Abstract

Pura is a ubiquitous DNA- and RNA-binding protein that is highly conserved in eukaryotic organisms. Pura binds preferentially to purine-rich single-stranded DNA. To date, biochemical analyses of Pura have indicated that this protein might be involved in diverse cellular functions, including DNA replication, transcription, translation and cell growth control (Gallia et al., 2000). However, the function of Pura in vivo remains unclear. In this study, Pura knock-out mice (KO) were generated using homologous recombination techniques in ES cells. The Pura gene locus was isolated from 129/SvJ Mouse Genomic Library and partially sequenced. A targeting vector using positive and negative selection was generated for homologous recombination and was finally electroporated into E14 ES cells for five times. Out of 2400 ES cell clones only one positive ES cell clone (KO 248) was identified by Southern-blot analysis. Pura KO mice generated from ES cell clone KO 248 displayed multiple phenotypes. Apparently, they were normal when born, but developed assorted abnormalities. 12-14 days after birth they developed a severe and continuous tremor phenotype similar to shiverer mice that lack myelin basic protein (MBP) (Chernoff, 1981). This tremor phenotype continued unabated until natural death occurred. Pura KO mice seemed to have a shorter life span since none of the Pura KO mice lived longer than 6 months while the wild type mice lived normally. Pura has been shown to be involved in developmental and transcriptional control of MBP (Haas et al., 1995), therefore the MBP levels in the brain stems of Pura KO mice were investigated. It could be shown that MBP levels in 16-day-old Pura KO mice were decreased by 30-40% as compared to heterozygotes and wild type. However, MBP levels of the 63-dayold Pura KO, heterozygous and wild type mice did not show significant difference. Although Pura KO mice showed a tremor phenotype comparable to shiverer mice and displayed reduced MBP levels at the beginning of brain myelination, they still possessed 60-70% of normal amounts of MBP, which has been shown to be sufficient to rescue the tremor phenotype in transgenic shiverer mice. These data suggest that the tremor phenotype of Pura KO mice may be due to some other reasons, which are still unclear. Strikingly, 70-day-old Pura KO mice displayed a steadily increasing obesity and significantly enlarged brain size with increased lipid content, suggesting aberrant fat metabolism or homeostasis, although the underlying cause(s) are as yet unknown. Brain stems of Pura KO mice were significantly enlarged and were about three times as heavy as those of the normal mice. The hippocampus and sensoric cortex of Pura KO mice were also enlarged but had decreased weight. Moreover, more lipids were observed not only in the brain of Pura KO mice but also in blood according to blood smear analysis. It has also been shown that female Pura KO mice had more fat droplets in blood than males. Apart from the phenotypes described above, Pura KO mice displayed a different weight development curve as compared to normal mice: With the onset of tremor phenotype, Pura KO mice started to loose their weight; the weight loss of the male KO mice beginning from postnatal day 11-15 was more severer than females and their weight recovery lasted longer than females; the female Pura KO mice are heavier than males while normal male mice are heavier than females, suggesting that this obesity effect may also be influenced by animal gender. These results indicate that Pura may influence or regulate the metabolism of fat in an unknown mechanism. In the second part of this study, a new, yet uncharacterized in vitro enzymatic activity of Pura was investigated. Pura displayed a helix destabilizing activity by unwinding a labeled 17-mer oligonucleotide annealed to ssM13 DNA. This activity was independent of ATP or any other nucleosides and lacks any directionality, distinguishing it therefore from common helicase activities. 16 truncated GST-Pura were generated by PCR for mapping of the domain responsible for the DNA unwinding activity. It was revealed that the central part of the protein from amino acids 54-215 was essential for this feature of Pura. The identical part of Pura was sufficient and essential for DNA binding activity. Therefore, DNA binding could not be separated from DNA unwinding capacity. In addition, from pull-down assays, it was shown that the same part of the protein was also sufficient and necessary for Pura selfassociation. Since Pura has been shown to influence many important cellular processes, the DNA unwinding activity of the protein may establish a mechanism of how Pura functions in different cellular processes, such as replication and transcription.