Look what the devil buried in Bavaria !

[GoodSarmatian]

The skeleton of an unknown dinosaur was recently found in Bavaria. And it's exceptionally well preserved.

From Spiegel Online

German paleontologists have discovered what they believe is the best-preserved dinosaur skeleton ever found in Europe. Some 98 percent of the fossil found in the southern state of Bavaria is intact, and it will soon be placed on display for a short time in Munich.

The discovery of young, unnamed dinosaur fossils is rare, but on Wednesday researchers in southern German state of Bavaria announced they had uncovered an almost perfect specimen. The flesh-eating member of the theropod subgroup, which walked on its hind legs, is among the best preserved specimens of its kind worldwide, said Oliver Rauhut, conservator of Bavaria's state paleontological and geological collections (BSPG) in Munich.

The fossil found in the central Bavarian community of Kelheim is about 98 percent complete, and also includes preserved bits of skin. "The around 135-million-year-old fossil is of outstanding scientific importance," dinosaur expert Rauhut told the German news agency DPA.
A number of similar fossils have been found in China, he said, but they are not as well-preserved. "From far away they often look complete," he said. "But up close one sees that the bone preservation is not that great."

Bits of preserved skin ! I know it's, but i want them to clone it ! Or just analyse it's DNA in case that'S even possible with fossiles.

 

[Cutlass]

That almost looks too good to be true....

 

[emzie]

It's not possible. DNA is unstable and without repair, degrades pretty quickly.

 

[GoodSarmatian]

It's not possible. DNA is unstable and without repair, degrades pretty quickly.

Can't we use frogs to fill in the gaps ?

 

[uppi]

Can't we use frogs to fill in the gaps ?

You would need enough intact DNA for there to be gaps to fill in and after 135 million years there is just nothing left.

Frogs would be a bad source anyway, as they're quite distant to dinosaurs (they're amphibians, not reptiles). In a hypothetical scenario where this would be possible, it would be much better to use DNA of the many species of the theropod subgroup which are living today.

 

[Winner]

In a hypothetical scenario where this would be possible, it would be much better to use DNA of the many species of the theropod subgroup which are living today.

A.K.A. birds.

--------
--------

Great discovery, of course (*remembers his childhood's fascination and fervent "study" of dinosaurs*).

About the possibility of cloning them - we're as of yet pretty unable to clone even relatively recently extinct species - mammoths, other Pleistocene megafauna, Neanderthals, etc. - despite the fact that we actually have some of their DNA. We've found whole mammoths frozen in Siberian permafrost, for heavens sake, and it still isn't enough.

The bits of "preserved skin" are the same as the rest of the fossil - a rock. Whatever biological matter was there is long gone (well, maybe there are some trace amounts left, but for all intents and purposes, it's gone). I have no idea where we could possibly find enough reasonably preserved dinosaur DNA which would enable us to even consider bringing some of their species back from the dead.

 

[emzie]

Well, if you can extract proteins from fossils, you can reconstruct DNA sequences. There's reported finds of proteins in fossilized bone (though you destroy the bone to find it).

 

[GoodSarmatian]

You would need enough intact DNA for there to be gaps to fill in and after 135 million years there is just nothing left.

Frogs would be a bad source anyway, as they're quite distant to dinosaurs (they're amphibians, not reptiles). In a hypothetical scenario where this would be possible, it would be much better to use DNA of the many species of the theropod subgroup which are living today.

I love it when I get a serious answer to a stupid joke question.


Link to video.

 

[Dachs]

Well, if you can extract proteins from fossils, you can reconstruct DNA sequences. There's reported finds of proteins in fossilized bone (though you destroy the bone to find it).

In the Jurassic Park novel, this was described as the "Loy extraction technique", but I didn't know it was actually a Thing (assumed Crichton kinda made it up or expanded on something that wasn't nearly up to spec as he made it seem). Could you say more about the real-world version?

 

[GoodGame]

In the Jurassic Park novel, this was described as the "Loy extraction technique", but I didn't know it was actually a Thing (assumed Crichton kinda made it up or expanded on something that wasn't nearly up to spec as he made it seem). Could you say more about the real-world version?

I've never read the novel, but I understand it was totally fictional (other than DNA can be taken from amber samples).

Out of curiosity, I tried to find the oldest samples from which DNA was extracted.

1.
5000 year old human bones: http://www.sfu.ca/~donyang/adnaweb/Yang DY AJPA1998.pdf
This uses PCR to amplify the DNA sequences.
I believe these are 100% bone tissue that hasn't undergone fossilization.

2. 25 million year old fossil of a termite preserved in amber (pretty relevant to the Jurassic Park story, but not relevant to the idea of a parasite containing prey DNA in it's body and then successfully purifying that):
http://www.sciencemag.org/content/257/5078/1933.short
It uses PCR again. They reported the DNA was very degraded, and in ultra small strands (250 base pairs which is very minute relative to an organism's genome). They looked for strands relevant to highly conserved cell machinery (ribosomes) to try to cross-compare it to modern day species.

3. 19,000 year old bird egg fossils:
http://www.springerlink.com/content/u802l6nu50m23422/#section=1017899&page=1&locus=10


4. Interest in sequencing the neaderthal genome is probably the main goal of some similarity to a Jurassic Park. http://www.sciencemag.org/content/328/5979/710.full
http://www.nature.com/emboj/journal/v28/n17/full/emboj2009222a.html
http://en.wikipedia.org/wiki/Neanderthal_genome_project


Modern genome mapping techniques (high throughput shotgun techniques) that use very small sequences of DNA to construct maps, might make full use of degraded DNA to get an acceptable map to work from?

AFAIK, there's no dinosaur genome project. But a chicken genome was completed, and work is on some lizards. Possibly that info could help with a dinosaur mapping project?

 

[MrCynical]

Well, if you can extract proteins from fossils, you can reconstruct DNA sequences.

Hate to break it to you, but whether a protein survives that kind of time frame tells you nothing about the lifespan of DNA (let alone intact sequences long enough to be useful). They're completely different molecules.

Out of curiosity, I tried to find the oldest samples from which DNA was extracted.

There are a number of examples of significant DNA sequences being recovered from samples that are tens of thousands of years old. That's still orders of magnitude below what's needed to recover dinosaur DNA. Your second example is the only one even close to the required time frame, and what they could recover what seriously limited. They reported most of the degraded DNA they could recover was in 250 base pair or shorter sections. A complete dinosaur genome would probably be on the order of a few billion base pairs. That's quite a jigsaw puzzle, and there's no guarantee you've got all of the pieces.

Reading through the paper also highlights another major problem. When dealing with small samples of DNA it's generally necessary to use a PCR to amplify it enough to be able to sequence. You cannot simply throw an unknown DNA sequence in a PCR tube and amplify it. This method requires primers (short sections of DNA about 25 - 50 bps in length which match sections on either side of the unknown DNA sequence you want to amplify), and without some clue what the sequence is there's no way to design them. In this paper they've found a neat way round this by amplifying the ribosome which is very highly conserved between species. They've chosen sections which show the least variation between species that still exist, and assumed that they were still the same 25 million years ago. Using these as primers, they were able to amplify a few hundred unknown bases between them and sequence them.

Without already knowing the exact sequence of millions of short sections of DNA throughout the genome of the dinosaur, there's no way to recover the whole sequence by this method. It only works for the ribosome because it has changed so little over the millions of years. The genomes of extant species such as chickens is not much help. We've no way of telling what has been conserved from their dinosaur ancestors and so might make a working primer sequence, and even if we did there's unlikely to be sufficient conserved sites between the species to be able to amplify any significant percentage of the ~250bp fragments from the amber.

I think in practice trying to recreate a dinosaur would be less a case of recovering original dinosaur DNA, and more reengineering the DNA of an extant species to produce something that resembles a dinosaur.

 

[Winner]

It seems that creating a time machine to sample 100% correct dinosaur DNA would be simpler

 

[GoodGame]

Reading through the paper also highlights another major problem. When dealing with small samples of DNA it's generally necessary to use a PCR to amplify it enough to be able to sequence. You cannot simply throw an unknown DNA sequence in a PCR tube and amplify it. This method requires primers (short sections of DNA about 25 - 50 bps in length which match sections on either side of the unknown DNA sequence you want to amplify), and without some clue what the sequence is there's no way to design them. In this paper they've found a neat way round this by amplifying the ribosome which is very highly conserved between species. They've chosen sections which show the least variation between species that still exist, and assumed that they were still the same 25 million years ago. Using these as primers, they were able to amplify a few hundred unknown bases between them and sequence them.

Without already knowing the exact sequence of millions of short sections of DNA throughout the genome of the dinosaur, there's no way to recover the whole sequence by this method. It only works for the ribosome because it has changed so little over the millions of years. The genomes of extant species such as chickens is not much help. We've no way of telling what has been conserved from their dinosaur ancestors and so might make a working primer sequence, and even if we did there's unlikely to be sufficient conserved sites between the species to be able to amplify any significant percentage of the ~250bp fragments from the amber.

I think in practice trying to recreate a dinosaur would be less a case of recovering original dinosaur DNA, and more reengineering the DNA of an extant species to produce something that resembles a dinosaur.

Yeah I think that is the consensus on your last point. It also definitely seems that knowledge of descendants' genomes and faith in conservation of sequences is very important to making educated guesses when piecing together a whole genome map. But I imagine there would be many drafts in which even if it were possible to manufacture an embryo and a surrogate mother, there would be many drafts until they got one that "kind of looks like" a raptor. And the irony is 'who ever actually saw a living raptor' to judge?

 

[emzie]

Hate to break it to you, but whether a protein survives that kind of time frame tells you nothing about the lifespan of DNA (let alone intact sequences long enough to be useful). They're completely different molecules.

Um? If you have a protein, you can, with a margin of error, figure out the original DNA sequence that codes it.

edit: I mean, molecular biology was never my strong suit but I'm under the impression CDMB covers what happens in nature, not what we're able to deduce through examination.

 

[peter grimes]

So the popular scientific press leads laymen like me to think that the epigenome may in fact be just as integral to the development of an organism as the genome itself.

If this is the case, then doesn't that mean that Mr Cynical's claim is barely half the story?

trying to recreate a dinosaur would be less a case of recovering original dinosaur DNA, and more reengineering the DNA of an extant species to produce something that resembles a dinosaur.

Simply recreating the sequence of base pairs would ignore methylation, right? For that matter, is methylation even preserved in amber or similar situations?

 

[GoodGame]

Um? If you have a protein, you can, with a margin of error, figure out the original DNA sequence that codes it.

edit: I mean, molecular biology was never my strong suit but I'm under the impression CDMB covers what happens in nature, not what we're able to deduce through examination.

You only get some codons from the protein. You lose any splicing patterns of the genes, and you also lose any genetic regulation information that is non-protein based. The complementary DNA from the protein sequence enough info to be of some use, like comparative evolution studies in a genome browser.

The main utility I imagine from finding a fossilized protein is the evolutionary studies, assuming you have high confidence in your identification of the species and the purity of the protein, and your confidence that the protein comes from the species you have identified. It'd be interesting to know how specific proteins may/may not have changed over evolution from dinosaurs to birds. It'd be also interesting to know how diverged a protein sequence is from from modern (and era) cousins.

 

[MrCynical]

Um? If you have a protein, you can, with a margin of error, figure out the original DNA sequence that codes it.

There are multiple codons for most amino acids, and there's no way to tell from just the protein which ones were present in the original DNA. You could design a DNA sequence that would give the same protein, but the sequence is unlikely to be a close match to the original (only two of the twenty common amino acids have unique codons). You'd scramble any information in alternatively spliced versions containing the DNA sequence, and regulatory proteins probably wouldn't be able to recognise their targets in that region. Also remember that the actual protein encoding DNA is only makes up a fraction of the total DNA, and the rest isn't just junk. There's all kinds of gene promoters and regulatory code which is critical, but isn't expressed as protein.

Simply recreating the sequence of base pairs would ignore methylation, right? For that matter, is methylation even preserved in amber or similar situations?

How important epigenetic factors are to the phenotype of an organism is rather uncertain at the moment, but there'd be no way to recover this information from the protein sequence. DNA methylation might be preserved in amber, but that doesn't necessarily mean we'd be able to recover that information. Similar problem with the PCR as before. The method will amplify the tiny amount of original DNA, but none of the copies will have any methylation regardless of whether or not it was present in the original. You'd lose any information of this type.

My point about recreating dinosaurs is that there is probably not enough recoverable information to assemble a complete original dinosaur genome (never mind any epigenetic information). An attempt to recreate a dinosaur would be more a case of taking the very few fragements we have, and piecing it out with educated guesswork from existing species. The resulting genome (which yes, might require some epigenetic tinkering) would almost certainly be nothing like those of the original dinosaurs, but might give an acceptable facsimile of the animals themselves. This is of course still well in the realms of science fiction for the moment.

 

[MCdread]

There are multiple codons for most amino acids, and there's no way to tell from just the protein which ones were present in the original DNA. You could design a DNA sequence that would give the same protein, but the sequence is unlikely to be a close match to the original (only two of the twenty common amino acids have unique codons). You'd scramble any information in alternatively spliced versions containing the DNA sequence, and regulatory proteins probably wouldn't be able to recognise their targets in that region. Also remember that the actual protein encoding DNA is only makes up a fraction of the total DNA, and the rest isn't just junk. There's all kinds of gene promoters and regulatory code which is critical, but isn't expressed as protein.

You could try to engineer all the gazillion sets of exons that give you that protein, but as you say, that'd be pretty useless beyond trying to put that in a plasmid and express it in a bacterium. Which would still be very cool of course. You could compare it to extant versions and see if you can spot any obvious differences and adaptations.

How important epigenetic factors are to the phenotype of an organism is rather uncertain at the moment, but there'd be no way to recover this information from the protein sequence. DNA methylation might be preserved in amber, but that doesn't necessarily mean we'd be able to recover that information.

Even if we could recover it, we'd only have a screenshot of how the patterns of methylation looked at one point in time in the life of that individual. And in one tissue. And there's more epigenetics going on beyond methylation of course.

 

[peter grimes]

And there's more epigenetics going on beyond methylation of course.

Oh? Please go on. I thought methylation = epigenome.

 

[uppi]

Yesterday, a new paper on reconstructing ancient DNA was published:

http://www.sciencemag.org/content/early/2012/08/31/science.1224344.abstract

We present a DNA library preparation method that has allowed us to reconstruct a high-coverage (30X) genome sequence of a Denisovan, an extinct relative of Neandertals. The quality of this genome allows a direct estimation of Denisovan heterozygosity, indicating that genetic diversity in these archaic hominins was extremely low. It also allows tentative dating of the specimen on the basis of “missing evolution” in its genome, detailed measurements of Denisovan and Neandertal admixture into present-day human populations, and the generation of a near-complete catalog of genetic changes that swept to high frequency in modern humans since their divergence from Denisovans.

They managed to reconstruct nearly the complete genome of a 70000-80000 year old Denisovan. So it might be possible to go for even older species (although its still a long way to non-avian dinosaurs).

Ethical problems aside*, would it be possible with this information to clone a Denisovan?

*Interesting question: If we were able to clone ancients humanoids, at what point would we grant them human rights?