How does skin adapt to repetitive mechanical stress to become load tolerant?

Abstract

Skin breakdown from mechanical stress application is a difficult health care problem for lower-limb amputees using prosthetic limbs. Post-operative treatments to encourage skin adaptation do exist, but are largely unsuccessful. Potentially, by understanding skin adaptation on a molecular level, appropriate biomolecules can be identified and then delivered to skin to encourage adaptation in at-risk patients. Based from a critical review of the literature, it is expected that adaptation occurs by forming new collagen fibrils with larger diameters as opposed to increasing diameters of existing fibrils. Small collagen fibril breakdown by stress activated metalloproteinases is expected to be followed by increased expressions of decorin, biglycan, fibromodulin, lumican, thrombospondin-2, and collagens I and III, facilitating formation of new fibrils with larger diameters. After remodeling, total collagen fibril cross-sectional area is expected to return to baseline values since increased collagen content would increase mass and be redundant towards the purpose of adaptation.

Introduction

Soft tissue breakdown resulting from localized mechanical stress is a serious healthcare problem affecting the elderly, the immobile, and the disabled. It is generally accepted that for static stresses, such as those experienced by the immobile, this breakdown initially occurs in the muscle layer [1], [2], [3]. However, where the tissue is subjected to abnormally high dynamic compressive stresses and shear stresses during ambulation, as in prosthesis users, the skin is the first to show signs of breakdown [4], [5], [6], [7], [8]. In a study of 120 patients with new transmetatarsal amputations (9), 27% developed skin breakdown and 48% of these occurred within the first three months after surgery. On VA patients who received permanent lower-extremity prostheses in the Bay Area (7), over a three-year period 41% of below-knee and 22% of above-knee amputees demonstrated skin irritation or breakdown. These studies show that skin breakdown affects a significant proportion of the amputee population. With over 110,000 amputations performed each year in the US [10], [11], breakdown is a serious and costly health care problem.

Although post-operative clinical treatments to encourage skin adaptation [12], [13], [14] do exist, for amputees the predominant methods for prevention focus on reducing prolonged stresses experienced by the tissue. The prevalence and incidence of breakdown clearly show that these methods have largely been unsuccessful. One method which has not been well-approached from a cellular or molecular perspective is to enhance skin load tolerance before potentially detrimental stresses are applied. If the bioprocesses of skin adaptation were understood on a cellular and molecular level then potentially appropriate biomolecules could be delivered to encourage skin adaptation in at-risk patients. The purpose of this paper is to present a hypothesis for how skin adapts with the intent to identify candidate biomolecules for therapeutic delivery.

Collagen is the major structural component of skin and it is suggested that adaptations occur in the collagen matrix in response to load application. Collagen fibril diameters increase in response to repetitive mechanical pressure [15], [16] and shear (17), allowing the tissue structure to withstand greater stresses [18], [19]. Even if the total fibril cross-sectional area remains the same, skin with large diameter fibrils has a greater ultimate strength than that with small diameter fibrils because of the reduced tensile stress build-up within the cross-sections (19). Mechanical strain has also been shown to increase type I collagen synthesis in the pulmonary system [20], [21], [22], [23].

It is important to note that in these adaptation studies [15], [16], [17] the tissue was natural and not scar tissue. This characteristic is crucial to recognize here because it points to the fact that skin adaptation (natural tissue) is distinctly different from wound healing (scar forming). More importantly, wound healing involves mainly the blood (e.g., macrophage cells) whereas skin adaptation involves mainly extracellular matrix proteins (as discussed below). In addition, soft tissue scar has inferior biomechanical properties and impaired function [24], [25] compared with adapted skin thus is highly undesirable in disability patients, particularly prosthesis users. In hypertropic scars collagen fibrils are narrower, more widely spaced, and more disorganized in orientation than in normal skin, forming whorls or nodules rather than parallel fibers [26], [27].

In order to establish which biomolecules are instrumental for adaptation, the pathway by which adaptation occurs as a result of stress application must be determined. There are three possible pathways (Fig. 1). Mechanical stress can stimulate: the enlargement of existing collagen fibrils (Path A); the formation of new collagen fibrils with the simultaneous degradation of old collagen (Path B); or a combination of the two (Path C). A review of the literature highlights several biomolecules that may play important roles, both in a constructive and degradative sense, at different stages of adaptation and provides insight into which pathway occurs.

Both proteoglycans and their associated glycosaminoglycans (GAGs) are relevant towards the constructive component of skin adaptation, i.e., collagen fibrillogenesis. The GAGs present in skin include dermatan sulphate, keratin sulphate, hyaluronan, and chondroitin sulphate, though as discussed below only the first two facilitate collagen fibril enlargement. In skin, the principle dermatan sulfate proteoglycans are decorin and biglycan (28), while the principle keratin sulphate proteoglycans are fibromodulin and lumican (29). Thrombospondin-2, also linked with collagen fibrillogenesis in skin, is also present though classifies as a member of a separate protein family.

Different GAGs are associated with different collagen architectures in soft tissue. It is clear from a number of comparative medicine studies (evolutionary adaptation) that tissues with elevated dermatan sulphate levels have larger collagen fibrils [30], [31], [32], [33], [34], [35], [36]. High hyaluronan levels are associated with tissues with small collagen fibrils while tissues with fibrils of intermediate size have elevated chondroitin sulphate levels (34).

Not only are different GAGs associated with different collagen architectures, they are also associated with different mechanical loading conditions. Tissues which are subjected to high tensile stresses have an increased percentage of dermatan sulphate proteoglycans [34], [37]. On the convex side of the rabbit flexor digitorum profundus tendon, which experiences tensile stresses, the fibril diameters were approximately 200 nm, and the majority of the GAG content was dermatan sulphate. On the concave part, which experiences compressive stresses, the collagen fibrils were smaller with a diameter of approximately 150 nm, and the majority of the GAGs were chondroitin sulphate [35], [38], [39].

Short-term biomechanical studies support the comparative medicine findings in that they demonstrate increased dermatan sulphate contents and larger collagen fibrils under enhanced tensile loading. When the flexor digitorum profundus tendon of a rabbit was transferred so that it experienced only tensile stresses, regions containing small diameter collagen fibrils were gradually replaced by thicker collagen fibrils, and the dermatan sulphate content elevated. These adaptive changes were reversed if the tendon was returned to its original position [35], [40], [41]. The mechanisms by which these adaptive processes occur are still unknown, though it is suggested that changes in GAG compositions precede changes in collagen architecture (34).

Though the role of some GAGs in adaptation is relatively well investigated, it is less clear exactly which proteoglycans are responsible for which architectural adaptations. Proteoglycans decorin, biglycan, fibromodulin, lumican, and thrombospondin-2 have been investigated. Decorin-null mice have fragile skin with loosely packed, irregular cross-sections and a collagen fibril morphology with greater variability in size and shape compared to the wild type (42). There is a significant reduction of proteoglycan granules and filaments around the collagen fibrils. Visser et al. [43], [44] and Korver et al. (45) observed increases in the production of decorin in dynamically loaded compared with statically loaded tissue. Biglycan, the other dermatan sulphate proteoglycan, is suggested as being another collagen-organizing macromolecule, although to a lesser extent than decorin (46). In a study by Svensson et al. (29), fibromodulin-null mice had disorganized collagen fiber bundles with irregular cross-sections and thinner fibers than the wild type. In addition there was a 3–5-fold increase in lumican concentration. In genetic studies, thrombospondin-2-null mice had fragile skin with a disorganized collagen network. Kyriakides et al. (47) suggested that like decorin, thrombospondin-2 might be another fibril-associated protein that regulates fibril diameter and fibrillogenesis.

In addition to these matrix proteins, collagen V, which associates with type I collagen, is proposed to affect the diameter of collagen fibrils. It acts as an inhibitor rather than a non-inhibitor like decorin, biglycan, fibromodulin, lumican, and thrombospondin-2. Disruption of type V collagen results in thicker and disorganized fibrils (48), and an increased presence of type V collagen results in significantly thinner fibrils (49). It has been suggested that type V collagen may either inhibit the lateral growth of type I fibrils due to the longer helix [50], [51], or form intermediates which act as additional nucleation sites for the assembly of type I collagen, resulting in more fibrils of reduced diameters.

Simultaneous degradation of collagen must occur if Path B or C takes place (Fig. 1). There are enzymes that are responsible for the degradation of different extracellular matrix components in other bioprocesses such as inflammation, cell migration, tumor invasion, and metastasis. Participating enzymes include three different types of proteases that can be subdivided by their active centers, namely the serine-, cystine-, and metalloproteinases. The metalloproteinases are responsible for collagen and proteoglycan degradation (52). In particular, interstitial collagenase catalyzes the initial step in the degradation of both collagen type I and collagen type III (53). Variations in the rate of collagen degradation depend on, among other features, the application of stress or strain to the fibrils [54], [55], the presence of proteoglycans (56), the primary structure of the collagen polypeptide subunits, on their state of aggregation, level of cross-linking [57], [58], and the presence of inhibitors [59], [60] such as collagen V. Therefore, if the collagen fibrils are degraded and replaced with new fibrils during adaptation (Fig. 1, Path B or C), the metalloproteinases would be expected to play a major role, and their concentrations would be expected to increase immediately after stress application.