1.Introduction:

Osteoarthritis (OA) is characterized by a progressive degradation of articular cartilage leading to loss of joint function. The molecular mechanisms regulating pathogenesis and progression of OA are poorly understood. Remarkably, chondrocyte differentiation process, addressed as chondrocytogenesis, during skeletal development by endochondral ossification is involved in the progression of OA. Articular chondrocytes in degradated cartilage exhibit progressively differentiation. Thus, regulation factors engaged in chondrocyte differentiation in both growth cartilage and permanent articular cartilage implicated in OA may be possible targets for novel disease-modifying therapies for OA.

Chondrocytogenesis is the process by which cartilage is formed from condensed mesenchymal cells at growth plate, the ends of long bone (epiphyseal plate) and distal end of pubis (apophyseal plate), where longitudinal bone growth occurs. Therefore, this process can determine the length and shape of bone morphology. The process starts the commitment from mesenchymal stem cells (MSCs) to subsequent proliferation of chondrocytes, and then differentiates into hypertrophic chondrocytes (1). It has been documented that various signaling pathways, growth factors, transcription factors (i.e.Sox9) and hormones (i.e. sex steroids) are involved in the regulation of chondrocytogenesis (2).

During puberty, it is well known that the sex differences in bone length as well as bone shape are apparently remarkable. From these physiological evidences, it can be hypothesized that sex steroids, especially androgens, might play a role in chondrocytogenesis. Most of androgens exert their functions as ligands specifically through the androgen receptor (AR), a member of nuclear receptor superfamily. AR is expressed in chondrocytes of rodents and humans. Hence, AR may perform important function on determination of bone morphology via direct effect on growth plate. Though, no significant difference in longitudinal growth of long bone between the conventional AR knock-out (ARKO) mice and WT littermates is observed, the ARKO mice suffer from endocrine disturbance, inferring that the phenotype on long bone may be not specifically according to androgen-AR signaling (3, 4). Thus, the chondrocyte-specific ARKO mice are required to determine whether AR takes the direct biological function in chondrocytes.

Meanwhile, so far, the researches of chondrocytogenesis-related factors have merely had access to the tip of an iceberg, which unable to clarify the chondrocyte differentiation- involved pathological mechanisms even the regulation mechanisms thoroughly. Lots of novel factors, especially transcriptional and epigenetic regulators should be considered particularly noteworthy.

Therefore, the purpose of this study is clarification of AR function together with identification of the novel transcription and/or epigenetic factors in chondrocytogenesis and determination their underlying molecular mechanisms in chondrocytes in vitro and in vivo.

2.Results:

1)Determination of the function of androgen receptor in chondrocytogenesis

By employing the Cre/loxP system, the male proliferative chondrocyte-specific ARKO mice (Col2a1-ARKO) and hypertrophic chondrocyte-specific ARKO mice (Col10a1-ARKO) were generated to uncover the biological function of AR in chondrocytogenesis. Unexpectedly, not only in cartilage but actually in testis, the deletion of AR gene was partially observed in both mice lines. According with the analyses associated with testis, the partial AR gene ablation in testis was unlikely to reflect on endocrine system. Thus, the generated chondrocyte specific ARKO mice were tolerating models for morphological analysis of skeletal tissues. In both chondrocyte-specific ARKO mice, there was no significant difference in long bone length and bone mass. However, the expanded proliferating zone with consequence of markedly increased pubis length and pelvic size were exhibited in Col2a1-ARKO, not in Col10a1-ARKO mice. This result suggested that AR may suppress the proliferation of chondrocytes in pubic bone not in long bone.

To determine the molecular mechanism of AR in chondrocytogenesis, chondrocyte differentiation model in vitro should be established. In micro mass culture system using C3H10T1/2 (10T1/2) cell line treated with BMP2, the mass cultured 10T1/2 cells differentiated into proliferative chondrocytes at day 5 followed by the differentiation into hypertrophic chondrocytes at day 9. In the chondrogenic 10T1/2 cells, AR was highly expressed in proliferative chondrocytes on both mRNA and protein levels. Moreover, the mRNA expression of proliferative chondrocyte-related genes was down-regulated by an AR ligand, dihydrotestosterone (DHT). These results suggested that liganded AR might be served as a novel negative regulator in early stage of chondrocytogenesis. According to the distribution and the function of AR in chondrocytogenesis, the early stage chondrogenic ATDC5 cell line was utilized to determine the mechanism by which AR negatively regulated early chondrocytogenesis. Interestingly, the result of luciferase assay with Col2a1 luciferase reporter containing Sox9 binding site, revealed that AR repressed transcriptional activity of Sox9 in a ligand-dependent manner. Furthermore, AR recruitment was detected in Sox9 binding motif in Col2a1 enhancer region under the DHT treatment. Meanwhile, acetylation on histone H3 and methylation on H3K4 residues were decreased at the same region in the presence of DHT via the analysis of ChIP-qPCR.

Together, these results suggested that liganded AR may suppress the transcriptional activity of Sox9, subsequently, down-regulate the expression of Col2a1 via alteration of histone modifications to repress the early chondrocytogenesis. This androgen-AR dependent mechanism might play a role in the determination of pelvic bone morphology.

2)Identification of novel transcriptional and/or epigenetic regulators in chondrocytogenesis

To identify novel transcriptional and/or epigenetic regulators in chondrocytogenesis, differentiation stage dependent gene expression profiles of the mass cultured 10T1/2 cells were analyzed by gene expression microarray. From the results of the differentially expressed candidates among differentiation stages followed by validation through qRT-PCR, autoimmune regulator (Aire), a transcription factor, was targeted, which mutations cause the autoimmune polyendocrinopathy- candidiasis-ectodermal dystrophy (APECED) accompanying with reversible metaphyseal dysplasia.

As a result of qRT-PCR, mRNA expression levels of chondrocyte-related genes were decreased in Aire stable knockdown mass cultured 10T1/2 cells and Aire-/- primary cultured chondrocytes. These data suggested that Aire might work as a novel positive regulator in chondrocytogenesis. To further clarify the function of Aire, bone morphogenetic protein 2 (BMP2) was focused, which is a member of the transforming growth factor β superfamily and can facilitate chondrocytogenesis via BMP2/Smads signaling pathway, as a possible target gene of Aire, because the expression of BMP2 is down-regulated in Aire-/- thymic epithelial cells (TECs). Interestingly, luciferase assay with BMP2 promoter suggested that Aire can directly up-regulated BMP2 expression in ATDC5 cells. Also, the early stage chondrocytogenesis in mass cultured 10T1/2 cells was induced by Aire overexpression even without BMP2 treatment. Furthermore, ChIP-qPCR assay was performed to define the molecular mechanism of Aire on BMP2 expression. As a result, Aire was recruited to Aire binding motif (T box, TTATTA) in BMP2 promoter in early stage of chondrocytogenesis, with increased H3K4me2, an active histone marker for gene expression. This modification was partially reduced with the inactivation of Aire. These data indicated that Aire may be a novel transcription factor in chondrocytogenesis via up-regulating BMP2 expression through the alteration of histone modification to accelerate the chondrocyte differentiation in early stage.

3.Conclusion:

In this study, we determined the physiological function of AR via chondrocyte-specific androgen receptor knock-out mice model and the molecular mechanism of AR in differentiated mass cultured 10T1/2 cell line or early chondrogenic ATDC5 cell line. Meanwhile, based on the analysis of microarray of gene expression during the chondrocytogenesis, we also identified Aire, which may be a novel transcription factor in chondrocytogenesis.

From the results of chapter 2 and 3, it had been observed that the increased chondrocyte proliferation in Col2a1-ARKO mice in vivo as well as the decreased expression of proliferative chondrocyte-related genes in DHT-dependent manner in vitro. It indicated that liganded AR may be regarded as a suppressor of early chondrocytogenesis, which was consistence with the smaller pelvis size in male than in female, however, contrary to that long bone is longer in male comparing with in female. It may be contributed to the different construction of two growth plates resulting in the increased expression of aromatase in long bone not in pubic bone. Hence, in male mice, most of androgens may be aromatized into estrogens, which stimulate the growth spurt to cover the negative regulation of AR via less of androgens on long bone growth.

Meanwhile, results from chapter 4 indicated that Aire prompted the early stage of chondrocyte differentiation by up-regulating BMP2. Aire may be recruited to Aire binding motif on BMP2 promoter via the recognition of hypomethylated H3K4 by PHD1 domain of Aire and associate with histone methyltransferases for methylation on H3K4 or co-regulators to activate the expression of BMP2. However, it has been published that the chondrocyte-specific BMP2 knock-out mice exhibit a severe chondrodysplasia phenotype, whereas, no significant phenotype is observed in bone in the reported conventional Aire knock-out mice. The possible reasons of this inconsistency are considered as follows: 1) conventional Aire knock-out mice suffering from immune deficiency may exhibit a growth factors-related disorder, such as IGF-1. 2) Aire may be engaged in transcriptional regulation of BMP2 as one of the mechanisms, which can be compensated by other factors.

Therefore, considering the temporal and spatial-dependent function of AR, other stages of chondrocyte-specific ARKO mice should be required to elucidate the AR biological function in long bone growth. Furthermore, in vitro, the specific AR-associated complexes involving HDACs/HDMs should be identified to further understand the mechanism of AR in early stage of chondrocytogenesis. Besides, generation of the chondrocyte-specific Aire knock-out mice and analyzed in early stage should be required to clarify the physiological function of Aire in bone growth in vivo. Further, to deep draw out the molecular mechanism of Aire, the PHD1 domain and Aire-involved complexes in early chondrocytogenesis should be a key point to be clarified.

Together, the activity of proliferative chondrocytes was suppressed by AR and activated by Aire. By the understanding of physiological and molecular function of them, it can bring us incredible wide insight in the fields of chondrocytogenesis regulation as well as the treatment of OA in the future.

本研究は、哺乳類の骨格形成およびヒトにおける関節疾患の発症や進展に深く関与している軟骨細胞分化を制御しうる転写関連因子群の役割の解明および新規軟骨細胞分化関連因子の同定および解析に取り組んだもので、論文は序論およびそれに続く3章からなる本論、総合討論で構成されている。

序論では、研究の背景と目的について述べた後、第2章では、軟骨特異的アンドロゲン受容体欠損マウスの作出および解析を行っている。哺乳類の性徴期には、長管骨の長さや骨盤の形態などの様々な性差が存在する。このような性差は、一般に性ホルモン作用により生じるとされている。男性ホルモン(アンドロゲン)および女性ホルモン(エストロゲン)は標的組織におけるアンドロゲン受容体(AR)およびエストロゲン受容体(ER)を介した標的遺伝子の発現制御により、その生理作用を発揮する。これまでにERは雌性型骨長において機能することが明らかとなっており、ARは雄性型骨形態の構築および軟骨細胞分化過程への関与が示唆されている。しかしながら、現在までの全身AR遺伝子欠損(ARKO)マウスの解析では、内分泌異常を伴う事で雄性型骨形態への関与が示されなかったことから、ARの軟骨細胞分化過程への直接作用は証明することができなかった。軟骨細胞分化過程は、間葉系幹細胞から、静止軟骨細胞、増殖軟骨細胞 、肥大軟骨細胞へと分化し、最終的に、アポトーシスへと誘導されるものである。この三つの軟骨細胞層は成長期の骨端成長板を構成することおよび軟骨分化過程が骨形態を決定することが知られている。そこで、本研究では軟骨細胞特異的ARKOの作出により、ARの軟骨細胞を介した骨長および骨盤型への直接作用の検証を試みた。Cre/loxPシステムによる増殖軟骨細胞特異的なARKO マウス(Col2a1-ARKO)および肥大軟骨細胞特異的なARKO マウス(Col10a1-ARKO)を作出した。これらの軟骨特異的なARKOマウスを用いて、雄性型骨形態、特に、長管骨や恥骨の骨長および恥骨の骨長決定した骨盤型の解析を行った。その結果、Col2a1-ARKOやCol10a1-ARKOマウスの長管骨およびCol10a1-ARKOマウスの恥骨や骨盤型は野生型と有意な差異は見出されなかった。しかしながら、Col2a1-ARKOマウスでは、有意な恥骨長の延長と骨盤腔の拡大を認め、雄マウスにおいて雌性型骨盤が観察された。恥骨軟骨細胞層を詳細に解析したところ、成長板全層および増殖軟骨層の厚さが野生型と比べて、Col2a1-ARKOマウスでは有意に増加していた。このことから、ARが軟骨細胞増殖を抑制することが示唆された。

第3章では、軟骨細胞分化細胞系を用いて、軟骨組織分化におけるAR分子機構を明らかにした。軟骨細胞、特にARKOマウスの解析により、成長板軟骨層厚さの増加は軟骨細胞の分化異常に起因すると考えられた。そこで、軟骨組織および軟骨細胞分化系を用いた結果、ARリガンド投与により、増殖軟骨分化マーカー遺伝子Col2a1および軟骨細胞増殖制御転写因子Sox9の発現上昇を認めた。また、ARがSox9に対する転写共役因子である可能性をレポーターアッセイにより検討した。その結果、ARがリガンド依存的にSox9転写能を抑制することを見出した。さらに、Sox9の標的遺伝子であるCol2a1のプロモーター領域におけるChIPアッセイの結果により、ARがリガンド依存的にリクルートされることを見出した。また、Col2a1の発現変化がプロモーター領域のヒストン修飾の変化に起因するものと考え、ChIPアッセイを行った。その結果、ARでは転写活性化のマークであるH3K4me3およびH3Acの修飾がリガンド依存的に抑制されていることが明らかになった。以上の結果より、ARはSox9と協調的に作用し、ヒストン修飾変動を介して、Sox9の転写能を減弱させ、その標的遺伝子であるCol2a1の発現を抑制するという分子機構を明らかにした。

第4章では、軟骨細胞分化系を用いて、マイクロアレイの結果により、Aireを新規軟骨細胞分化制御因子として同定、解析した。青年期のヒトで、Aire遺伝子の変異が可逆骨幹端異形成症(RMD)という病気に関連する報告がある。そこで、軟骨初代培養細胞および軟骨細胞分化系を用いて、Aireの軟骨細胞分化における関与を検討した。その結果、Aireは軟骨細胞分化を促進することが明らかとなった。さらに、Aireが軟骨細胞分化を促進するサイトカインであるBMP2の発現を促進することを確認した。そこで、BMP2のプロモーター領域を用いたレポーターアッセイおよびChIPアッセイを行ったところ、AireがBMP2のプロモーターにリクルートされて、直接的にBMP2の発現を促進することが示唆された。同時に、同領域においてH3K4me2修飾がAire依存的に増加することも見出した。以上の結果より、Aireは軟骨細胞分化過程における活性化因子として、ヒストン修飾変動を介して、直接的にBMP2の発現を促進するという分子機構を明らかにした。

続く総合討論では本研究を総括および考察し、今後の課題ならびに研究の方向性について述べている。

以上本論文は、骨盤形態形成におけるARの生理機能および軟骨分化に対するARの転写制御機構の解明、ならびに新規軟骨細胞分化関連因子としてAireの同定およびその機能解析を行ったもので、骨軟骨代謝領域においても研究の発展性が期待され、学術的に貢献するところが少なくない。よって、審査委員一同は、本論文が博士(農学)の学位論文として価値あるものと認めた。