It is more than 100 years since Robert Koch isolated Mycobacterium tuberculosis as the causative agent of one of the major chronic diseases affecting human populations. In the last third of the 19th century, the laboratories of Koch, Pasteur and their protégés firmly established the role of microorganisms in the aetiology of anthrax, rabies, cholera, diphtheria, typhoid, tetanus, plague and many other diseases. The germ theory was secure and accompanied by the widespread belief that such microorganisms could explain most diseases of human populations. Ronald Ross’s 1902 Nobel Prize for his confirmation of the malaria lifecycle 5 years previously, and Bruce’s turn-of-the-century work on trypanosomiasis, extended this optimism to parasitic infections. The advent of the era of antimicrobial therapy over the next 50 years gave rise to the possibility of eradication of the major communicable diseases (1).
Koch’s postulates, formalized in 1882, set the bar for establishing the microbial cause of disease; for most of the bacterial agents studied, these postulates were fulfilled, arguably culminating in the self-inoculation by Barry Marshall with Helicobacter pylori almost exactly 100 years later (2). Reversing decades of medical dogma that peptic ulcer disease had no microbial aetiology, his self-inoculation followed in the tradition going back to Pettenkoffer (who survived the Vibrio broth Koch sent him) and the discoverers of yellow fever, and including filarial infections (3). The use of live parasite inoculation as therapy had proceeded in parallel, from malariotherapy for neurosyphilis, nephrotic syndrome and HIV to helminth therapy for inflammatory bowel disease and asthma (4–9). Research on the potential role of microbes in the aetiology of diseases hitherto not associated with an infectious aetiology continues, most recently resulting in the implementation of routine immunization of young women against Human Papilloma Virus to prevent cervical cancer (10).
Some associations have not fulfilled these postulates. The putative link between Chlamydia infection and atherosclerosis is one example; Mycobacterium paratuberculosis as a causative agent of Crohn’s disease is another (11,12). Some have failed Koch’s postulates because of the inherent technical difficulties in culturing these organisms, and others because of the long latency between infection and chronic disease (for which Hill’s criteria offer a more appropriate test) and also the likely multifactorial nature of many of these diseases (12,13). Parasites, by virtue of their life cycles, are particularly challenging in this regard. Ultimately, pace Marshall, the proof of the pudding is not simply in the eating. For diseases with multiple aetiologies, the impact of eradication programmes will provide the most useful answers to the microbe-specific attributable fraction of the public health burden of these diseases, as commented by Ryan Wagner and Charles Newton in this work, with regard to neurocysticercosis (14). As a recent example, country-level helminth eradication programmes suggest in some instances only a small attributable fraction of disease caused by parasites that are known to cause chronic disease (15). Indeed, if anything, recent attention has focused on the ecological and immunological evidence to support a protective role for microorganisms against many chronic diseases, as reviewed in this journal in 2006 (16,17).
What of the mechanisms involved in chronic diseases? For diseases with undisputed associations, such as Schistosoma mansoni infection and portal hypertension, where granulomatous inflammation centred on helminth eggs is the hallmark of the pathological disease process, insights have been provided by research methods developed for the study of complex noncommunicable diseases. Genetic linkage studies of S. mansoni and of lymphatic filariasis (LF) are examples, harnessing the population-level variation in disease burden to further an understanding of immunopathogenesis (18–20). Opportunities for further cross-disciplinary insights should be afforded for other diseases associated with portal hypertension, or for diseases in which the pathology is dependent on tissue fibrosis for which other treatments may emerge (21). Treatment studies of S. mansoni-associated portal hypertension raised questions about the reversible nature of fibrosis; an early explanation was that type III collagen, not heavily cross-linked and therefore potentially resorbed, is a feature of early, reversible schistosomiasis (22). However, as highlighted in Zilton Andrade’s review, plastic casts of mouse liver vasculature have illustrated the role of angiogenesis and vascular remodelling in this condition (23). Hepatic stellate cells, the pericytes of the liver, may play a crucial role. The pro-angiogenic factors produced by schistosome eggs may provide a better understanding of tumorigenesis in S. haematobium-associated bladder cancer and other malignancies. Vascular endothelial-derived growth factors (VEGF) explored in schistosome-associated pathology may also mediate the lymphatic dilatation, which is the precursor to lymphoedema and hydrocoele in LF, as demonstrated in the review by Kenneth Pfarr et al. (20). Genetic polymorphisms in upstream promoters of VEGF-A correlate with disease manifestations. The role of the endosymbiont Wolbachia in this axis, and of its interplay with exogenous bacteria that contribute to chronic manifestations of LF, are the subject of ongoing studies.
The pathological role of tissue remodelling features in research into most chronic diseases, including those discussed in this work. The balance between matrix metalloproteinases and their inhibitors has been implicated in Chagasic cardiomyopathy and schistosomiasis, and once again models of non-communicable disease are providing insights (24,25). The absence of appropriate animal models, such as for chronic Chagasic cardiomyopathy, limits our ability to explore the mechanisms whereby parasites interact with these factors.
Attempts to explain associations between parasites and neurological disease are hampered by the limitations of epidemiological studies, which include inadequate case definitions, co-existence of multiple pathologies and huge population-wide variation in infection intensity. Even controlled trials may not be adequate, as the intervention may not be specific to the pathogen under scrutiny, as commented by Wagner and Newton (14). A better understanding of the immunological basis of epilepsy would help; recent insights, for example, on the role of autoantibodies in epileptogenesis are intriguing, given the literature on helminth-induced autoantibody production. That inflammation is principally induced after worm death (as in LF and other helminthiases) should provide avenues for exploration, in animal models, of down-regulation of immune responses in the brain by helminth antigens. By contrast, the epidemiological evidence associating toxoplasma infection with schizophrenia, appraised by Bob Yolken et al. in this issue, perhaps best fits a role for reactivation of latent infection, for which immunological insights might be afforded by studies of toxoplasma encephalitis in immunosuppressed animals and humans, and by further studies of parasite-strain-specific differences in immune response (26,27). Could the immune response associated with toxoplasma reactivation also have an effect on dopamine pathways, or might parasite-derived enzymes act directly?
Robust epidemiological evidence linking helminth infection to certain cancers has long been found, perhaps more than might be expected given the multifactorial nature of carcinogenesis. These strong associations have not helped clarify the pathogenesis behind this association. Does down-regulation of the immune response characteristic of chronic helminth infection limit tumour surveillance by the immune system (28)? Or does helminth-induced inflammation generate genotoxins, such as reactive oxygen species and N-nitroso compounds, or interact with exogenous toxins, well-described for aflatoxins in Hepatitis B-associated malignancy? (29). Birgitte Vennervald and Katja Polman summarize the evidence here (30). Insights from other disciplines exploring the role of inflammation in carcinogenesis may provide further understanding; might micro RNA be relevant (31)?
In summary, research on diseases with more speculative parasite associations are limited by the lack of good animal models, imprecise case definitions, impractical intervention studies and poor understanding of the immunopathogenesis of, for example, neuropsychopathology. For chronic diseases with established associations with parasitic infection, the latest research exploits tools for examining the roles of regulatory T cells and of pro-inflammatory cytokines, such as IL-17, in the immunology of parasite persistence, host inflammation and autoimmunity. Insights and therapies from non-communicable disease research, covering areas such as tissue remodelling and DNA methylation, may be important for parasite-associated chronic disease pathogenesis. Interactions between the immune response and these factors, and the role of genetic variation in both host and parasite, are ripe avenues for further research.