*2.6. Influence of AT1R Depletion on Bone Erosion in TNFtg Mice*

We next examined whether the deletion of AT1R could reduce bone destruction in the arthritic mice. Micro-CT analysis of the ankle joints revealed that TNFtg/AT1R−/<sup>−</sup> mice exhibited the same extent of severe bone loss as TNFtg mice (Figure 5A,B). Additionally, the BV reduction rate and the erosive volume (Ev/Rpv) of the talus in the TNFtg/AT1R−/<sup>−</sup> mice were comparable to those in the TNFtg mice (Figure 5C,D). These findings indicate that AT1R deficiency did not alleviate the destructive bone changes in the inflammatory joints of the arthritic mice. Histological analyses revealed that the extents of bone erosion and osteoclast formation were comparable between TNFtg and TNFtg/AT1R−/<sup>−</sup> mice (Figure 5E–G).

**Figure 5.** Influence of AT1R depletion on bone erosion in tumor necrosis factor-transgenic (TNFtg) mice. TNFtg mice were crossed with AT1R-deficient (AT1R−/−) mice. WT (*n* = 9), AT1R−/<sup>−</sup> (*n* = 7), TNFtg (*n* = 12), and TNFtg AT1R−/<sup>−</sup> (*n* = 8) male mice were analyzed at the age of 16 weeks. (**A**) Representative 3D micro-CT images of the talus. (**B**) Bone volume (BV) of the talus. (**C**) Rate of reduction in BV by

AT1R deficiency relative to that in the control mice of each genotype. (**D**) Eroded volume per repaired volume (Ev/Rpv) of the talus. (**E**) Representative images of tartrate-resistant acid phosphatase (TRAP) staining around the talus bones of indicated mice; original magnification ×40. (**F**) Eroded surface per bone surface (ES/BS) around the taluses. (**G**) The number of osteoclasts per bone surface (N.Oc/BS) around the taluses. Values are the mean ± SEM. n.s., not significant. \*, *p* < 0.05.

#### *2.7. E*ff*ect of AT1R Deficiency on Bone Properties of the Trabecular and Cortical Bones in TNFtg Mice*

AT1R−/<sup>−</sup> mice were previously reported to exhibit an increased trabecular BV and increased trabecular number and connectivity [16]. To examine the effect of AT1R deficiency on the bone volume of systemic bones in the arthritic condition, we analyzed the bone properties of the tibia and vertebra of the TNFtg arthritic mice using micro-CT (Figure 6A–C). The TNFtg mice exhibited a significant reduction in BV/TV of the tibia and the AT1R deficiency modestly alleviated the bone loss caused by arthritis, even though the difference between TNFtg and TNFtg/AT1R−/<sup>−</sup> mice was not statistically significant (Figure 6D). A similar insignificant tendency was observed in the vertebral trabecular bone (Figure 6F). In the tibia cortical bone, AT1R deficiency did not show any protective effect on bone loss (Figure 6E). The other analyzed parameters of the trabecular and cortical bones also indicated no significant effect of AT1R deficiency on bone properties (Figure A4). These findings suggest that the inhibition of endogenous Ang II has a limited protective effect on bone loss in arthritic mice.

**Figure 6.** Effect of AT1R deficiency on bone properties of the trabecular and cortical bones in tumor necrosis factor-transgenic (TNFtg) mice. TNFtg mice were crossed with AT1R-deficient (AT1R−/−) mice. WT (*n* = 9), AT1R−/<sup>−</sup> (*n* = 7), TNFtg (*n* = 12), and TNFtg AT1R−/<sup>−</sup> (*n* = 8) male mice were analyzed at the age of 16 weeks. Representative 2D micro-CT images of the tibia trabecular bone (**A**), the tibia cortical bone (**B**), and the trabecular bone of the spine (fifth lumbar vertebra) (**C**). (**D**) Bone volume per total volume (BV/TV) and reduction rate of the tibia trabecula bone. (**E**) Cortical thickness (Ct.Th) and reduction rate of the tibia midshaft. (**F**) Bone volume per total volume (BV/TV) and reduction rate of the fifth lumbar vertebral trabecular bone. Values are the mean ± SEM. n.s., not significant. \*, *p* < 0.05.

#### **3. Discussion**

In this study, we sought to clarify the impact of excessive Ang II and inhibition of the endogenous RAS on bone erosion and systemic bone loss in a TNF-mediated arthritic condition. We found that the administration of Ang II enhanced destructive bone changes in inflammatory joints without affecting the severity of inflammation. There was no noticeable synergistic effect of Ang II administration and inflammation on osteopenia of the tibia and vertebra in mice. Further, we found that AT1R deficiency had a minimal protective effect on bone erosion and systemic bone loss in the arthritis model.

Interestingly, we observed that the administration of Ang II aggravated joint destruction in the arthritic mice. Ang II has been reported to enhance systemic bone loss in murine osteoporosis models [11,17]. Ang II induces RANKL expression in osteoblasts and subsequently enhances osteoclastogenesis, resulting in systemic bone loss [6,11]. However, no previous studies have explored the role of the RAS in the development of bone erosion in an arthritis model. In rheumatoid arthritis, inflammatory cytokines such as TNF increase RANKL expression in synoviocytes and subsequently promote osteoclastic differentiation and activation, resulting in erosive bone changes in joints [2]. Our results demonstrate that excessive Ang II could exacerbate the TNF-induced inflammatory joint destruction associated with increased osteoclast formation.

The current study has important clinical implications for the management of rheumatoid arthritis. Our findings suggest that in patients in whom the local effect of Ang II is upregulated via increased imported Ang II from circulation, joint destruction can be promoted as a consequence of systemic RAS activation. Systemic activation of the RAS can be observed in several pathological conditions, such as renal artery stenosis, congestive heart failure, cardiac hypertrophy, chronic kidney disease, and obesity [18,19]. Such pathological conditions could be risk factors for progressive joint destruction in inflammatory arthropathies.

Although Ang II appeared to promote bone erosion in inflamed joints, its effects on systemic bones, represented by the tibia and vertebra, were found to be very limited. There are several possible explanations for this. Firstly, in the arthritic joints of mice, other inflammatory cytokines such as IL-1 and IL-6 are highly produced [2,20]. In addition to TNF, these other osteoclast-activating factors might play important synergistic roles in the Ang II-promoted bone erosion in joints. Secondly, the expression of AT1R was significantly increased in the arthritic joints (Figure 1A). This could contribute to hyper-responsiveness to Ang II, resulting in increased osteoclastic bone destruction in the joints. Thirdly, the exposure period to Ang II (4 weeks in this study) might be too short for this effector to exert an osteopenic effect on systemic bones. Indeed, a previous study showed a significant osteopenic effect of excessive RAS activation in 6-month-old Tsukuba hypertensive mice that were continuously exposed to excessive Ang II via transgenes encoding human renin and human angiotensinogen [17].

Since the expression of AT1R was increased in the arthritic joints of the TNFtg mice (Figure 1A), we assumed that AT1R deficiency would ameliorate bone erosion in this arthritis model. Contrary to our expectation, AT1R deficiency did not significantly improve the erosive bone changes in the TNFtg mice. These data indicate a limited role of the local RAS during the process of joint destruction in this arthritic model. The RAS might modulate bone mass only under pathological conditions with excessive systemic activation. Analyses of AT1R-deficient arthritic mice with excessive Ang II would be needed to verify this concept.

The limitation of our study is that the precise mechanisms through which Ang II enhances bone erosion are unclear. We have tested the effect of Ang II on osteoclast differentiation in murine primary bone marrow-derived macrophage cultures. Ang II stimulation did not promote osteoclast formation in the mono-culture of bone marrow-derived macrophage (Figure A5A,B), whereas Ang II enhanced osteoclast formation in the co-culture system with osteoblasts (Figure A5C,D). These data suggest that Ang II promotes osteoclast formation indirectly via stromal cells. In support of this notion, Ang II has been previously reported to induce RANKL expression in stromal cells [17]. Various cells can express RANKL in arthritic joints, synovial cells, osteoblasts, or osteocytes, which might be attributed to the

Ang II-mediated bone erosion. Other possibilities are that Ang II regulates angiogenesis in arthritic joints or that Ang II modulates cellular functions via the Ang II type 2 receptor, which reportedly regulates inflammation in the arthritic synovium [21]. Further research will be required to clarify the underlying mechanisms.

Another possible limitation of this study is the relatively small sample sizes which may have insufficient statistical power to detect a small difference in some comparisons. For instance, a statistically significant difference was not detected in the trabecular BV/TV of the tibia between H2O-treated WT (*n* = 6) and H2O-treated TNFtg (*n* = 4) mice (Figure 3D), although there is a statistically significant difference between WT (*n* = 9) and TNFtg (*n* = 12) mice in the AT1R−/<sup>−</sup> strain (Figure 6D). Post-hoc power analyses have shown that a larger sample size would be needed to detect a substantial difference in Figure 3D. Therefore, future studies with larger sample sizes would be necessary to detect a small but significant difference.

We previously reported that the RAS is involved in vascular damage and that AT1R blockers have potent vascular protective effects in an arthritis model [22]. Therefore, in patients with rheumatoid arthritis complicated by RAS-dependent hypertension, the blockade of the RAS might be beneficial not only to reduce blood pressure and vascular damage but also to prevent bone erosion.

In conclusion, this study provides novel insights into the pathophysiological function of Ang II in the regulation of inflammatory bone destruction. In patients with rheumatoid arthritis, the systemically activated RAS in concurrent pathological conditions could be involved in the progression of joint destruction in conjunction with increased local expression of AT1R. The effects of pharmacological inhibition of the Ang II-mediated pathway on bone erosion remain unclear but warrant further clinical examination.
