Skip to main content
  • Loading metrics

ATP6V1H Deficiency Impairs Bone Development through Activation of MMP9 and MMP13

  • Yihan Zhang ,

    Contributed equally to this work with: Yihan Zhang, Haigen Huang

    Affiliations Laboratory of Chemical Genomics, School of Chemical Biology and Biotechnology, Peking University Shenzhen Graduate School, Shenzhen China, Department of Molecular, Cell, and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America

  • Haigen Huang ,

    Contributed equally to this work with: Yihan Zhang, Haigen Huang

    Affiliation Laboratory of Chemical Genomics, School of Chemical Biology and Biotechnology, Peking University Shenzhen Graduate School, Shenzhen China

  • Gexin Zhao,

    Affiliations Department of Molecular, Cell, and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America, Department of Orthopaedic Surgery and Orthopaedic Institute for Children, David Geffen School of Medicine, University of California, Los Angeles, California, United States of America

  • Tadafumi Yokoyama,

    Affiliation Medical Genetics Branch, National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • Hugo Vega,

    Affiliation NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • Yan Huang,

    Affiliation NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • Raman Sood,

    Affiliation Zebrafish core, Translational and Functional Genomics Branch, National Human Genome Research Institute/NIH, Bethesda, Maryland, United States of America

  • Kevin Bishop,

    Affiliation Zebrafish core, Translational and Functional Genomics Branch, National Human Genome Research Institute/NIH, Bethesda, Maryland, United States of America

  • Valerie Maduro,

    Affiliation NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • John Accardi,

    Affiliation NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • Camilo Toro,

    Affiliation NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • Cornelius F. Boerkoel,

    Affiliation NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America

  • Karen Lyons,

    Affiliations Department of Molecular, Cell, and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America, Department of Orthopaedic Surgery and Orthopaedic Institute for Children, David Geffen School of Medicine, University of California, Los Angeles, California, United States of America

  • William A. Gahl,

    Affiliations Medical Genetics Branch, National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America, NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America, Office of the Clinical Director, National Human Genome Research Institute/NIH, Bethesda, Maryland, United States of America

  • Xiaohong Duan , (XD); (MCVM); (SL)

    Affiliation State Key Laboratory of Military Stomatology, National Clinical Research Center for Oral Diseases, Shaanxi Key Laboratory of Oral Diseases, Department of Oral Biology, Clinic of Oral Rare and Genetic Diseases, School of Stomatology, The Fourth Military Medical University, Xi’an, China

  • May Christine V. Malicdan , (XD); (MCVM); (SL)

    Affiliations Medical Genetics Branch, National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America, NIH Undiagnosed Diseases Program, Common Fund, Office of the Director, NIH and National Human Genome Research Institute, NIH, Bethesda, Maryland, United States of America, Office of the Clinical Director, National Human Genome Research Institute/NIH, Bethesda, Maryland, United States of America

  •  [ ... ],
  • Shuo Lin (XD); (MCVM); (SL)

    Affiliation Department of Molecular, Cell, and Developmental Biology, University of California Los Angeles, Los Angeles, California, United States of America

  • [ view all ]
  • [ view less ]


27 Feb 2017: Zhang Y, Huang H, Zhao G, Yokoyama T, Vega H, et al. (2017) Correction: ATP6V1H Deficiency Impairs Bone Development through Activation of MMP9 and MMP13. PLOS Genetics 13(2): e1006624. View correction


ATP6V1H is a component of a large protein complex with vacuolar ATPase (V-ATPase) activity. We identified two generations of individuals in which short stature and osteoporosis co-segregated with a mutation in ATP6V1H. Since V-ATPases are highly conserved between human and zebrafish, we generated loss-of-function mutants in atp6v1h in zebrafish through CRISPR/Cas9-mediated gene knockout. Homozygous mutant atp6v1h zebrafish exhibited a severe reduction in the number of mature calcified bone cells and a dramatic increase in the expression of mmp9 and mmp13. Heterozygous adults showed curved vertebra that lack calcified centrum structure and reduced bone mass and density. Treatment of mutant embryos with small molecule inhibitors of MMP9 and MMP13 significantly restored bone mass in the atp6v1h mutants. These studies have uncovered a new, ATP6V1H-mediated pathway that regulates bone formation, and defines a new mechanism of disease that leads to bone loss. We propose that MMP9/MMP13 could be therapeutic targets for patients with this rare genetic disease.

Author Summary

Osteoporosis, a major health problem worldwide, is characterized by low bone mineral density (BMD) and a propensity to fracture. Genetic factors are clearly the major determinants of BMD, but their identification and contribution to osteoporosis risk have been difficult to assess in humans. Genome-wide association studies (GWAS) have identified numerous sequence variants that influence BMD. The loci identified to date, however, account for only a small fraction of the total variation in BMD. Through analysis of a pedigree with an undiagnosed disease in which affected members have markedly low bone mass, we identify a novel and critical bone formation pathway, mediated through the gene ATP6V1H. Zebrafish lacking apt6vh1 demonstrate loss of bone mass, and exhibit increased mmp9 and mmp13 levels; inhibition of mmp9 and mmp13 led to the rescue of bone density defects. Here we show that happloinsufficiency of ATP6V1H is associated with osteoporosis in both humans and zebrafish. This study exemplifies the value of studying rare diseases to understand prevalent ones.


ATP6V1H is a member of the family of vacuolar ATPases (V-ATPases), and a component of protein complexes responsible for acidification of intracellular compartments in all eukaryotic cells [1]. Specifically, V-ATPases regulate protein degradation, mediate pH homeostasis, and facilitate cellular function and development through acidification. In human, mutations in ATP6B1, which encodes the β-subunit of the apical proton pump mediating distal nephron acid secretion, cause distal renal tubular acidosis, a condition characterized by impaired renal acid secretion, resulting in metabolic acidosis [2]. In osteoclasts, V-ATPases acidify the space that is in contact with the bone surface; this process is essential for bone resorption. Consequently, components of V-ATPases are potential therapeutic targets for the treatment of osteolysis [3]. Moreover, Atp6i-deficient mice exhibit severe osteopetrosis due to loss of osteoclast-mediated extracellular acidification; the osteopetrotic phenotype may be due to failure of osteoclast formation during development [4]. Using zebrafish model and patient phenotypic data, we show that haploinsufficiency of the V1H subunit of V-ATPases can result to osteoporosis.

Through the NIH Undiagnosed Diseases Program (UDP) [57], we identified a 48-year old woman and her 24-year old son who exhibited markedly low bone density. This phenotype segregated with a monoallelic mutation in ATP6V1H. Since osteoporosis is the opposite of the osteopetrosis seen in Atp6i-deficient mice and expected for reduced V-ATPase activity, we studied the phenotype of zebrafish harboring a null mutation in atp6v1h. Indeed, this animal model displayed reduced bone mass, and we demonstrated that the bone defect was mediated by increased activity of mmp9 and mmp13. Small molecule inhibitors of both proteins notably restored bone volume, suggesting that MMP9 and MMP13 could be therapeutic targets for treatment.



Patient 1 (P1; II.1 in Fig 1A) is a 48-year old woman and Patient 2 (P2; III.1 in Fig 1A) is her 24-year old son. Both exhibited short stature, bone pain and decreased bone density (Fig 1C, 1–5 and S1A Fig). The father of P1 (I.2 in Fig 1C) reportedly had the same features, with dilated frontal sinuses and decreased bone density (S1B and S1C Fig). Exome sequencing revealed NM_015941.2:c.1158_1159delinsTT (p.Lys386_Asn387delinsAsnTyr, named K386N/N387Y hereafter) in ATP6V1H that segregated with the osteoporotic phenotype (Fig 1B). In silico analysis for protein stability predicted that both amino acid changes due to the mutation (K386N and N387Y) can alter protein stability (I-Mutant2.0). To confirm this finding, we generated a stable cell line expressing K386N/N387Y and measured protein half-life. After protein synthesis inhibition by cycloheximide treatment, the mutant protein decayed rapidly compared to the wild-type protein, suggesting that the mutation produced an unstable protein (S2 Fig).

Fig 1. Clinical features of patients with ATP6V1H mutations.

The family comes from a three-generation pedigree; affected members are shaded in black (A); Chromatograms showing that the ATP6V1H mutation is present in affected family members, (II.1 and III.1 (B); DNA from I.1 is not available; (C) Radiographs (C) from Patient 1 (P1) (1, 2, and 3) and Patient 2 (P2) (4, and 5) reveal that P1 has scoliosis (C1) and increased radiolucency in tibia (C2). Bone scan shows multiple sites of increased bone mineral density uptake–tip of right scapula and multiple areas of thoracic and lumbar spine, sacrum, both sacroiliac joints, left humeral head, shaft of left femur, right tibia and right ribs (C3). Some increase in radiolucency especially in the lumbar spine (C4), is seen in P2.

Bone but not cartilage is defective in atp6v1h-deficient zebrafish

Based on the finding that K386N/N387Y mutant protein was unstable, we reasoned that loss of function mutation in an animal model could address the function of ATP6V1H. We obtained a null mutation of atp6v1h in zebrafish through CRISPR/Cas9 mediated gene knockout (S3 Fig and supplementary information, designated as atp6v1hla439). Homozygous mutants died at 10–11 days post fertilization (dpf). Up to dpf 6, the mutant embryos did not appear to have major abnormalities in morphology, but they failed to inflate their swim bladders (Fig 2). Although a retrovirus-induced insertional mutation of atp6v1h has been previously reported for zebrafish, no bone related phenotypes were analyzed [8]. In this study we showed that atp6v1h expression was largely confined to the head region, where cartilage and bone cells first develop, and this expression was completely lost in our mutant (S3 Fig), suggesting true loss of function. Analysis by calcein staining indicated that atp6v1h -/- embryos had greatly reduced or absent calcification of bone cells (Fig 2). Co-staining of bone and cartilage cells showed that cartilage cells were largely unaffected, while mineralized bone was nearly absent (Fig 2). We did not observe any obvious skeletal defects in wild type (+/+) or heterozygous (+/-) embryos, suggesting that complete loss of function for atp6v1h is required to produce the observed embryonic phenotype.

Fig 2. Bone but not cartilage cells are defective in atp6v1h-deficient zebrafish.

(A-B) shows an image of a 6 days post fertilization (dpf) wild type and mutant embryo with relatively normal gross morphology. (C-F) Day 6 and 8 wild type and mutant embryos showing reduction in bone detected by calcein staining (ventral view, bright fluorescent stains indicate calcified bone). In (G-H), double staining for cartilage and bone demonstrate reduction of mineralized bone (dark purple, arrows) but not cartilage (blue). A, C, E and G are wild type. B, D, F, and H are mutant.

We next determined if the ATP6V1H K386N/N387Y mutation had a dominant effect on zebrafish development. The affected region of the protein is fully conserved in zebrafish atp6v1h (S3 Fig). We generated cDNA constructs containing the same mutation or the wild type sequence as a control to evaluate the effects of the mutation. We next restored bone mineralization in the mutant zebrafish by injecting with wild type mRNA; injecting K386N/N387Y mutant mRNA did not restore mineralization (S5 Fig), suggesting that the change leads to a loss-of-function mutation. Additionally, injection of wild-type embryos with K3866N/N387Y mRNA had no obvious effect on embryonic development, suggesting that this mutation did not result in gain-of-function S5 Fig).

We then analyzed the phenotype of adult fishes that lacked atp6V1h in one allele (heterozygous, +/-). By eight months after birth, over 50% of the heterozygous adults showed a curved body (Fig 3). X-ray and micro-CT studies revealed that bone mineral density, bone volume and bone surface were all reduced but the later two reductions were more dramatic (S6 Fig). The most remarkable discover is that the vertebrae of heterozygous fish completely lacked calcified structure in the centrum cavity (Fig 3). Additionally, overall skeleton structures appeared smaller and had subtle but notable abnormalities. Furthermore, this adult phenotype was observed in the previously reported retroviral insertion mutant allele of zebrafish atp6v1h gene (S6 Fig) [8].

Fig 3. Micro-computed tomography scan of adult zebrafish.

Adult zebrafish heterozygous for atp6V1h mutation (+/-) show curved body, compared to the wild type (+/+) sibling (A). X-ray of the same +/- fish shows curved spine compared to the +/+ sibling (B). Vertebra of +/- fish has bristlier surface and smaller size (C). Vertebra of +/- fish lacks calcified structure in the centrum cavity (D). Individual vertebra of +/- fish shows bristlier surface and smaller size with a hollow centrum (E) (n = 3 each).

mmp9 and mmp13 are highly induced by atp6v1h-deficiency

To determine the molecular mechanism underlying the bone abnormalities associated with atp6v1h deficiency, we analyzed lineage-specific markers involved in different stages of osteoblast and osteoclast development. First, we determined if atp6v1h was expressed in osteoblast or osteoclast cells of zebrafish by performing high resolution double fluorescent in situ hybridization using atp6v1h probe with the osteoblast marker runx2a or osteoclast marker rank. This analysis showed that atp6v1h had broader co-expression with osteoclast marker rank but less with osteoblast marker runx2a (S4 Fig), which is consistent with the previous report that V-ATPase is enriched in osteoclast cells [1]. Quantitative PCR (qPCR) analysis was further performed to determine mRNA expression of major bone-specific marker genes, including runx2, osx, col1a2, osc, osn and alp for osteoblasts and rank, cstk and mmp9 for osteoclasts between homozygous mutant and sibling embryos. The expression of intermediate and late osteoblast genes in the atp6v1h-deficient zebrafish was marginally reduced (Fig 4A). RNA in situ hybridization (S7 Fig), showing a slight reduction in the expression of the osteoblast genes, confirmed the qPCR data. The expression of the osteoclast genes rank and cstk did not obviously differ from wild type, but mmp9 mRNA levels were vastly elevated in the atp6v1h-deficient zebrafish (Fig 4A). In situ hybridization demonstrated that the pattern and level of expression of the osteoclast marker rank were similar in the mutant and wild type zebrafish, and that the increased mmp9 mRNA expression occurred primarily in the head region of the zebrafish (Fig 4B), where mmp9 was previously identified to be expressed as an osteoclast marker [9,10]. Considering that we mainly detected mmp9 up-regulation in ATP6V1H-defficient mammalian osteoclast but not in osteoblast and chondrocytes (Fig 4C and S8 Fig), we hypothesize that mmp9 is mainly up-regulated in zebrafish osteoclasts. It remains to be determined if other cell types are also induced to express excess amount of mmp9.

Fig 4. Induction of mmp9/mmp13 and rescue of atp6v1h-deficient phenotype.

In (A), qPCR analysis of designated marker genes showing significant induction of mmp9 and mmp13 mRNA expression in atp6v1h mutant. Analysis of osteoclast cells by RNA whole mount in situ hybridization. atp6v1h mutant embryos showing significant increases of mmp9 and mmp13a signals in embryonic head skeleton (e/f/i/j wild type, g/h/k/l mutant), but osteoclast marker rank pattern is not changed (a/b wild type, c/d mutant) (B). In (C), analysis of Mmp9/Mmp13 expression in mouse osteoclast Raw264.7 cells after knockdown of Atp6v1h by siRNA is shown. When Atp6v1h expression was knocked down by siRNA (a), the expression of Mmp9 (b) and Mmp13 (c) were induced. Similarly, the induction of Mmp9 (d) and Mmp13 (e) was also observed in RANKL-induced differentiated mouse osteoclast Raw264.7 cells. In (D), MMP inhibition rescues the bone phenotype in Atp16V1h mutant fish as detected by calcein staining; a, wild type, 0.5% DMSO; b, mutant, 0.5% DMSO; c, mutant treated with 200uM MMP9 inhibitor II in 0.5% DMSO; d, mutant treated with 20uM MMP13 inhibitor in 0.5% DMSO; e, mutant treated with 50uM MMP9/13 inhibitor I in 0.5% DMSO. Images are ventral views of 6 dpf embryos.

The finding that mmp9 was up-regulated prompted us to analyze mmp2 and mmp13, previously reported to be involved in mmp9 related bone homeostasis [9,10]. Expression of mmp13a mRNA was significantly increased in atp6v1h-deficient zebrafish embryos, but that of mmp2 was not (Fig 4A). Zebrafish also have a gene named mmp13b that appears to be a pseudogene since there was no detectable expression by qPCR either in the wild type or atp6v1h-deficient embryos. Further analysis of mmp13a expression by RNA in situ hybridization revealed a pattern and intensity that was very similar to that of mmp9 (Fig 4B).

To determine if Mmp9 and Mmp13 expression was similarly induced in mammalian cells deficient of Atp6v1h, we performed siRNA knockdown experiments in mouse osteoclast, osteoblast and chondrocyte cell types. We found that both Mmp9 and Mmp13 were induced when Atp6v1h was knocked down in mouse osteoclast cells, but reduced in osteoblast and chondrocyte (Fig 4C and S8 Fig), suggesting a conserved action of Atp6v1h deficiency. To study if maturation was affected we performed TRAP staining on these osteoclast cells. This study revealed more TRAP staining in ATP6V1H deficient cells (S9 Fig). ATP6V1H siRNA groups have 0.41%±0.07 TRAP positive cells whereas scrambled controls have 0.13%±0.01 (S2 Table). Consistent to this finding, we found adult zebrafish +/- for atp6v1h appeared to have more TRAP staining in bone sections (S10 Fig). Overall, these data suggest that a higher osteoclast activity is induced by atp6v1h deficiency.

MMP9 and MMP13 inhibitors rescue bone abnormalities in atp6v1h mutants

Because abnormal expression of mmp9 and mmp13 occurs in human diseases such as cancer, arthritis and cardiovascular disorders, small molecule inhibitors specific to MMP9 and MMP13, respectively, or to both MMP9 and MMP13 were available [11,12]. We treated atp6v1h mutant zebrafish embryos with these inhibitors and found that an MMP13 inhibitor restored bone formation nearly toward normal while an MMP9 inhibitor had a reduced rescue effect (Fig 4D). Treatment with an inhibitor that has dual activity on both MMP9 and MMP13 rescued the bone phenotype to an extent similar to the inhibitor of MMP13 (Fig 4D). These findings suggest that increased mmp9 and mmp13 expression contribute significantly to the bone defect, and targeting MMP13 would likely alleviate this problem.


We describe the consequences of atp6v1h mutation in zebrafish, based upon the ascertainment of a family with 3 generations of osteoporosis segregating with a K386N/N387Y mutation in ATP6V1H. We discovered a profound reduction in the amount of mature bone in homozygous atp6v1h zebrafish mutants; cartilage formation was largely unaffected. Since injection of mRNA carrying the human point mutation failed to induce any abnormality in wild type embryos, we propose that K386N/N387Y is a loss-of-function mutation. Given that adult zebrafish heterozygous for atp6v1h deficiency has bone defects and the human mutation was transmitted in a heterozygous manner, this mutation can lead to haploinsufficiency of ATP6V1H protein in humans.

The cellular targets of ATP6V1H function remain to be determined. Osteoblast markers were only marginally affected, suggesting that ATP6V1H does not exert its primary effects on osteoblast formation. The skeletal abnormalities in atp6v1h mutant zebrafish were associated with dramatically increased expression of mmp9 and mmp13, and inhibition of MMP9 and MMP13 significantly restored bone volume. MMP9 and MMP13 are members of the matrix metalloproteinase family, which are important for remodeling of the extracellular matrix in diverse biological processes. The functions of MMP2, MMP9 and MMP13 in skeletal remodeling have been extensively investigated through genetic studies. While Mmp2-null mice have osteoporosis [13], Mmp9-null and Mmp13-null mice exhibit enlarged hypertrophic zones and osteopetrosis [14,15]. The double knockout of Mmp9 and Mmp13 leads to a more dramatic expansion in the length of the hypertrophic zone than in single knockouts, suggesting a synergistic relationship between Mmp9 and Mmp13 [16]. This is consistent with the findings in our atp6v1h zebrafish mutant, in which increased mmp9 and mmp13 expression combined to induce the bone defect.

MMP13 acts as a collagenase to degrade type 1 collagen, the main component of bone extracellular matrix. Collagenases degrade the osteoid layer covering the surfaces of bone to expose calcified bone matrix to osteoclasts [1720]. Significant induction of mmp13 in atp6v1h zebrafish mutant might, therefore, accelerate the degradation of type 1 collagen and provide more mineral contact for osteoclasts, without increasing osteoclast numbers (Fig 4B). Moreover, extracellular matrix has been demonstrated to play a role in osteoblast growth and differentiation [21]. Type I collagen influences osteoblast development and maintenance by modifying the expression of genes supporting cell growth, adhesion and extracellular matrix mineralization [21]. Thus, enhanced degradation of matrix collagen leads not only to increased bone resorption but also to suppression of bone formation [22]. This is consistent with our findings that expression of genes that label intermediate and mature stage osteoblasts is reduced relative to early markers of osteoblast formation (Fig 4).

In zebrafish and mammals, mmp9 is highly expressed in skeletal cells, especially in osteoclasts [9,2325]. Mmp9 can degrade denatured collagens. Therefore, increased mmp9 expression in atp6v1h mutant zebrafish would be predicted to contribute to ECM degradation. Since the expression level of other osteoclast genes, rank and ctsk, remained largely unchanged, we speculate that increased osteoclast numbers do not play a significant role in reducing bone density in atp6v1h mutants. However, increased maturation and activity of osteoclasts might be more relevant, as demonstrated by the increased differentiation revealed by TRAP staining in mouse ATP6V1H deficient pre-osteoclast cells and adult heterozygous atp6v1h zebrafish.

Apart from matrix degradation, Mmp9 also promotes the apoptosis of hypertrophic chondrocytes in the growth plate [14]. Hence, increased MMP9 activity in atp6v1h mutants may also be associated with accelerated loss of hypertrophic chondrocytes or abnormal apoptosis, contributing to bone loss. Interestingly, Mmp13 and Mmp9 induction are also found in patients with arthritis [26]; these individuals also have abnormal matrix, and could share the same pathogenetic mechanism. Inhibition of MMP9 and MMP13 might be considered as a therapeutic approach for treating these diseases.

Vacuolar ATPases have complicated structures and functions, which may explain why mutations in some subunits cause osteopetrosis while mutations in others cause osteoporosis. V-ATPase is composed of 14 different subunits arranged in two structural domains, V1 and V0. ATP hydrolysis occurs in the cytoplasmic V1 domain whereas proton translocation occurs through the membrane V0 domain [1]. Knockout of the a3 subunit of V0 directly blocks its function in proton translocation, thus causing an acidification defect; Atp6i-deficient mice exhibit severe osteopetrosis [2]. Subunit H of V1 activates ATP-driven proton pumping in the intact V-ATPase complex but, when V1 and V0 are disassociated, the H subunit inhibits the Mg-ATPase activity of the cytosolic V1 domain to prevent unnecessary ATP hydrolysis [27,28]. Isolated V1 complexes without subunit H show significantly higher levels of ATPase activity compared to the wild-type V1 complexes. Therefore, in addition to the acidification defect caused by loss of the H subunit, ATP hydrolysis is increased [29]. In yeast strains lacking subunit H, V1-V0 complexes are inactive and unstable, suggesting that the H subunit serves as an activator of the intact V-ATPase [30].

We propose a model of action in which mutation of the V0 subunit a3 causes reduced acidification, increasing bone density whereas mutation of the V1 subunit H causes excessive ATP hydrolysis, decreasing bone density through increased MMP9/MMP13 activity (Fig 5). Supporting this hypothesis is that knockdown of Atp6i in B16-F10 cells causes decreased expression of Mmp9 [31]. In our model, although acidification by V-ATPase is marginalized when ATP6V1H is mutated, the increase in MMP9 /MMP13 activity related to the constitutive Mg-ATPase activity of the cytosolic V1 sector, dominates, resulting in a net reduction in bone density. Enhanced ATP hydrolysis by cytosolic Mg-ATPase activity might activate signaling pathways to stimulate mmp9 and mmp13 expression, a topic for future research. In any event, the bone defect caused by atp6v1h deficiency through activation of mmp9/13 represents a novel finding that offers potential therapeutic entry points for related diseases, since MMP9 levels have been correlated to severity of osteoporosis [32].

Fig 5. Model of action.

A schema showing Vacuolar ATPase and its subunits. Subunit H (V1, cytosolic domain) is shown in green. Under physiologic conditions, ATP hydrolysis occurs in the cytosolic domain while proton translocation occurs through the membrane domain. The V1H subunit activates ATP-driven proton pumping in the intact V-ATPase complex, but when V1 and V0 are disassociated, inhibition of the Mg-ATPase activity of the V1 domain occurs to prevent unnecessary ATP hydrolysis. When V1H is mutated, significantly higher levels of ATPase activity compared to the wild-type V1 complexes is expected, causing increased matrix degradation, leading to decreased bone density via increased MMP9/MMP13 activity.

Materials and Methods

Ethics statement

Patients were enrolled in the NIH Undiagnosed Diseases Program [57] and in protocol 76-HG-0238, “Diagnosis and Treatment of Patients with Inborn Errors of Metabolism and Other Genetic Disorders”, approved by the NHGRI Institutional Review Board. All patients involved in this study gave written, informed consent. All zebrafish experiments were approved by the UCLA Animal Care and Use Committee and the NHGRI ACUC Committee.


To establish a mutant ATP6V1H zebrafish line, gRNA targeting the fourth exon of ATP6V1H was designed and shown to induce indels by injection with Cas9 mRNA in zebrafish. Several mutant lines were established using gRNA with the sequence 5’-GTGTGTCATCAATCAGGGTCAGG; the line used in this study carried a 17 bp insertion and produced a truncated peptide of 94aa, whereas the wild type ATP6V1H encodes a 463aa protein. Stable heterozygous carriers of the mutation were mated to produce homozygous mutants. Details of mutation generation and characterization are presented in the supplementary section and S3 Fig. Staining for bone and cartilage cells was performed as described [33,34].

Quantitative PCR

Total RNA was extracted in Trizol Reagent (Life Technologies) according to the manufacturer’s protocol. cDNA was synthesized using Superscript III (Life Technologies) and oligodT (Life Technologies) following the manufacturer’s protocol. Quantitative real-time PCR was performed using FastStart SYBR Green Master (Roche). Primer sequences are listed in S1 Table. Values of three independent samples (n = 50 each sample) are shown.

siRNA knockdown of Atp6v1h in mouse cells

See supplementary materials.

Whole-mount in situ hybridization

The probes listed in the figures were generated by PCR using zebrafish embryonic cDNA (Primers in S1 Table). Regular and double fluorescence whole-mount in situ hybridization was, respectively, performed as previously described [35,36].

Micro-computed tomography analysis and TRAP staining in adult zebrafish

For X-ray imaging, adult zebrafish were fixed in 10% neutral-buffered formalin for 24 hours and X-ray images were taken using Faxitron MX-20 System Digital Radiography system (Faxitron Bioptics, Tucson, Arizona). For micro-computed tomography (μ-CT), adult zebrafish were fixed in 10% neutral-buffered formalin for 24hours and mounted in 1% low-melt agarose (Sigma) in a plastic vial. Specimens were scanned using SkyScan1172 Micro-CT scanner (Bruker MicroCT, Kontich, Belgium) with the X-ray power at 55kVp and 185μA, and specimens were scanned at a 10 μm voxel resolution. A three-dimensional reconstruction was generated with NRecon software from the set of scans. The images were then analyzed using CTAn (v.1.14.4) and CTVol (v.2.2) softwares. The first caudal bones were assessed for bone mineral density (BMD), bone volume (BV), and bone surface (BS). For TRAP staining, fixed fish were decalcified and sectioned (5 um) after paraffin embedding. TRAP staining was performed using a kit (Sigma 387A-KT) according to the manufacturer's instructions.

Treatment with MMP inhibitors

MMP9 Inhibitor II (Cat# 444293), MMP13 Inhibitor (Cat# 444283) and MMP9/13 Inhibitor I (Cat# 444252) were purchased from EMD Millipore. Zebrafish were treated with designated concentrations of each inhibitor from 1 to 6 days post fertilization in 2 ml of fish water.

Supporting Information

S1 Fig. Additional clinical features.

  1. A. Table shows the levels of biochemistry findings from Patient 1 (P1) and Patient 2 (P2), and Bone mineral densitometry scores.
  2. B. Radiographic pictures from I.2 (father of P1) showing scoliosis and fractured fibula.
  3. C. Dual-energy x-ray absorptiometry (DXA) results of I.2.


S2 Fig. Mutant ATP6V1H protein is degraded more rapidly than wild type protein.

Cell lysates form HEK cells transfected with full length ATP6V1H wild-type or mutant (c.1158_1159delinsTT) cDNA and treated with cycloheximide were subjected to western blot. Figure shows ATP6V1H bands at 0 hour (baseline), 6 hours, 12 hours, and 24 hours. Protein loading is normalized with GAPDH. Experiments were done in three replicates.


S3 Fig. Generation and Genotyping of atp6v1h zebrafish mutants.

Protein alignment of ATP6V1H in human (NP_057025.2) and zebrafish (NP_775377.1) shows high homology (~85%). More importantly, the region where the mutation is located is highly conserved (A). Using CRISPR/Cas9, guide RNA (gRNA) targeting exon 4 of zebrafish atp6v1h was co-injected with Cas9 mRNA into zebrafish embryos and detected by T7 endonuclease digestion. Sequence of gRNA and target is shown (B). Several founders with indels were screened for germline transmission, and the one with 17bp insertion was used throughout the study; The 17bp insertion is predicted to produce a termination codon 94 amino acids after initiation of start codon (C). Genotyping done by polyacrylamide gel electrophoresis (PAGE) analysis demonstrating a 168 bp band for wild type allele, and 185 bp band for mutant allele; representative results for wild type (+/+), heterozygous (+/-), and homozygous (-/-) embryos are shown (D). Using probes specific for atp6v1h, RNA in situ hybridization analysis of atp6v1h showed expression in the head region in wild type embryos, while the expression is nearly absent in homozygous atp6v1h embryos, suggesting that the 17bp insertion may lead to nonsense-mediated decay (NMD) of mRNA and result in loss-of-function of this gene (E).


S4 Fig. Co-expression analysis of atp6v1h with osteoclast or osteoblast markers.

Double fluorescence RNA whole mount in situ hybridization with atp6v1h and rank or runx2a probes were performed on 60 hpf embryos and imaged by confocal microscopy. Upper panels are ventral views of merged lower magnification images of runx2a and atp6v1h or rank and atp6v1h, in which the boxes indicate the regions shown in the lower panels of higher magnification. The lower panels are images of individual colors (green fluorescence for probes of runx2a or rank, red fluorescence for atp6v1h) and merged color (runx2a + atp6v1h, rank+ atp6v1h). Arrowheads point to cells that are mainly positive for runx2a and arrows point to cells that appear positive both for rank and atp6v1h (co-expression: yellow fluorescence).


S5 Fig. Rescue of atp6v1h mutant zebrafish.

Bone staining of uninjected wild type and atp6v1h -/- embryos; loss of bone mineralization is seen in atp6v1h mutant embryos (A and B). atp6v1h mutant embryos at single-cell stage injected with wild type mRNA (C, 200 pg; D, 300 pg) were stained at 5dpf; representative figures demonstrate the rescue of bone phenotype, seen as increase in bone mineralization. atp6v1h mutant embryos injected with mutant mRNA containing the same K386N and N387Y mutations found in humans (E, 200 pg; F, 300 pg) at single-cell stage and stained at 5 dpf; representative figures show the absence of bone staining, indicating that the mutations create a non-functional gene. Wild type and mutant mRNA were injected into wild type embryos at single-cell stage and stained at 5 dpf (G and H); representative figures show no alteration in morphology and bone staining, suggesting the absence of gain-of-function resulting from overexpression of either wild type or mutant mRNA.


S6 Fig. Micro-CT analysis of heterozygous adult bone phenotype.

A. Micro-CT data of the first to fifth caudal vertebrae. Wild type sibling +/+. Heterozygous +/-. BMD, BV and BS designate bone mineral density, bone volume and bone surface (n = 3). B. Micro-CT images of bone in 8-month-old male adult wild type zebrafish and heterozygous allele of retrovirus insertion mutant, atp6v1hhi923+/- (n = 5 each).


S7 Fig. RNA in situ hybridization analysis of osteoblast and osteoclast gene markers.

In situ hybridization markers of osteoblast development (runx2a, runx2b, osx, osn, col1a2) show slight reduction between 5dpf siblings (wild type) or mutant (-/-).


S8 Fig. siRNA knockdown of Atp6v1h in mouse osteoclast, osteoblast and chondrocyte cells.

siRNA knockdown of atp6v1h was performed in mouse chondrocytes (ATDC5) and osteoblasts (MC3T3). Knockdown was confrmed by qPCR, showing reduced Atp6v1h expression as normalized with Gapdh (A and D). Mmp9 (B and E) and Mmp13 (C and F) in chondrocytes and osteoblasts are slightly reduced after Atp6V1h knockdown, suggesting that Atp6v1h may not have a profound impact on osteoblasts or chondrocytes.


S9 Fig. Analysis of osteoclast maturation by TRAP.

Differentiation of Raw264.7 (pre-osteoclast) cell was induced by RANKL (100ng/ml). Raw264.7 was transfected by siRNA anti-ATP6V1H 2 times, and osteoclast cells were identified by TRAP staining (red color). (A) The ATP6V1H siRNA groups have significantly more mature osteoclasts (red staining of large multi-nuclei cells) than the scrambled controls. Included here are sets of five randomly selected fields of images of ATP6V1H siRNA groups and scrambled controls. (B) The ratio of TRAP positive cells and total cells is significantly higher in ATP6V1H siRNA group. Specific data is listed in S2 Table.


S10 Fig. Analysis of osteoclast by TRAP in adult zebrafish.

Adult zebrafish (10 month old) of +/- for atp6v1h (line atp6v1hla439) and wild type sibling were fixed, sectioned and stained by TRAP. Atp6v1hla439+/- fish appear to have more TRAP staining compared to their sibling (n = 3, two images are shown here for wild type and +/- mutant). Scale bar: 50 um.


S1 Table. Table lists the primer sequences used in qPCR analysis and generation of probes for in situ hybridization.


S2 Table. Cell counts of TRAP-negative and positive Raw cells induced by RANKL.


S1 Methods. This section describes the methods used to generate supplementary figures in the study.



We thank Jason Ear for editing the paper and helpful discussion. We thank Linda Yuan for maintaining the fish used in this study. We also thank Dr. Alessandra Romani (Modena, Italy) and Dr. Cesari Carani (Endocrinology and Metabolism Unit, Department of Medicine, Endocrinology and Metabolism, Geriatrics University of Modena and Reggio Emilia, NOCSAE of Baggiovara, Modena, Italy) for referring the patient to the NIH-UDP Program.

Author Contributions

  1. Conceptualization: YZ HH CFB WAG XD SL MCVM.
  2. Formal analysis: YZ HH GZ TY HV CT VM JA KL WAG SL MCVM.
  3. Funding acquisition: SL WAG.
  4. Investigation: YZ HH GZ TY YH CT KL WAG SL MCVM.
  5. Methodology: YZ HH GZ HV CBF KL RS KB WAG SL MCVM.
  6. Project administration: KL CBF WAG SL MCVM.
  7. Resources: YZ HH GZ HV RS KB CT SL MCVM.
  8. Supervision: CBF RS KL WAG SL MCVM.
  9. Validation: YZ HH GZ TY YH KL XD VM JA RS KB SL.
  10. Visualization: YZ HH GZ TY CT KL XD WAG MCVM SL.
  11. Writing – original draft: YZ HH GZ KL WAG SL MCVM.
  12. Writing – review & editing: YZ HH GZ TY HV YH CT CBF RS KB VM JA KL WAG SL MCVM.


  1. 1. Qin A, Cheng TS, Pavlos NJ, Lin Z, Dai KR, et al. (2012) V-ATPases in osteoclasts: structure, function and potential inhibitors of bone resorption. Int J Biochem Cell Biol 44: 1422–1435. pmid:22652318
  2. 2. Karet FE, Finberg KE, Nelson RD, Nayir A, Mocan H, et al. (1999) Mutations in the gene encoding B1 subunit of H+-ATPase cause renal tubular acidosis with sensorineural deafness. Nat Genet 21: 84–90. pmid:9916796
  3. 3. Xu J, Cheng T, Feng HT, Pavlos NJ, Zheng MH (2007) Structure and function of V-ATPases in osteoclasts: potential therapeutic targets for the treatment of osteolysis. Histol Histopathol 22: 443–454. pmid:17290355
  4. 4. Li YP, Chen W, Liang YQ, Li E, Stashenko P (1999) Atp6i-deficient mice exhibit severe osteopetrosis due to loss of osteoclast-mediated extracellular acidification. Nature Genetics 23: 447–451. pmid:10581033
  5. 5. Gahl WA, Markello TC, Toro C, Fajardo KF, Sincan M, et al. (2012) The National Institutes of Health Undiagnosed Diseases Program: insights into rare diseases. Genet Med 14: 51–59. pmid:22237431
  6. 6. Gahl WA, Tifft CJ (2011) The NIH Undiagnosed Diseases Program: lessons learned. JAMA 305: 1904–1905. pmid:21558523
  7. 7. Tifft CJ, Adams DR (2014) The National Institutes of Health undiagnosed diseases program. Curr Opin Pediatr 26: 626–633. pmid:25313974
  8. 8. Nuckels RJ, Ng A, Darland T, Gross JM (2009) The vacuolar-ATPase complex regulates retinoblast proliferation and survival, photoreceptor morphogenesis, and pigmentation in the zebrafish eye. Invest Ophthalmol Vis Sci 50: 893–905. pmid:18836173
  9. 9. de Vrieze E, Sharif F, Metz JR, Flik G, Richardson MK (2011) Matrix metalloproteinases in osteoclasts of ontogenetic and regenerating zebrafish scales. Bone 48: 704–712. pmid:21185415
  10. 10. Nannuru KC, Futakuchi M, Varney ML, Vincent TM, Marcusson EG, et al. (2010) Matrix metalloproteinase (MMP)-13 regulates mammary tumor-induced osteolysis by activating MMP9 and transforming growth factor-beta signaling at the tumor-bone interface. Cancer Res 70: 3494–3504. pmid:20406980
  11. 11. Dufour A, Sampson NS, Li J, Kuscu C, Rizzo RC, et al. (2011) Small-molecule anticancer compounds selectively target the hemopexin domain of matrix metalloproteinase-9. Cancer Res 71: 4977–4988. pmid:21646471
  12. 12. Cheng M, De B, Pikul S, Almstead NG, Natchus MG, et al. (2000) Design and synthesis of piperazine-based matrix metalloproteinase inhibitors. J Med Chem 43: 369–380. pmid:10669564
  13. 13. Inoue K, Mikuni-Takagaki Y, Oikawa K, Itoh T, Inada M, et al. (2006) A crucial role for matrix metalloproteinase 2 in osteocytic canalicular formation and bone metabolism. J Biol Chem 281: 33814–33824. pmid:16959767
  14. 14. Vu TH, Shipley JM, Bergers G, Berger JE, Helms JA, et al. (1998) MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell 93: 411–422. pmid:9590175
  15. 15. Inada M, Wang Y, Byrne MH, Rahman MU, Miyaura C, et al. (2004) Critical roles for collagenase-3 (Mmp13) in development of growth plate cartilage and in endochondral ossification. Proc Natl Acad Sci U S A 101: 17192–17197. pmid:15563592
  16. 16. Stickens D, Behonick DJ, Ortega N, Heyer B, Hartenstein B, et al. (2004) Altered endochondral bone development in matrix metalloproteinase 13-deficient mice. Development 131: 5883–5895. pmid:15539485
  17. 17. Chambers TJ, Fuller K (1985) Bone cells predispose bone surfaces to resorption by exposure of mineral to osteoclastic contact. J Cell Sci 76: 155–165. pmid:4066784
  18. 18. Kahn AJ, Partridge NC (1987) New concepts in bone remodeling: an expanding role for the osteoblast. Am J Otolaryngol 8: 258–264. pmid:3324780
  19. 19. Zhao W, Byrne MH, Boyce BF, Krane SM (1999) Bone resorption induced by parathyroid hormone is strikingly diminished in collagenase-resistant mutant mice. J Clin Invest 103: 517–524. pmid:10021460
  20. 20. Everts V, Delaisse JM, Korper W, Niehof A, Vaes G, et al. (1992) Degradation of collagen in the bone-resorbing compartment underlying the osteoclast involves both cysteine-proteinases and matrix metalloproteinases. J Cell Physiol 150: 221–231. pmid:1734028
  21. 21. Lynch MP, Stein JL, Stein GS, Lian JB (1995) The influence of type I collagen on the development and maintenance of the osteoblast phenotype in primary and passaged rat calvarial osteoblasts: modification of expression of genes supporting cell growth, adhesion, and extracellular matrix mineralization. Exp Cell Res 216: 35–45. pmid:7813631
  22. 22. Onodera S, Nishihira J, Iwabuchi K, Koyama Y, Yoshida K, et al. (2002) Macrophage migration inhibitory factor up-regulates matrix metalloproteinase-9 and -13 in rat osteoblasts. Relevance to intracellular signaling pathways. J Biol Chem 277: 7865–7874. pmid:11751895
  23. 23. Sharif F, de Bakker MA, Richardson MK (2014) Osteoclast-like Cells in Early Zebrafish Embryos. Cell J 16: 211–224. pmid:24567948
  24. 24. Reponen P, Sahlberg C, Munaut C, Thesleff I, Tryggvason K (1994) High expression of 92-kD type IV collagenase (gelatinase B) in the osteoclast lineage during mouse development. J Cell Biol 124: 1091–1102. pmid:8132709
  25. 25. Takeshita S, Kaji K, Kudo A (2000) Identification and characterization of the new osteoclast progenitor with macrophage phenotypes being able to differentiate into mature osteoclasts. J Bone Miner Res 15: 1477–1488. pmid:10934646
  26. 26. Jackson MT, Moradi B, Smith MM, Jackson CJ, Little CB (2014) Activation of matrix metalloproteinases 2, 9, and 13 by activated protein C in human osteoarthritic cartilage chondrocytes. Arthritis Rheumatol 66: 1525–1536. pmid:24574263
  27. 27. Jefferies KC, Forgac M (2008) Subunit H of the vacuolar (H+) ATPase inhibits ATP hydrolysis by the free V1 domain by interaction with the rotary subunit F. J Biol Chem 283: 4512–4519. pmid:18156183
  28. 28. Parra KJ, Keenan KL, Kane PM (2000) The H subunit (Vma13p) of the yeast V-ATPase inhibits the ATPase activity of cytosolic V1 complexes. J Biol Chem 275: 21761–21767. pmid:10781598
  29. 29. Rizzo JM, Tarsio M, Martinez-Munoz GA, Kane PM (2007) Diploids heterozygous for a vma13Delta mutation in Saccharomyces cerevisiae highlight the importance of V-ATPase subunit balance in supporting vacuolar acidification and silencing cytosolic V1-ATPase activity. J Biol Chem 282: 8521–8532. pmid:17234635
  30. 30. Ho MN, Hirata R, Umemoto N, Ohya Y, Takatsuki A, et al. (1993) VMA13 encodes a 54-kDa vacuolar H(+)-ATPase subunit required for activity but not assembly of the enzyme complex in Saccharomyces cerevisiae. J Biol Chem 268: 18286–18292. pmid:8349704
  31. 31. Nishisho T, Hata K, Nakanishi M, Morita Y, Sun-Wada GH, et al. (2011) The a3 isoform vacuolar type H(+)-ATPase promotes distant metastasis in the mouse B16 melanoma cells. Mol Cancer Res 9: 845–855. pmid:21669964
  32. 32. Bolton CE, Stone MD, Edwards PH, Duckers JM, Evans WD, et al. (2009) Circulating matrix metalloproteinase-9 and osteoporosis in patients with chronic obstructive pulmonary disease. Chron Respir Dis 6: 81–87. pmid:19411568
  33. 33. Walker MB, Kimmel CB (2007) A two-color acid-free cartilage and bone stain for zebrafish larvae. Biotech Histochem 82: 23–28. pmid:17510811
  34. 34. Du SJ, Frenkel V, Kindschi G, Zohar Y (2001) Visualizing normal and defective bone development in zebrafish embryos using the fluorescent chromophore calcein. Dev Biol 238: 239–246. pmid:11784007
  35. 35. Jowett T (1999) Analysis of protein and gene expression. Methods Cell Biol 59: 63–85. pmid:9891356
  36. 36. Brend T, Holley SA (2009) Zebrafish whole mount high-resolution double fluorescent in situ hybridization. J Vis Exp.