Forum:DNA in amber

Forums: Index > Fossil Fuels > DNA in amber

Let is have a discussion about the topic: DNA in Amber

Lebanorhinus specimen

I am now calling the Cretaceous weevil sequence into question. You said that Martin and Gutierrez may be wrong because the sequence only shows 98% similarity to modern weevils. However, I read their paper, and they said this.

From the GenBank database we retrieved 30 homologous sequences belonging 
to coleoptera and otherrelated insect groups (table 1) and aligned them
using CLUSTAL W (Thompson, Higgins, and Gibson 1994).Maximum-likelihood 
pairwise distances (transition/transversion ratio52) were calculated with the 
programDNADIST (Felsenstein 1993). The outgroup was anodonate,Aeschna 
cyanea(X89481). Dipterans sequenc-es were discarded because of their high rate 
of nucleotide substitution (Friedrich and Tautz 1997). The dis-tance between 
the outgroup andLebanorhinuswas com-pared to the distances between the outgroup 
and otherinsects with a Wilcoxon nonparametric test. The difference was 
nonsignificant (p50.389), suggesting that theLebanorhinussequence is not ancient.
The distance be-tween the outgroup and the fossil weevil was 0.130 andthe mean 
distance between the former and the other in-sects was 0.127.One might think that
such a test would not discrim-inate a 120–135 MYA old sequence from an extant 
one; to counter that, we have estimated how much divergencewould have been lost
by a 18S rRNA that stoppedevolving 120–135 MYA. First, as an estimation of the
18S rRNA substitution rate, we have used the estimatedmetazoan substitution rate
given by Wray, Levinton, andShapiro (1996). For a period of 120 MYA, the 
expected divergence is 0.018. The mean distance between the out-group and extant
sequences is 0.127; then, the distancebetween the outgroup and an 120-Myr-old 
sequence should be 0.109, the test now being significant (p,0.0001).
We think that theLebanorhinussequence may bean extant beetle contamination. 
For instance, Waldenand Robertson (1997) recovered a beetle sequence when they
were amplifying DNA from an amber-entombed bee specimen, but no coleoptera was
studied before intheir laboratory.

I don't understand much of that, but I think what they're saying is that the Lebanorhinus specimen may have been contaminated by a modern beetle because the 18s rRNA gene doesn't have the expected amount of divergence for a DNA sequence that old. Did you consider this? And I did the BLAST again. The highest amount of similarity with extant beetles was 99%, not 98. I'm starting to think that it was indeed contaminated by an extant weevil, probably Hypera. I also heard that 18s rRNA is often used alone to make a tree because of how conserved it is. I made a tree again, and Hypera was indeed the closest relative. I feel deceived. User:Jurassic Park Treasury

I knew this for a long time, I've said what I think about it at the forum. Maybe we better discuss it here. There are two explanations for this:

• There was no endogenous DNA in the weevil sample. Of all the insects, fungi and microbes it could have been contaminated with, it was contaminated with weevil DNA.
• The Weevil DNA is from the Weevil sample.

I dare to say that unless you are biased against ancient DNA, explanation 2 in more likely. The sequence is very similar to modern DNA, but not identical. This DNA fragment was quite short, I dare to doubt how much we can extrapolate from it. The other fragment, probably from a symbiotic yeast, only had a 90% similarity with extant species. How do you explain that with modern contamination? BastionMonk (talk) 22:04, July 15, 2013 (UTC)

Correction: The yeast is 93% identical. However, you do have a point there. That sequence could be ancient.

But it isn't too far fetched that the prehistoric weevil could be contaminated be a modern one. Although you have pretty much zero respect for the critical studies of DNA in amber, you have to acknowledge that some contaminated sequences were actually from beetles. They could not have been from the amber insects because the insects being studied by the critics were bees and flies.

You asked me why this contaminated beetle wasn't sequenced by now. You have got to remember that DNA sequencing is not magic. We have not automatically sequenced every bug in existence. One of the Hypera sequences showing up in the BLAST list wasn't added until March 2013!

I am actually pretty divided on what to think. I want to believe in the ancient weevil DNA because it would be awesome. At the same time, however, my skeptic side is saying nay. I don't want to be wrong. However, even if it IS contaminated, it's not a complete loss, since the fungal sequence does seem ancient. Jurassic Park Treasury (talk) 01:37, July 16, 2013 (UTC)

Update: It seems that the only reason why the fungal sequence wasn't 100% similar to modern sequences is because the sequence has gaps. Otherwise, it only differs by 1 base pair. So the fungal sequence may be modern after all. Jurassic Park Treasury (talk) 04:22, July 16, 2013 (UTC)

Although you have pretty much zero respect for the critical 
studies of DNA in amber

I am actually pretty divided on what to think.

Don't get me wrong. I have respect for critical studies and stuff. I also don't know what to think. I KNOW something, or I DON'T know something. Is Cano's amber DNA endogenous? I don't know. I just dare to say that the case isn't closed. I am fed up the the global skepticism on paleo-DNA. Everywhere on the internet it says that DNA can't last longer than 100,000 years. However, DNA is extracted from 8 million years old microbes in Antarctic ice and a 700K years old horse's genome is fully sequenced. This shows me that the skepticism of the masses is unjustified.

If someone claims that it is proven that the Cano sequence is contamination, I raise doubts because the case isn't closed. If a group would claim the sequences are real, I would also raise doubts because I know there are indications it is not. I don't take sides, I only fight against those that take sides. You should do the same.

To be sure, we need a lot of more sequences from amber entombed samples. And I can't find reports about that. That could mean there is no DNA in ancient Amber, but I also can't find reports of DNA much younger amber or opal. So, I think there isn't much extraction done in the first place.

the only reason why the fungal sequence wasn't 100% similar to 
modern sequences is because the sequence has gaps.

Nice find! Gaps are even more unlikely to happen than substitutions. So, it CAN still indicate antiquity. BastionMonk (talk) 13:29, July 16, 2013 (UTC)

I was also wrong again. I looked at the fungus alignment again, and it did differ by more than one base pair.

From what I've heard, DeSalle also extracted DNA from amber wood gnats, but I can't find the sequences on GenBank or anywhere else, so I can't BLAST them.

"To be sure, we need a lot of more sequences from amber entombed samples."

Indeed. However, I doubt scientists will try anymore, thanks to Allentoft's DNA half-life study. Seriously, how hard is it to obtain some amber and attempt to extract DNA from it? Even if they fail, I doubt it would be much of a loss.

Jurassic Park Treasury (talk) 20:16, July 16, 2013 (UTC)

Here is the alignment.

I did a BLAST search with the 18S ribosomal RNA gene of an ostrich. It only had a 1% difference to the chicken version of a gene, and they belong to different clades! The Lebanorhinus and Hypera differ by three base pairs, the chicken and ostrich differ by two. So the sequence similarity doesn't prove contamination after all. Jurassic Park Treasury (talk) 08:57, July 19, 2013 (UTC)

I have a new criticism of the sequence: I BLASTed a supposed ancient termite sequence from amber, and as I said in the other forum, it did not match with any modern termite, only fruit flies and relatives. However, it only had a 92% similarity to modern fly DNA. Does that make the sequence ancient? Jurassic Park Treasury (talk) 08:40, September 13, 2013 (UTC)

None of them passed.

OK, here's a table from a 2006 paper showing whether paleo-DNA samples passed the authenticity test. None of them did. I would like your thoughts on this, as it seems interesting that none of the amber sequences were recovered from a clean lab, and that they don't seem to fit relative rates tests as noted by Martin on the weevil sequence. You still haven't answered that, you only repeated that it was unlikely for a weevil specimen to be contaminated by another weevil.

The paper in question.

Jurassic Park Treasury (talk) 05:19, March 7, 2014 (UTC)

The relative rates test does nothing more that calculating the evolutionary distance between the amber-sequence and a close relative. They try to recover the age from the mutation rates in the species. Tests like this rely on a huge amount of assumption from our part. These assumption have HUGE influence on the outcome. Let me give you an example:

Researchers tried to calculate when humans started to wear clothes. They thought they could do this by calculated at what time the common ancestor of the body louse lived. One team estimated that the evolutionary distance between all extant louses was around 107,000 years ago. However, a second group of researchers using similar genetic methods estimate that clothing originated around 540,000 years ago. (see Clothing#Origin_and_history)

In that 2006 paper they propose a brand new method for the relative test. Until that test has been used to predict the age of sequence for which the age is already known, we don't know its validity for sure. I've briefly scanned the paper, and I can't see them doing anything like that.

Furthermore, I think that the insect fragment is FAR to small for a good analysis. Even is this tiny fragment only have 3 basepair differences with Hypera, we don't know if that also was the case for the rest of it's billion basepair genome. BastionMonk (talk) 15:10, March 7, 2014 (UTC)

Sequence

Here I comprared the 18S ribosomal RNA gene sequence from the Cretaceous Lebanorhinus with extant Hypera zoilus (aka Hypera punctata):

Comparison.

I count 5 nucleotide differences. Could someone make the same comparison for multiple extant Weevil species whom the common ancestor lived in the Dinosaur age? I would like to see how many nucleotide differences they have. If that is far more than 5/300, the age of the Lebanorhinus-sequence is indeed very questionable. BastionMonk (talk) 15:10, March 7, 2014 (UTC)

Here is a sequence from a nemonychid, and a curculionid. Their common ancestor would have lived in the dinosaur age. Jurassic Park Treasury (talk) 19:56, March 7, 2014 (UTC)

How do you know their common ancestor lived in the Mesozoic? Anyway, in our 350 bp region there are only 3 differences between the species. However, one of them is an insertion, which is much rarer. BastionMonk (talk) 14:13, March 11, 2014 (UTC)

Lebanorhinus lived in the Mesozoic, and it was a nemonychid. Therefore, by the time it lived, the two families had already diverged. And between what two species? The ones I linked? If so, is the insertion a good thing or a bad thing? Jurassic Park Treasury (talk) 20:24, March 11, 2014 (UTC)

Yeah, I compared the two species you linked. Could maybe you try to compare more species? I would use the Clustal tool if I were you. I don't know what to think of the insertion, only that it is much rarer than a mutation. How rare differs per species (an possibly era). BastionMonk (talk) 15:38, March 24, 2014 (UTC)

It's difficult to find good 18s gene sequences from nemonychids on GenBank. Most of the sequences have long NNNNNNN gaps, making a proper comparison impossible. Here are two more nemonychid sequences, which still contain some Ns here and there: 1, 2. And a few curculionids: 1, 2 and 3. I'll compare them later. Jurassic Park Treasury (talk) 01:41, March 25, 2014 (UTC)

Yeast

I compared the Cano ITS sequence against GenBank. It has similarity with the Saccharomycetales species Meyerozyma and Pichia. However, it is very different from the extant sequences, there are huge gaps. I should investigate this more. BastionMonk (talk) 09:02, March 31, 2014 (UTC)

Reproducibility

One of the main points of Amber-DNA critics is that other teams couldn't reproduce Cano's results. I've not been able to find on what amber specimens they performed the extractions, but maybe this could be an explanation.

I stumbled upon this video of a amber-DNA critic:

DNA from Amber

Professor Timothy Rowe:

"We have actually CAT scanned some insects in amber, and they're very misleading. When you look at ... this bug that is trapped inside the amber, it looks like a beautiful specimen. But the only parts of it that are preserved are the actual exoskeleton, its outer surface... When that animal was alive it had a bacterial colony inside, (when it died) it just digested it all out. So it turns out that when we section these things, they are hollow inside. There is nothing left, except the outer skin."

Now, we know that this isn't true in all cases. Poinar found Sandflies with the guts intact, complete with reptile blood. However, according to this professor the insects are hollow in most cases. When the insect dies, internal microflora digests the soft inside of the insect. After that they eat each other until nothing is left.

So, could it be possible that all those people that tried to reproduce Cano's results were actally trying to extract DNA from empty exoskeletons? If true, it is no surprise they found nothing. Maybe Cano and Poinar were lucky, because their Weevil had still some tissue inside it. BastionMonk (talk) 21:30, July 31, 2013 (UTC)

It might also mean that DNA in amber might be rare, since intact insects are rare also.

I have a new idea for a study: Try extracting DNA from amber insects that are NOT completely hollow. This way, we might finally get an answer on whether DNA is preserved. It might be wise to try extracting DNA or proteins from the blood cells in the Cretaceous sandfly gut also. Jurassic Park Treasury (talk) 22:53, July 31, 2013 (UTC)

DNA from copal-entombed bees

Penney et al. (2013) tried to extract DNA from copal, a younger version of amber. In this case one amber sample was ~10,000 years old, the other <60. They failed to extract (almost) any DNA from the samples. In popular media this is seen as conclusive evidence that DNA can't be preserved in amber.

An example is the article Jurassic Park sequel announced, but scientists say dinosaurs not returning:

The research, published today in the journal The Public Library of Science ONE,
effectively debunks the theory popularised 20 years ago in the movie Jurassic
Park that dinosaurs could be reborn by extracting DNA from remains that had 
been preserved in amber.

I've just read the article, and I dare to say the case isn't closed.

Penney et al. only performed the extraction on two stingless bees. A test set with size 2 is far to small to conclusively prove anything.

Furthermore, they acknowledge the point I made above: the preservation of internal tissue in resin varies from sample to sample.

Certainly, the preservation and fossilization processes are not uniform for all  
copal and amber deposits, nor for all inclusions from a single deposit. For 
example, the recent application of X-ray computed tomography (CT) has shown that
in some instances internal organs are preserved in amber specimens 50 million 
years in age [35], whereas digital dissection of a spider preserved in amber of a
similar age revealed nothing substantial preserved internally [36].

I can not find anywhere in the article that they checked, before or after the extraction, if internal tissue was present in the bees. This sentence in their conclusion seems to prove my suspicions:

Unfortunately, it is rarely possible to draw any conclusions about the degree of
internal preservation using traditional light microscopy. Clearly, the better
preserved a copal specimen is internally, then the more likely it is that DNA
will survive.

future studies may benefit from identifying specimens with preserved internal 
morphology prior to attempting aDNA extraction, assuming the X-ray energy
levels required to scan the copal are maintained below a threshold
at which DNA is damaged.

I hope that future studies will first make CT scans of each specimen and select the insects with internal tissue. I also hope these studies will have large test sets. BastionMonk (talk) 13:25, September 12, 2013 (UTC)

I just hope that the negative results of poorly-done studies like this won't discourage further research into this subject. Jurassic Park Treasury (talk) 13:34, September 12, 2013 (UTC)

Me too. Although, I should note that this isn't a poorly-done study. They write that the same methods resulted in succesfull DNA extractions from dried museum samples. It is just that preservation in amber is more complex than we thought. The main point of this (and their previous) study was to compare destructive and non-destructive extraction methods. They showed that destructive extraction methods yield more and longer DNA fragments. BastionMonk (talk) 13:58, September 12, 2013 (UTC)

A critic on the fossil forum said this:

"The case on DNA from dinosaurs from amber is closed, the only debate is whether or not it was ever really open. Even when first reported there was wide spread skepticism among researchers, rightly justified as attempts to reproduce the initial study have always confirmed contamination.

However, no study is perfect and at the risk of providing traction for doubters I will say this. The DNA isolation used in this study was not optimal and I'm surprised that these authors chose that methodology for ancient DNA (aDNA). The QiaQuick DNA isolation kit used in this study relies on silica-membrane columns to bind DNA in the presence of EtOH during wash steps then eluting said DNA using molecular biology grade water or a suitable buffer. Such silica-membrane kits are extremely easy to use and, in combination with silica micro beads, have quickly become the industry standard. They do however have limitations that are not well publicized.

Foremost, with regards to this study, there is a size range for binding DNA to silica-membranes, that range is between 100bp -10kb. While 100 base pairs seems very small, in actuality it isn't. There is no reason to think that aDNA present in copal insects would not have fragmented beyond 100bp. This doesn't mean that DNA is unrecoverable, fragmented DNA can be sequenced and reconstructed reportedly down to 40bp -50bp. It just means that the kit these researchers used likely did not recover those smaller fragments.

In addition, silica-membrane technology has another limitation. Silica-membranes never release all of the DNA bound to them. Here is a chart from Qiagen's Qiaquick Handbook showing DNA recovery with different elution volumes. Notice that even with the maximum recommended elution volume recovery never exceeds ~90%. Sadly, the authors of this study followed a troubling trend in current molecular work by stating "DNA was then isolated from both the non-destructive and destructive extracts using a QiaQuick PCR purification kit (Qiagen)" as if simply stating that using a certain kit was sufficient to document their methodology. In doing so these authors did not publish their elution volume for this step so we cannot estimate DNA recovery. This is something that should have been addressed in review.

A better way to recover the DNA extracted from these specimens would have been to use a single tube phenol-chloroform isoamyl alcohol extraction method. The techniques involved are not as trivial as a kit extraction method, and they do require more training and caustic chemicals like chloroform; however, they would have recovered virtually all of the extracted DNA and what little was lost would not have been biased by fragment size." Jurassic Park Treasury (talk) 01:51, September 13, 2013 (UTC)

About the size issue

there is a size range for binding DNA to silica-membranes, that range is between
100bp -10kb. There is no reason to think that aDNA present in copal insects would
not have fragmented beyond 100bp. This doesn't mean that DNA is unrecoverable, 
fragmented DNA can be sequenced and reconstructed reportedly down to 40bp -50bp.
It just means that the kit these researchers used likely did not recover those
smaller fragments.

Did this person checked the results in the article? Most of the captured fragments had a length lower than 100 bp.

Age (years)	Method	No fragments	Length (bp)
<60	Non-destructive	30	25-100 (av. 49)
	Destructive	460	24–536 (av. 100)
+10,000	Non-destructive	54	34–44 (av. 39)
	Destructive	1335	33–47 (av. 40.36)

The other issues should be discussed in different sections of the forum. BastionMonk (talk) 08:47, September 13, 2013 (UTC)

Extraction tool

In The Real Jurassic Park, scientists are shown using a needle-like object to extract tissue from an insect in amber. What is the tool called? It's probably obvious, but I don't know. Jurassic Park Treasury (talk) 23:48, August 6, 2013 (UTC)

What is the circled tool called?

Coprolites from cockroaches

In 2013, a Cretaceous cockroach with coprolites inside was found in Lebanese amber. Jurassic Park Treasury (talk) 22:59, January 4, 2014 (UTC)

Bee DNA

Here is a paper that failed to reproduce Poinar's results of getting DNA from fossil bees in amber: [1]. Jurassic Park Treasury (talk) 07:29, March 1, 2014 (UTC)

Plant DNA

This is a sequence from an extinct legume leaf preserved in Dominican amber. When a BLAST search is done, it turns out that modern plants only have a 96% similarity. I don't know whether plant DNA is more resistant to decay than animal DNA, though. Jurassic Park Treasury (talk) 08:15, March 1, 2014 (UTC)

This sequence was recovered from a 40 MYA Pinus fragment from Pinus seeds entombed in Domenican Amber.

In the 2014 paper The never-ending story of geologically ancient DNA: was the model plant Arabidopsis the source of Miocene Dominican amber?, the authors debunked an alleged Miocene plant sequence, finding it was contamination from a common model plant. In the paper, they also find that the sequence differences from modern plants might have been the results of sequencing artifacts or manipulation. Is it possible that this is also the case with the weevil sequence? Jurassic Park Treasury (talk) 03:31, March 30, 2014 (UTC)

A final study

If we want to know if DNA is preserved in amber or not, I suggest that scientists should do a study like this. The study would use a large sample size containing both copal and amber from various locations and ages.

The samples should be organized somewhat like this:

• 20-50 pieces of Madagascan copal (>10,000 years).
• 25-50 pieces of Dominican amber (25 to 40 million years).
• 10-20 pieces of Canadian amber (75 million years).
• 15-25 pieces of New Jersey amber (90-100 million years).

The lab and equipment would need to be as sterile as possible to avoid contamination. Before attempting to extract DNA, the scientists would need to check the insect specimens for internal remains. And instead of using a QiaQuick kit to recover the DNA, the scientists would use phenol/chloroform extraction to recover it, which according to the critic, would extract virtually all of the DNA.

If this study was performed, and no endogenous DNA was found in the amber samples, it would be safe to say that amber does not preserve DNA.

Neither of us has the money or resources for such a study however. Even though a decent DIYBio lab can be built for a few hundred to a few thousand dollars, as seen here, such a lab would not be suitable for any ancient DNA extraction and would be very vulnerable to contamination due to a lack of sterility. And amber is very expensive to buy, so I don't think any scientists would want to spend so much money on the equipment in fear of failure. Jurassic Park Treasury (talk) 05:09, March 12, 2014 (UTC)