Endothelial progenitor cells (EPC) are still one of promising sources of regenerative medicine even though induced pluripotent stem (iPS) cells are becoming increasingly important in this field. The isolation of EPC was first reported by Dr. Asahara in 1997. They reported that the CD34 positive cells in the human adult peripheral mononuclear cell fraction differentiate into vascular endothelial cells, and demonstrated that these cells were derived from bone marrow and contribute to the postnatal vascular repair. Owing to this report, the interest in EPC as a potential source of cell therapy increased, and numerous basic and clinical studies of EPC have been conducted. However, their overall impact is still controversial, mainly because of the existence of some unresolved issues, including proper clinical trial designs, the optimal time point for cell transplantation, cell delivery methods for the target organs and the definition of EPC itself. In particular, there have been several definitions of 'EPC' in both pre-clinical and clinical studies from different researchers, and these were widely defined, ranging from bone marrow-derived MNCs to cells with specific surface marker expression. Thus, the exploration of the 'best EPC' for cell therapy and the appropriate classification of heterogeneous EPC are crucial for there to be steady progress in EPC studies for clinical applications. Recently, the definition of 'Early EPC' and 'Late EPC' was proposed according to the cell appearance and the timing of colony formation. Late EPC is defined as cells with a cobblestone shape, emerging relatively late in the MNC culture, producing more nitric oxide, and forming capillary-like tubes in vitro. Later, some researchers reported that the main mechanism by which Late EPC increased angiogenesis was via direct incorporation into pre-existing vascular networks. Therefore, Late EPC might be closer to the putative EPC.

Prompted by these reports, we focused on the Late EPC and tried to further investigate their functional features. We noticed that the timing of colony emergence of the Late EPC was widely distributed, from 10 to 30 days after the seeding of MNCs on the culture plates. Thus, we hypothesized that Late EPC consist of some heterogeneous cells with different properties and angiogenic potentials. In this study, the period of colony emergence of Late EPC was divided into three groups. Then, the cells were redefined according to these periods, and their angiogenic properties were compared to clarify the best EPC to be used as a source of cell therapies.

In the present study, various types of EPCs were originally defined. First, human peripheral blood MNCs (hMNCs) were cultured on human fibronectin-coated wells with EBM-2 medium for 3 to 7 days, and adherent clustered cells were defined as Early outgrowth cells (EOC). Next, human MNCs were cultured on rat tail collagen I-coated wells, and the cells of the emergent colonies with a cobblestone appearance from days 10 to 30 were defined as the Late EPC. Then, these were sub-divided into three groups as follows: 10 to 16 days as MOC (mid-term outgrowth cells), 17 to 23 days as LOC (late-term outgrowth cells), and 24 to 30 days as VOC (very late-term outgrowth cells). According to this definition, the angiogenic properties of EPCs were investigated by following several methods in vitro and in vivo.

To investigate the differences in cell surface antigen expression among the various types of EPCs, the expression of some important markers were evaluated by flowcytometry. Most cells of all four types were strongly positive for CD31 and CD105, and negative for CD133. On the other hand, the positive rate of some markers varied among four types of EPCs, e.g., EOC: CD34-CD14+CD45(high), MOC: CD34+CD14(low)CD45+, LOC: CD34+CD14-CD45(low) and VOC: CD34+CD14-CD45-. The proliferative potential of EPCs was assessed by the commercially available kit (MTS assay). LOC demonstrated significantly high proliferative activity with approximately 1.3 fold of other EPCs. Next, the in vitro angiogenic potential of EPCs was evaluated using Matrigel tube formation assay. Only the LOC could form thicker and tighter tube-like structures on the Matrigel, whereas MOC and VOC could form only a thin cord-like network. The EOC never formed capillary-like tubes in Matrigel. The total length of these structures was significantly longer in the LOC. To compare the mRNA expression levels of crucial markers for angiogenesis among EPCs, a quantitative Polymerase Chain Reaction (qPCR) analysis was performed. The mRNA expression level of eNOS in LOC was significantly higher than those in the other EPCs. The expression levels of Flk-1, Flt-1 and Tie-2 in LOC were also highest among EPCs whereas no statistical difference was observed. From these in vitro analyses, the different properties of four types of EPCs were clarified, and the superiority of proliferative activity and angiogenic potentials in LOC among four types of EPCs was suggested.

To demonstrate this superiority of LOC in vivo, several analyses using mouse hindlimb ischemic model were conducted. First, to confirm that the injected EPC subpopulation could directly incorporate into the mouse vasculature, GFP-labeled EPCs were intravenously injected into immunocompromised mice on days 1, 3, 5 and 7 after unilateral hind limb ischemia surgery (1×105 cells per injection). Ten days after surgery, GFP signals were detected in some capillaries of the LOC-injected mouse (in approximately 4% of capillaries) but were hardly detected in EOC-injected mouse capillaries. Next, we investigated the continuous therapeutic efficacy of EPC transplantation using the same model. Four weeks after surgery, the mice in the LOC-injected group showed significantly enhanced blood flow recovery as measured by Laser Doppler scanning (blood flow ratios of the ischemic/nonischemic leg: 0.94 ± 0.02 [LOC group] versus 0.74 ± 0.07 and 0.78± 0.09 [EOC and MOC groups], P<0.05). In addition, capillary density measured by histological staining of both anti-mouse CD31 and anti-human CD31 was highest in LOC-injected group. These results suggested that the ability of both enhancing recipient cell-derived capillary formation and direct incorporation of injected cells into a part of recipient capillary was greatest in the transplantation of LOC among groups.

From these in vitro and in vivo analyses, we concluded that Late outgrowth EPC emerging on days 17-23 in ex vivo culture of human PBMNCs are superior to other EPC subpopulations with regard to the angiogenic potentials as the source of cell therapy.

本研究は血管新生において中心的な役割を果たすと考えられる、血管内皮前駆細胞(以下、EPC)について詳細な検討を行ったものである。EPCは末梢血の単核球系細胞に含まれると考えられ、近年ではex vivoで培養を行った結果得られるEPCを、そのcolony出現時期に応じてearly EPCおよびlate EPCと分類し、その役割の相違について様々な研究がなされている。ここでは、ヒト末梢血由来の単核球を内皮細胞用の培地で培養し、培養開始からcolony出現するまでの期間に応じて、EOC (3-7days), MOC (10-16days), LOC (17-23days), VOC (24-30days)と分類し、in vitroおよびin vivoにおいて、それぞれの治療的血管新生能を調べ、以下のような結果を得られている。

1.本研究において定義した各EPCに関して、その表面マーカーをflowcytmetryで精査したところ、各EPCによってその陽性率は異なり、機能的な差異の存在や分化過程の差異の存在を示唆する結果となった。これは、便宜的に定義した分類の妥当性を示唆するものと言えるが、各表面マーカーの意義に関しては科学的に不明な点もあり、今後の精査が望まれる。

2.MTS assayを用いた細胞増殖能試験、およびMatrigel assayを用いた血管新生能試験において、各EPCの中でLOCが最も優れていることが示された。

3.定量的PCR法による、mRNA発現量を精査した結果、LOCにおいてeNOSの発現が他EPCに比し有意に亢進していることが示された。血管新生の過程におけるeNOSの役割に関しては不明な点も多いが、in vivoの実験でLOCが優れた血管新生能を示す、一つの根拠である可能性が示唆される。

4.EPC投与による治療的血管新生能を調べるため、免疫不全マウスを用いて下肢虚血モデルを作成し評価を行った。まず、GFP遺伝子を導入して標識したEPCの投与を行い、投与したEPCが実際に内皮細胞に分化し、毛細血管網の一部として下肢血流の回復に寄与しているか否かを精査した。その結果、他EPCに比し、LOC投与群で有意にその寄与率が高いことが示された。ただし、実際に投与したEPCが毛細血管に分化している割合は限られていると推測され、EPC投与による下肢血流の回復は、直接的な内皮細胞への分化だけでは説明がつかないことが示唆された。

5.次に、下肢血流の経時的・継続的な回復を評価するため、レーザードップラーによる下肢血流の評価、および毛細血管量の組織学的評価を行った。いずれにおいても、LOC投与群で他EPC投与群に比し優れた結果を示した。

以上、本論文はEPCと称される細胞群の中でも、内皮細胞用の培地で培養した後、17-23日目にcolonyを形成するEPCが、最も優れた血管新生能を有する可能性があることを示した。本研究の結果は、発展途上であるEPCを用いた治療的血管新生の研究に重要な貢献をなすことが期待され、学位の授与に値すると考えられる