|
Kresimir
Pavelic, Marijeta Kralj, Sandra Kraljevic, Mirela Sedic
Kresimir Pavelic, MD, PhD
Head of the Division of Molecular Medicine
Rudjer Boskovic Institute
Bijenicka cesta 54
POB 180
HR-10002 Zagreb
Croatia
 Download as a
PDF here
Functional genomics (transcriptomics and proteomics) is a
global, systematic and comprehensive approach to identification and
description of the processes and pathways involved in the normal
and abnormal physiological states. The most applied methods of
functional genomics today are DNA microarrays and proteomics
methods, primarily two-dimensional gel electrophoresis coupled with
mass spectrometry. Up to date interesting research has been carried
out, representing the milestones for future implementation of
functional genomics/proteomics in biomedicine. Still, further
systematic examination of differentially regulated genes and
proteins in tissues and fluids in healthy vs. diseased subjects
will be required. However, high-throughput technologies reflect
biological fluctuations and methodological errors. Large amount of
such different data challenges the performance and capacity of
statistical tools and softwares available at the moment. Yet,
further major developments in this field are pending and the
intellectual investment will certainly result in clinical
advances.
3.1 Introduction to
functional genomics
As an emerging discipline, functional genomics has many
different definitions, which depend on the research area.
Functional genomics (transcriptomics and proteomics) is a global,
systematic and comprehensive approach to identification and
description of the processes and pathways involved in the normal
and abnormal states. Why is it such an important experimental
approach nowadays? It is estimated that approximately 30% of the
open reading frames in a fully sequenced organism have unknown
function at the biochemical level and are unrelated to any known
gene. This is why, recently, the interest of researchers has
shifted from genome mapping and sequencing to determination of
genome function by using the functional genomics approach (Figure
1).
Figure 1. A single gene can give rise to multiple gene products.
RNA can be alternatively spliced or edited to form mature mRNA.
Besides, proteins are regulated by additional mechanisms such as
posttranslational modifications, compartmentalization and
proteolysis. Finally, biological function is determined by the
complexity of these processes.
Techniques of functional genomics include methods for gene
expression, profiling at the transcript level (differential
display, expressed sequence tags, serial analysis of gene
expression and DNA microarrays), as well as methods for proteome
analysis. Due to recent technological advances in fabrication of
very precise high-throughput instruments, a complex functional
genomics approach has become possible. Processing a large quantity
of experimental data requires powerful information systems �
bioinformatics, which encompasses development of new computational
methods and application of those methods to solve biological
problems. Bioinformatics has also a large service component in
which computational resources such as databases are operated for
the benefit of the research community. It also finds its
implications in other aspects of biomedical research involved in
functional genomics, such as laboratory information management
systems, medical record systems, as well as documentation of
clinical trial results for regulatory agencies.
All those high-throughput experimental procedures have enabled
and accelerated new and important basic discoveries, especially in
the field of molecular medicine. The ultimate aim is, however, to
bring these discoveries closer to clinics and therefore
substantially to improve daily clinical practice.
The benefits of the functional (integrative) genomics approach
and applications to biomedicine are likely to be:
a) the understanding of physiological or pathophysiological
processes,
b) identification of genes/proteins which can be used for
screening, diagnosis or monitoring disease severity and
c) the discovery of novel genes/proteins suitable for therapeutic
manipulation.
The most applied methods of functional genomics today are DNA
microarrays and proteomics methods, primarily two-dimensional
electrophoresis (2-DE) coupled with mass spectrometry (MS).
3.1.1 Introduction to
DNA arrays
DNA microarrays may be defined as miniaturized, systematic
immobilization of nucleic acid fragments derived from individual
genes on a solid support, which by specific hybridization, enables
the simultaneous analysis of thousands of genes in parallel.
Microarray technique is based on new technology in which DNA probes
are robotically spotted on miniature slides. The length of probes
varies from kilobases, for standard cDNA arrays used for RNA
expression analysis, to tens of basis in oligonucleotide arrays for
both RNA expression and DNA sequence analysis.
Nylon DNA arrays, often termed macroarrays are nylon-membrane
based cDNA arrays for broad-scale expression profiling. They
continue to be used because of their accessibility and flexibility.
This is due not only to somewhat expensive use of microarrays but
also to the fact that nylon DNA arrays offer a promising
alternative in a number of situations, particularly when small
amounts of sample are available or when specific pathways are being
studied, which is itself significant for a number of research and
clinical applications.
There are several different implementations of the DNA
microarray principle for expression measurement. Microarrays are
the most promising in terms of throughput and should eventually
allow simultaneous measurement of expression of the whole set of
genes. DNA microarrays already have had various applications:
messenger RNA expression profiling for improved disease
classification; genotyping of polymorphisms affecting disease
susceptibility; identification of genetic lesions within
malignancies; design and discovery of new therapeutic drugs; and
sequencing of DNA.
3.1.2 DNA arrays in
transcriptomics studies
DNA microarrays are the preferred and widely accepted method for
transcriptomic research as most disease processes are accompanied,
not only by characteristic macroscopic or histological changes, but
also by systematic changes in gene expression patterns. For some
pathological processes such as cancer, inappropriate gene
expression is a fundamental aspect of pathogenesis. For other
pathological processes, the gene expression programmes, both in
cells directly affected by a disease and in healthy cells
responding to the local and systemic effects of a disease, can
provide a detailed molecular picture of the pathogenic process.
When used for documentation of gene expression at a genome-wide
scale, microarray-based transcriptional profiling allows the
identification of a set of genes that defines differential
biological states (Figure 2). This is a crucial step in the
development of novel approaches for complete diagnosis of a
disease. Molecular classification of a disease combined with the
ability to place a certain patient into a specific genetic subtype,
holds a promise of a better comprehension of a disease and thus
development of individualized health care delivery. Microarray
analysis can help in prediction of disease outcomes and
prognosis.
Figure 2. Comparative gene expression analysis by using DNA
microarray.
Microarrays are also being applied increasingly to mutation
analysis (polymorphisms analysis) by minisequencing. Single
nucleotide polymorphisms (SNPs) are the commonest source of
mutation in man and can be used as markers in whole genome-linkage
analysis of families or in association studies of individuals in a
population. Single base differences between human genomes, or
polymorphisms underlie differences in susceptibility to or
protection from a host of diseases. Scientists believe SNP maps
will help them identify the multiple genes associated with complex
diseases such as cancer, diabetes, vascular disease, and some forms
of mental illness. These relationships are difficult to establish
with conventional gene-hunting methods because a single altered
gene may make only a small contribution to the disease.
For instance, when applied to pregnancy research, the DNA array
technique is extremely powerful. In that regard, interesting papers
were published on newly discovered cDNAs/proteins which play a role
in various stages of placentation mostly using mouse models and
purified cell-culture systems.
3.2 Proteomics
The term proteome was first introduced to describe the entire
PROTEin complement of a given genOME, i.e. the complete set of
proteins, which are expressed by the entire genome. Since the
proteome is quite dynamic and it changes along with the development
of an organism and with any alterations in the environment, it can
be referred to as the array of proteins expressed in a biological
compartment, such as cell, tissue or organ, under particular
environmental circumstances at a particular time. Proteomics is a
powerful tool for examining differential protein expression
comparing hundreds of proteins simultaneously.
The majority of our cells contain the same genome regardless of
the cell type, stage of the development, or environmental
conditions, unlike the proteome, that varies significantly under
these diverse circumstances due to different patterns of gene
expression and protein modification. There are more proteins in the
proteome in comparison with the genes in a genome; and it has been
estimated that the human proteome is at least an order of magnitude
more complex than the human genome, since it is assessed that there
might be as many as a million human proteins. For example, in spite
of the same genome, the same growth factor may have diverse
signaling pathways in different cell types or cell states, implying
that the study of the genome exclusively could not clarify the
molecular mechanisms of diseases, ageing and cell response to
external stimuli.
It has been postulated that the average number of protein forms
per gene is one or two in bacteria, three in yeast and three or
more in humans. Proteomics deals with the characterization,
identification and quantitative analysis of proteins in cells,
tissues and body fluids. It is particularly suitable for the
analysis of some body fluids, such as serum and urine that are
deprived of mRNA, and thus cannot be studied by mRNA analysis.
Proteomic studies also include the analyses of the variations in
the protein expression levels in cells or tissues under two
different conditions and identification of the main proteins likely
to have important functional roles in the response of the cells to
those conditions, for example, healthy versus diseased tissue or
cells. A protein found only in a diseased sample may represent a
useful drug target or diagnostic marker. These proteomic
comparative approaches are analogous to microarray experiments,
which examine whether genes are turned on or off under diverse
conditions.
Furthermore, proteomics studies the numerous possible
interactions among proteins as well as the molecular composition of
the particular cellular structures (organelles). Knowing exactly
which proteins interact with one another could help to determine,
for example, components of a particular enzymatic pathway.
Protein modifications not obvious from the DNA sequence, such as
isoforms and post-translational modifications (e.g. phosphorylation
and glycosylation) can be determined solely by the proteomic
studies. It is assessed that approximately 200 diverse types of
post-translational protein modifications occur.
One aspect of proteomics studies is directed towards determining
the subcellular localization of proteins in order to construct an
overall three-dimensional protein map of the cell, which could give
an insight into the regulation of protein function.
Taken together, the scope of proteomics is quite broad and its
practical application goes far beyond merely analyzing large
numbers of proteins in complex mixtures.
Figure 3. Practical approach to proteome research.
3.2.1 Short overview of proteomics methods
The method of polymerase chain reaction (PCR) amplification was
probably the largest advance in genome studies, which enabled many
applications for genome study to date. However, in protein-based
studies it may not be possible to exploit such a tool. For that
reason and because protein chemistry and enzymology differ
substantially from those of the DNA, proteomics is expected to be a
more complex task than genome sequencing. To meet this challenge,
many new and different technologies for the whole proteome
characterization must be applied synergistically.
A detailed overview of proteomics techniques could be found
elsewhere. In general, proteome studies include the following
experimental stages: isolation and separation of proteins from a
cell line, tissue or organism; characterization and identification
of the particular protein species; and the information storage in
databases (Figure 3). Two-dimensional polyacrylamide gel
electrophoresis is the most widely used method for the separation
of proteins in proteomics, which allows simultaneous analysis of
hundreds to thousands of gene products. The first dimension,
isoelectric focusing (IEF), is an electrophoretic method that
separates proteins according to their isoelectric points (pI),
which are defined as the specific pH at which the protein net
charge is zero. The second, perpendicular dimension, SDS-PAGE
(sodium dodecyl sulfate-polyacrylamide gel electrophoresis),
resolves polypeptides according to their molecular weights. The
ultimate result of two-dimensional gel electrophoresis (2DGE) is a
gel with spots corresponding to individual proteins. The most
critical step in 2-DGE is the sample preparation, which depends on
whether the aim of the study is to examine the complete protein
profile of the cell or only proteins present in the particular
cellular compartments (such as membranes, subnuclear structures,
small organelles and vesicles).
Subsequent to 2-DGE, the protein spots to be identified are cut
out of the gel and the proteins digested into shorter peptides by a
protease, usually trypsin. The peptide fragments are then analyzed
by mass spectrometer for the purpose of identification. The
resulting peptide masses make a so-called fingerprint, which is
characteristic for the particular protein. This fingerprint is
juxtaposed to theoretically expected peptide masses for each
protein entry in the database. Mass spectrometry measures the mass
of unknown molecules by ionizing, separating and detecting ions
according to their mass-to-charge ratios and consists of three main
components: an ion source, a mass-selective analyzer and an ion
detector. High-throughput protein identification from 2-DGE gels is
dominated by the use of matrix assisted laser desorption/ionization
method of sample ionization and time-of-flight mass analyzer
(MALDI-TOF).
If the peptide mass fingerprint fails to identify the protein,
it is possible to determine a small piece of amino-acid sequence
from one or more of the peptides by fragmentation of the peptide
ions in the mass spectrometer that uses quadrupole, orthogonal,
time-of-flight, mass analyzer (Q-TOF). The analysis of these new
masses provides some partial amino acid sequence from the peptides,
which is normally sufficient clearly to identify a protein.
Together with all broadly accepted standard proteomics methods,
a high-throughput, analytical tool for rapid identification of
differences between large numbers of samples has currently being
developed � protein arrays. It would be extremely powerful to
analyze hundreds or thousands of protein samples using a single
protein array and thus acquire large amounts of data. However,
making protein arrays is far more difficult than making DNA arrays.
This is mainly due to the fact that protein function depends on its
precise three-dimensional structure. Out of the narrow range of
environmental conditions, proteins denature. For protein arrays, it
is thus critical to choose the right capture agent on the basis of
its specificity and affinity for the target protein. The capture
agent is immobilized on the array surface and, when exposed to
samples containing complex mixture of proteins, it binds the target
protein. The proteins that remain bound to the capture agents on
the surface are usually detected by fluorescence and identified by
mass spectrometry. Protein arrays will possibly be used as
sensitive, throughput and robust technology only if they will
further mature and be constantly applied to authentic biological
samples. This technology is very useful in linking gene-array
expression data with protein discovery and, unlike its gene-based
counterpart, can be used to examine post-translational modification
of proteins.
3.2.2 Functional
genomics in practice: is it all that perfect?
In spite of constant technological improvements that are being
made in the field of functional genomics, there are still many
technical and statistical obstacles hampering their usage in
routine clinical practice.
When it comes to microarray technology, data analysis has been
criticized for inter-, and sometimes intra-laboratory variability.
Microarray experiments are quite complex and the data have proved
to be noisy, susceptible to systematic errors, dependent upon
sample heterogeneity (e.g. tissue) and easily affected by technical
problems (e.g. sampling error, DNA/RNA isolation method, variation
in protocols and handling).
When planning experiments with microarrays, some critical points
should be taken into account starting from standardization of all
methods and protocols, through the experimental design, and ending
with careful analysis and validation of the data. Before starting a
microarray analysis, the reported measurements should be normalized
or modified to make them comparable. The goal of normalization is
to adjust for effects that are due to variations in the technology
rather than the biology. Methods of interpretation of large sets of
microarray biologic data are still being developed. Questions arise
as to which method is considered the �right one� because of the
variety of the possible outcomes. However, patterns of gene
expression revealed by data analysis are just the beginning. In
many cases, greater biological understanding can be attained by
using expression data in conjunction with sequence data, pathway,
and biomedical text sources. The limiting steps in performing
microarray experiments are hence not only sample handling or the
analysis itself but also determination of what the obtained results
actually mean which depends, as previously mentioned, on the smart
use of bioinformatics tools that allow integrated analysis of
multiple data types (mRNA levels coupled with proteomic analysis)
resulting in the improvement of the identification of the clinical
endpoints biomarkers.
However, in comparison with microarray technology, proteomics
encounters some other specific problems. In particular, rather
small fractions out of the total number of proteins expressed by
eukaryotic cell can be routinely separated on 2-DGE gel. The reason
for this might be that some proteins simply fail to get into the
gel due to poor solubilization (hydrophobic membrane proteins,
nuclear proteins and proteins that tend to aggregate, e.g. tubulin
and keratins) or molecular weight size (very large and very small
proteins), others are not resolved by the pH gradient (basic and
acidic proteins), or even limitations in the sensitivity of the gel
staining method (disability to detect low-abundance proteins, in
particular those playing an important role in the cell cycle,
signal transduction and receptors).
Sample preparation is considered as the bottleneck of 2-DGE in
terms of quality and protein distribution, and its success lies in
the efficient extraction and solubilization of proteins of
interest. Unfortunately, there is no universal protocol adequate
for all proteins, so that each sample (cell culture, tissue or body
fluid) represents new challenge with respect to sample preparation.
An important issue in this respect is the removal of
non-proteinaceous particles that might interfere with 2-DGE (such
as salts, lipids, nucleic acids, polysaccharides) as well as of
abundant proteins (e.g. albumin in human serum or fibrinogen in
plasma) that might hide some low-copy proteins of biological
significance.
Another matter to be considered is the artifactual modifications
of proteins that might occur during 2-DGE, i.e. proteolysis that
occurs upon liberation of proteases subsequent to cell disruption
and carbamylation, caused by the degradation of urea used in sample
solutions to cyanate, which in turn reacts with the amine groups on
proteins and changes their net charge thus affecting IEF
separation.
Owing to the variability from gel to gel, resulting in the
discrepancy in the observed amounts and number of the individual
protein spots, some spots observed in one gel might not be
displayed in another gel of the same sample run in parallel.
Consequently, it is often difficult to come to a firm conclusion on
whether the particular spot on one 2-DGE gel actually matches the
same protein spot on a different gel. Also, biological variations
amongst samples render it difficult to establish normal protein
expression profiles that can be compared with the diseased
states.
Finally, 2-DGE is quite a manual technique and does not seem to
be easily adapted to automation for the purpose of high-throughput
analyses. Nevertheless, it proves to be a method of choice for
protein separation due to its excellent resolving power as well as
the ability to separate different protein isoforms, and we believe
it will continue to play an important role in future proteomics
research.
3.3 Conclusion
The sequencing of the human genome has provided the first look
at all genes. The next steps however require powerful
transcriptomics tools in order to verify and refine the predicted,
incomplete gene models as well as define new models. Combined with
complete coding sequences, protein sequences can be examined for
motifs, domains, and biochemical characteristics. Proteomics is
complementary to transcriptomics and it can be used to confirm the
existence of an individual gene thus serving to figure out the
total number of genes in a particular genome, so called �functional
annotation� of a genome. Future implementation of functional
genomics/proteomics in biomedicine will require a systematic
examination of differentially regulated genes and proteins in
tissues and fluids in healthy versus diseased subjects. However,
high-throughput technologies reflect biological fluctuations and
methodological errors. Large amount of such different data
challenges the performance and capacity of statistical tools and
softwares available at moment. Further major developments in this
field are pending and the intellectual investment will certainly
result in clinical advances.
References
- Bogusky MS, MW McIntosh: Biomedical informatics for proteomics.
Nature 2003; 422: 233
- Celis JE, M Kruhoffer, I Gromova, C Frederiksen, M Ostergaard,
T Thykjaer, P Gromov, J Yu, H Palsdottir, N Magnusson, TF Orntoft:
Gene expression profiling: monitoring transcription and translation
products using DNA microarrays and proteomics. FEBS Lett 480 (2000)
2
- Diehn M, AA Alizadeh, PO Brown: Examining the living genome in
health and disease with DNA microarrays. JAMA 283 (2000) 2298
- Dudda-Subramanya R, G Lucchese, D Kanduc, AA Sinha: Clinical
applications of DNA microarrays analysis. J Exp Ther Oncol 3 (2003)
297
- Garfin DE: Two-dimensional gel electrophoresis: an overview.
Trends Anal Chem 22 (2003) 263
- Gevaert K, J Vandekerckhove: Protein identification methods in
proteomics. Electrophoresis 21 (2000) 1145
- Godovac-Zimmermann J, LR Brown: Perspectives for mass
spectrometry and functional proteomics. Mass Spectrom Rev 20 (2001)
1
- Graves PR, TAJ Haystead: Molecular biologist's guide to
proteomics. Microbiol Mol Biol Rev 66 (2002) 39
- Kraljevic S, PJ Stambrook, K Pavelic: Accelerating drug
discovery. EMBO Reports 5 (2004) 837
- Lockhart DJ, EA Winzeler: Genomics, gene expression and DNA
arrays. Nature 405 (2000) 827
|