How was the nucleus formed first

Inside the nucleus

Light microscopy created the basis for research into the smallest units of life. Today, the combination of sophisticated coloring methods with super-resolution microscopy and 3D computer reconstructions allows a direct view of the spatial shape of the chromosomes in the cell nucleus, even individual gene locations become visible. Once again, scientists have to revise their ideas: Chromosomes don't fill the nucleus like tangled cables. And they change shape, not volume, when they are active. At the Institute for Applied Physics, Christoph Cremer and his team are working on the further development of display methods with which medical research could gain fundamental new knowledge, for example about the development of tumors.

Around three hundred years ago, the Englishman Robert Hooke invented the compound light microscope with which he could see subtleties that had hitherto remained hidden to the naked eye: strange subdivisions in plants, which he called "cellulae". At the time, however, no one could say what they meant. More detailed investigations were not possible because of the extremely poor image quality of the early microscopes. Microscopy resembled "reading in coffee grounds", and not much was missing, and the French Academy of Sciences, the most respected scientific authority of the time, would have condemned microscopy as unscientific, "since everyone could see what he wanted with it". At the beginning of the last century, the instruments were not much better, so that Goethe was almost right in his judgment that microscopes only confuse. However, during his lifetime there was a tremendous change. Light microscopes with significantly better image quality made a fundamental discovery possible, without which today's biology and medicine would be inconceivable: the knowledge that all living organisms, plants, animals and people are made up of tiny living subunits, namely cells. At first, the cells, which were a few thousandths of a millimeter in size, appeared to be quite simple structures, but the impression turned out to be completely wrong. In the middle of the last century it was discovered that most cells have a nucleus that is clearly different from the rest, the cytoplasm. It was only absent in dividing cells, in which instead thread-like structures about one to several micrometers in length became visible, the chromosomes. They had to be in the nucleus somehow, even if you couldn't see them there. Today we know that the chromosomes contain almost all of the information that is important for the life of the cells and for the structure of the living beings formed from them. In human cells, for example, genetic information is divided into 46 individual chromosomes. In each of them, the information is stored in a huge DNA molecule that would be several centimeters long when stretched out. The total genetic mass of a cell would correspond to around two meters.

How can 46 DNA "cables" with a total length of two meters be accommodated in a space like the cell nucleus, which is around two hundred thousand times smaller in diameter? Since until recently it was not possible to make the DNA threads of a single chromosome in the cell nucleus visible in the microscope with the help of chemical dyes, speculation flourished for around a century. Basically, biologists considered two explanations possible: One assumed that the long strands of DNA ran through a large part of the cell nucleus, similar to the tangled cables in an old telephone exchange. The other assumed that the DNA thread of every single chromosome was wound up in a kind of knot in the cell nucleus; the individual chromosomes only take up a small partial volume of the cell nucleus. Only a few years ago it was possible to make individual chromosomes in the cell nucleus directly visible with a light microscope and thus to decide the hundred year old controversial question of cell biology in favor of the second possibility. Again with the help of a new process, the "chromosomal suppression hybridization", in which chromosomes are marked with fluorescent dyes - also clearly called "chromosome painting" (see the article by Thomas Cremer and co-workers in Ruperto Carola 1/93).

Chromosomes are genetically active only in the cell nucleus, i.e. outside the division phase; But even here, depending on the cell type, only certain genes are read, only in this way a highly complex organism like the human body can be formed and maintained. Even entire chromosomes can be genetically "switched off", for example in women one of the two female sex chromosomes, the X chromosomes. In the nuclei of a particular cell, only one of the two X chromosomes is active. A generally accepted assumption is that the genetically inactive chromosome is many times more condensed than the genetically active one. Both chromosomes would have to occupy a different volume in the cell nucleus. The development of the "chromosome painting" method made it possible for the first time to test this concept, which was already several decades old, experimentally. To do this, it was necessary to determine the volumes of the active and inactive X chromosomes after staining using three-dimensional microscopy techniques.

How it works is illustrated by an example from photography: If you want to record the blossom of a cherry tree branch with a near-field lens, you have to bring it within a certain distance - for example between ten and twelve centimeters - otherwise the image will be blurred. The shallow depth of field can also be used to obtain a three-dimensional image of the entire branch by bringing the parts of the object one behind the other into the field of focus of the lens and taking one image at a time. Each time only a part of the flowers is sharply depicted. This process is known as "optical cutting". Put together all photos, all optical sections, result in a three-dimensional image of the branch. The smaller the depth of field, the more accurate the image will be. Optical sections also form the basis of three-dimensional microscopy. The different levels of the object, in our case a cell nucleus with specifically stained chromosomes, are brought into the focus area of ​​the microscope objective one after the other and "photographed". With conventional microscopes, which are essentially still constructed like the one developed by Robert Hooke, the optimum depth of field is about one micrometer, i.e. the spatial resolution is also about one micrometer. However, the chromosome territories in the cell nucleus that we want to investigate are only about three times the diameter of this size, so a useful determination of the volume is hardly possible.

In order to further reduce the depth of field and thus improve the distance resolution and the determination of volume, fundamentally new ways of high-resolution light microscopy had to be found. Based on the experimental experience that Thomas Cremer and I had gained in the construction and biological application of laser micro-irradiation equipment, we proposed the following procedure in 1978: The cell nucleus is scanned point by point by a strongly focused laser beam and stimulated to emit fluorescent light. An extremely sensitive light meter registers the fluorescent light point by point and composes the picture from it, similar to a television picture. The crucial "trick" is that there is a small hole in front of the light meter, a screen. The diameter of the diaphragm is chosen so that it largely only lets through light from the respective illuminated object point. This should make it possible to determine the spatial extent of small biological objects more precisely. Working groups at the European Laboratory for Molecular Biology in Heidelberg, EMBL, and at the University of Amsterdam realized a similar method a few years later, which they called "confocal laser scanning fluorescence microscopy". The optical sections obtained with such a microscope are approximately twice as "sharp", that is, the spatial distance resolution is now only about half a micrometer and thus only about a sixth of the diameter of a chromosome territory. In close cooperation with the pioneers of the new high-resolution microscopy at EMBL, members of Thomas Cremer's group at the Institute of Human Genetics and Anthropology and my own group have used confocal fluorescence microscopy to measure the volume and shape of the active and inactive X chromosomes in female human cells to be determined quantitatively. First of all, we recorded optical sections from each cell nucleus and then measured them using three-dimensional computer methods. We were very surprised when we saw that the active and inactive X chromosomes - contrary to popular belief - had almost the same volume. Various experiments and computer evaluation methods repeatedly confirmed this strange result. That meant that the degree of condensation, the density, of the chromosomes was not the decisive factor for their genetic activity.

Which size was decisive then? On the basis of other experimental observations and theoretical considerations, Thomas Cremer, Peter Lichter, German Cancer Research Center, and the author recently developed a new model for the structure of chromosomes in the cell nucleus. According to our model, the genetically active genes are on the surface of the chromosome territories. Inactive genes can be moved inside the chromosome. The surfaces of the chromosomes repel each other due to their negative electrical charge. As a result, a narrow space is created between them, in which the sometimes very large molecules can move, which are important for reading the genes, for their regulation and repair as well as for the replication of the chromosomal DNA. When many such molecules (or viruses) flow in, the initially narrow space can be considerably widened under certain circumstances.

Since different genes are active in different body tissues, the chromosomes would have to have a different spatial structure depending on the cell type and stage of development of the organism. Chromosomes with many active genes should have a larger surface area than those with few active genes. We tested this experimentally with the help of computer reconstruction and found that the active X chromosome has a much larger surface area than its inactive counterpart. Since they do not differ in terms of volume, this is of course only possible if the shape changes. In fact, the active X chromosome is longer and flatter than the more spherical inactive one. The remaining 44 chromosomes, which occur in pairs in every cell, have a very similar genetic activity and should therefore, according to our model, also have a similarly large surface area. To check this, we have fluorescently marked both chromosomes no. 7 in addition to the two X chromosomes. Optical sections and three-dimensional computer reconstruction confirmed our assumption. In order to examine the three-dimensional architecture more closely, we wanted to determine the location of certain genes in a chromosome. This is not only of fundamental scientific importance for the biology of the cell in general, but also for basic medical research. For example, our considerations have an immediate consequence for the location of "cancer genes", so-called oncogenes, which - especially if they are present in many copies - can lead to the unhindered division of a cell. Evil growth should only occur in cells with oncogenes on the surface of the chromosome. How the spatial arrangement of active genes actually looks and how it comes about is, however, completely unknown up to now. The assertion that active genes are located on the chromosome surface is currently being tested experimentally using the "c-myc oncogene", which is of central importance in many malignant cancers. The evaluation of our 3D reconstructions is not yet complete, but one thing is already clear: It will be very difficult to clearly decide which genes are directly on the surface and which are hidden inside. The spatial resolution of the current confocal microscope is still too limited. The methods described so far can only distinguish points that are not more than half a micrometer away from each other in the direction of the microscope axis and about 0.2 micrometers away from each other in the horizontal direction. This is sufficient to obtain initial information about the volume and shape of a territory, but precise explorations of the fine structure are very tedious. It would therefore be extremely desirable to further improve the three-dimensional resolution.

This could be achieved with so-called far-field microscopes. This would allow relatively "thick" objects, such as cell nuclei with a diameter of around ten micrometers, to be examined with a greatly increased spatial resolution without having to cut them into wafer-thin slices as before, and with "multifluorescence hybridization" it would be possible to have many at the same time Mark locations specifically with different colors. Such a requirement is easy to make; In every physics textbook, however, one can read that, because of the wave nature of light, a resolution higher than about 0.2 micrometers cannot be achieved with far-field microscopes. This is also true for microscope types, as they were known by Ernst Abbe, who around a hundred years ago in Jena put forward fundamental considerations about the resolving power of microscopes. But Ernst Abbe knew neither laser light nor holograms. If laser light is allowed to strike suitable holograms from all sides, it should in principle be possible to focus light much more strongly than with conventional lens systems.

In the 1970s Thomas Cremer and I published a proposal how this theoretical possibility could be used to build a confocal laser scanning microscope with increased resolution. We imagined letting laser light fall from all sides onto a hologram, and thus achieving a much sharper focus of the light at one point than with a normal microscope objective. The fluorescent light emanating from the object point in the focus would then be registered in a manner similar to that of a conventional confocal microscope. In order to obtain a three-dimensional image of the object, it should be pushed through the focus of the "4pi hologram" point by point.

Such a "4pi holographic microscope" has not yet been realized. Instead, the goal of an improved resolution has been approached in a different way. As part of his dissertation at the Institute for Applied Physics with Siegfried Hunklinger, Stefan Hell thought that a "4pi microscope" could be realized by bundling laser light simultaneously through two opposing microscope lenses onto a common focus. The light emanating from the object point located there is registered confocally with a light detector. A three-dimensional image is created when the object is moved point by point through the laser focus. Stefan Hell calculated that in this way the resolution in the direction of the optical axis, i.e. the center line through the two lenses, could be 0.1 micrometers and less, instead of at best 0.5 micrometers as before. Following his doctoral thesis, he was able to test his idea in initial experiments in Ernst Stelzer's group at EMBL in Heidelberg. With the first prototypes of the new microscope, he achieved experimental resolutions of 0.075 micrometers in the direction of the optical axis. Further theoretical considerations, which he is now making at the University of Turku in Finland, show that a value that is even smaller by a factor of two should be possible. In the super-resolution light microscopes he designed, the high resolution is achieved either in the direction of the optical axis or perpendicular to it. However, an increased resolution in all three spatial directions at the same time is of great importance in order to determine the spatial position of gene locations to one another with sufficient accuracy. One possibility to achieve this with the light microscopic methods presented so far is to rotate the object in a suitable manner and then to combine the images recorded under the different observation angles into a stereo image with increased spatial resolution. Similarly, we improve the "resolving power" of our eyes many times over in daily life by turning an object or walking around it, i.e. registering it from different viewing angles.

We try to implement this principle in "axial tomographic microscopy". There are still many detailed problems to be overcome; At present, the cell nuclei to be examined are placed in a thin quartz capillary one fifth of a millimeter in diameter. The capillary and its contents can be rotated at any angle under a high-performance microscope objective and one level of the object can be imaged with the highest possible resolution. Quantitative measurements show that spatial distance measurements can be carried out with a higher degree of accuracy than with conventional confocal fluorescence microscopy.

In a project of the Institute for Applied Physics applied for at the Deutsche Forschungsgemeinschaft, we plan to combine the axial tomographic rotating mechanism with a "wave field microscope", the optical structure of which is very similar to that of a "4pi microscope". In the long term we hope to achieve a spatial resolution in the range of about 0.04 micrometers by combining axial tomography with "4pi microscopy" and other possibilities of super-resolution microscopy, which corresponds to a tenth of the wavelength of the incident laser light.A "search volume" of the laser focus of 0.04 micrometers in all spatial directions would make it possible, in conjunction with suitable hybridization processes, to obtain hundreds of times more information about the structure of a chromosome territory than before. Even the spatial position of parts of individual genes could be researched. From an exact knowledge of the architecture in the cell nucleus, we expect fundamental new insights into the structure of chromosomes and their evolution, the basics of gene activity and the development of the whole organism. The knowledge of the connection between chromosome architecture and certain diseases, especially the development of malignant tumors, could have far-reaching significance for future medical research.

Numerous diploma students, doctoral students and scientists have contributed to the results described in the text. Particular mention should be made of: Axel Bischoff, Joachim Bradl, Dr. Michael Hausmann, Steffen Lindek, Bernd Rinke, Khan Saracoglu, Eva Weiland (Institute for Applied Physics); Roland Eils, Dr. Norbert Quien (Interdisciplinary Center for Scientific Computing and Graduate College "Modeling and Scientific Computing in Mathematics and Natural Sciences"); Stefan Dietzel, Christine Mefferts, Dr. Evelin Schroeck (Institute for Human Genetics, Group Prof. Thomas Cremer); Dr. Stefan Hell (European Molecular Biology Laboratory (EMBL) / Institute for Applied Physics, and Department for Medical Physics, University of Turku); Dr. Peter Lichter (German Cancer Research Center); as well as Prof. Michel Robert-Nicoud and his group (University of Grenoble).

Prof. Dr. Dr. Christoph Cremer
Institute for Applied Physics, Albert-Ueberle-Str. 3-5, 69120 Heidelberg,
Telephone (06221) 56 93 93