Remodelling of fibrillin-rich microfibrils by solar-simulated radiation: Impact of skin

Fibrillin-rich microfibrils (FRMs) constitute integral components of the dermal elastic fibre network with a distinctive ultrastructural 'beads-on-a-string' appearance that can be visualised using atomic force microscopy and characterised by measurement of their length and inter-bead periodicity. Their deposition within the dermis in photoprotected skin appears to be contingent on skin ethnicity, and influences the ultrastructure of papillary - but not reticular dermal FRMs. Truncation and depletion of FRMs at the dermal-epidermal junction of skin occurs early in photoageing in people with lightly pigmented skin; a process of accelerated skin ageing that arises due to chronic sun exposure. Accumulation of ultraviolet radiation (UVR)-induced damage, either by the action of enzymes, oxidation or direct photon absorption, results in FRM remodelling and changes to ultrastructure. In the current study, the direct effect of UVR exposure on FRM ultrastructure was assayed by isolating FRMs from the papillary and reticular dermis of photoprotected buttock skin of individuals of either black African or white Northern European ancestry and exposing them to solar-simulated radiation (SSR). Exposure to SSR resulted in significant reduction in inter-bead periodicity for reticular dermis-derived FRMs across both cohorts. In contrast, papillary dermal FRMs exhibited significantly increased inter-bead periodicity, with the magnitude of damage greater for African FRMs, as compared to Northern European FRMs. Our data suggest that FRMs of the dermis should be considered as two distinct populations that differentially accrue damage in response to SSR. Furthermore, papillary dermal FRMs derived from black African subjects show greater change following UVR challenge, when extracted from skin. Future studies should focus on understanding the consequences of UVR exposure in vivo , regardless of skin ethnicity, on the molecular composition of FRMs and how this UVR-induced remodelling may affect the role FRMs play in skin homeostasis. In this study we propose that incidental UVR exposure of photoprotected lightly-pigmented skin may induce sub-clinical damage to papillary dermal FRMs that induces remodelling of the FRMs. This damage may arise due to their close proximity to the DEJ and as a result of their low-level protection from melanin. Upon ex vivo SSR irradiation, these same FRMs are further remodelled and inter-bead periodicity significantly increases. In contrast, the papillary dermal FRM in highly-pigmented skin are well-protected by epidermal melanin and hence incidental UVR damage is potentially mitigated. As such, upon ex vivo SSR exposure these FRM are more susceptible to remodelling and inter-bead periodicity is increased to a greater magnitude. FRMs of the reticular dermis are largely protected from incidental UVR exposure in vivo due to their increased depth from the surface of the skin. Extraction of these FRM and subsequent SSR exposure causes a significant decrease to inter-bead periodicity as compared to unirradiated controls across both cohorts. We hypothesise that differences in spatial arrangement (papillary vs. reticular dermis), cellular origin (keratinocytes vs. fibroblasts), assembly and associated binding-proteins may drive FRM ultrastructural diversity and lead to their differential damage responses upon exposure to SSR. In this study we propose that incidental UVR exposure of photoprotected lightly-pigmented skin may induce sub-clinical damage to papillary dermal FRMs that induces remodelling of the FRMs. This damage may arise due to their close proximity to the DEJ and as a result of their low-level protection from melanin. Upon ex vivo SSR irradiation, these same FRMs are further remodelled and inter-bead periodicity significantly increases. In contrast, the papillary dermal FRM in highly-pigmented skin are well-protected by epidermal melanin and hence incidental UVR damage is potentially mitigated. As such, upon ex vivo SSR exposure these FRM are more susceptible to remodelling and inter-bead periodicity is increased to a greater magnitude. FRMs of the reticular dermis are largely protected from incidental UVR exposure in vivo due to their increased depth from the surface of the skin. Extraction of these FRM and subsequent SSR exposure causes a significant decrease to inter-bead periodicity as compared to unirradiated controls across both cohorts. We hypothesise that differences in spatial arrangement (papillary vs. reticular dermis), cellular origin (keratinocytes vs. fibroblasts), assembly and associated binding-proteins may drive FRM ultrastructural diversity and lead to their differential damage responses upon exposure to SSR. Cutaneous fibrillin-rich microfibrils (FRMs) should be considered as two distinct populations that differentially accrue damage in response to SSR. Furthermore, FRMs derived from black African skin show greater change following UVR challenge.


ABSTRACT
Fibrillin-rich microfibrils (FRMs) constitute integral components of the dermal elastic fibre network with a distinctive ultrastructural 'beads-on-a-string' appearance that can be visualised using atomic force microscopy and characterised by measurement of their length and interbead periodicity. Their deposition within the dermis in photoprotected skin appears to be contingent on skin ethnicity, and influences the ultrastructure of papillary -but not reticulardermal FRMs. Truncation and depletion of FRMs at the dermal-epidermal junction of skin occurs early in photoageing in people with lightly pigmented skin; a process of accelerated skin ageing that arises due to chronic sun exposure. Accumulation of ultraviolet radiation (UVR)-induced damage, either by the action of enzymes, oxidation or direct photon absorption, results in FRM remodelling and changes to ultrastructure. In the current study, the direct effect of UVR exposure on FRM ultrastructure was assayed by isolating FRMs from the papillary and reticular dermis of photoprotected buttock skin of individuals of either black African or white Northern European ancestry and exposing them to solar-simulated radiation (SSR). Exposure to SSR resulted in significant reduction in inter-bead periodicity for reticular dermis-derived FRMs across both cohorts. In contrast, papillary dermal FRMs exhibited significantly increased inter-bead periodicity, with the magnitude of damage greater for African FRMs, as compared to Northern European FRMs. Our data suggest that FRMs of the dermis should be considered as two distinct populations that differentially accrue damage in response to SSR. Furthermore, papillary dermal FRMs derived from black African subjects show greater change following UVR challenge, when extracted from skin. Future studies should focus on understanding the consequences of UVR exposure in vivo, regardless of skin ethnicity, on the molecular composition of FRMs and how this UVR-induced remodelling may affect the role FRMs play in skin homeostasis.

INTRODUCTION
Human skin is composed of three layers -a cell-rich epidermis; dermis; and deeper hypodermis. Within the dermis, a complex extracellular matrix (ECM) consisting of collagens, elastic fibres, proteoglycans and glycosaminoglycans work in concert to imbue skin with its biomechanical properties. The components of the elastic fibre network -elastin, fibrillin-rich microfibrils (FRMs) and microfibril-associated proteins -are the primary effectors of elasticity, enabling the skin to extend and recoil many times over the lifetime of an individual 1, 2 . During early development, the genesis of elastic fibres involves the deposition of tropoelastin (the soluble precursor of mature elastin) on a pre-formed template of FRMs 3 and, regardless of tissue type or species, these have a characteristic 'beads-on-a-string' ultrastructural appearance with an average inter-bead distance (or periodicity) of 56 nm 4 ; however, FRM and the composition of their accessory proteins is tissue-specific 5 .
The elastic fibre network forms a distinctive, highly ordered arrangement within the dermal ECM: at the dermal-epidermal junction (DEJ), superficial oxytalan fibres consist of cascades of discrete FRM bundles. These 'oxytalan' fibres coalesce with a fine network of elastin and fibrillin-rich 'elaunin' fibres within the papillary dermis, whilst in the reticular dermis, mature elastic fibres run in parallel to the DEJ 6, 7 . In young, healthy photoprotected skin we have previously identified that FRMs are differentially deposited in the papillary dermis of individuals from diverse ethnic backgrounds; with black African skin containing significantly more FRMs than both white Northern European and Far East Asian skin types 8 . Furthermore, it has recently been demonstrated that skin ethnicity also influences the ultrastructure of FRMs 9 . At birth, FRM ultrastructure is invariant; however, in adults from diverse ethnic backgrounds, there is a significant difference in ultrastructure for papillary dermal FRMs. In contrast,

Fibrillin microfibril isolation
Assemblies of fibrillin-rich microfibrils were isolated from the papillary dermis of adult buttock skin by cryosectioning bisected 6 mm skin biopsies en face to a depth of 400 µm. Next, 100 µm was cryosectioned from the skin biopsy and discarded; FRM were then isolated from the remaining reticular dermal fraction of the skin biopsy (

Atomic force microscopy and data processing
The ultrastructure of extracted FRMs were characterised by atomic force microscopy (AFM).
Using the Multimode 8 AFM (Bruker AFM Probes, Camarillo, California USA) fitted with ScanAsyst-Air cantilevers, randomly selected 10 x 10 µm locations were scanned at a rate of 1.97 Hz. The morphologic metrics assessed were the number of beads per FRM and inter-bead periodicity. Periodicity was determined by measuring the distance between individual beads (n = 1000) using WSxM scanning probe microscopy software and by routines written in Microsoft Visual Basic 6.0. Inter-bead periodicity is a widely used, reliable and quantitative marker for analysis of FRM ultrastructure 5,9,[19][20][21] .

Statistical analyses
Regression whether differences in periodicity existed between the two groups, controlling for participant age and gender. As the data were hierarchical -inter-bead periodicity is 'clustered' within FRM which, themselves are clustered with participants -we initially fitted a multi-level mixedeffects linear regression model, which takes account of both the potential correlation between periodicity measurements within the same FRM and the potential correlation of FRMs within the same participant (although the latter is likely to be smaller). As the sample of periodicity measurements has some positive skew and is highly kurtotic, a non-parametric bootstrapped standard error was also derived, using 200 replications and a random initial-value 'seed'.
Having run these models, it was apparent that there was some mis-specification, as there were inconsistencies in the estimated standard errors. We therefore decided to take account of the likely correlation between any randomly chosen periodicity measurement and its preceding one in the 'chain' (known as auto-regression). We therefore fitted separate population-averaged linear regression models, with auto-regressive order-1 correlations (AR1) and robust standard errors (as bootstrapping is not available for this type of model), to the original data, to logarithmic-transformed data and to square root-transformed data.

Buttock skin biopsies were dissected into papillary and reticular dermal components and FRMs
were extracted prior to irradiation with a single dose of SSR (15.4 J/cm 2 ), to allow characterisation of ultrastructural parameters (bead number per FRM and inter-bead periodicity; Figure 1). For both papillary and reticular dermal FRMs, SSR exposure did not induce significant changes to the number of beads per microfibril in either cohort. However, reticular dermal FRMs consistently had a greater number of beads per FRM as compared to those extracted from the papillary dermis (Table 1).
Next, the impact of SSR exposure on inter-bead periodicity was assessed for both papillary and reticular dermal FRMs. For FRMs extracted from the papillary dermis, SSR caused a significant increase in inter-bead periodicity compared to unirradiated controls for both cohorts (  Figure   2).
In contrast, for FRMs extracted from the reticular dermis, SSR exposure caused a significant reduction in inter-bead periodicity as compared to unirradiated controls across both cohorts. African response similar to that observed for the Northern European cohort (Figure 2).
Taken together, these findings show that experimental SSR does not cause significant changes to FRM bead number; however, it does induce changes to inter-bead periodicity. Furthermore, the direction of this difference appears to be dependent on the dermal compartment from which the FRMs are extracted (P < 0.001) but not from the ethnic group from which they are derived (P = 0.205). reporting of a single value may underestimate the differences between samples 9 . Therefore, this study employs microfibril-by-microfibril analysis of each FRM to provide detailed observations of the distribution of inter-bead periodicities.
Our experimental protocol was designed to assay the direct effect of SSR-exposure on FRM ultrastructure. This was achieved by irradiating the FRMs following their extraction from the skin and hence the usual events that would occur in a physiological setting -such as infiltration of immune cells into the skin 12 , photo-oxidation 13 or induction and release of MMPs 14 -were absent. The finding that bead number per FRM was unaffected by SSR exposure suggests that when fragmentation and truncation of FRM in photoexposed skin occurs, it is likely via a cellmediated process driven by the UVR-induced expression and/or activation of ECM proteases, such as MMPs [23][24][25] , and/or the liberation of reactive oxygen species (ROS) 26  Fibrillin-rich microfibrils are particularly susceptible to UVR damage due to their photochemical composition 30 ; superficial microfibril assemblies in the papillary dermis ('oxytalan' fibres) are devoid of elastin and susceptible, by means of their amino acid content, to be particularly susceptible to UVR 19 , whereas FRMs of the reticular dermis are largely protected, not only by their increased distance from the DEJ, but also by their combination with elastin (to form 'elaunin' fibres) 6,31 . Thus, differences in spatial arrangement, cellular origin, assembly and associated binding-proteins may drive FRM ultrastructural diversity and lead to differential responses upon exposure to SSR.
The ability of SSR to induce a more profound effect on papillary dermal FRM derived from African skin than those of Northern European skin was somewhat surprising. We hypothesise that these differences arise because FRM assemblies extracted from the papillary dermis of lighter pigmented subjects may have already accumulated damage in vivo. Papillary dermal FRMs appear particularly susceptible to photoexposure, due in part to their close proximity to the epidermis 10 . Excessive sun exposure, mainly in childhood, and various high-risk activities,     in vivo due to their increased depth from the surface of the skin. Extraction of these FRM and subsequent SSR exposure causes a significant decrease to inter-bead periodicity as compared to unirradiated controls across both cohorts. We hypothesise that differences in spatial arrangement (papillary vs. reticular dermis), cellular origin (keratinocytes vs. fibroblasts), assembly and associated binding-proteins may drive FRM ultrastructural diversity and lead to their differential damage responses upon exposure to SSR. . Solar-simulated radiation induces differential remodelling of FRMs in the dermis of ethnically diverse skin For FRMs extracted from the papillary dermis, SSR irradiation caused a significant increase to inter-bead periodicity compared to unirradiated controls for both African and European cohorts. However, the cumulative distributions revealed differences in the magnitude of each cohort's response to SSR, with the African response greater than that observed for the European cohort (a). In contrast, for FRMs extracted from the reticular dermis, SSR exposure caused a significant decrease to inter-bead periodicity as compared to unirradiated controls across both cohorts (b).
139x132mm (300 x 300 DPI) Figure 3. Schematic image summarising our proposed hypothesis of the differential remodelling of FRM that occurs in response to SSR exposure In this study we propose that incidental UVR exposure of photoprotected lightly-pigmented skin may induce sub-clinical damage to papillary dermal FRMs that induces remodelling of the FRMs. This damage may arise due to their close proximity to the DEJ and as a result of their low-level protection from melanin. Upon ex vivo SSR irradiation, these same FRMs are further remodelled and inter-bead periodicity significantly increases. In contrast, the papillary dermal FRM in highly-pigmented skin are well-protected by epidermal melanin and hence incidental UVR damage is potentially mitigated. As such, upon ex vivo SSR exposure these FRM are more susceptible to remodelling and inter-bead periodicity is increased to a greater magnitude. FRMs of the reticular dermis are largely protected from incidental UVR exposure in vivo due to their increased depth from the surface of the skin. Extraction of these FRM and subsequent SSR exposure causes a significant decrease to inter-bead periodicity as compared to unirradiated controls across both cohorts. We hypothesise that differences in spatial arrangement (papillary vs. reticular dermis), cellular origin (keratinocytes vs. fibroblasts), assembly and associated binding-proteins may drive FRM ultrastructural diversity and lead to their differential damage responses upon exposure to SSR.