DOWN (OR DONE) WITH THE NEANDERTHALS ... this time?

[Jacques Cinq-Mars]

This is already old stuff (!), but it certainly deserves the attention of people interested in the nature and timing of the biological component of the MP/UP in Europe.

Here is an abstract from a just released PNAS article, followed  a Discovery Channel piece dealing with the article in question.

The PNAS Abstract :

Published online before print May 12, 2003
Proc. Natl. Acad. Sci. USA, 10.1073/pnas.1130343100

Evidence for a genetic discontinuity between Neandertals and 24,000-year-old anatomically modern Europeans.

David Caramelli , Carles Lalueza-Fox {, Cristiano Vernesi , Martina Lari , Antonella Casoli , Francesco Mallegni , Brunetto Chiarelli , Isabelle Dupanloup , Jaume Bertranpetit , Guido Barbujani , and Giorgio Bertorelle

During the late Pleistocene, early anatomically modern humans coexisted in Europe with the anatomically archaic Neandertals for some thousand years. Under the recent variants of the multiregional model of human evolution, modern and archaic forms were different but related populations within a single evolving species, and both have contributed to the gene pool of current humans. Conversely, the Out-of-Africa model considers the transition between Neandertals and anatomically modern humans as the result of a demographic replacement, and hence it predicts a genetic discontinuity between them. Following the most stringent current standards for validation of ancient DNA sequences, we typed the mtDNA hypervariable region I of two anatomically modern Homo sapiens sapiens individuals of the Cro-Magnon type dated at about 23 and 25 thousand years ago. Here we show that the mtDNAs of these individuals fall well within the range of variation of today's humans, but differ sharply from the available sequences of the chronologically closer Neandertals. This discontinuity is difficult to reconcile with the hypothesis that both Neandertals and early anatomically modern humans contributed to the current European gene pool.

… and the Discovery Channel piece:

Study: Human DNA Neanderthal-Free
By Jennifer Viegas, Discovery News May 12, 2003.

Neanderthals did not contribute to the gene pool of modern humans, according to a recent study that compared the DNA of two ancient Cro-Magnons with that of four Neanderthals.

While Neanderthals and early humans coexisted in Europe for a few thousand years 40,000 years ago, the findings suggest they did not interbreed, an action that would have made Neanderthals a direct ancestor of modern humans.

The study also supports the "Out of Africa" theory. According to this view, modern humans evolved in East Africa and then spread into Europe and Asia through the Middle East.

<snip>

Alan Cooper, professor of zoology at Oxford University, agreed with the findings, but suggests that Neanderthals should not be ruled out just yet as direct human ancestors.

"There is still a remote possibility that only nuclear DNA was contributed, or that any Neanderthal mtDNA lineage has been lost during human population bottlenecks in the last glacial max, but the odds appear pretty slim," said Cooper.

Henry Harpending, professor of anthropology at the University of Utah, also hesitated to rule out the possibility of a human-Neanderthal connection. Harpending further believes that the multiregional mode of expansion theory still is plausible, as "mtDNA seems to have a different history than much of the rest of the genome."

For the full text, CLICK HERE.

For time being, I tend to agree with Cooper’s and Harpending’s cautionary comments.

Jacques Cinq-Mars

[lagarvelho]

Jacques and all:

First of all, I don't think this issue is going to be solved that easily.  As Harpending pointed out, mtDNA may have a different history than other DNA. I also think it's possible that you can take *any* gene or piece of DNA and trace a "unique" history for that particular piece of DNA.  

Besides which, I have a sinking feeling somebody else may well come up with quite different conclusions(depending on which theory they tend to support), based on different sorts of evidence, even genetic evidence.  

It's one of those "stay tuned" situations. . . .
Anne G

[Dale Hoogeveen]

Hi Anne,

I think you are exactly correct.  And likely, as you said, there are going to be a variety of versions before this all gets eventually settled.

Once we are able to integrate other genetic features the description of lineage histories will be no where near as simple as mtDNA-only analysis is currently making it, nor as single threaded.  

From my point of view that should be obvious from the disparate life histories and independant reproductive stategies of the various DNA materials involved in producing functional genotypes.  

I am impressed that the Harpending quote was included to balance the report.

Dutch

Jacques and all:

First of all, I don't think this issue is going to be solved that easily.  As Harpending pointed out, mtDNA may have a different history than other DNA. I also think it's possible that you can take *any* gene or piece of DNA and trace a "unique" history for that particular piece of DNA.  

Besides which, I have a sinking feeling somebody else may well come up with quite different conclusions(depending on which theory they tend to support), based on different sorts of evidence, even genetic evidence.  

It's one of those "stay tuned" situations. . . .
Anne G

[lagarvelho]

Dale:

<<I think you are exactly correct.  And likely, as you said, there are going to be a variety of versions before this all gets eventually settled.

Once we are able to integrate other genetic features the description of lineage histories will be no where near as simple as mtDNA-only analysis is currently making it, nor as single threaded.  

From my point of view that should be obvious from the disparate life histories and independant reproductive stategies of the various DNA materials involved in producing functional genotypes.  

I am impressed that the Harpending quote was included to balance the report.> >

I have the sinking feeling that the researchers thought (whether consciously or not), that "modern" humans were "different" from Neandertals, even before they did the study. It probably has something to do with the way they were trained, but of course I can't be sure.  But a lot of people want to believe these things for a variety of reasons.  Harpending is a strong supporter of the Eswaran study and also the idea that a "genetic bottleneck" occurred at some point in human history.  So he has his opinions too, and stated them.  As I said, I doubt that this is the last we'll be hearing about this issue.
Anne G

[Dale Hoogeveen]

Dale:

I have the sinking feeling that the researchers thought (whether consciously or not), that "modern" humans were "different" from Neandertals, even before they did the study. It probably has something to do with the way they were trained, but of course I can't be sure.  But a lot of people want to believe these things for a variety of reasons.  Harpending is a strong supporter of the Eswaran study and also the idea that a "genetic bottleneck" occurred at some point in human history.  So he has his opinions too, and stated them.  As I said, I doubt that this is the last we'll be hearing about this issue.
Anne G

Hi Anne,

I like the Eswaran approach as well, although not exactly for its conclusions as much as it use of dynamics which implies multiple vectors and even possibly multiple convergence points.  At the very least that makes complete or even universal dominant replacement quite unlikely, in my opinion.  I think that squares far better with the population dynmacs of most other organisms than most forms of replacement theory that I have seen, and I have not by far seen all of them or probably even most.  But then it doesn't seem to me that most replacement advocates have seen the dynamics of as many types of mammal populations as I have either, so that about squares that score, I think.

If I had Harpending across the table from me, I would tell him to look for his "genetic bottleneck" as an inherant part of the life cycle of the reproductive populations of mtDNA in every human generation first.  Second I would suggest he add random adult human female non-participation in reproduction which in humans is relatively large as a secondary and like I said random and so essentially unpredictable bottleneck factor, although there may be some way to estimate the effect over time, if we can determine how vastly increased reproductive success in current human populations messes up that calculation.   IIRC both Sarich and Wolpoff have made comments about that effect in regard to mtDNA and there are likely others.  I think that determining how non-participation can continue to be random over what is essentially the lifetime of a species would also need to be undertaken, since selection definitely tends to eliminate tendancies toward non-participation.  So there would almost certainly be an objection or two to that idea.  

There are other species with low levels of mtDNA variation at least somewhat resembling that in humans,  the wolf, I believe is one such, which also shows substantial adult female reproductive non-participation as part of the pack structure.  In that case mtDNA exclusion as a basis for new species doesn't seem to be getting off the ground.  At least persistant female reproductive non-particpation has some available examples in other mammals and may be showing a relationship between that a lack of mtDNA variation.  That needs more work.

To me the importantance of mitochondria is its function, its lineages being so severely limited and unrepresentative as to be of much less importance.  mtDNA is simple enough being a single fissioning lineage of a concise and generally quite uniform population that is always and only threaded along a single ancestral twig, by never more than one member of any generation throughout an entire ancestry, so that the nature of any and all excluded lineages cannot be sampled.  Pretty small and unrepresentive set of samples upon which to base exclusion of even other mtDNA strains, I think.  Especially since lineage replacement is immediate and complete in a single generation simply by using a foreign female.  

That should about take care of mtDNA lineages, I think.
(If only it were really that simple!)

And then we might be able to deal with chromosomal DNA in its own right and according to its own properties and lineages, eventually patching the multiple lineages together as some sort of convergence or set of convergences.  That definitely brings us back to multiple vector relationships and their interacting convergence points which is about as good a general way to describe the dynamics of a breeding population as any in a single sentence.  And to some sort of expanded "Eswaran-like" dynamic approach. (IIRC he used a limited, but carefully verified genetic set, in what seemed to me to be more of a demonstration of potential than anything else, which is why I find his conclusions less exciting than the process.)

I think your "sinking feeling" may be correct in that data and analysis may well be prelimited in such a way that preheld beliefs are upheld instead of being tested.  I will say that I do not believe that such a thing is intentional or conscious, if it exists, as we both suspect it does.

Peace

Dutch  

[Jacques Cinq-Mars]

Here is a frustratingly short series of comments by Erik Trinkaus on the significanve of the recent Caramelli & al paper which was mentioned in the first post in the TOPIC. It does, however, allude to the fact that, as in most other disciplines that make up the large palaeoanthropological field, molecular biology studies are far from being immune to problems of contamination. In a perverse way, I find this reassuring.

Jacques Cinq-Mars

THE SCIENTIST

Volume 17 | Issue 11 | 29 | Jun. 3, 2003
A Cro-Magnon Capulet?

Long ago, in what is now northwestern Europe, a Neanderthal Romeo and Cro-Magnon Juliet may have met, fallen in love, and had children--or not. Debate rages as to whether human ancestors migrating out of Africa displaced archaic humans like Neanderthals, or mixed with them. A new report lends credence to the displacement camp.

Geneticist Giorgio Bertorelle, University of Ferrara, Italy, purified and sequenced segments of mitochondrial DNA (mtDNA) from two 24,000-year-old Cro-Magnon specimens. He then compared them to previously published Neanderthal mtDNA (29,000-42,000 years old), more recent Cro-Magnon samples, and modern human sequences. He found that the Cro-Magnon samples were more similar to modern humans than to Neanderthals, and he concludes that interbreeding was either nonexistent or extremely rare. (D. Caramelli et al., "Evidence for a genetic discontinuity between Neandertals and 24,000-year-old anatomically modern Europeans," Proc Natl Acad Sci, e-pub ahead of print, doi:10.1073/pnas.1130343100, May 12, 2003).

But, do not dispel other possibilities, says anthropologist Erik Trinkaus of Washington University, St. Louis. Besides a small sample size, he says, contamination by modern DNA can't be ruled out. Moreover, Neanderthal DNA extraction protocols are inherently biased, presupposing that Neanderthal sequences found to be closely related to modern humans must be contaminated.

So did any star-crossed hominids find love? "Sure. Why not?" says Trinkaus. "They were all dirty and smelly. Most of the anatomic characteristics that we use to differentiate them on the basis of their skeletons would be invisible to the individuals concerned." In other words, love is blind.

--Jeffrey M. Perkel

[John Hawks]

But, do not dispel other possibilities, says anthropologist Erik Trinkaus of Washington University, St. Louis. Besides a small sample size, he says, contamination by modern DNA can't be ruled out. Moreover, Neanderthal DNA extraction protocols are inherently biased, presupposing that Neanderthal sequences found to be closely related to modern humans must be contaminated.

That's basically a message that a number of us have been making for awhile; considering that a number of DNA sequences have been recovered from Neandertal remains but not reported because of their similarity to modern sequences. These are assumed to be contamination, but largely because of a protocol that assumes that Neandertals cannot have modern-like sequences. I can't substantiate that they aren't contaminated--in fact, I suspect that they are. But arguments that we can somehow tell "from the sequence" that it is genuine are simply false--the best indicator of that is these 20,000+ year-old Cro-Magnon sequences that are either identical or only a base pair away from the modern human reference sequence.

Would the current Cro-Magnon sequences have been rejected as contamination if they had belonged to a Neandertal? Only double-blind tests would convince me. I would guess that these sequences are genuine, but I have also witnessed really strange cases of contamination that appear to defy all attempts at preventing them.

--John

[lagarvelho]

John and Jacques:

"That's basically a message that a number of us have been making for awhile; considering that a number of DNA sequences have been recovered from Neandertal remains but not reported because of their similarity to modern sequences. These are assumed to be contamination, but largely because of a protocol that assumes that Neandertals cannot have modern-like sequences. I can't substantiate that they aren't contaminated--in fact, I suspect that they are. But arguments that we can somehow tell "from the sequence" that it is genuine are simply false--the best indicator of that is these 20,000+ year-old Cro-Magnon sequences that are either identical or only a base pair away from the modern human reference sequence.

Would the current Cro-Magnon sequences have been rejected as contamination if they had belonged to a Neandertal? Only double-blind tests would convince me. I would guess that these sequences are genuine, but I have also witnessed really strange cases of contamination that appear to defy all attempts at preventing them."

It's hard for me to judge how important the "contamination" issue is, since each "side" in the issue flings the accusation at the other, that the bones whatever side was studying were "contaminated" by some unspecified outside material.  But it seems to me that not many people take into consideration the possibility that you can get the results with "modern" humans that you get, if, say, the Neandertal population was rather small and scattered, relative to the "modern" one.

As for the differences being insignifican to the people involved, as Trinkaus suggests, I have done (admittedly quite unscientific) experiments, that people don't "notice" the supposedly relevant differences between Neandertals and "moderns" either.
Anne G

[John Hawks]

It's hard for me to judge how important the "contamination" issue is, since each "side" in the issue flings the accusation at the other, that the bones whatever side was studying were "contaminated" by some unspecified outside material.  But it seems to me that not many people take into consideration the possibility that you can get the results with "modern" humans that you get, if, say, the Neandertal population was rather small and scattered, relative to the "modern" one.

My feeling is that contamination is mostly a red herring. Finding a modern-like sequence in a Neandertal would not show that they survived. Finding a Neandertal-like sequence in a more recent human would not show that interbreeding was important enough to affect the evolution of recent people. Although many of us would embrace these points, I don't think they would resolve anything.

We need to deal with the fact that known Neandertal mtDNA sequences are an outgroup to the set of known living human sequences. This can't be a result of Neandertal populations being small and scattered: the fact that the global population expanded greatly since the Neandertals should have preserved their sequences in the living population, not eliminated them.

We need a reason to explain why the proportion of Neandertal-like sequences in the global human population changed from some unknown (but nonzero) value 40,000 years ago to zero today. Two alternatives are population replacement and natural selection. The pattern of human mtDNA is consistent with selection. Although this finding does not refute that a replacement may have occurred, it does mean that mtDNA may not be able to test the replacement hypothesis for Neandertals.

[lagarvelho]

Dr. Hawks:

You bring up some good point that I hadn't considered until now(about modern-like and Neandertal-like sequences, and replacement v. natural selection).  Obviously, I'll have to think about this some more.

But I still appreciate your input,
Anne G

[Dale Hoogeveen]

Hi Dr Hawks,

Since mtDNA travels almost completely from mother to offspring, discarding father's contributions, it would travel over generations as complete units.  It seems to me that expecting fragments to remain, as one would in the case of chromosomal recombination, is unrealistic.  

mtDNA lineages only follow one possible ancestral path to any individual expression, and they follow that as a unit.  It is the ultimate in minority representation, a single ancestral source per generation.  Chromosomal lineages can and usually do come from several possible ancestral points, and are inevitably spread as fragments of the total chromosomal set with the expressions that are subject to selection heavily determined by dominance/recessive interaction between the paired chromosomes which is not present in mtDNA selection at all.

In the extreme case, it would only take one selection factor, from whatever source, that differentiated between females for mtDNA lineage replacement.  It is very likely that, whatever the case, male expression followed that female-only selection factor requiring only that it be compatible after the fact rather than contributory to that selection factor or factors.

If we follow the patrilocal pattern that appears to be common in Pan, and we seem to do that at least somewhat, then individual adventerous females would be all that was required for mtDNA replacement, especially in species such as ours where there is little mtDNA variability, provided that there was  any sort of selective advantage being conveyed by the type that was exporting its females.

Additionally that advantage may actually have been more effective when combined with the original local types than it might have been for its originators, hastening the spread of the introduced mtDNA strains, should female introduction have actually been the case in the first place.  (And I think that case was likely, if the patrilocal pattern holds at all for humans.)

I think that there is a lot more work to be done on human female-only selection factors before one can make any kind of definitive statement about inherant linkages between mtDNA lineages and any particular chromosomal lineages or even general chromosomal sets.  Furthermore I think that female only selection as a primary consideration has to provide an initial assumption of independance between mtDNA and chromosomal inheritance, which I think is currently reversed.  There is after all a 50% disassociation between the ongoing complete mtDNA lineage and the inherantly fragmented chromosomal lineage at each reproduction.

Dutch

It's hard for me to judge how important the "contamination" issue is, since each "side" in the issue flings the accusation at the other, that the bones whatever side was studying were "contaminated" by some unspecified outside material.  But it seems to me that not many people take into consideration the possibility that you can get the results with "modern" humans that you get, if, say, the Neandertal population was rather small and scattered, relative to the "modern" one.

My feeling is that contamination is mostly a red herring. Finding a modern-like sequence in a Neandertal would not show that they survived. Finding a Neandertal-like sequence in a more recent human would not show that interbreeding was important enough to affect the evolution of recent people. Although many of us would embrace these points, I don't think they would resolve anything.

We need to deal with the fact that known Neandertal mtDNA sequences are an outgroup to the set of known living human sequences. This can't be a result of Neandertal populations being small and scattered: the fact that the global population expanded greatly since the Neandertals should have preserved their sequences in the living population, not eliminated them.

We need a reason to explain why the proportion of Neandertal-like sequences in the global human population changed from some unknown (but nonzero) value 40,000 years ago to zero today. Two alternatives are population replacement and natural selection. The pattern of human mtDNA is consistent with selection. Although this finding does not refute that a replacement may have occurred, it does mean that mtDNA may not be able to test the replacement hypothesis for Neandertals.

[Dale Hoogeveen]

I think that it is important to view degree of mtDNA variability as an inherantly species-specific consideration that is closely related to female participation in reproductive strategy, likely based more on proportion of completed female fertility incidents than on  replacement of populations.  There is likely a species-specific maximum number of ongoing mtDNA lineages per breeding population with excessive lineages automatically discarded along with the unreproductive females that carry them.  In those species with larger numbers of uncompleted female fertility incidents for whatever reason that maximum number would be smaller and mtDNA variability would be reduced by the mechanics of its reproductive strategy alone.  Concealed ovulation in humans would produce an extreme difference even from Pan where estrus so vividly displays impending ovulation that few potential Pan female reproductive incidents are missed.  That makes Pan an inappropriate  control for human mtDNA calculations, despite close relationships and some otherwise mechanically similar behaviours.  The degree of Pan's inherant female participation in every generation is dramatically and automatically more extensive.  That alone accounts for the large difference in mtDNA lineage variation between current Homo and Pan, badly compromising all calculations where that difference is ignored.

Dutch

If we follow the patrilocal pattern that appears to be common in Pan, and we seem to do that at least somewhat, then individual adventerous females would be all that was required for mtDNA replacement, especially in species such as ours where there is little mtDNA variability, provided that there was  any sort of selective advantage being conveyed by the type that was exporting its females.

[John Hawks]

Since mtDNA travels almost completely from mother to offspring, discarding father's contributions, it would travel over generations as complete units.  It seems to me that expecting fragments to remain, as one would in the case of chromosomal recombination, is unrealistic.  

Quite true that mtDNA is all or nothing, but really the point is whether a selective or neutral explanation is true. Suppose we assume neutrality for mtDNA. For Neandertal mtDNA to survive in today's population, we require an unbroken chain of female descent from some Neandertal woman to some living woman. If mtDNA is neutral, the odds of this happening depend on the size of the human population and whether the population has been growing: large population size over time makes Neandertal mtDNA survival more likely (more women) and growth also makes it more likely (more children per woman).

Let's consider some conditions under which Neandertal mtDNA disappearance might be expected without selection. Approximately 1000 human generations have elapsed since Neandertal morphology disappeared. Suppose that at around the time of Neandertal disappearance, there were 50,000 females globally, and that 500 Neandertal females mated into this population, for an initial frequency of the Neandertal mtDNA type of 1 percent. Further assume that the population never grew in size. Under these assumptions there is about a 35 percent chance that Neandertal mtDNA would not be found in living people.

Notice that these are the kind of parameters that most replacement advocates would be very comfortable assimilating into their models of complete replacement. If the population was much larger, and growing, as we know to have been true, or if we assert that the Neandertals made up a more substantial part of our ancestry, it becomes quickly certain that Neandertal mtDNA should survive today, if continuity were true. In short, the neutral assumptions make the replacement model appear to be a very likely explanation of the mtDNA data, and this is why so many people find the argument so convincing.

In the extreme case, it would only take one selection factor, from whatever source, that differentiated between females for mtDNA lineage replacement.  It is very likely that, whatever the case, male expression followed that female-only selection factor requiring only that it be compatible after the fact rather than contributory to that selection factor or factors.

Exactly right, and since the mtDNA distribution of living humans shows clear signs of past selection, it follows that the molecule should tell us nothing about whether other Neandertal genes may have been replaced. In my view, there is little to be gained in further, more specific descriptions of how easy it would be for different social systems to cause mtDNA replacement; the hypothesis of selection is consistent with the data and more powerful than social factors. I would note that there is no necessity for this selection to be limited to, or even present in, females; selection on anyone will work. It is the strong probability of past selection that persuades me that the data cannot test continuity.

[Dale Hoogeveen]

(heavy snip)
It is the strong probability of past selection that persuades me that the data cannot test continuity.

Dr Hawks,

Very well put.

I would say that restriction to only a single mtDNA ancestor per generation (not even to all female ancestors which is an easy but very incorrect approach which vastly overestimates the preceding presence of surviving mtDNA lineages) makes the necessary continuity automatically very rare in the face of any selection, and statistically unrepresentative of the mtDNA strains of those who made chromosomal contributions, the vast majority of which are not testable at all, since they were completely discarded.  They disappear without a trace regardless of similarity or difference from the surviving strains.

That mtDNA strains of those that contribute chromosomal genetics must survive is an unprovable assumption.  That the vast majority of them go extinct is, however, demonstratable from the material itself, since there is only a 50% survival rate in every reproduction with no recombination to reconstitute or reintroduce discarded lineages.  That approaches zero very quickly over a relatively small number of generations even where there is mother/daughter continuity.  (Of course, it becomes zero automatically anytime that continuity is broken.)

It seems to me that the real story of mtDNA lineages is that of massive extinction with a very minimal survival rather than any kind of representative survival.

Dutch

[Dale Hoogeveen]

Sorry I messed up on the number of quote tags.  Hopefully this is what I meant to send in my previous post:

(heavy snip)
It is the strong probability of past selection that persuades me that the data cannot test continuity.

Dr Hawks,

Very well put.

I would say that restriction to only a single mtDNA ancestor per generation (not even to all female ancestors which is an easy but very incorrect approach which vastly overestimates the preceding presence of surviving mtDNA lineages) makes the necessary continuity automatically very rare in the face of any selection, and statistically unrepresentative of the mtDNA strains of those who made chromosomal contributions, the vast majority of which are not testable at all, since they were completely discarded.  They disappear without a trace regardless of similarity or difference from the surviving strains.

That mtDNA strains of those that contribute chromosomal genetics must survive is an unprovable assumption.  That the vast majority of them go extinct is, however, demonstratable from the material itself, since there is only a 50% survival rate in every reproduction with no recombination to reconstitute or reintroduce discarded lineages.  That approaches zero very quickly over a relatively small number of generations even where there is mother/daughter continuity.  (Of course, it becomes zero automatically anytime that continuity is broken.)

It seems to me that the real story of mtDNA lineages is that of massive extinction with a very minimal survival rather than any kind of representative survival.

Dutch