Neuroprotective Effect and Potential of Cellular Prion Protein and Its Cleavage Products for Treatment of Neurodegenerative Disorders Part II. Strategies for Therapeutics Development

Abstract

Introduction:

Cellular prion protein (PrP^C), some of its derivatives (especially PrP N-terminal N1 peptide and shed PrP), and PrP^C-containing exosomes have strong neuroprotective activities, which have been reviewed in the companion article (Part I) and are briefly summarized here.

Areas covered:

We propose that elevating the extracellular levels of a protective PrP form using gene therapy and other approaches is a very promising novel avenue for prophylactic and therapeutic treatments against prion disease, Alzheimer’s disease and several other neurodegenerative diseases. We will dissect the pros and cons of various potential PrP-based treatment options and propose a few strategies that are more likely to succeed. The cited references were obtained from extensive PubMed searches of recent literature, including peer-reviewed original articles and review articles.

Expert opinions:

Concurrent knockdown of cellular PrP expression and elevation of the extracellular levels of a neuroprotective PrP N-terminal peptide via optimized gene therapy vectors is a highly promising broad-spectrum prophylactic and therapeutic strategy against several neurodegenerative diseases, including prion diseases, Alzheimer’s disease and Parkinson’s disease.

1. Overview of protective extracellular PrP forms

The cellular prion protein (PrP^C) is a glycosylphosphatidylinositol (GPI)-anchored glycoprotein highly expressed in the nervous system and at various levels in lymphoid, muscle and other tissues^1–6.

The biological and pathological roles of PrP are complicated. On one hand, PrP^C is essential for both the replication and pathogenesis of the prion agents in prion diseases (PrD)^7–9, and it plays a key role in the pathogenesis of other neurodegenerative diseases [such as Alzheimer’s (AD) and Parkinson’s (PD)], primarily serving as a receptor for the cytotoxic oligomers of amyloid-β (Aβ), Tau, and α-synuclein (αSyn)^10–17. In addition, elevated PrP^C is associated with several cancers^18–22. On the other hand, cell surface full-length PrP^C seems to play an important role in many biological processes^5–6, such as myelin maintenance²³ and neurite outgrowth and neuroprotective signaling^24–26. Some other PrP forms and PrP processing are also protective or beneficial, which has been reviewed in a companion paper²⁷ and is briefly summarized below.

The PrP N-terminal fragment derived from α-cleavage of PrP^C, extracellular full-length PrP from shedding of cell surface PrP^C, and exosomes containing PrP^C have all shown beneficial and protective effects. The α-cleavage of PrP^C occurs at aa110–111 or aa111–112, yielding the membrane-attached C-terminal C1 fragment²⁸ and the N-terminal N1 peptide (88–89 amino acid residues) that is released into the extracellular space. A group of enzymes have been implicated in the α-cleavage of PrP^C, including ADAM8, ADAM9, ADAM10, and ADAM17, but all except ADAM8 are still controversial^29–35. It also appears that more than one enzyme is involved in the α-cleavage of PrP^C in the CNS and other tissues^31,33–34.

The N1 peptide is protective against reactive oxygen species^36–37. It also protects against toxic Aβ oligomers in AD through inhibiting Aβ oligomerization, accelerating fibril formation (thereby reducing levels of the oligomeric forms), and sequestering Aβ oligomers to prevent toxic signaling and alleviate long-term potentiation (LTP) inhibition^38–40. The C1 fragment has been reported to inhibit prion replication³⁶ and its negative activities^42–44 seems to be no significant concerns^36,41. Although N1 has not been directly shown to bind to and neutralize Tau and αSyn oligomers, the regions in full-length PrP required for binding to these toxic oligomers are fully contained within N1¹⁷, so it is reasonable to assume that N1 can neutralize these two oligomers as well.

Recombinant full-length PrP has been reported to induce neuritogenesis^45–46, and binds to and provides protection against toxic Aβ species^38,47. Transgenically expressed anchorless PrP, which is basically the same as the shed PrP, has also been reported to reduce LTP impairment in an AD mouse model⁴⁸.

Shed PrP is released to the extracellular space through cleavages by a sheddase at or near the C-terminus of PrP^C (aa228–231) or within the GPI anchor^30,49–51. ADAM10 appears to be the primary sheddase for PrP^{C 30,32,35,52}, and the PrP shedding activity of ADAM10 is modulated by ADAM9³⁰. While there is no direct evidence demonstrating the neuroprotective activity of shed PrP, shed PrP is expected to show similar protective effect as that of recombinant full-length PrP since it is very similar to recombinant full-length PrP except for its O-linked glycans. In addition, PrP shedding is presumed to be protective against prion diseases and other misfolding diseases since it is supposed to decrease cell surface PrP^C levels^32,53 thereby reducing the levels of the essential substrate for prion replication and a critical receptor for toxic protein oligomers^17,54–56.

PrP on exosomes, due to their extracellular location, will not mediate the pathogenesis of toxic oligomers, is a third form of protective extracellular PrP that should be safer for therapeutic purpose^57–58. PrP^C on exosomes has been shown to be protective against Aβ toxicity through accelerating Aβ₄₂ aggregation into its fibrillar form, reducing the uptake of synthetic Aβ, abolishing Aβ₄₂-induced apoptosis^57–59, and ameliorating Aβ-induced synaptic loss and LTP inhibition in rodent AD models^57,59. Since recombinant and cellular PrP can also bind to toxic Tau and αSyn oligomers¹⁷, it is reasonable to expect the same protective effect against these toxic oligomers from exosomal PrP.

It is very tempting to harness the protective and beneficial activities of some of the PrP forms and PrP processing described above for therapeutic and preventative treatments against PrD, AD, PD and a few other protein misfolding neurodegenerative diseases involving toxic protein oligomers. Here we present and evaluate various strategies for the development of prophylactic and therapeutic treatments against AD and other neurodegenerative diseases through elevating the extracellular levels of one of the protective PrP forms. Knocking down PrP is also a very attractive strategy to battle a number of neurodegenerative diseases where PrP^C plays a critical role in mediating toxicity, but this topic has been reviewed very recently⁶⁰ and will not be discussed in depth here.

2. Strategies for developing PrP-based therapies

We propose that augmenting the extracellular levels of protective PrP forms is an attractive and broad-spectrum strategy for effective treatment and prevention of many neurodegenerative diseases involving toxic oligomers of PrP, Tau, or α-synuclein, including prion diseases, AD, PD, and other tauopathies (such as progressive supranuclear palsy and frontotemporal lobar degeneration) and synucleinopathies (such as dementia with Lewy bodies, multiple system atrophy, and Lewy body variant of AD). Several PrP forms have been shown to, or are expected to, have protective activities when existing in the extracellular space, and their extracellular levels can be enhanced by a variety of approaches. These include (1) full-length PrP that is shed from cell surface, administered as recombinant PrP, or expressed as anchorless PrP from a gene therapy vector, (2) full-length PrP on exosomes, and (3) N1 peptide released by α-cleavage of PrP^C on cell surface, administered as a recombinant peptide, expressed as a secreted peptide from a gene therapy vector, or processed by α-cleavage of a fusion protein expressed from a gene therapy vector. Here we evaluate these strategies to elevate the extracellular levels of each of these protective PrP forms and propose the most feasible options for prophylactic and therapeutic purposes.

2.1. PrP N-terminal fragment-based strategies

The PrP N1 peptide protects cells against toxic Aβ oligomers, and it is likely effective against Tau and αSyn oligomers. No significant risks have been identified with extracellular N1, although its reported impact on cell cycle progression³⁶ needs to be assessed. These factors make extracellular N1 and similar extracellular PrP N-terminal peptides highly promising candidates for prophylactic and therapeutic treatments of AD and other neurodegenerative diseases. For simplicity, we will use PrP-N to represent N1 and similar extracellular PrP N-terminal peptides.

There are four main ways to elevate extracellular PrP-N: (1) direct administration of recombinant PrP-N into the ventricular system of the brain, central canal of the spinal cord, or hippocampus and other brain regions, (2) enhancing PrP^C α-cleavage activities, (3) CNS expression of a secretive form of PrP-N from a gene therapy vector, and (4) CNS expression of a cell membrane-anchored form of PrP from a gene therapy vector that will be subjected to α-cleavage to release N1 into extracellular space. Some of these strategies have been discussed in an earlier review⁶¹.

Enhancing extracellular PrP-N levels through direct administration of a recombinant or synthetic PrP-N peptide is very attractive because it appears to be straightforward and mature technologies exist (Figure 1). There have been reports showing that the cytotoxicity of Aβ aggregates is significantly reduced in vitro or in vivo after preincubation with synthetic PrP N-terminal peptides (PrP^23–50 or PrP^90–112)⁶² or recombinant N1 (PrP23–110)³⁹, although there has been no report of direct infusion with recombinant PrP-N in AD mouse models. In fact, there are many challenges common for the recombinant protein strategy: high costs, biostability, possible immune reactions especially after repeated injections, and detrimental effects on target and non-target cells. For CNS applications, there are also the increased risks of infection and pain associated with repeated invasive intracranial/intrathecal injections. In addition, the recombinant PrP-N strategy will be very costly and likely ineffective if an unmodified N1 peptide is used. The PrP N-terminal domain is unstructured. We and others have found that N1 is highly unstable and easily degraded in tissues^63–64, suggesting that exogenously administered recombinant N1 may have very short half-life in vivo. It will also require a large amount of the recombinant PrP-N, which is expensive to produce. It is possible to design synthetic modified short PrP-N peptides that have high stability and long half-life in vivo^65–67, but it will require a lot of experimentation to optimize and validate. Alternatively, fusion with a structured domain may stabilize the N1 peptide as suggested⁶⁴.

Figure 1. Strategies to develop PrP-based prophylactic and therapeutic treatments against AD and other neurodegenerative diseases.

Protective extracellular PrPs (full-length, N1 or similar N-terminal fragments [termed PrP-N]) can sequester toxic protein oligomers and neutralize their toxicities in AD and other neurodegenerative diseases. Extracellular PrP levels can be elevated through four approaches. First, direct administration of in vitro prepared PrP^C-rich exosomes, recombinant N1, PrP-N or full-length PrP. Second, expression from a gene therapy vector of the secreted N1 or PrP-N, a PrP α-cleavage enzyme, a PrP shedding enzyme, or a protein activating the above α-cleavage or shedding enzymes. Third, enhancing the α-cleavage or shedding of cell surface PrP^C through upregulation of the expression of an endogenous gene coding for an α-cleavage or shedding enzyme, or an endogenous gene whose encoded protein activates these endogenous enzymes. Fourth, direct activation of the enzymatic activity of an endogenous PrP^C α-cleavage or shedding enzyme. For simplicity, the third and fourth strategies have been simplified as a single concept of increasing α-cleavage or shedding for enhanced production of protective PrP forms in the diagram.

Elevating extracellular PrP-N levels using a gene therapy vector to express PrP-N in the CNS is another appealing strategy (Figure 1). The gene therapy techniques, especially with recombinant adeno-associated virus (rAAV) as the vector, have matured to the point of approved clinical applications for some single-gene diseases in the last few years^68–70. This strategy has the potential to achieve sustained elevated levels of extracellular N1 or PrP-N for years after a single injection of the gene therapy vector. It will also largely bypass the issue of N1 instability through continuous production and secretion of N1 or PrP-N and will likely achieve a stable elevated N1 or PrP-N concentration in the extracellular space, which may enhance the therapeutic benefits. Mohammadi et al.^63–64 successfully generated transgenic mice overexpressing N1 without the GPI-anchor signal peptide, but the transgene-expressed N1 was found to be trapped inside the cells due to impaired endoplasmic reticulum translocation, and no protection against Aβ-mediated synaptotoxicity was found in an AD mouse model. Although fusion of a structured domain to the C-terminus of N1 might overcome the intracellular translocation defect of N1⁶⁴, the potential side effects of the fusion tag need to be thoroughly studied. In comparison, adding a structured domain from PrP^C itself to the N1 fragment seems to be an attractive alternative since it may not only stabilize the N1 peptide but also solve the secretion defect of N1 without adding a foreign fusion tag. We have demonstrated the feasibility of this approach (unpublished data), but whether the added structured PrP domain will affect the protective activity or leads to undesirable biological activities is still unclear and must be investigated. The gene therapy approach itself also has significant risks, including immunogenicity that will cause inflammation and reduce effectiveness (for viral vectors), chromosomal integrations of the viral DNA (although uncommon for rAAV vectors) that could cause cancers, and high costs and significant difficulty in manufacturing of high-titer viral particles. These concerns can be largely addressed when a non-viral gene therapy vector is used, but the poor transduction efficiency and short duration of gene expression associated with non-viral vectors still need to be significantly improved before clinical use.

Extracellular PrP-N levels can also be augmented through heightened α-cleavage activities, but this strategy is hindered by the uncertainty of the identity of the protease(s) responsible for PrP^C α-cleavage. Although several reports implicated the involvement of ADAM8, ADAM9, ADAM10, and ADAM17^{29–32,34–35}, all but ADAM8 remain controversial. The identity of the true α-secretase(s) other than ADAM8 for PrP^C must be clarified first.

The selection of an optimal α-secretase ADAM for augmentation will be both a challenging task and a critical factor for outcomes. Of the known ADAMs implicated in PrP processing, ADAM10 was initially reported to contribute to the α-cleavage of PrP^{C 29,31,35,71}, and it seems to be an attractive candidate to be enhanced for a number of reasons described below. But the PrP^C α-cleavage activity of ADAM10 is strongly disputed and appears to be ruled out by later experiments with neuron-specific conditional ADAM10-knockout or ADAM10-overexpressing mice or cells^30,32,54. Nevertheless, ADAM10 remains an interesting protein for therapeutic purposes even if it is not involved in PrP^C α-cleavage. One of the reasons is that ADAM10 is the primary PrP sheddase⁵⁶ that produces an extracellular PrP form (termed shed PrP) with protective activities^38,45–48, and it plays a key role in the non-amyloidogenic α-cleavage of APP^72–73, making it a notable target for AD therapy development^72–74. Transgenic mice overexpressing ADAM10 in neurons showed increased neurotrophic N-terminal APP domain (APPsα), reduced soluble Aβ peptides, diminished Aβ plaques, and partially rescued LTP⁷⁴. Additionally, the increased levels of ADAM10 did not seem to cause significant detrimental phenotypic alterations^{29,54,74–75}, suggesting that enhancing APP α-cleavage activities through augmenting ADAM10 is a viable strategy. It is worth noting that high-level neuronal overexpression of ADAM10 was reported to cause more seizures and stronger neuronal damage and inflammation after kainate-induced epileptic seizures⁷⁶. In addition, ADAM10 is known to promote cancer⁷⁷ and it is a target for cancer and autoimmunity therapies⁷⁸, suggesting that enhancement of ADAM10 activity needs to be done carefully and kept at moderate levels. Moreover, transgenic neuronal overexpression of ADAM10 in mice led to a decrease in total PrP levels rather than an increase of specific PrP^C cleavage products as well as enhanced survival of the mice when challenged with prions⁵⁴. If the same is true when the ADAM10 activity is elevated through administration of recombinant ADAM10 or expression from a gene therapy vector, augmenting ADAM10 should still protect against AD, PD, and other neurodegenerative diseases through decreasing cell surface PrP^C levels. Whether ADAM8 and/or the yet-to-be identified other PrP^C α-cleavage enzyme(s) can be targeted to fight AD and other neurodegenerative diseases awaits further studies.

The PrP α-cleavage activities can be enhanced by different approaches (Figure 1): (1) infusion with a bioactive recombinant α-secretase, (2) intracranial expression of a α-secretase from a gene therapy vector, (3) stimulation of the expression of an endogenous α-secretase gene, and (4) enhancement of the activity of an endogenous α-secretase protein.

Infusion with a recombinant α-secretase appears to be a simple solution, but it could be complicated in reality. First, as discussed earlier, we need to select an α-secretase that will have a strong impact on the PrP^C α-cleavage activities in the CNS yet with minimal side effects, and it is critical to factor in the substrate spectrum of the α-secretase. All ADAMs implicated in PrP^C α-cleavage have many substrates⁷⁹ and enhanced shedding of some of the unintended substrates due to elevated activity of one PrP^C α-cleavage enzyme may have serious deleterious consequences, such as promoting inflammation or tumor metastasis. Second, we must find the best route for administration of the chosen recombinant α-secretase. The expression of α-secretase is usually region and cell type specific in the brain; conceptually, a broad increase of α-secretase activity in the CNS could lead to detrimental cleavages of some substrates in undesirable brain regions and/or cell types. This concern can be partially addressed via localized administration of the recombinant α-secretase to a proper brain region. Third, ideally a modified or variant form of the recombinant α-secretase that does not promote cancer progression can be designed for therapeutic purposes, since all ADAM enzymes implicated in PrP^C α-cleavage have been reported to play a role in promoting cancers^77–80. Administration of a recombinant precursor (inactive) form of the chosen α-secretase will likely be safer since natural cellular regulations may limit the activation of the recombinant precursor to brain areas where this α-secretase is normally expressed. Fourth, the common limitations of recombinant protein therapy described earlier also need to be addressed.

Increased PrP α-cleavage activities can also be achieved through expression of a α-cleavage enzyme from a gene therapy vector (Figure 1), which will come with all the benefits of the gene therapy approach, including sustained effect after a single treatment. Nevertheless, there are many issues to consider with this approach. First, same as for the recombinant protein approach, the choice of the α-secretase enzyme is critical and expression of its precursor form may be safer. Second, the selection/design of an optimal gene therapy vector is paramount that will enable efficient transduction of the CNS cells and effective and authentic cell-type specific expression of the α-secretase gene. The vector backbone and the gene expression cassette for the selected α-secretase need to be evaluated for efficiency of cell entry, specificity of cell targeting, and vector-carried α-secretase gene transcription activity in CNS cells. Furthermore, a choice must be made between integrating type of vectors (retroviral vectors) and non-integrating vectors (rAAV and non-viral plasmid DNA). Viral vectors are often chosen over non-viral vectors because of their high efficiency of cell entry and long duration of expression. Third, the route of gene therapy vector administration is also very important since it will affect the organ distribution as well as brain regional distribution of the gene therapy vector. Systemic administration (e.g. intravenous) will likely lead to broad expression of the α-secretase in peripheral organs (such as liver and lungs) that may have unpredictable consequences and would also require the vector to penetrate efficiently the blood-brain-barrier in order to reach the CNS. In contrast, direct injection of a gene therapy vector to the CNS, intrathecally or intracerebroventricularly, may address the disadvantages of systematic treatments, but it is much more invasive. Fourth, a proper dosage range needs to be determined to achieve adequate prophylactic or therapeutic protections with the least side effects and costs. Fifth, the previously discussed general risks associated with gene therapy also need to be addressed.

Modulation of endogenous α-secretase gene activity can also yield increased levels of extracellular N1 (Figure 1). This approach may be better than the other two approaches described above, because it will likely yield a more authentic expression profile thereby carrying less side effects. There have been reports on regulation of various putative α-secretase gene activities^29,81–82. Expression of dimeric PrP from a gene therapy vector may also be feasible since transgenic expression of a dimeric PrP has been shown to enhance PrP α-cleavage and appears to be safe^83–85. In addition, identifying small molecules that specifically activate a proper PrP α-cleavage enzyme through established conventional methods would also be worth trying^86–87.

Enhancing α-cleavage activities carries significant risks as detailed in the companion paper (Part I) and briefly summarized below. First, all ADAMs implicated in PrP^C a-cleavage (ADAM8, ADAM9, ADAM10, and ADAM17) have been reported to promote cancer^78,80. Second, enhancing the α-cleavage will also lead to increase levels of the C1 fragment, which has been reported to make cells more sensitive to staurosporine⁴² and implicated in primary myopathy involving p53 in an inducible transgenic mouse model overexpressing PrP in the skeletal muscles^43–44. However, the negative effects of C1 may be countered by N1 through down-regulation of p53 expression³⁶. In addition, transgenic mice overexpressing GPI-anchored C1 did not appear to exhibit gross neurological deficits or histological lesions⁴¹, suggesting that C1 may be pro-apoptotic only under apoptotic stimuli but not under normal conditions³³.

2.2. Exosomal PrP

Enhancing extracellular PrP levels through augmenting PrP-containing exosomes is another attractive alternative for PrP-based therapies since they have been reported to be protective against the toxicity of Aβ, Tau, and αSyn oligomers^57–59,88. The exosome approach has significant advantages over the recombinant PrP approach. Exosomes are capable of crossing the blood-brain barrier^89,90. Exosomes can be engineered for various purposes^91–92 including as a safe and effective delivery vehicle for anti-cancer agents⁹³. Exosomes have been successfully modified to carry a siRNA cargo and target brain cells, achieving a 62% knockdown of BACE1 (a therapeutic target in Alzheimer’s disease) in wild-type mice when administered intravenously⁹⁴. Several studies used exosomes to target brain cancers⁹³ and traumatic brain injuries⁹⁵.

There are two ways to achieve higher levels of PrP-carrying exosomes in the CNS: (1) CNS injection with PrP-containing exosomes generated from cultured cells and (2) enhancement of endogenous exosome generation, but only the former seems feasible at this point (Figure 1). Many methods have been developed for efficient isolation of exosomes^96–97, including from brain cell types^{59,88,98–99}. Exosomes carrying high levels of PrP^C may be obtained from cultured cells naturally or engineered to express PrP^C at high levels. It should be possible to design an exosomal PrP that is more protective and carries less side effects. The exosomes can also be further engineered to target brain cells for less invasive peripheral administration.

Elevating extracellular PrP levels through administration of PrP-carrying exosomes share some of the risks with the recombinant full-length PrP strategy, such as unwanted functions of the PrP protein, the need for repeated treatments, and likely high costs. In addition, exosomes carry many other proteins and lipids on its surface¹⁰⁰ as well as proteins, nucleic acids and other molecules internally, which could have detrimental effects^101–103. Several of the caveats associated with exosomes have been reviewed recently⁹¹. For example, exosomes from prion-infected cells have been reported to carry prion infectivity¹⁰⁴ and facilitate the replication and spread of prions^105–106. Careful selection of cells for exosome production and targeted engineering of the exosomes will reduce the risks.

2.3. Shed PrP or recombinant full-length PrP

Elevating extracellular PrP levels for therapeutic purposes can also be achieved through enhancing shedding of endogenous PrP^C or infusion with recombinant full-length PrP. Recombinant full-length PrP or transgenic expression of anchorless PrP that is secreted into extracellular space have shown protective effects^38,45–48. It is reasonable to expect similar protective activities from full-length PrP shed from PrP^C on cell surfaces. Enhancing PrP shedding was indeed found to be protective against prions in most reports^{30,53–54,56}.

Direct CNS infusion with recombinant full-length PrP seems simple (Figure 1), but there have been no published reports with this approach. Full-length PrP carries unique risks, including promotion of replication and spread of prions^107–109 and induction of CNS inflammation^110–111. In addition, the common caveats related to the use of recombinant proteins also apply.

Enhancing shedding of cell surface PrP^C is an alternative to direct infusion of full-length recombinant PrP. This strategy aims to augment PrP shedding from cell surface through enhancing the activity of ADAM10, the primary PrP sheddase, which is supposed to lead to increased shedding of PrP^C and α-cleavage of PrP^C and APP, all of which are protective^72–74. It can be achieved through four strategies that are very similar to those of the of α-secretase activity enhancement strategy (Figure 1): (1) infusion with a bioactive recombinant sheddase (ADAM10), (2) CNS expression of ADAM10 from a gene therapy vector, (3) stimulation of the expression of the endogenous ADAM10 gene, and (4) enhancing the activity of the endogenous ADAM10 protein. The last two approaches are of particular interest since they will likely preserve the authentic spatial expression profile of the endogenous ADAM10 gene, thereby minimizing potential side effects. Again, activating ADAM9 appears to be an attractive strategy since ADAM9 has been reported to promote shedding of active ADAM10⁸¹ and upregulate the α-cleavage of PrP^{C 29,81} and APP⁷². The same pros and cons described earlier for the α-secretase activity enhancement approach also apply here. Moreover, there are multiple potential risks associated with the ADAM10 enhancement approach, including promoting cancer and autoimmunity^77–78 as well as facilitating prion replication and inducing CNS inflammation due to increases level of shed PrP^108,109. Despite these concerns, it may still be possible to elevate ADAM10 activity safely since transgenic mice overexpressing ADAM10 appear largely normal^{29,54,74–75}.

3. Expert opinion

Extracellular PrP forms, including shed PrP, exosomal PrP and PrP N-terminal fragments (PrP-N), have strong and broad spectrum therapeutic potentials for several neurodegenerative diseases due to their protective activities against PrD, AD, PD and other protein misfolding neurodegenerative diseases.

We have summarized a variety of strategies to harness the power of protective PrP forms, including direct infusion with recombinant PrP or its N-terminal fragments, infusion of recombinant enzymes that release PrP or its N-terminal fragments from cell surfaces, infusion with exosomes carrying PrP, expression of a secreted form of PrP or its N-terminal fragments from a gene therapy vector, or enhancing the expression/activities of an endogenous a-secretase or sheddase (Figure 1). All of these strategies have their own strengths and weaknesses. Because N1 has shown multiple protective activities and the N1 peptide has no known side effect aside from its impact on cell cycle progression, the most promising approach seems to be expression of a well-designed PrP-N from an optimal gene therapy vector that can achieve sustained high-level expression and secretion of N1 or PrP-N in the CNS after a single injection. The second attractive approach is elevating ADAM9 activities that will activate ADAM10 to simultaneously enhance the α-cleavage of APP and the shedding of PrP^C. However, the cancer promoting effect of ADAM9¹¹² needs to be evaluated thoroughly to validate the safety of this approach and a safe and effective range of ADAM9 elevation must be established. The third appealing strategy is peripheral administration of engineered exosomes that target brain cells and carry an optimized PrP with strong protective activities and minimal hazardous activities. Both the exosomes and the PrP molecules they carry can be engineered to maximize the delivery efficiency and protective activities while minimizing unwanted side effects. These approaches, especially those delivered as gene therapies, also have the potential to be used safely as prophylactics for disease prevention when applied before disease onset or at preclinical stages. AD and other neurodegenerative diseases result from progressive damages to dendrites, axons and synapses, as well as loss of neurons. Neuronal loss in most CNS regions is irreversible but minimal during preclinical and early phases of the diseases. Therefore, early and accurate diagnosis of the neurodegenerative diseases, including the subtypes or strains, combined with early and prompt interventions (including with our proposed strategies), will be critical for favorable clinical outcomes. There will be significant challenges with these strategies, ranging from direct side effects of the elevated PrP form, the negative impact stemming from the molecules adopted to elevate the extracellular PrP forms (such as the problems associated with enhancing an ADAM enzyme activity), and hurdles and complications from the technologies employed (such as the difficulty in generating the required amount of clinical grade recombinant proteins or gene therapy vectors, immune response to the recombinant protein or vector, and potential carcinogenicity of the gene therapy vector).

In addition, knocking down endogenous PrP gene expression with antisense oligonucleotides or siRNA has shown great potential in delaying onset and slowing down progression in AD and PrD mouse models^60,113–120. It will likely be much more effective to use two gene therapy vectors to simultaneously overexpress a secreted PrP-N and knock down the endogenous PrP expression. This combined strategy should achieve strong neuroprotection (through PrP-N) and significant reduction of toxic signaling mediated by the cell surface PrP^C (through the knockdown of the endogenous PrP gene). We are in the process of testing this combined strategy for treatment and prevention of human prion diseases in humanized transgenic mouse models.

Article highlights:

• The extracellular prion protein (PrP) forms, such as shed PrP, exosomal PrP and PrP N-terminal fragments (PrP-N) can inhibit and neutralize key toxic molecules and prevent toxic signaling in prion diseases (PrD), Alzheimer’s disease (AD), Parkinson’s Disease (PD) and a few other related protein misfolding neurodegenerative diseases.

• Several strategies to harness the protective powers of the extracellular PrP forms are explored, including direct infusion with a recombinant PrP form, infusion of recombinant enzymes that release the PrP form from the cell surface (ADAM10 or ADAM8 or ADAM17), peripheral infusion with PrP-rich exosomes engineered to target the brain, or increased expression of a secreted form of PrP from a gene therapy vector, or enhancing the expression/activities of an endogenous secretase or sheddase.

• Expected challenges for each strategy are discussed, including unwanted increased biological activity of the PrP form, indirect side effects of increased PrP cleavage enzyme activity, difficulty in generating the sufficient recombinant proteins or gene therapy vectors for treatment, risk of immune responses to recombinant proteins or gene therapy vector, and potential carcinogenicity of the gene therapy vector.

• The optimal strategy to increase PrP-N levels consists of a gene therapy vector that can achieve sustained high-level expression and secretion of PrP-N in the CNS after a single injection.

• Another appealing strategy is to elevate the ADAM9 activity that will activate ADAM10 to enhance both the α-cleavage of APP and the shedding of PrPC.

• Simultaneous knockdown of cellular PrP expression and elevation of the extracellular levels of a PrP-N via optimized gene therapy vectors should be a highly promising broad -spectrum prophylactic and therapeutic strategy against several neurodegenerative diseases, including PrD, AD and PD.