Production of a model for Lesch-Nyhan syndrome in hypoxanthine phosphoribosyltransferase-deficient mice

Secondary loss of deoxyguanosine kinase activity in purine nucleoside phosphorylase deficient mice

The T-cell immunodeficiency associated with purine nucleoside phosphorylase (PNP) deficiency in man is believed to be due to the accumulation of dGTP which may be preferentially formed from deoxyguanosine in T-lymphocytes or their precursor cells. We found no evidence for dGTP accumulation in thymocytes or spleen leucocytes, < 1 nmol/10(9) cells, nor in erythrocytes, < 0.05 nmol/10(9) cells, of the B6-NPE- or B6-NPF PNP-deficient mice strains. There were no changes in purine or pyrimidine ribonucleotide pools. As these mice had been previously shown to excrete PNP nucleoside substrates, we examined the metabolism of deoxyguanosine. Deoxyguanosine kinase activity as compared to control mice was 6 to 52% for the B6-NPE mutant, 2 to 22% for the B6-NPF mutant. Fractionation of erythrocyte and liver lysates from the F mutation and the background strain, C57BL/6J, by anion exchange chromatography confirmed the secondary deficiency of deoxyguanosine kinase and demonstrated that this activity was distinct from adenosine kinase and two major peaks of deoxycytidine kinase activity. Mouse PNP, expressed and purified as a fusion protein, did not show evidence of being bifunctional and having deoxyguanosine kinase activity. Metabolic modelling revealed that the ratio of deoxyguanosine phosphorylation versus phosphorolysis was < 0.06 in control mice, and < or = 0.3 in lymphocytes of PNP-deficient mice. Were deoxyguanosine kinase not reduced in the PNP-deficient mice, all tissues of the B6-NPF mutant would preferentially phosphorylate deoxyguanosine at low substrate concentrations.

Gene targeting in the mouse

Mice with alterations to specific endogenous genes can be produced by gene targeting in embryonic stem cells. The field has developed rapidly over the past decade, so that large numbers of mice with different gene deficiencies have been generated. Knockout mice provide an ideal opportunity to analyse the function of individual mammalian genes and to model a range of human inherited disorders. This powerful approach has also identified numerous examples of gene redundancy and has highlighted the need to consider metabolic differences between man and mouse in disease modelling. More sophisticated gene-targeting methods are now being used to introduce subtle gene alterations. In the future, more refined genetic analysis and genome, rather than individual gene, alterations will be achieved by incorporating site-specific recombination into targeting strategies. Gene targeting could also make a contribution to improved protocols for gene therapy.

5'-flanking sequences of the human HPRT gene direct neuronal expression in the brain of transgenic mice

Total deficiency of hypoxanthine phosphoribosyltransferase (HPRT) in humans causes the neurological disorder Lesch-Nyhan syndrome. The HPRT gene is expressed at basal levels in all tissues but at higher levels in the brain, the relevance and mechanism of which is unknown. To determine if cis-acting DNA elements play a role in the tissue-differential pattern of expression, we generated transgenic mice carrying different sequences of the human HPRT (hHPRT) promoter fused to the bacterial lacZ gene. We show that a 1.6 kb fragment of the hHPRT promoter contains essential information to direct beta-galactosidase expression preferentially to the basal ganglia, cerebral cortex, hippocampus, and several other areas of the forebrain. At least two elements within the 1.6 kb fragment appear to be required for neuronal expression. A 182 bp element (hHPRT-NE) represents one of these sequences and is involved not only in conferring neuronal specificity but also in repressing transgene expression in non-neuronal tissues. These studies provide molecular insight into the mechanism of increased HPRT expression in the brain.

Murine genetic models of human disease

Until recently, good animal models of human disease have been available only in limited numbers, largely because of technical difficulties associated with transgenesis. As a consequence of recent rapid advances principally, but not exclusively, focused around the use of embryonic stem cells, it is now theoretically possible to model the genetic lesion underlying any human disease in the mouse. This has led not only to a better understanding of complex disease processes, such as those associated with malignancy, but, as in the cases of cystic fibrosis and duchenne muscular dystrophy, is now allowing the development of novel therapy regimes.

Mouse homologues of human hereditary disease

Details are given of 214 loci known to be associated with human hereditary disease, which have been mapped on both human and mouse chromosomes. Forty two of these have pathological variants in both species; in general the mouse variants are similar in their effects to the corresponding human ones, but exceptions include the Dmd/DMD and Hprt/HPRT mutations which cause little, if any, harm in mice. Possible reasons for phenotypic differences are discussed. In most pathological variants the gene product seems to be absent or greatly reduced in both species. The extensive data on conserved segments between human and mouse chromosomes are used to predict locations in the mouse of over 50 loci of medical interest which are mapped so far only on human chromosomes. In about 80% of these a fairly confident prediction can be made. Some likely homologies between mapped mouse loci and unmapped human ones are also given. Sixty six human and mouse proto-oncogene and growth factor gene homologies are also listed; those of confirmed location are all in known conserved segments. A survey of 18 mapped human disease loci and chromosome regions in which the manifestation or severity of pathological effects is thought to be the result of genomic imprinting shows that most of the homologous regions in the mouse are also associated with imprinting, especially those with homologues on human chromosomes 11p and 15q. Useful methods of accelerating the production of mouse models of human hereditary disease include (1) use of a supermutagen, such as ethylnitrosourea (ENU), (2) targeted mutagenesis involving ES cells, and (3) use of gene transfer techniques, with production of 'knockout mutations'.

Alzheimer's disease: molecular genetics and transgenic animal models

Disease-causing mutations in the amyloid precursor protein (APP) gene have been found on chromosome 21 during the last 2 years in some early onset Alzheimer's disease (AD) families. Genetic evidence shows that other genes than the APP are also involved in the aetiology of AD. Linkage to a loci on chromosome 14 has been found in early onset disease. The identification of APP mutation has led to the realization that APP mismetabolism is a central event in the aetiology and pathogenesis of the disease. Experiments to test this in transgenic mice have so far met with little success. There are many possible explanations for the problems to generate transgenic mice. These include the possibilities that mice are incapable of developing AD for reasons dependent on their APP sequence; and that appropriate regulation of APP gene is required for pathology to develop. Current attempts that seem promising to model the disease pathology are the use of homologous recombination to insert the pathogenic mutation and transfection of YACs into transgenic animals.

Age-dependent selection against hypoxanthine phosphoribosyl transferase-deficient cells in mouse haematopoiesis

The basis of a previously observed difference in the level of contribution of hypoxanthine phosphoribosyltransferase-deficient cells between the haematopoietic and non-haematopoietic tissues of chimaeric and heterozygous mice has been clarified by studying two populations of female mice that differ only in that one is heterozygous for a null allele at the hprt locus and the other is wild type at this locus. Both populations are heterozygous for an electrophoretic variant allele at the X-linked Pgk-1 locus, so that X-chromosome inactivation generates cells expressing different isozymes of phosphoglycerate kinase which can be assayed to monitor cell selection. The results show that hypoxanthine phosphoribosyltransferase deficiency itself, rather than an effect of another X-linked gene, causes a reduced level of contribution to haematopoietic tissues. Further, the extent of the depletion increases significantly with age, and this effect is due to a progressive reduction in the level of contribution to haematopoietic tissues rather than to an increase in the level of contribution to non-haematopoietic tissues.

Biochemical genetics: examples of life after cloning

The blend of biochemistry and molecular biology required to understand the pathogenesis of genetic disease is assuming an increasing role in research. We review three example of this inevitable post-cloning trend: first, the surprising relationship between mice with albino deletions and human hereditary tyrosinemia type I; second, the discovery that choroideremia is due to defect in prenylation; and third, fibrillin mutations in the Marfan syndrome.