Prion Disease: February 2000 Science News
Mad Cow Home ... Best Links ... Contact Researcher ... Science Index

Oxford reins in nvCJD tonsil modeller
Fibrillization of human prion 90-231
Optimal toxicity: 106-126
CJD profile in patients homozygous for PRNP E200K.
Disease-finding with GenMap00
A synthetic peptide initiates GSS in transgenic mice
Prion postdoc opening
NEJM 14-3-3 discussion
Current diagnostic criteria for CJD
Monitoring plasma processing steps
Carnoy's fixative improves Prp-res immunohistochemistry
Yeast 'prion' developments

Assessment of the prevalence of nvCJD through testing tonsils and appendices for abnormal prion protein.

Proc R Soc Lond B Biol Sci 2000 Jan 7;267(1438):23-9 
Ghani AC, Donnelly CA, Ferguson NM, Anderson RM
The objective of this study was to determine the age group or groups which will provide the most information on the potential size of the vCJD epidemic in Great Britain via the sampling of tonsil and appendix material to detect the presence of abnormal prion protein (PrP(Sc)).

A subsidiary aim was to determine the degree to which such an anonymous age-stratified testing programme will reduce current uncertainties in the size of the epidemic in future years. A cohort- and time-stratified model was used to generate epidemic scenarios consistent with the observed vCJD case incidence. These scenarios, together with data on the age distribution of tonsillectomies and appendectomies, were used to evaluate the optimal age group and calendar time for undertaking testing and to calculate the range of epidemic sizes consistent with different outcomes.

The analyses suggested that the optimal five-year age group to test is 25-29 years, although a random sample of appendix tissue from all age groups is nearly as informative. A random sample of tonsil tissue from all age groups is less informative, but the information content is improved if sampling is restricted to tissues removed from those over ten years of age.

Based on the assumption that the test is able to detect infection in the last 75% of the incubation period, zero detected infections in an initial random sample of 1000 tissues would suggest that the epidemic will be less than 870,000 cases. If infections are detected, then the model prediction suggests that both relatively small epidemics (800+ cases if one is detected or 8300+ if two are detected) and larger epidemics (21,000+ cases if three or more are detected) are possible.

It was concluded that testing will be most informative if undertaken using appendix tissues or tonsil tissues removed from those over ten years of age. Large epidemics can only be excluded if a small number of infections are detected and the test is able to detect infection early in the incubation period.

Comment (webmaster): While it is not expensive or time-consuming to test tonsils, that effort might be further optimized by the model if anyone were interested. This group has put out an extraordinary, even excessive, theoretical effort over the last years that has been severely limited from the get-go by lack of cooperation from those with good data on needed model parameters.

The senior author, Anderson, has just been suspended from as professor job at Oxford [27 Jan 00 Nature, page 353] supposedly over speculating about how a friend of the new zoology chair got her prestigous epidemiology appointment. Medline shows 16 papers over 11 years that Anderson co-authored with the appointee, who is also a published novelist.

The timing of the disclipinary move -- virtually the same day the abstract appeared on Medline -- conveniently falls just outside the time scope of the BSE Inquiry. The Sunday Times of 27.2.00, according to Dealler's web site. notes that Anderson is a member of SEAC send off work from Oxford University because of complaints from junior staff now appears to have other problems in terms of possible manipulation of funds. He is a member of the Wellcome Trust, a major owner of a research company from Oxford, and has received large amounts of funding from the Wellcome. The Wellcome did not know that he was an owner of the Biomedical and Health Sciences Consortium

Looking at the abstract above and considering dozens of previous incidents of stiffling research in TSE, one wonders what is really going on at Oxford, perhaps reining in of Anderson for pushing too hard for disclosure of the extent of nvCJD. Over 18 months has elapsed since Collinge's group released a very promising tonsil study.

This wasn't Anderson's first offense: getting the BSE data from MAFF a few years back after years of hassle led to a similarly unwelcome estimate of how many undetected cases of mad cow disease went directly into the human food chain. Theorists are less dependent less grant money and have to be contained indirectly through character or corruption charges.

Academic experimentalists, whose research is dependent on a perpetual stream of large grants, avoid "controversial" research or face the consequences -- London Zoo cut off its research chief in 1995 at the behest of MAFF/MRC after too many species were found with BSE. France threatened the same thing last year when zoo primates contradicted public posturing, much as happened in the US over mad mink apparently contracted disease from downer dairy.

The Internet has been a great nuisance, but one domestic web host was silenced by offering an insider position requiring confidentiality, another by a retirement package agreement, and still others by the Official Secrets Act. Britain briefly considered a plan to curtail access to certain URLs based in foreign countries having freedom of information but gave it up because of mirroring and email workarounds. TSE researchers in Britain seldom leak anything because mail, fax, and email transmissions are considered insecure.

Evidently we are in our 13th year of this rubbish:

SURREY KT6 7NF TELEPHONE 01-337 6611 Ext 525

Mr R Hallcock VI Centre TRURO 13 May I987

Dear Roger


Further to our telephone conversation this morning, I am now confirming
that the letter to the Veterinary Record which I cleared earlier in the
week should not be published. I explained to you that this condition had
been discussed by the CVO and the Director of CVL, and because of
possible effects on exports and the political implications it had been
decided that, at this stage, no account should be published. No doubt
there will be an opportunity for your case to be published in due
course.  [No publication by Hallcock ever appeared. -- webmaster]

Yours sincerely

Assistant Chief
Veterinary Officer 
cc: Dr Watson-Mr Pill

Aggregation and fibrillization of the recombinant human prion protein huPrP90-231

Swietnicki W, Morillas M, Chen SG, Gambetti P, Surewicz WK
"According to the "protein-only" hypothesis, the critical step in the pathogenesis of prion diseases is the conformational transition between the normal (PrP(C)) and pathological (PrP(Sc)) isoforms of prion protein. To gain insight into the mechanism of this transition, we have characterized the biophysical properties of the recombinant protein corresponding to residues 90-231 of the human prion protein (huPrP90-231).

Incubation of the protein under acidic conditions (pH 3.6-5) in the presence of 1 M guanidine-HCl resulted in a time-dependent transition from an alpha-helical conformation to a beta-sheet structure and oligomerization of huPrP90-231 into large molecular weight aggregates. No stable monomeric beta-sheet-rich folding intermediate of the protein could be detected in the present experiments. Kinetic analysis of the data indicates that the formation of beta-sheet structure and protein oligomerization likely occur concomitantly. The beta-sheet-rich oligomers were characterized by a markedly increased resistance to proteinase K digestion and a fibrillar morphology (i.e., they had the essential physicochemical properties of PrP(Sc)).

Contrary to previous suggestions, the conversion of the recombinant prion protein into a PrP(Sc)-like form could be accomplished under nonreducing conditions, without the need to disrupt the disulfide bond. Experiments in urea indicate that, in addition to acidic pH, another critical factor controlling the transition of huPrP90-231 to an oligomeric beta-sheet structure is the presence of salt."

Comment (webmaster): The apparent conflict mentioned in the abstract is to a paper from Collinge's group, Science 1999 Mar 19;283(5409):1935-7. It is important to get this straight (this paper has the advantage) because the intra-cellular milieau is reducing and that of a GPI protein oxidizing. An intact disulfide bond greatly constrains conformational change; indeed, that is why they are mainly restricted to proteins in the less homeostatic extra-cellular environment. Doppel even has a second disulfide further stabilizing this loop.

Clearly a lot of residues near an intact disulfide could not plausibly participate in beta sheet. The potential significance (down the road) is that conversion could take now place on the surface of the cell or in the debris of dead cells after exposure to external infectious agent or perhaps along the endocytic pathway, whereas had a reduced disulfide been necessary, the exogenous agent would need to make its way into the reducing interior.

We are seeing considerable consistency across papers with the efficacy of low pH, guanidine-HCl, urea, and sodium chloride. These could be viewed as chaotropes lowering some potential barrier that separates two conformational states. Here, salt and low pH alone just gave an aggragate, not the thioflavine T positive fibril. Small oligomeric intermediates perhaps make some sense in native tetramers such as transthyretin but not in monomeric proteins going on to form very long fibrils involving thousands of units -- what would be the natural stopping points in the highly cooperative assembly of a one-dimensional crystal? Beta structure, by its nature, is non-local; there are no proteins with a single beta strand. Note however in beta barrels -- but not in cross-beta -- that beta structure is self-limiting.

Thus the lack of observed stable intermediates in this paper make sense, offering no support to long-lived seeds as a separate entity. The 30 second dead time for negative ellipticity kinetics reported does not leave much wiggle room in conjunction with millisecond unfolding reported elsewhere. On the other hand, it is not immediate how to fully reconcile this with the many papers over the years interpreting various data as small oligomers, though the authors do this for their own recent JBC paper [below].

The pH values used are on the low side, dramatically slowing by pH 4.4. Nothing happened in two weeks at neutral pH. It is hard to fathom the significance of these time scales in a disease where even inherited forms can have onset in the fifth decade or later. The reaction was strongly dependent on protein concentration (supporting oligomerization) and not reversible upon neutralization. Protease K resistance accompanies the change. Molecular weight were quickly in the 400-2000k range, eventually exceeding the column sizing capability of 40,000k.

The authors here have gotten the protein under such consistently good control that one supposes they are looking at possible infectivity of their beta-sheet-rich oligomers. Recombinant 90-231 is of good physiological length and does not seem to suffer from its lack of covalent modifications. Would infectivity be a milestone or we would still wonder what about the conformation of the actual in vivo infectious fragment? Perhaps it would be more interesting to define residues participating in the the beta structure better via internal cross linking of side chains..

Membrane environment alters the conformational structure of the recombinant human prion protein.

J Biol Chem 1999 Dec 24;274(52):36859-65 
Morillas M, Swietnicki W, Gambetti P, Surewicz WK
. The prion protein (PrP) in a living cell is associated with cellular membranes. However, all previous biophysical studies with the recombinant prion protein have been performed in an aqueous solution. To determine the effect of a membrane environment on the conformational structure of PrP, we studied the interaction of the recombinant human prion protein with model lipid membranes. The protein was found to bind to acidic lipid-containing membrane vesicles. This interaction is pH-dependent and becomes particularly strong under acidic conditions. Spectroscopic data show that membrane binding of PrP results in a significant ordering of the N-terminal part of the molecule. The folded C-terminal domain, on the other hand, becomes destabilized upon binding to the membrane surface, especially at low pH. Overall, these results show that the conformational structure and stability of the recombinant human PrP in a membrane environment are substantially different from those of the free protein in solution. These observations have important implications for understanding the mechanism of the conversion between the normal (PrP(C)) and pathogenic (PrP(Sc)) forms of prion protein.
     Swietnicki W, et al.  
       Familial mutations and the thermodynamic stability of the recombinant human prion protein. 
       J Biol Chem. 1998 Nov 20;273(47):31048-52. 

     Swietnicki W, et al. 
       pH-dependent stability and conformation of the recombinant human prion protein PrP(90-231). 
       J Biol Chem. 1997 Oct 31;272(44):27517-20. 

Optimal toxicity: 106-126

Brown DR
Mol Cell Neurosci 2000 Jan;15(1):66-78
"In prion disease neurodegeneration requires deposition of the abnormal isoform of the prion protein (PrP(Sc)) within nervous tissue. In vitro PrP(Sc) has neurotoxicity that can be mimicked by peptides based on part of its sequence. In this investigation the region of the protein required for maximal neurotoxicity was precisely determined. The optimal neurotoxic peptide was found to contain amino acids 112-126 of the human sequence.

The sequence AGAAAAGA was found to be necessary but not sufficient for a neurotoxic effect. The AGAAAAGA peptide blocked the toxicity of PrP106-126, suggesting that this sequence is necessary for the interaction of PrP106-126 with neurons. These results suggest that targeting or use of the AGAAAAGA peptide may represent a therapeutic opportunity for controlling prion disease."

Altered toxicity of the prion protein peptide PrP106-126 carrying the Ala117-->Val mutation.

Biochem J 2000 Mar 15;346(Pt 3):785-791
Brown DR 
The inherited prion diseases such as Gerstmann-Straussler-Scheinker syndrome (GSS) are linked to point mutations in the gene coding for the cellular isoform of the prion protein (PrP(C)). One particular point mutation A117V (Ala(117)-->Val) is linked to a variable pathology that usually includes deposition of neurofibrillary tangles. A prion protein peptide carrying this point mutation [PrP106-126(117V)] was generated and compared with a peptide based on the normal human sequence [PrP106-126(117A)]. The inclusion of this point mutation increased the toxicity of PrP106-126 which could be linked to an increased beta-sheet content. An assay of microtubule formation in the presence of tau indicated that PrP106-126 decreased the rate of microtubule formation that could be related to the displacement of tau. PrP106-126 carrying the 117 mutation was more efficient at inhibiting microtubule formation. These results suggest a possible mechanism of toxicity for protein carrying this mutation via destabilization of the cytoskeleton and deposition of tau in filaments, as observed in GSS.

Comment (webmaster): This would be basic + bland again: KTNMKHMAGAAAAGAVVG. There have been quite a few similar studies of this region; it would be nice if someone had time to spread out the full texts on a table and pull it together. Three issues are conflated, the first the propensity of this 'stop transfer effector ' region to hang up in the endoplasmic reticulum resulting in wrong membrane topology, the second its ability to form the core of the cross-beta fibril, and the third a possible role in competing for still-hypothetical normal prion receptor.

The key point made here is that the subsequence AGAAAAGA was 'necessary but not sufficient' for toxiticity. Supposing, as recommended by William of Occam, that it is amyloid fibril itself that is toxic, this peptide could block the fibril extension site or block at the incoming monomer, possibly then tilting the balance in favor of the processes of fibril turnover. Prp-res deposits in diseased brain could be dated using racemation methods of Jeffrey Bada that work on this time scale.

The paper makes some interesting improvements over controls that simply scramble the overall amino acid composition, using terminally scrambled KTNMKHMGAAVLGGAVAGAGA, KTNMKHMAVGAGGGLVGGAAA, and KTNMKHMGAGVAGAGAVAG, as well as a methyalated histidine variant (with no effect). The combinatorics of 20 amino acids are such that very little of sequence space can actually be explored. The three scrambled peptides, despite being intended as controls, ended up being more toxic than wildtype fragment. Brown correlates this with the 310 million year absolute invariance of this domain, suggesting it is trapped in a cul de sac where all variation leads to lethal dominant negatives. But where did it come from?: no such peptide can be found in the complete yeast, nematode, or drosophila genomes though searches are difficult because of abundant glycine-rich proteins.

AGAAAAGA is not viewed as a beta breaker in the sense of similar peptides in Alzheimer but rather as exerting its effect through to binding of Prp-c or another protein. AGAAAAGA was found the minimum essential for fibril formation, Prp-sc specific toxity, and binding partners. The paper makes a good contribution in its experimental resolution of these issues.

CJD profile in patients homozygous for PRNP E200K.

Ann Neurol 2000 Feb;47(2):257-60
Simon ES, Kahana E, Chapman J, Treves TA, Gabizon R, Rosenmann H, Zilber N, Korczyn AD
"We identified 70 Creutzfeldt-Jakob disease patients with the previously described E200K mutation in the prion protein gene. The purpose of this study was to define the clinical features of E200K homozygous patients (n = 5), compared with heterozygotes. We found a statistically significant younger age at disease onset for the homozygous patients, although the average age at onset in this group was still in midlife. Disease features were not statistically different in the two groups. Possible explanations are discussed."

Comment (webmaster): Familial CJD is autosomal dominant with fairly high penetrance, meaning only one mutant chromosome causes disease in most carriers. Homozygotes are thus detected in patient screens only as the square of a very small number, except for E200K which involves intermarriage.

The expectation of the rogue prion fibril model is earlier onset of disease, attributable to protein dosage. The bad protein may or may not be able to effectively recruit normal prion protein from the other allele to the growing fibril or serve as seed for what become normal allele fibrils. If it cannot (or if it can but this caps or otherwise interferes with fibril extension), the effect of homozygosity on onset might be quite dramatic, doubling the amount of substrate and driving the equilibrium in the direction of fibril.

The paper above reports the expected effect in EE200KK but not a dramatic one. Taking onset as a rough proxy for fibril substrate concentration, this suggests that fibrils in heterozygotes are mixes of the respective monomers (or that 200K can seed a E200 fibril), consistent with similar symptoms of the same severity.

Another very interesting potential implication of the paper is that EE200KK does not leave the cell without normal prion function. As a toxic-gain-of-function disorder with no assay for normal function, it has never been possible to decide whether mutations are neutral with respect to normal prion function, a situation compounded by uncertainties involving doppel compensation and over-compensation (but note doppel lacks the copper binding domain). The simplest view of the data is that E200K enhances predisposition to fibril formation wilthout affecting normal function to a significant degree.

Various other double allele combinations are known:

-- Pocchiari found a del R2 trans to a V210I central Italian family 
-- Perry found a del R2 in an Alabama FAD, parent R2/R34 
-- Yamada found a del R2 in a Japanese family transto P105L 
-- Masullo found a homozygous del R2 in an Italian, adopted 
-- Vnecak-Jones found a del R34 mid-Tennessee with E200K 
-- Laplanche saw a del R34 in a Tunisian E200K family 
-- Bosque saw a del R34 in a Tennessee family with D178N 
-- Cervenakova saw a del R34/R3g34 in an African-Amer family 
-- Ghetti writes in 1996, 'no homozygous F198S has yet been seen in the Indiana kindred' 
-- Hitoshi found a V180I M232R double mutation in an 84 year old Japanese man 
Tranchant in Rev Neurol (Paris) 1991;147(4):274-8 found a double mutant A117V in a family of Alsatian origin in all patients analyzed and in 10 healthy family members leading to loss of the restriction site PvuII and to the replacement of an alanine by a valine, a litle confusing since GCA to GTA is the usual A117V, GCA to GCG the usual silent PvuII loss. Perhaps what is meant is simply A117V in the background of the PvuII polymorphism, ie the family is heterozygous for ala/val at codon 117. The full text may give the age of the 10 healthy family members. This mutation affects center of the palindromic region.

Gabizon observes in Am J Hum Genet 1993 Oct;53(4):828-35 that 'The identification of three Libyan Jews homozygous for the Lys200Arg mutation suggests frequent intrafamilial marriages, a custom documented by genealogical investigations." The clinical course was mentioned in passing in one patient but not compared in detail to heterozygotes.

Hsiao observed in N Engl J Med 1991 Apr 18;324(16):1091-7 that 'One patient was homozygous for the lysine mutation [E200K], and her clinical course did not differ from that of the patients heterozygous for the mutation.... The similarity of the clinical courses of the patient homozygous for this mutation and the patients heterozygous for it argues that familial Creutzfeldt-Jakob disease is a true dominant disorder. ' This case is probably the same as one mentioned by Gabizon.

Disease-finding with GenMap00

Mon, 7 Feb 2000 webmaster research
A major update has been made to GenMap00, which now orders all 329 GeneMap99 markers on human chromosome 20 and in fact assembles the sequence of the whole p arm, using blastn(htgs) and blastn(hum gen) with Unigene probes from GeneMap99. This involved quick annotations of 25Mbp of dna, both finished and unassembled. There are 74 characterized genes in this region and roughly equal numbers of novel genes and pseudogenes.

The older markers are important to locate because historically a lot of mapping of inherited disease took place relative to these markers. That research has stalled out because GeneMap98 nd GeneMap99 were rife with errors, leaving frustrated clinicians with too many genes to analyze in patient families.

Humans will not have very many genes after all, perhaps twice as many as a 1 millimeter nematode. That is completely clear from each of the chromosomes being annotated -- the average gene density does not exceed 10-12 per million bp. And many genes are clustered into closely related families: a very common pattern is for two copies of a retrotransposon to mediate unequal recombination, leading first to a tandom duplicate and then to higher order clusters of genes and pseudogenes, sometimes numbering in the hundreds.

The prion-doppel pair is a typical tandem duplicate. It is flanked by other genes expressed in the brain: ADRA1A adrenergic alpha 1A receptor upstream and and KIAA0168 and PCNA proliferating antigen to the centromere. Chromsome 20p also contains a larger tandem replicate feature involving cystatin genes nearer the centromere and a very complex cluster of 10-12 SIRPS alpha and beta genes in 20p13.

It does not seem likely that there will be additional copies of prion or doppel genes within chromosome 20 itself. With the whole genome half-sequenced, none show up on other chromosomes either, despite two earlier tetraploidization events in vertebrates. [It is feasible to locate syntenic regions in other human chromosomes based on the flanking genes whether or not prion-doppel themselves are sequenced yet or otherwise detectable.] Recalling the difficulties in recognizing doppel, highly diverged or seldom transcribed copies could be easily missed. Prion pseudogenes originating from retrotransposed mRNA are likely elsewhere in the genome, but given the long length of 3'UTR in this gene and that insertion begins at the poly A tail, these may never reach the coding region.

Mouse chromosome 2 has long been known to be the counterpart of human chr 20p12. GeneMap00 identifies the known orthologues, so it becomes possible for the first time to look for inversions. The TCF15 gene and a few others are evidently out of order but mapping error, rather than inversion, could account for this. Rodent genomes are very scrambled compared to the ancestral mammal and enough is known now of other species to assign inversions to the lineage in which it occurred (ie it would probably have been in rodents).

GenMap00 now contains a column with the mapping markers of various diseases in search of a gene as well as a mouse synteny column. Some chromosomes can now be completely assembled but no one is telling disease researchers, many of whom simply need a list of genes to screen in their patient set.

OMIM and GenBank both got started in the 'pre-computer' age and have not transitioned to the web as well as they might. Neither thought of itself as a database nor of the kinds of searches people would need to make later. Consequently both have terrible data structures.

For example with GenBank, try to extract the fraction of annotated genes have, like prion and doppel, a single open reading frame (no introns). Or what is the length distribution of 3' UTR (so critical to EST read length)? How common are processed pseudogenes? How many splice junctions are not the canonical GT|AG?

These are routine questions to ask of such a database to put a gene like prion or doppel into perspective -- but GenBank is not structured so it can respond. Some things can be extracted by dumb luck, eg the average size of a mammalian exon (by pulling out all entry lines containing the words 'mRNA' and 'join'). Otherwise, GenBank annotations are going to be very expensive to fix -- every day more entries pour in in the same old bad format.

OMIM has a primitive fulltext search without boolean capabilities but the whole file is only 24 megs (less that so it can be downloaded and indexed locally for searching. Even still, you cannot request, say, all amyloidoses where a protein sequence has been determined.

Something truly amazing that you cannot extract from OMIM except by opening individual entries: a list of mapped diseases for which the gene is not yet identified. Ironically, this is the whole point of both OMIM and the human genome project. A visitor should be able to extract this information as a quick and natural database query but cannot.

What one can do is go to OMIM's table of genes by chromosomal order, then open each of the 1793 records that has an entry in column 8 (disorder) and see if there is a DNA link. If not, then the disorder has been mapped fairly well but no one has been able to make the connection to the associated gene. (Except that crazily OMIM includes links to genes were later _excluded_.) The OMIM text usually gives the nearest STS markers and their lod scores. OMIM does make the Gene and Morbid maps available as clean flatfile databases.

How tedious is this (rather, why hasn't anyone done this)? Quite tedious -- one soon needs access to full texts in a thousand journals. And some diseases look quite easy from the genome annotation perspective, eg, pseudohypoparathyroidism type IB, yet are a mess: "...Two distinct genes, the Gs gene and the "PHP-Ib gene," are involved independently in the regulation of mineral ion homeostasis. However, these two genes would have to reside in the same or in adjacent, paternally imprinted regions and could be in close enough physical proximity to allow the sharing of regulatory elements and/or coding nucleotide sequences, as observed with other imprinted genes."

This year is the Golden Age for finding monogenic disease genes, truly the end of an era that began in 1949 with Linus Pauling and sickle cell. In 2002, DNA from everyone in the kindred will simply be sent off for micro-array microsatellite mapping followed by megabase contig sequencing. The computer will identify the errant gene within a few days. (Multi-factorial diseases are a different story.) Of course we still will not know the normal function nor have gene therapy. Note however that a defective adenovirus is a universal therapy just as a congo red analogue could be a universal amyloidosis therapy.

Meanwhile, a lot of research teams have worked long and hard with traditional methods and gotten frustratingly close to many disease genes. Their clinics are filled with identified family members in the pipeline who will inevitably come down with the disorder. Waiting until 2002 is not an option.

It is very eye-opening to learn that many disease researchers have no time to follow what is going on with genome sequencing projects (and have near-zero ability to gene-find in raw DNA). For example, the congenital hereditary endothelial dystrophy of cornea community apparently did not know about the corn1 syntenic mouse gene (instant model) and the neurodegeneration of brain with iron accumulation labs did not know about the chr 20p ferritin gene.

Actually it is far more efficient to have them stay focused on getting sample permissions from the kindred, doing pathological characterization, and mapping microsatellites, while other people assemble the chromosomes, place the markers, and annotate genome-wide. Sort of like laying track from both ends at the time the trans-continental railroad was being built, getting the gene corresponds to driving the golden spike.

One person could actually annotate the whole genome in a year. At 3.6 billion base pairs, this means looking at 69 million bp a week or 10 million bp a day for a year. That could be done (just barely) but not in any real depth, plus at the end of the year it would all have to be revisited. Further, it does not focus the energy where it is needed: after all, the ratio of mapped diseases (lacking genes) to mapped genes is only 1:20.

So it would be far better just to pull out all mapped diseases with their critical map regions from OMIM and Medline into a database and just annotate those, exactly as was done in GenMap00 for the short arm of chromosome 20p. In fact, a person could rate these in terms of clue quality from the phenotype and high lod score and start on the easy ones. (Many diseases do not have big labs dedicated to them and so are not necessarily hard to match to a gene, if a short list of local genes is available.)

Below, 10 neighboring genes on each side of prion doppel are checked to see whether they had counterparts in zebrafish and if so, how high the percent protein homology was. The average was 57% at the protein level, which is easily detectable. This is done by tblastn (zfish est) with human protein query. While almost all had fair to excellent length matches, few had more than 1-2 good ests, ie, the zebrafish EST collection is still very small and not saturated. More matches would likely have been obtained using 3' UTRs except that these are not adequately conserved.

Human prion, human doppel, and avian prion do not elicit any sort of match at all. Doppel barely generates enough transcript to be detectable in human ESTs; mammalian prion has a 1210 bp 3' UTR.

Zebrafish are the key to understanding normal prion function and the relationship of prion to doppel. One really has to wonder what TSE researchers were thinking of, to look for orthologues in drosophila and nematode but not in zebrafish (which is far easier, far more certain, and far more closely related). As noted earlier on this site, the zebrafish prion gene is very likely adjacent to KIAA0168 on linkage group 17 and 20. KIAA0168 is well-represented in zebrafish ESTs, allowing the prion-containing clone to be isolated.

SNAP25	Hs.84389	92	AW305507
PCNA	Hs.78996	91	AW233231
BMP2	Hs.73853	88	AI878481
CDS2	Hs.24812	79	AA495116
JAG1	Hs.91143	72	AI878310
11L	Hs.30524	63	AW280736
KIAA0952	Hs.7935	63	AI883008
HAO1	Hs.193640	51	AI617812
PLCB1	Hs.41143	50	AW280848
FTLL1	Hs.111334	49	AW184684
ADRA1A	Hs.557	48	AI461341
PLCB4	Hs.32539	47	AI588627
STK	Hs.21864	43	AW076957
CHGB	Hs.2281	38
PCSK2	Hs.93164	34	AW174525
SSTR4	Hs.226015	34	AI959485
THBD	Hs.2030	33	AI588519
CENPB	Hs.85004	28	AI722594
PRND	missing	-	-
PRNP	Hs.74621	-	-

A synthetic peptide initiates GSS in transgenic mice.

J Mol Biol 2000 Jan 28;295(4):997-1007
Kaneko K, Ball HL, Wille H, Zhang H, ... Baldwin MA, Cohen FE   
The molecular basis of the infectious, inherited and sporadic forms of prion diseases is best explained by a conformationally dimorphic protein that can exist in distinct normal and disease-causing isoforms. We identified a 55-residue peptide of a mutant prion protein that can be refolded into at least two distinct conformations. When inoculated intracerebrally into the appropriate transgenic mouse host, 20 of 20 mice receiving the beta-form of this peptide developed signs of central nervous system dysfunction at approximately 360 days, with neurohistologic changes that are pathognomonic of Gerstmann-Straussler-Scheinker disease. By contrast, eight of eight mice receiving a non-beta-form of the peptide failed to develop any neuropathologic changes more than 600 days after the peptide injections. We conclude that a chemically synthesized peptide refolded into the appropriate conformation can accelerate or possibly initiate prion disease.

Comment (webmaster): It is always hard to tell with this group whether a new or old result is being offered. These experiments by their nature have 2-3 year durations and other data has seen up to 4 years of publication delay. This one seems addressed to long-forgetten issues raised by opponents of protein-only.

Surprisingly it doesn't say anything in the abstract about sequential transmission from the mice affected in the first round, even though 240 days elapsed between the onset of symptoms and termination of control mice at 600 days. 'Pathognomic for GSS' means decisively GSS but what exactly could this mean in a mouse? Were there prion deposits and if so of what peptide? -- the lab certainly has a sufficient panel of antibodies and mass spectrometry tools at hand to determine this.

The beta form -- which doesn't seem to have a supporting structural determination despite its small size -- might simply be more stable to proteolysis and not be related to natural seed conformation; indeed in vivo peptides recovered from amyloid have always been longer. The fold might not be accessible to longer peptide because the latter can access the natural domain structure. Did it retain its fold after injection (despite chaperones) or was this simply inferred?

Diagnostic criteria for CJD

Listserve Sat, 29 Jan 2000 Univ.-Prof.Dr.Herbert Budka
According to Kretzschmar, H.A. et al .[Archives of Neurology 1996; 53(N9): 913-20] and and Otto, M et al. [British Medical Journal 1998 316(7131): 577-82]:
"Patients are classed as probably having Creutzfeldt-Jakob disease if they had rapidly progressive dementia of less than 2 years' duration and periodic sharp wave complexes on electroencephalography. They also have to have any two of the following symptoms: myoclonus; visual or cerebellar symptoms, or both; pyramidal or extrapyramidal signs, or both; or akinetic mutism.

Patients who fulfill the criteria for probably having the disease but did not have any abnormalities on electroencephalography are classed as possibly having the disease. Patients who do not fulfil the criteria for either possibly or probably having the disease are classed as having other diseases. Patients who had an identified pathological isoform of the prion protein in brain tissue on immunohistochemical analysis are classed as definitely having the disease."

Comment (H Budka, U. Vienna): But note that many vCJD cases as well as about 5% of the other CJD cases (Brown, al, Neurology 1984 Sep; 16(3): 295-304) do not fullfill this criteria.

These criteria are valid only for sporadic CJD. According to the criteria which we use in the European Surveillance study and which are also used by the WHO, there is NO time limit (shorter than 2 years of duration) for probable CJD, but ONLY for possible CJD. Moreover, the surveillance criteria for probable CJD were amended last year to add a positive 14-3-3 protein CSF test (only in this case: with a duration less than 2 years) as an alternative to the typical EEG.

Below are the surveillance criteria valid at present:

Diagnosis of Creutzfeldt-Jakob disease according to the EC Surveillance Group of Creutzfeldt-Jakob Disease in Europe

Sporadic CJD
	Definite CJD
		Neuropathological confirmation and/or
		Immunochemical confirmation - PrP positive (Western blot) and/or
		Presence of scrapie-associated fibrils (SAF)
	Probable CJD
		Progressive dementia
		Typical EEG and/or positive 14-3-3 protein CSF test (with duration less than 2 years)
		At least two out of four clinical features listed:
			Visual or cerebellar disturbance
			Pyramidal / extrapyramidal dysfunction
			Akinetic mutism
	Possible CJD
		Progressive dementia
		No EEG or atypical EEG
		At least two out of four clinical features listed:
			Visual or cerebellar disturbance
			Pyramidal / extrapyramidal dysfunction
			Akinetic mutism 
		Duration of less than two years
It is well known that these criteria do not cover all CJD. In a retrospective study of neuropathologically confirmed CJD, at least 14% [Hainfellner JA et al. J Neurol Neurosurg Psychiat 61: 139-142 (1996)] or, more recently, 19% [Radbauer C et al. Wien Med Wochenschr 148: 101-106 (1998)] did not fulfill clinical criteria for probable or possible CJD. This is why it is of utmost importance to have an autopsy in every suspect case. Of course neuropathological diagnosis nowadays includes immunocytochemistry.

Is there regional incidence of CJD? In my opinion this can be simply a consequence of small numbers. Epidemiology in Spain has caused confusion over the numbers of "suspect" cases. It makes little sense to use such figures or those of "referrals" to reference centres. Criteria for this vary greatly. Please note that the usual surveillance figures give only numbers for definite and probable cases.

The increase from 1993 in Spain is easily explained by improved surveillance, as was seen in many other countries. A regional difference might just reflect a consequence of small numbers or varying quality of surveillance, as we suspect in Austria where, in the last 6 years, the mortality of diagnosed cases was 3 times higher in Vienna and surrounding Lower Austria than in the rest of the country [Radbauer C et al. Wien Med Wochenschr 148: 101-106 (1998)].

Prion postdoc opening

01 Feb 00 webmaster
There is a prion research postdoc position available at the Sanders-Brown Center on Aging University of Kentucky in Lexington, KY in the Glenn Telling group, tel: (606) 257-1412; fax: (606) 323-5510. Some further backgroundis given at the lab web site. The announcement reads:

An opening for a postdoctoral position in prion research.

"The major goal of our group is to decipher the molecular and cellular mechanisms of prion pathogenesis using transgenic, cell culture and eukaryotic viral expression systems. These studies are not only expected to help devise effective therapies for these devastating conditions but should also continue to cast considerable insight into pathogenic mechanisms of more common CNS degenerative disorders including Alzheimer's disease, amyotrophic lateral sclerosis and Parkinson's disease.

Interested candidates should have a recent Ph.D. with successful research experience in cellular and/or molecular biology and must have U.S. citizenship or permanent residency status. Send curriculum vitae, a brief description of research experience, and names and telephone numbers of three references."

Glenn Telling
Assistant Professor
Department of Microbiology and Immunology
Sanders-Brown Center on Aging
University of Kentucky
Lexington, KY 40536-0230

NEJM 14-3-3 discussion

January 27, 2000 The New England Journal of Medicine Vol. 342, No. 4
CASE 28-1999: Creutzfeldt-jakob Disease

Martin Zeidler and Inga Zerr write that in Dr. Shinobu's discussion of Creutzfeldt-Jakob disease in the September 16 Case Records, (1) he implies that cerebrospinal fluid markers have little diagnostic value.

In fact, a positive test for 14-3-3 protein in cerebrospinal fluid has now been shown to be an accurate predictor of the disease. Traditionally, the most useful noninvasive study for the diagnosis of sporadic Creutzfeldt-Jakob disease has been electroencephalography, with a characteristic tracing of periodic sharp-wave complexes reported to have a sensitivity of 67 percent and a specificity of 86 percent.

However, the test for the 14-3-3 protein in cerebrospinal fluid has a sensitivity of 92 to 96 percent and a specificity of 93 to 100 percent. (2,3,4) The accuracy of this test led the World Health Organization and the European Union's Biomed 2 surveillance program for Creutzfeldt-Jakob disease to revise their clinical criteria for the diagnosis of sporadic Creutzfeldt-Jakob disease.

A positive cerebrospinal fluid test for the protein 14-3-3 is now given a diagnostic weight similar to that of characteristic findings on an electroencephalogram (Table 1). (5) The 14-3-3 protein can be detected in the cerebrospinal fluid in other conditions, (2,3,4) but they are usually easy to distinguish from Creutzfeldt-Jakob disease.

The protein has been shown to be stable at room temperature for prolonged periods, and specimens can therefore be sent by mail to special centers for testing; analysis can be completed in a few days. Many laboratories throughout the world are now routinely testing cerebrospinal fluid specimens from patients suspected of having Creutzfeldt-Jakob disease.


1. Case Records of the Massachusetts General Hospital (Case 28-1999). N Engl
J Med 1999;341:901-8.
2. Hsich G, Kenney K, Gibbs CJ Jr, Lee KH, Harrington MG. The 14-3-3 brain
protein in cerebrospinal fluid as a marker for transmissible spongiform
encephalopathies. N Engl J Med 1996;335:924-30.
3. Brandel J-P, Beaudry P, Delasnerie-Laupretre N, Laplanche J-L.
Maladie de Creutzfeldt-Jakob: valeur diagnostique de la detection de la
proteine 14-3-3 et du dosage de la NSE dans le liquide cephalo-rachidien.
Rev Neurol 1999;155:148-51.
4. Zerr I, Bodemer M, Gefeller O, et al. Detection of 14-3-3 protein in the
cerebrospinal fluid supports the diagnosis of Creutzfeldt-Jakob disease. Ann
Neurol 1998;43:32-40.
5. Human transmissible spongiform encephalopathies. Wkly Epidemiol Rec
Dr. Shinobu replies that the comments by Zeidler et al. convey several important points made in the original discussion, which was shortened for publication.

Elevations in the cerebrospinal fluid protein levels of lactic acid, A beta-protein, tau, ubiquitin, protein S-100, and neuron-specific enolase all have some demonstrated usefulness in differentiating cases of Creutzfeldt-Jakob disease from other neurodegenerative diseases, when there is a high a priori clinical suspicion of the disease. (1,2,3,4,5)

Nonetheless, detection of the 14-3-3 protein in the cerebrospinal fluid, in the appropriate clinical context, deserves special attention. The current criteria of the World Health Organization underscore the consensus on the usefulness of this marker.

Analysis of data being collected by the National Prion Disease Pathology Surveillance Center [apparently meaning the Gambetti lab -- webmaster] (which, among other things, tests frozen brain tissue and blood for the gene for prion protein and tests cerebrospinal fluid for the 14-3-3 protein) should help determine the incidence and specific characteristics of selected cases in which tests for the 14-3-3 protein are currently considered to have false negative results.


1. Kropp K, Zerr I, Schulz-Schaeffer WJ, et al. Increase of neuron-specific
enolase in patients with Creutzfeldt-Jakob disease. Neurosci Lett
2. Nooijen PT, Schoonderwaldt HC, Wevers RA, Hommes OR, Lamers KJ.
Neuron-specific enolase, S-100 protein, myelin basic protein and lactate in
CSF in dementia. Dement Geriatr Cogn Disord 1997;8:169-73.
3. Otto M, Wiltfang J, Tumani H, et al. Elevated levels of tau-protein in
cerebrospinal fluid of patients with Creutzfeldt-Jakob disease. Neurosci
Lett 1997;225:210-2.
4. Otto M, Stein H, Szudra A, et al. S-100 protein concentration in the
cerebrospinal fluid of patients with Creutzfeldt-Jakob disease. J Neurol
5. Manaka H, Kato T, Kurita K, et al. Marked increase in cerebrospinal fluid
ubiquitin in Creutzfeldt-Jakob disease.  Neurosci Lett 1992;139:47-9.

Monitoring plasma processing steps with a sensitive Western blot assay for the detection of the prion protein.

J Virol Methods 2000 Jan;84(1):77-89
Lee DC, Stenland CJ, Hartwell RC, Ford EK, Cai K,
Miller JL, Gilligan KJ, Rubenstein R, Fournel M, Petteway SR 
"Determining the risk of transmissible spongiform encephalopathy (TSE) transmission by blood or plasma-derived products requires sensitive and specific assays for the detection of either infectivity or a reliable marker for infectivity.

To this end, a Western blot assay that is both sensitive and reproducible for the detection of PrP(RES), a marker for TSE infectivity, was developed. Using the 263K strain of TSE as a model system, the Western blot assay proved to be sensitive, specific and quantitative over a 3-4 log dynamic range.

Compared to the rodent bioassay, the assay was shown to detect PrP(RES) down to approximately 10(3.4) IU/ml which is approximately 5-10 pg of PrP or approximately 10-20 ng brain equivalents. The Western blot was applied to monitor the partitioning of spiked PrP(Sc) through three plasma fractionation steps, cryoprecipitation, fraction I and fraction III, that are common to the purification of several human plasma-derived therapeutic products including albumin and immunoglobulins.

The results from these studies demonstrated 1 log, 1 log and 4 logs of PrP(Sc) partitioning away from the effluent fraction for the cryoprecipitation, fraction I and fraction III steps, respectively."

Comment (webmaster): This paper, which is beautifully written and a model of clarity, gives some idea what researchers at Bayer have been up to as far as developing prion assays for blood product safety. It is clear that the 3 steps of purification are having their effect but whether they think the amount left at the end is significant if injected (over a human lifespan) is not made clear.

The authors carefully discuss their pivotal assumption, that Prp-res is a proxy for infectivity. (Bioassays take too long to be useful in practical medical uses of blood products.)

Assuming this, how much Prp-res in final product is too much? Will a proxy for infectivity be accepted as a policy trigger for a catastrophically expensive recall? Zero-tolerance could become very inconvenient if further improvements in assay become possible, eg, if tiny amounts of residual Prp-res are used to amplify added Prp-sens prior to assay.

The sensitivity of the test here (which benefits from chemiluminscence on the blot membrane, with 3F4 monoclonal used to detect PSKPKTNMKHMAG) is not compared to capillary electrophoresis. That latter test, which promises great sensitivity and selectivity, has encountered considerably flak at meetings and lacks a serious assessment in print.

Mice have become increasingly problematic as a calibration device for determining what is an "acceptable" level of residual Prp-res. Partly, this stems from a 2 year lifespan versus a 75 year lifespan in an immunized child. Multiple copies of human prion might overcome this limitation to a certain extent; however, 15% of the time that Prp-res originates from a coding sequence mutant with a non-standard interaction (more potent?) with normal prion. Doppel in mouse seems regulated quite differently than in human; none of the mouse strains constructed to date have taken this into account.

Carnoy's fixative improves the immunohistochemistry of the proteinase K-resistant prion protein.

Brain Pathol 2000 Jan;10(1):31-7
Giaccone G, Canciani B, Puoti G, Rossi G, Goffredo D, Iussich S, Fociani P, Tagliavini F, Bugiani O
The neuropathological diagnosis of Creutzfeldt-Jakob disease relies on the immunohistochemical demonstration of the proteinase-K resistant form of the prion protein (PrPres) in the brain tissue.The antigenicity of PrPres is strongly reduced by the formalin solution widely used to fix the tissue, thus the PrPres immunoreactivity is inconsistently detectable in formalin-fixed tissue.

A better PrPres immunostaining can be obtained by using Carnoy's fixing solution, which is composed of ethanol, chloroform and acetic acid (6:3:1). PrPres can easily be extracted from Carnoy's-fixed, paraplast-embedded tissue. Accordingly, Carnoy's-fixed tissue can prior to immunolabeling be subjected to proteinase K and guanidine thiocyanate, which respectively eliminate the normal cellular form of prion protein and promote protein denaturation.

In comparison with the best protocols for formalin-fixed tissue (i.e.--hydrolytic autoclaving or autoclaving in distilled water followed by formic acid and guanidine thiocyanate), PrPres immunostaining carried out on sections cut from Carnoy's-fixed, paraplast-embedded tissue blocks and subjected to proteinase K and guanidine thiocyanate, proved more successful to detect and map both diffuse and focal PrPres immunoreactivity, and to correlate the immunoreactivity pattern with MV polymorphism at PRNP codon 129 and PrPres banding and glycosylation pattern revealed by Western blot.

'Prion' properties of the yeast

Medline 12 Feb 00
Comment (webmaster): There will come a time, hopefully soon, when yeast researchers will grow up and stop clinging to the apron strings of real prion research. The genes and proteins with congophilic fibril-forming properties studied to date in yeast and other fungi have no demonstrated relationship to any of the 20-odd mammalian amyloidoses. A great many proteins from phage to mammals have this property -- in fact, protein chemists now argue that almost all do under proper conditions.

The problem is a total lack of homology. Ironically, many human amyloidosis genes do have valid fungal orthologues (though the prion gene is not among them). However, the relationships are so faint that the domains critical for aggragation are not well preserved. There is no motivation for evolutionary conservation of this property -- after all, aggragation into fibrils is not a normal function of any human amyloidosis gene product.

Lack of homology diminishes the implications of research into chaperones -- are they involved at all in human amyloidosis and if so why should they be orthologues of the respective yeast chaperone? The same could be said for seeding mechanisms. The normal function of the sup35p domain remains as elusive as ever despite a commendable round of comparative sequencing; unsuspected cofactors could regularize the structure.

It is completely at odds with biochemical nomenclature to give a remote analogy (with obscure significance in nature) a name that deliberately misleads. What's next, shall we say an uncontrolled burst of cell division in E. coli is cancer, a ciliate photoreceptor defect a retinal disease, a hemocyanin aggragate is sickle cell anemia? Who wants all these extraneous entries when they search GenBank or Medline? The word prion has long been officially registered as the product of the PRNP gene. Grant money is not so short that desparate metaphors are necessary.

Prion researchers don't call their protein beta-amyloid. Could not sup35p equally well be called huntingtin? How would yeast researchers feel if a spinocerebellar ataxia protein was called [psi]?

Even minimal-decency terms such as 'prion-like' or 'prion-analogue' seem unwarranted for yeast sup35p. The deeper analogy structurally is not really to the aggregating domain of prion protein but rather to those of polyglutamine disorders, some far more significant medically than CJD. Even here the case for relevence has not been made in terms of pathological mechanism. Perhaps it would be better to use the inherent advantages of yeast manipulation with expressed bona fide mammalian amyloidoses proteins. [This is finally happening, see PNAS 2000 Feb 15;97(4):1589-94.]

This is not to say that research into fibril-forming non-homologous proteins of fungi is uninteresting in its own right. In fact, much faster progress is possible due to the advanced status of yeast as an experimental organism. That is precisely the point: this reasearch can and should stand on its own two feet, which is not as ersatz prion.

Here are 16 Medline yeast abstracts published in the last 90 days.

     Sondheimer N, et al. 
       Rnq1: an epigenetic modifier of protein function in yeast. 
       Mol Cell. 2000 Jan;5(1):163-72. 

[Comment (webmaster):    The abstract claims, improbably, that search criteria were developed and applied to the [yeast?] protein database to find other prion-like proteins. One of several reported, rnq1, is said validated.  However, rnq1 was already listed as an epigenetic factor by 1992 [see NP_009902] probably just reflecting a glutamine ladder due to its high glutamine/asparagine content, 126 of 405 residues, eg ..NQQQY NQQGQNNQQQ YQQQGQNYQH QQQGQQQQQG...  A very large number of such proteins is found at GenBank; indeed, human disease researchers conducted direct genomic searches years ago to identify more diseases in this class.]

     Santoso A, et al. 
       Molecular basis of a yeast prion species barrier. 
       Cell. 2000 Jan 21;100(2):277-88. 

     Kushnirov VV, et al. 
       Prion properties of the Sup35 protein of yeast Pichia methanolica. 
       EMBO J. 2000 Jan 17;19(3):324-331. 

     Li L, et al. 
       Creating a protein-based element of inheritance. 
       Science. 2000 Jan 28;287(5453):661-4. 

     Komar AA, et al. 
       The [URE3] yeast prion: from genetics to biochemistry. 
       Biochemistry (Mosc). 1999 Dec;64(12):1401-7. 

     Ter-Avanesyan MD, et al. 
       Prions: infectious proteins with genetic properties. 
       Biochemistry (Mosc). 1999 Dec;64(12):1382-90. 

     Kisselev LL. 
       Translation termination and yeast prions. 
       Biochemistry (Mosc). 1999 Dec;64(12):1337-41. .

     Lindquist I, et al. 
       Expanding the prion model for the yeast [PSI+] element: response. 
       Trends Microbiol. 2000 Jan;8(1):3. No abstract available.

     Masison DC. 
       Expanding the prion model for the yeast [PSI+] element. 
       Trends Microbiol. 2000 Jan;8(1):1-2. No abstract available.

     Eaglestone SS, et al. 
       Guanidine hydrochloride blocks a critical step in the propagation of the
       prion-like determinant [PSI(+)] of Saccharomyces
       Proc Natl Acad Sci U S A. 2000 Jan 4;97(1):240-4. 

     Serio TR, et al. 
       [PSI+]: an epigenetic modulator of translation termination efficiency. 
       Annu Rev Cell Dev Biol. 1999;15:661-703. Review. 

     Wickner RB, et al. 
       Prions in Saccharomyces and Podospora spp.: protein-based inheritance. 
       Microbiol Mol Biol Rev. 1999 Dec;63(4):844-61, table of contents. Review. 

     Burck CL, et al. 
       Translational suppressors and antisuppressors alter the efficiency of the Ty1 programmed translational frameshift. 
       RNA. 1999 Nov;5(11):1451-7. 

     Lansbury PT. 
       Bungee cord domains' balancing act. 
       Curr Biol. 1999 Nov 18;9(22):R845-R847. 

     Chernoff YO, et al. 
       Evidence for a protein mutator in yeast: role of the Hsp70-related chaperone ssb in formation, stability, and toxicity of the
       [PSI] prion. 
       Mol Cell Biol. 1999 Dec;19(12):8103-12. 

     Ma J, et al. 
       De novo generation of a PrPSc-like conformation in living cells. 
       Nat Cell Biol. 1999 Oct;1(6):358-361. 

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