Background: The proteins belonging to the collagen family are ubiquitous throughout the animal kingdom. The most abundant collagen, type 1, readily forms fibrils that convey the principal mechanical support and structural organization in the extracellular matrix of connective tissues such as bone, skin, tendon, and vasculature. An understanding of the molecular arrangement of collagen in fibrils is essential since it relates molecular interactions to the mechanical strength of fibrous tissues and may reveal the underlying molecular pathology of numerous connective tissue diseases.Results: Using synchrotron radiation, we have conducted a study of the native fibril structure at anisotropic resolution (5.4 Angstrom axial and 10 Angstrom lateral). The intensities of the tendon X-ray diffraction pattern that arise from the lateral packing (three-dimensional arrangement) of collagen molecules were measured by using a method analogous to Rietveld methods in powder crystallography and to the separation of closely spaced peaks In Laue diffraction patterns. These were then used to determine the packing structure of collagen by MIR.Conclusions: Our electron density map is the first obtained from a natural fiber using these techniques (more commonly applied to single crystal crystallography). It reveals the three-dimensional molecular packing arrangement of type I Collagen and conclusively proves that the molecules are arranged on a quasihexagonal lattice. The molecular segments that contain the telopeptides (central to the function of collagen fibrils in health and disease) have been identified, revealing that they form a corrugated arrangement of crosslinked molecules that strengthen and stabilize the native fibril.
Background: Type I collagen contains specific lysine and hydroxylysine residues that are critical in the formation of intermolecular cross-links crucial for the normal configuration and stability of the 67 nm axial repeat:of collagen fibrils in the extracellular matrix. The major cross-linkage sites are believed to occur between the non-helical terminal regions (telopeptides) and helical segments of adjacent collagen molecules. In this X-ray fibre diffraction study the tissue has been maintained in the hydrated fibrillar state,,whilst detailed structural information was obtained using highly collimated, synchrotron radiation.Results: The axial component of the X-ray diffraction patterns extends more than twice as far in reciprocal space than that of any already:published. The structure-factor phases were calculated using the multiple isomorphous addition method, avoiding model-based approaches, and produced an electron-density profile of the molecular arrangement projected on to the fibre axis to 0.54 nm resolution. This corresponds to the phasing of 124 orders of the meridional diffraction pattern.Conclusions: The axially projected electron-density profile and the electron-density difference maps showed that both the N- and C-terminal telopeptides are contracted structures. This profile puts narrow constraints on the possible conformations of the C-terminal telopeptide; the best fit to the electron-density profile is when the alpha 1 chains adopt a folded conformation with a Sharp hairpin turn around residues 13 and 14 of the 25-residue telopeptide. Our results reveal for the first time the location, parallel to the fibril axis, of the intermolecular cross-links in normal hydrated tissue. These cross-links are essential for the biological function of the tissue.
X-ray diffraction of rat tail tendon shows that type I collagen fibrils contain regions of three-dimensional crystalline arrays; where molecular packing is speculated to be by a staggered sheet or microfibril arrangement. The X-ray diffraction pattern also contains a significant amount of diffuse scatter indicative of static and thermal disorder in fibrils. Removal of the diffuse scatter from the equatorial region of X-ray diffraction patterns obtained using synchrotron radiation allowed the Bragg intensities to be viewed on a flat background. Indexing of Bragg peak intensity on the 10, −10, 0 –1, 01, −11 and 1 –1 row-lines of the triclinic unit cell have been used here to test possible sheet and microfibril packing arrangements. The relative translation of molecular segments in the gap and overlap regions as well as the telopeptide orientation have been investigated. A global search through combinations of molecular packing and molecular translation revealed that the sheet-type conformations cannot account for the observed low-angle off-meridional Bragg peak intensity distribution. A superior fit is obtained with D-staggered left-handed microfibril structures. The orientation of the telopeptides may indicate that there are interconnections between microfibrils that may explain the difficulty in isolating individual microfibrillar structures.
In this review, recent results from X-ray diffraction studies of tendon are used to develop an understanding of the molecular packing of type I collagen in tendon fibrils. These cover the definition of the unit cell as triclinic, the lateral architecture of molecular packing in a fibril and the molecular packing topology of a structure that gives good agreement with X-ray diffraction data. The proposed model is a 1D staggered left handed microfibril; the molecular orientation of the telopeptides indicates that there are interconnections between microfibrils that may explain the difficulty in isolating individual microfibrillar structures. This is the first structure that defines the absolute molecular packing of molecular segments based on X-ray diffraction data. These results are discussed in the light of direct and indirect evidence relating to molecular packing such as mineralization, natural crosslink position, and biomechanical evidence. The ability of the proposed structure to fulfill many of the structural and biochemical criteria point towards the structure providing a basis for a consensus model of collagen packing.
The chemical reactivity of collagen can be studied using neutron diffraction (a non-destructive technique), for certain reaction types. Collagen contains a number of lysine and hydroxylysine side chains that can react with aldehydes and ketones, or these side chains can themselves be converted to aldehydes by lysyl oxidase. The reactivity of these groups not only has an important role in the maintenance of mechanical strength in collagen fibrils, but can also manifest pathologically in the cases of aging, diabetes (reactivity with a variety of sugars) and alcoholism (reactivity with acetaldehyde). The reactivity of reducing groups with collagen can be studied by neutron diffraction, since the crosslink formed in the adduction process is initially of a Schiff base or keto-imine nature. The nature of this crosslink allows it to be deuterated, and the position of this relatively heavy scattering atom can be used in a process of phase determination by multiple isomorphous replacement. This process was used to study the following: the position of natural crosslinks in collagen; the position of adducts in tendon from diabetic rats in vivo and the in vitro position of acetaldehyde adducts in tendon.
Background/Aims-Maternal diet may have an effect on the health of the offspring in middle and later life. This study used the laboratory rat as an animal model to examine whether the fibre content of the maternal diet during pregnancy affected subsequent development of colonic diverticula in the offspring fed lifelong fibre deficient or higher fibre diets.Methods-The parents of experimental animals were fed either a diet that was known to predispose to colonic diverticulosis or a control diet for one month prior to mating. The offspring were fed one of these diets for 18 months. The incidence of colonic diverticulosis, submucosal collagen content, collagen solubility in weak acid, and the composition of intestinal contents were then measured.Results-Offspring of rats fed a higher fibre diet from higher fibre diet fed parents had 0% incidence of colonic diverticulosis. When offspring (regardless of parental diet) were fed a low fibre diet for life the acid solubility was lowered compared with rats fed lifelong higher fibre diet mean (SD) (0.044 (0.0007) v 0.073 (0.0015) sigmoid colon (ratio of soluble:insoluble collagen)); 21.1% had diverticulosis and there was reduced fibre fermentation. However when the diet of the parents of the fibre deficient diet fed rats was considered, the animals whose mothers had a fibre deficient diet had lower acid solubility (0.032 (0.0007)) and an increased incidence of colonic diverticulosis (42.1%) than the animals fed a fibre deficient diet from higher fibre diet fed parents (p<0.01 in all instances).Conclusion-Maternal diet and the subsequent nutrition of the progeny seem to be of importance in the development of colonic diverticulosis in the rat.
BACKGROUND--Changes in the structure and integrity of the colon dependent on collagen content and crosslinkage occur with age. AIMS--This study using an animal model examines colonic collagen content and crosslinkage over the lifetime of rats on fibre deficient and higher fibre diets. METHODS--Two groups of 20 rats were fed either a fibre deficient diet (1.7 g NSP (non-starch polysaccharide)/100 g) or a higher fibre diet (13.3 g NSP/100 g) for 18 months. Diverticula were identified by postmortem examination. Caecal and colonic contents were weighed and assayed for short chain fatty acids. Collagen solubility in weak acid was measured to give an indication of the nature and amount of crosslinks in the collagen of the bowel wall. RESULTS--The incidence of colonic diverticula was greater (42.1% fibre deficient rats; 0% higher fibre rats). Colonic collagen solubility index in fibre deficient rats was significantly lower than higher fibre diet fed rats (p < 0.001 in all four sections of the large bowel). Rats with diverticula had the lowest solubility index (p < 0.001 in all four sections of the large bowel). Higher fibre diet rats had increased caecal and colonic contents, caecal and colonic tissue wet weights, and greater caecal short chain fatty acids. Fibre deficient diet fed rats had more pathological abnormalities. CONCLUSIONS--This animal model permits a study of the relation between collagen crosslinkage and the development of colonic diverticulosis. A higher fibre diet protects against collagen crosslinking and this is related to a decreased incidence of diverticula.
Development of colonic diverticulosis is a function of age and declining colonic wall mechanical strength. The latter is partly a consequence of changes in the collagen structure. Collagen from unaffected human colons (n = 20, age range 20-80 years) and those with colonic diverticulosis (n = 5, age range 67-80 years) were obtained at necropsy. The total collagen content was measured as the hydroxyproline content and cross linkage by collagen solubility in weak acid was studied. The colonic total collagen content was constant with age (mean (SD) 15.8 (0.3) mg/100 mg wet weight of tissue). The acid solubility of the collagen, however, increased after the age of 40 years: at over 60 years, colonic diverticulosis was associated with an increased acid solubility ratio compared with values in unaffected colons (15.3 (0.2); compared with 9.2 (0.2), p < 0.001). The cross linking of colonic collagen increases with age. These changes seem to be a factor in the aetiology of colonic diverticulosis.
The X-ray diffraction pattern of tendon collagen can contain a number ofsharp Bragg peaks indicating three-dimensional crystallinity of the sample. Optimal diffraction images have been obtained with a high flux synchrotron X-ray source and a carefully maintained sample environment and staining techniques. The Bragg peaks are always superimposed on a diffuse background. This makes interpretation of data difficult and a number of conflicting models of collagen packing have been proposed. The removal of the diffuse scatter from the images allows the Bragg peaks to be seen on a relatively flat background. This was conducted by modelling the background points as a series of two-dimensional polynomial functions. The resultant set of observed Bragg reflections serves as an excellent basis to test the validity of two contradictory packing modes; (1) the triclinic model, Fraser et al., (2) the microfibril model, Kajava. From this it can easily be seen that the model proposed by Kajava is inappropriate, since there is limited agreement between predicted positions of reflections and the positions of observable reflections on film. The packing of collagen molecules on a triclinic lattice is favoured by this criterion.
Plasma fibronectin was purified by gelatin affinity chromatography in the absence of urea and studied by photon correlation spectroscopy. Polydispersity in the observed translational diffusion coefficient (D20,w) was minimized by subsequent gel permeation fast protein liquid chromatography (FPLC) on Superose 6, which separated fibronectin monomers (D20,w = 2.15 +/- 0.03 x 10(-7) cm2 sec-1; polydispersity 5.2%) from aggregates. Addition of heparin to FPLC-purified fibronectin, at physiological pH, ionic strength and temperature, induced fibronectin aggregation, as shown by an increase of up to 60% in the static light-scattering intensity. Additional changes induced by heparin were an approximate 40% decrease in D20,w and an increase in polydispersity to 33%. After removal of aggregates by FPLC, the translational diffusion coefficient for fibronectin monomers was unaffected by the presence of heparin, in conditions where fluorescence polarization with fluoresceinamine-labelled heparin showed that 80% of the available heparin binding sites on fibronectin were occupied. Small differences in the circular dichroism spectrum of gelatin affinity-purified fibronectin were observed before and after removal of aggregates by gel permeation FPLC, and similar changes were seen when heparin was added to FLPC-purified fibronectin, without subsequent removal of aggregates. The results demonstrate the importance of minimizing polydispersity in the biophysical analysis of fibronectin in solution. We conclude that heparin binding to monomeric fibronectin occurs without large-scale changes in the conformation of the fibronectin molecule, although the possibility of more extended conformations in aggregated forms of fibronectin cannot be excluded.
Glycation (non-enzymatic glycosylation) sites in the axial unit cell of diabetic tendon collagen were investigated by neutron diffraction. Samples of diabetic and control tendon were reacted with sodium borodeuteride and sodium cyanoborodeuteride. This facilitated deuteration at aldimine, aldol or ketoimine groups in the molecule. These are natural collagen cross-links and sites where non-enzymatic glycation had occurred. The introduction of a deuteron at specific locations allowed the diabetic glycation collagen to be treated as multiple isomorphous derivatives for neutron fibre diffraction. Neutron diffraction was conducted at the Institut Laue Langevin, Grenoble. Standard crystallographic refinement techniques (modified for axial projections) were used to determine the structure of the control (non-diabetic) and diabetic samples. The results are shown as difference maps, these indicate that glycation takes place at different rates within the collagen axial unit cell. The position of glycation correlates well with the position of hydroxylysine residues. The reactions of periodate with enzymatically attached sugars, proteoglycan, natural cross-links and glycation products lead to complications in map interpretation.
A 2D focusing SAXS camera has been build based on a circular Bragg-Fresnel lens. The camera has been tested at the Microfocus beamline at the European Synchrotron Radiation Facility. An undulator was used as radiation source. The focused beam size at the sample position of about 1,5*2mum2 and intensity of about 10(9) photons/s was obtained at a wavelength of 1.24A (10keV). A Molecular Dynamics image plate was used to detect the low -angle diffraction pattern. For native turkey leg tendon collagen intermediate areas between the calcified and non-calcified regions were analysed. More than 22 orders including the first order peak were clearly resolved.
Biochemical determinations of the collagen and elastin content in 50 mm3 samples of human lung are presented in relation to morphometric measurements of lung structure, as the amount of alveolar wall surface area per unit volume (A WUV), on adjacent slices. There were no differences in A WUV values, collagen content (determined as hydroxyproline) or elastin content (determined as isodesmosine) between upper and lower lobes within a single lung. In a study of 102 samples from 9 smokers lungs with no evidence of macro- or microscopic emphysema (as estimated by A WUV measurement), there was a negative correlation between A WUV and the amounts of collagen or elastin per unit volume of inflated lung. The correlation was stronger when collagen and elastin content were expressed per unit area of alveolar wall. The negative correlation is interpreted as representing either the anatomical variation within the complex hierarchy of normal lung structure or possibly low levels of fibrosis in response to cigarette smoking.
Using a Fresnel zone plate we demonstrate for the first time the direct visualization by x-ray microscopy of suboptical regularity in a biological specimen, namely the 65-nm axial periodicity of tendon collagen. This resolution test demonstrates a resolving power of about 20λ; a resolving power of <10λ is in prospect.
Alphaviruses, like many enveloped animal viruses, enter the cell by fusing with the cell membrane. This fusion occurs only in coated vesicles at a low pH. By using X-ray solution scattering of highly purified virus particles we have gained direct evidence that a drop in pH does not alter the structure of the virus core but does cause a significant change in the structure of the virus envelope. Thus these experiments give direct evidence to support the hypothesis that a reduction in pH causes a conformational change in the virus E protein, which enable it to promote fusion with the cell envelope and trigger virus infection.
Alphaviruses, like many enveloped animal viruses, enter the cell by fusing with the cell membrane. This fusion occurs only in coated vesicles at a low pH. By using X-ray solution scattering of highly purified virus particles we have gained direct evidence that a drop in pH does not alter the structure of the virus core but does cause a significant change in the structure of the virus envelope. Thus these experiments give direct evidence to support the hypothesis that a reduction in pH causes a conformational change in the virus E protein, which enable it to promote fusion with the cell envelope and trigger virus infection.
The translational diffusion coefficient (D20,w) of human platelet thrombospondin was measured by dynamic light-scattering. D20,w, measured in 20 mM-Hepes buffer, pH 7.4, containing 350 mM-NaCl and 2 mM-CaCl2, was 1.73(+/- 0.02) x 10(-7) cm2.s-1. After removal of bound Ca2+ by addition of EDTA, D20,w decreased to 1.56(+/- 0.04) x 10(-7) cm2.s-1; this was not a consequence of aggregation. D20,w showed little sensitivity to NaCl concentration between 130 and 550 mM. Through hydrodynamic analysis combining D20,w and other parameters taken from the literature, two major types of models for thrombospondin can be proposed: either classic compact models (i.e. low degree of hydration) such as prolate or oblate ellipsoids with a high axial ratio (greater than 20) or models of low axial ratio made of multiple subunits with significant cavities (i.e. high degree of hydration).