UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
UvA-DARE (Digital Academic Repository)
The volatile metabolome and microbiome in pulmonary and gastro-intestinal
disease
van der Schee, M.P.C.
Publication date
2015
Document Version
Final published version
Link to publication
Citation for published version (APA):
van der Schee, M. P. C. (2015). The volatile metabolome and microbiome in pulmonary and
gastro-intestinal disease.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
Chapter 1
Omics Research in Medicine
Marc P. van der Schee Departments of Respiratory Medicine, Pediatric Respiratory Medicine, Academic
Medical Center, University of Amsterdam, Amsterdam, The Netherlands Departments of Pediatric Pulmonology, Pediatric Gastroenterology, Gastroenterology
and Hepatology, VU University Medical Centre, Amsterdam, The Netherlands
Unpublished
Omics r
esear
Medical science generates knowledge through hypothesis driven research
The science of medicine defines disease as a dysfunction away from a state of
dynamic homeostasis or homeokinesis1. Consequently, detailed knowledge of
both normal bodily function and pathophysiological processes challenging this functioning is fundamental to developing interventions restoring homeostasis and with that health. Around 400 BC Hippocrates was the first to advocate a rational approach to medicine based on the concept of ὑπόθεσις or ‘hypothesis’ (to suppose) as suggested by Socrates. By abiding to this strict set of rules of hypothesis-driven research medical knowledge has grown exponentially resulting in its formal integration in daily clinical practice as evidence based medicine. Technological advancements in the twenty first century are currently pushing the boundaries of hypothesis driven research by providing researchers with tools to approach the science of medicine in an essentially different way through hypothesis generating research. This thesis focuses on examples of the ‘omics’ techniques that facilitate such an approach and explore their ability to contribute to an integrative systems biology approach to medical research.
The central dogma as a scaffold to describe health and disease
Paradoxically the foundation for hypothesis generating research lies in one of the pinnacles of hypothesis driven research; the description of the double helical shape of deoxyribose nucleic acid, or DNA, along with a plausible mechanism for the replication of the genetic material contained therein, as first described
in 1953 by Francis Crick and James D. Watson2. With that concept they laid the
foundation for the ‘central dogma’ as a model for the functioning of all living organisms in health and disease. The central dogma essentially describes the way in which DNA, our genotype, is transcribed into complex mixtures of RNA and subsequently translated into arrays of proteins functioning as effector molecules in our bodies. Through this elegant mechanism we can describe in what way our inherited genotype (nature) interacts with our environment (nurture) to develop into a certain phenotype. It is this dogma that is the current foundation used to describe pathophysiological processes underlying disease at the composite DNA, RNA, protein and metabolite levels.
Omics technologies enable hypothesis generating research
Recent advances in research techniques in molecular biology and chemical analytical chemistry have provided researchers with sophisticated tools that allow detailed qualification and quantification of the entirety of a biochemical molecular family (such as DNA, RNA or proteins) that are present in a single biological sample. Such research is different in its essence, because it is not primarily driven by a hypothesis other than; the obtained results contain relevant
information that allow discrimination of different phenotypes (for instance health and disease). As opposed to assessing a single or few targeted biomarkers, the molecular patterns captured by such ‘omics’-techniques may more accurately
represent the highly complex nature of health and disease3.
The implications of this can be visualized through an analogy with the Nazca desert in Peru; figure 1a depicts this desert from the ground and figure 1b from an airplane. This desert had already been studied for many years for remnants of the Nazca culture through hypothesis driven research. However it wasn’t until historian Paul Kosok flew across the area in 1940 that he was struck by the figures depicted in the desert. As you can appreciate from these pictures the monkey had never been noticed from the ground where all research to that date had been taking place. By analogy omics techniques may provide a birds eye overview of the landscape of, for instance, all the proteins in blood, allowing researchers to establish patterns and associations that would have been (partly) obscured with hypothesis driven research.
Figure 1. Visualization of omics approach
Such an approach firstly implies that omics technologies allow researchers to generate hypotheses about the complex biological mechanisms underlying disease phenotypes. This may help to guide hypothesis driven research to key factors driving pathophysiological processes. Second, this can help to generate novel, biology driven, phenotypes by creating clusters of subjects that have similar biomarker profiles (figure 2). These phenotypes can subsequently be associated with clinical traits such as: disease severity, progression, and treatment response. It is important to realize these biological phenotypes can help to differentiate health and disease in a probabilistic manor without requiring detailed knowledge of the exact mechanisms driving the changes in the underlying biomarkers. This requires a radically different way of approaching research results but is in fact
Figure 1. Nazca desert Peru from the ground (a) and from an airplane (b).
Omics r
esear
analogous to the way a physician uses his or her intuition to differentiate health from disease based on the patterns observed in the patients’ clinical presentation. The final implication of such an approach is that phenotypes based on underlying biological mechanisms need not necessarily match the currently established disease categorization that has been built over the last few centuries, mostly by classifying patients based on their clinical presentation. Omics technologies may very well point towards various disease mechanisms resulting in identical symptoms and vice versa may reveal identical disease processes underlying very distinct clinical presentations. Omics technologies therefore have the ability to progress medicine by providing personalized treatment modalities focused on restoring individual metabolic disturbances, rather than symptom suprression. Such personalized or p4 medicine is likely to significantly decrease side effects, reduce healthcare costs and improve therapeutic response if significant
challenges are overcome4.
Figure 2. Unsupervised multidimensional cluster analysis
Omics techniques span all biochemical families
Various omics techniques have been developed during the past years that span all possible biochemical families of the central dogma3. A short overview is provided here in order to place this thesis in the wider context of available techniques. Specific emphasis will be placed on the techniques applied in this thesis. The techniques are ordered according to the dominant flow of information through biological systems (figure 3). It is however important to realize that these domains cannot be considered to be separate entities; they are strongly intertwined and interact on multiple levels.
Figure 2. An example of a multidimensional cluster analysis by which subjects are grouped based on the similarity of their biological traits. Patients are represented by nodes, separate clusters are marked with different colors5.
Figure 3. Central ‘dogma’ of biology
Figure3. Graphical depiction of dominant flow of hereditary information through the human body and associated omics techniques.
Genomics
In genomics studies the entire genome of an individual is analyzed by sequencing its genetic code. In humans this was achieved for the first time by the human genome project initiated in 1990, which took an impressive 13 years to sequence
one entire genome6. This is in stark contrast with currently available next
generation sequencing methods that have a staggering throughput of roughly a
full genome a day7.
These techniques have enabled large Genome Wide Association Studies (GWAS) establishing many novel candidate disease genes through analysis of single-nucleotide polymorphisms. It was anticipated that common allelic variants would account for a substantial proportion of the population risk for those diseases constituting the major public health burden. Unfortunately, the reality is that despite great efforts, results to date do not live up to those expectations. Even in Crohn’s Disease, where strong associations with several loci have been
reported, the attributable risk that is explained by these genes is less than 10%8,9.
This suggests the majority of phenotypical differences between health and disease arise from sources downstream of the genetic code itself. Therefore many genomics studies have moved beyond case-control comparisons to include factors
such as environmental exposures10. The potential merits of such an approach are
illustrated by a study showing that common variants at the 17q21 locus were only associated with asthma in children who had both Rhinovirus induced wheezing
illness and expression of two genes at this locus11.
A specific type of genomics that has gained considerable attention over the past years is that of microbiome analysis. The human microbiome is the aggregate of the entire ecological community of microbes in a given body sample. The microbiome is composed of approximately 100 trillion microbial cells thereby outnumbering human cells 10 to 1. Primary sites of colonization with microbes are the skin,
Omics r
esear
gut and as recently discovered the (lower) airways12,13. Microbiome analysis is
performed by sequencing the hereditary information encoded in the gene for the
s16ribosomal RNA component of the small subunit of prokaryotic ribosomes14,
or by determining the length of the interspace regions of this gene15. This
region is core to cellular function and is therefore highly conserved throughout bacterial species providing a binding site for primers. The unique feature of this region however is that the 16s rRNA also contains regions hypervariable in both sequence and length that are species-specific and can be used for bacterial identification. Microbiome analysis can therefore provide a readout of all the
microbes contained in a specific sample (figure 4)16,17.
Similar to the human genome project, the human microbiome project sequenced over 5000 samples from upper airways, skin and feces. This resulted in the identification of over 10.000 microbial species that occupy humans as their
ecosystem12. In view of these numbers it is not surprising the microbiome has a
significant effect on human physiology. The microbiome can for instance provide essential metabolic functions that humans lack, supporting the hypothesis that
the microbiome has likely co-evolved with humans18.
Figure 4. Example readout of human microbiome
Our mutualism with the microbiome is a delicate balance; the immune systems controls and selects microbiota and likewise the microbiome exerts influence upon the immune system. This is supported by experimental data showing that germ free mice have an underdeveloped immune system and are more likely
to develop allergic airway disease19. Furthermore specific microbiota appear to
drive the differentiation of Th17 cells, an immune cells that has been implicated
in many immunity related diseases20. Finally, cross-sectional studies have
shown that difference in microbiome composition are associated with various
200 300 400 500 600 700 800 900 1000 1100 30000 25000 20000 15000 10000 5000 0 Fragment length (nc) Relative Abundance (RFU )
Figure 4. An example of a readout from the nasopharyngeal microbiome by ISpro. Every separate peak represents a microbial species with colors indicating the phylum (bacterial family) the strain belongs to. The height of the peak represents their abundance (see chapter 9).
diseases such as inflammatory bowel disease21, diabetes22 and even asthma13.
In case of asthma, disease development may even be driven by the gastro-intestinal microbiome which affects the immune system in such a way that it
predisposes to allergic airway inflammation23. Furthermore, gut microbiota
may induce an unfavorable response to airway microbial infections, inducing
allergic sensitization and respiratory disease24. Some of these effects appear
to be reversible in childhood upon introduction of a ‘healthy’ gut microbiome
but persist if adults are recolonized19. This points towards a clear window of
opportunity in early life during which microbial exposures affect the developing immune system, suggesting that especially studies of the microbiome in
childhood are important in view of possible disease modifying interventions25,26
(Chapter 9). Since the composition of the microbiome is highly individual and site
dependent12, the research challenge will be to correlate these complex patterns
longitudinally with relevant health outcomes.
Transcriptomics
DNA is transcribed into RNA depending on the epigenetic phenomena driving the configuration of DNA. The transcriptome is the entire set of RNA-molecules including messenger RNA, ribosomal RNA, transfer RNA, etcetera present in a specific biological sample. With the recent advent of next-generation RNA-sequencing (RNAseq) scientist have a high-throughput technology that enables quantification of changes in RNA in health, disease and in response to
interventions down to the single cell27. This clearly differentiates transcriptomics
from genomics for it allows a detailed overview of cellular activity of distinct cell populations rather than an overview of the full genetic code which is not cell specific. Because messenger RNA functions as a precursor for protein translation this holds important information pertaining to the phenotype. Therefore the transcriptome is the first ‘level’ (figure 3) at which both genetic predisposition and environmental exposures are integrated. This was illustrated by a study in asthma which investigated the effect of oral glucocorticoids on the transcriptomic profile of airway smooth muscle cells. This study revealed that steroids not only have anti-inflammatory properties but also improve airway hyper-responsiveness
by affecting gene expression of the airway smooth muscle layer28, which may
help to advance asthma therapy. Furthermore, transcriptomics analysis of non-inflamed mucosal gut biopsies in patients with inflammatory bowel disease identified down regulation of specific gene transcripts that may predispose to gut
inflammation, providing a possible screening tool29.
It is important to realize these results do not imply that the transcriptome translates directly to effector functions of the encoded proteins. Processes such as alternative splicing and post-translational modifications help to determine
Omics r
esear
the resulting protein products. Furthermore, as single messenger RNA can be transcribed multiple times. Consequently there is no strict correlation between the RNA and protein concentration. The proteome thus contains information regarding cell functioning that can’t be deducted directly from genomics or transcriptomics data.
Proteomics
Proteins are polypeptide effector molecules that are both building blocks for our cells and messenger molecules that orchestrate the functioning of cells, tissues and organisms. This means they are core to defining the phenotype of a cell, tissue and organism; making them the primary target of the majority of drugs to date. Analysis of these compounds can therefore yield key insights into the composition, regulation and functioning of cellular pathways and structural
complexes30. In cystic fibrosis a proteomics approach revealed that chaperone
proteins were responsible for the disrupted folding of the membrane chloride channel that underlies the disease. This initiated subsequent hypothesis driven research which revealed these chaperone molecules could be triggered to fold the chloride channel correctly thereby potentially providing a novel treatment
option31.
The major challenge of proteomics relative to transcriptomics is its lack of a detailed high throughput technique such as RNAseq. For current proteomic studies Liquid Chromatography and Mass-Spectrometry (LC-MS) is the dominant technique. Unfortunately these technologies identify only a fraction of all proteins
in a single experiment and still lack reliable quantitative measures32. Similar
to analysis of RNA, these experiments are further complicated by the dynamic nature of expression of the proteins creating considerable challenges to reliably
stabilize the samples32. Fortunately, efforts to build a library of the full human
proteome are underway33. In lack of a true omics technique such initiatives will
prove to be invaluable to ascertain a full understanding of molecular pathways by combining a multitude of experiments along with complementary molecular techniques. Such evidence can, in the future, help to identify target proteins for therapeutic interventions. To date however, the most global picture of cellular activity in a single experiment is still obtained through transcriptomics.
Metabolomics
Proteins, lipids, amino acids, drugs, food, etcetera all take part in the metabolic machinery that keeps our body running. The metabolome is the aggregate of the small molecules (<1500 Dalton) that form the raw materials of these metabolic
reactions and its resulting products34. Small changes in gene-expression or protein
them valuable candidate biomarkers. This is in fact well known; blood cholesterol and glucose are tell-tale signs of cardiac risk and diabetes, respectively.
Metabolomics studies are generally performed by chemical analytical techniques such as Gas or Liquid Chromatography coupled to Mass Spectrometry (LC-MS /
GC-MS) or Nuclear Magnetic Resonance (NMR) spectroscopy34. As with proteomics
these techniques are challenged by the fact that a single experiment does not allow characterization of the full spectrum of available molecules. Therefore, these experiments are preferentially performed with an a priori defined focus with respect to the target metabolites. This partly undermines the unbiased nature of the omics approach. More so than with every other omics technique discussed, metabolites can have extremely short half-life times making strict sample handling and a detailed analysis protocols essential.
A differentiator of metabolomics relative to (blood based) proteomics is the fact that the metabolome more readily contain markers originating from extracorporeal sources such as drugs and diet thereby enabling studies of bodily exposures
such as cigarette smoke35. Furthermore, metabolites are ultimately excreted by
our bodies with our urine, feces, skin or breath. This makes metabolomics an attractive technique from the patient’s perspective as the samples can be accessed non-invasively.
This is especially true for the analysis of volatile organic compounds (VOCs) which are the gaseous carbon based end products of metabolic processes such as lipid
peroxidation and glucose metabolism36. These VOCs can be analyzed from many
bodily samples such as urine, feces and breath. Examples of such approaches can be found in this thesis.
Thesis Aims
As can be inferred from the discussion above omics techniques have potential as diagnostic tools, for phenotyping, for monitoring and for the prediction of disease prognosis. In this these I aim to study the value of microbiomics and volatile metabolomics to address these clinical challenges, together with my co-workers. We aim to study this in infectious, neoplastic and chronic inflammatory conditions for both pulmonary and gastro-intestinal diseases. These studies serve to provide insight into the merits and pitfalls of metabolomics and microbiome studies in medicine helping to guide future experiments.
Technical validation
As with any novel technique, proper technical validation is core to assessing its value for clinical practice. The most predominantly used omics technique in
Omics r
esear
this thesis is electronic nose analysis of VOCs. This device uses promiscuous gas sensors that show cross-reactive interactions with the target volatiles (chapter 11). The resulting outcomes (VOC profile) can be used in pattern recognition algorithms but does not identify individual biomarkers. This poses specific challenges on the design of validation experiments. Exposing the sensor array to (a set of) specific compounds may show adequate reproducibility for that compound and/or chemical group. This does however not imply that the same behavior will be found for different classes of volatile metabolites. Validation of repeatability and reproducibility is therefore best assessed in the actual clinical study samples.
In a previous study on exhaled breath samples we have shown adequate reproducibility of VOC profiles in asthmatic patients upon repeating the full
breath collection and analysis manouvre39. In this study we furthermore
determined that the use of an inspiratory VOC-filter, for removal of ambient VOCs, improved the reliability of the measurement. Because such findings are highly application specific we will repeat reproducibility and validation experiments for novel applications in this thesis, namely exhaled VOC analysis in children (Chapter 8), fecal VOC analysis (Chapter 7) and offline VOC analysis (Chapter 10).
Disease diagnosis
Chapter 2, Aspergillosis: : Firstly we aim to test the diagnostic potential
of exhaled breath metabolomics in infectious disease. We will study this in patients with prolonged chemo-therapy induced neutropenia who are at risk of the development of invasive pulmonary aspergillosis infection. This severe and life threatening infection will serve as a model to assess whether profiling of exhaled biomarkers can help to identify active infection.
Chapter 3, Mesothelioma: To assess the diagnostic potential of exhaled breath
metabolomics in cancer in an intention to diagnose population, we aim to study patients with long term asbestos exposure and patients with malignant pleural mesothelioma. This study will serve as a model to determine whether exhaled biomarkers are affected by 1. long-term asbestos exposure and 2. the presence of mesothelioma.
Chapter 4, Lung Cancer: We aim to determine to what extend local and systemically
produced metabolites contribute to the volatile metabolome associated with lung tumors. To this end we will develop a technique to sample alveolar air during bronchoscopy, both at the tumor site and at the healthy contra-lateral side. This study aims to determine the relative contributions of local and systemic VOCs to volatile tumor biomarkers and thereby provide an indication whether diagnosis
of extra-pulmonary tumors by breath based metabolites may be feasible.
Chapter 5, Asthma: We aim to assess the diagnostic potential of exhaled metabolites for inflammatory disease. To this end we aim to compare the diagnostic potential of exhaled volatiles in asthma to that of the fraction of exhaled nitric
oxide FENO and sputum eosinophils, both during and without treatment. The
evidence generated in this study will help to determine the potential of exhaled breath metabolomics relative to other diagnostic tests and help to assess whether steroid therapy affects the assessment of inflammatory diseases.
Disease phenotyping and monitoring
Chapter 5, Asthma steroid responsiveness and disease activity: We aim to determine whether exhaled metabolomics can help to predict therapy response. To model this we aim to assess the exhaled biomarkers of patients prior to installment of a course of oral prednisone and compare those patients with and without a therapeutic response. As a second aim, we wish to assess whether the exhaled metabolome reflects disease activity and has value for disease monitoring. We will study this by correlating sputum eosinophils in asthma, as a marker of disease activity, with the exhaled metabolic profiles in the same patients. Furthermore, we aim to assess whether patients who experience loss of control (increase of symptoms) after withdrawal of therapy differ from those who remain stable. This study may provide insight whether monitoring of disease activity by means of the exhaled metabolome is feasible.
Chapter 6, Mucociliary clearance disease: We aim to assess the potential of exhaled metabolomics to differentiate between two clinically similar diseases that have distinct underlying pathophysiological processes; the mucociliary clearance diseases of cystic fibrosis and primary ciliary dyskinesia. This study can help to determine whether exhaled metabolites reflect the differences in the underlying disease process. To further elucidate the effect of disease activity we will also compare patients with and without an active exacerbation.
Chapter 7, Inflammatory bowel disease: Similar to chapter 6, we aim to assess the potential of the fecal volatile metabolome to differentiate between two clinically similar gastro-intestinal diseases with distinct pathophysiological presentations; Crohn’s disease and Ulcerative Colitis. We aim to do so both during active disease at first presentation and after induction of remission in order to assess whether disease activity influences the fecal metabolites. This will help to provide evidence whether the fecal volatile metabolome holds valuable information for disease diagnosis, phenotyping and monitoring.
Omics r
esear
Disease prognosis
Chapter 8 & 9, Detection of early signs of asthma: We aim to study whether omics techniques allow detection of early signs of asthma in childhood. To this end we will build a prospective community based cohort in which we will attempt to diagnose childhood asthma earlier than currently possible. As a first step, we will compare children with and without high-risk phenotypes for the development of asthma by means of:
Exhaled metabolomics (Chapter 8) Microbiomics (Chapter 9)
These studies aim to provide knowledge whether the chronic inflammation
that is likely present in pre-school children at high-risk to develop asthma37 can
be detected by omics techniques at an age where symptoms do not allow such
differentiation38.
Outlook and overview
Chapter 10, Multi-center volatile metabolomics studies: In order to initiate future multi-centre volatile biomarker studies we aim to assess the effects of storage and transport on volatile metabolite profiles.
Chapter 11, Exhaled metabolomics in pulmonary medicine: We aim to review the current evidence with respect to the value of analysis of exhaled metabolites in diagnosing, phenotyping and monitoring in pulmonary medicine.
Chapter 12, Volatile metabolomics in colorectal cancer: : We aim to review the evidence pertaining to the diagnosis of colorectal cancer by means of metabolomics analysis of both exhaled and fecal volatiles.
Chapter 13, General discussion: : I will assess whether the objectives set out
in this thesis have been met. I will discuss the merits and pitfalls of these omics techniques in context of these studies and will provide an outlook what potential ramifications omics techniques may have on medicine
References
1. Macklem, P. T. & Seely, A. Towards a definition of life. Perspect. Biol. Med. 53, 330–40 (2010).
2. Watson, J. & Crick, F. Molecular structure of nucleic acids. Nature 171, 737–738 (1953).
3. Wheelock, C. E. et al. Application of ’omics technologies to biomarker discovery in inflammatory lung diseases. Eur Respir J 42, 1–45 (2013).
4. Hood, L., Balling, R. & Auffray, C. Revolutionizing medicine in the 21(st) century through systems approaches. Biotechnol. journal, Spec. issue Syst. Biol. Pers. Med. 7, 992–1001 (2012).
5. Clark, L. Data-Visualization Firm’s New Software Autonomously Finds
Abstract Connections. at <http://www.wired.com/2013/01/data-viz-ayasdi-iris/>
6. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860–921 (2001).
7. Mardis, E. R. Next-generation DNA sequencing methods. Annu. Rev. Genomics Hum. Genet. 9, 387–402 (2008).
8. Weiss, K. M. & Terwilliger, J. D. How many diseases does it take to map a gene with SNPs? Nat. Genet. 26, 151–7 (2000).
9. Franke, A. et al. Genome-wide meta-analysis increases to 71 the number of confirmed Crohn’s disease susceptibility loci. Nat. Genet. 42, 1118–1125 (2010).
10. Lockett, G. a & Holloway, J. W. Genome-wide association studies in asthma; perhaps, the end of the beginning. Curr. Opin. Allergy Clin. Immunol. 13, 463–9 (2013).
11. Calışkan, M. et al. Rhinovirus Wheezing Illness and Genetic Risk of Childhood-Onset Asthma. New Engl J Med 368, 1398–407 (2013).
12. The Human Microbiome Project Consortium. Structure, function and
diversity of the healthy human microbiome. Nature 486, 207–14 (2012). 13. Hilty, M. et al. Disordered microbial communities in asthmatic airways. PLoS
One 5, e8578 (2010).
14. Kuczynski, J. et al. Experimental and analytical tools for studying the human microbiome. Nat. Rev. Genet. 13, 47–58 (2012).
15. Budding, a E. et al. IS-pro: high-throughput molecular fingerprinting of the intestinal microbiota. FASEB J. 24, 4556–64 (2010).
16. Woese, C. R. & Fox, G. E. Phylogenetic structure of the prokaryotic domain: the primary kingdoms. Proc. Natl. Acad. Sci. U. S. A. 74, 5088–5090 (1977). 17. Weisburg, W. G., Barns, S. M., Pelletier, D. A. & Lane, D. J. 16S ribosomal DNA
amplification for phylogenetic study. J. Bacteriol. 173, 697–703 (1991).
18. Gill, S. R. et al. Metagenomic analysis of the human distal gut microbiome. Science 312, 1355–1359 (2006).
Omics r
esear
19. Olszak, T. et al. Microbial exposure during early life has persistent effects on natural killer T cell function. Science 336, 489–93 (2012).
20. Ivanov, I. I. et al. Specific Microbiota Direct the Differentiation of IL-17-Producing T-Helper Cells in the Mucosa of the Small Intestine. Cell Host Microbe 4, 337–349 (2008).
21. Major, G. & Spiller, R. Irritable bowel syndrome, inflammatory bowel disease and the microbiome. Curr. Opin. Endocrinol. Diabetes. Obes. 21, 15–21 (2014). 22. Giongo, A. et al. Toward defining the autoimmune microbiome for type 1
diabetes. ISME J. 5, 82–91 (2011).
23. Trompette, A. et al. Gut microbiota metabolism of dietary fiber influences allergic airway disease and hematopoiesis. Nat. Med. 20, 159–66 (2014). 24. Huang, Y. J. & Boushey, H. a. The microbiome and asthma. Ann. Am. Thorac.
Soc. 11 Suppl 1, S48–51 (2014).
25. Bisgaard, H., Bønnelykke, K. & Stokholm, J. Immune-mediated diseases and microbial exposure in early life. Clin. Exp. Allergy 44, 475–81 (2014).
26. Gollwitzer, E. S. & Marsland, B. J. Microbiota abnormalities in inflammatory airway diseases: Potential for therapy. Pharmacol. Ther. 141, 32–39 (2014).
27. Shalek, A. K. et al. Single-cell transcriptomics reveals bimodality in
expression and splicing in immune cells. Nature 498, 236–40 (2013).
28. Yick, C. Y. et al. Glucocorticoid-induced changes in gene expression of airway smooth muscle in patients with asthma. Am. J. Respir. Crit. Care Med. 187, 1076–84 (2013).
29. Smith, P. J. et al. Mucosal Transcriptomics Implicates Under Expression of BRINP3 in the Pathogenesis of Ulcerative Colitis. Inflamm. Bowel Dis. 0, 1–11 (2014).
30. Cravatt, B. F., Simon, G. M. & Yates, J. R. The biological impact of mass-spectrometry-based proteomics. Nature 450, 991–1000 (2007).
31. Wang, X. et al. Hsp90 Cochaperone Aha1 Downregulation Rescues Misfolding of CFTR in Cystic Fibrosis. Cell 127, 803–815 (2006).
32. Mirza, S. P. & Olivier, M. Methods and approaches for the comprehensive characterization and quantification of cellular proteomes using mass spectrometry. Physiol. Genomics 33, 3–11 (2008).
33. Wilhelm, M. et al. Mass-spectrometry-based draft of the human proteome. Nature 509, 582–7 (2014).
34. Wishart, D. S. et al. HMDB: the Human Metabolome Database. Nucleic Acids Res. 35, D521–6 (2007).
35. Ahotupa, M., Bussacchini-griot, V., Camus, A. & Cedex, L. Rapid oxidative stress induced by N-Nitrosamines. Biochem. Biophys. Res. Commun. 146, 1047–1054 (1987).
36. Buszewski, B., Kesy, M., Ligor, T. & Amann, A. Human exhaled air analytics : biomarkers of diseases. Biomed Chrom 21, 553–566 (2007).
37. Saglani, S. et al. Early Detection of Airway Wall Remodeling and Eosinophilic Inflammation in Preschool Wheezers. Am J Respir Crit Care Med 176, 858–864 (2007).
38. Fouzas, S. & Brand, P. L. P. Predicting persistence of asthma in preschool wheezers: crystal balls or muddy waters? Ped Resp Rev 14, 46–52 (2013). 39. Dragonieri S, Schot R, Mertens BJ et al. An electronic nose in the discrimination
of patients with asthma and controls. J Allergy Clin Immunol 2007; 120:856–62.
Omics r
esear
