Mad Cow Home ... Best Links ... Contact Researcher ... Science Index

Doppel protein properties
5 alleles of human doppel
Doppel allele and sequence resourse
CpG hotspots for doppel
Odd new doppel EST
Index to doppel articles

Prion genomics round-up: a human exon 2 EST
Jag2: another prion gene on chromosome 14?
Late onset prion mutations: E196K, V203I, E211Q
Self-assembly of recombinant prion protein of 106 residues

Doppel protein properties

J Biol Chem Jun 6 2000 
L Silverman G, Qin K, Moore RC, Yang Y, Mastrangelo P, Tremblay P, Prusiner SB, Cohen FE, Westaway D
[[[Comment (webmaster): The study of doppel moves beyond genomics and mRNA splicing to experimentally determined properties of the doppel protein itself in this accepted JBC manuscript, which runs to 42 pages as a double-spaced pdf or 1.7 megabyte as email. Commendably, the authors distributed the paper widely as soon as this became appropriate. It is a most welcome addition to the doppel literature that signficantly buttresses previous bioinformatic flights of fancy; though the latter were right on target, experimental complexities are found in disulfide pairing, extent and tissue-dependence of diglycosylation, and other aspects.

Most intriguingly, the authors observed tissue-dependent expression ratios of doppel and prion proteins. Doppel is expressed at fairly high levels in testes and heart but not detectably in brain (except in the context of alternative splicing in certain strains). This was also inferred earlier to a certain extent from dbEST or SAGE, but doppel remains fairly poorly represented there and also in proprietary mRNA collections. Tissue or developmental specialization is commonly observed among diverging gene duplicates, eg fetal hemoglobin, and provides an oft-cited rationale for retention of both copies. Thus it is not too surprising that doppel is toxic when expressed in the wrong tissue (brain) in certain strains. Note too that genes such as doppel expressed at high levels in testes provide enhanced opportunity for pseduogene creation.

The prion/doppel tandem may be understood as follows: Both genes are a bit odd in possessing a single coding exon. Imagine now a single prion gene with multiple exons corresponding roughly to signal region, repeat, invariant core, globular domain, and GPI tail. Conventional alternative splicing could generate two distinct proteins in different mixes according to tissue or stage, one full length and one skipping exons 2 and 3 (that is, doppel-like in lacking the repeat and 106-126).

One sees that chimeric transcripts beginning at 5' UTR exon 2 in the actual prion-doppel tandem pair accomplish the same end result by an alternative mechanism to the usual splicing out small coding exons, a mechanism appropriate to tandem gene pairs with single coding exons. Tandem duplication thus rescues the opportunity for tissue-specific expression of differing protein forms without the need for internal intron creation, provided the 5' UTR splice acceptor can be regulated as per tissue. The mechanism does require fixation of a two domain deletion event in the doppel copy soon after the tandem pair forms. Neither Medline nor GenBank are structured to be convenient to pull out all single-exon genes and determine which other tandem pairs utilize this mechanism (which might be called tandem deletion splice pairing); the prion/doppel model may be fairly common in the final analysis.]]]

The paper provides significant information on specific doppel biochemistry issues. The level of experimental detail is remarkable -- someone could actually replicate the experiments based on it.

-- Secondary structure: recombinant mature unglycosylated oxidized doppel refolds into alpha-helical structure as anticipated from alignment to prion protein, though CD is incapable of determing helical boundaries and the issue of a comparable small 2-strand beta sheet wasn't resolved. The authors introduced a Ser at the amino-terminus of mouse doppel 27-154 to minimize proteolytic degradation in the E. coli cytoplasm, giving N-terminal sequence SRGIKHRF.... The actual signal peptide junction is uncertain: Laplanche reports residue 24 is also a candidate according to standard prediction tools. This is relevent given the observed human allele T26M. Secondary structure is unlikely to be affected by these subtleties.

-- Intra or inter-molecular disulfides: Little free cysteine occurs, one doppel disulfide should parallel that of prion protein, but can doppel dimerize through cross-links of the remaining disulfide? Tryptic peptides of the 'native' protein indeed indicated a disulfide bond between residues 109 and 143, analogous to the prion disulfide bond 178-213. While primary support was found for cys 95-cys 148 connectivity, 143-148 was also observed. Note that the geometry of GPI anchors make both an inter-doppelor prion-doppel disulfide dimer unfavorable. Given insufficient material direct from mice, the matter must rest here.

Mixed cross-linking makes direct crystallographic or nmr structure determination more difficult, though renaturation in the presence of disulfide isomerase could prevent artefacts. Rudi Glockshuber, an authority on protein disulfides and nmr structural determination, is thanked for a doppel plasmid in the acknowledgements, implying disulfide pairs and structural coordinates may soon be at hand. Note that a serious problem exists in building the second disulfide off doppel threaded to prion nmr coordinates: phe 175 is an absolute invariant in all species including turtle and all doppels' this deep interior residue edge-bonds to tyr 194 in most species and no doubt has the same relation to his194 of doppel. This leaves the fourth cysteine facing the wrong way on the back side of helix C, unable to form a disulfide pair, though 3D-PSSM apparently resolves this by a twist of helic C.

-- Mono and diglycosylatation: PNGaseF was used in immunoblots of Dpl-3 cell lysates to establish that both glycosylation sites are occupied in this system. Monoglycosylated protein was not observed prior to the PNGaseF digestion time series. Doppel oligosaccharide side-chains are resistant to digestion by endoglycosidase H (unpublished) but are sensitive to PNGaseF, ie a linkage indicating doppel processing in the ER and Golgi. The nature of the attached glycans were not established -- they may or may not be identical to the tetra-antennary prion glycan. Indeed glycans may vary by tissue, the authors note that doppel anomalously expressed in brain has a different banding pattern than in testes and N2a cells (pg 24). Whether either or both monoglycosylated or non-glycosylated forms occur in vivo is not settled; if so, these would lack the strain type significance seen in prion amyloid fibril, which is attributable to epitaxic fiber growth constraints and neural cell type glycosylation differences. Strain types are not relevent to doppel because --lacking the 106-126 region -- it has no special propensity for amyloid fiber formation.

-- GPI-anchor: While expected from distal homology to prion protein, bioinformatic support for a GPI recognition site in doppel is mediocre. An anchor would imply external cell surface location (though subsequent physiological clipping to exterior milieau occurs in some systems). Monolayer of intact Dpl-3 cells were incubated with the clipping enzyme PIPLC prior to collection and lysis; separately Triton X-114 phase-partitioning of PIPLC treated and non-treated Dpl-3 cells was done. While a long way from a direct mass spec demonstration of GPI, the bottom line is thata GPI anchor predominates over the alternative an inserted integral membrane tail. It was not established whether the somewhat unconventional gly 155 site indeed serves as the residue of GPI attachment -- a year has passed since the JMB group claimed to have comprehensively reviewed GPI attachments genome-wide. GPI-anchors are sometimes only a way station for proteins targeted to the extra-cellular medium (where a second disulfide would provide needed extra structrual stability). There is no evidence here that the GPI of doppel directs it to caveolae-like membrane domains.

-- Expression: Doppel is not significantly expressed in brain, indicating its principal physiological function lies elsewhere. Note that expression in heart or adult testes might still be associated with periferal neurons extending to these tissues. The prion protein is expressed in a wide range of tissues, including non-innervated blood cells, showing its role too is not really brain-specific. The brain is a stable site where decade-long processes requiring non-dividing cells can give rise to pathology unrelated to function or loss of function While doppel mRNA can be detected in both testis of wt mice and in the brains of still-asymptomatic Rcm0 mice by northern analysis, doppel protein is more abundant in the former location, suggesting (among other possibilities) that translation of doppel from chimeric mRNA in the brains of Rcm0 mice is less efficient.

-- Pathology: No experimental data is given; the authors speculate about a common pathway with certain rescuable prion deletion mutations involving a hypothetical membrane receptor mediating cell death. This paper also makes significant and unfortunate appeal to unpublished data such as endoglycosidase H experiments (pg 22), rat genomic structure (pg 3), computer modeling studies of the four Cys residues (pg 15), and inadequate disclosure of cross-reactivity of anti-doppel rabbit polyclonal Ab D7177 (page 19).

Glossary of mouse strains (none affect doppel coding region directly):
-- Rcm0 and Ngsk: Purkinje cell loss at 70 weeks, progressive ataxia. Deletion of prion gene and its 5'UTR splice acceptor; effect supressed by prion transgene.
-- ZrchI and ZrchII Prnp 0/0: mice remain healthy despite deletion of prion coding regions. No intergenic splicing as exon 3 splice acceptor intact.
-- FVB/N0: wildtype control, whether linc or sinc allele is not specified.
-- N2a: mouse neuroblastoma cell line from ATCC used to establish all cell lines, whether linc or sinc allele is not specified..
-- Dpl-3 and Dpl-7: stably transfected cell lines, which respectively express or don't express doppel mRNA or protein when assayed by northern and western blotting.

Highlights of the paper:

Doppel is an N-glycosylated GPI-anchored protein: expression in testis and ectopic production in the brains of Prnp{super0/0} mice predisposed to Purkinje cell loss.

The Prnd gene encodes a homolog of the cellular prion protein (PrPC) called doppel (Dpl). Up-regulation of Prnd mRNA in two distinct lines of PrP gene-ablated (Prnp0/0) mice, designated Rcm0 and Ngsk, is associated with death of Purkinje cells. Using recombinant Dpl expressed in E.coli and mouse neuroblastoma cells we demonstrate that wild type (wt) Dpl, like PrPC, adopts a predominantly alpha-helical conformation, forms intramolecular disulfide bonds, has two N-linked oligosaccharides, and is presented on the cell-surface via a GPI anchor.

Dpl protein was detected in testis of wt mice. Using Triton X-114 phase partitioning to enrich for GPI-anchored proteins, Dpl was detected in brain samples from Rcm0 Prnp0/0 mice but was absent in equivalent samples from wt mice and ZrchI Prnp0/0 mice, indicating that ectopic expression of this protein may cause cerebellar pathology in Rcm0 mice.

Biochemical and structural similarities between PrPC and Dpl documented here parallel the observation that ataxia Ngsk Prnp0/0 mice can be rescued by overexpression of wild-type PrP transgenes, and suggest that cell-surface PrPC can antagonize the toxic effect of Dpl expressed in the central nervous system.

The prion protein gene, Prnp, encodes PrP Sc , the major structural component of the etiological agent in disorders such as scrapie and Creutzfeldt-Jakob disease. In uninfected animals Prnp encodes a cellular protein denoted PrP C , which is thought to be converted to PrP Sc by a posttranslational mechanism during the course of prion infections. Though Prnp is conserved in mammalian species, the ZrchI and Npu lines of Prnp 0/0 gene-ablated mice remain healthy throughout development (1-3): this may indicate that PrP C serves a function that is not apparent in a laboratory setting, or perhaps that other molecules fulfill an overlapping function. An insight into this issue has emerged from large-scale sequencing (4).

While Prnp was originally considered to be unique within the mammalian genome it has recently been shown to have a paralog (5). The mouse gene Prnd is located 16 kB downstream of Prnp and encodes the 179 amino acid Dpl protein. Corresponding Dpl coding sequences are located downstream of Prnp in both humans and rats, and most likely in other mammals. Dpl has approximately 25% sequence identity with the C-terminal two-thirds of PrP C and is predicted to contain three alpha helices and two disulphide bonds, but is notably lacking an octapeptide repeat domain like the one present at the N-terminus of PrP C (5).

Unlike Prnp, abundant Prnd mRNAs are absent from the brains of wild-type (wt) animals, although species of 1.7 and 2.7 kB are present in high levels in both heart and testis (5). Interestingly, however, inappropriate expression of Prnd mRNA in the CNS is associated with a neurodegenerative syndrome. The Rcm0 and Ngsk lines of Prnp 0/0 mice display normal early development but develop a widespread loss of cerebellarPurkinje cells at around 70 weeks of age, which is accompanied by a progressive ataxia (5,6). These mice differ from asymptomatic Prnp 0/0 mice in that they carry chromosomal deletions which extend beyond the Prnp open reading frame (ORF) into intron 2 at the 5ı end, and extending into the 3ı untranslated region of the Prnp mRNA (encoded by exon 3) at the 3ı end.

These deletions eliminate the Prnp exon 3 splice acceptor site and result in the generation of high levels of Prnd transcripts in the brain by an unusual mechanism involving exon-skipping and intergenic splicing (5). Although it is possible that long-range perturbations in chromatin structure underlie the ataxic phenotype of the Rcm0 and Ngsk lines of Prnp 0/0 mice, the parsimonious interpretation of the available data is that Dpl overexpression is toxic to Purkinje cells. Lastly, it is noteworthy that disease in Ngsk Prnp 0/0 mice is abrogated by reintroduction of a wt Prnp transgene (7), providing genetic evidence for an interaction between PrP C and the molecules that cause Purkinje cell loss.

In this work we sought confirmation of previous predictions regarding the biochemical and structural properties of the Dpl protein. We demonstrate that recombinant Dpl refolded under standard conditions adopts a predominantly € -helical conformation, and that Dpl expressed in mouse neuroblastoma cells possesses two N-linked oligosaccharides and is attached to the cell surface via a GPI-anchor. These findings show that Dpl and PrP C resemble one another with respect to their post-translational modifications and sub-cellular localization. We also report that Dpl protein is present in the brains of Rcm0 mice, but is undetectable in the brains of wt FVB/N mice and the ZrchI line of PrP-knockout mice. Our findings are discussed in the context of the transgene rescue experiments indicating interactions between Dpl and PrP C .

First report of polymorphisms in the prion-like protein gene (PRND): Implications for human prion diseases

Neurosci Lett. 2000 Jun 2;286(2):144-148. 
Katell Peoc'h, Caroline Guerin, Jean-Philippe Brandel, Jean-Marie Launay, Jean-Louis Laplanche
Comment (webmaster): This is an important contribution that looks at doppel sequence in some 268 individuals, including 95 with sporadic CJD and others with various prion mutations, as well as 2 French nvCJD cases. There was no smoking gun linking any of the 5 alleles to CJD or other neurological disease, though doppel may yet have its own (recessive?) disease. One way to get overproduction of prion protein (accounting for an unknown fraction of sporadic CJD) is to have less prion mRNAs diverted to doppel(assuming that human, like mice, have chimeric transcripts).

The doppel literature remains still quite managable as this is just the second paper on it (not counting 28 webmaster articles and 20 graphics ). A long-awaited paper on chimeric mouse mRNAs is still unavailable: Li,A., Sakaguchi,S., Atarashi,R.... and Shigematu,K. "Identification of a novel gene encoding a PrP-like protein expressed as chimeric transcripts fused to PrP exon1/2 in ataxic mouse line with a disrupted PrP gene" Cell. Mol. Neurobiol. (2000), not out as of June 2000 issue, a journal with no functioning online site.

The most common allele M174T was described in Sept 99 by the webmaster based on comparison of various genomic and EST efforts, though allele frequencies could not be estimated accurately. T174M would be cleaved off in the event of a GPI anchor. In the comparable region of the prion gene, the seemingly innocuous M232R and P238S are both inexplicably linked to familial CJD. Even though doppel is not plausibly amyloidogenic, it is still premature to conclude that T174M is without its effects, though near-equal frequencies do not favor pathogenicity per se.

Laplanche's group also found provocative T26M and P56L coding polymorphisms. The first may disrupt cleavage of the signal peptide (whose actual location is a little ambiguous); the latter is a quite conserved proline (reminiscent of, though not the homologous counterpart to either P102L or P105L from familial CJD). These residues occur too early in the sequence to be modelled in 3D using homology threading to the prion gene so their effect on structure cannot be deduced. It would be helpful to have cow, sheep, and deer doppel sequences.

The paper also has an interesting section on CJD kindred tracking using doppel haplotypes as close-in downstream markers. These are convenient becausethe small doppel gene would be often characterized anyway. Some CJD alleles have arisen repeatedly in separate founders; doppel can sometimes sort this out. It turned out that different PRND alleles were linked to E200K mutations. It seems in France three distinct mutational events leading to E200K occurred independently from those in the Mediterranean and Slovakian patients.

For comparative purposes, a similarly sized pool of control EU blood donors would exhibit only one prion coding polymorphism and two silent ones -- coding alleles may more abundant in doppel suggesting less selective pressure:

poly   freq
M129V	.34
A117A	.05
G124G	.005

Highlights of article:

"The aim of this study was to investigate the possible involvement of genetic variation in the prion-like protein gene (PRND), which encodes the doppel protein (Dpl), in the etiology of human prion diseases. Patients with sporadic, infectious or genetic forms of human prion diseases and controls were systematically screened, using the single-strand conformational polymorphism method, for genetic variants of the PRND gene. Four polymorphisms in PRND (three structural changes, T26M, P56L and T174M and a silent polymorphism, T174T) were detected. No strong association was found between any of these polymorphisms and human prion diseases but certain PRND alleles may be useful markers for tracing the chromosomal ancestry of PRNP mutations. Although genetic variation in PRND does not seem to play a major role in the pathogenesis of prion diseases, this first report of PRND polymorphisms may open up new possibilities for investigating the involvement of such polymorphisms in other human diseases.

The first prion-like protein gene was recently discovered in mouse and called prnd [10]. The equivalent human prion-like protein gene (PRND, OMIM *604263) maps to chromosomal location 20p12-pter, 27 Kb downstream from the human prion protein gene (PRNP) [10]. Prion protein (PrP) is the key element in the physiopathology of human and animal prion diseases. These rare diseases are transmitted by infectious particles called prions [16]. Prion diseases may be sporadic, infectious, or genetic. They involved the accumulation in the brain of an abnormal conformer of PrP, a cell-surface protein thought to be implied in protecting the cell against oxidative stress [2]. The mouse and human PRND genes encode a PrP-like protein 179 amino acids long in mouse and 176 amino acids long in humans, called Doppel (Dpl). The mouse and human Dpl proteins are 73% identical to each other and about 25% identical to the carboxy-terminal part of the known prion proteins [10].

Structural predictions for Dpl suggest that it is anchored at the cell surface, like PrP, by glycosylphospatidylinositol anchor. However, the function of the protein remains elusive. PRNP and PRND are transcribed separately but chimeric transcripts are also generated by intergenic splicing suggesting that the synthesis of the corresponding proteins may be synchronized. Disruption of the prnp gene in two lines of mice has been found to result in the overexpression of prnd which is thought to be responsible for late-onset ataxia, loss of cerebellar Purkinje cells and demyelination [17, 10].

Reintroduction of a mouse prion protein transgene prevents the ataxic phenotype [13] suggesting either direct interaction between PrP and Dpl, or competition for a common receptor at the cell surface. Based on this possible link between Dpl and PrP, we looked for polymorphism and mutation in PRND. We assessed the likehood of such genetic variation being involved in the susceptibility of humans to prions and its possible effects on the phenotype of human prion diseases in their sporadic, infectious and genetic forms.

DNA from patients with neuropathologically confirmed prion diseases, mainly Creutzfeldt-Jakob disease (CJD), referred to the French CJD Surveillance Network, were studied. In each case, we screened the patientıs DNA for mutations in the PRNP gene and the genotype at the polymorphic codon 129 (M/V) was established. This polymorphism has consistently been shown to influence the susceptibility to prion diseases, both sporadic and infectious, and to modify the phenotypic expression of some inherited prion diseases [7, 8].

The patients studied included 95 sporadic CJD cases (46 PRNP 129MM, 24 129MV, 25 129VV), six iatrogenic CJD cases resulting from treatment with human growth-hormone (hGH, two PRNP 129MM, two 129MV, two 129VV) and two French cases of variant CJD (vCJD, two PRNP 129MM). In addition, we also included forty-eight patients with inherited prion diseases, carrying the following PRNP mutations: P102L (n=2), A117V (n=2), 129VD178N (n=3),129MD178N (n=2), V180I (n=1), E200K (n=31), V210I (n=6) and E211Q (n=1). These patients belonged to 42 different families, from which 29 asymptomatic relatives were also studied. One hundred and six Caucasian controls with no family history of neurologic disorders were also analyzed.

The 528-bp coding sequence of PRND (Genbank accession number AF106918) located in a single exon, was amplified by PCR using the following primers: forward 5'GACCCACCGCCGTTTCTCT3' (20938-20956) and reverse 5'CGTGGGTTTGGGGGAGAACA (21585-21566), in standard conditions at an annealing temperature of 58°C. The 648-bp PCR product was then cleaved by the restriction endonuclease Bgl I into three fragments (277, 186 and 185-bp) which were screened for nucleotide variations by the single-strand conformational polymorphism (SSCP) method using Genegel Excel 12.5 gels and a GenePhor system (Amersham-Pharmacia-Biotech, Uppsala, Sweden) with electrophoresis at 25 mA, 15 W at 15° C for 1h 45min. The gel was silver stained and PCR products with atypical electrophoretic migrations were directly sequenced on both strands by the dideoxy chain-termination method using an ALF Express automated laser fluorescence sequencer (Amersham-Pharmacia-Biotech, Uppsala, Sweden).

SSCP analysis of the PRND coding sequence was carried out for a total of 286 individuals comprising patients with prion diseases and their relatives and healthy controls. We observed several distinct electrophoretic patterns. Direct sequencing of the corresponding PCR products identified four nucleotide substitutions in the PRND coding sequence which occurred at CpG dinucleotides, known hot-spots of mutation (Figure 1, a). A C-to-T substitution at the second nucleotide of codon 26, led to the replacement of threonine by methionine in Dpl.

Two different prediction programs [11, 12] have predicted cleavage of the signal peptide after residues 23 and 26, respectively (Figure 1, b). It is therefore unclear whether residue 26 is in the signal peptide or the mature protein. According to von Heijne's method for predicting signal sequence recognition [18], the presence of a methionine instead of a threonine at position 26 should prevent peptide signal cleavage. This mutation may profoundly change the intracellular trafficking of the protein, leading to the accumulation of M26Dpl in the cytoplasm of PRND expressing cells.

A C-to-T substitution introducing a Msp I restriction site and a predicted proline-to-leucine change in Dpl was observed at the second position of codon 56. Amino acid 56 is part of a stretch of 25 residues that are highly conserved in mouse and rat Dpl [10]. Two nucleotide changes were observed at codon 174: i) a C-to-T substitution, introducing a Nla III restriction site and resulting in the replacement of threonine by methionine, at the second position of the codon, and ii) a G-to-A silent substitution (174 aca was called " T2174") at the third position of the codon. Amino acid 174 is part of the the carboxy-terminal hydrophobic tail of Dpl, predicted either to form part of transmembrane domain or to be removed before attachment of the GPI anchor, probably at glycine 152 (equivalent to G155 in mouse, [10]).

Based on these four polymorphisms, we were able to define five different PRND alleles in controls and patients: T26 P56 M174, T26 P56 T174, T26 P56 T2174, M26 P56 T174 and T26 L56 M174. Genotype distribution and allele frequencies were determined from the SSCP profiles and by direct sequencing and Msp I and Nla III digestion (Table 1).

Both controls and sporadic CJD patients showed Hardy-Weinberg equilibrium for genotype distribution. No significant difference was found between sporadic CJD and controls in PRND genotype distribution (c2= 5.82, P= 0.12, 2 df.) or allele frequency (c2= 2.74, P= 0.25, 2 df.) suggesting that PRND polymorphisms do not increase susceptibility to sporadic CJD. No preferential association between the three genotypes at PRNP codon 129 and PRND genotypes was identified in patients with sporadic CJD (c2= 0.22, P= 0.99, 4 df.).

In particular, three of the seven patients carrying the PRND M26P56T174 allele were PRNP 129MM, two 129MV and two 129VV. Only the three patients heterozygous for T26L56M174 were all homozygous for methionine at PRNP codon 129, indicating possible linkage between PRNP 129M and this allele. These data show that the strong association between sporadic CJD and the homozygosity for a methionine at PRNP codon 129 [14, 8] is not due to linkage disequilibrium between the PRNP 129M allele and a recessive mutation in the coding sequence of the neighboring PRND gene.

We also explored the possible effect of PRND polymorphisms on the clinical phenotype of the sporadic CJD form, within the limits set by the low frequency of some alleles. The seven patients with the PRND M26P56T174 allele had a mean age at onset of 71 years (range 51-80) with a mean disease duration of 5 months (range: 2-15). These characteristics did no differ significantly from those of the three T26L56M174 carriers (mean age at onset: 70 yrs. (range: 69-72); mean duration of disease: 2 mo. (range: 2-6) or the other sporadic patients (67 yrs. (23-85); 4 mo. (1-74); n=85)(P= 0.42 and P= 0.34 for age and disease duration, respectively, KW test). The clinical signs of these patients were unremarkable and undistinguishable from those of classic sporadic CJD.

The number of patients with infectious prion diseases (iatrogenic or vCJD due to contamination by the agent of bovine spongiform encephalopathy (BSE)) was small. However, the different PRND genotypes observed in these patients, would suggest that an association between a particular PRND genotype and a greater susceptibility to exogenous prions is unlikely.

We also analyzed PRND in a subtype of prion diseases, inherited forms. Inherited prion diseases have been linked to or associated with an increasing number of mutations in PRNP [4,15]. PRND alleles linked to the mutant PRNP genes were unambiguously identified in 25 of 42 families carrying various PRNP mutations (Table 2). Most of the mutations studied were co-transmitted with the frequent PRND T26P56M174 or T26P56T174 alleles with the notable exception of the E200K mutation.

This mutation is present in about 70% of patients with inherited prion diseases and is responsible for familial clusters of CJD, mainly in Mediterranean Jews [5] and Eastern European non-Jewish families [6]. The rare PRND T26L56M174 allele was co-transmitted with the E200K mutation in five families (10/31 patients) analyzed but, as in sporadic cases, seemed to have no effect on the age at onset (mean: 64.5 yrs. (range: 46-81), n= 10, vs 63 (range: 47-78) in other patients, n= 21, P= 0.49, KS test) nor the disease duration (5.5 mo.(1-14) vs 4.5 mo.(2-32), n= 21, P= 0.60, KS test). There was no significant difference in clinical features between E200K patients, carrying and not carrying the PRND T26L56M174 allele.

As three different PRND alleles co-segregated with the disease in families with different geographical origins, we also explored the usefulness of PRND polymorphisms as potential markers of the chromosomal ancestry of the E200K mutation (Table 2). All patients with Tunisian, Libyan or Greek ancestry carried the PRND T26P56M174 allele and all belonged to Mediterranean Jewish families. The same allele was also present in a Spanish non-Jewish patient.

This association is consistent with recent studies with polymorphic microsatellite markers showing a common chromosomal ancestry for the PRNP E200K mutation in Spanish and Mediterranean Sephardic patients [9]. In contrast, two different PRND alleles were co-transmitted with the PRNP E200K mutation in families of French ancestry: T26P56T174 in three families and T26L56M174 in five families. The PRND T26L56M174 allele was found in five previously unconnected families, three of which have already been reported [1, 3], probably due to a founder effect. As PRND polymorphisms are the closest known polymorphic markers to PRNP, our data suggest that the PRNP E200K mutation in French patients may result from at least three different mutational events that occurred independently from those in Mediterranean and Slovak populations.

In summary, we have described four new polymorphisms in the PRND coding sequence. It is currently unkown whether these polymorphisms affect mRNA stability or Dpl post-translational modification. However, variations in the PRND gene does not seem to play a major role in the pathogenesis of human prion diseases, either sporadic, infectious or inherited. Further investigation is required to assess the potential consequences for prion diseases of changes in PRND gene expression. The function of Dpl is unknown and the possibility that there are Dpl-associated diseases unrelated to prions cannot be excluded. Our identification of single nucleotide polymorphisms in PRND should make it possible to determine the involvement of this gene in other human diseases, using a candidate gene strategy.

Doppel allele resourse

6 Mar 00 webmaster.  Doppel counterpart to familial CJD point mutation database.
Color key:disease(0)*stop (0)silent(1)neutral(3)**based on consensus opinion
Initial Final Initial AA Final AA Codon Reference
cagACGagg1 ATG thr met T26M Laplanche (Neurosci Lett. 2000;286(2):144-148.)
cgcCCGgga1 CTG pro leu P56L Laplanche (Neurosci Lett. 2000;286(2):144-148.)
ctcACGgtg1 ACA thr thr T174T2 Laplanche (Neurosci Lett. 2000;286(2):144-148.)
ctcATGgtg1 ACG met thr M174T webmaster

0. Laplanche et al.found no statistical association of any of 5 doppel alleles to sporadic CJD, iatrogenic CJD or nvCJD (two French cases examined); they note T26M may not be functional as it near the inferred signal peptide boundary possibly affecting processing to mature protein; P56L disrupts a conserved doppel proline which may have a special structural role as a imino acid not conservatively replaced by leucine.

1. Spontaneous deamidation of methylated C in CpG of DNA enhances mutational frequency (hotspots). This is seen in the prion gene as well.

2. This may be a secondary mutation within in the T174 population, ie, M174T created a CpG hotspot that then further mutated to the second threonine codon (as well as back to met). Mouse is ATT isoleucine at 174, requiring a double substitution from either threonine codon but only a single transition from methionine. Anthropoid ape sequences might resolve this issue

Doppel Haplotype Frequencies
Key: T:ACA in place of ACG
 T26M: ACG->ATG   P56L: CCG->CTG  
T174M: ACG->ATG  T174T: ACG->ACA

Doppel reference sequences:

>doppel ancestral sequence

>dopMus U29187

>dopRat [taken from JMB figure]

>dopHuman U29185 (AL133396) T174; AF106918 M174

>dopovi AJ251331 sheep

>dopbos AJ251332 cow

>dopMus CDS 
        1 atgaagaacc ggctgggtac atggtgggtg gccatcctct gcatgctgct tgccagccac
       61 ctctccacgg tcaaggcaag gggcataaag cacaggttca agtggaaccg gaaggtcctg
      121 cccagcagcg gcggccagat caccgaagct cgggtagctg agaaccgccc aggagccttc
      181 atcaagcaag gccggaagct ggacatcgac tttggagcag agggcaacag gtactacgcg
      241 gctaactatt ggcagttccc tgatgggatc tactacgaag gctgctctga agccaacgtg
      301 accaaggaga tgctggtgac cagctgcgtc aacgccaccc aggcggccaa ccaggctgag
      361 ttctcccggg agaagcagga tagcaagctc caccagcgag tcctgtggcg gctgatcaaa
      421 gagatctgct ccgccaagca ctgcgatttc tggctggaaa ggggagctgc gcttcgggtc
      481 gccgtggacc aaccggcgat ggtctgcctg ctgggtttcg tttggttcat tgtgaagtaa

>dopHum CDS
        1 atgaggaagc acctgagctg gtggtggctg gccactgtct gcatgctgct cttcagccac
       61 ctctctgcgg tccagacgag gggcatcaag cacagaatca agtggaaccg gaaggccctg
      121 cccagcactg cccagatcac tgaggcccag gtggctgaga accgcccggg agccttcatc
      181 aagcaaggcc gcaagctcga cattgacttc ggagccgagg gcaacaggta ctacgaggcc
      241 aactactggc agttccccga tggcatccac tacaacggct gctctgaggc taatgtgacc
      301 aaggaggcat ttgtcaccgg ctgcatcaat gccacccagg cggcgaacca gggggagttc
      361 cagaagccag acaacaagct ccaccagcag gtgctctggc ggctggtcca ggagctctgc
      421 tccctcaagc attgcgagtt ttggttggag aggggcgcag gacttcgggt caccatgcac
      481 cagccagtgc tcctctgcct tctggctttg atctggctca cggtgaaata a

>dopovi AJ251331
        1 atcaagtgga accggaaggt cttgccaagt acctcccagg tcacggaggc ccacactgcg
       61 gaaatccgcc caggggcctt catcaagcaa ggccgaaagc tggatatcaa ctttggagtg
      121 gagggcaata ggtactatga ggccaactat tggcagtttc ctgacggcat ccattacaac
      181 ggctgctccg aggccaatgt caccaaggaa aagtttgtca ccagctgcat taatgccacc
      241 caggtggcaa atcaagagga actgtcccgt gagaaacaag acaacaagct ttaccagcgg
      301 gtcctgtggc agctgatcag ggagctctgc tccatcaagc actgtgactt t

>dopbos AJ251332
        1 atcaagtgga accggaaggt cttgccaagt acctcccagg tcacggaggc ccgcactgcg
       61 gaaatccgcc caggggcctt catcaagcaa ggccgaaagc tggatatcga ctttggagtg
      121 gagggcaata ggtactatga ggccaactat tggcagtttc ctgacggcat ccattacaac
      181 ggctgctcca aggccaatgt caccaaggaa aagtttatca ccagctgcat taatgccacc
      241 caggcggcga atcaagagga actgtcccgt gagaaacaag acaacaagct ttaccagcgg
      301 gtcctgtggc agctgatcag ggagctctgc tccaccaagc actgtgactt 
Additional specifically sought mRNA sequences are available representing various splicing scenarios: AF165165 AF165166

Doppel ESTs are rare and generally involve 3' UTR. Unfortunately, the word 'doppel' and symbol PRND are not being used consistently in GenBank entries; searches must be done with 'prion protein-like'. There are 8 entries as of 1 June 00.

Gene models of the prion-doppel complex in various species will be available here shortly.

CpG hotspots for doppel

3 Mar 00 webmaster research
Some progress can be made in predicting doppel mutations that will be found in anticipated patient and control screens if one recalls:

"The fifth base in human DNA, 5-methylcytosine, is inherently mutagenic. This has led to marked changes in the distribution of the CpG methyl acceptor site and an 80% depletion in its frequency of occurrence in vertebrate DNA. The coding regions of many genes contain CpGs which are methylated in sperm and serve as hot spots for mutation in human genetic diseases. Fully 30-40% of all human germline point mutations are thought to be methylation-induced even though the CpG dinucleotide is under-represented and efficient cellular repair systems exist." [Bioessays 1992 Jan;14(1):33-36]

Human doppel has 20 uniformly distributed CpG sites in its coding region, a somewhat higher density than the longer prion protein which has 23, as expected from doppel's more rapid rate of evolution. Doppel and prion have very similar C+G percentages (59.9 and 58.0, resp.) leading to base-composition expectations of CpG occurences of 47.6 and 61.6 instances, and thus to a higher depletion level of CpG in prion protein.

Using two public genomic sequences, the webmaster described the first doppel allele, T174M, on 28 Sep 1999. This involved a CpG site. Still-rare doppel coding region ESTs do not give any further alleles as of 1 Mar 00. No variants have been seen yet in mouse genomic or EST data; rat doppel genomic sequence is implausibly still a tightly guarded petty secret despite publication of the amino acid sequence in a JMB graphic submitted May 1999 (and the surge in nvCJD cases).

CpG sites generally mutate to TpG or CpA after deamination. The 20 sites then give 40 predicted changes (shown in the graphic below as the dopHSCT and dopHSCA lines; green arrows are CpG sites; red arrows predict non-conservative amino acid substitutions, the best candidates for loss of doppel function). Doppel amino acids conserved in rat, mouse, and human but changed non-conservatively (in the PAM matrix sense) by a CpG mutation are the best candidates for loss-of-function change, especially in very ancient conserved regions (eg, alignable and conserved in vertebrate prion). Dopple has essentially the same fold as prion protein in alignable regions; note CpG codons 37, 55, 56, 153, 156, and 174 are not in threadable structural regions.

Mutations to cysteine might present special problems in a protein having two disulfides and an unpaired cysteine already, giving rise to cross-linked or mis-linked protein despite the best efforts of disulfide isomerase. Mutations that disrupt secondary structure might also cause problems; note however amino acids in these regions are not especially conserved in doppel alignments. Mutations affecting signal or GPI processing, while not affecting structured regions, could derail correct processing, causing the protein not to reach the destination where it normally functions (and subject it to protein to rapid degradateion). Doppel is unlikely to be amyloidogenic so there is no evident reason for mutations also to cluster in the distal helix or be autosomal dominant (except possibly indirectly through toxic gain of function in prion protein).

Based on these considerations, the CpG mutations asterisked below may be the first to be found in doppel disease (which could be very rare if autosomal recessive). R55C and R64C each create a free cysteine at a conserved site. P56L could also be troublesome since it occurs in a highly conserved stretch and proline often has a unique structural role (unknown in this instance). G71R disrupts the highly conserved interior core with an unacceptable charge and size change and is the single safest prediction. R134W occurs in the third helix and places a bulky hydrophobic residue on the exterior where it might also interfere with protein-protein interactions and also creates a WW pair potentially competing for structure.

R37W*   E73K
R37Q    D87N
R55C*   G93S
R55H    A114V
P56L*   A115V
R64C*   R134W* 
R64H    R134Q
D67N    A153T
G71R**  A153T
R156W   R156Q
It is important to keep in mind the distinction between a mutant and an allele. The former is a unique event in an individual whereas the latter is an event later fixed a population to a detectable percentage either through neutral drift or balancing selection. CpG hotspots enhance mutations but codons with favorable potential as positive alleles seldom can arise from CpG; on the other hand, for neutral drift there are many more opportunities from CpG events for fixation plus multiple independent occurences of the same event. Since doppel is evolving more rapidly than prion protein (suggesting less selective pressure), alleles will be more abundant in screens. There will also be more rapid establishment of mutations to CpG.

The best studied human gene with respect to actual founding mutations is probably coagulation factor IX. [See Hum Gen 105:629-640 1999.] Since CpG mutations primarily occur during DNA replication and germline replication takes place in 15-fold excess in male germline cells, a marked paternal bias of 6.6:1 occurs in the case of CpG transitions. In this gene, 15% of all mutations were CpG, 56% were transitions, 25% transversions, and 4% small deletions.

Detectability of doppel disease in humans is really driven more by autsomal dominance. Here again, the CpG sites themselves may not be able to give rise to such mutations. Many genes simply do not have the potential for autosomal dominant mutations because a second normal allele (or a secondary gene) can compensate through regulatory feedback for a defective allele. Note that the reported mouse ataxia is a (recessive?) double knockout mutation of the prion gene itself causing overproduction of doppel in the absence of prion protein, an unlikely event to arise spontaneously in humans.

Doppel polymorphisms might manifest themselves not as a stand-alone disease but through modulatory effects on prion disease. At this time, there is no real understanding of the normal roles of either gene, much less of functional interactions with each other or unrelated proteins. If a modulatory role is found, the existence of several moderately common doppel alleles would compound the combinatorics of sorting out haplotypes. In a rare disease, it would be difficult to find statistically adequate numbers of patients with the same prion+doppel genotype. One certain benefit of doppel alleles would be a close-in marker to identify kindreds and distinct founder events in P102L, D178N, and E200K. Alleles may also allow detection of unsuspected kindreds among nvCJD cases.

Note, however, if deleterious single doppel alleles are tolerated and moderately common, then compound heterozygotes are more common than the extremely rare homozygotes. That is, a deleterious allele with frequency 1 per million occurs as a homozyogote with frequency 1 per trillion and so is effectively unobservable. However, if that allele is tolerated it may drift to higher frequencies along with deleterious changes at other codons on separate chromosomes. This raises to odds beyond just the set of homozygotes (sum of squares of individual frequencies) to compound heterozygotes having different deleterious alleles on the two chromosomes (sum of 2 x freq1 x freq2). For many genes on autosomal chromosomes, these are the most commonly observed type of mutation.

Finally, it is easy to enumerate CpG sites in non-coding regions. However, predicting the effects of these is much harder than in proteins (where it is already hard enough). The key role of overproduction in mouse ataxia suggests consideration of regulatory or splice mutations. The CpG island of the prion gene itself is protected; doppel lacks such a feature in the vicinity of its own promoter.

Odd new doppel EST

Webmaster analysis of 25 May 00 GenBank doppel-related EST entry
The doppel mouse mRNA sequence below just appeared on GenBank; it is comprised of 3' UTR and does not reach into coding sequence. The first 31 bp do not match anything, including the chimeric prion exon 2 splice junction. Oddly, the match is only 92% identity to known doppel transcripts such as AF192384 or U29187 (18 scattered mismatches) or to the mouse EST AW230467. ESTs today from experienced high-throughput labs do not have an 8% error rate, rather 0-1 errors over such a 269 bp stretch..

From mouse genomic U29187, the match of 30-140 of the new sequence is somewhat suspicious: positions 37557 - 37793 represent a retrotransposon [c(37553..37665) PB1D9 SINE/Alu ] that can be part of normal doppel mRNA in the 3' UTR region. There may be many divergent copies of this Alu within the mouse genome.

The two options, discounting outright sequencing error, are a second mouse doppel gene (or transcriped pseudogene) expressed in neonate skin or alternately, merely another member of the PB1D9 family that has diverged to the 92% match level. The second possibility is inconsistent with the high quality but imperfect match to 129 bp of authentic non-retrotransposon doppel mRNA following the PB1D9. No false polyA primer is evident in the vicinity in U29187.

Either possibility is consistent with the 32 bp of non-matchable leading sequence [which do not match mouse prion exon 2 splice donor] and the 92% overal match. The gene scenario cannot distinguish between gene and pseudogene since the EST does not reach the coding sequence. Doppel has such low levels of expression in testes that a pseudogene seems implausible. It could represent a fairly recent gene doubling (because PB1D9 would not have inserted twice in the same spot), in tandem or in translocation, so one that might occur only in the rodent lineage. There is not enough complete or incomplete (htgs) mouse genomic sequence to confirm such an origin. There are no further overlapping matching ESTs allowing unigene construction. No more can be done without further experimental input.

BB001678 269 bp mRNA EST 25-MAY-2000
RIKEN full-length enriched, 0 day neonate skin Mus musculus cDNA 

   1 cccccgccgg ttccctatga ctgtcagtat ttttgttttc aagatagaaa tttttctaaa
  61 tagcccagct tgaccccata atcatgatct tctgccctac gcctcctgga tgctgggatt
 121 gcagacatgc caccaagccc aatatgtgtt tacttttgat ttcccctggc cctatgtgca
 181 tgtgtgtttt taggcttttt aagatgctca ttgcgcacag agttttttgc aaacactgtt
 241 tgtttttaaa acccactttg gtaaaagcg

BB001678 repeat-masked 30-140 for PB1D9, a SINE/Alu:


BB001678 alignment target: U29187 Mus musculus long incubation doppel genomic:

Query: 32    tttgttttcaagatagaaatttttctaaatagcccagcttgaccccataatcatgatctt 91
             ||||||||| |||||||||||||||||  |||||||| ||||||||||||||||||||||
Sbjct: 37557 tttgttttctagatagaaatttttctatgtagcccaggttgaccccataatcatgatctt 37616

Query: 92    ctgccctacgcctcctggatgctgggattgcagacatgccaccaagcccaatatgtgttt 151
             ||| |||  |||||||| ||||||||||||||||||||||||||||||||||||||||||
Sbjct: 37617 ctggcctctgcctcctgaatgctgggattgcagacatgccaccaagcccaatatgtgttt 37676

Query: 152   acttttgatttcccctggccctatgtgcatgtgtgtttttaggctttttaagatgctcat 211
             |||| |||| ||||||||||||||||||||||||||| |||||| | ||||||||||||
Sbjct: 37677 acttctgatgtcccctggccctatgtgcatgtgtgttcttaggcatcttaagatgctcac 37736

Query: 212   tgcgcacagagttttttgcaaacactgtttgtttttaaaacccactttggtaaaagc 268
             ||| ||||||| | | |||||||||||| ||||||||||||||||||||||||||||
Sbjct: 37737 tgcacacagagctctctgcaaacactgtctgtttttaaaacccactttggtaaaagc 37793

Doppel directory

30 May 00 webmaster
Directory shows articles and  associated graphics (***) in reverse chronological order.
New doppel neighbor: a tough little gene
Doppel transcripts as of Aug 00: rodents outnumber human
Doppel polymorphisms in nvCJD
Doppel: 5 human alleles
Doppel allele and sequence resourse
CpG hotspots for doppel
***Doppel CpG hotspots relative to conserved codons

Alternative splicing across prion-doppel tandem duplication
*** new mouse doppel chimeric mRNAs
*** CDC42 partial tandem
*** gorilla CEL tandem
Doppel disease: paralogue compensation

Doppel delay
Origin of doppel chimeric transcripts
*** mechanistic origin of tandem duplication
*** evolution of doppel signal region
*** prion-doppel duplication within vertebrate evolution
*** ancestral doppel sequence
*** Doppel disulfide position and effects on stability
First human doppel allele found: M174T
Doppelcross: cis-trans genetic issues
Alignments: getting them straight
Doppel disulfide in 3D (and related issues)
Glycosylation sites in bird prion and mammalian doppel
Candidate exon 2 switch alleles
Intergenic exon 1 is within a Mer115 insertion element

Doppel prion upregulated in knockouts
Better chr 20 microsatellites
*** status of chromosome 20
*** nested retrotranspositions 5' of prion gene
Did a tandem gene doubling created the prion superfamily?
Details on RPS4X pseudogene
Massive parasitism of chromosome 20
Have any neighboring genes been missed?

Prnd and Prnp gobbled up by human genome project
*** making sense of Blast(est) matches
*** doppel human and mouse comparative annotation
Which came first, the prion or the ghost?
*** timing of divergence of prion and doppel
Ghost prion promoters
Bridging the gap
Pseudogene prion neighbor: isopentenyl diphosphate isomerase psIPP
*** chromosome 20 contigs relative to prion and doppel
Prion paralogue found adjacent to main gene!
*** doppel exons, introns, repeats, and ESTs
*** Doppel/prion alignments: blast
*** Clustalw alignment
*** doppel secondary structury predictions
Long and short incubation mouse prion genes compared
Zebrafish RH panel locates prion gene
*** Prion gene probable location on LG17 of zebrafish
Anti-prion 4.5 kb mRNA: a harmless prion pseudogene?

Prion genomics round-up: human exon 2 EST expressed in amygdala?

Sun, 7 May 2000 webmaster
Mammalian prion genes have two 5' untranslated exons. Humans have always been a bit of a mystery in that all mRNAs to date fail to contain exon 2 (as identified on the basis of strong homology). Interest in exon 2 is heightened by the discovery that trans-splicing to doppel takes off at the exon 2 splice donor in mouse.

Checking dbEST, which now has 1,856,102 human entries, with human exon 2 probe on 7 May 00 turned up AL157616, a 607 bp mRNA EST posted 24 Feb 2000 by the German CancerResearch Center. Interestingly, the tissue source for the mRNA was amygdala.

This EST begins 1 bp before the start of the 99 bp putative exon 2, contains all of it, and then continues through the splice donor site for another 507 bp of adjacent intron 2, that is, from 15389-15995 in human prion GenBank entry U29185.

This is not an ideal situation for analysis even though exon 1 ends in the G nucleotide that precedes exon 2 -- that is not enough to conclude that splicing took place between exon 1 and 2. There is no evidence of a splice to exon 3 or to doppel. The quality of the EST is excellent: 100% agreement on exon 2 and only two distal errors overall over 607 bp.

Very few ESTs extend into exon 1: AL119735 begins at position 84 of exon 1 and continues into exon 3; AU077172 begins at position 91 and also skips to exon 3; AL119841 barely catches exon 1 at 133 and continues in exon 3. The problem of course is that EST reads are fairly short and the prion gene has a very long 3' UTR tail.

One possibility is an artefact due to a happenstantial polyA signal within intron 2 that lead to false priming; along these lines, the known sequence of the EST is immediately followed by ...GAATA GTGTGTGCAG CAATTCTCAC AGCTAGGAGA ATTTTTTTT. Intron 2 has no retrotransposons in this region; the nearest element, MER5B, is found at 16201-16334.

In conclusion, while this EST suggests tissue-specific use (in the amygdala and possibly elsewhere) of human prion exon 2 occurs (and thus accounts for its conservation), further experimental work on the clone is needed. As only the 5' sequence of the (readily available) clone insert was determined, it could be sequenced it further to see if later (alternative) splicing occurs. A more informative approach would explicitly search for exon 2 expression in amygdala, eg, determine 1-3 to 1-2-3 mRNA ratios.

Epithelial and endothelial expression of the green fluorescent protein reporter gene under the control of bovine prion protein (PrP) gene regulatory sequences in transgenic mice

May 2, 2000 PNAS  
Catherine Lemaire-Vieille, Tobias Schulze,...and Jean-Yves Cesbron
The identification of the cell types expressing PrPc is necessary to understanding how the agent replicates and spreads from peripheral sites to the central nervous system. To determine the nature of the cell types expressing PrPc, a green fluorescent protein reporter gene was expressed in transgenic mice under the control of 6.9 kb of the bovine PrP gene regulatory sequences. It was shown that the bovine PrP gene is expressed as two populations of mRNA differing by alternative splicing of one 115-bp 5' untranslated exon in 17 different bovine tissues. [confirming and extending a 1997 report by M Horiuchi et al. in BBRC.

The analysis of transgenic mice showed reporter gene expression in some cells that have been identified as expressing PrP, such as cerebellar Purkinje cells, lymphocytes, and keratinocytes. In addition, expression of green fluorescent protein was observed in the plexus of the enteric nervous system and in a restricted subset of cells not yet clearly identified as expressing PrP: the epithelial cells of the thymic medullary and the endothelial cells of both the mucosal capillaries of the intestine and the renal capillaries. These data provide valuable information on the distribution of PrPc at the cellular level and argue for roles of the epithelial and endothelial cells in the spread of infection from the periphery to the brain. Moreover, the transgenic mice described in this paper provide a model that will allow for the study of the transcriptional activity of the PrP gene promoter in response to scrapie infection.

Other genomic updates: This paper also provided 3500 bp of new upstream bovine sequence AF163764, which did not extend past known sheep sequence and was 94% identical to it on the overlap. The bovine upstream sequence had same set of nested retrotransposons as sheep. Some of these elements are seen in human as well; rodent 5' remains a puzzle upstream of the promoter.

That is, sheep, cow, and human are alignable for long distances whereas mouse is not; using non-repeat mouse regions against finished and unfinished human genome gives nothing. Either a large deletion occurred in mouse or a translocation occurred of a piece homologous to some other human chr than 20.

No zebrafish prion EST has emerged and the prion gene remains unmapped in the April 2000 Genomics Research consortium map zebrafish article. However a very exciting development related to tracing the prion gene back to fish has taken place: a French group has already sequenced 1/3 of the compact genome (fewer retrotransposons than mammals) of a fresh water pufferfish: Roest Crollius H., et al. Human gene number estimate provided by genome wide analysis using Tetraodon nigroviridis DNA sequence. Nat. Genetics. (2000) 25:235-238.

The sequence of human chromosome 21, in conjunction with chromosome 22, supports a very low total number of human genes, under 40,000 in total. Chromosome 20p, which contains the prion gene, also has a very low gene density and predicts a 38,000 total. The entire chromosome 21 had only 225 apparent genes. There were numerous cases of tandem duplication families and 59 pseudogenes. Long stretches of DNA had no gene candidates whatsoever; some 42,000 parasitic elements occur overall. The annotation philosophy was very stringent: single exon genes were discarded as well as genes and pseudogenes with long 3' or 5' UTR; the paper failed to provide gene models. Still, the detection methods did find 95% of the genes already known to be present. Most noteworthy:

"We have found some paralogous regions between chromosome 21 and other human chromosomes, which were also pointed out by metaphase FISH analysis of the corresponding genomic clones. For example, a 100-kb region of clone B15L0C0 located on 21p is shared with chromosomes 4, 7, 20 and 22. A second homologous region of 50 kb on 21q between 15,530 and 15,580 kb is shared with a segment on chromosome 16 between the genes 44M2.1 and 44M2.2. More details on these regions can be found at"

However, no further mention of synteny to chromosome 20 is found at that site. The sequence is given: HSB15L0C0 is 80692 bp in length with accession AL078471. The webmaster took that DNA sequence, masked it for repeats (total interspersed repeats: 38429 bp or 47.62 %), and compared both masked and unmasked versions to finishedand unfinished chromosome 20 sequence in an attempt to pursue the region or paralogy. No match was found; the authors, contacted for an explanation, were attending the Cold Springs Harbor genome meeting.

A region surrounding the attractin gene (slightly telomeric to the prion gene) on chromosome 20p13 has been analyzed for alternative splicing in the May 23, 2000 PNAS by Tang et al. The 30 exons straddle unfinished contigs of AL353193 (182,242 bp) in addition to a large finished stretch, NT_003217 (consisting of AL109805 , 73666 bp, which has SF3A3 pseudogene and 4 exons of the ATRN gene spanning LTGSSGFVTD...PHCTDNCGFP and AL132773, 104907 bp, which has the fully annotated ATRN gene from the Sanger Centre dated 4 May 00 extending from positions 2..103785 of length 970 amino acids begingging ISNSSDTVECEC.) Exon 1 has a perfect match to 45325-45813 of assembly fragment 00099 which extends from 9693-69,069 giving 59,377 bp of embedding genomic for exon 1.

Very close homologues for the attractin gene occur on chr 2, 10 (KIAA0534), and 11. This raises the possibility of a regional duplication of the chr 20p region and more copies of the prion and doppel genes. The origin and extent of regions paralogous to ATRN has not been clarified yet.

Including promoters and alternative poly A sites, the attractin gene is thus sprawled out over a very large region of chromosome 20, perhaps 250,000 base pairs. This is consistent with the low gene density seen in the prion-doppel region and with total gene estimates for the entire human genome derived from the completed sequences of chr 21 and 22.

Some further neighboring genes of the prion-doppel region are now annotated in finished contigs:

 AL121755 . Human DNA sequence...[gi:7406634] 
 HSDJ680N4  143892 bp    DNA             PRI       02-MAY-2000
 Human DNA sequence from clone RP4-680N4 on chromosome 20. Contains
 part of the CDS2 (CDP-diacylglycerol synthase (phosphatidate
 cytidylyltransferase) 2) gene, the UBE2D3 (ubiquitin-conjugating
 enzyme E2D 3 (homologous to yeast UBC4/5)) gene, part of a
 G-protein coupled receptor gene

 Human DNA sequence from clone RP5-1022P6 on chromosome 20 Contains part of the gene for KIAA1434
 protein, a pseudogene similar to ribosomal protein S18, a pseudogene similar to Eukaryotic initiation factor 4E, ESTs, STSs, GSSs and CpG Islands, complete sequence

Grandiose claims were made in JMB (pp. 523-544 doi:10.1006/jmbi.2000.3741) for new software 3D-PSSM that supposedly improved on the ability to recognize distant protein homologies:

"A method (three-dimensional position-specific scoring matrix, 3D-PSSM, to recognise remote protein sequence homologues is described. The method combines the power of multiple sequence profiles with knowledge of protein structure to provide enhanced recognition and thus functional assignment of newly sequenced genomes. The method uses structural alignments of homologous proteins of similar three-dimensional structure in the structural classification of proteins (SCOP) database to obtain a structural equivalence of residues. These equivalences are used to extend multiply aligned sequences obtained by standard sequence searches...."

The program, while it might have merit in some situations, performed poorly on various prion and doppel proteins, failing to identify as many homologues as a naive Blast search.

Wild claims are also being made in Science for imminent completion of a draft public human genome. The webmaster could find no evidence in the overnight public databases that completion of chromosome 20p is anywhere in sight. Many contigs contain large numbers of unassembled 1-2 kilobase fragments that are useless for gene discovery. The NCBI web page for chr 20 has seen little if any updating in the last six months. The Sanger Centre chr 20 web site has not been updated for news for years and its data notebooks remain klutzy and nearly useless to outsiders. The Dept of Energy abandoned its work on 3 chromosomes to pursue other interests, leaving behind incomplete coverage and no gene annotation; a press release incongrously celebrated finishing the job.

Proprietary EST and genomic data: Various biotech startups have taken out full page ads in Nature and Science, claiming huge private collections of data and powerful genomic software. Since the data is secret and few if any peer-reviewed publications occur, it is difficult to evaluate the claims. It can cost up $5,000,000 to gain full access to the data for twelve months. Sometimes the company offers a few free trials on its website; the webmaster tested various sites with doppel-related mRNAs.

1. allows visitors who briefly register to perform 6 free LabOnWeb analyses, including iRace elongation, functional analysis (SAS) and expression levels based on public SAGE libraries. Turn-around is slow, basically a whole day. Mostly it returns a small subset of what could have been found within minutes -- in better detail -- by a conventional GenBank Blast search. the 'lab notebook' concept is cute but works poorly as non-standard html. It was not able to extend any of the doppel test sequences fed it, though this is easily done at GenBank; it extended prion exon 2 but missed the chimeric extensions. The convenience it offers is that it bundles a large number of Blast searches, saving the visitor the boredom of many separate submissions.

2. On 31 May 00, offered broken-down beta software that could not even process guest registration. TIGR wants visitors to print, sign, and fax in a non-profit agreement to get a flatfile database of gene indexes, whatever these are. However, the Blast service works fine though it seems merely to return old GenBank sequences under new names. There is no question that the company has a lot of high quality genomic sequence, right now it isn't doing working scientists any good.

3. claims 3.9 million human ESTs, 50,000 Incyte-unique genes not available in the public-domain, 25,000 full-length or 5' complete gene sequences, and 837 Incyte-unique cDNA libraries. Maybe they do and maybe they don't. The fasta window does not allow numbers so the sequence must be cleaned, showing that nobody really uses it regularly. It is not automated: "an Incyte application specialist [salesman] will determine the best gene view, full-length clone, and most 5' EST view and clone that have an exact match to your probe." Then you have to confirm an order by telephone before getting to see anything -- it is all about money, not science.

The company didn't convince the webmaster that they know what they are doing, a Feb 2000 article stating confidently "a total estimate about 1600-1900 transcription units is roughly consistent with the expected number of genes believed to be on chr 22 given its size and gene density" -- meanwhile, back in the real world, the chr 22 sequencing team found only 545 genes for the completed sequence. They don't seem to understand that non-overlapping ESTs easily arise from a single gene and so are totally unsuitable for counting genes, as are CpG islands. The company has failed to release its version of chr 22 gene annotation for public scrutiny. Perhaps the focus is merely to throw as many press releases and patent applications as possible on the table. Their set of links is worth a look however.

4. is similar to LabonWeb but has more sophisticated bioinformatic software. Their interminably slow slide show explaining clustering and alignment protocols seems directed at stockbrokers, not scientists. The analysis of chromosome 22 supposedly identified 1,485 genes and 2,700 alternative splice forms. But nobody gets to see these; discrepancies with the published chr 22 paper are not explained. It seems that they mainly relied on consensus between multiple gene prediction programs. In the webmaster's experience, this is a poor idea: GenScan is the only gene prediction program that works worth a damn and even it is 50% and more false positives.

But the alternate splice claims could be important. Again, it is just hot air until methods, data, and results undergo public review. It raises the question of whether paid subscribers will attempt to publish scientific papers, patents, or drug approvals based on data that readers, examiners, and regulators have no access to -- not a viable proposal. GeneTool Lite 1.0 and PepTool 1.1 are available for free for Mac and PC. The first is a 16 meg download; the latter 7.6. Often it is better to stay on the web and simply move from one preferred tool to another, rather than stay in a limited proprietary environment. The large web site offers nothing on costs -- a sales rep must be contacted (no information is given by email either, you must speak with a salesman). Guest access doesn't allow any work to get done.

On balance, the concept of privatization of genome products is working very poorly for the research community and public. They download huge government data sets, add sequences of their own to them, and repackage with enormous costs of access that will have to resurface many times over in the cost of medicines. What is the purpose of annotating the whole human genome but only letting large drug companies look at it? The current arrangement is extremely divisive with few if any of the scientific players having access to the same information; the quality of that information having never been openly evaluated. The underlying biology is secondary. It is extremely wasteful of effort not to pool all the separate EST collections. In summary, unless there is open data release, there is very little that is useable or relevent to the mainstream academic scientist who will have to rely on duplicative government initiatives.

Late onset prion mutations: E196K, V203I, E211Q

Human Mutation [Mutation in Brief section] #203 (2000): free pdf
Katell Peoc'h, Philippe Manivet, Patrice Beaudry, Françoise Attane, *** Jean-Louis Laplanche
"Inherited prion diseases are characterized by mutations in the PRNP gene encoding the prion protein (PrP). As the other sporadic or infectious prion disease forms, they are almost all characterized by the accumulation in the brain of an abnormal misfolded form of the patientıs PrP. Brain extracts can often transmit the disease once inoculated in a recipient animal. Inherited prion diseases with Creutzfeldt-Jakob disease (CJD) phenotype are autosomal forms, although sporadic cases have been reported.

We report three novel mutations of the PRNP gene in unrelated patients with clinical and histopathologic features of CJD. The three mutations were missense: codon 635 G to A (E196K), 656 G to A (V203I) and 680 G to C (E211Q).

Familial history of neurologic disorders was evidenced for patients carrying the E196K and E211Q mutations. E196K would be predicted to have more severe effects on protein stability than V203I and E211Q. These mutations expand the spectrum of mutations in PRNP and reduce the proportion of CJD patients in whom genetic alterations have not been found.

Creutzfeldt-Jakob disease was clinically diagnosed in three patients according to the criteria recently established by the EU collaborative group on CJD (1998). The 14-3-3 CJD marker in the cerebrospinal fluid was positive. The diagnosis of CJD was confirmed in the three cases by brain examination at autopsy and, in addition, by the detection of the abnormal PK-resistant form of PrP in the brain extracts from patients 1 and 2.

Patient 1 (E196K M129M) was a French 69-year-old woman who developed in three to four weeks behavioural abnormalities. She became anorexic and mute with emotional lability and inappropriate giggling. The patient was then bedridden and developed choreo-athetoid movements. No periodic activity was recorded at EEG. The patient died 12 months after the first recorded symptoms. Her brother had died aged 64 after a three-month history of rapidly progressive dementia with motor impairment that apparently began 10 months earlier with gradually progressive mental deterioration without neurological signs. EEG presented a diffuse slowing one-month before death. The diagnosis of Alzheimer disease was considered. No autopsy was performed. The probandıs mother died aged 77 and was reported to be affected by the same disease.

Patient 2 (V203I M129M c421g) was a 69-year-old male Italian patient initially hospitalized to evaluate a 10- day history of monocular diplopia and dizziness. Fourteen days later, the patient presented a sudden onset of confusion and dramatic visual hallucinations. Examination showed tremor, cerebellar gait, coordination deficit, multidirectional nystagmus and myoclonus. EEG demonstrated a periodic activity. The patient died 25 days after admission. There was no family history of neurologic disorders. His father and mother died at age 60 and 62 from cardiac failure and traffic accident, respectively.

Patient 3 (E211Q M129M c421g) was a French 69-year-old man who initially presented with behavioural abnormalities and a cerebellar ataxia preceding dementia with myoclonus. A periodic activity was recorded at EEG. The patient died 6 months after the onset. A 64-year-old brother developed at about the same time a rapidly progressive dementia with cerebellar signs, gait instability and prominent pyramidal syndrome and died 32 months later. A sister had died at age 75 after a one-year history of dementia with cerebellar syndrome and abnormal movements. Unsteady gait and tremor were reported in a second brother before he died suddenly at age 47. Their mother and father died at age 64 and 86, respectively, cause unknown.

These three novel mutations were neither found in 120 Caucasian healthy controls nor in 345 patients with sporadic or familial prion diseases and 224 patients with various neurological disorders (1378 chromosomes [all authors are French so probably a French patient set -- webmaster])."

Comment (webmaster): This excellent account from the Laplanche group in France was published in rapid-appearance format (11 Jan 00 submission!) with barrier-free access to the pdf. The account is quite modern in that residue 3D exposure, conserved salt bridges, and species conservation were considered in evaluating the mutations. This integrated approach is especially important in evaluating late-onset mutations where it is very difficult (even impossible) to assemble enough of a kindred to determine a familial pattern.

Note also that complete haplotypes were reported for all patients, including even details like the silent G124G (0.5% polymorphism) on the same chromosome as 211Q (which of course is ideal for kindred tracking). Haplotypes will become an increasingly complex issue if cis effects (chimeric splicing) couple prion alleles to doppel alleles on the same chromosome, as they do in mice.

The testable kindreds were quite small but it appears that E196K and E211Q are very likely causative. Both are strong evolutionary invariants; In birds, codon 211 is still glu in 8 species, codon 203-205 are deleted, and codon 196 is in a non-alignable region. In doppel, 211 is arguably conserved (ELCSIK***). A 3D model of doppel might support conservation of internal salt bridges even where amino acid identity is lost.

V203I is less certain: the Italian kindred was small; the patient had no family history of neurologic disorders and very late onset (69); CpG sites are mutational hotspots. The mutation may be weakly causative. However, codon 203 is methionine in canids; val in primates, most rodents, and early-branching artiodactyls; and isoleucine in most ruminants and cat. In other words, it is quite conservative within old taxonomic lineages but admits non-pathogenic isoleucine. This situation is similar to N171S.

The distribution of pathogenic mutations, as noted in the article, continues a puzzling strong distal clustering (only A117V is known in the core amyloid region). The authors, like many others, find thermodynamic instability per se implausible as a general explanation. It is hard to envision a distal charge-retaining substitution of glutamic acid to lysine at 196 impacting membrane topology at the ER. Some mutations are surface but others are interior and could not interact with other proteins prior to significant structural denaturation. Indeed, several distinct mechanisms may be operarative (as further suggested by repeat region anomalies).

The 3 mutations, along with an older silent mutation H177H [Ripoll (1993) Neurology 43:1934 ], two of which are CpG, have now been added to the central collection of point mutations. There are currently 21 missense and 2 stop mutations causative for CJD, 4 neutral/modulatory alleles, and 8 silent mutations in the coding region of the prion gene.

Three conclusions can be drawn from the remarkable growth of the mutation set in the last 12 months [11 additions from Windl, Finckh, and Laplance groups]: the set of sites is far from saturation, methods of detection are greatly improved (eg, earlier surveys missed mutations), and sporadic CJD has been significantly over-estimated.

Turning this around, a saturated set of mutations (meaning large high quality surveys find only further kindreds of known sites) would suggest certain CpG mutations are not pathogenic. These CpG sites were last considered in January 1998. At that time, 6 of 21 CpG sites had been seen in the patient set. Since then V203I, D202N, T188K, T188R, and H177H have been observed, giving total saturation of all distal CpG sites in human prion. R208C is the only missing prediction; R208H was found in 1996. An unpaired cysteine could lead to cross-linking of monomers in the oxidizing cell-surface environment.

Despite this, 6 opportunities in the first alpha helix and loop have not manifested themselves: R148C/H, R151C/H, and R156C/H. This is the CpG restatement of observed distal clustering. A fairly strong statistical case could be made that these hotspot mutations must occur but are not pathogenic. [What are the odds that 5/5 newly observed hotspot mutations would all be distally located by chance when an equal opportunity for proximal location existed, (1/2)exp5 or 3%.] A similarly-sized set of non-observed CpG sites are seen 5' to the repeat region; again one might argue that changes here have occurred but are not pathogenic (though asymptomatic individuals might yet be infections).

An alternative explanation is that these mutations do in fact occur but are lethal in utero or have such early onset means that the founder dies off without children. Since there would be no history of dementia in a founder born, say, in 1985, cases would unlikely to be screened. (Indeed, the symptoms might not be neurological.) In this view, CpG hotspots in the pre-repeat and first helix are a mix of super-causative and non-causative, never falling into the detectable range for age of onset (20-70). As mutational sites continue to accrue in the distal clustering pattern, this view becomes increasingly untenable statistically.

There is not any codon known with consistently early onset of familial CJD. More typically, it is 35-65 years (though Laplanche reported earlier a kindred in its fifth generation with onset 21-34, average 28, a 8x repeat). So is prion disease an inherently slow process, or one that takes place fairly rapidly but only in aging brain, or does earler onset go unrecognized?

Similar issues are raised by stop codons Y145x and Q160x. The genetic code is structured so that only an arginine codon, CGA, can give rise to a stop codon, TGA, via a CpG hotspot (irregardless of primary sequence adjacency) using C to T and G to A as standard hotspot changes. Needless to say, this arginine codon is severely depleted in humans (and absent in prion protein). However, 5 of 60 codons that could mutate to a stop codon lie between codon 145 and 160; thus E146x, Y149x, Y150x, E152x, Y157x, Y162x, and Y163x are prime candidates for future causative stop mutations (which differ from simple loss-of-function effect). Y128x and G131x are the only candidates upstream (assuming the amyloid core is required) for minimal mini-prions. Because they are not hotspots, a large familial CJD sample size is needed.

Note finally the lack of mutations that prevent the signal peptide from being cleaved. The nascent peptide would not be transported to the endoplasmic reticulum, possibly leading to overproduction and accumulation. Two adjacent CpG sites exist in conserved arginine and proline codons 3-4 residues from the cut site, MANLGCWMLVLFVATWSDLGLC KKRPKPGGWNTGG. Following a suggestion of JL Laplanche, in silico mutations (RP to CP, HP, RL) can be tested using online splice tools such as PSORT. However, these all have identical McGeoch/von Heijne scores to wildtype. The signal peptide in fact is quite robust to many single amino acid changes.

CpG predictions for doppel are given here; the bovine gene is analyzed elsewhere with homology to human pathogenic mutations taken into consideration. (Note that no known human mutation has early enough onset relative to a bovine lifespan. It is not clear how incubation times for familial TSE should be scaled, or even if they should be scaled, to an animal's life expectancy or metabolic rate. Lab rodent lifespans are 2-3 years, cows 25, humans 75.)

All the surveys to date have been restricted to the coding region despite frequent splice site mutations known in many other diseases and the very convenient length and location of the 5' UTR prion splice junction. It is worth noting that if the splice acceptor of human exon 3 is competing with a downstream doppel splice acceptor (as in mice), less mRNA diverted to doppel implies more prion mRNA, leading possibly to more prion protein. Overproduction could then drive amyloid aggragation by the mass action principle leading to earlier onset of disease (eg, during the lifespan).

Until this situation is resolved, the existence of spontaneous CJD remains in limbo. All CJD may ultimately be genetic in origin, iatrogenic and dietary CJD being interpreted as passaging of an initial mutational case. However, at this time, only 1 case in 7 of sporadic CJD can be allocated. For this reason, the term 'classical' CJD should never be used.

It is very difficult to find and map secondary genes because very subtle differences in genetic background give rise to very large phenotypic differences, as can be seen in a P102L case in monozygotic twins [Hamasaki et al. Lancet 352 1998]. The authors note identical twins are not quite genetically identical due to a dozen or so mutations in each round of cell division since fertilization. The first twin had late onset (58) and very slow progression (8+ years); the second twin may simply be progressing a few percent slower despite the monozygosity.

Molecular analysis of PRNP gene in Polish population and in CJD.

Folia Neuropathol 1999;37(4):277-80
Bratosiewicz J, Kordek R, Kulczycki J, Botts G, Liberski PP
In our study we have examined allelic variation of codon 129 among the Polish population as well as Polish and Dutch CJD cases. The open reading frame of the PrP gene was amplified using the polymerase chain reaction (PCR). PCR product was digested with Nsp I and Mae II endonucleases and separated by 2% agarose gel electrophoresis and, finally, sequenced by the Sanger dideoxy-mediated chain-termination method. To obtain population data we have screened 109 unrelated Polish adults. There were 45% of methionine homozygotes, 16% of valine homozygotes and 39% of heterozygotes. Among Polish CJD cases, 75% were methionine homozygous, 12.5% were valine homozygous and 12.5% were heterozygous, whereas among Dutch CJD cases it was 29% of Met/Met and 71% of Met/Val genotypes.

Characterization, Chromosomal Localization, and the Complete 30-kb DNA Sequence of the Human Jagged2 (JAG2) Gene.

Genomics 2000 Jan 1;63(1):133-138 
Deng Y, Madan A, Banta AB, Friedman C, Trask BJ, Hood L, Li L
The genomic sequence of the human Jagged2 (JAG2) gene, which encodes a ligand for the Notch receptors, was determined. The 30-kb DNA sequence spanning the JAG2 gene contains 26 exons and a putative promoter region. Several potential binding sites for transcription factors, including NF-kappab, E47, E12, E2F, Ets-1, MyoD, and OCT-1, were found in the human JAG2 promoter region. The JAG2 gene was also mapped to the chromosomal region 14q32 using fluorescence in situ hybridization.

Comment (webmaster): The abstract does not make it at all clear that what they found -- apparently a region of extended synteny between chromosome 20p and chr 14q32. Jag1 (Alagille syndrome) is located on chr 20p just centromeric to Snap25 and prion/doppel; the report here established a second gene, Jag2, on chr 14q32 with an identical 26 exon/intron structure and 62% identity at the protein level.

This means that the second Jag gene did not arise, as so many paralogous copies do, by retro-transcription of mRNA or local tandem duplication. Instead it represents 30kb of chromosomal transfer (minimally) or the relic of a tetraploidization event. The question arises immediately as to the syntenic boundaries, that is, does chr 14q32 extend to prion'-doppel'-Snap25'-Jag2 ? (Oddly this is not discussed although one of the authors did the original sequencing of the prion gene a decade ago and was involved in doppel more recently.)

This putative synteny could be pursued by blasting each of the serially-ordered chr20p genes established by GenMap00 against a chr 14 finished database as well as against unfinished sequence (htgs database); very long delays at the NCBI server make this tedious. Aggravatingly, GenBank accepted tens of thousands of htgs entries from genome project labs (notably Wustl and Whitehead Institute) that failed to state which chromosome they originated from, making it necessary to email the lab repeatedly for identifications.

Chromosome 14 is 95 million bp in length; 29 mbp (31.7%) are finished. The genomic sequences here are found at AF111170 (148 kb) and NT_000641. The annotation suggests 3 genes in the contig: transcription factor TFIIIB, followed by an unknown protein, then Jag2.

The prion gene was supposed established as single-copy in the late 1980's by hybridization experiments, later proven unsatisfactory by the discovery of the tandem duplicate doppel. This also cast into doubt experimental claims that no homologue exists in nematode or drosophila (genome available very shortly).

If, and this is a large multiple if, a 65% homologue of prion gene functioned on chr 14, what would be the signficance to CJD? Possibly none, possibly a lot, depending on what features of structure, function, and interaction had been conserved. A chr 14 locus would not have emerged from familial CJD studies; kindreds are found today not by positional mapping but by screening the prion gene in sporadic CJD or assorted neurological ailments. A defect in the hypothetical prion paralogue on chr14 would simply be ascribed to sporadic CJD.

Self-Assembly of Recombinant Prion Protein of 106 Residues

Biochemistry 2000 Mar 14;39(10):2792-2804
Baskakov IV, Aagaard C, Mehlhorn I, Wille H, Groth D, Baldwin MA, Prusiner SB, Cohen FE
The central event in the pathogenesis of prion diseases is a profound conformational change of the prion protein (PrP) from an alpha-helical (PrP(C)) to a beta-sheet-rich isoform (PrP(Sc)). The elucidation of the mechanism of conformational transition has been complicated by the challenge of collecting high-resolution biophysical data on the relatively insoluble aggregation-prone PrP(Sc) isoform.

In an attempt to facilitate the structural analysis of PrP(Sc), a redacted chimeric mouse-hamster PrP of 106 amino acids (MHM2 PrP106) with two deletions (Delta23-88 and Delta141-176) was expressed and purified from Escherichia coli.

PrP106 retains the ability to support PrP(Sc) formation in transgenic mice, implying that it contains all regions of PrP that are necessary for the conformational transition into the pathogenic isoform [Supattapone, S., et al. (1999) Cell 96, 869-878]. Unstructured at low concentrations, recombinant unglycosylated PrP106 (rPrP106) undergoes a concentration-dependent conformational transition to a beta-sheet-rich form. Following the conformational transition, rPrP106 possesses properties similar to those of PrP(Sc)106, such as high beta-sheet content, defined tertiary structure, resistance to limited digestion by proteinase K, and high thermodynamic stability.

In GdnHCl-induced denaturation studies, a single cooperative conformational transition between the unstructured monomer and the assembled beta-oligomer was observed. After proteinase K digestion, the oligomers retain an intact core with unusually high beta-sheet content (>80%). Using mass spectrometry, we discovered that the region of residues 134-215 of rPrP106 is protected from proteinase K digestion and possesses a solvent-independent propensity to adopt a beta-sheet-rich conformation.

In contrast to the PrP(Sc)106 purified from the brains of neurologically impaired animals, multimeric beta-rPrP106 remains soluble, providing opportunities for detailed structural studies.

Comment (webmaster): This paper continues an interesting concept that seeks to fix the relative unsuitability of prion protein among amyloidogenic proteins as a group (which often have normal solubility). Deleting the repeat region seems fairly minor compared to del 141-176, which affects helix 1, core internal packing, and beta strand 2. Perhaps a better choice would have been one of the two stop mutation genes that seem to cause CJD-like diseases, at 145 and 160, and suggest a non-essential role for distal peptide. A repeat deletion here might give rise to very short mini-prion which would simplify further structural studies.

The abstract does not speak to congo red binding to the multimer nor to what limits its degree of aggragation; cross-beta fibrils of course admit indefinite extension. Even a small multimer seems beyond the size range for nmr and an unlikely candidate for xray crystallography, so it is unclear how forthcoming structural studies will produce 3D coordinates.

Mad Cow Home ... Best Links ... Contact Researcher ... Science Index