Originally Posted by samclem
My question is, how are proteins able to survive in usable form for 68 million years inside a dinosaur bone, at least enough to enable scientists to do DNA tests, confirming a link between birds/dinosaurs?
I never took biology, so this kind of thing is always interesting to me.
The bones were found in a site called Hell Creek. I can't speak for the preservation conditions necessary to keep soft tissue around, but apparently this is currently being actively researched
On the other hand, I can talk about the methods they used to analyze the sample. The current mass spectrometric methods used for proteomics are very, very sensitive. A modern instrument has attomole sensitivity - that's 105
molecules (a mole is 1023
molecules). The particular protein that they detected in the T. rex sample was collagen, the most abundant protein in the body. Also, they were looking at protein fragments, not whole proteins. That technique is called 'bottom-up proteomics', and is far more sensitive than trying to detect whole proteins ('top-down proteomics'), though it gives less information about what was in the sample.
From the materials and methods section of their original paper (published last year, by the way - the recent paper was just a phylogenetic analysis), they took ~2.5g of bone from the fossils (though they mentioned that the T. rex bones required significantly more extract - but they never said exactly how much). They took this stuff, purified it through a number of methods, digested it with an enzyme that cut it at predictable locations (trypsin) then ran it through a separation platform (liquid chromatography). The separations were done in a microcapillary, so that should tell you how tiny the volumes we're talking about here are. They then electrosprayed the resulting solution into an ion trap mass spectrometer (crappy mass accuracy, extremely high sensitivity and speed) and measured their mass-to-charge ratios. They then fragmented the detected ions by collisions with helium gas (collision-induced dissociation) and again measured their mass-to-charge ratios.
By inferring the masses of the intact peptides and the masses of various fragments, it's possible to reconstruct the sequence of the intact peptides. Then, by matching these peptides to databases of known protein sequences, it's possible to reach an identification of exactly what they are. The sequences matched most closely to bird collagen.