Letters 134

Human endogenous retroviruses in health and disease In his excellent and comprehensive review on human endogenous retroviruses (HERVs), Frank Ryan (December 2004, JRSM1) introduces a subject that could well become of central importance in 21st century medicine. To take one example, HERVs could provide the key to understanding how environmental changes associated with improved hygiene in the industrialized nations appear to be associated with the upsurge of certain diseases, notably autoimmunity, allergy and cancer. In this context, the risk of childhood leukaemia has been linked to hygiene-related factors,2 and an extensive multicentre study showed that the risk of melanoma is reduced by certain vaccinations—BCG and vaccinia—as well as some uncommon but serious infectious diseases early in life.3,4 Furthermore, the prognosis of melanoma patients who had previously received BCG and/or vaccinia was significantly better than that of unvaccinated patients.5 There is evidence that many cases of melanoma are associated with the expression of HERV-encoded proteins that induce malignant change by altering the intracellular redox potential. The ‘Achilles’ heel’ of this process is that a HERV-derived nonapeptide or decapeptide epitope, HERVK-MEL, is expressed on the surface of the cell undergoing malignant transformation, in principle enabling it to be recognized by the immune system.6 We have shown that BCG, vaccinia and the agents causing the infectious diseases that afford protection against melanoma have analogues of the HERV-K-MEL epitope and we have therefore postulated that these vaccinations or infections generate populations of cross-reactive T cells able to engage in immunosurveillance and effect repair or destruction of melanocytes expressing this antigen.7 Accordingly, improvements in hygiene in the industrially developed countries resulting in less infectious diseases, together with the abandonment of vaccinia vaccination in all such countries and BCG in many, could be having a deleterious effect on the development of effective immunosurveillance against at least one type of cancer. Further epidemiological studies to determine whether malignant disease other than melanoma and leukaemia are affected by hygiene-related factors, as well as laboratory studies to confirm the molecular basis of carcinogenesis and immunosurveillance, are required. Such studies could lead to the rational development of effective preventive and immunotherapeutic agents based on the relevant HERVencoded epitopes or on micro-organisms, including some currently available vaccines, which induce cross-reacting immune responses to such epitopes. John M Grange1 Bernd Krone2 Klaus Kölmel2 Centre for Infectious Diseases and International Health, University College