Thrombocytopenia with absent radii (TAR) syndrome: from hemopoietic progenitor to mesenchymal stromal cell disease?

Routine laboratory characteristics 

Patients usually present with severe symptomatic thrombocytopenia, (platelet count <10 × 109/L) in the first week of life. Bone marrow megakaryocytes may be decreased, absent, or immature, with small, basophilic, and vacuolated cells, while erythropoiesis is normal or stimulated, so that anemia, when present, is very likely due to bleeding episodes. Leukocytosis may be present and precedes thrombocytopenia, white blood cell count being >35 × 109/L with a left shift and eosinophilia in about 50% of patients. Bleeding is most frequent during the first 1 to 2 years of life, with increased mortality due to intracranial hemorrhage [4]. Frequency of hemorrhagic episodes decreases with age in agreement with a rise in the platelets, the function of which is probably normal in the majority of patients with TAR syndrome [4]. Laboratory and clinical features distinguishing TAR syndrome from related disorders are reported in Table 2.

Table 2. Laboratory and Clinical features distinguishing thrombocytopenia-absent radius syndrome from related disorders

		Hematological	Skeletal	Other	Prognosis	Inheritance
	TAR	Isolated hypomegakaryocytic thrombocytopenia; onset in infancy; other cell lines normal	Bilateral absence of radii; thumbs always present; lower extremities less involved	Cardiac abnormalities	Excellent if after first year	Probably autosomic recessive
	Fanconi's anemia	Anemia; onset of thrombocytopenia preceding aplastic anemia	Absent thumbs; Terminal skeletal defects	Renal anomalies	Poor without transplant; high risk of malignancies	Autosomal recessive; chromosomal breaks, gaps and rearrangements
	Holt-Oram	Normal	Thumb and upper extremity defects	Cardiac anomalies	Good	Autosomal dominant
	Roberts’/SC-phocomelia	Not significant	Tetraphocom; thumbs sometimes present craniofacial anomalies	Multiple malformations	Failure to thrive	Autosomal recessive
	Diamond-Blackfan	Hypoplastic anemia	Multiple

Therapy 

When necessary to control bleeding, treatment is based on platelet support, while use of prophylactic transfusions is restricted to patients at high risk of clinically significant hemorrhage. In these cases leukocyte-reduced platelet concentrates or random single-donor platelets reduce risk of human leukocyte antigen and alloimmunization. Splenectomy may be effective for treatment of thrombocytopenia in adults, while bone marrow transplantation is an extreme option for patients whose bleeding cannot be adequately controlled.

Megakaryocyte progenitors 

Colony growth studies in the TAR syndrome have been made difficult by the low number of patients and refusal of parents to give permission for bone marrow sampling in some patient populations. Moreover, different cytokine combinations have been used [11–13] and, on one occasion, the combination of interleukin (IL)-3, IL-6 and stem cell factor was more efficient than pegylated recombinant human megakaryocyte growth and development factor alone or plus stem cell factor in inducing colony-forming unit-megakaryocyte (CFU-MK) growth from a population of CD34+ cells [14]. Hence, it is not surprising that results vary from no MK colony growth 11, 15 to an increase in megakaryocytic progenitors, with a much lower number of cells per colony and small size of individual megakaryocytes compared to controls 12, 13. We studied (unpublished data) CFU-MK growth in four patients and found normal growth in all cases, the only noteworthy difference from normal controls being the MK colony size, a finding also reported by other authors [12–14]. The overall data on colony growth suggest that thrombocytopenia could be due to a decreased response to thrombopoietin (TPO) that affects both proliferation and differentiation [14] and, as al-Jefri et al. [13] suggested, reduction in frequency of MK progenitors and the suboptimal size of colonies are compatible with defective signal transduction in the TPO pathway. In the meantime, the increased number of granulocyte-monocyte colony forming units (CFU-GM) and erythroid colonies reported in two articles 11, 12 indicates that granulopoiesis and erythropoiesis are unaffected in the TAR syndrome. Table 3 summarizes the results of MK colony growth studies.

Table 3. Megakaryocytic colony formation from thrombocytopenia-absent radius (patients) and normal (controls) bone marrow progenitors studied by different research groups

	First author	CFU-MK/105 cells		Cell type	No. of		Stimulus
		Patients	Controls		Patients	Controls
	Sekine [12]	9	17	MNAC	1	12	IL-3, SCF, EPO, GM-CSF + TPO
		125	78
	Ballmaier [11]	ND∗	ND∗	MNAC	1	12	IL-3, IL-6, EPO, GM-CSF, SCF + TPO
		0	11
	al-Jefri, et al. [13]	4	20	MNAC	3	6	IL-3, IL-6, GM-CSF, SCF + PEG-rHuMGDF
		8	15
	Bagnara	42	21	MNAC	4	10	IL-3, GM-CSF + TPO
		62	50
	Homans [15]	0	ND	MNAC	1	2
		0	ND
	Letetsu [14]	20	175	PB-CD34+ (2 × 103)	4	2	SCF, IL-3, IL-6, SCF + PEG-rHuMGDF
		5	275

CFU-MK = colony-forming unit megakaryocyte; EPO = erythropoeitin; GM-CSF = granulocyte-macrophage colony-stimulating factor; IL = interleukin; MNAC = mononuclear nonadherent cells; ND = not determined; PEG-rHuMGDF = pegylated recombinant human megakaryocyte growth and development factor; SCF = stem cell factor; TPO = thrombopoietin.

∗ High number of megakaryocytic colonies.

TPO and TPO receptor 

TPO, the primary regulator of thrombopoiesis [16], although not lineage-restricted 17, 18, 19, and its receptor MPL, a member of the hematopoietic growth factor receptor superfamily codified by the proto-oncogene c-mpl 20, 21, have been studied in TAR syndrome patients. Ballmaier et al. [11] found that TPO serum levels were elevated in two separate series of TAR patients (six cases) 11, 22, on the basis of a bioassay using 32D cell line transfected with human c-mpl. Strauss et al. [23] and our group (unpublished data) obtained similar results by a sandwich enzyme-linked immunosorbent assay. In two case reports, serum levels of TPO in patients with TAR were comparable with those of an age-matched control 12, 24, in spite of marked thrombocytopenia. In conclusion, the data are few and far between, but suggest that defective TPO production is not likely to induce thrombocytopenia, in agreement with the stability of TPO mRNA expression and protein levels in thrombocytopenic patients [22]. Studies on other cytokines regulating megakaryocytopoiesis, such as IL-6, leukemia inhibitory factor, and IL-11 have not proved informative, because only IL-11 was found elevated in three of five TAR patients as compared to normal sera [11]. The TPO receptor, c-mpl, plays a crucial role in MK differentiation, so many studies have focused on mutations possibly leading to defective binding of the ligand and possibly affecting both CFU-MK proliferation and MK maturation. In 1998, Strippoli et al. [25] investigated four unrelated TAR patients sequencing each of the 12 exons of the c-mpl gene including intronic flanking sequences plus the 5′ promoter region. The findings of GGCC inversion at position 1175 and C insertion at position 1252 revealed in those patients and 13 normal subjects indicate that they are common, at least in the Italian population, while Letestu et al. [14] did not detect any mutation or polymorphism in eight patients. These results suggest that abnormalities of the c-mpl gene do not play a role, but changes in MPL (detected by Western blot analysis) and in c-mpl RNA levels may be involved in the defective megakaryopoiesis of TAR syndrome [14]. The molecular weight of platelet c-mpl was found in the range of normal controls [11], but conflicting data regarding its expression on the platelet surface have been reported 11, 14. Moreover, the absence of in vitro reactivity to recombinant TPO of platelets from TAR patients, as reported by Ballmeier et al., and the presence of defective platelet aggregation observed in our laboratory in one patient and his sibling [25] suggest that signal transduction in TAR patients is altered.

Other genetic studies 

Attention has also been trained on homeobox genes (HOX). Expression analysis and gain- or loss-of-function studies have shown that HOX genes play an important role in the regulation of early stages of hematopoiesis, including self-renewal, and also in leukemogenesis. In particular, HOXA10 may be related to megakaryocytopoiesis (its overexpression stimulates mouse MK development in vitro [26]) and targeted disruptions of HOX genes result in abnormal development of the mouse radius. However, the hypothesis that HOXA10 homeobox gene is a candidate for the thrombocytopenia of TAR syndrome cannot be maintained, as no genomic rearrangements or deletions have been found at the HOXA10, HOXA11, or HOXD11 loci by direct sequencing of overlapping polymerase chain reaction products and Southern blotting [26].

Deletion in chromosome 1q21.1 is associated with congenital heart disease [27] and TAR syndrome [28]. Klopocki et al. [28] described a common interstitial microdeletion of 200 kb on the same chromosome in all 30 investigated patients with TAR syndrome, detected by microarray-based comparative genomic hybridization. Analysis of the parents revealed that this deletion occurred de novo in 25% of patients. However, deletion on chromosome 1q21.1 is considered necessary but not sufficient to give rise to a TAR phenotype and a second unknown mutant gene is required [28]. Transforming growth factor-β2 (TGF-β2) gene is also located on chromosome 1q although in a different position (1q41). This multifunctional growth factor is the prototype of an enormous super family of structurally and functionally similar cytokines that play a key role in embryo development, osteogenetic differentiation in normal physiology, and the pathogenesis of several diseases 29, 30, 31. The effects of targeted TGF-β2 disruption genes were studied by Sanford et al. [32], who found that TGF-β2-null mice exhibit perinatal mortality and a wide range of developmental defects (e.g., cardiac, lung, craniofacial, limb, spinal column, eye, inner ear, and urogenital) and do not overlap phenotypically with TGF-β1-; and TGF-β3-null mice, indicating numerous noncompensated functions between the TGF-β isoforms. We carried out a complete mutational screening of the TGF-β2 coding region and promoter region by DNA sequencing on four TAR patients deriving from three families. Two mutations were identified in the 5′ end noncoding sequence of TGF-β2 gene: an A–C substitution at position 2114 (Fig. 1A) and an A–G substitution at position 2339 on the M87843 Genebank sequence, which were not present in 60 normal controls (Fig. 1B). Automatic sequencing of TGF-β2 gene in TAR samples showed many polymorphisms, as observed upon comparing electropherograms obtained by DNA patient analysis with the M87843 Genebank sequence. Moreover, sequence analysis of the gene promoter region, codifying region and exon-intron junctions showed 65 polymorphisms, but they were also found to be frequent in the control population [33].

View Large Image Download to PowerPoint

Figure 1 Sequencing electropherograms (ABI–PRISM 377 instrument) of transforming growth factor β2 (TGF–β2) gene. Two mutations were identified: an A−C substitution at position 2114 (A) and an A−G substitution at position 2339 (B) as compared to the sequence M87843 Genebank accession number.

The specific region of the TGF-β2 gene with a potential regulatory activity, including the two identified mutations (Fig. 1), was amplified from the genomic DNA of both the mutated and the wild-type samples. By luciferase assay, we verified that there were no differences in gene expression induced by mutated and wild-type fragments (see Fig. 2) [34]. Whether TGF-β1, a potent inhibitor of megakaryopoiesis, TGF-β3 (a pivotal factor in chondrogenesis) and their receptors may also play a role in the physiopathology of TAR syndrome is still unknown. We isolated a sufficient number of adherent cells from a TAR bone marrow sample and characterized them by flow cytometry as described previously 35, 36. Phenotype and differentiation capabilities are somewhat different in adherent cells from bone marrow samples of TAR syndrome and mesenchymal stromal cells (MSCs). MSCs from normal bone marrow samples express CD73 and CD105 antigens [34], while TAR syndrome-adherent cells expressed CD73 (Fig. 3B) but did not express endoglin (CD105), part of the receptor complex for TGF-β1 and TGF-β3 (Fig. 3A). Endoglin has a pivotal role in development of the cardiovascular system, at least in a mouse model [37], and also in vascular remodeling. Interestingly, in humans endoglin may be involved in the autosomal dominant disorder hereditary hemorrhagic telangiectasia type 1 [38]. Moreover, in our experience isolated TAR syndrome stromal cells were capable of normal osteogenic differentiation (Fig. 3C, D), but failed to show chondrogenic differentiation.

View Large Image Download to PowerPoint

Figure 2 Histograms obtained by luciferase assay (Victor2 Multilabel Counter; Perkin Elmer). No differences in gene expression were detected between mutated and wild–type fragments. CPS = counts per second.

View Large Image Download to PowerPoint

Figure 3 Flow cytometric characterization of mesenchymal stromal cells (MSCs) from one case of thrombocytopenia–absent radius (TAR) syndrome (A, B). (A) Negative reaction to SH2 antibody binding CD105 antigen; (B): positive reaction to SH3 antibody binding CD73 antigen. Cultured MSCs from TAR patient were tested for the ability to differentiate in vitro toward osteogenic lineage (C, D). Osteogenesis was indicated by the increase in alkaline phosphatase (C) and calcium deposition (D) in histochemistry staining.