Some Questions About ERV Evidence for Common Descent

[swamidass (S. Joshua Swamidass)]

A question came to me over email, and the questioner agreed to let me post the article here. How would the scientists here answer him?

I recently read a paper that states that many the ERV insertions do not match the accepted pattern of common decent. And that when they do match it’s because of the insertions are not at all random but very specific. So, do matching ERVs really provide evidence for common ancestry? Or do they often match because the insertions are in fact not random, but target very specific locations in the genome? I may be totally missing the point, I’m pretty new to this subject.

The relevant papers and articles are below.

Do Shared ERVs Support Common Ancestry? | Evolution News

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1054887/pdf/pbio.0030110.pdf

“We performed two analyses to determine whether these 12 shared map intervals might indeed be orthologous. First, we examined the distribution of shared sites between species (Table S3). We found that the distribution is inconsistent with the generally accepted phylogeny of catarrhine primates. This is particularly relevant for the human/great ape lineage. For example, only one interval is shared by gorilla and chimpanzee; however, two intervals are shared by gorilla and baboon; while three intervals are apparently shared by macaque and chimpanzee. Our Southern analysis shows that human and orangutan completely lack PTERV1 sequence (see Figure 2A). If these sites were truly orthologous and, thus, ancestral in the human/ape ancestor, it would require that at least six of these sites were deleted in the human lineage. Moreover, the same exact six sites would also have had to have been deleted in the orangutan lineage if the generally accepted phylogeny is correct. Such a series of independent deletion events at the same precise locations in the genome is unlikely (Figure S3).

[…]

Several lines of evidence indicate that chimpanzee and gorilla PTERV1 copies arose from an exogenous source. First, there is virtually no overlap (less than 4%) between the location of insertions among chimpanzee, gorilla, macaque, and baboon, making it unlikely that endogenous copies existed in a common ancestor and then became subsequently deleted in the human lineage and orangutan lineage. Second, the PTERV1 phylogenetic tree is inconsistent with the generally accepted species tree for primates, suggesting a horizontal transmission as opposed to a vertical transmission from a common ape ancestor. An alternative explanation may be that the primate phylogeny is grossly incorrect, as has been proposed by a minority of anthropologists.”

“Endogenous retroviruses may arise within genomes by at least two different mechanisms: retrotransposition from a pre-existing endogenous retrovirus (intraspecific transmis- sion) or infection and integration via an exogenous source virus (horizontal transmission). Many cross-species trans- missions have been documented and frequently manifest themselves as inconsistencies in the presumed phylogeny of closely related species . During the 1970s and 1980s, Benve-niste and colleagues identified, by DNA hybridization and immunological cross-reactivity, several retroviral elements that could be found among more diverse primate/mammalian species but not necessarily among more closely related sister taxa [12,13,14]. Lieber and colleagues, for example, reported the isolation of a particular class of type C retroviruses from a woolly monkey (SSV-SSAV) and gibbon ape (GALV) but not the African great apes [13]. These viruses shared antigenic properties with previously described type C activated endogenous retroviruses of the Asian feral mouse Mus caroli. Cross-species infection from murines to primates was proposed as the likely origin of the retrovirus. A related endogenous retrovirus was subsequently identified in the koala, suggesting a zoonotic transmission from placentals to mammals [15]. Evidence of horizontal transmission for other families of retrovirus has been reported among classes of species as distantly related as avians and mammals.”

[swamidass (S. Joshua Swamidass)]

@ColtCorrea you might want to look at this. @evograd and @davecarlson, I recall we discussed this recently…

[Barry_Desborough1 (Barry Desborough)]

From the email,

“I recently read a paper that states that many the ERV insertions do not match the accepted pattern of common decent. And that when they do match it’s because of the insertions are not at all random but very specific. So, do matching ERVs really provide evidence for common ancestry? Or do they often match because the insertions are in fact not random, but target very specific locations in the genome? I may be totally missing the point, I’m pretty new to this subject.”

We know that ERV integrations that are not locus-specific do not match the accepted pattern of common descent, because they derive from independent endogenizations. This is how we tell that they are not evidence for common descent! Those that are locus specific (commonly located) ARE evidence for common descent. McClatchie glosses over these, which are the vast majority of ERVs in Homo and Pan. Target specificity is discussed in my blog here. Veritas: ERV FAQ: Don't retroviruses target particular locations in the DNA? Doesn't this explain corresponding ERVs?

Your correspondent quotes Yohn et. al.

"Endogenous retroviruses may arise within genomes by at least two different mechanisms: retrotransposition from a pre-existing endogenous retrovirus (intraspecific transmis- sion) or infection and integration via an exogenous source virus (horizontal transmission). Many cross-species trans- missions have been documented and frequently manifest themselves as inconsistencies in the presumed phylogeny of closely related species . During the 1970s and 1980s, Benve-niste and colleagues identified, by DNA hybridization and immunological cross-reactivity, several retroviral elements that could be found among more diverse primate/mammalian species but not necessarily among more closely related sister taxa [12,13,14]. Lieber and colleagues, for example, reported the isolation of a particular class of type C retroviruses from a woolly monkey (SSV-SSAV) and gibbon ape (GALV) but not the African great apes [13]. These viruses shared antigenic properties with previously described type C activated endogenous retroviruses of the Asian feral mouse Mus caroli. Cross-species infection from murines to primates was proposed as the likely origin of the retrovirus. A related endogenous retrovirus was subsequently identified in the koala, suggesting a zoonotic transmission from placentals to mammals [15]. Evidence of horizontal transmission for other families of retrovirus has been reported among classes of species as distantly related as avians and mammals.”

Again, ERVs that are present in different species, but not in corresponding loci are evidence of independent endogenizations, not endogenizations in common ancestors. Independent endogenizations are evidence of spillover events. But They are NOT evidence against the conclusion that correspondingly located ERVs ARE powerful evidence for endogenizations in common ancestors. Click on “ERV FAQ” at the top of my ERV blog pages for the full story.

[Mod edit to mark quotes]

[Barry_Desborough1 (Barry Desborough)]

Do Shared ERVs Support Common Ancestry? from Evolution "News"

[Michael_Okoko]

This article might be helpful

Endogenous retroviruses and the evidence for evolution

[John_Harshman]

Correct me if I’m wrong, but that paper seems to be mostly talking about ERV families, not particular insertions at particular sites. There is apparently nothing to see here except the peril of paying attention to Evolution News. The little bit about specific insertions shows that the insertions are not orthologous in the different species, i.e. they happened fairly close to each other but not at exactly the same sites. This is not incongruence with phylogeny; it’s an attempt at poor character coding by EN&V.

Now, a proper way to determine insertion homology is to sequence through the site to determine if the insertions happen (or don’t) at identical spots, which is also a way to distinguish whether absence of an ERV is a deletion or just lack of insertion.

[Faizal_Ali]

My simple way of looking at it:

If you are reading an book or article, and find entire sentences or paragraphs that are identical verbatim to those in a work that was published earlier, you would be correct in concluding plagiarism.

It would not matter that certain combinations and permutations of words are more likely than others in the English language, nor that many of the sentences in the later work do not appear in the earlier one.

[thoughtful (Valerie (non-scientist))]

@Faizal_Ali

If you are reading an book or article, and find entire sentences or paragraphs that are identical verbatim to those in a work that was published earlier, you would be correct in concluding plagiarism.

I don’t think that’s a good analogy to use. Say you were reading an instruction manual. If you had another instruction manual for a similar but slightly different piece of furniture, sections of the instructions may be identical, where the structure of the furniture is identical.

[RonSewell (RonS)]

@thoughtful

I don’t think that’s a good analogy to use. Say you were reading an instruction manual. If you had another instruction manual for a similar but slightly different piece of furniture, sections of the instructions may be identical, where the structure of the furniture is identical.

But then the identical sections are still “plagiarized”. A basic short sentence such as “There are no user serviceable parts” may be coincidental, but an entire paragraph is unlikely. If there are typo’s or the manual is from, ahem, a non English speaking country, there may irregularities of grammar and diction, and those would be watermarks if copied intact. So while analogies have their limits and only serve to illustrate, @Faizal_Ali’s is reasonable here.

Furthermore, are not analogies from genome to language promoted by YEC with boundless enthusiasm?

[Faizal_Ali]

That is the equivalent of sequences being similar between species because they perform the same function, which is the creationist hypothesis. That does not apply to ERV’s, of course. I also mentioned, in my analogy, that certain combinations of words tend to occur in the English language, without necessarily indicating plagiarism. If someone writes “Be that as it may…”, he is simply using a phrase that in the public domain. But if he writes “It was the best of times, it was the worst of times…” we know where he got that from.

[Mercer (John Mercer)]

@thoughtful

I don’t think that’s a good analogy to use. Say you were reading an instruction manual. If you had another instruction manual for a similar but slightly different piece of furniture, sections of the instructions may be identical, where the structure of the furniture is identical.

The analogy for sequence data, though, is that we have identical structures in different species with nonfunctional differences (“silent” mutations) in their instructions. That’s totally inconsistent with design.

[T_aquaticus]

@swamidass

“We performed two analyses to determine whether these 12 shared map intervals might indeed be orthologous. First, we examined the distribution of shared sites between species (Table S3). We found that the distribution is inconsistent with the generally accepted phylogeny of catarrhine primates. This is particularly relevant for the human/great ape lineage. For example, only one interval is shared by gorilla and chimpanzee; however, two intervals are shared by gorilla and baboon; while three intervals are apparently shared by macaque and chimpanzee. Our Southern analysis shows that human and orangutan completely lack PTERV1 sequence (see Figure 2A). If these sites were truly orthologous and, thus, ancestral in the human/ape ancestor, it would require that at least six of these sites were deleted in the human lineage. Moreover, the same exact six sites would also have had to have been deleted in the orangutan lineage if the generally accepted phylogeny is correct. Such a series of independent deletion events at the same precise locations in the genome is unlikely (Figure S3).

I am familiar with that paper, so I will comment.

The ERV’s in this case were caused by related viruses, and their insertions are called PtERV1 insertions, probably because they were found first in the Pan troglodytes (Pt) genome. As it turns out, PtERV1 insertions are a really good test of common ancestry. Since we find many PtERV1 insertions in the chimp and gorilla genomes and not in the human lineage this would lead us to predict that the retrovirus responsible for the PtERV1 insertions infected those lineages after the split of the chimp and human lineages. Therefore, the vast majority, if not all, of these insertions should be found at different locations in the chimp and gorilla lineages because they are independent insertions. So what did the authors of that paper find?

Based on an analysis of 1,467 large-insert clones, we mapped 299 retroviral insertion sites among the four species (Figure 3; Table S2). A total of 275 of the insertion sites mapped unambiguously to non-orthologous locations (Table 2), indicating that the vast majority of elements were lineage-specific (i.e., they emerged after the divergence of gorilla/chimpanzee and macaque/baboon from their common ancestor).
Lineage-Specific Expansions of Retroviral Insertions within the Genomes of African Great Apes but Not Humans and Orangutans

For those who are unfamiliar with BAC clones, these are plasmids that carry a large chunk of DNA, from 50,000 base pairs to 300,000. 275 of the PtERV1 insertions in chimps and gorillas were not within tens of thousands of base pairs of each other in the respective genomes. Of the 24 remaining PtERV1 insertions they can only say that they are within tens of thousands of base pairs. However, they can’t use their BAC clone method to determine if the PtERV1 insertions are at the same base (i.e. orthologous). This is what the author meant by “within the limits of the BAC-based end sequencing mapping approach”:

Within the limits of this BAC-based end-sequencing mapping approach, 24 sites mapped to similar regions of the human reference genome (approximately 160 kb) and could not be definitively resolved as orthologous or non-orthologous (Table S3). We classified these as “ambiguous” overlap loci (Figure 3). If all 24 locations corresponded to insertions that were orthologous for each pair, this would correspond to a maximum of 12 orthologous loci. The number of non-orthologous loci was calculated as 275/287 (275 + 12) or 95.8%. This is almost certainly a lower-bound estimate owing to the limitation of our BAC-based mapping approach to refine the precise locations of the insertions.

To determine if these insertions are truly orthologous you would need 1 base resolution, such as that found in genome sequencing. Unfortunately, at the time of publication they didn’t have the sequence for the gorilla genome. However, they did have some ambiguous PtERV1 insertions from the chimp and macaque genomes, and this is what they found:

For the three intervals putatively shared between macaque and chimpanzee, we attempted to refine the precise position of the insertions by taking advantage of the available whole-genome shotgun sequences for these two genomes. For each of the three loci, we mapped the precise insertion site in the chimpanzee and then examined the corresponding site in macaque (http://www.ncbi.nlm.nih.gov). In one case, we were unable to refine the map interval owing to the presence of repetitive rich sequences within the interval. In two cases, we were able to refine the map location to single basepair resolution (Figures S4 and S5). Based on this analysis, we determined that the sites were not orthologous between chimpanzee and macaque. It is interesting to note that this level of refined mapping in chimpanzee revealed 4- to 5-bp AT-rich target site duplications in both cases. These findings are consistent with an exogenous retrovirus source since proviral integrations typically target AT-rich DNA ranging from 4 to 6 bp in length [24]. Although the status of the remaining overlapping sites is unknown, these data resolve four additional sites as independent insertion events and suggest that the remainder may similarly be non-orthologous. This apparent independent clustering of retroviral insertions at similar locations may be a consequence of preferential integration bias or the effect of selection pressure against gene regions, limiting the number of effective sites that are tolerated for fixation.

Nowhere in the paper were they able to find an unambiguous example of an orthologous PtERV1 insertion shared by apes/primates that was inconsistent with the standard phylogeny. At the time of the paper’s publication there were still ambiguous PtERV1 insertions, but it would be interesting to revisit this data with more complete data.

TL;DR: None of the insertions were inconsistent with common ancestry which should have been apparent by reading the title of the paper: " Lineage-Specific Expansions of Retroviral Insertions within the Genomes of African Great Apes but Not Humans and Orangutans"

None of those insertions were unambiguously found at the same base in those genomes, and over 95% were confirmed to be at different locations as we would expect from insertions that happened after these lineages split off from one another. Therefore, PtERV1 insertions DO match the pattern expected from common descent.

[T_aquaticus]

Addressing the specific questions posed by the email:

@swamidass

So, do matching ERVs really provide evidence for common ancestry? Or do they often match because the insertions are in fact not random, but target very specific locations in the genome?

As with almost all of the evidence for evolution, it is the pattern of both similarities and differences which points to common ancestry and evolution. That pattern is a nested hierarchy or phylogeny. A lot of what we talk about is couched in terms of phylogeny, and ERV’s are no different.

PtERV1 insertions are actually a perfect test for your question. Some ID/creationists claim that ERV’s match because independent insertions will produce the same insertions at the same base in each of the genomes. Common descent and evolution can make a different prediction. Here is the standard phylogeny for primates:

You will notice that gorillas branched off of the chimp/human lineage first followed by the chimp and human lineages splitting. Therefore, if the PtERV insertions happened before the gorilla lineage split off then they should be found in the genomes of all three species. They aren’t. Instead, PtERV1 insertions are found in the chimp and gorilla genomes. This means those insertions had to occur after all three lineages were on their own. Common ancestry would predict that these insertions happened independently in the chimp and gorilla genome which would also mean that they should NOT be found at the same location in the chimp and gorilla genomes due to the random insertion of retroviruses into the host genome.

ID/creationism prediction: PtERV1 insertions should be found at the same location.
Common ancestry prediction: PtERV1 insertions should NOT be found at the same location.

What do we observe? All or nearly all of the PtERV1 insertions are not found at the same location in the chimp and gorilla genomes, exactly what common ancestry predicts and opposite of what ID/creationism predicts.

Addendum:
If memory serves, there is one possible PtERV1 insertion that scientists have found in the human genome. I have no idea if it is orthologous to any chimp insertions, but I should probably look that up. Nonetheless, it is pretty clear that there are few if any PtERV1 insertions in the human genome compared to the hundreds found in the chimp and gorilla genomes. It is interesting that this virus seems to have infected members of the gorilla and chimp lineages but not the human lineage (or the orangutan lineage for that matter). Whether this is due to host specificity or geography is not known.

[ColtCorrea]

Great discussion everyone. Thanks for your contributions.

[John_Harshman]

It’s important to recognize the difference between ERV insertions and ERV families. ERV families are easily spread (by means of retroviruses that are not incorporated into the host genome) from species to species, and so the presence of similar families in different species is not indicative of spread by common descent in host genomes. But ERV insertions at particular, exact sites in different species are most easily explained by inheritance from insertions into the genomes of the common ancestor of those species.

[Deb0036 (Deborah Ellen Dean)]

Can you describe how and when the split happened? Is there ERV evidence for the half-human, half-chimp that would’ve been the 2nd step? Where did the human (man) come from that breeded with the chimp, originally? How can you show that increased precision and complexity came out of nothing when we don’t see evidence of this happening in any time in which man could record this? I will look through your site, you may address some of these issues.

[John_Harshman]

Lots of questions here. If you removed the chip from your shoulder, you might get better answers, and might ask better questions.

@Deb0036

Can you describe how and when the split happened?

Sorry, what split? I think you may be confused about all this.

@Deb0036

Is there ERV evidence for the half-human, half-chimp that would’ve been the 2nd step?

There is no such thing. It isn’t clear what second step you mean.

@Deb0036

Where did the human (man) come from that breeded with the chimp, originally?

Nobody did any such thing.

@Deb0036

How can you show that increased precision and complexity came out of nothing when we don’t see evidence of this happening in any time in which man could record this?

What increased precision and complexity? This thread is about ERV insertions as evidence for common descent. Common descent does not involve humans mating with chimps. And humans evolved their “increased precision and complexity”, if that’s what it is, long before the invention of writing.

[glipsnort (Steve Schaffner)]

@T_aquaticus

With large BAC clones they were able to determine that only 24 insertions out of 299 were at different places in the genome.

I think this sentence didn’t come out the way you intended.

[T_aquaticus]

@Deb0036

Can you describe how and when the split happened?

The split between the chimp and human lineages occurred around 5 million years ago as suggested by fossil evidence and the number of mutations that separate the chimp and human lineages. The split was due to speciation where one population becomes two populations which causes different mutations to accumulate in each of the separate populations. Speciation is a pretty big topic, but you can get a start here:

Speciation

@Deb0036

Is there ERV evidence for the half-human, half-chimp that would’ve been the 2nd step?

The ERV evidence demonstrates that we share a common ancestor with chimps which means there had to be generations between us and that common ancestor. That ancestor would not have been a chimp or a human. Chimps are our cousins with whom we share a common ancestor.

@Deb0036

How can you show that increased precision and complexity came out of nothing when we don’t see evidence of this happening in any time in which man could record this?

With evolution, precision and complexity does not come out of nothing. They come from modifications of earlier generations.

It would be helpful to focus on questions about ERV’s in this thread, but feel free to start a new thread on a specific topic if you still have questions.

[Faizal_Ali]

@T_aquaticus

The split between the chimp and human lineages occurred around 5 million years ago as suggested by fossil evidence and the number of mutations that separate the chimp and human lineages.

I’ll just emphasize this point, even though it is not exactly on topic for this thread, because I think it is often not sufficiently appreciated.

We can come up with a pretty accurate estimate of when the most recent common ancestor of humans and chimps existed from the fossil record. And can also estimate this from the number of genomic differences that differentiate us from chimps (which under common descent would have resulted from mutations) and the observed mutation rate.

The estimates by these two very different methods come up with the same answer.

That would be quite an amazing coincidence if common ancestry was not true.