ATP6AP2-to-MMP14, a key pathway for osteoblast to osteocyte transition
Curation statements for this article:-
Curated by eLife
eLife assessment
This manuscript is of interest to readers in the field of bone biology. It identifies a novel role for the vacuolar ATPase accessory protein ATP6AP2 within the osteoblast lineage and shows that loss of ATP6AP2 in the mature osteoblast results in disorganized bone formation. A similar, but milder, bone disorganization phenotype is also observed when this gene is knocked out in osteocytes. The authors show that this bone phenotype is partially rescued via restoration of MMP14 action.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Osteocytes, derived from osteoblast (OB)-lineage, occupy lacunae within bone matrix and exhibit unique morphology with dendrite-like projections to form an inter-connected network. Such a network is essential for osteocytes to monitor and orchestrate bone homeostasis. Thus, it is of considerable interest to investigate how osteocytes are formed and how they built the network. Here we provide evidence for ATP6AP2, an accessory subunit of V-ATPase, in OB-lineage cells to be critical for OB-to-osteocyte transition, and osteocyte distribution, maturation, and morphogenesis. Mice (ATP6AP2 Ocn-Cre ) that selectively deplete ATP6AP2 in OB-lineage cells results in altered osteocyte distribution and morphology, impaired osteocyte maturation and dendrite-like processes, increased osteocyte cell death and cortical woven bone formation. Further mechanistic studies identify MMP14 (matrix metalloproteinase-14) as a critical downstream of ATP6AP2 for osteocyte differentiation. ATP6AP2 interacts with MMP14 and promotes MMP14 surface distribution largely in immature- or osteoid-osteocytes, where this pathway regulates bone matrix remodeling and osteocyte differentiation. Expression of MMP14 into ATP6AP2 knock out OB-lineage cells in the mouse cortical bone could diminish the deficits in osteocyte maturation, dendrite-like process formation, and survival. These results thus demonstrate an un-recognized function of ATP6AP2 in promoting OB-to-osteocyte transition and uncover a pathway from ATP6AP2-to-MMP14 in osteocyte differentiation.
Article activity feed
-
Author Response
Reviewer #1 (Public Review):
“Overall this is an interesting study of the function of ATP6AP2 in the osteoblastic lineage. This gene is unstudied in the osteoblast, despite its known role in WNT signaling. In this study, the authors first show that loss of this gene in mature osteoblasts results in a strong cortical bone phenotype, with reduced osteocyte numbers and disorganized collagen. This phenotype is not present at birth but progressively worsens as the animals reach weaning age. In the compact bone, they show that loss of ATP6AP2 results in osteocytes largely devoid of dendritic processes. Loss of this gene starting at the osteocyte stage results in a milder phenotype. They then show that the osteocytes presenting have reduced MMP14 and that partial restoration of MM14 attenuates the severity of the cortical …
Author Response
Reviewer #1 (Public Review):
“Overall this is an interesting study of the function of ATP6AP2 in the osteoblastic lineage. This gene is unstudied in the osteoblast, despite its known role in WNT signaling. In this study, the authors first show that loss of this gene in mature osteoblasts results in a strong cortical bone phenotype, with reduced osteocyte numbers and disorganized collagen. This phenotype is not present at birth but progressively worsens as the animals reach weaning age. In the compact bone, they show that loss of ATP6AP2 results in osteocytes largely devoid of dendritic processes. Loss of this gene starting at the osteocyte stage results in a milder phenotype. They then show that the osteocytes presenting have reduced MMP14 and that partial restoration of MM14 attenuates the severity of the cortical phenotype.”
Strengths
“This study uses cutting-edge microscopy to thoroughly characterize how and where the loss of ATP6AP2 in either the mature osteoblast or the osteocyte results in disorganized bone. Innovative proteomics techniques are used to identify cell surface proteins, including MMP14 that may mediate this phenotype. Two cre-drivers are used to determine when in the osteoblast-osteocyte lineage this gene has the maximum effect. Lastly, in vivo lentivirus replacement is used to test if the replacement of MMP14 can rescue the phenotype. This latter experiment solidifies the importance of MMP14 as a major player in the downstream sequela of ATP6AP2 action.”
Weaknesses
“Unfortunately, all of the histology is conducted on demineralized bone, and counts of osteoblasts and osteoclasts on the bone surface are not presented. This reduces the ability to interpret all downstream work. As such, the extent of the mineralization defects is difficult to interpret. Much of this paper is focused on the osteocyte, which is curious as the phenotype of the mature osteoblasts ATP6AP2 knockout mice is so much more severe than that of the osteocyte ATP6AP2 knockout mice. While it is clear how MMP14 was identified as being deficient in the mature osteoblasts ATP6AP2 knockout cells, it is not obvious how this gene became the sole focus of the remainder of this paper. This phenotype progresses as the mice become ambulatory and therefore weight bearing on their limbs. This could partially explain the presentation of the mouse phenotype, but this is not discussed.”
Good suggestions! We have performed the suggested experiments on mineralized bone sections, and quantified both osteoblasts and osteoclasts on the bone surface.
The results, shown in Fig. 1A-B above, demonstrated increased osteoclast numbers in both trabecular and endocortical bone surfaces in ATP6AP2 mutant mice, which were accompanied with elevated bone resorption (see Fig. 1C, measured by serum levels of PYD). However, upon bisphosphonates (alendronate) treatment, an inhibitor of osteoclastic activity, the trabecular bone mass was restored, but little effect on the cortical bone phenotype in the mutant mice (Fig. 2A-F). These results thus suggest an osteoclast activity independent cortical bone phenotype in the mutant mice.
We thus further investigated the cortical bone phenotypes and the osteoclast independent underlying mechanisms. Whereases no significant change in the number of osteoblasts was detected in the metaphysis region of femur in ATP6AP2Ocn-cre mice (see Fig. 6A-B below), we did detect mineralization deficit in the mutant mice by both in vivo and in vitro experiments (see Fig. 5 and Fig. 7). These results suggest that the increased cortical woven bone in the mutant mice is likely due to an impairment in the replacement of woven bone with the mineralized cortical bone matrix.
Additionally, the expression levels of ATP6AP2 in osteocytes appeared to be similar to that in osteoblasts and BMSCs (Fig. 8). The phenotype of osteocyte ATP6AP2 knockout mice (ATP6AP2DMP-Cre) appeared to be weaker than that of the BMSCs/osteoblasts ATP6AP2 knockout mice (ATP6AP2OCN-Cre) led us to speculate that ATP6AP2 in Ocn-Cre+ osteoblastic cells (e.g., immature osteocytes) may play a more critical role than that in DMP1-Cre+ mature osteocytes in regulating cortical bone matrix remodeling and osteocyte development.
These results and points, in line with our model, will be included into a revised manuscript.
Reviewer #3 (Public Review):
“In this work, the authors have assessed the bone phenotype of a mouse with targeted ablation of the vacuolar ATPase accessory protein ATP6AP2 in the osteoblast lineage. They observe a clear increase in cortical thickness, but the cortex is highly porous and contains remnant cartilage as well as extensive woven bone. They then follow this by suggesting that one cause of this phenotype may be a change in the surface expression of the protein MMP14, a matrix metalloproteinase, known to be involved in bone matrix degradation, at least in osteoclasts. They provide evidence that this protein may also regulate matrix degradation surrounding osteocytes and an increase in this protein in osteocytes lacking ATP6AP2 may be a cause of the initial phenotype described.”
While the phenotype described is very dramatic, the interpretation that it reflects a defect in osteoblast to osteocyte transition is questioned by this reviewer. The phenotype appears to be an osteopetrosis, including a lack of remodelling of the cortex. Cartilage and woven bone are not replaced effectively by lamellar bone. The bone contains ample osteocytes, but they are the osteocytes typical of woven bone, with rounded cell bodies, disordered organisation, low sclerostin expression, and short dendritic processes. The defect in the ATP6AP2 mice is a lack of cortical remodelling during cortical consolidation (for review see PMID: 34196732). Cartilage and woven bone remnants, which are normally remodelled as cortical bone matures, remain in the cortex until adulthood. It is not clear whether this results from reduced or increased remodelling of the cortex, but it is not because the osteoblasts cannot form osteocytes.
Some of the data is very challenging to interpret because of low sample numbers (n=4 for much of the analysis), and lack of detail as to the sex of the animals. Regions used for imaging, histomorphometry, and dynamic histomorphometry all need to be defined throughout the work. Since the cortex differs dramatically by site, and by distance from the growth plate (due to the different stages of maturation) this is critical. Some methods are not defined, although they could be of great use to the field (e.g. the method for assessing bone degradation by MMP14).
Good suggestions! We will describe the results more precisely in a revised manuscript.
-
eLife assessment
This manuscript is of interest to readers in the field of bone biology. It identifies a novel role for the vacuolar ATPase accessory protein ATP6AP2 within the osteoblast lineage and shows that loss of ATP6AP2 in the mature osteoblast results in disorganized bone formation. A similar, but milder, bone disorganization phenotype is also observed when this gene is knocked out in osteocytes. The authors show that this bone phenotype is partially rescued via restoration of MMP14 action.
-
Reviewer #1 (Public Review):
Overall this is an interesting study of the function of ATP6AP2 in the osteoblastic lineage. This gene is unstudied in the osteoblast, despite its known role in WNT signaling. In this study, the authors first show that loss of this gene in mature osteoblasts results in a strong cortical bone phenotype, with reduced osteocyte numbers and disorganized collagen. This phenotype is not present at birth but progressively worsens as the animals reach weaning age. In the compact bone, they show that loss of ATP6AP2 results in osteocytes largely devoid of dendritic processes. Loss of this gene starting at the osteocyte stage results in a milder phenotype. They then show that the osteocytes presenting have reduced MMP14 and that partial restoration of MM14 attenuates the severity of the cortical phenotype.
Strengths
Th…
Reviewer #1 (Public Review):
Overall this is an interesting study of the function of ATP6AP2 in the osteoblastic lineage. This gene is unstudied in the osteoblast, despite its known role in WNT signaling. In this study, the authors first show that loss of this gene in mature osteoblasts results in a strong cortical bone phenotype, with reduced osteocyte numbers and disorganized collagen. This phenotype is not present at birth but progressively worsens as the animals reach weaning age. In the compact bone, they show that loss of ATP6AP2 results in osteocytes largely devoid of dendritic processes. Loss of this gene starting at the osteocyte stage results in a milder phenotype. They then show that the osteocytes presenting have reduced MMP14 and that partial restoration of MM14 attenuates the severity of the cortical phenotype.
Strengths
This study uses cutting-edge microscopy to thoroughly characterize how and where the loss of ATP6AP2 in either the mature osteoblast or the osteocyte results in disorganized bone. Innovative proteomics techniques are used to identify cell surface proteins, including MMP14 that may mediate this phenotype. Two cre-drivers are used to determine when in the osteoblast-osteocyte lineage this gene has the maximum effect. Lastly, in vivo lentivirus replacement is used to test if the replacement of MMP14 can rescue the phenotype. This latter experiment solidifies the importance of MMP14 as a major player in the downstream sequela of ATP6AP2 action.
Weaknesses
Unfortunately, all of the histology is conducted on demineralized bone, and counts of osteoblasts and osteoclasts on the bone surface are not presented. This reduces the ability to interpret all downstream work. As such, the extent of the mineralization defects is difficult to interpret. Much of this paper is focused on the osteocyte, which is curious as the phenotype of the mature osteoblasts ATP6AP2 knockout mice is so much more severe than that of the osteocyte ATP6AP2 knockout mice. While it is clear how MMP14 was identified as being deficient in the mature osteoblasts ATP6AP2 knockout cells, it is not obvious how this gene became the sole focus of the remainder of this paper. This phenotype progresses as the mice become ambulatory and therefore weight bearing on their limbs. This could partially explain the presentation of the mouse phenotype, but this is not discussed.
-
Reviewer #2 (Public Review):
The authors examine the effects of depletion of an accessory subunit of the V-ATPase, ATP6AP2, using recombination of a floxed gene with osteocalcin promoter cre recombinase. Major findings are that defects and death in osteocytes occur, with mass spectrometry sequencing showing that matrix metalloproteinase, MMP14, which is involved in collagen remodeling in a number of other contexts, regulates bone matrix remodeling and osteocyte differentiation downstream of ATP6AP2. Further, ATP6AP2 depletion was counteracted in part by direct expression of MMP14 in ATP6AP2 depleted osteoblast-lineage cells.
Major strengths of the work include a clear description of methods and most results, as well as a concise and clear discussion.
- There is an extensive description of the bone with a detailed discussion of micro …Reviewer #2 (Public Review):
The authors examine the effects of depletion of an accessory subunit of the V-ATPase, ATP6AP2, using recombination of a floxed gene with osteocalcin promoter cre recombinase. Major findings are that defects and death in osteocytes occur, with mass spectrometry sequencing showing that matrix metalloproteinase, MMP14, which is involved in collagen remodeling in a number of other contexts, regulates bone matrix remodeling and osteocyte differentiation downstream of ATP6AP2. Further, ATP6AP2 depletion was counteracted in part by direct expression of MMP14 in ATP6AP2 depleted osteoblast-lineage cells.
Major strengths of the work include a clear description of methods and most results, as well as a concise and clear discussion.
- There is an extensive description of the bone with a detailed discussion of micro computed tomography and staining results.
- Interesting findings include retention of woven bone, and labeling for secondary indicators including cleaved caspase 3, RunX2, and sclerostin.
- Osteocyte tomato labeling of the ATP6AP2Ocn-cre animals is a very good confirmation of the histomorphometric analysis.
- The KI67 labeling of proliferative cells is very interesting but should be introduced more clearly. Similarly, cleaved caspase 3 is very useful but a sentence stating why this was done would be useful for clarity.
- Interaction of ATP6AP2 directly with MMP14 is very interesting and useful in wrapping up the paper.Weaknesses include:
- When introducing assays, a brief description of why this is done would make the paper more accessible.
- The reviewer would like to see a clearer description of the depletion of ATP6AP2 by cre-lox recombination.
- Results showing calcein deposition, not on the surface of the cortical bone requires more data to strengthen this finding.
- Retention of woven bone suggests a defect in resorption, but a clear description of the resorbed area is not seen.The reviewer is enthusiastic about the manuscript.
-
Reviewer #3 (Public Review):
In this work, the authors have assessed the bone phenotype of a mouse with targeted ablation of the vacuolar ATPase accessory protein ATP6AP2 in the osteoblast lineage. They observe a clear increase in cortical thickness, but the cortex is highly porous and contains remnant cartilage as well as extensive woven bone. They then follow this by suggesting that one cause of this phenotype may be a change in the surface expression of the protein MMP14, a matrix metalloproteinase, known to be involved in bone matrix degradation, at least in osteoclasts. They provide evidence that this protein may also regulate matrix degradation surrounding osteocytes and an increase in this protein in osteocytes lacking ATP6AP2 may be a cause of the initial phenotype described.
While the phenotype described is very dramatic, the …
Reviewer #3 (Public Review):
In this work, the authors have assessed the bone phenotype of a mouse with targeted ablation of the vacuolar ATPase accessory protein ATP6AP2 in the osteoblast lineage. They observe a clear increase in cortical thickness, but the cortex is highly porous and contains remnant cartilage as well as extensive woven bone. They then follow this by suggesting that one cause of this phenotype may be a change in the surface expression of the protein MMP14, a matrix metalloproteinase, known to be involved in bone matrix degradation, at least in osteoclasts. They provide evidence that this protein may also regulate matrix degradation surrounding osteocytes and an increase in this protein in osteocytes lacking ATP6AP2 may be a cause of the initial phenotype described.
While the phenotype described is very dramatic, the interpretation that it reflects a defect in osteoblast to osteocyte transition is questioned by this reviewer. The phenotype appears to be an osteopetrosis, including a lack of remodelling of the cortex. Cartilage and woven bone are not replaced effectively by lamellar bone. The bone contains ample osteocytes, but they are the osteocytes typical of woven bone, with rounded cell bodies, disordered organisation, low sclerostin expression, and short dendritic processes. The defect in the ATP6AP2 mice is a lack of cortical remodelling during cortical consolidation (for review see PMID: 34196732). Cartilage and woven bone remnants, which are normally remodelled as cortical bone matures, remain in the cortex until adulthood. It is not clear whether this results from reduced or increased remodelling of the cortex, but it is not because the osteoblasts cannot form osteocytes.
Some of the data is very challenging to interpret because of low sample numbers (n=4 for much of the analysis), and lack of detail as to the sex of the animals. Regions used for imaging, histomorphometry, and dynamic histomorphometry all need to be defined throughout the work. Since the cortex differs dramatically by site, and by distance from the growth plate (due to the different stages of maturation) this is critical. Some methods are not defined, although they could be of great use to the field (e.g. the method for assessing bone degradation by MMP14).
-
