Transcript Nutritional Genomics and human health

Use of metabonomics in
human health
Inbal Dangoor
Introduction
• Since the completion of the human genome sequencing
project, the main goals of functional genomics have
been to determine the function of the products of
newly identified genes, as well as to determine those
that might be therapeutically targeted.
• Combined effort of different ‘omics’ approaches.
• Traditional medicine measure metabolites because of
their fundamental regulatory importance in
biochemical pathways.
• However, metabonomics leave behind the reductionist
method of investigating single component effect on a
biological system for a more holistic approach of
exploring the molecular details of multiple factors on
an entire biological organism.
• Our knowledge on various cellular
conditions during disease situations can be
translated to promote human health
through the use of metabolomics.
• For instance, Tumor cells possess the
potential for proliferation, differentiation,
cell cycle arrest, and apoptosis.
There is a specific metabolic phenotype
associated with each of these processes
that is characterized by the production of
energy and special substrates necessary
for the cells to function in that particular
state.
The future (hopefully…)
• Personalized medicine throughout the life time
of each individual.
• Protect health and prevent disease by
prediction of future problems, rather than
solely diagnosing and reversing existing
disease.
• The main tools: identify biomarkers and
evaluate metabolic disorders early enough.
• The single biomarker approach to disease
diagnosis (e.g., cholesterol for potential heart
disease) will be replaced by direct assessment
of health and, most importantly, assessment
will be comprehensive.
Definitions
• Metabolomics: “the complete set of metabolites/lowmolecular-weight intermediates, which are context
dependent, varying according to the physiology,
developmental or pathological state of the cell,
tissue, organ or organism”.
• Metabonomics: “the quantitative measurement of the
multivariate metabolic responses of multicellular
systems to pathophysiological stimuli or genetic
modification”.
Applications and achievements of
metabonomics in health so far
• Diagnosis and classification of diseases (tumor types)
• Disease state (time course progression)
• Learn pathological mechanisms and identify new
biomarkers
• Responses to treatment (efficiency, toxicity), and drug
design (decrease development time)
• Generating (partial) databases: HumanCyc,NIST, tumor
metabolome database, METLIN (human biofluid),
Interpret NMR database (metabolic profiles of
tumors).
Metabolic biomarkers of tumors
Griffin et al., Nature reviews 4, 2004.
The major metabolic disorders screened for in newborns using MS
Want et al., ChemBioChem 6, 2005.
Advantages of metabonomics
in human health
•
•
•
•
•
Cheap (1$ apiece ??)
Fast
High-throughput
Fully automated
Ideal for following rapidly changing
phenotype
• Minimally invasive screening
• Broad view of a patient status
Methods
The biological material:
-Blood serum
-Blood plasma
-Urine
-Tissue extracts or biopsy - invasive
-CSF (cerebrospinal fluid) - invasive
-Milk
-Saliva
Techniques:
-NMR (very common due to many pioneering works)
-GCMS, LCMS (increased sensitivity)
-HRMAS 1H NMR (High resolution magic angle spinning, intact tissues)
-MRS (magnetic resonance spectroscopy, in tissues within living organisms)
-FT-IR
-TLC
-Metabolite array
Data mining:
Pattern recognition software is needed to associate specific profiles with
different cell types, tumor types or a stage of treatment.
Example 1 - Diagnosis
• Screening human populations for coronary
artery disease, a common metabolic disorder,
using NMR of serum, instead of expensive and
invasive techniques such as angiography.
• NMR spectra of individuals who have been
extensively characterized by angiography and
algorithms based on conventional risk factors.
This allowed the direct comparison of pattern
recognition techniques with standard means.
Discrimination between healthy and sick samples
healthy
sick
Pattern recognition model was generated using multivariate
classification, PLS-DA (Partial least squares-discriminant
analysis)
Without (d) or with (f) OSC (orthogonal signal correction, used
to optimize the separation by subtracting variation)
The spectral regions that contribute to the discriminations
were lipids (LDL, HDL, VLDL) and choline.
The derived model predicted correctly 92% of unknown
samples.
Brindle et al., nature medicine 8, 2002.
Comparison of patients with different severity of coronary atherosclerosis
Brindle et al., nature medicine 8, 2002.
Example 2 - metabolic profiles of
cancer cells
• HIF-1 (hypoxia inducible factor) is a heterodimeric
transcription factor that is upregulated in several
cancer types. It cause an increase in glycolysis rate,
growth factor production and angiogenesis induction.
• Studying the effect of HIF-1 deficiency on tumor
metabolism and growth using.
• Observed reduced rates of growth in HIF-1 deficiency
cells.
• Results: increase of phosphorylated cell membrane
constituents and reduction of ATP content because of
reduced rate of glycolysis.
• This is a good example for the use of metabolomics in
studying metabolic pathways that could be targeted
therapeutically.
Catabolism of phospholipid membrane components is
elevated in vivo in deficient Hepa c4 tumors
PDE
phosphodiester
PDE/Pi
Higher ratio
Griffiths and Stubbs, Advan Enzyme Regul 43, 2003.
Alteration in metabolic pathways in
HIF- deficient tumors
choline
glycolysis
3-phosphoglycerate
glycine
lipids membranes
PDE/Pi
Nucleotide synthesis
ATP
betaine
Example 3 - NMR spectra at different time points
after gene therapy induced apoptosis in tumors
• Gene therapy approach was carried
in order to induce apoptosis.
• Apoptosis in glioma cells might be
specifically associated with an
increase in the amount of
unsaturated lipids.
• Investigate changes in lipid
biochemistry: distinguishes the
chemical groups present — for
example, the identity and amount of
specific lipids.
• Lipids that are released as part of
the apoptotic process can be
monitored and provide non-invasive
way to monitor effects of cancer
treatment in patients in vivo.
Polyunsaturated lipids
Griffin et al., Nature reviews 4, 2004
Example 4 - personalized
medicine
•
•
•
Administration of galactosamine hydrochloride to a group of ten rats.
Using NMR they observed liver effects that were very variable, enough
that the rats could be classified as either ‘responders’ or ‘nonresponders’.
PCA analysis showed some discrimination between responder and nonresponder groups in terms of their pre-dose metabolite profiles.
PCA (predose)
Andrew et al., Nature letters 440, April 2006.
Information on individual responses to xenobiotics might be
contained in the metabolite patterns of pre-dose biofluids
•Mole ratio of paracetamol glucuronide to paracetamol (G/P) was found to be
the most convincingly predicted of the various postdose metabolite
quantities.
•PLS model showed positive correlation (r = 0.48) between G/P and the
integral of the 5.06–5.14 region of the pre-dose NMR spectra, and negative
correlations (r = -0.56 and -0.54) between G/P and the integrals of the
8.98–9.10 and 0.50–0.86 regions of the pre-dose spectra.
•The predictive model for G/P is statistically significant.
NMR spectra
PLS model
R<0
R>0
R<0
predose
postdose
Andrew et al., Nature letters 440, April 2006.
Metabolomics in human nutrition
• Nutrients are traditionally known as chemical
substances that are consumed from food and needed
by the body for growth, maintenance, and repair of
tissue.
• Essential nutrients are not formed metabolically
within the cell and must be present in food that is
ingested, whereas nonessential nutrients can be
synthesized by the cell.
• Nutrients can modulate many pathways, for example:
DNA repair, altering carcinogen activation,
angiogenesis and signal transduction.
• They can cause diseases or prevent them!
• Diseases of modern civilization, such as diabetes,
heart disease, and cancer, are known to be influenced
from dietary patterns.
• Experimental evidence indicates that dietary
constituents, particularly phytochemicals and some
minerals and vitamins, can modulate the complex
multistep, multistage carcinogenesis process at the
initiation, promotion, and progression phases of
neoplasia.
• Examples: flavons (heart disease), stannols
(cholesterol metabolism), soy-based estrogen
analogues (cancer).
• The study of global metabolic profiles is only
starting, thus current reports are mostly theoretical.
Cancer prevention using appropriate
diet in children
Abnormal cell growth
Liang et al., American Society for Nutritional Sciences, 2003
Developing a protocol for metabolite extraction
from human serum for LC-MS studies
Deproteinization
• Some metabolites are noncovalently bound to
proteins, in addition, proteins can dominate the LC
analysis and cause signal suppression of less
abundant metabolites. Therefore, the applicability
of an extraction method may be determined by
the number of metabolites recovered as well as
the efficiency of protein removal.
• Deproteinization can be achieved using organic
solvents such as methanol and acetonitrile,
lowering pH with acid, or denaturation using heat.
• Strong acids adversely affect chemically unstable
compounds, and too much salt in samples can impair
the MS analysis.
Organic solvents were the most efficient and reproducible
for both metabolite recovery and protein precipitation
6 kD protein
•
•
•
LC-MS chromatograms of
serum samples obtained by
different extraction methods.
Avg. % RSD represents
average RSD value (relative
standard deviation) from six
replicates for all detected
features within each method.
RSD is an average of several
thousand metabolite features,
and individual metabolite RSD
values can vary greatly within
each method.
Organic
solvents
Missing
metabolites
Heat
Acid
Want et al., Anal Chem 78, 2006.
Number of Reproducible Features
A reproducible Feature is defined as a metabolite feature that was
detected in at least five out of six LC-MS runs for a given method.
MeOH
Heat+Acid
Want et al., Anal Chem 78, 2006.
Current limitations of metabonomics
in health studies
• Variations between samples due to extraction
methods, diet, sex, stresses, age, tissue type, natural
microflora and infections, genetic background etc…
Therefore it is not possible to add isotope-labeled
internal standard for all detected compounds, which
would increase the accuracy and reproducibility of
the analysis.
• The number of metabolites that can be detected.
High concentrated metabolites dominant the spectra
(need more sensitivity).
• Lack of comprehensive databases (in progress).
Technical improvements
• Improved sensitivity by combined use of NMR and LC (less coresonant peaks, detect low concentrated compounds).
• 13C NMR spectroscopy, which has high resolution but low
sensitivity, will be usable due to Cryoprobes (4oK). Cooling the
NMR probe will allow acquisition of the spectra, and improved
sensitivity.
• Metabolite array: 96-well plate assay system for phenotyping
(different assay mixtures).
• Increased automation will allow the rapid generation of
databases to assist in patient screening.
• Increased use of HRMAS 1H NMR spectroscopy: ideal for
following key metabolic events in the cytosol, and can be applied
directly to the tissue.
• Improved statistical tools.
• Protocols optimization from specific biofluid for precise goals.
Future directions
• Design different clinical experiments using whole-plant food
extracts and high throughput assays to determine the
mechanistic health benefits derived from fruits and
vegetables.
• Genetically modified food that will be designed to treat
specific disease (bonus: overcome the population fear from
transgenes…)
• Optimal: personalized medicine that will identify for each
person his health condition and predict future complications.
• If an individual is assessed before the development of disease,
analyzes must be able to predict the likelihood of future
diseases within the context of an individual’s overall, leading to
recommendation of appropriate means to avoid deleterious
health directions, with minimal side effects.
Conclusions
• Metabolic analysis is a promising approach to identify
biomarkers that could be used in the non-invasive monitoring
of the human health, by detecting early stages of illness and
applying the right treatment.
• The advantage of metabonomics: get a global view of multiple
effectors that are related to specific physiological status.
• Protect health and prevent disease by prediction of future
problems, rather than solely diagnosing and reversing
existing disease.
• Personalized medicine throughout the life time of each
individual, that will eventually improve the health of the
whole population.
THANK YOU FOR YOUR PATIENCE
AND TAKE CARE OF YOUR
HEALTH!!!