Notes from March 1, 2012 – Day 1 of the prion workshop at Koch Institute

Much of the afternoon was spent discussing all the things we don’t know about PrP and prions:

  • The role of aggregates.  It is not known whether the protein aggregates formed in human prion diseases are beneficial (potentially by taking harmful prions out of circulation), harmful, or neither.  PrP can and does form multiple types of aggregates– amyloid (which is highly organized and can be stained with Congo Red) or non-amyloid (which is more disordered), and can be smooth, or form fingers/tendrils, or spikes.  But prion scientists cannot totally agree on where PrP ends up physically located in prion disease– is it in the synapse, on the membranes, etc.  It is too mobile, and our ways of trying to measure it (for instance with tagging or staining) change it too much for the readouts to be wholly believable.  (An interesting aside: most proteins inside the aggregates appear to be correctly folded).
  • The reason for late onset.  Prion diseases, and neurodegenerative diseases generally, are interesting because the proteins necessary for disease are present throughout a person’s (or other organism’s) life, yet the disease state make take the better part of one’s life to occur.  Understanding this mechanism better would help lead to ways to intervene delay onset.
  • The reason for differential expression across tissues.  It’s well known that different hereditary prion diseases have different phenotypes and impact different brain tissues.  fCJD is known for regular, amorphous aggregates, and spongiosis in the cortex, while FFI exhibits little spongiosis and sometimes no aggregates at all, and affects the thalamus most strongly, with noticeable effects in the hindbrain and brain stem as well.  GSS forms plaques with little spongiosis and affects yet other parts of the brain.  All of this is true even though these mutations affect the same gene.  Interestingly, CAG codon repeats can cause different neurodegenerative diseases depending upon which gene they are located in (for instance, in HTT they cause Huntington’s) yet the disease they cause is not necessarily observed in the tissue where that gene is most expressed.
  • The nature of sporadic prion disease.  While sporadic cases account for about 85%+ percent of human prion disease, some researchers dispute that these cases are truly sporadic and could actually represent an infectious origin that we’re unable to confirm through evidence of prions in the lymphatic system.  But since Alzheimer’s disease can occur sporadically there is no clear reason to disbelieve the possibility that prion disease could do so as well.
  • What part of the prion protein is toxic.  A recent paper from folks at Scripps finds that PrP monomers are in fact the most toxic form of PrP, while other researchers have suggested that oligomers are more toxic.  There is  no consensus on this point.

We also discussed some experimental results that cannot yet be throughly explained.  Prion-infected brain homogenates were injected into mice with and without spleens, with the following results:

injection into brain injection elsewhere
mice with spleen Prions cleared from brain, but later reappear in spleen, and then reappear in brain and kill mouse. Prions appear in spleen and then migrate to brain and kill mouse.
mice without spleen Prions cleared from brain, but later reappear in brain and kill mouse. It’s not known where they go in between. Mice seem to never become infected.

These results demonstrate the ability of the brain to clear some amounts of infectious prion material, and also show that prions can migrate elsewhere in the body and come back to infect the brain later. They also suggest a role of the lymphatic system in providing a route for prions elsewhere in the body to infect the brain (the spleen creates migratory cells). But it’s not known how the brain clears the initial dose of prions, or where they go to “hide” for a later attack.

The results may also have to do with the short-lived nature of microglia– these cells are transported out of the brain for breakdown in the spleen on a monthly-or-so cycle, at which point the spleen may become infected.

It was asked whether an FFI genotype mouse infected with CJD brain homogenate would develop FFI or CJD.  (Relatedly, the possibility of mice heterozygous for FFI and CJD alleles was raised).  This experiment hasn’t been done but similar experiments have been done with sheep and goat scrapie and the evidence is that one type of prion will “win out” over the other and there will only be one disease phenotype observed.

We also learned some things that are known about PrP:

1. Mice with PrP recover more readily from strokes than PrP knockouts, by a lot.  Knocking out PrP more than doubles the infarc volume of mice suffering strokes.  It is believed therefore that PrP must have a role either with radicals and oxygen control, and/or must function as an anti-apoptotic.  It’s also been observed that PrP expression is heightened in cancer cells, lending credence to the anti-apoptotic theory.  It is possible that some (not all) of the harmfulness of prion disease is simply do to a loss of function, as PrP is unable to perform its anti-apoptotic duties.  However, it’s also been observed that PrP knockout mice and cows do well in life– “unless something happens.”  There is also an interesting literature on the role of PrP in Alzheimer’s disease– some studies suggest that the presence of PrP (not disease PrP, just any PrP) is required for Alzheimer’s disease.  (But no human without PrP has ever been observed– in fact, no mammal of any kind naturally lacking in PrP has ever been identified).

2. In order to get prion disease you need to express PrP.  Knockout mice don’t get sick even if you inject prion-infected brain homogenate into their brains.

An “infectious unit” is defined as the smallest amount of an infectious agent that will cause an infection (I think defined specifically as two LD50s), and mice can be injected with 105 or 107 infectious units, which causes them to have onset after 180 days or 150 days respectively, and they’ll all die within one week of onset.

Prions are hard to “kill” – infectivity remains after boiling and actually appears to get stronger after cleaning with benzene, alcohol or aldehyde because these are cross-linking agents that actually stabilize the misfolded conformation. Silver is also a catalyst for conformational change and might be involved. A famous story of iatrogenic transmission of prions in 1974 involved silver electrodes “cleaned” with alcohol and formaldehyde vapor.  However, prions can be “killed” by strong acids or bases.  NaOH and bleach are both up to the job.

Prions do, however, exhibit a strong species barrier.  In the UK in the 1990s it is believed that hundreds of millions of infectious units of prions were consumed by probably hundreds of thousands or millions of people, and yet only a few hundred people got sick.  It remains to be seen whether some people are experiencing a long incubation period and might yet become sick, but it is known that it is difficult (not impossible, but difficult) for mammals to become infected with prions from a different mammalian species.  Walker Jackson’s FFI knock-in mice, which we’ve written about previously, have just 2 of their PrP amino acids “humanized”, but their PrP still differs from true human PrP in about 20 amino acids.

We also learned a great deal more about those mice.  They were created using a gene-targeted (knock-in) approach making use of homologous recombination.  The technique is described in the paper’s supplement.  This is as opposed to a random integration approach, where the new gene might be incorporated anywhere in the genome and therefore might not be expressed in the same tissues as the original gene was.

Finally, we discussed several possible next steps in working to understand, and develop treatments for, FFI:

  • Improve the use of imaging (MRI, and MRS which apparently is what people are calling NMR these days) to detect preclinical indications of prion formation.
  • Investigating ways to improve mouse models by making age of onset earlier (so that therapeutics can be tested more quickly) and less variable (so that fewer mice are needed to establish statistical significance).
  • Test the impact of dietary factors on age of onset.  (FYI- there is evidence that periodic fasting delays Huntington’s onset in mice)
  • Identify pathways related to PrP and a set of drugs that act on those pathways that could be tested for some impact on the disease (cell growth vs. apoptosis; protein synthesis and degradation / autophagy; neurotransmitters…)