Reviewer 1: Jerzy Jurka, Genetics Information Institute
This is a straightforward paper exploring important implications of prion-mediated heredity for the Central Dogma.
I have only some minor comments:
(1) It would be useful to quote the article Alain E. Bussard on the same subject [75
], and to highlight the major new arguments introduced in the current article.
Author’s response:Bussard’s article appears to be something of a misnomer in that the author puts the Central Dogma in the title but does not really examine it with respect to the properites of prions. His article instead explores the Lamarckian features of the prion inheritance and certainly is of interest for its historical aspects.
(2) I believe that the author should also comment on RNA editing in the “look-ahead” section; e.g. [76
Author’s response:This comment is appreciated, an interesting point is brought up here. Although as far as RNA editing or transcriptional errors are involved, the actual mutating entity is RNA rather than protein, the net effect on heredity is the same as with translation errors. Effectively, a version of the look-ahead effect indeed could be engendered by RNA editing, with stochastic editing creating variation that can be subsequently fixed in the genome through convergent genetic variation. I included a statement to that effect in the revised version of the manuscript. The same pertains to errors of transcription. Unfortunately, the report of extensive editing by Li et al.
]has been compromised to such an extent
]that it has become difficult to interpret these observations.
(3) Finally, I recommend including the article by Sergey G. Inge-Vechtomov et al.
], that includes a unique historical perspective on “non-inherent variability” dating back to Kirpichnikov.
Author’s response:I cited the article by Inge-Vechtomov et al. as a review but my reading again is that this is about protein-based heredity not violations of the Central Dogma.
Reviewer 2: Pierre Pontarotti, Universités de Provence et de la Méditerranée
In this review/outlook, the author studied the assumption that the information could be originated from protein to protein and from phenotypes to DNA. The author also highlighted that “violation of the central dogma” and the “epigenetics trans generational inheritance” are two different phenomena, even if sometime, they could be connected. Although this point seems to be obvious, this clarification is essential (I meet several scientists mixing the two concepts). In my opinion, I think that the main question asked by the author: “Does the central dogma still stand?” is opportune. Investigators really need to transgress scientific dogmas. But we will still need to propose robust approaches to test new hypotheses.
Reviewer 3: Juergen Brosius, University of Muenster
The author questions the Central Dogma of Molecular Biology, because structural modification of prion-like proteins might have a (more global) effect on the expression and even structure of gene products. The idea is based on data obtained with studies of yeast prion-like proteins, for example, the Sup35 protein that normally acts as a translation termination factor. However, when sequestered in amyloid, hence insoluble and inactive, its deficiency allows for readthrough of termination codons in messenger RNAs.
The author places such strategies of increasing evolvability into the category of quasi-Lamarckian evolution. Like most phenomena in biology, similarity to the Lamarckian mode of evolution lies somewhere on a continuum ranging from barely apparent to very strong. In my opinion, the possible elongation of polypeptides is at the very weak end of the continuum of Lamarckian mode of evolution, not too remote from random mutations of nucleotides, since this process is almost equally non-directed.
Author’s response:I am not going to strongly argue this point. Thea ‘quasi-Lamarckian’ mechanisms certainly belong to the continuum of evolutionary phenomena, from stochastic to deterministic ones
]. The Lamarckian character of this readthrough is not the focus of the present article which is about information transfer from protein to genome; the readthrough seems to capacitate such transfer.
In addition, most S. cerevisiae 3′-UTRs tend to be short, typically in the size range of 50 to 200 nucleotides, with a median length of 121 nt [80
]. Should a C-terminal ORF extension truly be part of a ‘look ahead effect’ [50
] for times of stress, might one not observe distal to the bona fide stop codons a slightly higher conservation of the first two positions in the respective codons?
Author’s response:This is a really, really interesting idea. Yes, in principle, one should expect some degree of purifying selection in the sequences downstream of stop codons. However, because readthrough is not frequent, the effect could be quite weak so that its detection would require sophisticated statistical analysis of large data sets. To the best of my knowledge, no one has shown that such conservation does not exist (a difficult task as well). This seems to be well worth investigating.
Furthermore, messenger RNAs transcribed from genes containing mutations that generate aberrant extended 3′ untranslated regions are degraded by nonsense mediated decay (NMD) [81
]. Even if NMD was suppressed by an additional stress induced mechanism, the C-termini of proteins usually are the least conserved parts of a protein and often can be altered or extended without functional consequences [84
]. Once more, the action of prions apparently do not have an effect on their own expression or C-terminal extension. Due to this undirected nature, I would place prion formation to the weak end of the continuum concerning quasi-Lamarckian mode of evolution. Something similar could be said about modification of nucleic acids, most prominently methylation of DNA in control regions of genes, as this process seems to be not specifically directed. However should the link between prenatal nutrient deprivation in humans and adiposity in later life - in conjunction with the findings that reduced methylation of the insulin like growth factor II gene (IGF2) is the underlying molecular mechanism for this effect – become substantiated, at least some of these epigenetic effects due to methylation changes could be placed closer to the opposite end of the continuum [85
]. Animal studies will be essential to rigorously test these observations initially made in human populations. The case of the prokaryotic CRISPR-cas system of defense against mobile elements including plasmids and viruses is an interesting and much stronger case as covered by the author in a previous publication [67
]. Nevertheless, a stochastic event and not the need for viral defense lead to integration and antisense transcription of part of the invader’s genome. The fortuitous beneficiary effect of, e.g., antiviral protection was of selective advantage and became fixed. As mentioned in the review of this Koonin/Wolf article, in my view, the examples that most closely resemble Lamarckism or quasi-Lamarckism stem from several ongoing transitions in our own lineage, namely vertical and horizontal transmission of memes [88
] and at the level of genes through our potential (not yet realized) to direct acquired knowledge about genotype/phenotype relationships into our own genome in a precise and specific manner via genetic engineering [90
Author’s response:These comments are appreciated. I find memetics to be of much interest and promise. However, in my view, this field of enquiry is outside evolutionary biology sensu strictu.
The Central Dogma of Molecular Biology is, in my opinion, still untouched as there is no reverse translation. This would change if a mutation in a given protein including a translational readthrough beyond the stop codon directly would lead to a nucleotide change that converts said stop codon into one that encodes an amino acid. My own problem with the Central Dogma of Molecular Biology is different, more trivial, and based on the depiction of DNA and not RNA topping the hierarchy [93
Author’s response:This is a key conceptual point that is addressed in the main text of the present article but is worth pondering again. True, to the best of our knowledge, there is no reverse translation but this is not what the Central Dogma is about. Quoting Crick
]: ‘The central dogma of molecular biology deals with the detailed residue-by-residue transfer of sequential information. It states that such information cannot be transferred back from protein to either protein or nucleic acid.’ So Crick was fully explicit in formulating the Central Dogma as a ‘law’ of information transfer in biological systems not as a statement about specific reaction paths. It is remarkable that, although evolution failed to find ways to reverse transcription, it has found means to circumvent this irreversibility through completely different mechanism, and so after all, to reverse the direction of the information flow. As for the “more trivial” aspect, it is certainly indisputable that the diversity of RNA roles in biology, in particular in shaping genomes
via retrotransposition, was vastly under-appreciated 40 years ago (and might not be fully appreciated yet).