The replication machinery of LUCA
A Von Neumann self-replicating machine has never been constructed because it is too complicated.
R. A. Freitas (2004): Penrose, quoting Kemeny, complained that the body of the von Neumann kinematic machine “would be a box containing a minimum of 32,000 constituent parts and comprise 150,000 [bits] of information.” Macroscale kinematic replicators will require a great deal of effort to design and to build, which may explain why so few working devices have been constructed to date,* despite popular interest.
Man, with all its intelligence, has failed. But, if abiogenesis is true, the emergence of self-replicating cell with a minimum of one million bits of information happened from randomly distributed, nonreplicating components by entirely non-intelligent unguided means.
E.V. Koonin (2020): Origin of DNA replication is an enigma because the replicative DNA polymerases (DNAPs) are not homologous among the three domains of life..DNA replication is a central process for all living cells. Therefore, it is astonishing that the key enzymes involved in DNA replication, in particular, the replicative DNA polymerases (rDNAP), are unrelated among the 3 domains of life, Bacteria, Archaea, and Eukarya.
Comment: This might be astonishing under a evolutionary model with a universal common ancestor, but not under the intelligent design hypothesis, where the designer made the different replication machinery tailored to each of the three domains of life.
This diversity of the replication machineries sharply contrasts with the conservation of the proteins involved in the other key processes of information transfer, namely, transcription and translation, as well as some key metabolic processes, such as nucleotide biosynthesis. The lack of conservation of the rDNAPs and some other key components of the replication machinery, such as helicases and primases, complicates the reconstruction of the replicative apparatus of the ancestral life forms.
There are several families of DNA polymerases that are involved in replication, repair, or both types of processes. The replicative DNAPs of bacteria, archaea, and eukaryotes belong to 3 distinct protein families, and the core catalytic domains of these 3 DNAPs are unrelated to each other, i.e., adopt different protein folds as their catalytic cores and therefore are unlikely to share common ancestry.
The great majority of dsDNA viruses that infect either prokaryotes or eukaryotes and encode their own rDNAPs have the B family polymerase (PolB) that is also responsible for the replication in eukaryotes (Table above). Archaea encode multiple PolB copies, and with the exception of members of the order Crenarchaeota and some thermophilic members of the Thaumarchaeota, also the distinct family D DNAP (PolD). In archaea that possess both DNAPs, it has been recently demonstrated that PolD, rather than PolB, is responsible for the synthesis of both DNA strands. The structure of PolD has been recently solved, resulting in a surprising discovery that the catalytic core of PolD is homologous to that of the large subunits of the DNA-directed RNA polymerases (RNAPs) that are responsible for transcription in all three domains of life and many large DNA viruses. These findings seem to shed unexpected light on the evolution of the replication machineries in the three domains of life as well as viruses. They might even help to infer the nature of the replication machinery in the LUCA suggesting an evolutionary scenario in which PolD takes the central stage as the ancestral replicative polymerase. 1
Massimo Di Giulio (2021): Cellular replicative DNA polymerases (Pols) are classified into three families of non-homologous enzymes, PolD, PolC, and PolB, which synthesize DNA at the replication forks of archaea, bacteria and eukaryotes, respectively. The primases and the main helicases involved in replication also appear to be non-homologous. In particular, while euryarchaeota possess PolD, crenoarchaeota, some thaumarchaeota and eukaryotes have, on the contrary, PolB, which results be present in all domains of life and in viruses. In particular, the diversity of PolBs in archaea is striking. This led to consider that the archaeal ancestor might have had other PolBs. Nevertheless, it would appear that only PolDs are directly involved in DNA replication in archaea, while PolBs perform this function in crenoarchaeota. In the case of eukaryotes, the wide distribution of PolBs in most eukaryotic supergroups would seem to indicate that the DNA of the eukaryotic ancestor was replicated by this enzyme. On the other hand, as far as bacteria are concerned, it would seem that their ancestor did not have PolB, so that this ancestor replicated its DNA by PolC, because this enzyme is present in all phyla of contemporary bacteria. Thus, the phylogenetic distribution of DNA polymerases led to the following suggestions. A first hypothesis is that despite the presence of all these DNA polymerases in archaea, the true DNA polymerase of archaea is PolD - which was present in the LUCA - and that, instead, PolB was introduced into crenoarchaeota and eukaryotes late, and only later than the LUCA stage. Another conjecture is that the LUCA might have replicated its DNA with PolB, that is to say, the LUCA would have had a PolB-based DNA replication machinery. Other interpretations concern the hypothesis that the LUCA had an RNA genome. This would have the advantage of explaining the remarkable heterogeneity of the replication apparatuses of archaea, bacteria and eukaryotes because the replication is not homologous among the domains of life, implying a late RNA- > DNA transition. Namely, the late appearance of DNA only in ancestors of the major phyletic lineages would be in full agreement with the fact that three completely different replicative DNA polymerases are actually operative in the three different domains of life.
Why the late origin of DNA would imply that the corresponding evolutionary stage belonged to a progenote: some criticisms of what reported in the literature
The RNA world hypothesis suggests that RNA appeared, as genetic material, before DNA. If so, then the RNA- > DNA transition, regardless of the evolutionary stage in which it occurred, should have caused quite a few problems for the biological entity in which it took place, in the sense that a high “noise” should have been associated with this transition. Indeed, it seems legitimate to me to think that the RNA- > DNA transition is to be considered as one of the major evolutionary transitions because it would ultimately lead to the birth of chromosomes. However, it seems to me that the passage from RNA to DNA must have necessarily generated a high noise which would seem to be, not only more easily tolerated by a protocellular stage compared to the cellular one, but indeed it would seem to be precisely expression, specifically, of such a protocellular stage. In other words, the birth of the deep fundamental traits such as the genetic code, the ribosome, the cell membrane, the DNA - that is, the establishment of the cellular state - would seem that it must necessarily not only come from a condition, i.e. to have passed through an evolutionary stage with high noise, but to be in itself an expression of a high level of protocellular noise, reflecting precisely the incompleteness of the evolution of the different “apparatuses” and such as to identify that stage as that of a progenote. All this simply because it was associated with the primary birth of fundamental genetically profound traits of the cell. That is to say, at the origin of all genetically deep cellular traits we would necessarily have to associate an evolutionary stage with a high protocellular noise simply because these traits were originating for the first time. The primary origin, per se, would necessarily seem to imply the presence of a high noise because we cannot think of these origins as having occurred in “cellular” conditions because cells did not yet exist (!) but were indeed originating, obviously implying a time and modalities typical of a progenote not certainly those of a complete cell.
Thus, if we recognize DNA as a genetically deep fundamental trait, then we should automatically identify the corresponding evolutionary stage as that which belonged to a progenote. Furthermore, if the RNA- > DNA transition consisted, for example, of transforming a genome made of pieces of RNA (i.e., fragmented) into a continuous genome made of DNA, then this should have generated a high noise not only because part of the genetic information contained in RNAs should have been lost but, above all because this transition would appear quite complex implying the presence of a high transitional noise. In conclusion, it could turn out that if the replication apparatus of the life domains were not homologous, then we should attribute to ancestors of these domains the qualification of progenotes, precisely because the origin of the replication apparatus should be a genetically deep fundamental trait, among other things attributable to one of the major evolutionary transitions that occurred in the evolution of cellular structures. Thus, I find it extremely curious that Forterre though acknowledging that models involving the transformation of one domain into another “have little credibility”, nevertheless he uses the expression Last Universal Cellular Ancestor - on the contrary, here I use the other expression, Last Universal Common Ancestor (= LUCA) - implying that LUCA would be a cell but not a protocell. Since the LUCA was either a cell or a protocell, then the expression last universal cellular ancestor should not be used by Forterre as he believes that it is impossible to transform one cellular domain into another completely different. That is to say, if LUCA had been a cell - as understood in the expression last universal cellular ancestor - then it could not be transformed into different cellular domains, because the transformation of one domain into another - that is, the transformation of a fundamental cell type into an another fundamental type - according to Forterre and other authors, has little credibility. More generally, I also find it unlikely that Koonin et al. (2020) while recognizing that the replication apparatus among the domains of life is not homologous, because the replicative DNA polymerases are not homologous among bacteria, archaea and eukaryotes, nevertheless they also use the expression last universal cellular ancestor meaning that the LUCA was a cell. They, reinforcing the hypothesis that the origin of DNA replication is a very late event - namely, probably occurred in the main cellular lineages because the replication apparatus is not homologous among the domains of life – what would contradict the hypothesis that LUCA was a cell. Indeed, as shown above, a late origin of DNA in the main cellular lineages would instead imply a progenotic LUCA, because the late appearance of the DNA - that is to say, a fundamental trait that is genetically profound - would be associated with a high noise that would probably define that this evolutionary stage belonged to a progenote. [(Obviously, what has already been said above also applies here, namely, that the transformation of a cellular LUCA into the different cellular domains - archaea, bacteria and eukaryotes - is not evolutionarily hypothesized, that is, permitted. All this has also led to the suggestion that LUCA might have had an RNA genome, which would imply late RNA- > DNA transitions in the main lineages, which in turn would imply the presence of a late progenotic stage in ancestors of the life domains because a high transitional noise, typical of progenotes, should be associated with the RNA- > DNA transition. On the other hand, it would not be expected in the context of a cellular LUCA – as, on the contrary, argued by Koonin et al. - because the modification of a genetically deep trait, in a cellular stage, would absolutely not be expected. However, displacements have been observed among proteins of the replication apparatus of viruses, plasmids and their hosts. Although displacements of individual proteins of the replication apparatus have certainly occurred in a cellular context, nevertheless the fact that we are able to clearly recognizing the different fundamental types of replication apparatuses, would imply that these apparatuses have remained relatively stable throughout the course of evolution. On the other hand, it would seem to strengthen the hypothesis that the model of Koonin et al. should be replaced by a model based on a progenotic state for LUCA - but not a genotic state - because the presumed stability of the replication apparatus would imply that its fixation occurred in a progenote but not in a complete cell because in a cell its change would not - by hypothesis - evidently be expected because it is too radical and therefore not tolerated. Finally, I believe that the observation that unlike DNA polymerases, RNA polymerases are homologous among the domains of life, supports the hypothesis that DNA appeared very late in cellular evolution, that is to say, in ancestors of the main lineages of life. Indeed, while it would seem natural to believe that since the function of RNA polymerases is linked to the transcription of DNA, this would imply that the DNA must have already been present when the transcription evolved and, hence, the RNA polymerases originated. Therefore, it would seem a contradiction that RNA polymerases are homologous while DNA polymerases are not; precisely because if DNA had to already be present - to be transcribed - then it would be expected that DNA replication should have originated before transcription and therefore that DNA polymerases should have been older than RNA polymerases. This would have the consequence that DNA polymerases should be homologous because they are older than RNA polymerases. In other words, DNA polymerases, according to this reasoning - being older than RNA polymerases - would impose their own homology relationship precisely because the longer time available for the evolution of a character would imply a more extensive homology relationship in organisms. This is because a longer time would have determined a deeper origin of that character and therefore a more widespread homology relationship for that particular character, on the tree of life. This reasoning is correct but it does not take into account “the fact” that it is the RNA that evolutionarily appeared before the DNA.
Conclusion
The late origin of replicative DNA polymerases together with the late origin of tRNA genes would corroborate the hypothesis of the late appearance of DNA, which would imply that ancestors of the domains of life were progenotes. The lack of homology among the replicative DNA polymerases of the three life domains, implying a late origin of the DNA - in the main lineages - would be in full agreement with the late origin of tRNA genes which would also imply, equivalently, a late appearance for DNA only in the domains of life. Therefore, there would be two completely independent observations converging towards the same conclusion of the late appearance of DNA independently and only in the domains of life. This in turn would imply a widespread progenotic state during the formation of the main lineages of the tree of life. This example of consilience would seem to represent a formidable corroboration in favor of the hypothesis that the LUCA and ancestors of the domains of life were progenotes. 2
My comment: If LUCA was a progenote, then the transition to the three domains of life would mean convergent evolution of DNA, and the replication machinery in three separate events. And in parallel, as well, the evolution of DNA replication in Viruses. The attempt to keep the model of a LUCA creates enormous difficulties, and hurdles, that are enormously steep to overcome by evolution, and are therefore, an unlikely scenario. Since Koonin et al already openly admit polyphyly of Viruses, why not admit that life was also created separately - by God? and there was never a universal common ancestor?
1. Eugene V. Koonin: The replication machinery of LUCA: common origin of DNA replication and transcription 09 June 2020
2. Massimo Di Giulio: The late appearance of DNA, the nature of the LUCA and ancestors of the domains of life 19 December 2020
A Von Neumann self-replicating machine has never been constructed because it is too complicated.
R. A. Freitas (2004): Penrose, quoting Kemeny, complained that the body of the von Neumann kinematic machine “would be a box containing a minimum of 32,000 constituent parts and comprise 150,000 [bits] of information.” Macroscale kinematic replicators will require a great deal of effort to design and to build, which may explain why so few working devices have been constructed to date,* despite popular interest.
Man, with all its intelligence, has failed. But, if abiogenesis is true, the emergence of self-replicating cell with a minimum of one million bits of information happened from randomly distributed, nonreplicating components by entirely non-intelligent unguided means.
E.V. Koonin (2020): Origin of DNA replication is an enigma because the replicative DNA polymerases (DNAPs) are not homologous among the three domains of life..DNA replication is a central process for all living cells. Therefore, it is astonishing that the key enzymes involved in DNA replication, in particular, the replicative DNA polymerases (rDNAP), are unrelated among the 3 domains of life, Bacteria, Archaea, and Eukarya.
Comment: This might be astonishing under a evolutionary model with a universal common ancestor, but not under the intelligent design hypothesis, where the designer made the different replication machinery tailored to each of the three domains of life.
This diversity of the replication machineries sharply contrasts with the conservation of the proteins involved in the other key processes of information transfer, namely, transcription and translation, as well as some key metabolic processes, such as nucleotide biosynthesis. The lack of conservation of the rDNAPs and some other key components of the replication machinery, such as helicases and primases, complicates the reconstruction of the replicative apparatus of the ancestral life forms.
There are several families of DNA polymerases that are involved in replication, repair, or both types of processes. The replicative DNAPs of bacteria, archaea, and eukaryotes belong to 3 distinct protein families, and the core catalytic domains of these 3 DNAPs are unrelated to each other, i.e., adopt different protein folds as their catalytic cores and therefore are unlikely to share common ancestry.
The great majority of dsDNA viruses that infect either prokaryotes or eukaryotes and encode their own rDNAPs have the B family polymerase (PolB) that is also responsible for the replication in eukaryotes (Table above). Archaea encode multiple PolB copies, and with the exception of members of the order Crenarchaeota and some thermophilic members of the Thaumarchaeota, also the distinct family D DNAP (PolD). In archaea that possess both DNAPs, it has been recently demonstrated that PolD, rather than PolB, is responsible for the synthesis of both DNA strands. The structure of PolD has been recently solved, resulting in a surprising discovery that the catalytic core of PolD is homologous to that of the large subunits of the DNA-directed RNA polymerases (RNAPs) that are responsible for transcription in all three domains of life and many large DNA viruses. These findings seem to shed unexpected light on the evolution of the replication machineries in the three domains of life as well as viruses. They might even help to infer the nature of the replication machinery in the LUCA suggesting an evolutionary scenario in which PolD takes the central stage as the ancestral replicative polymerase. 1
Massimo Di Giulio (2021): Cellular replicative DNA polymerases (Pols) are classified into three families of non-homologous enzymes, PolD, PolC, and PolB, which synthesize DNA at the replication forks of archaea, bacteria and eukaryotes, respectively. The primases and the main helicases involved in replication also appear to be non-homologous. In particular, while euryarchaeota possess PolD, crenoarchaeota, some thaumarchaeota and eukaryotes have, on the contrary, PolB, which results be present in all domains of life and in viruses. In particular, the diversity of PolBs in archaea is striking. This led to consider that the archaeal ancestor might have had other PolBs. Nevertheless, it would appear that only PolDs are directly involved in DNA replication in archaea, while PolBs perform this function in crenoarchaeota. In the case of eukaryotes, the wide distribution of PolBs in most eukaryotic supergroups would seem to indicate that the DNA of the eukaryotic ancestor was replicated by this enzyme. On the other hand, as far as bacteria are concerned, it would seem that their ancestor did not have PolB, so that this ancestor replicated its DNA by PolC, because this enzyme is present in all phyla of contemporary bacteria. Thus, the phylogenetic distribution of DNA polymerases led to the following suggestions. A first hypothesis is that despite the presence of all these DNA polymerases in archaea, the true DNA polymerase of archaea is PolD - which was present in the LUCA - and that, instead, PolB was introduced into crenoarchaeota and eukaryotes late, and only later than the LUCA stage. Another conjecture is that the LUCA might have replicated its DNA with PolB, that is to say, the LUCA would have had a PolB-based DNA replication machinery. Other interpretations concern the hypothesis that the LUCA had an RNA genome. This would have the advantage of explaining the remarkable heterogeneity of the replication apparatuses of archaea, bacteria and eukaryotes because the replication is not homologous among the domains of life, implying a late RNA- > DNA transition. Namely, the late appearance of DNA only in ancestors of the major phyletic lineages would be in full agreement with the fact that three completely different replicative DNA polymerases are actually operative in the three different domains of life.
Why the late origin of DNA would imply that the corresponding evolutionary stage belonged to a progenote: some criticisms of what reported in the literature
The RNA world hypothesis suggests that RNA appeared, as genetic material, before DNA. If so, then the RNA- > DNA transition, regardless of the evolutionary stage in which it occurred, should have caused quite a few problems for the biological entity in which it took place, in the sense that a high “noise” should have been associated with this transition. Indeed, it seems legitimate to me to think that the RNA- > DNA transition is to be considered as one of the major evolutionary transitions because it would ultimately lead to the birth of chromosomes. However, it seems to me that the passage from RNA to DNA must have necessarily generated a high noise which would seem to be, not only more easily tolerated by a protocellular stage compared to the cellular one, but indeed it would seem to be precisely expression, specifically, of such a protocellular stage. In other words, the birth of the deep fundamental traits such as the genetic code, the ribosome, the cell membrane, the DNA - that is, the establishment of the cellular state - would seem that it must necessarily not only come from a condition, i.e. to have passed through an evolutionary stage with high noise, but to be in itself an expression of a high level of protocellular noise, reflecting precisely the incompleteness of the evolution of the different “apparatuses” and such as to identify that stage as that of a progenote. All this simply because it was associated with the primary birth of fundamental genetically profound traits of the cell. That is to say, at the origin of all genetically deep cellular traits we would necessarily have to associate an evolutionary stage with a high protocellular noise simply because these traits were originating for the first time. The primary origin, per se, would necessarily seem to imply the presence of a high noise because we cannot think of these origins as having occurred in “cellular” conditions because cells did not yet exist (!) but were indeed originating, obviously implying a time and modalities typical of a progenote not certainly those of a complete cell.
Thus, if we recognize DNA as a genetically deep fundamental trait, then we should automatically identify the corresponding evolutionary stage as that which belonged to a progenote. Furthermore, if the RNA- > DNA transition consisted, for example, of transforming a genome made of pieces of RNA (i.e., fragmented) into a continuous genome made of DNA, then this should have generated a high noise not only because part of the genetic information contained in RNAs should have been lost but, above all because this transition would appear quite complex implying the presence of a high transitional noise. In conclusion, it could turn out that if the replication apparatus of the life domains were not homologous, then we should attribute to ancestors of these domains the qualification of progenotes, precisely because the origin of the replication apparatus should be a genetically deep fundamental trait, among other things attributable to one of the major evolutionary transitions that occurred in the evolution of cellular structures. Thus, I find it extremely curious that Forterre though acknowledging that models involving the transformation of one domain into another “have little credibility”, nevertheless he uses the expression Last Universal Cellular Ancestor - on the contrary, here I use the other expression, Last Universal Common Ancestor (= LUCA) - implying that LUCA would be a cell but not a protocell. Since the LUCA was either a cell or a protocell, then the expression last universal cellular ancestor should not be used by Forterre as he believes that it is impossible to transform one cellular domain into another completely different. That is to say, if LUCA had been a cell - as understood in the expression last universal cellular ancestor - then it could not be transformed into different cellular domains, because the transformation of one domain into another - that is, the transformation of a fundamental cell type into an another fundamental type - according to Forterre and other authors, has little credibility. More generally, I also find it unlikely that Koonin et al. (2020) while recognizing that the replication apparatus among the domains of life is not homologous, because the replicative DNA polymerases are not homologous among bacteria, archaea and eukaryotes, nevertheless they also use the expression last universal cellular ancestor meaning that the LUCA was a cell. They, reinforcing the hypothesis that the origin of DNA replication is a very late event - namely, probably occurred in the main cellular lineages because the replication apparatus is not homologous among the domains of life – what would contradict the hypothesis that LUCA was a cell. Indeed, as shown above, a late origin of DNA in the main cellular lineages would instead imply a progenotic LUCA, because the late appearance of the DNA - that is to say, a fundamental trait that is genetically profound - would be associated with a high noise that would probably define that this evolutionary stage belonged to a progenote. [(Obviously, what has already been said above also applies here, namely, that the transformation of a cellular LUCA into the different cellular domains - archaea, bacteria and eukaryotes - is not evolutionarily hypothesized, that is, permitted. All this has also led to the suggestion that LUCA might have had an RNA genome, which would imply late RNA- > DNA transitions in the main lineages, which in turn would imply the presence of a late progenotic stage in ancestors of the life domains because a high transitional noise, typical of progenotes, should be associated with the RNA- > DNA transition. On the other hand, it would not be expected in the context of a cellular LUCA – as, on the contrary, argued by Koonin et al. - because the modification of a genetically deep trait, in a cellular stage, would absolutely not be expected. However, displacements have been observed among proteins of the replication apparatus of viruses, plasmids and their hosts. Although displacements of individual proteins of the replication apparatus have certainly occurred in a cellular context, nevertheless the fact that we are able to clearly recognizing the different fundamental types of replication apparatuses, would imply that these apparatuses have remained relatively stable throughout the course of evolution. On the other hand, it would seem to strengthen the hypothesis that the model of Koonin et al. should be replaced by a model based on a progenotic state for LUCA - but not a genotic state - because the presumed stability of the replication apparatus would imply that its fixation occurred in a progenote but not in a complete cell because in a cell its change would not - by hypothesis - evidently be expected because it is too radical and therefore not tolerated. Finally, I believe that the observation that unlike DNA polymerases, RNA polymerases are homologous among the domains of life, supports the hypothesis that DNA appeared very late in cellular evolution, that is to say, in ancestors of the main lineages of life. Indeed, while it would seem natural to believe that since the function of RNA polymerases is linked to the transcription of DNA, this would imply that the DNA must have already been present when the transcription evolved and, hence, the RNA polymerases originated. Therefore, it would seem a contradiction that RNA polymerases are homologous while DNA polymerases are not; precisely because if DNA had to already be present - to be transcribed - then it would be expected that DNA replication should have originated before transcription and therefore that DNA polymerases should have been older than RNA polymerases. This would have the consequence that DNA polymerases should be homologous because they are older than RNA polymerases. In other words, DNA polymerases, according to this reasoning - being older than RNA polymerases - would impose their own homology relationship precisely because the longer time available for the evolution of a character would imply a more extensive homology relationship in organisms. This is because a longer time would have determined a deeper origin of that character and therefore a more widespread homology relationship for that particular character, on the tree of life. This reasoning is correct but it does not take into account “the fact” that it is the RNA that evolutionarily appeared before the DNA.
Conclusion
The late origin of replicative DNA polymerases together with the late origin of tRNA genes would corroborate the hypothesis of the late appearance of DNA, which would imply that ancestors of the domains of life were progenotes. The lack of homology among the replicative DNA polymerases of the three life domains, implying a late origin of the DNA - in the main lineages - would be in full agreement with the late origin of tRNA genes which would also imply, equivalently, a late appearance for DNA only in the domains of life. Therefore, there would be two completely independent observations converging towards the same conclusion of the late appearance of DNA independently and only in the domains of life. This in turn would imply a widespread progenotic state during the formation of the main lineages of the tree of life. This example of consilience would seem to represent a formidable corroboration in favor of the hypothesis that the LUCA and ancestors of the domains of life were progenotes. 2
My comment: If LUCA was a progenote, then the transition to the three domains of life would mean convergent evolution of DNA, and the replication machinery in three separate events. And in parallel, as well, the evolution of DNA replication in Viruses. The attempt to keep the model of a LUCA creates enormous difficulties, and hurdles, that are enormously steep to overcome by evolution, and are therefore, an unlikely scenario. Since Koonin et al already openly admit polyphyly of Viruses, why not admit that life was also created separately - by God? and there was never a universal common ancestor?
1. Eugene V. Koonin: The replication machinery of LUCA: common origin of DNA replication and transcription 09 June 2020
2. Massimo Di Giulio: The late appearance of DNA, the nature of the LUCA and ancestors of the domains of life 19 December 2020
