Genes and their metabolites

The company metanomics determines metabolite profiles and feeds them into enormous databases. These profiles open up completely new perspectives in the fields of plant biotechnology, medicine and pharmacological research.

Text: Catarina Pietschmann

Zoom 1294678240

They came up with an unusual idea and found a way to put it into practice. They provided the proof of concept – and it worked. And then they did the necessary calculations: How long would it take to switch off, individually, all of the approximately 30,000 genes found in thale cress, Arabidopsis thaliana, and to see how its metabolism changes as a result? Even the most optimistic projections indicated that it would take years. So now what?

In 1996, Lothar Willmitzer, Director at the Max Planck Institute of Molecular Plant Physiology in Golm, near Potsdam, found himself confronted with an awkward question: Is it really the task of a basic research institute like Max Planck to develop a high-throughput process? The answer was a clear no. The parallelization of processes is the concern of industry, he decided. And thus the idea of setting up a company was born.

What is so interesting about plant metabolism? If humans and plants have one thing in common, it is the fact that they both have rather complicated metabolic systems. The human metabolic system has been well researched, mainly because changes in this system can cause serious illnesses or, conversely, be caused by them, as in the case of diabetes.

In addition, the human metabolism does not have the variety of secondary compounds found in plants. Far less is known about metabolic processes of plants. To complicate matters further, there is enormous species diversity in the plant world, and the metabolites can differ significantly from species to species: some produce high-quality oils, others are rich in vitamins and flavors, while yet others mainly form sugar or starch. Which is a good thing, since plants are the primary basis of our nutrition, and life would be miserable if corn-on-the-cob, thyme, strawberries and grapefruit all tasted the same. It would be possible to tell the difference between garlic and vanilla based on their physical appearance, but not from their smell.

The consequences do not bear thinking about: the cuisine of Thailand would be the same as that of Italy. Restaurants would not exist – what for? Eating would not be a pleasurable activity, but merely a means of nutrition intake – an annoying but necessary process. Instead of grocery stores, we might have depots everywhere that we would have to stop by regularly to cheerlessly consume some nondescript pureed mush. Evolution might even have equipped us with a practical proboscis for the purpose. Fortunately, things turned out differently: a strawberry is a strawberry. And its contents are completely different in composition than those of a kernel of corn.

Metabolite profiling reveiling all the genes metabolites

The nutrient content, flavor, shape, color, size and yield of a plant can be influenced through breeding. This is, however, a protracted process. Too protracted to enable the resolution of the problem of world hunger against the backdrop of the ever-increasing global population. Plant biotechnology facilitates much faster development of crop plants that are more productive, nutrient-rich, or even stress-resistant – which means that they can thrive on barren or saliferous soils, withstand longer periods of drought, or get by with very little light.

Original 1294684574
With a particular gene missing, knockout plants are no longer able to produce certain metabolites (left). A comparison with a wild-type (right), show the metabolites of the deactivated gene.
With a particular gene missing, knockout plants are no longer able to produce certain metabolites (left). A comparison with a wild-type (right), show the metabolites of the deactivated gene.

But why work through a plant’s entire genome for this? The Max Planck Institute of Molecular Plant Physiology had been established just shortly before all this happened, in 1994. Its purpose was to test the biosynthetic pathways in plants in order to understand how the formation and transport of metabolic products and their storage in the leaf, flower or fruit work. “At the time, we concentrated on the biosynthesis of starch in the potato, and identified genes that play an important role in that process,” explains Willmitzer. The researchers altered the activity of this gene in the hope that the next generation of plants would produce more and higher-quality starch. But they generated less starch and did other nonsensical things as well. But why?

“Research at the frontiers of knowledge” is the motto of the Max Planck Society. And this was just such a frontier. “Genes influence each other reciprocally. We realized that, in order to make any progress, we would first have to characterize a plant completely at the metabolic level.” The complete sequencing of the genome of Arabidopsis, the inconspicuous weed that biologists use as a model, would soon be carried out. And the most gripping question in science at the time was: Which gene performs which function? “We thought, ‘That’s great. We’ll be able to see how the metabolism changes if we switch off each gene individually,’” recalls Willmitzer.

Up to then, it was standard practice to fully work through all biochemical levels, from the gene to the DNA and the messenger RNA, and from protein biosynthesis right down to the metabolic products. Willmitzer and his team, however, embarked on a new path that was not only shorter, but that also yielded, as it would turn out, far more comprehensive insights. They started at the top – with the gene – and looked only at what came out at the end. That was the chaotic cocktail of metabolites: sugar, sugar alcohols, amino acids, fatty acids, vitamins, flavors and colorings, and much more.

Original 1294678242

The procedure sounds relatively easy: the relevant gene is switched off using molecular biology methods. To this end, it is replicated and smuggled into the flowering plant with the help of a transporter, a bacterium. A small leaf from the daughter plant, which now carries the switched-off gene, is later homogenized, the contents extracted and, finally, analyzed using a combination of gas chromatography and mass spectrometry. Process completed. The result is a diagram with more than 350 peaks representing an equal number of substances. There is also a mass spectrum for each peak, from which, for example, the molecular weight can be read.

A paradigm change in the analytical approach was in order – it was not about isolating all of these substances, but about the big picture, a unique metabolite pattern that corresponds precisely to the switched-off gene. Only 60 percent of the substances are even known. But that is irrelevant. If you want to know whether a certain vitamin or special amino acid is among them, all you have to do is include the corresponding reference substance. The process is known as metabolite profiling. The new method was developed.

metagenomics - a Max Planck Society spin-off

So what happened next? The Max Planck Society gave the green light for the spin-off. Willmitzer and Trethewey drew up a business plan in 1997. “Initially, we expected an investment requirement of 30 million deutschmarks for five years.” Then the talks with potential investors began. Willmitzer, who recognized the potential in the method early on, realized that they were still facing a very long process before the concept would actually be used commercially.

Original 1294678243

The company’s first objective had to be the development of a highly efficient technology platform. The second was the systematic analysis of the Arabidopsis genome and the development of the metabolite database. The planning of marketable products had to wait until phase three: the transfer of the acquired expertise to crop plants – that is, testing selected genes in corn, rice, soybeans, rape or cotton – and producing individual packages of knowledge for industrial clients.

Was it courageous risk-taking or foresight? Perhaps a combination of the two: BASF authorized an investment of 50 million deutschmarks. They then approached the Federal Ministry of Education and Research, which also agreed to participate and contributed a few million more for technology development. The Max Planck Society reviewed the agreements, a location was sought, and metanomics was launched in the Charlottenburg neighborhood of Berlin in 1998 with four researchers from the institute in Golm, a few technical assistants (TA), and Arno Krotzky, the Managing Director, who BASF brought into the project.

The Max Planck Society did not want to get in on the act itself. Willmitzer would have liked to have had it on board – as a neutralizer, so to speak. “But it did something else, something quite crucial for us: with the establishment of the institute in Golm, it gave us free rein to research new things and made comprehensive funding available for this purpose.” It is unlikely that they would have made such progress at a university institute: the financial risk was too great. “At the Max Planck Society, we had an environment where this was not an issue.” The mere fact that they were allowed to create a spin-off is something that Willmitzer values highly. “It can’t be done part-time, working in the evening. We were already devoting one full day a week to it from the start.”

Today, metanomics has 110 employees, a third of them scientists. The concept worked. The development of the technology platform progressed rapidly; no new equipment had to be developed, but processes and a lot of software, as most of the processes were to be automated – from cloning, extraction and analysis to feeding the comprehensive files into the database. Willmitzer, who stayed on with metanomics for a time as a consultant, adds, not without a certain pride: “After just one year, they managed to get through almost one hundred genes per week. But not with a hundred people – just one scientist working with four technical assistants!”

The number of genes that are now processed each day is not disclosed. Just this much: metanomics works in two shifts with the help of vigilant robots that send a text message if a problem arises. The analysis laboratory, in which several dozen GC-MS machines work away buzzing and humming, are generally deserted. Chemistry TAs track the work of their fully-automated colleagues on monitors in the quiet office. They enter the laboratory only for maintenance purposes and to exchange the samples.

Original 1294678245

Fifty-five thousand genes have now been tested – both by switching off existing genes and by deploying new ones in Arabidopsis. “The latter process is very interesting, as it allows useful new characteristics to be introduced into the plant. For example, the ability to activate new metabolic pathways or produce new substances,” explains Trethewey. Attention is now focused on crop plants. Why didn’t they work on corn from the outset? For practical reasons: Arabidopsis is small and easy to grow in the greenhouse and, unlike, corn, takes only a few weeks from germination to seed formation. Unfortunately, the genes in different plant species are not identical. The challenge now is to identify the crop plants’ most interesting genes from the existing data and the information provided in the relevant literature and to generate their metabolite profiles.

Every patient has his individual metabolite profile

Apart from the issue of efficiency, how does working in a company differ from working in an institute or university? Trethewey ponders the question briefly. “If we get an unexpected result, we always have to look at it in the context of our objective: are we more likely to achieve our objective now, or would it make sense to go off on a tangent with a different aim? Industry is always highly aware of such processes, but this is not always the case in basic research.”

Original 1294678246

“Not everyone reacts to a drug in the same way. For clinical tests, it would be helpful to know whether or not a subject will be among the responders,” Trethewey explains. This can be tested using a metabolite pattern in the patient’s blood or urine. The procedure is the same – only the sample preparation differs for the red and the green profiling.

“Big or small, healthy or diseased, yellow, green or speckled, the metabolic products are closest to the phenotype,” explains Willmitzer. Especially in medicine, it is important to be able to differentiate between state A and B. “And it has emerged that this can be done surprisingly well using metabolite profiles.” Together with the Charité hospital in Berlin, his Max Planck team has produced profiles of 100 kidney cancer patients – for both the tumor and healthy kidney tissue. “It was no surprise to discover that they can be clearly differentiated using the method,” he explains. However, it was astonishing that just two substances out of hundreds were sufficient to do this. And that these are substances that would never have been considered as cancer markers before. Now numerous other examples involving the use of the method can be found in medicine.

The use of this method could also lead to significant progress in the context of everyday clinical laboratory medicine. “At present, only a few metabolites, like glucose and cholesterol, are observed in blood tests. The diagnostic validity is relatively low. But we see thousands of these compounds! If they are measured regularly, pathological changes can be identified long before they have manifested themselves. “A regular metabolite check by the family doctor? Why not? And healthy patients could be given tips on how to feel even better.

The method is now well established in science. Researchers throughout the world use it to delve deep into the system biology of various organisms. At the Max Planck Institute in Golm, it is used to track the changes in a plant’s metabolism minute by minute – from sunrise to deep in the night, at high temperatures, in frosty conditions or during periods of nutrient deprivation.

Richard Trethewey’s vision for the future also includes, among other things, new drugs that are based on a more detailed understanding of metabolism. “Metabolic analysis holds enormous potential and has historically been underestimated. We are currently experiencing an absolute renaissance in medical science. Despite this, it remains a challenge to complete the cycle from scientific insight to an actual product.”

Whether green or red profiling – the foresightedness of science and industry will pay off. As Victor Hugo said: “There is nothing more powerful than an idea whose time has come.”



Gas chromatography

The technique separates mixtures of substances: Depending on their boiling point and their polarity, different substances are bound for different lengths of time to the special coating of a capillary up to 60 meters in length. The substances can then be analyzed with a mass spectrometer.

Mass spectrometry

The method provides information about the mass, or more precisely the ratio of charge to mass in ionized molecules and their fragments. Depending on this ratio, different particles are accelerated to different speeds in an electric field and separately collide with a detector.


To duplicate a DNA sequence, the sequence is injected into bacteria, it reproduces with the bacteria and is subsequently isolated.


Method of isolating substances from mixtures. To achieve this, plants are chopped up mechanically and the metabolic products washed out with solvents.

Go to Editor View