DNA Structure
Through the eyes of an enzyme
When Galileo noted - through his telescope - that the earth was round rather than flat, the universe started to make more sense. Understanding a whole system like the universe is impossible without architectural insight. This is very much the case when it comes to studying the inner life of a cell. Figuring out how cells work relies on a clear picture of what they look like. Molecular investigators work on the biological blueprints that cell biologists and geneticists can use to model the mechanics.
When Karl-Peter Hopfner (University of Munich, Germany) puts on his molecular glasses, he is trying to see the cell nucleus through the eyes of an enzyme. "Enzymes see a vast amount of DNA in the nucleus," which means they need to be able to distinguish between healthy and damaged DNA. Why, for example, is smoking toxic and how does it promote cancer?
Smoke forms DNA adducts, tiny chemical bolt-ons to the double-helix, which can damage the way our cells operate if not removed. "If you are smoking actively or passively," explains Karl-Peter, "the DNA repair machinery can't work properly. A DNA adduct itself isn't often a problem. The problem occurs when enzyme function is disturbed, adducts aren't removed, or there are subsequent changes in the DNA sequence.
One very important enzyme is the kind involved in making new DNA before cells divide. "DNA polymerases can make mistakes causing mutations or breaks in the DNA backbone. This genome instability can lead to all sorts of mayhem in the cell; different genes get activated and cell physiology changes dramatically. "In order to understand the process, we need to understand the chemistry underlying how enzymes recognise lesions."
Using the knowledge
Thwarting the DNA repair machinery can be a useful anti-cancer strategy. The drug, cisplatin, used in cancer chemotherapy, interferes with normal cellular processes by causing DNA lesions. These lesions form crosslinks between DNA bases that can interfere with several cellular processes. The DNA repair or surveillance system kicks in and cells are killed off. While the drug helps get rid of cancer cells, normal cells suffer too, which is why patients receiving chemotherapy lose all their hair.
Big problems arise when cancer cells find a way around the drug. "For instance, cancer cells try to make polymerases that can replicate over cisplatin," Karl-Peter explains, so they can survive drug treatment, because the DNA repair checks are evaded, and cells aren't sent to the recycle bin. Structural biologists want to understand how such enzymes can recognise the cisplatin lesion. What is the active site chemistry of the interaction? "Understanding this", explains Karl-Peter, "you could design other chemotherapeutics that this polymerase cannot replicate over. This is the kind of contribution that we could make."
Deciphering the world within
So how does Karl-Peter's work provide a basis for new treatments? "We provide the toolbox for modulating enzyme function," he confirms. "Take this polymerase, for example. If you have the crystal structure you can start to think about how to design a drug that can interact with it."To get the crystal structure, you need to purify the protein. In fact, around 90% of his time is spent trying to purify proteins to 'grow' crystals for analysis.
 "We try to derive a three-dimensional picture of all the atoms in a molecule like an enzyme. The most powerful and best way is to use X-ray crystallography. We use the interaction of X-ray photons with electrons in a molecule." X-rays (photons) are literally bombarded against the crystal and are diffracted off the electron 'clouds' surrounding atoms.
As Karl-Peter explains, collecting these diffraction patterns is a task in itself. "In order to resolve the distances between atoms, you need a means to measure the waves between atoms, which are separated by distances of around 1 angstrom", a hundred-billionth of a meter. There is currently no such thing as an X-ray microscope because of the difficulty in making a lens that can capture the tiny X-ray waves.
"So why do we need a crystal? We use diffraction patterns from millions of molecules because we can't reconstruct the patterns through a lens. We need to grow crystals, because the atoms are all orientated in the same way." Millions of molecules are needed to get enough 'signal' for the structural information. Owing to the lens problem, "half of the information (phases of light) is lost, so we need to run computer programmes to get the lost phases back." Another difficult job. In fact, admits Karl-Peter, "this is often a PhD thesis!"
Once diffracted waves are run through a computer programme, "We get the electron density distribution and model the protein structure, something we do nowadays at synchrotrons." A synchrotron is an enormous electron accelerator, which is used to produce high-intensity X-ray beams. There is one in Grenoble, an impressive 1 km across, which cost over a billion euros to build. Synchrotrons "accelerate electrons to almost the speed of light."
Although there are several in Europe, you need to apply for time to use them. This is often funded by a grant for one day's worth of measurement. Karl-Peter and his team make regular trips to Grenoble carrying their little 0.1mm wide crystals in liquid nitrogen. The one disadvantage of the X-ray method of deciphering protein structure is that, "you need lots and lots of protein." Although 10mg doesn't sound like much, "it takes a lot of time to purify this much protein." Also, you often want to verify X-ray crystallographic models by complementary analysis. For instance, electron-microscopy is useful, because you need 100-fold less protein.
Through deciphering the molecular architecture of the cell, We can really envision the chemistry of DNA repair." With a clear picture of the tools used by cells we can improve therapeutic strategies. As Karl-Peter notes, 'slash-hammer' methods of killing cancer cells are not specific and kill every cell that divides. Harnessing the ingenuity of viruses, for example, which naturally reprogramme cells, holds promise, but without having a clear model of the molecular infrastructure, strategies will remain unrefined.
Texts by Brona McVittie, Science Writer, London, UK.
| 
