Spray Painting Cancer

I’m pretty certain that anyone reading this will be fully aware that one of the biggest problems in cancer is spotting the blighters. We have, of course, X-ray detection (as in mammography), CTs and MRI scans, all so familiar we need not bother to define them, and there’s also a variety of sampling methods for specific cancers (e.g., the Pap test for cervical cancer). But, useful though all these are, the plain fact of the matter is that none are ideal and in particular the pictures created by imaging methods are very limited in sensitivity. Put another way, they won’t pick something up until it is quite large – a centimeter in diameter – meaning that the abnormal growth is already quite advanced.

Cunning Chemistry

Needless to say, much inspiration and perspiration is being applied to this matter and what has been really exciting over the last ten years or so is the way very smart chemists are collaborating with clinicians to come up with new ways of looking at the problem. One of these clever tactics is being developed in the University of Tokyo using a different type of imaging ‘reporter’ that signals its presence by fluorescing. Fluorescence occurs when a molecule absorbs light and becomes ‘excited’ before relaxing back to its ‘ground state’ by giving off a photon. Fluorescent molecules (fluorophores) are much used in biology because the background signal is often very low so the high signal-to-noise ratio gives excellent sensitivity.

Spray Paint scheme

The cell-surface enzyme GGT converts the small molecule  gGlu-HMRG to a fluorescent form (HMRG) that is then taken up by the cell. GGT is only found on tumor cells so they light up and normal cells do not

Fortunately we don’t need to know how the chemists did it – merely to say that Yasuteru Urano and his colleagues came up with a small molecule (called gGlu-HMRG for short) that does not give off light until a small fragment is chopped off its end, whereupon it changes shape: this flips the switch that turns on fluorescence. The cutting step needs an enzyme that is found on the surface of various cancer cells but not in normal tissue (GGT for short).

Joining Forces

To show that there was real mileage in their idea they followed the time-honored blue-print of cancer research, showing first that it works on tumor cells grown in the lab (and, equally important, that it doesn’t highlight normal cells), before moving to mouse models of ovarian tumors. The later is where chemists meet clinicians because an endoscope is required (quite a small one) – a flexible tube for looking inside the body – devices now so sophisticated that they can incorporate a fluorescence camera.

In the final synthetic step the cunning chemists formulated a spray-on version of their probe molecule so that it can be dispensed during endoscopy or surgery – a bit like an underarm deodorant. Now it’s easy: find suspect tissue, give it a squirt of gGlu-HMRG, wait a few minutes and see if it lights up. The answer is, of course, that in their ovarian cancer model the spray-on graffiti lights up within 10 minutes of sticking to a tumor cell and can detect clumps of cells as small as 1 millimeter in diameter – a terrific advance in terms of sensitivity. The brief time taken for the signal to be visible after the probe has been applied means that within the same procedure it could be used to guide surgeons in removing small tumor masses.

The Tokyo system is not the only one under development. My colleague Andre Neves at the Cambridge Cancer Centre, another of these fiendishly clever chemists, is working on a parallel line using different fluorophores that can be topically applied to the lining of the intestine. The goal here is, of course, the early detection of colon tumors. Yet other approaches use molecules that accumulate preferentially in tumor cells and respond to light in the near-infrared region of the spectrum (800 nm to 2500 nm wavelength, compared to just under 500 nm for gGlu-HMRG), giving an even better signal-to-noise ratio.

This is, as Mr. Churchill might have pointed out, not even the beginning of the end of this story. But it is one more small and innovative step forward. Not all cancers even of the same type will be detectable by a given probe because they vary so much in the genes they express but the ingenuity of the chemists gives hope that a substantial panel of ever more sensitive reporters will emerge. It is also true that endoscopy is unlikely to gain widespread popularity as a routine screening method. However, these advances, moving us to detection at ever earlier stages may become very powerful as a follow-up test, combined with the capacity for simultaneous treatment, when tumor cells have been detected in more comfortable screens, for example as circulating cells in small blood samples, an immensely exciting prospect to which we will return in a later episode.


Urano, Y., Masayo Sakabe, Nobuyuki Kosaka, Mikako Ogawa, Makoto Mitsunaga, Daisuke Asanuma, Mako Kamiya, Matthew R. Young, Tetsuo Nagano, Peter L. Choyke, and Kobayashi, H. (2011). Rapid Cancer Detection by Topically Spraying a γ-Glutamyltranspeptidase–Activated Fluorescent Probe. Science Translational Medicine 3, 110ra119.


Shi, C. (2012). Comment on “Rapid Cancer Detection by Topically Spraying a γ-Glutamyltranspeptidase–Activated Fluorescent Probe. Science Translational Medicine 4, 121le1.



The Creation of Cancer

Where do cancers come from?’ One of those dreaded childish questions – so best to get your thinking in first, rather than trying to answer on the hoof in the face of that unblinking stare of expectation. In the beginning, as you might say, we need a hand-wavy word on how DNA ‘makes proteins’, why they’re important (‘Proteins R Us’, in short) and what can go wrong with them.

DNA double-helix

The double helix of DNA

In 1953 Watson and Crick worked out the structure of DNA. It holds, of course, the secret of life and you might observe that it has the appropriate shape of a spiral staircase to nowhere. The ‘genetic code’ is the order of thousands of small bits that are linked together to make the very long molecules of DNA. These bits contain smaller bits called bases – four of them (A, C, G and T) – and they’re firmly stuck together so that each DNA molecule is pretty stable. In addition, bases in one DNA can stick to those in a second strand – hence the double-helix.


DNA encodes proteins

The essence of life is the transformation of the genetic code into the corresponding sequence of the building blocks that make proteins. The blocks are amino acids, stitched together to make proteins in much the same way as DNA is built from its base-containing units. There are 20 different types that can be glued together in any order, a typical protein containing a thousand amino acids. They tell the protein how to fold up into its final shape – a 3D structure unique for each protein. Many proteins are blobs (like balls of string) but, as you’d guess given that they do everything, they come in all shapes and sizes—cables, sheets, coils, bridges, etc. The idea then is fairly simple: flexible protein chains fold themselves into their working shape – and individual shapes enable proteins to do specific jobs. A simple sum can show that a limitless variety of proteins can be made: they are the machines of life that make all living things work and they have created all the species of life on earth.


Proteins make life possible because the exquisite choreography that generates their shape creates localised regions (sticky bits, clefts, cavities, etc.) for interactions with other molecules. These confer amazing versatility: proteins can ‘talk’ to each other and form relay teams that transmit information from one part of a cell to another, they can generate movement (as in muscles), and bring molecules together (e.g., when they act as enzymes driving chemical reactions that otherwise would not occur). But, as we all know, mistakes can happen even in the best-run enterprises. Mistakes in proteins arise from mutations – changes in the DNA code. Many diseases result from single base alterations: if that changes an amino acid the result can be a protein with dramatically altered function. A well-known example is cystic fibrosis: a protein made in the lung has one abnormal amino acid: the effect on its activity causes a build-up of mucus that makes breathing difficult and is a target for fatal infections.

Mutations and cancer

Cancers are also caused by mutations but they’re a bit more complicated, being driven by groups of mutations, rather than by one event. For most cancers these are picked up as we go through life – so the creation of a cancer is a slow process. Most don’t appear until we are over 60 years of age – collecting a suitable hand of mutations takes time. Because several critical mutations are required you’d guess that what tumour cells are up to is evolving a number of tactics for outsmarting their normal counterparts on the survival front. Indeed they are. They multiply in an unregulated way (because they ignore signals that control normal cells), side-step protective mechanisms that usually kill abnormal cells, divert nutrients from normal tissue to themselves, and make new blood vessels for the delivery of food and oxygen. Perhaps most amazingly of all, they seduce and subvert cells of the immune system: these begin by trying to eliminate the tumour but end up playing a key role in its growth – a sort of co-operative corruption.

All this is why cancer needs several mutations, and these are part of a wider genetic mayhem that will kill most cells – because essential survival genes are damaged. The cells that emerge as tumour precursors are molecular freaks in that they’ve both survived and picked up a bag of dirty tricks with which to out-compete their normal brethren. So, molecularly speaking, cancers are rare events. What’s more, there’s no forethought, no premeditation at work here. If the expression ‘unintelligent design’ conveys random chance in a game of genetic roulette then it’s an excellent descriptor of cancer evolution.

Stop me if you’ve heard it

If all this is beginning to sound familiar, so it should. It’s a completely undirected process that usually fails – but when it succeeds represents an extraordinary triumph of the flexibility of DNA and hence the adaptability of cells. Familiar, of course, because it’s a form of evolution that parallels the emergence of new species.

Tree of life

In the revolution started by unveiling the structure of DNA, the biggest advance has been finding a way to work out the order of bases – the genetic code. The first complete human DNA sequence came in 2003. Since then astonishing technical advances have led to thousands of tumours and hundreds of different species being sequenced. From this you can estimate when new species arose and draw a map of the evolution of all major forms of life on earth from a single, common ancestor. The time scale is incomprehensibly vast, but the picture is stunning in its simplicity, showing how everything is related – bugs, plants, fungi and humans – and how that family has emerged over nearly four billion years. This would have delighted Charles Darwin who, in 1859, was able to define evolution by natural selection only on the basis of what he could see. Molecular biology has now revealed its foundations.

Cancer evolution

In many ways tumours do indeed behave like new species: through the acquisition of mutations they out-compete normal neighbours and establish new niches in which to survive and prosper. But tumours are not new organisms: they’re normal cells that have gone off the rails – been hijacked, if you will, by delinquent genes. The big difference is the brief time scale over which tumours develop compared with the almost infinitely slow, step-wise testing of novel genetic variants in species evolution. So becoming a tumour is a very chancy business – but it’s a lot less fraught than making a new form of life. They take any short-term growth advantage conferred by a mutation without concern for the consequences.

Short trials and lots of errors

When a cell picks up its first growth-promoting mutation it has taken an irreversible step towards a life of crime. It’s become a high roller in the cellular casino, addicted to roulette of the Russian variety, and no amount of genetic counselling will reform it. If only it could think, how our tumour cell would long for a guiding hand – a more knowing form of life that could steer its orgies of DNA destruction toward survival. Alas! Like every other life form, tumours are in thrall to the random creator called chemistry. In a tiny few the dice fall favourably and they grow to rule their kingdom – briefly. Oh for an intelligent brain to design them not to kill their life-support system! Like cellular spaceships seeking immortality in the celestial wastes without the know-how to reach escape velocity, they can only burn brightly before crashing. Tumours are indeed a microcosm of evolution, working on an abbreviated time-scale – they’re dynamic Darwinism.