Collagen VI involvement in Ullrich syndrome

Patients and methods.

The patients included in this study were assessed and followed up at the Dubowitz Neuromuscular Center at Hammersmith Hospital, London, UK. Patients were included if 1) they presented with clinical features compatible with Ullrich’s diagnostic criteria—namely, proximal joint contractures with severe hyperlaxity of distal joints presenting in early infancy, absence of severe mental retardation, and myopathic or dystrophic changes on muscle biopsy; and 2) muscle biopsy samples were available for collagen VI studies or were part of informative families for linkage to collagen VI.

Muscle biopsies were processed according to standard histologic and histochemical techniques.13 A panel of proteins was also examined immunohistochemically, including dystrophin, sarcoglycans, laminin α2, α5, β1, and γ1 chains, fetal, fast, and slow myosin,14 and collagen V1 (Clones 1944 and 3303, Chemicon, Harrow, UK). The assessment of collagen status was performed blindly to the linkage analysis by two independent observers.

Genomic DNA was extracted from whole blood using standard procedures. Linkage to the two collagen 6 loci (COL6A1/A2 on chromosome 21q22.3 and COL6A3 on chromosome 2q37) was assessed by genotyping subjects for dinucleotide repeat markers either flanking the loci or contained within it. For the COL6A1/2 locus, we used one flanking marker, D21S1259, and two intragenic markers, Col61 and Col6A2.15

For the COL6A3 loci, flanking markers D2S206, D2S345, and D2S338 were used in addition to the Col6A3 intragenic marker.16 Primers amplifying these markers were purchased from Invitrogen (Paisley, UK), with the forward primer modified at the 5′ end by the addition of either a FAM, HEX, or NED fluorescent label. PCR primer sequences were as follows: D21S1259, forward GGGACTGTAATAAATATTCTG, reverse CACTGGCTCTCCTGACC; Col6A1, forward GATCCCAGCTACTCGGGA, reverse CACAACTCCACAGTGACTGT; Col6A2, forward CCCAGATCCAGCCTGATCTG, reverse GATGCAGGCTGGTGAAGCCC;D2S206, forward TTAAAAATTAAGTAG-GCTTTTGGTT, reverse GTCCTCATGTGTTTATGCTGT; D2S345, forward GGAAGCCACCATGAAT, reverse AGATCAACAGACATAACCCA; D2S338, forward AGAGCCCAAAACAAAACTTCC, reverse TGGAATTTTGATT-TTCAGATTTG; and Col6A3 forward AGGTCACAAAGGCTCTTGGC, reverse TTGCCCTCCAAGTCCTTCC. PCR products were amplified using AmpliTaqGold DNA polymerase (PE Biosystems, Warrington, UK) on a 9700 thermal cycler (PE Biosystems) and separated on a 5% denaturing gel (Amresco, Luton, UK) in a ABI 377A automated DNA sequencer (ABI, Warrington, UK). Gels were analyzed using Genescan v3.01 and Genotyper v2.01 software (ABI).

We systematically collected data regarding age at and mode of onset, muscle power, pattern and severity of contractures, functional abilities, and progression of disease. Data on cardiac function by means of EKG and echocardiograms, as well as respiratory function (forced vital capacity and, when appropriate, overnight oxygen saturation monitoring), were also collected.

Results.

Fifteen patients fulfilled the inclusion criteria. Their ages ranged from 3 to 23.5 years. Eleven were sporadic cases and four were familial cases. Two of the four familial cases were first cousins, both born to consanguineous parents (Patients 7 and 8), the other two were twin sisters (Patients 12 and 13). Three of the sporadic cases were from consanguineous families (Patients 1, 2, and 15). Details of the cohort studied are provide in the table.

Muscle biopsy.

Muscle biopsies were available in 11 of the 15 cases and reports from previous biopsies performed in other centers were available in the other four. All patients showed variation in fiber size affecting both fast and slow fibers and type 1 predominance. Eight of the 15 patients also showed increased endomysial connective tissue, necrosis, or evidence of muscle fiber regeneration by the presence of fibers containing fetal myosin.

Of the 11 children assessed for collagen VI by immunocytochemistry, five had normal sarcolemmal and endomysial expression of collagen VI, two showed only very slight traces, three showed a marked reduction (particularly at the sarcolemma), and one showed a mild reduction. Some of this reduction may have been due to poor preservation of a sample stored for many years (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Pedigrees of three consanguineous families showing linkage data. Families 1 and 2 are consistent with linkage to COL6A1/A2 locus on chromosome 21q22.3, whereas Family 3 is consistent with linkage to COL6A3 on chromosome 2q37.

Dystrophin, sarcoglycans, and laminin α2, α5, β1, and γ1 chains were all normal.

Genetic analysis.

Of the 15 families in the cohort, six were informative enough to undertake linkage analysis to the collagen 6 loci. Initial investigations had previously excluded linkage to the other known CMD loci. Two consanguineous families were consistent with linkage to the COL6A1/2 locus and one other consanguineous family to COL6A3 (figure 2). The remaining three informative families were excluded from both loci (see the table).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Control muscle stained for collagen V1 using antibodies 1944 (A) and 3303 (F). Patients 4 (C and H) and 5 (D and I) all showed a reduction in staining with both antibodies relative to control muscle (A and F), whereas Patients 1 (B and G) and 3 (E and J) showed almost absent collagen V1.

The cohort was subdivided in two subgroups according to whether genetic or immunocytochemical data suggested primary collagen VI involvement (see the table).

The first group included six children with reduced collagen VI in muscle on immunohistochemistry. Three of the six came from informative families and genetic studies suggested linkage to one of the collagen VI loci.

The second group included nine children. Five of the nine had muscle samples available for immunohistochemical studies that showed normal collagen VI labeling. One of the five families was informative and linkage to all the collagen VI loci was excluded.

In three patients, muscle samples were not available but genetic studies excluded linkage to the 3 collagen VI loci. The remaining patient was the similarly affected twin sister of one of the patients in whom collagen was found to be normal on muscle biopsy.

Discussion.

All the cases described in this article presented with a combination of proximal contractures, distal hyperextensibility, and abnormal muscle biopsy, as originally described by Ullrich.1 All patients shared a similar clinical history, with onset in the first months of life, a characteristic facial appearance, early rigidity of the spine and respiratory involvement, and absence of severe mental retardation, which makes them easily recognizable.

After recent evidence of collagen VI involvement in UCMD,10-12⇓⇓ we screened all our cases with immunocytochemistry or, in informative families, performed linkage analysis to the three collagen VI loci. We found that only six of the 15 patients studied (40%) had evidence of primary collagen VI involvement, whereas this could not be demonstrated in the remaining nine. Because several patients had a muscle biopsy performed in other centers and in many cases they were from small noninformative families, both genetic tests and immunohistochemistry data were only available in four families. In all four families, there was concordance between the results obtained with the two techniques in the patients with and without collagen VI deficiency.

We found 1) Collagen VI involvement has been confirmed in six families with UCMD; conversely, its involvement has been definitively excluded by linkage in three families. In the remaining five cases, the normal expression of collagen made the primary role of collagen VI unlikely. 2) The clinical features of patients with collagen VI involvement were very similar to those observed in patients without evidence of collagen VI involvement.

Two-thirds of our patients with collagen VI involvement never achieved independent ambulation. However, we also observed that more than 50% of the patients without evidence of collagen VI involvement also never achieved independent ambulation. The maximal functional ability was associated with the age and presentation at onset. The children with more severe contractures at birth and with hip dislocation were also the ones who never achieved independent ambulation, whereas the patients with hypotonia and motor delay in the first months of life generally achieved independent ambulation. Those who achieved independent ambulation usually only showed a mild motor delay and were able to walk before age 2 years. The variability of our findings is in keeping with the two studies that recently described patients with the Ullrich phenotype and recessive mutations in collagen VI.11,12⇓ Whereas in the first study the three patients described were able to walk independently by age 3 years,11 in the other study both children never achieved independent ambulation.

Respiratory impairment was present in both groups, irrespective of the collagen VI status. Previous studies reported that patients affected by UCMD generally experience early severe respiratory infections.3 In our cohort all the children older than age 5 years already had a forced vital capacity below 40% and seven of the 10 patients older than 11 years developed nocturnal hypoventilation and required night-time ventilatory support with noninvasive, facial biphasic inspiratory pressure ventilation. In some cases ventilation had been required at a much younger age. Only one of our patients died, at age 12 years (Patient 6, with collagen VI reduction). Although her vital capacity was low (18%), repeated and recent sleep studies were normal. In her case, a sudden death of cardiac origin raises the question of possible cardiac involvement.

The severity of motor impairment did not correlate with the severity of other clinical signs such as rigidity of the spine, failure to thrive, or respiratory impairment, which were found in most patients. All of these signs were rapidly progressive in all patients, irrespective of the maximal function ability or collagen status.

In agreement with previous reports,2,3,8⇓⇓ we also found the muscle pathology to be variable, ranging from nonspecific, mild myopathic changes to more severe dystrophiclike changes. Fast fibers were not preferentially affected. The severity of the pathologic findings did not always correlate with the severity of the clinical phenotype because dystrophic changes could be observed in both ambulant and nonambulant children.