Cancer as a model case
Hans Lehrach and his colleagues comprise one of the few research teams in the world to work on computer models that aim to reconstruct all processes in a cancer cell and a healthy cell at the same time. Their computers analyse the cells’ countless interactions and chain reactions and derive from them a network of genes, proteins and mRNAs.
Text: Klaus Wilhelm
When Hans Lehrach thinks about his vision of the cancer medicine of the future, it is clear that analysis of tumour proteins has a crucial role to play. "The fact is, the really important biological processes take place at protein level", says the Director at Berlin's Max Planck Institute for Molecular Genetics. Together with oncologists at Charité university hospital, he hopes to fundamentally change cancer treatment. "We want to predict the prospects of success of a treatment for a given patient, and to adjust their therapy on an individual basis", explains Lehrach. Their goal: to either cure the cancer or turn it into a chronic illness.
Proteins play a decisive role in cancer. They are involved in most metabolic processes and signalling pathways in the cell and one of their functions is to control cell division – a finely balanced process that spirals out of control during cancer development. How many of the approx. 10,000 proteins in a cell are involved remains an unanswered question. The building plans for proteins are found in the genes and often, though by no means always, follow the rule of one gene to one protein. When the cell's internal machinery "reads" a gene, as the experts say, a short-lived molecule of messenger RNA (or mRNA) is generated as an intermediary. This acts as a template for the production of the protein in question.
In addition to their own results, new data is constantly flowing into the team's model from scientists around the world, as they garner more information about the genome, transcriptome and proteome of healthy and malignant cells. "So far, though, we can only simulate some of the molecular signalling pathways in the cell", says Lehrach. Thanks to an explosion in technological advances during the last ten years, genes and mRNAs can now be analyzed much faster. New DNA and RNA sequencing methods, for example, reveal even the tiniest genetic mutations in cancer cells. There may be hundreds of these mutations in a single tumour, which explains why no two tumours are the same.
Some mutations affect the proteins directly, making them function less effectively or not at all, while in other cases they become more active than normal. Alternatively, they may be produced in excessive or insufficient quantities.
Proteomics aims to analyze these kinds of differences between cells; but other changes are meaningful as well. "Following their synthesis, proteins are often chemically modified and may be provided with appendages from a phosphate, methyl or acetyl group. This has a major influence on their function", explains Markus Ralser of the Max Planck Institute for Molecular Genetics.
These post-translational modifications are often temporary and flexible. However, even the life expectancy of the proteins themselves varies according to the cell's needs: If a protein is frequently needed, it decays quickly and is dismantled. If a cell does not need certain proteins, on the other hand, these remain in an inactive state for a long period. "So if I only analyze mRNAs, I have no way of knowing which proteins are actually active", says Ralser, who is now also leading a research group in Cambridge, England. Only proteomics can provide the answer.
Proteins, however, are far more difficult to study than genes, and not only because of their more complex structure involving 20 amino acids. Researchers can replicate DNA and RNA almost at will with the help of natural enzymes, and this facilitates analysis. Proteins cannot be replicated, though, so researchers have to make do with unimaginably minute amounts – in the order of thousandths of milligrams. Such scales of magnitude mean that extremely sensitive analytical techniques are required.
Mass spectrometry is one such highly sensitive technique in which enzymes chop all proteins from a tissue sample into about a million peptide fragments, always at the same points. Nano-chromatography is used to separate out these fragments. The peptides are then given an electrical charge and the electrical field of the mass spectrometer deflects them to different paths in function of their mass. By recording the time of flight until the peptides hit the sensor of the mass spectrometer, the scientists calculate the mass of the peptides and assign them to certain proteins. "Mass spectrometry allows us to analyse complex protein mixtures and also to detect a single peptide among a million peptides", stresses Ralser.
On the one hand, the Max Planck scientists in Berlin want to discover as many proteins as possible in their samples of diseased and healthy tissue. They can then determine how many proteins are present and where they occur in the cell. On the other hand, they have to constantly filter out individual proteins and examine them separately – and these two aspects of their work have always hampered each other to date. "In the foreseeable future, we will be able to combine the two approaches", explains Ralser.
Their method now works so well that the Max Planck team recently managed to resolve the issue of what lay behind the world's rarest disease. Ribose-5-phosphate isomerase deficiency is only known to affect a single patient in the world today. He suffers from a degeneration of brain tissue that leads to symptoms of memory loss, movement disorders and below-average mental performance. The disorder is caused by a genetic defect that even the high-tech methods of genomics cannot analyze.
The patient inherited a wholly defective gene for isomerase from his father, along with a second gene from his mother with a peculiar defect: It causes the enzyme to be produced in some body tissues, but not in others - such as the brain. "The enzyme is regulated in different ways in different tissues", says Markus Ralser. Using mRNA, the researchers could only partially infer whether the enzyme was present in the brain; it took mass spectrometry to clarify the situation.
The results of proteomics in cancer research have been similarly instructive in Hans Lehrach’s laboratory. As early as the first half of the 20th century, the famous biochemist Otto Warburg studied the metabolism of cancer tissue and found that almost all cells in malignant tumours consume less oxygen than healthy cells. At the same time, they release more lactic acid - a phenomenon known as the "Warburg effect". Since then, experts have discovered that the enzyme pyruvate kinase is involved in the switch to tumour metabolism. But how? Opinions on this point are divided.
In 2008, it appeared that an answer had finally been found. Two research teams had discovered that pyruvate kinase occurs in different variants in tumours and in healthy tissue. It was a discovery of far-reaching significance: If the tumour variant could be converted back to the healthy form, it might be possible to slow or even stop the growth of the cancer. Two pharmaceutical companies were excited about this idea and began to pursue it, assigning billions of euros to the project.
After a proteomics study by Lehrach’s team, however, the companies shifted their focus. The scientists examined healthy and diseased tissue from 50 renal cancer patients and analysed the respective variants of the pyruvate kinase protein using mass spectrometry. The result was clear: Although the gene in healthy cells differed from the gene in cancer cells, the same protein occurred in both. "This demonstrates the importance of proteomics to medicine", says Ralser.
Since the biochemist's team saved the two firms from a bad investment, he has been collaborating with the companies himself. Hans Lehrach, too, maintains close relationships with industry and has founded a new company with a view to accelerating the implementation of his vision for cancer medicine. He hopes for an extra boost from the planned purchase of a new mass spectrometer that will be able to quantify thousands of proteins with all their modifications at one go.
Working with the Charité Comprehensive Cancer Centre, Lehrach's team is already analysing the genes and proteins of 20 patients with malignant melanoma. They intend to generate computer models of the cancer cells of each patient, complete with all individual molecular alterations and characteristics – digital tumours, in a manner of speaking. The computer will then use all this information to simulate the processes in cancer cells, together with the effects of different medicines. This might enable doctors to calculate the ideal cocktail of drugs, with maximum impact and minimum side effects. It could also lead to the development of new drugs. "For now, we must prove that our approach works", asserts Hans Lehrach. "In five years, we want to have reached the stage of introducing it into medical practice."