How to draw genes
Candidate of Chemical Sciences O. BELOKONEVA, "Science and Life" No. 8-2008
It has already been several years since the genome of humans and some other living beings was completely decoded. Decoding the genome has posed an even more difficult task for the scientific community — to understand what functions are performed by sections of DNA called genes. Scientists of the United Europe have created a genetic atlas that will answer this and many other questions of the new stage of genomic research.
In 2003, the human genome was completely decoded. This means that scientists now know the sequence of more than three billion nucleotides in a human DNA molecule. Data on the sequencing of the genome of drosophila, nematodes, E.coli bacteria, mice have been published, work has begun on decoding the genome of chimpanzees. It would seem that knowing the chemical structure of DNA, we will be able to answer all the questions about how genes are arranged. But this is far from the case. The DNA molecule really consists of certain sections — genes responsible for the synthesis of protein molecules. But a gene sequence is not a genome yet. For example, only 25% of the human genome consists of "real" genes. Two-thirds of the genome are regulatory regions, "meaningless" sequences, genes can overlap, one gene is often responsible for the synthesis of several proteins at once, etc. As a result, instead of the 100 thousand human genes predicted in the 90s of the last century, in 2003, scientists identified only about 20 thousand semantic genetic sites. Moreover, it is not always clear where one gene ends and another begins, how these genes work and which proteins are responsible for the synthesis. Therefore, after decoding the genome, an abyss of the unknown opened up to molecular biologists. Now scientists will have to look for meaning in "meaningless" sections of DNA, identify new genes, study the mechanism of regulation of already known genes and determine their functions.
Every cell of the organism of a living being contains the same DNA, the same genes. Nevertheless, the protein composition of a cell, for example, cartilage, is obviously different from the proteins of liver or brain cells. What's the matter? Yes, because not all proteins encoded in the DNA structure are synthesized in the cell, but only the necessary ones. Simplistically, the collagen protein gene is activated (or, as they say, expressed) in connective tissue, but "sleeps" in the liver; the adrenaline gene is "working" in the adrenal glands, but "resting" in brain cells, etc. The study of the function of genes is engaged in functional genomics, which received a huge impetus to development after the decoding of the genome.
At the current stage of development of functional genomics, it has become possible to determine which genes are expressed in different parts of the body — "work" and which "sleep". And not just to determine, but to obtain three—dimensional images of the distribution of gene activity in all organs and tissues - to create so-called atlases of gene expression. Scientists use laboratory mouse embryos as a model, since the mouse genome sequenced in 2004 consists of almost the same number of genes as the human genome. The physiology of mice is similar to that of primates, and human embryonic development and genetic diseases can be modeled on mice.
In 2005, Professor Gregor Eichel developed a special automated technology for creating genetic atlases, which he called "genepaint" (English — to draw a gene). Now Professor Eichel is the director of the Department of Functional Genomics at the Max Planck Institute of Biophysical Chemistry in Göttingen (Germany), where the coordination center of the now pan—European Atlas of gene expression project is located. In the Eureexpress international postgenomic project (http://www.genepaint.org /), in addition to the Göttingen Institute, several research centers participate — in Berlin, Naples, Geneva, Strasbourg. Two more centers — in Edinburgh and Zurich — are forming a database of the genetic atlas. 
The development cycle of the mouse embryo is 19 days. 14.5- and 17-day-old mouse embryos are taken for genetic analysis. They are frozen, fixed in paraffin. Then, the thinnest sections with a thickness of no more than 20 microns are prepared from the obtained samples, which are placed on a conventional preparative glass for examination under a microscope.
How is it possible to get an image of the activity of a particular gene? To do this, an RNA sample of one of 20 thousand mouse genes is synthesized in the laboratory, since the mouse genome has been completely decoded, and a sample of fetal tissue is processed with it. If the gene under study is active in this sample, mRNA molecules complementary to the RNA sample are present in the tissue. As a result of their interaction, the RNA molecule of the sample is firmly "intertwined" with the mRNA of the tissue — hybridization occurs.
The RNA sample is marked with special molecular markers, which, when treated with certain chemical agents, give a color reaction, therefore, at the site of gene expression, the tissue turns blue.
The slice is photographed in a conventional light microscope. However, the slices are too large to photograph them completely, so the sample sections are photographed sequentially. All the received images are assembled into a mosaic image, which after processing is saved in tiff format. A digital photograph of the stained slice, together with metadata — the conditions of the hybridization reaction, the properties of the embryo slice, the structure of the RNA sample — enters the Eurepress database and becomes available to the entire scientific community on the Internet.
The procedure for processing and photographing the slice is completely robotic. The Max Planck Institute of Biophysical Chemistry receives data on the localization of 60 genes per week, in five other institutes — twice as much. From January 2005 to July 2008, 15 thousand genes were scanned - 3/4 of the entire mouse genome.
More than 250 thousand high-resolution images are stored in the Eureexpress computer database. Now the database volume is more than 20 terabytes, and every month the database grows by another terabyte. The size of each image of a single slice can exceed 100 MB. In order to view such images on the Internet, they are stored on the server in a special Zoom Image Server format. The program allows you to dynamically load the areas of interest in the image, so that it becomes possible to increase any section of the slice to its maximum resolution.
The goal of the project is to obtain an atlas of the expression of all 20 thousand genes in the mouse embryo. With the help of such a genetic atlas, it will be possible to determine at what stage of embryo development and in what place a particular gene is active. This is very important for understanding the physiological function of a gene and its corresponding protein. Also, with the help of the atlas, it will be possible to compare the activity of different genes, not only mouse, but also human, in norm and pathology. The new database will certainly advance knowledge in functional genomics and contribute to the identification of human disease genes.
Portal "Eternal youth" www.vechnayamolodost.ru16.10.2008