Prions

 

This lecture borrowed heavily from the following sources, which are provided to you (reading them is recommended):

Stanley Prusiner's Prix Nobel lecture

1999 JBC mini-reviews about prions:

Wickner et al. (prions in fungi)

Liebman & Derkatch (the PSI system in yeast)

Weissman (prion pathology)

 

Diseases caused by prions include:

Kuru (Fore people of New guinea)

Creutzfeldt—Jakob disease (CJD):

iatrogenic, variant, familial, sporadic

Gerstmann—Sträussler—Sheinker (GSS) disease

fatal familial insomnia (FFI)

fatal sporadic insomnia (FSI)

 

scrapie — sheep

bovine spongiform encephalopathy (BSE) — cows

transmissible mink encephalopathy (TME) — mink

chronic wasting disease (CWD) — mule, deer, elk

feline spongiform encephalopathy (FSE) — cats

 

They are similar in that they cause a variety of neurodegenerative diseases.

 

Different modes by which disease presents itself

    1. transmissible, but with very long incubation times
    2. some exhibit a genetic characteristic (genetic predisposition among certain families)
    3. some cases seem "sporadic" — no risk factors in terms of genetics or exposure to infection

 

Strange properties of prion infective agent:

    1. very resistant to UV and ionizing radiation
    2. resistance to formalin or heat
    3. very long incubation times ("slow viruses")
    4. immune system sometimes not activated against a foreign agent
    5. resistant to DNase/RNase, but sensitive to protease

 

The Prion hypothesis:

    1. the infective agent is composed solely of protein (the prion protein = PrP)
    2. this protein corresponds to a protein naturally found in the organism (in the brain) — PrPC
    3. the disease-causing form is an alternate conformation — PrPSc
    4. PrPSc can cause the production of more PrPSc, most likely by converting PrPC to PrPSc.

 

  1. properties of prions
  2. experiments to test hypothesis
  3. generalization of the concept — prions in yeast
  4. thermodynamics/kinetics/folding & chaperones

 

Properties of prions

 

  1. Species barrier: prions from one species are either unable to infect another species or the incubation time is significantly longer.
  2. Strains: Existence of different forms, from a single species, that vary in terms of incubation time, area of plaque formation, form of PrPC, etc.

 

Experiments to test prion hypothesis:

Predictions & test

 

  1. Cellular and prion forms should be different:
    a) prion form is more protease-resistant
    (often has protease-resistant core)

    b) CD & FTIR spectroscopy indicate different distributions of 2° structure:
    PrPC: rich in     a-helix, little b-sheet
    PrPSc: more     b-sheet (40%) than a-helix (30%)
  2. PrPSc (protease-resistant form) should accumulates accumulates in affected areas of brain:
    is main component of plaques — found as fibrils
  3. Species barrier should be due to differences in amino acid sequence of PrP:
    Can be broken by expressing the relevant PrP gene in the other organism (expressed hamster PrP in mice and this allowed infection by hamster prion — the infectious agent that came out of the mouse was made of mouse PrPSc, but had characteristics of hamster prion disease)
  4. Loss of PrP gene should block progression of disease:
    Mice lacking the PrP gene were made — they have little obvious phenotype, but they are now resistant to infection by prions.
  5. Expression of PrP genes in the knock-out mice changes species specificity.
  6. Chimeric PrP genes exhibit novel prion characteristics.
  7. Different strains can be generated by passaging different prions in the mouse containing only a single PrP gene — thus, the same polypeptide can adopt more than 1 alternate conformation, which explains the existence of multiple strains.
  8.  

The Conversion Process:

1) Thermodynamic or kinetic?

2) Displays conservation of strain — templated protein folding or refolding.

3) The prion form always tends to form aggregates of some kind

 

Broadening the definition:
Prions in yeast

Prions can now be defined as:

Proteins that can assume >1 conformation, one of which can cause (by stabilization and/or catalysis) the production of more polypeptides with that conformation in an entirely post-translational reaction.

The disease may be caused by loss of the "cellular conformation" (PrPC) or interference of the "prion conformation" (PrPSc).

 

2 systems have been found in bakers' yeast:

1) [URE3] —> prion form of Ure2p

(Ure2p is transciptional regulator of nitrogen metabolism.)

 

2) [PSI+] —> prion form of Sup35p

(Sup35p is eRF3.)

The phenotypes are both caused by loss of active conformation.

 

[URE3] system

Ure2p is a negative regulator of transcription of genes that encode enzymes involved in uptake and utilization of poor nitrogen sources. ure2 mutants constitutively express these genes, even in the presence of a good N source (like NH4+). [URE3] mutants show up at a higher frequency — they have the same phenotype, but it is dominant and behaves like a cytoplasmic element (i.e. non-Mendelian, indicated by the brackets).

Interestingly, the URE2 gene is required for maintenance of [URE3]. This can be shown by doing the following genetic crosses:

1) ure2 x [URE3] —> diploid has Ure phenotype
Expected because [URE3] is dominant

2) Cause diploid to sporulate —> all 4 progeny have Ure phenotype (expected — [URE3] segregates 4:0)

3) Cross each of the progeny to a URE2 strain. The resulting diploids of half of them still have Ure phenotype (thus have [URE3]). The diploids of the other half are Ure-. Thus

 

This last feature had been observed before — several intracellular viruses in yeast required nuclear genes for their maintenance, for example (the same has also been seen for mitochondrial genome). Reed Wickner, who worked on these yeast viruses, found that the ure2/[URE3] relationship did not fit this paradigm. He came up with a list of genetic criteria for a cellular prion, and found that [URE3] fulfilled them all.

 

Genetic criteria for prions

  1. Reversible curing: strains that have spontaneously lost the prion phenotype should be able to regain it again at the same relatively high rate (~10-6) as it was initially found.
  2. Overproduction causes higher frequency of prion phenotype: if the phenotype is caused by an aggregation of an alternate conformation of the protein, then higher amounts of the protein would cause a greater frequency of the critical event that starts this aggregation.
  3. Logical relationship between prion phenotype and the mutant gene phenotype: if the prion conformation is inactive, then the prion phenotype should resemble loss of (or lower activity of) the respective protein.

 

[PSI] system

Sup35p is eRF3 — it is required for eRF1 and eRF2 to terminate translation. sup35 mutants show an increased level of nonsense codon suppression by any nonsense suppressor tRNA, presumably because termination would compete less well with nonsense suppression. [PSI] mutants have the same phenotype, but it is dominant and behaves like a cytoplasmic element.

Some sup35 mutants are unable to maintain [PSI].
(Null mutants in this gene are not available, because the gene is essential.)

It was later shown that PSI fulfilled Wickner's genetic criteria for a prion.

Thus, the idea is that the prion form of Sup35p (eRF3) forms an aggregate that of inactive protein, reducing the level of active eRF3, which gives rise to the Sup phenotype — [PSI].

Interestingly, the Sup phenotype is not dominant in vitro — if one mixes translation extracts from wild-type and [PSI] cells, it behaves like wild-type (i.e. no suppression). Thus, formation of the aggregate in vitro is not fast enough to inactivate the eRF3 from the wild-type extract.

 

More work on yeast prions

  1. Interaction with chaperones: Overproduction of Hsp104 (heat shock protein — chaperone) will cure cells of [PSI].
    Explanation: Hsp104 is involved in disaggregation of proteins that aggregate during heat shock. When overproduced, it can break up the Sup35p aggregates. (However, heat shock does not cure cells of [PSI], probably because all of the denatured proteins compete with Sup35p aggregates for Hsp104.)

    Furthermore, certain mutants in Hsp104 do not allow formation or propagation of [PSI].
    Explanation: Hsp104 is involved in allowing conformational change to prion form OR it is required to break up large aggregates sufficiently that they can enter the daughter cell.
  2. In both cases (Ure2p and Sup35p), the domain that causes prion formation can be separated from the domain required for their cellular functions. The prion domains of both proteins are smaller regions at the N-terminus rich in Asn and Gln.
    Overexpression of the prion domain of Ure2p is sufficient to generate [URE3], and is required for maintenance of [URE3]; mutants lacking the prion domain can regulate transcription, but are insensitive to the presence of the prion state. (If you express them separately, the N-terminus forms aggregates, but the C-terminus remains dispersed and active.)
  3. Both Ure2p and Sup35p have been observed to be in an aggregated form in vivo, only in [URE3] or [PSI] cells, respectively (otherwise, they are evenly distributed). Also, one can observe formation of Ure2p fibrils in vitro if one seeds a solution of Ure2p with the prion domain. They resemble amyloid in appearance, and the fact that they have a lot of b-sheet character.

 

The [Het-s] system: Useful prions

The hyphal fungus Podospora will only fuse with individuals that are closely related genetically. This is not a mating event, but formation of a syncitial colony (fused cells) — not a limitation of genetic diversity, but limitation of the spread of viruses. There are ≥ 8 het genes that control the "heterokaryon incompatibility" reaction — when fusion occurs between colonies that differ in any one of these genes, the fusion rapidly degenerates and barriers are formed against future fusion events.

One of the het genes (het-s) appears to use a prion conformation in its normal role:

    1. het-s can fuse with het-s,
      and het-S can fuse with het-S,
      but het-s cannot fuse with het-S
    2. het-s exists in 2 forms:
      [Het-s] — behaves normally
      [Het-s*] — compatible with both het-s and het-S
    3. [Het-s] behaves as a non-Mendelian element, while [Het-s*] behaves as its absence. Loss of het-s gives [Het-s*] and precludes maintenance of [Het-s].
    4. [Het-s] fulfills 2/3 of the genetic criteria:
      a) It can be lost spontaneously (i.e. converted to [Het-s*]), but can also be regained spontaneously at a low frequency
      b) The frequency can be raised by overexpression of het-s gene.