Folding art in the nucleus
Scientists from Berlin and Naples develop methods for predictions about the three-dimensional structure of DNA
The medical scientist Stefan Mundlos and his team are interested in the molecular causes of congenital malformations, especially of the skeletal system. In "brachydactyly", for example, the bones of the fingers are severely shortened; in "syndactyly", the bones of different fingers are fused. "Only until recently, scientists thought that such changes could only be caused by mutations within genes. However, the genes of many patients were completely unchanged. That's why we couldn’t get a picture about how these diseases actually occur," says Mundlos. "But now, we start to understand what is happening." Using innovative technologies to analyze chromatin, the scientists found out that the malformations were obviously based on a previously unknown type of genetic defect.
Chromatin consists of the genetic material DNA and the proteins associated to it. Strung together, the DNA molecules of the chromosomes would correspond to a thread with the length of two meters. This needs to fold in order to fit into the nucleus with a size of only a thousandth part of a millimeter. Based on the results of new analysis approaches, scientists assume that the folding process is guided by a genetic master program that determines the exact three-dimensional structure of the chromatin.
The three-dimensional folding of the DNA does not only serve the compact packaging, but also influences the regulation of gene activity directly. The folding of the linear DNA results in spatially separated regions, the so-called TADs (topologically associated domains). Simultaneously, it causes a close proximity of exactly defined DNA regions that are far apart from each other on the elongated DNA thread. "Inside a TAD, the chromatin folds very strongly and the elements within it interact intensely, while there are little to no interactions between different TADs," Mundlos explains.
Using elaborate Hi-C technology, the team of the MPIMG Research Group "Development & Disease" quantified the parts of the chromosomes with many contacts to each other, due to the 3D-folding of the DNA. Based on these “contact points”, the entire 3D structure of the DNA could be calculated and visualized. Meanwhile, the scientists assume that the cell is folding its DNA in a highly controlled manner, because “the resulting contacts between distant DNA regions are necessary to switch genes on or off”, Mundlos says. Therefore, the correct 3D structure is a prerequisite for a controlled gene regulation.
One kind of tools to switch genes on or off are so-called enhancers. Enhancers are DNA regions that increase the activity of genes exactly when needed. They are distributed in large quantities across the genome and could be located between or within genes. With correct folding of the DNA, enhancers which might actually be located in great distance will be moved to immediate vicinity to the control regions of genes. Thus, only the organization of the DNA in TADs as functional units will determine which enhancer belongs to which gene.
To find out whether large structural variations could actually alter the configuration of TADs and thus cause disease, the MPIMG researchers used the example of hand malformations in human and mice caused by mutations in the region of the EPHA4 gene. Their results: In brachydactyly, DNA segments neighboring the gene are missing, while in syndactyly a duplication and an inversion of DNA segments occurs. The gene itself remained unchanged. But interestingly, says Mundlos, "all alterations affect the boundaries of the TAD containing the EPHA4 gene." Their new finding: These malformations develop, because the mutations disarrange the three-dimensional structure of the chromatin, resulting in “wrong neighborhoods”, which normally wouldn’t occur. Subsequently, the wrong enhancer elements can switch on muscle genes like PAX3 during embryonic development at the wrong time and in the wrong tissue, thus resulting in malformations.
However, without the elaborate Hi-C technology, it is difficult to predict, whether a mutation firstly induces a wrong contact point, secondly, if this changes the regulation of the gene, and thirdly, if misregulation causes a disease. But this is exactly, what physicians like Stefan Mundlos, who also heads the Institute of Medical Genetics and Human Genetics at the Charitè – Universitätsmedizin Berlin, are interested in. Searching for standardized prediction methods, the Berlin group teamed up with Mario Nicodemi from the University of Naples. The specialist for theoretical physics in molecular biology takes DNA as a polymer, a molecule composed of many repeated subunits. The folding of polymers follows specific physical laws so that it can be modeled in the computer. Nicodemi designed a computer model for the chromatin. In this model, it is possible to ad libitum remove, duplicate or invert segments of DNA – and to observe how the three-dimensional folding and the contact points between DNA regions change and which effects follow. In experiments with cells from humans and mice, the researchers tested, whether Nicodemi's predictions are correct. "And they are very good," summarizes Mundlos the result of the new study, published in Nature Genetics.
Chromosome mutations are common findings in routine genetic diagnostics. However, it is difficult to predict, which changes cause disease and which don’t. Calculating the folding of chromosomes in silico can provide valuable information for the interpretation of these changes, thereby improving the diagnosis and genetic consultation of affected patients and their families.