Proteomic Analysis Identifies HSP70 As a Novel Target Therapy to Polycythemia Vera

Abstract 2827

Background Proteomic screening is a useful tool to find new therapeutic targets. Genetic molecular studies are frequent in myeloproliferative neoplasms (MPN), however proteomic screening studies are limited.

HSP70 is a Heat Shock protein related to apoptosis supression and erythroid differenciation through GATA-1. Additionally, other HSP, as HSP90. have been found as novel therapies in Polycythemia Vera (PV) and Essential thrombocythemia (ET) (Marubayashi et al, Journal of Clinical Investigation, 2010Oct).

Aims To analyze the phenotypic divergence between PV and ET by a proteomic screening, and to validate these results by protein expression analysis and in vitro model of MPN, with the aim of identifying alternative routes for targeted therapy.

Methods Seventy-one MPN diagnosed by WHO criteria were included in the study: 25 PV, 24 ET JAK2V617F, 12 ETJAK2 Wild Type (JAK2WT) negative and 11 Primary Myelofibrosis (PMF). Additionally 24 healthy subjects were used as controls.

First of all, granulocytes from whole venous peripheral blood were isolated and the corresponding cytosolic protein fraction was extracted. Cytosolic proteomes of 10 PV, 10 ET and 10 healthy subjects were analyzed using 2D-DIGE gels followed by MALDI-TOFTOF mass spectrometry (MS) analysis of the spots of interest. Results were analyzed with DeCyder v7.0 and Mascot software. Secondly, bone marrow biopsies (BMB) of 55 MPN patients (11 PV, 11 ET JAK2V617F, 11 ET JAK2WT, 11 PMF and 11 healthy controls) were selected to perform immunohistochemistry (IHC) with anti-HSPA1A (HSP70), anti-SERPINB1 and anti-LTA4H. Afterwards, western blot of these three proteins was performed.

Finally, an in vitro model of MPN was employed. Mononuclear cells from 4 PV, 4 ET and 3 healthy donors were extracted and seeded in Methocult with IL-3, SCF and EPO. A HSP70 inhibition assay was performed by the drug KNK437 at 100μM, 50μM and 10μM. Results were analyzed by BFU-E count, viability study by trypan blue and flow cytometry (FCM) employing anti-CD45, anti-CD41, anti-CD34, anti-CD71, and Annexin antibodies.

Intracellular proteins including phospho-proteins p38, P-p38, MEK, P-MEK, STAT1, P-STAT1, AKT1 and P-AKT were studied by cytometric bead array multiplexed bead-based immunoassay (CBAs) technique. The Mann-Whitney non-parametrical statistical hypothesis test was used to assess the statistical significance of our results.

Results 2D-DIGE analysis found 112 spots with statistically significant differences in protein expression between PV and ET samples. On the other hand, 241 spots showed differential expression between ET and healthy donors, and 229 between PV and healthy donors. We identified, by MS, 67 proteins differentially expressed between PV and ET, 20 between PV and controls and 6 between ET and controls. A large number of these proteins were metabolic and citoskeleton proteins, as Lactotranferrin, Enolase, Actin, etc. However, three spots were especially interesting according to our hypothesis: SERPINB1, LTA4H and HSPA1A (HSP70); the last, chaperone related with GATA-1 and erythroid differentiation. IHC showed over-expression of HSPA1A in granulocytes of PV (72% positives patients, 80% positive granulocytes) compared to ET (50% positive patients, 30% positive granulocytes). Finally, significant differences in inhibition of BFU-E growth and cell proliferation were found between cultures treated with HSP70 inhibitor, KNK437 (100μM and 50μM) versus cultures without treatment (P=0.029). FCM of BFU-E cultures pointed to a dramatic increase of erythroid apoptotic cells. CBAs analysis showed a decrease of P-STAT1 and P-MEK in cells of cultures under KNK437 treatment.

Conclusions Our results suggest that HSP70 could play a key role in erythroid survival and differentiation through JAK-STAT and MAPK pathways. These molecules might contribute to their phenotypic divergence. Finally, HSP70 could be a new therapeutic target, and KNK437 a novel PV treatment.