Prion Disease: A Tale of Folds and Strains

Abstract

Research on prions, the infectious agents of devastating neurological diseases in humans and animals, has been in the forefront of developing the concept of protein aggregation diseases. Prion diseases are distinguished from other neurodegenerative diseases by three peculiarities. First, prion diseases, in addition to being sporadic or genetic like all other neurodegenerative diseases, are infectious diseases. Animal models were developed early on (a long time before the advent of transgenic technology), and this has made possible the discovery of the prion protein as the infectious agent. Second, human prion diseases have true equivalents in animals, such as scrapie, which has been the subject of experimental research for many years. Variant Creutzfeldt–Jakob disease (vCJD) is a zoonosis caused by bovine spongiform encephalopathy (BSE) prions. Third, they show a wide variety of phenotypes in humans and animals, much wider than the variants of any other sporadic or genetic neurodegenerative disease. It has now become firmly established that particular PrP^Sc isoforms are closely related to specific human prion strains. The variety of human prion diseases, still an enigma in its own right, is a focus of this article. Recently, a series of experiments has shown that the concept of aberrant protein folding and templating, first developed for prions, may apply to a variety of neurodegenerative diseases. In the wake of these discoveries, the term prion has come to be used for Aβ, α‐synuclein, tau and possibly others. The self‐propagation of alternative conformations seems to be the common denominator of these “prions,” which in future, in order to avoid confusion, may have to be specified either as “neurodegenerative prions” or “infectious prions.”

The Prion Protein

In search for the scrapie agent, Bolton et al reported a partially protease‐resistant protein co‐purified with infectivity 11. Based on the N‐terminal amino acid sequence of this protein, denoted as prion protein (PrP) 136, the PrP gene (Prnp) was identified by two independent groups 24, 113. Unexpectedly, PrP is encoded by the host, and therefore, its expression is not disease‐specific. Based on these and numerous subsequent studies, the concept emerged that a conformational transition of the cellular PrP (PrP^C) into an aberrant conformation (PrP^Sc for Scrapie prion protein) is the disease‐specific feature 134.

Notably, PrP^C is found in all tetrapods and birds 42, 55, 81, 152, 164, 167, 170, 178, 192. Moreover, high‐resolution structures of PrP^C from different mammalian species, chicken, turtle and frog are available, indicating a conserved modular composition: the N‐terminal domain of PrP is flexibly disordered, followed by a highly structured C‐terminal domain of three α‐helical regions and a short, two‐stranded β‐sheet 38, 144, 145. Notably, the structures of the PrP C‐terminal domain of different tetrapod classes show extensive similarities, suggesting a similar physiological activity 20, 94. Unfortunately, the structure linked to the neurotoxic and/or infectious species of PrP^Sc remains elusive.

Biogenesis and Maturation of PrP ^C

Similar to other proteins destined for the secretory pathway, the biogenesis of PrP^C is initiated on free ribosomes in the cytosol. The N‐terminal signal sequence, which is recognized and bound by the signal recognition particle (SRP), mediates targeting of the ribosome‐polypeptide complex to the Sec61 translocon and initiates translocation of the protein into the lumen of the endoplasmic reticulum (ER) 49, 50, 107, 181, 182, 183. Interestingly, an incomplete or modified ER import can lead to the formation of neurotoxic PrP species. One of them is a transmembrane form of PrP, with the C‐terminus facing the ER lumen (^ctmPrP). Increased synthesis of ^ctmPrP has been linked with progressive neurodegeneration both in Gerstmann–Sträussler–Scheinker syndrome (GSS) patients with an A117V mutation and in transgenic mice 58, 168. In addition, it has been shown that cytosolic (cyto)PrP has neurotoxic potential 96, 185. This form can be generated by alternative initiation of translation from a downstream start codon, which eliminates essential parts of the ER signal peptide 68, 93. Alternatively, cytoPrP results from aborted ER import or retrograde transport of PrP 23, 60, 62, 82, 95, 140, 142. Based on studies in transgenic mouse models, neither ^ctmPrP nor cytoPrP seems to transmit a prion disease, emphasizing the idea that PrP can adopt a neurotoxic conformation that is not infectious.

During and after translocation into the ER lumen, PrP is subjected to a series of co‐ and post‐translational modifications. After cleavage of the N‐terminal ER signal peptide, core glycans are transferred into two asparagine residues located in the globular C‐terminal domain (N180 and 196 in murine PrP; N181 and 197 in human and Syrian hamster PrP) 54. Finally, a glycosylphosphatidylinositol (GPI) anchor is attached to serine 230 after PrP is fully translocated into the ER lumen 166. Glycoprotein‐processing enzymes in the Golgi complex convert the core glycans of PrP^C into complex structures [reviewed in 59, 75, 126]. Mature PrP^C and PrP^Sc contain the same set of at least 52 different bi‐, tri‐ and tetra‐antennary glycans, although with different relative proportions of individual saccharides 39, 147, 169. The attachment of two partially sialylated oligosaccharide moieties generates a negative electrostatic field, which covers the whole surface of helices 2 and 3 197. Regarding a possible physiological or pathological relevance, PrP glycosylation has been shown to have an impact on the conformational transition of PrP^C into PrP^Sc, to influence the selective neuronal targeting of PrP^Sc and to contribute to the phenomenon of prion strain diversity 29, 30, 31, 77, 84, 90, 118, 133, 174, 191. However, studies in cultured cells, in vitro systems and transgenic mice revealed that neither the N‐linked glycans nor the C‐terminal GPI anchor is required for the generation of infectious prions.

The GPI anchor targets mature PrP^C to the outer leaflet of the plasma membrane. Internalization of cell surface PrP^C may occur via coated pits or caveolae‐dependent structures 99, 131, 162, 171 and is mainly mediated by the unstructured N‐terminus 70, 112, 163.

The physiological function of PrP^C has remained elusive. Various binding partners have been suggested such as the neural cell adhesion molecule 153. A large number of experiments have shown that the N‐terminal histidine residues bind Cu⁺⁺ 15, 32, 61, 78, 139, but a definite answer to the function of copper binding of PrP is lacking. Cell culture experiments have hinted at a role in resistance to oxidative stress, which has been difficult to define on the molecular level. Experimental infarct models in mice have shown that occlusion of the internal carotid artery has far more deleterious effects in PrP^C knockout (Prnp‐/‐) mice than in normal controls 102, and that this protective function of PrP^C is linked to the N‐terminus 108.

Aberrant PrP Folding is Linked to Infectivity and Toxicity

In prion diseases, neurodegeneration is accompanied by the formation of an infectious particle denoted as prion. Because recombinant PrP expressed in and purified from bacteria and subsequently misfolded in vitro can transmit the disease 71, 88, 97, 184, it can be inferred that the essential component of an infectious prion is an aberrantly folded PrP conformer. This does not exclude the possibility that auxiliary components, such as lipids, can modulate PrP^Sc propagation 33, 34. Recently, it was suggested that other proteins associated with neurodegenerative diseases, such as Aβ, τ and α‐synuclein, may have a prion‐like activity 135. More details on this intriguing topic are given by Eisele, and Clavaguera et al, George et al (this issue). In the majority of prion diseases, there is a correlation between the accumulation of proteinase K (PK)‐resistant PrP^Sc, formation of infectious prions and neurodegeneration. However, several pathogenic PrP mutants, which can induce neuronal cell death in transgenic mice, lack the ability to form infectious prions and do not adopt a PK‐resistant conformation 6, 26, 40, 58, 92, 110, 161. Vice versa, propagation of infectious prions and PK‐resistant PrP can occur in the absence of clinical signs 13, 25, 98. Thus, two partially independent pathways seem to exist, one leading to the propagation of infectious prions and another one that mediates neurotoxic signaling 150.

While we are far from understanding the mechanisms underlying the neurotoxic activity of PrP^Sc and other pathogenic protein conformers, it was interesting to see that different mouse and cell culture models revealed an intriguing activity of PrP^C in neuronal cells to serve as a cell surface receptor that mediates toxic signaling of both PrP^Sc and Aβ 1, 4, 5, 8, 13, 19, 27, 41, 51, 85, 87, 98, 141, 143, 177, 195.

Transmission Barrier and Prion Strains

Prion diseases are highly infectious provided that the amino acid sequences between PrP^Sc in the inoculum and PrP^C expressed in the infected organism are identical. As a consequence, PrP^Sc molecules derived from one species fail to or only inefficiently transmit the disease to an organism expressing heterologous PrP^C molecules, a feature denoted as a species or transmission barrier 28, 109, 137, 154, 155. In a seminal experiment, Scott et al showed that a transmission barrier can be overcome by transgenic expression of PrP^C that is homologous to the PrP^Sc present in the inoculum 154. Based on this feature, it is possible to study different mammalian prion diseases in transgenic mice expressing the respective PrP^C protein [reviewed in 156, 175, 179].

Interestingly, a transmission barrier is not only seen between different species. After inoculated with the ME7 scrapie isolate, VM mice succumb to disease after 280 days compared to incubation periods between 140 and 180 days in eight other mouse lines 36. Later, it was shown that the differences in incubation time are linked to a polymorphism in the mouse PrP gene, resulting in the exchange of two amino acids 186. In humans, the polymorphism at codon 129 (MM, VV, MV) is a strong genetic risk factor for Creutzfeldt–Jakob disease (CJD) (see below).

Interestingly, differences in incubation time can also occur when the invading PrP^Sc and the host PrP^C have identical amino acid sequences 127. This phenomenon is caused by the existence of prion strains that have the ability to generate distinct incubation times and neuropathological lesions in homologous hosts. Initially used as an argument to challenge the protein‐only hypothesis and to postulate prion‐specific nucleic acids, it is now generally assumed that strain properties are linked to distinct PrP conformations. As structural information for PrP^Sc is not available at high resolution, this concept is currently based on distinct biochemical and biophysical properties of different strains 3, 9, 29, 83, 89, 128, 129, 148, 176. Along this line, it is tempting to speculate that the transmission barrier is also caused by distinct conformations of PrP^Sc from various species.

Prion Diseases

Considering the fact that humans have been exposed to animal prions for at least centuries, the importance of the species barrier can hardly be overestimated. Scrapie in sheep was first described in the 1700s and scrapie prions have certainly been ingested by humans for centuries. It appears that ovine and caprine scrapie prions have not passed the species barrier to humans. BSE prions, on the other hand, have crossed the species barrier and are the infectious agents of variant CJD, a human prion disease with distinct clinical and pathological characteristics. Chronic wasting disease, rampant and highly contagious among cervids in North America, does not seem to cross the species barrier to humans, but the identification of multiple strains and the potential for agent evolution warrants caution 35, 151, 173, 188. A description of prion diseases in animals is beyond the scope of this review. Suffice it to consider that prion diseases in animals—like in humans—may be sporadic and spontaneous diseases whose next eruption and strain properties are impossible to predict.

Human Prion Diseases

Human prion diseases can be idiopathic (this is the vast majority of cases), hereditary (familial, genetic) and acquired (see Table 1). Central nervous system (CNS) tissue of many genetic cases has been experimentally transmitted to non‐human primates and (transgenic) mice.

Table 1.

Human prion diseases. Abbreviations: BSE = bovine spongiform encephalopathy; sCJD = sporadic Creutzfeldt–Jakob disease; sFI = sporadic fatal insomnia; VPSPr = variably protease‐sensitive prionopathy; gCJD = genetic CJD; GSS = Gerstmann–Sträussler–Scheinker disease; FFI = fatal familial insomnia; vCJD = variant CJD

	Etiology
Idiopathic
sCJD	Unknown
sFI	Unknown (corresponds to CJD MM2T with thalamo‐olivary atrophy)
VPSPr	Unknown
Genetic (hereditary, familial)
gCJD	>20 different PRNP mutations
GSS	Certain PRNP mutations
FFI	PRNP mutation D178N associated with M129
Acquired
Iatrogenic CJD	Through treatment with contaminated human growth hormone, dura mater, cornea, etc.
vCJD	Infection with BSE prions
kuru	Ritualistic cannibalism (historical, New Guinea)

Polymorphisms of the human prion protein gene (PRNP)

PRNP is located on human chromosome 20; it encodes a protein of 253 amino acids 81, 138. An amino acid polymorphism at codon 129 of PRNP, which encodes either methionine (M) or valine (V), is a strong genetic risk factor for human prion diseases 114. The M allele shows a frequency of >0.9 in Asians and 0.55 in Africans 165. In Caucasian populations, about 51% are MV heterozygous, 38% are MM homozygous and 11% are VV homozygous 86, 189, 190. MM homozygotes are significantly overrepresented in sCJD. Among the 300 cases from Europe and the USA, 71.6% were MM, 11.7% were MV and 16.7 were VV 122. All confirmed vCJD cases have been methionine homozygotes; there may be one possible case of vCJD reported as a heterozygote 64. In addition to being a risk factor, codon 129 is a strong modifier of the clinical and pathological phenotypes of sCJD and gCJD (see below). Lewis et al showed that recombinant PrP‐M129 has a stronger propensity to form amyloid fibrils than V129 91; however, a thermodynamic understanding of this phenomenon is lacking.

Two other polymorphic codons of PRNP (127 and 219) may confer resistance to kuru and sCJD 106, 158, 159. The G127V polymorphism has been proposed to reduce the relative risk to developing kuru in exposed individuals. The E219K variant was reported in about 6% of the Japanese population but has not been identified in CJD. Studies in transgenic mice have found that PrP‐219 K was not converted to PrP^Sc and inhibited the conversion of co‐expressed wild‐type PrP 130. The loss of one octarepeat was reported with no influence on disease susceptibility 115, while the loss of two octarepeats has been associated with CJD 7, 21. A number of silent polymorphisms with no recognizable influence on disease susceptibility have been reported 193.

Genome‐wide association studies (GWAS)

Two GWAS on vCJD have identified the PRNP locus as strongly associated with risk and additional factors whose functional analysis in this context is still ongoing 104, 149. A large GWAS of human prion diseases has recently been performed on sCJD, vCJD, iatrogenic CJD, inherited prion disease and kuru 105. Two thousand disease samples and more than 6000 controls were investigated. Only the PRNP locus was highly associated with risk in all etiological and geographical groups, driven by the coding variation at codon 129, whereas no other locus of all human prion diseases achieved genome‐wide significance. A single‐nucleotide polymorphism (SNP) in the CHN2 gene was associated with vCJD but no other group of prion disease. Overall, 14 SNPs (two at PRNP) were associated, but with the exception of PRNP, all others showed only moderate effects, underscoring the need for further association studies to provide definitive evidence. Previously proposed prion disease risk factors, particularly for vCJD, were not confirmed by this GWAS; this may be explained by the rarity of prion disease and the accordingly small sample size of vCJD in those studies.

Types and strains of sporadic CJD

With the advent of molecular characterization of human prion diseases, well‐defined types of sporadic and acquired CJD have become apparent. First, it was shown that the two genetic PrP variants encoding either methionine (M) or valine (V) at codon 129 not only define susceptibility to sporadic and acquired CJD but also determine the pathology 37. Subsequently, it was realized that PK‐resistant PrP in humans can be observed in two different forms (PrP^Sc type 1 migrating in Western blots at 21 kDa and type 2 migrating at 19 kDa) distinguishable by two preferential PK digestion sites at amino acids 96 and 85 that also affect pathological and clinical appearance of the disease 118. This has resulted in the recognition of six types of sporadic CJD, which have been well characterized with regard to their molecular composition, that is, genotype at codon 129 (MM, MV or VV) and PrP^Sc type 1 or 2, and pathological as well as clinical phenotype in the following way (see Table 2 and Figure 1): (i) The MM/MV1 type, which in its pure form comprises 40% of all sCJD cases, shows small vacuoles, synaptic PrP^Sc deposition, and is clinically characterized by rapid cognitive decline, typical periodic sharp wave complexes (PSWC) in electroencephalogram (EEG) tracings and positive 14‐3‐3 proteins in the CSF. (ii) The VV2 type (15% of sCJD cases) shows plaque‐like deposits and perineural staining in the cortex as well as plaque‐like deposits in the cerebellar cortex and white matter. Clinically, gait ataxia is prominent in early stages. (iii) The MV2K type is found in 8% of all sCJD patients and shows kuru plaques on histological examination. Clinically, dementia and ataxia are characteristic features. (iv) The MM2 C type is very rare in its pure form (1% of all sCJD cases); its hallmark is widespread, large, confluent vacuoles in the cerebral cortex with perivacuolar and coarse focal PrP^Sc staining. (v) The MM2T type is also very rare (about 1%) and is indistinguishable from FFI, which is caused by a D178N mutation in coupling with 129 M. Clinically, insomnia is prominent, while pathologically there is atrophy of the thalamus and the inferior olive, with only minor changes in other regions of the brain. (vi) The VV1 type is also very rare (1%) and affects young patients (average age of 39 years). Pathologically, there is severe spongiform change in corticostriatal regions, but only very faint punctate PrP^Sc staining on immunohistochemistry.

Table 2.

Terminology of Creutzfeldt–Jakob disease 121, 124). Abbreviations: CJD = Creutzfeldt–Jakob disease; sFI = sporadic fatal insomnia; vCJD = variant CJD

Molecular type	Histological type “CJD with …”
MM/MV 1	Diffuse synaptic deposits
VV 2	Perineuronal and cerebellar plaque‐like deposits
MV 2K	Kuru plaques
MM 2C	Cortical confluent vacuoles
MM 2T (sFI)	Thalamo‐olivary atrophy
VV 1	Cortiostriatal synaptic deposits
MM 2V (vCJD)	Florid plaques
MM/MV1+2C	Mixed diffuse synaptic deposits and cortical confluent vacuoles
MV 2K+C	Mixed Kuru plaques and cortical confluent vacuoles
Various	Atypical

Figure 1.

A. Hematoxylin and eosin (H&E) stained section of a MM1 case showing spongiform change with typical small vacuoles. Frontal cortex. B. Diffuse synaptic PrP deposits in the cerebellum of an MM1 case. IHC with an antibody against PrP (L 42). C. Confluent vacuoles typical of cortical areas with MM2. H&E, frontal cortex. D. Perivacuolar immunohistochemical PrP staining in a mixed MM1/MM2 case. IHC with an antibody against PrP (L 42). E. Kuru plaque in the cerebellar cortex of an MV2 case. H&E stain. F. Cerebellar plaque‐like PrP deposits in the granule cell layer and the adjacent white matter.

In addition to the six pure phenotypes discussed earlier, it has become apparent that up to 30% of sCJD cases can be “mixed,” with co‐occurrence of PrP^Sc types and parallel pathological phenotypes 125. Type MM/MV1+2C is the commonest mixed type (43% of all MM cases) and pathologically shows large confluent vacuoles in some areas, in addition to the small vacuolar change typical of MM/MV1. Type VV2+1 is only recognized by Western blotting, as the histological appearance is very similar to the pure VV2 type. In contrast, types MV2K+2C and MM2T+2C can only be diagnosed on histological grounds (both have PrP^Sc type 2 pattern in Western blots).

There is only a relatively small number of CJD cases, mostly single cases of CJD with pathological changes that do not fit into the above‐mentioned classification system. These include one CJD MM1 case with kuru plaques 66, three CJD MM1 cases with long duration and PrP‐positive plaques in the white matter 72, one MM case with PrP^Sc lacking the doubly glycosylated band 48, one case with widespread PrP deposition and an unusual PrP^Sc type 103, and others 52. One has to assume that many similar cases are unpublished. However, considering the thousands of CJD cases that have been investigated in recent years, these cases must be considered as exceptionally rare; their etiology and pathogenesis remains unclear, genetic factors and the possibility of an infectious origin must be considered.

An inter‐laboratory trial has shown that PrP^Sc types 1 and 2 as seen on Western blots could be reliably classified in seven different European laboratories 123, whereas subtle further distinction requires high‐sensitivity gel electrophoresis protocols 111. Similarly, histological typing was assessed in surveillance centers of Europe and the USA. While there was full agreement on vCJD and the common sCJD types, rare types were not recognized by all participants 121.

Typing of CJD by Western blots and histology has not resolved the question of whether the described sCJD types correspond to prion strains. Strains are classically defined as isolates of infectious prions characterized by distinctive clinical and neuropathological features that are faithfully recapitulated upon serial passage within the same host genotype 17. The existence of distinct human strains was demonstrated in transmission experiments 18, 63, 76. To tackle the problem of human prion strains and PrP isoforms, Parchi et al investigated archival cases of the National Institutes of Health (NIH) series of human sCJD, iCJD and kuru that had been transmitted to non‐human primates 119. Transmission properties, Western blot, analysis of PrP and histopathology of the MM1 phenotype were significantly different from CJD VV2 and MV2 prions, which showed that there are at least two distinct prion strains causing the common sCJD types. However, the historic NIH series was not complete and did not include a case of the very rare VV1 type. Bishop et al transmitted sCJD types to transgenic mice expressing human 129MM, 129VV and 129MV. They were able to distinguish four distinct strains, MM1 and MV1 showing identical transmission properties, MV2 and VV2 being very similar, and MM2 as well as VV1 being different from each other and the other strains. MM2T (sporadic fatal insomnia) was not included in their study 10. Thus, the various PrP^Sc protein types provide a molecular signature for particular transmissible prion strains.

Variably protease‐sensitive prionopathy (VPSPr)

In 2008, Gambetti et al described 11 codon 129 VV homozygous patients affected by a disease that was different from hitherto known human prion diseases in terms of clinical and pathological features as well as the PrP species identified in the brains of affected patients, which did not show the resistance to treatment with proteases typical of other prion diseases. Hence, this novel type of disease was designated “protease‐sensitive prionopathy (PSPr)” 44. In 2010, the same group reported 15 new cases with similar features, which included cases of MM homozygotes and MV heterozygotes 196. The newly described cases presented diminished and dissimilar sensitivity to protease digestion and the disease was renamed variably protease‐sensitive prionopathy (VPSPr). The emerging features of this disease include a clinical presentation with behavioral and mood changes, aphasia, cognitive impairment, ataxia and parkinsonism. Histopathology shows relatively large vacuoles, and immunohistochemistry reveals microplaques in the molecular layer of the cerebellum as well as target‐like rounded formations of clusters of granules that increase in size toward the center. PrP as seen on Western blots is most distinctive, showing a ladder‐like appearance after pK treatment commonly consisting of five bands migrating between 27 and 7 kDa. The 7‐kDa band showed high pK resistance. Sensitivity to proteinase K was highest in VV cases, intermediate in MV and lowest in MM.

Similar cases have been identified since then by other groups 56, 57, 67, 146. Krebs et al had described a surgical biopsy case, which in retrospect must be regarded as a case of VPSPr in the year preceding Gambetti's publication 79. That case was VV homozygous; the Western blot showed a 6.2‐kDa band of pK‐resistant PrP, the histopathology and IHC were quite typical of VPSPr. The patient was a 66‐year‐old woman with a 3‐year history of spasticity followed by dementia.

To date, transmissibility of this disease to laboratory animals has not been reported. The real nature of VPSPr, and in particular its nosological position related to sCJD, remains to be established.

Familial/Genetic Human Prion Diseases

About 10%–15% of individuals developing a prion disease harbor either a point mutation or an octapeptide repeat insert mutation of PRNP; genetic human prion diseases are transmitted in an autosomal dominant fashion. These mutations are remarkable for two reasons. First, they confer a genetic disease that is experimentally transmissible to laboratory animals (in fact, this has only been shown as a “proof of principle” for some of the mutations). Second, they cause a large variety of phenotypes—much larger than other heritable neurodegenerative disease—that include GSS, fatal familial insomnia (FFI) and genetic CJD (gCJD).

At present, more than 20 CJD‐associated point and octapeptide repeat mutations of PRNP are known 22. Some of them have been noted worldwide in many patients, particularly E200K, while others are rare or very rare. By and large gCJD reproduces the phenotypic spectrum of sCJD, with the codon 129 genotype and PrP^Sc properties mainly determining the phenotype. The particular PRNP mutations seem to have a lesser influence on the phenotype. In gCJD, the codon 129 polymorphism mainly acts in cis with the mutation, while the normal allele is less significant. Particular phenotypes are seen in E200K cases that may show stripe‐like, coarse, granular PrP deposits in the molecular layer of the cerebellum, patchy PrP deposits in the cerebellum in cases with additional octarepeats, and others 22.

Fatal familial insomnia differs from gCJD in a number of ways 45. It is characterized by disturbances of the sleep–wake cycle, autonomic activation, and cognitive and motor signs that are associated with thalamic atrophy. The pathology is dominated by a severe atrophy of the anterior ventral, mediodorsal and pulvinar thalamic nuclei as well as atrophy of the inferior olives. Spongiform degeneration may be minimal and limited to the entorhinal cortex, particularly in cases with a short clinical course. The D178N mutation has been associated with two different phenotypes related to the codon 129 genotype of the mutant allele 45. The D178N‐129M genotype is associated with PrP^Sc type 2 and the FFI phenotype, D178N‐129V is associated with PrP^Sc type 1 and a phenotype that is similar to the VV1 type of sCJD.

GSS is clinically characterized by ataxia and/or dementia; the pathology shows large, PrP‐positive plaques and a variable degree of spongiform change. It is most often caused by the P102L mutation of PRNP, but other mutations, at codons 105, 117, 131, 198, 202, 212, 217, 218 and insert mutations have been found (116, 132) 2. On Western blots, PrP often shows a 7‐ to 8‐kDa or 10‐kDa band, which was reported to be present in brain regions showing PrP‐positive multicentric plaques 120. Hyperphosphorylated tau pathology is a feature of some mutations (F198S, Q217R, Y218N) 2, 46, 47.

Acquired Human Prion Diseases

vCJD

In 1996, 11 years after the identification of BSE, a novel prion disease in cattle, the UK National CJD Surveillance Unit in Edinburgh reported a new form of CJD in humans, which is now known as vCJD 187. Epidemiological evidence suggested that vCJD was caused by BSE prions. Experimental transmission studies showed that BSE prions and the vCJD agent were closely related, whereas both of them were significantly different from all types of sCJD infectious agents 18, 63, 157. All definite vCJD cases that have been tested were codon 129 MM homozygotes; one possible case has been reported in a heterozygote 64, 69. To date, 176 cases have been diagnosed in the UK, 49 additional cases in 11 other countries. Four probable cases of vCJD infection have been identified after transfusion of red blood cells from asymptomatic donors that later developed vCJD.

vCJD prions are different from other human prion types in a number of ways. Western blots of PrP^Sc show a predominance of the diglycosylated form that has been termed 2B PrP^Sc and that differs significantly from the PrP^Sc types seen in all other human prion diseases 117. vCJD starts at a relatively earlier age, there are prominent early psychiatric symptoms, often depression and anxiety, dys‐ or paresthesia, ataxia, myoclonus, chorea and dystonia. Death occurs 14 months after disease onset. After oral exposure to BSE, prions replicate within lymphoid tissues. The microscopic appearance of the disease in the brain is predominated by florid plaques, widespread accumulation of PrP^Sc and thalamic gliosis 65.

Iatrogenic CJD

The major sources for iatrogenic accidental transmission of CJD were contaminated human growth hormone (226 cases) and human dura mater grafts (228 cases) 16. Only occasional cases with long incubation times exceeding 20 years are observed today; thus, the era of iatrogenically or accidentally transmitted CJD (iCJD) is coming to its end.

The sources of infection in iCJD were human strains, the recipients—humans. It is not surprising therefore that the clinical and pathological features of iCJD by and large mirror those of sCJD. However, there are some noticeable exceptions. First, although it has been shown that codon 129 MM homozygosity is associated with an increased risk to develop sCJD and iCJD, UK growth hormone recipients have a strong prevalence of VV homozygotes 12. It has been speculated that a different prion strain might have been disseminated, possibly from a VV homozygote. Second, in contrast with the clinical appearance of sCJD, dementia in iCJD of hormone recipients most often never occurred or was only a component of the late clinical disease 16. Last and most surprisingly, among the more than 200 cases of dura mater‐related cases worldwide, a small minority with divergent pathology has been observed in Japan, France and Germany 74, 80, 160, 172. Histopathology showed florid plaques otherwise only seen in vCJD, albeit at a much lower frequency. At least in one case PrP^Sc showed a banding pattern that was slightly different from sCJD 80. These cases occurred in MM homozygotes; on EEG examination, they lacked PSWC. It is likely but not quite clear whether all six reported cases were caused by the same source. However, it stands to reason that they represent a CJD strain that is different from all hitherto known human prion strains.

In a recent survey, Yamada et al found that in Japan the plaque‐type (not necessarily showing “florid” plaques) accounted for 52% (14 of 27 cases) of the pathologically diagnosed iCJD cases after dura mater implants; all of these cases were codon 129 homozygous for methionine 194. Based on the results of transmission experiments to transgenic mice, Kobayashi et al have suggested that plaque‐type iCJD may be caused by transmission of sporadic CJD‐VV2 or MV2 prions to persons with a codon 129 MM genotype 73.

Kuru

Kuru, a historical epidemic disease of the Eastern Highlands of New Gunea, which was transmitted by the ritualistic consumption of dead relatives, was the first human prion disease transmitted experimentally to non‐human primates 43. The clinical appearance of kuru has been described as somewhat diverging from sCJD, with progressive ataxia being prominent and dementia being less distinct and often only appearing in late stages, features that are reminiscent of iCJD. It has therefore been hypothesized whether kuru might represent a distinct human prion strain.

For a comparison of strains, Wadsworth et al have undertaken to transmit prions from kuru, sCJD, vCJD and iCJD to wild‐type mice and transgenic mice expressing human PrP 180. Western blot signatures of PrP, transmission rates and immunohistochemistry of kuru and sCJD of the appropriate genotypes by and large showed large similarities.

It has been noted that kuru shows considerable similarities to sCJD VV2 or sCJD MV 2K 14, 53, 100, 101, 117. Using archival tissues from transmission experiments, Parchi et al recently provided evidence that strengthened the theory that kuru originated from cannibalism of an individual with sCJD VV2 or MV 2 K, as no indication of MM1 prions could be found, even in subjects carrying the MM or MV genotype that had succumbed to kuru 119.

Conclusions

Since the 1980s, when the hypothesis of an infectious proteinaceous agent of scrapie was first propounded, prion research has led the way to understanding protein misfolding and protein aggregation diseases. Although it has been established that particular PrP^Sc isoforms are closely related to specific human prion strains, neither the multitude of conformational differences of PrP^Sc nor their way of conferring strain properties are understood. The self‐propagation of alternative conformations by templating that we have come to learn from infectious prions seems to be a common theme of a number of other proteins that include Aβ, tau and α‐synuclein. Tempting as it may be to subsume all of these under the term of “prions” for their mechanistic similarities, it may be wise at present to keep them apart as “infectious prions” and “neurodegenerative prions” until the various forms of self‐propagation of alternative conformations as well as the mechanisms of cell‐to‐cell propagation and infectivity have been clearly defined. Nevertheless, there is a reason for hope that new inroads into understanding the mechanisms of pathogenesis, diagnosis and therapy of one of these diseases may also benefit the others.

Conflict of interest

The authors declare that they have no conflict of interest related to this article.