Rapid screening for the detection of HLA-B57 and HLA-B58 in prevention of drug hypersensitivity.

HLA-B57 and HLA-B58 are major histocompatibility class (MHC)-I allotypes that are potentially predictive of important clinical immune phenotypes. HLA-B*5701 is strongly associated with hypersensitivity to the HIV drug abacavir, liver toxicity from the antibiotic flucloxacillin and is a marker for slow progression of HIV AIDS. HLA-B*5801 is associated with hypersensitivity to allopurinol used to treat hyperuricaemia and recurrent gout. Here we describe a monoclonal antibody (mAb) specific for HLA-B57 and HLA-B58 that provides an inexpensive and sensitive screen for these MHC-I allotypes. The usefulness of HLA-B57 screening for prediction of abacavir hypersensitivity was shown in three independent laboratories, including confirmation of the mAb sensitivity and specificity in a cohort of patients enrolled in the PREDICT-1 trial. Our data show that patients who test negative by mAb screening comprise 90%-95% of all individuals in most human populations and require no further human leukocyte antigen (HLA) typing. Patients who test positive by mAb screening should proceed to high-resolution typing to ascertain the presence of HLA-B*5701 or HLA-B*5801. Hence, mAb screening provides a low-cost alternative to high-resolution typing of all patients and lends itself to point-of-care diagnostics and rapid ascertainment of low-risk patients who can begin immediate therapy with abacavir, flucloxacillin or allopurinol.

Novel strategy for identification of candidate cytotoxic T-cell epitopes from human preproinsulin.

We describe a strategy for identifying ligands of human leukocyte antigen (HLA) class I molecules based on a peptide library-mediated in vitro assembly of recombinant class I molecules. We established a microscale class I assembly assay and used a capture ELISA to quantify the assembled HLA-peptide complexes. The identity of the bound ligands was then deduced by mass spectrometry. In this method, HLA complexes assembled in vitro in the presence of components of a mixture of peptides were immunoprecipitated and the bound peptide(s) identified by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry. This process of epitope extraction is robust and can be used with complex mixtures containing in excess of 300 candidate ligands. A library of overlapping peptides representing all potential octamers, nonamers and decamers from human preproinsulin was synthesized using unique library chemistry. Peptides from the library were used to initiate assembly of recombinant HLA-B8, HLA-B15 and HLA-A2, facilitating the identification of candidate T-cell epitopes from preproinsulin.

The GRIP domain is a specific targeting sequence for a population of trans-Golgi network derived tubulo-vesicular carriers.

Vesicular carriers for intracellular transport associate with unique sets of accessory molecules that dictate budding and docking on specific membrane domains. Although many of these accessory molecules are peripheral membrane proteins, in most cases the targeting sequences responsible for their membrane recruitment have yet to be identified. We have previously defined a novel Golgi targeting domain (GRIP) shared by a family of coiled-coil peripheral membrane Golgi proteins implicated in membrane trafficking. We show here that the docking site for the GRIP motif of p230 is a specific domain of Golgi membranes. By immuno-electron microscopy of HeLa cells stably expressing a green fluorescent protein (GFP)-p230GRIP fusion protein, we show binding specifically to a subset of membranes of the trans-Golgi network (TGN). Real-time imaging of live HeLa cells revealed that the GFP-p230GRIP was associated with highly dynamic tubular extensions of the TGN, which have the appearance and behaviour of transport carriers. To further define the nature of the GRIP membrane binding site, in vitro budding assays were performed using purified rat liver Golgi membranes and cytosol from GFP-p230GRIP-transfected cells. Analysis of Golgi-derived vesicles by sucrose gradient fractionation demonstrated that GFP-p230GRIP binds to a specific population of vesicles distinct from those labelled for beta-COP or gamma-adaptin. The GFP-p230GRIP fusion protein is recruited to the same vesicle population as full-length p230, demonstrating that the GRIP domain is solely proficient as a targeting signal for membrane binding of the native molecule. Therefore, p230 GRIP is a targeting signal for recruitment to a highly selective membrane attachment site on a specific population of trans-Golgi network tubulo-vesicular carriers.

The Golgi-targeting sequence of the peripheral membrane protein p230.

Vesicle transport requires the recruitment of cytosolic proteins to specific membrane compartments. We have previously characterised a brefeldin A-sensitive trans-Golgi network-localised protein (p230) that is associated with a population of non-clathrin-coated vesicles. p230 recycles between the cytosol and the cytoplasmic face of buds/vesicles of trans-Golgi network membranes in a G protein-regulated manner. Identifying the mechanism responsible for Golgi targeting of p230 is important for the elucidation of its function. By transfection of COS cells with deletion mutants of p230 we here demonstrate that the C-terminal domain is necessary for targeting to the Golgi. Furthermore, the C-terminal 98 amino acid domain of p230 attached to the green fluorescent protein (GFP-p230-C98aa) was efficiently Golgi-localised in transfected COS cells. Deletion mutants of GFP-p230-C98aa together with alanine scanning mutagenesis identified a minimum stretch of 42 amino acids that is essential for Golgi targeting, suggesting that the conformation of the domain is critical for efficient targeting. In COS cells expressing high levels of GFP-p230-C98aa fusion protein, endogenous p230 was no longer associated with Golgi membranes, suggesting that the GFP fusion protein and endogenous p230 may compete for the same membrane target structures. The Golgi binding of GFP-p230-C98aa is brefeldin A-sensitive and is regulated by G proteins. These studies have identified a minimal sequence responsible for specific targeting of p230 to the Golgi apparatus, which displays similar membrane binding characteristics to wild-type p230.

A novel Golgi-localisation domain shared by a class of coiled-coil peripheral membrane proteins.

The mechanism by which peripheral membrane proteins are targeted to the cytoplasmic face of the Golgi apparatus is poorly understood. Previously, we have identified a carboxy-terminal domain of the trans-Golgi-network (TGN) protein p230 that is responsible for Golgi localisation [1]. Here, we report the identification of a similar Golgi-localisation domain (GLD, also termed the 'GRIP' domain - see the paper by Munro and Nichols elsewhere in this issue) in a family of putative peripheral membrane proteins from lower and higher eucaryotes. The majority of family members have a domain structure similar to that of p230, with extensive coiled-coil regions (>80%) and the potential GLD located in a non-coiled-coil domain at the carboxyl terminus. Previously reported proteins in this family include human golgin-97 and Saccharomyces cerevisiae Imh1p. By constructing chimeric cDNAs encoding carboxy-terminal regions of these family members fused to green fluorescent protein (GFP), we have directly demonstrated that the GLD of p230, golgin-97, the newly identified human protein GCC1p and yeast Imh1p functions as a Golgi-targeting domain in transfected mammalian cells. Site-directed mutagenesis of the GLDs identified two conserved aromatic residues that are critical for the function of this targeting domain. Endogenous p230 was displaced from the Golgi membranes in transfected cells expressing high levels of GFP fused to the GLD of either p230 or golgin-97, indicating that different GLDs interact with similar membrane determinants. Thus, we have identified a family of coiled-coil proteins that share a domain shown to be sufficient for the localisation of peripheral membrane proteins to the Golgi apparatus.

Avidity for antigen can influence the helper dependence of CD8+ T lymphocytes.

To examine the influence of avidity on the helper dependence of CD8+ T cells, purified CD8+ T cells from transgenic mice expressing a TCR specific for H-2Kb were tested for their responsiveness to spleen cells expressing different densities of H-2Kb. High avidity interactions resulted in the induction of strong proliferative responses by purified CD8+ T cells. These cells synthesized their own IL-2 and IL-2R. By contrast, low avidity interactions failed to induce proliferation or detectable IL-2 synthesis, although IL-2R were induced and addition of rIL-2 induced a proliferative response. These results are consistent with a model in which TCR avidity for Ag can determine the helper dependence of a CD8+ T cell.

Impaired expression of chimaeric major histocompatibility complex transgenes associated with plasmid sequences.

Plasmid vector sequences were retained (vector+), or removed (vector-) from hybrid major histocompatibility complex gene constructs prior to microinjection of fertilized ova for the production of transgenic mice. In transgenic mice containing integrated vector+ gene constructs, low levels of class II cell surface determinants were detected on splenocytes from only two out of six independent lines. Class II membrane determinants were not detectable on splenocytes from the remaining four vector+ transgenic lines. Expression of transgene products did not correlate with transgene copy number which ranged from 1-10 copies. Low levels of mRNA transcripts were detected in thymic mRNA from vector+ lines. In contrast, high levels of thymic and splenic mRNA transcripts were detected in offspring from all four vector- transgenic lines. Spleen cells from the vector- transgenic animals also expressed high levels of the hybrid major histocompatibility complex transgene products. These results implicate plasmid vector sequences in the inhibition of expression of the hybrid class II-class I major histocompatibility complex genes in transgenic mice. This putative inhibition of transgene expression presumably occurs at the level of gene transcription.

Class II-restricted presentation of an endogenously derived immunodominant T-cell determinant of hen egg lysozyme.

An in vitro model was used to investigate the potential for different structural forms of endogenous antigen to be processed and presented by major histocompatibility complex class II molecules. For this purpose the class II-restricted presentation of an immunodominant epitope of hen egg lysozyme [HEL-(46-61)] was studied in class II-positive B-lymphoma cells (M12.C3) transfected with genes encoding HEL molecules either (i) secreted in high (hi) or low (lo) amounts as soluble antigen [sHEL(hi/lo)], (ii) localized within the endoplasmic reticulum (ER)/salvage compartment (ER-HEL), or (iii) anchored on the cell surface as an integral membrane protein (mHEL). The corresponding sHEL, ER-HEL, and mHEL gene products were expressed as predicted except that HEL determinants accumulated in the culture supernatant as well as on the cell membrane of mHEL-transfected cells. Class II-positive cells endogenously expressing all three forms of HEL antigen constitutively presented the immunodominant HEL-(46-61) determinant with differential efficiency (mHEL, sHEL greater than ERHEL) to a class II-restricted T hybridoma. A second T hybridoma recognized endogenous HEL-(46-61) determinants constitutively presented on sHEL(hi) and mHEL transfectants but not on sHEL(lo) or ERHEL transfectants. The formation of HEL-(46-61)/I-Ak complexes in the ERHEL and sHEL(lo) transfectants was therefore limiting. Mixing experiments with different antigen-presenting cells indicated that the HEL-(46-61) determinant was derived from endogenous antigen rather than by reuptake of shed or secreted HEL determinants. We conclude that MHC class II molecules can present some antigenic determinants derived from endogenous proteins that are sequestered in the ER/salvage compartment as well as distally transported in the form of secretory or membrane antigens.

Prevention of diabetes in non-obese diabetic I-Ak transgenic mice.

The non-obese diabetic (NOD) mouse develops insulin-dependent diabetes mellitus (IDDM) with mononuclear cell infiltration of the islets of Langerhans and selective destruction of the insulin-producing beta-cells, as in humans. Most infiltrating cells are T lymphocytes, and most of these carry the CD4 antigen. Adoptive transfer of T cells from diabetic NOD mice into irradiated NOD or athymic nude NOD mice induces diabetes. Susceptibility to IDDM in NOD mice is polygenic, with one gene linked to the major histocompatibility complex class II locus, which in NOD mice expresses a unique I-A molecule but no I-E. Speculation exists as to the role of the I-A molecule in the diabetes susceptibility of NOD mice, especially regarding the significance of specific unique residues. To examine the role of the NOD I-A molecule in IDDM pathogenesis, we made NOD/Lt mice transgenic for I-Ak by microinjecting I-Ak alpha- and beta-genes into fertilized NOD/Lt eggs. Insulitis was markedly reduced and diabetes prevented in NOD/Lt mice expressing I-Ak.

Structure function analysis of in vitro mutated CD4 and major histocompatibility complex class II gene products.

The Class II major histocompatibility molecules are implicated in the initiation of antigen-driven autoimmune responses by CD4-positive T cells. In order to study the structure and function of CD4 and MHC Class II molecules, strategies were developed with the intent of generating secreted forms of these molecules by in vitro mutagenesis of the respective genes. A full length cDNA encoding an expressible human CD4 molecule was mutagenized to introduce a premature stop codon corresponding to residue 367 located 8 amino acids amino terminal to the start of the predicted trans-membrane region. Following DNA-mediated gene transfer of the mutant gene, secreted CD4 was detected in the supernatant of transiently transfected COS-1 cells. Surface expression of the membrane-bound form of CD4 was detected under the same conditions. In an attempt to create a secreted form of the mouse Class II molecule I-Ak the exons encoding the connecting stalk, transmembrane and cytoplasmic domains of both the alpha and beta chains were replaced by the corresponding exons from the gene encoding a secreted Class I-like molecule, Q10b. Transfer of these genes into mouse L cells failed to generate detectable secreted I-A molecules. In view of the secretion of CD4 reported in other mutagenesis studies, it is concluded that very subtle differences in the structure of the COOH-terminus can influence the folding, solubility or transport of the CD4 molecule. In addition, the assembly of heterodimeric Class II molecules may require membrane anchorage of the separate chains or some other contribution from the COOH-terminal domains of the alpha and beta chains.

The extracellular domains of MHC class II molecules determine their processing requirements for antigen presentation.

We have evaluated the relative contributions of the extracellular and cytoplasmic domains of MHC class II molecules in determining the Ag-processing requirements for class II-restricted Ag presentation to T cells. Hybrid genes were constructed to encode a heterodimeric I-Ak molecule in which the extracellular portion of the molecule resembled wild type I-Ak but where the connecting stalk, transmembrane and cytoplasmic domains of both the alpha- and beta-chain were derived from the class I molecule H-2Dd. Mutant I-Ak molecules were expressed as heterodimeric membrane glycoproteins reactive with mAb specific for wild type I-Ak. Fibroblast and B lymphoma cells expressing either wild type or mutant I-Ak molecules were able to process and present hen egg lysozyme (HEL) and conalbumin to Ag-specific, I-Ak-restricted, T cell hybridomas or clones. The mutant-expressing cells presented native and peptide Ag less efficiently than the wild type-expressing cells, suggesting that the disparity in presentation efficiency was not due to a difference in Ag processing. CD4 interaction was intact on the mutant I-Ak molecules. Presentation of native Ag by mutant and wild type-I-Ak-expressing cells was abolished by preincubation with chloroquine, or after paraformaldehyde fixation. After transfection of a cDNA encoding the gene for HEL, neither mutant nor wild type-I-Ak-expressing cells presented endogenously synthesized HEL to a specific T hybrid. Newly synthesized mutant I-Ak molecules were associated with invariant chain. These data demonstrate the ability of hybrid class II molecules to associate intracellularly with invariant chain and degraded foreign Ag in a conventional class II-restricted processing pathway indicating that the extracellular domains of class II molecules play a dominant role in controlling these Ag-processing requirements.