Sometimes it is necessary to revisit established ideas that are widely circulated as dogma. Investigators with University College London affirmed this recently when studying a variety of collagen-based samples — including rat-tail extracts, engineered collagen sheets used in implants, and historical parchments made from animal skin — with atomic force microscopy. “The idea was to review the morphology of the collagen itself, but at the nanoscale,” explained researcher Laurent Bozec. “After a few images, we found the well-known D-banding — a consistent periodicity along the fibril — and wanted to understand where that characteristic banding was coming from.” Researchers have used atomic force microscopy (AFM) to uncover the morphology of collagen fibrils — as seen in the optical image in (a). For decades, investigators have subscribed to a model of fibril structure that explains the periodicity in some fibrils as a staggered repeat of collagen molecules, but researchers at University College London have confirmed a twisted, ropelike structure. Here, AFM images of collagen fibrils from bovine digital tendons (c) are contrasted with those from rat tails (d), which exhibit the same banding but which are much straighter . Panel (b) is a lower-magnification AFM image of the tendon collagen fibrils. Images reprinted with permission of Biophysical Journal. The problem was that most everyone who studies such things — including first-year medical students — already knows where it comes from. Or they think they do, anyway. Hodge and Petruska proposed a model of collagen fibril formation more than 40 years ago and, although it leaves open several structural and functional questions, this model remains more or less the final word on the topic. “We didn’t want to take any early studies for granted,” Bozec explained, “so we had to come up with our own model and see where it fit in the current literature.” The accepted wisdom suggests that collagen fibrils are formed through staggered repeats of individual molecules — leading to the consistent periodicity along the fibrils. This model is essentially two-dimensional, however. Other investigations have sought to address the resulting deficiencies — for example, one study posited a layered, spiral arrangement of collagen molecules. But studies have failed to explain a number of characteristics, including the preservation of D-banding regardless of fibril diameter and the fact that fibrils up to 10 nm in diameter form into fibrils of greater diameter — and thus into macroscale objects such as tendons. Mechanical modeling further supported the ropelike structure of the fibrils. Image (a) shows patterns for a single strand (top) and the corresponding two-ply (bottom), for a fibril with no intrinsic twist. Image (b) shows a pattern for a six-ply fibril with an intrinsic twist along with a corresponding AFM image. Such modeling helped make experimentally verifiable predictions about the fibrils’ behavior. More recent studies have revealed a possible ropelike structure to the fibrils. Unlike the theories from the earlier studies, this would satisfy an established dictum in biology: A tissue’s form follows its function. Hence, the twisted rope features of the fibrils — the source of periodicity — may be crimps resulting from the relaxation of subcomponents of the fibrils that contribute in some way to their mechanical features. In a Biophysical Journal paper published online Oct. 6, the University College London researchers reported a study in which they used atomic force microscopy to confirm the ropelike model of fibril periodicity. They sought to understand the architecture of the collagen fibril by trying to open the fibril itself. “Other studies have shown some degree of success by looking at the different stages of the fibrillogenesis and deducing the overall architecture,” Bozec said. “We took an intact fibril and looked at it under a microscope as it unravels.” “Scientists believe the ‘messy’ structure they sometimes see is just wrong,” he added. “We decided to look deeper inside those abnormalities and discovered clues that helped us to understand the fibrillar architecture.” To this end, they used atomic force microscopes made by Veeco Instruments Inc. of Santa Barbara, Calif., and by JPK Nanowizard AG of Berlin to record topologic (height) and error signal (deflection) images of native bovine digital tendon and rat-tail (or flexor digitalis) tendon collagen fibrils. They operated the microscopes in contact mode using cantilever tips also made by Veeco Instruments. The images provided very strong evidence of a ropelike structure — suggesting that many investigators must re-evaluate their existing structural data sets, as well as their long-held beliefs about periodicity. The new model also offers experimentally verifiable predictions. For example, if fibrils’ ropelike structures are a response to their mechanical needs, then fibrils exposed to tension should be more inclined to twisting during formation. All of which has important implications for understanding the biomechanics of the human body, and clinical implications in particular. For example, investigators are working to develop collagen-based scaffolds for tissue-engineering applications. The insights afforded by the present study suggest that twisted fibrils should be used for such scaffolds in arteries and other mechanically stressed locations, because the fibrils are essentially designed to meet the needs of scaffolds in these locations. Questions remain, however. Regardless of topology or dimensions, the fibrils had one thing in common: All exhibited D-banding periodicity of 69.6 ±2.9 nm. What accounts for the apparent disconnect between topology and D-banding periodicity? And how do single molecules come together to form this periodicity independent of fibril diameter? To address these and other questions, the researchers are working to refine the model and to look deeper inside the fibril with atomic force microscopy. The model is currently incomplete, Bozec said, because it simplifies the triple-helical structure of the molecule to that of a patterned cylinder. “We need to use a more realistic base unit for an even closer-to-reality model.”