Prion creation theory myth checks out -- sort of
Griffith -- the other papers
Key 1967 paper of Alper et al.
Earliest papers of Prusiner

It turns out that Griffith was indeed close to the target with his second mechanism (below) -- it still sounds fairly good 30 years later. Most amusingly, Griffith himself does not take credit for this idea, but clearly attributes it to a 1957 paper of the Penroses {Nature #4571 p1183 1957], who are said to have elaborated on a 1922 paper by A Gratia (Brit Med J ii 296 1922). It can quickly be seen from a PubMed search that Griffiths' idea had no any follow-up or support of any kind.

The Penrose paper is a short letter with no citations. Wooden blocks of two types in a one-dimensional rack assemble into "dimers" when shaken, but only if and as directed by the configuration of dimer it is seeded with. The dimers in the example cannot form higher order aggregates or dissociate. The 1922 journal is not readily available.

The Gratia paper is part IV of a set of discussions on the topic of 'bacteriophage.' It consists of a few arguments regarding why bacteriophages need not be living organisms. One argument involves an analogy between the spread of fire and the reproduction of living organisms, and the other (the probable source of the 'prion hypothesis' citation) discusses how a small fraction of thrombin produced in one test tube can be used to cause plasma in another tube to clot, after which a small fraction of the second product can be used to clot another tube, ad infinitum. While one can see how this discussion is part of the intellectual history of the prion hypothesis, it does not suggest that Gratia orininated the concept.

Griffith was clearly not familiar with the biochemistry of protein oligomers and diseases of protein aggregation, which were already well-developed subjects in 1967. Looking back at 1967 undergraduate biochemistry texts, we can confirm that Griffith's reaction chains, thermodynamic equations, and potential barrier theory of dimer conformational change were commonly taught to college students in that year. Griffith's free energy driving force, for example, is exactly that of a coupled ATPase.

This is a key point, so worth repeating: all three of Griffith's mechansims were unoriginal -- all standard ideas covered in 1967 protein chemistry textbooks. Griffith's work was original only in the sense that he was not familiar with the scientific literature outside his field.

Now Griffith should be credited more often than he is for writing a prescient mini-review article on protein 'self-replication' with possible application to the scrapie mechanism of replication. [The protein-only nature of the agent he attributes to Pattison and Alper.] And batting one for three (on the models he found in the literature) is better than most theorists do -- make that one for twenty, looking at his other papers. Experimentalists would say, tell us which one of the twenty ideas is the good one.

Prusiner clearly and explcitly cites the key paper of Griffith as reference 32 in a 1982 paper, Science 216 136-144 1982 . Griffith is cited in the fourth section, 'Hypothetical Structures of the Scrapie Agent,' in a long review article, but for protein-only, not for mechanistic details. The passage reads:

...Hypotheses on the chemical structure of the scrapie agent have included: sarcosporidia parasite, "filterable" virus, small DNA virus, replicating protein [Griffith cited], replicating abnormal polysaccharide within membrane....

It can also be seen that Griffith published in short order eight other wildly speculative papers on neurobiology [DNA ticketing theory of memory, neural organization underlying conscious thought, etc .. see below], none of which ever went anywhere. It is abundantly clear that he was not what a mathematician would consider a mathematician; rather it is applied mathematics/bioengineering centered on process control theory. None of this bears on the merits of the 1967 scrapie paper.

Another minor myth -- that Prusiner's group came across the Griffith paper by juxtapositional accident as Bolton was xeroxing the paper just ahead of it -- checks out. That would be the eminently forgettable article in Nature 215 1041-43 1967 'Nature of the Scrapie Agent' by RA Gibbons and GD Hunter (agent is an altered 3D structural membrane pattern). You cannot xerox the last page without also xeroxing the first column of three of "Self-replication and Scrapie' by JS Griffith of the Math Dept., Bedford College, London.

There does not exist an earlier paper by Prusiner in which he cites the 1967 Gibbons and Hunter paper but not the adjacent Griffith paper -- see table below. Prusiner's group has published 300 peer-reviewed experimental papers on prions. True, there is a mind-numbing repetitiveness in the introduction/history to prion papers from all authors, as no one can be bothered to actually look up the history and each safely copies the cites from the last, but this is true for every subject in molecular biology and can't be taken as a conspiracy to short-change Griffith.

Oops, I meant short-change Gardia. Penrose, and Lehninger.

It would make more sense to critique Prusiner -- or even better, Griffith -- for not citing the sickle cell hemoglobin papers -- they had a huge mass of experimental backing and are much more germaine to rogue prion formation than a packet of shotgun speculation and guesswork .

But it would make even better sense not to critique Prusiner's bibliographic practises at all -- on the whole, they are very thorough; in the worse case scenario, they are comparable to everyone else's. Credit has to mainly go to whoever establishes the experimentally-demonstrated correct sequence of events -- and we're not quite there yet.

And the devil is always in the details: why the alpha to beta shift, which residues are critical and why, can this component of the species barrier be predicted, can the rogue conformer be differentially assayed, how critical is Hsp 70, where does the change occur, are CLDs integral, how do we intervene, and so on. You won't find even a wiff of any of this in Giffith.

Prusiner's problem is not Griffiths but the fact that there simply isn't anything paradigm-breaking about prion replication in the final analysis. It all comes about via the central dogma of molecular biology: DNA to RNA to protein to oligomer. The gene is very ordinary in structure and sequence. Prion protein undergoes a reversible quaternary conformational shift, but so do the vast majority of known proteins. Stability of the recruited state is a matter of energy levels and context.

So the only potentially interesting aspects are the generality of the beta structure, the normal role, the origin of toxicity, and the mechanism of transmissional propagation (which could be boring if attributable to the GPI lipid anchor migrating up neural processes). The only unusual aspect is the net result of confluencing so many aspects, each fairly unremarkable in itself, into a single protein. Poor solubility, lack of assays, and lab hazard conspired to make it hard to work out.

Gunther Stent used to say that each subject in science progressed through its romantic, academic, and pedantic phases. Prion research is transitioning into the academic phase in 1997.

So what is really in the text of this Giffith paper?

Paragraph 1 reviews the data of 1967 (agent mol wt ~20k, probably no nucleic acid, seen in goats, sheep, rats, mice, hamsters). At least three distinct ways in which proteins might self-replicate are said to be possible so "there is no reason to fear that the existence of a protein agent would cause the whole theoretical structure of moledcular biology to come tumbling down."

Paragraph 2-8 deal with the first protein self-replication mechanism. This is an unoriginal variant of 1967 bacterial operon theory a la Monod and Jacob, involving self-regulation of gene expression. A mutated control region of the scrapie gene does not allow the scrapie agent to be expressed, unless an exogenous scrapie agent can come in and act as inducer for itself or via a small metabolite that it makes.

Paragraph 9-14 deal with the second protein self-replication mechanism. It is adapted from a proposal by Roger and LS Penrose, Nature 179 1183 1957, which in turn implements the 1922 idea of A Gratia (Brit MedJ ii 296 1922). This involves monomers, dimers, trimers, tetramers, associations, dissassociations, normal conformers, favorable free energies, conformational change catalyzed by oligomers, an isozyme-like theory of scrapie strains, and talk of 'condensation nuclei" and new conformations formed on "templates." The reaction chains, equations, and potential barriers could be found in minor variation in any undergraduate biochemistry textbook of 1967, such as Lenninger, albeit not in the context of scrapie..

Paragraph 15 deals with the third protein self-replication mechanism. Prion prion induces an antigen that happens to have the same covalent structure as itself. Griffith notes prions are not antigenic.

Paragraph 16 dismisses mechanism 3 and concludes that the agent might be a protein that an animial "is genetically equipped to make, but which it either does not normallyu make or does not make in that form. It may be passed between animals but be actually a distinct protein in different species. Finally, ... there is the possibility of spontaneous appearance of the disease in previously healthy animals."

There are 14 references, fairly standard, except as noted.

     Griffith JS, 1969
      DNA ticketing theory of memory. 
      Nature 223, 580-582 (1969) 

    Griffith JS, 1968
      The unification of neural activity. 
      Proc. R. Soc. Lond., B, Biol. Sci. 171, 353-359 (1968) 

    Griffith JS, 1968
      Mathematics of cellular control processes. II. Positive feedback to one gene. 
      J. Theor. Biol. 20, 209-216 (1968) 

    Griffith JS, 1968
      Mathematics of cellular control processes. I. Negative feedback to one gene. 
      J. Theor. Biol. 20, 202-208 (1968) 

    Griffith JS, 1967
      Self-replication and scrapie. 
      Nature 215, 1043-1044 (1967) 

    Griffith JS, 1967 
      Neural organization underlying conscious thought. 
      Nature 214, 345-349 (1967) 

    Griffith JS, 1966
      An analysis of spontaneous impulse activity of units in the striate cortex of
      unrestrained cats. 
      J. Physiol. 186, 616-634 (1966) 

    Griffith JS, 1966
      A theory of the nature of memory. 
      Nature 211, 1160-1163 (1966) 

    Griffith JS, 1965
      A field theory of neural nets. II. Properties of the field equations. 
      Bull Math Biophys 27, 187-195 (1965) 

Patison IH and Jones, KM Vet Rec 80 2 196

Alper T, Haig, DA, and Clarke, MC BBRC22 2 278 1966

Does the agent of scrapie replicate without nucleic acid?

Nature 214 (90): 764-766 (1967) 
Alper T, Cramp WA, Haig DA, Clarke MC

Scrapie resistance was complete at dosages that knocked the control viability down 3 logs in activity, giving rise to the oft-cited Fig1 and 2 of this paper. Residual scrapie titer was measured by serial dilution, using intra-cranial injection in 7-8 mice at every dilution.

Viroids and prions.

Proc Natl Acad Sci U S A 79 (17): 5220-5224 (1982)
Diener TO, McKinley MP, Prusiner SB

Novel proteinaceous infectious particles cause scrapie.

Science 216 (4542): 136-144 (1982) 
Prusiner SB

1. The Twort phenomenon and the d'Herelle phenomenon are identical. They are two different aspects of one and the same phenomenon: the transmissible lysis of bacteria. When the " dissolving material " of Twort found in diseased agar cultures of micrococci obtained from vaccinia lymph is transplanted into a young broth culture of staphylococci a dissolution. of the latter occurs, and the filtrate of the dissolved culture exhibits all the characteristics of a typical staphylococcus bacteriophage according to the definition of d'Herelle. On the other hand, typical staphylococcus bacteriophage could be obtained also by other means--namely by the leucocytic exudate technique of Bordet and Ciuca, or by the puncture of a subcutaneous abscess. When small amounts of this staphylococcus lytic agent are introduced in melted agar which is afterwards slanted and seeded with sensitive staphylococci a culture results, apparently normal at the beginning, but which, a little later, turns into the typical glassy transparent material of Twort. In other words, the Twort phenomenon leads to the d'Herelle phenomenon, and, inversely, the d'Herelle phenomenon leads to the Twort phenomenon. 2. There are no unquestionable proofs that the bacteriophage is a living organism. The assumption of the: bacteriophage being a filterable virus for bacteria was suggested by two main facts: (a) The power of reproduction possessed by the lytic agent, and (b) the localization of the lysis to certain round spots of clarification when a very diluted lytic agent is poured over the surface of an agar culture of sensitive bacteria. Although easily explained by the virus theory, yet both facts are not unquestionable proofs of the living nature of the bacteriophage, because they are by no means exclusive features of living beings. Fire is not living, and yet fire is endowed with power of reproduction. When once lighted, thanks to an initial impu1sion such as an electric spark or the mere striking of a match, it can be indefinitely reproduced if fuel is provided. A still more striking, because more biological, example is found in blood coagulation. Suppose a series of test tubes containing a stable plasma--bird's plasma, for instance, --which will remain indefinitely fluid. To the first tube we add just a few cubic centimetres of distilled water. As a result of that initial thromboplastic action, which does not need to be repeated in the future, thrombin suddenly appears in the first tube and the plasma clots. If a few drops of the exudate serum in the first tube are pipetted off and poured in the next tube, this second tube clots, in its turn, with a new regeneration of thrombin, which, transferred in the third tube, brings about thc coagulation of that tube with again a new production of thrombin, and so on indefinitely. In this way we realize the transmissible coagulation of blood in series, with the continuous regeneration of thrombin, and thrombin is not a living being. The localization of the lytic action of diluted bacteriophage can be explained by the hypothesis of a chemical substance as well, It must be kept in mind that a culture is not a homogeneous whole, but made up of organisms showing allkinds of qualitative and quantitative individual differences--that is, as far as their susceptibility to the lytic agent is concerned. When a very concentrated lytic agent is poured over the surface of an agar culture an almost complete dissolution occurs, with the exception of just a few organisms resistant enough to overwhelm the strong action of this concentrated lytic agent. On the other hand, when a diluted lytic agent is used only the few extremely sensitive bacteria will be influenced, and each of them becomes a centre of regeneration of the lytic agent, which, diffusing evenly in every direction, produces perfectly round spots of clarification very often surrounded by a kind of halo of diffusion. Between these two extreme conditions all kinds of intermediate degrees exist. Further, any substance, living or not, is composed of particles, molecules, atoms, or ions. When we pour out a glass of soda water, there appear on the wall of the glass small round bubbles of gas, the size of which increase exactly as the so-called colonies of bacteriophage, and yet gas is not a virus. 3. The idea of the bacteriophage being a product of bacterial activity is suggested by the close parallelism existing between, the regeneration of the lytic agent on the one hand, and activity of growth of the bacteria on the other hand. No regeneration ever occurs in dead cultures, nor in living cultures when put in such conditions that they cannot grow-- in saline emulsions of bacteria, for instance, or at low temperature. A slight lysis, with but a small regeneration of lytic agent, is induced in the slow-growing culture of B. coli in a synthetic medium. On the contrary, an abundant regeneration occurs in a fast-growing culture in broth. A recently seeded broth culture to which is added just a trace of lytic agent will not be inhibited ; but a few hours later, at the very moment the culture reaches its acme of growth, a rapid dissolution occurs with abundant regeneration of lytic agent. 4. The conception of the bacteriophage being a chemical substance is favored by the chemical-like affinity existing between a given lytic agent and the corresponding susceptible strain I first observed that small amounts of lytic agents lose a certain part of their activity when put together with too thick emulsions of sensitive bacteria. Bordet, with a different technique, could even obtain the complete disappearance of traces of lytic agent in the same condition. Still more convincing are the results of Yaumain and of Da Costa, who observed the absorption of relatively important amount of lytic agent by dead emulsions of the corresponding sensitive bacteria. This specific affinity which is the necessary condition for a lytic agent for inducing the dissolution of a given bacterium is not favourable to the virus theory, because we question how a virus could be definitely fixed by dead bacilli, which, however, it is unable to attack. 5. The bacteriophage is not one and the same antigen. Several lytic agents showing antigenic specificity must be considered. The coli lytic agent can be completely neutralized by proper amounts of corresponding coli antilytic serum, but is not at all affected by staphylococcus antilytic serum, which, on the other hand, is only able to neutralize staphylococcus lytic agent and not coli lytic agent. This neutralization reaction is thus specific, and demonstrates the plurality of the bacteriophage. The non-specific results obtained with the alexin fixation reaction and advocated by d'Herelle in favour of the unicity of the bacteriophage, are of no value, because they are vitiated, as can be easily demonstrated, by thc presence in the bacteriophage of bacterial dissolution products which have lost their specificity and play therefore the role of common antigen between different lytic agents. Br Med J 1922 ii 296-297