 region of Q9m, (P3) 5⬘-GCGCACGGATC
CCCCAAAGGCACATGTGACCCATC-3⬘ (Q9m ␣3 N-terminal region
in bold, BamHI site underlined), and (P4) 5⬘-CTGCAGCTCGAGT
CATGCTGGAGCTGGAGCACAGTCCCC-3⬘ (Q9m C terminus in
bold, stop codon in bold italics, and XhoI site underlined); for the ␣3 region
of Q7s (P3) and (P5) 5⬘-CCAATCGAATTCGCTGGAGCTGGAGCA
CAGTCCCC-3⬘ (Q7s C terminus in bold, EcoRI site underlined). The
Q9m and Q7s fragments were cloned into plasmid pIC20H and the Q10
fragment into Bluescript II KS(⫺) (Stratagene, San Diego, CA), respectively, using the indicated underlined sites. To add sequences encoding six
histidines followed by a stop codon (in italics ⫽ 6xHis-tag), we inserted a
synthetic linker 5⬘-AGCGAATTCACATCACCATCACCATCACTGACT
GCAC-3⬘ at the C terminus of the Q7s fragment (in bold) using the underlined EcoRI and XhoI sites. The DNA structures of all fragments were
verified by sequencing. Recombinant MQ10 and SQ10 clones were constructed by combining the Q10 fragment with Q9m or Q7s6xHis-tag DNAs
in vector pIC20H in the following configuration: SalI–Q10 –BglII/BamHI–
Q9m–XhoI or SalI–Q10 –BglII/BamHI–Q7s6xHis-tag–XhoI. SalI/XhoI
fragments containing full-size hybrid Q10 genes were cloned subsequently
into the XhoI site of vector pBJ5 behind the ubiquitous SR␣ promoter (33)
(pBJ5/MQ10 and pBJ5/SQ106xHis-tag). These plasmids were used to
transfect various cell lines.
The murine cell lines RMA and its TAP2-deficient mutant, RMA-S (37),
were transfected with linearized Q10 constructs and pHEKneo vector
(G418 resistance marker) by electroporation as described previously (38).
Transfectants expressing the highest levels of MQ10 were selected by flow
cytometry with mAb 46. Clones secreting the highest levels of SQ10 were
identified by a two-Ab sandwich ELISA (see below) and further characterized by immunoprecipitation with mAb 46. Transfectants were propagated in the presence 0.15 mg/ml of active G418 (Fisher Scientific, Pittsburgh, PA). Large scale cultures of RMA/MQ10-positive cells were grown
in the Laboratory of Cellular and Developmental Biology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of
Health (Rockville, MD) under the direction of Dr. J. Shiloach. Large scale
cultures of RMA/SQ10-positive cells were grown to saturation (2 ⫻ 106/
ml) in Fenwal Lifecell TC flasks of 3-liter capacity (Baxter Scientific Products, McGaw Park, IL) at the University of Texas Southwestern Medical
Center, Dallas. The rat YB2/0 (39) and human C1R (40) cell lines were
obtained from ATCC (ATCC CRL 1662) and Dr. J. Forman (University of
Texas Southwestern Medical Center, Dallas), respectively. The RMA and
RMA-S transfectants expressing MQa-2 and SQa-2 were described elsewhere (38, 41).
1908
Peptide sequence analysis
All mass spectrometric data were acquired on an API 300 triple quadrupole
mass spectrometer (PE-SCIEX, Toronto, Ontario, Canada) equipped with
a MicroIonSpray source as previously reported (42). The program MS-Tag,
written by Karl Clauser and Peter Baker, and available on the worldwide
web at http://prospector.ucsf.edu, was used to match collision-activated
dissociation (CAD) spectra against the protein sequence databases available at the web site. N-terminal amino acid sequence analysis was performed by standard automated Edman degradation.
Peptide synthesis
Peptides were synthesized as described previously (41, 43). Purity and
sequence of the synthetic peptides was established by analytical RP-HPLC
and mass spectrometry.
Protein analysis
Mass spectrometric analysis of the proteins retained by the ultrafiltration
membrane, intact or after deglycosylation, was performed after purification
on a narrow-bore Vydac C4 column (150 ⫻ 2.1 mm, 5 ␮m, 330 Å pore
size) using the gradient described above for peptide separation. Samples
were directly injected into the mass spectrometer ion source by infusion at
1 ␮l/min.
Deglycosylation of SQ10His heavy chain was conducted in 0.5% 3-[(3cholamidopropyl)dimethylammonio]-1-propanesulfonate, 100 mM TrisHCl (pH 8.0), and 0.1 mM DTT. The sample was incubated with 0.5 U of
N-glycosidase F (Boehringer Mannheim, Indianapolis, IN) at 37°C for 16 h
and subsequently purified by RP-HPLC for mass spectrometric analysis.
Analysis of peptide binding to class I molecules
ELISA-based peptide-induced stabilization assays were conducted by a
modification of a method described earlier (41). Briefly, graded concentrations of synthetic peptides were added to cell lysates of RMA-S cells
expressing MQ10 molecules (transfectant P25-1). The mixtures were kept
on ice for 16 h, followed by an 80-min incubation at 42°C. The presence
of conformationally stable MQ10 serologic epitopes was detected by sandwich ELISA with mAb 46 and S19.8 Abs as described above in “Measurement of expression levels and stability of Q10 molecules by ELISA”.
FIGURE 1. Structure of hybrid MQ10 and SQ10 constructs and proteins. Exons originating from Q10 (black boxes) and Qa-2 genes (open
boxes) are denoted as leader (L), ␣1, ␣2, ␣3, transmembrane (TM), and
cytoplasmic (cyt). GPI anchor of MQ10 is denoted as GPI and 6xHis-tag
of SQ10 is denoted as 6His. Positions of primers used for PCR amplification of Q10 and Qa-2 cDNAs are indicated by P1-P5 arrows. Note that
the Q10 promoter has been replaced with the SR␣ that allows expression
in all cell types.
Results
Construction of cell lines expressing hybrid Q10 proteins
To perform direct biochemical analysis of endogenously synthesized Q10-binding ligands, it is necessary to isolate sufficient
quantities of the relevant class I complexes. Since there are currently no known mAbs that would allow purification of wild-type
Q10 from serum, we cloned Q10 cDNA and expressed Q10 molecules as class I hybrid proteins in transfected tissue-cultured cell
lines (Fig. 1). The two putative ligand-binding domains (␣1 and
␣2) of Q10 were fused to the ␣3 domain and C-terminal portion of
another Q region protein, Qa-2 (44 – 46). The Qa-2 proteins exist
in two isoforms: membrane-bound Qa-2 attached to cell surface
via GPI moiety (MQa-2) and soluble Qa-2 derived from the same
gene by alternative splicing (5, 46) (SQa-2). The choice of the ␣3
domain in the Q10 hybrids was dictated by the high homology
between the Q10 and Qa-2 sequences (28, 45), by availability
of multiple mAbs recognizing unique Qa-2 epitopes on the ␣3
domain (34, 35, 38), and by previous studies showing that the
shuffling of Qa-2 domains with other class I domains does not
disturb the conformation of the ␣1␣2 portion of hybrid complexes
(47, 48).
Two forms of Q10 hybrid genes were constructed: SQ10, encoding soluble form of Q10, and MQ10, encoding membranebound, GPI-linked Q10 protein. The structure of the predicted hybrid Q10 genes and proteins is depicted in Fig. 1. The hybrid
constructs were transfected into lymphoid-derived cell lines: murine (RMA), rat (YB2/0), and human (CIR) cell lines. High levels
of Q10 proteins were detected in every case, suggesting that assembly of these complexes is not limited by the lack of appropriate
ligands or chaperones (data not shown and see below). We were
unable to perform any studies with transfected liver cell lines because these cells could not be grown to the densities necessary for
biochemical characterization of the hybrid molecules (46).
To verify the integrity of Qa-2 conformational epitopes on the
␣3 domain of MQ10, the transfected murine RMA cells were
stained with six anti-␣3 Qa-2 mAbs (data not shown). All reacted
with MQ10 as well as with control wild-type Qa-2-positive cells.
As expected, anti-␣1␣2 Qa-2 Abs did not react with MQ10.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
isolated by a modification of a method previously described for other membrane-bound class I complexes (42). A total of 1010 MQ10-transfected
RMA cells were lysed in 20 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1%
3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate, 0.25% sodium deoxycholate, 1 mM PMSF, 100 mM iodoacetamide, 5 ␮g/ml aprotinin, 10 ␮g/ml leupeptin, 10 ␮g/ml pepstatin A, 5 mM EDTA, and 0.04%
sodium azide. After centrifugation, the cell lysate was loaded onto a column of inactivated Sepharose CL-4B and subsequently onto a Sepharose
CL-4B column to which mAb 46 had been coupled (36). After extensive
washing, MQ10 complexes were eluted with 10% acetic acid. The released
peptides were isolated by centrifugation through an Ultrafree-CL 5kDa
microconcentrator (Millipore, Bedford, MA) and concentrated by lyophilization to 250 ␮l. Peptides were separated by RP-HPLC as described
elsewhere (42). Individual fractions were collected, dried, and stored at
⫺20°C before mass spectral analysis.
SQ10 molecules were purified from 50 liters of supernatant of RMA/
SQ10 transfectants (P29-3.4) collected over a 3-wk period. The medium
collected from cells was supplemented with 0.2% w/v sodium azide and
stored at 4°C. The pooled supernatant was concentrated to 4 liters by ultrafiltration using a hollow fiber cartridge with a 30-kDa cutoff (model
UFP-3-C-5; A/G Technology, Needham, MA). The concentrate was spun
at 14,000 ⫻ g for 2 h and filtered through a 0.2-␮m membrane. Tris-HCl
was added to a final concentration of 0.1 M, and the pH was adjusted to 8.0.
SQ10 proteins containing the 6xHis-tag were purified by metal affinity
chromatography. Briefly, 30 ml of Ni-NTA Sepharose beads (Qiagen,
Chatsworth, CA) was stirred gently overnight at 4°C with the concentrated
sample containing recombinant SQ10. Sepharose beads suspension was
transferred to a chromatography column and extensively washed. The
SQ10 was eluted from the column with 2 column volumes of PBS and 250
mM imidazole. To achieve prompt neutralization, fractions (5 ml) were
collected in tubes containing 0.5 ml of 1 M Tris-HCl (pH 7.0). Fractions
containing SQ10 were identified by sandwich ELISA. SQ10 molecules
were further affinity purified with a mAb M46 column, as described above,
for the membrane-bound isoform of the protein. The SQ10 complexes were
eluted from affinity columns and allowed to dissociate by treatment of the
slurry with 10% acetic acid. The mixture of released peptides was separated from the high m.w. and RP-HPLC was purified as described above
for MQ10.
PEPTIDE LIGANDS OF Q10
The Journal of Immunology
1909
noprecipitates, suggesting that these two types of class I chains
differ in their ability to associate with endogenous ␤2m. There are
no Q10/Qa-2 differences involving the putative interdomain contact residues affecting the interaction of the ␣3 domain with ␤2m
or between the ␣3 domain with the ␣1␣2 domains, assuming homologous interactions to those found by x-ray crystallography for
HLA-A2 (49). Therefore, it is likely that the observed lower affinity of Q10 for murine ␤2m and its replacement with bovine ␤2m
(see high levels of unlabeled ␤2m recovered from Q10 complexes
in Fig. 2B and see below) is caused by Q10 specific residues in the
␣1 and ␣2 domains that contact ␤2m directly. These may involve
residues 6 and 9 of ␣1 and residue 116 of ␣2.
Partial TAP dependence of MQ10 membrane expression
The m.w. of MQ10 and SQ10 hybrids, their association with
␤2m, and GPI attachment of MQ10 were tested as follows. RMA
cells transfected with hybrid Q10 constructs and control Qa-2
genes were biosynthetically labeled with [35S]methionine and cysteine and, where appropriate, treated with PI-PLC, which specifically cleaves GPI-linked molecules and releases them from the cell
surface (Fig. 2A). The supernatants containing SQ10, PI-PLCcleaved MQ10, and control Qa-2 molecules were immunoprecipitated with anti-␣3 Qa-2 Ab MAb 46 (34). As predicted, all of the
analyzed Q10 and Qa-2 molecules migrated in SDS-polyacrylamide gels with apparent mass of 39 – 40 kDa (Fig. 2A). The SQ10
heavy chain migrates slightly faster in SDS gel than MQ10 released by PI-PLC treatment, which is in agreement with similar
observations for SQa-2 and MQa-2 (5). The 39- to 40-kDa weight
estimates are compatible with two carbohydrate moieties attached
to the mature Q10 and Qa-2 proteins at the putative N-linked glycosylation positions at residues 86 and 256 (3).
Both SQ10 and PI-PLC-cleaved MQ10 coimmunoprecipitated
with biosynthetically labeled murine ␤2m (Fig. 2A). The relative
proportions of radioactively labeled murine ␤2m precipitated with
Q10 heavy chain were lower than those observed in Qa-2 immu-
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
FIGURE 2. Hybrid Q10 proteins have the predicted m.w. and are associated with ␤2m. A, Biochemical analysis of Q10 proteins expressed in
transfected RMA cells. MQ10- and SQ10-transfected and control cells
were biosynthetically labeled for 4 h with [35S]methionine and cysteine,
and supernatants collected from an equal number of cells were analyzed by
immunoprecipitation with mAb 46. The precipitated molecules in lanes
1–10 originated from supernatants of the following cells: lanes 1 and 2, two
independent RMA transfectants of SQ10 (P29-4.4 and P29-4.1); lane 3,
RMA cells transfected with soluble Qa-2 carrying the 6xHis-tag at the C
terminus (P29-3.4); lane 4, RMA cells synthesizing soluble Qa-2 without
the 6xHis-tag (P1-6.60); lane 5, PI-PLC-treated RMA cells expressing
MQ10 (M15-3); lane 6, RMA cells expressing MQ10 (M15-3) that were
not treated with PI-PLC; lane 7, PLC-treated RMA transfectants expressing membrane-bound Qa-2 (P5-5.6); lane 8, RMA transfectants expressing
membrane-bound Qa-2 (P5-5.6) that were not treated with PI-PLC; and
lanes 9 and 10, control untransfected RMA cells treated (9) and untreated
(10) with PI-PLC. B, Coomassie blue-stained and SDS-PAGE-resolved
purified SQ10. The heavy chain and ␤2m are indicated by arrows. The
additional bands correspond to heavy and light Ig chains which were also
released from the mAb 46 immunoaffinity column. Coomassie blue-stained
␤2m may include murine and bovine species, whereas radioactively labeled
␤2m in Fig. 2A corresponds to endogenously synthesized murine ␤2m only.
Multiple studies have demonstrated that mutations in the TAP
genes, that direct synthesis of the peptide transporter molecules in
the class I Ag presentation pathway, lead to reduced levels of
classical class I Ags on the cell surface (37, 50). This phenotype is
thought to result from limiting quantities of peptide ligands delivered to the endoplasmic reticulum in TAP-negative mutants and
the resulting instability of “empty” class I-␤2m complexes. In most
cases, the decrease in membrane expression can be reversed by
low temperature (⬃26°C), which stabilizes peptide-free class I
complexes that reach the cell surface in TAP-negative cells.
To address the question of whether MQ10 associates with TAPdelivered peptides, we introduced MQ10 into TAP2-negative
RMA-S cells and compared its expression to the parental RMA
cells by FACS staining (Fig. 3). The control H-2Kb Ag coexpressed on MQ10 tranfectants displayed classical TAP-dependent
behavior: ⬃4-fold reduced expression in RMA-S vs RMA cells at
37°C and 42°C and ⬃8-fold induction of H-2Kb levels in RMA-S
cells at 26°C (38). Qa-2 Ag showed a more drastic reduction of
surface levels in RMA-S compared with RMA cells (12–14-fold)
at 37°C and 42°C (38). This expression was enhanced only weakly
at 26°C, in agreement with our previous data showing that most of
the empty Qa-2 fail to reach the cell surface in TAP2-negative
cells and accumulate intracellularly (41). MQ10 expression
showed a TAP2-dependent phenotype intermediate between
H-2Kb and Qa-2. Surface MQ10 levels were reduced (⬃5-fold in
RMA-S vs RMA cells) and were only weakly inducible at 26°C.
Thus, compared with wild-type Qa-2, MQ10 contains a somewhat
larger fraction (⬃20%) of heat-stable molecules that reach the cell
surface in a TAP2-independent fashion in RMA-S cells. This phenotype is most likely controlled by the structural properties of the
␣1 and ␣2 domains of Q10.
To confirm that the majority of MQ10 complexes in TAP2negative cells remain intracellular and behave as heat-unstable
empty molecules, we analyzed MQ10 from lysates of RMA-S
transfectants using conformation-dependent ELISA. The data in
Fig. 4 show that almost all MQ10 molecules, as well as the control
H-2Kb and Qa-2 molecules released from lysates of transfected
RMA-S cells, loose conformational epitopes upon incubation at
42°C for 80 min, whereas the majority of RMA-expressed complexes are stable under the same conditions. The partial loss of
conformational epitopes (in ⬃30% of the RMA-derived complexes) may be indicative of empty molecules, which accumulate
intracellularly because they have not been loaded with peptides or,
alternatively, may reflect the fact that some molecules associate
with peptides of low affinity that are released upon heat shock.
Taken together, these observations are consistent with the notion
that the majority (⬃80%) of mature MQ10 molecules require a
functional TAP pathway for cell surface expression. This property
suggests that ␣1␣2 of MQ10 molecules are peptide loaded.
1910
PEPTIDE LIGANDS OF Q10
Isolation and sequencing of peptides associated with membranebound and soluble Q10 proteins
The MQ10 and SQ10 complexes expressed in RMA cells were
purified by immunoaffinity chromatography and the sequences of
several endogenously bound peptides were determined by tandem
mass spectrometry (MS/MS).
Because of the different properties of the secreted and membrane-bound class I Ags, the two Q10 complexes were purified
using slightly different approaches (see Materials and Methods).
The SQ10 complexes were purified from tissue culture medium
using metal affinity chromatography followed by immunoaffinity
chromatography using the anti-␣3-specific mAb 46. In our previous studies with human MHC class I molecules, we routinely
quantitated the amount of purified complex by measuring the concentration of ␤2m that was retained by the ultrafiltration membrane
used to separate the peptides from intact proteins. The intact proteins retained in the high m.w. fraction were separated by RPHPLC, and the amount of ␤2m present was estimated by both
Edman degradation and absorbance at 280 nm. In the SQ10 preparation both the ␤2m and the SQ10 heavy chain were readily detected, allowing us to estimate that about 4 nmol of complex had
been purified. Electrospray ionization mass spectrometry (ESI/
MS) analysis of the fraction containing the SQ10 heavy chain
yielded a molecular mass of 38,874 Da, about 4850 mass units
greater than expected for the unglycosylated molecule (molecular
mass, 34,026 Da). This mass difference is most likely due to Nlinked carbohydrate moieties at Asn86 and Asn256, and it could be
accounted for by two triantennary carbohydrate structures. After
digestion with N-glycosidase F, in fact, two components with molecular masses of 36,518 and 34,170 Da, respectively, were detected, most likely corresponding to a partially and a completely
deglycosylated form of the protein. When the fraction containing
␤2m was analyzed by Edman sequencing, a mixed sequence was
obtained, indicating that about 20% murine ␤2m and 80% bovine
␤2m was present; ESI/MS analysis detected only bovine ␤2m (molecular mass, 11,632). The preferential association of the SQ10
heavy chain with bovine ␤2m is consistent with the observation
that murine ␤2m undergoes exchange with other species of ␤2m
present in the medium (Fig. 2). The long period of incubation of
the SQ10 complex in the FCS-supplemented tissue culture supernatants before purification may account for the observed high proportion of bovine ␤2m in SQ10 complexes.
MQ10 hybrid molecules were immunoaffinity-purified using
mAb 46 specific for the ␣3 domain of Qa-2. In this preparation, to
our surprise, no ␤2m or MQ10 heavy chain was recovered from the
ultrafiltration membrane, making it impossible to quantify the
amount of class I complexes that had been purified, even though
peptides could easily be detected (see below).
The peptides associated with both MQ10 and SQ10 were acid
extracted and separated by narrow-bore HPLC. The HPLC profile
of the MQ10-associated peptides is shown in Fig. 5A. An enlarged
view of the region containing the majority of the eluted peptides is
shown in Fig. 5B. The anticipated peptide-containing fractions
were analyzed by ESI/MS. The ESI/MS analysis showed the presence of at least 50 peptides for MQ10 and 110 peptides for SQ10
(data not shown) whose molecular mass fell into the mass range
appropriate for 8 –11-mer peptides, many of which were present in
both samples. The larger number of peptide signals detected for
SQ10 may reflect the ability of the soluble form to bind a larger
number of peptides. Alternatively, because we were not able to
quantify the amount of MQ10 complexes purified, it might simply
indicate that a larger quantity of purified SQ10 complex was available for analysis.
CAD analysis (51) performed on selected ions present in both
the MQ10- and SQ10-purified material defined plausible amino
acid sequences for six peptides (Table I). A representative CAD
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
FIGURE 3. Staining of MQ10 and control MQa-2 and H-2Kb molecules on transfected RMA (A) and RMA-S (B) cells at 26, 37, and 42°C. MQ10
transfectants correspond to sorted, mixed populations of RMA and RMA-S clones expressing the top 5% of MQ10 levels from each transfected cell
line. Fluorescence peak channels for MQ10/RMA (mAb 46) were 1715 (26°C), 1703 (37°C), and 1911 (42°C); for MQ10/RMA-S (mAb 46) were
437 (26°C), 365 (37°C), and 352 (42°C); for MQa-2/RMA (mAb 46) were 1433 (26°C), 1596 (37°C), and 1590 (42°C); for MQa-2/RMA-S (mAb
46) were 154 (26°C), 133 (37°C), and 111 (42°C); for H-2Kb/RMA (mAb Y3) were 1286 (26°C), 626 (37°C), and 523 (42°C); for H-2Kb/RMA-S
(mAb 46) were 1197 (26°C), 178 (37°C), and 149 (42°C); FITC only, ⬃15 (37°C).
The Journal of Immunology
1911
FIGURE 4. Intracellular MQ10 molecules are unstable at high temperature in RMA-S cells. The concentrations of folded MQ10, H-2Kb, and
MQa-2 molecules were measured in cell lysates of transfected RMA (A)
and RMA-S cells (B) using conformation-dependent sandwich ELISA. The
prechilled lysates were heat shocked at 42°C for the indicated time intervals. The data were standardized so that the time point 0 (no heat shock)
corresponds to 100% expression of the conformational epitopes on folded
molecules at 4°C.
spectrum is shown in Fig. 6 for the peptide with mass:charge ratio
(m/z) of 920.2 (peptide 5, Table I). By this means, complete sequences were obtained for several peptides. For peptide
QGVQXXDF (peptide 5) assignment of Q vs K (which are of
nearly identical mass) was made by MS analysis following acetylation. This derivatization resulted in a single 42 mass unit shift,
leading us to conclude that the N-terminal amino group is the only
amino group present in the peptide. Four of the six peptides were
eight amino acids long, and two were nine amino acids long, suggesting that depending on the sequence, Q10 preferentially binds
octamers but occasionally nonamers, similar to H-2Kb, -Kk (2) and
HLA-B8 (52). All six of the peptides contained Gly at P2, five of
the six peptides contained Lxx at P6, and at P␻ all peptides contained a hydrophobic residue: either Phe (in four sequences) or
Lxx. The other positions of the peptides were more variable.
One of the largest peaks in the absorbance trace, at 34 min (Fig.
5), contained peptide TGTETXYF. Unlike most of the other peaks
(some of the larger of which were present both in the nonspecific
material eluted from glycine-Sepharose and in the Q10 mAb eluate), the peak at 34 min had an UV spectrum with a maximum at
278 nm, typical for peptides containing tyrosine residues, leading
to the conclusion that the large absorbance of the material in this
peak is largely due to peptide TGTETXYF and not to unrelated
molecules. On the basis of absorbance and Edman degradation
data (which matched the MS/MS sequence with Leu at P6),
we concluded that there was about 250 –300 pmol of peptide
TGTETLYF in the MQ10 preparation and about 1 nmol in the
SQ10 preparation; all other peptides were 60- to 100-fold less
abundant. These estimates suggest that the TGTETLYF peptide
may constitute as much as ⬃25% of the total Q10 ligand pool.
Synthetic peptides corresponding to the sequences of the
constitutively bound peptides form complexes with MQ10 in
vitro
When gene and protein sequence databases were searched for possible parent proteins of the Q10-specific peptides, four peptides
from Table I matched murine protein sequences. Interestingly, all
of these putative proteins correspond to fairly abundant polypeptides. Peptide 1 is identical to an octameric sequence present
within two distinct proteasome subunits: constitutively expressed
PSMB5 (53) and IFN-␥ regulated LMP7 (54). Peptide 2 is homologous to ribophorin (accession number D31717.1) and peptide 6 to
cytochrome c oxidase (accession number P43024). The putative
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
FIGURE 5. RP-HPLC separation of MQ-10 associated peptides. A, RPHPLC chromatogram of peptides eluted from MQ10 complexes following
immunoaffinity purification. B, Enlarged view of the elution profile. Fractions eluted between 20 and 45 min were analyzed by ESI/MS. Superimposed on the profile for the MQ10-associated peptides (top line) is the
corresponding region of the chromatogram for material obtained upon elution from a nonspecific adsorbent consisting of glycine coupled to Sepharose (bottom line).
1912
PEPTIDE LIGANDS OF Q10
Table I. Sequences of Q10-binding peptides are homologues to endogenous murine proteins
Position
Peptide
m/z
1
2
3
4
5
6
7
8
1
2
3
4
5
6
848.1
831.3
932.0
843.1
920.2
966.2
H
V
T
Xa
Q
V
G
G
G
G
G
G
T
I
T
A
V
V
T
T
E
A
Q
S
T
N
T
X
X
M
L
V
L
X
X
L
A
D
Y
G
D
N
F
L
F
D
F
V
P6
L/V
P7
9
Protein of Origin
Proteasome subunits
Riboforin
TCR V␤ chain
X
F
Cytochrome C oxidase
Motif
P1
a
P2
G
P3
P4
P5
P8/P␻
F/L
X indicates Leu or Ile residues which are not distinguishable by tandem MS.
FIGURE 6. CAD mass spectrum of peptide ions at m/z 920.2 (peptide
5). Predicted masses for fragment ions of types b and y (51) are shown
above and below the deduced sequence, respectively. Ions observed in the
spectrum are underlined. Interpretation of CAD spectra is fully explained
elsewhere (51). Because Ile and Leu are of identical mass, they cannot
be differentiated on the triple quadrupole instrument and are specified here
as Lxx.
To verify that the sequences obtained in this study represent
genuine endogenous peptides that can associate specifically with
Q10, the synthetic homologues were tested in an in vitro peptidebinding assay. The assay measured the peptide-dependent stabilization of MQ10 epitopes on the ␣3 domain (recognized by mAb
46) and murine ␤2m (recognized by mAb S19.8) by sandwich
ELISA. The signal:background ratio of this assay is lower than
the one observed with MQa-2 (data not shown). This effect may
be explained by preferential displacement of murine ␤2m from
MQ10 heavy chain by bovine ␤2m and/or by higher background
of “temperature-resistant” MQ10 complexes formed in transfected RMA-S cells. Five of the six synthetic peptides stabilized
the MQ10/␤2m complexes over a wide range of peptide concentrations: 100 ng to100 ␮g, as shown in Figs. 7 and 8. Although the
peptide stabilization assay cannot be regarded as a rigorous measurement of peptide affinity, the half-maximal and maximal points
FIGURE 7. Stabilization of MQ10 conformational epitopes with Q10
synthetic peptides. Group A peptides correspond to synthetic homologues
of Q10 peptides listed in Table I, B is a poly(A) nonamer AAAAAAAAA,
group C peptides are (left to right): RYWAIRTRS, FRYNGLIHL,
RYWATRSGG, and GRIDKPILK. All peptides have been used at saturating concentrations of 20 ␮g/ml. The values shown correspond to the
mean of triplicate experiments and are expressed in arbitrary units relative
to the internal ELISA standard.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
source protein of the most abundant peptide (peptide 3) corresponds to TCR V ␤-chain (55) and is the only source protein which
would not be normally expressed in liver cells. Its presence in the
Q10-transfected RMA cells is consistent with the lymphoma phenotype of the parental line. All of the putative source proteins are
expressed intracellularly, suggesting that their peptide components
were introduced into the Q10 grooves by the class I Ag presentation pathway and were not incorporated into the complexes during
the purification procedure from extracellular sources such as tissue
culture medium. The alignment of peptide sequences 1, 2, and 6
with their putative sources allowed us also to assign Leu and Ile for
these three peptides.
Peptides corresponding to the six sequences reported in Table I
were synthesized using Leu as the default amino acid in positions
where Leu could not be distinguished from Ile (reported as X in
Table I). In each case, CAD fragmentation spectra were identical
to the spectra derived from the Q10-associated peptides, confirming that the deduced sequences were concordant by this criteria.
However, when the RP-HPLC retention times of the six synthetic
peptides were compared with those obtained for the endogenous
Q10 peptides, peptides 4 and 5 showed a higher retention time than
expected, likely due to the presence of Ile instead of Leu at some
positions. Due to a large number of potential permutations of Leu
and Ile in peptides 4 and 5, we have not synthesized additional
candidate peptides.
The Journal of Immunology
1913
FIGURE 8. Stabilization of conformational epitopes on MQ10 hybrid
molecules with titered synthetic peptides. The experiments were performed
under the same conditions as those in Fig. 7, except that a wide range of
peptide concentrations was used to stabilize MQ10 complexes. The values
shown correspond to the means of triplicate experiments and are expressed
in arbitrary units relative to the internal ELISA standard.
of concentration curves in Fig. 8 do not give any indication that the
dominant TGTETLYF peptide binds Q10 better than other titered
peptides. Hence, we conclude that the peptide-binding affinity of
TGTETLYF is comparable to other peptides tested by ELISA approach. Peptide 4, LGAALLGDL, was consistently negative in our
assay (comparable to negative controls in Fig. 7). This peptide
contains four Leu residues synthesized as default amino acids in
positions in which Leu could not be distinguished from Ile in the
endogenous Q10 sequence.
Q10-associating peptides display classical peptide-binding motif
MS sequencing of MQ10- and SQ10-eluted peptides suggested
that these molecules bind a heterogeneous mixture of diverse, endogenously synthesized ligands, which occupy as much as ⬃75%
of all Q10 receptors. The remaining ⬃25% of Q10 molecules are
filled with a single peptide species TGTETLYF. This dual affinity
of Q10 molecules prompted us to examine sequence requirements
of the Q10 ligands for binding to Q10 groove. We reasoned that a
groove that is severely biased toward accepting peptides with defined sequences will be less efficient in associating with mutant
peptides which carry single residue substitutions along the entire
length of the peptide. This effect has been observed for Qdm peptide which is the dominant peptide in Qa-1b molecule (14, 56). If,
on the other hand, the groove can accommodate many diverse
peptides then loss/reduction of binding will be observed only when
the test peptide is mutated at the classically defined “anchor”
positions.
To address this question, we selected peptide 1, HGTTTLAF
(homologous to subunits of proteasome), for the analysis. This
peptide, unlike TGTETLYF (homologous to TCR V ␤-chain),
would be present in liver cells and is identical to TGTETLYF in
five of eight positions. A series of synthetic peptides substituted by
Ala or Ser at each of the positions was synthesized (see legend to
Fig. 9), and the peptides were used in the ELISA sandwich peptide-binding assay. The results of the binding experiments indicated that only three peptide residues could not tolerate being replaced with Ala for efficient binding to MQ10: Gly at P2, Leu at
P6, and Phe at P8. Substitutions of these residues with Ala led to
either reduction or loss of binding comparable to negative control
peptides VSV, L19, and NP (Fig. 9). The three anchor residues
correspond to the conserved residues determined from the MS sequencing (Table I). Thus, we conclude that the majority of the
peptides associating with the Q10 groove display a classical peptide-binding motif similar to the diverse repertoire of ligands that
occupy the H-2K, H-2D, or HLA-A and -B grooves.
Discussion
Recent research into the functions of the nonclassical class Ib
MHC Ags led to the recognition of their diversity and heterogeneous properties; however, despite the fact that much has been
learned about immunological features of murine M3, Qa-1, and
human CD1, HLA-E, and HLA-G Ags, the great majority of the
class Ib molecules remain uncharacterized. One such molecule is
the soluble, liver-specific Q10 protein that is expressed in a wide
variety of inbred and wild mouse strains.
In an attempt to learn about Ag-presenting functions of Q10
proteins, we analyzed peptide ligands constitutively associated
with the Q10 ␣1␣2 domains. Because of the technical limitations
imposed by the necessity to produce large amounts of this protein,
we expressed and analyzed hybrid Q10/Qa-2 molecules in lymphoid-derived cells. The results of the MS sequencing of the Q10associated ligands revealed that they are very similar to the processed protein fragments eluted from classical class I Ags. As is
the case with H-2Kb or H-2Kk peptides (2), the majority of the Q10
ligands are octameric (although nonamers were also detected). The
Q10 peptides carry a peptide-binding motif typical of the class Ia
motifs. The conserved residues include a hydrophobic (Phe or Leu)
dominant anchor at P␻ and two additional invariant residues: Gly
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
FIGURE 9. Identification of anchor positions in HGTTTLAF peptide.
The wild-type (WT) and mutant synthetic peptides substituted at different
residues by Ala (1A-6A, 8A) or by Ser (7S) were tested in peptide stabilization assay for binding to MQ10. The experiment was performed as in
Fig. 7. The dark and light bars correspond to ELISA values measured with
20 ␮g and 4 ␮g of the synthetic peptides, respectively. The sequences of
negative control peptides VSV, L19, and NP are listed elsewhere (41). The
values shown correspond to the means of triplicate experiments.
1914
iments of TGTETLYF-related peptide identified only three conserved anchor residues, at P2, P6, and P8, that are the only prerequisition for peptide binding. In cases where the cleft is
preferentially occupied by a single peptide (56, 60), all peptide
positions affect binding efficiency to a detectable degree.
Liver cells, in which Q10 is normally synthesized, do not express TCR. Thus, any potential bias of Q10 groove for TCR-derived peptide cannot be easily rationalized, particularly because the
region of homology corresponds to the highly variable CDR3 region embedded within the TCR cleft. Nevertheless, we cannot exclude the possibility that the Q10 groove binds under some circumstances, in lieu of normally processed peptide, a fragment of
␤-chain looping of the TCR complex on T cells. A precedent for
this interaction was recently reported for class II MHC and the
TCR ␣-chain (61).
The nature of T cell-mediated recognition of Q10 has been addressed before (59, 62). Since Q10 protein shares structural features with many class I MHC proteins (59, 63), alloreactive CTLs
raised against Q10 ␣1␣2 domains cross-react on multiple class I
MHC proteins (59, 62). This property may allow Q10 to interact
with a broader range of TCRs than is normally expected for classical class I Ags. Whether such interactions occur in vivo and
whether they lead to apoptosis of T cells, as reported for soluble
classical class I proteins interacting with TCRs (64), remains to be
established. In this regard, it is of interest that liver is the major
organ in which T cell death occurs (65).
Association of the Q10 groove with a diverse array of peptides
and the similarity of these complexes with classical class I MHC
raises another question, namely, whether these proteins can interact with receptors on NK cells. Many different families of NK, B
cell, and monocyte receptors recognizing classical and nonclassical (Qa-1, HLA-E, HLA-G) class I complexes were identified in
the recent years (17–19, 66). Some of these receptors bind class I
complexes in peptide-dependent fashion while other associations
are peptide independent. Because liver is very rich in NK cells, it
is of interest to examine whether Q10 proteins engage in specific
interactions with NK cell receptors. New experimental approaches
such as tetramer staining (67) may allow us to address those issues
in the near future.
Acknowledgments
We thank Maile Henson for tissue culture and help with characterization of
the transfectants and Dr. Ming Chen for screening of the protein data banks
with Q10 peptide sequences.
References
1. Engelhard, V. H. 1994. Structure of peptides associated with class I and class II
MHC molecules. Annu. Rev. Immunol. 12:181.
2. Rammensee, H. G., T. Friede, and S. Stevanoviic. 1995. MHC ligands and peptide motifs: first listing. Immunogenetics 41:178.
3. Stroynowski, I. 1990. Molecules related to class-I major histocompatibility complex antigens. Annu. Rev. Immunol. 8:501.
4. Flaherty, L., E. Elliott, J. A. Tine, A. C. Walsh, and J. B. Waters. 1990. Immunogenetics of the Q and TL regions of the mouse. Crit. Rev. Immunol. 10:131.
5. Ulker, N., K. D. Lewis, L. E. Hood, and I. Stroynowski. 1990. Activated T cells
transcribe an alternatively spliced mRNA encoding a soluble form of Qa-2 antigen. EMBO J. 9:3839.
6. Ishitani, A., and D. E. Geraghty. 1992. Alternative splicing of HLA-G transcripts
yields proteins with primary structures resembling both class I and class II antigens. Proc. Natl. Acad. Sci. USA 89:3947.
7. Shawar, S. M., J. M. Vyas, J. R. Rodgers, and R. R. Rich. 1994. Antigen presentation by major histocompatibility complex class I-B molecules. Annu. Rev.
Immunol. 12:839.
8. Stroynowski, I., and K. Fisher-Lindahl. 1994. Antigen presentation by non-classical class I molecules. Curr. Opin. Immunol. 6:38.
9. O’Callaghan, C. A., and J. I. Bell. 1998. Structure and function of the human
MHC class Ib molecules HLA-E, HLA-F and HLA-G. Immunol. Rev. 163:129.
10. Maher, J. K., and M. Kronenberg. 1997. The role of CD1 molecules in immune
responses to infection. Curr. Opin. Immunol. 9:456.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
at P2 and Leu/Val at P6. All three of these positions influence
binding of Q10 synthetic peptide homologues to Q10 groove. The
residues found at P2, P6, and P␻ on Q10 ligands have been reported to serve as anchors in peptides eluted from other class I
MHC Ags (2). This is not surprising considering the fact that the
predicted geometry of the Q10 groove is very similar to HLA-A2.
Although Q10 ␣1␣2 domains contain a number of unique substitutions that are not commonly found in other class I MHC proteins
(at positions 24, 75, 89, 90, 102, 109, 137, 162, and 176), only one
of them, Ile 24, is located at a position predicted to face the peptide-binding groove.
One slightly unusual feature of Q10 peptides is that the invariant
Gly at P2 is followed by a variant amino acid at P3. The only other
class I molecules for which Gly has been deduced to be critical for
binding are H-2Dd, where Gly at P2 is nearly invariably paired
with Pro at P3 (57) and HLA-B51, where Gly at P2 is almost
always paired with an aromatic residue at P3 (2). Because Gly has
no side chain and therefore cannot function as an anchor residue
directly, but instead promotes local flexibility and destabilization
of the peptide binding, it is possible that all four amino acids found
at P3 (Ala, Ile, Val, and Thr) play an important role in anchoring
of peptides to Q10 groove. The Gly anchor residue at P2 would be
expected to correlate with large side chains in the B pocket of the
peptide-binding cleft. The only unusual B pocket residue is Ile-24,
which is only found in Q8 (45), whose motif has not been determined, and in Qa-1, which is occupied predominantly with a single
peptide species carrying Met at P2 (14).
The nature of the putative source proteins giving rise to Q10
peptides warrants some discussion. All identified sequences
matched intracellular murine proteins: the LMP7 and PSMB5 proteasomal subunits, ribophorin, cytochrome c oxidase, and TCR V
␤-chain. This is in agreement with the notion that the peptides
bound to Q10 groove originated from cytoplasmic proteins and
were delivered to the complex by components of classical class I
MHC Ag presentation pathway. Consistent with this interpretation
we found that mutation of the TAP2 gene led to significant reduction of heat-resistant, peptide-filled Q10 molecules expressed on
the surface of RMA-S cells. The small proportion of thermally
stable MQ10 on RMA-S cells was comparable to H-2Kb expressed
in the same background and may correspond to MHC complexes
loaded by TAP1/TAP1 homodimers (58).
The identification of the degraded product of TCR V ␤-chain as
the most abundant peptide in the Q10 groove in RMA cells was the
unexpected finding of this study. Edman degradation and absorbance data suggested that peptide TGTETLYF occupied as much
as 25% of RMA-expressed Q10 molecules, whereas the remaining
75% of Q10 grooves were filled with a highly heterogeneous mixture of low abundance peptides. Although it is possible that the
homology of this peptide to TCR V ␤-chain from EL4 cells (55) is
serendipitous, it is more likely that it reflects the precursor-product
relationship because RMA is a lymphoma cell line and may express the same TCR V ␤-chain as EL4. Peptide-binding studies
reported here demonstrated that TGTETLYF has similar binding
affinity to Q10 cleft as other peptides examined in this study.
Hence, we conclude that overrepresentation of this peptide in
MQ10 grooves may have been brought about by preferential processing of this peptide or its enhanced delivery to the endoplasmic
reticulum, rather than preferential binding to MQ10. Two other
lines of evidence support the conclusion that Q10 does not have a
highly “specialized” binding groove. First, the modeling studies of
Q10 reported previously (59) suggested that the Q10 groove is
very similar to HLA-A2, although it may be somewhat shallow
due to the presence of multiple bulky residues (Tyr at 99, 155, 156,
and 159 and Trp at 97 and 167). Second, alanine-scanning exper-
PEPTIDE LIGANDS OF Q10
The Journal of Immunology
39. Kilmartin, J. V., B. Wright, and C. Milstein. 1982. Rat monoclonal antitubulin
antibodies derived by using a new nonsecreting rat cell line. J. Cell Biol. 93:576.
40. Edwards, P. A., C. M. Smith, A. M. Neville, and M. J. O’Hare. 1982. A humanhybridoma system based on a fast-growing mutant of the ARH-77 plasma cell
leukemia-derived line. Eur. J. Immunol. 12:641.
41. Tabaczewski, P., E. Chiang, M. Henson, and 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.
42. Zappacosta, F., F. Borrego, A. G. Brooks, K. C. Parker, and J. E. Coligan. 1997.
Peptides isolated from HLA-Cw*0304 confer different degrees of protection from
natural killer cell-mediated lysis. Proc. Natl. Acad. Sci. USA 94:6313.
43. Parker, K. C., M. DiBrino, L. Hull, and J. E. Coligan. 1992. The ␤2-microglobulin dissociation rate is an accurate measure of the stability of MHC class I heterotrimers and depends on which peptide is bound. J. Immunol. 149:1896.
44. Mellor, A. L., J. Antoniou, and P. J. Robinson. 1985. Structure and expression of
genes encoding murine Qa-2 class I antigens. Proc. Natl. Acad. Sci. USA 82:
5920.
45. Devlin, J. J., E. H. Weiss, M. Paulson, and R. A. Flavell. 1985. Duplicated gene
pairs and alleles of class I genes in the Qa2 region of the murine major histocompatibility complex: a comparison. EMBO J. 4:3203.
46. Stroynowski, I., M. Soloski, M. G. Low, and 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.
47. Straus, D. S., I. Stroynowski, S. G. Schiffer, and 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.
48. Stroynowski, I., S. Clark, L. A. Henderson, L. Hood, M. McMillan, and
J. Forman. 1985. Interaction of ␣1 with ␣2 region in class I MHC proteins
contributes determinants recognized by antibodies and cytotoxic T cells. J. Immunol. 135:2160.
49. Bjorkman, P. J., M. A. Saper, B. Samraoui, W. S. Bennett, J. L. Strominger, and
D. C. Wiley. 1987. Structure of the human class I histocompatibility antigen,
HLA-A2. Nature 329:506.
50. Attaya, M., S. Jameson, C. K. Martinez, E. Hermel, C. Aldrich, J. Forman,
K. F. Lindahl, M. J. Bevan, and J. J. Monaco. 1992. Ham-2 corrects the class I
antigen-processing defect in RMA-S cells. Nature 355:647.
51. Hunt, D. F., J. R. Yates, 3rd, J. Shabanowitz, S. Winston, and C. R. Hauer. 1986.
Protein sequencing by tandem mass spectrometry. Proc. Natl. Acad. Sci. USA
83:6233.
52. Malcherek, G., K. Falk, O. Rotzschke, H. G. Rammensee, S. Stevanovic,
V. Gnau, G. Jung, and A. Melms. 1993. Natural peptide ligand motifs of two
HLA molecules associated with myasthenia gravis. Int. Immunol. 5:1229.
53. Kohda, K., Y. Matsuda, T. Ishibashi, K. Tanaka, and M. Kasahara. 1997.
Structural analysis and chromosomal localization of the mouse Psmb5 gene
coding for the constitutively expressed ␤-type proteasome subunit. Immunogenetics 47:77. 32
54. Meinhardt, T., U. Graf, and G. J. Hammerling. 1993. Different genomic structure
of mouse and human Lmp7 genes: characterization of MHC-encoded proteasome
genes. Immunogenetics 38:373.
55. Shi, Y., A. Kaliyaperumal, L. Lu, S. Southwood, A. Sette, M. A. Michaels, and
S. K. Datta. 1998. Promiscuous presentation and recognition of nucleosomal
autoepitopes in lupus: role of autoimmune T cell receptor ␣ chain. J. Exp. Med.
187:367.
56. Kurepa, Z., C. A. Hasemann, and J. Forman. 1998. Qa-1b binds conserved class
I leader peptides derived from several mammalian species. J. Exp. Med. 188:973.
57. Corr, M., L. F. Boyd, E. A. Padlan, and D. H. Margulies. 1993. H-2Dd exploits
a four residue peptide binding motif. J. Exp. Med. 178:1877.
58. Gabathuler, R., G. Reid, G. Kolaitis, J. Driscoll, and W. A. Jefferies. 1994. Comparison of cell lines deficient in antigen presentation reveals a functional role for
TAP-1 alone in antigen processing. J. Exp. Med. 180:1415.
59. Mann, D. W., E. McLaughlin-Taylor, R. B. Wallace, and J. Forman. 1988. An
immunodominant epitope present in multiple class I MHC molecules and recognized by cytotoxic T lymphocytes. J. Exp. Med. 168:307.
60. O’Callaghan, C. A., J. Tormo, B. E. Willcox, V. M. Braud, B. K. Jakobsen,
D. I. Stuart, A. J. McMichael, J. I. Bell, and E. Y. Jones. 1998. Structural features
impose tight peptide binding specificity in the nonclassical MHC molecule
HLA-E. Mol. Cell 1:531.
61. Qadri, A., J. Thatte, and S. E. Ward. 1999. Characterization of the interaction of
a TCR ␣ chain variable domain with MHC II I-A molecules. Int. Immunol. 11:6.
62. Mann, D. W., and J. Forman. 1988. A third class I major histocompatibility
complex antigen encoded by a gene in the D region of the H-2d haplotype recognized by cytotoxic T lymphocytes. Immunogenetics 28:38.
63. Pease, L. R., D. H. Schulze, G. M. Pfaffenbach, and S. G. Nathenson. 1983.
Spontaneous H-2 mutants provide evidence that a copy mechanism analogous to
gene conversion generates polymorphism in the major histocompatibility complex. Proc. Natl. Acad. Sci. USA 80:242.
64. Zavazava, N., and M. Kronke. 1996. Soluble HLA class I molecules induce
apoptosis in alloreactive cytotoxic T lymphocytes. Nat. Med. 2:1005.
65. Crispe, I. N., and W. Z. Mehal. 1996. Strange brew: T cells in the liver. Immunol.
Today 17:522.
66. Colonna, M., F. Navarro, T. Bellon, M. Llano, P. Garcia, J. Samaridis,
L. Angman, M. Cella, and M. Lopez-Botet. 1997. A common inhibitory receptor
for major histocompatibility complex class I molecules on human lymphoid and
myelomonocytic cells. J. Exp. Med. 186:1809.
67. Altman, J. D., P. A. H. Moss, P. J. R. Goulder, D. H. Barouch,
M. G. McHeyzer-Williams, J. I. Bell, A. J. McMichael, and M. M. Davis. 1996.
Phenotypic analysis of antigen-specific T lymphocytes. Science 274:94.
Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017
11. Ghetie, V., and E. S. Ward. 1997. FcRn: the MHC class I-related receptor that is
more than an IgG transporter. Immunol. Today 18:592.
12. Feder, J. N., A. Gnirke, W. Thomas, Z. Tsuchihashi, D. A. Ruddy, A. Basava,
F. Dormishian, R. Domingo, Jr., M. C. Ellis, A. Fullan, et al. 1996. A novel MHC
class I-like gene is mutated in patients with hereditary haemochromatosis. Nat.
Genet. 13:399.
13. Fisher-Lindahl, K., D. E. Byers, V. M. Dabhi, R. Hovik, E. P. Jones, G. P. Smith,
C. R. Wang, H. Xiao, and M. Yoshino. 1997. H2–M3, a full-service class Ib
histocompatibility antigen. Annu. Rev. Immunol. 15:851.
14. DeCloux, A., A. S. Woods, R. J. Cotter, M. J. Soloski, and J. Forman. 1997.
Dominance of a single peptide bound to the class I(B) molecule, Qa-1b. J. Immunol. 158:2183.
15. Braud, V., E. Y. Jones, and A. McMichael. 1997. The human major histocompatibility complex class Ib molecule HLA-E binds signal sequence-derived peptides with primary anchor residues at positions 2 and 9. Eur. J. Immunol. 27:
1164.
16. Kastner, D. L., R. R. Rich, and F. W. Shen. 1979. Qa-1-associated antigens. I.
Generation of H-2-nonrestricted cytotoxic T lymphocytes specific for determinants of the Qa-1 region. J. Immunol. 123:1232.
17. Vance, R. E., J. R. Kraft, J. D. Altman, P. E. Jensen, and D. H. Raulet. 1998.
Mouse CD94/NKG2A is a natural killer cell receptor for the nonclassical major
histocompatibility complex (MHC) class I molecule Qa-1(b). J. Exp. Med. 188:
1841.
18. Borrego, F., M. Ulbrecht, E. H. Weiss, J. E. Coligan, and A. G. Brooks. 1998.
Recognition of human histocompatibility leukocyte antigen (HLA)-E complexed
with HLA class I signal sequence-derived peptides by CD94/NKG2 confers protection from natural killer cell-mediated lysis. J. Exp. Med. 187:813.
19. Braud, V. M., D. S. Allan, C. A. O’Callaghan, K. Soderstrom, A. D’Andrea,
G. S. Ogg, S. Lazetic, N. T. Young, J. I. Bell, J. H. Phillips, et al. 1998. HLA-E
binds to natural killer cell receptors CD94/NKG2A, B and C. Nature 391:795.
20. Joyce, S., P. Tabaczewski, R. H. Angeletti, S. G. Nathenson, and I. Stroynowski.
1994. A nonpolymorphic major histocompatibility complex class Ib molecule
binds a large array of diverse self-peptides. J. Exp. Med. 179:579.
21. Lee, N., A. R. Malacko, A. Ishitani, M. C. Chen, J. Bajorath, H. Marquardt, and
D. E. Geraghty. 1995. The membrane-bound and soluble forms of HLA-G bind
identical sets of endogenous peptides but differ with respect to TAP association.
Immunity 3:591.
22. Stroynowski, I., and J. Forman. 1995. Novel molecules related to MHC antigens.
Curr. Opin. Immunol. 7:97.
23. Kvist, S., and P. A. Peterson. 1978. Isolation and partial characterization of a
␤2-microglobulin-containing, H-2 antigen-like murine serum protein. Biochemistry 17:4794.
24. Kimura, F., T. Asano, M. Murai, H. Nakamura, M. Yamamoto, M. Ohtsubo,
M. Ishizawa, Y. Koyanagi, and N. Yamamoto. 1993. Diversity in mRNA encoding soluble form MHC class I-like molecules in human renal tissues. Transplant.
Proc. 25:167.
25. Puppo, F., M. Scudeletti, F. Indiveri, and S. Ferrone. 1995. Serum HLA class I
antigens: markers and modulators of an immune response? Immunol. Today 16:
124.
26. Kress, M., D. Cosman, G. Khoury, and G. Jay. 1983. Secretion of a transplantation-related antigen. Cell 34:189.
27. Lew, A. M., W. L. Maloy, and J. E. Coligan. 1986. Characteristics of the expression of the murine soluble class I molecule (Q10). J. Immunol. 136:254.
28. Mellor, A. L., E. H. Weiss, M. Kress, G. Jay, and R. A. Flavell. 1984. A nonpolymorphic class I gene in the murine major histocompatibility complex. Cell
36:139.
29. Marine, J. B., Y. Shirakata, S. A. Wadsworth, J. J. Hooley, D. E. Handy, and
J. E. Coligan. 1993. Role of the Q10 class I regulatory element region 1 in
controlling tissue-specific expression in vivo. J. Immunol. 151:1989.
30. David-Watine, B., C. Transy, G. Gachelin, and P. Kourilsky. 1987. Tissue-specific expression of the mouse Q10 H-2 class-I gene during embryogenesis. Gene
61:145.
31. Girgis, K. R., J. D. Capra, and I. Stroynowski. 1995. Nucleotide sequences of
H2g7 K and D loci of nonobese diabetic mice. Immunogenetics 41:386.
32. Watts, S., A. C. Davis, B. Gaut, C. Wheeler, L. Hill, and R. S. Goodenow. 1989.
Organization and structure of the Qa genes of the major histocompatibility complex of the C3H mouse: implications for Qa function and class I evolution.
EMBO J. 8:1749.
33. Gastinel, L. N., N. E. Simister, and P. J. Bjorkman. 1992. Expression and crystallization of a soluble and functional form of an Fc receptor related to class I
histocompatibility molecules. Proc. Natl. Acad. Sci. USA 89:638.
34. Hasenkrug, K. J., J. M. Cory, and J. H. Stimpfling. 1987. Monoclonal antibodies
defining mouse tissue antigens encoded by the H-2 region. Immunogenetics 25:
136.
35. Tada, N., S. Kimura, A. Hatzfeld, and U. Hammerling. 1980. Ly-m11: the H-3
region of mouse chromosome 2 controls a new surface alloantigen. Immunogenetics 11:441.
36. Harlow, E., and Lane, D., eds. (1988) Antibodies. Cold Spring Harbor Lab. Press,
Plainview, NY.
37. Ljunggren, H. G., S. Paabo, M. Cochet, G. Kling, P. Kourilsky, and K. Karre.
1989. Molecular analysis of H-2-deficient lymphoma lines: distinct defects in
biosynthesis and association of MHC class I heavy chains and ␤2-microglobulin
observed in cells with increased sensitivity to NK cell lysis. J. Immunol. 142:
2911.
38. Tabaczewski, P., and I. Stroynowski. 1994. Expression of secreted and glycosylphosphatidylinositol-bound Qa-2 molecules is dependent on functional TAP-2
peptide transporter. J. Immunol. 152:5268.
1915