Tackling the genetic onset of Down syndrome

Chromosome 21 is the smallest of the 23 pairs of chromosomes in humans, yet it is responsible for Down syndrome—the most common genetic mental retardation. Down syndrome (DS) is caused by the erroneous replication of human chromosome 21 (HC21), which results in three copies of the chromosome instead of the normal pair of two.

The most common cause of this ‘trisomy’ on HC21 is the failure of the chromosome pair to divide in an egg cell—often linked to advanced maternal age. Such an egg cell has two copies of HC21, and when fertilized, accepts another copy of HC21 from the sperm cell, resulting in a total of three, instead of the normal two, copies of the chromosome.

Unlocking the molecular pathology of trisomy 21 is greatly anticipated. Worldwide, it is estimated that up to one in every 700 babies is born with DS, and there are no specific therapeutic treatments. Yet little is known as to what determines the various phenotypes associated with the disorder. Patients typically suffer from neurological and behavioral difficulties, including language delays and attention difficulties, and some also face immunological, digestive and cardiac problems.

The severity of mental retardation differs by patient and age—sometimes the symptoms are alleviated with age, while for others the symptoms become worse, developing into conditions such as Alzheimer’s disease. “We’d like to understand the molecular pathways responsible for the disease so that we can contribute to the development of effective therapies in the future,” says Kazuhiro Yamakawa, head of the Neurogenetics Laboratory at the RIKEN Brain Science Institute in Wako, Saitama.

In fiscal 2008, Yamakawa’s team secured a grant from the President’s Fund under the category of ‘challenging research’ for a two-year project aimed at developing a highly efficient system to generate transgenic mouse lines for DS research. Their goal is to establish a high-throughput system to generate partial-trisomic DS mouse models and to identify the gene or genes responsible for DS features. In the 14-member laboratory, which also studies epilepsy, Yamakawa and four young researchers participate in this exciting project.

Down syndrome mouse models and lines of research

HC21 carries approximately 360 genes and contains the Down syndrome critical region (DSCR) that many researchers believe is key to the occurrence of DS. HC21 is orthologous (similar by shared ancestry) to part of mouse chromosome 16 (MC16), and several fortuitously generated DS mouse models with partial MC16 have been reported, including Ts1Cje, which has approximately 100 genes and is responsible for milder learning defects; and Ts2Cje, which has approximately 140 genes, partially shared with Ts1Cje, and is responsible for learning defects. Both Ts1Cje and Ts2Cje trisomic segments contain DSCR.

Some researchers believe that the imbalance in chromosomal number for HC21 induces DS by affecting the expressions of genes on the overall genome. Yamakawa and several other groups, however, support a different idea that the initial switch for the disorder is the dosage-dependent overexpression of genes located only on HC21. In 2004, Yamakawa’s team found that the expression levels of genes in the trisomic region of MC16 in Ts1Cje mice were increased by 50%, whereas the levels of other genes on other chromosomes or the normal euploid region of MC16 were almost the same as in normal mice1. Yamakawa adds that many of these overexpressed genes may not result in substantive damage, so identification of a few ‘master’ genes is critical. Some of those genes might also work collectively to contribute to the disorder.

In attempts to identify such master genes, several transgenic mouse models overexpressing individual candidate genes in distinct systems have been developed in other laboratories. However, even though some of the mice models displayed DS-like phenotypes, such as learning disability, the expressions of HC21 or MC16 genes in those models were too excessive (much more than 1.5-fold), ectopic (out of place) or occurred in the wrong developmental timings, making it difficult to compare the results. To improve on this, Yamakawa and his colleagues established a common platform in which heterozygous knockout mice for the HC21-orthologous MC16 candidate gene were mated with partial trisomy MC16 mice such as Ts1Cje. This procedure ensures that the number of copies of the candidate gene is returned to two, while other genes on the trisomic segment retain three copies. Using this system, it is possible to investigate whether any of the DS-like abnormal phenotypes are improved, allowing the contribution of each gene to the biochemical, biophysical or behavioral DS abnormal parameters to be compared impartially. The team has already identified and reported a large number of parameter abnormalities, including decreases in mitochondrial membrane potential and adenosine triphosphate production, increases in reactive oxygen species and kinase activities in Ts1Cje, and enlarged brain ventricles, impaired developmental and adult neurogenesis in Ts1Cje and Ts2Cje. These DS mouse models also display behavioral abnormalities indicative of learning defects. Knockout mice for more than ten HC21-orthologous MC16 genes have been generated, including a model for Dscam—a neural cell adhesion molecule that Yamakawa’s team has long been investigating as a promising candidate for DS mental retardation2. The team is now characterizing mice obtained by mating these model mice with Ts1Cje or Ts2Cje, and evaluating the role of each gene in the DS-like abnormal phenotypes.

A highly efficient system for generating Down syndrome mouse models

Although Yamakawa’s strategy is promising, the generation of many knockout mice is still a daunting task. To implement large-scale analysis more effectively, Yamakawa's group is now establishing a new system that will allow mice to be generated with high efficiency and with free design of partial trisomic segments harboring gene knockouts as desired. In their approach, a selection marker is introduced into MC16 in a mouse primary fibroblast, which is then processed into ‘microcells’, each of which contains, on average, a single chromosome string fused to a special ‘X’ cell. In the cells containing MC16, which are separated out, MC16 can be manipulated to achieve the desired chromosomal segments and gene knockouts with high efficiency. The designed cell containing recombinant MC16 is once again processed into microcells, and then fused into a mouse embryonic stem cell to generate a partial trisomic MC16 mouse. “This system will enable us to not only accelerate our research but also implement experiments that have been impossible, such as the inactivation of multiple genes simultaneously,” Yamakawa says.

Atsushi Shimohata, a member of the technical staff at Yamakawa’s laboratory, and Kenji Amano, a research associate, are currently devoting much of their endeavors to establishing this new system and to removing unnecessary segments from MC16 within the fused cell. Ei-ichi Takaki and Sachie Asada, both research scientists in Yamakawa’s laboratory, are concurrently attempting to establish backup systems, including another type of mouse generating system and a high-efficiency in vitro screening system.

Within 3–5 years, Yamakawa hopes to establish these systems and identify several critical DS genes to elevate DS research to the next level. “We hope that our great passion for this project will eventually lead to alleviating patients’ conditions,” Yamakawa says.

Journal information
1. Amano, K., Sago, H., Uchikawa, C., Suzuki T., Kotliarova, E. S., Nukina N., Epstein J. C. & Yamakawa, K. Dosage-dependent over-expression of genes in the trisomic region of Ts1Cje mouse model for Down syndrome. Human Molecular Genetics 13, 1333–1340 (2004).

2. Amano, K., Fujii, M., Arata, S., Tojima, T., Ogawa, M., Morita, N., Shimohata, A., Furuichi, T., Itohara, S., Kamiguchi, H., Korenberg, J. R., Arata, A. & Yamakawa, K. DSCAM deficiency causes loss of pre-inspiratory neuron synchroneity and perinatal death. Journal of Neuroscience 29, 2984–2996 (2009).