The backwards evolution of the prion concept

Sometimes speculations leap so far ahead of evidence that they develop a detached life of their own. When popularized, unfounded beliefs can also be difficult to erase. Through much of the Middle Ages it was believed that plagues were a consequence of divine retribution. Yet most people in the USA today realize that plagues are caused by infectious organisms. Eradication of these organisms is both preventive and curative. Barrier experiments during the 1700s first proved that rats and other forms of visible life did not arise spontaneously. However, the idea of spontaneous generation of microscopic forms of life, such as bacteria and viruses, persisted until similar barrier and filtration experiments of Pasteur and others were undertaken. These showed a pre-existing “seed” or organism had to be present to generate new life. Furthermore, modern molecular genetics clearly shows infectious organisms all need nucleic acid for true replication (making many new copies), and viral and bacterial genomes can be propagated solely by using their nucleic acid (DNA or RNA) sequences. Moreover, simply modifying these sequences alters the strain-specific properties of infectious agents.

A hypothetical new form of life called a “prion” (an acronym for a protein infectious particle, later identified as the host protein “PrP”) is claimed to have remarkable features that are most reminiscent of ancient beliefs. Yet there is no direct evidence for many prion claims. These include the spontaneous generation of infectivity by a normal mammalian protein. Spontaneous conversion was a convenient belief (and legal cover) for governments dealing with practices that led to the epidemic outbreak of mad cow disease. Unlike spontaneous events that should occur with some type of mathematical frequency, the characteristics of the specific mad cow infectious agent had never been seen before. Additionally, the disease occurred only in places exposed to contaminated products (i.e., not spontaneously made by the host). Removing infected materials dramatically and abruptly diminished the UK cow epidemic. Similarly, new kuru infections disappeared after the cessation of ritual cannibalism in New Guinea.

Perhaps more disingenuous, in view of the experimental data, are the statements that TSE infectious agents, “because they are prions”, contain no nucleic acids (DNA or RNA). The prion concept was introduced in 1982 (PMID: 6801762) despite the fact that the preparations treated to abolish nucleic acids had enough DNA to code for a million human beings (reviewed in PMID: 12828865). In fact, the only two subsequent “peer-reviewed” experimental papers (from 1993 and 2005) to conclude that there are no nucleic acids of >50 bases in TSE infectious fractions are written by S. Prusiner’s prion group. Neither of these reports used modern sensitive end labeling, or established RT-PCR methods for nucleic acid detection (PMID: 16051871). Nor were controls such as enteroviral particles added to monitor nucleic acid extraction. Moreover, detailed RT-PCR methods published in 1994 (PMID: 8152913) had already demonstrated endogenous virus with >5,000 bases co-migrates with infectious TSE particles during purification. These TSE and viral particles were resistant to exhaustive nuclease digestion. It is also notable that both the R. Marsh (PMID: 1972202) and C. Weissmann groups independently uncovered nucleic acids of >300 bases in Prusiner’s purified prion preparations that were presumably free of nucleic acids. Several groups in France have also independently reported nucleic acids in their own infectious preparations (PMID: 18422484, and cited in http://www.sciencemag.org/cgi/eletters/327/5967/869#12963).

The belief that an infectious protein, rather than a covert nucleic acid, must code for the wellestablished different TSE agent strains (sporadic CJD, kuru, BSE and various scrapie isolates) excludes the objective evidence of nucleic acids in infectious preparations. Recent claims that prion protein mutates and evolves (PMID: 20044542), instead of a nucleic acid, also replays the discredited Tobacco Mosaic Virus story where a self-catalyzed crystalline protein was said to be the causal infectious agent. This Nobel Prize work simply ignored the viral nucleic acid present in the preparations that coded for the viral replication.

We continue to work on the more parsimonious concept that the fundamental infectious particle causing CJD, scrapie and other TSEs is a small virus of ~25nm, as determined experimentally by independent size, density and ultrastructural studies (PMID: 17044041; PMID: 17267596). PrP acts as a necessary host receptor or scaffold for this virus, and pathologic PrP amyloid is formed late in infection as a result of this interaction. Notably, microglia that have negligible PrP and no detectable PrP amyloid have shown very high levels of infectivity (PMID: 12368333). Thus we continue to do work on TSEs to understand this group of infectious agents and the ways they cause dementia. There are far more parts of this puzzle than the prion protein. The following table lists essential prion claims that are contradicted by experimental data. The findings continue to implicate a 25nm diameter virus as the cause of TSEs.

Claim	Experiment	Comment
1) “PrPSc (PrP amyloid, PrP-res) is proportional to titer”	FALSE	diverse data from many labs
2) “Procedures that modify or hydrolyze PrPSc inactivate prions”	TRUE	They also inactivate viruses
3) “No evidence exists for a virus-like particle”	FALSE	discrete ~25nm viral particle
4) “Transmissible particles are devoid of nucleic acid”	FALSE	all infectious preps contain long nucleic acids
5) “PrP gene mutations cause formation of PrPSc”	FALSE	toxic pathology, but no PrPSc
6) “PrP gene mutations cause transmissible disease”	not reproducible	Contamination with lab scrapie 263K
7) “Prion diversity is enciphered by PrPSc” conformation	FALSE	changing PrP-res folding has no effect on strain or titer
8) TSE “strains can be generated by passing through hosts with different PrP genes”	rarely	Strains typically keep their unique identity in different species (as BSE)
9) “No sign of an immune response to foreign agent”	FALSE	Early innate immune responses before detectable PrP-res
10) “Accumulation of PrPSc associated with pathology”	TRUE	PrPSc is a late response to infection
11) Protein X postulated to bind PrP for transmission	X not found	PrP itself not infectious; X is probably a virus
12) “Prions defy the rules of protein structure”	TRUE	possibly also of thermodynamics
13) CJD infectious agent arises spontaneously	No evidence	An idea predating evolution

TABLE I: Major arguments and the “wealth of data…for the fundamental principles of prion biology”. PrPsc the presumably “infectious scrapie or prion form” is revealed by limited protease digestion of tissue in a test tube and is equivalent to PrP amyloid, PrP-res, aggregated PrP. Tissue culture studies have shown that altering the PrP profile does not change the strain properties. The bands are tissue and cell-type specific, and unlike the infectious agent do not breed true (e.g., PMID: 15161970). Below is a picture of infected neuronal cultures and they can also show the pathological prion protein (in red) unlike mock-infected cultures (inset). Notably, despite the high level of pathological PrP in these cells they show no toxic response. Thus PrP itself may not be the ultimate or sole cause of neurodegeneration. These cells give reliable, rapid assays of the agent.

The ultrastructural 25nm particles have been found in more purified infectious TSE preparations. They do not bind PrP antibodies, unlike isolated amyloid fibrils (PMID: 17044041). These virus-like particles can also be identified in intact cells with high levels of infectivity (PMID: 17267596), and can be seen in compact arrays, like many conventional virions. An array is depicted here inside the cell and shown in a corresponding 3-D projection from different angles. These particle arrays are unrelated to the PrP amyloid fibrils in the cytoplasm identified structurally and by PrP antibodies.