Background:

Myocardial infarction (MI) is the most common cause of cardiac morbidity and mortality, and left ventricular remodeling after MI is still a serious problem even in the era in which emergent cardiac catheterization is available world-wide. Ventricular remodeling is linked to heart failure progression and is associated with poor prognosis following MI. Accordingly, it is of critical importance to develop therapeutic strategies that will effectively inhibit the development of ventricular remodeling and failure after MI.

BCRP1/ABCG2 is a member of the ATP-binding cassette transporter family originally identified by its ability to confer drug resistance in tumor cells by active efflux of multiple drugs. This protein has been shown to be expressed in various normal organs, and has been suggested to regulate several tissue defense mechanisms. Previous studies have suggested that BCRP1/ABCG2 may have functions in tissue defense in the diseased heart. However, the physiological significance of its expression in cardiac injury has not yet been fully elucidated. This study was performed to clarify the impact of BCRP1/ABCG2 expression on cardiac repair after MI.

Methods and Results:

First, I confirmed the expression of BCRP1/ABCG2 mRNA in murine heart by reverse transcription polymerase chain reaction (PCR). Immunohistochemistry showed that BCRP1/ABCG2 was mainly expressed in endothelial cells of microvessels in the heart.

To clarify the impact of BCRP1/ABCG2 expression on cardiac repair after MI, I induced MI in 8- to 12-week-old wild-type (WT) and Bcrp1/Abcg2 knock-out (KO) mice by ligating the left anterior descending artery. The survival rate up to 28 days after MI was significantly lower in KO mice than in WT mice (KO 28.3% versus WT 74.5%, p<0.0001) mainly due to cardiac rupture in 4 to 6 days after MI. Echocardiography showed that dilatation of the left ventricle (LV), thickening of the non-infarcted area, and ejection fraction were worse in KO mice than in WT mice at 28 days after MI. Hemodynamic measurements by the micro pressure transducers through the right carotid artery demonstrated that LV function was more deteriorated in KO mice than those in WT mice at 28 days after MI. Heart weight to body weight ratio was greater in KO mice than in WT mice, and infarct size at 28 days after MI, assessed by sirius red staining, was significantly larger in KO mice than in WT mice, although initial area at risk and initial infarct size did not differ between the 2 groups. Myocyte cross-sectional area (CSA) and collagen volume fraction (CVF) in the non-infarcted area at 28 days after MI, assessed by hematoxilin-eosin (H&E) and sirius red staining, respectively, were greater in KO mice than in WT mice. These results indicated that ventricular remodeling after MI was exaggerated in KO mice compared with WT mice.

As angiogenesis and recruitment of macrophages and myofibroblats play an important role in would healing process, capillary, macrophages and myofibroblasts density in the peri-infarction area were assessed by immunohistochemistry with anti-CD31 antibody, anti-Mac-3 antibody and anti-α-smooth muscle actin antibody, respectively. At 5 days after MI, capillary, macrophages and myofibroblasts density in the peri-infarction area were significantly reduced in KO mice compared with WT mice. I also assessed cytokine mRNA expression in the peri-infarction area by real time PCR. At 5 days after MI, gene expression of pro-inflammatory, angiogenesis-related and fibrosis-related cytokines was comparable or higher in KO mice than in WT mice, whereas cytokine mRNA expression levels at baseline were comparable between WT and KO mice.

To assess the impact of BCRP1/ABCG2 expression in microvascular endothelial cells of the heart, in vitro experiments with human microvascular endothelial cells from the heart (HMVEC-Cs) were performed. MTS assay showed that pharmacological inhibition of BCRP1/ABCG2 with fumitremorgin C (FTC) resulted in impaired survival of human microvascular endothelial cells from the heart under oxidative stress, although BCRP1/ABCG2 inhibition did not alter the expression pattern of ICAM-1, VCAM-1 and eNOS even under oxidative stress. Flow cytometry analysis demonstrated that protoporphyrin IX concentration was higher in HMVEC-Cs with FTC than in those without FTC, which might exaggerate oxidative stress in HMVEC-Cs.

Discussion:

In the present study, I found that BCRP1/ABCG2 is expressed mainly in the endothelial cells of microvessels in the heart. I also demonstrated that genetic disruption of BCRP1/ABCG2 deteriorated mortality and cardiac remodeling after MI. In KO mice, angiogenesis and recruitment of macrophages and myofibroblasts were impaired. In vitro experiments showed that BCRP1/ABCG2 played an important role in survival of microvascular endothelial cells of the heart under oxidative stress possibly by preventing from accumulation of intra-cellular protoporphyrin IX.

In this study, I found that BCRP1/ABCG2 is essential for microvascular endothelial cell survival under oxidative stress, which is important for angiogenesis in damaged tissues. This result may explain why angiogenesis was impaired in KO mice, although the expression levels of angiogenesis-related cytokines were higher in KO mice. Previous studies suggested the mechanisms by which BCRP1/ABCG2 expression protect cell death. BCRP1/ABCG2 has been demonstrated to efflux protoporphyrin IX. The regulation of porphyrins and heme within a cell is important because the accumulation of heme within cell can ultimately lead to the accumulation of iron and the production of cell-damaging reactive oxygen species by the Fenton reaction. This may explain why an inhibition of BCRP1/ABCG2 lead to impaired survival of microvascular endothelial cells under oxidative stress, but not under normal condition.

Impaired angiogenesis might lead to delayed cardiac repair after MI. I also found that macrophage recruitment was impaired in KO mice at 5 days after MI, although the expression levels of pro-inflammatory cytokines were comparable or higher in KO mice than in WT mice. The previous study showed that loss of BCRP1/ABCG2 does not affect hematopoiesis. In addition, my in vitro experiments showed that BCRP1/ABCG2 had no effect on expression of adhesion molecules such as ICAM-1 and VCAM-1 in microvascular endothelial cells. Reduced capillary density, therefore, might lead to the impaired macrophage recruitment from blood stream. As macrophages play an important role in clearance of necrotic cardiomyocytes, impaired macrophage recruitment might lead to impaired cardiac healing and also lead to fragility of myocardium.

In the present study, the number of myofibroblasts was reduced in KO mice compared with WT mice at 5 days after MI. As myofibroblasts play a pivotal role in strengthening wound in healing process, impaired recruitment of myofibroblasts may explain why cardiac rupture was more often observed in KO mice in 4 to 6 days after MI. However, my data showed that transforming growth factorβ1 (TGF-β1), a cytokine which has been shown to induce differentiation of fibroblasts into myofibroblasts, was highly expressed in KO mice. Although these results appear discrepant, impaired cytokine balance by impaired angiogenesis may explain this discrepancy because TGF-β is a pleiotropic and multifunctional cytokine, known to exert diverse and often contradictory cellular effects on all cell types.

In the previous study, a PPARγ agonist has been shown to regulate Bcrp1/Abcg2 expression positively. In addition, a PPARγ agonist has been demonstrated to ameliorate ventricular remodeling after MI in mice. These results suggest that a PPAR gamma agonist may improve cardiac healing, in part, through up-regulation of BCRP1/ABCG2 expression. Together with these results, my findings suggest that up-regulation of BCRP1/ABCG2 might be a promising strategy for treatment of MI.

Conclusions:

I demonstrated that BCRP1/ABCG2 plays a pivotal role in cardiac repair after MI via protection of microvascular endothelial cells. My results suggest that BCRP1/ABCG2 may be of interest for a therapeutic target to improve clinical outcomes after MI.

本研究はABCトランスポーターの一つであるBCRP1/ABCG2の心臓における発現意義を明らかにするため、マウス心筋梗塞モデルを用いて心筋梗塞後の組織修復におけるBCRP1/ABCG2の役割の解明を試みたものであり、下記の結果を得ている。

1. RT-PCRの結果、心臓におけるBCRP1/ABCG2の発現を確認した。免疫組織染色にて、BCRP1/ABCG2は主に微小血管内皮細胞に発現していることが示された。

2. 野生型マウス(WTマウス)及びBcrp1/Abcg2ノックアウトマウス(KOマウス)の冠動脈(前下行枝)を結紮することにより、心筋梗塞を作成し、生存率、組織学的所見、心臓超音波検査所見、カテーテル検査所見を比較した。生存率は、WTマウスと比較してKOマウスで有意に低く、また心筋梗塞後一週間以内の心破裂のイベントが有意に多かった。心筋梗塞作成前における組織学的所見、心臓超音波検査所見、カテーテル検査所見では、両者に差は認めなかったが、心筋梗塞後28日目において、組織学的所見及び心臓超音波検査における心筋梗塞後の左室リモデリングはWTマウスに比較してKOマウスで有意に高度であり、心臓超音波検査所見及びカテーテル検査所見上の心機能も有意に低下していた。以上より、BCRP1/ABCG2は心筋梗塞後の生存及び左室リモデリングにおいて重要な役割を果たしていることが示された。

3. KOマウスにおいて心破裂を最も高頻度に認めた心筋梗塞後5日目の心臓組織においてその組織修復過程を評価するため、免疫組織染色にて心筋梗塞境界領域における血管密度、マクロファージ密度、筋線維芽細胞密度を比較したところ、WTマウスに比較してKOマウスにおいて有意に各密度は低下していた。また、心筋梗塞後28日目における心筋梗塞境界領域での血管密度も比較したところ、WTマウスに比較してKOマウスにおいて有意に低下していた。以上の所見は、KOマウスにおいて心筋梗塞後の組織修復が障害されていることを示唆する所見であり、BCRP1/ABCG2が心筋梗塞後組織修復において重要な役割を果たしていることが示された。

4. 心筋梗塞作成前、心筋梗塞後5日目及び心筋梗塞後28日目において、各種サイトカインの発現をreal time PCRにて比較したところ、心筋梗塞作成前においてはWTマウス及びKOマウス間に差を認めなかったが、心筋梗塞後5日目における血管新生関連サイトカイン、炎症性サイトカイン及び線維化関連サイトカインの発現はWTマウスと比較してKOマウスにおいて不変もしくは有意に増加していた。心筋梗塞後28日目においては、VEGFA及びAngiopoietin-1の発現が、WTマウスに比較してKOマウスにおいてそれぞれ有意に低下及び増加していたが、それ以外のサイトカインの発現に関しては有意差を認めなかった。以上の結果は、KOマウスにおける心筋梗塞後組織修復の障害が、サイトカインの発現が障害されることによって引き起こされるのではなく、BCRP1/ABCGを発現している細胞、すなわち微小血管内皮細胞の障害によるものである可能性を示唆する結果であった。

5. RT-PCRの結果、ヒト微小血管内皮細胞にBCRP1/ABCG2が発現していることが確認された。

6. ヒト微小血管内皮細胞において、BCRP1/ABCG2の特異的な阻害薬であるfumitremorgin Cを使用してBCRP1/ABCG2の機能を阻害し、過酸化水素による酸化ストレス下及び低酸素状態での細胞の生存率をMTS assayにより測定した。負荷が無い状態では、BCRP1/ABCG2の阻害の影響はなかったが、酸化ストレス下ではBCRP1/ABCG2阻害により、細胞の生存率が有意に低下していた。低酸素状態ではBCRP1/ABCG2阻害による影響は認めなかった。以上より、BCRP1/ABCG2はヒト微小血管内皮細胞の酸化ストレス下における生存に重要な役割を果たしていることが示された。

7. ヒト微小血管内皮細胞において、BCRP1/ABCG2阻害の有無による、TNF-α負荷後のICAM-1, VCAM-1, eNOSの発現を比較したが、有意差は認めなかった。

8. 過去の文献において、BCRP1/ABCG2の輸送する基質として、ヘムの前駆体であるprotoporphyrin IXが報告されており、BCRP1/ABCG2の阻害によりprotoporphyrin IX及びヘムの蓄積が生じ、Fenton反応により酸化ストレスを増悪させる原因となりうることが示唆されている。ヒト微小血管内皮細胞における、BCRP1/ABCG2のprotoporphyrin IXの動態に与える影響を評価するため、BCRP1/ABCG2阻害下の細胞内protoporphyrin IX濃度をFACS解析にて評価したところ、ヒト微小血管内皮細胞においても、BCRP1/ABCG2阻害によりprotoporphyrin IXが細胞内に蓄積することが明らかとなった。以上より、BCRP1/ABCG2はprotoporphyrin IXを排出することにより、酸化ストレス下での微小血管内皮細胞の生存に重要な役割を果たしていることが示唆された。

以上、本論文は、ABCトランスポーターであるBCRP1/ABCG2が微小血管内皮細胞を酸化ストレスから保護することにより、心筋梗塞後の組織修復において非常に重要な役割を果たしていることを明らかにした。本研究はこれまで心臓における発現意義が不明であったABCトランスポーターであるBCRP1/ABCG2の役割の解明に重要な貢献をなすと考えられ、学位の授与に値するものと考えられる。