Guest Contributor posted Entry 1247 on August 30, 2005 10:58 PM.
Trackback URL: http://www.pandasthumb.org/cgi-bin/mt/mt-tb.fcgi/1245

by Mike Syvanen

[Dr. Michael Syvanen is a professor studying molecular genetics in the Department of Medical Microbiology at the University of California, Davis, and has been an advocate since the early-80s of an idea that has gained considerable support over the last few years - that much evolution is not tree-shaped, but net-shaped. That is, that genes cross taxonomic lineages. Since many attacks on evolution claim we should “teach the controversy”, we at Panda’s Thumb thought it might be nice to present an *actual* controversy in science. Discussion is welcomed. Here, at least.]

It has been over 30 years since the suggestion that horizontal gene transfer (HGT) may have been a factor in the evolution of life entered the literature. Initially these speculations were based on discoveries made in medical microbiology; namely that genes for resistance to antibiotics were found to move from one bacterial pathogen to another. This discovery was so unexpected and contrary to accepted genetic principles that though announced in Japan in 1959 (1,2) it was not generally recognized in the west for another decade. Speculations that HGT may have been a bigger factor in the evolution of life was inviting because it offered broad explanations for a variety of biological phenomena that have interested and puzzled biologist for over the last century and a half. These were problems that had been raised by botanists that have puzzled over the evolution of green plants (3) as well as by paleontologists that recorded macroevolutionary trends (4) in the fossil record that were often difficult to reconcile with the New Synthesis that merged Darwin’s thinking with Mendelian genetics. However, outside of the field of bacteriology this exercise did not really attract that much attention until the late 1990s at which time there was a major influx of data indicating that HGT had been very pervasive in early life. Namely, complete genome sequences began to appear. Simple examination of these sequences showed beyond any doubt that horizontal gene transfer was indeed a major factor in the evolution of modern bacterial, Archael and Eukaryotic genomes.

As an example for how profound the notion of HGT has changed our thinking concerns the notion of the last universal common ancestor (LUCA). This is an idea that was central to the hypothesis that life shared common ancestors. Though the idea of common ancestry remains valid (indeed evidence for common ancestry is everywhere in the sequence of our genes) there is no longer a need to postulate that all life evolved from a single last universal common ancestor. Rather, we can entertain common descent from multiple ancestors.

luca

Figure 1 Universal tree of life and two alternatives.

Bacteria contain many deeply rooted branchs, here we include two groups which are shown as the gram (-) or more accurately known as proteobacteria and gram (+) or the low GC gram (+) bacteria. A shows the so called universal tree that is supported by the rRNA sequences. B shows the relationships found between a very large number of genes involved in metabolism and biosynthesis. C simply shows the remaining four taxa relationship for which very few genes seem to follow.

The notion that all life passed through a single interbreeding bottleneck is still probably believed to be true by most people who think about this problem. The reason is simple. There are many genes involved in information processing (i.e. DNA replication, RNA transcription and protein synthesis) whose molecular homologs are found in all three major domains of life. Furthermore, when the sequences of these genes are submitted to phylogenetic analysis they more or less support the following relationship – the Archaea and Eukaryotes define a branch to the exclusion of a bacterial branch and a single line links both. Figure 1a shows this relationship. The figure shows an unrooted tree with four of life’s major groups. These are the Archaea, the Eukaryotes and two of the major groups of bacteria. The Archael/Eukaryote branch, by definition, implies the existence of a common ancestor for these two groups and further we can infer that a point on the line leading to the bacterial branch that represents the last common ancestor of all life. Thus we can say there is empirical support for the existence of the last common ancestor. I mentioned above that this scenario is more or less supported by the informational genes. The striking finding is that other genes common to the three major kingdoms frequently show exceptions to these relationships.

When it comes to the genes for energy metabolism Eukaryotes and gram-negative bacteria are usually more closely related to one another than they are to the Archaea and other bacteria (as in Fig 1B). These genes are thought to have become associated with the Eukaryotic cell through the endosymbiote that eventually gave rise to the mitochondrion (5,6,7). In green plants we can also trace the ancestry of many genes involved in carbon fixation, photosynthesis as well as other metabolic processes to cyanobacteria, the endosymbiote host that gave rise to the chloroplast. For many of the biosynthetic pathways the relevant genes yield even more complex relationships. Thus we have arrived at the current situation that is accepted by most – there remain a few genes (almost all associated with the most basic genetic informational processing) that reflect an evolutionary history that goes back to some very primitive LUCA, but that superimposed over the remnants of that primitive ancestor in modern genomes are numerous examples of subsequent horizontal gene transfer events.

The above is a good model and it requires good reasons to reject it. To begin, not all of the informational molecular homologues support the simple phylogenetic pattern outlined above. Even here there are some exceptions. These exceptions have been dealt with in one of two ways – first in some cases it can be argued that there is insufficient amount of sequence to rigorously support the true branch relationships (i.e. sequence noise or homoplasy is hiding the true pattern) or alternatively, these are informational genes that also have been involved in HGT events. Though some of the cases are still open to debate there are a number of cases where it is simplest to conclude that some of the informational genes have been involved in HGT events, this is especially true for some of the amino acid-tRNA charging enzymes (8); these enzymes are intimately involved in translating the genetic and hence are central to information processing. Once we reach this point then it is no longer possible to argue that biochemically complex processes such as protein synthesis are too complicated to have their genes being involved in HGT events; a position that was held at least up until 1998. In fact Woese (9) suggested there existed in the very primitive cells a less functionally constrained protein synthesis machinery that permitted some HGT events of these components thereby accounting for the few exceptions. In this formulation a LUCA at least implicitly remains in the model. But evidence for the LUCA is greatly reduced, at least with respect to the number of genes found in modern genomes that can be directly traced back to the LUCA via exclusive vertical evolution. In 1982 it was automatic to assume that because a biochemical process was found in all of modern life, than that process must represent evidence for the one interbreeding population of the LUCA. Now we know that many of the universal biochemical processes have moved horizontally multiple times. Thus today we have a greatly truncated LUCA from what we believed just a decade ago.

When speculating on the nature of the LUCA it is generally accepted that it must have contained the modern universal genetic code since that is a feature shared by all life. However, even if we accept the existence of this LUCA there are a variety of reasons to believe that the LUCA itself was the product of an evolutionary process that employed horizontal transfer events; this is so especially with respect to the evolution of the genetic code. It is very difficult to see how the modern genetic code could have evolved in a sequential fashion; rather the code must have evolved on separate occasions and become fused into single lineages. This problem is illustrated by considering the case of lysine-tRNA charging enzyme genes found in modern life. All life has two different completely nonhomologous enzymes. If the modern genetic code evolved in a sequential fashion, then we would have to imagine a situation where a lineage that carried one of the two enzymes evolved the second. The raises the question: what selective pressure could possibly to account for the emergence of this second enzyme when it already has one? It is much simpler to believe that the lysine enzyme evolved independently in two different lineages, which then fused to give rise to the ancestor of modern life. This is not a radical idea. Of course, if HGT is common to life after the time of LUCA then it seems not unreasonable to assume that it was common to life before the LUCA. At this point we come to the following model for evolution of life if we try to preserve the LUCA. We have multiple lineages of pre-LUCA life that are linked together by HGT events into a netted or reticulate evolutionary pattern. This leads to the LUCA. The LUCA diversifies into its many modern lineages and then these lineages are again reticulated. We then end up with a topological model that looks like an hourglass. Namely a net above that bottlenecks to the LUCA which then diversifies and yields a net below. At this point the principle of parsimony should kick in. Why encumber our model with this bottleneck. It is not only no longer necessary but is now an exceptional assumption.

There is another reason that we should jettison the LUCA. This has to do with the finding that many of the universal genes including a number that make up the genetic code, are younger than are the major kingdoms of life. That is we can be reasonably sure that life forms resembling Archaea, bacteria and some kind of primitive Eukaryote existed before 1.5 and likely before 2 billion years ago. However, parts of the genetic code are younger than that. (see papers 8 and 9 at ref 11 for the documentation). The simplest explanation is the genetic code continued to evolve after modern life diversified. If so, then the only reasonable explanation for this is that these younger members of the genetic code must have achieved their current modern and universal distribution via HGT events. Once we accept that something as complex as the genetic code can evolve and spread by HGT events, it strongly suggests that a gene encoding any function could also.

There are deep ideological reasons for believing in a LUCA that explain the reluctance of many to abandon it. In fact this reason is built directly into the most basic model of modern biology, i.e. the tree of life. The only figure in Darwin’s “Origin of Species” happens to be a tree that inevitably maps back to a single trunk. Indeed the algorithms used in phylogenetic analysis can only find a single trunk, which, of course, is how they are designed. All practicing biologists are aware of the limitations of phylogenetic modeling with its built in assumptions, but nevertheless these assumptions do cause confusion. For example, let me pose a question and ask how often there was confusion when thinking about mitochondrial eve? Isn’t it a common misperception to think at some point that all of human life could be mapped back to a single woman? When in fact all we can say is that the only surviving remnant of that distant ancestor is her mitochondrial genome, and it is extremely unlikely that any of her other genes survive in any human populations. Because of the phenomena of sexual reproduction and recombination we share genes with multiple ancestors with no need to hypothesize any individual ancestor from whom we have descended. The same reasoning should apply to the evolution of all life; because of the phenomena of horizontal gene transfer we share genes with multiple ancestors with no need to hypothesize individual species from whom we have descended (10).

Michael Syvanen

UCDavis

June 23, 2005

1. Ochiai, K., Yamanaka, T Kimura K and Sawada, O (1959) Inheritance of drug resistance (and its tranfer) between Shigella strains and Between Shigella and E.coli strains. Hihon Iji Shimpor 1861: 34 (in Japanese).

2. AKIBA T, KOYAMA K, ISHIKI Y, KIMURA S, FUKUSHIMA T. On the mechanism of the development of multiple-drug-resistant clones of Shigella. Jpn J Microbiol. 1960 Apr;4:219-27.

3. Went FW 1971 Parallel Evoluion Taxon 20:197-226

4. Reanney D 1976 Extrachromosomal elements as possible agents of adaptation and development. Bacteriol Rev. 40:552-90

5. Golding GB, Gupta RS. 1995 Protein-based phylogenies support a chimeric origin for the eukaryotic genome. Mol Biol Evol. 12:1-6.

6. Gogarten JP, Doolittle WF, Lawrence JG 2002 Prokaryotic evolution in light of gene transfer. Mol Biol Evol. 19:2226-38.

7. Doolittle WF. 1999 Lateral genomics. Trends Cell Biol. 9:M5-8.

8. Brown JR, Doolittle WF. 1999 Gene descent, duplication, and horizontal transfer in the evolution of glutamyl- and glutaminyl-tRNA synthetases. J Mol Evol. 49:485-95.

9. Woese C. (1998) The universal ancestor.Proc Natl Acad Sci U S A. 95:6854-9.

10. Zhaxybayeva O, Gogarten JP 2004 Cladogenesis, coalescence and the evolution of the three domains of life. Trends Genet. 20:291.

11. This site has a number of HGT papers in .pdf by the author :
(http://www.vme.net/hgt/)

Commenters are responsible for the content of comments. The opinions expressed in articles, linked materials, and comments are not necessarily those of PandasThumb.org. See our full disclaimer.

Comment #45820

Posted by BC on August 31, 2005 12:08 AM (e)

The same reasoning should apply to the evolution of all life; because of the phenomena of horizontal gene transfer we share genes with multiple ancestors with no need to hypothesize individual species from whom we have descended (10).

Much of the paper deals with LUCA, which would be a very ancient species. But, the conclusion, “we share genes with multiple ancestors with no need to hypothesize individual species from whom we have descended” doesn’t necessarily follow. For example, while HGT might’ve occured a great number of times in the last 2 billion years, is it reasonable to say that much, if any, has occured in humans or chimps since the human-chimpanzee clade? Further, since HGT seems to occur most readily with single-celled organisms, this means that HGT since the human-chimpanzee clade (~5 million years ago) would be even less common (on a per-year basis) than it was, say, 1 billion years ago. Thus, it would still be entirely valid to “hypothesize individual species from whom we have descended”, provided that we aren’t reaching too many tens/hundreds of millions of years into the past.

Comment #45824

Posted by H. Humbert on August 31, 2005 12:47 AM (e)

Ok, maybe not chimpanzee to human, but why not bacteria to human and chimpanzee?

I’m not formally educated in biology, so I’m asking as a layman. Is it theoretically possible for HGT to occur between single-celled organisms and the individual cells of multi-cellular organisms?

Comment #45827

Posted by Joseph O'Donnell on August 31, 2005 1:33 AM (e)

Is it theoretically possible for HGT to occur between single-celled organisms and the individual cells of multi-cellular organisms?

Possibly. The problem occurs in that this occurs in areas where bacterial genes are unlikely (in fact, almost no chance) of being passed on such as gut tissue. We probably do pick up the odd gene from bacteria or even the food we eat, but these are mostly irrelevant rare events and as they don’t affect gametes there is likely to be no evolutionary benefit.

Bacteria have been known to trade genes with plants however, such as those bacteria that form root nodules for nitrogen fixation.

Comment #45829

Posted by Hyperion on August 31, 2005 1:49 AM (e)

I’m not formally educated in biology either, so this may be a very stupid question, but even if we rule out a more recent LUCA, such as a unifier between Archea, Bacteria, and Eukaryota, wouldn’t we still have a LUCA somewhere farther back, even if it was the first true DNA-based life form?

Or is the argument that HGT implies non-causal interrelations without true descent?

Or does this imply that the first true organisms may have arisen more than once, but later gene transfers have combined their genome, thus making ancestry meaningless?

Comment #45837

Posted by Pastor Bentonit on August 31, 2005 3:48 AM (e)

H. Humbert wrote:

I’m not formally educated in biology, so I’m asking as a layman. Is it theoretically possible for HGT to occur between single-celled organisms and the individual cells of multi-cellular organisms?

It is not only theoreticaly possible, it happens all the time.

Moreover, it is very likely that the plant hormone genes present in the bacterial genome (in the Ti plasmid) have been historically acquired by HGT.

Hope that helps a lot.

Cheers,

/The Rev

Comment #45839

Posted by SEF on August 31, 2005 4:04 AM (e)

More likely RNA-based!

I think the issue is more one of when exactly something becomes recognisable as “life”. If a lot of the chemical reactions in some primordial soup are beginning to form feedback loops which nonetheless still exchange components you probably wouldn’t label them as alive. By the time you did label something as alive, there might be more than one type of it, though still related by sharing many of those original component cycles.

You probably wouldn’t call the Krebs cycle alive. Nor pyruvic acid oxidation. Yet several such things could have been going on in clusters of reactions which were going to become things you might define as life.

Comment #45856

Posted by Left_Wing_Fox on August 31, 2005 7:27 AM (e)

My biology training ended in 1994 when I couldn’t keep a lab book, but…

Is there some modern thinking that postulates that viruses might serve as an HGT mechanism between even modern species? I seem to remember reading somewhere that viruses occasionally incorporate some of the genetic code of the host cells when it hijacks them for replication, pass those genes on when jumping the species barrier, and then leave some of that genetic information behind in survivors of the viral infection.

Or am I just reading too much science fiction? ;)

Comment #45866

Posted by ruidh on August 31, 2005 9:12 AM (e)

Another possiblility? Early life on earth was “designed” by aliens with a limited toolkit. So, they reused some of their basic building blocks in several different organisms. The HGT mechanism was apparently some kind of presensile pasta-like substance in a biology lab.

Unfortunately, he got a B- from his instructor for his project.

Comment #45869

Posted by qetzal on August 31, 2005 9:25 AM (e)

So, I guess there are three ways for new genes to appear in a lineage. One is ‘intra-lineal’ and corresponds to the familiar evolutionary process of mutation plus natural selection. This obviously continues today.

Two are ‘inter-lineal.’ HGT results in one or more genes moving between lineages. This also continues today. Certainly in bacteria, as evidenced by horizontal spread of antibiotic resistance genes. Probably also in metazoa, e.g. via things like retroviruses (although it’s probably much rarer, since the HGT recipient must be a reproductive cell).

Then there’s endosymbiosis, which is genetically equivalent to complete fusion of two previously independent genomes. This clearly gave rise to (at least) chloroplasts and mitochondria. This also probably still occurs, as there are numerous examples of obligate intracellular parasites, some of which may ultimately lead to evolution of new ‘post-endosymbiotic’ species.

Still, isn’t it likely that all life can still be traced back to a single organism if one follows descent via cellular reproduction?

The simplest explanation is the genetic code continued to evolve after modern life diversified. If so, then the only reasonable explanation for this is that these younger members of the genetic code must have achieved their current modern and universal distribution via HGT events. Once we accept that something as complex as the genetic code can evolve and spread by HGT events, it strongly suggests that a gene encoding any function could also.

So these younger members of the (now) universal code evolved in some unidentified lineage, and spread by HGT to the already diverged ancestors of the Eukaryotes, Archae, and Bacteria. Either they spread to all lineages in existence at the time, or (more likely, as the author suggests) lineages that did not pick up these genes died out.

Still, the ancestral Euk, Arch, and Bact that picked up these genes can presumably also be traced back to a common cellular ancestor. Whether a LUCA really existed seems to me to depend on what one means by a LUCA. Based on the info, it seems likely there was no single LUCA that is ancestral to every extant gene, but it still seems to me that a ‘cellular’ LUCA must have existed.

We could readily re-draw the tree of life to depict all of this. Solid lines could represent descent via cellular reproduction. These would still look mostly like the familiar branching tree or bush diagrams. Rarely, two solid branches would merge again, corresponding to an endosymbiotic event.

In addition, there would be a large number of dotted lines interconnecting various branches. These would represent HGT events between lineages. (I’m almost certain I’ve already seen this type of diagram, probably here or on Pharyngula.)

The “base” of the solid lines would still represent a cellular LUCA.

Comment #45880

Posted by Jim Harrison on August 31, 2005 10:40 AM (e)

Horizontal gene transfer can occur whenever you and your mate get horizontal.

Comment #45887

Posted by Bayesian Bouffant, FCD on August 31, 2005 11:13 AM (e)

More likely RNA-based!

Brain-twisting stuff. I believe all cellular life that we know of uses DNA; the only current (that we know of) “life” without DNA would be RNA viruses. It would be great if someone found a DNA-free cell.

Presumably those early separate branches all came from the same original trunk, but it may not be possible to verify that.

Every mass extinction is a bottleneck. Every major selective advantage (development of protein synthesis, of cell membranes, of DNA, …) could have led to a ‘sweep’ of the existing population of life; coupled with the mass extinctions this leads to the possibility of multiple bottlenecks.

Um… so what? Does this lead to any actual experiments? Is it possible to see back beyond such events? We need some very clever people thinking about such questions.

Comment #45888

Posted by Timothy Chase on August 31, 2005 11:16 AM (e)

BC Wrote:
Much of the paper deals with LUCA, which would be a very ancient species. But, the conclusion, “we share genes with multiple ancestors with no need to hypothesize individual species from whom we have descended” doesn’t necessarily follow. For example, while HGT might’ve occured a great number of times in the last 2 billion years, is it reasonable to say that much, if any, has occured in humans or chimps since the human-chimpanzee clade? Further, since HGT seems to occur most readily with single-celled organisms, this means that HGT since the human-chimpanzee clade (~5 million years ago) would be even less common (on a per-year basis) than it was, say, 1 billion years ago. Thus, it would still be entirely valid to “hypothesize individual species from whom we have descended”, provided that we aren’t reaching too many tens/hundreds of millions of years into the past.

From the context of the paper, it should be fairly clear that what he is speaking of is the species of the Last Universal Common Ancestor. In essence, what he is arguing is that there may have been several species which all existing extant species share at least some of their genetic heritage from. In this view, the Last Universal Common Ancestor of all extant species would have been an ecological system or community of species where between which lateral gene transfer occured rather than a specific species itself.

With regard to lateral (“horizontal”) gene transfer in the human-chimpanzee clade, there has been plenty of it – in the form of retroviruses. During an epidemic of some exogenous retrovirus, retroviruses may be able to enter the germline, lose their ability to be transmitted through infection, and become fully endogenous – being preserved from generation to generation in the genome in proviral (DNA) form. Thus, for example, there exist roughly 50 copies of the human endogenous retrovirus HERV-K in the current day human haploid (gamete) genome, which have entered the genome at various points in the past 4-30 MYA. But this is just one example. In fact, there exist approximately 30,000 endogenous retroviruses in the human haploid genome.

Endogenous retroviruses can be particularly useful for evolutionary biologists. For example, just by tracking the presence of a few endogenous retroviruses and their mutations, it is possible to construct a fairly complete phylogenetic tree for the old world monkeys. Likewise, endogenous retroviruses made possible the identification of the hippo as the closest extant land mammal to which whales are related.

However, they can also be particularly useful as far as the host is concerned. For example, it appears that endogenous retroviruses are responsible for creating a barrier to the mother’s immune system in the placenta which protects the embryo in eutherian mammals. This made possible later encephalization which reached two pinnacles in primates (and particularly, humans) and cetacea. Moreover, a number of endogenous retroviruses are expressed in embryonic tissue development in the kidneys, testes, lungs, nervous system, and other organs. Much if not all of the retroelements in the host genomes are retroviral in origin, and play a very significant role in the consequent evolution of the host. For example, retrotransposons are responsible for gene duplication. (Similarly, transposons (which are responsible for transposition) are generally thought to have originated as sDNA (single stranded DNA) viruses.) One current hypothesis is that the ancestors of retroviruses may have been responsible for the transition from the RNA World to the DNA World.

Periodic Explosive Expansion of Human retroelements Associated with the Evolution of the Hominoid Primate
Tae-Min Kim, Seung-Jin Hong, Mun-Gan Rhyu

“Five retroelement families, L1 and L2 (long interspersed nuclear element, LINE), Alu and MIR (short interspersed nuclear element, SINE), and LTR (long terminal repeat), comprise almost half of the human genome….”

J Korean Med Sci 2004; 19: 177-85
http://jkms.kams.or.kr/2004/pdf/04177.pdf

Periodic Explosive Expansion of Human retroelements Associated with the Evolution of the Hominoid Primate
Tae-Min Kim, Seung-Jin Hong, Mun-Gan Rhyu
J Korean Med Sci 2004; 19: 177-85
http://jkms.kams.or.kr/2004/pdf/04177.pdf

Constructing primate phylogenies from ancient retrovirus sequences
Welkin E. Johnson and John M. Coffin
Proc. Natl. Acad. Sci.
Vol. 96, pp. 10254-10260, August 1999
http://www.pnas.org/cgi/content/abstract/96/18/10254

Phylogenetic relationships among cetartiodactyls based on insertions of short and long interersed elements: Hippopotamuses are the closest extant relatives of whales
Masato Nikaido, Alejandro P. Rooney, and Norihiro Okada
Proc. Natl. Acad. Sci.
Vol. 96 pp. 10261-10266, August 1999
http://www.pnas.org/cgi/content/abstract/96/18/10261

Human endogenous retroviruses in health and disease: a symbiotic perspective
Frank P. Ryan
J R Soc Med 2004; 97:560-565
December 2004
http://www.rsm.ac.uk/new/pdfs/j_art_dec04.pdf

Expressions and Functions of Human Endogenous Retroviruses in the Placenta: An Update
A. Muir, A. Lever and A. Moffet
Placenta (2004), 25, Supplement A, Trophoblast Research, Vol. 18 S16-S25
Accepted 5 January 2004

Join “Debunk Creation”
http://groups.yahoo.com/group/DebunkCreation/

Comment #45889

Posted by Bayesian Bouffant, FCD on August 31, 2005 11:23 AM (e)

In a world with HGT, use of a common genetic code presumably provides a selective advantage. There may be disadvantages as well, such as susceptibility to viruses. Thus, one version of the genetic code could sweep due not to any technical superiority, but due to a ‘market share’ victory.

I’m sure anyone with a computer understands what I’m talking about.

Comment #45891

Posted by Timothy Chase on August 31, 2005 11:47 AM (e)

Bayesian Bouffant wrote:

In a world with HGT, use of a common genetic code presumably provides a selective advantage. There may be disadvantages as well, such as susceptibility to viruses. Thus, one version of the genetic code could sweep due not to any technical superiority, but due to a ‘market share’ victory.

Actually, it appears that we have developed a defense against potentially harmful transpositions in the form of small nuclear RNA (snRNA) which evolved from retrotransposons, where the retrotransposons themselves would have evolved from endogenous retroviruses.

Join “Debunk Creation”
http://groups.yahoo.com/group/DebunkCreation/

Comment #45892

Posted by Jim Harrison on August 31, 2005 12:03 PM (e)

However the universality of the genetic code came about, computer simulations suggest that the current code is superior to all or almost all of the zillions of alternative codes. Conway Morris discusses the evidence in Life’s Solution(new ed. 2004). Andreas Wagner provides a more up-to-date run down of the evidence in Robustness and Evolvability in Living Systems (2005)

Comment #45895

Posted by Timothy Chase on August 31, 2005 12:49 PM (e)

[Dr. Michael Syvanen is a professor studying molecular genetics in the Department of Medical Microbiology at the University of California, Davis, and has been an advocate since the early-80s of an idea that has gained considerable support over the last few years - that much evolution is not tree-shaped, but net-shaped. That is, that genes cross taxonomic lineages. Since many attacks on evolution claim we should “teach the controversy”, we at Panda’s Thumb thought it might be nice to present an *actual* controversy in science. Discussion is welcomed. Here, at least.]

Other *actual* controversies which might be considered would include punctuated equilibria theory vs. gradualism, and the importance of symbiosis in evolution (e.g., “Darwin’s Blindspot” by Frank Ryan).

Comment #45920

Posted by Peter S. on August 31, 2005 3:12 PM (e)

In regards to your papers, wouldn’t it be simpler to postulate that, instead of the genetic code itself having adapted over time, the mechanisms for processing arganine and tryptophan evolved over time, and the original mechanisms were replaced by superior ones using HGT. It seems to me also that both of these processes should have left some genetic evidence around. In the first case, there should be odd organisms lurking in corners that still don’t use arganine and tryptophan, and in the second, these corner-lurkers should still be using the old, inferior, versions of these enzymes.

Even if this isn’t the case, it seems to me that adding a new amino acid to an already existing genetic code really should have left a detectable signature, as substituting the new amino acid in older genes shouldn’t always happen universally.

Comment #45921

Posted by Dave Thomas on August 31, 2005 3:14 PM (e)

Gene Transfer from Bacteria to Beetle

Nature reported on Sept. 27th, 2002 that

Researchers think they have caught a set of bacterial genes that jumped ship and relocated to the genome of a Japanese beetle. They could be the first to witness natural horizontal gene transfer between a bacterium and an animal. Although many researchers suspect this sort of gene movement occurs, no one had stumbled across evidence as direct as this before….

Source: http://www.nature.com/nsu/020923/020923-11.html
(subscription)

Comment #45933

Posted by Mike Syvanen on August 31, 2005 5:05 PM (e)

To answer some of the above queries.

BC wrote: But, the conclusion, “we share genes with multiple ancestors with no need to hypothesize individual species from whom we have descended” doesn’t necessarily follow.

response: Of course you are correct. I overstated the case here. Unclear writing on my part.

H. Humbert asked: “Is it theoretically possible for HGT to occur between single-celled organisms and the individual cells of multi-cellular organisms?

response: Most certainly yes. We do it all the time in the lab and there is evidence it has happened naturally.

Hyperion asked: wouldn’t we still have a LUCA somewhere farther back, even if it was the first true DNA-based life form?

response: Lets say there were multiple RNA based life forms and that one of them began the DNA experiment. If that experiment were shared with the other members than the answer would be no, we still do not have a LUCA.

Quetzal wrote: Still, the ancestral Euk, Arch, and Bact that picked up these genes can presumably also be traced back to a common cellular ancestor. Whether a LUCA really existed seems to me to depend on what one means by a LUCA. Based on the info, it seems likely there was no single LUCA that is ancestral to every extant gene, but it still seems to me that a ‘cellular’ LUCA must have existed.

Response: Well no. At what point, must it have existed?

T. Chase wrote: For example, while HGT might’ve occured a great number of times in the last 2 billion years, is it reasonable to say that much, if any, has occured in humans or chimps since the human-chimpanzee clade?

response: Of course there are common ancestors to modern species for which we do not need to cite any cases of HGT to explain their differences. Human and Chimps are probably an example. It is the LUCA that is doesn’t exist anymore.

Bayesian B wrote: There may be disadvantages as well, such as susceptibility to viruses.

response: Yes there is that. In the bacterial world there are specific mechanisms to repell invading DNA (called restriction systems). Just like we have mechanisms to avoid replication errors (ie DNA repair) that does not mean that mutations are not important in evolution.

J Harrison pointed out: However the universality of the genetic code came about, computer simulations suggest that the current code is superior to all or almost all [other codes].

response: I have seen these results and they are interesting indeed. Getting very speculative, I think that before the consolidation of the current universal code there were multiple codes that had emerged from the soup and that these competed with one another. So not only was there selection for a universal code, but there was selection for an efficient code.

Peter S asked: wouldn’t it be simpler to postulate that, instead of the genetic code itself having adapted over time, the mechanisms for processing arganine and tryptophan evolved over time, and the original mechanisms were replaced by superior ones using HGT.

response: This possiblity is not excluded by the data I have examined and I considered it before. However, I do not agree that it is a ‘simpler’ postulate at least for arg and trp. However, by the same criteria that date arg and trp biosynthetic pathways as new, three genes in the purine biosynthetic pathway appear to be young. In this case, your explanation is probably closer to the truth.

Comment #45939

Posted by Paley's Ghost on August 31, 2005 5:33 PM (e)

Researchers think they have caught a set of bacterial genes that jumped ship and relocated to the genome of a Japanese beetle. They could be the first to witness natural horizontal gene transfer between a bacterium and an animal. Although many researchers suspect this sort of gene movement occurs, no one had stumbled across evidence as direct as this before…. wrote:

Evolutionists have always used the existence of the same genes in a variety of organisms as proof of their amoral ontology, but once again they have been disproved by real evidence. The genetic similarities that they use to build their anti-God “Tree of Life” represent nothing more than the fact organisms occasionally eat each other. If you eat steak and then some cow genes are in you, evolutionists think that proves you came from cows–what a bunch of stupid, amoral left-wing ideologues! In a peer-reviewed paper written by evolutionists themselves1, they admit,

Since Darwin's Origin of Species, reconstructing the Tree of Life has been a goal of evolutionists, and tree-thinking has become a major concept in evolutionary biology. Practically, building the Tree of Life has proven to be tedious. Too few morphological characters are useful for conducting phylogentic analyses at the highest taxonomic level. . .For this reason, tree-makers expects a lot from gene comparisons. . .Heat map analyses were used to investigate the congruece or orthologues. . .We conclude that we simply cannot dtermine if a large portion of the genes have a common history. . .We argue that representaions other than a tree should be investigated in this case. wrote:

The authors go on to explain how mainstream evolutionists who are largely ignorant of mathematics beyond High School algebra “prove” genetic relations using Monte Carlo methods, that, like Richard Dawkins weasel program, assume the conclusion they are trying to prove. The authors use Heat Map analyses, which are a cruder and more simplistic version of Dembski’s Design inference. While these analyses disprove evolutionism as currently understood, they do not prove intellegent design theory as Dembski has done.

Evolutionists on this message board have even assterted that their religion could be disproved by demonstrating the genome is not tree-like. Well, it has been done! Watch the evolutionists squirm to evade reality, but the people of Kansas see right through them!

1Eat your hearts out!

Comment #45945

Posted by 'Rev Dr' Lenny Flank on August 31, 2005 5:56 PM (e)

Evolutionists on this message board have even assterted that their religion could be disproved by demonstrating the genome is not tree-like. Well, it has been done! Watch the evolutionists squirm to evade reality, but the people of Kansas see right through them!

“Waterloo !!!! Waterloo !!!!” Waterloo !!!!!!”

Nice parody, whoever wrote this.

But next time, to make it sound even *more* like a typical uneducated fundie, you need to add MORE CAPS and LOTS!!!!!!!!! of exclamation points.

Some spelling errors wouldn’t hurt, either.

Comment #45949

Posted by Mike Syvanen on August 31, 2005 6:11 PM (e)

Lenny

Please do not respond to the crazies in this particular thread. I realize that Panda’s Thumb most definitely argues with them, I was hoping to use this exchange to clarify some ideas that I know well meaning and educated folks have dificulty understanding.

Comment #45952

Posted by Bayesian Bouffant, FCD on August 31, 2005 6:18 PM (e)

Here’s some recent work:

Evolution of the Genetic Triplet Code via Two Types of Doublet Codons

Huan-Lin Wu, Stefan Bagby, Jean M.H. and den van Elsen
Journal of Molecular Evolution Volume 61, Number 1, Pages: 54 - 64, 19 July 2005

Abstract Explaining the apparent non-random codon distribution and the nature and number of amino acids in the ‘standard’ genetic code remains a challenge, despite the various hypotheses so far proposed. In this paper we propose a simple new hypothesis for code evolution involving a progression from singlet to doublet to triplet codons with a reading mechanism that moves three bases each step. We suggest that triplet codons gradually evolved from two types of ambiguous doublet codons, those in which the first two bases of each three-base window were read (‘prefix’ codons) and those in which the last two bases of each window were read (‘suffix’ codons). This hypothesis explains multiple features of the genetic code such as the origin of the pattern of four-fold degenerate and two-fold degenerate triplet codons, the origin of its error minimising properties, and why there are only 20 amino acids.

I suppose its possible that the two proposed doublet codes evolved in separate life forms (may not even have been cells back then?)

Comment #45953

Posted by C.J.O'Brien on August 31, 2005 6:24 PM (e)

A highly intelligent, deeply moral, (and pig-ignorant)right-wing ideologue wrote:

The genetic similarities that they use to build their anti-God “Tree of Life” represent nothing more than the fact organisms occasionally eat each other. If you eat steak and then some cow genes are in you, evolutionists think that proves you came from cows—what a bunch of stupid, amoral left-wing ideologues!

Wow.
So, take a gene that codes for a given protein. Do what’s called parsimony analysis using that gene and two species of arbitrary relatedness that share the gene. You come up with a phylogenetic tree indicating the degree of hte two species’ relatedness.
Now look at a different gene for a different protein. Do the parsimony analysis.
Marvel in awe at just how stupid your “meat’n’potatoes” hypothesis is.

Comment #45961

Posted by Paley's Ghost on August 31, 2005 6:57 PM (e)

Wow. So, take a gene that codes for a given protein. Do what’s called parsimony analysis using that gene and two species of arbitrary relatedness that share the gene. You come up with a phylogenetic tree indicating the degree of hte two species’ relatedness. Now look at a different gene for a different protein. Do the parsimony analysis. Marvel in awe at just how stupid your “meat’n’potatoes” hypothesis is. wrote:

Unlike fools like you who can barely comprehend high school algebra, you assume the only way for a creature to have a certain gene is for that creature to be related to another creature who has this gene. You’re committing the fallacy of affirming the consequent. You assume becuse there are puddles of water in the streets, it rained last night when what really happened was a bunch of commie hippies from the ACLU had an all night vigil in protest of their evolutionary lies challenged in public schools and peed on everything. Even techniques of mathematical analysis occasionally used by evolutionists show this not to be the case.

Comment #45965

Posted by Russell on August 31, 2005 7:17 PM (e)

I’m not seeing Figure 1. Is it just my browser?

Comment #45969

Posted by Flint on August 31, 2005 7:57 PM (e)

Paley’s Ghost might have a point. I’ve noticed that human vegetarians tend to look a little greenish at times, and some of them indeed put down roots. Some of them branch out into other areas, and most of them leave (sooner or later).

Comment #45973

Posted by 'Rev Dr' Lenny Flank on August 31, 2005 8:06 PM (e)

Please do not respond to the crazies in this particular thread. I realize that Panda’s Thumb most definitely argues with them, I was hoping to use this exchange to clarify some ideas that I know well meaning and educated folks have dificulty understanding.

Well-meaning and educated? IDers?

My, you are quite the optimist, huh.

Comment #45977

Posted by Hyperion on August 31, 2005 8:19 PM (e)

Nice parody, whoever wrote this.

But next time, to make it sound even *more* like a typical uneducated fundie, you need to add MORE CAPS and LOTS!!!!!!!!! of exclamation points.

Some spelling errors wouldn’t hurt, either.

Also smaller words, less commas and more run-on sentences, more dangling participles, and of course, one must remember to use at least one or two words incorrectly, such as infer when they mean imply, or preferably even using a word that means the exact opposite of what they want.

Mike: Thanks for clearing that up for me, now I think I see what you mean. Is there any way to differentiate between evolutionary change due to LGT/HGT and traditional evolutionary change through descent, in terms of constructing phylogenic relationships? I mean, are we looking at a bush with occasional strands connecting various branches, or do the horizontal connections themselves completely obscure the branches?

Comment #45982

Posted by Reed A. Cartwright on August 31, 2005 8:59 PM (e)

I’ve added figure 1.

Comment #45983

Posted by C.J.O'Brien on August 31, 2005 9:35 PM (e)

In the interest of keeping things on-topic (and sorry, I baited the troll before M. Syvanen requested that we not), I started a thread on After the Bar Closes so Paley’s Ghost can flaunt his superior intellect.

/PSA

Comment #46018

Posted by ts (not Tim) on September 1, 2005 5:04 AM (e)

Lenny

Please do not respond to the crazies in this particular thread. I realize that Panda’s Thumb most definitely argues with them, I was hoping to use this exchange to clarify some ideas that I know well meaning and educated folks have dificulty understanding.

Lenny has expressed his view on this matter in http://www.pandasthumb.org/archives/2005/08/the_faith_that.html#c45635

Nobody wants to listen to deadly-dull lectures on “pre-biotic polymer chemosynthesis” or “the homology between type III secretory apparatus and the bacterial flagellum” (yawn). This isn’t a science symposium. Don’t treat it as one. You’ll just bore everyone to death.

Comment #46104

Posted by mike syvanen on September 1, 2005 4:42 PM (e)

ts(not tim)

You don’t get it. Lenny is talking about arguing with creationist, I am trying to explain a real scientific hypothesis to a group of people who are interested in understanding this hypothesis. Two completely different problems

Comment #46106

Posted by Steviepinhead on September 1, 2005 5:02 PM (e)

Mike, I very much appreciated your explication, as I’m sure did most of us lurking out here. I for one am just as much interested in the science-fact posts as I am in the political-debate posts.

I thus have every sympathy for your hope to to be able “to explain a real scientific hypothesis to a group of people who are interested in understanding this hypothesis.”

However, you do need to realize that, as a practical matter, even when you have contributed a very science-oriented post and even when you have requested science-oriented comments, it’s just very difficult to constrain the discussion in that manner. While this is certainly a pro-evolutionary blog, it’s one that enables anonymous comments and one that has also traditionally encouraged very free-ranging discussion in those comments…

Comment #46114

Posted by 'Rev Dr' Lenny Flank on September 1, 2005 7:22 PM (e)

I’m, uh, just sitting here, shutting up. :>

Comment #46116

Posted by 'Rev Dr' Lenny Flank on September 1, 2005 7:47 PM (e)

As an example for how profound the notion of HGT has changed our thinking concerns the notion of the last universal common ancestor (LUCA). This is an idea that was central to the hypothesis that life shared common ancestors. Though the idea of common ancestry remains valid (indeed evidence for common ancestry is everywhere in the sequence of our genes) there is no longer a need to postulate that all life evolved from a single last universal common ancestor. Rather, we can entertain common descent from multiple ancestors.

Well, I’ll help steer the conversation back to the topic of the thread.

I’m not a geneticist, so perhaps I’m just horridly ignorant. But it seems to me that even if we accept a rather large amount of lateral gene transfer, right from the beginning, that would not eliminate a Last Common Ancestor. After all, the transferred genes would still all come from the same original life, though via different branches from it. And in the very beginning, it still seems as though there would be still be one original self-replicating life form, which certainly may have rapidly split to form other branches that then busily exchanged genes with each other. Yes? No?

Or is the postulate one stating that there were several separate life forms at the beginning which arose simultaneously, none of them related in any way by descent from the others? If so, there wouldn’t seem to be any requirement for them to all possess the same genetic code, no? (if they did share the same genetics, that would indicate that they shared a common ancestor, yes?). Therefore, it seems to me, one of these various genetic mechanisms (whatever they might be) would come to dominate the others, drive them out, and become the universal ancestor of everything alive today, even if it did exchange genes previously with its now-extinct compatriots.

So while the different INDIVIDUAL GENES might not have the same universal common ancestor, the ORGANISMS AS A WHOLE still would.

Yes? No?

Comment #46128

Posted by Jim Harrison on September 1, 2005 10:43 PM (e)

Physicists understand the trajectories of things as world lines in space-time. In the biological version of this representation, living things appear as world tubes in space-time, continuous regions that remain in an unbroken state of thermodynamic tension with their surroundings (more ATP than ADP, for example). Most of the tubes are exceedingly fine because they are prokaryotes, the dominant life form on the planet; but even the thicker tubes–the eurkaryotes–thin drastically at regular intervals with the alteration of hapoloid and diploid forms. Almost all of the tubes are dead ends. Asexual reproduction results in bifurcating tubes, sex in both bifurcating and anastomosing tubes. The novelty of HGT via viruses or the exchange of naked DNA between bacteria is that the genetics of one tube gets altered by another without a living, that is, continuously metabolizing, link. Presumably the tangle of threads that make up the biosphere sprang from a less diferentiated and perhaps more complexly connected mass.

I have no idea what the hcck this topological image of life on Earth is good for, but I find it somehow edifying to imagine. I’m trying to extend the idea by integrating into it the notion of niche construction since, considered thermodynamically, the margins of a living thing are not very clear cut. Living things typically either construct or chose intimate environments that are more orderly and high energy than their more distant surroundings. Which converts the originally picture of life on earth from a giant mycellum to a more complicated structure something like a tangle of insulated wires embedded in anisotropic jello.

Submitted to prove that you don’t have to be an ID supporter to have bizarre ideas.

Comment #46143

Posted by ts (not Tim) on September 2, 2005 12:56 AM (e)

ts(not tim)

You don’t get it. Lenny is talking about arguing with creationist, I am trying to explain a real scientific hypothesis to a group of people who are interested in understanding this hypothesis. Two completely different problems

That’s what I do get, and Lenny seems not to. Thus, when you say “I was hoping to use this exchange to clarify some ideas that I know well meaning and educated folks have dificulty understanding”, Lenny responds “Well-meaning and educated? IDers? My, you are quite the optimist, huh”. Lenny not only doesn’t want this this forum to be anything more than a place to bash ID trolls, but he doesn’t even recognize when they aren’t the ones being addressed.

Comment #46168

Posted by 'Rev Dr' Lenny Flank on September 2, 2005 7:14 AM (e)

(ignoring ts’s dick-waving, yet again)

(Re-asking my previous question):

I’m not a geneticist, so perhaps I’m just horridly ignorant. But it seems to me that even if we accept a rather large amount of lateral gene transfer, right from the beginning, that would not eliminate a Last Common Ancestor. After all, the transferred genes would still all come from the same original life, though via different branches from it. And in the very beginning, it still seems as though there would be still be one original self-replicating life form, which certainly may have rapidly split to form other branches that then busily exchanged genes with each other. Yes? No?

Or is the postulate one stating that there were several separate life forms at the beginning which arose simultaneously, none of them related in any way by descent from the others? If so, there wouldn’t seem to be any requirement for them to all possess the same genetic code, no? (if they did share the same genetics, that would indicate that they shared a common ancestor, yes?). Therefore, it seems to me, one of these various genetic mechanisms (whatever they might be) would come to dominate the others, drive them out, and become the universal ancestor of everything alive today, even if it did exchange genes previously with its now-extinct compatriots.

So while the different INDIVIDUAL GENES might not have the same universal common ancestor, the ORGANISMS AS A WHOLE still would.

Yes? No?

Comment #46169

Posted by Dave Harmon on September 2, 2005 7:15 AM (e)

Hmm. Are you sure that the “introduction” of a LUCA bottleneck really violates parsimony? It occurs to me that early evolution suggests a number of developments what would have naturally provided a whole series of bottlenecks by way of “sweeping the field”. Some of the possibilities:

1) The first reproducing system(s) would have spread via exponential growth, thus tending to displace alternatives.

2) The first *self-contained* reproductive system(s) (protocell) would have done much the same; not depending on the “soup” for the more complex components would be a drastic advantage.

3) Likewise for the first “predatory” cell(s), capable of killing its competitors and swiping their biomaterial. Note that the early versions of this would include any genetic material, (LGT in spades!), but some of that would then fall by the wayside.

4) The “oxygen crisis”, posed by the appearance of photosynthesis, would have knocked off a goodly number of lineages, arguably to bottleneck levels (in affected areas).

5) Parallel to all of the above, the occasional mixing of separated “life-pools”, would produce rapid consolidation through both LGT and traditional competition.

So, even if the early descent was net-like, it’s still liable to have enough bottlenecks to produce the rough appearance of an LUCA in the genetic record, even if there was never a single species that truly matched the definition.

Comment #46177

Posted by 'Rev Dr' Lenny Flank on September 2, 2005 7:52 AM (e)

So, even if the early descent was net-like, it’s still liable to have enough bottlenecks to produce the rough appearance of an LUCA in the genetic record, even if there was never a single species that truly matched the definition.

Yes. This all relates to my question above.

Can anyone help us out here?

Comment #46183

Posted by qetzal on September 2, 2005 9:23 AM (e)

Mike Syvanen wrote:

Quetzal wrote: Still, the ancestral Euk, Arch, and Bact that picked up these genes can presumably also be traced back to a common cellular ancestor. Whether a LUCA really existed seems to me to depend on what one means by a LUCA. Based on the info, it seems likely there was no single LUCA that is ancestral to every extant gene, but it still seems to me that a ‘cellular’ LUCA must have existed.

Response: Well no. At what point, must it have existed?

Sorry. Poor writing on my part.

What I meant was, a cellular LUCA must have existed unless, of course, cellular life arose more than once. This is essentially the same thing Lenny is saying.

You said in your post that we “should jettison the [concept of] the LUCA” as an exceptional and unnecessary assumption. You indicate that current data actually argues against the ‘primitive LUCA with superimposed HGT’ model. But is the evidence really that strong?

You give examples of genes that are universal, yet appear to have been acquired via HGT more recently than other universal genes. Does the data suggest that most universal genes were acquired at different times, or just some? If we look at all the universal genes for which good data is available, what fraction are at least still consistent with the LUCA hypothesis?

If that fraction is substantial, then I can’t see why a LUCA is an unnecessary and extraordinary assumption. Seems like it would still be the most parsimonious explanation for those genes.

Given what you’ve presented, I’m persuaded that the LUCA hypothesis is much less certain than I previously assumed. I’m not yet persuaded that it should be rejected. I will try to read more of the papers on your web site (I’ve only read the Trends paper so far), but I look forward to your further comments here as well.

Comment #46251

Posted by mike syvanen on September 2, 2005 3:30 PM (e)

Syntax Error: mismatched tag 'kwickxml'

Comment #46253

Posted by mike syvanen on September 2, 2005 3:38 PM (e)

Syntax Error: mismatched tag 'kwickxml'

Comment #46256

Posted by mike syvanen on September 2, 2005 3:49 PM (e)

Response to Lenny.

No you are not horridly ignorant. I am proposing something that goes directly against much of what biologists have thought since 1859. (Though those people who thought about evolution before then would not have any trouble with this idea).

I am arguing that it is no longer necessary to postulate the existence of LUCA. Part of this is that it seems likely that there were multiple life forms involving replicating cells that emerged from abiotic chemistry. Of course most of those became extinct as predation was first invented. However, modern life contains evidence as containing more than one of primitive life forms. As I argued, the two non-homologous lysine tRNA charging enzymes provide the evidence. If we are unable to explain the emergence of the first modern looking cell without citing fusions of independent, and since we know that HGT has been prevalent since it therefore follows that citing a LUCA is no longer necessary. There is nothing in what I have written to suggest that all life could not have possibly passed through a single bottleneck.

Lenny asked “Or is the postulate one stating that there were several separate life forms at the beginning which arose simultaneously, none of them related in any way by descent from the others?”

Yes and no. Yes there were several separate life forms at the begining but no it is not necessary to say none of them are unrelated to one another.

Let me try to explain this again by considering human evolution. Somewhat arbitrarily let us go back say 10 million years. There was some kind of hominoid back then that proved ancestral to humans, chimps and gorillas. Let us say that it had a total population of 100,000. A beneficial mutation occurs that via the mechanism of natural selection, this mutation eventually sweeps the population, which then characterizes the subsequent descendent population. Also imagine that elsewhere in the population, in a line independent of the first beneficial mutation, a second mutation occurs that also sweeps the population. Is that possible to have two independent, simultaneous, mutational events become fixed in the descendant population? Of course it is. Sexual reproduction and genetic recombination will make this process very likely. In fact, after a few thousand generations it is extremely likely that the only genes from that first beneficial mutant that survive in the population are those that are linked to the original beneficial mutation. Thus we do not need to postulate the existence of any bottleneck to account for the population change nor do we need to point to any specific individual in that ancestral population who is a unique ancestor.

This scenario does not deny the possibility of a small number of founder individuals, but it is not necessary to postulate their existence. Evolution would work quite well without needing to cite founders. In the case of human evolution, we can imagine the total population as never having to even decline to move from that ancestral form 10 million years ago to modern humans. In fact I would further argue that evolution would be accelerated if bottlenecks were avoided. That is, the more breeding individuals we have the greater the probability that one of them will spin off some beneficial mutation. Bottlenecks are bad because they reduce the genetic diversity increases the chances of population survival.

To summarize, if you can accept this analogy, it is a simple logical move to use this to understand my arguments against LUCA.

I hope this helps and I am not just repeating myself.

Comment #46265

Posted by mike syvanen on September 2, 2005 4:48 PM (e)

Response to Harmon

Your 1 and 2 seem to be one point. These two and number 3 certainly could have resulted in a bottleneck but if so these events would have happened long before anything resembling a modern genetic code had appeared. This is going to far back. LUCA is a hypothesized ancestor for which there is purported to be evidence in the homology of existing genes. My point is that there is no need to cite the existence of a LUCA to account for extant homologies.

4 The oxygen catastrophe would have certainly caused mass extinction but we can be reasonably certain that at least six lineages survived - ie eukaryotes, two clades of archaea and at least three bacterial.

5 and concluding paragraph. Not sure that I even disagree with these comments but do not see how they are criticisms of my postulate.

Comment #46283

Posted by ts (not Tim) on September 2, 2005 5:53 PM (e)

Lenny wrote:

(ignoring ts’s dick-waving, yet again)

I responded directly to a false charge by Mike Syvanen. You would have ignored it by not commenting; instead, you do, again, exactly what you accuse me of.

Comment #46288

Posted by 'Rev Dr' Lenny Flank on September 2, 2005 6:31 PM (e)

(ignoring ts’s dick-waving, yet AGAIN)

I am arguing that it is no longer necessary to postulate the existence of LUCA. Part of this is that it seems likely that there were multiple life forms involving replicating cells that emerged from abiotic chemistry. Of course most of those became extinct as predation was first invented. However, modern life contains evidence as containing more than one of primitive life forms.

OK, but here’s the part where I think I am not-understanding; if there were multiple life forms at the beginning, and they exchanged genes, and then subsequently one of these forms went on to dominate “whether through predation or whatever), then all subsequent life would still be descended from it, and it would hence still be a Last Common Ancestor – even IF it had previously exchanged genes with now-extinct contemporaries.

So I guess my question is (and maybe I am just missing your answer) is there an indication that any existing genetic lines MUST have come from a totally separate unrelated first ancestor, and did NOT just result from lateral transfer between different lineages that themselves descended from an earlier common ancestor?

In other words, while the individual gene lines may have come from different original progenitors (perhaps not even related to each other), if one of these original life forms went on to outcompete the others, then all existing life would STILL be descended from it (even if all the GENES they carry might come from different contemporaries).

Yes? No?

Comment #46297

Posted by mike syvanen on September 2, 2005 7:11 PM (e)

response to quetzal

What I meant was, a cellular LUCA must have existed unless, of course, cellular life arose more than once. This is essentially the same thing Lenny is saying. You said in your post that we “should jettison the [concept of] the LUCA” as an exceptional and unnecessary assumption. You indicate that current data actually argues against the ‘primitive LUCA with superimposed HGT’ model. But is the evidence really that strong? wrote:

No the evidence is really not that strong. There are a few reasonably good exceptions upon which I base the argument. But when these few examples are compared to extent of support for LUCA, they serve to undermine the whole concept.

You give examples of genes that are universal, yet appear to have been acquired via HGT more recently than other universal genes. Does the data suggest that most universal genes were acquired at different times, or just some? If we look at all the universal genes for which good data is available, what fraction are at least still consistent with the LUCA hypothesis?

If that fraction is substantial, then I can’t see why a LUCA is an unnecessary and extraordinary assumption. Seems like it would still be the most parsimonious explanation for those genes. wrote:

Many genes for information processing but certainly not all support the universal tree (fig 1a) hypothesis. However a majority of the rRNA charging enzymes do not. Outside of this group, I found that 24 out of 27 arbitrarily chosen genes supported either alternative trees (b or c) or were so recent to yield star phylogenies.

I am not denying that there exist a small number of genes, abeit important ones, that support tree 1 A, but the more I thought this through I realized that this dwindling number doesn’t require the existence of a LUCA. This follows from thinking about the implications of the fact that it is likely that HGT events are more probable (ie mechanistically simpler) between closely related species than between distantly related species. That is, homologies can be reinforced after speciation events via HGT.

Comment #46331

Posted by mike syvanen on September 3, 2005 1:21 AM (e)

Lenny pointed out

OK, but here’s the part where I think I am not-understanding; if there were multiple life forms at the beginning, and they exchanged genes, and then subsequently one of these forms went on to dominate “whether through predation or whatever), then all subsequent life would still be descended from it, and it would hence still be a Last Common Ancestor — even IF it had previously exchanged genes with now-extinct contemporaries. wrote:

We are dealing with a multi-dimensional problem here and my answers to one specific question may raise ambiguities with respect to a second question. But to directly address your question, let me answer it with a possible scenario.

The first predator life form appeared. This provided it with tremendous selective advantage. Further imagine that there are competing life forms that carry beneficial traits independent of predation (eg an enzyme that converts d amino acids from the primeval soup into l amino acids). If one of the predator’s offspring borrowed that metabolic trait and that led to its ability to out-compete its siblings then we can argue that the new predator has descended from two different ancestors.

Hope this helps. But in any case thanks for the interesting questions.

Mike Syvanen

Comment #46334

Posted by ts (not Tim) on September 3, 2005 2:25 AM (e)

(ignoring ts’s dick-waving, yet AGAIN)

Yet again you prove yourself a hypocritical liar.

Comment #46347

Posted by 'Rev Dr' Lenny Flank on September 3, 2005 6:23 AM (e)

(Once again ignoring ts’s silly dick-waving)

We are dealing with a multi-dimensional problem here and my answers to one specific question may raise ambiguities with respect to a second question. But to directly address your question, let me answer it with a possible scenario.

The first predator life form appeared. This provided it with tremendous selective advantage. Further imagine that there are competing life forms that carry beneficial traits independent of predation (eg an enzyme that converts d amino acids from the primeval soup into l amino acids). If one of the predator’s offspring borrowed that metabolic trait and that led to its ability to out-compete its siblings then we can argue that the new predator has descended from two different ancestors.

Yes, I get that part. What I am asking is this; if that new predator or its offspring (and at this stage I presume it would be single-cellular) then went on to dominate all the other contemporary life forms and drive them out, for whatever reason, then all life afterwards would still be descended from it, and only it (the other contemporary lifeforms having gone extinct). Therefore, it would still be the Last Common Ancestor of all existing life, even if it itself were not the sole original life form, and even if it itself had multiple genetic ancestors.

See what I’m saying? The GENE lines would have multiple ancestors, but the ORGANISMS would all still have the same common ancestor (which, itself, could contain different gene lines from any number of previous ancestors).

So then the “tree” of life would turn into a “mangrove” of life —- lots of different genetic lines would join in the Last Common Ancestor, which would then itself branch off in to all the current lifeforms.

The question would then become, what does the very bottom of the “mangrove roots” look like? Was there one original life form, which then diverged genetically and later exchanged genes between all these diverging branches before one of them went on to dominate and produce all existing life? Or was there *more than one independently-appearing original life form*, which then exchanged genes with each other to form the “mangrove roots”, before one of those “roots” went on to dominate and produce all the subsequent branches? I’m not sure how one would be able to tell, but it would seem to be more parsimonious to presume that there was, originally, just one life form, which diverged over a long time to produce differing branches, and these differing branches came to be dominated by just one, the Last Common Ancestor which diverged to produce everything we see today.

So how would we tell which is the case?

Maybe I can be more clear if I use some graphics:

Scenario 1 (http://www.geocities.com/lflank/scenario1.bmp)
Scenario 1, it seems to me, would be that there was more than one original life form, which then exchanged genes with each other through lateral transfer, and one of these branches comes to dominate the others and produce all currently existing life. Hence, all current life would share the same common geneological ancestor, but the ancestry of individual genes would pass through this Last Common Ancestor and trace back to different original life forms.

Scenario 2 (http://www.geocities.com/lflank/scenario2.bmp)
Scenario 2, it seems, would involve a single original life form, which diverged over time to form descendent branches which exchanged genes with each other through lateral transfer, and then one of those branches came to dominate the others, thus becoming the Last Common Ancestor of all existing life. Once again, all life today would be geneologically descended from a common ancestor, but each gene line would trace its ancestry through the Common Ancestor to different branches of the original life form (and then finally back to the single life form, which would in essence be another Last Common Ancestor).

A variation would be scenario 3 (http://www.geocities.com/lflank/scenario3.bmp)
where, if there were one original life form or several, the divergent branches would exchange genes through lateral transfer, but this time more than one branch survives to each produce different portions of the existing “tree of life”. In this scenario, existing life would not share a single common ancestor, and whether or not there was any Last Common Ancestor would depend on whether the original life consisted of just one form or several independently-appearing ones.

So, how do we tell which scenario is correct, and which sceniario does current evidence indicate?

And thanks for the interesting conversation. :>

Comment #46601

Posted by syvanen on September 5, 2005 1:24 PM (e)

response to Lenny

I should reiterate one point before proceeding into a much more speculative area. My point is that existing homologies that link all life can easily be explained without the postulate that there existed LUCA. You are talking about something that lies, at least for now, beyond existing evidence.

However, there are some conjectures that we can pursue. It is not unreasonable to suppose that the first predator was some kind of primitive eukaryotic cell (ie without mitochondria, chloroplast or even a nucleas as we understand it). It was some kind of very large cell, by comparison to bacteria, that had an ability to injest bacteria via endocytosis. Today there is no bacteria that has that ability. The closest thing to predatory bacteria are the myxobacteria that excrete digestive enzymes and absord the digested polymers. This primitve eukaryote, which has been named the kronocyte, likely had a very primitive metabolism as judged by the likely fact that today most all of its metabolic activities seem to have been derived either from cyanobacteria or proteobacteria. Its informational genes have more or less been acquired from the archaea. All that remains from the original kronocyte are those genes that code for the proteins used for mobility and endocytosis (actin, tubulin, myosin and so on) plus some RNA enzymes,a few ribosomal proteins, calmodulin, some proteolytic mechanisms, etc.

If this conjecture is correct and if bacteria were the first life on earth, then we can easily conclude that this first predator did not replace previous life. In fact, given that photosynthetic bacteria were the likely cause of the oxygen crises we can conclude that incorporation of bacterial genes into the kronocyte became essential for its own survival.

Also the idea that the first predator would erase all previous life doesn’t make ecological sense. A predator that eats 100% of its prey, turns out to a very hungry predator.

In any case, no matter how I play out these various scenarios, I have yet to hit on a situation that really requires any kind of unique bottleneck. This, of course, does not preclude such a possibility, simply that it is not a needed postulate.

Comment #46616

Posted by 'Rev Dr' Lenny Flank on September 5, 2005 3:11 PM (e)

In any case, no matter how I play out these various scenarios, I have yet to hit on a situation that really requires any kind of unique bottleneck. This, of course, does not preclude such a possibility, simply that it is not a needed postulate.

But the fact that all life uses the same genetic code would indeed seem to indicate that all life is descended from a single source. Unless there were multiple independent life forms at the beginning who all happened to hit upon the very same genetic code … ?

As for predators erasing all previous life, that wasn’t what I said. ANYTHING that gave a selective advantage could eliminate competitors and erase all previous life — better replication, more efficient genetic system with fewer mistakes, whatever. All that’s required is for one genetic line, for whatever reason, to reproduce better than all the rest. And since there’s only one system of genetics left today, that indicates that either there only ever WAS one, or any other ones that might have existed disappeared, leaving only the one that we have today.

Yes, no?

Comment #46648

Posted by Henry J on September 5, 2005 5:37 PM (e)

Re “My point is that existing homologies that link all life can easily be explained without the postulate that there existed LUCA.”

I’d think it would depend on whether a majority of the DNA supports a consistent hierarchy. If say 90% or more of the genes generate the same hierarchy, that would seem to me to imply a single common ancestor. So what’s needed to answer the question is genetic sequencing of a variety of species from all the known domains.

Henry

Comment #46670

Posted by ts (not Tim) on September 5, 2005 9:16 PM (e)

Lenny wrote:

But the fact that all life uses the same genetic code would indeed seem to indicate that all life is descended from a single source. Unless there were multiple independent life forms at the beginning who all happened to hit upon the very same genetic code … ?

Mike already addressed this:

Getting very speculative, I think that before the consolidation of the current universal code there were multiple codes that had emerged from the soup and that these competed with one another. So not only was there selection for a universal code, but there was selection for an efficient code.

Comment #46689

Posted by Vincent Torley on September 6, 2005 2:10 AM (e)

I have a query for Professor Syvanen. Professor, I would like to apologize at the outset if I have misunderstood anything you have written, as my background is mainly in philosophy, not science (I do have a B.S., but I obtained it in 1981 and my major was mathematics, not biology). My query basically boils down to this: exactly when is it legitimate to invoke homologies between organisms at the genetic level, as an argument for their common ancestry?

In your article, you wrote:

“In 1982 it was automatic to assume that because a biochemical process was found in all of modern life, than that process must represent evidence for the one interbreeding population of the LUCA. Now we know that many of the universal biochemical processes have moved horizontally multiple times.”

If I understand you rightly, you are saying that there are two possible explanations for the profound genetic and biochemical similarities between living organisms: either a “vertical” or historical explanation (common descent) or a “horizontal” one (horizontal gene transfer). If that is the case, then the fact that all living organisms share certain common genetic and biochemical features need not indicate their common ancestry. It may simply indicate the vast antiquity of their distinct lineages: they have been around for long enough to swap genetic information, and for more successfully reproducing genes to have displaced less viable alternatives.

At the same time, you robustly affirm that “the idea of common ancestry remains valid (indeed evidence for common ancestry is everywhere in the sequence of our genes)” even though “there is no longer a need to postulate that all life evolved from a single last universal common ancestor.”

All this leaves me scratching my head, for three reasons. First, I remember learning about evolution in high school and reading about it at university, and the thing that impressed me most was the unity and simplicity of the hypothesis. As Poo Bah declares in “The Mikado”, “I can trace my ancestry back to a primordial protoplasmic globule.” (I have omitted the word “atomic” from the quote, as it makes no sense in this context.) The alternative explanation (“Well, the Creator could have used the same design to make all organisms”) struck me as utterly contrived and ad hoc - in other words, the kind of explanation that scientists should rightly ignore. Historical explanations have a place in science; divine fiat does not. What worries me is that now, instead of a beautiful tree, we have an ugly net with just as many branches at the top as at the bottom. Your evolutionary account isn’t simple any more, and it no longer even attempts to unify organisms.

My second concern is that the genetic similarity between organisms is supposed to be the strongest argument for evolution, but it seems to me that you have greatly weakened it. Most of the other textbook arguments for evolution (e.g. missing links such as Archaeopteryx and the cynodonts; embryological similarities; and convergent evolution) apply mainly to animals, which represent only a tiny twig on the evolutionary tree (or net, as the case may be). While they suggest strongly that all animals had a common ancestor, they cannot establish common ancestry for all organisms. However, if the genetic similarities may be due to HGT, then the argument that all organisms share a common ancestry is robbed of much (if not all) of its force.

My third and final worry is that in your response to one correspondent (J. Harrison) you seemed to concede his/her claim that the current genetic code “is superior to all or almost all” other codes, before suggesting that there were once multiple codes which competed with one another, generating selection for an efficient universal code. What worries me is that for many years, textbooks have featured the argument that there was nothing special about the code all living things now possess - any number of viable alternatives would work equally well. But if there is indeed something inherently superior about the current code, then the best explanation for the universality of the genetic code is no longer an historical one (“the first life-form possessed it, so we have it too”) but a formal one (“other designs don’t work as well”). Certainly, the two explanations need not be mutually exclusive; however, the former seems to render the latter redundant.

Since my forte is philosophy (not science), I cannot tell you whether common ancestry is a good hypothesis or not. What I can tell you is that the latest scientific research makes it a lot harder to argue for - especially to your average John and Jane Doe, who have no more than a high school science education.

One piece of evidence that still seems to point strongly towards common ancestry is the evidence for endosymbiosis: structures in our cells (such as mitochondria) which appear to bear the hallmarks of cellular cannibalism in our distant past. I would like to ask you: does Margulis’ endosymbiotic explanation command wide assent among contemporary biologists?

Thank you for taking the time to read my lengthy missive.

Comment #46769

Posted by mike syvanen on September 6, 2005 3:02 PM (e)

response to Lenny:

ANYTHING that gave a selective advantage could eliminate competitors and erase all previous life —- better replication, more efficient genetic system with fewer mistakes, whatever. All that’s required is for one genetic line, for whatever reason, to reproduce better than all the rest. wrote:

What I have been arguing is that when such a more efficient genetic system arises and threatens to replace all previous life then those endangered competitors are faced with two choices: 1) adapt (ie most effectively by borrowing the trait from its competitor or 2) face extinction. Since evolution of complex traits works by small step by step changes, this borrowing process (HGT) is not that difficult.

Comment #46776

Posted by mike syvanen on September 6, 2005 3:36 PM (e)

response to Victor:

What worries me is that now, instead of a beautiful tree, we have an ugly net with just as many branches at the top as at the bottom. Your evolutionary account isn’t simple any more, and it no longer even attempts to unify organisms. wrote:

True it makes it a little more complicated, but I wouldn’t call it ugly. But with respect to unity, I am proposing even greater biological unity. Not only does unity remain from common descent in the past, but the unity continues to the present in each species ability to read and decipher foreign genes.

However, if the genetic similarities may be due to HGT, then the argument that all organisms share a common ancestry is robbed of much (if not all) of its force. wrote:

Common ancestry remains very central in my way of thinking. The only real difference is that we are related to more common ancestors than previously believed.

What worries me is that for many years, textbooks have featured the argument that there was nothing special about the code all living things now possess - any number of viable alternatives would work equally well. But if there is indeed something inherently superior about the current code, then the best explanation for the universality of the genetic code is no longer an historical one (”the first life-form possessed it, so we have it too”) but a formal one (”other designs don’t work as well”). Certainly, the two explanations need not be mutually exclusive; however, the former seems to render the latter redundant. wrote:

Yes that was called the frozen accident hypothesis and I thought it was a reasonable idea. Unfortunately, such are the dangers with science, once we think we understand a problem, new evidence appears that force us to consider alternatives. The text books will have to be changed because of these new studies that seem to show that the code has been (or nearly so) optimized.

Though I didn’t predict this new developement, I found it easy to incorporate it into my thinking on how the code evolved.

One piece of evidence that still seems to point strongly towards common ancestry is the evidence for endosymbiosis: structures in our cells (such as mitochondria) which appear to bear the hallmarks of cellular cannibalism in our distant past. I would like to ask you: does Margulis’ endosymbiotic explanation command wide assent among contemporary biologists? wrote:

Endosymbiosis just happens to be one of the mechanisms that promotes HGT that was important to eukaryotic evolution. Just another mechanism that is fully consistent with everything I have been saying. And yes endosymbiotic theory is today nearly universally accepted.

As a point of history, the idea is not Margulis’s, she simply resurrected Mereschovski’s ideas that were published in 1910 and widely recognized as worthy of discussion up into the 30’s. Somehow they were largely forgotten until she revived them in the 1970’s. The big breakthrough came in 1976 when Woese showed the close relationship between the ssRNA from mitochondria and chloroplast to the alpha proteobacteria and cyanobacteria respectively.

In general you should keep in mind that these problems yor are grappling with arise primarily because the full incorporation of the notion of HGT into modern evolutionary theory, will require some substantial changes in how we think about evolution. New paradigms and all that, if you like, but the thing that makes science fun.

Comment #46850

Posted by harryinharryout on September 7, 2005 11:22 AM (e)

“This problem is illustrated by considering the case of lysine-tRNA charging enzyme genes found in modern life. All life has two different completely nonhomologous enzymes. If the modern genetic code evolved in a sequential fashion, then we would have to imagine a situation where a lineage that carried one of the two enzymes evolved the second. The raises the question: what selective pressure could possibly to account for the emergence of this second enzyme when it already has one? It is much simpler to believe that the lysine enzyme evolved independently in two different lineages, which then fused to give rise to the ancestor of modern life.”

If all life has two nonhomologous enzymes what is the selective pressure to keep both functional? What ever that selective pressure is would account for sequential evolution of both.

Comment #46963

Posted by mike syvanen on September 8, 2005 12:08 AM (e)

If all life has two nonhomologous enzymes what is the selective pressure to keep both functional? What ever that selective pressure is would account for sequential evolution of both.

Well no. That is not how we think about the evolution of novelty.

Comment #47207

Posted by Dave Harmon on September 9, 2005 4:22 PM (e)

Previously and here, I’m not necessarily contesting your (Professor Synanen’s) points, so much as trying to understand the implications. Certainly I find the idea interesting; I’m not sure it isn’t moot, but the way to decide that is to explore the idea, and see what insights come out of it.

Mike Syvanen wrote:

Your 1 and 2 seem to be one point. These two and number 3 certainly could have resulted in a bottleneck but if so these events would have happened long before anything resembling a modern genetic code had appeared.

Well, you’ve confirmed in other comments that you’re not trying to disprove the LUCA, so much as to “dethrone” it. I don’t think, though, that these first three stages (replication, containment, incorporation) would be “long before” the modern genetic code. Rather, I would suspect that these stages would be exactly when this code developed and became “universal”. The various oddities found in coding, and occasional parallelism in basic functions, would then be explicable as artifacts from this early merger. Of course, the survivors at this point might well represent several primal lineages, each coalescing from the original pool of protobiotic systems. Given the advantage of being able to freely absorb each other’s material (and genes), their coding and metabolic commonalities would be homogenized by LGT-plus-selection. But these lineages might well represent (say) distinct competitive strategies in the primal ecology, so that each would turn the common “toolkit” to separate ends, and be shaped by progressively different selection pressures.

It is not unreasonable to suppose that the first predator was some kind of primitive eukaryotic cell (ie without mitochondria, chloroplast or even a nucleas as we understand it). It was some kind of very large cell, by comparison to bacteria, that had an ability to injest bacteria via endocytosis. Today there is no bacteria that has that ability. The closest thing to predatory bacteria are the myxobacteria that excrete digestive enzymes and absord the digested

All this would be even before the kronocyte stage, which I would expect to be a later innovation. As a side note, the likely reason we don’t have modern eukaryotes which can do this, is that over time, the smaller lineages “learned” to retaliate against such “kidnappers”. Indeed, this might well have been caused by the occasion of those capturing cells “learning” to digest their prey (and its genome). If a modern eukaryote tried to capture modern bacteria in the “old” way, it would eventually grab something able to break out of confinement, and eat the larger cell from within.

Also the idea that the first predator would erase all previous life doesn’t make ecological sense. A predator that eats 100% of its prey, turns out to a very hungry predator.

The kronocyte itself shows what’s wrong with that argument – at the earlier stage I was discussing, predation would include absorbing features from the “prey”, so that elements of both would remain “in play” – and consequently, the result wouldn’t necessarily be an obligate predator, any more than modern eukaryotes need to continue capturing bacteria. Some of the resulting descendants would indeed be obligate predators, but others would have collected traits by which they could defend themselves, thus “opting out” of the next round(s) of consolidation. Eventually this would develop into the modern predator/prey balance.

In closing, I’d just like to thank you for your contribution, and your thoughtful discussion of various comments. Yeah, the trolls are a PITA, but that’s the price of open forums.