|
Epigenetics and Beyond
ISIS Report 19/01/09
From Genomics to Epigenomics
Decades of sequencing and dissecting the human genome have confirmed that
the real causes of ill health are environmental and social
It is not the genetic messages encoded in genomic DNA but environmentally-induced
epigenetic modifications that overwhelmingly determine people’s health and well-being
Dr. Mae-Wan Ho
A fully referenced version of this article is posted on ISIS members’ website.
Details here
An electronic version of the full report can be downloaded from the ISIS
online store. Download Now
The Human Genome Project failed to deliver
Some of us had predicted that the US$ 3 billion project to sequence the human
and other genomes would fail to deliver its extravagant promises [1, 2] (Genetic
Engineering Dream or Nightmare, ISIS publication; Human Genome -The Biggest Sellout.in
Human History, ISIS Report); and we were right [3] (Why Genomics Won't Deliver, SiS
26).
The Human Genome Project was followed by HapMap, a public-private
research consortium dedicated to finding genetic variants that predispose
people to common illnesses such as cancer, Alzheimer’s and cardiovascular
disease. HapMap was launched in Washington in 2002 [4], involving scientists
and funding agencies from Japan, the UK, Canada, China, Nigeria, and the US.
It would cost US$100million and take three years to complete. Francis Collins,
who headed the Human Genome Project, and now Director of the US National Human
Genome Research Institute (NHGRI), said: “The HapMap will provide a powerful
tool to help us take the next quantum leap toward understanding the fundamental
contribution that genes make to common illnesses like cancer, diabetes and
mental illnesses.” Companies like Affymetrix and Illumina developed powerful
gene chips for scanning the human genome. Medical statisticians designed the
genome-wide association study, a robust method for discovering ‘true’ disease
genes and avoid the many false positives that have dogged the field [5].
In 2006, Elias Zerhouni, director of the US National Institutes
for Health predicted that: “comprehensive, genomics-based health care will
become the norm, with individualized preventive medicine and early detection
of illnesses [6]. A year later, AmpliChip announced the new era of pharmacogenomics
worldwide, as its test for cytochrome P450 genes can help drug providers prescribe
selective serotonin reuptake inhibitors in the treatment of adults with depression
[7]. The era of “genomics medicine” has arrived [8]; or has it?
The 1000 Genomes Project to the rescue
The reality behind the hype is something else. The lack of progress is such
that in January 2008, the 1000 Genomes Project was announced [8]; its aim
was to sequence at least 1 000 individual human genomes, and to look, again,
for genetic susceptibilities to common diseases “at a resolution unmatched
by current resources.” Some of the HapMap organisations have committed major
support to the new project: Beijing Genomics Institute in Shenzhen, China,
the Wellcome Trust Sanger Institute in Cambridge, UK, and the NHGRI. Three
US sequencing companies joined the consortium in June 2008: 454 Life Sciences,
a Roche company in Branford, Conn, Applied Biosystems, an Applera Corp business
in Foster City, California, and Illumina Inc., in San Diego, California.
The genomes of any two humans are more than 99 percent identical.
It is hoped that the small fraction of genetic material that varies among
people holds valuable clues to individual differences in disease susceptibility,
response to drugs and sensitivity to environmental factors.
The 1000 Genomes Project is to build upon the HapMap comprehensive
catalogue of human genetic variation organized into blocks called haplotypes.
The HapMap catalogue laid the foundation for the recent [8] “explosion of
genome-wide association studies that have identified more than 130 genetic
variants linked to a wide range of common diseases, including type 2 diabetes,
coronary artery disease, prostate and breast cancers, rheumatoid arthritis,
inflammatory bowel disease and a number of mental illnesses.”
However, the HapMap catalogue only identifies genetic variants
present at a frequency of 5 percent or greater, while the 1000 Genomes Project
catalogue will map many more details of the human genome and identify variants
present at a frequency of 1 percent across most of the genome, and down to
0.5 percent or lower within genes. Francis Collins said it is like building
bigger telescopes; “the results of the 1000 Genomes Project will give us greater
resolution as we view our own genetic blueprint. We’ll be able to see more
clearly than before and that will be important for understanding the genetic
contributions to health and illness.” The project is estimated to cost around
$60 million.
By June 2008, the 1000 Genomes Project has generated such vast
quantities of data that the information is taxing the current capacity of
public research databases. But information is not knowledge; genome sequences
are telling us next to nothing on disease susceptibilities.
The “genomics medicine” that never was nor will be
By September 2008, David B. Goldstein at Duke University, a leading young
population geneticist known partly for his research into the genetic origins
of the Jews, said the effort to pin down disease susceptibility genes is not
working.
There is absolutely no question that for the whole hope of personalized
medicine, the news has been just about as bleak as it could be,” he told the
New York Times [5]. The HapMap and other techniques developed to make
sense of the human genome was a “tour de force”, but has produced only a handful
of genes accounting for very little in explaining genetic predisposition to
diseases: for schizophrenia and bipolar disorder, almost nothing, for type
2 diabetes, 20 variants that explain only 2 to 3 percent of familial clustering,
and so on.
The reason for this disappointing outcome, in his view, is that
natural selection has been far more efficient at eliminating disease-causing
variants than people thought, so these variants are rare. It takes large,
expensive studies with hundreds of patients in different countries to find
even common disease variants, so rare variants are simply beyond reach.
It’s an astounding thing,” said Goldstein, “that we have cracked
open the human genome and can look at the entire complement of common genetic
variants, and what do we find? Almost nothing. That is absolutely beyond belief.”
Goldstein is not alone in this bleak assessment of genomics.
Concern has been raised for several years over commercially available gene
tests offered to consumers, especially ‘predictive genomic profiling’ testing
for variants in different combinations of genes for risks to illnesses such
as lung cancer, type 2 diabetes or cardiovascular disease that are supposed
to give people personalised nutrition and other life-style health recommendations.
Recently, researchers at Erasmus MC University Medical Center
Rotterdam in The Netherlands critically appraised these genomic profiling
now offered online by at least seven companies testing for variants in 56
genes. For 24 of the genes, there were no available studies to show that the
profiling was useful in the general population. Of the remaining, only variants
in 25 genes showed significant associations with risks in 28 diseases, but
the associations were generally modest, and many of associations were with
diseases unrelated to the condition for which the profiling was intended [10].
These weak associations most certainly do not mean that people carrying
‘high’ risk variants will definitely develop the disease, nor do they give licence
to those carrying ‘low’ risk variants to adopt unhealthy lifestyles with impunity.
As one critic commented [11], the genetic information provided by such direct
to consumer genomics is “nearly all, to varying degrees, inaccurate, misleading
or merely useless.”
The real reasons genomics profiling fail, however, is not due to lack of data,
or that natural selection is so effective in eliminating deleterious variants.
It is the genomics project itself that is misguided.
Genetics to epigenetics
Critical voices had been raised against the genomics projects from within
the scientific establishment since 2003; and soon afterwards, it became clear
why genome sequences could tell us little about disease susceptibility, and
much less, how to make designer babies. That’s basically because the genome
is fluid and dynamic, and impossible to pin down; the actions are predominantly
in the ‘hidden’ parts of the genome that don’t code for proteins, especially
in epigenetic processes in response to the environment [3].
Even the conventional gene sequences that constitute only 1.5
percent of the genome are far from simple, as revealed by the findings of
project ENCODE (Encyclopedia of DNA elements) organised by the NHGRI, and
published in July 2007 [12]. ENCODE involved a consortium of 35 research groups
that went through 1 percent of the human genome with a fine-tooth comb to
find out exactly how genes work, and came up with some major surprises.
As Barry Patrick wrote of the ENCODE findings in Science News
[13]: “genes are proving to be fragmented, intertwined with other genes, and
scattered across the whole genome.”
Indeed, within the human and other mammalian genomes, coding
sequences are in bits (exons) separated by non-coding introns; and exons contributing
to a single protein could be in different parts of the genome. Coding sequences
of different proteins frequently overlap. Regulatory signals are similarly
scattered upstream, downstream, within the coding sequence or in some other
distant part of the genome [14] (see GM is Dangerous and Futile,
SiS 40). The potential repertoire of proteins that can be made by combining
different exons is perhaps a million times larger than the official number
of about 20 000 genes identified in the human genome. Which exons are recruited
to make specific proteins depends entirely on the environmental contexts.
Genome DNA sequences therefore really determine very little;
it is our individual environmental experiences that overwhelmingly shape our
own health as well as the health of our offspring, and possibly, our offspring’s
offspring.
Epigenetic inheritance not due to genomic DNA
The new discipline of epigenetics is the study of inheritance ‘outside’
genetics, i.e., not due to the DNA of the genome. This definition is the best
I can think of that covers all examples to-date described in this series [15]
(Epigenetics and
Beyond series, SiS 41). It reveals how distinctly different proteins
are assembled from separate exons; how specific genes are marked to be expressed
or not, according to environmental context, how messages transcribed are altered,
and even recoded in the genome; all of which were unthinkable to most people
just a few years ago. These findings violate fundamental tenets of heredity,
i.e., genetic determinism, that have dominated biology for a hundred years:
the firm belief that the environment can never directly affect the genes, and
characters acquired during one’s life time cannot be inherited.
Epigenetics has put an end to genetic determinism; but by no mean supports
environmental determinism. The hallmark of epigenetic inheritance is its dynamism
and plasticity. Although the environmental epigenetic influence persists for
varying periods of time, and can be transmitted across generations, it can also
be reversed, or changed further by altering the environment in an appropriate
way [16] (see Caring
Mothers Strike Fatal Blow against Genetic Determinism, SiS 41).
Epigenetics is spawning its own databases to top all databases
Faced with the ever-expanding
molecular complexities discovered soon after the human genome was sequenced,
‘systems biology’ was invented in academic institutions to curate a
series of ‘bio-informatics’ databases all ending in ‘omics’: ‘transcriptomics’,
for all RNA transcripts, the vast majority not coding for protein; ‘proteomics’,
for all proteins translated; and ‘metabolomics’ for all metabolites made by
chemical reactions in the body; in the vain hope that the true meaning of
life will emerge from the data deluge [17] (No
System in Systems Biology, SiS 21).
Now, more than five years later, ‘epigenomics’ is
jostling for its own databases to top all databases. The Human Epigenome Pilot
Project has begun with a consortium that includes the Wellcome Trust Sanger
Institute in the UK, Epigenomics AG, a transatlantic biotech company with
headquarter in Berlin, Germany, and the Centre National de Génotypage, a French
Government research institute [18]. It aims to identify DNA methylation variable
positions (MVPs) in the human genome. DNA methylation is only one among dozens
of epigenetic mechanisms that alters the gene expression states or genetic
messages in cells and tissues. The epigenome is an ever-changing, ever-evolving
entity; there being potentially as many epigenomes as cells or tissues within
a single organism, depending on the micro-environmental contexts. Indeed,
monozygotic (genetically identical) twins do not always show the same disease
susceptibilities; and it has been reported that young twins have similar DNA
methylation whereas older twins differ considerably in both the amounts and
patterns of DNA methylation [19], and most likely in other epigenetic markers
acquired during the different individual experiences of each twin.
The epigenome of MVPs alone is unlikely to predict
disease susceptibility as promised. At best, epigenomics may suggest appropriate
environmental interventions for individuals after extensive and costly ‘epigenomic
typing’; and it is not clear that is feasible. Even if the capacity of current
public databases could be expanded to accommodate these colossal catalogues,
and all existing scientists were to be deployed in annotating and servicing
the databases, it might still take more time than the age of the universe
to search through them all.
Epigenetics confirms that the causes of ill health are overwhelmingly environmental
and social and must be addressed by appropriate policies
Epigenetics is a new and exciting discipline,
but the last thing it calls for is yet more mind-numbing cataloguing exercises
and databases.
It has long been recognized that in stark contrast to the subtle effects of
susceptibility genes, environmental effects swamp out even large genetic differences
[1].
For example, toxic agents in the environment were found to scramble genome
sequences to produce new transcripts linked to a range of chronic illnesses
such Gulf War Syndrome, chronic fatigue syndrome, autoimmune diseases and leukaemia
[20] (see Health and the Fluid Genome
series, SiS 19). A new subdiscipline of Epigenetic
Toxicology [21] (SiS 41) has emerged in recognition that toxic agents
can have heritable epigenetic effects not only on individuals exposed, but also
on their offspring
I can do no better than
repeat my earlier warning that preoccupation with genomics and other ‘bio-informatics’
databases can only distract us from addressing the real causes of ill health
[1-3], which are predominantly environmental and social, as all the findings
in the new discipline of epigenetics are telling us in no uncertain terms.
To keep our genome, and much more so, our epigenome, healthy, we need a balanced
ecosystem free from pollutants, we need to move away from industrial monoculture
to a biodiverse, sustainable agriculture [22] (Food Futures Now *Organic *Sustainable
*Fossil Fuel Free , ISIS publication). Sustainable agriculture free from
chemical inputs and consumed locally is the only way to overcome both macronutrient
and micronutrient deficiencies that compromise our physical and mental health,
and our natural immunity against infectious diseases [23] (Unraveling AIDS, ISIS publication)
We also need social policies that guarantee equal opportunities for all, and
prevent the environmental deprivations we now know to have devastating epigenetic
effects across several generations [16, 21].
| |
|
