Antigen-presentation of non-peptidic antigens lipid trafficking and loading

T cells recognize a broad variety of antigens, including peptides, lipids and non-peptidic phosphorylated metabolites. Clarification of the rules rendering non-peptidic molecules immunogenic is essential to understand and to influence the reactions of the immune system to this class of substances in health and disease. Despite recent advances in research about immune responses to non-peptidic compounds, important issues remain unanswered. Molecular mechanisms governing the immunogenicity of non-peptidic ligands such as their cell internalization, trafficking within intracellular organelles, association with dedicated antigen-presenting molecules, induction of central and peripheral tolerance, and finally their role in autoimmune diseases as well as in protection during infections are unknown to date.
The aims of this thesis were to assess some of the immunological functions and cell biological rules governing the immunogenicity of non-peptidic antigens, with particular emphasis on cell trafficking of non-peptidic antigens and antigen-presenting molecules. It focused on (i) the antigen reactivity and presence of human invariant natural killer T (iNKT) cells in diseases, (ii) the role of CD1a trafficking in lipid antigen presentation by this protein, and (iii) the requirements of membrane translocation of phosphorylated mevalonate metabolites that stimulate human T cell receptor (TCR) gamma-delta cells.
With the development of alpha-galactosylceramide (alpha-GC)-loaded soluble CD1d dimers, which specifically interact with the TCR of iNKT cells, we have the perfect tool in our hands to perform detailed studies on iNKT cells. Analysis of the iNKT cells in blood unveiled large differences in their fluorescence intensity suggesting the presence of semi-invariant iNKT TCR with large disparities in the affinity for the alpha-GC-CD1d complex. Unexpectedly, established iNKT cell clones showed no correlation between CD1d dimer-staining levels and alpha-GC reactivity, indicating that additional mechanisms control responsiveness of iNKT cells, at least to this lipid antigen.
The identification of lipid antigens stimulating exclusively some desired functions in human iNKT cells might lead to new medical therapies or vaccines. To screen a variety of synthetic lipids for their capacity to activate iNKT cells, we devised an in vitro model based on plastic-bound CD1d. Piperidinones, molecules with a ceramide- or sphingosine-like structure, revealed that a single lipid tail is sufficient to form stimulatory complexes with CD1d. Interestingly, piperidinones preferentially induce TH1-like cytokines, predicting a possible role as novel leader molecules to functionally direct iNKT cell responses deployable in clinical therapies.
The balance of proinflammatory TH1 to regulatory TH2 cytokines is well-known to be decisive for the outcome of many diseases. Atherosclerosis (ATH) is a chronic inflammatory disease characterized by lipid accumulation in plaques. The disease is complicated by cardiovascular events provoked by plaque rupture or erosion. Because inflammation participates in lesion progression and rupture of plaques, the identification of its causes and of the culprit leukocyte populations involved in plaque destabilization is crucial for effective prevention of cardiovascular events. We used CD1d dimers to detect and characterize iNKT cells in ATH patients. We found that, in human atherosclerotic lesions, the abundance of CD1d-positive antigen-presenting cells (APC) and of iNKT cells correlates with disease severity and activity. CD1d-positive cells colonize advanced plaques in symptomatic patients and are most abundant in plaques with concomitant signs of ectopic neovascularization. In plaques, the frequency of iNKT cells among total T cells exceeds the one in blood. After having successfully isolated iNKT cell lines from plaque tissue, we showed that they promptly release proinflammatory cytokines upon lipid antigen stimulation and promote endothelial cell migration and microvascular sprout formation in vitro. This functional proangiogenic activity is ascribed to interleukin-8 released by iNKT cells after lipid recognition. These findings introduce iNKT cells as novel candidates to induce plaque neovascularization and destabilization in human ATH. Targeting iNKT cells could lead to late stage ATH treatment.
Another approach to understand the role of lipid-specific immune responses is to investigate the molecular rules of lipid-CD1 complex formation. Lipids distribute, due to their physicochemical properties or with the help of specific transporters and lipid transfer proteins, to different intracellular compartments and membrane domains. Thus, it is advantageous for the immune system to utilize multiple CD1 isoforms, each with a distinct trafficking pattern, to facilitate sampling of lipid antigens localized in various membranes. Several studies have addressed trafficking of CD1 isoforms. However, the molecular mechanisms are known in only a few cases.
We identified invariant chain (Ii) and lipid rafts as key regulators of CD1a organization on the surface of APC and of its immunological function as antigen-presenting molecule. Colocalization of CD1a with Ii is dependent on raft integrity and CD1a internalization is increased by Ii. The localization of CD1a in lipid rafts is functionally relevant as raft disruption inhibits CD1a-restricted antigen presentation.
Moreover, we found that CD1a is internalized independently of clathrin and dynamin and that it follows a Rab22a- and adenosine diphosphate ribosylation factor (ARF) 6-dependent recycling pathway, similar to other clathrin-independent cargo. Posttranslational S-acylation of the CD1a cytoplasmic tail may occur but neither determines the rate of internalization nor recycling nor its localization to detergent-resistant membrane microdomains. These findings place CD1a close to major histocompatibility complex (MHC) class I in its trafficking routes although CD1a loads lipids in recycling endosomes and not in the endoplasmic reticulum as MHC class I.
Strikingly, the glycolipid antigen sulfatide was found localized predominantly to early and recycling endosomes where CD1a is located. Swapping the cytoplasmic tail of CD1a for the one of CD1b and hence targeting the CD1a protein to the late endosomal and lysosomal compartments decreases its capacity to present sulfatide and shortens the half-life of stimulatory complexes. Thus, the physiological intracellular trafficking route of CD1a is critical for efficient presentation of lipid antigens that traffic through the early endocytic and recycling pathways.
Intracellular trafficking of another class of non-peptidic antigens, namely the phosphorylated metabolites which stimulate human TCR gamma-delta cells expressing the Vgamma9/Vdelta2 heterodimer, was examined. These T cells recognize a family of structurally related compounds produced in the eukaryotic mevalonate and prokaryotic methylerythritol phosphate (MEP) pathways. The endogenous self-ligands are generated within the cytoplasm and must cross the membrane in order to associate with dedicated antigen-presenting molecules, which remain unknown at present.
Using an in vitro transport assay, we demonstrated that the multidrug resistance-associated protein (MRP) 5 transporter is involved in membrane translocation of antigenic phosphorylated metabolites. Confocal microscopy illustrated that MRP5 is located in membranes of both endoplasmic reticulum and early endosomes. Both the intracellular localization and active role in antigen transport confer an immunological function to MRP5, resembling that of TAP (transporter associated with antigen processing) transporters involved in peptide antigen translocation. This indicates a similar strategy used for antigen presentation to TCR alpha-beta and gamma-delta T cells.
In conclusion, these studies have underlined the physiological relevance of T cells recognizing non-peptidic ligands and have revealed unanticipated molecular mechanisms controlling the efficient presentation of such antigens.