Abnormal proplatelet formation and emperipolesis in cultured human megakaryocytes from gray platelet syndrome patients

The Gray Platelet Syndrome (GPS) is a rare inherited bleeding disorder characterized by deficiency of platelet α-granules, macrothrombocytopenia and marrow fibrosis. The autosomal recessive form of GPS is linked to loss of function mutations in NBEAL2, which is predicted to regulate granule trafficking in megakaryocytes, the platelet progenitors. We report the first analysis of cultured megakaryocytes from GPS patients with NBEAL2 mutations. Megakaryocytes cultured from peripheral blood or bone marrow hematopoietic progenitor cells from four patients were used to investigate megakaryopoiesis, megakaryocyte morphology and platelet formation. In vitro differentiation of megakaryocytes was normal, whereas we observed deficiency of megakaryocyte α-granule proteins and emperipolesis. Importantly, we first demonstrated that platelet formation by GPS megakaryocytes was severely affected, a defect which might be the major cause of thrombocytopenia in patients. These results demonstrate that cultured megakaryocytes from GPS patients provide a valuable model to understand the pathogenesis of GPS in humans.

platelet α -granules and spleen enlargement. Myelofibrosis was demonstrated in older animals 12 . Nbeal2 −/− mice were used to extensively study the role of platelet α -granules constituents in hemostasis, thrombosis, thrombo-inflammatory disease states, megakaryocyte survival and development, platelet production, tissue reconstitution after injury, development of myelofibrosis and cancer metastasis propagation. However, some differences among the three Nbeal2 −/− mouse strains were reported with regard to megakaryocyte development and differentiation, proplatelet formation and α -granules content. Therefore, whether Nbeal2 loss of function in mice affects megakaryopoiesis and/or proplatelet formation, and how it contributes to thrombocytopenia is unclear.
To investigate the impact of mutations in NBEAL2 gene on human thrombopoiesis, we enrolled four GPS patients, whose clinical features and NBEAL2 mutations have been previously described [13][14][15] . We obtained in vitro differentiated megakaryocytes from peripheral blood or bone marrow hematopoietic progenitor cells of the four GPS patients and five controls and evaluated megakaryocyte maturation and function and proplatelet formation.

Results
NBEAL2 mutations do not affect megakaryocyte differentiation by human hematopoietic progenitors. We differentiated human megakaryocytes in vitro starting from peripheral blood or bone marrow hematopoietic progenitor cells of GPS patients with mutated NBEAL2 and healthy controls. Hereinafter, the patient carrying the c.1253del, c.3584G> A and c.5720+ 1G> A (r.5720_5721ins148) mutations will be identified as #1; the patient carrying the c.5572C> T, c.6652G> T and c.7033C> T mutations will be identified as #2; the two patients carrying the c.2187C> A homozygous mutations will be identified as #3 (Table 1). After 14 days of culture, in vitro megakaryocyte differentiation and output of CD61 + CD42b + megakaryocytes were similar to those of healthy control samples (Fig. 1a,b). Further, maturation stages of CD61 + megakaryocytes, classified I to IV according to standard morphological criteria 16 , were also similar in GPS patients and controls (Fig. 1c). No differences were observed among patients #1, #2 and #3, thus indicating that all the analyzed NBEAL2 mutations did not affect either the differentiation or the maturation of megakaryocytes. The results from both peripheral blood-and bone marrow-derived megakaryocytes were perfectively comparable (not shown).
Reduced α-granule content in human megakaryocytes from GPS patients. Despite normal differentiation, three of the most abundant proteins normally contained in α -granules, von Willebrand factor, thrombospondin and P-selectin, were markedly reduced in in vitro differentiated human megakaryocytes from both #1 and #2 GPS patients compared to controls (Fig. 2a-c). No differences were observed between patients, thus suggesting that the storage of these proteins was compromised in GPS megakaryocytes regardless of the type of mutation. Consistently, #1 and #2 GPS-derived megakaryocytes stimulated with protease activated receptors (PARs)-activating peptides (APs) exposed less P-selectin compared to control megakaryocytes (Fig. 2d). Interestingly, these results are in agreement with those we have recently observed in platelets from peripheral blood of patients #1 and #2 15 , and demonstrate that defective P-selectin mobilization is a distinctive feature that gray platelets inherit directly from their progenitors megakaryocytes before their release into bloodstream.
Evidence of emperipolesis in human megakaryocytes from GPS patients. Bone marrow megakaryocytes from gray platelets syndrome patients display intact cells within their cytoplasm, as a consequence of a rare biological process that is named emperipolesis 6,15,17 . Emperipolesis, present in less than 2% of megakaryocytes in normal bone marrow biopsies, has been shown in 38% to 65% of megakaryocytes in GPS bone marrow biopsies 15 . However, whether this abnormality is directly dependent on NBEAL2 mutation rather than a consequence of compromised bone marrow homeostasis, associated with the progressive myelofibrosis, has never been investigated. We found that several megakaryocytes from #1 and #2 GPS patients displayed the presence of intact cells within their cytoplasm, likely as a consequence of emperipolesis, while this abnormality was not observed in any of the mature megakaryocytes from control subjects (Fig. 3). The presence of cells other than megakaryocytes in our in vitro culture conditions is not surprising given the presence of a mixed population of cells in the culture media before megakaryocyte population is enriched through an albumin gradient [18][19][20][21] , which could grow and be engulfed within the megakaryocyte cytoplasm during the course of differentiation. To the best of our knowledge, none of the previously published studies of megakaryocytes from patients with other forms of thrombocytopenia or healthy subjects, grown under similar experimental conditions, revealed the presence of intact cells in cultured megakaryocytes 20,[22][23][24] . Therefore, our study provides the first evidence that emperipolesis occurs in human GPS megakaryocytes as a consequence of an intrinsic cellular defect, rather than of abnormalities of the bone marrow microenvironment.
Calcium signaling is maintained in human megakaryocytes with mutated NBEAL2 gene. We have recently demonstrated that constitutively released adenosine diphosphate (ADP) by human mature megakaryocytes promotes the activation of Store-Operated Calcium Entry (SOCE) which in turn is responsible for the regulation of platelet formation and interaction with the components of the extracellular matrix 19,25 . We explored SOCE functionality in GPS megakaryocytes from #1 and #2 patients by depleting the stores with cyclopiazonic acid (CPA, 10 μM). The treatment evoked a transient rise in intracellular calcium (Ca 2+ ) concentration, because of passive emptying of endoplasmic reticulum, followed by influx of external Ca 2+ through activated SOC channels (Fig. 4a). Analysis of the extent of both Ca 2+ release and entry demonstrated no significant difference between control-and GPS-derived megakaryocytes (Fig. 4b). Consistent with this data, megakaryocytes from #1 and #2 GPS patients normally responded to ADP stimulation, with no significant differences in the extent of Ca 2+ mobilization compared to control megakaryocytes (Fig. 4c,d). All together, these data indicate that Ca 2+ homeostasis is normally maintained in GPS megakaryocytes.

Human megakaryocytes from GPS patients show dramatically impaired proplatelet formation.
An efficient Ca 2+ signaling is fundamental in order to guarantee proper megakaryocyte interaction with extracellular matrix components of the bone marrow environment 25 , which is in turn relevant for the control of platelet release 26,27 . Therefore, based on the demonstration of a normal Ca 2+ homeostasis in GPS megakaryocytes, we expected that they could normally interact with extracellular matrix components. Surprisingly, the analysis of GPS megakaryocyte interaction with type I collagen, an extracellular matrix component that promotes megakaryocyte spreading by supporting cytoskeleton contractility 27 , revealed abnormal actin stress fibers formation and microtubule assembly and, consequently, defective megakaryocyte spreading (Fig. 5). Similar results were observed in samples from #1 and #2 patients ( Fig. 5v-xii). Further, the analysis of proplatelet formation on fibronectin 28,29 , revealed that megakaryocytes derived from either #1, #2 or #3 GPS patients displayed shorter proplatelet branches compared to control megakaryocytes (Fig. 6a). Similar results were obtained from both peripheral blood-( Fig. 6ai-vi,viii-xiii) and bone marrow-derived megakaryocytes (Fig. 6avii,xiv). The percentage of proplatelet forming megakaryocytes was markedly reduced in diseased megakaryocytes compared to controls, while no significant differences were observed among #1, #2 and #3 samples (Fig. 6b). The tips, terminal ends of proplatelet branches, are the site of the assembly of nascent platelets 30 . Consequently, the number of platelets that are formed depends on both the percentage of proplatelets and the number of their bifurcations in order to ensure the highest number of tips per single megakaryocyte. Proplatelets from GPS megakaryocytes displayed abnormal architecture with significant reduction in the absolute number of bifurcations compared to healthy controls (Fig. 6c). These results were consistent with median platelet number of GPS patients, which was significantly lower in #1, #2 and #3 patients (56 × 10 3 platelets/μl, range: 30-65) than in healthy controls (294 × 10 3 platelets/ μl, range: 200-380). Moreover, we observed the presence of giant proplatelet tips and released platelets in cell cultures from #1 and #2 GPS patients (Fig. 7), consistent with the findings that these patients displayed a significant macrothrombocytopenia (mean platelet volume: 3.9 μm, range: 3.5-4.3) with respect to healthy controls (mean  platelet volume: 2.4 μm, range: 2-2.7). All together these data suggest that altered megakaryocyte interaction with extracellular environment, together with a severe defect in proplatelet formation and branching, may be the major causes of reduced platelet count and increased platelet size that characterize GPS.

Discussion
To the best of our knowledge, this the first report on megakaryopoiesis and proplatelet formation in cultured megakaryocytes from GPS patients carrying different autosomal recessive NBEAL2 mutations ( Table 1). The observed results were, therefore, independent of the mutations or of their location to a particular domain of the gene.  Control (CTRL)-(i-iv) and GPS-derived megakaryocytes (v-xii) at the end of differentiation were plated on type I collagen-coated cover-slips, at 37 °C in a 5% CO 2 atmosphere. After 16 hours adherent cells were fixed and stained for immunofluorescence analysis of the cytoskeleton components actin (red = TRTIC-phalloidin) and tubulin (green = α -tubulin). Nuclei were counterstained with Hoechst 33258 in blue (scale bar = 20 μm).
NBEAL2 mutations did not affect either the differentiation or the maturation of human megakaryocytes. Similar results were obtained in different patients, using different sources (peripheral blood and bone marrow) of hematopoietic progenitor cells to differentiate along the megakaryocytic lineage in culture. Our results are in  agreement with the clinical findings that, despite the presence of myelofibrosis, the number and the maturation of megakaryocytes are not abnormal in patients with GPS 2 . Nbeal2 −/− mouse models, to this regard, displayed different behaviors, which did not clarify the role of Nbeal2 in megakaryocyte differentiation [10][11][12] . In the bone marrow of Nbeal2 −/− mice, compared to control mice, the megakaryocyte number was increased in one strain 11 , slightly increased in older mice in another strain and similar to control mice in the third one 12 . Cultured megakaryocytes from bone marrow revealed normal differentiation and number of high ploidy megakaryocytes in one model 12 and reduced ploidy together with megakaryocytes arrested in the earliest stage of differentiation in another model 10 . Our findings in GPS patients show that NBEAL2 in humans does not affect the megakaryopoiesis, regardless of the type of mutations affecting the patients.
We also found a marked reduction of α -granule cargo and membrane proteins in mature megakaryocytes from #1 and #2 GPS patients compared to controls. These results are consistent with previous findings in the bone marrow biopsies of GPS patients 15,31,32 and in the bone marrow of two Nbeal2 −/− mouse models 10,11 . Moreover, they are in agreement with the demonstration that NBEAL2 is involved in the biogenesis/trafficking of granules in megakaryocytes [7][8][9] . Altogether these findings, although do not clarify whether the cause of reduced α -granule proteins is a hampered granule synthesis or a lack of retention of α -granules within the megakaryocytes, do confirm that the lack of α -granule proteins is present at the level of megakaryocytes (both culture-derived in vitro and in the bone marrow in vivo), and does not occur in platelets upon release of α -granules content into the circulation.
Emperipolesis was found in in vitro cultured GPS megakaryocytes from #1 and #2 patients. Emperipolesis, specifically the presence of leukocytes within the cytoplasm of megakaryocytes, has been already described as a prominent feature in bone marrow samples from GPS patients 17 , carrying both NBEAL2 15 and GFI1B mutations 6 , and in all three Nbeal2 −/− mice 10-12 . These consistent findings suggest that this is a distinctive feature of GPS independent of the genotype and might be crucial for the development of the platelet phenotype. Our results on cultured megakaryocytes demonstrate that emperipolesis is intrinsic to these cells, independent of influence from microenvironment and/or other cells. Although the mechanism that favors emperipolesis in GPS is still far to be elucidated, some evidences suggest a mechanism similar to that described in primary myelofibrosis in humans, involving the mislocalization of P-selectin on megakaryocyte membrane favoring the interaction with leukocytes through the PSGL-1 countereceptor and their trapping within the megakaryocyte cytoplasm 33,34 . Also, a proinflammatory phenotype of Nbeal2 −/− megakaryocytes overproducing and releasing chemoattractans has been hypothesized as the cause of this phenomenon 12 . Whatever the mechanism involved in this phenomenon, Guerrero et al. demonstrated that engulfment of cells within the megakaryocytes does not affect their survival and apoptosis 12 .
Ca 2+ homeostasis was normal in #1 and #2 GPS megakaryocytes. Conversely, we report for the first time a dramatic defect in proplatelet formation and megakaryocyte interaction with extracellular matrix. GPS megakaryocytes, from all studied mutations and from both peripheral blood and bone marrow samples, displayed shorter proplatelet branches, a significant reduction in the absolute number of their bifurcations and abnormal architecture of proplatelets when compared to control megakaryocytes. Moreover, the percentage of proplatelet forming megakaryocytes was markedly reduced compared to controls, independently of genotype, in agreement with results obtained in Nbeal2 −/− mice by Kahr et al. 10 . Unaltered proplatelet formation was reported in the other two Nbeal2 −/− mouse models 11,12 . We also found altered GPS megakaryocyte interaction with type I collagen, a component of bone marrow extracellular matrix responsible for control of proper platelet formation through regulation of cytoskeleton contractility. It is well known that the release of cellular fibronectin, stored in α -granules, by mature megakaryocytes adhering on type I collagen is the major determinant for the maintenance of the tight interaction between cells and the substrates, which in turn favors the cytoskeleton contractility 27 . Thus, impaired α -granule assembly may be the cause for the altered megakaryocyte contractility in GPS patients.
Defective proplatelet formation together with altered megakaryocyte interaction with extracellular environment may be the major causes of congenitally reduced platelet count in patients with GPS. Other factors partially contributing the worsening of macrothrombocytopenia with age are splenomegaly and myelofibrosis. However, spleen removal causes a mild increase of the platelet count 3 , thus suggesting that splenomegaly is not the major cause of thrombocytopenia. Myelofibrosis is progressive in GPS and can account for the progressive reduction of platelet count as patients become older, but not for thrombocytopenia during childhood, when myelofibrosis can be absent 3 . By showing in vitro reduced proplatelet formation in cultured megakaryocytes from GPS patients, we provide a new mechanism for the thrombocytopenia in these patients.
In conclusion, this is the first systematic analysis of human, cultured GPS megakaryocytes and represents a comparator to the extensive work done with the Nbeal2 −/− mice. Some differences and some similarities between NBEAL2 deficient megakaryocytes from GPS patients and Nbeal2 deficient megakaryocytes from mouse models are reported. We demonstrate that cultured human megakaryocytes from GPS patients with autosomal recessive mutation of NBEAL2 undergo normal differentiation and maturation and display, instead, severely impairment of both proplatelet formation and interaction with extracellular matrix. These two defects might be the major causes for the thrombocytopenia. We provide first evidence that emperipolesis, a distinctive feature of GPS, is a megakaryocyte-dependent phenomenon. Decreased content of α -granule proteins was found in cultured megakaryocytes, confirming that this phenomenon occurs at the level of megakaryocytes and not in circulating platelets. Altogether these findings demonstrated that in vitro cultured megakaryocytes represent a valuable tool in understanding human GPS and it might of potential use to dissect the function of NBEAL2 at the molecular and cellular level to determine which trafficking steps are affected in the disease.
Patients. Clinical and genetic features of patients 1 and 2, from Italy and of patients 3A and 3B (brothers), from Argentina, have been already reported [13][14][15] . Briefly, at the time of this study patients 1, 2, 3A and 3B were, respectively, 23,8,68 and 73 year-old. All patients had similar degree of macrothrombocytopenia, as reported in the result section, whereas leukocyte and red blood cells counts were not affected. All the details regarding the type of NBEAL2 mutations are reported in Table 1.
Blood (all patients) and bone marrow (patient 3A) samples were obtained with informed consent of the patients or of the parents in case of minor age. Four healthy volunteers were recruited as controls for peripheral blood samples. Bone marrow sample from one patient with Ewing sarcoma, who underwent bone marrow biopsy for staging of the disease and without bone marrow involvement, was used as a control for bone marrow sample. Cultures of peripheral blood and of bone marrow samples of patients and controls were run in parallel in each experiment. The study was performed in accordance with the principles of the Declaration of Helsinki. The study was approved by the ethical committees of Policlinico Agostino Gemelli, Rome, Italy and of Instituto de Investigaciones Médicas Alfredo Lanari, CONICET, Buenos Aires, Argentina, involved in the management of enrolled patients.
Differentiation of human megakaryocytes and morphological analysis. CD45 + from peripheral blood or CD34 + from bone marrow samples were separated by immunomagnetic bead selection (Miltenyi Biotec, Bologna, Italy) and cultured in Stem Span medium supplemented with 10 ng/mL thrombopoietin (TPO), interleukin (IL)-6 and IL-11 at 37 °C in a 5% CO 2 fully humidified atmosphere, as previously described 18,20,22 . At the end of the culture (14 th day) 150 × 10 3 cells were collected, cytospun on glass coverslips and stained with a primary antibody against CD61 (1:100) to evaluate megakaryocyte differentiation. Nuclei were stained with Hoechst 33258 (1:10000). CD61 + megakaryocytes were assigned to distinct stages of maturity according to standard morphological criteria 16 , as previously described 20,21,36 . Specifically, in the early maturation stage (I) megakaryocytes present the lowest cytoplasmic/nuclear ratio, compact nucleus and small size. Successive stages (II, III) are identified by progressive cytoplasmic mass increase and highly lobulated nuclei. Fully differentiated megakaryocytes (stage IV) present the highest cytoplasmic/nuclear ratio. For each specimen, at least 100 megakaryocytes were evaluated.
For the analysis of the α -granules content, at day 14 of the culture megakaryocytes were harvested, seeded onto glass coverslips, previously coated with Poly-L-Lysine, and allowed to adhere by centrifugation at 240 × g. Cells were then fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with anti-von Willebrand Factor (1:100), anti-thrombospondin (1:50) or anti-P-Selectin (1:100). Nuclei were stained with Hoechst 33258 (1:10000). The fluorescence intensity was quantified by Image J software. Values are reported as arbiter units (a.u.). For each specimen, at least 100 megakaryocytes were evaluated.
For the analysis of emperipolesis, after fixation, samples were stained with hematoxylin and eosin. For each specimen, at least 200 megakaryocytes were evaluated. Results are expressed as percentage of total cells analyzed.
For all experiments the coverslips were mounted onto glass slides with ProLong Gold antifade reagent (Invitrogen, Milan, Italy) and images acquired by Olympus BX51 microscope (Olympus, Deutschland GmbH, Hamburg, Germany).
Flow cytometry analysis. 200 × 10 3 cells were collected at the end of the culture (14 th day) and centrifuged at 250xg for 7 minutes. Cells were then suspended in phosphate buffer saline (PBS) and stained with a FITC-conjugated antibody against human CD61 and a PE-conjugated antibody against human CD42b, at room temperature, in the dark for 30 minutes. The number of CD61 + CD42b + megakaryocytes (megakaryocyte output) from GPS patients was calculated and evaluated with respect to those of healthy controls.
For analysis of P-Selectin exposure samples were stimulated with 25 μM of both PAR-1 and PAR-4 activating peptides for 20 minutes, at 37 °C, in presence of 2 μM CaCl 2 and 2 μM MgCl 2 . Cells were then stained with APC-conjugated antibody against human P-Selectin antibody for additional 15 minutes. In all described experiments, after incubation, samples were immediately analyzed by a Beckman Coulter Navios flow cytometer. Non-stained samples were used to set the correct analytical gating. Off-line data analysis was performed using Beckman Coulter Navios software package.
Evaluation of cell adhesion and proplatelet formation on fibronectin and type I collagen. In order to analyze megakaryocyte adhesion and proplatelet formation onto different extracellular matrix components, 12 mm glass cover-slips were coated with 25 μg/ml fibronectin or 25 μg/ml type I collagen, overnight at 4 °C, as previously described. At the end of the culture megakaryocyte population was enriched through a BSA gradient as previously described 20 . 1 × 10 5 megakaryocytes were harvested and allowed to adhere at 37 °C and 5% CO 2 . After 16 hours, adhering cells were visualized either by phase contrast images by an Olympus IX53 microscope (Olympus, Deutschland GmbH, Hamburg, Germany) or fixed in 4% PFA, permeabilized with 0.1% Triton X-100, and stained for immunofluorescence evaluation by an Olympus BX51 microscope (Olympus, Deutschland GmbH, Hamburg, Germany). Briefly, for immunofluorescence staining samples were stained with anti-α -tubulin antibody (1:700), TRITC-conjugated phalloidin (1:5000) and/or CD61 (1:100), as previously described 25,27 . Nuclei were stained with Hoechst 33258 (1:10000). The cover-slips were mounted onto glass slides with ProLong Gold antifade reagent (Invitrogen, Milan, Italy). At least 50 fields per sample were analyzed. Cytoskeleton reorganization was recognized in cells displaying microtubule and stress fibers assembly. Proplatelet forming-megakaryocytes were identified as cells displaying long filamentous structure ending with platelet-sized tips. The percentage of proplatelets from GPS patients was compared with respect to those of healthy controls.
Measurements of intracellular calcium concentration. The analysis of Ca 2+ flows was performed as previously described 25,37 . 12 mm glass coverslips were coated with 25 μg/ml fibronectin, overnight at 4 °C. Mature megakaryocytes were plated onto substrate-coated cover-slips in 24-wells plates (1 × 10 5 cells/well). After 60 minutes at 37 °C and 5% CO 2 , megakaryocytes were loaded with 4 μM fura-2 AM in PSS. After 30 minutes, the coverslip was fixed to the bottom of a Petri dish and the cells were observed using an upright epifluorescence Axiolab microscope (Carl Zeiss). Custom software, working in the LINUX environment, was used to drive the camera (Extended-ISIS Camera; Photonic Science) and the filter wheel and to measure and plot on-line the fluorescence from single cells enclosed in rectangular regions of interest (ROI). Intracellular Ca 2+ concentration ([Ca 2+ ] i ) was monitored by measuring, for each ROI, the ratio of the mean fluorescence emitted at 510 nm when exciting alternatively at 340 and 380 nm (shortly termed "RATIO"). An increase in [Ca 2+ ] i causes an increase in the ratio. The experiments were performed at room temperature.
Statistic. Megakaryocyte cultures were performed three times from peripheral blood of patient 1, two times from peripheral blood of patient 2, two times from peripheral blood and one time from bone marrow of patient 3A, one time from peripheral blood of patient 3B and each experiment was independently replicated at least 3 times, unless specified otherwise. Control samples from healthy volunteers (peripheral blood samples) and from one patient with Ewing sarcoma and no bone marrow involvement (bone marrow sample) were always performed in parallel. For all the experiments values are expressed as mean ± SD of all independent experiments. Student's t-test was performed for statistical analysis. Values of at least p < 0.05 were considered statistically significant.