Scientists in Leipzig, Germany, and St. Louis are reporting that they have been able to recover the bone protein osteocalcin from 75,000 year-old Neanderthal bones and even to obtain a protein sequence. They were then able to compare the protein sequence from Neanderthals to that of other primates including humans and gorillas. Interestingly, the ninth amino acid of the protein, identical in humans and Neanderthals, was different in gorillas, a difference the authors attribute to differences in diet among the species. We omnivores don't get as much vitamin C as gorillas do.
Some of these scientists have been involved in ongoing efforts to extract DNA from ancient material. The work on sequencing proteins, which has been made possible by big advances in mass spectrometry , a technique in which proteins are shattered and the masses and sequences of the fragments are measured. For the purposes of comparing species (especially very closely related species), DNA analysis would be preferred, because DNA sequences can have differences which are not apparent in the protein sequences. The specific modification reported here, hydroxylation of the proline, would of course have been missed in DNA work.
Orbis-quintus has a nice discussion, including links describing the site in Iraq where the bones were found (a Neanderthal burial ground?), and Zinken , who must have the world's fastest RSS reader, has a second post about the potential use of this method for retrieving REALLY old protein.