Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
T. Wisniewski, New York University School of Medicine, Departments of Neurology, Psychiatry and Pathology, Millhauser Laboratories, Room HN419, 560 First Avenue, New York, NY 10016, USA Fax: +1 212 263 7528 Tel: +1 212 263 7993 E-mail: email@example.com
Alzheimer's and prion diseases belong to a category of conformational neurodegenerative disorders [Prusiner SB (2001) N Eng J Med344, 1516–1526; Sadowski M & Wisniewski T (2007) Curr Pharm Des13, 1943–1954; Beekes M (2007) FEBS J274, 575]. Treatments capable of arresting or at least effectively modifying the course of disease do not yet exist for either one of these diseases. Alzheimer's disease is the major cause of dementia in the elderly and has become an ever greater problem with the aging of Western societies. Unlike Alzheimer's disease, prion diseases are relatively rare. Each year only approximately 300 people in the USA and approximately 100 people in the UK succumb to various forms of prion diseases [Beekes M (2007) FEBS J274, 575; Sigurdsson EM & Wisniewski T (2005) Exp Rev Vaccines4, 607–610]. Nevertheless, these disorders have received great scientific and public interest due to the fact that they can be transmissible among humans and in certain conditions from animals to humans. The emergence of variant Creutzfeld–Jakob disease demonstrated the transmissibility of the bovine spongiform encephalopathy to humans [Beekes M (2007) FEBS J274, 575]. Therefore, the spread of bovine spongiform encephalopathy across Europe and the recently identified cases in North America have put a large human population at risk of prion infection. It is estimated that at least several thousand Britons are asymptomatic carriers of prion infections and may develop variant Creutzfeld–Jakob disease in the future [Hilton DA (2006) J Pathol208, 134–141]. This delayed emergence of human cases following the near elimination of bovine spongiform encephalopathy in the UK may occur because prion disease have a very prolonged incubation period, ranging from months to decades, which depends on the amount of inoculum, the route of infection and the genetic predisposition of the infected subject [Hilton DA (2006) J Pathol208, 134–141]. Therefore, there is a great need for effective therapies for both Alzheimer's disease and prion diseases.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Alzheimer's disease (AD) and prion disease belong to a category of conformational disorders showing substantial overlap in pathologic mechanism [1–3]. The basic pathomechanism in both disorders is related to a conformational change of normally expressed proteins: amyloid-β (Aβ) in AD and the prion protein (PrP) in prion disease (Fig. 1) [4,5]. This occurs without an alteration in the amino-acid sequence of the proteins. Aβ is a 40–43 amino acid peptide, which, in AD, self-assembles into toxic oligomers and fibrils that accumulate in the brain, forming plaques and deposits in the walls of meningocephalic vessels [6,7]. The same peptide can be detected in most physiological fluids, such as serum or cerebrospinal fluid, where it is called soluble Aβ (sAβ) . PrPC (C-cellular) is a 209 amino acid, cell membrane anchored protein expressed at highest levels by neurons and follicular dendritic cells of the immune system. In the setting of prion disease, this protein undergoes a transformation to toxic PrPSc (Sc-scrapie) [8–10]. Fibrillar Aβ and PrPSc have a high β-sheet content which renders them insoluble, resistant to proteolytic degradation and toxic to neurons. Neurological symptoms in AD and prion disease are directly related to loss of neurons and synaptic connections. Oligomeric and fibrillar Aβ can be directly neurotoxic and/or can promote formation of neurofibrillary tangles . Both fibrillar Aβ and PrPSc are capable of forming amyloid deposits. The presence of amyloid deposits is necessary for making the diagnosis of AD [11,12]. Abundant amyloid deposits composed of PrPSc (full length or fragments) are a neuropathological hallmark of variant Creutzfeld–Jakob disease (vCJD), Gerstmann–Sträusler–Scheinker syndrome (GSS), and kuru . They are also present in 10% of sporadic CJD (sCJD) cases .
A number of proteins may actively promote the conformational transformation of these disease specific proteins and stabilize their abnormal structure. Examples of such proteins in AD include apolipoprotein E (apoE), especially its E4 isoform [13,14], α1-antichymotrypsin (ACT)  or C1q complement factor [16,17] (Fig. 1). In their presence, the formation of Aβ fibrils in a solution of sAβ is much more efficient [13,15]. These ‘pathological chaperone’ proteins have been found histologically and biochemically in association with fibrillar Aβ deposits  but not in preamyloid aggregates, which are not associated with neuronal loss . Similarly, in prion disease, extensive data points toward the existence of an unidentified protein X actively involved in the conversion of PrPC into PrPSc.
AD and prion diseases exist as sporadic and inherited illnesses. In addition, prion disease can be transmitted from one subject to another. In experimental model settings, some evidence also exists for the infectivity of AD [21,22]. An important event in the pathomechanism of AD is thought to be reaching a critical concentration of sAβ and/or chaperone proteins in the brain, at which point the conformational change occurs . This leads to the formation of Aβ aggregates, initiating a neurodegenerative cascade. In sporadic AD, this occurs due to an age-associated overproduction of Aβ, impaired clearance from the brain, and/or influx into the central nervous system (CNS) of sAβ circulating in the serum . Inherited forms of AD are associated with various genetic defects, resulting in overproduction of total sAβ, or more fibrillogenic Aβ 1–42 species .
Sporadic prionoses like sCJD are thought to result from the spontaneous conversion of PrPC into PrPSc. The mechanisms that stabilize PrPC structure are largely unknown but, once PrPSc assumes its pathological conformation, it can bind to PrPC and induce a conformation change. This starts a self-perpetuating vicious cycle allowing PrPSc to replicate without DNA, using the host cell's PrPC as a template [9,26]. Most inherited prionoses such as GSS or inherited forms of CJD are the result of a point mutation in PrPC that increases the propensity for it to assume an abnormal conformation. Virtually all genetic defects implicated in familial forms of AD and prionoses are inherited in an autosomal dominant fashion. Unlike AD, prionoses can be easily transmitted between subjects of the same species. Transmissibility of prionoses between different species is generally ineffective due to differences in the PrP sequence. The phenomenon protecting one species from acquiring a prion disease from another is called ‘the species barrier’. Therefore, scrapie (a prionosis affecting sheep) is not transmissible to humans. The species barrier does not provide absolute protection; therefore, transmission of scrapie to cattle and transmission of bovine spongiform encephalopathy (BSE) from cattle to humans results in the emergence of vCJD. In transmissible prionoses, exogenous PrPSc present in the inoculum is responsible for the conformational transformation of host PrPC. Upon entering an organism, PrPSc initially replicates within the lymphoreticular organs, including the spleen, lymph nodes and tonsils, for months to years prior to neuroinvasion and the onset of neurological symptoms. Therefore, infected but asymptomatic individuals are a reservoir of infectious material. This occurs because PrPC is expressed by follicular dendritic cells and other lymphoid cells . Accumulation of PrPSc in the lymphatic organs of presymptomatic humans infected with BSE has been demonstrated by immunohistochemistry . PrPSc replication is possible because it does not elicit an immune response . This is related to the inability of the immune system to distinguish between PrPC and PrPSc.
Vaccination approaches for AD
Vaccination was the first treatment approach demonstrated to have genuine impact on disease process, at least in animal models of AD. Vaccination of AD transgenic (Tg) mice with Aβ1–42 or Aβ homologous peptides coinjected with Freund's adjuvant prevented the formation of Aβ deposition and, as a consequence, eliminated the behavioral impairments that are related to Aβ deposition [30–35]. Similar effects on Aβ load and behavior have been demonstrated in AD Tg mice by peripheral injections of anti-Aβ monoclonal serum indicating that the therapeutic effect of the vaccine is based primarily on eliciting a humoral response [36,37]. The striking biological effect of the vaccine in preclinical testing and the apparent lack of side-effects in AD Tg mice encouraged Elan/Wyeth to launch clinical trials with a vaccine designated as AN1792 which contained preaggregated Aβ1–42 and QS21 as an adjuvant. This type of vaccine design was aimed to induce a strong cell-mediated immune response because QS21 is known to be a strong inducer of Th-1 lymphocytes . The initial safety testing of AN1792 in phase I of the trial did not demonstrate any adverse effects. The phase II of the trial was prematurely terminated when 6% of vaccinated patients manifested symptoms of acute meningoencephalitis [38,39]. An autopsy performed on one of the affected patients revealed an extensive cytotoxic T-cell reaction surrounding some cerebral vessels; however, analysis of the Aβ load in the brain cortex suggested that Aβ clearance had occurred . It appeared that the immune reaction triggered by AN1792 was a double-edge sword, where the benefits of a humoral response against Aβ were overshadowed in some individuals by uncontrolled cytotoxicity . Not all patients who received AN1792 responded with antibody production. The majority mounted a humoral response and showed a modest but statistically significant cognitive benefit demonstrated as an improvement on some cognitive testing scales compared to baseline and a slowed rate of disease progression compared to patients who did not form antibodies . The follow-up data from the ‘Zurich’s cohort', who are a subset of the Elan/Wyeth trial followed by Dr Nitsch's group [42,43], indicated that the vaccination approach may be beneficial for human AD patients but that the concept of the vaccine has to be redesigned.
It appears that a humoral response elicited by the vaccine has at least two mechanisms of action and both of these are thought to be involved in amyloid clearance [44,45]. Conformational selective anti-Aβ serum may target Aβ deposits in the brain  leading to their disassembly [46,47] and elicit Fc mediated phagocytosis by microglia cells. The second mechanism by which anti-Aβ serum likely prevents Aβ deposition is the creation of a ‘peripheral sink’ effect, where the removal of excess sAβ circulating in the blood stream leads to sAβ being drawn out from the brain [31,34, 47,48]. This peripheral sink mechanism is likely to be the dominant means of reducing Aβ peptides in the brain.
The cause(s) for the toxicity in 6% of the Elan trial patients are not entirely known; however, from the available clinical and limited autopsy data, it is thought that an excessive Th-1 cell-mediated response within the brain was to blame . The concept of a redesigned AD vaccine puts emphasis on avoiding this cell-mediated response in the following ways: (a) avoiding stimulation of Th-1 lymphocytes so the vaccine could potentially elicit a purely humoral response; (b) using nontoxic and nonfibrillogenic Aβ homologous peptides, so that the immunogen can not produce any direct toxicity; and (c) enhancing the peripheral sink effect rather than central action.
Passive transfer of exogenous anti-Aβ monoclonal serum appears to be the easiest way to fulfill the goal of providing anti-Aβ serum without risk of uncontrolled Th-1 mediated autoimmunity. AD Tg model mice treated this way had a significantly reduced Aβ level and demonstrated cognitive benefit [36,37]. The major drawbacks of this approach are the high cost, limited half-life of monoclonal antibodies (2–21 days depending on class and isoform) and the potential for inducing serum sickness with resultant complications such as renal failure or lymphomas. Nevertheless, clinical trials for passive immunization trials are underway. Alternative approaches for passive immunization which are less likely to be associated with toxicity, are use of Fv fragments or mimetics of the active antibody binding site.
Another potential source of toxicity in association with passive immunization is cerebral hemorrhage. The mechanism of this hemorrhage is thought to be inflammation in association with cerebral amyloid deposits (congophilic amyloid angiopathy; CAA) that weakens the blood vessel wall. Several reports have shown an increase in microhemorrhages in different AD mouse models following passive intraperitoneal immunization with different monoclonal antibodies with high affinity for Aβ plaques and CAA [50–52]. The risk of microhemorrhage following active immunization in animal models has not been fully assessed. It has not been a problem in our own active immunization studies [34,35], but has been reported in one study . Furthermore, the clinical trial data from the limited number of autopsied cases suggests that vascular amyloid was not being cleared and that hemorrhage may have been increased [54–56]. In one of these autopsies, numerous cortical bleeds, which are typically rare in AD patients, were evident . In addition, the association of T lymphocytosis and cuffing with the cerebral vessel Aβ in these autopsies suggests a potential role of CAA and an excessive Th-1 response in the genesis of the inflammatory side-effects . This is an important issue because CAA is present in virtually all AD cases, with approximately 20% of AD patients having ‘severe’ CAA . Furthermore CAA is present in approximately 33% of cognitively normal elderly, control populations [59–61].
Understanding the antigenic profile of Aβ peptide, allows engineering of modifications that favor a humoral response and reduce the potential for a Th-1 mediated response. This approach has been termed altered peptide ligands. Computer models have predicted that Aβ1–42 has one major antibody binding site located on its N-terminus and two major T-cell epitopes located at the central and C-terminal hydrophobic regions encompassing residues 17–21 and 29–42, respectively [62–64]. Therefore, their elimination or modification provides a double gain by eliminating toxicity, as well as the potential for T-cell stimulation. Sigurdsson et al.  immunized AD Tg mice with K6Aβ1–30[E18E19], a nontoxic Aβ-homologous peptide, where the first above mentioned T-cell epitope was modified and the second removed. Polyamino acid chains coupled to its N-terminus aimed to increase the immunogenicity and solubility of the peptide. AD Tg mice vaccinated with this peptide produced mainly an IgM class antibodies and low or absent IgG titer. These animals showed behavioral improvement and a partial reduction of Aβ deposits [34,35]. One of the advantages of this design is that IgM, with a molecular mass of 900 kDa, does not penetrate the blood–brain barrier (BBB) and therefore is unlikely to be associated with any immune reaction in the brain. Like passive immunization, this type of vaccine focuses its mechanism of action on the peripheral sink. Furthermore, the IgM response is reversible because it is T-cell independent; hence memory T-cells that could maintain the immune response are not generated. Therefore, this vaccine method may potentially be safer than typical active immunization.
Mucosal vaccination can be an alternative way to achieve a primarily humoral response. This mechanism is based on the presence of lymphocytes in the mucosa of the nasal cavity and of the gastrointestinal tract. This type of response produces primarily S-IgA antibodies but, when the antigen is coadministrated with adjuvants such as cholera toxin subunit B or heat-labile Escherichia coli enterotoxin, significant IgG titer in the serum may be achieved [65,66]. A marked reduction of Aβ burden in AD Tg mice immunized this way using Aβ as an antigen has been already demonstrated [66,67]. Interestingly, this type of mucosal immunization has recently been shown to be highly effective for prion infection [68,69,70]. This promising approach requires further exploration, especially using nonfibrillar and nontoxic Aβ homologous peptides as an antigen. Mucosal immunization offers a great potential advantage in that a more limited humoral immune response can be obtained, with little or no cell-mediated immunity.
Inhibition of Aβ fibrillization
Formation of Aβ fibrils and deposition of Aβ in the brain parenchyma or in the brain's vessels occurs in the setting of increased local Aβ peptide concentrations . Initially, conditions do not favor aggregation of fibrils; however, once a critical nucleus has been formed, aggregation with fast kinetics is favored. Any available monomer can then become entrapped in an aggregate or fibril. Several compounds, such as Congo red , anthracycline , rifampicin , anionic sulphonates , or melatonin , can interact with Aβ and prevent its aggregation into fibrils in vitro, thereby reducing toxicity. It has been further identified that certain nonfibrillogenic, Aβ homologous peptides can bind to Aβ and break the formation of β-sheet structure [77–80]. Therefore, these peptides were termed β-sheet breakers. Several modifications were used to extend serum half-life and increase BBB permeability of these peptides. Permanne et al. , using a BBB permeable five amino-acid long peptide (iAβ5), were able to demonstrate a reduction of Aβ load in AD Tg mice that received this peptide compared with age-matched control group which received placebo. Of interest, a similar concept of β-sheet breakers has been shown to be applicable to prion disease .
Extensive evidence suggests that the most toxic forms of Aβ are oligomeric aggregates . There is also evidence implicating oligomeric aggregates in the mediation of PrPSc toxicity and infectivity [84,85]. Recently, compounds and antibodies have been developed that specifically target Aβ oligomers [86–88]. Similar approaches are being developed for prion oligomers.
Aβ homologous peptides can aggregate and form fibrils spontaneously in vitro; however, in vivo this process appears more dependant on the presence of Aβ pathological chaperones. This group of proteins promotes conformational transformation at certain concentrations by increasing the β-sheet content of these disease specific proteins and stabilizes their abnormal structure [89,90]. Examples of such proteins in AD include apoE, especially its E4 isoform [18,91], ACT  or C1q complement factor [21,22]. In their presence, the formation of Aβ fibrils in a solution of sAβ monomers becomes much more efficient [18,20]. These ‘pathological chaperone’ proteins have been found histologically and biochemically in association with fibrillar Aβ deposits [23,89,92,93] but not in preamyloid aggregates that are not associated with neuronal toxicity [24,94]. Inheritance of the apoE4 isoform has been identified as the major identified genetic risk factor for sporadic, late-onset AD  and correlates with an earlier age of onset and greater Aβ deposition, in an allele-dose-dependent manner [19,95,96]. In vitro, all apoE isoforms can propagate the β-sheet content of Aβ peptides promoting fibril formation , with apoE4 being the most efficient . The critical dependence of Aβ deposition in plaques on the presence of apoE has also been confirmed in AD Tg APPV717F/apoE–/– mice which have a delayed onset of Aβ deposition, a reduced Aβ load, and no fibrillar Aβ deposits. Compared to APPV717F/apoE+/+ Tg mice, APPV717F/apoE+/– mice demonstrate an intermediate level of pathology [97–100]. Neutralization of the chaperoning effect of apoE would therefore potentially have a mitigating effect on Aβ accumulation. ApoE hydrophobically binds to the 12–28 amino acid sequence of Aβ, forming SDS insoluble complexes [101–103]. Ma et al.  have demonstrated that a synthetic peptide homologous to 12–28 amino-acid sequence of Aβ can be used as a competitive inhibitor of the binding of full length Aβ to apoE, resulting in reduced fibril formation in vitro and increased survival of cultured neurons. The introduction of several modifications to Aβ12–28 by replacing a valine for proline in position 18, making this peptide nontoxic and nonfibrillogenic, as well as end-protection by amidation and and acetylation of the C- and N-termini, respectively, to increase serum half-life, have allowed us to use this peptide therapeutically in the APPK670N/M671L/PS1M146L double Tg mice model. Tg mice treated with Aβ12–28P for 1 month demonstrated a 63.3% reduction in Aβ load in the cortex (P = 0.0043) and a 59.5% (P = 0.0087) reduction in the hippocampus comparing to age-matched control Tg mice that received placebo [105,106]. The treated Tg mice also had a cognitive benefit [105,106]. No antibodies against Aβ were detected in sera of treated mice; therefore, the observed therapeutic effect of Aβ12–28P cannot be attributed to an antibody clearance response. This experiment demonstrates that compounds blocking the interaction between Aβ and its pathological chaperones may be beneficial for treatment of Aβ accumulation in AD [14,105,106]. Whether similar approaches can be used for prion disease remains to be determined.
Interest in prion disease has greatly increased subsequent to the emergence of BSE in England and the resulting appearance of vCJD in human populations. BSE arose from the feeding of cattle with prion contaminated meat and bone meal products, whereas vCJD developed following entry of BSE into the human food chain [107,108]. Since the original report in 1995, a total of 201 probable or confirmed cases of vCJD have been diagnosed, 165 in Great Britain, 21 in France, four in Ireland, three in the USA, two in the Netherlands and one each in Italy, Canada, Japan, Saudi Arabia, Portugal and Spain. Most of the patients from these countries resided in the UK during a key exposure period of the UK population to the BSE agent. It has proven difficult to predict the expected future numbers of vCJD. Mathematical analysis has given a range from 1000 to approximately 136 000 individuals who will eventually develop the disease. This broad range reflects a lack of knowledge regarding the time of incubation and the number of patients who could be infected from a given dosage of BSE agent. Because the vCJD agent is present at high levels in the lymphatic tissue, screening for PrPSc was performed on sections from lymph nodes, tonsils, and appendices archives in the UK. Three out of 12 674 randomly selected cases showed evidence of subclinical infection, leading to a prediction that approximately 4000 vCJD further cases may occur in the UK . However, there is much uncertainty about such a prediction because it is not known whether all subclinical infections will progress and also whether such screening of lymphoid tissue would capture all subclinical cases. The initially predicted epidemic of vCJD does not seem to be materializing because the number of cases in the UK has declined from a peak of 28 in 2000 to five cases in 2006 . A complicating factor for estimating future numbers of vCJD is the documentation of several transfusion associated cases. These occurred after incubation periods of 6–8 years. One of these disease associated donations was made more than 3 years before the donor became symptomatic, suggesting that vCJD can be transmitted from silently infected individuals . The estimated risk for new cases of vCJD in other European countries looks more optimistic. In the UK, 200 000 cases of BSE were reported (it is estimated that four times this number entered the food chain), compared to approximately 5600 BSE cases in other European countries (with the highest numbers being 1590, 1030 and 986 in Ireland, Portugal and Frances, respectively). This suggests a significantly lower exposure of these populations to BSE prions. A few cases of BSE have also been reported in other parts of the world, such as Japan, the USA and Canada.
Of greater concern in North America is chronic wasting disease (CWD). This disease is now endemic in Colorado, Wyoming and Nebraska and continues to spread to other parts in the USA, initially in the Midwest but now detected as far East as New York State [111,112]. Most vulnerable to CWD infection are white tailed deer and the disease is now found in areas with a large population of these animals, which indicates that its prevalence can be expected to increase substantially in the future. The occurrence of CJD among three young deer hunters from this same region raised the speculation of transmission of the CWD to humans . However, autopsy of these three subjects did not reveal the extensive amyloidosis characteristic of vCJD and CWD . However like BSE, CWD is transmissible to nonhuman primates and transgenic mice expressing human PrPC[115,116]. Therefore, the possibility of such transmission needs to be closely monitored. CWD is similar to BSE in that the peripheral titers of the prion agent are high. PrPSc has been detected in both muscle and saliva of CWD infected deer [117,118].
Vaccination as a therapeutic approach for prionoses
The prion protein is a self-antigen; hence, prion infection is not known to elicit a classical immune response. In fact, the immune system is involved in the peripheral replication of the prion agent and its ultimate access to the CNS [29,68]. This involvement is further supported by the observation that immune suppression with, for example, splenectomy or immunosuppressive drugs, increases the incubation period. This interval, during which time the prion agent replicates peripherally, without producing any symptoms, is quite long, lasting many months in experimental animals and up to 56 years in documented human cases associated with cannibalistic exposure to the prion agent . Lymphatic organs such as the spleen, tonsils, lymph nodes or gut associated lymphoid tissue contain high concentrations of PrPSc long before PrPSc replication starts in the brain [27,120,121]. Cells found to be particularly important for peripheral PrPSc replication are the follicular dendritic cells (DC) and the migratory bone-marrow derived DC [121,122]. DC from infected animals are capable of spreading the disease . An emerging therapeutic approach for prion infection is immunomodulation [68,70,123].
Currently, there is no treatment that would arrest and/or reverse progression of prion disease in nonexperimental settings, although many approaches have been tried . Partly due to the success in AD models discussed above, similar experiments with anti-PrP serum were initiated in prion infectivity culture models as well as active and passive immunization studies in rodent models. Earlier in vivo studies showed that infection with a slow strain of PrPSc blocked expression of a more virulent fast strain of PrP, mimicking vaccination with a live attenuated organism . In tissue culture studies, anti-PrP serum and antigen binding fragments directed against PrP were shown to inhibit prion replication [126–128]. Although we first demonstrated that active immunization with recombinant PrP delayed the onset of prion disease in wild-type mice, the therapeutic effect was relatively modest and, eventually, all the mice succumbed to the disease . This limited therapeutic effect may be explained by the observation that antibodies generated against prokaryotic PrP often do not have a high affinity towards PrPC, although, in our studies, the increase in the incubation period correlated well with the antibody titers against PrPC. Our follow-up passive anti-PrP immunization study confirmed the importance of the humoral response, showing that anti-PrP serum is able to prolong the incubation period . Subsequently, other investigators, using a much higher antibody dosage, were able to completely prevent disease onset in mice exposed to PrPSc provided that passive immunization was initiated within 1 month of exposure . This type of approach could be used immediately following accidental exposure in humans to prevent future infection. However, passive immunization has not been found to be effective closer to the clinically symptomatic stages of prion infection. Also, passive immunization would be an approach that is too costly for animal prion diseases.
In the development of immunotherapeutic approaches targeting a self-antigen, designing a vaccine avoiding auto-immune related toxicity is a major concern. The emerging data from AD targeting immunization is that toxicity is due to excessive cell-mediated immunity within the CNS, whereas the therapeutic response is linked to humoral immunity. In addition, toxicity could be partially related to the immunogen and/or to the adjuvant used; in the human AD vaccination trial, fibrillar Aβ1–42 was used as an immunogen. This peptide is well characterized to be toxic. Hence, we have been promoting the use of nonamyloidogenic derivatives as immunogens for protein conformational disorders, including AD and prion disease [31,34,38]. How significant an issue direct toxicity of the immunogen may be for prion vaccination remains unclear. Unlike the Aβ peptide used for vaccination in AD models, direct application of recombinant PrP has not been shown to be toxic. However, this issue has not been investigated as thoroughly as in the Alzheimer's field and remains controversial. Several lines of evidence suggest that intracellular accumulations of PrPSc promote neurodegeneration .
A potential ideal means of using immunomodulation to prevent prion infection is by mucosal immunization. One important reason for this is that the gut is the major route of entry for many prion diseases such as CWD, BSE and vCJD. Furthermore, mucosal immunization can be designed to induce primarily a humoral immune response, avoiding the cell-mediated toxicity that was seen in the human AD vaccine trial. In addition, mucosal vaccination has the advantage that it is unlikely to induce significant immune response within the brain. Although it has been shown that reduced levels or absence of CNS PrPC by, for example, conditional ablation by genetic manipulation of neuronal PrPC can prevent clinical prion infection, it is likely that the immunological targeting of neuronal PrP would be associated with inflammatory toxicity. Recently, we have been developing prion vaccines that target gut associated tissue, the main site of entry of the prion agent. One of our approaches is to express PrP in attenuated Salmonella strains as a live vector for oral vaccination, which has resulted in prevention or significant delay of prion disease in mice . Live attenuated strains of Salmonella enterica have been used for many years as vaccines against salmonellosis and as a delivery system for the construction of multivalent vaccines with a broad application in human and veterinary medicine . A main advantage for this system is that the safety of human administration of live attenuated Salmonella has been extensively confirmed in humans and animals [136,137]. Ruminants and other veterinary species can be effectively immunized by the oral route using attenuated Salmonella, to induce humoral mucosal responses [138,139]. We are currently exploring ways to increase the efficacy even further. In these studies, the mucosal IgA anti-PrP titer correlates well with the delay or prevention of prion infection, further supporting the importance of the humoral response for the therapeutic effect. Salmonella target M-cells, antigen sampling cells in the intestines, which may also be important for uptake of PrPSc[27,68,121]. Hence, this approach is more targeted than prior vaccination studies, likely explaining the improved efficacy. By exploring other strains of attenuated Salmonella, using different bacteria or oral adjuvants, and/or by altering the expression levels or sequence of the PrP antigen, it is likely that the percentage of uninfected animals can be improved. Our recent work utilizing this approach indicates that complete protection to clinical prion infection via an oral route is possible. Overall, this approach holds great promise as an inexpensive prophylactic immunotherapy to prevent the spread of prion disease, particularly in animals at risk and perhaps eventually in certain high risk human populations.
Metal chelation for prion and AD
Metal chelation is emerging as an important therapeutic approach for AD, which is currently in clinical trial [140,141]. This approach for AD is reviewed elsewhere in this minireview series. Importantly, modulation of metal levels, in particular copper, has been shown to be important for the conversion of PrPC to PrPSc, highlighting another similarity between AD and prion diseases . Copper binding is thought to be part of the normal function of PrPC[142–144]. The binding of copper to PrPC gives the complex antioxidant activity [145,146]; hence, it has been suggested that the reduced copper binding of PrPSc with a consequent reduction of antioxidant activity is part of the pathogenesis of prion disease . This hypothesis has been supported by the finding that copper is reduced up to 50% in the brains of sporadic CJD patients . How copper binding influences the PrPC to PrPSc conversion is complex [10,149]. We were the first to show that, similar to studies in AD Tg models, metal chelation can be used therapeutically  in prion infection. Our studies indicated that penicillamine, a copper chelator, prolongs the incubation period of scrapie in mice . Consistent with this observation, the presence of copper has also been shown to stabilize the PrPSc conformation using preformed fibrils [151–158], as well as to induce aggregation of the prion peptide 106–126 . Some tissue culture studies of prion infection have also suggested that copper chelators are suitable candidates for antiprion drugs . However, there are conflicting reports indicating that the interaction between copper and PrP is likely to be quite complex. For example, copper has been shown to inhibit the in vitro conversion of recombinant PrP into amyloid fibrils but, also in contrast, to enhance the proteinase K resistance of preformed fibrils . These findings indicate that copper may have a dual and opposite effect on prion propagation. It may both inhibit prion replication and prevent clearance of potentially infectious forms of the prion protein. Furthermore, copper treatment has also been shown to inhibit PrPSc amplification in reactions where brain derived PrPC was used as a seed , as well as delaying the onset of clinical disease in scrapie infected hamsters . In addition, it has been shown that physiological levels of copper promote internalization of PrPC. The interaction between PrPC and copper was found to be the overriding factor in stimulating the internalization response with other metals showing no effect. The decrease in detectable levels of PrPC at the cell surface following copper treatment was found to be the result of internalization rather than loss into the surrounding environment . Such internalization would limit the exposure of PrPC to conversion from exogenous PrPSc; however, because cytoplasmic forms of PrP have been linked to neurodegeneration , increased internalization could also be deleterious in some settings. Copper has also been shown to have immunomodulatory effects  and, as discussed earlier, the immune system can have profound effects on prion infection. Hence, it appears that the deleterious or beneficial role of copper in prion infection might vary depending on which function predominates under the distinct experimental conditions being used. Nevertheless, it is clear that a greater understanding of the role of metal binding in prion infection presents a therapeutic opportunity.
Immunization appears to be an effective therapeutic method for prevention of Aβ deposition and cognitive decline in AD, provided that cell-mediated auto-immune toxicity can be avoided. The second generation AD vaccines, which are under development, are based on nontoxic and nonfibrillar Aβ homologous peptides that are modified to eliminate the potential for inducing cellular immunity, and elicit primarily a humoral response. Other related approaches include direct administration of antibodies that target Aβ. These interventions would likely favor a peripheral sink effect, clearing soluble Aβ from the blood stream and inducing efflux of Aβ from the brain. Additional potentially synergistic therapeutic approaches for AD would include blocking the interaction of Aβ with its ‘pathological chaperones’ such as apoE, as well as use of β-sheet breaker compounds. Immunization approaches could be used for sporadic AD, familial AD, and AD associated with Down's syndrome. The effectiveness of treatment would depend on its initiation early in the disease course. Therefore, such a treatment needs to coincide with the development of a procedure for the detection and monitoring of Aβ deposits.
Both active and passive immunization appear to be effective in prevention of prion infections in animal models. Further studies are needed to develop specific protocols applicable for human use. Active immunization, using especially mucosal immunization could be used to prevent spread of BSE through the oral route, whereas passive immunization protocols would be more appropriate for subjects accidentally infected with prion contaminated material (e.g. blood transfusion or organ transplant). Effective immunization for prion infections works through prevention of entry of PrPSc via the gut and/or neutralization of PrPSc replicating in the peripheral lymphoreticular system.
Metal chelation is another promising therapeutic approach for AD, which is currently undergoing clinical trials. Similar approaches are just emerging for prion diseases. However, a greater understanding of the role of copper and other metals in the PrPC to PrPSc conversion is needed before this therapeutic strategy can be effectively harnessed for prion infection.
This manuscript is supported by NIH grants: AG15408, AG20245, AG20197 and the Alzheimer's Association.