Molecular goldsmith

São Paulo network overcomes the high risk of a fiasco and clarifies the structure of 52 proteins

Shaker Chuck Farah, a Canadian biochemist who has lived in Sao Paulo for the last fifteen years, points out on the computer screen a group of concentric circles defined by a subtle variation of gray tones. Black points speckle in the center of the monitor and form an image that reminds one of a target riddled with bullets. This is the light signature of a protein produced by Xanthomonas axonopodis pv citri, the bacterium that produces citrus canker egated chlorosis (CVC), one of the worst pests in the national citrus industry. On passing through a crystal of the protein, a beam of X-rays undergoes deviation and registers on the detector points that permit the identification of the 3-dimensional structure of this molecule. By way of this technique known as X-ray diffraction, the biochemist’s team at the University of Sao Paulo (USP) have defined, atom by atom, the structure in relief of YaeQ: its 182 essential blocks (amino acids) that group together and take the shape of a barrel without lids and with a gap at the top.

The YaeQ is only one of the 52 molecules whose spatial structure was described over the last two years by researchers from the Structural Molecular Biology Network (Smolbnet). Established in December of 2000 under the coordination of the biochemist Rogerio Meneghini, with FAPESP’s support and that of the National Synchrotron Light Laboratory (LNLS), this network is now beginning to reap the main results of the analysis of proteins after four and a half years of work. And, firstly it was necessary to train biologists, biochemists and medical doctors from the twenty São Paulo laboratories in the techniques adopted to discover the 3-dimensional form of these molecules, essential for the composition and the working of living organisms.

Two characteristics of this network made the task more difficult. The first was that the teams exercised distinct activities one from the other. The group led by the medical doctor Ismael Silva, for example, studied proteins linked to the emergence of gynecological cancers, whilst the team headed by the biologist Luis Eduardo Soares Netto investigated a group of proteins that protect the cells from free radicals. As well, the biologist Carla Columbano de Oliveira analyzed the structure of the protein components of a conglomerate – the exosome – existing in the nucleus of cells, responsible for the quality control of the molecules of ribonucleic acid (RNA) that, among other tasks, copy the information of genes and guide the production of other proteins. The second complication was that almost all of the groups presented little or no knowledge about the techniques of the determination of protein structures. “Our objective was to teach the process of investigation of the 3-dimensional form of proteins to the research teams that they frequently come across with the need to know the structure of these molecules”, says Meneghini, who from 1997 until 2004 directed the Structural Molecular Biology Center (Cebime) of the LNLS.

Multiple techniques
The effort was worth it. Over the last two years the different teams determined the structure of 24 proteins through X-ray diffusion – the main technique used in the study of the crystals of these molecules, protein crystallography. Among them, there are two important proteins that inhibit in different stages the coagulation of blood, and, in the future, could well be applied in the treatment of a heart attack or a thrombosis. The researchers also arrived at the 3-dimensional form of a further 14 proteins with a second technique that uses X-rays, but is less precise than the previously mentioned technique: Small Angle X-Ray Scattering (SAXS). By way of a third technique, nuclear magnetic resonance (NMR), they have unveiled the structure of a further 14 molecules.

Presented in fifty articles published in international magazines, these results were considered extremely good during the evaluation carried out in October of 2004 by an independent commission, which included specialists from the Pasteur Institute, in France, the University of Oxford, in the United Kingdom, and from the National Brookhaven Laboratory, in the United States. “The level of success from the production of clones up to the determination of the structures is comparable with those of international projects”, the foreign evaluators stated. They also reinforced the need to install at the LNLS another line of X-rays, which supplies more intense radiation in specific wavelengths. Essential in the identification of the structure of unknown proteins, this line, named MAD (Multi-wavelength Anomalous Dispersion), should enter into operation by the end of this year.

It really was necessary to establish a network such as this one. Up until the end of the decade of the 1990s one count could on the fingers of one hand the Brazilian laboratories dedicated to protein crystallography, with the highlight being the group headed by the crystallographer Glaucius Oliva, from the Physics Institute of Sao Carlos, linked to USP, and the Cebime itself. Reducing this backwardness, the teams of the 20 laboratories on their own are capable of carrying out five of the six stages of the process. “On being judged by the results obtained, this program gave an important stimulus to the development of this area in the country”, comments Meneghini, who currently is a researcher at the Latin American and Caribbean Center on Health Science Information (Bireme).

Overshadowed by the notoriety conquered by the sequencing of the genome of 180 organisms, crystallography is now re-assuming throughout the world the primordial role that it performed in the study of the function of proteins. The explanation: the spatial configuration of these molecules determines the role that it will exercise. A protein is produced in the interior of cells through the growth of one amino acid upon another, in a long string essentially composed of the atoms of hydrogen, carbon, nitrogen and oxygen. As soon as it is formed, this string begins to suffer torsions and doubles over itself, assuming a spatial formation because of the attraction and repulsion exercised by the electrical charge on determined stretches of the molecule.

The result of this solitary dance are molecules in the form of a string, of a ball or of an hour-glass. Like a key that opens up only one lock, the proteins show a spatial structure so specific that in general they only integrate with other molecules in a complementary form. “It was the 3-dimensional form of DNA that indicated to James Watson and Francis Crick how this molecule behaved”, comments Farah. And what does the form of a barrel with a gap count for in the eyes of the  biologist from USP with respect to the working of this protein? Farah still does not know. A comparison with other proteins suggests that the YaeQ is something completely different from what is currently known. “This is our next challenge”, says the USP biologist. “We’re following the pathway opposite to that traditionally trodden”, says Farah. In general, crystallography is used to unveil the structure of proteins with unknown functions, but, with the termination of various genomes, the number of proteins whose functions were still not identified grew.

Successes and frustrations
More artesian than the sequencing of genomes, the determination of the 3-dimensional form of proteins is a torturous task, almost always made up of six stages and with a low level of success, on average 5%. The problem is that every stage presents hurdles, with an extra difficulty: there is no way of anticipating which stage will go wrong nor for what reason it does not work, marking a return to the empirical tests of trial and error. The first stage is to choose the gene responsible for the production of a protein and to discover the sequence of the pairs of nitrogenous bases (adenine, thymine, cytosine and guanine) of which they are made. Next, there is an attempt to copy the gene of the protein from which it is desired to obtain a crystal.

When everything runs well, the researchers insert the chosen gene into genetic material of a bacterium or a yeast, which should produce the protein in a quantity sufficient for the following stages. Through the work done by these microorganisms, there is nothing to do apart from waiting. The next stage is to separate the protein to be studied from that manufactured by the bacterium or by the yeast – the filtration can take weeks. The few milligrams of the purified protein are diluted into different salt and alcohol concentrations, deposited in dozens of closed recipients and distributed over rectangular acrylic plates, of size a little more than a packet of cigarettes. And once again it is necessary to wait.

In a process, little understood by the researchers, the millions of copies of the protein begin to lose water and move closer to each other, forming a block in a solid state. With a bit of luck, the protein molecules set themselves up at the same distance from each other and in the same direction, much like a squadron of soldiers lining up when awaiting their commander’s instructions. This is the protein crystal. However, if the distances between one molecule and the next are irregular or the direction of the copied proteins is not homogenous, a deformed solid comes about in the bottom of the recipient. And there is no absolute recipe. “Each protein needs specific conditions – a bit more of the determined salt or a bit less of a certain alcohol – in order to form the crystal”, comments the biologist Andrea Balan, from the Biomedical Sciences Institute of USP, who, in partnership with Luis Carlos Ferreira, investigated the spatial form of a grouping of 18 proteins of Xanthomonas citri involved in the transportation of nutrients to the interior of the bacterium.

A jigsaw puzzle
Only then do the diffraction tests at the LNLS, in Campinas, begin. In an unimposing concrete building, powerful magnets accelerate electrons at a velocity close to that of light in the interior of a circular tunnel. Each time they are deviated in order to maintain a circular trajectory, the electrons liberate a highly intense light, synchrotron light, composed of radiations that go from infrared to gamma rays. In a laboratory connected to this ring of the synchrotron light, the researchers control, by way of a computer, the direction of the crystal exposed to the X-rays and carry out hundreds of images similar to those exhibited by Farah, each one at a different angle. A computer program analyzes the images and generates an outline of the protein. Knowing the sequences of the protein’s amino acids, the researchers initiate a real game of a jigsaw puzzle that can take months of work: they experiment, one by one, the amino acids until they discover their specific position in the coiled up protein.

What is complicated today had previously been much more difficult. In 1937, when the Austrian biochemist Max Ferdinand Perutz began to make use of X-ray diffraction to investigate the structure of hemoglobin, the protein that transports oxygen in the blood, there were no computers. The work was all manual: the images were made with X-ray equipment much less powerful, impressed upon glass plates and set in aligned frames in the laboratory before calculating on paper the atoms’ positions. Perutz was determined to reveal the 3-dimensional form of hemoglobin during his doctorate at the prestigious Cavendish Laboratory, of Cambridge University, but did not manage it. The spatial structure of hemoglobin was only clarified in 1959, almost two decades after the completion of his doctorate degree, a piece of research that won him the Nobel Prize in Chemistry in 1962, sharing the prize with the British researcher John Kendrew.

Following the recipe – but with the help of computers –, Farah’s team studied some 35 of the 1,700 genes of Xanthomonas citri responsible for the manufacture of proteins as yet unknown and determined the form of two proteins by way of X-ray diffraction. Half way along the road, however, the USP biochemist had the luck meeting up with Ana Paula Valente and Fábio Almeida, both from the Federal University of Rio de Janeiro (UFRJ), which had developed a form of protein analysis by way of nuclear magnetic resonance (NMR) without passing through the stage of purification. Farah adapted the use of NMR in order to analyze the stability of proteins – when more stable, or that is, doubled over themselves, but they easily form crystals – and saved months of work that possibly might not have been successful.

Other teams took different paths. Instead of starting from groups of proteins of a microorganism in order to identify the structure of a few of them, leaving aside what did not go well, they chose to investigate the spatial form of proteins such as those that have already been worked upon. This was the decision of the biologist Sergio Schenkman, from the Federal University of Sao Paulo (UNIFESP), who for some twenty years study the proteins of the life cycle of the protozoa Trypanosoma cruzi, responsible for Chagas’s disease, and of its transmitter in South America, the insect Triatoma infestans, or the kissing bug. Biologist Schenkman’s team detailed the structure of two proteins within the digestive system of the kissing bug that act in distinct phases of the sequence of reactions in blood coagulation: one inhibits the action of thrombin and the other impedes the working of factor XIIa. Both present potential applications in the treatment of problems brought about by the coagulation of blood, such as a thrombosis. “These proteins must impede the coagulation of blood in the digestive tract of the insect, which feeds itself once every 15 or 20 days”, the researcher explains.

Via nuclear magnetic resonance, Schenkman has identified the spatial structure of a third molecule, corresponding to the terminal portion of trialisina, a protein of the kissing bug that opens the cell’s pores and kills it. Possibly the insect injects this protein into the skin at the moment of the bite, opening the way for sucking the blood. Analysis from the UNIFESP team and LNLS have permitted the mapping of regions of the terminal stretch of the trialisina important for the activities of this molecule. Based on this information, the researchers believe that it will be possible to design peptides with anti-microbe action. Different from genomic studies, the crystallography of proteins is advancing one step at a time.

The Project
Structural Molecular Biology Network (Smolbnet); Coordinator
Rogerio Meneghini – Bireme; Investment R$ 13,036,329.12 (FAPESP)