Prion Disease: Doppel Nmr Structure
Mad Cow Home ... Best Links ... Contact Researcher ... Science Index

Doppel nmr structure
Doppel expresion level and onset of ataxia
Strain typing of French nvCJD; sheep scrapie strain matches sporadic CJD
Another case of human intergenic splicing: P2Y11 and SSF1
Fourth Alzheimer gene found
Xanthinuria II: another bovine disease that humans get
Dehydrated beta-sheet structure for yeast amyloid
A 7k PrP fragment is the major amyloid protein in GSS A117V
Membrane topology influences N-glycosylation of the prion protein
Joe Gibbs, top neurological researcher, is dead at 76

Doppel nmr structure

27 Feb 01 PNAS
H Mo, R Moore, FE Cohen, D Westaway, SB Prusiner, PE Wright HJ Dyson
PDB coordinates: 1I17 use #16 of 20 structures or Amber mean structure
Comment: This welcome experimental structure for doppel is an important development, though in some ways it is a disappointment in that the structure largely confirms secondary and tertiary predictions made back on 29 Aug 99. These predictions were hardly rocket science: the doppel fold is basically that of the prion protein because of conserved anchors (such as the hydrophobic core and disulfide) and an adequate level of primary sequence alignment (folds are commonly conserved down to 5-7% identity, whereas prion and doppel approach 25%). Warwicker also posted some months back molecular dynamics models for doppel and the second disulfide etc. were verified in several earlier biochemistry papers

Now the experimentally determined structure did have some interesting twists. The mouse doppel nmr paper began by studying a recombinant complete mature protein, residues 26-157, but the doppel structure turned out to be determinable only for 51-157, a result reminiscent of the prion protein even though the amino termini are not alignable. However residues 32-38 of doppel showed some signs of associating with the main globular domain.

While the secondary and tertiary structures are quite similar to the prion protein, two slight differences occur in doppel: the beta sheet plane is parallel, not perpendicular, to helix B and C axes, and helix B has a kink between helical sections at A116-N117. The second disulfide pair unsurprisingly imparts more rigidity to the affected loop in doppel.

The mouse doppel beta strands are at AFI and DGI (one letter amino acid codes), both exactly conserved in the 6 species sequenced so far. The AFI is at the expected homologous position; the DGI is upshifted 2 residues from the expected IYY homologous second beta strand in prion protein (possibly thus accounting for the parallel beta plane). The alpha helices are helix A: AEGNRYYAAN (as expected except for 5' A) , helix B: TKEMLVTSCVNATQAANQAEFSREK (TKE and REK not expected), and helix C: KLHQRVLWRLIKEICSA (short KHC).

Aligning published doppel sequences, including ESTs, and inferring the ancestral mammalian doppel, we see doppel quite conserved yet diverging twice as fast as prion protein across the same species set:

The authors observe, like others before them, that surface residues of helix A facing away from the protein are conserved. Since these do not contribute to globular domain stability, this feature is likely involved in a protein-protein interaction. Whether that partner is another copy of doppel, the prion protein itself, or some signaling cascade protein remains to be determined.

Is the amino terminus really unstructured, or is this an artefact of renatured recombinant protein, nmr buffer conditions, or missing small or large ligand? Could doppel and prion have two natural fuknctioning conformations in addition to the cross-beta amyloid conformation not accessible to doppel?

While the authors can only report what they observed, it is most unlikely that this domain is unstructured in the native state: the conservation of residues across 6 species, in a rapidly evolving protein, is phenomenal, fully comparable to deep internal conserved residues. What selective forces could provide such strong constraints in different lineages over hundreds of millions of years on extended amino acids if there was no packing?

::******:*****.***:. *:***:.** ***** 6 species, invariant residues

Thus, as with prion protein, nmr may be producing rapid but wrong results: perhaps we would be better off with xray crystallography of native protein released from the cell surface in capturing more subtle packing. Nmr coordinates even of the stable domain are adversely influenced by edge and surface effects -- if 26-50 is not folded in properly, of course that influences the rest of the structure, especially if this region normally docks to make a more globular protein. Indeed, rumors are swirling around of a fourth helix in the 39-52 region found separately by French and English groups.

Now the signal peptide of any protein possessing one is an example of a validated extended region, be it short-lived. Provided it satisfies somewhat generic hydrophobic and helical constraints that allow it to lie in the pocket of the processing system (known from xray structure), it can be processed correctly at the endoplasmic reticulum. Primary sequence conservation is moderate across species for a given protein's signal region, yet no post-met invariant residues exist across signal peptides within a species despite the common processor pocket.

Doppel signal has 9 conserved residues in its 26 residue signal across 6 species. These are eroding roughly at a rate of one residue per new species sequenced. The "unstructured" region of doppel has 18 invariant residues over the following 26 mature residues with no erosion from the most recent sequence (pig).

The signal region of prion protein has been determined in 45 genera of mammals. (Alleles are found in gorilla, kudu, and squirrel monkey; 8 sequences are partial.) This data is available as a comma delimited database presorted by phylogenetic adjacency. Note Figure 1 of the paper has the signal peptide misgapped.

The most parsimonious explanation here is that a 6 base pair deletion occurred in the common ancestor prior to rodent divergence at 100 myr. That deletion persisted for millins of years as a polymorphism but only won out in rodent, primate, and lagomorph lineages, with carnivores and artiodactyls retaining the ancestral length.seen in marsupial, avian, and turtle outgroups.

The alignment of doppel and prion in Figure 1 of the paper is also less than optimal in that doppel actually has no counterpart to the 106-126 amyloidogenic region of prion. Reliable homological alignment of these proteins became possible with turtle sequence along with prediction of beta sheet and alpha helices.

The alignment of the prion-doppel signal region from representative genera to the left is color-coded for amino acid polarity and residue size. (Cysteine is put in with aromatics in view of how its codons are placed within the genetic code.) A central hydrophobic region is flanked by charged and polar regions.

One unresolved deeper issue not discussed in the paper is whether mouse doppel is actually closer (more coupled in sequence) to mouse prion than, say, mouse doppel and cow prion (or even mouse doppel and chicken prion). The former scenario is expected if a domain in both proteins interacted with a common third party within mouse, forcing co-evolution to this common constraint (itself evolving). The interest here is in the mature proteins; the signal and GPI domains are well-known to interact with species-specific but orthologous processing systems. For example, if prior to the tandem duplication 500 million years ago, the ancestral protein was a Jacob-Monod dimer, that binding might be retained subsequent to duplication as a quasi-twofold symmetry in the heterodimer giving rise to retention of primary sequence coordination.

This issue is potentially of huge significance to normal function: a better determined fold class might allow recognition of a larger superfamily (via Dali) in which primary sequence homology is lost but the fold retained -- some member of this family might have a known function.

No ready answer can be inferred from primary sequence comparisons -- mouse prion as a whole is equidistant from the 6 known doppels. Individual domains such as helix A are too short to afford statistically significant differences. If more experimental data were available, one might compute root mean square differences between all (prion, doppel) pairs hoping that same-species comparisons might consistently correlate best.

It is not clear whether the Swiss nmr group is also determining mouse or instead human doppel. In human, the issues are the extra glycine insertion in mouse only and the 2 residues deleted in human helix B. It is not a structurally trivial matter to delete 2 residues in a helix that naturally has 3.4 residues per turn because charged or polar exterior side chains are rotated to the interior (and vice versa with apolar). Humans have the potentially harmful P56L allele found by Laplance; as expected this invariant proline terminates the beta strand in the nmr structure, which along with the adjacent glycine suggests a conserved sharp bend.

The tandem duplication event will likely be dated soon by genomics projects in 3 species of fish, skate, and a tunicate. This may also resolve which protein lineage gained/lost the repeat region and invariant neurotoxic domain. Because homology between mammalian doppel and prion is barely detectable by sequence alignment (though strongly affirmed by tandem chromosomal position), it is still possible that even more weakly aligning orthologues are to be found in nematode and fruit fly.

Ataxia and Purkinje cell loss in null mice inversely correlated with brain doppel level

EMBO Journal, Vol. 20, No. 4 pp. 694-702, 2001
Daniela Rossi, Antonio Cozzio1, ... Adriano Aguzzi and Charles Weissmann

PrP knockout mice in which only the open reading frame was disrupted (Zurich I) remained healthy. However, more extensive deletions resulted in ataxia, Purkinje cell loss and ectopic expression in brain of Doppel (Dpl), encoded by the downstream gene, Prnd. A new PrP knockout line, Zurich II, with a 2.9 kb Prnp deletion, developed this phenotype at 10 months (50% morbidity). A single Prnp allele abolished the syndrome. Compound Zurich I/Zurich II heterozygotes had half the Dpl of Zurich II mice and developed symptoms 6 months later. Zurich II mice transgenic for a Prnd-containing cosmid expressed Dpl at twice the level and became ataxic 5 months earlier. Thus, Dpl levels in brain and onset of the ataxic syndrome are inversely correlated....

Here, we describe the generation of a further PrP knockout line, hereafter called Prnp­­ Zurich II, in which the PrP-encoding exon and its flanking regions were replaced by a loxP site. The homozygous mice, like their Prnp­­ Nagasaki counterparts, developed progressive ataxia and age-dependent Purkinje cell loss starting at 5­6 months, with half the mice affected by 10 months, and showed ectopic expression of two Prnd-derived mRNAs and Dpl in brain. A single wild-type PrP allele fully corrected the deleterious phenotype in agreement with previous reports (Sakaguchi et al., 1996; Nishida et al., 1999). The F1 offspring of a cross between the Zurich I and Zurich II lines showed partial complementation, in that they remained healthy 6 months longer than the Zurich II mice, despite the absence of a PrP ORF. The level of Prnd-derived mRNAs and Dpl in brain was half that in Zurich II mice.

Introduction into Zurich II mice of a cosmid comprising both the Prnd gene and a Prnp locus lacking the PrP ORF resulted in accelerated development of the ataxic phenotype and was associated with increased levels of Prnd-derived transcripts and Dpl in brain. These results suggest that the allele with deletions extending beyond the PrP ORF is pathogenic in a dose-dependent fashion and that this pathogenicity is due to ectopic expression in brain of Dpl (Moore et al., 1999; Li et al., 2000) rather than to the absence of PrP. N-proximally truncated PrP, which resembles Dpl, also causes ataxia that, as in the case of ectopic Dpl expression, is counteracted by full-length PrP (Shmerling et al., 1998).

Five independent PrP knockout mouse lines have been reported. Three of these show cerebellar symptoms and loss of Purkinje cells on ageing, namely the Prnp-/- Nagasaki mice, the Rcm0 mice and the Prnp-/- Zurich II mice, while the Prnp-/-Edinburgh and the Prnp0/0 Zurich I mice do not. The strategies used to abolish PrP differed in an important respect: in the lines remaining healthy, PrP expression was abrogated either by placing an insert within the PrP coding region (the Edinburgh mice) or by replacing the coding region between codons 3 and 188 by a neo cassette. In contrast, the Nagasaki, Rcm0 and Zurich II lines were generated by deleting not only the ORF, but also 5' flanking sequences extending into the second intron and 3' non-coding sequences.The three lines have in common the loss of 270 bp upstream of the PrP reading frame and of 450 bp downstream. Whilst the deleted sequences in the Nagasaki mice were replaced by a neo cassette, which, at least in some cases, causes an abnormal phenotype (Fiering et al., 1995), those in the Zurich II mice were replaced by a 34 bp loxP sequence, which is not known to cause deleterious effects.

Although the cerebellar phenotype resulting from extended deletions in the PrP gene can be rescued by a wild-type PrP allele, it is clearly not caused by the absence of PrP, because the Zurich I and the Edinburgh mice remain healthy despite their lack of PrP (Weissmann, 1996). In the Edinburgh mice no PrP-specific mRNA was detected, so that significant levels of any fusion protein containing PrP sequences are unlikely. In the Zurich I mice the neo cassette was inserted between the third and the 188th codon of the PrP sequence; although its coding and 3' non-coding sequence were in-frame with the 67 residual PrP codons, there were two termination codons in between, which would preclude read-through. The relevant DNA segment from the Zurich I mice currently in use was resequenced and the presence of the termination codons confirmed (data not shown). Therefore, the presence of a PrP fragment or a fusion product is not responsible for maintaining the normal phenotype in either of the two lines.

The fact that the knockout lines showing the cerebellar phenotype lack sequences flanking the PrP ORF suggested that critical information was partly or entirely located in these regions. The pathological phenotype could come about either by loss of function, if these flanking regions controlled the formation or encoded part or all of some essential protein or RNA, or by gain of function, if the extended deletion resulted in the production of a deleterious product. The finding that introduction of a wild-type Prnp allele, either by breeding or as a transgene, abrogated the ataxic phenotype could be accommodated by either explanation: because the Prnp allele contains the flanking sequences, it could supply the missing function conjectured by the loss-of-function hypothesis. Alternatively, within the framework of the gain-of-function hypothesis, PrP might overcome the pathogenic effect of a postulated deleterious product.

The discovery of Prnd, the gene encoding Dpl, and its expression in the brains of Zurich II, Nagasaki and Rcm0 mice, suggested that the PrP knockout alleles in these animals give rise to a deleterious product. Analysis of brain-derived cDNAs indicated that in wild-type mice Dpl mRNA is very weakly expressed, mainly from a promoter upstream of exon 1 of Prnd, whilst the strong expression in Nagasaki and Rcm0 mice is due to chimeric RNAs that originate at the Prnp promoter, run all the way across and past the Prnd ORF and are processed by one or more splicing events that link the 3' end of the second PrP exon directly or indirectly to the Dpl-encoding exon (Moore et al., 1999; Li et al., 2000). This intergenic splicing, which was also detected by PCR at very low levels in wild-type mice, is greatly enhanced in the ataxic mice because the splice acceptor site upstream of the PrP-encoding exon is deleted, thus diverting the splice to a downstream acceptor site. Prnd-specific mRNA was expressed undiminished in brain of Nagasaki mice cured by the introduction of a PrP-expressing transgene (Nishida et al., 1999).

The hypothesis that expression of Dpl in brain is responsible for the ataxic syndrome is supported not only by the fact that in Zurich I knockout mice containing a single Zurich II allele onset of ataxia and Purkinje cell degeneration is retarded, but also by the finding that Dpl expression in the brain at twice the level of that in Zurich II mice, found in mice transgenic for a cosmid devoid of the PrP ORF but containing Prnd, accelerates appearance of the symptoms.

Why should overexpression of Dpl cause ataxia and concurrent overexpression of PrP restore normal function? Shmerling et al. (1998) found that introduction into Zurich I Prnp0/0 mice of an amino-proximally truncated transgene encoding PrP devoid of the octa repeats and the conserved 112­126 region (PrP32­134) leads to ataxia and degeneration of the cerebellar granule cell layer within weeks of birth. Moreover, introduction of a single wild-type PrP allele prevented the disease.

They proposed that PrP interacts with a ligand to elicit an essential signal and that a conjectured PrP-like molecule with lower binding affinity can fulfil the same function in the absence of PrP. According to this hypothesis, in PrP knockout mice the truncated PrP could interact with the ligand, displacing the PrP-like molecule, without, however, eliciting the survival signal. If PrP has the higher affinity for the ligand, it would displace its truncated counterpart and restore function. Because Dpl resembles the truncated PrP, it might cause disease by the same mechanism (Moore et al., 1999; Silverman et al., 2000).

However, because the promoter used to express the truncated PrP is active in granule cells but not in Purkinje cells, while the wild-type PrP promoter directing the ectopic expression of Dpl is active in Purkinje cells (Li et al., 2000), the cellular targets may be different. Indeed, targeting the truncated PrP to Purkinje cells causes ataxia and degeneration of Purkinje cells (E.Flechsig, R.Leimeroth and C.Weissmann, unpublished data).

The wild-type Prnp allele or a Prnp-containing cosmid gives rise to PrP expression in Purkinje cells that may counteract the conjectured deleterious effect of Dpl. Interestingly, the Tga20 transgene also has this beneficial effect, although there is no detectable expression in Purkinje cells. Perhaps PrP expressed on the synapses of neighbouring cells can supply the protective effect or can be physically transferred, as shown for other GPI-linked proteins (Kooyman et al., 1995; Anderson et al., 1996; Brunschwig et al., 1999; McHugh et al., 1999). Interestingly also, the ataxia provoked by expression of truncated PrP in Purkinje cells can be abrogated by the Tga20 allele, perhaps by the same mechanism (E.Flechsig, R.Leimeroth, I.Hegyi and C.Weissmann, unpublished data).

It would thus seem that chronic ablation of PrP per se has only quite modest effects such as alterations in circadian activity rhythms and sleep patterns (Tobler et al., 1996) and demyelination in the peripheral nervous system with old age (Nishida et al., 1999). The hypothesis that expression of Dpl in Purkinje cells causes cell death and ataxia and that the deleterious effects can be counteracted by concomitant expression of PrP is attractive and strongly supported by our results.

Cotranscription and intergenic splicing of human P2Y11 and SSF1genes

J Biol Chem February 5, 2001
Didier Communi, Nathalie Suarez-Huerta, Danielle Dussossoy, Pierre Savi, and Jean-Marie Boeynaems
The P2Y11 receptor is an ATP receptor positively coupled to the cAMP and phosphoinositide pathways. SSF1 is a Saccharomyces cerevisiae nuclear protein which plays an important role in mating. The gene encoding the human orthologue of SSF1 is adjacent to the P2Y11 gene on chromosome 19. During the screening of placenta cDNA libraries, we isolated a chimeric clone resulting from the intergenic splicing between the P2Y11 and SSF1 genes. The fusion protein was stably expressed in CHO-K1 cells, where it generated a cAMP response to ATP qualitatively undistinguishable from that of the P2Y11 receptor. According to both Western blotting and cAMP response, the expression of the fusion protein in the transfected cells was clearly lower than that of the P2Y11 receptor. Both P2Y11 and SSF1 probes detected a 5.6 kb messenger RNA with a similar pattern of intensity in each of 11 human tissues.

The ubiquitous presence of chimeric transcripts and their upregulation during granulocytic differentiation indicate that the transgenic splicing between the P2Y11 and the SSF1 genes is a common and regulated phenomenon. There are very few examples of intergenic splicing in mammalian cells, and this is the first case involving a G-protein coupled receptor.

Comment (webmaster): The situation with these tandem genes may be visualized at the UCSC genome browser. One sees that there are only about 300 bp separating the genes, rather short for both a polyA in 3'UTR as well as a promoter for the following gene. Prion and doppel are separated by some 26,000 bp in human

Fourth Alzheimer gene found

7 Jan 01 webmaster add-on to 21 Dec 00 Science articles
Opinion (webmaster): One of the great mysteries of CJD has been the lack of work on other genetic loci involved. Other than the recent QTL paper from Carlson, which partly delineated two new sites in mouse, we have no idea what might be involved. This is particularly relevant to nvCJD because -- beyond the met/met at codon 129 of the prion gene -- the victim set is puzzling, suggesting other loci important to susceptibility.

A fair proportion of "sporadic" CJD may actually be familial -- either non-coding regions of the prion gene [newly surveyed in a Taiwanese study of schizophrenia, n=62] or attributable to some other loci entirely. Family hisories are often uninformative, due to late onset and little liklihood of diagnosis in bygone eras. The Laplanche and Collinge groups found no support for a role for doppel in either moderating or mitigating CJD.

In Alzheimer, researchers have found three distinct genes: beta-amyloid precursor protein (APP) on chr 21q21.3, presenilin 1 on chr 14q24.3, and presenilin 2 on chr 1q31 cause the rare, early-onset autosomal dominant form of Alzheimer's disease.

These mutations all affect APP metabolism, leading to over-production of Abeta42 peptide, the amyloidogenic peptide villain. In CJD mouse models, over-production of prion protein analogously results in earlier onset, higher susceptibility to outside challenges, and so on, probably attributable to a mass action effect. It is thus important to know of alleles of other genes that could lead to prion overproduction. These do not necessarily lead to disease characteristic of the affected gene but rather could be manifested in prion gene disease.

In the 22 Dec 00 issue of Science, three back-to-back papers reported a fourth Alzheimer locus on chr 10q23.1, already reviewed at OMIM on 4 Jan 01.. However, they were only able to partly map the gene to a region of several million bp, with IDE (insulin degrading enzyme; insulysin; insulinase) being mentioned as in linkage disequilibrium. IDE was suggested as a candidate gene because both the presenilins are proteases that act on APP to produce more of the longer amyloidogenic peptide. IDE has also come up as a protease with possibility specificity for degrading amyloidogenic proteins -- in this scenario, reduced IDE leads to reduced clearance of Alzheimer amyloid, tilting the equilibrium in the direction of pathogenicity.

None of the papers did any genomics work (despite 52 total authors!) involving this region of the nearly finished human genome, so the webmaster looked at modern finished genome assemblies (1, 2) to see if the gene could be found. First, the gene order in the neighborhood of the best LOD score is readily found from the UCSC genome browser or LocusLink finished contig for IDE, namely NT_024129.

TAF172 .. PPP1R5 .. FLJ20445 .. KIAA0940 .. FLJ20479 .. HHEX=PRHX .. KNSL1 .. IDE

Looking up this gene region in OMIM or reviewing partly mapped unassigned dementia loci in MorbidMap does not turn up any plausible candidates, eg a dementia locus not necessarily recognized as Alzheimer.

Another easy thing to check is whether any of the 3 known causative genes have a homologous copy in the general vicinity of chr 10q23, the idea being that paralogous genes are often associated with related diseases. This amounts to looking up the proteins, concatenated into a single probe to save on queries, and tblastning against finished and unfinished human genome.

This turns up fairly good matches on chr 9, 11, and 19 (in addition to recovering the 3 known genes) but no candidates on chr 10. These paralogs might be worth pursuing in their own right for related diseases, but the bottom line is that the chr 10 loci is an altogether new gene family involved in AD. That's a result not found in the three Science papers despite the plethora of authors -- genomics has not percolated down to the laboratory level.

Next, even though the genes in the vicinity of IDE on chr 10q had no known associated disease, it is possible that their close homologues elsewhere in the genome would give some clue as to what sort of disease they might cause.

Because the three Science papers likely held back what they really knew about the gene's locus for a second publication, only IDE itself was checked, rather than 10 genes on each side. This gene turns out to be single copy, rather like the prion-doppel locus, with no closely related proteins or even related domains (despite 1019 aa) anywhere in the genome.

However, the distantly related NRD1 nardilysin (N-arginine dibasic convertase chr 1p32.2-p32.1) at 33% is another representative of this ancient family, eg drosophila 46% IDE, yeast or E. coli protease III.

It can also be productive to look for mouse models of disease on the syntenic mouse chr 19 and see if a disease has been assoicated with IDE in mouse. Actually the diabetic GK rat have a known IDE problem at the NIDDM1B locus as seen in OMIM 146680 even though this is not a known human diabetes locus. It is not clear whether dementia in rats would be readily apparent.

Degradation of amylin by insulin-degrading enzyme.

J Biol Chem 2000 Nov 24;275(47):36621-5
Bennett RG, Duckworth WC, Hamel FG
A pathological feature of Type 2 diabetes is deposits in the pancreatic islets primarily composed of amylin (islet amyloid polypeptide). Although much attention has been paid to the expression and secretion of amylin, little is known about the enzymes involved in amylin turnover. [This should not be confused with a second amyloidosis that involves the insulin protein itself -- the first documented infectious protein to cross the species barrier. -- webmaster]

Recent reports suggest that insulin-degrading enzyme (IDE) may have specificity for amyloidogenic proteins, and therefore we sought to determine whether amylin is an IDE substrate... The data strongly suggest that IDE is an amylin-degrading enzyme and plays an important role in the clearance of amylin and the prevention of islet amyloid formation.

In addition to insulin, a number of other proteins have been identified as IDE substrates, including proteins structurally related to insulin such as proinsulin, insulin-like growth factor II, and relaxin, and seemingly unrelated peptides, such as atrial natriuretic peptide (ANP) and glucagon. However, insulin, ANP, and glucagon all contain regions that can form beta-pleated sheets, and thus are amyloid-forming peptides (9).

Thus, it has been proposed that, rather than displaying specificity for a primary sequence motif, IDE is specific for amyloidogenic peptides (8, 10). Indeed, recently the Alzheimer's beta-amyloid peptide was shown to be degraded by IDE, and IDE was implicated in regulating extracellular levels of beta-amyloid peptide.

Because IDE is an amylin-degrading enzyme, it may play an important role in amylin homeostasis. It is important to remember that in species expressing amyloidogenic amylin (including humans), amylin is continually present, yet does not normally aggregate into amyloid deposits. Therefore, the mere presence of amylin does not predicate islet amyloid formation. This has been explored further in studies using transgenic animals. Although some lines of transgenic mice overexpressing human amylin displayed amyloid deposits (17, 18), others did not (3, 19-21), suggesting that elevated amylin alone may not be sufficient for amyloid formation. Therefore, a perturbation of some other element of amylin processing, such as amylin degradation, may be involved.

Amylin from a number of species, including human, spontaneously aggregates into amyloid fibrils as a result of -pleated sheet formation around residues 20-29 of the primary sequence. Rodent amylin, on the other hand, does not form amyloid, presumably because proline residues in this area of the molecule alter secondary structure (22). In the studies presented here, IDE degraded both human and rat amylin. It would follow that the recognition motif for IDE is not the -pleated sheet region itself, but the structure of the peptides in the non-amyloid-forming state. Indeed, most if not all studies of IDE and insulin have been performed under conditions in which insulin does not spontaneously form amyloid fibrils. The role of IDE may, therefore, be in the cleavage of peptides with the potential to form amyloid aggregates.

In this case, IDE can be thought of as a scavenger of amyloidogenic peptides. Normally, a balance exists between deposition and degradation of the amyloidogenic peptide. When the levels of the peptide exceed the capacity of IDE to degrade them, either by increased expression of the peptide, or decreased expression or enzymatic activity of IDE, the balance is shifted from degradation to deposition. In the case of Type 2 diabetes, both insulin and amylin secretion are increased due to peripheral insulin resistance. Because IDE has approximately 4-fold greater affinity for insulin than for amylin, amylin degradation will be proportionately impaired. The increased production and relative decrease in degradation may allow sufficient accumulation of amylin to cause islet amyloid formation.

The site of both synthesis of amylin and deposition of islet amyloid, the pancreatic beta cells, is a logical site of amylin degradation. An insulin-degrading enzyme consistent with the properties of IDE has been reported in islets and in a beta cell line (23). Furthermore, recent studies in our laboratory have suggested that IDE is responsible for amylin degradation in a beta cell line and that inhibition of IDE increases the formation of amyloid by exogenous human amylin (24). If this proves to be the case, then IDE would be a critical player in the prevention of amyloid formation by amylin and perhaps other amyloidogenic peptides. A new area of study focusing on turnover of amyloidogenic peptides could lead to novel therapeutic approaches in the treatment of amyloid diseases.

Amyloidogenic determinant as a substrate recognition motif of insulin-degrading enzyme.

FEBS Lett 1998 May 8;427(2):153-6
Kurochkin IV 
Insulin-degrading enzyme (IDE) is an evolutionarily conserved neutral thiol metalloprotease expressed in all mammalian tissues whose biological role is not well established. IDE has highly selective substrate specificity. It degrades insulin, glucagon, atrial natriuretic peptide, transforming growth factor alpha but does not act on related hormones and growth factors. The structural properties determining whether a peptide is an IDE substrate are essentially unknown. The reported cleavage sites are not consistent with simple peptide-bond recognition and it was proposed that IDE recognizes in its substrates some elements of tertiary structure.

We noticed that although IDE substrates are functionally unrelated, the majority of them share a specific property, an ability to form under certain conditions amyloid fibrils. Utilizing the residue pattern recognition procedure, this study reveals a common motif in the sequences of IDE substrates, HNHHHPSH, where H is wholly or partly hydrophobic character, N is small and neutral, P is polar, and S is polar and/or small amino acid residue. It is proposed that this sequence motif predetermines a structure recognized by IDE.

The identified motif appears to be essentially the same as the proposed earlier consensus sequence for amyloid-forming peptides [Turnell and Finch, J. Mol. Biol. 227 (1992) 1205-1223]. The study suggests that IDE may play a role in elimination of potentially toxic amyloidogenic peptides.


10 Mar 01 webmaster research
The near-completion of the human genome allows a very accurate compilation of codon use within the redundancies of the genetic code (taken from Nature genome article, fig 34 whole proteome data). By taking the most abundant of the redundant codons for each amino acid, it becomes possible to reverse-translate a protein sequence into DNA (the amino acids with 6 codons must be taken into account correctly).

Why is this useful? Sometimes, a new paper shows a protein sequence but the DNA is not yet posted at GenBank (eg, rat doppel). Retrotranslation is good enough to search a genome by blastn. Sometimes, it is desired to locate a protein by tBlastn but a given web site may only offer Blastn: the retrotranslated gene then suffices.

For 46.1% of the time on average, retrotranslation gets all three codon positions right; overall the result is 78.8% accurate. For the prion protein, retrotranslation gave a gene with 86% identity with the authentic human gene, suggesting either a compositional effect or that the most common codons are used more frequently than the genomewide average.

Data underlying the protein-to-DNA inversion table, implemented as an open source TextSpresso filter.

1aa,3aa,prefered codon, use per 10k, other codons,total aa codon use, wrong letters

Dehydrated beta-sheet structure for yeast amyloid

PNAS 20 Feb 01
Melinda Balbirnie, Robert Grothe, and David S. Eisenberg
X-ray diffraction and other biophysical tools reveal features of the atomic structure of an amyloid-like crystal. Sup35, a prion-like protein in yeast, forms fibrillar amyloid assemblies.

We have identified a polar peptide from the N-terminal prion-determining domain of Sup35 that exhibits the amyloid properties of full-length Sup35, including cooperative kinetics of aggregation, fibril formation, binding of the dye Congo red, and the characteristic cross- x-ray diffraction pattern. Microcrystals of this peptide also share the principal properties of the fibrillar amyloid, including a highly stable, betabeta-sheet-rich structure and the binding of Congo red.

The x-ray powder pattern of the microcrystals, extending to 0.9-Å resolution, yields the unit cell dimensions of the well-ordered structure. These dimensions restrict possible atomic models of this amyloid-like structure and demonstrate that it forms packed, parallel-stranded betabeta-sheets.

The unusually high density of the crystals shows that the packed betabeta-sheets are dehydrated, despite the polar character of the side chains. These results suggest that amyloid is a highly intermolecularly bonded, dehydrated array of densely packed betabeta-sheets. This dry betabeta-sheet could form as Sup35 partially unfolds to expose the peptide, permitting it to hydrogen-bond to the same peptide of other Sup35 molecules. The implication is that amyloid-forming units may be short segments of proteins, exposed for interactions by partial unfolding.

Much of interest is not defined in the model of Fig. 4. This includes all atomic details of the sheet faces that pack against one another, including: the registration of the sheets against each other (along a and c); the precise angle of the strands to c; and the nature of the hydrogen-bonded networks between the side chains. Determining these details is now in progress by electron diffraction and further x-ray diffraction.

A persistent puzzle about amyloids, as well as polyglutamine aggregates and the transformed or scrapie forms of prions has been their unusual stability. Our observation of the densely packed, nearly anhydrous, highly hydrogen-bonded structure for GNNQQNY, also a highly stable structure, offers an answer to this puzzle.

Our proposed hydrogen-bonded sheet structure has essential features in common with Perutz et al.'s proposal of a polar zipper for polyglutamine aggregates (30), and indeed, GNNQQNY is glutamine and asparagine rich. This dry beta-sheet model shares with the polar zipper model the hydrogen-bonding of backbone and side chains. But in the dry beta-sheet model, there are fully formed beta-sheets. This dry beta-sheet model might be expected to be particularly stable because it expels most of the water molecules that ordinarily would be expected to hydrate the polar side chains of asparagine, glutamine, and tyrosine. Dunitz (33) noted that each water molecule that is localized within a solid aggregate has an associated entropy cost of up to 2 kcal/mole at 37°C. Because each Asn or Gln residue has the capacity to form hydrogen bonds with up to seven water molecules and the Tyr residues with five, the almost complete exclusion of water from the dry beta-sheet could account for a major component of the stability of the dry beta-sheet structure. The enthalpic contribution to water-residue hydrogen bonding would be compensated by interresidue hydrogen bonding. This dry beta-sheet can accommodate polar sequences and apolar sequences. Reopening the dry sheet must be a highly cooperative process of breaking many weak bonds, probably accounting for the unusual stability of amyloid.

Abeta amyloid fibrils possess a core structure highly resistant to hydrogen exchange

PNAS 2000 Dec 5;97(25):13597-601
Kheterpal I, Zhou S, Cook KD, Wetzel R. 
We describe here experiments designed to characterize the secondary structure of amyloid fibrils of the Alzheimer's amyloid plaque peptide Abeta, using hydrogen-deuterium exchange measurements evaluated by mass spectrometry. The results show that approximately 50% of the amide protons of the polypeptide backbone of Abeta(1-40) resist exchange in aqueous, neutral pH buffer even after more than 1, 000 h of incubation at room temperature.

We attribute this extensive, strong protection to H-bonding by residues in core regions of beta-sheet structure within the fibril. The backbone amide hydrogens exchange at variable rates, suggesting different degrees of protection within the fibril. These data suggest that it is unlikely that the entire Abeta sequence is involved in H-bonded secondary structure within the amyloid fibril. Future studies using the methods described here should reveal further details of Abeta fibril structure and assembly. These methods also should be amenable to studies of other amyloid fibrils and protein aggregates.

Backbone amide protons in globular proteins occupy a variety of structures that exhibit varying degrees of protection against hydrogen exchange that span at least 6 orders of magnitude of protection factors (17). Core portions of beta-sheet are more highly protected than edge regions of sheets and alpha-helix, which in turn are more protected than loop and random coil regions.

In this paper we show that all of the backbone amide hydrogens of monomeric A undergo rapid HX consistent with the absence of protective structure within the molecule. On the contrary, A incorporated into fibrils undergoes much slower exchange with more complex kinetics with at least three classes of backbone amides in fibrils: those that exchange as rapidly as the backbone amide hydrogens of the monomer, those that exchange at intermediate rates, and those that do not exchange even after 1,000 h of exposure to D2O.

At least 50% of the A peptide backbone resides in this highly protected, rigid core structure in the fibrils. Such high extents of protection over extended exchange times generally are not observed in HX studies on globular proteins.

Even myoglobin can form amyloid

Staff Reporter BBC--Saturday 10 March 2001
Scientists have made a breakthrough in the search for cures for the brain diseases Alzheimer's and CJD. They have discovered that how ordinary proteins within the body turn into the insoluble plaques believed to be at least partly responsible for the diseases.

Researchers at Oxford University's Centre for Molecular Sciences (OCMS) carried out a series of tests which exposed myoglobin - an essential protein which stores oxygen in the body's muscle cells - to a variety of different environments, including changes in temperature and pH. They found that the protein could adopt two distinct and highly-organised forms:

-- a compact folded structure typical of most proteins in the body

-- an insoluble thread-like structure typical of deposits found in the brains of many patients suffering from neurodegenerative disorders such as Alzheimer's

The researchers believe that the compact fold assumed by most proteins is an evolutionary adaptation which ensures that the thread-like structure is never formed under normal circumstances by the components that make up the proteins. Adoption of this structural form therefore prevents the formation of harmful deposits within the carefully regulated environment of the cells.

However, when this environment is disrupted, perhaps by age, genetic mutations or the ingestion of harmful material, the proteins can lose their folds and assume the alternative structure....

Inhibition of interactions and interconversions of prion protein isoforms by peptide fragments from the C-terminal folded domain

J. Biol. Chem February 1, 2001
Motohiro Horiuchi, Gerald S. Baron, Liang-Wen Xiong, and Byron Caughey
The formation of protease-resistant prion protein (PrP-res or PrPSc) involves selective interactions between PrP-res and its normal protease-sensitive counterpart, PrP-sen or PrPC. Previous studies have shown that synthetic peptide fragments of the PrP sequence corresponding to residues 119-136 of hamster PrP (Ha119-136) can selectively block PrP-res formation in cell-free systems and scrapie-infected tissue culture cells.

Here we show that two other peptides corresponding to residues 166-179 (Ha166-179) and 200-223 (Ha200-223) also potently inhibit the PrP-res induced cell-free conversion of PrP-sen to the protease-resistant state. In contrast, Ha121-141, Ha180-199 and Ha218-232 were much less effective as inhibitors.

Mechanistic analyses indicated that Ha166-179, Ha200-223 and peptides containing residues 119-136 inhibit primarily by binding to PrP-sen and blocking its binding to PrP-res. Circular dichroism analyses indicated that Ha117-141 and Ha200-223, but not non-inhibitory peptides, readily formed high beta sheet structures when placed under the conditions of the conversion reaction. We conclude that these inhibitory peptides may mimic contact surfaces between PrP-res and PrP-sen and thereby serve as models of potential therapeutic agents for transmissible spongiform encephalopathies.

A 7k PrP fragment is the major amyloid protein in GSS A117V

J. Biol. Chem. 2001 276(8): p. 6009-6015
Fabrizio Tagliavini,..., Blas Frangione, and Frances Prelli
GSS is a cerebral amyloidosis associated with mutations in the prion protein. The aim of this study was to characterize amyloid peptides purified from brain tissue of a patient with the A117V mutation who was Met/Val heterozygous at codon 129, Val129 being in coupling phase with mutant Val117.

The major peptide extracted from amyloid fibrils was a ~7-kDa PrP fragment. Sequence analysis and mass spectrometry showed that this fragment had ragged N and C termini, starting mainly at Gly88 and Gly90 and ending with Arg148, Glu152, or Asn153. Only Val was present at positions 117 and 129, indicating that the amyloid protein originated from mutant PrP molecules. In addition to the ~7-kDa peptides, the amyloid fraction contained N- and C-terminal PrP fragments corresponding to residues 23-41, 191-205, and 217-228.

Fibrillogenesis in vitro with synthetic peptides corresponding to PrP fragments extracted from brain tissue showed that peptide PrP-(85-148) readily assembled into amyloid fibrils. Peptide PrP-(191-205) also formed fibrillary structures although with different morphology, whereas peptides PrP-(23-41) and PrP-(217-228) did not. These findings suggest that the processing of mutant PrP isoforms associated with Gerstmann-Sträussler-Scheinker disease may occur extracellularly. It is conceivable that full-length PrP and/or large PrP peptides are deposited in the extracellular compartment, partially degraded by proteases and further digested by tissue endopeptidases, giving rise to a ~7-kDa protease-resistant core that is similar in patients with different mutations. Furthermore, the present data suggest that C-terminal fragments of PrP may participate in amyloid formation.

PrP(Sc) typing by N-terminal sequencing and mass spectrometry.

Arch Virol Suppl 2000;(16):209-16
Chen SG, Zou W, Parchi P, Gambetti P
The heterogeneity of the clinicopathological phenotype in human prion diseases is associated with the presence of the different forms of the abnormal prion protein, PrP(Sc). We have previously shown that PrP(Sc) in FFI and a subtype of familial CJD linked to the D178N mutation can be distinguished by their difference in gel mobility following proteinase K (PK) treatment. To further characterize the structural difference of PrP(Sc) in familial prion diseases, N-terminal sequencing and mass spectrometry were used to identify the protease cleavage sites in PrP(Sc) extracted from affected brains.

We found that the main PK cleavage sites of PrP(Sc) are located at residue 97 in FFI, and residue 82 in both CJD178 and a GSS subtype linked to the P102L mutation. The differential accessibility to protease in the native PrP(Sc) suggests that PrP(Sc) exist as distinct conformers in different disease states.

Membrane topology influences N-glycosylation of the prion protein

EMBO Journal, Vol. 20, No. 4 pp. 703-712, 2001
Adrian R. Walmsley, Fanning Zeng and Nigel M. Hooper
The glycosylation state of the glycosyl-phosphatidylinositol (GPI) anchored cellular prion protein (PrPC) can influence the formation of the disease form of the protein responsible for the neurodegenerative spongiform encephalopathies.

We have investigated the role of membrane topology in the N-glycosylation of PrP by expressing a C-terminal transmembrane anchored form, PrP-CTM, an N-terminal transmembrane anchored form, PrP-NTM, a double-anchored form, PrP-DA, and a truncated form, PrPGPI, in human neuroblastoma SH-SY5Y cells.

Wild-type PrP, PrP- CTM and PrP-DA were membrane anchored and present on the cell surface as glycosylated forms. In contrast, PrP-NTM, although membrane anchored and localized at the cell surface, was not N-glycosylated. PrPGPI was secreted from the cells into the medium in a hydrophilic form that was unglycosylated. The 4-fold slower rate at which PrPGPI was trafficked through the cell compared with wild-type PrP was due to the absence of the GPI anchor, not the lack of N-glycans. Retention of PrPGPI in the endoplasmic reticulum did not lead to its glycosylation. These results indicate that C-terminal membrane anchorage is required for N-glycosylation of PrP.

Glycosylation of membrane and secretory proteins involves a series of steps involving a number of oligo- and monosaccharide transferases. This process is initiated in the lumen of the endoplasmic reticulum, concomitant with protein translation and translocation by the transfer of a core N-linked glycosylation unit, Glc3Man9GlcNAc2, onto acceptor Asn residues in the tripeptide sequon Asn-Xaa-Ser/Thr. This initial step is catalysed by the enzyme oligosaccharyltransferase, which is a component of the protein translocation machinery in the ER membrane. Thus, N-linked glycosylation is considered to be a co-translational event....

As the newly glycosylated protein is transported from the ER to the cis-Golgi, trimming and addition of other sugar residues occur. N-linked glycosylation is often essential for the folding, stability, intracellular transport, secretion and function of glycoproteins. The addition of the GPI anchor to the C-terminus of membrane proteins occurs efficiently and rapidly within 1 min of translocation of the polypeptide chain. A transamidase enzyme, Gpi8p, is involved in this process, and is closely associated with the translocon apparatus in the membrane of the ER. Thus, addition of the GPI anchor is likely to occur after glycosylation of available Asn residues by oligosaccharyltransferase, and would not be expected to interfere drastically with N-glycosylation of the protein....

Our data imply that with PrP, glycosylation and membrane anchorage are co-operative processes. This is a somewhat surprising finding given that the two N-glycosylation sequons in PrP will become accessible to oligosaccharyltransferase after the N-terminal signal peptide, and before the C-terminus of the polypeptide, have been translated and translocated through the ER membrane. One possibility is that the anchorless construct is initially glycosylated and then the lack of membrane anchorage causes it to be deglycosylated. However, this seems unlikely as retention of the protein within the ER did not lead to increased glycosylation, and numerous other anchorless proteins are fully glycosylated.

Alternatively, as all the enzymes involved in glycan processing that have been characterized to date are membrane bound, membrane anchorage may affect the access of these enzymes to the protein.

However, neither of these possibilities would explain why the N-terminally anchored PrP-NTM is unglycosylated. Another possibility is that N-glycosylation occurs after the entire polypeptide chain has been translocated, such that C-terminal rather than N-terminal membrane anchorage somehow influences whether oligosaccharyltransferase recognizes and modifies the N-linked glycan sequons.

As the glycosylation state of PrP is known to influence its conversion into the infectious PrPSc isoform (Taraboulos et al., 1990; Kocisko et al., 1994; Lehmann and Harris, 1997; Nitrini et al., 1997), the present study indicates that membrane topology may influence this process. Furthermore, it has recently been reported that certain inherited PrP mutations appear to cause neurodegeneration in the absence of PrPSc, working instead by favoured synthesis of a transmembrane form of PrP (Hegde et al., 1999).

Strain typing of French nvCJD; sheep scrapie strain matches sporadic CJD

PNAS 10.1073/pnas.041490898 
CI Lasmézas,Jean-Guy Fournier,V Nouvel, HBoe*, D Marcé, ...J Ironside, Moira Bruce, D Dormont, and JP Deslys
Comment (webmaster): This is an important paper with carefully conducted research and numerous public health implications, all beautifully written up. On the whole, it strongly validates existing concerns of secondary human-to-human transmission of nvCJD, that the diagnostic signature will remain the same, even in older cases, with onsets more rapid. The data also supports further a precautionary approach on scrapie-to-human, at least for some strains.


nvCJD disease has raised concerns about the possibility of an iatrogenic secondary transmission to humans, because the biological properties of the primate-adapted BSE agent are unknown. We show that (i) BSE can be transmitted from primate to primate by intravenous route in 25 months, and (ii) an iatrogenic transmission of nvCJD to humans could be readily recognized pathologically, whether it occurs by the central or peripheral route.

Strain typing in mice demonstrates that the BSE agent adapts to macaques in the same way as it does to humans and confirms that the BSE agent is responsible for nvCJD not only in the United Kingdom but also in France.

The agent responsible for French iatrogenic growth hormone-linked CJD taken as a control is very different from nvCJD but is similar to that found in one case of sporadic CJD and one sheep scrapie isolate....

...Transmission of BSE to macaques provided the first experimental evidence as it produced a disease close to nvCJD in humans . Strain typing in inbred mice (consisting of measuring the incubation period and establishing lesion profiles corresponding to the strain-specific distribution of brain vacuolation) allows reliable identification of TSE strains.

This method, together with biochemical methods, has revealed a single phenotype for the agents of BSE and the British cases of nvCJD . Mice expressing only the bovine prion protein (PrP) were highly susceptible to nvCJD and BSE, which induced the same disease. Thus, it is now well established that BSE has caused nvCJD, probably by alimentary contamination.

In this respect, the finding of abnormal PrP labeling in the gastrointestinal tract and lymphatic tissues of orally BSE-contaminated lemurs shows that the BSE agent can infect primates by the oral route.... Unlike sporadic CJD (sCJD) and iatrogenic CJD (iCJD) linked to the administration of contaminated growth hormone extracted from human hypophyses, in nvCJD, the infectious agent seems to be widely distributed in lymphoid organs, as pathological PrP (PrPres) can be detected in tonsils, lymph nodes, spleen, and appendix even in the preclinical phase of the disease.

This raises a public health issue with regard to the risk of iatrogenic transmission of nvCJD through surgical instruments, grafts, blood transfusion, or parenteral administration of biological products of human origin. However, this risk is difficult to assess, because it largely depends on factors such as the virulence of the BSE agent adapted to primates and the efficiency of secondary transmission to humans by a peripheral route such as the i.v. one.

A further issue is whether nvCJD accidentally acquired from humans would be recognized. The latter poses the question of a phenotypic variation of the BSE agent after successive transmissions in humans: does it retain its strain characteristics, and does it induce a pathology similar to that observed in the previous host? A 9-year history of transmission of BSE to primates and mice enables us today to clarify a number of these important points.

Although BSE has mainly affected the U.K., two definite cases and one probable case of nvCJD have now been reported in France in people who have never resided in the UK.

We strain-typed the first of these cases to establish its origin. Strain typing in C57BL/6 mice of BSE, French, and British nvCJD was compared with that of BSE passaged in nonhuman primates, thus allowing us to study the effect of serial passages in primates.

Comparisons were also made with French cases of sCJD and iCJD and two strains of scrapie (one of French and one of U.S. origin). Our findings provide experimental demonstration that the same agent, namely that responsible for the cattle disease BSE, has caused nvCJD both in France and in the U.K., in line with biochemical data and with the fact that, until 1996, about 10% of the beef consumed in France was imported from the U.K.

We found that the BSE agent in nonhuman primates is similar to that causing nvCJD in humans and tends to evolve rapidly toward a primate-adapted variant. Furthermore, we showed that the strain responsible for iCJD is closely related to that of one patient with sCJD, and, more unexpectedly, that these agents were similar to the French scrapie strain studied (but different from the U.S. scrapie strain). This finding requires a cautious interpretation for several reasons, not least because of the inevitably limited number of TSE strains that can be studied by such a cumbersome method as strain typing. Nonetheless, it also prompts reconsideration of the possibility that, in some instances, sheep and human TSEs can share a common origin.

Inocula. Transmissions to C57BL/6 mice were set up from the first French patient with nvCJD, two British patients with nvCJD, BSE cattle, and macaques experimentally infected with BSE (first and second passage).... The French patient was a 26-year-old man who died of CJD in 1996. This case has been classified as vCJD on the basis of the observation of florid plaques and PrPres typing. One British nvCJD patient was a 42-year-old woman who had sensory symptoms at onset of the disease; the other was a 31-year-old man whose clinical history began with memory impairment.

Cynomolgus macaques (Macaca fascicularis) were inoculated intracerebrally (i.c.) with brain of BSE-infected cattle from the U.K. They were killed 3 years later at the terminal stage of the disease, and the brain of the youngest was used for inoculation of a second set of macaques. [The macaques could hardly be sourced in France given the contamination of monkey chow; the animals here were purchased in Mauritius. However, the status of the first inoculated macaque is not clear. None were genotyped. -- webmaster]

[Old world monkeys diverged tens of million years ago from the great ape lineage. The M. fascicularis mature prion protein differs from human at 9 locations. "Adaptation" means that the sequence of the recruiting prion protein has come to match that of the recruited polypepetide. Results are hardly transferable across all primate species based on study of one. -- webmaster]

As controls, primary transmissions were also performed from two French patients affected by sCJD and iCJD (linked to the subcutaneous injection of pituitary-extracted growth hormone), respectively, and from two Romanov sheep of the same flock at the clinical stage of scrapie....

[Their cumbersome method needs to be streamlined or abandoned: it prevented them from looking at an adequate number of controls. One wonders about its resolving power, indeed with iatrogenic = French scrapie = sporadic CJD ‚ a US scrapie, the webmaster wanted to see a whole lot more sporadic CJD and scrapie that didn't look like these. However: " One of us reported the absence of similarity between sCJD (six cases) and U.K. scrapie (eight cases) in transmission characteristics in mice (1997 paper, difficult to compare to matches here as methods may have differed).

Western blots were done but again in a way with little resolving power compared to what Prusiner's lab has done. While in vitro conversion supports scrapie to human transmission, no one could conclude from the data in this paper that the obvious scenario had in fact occurred: scrapie transmitted to human by diet with secondary passage to human through pituitary growth hormone, all with retention of strain type. -- webmaster]

Both human patients were methionine homozygotes at codon 129 of the PrP gene, like the patients with nvCJD. They had been classified as type 1 (corresponding to high molecular weight PrPres). The sheep were VRQ/ARQ heterozygotes at codons 136, 154, and 171. Scrapie appeared suddenly in this experimental Romanov flock after oral challenge with nematode parasites in 1994 . The mouse-adapted C506 M3 scrapie strain was used as a further control. It is derived from a U.S. case of scrapie in a Cheviot sheep.

Comment (webmaster): Some people would read the passage below as saying alternatively that faithful passaging of strain types is a bit of a myth:

"As far as the evolution of the BSE agent in primates is concerned, we observed an interesting phenomenon: at first passage of BSE in macaques and with vCJD, there was a polymorphism of the lesion profile in mice in the hippocampal region, with about half of them harboring much more severe vacuolation than the mice inoculated with cattle BSE. At second passage, the polymorphism tended to disappear, with all mice showing higher vacuolation scores in the hippocampus than cattle BSE mice. This observation suggests the appearance of a variant of the BSE agent at first passage in primates and its clonal selection during second passage in primates. The lesion profiles showed that it was still the BSE agent, but the progressive appearance of a "hippocampal signature" hallmarked the evolution toward a variant by essence more virulent to primates. "

Xanthinuria II: another bovine disease that humans get

Tue, 9 Jan 2001 webmaster research
While we normally think of rodents providing model diseases relevant human genes, here is a situation where a xanthinuria in cattle actually facilitated identification of a human disease, type II xanthinuria.

The availability of the nearly complete human genome, and its accessibility via the UCSC Genome Browser, has hugely accelerated disease gene discovery (the webmaster alone has pinpinted 4-5 monogenic disorders in the last months), though some 400 diseases remain only unsatisfactorily mapped and conversely housands of seemingly important genes have no disease candidates (eg, loss of normal prion or doppel function).

The xanthinurias are part of a much broader class of human disease involving either molybdenum enzymes per se (notably xanthine dehydrogenase, aldehyde oxidase and sulfite oxidase) or the biosynthesis of a remarkable organomolybdenum cofactor family based on a tricyclic pyranopterin (MPT; MoCo;MGD, sometimes an attached GDP or CTP) with a cis-dithiolene group responsible for molybdenum ligation in some enzymes

This gives rise to an unusual situation whereby spectrum of enzymes affected by a given mutation depends on exactly which step of cofactor biosynthesis is affected relative to which enzymes used the various specialized forms. Thus in humans, aldehyde oxidase and xanthine oxidoreductase use a cyanolyzable final form of cofactor, whereas sulfite oxidase does not, meaning certain mutations in cofactor biosynthesis only impact the first two enzymes.

The molybdenum cofactor family is universal in that it accounts for all biological use of molybdenum and ubiquitous in that all cellular forms of life use it. One might suppose, quite wrongly, that this biosynthesis of this complex cofactor is too complex for mammals and that it would be a vitamin -- after all, 15 genes are allocated to its biosynthesis in E.coli (even though only 10 enzymes ultimately use it).

Xanthinuria I involves mutations in XDH, meaning AOX1 is unaffected; thus allopurinol can still be oxidized but xanthine cannot. In xanthanuria II, both capabilities are lost. This class of molybdo-enxyme has a cyanolysable Mo=S ligand, unlike sulfite oxidase which has an oxygen in place of the sulfur. Xanthinuria results from a mutation in the sulferase needed to insert this sulfur in the XDH/AOX1 class of enzyme. Unlike sulfite oxidase, XOR and AOX (which are 48% paralogous, both 35 coding exons, AOX evolving more rapidly) use FAD.

The starting point for assignment of human xanthanuria II is a paper by Wanabe et al who reported a deletion causing bovine xanthinuria type II affected the structural gene for molybdopterin cofactor sulfurase (homologous to drosophila ma-1 and Aspergillus hxB), mapped to BTA 24q13.1-13.3 and human 18q12 by Amrani et al. This enzyme contain a domain common to pyridoxal phosphate-dependent cysteine transulphurases: the PP cofactor binding site as well as a conserved cysteine, the likely sulphur donor.

Now bovine chr 24 is syntenic only to human chr 18; as seen below gene order is adequately conserved in the vicinity of the bovine gene MCSU. The bovine molybdopterin cofactor sulfurase proves 80% homologous for its full length to a single-copy unannotated gene FLJ20733 on chr 18q12.2. In humans, xanthinuria type II is fairly well studied (eg, cyanolysable-restricted) but not mapped at all.

Known human molybdoenzymes and associated deficiency diseases:


chr 12q13.2
chr 2p22.3
chr 2q33.1 

chr 18q12.2
chr 6p21.3
chr 5q11
chr 14q24.1

sulfite oxidase
xanthine oxidase
aldehyde oxidase NT_005229

molybdopterin cofactor sulfurase
molybdopterin Z synthase
molybdopterin dithiolene synthase
gephyrin molybdate addition





sulfocysteinuria-neurological abnormalities
xanthinuria I-xanthine calculus-adult respiratory syndrome
candidate gene for amyotrophic lateral sclerosis ALS2

xanthinuria II
bicistronic comp. group A neonatal seizures 
bicistronic B neonatal seizures 
hyperekplexia-Kok disease-stiff man autoimmune syndrome

Cattle-Human synteny supports xanthinuria II assignment
cattle chr gene human range1 range2
MBP 24 MBP 18 q23 q23
CYB5 24 CYB5 18 q22.3 q23
PAI2 24 PAI2 18 q22.1 q22.1 chr18:62362593-62378771
GRP 24 GRP 18 q21.32 q21.33 chr18:68085096-68095649
FECH 24 FECH 18 q21.1 q21.31 chr18:69539023-69576136
*MCSU 24 FLJ20733 18 q12.2 q12,2 chr18:36393432-36529604*
TTR 24 TTR 18 q12.1 q12.1 chr18:31657661-31664605
DSC2 24 DSC2 18 q12.1 q12.1 chr18:30793638-30981584
CDH2 24 CDH2 18 q12.1 q12.1 chr18:27474523-27698341
TYMS 24 TYMS 18 p11.31 p11.22
YES1 24 YES1 18 p11.31 p11.22
MC2R 24 MC2R 18 p11.2 p11.2
ADCYAP1 24 ADCYAP1 18 p11 p11
It is straightforward to workeout the 15 exon structure of the human gene, its alternative splicing (none in 12 ESTs), tissues of expresssion (placenta, brain, prostate, colon, heart pool, breast, salivary gland), presence and alternate splicing in other mammals (49 further ESTs in mouse and cow AB036422 which have similar exon structure except for 210 bp missing in exon 8 and 5'UTR ) and genomic neighbors (DTNA-dystrobrevinA, ZNF24, FLJ10656, LIV1, FLJ10879, FLJ20733).

Are there other human enzymes that should be listed above?

-- The molybdenom cofactor is degraded to urothione (via phospho-norurothione and norurothione) and secreted in urine. The responsible enzyme(s) have never been determined; there is no known disease that results from a defect in molybdopterin catabolism. It may be possible to find this in the completed genome by developing a molybdopterin binding motif search capability.

-- No human gene is known for molybdenum uptake. In other organisms, this involves the ABC superfamily of transporters with components of this including a periplasmic binding protein ModA , an integral membrane protein ModB, and an energizer protein ModC, regulated by a repressor protein, ModE. Molybdate transport is tightly coupled to utilization in E. coli , while other organisms appear to have a molybdenum storage protein, Mop (1FR3).

-- Searching the whole human genome with tblastn of the above protein probes does not turn up any closely related genes on other chromosomes.

-- Searching the whole human genome with tblastn of molybdoprotein probes from other organisms does not turn up any closely related genes on other chromosomes.

-- New molybdenum (and tungstate) enzymes are still being discovered. For example, in E. coli there may be a whole new class of enzymes affectuating 6-hydroxylaminopurine reductase; as of 11 Jan 2001 the gene had not been identified (pers. correspondence). It would be of special interest for eukaryotes as it operates aerobically.

-- No effective ProSite search signature exists for the molybopterin binding site (unlike pyridoxal phosphate). It is unclear at this time how many distinct superfamilies exist. More promising for uptake and even storage is a molybdate binding domain seen in Mop and other proteins.

-- Although a molybdo bacterial enzyme reduces trimethylamine N-oxide to trimethylamine, trimethylamine oxidation in humans does not utilize this cofactor. A patient with trimethylamine oxidase deficiency (fish odor syndrome) had normal levels of urothione and defective FMO4 on chr1q23 which utilizes FAD and NADPH and is insensitive to tungstate.

-- Mice have a triple tandem duplication of aldehyde oxidase resulting in AOH1 and AOH2 for which there is no counterpart in humans (no EST support, blastp, or genomic tblastn support). The mice enxymes seem to have specialized to the male specific sex steroid hormone pathway.

Molybdenum metabolism in E.coli may suggest human counterparts

mog b0009 putative molybdochetalase in molybdopterine biosynthesis moaB

modF b0760_1|2 putative ABC superfamily (atp_bind),molybdenum transporter

modE b0761 transcriptional repressor of modABCD operon (molybdate uptake) and moa regulation

modA b0763_1 ABC superfamily (periplasm), molybdate transporter (1stmodule)
modB b0764_2 ABC superfamily (membrane), molybdate transporter (2nd module)
modC b0765_1 ABC superfamily (atp_bind), molybdate transporter (1st module)

moaA b0781_1 molybdopterin biosynthesis, protein A (1st module)
moaC b0783 molybdopterin biosynthesis, protein C

moaB b0782 molybdopterin biosynthesis, protein B
moaD b0784 molybdopterin biosynthesis
moaE b0785 molybdopterin converting factor, subunit 2

moeB b0826_1 molybdopterin biosynthesis (1st module)
moeA b0827 molybdopterin biosynthesis protein

mobB b3856 molybdopterin-guanine dinucleotide biosynthesis protein B GTP-binding
mobA b3857 molybdopterin-guanine dinucleotide, GMP attachment step
So far, ten E. coli molybdoenzymes have been described (all seem to use molybdopterin guanine dinucleotide, MGD:

DMSO reductase
TMAO (trimethylamine-N-oxide) reductase
biotin sulfoxide reductase (BisC) aerobic
three nitrate reductases
... NRA
... NRZ aerobic
... NAP
three formate dehydrogenases
... FDH-N
... FDH-H
... FDH-O aerobic
6-Hydroxylaminopurine (HAP) reductase aerobic

Joe Gibbs, a Top Neurological Researcher, Is Dead at 76

Sat, 10 Mar 2001 NY Times By SANDRA BLAKESLEE 
Dr. Joe Gibbs, an expert on neurological diseases who helped show that maladies like mad cow disease and scrapie are infectious rather than genetic, died on Feb. 16 at a hospital in Washington, his hometown. He was 76. The cause was a heart attack, his family said.

A career scientist at the National Institute of Neurological Disorders and Stroke, in Bethesda, Md., Dr. Gibbs ran a laboratory there specializing in a mysterious class of disorders known as transmissible spongiform encephalopathies, or T.S.E.'s.

The disease has different names in different species: mad cow disease in cattle, scrapie in sheep and goats, chronic wasting disease in deer and elk, and several forms of Creutzfeldt- Jakob disease, or C.J.D., in humans. One of the human diseases, kuru, is contracted by eating brain tissue of humans who have fallen victim to the disorder. Another, called new variant C.J.D., is acquired by eating the nervous tissues of cattle infected with mad cow disease.

In the early 1960's, Dr. Gibbs began a long collaboration with Dr. D. Carleton Gajdusek, a virologist who had spent several years in the jungle highlands of Papua New Guinea studying an unidentified disease that was killing half of all the women and children of the Fore tribe. Victims trembled, staggered, laughed uncontrollably, became progressively paralyzed with frozen, masklike smiles on their faces and then died. The Fore called the affliction kuru.

Kuru was a major medical mystery. All the classic symptoms of infection ‹ fever, inflammation, changes in white blood cells or antibody response ‹ were absent. Further, while most bacterial or viral diseases produce illness within days or hours, kuru was believed to take years (a suspicion later confirmed). As a result, most experts concluded that it must be hereditary.

But Dr. Gajdusek and Dr. Gibbs had a hunch that the cause was instead a slow virus.

The Gibbs-Gajdusek team carried out experiments to test that hypothesis. First, the researchers established the Patuxent Wildlife Center in what was then the Maryland countryside, where in 1963 Dr. Gibbs injected the ground-up brains from the bodies of kuru victims into chimpanzee brains, and waited. Two years later the animals displayed the signs of kuru.

Eventually the transmission among the Fore was traced to the practice of eating the brains of dead relatives, a ceremonial activity that had been undertaken by women and children of the tribe not long before. In 1976, Dr. Gajdusek shared the Nobel Prize in Medicine, for proving that kuru was caused by an infectious agent ‹ whose identity remains uncertain to this day ‹ rather than heredity.

"There are lots of people who felt that Joe Gibbs should have shared that prize," Dr. Richard Johnson, a neurologist at the Johns Hopkins School of Medicine who was Dr. Gibbs's friend for over 40 years, said this week.

After his success with kuru and chimpanzees, Dr. Gibbs injected ground brain tissue from the bodies of C.J.D. victims into monkeys' brains, and within a year the animals had developed the disease. Meanwhile, efforts to infect animals with brain tissue taken from patients suffering from other neurological diseases, including Alzheimer's and Parkinson's, did not work.

As with kuru, the identity of the infectious agents in other transmissible spongiform encephalopathies remains uncertain. But a few years after Dr. Gibbs's later experiments, one of his students, Dr. Stanley B. Prusiner, developed a new theory about them.

Called the protein-only hypothesis, this theory holds that diseases like new variant C.J.D. and kuru are caused when one of the body's own proteins, called a prion, misfolds to form an abnormal protein that cannot be broken down by the body. Moreover, the theory holds, the misfolded protein acts like a seed crystal, causing other, healthy proteins to misfold and bringing on destruction of brain tissue. For his work related to this hypothesis, Dr. Prusiner won the Nobel Prize for Medicine in 1997.

Dr. Gibbs often told friends that a third Nobel Prize was yet to be won for work on the mysteries of T.S.E.'s. He suggested that another molecule, perhaps an exceptionally small viral particle, might join the abnormal protein to cause the disease.

But if his suggestion is wrong and the disease arises spontaneously, he said, then T.S.E.'s can never be eradicated. Rather, they could be expected to occur in roughly one of every million people, cows, chickens, pigs, sharks and so on throughout the world forever.

Clarence Joseph Gibbs was born on Dec 10, 1924, in Washington. After graduating from St. John's College High School there, he joined the Navy, where he was on active duty until 1946 as a pharmacist's mate.

He earned his bachelor's, master's and doctoral degrees from the Catholic University of America, where he later became a trustee. He also rose to the rank of captain in the Naval Medical Reserve.

Dr. Gibbs is survived by a brother, Edward C. Gibbs Sr. of Camp Springs, Md., and two sisters, Lorraine Morrison, also of Camp Springs, and Margaret Brooks of Waldorf, Md.

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