ITCH Deficiency Promotes an MPN Phenotype with Eosinophilia, Inflammation, Mislocalization and Overexpression of the Transcription Factor NF-E2

Background: We have described overexpression of the transcription factor NF-E2 in MPN patients and shown that elevated NF-E2 levels cause a MPN phenotype in transgenic mice. This includes thrombocytosis, leukocytosis, splenomegaly as well as an expansion of the stem- and progenitor cell compartments in the bone marrow.

Recently, we have shown that, counterintuitively for a transcription factor, NF-E2 is located exclusively in the cytoplasm in the vast majority of erythroid cells in the bone marrow (85%). Patients with PMF show a statistically highly significant elevation in the proportion of cells displaying nuclear NF-E2 compared to either healthy controls or ET and PV patients. However, the molecular mechanisms regulating the subcellular localization of NF-E2 and its aberrant localization in PMF remain to be investigated.

The E3 ubiquitin ligase ITCH has been postulated to stabilize and retain NF-E2 in the cytosol by protein-protein interaction and subsequent ubiquitinylation. The phenotype of ITCH deficient mice, however, has only been described briefly: animals display splenomegaly and an expansion of the stem cell compartment. The effect of ITCH deficiency on peripheral blood counts and on NF-E2 activity has not been determined.

Aims: To characterize the phenotype of ITCH deficient mice and investigate the effect of ITCH deficiency on NF-E2 localization and activity.

Methods: The peripheral blood and bone marrow of ITCH knock out mice as well as of heterozygous and wild-type control animals was analyzed: CBCs were determined every four weeks, stem- and progenitor populations in the bone marrow were assessed by 7-color FACS. Expression levels of NF-E2 and its targets genes were measured by quantitative PCR. Plasma cytokine concentrations were measured by Cytometric Bead Array. To determine the subcellular localization of NF-E2, immunohistochemical stainings of ITCH knock out BMs and wild-type controls were conducted.

Results: At several consecutive time points ITCH knock out mice displayed a statistically significant elevation in WBC compared to heterozygous and wild-type littermates. Interestingly, both the percentage and the absolute number of eosinophils were significantly increased, some animals presenting with a drastic eosinophilia, the differential containing over 60% eosinophils. Furthermore, ITCH knock out mice display a significant decrease in platelet count, accompanied by an increase in platelet mass and volume, indicative of giant platelets. In the bone marrow ITCH deficient mice show a significant increase in the absolute number of Common Myeloid Progenitors (CMP).

NF-E2 expression levels in the peripheral blood as well as in the bone marrow were highly statistically significantly increased compared to the levels measured in wild-type or heterozygous control mice. Consequently, the NF-E2 target gene Thromboxane Synthase A was statistically significantly overexpressed in peripheral blood of ITCH knock out mice.

Plamsa concentrations of the inflammatory cytokines INF-γ and TNF were statistically significantly elevated, reaching two to threefold higher levels in ITCH knock out mice compared to wild-type littermates.

Lastly, NF-E2 subcellular localization was altered in ITCH deficient mice, which display a significant increase in the proportion of megakaryocytes positive for nuclear NF-E2.

Summary/Conclusions: Our data identify the E3 ubiquitin ligase ITCH as a regulator of NF-E2 activity. Impaired ITCH activity leads to both an NF-E2 overexpression and an increased nuclear NF-E2 localization that together drive overexpression of NF-E2 target genes. Furthermore, ITCH deficiency leads to higher inflammatory cytokine levels, comparable to those seen in PMF patients. All of these factors contribute to the resulting myeloproliferative phenotype with eosinophilia. Our data provide the first pathophysiological explanation of the pathognomonic symptom of ITCH deletion: pruritus in "itchy" mice. Moreover, given the aberrant NF-E2 localization in PMF patients, our data provide a possible mechanism and underscore the role of elevated NF-E2 activity in the pathophysiology of myeloproliferative neoplasms.