The p-i Concept: Pharmacological Interaction of Drugs With Immune Receptors

Abstract

The immune response in drug hypersensitivity is normally explained by the hapten hypothesis. It postulates that drugs with a molecular weight of less than 1000 D are too small to cause an immune response per se. However, if a chemically reactive drug or drug metabolite binds covalently to a protein and thus forms a so-called hapten-carrier complex, this modified protein can induce an immune response. This concept has recently been supplemented by the p-i concept (or pharmacological interaction with immune receptors), which postulates that some drugs that lack hapten characteristics can bind directly and reversibly (noncovalently) to immune receptors and thereby stimulate the cells. For example, a certain drug may bind to a particular T-cell receptor, and this binding suffices to stimulate the T cell to secrete cytokines, to proliferate, and to exert cytotoxicity. The p-i concept has major implications for our understanding of drug interaction with the specific immune system and for drug hypersensitivity reactions. It is based on extensive investigations of T-cell clones reacting with the drug and recently of hybridoma cells transfected with the drug-specific T-cell receptor for antigen (TCR). It is a highly specific interaction dependent on the expression of a TCR into which the drug can bind with sufficient affinity to cause signaling. Small modification of the drug structure may already abrogate reactivity. Stimulation of T cells occurs within minutes as revealed by rapid Ca influx after drug addition to drug-specific T-cell clones or hybridoma cells, thus, before metabolism and processing can occur. As the immune system can only react in an immunologic way, the symptoms arising after drug stimulation of immune receptors imitate an immune response after recognition of a peptide antigen, although it is actually a pharmacological stimulation of some T cells via their TCRs. Clinically, the p-i concept could explain the sometimes rapid appearance of symptoms without previous sensitizations and the sometimes chaotic immune reaction of drug hypersensitivity with participation of different immune mechanisms while normal immune reactions to antigens are highly coordinated. Nevertheless, because the reactions lead to expansion of drug-reactive cells, many features such as skin test reactivity and stronger reactivity upon reexposure are identical to real immune reactions.

Figure 1

The p-i concept. A schematic representation of the p-i concept in comparison with the hapten concept. A, In the p-i concept, the drug (eg, SMX) binds to the TCR and provides some initial signal. In most instances, this signal is insufficient to induce T-cell activation with cytokine synthesis and proliferation. The signal is strengthened by the additional interaction with HLA molecules. It seems to be rather independent of the type of peptide embedded and sometimes even from the HLA allele. Drug binding and HLA interactions stimulate T cells like a normal peptide/HLA complex. Because this way of T-cell stimulation does not follow the rules of the development of a normal immune response, the subsequent activations appear clinically and upon in vitro analysis often chaotic and uncoordinated. B, In the hapten hypothesis, the hapten-modified peptide is recognized and stimulates T cells. The hapten may also have the ability to activate the innate immune system, for example, dendritic cells.

Figure 2

HLA-allele association and p-i concept. Some drugs elicit severe, often CD8-dominated, immune reactions in patients with a certain HLA-B alleles (carbamazepine and HLA-B*1502, allopurinol with HLA-B*5801, abacavir with HLA-B*5701). If these drugs stimulate via the p-i concept (as well documented for carbamazepine [CBZ], but not abacavir or allopurinol), then the strong HLA association would be explained by the ability of only the selected HLA allele to supplement the activation caused by the respective drug. In the example given, the HLA alleles B*0701 or B*2701 would be unable to supplement this initial signal by CBZ. However, none of the existing hypotheses can explain the finding that some of these alleles are found to be involved only in certain human races (eg, CBZ and HLA-B*1502 in Han Chinese). SJS indicates Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis.