Crossref
journal-article
American Association for the Advancement of Science (AAAS)
Science (221)
Abstract
Mouse CD1d1, a member of the CD1 family of evolutionarily conserved major histocompatibility antigen–like molecules, controls the differentiation and function of a T lymphocyte subset, NK1 + natural T cells, proposed to regulate immune responses. The CD1d1 crystal structure revealed a large hydrophobic binding site occupied by a ligand of unknown chemical nature. Mass spectrometry and metabolic radiolabeling were used to identify cellular glycosylphosphatidylinositol as a major natural ligand of CD1d1. CD1d1 bound glycosylphosphatidylinositol through its phosphatidylinositol aspect with high affinity. Glycosylphosphatidylinositol or another glycolipid could be a candidate natural ligand for CD1d1-restricted T cells.
References
35
Referenced
330
-
L. H. Martin F. Calabi C. Milstein Proc. Natl. Acad. Sci. U.S.A. 83 9154 (1986).
(
10.1073/pnas.83.23.9154) 10.1126/science.277.5324.339- Bendelac A., et al., ibid. 268, 863 (1995). / ibid. by Bendelac A. (1995)
-
Exley M., Garcia G., Balk S. P., Porcelli S., J. Exp. Med. 186, 109 (1997).
(
10.1084/jem.186.1.109) / J. Exp. Med. by Exley M. (1997) -
Sykes M., J. Immunol. 145, 3209 (1990).
(
10.4049/jimmunol.145.10.3209) / J. Immunol. by Sykes M. (1990) 10.1126/science.270.5243.1845-
Baxter A. G., Kinder S. J., Hammond K. J. L., Scollay R., Godfrey D. I., Diabetes 46, 572 (1997).
(
10.2337/diab.46.4.572) / Diabetes by Baxter A. G. (1997) 10.1126/science.275.5302.977-
Chen Y.-H., Chiu N. M., Mandal M., Wang N., Wang C.-R., Immunity 6, 459 (1997).
(
10.1016/S1074-7613(00)80289-7) / Immunity by Chen Y.-H. (1997) - S. K. Mendiratta et al. ibid. p. 469.
-
Arase H., Arase N., Agasawara K., Good R. A., Onoe K., Proc. Natl. Acad. Sci. U.S.A. 89, 6506 (1992).
(
10.1073/pnas.89.14.6506) / Proc. Natl. Acad. Sci. U.S.A. by Arase H. (1992) - Brooks E. G., et al., ibid. 90, 11787 (1993). / ibid. by Brooks E. G. (1993)
-
Lantz O., Bendelac A., J. Exp. Med. 180, 1097 (1994).
(
10.1084/jem.180.3.1097) / J. Exp. Med. by Lantz O. (1994) - Brutkiewicz R. R., Bennink J. R., Yewdell J. W., Bendelac A., ibid. 182, 1913 (1995). / ibid. by Brutkiewicz R. R. (1995)
-
S. Joyce P. Tabaczewski R. H. Angeletti S. G. Nathenson I. Stroynowski ibid. 179 577 (1994).
(
10.1084/jem.179.2.579) - Lacking a suitable CD1-specific antibody or antiserum we exploited the observation that the NKT cell-activation function of CD1d1 is TAP independent and β 2 -microglobulin (β 2 m) dependent (14). CD1d1 was affinity purified with a mAb to human β 2 m from TAP-deficient human T2 cells infected with a recombinant vaccinia virus (VV) expressing the CD1d1 cDNA (3 14). The T2 cells are not completely devoid of endogenous class I molecules bearing low molecular weight ligands (32); these were first removed from detergent extracts by using a mixture of mAbs to HLA-A -B and -C. About 10 10 T2 cells infected with VV-CD1d1 at a multiplicity of infection of ∼3 for 4 to 5 hours were detergent solubilized and HLA class I was isolated by using protein A–Sepharose (Repligen) coupled with a mixture of specific mAbs (HB95 HB116 HB118 and HB120; a kind gift from T. Mohanakumar) followed by the purification of CD1d1 by using a specific mAb to human β 2 m (HB28; ATCC). HLA class I and CD1d1–associated ligands were isolated separated and fractionated by RP-HPLC as described (15). Briefly sample was injected in buffer A [0.06% trifluoroacetic acid (TFA) in water] onto a reversed-phase C 18 column (1.0 × 250 mm; Alltech) and eluted immediately by using a buffer B (0.05% TFA containing acetonitrile) gradient starting at 0% at 10 min to 37 70 90 and 100% at 73 105 115 and 120 min respectively at 50 μl/min.
- S. Joyce et al. unpublished data.
- Second-dimension RP-HPLC was performed with a 1:1 mixture of C 18 and cation-exchange column (1.0 × 250 mm; Alltech) by achieving 15 60 and 100% buffer B at 15 105 and 125 min respectively at 50 μl/min.
- Mass spectra were acquired on a Kratos analytical MALDI-4 mass spectrometer equipped with a curved-field reflectron and a nitrogen laser. About 0.3 to 1.0 μl of each fraction (30 to 40 μl) either directly or after concentrating to ∼15 μl was applied onto the sample probe. Matrix saturated α-cyano-4-hydroxycinnamic acid in 45% ethanol containing 8.8% formic acid (∼300 nl) was then applied to the sample. A replicate of the sample was similarly spotted except that to enhance ion signal 300 nl of saturated ammonium sulfate was added before matrix (20). Spectra were acquired in positive and negative modes as well as in linear and reflectron modes after application of a 20-kV accelerating voltage. Because different fragments form stable positive and negative ions when derived from the same molecule positive and negative spectra provide complementary structural information. The mass analyses reported here are within a mass accuracy of 0.5 dalton.
-
Woods A. S., et al., Anal. Biochem. 226, 15 (1995).
(
10.1006/abio.1995.1185) / Anal. Biochem. by Woods A. S. (1995) -
Heller N., Cotter R. J., Fenselau C., Uy O. M., Anal. Chem. 59, 2806 (1987).
(
10.1021/ac00150a018) / Anal. Chem. by Heller N. (1987) -
Heller N., Murphy C., Cotter R. J., Fenselau C., Uy O. M., ibid. 60, 2787 (1988).
(
10.1103/PhysRevLett.60.2235) / ibid. by Heller N. (1988) - Spent tissue culture supernatant containing sCD1d1 (∼2 liters; to be described elsewhere) and Db-sol (∼10 liters; to be described elsewhere) were concentrated by tangential flow filtration (Pall Filtron). CD1d1 was Ni-NTA–affinity purified (Qiagen's method) and Db-sol was immunoaffinity purified with B22-249 a H-2D b –specific mAb-coupled protein A–Sepharose (Repligen). Low molecular weight ligands were isolated from ∼3.0 mg of sCD1d1 and Db-sol and fractionated by C 18 RP-HPLC. Sample elution was initiated after the injection front returned to zero by increasing buffer B concentration to 37 70 90 and 100% at 63 95 105 and 110 min respectively at 50 μl/min.
-
Homans A. W., et al., Nature 333, 269 (1988).
(
10.1038/333269a0) / Nature by Homans A. W. (1988) - About 3 to 5 × 10 7 sCD1d1 and Db-sol cells were labeled with 62.5 μCi of [5 6 8 9 11 12 14 15- 3 H]arachidonic acid 250 μCi of [1- 3 H]ethanolamine-HCl or 250 μCi of d -[2- 3 H]mannose (American Radiolabeled Chemicals). [ 3 H]Mannose labeling was done in glucose-free and l -glutamine–free medium supplemented with 5% dialyzed fetal bovine serum dextrose (0.5 mg/ml) sodium pyruvate (1.1 mg/ml) antibiotics nucleosides and nonessential amino acids for 18 to 24 hours at 37 o C. CD1d1 and Db-sol in the culture supernatant were purified with His-Trap Ni–Sepharose and B22-249–coupled protein A–Sepharose columns respectively after preclearing with Hi-Trap protein A–Sepharose according to the manufacturer (Pharmacia Biotech). Radioactivity in each fraction was monitored with a scintillation counter (Beckman). [ 3 H]Mannose-labeled sCD1d1 and Db-sol–associated ligands were separated from the heavy and light chains by Microcon-10 (Amicon) filtration to specifically monitor GPI-associated radioactivity.
- sCD1d1 and H-2D b (about 4 μM) in triplicate were mixed with 1.8 μCi of l -α-[ myo -inositol-2- 3 H]PI (11 Ci/mmol; DuPont-NEN) in 100 μl of phosphate-buffered saline at 37°C for about 18 hours. sCD1d1- and H-2D b –bound [ 3 H]PI were separated from free [ 3 H]PI by Microcon-10 filtration and radioactivity in the retained solution was measured in a scintillation counter. H-2D b reconstituted in vitro from heavy and light chains produced in Escherichia coli and with a H-2D b –binding peptide Gly-Ala-Ile-Ser-Asn-Met-Tyr-Ala-Met derived from glutamic acid dehydrogenase was used as the control for binding specificity. The in vitro reconstituted H-2D b was generously provided by E. Palmieri and S. G. Nathenson. For Scatchard analysis various concentrations of purified sCD1d1 in duplicate were mixed with 1.8 μCi of [ 3 H]PI (∼1.6 μM) in 100 μl of 20 mM phosphate buffer pH 7.4. After incubation at 37 o C for ∼18 hours sCD1d1-[ 3 H]PI complexes were separated from free [ 3 H]PI by Microcon-10 filtration and radioactivity in the retained solution was measured.
- Clearly GPI is the major ligand identified under the conditions described here for the isolation RP-HPLC fractionation and MALDI-MS analysis (Fig. 1). To determine whether GPI is the major or the only natural ligand of CD1d1 we estimated the percent of CD1d1 occupied by GPI from the ratio of [ 3 H]mannose-labeled heavy chain to [ 3 H]mannose-labeled GPI. Considering that d -[2- 3 H]mannose converts mostly to d -[2- 3 H]fucose and rarely to other sugars (33) there are about seven times as many mannoses and fucoses in the heavy chain as in GPI (17). Thus ∼65% of CD1d1 is occupied by GPI before accounting for losses incurred during the purification steps. Assuming 65 to 70% recovery of the ligand [based on peptide recoveries from class I molecules (15 34)] then >90% of CD1d1 is occupied by GPI.
10.1126/science.7542403-
Udenfriend S., Kodukula K., Annu. Rev. Biochem. 64, 563 (1995).
(
10.1146/annurev.bi.64.070195.003023) / Annu. Rev. Biochem. by Udenfriend S. (1995) 10.1126/science.273.5273.349-
Wolfe P. R., Ploegh H. L., Annu. Rev. Cell Biol. 11, 267 (1995).
(
10.1146/annurev.cb.11.110195.001411) / Annu. Rev. Cell Biol. by Wolfe P. R. (1995) -
Henderson R. A., et al., Science 255, 1264 (1992).
(
10.1126/science.1546329) / Science by Henderson R. A. (1992) -
Varki A., Methods Enzymol. 230, 16 (1994).
(
10.1016/0076-6879(94)30004-6) / Methods Enzymol. by Varki A. (1994) -
Jardetzky T. S., Lane W. S., Robinson R. A., Madden D. R., Wiley D. C., Nature 353, 326 (1991).
(
10.1038/353326a0) / Nature by Jardetzky T. S. (1991) - We thank M. H. Hunsinger and J. M. Weaver for technical assistance; T. Mohanakumar E. Palmieri and S. G. Nathenson for generously providing affinity-purified anti-HLA class I mAb and reconstituted H-2D b respectively; and A. K. Menon for helpful discussions. Supported by grants from NIH (S.J. S.P.B. and R.J.C.) the Juvenile Diabetes Foundation International (S.J.) the American Cancer Society (S.J.) and the National Research Council–NIH (R.R.B.).
Dates
| Type | When |
|---|---|
| Created | 23 years, 1 month ago (July 27, 2002, 5:35 a.m.) |
| Deposited | 1 year, 7 months ago (Jan. 12, 2024, 9:55 p.m.) |
| Indexed | 4 weeks ago (July 30, 2025, 11:18 a.m.) |
| Issued | 27 years, 5 months ago (March 6, 1998) |
| Published | 27 years, 5 months ago (March 6, 1998) |
| Published Print | 27 years, 5 months ago (March 6, 1998) |
@article{Joyce_1998, title={Natural Ligand of Mouse CD1d1: Cellular Glycosylphosphatidylinositol}, volume={279}, ISSN={1095-9203}, url={http://dx.doi.org/10.1126/science.279.5356.1541}, DOI={10.1126/science.279.5356.1541}, number={5356}, journal={Science}, publisher={American Association for the Advancement of Science (AAAS)}, author={Joyce, Sebastian and Woods, Amina S. and Yewdell, Jonathan W. and Bennink, Jack R. and De Silva, A. Dharshan and Boesteanu, Alina and Balk, Steven P. and Cotter, Robert J. and Brutkiewicz, Randy R.}, year={1998}, month=mar, pages={1541–1544} }