Comparing Clinical and Genetic Characteristics of De Novo and Inherited COL1A1/COL1A2 Variants in a Large Chinese Cohort of Osteogenesis Imperfecta

Abstract

Purpose: Nearly 85%-90% of osteogenesis imperfecta (OI) cases are caused by autosome dominant mutations of COL1A1 and COL1A2 genes, of which de novo mutations cover a large proportion, whereas their characteristics remain to be elucidated. This study aims to compare the differences in clinical and genetic characteristics of de novo and inherited COL1A1/COL1A2 mutations of OI, assess the average paternal and maternal age at conception in de novo mutations, and research the rate of nonpenetrance in inherited mutations.

Materials and methods: A retrospective comparison between de novo and inherited mutations was performed among 135 OI probands with COL1A1/COL1A2 mutations. Mutational analyses of all probands and their family members were completed by Sanger sequencing. A new clinical scoring system was developed to assess the clinical severity of OI quantitatively.

Results: A total of 51 probands (37.78%) with de novo mutations and 84 probands (62.22%) with inherited mutations were grouped by the results of the parental gene verification. The proportion of clinical type III (P<0.001) and clinical scores (P<0.001) were significantly higher in de novo mutations. Missense mutations covered a slightly higher proportion of de novo COL1A1 mutations (46.34%) compared with inherited COL1A1 mutations (33.33%), however, lacking a significant difference (P=0.1923). The mean BMD Z/T-score at the lumbar spine in de novo mutations was -2.3 ± 1.5, lower than inherited mutations (-1.7 ± 1.8), but lacking statistical significance (P=0.0742). There was no significant difference between the two groups in OI-related phenotypes (like fracture frequency, blue sclera, and hearing loss) and biochemical indexes. In de novo mutations, the average paternal and maternal age at conception was 29.2 (P<0.05) and 26.8 (P<0.0001), respectively, which were significantly younger than the average gestational age of the population. Additionally, 98.04% of pedigrees (50/51) with de novo mutations were spontaneous conception. The rate of nonpenetrance of parents with pathogenic variants in the inherited mutation group was 25.64% (20/78).

Conclusions: Our data revealed that the proportion of clinical type III and clinical scores were significantly higher in de novo mutations than in inherited mutations, demonstrating that de novo mutations are more damaging because they have not undergone purifying selection.

Keywords: COL1A1; COL1A2; clinical scoring system; de novo; inherited; osteogenesis imperfecta.

Figure 1

Flowchart of study cohort selection.

Figure 2

X-ray radiography of the lower extremities of six de novo and two inherited mutation probands with clinical type III, and diagrams showing mutation loci identified in this study. (A, B) X-ray radiography of the femurs of the probands (p68 and p56) with inherited mutations. (C–H) X-ray radiography of the lower extremities of the probands (P9, P3, P49, P43, P13, and P26, respectively) with de novo mutations. (I) Mutation loci in COL1A1 gene identified in two groups. (J) Mutation loci in COL1A2 gene identified in two groups.

Figure 3

Genotype-phenotype correlations in de novo and inherited mutations. (A, B) Frequency of clinical type for each mutation type in De novo and Inherited. (C–F) The bubble plots of the clinical score of available probands with de novo or inherited COL1A1/COL1A2 mutations. X-axis represented the mutation site on the COL1A1/COL1A2 gene; Y-axis was the clinical score; the colors of the bubble represented different mutation effects (missense, nonsense, frameshift, and splice, respectively); the size of the bubble indicated the mutation frequency of each mutation site in the de novo or inherited mutation group. The bubble was higher and bigger; the phenotype was more severe, and the mutation site was more frequent.