Blood is one of the most easily accessible body fluids that contains a wealth of information about the state of human health. Mining this information can provide details about each individual’s homeostatic state and states of pathology, both one-time and chronic. In view of challenges offered in the arena of blood diagnostics, the NHLBI organized a Working Group to identify the future of blood diagnostics which is within the mission of HLB Institute. The Workshop was organized by Program Director Pankaj Qasba along with co-Chairs, Drs. John Stamatoyannopoulos (University of Washington), Garry Nolan (Stanford) and Shankar Subramaniam (UC San Diego). The detailed Workshop Program is presented here. Leaders in the areas of diagnostic technologies, blood measurements, and health leaders were invited to participate in the workshop.
The primary goal of the Workshop was to identify key challenges in blood diagnostics and potential applications to human health. Specific issues raised for discuss included:
- What is the smallest volume of blood that would be sufficient to identify key molecules that discriminate normal and diseased states?
- What are best ways of extracting and preserving blood without expensive storage strategies?
- What is the dynamic range of measurements of metabolites, proteins, nucleic acids and other molecules that can be measured in the smallest volumes of blood?
- What technologies are available for extracting and identifying cells from blood?
- What are the technologies currently available for molecular measurements in blood?
- What informatics and modeling strategies are available or need to be developed to mine the data from blood to provide knowledge of human physiology?
- How can we carry out longitudinal monitoring that would provide information about the steady state and variations from the steady state?
- How can one assess the influence of diet, drugs and environment on alterations in blood composition?
- Can blood-type be an indicator of susceptibility to human pathologies?
- Can we create a catalog of normal and deviant measurements in blood that can used to provide insights into personalized and precision medicine?
The discussions at the Workshop provided significant insights into the above challenges. Several nascent cell and molecule based blood diagnostic technologies were presented. Applications to key human pathologies including liver disease, vascular diseases and alterations post-vaccination were discussed. The recommendations emphasized the need for development of further non-invasive molecular and cellular detection technologies, the imminent need for informatics and computational strategies to characterize human health from blood measurements and the need to develop a catalog of blood measurements. It was suggested if a large number of metabolites that vary with the state of human health were to be identified, the notion of “metabotyping” an individual can become a reality. We present the details of the Workshop below.
Cells and Cell-based Markers in Blood:
Current diagnostic paradigms use either imaging techniques or solid tissue biopsies or a combination thereof for managing a variety of disease states. Single cell analysis technologies are coming into their own, notably recent progress in optical and next-generation sequencing-based approaches for RNA, DNA, and protein analyses. Single-cell analysis is particularly relevant for the analysis of blood due to the extreme diversity of circulating cells. Advanced molecular analysis of single cells presents key opportunities to characterize rare cell populations as defined by cell type, origin, or function and relieve requirements for physical purification of target cells in some workflows. In its early development, single-cell analysis has been heavily reliant on expensive instrumentation and specialized technology platforms such as microfluidic platforms. Continuing advances in DNA-encoded technologies such as molecular barcodes and microfabrication-independent approaches including in situ molecular processing and diffusion-constrained reaction matrices promise to make high-throughput single-cell routine and reliable. A particularly exciting direction is the confluence of single-cell dynamic functional assays with detailed molecular characterization of individual cells.
In addition to molecules that are diagnostic, blood contains rare cells and nanoscale vesicles that can serve as excellent markers for normal and pathophysiology. Current tools are not adequate to detect nano vesicles and a systematic characterization of these vesicles would be valuable. The detection and deep analysis of extremely rare cell types from the blood is both a great technology driver, and can provide a rich set of biomarkers and, with the advent of cell-based therapies, such analyses can also provide guidance for designing such therapies. In particular, antigen-specific CD4 T cells are probably one of the toughest class of cells to quantitate, and they are intimately involved in auto-immune disorders. There are reasonably good (albeit very new) ways to search for antigen-specific CD8 T cells from the blood, but CD4 cells which recognize MHC Class II, are much more challenging. Second, we all view as blood as a window into disease and health, but the intimate relationships between cells, protein, metabolites, etc., in the blood, and the originating diseased or healthy tissue, is an area that can be deeply mined. Most importantly several of the detection methods can be developed as point-of-care diagnostic tools.
Recent developments in regenerative medicine show that cells in blood can serve as excellent starting points for de-differentiation into induced pluripotent stem cells. This would be revolutionary if cells in blood can be transformed directly or through iPSCs into multiple tissues. These tissues can be used to a) identify molecular components that are tissue-specific and also present in blood, b) assess normal versus diseased states through targeted perturbations using CRISPR-Cas methods and c) screen drugs for efficacy and dose requirements. This will essentially bring blood as the most accessible human tissue for assessment of a host of human diseases in a personalized manner. Further, this would help in the development of novel therapeutic interventions.
Molecular markers from Blood:
Blood cell epigenomes as windows into central disease processes. Blood leukocytes traverse the vasculature and thus gain exposure to every body tissue bed. Additionally, different types of immune cells may transgress into and out of tissue beds. The epigenomes of blood cells are plastic, and can be altered by exposure to different environmental stimuli, raising the possibility that the epigenomes of circulating blood cells may record systematic ‘imprints’ of disease states affecting central organs and tissues. This is supported by recent studies showing associations between central disease processes such as schizophrenia and epigenomic profiles (e.g., DNA methylation) of peripheral blood mononuclear cells. However, the ability of the epigenomes of blood cells to acquire imprints of central disease states is currently largely unexplored.
Cellular and organismal origins of extracellular nucleic acids circulating in the blood of normal individuals. Circulating nucleic acids have been shown to contain subpopulations that derive from the host nuclear genome, the host mitochondrial genome, and microbial and plant genomes. Of these the most abundant category are those deriving from host cells and tissues. In principle, sequencing and mapping of circulating DNA fragment populations to the reference genome can be compared with reference maps of chromatin structure to infer both the cell/tissue of origin of the circulating fragments, as well as gene expression states within that cell/tissue. However, this and other analyses are complicated by the fact that many circulating nucleic acids are small species that complicate accurate sequence alignment. Methods and approaches are therefore needed for (i) systemic assignment of nucleic acids to specific positions in host or exogenous genomes, and (ii) the systematic assignment of circulating nucleic acids to a host cell or tissue type of origin.
Need for Technology:
Molecular based approaches to develop glycan signatures associated with CVD. A full catalog of glycans and glycoproteins in blood, their reference range, variations among the human populations, and disease-specific structural characterization is yet unestablished. There is growing evidence that such glycan-based biomarkers may play an important role in the early detection of blood disorders and cardiovascular diseases. These technologies can also amplify the measurement of proteomics based signals, since it is already established that several acute phase, complement and coagulation proteins are dysregulated during cardiovascular diseases.
The establishment of glycomics and glycoproteomics methods in normal individuals can set the stage for comparison of the glycome between normal and disease individuals. This is important for several reasons: 1. This can lead to the discovery of novel glycan-biomarkers, that can be translated to the clinic for both prognostic and diagnostic value. 2. It will enable the integration of glycomics data with multi-OMIC approaches that perform studies at the gene, protein and metabolome levels. 3. In the long run, such multi-OMICs data can result in systems biology based approaches that provide a link between glycan structure discover and basic science pathway information they regulate. Such studies can thus catalyze both translational research and basic science understanding that is critical for future efforts to improve human health.
The human metabolome is comprised of approximately 4,000 endogenous low molecular weight metabolites that map to known biochemical pathways, with the identification of an increasing number (~ 6,000 to date) of exogenous metabolites that are derived from plants and foods, exposure to drugs, chemicals, or nutritional supplements (Metabolomics Workbench, http://www.metabolomicsworkbench.org). Investigations have revealed perturbations in metabolomics profiles (or the metabotype) that correlate with many disease phenotypes (e.g., cancer, immunology, mental health, stress, diabetes, obesity, cardiovascular disease, lung and airway disease), as well as demographic and lifestyle factors [Quehenberger et al., 2010; Gorden et al., 2015]. In addition to disease diagnostics, metabolomics is being used to reveal biomarkers for the early detection of disease, to monitor treatment, to identify druggable or nutritional targets, and to provide insights to link exposure to disease outcomes. The ability to apply metabolomics to the analysis of relatively non-invasive biological fluids and excreta (e.g., blood, serum, plasma, urine, saliva, feces) makes it ideally suited to develop biomarkers or signatures (suites of markers) associated with an individual’s disease state, as well as to provide an indication of an individual’s state of wellness. Clinical panels that are used in routine physicals are comprised of biomarkers that have the potential to reveal the presence of disease, but these biomarkers fail to detect disease early on. Routine clinical testing involves the assessment of the fit of each measured analyte into an acceptable range. A more reliable approach would be to assess the overall pattern of changes in levels of these analytes in order to understand overall metabolic perturbations to improve early detection and diagnosis. Metabolomics provides a means to move past the point by point comparison of a diagnosis with an acceptable range, to providing a signature of analytes that can improve our ability to assess health and wellness at an individual level.
Disease based biomarkers will require stringent metabolite identification in studies designed for discovery, replication, and validation. Study designs may be at the population level, but because we know that metabolomics profiles are influenced by demographics (e.g., gender, age, ethnicity), lifestyle factors (e.g., place or residence, smoking status, nutritional intact), by physiological states (e.g., body mass, pregnancy status, lactation), and psychological stressors- it is likely that most disease based biomarker studies will be conducted using cross sectional study designs. Furthermore, there is a critical need to develop biomarkers that serve sub-populations of women, the elderly, and minorities who may be more vulnerable to disease risks due to windows of susceptibility, and adverse exposure scenarios. Biomarkers only studied at the population level may swamp out the ability to serve underrepresented subgroups.
Clinotyping: Health and wellness are not just a factor of genes, proteins, or biomolecules in the blood. Psychological stressors, nutrition, and social interactions – even income and insurance- all are known to have links to our health and wellness. All of these factors should be integrated to derive the most important markers that collectively mark an individual’s health. It is the intersection of the biological psychological and social interactions that drive health, and all of these are important metadata to needs to be modeled to derive signatures and markers of health.
Metabotype of Blood type:
Determine the relationship between blood type (which is inherited) – the glycoproteins on blood-- with the metabotype- and the onset and progression of disease may lead to the development of drug targets or nutritional strategies at the sub-population level.
A comprehensive, open-access reference catalog of blood cell types. Owing to its accessibility, blood has been the most studied human organ. However, in spite of decades of work, recent technological advances have revealed that the diversity of blood cell types is far greater than previously appreciated. This diversity encompasses not only distinct cell populations that were previously agglomerated into broader compartments defined by a limited number of cellular markers, but also rare cell populations such circulating tumor cells. A comprehensive catalog of blood cell types, together with accurate information on their prevalence in normal individuals, would be an invaluable reference resource for diverse investigations into both basic biology and blood alterations in disease.
Sud M, Fahy E, Cotter D, Azam K, Vadivelu I, Burant C, Edison A, Fiehn O, Higashi R, Nair KS, Sumner S, Subramaniam S.
Nucleic Acids Res. 2016 Jan 4;44(D1):D463-70. doi: 10.1093/nar/gkv1042.
Quehenberger O, Armando AM, Brown AH, Milne SB, Myers DS, Merrill AH, Bandyopadhyay S, Jones KN, Kelly S, Shaner RL, Sullards CM, Wang E, Murphy RC, Barkley RM, Leiker TJ, Raetz CR, Guan Z, Laird GM, Six DA, Russell DW, McDonald JG, Subramaniam S, Fahy E, Dennis EA.
J Lipid Res. 2010 Nov;51(11):3299-305. doi: 10.1194/jlr.M009449.
Gorden DL, Myers DS, Ivanova PT, Fahy E, Maurya MR, Gupta S, Min J, Spann NJ, McDonald JG, Kelly SL, Duan J, Sullards MC, Leiker TJ, Barkley RM, Quehenberger O, Armando AM, Milne SB, Mathews TP, Armstrong MD, Li C, Melvin WV, Clements RH, Washington MK, Mendonsa AM, Witztum JL, Guan Z, Glass CK, Murphy RC, Dennis EA, Merrill AH Jr, Russell DW, Subramaniam S, Brown HA.
J Lipid Res. 2015 Mar;56(3):722-36. doi: 10.1194/jlr.P056002.