Cell Tissue Res (2010) 339:155–165 DOI 10.1007/s00441-009-0874-y

REVIEW

Unexpected matrix diseases and novel therapeutic strategies Claudia Nicolae & Bjorn R. Olsen

Received: 12 August 2009 / Accepted: 2 September 2009 / Published online: 8 October 2009 # Springer-Verlag 2009

Abstract Within the framework of a broad definition of the extracellular matrix (ECM), this review discusses three genetic disorders in which major pathogenetic features have been traced back to alterations in the levels/activities of matrix components. In each case, disease-associated alterations are found both intra- and extracellularly. The nature of the ECM involvement is surprising, offers an exciting therapeutic opportunity, and deepens our understanding of ECM-cell interactions. The first of these disorders, cherubism, is a case of inflammatory bone loss in the jaws of children for reasons that are surprisingly systemic in nature, considering the local nature of the disease. The primary defect involves an intracellular signaling molecule, but a major pathogenetic component and therapeutic target of the disease is the extracellular cytokine tumor necrosis factor alpha. The second disorder, Knobloch syndrome, is caused by recessive mutations in collagen XVIII. Although this protein has been classified as belonging to a group of structural macromolecules, the consequence of the mutations is impairment of cellular metabolism. The third disorder, infantile hemangioma, is a common tumor of capillary endothelial cells in infancy. The tumor appears within a few days/weeks after birth, grows rapidly over several months, and regresses over several years. The proliferative phase is the result of constitutively high levels of vascular endothelial cell growth factor (VEGF)dependent signaling through VEGF receptor 2 (VEGFR2), but recent studies have led to the surprising conclusion that

C. Nicolae : B. R. Olsen (*) Department of Developmental Biology, Harvard School of Dental Medicine, Boston, MA 02115, USA e-mail: [email protected]

abnormalities in a cell-surface receptor complex controlling expression of the VEGF decoy receptor VEGFR1 is the underlying cause. Keywords Extracellular matrix . Cherubism . Infantile hemangioma . Knobloch syndrome

Introduction Extracellular matrix (ECM) is frequently described as a scaffold of macromolecular complexes that surrounds all cells, but this narrow definition is clearly insufficient as a description of its perception by cells. From a cellular perspective, the ECM is all the material that separates a cell from its neighbors, including peptides, small and large proteins, and polysaccharides representing numerous functions across the spectrum of maintenance and regulation of tissue architecture, cellular proliferation and differentiation, cell survival, and apoptosis. Thus, structural macromolecules, cytokines, growth factors, bioactive peptides, and cell-surface molecules are all components of the ECM from a cellular point of view. This broad definition of ECM is adopted as a basis for this review for three reasons. First, it represents the most logical basis for research in the field of ECM biology. Research in this field rests on a solid foundation of macromolecular biochemistry and structural investigations of matrix components, but the classification of ECM components on the basis of “structural” and “regulatory” distinctions is now generally accepted to be no longer useful. Second, it provides an unbiased framework for classifying diseases involving major abnormalities in the levels of expression and/or activities of molecules within the extracellular milieu of cells as matrix diseases,

156

independently of traditional thoughts as to what these activities are. Third, it helps stimulate creative thinking about novel therapeutic targets and avoids the “channeling” effects that functional labels have on thoughts related to the treatment of diseases. A good example of the limiting effects of such labels is provided by the history of abnormalities in so-called “structural” macromolecules (including fibrillar collagens or fibrillin, the components of collagen fibers and microfibrils, respectively) for which therapeutic ideas were limited, for some time, to strategies aimed at correcting the “structural” deficiency, until it became clear that a better therapeutic answer may come from attempts to address alterations in signaling functions. Based on this broad definition of ECM, we will discuss, in this review, a selected set of genetically based disorders in which major pathogenetic features can be traced back to alterations in the levels and/or activities of matrix components. In each case, disease-associated alterations can be found in both intracellular and extracellular processes, and the nature of the ECM involvement has come as both a surprise and an exciting therapeutic opportunity. Finally, studies of the pathogenetic mechanisms in the three discussed disorders are opening doors to a deeper understanding of the processes by which ECM and cells interact. The first of these disorders, cherubism, is a case of tissue remodeling “gone haywire” in the jaws of children for reasons that have turned out to be surprisingly systemic in nature, considering the local nature of the disease. In addition, although the primary defect has been discovered to involve an intracellular signaling molecule, the major pathogenetic component and probably the most significant therapeutic target of the disease is the extracellular cytokine tumor necrosis factor alpha (TNFα). The second disorder, Knobloch syndrome, is an example of a matrix disease in which recessive mutations in a matrix component, collagen XVIII, cause defects that have many of the hallmarks of cellular signaling abnormalities. Although this collagen initially was placed in the category of so-called structural macromolecules, subsequent studies have provided little support for a structural role of this collagen. The third disorder, infantile hemangioma, is an intriguingly common (up to 10% of Caucasian infants) tumor of capillary endothelial cells in infancy and early childhood. The tumor usually appears within a few days/weeks after birth, grows rapidly over several months (proliferative phase), and then regresses over a period of several years (involuting phase) until it has reached a state of complete involution. During the proliferative phase, endothelial cell proliferation is the result of constitutively high levels of vascular endothelial cell growth factor (VEGF)-dependent signaling through VEGF receptor 2 (VEGFR2), and recent studies have led to the surprising conclusion that this is caused by abnormalities in a cell-surface receptor complex in which a novel

Cell Tissue Res (2010) 339:155–165

extracellular signaling function of VEGFR2 plays an important role.

Research on cherubism—a journey of discovery and surprises leading to an extracellular therapeutic target First described by Jones in 1933 (Jones 1933), cherubism is a dominantly inherited syndrome that is characterized by excessive bone resorption in the upper and lower jaws and replacement of the bone with inflammatory and fibrous tissue. This causes a characteristic swelling of the lower face and a retraction of the lower eyelids, giving the affected child a “cherubic” appearance. The first pathological facial changes appear between the ages of 2–4 years. The disease progresses during childhood and usually regresses after puberty. In addition to the aesthetic deformity and its psychological consequences, impairment of mastication and of speech functions and permanent dental complications are present. As a result of gene mapping studies in several families with cherubism, the gene region responsible for the disease was located to chromosome 4p16.3 (Mangion et al. 1999; Tiziani et al. 1999). Extensive bone remodeling within the mandible and/or maxilla and a possible connection with tooth development led to the initial hypothesis (wrong, as it turned out) that mutations in FGFR3 or MSX1, located in this chromosomal region, could cause this disease (Mangion et al. 1999). Considering both the temporal and spatial association of the clinical course of cherubism with tooth development and/or eruption, cherubism was also suggested to be caused by defective signaling within the mandible and maxilla, triggered by the eruption of secondary teeth and transmitted through the ECM (Tiziani et al. 1999). Given the jaw-restricted nature of the soft tissue and bone lesions in cherubism patients, the disorder was surprisingly discovered to be caused by heterozygous germline mutations in SH3BP2, a gene that encodes a widely expressed signaling scaffold/adapter protein (Ueki et al. 2001). SH3BP2 contains three modular domains, an N-terminal pleckstrin homology (PH) domain, a 10-amino-acid Src-homology (SH3)-binding region, and a C-terminal SH2 domain. Initially identified in a screen for proteins that bind to the SH3 domain of the tyrosine kinase c-Abl (Cicchetti et al. 1992; Ren et al. 1993), SH3BP2 can form complexes with a number of signaling proteins, such as Syk kinase, Zap-70, LAT, phospholipase C γ1 (PLC-γ1), 14-3-3, Grb2, Cbl, Fyn, Vav1, and Vav2 (Deckert et al. 1996, 1998; Foucault et al. 2003, 2005). Mutations leading to single amino acid substitutions have all been mapped to a six-amino-acid residue-stretch between the PH and SH2 domains encoded by exon 9 (Imai et al. 2003; Li and Yu 2006; Lietman et al. 2006; Lo et al.

Cell Tissue Res (2010) 339:155–165

2003). The missense mutations in SH3BP2 do not abolish its function, as deletion of the chromosomal region in which this gene is located does not cause cherubism, but instead the unrelated Wolf-Hirschhorn syndrome (Zollino et al. 2000). To obtain insights into SH3BP2 function and the pathogenetic consequences of the mutations, we introduced the most common mutation found in cherubism, a proline to arginine substitution (P148R in humans; P146R in mice), into the mouse Sh3bp2 locus (Ueki et al. 2007). Heterozygous knockin mice exhibited osteopenia with increased numbers of osteoclasts in bone but did not show swelling of soft facial tissues and local lymph nodes as seen in patients with cherubism. In contrast, homozygotes displayed swollen eyelids and snouts starting about 6 weeks after birth. They also developed severe systemic bone loss with increased numbers of osteoclasts and inflammatory macrophage-rich infiltrates in skeletal elements and internal organs, including lymph nodes. Most homozygotes died before 30 weeks of age. Further studies showed that the SH3BP2 mutation makes myeloid cells abnormally sensitive to cytokines that stimulate their differentiation to macrophages and osteoclasts (Ueki et al. 2007). In addition, mutant macrophages were found to overproduce tumor necrosis factor α (TNFα) in response to macrophage colony-stimulating factor (M-CSF). The increased osteoclastogenesis in the mutant mice resulted from a RANKL (receptor activator of NF-kB ligand)-dependent increase in the levels and activity of the transcription factor NFATc1 (nuclear factor of activated T cells c1; Aliprantis et al. 2008). These data demonstrate that cherubism is not a local disease, in spite of the jaw-restricted lesions, but a systemic disorder of hematopoietic myeloid cells (Fig. 1). Consistent Fig. 1 Cellular and molecular defects in cherubism. Missense mutations in the gene for SH3BP2 (a widely expressed signaling scaffold/adapter protein) increase the sensitivity of monocytes to macrophage colony-stimulating factor (MCSF) and receptor activator of NF-kB ligand (RANKL). As a result, the formation of macrophages and osteoclasts is stimulated. In addition, production of tumor necrosis factor α (TNFα) by macrophages is upregulated. Since TNFα stimulates the production of M-CSF and RANKL by stromal cells, this creates a positive feedback loop and induces a high level of inflammatory bone loss (modified from Ueki et al. 2007)

157

with this unexpected conclusion was the finding that fetal liver transplants from SH3BP2 homozygous mutant donors into γ-irradiated wild-type mice induced the complete disease phenotype in the recipients (Ueki et al. 2007). A comparison of the histological and clinical features of cherubism in humans and mice indicates that, whereas heterozygous knockin mice develop mild osteoclastdependent osteopenia, the disease process in the jaws of patients is not simply osteopenia, but a dramatic inflammatory process, similar to that seen (systemically) in homozygous knockin mice. This obviously raises the question of why a systemic inflammatory process that, in mice, is seen only in homozygotes is restricted to the jaws of heterozygous patients? The answer to this question is not known. However, it may be related to the high rate of growth and remodeling of the jaws in humans during the first 4–5 years of life. During this period, the jaw bones (in contrast to the rest of the skeleton) not only grow to about 80% of their adult size, but also undergo an extremely high degree of remodeling as they change their shapes and accommodate the growth and eruption of teeth. These processes require a high level of cytokine (M-CSF and RANKL)-dependent osteoclastic differentiation from myeloid precursors. In individuals who are heterozygous for a cherubism mutation, this environment may be sufficient also to drive macrophage differentiation to levels that, in mice, are only seen in homozygotes. Finally, one may ask what aspects of cherubism justify its inclusion among matrix disorders. The answer is simple: from a therapeutic point of view, cherubism in humans is probably a matrix disease. The inflammatory component of the homozygous mouse phenotype is entirely dependent on

158

increased expression of TNFα, and elimination of TNFα in homozygous cherubism mice rescues the early lethality and reduces the phenotype to that of heterozygous mice (Ueki et al. 2007). Since TNFα is a trimeric type II transmembrane protein that can be released as a soluble protein via proteolytic cleavage by the metalloprotease TNFαconverting enzyme TACE/ADAM17 (Black et al. 1997; Bradley 2008; Kriegler et al. 1988; Tang et al. 1996), it is as much a matrix protein as a large number of other proteins that are found in both membrane-bound and soluble forms in the extracellular space. In addition, the actions of TNFα are to a large extent directed at or coupled to changes in cell-matrix interactions, including the induction of many matrix metalloproteinases (Clark et al. 2008; Fanjul-Fernandez et al. 2009), the inactivation and modulation of the cell-surface expression of integrins (Gao et al. 2000; Rotundo et al. 2002), the downregulation of lysyl oxidase (Rodriguez et al. 2008), the upregulation of tissue transglutaminase expression (Chen et al. 2000), and the control of cell proliferation, migration, and apoptosis (Chen et al. 2007). A major therapeutic target for cherubism is therefore an ECM protein, and biological drugs already in clinical use for the treatment of rheumatoid arthritis (Brennan and McInnes 2008) and other immune disorders, such as monoclonal antibodies specific for TNFα or fusion proteins of human soluble TNFR and the Fc fragment of human IgG, may prevent most of the facial tissue destruction in cherubism patients.

Knobloch syndrome—an organ-restricted matrix disease with developmental and age-dependent cellular abnormalities The cherubism story provides an example of the way that intracellular signaling abnormalities in a specific cell type can result in disease-causing changes in the level/activity of an extracellular, therapeutically accessible, regulator of cellular activity. The next story, Knobloch syndrome, shows how abnormalities in an extracellular component can lead to intracellular metabolic changes. As in the case of cherubism, it also illustrates the power of combining human and mouse genetic approaches when analyzing the pathogenetic basis of human disease. Knobloch syndrome is an autosomal recessive disorder with phenotypic features that are primarily restricted to the eye, including vitreoretinal degeneration with retinal detachment, macular degeneration and high myopia, in addition to occipital (or more rarely frontal) encephalocele (Knobloch and Layer 1972). The gene responsible for the disorder was mapped to chromosome 21q22.3 (Sertié et al. 1996) and a mutation generating a premature stop codon within the COL18A1 gene was identified (Sertié et al. 2000). Given the ocular-restricted

Cell Tissue Res (2010) 339:155–165

phenotype, this was quite surprising, since collagen XVIII is a major collagen/proteoglycan (Dong et al. 2003; Halfter et al. 1998) component of almost all vascular and epithelial basement membranes in a variety of tissues (Muragaki et al. 1995; Saarela et al. 1998a, 1998b). Three isoforms of the protein are produced as a consequence of transcription from two different promoters and alternative splicing of the primary transcript (Muragaki et al. 1995; Rehn et al. 1996). The isoforms differ in their N-terminal noncollagenous domains (NC11), but they share a common triple-helical region and a C-terminal (NC1) domain. A 20-kDa proteolytic fragment from the C-terminal end of NC1 is the heparin-binding protein endostatin, which has been found to have significant anti-angiogenic activity both in vitro and in vivo (O’Reilly et al. 1997; Read et al. 2001; Yamaguchi et al. 1999). In patients with Knobloch syndrome, premature stop codons, insertions, and deletions have been identified in various exons and introns of COL18A1 (Suzuki et al. 2009). That these mutations represent loss-of-function mutations has recently been confirmed by the loss of immunodetectable collagen XVIII in skin biopsies from Knobloch syndrome patients (Suzuki et al. 2009). All patients have severe ocular abnormalities that progress to bilateral blindness. The age-dependent progression of the disease suggests that collagen XVIII is important for both the development of the eye and the maintenance of visual function and indicates that long-term studies in an animal model are required for understanding collagen XVIII function and the pathogenetic mechanisms underlying Knobloch syndrome. Collagen XVIII null mice, generated and characterized by our group, represent such a model. Collaborative studies have shown that collagen-XVIII-deficient mice exhibit a delay in the normal regression of hyaloid vessels, abnormal outgrowth of retinal vessels, and abnormalities in the iris and ciliary body during the early postnatal period (Fukai et al. 2002; Marneros et al. 2004; Marneros and Olsen 2003; Ylikärppä et al. 2003). With age, homozygous mutant mice exhibit progressive loss of vision, accumulation of abnormal protein deposits between the retinal pigment epithelium (RPE) and the underlying Bruch’s membrane, similar to that found in age-related macular degeneration in humans, and loss of RPE structure and functions (Fukai et al. 2002; Marneros et al. 2004; Marneros and Olsen 2003). Normal RPE cells have their apical and basal plasma membrane surfaces enormously expanded by the formation of tightly packed microvilli (Fig. 2). On the side facing the outer segments of the photoreceptors, these microvilli are normally in close contact with the photoreceptors (about 30 photoreceptors per RPE cell) ensuring a high rate of phagocytic removal (a daily phagocytic burst with about 2000 photoreceptor membrane discs digested by each RPE cell) of the spent outer segments by the RPE. In aging

Cell Tissue Res (2010) 339:155–165

159

Col18a1 null eyes, the intimate contact between the two cell types is lost, phagocytosis of spent photoreceptor outer segments is impaired, and spent outer segments accumulate in the area between the photoreceptors and the RPE (Marneros et al. 2004). On the side facing the underlying Bruch’s membrane, the microvilli provide an enormous surface area for the transport of materials from the RPE into the fenestrated capillaries of Bruch’s membrane and for the uptake of vitamin A from the blood. In the RPE, the all-trans-retinol is converted to 11-cis-retinal and transported to the photoreceptors for the generation of rhodopsin. In aging collagen XVIII mutant mice, the microvilli are lost and replaced by deposits of basementmembrane-like material (Marneros et al. 2004; Fig. 2). Given these changes on both sides of the RPE epithelial layer, retinal levels of rhodopsin in aged mutant mice are unsurprisingly reduced to 50% of levels in control eyes, and the level of RPE65 protein, essential for the formation of 11-cis-retinal, is only a third in mutant eyes compared with age-matched controls. However, a surprising aspect is that the absence of a basement membrane component on one side of the RPE cells should have such dramatic

effects on the other side where the cells interact with the photoreceptors. This raises important questions regarding the pathogenetic mechanisms of Knobloch syndrome and the phenotypes of the mouse model of the disease. One question is whether the phenotypes, at different ocular sites, are all consequences of the local loss of collagen XVIII from the basement membranes at these sites. Whereas some of the early postnatal phenotypes in the anterior portion of the eye, such as the delayed hyaloid vessel regression observed in Col18a1 null mice and documented in a patient with Knobloch syndrome (Duh et al. 2004; Fukai et al. 2002), the rupture and separation of the anterior and posterior iris epithelial layers, and the flattening of the ciliary epithelial basal surface (Marneros and Olsen 2003; Ylikärppä et al. 2003), might be caused by local loss of collagen XVIII, the pathogenetic mechanisms are unclear. Could they be associated with loss of the heparan sulfate side chains in collagen XVIII? Might they be consequences of the loss of the collagen region with its multiple short triple-helical domains? Do the mechanisms result from the local loss of the heparan-sulfate-binding Cterminal endostatin domain of collagen XVIII?

Fig. 2 Age-dependent retinal abnormalities in collagen-XVIIIdeficient eyes. Left Light microscopy of a section through a normal mouse retina (a), showing the pigmented epithelial cell layer (RPE) separating the photoreceptor outer segments from the vascular (and pigmented) choroid (boxed area regions shown at higher magnification middle and right). Middle Electron micrographs of a normal

mouse eye showing the apical infoldings of the RPE surrounding the photoreceptor outer segments (b) and the basal infoldings at the basal lamina separating the RPE from the choroid (d). Right Electron micrographs from a collagen-XVIII-deficient mouse eye showing abnormalities in the apical (c) and basal (e) infoldings of the RPE

160

Furthermore, whereas age-dependent phenotypes, including the flattening of the basal surface of the RPE, the accumulation of basal laminar deposits, and the loss of RPE phagocytic activity (Marneros et al. 2004) could also result from the local loss of collagen XVIII as a basement membrane component, they might instead be the consequences of systemic alterations induced by the loss of collagen XVIII from basement membranes in all organs. One such systemic alteration would be the loss of circulating levels of endostatin or a systemic change in other circulating factors as a result of loss of collagen XVIII in all epithelial and endothelial basement membranes. Support for the possibility that at least some phenotypic consequences of homozygous loss-of-function mutations in collagen XVIII are attributable to the systemic loss of endostatin and not a local collagen XVIII deficiency comes from recent studies in which we have compared laserinduced choroidal neovascularization (CNV) in wild-type and collagen XVIII knock-out mice (Marneros et al. 2007). The data have demonstrated a significant increase in the size of CNV lesions in Col18a1 null mice compared with control mice. In the mutant mice, the lesions occur as large confluent areas of vessels with increased permeability as compared with small, clearly circumscribed areas in controls. Extremely interesting has been the finding that the restoration of the physiological levels of circulating endostatin by administration of recombinant endostatin to the mutant mice rescues the abnormal CNV response and reduces the lesions to those observed in control mice. In addition, the administration of endostatin to control mice after laser treatment reduces the CNV response to barely detectable lesions. This finding suggests that endostatinbased biological agents might usefully complement antiangiogenic therapy in the CNV-associated “wet” form of age-related macular degeneration or more generally in the treatment of angiogenesis-dependent pathological processes. Should further studies indicate that age-related loss of RPE function in Col18a1 null mice is a consequence of the loss of circulating endostatin rather than a deficiency of collagen XVIII within Bruch’s membrane, then endostatin-based therapy may be of value in patients with Knobloch syndrome. However, before endostatin-based therapeutic agents are further pursued, additional studies are needed to improve our understanding of the similarities and differences between the binding properties and signaling functions of the trimeric endostatin-containing NC1 domain within the intact collagen XVIII protein and the properties and signaling activities of the monomeric circulating form of endostatin. Finally, more work is required to characterize the signaling abnormalities in RPE cells that lead to age-dependent vision loss in Knobloch mice and the mechanisms underlying the

Cell Tissue Res (2010) 339:155–165

abnormal accumulation of basement membrane material under the RPE in these mice. Insights into these mechanisms may well open the door to the identification of new targets for therapies to prevent the formation of similar deposits in the common “dry” form of age-related macular degeneration in humans.

Research on infantile hemangioma–a story of basic cellular control mechanisms revealed through the lens of a capillary tumor The next story represents an example of the way that the discovery of a regulatory network linking an endothelial cell-signaling abnormality to defects in cell-surface matrix proteins was made possible by genetically inspired approaches. Infantile hemangiomas are benign capillary endothelial tumors (mostly solitary and sporadic) affecting up to 10% of Caucasian infants, with a lower incidence in infants of Asian and African descent (Boye and Olsen 2009; Hidano and Nakajima 1972; Mulliken and Glowacki 1982). They are more frequent in females than in males. The lesions are not apparent at birth but appear during the first few weeks of life and expand rapidly in the first year. After an intensive proliferating phase, they slowly involute by apoptosis during the next 7–10 years. This unique life cycle concludes with the complete regression of the tumor and the replacement of the proliferating endothelium with adipocyte-rich fibrous tissue. Almost 80% of hemangiomas are cutaneous lesions in the neck and head region and some can interfere with feeding, respiration, or vision and require therapeutic intervention. Current treatment options include corticosteroids (Boon et al. 1999), a chemotherapeutic agent, vincristine (Enjolras et al. 2004; Fawcett et al. 2004), interferon α (Ezekowitz et al. 1992), and propranolol (Ezekowitz et al. 1992; Leaute-Labreze et al. 2008), but these therapies are of variable success and are not based on a clear understanding of the mechanism of action. Several hypotheses for cellular and molecular mechanisms responsible for the appearance and rapid growth of hemangiomas have been put forward. One of these hypotheses, based on similarities between the expression patterns of hemangioma- and placenta-derived endothelial cells, is that hemangiomas represent aberrant differentiation toward the placental microvascular phenotype or represent the growth of embolized placental endothelial cells (Barnes et al. 2005; North et al. 2001). Another hypothesis is that proliferating hemangiomas represent clonal expansions of endothelial cells with defects (for example, somatic mutations) in regulators of endothelial cell proliferation. This is based on the finding that X-chromosome inactivation patterns are non-

Cell Tissue Res (2010) 339:155–165

random (skewed) in endothelial cells isolated from hemangiomas of female patients as compared with the random patterns in non-endothelial cells isolated from the same lesions (Boye et al. 2001). Taking advantage of this clonal nature of hemangioma endothelial cells, we compared the properties of such cells isolated from nine unrelated patients with the properties of several different control endothelial cells in culture (Jinnin et al. 2008). A characteristic feature of all cultured hemangioma endothelial cells is that their level of proliferation is as high in the absence of exogenous VEGF-A as it is in control endothelial cells in the presence of VEGF-A. This activation of proliferation in hemangioma cells is not the result of increased levels of endogenous VEGF-A in these cells but is a consequence of increased phosphorylation of the receptor tyrosine kinase VEGFR2. A number of signaling molecules within pathways that are downstream of VEGFR2 have been found to be constitutively phosphorylated/activated in hemangioma endothelial cells. As expected, the expression of several downstream target proteins are also upregulated, including the transcription factor HIF1α and its target GLUT-1, a glucose transporter previously demonstrated to be upregulated in hemangioma endothelial cells in vivo and found to be useful as an immunohistochemical marker for hemangioma (North et al. 2000). Therefore, VEGFR2-mediated signaling is constitutively activated in hemangioma endothelium. This constitutive phosphorylation/activation can be reduced to control levels when neutralizing antibodies against VEGF-A are added to hemangioma cell cultures, indicating that VEGF-A is somehow more active in hemangioma endothelial cells than in control cells (Jinnin et al. 2008). The reason for the increased activity of VEGF-A hemangioma endothelium became apparent when hemangioma endothelial cells were found to have abnormally low expression levels of the VEGF-A decoy receptor VEGFR1 compared with control endothelial cells, both in vitro and in vivo. In turn, this was found to be caused by extremely low levels of activation/nuclear translocation of NFATc2, a transcription factor that turns out to be a critical activator of the VEGFR1 gene promoter (Jinnin et al. 2008). Finally, the lack of NFATc2 activation was traced back to an abnormality in an endothelial cellsurface complex, consisting of VEGFR2, β1 integrin, TEM8 (tumor endothelial marker 8), and other unidentified proteins (Jinnin et al. 2008). The discovery of this complex was made possible by a screen for somatic and/or germline mutations in endothelial cells from the lesions of several unrelated hemangioma patients. Based on evidence supporting a genetic component in hemangioma formation, such as different frequencies in different ethnic groups (Chiller et al. 2002; Haggstrom et al. 2007), rare cases of familial hemangioma

161

with transmission patterns suggesting autosomal dominant inheritance (Blei et al. 1998), and our own studies of clonality of hemangioma endothelial cells (Boye et al. 2001), we sequenced 24 candidate genes and found heterozygosity for germline mutations in two genes in three of nine hemangioma patients. In one patient, a missense mutation (Ala to Thr) in the transmembrane domain of TEM8 could be shown to have a dominant negative effect on the ability of endothelial cells to activate NFATc2mediated VEGFR1 expression. Viral-mediated expression of mutant TEM8 was sufficient to induce a hemangioma phenotype in control endothelial cells, whereas viralmediated expression of wild-type TEM8 in hemangioma endothelial cells, even cells from patients with no mutations in TEM8, was sufficient to induce a control endothelial phenotype. In two other patients, a Cys to Arg missense mutation was found in the extracellular domain of VEGFR2. This mutation had no effect on receptor synthesis, membrane trafficking, and VEGF-dependent phosphorylation, but it eliminated the ability of the receptor to induce VEGFR1 transcription following viral-mediated expression in hemangioma endothelial cells. This suggests that the Cys to Arg missense mutation in VEGFR2 is a disease-causing loss-of-function mutation in the context of hemangioma. The effects of the mutations in TEM8 and VEGFR2 strongly implied a functional link between the two proteins, and additional data also indicated a role for β1 integrin. TEM8, a transmembrane receptor also known as anthrax toxin receptor 1, has an integrin-like extracellular domain that may bind ECM components such as collagen I (Hotchkiss et al. 2005) and the C5 domain of α3(VI) collagen (Nanda et al. 2004) and might function as an adhesion molecule connecting extracellular ligands to the actin cytoskeleton in the support of cell spreading (Werner et al. 2006). When control endothelial cells are stimulated with a β1 integrin ligand or VEGF-A, the VEGFR2/β1 integrin/TEM8 complex induces the activation of NFATc2 and VEGFR1 transcription in a TEM8-dependent manner (Jinnin et al. 2008). By controlling the level of expression of the VEGF decoy receptor VEGFR1, the function of the complex in normal cells is therefore to put a brake on VEGFR2-dependent VEGF signaling when cells are attached to ECM or when levels of VEGF-A are increased. The role of TEM8 within the complex may be to integrate input from the interaction of VEGF-A with VEGFR2 and of matrix ligands with β1 integrin. Interestingly, the TEM8 primary transcript gives rise to three different mRNA splice variants, and one of these variants (variant 3) lacks the transmembrane and cytoplasmic domains of the full-length protein (Bradley et al. 2001). This variant has the same dominant negative effect on VEGFR1 expression as mutant TEM8, raising the possibility that cells may be able to

162

control the efficiency of the VEGFR2/β1 integrin/TEM8 complex by regulating the expression level of variant 3. In hemangioma endothelial cells, the ability of the complex to activate NFATc2 and VEGFR1 is compromised, and β1 integrin appears to be “trapped” in an inactive conformation (Jinnin et al. 2008; Fig. 3). This is the case for all hemangioma endothelial cells that we have analyzed, even those derived from patients in whom we have found no mutations in TEM8 or VEGFR2. Thus, we believe that hemangioma-associated mutations are likely to be found in other components or regulators of the VEGFR2/β1 integrin/ TEM8 complex. The TEM8 and VEGFR2 mutations affect the ability of the VEGFR2/β1 integrin/TEM8 complex to induce NFATc2-dependent VEGFR1 transcription and thus negatively control VEGFR2-dependent VEGF-A signaling. Since the mutations are germline mutations, they do not explain the clonality of endothelial cells within hemangioma lesions. They must therefore be considered risk factor

Cell Tissue Res (2010) 339:155–165

mutations that require a local triggering event to induce the local lesion. This trigger may be the result of a somatic mutation, as has recently been found in another vascular disorder, venous malformations (Vikkula et al. 1996). In familial venous malformations, the abnormal localized growths of cutaneous venous channels are associated with heterozygous activating germline mutations in the endothelial tyrosine kinase receptor TIE2 and somatic loss-offunction mutations in the wild-type allele of the same gene (Limaye et al. 2009). Thus, all endothelial cells in an affected individual are heterozygous for the TIE2 germline mutation, and this mutation is required for the lesion to develop but is not sufficient, because lesions develop only when the wild-type allele is inactivated by a somatic mutation. Although a full understanding of all the mechanistic details in proliferating hemangiomas will require additional studies, and although little is known about the processes that result in hemangioma involution, the recent work provides a logical basis for the treatment of rapidly growing and clinically problematic hemangiomas. Since the direct cause of endothelial cell activation in the tumor is VEGFA-dependent VEGFR2 activation, local administration of neutralizing antibodies against the extracellular target VEGF-A, already in clinical trials for targeting angiogenesis in malignant tumors, is now also being considered for hemangioma.

Concluding remarks Based on our own research experience, we have tried here to provide case studies that may help the reader rethink the concepts of ECM and matrix disease. Many other stories of exciting discoveries and unexpected therapeutic potential could have been added, but for space reasons, we have to invite the reader to continue this exploration on her (his) own. In doing so, we suggest that time spent with recent work on Marfan syndrome (Habashi et al. 2006; Matt et al. 2009; Olsen 2006), collagen-VI-based myopathy (Angelin et al. 2007; Merlini et al. 2008; Olsen 2007; Palma et al. 2009; Tiepolo et al. 2009), and Alzheimer’s disease (Bellucci et al. 2007; Cheng et al. 2009; Osada et al. 2005; Soderberg et al. 2005; Tobinick and Gross 2008) will be interesting. Fig. 3 Molecular abnormalities in hemangioma endothelial cells. Missense mutations in either tumor endothelial marker 8 (TEM8) or VEGF receptor 2 (VEGFR2) result in the inactivation of β1 integrin and the suppression of the ability of the VEGFR2/β1 integrin/TEM8 complex to activate nuclear factor of activated T cells (NFAT) in hemangioma cells. Reduced translocation of NFAT to the nucleus leads, in turn, to reduced transcription of the VEGF-A decoy receptor (VEGFR1) and low levels of VEGFR1 at the cell surface. VEGFmediated signaling through VEGFR2 is therefore constitutively high (modified from Boye and Olsen 2009)

References Aliprantis AO, Ueki Y, Sulyanto R, Park A, Sigrist KS, Sharma SM, Ostrowski MC, Olsen BR, Glimcher LH (2008) NFATc1 in mice represses osteoprotegerin during osteoclastogenesis and disso-

Cell Tissue Res (2010) 339:155–165 ciates systemic osteopenia from inflammation in cherubism. J Clin Invest 118:3775–3789 Angelin A, Tiepolo T, Sabatelli P, Grumati P, Bergamin N, Golfieri C, Mattioli E, Gualandi F, Ferlini A, Merlini L, Maraldi NM, Bonaldo P, Bernardi P (2007) Mitochondrial dysfunction in the pathogenesis of Ullrich congenital muscular dystrophy and prospective therapy with cyclosporins. Proc Natl Acad Sci USA 104:991–996 Barnes CM, Huang S, Kaipainen A, Sanoudou D, Chen EJ, Eichler GS, Guo Y, Yu Y, Ingber DE, Mulliken JB, Beggs AH, Folkman J, Fishman SJ (2005) Evidence by molecular profiling for a placental origin of infantile hemangioma. Proc Natl Acad Sci USA 102:19097–19102 Bellucci C, Lilli C, Baroni T, Parnetti L, Sorbi S, Emiliani C, Lumare E, Calabresi P, Balloni S, Bodo M (2007) Differences in extracellular matrix production and basic fibroblast growth factor response in skin fibroblasts from sporadic and familial Alzheimer’s disease. Mol Med 13:542–550 Black RA, Rauch CT, Kozlosky CJ, Peschon JJ, Slack JL, Wolfson MF, Castner BJ, Stocking KL, Reddy P, Srinivasan S, Nelson N, Boiani N, Schooley KA, Gerhart M, Davis R, Fitzner JN, Johnson RS, Paxton RJ, March CJ, Cerretti DP (1997) A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells. Nature 385:729–733 Blei F, Walter J, Orlow SJ, Marchuk DA (1998) Familial segregation of hemangiomas and vascular malformations as an autosomal dominant trait. Arch Dermatol 134:718–722 [erratum in Arch Dermatol 134:1425] Boon LM, MacDonald DM, Mulliken JB (1999) Complications of systemic corticosteroid therapy for problematic hemangioma. Plast Reconstr Surg 104:1616–1623 Boye E, Olsen BR (2009) Signaling mechanisms in infantile hemangioma. Curr Opin Hematol 16:202–208 Boye E, Yu Y, Paranya G, Mulliken JB, Olsen BR, Bischoff J (2001) Clonality and altered behavior of endothelial cells from hemangiomas. J Clin Invest 107:745–752 Bradley JR (2008) TNF-mediated inflammatory disease. J Pathol 214:149–160 Bradley KA, Mogridge J, Mourez M, Collier RJ, Young JA (2001) Identification of the cellular receptor for anthrax toxin. Nature 414:225–229 Brennan FM, McInnes IB (2008) Evidence that cytokines play a role in rheumatoid arthritis. J Clin Invest 118:3537–3545 Chen R, Gao B, Huang C, Olsen B, Rotundo RF, Blumenstock F, Saba TM (2000) Transglutaminase-mediated fibronectin multimerization in lung endothelial matrix in response to TNF-alpha. Am J Physiol Lung Cell Mol Physiol 279:L161–L174 Chen CC, Young JL, Monzon RI, Chen N, Todorovic V, Lau LF (2007) Cytotoxicity of TNFalpha is regulated by integrinmediated matrix signaling. EMBO J 26:1257–1267 Cheng JS, Dubal DB, Kim DH, Legleiter J, Cheng IH, Yu GQ, Tesseur I, Wyss-Coray T, Bonaldo P, Mucke L (2009) Collagen VI protects neurons against Abeta toxicity. Nat Neurosci 12:119–121 Chiller KG, Passaro D, Frieden IJ (2002) Hemangiomas of infancy: clinical characteristics, morphologic subtypes, and their relationship to race, ethnicity, and sex. Arch Dermatol 138:1567–1576 Cicchetti P, Mayer BJ, Thiel G, Baltimore D (1992) Identification of a protein that binds to the SH3 region of Abl and is similar to Bcr and GAP-rho. Science 257:803–806 Clark IM, Swingler TE, Sampieri CL, Edwards DR (2008) The regulation of matrix metalloproteinases and their inhibitors. Int J Biochem Cell Biol 40:1362–1378 Deckert M, Tartare-Deckert S, Couture C, Mustelin T, Altman A (1996) Functional and physical interactions of Syk family kinases with the Vav proto-oncogene product. Immunity 5:591–604

163 Deckert M, Tartare-Deckert S, Hernandez J, Rottapel R, Altman A (1998) Adaptor function for the Syk kinases-interacting protein 3BP2 in IL-2 gene activation. Immunity 9:595–605 Dong S, Cole GJ, Halfter W (2003) Expression of collagen XVIII and localization of its glycosaminoglycan attachment sites. J Biol Chem 278:1700–1707 Duh EJ, Yao YG, Dagli M, Goldberg MF (2004) Persistence of fetal vasculature in a patient with Knobloch syndrome: potential role for endostatin in fetal vascular remodeling of the eye. Ophthalmology 111:1885–1888 Enjolras O, Breviere GM, Roger G, Tovi M, Pellegrino B, Varotti E, Soupre V, Picard A, Leverger G (2004) Vincristine treatment for function- and life-threatening infantile hemangioma. Arch Pediatr 11:99–107 Ezekowitz RA, Mulliken JB, Folkman J (1992) Interferon alfa-2a therapy for life-threatening hemangiomas of infancy. N Engl J Med 326:1456–1463 Fanjul-Fernandez M, Folgueras AR, Cabrera S, Lopez-Otin C (2009) Matrix metalloproteinases: evolution, gene regulation and functional analysis in mouse models. Biochim Biophys Acta (in press) Fawcett SL, Grant I, Hall PN, Kelsall AW, Nicholson JC (2004) Vincristine as a treatment for a large haemangioma threatening vital functions. Br J Plast Surg 57:168–171 Foucault I, Liu YC, Bernard A, Deckert M (2003) The chaperone protein 14–3-3 interacts with 3BP2/SH3BP2 and regulates its adapter function. J Biol Chem 278:7146–7153 Foucault I, Le Bras S, Charvet C, Moon C, Altman A, Deckert M (2005) The adaptor protein 3BP2 associates with VAV guanine nucleotide exchange factors to regulate NFAT activation by the B-cell antigen receptor. Blood 105:1106–1113 Fukai N, Eklund L, Marneros AG, Oh SP, Keene DR, Tamarkin L, Li E, Pihlajaniemi T, Olsen BR (2002) Lack of collagen XVIII/ endostatin results in eye abnormalities. EMBO J 21:1535–1544 Gao B, Curtis TM, Blumenstock FA, Minnear FL, Saba TM (2000) Increased recycling of (alpha)5(beta)1 integrins by lung endothelial cells in response to tumor necrosis factor. J Cell Sci 113:247–257 Habashi JP, Judge DP, Holm TM, Cohn RD, Loeys BL, Cooper TK, Myers L, Klein EC, Liu G, Calvi C, Podowski M, Neptune ER, Halushka MK, Bedja D, Gabrielson K, Rifkin DB, Carta L, Ramirez F, Huso DL, Dietz HC (2006) Losartan, an AT1 antagonist, prevents aortic aneurysm in a mouse model of Marfan syndrome. Science 312:117–121 Haggstrom AN, Drolet BA, Baselga E, Chamlin SL, Garzon MC, Horii KA, Lucky AW, Mancini AJ, Metry DW, Newell B, Nopper AJ, Frieden IJ (2007) Prospective study of infantile hemangiomas: demographic, prenatal, and perinatal characteristics. J Pediatr 150:291–294 Halfter W, Dong S, Schurer B, Cole GJ (1998) Collagen XVIII is a basement membrane heparan sulfate proteoglycan. J Biol Chem 273:25404–25412 Hidano A, Nakajima S (1972) Earliest features of the strawberry mark in the newborn. Br J Dermatol 87:138–144 Hotchkiss KA, Basile CM, Spring SC, Bonuccelli G, Lisanti MP, Terman BI (2005) TEM8 expression stimulates endothelial cell adhesion and migration by regulating cell-matrix interactions on collagen. Exp Cell Res 305:133–144 Imai Y, Kanno K, Moriya T, Kayano S, Seino H, Matsubara Y, Yamada A (2003) A missense mutation in the SH3BP2 gene on chromosome 4p16.3 found in a case of nonfamilial cherubism. Cleft Palate Craniofac J 40:632–638 Jinnin M, Medici D, Park L, Limaye N, Liu Y, Boscolo E, Bischoff J, Vikkula M, Boye E, Olsen BR (2008) Suppressed NFATdependent VEGFR1 expression and constitutive VEGFR2 signaling in infantile hemangioma. Nat Med 14:1236–1246

164 Jones WA (1933) Familial multilocular cystic disease of the jaws. Am J Cancer 17:946–950 Knobloch WH, Layer JM (1972) Clefting syndromes associated with retinal detachment. Am J Ophthalmol 73:517–530 Kriegler M, Perez C, DeFay K, Albert I, Lu SD (1988) A novel form of TNF/cachectin is a cell surface cytotoxic transmembrane protein: ramifications for the complex physiology of TNF. Cell 53:45–53 Leaute-Labreze C, Dumas de la Roque E, Hubiche T, Boralevi F, Thambo JB, Taieb A (2008) Propranolol for severe hemangiomas of infancy. N Engl J Med 358:2649–2651 Li CY, Yu SF (2006) A novel mutation in the SH3BP2 gene causes cherubism: case report. BMC Med Genet 7:84 Lietman SA, Kalinchinko N, Deng X, Kohanski R, Levine MA (2006) Identification of a novel mutation of SH3BP2 in cherubism and demonstration that SH3BP2 mutations lead to increased NFAT activation. Hum Mutat 27:717–718 Limaye N, Wouters V, Uebelhoer M, Tuominen M, Wirkkala R, Mulliken JB, Eklund L, Boon LM, Vikkula M (2009) Somatic mutations in angiopoietin receptor gene TEK cause solitary and multiple sporadic venous malformations. Nat Genet 41:118–124 Lo B, Faiyaz-Ul-Haque M, Kennedy S, Aviv R, Tsui LC, Teebi AS (2003) Novel mutation in the gene encoding c-Ablbinding protein SH3BP2 causes cherubism. Am J Med Genet 121A:37–40 Mangion J, Rahman N, Edkins S, Barfoot R, Nguyen T, Sigurdsson A, Townend JV, Fitzpatrick DR, Flanagan AM, Stratton MR (1999) The gene for cherubism maps to chromosome 4p16.3. Am J Hum Genet 65:151–157 Marneros AG, Olsen BR (2003) Age-dependent iris abnormalities in collagen XVIII/endostatin deficient mice with similarities to human pigment dispersion syndrome. IOVS 44:2367–2372 Marneros AG, Keene DR, Hansen U, Fukai N, Moulton K, Goletz PL, Moiseyev G, Pawlyk BS, Halfter W, Dong S, Shibata M, Li T, Crouch RK, Bruckner P, Olsen BR (2004) Collagen XVIII/ endostatin is essential for vision and retinal pigment epithelial function. EMBO J 23:89–99 Marneros AG, She H, Zambarakji H, Hashizume H, Connoly EJ, Kim I, Gragoudas ES, Miller JW, Olsen BR (2007) Endogenous endostatin inhibits choroidal neovascularization. FASEB J 21:3809–3818 Matt P, Schoenhoff F, Habashi J, Holm T, Van Erp C, Loch D, Carlson OD, Griswold BF, Fu Q, De Backer J, Loeys B, Huso DL, McDonnell NB, Van Eyk JE, Dietz HC (2009) Circulating transforming growth factor-beta in Marfan syndrome. Circulation 120:526–532 Merlini L, Angelin A, Tiepolo T, Braghetta P, Sabatelli P, Zamparelli A, Ferlini A, Maraldi NM, Bonaldo P, Bernardi P (2008) Cyclosporin A corrects mitochondrial dysfunction and muscle apoptosis in patients with collagen VI myopathies. Proc Natl Acad Sci USA 105:5225–5229 Mulliken JB, Glowacki J (1982) Hemangiomas and vascular malformations in infants and children: a classification based on endothelial characteristics. Plast Reconstr Surg 69:412–422 Muragaki Y, Timmons S, Griffith CM, Oh SP, Fadel B, Quertermous T, Olsen BR (1995) Mouse Col18a1 is expressed in a tissue-specific manner as three alternative variants and is localized in basement membrane zones. Proc Natl Acad Sci USA 92:8763–8767 Nanda A, Carson-Walter EB, Seaman S, Barber TD, Stampfl J, Singh S, Vogelstein B, Kinzler KW, St Croix B (2004) TEM8 interacts with the cleaved C5 domain of collagen alpha 3(VI). Cancer Res 64:817–820 North PE, Waner M, Mizeracki A, Mihm MC Jr (2000) GLUT1: a newly discovered immunohistochemical marker for juvenile hemangiomas. Hum Pathol 31:11–22

Cell Tissue Res (2010) 339:155–165 North PE, Waner M, Mizeracki A, Mrak RE, Nicholas R, Kincannon J, Suen JY, Mihm MC Jr (2001) A unique microvascular phenotype shared by juvenile hemangiomas and human placenta. Arch Dermatol 137:559–570 O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E, Birkhead JR, Olsen BR, Folkman J (1997) Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell 88:277–285 Olsen BR (2006) From the editor’s desk. Matrix Biol 25:269–270 Olsen BR (2007) From the editor’s desk. Matrix Biol 26:145 Osada Y, Hashimoto T, Nishimura A, Matsuo Y, Wakabayashi T, Iwatsubo T (2005) CLAC binds to amyloid beta peptides through the positively charged amino acid cluster within the collagenous domain 1 and inhibits formation of amyloid fibrils. J Biol Chem 280:8596–8605 Palma E, Tiepolo T, Angelin A, Sabatelli P, Maraldi NM, Basso E, Forte MA, Bernardi P, Bonaldo P (2009) Genetic ablation of cyclophilin D rescues mitochondrial defects and prevents muscle apoptosis in collagen VI myopathic mice. Hum Mol Genet 18:2024–2031 Read TA, Sorensen DR, Mahesparan R, Enger PO, Timpl R, Olsen BR, Hjelstuen MH, Haraldseth O, Bjerkvig R (2001) Local endostatin treatment of gliomas administered by microencapsulated producer cells. Nat Biotech 19:29–34 Ren R, Mayer BJ, Cicchetti P, Baltimore D (1993) Identification of a ten-amino acid proline-rich SH3 binding site. Science 259:1157–1161 Rehn M, Hintikka E, Pihlajaniemi T (1996) Characterization of the mouse gene for the alpha 1 chain of type XVIII collagen (Col18a1) reveals that the three variant N-terminal polypeptide forms are transcribed from two widely separated promoters. Genomics 32:436–446 Rodriguez C, Alcudia JF, Martinez-Gonzalez J, Raposo B, Navarro MA, Badimon L (2008) Lysyl oxidase (LOX) down-regulation by TNFalpha: a new mechanism underlying TNFalpha-induced endothelial dysfunction. Atherosclerosis 196:558–564 Rotundo RF, Curtis TM, Shah MD, Gao B, Mastrangelo A, LaFlamme SE, Saba TM (2002) TNF-alpha disruption of lung endothelial integrity: reduced integrin mediated adhesion to fibronectin. Am J Physiol Lung Cell Mol Physiol 282:L316–L329 Saarela J, Rehn M, Oikarinen A, Autio-Harmainen H, Pihlajaniemi T (1998a) The short and long forms of type XVIII collagen show clear tissue specificities in their expression and location in basement membrane zones in humans. Am J Pathol 153:611–626 Saarela J, Ylikarppa R, Rehn M, Purmonen S, Pihlajaniemi T (1998b) Complete primary structure of two variant forms of human type XVIII collagen and tissue-specific differences in the expression of the corresponding transcripts. Matrix Biol 16:319–328 Sertié AL, Quimby M, Moreira ES, Murray J, Zatz M, Antonarakis SE, Passos-Bueno MR (1996) A gene which causes severe ocular alterations and occipital encephalocele (Knobloch syndrome) is mapped to 21q22.3. Hum Mol Genet 5:843–847 Sertié AL, Sossi V, Camargo AA, Zatz M, Brahe C, Passos-Bueno MR (2000) Collagen XVIII, containing an endogenous inhibitor of angiogenesis and tumor growth, plays a critical role in the maintenance of retinal structure and in neural tube closure. Hum Mol Genet 9:2051–2058 Soderberg L, Dahlqvist C, Kakuyama H, Thyberg J, Ito A, Winblad B, Naslund J, Tjernberg LO (2005) Collagenous Alzheimer amyloid plaque component assembles amyloid fibrils into protease resistant aggregates. FEBS J 272:2231–2236 Suzuki O, Kague E, Bagatini K, Tu H, Heljasvaara R, Carvalhaes L, Gava E, Oliveira G de, Godoi P, Oliva G, Kitten G, Pihlajaniemi T, Passos-Bueno MR (2009) Novel pathogenic mutations and skin biopsy analysis in Knobloch syndrome. Mol Vis 15:801–809

Cell Tissue Res (2010) 339:155–165 Tang P, Hung MC, Klostergaard J (1996) Human pro-tumor necrosis factor is a homotrimer. Biochemistry 35:8216–8225 Tiepolo T, Angelin A, Palma E, Sabatelli P, Merlini L, Nicolosi L, Finetti F, Braghetta P, Vuagniaux G, Dumont JM, Baldari CT, Bonaldo P, Bernardi P (2009) The cyclophilin inhibitor Debio 025 normalizes mitochondrial function, muscle apoptosis and ultrastructural defects in Col6a1(-/-) myopathic mice. Br J Pharmacol 157:1045–1052 Tiziani V, Reichenberger E, Buzzo CL, Niazi S, Fukai N, Stiller M, Peters H, Salzano FM, Raposo do Amaral CM, Olsen BR (1999) The gene for cherubism maps to chromosome 4p16. Am J Hum Genet 65:158–166 Tobinick EL, Gross H (2008) Rapid improvement in verbal fluency and aphasia following perispinal etanercept in Alzheimer’s disease. BMC Neurol 8:27 Ueki Y, Tiziani V, Santanna C, Fukai N, Maulik C, Garfinkle J, Ninomiya C, Raposo do Amaral CM, Peters H, Habal M, Rhee-Morris L, Doss JB, Kreiborg S, Olsen BR, Reichenberger E (2001) Mutations in the gene encoding c-Ablbinding protein SH3BP2 cause cherubism. Nat Genet 28:125–126 Ueki Y, Lin CY, Senoo M, Ebihara T, Agata N, Onji M, Saheki Y, Kawai T, Mukherjee PM, Reichenberger E, Olsen BR (2007) Increased myeloid cell responses to M-CSF and RANKL cause

165 bone loss and inflammation in SH3BP2 “cherubism” mice. Cell 128:71–83 Vikkula M, Boon LM, Carraway K, Calvert JT, Diamonti AJ, Goumnerov B, Pasyk KA, Marchuk DA, Warman ML, Cantley LC, Mulliken JB, Olsen BR (1996) Vascular dysmorphogenesis caused by an activating mutation in the receptor tyrosine kinase TIE2. Cell 87:1181–1190 Werner E, Kowalczyk AP, Faundez V (2006) Anthrax toxin receptor 1/tumor endothelium marker 8 mediates cell spreading by coupling extracellular ligands to the actin cytoskeleton. J Biol Chem 281:23227–23236 Yamaguchi N, Anand-Apte B, Lee M, Sasaki T, Fukai N, Shapiro R, Que I, Lowik C, Timpl R, Olsen BR (1999) Endostatin inhibits VEGF-induced endothelial cell migration and tumor growth independently of zinc binding. EMBO J 18:4414–4423 Ylikärppä R, Eklund L, Sormunen R, Utriainen A, Kontiola A, Määttä M, Muona A, Fukai N, Olsen BR, Pihlajaniemi T (2003) Lack of type XVIII collagen results in anterior ocular defects. FASEB J 17:2257–2259 Zollino M, Di Stefano C, Zampino G, Mastroiacovo P, Wright TJ, Sorge G, Selicorni A, Tenconi R, Zappala A, Battaglia A, Di Rocco M, Palka G, Pallotta R, Altherr MR, Neri G (2000) Genotype-phenotype correlations and clinical diagnostic criteria in Wolf-Hirschhorn syndrome. Am J Med Genet 94:254–261