Field effect sensing for protein microarrays

我们由数以万亿计的细胞组成，每个细胞由数十亿蛋白质组成。为了了解我们的身体是如何工作的，我们需要了解细胞中的蛋白质如何相互作用。这种理解的第一步是对人类基因组进行测序，这是一组基因，其DNA编码了我们希望了解的蛋白质。这使得我们能够设计探针--在某些情况下使用人类DNA本身--可以在显微镜载玻片上以非常高的密度打印。我们可以将针对每个基因的探针打印到一张显微镜载玻片上，这样就可以询问在细胞生命的每个阶段，哪些基因被开启或关闭，以产生RNA。这种能力已经创造了对生物学的新理解，并导致了一种关于细胞如何工作的新方式。尽管令人印象深刻，但这仍然描绘了一幅非常不完整的工作细胞机制的图景，因为尽管它们告诉了我们许多关于正常基因作为RNA的表达，但它们没有告诉我们任何关于蛋白质是否由这种RNA构成的信息(并不总是)，也没有告诉我们其基因携带突变导致疾病的蛋白质的表达。为了纠正这一点，我们需要打印能够识别所有不同形式表达蛋白的探针。使用抗体或人类蛋白质本身作为探针，已经尝试了这一点。虽然这很有意义，因为抗体和蛋白质都已经进化到识别每种蛋白质的特定亚型，但事实证明这很困难，原因有两个。首先，由于抗体和蛋白质非常脆弱，将它们打印到玻片上会导致它们失去识别细胞提取液中目标蛋白质的能力。其次，蛋白质是微小的，检测两者之间的相互作用相应地很难。到目前为止，最好的方法是在细胞提取物中的蛋白质上附着一种染料，这样，如果印刷的蛋白质识别出提取物中的目标蛋白质，它就会发光。这是有问题的：连接染料会改变蛋白质，我们面临的风险是，它可能不再被我们的探测器识别，甚至可能被错误的探测器识别。我们计划逐一解决这些困难。首先，我们可以使用人造抗体库，这些抗体库具有识别细胞蛋白质的能力，但非常稳定，可以打印到玻片上，而不会失去这种能力。我们知道，对于我们尝试过的第一个蛋白质来说，这是正确的，这个提议的一部分将允许我们问为什么这是可行的，以及我们是否可以从这些蛋白质的行为中学习有助于我们处理其他不太稳定的蛋白质的想法。其次，我们计划扩展我们的发现，即当蛋白质印在电极表面时，它们以可测量的方式改变电极的电学性质。然后，如果第二种蛋白质(来自细胞提取物)与第一种蛋白质结合，我们就会得到另一种变化--因此，如果我们有许多电极，每个电极携带一种受体蛋白质，我们将能够依次测量每种细胞蛋白质的结合，从而在任何给定的时间测量细胞中每种蛋白质的行为和相互作用。这将使我们能够深入了解细胞中生命的分子基础。

We are made up of trillions of cells, and each cell is made of billions of proteins. To understand how our bodies work, we need to understand how the proteins in a cell interact. The first step in this understanding was the sequencing of the human genome, the set of genes whose DNA encodes the proteins we wish to understand. This has allowed us to design probes- in some cases using human DNA itself- that can be printed in very high densities on microscope slides. We can print a probe that is specific for every gene onto a single microscope slide, and so ask which genes are being turned on, or off to make RNA or not, at every stage in the life of a cell. This ability has already created a new understanding of biology, and led to a new way of thinking about how cells work. Though impressive, this still paints but a very incomplete picture of cellular machineries at work, because although they tell us a lot about the expression of normal genes as RNA, they do not tell us anything about whether or not protein is made from that RNA (it isn¿t always), nor do they tell us about the expression of proteins whose genes carry mutations that will lead to disease. To rectify this, we would need to print probes that recognise all of the different forms of expressed proteins. This has been attempted, using either antibodies or human proteins themselves as the probes. While this makes a lot of sense, because both antibodies and proteins have evolved to recognise specific isoforms of each protein, it has proved difficult, for two reasons. First, because antibodies and proteins are very fragile, printing them onto glass slides causes them to lose their ability to recognise their target proteins in cell extracts. Second, proteins are minute, and detecting the interaction between two is correspondingly hard. So far, the best way of doing this is to attach a dye to the proteins in cell extracts, so that the printed protein lights up if it recognises its target protein in the extract. This is problematic: attaching the dye can change the protein, and we run the risk that it may no longer be recognised by our probes or may even be recognised by the wrong ones. We plan to address each of these difficulties. First, we have access to libraries of artificial antibodies that have the ability to recognise cellular proteins, but are very stable and can be printed onto a slide without losing this ability. We know that this is true for the first ones we have tried, and part of this proposal will allow us to ask why this works, and whether we can learn from the behaviour of these proteins ideas that will help us with other, less stable proteins. Second, we plan to expand upon our finding that when proteins are printed on the surface of an electrode, they change the electrical properties of the electrode in a measurable way. Then, if a second protein (from a cell extract) binds to the first protein, we get another change- so if we have many electrodes, each carrying one receptor protein, we will be able to measure the binding of each cell protein in turn, and thus the behaviour of and interactions between, every protein in a cell at any given time. This will allow us to gain intimate insights into the molecular basis of life in a cell.