Proplatelet generation in the mouse requires PKCε-dependent RhoA inhibition

Defective platelet numbers can compromise proper wound healing and cause bleeding, whereas excessive platelet production or activation can lead to thrombosis. As a result, platelet homeostasis is a tightly regulated process. The process of generating thousands of platelets from a single megakaryocyte is characterized by several morphologically distinct stages that involve extensive cytoskeletal remodeling. In particular, megakaryoctyes, following polyploidization and prior to platelet release, form elongated cellular processes called proplatelets (proPLT).¹  Perturbations in proPLT formation lead to alterations in platelet morphology and thrombocytopenia,^2,3  emphasizing the importance of this step.

The marginal microtubule band, actin-based cytoskeleton, and spectrin-based membrane skeleton are all prominent components of the cytoskeletal remodeling that occurs during platelet formation.⁴  Specifically, dynamic tubulin polymerization is required for the proper platelet morphology, assembly of the marginal microtubule coil, proPLT extensions and microtubule band formation.^4-8  Various mutations in genes that regulate actin polymerization and turnover cause thrombocytopenia due to altered formation and morphology of proPLT extensions, increased platelet size, and defective platelet release.^9,10  The spectrin cytoskeleton membrane influences platelet formation by supporting the megakaryocyte membrane invagination that is critical to the production of proPLT extensions.¹¹ 

In addition to these cytoskeletal pillars, the Rho/Rho-associated, coiled-coil containing protein kinase (ROCK) pathway regulates platelet production. Overexpression of a dominant-negative RhoA boosts proPLT formation, whereas constitutive RhoA activation abrogates this process.¹²  Moreover, blocking RhoA activity is necessary for microtubule-driven formation of proPLT cytoplasmic extensions.¹³  RhoA-deficient mice display macrothrombocytopenia, suggesting that RhoA activity must be temporally regulated for successful PLT production.¹⁴  There are implications that RhoA signaling influence platelet formation by regulating cytoskeleton components. For example, it has been demonstrated that the correlation between PLT size and changes in tubulin polymerization involves a RhoA/Rho-dependent regulation of cytoskeletal microtubule ring remodeling.^2,15  Furthermore, RhoA contributes to actin function during actomyosin contractility and stress fiber formation.¹⁶  Additionally, the Rho/ROCK pathway is related to the late failure of cytokinesis responsible for the endomitotic process in megakaryocytes.¹⁷ 

We recently showed that expression of protein kinase Cepsilon (PKCε), a member of the protein kinase C family of serine/threonine kinases, changes throughout the course of thrombopoietin (TPO)–induced megakaryocyte (MK) differentiation from human CD34^pos cells.^18,19  PKCε has been implicated in both the regulation of cytoskeleton remodeling and the RhoA pathway.²⁰  Specifically, PKCε, unlike other PKC family members, contains an F-actin-binding region that promotes the formation of F-actin filaments by preventing depolymerization and increasing the rate of actin filament elongation.²¹  During mitotic cytokinesis of several cell types (ie, COS-7, HeLa, MEF, HEK 293), PKCε assembles with 14-3-3zeta, and together they transiently accumulate at the actomyosin ring with RhoA and is required to dissociate both itself and RhoA from the midbody.²²  In neurons, PKCε promotes neurite outgrowth, which can be blocked by forced activation of the ROCK/RhoA pathway, indicating that PKCepilson may attenuate RhoA/ROCK pathway activity in the process.²³ 

On these bases we hypothesized that PKCε regulates proPLT formation by modulating RhoA activity. Indeed, our data show that PKCε shut down results in RhoA expression levels that are incompatible with normal PLT generation.

Mouse fetal liver cells (FLC) were isolated from embryos at day 13.5. Single-cell suspensions were prepared as previously described²⁴  and cultured for 4 days in Dulbecco’s modified Eagle medium (Invitrogen, Carlsbad, CA), supplemented with 10% fetal calf serum, 2 mmol/L l-glutamine, 50 U/mL penicillin, and 50 µg/mL streptomycin at 37°C and 5% CO₂ in the presence of 0.1 µg/mL purified recombinant mouse TPO.²⁵  All experiments were complied with institutional guidelines approved by the Children’s Hospital Animal Care and Use Committee and the Institutional Animal Care and Use Committee.

FLC cultures were layered on a single-step gradient (1.5% to 3.0% bovine serum albumin [BSA]) at different times of differentiation (days 2, 3, 4, 5), and MKs were allowed to sediment for 30 minutes. For proPLT production analysis, the isolated MKs were resuspended in fresh media and cultured for an additional 24 hours, during which proPLT production was readily observed. For some experiments, MK cultures were layered on a second single-step gradient (1.5% to 3.0% BSA) on day 5. MKs were allowed to sediment for 30 minutes, during which intermediate stages in PLT production were resolved within different fractions of the gradient.²⁶  Round MKs (rMK) sedimented on the bottom, and proPLT-forming MKs (ppfMK) localized to the BSA fraction, whereas released proPLTs and individual PLTs remained in the top fraction.

Mouse MKs, culture intermediates, and PLTs were purified and probed as previously described.²⁷  In brief, samples were fixed in 4% formaldehyde, centrifuged onto 1 µg/mL poly-l-lysine–coated coverslips, permeabilized, and blocked⁸  before antibody labeling. Samples were incubated with mouse monoclonal antibodies against α and β tubulin (Sigma-Aldrich, St. Louis, MO) and a rabbit monoclonal antibody against PKCε (Novus Biologicals, Littleton, CO). They were treated with secondary goat anti-mouse and goat anti-rabbit antibodies, conjugated to an Alexa Fluor 568 and to an Alexa Fluor 488 (Invitrogen), respectively; 4,6 diamidino-2-phenylindole (DAPI) (Sigma-Aldrich) was used to label the MK nucleus. Samples were examined with a microscope (Axiovert 200, Carl Zeiss, Inc.,Oberkochen, Germany) equipped with a 63× numerical aperture 1.4 oil immersion objective. Images were obtained using a charge-coupled device (CCD) camera (Hamamatsu Photonics, Hamamatsu, Japan) and analyzed using the MetaMorph image analysis software (MDS Analytical Technologies, Sunnyvale, CA).

Further images were acquired with a confocal microscope Olympus FluoView™ FV10i (Center Valley, PA), with a 60× water corrected objective. Series of x-y images were collected along the z-axis at 0.1- or 0.5-µm intervals. Three individual images were averaged to generate a representative image for each sample plane. At least 7 averaged images were collected at 0.5-µm steps along the z-axis of each sample for both fluorescence channels. Image analysis was generated using ImageJ software (National Institutes of Health, Bethesda, MD).

Cell perimeter and diameter were analyzed using the MetaMorph software thresholding, integrated morphometry analysis, and software calipers tools and by ImageJ software. Experiments represent 1 standard deviation about the mean for at least 3 independent cultures.

For short hairpin RNA (shRNA)–based gene silencing, pLKO.1 lentiviral vector encoding shRNA against mouse PKCε (NM_011104; shRNA59 and shRNA60) were obtained from Open-Biosystem (Thermo Scientific, Waltham, MA). As control (shRNACT), we used the MISSION pLKO.1-puro Non-Target shRNA Control Plasmid, containing a shRNA insert that does not target any known genes from any species (Sigma-Aldrich). The shRNA expressing viruses were produced in 293TL cells according to standard protocols. Mouse FLCs were infected at day 0 of culture and then cultured with TPO in the presence of puromycin, to select infected, puromycin-resistant cells.

PKCε was downmodulated in primary MK progenitors infecting FLCs with specific shRNA at day 0 of culture. Infected cells were cultured with TPO in the presence of puromycin, for cell selection. At day 4 of culture, infected rMKs were isolated by BSA gradient and cultured for an additional 24 hours in the presence of TPO and puromycin. At day 5 of culture, cells were analyzed for proPLT formation by light microscopy and then fixed for immunofluorescence staining.

Aliquots from each culture were collected for western blot analysis.

At day 4 of culture, infected rMKs were isolated by BSA gradient and cultured for an additional 24 hours in the presence of TPO and puromycin, with or without C3 transferase RhoA inhibitor (Cytoskeleton, Inc., Denver, CO). At day 5 of culture, cells were analyzed for proPLT formation by light microscopy and then fixed for immunofluorescence staining, and a fixed volume of culture supernatants were collected for PLT count by flow cytometry.

PLT production in culture was quantified as previously described.^19,28  Briefly, fixed volumes of the culture supernatants were collected, incubated with anti-CD41-RPE and Calcein AM (Sigma-Aldrich) (to exclude fragments), and added to a fixed volume of calibration beads (DAKO, Glostrup, Denmark) (size of beads of known concentration). CD41^pos/Calcein AM^pos PLTs were quantified by flow cytometry and normalized to bead counts to determine absolute PLT counts.

Cultured cells were counted, and 1.5 × 10⁶ cells were collected on days 0, 2, 3, 4, and 5, washed in phosphate-buffered saline, and centrifuged at 200 g for 10 minutes. Pellets were resuspended in a cell lysis buffer (50 mM Tris-HCl, pH 7.4; 1% NP-40; 0.25% sodium deoxycholate; 150 mM NaCl; 1 mM EDTA; 1 mM phenylmethylsulfonyl fluoride; 1 mM Na₃VO₄; 1 mM NaF), supplemented with fresh protease inhibitors, and protein concentration was determined using a Pierce BCA protein assay kit (Thermo Scientific). Thirty micrograms of proteins from each sample were run on 4% to 20% sodium dodecyl sulfate–acrylamide gels (Bio-Rad, Hercules, CA), blotted onto nitrocellulose membranes, blocked, and incubated with specific primary antibodies diluted as described in manufacturers’ protocols. Specifically, rabbit polyclonal anti-PKCε antibody (catalog number: 06991; Merck Millipore, Darmstadt, Germany) was used at the concentration of 1 μg/mL. Mouse monoclonal anti-PKCδ antibody (BD Pharmingen, San Jose, CA) was diluted 1:500, rabbit polyclonal anti-PKCθ antibody (Cell Signaling, Danvers, MA) was diluted 1:1000, rabbit polyclonal anti-PKCα antibody (Cell Signaling) was diluted 1:1000, and rabbit polyclonal anti-14-3-3zeta antibody (catalog number: AB9746; Merk Millipore) was diluted 1:500. A mouse monoclonal anto-RhoA antibody (Cytoskeleton, Inc.) was diluted 1:500, and mouse monoclonal anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (catalog number: MAB374; Merk Millipore), 1:5000 dilution) used as a protein-loading control.

Membranes were washed and incubated for 1.5 hours at room temperature with 1:5000 peroxidase-conjugated anti-rabbit or with 1:2000 peroxidase-conjugated anti-mouse IgG (Pierce) at room temperature and resolved by ECL Supersignal West Pico Chemiluminescent Substrate detection system (Pierce). Protein densitometric analyses from western blot assays were performed by using the Image J software system, normalized for GAPDH expression levels.

Statistical analyses were performed by using analysis of variance (ANOVA) and Tukey tests.

In agreement with our previous studies in human megakaryocyte coltures,¹⁸  PKCε protein expression increases during the early phases of MK differentiation (Figure 1A). However, unlike human megakaryopoiesis, mouse PKCε remains highly expressed throughout the entire maturation process, including PLT release (relative densitometric measurements of the samples tested are shown in Figure 1B).

Both morphological and confocal immunofluorescence analyses of differentiating MKs show that progressively enlarged, multinucleated cells stably express PKCε and α/β-Tubulin, the main cytoskeletal protein comprising the marginal tubular coil,⁵  and that PKCε colocalizes with α/β-Tubulin dimers at the marginal tubular coil of proPLTs (Figure 1C). Indeed in the cytoplasm, PKCε staining appears diffuse, with no evident overlapping with α/β-Tubulin. On the contrary, a strong overlap of PKCε with α/β-Tubulin is present along all the proPLT extensions and the marginal tubular coils, as shown in the composite panel of Figure 1C.

We then sought to determine whether PKCε expression was necessary for proplatelet production. Therefore we used recombinant lentiviral vectors to introduce and stably express shRNA molecules that specifically target PKCε into FLC at day 0 of culture. Analysis of megakaryocyte cultures at day 5 postinfection revealed that abrogation of PKCε was specific, because it did not modify either the expression of the other novel PKC isoforms or that of PKC α (Figure 2A-B). Morphological analyses of isolated megakaryocytes revealed that inhibition of PKCε expression resulted in the production of more aberrant proPLTs in comparison with control cultures (Figure 2C). The shRNA60 was more effective than shRNA59 in proPLT inhibition. However, although few residual branched protrusions could still be observed, proPLTs generated by both shRNA-infected cells were characterized by few abortive branches with some residual PKCε staining and virtual absence of the discoid structures. On the contrary, shRNACT proPLTs, characterized by the presence of discoid structures, expressed PKCε both in the cytoplasm and in the marginal tubular coil, where it appears particularly bright (Figure 2D). Moreover, we observed a greater than 70% reduction in platelet numbers in cultures in which we introduced PKCε shRNAs, and the remaining released PLTs were bigger than normal with a blurred tubulin-formed marginal tubular coil (Figure 2D insert), consistent with a defective cytoskeletal organization. Similar results were observed also by the inhibition of PKCε expression using specific small interfering RNA (siRNA) (supplemental data found on the Blood Website).

In certain cell types, PKCε cooperates with the 14-3-3zeta isoform to inhibit RhoA activity and drive cell division.²²  RhoA levels are also downmodulated in the later phases of thrombopoiesis to allow MK polyploidization²⁹  and cytoskeletal rearrangement.³⁰  Western blot analyses of TPO-treated FLCs over the course of MK differentiation revealed that active and cleaved RhoA levels diminish at day 5 post–TPO stimulation, when rMK start producing proPLTs. On the contrary, the expression of both PKCε and 14-3-3zeta escalates within the first 2 days of TPO induction and then remains stable for the duration of MK differentiation (Figure 3A-B).

Next, we examined whether modulating PKCε expression impacted RhoA activity during proPLT formation. Though abrogation of PKCε expression in primary MK precursors did not affect 14-3-3zeta levels, we did observe a substantial increase of RhoA expression (Figure 3C). Figure 3D shows the cumulative densitometric data from 4 unrelated experiments.

Finally, we set out to test whether the proPLT reduction mediated by PKCε inhibition was due to inappropriately high RhoA expression and activity. Therefore we analyzed the proPLT formation from PKCε knock-down and control samples in the presence or absence of a pharmacological RhoA inhibitor (Figure 4A,). As shown in Figure 4B, using a well-established flow cytometric method,^19,28  we also counted the absolute number of platelets in the culture medium. Flow cytometric analysis of functional platelets—positive for surface expression of CD41 and Calcein AM—showed that pharmacological inhibition of RhoA dramatically increased PLT formation and release in both control and PKCε-depleted samples (Figure 4B; a2 vs b2, c2 vs d2, and e2 vs f2; Figure 4C). Additionally, RhoA inhibition restored normal proPLT morphology in cells depleted of PKCε expression (Figure 4D-F). Figure 4F shows the size distribution within the platelet populations. Ninety-five percent of the shRNACT PLTs had a diameter smaller than 3.8 μm, whereas only 42% of the shRNA59 and 35% of the shRNA60 showed a diameter comparable to that of the control. Instead, in the presence of the RhoA inhibitor, 80% of shRNA59 and 75% of shRNA60 PLT populations returned to a size similar to that of the control.

Collectively, these data show that PKCε downregulates RhoA activity to promote proPLT formation and platelet release.

The process of platelet formation from megakaryocytes is characterized by distinct morphological stages that require extensive cytoplasmic remodeling. We had previously demonstrated that PKCε levels are specifically modulated by TPO in order to drive human hematopoietic cell precursors to PLT-producing MK.^18,31  We now show here that PKCε expression, which is virtually absent in mouse FLC, progressively increases in TPO-treated cells, and its levels remain high also in ppfMK and isolated proPLT. In proPLT, PKCε colocalizes with α/β-tubulin, which forms the marginal tubular coil, suggesting a functional correlation with the cytoskeleton reorganization events that accompany MK diffentiation and proplatelet formation. In support of this hypothesis, we observed that downregulation of PKCε expression by specific siRNAs resulted in lower proplatelet numbers and larger diameter platelets in culture. Because the efficiency of siRNA transfection is not homogeneous, we used recombinant lentiviral vectors that coexpress PKCε-targeting shRNAs and a puromycin-resistant cassette, thus allowing for the enrichment of PKCε-depleted cells. As expected, we detected a strong numerical reduction of proPLT formation with the remaining proPLTs lacking the typical organization in connected discoid structures.

It is well known that PKCε has a key role in neurite outgrowth,²³  a process that shares aspects with proPLT formation, such as the involvement of the RhoA pathway²³  which in turn is regulated by the complex PKCε/14-3-3zeta during MK polyploidization.²²  For this reason, we subsequently asked whether PKCε regulates proPLT formation and PLT release via inhibition of RhoA.

Indeed, we found that the active form of RhoA was downregulated at the terminal phases of MK differentiation, just before proPLT production, whereas PKCε and 14-3-3zeta expression did not change. PKCε knockdown by specific shRNA resulted in a significant increase of the active RhoA expression, along with hampered proPLT formation and PLT production. Accordingly, PKCε-depleted precursors treated with the pharmacological RhoA inhibitor generated normally shaped proPLT, with long cytoplasmic extensions characterized by a network of discoid structures, and restored PLT generation both in terms of number and size. Pleines et al¹⁴  have previously described the effects of RhoA knockdown on PLT size, showing that RhoA^−/− mice are macrothrombocytopenic with high ploidy MKs. Given the role of RhoA in a plethora of signaling pathways, the RhoA^−/− mouse might well be considered as a good indicator of the relevance of RhoA in megakaryopoiesis, but probably is not ideal for assessing the impact of compensatory mechanisms that represent unpredictable variables in a highly specific setting as proPLT formation. Our data show that inhibiting RhoA specifically at day 4 (preterminal phase) of proPLT formation carefully avoids perturbations of the earlier phases of cell commitment and differentiation.

These data show for the first time that PKCε has a key role in mouse PLT formation, by modulation of RhoA activity and interaction with the marginal tubular coil. These results represent a valid model for unraveling the molecular regulation of PLT generation. In fact, though PKCε is not expressed in late human megakaryocytopoiesis and its forced expression impairs PLT generation¹⁸  and function,³²  it must be considered that human—but not mouse—platelets express other novel PKC isoforms (namely PKCδ). The goal now will be to understand whether these 2 PKC isoforms parallel in the 2 species, with obvious consequences on our approach to human PLT disorders such as thrombocytopenia or thrombocytemia.

The authors are grateful to Matthew Devine, Luciana Cerasuolo, Vincenzo Palermo, Domenico Manfredi, and Davide Dallatana for technical support.

This work was supported by Regione Emilia-Romagna Area 1—Strategic Program 2010-2012, and Fondo per gli Investimenti per la Ricerca di Base-accordi di programma 2010 (Italian-Ministry of the University and Scientific and Technological Research/Ministry of Education, University and Research, Ministero dell'Istruzione, dell'Università e della Ricerca), RBAP10KCNS_002.

Contribution: G.G., P.M., M.V., and J.E.I. designed the experiments and interpreted the data; C.C., G.G., E.M., F.F., and A.N. performed the experiments; J.N.T., J.E.I., and S.M.S. revised the manuscript; M.V. and G.G. wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Marco Vitale, Department of Biomedical, Biotechnological, and Translational Sciences, Operative Unit of Anatomy and Histology, University of Parma, Ospedale Maggiore, Via Gramsci, 14, I-43126 Parma, Italy; e-mail: marco.vitale@unipr.it.