Real Time Research
Mad Cow Home ... Best Links ... Search this site

Prions studied with panel of new monoclonals
FFI symposium conclusions
Prion termini in P102L GSS
What got modelled in 3D Crunch?
What did 3D Crunch do with opossum and chicken?
Genetic mechanism causing CJD insertion mutants
Yeast prion polyglutamine
Human Analogue of scrapie-responsive Gene ScRG-1

Prion termini in P102L GSS

PNAS Vol. 95, Issue 14, 8322-8327, July 7, 1998
Parch P and 11 co-authors
This is a very substantial contribution that clarifies the underlying biochemical basis for phenotypic and pathological heterogeneity in P102L GSS and possibly more broadly. [GSS is taken as the P105L, A117V, Y145stop, F198S, Q217R and an extra repeat mutations.] Basically it comes down to the neuropathology of prion diseases largely depends on the in vivo. fragment of PrP-res.

After all these years, it still was not known what prion protein fragments accumulate, or even if the wild type allele itself was implicated. After quite a bit of hard work, they conclude here that the in vivo [non protease K] termini, glycosylation states, and correlated pathology are:

8k, unglycosylated, found in all subjects with PrP-positive GSS multicentric amyloid plaques, ragged N- and C-termini:

74- , 78-, 80- , 82- to -147, -148, -149, -150, -151, -152, and -153, essentially all from mutant allele.

73: W-GQPH-GG-GW-GQPHGGGWGQGGGSQWNKLSKPKTNMKHMAGAAAAGAVVGGLGGYMLGSAMSRPIIHFGSDYED-R-Y-Y-R-E-N-M: 154

Note that the N-terminal protease is cleaving proximally to a glycine and that the C-terminal nick is non-specific to the end of the non-globular domain. These proteases are not purified or further characterized in the article. The termination mutation, Y145stop, sems to be affirmed by this work; UCSF had been hoping it would quietly go away because there is nothing left of the globular domain to form beta structure.

19k and 21k, glycosylated type I, five of the seven subjects correlated with spongiform degeneration and "synaptic" pattern of PrP deposition: 78-, 80- , 82- to full length C-terminus.

Addressing those cases where PrP-res wasn't found, they write:

"Diagnosis of prion diseases increasingly relies on immunoblot detection of PrP-res (23). In at least one GSS subtype, however, PrP-res has been reported to be undetectable, despite the demonstration of numerous PrP positive amyloid deposit (24). Our data, which show that, in GSS P102L subjects without significant spongiform degeneration, the 8-kDa fragment is the only detectable PrP-res fragment, underline the importance of the search of smaller PrP-res peptides in addition to PrP 27-30. A systematic search for small PrP-res peptides in all subtypes of GSS will be of paramount importance and should answer the question of whether PrP-res fragments (that resist a harsh PK digestion) are indeed present in all forms of naturally occurring prion diseases. "

On transmissibility of GSS,

"Thus, all subtypes of human prion diseases associated with the formation of either type 1 or type 2 PrP-res have been shown to be transmissible. In contrast, we are aware of only one case of GSS (case 7 in this series) that transmitted the disease and yet lacked significant spongiform degeneration and detectable PrP-res type 1 (in two brain samples). The disease could be transmitted to rodents but not to primates (26, 29, 30). Taken together, these data indicate that the formation in vivo of distinct truncated PrP-res fragments not only correlated with heterogeneous phenotypes, but also with other fundamental properties of prions such as infectivity. "

On what this has to do with the basis of strains,

"The reason why, in a subset of subjects with the P102L mutation, PrP-res type 1 is not detectable is puzzling. It may depend on yet-unidentified genetic factors of the host or may represent a strain-specific phenomenon. Transmission studies, aimed to determine whether brain homogenates from subjects with GSS P102L will induce heterogeneous phenotypes in the recipient syngenic animals, may provide significant clues to answer the question. Similarly, neuropatholopic examination and PrP-res typing of animals inoculated with GSS P102L that developed disease will be important to unravel the critical factors that lead to the formation of the 8-kDa amyloidogenic fragment in GSS P102L. Because all 10 cases of GSS P102L transmitted to date have induced in three different animal species a spongiform encephalopathy indistinguishable from that caused by the inoculation of homogenates from sporadic CJD subject with a typical phenotype (28-30), it seems likely that the formation of amyloid plaques and the 8-kDa PrP-res fragment are specifically linked to the presence of the mutation. The determination of whether the PrP-res that forms in the infected animals includes the 8-kDa peptide will give the definitive answer and will further elucidate the causative role of this and similar peptides in amyloid formation. "

Prion protein expression in different species: Analysis with a panel of new mAbs

PNAS Vol. 95, Issue 15,  8812-8816, July 21, 1998
Gianluigi Zanusso and 13 co-authors (received for review bys SBP on April  1, 1998
Commentary (webmaster): This is a straighforward paper with useful results and more to come. Six new monoclonal antibodies to normal prion were characterized as to epitope region, glycosylation, rogue prion, and cross-species reactivity.

More discussion (or uniform testing)is in order of 3 mAb found earlier by other researchers as well as latin binomials for species used such as the meaningless "squirrel."

I couldn't help wondering, if the first mAb was found in 1987 and the disease is costing so many billions, why are we just getting to this now? The possible useful monoclonal classes does not seem saturated with these six. Why not just sequence the squirrel? The panel could have been easily run against key species like frog and fish.

By immunizing prion knockout mice (Prnp/) with recombinant murine prion protein (PrPc), we obtained a panel of mAbs specific for murine PrPc. These mAbs can be applied to immunoblotting, cell surface immunofluorescent staining, and immunohistochemistry at light and electron microscopy. These mAbs recognize both the normal (PrPc) and protease-resistant (PrPres) isoforms of PrP. Some mAbs are species restricted, while others react with PrP from a broad range of mammals including mice, humans, monkeys, cows, sheep, squirrels, and hamsters. Moreover, some of the mAbs selectively recognize different PrP glycoforms as well as the metabolic fragments of PrPc. These newly generated PrPc antibodies will help to explore the biology of PrPc and to establish the diagnosis of prion diseases in both humans and animals.

0. 3F4 epitope between amino acids 109 and 112, does not recognize the C-terminal PrPc fragments does not detect PrPc in mouse, cow, sheep, etc. [Kascsak 1987].

0. Prionics 15B3 Nature. 1997 Nov 6; 390(6655): 74-77. epitope not contiguous in normal.

0. O'Rourke mAb F89/160.1.5 J Clin Microbiol. 1998 Jun; 36(6): 1750-1755. reacts with prion protein in tissues from sheep, cattle, mule deer, and elk with TSE, recognizes a conserved epitope on the prion protein in formalin-fixed, paraffin-embedded sections after hydrated autoclaving.

I. mAb 8H4, 8C6, 9H7, and 2F8 recognize full-length and truncated forms of PrPc; react equally with the three known PrPc glycoforms, plus forms truncated at the N terminus. epitope located in the C-terminal region of PrPc between residues 145 and 22, including 166 and 189.

II. mAb 5B2 reacts only with the full-length PrPc; binds specifically to a synthetic peptide corresponding to amino acids 23-40, immunoelectron microscopy works in neuroblastoma cells fixed in glutaraldehyde.

III. mAb 6G9 is glycosylation specific, selectively failsing to recognize the highly glycosylated form of PrPc, epitope close to N-glycosylation site at residue 181. 6G9 fails to recognize PrPc in the squirrel suggesting an amino acid substitution near 182

PrPc is distributed in the intracellular compartment with a Golgi-like distribution.

Chicken monoclonal antibodies against synthetic bovine prion protein peptide.

J Vet Med Sci 1998 Jun;60(6):777-779
Matsushita K, Horiuchi H, Furusawa S, Horiuchi M, Shinagawa M, Matsuda H
Chicken monoclonal antibodies (mAbs) were developed against bovine prion protein (PrP) peptide. Chickens immunized with bovine PrP peptide B204 (amino acid residues 204-220) coupled to keyhole limpet hemocyanin produced specific antibodies to the peptide as determined by an enzyme-linked immunosorbent assay (ELISA) using the B204 peptide coupled to ovalbumin as target antigen.

From a fusion experiment using the chicken fusion partner cell line MuH1 and immune spleen cells, 19 mAbs reactive with B204 were generated. These mAbs were subdivided into five groups based on competitive ELISA using B204 and four 10-amino acid peptides. These five groups included all combinations expected based on comparison of amino acid sequences among the five species, bovine, mouse, human, sheep and hamster, examined. These results indicate that the chicken mAb system is a suitable technique for immunological analysis of PrP in mammals.

What got modelled in 3D Crunch?

6 July 98 webmaster
Newly released Swiss 3D Crunch data provides structural data for all the prion sequences carried by SwissProt. These may help understand how normal prion 3D structure has changed over time in various lineages.

3D Crunch ran for 41 days on a a 64-processor Silicon Graphics CRAY Origin2000 server, analyzing 200,000 public protein sequences to predict 64,561 new 3D protein structures, exceeding their target of 50,000. The 3DCrunch result server provides coordinates by instant (but massive email) in layers, ie mouse reference on the bottom, result structure on top, that are convenientely viewed in the free Swiss PDB Viewer. Both 2PRP.pdb and 1AG2.pdb were used as reference structures.

The Swiss search engine returns a number of irrelent entries using 'prion' so it is best to query with the modelling template. The process produced 48 models, all known prions, notably:

P04156_C00001     3D structure of human prion
Q01880_C00001     3D structure of cow prion
P23907_C00001     3D structure of sheep prion

P51780_C00001     3D structure of opossum prion
P27177_C00001     3D structure of chicken prion
Swiss ACResiduesSpecies
P27177104-213chicken
P5178096-235PRIO_TRIVU
P0415690-231PRIO_HUMAN
P7844683-224HOMO SAPIENS
Q1521682-223HOMO SAPIENS
Q1522174-215HOMO SAPIENS
P2390793-234PRIO_SHEEP
Q0188093-234PRP2_BOVIN
P10279101-242PRIO_BOVIN
O0282570-211Odo. hemionus
O0284170-211Odo. hemionus
O1875493-234PRIO_FELCA
O190163-123PRIO_FELCA
P04273124-226PRIO_MESAU
P0492589-230PRIO_MOUSE
P1385290-231PRIO_RAT
P97895113-215Mes. auratus
Q60468no modelPRIO_CRIMI
Q60506no modelPRIO_CRIGR
P40242101-242RP1_TRAST
P4024393-234PRP2_TRAST
P4024494-235PRIO_MUSVI
P5211494-235PRIO_MUSPF
P4024582-223PRIO_AOTTR
P4024674-215PRIO_ATEGE
P4024789-230PRIO_CALJA
P4024883-224PRIO_CALMO
P4024989-230PRIO_CEBAP
P4025082-223PRIO_CERAE
P4025190-231PRIO_COLGU
P4025290-231PRIO_GORGO
Q2841990-231PRIO_GORGO
P4025390-231PRIO_PANTR
P4025490-231PRIO_MACFA
P4025583-224PRIO_MANSP
P4025690-231PRIO_PONPY
P4025790-231PRIO_PREFR
P4025897-238PRIO_SAISC
P4785293-234PRIO_ODOHE
P4992794-235PRIO_PIG
P5144689-230PRIO_ATEPA
P5211393-234PRIO_CAPHI
P7914193-234PRIO_CAMDR
P7914293-234PRIO_CEREL
Q9514575-216PRIO_CERAT
Q9517283-224PRIO_CERMO
Q9517383-224PRIO_CERMO
Q9517483-224PRIO_CERPA
Q9517683-224PRIO_CERTO
Q9520075-216Macaca sylvanus
Q9521189-230PRIO_RABIT
Q9527075-216PRIO_THEGE

What did Crunch do with opossum and chicken?

Output for structure of marsupial prion:

SEQALI P51780 1 NWGQGGYNKW KP-DKPKTNL KHVAGAAAAG AVVGGLGGYM LGSAMSRPVI SEQALI 12PRP 1 GQGGGTHNQW NKPSKPKTNM KHMAGAAAAG AVVGGLGGYM LGSAMSRPMM SEQALI 11AG2 1 GLGGYM LGSAMSRPMI SEQALI ****** ********.. SEQALI P51780 s ss SEQALI 12PRP s ss SEQALI 11AG2 s ss SEQALI SEQALI SEQALI P51780 50 HFGNEYEDRY YRENQYRYPN QVMYRPIDQY SSQNNFVHDC VNITVKQHTT SEQALI 12PRP 51 HFGNDWEDRY YRENMNRYPN QVYYRPVDQY NNQNNFVHDC VNITIKQHTV SEQALI 11AG2 17 HFGNDWEDRY YRENMYRYPN QVYYRPVDQY SNQNNFVHDC VNITIKQHTV SEQALI ****. **** **** **** ** ***.*** ..******** ****.**** SEQALI P51780 sss SEQALI 12PRP hhhhhh hhh sss hhhhhhhh hhhhhhhhhh SEQALI 11AG2 hhhhhhh hhhh sss hhhhhhhh hhhhhhhhhh SEQALI SEQALI SEQALI P51780 100 TTTTKGENFT ETDIKIMERV VEQMCITQYQ AEYEAAAQRA Y SEQALI 12PRP 101 TTTTKGENFT ETDIKIMERV VEQMCTTQYQ KESQAYYDGR R SEQALI 11AG2 67 TTTTKGENFT ETDVKMMERV VEQMCVTQYQ KESQAYY--- - SEQALI ********** ***.*.**** ***** **** * .*

Output for structure of chicken prion:

SEQALI P51780 SEQALI 12PRP hhhhhh h hhhhhhhhhh hhhhhhhhhh hhhhhhh SEQALI 11AG2 hhhhh h hhhhhhhhhh hhhhhhhhhh hhhhhhh SEQALI P27177 1 SSGGSYHN-- QKPWKPPKTN FKHVAGAAAA GAVVGGLGGY AMGRVMSGMN SEQALI 12PRP 1 GQGGGTHNQW NKPSKP-KTN MKHMAGAAAA GAVVGGLGGY MLGSAMSRPM SEQALI 11AG2 1 GLGGY MLGSAMSRPM SEQALI ***** .* ** SEQALI P27177 sss SEQALI 12PRP sss SEQALI 11AG2 sss SEQALI SEQALI SEQALI P27177 49 YHFDSPDEYR WWSENSARYP NRVYYRDYSS PVPQDVFVAD CFNITVTEYS SEQALI 12PRP 50 MHFGNDWEDR YYRENMNRYP NQVYYRPVDQ YNNQNNFVHD CVNITIKQHT SEQALI 11AG2 16 IHFGNDWEDR YYRENMYRYP NQVYYRPVDQ YSNQNNFVHD CVNITIKQHT SEQALI ** . * * ** *** *.**** *. ** * * ***. . . SEQALI P27177 sss SEQALI 12PRP hhhhh hhhh sss hhhhhhh hhhhhhhhhh SEQALI 11AG2 hhhhhh hhhhh sss hhhhhhh hhhhhhhhhh SEQALI SEQALI SEQALI P27177 99 IGPAAKKNTS EA SEQALI 12PRP 100 VTTTTKGENF TE SEQALI 11AG2 66 VTTTTKGENF TE SEQALI . ..* . SEQALI P27177 SEQALI 12PRP hhhhhhh hh SEQALI 11AG2 hhhhhh hh

Origin of extra prion repeat units

webmaster draft document last modified 13 August 98
The prion repeat region experiences mutations in which the repeat unit itself is the unit of genetic change. Deletions of a single unit are common polymormphisms both in humans and other species and do not seem linked to CJD. Insertions are much more varied, ranging up to insertion of 9 extra units and commonly include associated point mutations. Exhaustive Medline searches, foreign-language journal searches, back searches from article bibliography, and correspondence with specialists produced the following set of insertion mutations worldwide with distinct pedigrees, current to 13 August 98:

PatternSlipsExtraTotal BPModificationFamilyCitation
1 2 2.......................3 4
00 5 0---wild type+Kretzschmar 1986
1 2 2.....................2 3 4
11 6 24---French 'GF'+LaPlanche 1995
1 2 2 3...................X X 4
22 7 48X=R2aUS Allentown +Goldfarb 1993
1 2 2 3 .2 .............2 2 3 4
24 9 96---French 'JC'+LaPlanche 1995
1 2 2 3..2 3..............2 3 4
24 9 96---US 'Hay'+Goldfarb 1991
1 2 2 ...2 2 ...........2 2 3 4
24 9 96---British+Campbell 1996
???
24 9 96---Japanese-Isozaki 1994
1 2 2 3..2 Y............2 2 3 4
2510120Y=R3gUS 'Kel'+Goldfarb 1991
1 2 2 X..2 X ...........2 2 3 4
2510120X=R2aUS+Cervenakova 1995
1 2 2 3 .2 2 ...........2 2 3 4
2510120---US/Ukrainian+Cochran 1996
1 2 2 ...2 3 2 Y........2 2 3 4
3611144Y=R3gBritish 'A'+Poulter 1992
1 2 2 3..2 Y 2 Y......... 2 3 4
3611144Y=R3gEnglish+Nicholl 1995
1 2 2 ...2 2 2 2.. .....2 2 3 4
3611144---Basque+Capellari 1997
1 2 2 Y .2 2 Y .........2 2 3 4
3611144Y=R3gJapanese+Oda 1995
1 2 2 Y .2 2 Y .........2 2 3 4
3611144Y=R3gUS+Cervenakova 1995
1 2 Z 3..2 3 2 3 .......2 Y 3 4
3712168Z=R2c,Y=R3gUS/N English 'Ald'+Brown 1992
???
3712168---Japanese-Tateishi 1991
1 2 2 Y..3 2 2 2 2 .....2 2 3 4
4813192Y=R3gDutch 'A'+van Gool 1995
1 2 2 3..2 2 2 2 2 .....2 2 X 4
4813192X=R2aBreton 'Che'+Goldfarb 1992
1 2 2 3..2 Y X 2 2 2 Y... 2 3 4
4914216X=R2a,Y=R3gBritish+Owen 1992
1 2 2 3 ..2 3 Y 2  2 3....2 3 4
4914216X=R2a,Y=R3gGerman+Kraseman 1995
Legend: + full text examined; annotated references are given below. The only duplication, for 6x=144, is shown in red; 19 distinct variants have been reported. The Japanese data will be available shortly -- more duplication would markedly affect estimates of ultimate number of variants. Note these families are10% of total data and independent pedigrees. The number of slips is calculated as per a model. Modified repeats , their terminology, and slippage ambiguity zones are defined in the following graphic.

Mechanism by which insertion mutations are generated

The genetic mechanism proposed here is (multi-stage) replication strand slippage. Daughter strand slippage is the likliest mechanism for insertions; template strand slippage for deletions. The imbalance of reciprocal recombinants (notably at length one), complexity of some variants, and generational stability are implausibly implemented via unequal recombination. Deletions of one repeat vastly outnumber insertions of one repeat (despite roughly equal ascertainment and lack of lethality): recombination requires equal numbers of reciprocal recombinants.

However, some as-yet-unimagined mechanism could be responsible. For that reason, I develop below a number of specific predictions of the replication slippage model. If other mechanisms are developed -- provided they make differing predictions -- additional data could discriminate amongst them. While slippage has been proposed in many genetic settings, multi-stage slippage within a single event is apparently novel. Of course, the human prion gene repeat region provides a favorable mileau for slippage and interest in CJD has enabled an intensive screen.

The basic idea is that as DNA replication procedes through the prion gene, the daughter strand experiences slippage facilitated by complementarity to an earlier template repeat. Repeats, by their nature, are still complementary when misaligned by an integral multiple of 24 bp (the principal ambiguity zone is shown in dark blue above -- slippage need not respect repeat boundary end points). Exonuclease editing 3' to 5' is available to remove mismatches -- partial editing produces chimeric repeats. Progress of the replication fork may be affected by a unique feature of the prion gene, the conserved hairpin C whose function at the level of mRNA or DNA remains unknown.

Hairpin C takes up most of R1:
AACCGCTACCCACCTCAGGGCGGTGGTGGCTGGGGGCAG
 N  R  Y  P  P  Q  G  G  G  G  W  G  Q

When replication resumes with slipped complementarity in place, the result is extra repeats. (See graphics and animations.) Longer complex repeats require multiple rounds of stalling, slipping, editing, and resuming replication until finally replication run-out occurs. This takes place within a single event (ie, not over generations). Note that the direction of replication in the opposite sense from transcription fits stalling as the fork reaches the hairpin; alternately, replication in the same sense as transcription fits hairpin snap-back after the replication fork passes through the hairpin, causing slippage.

Observations on insertion events

One odd feature of the data on insertions is that nearly every mutational event has resulted in a different outcome; complex events are as well represented as simple ones. (There is moderate detection bias against shorter insertions because usually dementia cases are sequenced, which favors longer repeats.) Though solid statistics are hard to come by, repeat insertions differ from repeat deletions. With deletions, 5 classes are seen with one predominating and some signs of type saturation. Point mutations are created similarly in the slippage model of deletions; only R2c is seen (along with fusions).

Multi-stage slippage is progressive. That is, the first round of extra repeats strongly influences the second round, and so on, because it is the earlier round that must hybridize to template before strand synthesis can continue. This feature of the process drives iterated internal repeat patterns, eg R 1 2 2 3 2 3 2 3 4. The template, of course, never changes.

Note that CJD insertional disease, unlike polyglutamine repeat disorders, shows no indication of 'anticipation.' That is, a given insertional event is stable over generations: long repeats do not become longer. There is further no data supporting unequal recombination or gene conversion in progeny. The reason for this is probably the rarity of these events: an initial event does in fact provide more fodder as more repeats equals more opportunity for subsequent slippage events, but these still remain quite unlikely given the small numbers involved. Thus a weak version of anticipation is predicted here: the signature is a wing of the pedigree wherein descendants have even more insertions (or less -- deleted repeats -- for that matter).

There is every reason to doubt that the present set of known insertions is close to saturating what is possible. If 190 or 1900 pedigrees were known, instead of 19, the list of distinct insertions would surely be longer. [There is no data from India, Africa, South America, or mainland Asia.] Yet the possibilities are far from arbitrary concatenations of wildtype repeats. While specific single historical events are outside the reach of experimental methods, it is plausible in modelling to take each as the simplest possible event, ie, by invoking the minimal necessary number of slippage events possible, as shown in column 2 of the above table.

Note that the initial and terminal repeats R1 and R4 are only peripherally involved; repeats seen are invariably internal ones, involving multiples of R2 and R3. This is predicted by the slippage model to hold for subsequent data, for both insertions and deletions, in all species. Roughly, replication begins normally (initial run), then the first slippage occurs as enough potential for this is acquired. Once slippage has been set in motion, multiple slippages are enabled. The event usually terminates in a 'run-out' through the distal repeats of the form R2 R2 R3 R4. The process does not allow for extra R1 or R4.

The largest ambiguity zone is comprised of the last 6 codons of R1 and all of R2 and R2', for 66 base pairs. (This number varies by species.) A slippage of 24 bases anywhere within this zone, not necessarily in formal register with repeat starts, results in perfect complementarity. However, R3 and R4 have substantial potential for misalignment as well, especially if 1-2 mispairings are tolerated in the secondary ambiguity zone.

Point variants are generally not separate events; here they are coupled to the extra repeat generation mechanism. All variants are transitions; however very little of the potential for neutral third codon change is realized: the same changes occur over and over again in different families with varying repeats. The traditional notation of the point variants is potentially misleading: R2a might equally well be called R3c or an R23 or R43 chimera; its interpretation might be context-dependent. R3g might equally well be called R2c' or an R32 reverse chimera. R2c has no explanation other than an uncoupled point transition. It has been reported only once and is unlikely to represent sequencing error. In the allele collection, repeat R2a occurs 5 times, R3g 12 times. R2c occurs again in deletions in two pedigrees.

Even though single unit deletions [leaving R1234] are common polymorphisms affecting 1-2% of the population, the case cannot be made that any of the extra repeat insertions began from this initial state, not surprising since only 20 pedigrees are known. [Cases are known of a point mutation, eg D178N, compounded with a deletion allele.] Further, there are no cases with an insertion on one allele and a deletion on the other (which in analogy to M129V might slow onset). The signature of a slippage upon a deletion would be recognizable in some instances: R123234 for example could only come from R1234. While a prediction of the slippage model, such events could be especially rare if four repeats has less tendency for slippage than the longer wildtype five repeats.

Disentangling slippage from editing

It is advantageous to disentangle strand slippage per se from point modifications (which arise from 'optional' editing). That is, does the observed collection of insertions collapse to a simpler set of plain slippages if point modifications are ignored? Because the point modifications are ambiguous (ie R2a might drop to R2 or R3 depending on context) this is model-dependent.

The origin of R2a is shown below in excruciating detail:

(to be continued)

How many different insertion mutations will eventually be found?

So far, 19 different insert mutations have been found in 20 kindreds. That is, only in 1 case has the insert proved to be one already known from an earlier family. Suppose that:
a. There is no bias in ascertainment, eg, an insert of one repeat is as readily detected as an insert of 9 repeats. (This was not true in the early days of sequencing because of PCR differential amplification, plus the 9 repeat is favored in atypical dementia or CJD sampling pools.)
b. Each insert is present at the same frequency in the human population (despite substantial differences in event complexity, this seems to fit the data).

The question is then, given these initial results and assumptions, how many different classes of insertion mutants will ultimately be found?

This is no different than fishing in a lake with N species all present in large number. If N=19, we would certainly not expect to catch all 19 species with only 20 castings. Viewing it as the birthday problem, there is a 50-50 chance in a group of 20 people of two with the same birthday (on a planet with N=280 days per year). Such an extreme result has only a 5% chance of occurence when N=70.

The best single estimate of N is fairly soft and critically dependent on the occurences of repeats, so the range of reasonable N is quite large. There is the issue of robustness in regards to the next data points. But the bottom line is safe: there must be many dozens of repeat insert variations that have not been detected so far.

At a fixed number of extra repeats, there can be various possibilites:

1 2 3 4 5 6 7 8 9  number of extra repeats
1 1 0 3 3 4 2 2 2  number of independent pedigrees in each class
? ? ? ? ? ? ? ? ?  number of possible mutational classes at each length
0 0 0 0 1 0 0 0 0  number of known independent duplications
It is hard to say from this whether, say, 3 extra repeats is forbidden or whether some cap exists on maximum repeat length. However, we are not seeing 12 or 15 or 39 extra repeats: the data are clustered in the 1-9 range centered on 4 extras, suggesting that more than 12-13 extra repeats will be quite rare.

The repeat insertion collection plus hairpin C (R1) together predict the direction of replication through the prion gene (not necessarily the same as the direction of transcription, of course).

Multi-stage replication slippage: While replication slippage is a well-known mutational mechanism in all life kingdoms and organelles, multi-stage slippage within a single genetic event is apparently novel and not documented before (though difficult to search the literature for). Deletions occur by a dual mechanism in the template strand and have possible outcomes mostly limited by the ambiguity zones. (In polyglutamine CAG disorders, multi-stage slippage may occur but is harder to discern because of ambiguity and events in successive generations.)

Notice a certain upside bias to the data: it may not be realized initally that widely separated families with the same mutation may ultimately prove to be in the same pedigree. Parentage is not always what it seems: there is adoption and children sired outside of marriage. These three effects can only serve to reduce the adjusted incidence of de novo insertion events giving rise to the same mutation. Since there is really no data except for Europeans and Japanese, the best scenario for a bona fide duplication is a repeat found in both populations.

It is instructive to 'collapse' the table of mutations to see the level of duplication when modified bases are ignored. This is somewhat ambiguous, but the 5X mutations could coalesce, as could two of the 6X, 7X, and 9X. The motivation would be to separate slippage events from editing events. Pure slippage provides a lower bound that is still quite high but showing signs of saturation. The idea here is to determine a set of event generators and to describe specific insertions allowed by their iteration. Examples of generators are:

IR: initial run, usually 122 or 1223
RO: final run-out, usually 2234, requires slippage into hairpin C
R3g: slippage of R3 with partial edit
R2a: slippage of R4 with partial edit

The main event: predicting the final list of allowed and forbidden mutations.

The idea is to test the predictive content of the multi-stage replication slippage proposal by furnishing a model that generates a list of allowable and non-allowable mutations. If different hypotheses predicting different lists were on the table, these alternatives become experimentally discernable. At the current rate of study, it could be 15 years before sufficient numbers of cases could be collected; large western and Japanese pedigrees have probably mostly been found already.

The list generator here is based on the notion that replication runs can be identified (equivalently, slippage event boundaries identified). This is inferred from collinearity, complementarity, 5' to 3' replication, the master template order, and visual inspection of the data. Example: 1 2 2 3 2 3 2 3 2 Y 3 4 is readily broken into runs.

Going through the full set of known mutations, a rather limited set of unique runs is obtained. These may well be a full saturated set of allowable runs even thought the global mutations are far from saturated. The set of allowable runs is then taken as generators of the set of allowable mutations.

The point mutations R3g and R2c are treated somewhat differently, as potential modifiers of certain runs. These are modelled as arising from partial editing of slight mismatches caused by slippage to imperfect complementarity. The data indicate that the R3g and R2c processes -- which occur over and over -- may represent a saturated set of the main modifiers.

The list of possible mutations is then written as strings of generators and modifiers. It is easiest to first generate unmodified strings and then have modifiers operate on these objects if proper substrates.

Note that the independent Japanese mutations could profoundly impact estimates of the total number of possible mutations if they come up as duplicates. Duplication seems more likely at 4X since there are far fewer possibilities than at 7X. Duplications can be weighed probabalistically after the list of allowed mutations is established.

Oda et al raise an interesting issue in comparing clinical and pathological differences between English and Japanese lineages of slightly 6x=144 inserts, R1222323g2234 and R1223g223g2234, both M129M E219E: can different phenotypes really be attributed to differences in the primary structure of the prion gene? Note that these are silent mutations affecting only the gene and mRNA; all 6x=144 repeat patterns are the same at the protein level. But even here, the repeat region may cleaved off by endopeptides and not even be present in the rogue conformer. Though appropriate antibodies are readily available, few investigators have determined the amino terminus of plaque prion or even whether wildtype allele participated; the amino terminus is still present at least in some situations [GSS: Kitamoto T et al Brain Res 545:319-321 (1991)]

The incidence of repeats mutation depends on its definition and the diligence with which pedigrees are unravelled. If a mutation occurred in 1794 , giving rising to 221 living descendants today (roughly half of whom will be affected), should this count as 1 event or 110? The former is more useful for purposes of understanding the molecular biology underlying the event; the latter for partitioning contemporary cases of CJD. Several insertions exhilbit no family history, attributable to a de novo event, lack of older records, low penetrance, or short innocuous repeat length. These cases include the Japaneses 4x and 7x cases, the 4x cirrhosis case with untestable illegimate father, and the English 9x case.

There is a trend to describe extra-insert cases with the acronym, OPRI (octapeptide repeat insertions). Previously, some cases had been lumped with GSS, which is already an unfortunate mixture of unrelated alleles. No comparable notation has been used for seemingly neutral deletions, though OPRD might serve. It would be far better if each molecular variant carried its own name, eg, it is a major irritant to other researchers if a Medline abstract mentions only OPRI or GSS without specifying R1222a2a34 or P102L. Haplotypes should also be given, ie GSS P102L 5Rdel V129M K219E. (Within large pedigrees where every individual is not genotyped, the disease-carrying chromosome is unlikely to have undergone recombination whereas codon 129 and 219 have a good chance of varying through outbreeding in European and Japanese lineages, respectively. (to be continued)

Insertion alleles in other species

Repeats in 80 species are tabulated in a separate archive. Eight species have single extra inserts; seven have deletions. All examples involve a single repeat unit. In some cases, the extra insert is a polymorphism, with other individuals exhibiting the number of repeats expected from phyllogenetic considerations. Note that each species has a distinct ambiguity zone and that the slippage event has occurred at various places, eg X43 denotes an insertion after the third repeat.
length genus species common name
X43 Gazella subgutturosa goitered gazelle
X34 Lama glama lama
X34 Saimiri sciureus squirrel monkey
X23 Bison bonasus wisent
X23 Bos primigenius taurus watussi
X23 Bos taurus bovine, long
X23 Tragelaphus strepsiceros greater kudu
X21 Giraffa camelopardalis giraffe

D34 Theropithecus gelada gelada baboon
D34 Macaca sylvanus barbary macaque
D34 Cercopithecus aethiops green monkey/grivet
D34 Cercopithecus dianae dianae monkey
D34 Cercocebus aterrimus mangabey
D23 Pongo pygmaeus orangutan
D12 Ateles geoffroyi black-handed spider monkey

Acknowledgements

I thank M.Kubo NIAH, Japan, T. Kitamoto, L. Cervenakova, and LG Goldfarb for helpful correspondence and difficult-to-obtain documents.

References:

+ Goldfarb LV, Cervenakova L, Brown P, Gajdusek DC
In Transmissible Subacute Spongiform Encephalopathies pg 425-431
I Court, H Dodet eds.  Elsevier, Paris 1996
Table II was the starting point for the update presented here.

+Kretzschmar H.A., Stowring LE, Westaway D, Stubblebine WW, Prusiner SB, DeArmond SJ
Molecular cloning of a human prion protein cDNA
DNA 5:315-324(1986)
Human prion first sequenced.

+Laplanche JL, Delasnerie-Laupretre N, Brandel JP, Dussaucy M, Chatelain J, Launay JM
Two novel insertions in the prion protein gene in patients with late-onset dementia.
Hum Mol Genet. 1995 Jun; 4(6): 1109-1111, no abstract.
1X=24, R122234 in patient GF, 4x=96, R1223322234 in patient JC, Paris

+Goldfarb LG, Brown P, Little BW, Cervenakova L, Kenney K, Gibbs CJ Jr, Gajdusek DC
A new (two-repeat) octapeptide coding insert mutation in Creutzfeldt-Jakob disease.
Neurology. 1993 Nov; 43(11): 2392-2394.
2X=48, R1222a2a34 1 family, NIH, Bethesda.

-Dementia [Japanese] 1994 8:363-371
Isozaki, E et al.
CJD presenting as frontal lobe dementia associated with a 4X=96 base insertion.
4X=96 article in Japanese, details unavailable

+Goldfarb LG, Brown P, McCombie WR, Goldgaber D, Swergold GD, Wills PR, Cervenakova L, Baron H, Gibbs CJ, Gajdusek DC
Transmissible familial Creutzfeldt-Jakob disease associated with five, seven, and eight extra octapeptide coding repeats in the PRNP gene.
Proc Natl Acad Sci U S A. 1991 Dec 1; 88(23): 10926-10930.
4X=96 R122323234 hepatic cirrhosis Hay, 5X=120 R122323g2234 American Kel, 6X=144 R1222323g2234 large English A family, 7X=168 R122c3232323g34 American Ald family, 8X=192 R122322222222a4 French Che family; 1 family each, screen of 535, recombination model, NIH, Bethesda

+Campbell TA, Palmer MS, Will RG, Gibb WR, Luthert PJ, Collinge J
A prion disease with a novel 96-base pair insertional mutation in the prion protein gene.
Neurology. 1996 Mar; 46(3): 761-766.
4X=96 R122222234, 1 case, London. Not same as hepatic 4X case.

+Cochran EJ, Bennett DA, Cervenakova L, Kenney K, Bernard B, Foster NL, Benson DF, Goldfarb LG, Brown P 
Familial CJD with a five-repeat octapeptide insert mutation. 
Neurology. 1996 Sep; 47(3): 727-733
5X=120 R122323g2234, 1 case of American Kel family of Ukranian origin, Chicago

+Am J Hum Genet (1995) 57 A 209. abstract, article never appeared 
L.Cervenakova, LG Goldfarb, P Brown, K. Kenney, EJ Cochran, DA Bennett, R. Roos, DC Gajdusek
Three new PRNP genotypes associated with familial CJD. [Abstract only]
5X=120 Kel family,  5X=120 >R122a22a2234 new;
6X=144 same sequence as Japanese, different pedigree.  all families 129 met 
 
+Poulter M, Baker HF, Frith CD, Leach M, Lofthouse R, Ridley RM, Shah T, Owen F, Collinge J, Brown J.
Inherited prion disease with 144 base pair gene insertion. 1. Genealogical and molecular studies.
Brain. 1992 Jun; 115( Pt 3): 675-685.
6X=144 R1222323g2234, 1 case from 4 families dating to 1794, London.

-Nicholl D, Windl O, de Silva R, Sawcer S, Dempster M, Ironside JW, Estibeiro JP, Yuill GM, Lathe R, Will RG
Inherited CJD in a British family associated with a novel 144 base pair insertion of the prion protein gene.
J Neurol Neurosurg Psychiatry. 1995 Jan; 58(1): 65-69.
6X=144, R122323g23g234 1 case not earlier British kindred, Crumpsall, UK.

+Capellari S, Vital C, Parchi P, Petersen RB, Ferrer X, Jarnier D, Pegoraro E, Gambetti P, Julien J
Familial prion disease with a novel 144-bp insertion in the prion protein gene in a Basque family.
Neurology. 1997 Jul; 49(1): 133-141.
6X=144 R12222222234 non-congophilic, 1 family, Basque.

+Oda T, Kitamoto T, Tateishi J, Mitsuhashi T, Iwabuchi K, Haga C, Oguni E, Kato Y, Tominaga I, Yanai K.
Prion disease with 144 base pair insertion in a Japanese family line.
Acta Neuropathol (Berl). 1995; 90(1): 80-86.
6X=144 R1223g223g2234 M129M E219E, Japanese lineage, Ibaragi Prefecture, ~1907, not congopohilic.

-Clinton J, Forsyth C, Royston MC, Roberts GW
Synaptic degeneration is the primary neuropathological feature in prion disease: a preliminary study.
Neuroreport. 1993 Jan; 4(1): 65-68.
6X=144, 2 cases, London.

-Tateishi J.
[Recent advances in the research of Creutzfeldt-Jakob disease (CJD) and Gerstmann-Strussler syndrome].
Rinsho Shinkeigaku. 1991 Dec; 31(12): 1306-1308. Review. Japanese.
7X=168, 1 case in Japanese woman, Kyushu.

+Brown P, Goldfarb LG, McCombie WR, Nieto A, Squillacote D, Sheremata W, Little BW, Godec MS, Gibbs CJ, Gajdusek DC
Atypical CJD in an American family with an insert mutation in the PRNP amyloid precursor gene.
Neurology. 1992 Feb; 42(2): 422-427.
7x=168 R122c32323223g34 American Ald family of northern English origin, NIH, Bethesda

-Dementia [Japanese] 1994 8:380-390
Mizushima, S et al.
A case of presenile dementia with a 7X=168 base pair insertion
7X=168 article in Japanese, details unavailable

+van Gool WA, Hensels GW, Hoogerwaard EM, Wiezer JH, Wesseling P, Bolhuis PA.
Hypokinesia and presenile dementia in a Dutch family with a novel insertion in the prion protein gene.
Brain. 1995 Dec; 118( Pt 6): 1565-1571.
8X=192 R1223g322222234, 1 Dutch case, different from other 8X case, Amsterdam.

-Goldfarb LG, Brown P, Vrbovska A, Baron H, McCombie WR, Cathala F, Gibbs CJ Jr, Gajdusek DC
An insert mutation in the chromosome 20 amyloid precursor gene in a GSS family. 
J Neurol Sci. 1992 Sep; 111(2):189-194. 
8X=192 R122322222222a4 congophilic, 1 French-Breton Che family

+Krasemann S, Zerr I, Weber T, Poser S, Kretzschmar H, Hunsmann G, Bodemer W
Prion disease associated with a novel nine octapeptide repeat insertion in the PRNP gene.
Brain Res Mol Brain Res. 1995 Dec 1; 34(1): 173-176.
9X=216 R1223233g22a23234, 1 case, Gottingen.

+Owen F, Poulter M, Collinge J, Leach M, Lofthouse R, Crow TJ, Harding AE
A dementing illness associated with a novel insertion in the prion protein gene.
Brain Res Mol Brain Res. 1992 Mar; 13(1-2): 155-157.
9X=216 R122323g2a2223g234 1 case, Manchester, UK.

Redundant, preliminary, further study same kindred, clinical, reviews:

+Collinge J, Brown J, Hardy J, Mullan M, Rossor MN, Baker H, Crow TJ, Lofthouse R, Poulter M, Ridley R
Inherited prion disease with 144 base pair gene insertion. 2. Clinical and pathological features.
Brain. 1992 Jun; 115( Pt 3): 687-710.
6X=144 R1222323g2234, the large family, London.

+Owen F, Poulter M, Shah T, Collinge J, Lofthouse R, Baker H, Ridley R, McVey J, Crow TJ
An in-frame insertion in the prion protein gene in familial Creutzfeldt-Jakob disease.
Brain Res Mol Brain Res. 1990 Apr; 7(3): 273-276.
6X=144 R1222323g2234, start of big kindred, Middlesex, UK.

+Owen F, Poulter M, Collinge J, Leach M, Shah T, Lofthouse R, Chen YF, Crow TJ, Harding AE, Hardy J.
Insertions in the prion protein gene in atypical dementias.
Exp Neurol. 1991 May; 112(2): 240-242.
6X=144 in 4 related families, 9x=216, 5 insertions found in 101 atypical dementias, details reported again in separate papers.  Middlesex, UK

+Owen F, Poulter M, Lofthouse R, Collinge J, Crow TJ, Risby D, Baker HF, Ridley RM, Hsiao K, Prusiner SB
Insertion in prion protein gene in familial Creutzfeldt-Jakob disease.
Lancet. 1989 Jan 7; 1(8628): 51-52, no abstract.
6X=144 one family, preliminary version of paper.

+Duchen LW, Poulter M, Harding AE
Dementia associated with a 216 base pair insertion in the prion protein gene.
Brain. 1993 Jun; 116( Pt 3): 555-567.
9X=216 R122323g2a2223g234 clinical aspects of Owen 1992, London

-Watanabe R, Duchen LW
Cerebral amyloid in human prion disease.
Neuropathol Appl Neurobiol. 1993 Jun; 19(3): 253-260.
9X=216 probably Duchen case R122323g2a2223g234, 1 case, London.

-Goldgaber D, Goldfarb LG, Brown P, Asher DM, Brown WT, Lin S, Teener JW, Feinstone SM, Rubenstein R, Kascsak RJ
Mutations in familial CJD and GSS syndrome.
Exp Neurol. 1989 Nov; 106(2): 204-206.
-----, screen found no cases of previously described insert, probably Owen's 6x=144.

+Collinge J, Palmer MS
Molecular genetics of human prion diseases. 
Philos Trans R Soc Lond B Biol Sci. 1994 Mar 29; 343(1306): 371-378. Review. 
Reviews, new 4x=96 discussed

+Brown P, Goldfarb LG, Cathala F, Vrbovska A, Sulima M, Nieto A, Gibbs CJ, Gajdusek DC 
The molecular genetics of familial Creutzfeldt-Jakob disease in France. 
J Neurol Sci. 1991 Oct; 105(2): 240-246.
Reviews and has table of inserts and terminology

+Kenney K, Brown P, Little B
Insert mutation in Creutzfeldt-Jakob disease.
Neurology. 1995 Jul; 45(7): 1428, letter, no abstract.
2x=48 R1222a2a34, same family, claiming sporadic CJD not insert

+Vital C, Gray F, Vital A, Parchi P, Capellari S, Petersen RB, Ferrer X, Jarnier D, Julien J, Gambetti P
Prion encephalopathy with insertion of ocatpeptide repeats: the number of repeats determines the type of cerebellar deposits
Neuropath App Neurob 24: 125-130 1998
study of deposits in Basque 6x [non-congophilic] and Breton 8x cases [congophilic].

Uncertain:

-Neuropathol Appl Neurobiol 1998 Apr;24(2):125-130 Prion encephalopathy 
with insertion of octapeptide repeats: the number of repeats determines 
the type of cerebellar deposits.

-Neuropathol Appl Neurobiol 1993 Jun;19(3):253-260 Cerebral amyloid in 
human prion disease. Watanabe R, Duchen LW 
Clinical and neuropathological features of 216 base-pair insertion 

-Am J Pathol 1992 Aug;141(2):271-277 The primary structure of the prion 
protein influences the distribution of abnormal prion protein in the 
central nervous system.Kitamoto T, Doh-ura K, Muramoto T, Miyazono M, 
Tateishi JDepartment of Neuropathology, Kyushu University, Fukuoka,  
Insertional polymorphism, a pointmutation in codon 102 or 117/129, and codon 129 
(Val129) result in plaque-type PrP accumulations.

-Am J Hum Genet 1991 Dec;49(6):1351-1354 Presymptomatic detection or 
exclusion of prion protein gene defects in families with inherited prion 
diseases.Collinge J, Poulter M, Davis MB, Baraitser M, Owen F, Crow TJ, 
Harding AE
After counseling, PrP gene analysis was performed in two families with a 144-bp insert

A critical role for amino-terminal glutamine/asparagine repeats in the formation and propagation of a yeast prion.

Cell 1998 Jun 26;93(7):1241-1252 
DePace AH, Santoso A, Hillner P, Weissman JS
Comment: This is a remarkable paper worth a close look.

DePace et al. sequenced 41 yeast prion mutants after screening 45,000 colonies, finding that all of them clustered in the short asn/gln repeat region of yeast prion. "This work differs fundamentally from previous genetic studies of mammalian amyloid disorders that have relied on identification of naturally occurring mutations which accelerate disease. Such mutations typically increase the propensity to form amyloids by destabilizing the native state and thus do not report directly on the requirements for amyloid formation."

They found that simple test tube rotation profoundly accelerated conversion kinetics, as detected by Congo red binding, with seeded reactions going to completion in as little as 45 min and unseeded reactions within 180 min, just as in sickle cell. Prion researchers have not tried this technique, which presumbably gently aligns fibrils. They also can follow aggregation in vivo using green fluorescent protein and replace the N-terminal Gln/Asn region with pure polyGln.

Their dominant-negatives were interpreted not as Protein X chaperone effects but as heteropolymers with WT protein that fail to effectively promote conversion of either WT or mutant protein. Mutants can be recruited into WT aggregates in vivo; in vitro WT and AS17R amyloids are mutually defective in seeding each other's conversion.

Finally they have huntingtin protein going in yeast with pathogenically expanded polyGln repeats fused to GFP forming aggregate.

Two points of potential confusion: in yeast, the action is concentrated in the Sup 35 repeat region. This corresponds to the poly-gln of the various repeat amyloid disorders. However, this region does not correspond to the prion repeat region even though the composition is similar. Mammalian repeats can be deleted with retention of disease; the action is in 106-126 which is not asn/gln rich. Secondly, extra-repeat CJD insertions are not analogous to extra CAG repeats. So while yeast also form congophilic cross-beta, the closer analogy is to Huntington's disease, SCA etc.. -- webmaster The yeast [PSI+] factor propagates by a prion-like mechanism involving self-replicating Sup35p amyloids. We identified multiple Sup35p mutants that either are poorly recruited into, or cause curing of, wildtype amyloids in vivo. In vitro, these mutants showed markedly decreased rates of amyloid formation, strongly supporting the protein-only prion hypothesis.

Kinetic analysis suggests that the prion state replicates byaccelerating slow conformational changes rather than by providing stable nuclei.

Strikingly, our mutations map exclusively within a short glutamine/asparagine-rich region of Sup35p, and all but one occur at polar residues. Even after replacement of this region with polyglutamine, Sup35p retains its ability to form amyloids. These and other considerations suggest similarities between the prion-like propagation of [PSI+] and polyglutamine-mediated pathogenesis of several neurodegenerative diseases.

Virtually all denatured proteins have a strong propensity to form amorphous aggregates. This aggregation is largely driven by the association of hydrophobic regions that are normally buried in the native structure (for review, see Jaenicke and Seckler, 1997 ). In contrast to these more frequent disordered aggregates, some proteins form ordered aggregates called amyloid fibrils. Amyloidogenic proteins show no obvious sequence similarity, nor do their native folds resemble one another. Yet despite this diversity amyloid fibrils appear to share a similar architecture.

Amyloid fibrils are cross ß-sheet structures in which the individual ß strands run perpendicular to the long axis of the fibril, whereas the faces of the ß sheets extend parallel to it (Lansbury et al., 1995 ; Sunde and Blake, 1997 ). These fibrils are roughly 100 Å in diameter and are typically composed of 4-6 interwoven protofilaments. Recent X-ray diffraction studies suggest that the ß sheets composing the individual protofilaments are twisted, resulting in a helical conformation propagating along the amyloid axis (Sunde et al., 1997 ; see, however, Lazo and Downing, 1998 ). The ordered structure of amyloids allows them to bind the dye Congo red at regular intervals, leading to a characteristic red-green birefringence under polarized light. Despite these shared structural features, electron microscopy (EM) and atomic force microscopy (AFM) (Harper et al., 1997a ) analyses indicate that there are ultrastructural variations (e.g., differences in the number and packing of the protofilaments) between the amyloids. It is also possible that individual protein monomers retain partial native structure within the amyloid (Liu et al., 1998 ).

To date, 20 proteins have been found to form amyloids associated with human disease. These include the mammalian prion protein, PrP, the infectious protein implicated as the causative agent of transmissible spongiform encephalopathies (see Caughey and Chesebro, 1997 ; Prusiner et al., 1998 ). PrP forms amyloid fibrils that, at least under some conditions, are associated with the infectious agent. Similarly, some non-prion neurodegenerative diseases, such as Alzheimer's and Huntington's disease, also involve amyloid formation. In Huntington's, as well as several other neurodegenerative diseases, amyloid formation appears to be caused by expansion of a polyglutamine (polyGln) tract (e.g., Paulson et al., 1997 ; Scherzinger et al., 1997 ). Finally, systemic amyloidoses result from the aggregation of a number of proteins such as lysozyme and transthyretin (TTR) (Wetzel, 1997 ).

Despite active research, many basic questions about the conversion from native state to amyloid fibril remain unanswered. For example, little is known about what stabilizes amyloid structures. In particular, how sensitive is amyloid formation to changes in primary sequence? Are amyloids, like amorphous aggregates, principally stabilized by hydrophobic interactions? The mechanism of propagation of infectious prion diseases and its relationship to that of noninfectious amyloidoses is also poorly understood (Harper and Lansbury, 1997 ; Prusiner et al., 1998 ). For example, what is the nature of the rate-limiting step in de novo formation of amyloids, and how do preformed fibrils accelerate this process?

Lastly, our understanding of the role of cellular factors, such as molecular chaperones, in promoting formation of amyloids is incomplete (Chernoff et al., 1995 ; Kaneko et al., 1997 ). In particular, do in vitro conversion reactions using purified proteins accurately reproduce the propagation of disease or infection in vivo?

Sup35p is composed of three domains . The N-terminal prion determining domain (PrD) is necessary and sufficient for propagation of aggregates. Deletion of this domain allows Sup35p to remain soluble even in [PSI+] cells (Paushkin et al., 1996 ). Conversely, fusion of the PrD to GFP confers aggregation of the fusion protein in a [PSI+]-dependent manner

We have developed a genetic screen to identify PrD mutants that are defective in amyloid formation and propagation. This screen is based on the fact that such mutants will remain soluble even in a [PSI+] yeast cell that contains WT Sup35p aggregates, thereby conferring an antisuppressor (ASU) phenotype The ASU mutants were then examined for their ability to cause permanent conversion to a [psi-] state. Such "curing" alleles are referred to as Psi No More (PNM) mutants We screened 15,000 colonies for ASU phenotypes and 30,000 colonies for PNM phenotypes. In this pool, we estimate that every point mutant is represented 5- to 10-fold. In total, 28 ASU and 13 PNM mutants were sequenced.

Unexpectedly, all ASU and PNM mutants contained changes in a short region at the extreme N terminus of the PrD between residues 8 and 24 (Figure 2A). The failure to identify mutations outside of this region was not due to lack of diversity in the original mutant pool, as a number of mutations that did not cause the ASU phenotype were found throughout the PrD. Nor was it because the residues outside this region were unnecessary for PrD function: deletion of residues 53-124 of the PrD resulted in an ASU phenotype (data not shown), and a previously described chromosomal PNM mutant maps outside this region

Moreover, all of the mutations occurred in Gln or Asn residues with the exception of a single Gly and a single Ser mutant. Virtually all mutations resulted in a change to charged amino acids. Particularly striking is the predominance of Gln or Ser to Arg mutations.

WT PrD-GFP converts from a diffuse fluorescence pattern in [psi-] cells (Figure 3A) to a punctate pattern in [PSI+] (Figure 3B), making it possible to monitor the aggregation state of PrD fusion protein in vivo. Each of the ASU and PNM mutants was fused to GFP and put under control of the inducible CUP1 promoter. Use of an inducible promoter minimizes secondary effects that prolonged expression of mutants might have on the solubility of the WT Sup35p. In contrast to WT PrD-GFP, which had uniformly punctate patterns in [PSI+] cells, each mutant led to a mixed population of fluorescent patterns: in individual cells, the fusion proteins either appeared entirely soluble (Figure 3C), in discrete foci (Figure 3D), or more commonly in both (Figure 3E).

Rotation also caused a dramatic increase in the conversion kinetics, as detected by Congo red binding, with seeded reactions going to completion in as little as 45 min and unseeded reactions within 180 min . Similar enhancement in the kinetics of polymerization is observed when hemoglobin S is subjected to rotation forces

Taken together, these observations indicate that even after replacement of the N-terminal Gln/Asn region with a polyGln stretch, the PrD retained its ability both to form new aggregates and to be recruited into existing aggregates.

This work differs fundamentally from previous genetic studies of mammalian amyloid disorders that have relied on identification of naturally occurring mutations which accelerate disease. Such mutations typically increase the propensity to form amyloids by destabilizing the native state and thus do not report directly on the requirements for amyloid formation

Two classes of mutants that lead to increased levels of soluble protein were identified: the first (ASU) inhibited incorporation into aggregates without irreversibly preventing propagation of WT amyloids, and the second (PNM) resulted in curing of the [PSI+] state in addition to its ASU phenotype.

Even more striking is the recently sequenced SUP35 gene from Candida albicans, which contains 55 Gln residues in its putative PrD domain, including long stretches of pure Gln repeats near the N terminus. These observations suggest that the amino acid composition of a protein, as much as its exact primary sequence, determines its amyloidogenicity.

. Our dominant-negative Sup35p alleles could also result from altered recognition by a molecular chaperone such a HSP104. Alternatively, our Sup35p mutants might form heteropolymers with WT protein that fail to effectively promote conversion of either WT or mutant protein. Consistent with this proposal, GFP analysis reveals that the mutants can be recruited into WT aggregates in vivo. Moreover, in vitro WT and AS17R amyloids are mutually defective in seeding each other's conversion.

A prediction of the self-replicating amyloid model is that a cell will undergo a stochastic conversion event, after which newly made protein will be recruited into amyloid fibrils. Consistent with this, we found in a related set of studies that huntingtin protein containing pathogenically expanded polyGln repeats fused to GFP is initially soluble when expressed in yeast but eventually forms aggregates that appear to propagate stably (A. S. and J. S. W., unpublished data). We are currently using the in vitro assay to directly examine the propagation of huntingtin amyloids.

Structure and Replication of Yeast Prions

10 July 98 Cell
 Vitaly V. Kushnirov and Michael D. Ter-Avanesyan
...A Role for the Hsp104 Protein
The chaperone Hsp104p plays an important role in propagation of the [PSI+] determinant. Increased levels of Hsp104p may cause [PSI+] elimination, but surprisingly, the lack of Hsp104p also cures [PSI+] (Chernoff et al., 1995 ). Based on these results, two models have been suggested for the role of Hsp104p in propagation of the [PSI+] state. According to the first model:

"Hsp104p is required because it promotes a conformational `transition' state in Sup35 that facilitates its folding into the prion-like [PSI+] structure. When Hsp104p concentrations are too high, however, the transition state conformers generated by Hsp104p are too broadly dispersed to promote assembly into a substructure that is competent for association with preexisting [PSI+] elements" (Lindquist, 1997 ). Alternatively, we proposed that Hsp104p cleaves Sup35pPSI+ aggregates into smaller pieces, which is necessary for their stable inheritance in cellular divisions (Paushkin et al., 1996 ).

The demonstration that Sup35p can form amyloid-like fibers in vitro suggests a similar fibrous structure for the Sup35pPSI+ aggregates in vivo. This provides additional support for the second model, which we will discuss further below.

Hsp104p was not required for the polymerization of Sup35p into fibers in vitro in contrast to the strong requirement for Hsp104p for the [PSI+] propagation in vivo. This suggests that Hsp104p is not essential for the Sup35p prion conversion, but is required for certain aspects encountered only in vivo, such as, for example, the inheritance of the prion state. It was demonstrated that Hsp104p dissolves aggregates of heat-denatured protein (Parsell et al., 1994 ). Overexpression of Hsp104p in [PSI+] cells caused an increase in levels of soluble Sup35p and a gradual decrease of aggregated Sup35p (Paushkin et al., 1996 ). This suggests that Hsp104p may act on Sup35pPSI+ complexes in the same way as it acts on denatured protein aggregates. Probably, Sup35pPSI+ aggregates show some properties of denatured protein, in contrast to other cellular macrostructures and microfilaments not recognized by Hsp104p. The action of Hsp104p on Sup35pPSI+ fibers at any random position would simply cut them into smaller pieces.

This may occur through refolding of one or several Sup35p molecules to a soluble conformation at a particular site. The "destructive" function of Hsp104p toward Sup35p fibers may be balanced by their ability to grow by the accretion of soluble monomers. Overexpressed Hsp104p may disrupt the balance and dissolve all Sup35p aggregates. However, at moderate levels Hsp104p would only facilitate the propagation and inheritance of prion aggregates due to the following considerations.

First, since Sup35p filaments can grow only at their ends, the overall polymerization rate is proportional to the number of these ends. The cuts by Hsp104p could create new free ends, thus accelerating polymerization (Figure 1).

Second, it is possible that the growth surfaces of the aggregate would become blocked due to fortuitous interactions with other proteins or proteolytic fragments of the same protein. The polymerization would be continued only if new growth faces would be created by Hsp104p cleavage at internal sites of the fiber.

Third, the cleavage of Sup35p aggregates by Hsp104p would be required for the inheritance of aggregates in frequently dividing yeast cells. Without it the aggregates could grow in size, but their number would not increase, which would eventually preclude their segregation to daughter cells in cellular divisions.

Thus, although Hsp104p is not essential for the Sup35p prion-like conversion, it could increase the overall Sup35p polymerization rate and ensure the segregation of Sup35pPSI+ aggregates in cellular divisions. It should be noted that the proposed mechanism of Hsp104p action is only applicable to filamentous aggregates and is unlikely to work in the case of globular aggregates, which could not be cut so easily into smaller pieces.

The acceleration of fibril formation by fragmentation was observed in several cases for both amyloid and nonamyloid fibrils. For example, it was found that the polymerization of purified Sup35p was dramatically accelerated simply by rotation of the samples during incubation (DePace et al., 1998 ). The acceleration was explained by shearing of the Sup35p fibrils by hydrodynamic forces. This appears to be the most likely explanation, even though such forces in a rotating tube should be fairly mild.

Hsp104p is the only chaperone capable of disaggregating heat-denatured proteins in yeast (Parsell et al., 1994 ). Since Hsp104p is essential for the [PSI+] maintenance, one might assume that Hsp104p could be the only or the most efficient enzyme capable of breaking filamentous aggregates and that this function might not be substituted by other cellular enzymes, such as other chaperones or proteases. ...

Characterization of the Human Analogue of a Scrapie-responsive Gene

J Biol Chem 1998 Jul 17;273(29):18015-18018
Dron M, Dandoy-Dron F, Guillo F, Benboudjema L, Hauw JJ, Lebon P, Dormont D, Tovey MG
We have recently described a novel mRNA denominated ScRG-1, the level of which is increased in the brains of Scrapie-infected mice (Dandoy-Dron et al. (1998) J. Biol. Chem. 273, 7691-7697). The increase in ScRG-1 mRNA in the brain follows the accumulation of PrPSc, the proteinase K-resistant form of the prion protein (PrP), and precedes the widespread neuronal death that occurs in late stage disease.

In the present study, we have isolated a cDNA encoding the human counterpart of ScRG-1. Comparison of the human and mouse transcripts firmly established that both sequences encode a highly conserved protein of 98 amino acids that contains a signal peptide, suggesting that the protein may be secreted. Examination of the distribution of human ScRG-1 mRNA in adult and fetal tissues revealed that the gene was expressed primarily in the central nervous system as a 0.7-kilobase message and was under strict developmental control.

FFI symposium conclusions

Medline article group of 21 July 98
Gambetti P, et al. Conclusions of the symposium. Brain Pathol. 1998 
Jul; 8(3): 571-575.

Pocchiari M, et al. Recent Italian FFI cases. Brain Pathol. 
1998 Jul; 8(3): 564-566. No abstract available.

Budka H. Fatal familial insomnia 
around the world. Introduction. Brain Pathol. 1998 Jul; 8(3): 553. 

Velayos JL, et 
al. Afferent projections to the mediodorsal and anterior thalamic nuclei 
in the cat. Anatomical-clinical correlations. Brain Pathol. 1998 Jul; 
8(3): 549-552. 

Parchi P, et al. Molecular pathology of fatal familial insomnia. Brain 
Pathol. 1998 Jul; 8(3): 539-548. 

Dorandeu A, et al. Neuronal apoptosis in fatal 
familial insomnia. Brain Pathol. 1998 Jul; 8(3): 531-537. 

.Montagna P, et al. 
Clinical features of fatal familial insomnia: phenotypic variability in 
relation to a polymorphism at codon 129 of the prion protein gene. Brain 
Pathol. 1998 Jul; 8(3): 515-520. 

Prusiner SB. The prion diseases. Brain Pathol. 
1998 Jul; 8(3): 499-513. 

Giese A, et al. Role of microglia in neuronal cell death in 
prion disease. Brain Pathol. 1998 Jul; 8(3): 449-457. 

McLean CA, et al. Comparative 
neuropathology of Kuru with the new variant of Creutzfeldt-Jakob 
disease: evidence for strain of agent predominating over genotype of 
host. Brain Pathol. 1998 Jul; 8(3): 429-437. 

Abstract snippets:

On the basis of twenty-one kindreds and three cases from uninformative families, the Symposium has confirmed that fatal familial insomnia (FFI) is genotypically and phenotypically distinct and, likely, the third most common inherited prion disease. The genotype is D178N plus M129. D178N plus V129 could be considered a form of CJD. A distinctive feature of the FFI PrPres is the underrepresentation of the unglycosylated PrPres form.

A condition lacking the D178N mutation and pathologically identical to FFI has been reported. Presence of sleep, autonomic and endocrine abnormalities needs to be demonstrated to identify this condition as a sporadic form of FFI.

The normal allele further modifies the FFI phenotype determining patient subpopulations of 129 homozygotes and heterozygotes: disease duration is generally shorter, insomnia more severe and histopathology more restricted to the thalamus in the homozygotes than in the heterozygotes. The allelic origin of PrPres fails to explain this finding since in both cases FFI PrPres is expressed only by the mutant allele.

Mad Cow Home ... Best Links ... Search this site