Multiplexed, Quantitative Fluorescence Imaging in Tumor Sections

 描述（由申请人提供）：癌症的精准医学需要了解特定患者肿瘤中的驱动突变。最近的单细胞遗传学研究表明，驱动突变和癌细胞亚型在单个患者中具有高度异质性。因此，临床医生将需要采用实验方法，允许在特定患者的肿瘤切片内原位观察许多单个细胞中的各种分子分析物。目前的肿瘤切片分析不是高度多路复用的，并且通常仍然限于~4-5种分析物，或7种多光谱成像。最近的技术已经将这个数字推到30 - 60，但是它们需要昂贵的设备和/或试剂，复杂的分析或显著增加的测定时间，所有这些都将妨碍它们在许多临床病理学和临床前研究实验室中的实际使用。因此，仍然非常需要在肿瘤切片中进行多重测量但可广泛获得且具有成本效益的技术。我们专注于解决这一需要与易于采用，但新的多光谱荧光为基础的方法。它基于这样的假设，即可以利用组合学的力量，通过以新的方式排列广泛的可用荧光团，大大增加混合物中可定量分析物的数量。我们称我们的方法组合荧光光谱成像（CoFSI）。CoFSI仅需要执行多通道荧光激发和发射光谱扫描的能力，这在大多数板/载玻片读取器和许多显微镜中广泛可用且易于实现。数据分析涉及一种简单、快速的计算技术，称为线性分解。初步的模拟研究表明，从16个现有的荧光蛋白构建的123个不同的CoFSI探针的浓度可以同时估计具有良好的准确度和精度，跨越3个数量级的浓度与大量的激发通道，和48个探针与6个激发通道。受可用肿瘤切片成像设备和Alexa染料限制的类似模拟研究表明，25次同时测量是可能的。初步试验表明，混合物中的七种不同CoFSI探针水平可以仅使用光谱的蓝黄部分准确和精确地测量。该方案进一步通过实验测试了CoFSI的局限性，并将CoFSI应用于肿瘤切片成像，有两个目的：（1）同时定量混合物中48种荧光探针的水平;（2）定量肿瘤切片中25种分析物的空间分布。如果成功，CoFSI可以将定量荧光多路复用增加至少~5至10倍，同时依赖于标准实验室资源和简单的分析。CoFSI也与其他难以复用的技术可行地兼容，例如高含量筛选、活细胞成像和体内啮齿动物成像，因此可能具有广泛的影响。

 DESCRIPTION (provided by applicant): Precision medicine for cancer requires knowledge of the driver mutations in a particular patient's tumor. Recent single cell genetic studies have shown that driver mutations and cancer cell subtype are highly heterogeneous within a single patient. Thus, clinicians will need to employ experimental methods that allow observation of a variety of molecular analytes across many individual cells within a particular patient's tumor section in situ. Current tumor section analyses are not highly multiplexed and typically remain limited to ~4-5 analytes, or 7 with multi-spectral imaging. Recent technologies have pushed this number into 30s-60s, but they require expensive equipment and/or reagents, sophisticated analyses or markedly increased assay time, all of which would preclude their practical use in many clinical pathology and preclinical research laboratories. Thus, there remains a significant need for technologies that multiplex measurements in tumor sections but are widely accessible and cost-effective. We focus on addressing this need with a readily-adoptable but novel multi- spectral fluorescence-based method. It is based on the hypothesis that the power of combinatorics can be harnessed to vastly increase the number of quantifiable analytes in a mixture by permuting the wide array of available fluorophores in new ways. We term our approach combinatorial fluorescence with spectral imaging (CoFSI). CoFSI only requires the ability to perform multi-channel fluorescence excitation and emission spectral scanning, which is widely available and easy to implement in most plate/slide readers and many microscopes. Data analysis involves a straightforward, fast computational technique called linear unmixing. Preliminary simulation studies suggest that the concentrations of 123 different CoFSI probes constructed from 16 existing fluorescent proteins can be estimated simultaneously with good accuracy and precision across 3 orders of concentration magnitude with a large number of excitation channels, and 48 probes with 6 excitation channels. Similar simulation studies constrained by available tumor section imaging equipment and Alexa dyes suggest 25 simultaneous measurements are possible. Initial pilot experiments demonstrate that seven different CoFSI probe levels in a mixture can be measured both accurately and precisely using only the blue-yellow part of the spectrum. This proposal further tests the limits of CoFSI experimentally, and applies CoFSI to tumor section imaging, with two Aims: (1) Quantify the Levels of 48 Fluorescent Probes in a Mixture Simultaneously; and (2) Quantify the Spatial Distribution of 25 Analytes in Tumor Sections. If successful, CoFSI can increase quantitative fluorescence multiplexing at least ~5 to 10-fold while relying on standard lab resources and straightforward analyses. CoFSI is also feasibly compatible with other difficult-to-multiplex technologies such as high content screening, live-cell imaging, and in vivo rodent imaging, and thus may have broad impact.