Canine osteosarcoma (OS) is an aggressive primary bone tumor, accounting for up to 85% of malignancies originating from the skeleton. Despite of advances in management modalities for canine OS, prognosis remains poor because almost 90% of dogs develop pulmonary metastasis which is the main cause of death.

Xenograft models of canine OS cells injected into immunocompromised mice are necessary for greater understanding of the biology of metastasis to improve the outcome for canine OS patients. These models can demonstrate the interaction between cancer cells and the surrounding microenvironment that is necessary for metastasis. Moreover, they also provide opportunities to evaluate novel therapeutics.

Ezrin (cytovillin/p81/80k/Villin-2), a member of the ezrin-radixin-moesin (ERM) proteins family, was identified as a key molecule during the onset and progression of the metastatic cascade in a murine OS model. Phosphorylated ezrin, an active form, was dynamically regulated in the process of murine and human OS metastasis by protein kinase C (PKC). Ezrin activation allows the tumor cells to interact with tumor microenvironment through the link between the transmembrane receptor and the actin cytoskeleton. This process is essential for many fundamental cellular processes and integration of membrane transport with signaling pathways, such as MAPK-signaling pathway. Recently, most studies report that ezrin is associated with tumor progression and metastasis in human and murine OS. Moreover, the mechanism of ezrin activity with tumor cellular processes during metastatic cascade is still unclear.

The objective of this study was to determine the relationships between malignant behaviors and ezrin activity in canine OS.

In chapter 1, expression of ezrin and phosphorylated ERM (p-ERM) was investigated in primary tumor tissues of canine OS patients. Immunohistochemistry for ezrin and p-ERM was performed on primary tumor tissues of 15 canine OS patients surgically collected at Veterinary Medical Center of the University of Tokyo between 2006 and 2010. Correlations of the expression of ezrin and p-ERM with the clinical data and proliferation index (PI) on immunohistochemistry of Ki-67 were evaluated.

The result demonstrated that ezrin and p-ERM expression was found in all primary tissues of canine OS. High expression of ezrin and p-ERM was observed in 13 of 15 patients. The expression of ezrin was significantly correlated with p-ERM (r=0.9932, p=0.0000), but not with proliferation index (r=0.2640, p=0.3417). There were no significant correlations between high expression of ezrin and p-ERM and age, gender, breed, body weight, primary location site, histological type, serum ALP, and lung metastasis. The survival curves of patients with high and low ezrin, p-ERM, and Ki-67 expressions did not show significant differences; however, the survival curve of mice with high expression of ezrin and p-ERM tended to be worse than those with the low expression. The results suggested that expression of ezrin and p-ERM may be associated with the malignant behavior of canine OS.

In chapter 2, the effects of microenvironment at transplantation sites on the growth and metastasis in a canine OS cells-xenografted mouse model were analyzed.

In section 1, to investigate the expression of ezrin and p-ERM on 3 canine OS cell lines (HMPOS, OOS, and CHOS), Western blot analysis and immunocytochemistry were performed. Ezrin was expressed on all these cell lines, however p-ERM was not expressed on CHOS cell line by both Western blot analysis and immunocytochemistry.

Those 3 OS cell lines were transplanted via subcutaneous (SC), intratibial (IT), and intravenous (IV) injection into 5-week-old female BALB/c nude mice with 5 mice per group. The primary tumor volumes, number of lung metastatic nodules, and expression of ezrin and p-ERM were evaluated on immunohistochemistry of primary tumors and lung metastatic tissues at 1, 2, and 4 weeks after transplantation.

IT xenografts exhibited greater potential for developing primary masses and pulmonary metastasis than SC xenografts. The expression of ezrin and p-ERM in the primary tumors of IT-xenografted mice was significantly higher (p<0.05) than those in SC-xenografted mice with HMPOS and OOS cells. In IT and IV xenografts, lung micrometastases along with p-ERM overexpression were found in mice xenografted with HMPOS and OOS cells after 1 week. But decreased p-ERM expression was found at the later time points in these mice. The results suggested that the orthotopic transplantation site was associated with the primary tumor growth and the spontaneous metastasis of canine OS. In addition, ezrin phosphorylation may be involved in the early metastatic mechanism of canine OS cells.

In section 2, relationships between ezrin and p-ERM expression and PI, Ras/Raf/ERK MAPK pathway, and PKCα in tissues developed in xenografted mice were evaluated.

The expressions of molecules in Ras/Raf/ERK MAPK pathway (Ras, B-Raf, C-Raf, p-ERK1/2, and ERK1/2) and PKCα were investigated in OS cell lines on immunocytochemistry and Western blot analysis. In addition, Ki-67, Ras, B-Raf, C-Raf, p-ERK1/2 and PKCα in xenograft mouse tissues were investigated on immunohistochemistry.

High metastatic HMPOS and OOS cells expressed high B-Raf and C-Raf but low p-ERK1/2 when compared to non-metastatic CHOS cells. The PI, Ras, C-Raf, and p-ERK1/2 in primary tissues of IT-xenografted mice increased in a time-dependent manner and were significantly higher (p<0.05) than those of SC- xenografted mice in the later progression time after transplantation, relating to tumor growth. These results suggested that ezrin in phosphorylated form may be involved in a short time at the beginning step for signal transduction of MAPK pathway and then dephosphorylated form may be involved in the later time for proliferation of OS.

The expression of PKCα in primary tissues of IT-xenografted mice was significantly higher (p<0.05) than that of SC-xenografted mice at the earlier time of progression and was significantly correlated with expression of ezrin and p-ERM. Moreover, expression of p-ERM and PKCα was low in the central portions of the primary and metastatic lesions in progression time at 4 weeks after transplantation. This result suggested that activated ezrin may be associated with PKCα.

In chapter 3, the effect of suppression of p-ERM in canine OS cells by a PKC inhibitor was investigated in both in vitro and in vivo orthotopic xenografted mouse model.

Chelerythrine (CHE) was used as a PKC inhibitor. Cell viability MTT assay was performed for HMPOS, OOS, and CHOS cell lines treated with CHE at 0.5, 1, 1.25, 2.5, 5, 10 and 20 μM for 24 hours. Cell motility using wound-healing, transwell, and xCELLigence real time migration assays as well as cell invasion using matrigel invasion and xCELLigence real time invasion assays were also performed for those cells treated with the dosage of CHE without affecting cell viability. The expression of PKCα, ezrin, p-ERM, Ras, B-Raf, C-Raf, p-ERK1/2 and ERK1/2 was investigated in HMPOS cells by Western blot analysis.

The IC50 values of CHE were 7.42, 7.10, and 7.56 μM for HMPOS, OOS, and CHOS cell lines, respectively. CHE had the anti-migratory and anti-invasive effects in all cell lines exposed to CHE at 1 and 5 μM, respectively. HMPOS cells exposed to CHE showed suppression of PKCα and p-ERM with increased p-ERK1/2 in a time-dependent manner.

In in vivo experiment, IT-xenografted mouse model with HMPOS cell line was used. Mice were divided into 3 groups (n = 5 mice/group). The control group received no treatments. The treatment 1 group received 5 mg/kg CHE i.p. at day 0, 2, and 4 after transplantation. The treatment 2 group received 5 mg/kg CHE i.p. at day 8, 10, and 12 after transplantation. At 1, 2, and 4 weeks after transplantation, the primary tumor volume and lung metastatic nodules were measured, and immunohistochemistry was performed for ezrin, p-ERM, p-ERK1/2, and PKCα in primary tumors and lung metastatic tissues.

Lung metastasis was not detected in the treatment 1 group at 1 week after transplantation. At 4 weeks after transplantation, number of metastatic nodules of mice in the control group was significantly higher (p<0.05) than those of treatment 1 and 2 groups. However, the tumor volumes were not significantly different between the control and treatment groups. The expressions of p-ERM and PKCα in primary and lung metastatic lesions at 2 weeks after transplantation in the treatment 1 and 2 groups were significantly lower (p<0.05) than that in the control group. These results suggested that decreased lung metastatic potential was associated with the regulation effect of p-ERM by PKC through suppression of migration and invasion of canine OS cells.

In conclusion, ezrin and p-ERM may play an important role in the proliferation and the earlier phase of lung metastasis in canine OS. These functions may be due to increase in motility and invasion of OS cells. The high expressions of ezrin and p-ERM may suggest shorter survival time in OS patients. In addition, a PKC inhibitor may suppress their expressions and decrease the incidence of lung metastasis in canine OS.

犬骨肉腫は悪性度の高い骨原発腫瘍であり、約90%の症例は肺転移によって死亡する。本腫瘍の転移機構に関しては様々な研究がなされているが、その詳細は不明である。

エズリン(cytovillin/p81/80k/Villin-2)は、ezrin-radixin-moesin (ERM)蛋白ファミリーに属し、マウス骨肉腫では、その進行や転移の鍵になる分子として知られている。エズリンはリン酸化されて活性化され、また、protein kinase C (PKC) により転移に関わる機能が調整されている。しかし、犬の骨肉腫に関する報告はほとんどない。本研究の目的は、犬の骨肉腫における悪性挙動とエズリンの関連を追及することにある。

第1章では、犬の骨肉腫症例の腫瘍組織におけるエズリンとそのリン酸化分子(リン酸化ERMとして測定:p-ERM)の発現を検討した。すなわち、15頭の犬骨肉腫組織に対し免疫染色を行い、これらの発現と症例の臨床データとの相関を検討した。

その結果、すべての組織でエズリンとp-ERMの発現が認められ、そのうち13例では強く発現していた。エズリンの発現はp-ERMの発現と有意に相関した。腫瘍の組織型などの臨床データとエズリン、p-ERMの発現は相関しなかったが、生存期間については負の相関を示した。このことから、これらの発現と犬骨肉腫の悪性挙動との間には何らかの関連があると示唆された。

第2章では、ヌードマウスモデルを用い、犬骨肉腫細胞株を同所性(脛骨)あるいは異所性(皮下)に移植し、その後の腫瘍発育や転移とエズリン、p-ERMの発現との関連について検討した。

セクション1では、当研究室で樹立した3骨肉腫細胞株(HMPOS, OOS, CHOS)における発現について免疫細胞化学ならびにウェスタンブロットによって検討した。その結果、エズリンはいずれの細胞株にも発現していたが、p-ERMはHMPOS, OOSに発現が見られるものの、比較的悪性度の低いCHOSには発現しなかった。ついで、これらの細胞株をヌードマウスの脛骨あるいは皮下に移植した。

脛骨移植(IT)群では、皮下移植(SC)群より原発巣の発育が早く、転移も多かった。また、IT群の原発組織におけるエズリンとp-ERMの発現は、SC群より有意に高かった。IT群では、HMPOSとOOSで肺転移が見られ、これらの組織では、移植後1週目でp-ERMの過剰発現が認められた。しかし、移植後時間の経過とともに、p-ERMの発現は低下した。これらの結果から、同所性移植ではより早く原発巣の増殖が見られ、早期に肺転移の生じることが確認された。さらに、エズリンのリン酸化は、転移の初期の機序に関わっていることが示唆された。

セクション2では、これらの細胞株ならびにヌードマウス移植モデルにおけるエズリン、p-ERMと、Ras/Raf/ERK MAPK経路の各分子ならびにPKCαの発現を免疫組織化学ならびにウェスタンブロットで検討した。

IT群の原発組織においては、PI, Ras, C-Rafおよびp-ERK1/2の発現が経時的に増加し、4週後ではSC群より有意に高かった。これらの結果は、リン酸化されたエズリンは、MAPK 経路のシグナル伝達の初期段階に関与し、脱リン酸化されたエズリンは骨肉腫の増殖の後期に関与することを示唆するものと思われた。一方、IT群の原発巣におけるPKCα発現は、初期にはSC群より有意に高く、かつエズリンおよびp-ERMの発現と有意に相関していた。

第3章では、in vitroおよび同所性移植モデルを用い、PKC阻害剤によるp-ERMの抑制効果を検討した。阻害剤としては、chelerythrine(CHE)を用いた。

まず、3種の細胞株の細胞運動能と細胞浸潤能を測定した。同時に、HMPOS細胞に関しては、CHE処理後のPKCα、エズリン、p-ERM, Ras, B-Raf, C-Raf, p-ERK1/2, ERK1/2の発現をウェスタンブロットで測定した。

その結果、CHEはこれらの細胞株に毒性を示さない用量で運動能、浸潤能を抑制し、同時に、PKCα、p-ERMの発現を経時的に抑制した。次に、HMPOS細胞をIT 移植し、これらのマウスにCHEを投与した。その結果、CHE処理マウスでは無処置のマウスに比較し、原発巣の腫瘍体積に有意な差がなかったものの、肺転移結節数が有意に低下した。

以上から、エズリンおよびp-ERMは犬の骨肉腫において、腫瘍の転移初期に大きな役割を果たしていること、この作用には腫瘍細胞の運動能、浸潤能を介している可能性が示唆された。また、これらの発現する骨肉腫症例では、予後が短い可能性があり、さらに、PKC阻害剤によるこれらの発現の抑制により、少なくとも肺転移が抑制される可能性が示された。

以上要するに、本研究は犬骨肉腫の転移機構におけるエズリンの役割の一端を証明したものであり、学術上臨床上その貢献するところは少なくない。よって審査委員一同は本論文が博士(獣医学)の論文として価値あるものと認めた。