A novel in vitro model of tumor angiogenesis.
A novel in vitro model of tumor angiogenesis.
复制标题
肿瘤血管生成的新型体外模型。
DOI：
10.1007/bf02577521
复制
发表时间：
2000
期刊：
影响因子：
--
通讯作者：
RayChaudhury,A
中科院分区：
文献类型：
--
作者：
Kozien,D;Gerol,M;Hendey,B;RayChaudhury,A
编辑先生：大量证据表明，实体瘤的生长依赖于血管生成（Hanahan和Folkman，1996）。已经开发了几种模型来阐明肿瘤诱导的血管生成的机制。血管生成的体内模型虽然提供了关于特定药剂是否是血管生成的或抗血管生成的极好信息，但对于详细的细胞和分子研究并不方便，因此需要开发体外模型。两种最流行的体外模型利用基质胶或胶原凝胶结合血管生成剂，诱导内皮细胞（EC）形成微血管。然而，这两种模型都是繁琐的设置，并具有有限的适用性，研究肿瘤血管生成，由于在这些环境中开发容易分析的共培养物的不便。最近已经开发了几种血管生成的体外模型以改进这些系统（Wolff等人，1993; Chopra等人，1997; Ment等人，1997），包括使用用于细胞生长的旋转血管壁。我们在此描述了一种肿瘤血管生成模型，该模型是二维的，易于建立和分析，并且成本低廉，其中肿瘤细胞而不是添加的血管生成剂诱导小管形成。在该模型中，如图1所示，24孔培养皿的威尔斯孔（或组织培养载玻片的室）完全覆盖有溶解在Dulbecco改良Eagle培养基中的1%低熔点琼脂糖。琼脂糖凝胶化后，通过铲取除去一部分（约70%面积）琼脂糖，并将人乳腺癌MCF-7细胞（美国典型培养物保藏中心[ATCC]，Rockville，MD）以汇合浓度（约300，000个细胞/cm 2）接种在暴露区域中。细胞粘附并铺展在裸露区域上，并且不在剩余琼脂糖下爬行。在癌细胞粘附并扩散（通常过夜）后，除去剩余的琼脂糖，吸出培养基，并将牛肺动脉EC（BPAEC，来自ATCC）以汇合浓度（~ 200，000个细胞/cm 2）铺板。BPAEC不粘附在癌细胞的顶部（MCF-7细胞粘附在EC单层上，因此首先接种MCF-7细胞是至关重要的），因此它们仅粘附在约30%的表面积中，其中剩余的琼脂糖阻止MCF-7细胞接种，产生分离的共培养物（图2a）。通过将细胞与含有DiI-Ac-LDL的培养基一起孵育，证实了这些共培养物的分离性质（BTI，斯托顿，MA），一种被EC掺入但不被MCF-7细胞掺入的荧光标记物（图2b）。在该装置中，EC单层在约3天后经历破裂，细胞表现出细长的形态以及重叠和分散的分布（图2c），并在7-10 d形成微血管网。如图3（图1-3）所示，这些”微血管”形成分支图案并呈现圆柱形，这两个特征与真正的毛细血管一致。相反，在对照共培养物中，当非恶性正常培养的人乳腺EC或3 T3细胞
Dear Editor: Overwhelming evidence suggests that the growth of solid tumors is angiogenesis-dependent (Hanahan and Folkman, 1996). Several models have been developed to elucidate the mechanism of tumor-induced angiogenesis. In vivo models of angiogenesis, while pro-viding excellent information on whether or not a particular agent is angiogenic or antiangiogenic, have not been convenient for detailed cellular and molecular studies, necessitating the development of in vitro models. Two of the most popular in vitro models utilize either matrigel or collagen gels in conjunction with angiogenic agents, to induce endothelial cells (EC) to form microvessels. However, both models are cumbersome to set up and have limited applicability for studying tumor angiogenesis due to the inconvenience of developing easy to analyze cocultures in these environments. Several in vitro models of angiogenesis have recently been developed to improve upon these systems (Wolff et al., 1993; Chopra et al., 1997; Ment et al., 1997), including using rotating vessel walls for cell growth. We describe here a model of tumor angiogenesis that is two-dimen-sional, easy, and inexpensive to set up and analyze, and where tumor cells and not added angiogenic agents induce tubule forma-tion. In this model, schematically shown in Fig. 1, wells of a 24-well dish (or chambers of tissue culture slides) were completely covered with 1% low-melt agarose dissolved in Dulbecco modified Eagle medium. After the agarose gelled, a fraction (-70% by area) of the agarose were removed by scooping and human breast cancer MCF-7 cells (American Type Culture Collection [ATCC], Rockville, MD) were plated in the exposed area at confluent concentrations (-300,000 cells/cm2). The cells adhered and spread on the bare area and do not crawl under the remaining agarose. After the cancer cells had adhered and spread (usually overnight), the rest of the agarose was removed, media aspirated, and bovine pulmonary ar-terial EC (BPAEC, from ATCC) were plated at confluent concentra-tions (-200,000 cells/cm2). BPAEC do not adhere on top of the cancer cells (MCF-7 cells adhere to EC monolayer, so it is critically important to plate MCF-7 cells first), so they adhered only in-30% of the surface area where MCF-7 cells were prevented from plating by the remaining agarose, giving rise to a segregated coculture (Fig. 2a). Segregated nature of these cocultures was confirmed by incu-bating cells with media containing DiI-Ac-LDL (BTI, Stoughton, MA), a fluorescent marker incorporated by EC but not by MCF-7 cells (Fig. 2b).In this setup, the EC monolayer underwent disruption after about 3 d, with cells exhibiting an elongated morphology and an overlapping and scattered distribution (Fig. 2c), and eventually formed mi-crovascular networks in 7-10 d. As shown in Fig. 3 (panels 1-3), these" microvessels" formed branching patterns and appeared cylindrical, both characteristics consistent with true capillaries. In contrast, absolutely no tube formation was seen in control cocultures when nonmalignant normal cultured human breast ECs or 3T3 cells