Getting to the matter of the heart

What can a three-generational study in a small American town tell us about cardiovascular disease? The potential is endless says George Lipscomb - as long as the techniques used to mine this unique biobank are sound

The Framingham Heart Study (FHS) is one of the most well known and prestigious clinical studies in medicine, and has helped to shape our understanding of cardiovascular disease (CVD) for over 60 years. This multi-generational study was started in 1948 by the National Heart, Lung, and Blood Institute (NHLBI), and originally included more than 5,000 adult residents of the town of Framingham, Massachusetts. Now on its third generation of participants, the project is run in collaboration with Boston University and continues to provide valuable data and scientific understanding on the epidemiology and progression of CVD.

As part of the study’s prospective design, a large biobank has been created, housing frozen serum, plasma and urine samples from each participant. In addition to traditional clinical chemistry and biochemistry studies, these samples are now being used in a range of genomic and proteomic investigations, employing state-of-the-art techniques to ensure the best use of this finite and valuable resource. As part of the FHS’ Systems Approach to Biomarker Research in Cardiovascular Disease (SABRe CVD) initiative, the NHLBI and Boston University recently sought research collaborators to develop new tests for the validation of novel circulating biomarkers for CVD, focusing on quantitative measurement of a large number of potential biomarkers using minimal sample volume. The aim of this five year project is to promote research into these biomarkers for the development of new diagnostic tests, to help identify individuals at high risk of CVD, and as potential therapeutic targets. It is hoped that by identifying the proteomic signature of CVD risk factors, such as atherosclerosis and metabolic syndrome, the disease promoting pathways associated with these factors can be elucidated, providing novel targets for the treatment and prevention of disease.

Thanks to its forward looking design, the FHS is ideally positioned to enable identification of novel risk factors for CVD. Prospective circulating biomarkers have been selected for the project by the NHLBI and Boston University, based on expert consensus through careful review of the literature. Factors for selection included: previously identified biomarkers for atherosclerosis and/or metabolic syndrome; products of genes implicated in atherosclerosis and/or metabolic syndrome; components of pathways implicated in atherosclerosis and/or metabolic syndrome; and genes associated with the phenotypes of interest.

Based on these criteria, an initial panel of approximately 180 potential markers has already been established. This diverse collection includes targets of interest from both a discovery and verification standpoint, as well as markers which may have specific clinical value for diagnostic or therapeutic applications. This panel of prospective markers will be continuously reviewed over the five year duration of the project, ensuring both biological and clinical relevance is maintained.

Over the course of the study, material from 7,000 individuals of the ‘offspring’ and ‘third generation’ cohorts will be quantitatively analysed for the selected biomarkers. There are a number of hurdles which must be overcome to ensure accurate, high resolution profiling. The first, and most obvious, challenge is the limited availability of material for analysis. The FMS biobank is an extremely valuable resource, and the linear nature of the study means that all samples within the biobank are irreplaceable. As such, it is vital to maximise data acquisition while minimising sample consumption.

While multiplexing technologies offer at least a partial solution to this issue, the situation is further complicated by the relative concentrations of each potential biomarker. While many established biomarkers occur in relatively high concentrations within plasma, the detection and quantification of low level markers must not be compromised by the relative abundance of these more prevalent molecules. To minimise the impact of this issue, candidate markers have been prioritised according to clinical or biological relevance, and the analytical workflow has been carefully designed to work across a large dynamic range without introducing bias towards commonly occurring markers.

Another aspect of marker validation which must be considered is the dynamic nature of the plasma samples. As is true of all proteomic analyses, the biochemical signature of the sample is not static unless kept frozen at very low temperatures and, even then, some degradation may still occur. Therefore, during the course of analysis, each aliquot of biological material undergoes a series of small but significant changes, including enzymatic activity – such as post-translational modifications and proteolysis – and physical degradation. Accurate and truly representative analysis therefore requires an expert understanding of the biology of the system and the likely consequence of these processes on the results obtained, as well as the sophisticated technical capabilities required to limit their impact.

ELISA techniques are commonly used for detection and quantitation of a wide range of biomolecules, exploiting the high specificity potential of antibody biorecognition to distinguish between biologically and chemically similar species. Sandwich ELISA systems offer excellent analyte specificity and sensitivity over a broad dynamic range, using pairs of antibodies corresponding to discreet epitopes to allow reliable detection of the biomarkers of interest across a large concentration range. However, this technique relies on high affinity, high fidelity antibodies to ensure that only the target molecule is detected, and to avoid masking of low level markers by more prevalent molecules. This makes optimisation and purification of antibodies highly specific to each potential biomarker vital, and requires the capacity to develop new antibodies for novel biomarker targets.  

Once suitable antibody pairs have been selected, multiplexing offers a powerful solution to deal with limited sample availability, allowing simultaneous quantitation of numerous analytes in a single aliquot of material. Commercially available, bead-based multiplex ELISAs offer a robust, validated method for parallel detection of up to 100 distinct biomarkers at picogram per millilitre concentrations (Figure 1), minimising sample requirements for analysis.

The ongoing nature of the FHS provides a unique opportunity for multi-generational investigation of biomarkers for CVD, however, the restricted availability of biobanked samples makes development of a robust, reliable and highly sensitive methodology essential.

Selection and optimisation of antibodies by singleplex ELISA is necessary prior to multiplex assay development. In addition, low abundance biomarkers may require enrichment to improve the signal-to-noise ratio. This complex process requires extensive experience in both protein purification and antibody development, and necessitates both purified antigens and non-FHS plasma samples for validation.

Following validation of individual antibodies, a pilot-scale study of the multiplex assay format will be conducted using four previously characterised markers of CVD, first on non-FHS material, then with plasma samples from the FHS biobank. Only once this system has been rigorously tested and validated will full scale quantitation of novel biomarkers for CVD in the FHS samples begin.

Quantitative data generated from this project will be correlated with phenotypic, genotypic and clinical information as part of the SABRe CVD initiative, providing new insight into the genetic and environmental factors contributing to CVD. In addition, antibody sets developed for identification of CVD biomarkers will be available for research into new diagnostic and therapeutic strategies for individuals at high risk of CVD.

Author: George Lipscomb is Global Market Segment Manager for Proteomics, Sigma Life Science