Bleeding syndromes that arise through an inherited defect of platelet production constitute a heterogeneous group of rare platelet disorders of growing importance.1, 2 Some, including the Bernard-Soulier syndrome (BSS) and Wiskott-Aldrich syndrome (WAS), associate a low circulating platelet count with a deficiency in a known functional protein (Table 1). In others, platelet dysfunction has not been shown and the genetic cause lies in the inability of megakaryocytes (MK) to mature and to produce platelets in sufficient numbers. In congenital amegakaryocytic thrombocytopenia, there is an increased tendency towards the development of leukemia, while in others such as the Jacobsen syndrome, the defects extend outside megakaryocytopoiesis and interfere with the development and/or functioning of major organs. In many of these rare diseases, the low platelet count is accompanied by changes in platelet morphology including the presence of enlarged or giant platelets. The elucidation of the genetic basis of familial thrombocytopenias is providing basic knowledge of how MK develop from the pluripotent hematopoietic stem cell (HSC) under the influence of thrombopoietin (TPO) and other cytokines. This short review will mainly deal with the biology and genetics of inherited thrombocytopenias.
Over 30 years ago, a series of 145 subjects from Italy and the Balkan peninsula were reported to have what was termed Mediterranean macrothrombocytopenia.3 The diagnostic criteria included a moderately low platelet count (70,000–150,000/μL), increased mean platelet volume and mild bleeding. These patients mostly had autosomal dominant inheritance. A series of unrelated Italian families was subsequently studied by linkage analysis and mutation screening.4 In six of them, a heterozygous A156V missense substitution was identified in GPIBα while in eight of ten patients GPIb-IX density on platelets was at levels reduced to those of BSS heterozygotes. This is somewhat enigmatic, for BSS is classically a disorder with autosomal recessive inheritance and an increased percentage of large platelets in obligate carriers is not an absolute rule.5 It is possible that another as yet unidentified factor contributes to the Mediterranean macrothrombocytopenia phenotype.
DiGeorge or velocardiofacial syndrome
Although this disorder can show autosomal recessive inheritance, in most patients it is acquired. The phenotype is linked to a monoallelic chromosome 22q11.2 microdeletion. Phenotypic features include conotruncal cardiac abnormalities, learning disabilities, velopharyngeal insufficiency, immunodeficiency, facial dysmorphism and thymic hypoplasia. Studies on mutant mice suggest that a haploin-sufficiency of a single gene, TBX1 (encoding a T-box containing transcription factor), largely accounts for the phenotype.6 Surveys of patients with DiGeorge syndrome suggest that mild thrombocytopenia and platelets of increased size affect about 20% of patients.7 Adjacent to TBX1 is the GPIBB gene, and its deletion can give rise to BSS when accompanied by a pathological mutation on the second allele.8 Defining the factors that give rise to giant platelets in DiGeorge syndrome would help clarify the origin of enlarged platelets in inherited thrombocytopenias. Bleeding is not a noted characteristic of DiGeorge syndrome patients.
MYH9-related thrombocytopenia syndromes
May-Hegglin anomaly (MHA), Fechtner syndrome, Sebastian platelet syndrome and Epstein syndrome constitute a group of related disorders with autosomal dominant inheritance and giant platelets.1, 2, 9 Thrombocytopenia is often mild; bleeding is infrequent and rarely life-threatening. Myocardial infarction has been reported in elderly patients. Sometimes, however, thrombocytopenia can be very severe. Phenotypic variations depend on the association of giant platelets with variable combinations of Döhle-like bodies in leukocytes, nephritis, sensorineural hearing loss and cataracts. In Epstein syndrome, neither easily detectable leukocyte inclusions nor cataracts occur but giant platelets, kidney disease and deafness are all present.9, 10 Significantly, mature MK from bone marrow biopsies show distinct morphological abnormalities proving that the defect manifests prior to platelet production.11
Molecular studies have shown that defects in the MYH9 gene encoding the non-muscle myosin heavy-chain IIA (myosin-IIA)12, 13 characterize all four syndromes. The same mutations often recur within unrelated families, and although haplotype analysis can identify common ancestors, a widespread geographical distribution makes this unlikely in many cases. Somatic mutations have already been reported. Myosin is a hexameric enzyme composed of two heavy chains and two pairs of light chains. The heavy chain aminoterminus forms a globular head that binds to actin and ATP, has ATPase activity, and is required for motor activity.14 The C-terminal α-helical domain features a coiled-coil and a single rod-like tail that allows the molecules to polymerize into bipolar filaments. Myosin-IIA occurs in platelets, monocytes and granulocytes and in organs such as kidney, eye and ear.10, 15 Of importance is the fact that some mutations are repeated in different phenotypes suggesting that the diseases are not truly mono-genic. An example is an exon 20 mutation (D1424N) in the coiled coil found first in May-Hegglin anomaly but then in Fechtner syndrome and also in Sebastian platelet syndrome.13, 16 Another is an exon 16 mutation (R702C) that has been located in Epstein syndrome and Fechtner syndrome as well as a variant of Alport syndrome.16 Intriguingly, these mutations give rise to amino acid substitutions in, respectively, the rod and head domains of myosin-IIA, suggesting that MYH9-related disease is expressed independently of the site of the structural change. The question of the role of haploinsufficiency must therefore be asked.
Recent studies on mouse models imply that myosin-IIA complexes and their upstream signaling pathways regulate the timing of proplatelet formation.17 The latter was premature in MK derived from Myh9−/−embryonic stem cells, while raised expression of myosin-IIA activity or constitutive phosphorylation of regulatory myosin light chain (MLC) led to reduced proplatelet formation. MLC phosphorylation in MK is controlled by Rho-associated kinase (ROCK) and ROCK inhibition was shown to promote proplatelet formation.17, 18 In vivo, this could mean that a lifting of the restraints imposed on pro-platelet formation by the Rho-ROCK-MLC-myosinIIA pathway is the key trigger for proplatelet formation at the sinusoids. An extracellular stimulus for platelet production could be provided by stromal cell-derived factor (SDF-1 or CXCL12) a chemoattractant known to dampen Rho activity.
Pathologically decreased MLC phosphorylation in MKs in MYH9-related disease could slow MK migration towards the sinusoids as well as giving rise to premature proplatelet formation. Despite these advances, the reasons for phenotypic variability are still unclear. Do other genetic and/or environmental factors also intervene as suggested, for example, by studies on a large family with Fechtner syndrome in which five out of ten affected members showed no signs of renal lesions at the time of study.19 One possible explanation is that additional II-B and II-C myosin isoforms can compensate for the malfunction of defective II-A isoform when expressed in the same tissues.15 The potential role of fibulin-1, encoding an extracellular matrix protein as a disease modifier, has recently been underlined.20
Type 2B von Willebrand disease (Type 2B VWD)
In this disorder, bleeding occurs through the spontaneous binding of mutated plasma VWF to the platelet GPIb receptor, thereby preventing its adhesive function. Although reportedly not the general rule, both giant platelets and thrombocytopenia occur in type 2B VWD. In rare families, severe macrothrombocytopenia is associated with circulating platelet aggregates.21 Studies on a French family with an R1308P VWF substitution have shown that such platelets appear to be produced prematurely and show signs of apoptosis with an altered expression of Ca2+-ATPases and signs of caspase-3-mediated poly (ADP-ribose) polymerase (PARP) hydrolysis.21 Interestingly, culture of CD34+ cells from the peripheral blood of a member of this family resulted in an unusual surface expression of the mutant VWF on mature MK and intertwined pro-platelets suggesting that platelet production is perturbed. Early association of neosynthesized VWF with GPIb may be interfering with the normal maturation and/or migration of MK. A similar phenotype with circulating agglutinated platelets was previously reported in the Montreal platelet syndrome.1, 2
Thrombocytopenia with platelets of normal size
In one large family, autosomal dominant and lifelong moderate to severe thrombocytopenia was associated with normal sized platelets and a late block in terminal differentiation of MK.22 Clinical manifestations include a propensity to easy bruising and increased bleeding at times of hemostatic stress. Myeloid or erythroid cells are not affected and there is no progression to aplastic anaemia. A novel putative kinase, FLJ14813, was identified as a candidate gene on chromosome 10 in this disorder.23
Familial thrombocytopenia with a predisposition to acute myelogenous leukemia
Studies on another large kindred linked a bleeding tendency to an autosomal dominant disorder of platelet production and function, and a propensity to develop myeloid leukemia.24 Thrombocytopenia was moderate and platelet size normal. Linkage to markers on chromosome 21q identified an 880-kb interval containing the disease gene. Further analysis on the above and other families revealed nonsense mutations, missense mutations or intragenic deletion of one allele of the hematopoietic transcription factor RUNX1 (CBFA2, AML1) gene, abnormalities that cosegregated with the disease in affected families.25, 26 RUNX1 is thought to act as a tumor suppressor. The haplodeficiency and missense or nonsense mutations interfering with DNA binding appear to lead to decreased CFU-MK and insufficient production of platelets from birth. Mutated RUNX1 may heterodimerize with normal protein and lead to loss of function. The propensity to develop leukemia requires that patients have a higher tendency to develop a second mutation either in CBFA2 or another gene.26 This may be aided by the presence of an expanded population of undifferentiated HSC. When platelets of one such patient were examined, impaired platelet aggregation, secretion, protein phosphorylation of plekstrin and MLC, and GPIIb-IIIa activation were accompanied by a decreased platelet expression of myosin light chain regulatory polypeptide gene and other genes by platelet expression profiling.27
Gray platelet syndrome
A mild bleeding disorder for which both autosomal recessive and autosomal dominant inheritance have been described, gray platelet syndrome (GPS) is characterized by the platelet’s inability to store α-granule proteins.28 Platelets are enlarged in this disorder but not giant. Thrombocytopenia is moderate, and the absence of α-granule contents gives the platelets a typical gray appearance on blood smears. It remains controversial whether the defect extends to neutrophils.28 A feature of most patients is the early onset of myelofibrosis (which remains stable), a finding attributed to the inability of MK to store newly synthesized platelet-derived growth factors which, as a result, are released into the marrow. There is a tendency for secretion-dependent platelet aggregation to be abnormal. Collagen-induced platelet aggregation is particularly affected in a cohort of GPS patients with an acquired deficiency of the collagen receptor, GPVI.29 This probably results from an aberrant metalloprotease activity. Electron microscopy has shown that the α-granules in gray platelets and megakaryocytes are small and almost empty rather than absent, many vacuoles are to be seen. Residual α-granule proteins can be detected in the surface-connected canalicular system (SCCS) and the basic defect appears to involve packaging or storage of the α-granule contents. Dense granules and their contents are normal. Emperipolesis (passage of other blood cells through megakaryocytes) has also been described in GPS perhaps linked to an abnormal surface expression of P-selectin.30 Probably GPS is a heterogeneous disorder with more than one molecular cause. X-linked GPS due to a GATA1 Arg216Gln mutation has been reported.31 However this mutation has previously been associated with X-linked thrombocytopenia with thalassemia (XLTT) (see later in the text).
Wiskott-Aldrich syndrome (WAS), an X-linked recessive disease characterized by moderate to severe thrombocytopenia, is characterized by a predisposition to infection, and eczema due to immune deficiency.1, 2 A milder form without the immune problems is known as hereditary X-linked thrombocytopenia (XLT). WAS platelets are small and show a decreased aggregation response. The disease is not exclusive to platelets, since T lymphocytes also show defective function. The gene responsible for WAS has been cloned and encodes a 502 amino acid protein (p) termed WASp that is selectively expressed in HSC lineages. Mutations or other genetic defects localize throughout the WASP gene and result either in the decreased expression of WASp or its absence.32 Mutations in exons 1, 2 and 3 were initially reported to results more likely in XLT. However, this may be due to a high prevalence of missense mutations in this region; a comprehensive phenotype-genotype study in Japan suggested that the clinical phenotype depended on the presence or absence of WASp.33 WASp-negative patients show a much higher susceptibility to infection and eczema as well as having a greater tendency for malignancies.
WASp is a multifunctional protein involved in signal transduction, possessing tyrosine phosphorylation sites and adapter protein function. It has a subtle role in cytoskeleton formation, and stimulates actin assembly by the Arp2/3 complex.34 Studies on a murine model have identified a critical role for WASp during murine platelet biogenesis.35 The absence of WASp led to a defective MK interaction with collagen, a lack of formation of actin-rich podosomes and of SDF-1-induced MK migration. It was concluded that platelets were released prematurely ectopically into the marrow space. So, more evidence has been obtained pointing to an abnormal interaction of MK with their environment in the generation of familial thrombocytopenias. The finding that lentiviral vectors targeting WASp expression to hematopoietic cells efficiently transduce and correct cells from WAS patients, offers long-term hope for gene therapy in WAS.36 Stem cell transplantation is currently performed.
The Paris-Trousseau syndrome is a rare disorder in which low platelet production and a mild hemorragic tendency are associated with a haplodeficiency of chromosome 11 (deletion at 11q23.3–24).37 Thrombocytopenia can be chronic (<50×109/L) despite normal platelet survival. The platelet aggregation response may be abnormal. Paris-Trousseau syndrome is often said to have autosomal dominant inheritance. Most reports concern children. Bone marrow dysmegakaryopoiesis is a constant feature; circulating platelets are often enlarged and some have characteristic giant α-granules. Morphologically distinct populations of MK are present; one composed of normal cells and the other of large numbers of small immature cells with arrested maturation. Occasionally, the platelet count may rapidly increase after birth and even normalize. The giant α-granules may represent granule fusion after platelet release. Paris-Trousseau syndrome is a variant of the much more frequently encountered Jacobsen syndrome in which a 11q.23 deletion can give rise to congenital heart defects, trigonocephaly, mental retardation, respiratory infections and malfunctions of multiple organs.37, 38 Pancytopenia and/or thrombocytopenia are also seen in some but not all patients with Jacobsen syndrome. A feature of the deletion that occurs within 11q23-q24 is that it affects but one allele. Of variable length, the deletion includes two ETS transcription factor genes, ETS-1 and FLI-1 that control, among other things, platelet membrane glycorotein expression.39, 40 Lentivirus-mediated overexpression of FLI-1 in CD34+ cells from a Paris-Trousseau syndrome patient restored megakaryocytopoiesis in vitro, thereby proving that FLI-1 hemizygous deletion contributes to the hematopoietic defects.41 The authors elegantly showed that FLI-1 expression is transiently monoallelic in CD41+CD42−progenitors from normal donors, while it is predominantly biallelic in the other stages of megakaryocytopoiesis. A half-reduction in FLI-1 gene dosage generated a subpopulation of CD41+CD42− cells completely lacking FLI-1 transcription. The decreased FLI-1 protein was suggested to prevent megakaryocyte differentiation, and to explain the presence of a subpopulation of micromegakaryocytes that fail to reach the platelet production stage.
Congenital amegakaryocytic thrombocytopenia (CAMT)
Here, severe thrombocytopenia at birth may develop into a pancytopenia. With autosomal recessive inheritance, affected patients have low numbers of MK in their marrow from birth, but show no physical abnormalities. The defects concern the incapacity of TPO to fulfill its normal thrombopoietic role as a result of abnormalities in the c-MPL gene that encodes the TPO receptor (c-MPL).1, 2 Ballmaier et al.42 defined heterogeneity in congenital amegakaryocytic thrombocytopenia, distinguishing patients with a severe form and early development into pancytopenia, and a second group with mild thrombocytopenia in the first year of life but which then worsens with pancytopenia occurring in later life. Of eight point mutations detected in their study, frameshift or nonsense mutations predicted a complete loss of c-Mpl in five patients with severe disease. Missense mutations in three patients possessing residual c-Mpl were associated with a slower progression of the disease. The same authors have more recently extended their findings to a larger group of patients.43 However, in this issue of the journal, Savoia et al.44 have challenged this conclusion. The authors diagnosed five CAMT patients, but failed to find a correlation between the type of mutation and the clinical course. Furthermore, they provided evidence that elevated levels of the inhibitory cytokines, TNF-α and IFN-γ, probably contribute to the pancytopenia. Importantly, these results suggest that bone marrow transplantation should not be delayed in children with mutations that might predict residual activity of c-Mpl.
Amegakaryocytic thrombocytopenia with radio-ulnar synostosis
A small number of families have been reported in which family members have an association of bone marrow failure and radio-ulnar synostosis (and other skeletal abnormalities), with what appeared to be autosomal dominant inheritance.45, 46 In some children, symptomatic thrombocytopenia with bleeding required correction by bone marrow or umbilical cord stem-cell transplantation. Marrow studies showed few or no MK. Pancytopenia developed in some but not all individuals. A heterozygous mutation was found in exon 2 of the homeobox gene, HOXA11. This mutation was only found in affected individuals and occurred in a domain critical for DNA binding. More recent studies have shown that this mutation gives rise to a truncated HOXA11 protein with defective DNA-binding ability.47 Furthermore, it appears that the mutation interferes with the early stages of MK differentiation. However, it is so far unclear whether both the skeletal and hemato-logical problems relate to HOXA11 gene deficiency although a predisposition to pancytopenia suggests an abnormality at the level of the HSC.
GATA-1-linked thrombocytopenia with dyserythropoiesis
A special category of familial thrombocytopenia segregates with dyserythropoiesis with or without anemia. Immunodeficiency and eczema are absent, platelets are often large.1, 2 Thrombocytopenia ranges from moderate to severe (10,000–40,000 platelets/μL); splenomegaly, reticulocytosis, and unbalanced hemoglobin chain synthesis resembling β-thalassemia minor may be present. Bleeding times are prolonged and platelet function moderately affected. The gene responsible for this combination of defects was first mapped to the X-chromosome (Xp11-12). The transcription factor GATA-1 together with its cofactor FOG-1 (Friend of GATA) regulates erythropoiesis and megakaryocytopoiesis. The GATA-1 gene is located at Xp11–12 and was found to be mutated in a series of families with XLT and dyserythropoiesis.48–50 GATA-1 contains two zinc fingers, the C-terminal of which accounts for sequence-specific DNA binding and the N-terminal for both stabilization of DNA binding and for the interaction with FOG-1.
In general, MKs and platelets from these patients display few α-granules (see Section on the GPS) and an abor-mal membrane development. Erythrocytes are abnormal in size and shape. In one family two affected male half siblings were both severely anemic and thrombocytopenic at birth and thereafter.49 A V205M substitution was found in GATA-1 that abrogated its interaction with FOG-1. Such findings underscored the importance of FOG-1:GATA-1 associations in both MK and erythroid development. Freson et al.50 studied a family with XLT without anemia (but with some dyserythropoietic features). A D218G substitution in GATA-1 affected its interaction with FOG1 although GATA-1 binding to DNA was normal. The bone marrow showed an increased number of large MK. Structurally abnormal and enlarged platelets and dysmorphic red cells were present. Semiquantitative RNA analysis revealed a low transcription of the GATA-1 target genes, GPIb‚ and GPIX; and a reduced transcription of late MK maturation markers. Yu et al.51 reported a R216Q substitution in the N-terminal finger of GATA-1 in the family that originally led to the description of XLT, and in which red cell abnormalities were consistent with β-thalassemia (XLTT). Studies with recombinant GATA-1 showed that the mutation destabilized its binding to palindromic DNA sites but did not affect FOG-1 binding. Overall, a defective interaction of GATA-1 with either DNA or FOG-1 can disrupt hematopoiesis and give rise to disorders sharing common themes but with diverging phenotypes. Defects in GATA-1 also extend to Down syndrome acute megakaryoblastic leukemia.48 The role of GATA-1 was confirmed by transgenic mouse models, with immature low ploidy megakaryocyte progenitors proliferating profusely in vitro as a result of a unique megakaryocyte differentiation arrest, a finding associated with dyserythropoietic anemia. Transgenic rescue of GATA-1 deficient mice with GATA-1 lacking a FOG-1 association site (GATA-1V205G) phenocopied patients with a V205 mutation.52
Thrombocytopenia with absent radii
Thrombocytopenia with absent radii (TAR) syndrome is a rare congenital defect associating CAMT-like thrombocytopenia and osteodysgenesis, namely shortened (or absent) forearms due to bilateral radial aplasia.53 Although other skeletal anomalies can be present, hands and fingers are unaffected. Inheritance appears to be autosomal recessive. Thrombocytopenia is very severe at birth, but platelet numbers increase during childhood and platelet count can be near normal in adulthood. Intracranial hemorrhage is a particular risk in early life when mortality is at its greatest. TPO levels are elevated in the serum, and platelets of TAR patients failed to respond to recombinant TPO as measured by testing TPO synergism to suboptimal concentrations of platelet activators.54 Studies on in vitro MK differentiation and expression of the c-MPL gene in six patients showed a profound defect in MK progenitors.55 This was associated with a blockage in MK differentiation with cells expressing GPIb without GPIIb-IIIa. Megakaryocyte differentiation was poorly stimulated by TPO or other cytokines and this was associated with a decrease in c-MPL transcripts and TPO receptor protein. However, screening the c-MPL gene has so far failed to show mutations and a defect in signal transduction has been proposed to explain the disorder.
We have briefly covered some of the major causes of inherited thrombocytopenias. However, an important question concerns the true abundance of inherited diseases with a low platelet count, so often falsely diagnosed as immune thrombocytopenic purpura. A recent survey of patients with macrothrombocytopenias in our Reference Center in Bordeaux revealed that in a cohort of 38 unrelated patients, 27 did not fall within the categories covered by this review, thereby suggesting other molecular causes. Australian studies on an N-ethyl-N-nitrosourea mutagenesis screen in mice also point to a wider range of causes of thrombocytopenia.56–58 For example, the mouse models have shown that programmed anuclear cell death delimits platelet life span and suggests that pro-survival Bcl-xL is another candidate gene for mutations in patients with congenitally shortened platelet life span and normal sized platelets.56 Also to be looked out for are chaperone or enzyme mutations affecting platelet glycoprotein glycosylation and where thrombocytopenia can be associated with kidney disease.57 Finally, mutations in the cofilin partner Aip/Wdr1 cause autoinflammatory disease and macrothrombocytopenia.58 These are exciting times and much work remains to be done.
- Copyright© Ferrata Storti Foundation