Science via AP

These undated photos provided by the journal Science show demineralized fragments of tissues lining the marrow cavity of a Tyrannosaurus rex femur. The find included what appear to be blood vessels, and possibly even cells.

Scientists have discovered how bits of the protein collagen from dinosaurs that died millions of years ago might have survived in fossilized bones.

The new research shows the rope-like connective tissue is twisted in such a way that certain parts are shielded from biological degradation and could survive for millions of years. 

James San Antonio, a biochemist at Orthovita, a medical-implant manufacturer, and colleagues looked at bits of collagen extracted from dinosaur bones and compared them to where those bits show up in collagen fibers of rats and humans.

They found that all 11 pieces analyzed came from the innermost parts of the fibers where they would have been protected from degradation by enzymes and the elements.

What's more, several of the molecules contain amino acids that are water-hating and contain few acidic amino acids, which make them less vulnerable to degradation by enzymes and water.

"We were rather pleasantly surprised and taken aback," Orgel told ScienceNews, adding that if any molecules were to survive millions of years, it would be the sheltered kinds they observed. 

If so, the finding helps explain how researchers were able to recover and sequence soft tissue from a Tyrannosaurus rex and a duck-billed Hadrosaur.

Those studies are controversial since many scientists believe there's no way soft tissue could survive millions of years and thus what has been found and sequenced is actually contamination from more recent times.

The new finding, reported in PLoS One, has yet to convince the skeptics.

"I'm an old protein cynic," Matthew Collins of the University of York in England, told ScienceNews. "Obviously, we would like to see a lot of samples from lots of labs. I'm not convinced yet."

The study was based on data from the initial reports. And that data, geneticist Stephen Salzberg at the University of Maryland has pointed out in a comment at PLoS One "are contaminants from modern species," notes Nature News.

The new study's authors refute the criticism, saying that if the fibers were contaminants, they should be from random parts of the structure. All the bits they found were from the most protected regions.

So, the controversy remains, but the new study does show how these proteins could survive millions of years. And if they did, scientists believe they'll be able to gain insight on how dinosaurs lived.

More on dinosaur proteins:

• T. rex analysis supports dino-bird link 
• Proteins could reveal new dinosaur secrets 
• Gunk in T. rex fossil confirms dino-bird lineage 
• Scientists recover T. rex soft tissue 
• Dinosaur mummy reveals organic molecules 

Nicholas Longrich / Yale University

An illustration depicts a newly discovered dinosaur called Titanoceratops, which predates the better-known Triceratops.

Paleontologists report that a massive horned dinosaur was roaming the American Southwest 5 million years before the well-known Triceratops ... or was that a Torosaurus?

The newly named species, Titanoceratops, weighed nearly 15,000 pounds and had an 8-foot-long skull. It lived during the Cretaceous period, around 74 million years ago.

The finding, accepted for publication in the journal Cretaceous Research, suggests that the large horned dinosaurs evolved their large size earlier than previously thought, reports Yale University paleontologist Nicholas Longrich.

The paleontologist got an inkling about the existence of this dinosaur while searching through scientific papers that described a partial skeleton discovered New Mexico in 1941. The skeleton was identified as Pentaceratops, a common species to the area, and was reconstructed as one for display at the Oklahoma Museum of Natural History in 1995.

"When I looked at the skeleton more closely, I realized it was just too different from the other known Pentaceratops to be a member of the species," Longrich said in a news release.

Nicholas Longrich / Yale University

The skull on the left is the Titanoceratops skull, the missing parts of which were reconstructed to look like a Pentaceratops. The illustration on the right shows the missing parts of the frill (shaded).

Instead, he says the dinosaur likely weighed twice as much as an adult Pentaceratops. It was similar to Triceratops, but with a thinner frill, longer nose and slightly bigger horns.

He suspects that Titanoceratops is the ancestor of both Triceratops and Torosaurus. "This skeleton is exactly what you would expect their ancestor to look like," he said.

More work is needed, however, before the assignment to a new species is confirmed. As pointed out by Brian Switek on the Smithsonian's Dinosaur Tracking blog, members of the Dinosaur Mailing List are debating whether Pentaceratops and Titanoceratops are different growth stages of a single species.

"The animal Longrich has named Titanoceratops certainly did exist," Switek writes, "but as with any other species, the animal's name is a scientific hypothesis that will likely be discussed and debated in years to come."

The discussion parallels the debate over whether Triceratops and Torosaurus fossils actually represent the juvenile and adult forms of the same animal. Some experts suggest that the name "Triceratops" (or more likely "Torosaurus") should go extinct, alongside "Brontosaurus." Others, however, insist that Triceratops and Torosaurus were truly different breeds of horned dinosaurs.

So how do you settle the debate? The best way is to find more fossils, and especially fossilized frills. That's exactly what Longrich is hoping for in Titanoceratops' case.

More stories on horned dinosaurs:

• Puzzling three-horned dino is adult Triceratops
• Mini Triceratops first horned dino from South Korea
• Triceratops had horn-to-horn battles with kin
• Dinosaur had horns the size of baseball bats

John Roach is a contributing writer for msnbc.com. Connect with the Cosmic Log community by hitting the "like" button on the Cosmic Log Facebook page or following msnbc.com's science editor, Alan Boyle, on Twitter (@b0yle).