JAK2: how many faces in MPDs?

Polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF) share a valine for phenylalanine substitution (V617F) in the JH2 domain of JAK2. JAK2 is a cytoplasmic tyrosine kinase that associates with homodimeric type I cytokine receptors such as EPOR, MPL, and GCSFR, and serves to transduce signals generated after ligand binding to the appropriate cytokine receptor. The V617F mutated protein is constitutively active, thus conferring cytokine hypersensitivity and cytokine-independent growth to hematopoietic cells, and reproducing an MPD phenotype in murine bone marrow (BM) transplantation assay.¹  The large majority of PV patients, and 50% to 70% of ET or PMF patients, harbor the V617F allele. There are several possible explanations for the occurrence of a single disease-associated allele in related, though clinically distinct, MPDs; they may encompass (but not be limited to) the different hierarchical levels of the hematopoietic stem/progenitor cells that eventually acquired the mutation, reciprocal variations in the amount of mutant and wild-type allele, and individual genetic backgrounds of the patients.

SNPs are the most frequently occurring genetic variation in the human genome, estimated to exceed 9 million bases, and contribute to heritable interindividual differences in complex phenotypes. Development of high-throughput genotyping methods has expanded their use to include identification of genomic regions contributing to cancer phenotype.²  Pardanani and colleagues have used an association-based study design to compare allele frequency at polymorphic loci in candidate genes involved in the JAK-STAT signaling pathway in MPD patients. They found that 3 polymorphic alleles in the JAK2 gene itself were differentially represented in patients with PV or ET, while 3 additional SNPs were associated with PV only. Of note, association of the 3 SNPs with PV or ET remained after stratifying for JAK2V617F mutational status. Conversely, genotyping of other candidate genes—EPOR, MPL, and GCSFR—did not produce additional informative SNPs.

What are the implications of these findings, and where can they lead? First, they provide a proof-of-principle that genetic variations other than the V617F allele contribute to phenotypic diversity in MPDs, although those SNPs do not necessarily represent disease-predisposing alleles. Due to an anticipated relatively small effect of modifier allele(s) in complex phenotypes, multiple testing corrections are necessary to account for the coincidental concurrence of other individual genetic variables, themselves acting as modifiers.³  The recent founding of international collaborative groups focused on MPDs represents an ideal platform for these kinds of studies, including replication of current findings in an independent patient set. Second, researchers are urged to come back to JAK2 to study functional correlates of the SNPs characterized by Pardanani and colleagues. When SNPs do occur, they could result in abnormal expression, stability, or function of a protein, and directly contribute to phenotypic variations. The task is now to assess whether and how these SNPs in JAK2 influence cellular functions,⁴  providing a rationale for their preferential association with unique MPD phenotypes.

The coexistence of a catalytic and a negative autoregulatory domain in the same mol-ecule originally led to the acronym JAK2: Janus Kinase 2, in reference to the Roman God Janus who is depicted with 2 faces to signify beginnings and endings. Playfully, someone also referred to JAK2 as simply as “Just Another Kinase.”⁵  But findings from the work of Pardanani and colleagues suggest that 2 faces are still not enough to decipher the whole story of JAK2 in MPDs.