Regulation of platelet biogenesis: insights from the May–Hegglin anomaly and other MYH9-related disorders

Authors


Ramesh A. Shivdasani, Dana-Farber Cancer Institute, 44 Binney Street, Dana 720, Boston, MA 02115, USA.
Tel.: +617 632 5746; fax: +617 582 7198.
E-mail: ramesh_shivdasani@dfci.harvard.edu

Abstract

Summary.  Megakaryocyte (MK) maturation culminates in release of blood platelets through proplatelet extensions. MKs presumably delay elaborating proplatelets until synthesis of platelet constituents is complete. Recent insights from investigation of a classic human congenital macrothrombocytopenia, the May–Hegglin anomaly, and related MYH9-associated disorders shed new light on underlying mechanisms. The findings reviewed in this article implicate myosin IIA, the non-muscle myosin heavy chain product of the MYH9 gene, in restraining proplatelet formation until MKs achieve terminal maturity. Loss of myosin IIA function, through dominant inhibitory mutations in humans, targeted gene disruption in mice, or manipulation of cultured MKs, seems to accelerate proplatelet formation. The resulting process is inefficient and produces platelets that vary widely in size, shape and content. Several lines of evidence suggest that the Rho-ROCK-myosin light chain pathway restrains proplatelet formation through myosin IIA. These findings illustrate that mammalian thrombopoiesis is complex and subject to both positive and negative regulation.

Brief history

The year 2009 marks the centennial of May’s original description of a patient with giant platelets and leukocytic inclusions [1], a rare congenital disorder that Hegglin characterized further in 1945, when he recognized the familial inheritance and likened the inclusion bodies to those that Döhle had described more than 30 years earlier [2]. In the eponymous tradition of the early 20th century, this rare platelet disorder carried the name May–Hegglin anomaly, a fitting moniker for a condition with mild clinical manifestations and curious cellular correlates. Although sub-clinical symptoms probably lead to significant under-reporting, about 100 affected families are recorded and reveal an autosomal dominant pattern of inheritance. Typical signs include mild petechiae or ecchymoses and an exaggerated tendency to bleed during surgery. Clinical diagnosis relies on the hallmark triad of thrombocytopenia, large platelets, and leukocyte Döhle bodies. Families with additional clinical manifestations were for many years classified differently: macrothrombocytopenia combined with nephritis and sensorineural deafness constitutes the Epstein syndrome [3] and further addition of presenile cataracts the Fechtner syndrome [4]. The May–Hegglin anomaly can co-occur with features of Alport syndrome and leukocytic inclusions have subtly different morphology in the Sebastian syndrome [5]. Hereditary macrothrombocytopenia is the common thread in all these conditions.

The obvious heritability prompted a substantial effort to identify the responsible gene. In 2000, three groups implicated MYH9 [6–8], the gene encoding one isoform of non-muscle myosin IIA heavy chain, a motor protein with important functions in cell motility, cytokinesis, and cell shape [9–11]. In a satisfying resolution of related but seemingly diverse disorders, gene sequencing revealed MYH9 mutations not only in patients with the May–Hegglin anomaly but also in those with the Epstein, Fechtner, Sebastian, and variant Alport syndromes. These conditions are now all classified under the umbrella of MYH9-related disorders, a unified nosologic entity whose clinical continuum ranges from mild macrothrombocytopenia to a severe form complicated by hearing loss, cataracts, and renal failure [12]. Detailed genotyping has failed to identify an invariant clinical association with any of the 27 known MYH9 mutations spread across different regions of the coding sequence. However, mutations in the motor domain, especially amino acid residue 702, cause severe thrombocytopenia and early onset of nephritis and deafness, whereas mutations in the tail domain confer higher platelet counts and reduced risk of other complications [13]. The basis for such heterogeneity remains unknown.

Platelet and megakaryocyte studies in MYH9-related disorders

MYH9-related disorders are characterized by reduced platelet numbers with varying proportions of large platelets and consequent heterogeneity in mean platelet volume [14]. Although platelet function tests also give inconsistent results, reflecting small sample size and large individual variation, investigators agree that they function quite normally, except for slightly reduced adhesion to collagen I, and that platelet survival in the circulation is not compromised [14]. Biochemical analysis reveals altered resting composition and agonist-induced reorganization of the platelet cytoskeleton [15] and MYH9 aggregates within leukocyte inclusions [16], resulting perhaps from protein misfolding; however, these studies provide few insights into disease pathogenesis. If the bleeding tendency in affected individuals can be attributed largely to thrombocytopenia and the function of morphologically defective platelets is overtly normal, then the primary defect is likely to be in the process of platelet assembly and release, a possibility hematologists have considered since the early 1970s [17].

At the end of a long course of maturation, MKs expend their entire cytoplasm into long, beaded projections known as proplatelets; nascent blood platelets are assembled at the distal tips of these dynamic structures [18]. Patients affected by MYH9-related disorders have normal numbers of bone marrow MKs, implying that MK differentiation from multilineage progenitors remains unaffected. Although the aggregate impression from ultrastructural analysis of mature MKs is one of impaired cytoplasmic organization, the functional link between these distortions and platelet defects remains unclear. In wild-type mice, myosin-IIA is distributed uniformly throughout the MK cytoplasm, without obvious association with recognized structures or apparent sub-cellular organization [19]. Platelet formation is notoriously difficult to monitor in human samples; furthermore, MYH9-related disorders are rare and patients do not require repeated bone marrow evaluation. Thus, the extent and nature of proplatelet defects are uncertain, although the presence of significant platelet numbers, many with normal morphology, suggests that essential elements of proplatelet formation are probably intact.

Genetically engineered mice frequently inform the study of inherited human disorders. Conventional Myh9 knockout mouse embryos cannot implant in the uterus and accordingly die very early [20], severely limiting investigation of the role of myosin IIA in thrombopoiesis. Gachet and colleagues recently circumvented this limitation by engineering conditional disruption of the mouse Myh9 gene in the MK-platelet lineage. These mice have 70% fewer platelets than normal and show wide variation in platelet size, shape and content of rough endoplasmic reticulum, replicating key features of the May–Hegglin anomaly; they are also defective in clot retraction and outside-in signaling [21,22]. As described below, we used MKs differentiated from conventional Myh9−/− mouse embryonic stem (ES) cells to address the basis of defective thrombopoiesis [19]. Surprisingly, both approaches suggested that myosin-II deficiency accelerates rather than hampers proplatelet formation, leading apparently to premature release of platelets before MKs complete synthesis and proper organization of all the necessary components. Such observations are starting to yield glimpses into the basis for MYH9-related platelet disorders and, although much remains to be learned, also into critical aspects of normal platelet biogenesis.

In summary, thrombocytopenia in the MYH9-related disorders probably reflects an end-stage defect in mature MKs, perhaps akin to the poorly understood role of the surface receptor GPIb in another inherited macrothrombocytopenia, the Bernard–Soulier syndrome [23]. Albeit rare, such genetic disorders can provide vital insights into normal thrombopoiesis; these in turn can be applied to manipulate the process in efforts to alleviate reduced platelet numbers from diverse causes.

Role of myosin IIA in MK maturation and thrombopoiesis

The myosin-IIA complex contains two heavy chains that interact with actin to generate intracellular force and two pairs of light chains that regulate motor activity. The N-terminus of the heavy chain carries ATP-binding sites and a catalytic domain that generates energy for motor activity; the C-terminus contains sequences responsible for dimerization and interface with other binding partners. As an abundant motor protein and the only representative of its family within MKs [24], myosin-IIA is an ideal candidate to mediate the extensive cytoskeletal reorganization and cellular morphogenesis that accompany proplatelet extension and platelet release. Although actin–myosin motor function has been studied in other cell types, yielding general insights, the particulars of activity in MKs and of mutations specific to the MYH9-related disorders are just starting to be understood. Paracrystal morphology, judged by negative-stain electron microscopy, offers one model system for myosin-II assembly into bipolar filaments [25]. Franke and colleagues recently showed that wild-type tail fragments generate ordered paracrystal arrays, whereas mutant fragments aggregate aberrantly and disrupt assembly of wild-type filaments. In their studies, mutations in the MYH9 rod domain interfered with lateral intermolecular associations and high-order assembly, correlating with the dominant effects that the same mutations seem to exert in patients [25]. Such studies serve as an important adjunct to cell-based analyses and will ultimately provide a structural basis for myosin-IIA function in MKs.

We initially assumed that myosin-IIA motor activity might drive cytoplasmic reorganization during proplatelet formation (PPF) and that loss of function would either arrest or significantly impair the process. It therefore came as a surprise when MKs differentiated in vitro from Myh9-null ES cells elaborated proplatelets earlier and more profusely than their wild-type counterparts, providing the first hint that myosin IIA might function to inhibit rather than promote PPF [19]. Confirming this idea, blebbistatin, a specific inhibitor of myosin ATPase activity, also enhanced PPF in MKs cultured from mouse fetal liver [19]. Finally, MKs cultured from mice with conditional depletion of myosin-IIA are 2–3 times more efficient at producing proplatelets and these structures show large bulbous tips that are vaguely reminiscent of dysmorphic platelets in the human disorder [22]. These observations prompted us to revise our view on the potential role of myosin-IIA in thrombopoiesis, from an enabler of thrombopoiesis to an inhibitor, but they also begged the question of why MKs adopt such a baroque and counter-intuitive mechanism.

We believe the answer may lie in the likelihood that MKs must avoid platelet release until there is abundant and proportional synthesis of the many molecules and organelles that need to be packaged correctly within a platelet. We propose that because premature assembly generates fewer and defective platelets, a myosin-IIA-mediated pathway restrains thrombopoiesis until the appropriate time. One corollary of this model is that prematurely released platelets evade adequate control over their final size, shape and contents. Indeed, like platelets seen in the human disorder, those derived from Myh9 mutant mouse MKs are markedly heterogeneous in size and content, with a mixed population of elliptical and spheroid cells, some excessively and others sparsely loaded with organelles [22]. Our model suggests that thrombopoiesis may be impaired by both gain and loss of function mutations in MYH9. Loss could trigger PPF prematurely, whereas activating mutations may limit the quantitative scale of platelet production.

Spatial considerations may also apply. Most mature MKs in the bone marrow lie in intimate contact with the sinusoidal endothelium [26], an observation we recently corroborated with live fluorescence imaging in mice using intravital two-photon microscopy [27]. Explaining their observations in a mutant mouse model of Wiskott–Aldrich thrombocytopenia, Sabri and colleagues recently postulated that platelet release occurs ectopically in marrow interstitium instead of the sinusoidal vasculature [28]. Thus, MK migration into the vicinity of bone marrow sinusoidal endothelium seems to represent a critical step, possibly providing a facilitating environment or even the signal to lift myosin IIA-imposed restraint on PPF. This admittedly speculative argument raises the possibility that myosin IIA responds to critical signals in the bone marrow microenvironment that suppress its function and thereby trigger timely PPF. Many elements of this pathway remain untested and the truth is likely complex. We and others are, however, attempting to connect the dots in a way that sheds light on normal and abnormal thrombopoiesis.

Elucidation of a Rho-ROCK-MLC-myosin IIA pathway in platelet assembly and release

Myosin IIA motor activity is determined mainly by phosphorylation of its regulatory light chains (MLC). Phosphorylation at positions 18 and 19 enables both heavy chain binding to actin filaments and motor activity. Accordingly, expression of a phosphomimetic D18D19 mutant MLC leads to gain of myosin IIA function. In our experiments, forced DD-MLC expression in mouse hematopoietic progenitor cells had no effect on MK specification or early differentiation but completely arrested PPF, providing direct support for the idea that myosin IIA regulates thrombopoiesis by restricting PPF [19]. A converse mutation, A18A19-MLC, which resists phosphorylation and should compete with endogenous MLC, released the inhibitory activity and greatly enhanced PPF.

Among the three common kinases known to phosphorylate MLC, p21 activated kinase, MLC kinase and ROCK, our published studies suggest that the latter is the most relevant to myosin IIA function in restraining PPF [19]. Both ROCK and its immediate upstream activator, the small GTPase Rho, stimulate myosin IIA activity and attenuate PPF. Extending these in vitro observations, we used mouse bone marrow transplantation to demonstrate that activation of the Rho-ROCK-MLC-Mysoin IIA pathway reduces platelet production in vivo. Independent evidence that Rho and ROCK inhibit PPF in cultured human MKs supports this idea [29].

MYH9-associated disorders show autosomal dominant inheritance, and because mutations are confined to a single allele, the platelet defects must reflect either gene haploinsufficiency or dominant inhibitory effects of the mutant allele. As MKs derived from Myh9+/− mice or ES cells extend proplatelets normally, gene dosage is unlikely to be a limiting factor in humans, but may explain some truncation mutants. Our D18D19-MLC mutant, which activates myosin-IIA, abrogated PPF in a Myh9-dependent manner [19] and, together with the paracrystal studies cited above [25], hints that the major MYH9 mutations exert a dominant-negative effect. Taken together, these findings indicate that myosin IIA normally provides a brake on thrombopoiesis; mutations in the MYH9-related disorders act dominantly to release this brake, probably by interfering with native actin–myosin complexes in the cell periphery.

Insights into platelet biogenesis previously accrued largely from study of positive regulators such as the cytokines thrombopoietin and interleukin-6, transcription factors GATA1, Fli1, Runx1 and NF-E2, and structural proteins like β1 tubulin, whereas pharmacologic agents helped decipher the central role of the microtubule and actin cytoskeletons (reviewed in [30]). The Rho-ROCK-MLC-myosin IIA pathway illustrates negative regulation and highlights complexities in the process of platelet assembly and release.

Extracellular signals for proplatelet formation: closing the loop?

The complement of extraneous signals that initially restrain and subsequently trigger PPF is poorly understood; indeed, several may act individually or collectively to impose and lift temporal repression through myosin IIA and other pathways. Activation of integrin α2bβ1 inhibits PPF and requires intact ROCK activity for this effect [31]; this result implicates the marrow microenvironment and signaling by a cell surface receptor in limiting PPF. Another such signal may be Stem cell-derived factor 1 (Sdf-1, also called CXCL12), a known MK chemoattractant [32,33] and an extracellular ligand that signals through Rho [34]. We reported that mature MKs reduce cellular Rho activity in response to Sdf-1 [19], a change that would reverse inhibition of PPF by ROCK and phospho-MLC; Sdf-1 has also been shown to enhance PPF in cultured human MKs [32]. Sdf-1, which is enriched in marrow sinusoids [27], is ideally positioned to attenuate MK Rho activity and hence drive intracellular events that trigger PPF. We are currently investigating this and other related possibilities, to build on the premise that macrothrombocytopenia in MYH9-related disorders reflects precocious PPF because a physiologic barrier to this process is breached when myosin IIA activity is diminished.

Acknowledgments

Our work on megakaryocytes and platelet biogenesis is supported by grant R01HL63143 from the National Institutes of Health.

Disclosure of Conflict of Interests

The authors state that they have no conflicts of interest.

Ancillary