Constitutive STAT5 phosphorylation in CD34+ cells of patients with primary myelofibrosis: Correlation with driver mutation status and disease severity
Constitutive STAT5 phosphorylation in CD34+ cells of patients with primary myelofibrosis: Correlation with driver mutation status and disease severity
Carlotta Abb? 0 1
Rita Campanelli 0 1
Paolo Catarsi 0 1
Laura Villani 0 1
Vittorio Abbonante 1
Melania Antonietta Sesta 1
Giovanni Barosi 0 1
Vittorio Rosti 0 1
Margherita MassaID 1
0 Center for the Study of Myelofibrosis, Laboratory of Biochemistry, Biotechnology and Advanced Diagnosis, IRCCS Policlinico San Matteo Foundation , Pavia , Italy , 2 Laboratory of Biochemistry, Biotechnology and Advanced Diagnosis, IRCCS Policlinico San Matteo Foundation, Pavia, Italy and Department of Molecular Medicine, University of Pavia , Pavia , Italy , 3 Laboratory of Biochemistry, Biotechnology and Advanced Diagnosis, IRCCS Policlinico San Matteo Foundation , Pavia , Italy
1 Editor: Daniel Thomas, Stanford University , UNITED STATES
Primary Myelofibrosis (PMF) is a myeloproliferative disorder associated with JAK2V617F, Calreticulin (CALR) indels, and MPLW515L/K mutations activating the tyrosine kinase JAK2 and its downstream signaling pathway. The nature of signaling abnormalities in primary cells from PMF patients is poorly understood, since most of the work has been performed in cell lines or animal models. By flow cytometry we measured constitutive and cytokine induced phosphorylation of STAT5, STAT3, and ERK1/2 in circulating CD34+ cells from 57 patients with PMF (20 with prefibrotic-PMF) and 13 healthy controls (CTRLs). Levels of constitutive and TPO induced p-STAT5, and IL6 induced p-STAT3 were higher in patients than in CTRLs. Constitutive p-STAT5 values were lower in CALR than homozygous JAK2V617F mutated CD34+ cells from PMF patients. Moreover, constitutive p-STAT5 and IL6 induced p-STAT3 values correlated directly with circulating CD34+ cell number/L, and inversely with the frequency of circulating CD34+ cells expressing CXCR4. Constitutive p-STAT5 values of CD34+ cells were also inversely correlated with hemoglobin levels. When the patients were divided according with presence/absence of JAK2V617F mutation, all the correlations described characterized the JAK2V617F+ patients with prefibrotic-PMF (P-PMF). In conclusion, increased constitutive p-STAT5 and IL6 induced p-STAT3 values in circulating CD34+ cells characterize patients with PMF. Constitutive p-STAT5 and IL6 induced p-STAT3 values correlate with circulating CD34+ cell number/L, the frequency of circulating CD34+ cells expressing CXCR4 and hemoglobin levels within the prefibrotic JAK2V617F+ patient population. Our data point toward a complex activation of STAT5-dependent pathways in the stem/progenitor cell compartment, that characterize the phenotypic diversity of PMF.
Data Availability Statement: All relevant data are
within the manuscript or provided as supporting
Funding: This work was supported
by: Associazione Italiana Ricerca sul Cancro (AIRC)
project number 10005 ?Special Program Molecular
Clinical Oncology 5x1000? to Associazione Italiana
per la Ricerca sul Cancro Gruppo Italiano Malattie
Mieloproliferative (AGIMM) (A detailed description
of the AGIMM project is available at http://www.
progettoagimm.it), received by GB; by Ricerca
Corrente Fondazione Istituto di Ricovero e Cura a
Carattere Scientifico (IRCCS) Policlinico San
Matteo, project number 874, code number
08054517, received by VR (Vittorio Rosti) (www.
sanmatteo.org); and by Ricerca Finalizzata Italian
Ministry of Health, project number
GR-201602363136, received by VA (Vittorio Abbonante).
The funders had no role in study design, data
collection and analysis, preparation of the
manuscript or decision to publish.
Primary Myelofibrosis (PMF) is a clonal disorder of the hematopoietic progenitor cells
(HPCs) characterized by bone marrow (BM) fibrosis, increased number of peripheral blood
(PB) CD34+ cells, splenomegaly, and increased risk of leukemic transformation [
PMF has been associated with driver mutations such as JAK2V617F, Calreticulin (CALR)
indels, and MPLW515L/K [
]. JAK2 is a non-receptor tyrosine kinase that acts as an important
signal transducer in cytokine signaling and promoting growth, survival, and differentiation of
various cell types [
]. All the driver mutations activate the cytokine/receptor JAK2 pathway
and its downstream signaling such as the signaling transducer and activator of transcription
(STAT1, 3, and 5), phosphatidylinositol 3-kinase (PI3K)/ AKT/mTOR, and the MAPK/
extracellular signal-regulated kinase (ERK) pathways.
Despite their importance, the nature of signaling abnormalities in primary cells from
myeloproliferative neoplasm (MPN) patients is poorly understood. A number of studies, performed mainly
in animal models or using cell lines have reported the role of JAK2 and MPL mutations in
activating the STAT5, STAT3, PI3K/AKT, and RAS-MAPK-ERK pathways [
]; although relevant,
these data may not be representative of the physiologic/pathologic environment in humans. On the
other hand, the results of studies performed with patients are not conclusive, possibly due to both
non homogeneous methods of analysis and patient populations [
]. The evaluation of the
signaling pathways that may be altered by the driver mutations characterizing PMF may help in
understanding both the disease phenotype and the cell subset where possible abnormalities take place.
It has been reported an increased constitutive signaling of p-ERK in BM CD34+ HPCs and
of p-STAT5 and p-STAT3 in BM CD34- cells from patients with PMF; no significant
correlation was described between these signaling pathways and mutant JAK2 allele burden [
addition, gene expression profile in granulocytes from a limited number of patients with PMF
showed an increased expression of key signaling intermediates downstream of JAK2 (i.e.
STAT5B) in homozygous JAK2V617F mutant patients [
]. JAK2V617F mutation was
associated with significantly increased levels of phosphorylated STAT5 in hemopoietic cells, most
marked in megakaryocytes, of patients with MPNs [
]. Abnormal nuclear megakaryocytic
staining for p-STAT5 expression, previously associated with the JAK2V617F mutation, was
also associated with MPLW515L/K mutation [
In this study we perform a flow cytometry assay to measure constitutive and cytokine
induced phosphorylation of STAT5, STAT3, and ERK1/2 in circulating CD34+ cells from
patients with PMF and healthy controls (CTRLs) comparable for age. We investigate the
correlations of the phosphorylation patterns with mutated JAK2, CALR, and MPL genotypes and
clinical disease parameters.
PB of 57 consecutive subjects with PMF was collected from the Center for the Study of
Myelofibrosis, IRCCS Policlinico San Matteo Foundation, Pavia (Italy). All the patients included in the
study gave a written informed consent for PB collection and participation in research studies
related to their disease. We excluded subjects receiving disease-modifying drugs before or on the
date of base-cohort entry (hydroxyurea, interferon, ruxolitinib, corticosteroids,
immunosuppressive agents, and danazol). Diagnosis of overt PMF or prefibrotic-PMF (P-PMF) was based on
review of BM biopsy samples by an expert pathologist according with the 2016 WHO criteria
]. Data on medical history, concomitant diseases and drugs were obtained by interview and
recorded in the data-base. Patients were assigned a prognostic score based on the International
2 / 17
Prognostic Scoring System (IPSS) that includes a number of variables: age (older than 65 years),
the presence of constitutional symptoms, hemoglobin <10 g/dL, WBC count >25?109/L and the
presence of blasts in the PB. These factors define four risk groups: low risk (no factors),
intermediate risk-1 (one factor), intermediate risk-2 (two factors), high risk (three or more factors) [
additional prognostic score has been considered, the dynamic IPSS (DIPSS) that utilizes the same
prognostic variables used in IPSS but can be applied at any time during the course of the disease
]. Spleen index is the product of the longitudinal by the transversal spleen axis, the latter
defined as the maximal width of the organ. PB samples from 13 CTRLs comparable for age were
studied. The study was approved by the local Ethic Committee.
Genotypes and allele burden
All subjects had genetic analyses for driver mutations in blood granulocytes. JAK2V617F and
MPLW515 mutations were detected by real-time quantitative PCR. High Resolution Melting
analyses followed by bi-directional Sanger sequencing was also used to test for MPL mutations.
CALR indels were identified by High Resolution Melting analysis. The assessment of A3669G
polymorphism of the corticosteroid receptor was performed by High Resolution Melting
analysis and confirmed by direct sequencing [
Flow cytometry analysis
We measured constitutive and cytokine induced phosphorylation of STAT3, STAT5, and
ERK1/2 in the PB CD34+ cells by a phospho-specific flow cytometry assay. The signaling
activity was measured in 400 ?l of EDTA anticoagulated PB cells before (to measure constitutive
phosphorylation) and following incubation with recombinant IL6 (Miltenyi BiotecGmbH,
Bergisch Gladbach, Germany), thrombopoietin (TPO-Peprotech, Rocky Hill, NJ, USA), or
phorbol 12-myristate 13-acetate (PMA-Sigma Aldrich, St. Luis, MO, USA) for 15 minutes at
37?C, according with manufacturer?s instructions. Red cells were lysed and white cells fixed by
adding pre-warmed 1X Lyse/Fix buffer (BD Biosciences, San Jose, CA, USA) for 15 minutes at
37?C. PB cells were washed in Phosphate Buffered Saline (PBS) and permeabylized by ice cold
Perm buffer III (BD Biosciences) for 30 minutes at 4?C. Ten microliters of anti-human
CD34-FITC (BD Biosciences) were added to all the conditions. Antibodies used for
phosphoprotein detection (p-STAT3, p-STAT5, p-ERK1/2) were Alexa Fluor647-conjugated (20 ?l)
(BD Biosciences). Isotype controls were used at the same concentration as test antibodies.
Incubation was performed at room temperature for 1 hour in the dark. The same staining
procedure was adopted to assess the total STAT5 protein in CD34+ cells, by adding 10 ?l of
antihuman CD34-PE (BD Biosciences) and 10 ?l of anti-human STAT5-FITC (Thermo Fisher
Scientific, Rockford, IL, USA). PB cells (5x105-1.5x106) were acquired by a Navios flow cytometer
(Beckman Coulter, Inc, Brea, CA), and analysed by Kaluza flow analysis software (Beckman
Coulter). Based on their fluorescence, CD34+ cells were electronically gated. The median
fluorescence intensity (MFI) values were obtained by dividing the MFI of the specific antibody by
the MFI of the isotype control. ?MFI? in the manuscript will refer to this ratio.
One hundred microliters of EDTA anticoagulated blood was stained with anti-human
CD34-FITC and anti-human CXCR4-PE (BD Biosciences). After red cell lysis the samples
were centrifuged and the pellets resuspended in 300 ?L of saline. Results were expressed as
percentage of CD34+ cells co-expressing CXCR4.
Human CD34+ cells were lysed for 30 minutes at 4?C in HEPES-glycerol lysis buffer (50 mM
Hepes, 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1.5 mM MgCl2, 1 mM EGTA)
3 / 17
containing 1 ?g/mL leupeptin and 1 ?g/mL aprotinin. Afterwards, samples were clarified by
centrifugation at 15700g at 4?C for 15 min. Laemmli sample buffer was then added to
supernatants. Samples were heated at 95?C for 3 minutes, separated by electrophoresis on 12% sodium
dodecyl sulfate-polyacrylamide gel and then transferred to polyvinylidine fluoride membranes.
Membranes were probed with primary antibodies, washed 3 times with PBS and Tween 0.1%
and incubated with peroxidase-conjugate secondary antibodies. Membranes were visualized
using Immobilon western chemiluminescent HRP substrate (Millipore, Burlington, MA,
USA) and images were acquired by UVITEC Alliance Mini HD9 (Eppendorf, Hamburg,
Germany), and the protein levels detected were quantified using UVITEC NineAlliance 1D
The groups were compared by means of Mann-Whitney U-test for unpaired samples or
Wilcoxon for paired samples when indicated; ANOVA of Kruskall-Wallis and Bonferroni?s
correction as a post hoc test were used for multiple comparison. Spearman correlation test was
used when indicated. All computations were performed with STATISTICA software (StatSoft,
Inc. Tulsa, OK, USA).
Patients and controls
Two subjects (3.5%) had constitutive and cytokine induced analysis of STAT5, STAT3, and
ERK1/2 signaling activity determined at diagnosis and 55 (96.5%) after diagnosis (median
time from diagnosis 82.0 months, range, 1?389 months). Demographic, clinical, biological,
and molecular variables of the analyzed population are displayed in Table 1. None of the
patients was receiving therapy at time of sampling.
Measurement of constitutive and cytokine induced STAT3, STAT5, ERK1/
2 phosphorylation by flow cytometry
The detection of intracellular signaling pathways by flow cytometry has been shown to be
], and it was found to be of comparable sensitivity to Western blotting [
performed this technique acquiring 5x105-1.5x106 PB cells, and evaluating the signaling
activity in electronically gated CD34+ cells. We assessed the reproducibility of our assay testing 4
patients in 2 separate occasions (median months 4, range 1?24) during a stable phase of
disease; the values obtained were comparable in both constitutive (p = 0.87) and cytokine induced
condition (p = 0.79) (S1 Fig).
From now on, the acronym ?MFI? refers to the ratio: MFI of specific antibody / MFI of the
isotype control. The constitutive phosphorylation of STAT5 (p-STAT5) in circulating CD34+
cells was significantly higher (p = 0.002) in patients with PMF (n = 57) than in CTRLs (n = 13)
(Fig 1A), while p-STAT3 (MFI 1.0, range 0.04?2.4) and p-ERK1/2 (MFI 1.5, range 0.3?3.1)
were comparable to those of circulating CD34+ cells from CTRLs (p-STAT3 MFI 1.0, range
0.3?1.6; p-ERK1/2 MFI 1.4, range 1.0?1.9). The TPO induced p-STAT5 and IL6 induced
pSTAT3 were significantly higher (p = 0.00003 and p = 0.0003, respectively) in CD34+ cells
from patients with PMF than in CTRLs (Fig 1B and 1C, respectively), while PMA induced
pERK1/2 was comparable in patients (MFI 1.57, range 1.1?3.6) and CTRLs (MFI 1.4, range 1.0?
2.0). Similar results were obtained when patients with P-PMF (n = 20) with degree of fibrosis
(0?1) were separately compared to CTRLs (p-STAT5: p = 0.02, TPO induced p-STAT5:
p = 0.0006, IL6 induced p-STAT3: p = 0.007).
We analyzed the determinants of elevated constitutive and cytokine induced p-STAT5 and the
IL6 induced p-STAT3 by correlating their MFI values with parameters whose causal role on
signaling activation was plausible. We found no significant correlations with age, sex, body mass index,
or presence/absence of the A3669G polymorphism of the corticosteroid receptor (S2?S5 Figs).
Cytofluorimetric plots for CD34+ cell gating and histograms for constitutive p-STAT5,
TPO induced p-STAT5 and IL6 induced p-STAT3 from a representative patient with PMF
(Fig 2A?2D) and a representative CTRL (Fig 2E?2H) are shown.
To investigate the total STAT5 expression, we stained PB cells from patients (n = 5) and
CTRLs (n = 5) with a PE-conjugated anti-human CD34 and a FITC-conjugated anti-human
total STAT5 antibodies. These experiments showed that the MFI values of total STAT5 were
comparable in patients and CTRLs (Fig 3A). This finding was confirmed by Western blot
analysis (Fig 3B and 3C). These data indicate that the increase of p-STAT5 MFI values in patients
with PMF are not related to differences in the total STAT5 protein expression.
Driver mutations and p-STAT5 signaling of circulating CD34+ cells from
patients with PMF
When patients were divided according with their genotype and compared to CTRLs (ANOVA
of Kruskall-Wallis p = 0.005), the constitutive p-STAT5 MFI values in circulating CD34+ cells
from JAK2V617F+ (n = 34) and CALR+ (n = 19) patients with PMF were higher than those of
CTRLs (p = 0.027 and p = 0.01, respectively); in MPL+ patients (n = 4) the p-STAT5 MFI
values, possibly due to the limited number, were increased but not significantly (p = 0.06) higher
than CTRLs (Fig 4A).
In addition within the JAK2 mutant patients, the homozygous ones had CD34+ p-STAT5
MFI values higher than those of CTRLs and of patients with CALR+ genotype (p = 0.003 and
p = 0.03, respectively), while the heterozygous had p-STAT5 MFI values comparable to those
found in CD34+ cells of CTRLs and of CALR+ patients (ANOVA of Kruskall-Wallis p = 0.005)
5 / 17
Fig 1. STAT3/STAT5 constitutive and cytokine induced phosphorylation in CD34+ cells. Evaluation of constitutive
p-STAT5 (A), TPO induced p-STAT5 (B), and IL6 induced p-STAT3 (C) in circulating CD34+ cells of patients with
primary myelofibrosis (PMF) and healthy subjects (CTRLs). Median values are shown as solid lines.
Further investigating the JAK2 mutant patients, we evidenced a significant (R = 0.35,
p = 0.04) direct correlation between constitutive p-STAT5 MFI values and the mutant allele
burden (Fig 5A). We found that the significant correlation between constitutive p-STAT5 MFI
values in CD34+ cells and the mutant allele burden was present in patients with P-PMF
(n = 13, R = 0.85, p = 0.0002), but not in patients with overt PMF (n = 21, R = -0.055, p = 0.8)
(Fig 5B and 5C, respectively).
No significant correlation was found between the genotype or allele burden and the TPO
induced p-STAT5 or IL6 induced p-STAT3 MFI values in CD34+ cells (S6 and S7 Figs). No
difference was found in the signaling pathways tested in circulating CD34+ cells between
CALR+ patients expressing the 52-bp deletion (type 1) (n = 14) or the 5-bp insertion (type 2
mutation) (n = 5), as the most frequent variants (S8 Fig).
Anemia, CD34+ cell number, and CD34+CXCR4+ cell frequency are
associated with altered constitutive p-STAT5 or IL6 induced p-STAT3
signaling pathways in patients with PMF
In patients with PMF the CD34+ cell constitutive p-STAT5 MFI values were significantly
inversely correlated with hemoglobin levels (R = -0.28, p = 0.037), one of the most relevant
hematological and clinical variable, whose variations portrait disease progression; no
correlation was found with IPSS, DIPSS, WBC, monocyte, or platelet count, frequency of circulating
blood blasts, serum LDH, spleen size and degree of BM fibrosis (S1 Table).
Fig 2. STAT3/STAT5 constitutive and cytokine induced phosphorylation analysis by flow cytometry. Cytofluorimetric plots of CD34+ cell gating in a
representative patient with primary myelofibrosis (PMF) (A) and a healthy subject (CTRL) (E). Overlay histograms show the isotype control (light grey) and
constitutive p-STAT5 (B and F), TPO induced p-STAT5 (C and G) and IL6 induced p-STAT3 (D and H) signals.
7 / 17
Fig 3. STAT5 evaluation. Assessment of STAT5 by flow cytometry (A) in CD34+ cells from patients with primary myelofibrosis (PMF) (n = 5) and healthy subjects
(CTRLs) (n = 5). Western blot analysis of STAT5 in CD34+ cells derived from CTRLs and patients with primary myelofibrosis (PMF); membranes were probed with
anti-STAT5 antibody (CST 94205) and with anti-?-tubulin antibody (Abcam ab52866) as loading control (B for one representative patient and CTRL). Densitometric
analysis of STAT5/?-tubulin ratio (n = 5 CTRLs, n = 5 PMF) (C).
Among the major biological parameters of disease progression, constitutive p-STAT5 MFI
values were significantly inversely correlated with both the frequency of CD34+CXCR4+ cells
(R = -0.47 p = 0.0007), whose reduction marks the increase of the disease severity [
], and the
serum cholesterol concentration (R = -0.51, p = 0.009); in addition, constitutive p-STAT5 MFI
values were significantly directly correlated (R = 0.38, p = 0.005) with the number/L of
circulating CD34+ cells.
IL6 induced p-STAT3 MFI values of CD34+ cells from patients with PMF were significantly
directly correlated with the CD34+ number/L (R = 0.36, p = 0.008), and significantly inversely
correlated with the frequency of CD34+CXCR4+ cells (R = -0.31, p = 0.027). No significant
correlation was found when we considered the MFI values of TPO induced p-STAT5 (S1 Table).
Constitutive p-STAT5 or IL6 induced p-STAT3 MFI values and markers of disease
progression were separately performed in the JAK2V617F and CALR mutants (n = 34 and n = 19,
The correlations described in the whole population of patients with PMF characterized the
JAK2V617F+ patients (Fig 6A?6C). In fact, as shown in Table 2, in the CALR mutant subjects
we evidenced only a significant inverse correlation of constitutive p-STAT5 MFI values with
the cholesterol serum levels (R = -0.97, p = 0.0004).
Moreover, among the JAK2V617+ patients, those with P-PMF determined the significant
correlation of constitutive CD34+ cell p-STAT5 MFI values with the hemoglobin levels and the
PB CD34+CXCR4+ cell frequency (Table 3). At variance, a significant correlation of
constitutive p-STAT5 MFI values with the absolute number of circulating CD34+ cells was present in
JAK2V617+ patients either with P-PMF or PMF (Table 3). Only patients with overt PMF
showed a correlation of TPO-induced p-STAT5 and IL6 induced p-STAT3 MFI values with
the absolute number of circulating CD34+ cells (S2 Table).
8 / 17
Fig 4. Constitutive STAT5 phosphorylation and disease genotypes. Constitutive p-STAT5 signal in circulating
CD34+ cells of healthy subjects (CTRLs) and of patients with primary myelofibrosis (PMF) divided according with the
genotype (A) and according with the genotype and the allele burden within the JAK2+ patients (B). Median values are
shown as solid lines.
In this study, we have investigated the constitutive and cytokine induced p-STAT3, p-STAT5
and p-ERK1/2 MFI values in circulating CD34+ cells of patients with PMF, taking advantage
of a cytofluorimetric method. This approach allowed us to achieve a number of goals; first,
9 / 17
Fig 5. Constitutive STAT5 phosphorylation correlates with JAK2V617 allele burden. Significant direct correlation
between constitutive p-STAT5 MFI values in PB CD34+ cells and JAK2 mutated allele burden in patients with PMF
(A). The same correlation is shown in patients with prefibrotic-PMF (P-PMF) (B) and overt primary myelofibrosis
acquiring up to 1,5x106 PB cells we are able to analyse very low number of circulating CD34+
cells in both CTRLs and patients. Second, by electronically gating the CD34+ cells during the
analysis, we avoided immune-magnetic bead cell selection and the possible mechanical
activation of them during the procedure. Third, we avoided the necessity of great quantity of blood
necessary to quantify STAT phosphorylation in CD34+ cells by Western blotting, allowing the
detection also in anemic patients.
10 / 17
Fig 6. Correlations between constitutive p-STAT5 MFI values and disease parameters. Significant direct
correlation between constitutive p-STAT5 MFI values in PB CD34+ cells and hemoglobin levels (A), CD34+ expressing
CXCR4 frequency (B), and the absolute number of circulating CD34+ cells (C) in JAK2+ patients with PMF.
We demonstrate that CD34+ cells of patients with PMF have higher constitutive p-STAT5,
and cytokine induced p-STAT5 and p-STAT3 MFI values than those of CTRLs. We suggest
that the increase of constitutive p-STAT5 is actual, since the presence of total STAT5 protein is
comparable in patients and CTRLs by both flow cytometry and Western blotting. The pattern
of phosphorylation does not completely reflect what has been reported in PB CD34+ cells and
granulocytes of patients with PMF, where constitutive p-STAT3, but not constitutive
p11 / 17
STAT5, has been found increased compared to CTRLs [
]. In addition, Anand et al
 described higher values of constitutive p-ERK1/2 in CD34+ cells from BM of MF patients,
while Teofili et al, using an immune-histochemical approach, did not find any difference in
both constitutive p-STAT5 and p-STAT3 in BM CD34+ cells of patients with MF compared to
subjects without a myeloproliferative disease [
]. An explanation of the difference between our
results and published data is not obvious; however, the detection method, the cell type analysis,
and the disease phenotype could be responsible for the discrepancies. More importantly,
constitutive p-STAT3 levels have been associated in MPNs to an increase of inflammatory
cytokine secretion rather than to a proliferative phenotype [
]. The lack of increase of
constitutive p-STAT3 MFI values may reflect a low level of inflammatory component in our
cohort of patients, who showed plasma high sensitive C reactive protein levels (mean 0.3mg/
dL, standard deviation 0.3) well below the values (mean 0.6 mg/dL, standard deviation 1.3)
that we observed in a larger cohort (n = 526) of patients with PMF . Anyhow, the
significantly increased values of IL6 induced p-STAT3 indicate that this pathway is more prone to
activation in patients with PMF, even in a stable phase of the disease, than in CTRLs.
A second, original finding of our work is that constitutive p-STAT5 MFI values differ
according to the genotype, with homozygous JAK2- mutated CD34+ cells showing higher
constitutive p-STAT5 MFI values than the CALR+ ones. It is possible that the different molecular
mechanism of activation of the JAK/STAT pathway in CALR- mutated cells compared to JAK2
homozygous- mutation explains this quantitative difference. Similar differences were found
between JAK2- mutated and CALR- mutated patients with MPN affected by ruxolitinib
withdrawal syndrome [
]. We also found that the allele burden of JAK2V617F- mutated CD34+
cells directly correlates with the constitutive p-STAT5 MFI values of patients with PMF at
variance with a previous report performed in a limited number of patients [
]. Corroborating our
observation, the granulocyte gene expression profiling obtained from patients with MPN
showed that the JAK/STAT pathway activation signature was strikingly higher in patients with
n = 13
n = 21
a high JAK2V617F allele burden [
]. It is difficult to understand why the correlation between
JAK2V617F- mutated CD34+ cells and the constitutive p-STAT5 MFI values becomes stronger
within patients with P-PMF; however, since it has been reported that the allele burden is
correlated with the disease severity in both JAK2- mutated patients with PMF [
], and in
JAK2mutated animal models [
], the correlation of the JAK2- mutated allele burden with
constitutive p-STAT5 MFI values support the concept that the phosphorylation of this transcription
factor has a role in fueling the disease progression.
It has to be considered that we have measured constitutive p-STAT5 values in CD34+ cells,
while we have correlated them with the JAK2V617F allele burden assessed in granulocytes. An
unbalanced amplification of the JAK2- mutated versus non-mutated hematopoiesis in the PB
could underestimate or bias the correlation [
]; however, available data indicate that in MF, at
variance with PV, the average mutant allele burden measured in the hematopoietic
progenitor/stem cells remains stable during differentiation and is comparable to that of mature
hematopoietic cells [
]. As a third novel finding of our research, we report that constitutive
pSTAT5 activation inversely correlates with hemoglobin levels in patients with PMF. However,
when this correlation was investigated considering patients with P-PMF (fibrosis 0?1
according WHO 2016) vs patients with overt PMF (fibrosis 2?3) the inverse correlation with
hemoglobin was evident only for prefibrotic patients. Similarly, we observed that constitutive
pSTAT5 MFI values inversely correlated with the frequency of circulating CD34+ cells
expressing CXCR4 in patients with P-PMF but not in those with overt PMF. A potential interpretation
of these observations is that in the prefibrotic disease the capacity of p-STAT5 of determining
the phenotype is more pronounced than in the overt form, where it is likely that other
pathogenetic determinants can confound the role and the effect of this transcription factor on the
disease phenotype. As a matter of fact, these data, despite their interpretation, brings evidence to
the concept that prefibrotic and overt PMF are distinct biological entity [
]. In physiological
conditions, the binding of erythropoietin to its receptor activates the JAK/STAT5 pathway,
resulting in the proliferation/amplification of the erythroid compartment. However, in PMF it
is not unfrequent that patients present with normal or reduced hemoglobin levels, suggesting
that in the stem/progenitor cell compartment, despite the p-STAT activation, erythroid
differentiation is affected and does not depend only on the EpoR/JAK2/STAT5 pathway activation.
This is consistent with the observation that in mouse models the absence of STAT5 does not
affect the response to Epo [
]. It is therefore possible that in PMF hematopoiesis, at the stem/
progenitor cell level, constitutively increased p-STAT5 induces the activation of genes that
specifically counteract the EpoR pathway [
]; in particular, Interferon (IFN)-? has been shown
to be a pro-inflammatory protein that signals through the JAK/STAT pathway by binding of
activated STAT5 to different target genes [
], such as those of the apoptotic pathway, that
could be responsible of the reduction of hemoglobin levels [
Finally, we found that constitutive p-STAT5 MFI values were directly correlated with the
frequency of circulating CD34+ cells, but inversely correlated with the CXCR4 expression on
their membrane. Abnormal CD34+ cell mobilization is a well known specific biological feature
of patients with PMF [
]. We have also previously shown that in patients with MF a reduced
number of CD34+CXCR4+ circulating cells correlates with disease severity [
signaling, which occurs following the binding with its ligand SDF-1?/CXCL12, is known to
involve p-STAT5 [
], thus the constitutive activation of p-STAT5 could result in a reduced
expression of the CXCR4 on the cell membrane representing a potential mechanism of CD34+
cell mobilization from the BM into the PB. However, this cannot be the unique factor affecting
CD34+ mobilization, since constitutive p-STAT5 is also active in progenitor cells of PV and
ET patients [
], in whom this phenomenon is not detectable, at least in the chronic phase of
the disease. To be noted, another transcription factor, nuclear factor erythroid 2-related factor
13 / 17
2 (Nrf2) has already been reported to have a direct role on the HPC CXCR4 expression [
and has been shown to be involved in the regulation of CD34+ cell retention and homing to
the BM in MF, characterized by the JAK2V617F-induced oxidative stress [
In summary, we have shown here that constitutive activation of the JAK/STAT pathway in
circulating CD34+ cells of patients with PMF results in the increase of constitutive p-STAT5.
We have also shown that p-STAT5 values are directly correlated with JAK2V617F allele burden
and mark disease severity, both in term of hemoglobin reduction and of CD34+ cell
mobilization in patients with P-PMF. Our data point toward a complex activation of STAT5-dependent
pathways in the stem/progenitor compartment that characterizes PMF and its phenotypic
diversity, and that will be matter of future investigations.
S1 Fig. Assessment of the reproducibility of STAT-ERK1/2 phosphorylation by flow
cytometry. Four patients with primary myelofibrosis (PMF) were tested in 2 separate
occasions (grey and black columns), during a stable phase of disease, for the constitutive p-STAT5,
p-STAT3, and p-ERK1/2 signaling (A, B, C) and the TPO induced p-STAT5 (D), IL6 induced
p-STAT3 (E) and PMA induced p-ERK1/2 (F) signaling in circulating CD34+ cells.
S2 Fig. Correlation between age and constitutive p-STAT5 (A), TPO induced p-STAT5 (B),
or IL6 induced p-STAT3 (C) MFI values in circulating CD34+ cells.
S3 Fig. Correlation between sex and constitutive p-STAT5 (A), TPO induced p-STAT5 (B), or
IL6 induced p-STAT3 (C) MFI values in PB CD34+ cells of patients with PMF.
S4 Fig. Correlation between body mass index (BMI) and constitutive p-STAT5 (A), TPO
induced p-STAT5 (B), or IL6 induced p-STAT3 (C) MFI values in PB CD34+ cells of patients
S5 Fig. Correlation between A3669G polymorphism of the corticosteroid receptor and
constitutive p-STAT5 (A), TPO induced p-STAT5 (B), or IL6 induced p-STAT3 (C) MFI values in
PB CD34+ cells of patients with PMF.
S6 Fig. TPO induced p-STAT5 (A) and IL6 induced p-STAT3 (B) MFI values in PB CD34+
cells of patients with PMF divided according with the genotype. Median fluorescence intensity
(MFI) median values of patients with different genotype (solid lines) and of healthy subjects
(dotted line) are shown.
S7 Fig. Correlation between JAK2V617F allele burden and TPO induced p-STAT5 (A), or IL6
induced p-STAT3 (B) MFI values in PB CD34+ cells of patients with PMF.
S8 Fig. Constitutive p-STAT5 (A), TPO induced p-STAT5 (B) or IL6 induced p-STAT3 (C)
MFI values in PB CD34+ cells of CALR+ patients expressing the 52-bp deletion (type 1
mutation) or the 5-bp insertion (type 2 mutation).
14 / 17
S1 Table. Correlations between p-STAT5, TPO induced p-STAT5, and IL6 induced
pSTAT3 pathways tested in circulating CD34+ cells of patients with PMF and disease
S2 Table. Correlations between TPO induced p-STAT5, and IL6 induced p-STAT3 values
tested in circulating CD34+ cells of JAK2V617F+ patients with PMF and disease
Conceptualization: Giovanni Barosi, Vittorio Rosti, Margherita Massa.
Data curation: Carlotta Abb?, Rita Campanelli.
Formal analysis: Paolo Catarsi, Margherita Massa.
Investigation: Carlotta Abb?, Paolo Catarsi, Laura Villani, Melania Antonietta Sesta.
Methodology: Rita Campanelli.
Project administration: Vittorio Rosti, Margherita Massa.
Resources: Laura Villani, Vittorio Abbonante.
Validation: Carlotta Abb?, Vittorio Abbonante.
Writing ? original draft: Vittorio Rosti, Margherita Massa.
Writing ? review & editing: Giovanni Barosi.
15 / 17
16 / 17
1. Barosi G , Viarengo G , Pecci A , Rosti V , Piaggio G , Marchetti M et al. Diagnostic and clinical relevance of the number of circulating CD34(+) cells in myelofibrosis with myeloid metaplasia . Blood . 2001 ; 98 : 3249 - 3255 . https://doi.org/10.1182/blood.v98. 12 .3249 PMID: 11719361
2. Nangalia J , Green AR . Myeloproliferative neoplasms: from origins to outcomes . Blood . 2017 ; 130 : 2475 - 2483 . https://doi.org/10.1182/blood-2017 -06-782037 PMID: 29212804
3. Vainchenker W , Constantinescu SN . JAK/STAT signaling in hematological malignancies . Oncogene . 2013 ; 32 : 2601 - 2613 . https://doi.org/10.1038/onc. 2012 .347 PMID: 22869151
4. James C , Ugo V , Le Coue?dic JP , Staerk J , Delhommeau F , Lacout C et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera . Nature . 2005 ; 434 : 1144 - 1148 . https://doi.org/10.1038/nature03546 PMID: 15793561
5. Pikman Y , Lee BH , Mercher T , McDowell E , Ebert BL , Gozo M et al. MPLW515L is a novel somatic activating mutation in myelofibrosis with myeloid metaplasia . PLoS Med . 2006 ; 3 : 270 .
6. Kota J , Caceres N , Constantinescu SN . Aberrant signal transduction pathways in myeloproliferative neoplasms . Leukemia . 2008 ; 22 : 1828 - 1840 . https://doi.org/10.1038/leu. 2008 .236 PMID: 18769448
7. Ro?der S , Steimle C , Meinhardt G , Pahl HL . STAT3 is constitutively active in some patients with Polycythemia rubra vera . Exp Hematol . 2001 ; 29 : 694 - 702 . PMID: 11378264
8. Teofili L , Martini M , Cenci T , Petrucci G , Torti L , Storti S et al. Different STAT-3 and STAT-5 phosphorylation discriminates among Ph-negative chronic myeloproliferative diseases and is independent of the V617F JAK-2 mutation . Blood. 2007 ; 110 : 354 - 359 . https://doi.org/10.1182/blood-2007 -01-069237 PMID: 17376889
9. Anand S , Stedham F , Gudgin E , Campbell P , Beer P , Green AR et al. Increased basal intracellular signaling patterns do not correlate with JAK2 genotype in human myeloproliferative neoplasms . Blood . 2011 ; 118 : 1610 - 1621 . https://doi.org/10.1182/blood-2011 -02-335042 PMID: 21653937
10. Rampal R , Al-Shahrour F , Abedl-Wahab O , Patel JP , Brunel JP , Mermel CH et al. Integrated genomic analysis illustrates the central role of JAK-STAT pathway activation in myeloproliferative neoplasm pathogenesis . Blood . 2014 ; 123 : 123 - 133 .
11. Grimwade LF , Happerfield L , Tristram C , McIntosh G , Rees M , Bench AJ , et al. Phospho-STAT5 and phospho-Akt expression in chronic myeloproliferative neoplasms . British Journal of Haematology . 2009 ; 147 : 495 - 506 . https://doi.org/10.1111/j.1365- 2141 . 2009 . 07870 . x PMID : 19747364
12. Glembotsky AC , Korin L , Lev PR , Chazarreta CD , Marta RF , Molinas FC et al. Screening for MPL mutations in essential thrombocythemia and primary myelofibrosis: normal Mpl expression and absence of constitutive STAT3 and STAT5 activation inMPLW515L-positive platelets . European Journal of Haematology . 2010 ; 84 : 398 - 405 . https://doi.org/10.1111/j.1600- 0609 . 2010 . 01421 . x PMID : 20113333
13. Arber DA , Orazi A , Hasserjian R , Thiele J , Borowitz MJ , Le Beau MM et al. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia . Blood . 2016 ; 127 : 2391 - 2405 . https://doi.org/10.1182/blood-2016 -03-643544 PMID: 27069254
14. Cervantes F , Dupriez B , Pereira A , Passamonti F , Reilly JT , Morra E et al. New prognostic scoring system for primary myelofibrosis based on a study of the International Working Group for Myelofibrosis Research and Treatment . Blood . 2009 ; 113 : 2895 - 2901 https://doi.org/10.1182/blood-2008 -07-170449 PMID: 18988864
15. Passamonti F , Cervantes F , Vannucchi AM , Morra E , Rumi E , Pereira A , et al. A dynamic prognostic model to predict survival in primary myelofibrosis: a study by the IWG-MRT (International Working Group for Myeloproliferative Neoplasms Research and Treatment) Blood. 2010 ; 115 : 1703 - 1708 . https://doi.org/10.1182/blood-2009 -09-245837 PMID: 20008785
16. Poletto V , Rosti V , Villani L , Catarsi P , Carolei A , Campanelli R et al. A3669G polymorphism of glucocorticoid receptor is a susceptibility allele for primary myelofibrosis and contributes to phenotypic diversity and blast transformation . Blood . 2012 ; 120 : 3112 - 3117 . https://doi.org/10.1182/blood-2012 -05-433466 PMID: 22879541
17. Hamilton A , Elrick L , Myssina S , Copland M , J?rgensen H , Melo JV et al. BCR-ABL activity and its response to drugs can be determined in CD34+ CML stem cells by CrkL phosphorylation status using flow cytometry . Leukemia . 2006 ; 20 : 1035 - 1039 . https://doi.org/10.1038/sj.leu. 2404189 PMID: 16572205
18. Rosti V , Massa M , Vannucchi AM , Bergamaschi G , Campanelli R , Pecci A et al. The expression of CXCR4 is down-regulated on the CD34+ cells of patients with myelofibrosis with myeloid metaplasia . Blood Cells Mol Dis . 2007 ; 38 : 280 - 286 . https://doi.org/10.1016/j.bcmd. 2007 . 01 .003 PMID: 17350297
19. Mesa RA , Tefferi A , Lasho TS , Loegering D , McClure RF , Powell HL et al. Janus kinase 2 (V617F) mutation status, signal transducer and activator of transcription-3 phosphorylation and impaired neutrophil apoptosis in myelofibrosis with myeloid metaplasia . Leukemia . 2006 ; 20 : 1800 - 1808 . https://doi. org/10.1038/sj.leu. 2404338 PMID: 16871275
20. Levine RL , Wadleigh M , Cools J , Ebert BL , Wernig G , Huntly BJP et al. Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis . Cancer Cell . 2005 ; 7 : 387 - 397 . https://doi.org/10.1016/j.ccr. 2005 . 03 .023 PMID: 15837627
21. Verstovsek S , Kantarjian H , Mesa RA , Levy RS , Gupta V , Di Persio JF et al. Safety and efficacy of INCB018424, a JAK1 and JAK2 inhibitor, in myelofibrosis . N Engl J Med . 2010 ; 363 : 1117 - 1127 . https://doi.org/10.1056/NEJMoa1002028 PMID: 20843246
22. Yu H , Pardell D , Jove R. STATs in cancer inflammation and immunity: a leading role for STAT3 . Nat Rev Cancer . 2009 ; 9 : 798 - 809 . https://doi.org/10.1038/nrc2734 PMID: 19851315
23. Kleppe M , Kuok M , Koppikar P , Riester M , Keller M , Bastian L et al. JAK STAT pathway activation in malignant and non malignant cells contributes to MPN pathogenesis and therapeutic response . Cancer Discov . 2015 ; 5 : 316 - 331 . https://doi.org/10.1158/ 2159 - 8290 .CD- 14 -0736 PMID: 25572172
24. Barosi G , Massa M , Campanelli R , Fois G , Catarsi P , Viarengo G et al. Primary myelofibrosis: Older age and high JAK2V617F allele burden are associated with elevated plasma high-sensitivity C-reactive protein levels and a phenotype of progressive disease . Leuk Res . 2017 ; 60 : 18 - 23 . https://doi.org/10. 1016/j.leukres. 2017 . 06 .004 PMID: 28622624
25. Tvorogov D , Thomas D , Liau NPD , Dottore M , Barry EF , Lathi M et al. Accumulation of JAK activation loop phosphorylation is linked to type I JAK inhibitor withdrawal syndrome in myelofibrosis . Sci. Adv . 2018 ; 4:eaat3834 . https://doi.org/10.1126/sciadv.aat3834 PMID: 30498775
26. Akada H , Yan D , Zou H , Fiering S , Hutchinson RE , Mohi MG . Conditional expression of heterozygous or homozygous JAK2V617F from its endogenous promoter induces a polycythemia vera-like disease . Blood . 2010 ; 115 : 3589 - 3597 . https://doi.org/10.1182/blood-2009 -04-215848 PMID: 20197548
27. Anand S , Stedham F , Beer P , Gudgin E , Ortmann CA , Bench A et al. Effects of JAK2 mutation on the hematopoietic stem and progenitor compartment in human myeloproliferative neoplasms . Blood . 2011 ; 118 : 177 - 181 . https://doi.org/10.1182/blood-2010 -12-327593 PMID: 21562050
28. Teglund S , McKay C , Schuetz E , van Deursen JM , Stravopodis D , Wang D et al. Stat5a and Stat5b proteins have essential and non essential, or, redundant roles in cytokine responses . Cell . 1998 ; 93 : 841 - 850 . https://doi.org/10.1016/s0092- 8674 ( 00 ) 81444 - 0 PMID: 9630227
29. Moucadel V , Constantinescu SN . Differential STAT5 signaling by ligand-dependent and constitutively active cytokine receptors . J Biol Chem . 2005 ; 280 : 13364 - 13373 . https://doi.org/10.1074/jbc. M407326200 PMID: 15677477
30. Morales-Mantilla DE , King KY . The Role of Interferon-Gamma in Hematopoietic Stem Cell Development, Homeostasis, and Disease . Curr Stem Cell Rep . 2018 ; 4 : 264 - 271 . https://doi.org/10.1007/ s40778-018 -0139-3 PMID: 30148048
31. Saha B , Jyothi Prasanna S , Chandrasekar B , Nandi D . Gene modulation and immunoregolatory roles of interferon gamma . Cytokine . 2010 ; 50 : 1 - 14 . https://doi.org/10.1016/j.cyto. 2009 . 11 .021 PMID: 20036577
32. Selleri C , Maciejewski JP , Sato T , Young NS . Interferon-gamma constitutively expressed in the stromal microenvironment of human marrow cultures mediates potent hematopoietic inhibition . Blood . 1996 ; 87 : 4149 - 4157 . PMID: 8639773
33. Zeng W , Miyazato A , Chen G , Kajigaya S , Young NS , Maciejewski JP . Interferon-gamma-induced gene expression in CD34 cells: identification of pathologic cytokine-specific signature profiles . Blood . 2006 ; 107 : 167 - 175 . https://doi.org/10.1182/blood-2005 -05-1884 PMID: 16131564
34. Vila-Coro AJ , Rodriguez-Frade JM , Martin de Ana A , Moreno-Ortiz MC , Martinez-A C , Mellado M. The chemokine SDF-1 triggers CXCR4 receptor dimerization and activates the JAK/STAT pathway . FASEB J . 1999 ; 13 : 1699 - 1710 . PMID: 10506573
35. Tsai JJ , Dudakov JA , Takahashi K , Shieh JH , Velardi E , Holland AMet al. Nrf2 regulates haematopoietic stem cell function . Nat Cell Biol . 2013 ; 15 : 309 - 316 . https://doi.org/10.1038/ncb2699 PMID: 23434824
36. Hasselbalch HC , Thomassen M , Hasselbalch Riley C , Kjaer L , Larsen TS , Jensen MK et al. Whole blood transcriptional profiling reveals deregulation of oxidative and antioxidative defence genes in myelofibrosis and related neoplasms. Potential implications of downregulation of Nrf2 for genomic instability and disease progression . PLoS One . 2014 ; 9: e112786 . https://doi.org/10.1371/journal.pone. 0112786 PMID: 25397683