Genotype-phenotype investigation of 35 patients from 11 unrelated families with camptodactyly-arthropathy-coxa vara-pericarditis (CACP) syndrome

Abstract

Background: The camptodactyly-arthropathy-coxa vara-pericarditis syndrome (CACP) is a rare autosomal recessive condition characterized by camptodactyly, noninflammatory arthropathy, coxa vara, and pericarditis. CACP is caused by mutations in the proteoglycan 4 (PRG4) gene, which encodes a lubricating glycoprotein present in the synovial fluid and at the surface of articular cartilage.

Methods: In the present study, we compared the clinical and molecular findings of CACP syndrome in 35 patients from 11 unrelated families. In 28 patients, whole exome sequencing was used to investigate genomic variations.

Results: We found that camptodactyly of hands was the first symptom presented by most patients. Swelling of wrists, knees, and elbows began before 4 years of age, while the age of joint involvement was variable. Patients reported an increased pain level after the age of 10, and severe hip involvement developed after 20 years old. All patients presented developmental coxa vara and seven patients (~22%) had pleural effusion, pericarditis, and/or ascites. We identified nine novel genomic alterations, including the first case of homozygous complete deletion of exon 1 in the PRG4 gene.

Conclusion: With this study, we contribute to the catalog of CACP causing variants. We confirm that the skeletal component of this disease worsens with age, and presents the potential mechanisms for interfamily variability, by discussing the influence of a modifier gene and escape from nonsense-mediated mRNA decay. We believe that this report will increase awareness of this familial arthropathic condition and the characteristic clinical and radiological findings will facilitate the differentiation from the common childhood rheumatic diseases such as juvenile idiopathic arthritis.

Keywords: PRG4; NGS; camptodactyly-arthropathy-coxa vara-pericarditis; genotype-phenotype correlation; lubricin; noninflammatory arthropathy; nonsense-mediated mRNA decay.

Figure 1

(a) Variable degree of camptodactyly and large joints involvement of the patients at different ages. (b) Cystic radiolucent lesion on wrist and variable degrees of lumbar lordosis. (c) Ascites in patient 16 from family 3. (d) Hand X‐ray of patient 5 revealed cystic radiolucent lesion on distal metaphysis of ulna. (e) Pelvis imaging of the patients at different ages showed narrowing acetabular space and irregularity of femoral capitis with aging, osteoporosis, short femoral neck, and mild–moderate coxa vara

Figure 2

Structure of PRG4 protein and mutations identified in our cohort. (a) Functional domains of PRG4 protein and the mutations identified in the present study. SO domains, somatomedin B‐like domains; HX repeats, hemopexin‐like repeats; Chon_Sulph_att : chondroitin sulfate attachment site. (b) The graph represents the relative normalized copy number variation of PRG4 DNA for primer pairs Ex1 (located on exon 1), E1I1 (located between exon 1 and 2) and Ex8–9 (located between exon 8 and 9), respectively. Patient 20 from family 6 presents a homozygous deletion of exon 1 detected by primer pairs Ex1 and Ex1I1, while primer Ex8–9 shows no variation in copy number. The parents of patient 20 are heterozygous for the deletion identified in patient 20. Families 5 and 14 do not show any copy number variation. (c) Sanger sequencing results for the 17‐bp deletion in exon 10 (p.1306fs) segregating in family 2. The affected three siblings all present the 17‐bp deletion (homozygous profile), both parents carry one copy of the deletion (heterozygous profile) and the control DNA show two copies of the wild‐type sequence

Figure 3

Correlations of PRG4 mutations and clinical features. (a) Binominal test used to determine gender bias compared to a theoretical ratio of 1:1 gave a significant p‐value. Total number of published cases of CACP patients were used. (b) Fisher's exact test was used to determine gender bias linked to CACP disease. (c) Cumulative distribution of mutations across the coding region of the PRG4 gene. The expected and observed percentages were calculated according to the sequence length (see Appendix S2). (d) Correlation between Phenoscore and age of the patient. Each circle represents a patient. For each patient, the family number is indicated in the circle. The star indicates the presence of extraskeletal features like pericarditis, ascites, pleurites, MVP (mitral valve prolapses), or MR (mitral regurgitation). The graph shows a significant correlation between the aging process and the severity of CACP (Spearman r = 0.8614, p = 3.23 × 10⁻⁰⁸)

Figure 4

Homo sapiens proteoglycan 4 (PRG4) transcript variant A, mRNA (NM_005807) annotated cDNA sequence. Alternative exons are depicted on the cDNA sequence with alternative black and blue color (e.g., exon 1/2). Start codons are identified in red at the cDNA (160–162) and protein level (M54). All the reported mutations are displayed at cDNA and/or protein level. Previously reported deletions and insertions (3690del5) and point mutations (c.3648C≥A) and mutations discovered in the present study (c.3918del17)