 Biological systems display the ability to self-organize. This phenomenon is easy to appreciate at the macroscopic level, but a striking example of self-organization is demonstrated by cells in which functionally related molecules assembled form macromolecular machines that orchestrate biological events. These machines form and disassemble at distinct cellular addresses in a time-resolved manner. The notion of molecular organization in vivo was presented as early as 1930 when it was predicted that proteins formed a three-dimensional mosaic extending throughout the cell. But despite such a compelling hypothesis, a drastically different notion prevailed and the cell was considered to be a bag of enzymes in which biological events occurred as a result of the random collision of biomolecules. So, although efforts to understand the basis of molecular organization in vivo is relatively recent, a major goal of protein science now is to capture macromolecular machinery directly in vivo as they function, but the complex nature of the cell makes these studies incredibly difficult. One way to deal with cellular complexity is to take cells and breast them open so we can capture the cell interior. The resulting sample is called a cell extract, and it's a physiologically relevant solution to study protein behavior. To prepare cell extracts, we take 50 mils of an E. coli culture and harvest the cells by centrifugation. The resulting cell pellet is then resuspended in one mil of a physiologically relevant buffer. The cells are lised by sonication and the cell debris removed by centrifugation. And the resulting soluble portion is the cell extract. Obviously, lysis disrupts cell walls and membranes, but it also shears DNA and it disrupts the extensive macromolecular networks that wire the cytoplasm. But it's likely that similar physical chemical forces drive macromolecular assembly in vivo and in concentrated cell extracts. And in our protein science article, we describe the interactions of a protein in concentrated E. coli extracts. So Kira has already elegantly explained the problem. What we want to do is to somehow observe proteins inside cells and understand a little bit more about their interactions and how they actually function together in such a crowded, complex environment. In cell NMR spectroscopy has emerged as one of the techniques to address this question. But the big problem or the prevailing issue is that most proteins are undetectable by in cell NMR. So probably everybody is familiar with the HSQC spectrum, which is really just a fingerprint of a protein. And it's a lovely way of observing almost every amino acid residue in the protein. And here you can see a HSQC spectrum of a typical folded protein, lots of resolved peaks, easy to figure out and to assign. But when you then try to observe that same protein inside a simple bacterium like E. coli, the spectrum turns out to be this kind of disaster where there are very few peaks that we can actually follow. So this is the problem that we're trying to solve. And we're trying to answer that question, of course. Well, why are proteins that may not be so sticky under in vitro conditions, but when you pop them inside the cell or look at them inside the cell, they are again these sticky molecules? So what we've been trying to do then is to circumvent the problem, dissecting cells and looking at cell extracts and trying to find out where and when the protein becomes visible. So, for example, GB1 is considered as a biologically inert protein. It's relatively unsticky. And there are numerous papers about in cell NMR and GB1 in different types of cells, both in E. coli and in mammalian cells. What we did was take GB1 and fuse it with the sticky motif of TAT, so this arginine-rich TAT motif which is readily involved in numerous types of interactions. And of course, when you make that fusion protein, as has been shown by others, you produce now a molecule that likes to stick and bind to anything inside the cell. Taking the NMR dissection method, then we've been able to lyse the cells and then add exogenous components like enzymes and ions and observe at which stage the protein becomes visible. We can easily observe in the SEC where our protein of interest elutes and this provides information on the potential complexes that that protein is part of. So, ideally, we can change the conditions in our SEC and alter the nature of those complexes and therefore learn about the types of interactions that hold the protein to other macromolecules in the extract. So, science is all about being exact and we want to know the answer and we're always very certain about our techniques and how great they are and so on. But of course, we must also use other approaches and the artist's impression of the cellular interior has been very valuable in this regard. At some stage in the future, we will have ways of observing the entire cell and looking at every molecule of interest but for now, we're using the likes of David Goodcell's image or Adrian Elcock's simulations to get a glimpse at the macromolecular assemblies inside the cell and it's important to recognize the contribution of art to the development of our ideas. I think there's a way forward we can jump to some new interpretation if we work together with art.