Osteogenesis Imperfecta (OI), commonly known as “brittle bone disease,” is a rare inherited connective tissue disorder affecting approximately 1 in 15,000 to 20,000 live births globally, primarily characterized by bone fragility, recurrent fractures, and skeletal deformities due to defects in type I collagen synthesis or processing (Marini et al., 2017). OI places significant burdens on patients, especially in the pediatric population where hospitalization rates are 8.4 times higher than the general population, and the majority of hospitalizations are due to fragility fractures (Storoni et al., 2022; Abraham et al., 2025). The disorder manifests a broad phenotypic spectrum, ranging from mild manifestations to perinatal lethality, and is classified according to the conventional Sillence system (types I–IV) (Forlino and Marini, 2016). Over 90% of OI cases stem from pathogenic variants in genes involved in collagen biosynthesis, predominantly COL1A1 and COL1A2, though at least 18 other genes contributing to collagen modification, folding, or osteoblast function have been identified, with novel mutations continuously reported (Jovanovic and Marini, 2024; Yamaguti et al., 2023). This genetic complexity, compounded by population-specific mutation spectra and numerous undiscovered variants, hinders universal therapeutic strategies, necessitating ongoing exploration of new mutations.

COL1A1 and COL1A2 code for the α1 and α2 chains of type I procollagen, a key extracellular matrix protein in bone tissue. Pathogenic variants in these genes lead to OI through two principal mechanisms. The first involves quantitative defects, or haploinsufficiency, commonly caused by nonsense, frameshift, or splicing mutations that induce mRNA degradation and reduce normal collagen synthesis by about half; such mutations typically correlate with milder type I OI. The second mechanism entails qualitative defects characterized by dominant-negative effects. These often stem from missense mutations that replace glycine within the characteristic Gly-X-Y repeats, for example, with serine or aspartate, thereby disrupting triple-helix assembly and resulting in more severe OI forms, including types II through IV (Marini et al., 2017; Forlino and Marini, 2016; Jovanovic and Marini, 2024; Yamaguti et al., 2023). Recent studies highlight significant population-specific heterogeneity in OI genetics, as COL1A1/COL1A2 mutation rates vary markedly between groups, reaching approximately 90% in Western (Zhytnik et al., 2017) and Japanese cohorts (Ohata et al., 2019) while being considerably lower in Indian (47.6%) (Selina et al., 2025) and sporadic Chinese cases (40%) (Ju et al., 2020). Ongoing research continues to identify novel and recurrent pathogenic variants, such as deep intronic and splice-site mutations that disrupt collagen processing, as well as helical glycine substitutions strongly associated with severe phenotypes (Jovanovic and Marini, 2024; Schouw et al., 2025).These variants have had clinical implications by facilitating definitive diagnosis and prognosis prediction with genetic testing in complex clinical scenarios. However, the ongoing discovery of novel mutations hinders personalized treatment and genetic counseling.

Although there is now knowledge of the pathogenic variants in the COL1A1 gene, the challenge for clinicians still exists in classifying the pathogenicity of novel or rare variants (Jovanovic and Marini, 2024). The clinical geneticist cannot depend on the bioinformatics prediction tools to predict the effect of the variant since they are often inaccurate in predicting the effect on splicing and the protein. In addition, the genomic databases worldwide are biased towards the European population, resulting in a knowledge gap and bias of pathogenicity of variants that are found outside Europe. Hence, according to the American College of Medical Genetics and Genomics (ACMG) guidelines, these novel variants must be functionally validated, especially in a particular population (Richards et al., 2015).

In this study, we report the identification of a novel heterozygous COL1A1 intron 2 splice donor variant, c.298 + 1G>A, in a Chinese OI cohort through whole-exome sequencing. This variant, which was absent from major population databases, was functionally characterized through reverse transcription-polymerase chain reaction (RT-PCR), revealing an aberrant transcript, and confirmed with a minigene splicing assay. This finding enriches the genetic and phenotypic spectrum of OI in the Chinese population, providing a crucial basis for precise molecular diagnosis and genetic counseling for individuals in underrepresented groups.

In this study, we identified and functionally characterized a novel heterozygous splice-site variant (c.298 + 1G>A) in the COL1A1 gene in a Chinese family with OI. By combining detailed clinical phenotyping, genetic sequencing, bioinformatic prediction, and minigene splicing assays, we provide strong evidence supporting its pathogenicity. This finding not only broadens the known mutational spectrum of COL1A1 but also provides crucial insights into genotype–phenotype correlations in OI, which may facilitate more accurate diagnosis and targeted genetic counseling for affected individuals and their families.

In this Chinese family, classic clinical features of OI consistent with type I (mild) under the Sillence classification were observed. Key manifestations included recurrent fractures, blue sclerae, and the absence of severe skeletal deformities (Marini et al., 2017). These symptoms typically first appeared in infancy or childhood. While fracture frequency often decreases after adolescence, it may increase again during adulthood, particularly in postmenopausal women (Ralston and Gaston, 2020; Arshad and Bishop, 2021). The clinical presentation in this family aligns closely with the characteristics of mild OI as documented in the literature (Marini et al., 2017; Han et al., 2020; Huang et al., 2025), thereby excluding more severe forms such as type II (perinatal lethal) or type III (progressively deforming) OI.

Genetic analysis revealed that affected family members carried a heterozygous splice-site mutation in the COL1A1 gene, which encodes the alpha1 chain of type I collagen—the primary structural protein in connective tissues such as bone, skin, tendons, and sclera (Dalgleish, 1997). Pathogenic mutations in COL1A1 account for over 90% of all OI cases (Yamaguti et al., 2023; Zhytnik et al., 2017; Selina et al., 2025). Mutations in this gene are broadly categorized into two types: those leading to reduced collagen production (quantitative defects or haploinsufficiency) and those causing structurally abnormal collagen (qualitative defects or dominant-negative effects) (Jovanovic and Marini, 2024; Ju et al., 2020). Clinically, mild OI is commonly associated with quantitative defects, while severe or lethal OI often results from qualitative defects. The mild phenotype in this family strongly suggests that the novel c.298 + 1G>A splice-site mutation most likely leads to a quantitative defect. Specifically, we proposed that aberrant splicing triggered by this mutation results in transcripts that are degraded by nonsense-mediated mRNA decay (NMD), reducing the production of normal type I collagen.

To molecularly validate the pathogenic mechanism of the c.298 + 1G>A variant, a minigene splicing assay, which is considered the gold standard for functionally assessing splice-site variants, was employed (Anna and Monika, 2018). The experimental results strongly support the initial hypothesis. While the wild-type construct exhibited normal splicing patterns, the c.298 + 1G>A mutant led to the production of multiple aberrant transcripts. These abnormal splicing events, characterized by partial and complete exon skipping, disrupted the open reading frame and introduced a premature termination codon. Crucially, the most abundant transcripts identified in both minigene systems (band d in Figure 4 and band b in Figure 5) consistently introduced premature termination codons. Transcripts harboring premature termination codons are typically targeted for degradation via NMD, effectively preventing the synthesis of truncated protein from the mutant allele (Slayton et al., 2000; Willing et al., 1996). This “null allele” consequence provides a definitive molecular explanation for the mild OI phenotype observed in this family. By eliminating potentially deleterious α1 (I) chains, the NMD machinery prevents the accumulation of misfolded pro-collagen that would otherwise compromise the triple-helix assembly through a dominant-negative effect. Consequently, while the total output of type I collagen is reduced (haploinsufficiency), the molecules successfully secreted into the extracellular matrix remain structurally normal (Popp and Maquat, 2013). This quantitative deficiency leads to the characteristic bone fragility of Type I OI without the severe skeletal deformities associated with qualitative defects. These functional findings allow for the reclassification of the c.298 + 1G>A variant from a variant of uncertain significance (VUS) to a pathogenic mutation, establishing a robust genotype-phenotype correlation.

Although located in an intron, this mutation ultimately affects exon composition—a change particularly consequential in genes like COL1A1, where many exons encode Gly-X-Y repeats that form the triple-helical domain of type I collagen (Gensure et al., 2005; Pollitt et al., 2006). Notably, in-frame exon skipping—which does not cause a frameshift—can still disrupt the (Gly-X-Y)n repeats and impair triple-helix formation, often resulting in dominant-negative effects and severe phenotypes (Pollitt et al., 2006). However, both the minigene results and the mild clinical presentation in this family suggest that the c.298 + 1G>A mutation leads predominantly to frameshift-inducing splicing errors and subsequent NMD, rather than in-frame skipping. This supports the conclusion that the variant causes a quantitative rather than a qualitative collagen defect.

The findings of this study are consistent with previously reported COL1A1 splice-site variants and add to the known mutational spectrum of the gene. For instance, several studies have documented that splice-site mutations in COL1A1—such as c.3814 + 1G>T reported by Han et al. (2020) and c.3531 + 1G>T identified by Huang et al. (2025)—frequently lead to aberrant splicing and are associated with mild OI type I, largely through mechanisms involving mRNA mis-splicing and nonsense-mediated decay (NMD), resulting in haploinsufficiency. Similarly, the c.298 + 1G>A variant described here induces exon skipping, corroborating the well-established pathogenic mechanism of quantitative collagen defects in mild OI. The specific location of this variant at the donor site of intron 2, which corresponds to an early region of the COL1A1 gene, is particularly significant. Splice-site mutations in early exons, such as exons 1–5, that lead to NMD are hallmark causes of Type I OI. Our findings align with other early-region mutations where frameshift-inducing errors consistently result in mild phenotypes, in contrast to late-gene mutations that might bypass NMD and produce structurally defective collagen (Gökmen et al., 2024). This study identifies c.298 + 1G>A as a novel pathogenic variant in the Chinese population, further enriching the genotypic spectrum of OI. This mechanism aligns with the outcomes of numerous functional studies, including minigene assays performed in other cohorts, strengthening the claim that splice defects leading to in-frame disruptions and NMD are a common cause of type I OI. However, it is noteworthy that not all splice variants yield mild phenotypes; exceptions such as the deep intronic variant c.2451 + 77C>T reported by Schouw et al. (2025) can cause severe perinatal OI type II through in-frame insertions that disrupt the triple helix. The consistency between our functional results and those of previous studies underscores the utility of minigene assays in validating splicing defects and supports the reclassification of VUSs into pathogenic variants. Overall, this study reinforces the notion that splice-altering variants in COL1A1 commonly operate through loss-of-function mechanisms and expand the genotypic and phenotypic correlations essential for diagnostic and counseling purposes.

Although this study provides compelling evidence, several limitations should be acknowledged. First, the analysis is based on a single family, which may limit the generalizability of the findings. Identification of the same variant in broader populations would strengthen the genetic evidence. Second, although the minigene assay in HEK293T cells accurately detected splicing defects, this in vitro model utilizes a non-bone-related cell line that may not fully replicate the specific splicing environment of osteoblasts. Furthermore, it could not directly evaluate the actual production and secretion of type I collagen in patient-derived cells. Future studies should analyze RNA and protein extracted from patient fibroblasts or osteoblasts to directly assess transcript abundance and collagen defects under physiological conditions, thereby providing more comprehensive mechanistic insights and verifying the in vivo relevance of our findings. Finally, while the identification of predominantly frameshifted transcripts supports a haploinsufficiency model, other potential pathogenic mechanisms, such as the potential for minimal protein products to escape nonsense-mediated decay, cannot be entirely excluded. Therefore, the generalizability of these conclusions should be interpreted with prudence, as the pathogenic process may be influenced by specific genetic or environmental modifiers within this family.

This work highlights potential therapeutic avenues rooted in gene-editing strategies. Emerging techniques, such as CRISPR/Cas9, offer promising opportunities for precise correction of disease-causing mutations (Jung et al., 2021; Cao et al., 2023). Future studies should explore the use of CRISPR/Cas9 to directly repair the c.298 + 1G>A variant in patient-derived cells or iPSCs, thereby restoring normal COL1A1 splicing and collagen production. This approach may provide a curative strategy for individuals with this form of OI, moving beyond symptomatic management toward molecular correction (Chaugule et al., 2025; Deguchi et al., 2021).

In summary, this study identified and functionally validated a novel splice-site variant (c.298 + 1G>A) in the COL1A1 gene in a Chinese family with type I OI. Minigene assays demonstrated that this mutation disrupts canonical splicing, most likely leading to a quantitative defect in type I collagen production. These findings highlight the importance of integrating clinical phenotyping, genetic analysis, and functional studies for accurate diagnosis and pathogenic interpretation of variants in rare genetic disorders.