Getting the bigger picture

Technology has allowed immunology to develop at a vast rate. But there is still room for simple techniques to generate important data – indeed simplicity can be vital, especially when it comes to the immunoepidemiological study of malaria.

IMMUNOEPIDEMIOLOGY is an important part of the study of infectious diseases.  By combining immunology with epidemiology, we study not only the distribution and frequency of infection and disease but we also uncover how types and levels of immunity vary over time, in relation to exposure to infection or clinical presentation.  At the individual level, this can reveal how factors such as frequency of infection or genetic variation affect levels of immunity and, at the community level, can reveal how immune responses influence the prevalence of infection and the burden of disease. 

Immunologists have discovered much about host/parasite interactions by using laboratory animal models of human infections. However, to fully comprehend these complex relationships we not only need to be able to understand the immunological processes involved, but also how these processes influence, or are influenced by, variables they face in the real world. Studying an immunological process at a fixed point in time, in a controlled environment, does not tend to give a clear picture of the constant interaction and evolution between the parasite, the host and the environment which they both inhabit. For instance, when studying immunology in laboratory bred animals it is possible to change an individual variable and monitor the effects over a period of time. However, this is not possible when dealing with a human population and it is difficult to simulate the heterogeneity of the real world in controlled laboratory environments. In humans, variables such as age, nutritional status, genetic make-up and previous exposure to infection can have a profound effect on the immune response to a pathogen; frequency of exposure to a pathogen can in turn be affected by  environmental factors such as physical geography, climate, weather, housing and so on. Environmental factors play a particularly prominent role in transmission of vector borne diseases such as malaria, filariasis, schistosomiasis or arboviruses. In order to determine the impact of, or avoid confounding by, at least some of these variables, a large population needs to be studied over a lengthy period of time. Mathematical models can also be used to assess, or adjust for, the impact of some of these variables

Malaria affects 40 % of the world's population and causes up to 500 million clinical cases and over one million deaths each year. With children bearing the greatest burden of this mortality, it is estimated that every 30 seconds, a child somewhere dies of malaria. Although the human immune response to malaria has been the subject of much study over many years, a clear understanding of the mechanisms involved in the development of acquired protective immunity to the disease is still lacking. Without this knowledge, the development of an effective vaccine remains stalled.

Malaria is far and away the best described example of a disease exerting selective pressure on the genetic make-up of an exposed population1 (the protection afforded against malaria counterbalances the deleterious effects of genes such as Haemoglobin S, responsible for sickle cell anaemia, alpha- and beta-thalassaemia and G6PD-deficiency), and other malaria-protective polymorphisms are likely to be found at high frequency in affected populations. If we can identify these new genes and match protective polymorphisms to immune function we may gain insight into why some individuals suffer from life-threatening complications such as cerebral malaria even though most malaria cases resolve without major clinical consequences; in the longer term this information may allow us to design novel treatments and vaccines.

In addition to genetic background, another factor which is of major importance in influencing the pattern of malaria seen in a community is the intensity of malaria transmission by infected mosquitoes. Being able to accurately estimate the level of malaria transmission in a given area would mean the burden of malarial disease could be more precisely predicted and the success (or otherwise) of malaria control strategies could be more reliably calculated. 

We are currently using immunoepidemiological approaches to address both of these issues – i.e. the impact of host genetics, and of variation in malaria transmission ,on immune responses to malaria.  Both projects require the development and application of robust, transportable (and preferably cheap and technologically simple) methods to measure malaria-specific immune responses. As a first step we have been focussing primarily on antibody responses as they can be accurately measured with simple ELISA assays and there is a 20 year history of using recombinant proteins as surrogates for native antigen.

Classical ways of measuring malaria transmission can be time consuming and

Transmission by infected mosquitoes is a major factor in the pattern of malaria in a community

subjective. Estimating disease prevalence from hospital records is notoriously unreliable due to over-diagnosis of malaria in patients and the fact that large numbers of the community do not use health facilities. The classical way of measuring the transmission of malaria is by calculating the entomological inoculation rate, EIR (the average number of infectious bites, per person, per year). This involves estimating the number of infectious mosquitoes (always a small proportion of the total) and the average number of bites per person throughout the year. However, this method is labour intensive and complex due to seasonal variation in transmission, irregular spatial distribution of mosquitoes and heterogeneity in individual biting rates. For example, Smith et al. 2 estimate that 20% of people get 80% of the mosquito bites. Measuring EIR is also time consuming and requires specialist equipment and training.
Another common way to estimate malaria transmission is from the proportion of individuals in a population carrying parasites in their blood. However measuring parasite prevalence in blood samples by microscopy can also give misleading results due to host immunity (which reduces parasite numbers, lowering the estimate of transmission), seasonal variation, and varying precision of slide-reading. We have therefore developed a serological approach to estimating the transmission of malaria 3 which we hope will result in simpler, more accurate and more robust estimates of disease burden which, in turn, will facilitate the evaluation of control strategies and help in the design of new ones.

In this Wellcome Trust-funded study, we are estimating the age-specific prevalence of antibodies to specific malaria antigens within the population, and using a simple mathematical model of the change in seropositivity rates with age to estimate the transmission rate. Seroprevalence data is less affected by seasonal variations in transmission than parasite rates as malarial antibodies remain in the blood for a relatively long time, compared with the lifespan of the vector and half-life of individual malaria infections. The objective of the study is to validate our method of estimating transmission by testing samples from a wide variety of different malaria endemic settings and comparing our estimates with the “gold standard” estimate (i.e. EIR). Some samples are being collected into a serum repository for testing at LSHTM; others are being tested - with a standardised methodology and reagents - by collaborating laboratories in Africa, Asia and South America who wish to generate the data themselves.

MalariaGEN (the Malaria Genetic Epidemiology Network, www.malariagen.net) is a large multi-national project funded through the Grand Challenges in Global Health Initiative (www.gcgh.org). Its objective is to bring together experts from around the globe to explore and identify critical mechanisms of protective immunity against malaria which could lead to successful malaria vaccine development. Many of these collaborating institutions are in malaria-endemic countries in Africa and Asia; Kenya, Tanzania, Mali, Senegal, Burkina Faso, Sudan, Sri Lanka, Thailand and Vietnam. Although a number of genetic polymorphisms have already been associated with susceptibility to, or protection from, severe malaria, it has been difficult up to now to reconcile many of these with a clear mechanism of protection. As part of one of its three major work programs, MalariaGEN has established a repository for serum and plasma samples at NIBSC. Our project has three main aims : to develop the repository as a resource for testing hypotheses regarding the function of candidate genes; to develop a series of robust and transferable methods which reflect important aspects of immune function; and to make these standardised techniques available to partner laboratories as part of the investigation of the association between genes and immune function. The first step in this process has been to develop immunoassays which can be of use in identifying genes that influence antibody production to malaria antigens.

For both these projects we need to develop assays that generate accurate quantitative and qualitative measures of antibody concentration, specificity, class and subclass. As investigating malaria transmission and genetic association is a long term task, it is very important that the data obtained is reproducible between samples and experiments. As the burden of malaria is carried on the shoulders of the developing countries of the world it is imperative that the equipment used and the expertise and time needed are kept to a minimum. Maintaining and servicing complex equipment in less resource-rich countries is often expensive and difficult to organise and involves long response times. Methods need to simple, as well as reproducible and robust.  Previous techniques of measuring antibody prevalence, such as radioimmunoassay and immunofluorescence have proven expensive and, in the latter case, is limited by the subjective nature of the data. We have therefore focussed on ELISAs, which are relatively low-tech, have a reasonably high throughput and are relatively inexpensive to run. They require simple equipment and are reproducible. 

Since we have taken this deliberate decision to develop methods that are as simple, and require as modest resources as possible, automation must go out of the window – we do not use plate stackers or robots, and our plate washers have two highly versatile articulated arms, but come packaged in a lab coat (we find these much more personable and easier to program than commercial robots). The only essential piece of electrical equipment is a plate reader. For the fitting of prevalence data, access to a computer running Excel is necessary at present, but we intend to develop independent software tools which are not reliant on proprietary packages. Our goal, for the Wellcome Trust funded project, is to develop a simple assay kit that can be transported to laboratories all over the world that will produce reproducible, consistent data for the levels of specific anti-malarial antibodies in sera from malaria endemic populations. A consistent way of measuring seroprevalence of antibodies in populations in different countries will result in large amounts of comparable data which can be fitted to a mathematical model which allows us to calculate a number (?) representing the force of infection, which previous results have shown to correlate very closely with EIR.  For the MalariaGEN project, the assays have the additional dimension of being chosen to shed as much light as possible on the general function of the humoral system. But as one of the requirements for MalariaGEN sites is a detailed knowledge of local malaria transmission, we foresee that we may be able to cross-fertilise the two projects by using seroprevalence data to estimate transmission at MalariaGEN sites while using independent estimates of transmission intensity, where they exist, to support the model we are developing within the Wellcome Trust project.

As a methodological simplification, rather than using serum or plasma samples, which require specialist storage conditions, we are currently looking at using blood spots collected onto filter paper from finger pricks. Blood spots are extensively used for the collection of samples in clinical chemistry and have been used on numerous occasions to monitor antibodies to viral pathogens as well as, occasionally, to malaria. Furthermore, blood spots are routinely collected for genotyping malaria parasites and can be stored desiccated at 4°C for extended periods. Many immunoassays can be carried out on the antibodies contained in a single spot. We are in the final stages of validating standard operating procedures for blood spot collection, storage and use.

Both projects have at their core the maintenance of a repository of samples for future use. Most research departments involved in large field studies will be familiar with the situation where, at the end of a study, large sample sets finish up as an unwelcome - but potentially valuable - burden on scarce freezer space. Some time ago LSHTM received a grant from the MRC to establish a secure repository for plasma and serum samples which had accumulated within the School (www.lshtm.ac.uk/malaria/repository.htm), and a similar repository is being set up at NIBSC to maintain the samples which will accrue from the MalariaGEN project. In both cases, samples are aliquoted into individually bar-coded tubes, each of which is located to a bar-coded rack. Access to each sample– when, who by, how often and how much - can therefore be simply logged using a bar code scanner and simple desk top computing software. Access to repository samples is controlled through a supervisory committee which reviews the scientific value of the potential use of the sample, ensures that the proposed use is compatible with the ethical approval and informed consent given for the original study and that sample storage complies with current UK and EC legislation.

The technical environment in which immunological research is carried out has steadily been increasing in sophistication and complexity. These two projects, however, are predicated on the thesis that there is still a place for simple techniques to generate novel and important data if focussed on interesting questions and interpreted in fresh ways, especially when coupled to the power of genomic screening. We are looking forward to an interesting three years following the threads of these topics towards their conclusion.

References
1. Kwiatkowski, D.P. (2005) How malaria has affected the human genome and what human genetics can teach us about malaria. Am J Hum Genet 77, 171-192
2. Smith, D.L., et al. (2005) The entomological inoculation rate and Plasmodium falciparum infection in African children. Nature 438, 492-495
3. Drakeley, C., et al. (2005) Estimating Medium and Long term trends in malaria transmission using serological markers of malaria exposure. Proceedings of the National Academy of Sciences 102, 5108-5113

Acknowledgements
We would like to thank our colleagues, particularly Dr C Drakeley and Professor D Kwiatkowski, for support and inspiration. MalariaGEN is funded by the Bill and Melinda Gates Foundation (via the Foundation for the National Institutes of Health) and by the Wellcome Trust. The research on Serological Indicators of Transmission is funded by a grant from the Wellcome Trust.

By Jackie Cook, Paul Risley, Elaenor Riley and Patrick Corran.
(from left) Jackie Cook has a BSc in Biology with English from Keele University and is a Research Assistant at the London School of Hygiene and Tropical Medicine (LSHTM).  Paul Risley is a  Scientist in the Division of Biotherapeutics at NIBSC and has been working on the MalariaGEN project since November 2006. Eleanor Riley is Professor of Infectious Disease Immunology in the Department of Infectious and Tropical Diseases at LSHTM. Patrick Corran is a Senior Scientist in the Division of Biotherapeutics at NIBSC, on part-time secondment to LSHTM