Small critical RNAs in the scrapie agent
Correspondence: (Login to view email address)
- CEA-Institute of Emerging Diseases and Innovative Therapies, Route du Panorama, Bat. 60, F-92265 Fontenay-aux-Rose, France
- U974 INSERM, UMRS 7215 CNRS, Institut Myologie, Hôpital Pitié-Salpêtrière, 75013 Paris, France
- Laboratoire Virologie, Hôpital Saint Vincent de Paul, 75014 Paris, France
PDF (1.4 MB)
- Document Type:
- Manuscript
- Date:
- Received 16 June 2009 14:40 UTC; Posted 16 June 2009
- Subjects:
- Microbiology, Neuroscience
- Abstract:
Unconventional infectious agents cause transmissible spongiform encephalopathy (TSE) diseases including scrapie and bovine spongiform encephalopathy (BSE) in animals and Creutzfeldt-Jakob disease in humans. The protein only hypothesis claims that the TSE agent is composed solely of the protein called prion (PrPsc)1. This protein is the misfolded form of a host-encoded cellular protein, PrPc exerting presumably a vital role at the synapse2. Even though now widely accepted, the prion concept fails to provide in certain circumstances3-6, a satisfying interpretation of the infectious phenomenon. Using the 263K scrapie-hamster model, we conducted a transmission study to search for a putative prion-associated factor indispensable for infectivity. Here we show that innocuous recombinant prion protein (recPrP) was capable, in a reproducible manner, of transmitting scrapie disease when the protein was β–sheet converted in a solution containing PrPsc-derived RNA material. Analysis of the PrP-RNA mixture revealed the association of recPrP with two prominent populations of small RNA molecules having an average length of about ~27 and ~55 nucleotides. We conclude that the nature of the TSE agent seems to be composed of a nucleoprotein molecular complex, in which informative RNA molecules of small sizes are associated with the misfolded prion protein (PrPsc).
Discussion
- Votes:
-
33 votes
- Comments:
-
16 comments
I agree with Adriano Aguzzi’s comments. In addition I would note that if one wishes to make the case that a cofactor such as RNA is required to generate infectious particles, it would be appropriate to quote the prior work of S.Supattapone and his colleagues: They generated infectious prions from purified PrPC using the PMCA technique and showed that addition of a polyanion such as poly(A) was essential for the generation of infectivity (Deleault et al. PNAS 104, 9741 (2007)). Supattapone’s results suggest that the role of the polyanion is structural rather than informational and there seems to be no urgent reason to resuscitate the virino hypothesis at this stage.
I also object to being misquoted. The authors claim that I dismissed the possibility that an RNA molecule might be part of the prion when they say “The possibility of such RNAs interacting with PrP has been evoked previously, but was immediately rejected because of the absence of data establishing a link between the RNA substrate and the infectious phenomenon 13,14. In fact, in the article quoted (ref. 14; Weissmann C. Nat Rev Microbiol. 2004 :861-71) I state: “The discovery of siRNAs and microRNAs, which would have escaped notice in earlier analyses of prion preparations, owing to both their size and their host origin, provides candidates for the hypothetical co-prion that has been proposed by the UNIFIED THEORY18.”(Disclosure: I am the Product Manager for Nature Precedings). Nature Precedings is not designed to bypass the peer-review process, but to act as a complement. The ability of preprint servers to complement the traditional peer review and editorial process has already been demonstrated with the use of the preprint server arXiv in the physical sciences. Many of the articles later peer-reviewed and published in Nature Physics are first posted on arXiv prior to (or concurrently with) journal submission. We believe that the use of arXiv benefits the physics community at large and helps to accelerate the research cycle, and many authors believe feedback they receive from the service benefits their research. Nature Precedings is intended to bring the same benefits to the biomedical community.
One of the benefits is that authors are able to receive (hopefully) helpful comments from others who may not have participated in the traditional peer-review process, thereby leading to more robust (and open) research.
This paper is the first succesful TSE agent reconstitution experiment showing infectivity of RNA isolated form TSE infected tissue, much alike to the virus reconstitution experiments that were required to demonstrate the infectivity of tobacco mosaic virus RNA which (the virus) at some point was beleived to be a protein.
To further make the point of RNA being infectious and playing informational rather than structural role I suggest authors to do the following experiments:
1. Isolate RNA from Strain 1 and mix it with recPrP and inoculate into the animals brain. Do same thing with RNA isolated from Strain 2. Compare the resulting disease. RNA from strain 1 should give rise to Strain 1 TSE agent and RNA from Strain 2 should generate Strain 2 TSE agent. Taking into account the possibilty of pathology modification observed in the first passage the strain should be evaluated in the second and further passage animals.
Also to ease the strain discrimination such study should be done using other well caracterised TSE models like mouse scrapie strains such as RML,ME7,22L, 301C which are now well defined by IHC and CPA both in C57Bl/6 and Tga20 mice “Prion Strain Discrimination Based on Rapid In Vivo Amplification and Analysis by the Cell Panel Assay” http://www.plosone.org/article/info:doi/10.1371/journal.pone.00057302. To demonstrate that recPrP has only structural but not informational role in TSE agent, RNA from infected brain should be mixed with nucleic acid binding proteins other than PrP. My first proposal would be NCP7 – retroviral nucleocapsid protein. Then the infectivity of such complex should be tested.
Ultimately similar experiment could be done by simply mixing RNA with lipofectamine or other trasfecting agents like peptides.Also SAF preparations generally have less infectivity in them than unprocessed total brain homogenate, so the total RNA isolated from brain before SAF preparation might have even more infectivity.
RNAse A depending on salt concentration can also cleave double stranded RNA. I suggest to use RNAse III that digests only dsRNA to define the RNA species more precisely.
At the end I would like to comment on possible siRNA nature of TSE agent RNA. As I mentioned in my abstract (supplement of Brain Pathology 2006, absrtacts of ICNP, page 54, abstract 117: PRION OR VIRINO OR JUST A SMALL INTERFERING
RNA SILENCING NEURONAL SURVIVAL GENES) the strict homology of
TSE agent specific nucleic acid (siRNA) to host gene(s) that it is targeting and silencing explains
the failure to identify any “foreign” (non-host encoded) nucleic
acid sequences in infected tissues.
I propose satellite viral derivation of such siRNAs (owing to nucleic acid sequence homology with host genes). Satellite RNA replication would be supported by cryptic helper virus dormantly and persistently infecting all TSE susceptible species.The publication to Nature Precedings was not our first choice. The data was first submitted to Nature that decided, after almost one month of consideration, to reject the manuscript without sending it for peer-review. During this delay, Nature suggested the possibility that we submit the manuscript in Nature Precedings, an online preprint server that preserves the right of the authors to publish it a peer-reviewed journal subsequently. Because of the sensitivity of the data and to disseminate the results quickly to the whole scientific community we seized this opportunity, which is not so different than presenting data at a scientific conference.
As corresponding author I would like to make some comments . First of all to thank Nature Precedings to have opened this space of freedom. Freedom is the master word for a researcher. Of course peer-review is useful, but everyone has, at least once, lived the injustice of the emotional reactions of a unfair peer-reviewer and many of us are aware of work published in prestigious journals that came from unreproducible results. Reproductibility represents yet a crucial element in experimental research to validate results.We chose this option for our own work to further legitimate the interpretation of the data. If controls are missing, it is not by intellectual laziness because we carefully surveyed a possible contamination in several critical animal groups during a voluntary uncommon long incubation time 545 (1st experiment) /367 (2nd experiment) days.We though that to be sufficiently convincing for the reader. So I apologize to the colleagues for this truncated information because we have also performed other experiments not described in this pre-print version. Using a different protocol, we obtained a residual contamination that gave a 263K lesion profile. Also we performed second passage experiments where we noted for one animal, an infectious agent clearly different than the 263K suggesting it was born from recombinant PrP. This type of result with recPrP has never been obtained with the PMCA technique and it is not easy to compare our strategy with this approach. Concerning the attractive « UNIFIED THEORY » I would like just to specify that about the co-prion « it (the theory) denies that this nucleic acid is required for infectivity » what is precisely the heart of the debate here
I welcome the forum of Nature Precedings and appreciate its characterization as a digital scientific conference, of sorts. I am very pleased to see these data and have the opportunity to comment.
An obvious but time-consuming experiment is to perform secondary passage of this material. Only then could the reconstituted material be characterized as infectious. We have already seen transgenic/recombinant forms of PrP that evoke pathology without giving rise to a bona fide infectious agent.
Much time has elapsed since the authors isolated these SAF-associated RNAs and the lack of data as to their sequences (or even whether they are double stranded) also makes me uncomfortable.
This refers to the above mentioned virus reconstitution experiments which are very relevant to the current TSE agent reconstitution experiments.
Virus reconstitution and the proof of the existence of genomic RNA. http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=10212937
This paper presents interesting data that deserves attention and follow-up. Also refreshing to see Nature permitting a non-prion interpretation in its (electronic) pages.
Heino Diringer
The principal strategy to characterize an infectious agent remains purification and identification of components, followed by an analysis of their relationship to the transmissibiliy of disease. This has been carried out in the publication by Simoneau et al. which, although not apparently peer-reviewed, provides convincing evidence supporting a novel hypothesis.
The experimental results prove that two small RNA fractions are necessary for the transmission of transmissible spongiform encephalopathies (TSEs). This is the first time that a reconstitution experiment, involving a specific nucleic acid fraction from hamster scrapie and recombinant ‘so-called’ prion protein, has resulted in TSE transmission without the necessity for unnatural sonication.
The clear methodology, the inclusion of appropriate controls, the expected low transmission rate, the fact that the pathology of the induced hamster scrapie was distinct from the starting strain (a crucial control) and the reproduction of the biochemical results in two independent experiments in combination indicate that the conclusions are soundly based.
The paper would be improved by a more detailed presentation of the purification protocol for SAF and treatment with RNase and DNase later in the purification process may have yielded a better starting material for the extraction of nucleic acid. Nevertheless the results warrant priority publication as the data can be easily tested by other groups and may lead to a more defined understanding of the transmissibility of TSEs. The authors should be congratulated to their interesting results.
It seems to me that this kind of communication of new data may sometimes be preferable to the publication of a peer-reviewed sensation (see Science 216: 136 -144, 1982), which turned out to be a false interpretation of an extraordinarily badly designed biochemical experiment (see Bioscience Reports 3: 563 -568, 1983).
COMMENTS ON THE FRENCH PAPER
I regard the French paper as very interesting. Of course, the possibility of contamination has always to be considered. However, the paper is incomplete until the authors include a section on how they interpret PMCA data in the light of their conclusions. Their work is well conceived and carefully conducted and will become very important in rebalancing TSE research – Laura Manuelidis makes a very valid point because numerous prion sceptics are known to have been coerced by editors of various journals either to write in prionese or to withdraw their document.
—————————————————————————————————————————————————————————-
Charles Weissmann’s reaction that “there seems to be no urgent reason to resuscitate the virino hypothesis at this stage.” requires a response at this juncture. As the originator of the virino hypothesis, it is essential to correct the confusion about virinos in his TSE papers and their diagrams (1). The term virino has never implied that an informational nucleic acid (a ‘viro’) separates from its associated protecting PrP for entrance into host cells or that it is detached from PrP for replication. The virino hypothesis has not yet involved the question, if or when, the viro, during its lifecycle, may separate from its donor-coded dimeric PrP. The viro is probably a small RNA.The virino concept originated during the 1970s from genetical and pathogenesis experiments that initially yielded a replication site hypothesis in which different TSE strains compete when injected at different times (2). A numerically limiting stage in the process of agent replication appears to be involved, but a protein-only, prion copying cascade, is open ended without any bottleneck. A prion explanation of competition is still awaited.
The murine gene, sinc (syn. Prnp), exerts exceptionally precise control of the incubation period of the many scrapie strains (>20) and the complex differences between strains, including reversal of allelic effects, This precision was interpreted as a short, direct pathway between its protein (PrP) and agent replication, and thence the strict incubation dynamics. Our 1971 conjecture that this protein could be the agent’s bonding and replication site, could also imply that a host-coded informational molecule normally occupies the bonding site and that the intruder displaces it (3).
The recurrent objection to the virino hypothesis has been that no nucleic acid has yet been isolated from infective PrP preparations. There is a well established likely explanation. Several groups have shown that it is difficult to detect NAs associated with PrP when using standard NA methods (4). The very positive breakthrough, effectively endorsing the virino hypothesis has been provided by the group led by Jean-Luc Darlix, using a very unexpected approach. They proved that PrP is an RNA chaperone protein that could be interchanged with the NCp7 nucleocapsid protein of Aids virus, where it provided the requisite wide repertoire of nucleic acid processing properties (5).
The following quotation is from Charles Weissmann’s 2004 Nature review of TSEs : ”... the finding that a particular prion strain can be transmitted between two species without changing strain-specific properties, even though the PrP sequences of the two species are very different, is unexpected because it implies that the same conformation can be imposed on PrP molecules with different amino acid sequences.” (1) If he is still able to accept this as compatible with a prion hypothesis, it is akin to adding another epicycle layer to prion theory, rather than accepting its end. The discovery that PrP is an RNA chaperone appears to herald the validation of virinos and the obituary of prions. The real virino hypothesis is very much alive, but not noisy.
(1) Weissmann, C. (2004) The state of the prion. Nature Reviews Microbiology 2, 861 (2) Dickinson, A.G. et al.(1975) Extraneural competition between different scrapie agents leading to loss of infectivity. Nature 253, 556.
(3) Dickinson, A.G. & Meikle, V. M. (1971) Host-genotype and agent effects in scrapie incubation: change in allelic interaction with different strains of agent. Molec. gen. Genet. 112, 73.
(4) Silva, J.L. et al. (2008) Intriguing nucleic-acid-binding features of mammalian prion protein. Trends biochem. Sci. 552,1.
(5) Gabus, C. et al. (2001) The prion protein has RNA binding and chaperoning properties characteristic of nucleocapsid protein NCp7 of HIV-1 J. biol. Chem. 276, 19301Hugh Fraser. This is an important paper. The protein-only hypothesis has become stale. The causative agent (& the many strains)of the TSEs obey the rules of biology:- genetic information/diversity, mutation, exclusion competition between strains, (all Dickinson’s work); the protein-only hypothesis fails on all of these criteria, whereas the “virino” meets these challenges. This paper opens new and exciting possibilities on the chemistry of the informational molecule that has been elusive for so long.
I haven’t worked in this field for more than twenty years so I don’t feel qualified to judge the quality of this paper. But it seems to me that if its findings are verified and extended they good go a long way towards vindicating the assertion of Alan Dickinson and his team (with which I was privileged to work) that what ever proteins might be implicated in the pathogenesis of this class of diseases there must also be the involvement of a replicating informational molecule to account for the facts of a considerable number of transmission studies.
George OutramMy comments on this report by Simoneau et al. are rather tardy but, until a few days ago, I wasn’t aware of it, nor of the existence of Nature Precedings. To echo some of the previous comments on this report, I respect the decision by Simoneau et al. to post their paper, and congratulations to Nature for offering this opportunity.
Here, I would like to expand on Charles Weissmann’s comment, which points to the highly significant finding of this and many other studies (reviewed in [1,2], see also [3]) that nucleic acid molecules (NAs), such small RNA molecules, are involved in TSE.
Assuming that NAs do participate in TSE, then the role of these NAs should be evaluated (both, conceptually and experimentally) under the following two overall premises regarding (I) their function, and the time of the “life cycle” of the TSE agent when this function is exercised, and (II) their physical association with the “TSE infectious units” (TIUs):
I. NAs have: (a) structural function, (b) informational function, or (c) both, and this function is exercised (d) during the assembly of TIUs, (e) during the transmission to a new host (i.e. a new host organism or a new cell in the same host), or (d) during the multiplication of TIUs by serving as templates.
II. NAs are: (a) essential components of the TIUs, (b) they are essential chaperons during the assembly of new TIUs, but are not included in the TIUs, (c) they are non-essential chaperons, but they do increase the rate of generating new TIUs, or (d) NAs are included in TIUs nonspecifically, as bystanders, which means that their removal from TIUs will have no effect.
A combination of all these possibilities leads to a rather complex set of hypothetical models for the role of the NA in TSE. Considering that, historically, the TSE research field has been plagued by imprecise and sometimes rhetorical use of language, terms, and concepts, to an extent that it has been the subject of research studies by our colleagues in other disciplines [4], it is critical that the these models are clearly defined when designing TSE research studies or when interpreting the results of these studies. Otherwise, the long history of misunderstandings and the associated unconstructive discourse in the TSE field (see [4-6]) will continue to constrain progress.
In evaluating the study by Simoneau et al. and those by others who have shown that NAs participate in TSE (reviewed in [1,2]), I assume that the results are real and not due to contamination or other artifacts. The study by Simoneau et al. shows “that innocuous recombinant prion protein (recPrP) was capable, in a reproducible manner, of transmitting scrapie disease when the protein was β–sheet converted in a solution containing PrPsc-derived RNA material.” As mentioned above, there is strong evidence that NAs increase the efficiency of generating TIUs in vitro (reviewed in [1,2], see also [3]). However, unlike in other studies, the NAs in this study by Simoneau et al. were RNA molecules that were apparently released from the TIUs (specifically, from SAF preparations purified from brains of hamsters infected with the 263K scrapie strain).
The study by Simoneau et al. is highly significant from two perspectives. First, this is one of the few successful studies (maybe the first one, considering that other similar studies [7] apparently have yet to be reproduced in independent laboratories) showing that recombinant PrP can be a substrate for generating TIUs in presence of NAs. As pointed out before [6], failure to demonstrate that recPrP can be a substrate for generating TIUs has been one of the main arguments against the protein-only nature of the TSE pathogens as currently defined by the prion hypothesis. Interestingly, the study by Simoneau et al. demonstrates that recPrP can be a substrate for generating TIUs and, therefore, could be interpreted as strong experimental evidence for the prion hypothesis. Ironically, however, the authors of this study and many of the TSE researchers posting comments here interpret theses results as evidence against the prion hypothesis.
The other significant findings of the study by Simoneau et al. were that the TIUs (i.e. the SAF preparations) purified from brains of hamsters infected with the 263K scrapie strain apparently contain NAs, specifically two classes of small single stranded RNA (ssRNA) molecules averaging 27 and 55 nucleotides, and that these ssRNA can induce in vitro in the presence of recPrP the formation of new TIUs.
However, much of the interpretation of these results and the main conclusions by Simoneau et al. don’t do justice to these significant findings. Neither does the fact that the authors failed to include in their study a key reconstitution control experiment using non-specific, randomly synthesized RNA molecules with a molecular weight similar to that of the two classes of small ssRNAs (see below). Moreover, there is no statistical evaluation of the results.
Let’s consider first the statistical aspect. As an example, assume that the role of the small ssRNAs found in this study is to enhance the formation of TIUs by a factor of one hundred. Considering that only a relatively small percentage of the animals inoculated with the β recPrP-RNA complex developed TSE (5 of a total of 24 mice;20.8%) then, statistically, in order to see one individual animal with TSE in the control group inoculated only with β recPrP, this control group would need to contain about 500 mice. Suffice to say that this study did not use such a large control group. Therefore, this study does not demonstrate that the small RNA molecules are essential for generating TIUs, or for TSE infectivity, as concluded by the authors. It is possible that the role of ssRNA molecules is limited strictly to increasing the rate of generating new TIUs.
Considering the numerous studies indicating the participation of NAs, such small ssRNAs, in TSE, evaluating their specificity is critical. As mentioned above, failure to include a reconstitution control experiment using non-specific, randomly synthesized RNA molecules with a molecular weight similar to that of the two classes of small ssRNAs is probably the major weaknesses of this study by Simoneau et al. This relatively straightforward experiment would have given us at least some insights into the answers to this critical issue.
The authors mentioned in their report that they were in process of sequencing the two classes of RNA molecules, and this would ultimately answer the question about their specificity. Because several months have passed since, apparently, the authors started this sequencing work, they probably have an answer, and I hope they will announce it in a response to this comment. However, I’ll proceed addressing this issue based on the experimental evidence published in previous studies (reviewed in [1,2]; see also [3]), which showed that nonspecific NA (including polyanion molecules) can induce the conversion of PrP into TIUs under experimental conditions, apparently, even in the absence of seeding TIUs. Based on this evidence, it is likely that the sequence of the two classes of ssRNA molecules found in the study by Simoneau et al. is non-specific.
However, in order to avoid potential misconceptions, the use of the term “non-specific” requires clarification. Similar to many other biochemical interactions, it would be expected that, based on their length and their secondary structure, some ssRNA molecules might be more efficient in facilitating the formation of new TIUs than others. From this perspective, it is likely that if the NAs, such as ssRNA, do indeed participate in TSE, then their specificity is at the best structural, not sequence-based.
This is already a very long and rather involved comment, so I’m restraining from further discussing the potential participation of NAs (in vitro and in vivo) in the TSE phenomenon as “structural” or “informational” molecules according to the many models outlined above, but before concluding, I would like to point out that studies addressing the fundamental nature of TSE pathogens, such as this study by Simoneau et al., would benefit from integrating the in vitro and in vivo experimental findings to those observed in natural TSE. For instance, we know for at least a decade that TSE, and therefore TIUs, can be generated in the absence of previous TIUs, such as in spontaneous TSE associated with genetic mutations in the PrP gene. In the context of the study by Simoneau et al. this argues against involvement of sequence-specific RNA sequence in TSE, but not against non-specific RNA participation.
Another point I would like to make is that because a “genetic element” or any other entity are found in all the host cells, it doesn’t necessarily mean that they are host elements, as Simoneau et al. state in their paper. As I discussed, in a previous publication [8] and in an unpublished manuscript entitled “Endogenous viral etiology of prion diseases,” which I have heavily circulated during the last two years among researchers in TSE and related fields (please e-mail me at: cib1@cdc.gov, if you want a copy), some elements, such as endogenous viruses, can be transmitted vertically through the germ-line, and therefore they would be present in the all cells of an organism.
In conclusion, I believe that the study by Simoneau et al. is highly significant. However, the paper could be improved by addressing some of the issues in this and other posted commentaries.
Reference List
[1] Silva JL, Lima LM, Foguel D, Cordeiro Y. Intriguing nucleic-acid-binding features of mammalian prion protein. Trends Biochem Sci 2008; 33(3): 132-40.
[2] Nandi PK. Prions at the crossroads: the need to identify the active TSE agent. Bioessays 2004; 26(5): 469-73.
[3] Deleault NR, Harris BT, Rees JR, Supattapone S. Formation of native prions from minimal components in vitro. Proc Natl Acad Sci U S A 2007; 104(23): 9741-6.
[4] Reeves C. An Orthodox Heresy: Scientific Rhetoric and the Science of Prions. Science Communication 2002; 24: 98-122.
[5] Priola SA, Chesebro B, Caughey B. Biomedicine. A view from the top—prion diseases from 10,000 feet. Science 2003; 300(5621): 917-9.
[6] Aguzzi A, Heikenwalder M. Prion diseases: Cannibals and garbage piles. Nature 2003; 423(6936): 127-9.
[7] Legname G, Baskakov IV, Nguyen HO, Riesner D, Cohen FE, DeArmond SJ, Prusiner SB. Synthetic mammalian prions. Science 2004; 305(5684): 673-6.
[8] Bandea CI. From prions to prionic viruses. Med Hypotheses 1986; 20(2): 139-42.In my previous comment on the report by Simoneau et al., I outlined some of the potential models for the participation of nucleic acids (NAs) in TSE, but stopped short of evaluating their merit. Following a stimulating, informal correspondence with some colleagues in this field, I would like to address the potential role of NA TSE in more detail and from a new perspective.
For almost half of century, one of the primary objectives in TSE research, and the foundation of the different views about the nature of TSE pathogens, has concerned the role of NAs in TSE. The progress in the TSE field has been tremendous, but despite numerous studies and very large amounts of experimental data, we still have major opposing views on the nature of TSE pathogens although, as metaphorically described by Dr. Dickenson (see comment above), some are noisier than others. This situation is unprecedented in the history of science, and the reasons for this situation might be surprising.
In searching for an explanation, we might first ask the legitimate question of why different groups of highly regarded scientists would interpret the same experimental data in such disparate ways. To give tribute to these TSE researchers, I do not believe that so many of them have been wrong for such a long period. So what happened? A sensible explanation might be that, partially, all these opposing views are correct and, therefore, all the scientists involved are partially right: the proponents of the prion hypothesis are right, those of the virino hypothesis are right, and the near extinct breed of scientists sustaining a viral etiology of TSE are right. Is this possible? Can we integrate all these apparently disparate views into a unifying scenario about the nature of TSE pathogens based on conventional biological and evolutionary principles and be consistent with the experimental evidence? I believe we can, but something has to change, specifically our traditional, reductionist approach in evaluating the TSE phenomenon.
Historically, the studies on infectious diseases, particularly viral diseases, have focused primarily on the transmissible infectious units (IUs) in the life cycle of pathogens. Unfortunately, this practical approach has obscured the fundamental biological principle that all pathogens have a life cycle and that, in order to define the nature of these infectious pathogens, all the stages of their life cycle need to be clearly defined and integrated into a coherent view. Certainly, the stages in the life cycle of some pathogens, such as the TSE pathogens, might not be as distinct as those of others, but nonetheless it is critical that these stages are well defined when designing experiments and, especially, when interpreting the results. Failure to do so could lead to interpreting experimental data form different perspectives, leading to “correct” but conflicting views, and that is what I believe has happened in the TSE field.
Moreover, focusing primarily on the IU stage in the life cycle of an infectious pathogen could lead the expectation that the IUs should contain all, or at least the major “structural” and “informational” molecules of the pathogen. However, this doesn’t need to necessarily be the case. For instance, vertically transmitted pathogens such as endogenous viruses are present in all the host cells and, therefore, their IUs do not need to carry all their components (including the genome) to a new host cell. Of course, this begs some obvious questions, such as: if already present in host why would these pathogens still produce IUs, or if these pathogens are present in the host, why don’t they cause disease. I have addressed these essential questions at length in a recent paper (“Endogenous viral etiology of prion diseases;” please e-mail me at: cib1@cdc.gov, if you want a copy) so I’ll not discuss it here.
Based on these conventional biological principles and supported by evolutionary evidence, I proposed an endogenous TSE virus model which is consistent with the current experimental evidence and integrates all three major historical views about the nature of TSE pathogens – the conventional virus theory, the prion hypothesis, and the virino concept – into a unifying scenario. And, in context of this scenario, the NAs play a critical role in TSE.
First of all, similar to all endogenous viruses, the TSE endogenous viruses have a genome, a reduced viral genome, that is vertically transmitted through the germ-line. These TSE endogenous viruses are symbiotic elements that can inadvertently produce transmissible, “infectious” viral particles. The existence of a viral genome and the production of viral particles should satisfy researchers supporting the conventional virus view on the nature of TSE pathogens.Second, definitely the viral particles produced by the TSE endogenous viruses do not contain the viral genome. However, this is a very common phenomenon in the life cycle of most if not all viral species, which can produce viral particles made primarily of protein and lacking the viral genome. This conforms to the fundamental argument of the prion hypothesis that the transmissible particles do not contain a viral genome and are made primarily of protein.
Third, it is very likely that non-genomic NAs, such as small RNA molecules, do participate as chaperones in the assembly of transmissible TSE viral particles. During this process, it is likely that these RNA molecules participate as “structural elements” by facilitating the assembly of TSE viral particles. In doing so, it is possible that these NAs also play an “informational role” by dictating the rate or the pattern of the assembly of the viral particles, thereby dictating the strain characteristics of the TSE pathogens (this interesting possibility requires extensive discussion, and I will leave it for futures encounters). Certainly, in my view, this aspect in the life cycle of the endogenous TSE viruses conforms to the ideas behind the virino concept.
In summary, the endogenous TSE virus model integrates the major historical views about the nature of TSE pathogens into a unifying scenario, and explains the participation of NAs in TSE. From the perspective of those who see the glass half full, this might be a win-win situation for everyone.
In his comments on the report by Simoneau et al. a few months ago, Alan Dickinson raised a highly relevant point: “the paper is incomplete until the authors include a section on how they interpret PMCA data in the light of their conclusions.” The PMCA technology has reinvigorated the studies on the nature of TSE pathogens, and the PMCA data needs to be explained in context of all models on their nature.
Since my last comment, I’m happy to announce that I posted the paper presenting an endogenous TSE virus model in Nature Proceedings (1). Here, I would like to summarize this model, explain the PMCA data in context of this model, and discuss the potential role played by nucleic acid molecules (NAs), such as small RNA molecules, in the PMCA.
In the endogenous TSE virus model, the PRNP is the gene of a symbiotic endogenous virus that entered the genome of vertebrates very early in their evolution. Similar to other symbiotic endogenous viral genes, such as the murine endogenous viral gene Fv1, the PRNP encodes an antiviral protein, the PrP, which blocks the life cycle of various pathogenic viruses. As it would be expected, the PRNP is expressed primarily in tissues that are not under normal immune surveillance, such as the nervous system.
The endogenous viral PrP inhibits the life cycle of pathogenic viruses by interacting with their components. Under the selective pressure to perform this defensive function against various viruses, the PrP has evolved the property of folding into diverse isomeric conformations. Considering its potential cellular signaling properties, it is conceivable that PrP performs an additional antiviral protective mechanism by triggering the death of virus-infected host cells through apoptosis.
Similar to immune components, the PRNP was strongly selected against pathogenic expression. However, do to mutations in the PRNP (both, germ-line and somatic cell mutations), or to epigenetic factors, occasionally, the endogenous viral PrP self-assembles into transmissible, virus-like particles (TSE-VPs). The generation of TSE-VPs is an extraordinary rare event, with a rate that depends on the viral PRNP allele and other factors. However, once a TSE-VP is produced, it is recognized by other native PrP molecules as a genuine pathogenic virus. This triggers their antiviral protective function that, as explained in the paper, leads to generation of new TSE-VPs in a perpetual cycle.
The rate of this cycle depends on the isomeric conformation of the PrP within the transmissible TSE-VPs, which very likely is used as a scaffold for the assembly of new TSE-VPs, on the host’s PRNP alleles and, possibly, on other host factors, such small RNA molecules that act as chaperons in the assembly of new TSE-VPs. Basically, this is also what happens in the PMCA assays during the in vitro assembly of new TSE transmissible units (2).
As previously discussed (1), the self-assembly of virus particles from their components in vitro is a well established fact. Moreover, production of transmissible virus particles lacking the viral genome and made primarily of protein is a common event in the life cycle of many viruses. And, it would be a relatively easy task to experimentally simulate the endogenous TSE virus model in vivo, by engineering endogenous viruses that produce transmissible, protein-only virus particle that after entering new host cells would activate the native endogenous virus to produce new similar protein-only transmissible virus particles. Obviously, taking the production of these protein-only virus particles out of the context of the viral cycle and asserting that their production is independent of a genetic component (i.e. the viral genes) would be a major scientific flaw.
The role of NAs, such as small RNAs, in the assembly of new TSE-VPs, or TSE infectious units (TIUs), in vivo or in the PMCA assays is not known (see comments above). Most PMCA assays use brain tissue homogenates as a PrP source. Because these homogenates contain all the cellular components, including relatively large quantities of NAs, most of these PMCA studies did not provide new information on the nature of TSE pathogens (as some of these studies implied) or on the potential participation of NAs in the assembly of new TIUs. However, these PMCA studies demonstrated that TIUs can be generated de novo in a cell free system, which opened the door for the eventual purification of the components necessary their assembly.
Indeed, in a PMCA study using partially purified fractions of PrP, it was shown that these fractions can sustain the formation of new TSE transmissible units only in the presence of NAs, such as synthetic polyanions (3). This study is strong evidence for the participation of NAs in the in vitro generation of TIUs. The results of this study are consistent with the endogenous TSE virus model and inconsistent with the prion hypothesis. It remains to be seen to what extent various NA molecules, such as RNAs with specific nucleotide sequence or with specific secondary structures can affect the rate at which PrP assembles into new TIUs. These experiments can be performed with purified PrP fractions or with tissue homogenates that are free of native NAs, and reconstituted with NAs, such as specific RNAs.
In the endogenous TSE virus model, the generation of new TIUs is regarded as an intrinsic property and activity of the native PrP molecules, rather than of the incoming TIUs. This perspective accommodates not only infectious TSE, but also provides the scientific rationale for understanding inherited TSE, such as familial Creutzfeldt-Jakob disease, and spontaneous TSE, which are not triggered by infectious prions. Therefore, the property of the PrP to form TIUs is not a function of, or the result of the replicating activity of a preexisting TIU. Clearly, the TSE pathogens are not a case of replicating proteins as proposed in the prion hypothesis.
The endogenous TSE virus model could lead to a major paradigm shift in TSE field, as well as to a better understanding of the fundamental relationship between genes and proteins, between genotype and phenotype, while conforming to the central dogma of molecular biology.
References:
1. Bandea, Claudiu. Endogenous Viral Etiology of Prion Diseases. Available from Nature Precedings (http://hdl.handle.net/10101/npre.2009.3887.1), 2009.
2. Castilla,J., Morales,R., Saa,P., Barria,M., Gambetti,P., and Soto,C. 2008. Cell- free propagation of prion strains. EMBO J 27:2557-2566.
3. Deleault NR, Harris BT, Rees JR, Supattapone S. Formation of native prions from minimal components in vitro. Proc Natl Acad Sci U S A 2007; 104(23): 9741-6.
- (Login to share with a colleague)
Additional information
- License:
- This document is licensed to the public under the Creative Commons Attribution 3.0 License
- How to cite this document:
-
Simoneau, Steve, Ruchoux, Marie-Madeleine, Vignier, Nicolas, Lebon, Pierre, Freire, Sophie, Comoy, Emmanuel, Deslys, Jean-Philippe, and Fournier, Jean-Guy. Small critical RNAs in the scrapie agent. Available from Nature Precedings <http://hdl.handle.net/10101/npre.2009.3344.1> (2009)
- Version info:
-
Other versions of this document in Nature Precedings
None.
Other versions of this document elsewhere on the web
None known.
Adriano Aguzzi on 19 June 2009 10:36 UTC
Peer-review is not primarily meant to be a hassle for scientists eager to publish their findings. It is also a mean of protecting authors, e.g. by pointing out logical flaws, missing controls, alternative interpretations, etc. Not only does the process improve the quality of published science, but it also helps authors avoiding potential embarassment.
This paper appears to represent a good case in point. A plausible interpretation is that the traces of infectivity detected by the authors represent residual contamination stemming from the 263K inoculum. There would be valid ways to test for that, however no such efforts were undertaken. Experienced referees would have advised the authors to perform these tests before going public.
Time will tell whether the bypassing of the peer-review process, in this case, will have done any good to science at large and to the authors specifically.