Hazardous Haplotypes in sheep
Mad Cow Home or Best Links

Squalene cyclase and QW motif: relevence to prion repeat structure?
Hazardous Haplotypes in sheep
Coordinates for viewing sheep alleles threaded on mouse
Cat prion protein predicted
Predicting cow polymorphisms: bad apple bovine theory
Neurological disorder, unknown, ibex and chamois
Neuronal vacuolation in raccoons

Hazardous Haplotypes in sheep

Notes by webmaster  Last Rev. 15 Sept 1997
There are many implications of molecular phylogeny for the prion gene. Here, I consider alleles in sheep in the phylogenetic context, with a view to defining the evolutionary wild type, polymorphisms, mutations in highly variable regions of the prion gene, and mutations in crucial structural regions. The last class is one to avoid in breeding sheep. These classes are easily assigned by alignment with 75 species of vertebrates for which the prion gene has been determined (many new), using ClustalW (with default settings; gapping is not an issue here).

Sheep alleles do not correspond at all with what is found in wild animals, so their prion gene variants must be largely an artefact of domestication and inbreeding, ie these alleles have not reached their current abundance through selection or Kimura drift. The mutations are not found in the usual hotspots (loops with non-critical non-selected function) but always in essentially invariant positions that could reasonably be expected to destabilize secondary and tertiary structure, ie be prime candidates for sporadic TSE susceptibility or recruitment, though alleles don't speak directly to species barrier issues. The changes are probably moderate in terms of effect on normal function as all are common PAM substitutions (ignoring position invariance).

Of course, then there is the question of what happens when two of these bad alleles recombine to form what is surely a 'hazardous haplotype.'

When the side chain nmr coordinates are released on 8 Oct 97, we can 'thread' these alleles onto the known structure to better estimate their impacts. A lot depends on their interaction with residues not contiguous in the primary sequence. Normal sheep differ from mice in 8 positions already [below] in the 121-223 region, so the threading just of wild type sheep along would have to be accurate enough that the effect of the mutant sheep allele could be seen.

In the meantime, I located the sheep alleles and wild type differences in distinct colors on an interactive 3D image of mouse prion main chain, to at least show whether exterior or interior residues are at issue:

Animal husbandry has evidently created an unfit animal. Better to start over and breed back to the wildtype gene. One might wonder, since the bad alleles seem to occur in New Zeeland and Australia, what the adjusted incidence of scrapie might be in those countries. Imports are screened very carefully for scrapie but not genotypes; in the old days, this was not possible, now it is too late. Alex Bosser's GenBank notation for alleles is used, as amended.

The bottom line is:

SHPRP_wt    : no unusual features whatsoever 
SHPRP_M112T : destabilizing    
SHPRP_A136V : destabilizing     
SHPRP_M137T : destabilizing
SHPRP_L141F : less worrisome  
SHPRP_R154H : destabilizing
SHPRP_Q171H : destabilizing  
SHPRP_Q171R : destabilizing   
SHPRP_R211Q : possibly destabilizing, but a persistent polymorphism

>bossers_SHPRP_wt  
MVKSHIGSWILVLFVAMWSDVGLCKKRPKPGGGWNTGGSRYPGQGSP
GGNRYPPQGGGGWGQPHGGGWGQPHGGGWGQPHGGGWGQPHGGGGWGQGGSHSQWNKPSKPKTNMKHVAG
AAAAGAVVGGLGGYMLGSAMSRPLIHFGNDYEDRYYRENMYRYPNQVYYRPVDQYSNQNNFVHDCVNITV
KQHTVTTTTKGENFTETDIKIMERVVEQMCITQYQRESQAYYQRGASVILFSSPPVILLISFLIFLIVG
Sheep versus mouse (nmr was long mouse) wild type differences.
L141M: one of the sheep alleles is L141F
F143L short mouse only
Y148W
V187I
T193V
I206V
I208M
I218M
R223K
The 'wild type' sequence is indeed what is expected from other artiodactyls and ferrungulates. There is nothing noteworthy in the sheep sequence until the beginning of the STE invariant region. Threonine in position M112T is a fairly conservative substitution but definitely an anomaly. Leucine is tolerated in rat and marsupial and phenylalanine in birds in place of the mammalian norm of methionine.
KPSKPKTN T KH   SHPRP_112T 
KPSKPKTN M KH   invariant placental mammal
KPSKPKTN L KH   marsupial
KPPKSKTN F KH   consensus avian
The next change is also unfortunate: shortly after the first beta strand (which has been a fixture for 310 million years +). Here we find two fairly conservative substitutions but in a region that has never tolerated any subsitution (32/38 consecutive residues invariant back to the amniote divergence):
YMLG S V M SRP   SSHPRP_136V  
YMLG S A T SRP   SHPRP_137T 
YMLG S A M SRP   invariant placental mammal
YMLG S A M SRP   marsupial
YAMG R A M SGM   consensus avian
The change L141F in sheep and to I142M in goat are less worrisome because L141M, L141I, and L141V and I142L are seen. The region overall has seen rapid change and is not alignable with avian sequences.

The change at position 154 is bad news as well, an unprecedented change in the first helix. Arginine to histidine are both basic and the preceding pair of tyrosines can be changed to tryptophan (in birds). Curiously, the nyala (which was susceptible to BSE) has a comparable change on the opposite side.

DYED RYYH ENMYRYP SHPRP_154H
DYED HYYR ENMYRYP nyala
DYED RYYR ENMYRYP invariant placental mammal 
DYED RYYR ENMYRYP marsupial
The changes Q171H and Q171R occur along with Q171E in two other species and a deletion of this residue in birds. Changing glutamine to histidine or arginine is conservative. Otherwise, the immediate region has been stable for several hunderd million years. It has not been assigned any secondary structure.

The change at R211Q is also seen in goat. Of course, goats are susceptible to scrapie too. This a highly conserved position that only changes at the basal level of birds to another basic amino acid, lysine. There is a known CJD mutation one residue over at V213I. How can different species such as sheep and goat have the same rare allele? One possibility is that the change took place in a recent common ancestor, became a common polymorphism, won out in goats but lost out in sheep even though it persisted. This is called an unsorted lineage, where the gene tree does not correspond perfectly to the species tree. If more goat prions were sequenced, R211R might also be found.

Cat prion protein predicted

Tue, 16 Sep 1997 webmaster
The cat prion is not too difficult to predict given dog-dingo-wolf, horse-zebra, camel-lama-pig, mink-ferret, and rodent-primates. The sequence below will be accurate to the 99% confidence level [giving partial credit for alternatives. The bottom line is, it is just about impossible to get transmission to cat and mink without getting it to dog as well; there is not going to be an additional species barrier here.

>cat_hypothetical
MVKSHIGSWLLVLFVATWSDIGFCKKRPKPGGGWNTGGSRYPGQGSPGGN 
RYPGQGGG-WGQPHGGGWGQPHGGGWGQPHGGGWGQPHGGGGWGQGGGSH P G 
GQWGKPSKPKTNMKHVAGAAAAGAVVGGLGGYMLGSAMSRPLIHFGNDYE N
DRYYRENMYRYPNQVYYRPVDQYSNQNNFVHDCVNITVKQHTVTTTTKGE D K
NFTETDMKIMERVVEQMCVTQYQKESEAYYQRGASAILFSPPPVILLISL V I 
LILLIVG
I have also worked out crocodile prion sequence. This exercised me considerably -- the sauropsid taxonomy is unsettled and has to be worked out first; plus the mammals are a distant outgroup. The key idea is that the characteristic time scale associated with rate of change at each codon position has to be matched with the characteristic time scales of the divergences. In essence, the metric phylogenetic tree is used like a dirac delta function, basically fourier-transforming the time-dependent sequence function (of codon position) to properly weight the potential synaptomorphy characters. It is all downhill after that.

Predicting cow polymorphisms: the one bad apple bovine theory of BSE

 Wed, 17 Sep 1997 -- Webmaster
Some uncommon allele or mutation may have started it all, either some exceptionally scrapie-sensitive or self-nucleating. Cattle prion genes have not been sequenced as much as one would think, by SSCP or DSSG especially. As far as I can recall, the main thing that has turned up has been 5 and 6 repeat cattle. The 6 repeat has been around for a couple of million years, going by wisent, watussi, and banteng, maybe longer depending on unsequenced nilgai and cape buffalo and American bison. (See the tree relationships.)

There is an interesting if obscure phenomenon in evolutionary theory called 'lineage sorting' that has a practical application to predicting cattle polymorphisms. In lineage sorting, alleles persist across speciation events, causing the gene tree to not quite match the species tree, as alleles later variously win out. I wouldn't be a bit surprised to see some longer inserts as well. These could be the one bad apple that spoiled the barrel. Here's 8 alleles of cattle prion I would expect to find under this concept, though I am not saying, au contraire, that these are necessarily exciting for BSE nor the long-sought bad apples:

RYPPQGGG-W
RYPGQGGG-W
RYPGQGGGGW
THSQWNK
SHGQWNK
THGQWNK
GSDYED
QNSFV
(You can locate yourself unambiguously within the bovine prion gene using these fragments.)

Prp genetics in sheep and the implications for scrapie and BSE

Trends in Microbiology 1997; 5(N8): 331-4 1997
Hunter,N.
The strong links between PrP genotype and the occurrence of scrapie in sheep strengthen evidence supporting the central importance of the PrP protein in the development of transmissible spongiform encephalopathies, despite the fact that the cattle PrP gene has, so far, failed to show any association between PrP alleles and susceptibility to BSE.

From gene to organismal phylogeny: reconciled trees and the gene tree/species tree problem.

Mol Phylogenet Evol 1997 Apr;7(2):231-240
Page RD, Charleston MA
The processes of gene duplication, loss, and lineage sorting can result in incongruence between the phylogenies of genes and those of species. This incongruence complicates the task of inferring the latter from the former. We describe the use of reconciled trees to reconstruct the history of a gene tree with respect to a species tree. Reconciled trees allow the history of the gene tree to be visualized and also quantify the relationship between the two trees. The cost of a reconciled tree is the total number of duplications and gene losses required to reconcile a gene tree with its species tree. We describe the use of heuristic searches to find the species tree which yields the reconciled tree with the lowest cost. This method can be used to infer species trees from one or more gene trees.

Squalene cyclase and the QW motif

Fri, 19 Sep 1997 ...  webmaster
Tthe article yesterday in Science on the structure and function of squalene-hopene cyclase suggests that the QW motif (glutamine-tryptophan octamer repeat) possibly works in the same way structurally as the similarly spaced and composited glutamine-tryptophan motif octamer repeat of prion protein.

This could be a something big or a some big nothing.

The results of the Swiss nmr group are very likely experimentally reproducible, accurately interpreted in the context considered, and represent a lot of hard work and an important contribution to our knowledge of prion structure, but I don't believe for a minute that a structure that has been strictly conserved for 310,000,000 years is just flopping around as a random coil in the plasma waiting to be cleaved off by a protease. In this case, you would get a rate of change like a fibrinopeptide -- two orders of magnitude faster. Not observed.

The protein studied by the German group contains eight QW-sequence motifs, five also present in the paralogous oxo-squalene-cyclases. Seven of the eight motifs assume virtually identical polypeptide conformations, PDB accession code 1SQC, SWISSPROT|P33247|.

The side chains of Q and W are stacked, forming hydrogen bonds with the amino end of the adjacent outer barrel helix and with the carbonyl end of the preceding outer barrel helix, respectively. The outer ring of alpha helices is stabilized by the QW-motifs. It does not align overall to prion protein; its repeats are adjacent and downstream of the respective helix. This would be, if anything, convergence -- independent reinvention of the same short motif.

The repeats look like:

QNPDGGWG
QMADGGWG
QRPDGGWD
QLPNGDWPQ
QITVPGDWA
QKPDGSWFG
Prion protein direct repeats look like, in this light:
PQGGGGWG
QPHGGGWG
QPHGGGWG
QPHGGGWG
QPHGGGGWG
QGGTHGQW
Bird prion has a little different pattern, biphasic in proline, and a hexamer, not octamer, more N than Q and more Y than W, clearly a good rigid structure here of some kind, 18 consecutive repeats of Pxy Pzw. Should have the common ancestor reconciled with mammals shortly, with the new turtle sequences.
PSGGGWG
AGSHRQ
PSYPRQ
PGYPHN
PGYPHN
PGYPHN
PGYPHN
PGYPQN
PGYPHN
PGYPGWG
QGYNPSSG
GSYHNQKPW
How exactly are how the side chains of Q and W are stacked, parallel or anti-parallel, glutamine lying along the face of the tryptophan or along its edge? (Tryptophans themselves stack face-to-edge if you give them half a chance.) And what is hydrogen bonding to which part of which helix?

I will post here the QWs in interactive 3D; the coordinates are not on hold but still under review. The expert on QW repeats is Dr. Georg Schulz.

The issues are that:

(1) the prion protein has 3 alpha helices and a short anti-parallel beta but the QW are in a direct repeat upstream of the the secondary structure. In squalene cyclase protein, they are helix-contiguous and downstream, at least of the first helix bound. [There whole protein is ancient internal repeats.] There is another intermediate potential helix region, 109-122, so maybe helix-QW-helix-QW-helix-QW-helix-QW could work.

2. The prion protein is membrane-bound through a GPI anchor already. Normal function is unknown, large enough to be an enzyme. There are enough hydrophobic and charged residues unassigned to form a plateau, cavity, and monotopic membrane protein.

3. I wonder if several QWs would together form a periodic structure of their own, perhaps a ring or circular pore or gate or structural grid to organize helix attachment. This gets interesting because there are extra QW insertional mutations in cow, chicken, lama, and people. The latter, up to 9 extra QW regions, by itself causes CJD. It could make the pore too large or have the wrong QW at the helix, no longer stabilizing it. CJD is not a loss of function disease, it is a loss of structure disease. There is protease clipping and loss of QW region during normal turnover, which is when the disease oligomer is formed.

Structure and Function of a Squalene Cyclase

K. Ulrich Wendt, Karl Poralla, Georg E. Schulz
Science 277, # 5333, Issue of 19 September 1997, pp. 1811-1815

A specific amino acid repeat in squalene and oxidosqualene cyclases.

Trends Biochem Sci 1994 Apr;19(4):157-158 [no abstract]
 Poralla K, Hewelt A, Prestwich GD, Abe I, Reipen I, Sprenger G
>sheep_wildtype_prion
MVKSHIGSWILVLFVAMWSDVGLCKKRPKPGGGWNTGGSRYPGQGSPGGNRYPPQGGGGWGQPHGGGWGQPHGGGWGQPH
GGGWGQPHGGGGWGQGGSHSQWNKPSKPKTNMKHVAGAAAAGAVVGGLGGYMLGSAMSRPLIHFGNDYEDRYYRENMYRY
PNQVYYRPVDQYSNQNNFVHDCVNITVKQHTVTTTTKGENFTETDIKIMERVVEQMCITQYQRESQAYYQRGASVILFSS
PPVILLISFLIFLIVG

>gi|1351110|sp|P33247 squalene cyclase
HOPENECYCLASEMAEQLVEAPAYARTLDRAVEYLLSCQKDEGYWWGPLLSNVTMEAEYVLLCHILDRVDRDRMEKIRRY
LLHEQREDGTWALYPGGPPDLDTTIEAYVALKYIGMSRDEEPMQKALRFIQSQGGIESSRVFTRMWLALVGEYPWEKVPM
VPPEIMFLGKRMPLNIYEFGSWARATVVALSIVMSRQPVFPLPERARVPELYETDVPPRRRGAKGGGGWIFDALDRALHG
YQKLSVHPFRRAAEIRALDWLLERQAGDGSWGGIQPPWFYALIALKILDMTQHPAFIKGWEGLELYGVELDYGGWMFQAS
ISPVWDTGLAVLALRAAGLPADHDRLVKAGEWLLDRQITVPGDWAVKRPNLKPGGFAFQFDNVYYPDVDDTAVVVWALNT
LRLPDERRRRDAMTKGFRWIVGMQSSNGGWGAYDVDNTSDLPNHIPFCDFGEVTDPPSEDVTAHVLECFGSFGYDDAWKV
IRRAVEYLKREQKPDGSWFGRWGVNYLYGTGAVVSALKAVGIDTREPYIQKALDWVEQHQNPDGGWGEDCRSYEDPAYAG
KGASTPSQTAWALMALIAGGRAESEAARRGVQYLVETQRPDGGWDEPYYTGTASPGDFYLGYTMYRHVFPTLALGRYKQA
IERRR

>chic_prusiner
MARLLTTCCLLALLLAACTDVALSKKGKGKPSGGGWGAGSHRQP
SYPRQPGYPHNPGYPHNPGYPHNPGYPHNPGYPQNPGYPHNPGYPGWGQGYNPSSGGS
YHNQKPWKPPKTNFKHVAGAAAAGAVVGGLGGYAMGRVMSGMNYHFDRPDEYRWWSEN
SARYPNRVYYRDYSSPVPQDVFVADCFNITVTEYSIGPAAKKNTSEAVAAANQTEVEM
ENKVVTKVIREMCVQQYREYRLASGIQLHPADTWLAVLLLLLTTLFAMH

Neurological disorder, unknown, ibex and chamois

A ProMED-mail post Tue, 30 Sep 1997 08:16:04 -0400 (EDT)
From: Nicolas Troillet
The authorities of the Canton Valais (Western Switzerland) have recently been notified of an outbreak of neurological disease among wild chamois and "bouquetins" in a remote region of the Alps. Since September 15, hunters observed at least seven animals suffering from a progressive palsy affecting first the posterior limbs and leading to death in about a week. Autopsies performed until now failed to reveal any etiology but a viral encephalitis is suspected by veterinarians who are still working on this problem.

The question has been asked of a possible transmission to humans (it is the hunting season) or domestic animals (cattle might be in contact with affected wild animals). Does somebody out there have an idea for the etiology, or any suggestion about public health measures that should be undertaken?

Neuronal vacuolation in raccoons (Procyon lotor).

Vet-Pathol. 1997 May; 34(3): 250-2
Hamir-AN; Heidel-JR; Picton-R; Rupprecht-CE
Microscopic vacuolar changes in neuronal perikaryon are described in two free-ranging raccoons (Procyon lotor) from different geographic locations in the United States. Both animals were negative for rabies and scrapie-associated antigens. Microscopically, lesions were not seen in the neuropil. Neuronal vacuolations have previously been documented in brains of normal animals and in diseases such as rabies and prion-associated encephalopathies. Although experimental transmission of a spongiform mink encephalopathy has been documented in raccoons, a naturally occurring spongiform encephalopathy has not been described in this species. The presence of neuronal vacuolations in the raccoons is novel and requires further investigation to elucidate the mechanism of this phenomenon.

Mad Cow Home or Best Links