PMC

(A–C): Hematocrit (A), blood hemoglobin (B), and reticulocyte counts (C) from normal (–) B6 mice (green), B6 recipients of B6 WT BM transduced with retrovirus expressing murine JAK2 WT (blue) or JAK2 V617F (red), and B6 Lyn ^−/− Hck ^−/− Fgr ^−/− BM transduced with retrovirus expressing JAK2 V617F (orange).
The difference between recipients of JAK2 V617F-transduced Lyn ^−/− Hck ^−/− Fgr ^−/− BM and recipients of JAK2 WT-transduced WT BM was significant (unpaired t-tests) for hematocrit (P = 0.0009), hemoglobin (P = 0.0007), and reticulocytes (P = 0.0068), while the corresponding differences between recipients of JAK2 V617F-transduced BM from WT and Lyn ^−/− Hck ^−/− Fgr ^−/− donors were not significant.
(D) Western blot analysis of extracts of primary myeloerythroid cells from individual normal (lanes 1–3) B6 mice, recipients of WT BM transduced with JAK2 WT retrovirus (lanes 4–5), recipients of WT BM transduced with JAK2 V617F retrovirus (lanes 6–9), and recipients of Lyn ^−/− Hck ^−/− Fgr ^−/− BM transduced with JAK2 V617F retrovirus (lanes 10–14).
The membrane was immunoblotted with antibody recognizing the phosphorylated activation loop tyrosine (Y146 homolog) of c-Src, Lyn, Hck, Fyn, Lck, and Yes (top panel), and subsequently blotted with antibody recognizing total c-Src, Fyn, Yes, and Fgr (bottom panel).

(A–B): Western blot analysis of primary myeloerythroid cell extracts from control (normal) mice and recipients of JAK2 WT- and JAK2 V617F-transduced BM.
(A) Activated JAK2 (anti-pY1007/10008, top panel) and total JAK2 (middle panel) blots.
Note that JAK2 V617F is modestly overexpressed and constitutively active.
Anti-actin immunoblot demonstrating equivalent protein loading is at the bottom.
(B) Constitutive activation of Stat5, Gab2, and ERK, with overexpression of Bcl-X_L in spleen cells from JAK2 V617F recipients.
Primary myeloerythroid cell extracts from mice in (A) were immunoblotted with the indicated antibodies against total or phosphorylated Stat5, Gab2, ERK, and against Bcl-X.
(C) Oligo- to polyclonal MPD in recipients of JAK2 V617F-transduced BM.
Genomic DNA from spleens of individual JAK2 V617F recipients sacrificed at the indicated times post-transplant were analyzed with a radioactive GFP probe that detects a distinct band from each proviral integration event.
For comparison, DNAs from control Ba/F3 cell lines each containing a single provirus (Con1 and Con2) and spleen DNA from a representative mouse with BCR-ABL-induced CML-like MPD are included at left, along with DNA size markers in kb.

(A) Peripheral blood leukocyte counts for the three cohorts in Balb/c and B6 backgrounds as in .
The difference in leukocyte counts between JAK2 V617F and JAK2 WT recipients was significant for the Balb/c cohort (P = 0.0039) but not for B6 (P = 0.0542, unpaired t-test).
The hatched portion of the histogram represents the percentage of the total leukocyte count comprised by neutrophils.
(B) Wright-Giemsa-stained cytospins of peripheral blood leukocytes from representative mice with BCR-ABL-induced CML-like MPD (left) or JAK2 V617F-induced PV-like MPD (right).
Note the predominance of mature neutrophils and lack of immature myeloid elements (myelocytes and promyelocytes) in the JAK2 V617F recipient.
(C) Platelet counts for the groups in (A).
There was no significant difference in platelet counts between the three cohorts in either stain.
(D) GFP fluorescence (left panels) and phase contrast images (right panels) of purified megakaryocytes from recipients of JAK2 V617F-transduced BM (top panels) or normal control mice (bottom panels).
Note the GFP fluorescence in the majority of megakaryocytes from JAK2 V617F recipients.
(E) Tail bleeding time (Balb/c cohort) for the three groups.
The mean bleeding time (bar) of JAK2 V617F recipients was significantly longer than that of vector or JAK2 WT recipients (P = 0.0328 and P = 0.0002, respectively, unpaired t-test).

(A) Hematocrit (left panel) and reticulocyte counts (right panel) of cohorts of mice treated with twice daily oral gavage with 100 mg/kg imatinib (n = 3), 10 mg/kg dasatinib (n = 4), or vehicle (n = 2), determined before initiation of therapy (white bars, “pre”) or after 2 weeks of treatment (black bars, “post”).
The hematocrit and reticulocyte count were significantly decreased in response to imatinib therapy (P = 0.0267 and P = 0.0053, respectively, unpaired t-test), while imatinib had no effect on these parameters in normal mice (data not shown).
(B) Inhibition of Ba/F3 parental cells grown in IL-3 (green squares) or Ba/F3 cells expressing BCR-ABL (magenta triangles) or JAK2 V617F (red circles) grown without IL-3 by JAK Inhibitor I (left) or AG-490 (right).
Note that both JAK Inhibitor I and AG-490 inhibit IL-3-dependent proliferation of parental Ba/F3 cells, but only AG-490 inhibits the growth of BCR-ABL-expressing cells, as previously reported .
(C) Hematocrit (left panel) and reticulocyte counts (right panel) of mice treated with continuous parenteral administration of 300 µg/day AG-490 (n = 3) or vehicle (n = 2), determined before initiation of therapy (white bars), or after 2 weeks of treatment (black bars).
The hematocrit and reticulocyte count were significantly decreased in response to AG-490 therapy (P = 0.0134 and P = 0.0374, respectively, unpaired t-test).
There was no significant effect of AG-490 on the hematocrit or reticulocyte count of recipients of JAK2 WT-transduced BM (data not shown).

(A–C): Hematoxylin and eosin stains (magnification 500×) of BM (A), spleen (B), and liver (C) from representative polycythemic recipients of JAK2 V617F-transduced BM.
In the spleen, a lymphoid follicle (f) and an area of erythroid hyperplasia (e) are indicated.
The same tissues from recipients of JAK2 WT-transduced BM showed no pathological changes compared with normal controls (data not shown).
(D) Wright-Giemsa stains (top panels) and acetylcholinesterase stains (bottom panels) of purified megakaryocytes from recipients of JAK2 V617F-transduced BM (right panels) or normal control mice (left panels).
Note the preponderance of small megakaryocytes from JAK2 V617F recipients, some of which are undergoing proplatelet formation (arrowheads), accompanied by abnormal mitoses with low apparent ploidy (insert).
(E) Flow cytometric analysis of BM and spleen from a representative recipient of JAK2 V617F-transduced BM, stained with the indicated hematopoietic lineage antigens.
The mean fluorescence intensity of GFP expressed from JAK2 retroviral provirus was reproducibly and significantly lower than the GFP fluorescence of a comparable BCR-ABL retrovirus (data not shown).
Note the shift of GFP^+/lo erythropoiesis (TER-119⁺) from BM to spleen and the expression of GFP in low abundance CD41⁺ megakaryocytes.
(F) Flow cytometric assessment of erythrocyte differentiation , assessed by expression of transferrin receptor (CD71) and TER-119 in splenocytes of recipients of JAK2 WT-transduced (top) and JAK2 V617F-transduced (bottom) BM.
Note the increased (3.5-fold) population of CD71^+/loTER-119⁺ erythroblasts in JAK2 V617F recipients.

(A–C): Hematocrit (A), blood hemoglobin (B), and reticulocyte counts (C) from cohorts of Balb/c or B6 mice transplanted with syngeneic BM cells transduced with empty vector (green, n = 2), or retrovirus expressing murine JAK2 WT (blue, n = 10) or JAK2 V617F (red, n = 15).
In the case of the B6 cohorts, untransplanted mice (“Normal”, n = 4) were used instead of empty vector controls.
The difference between JAK2 V617F and JAK2 WT recipients was significant (unpaired t-test) for hematocrit (P<0.0001 for Balb/c and P = 0.0071 for B6), hemoglobin (P = 0.0011 for Balb/c and P = 0.0024 for B6), and reticulocytes (P<0.0001 for Balb/c and P = 0.0195 for B6), while there were no significant differences between vector and JAK2 WT recipients.
(D) Reticulocyte stains of peripheral blood from recipients of BM transduced with JAK2 WT (left) or JAK2 V617F (right).
(E–F): Erythrocyte survival curves (E) and red cell mass determination (F) for the three groups (Balb/c background).
The red cell mass of JAK2 V617F recipients was significantly greater than that of recipients of vector- or JAK2 WT-transduced BM (P = 0.0068 and P = 0.0009, respectively, unpaired t-test).
(G) Plasma Epo levels for the three groups (B6 background).
The mean Epo level (bar) of JAK2 V617F recipients was significantly lower than that of recipients of vector- or JAK2 WT-transduced BM (P = 0.0048 and P = 0.0151, respectively, unpaired t-test).
(H) Number of CFU-E colonies from normal BM (green) or spleen of JAK2 V617F recipients (red), expressed as percent of maximal colony number (Balb/c background).
Data for JAK2 WT recipients were similar to normal mice (data not shown).
All studies were carried out between 6 and 10 weeks post-transplantation.

(A) Box-style plots of hematocrit (red squares, left axis) and reticulocyte counts (blue triangles, right axis) in a cohort (n = 12) of Balb/c recipients of syngeneic JAK2 V617F-transduced BM, followed for over eight months after transplantation.
Similar data were observed for a B6 cohort (data not shown).
(B) Increasing fibrosis (demonstrated by reticulin staining) in spleen (left panels) and BM (right panels) of representative JAK2 V617F recipients at about 3 months (middle panels) and 7 months (bottom panels) after transplantation.
Note the marked increase in reticulin staining at 7 months in the JAK2 V617F recipients, but not in recipients of JAK2 WT-transduced BM (top panels).
(C): Efficient transfer of the PV-like MPD by transplantation of BM from primary mice sacrificed either in the early, polycythemic phase (left, n = 3, sacrificed 72–167 days post-transplant) or the late, myelofibrotic phase (right, n = 2, sacrificed 208 days post-transplant), to lethally irradiated syngeneic secondary recipients (n = 6 for early phase and n = 4 for late phase).
The graphs depict mean hematocrit (black, left axis), reticulocyte count (white, right axis) and peripheral blood leukocyte count (grey, right axis) of the donors at the time of sacrifice, and of the recipients at day 30–70 post-transplant.
For transplants performed in the late phase of the disease, the hematocrit and reticulocyte counts of recipients were significantly higher than of the donors (P = 0.0407 and P = 0.0337, respectively, unpaired t-test), while there was no significant difference between donors and recipients transplanted in the early phase.