Multicentric osteolysis, nodulosis, and arthropathy (OMIM no. 605156) is a rare osteolytic syndrome, which was originally described by Al-Aqeel et al3
and Al-Mayouf et al4
in a number of Saudi Arabian families. This condition is primarily characterized by the development of severe osteoporosis and flexion contractures of interphalangeal joints, with fusiform swelling and tenderness of the hands and feet, palmar and plantar subcutaneous nodules in the first years of life, and distinctive facies.3, 4
A number of additional nonosseous features, including hirsutism, have also been described.3
Given the rarity of the syndrome, it is often difficult to prove causation of the novel nonosseous features unless confirmed in additional families.
In this report, we describe progressive distal arthropathy, palmar and plantar subcutaneous nodules, facial changes, osteolysis of bones of hands and feet, and hirsutism in a 6-year-old boy (IV-2) and a 4-year-old girl (IV-6) who were second cousins. The 4-year-old girl (IV-6) demonstrated a more severe phenotype than that of her male cousin (IV-2). Both developed symptoms in the first year of life. Patient 1 (IV-2) had fusiform painful swellings on all fingers and camptodactyly of the fourth finger of the left hand and the fifth finger of the right hand, whereas patient 2 (IV-6) had painful fusiform swellings with hyperextension of the metacarpophalangeal joints in all fingers and flexion contractures of the interphalangeal joints. Both children had similar facial features, which were notable for proptosis, hypertelorism, bulbous nose, thick lower lip, gum hypertrophy, and large and low-set ears. Radiological evaluation revealed a generalized osteopenia, fusiform swelling of fingers with flexion of interphalangeal joints, and osteolysis of carpal and tarsal bones.
The patients were started on a trial of prednisolone therapy and, after 6 months, the number of painful swellings and nodules decreased and camptodactyly of the fingers improved in patient 1. This therapeutic intervention was, however, unsuccessful in patient 2, who had a more severe phenotype. We restarted prednisolone therapy and reexamined her after 1 year.
In addition, and possibly related to the syndrome, patient 1 (IV-2) had attention deficit hyperactivity disorder (ADHD), whereas patient 2 (IV-6) had corneal clouding. Corneal clouding and gum hypertrophy has not been reported previously in the MONA literature ().
Table 1 Comparison of clinical and radiological features of the present cases with those originally reported by Al Aqeel et al and Al Mayouf et al3, 4
Most striking, however, was that 3 children in this consanguineous family all shared cardiac defects. Patient 1 (IV-2) had a BAV and his older sister, who on the basis of history, facial features, and radiological examination we believe to also have shared the same MMP-2 mutation, died from complications associated with TGA. Patient 2 (IV-6) had both an ASD and VSD. We are aware of only one other reported child with MONA who has been described as having ‘heart disease';4
however, no information was provided on the nature of the defect. Given the preponderance of congenital heart defects clustering within this single family, and despite the apparently discordant anatomic features – TGA, BAV, ASD, and VSD – we nonetheless suggest that this cardiac association may represent a previously unrecognized component of MONA. At this time, however, the less likely possibilities that this family may also have had an inherited form of BAV or a secondary genetic defect associated with cardiac defects cannot be ruled out.
As has been recently reviewed,14
the increasing recognition of single gene defects underlying congenital heart disease is consistent with our hypothesis that the range of cardiac features in this family may be related to a single gene defect. For example, mutations in a number of cooperative transcription factors, including TBX5,15
have now been shown to result in cardiac septation defects during early development. Of these, mutations in TBX5 have also been associated with outflow tract defects, including TGA.18
In addition to transcription factors, mutations in signaling molecules also result in cardiac defects. In particular, mutations in Notch1, which also plays an important role in osteoblast differentiation,19
have been associated with multiple defects in valvulogenesis, including the development of BAV.20
Finally, mutations in genes along a single signaling pathway can also result in cardiac defects. Mutations in the ras-mitogen-activated protein-kinase signaling pathway have been demonstrated to underlie Noonan's syndrome, an autosomal dominant cardio-facio-cutaneous syndrome, associated with septation defects and pulmonary valve defects (reviewed by Gelb and Tartaglia21
Although an increasing literature is establishing an association between the MMPs and cardiac function, most have focused on their possible role in coronary and carotid atherosclerosis (reviewed by Abilleira et al22
). Despite the recent demonstration that MMP-2-deficient mice display features of MONA,11, 12
no reports have described cardiac defects in these mice. If cardiac defects do occur in MMP-2-deficient mice, early lethality may result and possibly explain our finding of lower-than-expected numbers of surviving Mmp2
−/− mice obtained in heterozygous crosses (Mosig RA and Martignetti JA, personal communication, 2008). Two studies, on the basis of MMP-2 inhibition, also support our hypothesis for a role of MMP-2 in cardiac formation. First is a study examining developing chick hearts following targeted inhibition of MMP-2. In these studies, at specific early developmental time points, either whole chick embryos were exposed to MMP-2 neutralizing antibodies or these antibodies were directly microinjected in the chick heart fields. Beating cardiac tissues developed, but were differentially associated with defects in ventral tissue fusion and directional defects in looping.23
Second, inhibition of MMP-2 also blocked the epithelial–mesenchymal transition, migration, and invasion of explanted mouse atrioventricular canal cells involved in the endocardial cushion formation necessary for proper heart development.24
Although the defects in chick and mouse do not prove the causality of the defects identified in our patients, they are nonetheless supportive for a role of MMP-2 in heart organogenesis. On the basis of these accumulated findings, and until additional MMP-2-deficient patients with cardiac defects are identified or MMP-2-deficient mice are shown to possess similar developmental defects, we suggest that newly identified patients be carefully screened for cardiac disease.
The inactivating mutation identified in this family is also of interest. The majority of previously reported MMP2
mutations have been postulated to directly affect enzymatic function through disruption of either the catalytic site itself or activation mechanism through mutations in the cysteine-switch domain.6, 7, 8, 9, 10
In general, the hemopexin domain is responsible for dimerization, activation, inhibition, and other functions for numerous proteins across many species.25
Specifically, the MMP-2 hemopexin domain is believed to be critical to a number of higher-order interactions. The two deleted hemopexin domains are believed critical for MMP-2 activation and inhibition, as they are responsible for interaction of several molecules, most notably TIMP-2.13
The activation mechanism of MMP-2 requires binding of the hemopexin domain by TIMP-2 tethered to MTI-MMP, which coordinates the cleavage of the MMP-2 propeptide by a second MTI-MMP molecule.26
Previous studies using site-directed mutagenesis demonstrated that the third and fourth hemopexin repeats are largely responsible for this interaction.26
Specifically, hemopexin domain 3 interacts with the C-terminal ‘tail' of TIMP-2, and hemopexin domain 4 interacts with the GH loop of TIMP-2. A number of other functions have also been assigned to this region and include the following: the C-terminal PEX domain's inhibitory interaction with integrin α
3 on the surface of angiogenic blood vessels;26
requirement of the hemopexin domain for cleavage of collagen type 1, the largest constituent of bone ECM;27, 28
and cleavage of both monocyte chemoattractant protein-3 and stromal-derived factor-1 into antagonists.29, 30
Taken together, these findings in a Turkish family with MONA not only highlight a potential cardiac involvement to the syndrome, but also highlight the importance of the terminal portion of the MMP-2 domain. In turn, the clinical findings may prove helpful in providing better prospective patient care and in understanding the physiologic role of MMP-2 in human development. The molecular findings may clarify which substrates and MMP-2 activities are critical in bone, joint, and possibly cardiac biology, and in the development of this syndrome.