Allergen isoforms for immunotherapy: diversity, degeneracy and promiscuity

Speci®city for antigen is the hallmark of adaptive immunity and is conferred by cells of the lymphocyte lineage. Aberrant immune responses lead to disease in an antigenspeci®c fashion. In order to avoid the generalized immunosuppressive effects of treatments such as corticosteroids and cyclosporin, efforts have been increasingly focused on devising therapeutic strategies which are equal in speci®city to the disease-initiating event. In the early 1980s Lamb and colleagues demonstrated the induction of antigen-speci®c non-responsiveness following challenge of puri®ed T-cell clones with high doses of speci®c peptide [1]. The concept of high dose peptide antigen for immunotherapy is attractive, particularly in the context of allergic diseases where conventional desensitizing therapy has been confounded by the presence of anaphylactogenic immunoglobulin (Ig) E antibodies. Peptide-induced tolerance has been demonstrated in a number of animal models of disease such as experimental autoimmune encephalomyelitis (EAE) [2,3], collageninduced arthritis [4,5], diabetes [6] and more recently, in models of allergic disease. Briner and colleagues sensitized mice to the cat allergen Fel d 1 and subsequently demonstrated the ability of allergen-derived peptides to downregulate T-cell cytokine and antibody production [7]. Hoyne and colleagues demonstrated the ability of peptides from the house dust mite allergen Der p 2 to downregulate T-cell and antibody responses to challenge with intact protein and also to prevent sensitization by prior administration [8]. More recently, attempts have been made to translate ®ndings in animal systems into viable therapies for human disease. Norman and colleagues evaluated the ability of two peptides from Fel d 1 to downregulate allergic responses to cat in 95 sensitized individuals in a double-blind placebocontrolled trial [9]. Ef®cacy was demonstrated but fell below expected levels resulting in cessation of the programme. More recently, the pendulum of future therapy for allergic diseases has swung in favour of production of recombinant allergens [10]. In particular, interest has focused on allergen isoforms and their genetically modi®ed counterparts [11± 14]. These molecules display reduced IgE binding whilst retaining T-cell epitopes and thus hold promise for safe desensitizing allergen immunotherapy. However, despite the current interest in recombinant allergen and isoformbased therapy, the immunology underlying responses to these proteins, particularly within the context of natural environmental exposure, is complex. In the current issue of the Journal, WuÈrtzen and colleagues [15] have investigated the epitope speci®city of T-cell lines and clones reactive with the group 5 grass allergen of Phleum pratense. Phl p 5 has two major isoforms which share regions of profound sequence homology but also differ considerably in some areas. Within the disparate regions of the two isoforms, WuÈrtzen and colleagues have identi®ed isoform-speci®c T-cell epitopes. The signi®cance of this ®nding should be placed in the context of the large numbers of allergen proteins which have multiple isoforms. These appear particularly to be derived from plant species and are less common among animal dander and mite allergens. The authors suggest that since modulation of the speci®c T-cell response appears to be pivotal in successful immunotherapy, recombinant and/or isoform-based whole allergen immunotherapy will need to take such diversity into account. Unlike animal models, the human population is heterogeneous in terms of major histocompatibility complex (MHC) expression. Since the expressed MHC alleles dictate which sequences will be Tcell epitopes, the numbers of potential epitopes within a single allergen molecule is very large. Ultimately, the numbers of isoforms required for a single therapeutic preparation to be ef®cacious at the population level, may make this approach impractical. The same argument is brought to bear on the potential of allergen-derived peptide sequences for therapy of allergic disease mediated by allergens with multiple isoforms. Whilst these arguments are compelling there are, at least conceptually, counter-arguments which can be applied based on some recent observations about the way in which T-cell receptors and MHC molecules interact with peptides and also the way in which tolerance induced by a region of a molecule, can effect tolerance to the entire molecule. The binding of peptide sequences to MHC molecules is dependent upon peptide sequence and charge, and the shape of the peptide-binding groove. Individual MHC molecules clearly have the capacity to bind large numbers of different peptide sequences [16]. Equally, certain peptide sequences have the capacity to bind almost universally to MHC molecules [17]. It follows therefore that many antigenderived peptides recognized by T cells can bind to more than one MHC molecule and are thus `promiscuous' in this regard. Moreover, whilst some peptides bind promiscuously to MHC molecules, it has been demonstrated that T-cell