MILL (MHC class I-like located near the leukocyte receptor complex) is a family of MHC class I-like molecules encoded outside the MHC, which displays the highest sequence similarity to human MICA/B molecules among known class I molecules. In the present study, we show that the two members of the mouse MILL family, MILL1 and MILL2, are GPI-anchored glycoproteins associated with β2-microglobulin (β2m) and that cell surface expression of MILL1 or MILL2 does not require functional TAP molecules. MILL1 and MILL2 molecules expressed in bacteria could be refolded in the presence of β2m, without adding any peptides. Hence, neither MILL1 nor MILL2 is likely to be involved in the presentation of peptides. Immunohistochemical analysis revealed that MILL1 is expressed in a subpopulation of thymic medullary epithelial cells and a restricted region of inner root sheaths in hair follicles. The present study provides additional evidence that MILL is a class I family distinct from MICA/B.

Classical MHC class I molecules, also known as class Ia, are heterodimeric glycoproteins made up of a transmembrane-type H chain and β2-microglobulin (β2m).3 They bind small peptides primarily derived from cytosolic proteins in a groove comprised of the α1 and α2 domains and present them to CD8+ T cells, thereby enabling the immune system to destroy abnormal cells that synthesize viral or other foreign proteins (1). Class Ia molecules are almost ubiquitously expressed and their H chains exhibit an extraordinary level of polymorphism (2).

By contrast, class I molecules, collectively called nonclassical class I or class Ib, are usually oligomorphic or monomorphic, and do not necessarily bind peptides (3, 4, 5). Many class Ib molecules have a more restricted tissue distribution than class Ia molecules. Although the majority of class Ib molecules form complexes with β2m, MICA/B (MHC class I-related chains A and B) (6), zinc-α2-glycoprotein (7), the endothelial protein C receptor (8), and the RAE-1 (retinoic acid early inducible-1) family of class Ib molecules (9) are not associated with β2m. Furthermore, a significant proportion of class Ib genes (the genes coding for the H chains of class Ib molecules) are located outside the MHC region (5). Accumulated evidence indicates that class Ib molecules have diverse functions ranging from specialized Ag presentation (10, 11, 12) to the activation of NK cells (13, 14), transport of IgG (15), pheromone detection (16, 17), and lipid mobilization and catabolism (18).

Recently, we identified a new family of class Ib genes designated Mill (MHC class I-like located near the leukocyte receptor complex) in mice (19) and rats (20). The two members of the Mill family, Mill1 and Mill2, are located close to the leukocyte receptor complex, thus outside the MHC. Mill1 and Mill2 show only limited levels of polymorphism and are transcribed at low levels in most adult tissues. RT-PCR analysis showed that Mill1 is transcribed in selected tissues such as neonatal thymus and skin whereas Mill2 is transcribed more ubiquitously at low levels. Predicted MILL1 and MILL2 molecules are glycoproteins with three extracellular domains (α1 to α3), but their α1 and α2 domains lack many of the residues essential for the docking of peptides, suggesting that MILL molecules do not bind peptides. Phylogenetically, MILL1 and MILL2 are most closely related to MICA/B among known class I molecules. Because rodents lack the MICA/B family and conversely, humans do not have the MILL family, we suggested previously that MILL might be a functional substitute for MICA/B (19).

In the present study, we show that MILL1 and MILL2 are GPI-anchored glycoproteins associated with β2m. Consistent with the absence of critical residues required for the docking of peptides (19), cell surface expression of MILL1 and MILL2 did not require TAP molecules. Immunohistochemical analysis revealed that MILL1 is expressed in a subpopulation of thymic medullary epithelial cells and a restricted region of inner root sheaths in hair follicles. The ability to form complexes with β2m, anchorage to the membrane by GPI, and unique expression patterns all provide further evidence that MILL is a class I family distinct from MICA/B.

The mouse T lymphoma cell line RMA (H2b-positive) and its TAP2-deficient mutant RMA-S (H2b-negative) (21) were obtained from Dr. Kärre (Karolinska Institute, Stockholm, Sweden). Cells were maintained in RPMI 1640 medium (Invitrogen) supplemented with 10% (v/v) heat-inactivated FBS at 37°C and 5% CO2.

Anti-FLAG mAb M2 (F3165) was purchased from Sigma-Aldrich. Goat polyclonal Ab to mouse β2m (sc-8361) was purchased from Santa Cruz Biotechnology. Anti-human pan-cytokeratin mAb AE1/AE3 (M1590) and anti-human hair shaft cytokeratin mAb AE13 (ab16113) were purchased from DakoCytomation and Abcam, respectively. Mouse anti-H2-Kb mAb (clone AF6–88.5) and anti-CD45 mAb (clone 30-F11) were from BD Pharmingen. The Abs used as secondary reagents were as follows: FITC-labeled goat anti-mouse IgG, F(ab′)2 fragment (IM0819; Beckman Coulter), FITC-labeled swine anti-rabbit Ig, F(ab′)2 fragment (F0054; DakoCytomation), HRP-conjugated sheep anti-mouse IgG (NA931; Amersham Biosciences), HRP-conjugated donkey anti-rabbit IgG (NA934; Amersham Biosciences), HRP-conjugated donkey anti-goat IgG (sc-2056; Santa Cruz Biotechnology), Alexa Fluor 594-conjugated goat anti-rabbit IgG (A11072; Molecular Probes), and Alexa Fluor 488-conjugated goat anti-mouse IgG (A11001; Molecular Probes). Isotype-matched mouse IgG1 Ab (PP100) and pooled normal rabbit serum (CL1000) were purchased from Chemicon International Inc. and Cedarlane Laboratory Ltd., respectively.

The α1-α3 domains of MILL1 and MILL2 with 6 × His tags at their N termini were expressed in Escherichia coli strain M15 using the pQE30 expression vector following the instructions of the manufacturer (Qiagen). Briefly, the DNA fragments encoding the α1-α3 domains of mouse MILL molecules were amplified by PCR using the BALB/c-derived Mill plasmid cDNA (19) as templates. The primer sequences were 5′-TTGCGAGCTCCACACTCTGCGCTATGACCT-3′ (with a SacI site at its 5′-end) and 5′-CCCAAGCTTATATTGTGGTTGCCGTGCTT-3′ (with a HindIII site at its 5′-end) for MILL1 and 5′-GTGGATCCACCCACACTCTGCGCTATAA-3′ (with a BamHI site at its 5′-end) and 5′-CCCAAGCTTCATCCTGACTGTCCTCAGCA-3′ (with a HindIII site at its 5′-end) for MILL2. PCR products digested with SacI/HindIII for MILL1 and BamHI/HindIII for MILL2 were ligated into SacI/HindIII- and BamHI/HindIII-digested pQE30, respectively. After transformation into M15, recombinant proteins were induced by adding isopropyl-1-thio-β-D-galactopyranoside to a final concentration of 1 mM. E. coli cells were harvested and lysed in buffer B (100 mM NaH2PO4, 10 mM Tris-HCl, 6 M guanidine hydrochloride, pH 8.0), and lysates were centrifuged at 10,000 × g for 20 min at room temperature. Ni-NTA acid resins were added to supernatants and mixed gently by shaking. Resin-lysate mixtures were loaded into an empty column and washed twice with buffer C (100 mM NaH2PO4, 10 mM Tris-HCl, 6 M guanidine hydrochloride, pH 5.9). Recombinant proteins were eluted by buffer D (100 mM NaH2PO4, 10 mM Tris-HCl, 6 M guanidine hydrochloride, pH 4.5), separated by preparative SDS-PAGE, eluted and concentrated. Purified recombinant proteins (200 μg per rabbit) were mixed with CFA and injected into rabbits. After 2, 4, and 6 wk, the animals were boosted with the same amount of recombinant proteins mixed with IFA. Whole bloods were collected and antisera prepared 1 wk after the last boost.

Mouse MILL molecules have an insertion of amino acids between the leader peptide and the α1 domain (19). The coding regions of mouse MILL1 and MILL2 excluding this inserted sequence and the leader peptide were obtained by PCR using the Mill plasmid cDNA (19) as templates. The primer sequences were 5′-CCAAGCTTGAACCCCACACTCTGCGCTA-3′ (with a HindIII site at its 5′-end) and 5′-GTGGATCCCTACCAACACTGTAGAAAAGAGC-3′ (with a BamHI site at its 5′-end) for MILL1 and 5′-CCAAGCTTACCCACACTCTGCGCTATAA-3′ (with a HindIII site at its 5′-end) and 5′-GTGGATCCTCAGTTGGCTCTGGCCAGTG-3′ (with a BamHI site at its 5′-end) for MILL2. After digestion with HindIII/BamHI, the PCR products were ligated to the HindIII/BamHI-digested pFLAG-CMV-3 expression vector carrying a preprotrypsin leader sequence (Sigma-Aldrich). These constructs, designated MILL1-pFLAG-CMV-3 and MILL2-pFLAG-CMV-3, respectively, enabled the expression of MILL molecules with an N-terminal FLAG tag. In all cases, the integrity of expression constructs was verified by sequencing. DNA for transfection was isolated with the plasmid purification kit purchased from Qiagen.

To establish stable cell lines expressing MILL molecules, RMA and RMA-S cells were transfected with linearized MILL1-pFLAG-CMV-3 or MILL2-pFLAG-CMV-3 plasmids by electroporation at 250 V, 950 μF with Gene Pulser II according to the instructions of the manufacturer (Bio-Rad). Neomycin-resistant cells were selected by treatment with G418 (600 and 800 μg/ml for RMA and RMA-S, respectively) and clones exhibiting high levels of MILL expression were expanded; expression of MILL proteins was monitored by flow cytometry and immunoblotting with anti-FLAG and anti-MILL Abs.

For cell surface staining, single cell suspensions (1 × 106 cells) were washed with ice-cold PBS (pH 7.4) and incubated in 100 μl of PBS (pH 7.4) containing 0.1% NaN3 with 1 μg of mAb or isotype controls for 30 min on ice. After washing with ice-cold PBS (pH 7.4), cells were incubated in 100 μl of PBS (pH 7.4) containing 0.1% NaN3 with the FITC-conjugated F(ab′) fragment of goat anti-mouse IgG or F(ab′)2 fragment of swine anti-rabbit Ig (1:200 dilution). Subsequently, cells were washed with ice-cold PBS (pH 7.4) and analyzed by EPICS ALTRA (Beckman Coulter). Data were analyzed with EXPO32 software (Beckman Coulter).

For purification of FLAG-tagged MILL proteins, RMA-MILL1 and RMA-MILL2 stable transfectants (1 × 108 cells) were solubilized by 1 ml of ice-cold lysis buffer (50 mM Tris-HCl, 1 mM EDTA, 150 mM NaCl, 1% Triton X-100, 0.2 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, 20 μM leupeptin, 1 μM pepstatin, pH 7.5). After incubation for 30 min at 4°C, cell lysates were centrifuged at 13,000 × g for 10 min at 4°C to remove cell nuclei and insoluble proteins. Cleared lysates were incubated with protein G-Sepharose beads (Amersham Biosciences) at 4°C for 1 h. Supernatants were incubated with anti-FLAG mAb-coupled protein G-Sepharose beads at 4°C for 1 h. After washing 4 times with lysis buffer, immunoprecipitated proteins were eluted by 0.1 M glycine-HCl (pH 3.0), and immediately neutralized by adding 0.1 M Tris-HCl (pH 9.0). Eluted proteins were denatured and treated with 500 U/μl peptide:N-glycosidase F (PNGase F; New England Biolabs) at 37°C for 18 h.

To detect MILL proteins and β2m, samples were incubated in 1 × SDS sample buffer at 95°C for 10 min. Denatured proteins were separated on 12% SDS-PAGE and transferred to Hybond-P polyvinylidene difluoride membranes (Amersham Biosciences) using a semidry blotter (Bio-Rad) at 15 V for 45 min. The blotted membranes were incubated with 5% skim milk or 3% BSA in PBS (pH 7.4) containing 0.1% Tween 20 (PBST) at room temperature for 60 min and then incubated with 1/500 diluted antisera or 1 μg/ml of Ab in PBST at room temperature for 60 min. After washing twice with PBST, the membranes were incubated with 1/25,000 diluted HRP-conjugated anti-mouse, rabbit or goat IgG Abs. After washing three times with PBST, positive bands were visualized using the ECL-Plus (Amersham Biosciences) or the Super Signal West Dura detection system (Pierce).

RMA-MILL and RMA-S-MILL cells were washed with PBS (pH 7.4) and treated with 1 U/ml PI-PLC (Sigma-Aldrich) in PBS (pH 7.4) at 37°C for 1 h. Subsequently, cells were washed with ice-cold PBS (pH 7.4) and used for flow cytometric analysis.

Cell surface MILL proteins expressed on the RMA-MILL stable transfectants were purified by PI-PLC treatment and immunoprecipitation with anti-FLAG Ab-coupled protein G-Sepharose beads. Immunoprecipitates were subjected to immunoblotting using anti-FLAG and anti-mouse β2m Ab.

cDNA encoding the ectodomains of MILL1 and MILL2 were amplified by PCR using the Mill plasmid cDNA (19) as templates. Primers used were 5′-CATTAATGGACAACCAAAGACTGGTG-3′ (sense) and 5′-TCCCCCGGGGGCAGCAGGTTCATTGATA-3′ (antisense) for MILL1, and 5′-CCATATGTCCAGCATCCAAGGAACC-3′ (sense) and 5′-AAAAGTACTGACAGCTGTCTGCATGATG-3′ (antisense) for MILL2. These primers contained AseI, SmaI, NdeI, or ScaI restriction enzyme sites indicated by underlines. The PCR-generated cDNA fragments of MILL1 and MILL2 were cloned into the bacterial expression vector pET3cN-bio, which was designed to express a recombinant protein with an N-terminal enzymatic biotinylation signal (22), to construct MILL1-pET3cNbio and MILL2-pET3cNbio, respectively. Rosetta (DE3) strain of E. coli (Novagen, Merck) was transformed with MILL1-pET3cNbio or MILL2-pET3cNbio. Expression of soluble MILL1 or MILL2 was induced with 1 mM isopropyl-1-thio-β-D-galactopyranoside, and MILL proteins were refolded from the purified inclusion bodies by dilution as described previously (22). To examine effects of β2m on refolding, C57BL/6-derived β2m (β2mb), similarly expressed in E. coli, was included in the refolding mixture at the molar ratio of 1:2 (MILL/β2m). Refolded soluble MILL1 and MILL2 proteins were purified by anion-exchange column chromatography and gel-filtration chromatography. In anion-exchange chromatography on a UNO Q-6 column using 20 mM Tris-HCl buffer (pH 8.5) as a mobile phase, soluble MILL1 or MILL2 refolded in the presence of β2m was eluted in the approximately 250 mM Cl fraction by a 0–500 mM NaCl gradient. The gel-filtration column chromatography was performed on a Superdex 75 10/30 column (Amersham Biosciences) equilibrated with 25 mM Tris-HCl buffer (pH 8.0) containing 150 mM NaCl at the flow rate of 0.5 ml/min. The column was calibrated with gel-filtration standards from Bio-Rad.

For immunostaining, frozen sections prepared from 3-day-old, 10-day-old, and 6-wk-old male BALB/c mice were fixed using cold acetone for 5 min, washed with PBS, stained by the standard method (23), and then mounted in fluorescent mounting medium (DakoCytomation). Immunofluorescence was detected using a fluorescence microscope (ECLIPSE E600; Nikon). To evaluate the specificity of staining, the antiserum against MILL1 was diluted 1/40 with PBS to a final volume of 1 ml and absorbed with 5 × 107 RMA-MILL1 or RMA cells at 4°C overnight. The preabsorbed antiserum was diluted 1/80 with PBS and used for immunostaining. All experiments using animals have been reviewed and approved by the institutional review committee of Hokkaido University Graduate School of Medicine.

Thymi were dissected from 4-wk-old C57BL/6 and β2m-deficient mice. Breeding pairs of the β2m-deficient strain, B6.129P2-B2mtm1Unc/J (stock no. 002087), were purchased from The Jackson Laboratory, and their progenies were produced at Kinki University School of Medicine. Thymic stromal cells were enriched as described (24). Briefly, thymic fragments were digested in RPMI 1640 medium containing collagenase D and DNase I (both obtained from Roche) at 37°C for 15 min. After repeating this procedure 3 times, cells were pooled and stained with mAb for CD45. CD45-negative fractions containing stromal cells including thymic epithelial cells were subjected to flow cytometric analysis.

To facilitate biochemical analysis, we transfected FLAG-tagged expression plasmids into the mouse T lymphoma cell line RMA and established stable cell lines, RMA-MILL1 and RMA-MILL2, expressing N-terminally FLAG-tagged MILL1 and MILL2 molecules, respectively (Fig. 1,A). Cell surface expression of MILL1 and MILL2 was confirmed by flow cytometry using the anti-FLAG Ab as well as the rabbit antisera generated against bacterially expressed MILL1 and MILL2 molecules. The specificity of our rabbit antisera was further confirmed by Western blot analysis of whole cell lysates (Fig. 1,B). The anti-FLAG Ab detected two major bands of 48 and 41 kDa in RMA-MILL1 cells. The band of 41 kDa was nonspecific because it was detected in untransfected RMA cells. A major band of 48 kDa and a minor band of 44 kDa were detected by the anti-MILL1, but not anti-MILL2, antiserum (Fig. 1,B). In RMA-MILL2 lysates, the anti-FLAG Ab detected bands of 43 and 41 kDa (Fig. 1,B, top), which were also detected with the anti-MILL2, but not anti-MILL1, antiserum (Fig. 1,B, bottom). Thus, the band of 41 kDa detected by the anti-FLAG Ab in RMA-MILL2 cells presumably represents doublets containing both specific and nonspecific signals. We also expressed MILL1 and MILL2 molecules on RMA cells using their endogenous signal peptides and performed cytometric analysis using the MILL-specific rabbit antisera. We obtained staining patterns similar to those shown in Fig. 1 A (data not shown).

FIGURE 1.

MILL1 and MILL2 are cell surface glycoproteins with N-linked sugars. A, Untransfected RMA cells and the transfected cell lines, RMA-MILL1 and RMA-MILL2, which stably express MILL1 and MILL2, respectively, were incubated with anti-FLAG mAb and FITC-conjugated goat anti-mouse IgG, anti-MILL1 antiserum (1/500 dilution) and FITC-labeled swine anti-rabbit Ig, or anti-MILL2 antiserum (1/500 dilution) and FITC-labeled swine anti-rabbit Ig (from the top to the bottom, shaded histograms). Negative control staining (open histograms) was obtained using an isotype-matched control Ab (top three panels) or normal rabbit serum (all other panels). Stained cells were analyzed by flow cytometry. B, Whole cell lysates of RMA-MILL1 and RMA-MILL2 were separated on 12% SDS-PAGE and subjected to immunoblotting using anti-FLAG mAb (top), anti-MILL1 antiserum (middle), or anti-MILL2 antiserum (bottom). Signals were detected by HRP-conjugated secondary Ab and ECL-Plus reagents. Nonspecific bands are indicated by asterisks. C, MILL1 and MILL2 proteins were immunoprecipitated with anti-FLAG mAb-coupled protein G-Sepharose beads from RMA-MILL1 and RMA-MILL2 cell lysates, respectively. After digestion with PNGase F at 37°C for 18 h, samples were separated on 12% SDS-PAGE and subjected to immunoblotting. MILL1 was detected by the rabbit anti-MILL1 antiserum and MILL2 by the rabbit anti-MILL2 antiserum. Signals were detected by HRP-conjugated secondary Ab and ECL-Plus reagents.

FIGURE 1.

MILL1 and MILL2 are cell surface glycoproteins with N-linked sugars. A, Untransfected RMA cells and the transfected cell lines, RMA-MILL1 and RMA-MILL2, which stably express MILL1 and MILL2, respectively, were incubated with anti-FLAG mAb and FITC-conjugated goat anti-mouse IgG, anti-MILL1 antiserum (1/500 dilution) and FITC-labeled swine anti-rabbit Ig, or anti-MILL2 antiserum (1/500 dilution) and FITC-labeled swine anti-rabbit Ig (from the top to the bottom, shaded histograms). Negative control staining (open histograms) was obtained using an isotype-matched control Ab (top three panels) or normal rabbit serum (all other panels). Stained cells were analyzed by flow cytometry. B, Whole cell lysates of RMA-MILL1 and RMA-MILL2 were separated on 12% SDS-PAGE and subjected to immunoblotting using anti-FLAG mAb (top), anti-MILL1 antiserum (middle), or anti-MILL2 antiserum (bottom). Signals were detected by HRP-conjugated secondary Ab and ECL-Plus reagents. Nonspecific bands are indicated by asterisks. C, MILL1 and MILL2 proteins were immunoprecipitated with anti-FLAG mAb-coupled protein G-Sepharose beads from RMA-MILL1 and RMA-MILL2 cell lysates, respectively. After digestion with PNGase F at 37°C for 18 h, samples were separated on 12% SDS-PAGE and subjected to immunoblotting. MILL1 was detected by the rabbit anti-MILL1 antiserum and MILL2 by the rabbit anti-MILL2 antiserum. Signals were detected by HRP-conjugated secondary Ab and ECL-Plus reagents.

Close modal

Deduced MILL1 and MILL2 molecules have three potential N-linked glycosylation sites, respectively (19). To examine glycosylation status, we isolated MILL molecules from the stable transfectants by immunoprecipitation with the anti-FLAG Ab, removed N-linked glycans with PNGase F and performed immunoblot analysis with the MILL-specific antisera (Fig. 1,C). We obtained two bands of 44 and 48 kDa for non-treated MILL1, and a single band of 38 kDa for PNGase F-treated MILL1 (Fig. 1,C, top). Similarly, we obtained two bands of 41 and 43 kDa for non-treated MILL2, and a single band of 35 kDa for PNGase F-treated MILL2 (Fig. 1 C, bottom). The expression constructs used for stable transfection predicted Mr of 39280.83 and 35013.86 for the protein moieties of N-terminally flagged MILL1 and MILL2 molecules, respectively. Thus, the sizes of deglycosylated products agreed well with theoretical expectations. These results indicate that MILL1 and MILL2 are cell surface glycoproteins with N-linked sugars.

We initially assumed that MILL1 and MILL2 were transmembrane proteins (19, 20). However, different prediction algorithms yielded inconsistent results concerning the presence or absence of transmembrane regions. Subsequent sequence analysis using the software ‘big-PI Predictor’ (25) suggested that MILL1 and MILL2 were likely GPI-anchored proteins. To examine this possibility, RMA-MILL1 and RMA-MILL2 cells were treated with PI-PLC, stained with the anti-FLAG Ab and examined by flow cytometry. In both RMA-MILL1 and RMA-MILL2 cells, cell surface staining was reduced markedly by PI-PLC treatment (Fig. 2, right panel). By contrast, cell surface staining with the H2-Kb Ab was not affected by similar treatment (Fig. 2, left panel), consistent with the fact that H2-Kb is an integral membrane protein. These results indicate that MILL1 and MILL2 are GPI-anchored cell surface proteins.

FIGURE 2.

MILL1 and MILL2 are GPI-anchored proteins. RMA-MILL1 and RMA-MILL2 cells (top and bottom, respectively) were incubated with 1 U/ml PI-PLC (shaded histograms) or PBS (open histograms). Subsequently, cells were stained with anti-H2-Kb (left) or anti-FLAG (right) mAb. An FITC-conjugated F(ab′)2 fragment of goat anti-mouse IgG was used as a secondary Ab. Stained cells were analyzed by flow cytometry.

FIGURE 2.

MILL1 and MILL2 are GPI-anchored proteins. RMA-MILL1 and RMA-MILL2 cells (top and bottom, respectively) were incubated with 1 U/ml PI-PLC (shaded histograms) or PBS (open histograms). Subsequently, cells were stained with anti-H2-Kb (left) or anti-FLAG (right) mAb. An FITC-conjugated F(ab′)2 fragment of goat anti-mouse IgG was used as a secondary Ab. Stained cells were analyzed by flow cytometry.

Close modal

RMA-S is a variant derived from RMA cells (21) that lacks functional TAP molecules because of a defective TAP2 subunit (26, 27). At 37°C, classical class I molecules are barely expressed on the surface of RMA-S cells because empty class I molecules (class I molecules without peptides) are thermodynamically unstable. However, RMA-S cells express empty class I molecules when they are cultured at lower temperatures (28). To examine whether surface expression of MILL requires TAP, we transfected RMA-S cells with MILL expression plasmids and established stable transfectants. These cells were cultured at 25°C or 37°C and stained with the anti-FLAG or anti-H2-Kb Ab. As expected, endogenous H2-Kb molecules were expressed on RMA-S cells at the level comparable to that expressed on RMA cells when these cells were cultured at 25°C (Fig. 3, A and B, left panel, open histograms). However, expression of H2-Kb on RMA-S cells was reduced markedly when the cells were cultured at 37°C (Fig. 3, A and B, left panel, shaded histograms). By contrast, the expression levels of MILL1 and MILL2 detected by the anti-FLAG Ab were nearly the same regardless of whether the RMA-S cells were cultured at 25°C or 37°C (Fig. 3, A and B, right panel). These results indicate that cell surface expression of MILL molecules is TAP-independent.

FIGURE 3.

Cell surface expression of MILL molecules does not require functional TAP molecules. A, RMA-MILL1 and RMA-S-MILL1 cells were cultured at 25°C (open histograms) or 37°C (shaded histograms) for 18 h. Cells were incubated with anti-H2-Kb (left) or anti-FLAG (right) and then treated with FITC-conjugated F(ab′)2 fragments of goat anti-mouse IgG. Stained cells were analyzed by flow cytometry. B, RMA-MILL2 and RMA-S-MILL2 cells were treated in the same manner as in A, and cell surface expression of MILL2 was monitored by flow cytometry.

FIGURE 3.

Cell surface expression of MILL molecules does not require functional TAP molecules. A, RMA-MILL1 and RMA-S-MILL1 cells were cultured at 25°C (open histograms) or 37°C (shaded histograms) for 18 h. Cells were incubated with anti-H2-Kb (left) or anti-FLAG (right) and then treated with FITC-conjugated F(ab′)2 fragments of goat anti-mouse IgG. Stained cells were analyzed by flow cytometry. B, RMA-MILL2 and RMA-S-MILL2 cells were treated in the same manner as in A, and cell surface expression of MILL2 was monitored by flow cytometry.

Close modal

To examine whether MILL molecules are associated with β2m in vivo, we performed coimmunoprecipitation analysis. After treatment of RMA-MILL1 and RMA-MILL2 cells with PI-PLC, the MILL molecules released into the supernatants were immunoprecipitated with the anti-FLAG Ab and subjected to immunoblotting analysis using the anti-FLAG and anti-mouse β2m Ab (Fig. 4). We found that β2m was coimmunoprecipitated with both MILL1 and MILL2, indicating that MILL molecules are associated with β2m on the cell surface.

FIGURE 4.

Cell surface-expressed MILL1 and MILL2 molecules are associated with β2m. RMA-MILL1 and RMA-MILL2 cells were treated with PI-PLC and soluble MILL proteins were purified by immunoprecipitation with anti-FLAG mAb-coupled protein G-Sepharose beads (right). Precipitated samples were separated on 12% (top) or 14% (bottom) SDS-PAGE and subjected to immunoblotting analysis. MILL1 and MILL2 were detected by anti-FLAG mAb (top) whereas mouse β2m was detected by anti-mouse β2m Ab (bottom). Signals were detected by HRP-conjugated secondary Ab using the Super Signal West Dura kit. An asterisk indicates mouse IgG H chains.

FIGURE 4.

Cell surface-expressed MILL1 and MILL2 molecules are associated with β2m. RMA-MILL1 and RMA-MILL2 cells were treated with PI-PLC and soluble MILL proteins were purified by immunoprecipitation with anti-FLAG mAb-coupled protein G-Sepharose beads (right). Precipitated samples were separated on 12% (top) or 14% (bottom) SDS-PAGE and subjected to immunoblotting analysis. MILL1 and MILL2 were detected by anti-FLAG mAb (top) whereas mouse β2m was detected by anti-mouse β2m Ab (bottom). Signals were detected by HRP-conjugated secondary Ab using the Super Signal West Dura kit. An asterisk indicates mouse IgG H chains.

Close modal

To examine whether MILL1 and MILL2 can associate with β2m in vitro, we expressed the extracellular domains of MILL1 and MILL2 in E. coli and refolded them in the presence or absence of mouse β2m. MILL1 could be successfully refolded only in the presence of β2m (Fig. 5,A, top panel), and MILL1 and β2m were eluted in the same fractions in gel filtration chromatography as revealed by SDS-PAGE analysis (Fig. 5,B, bottom half, top panel). Although MILL2 was able to form soluble proteins when it was refolded in the absence of β2m, β2m appeared to improve the efficacy of refolding, consistent with our other results (Fig. 5,A, bottom panel). MILL2 refolded in the presence of β2m was eluted earlier in gel filtration chromatography than that refolded in the absence of β2m (Fig. 5,A, bottom panel), indicating that MILL2 molecules refolded in the presence of β2m were associated with β2m, which was further confirmed by SDS-PAGE analysis (Fig. 5,B, bottom panel). MILL1 and MILL2 refolded in the presence of β2m were purified by anion-exchange chromatography followed by gel filtration chromatography. The purified MILL1 and MILL2 proteins contained β2m as a subunit (Fig. 5,C). These results indicate that efficient refolding of MILL1 and MILL2 requires β2m as a subunit. The molecular masses of MILL1/β2m and MILL2/β2m complexes estimated by gel filtration chromatography (Fig. 5,A) and the relative intensities (3:1) of the MILL1 and MILL2 bands to the β2m bands in the purified MILL1/β2m and MILL2/β2m complexes (Fig. 5 C) indicate that the MILL1 or MILL2 polypeptide and β2m bind at a 1:1 ratio.

FIGURE 5.

Refolding of MILL1 and MILL2 requires β2m. A, Bacterially expressed extracellular domains (α1–α3) of MILL1 and MILL2 were refolded in the presence (continuous line) or absence (broken line) of β2m and subjected to gel filtration chromatography on Superdex-75. Filled arrowheads indicate the peaks of MILL1 and MILL2 proteins associated with β2m. Arrows indicate the peaks of free β2m. An open arrowhead indicates the peak of MILL2 refolded in the absence of β2m. B, The fractions from gel filtration chromatography were analyzed on SDS-PAGE, and the gels were stained with silver staining. The top and bottom halves of each panel indicate fractionation of the samples refolded in the absence and presence of β2m, respectively. C, Coomassie brilliant blue-stained SDS-PAGE gel of in vitro refolded MILL1 and MILL2 molecules purified by sequential chromatography.

FIGURE 5.

Refolding of MILL1 and MILL2 requires β2m. A, Bacterially expressed extracellular domains (α1–α3) of MILL1 and MILL2 were refolded in the presence (continuous line) or absence (broken line) of β2m and subjected to gel filtration chromatography on Superdex-75. Filled arrowheads indicate the peaks of MILL1 and MILL2 proteins associated with β2m. Arrows indicate the peaks of free β2m. An open arrowhead indicates the peak of MILL2 refolded in the absence of β2m. B, The fractions from gel filtration chromatography were analyzed on SDS-PAGE, and the gels were stained with silver staining. The top and bottom halves of each panel indicate fractionation of the samples refolded in the absence and presence of β2m, respectively. C, Coomassie brilliant blue-stained SDS-PAGE gel of in vitro refolded MILL1 and MILL2 molecules purified by sequential chromatography.

Close modal

To determine the tissue distribution of MILL1 and MILL2 molecules, we first performed Western blot analysis using the antisera for MILL1 and MILL2 against a panel of tissues isolated from adult and neonatal mice. These experiments yielded no bands in any tissues, presumably because the expression levels of MILL1 and MILL2 are low (data not shown). Our previous RT-PCR analysis (19) indicated that Mill1 was transcribed in selected tissues including neonatal thymus and skin. We therefore examined expression of MILL1 in these tissues (Fig. 6). Staining was observed in a subpopulation of medullary epithelial cells in the neonatal thymus (Fig. 6,A). These MILL1-positive cells were also detectable in the thymus of adult mice (data not shown). In the skin of 3-day-old mice, cells stained with the anti-MILL1 antiserum were found in the hair follicle (Fig. 6,B, left). However, these cells became undetectable in the skin of 10-day-old (Fig. 6,B, right) or 6-wk-old (not shown) mice. To more precisely address the locations of cells stained with the anti-MILL1 antiserum, we performed immunohistochemical staining of hair shafts and outer root sheaths (Fig. 6 C). Cells stained with the anti-MILL1 antiserum were located outside the hair shaft (stained green with AE13 mAb), but inside the outer root sheath (stained green with AE1/AE3 mAb). Thus, positively stained cells are located in the inner root sheath. Because not all regions of inner root sheaths were stained with the antiserum, MILL1 seems to be expressed in a restricted region of the inner root sheath. To confirm the specificity of staining, we prepared anti-MILL1 antiserum preabsorbed with RMA-MILL1 or RMA cells. Preabsorption of the antiserum with RMA-MILL1 cells almost eliminated staining in thymic epithelial cells and hair follicles whereas staining was retained when the antiserum was preabsorbed with RMA cells (data not shown). Mill2 is transcribed almost ubiquitously at low levels (19). We stained several tissues including neonatal thymus and skin as well as adult aorta, uterus, heart, kidney and spleen with the antiserum for MILL2 (1/200 dilution). Although this antiserum, when used at this dilution, was capable of staining RMA-MILL2 cells grown in vivo in C57BL/6 mice, we were unable to obtain any positive staining for MILL2 in any of the tissues (data not shown).

FIGURE 6.

MILL1 is likely expressed in thymic medullary epithelial cells and hair follicles. A, Thymic tissue sections obtained from 3-day-old mice were blocked by incubation with normal goat serum (1/500 dilution), reacted with AE1/AE3 and anti-MILL1 (1/400 dilution) followed by staining with Alexa Fluor 488-conjugated goat anti-mouse IgG (1/300 dilution) and Alexa Fluor 594-conjugated goat anti-rabbit IgG (1/300 dilution). Aa, a low-power photo micrograph of the thymic medulla (original magnification ×100); AbAd, a high-power magnification (original magnification ×400). Images for MILL1 (stained red, Ab) and AE1/AE3 (stained green, Ac) as well as the merged image (Ad) were obtained with a Nikon ECLIPSE E600 microscope. B, Skin tissues of 3-day-old (Ba and Bc) and 10-day-old (Bb and Bd) mice. Upper panels, H & E (original magnification ×100). Lower panels, H & E (original magnification ×400). The inset in Bc shows staining with the anti-MILL1 antiserum (original magnification ×100). Staining was done as described in the legend to C. C, MILL1 is likely expressed in cells of the inner root sheaths. In Ca and Cb, tissue sections were blocked by incubation with normal goat serum (1/500 dilution), reacted with AE13 (1/1000 dilution) and anti-MILL1 (1/400 dilution) and stained with Alexa Fluor 488-conjugated goat anti-mouse IgG (1/300 dilution) and Alexa Fluor 594-conjugated goat anti-rabbit IgG (1:300 dilution). In Cc and Cd, AE13 was substituted by AE1/AE3. MILL1 is stained red. The hair cortex (Ca and Cb) and outer root sheaths (Cc and Cd) are stained green. In Ca and Cc, hair shafts were sectioned parallel to the long axis. Cb and Cd show cross sections of hair shafts. Arrows in Ca and Cb indicate the hair cortex, whereas those in Cc and Cd indicate outer root sheaths. Original magnification ×400.

FIGURE 6.

MILL1 is likely expressed in thymic medullary epithelial cells and hair follicles. A, Thymic tissue sections obtained from 3-day-old mice were blocked by incubation with normal goat serum (1/500 dilution), reacted with AE1/AE3 and anti-MILL1 (1/400 dilution) followed by staining with Alexa Fluor 488-conjugated goat anti-mouse IgG (1/300 dilution) and Alexa Fluor 594-conjugated goat anti-rabbit IgG (1/300 dilution). Aa, a low-power photo micrograph of the thymic medulla (original magnification ×100); AbAd, a high-power magnification (original magnification ×400). Images for MILL1 (stained red, Ab) and AE1/AE3 (stained green, Ac) as well as the merged image (Ad) were obtained with a Nikon ECLIPSE E600 microscope. B, Skin tissues of 3-day-old (Ba and Bc) and 10-day-old (Bb and Bd) mice. Upper panels, H & E (original magnification ×100). Lower panels, H & E (original magnification ×400). The inset in Bc shows staining with the anti-MILL1 antiserum (original magnification ×100). Staining was done as described in the legend to C. C, MILL1 is likely expressed in cells of the inner root sheaths. In Ca and Cb, tissue sections were blocked by incubation with normal goat serum (1/500 dilution), reacted with AE13 (1/1000 dilution) and anti-MILL1 (1/400 dilution) and stained with Alexa Fluor 488-conjugated goat anti-mouse IgG (1/300 dilution) and Alexa Fluor 594-conjugated goat anti-rabbit IgG (1:300 dilution). In Cc and Cd, AE13 was substituted by AE1/AE3. MILL1 is stained red. The hair cortex (Ca and Cb) and outer root sheaths (Cc and Cd) are stained green. In Ca and Cc, hair shafts were sectioned parallel to the long axis. Cb and Cd show cross sections of hair shafts. Arrows in Ca and Cb indicate the hair cortex, whereas those in Cc and Cd indicate outer root sheaths. Original magnification ×400.

Close modal

To examine whether cell surface expression of MILL1 requires β2m, we isolated thymic stromal cells from 4-wk-old β2m-deficient mice and stained with the anti-MILL1 antiserum (Fig. 7). Cell surface expression of MILL1 was almost completely abrogated in the β2m-deficient mice compared with the adult C57BL/6 mice, indicating that cell surface expression of MILL1 is β2m-dependent.

FIGURE 7.

Cell surface expression of MILL1 requires β2m. Thymic stromal cells isolated from C57BL/6 (left panel) and β2m-deficient (right panel) mice were stained with normal rabbit serum (open histograms) or the rabbit anti-MILL1 antiserum (shaded histograms). An FITC-labeled swine anti-rabbit Ig was used as a secondary Ab. Stained cells were analyzed by flow cytometry.

FIGURE 7.

Cell surface expression of MILL1 requires β2m. Thymic stromal cells isolated from C57BL/6 (left panel) and β2m-deficient (right panel) mice were stained with normal rabbit serum (open histograms) or the rabbit anti-MILL1 antiserum (shaded histograms). An FITC-labeled swine anti-rabbit Ig was used as a secondary Ab. Stained cells were analyzed by flow cytometry.

Close modal

MILL is the latest addition to the growing list of mammalian MHC class I families encoded outside the MHC region. Our previous work has revealed several unique features of this class I family (19, 20). First, not all mammalian species have the MILL family; although mice and rats have this family, it is absent in humans. Because MILL apparently arose before the radiation of mammals, humans seem to have lost this class I family. Second, unlike all other class I genes, the genes coding for mouse MILL have an exon between those coding for the signal peptide and the α1 domain. Third, the MILL family is phylogenetically most closely related to the MICA/B family among known class I families. Because the MILL family is absent in humans, and conversely, mice and rats lack the MICA/B family, we suggested that MILL might serve as a functional substitute of MICA/B in rodents (19). Fourth, deduced MILL molecules lack most of the residues required for the docking of peptide termini, suggesting that they are unlikely to bind peptides. Fifth, RT-PCR analysis indicated that the members of the MILL family are poorly transcribed in most adult tissues, suggesting a role other than conventional Ag presentation. Sixth, sequence comparison of rat and mouse MILL molecules revealed that Mill is one of the most rapidly evolving class I gene families, and that, in both Mill1 and Mill2, non-synonymous substitutions occur more frequently than synonymous substitutions in the α1 domain whereas the opposite is the case in the α2 and α3 domains, suggesting that the α1 domain may be under positive selection (20). Taking all of these points into consideration, we suggested that MILL may perform specialized immune functions required only in certain species or some redundant functions, part of which are executed by other molecules (20).

In the present study, we set out to perform a biochemical characterization of mouse MILL molecules. Consistent with the absence of key residues required for the docking of peptides, we found that cell surface expression of MILL1 and MILL2 does not require functional TAP molecules (Fig. 3). Furthermore, the extracellular domains of MILL1 and MILL2 expressed in E. coli could be efficiently refolded in the absence of peptides under standard class I refolding conditions when β2m was added into the mixture (Fig. 5). This is in contrast to the fact that refolding of recombinant class Ia molecules isolated from purified bacterial inclusion bodies requires the presence of a peptide ligand and is reminiscent of the behaviors of certain class Ib molecules, the refolding of which is ligand-independent (29, 30, 31). Taken together, it is likely that the MILL family of class I molecules performs functions other than the presentation of peptides.

Two observations made in this work were rather unexpected. First, we initially assumed that, like most other class I family members, MILL1 and MILL2 were integral membrane proteins with a transmembrane region (19, 20). Contrary to this assumption, MILL1 and MILL2 turned out to be GPI-anchored proteins (Fig. 2). The occurrence of GPI anchors is not unprecedented for class I molecules because most if not all members of RAE-1 and ULBP families as well as a large proportion of Qa-2 molecules are GPI-anchored (32, 33, 34, 35). Like other GPI-anchored proteins (36, 37), MILL may be primarily located in lipid rafts. Second, we assumed that MILL1 and MILL2 were unlikely to be associated with β2m because they lack many of the residues known to interact with β2m in classical class I molecules (19). Our present work demonstrates that both MILL1 and MILL2 are associated with β2m on the cell surface (Fig. 4). A similar unexpected association with β2m was previously observed for MR1; this class Ib molecule lacks many of the phylogenetically conserved motifs implicated in β2m association in class Ia molecules (38), yet biochemical studies have revealed that it associates with β2m (38, 39). We also found that β2m promoted refolding of bacterially produced MILL ectodomains in vitro (Fig. 5). Hence, β2m appears to constitute an integral component of MILL class I molecules. Consistent with this, cell surface expression of MILL1 on thymic stromal cells was almost completely abrogated in β2m-deficient mice, indicating that cell surface expression of MILL1 requires β2m (Fig. 7). Given the overall structural similarity of MILL1 and MILL2 (19), and their shared biochemical properties (Figs. 2–5), it seems reasonable to assume that MILL2 also requires β2m for cell surface expression. Because the refolding experiments showed that MILL2, but not MILL1, was able to form soluble proteins in the absence of β2m, albeit much less efficiently than in the presence of β2m (Fig. 5), β2m might not be an absolute requirement for cell surface expression of MILL2. Human CD1d molecules, normally associated with β2m, can be expressed on the surface of intestinal epithelial cells in a β2m-independent manner (40, 41), indicating that the requirement for β2m can differ depending on tissues. Therefore, it will be necessary to identify cells or tissues where MILL2 is physiologically expressed to determine whether cell surface expression of MILL2 requires β2m in vivo, and if it does, whether β2m is absolutely required.

Immunohistochemical analysis showed that MILL1 is expressed in a subpopulation of thymic medullary epithelial cells and a restricted region of inner root sheaths in hair follicles (Fig. 6). Expression in the thymus is suggestive of an immunological role for MILL1. Totally unexpected was the observation that some inner root sheath cells in 3-day-old, but not 10-day-old or 6-wk-old, mice were stained with the antiserum for MILL1, although we cannot rule out the possibility that our anti-MILL1 antiserum cross-reacts with epitopes on unrelated molecules in hair follicles. Hair follicles have been proposed to enjoy immune privilege (42, 43). Thus, MILL1 may somehow be involved in the establishment and maintenance of immune privilege in hair follicles. On the other hand, we have thus far been unable to identify cells expressing MILL2 proteins despite the fact that the Mill2 gene is ubiquitously transcribed at low levels. Thus, expression of MILL2 proteins might be translationally regulated or MILL2 proteins might be expressed at detectable levels only in highly specialized cells as recently demonstrated for certain class I molecules (16, 17). It is also possible that expression of the MILL family is enhanced by certain stimuli or under pathologic conditions. To fully understand the expression patterns of the MILL family, more detailed analysis is required.

In conclusion, this study highlights the biochemical differences between the MILL and MICA/B families of class I molecules. MILL1 and MILL2 are TAP-independent, β2m-associated glycoproteins attached to the cell surface by GPI anchors. In contrast, MICA and MICB are TAP-independent, transmembrane proteins that do not associate with β2m (44). These two families of class I molecules also differ in their expression patterns. MICA and MICB are stress-inducible class I molecules usually not expressed on the surface of normal cells (44). In contrast, expression of Mill1 or Mill2 mRNA is not inducible by heat shock (our unpublished observation), and the expression in hair follicles seems unique to the MILL family. Furthermore, our preliminary work indicates that NK cells are not stained with MILL tetramers. All of these observations argue against the possibility that MILL is a functional substitute of MICA/B in rodents. Generation of knockout mice may provide a clue for understanding the biologic function of the MILL family.

We gratefully acknowledge Dr. Taeko Nagata and Kaori Kuno for technical assistance.

The authors have no financial conflict of interest.

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-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, Uehara Memorial Foundation, the Naito Foundation, and the Takeda Science Foundation.

3

Abbreviations used in this paper: β2m, β2-microglobulin; MILL, MHC class I-like located near the leukocyte receptor complex; PI-PLC, phosphatidylinositol-specific phospholipase C; PNGase F, peptide:N-glycosidase F.

1
Bjorkman, P. J., P. Parham.
1990
. Structure, function, and diversity of class I major histocompatibility complex molecules.
Annu. Rev. Biochem.
59
:
253
-288.
2
Klein, J..
1986
.
Natural History of the Major Histocompatibility Complex
John Wiley & Sons, New York.
3
Klein, J., C. O’hUigin.
1994
. The conundrum of nonclassical major histocompatibility complex genes.
Proc. Natl. Acad. Sci. USA
91
:
6251
-6252.
4
Rodgers, J. R., R. G. Cook.
2005
. MHC class Ib molecules bridge innate and acquired immunity.
Nat. Rev. Immunol.
5
:
459
-471.
5
Radosavljevic, M., S. Bahram.
2003
. In vivo immunogenetics: from MIC to RAET1 loci.
Immunogenetics
55
:
1
-9.
6
Bahram, S., M. Bresnahan, D. E. Geraghty, T. Spies.
1994
. A second lineage of mammalian major histocompatibility complex class I genes.
Proc. Natl. Acad. Sci. USA
91
:
6259
-6263.
7
Araki, T., F. Gejyo, K. Takagaki, H. Haupt, H. G. Schwick, W. Buergi, T. Marti, J. Schaller, E. Rickli, R. Brossmer, et al
1988
. Complete amino acid sequence of human plasma Zn-α2-glycoprotein and its homology to histocompatibility antigens.
Proc. Natl. Acad. Sci. USA
85
:
679
-683.
8
Fukudome, K., C. T. Esmon.
1994
. Identification, cloning, and regulation of a novel endothelial cell protein C/activated protein C receptor.
J. Biol. Chem.
269
:
26486
-26491.
9
Zou, Z., M. Nomura, Y. Takihara, T. Yasunaga, K. Shimada.
1996
. Isolation and characterization of retinoic acid-inducible cDNA clones in F9 cells: a novel cDNA family encodes cell surface proteins sharing partial homology with MHC class I molecules.
J. Biochem.
119
:
319
-328.
10
Fischer Lindahl, K., E. Hermel, B. E. Loveland, C. R. Wang.
1991
. Maternally transmitted antigen of mice: a model transplantation antigen.
Annu. Rev. Immunol.
9
:
351
-372.
11
Treiner, E., L. Duban, S. Bahram, M. Radosavljevic, V. Wanner, F. Tilloy, P. Affaticati, S. Gilfillan, O. Lantz.
2003
. Selection of evolutionarily conserved mucosal-associated invariant T cells by MR1.
Nature
422
:
164
-169.
12
Vincent, M. S., J. E. Gumperz, M. B. Brenner.
2003
. Understanding the function of CD1-restricted T cells.
Nat. Immunol.
4
:
517
-523.
13
Bahram, S..
2000
. MIC genes: from genetics to biology.
Adv. Immunol.
76
:
1
-60.
14
Cerwenka, A., L. L. Lanier.
2001
. Natural killer cells, viruses and cancer.
Nat. Rev. Immunol.
1
:
41
-49.
15
Simister, N. E., K. E. Mostov.
1989
. An Fc receptor structurally related to MHC class I antigens.
Nature
337
:
184
-187.
16
Ishii, T., J. Hirota, P. Mombaerts.
2003
. Combinatorial coexpression of neural and immune multigene families in mouse vomeronasal sensory neurons.
Curr. Biol.
13
:
394
-400.
17
Loconto, J., F. Papes, E. Chang, L. Stowers, E. P. Jones, T. Takada, A. Kumanovics, K. Fischer Lindahl, C. Dulac.
2003
. Functional expression of murine V2R pheromone receptors involves selective association with the M10 and M1 families of MHC class Ib molecules.
Cell
112
:
607
-618.
18
Todorov, P. T., T. M. McDevitt, D. J. Meyer, H. Ueyama, I. Ohkubo, M. J. Tisdale.
1998
. Purification and characterization of a tumor lipid-mobilizing factor.
Cancer Res.
58
:
2353
-2358.
19
Kasahara, M., Y. Watanabe, M. Sumasu, T. Nagata.
2002
. A family of MHC class I-like genes located in the vicinity of the mouse leukocyte receptor complex.
Proc. Natl. Acad. Sci. USA
99
:
13687
-13692.
20
Watanabe, Y., T. Maruoka, L. Walter, M. Kasahara.
2004
. Comparative genomics of the Mill family: a rapidly evolving MHC class I gene family.
Eur. J. Immunol.
34
:
1597
-1607.
21
Karre, K., H. G. Ljunggren, G. Piontek, R. Kiessling.
1986
. Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy.
Nature
319
:
675
-678.
22
Wada, H., N. Matsumoto, K. Maenaka, K. Suzuki, K. Yamamoto.
2004
. The inhibitory NK cell receptor CD94/NKG2A and the activating receptor CD94/NKG2C bind the top of HLA-E through mostly shared but partly distinct sets of HLA-E residues.
Eur. J. Immunol.
34
:
81
-90.
23
Baba, T., A. Ishizu, H. Ikeda, Y. Miyatake, T. Tsuji, A. Suzuki, U. Tomaru, T. Yoshiki.
2005
. Chronic graft-versus-host disease-like autoimmune disorders spontaneously occurred in rats with neonatal thymus atrophy.
Eur. J. Immunol.
35
:
1731
-1740.
24
Gray, D. H., A. P. Chidgey, R. L. Boyd.
2002
. Analysis of thymic stromal cell populations using flow cytometry.
J. Immunol. Methods
260
:
15
-28.
25
Eisenhaber, B., P. Bork, F. Eisenhaber.
1999
. Prediction of potential GPI-modification sites in proprotein sequences.
J. Mol. Biol.
292
:
741
-758.
26
Attaya, M., S. Jameson, C. K. Martinez, E. Hermel, C. Aldrich, J. Forman, K. F. Lindahl, M. J. Bevan, J. J. Monaco.
1992
. Ham-2 corrects the class I antigen-processing defect in RMA-S cells.
Nature
355
:
647
-649.
27
Yang, Y., K. Fruh, J. Chambers, J. B. Waters, L. Wu, T. Spies, P. A. Peterson.
1992
. Major histocompatibility complex (MHC)-encoded HAM2 is necessary for antigenic peptide loading onto class I MHC molecules.
J. Biol. Chem.
267
:
11669
-11672.
28
Ljunggren, H. G., N. J. Stam, C. Ohlen, J. J. Neefjes, P. Hoglund, M. T. Heemels, J. Bastin, T. N. Schumacher, A. Townsend, K. Karre, et al
1990
. Empty MHC class I molecules come out in the cold.
Nature
346
:
476
-480.
29
Li, P., G. McDermott, R. K. Strong.
2002
. Crystal structures of RAE-1β and its complex with the activating immunoreceptor NKG2D.
Immunity
16
:
77
-86.
30
Steinle, A., P. Li, D. L. Morris, V. Groh, L. L. Lanier, R. K. Strong, T. Spies.
2001
. Interactions of human NKG2D with its ligands MICA, MICB, and homologs of the mouse RAE-1 protein family.
Immunogenetics
53
:
279
-287.
31
Wingren, C., M. P. Crowley, M. Degano, Y. Chien, I. A. Wilson.
2000
. Crystal structure of a γδ T cell receptor ligand T22: a truncated MHC-like fold.
Science
287
:
310
-314.
32
Nomura, M., Z. Zou, T. Joh, Y. Takihara, Y. Matsuda, K. Shimada.
1996
. Genomic structures and characterization of Rae1 family members encoding GPI-anchored cell surface proteins and expressed predominantly in embryonic mouse brain.
J. Biochem.
120
:
987
-995.
33
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.
34
Cerwenka, A., A. B. Bakker, T. McClanahan, J. Wagner, J. Wu, J. H. Phillips, L. L. Lanier.
2000
. Retinoic acid early inducible genes define a ligand family for the activating NKG2D receptor in mice.
Immunity
12
:
721
-727.
35
Cosman, D., J. Mullberg, W. Chin, R. Armitage, W. Fanslow, M. Kubin, N. J. Chalupny.
2001
. ULBPs, novel MHC class I-related molecules, bind to CMV glycoprotein UL16 and stimulate NK cytotoxicity through the NKG2D receptor.
Immunity
14
:
123
-133.
36
Horejsi, V., K. Drbal, M. Cebecauer, J. Cerny, T. Brdicka, P. Angelisova, H. Stockinger.
1999
. GPI-microdomains: a role in signalling via immunoreceptors.
Immunol. Today
20
:
356
-361.
37
Pizzo, P., A. Viola.
2004
. Lipid rafts in lymphocyte activation.
Microbes Infect.
6
:
686
-692.
38
Miley, M. J., S. M. Truscott, Y. Y. Yu, S. Gilfillan, D. H. Fremont, T. H. Hansen, L. Lybarger.
2003
. Biochemical features of the MHC-related protein 1 consistent with an immunological function.
J. Immunol.
170
:
6090
-6098.
39
Yamaguchi, H., K. Hashimoto.
2002
. Association of MR1 protein, an MHC class I-related molecule, with β2-microglobulin.
Biochem. Biophys. Res. Commun.
290
:
722
-729.
40
Balk, S. P., S. Burke, J. E. Polischuk, M. E. Frantz, L. Yang, S. Porcelli, S. P. Colgan, R. S. Blumberg.
1994
. β2-microglobulin-independent MHC class Ib molecule expressed by human intestinal epithelium.
Science
265
:
259
-262.
41
Somnay-Wadgaonkar, K., A. Nusrat, H. S. Kim, W. P. Canchis, S. P. Balk, S. P. Colgan, R. S. Blumberg.
1999
. Immunolocalization of CD1d in human intestinal epithelial cells and identification of a β2-microglobulin-associated form.
Int. Immunol.
11
:
383
-392.
42
Niederkorn, J. Y..
2003
. Mechanisms of immune privilege in the eye and hair follicle.
J. Investig. Dermatol. Symp. Proc.
8
:
168
-172.
43
Paus, R., B. J. Nickoloff, T. Ito.
2005
. A ‘hairy’ privilege.
Trends Immunol.
26
:
32
-40.
44
Groh, V., S. Bahram, S. Bauer, A. Herman, M. Beauchamp, T. Spies.
1996
. Cell stress-regulated human major histocompatibility complex class I gene expressed in gastrointestinal epithelium.
Proc. Natl. Acad. Sci. USA
93
:
12445
-12450.