Transgenic Studies of Prion Diseases

Introduction

This topic reviews studies that involve the manipulation of prion protein (PrP) genes by transgenesis in mice. These consist of two approaches: PrP gene knockout and gene replacement using homologous recombination in embryonic stem cells; and microinjection of transgenes into fertilized embryos. These studies have provided important insights into the pathogenesis of prion diseases including the molecular basis of prion strains and species barriers. Transgenic approaches have also provided important information about the mechanism by which human prion diseases can be both genetic and infectious. Despite these advances, our understanding of these unique pathogens is far from complete. Transgenic approaches will doubtless remain the cornerstone of investigations into the prion diseases, and will be important in the development of therapeutic agents in coming years.

Knockout and Gene Replacement Studies

Some of the most compelling evidence to date for the protein-only hypothesis of prion replication derives from experiments with knockout transgenic mice. Because PrPC is the source of PrPSc, the model predicts that elimination of PrPC would abolish prion replication. To test this, the mouse PrP gene, referred to as Prnp, was disrupted by homologous recombination in embryonic stem cells. Stem cells containing the disrupted Prnp gene were introduced into mouse blastocysts, and knockout mice were established (1-3). Unlike wild-type mice, the resultant homozygous null mice, referred to as Prnp0/0, which express no PrPC, fail to develop the characteristic clinical and neuropathologi-cal symptoms of scrapie after inoculation with mouse prions, and do not propagate prion infectivity (3-6), while mice that are hemizygous for PrP gene ablation, and therefore expresses one-half the normal level of PrPC, have prolonged incubation times (5-7).


Two Prnp0/0 lines, in which the PrP coding sequence was disrupted, were independently generated in Zurich and Edinburgh. Contrary to expectation, these mice developed normally, and suffered no gross phenotypic defects (1,2). These results raised the possibility that adaptive changes that compensate for the loss of PrPC function occur during the development of Prnp0/0 mice. To test this hypothesis, transgenic mice were produced in which expression of transgene-expressed PrPC could be experimentally regulated. Using the tetra-cycline gene-response system, mice were produced that co-express a tetracy-cline (TET)-responsive transactivator (tTA), and a tTA-responsive promoter that drives PrP expression (8). The tTA consists of the TET repressor fused to the transactivation domain of herpes simplex virus VP16, and binds specifically and with high affinity to the TET operator (tetO). Binding of tTA normally activates transcription of the PrP gene, but binding of doxycycline (DOX), a TET analog, to tTA prevents the tTA protein from binding to tetO, which in turn prevents PrP gene expression. Thus, PrPC is expressed in the absence of DOX, but not in its presence. Repressing PrPC expression by oral administration of DOX was not deleterious to adult mice. However, since DOX treatment did not completely inhibit PrPC expression in these mice, it is not clear whether this residual expression masks the true phenotype of Prnp0/0 mice.

A third line of gene-targeted Prnp010 mice generated in Nagasaki also developed normally, but, unlike the Prnp0/0 mice created in Zurich and Edinburgh, they showed progressive ataxia and cerebellar Purkinje cell degeneration at about 70 wks (9). Further histologic studies of these mice also revealed abnormal myelination in the spinal cord and peripheral nerves (10). These Prnp010 mice were successfully rescued from demyelination and Purkinje cell degeneration by introduction of a transgene encoding wild-type mouse PrPC, as a result of mating to Tg(MoPrP-A)4053/FVB mice (11). Resolving the molecular basis of the phenotypic differences in the different Prnp010 lines is important because the results may have widespread applications for understanding general mechanisms of neurodegeneration. The recent discovery of the Doppel (Dpl) locus, referred to as Prnd, 16 kb downstream of Prnp provided an important clue to this conundrum (12). Dpl is upregulated in the central nervous system (CNS) of Prnp010 mice that develop late-onset ataxia, suggesting that Dpl may provoke neurodegeneration.

Several other phenotypic defects have also been investigated in Prnp0/0 mice including altered circadian rhythms and sleep patterns (13), alterations in superoxide dismutase activity (14) and defects in copper metabolism (15). Electrophysiological studies suggested that y-aminobutyrio acid A receptor-mediated fast inhibition and long-term potentiation are impaired in hippocampal slices from Prnp0/0 mice (16,17), although these defects could not be identified in other studies (18,19).

Studies with inbred strains of mice demonstrate that a scrapie incubation time locus in mice, referred to as Prn-i, was genetically linked to the mouse PrP gene, Prnp (20). Inbred strains of mice with long and short scrapie incubation periods harbor distinct Prnp alleles, referred to as Prnpa and Prnpb, which differ at codons 108 (Leucine to Phenylalanine) and 189 (Threonine to Valine) (21). In order to precisely define the effects of these PrP polymorphisms on prion incubation times, the coding sequence of the endogenous Prnpa gene in embryonic stem cells was replaced with the Prnpb coding sequence by homologous recombination. Studies with the resulting transgenic mice confirmed, as predicted, that Prn-i and Prnp are the same gene, and that amino acid differences at residues 108 and/or 189 in PrP-A and PrP-B modulate scrapie incubation times (22). A mutation equivalent to P102L in the human PrP gene, associated with Gerstmann-Straussler Scheinker disease (GSS), has also been introduced into Prnp by gene targeting (23).

Microinjection Transgenic Mice

The majority of transgenic studies of prion diseases have involved the incorporation of wild type or mutant PrP genes from different species into the genome of fertilized mouse embryos by DNA microinjection. Various chi-meric gene constructs, incorporating PrP gene sequences from mouse and other species, have also been used to produce transgenic mice. The seminal transgenic experiments utilized cosmid clones containing PrP gene sequences isolated from Syrian hamster (SHa) and the I/lnJ strain of mice (24,25) and this approach was also used to produce transgenic mice expressing sheep PrP (26). The cos.Tet vector is a modification of the SHa cosmid vector and contains a 43-kb DNA fragment encompassing the PrP gene and approx 24 and 6 kb of 5′ and 3′ flanking sequences, respectively (27). The vector is designed to allow the convenient insertion of PrP coding sequences. A plasmid expression vector, based on the PrP gene derived from the I/lnJ PrP cosmid (phgPrP, or the "half-genomic" construct) has also been used to produce transgenic mice (28). A modified version of phgPrP, (MoPrP.Xho), has been produced with a unique XhoI site to allow the insertion of coding sequences downstream of exon 2 in Prnp (29). SHa and mouse PrP gene constructs, in which all intron sequences are removed (so-called "minigene constructs), fail to express PrP in the CNS, demonstrating the requirement for at least the smaller intron for efficient expression (24,28).

A transgenic expression vector (pMo53), has been engineered to incorporate the advantages and overcome the problems of the cosmid and plasmid vectors (30). The vector contains 5′ -flanking sequences derived from the I/lnJ PrP gene, which have been shown to directly express a chloramphenicol acetyl transferase reporter gene in mouse neuroblastoma Neuro-2A cells (31). The 1.2-kb 3′-flanking region consists of the polyadenylation signal from Prnp. The vector is designed to accept open reading frame cassettes at unique restriction sites immediately downstream from exon 2. The construct also allows the convenient replacement of the promoter region for ectopic expression studies.

Transgenic Models of Inherited Prion Diseases

Approximately 10-20% of human prion disease is inherited with an autoso-mal dominant mode of inheritance. More than 20 different missense and insertion mutations, which segregate with dominantly inherited neurodegenerative disorders, have been identified in the coding sequence of PRNP. Five of these mutations are genetically linked to loci controlling familial Creutzfeldt-Jakob disease (CJD), GSS, and fatal familial insomnia (FFI), which are all inherited human prion diseases that can be transmitted to experimental animals.

Additional compelling evidence for the protein-only hypothesis came from studies on transgenic mice that express a proline to leucine mutation at codon 101 of mouse PrP, equivalent to the human GSS P102L mutation. These mice (Tg[MoPrP-P101L]) spontaneously developed clinical and neuropathological symptoms similar to mouse scrapie at between 150 and 300 d of age (11,32). After crossing the mutant transgene onto the Prnp0/0 background, the resulting Tg(MoPrP-P101L) Prnp010 mice displayed a highly synchronous onset of illness at ~145 d of age, which shortened to ~85 d upon breeding to homozygosity for the transgene array. In addition, Tg(MoPrP-P101L) Prnp0/0 mice had increased numbers of PrP plaques and more severe spongiform degeneration (11). In contrast, transgenic mice overexpressing wild-type mouse PrP at equivalent levels did not spontaneously develop neurodegenerative disease, although they had highly reduced mouse scrapie incubation times after inoculation with mouse prions. The serial propagation of infectivity from the brains of spontaneously sick Tg(MoPrP-P101L) mice to indicator Tg196 mice that express low levels of mutant protein, and do not otherwise develop spontaneous disease, demonstrated the production of infectious prions in the brains of these spontaneously sick mice (11,33). Prion infectivity from brain extracts of humans expressing the P102L GSS mutation was also propagated in transgenic mice expressing a chimeric mouse-human PrP gene with the P101L mutation (34). A synthetic peptide spanning residues 89-143, carrying the P101L mutation that was refolded into a P-sheet conformation, produced clinical signs of neurological dysfunction as well as neuropathological characteristics of prion disease, after ~360 d in 20/20 inoculated Tg196 mice. By contrast, Tg196 mice receiving a substantially larger inoculum of the peptide in a non-^-sheet conformation exhibited no convincing evidence of experimental prion disease (35).

Unlike Tg(MoPrP-P101L) mice, gene-targeted mice, homozygous for the mouse PrP proline to leucine mutation at codon 101 did not spontaneously develop neurodegenerative disease (23). This result is perhaps not unexpected, since previous studies demonstrated that a threshold level of expression of P102L PrP was critical for the manifestation of spontaneous neurological disease in transgenic mice (11,32). Inoculation of gene-targeted MoPrP-P101L mice with prions from a patient with GSS, produced disease in 288 d. Disease was subsequently transmitted to wild-type mice at 226 d, and to gene targeted MoPrP-P101L mice after 148 d. Transmission of additional GSS cases will be important because previous studies suggested that the ability of some cases of GSS, but not others to transmit to wild-type mice, might be the result of strain effects (36).

Continued characterization of the various genetically programmed prion diseases in transgenic mice will provide the basis for studying the molecular mechanisms of phenotypic variability in these conditions. In contrast to Tg(MoPrP-P101L) mice, transgenic mice overexpressing a mutant mouse PrP gene with a glutamate to lysine mutation at codon 199, equivalent to the codon 200 mutation linked to familial CJD (E200K), did not spontaneously develop neurological disease (37). A mutation associated with GSS, in which the Tyrosine residue at codon 145 is mutated to a stop codon, was also modeled in transgenic mice. Expression of this truncated version of mouse PrP could not be detected in high copy number lines of these transgenic mice, referred to as Tg([MoPrP144#]), and neither uninoculated Tg(MoPrP144#) mice nor mice inoculated with mouse RML scrapie developed symptoms of neurodegenerative disease (38). Expression of a mouse PrP version of a nine-octapeptide insertion associated with prion dementia produced a slowly progressive neurological disorder in transgenic mice (39). At this stage it is not known whether infectious prions are produced in the brains of these mice.

Certain examples of prion disease, including transgenic models of GSS, occur without accumulation of protease-resistant PrPSc (11,32). Moreover, the time course of neurodegeneration is not equivalent to the time course of PrPSc accumulation in mice expressing low levels of PrPC (7). Thus, it appears that accumulation of protease-resistant PrPSc may not be the sole cause of pathology in prion diseases. An alternative mechanism of PrP-induced neurodegeneration arose from transgenic studies of mutant forms of PrP that disrupt PrP biogenesis in the endoplasmic reticulum (40,41). Transgenic mice expressing mutations in the so-called "stop transfer effector region" between residues lysine 104-methionine 112 and the hydrophobic TM1 region between residues alanine 113-serine 135, spontaneously develop neurodegenerative disease and accumulate an aberrant form of PrP, termed "CtmPrP". Accumulation of Ctmprp is also associated with a form of GSS in humans that segregates with the codon 117 mutation of PRNP. Accumulation of CtmPrP as a cause of neurodegeneration is not exclusive to genetically programmed prion diseases. An elegant series of experiments in transgenic mice demonstrated that the effectiveness of PrPSc in causing neurodegeneration in transmissible prion diseases depends on the predilection of the host to accumulate CtmPrP (41).

Transgenic Studies of Prion Species Barriers

The species barrier describes the difficulty with which prions from one species cause disease in another. In experimental studies, the initial passage of prions between species is associated with prolonged incubation times, with only a few animals developing illness. On subsequent passage in the same species, all animals become ill after greatly abbreviated incubation times. Prion species barriers have been eliminated by expressing PrP genes from other species or artificially engineered chimeric PrP genes in transgenic mice.

As a result of the species barrier, wild-type mice are normally resistant to infection with Syrian hamsters (SHa) prions. The seminal transgenic experiments by Scott et al. (24) that were designed to probe the molecular basis of the species barrier, demonstrated that expression of SHa PrPC in transgenic (Tg[(SHaPrP]) mice, rendered them susceptible to SHa prions, and produced CNS pathology similar to that found in Syrian hamsters with prion disease. These studies were extended to show that the incubation period of SHa prions was inversely related to the level of expression of transgene-encoded PrPC (42). Inoculation of Tg(SHaPrP) mice with mouse prions resulted in propagation of prions pathogenic for mice; inoculation with SHa prions resulted in the propagation of prions pathogenic for Syrian hamsters. These studies provided important clues about the mechanism of prion propagation involving association and conformational conversion of PrPC into PrPSc, and suggested that, for optimum progression of the disease, the interacting species should be identical in primary structure. SHa PrP differs from mouse PrP at 16/254 amino acid residues (43,44). Chimeric SHa/mouse PrP transgenes produced prions with new properties. The MH2M transgene carries five amino acid substitutions found in SHa PrP lying between codons 94 and 188. Tg(MH2M) mice generated prions with an artificial host range, so that infectiv-ity produced by inoculation with SHa prions could be passaged from Tg(MH2M) mice to wild-type mice, and infectivity produced by inoculation with mouse prions could be passaged from Tg(MH2M) mice to Syrian hamsters (45).

The infrequent transmission of human prion disease to rodents is also an example of the species barrier. Based on the results with Tg(SHaPrP) mice, it was expected that the species barrier to human prion propagation would be abrogated in transgenic mice expressing human PrP. However, transmission of human prion disease was generally no more efficient in transgenic mice expressing high levels of transgene-expressed human PrPC than in non-transgenic mice. In contrast, propagation of human prions was highly efficient in transgenic mice expressing a chimeric mouse-human PrP gene (Tg[MHu2M]) in which the region of the mouse gene between codons 94 and 188 was replaced with human PrP sequences (46). The barrier to CJD transmission in Tg(HuPrP) mice could be abolished by expressing HuPrP on a Prnp0/0 background, demonstrating that mouse PrPC inhibited the transmission of prions to transgenic mice expressing human PrPC, but not to those expressing chimeric PrP (34).

To explain these and other data, it was proposed that the most likely mediator of this inhibition is an auxiliary non-PrP molecule, provisionally designated "protein X," which participates in the formation of prions by interacting with PrPC to facilitate conversion to PrPSc (34). It has been proposed that protein X is bound to a form of PrP, referred to as "PrP*," which exists in equilibrium with PrPC (47). The PrP*-protein X complex interacts with PrPSc, which induces a conformational change in PrP*, the end result being two molecules of PrP with the infectious PrPSc conformation, which are free to induce confor-mational changes in additional PrP* molecules during the infectious cycle. Although protein X has been postulated from genetic arguments, factors that interact with PrPC, and are involved in its conversion to PrPSc, await identification and characterization. To date, no less than 12 proteins have been identified as potential PrPC ligands (reviewed in ref. 48), but in no case has physiological relevance been confirmed.

In agreement with these experimental observations, the three-dimensional structure of recombinant PrP, derived from nuclear magnetic resonance spectroscopy, indicates two potential species-dependent recognition sites for protein-protein interactions on opposite molecular surfaces in the structured C-terminal region of PrP (49-51). A refinement of this model classified amino acid residues that differ between species according to their locations in the three-dimensional structure of PrP and the chemical properties of the amino acid residues (52). The region containing so-called "class A" residues is suggested as the binding site for protein X; the variable "class C" residues are predicted to be involved in interactions between PrPC and PrPSc. A third region, consisting of "class B" residues, constitutes an internal hydrophobic core that may affect structural transformations following PrPSc-PrPC interactions.

Based on the success of Tg(MHu2M) mice, transgenic mice expressing a chimeric mouse-bovine PrP construct (MBo2M) were produced. Both the MHu2M and MBo2M chimeras were constructed by exchanging homologous regions between codons 94 and 188 of human and mouse PrP and bovine and mouse PrP coding sequences, using common restriction enzyme sites. However, although the nine amino acids from the human PrP coding sequence in this region of MHu2M are sufficient to confer susceptibility to CJD prions in transgenic mice, the eight amino acids from bovine PrP in this region of MBo2M are not sufficient to confer susceptibility to bovine songiform encephalopathy (BSE) prions, since Tg(MBo2M) mice did not develop disease after challenge with BSE (53). Transgenic mice expressing bovine PrP developed disease after inoculation with BSE, albeit with long incubation times between 250 and 300d (53).

Because of species-specific differences between mouse and bovine PrP, amino acid differences occur at residues 183 and 185 in a-helix 2 of MBo2M PrP, which are not present in MHu2M. The mouse and bovine PrP sequences also differ at position 202 in a-helix 3; the mouse and human sequences are equivalent in this region. All ungulate species that have succumbed to BSE infection encode valine at 183 and isoleucine at 202, raising the intriguing possibility that these residues may account for the differences in susceptibility of Tg(MHu2M) and Tg(MBo2M) mice to prion infection. The effects of exchanges at these positions on the capacity of transgenic mice to propagate BSE prions are currently being tested (30).

Transgenic Mice and Prion Strains

The degree of homology between PrP molecules in the host and inoculum is an important determinant of the species barrier, but an equally important component affecting prion transmission barriers is the strain of prion. Prion incubation times, profiles of neuropathological lesions in the CNS, and patterns of PrPSc deposition in the brain are features that have been used to characterize prion strains in inbred mice, hamsters, and transgenic mice (37,54-56). The importance of strain effects and species barriers is highlighted in the case of BSE, which has an unusually broad host range. As a result, BSE prions from cattle have caused disease in humans in the form of variant CJD (vCJD). The most convincing evidence that vCJD is the manifestation of BSE in humans has arisen from transmissions of vCJD prions to transgenic mice expressing bovine PrP, which produced incubation periods, neuropathology, PrPSc distribution, and PrPSc conformations that were identical to those produced by inoculation of BSE prions (57).

The passage of vCJD and BSE prions to inbred strains of mice has also been used to contend that prions causing BSE and vCJD are the same prion strain (55). These experiments are complicated by the transmission of prions across species barriers that prolong incubation times. A characteristic banding pattern of PrPSc glycoforms found in vCJD patients and BSE infected animals distinguishes vCJD PrPSc from the patterns observed in classical CJD (58,59). The predominance of diglycosylated PrPSc in both BSE and vCJD brains has also been used as an argument for the two diseases being caused by the same prion strain (58,59). Transgenic mice expressing mutations at one or both glycosylation consensus sites have been studied to investigate the role of the Asn-linked oligosaccharides of PrP (60). Mutation of the first site altered PrPC trafficking and prevented infection with two prion strains. Deletion of the second site did not alter PrPC trafficking, but permitted infection with one prion strain and altered the pattern of PrPSc deposition.

Studies of different strains of transmissible mink encephalopathy, a prion disease of captive mink, suggested that different strains may be represented by different conformational states of PrPSc (61). Evidence supporting this concept emerged from transmission studies of inherited human prion diseases (37). Expression of mutant prion proteins in patients with FFI and familial CJD result in variations in PrP conformation reflected in altered proteinase K cleavage sites that generate PrPSc molecules with molecular weight of 19 kDa in FFI and 21 kDa in fCJD(E200K) (62). Extracts from the brains of FFI and fCJD(E200K) patients transmitted disease to Tg(MHu2M) mice after about 200 d on first passage, and induced formation of 19 and 21 kDa PrPSc, respectively (37). Upon second passage in Tg(MHu2M) mice, these characteristic molecular sizes remain constant, but the incubation times for FFI and fCJD prions diverge (63). These results indicate that PrPSc conformers function as templates in directing the formation of nascent PrPSc and provide a mechanism to explain strains of prions in which diversity is enciphered in the tertiary structure of PrPSc.

A sporadic form of fatal insomnia (SFI) has recently been described (56,64). Although patients with SFI have symptoms and neuropathological profiles indistinguishable from patients with FFI, they do not express the D178N mutant form of human PrPC. SFI prions transmitted to Tg(MHu2M) mice were found to produce similar incubation periods and a pattern of neuropathology identical to transgenic mice infected with FFI prions (56). Analysis of PrPSc demonstrated equivalent conformations associated with SFI and FFI. These findings imply that the conformation of PrPSc not the amino acid sequence, determines the strain-specified disease phenotype.

Structure-Function Studies of PrP

The finding that the introduction of PrP transgenes into Prnp0/0 mice restores susceptibility to scrapie opened the possibility for assessing whether a modified PrPC molecule remains functional, at least insofar as it continues to be eligible for supporting prion propagation (65). Experiments in cell culture showed that deletion of the PrP sequence encoding residues that are removed from the N-terminus of PrPSc by limited proteolysis did not prevent the acquisition of protease resistance and PrPSc formation (66). To further investigate the role of this region, a series of transgenic mice, expressing N-terminal deletions of varying extent, were produced. Deletions between codons 69 and 84, 32 and 80, 32 and 93 or 32 and 106 of the PrP coding sequence were able to restore susceptibility to scrapie in Prnp0/0 mice (28,67), but deletions between codons 32 and 121 or 32 and 134 caused ataxia and degeneration of the granular layer of the cerebellum within 2-3 mo after birth (67). This defect was overcome by the co-expression of wild-type MoPrP, leading to the suggestion that truncated PrP may compete with a functionally similar non-PrP molecule for a common ligand.

A series of PrP coding sequence deletions, based on putative regions of secondary structure, were also expressed in transgenic Prnp0/0 mice (38). These deletions were engineered in a modified PrP construct that lacks amino acid residues 23-88. Transgenic mice with additional deletions between codons 95 and 107, 108 and 121 and 141 and 176 remained healthy; transgenic mice with deletions at the C-terminus between codons 177 and 190 and 201 and 217, which disrupted the penultimate and last a-helix, showed neuronal cytoplasmic inclusions of PrP-derived deposits and spontaneously developed fatal CNS illnesses at 90-227 d of age.

Two of these deletion constructs were further characterized in transgenic mice, regarding their ability to support prion replication (68). Transgenic mice in which residues 23-88 were deleted remained resistant to infection, and this block to prion propagation was alleviated by further deleting residues between 141 and 176. In both cases, the block to prion propagation was overcome by co-expression of wild-type MoPrP.

Ectopic Expression Studies

Although the pathological consequences of prion infection occur in the CNS, PrPC has a wide tissue distribution, and the exact cell types responsible for agent propagation and pathogenesis are still uncertain. PrP is expressed at highest levels in the CNS, but substantial amounts of PrP can be found in many tissues (69). Similarly, although the highest titers of infectious prions are found in the CNS, prions do accumulate in other organs, particularly in the spleen and other tissues of the reticuloendothelial system (70). In the CNS, PrP is expressed in neurons throughout the life of the animal, with levels of PrP mRNA varying among different types of neurons (44).

Transgenic mice expressing heterologous transgenes with cell-type-specific promoter/enhancer sequences linked to PrP coding sequences have been introduced into Prnp0/0 mice in order to study the ability of specific cell types to support prion propagation. Transgenic mice in which the neuron-specific eno-lase promoter regulated SHa PrP expression indicated that neuron-specific expression PrPC was sufficient to mediate susceptibility to hamster scrapie after intracerebral inoculation (71). PrP is also normally expressed in astrocytes and oligodendrocytes throughout the brain of postnatal hamsters and rats (72). The level of glial PrP mRNA expression in neonatal animals is comparable to that of neurons, and increases twofold during postnatal development. Astrocytes have been found to be the earliest site of PrPSc accumulation in the brain (73), suggesting that these cells may play an important role in scrapie propagation and/or pathogenesis or that astrocytes themselves may be the cells in which prion replication occurs. Transgenic mice expressing SHa PrP under the control of the astrocyte-specific glial fibrilary acidic protein accumulated infectivity and PrPSc to high levels and developed disease after ~220 d (74).

The interferon regulatory factor-1 promoter/E^ enhancer, lck promoter, and albumin promoter/enhancer have been used to direct PrP expression to the spleen, T lymphocytes and liver, respectively (75). High prion titers were found in the spleens of inoculated transgenic mice expressing PrP under the control of the interferon regulatory factor-1 promoter/E^ enhancer, while mice expressing PrP under the control of the lck and albumin promoters failed to replicate prions. Finally, although previous reports found little or no prion infectivity in skeletal muscle, two types of transgenic mice, in which expression of PrPC is directed exclusively to muscle under the control of the muscle creatine kinase and chicken a-actin promoters demonstrated that this tissue is capable of propagating prion infectivity (76). That muscle is competent to propagate prions raises the question of whether meat from prion disease infected animals may carry sufficient titers of prions to transmit disease to humans.

Conclusions and Future Prospects

While certain aspects of PrPC to PrPSc conversion can be studied using in vitro systems, many ambiguities remain, and workers continue to rely heavily on in vivo analysis for studying prions and the prion diseases. Of the two general transgenic approaches, transgenesis by pronuclear microinjection has been the more informative for studying the biology of prion diseases. Although this approach results in lines with variable copy number and expression levels, these effects are easily controlled and offer a degree of flexibility that is not possible using gene replacement methods. In addition to providing insights into mechanisms of prion propagation, this approach has also resulted in rapid and sensitive infectivity assays for human and animal prions. Knockout approaches have yielded crucial information about the requirement of PrP expression for prion replication, but they have been less informative than expected in answering questions about the normal function of PrP. The recent discovery of Prnd offers exciting new approaches in this respect and the phenotypes of Prnd and double Prnp/Prnd knockouts are eagerly awaited.

Although transgenic studies demonstrated that species-specific amino acid differences influence the ability of prions from one species to cause disease in another species, studies in transgenic mice have also shown that strain diversity results from the ability of PrPSc to impart different tertiary structures to PrPC by conformational templating. A major goal of future studies will be to determine the interplay between PrP primary structure and conformation in determining prion transmission barriers. In light of BSE transmission to humans, understanding the risk of prion infections from other sources is clearly of paramount importance. Transgenic models of other naturally occurring prion diseases, such as chronic wasting disease of deer and elk, are required to better understand modes of transmission and the pathogenesis of disease, as well as to improve diagnosis and control of these diseases. In addition, the complex genetics of scrapie susceptibility in sheep presents many opportunities for transgenic investigations.

Next post:

Previous post: