Quantification of Collagen Ultrastructure after Penetrating Keratoplasty – Implications for Corneal Biomechanics

Purpose

To quantify long-term changes in stromal collagen ultrastructure following penetrating keratoplasty (PK), and evaluate their possible implications for corneal biomechanics.

Methods

A pair of 16 mm post-mortem corneo-scleral buttons was obtained from a patient receiving bilateral penetrating keratoplasty 12 (left)/28 (right) years previously. Small-angle x-ray scattering quantified collagen fibril spacing, diameter and spatial order at 0.5 mm or 0.25 mm intervals along linear scans across the graft margin. Corresponding control data was collected from two corneo-scleral buttons with no history of refractive surgery. Wide-angle x-ray scattering quantified collagen fibril orientation at 0.25 mm (horizontal)×0.25 mm (vertical) intervals across both PK specimens. Quantification of orientation changes in the graft margin were verified by equivalent analysis of data from a 13 year post-operative right PK specimen obtained from a second patient in a previous study, and comparison made with new and published data from normal corneas.

Results

Marked changes to normal fibril alignment, in favour of tangentially oriented collagen, were observed around the entire graft margin in all PK specimens. The total number of meridional fibrils in the wound margin was observed to decrease by up to 40%, with the number of tangentially oriented fibrils increasing by up to 46%. As a result, in some locations the number of fibrils aligned parallel to the wound outnumbered those spanning it by up to five times. Localised increases in fibril spacing and diameter, with an accompanying reduction in matrix order, were also evident.

Conclusions

Abnormal collagen fibril size and spatial order within the PK graft margin are indicative of incomplete stromal wound remodelling and the long term persistence of fibrotic scar tissue. Lasting changes in collagen fibril orientation in and around PK wounds may alter corneal biomechanics and compromise the integrity of the graft-host interface in the long term.     

Variations in collagen fibril structure in tendons

Variation of collagen fibril structure in tendon was investigated by x-ray diffraction. Anatomically distinct tendons from single species, as well as tendons from different species, were examined to determine the variations that exist in both the axial and lateral structure of the collagen fibrils. The meridional diffraction is derived from the axial collagen fibril structure. Anatomically distinct tendons of a particular species give meridional patterns that are indistinguishable within experimental error. The meridional diffraction patterns from tendons of different mammals are similar but show small species-specific variations, most noticeably in the 14th–18th orders. Tendons of birds also give meridional patterns that are similar to each other, but the avian patterns differ considerably from the mammalian ones. Avian tendons give stronger odd and weaker even low orders, a feature consistent with a reduced gap:overlap ratio, and have a distinctive intensity pattern for the higher meridional orders. Interpretation of these differences has been approached using biochemical data, diffraction by reconsituted fibers of purified collagen, and Fourier transform analysis. From these methods, it appears that the variations observed in the lower orders (2nd–8th) and in the higher orders (29th–52nd) are probably related to differences in the primary structure of the Type I collagen found in the different species. The variations observed in the 14th–18th orders appear not to be related to features within the triple-helical domain of the molecule. Equatorial diffraction yields information on the lateral packing of collagen molecules in the fibrils, and considerable variation was seen in different tendons. Rat tail tendon gives sharp Bragg reflections, demonstrating the presence of a crystalline lateral arrangement of molecules in the fibril. For the first time, sharp lattice reflections similar to those in rat tail tendon have been observed in nontail tendons, including rat achilles tendon, rabbit leg tendon, and wing and leg tendons of quail. In the rabbit and quail tendons, one of the strong equatorial reflections characteristic of the rat tendon pattern, at 1.26 nm, was absent. The positions of the equatorial maxima, which are a measure of intermolecular spacing, varied considerably, being smallest in the specimens displaying crystalline packing. The intermolecular distance in chiken and turkey leg tendons is longer than that found in mammalian tendons, or in avian wing tendons, which supports the hypothesis that a larger intermolecular spacing is characteristic of tendons that calcify. Thus, x-ray diffraction indicates there are reproducible differences in both the axial and lateral structure of collagen fibrils among different tendons. This work on tendon, a tissue containing almost exclusively Type I collagen as its major component, should serve as a basis for analyzing the structure of other connective tissues, which contain different genetic types of collagen and larger amounts of noncollagenous components.     

Radial packing, order, and disorder in collagen fibrils.

Collagen fibrils resemble smectic, liquid crystals in being highly ordered axially but relatively disordered laterally. In some connective tissues, x-ray diffraction reveals three-dimensional crystallinity in the molecular packing within fibrils, although the continued presence of diffuse scatter indicates significant underlying disorder. In addition, several observations from electron microscopy suggest that the molecular packing is organized concentrically about the fibril core. In the present work, theoretical equatorial x-ray diffraction patterns for a number of models for collagen molecular packing are calculated and compared with the experimental data from tendon fibrils. None of the models suggested previously can account for both the crystalline Bragg peaks and the underlying diffuse scatter. In addition, models in which any of the nearest-neighbor, intermolecular vectors are perpendicular to the radial direction are inconsistent with the observed radial orientation of the principal approximately 4 nm Bragg spacing. Both multiple-start spiral and concentric ring models are devised in which one of the nearest-neighbor vectors is along the radial direction. These models are consistent with the radial orientation of the approximately 4 nm spacing, and energy minimization results in radially oriented crystalline domains separated by disordered grain boundaries. Theoretical x-ray diffraction patterns show a combination of sharp Bragg peaks and underlying diffuse scatter. Close agreement with the observed equatorial diffraction pattern is obtained. The concentric ring model is consistent with the observation that the diameters of collagen fibrils are restricted to discrete values.     

Synchrotron x-ray diffraction studies of the cornea, with implications for stromal hydration.

The intermolecular and interfibrillar spacings of collagen in bovine corneal stroma have been measured as a function of tissue hydration. Data were recorded from low- and high-angle x-ray diffraction patterns obtained using a high intensity synchrotron source. The most frequently occurring interfibrillar spacing varied from 34 nm in dry corneas to 76 nm at H = 5 (the hydration, H, is defined as the ratio of the weight of water to the dry weight). The most frequently occurring intermolecular Bragg spacing increased from 1.15 nm (dry) to approximately 1.60 nm at normal hydration (H approximately 3.2) and continued to increase only slowly above normal hydration. Most of the increase in the intermolecular spacing occurred between H = O and H = 1. Over this hydration range the interfibrillar and intermolecular spacings moved in tandem, which suggests that the initial water goes equally within and between the fibrils. Above H = 1 water goes preferentially between the fibrils. The results suggest that, even at normal hydration, water does not fill the interfibrillar space uniformly, and a proportion is located in another space or compartment. In dried-then-rehydrated corneas, a larger proportion of the water goes into this other compartment. In both cases, it is possible to postulate a second set or population of fibrils that are more widely and irregularly separated and therefore do not contribute significantly to the diffraction pattern.     

Development of collagen fibril organization and collagen crimp patterns during tendon healing

Normal tendon comprises coaxially aligned bundles of crimped collagen fibres each of which possesses a fibrillar substructure. In acute traumatic injury this level of organization is disrupted and the mechanical function of the tendon impaired. During repair, a degree of recovery of the fibrillar structure takes place. In this tudy we have assessed the re-establishment of tendon organization after injury on the basis of the collagen fibril diameter distribution and the collagen crimp parameters. Crimp became undetectable following injury but one month later was present throughout the tissue. At this time the periodicity was greatly reduced by comparison with that of the normal tendon and normal values were not re-established within 14 months following injury. Collagen fibril diameters remained abnormally small over this same period of time. In particular, fibrils of diameters in excess of 100 nm, commonly found in normal and contralateral tendons, were totally absent from the observed distributions in the healing tendons. Such large diameter fibrils often account for as much as 50% of the total mass of collagen present in the uninjured tissue. Thus the mechanical properties of the healing tendon may remain significantly different from those of normal tendon for a minimum time of 14 months after injury.     

A study of dense mineralized tissue by neutron diffraction

Neutron diffraction studies of mineralized tissue show a close relationship between the wet state equatorial diffraction spacing and wet tissue density expressable as a second-order polynomial. The molecular fractional shrinkage when the tissue is dried shows a straight line dependence on wet tissue density with a correlation of 0.98. Since the dry state equatorial diffraction spacing is much less than for the corresponding wet state, even in fully mineralized bone, the collagen molecules must be displaced through a mineral-free volume while drying. The mineral can only be located within the available volume of the dried tissue whether intra- or extrafibrillar. The dimension of the dry state equatorial spacing for each of the tissues examined is close to that of dried tendon collagen. It appears unlikely that hydroxyapatite crystallites can be accommodated radially between collagen molecules in bone if the packing is like that of dried tail tendon collagen. The only mineral within the fibrils must be in the intermolecular gaps. It is estimated on the basis of the volume of the axial intermolecular gaps and the minimum extrafibrillar volume that the intrafibrillar mineral can be no more than 20% of the total mineral and may be less than 10%.     

Mineralization of type I collagen

It was previously found that the lateral spacing of the collagen molecules in wet mineralized tissues is exactly proportional to the inverse wet density. Several properties were investigated and the same type of relationship was observed each time. A possible explanation is offered. It is hypothesized that mineral is deposited initially in the extrafibrillar space so as to isolate the fibrils. Further deposition reduces the net free fibril volume thereby decreasing the spacing between collagen molecules. The linear relationship is derived from density considerations together with limitations on the collagen packing structure described as the generalized packing model. Three experimental situations were studied: lateral spacing wet tissue versus density; lateral spacing dry tissue versus density; and lateral spacing versus water content. The observed variations of the spacing can be attributed to a structure where the mass of the tissue remains constant but the volume decreases linearly with increasing mineral content.     

Synchrotron X-ray diffraction study of corneal stroma

Using a synchrotron X-ray source, it has been possible to record a low-angle diffraction pattern from fresh bovine corneal stroma.The pattern can be interpreted as arising from the short-range order packing of collagen fibrils in lamellae. Model calculations suggest that the positions of the fibrils remain correlated over distances corresponding to, at most, three fibril diameters (~ 120 nm). These results support theories of transparency of the cornea based on short-range order.Further, a study of the fibril spacing as a function of hydration confirms that water uptake occurs largely between the lamellae and in regions devoid of collagen fibrils, and shows that the fibril diameter increases with hydration.     

Collagen fibril network and elastic system remodeling in a reconstructed skin transplanted on nude mice.

Wound healing of deep and extensive burns can induce hypertrophic scar formation, which is a detrimental outcome for skin functionality. These scars are characterized by an impaired collagen fibril organization with fibril bundles oriented parallel to each other, in contrast with a basket weave pattern arrangement in normal skin. We prepared a reconstructed skin made of a collagen sponge seeded with human fibroblasts and keratinocytes and grown in vitro for 20 days. We transplanted it on the back of nude mice to assess whether this reconstructed skin could prevent scar formation. After transplantation, murine blood vessels had revascularized one-third of the sponge thickness on the fifth day and were observed underneath the epidermis at day 15. The reconstructed skin extracellular matrix was mostly made of human collagen I, organized in loosely packed fibrils 5 days after transplantation, with a mean diameter of 45 nm. After 40-90 days, fibril bundles were arranged in a basket weave pattern while their mean diameter increased to 56 nm, therefore exactly matching mouse skin papillary dermis organization. Interestingly, we showed that an elastic system remodeling was started off in this model. Indeed, human elastin deposits were organized in thin fibrils oriented perpendicular to epidermis at day 90 whereas elastic system usually took years to be re-established in human scars. Our reconstructed skin model promoted in only 90 days the remodeling of an extracellular matrix nearly similar to normal dermis (i.e. collagen fibril diameter and arrangement, and the partial reconstruction of the elastic system).     

The regulation of size and form in the assembly of collagen fibrils in vivo

A possible mechanism for regulating the lateral growth of collagen fibrils in vivo is considered. A growth inhibitor associated with a particular part of the long semiflexible collagen molecule restricts that part of the molecule to the surface of the growing assembly. Lateral accretion ceases when these inhibitors form a complete circumferential layer around the fibril surface. Cell-mediated removal of the inhibitors allows lateral growth to proceed to a second limiting layer, and so on to subsequent limiting layers. In this way, cycles of inhibitor removal and limited lateral accretion permit growth to be synchronized over large populations of fibrils. Observed diameter distributions in bundles of embryonic and neonatal fibrils are those expected from a mechanism of this kind. The mechanism depends on the existence of axial order (D-periodicity) in fibrils, but not on any specific lateral packing of molecules. Rather, contacts between newly assembled molecules are presumed to be partly fluid-like in lateral directions (except where covalent cross-links have formed). Some initial fluidity in lateral packing prior to cross-linking does not preclude the subsequent emergence of quasi-crystalline packing as cross-links form. The cylindrical shape of fibrils in vivo may also be attributable in part to fluidity of intermolecular contacts at the growing surface.     

Collagen fibril formation in a wound healing model

Control of tissue composition and organization will be a key feature in the development of successful products through tissue engineering. However, the mechanism of collagen fibril formation, growth, and organization is not yet fully understood. In this study we have examined collagen fibril formation in a wound healing model in which the newly formed fibrils were kept distinct from preexisting tissue through use of a porous tubular biomaterial implant. Samples were examined after 4, 6, 14, and 28 days by light microscopy, in situ hybridization, and immunofluorescence microscopy. These showed a normal wound healing response, with significant collagen formation at 14 and 28 days. Individual collagen fibrils were isolated from these samples by gentle extraction in a gentamicin-containing buffer which allowed extraction of a large proportion of intact fibrils. Examination by transmission electron microscopy showed that approximately 80% of the intact fibrils showed a single polarity reversal, with both ends of each fibril comprising collagen amino-terminal domains; the remaining fibrils had no polarity reversal. All fibrils had similar diameters at both time points. Immunoelectron microscopy showed that all labeled fibrils contained both type I and III collagens. These data indicate that this wound healing model provides a system in which collagen fibril formation can be readily followed.     

X-ray diffraction study into the effects of liming on the structure of collagen

The manufacture of parchment from animal skin involves processes that remove hair, fats, and other macromolecules. Although it is well understood that the collagen fibers "open up" during processing, this study uses small and wide-angle X-ray diffraction to measure quantitatively the changes induced at the nanoscopic and microscopic levels. The axial rise per residue distance within the collagen molecules is unaffected by salt and lime treatments. Salting of the hides appears to remove noncollagenous materials. The intermolecular lateral packing distance between the hydrated collagen molecules (1.4 nm) increases after salting ( approximately 1.5 nm) and liming ( approximately 1.55 nm); drying is responsible for a reduction to approximately 1.2 nm in all samples. The axial staggered array (d spacing) is reduced by 1 nm after liming and is unaffected by drying. The average fibril diameter increases from 103.2 to 114.5 nm following liming, and the fibril-to-fibril distance increases from 122.6 to 136.1 nm.     

Fibrous long spacing collagen ultrastructure elucidated by atomic force microscopy.

Fibrous long spacing collagen (FLS) fibrils are collagen fibrils in which the periodicity is clearly greater than the 67-nm periodicity of native collagen. FLS fibrils were formed in vitro by the addition of alpha1-acid glycoprotein to an acidified solution of monomeric collagen and were imaged with atomic force microscopy. The fibrils formed were typically approximately 150 nm in diameter and had a distinct banding pattern with a 250-nm periodicity. At higher resolution, the mature FLS fibrils showed ultrastructure, both on the bands and in the interband region, which appears as protofibrils aligned along the main fibril axis. The alignment of protofibrils produced grooves along the main fibril, which were 2 nm deep and 20 nm in width. Examination of the tips of FLS fibrils suggests that they grow via the merging of protofibrils to the tip, followed by the entanglement and, ultimately, the tight packing of protofibrils. A comparison is made with native collagen in terms of structure and mechanism of assembly.     

X-ray diffraction studies of the corneal stroma.

A low-angle diffraction pattern has been obtained from corneal stroma. This pattern arises both from the arrangement of the collagen fibrils and from the packing of the tropocollagen molecules along the axes of the fibrils. The spacing arising from the packing of the fibrils increases homogeneously on swelling although the tissue as a whole swells only radially referred to the intact eye. The necessary rearrangement of the fibrils for this type of swelling to occur might result in the formation of regions devoid of collagen fibrils and the water not in the lattice of collagen fibrils could be synonymous with the lakes postulated by Benedek (1971) to explain the loss of transparency on swelling.The spacings due to the packing of the tropocollagen molecules are unusual in that, although they index as the third and fifth orders of the well-known 66 nm repeat, the first order of this spacing is absent. Calculation of the Patterson function for corneal collagen leads to peaks in electron density separated by distances of 0.38 and 0.24 of the repeat distance.     

Structure of type I and type III heterotypic collagen fibrils: an X-ray diffraction study

The molecular packing arrangement within collagen fibrils has a significant effect on the tensile properties of tissues. To date, most studies have focused on homotypic fibrils composed of type I collagen. This study investigates the packing of type I/III collagen molecules in heterotypic fibrils of colonic submucosa using a combination of X-ray diffraction data, molecular model building, and simulated X-ray diffraction fibre diagrams. A model comprising a 70-nm-diameter D- (approximately 65 nm) axial periodic structure containing type I and type III collagen chains was constructed from amino acid scattering factors organised in a liquid-like lateral packing arrangement simulated using a classical Lennard-Jones potential. The models that gave the most accurate correspondence with diffraction data revealed that the structure of the fibril involves liquid-like lateral packing combined with a constant helical inclination angle for molecules throughout the fibril. Combinations of type I:type III scattering factors in a ratio of 4:1 gave a reasonable correspondence with the meridional diffraction series. The attenuation of the meridional intensities may be explained by a blurring of the electron density profile of the D period caused by nonspecific or random interactions between collagen types I and III in the heterotypic fibril.     

Neutron diffraction studies of collagen in fully mineralized bone

Neutron diffraction measurements have been made of the equatorial and meridional spacings of collagen in fully mineralized mature bovine bone and demineralized bone collagen, in both wet and dry conditions. The collagen equatorial spacing in wet mineralized bovine bone is 1.24 nm, substantially lower than the 1.53 nm value observed in wet demineralized bovine bone collagen. Corresponding spacings for dry bone and demineralized bone collagen are 1.16 nm and 1.12 nm, respectively. The collagen meridional long spacing in mineralized bovine bone is 63.6 nm wet and 63.4 nm dry. These data indicate that collagen in fully mineralized bovine bone is considerably more closely packed than had been assumed previously, with a packing density similar to that of the relatively crystalline collagens such as wet rat tail tendon. The data also suggest that less space is available for mineral within the collagen fibrils in bovine bone than had previously been assumed, and that the major portion of the mineral in this bone must be located outside the fibrils.     

The extracellular matrix of lip wounds in fetal, neonatal and adult mice.

Wound healing in the fetus occurs rapidly, by a regenerative process and without an inflammatory response, resulting in complete restitution of normal tissue function. By contrast, in the adult, wounds heal with scar formation, which may impair function and inhibit further growth. The cellular mechanisms underlying these differing forms of wound healing are unknown but the extracellular matrix (ECM), through its effects on cell function, may play a key role. We have studied the ECM in upper lip wounds of adult, neonatal and fetal mice at days 14, 16 and 18 of gestation. The spatial and temporal distribution of collagen types I, III, IV, V and VI, fibronectin, tenascin, laminin, chondroitin and heparan sulphates were examined immunohistochemically. Results from the fetal groups were essentially similar whilst there were distinct differences between fetus, neonate and adult. Fibronectin was present at the surface of the wound in all groups at 1 h post-wounding. Tenascin was also present at the wound surface but the time at which it was first present differed between fetus (1 h), neonate (12 h) and adult (24 h). The time of first appearance paralleled the rate of wound healing which was most rapid in the fetus and slowest in the adult. Tenascin inhibits the cell adhesion effect of fibronectin and during development the appearance of tenascin correlates with the initiation of cell migration. During wound healing the appearance of tenascin preceded cell migration and the rapid closure of fetal wounds may be due to the early appearance of tenascin in the wound. Collagen types I, III, IV, V and VI were present in all three wound groups but the timing and pattern of collagen deposition differed, with restoration of the normal collagen pattern in the fetus and a scar pattern in the adult. This confirms that lack of scarring in fetal wounds is due to the organisation of collagen within the wound and not simply lack of collagen formation. The distribution of chondroitin sulphate differed between normal fetal and adult tissues and between fetal and adult wounds. Its presence in the fetal wound may alter collagen fibril formation. No inflammatory response was seen in the fetal wounds. The differences in the ECM of fetal and adult wounds suggests that it may be possible to alter the adult wound so that it heals by a fetal-like process without scar formation, loss of tissue function or restriction of growth.     

Ultrastructure of the corneal stroma: a comparative study.

Using a high intensity synchrotron x-ray source, we have recorded diffraction over a range of angles from the corneas of a wide variety of species. The results show that the interfibrillar Bragg spacing varies from 39 nm to 67 nm, the fibril diameter varies from 24 nm to 43 nm, but in the species studied intermolecular Bragg spacing is constant (1.58 +/- 0.03 nm). Using these data, a number of other structural parameters were calculated including the interfibrillar volume, V, and the surface-to-surface fibril separation, S. Large variations were found, particularly between aquatic and terrestrial animals. We found that the parameter which appears to be most constant throughout the species was the volume fraction, that is, the proportion of the tissue occupied by the hydrated fibrils. Ignoring the volume of the stroma occupied by cells, the tissue fibril volume fraction was (28 +/- 3)% for both aquatic and land animals. The observation of a constant volume fraction led us to propose a simple model in which collagen molecules and interfibrillar glycosaminoglycans occur in a fixed ratio in all the species--thus species with narrow fibrils have fewer interfibrillar glycosaminoglycans and the fibrils are thus more closely spaced, and vice versa. This model agrees with many of the experimental data on corneal composition and on the physical properties of the tissue reported in the literature.     

The intermolecular space of reconstituted collagen fibrils

The extent, geometry and heterogeneity of the intermolecular space of hydrated, purified and reconstituted steer skin collagen fibrils has been characterized. The extent of the space has been assessed experimentally by an X-ray diffraction method and a new physical chemical technique, and found to be 1.14 ml per gram collagen. A theoretical model relating the intermolecular space to X-ray diffraction parameters has been presented, and this suggests that the geometry of the intermolecular space arises from a near-hexagonal packing of the collagen molecules. On the basis of an assumed microfibrillar packing model and a geometric construction of the shape of a collagen molecule, the distribution of the space within reconstituted collagen fibrils has been characterized as follows: 0.13 ml of the intermolecular space/g collagen can be attributed to the helical groove of the collagen molecules per se and 1.01 ml/g is interstitial; 0.66 ml/g is present in the form of “pores” (hexagonally-closed packed spaces), whereas 0.48 ml/g is present in the form of “holes” (hexagonal volume defects); 0.73 ml/g of the intermolecular space is associated with a region of the collagen fibrils where holes are localized and 0.41 ml/g is attributable to the regions of the fibril in which pores only are present.     

Dinosaur peptides suggest mechanisms of protein survival

Eleven collagen peptide sequences recovered from chemical extracts of dinosaur bones were mapped onto molecular models of the vertebrate collagen fibril derived from extant taxa. The dinosaur peptides localized to fibril regions protected by the close packing of collagen molecules, and contained few acidic amino acids. Four peptides mapped to collagen regions crucial for cell-collagen interactions and tissue development. Dinosaur peptides were not represented in more exposed parts of the collagen fibril or regions mediating intermolecular cross-linking. Thus functionally significant regions of collagen fibrils that are physically shielded within the fibril may be preferentially preserved in fossils. These results show empirically that structure-function relationships at the molecular level could contribute to selective preservation in fossilized vertebrate remains across geological time, suggest a 'preservation motif', and bolster current concepts linking collagen structure to biological function. This non-random distribution supports the hypothesis that the peptides are produced by the extinct organisms and suggests a chemical mechanism for survival.