WHITE PAPER: NATURE, NURTURE, OR BOTH?

October 2010

By Zhenya Senyak

Eileen Wiggins – Out of the blue

Eileen Wiggins is a practical woman, straightforward, no nonsense. So when she started feeling tired most all the time and losing a little weight – not that she couldn’t afford to drop a few pounds here and there but for her age she was doing a lot better than some others she knew – she headed straight for the downtown clinic for testing.

The way Eileen Higgins remembers it, her first reaction when she was told she had polycythemia vera was shock. PV? What is it and why does she have it? Her questions quickly piled up: medication, therapies, life expectancy… but mostly she just plain couldn’t’t believe it. While listening to her doctor review the diagnosis and treatment options, in the back of her mind that single question nagged at her. Where on earth did this come from?

It’s not only a good question, but the answer might well provide a cure for the mysterious bone marrow neoplasms clustered together as MPDs. In all its forms – polycythemia vera, essential thrombocythemia, myelofibrosis, leukemias – the MPDs are a clonal, genetic disease affecting the pluripotent hematopoietic stem cells in the bone marrow. The result is over-production of one or more blood lines — red blood cells, platelets, or leukocytes – with attendant clinical issues.

The cause is elusive. By far the greatest number of patients, perhaps as many as 93%, have no known family history of an MPD. The cause of MPD in these sporadic patients could be somatic response to environmental events, radiation or medication, chemicals or maybe a simple error in the normal course of DNA transcription. Among the familial MPD patients, those who have other members of their first degree family with one or another MPD, the connection is so far not much clearer. There is an assumed germline transmission of the MPD since it is presumed to be inherited but perhaps what’s inherited is not the MPD itself but simply the genetic preference to produce a mutation that would contribute to the onset of the disease.

Worldwide, teams of scientists, researchers and clinicians are working to answer that very question. The answer lies somewhere in the billions of transactions that take place each minute within the body’s elaborate system of cellular birth and death scripted by the DNA within the nucleus of each cell. .

The search for the cause and cure of MPDs is a compelling medical detective story. 

For more than a half-century, molecular biologists and hematologists have pursued the cause and cure of myeloproliferative disorders, occasionally coming so close that some announced imminent victory. It’s a story filled with clues, false leads, high tech equipment, high speed chases of candidate genes and exhausting interrogations of the human genome. There have been a few heroes and many casualties along the way. At stake is not only the cure for MPDs but the chance to take a major step forward in understanding the genetic mechanisms of cancer itself.

The opening credits. Although investigators early in the 20th Century began to consider the proliferating blood diseases as closely related, it was William Dameshek in 1951 who first coined the term MPD and described the idea of the related trilineal blood disorders, as “myeloproliferative syndromes, including PV, ET, MF and erythroleukemia.”.(His paper appeared in Blood which he co-founded.)

A breakthrough. A major development that would lend urgency to the sporadic vs. familial debate was the discovery in 1960 by Peter Nowell and David Hungerford of the Philadelphia Chromosome, a chromosomal abnormality found in chronic myeloid leukemia (CML) patients, It would take another 13 years and the development of new staining and chromosome banding techniques before Janet Rowley would conclude that the Philadelphia chromosome was the result of a reciprocal translocation between chromosomes 9 and 22, thus setting the stage for its later description as an oncogenic BCR–ABL mutation.

In 1996 Brian Druker discovered imatinib, a compound that binds to a receptor site in the BCR-ABL tyrosine kinase effectively shutting it down. This was the silver bullet that changed chronic myeloid leukemia, one of the most dreaded blood diseases, from a death sentence to a manageable chronic illness.

So far, however, these discoveries did not establish a family connection. The genetic mutations known to cause an MPD were considered somatic, acquired chromosomal events that were not transmitted via the germline to succeeding generations.

Help from the labs. The same year Brian Druker published his findings in the New England Journal of Medicine (2001) with Nicholas Lydon and Charles Sawyer, the Human Genome Project (HGP) announced its sequencing of all 3 billion base pairs in the human genome was 90% complete. With this convergence of biotechnology the pace of discovery was about to pick up.

The difficulty at first would seem insurmountable. Our chromosomes live a shadowy life, undifferentiated in a sea of nucleic acid until the cell divides. The histones then release their iron grip on the coiled DNA that unfurls to be read, transcribed and reproduced. Only then, in 22 pairs, plus an X and Y do the chromosomes emerge, from the soupy sea of nucleic acid and carry on the work of reproduction, bearing the history of all our ancestors, male and female back to the time there was neither male nor female.

In each nucleated cell a strand of DNA, nearly 6 feet long, is folded and compressed with the code for our complete genome. The code is so complete that biologists could conceivably lift that cell and recreate a perfect clone of who we are…complete with our MPD. For somewhere along that double intertwined line of crackling, rotating, reproducing ladder of nucleotides lies the code that creates, stimulates, or predisposes the proliferation of one or more lines of our hematopoietic cells.

The line-up. With completion of the HGP, theoretically all we need do is line up a perfectly normal completely sequenced human genome and compare it with the full genome of an MPD patient and thereby locate the genetic anomaly that is causing the MPD. It’s a far more complex process since there are multiple chemical and biological interactions playing contributory roles. And then there are those millions of SNPs (pronounced Snips) – single nucleotide polymorphisms – along the way, slight transpositions, deletions or additions of a single base to the DNA strand that jolt the protein production line. The high frequency of SNPs, about 10 million or an estimated 0.3% of the 3 billion + bases in the human genome, also help MPD researchers zero in on suspected genes acting as markers.

The informers. SNPs can be useful, harmful, or meaningless but they can also act as informers and help identify short strands of DNA — haplotypes — that contain targets of interest to investigators. These haplotype-tagging SNPs (htSNPs) are critical when questioning large scale data sets such as the haplotype map that grew out of the HGP. (A haplotype is a portion of associated DNA, a group of alleles on a chromatid that are transmitted as a unit.) Haplotypes are most useful in identifying DNA activities and genetic anomalies economically since recognizing just a few alleles within the haplotype permits an investigator to identify the entire haplotype housing a target gene to be isolated and studied.

A new suspect. Consider the JAK2V617F mutation. In 2005, no fewer than five independent laboratories nearly simultaneously discovered the JAK2 mutation. It’s not entirely coincidental. Since the discovery of the BCR-ABL oncogene as a target for the development of imatinib, other tyrosine kinases received renewed attention. (JAK2 like ABL is active in signaling transcription.)

As reported by Robert Kralovics, Radek Skoda and others in the New England Journal of Medicine “A Gain-of-Function Mutation of JAK2 in Myeloproliferative Disorders,” the discovery was made on the short arm (p) of Chromosome 9 that included the Janus kinase 2, a protein known to be active in hematopoiesis, They found evidence of a homozygous transversion mutation, causing phenylalanine (F) to be substituted for valine (V) at position 617 of JAK2 on both copies of the gene. The mutation, V617F, appeared in nearly all patients with PV, more than half of those with MF, and about a quarter of ET patients. .

Heterozygosity is the normal genetic balancing act that occurs when two slightly different copies of a gene – one from each parent – provide insurance against pernicious dominance. Homozygous transversion results in LOH, loss of heterozygosity, and leaves the dominant mutated gene — in this case JAK2v617F– with unfettered license to stimulate proliferation of hematopoeic cells. The total number of mutated JAK2v617F cells resulting from LOH, the allele burden, is closely associated with MPD clinical symptoms.

The authors concluded that patients with the V617F mutation had a significantly longer duration of disease and higher rate of complications. Essentially, this point mutation caused the signal transduction pathways used by JAK2 to remain in the ON position at all times since there was no longer a countervailing normal allele.

With this discovery you could almost hear the sounds of champagne corks popping in labs and clinics.

This could be it! It appears the suspected perpetrator has been identified and surrounded. Once the target is identified the silver bullet will surely find its mark. Isn’t that what happened with BCR-ABL? Isn’t this one of the promised payoffs of the human genome project?

In this heady atmosphere JAK2v617F inhibitors were cooked up in multiple labs to the delight of shareholders and MPD patients alike. As the race to knock out the JAK2 mutation was on, candidate compounds were put on trial by Incyte, TargeGen, Cephalon, AstraZeneca, Exelixis and Cytopia — and a feverish pace of development and professional interaction and publishing followed.

Since the JAK2 mutation was clearly present in the majority of MPD patients and almost universally present among PV patients, it’s understandable that the drive to develop and test JAK2 inhibitors would consume much of the MPD research community’s energy over the following four years.

By 2009 it was clear that nearly all JAK2v617F inhibitors were more or less effective in relieving some of the common clinical MPD symptoms such as splenomegaly, pruritis and night sweats. Some JAK2 inhibitors also appear to reduce the mutated JAK2V617F allele burden but is that enough to affect the clone itself or halt progression of the MPD? In 2010, in the midst of continuing trials of the next generation of JAK2v617F inhibitors, the jury is still out. Researchers and patients are encouraged by initial results but few expect these inhibitors, alone, to be the magic bullet.

Reviewing the evidence. The JAK2 mutation remains an object of extreme interest for investigators. In response to needs of the organism, the Janus kinase signals hematopoietic surface receptors to initiate cell reproduction. The mutated Janus kinase, however, produces a sustained signal. That’s like a switch locked in the on position, similar to the effect of the BRC-ABL fusion protein. The presence of this mutation in almost all cases of polycythemia vera and in nearly half the cases of ET and MF strongly suggests JAK2v617F plays some role in MPDs. But what?

With hindsight, it’s easy to see there always was a problem in establishing JAK2v617F. as the single MPD causing genetic mutation. For one thing, it was associated with three different diseases that followed different courses. More importantly, many if not most patients with non-PV MPDs were negative for the mutation and even some PV patients did not have it. There seems, however, little doubt that JAK2v617F. plays some significant role in the initiation or proliferation of MPDs.

There remained a fundamental question not answered by this research. It was known the JAK2v617F mutation showed up in hematopoietic cells but was it a somatic event or was it inherited through the germline?

The idea that MPDs were caused by an inherited pre-disposing factor were shortly to get support from different corners of the MPD community.

En famille? Peut-etre, non..ou oui. A year after discovery of the JAK2 mutation, a French group (Albert Najman, Francois Delhommeau and others) published a paper in Blood, “Genetic and clinical implications of the Val617Phe JAK2 mutation in 72 families with myeloproliferative disorders.”

While concluding “the evidence does not support the existence of germ-line JAK2 mutation as a predisposing factor (for MPD)” the investigators held open the door to the “pre-disposing factor” school. They wrote “…determination of the JAK2 genotype may contribute to the search for genetic determinants favoring the development of MPD “

Najman, Delhommeau, et. al, by also finding the mutation in NK (natural killer) cells, confirmed its impact on a multi potent hematopoietic progenitor cell. Unsurprisingly, they associate the homozygous mutation with a proliferative advantage and higher risk for hematological complications.

The Swedish Connection. With publication of the very large scale Swedish population study in 2008 we start coming to a consensus on a refined hypothesis for the etiology of MPDs..

Using the Swedish health registry during the 1956-2005 period, researchers Ola Landgren , Magnus Bjorkholm and others from the Swedish Myeloproliferative Study Group and the National Cancer Institute, NIH, analyzed the records of nearly 25,000 relatives of 11,000 MPDs patients in Sweden, searching for patterns of disease among first degree relatives of MPD patients. They documented a five to seven-fold elevated risk of MPDs among first-degree relatives of MPD patients. In the absence of any evidence of direct inheritance of the disease, they suspect predisposition due to strong, shared susceptibility genes. They conclude the JAK2 mutation is not an early germline predisposing factor but rather a facilitator of proliferative advantage.

The conclusions are straightforward but the data is a little problematical. As Johns Hopkins’ Jerry Spivak points out, the data is a bit skewed since the Swedish hematologists switched from using the PVSG diagnostic criteria to the World Health Organization MPD diagnostic criteria. There is another concern. Since the data collection extends back to 1956, the accuracy of diagnosis in the absence of access to patients or clinical records, could be an issue. (The Hazleton, Pennsylvania PV cluster studies completed last year reported substantial over- and under-reporting of MPDs and demonstrated the significance of shifting diagnostic criteria.) However, the sheer size of the Swedish dataset appears to make these concerns less significant.

The Nature Genetics papers. Three papers published in the March, 2009 Nature Genetics report on studies that suggest the JAK2 mutation facilitates MPDs, either by inducing hypermutability on the JAK2 locus or by transmitting an inherited genetic predisposition. This is the gray area where acquired and inherited traits come together.

Nicholas Cross, University of Southampton School of Medicine, was lead author of the first report . Analyzing SNPs at or near the JAK2 locus in a study of 92 haplotypes, investigators discovered those designated “1” and and “46” differed by only one SNP. The association of this haplotype with JAK2 mutation was observed in all three disease entities.

Dr. Robert Kralovics, the Center for Molecular Medicine of the Austrian Academy of Sciences, lead author of the second paper, believes the hypermutability of JAK2 might have germline origin Some haplotypes favor mutation, he reasons. Somatic mutation in genes occurs randomly therefore mutations occur with equal frequency in alleles inherited from each parent. However 93 of 109 individuals heterozygous for a particular SNP (RS 12343867) within JAK2 developed the JAK2v617F mutation in the gene containing the C allele of this SNP.

Patients who were homozygous or heterozygous for the C alleles of RS12343867 had a greater probability of being JAK2v617F than did subjects homozygous for the “T” allele. This, says Kralovics. indicates that a specific JAK2 haplotype confers susceptibility to JAK2v617F positive MPN.

It’s noteworthy that the “C” allele of this SNP is part of Nicholas Cross’ haplotype 46/1.

The third paper, authored by Robert Klein and Ross Levine, both of Sloan Kettering, cites increased family risk to acquire MPD as a possibility for inherited genetic predisposition. Using genome wide SNP array data from patients with PV or ET, the investigators found four SNPs that were “significantly enriched” in these patients. The G allele of one SNP was a constituent of haplotype 46/1. According to Levine, “The other three SNPs of interest were non-JAK2 germline variants that likely contribute to MP(D) predisposition and phenotype.”

The Harvard familial MPD study – In search of the haplotype.

MPD Foundation grantees, Dr. Ben Ebert, of the Broad Institute and Dr. Ann Mullally of Harvard Medical School are part of the group studying the causes of familial MPDs. They are trying to determine the inherited genetic basis for the disease within families with multiple instances of MPD among first degree relatives.

One of the first steps is to recruit a cohort of Ashkenazi Jews with familial MPD. The reason is simple, European (Ashkenazi) Jews are diagnosed with MPDs at a higher rate than the general population. An Israeli study concludes: “The incidence (of MPD) in Ashkenazi Jews originating from eastern and central Europe, was 10 and 20 folds higher than in Sephardic Jews and Arabs respectively.”

Researchers believe this might be another instance of the Founder Effect. Well-defined population groups living in relative isolation over long periods of time share more common genes and phenotypes. It is suspected that an inherited gene, perhaps common to a mutation in a single ancestor, the Founder, might be triggering the disease, in this case of a higher incidence of MPDs

This study of Ashkenazi Jews is an opportunity to sort out the division between somatic and germline MPD influences within a group that had been largely sequestered from the general European population for at least 500 years.

Dr. Ann Mullally believes it is “possible we’ll find something new about the biological development of MPDs because we’re taking an unbiased genomic approach to this research. We have no candidate gene. We couldn’t possibly sequence the whole genome of everyone in the cohort but we could do more detailed sequencing of a likely subset of the group.”

And to reduce search time, the research group will turn to the Informers, SNPs and haplotypes, those short strands of related DNA, to locate those likely subsets of the group.

“We know,” says Dr. Mullally, “that the JAK2 mutation is not inherited but our study showed the inheritance of a specific haplotype can predispose you to acquire the mutation. There may well be overlap between what’s inherited and predisposed and what’s acquired that causes the MPD.”

As one of the newest MPD patients, Eileen Wiggins is watching the hunt unfold. Now, with her PV stabilized under hydroxyurea, she is no longer an innocent bystander. She’s part of the MPD community closely following the biomedical investigators close in on the inherited and somatic causes of myeloproliferative neoplasms. “So much has been discovered in the past few years I feel certain we are close to finding a cure for MPDs.”