An ADAMTSL2 founder mutation causes Musladin-Lueke Syndrome, a heritable disorder of beagle dogs, featuring stiff skin and joint contractures

Abstract

Background: Musladin-Lueke Syndrome (MLS) is a hereditary disorder affecting Beagle dogs that manifests with extensive fibrosis of the skin and joints. In this respect, it resembles human stiff skin syndrome and the Tight skin mouse, each of which is caused by gene defects affecting fibrillin-1, a major component of tissue microfibrils. The objective of this work was to determine the genetic basis of MLS and the molecular consequence of the identified mutation.

Methodology and principal findings: We mapped the locus for MLS by genome-wide association to a 3.05 Mb haplotype on canine chromosome 9 (CFA9 (50.11-54.26; p(raw) <10(-7))), which was homozygous and identical-by-descent among all affected dogs, consistent with recessive inheritance of a founder mutation. Sequence analysis of a candidate gene at this locus, ADAMTSL2, which is responsible for the human TGFβ dysregulation syndrome, Geleophysic Dysplasia (GD), uncovered a mutation in exon 7 (c.660C>T; p.R221C) perfectly associated with MLS (p-value=10(-12)). Murine ADAMTSL2 containing the p.R221C mutation formed anomalous disulfide-bonded dimers when transiently expressed in COS-1, HEK293F and CHO cells, and was present in the medium of these cells at lower levels than wild-type ADAMTSL2 expressed in parallel.

Conclusions/significance: The genetic basis of MLS is a founder mutation in ADAMTSL2, previously shown to interact with latent TGF-β binding protein, which binds fibrillin-1. The molecular effect of the founder mutation on ADAMTSL2 is formation of disulfide-bonded dimers. Although caused by a distinct mutation, and having a milder phenotype than human GD, MLS nevertheless offers a new animal model for study of GD, and for prospective insights on mechanisms and pathways of skin fibrosis and joint contractures.

Figure 1. Phenotypic presentation of MLS in Beagle dogs.

(A) Non-affected (left) and affected (right) Beagles matched by age, sex, and variety. Note short stature of affected dog. (B) Affected dog showing ear crease (arrow) and unusual stiff sitting posture. (C) Same dog as in B showing hyper-digitigrade standing (arrow), which is also noticeable in panel B. The typical facial appearance is seen in B and C. (D). Histology of skin from an MLS case compared to skin from an unaffected dog (WT). Stain: Masson trichrome (collagen = blue). Note the increased intensity and extent of blue staining corresponding to collagen and the presence of collagen strands extending from the dermis to underlying tissue in MLS skin. E, epidermis, D, dermis, SC, subcutaneous fat.

Figure 2. MLS is caused by an ADAMTSL2 mutation.

(A) Statistical significance scores from genome-wide case-control analysis of microsatellite-based genotype data are shown. Included are results for loci used in fine-scale mapping of CFA 9 (n = 23, depicted in blue). Multiple p-values estimated by permutation (10⁷ iterations) were saturated for statistical significance at 10⁻⁷ (red). (B) Fine-mapping revealed a six-marker haplotype shared homozygous among all affected dogs (Table S2). The haplotype encompassed a strong candidate gene, ADAMTSL2 (depicted in red). (C) DNA sequence analysis identified the c.660C>T mutation. Sequence traces from PCR products amplified from genomic DNA are shown for a non-affected (Wt) and affected (Aff) Beagle. The transition (C > T) converts an arginine residue to a cysteine (R221C).

Figure 3. The R221C mutation leads to aberrant ADAMTSL2 dimerization and reduced protein levels.

(A). Wild-type (WT) or R221C murine ADAMTSL2 was used to transfect HEK293F, COS-1 or CHO-K1 cells (each in triplicate). Conditioned media were run under reducing or non reducing conditions on 6% SDS-polyacrylamide gels, and were analyzed by Western blot using an anti-myc antibody to detect ADAMTSL2. In each cell type, both wildtype and mutant ADAMTSL2 migrated under reducing conditions at 140 kDa, corresponding to the monomeric (M) size of glycosylated ADAMTSL2. The 120 kDa species indicated by a double asterisk in CHO K1 medium, corresponds to an unglycosylated form (confirmed by peptide N-glycanase F digestion, data not shown). Under non-reducing conditions, wildtype ADAMTSL2 migrated predominantly at 140 kDa, although two quantitatively minor species >250 kDa were also observed. The 280 kDa band may correspond to a dimer (D). In contrast, R221C ADAMTSL2 migrated in all 3 cell types exclusively as a 280 kDa dimer (D). (B). HEK293F, COS-1 and CHO K1 cells were transiently transfected as described above, and conditioned media and cell lysates were run on 7.5% SDS-polyacrylamide gels under reducing conditions. ADAMTSL2 protein was detected with an anti-myc antibody in both conditioned medium (CM) and cell lysate (CL). In addition, cell lysates were probed with an anti-GAPDH antibody to control for variations in cell numbers. The western blots shown are representative of at least three independent experiments for each cell type. Note that in conditioned media, the level of R221C ADAMTSL2 was reduced compared to wild type ADAMTSL2, whereas the levels in cell lysates were not affected. (C). ECL signal following anti-myc immunoblotting of conditioned media and cell lysates were quantified by densitometry. ADAMTSL2 signal in both conditioned medium and cell lysate was normalized to the GAPDH signal. In conditioned media, the level of mutant protein was significantly reduced, whereas no significant difference was observed in the cell lysates (Student t-test).