Biology of the prion gene complex.

The prion protein gene Prnp encodes PrPSc, the major structural component of prions, infectious pathogens causing a number of disorders including scrapie and bovine spongiform encephalopathy (BSE). Missense mutations in the human Prnp gene, PRNP, cause inherited prion diseases such as familial Creutzfeldt-Jakob Disease. In uninfected animals, Prnp encodes a GPI-anchored protein denoted PrPC, and in prion infections, PrPC is converted to PrPSc by templated refolding. Although Prnp is conserved in mammalian species, attempts to verify interactions of putative PrP-binding proteins by genetic means have proven frustrating in that two independent lines of Prnp gene ablated mice (Prnp0/0 mice: ZrchI and Npu) lacking PrPC remain healthy throughout development. This indicates that PrPC serves a function that is not apparent in a laboratory setting or that other molecules have overlapping functions. Shuttling or sequestration of synaptic Cu(II) via binding to N-terminal octapeptide residues and (or) signal transduction involving the fyn kinase are possibilities currently under consideration. A new point of entry into the issue of prion protein function has emerged from identification of a paralog, Prnd, with 25% coding sequence identity to Prnp. Prnd lies downstream of Prnp and encodes the Dpl protein. Like PrPC, Dpl is presented on the cell surface via a GPI anchor and has three alpha-helices: however, it lacks the conformationally plastic and octapeptide repeat domains present in its well-known relative. Interestingly, Dpl is overexpressed in two other lines of Prnp0/0 mice (Ngsk and Rcm0) via intergenic splicing events. These lines of Prnp0/0 mice exhibit ataxia and apoptosis of cerebellar cells, indicating that ectopic synthesis of Dpl protein is toxic to CNS neurons: this inference has now been confirmed by the construction of transgenic mice expressing Dpl under the direct control of the PrP promoter. Remarkably, Dpl-programmed ataxia is rescued by wt Prnp transgenes. The interaction between the Prnp and Prnd genes in mouse cerebellar neurons may have a physical correlate in competition between Dpl and PrPC within a common biochemical pathway that, when misregulated, leads to apoptosis.     

Spatial and temporal down-regulation of transgene expression using the TRSID-silencer in mice: application to Prnp

Spatial and temporal control of ovine prion protein (Prnp) gene expression was achieved in mice using two transgenes: a Prnp minigene with tet-operator sequences inserted 5' to exon 1 and a mouse neurofilament genomic clone carrying the chimeric-repressor TRSID cDNA. In bi-transgenic mice, ovine PrP(C) expression could be reversibly controlled in neuronal cells by doxycycline treatment whereas it remains constant in other cell types. Overall, this model opens opportunities to assess the involvement of cell types in prion diseases and PrP physiological function. It demonstrates the potentiality of the TRSID-silencer to precisely control temporal and spatial gene expression in vivo.     

Doppel is an N-glycosylated, glycosylphosphatidylinositol-anchored protein. Expression in testis and ectopic production in the brains of Prnp(0/0) mice predisposed to Purkinje cell loss

The Prnd gene encodes a homolog of the cellular prion protein (PrP(C)) called doppel (Dpl). Up-regulation of Prnd mRNA in two distinct lines of PrP gene ablated (Prnp(0/0)) mice, designated Rcm0 and Ngsk, is associated with death of Purkinje cells. Using recombinant Dpl expressed in Escherichia coli and mouse neuroblastoma cells we demonstrate that wild type (wt) Dpl, like PrP(C), adopts a predominantly alpha-helical conformation, forms intramolecular disulfide bonds, has two N-linked oligosaccharides, and is presented on the cell surface via a glycosylphosphatidylinositol anchor. Dpl protein was detected in testis of wt mice. Using Triton X-114 phase partitioning to enrich for glycosylphosphatidylinositol-anchored proteins, Dpl was detected in brain samples from Rcm0 Prnp(0/0) mice but was absent in equivalent samples from wt mice and ZrchI Prnp(0/0) mice, indicating that ectopic expression of this protein may cause cerebellar pathology in Rcm0 mice. Biochemical and structural similarities between PrP(C) and Dpl documented here parallel the observation that ataxic Ngsk Prnp(0/0) mice can be rescued by overexpression of wild-type PrP transgenes, and suggest that cell surface PrP(C) can antagonize the toxic effect of Dpl expressed in the central nervous system.     

The prion gene complex encoding PrP(C) and Doppel: insights from mutational analysis

The prion protein gene, Prnp, encodes PrP(Sc), the major structural component of prions, infectious pathogens causing a number of disorders including scrapie and bovine spongiform encephalopathy (or BSE). Missense mutations in the human Prnp gene cause inherited prion diseases such as familial Creutzfeldt-Jakob disease. In uninfected animals Prnp encodes a glycophosphatidylinositol (GPI)-anchored protein denoted PrP(C) and in prion infections PrP(C) is converted to PrP(Sc) by templated refolding. Though Prnp is conserved in mammalian species, attempts to verify interactions of putative PrP binding proteins by genetic means have proven frustrating and the ZrchI and Npu lines of Prnp gene-ablated mice (Prnp(0/0) mice) lacking PrP(C) remain healthy throughout development. This indicates that PrP(C) serves a function that is not apparent in a laboratory setting or that other molecules have overlapping functions. Current possibilities involve shuttling or sequestration of synaptic Cu(II) via binding to N-terminal octapeptide residues and/or signal transduction involving the fyn kinase. A new point of entry into the issue of prion protein function has emerged from identification of a paralogue, Prnd, with 24% coding sequence identity to Prnp. Prnd lies downstream of Prnp and encodes the doppel (Dpl) protein. Like PrP(C), Dpl is presented on the cell surface via a GPI anchor and has three alpha-helices: however, it lacks the conformationally plastic and octapeptide repeat domains present in its well-known relative. Interestingly, Dpl is overexpressed in the Ngsk and Rcm0 lines of Prnp(0/0) mice via intergenic splicing events. These lines of Prnp(0/0) mice exhibit ataxia and apoptosis of cerebellar cells, indicating that ectopic synthesis of Dpl protein is toxic to central nervous system neurons: this inference has now been confirmed by the construction of transgenic mice expressing Dpl under the direct control of the PrP promoter. Remarkably, Dpl-programmed ataxia is rescued by wild-type Prnp transgenes. The interaction between the Prnp and Prnd genes in mouse cerebellar neurons may have a physical correlate in competition between Dpl and PrP(C) within a common biochemical pathway that when mis-regulated leads to apoptosis.     

Tissue-specific, high level expression of the rat whey acidic protein gene in transgenic mice.

The importance of intragenic and 3' flanking sequences in the control of the temporal, hormonal and tissue-specific expression of milk whey acidic protein (WAP) has been demonstrated in transgenic mice. Mouse lines carrying a 4.3 kb genomic clone containing the entire rat WAP gene minus 200 bp of the first intron with 0.949 kb of 5' and 1.4 kb of 3' flanking DNA were generated. In eight of nine independent lines of mice analyzed, WAP transgene expression was detected at levels ranging from 1% to 95% (average, 27%) of the endogenous gene. The transgene was expressed preferentially in the mammary gland. Although developmentally regulated during pregnancy and lactation, the temporal pattern of WAP transgene expression differed from the endogenous gene. A precocious increase in expression of the transgene was detected at 7 days of pregnancy, several days earlier in pregnancy than the major increase observed in endogenous mouse WAP mRNA. The rat WAP transgene was translated and secreted into the milk of transgenic mice at levels comparable to the endogenous mouse WAP. This is the first report of a gene that is negatively regulated in dissociated cell cultures as well as in transfected cells, yet is expressed efficiently in the correct multicellular environment of the transgenic mouse.     

Deletion analysis of the human CD2 gene locus control region in transgenic mice.

Genomic sequences located at the 3' flanking region of the human CD2 gene confer high level tissue-specific, position-independent expression of the gene when introduced in the germ line of mice. In order to further characterize these sequences a range of deletions, from the 3' end were produced and transgenic mice were generated with the human CD2 (hCD2) gene linked to these deleted fragments. This allowed us to establish the minimum sequences necessary for the copy-dependent transgene expression. 2.1 kb or 1.5 kb of 3' flanking sequences linked to a hCD2 mini-gene is sufficient to allow T-cell specific, copy-dependent, integration-independent expression in transgenic mice. 1.1 kb of 3' sequences results in the gene being expressed in a T-cell specific manner, but copy-dependent, integration-independent expression was not observed in a small number of transgenic animals. 0.2 or 0.5 kb of 3' flanking sequences were insufficient to allow expression above the level previously found with a human CD2 gene which lacked 3' flanking sequences. We conclude that the Locus Control Region (LCR) effect is caused by 1.5 kb of flanking sequences immediately 3' to the polyadenylation signal of the gene.     

Ectopic expression of chloramphenicol acetyltransferase (CAT) in the cerebellum in mice transgenic for a carbonic anhydrase II promoter-CAT construct that is without apparent phenotypic effect

We have developed six transgenic lines of mice with constructs containing presumptive 5' regulatory regions of carbonic anhydrase II (CA II). Four of the lines contained 1,100 bases of the 5' flanking region of the human CA II gene, and two transgenic lines resulted from a construct containing 500 bases of the 5' flanking region of the mouse CA II gene. Tissue-specific expression of the chloramphenicol acetyltransferase (CAT) gene was not obtained in any of the transgenic lines. One of the transgenic lines was found to have high levels of expression of CAT in cerebellum. This expression persisted through multiple generations and was independent of the parental origin of the transgene. On the assumption that the expression was due to the insertion of the transgene in or near a gene expressed normally in cerebellum, homozygous mice were bred for the transgenic insert to see if a mutation might have been induced. Homozygous mice were found and seemed to be normal in all aspects of their phenotype studied. Thus, in this case, neither the insertion of the gene nor the ectopic expression of CAT seemed to be harmful to the animals.     

Secretion of biologically active human granulocyte colony-stimulating factor (G-CSF) in milk of transgenic mice

Two constructs were devised, containing the full-length gene of the human granulocyte colony-stimulating factor (G-CSF) fused with the 5' and 3' flanking promoter sequences of bovine alpha-S1-casein gene. Both constructs contained a 1518-bp fragment that included exons 18 and 19 and 320 bp of the 3' flanking region of bovine gene @CSN1S1, but differed in size of the 5' flanking sequences, which were of 721 bp, and exon 1 in construct pGCm1 and 2001 bp and exon 1 and intron 1 in construct pGCm2. With both constructs, transgenic mice were produced. The transgene expression was assessed using RT-PCR and immunochemically from the production of human G-CSF in milk of lactating females. Secretion of human G-CSF into the milk varied in a wide range, from 0.8 microg/ml to over 1 mg/ml, in mice with construct pGCml and was low (up to 60 microg/ml) or absent in mice with construct pGCm2. G-CSF glycosylation was incomplete in mice with transgene pGCml and complete in mice with pGCm2. G-CSF of transgenic mouse milk was shown to stimulate the formation and growth of granulocyte-containing colonies in human umbilical blood cell culture and be close or identical in physiological activity to the natural human G-CSF.     

A potato Sus3 sucrose synthase gene contains a context-dependent 3' element and a leader intron with both positive and negative tissue-specific effects.

To examine which sequences are involved in regulating the potato sucrose synthase gene Sus3-65, we examined a series of deletion and substitution constructs in transgenic potato and tobacco plants. In a construct containing 3.9 kb of 5' flanking region, substitution of the native 3' sequence with the nopaline synthase 3' sequence and deletion of the leader intron did not significantly affect expression in vegetative tissues. However, in a construct containing only 320 bp of 5' flanking region, these changes had marked effects. Replacing the native 3' sequences with nopaline synthase 3' sequences caused a six- to 20-fold increase in expression in vascular tissue, and removing the leader intron almost completely abolished expression in potato plants. Surprisingly, removal of the leader intron from either the full-length construct or a construct containing only 320 bp of 5' flanking sequence reduced expression in vascular tissue of tobacco anthers at later stages of development but increased expression in pollen by more than 100-fold.     

High-level tuber expression and sucrose inducibility of a potato Sus4 sucrose synthase gene require 5' and 3' flanking sequences and the leader intron.

The 3.6 kb of 5' flanking sequence, leader intron, and 0.7 kb of 3' sequence from the potato sucrose synthase gene Sus4-16 are sufficient to direct high-level expression in developing tubers, in basal tissues of axillary buds and shoots, and in meristems and caps of roots, and to confer sucrose inducibility in leaves. By examining a series of deletion and substitution constructs in transgenic potato plants, we found that this pattern of expression requires 5' flanking sequences both upstream and downstream of position -1500 and that sequences between positions -1500 and -267 are essential for sucrose induction. Replacement of the native 3' sequence with the nopaline synthase 3' sequence resulted in the loss of sucrose inducibility and of expression in basal tissues of axillary buds. A general decrease in expression in other tissues was also observed. Removal of the 1612-bp leader intron also had a dramatic effect on both the pattern and level of expression.