Osteopetrosis (OPT) is a rare genetic disorder of bones, first reported in early 1904 [1]. Abnormally dense bone mass is a characteristic feature of osteopetrosis caused by a defective Osteoclastic bone resorption [2]. The condition lead to bone fragility and an increased vulnerability to fractures [3]. Four sub-types of OPT are described including infantile malignant OPT (OPTB1), adult/ benign OPT, intermediate OPT and carbon anhydrase type II deficiency syndrome [4]. However, Syndromic forms of osteopetrosis such as carbonic anhydrase 2 deficiency are not considered as classic autosomal recessive osteopetrosis (ARO) due to the mild presentation of the bone defect in affected patients [2]. Modes of inheritance further split osteopetrosis into autosomal dominant and autosomal recessive categories [5]. However, infantile malignant OPT, also known as autosomal recessive osteopetrosis type 1, remains the most common and best identified subtypes with an average disease frequency of 1/ 250,000 (1/200,000–1/300,000). The highest frequencies of infantile malignant OPT has been reported in populations of Costa-Rica (3.4/100,000) in Central American and Chuvash (1/3879 newborns) in Russia owing to a founder effect, high parental consanguinity or geographic isolation [2, 6].

Infantile malignant OPT is associated with intermediate and severe clinical phenotypes associated with reduced or complete loss of bone resorption capability of osteoclasts [7]. Impaired osteoclastic bone resorption leads to retention of old bones, hence an increase in bone density and obstruction of internal cavities containing vital tissues/organs such as bone marrow, nervous system etc. Distinctive symptoms include short stature, bone deformities and pathological fractures that may accompanied by severe hematological and neural failures [8]. In comparison with autosomal dominant benign OPT that occurs in late ages and with less severe clinical consequences, autosomal recessive infantile malignant OPT involves an early onset and lethal outcome [9]. The mortality rate of autosomal recessive OPT is significantly high during the first 2 years of life. Bone marrow failure, visual impairment and profuse infections before the age of 3 months are the major contributing factors to death [7].

Bone resorption depends on osteoclast secretion of hydrogen ions (H+), chloride ions (Cl−) and matrix-degrading protease, cathepsin K, into the resorption lacunae. The process mainly involves the breakdown of crystalline hydroxyapatite and collagen-rich organic bone matrix. The breakdown of both components requires acidic conditions, because hydroxyapatite is solubilized in acidic solution and the collagen-degrading protease, cathepsin K, is optimally active at acidic pH. Therefore, the bone surface acidic microenvironment beneath the ruffled border is a fundamental and crucial step for bone resorption. A continuous expulsion of H+ and Cl − through combined actions of V-ATPase proton pump and chloride channel in the ruffled border of osteoclasts serves to achieve this acidic microenvironment in resorptive lacunae. The degraded bone matrix is endocytosed from the resorption lacunae and finally released into the extracellular environment [10, 11]. The proton pump function of multi-subunit V-ATPase is tightly regulated by different genes whose pathogenic variants are frequently reported in impaired proton pump regulation. To date, 11 osteopetrosis genes have been identified including 7 genes associated with ARO; TCIRG1, CLCN7, OSTM1, SNX10, TNFSFR11A, TNFSF11 and PLEKHM1 [2, 5]. Four of the known ARO genes, TCIRG1, CLCN7, OSTM1 and SNX10 are significantly expressed in ruffled border of mature osteoclasts and are largely known for their involvement in autosomal recessive osteopetrosis. Specific proteins encoded by these genes are part of the osteoclast-specific enzymatic system that significantly contributes in bone dissolution process. The TCIRG1 encodes a3 subunit of proton pump, is the most frequently mutated gene responsible for >50% of ARO cases. The CLCN7 gene encodes chloride channel (CLC-7) to provide electro-neutrality during acidification process, account for approximately 13–16% of ARO cases. Mutations in a recently cloned OSTM1 gene which encode transmembrane protein 1, lead to severe phenotypes and account for 2–6% ARO cases. Whereas, SNX10 gene encodes sorting nexin 10, is accounting for almost 4% ARO cases. Overall, approximately >70% malignant infantile osteopetrosis cases emerge from mutations in these four genes. The remaining 3 gene are less frequently described in ARO patients with TNFSFR11A in <1–4% of cases, TNFSF11 in <1–3% of cases and only 2 cases of PLEKHM1 have been reported in the literature so far [2].

Though the incidence of disorder is worldwide without any gender specificity, it is frequently reported in ethnic groups with increased consanguinity [7]. Pakistan is also a country where consanguinity is being practiced over hundreds of years by highly conserved ethnic groups [12]. As a result, variety of inherited bone disorders including osteopetrosis can easily be identified. Unfortunately, no significant work has been carried out so far to identify molecular basis of osteopetrosis in Pakistani population. In the present study, a multi-generational consanguineous Pakistani family suffering from infantile osteopetrosis was clinically and genetically characterized. Genetic screening revealed a novel homozygous deletion mutation in V-ATPase a3 subunit associated with autosomal recessive osteopetrosis in this Pakistani family.

Bone resorption process requires a constant secretion of acid (H⁺) into the resorption lacunae mediated by V-ATPase proton pump [10]. The V-ATPase proton pump is multi-subunits membrane complex consisting of two main domains; a cytosolic hydrolytic domain (V1) that mediates ATP hydrolysis and a transmembrane proton translocation domain (V0) that facilitates extracellular acidification of organelles [10, 26]. Mutations, deletions or genes knockdown of different subunits of the V-ATPase complex have been described to be implicated in impaired osteoclastic bone resorption, leading to severe osteopetrosis [27]. Approximately 50% of the patients with recessive infantile malignant osteopetrosis have mutations to a3 subunit of V-ATPase protein complex, encoded by TCIRG1 gene [28]. To date, over hundred pathogenic variations of TCIRG1 gene including twenty small deletion mutations have been reported to be associated with impaired a3 subunit function, hence infantile malignant osteopetrosis (HGMD; www.hgmd.org).

In the present study we identified a homozygous deletion mutation (c. 624delC) in exon 6 of the TCIRG1 gene and to the best of our knowledge it has not been reported before. The deletion led to a frame shift that created a stop codon at position downstream next to the point of deletion. Hence, resulted premature truncated protein consists of 208 aa instead of normal 830 aa (p.P208PfsX1). In other words there is a loss of 622 aa at C-terminal domain of a3 subunit. The occurrence of deletion in homozygous state in all affected subjects, heterozygous state in both carrier parents and absence in a screen of 100 ethnically matched healthy controls confirms its association with infantile malignant osteopetrosis in Pakistani family.

In silico characterization of the deletion mutation revealed significant consequences on the secondary structure features of mutant type protein (see Additional file 1: Figure S3). Premature termination not only led to an abnormal shorter protein of 208 aa but also resulted in atypically reduced numbers of structural motives like alpha helices, strands and coils. Structural changes might have resulted in amino acid re-arrangement and conformational changes so that the receptor-ligand interaction trend of mutant a3 subunit is predicted totally different from that of wild-type a3 subunits (Fig. 4c-d). The 2-dimensional protein-ligand interaction complexes predicted hydrogen bonding between Tyr583 of wild-type receptor and Gly40 of ligand molecule ATPV1B1. Whereas, in case of mutant protein, the residues Cys25, Arg28 and Asp54 are involved in hydrogen bonding with Asp26 and Ala69 residues of protein-ligand (Fig. 5a-b). Significantly dislocated interaction sites and amino acid residues have also disturbed the hydrophobic interactions that are equally important in receptor-ligand interactions. Location of interaction sites and number of interactive pockets, involvement of different amino acid residues as well as hydrogen bonding pattern are in Table 1.

Topological studies propose a two-domain structure for “a3 subunit” with an N-terminal cytosolic domain necessary to target V-ATPase to the plasma membrane of osteoclasts, and a membrane integral C-terminal domain that make structural frame work of V0 domain and affects the coupling of proton transport and ATP hydrolysis [10, 29]. In the present study, loss of 622 aa at C-terminal of a3 subunit strongly suggests impaired proton coupling, transportation and ATP hydrolysis even if it is supposed that the N-terminal with 208 aa targets V-ATPase to the plasma membrane of osteoclasts. It can also be hypothesized that partially translated a3 subunit may have been degraded by “nonsense mediated mRNA decay” [26], resulting in a complete loss of a3 subunit. In either situation, there is a strong possibility of V-ATPase complex assembly disruption, leading to clinical phenotypes of osteopetrosis.

In summary, the mutant a3 isoform might have caused V-ATPase complex assembly to be disrupted, resulted in osteoclast failure to maintain extracellular acidification and hence defective bone resorption attributed to infantile malignant osteopetrosis. Our results, supported by SNP genotyping, Sanger sequencing, allele-specific PCR and bioinformatics have not only expanded the spectrum of TCIRG1 pathogenic variations but also the body of evidences supporting the role of a3 isoform in osteopetrosis. Moreover, our findings may be supportive for prenatal genetic screening and early diagnosis of osteopetrosis.