The Genome Report

Sunday, August 16, 2009

INDELS, gene expression and hereditary cancer

INDELs (the insertion or deletion of nucleotides) comprise an emergent and increasingly important class of sequence variants in the human genome. While there is no fixed limit as to the number of nucleotides that can be inserted or deleted, the vast majority are around 1-4 nucleotides. Although initially, it was thought that both insertions and deletions arose through the same mechanisms, recent research has noted that motifs indicative of replication events were frequently found near insertions, while recombination motifs were more commonly found near deletions, which has now led to the conclusion that they arise independently.

INDELS are important since they play a pathological role in many human genetic diseases. In most cases that have been studied, INDELS occur in the open reading frames of mRNAs. This alteration can lead to termination of protein translation and loss of gene function. For example, out of frame insertions in the ribosomal gene RPS19 lead to Diamond-Blackfan Anemia Disease. In other diseases, the insertion of a triplet repeat preserves the frame, but leads to the reiteration of an amino acid. For example, in Kennedy’s disease the insertion of CAG repeats leads to a polyglutamine segment in the Androgen Receptor.

However, the recent comprehensive sequencing of the human genome has shown that INDELs are much more densely populated in the non-protein encoding regions of the human genome. Since these regions contain gene regulatory elements, this suggests that INDELs may aberrantly regulate gene expression. In this report, using a bioinformatics approach we have examined whether INDELs might regulate expression by creating novel targets for microRNAs. MicroRNAs are small RNAs consisting of twenty-one to twenty-two nucleotides. To date, around 800 microRNAs have been cloned and sequenced from the human genome. MicroRNAs down regulate gene expression by annealing to complementary sequence in the untranslated regions of mRNA. Thus, an INDEL could either provide critical nucleotides to support the binding of the microRNA or delete nucleotides that preclude the contiguous binding of the seed sequence.

We reasoned that if INDELs indeed could create microRNA target sites and down regulate expression, this might be readily apparent in genes whose lack of expression leads to cell proliferation. Thus, we examined the INDELs found in the untranslated regions of BCRA1 mRNA. BRCA1 mRNA encodes a human tumor suppressor protein whose loss of expression confers a significant risk of cancer. Using RNA: RNA interaction prediction software, we found an INDEL (out of a total of 4 reported INDELs) that was capable of significantly strengthening a microRNA target site. As shown in figure 1 the deletion of two residues (CU) creates a contiguous sequence fully complementary to the seed sequence of microRNA.

Although this concept needs to be functionally tested, it is interesting to reflect on its possible disease significance. Such a phenomena might explain the well known clinical heterogeneity observed in patients who test positive in the BCRA1 diagnostic test. Although the population who harbor a structural mutation in one allele of BCRA1 has a very high overall risk of breast and ovarian cancer, some patients get cancer whilst others don’t. The molecular mechanism that underlies this heterogeneity is not well understood.

Our observation that INDELs could create microRNA target sites suggests a plausible mechanism. BRCA1 positive patients that get cancer might do so because their “normal “ allele is in fact abnormal and contains an INDEL that can down regulate its expression. Thus, activation of the microRNA will lead to down regulation of BCRA1 and tumorigenesis. Importantly, the activation of the microRNA may itself be regulated by other cellular and environmental factors. Thus, this may provide an important link to environmental factors (hormones, carcinogens) that may promote tumorigenesis in the BRCA1 positive population. In sum, sequencing the noncoding regions of the ‘normal’ allele might provide answers to why some carriers of BRCA1 mutations get cancer, and others don’t.

Guest Contributor : Elizabeth Dietz

Tuesday, July 21, 2009

Soon, we will all be VOUS – The need for a cancer oriented ENCODE program

The BRCA1/BRCA2 test is now considered the standard of care for identifying breast and ovarian cancer risk. Indeed, women who have been tested for BRCA1/BRCA2 status must view the current discourse on the possible implications of personalized genetic medicine with incredulity. It is very clear that women who inherit a dysfunctional form of the BRCA1 or BRCA2 gene have a much higher risk of developing breast or ovarian cancers. A positive BRCA1/BRCA2 test triggers the individual, their genetic counselor and oncologist to explore drastic preventative measures, such as an oophrectomy or a mastectomy. Regardless of whether or not the patient chooses to undergo surgery, a positive test will prompt regular checkups and greatly increased scrutiny. However it is unfortunate that comparatively little attention has been paid to this area. For example, the correlation between cancer risk and the specific mutation (for example missense or nonsense) has not been clarified. This is likely because the number of individuals that harbor informative mutations are still a small percentage of the total number of breast and ovarian cancer patients. However, to the individual with a mutation, the relevance is immediate and of critical importance. For example, it has recently been discovered in the U.K. that women with mutations in BRCA2 specifically respond to therapy that inhibits the DNA damage response pathway.
But what if test findings tell an individual that they have a variant of unknown significance? Becoming increasingly more common among BRCA1/BRCA2 test results, variants of unknown significance (VOUS), or unclassified variants, give the patient no concrete sense for their cancer risk. For individuals that decide genetic testing is the right route to assess their cancer risk, possessing the knowledge that their BRCA1 gene has a significant mutation may be bad news, but is at least a concrete picture of their cancer risk. Receiving news of a variant of unknown significance that could be inconsequential or life threatening is quite devastating.
This problem is likely to be significantly compounded in the near future. At present, it is only women who have an obvious family history of cancer that are tested, However, it is clear that there is significant public interest in DNA sequence analysis and personalized medicine. The whole scale adoption of whole genome analysis will undoubtedly provide a deluge of new sequence variants. Soon, everyone will be VOUS! The majority of test results that come back will show presence of variants of unknown significance. Importantly most of these variants will reside in the untranslated regions of BCRA1/2 and will influence the expression of BRCA1/2 rather than the structure of the proteins. For these variants, conventional risk assessment will be complicated by the unknown factors (oncogenes, hormones, environmental factors) that will likely modulate the effects of these variants upon the expression of BRCA1/2. Clearly, what is needed is an ENCODE type program, but one that is devoted to finding functional assays that can independently signal the risk of breast and ovarian cancer.

Caroline Meade - Guest Contributor

Tuesday, July 14, 2009

How do mutations arise in RNA viruses?

As we understand more about the human genome, it becomes equally important to understand its relationship with the genomic environment. In the case of RNA viruses such as SARS, Hepatitis C or Influenza, an inappropriate interaction with a viral genome can result in death. The human genome defends itself by to engineering specific immunological factors (T cells, antibodies) that selectively target viral antigens. However, RNA viruses can rapidly mutate, escape immunological defenses and even become resistant to therapeutic drugs. A contemporary example is the ability of the Flu virus to mutate a single amino acid, rendering it resistant to Tamiflu. Most of our fundamental knowledge in genomics and molecular biology has accrued from the study of model virus systems. However, there is a remarkable lacuna in our understanding of the molecular mechanisms of RNA viral mutation. The current dogma is that such errors arise because of the lack of a proof reading function during replication. In short, a default concept. However, given that it is advantageous to mutate, one suspects that there are viral factors that actively catalyze mutagenesis. Perhaps the somatic hypermutation of the immunoglobulin genes is an attractive metaphor. In any event, much more attention should be devoted to investigating these mechanisms. It would be tragic if the current H1N1 pandemic further illuminates our lack of knowledge

Friday, July 10, 2009

Novel connections between the “old” protein world and the “ new” RNA world.

Amidst the increasing attention given to the economical sequencing of the human genome, one inconvenient truth remains largely ignored. Virtually all of biomedical science and pharmaceutical industry has focused on merely 2% of the genome, the segment of the genome that encodes proteins. Until recently the other 98% had been somewhat famously described as Junk. However, it is now clear that non-coding RNAs such as microRNA, esiRNAS, piRNAs that reside in the “junk”, have critical cellular functions.
It has been known for a long time that it is the noncoding segment of the genome that is most variable between human individuals. Thus, when the deluge of human sequence information arrives, the successful interpretation of these variants will be a challenge. However, noncoding RNA genes appear to exert their effect by their complementarity to other nucleic acid sequences and therefore their interactions are much more predicable than their protein counterparts
For example investigators have begun to explore the possibility that variants in human sequence may affect phenotype by modulating microRNA target sites in messenger RNA. Indeed, Dr Francis Collins, the new NIH director has recently written a timely and compelling review of this area.
Importantly, it is critical to realize that these human sequence variants are “plastic”. Their ability to increase the risk of disease is not “cast in stone’ but will depend upon the expression of the noncoding RNA. It is hoped that an understanding of the factors that regulate such noncoding RNAs will identify changes in life style that may prevent disease.

Sunday, June 28, 2009

A Call for a Grass Roots Genomics Movement

Dicer and Familial Lung Cancer

The recent report in Science that mutations in the Dicer gene underlie Familial Pleuropulmonary Blastoma has all the hallmarks of a classic scientific discovery. Perseverance, dedication, compassion and brilliant scientific insight combined to yield a spectacular advance in cancer genetics. Dicer is an enzyme that is essential to the expression of a new class of non-coding genes called microRNAs. These microRNAs are powerful pleotropic effectors capable of dramatically altering the growth and differentiation of human cells. The authors show that mutations in the Dicer gene lead to this pediatric lung cancer. Thus, this is another remarkable example of the burgeoning linkage between the new biology of the non-coding human genome and disease. Indeed, one suspects that mutations in Dicer may well be associated with other more common human cancers.

The Need for a Grass Roots Genomics Movement
However, one begins to suspect that, in the future, such advances will only be made by a “ grassroots genomics movement”. It is a sad fact that many families, like the one in the study published in Science, suffer in silence from the tragically predictable consequences of their hereditary disease. Unfortunately, in general, such families remain unknown to the general biomedical research community. Firstly this is an information problem. Unless, the family represents an index case, their history is unlikely to be described in the literature. Secondly, it is because the major funding agencies believe that the study of rare disease will not have a significant impact on human health. Indeed, most federal research grant applicants assert the significance of the proposed studies by indicating the prevalence of the disease. Compounding this problem is the reluctance of most research grant applicants to propose to do truly innovative research. This problem has been recently discussed in an excellent article in the NY Sunday Times. Perhaps the time has come for a “grass roots” family based genomics movement. Much of the personalized genome discussion has focused on the DNA sequence of an individual. Now is the time to focus on genomic studies of families. With the dramatic decrease in the cost of whole genome sequencing, families with hereditary disease can now take the initiative and sequence their genomes. The alternative is to painfully wait until their hereditary disease attracts the attention of biomedical research.
This movement should not be confined to families with rare hereditary disease. The factors that regulate the penetrance of mutations in genes that are already known to underlie common genetic diseases (for example the BRCA1 gene and cancer) are far from clear. Family initiated sequence analysis studies may identify such factors and thereby refine the diagnosis.

Can it be done?
Some might say that this might be a difficult agenda. Indeed, the sequencing of the first human genome required the talents of a generation of the finest molecular biologists and considerable federal funding. However, many of the barriers have now been broken. Collection of DNA no longer requires a skilled phlebotomist and can be achieved by a simple cheek swab. The family will not need to operate or bear the cost of running a research lab. There are now several genomic sequence vendors who provide a sequencing service for around $20,000 per sample and an insight into the relevance of any sequence variants. The cost of sequencing 20 family members (including those affected and unaffected) is therefore much less than a typical R01 federal research grant. Families could form an “institute without walls” and directly apply for federal research grants. Their ability to provide a hereditary disease database might provide compelling preliminary data. As is typical of many grant applications, the collaboration with an established medical geneticist will provide both skills in grant preparation and in the interpretation of the sequence data.

It’s good science
Although it may be difficult at present to interpret the genome of an individual, the presence of a putative disease variant within a family should be much more readily tractable. It is important to point out that successful gene identification studies on families typically require a much smaller N than that required for GWAS studies on more heterogeneous populations. It is very likely that the identification of disease variants will garner the interest of experts in the biology of that gene and accelerate the development of therapeutic agents.

Thus, in sum, this movement will not only enhance a family’s ability to make sound healthcare decisions but also advance science.

Pages