[Background and Aim]

Colorectal-cancer is one of the most common malignant neoplasms. Pathogenesis of colorectal cancer is thought to be induced by genetic alterations accumulated under chronic inflammatory conditions in the large bowel. Effective preventive or therapeutic modalities against colorectal cancer are currently unavailable. Novel approaches such as cancer immunotherapy that target tumor associated-antigens offer an attractive aternative to the conventional cancer therapies, with the expectation of fewer side effects, and better prevention of metastasis and recurrence, MUC1 is a tumor associated-antigen over-expressed on various human colon adenocarcinoma and their precursor lesions. Immune response to MUC1 was detected inpatient with pancreatic, breast, and ovarian tumors with the presence of antibodies and T cells specific for MUC1, and it was previously showed correlated to the better prognosis. Because of the highly increased expression and altered glycosylation in tumors, MUC1 is considered to be an effective target of cancer immunotherapy. Several studies aim to develop immunotherapy targeting on tumor associated antigens has been extensively conducted, We have previously reported that the growth of murine colon carcinoma SL4 cells-expressing MUC1 in the orthotropic site and lung metastasis of B16-F10 melanoma cells-expressing MUC1 were suppressed by vaccination with MUC1 DNA in C57BL/6 mice. However, those studies were conducted on transplantable models, which cannot represent the natural condition of tumor development. Therefore, it is very important to test the efficacy of MUC1 targeted immunotherapy in carcinogenesis model.

In this study, I employed MUC1 transgenic (MUC1. Tg) mice, which express endogenous MUC1 antigen, and have a similar level and pattern of MUC1 expression in cells and organs as seen in humans, to establish colitis-associated colorectal carcinogenesis model. As a MUC1 targeted therapy, I used MUC1 plasmid DNA (MUC1 DNA), which contained the full-length human MUC1 cDNA, combined with purified CD11c + bone-marrow derived dendritic cells (DCs) to increase its efficacy. The long-term goal of this study is to develop a MUC1 specific immunotherapy that targets tumor cells expressing MUC1, and thus preventing the development of colorectal tumor or inhibits its growth. The specific aims of this research are to establish colitis-associated colorectal carcinogenesis in MUC1. Tg mice, to evaluate the vaccination efficacy of MUC1 DNA combined with DCs (MUCl+DCs) in the prevention of colitis-associated colorectal cancer, and to elucidate its mechanism.

[Methods and Results]

1. Colitis-associated colorectal carcinogenesis model was established in MUC1. Tg mice.

1.1. Colitis-associated colorectal carcinogenesis in MUC1. Tg.

Azoxymethane-dextran sulfate sodium (AOM-DSS) colon carcinogenesis model is an experimental model to generate colitis-associated colon carcinogenesis in rodents. in this model, a single dose of AOM, a genotoxic carcinogen, will initiate tumor formation in large bowel of rodents, followed by periodical administrations of DSS which induces inflammation that promotes carcinogenesis. After extensive optimization, I set up the condition for AOM-DSS colorectal carcinogenesis protocol to obtain spontaneous MUC1-expressing colorectal tumor growth in MUC1. Tg mice. Periodical administrations of 1% DSS after a single dose of 10μg/kg body weight AOM was effective to induce colitis-associated colorectal carcinogenesis in MUC1. Tg (100%-tumor incidence). The number of tumor per mouse was 11.25±3.57 and tumor size was 3.50±3.54mm2 respectively. Macroscopically, nodular and polypoid colonic tumors were observed in the distal colon of mouse and these tumors were histopathologically tubular adenoma, which can be characterized with various degrees of dysplasia, and squamous metaplasia with atypia.

1.2. Expression of MUC1 glycoforms in tumors induced by AOM-DSS.

To characterize MUC1 expression in the tumor developed in AOM-DSS colon carcinogenesis model, I performed immunohistochemical analysis by using various anti-MUC1 antibodies, which recognize MUC1 in different glycoforms. I found that 100% of adenomas (n=36) were positively recognized by chicken anti-MUC1 cytoplasmic tail polyclonal antibody (anti-CTP) which binds to MUC1 cytoplasmic tail, independent of glycoforms;91.66% of adenomas showed strong binding to monoclonal antibody (mAb) MY. 1E12 which recognize sialylated MUC1;80% of adenomas were recognized by mAb HMFG1 which recognize glycosylated MUC1; and 91.18% showed binding to mAb HMFG2, which recognize poorly glycosylation MUC1. In the case of squamous metaplasia (n=10), all tumor tissues were recognized by all anti-MUC1 antibodies. Various MUC1 glycoforms are reported to be expressed in human colon adenomas and adenocarcinomas. The levels were correlated to the degree of malignancy and were useful in determining the prognosis of the disease. In the present study, I found that tumors developed in MUC1. Tg by AOM-DSS treatment also expressed various MUC1 glycoforms, similar to the MUC1 expression pattern in human. These results indicate that AOM-DSS colitis-associated colorectal carcinogenesis model in MUC1. Tg mice can be used as a suitable experimental model to evaluate the efficacy of immunotherapy targeting MUC1 expressing tumor.

2. Vaccination with MUCl+DCs suppressed colorectal carcinogenesis in MUC1. Tg mice.

2.1. Effect of MUC1 vaccination on tumor development.

To test the efficacy of MUC1 targeted immunotherapy, I next examined the preventive effect of MUC1 DNA vaccine in AOM-DSS colitis-associated colorectal carcinogenesis model. Mice were immunized three times at weekly intervals with 100μg DNA plasmid mixed with or without 1x105 DCs per mouse, followed by AOM-DSS administration. Vaccination with MUC1+DCs reduced tumor incidences in AOM-DSS treated MUC1. Tg mice to 37.5% reduction and decreased number of tumor per mouse in vaccinated MUC1. Tg mice compared to the treatment with control plasmid (pcDNA), MUC1 DNA plasmid (MUC1), DCs, or pcDNA combined with DCs (pcDNA+DCs). These results strongly indicate that combining MUC1 and DCs is critical to induce effective immunotherapy to inhibit tumor development in colitis-associated colorectal carcinogenesis model.

2.2. Effect of MUC1 vaccination in different types of tumors.

To further determine whether MUC1 targeted immunotherapy was effective against particular type of tumor, I characterized the tumor by histological examination into adenoma with low-grade dysplasia, adenoma with high-grade dysplasia and squamous metaplasia with atypia. The development of all types of tumors was suppressed in MUC1+DCs-vaccinated mice. Adenoma with low-grade dysplasia observed in MUC1+DCs-vaccinated mice, showed reduced levels of MUC1. However adenoma with high-grade dysplasia and squamous metaplasia with atypia that were rarely found in MUC1+DCs-vaccinated mice maintained the levels of MUC1 similar to control groups. These results show that vaccination with MUC1+DCs are effective regardless of the malignancy of tumor in AOM-DSS colon carcinogenesis model, and MUC1 expression were maintained even in tumors developed in MUC1+DCs-vaccinated mice.

3. Vaccination with MUC1+DCs induced MUC1-specific immune response.

3.1. Induction of antigen-specific antibody production after MUC1 vaccination.

To elucidate the mechanism of the MUC1+DCs vaccine effects against colon cancer development, I evaluated anti-MUC1 immune responses elicited by vaccination with MUC1+DCs.For the measurement of humoral immune responses, I determined MUCI-specific antibody production in the sera after MUC1 vaccination. There was a significant increase in MUC1-specific total Ig (IgG, A, and M) production in the sera by ELISA in MUC1+DCs-vaccinated group, compared to before vaccination, and vaccination with pcDNA, MUC1, and DCs alone.MUC1+DCs-vaccinated micesignificantlyhave higher MUC1-specific antibody production compared to pcDNA+DCs-vaccinated mice.

3.2. Induction of antigen-specific cellular response after MUC1 vaccination.

I next examined whether vaccination with MUC1+DCs induced cellular immune response against MUC1. MUC1-specific IFN-y secretion by splenocytes from MUC1+DCs-vaccinated mice was quantified by ELISPOT assays. MUC1+DCs-vaccinated mice showed significantly higher frequency of MUC1-specific IFN-y producing cells compared to control group (p=0.0067). These results strongly suggest that MUC1-specific IFN-y producing T cells are elicited in MUC1.Tg mice vaccination with MUC1+DCs. Collectively, these results indicate that MUC1-specific cellular immune responses are responsible for the inhibition of tumor development by vaccination with MUC1+DCs in AOM-DSS colitis-associated colorectal carcinogenesis.

[Conclusions]

In the present study, I applied carcinogen-initiated and inflammation-driven colon carcinogenesis to MUC1. Tg, and demonstrated for the first time that the tumors expressed MUC1. I clearly demonstrated that significant reduction of carcinogenesis was observed when DCs were combined with MUC1 DNA vaccine and administrated to these mice prior to tumor development. The suppression of tumor induction was likely to be due to MUC1-specific immune responses. These results implicate the potential of this novel immunotherapy approach as a novel therapy to prevent colorectal carcinoma growth in clinic.

This study is the first to report the efficacy MUC1 DNA combined with DCs in the prevention of colitis-associated colorectal carcinogenesis in MUC1. Tg mice. This vaccine strategy also highlight the beneficial contribution of DCs to enhanched antitumor immunity of MUC1 DNA vaccine.

This study provides supporting preclinical evidence on the efficacy of MUC1 DNA and DCs vaccination to prevent colitis-association colorectal carcinogenesis. These findings support additional investigation on combination of MUC1 DNA and DCs for the prevention of colorectal cancer in a human. The results of this study also support our previous reports on the suppression of the MUC1-transfected colon tumor growth and MUC1-transfected lung metastasis by using MUC1 DNA vaccine in wild type mice. Future study should be conducted to elucidate the detailed mechanism on how this vaccination break the tolerance, and how to increase the potency of this vaccine.

「Prevention of Experimental Colitis-associated Colorectal Carcinogenesis by Vaccination with MUC1 DNA(MUC1 DNAワクチンによる大腸発癌の抑制)」と題する本論文では、ヒトの癌抗原であるムチン1(MUC1)のトランスジェニックマウス(Tg)において、変異原である発がん物質と炎症惹起物質の投与で大腸に誘導された腫瘍組織にMUC1が発現すること、MUC1 DNAと樹状細胞を混合したワクチンにより発がんが抑制されること、このワクチンによりMUC1.TgにおいてMUC1特異的なTリンパ球が増殖及び活性化していること、の三つの重要な発見が述べられている。それぞれが、一つの章になっており2、3、及び4章をなし、その他に全体を通しての序論(1章)と結論(5章)が述べられている。

第1章の序論では、大腸がんとその治療について、MUC1という腫瘍抗原について、及び大腸発がんモデルについて、背景が述べられている。重要な点は、ワクチンによって発がんを抑制する研究は前例が少なく、再現性や定量性の高い結果が得られれば非常に価値が高いと予想されたことである。

第2章ではMUC1.Tgに発がん物質であるアゾキシメタンを腹腔内投与した後、デキストラン硫酸ナトリウム塩(DSS)を繰り返し経口投与して大腸に炎症を惹起し、それによって腫瘍を形成させる条件を最適化し、高い再現性を得たことが述べられている。最初のアゾキシメタン投与後69日目に見られた平均腫瘍数は11.25個、腫瘍径は3.5mmであった。ヒト組織ではMUC1は多くの器官の腺上皮に発現しており、がん化した腺上皮ではグリコフォームが異なることが知られている。また腸管では発現が見られないが、大腸がんでは高頻度で発現する。MUC1.TgにおけるMUC1の発現分布はヒトにおけるそれと一致するが、腸管に腫瘍を形成させてMUC1の発現と分布を検証した例は今までなかった。上記の最適化された条件で、MUC1.Tgに発癌物質と炎症惹起物質を投与することにより大腸に腫瘍を発生させMUC1の発現を検証した。遠位に多発する腫瘍は扁平上皮異型の形態を示し、100%の細胞に対して高い抗MUC1抗体の結合性が見られ、種々のグリコフォームに特異的な抗体がいずれも結合した。これに対して近位に多発した腫瘍は組織学的には腺腫様の形態を示し、抗MUC1抗体の結合は80-90%の細胞に対して結合性があったが、細胞により染色度は多様で濃淡があった。これらの結果より、この炎症を伴う発がんモデルでMUC1を標的とするワクチンによる発がん抑制が可能であると判断した。

第3章では、この炎症を伴う発がんモデルにおいて、あらかじめMUC1 DNAを含むプラスミドを骨髄由来樹状細胞と混合して投与した際の腫瘍の数と大きさについて調べた結果が述べられている。MUC1 DNAを含むプラスミド100μgと105の樹状細胞を投与すると、コントロールプラスミドと樹状細胞を投与した群に比して、腫瘍を持っ個体数が100%から62.5%に減少した。マウスあたりの腫瘍数及び腫瘍がある場合の腫瘍サイズにも有意な減少が見られた。他のコントロールグループとして、ワクチン否投与、コントロールプラスミド、MUC1 DNAのみ、樹状細胞のみも用いたが、いずれも効果が見られなかった。MUC1 DNAと樹状細胞のワクチンの投与を受けたマウスに腫瘍が発生したケースでは、抗原であるMUC1の発現が腫瘍細胞において低下、または消失したために拒絶されなかったという可能性を、これらの腫瘍におけるMUC1発現を比較することによって検証した。上記の様に扁平上皮異型の形態を示す腫瘍にはMUC1が高発現であり腺腫様の腫瘍ではMUC1の発現が多様でまだらであったが、何れのケースでも免疫した個体における腫瘍の形成は見られ、またMUC1発現レベルが変化したりMUC1を発現しない細胞数が増加するということはなかった。以上から、ワクチン投与を行った個体に発生したがんを形成しているがん細胞免疫回避バリアントではないことが明らかになった。

第4章では実際にMUC1特異的な免疫応答が起きているかどうかを免疫学的な方法で検証した結果が述べられている。先ず、DNAワクチンの投与により炎症が抑制された、すなわち自然免疫が抑制されたことによって腫瘍形成が阻害されたとの可能性をについて可否を問うため、上記と同じ条件でプラスミドの投与を受けたマウスにDSSのみを投与し、腸管における炎症の重篤度を比較した。その結果MUC1 DNAと樹状細胞のワクチンを投与されたマウスにおいて影響が見られないことを確認した。ワクチンを投与したマウス中のMUC1に結合活性を持つ免疫グロブリン量を定量すると、結果MUC1 DNAと樹状細胞を投与したグループではその他のグループに比して有意に高レベルが検出された。また、エリスポットアッセイによりMUC1に応答してIFN-gを産生する細胞数が有意に増加していた。従って、MUC1特異的な免疫応答によって、変異原である発がん物質と炎症惹起物質の投与でMUC1.Tgの大腸に誘導された腫瘍細胞が拒絶されたことが確実と考えられた。

以上の様に本研究では、がん患者において免疫応答が見られることがありその事がより良い予後と相関することが知られているがん抗原であるMUClを、ヒトMUC1プロモーターによって発現するトランスジェニックマウスにおいてワクチンによる発がんの抑制に成功した。用いたワクチンはMUC1 DNAと骨髄由来樹状細胞を混合したもので、類似のものをヒト用に容易に作成できる。また、ヒトにおいてMUC1が正常上皮には発現しないががん組織には発現することが知られている臓器である大腸において、変異原である発がん物質と炎症惹起物質の投与種類の異なる複数の腫瘍が発生し、これらの細胞においてMUC1が発現することを初めて確認した。これらの結果はいずれも極めて新規性が高くまた前臨床研究として高いインパクトを持つものである。研究内容は、腫瘍学、免疫学、がん治療学、糖鎖生物学などに大きく寄与し、本研究を行ったムルワンティ,レトノは博士(薬学)の学位を得るにふさわしいと判断した。