Increased Sequestration of Matrix Metalloproteinases in Ageing Human Bruch's Membrane: Implications for ECM Turnover

Abstract

Purpose.: The ageing of Bruch's membrane is associated with progressive reduction in the degradation of the capacity for ECM turnover mediated by the matrix metalloproteinase (MMP) system. In this study, the free and bound pools of all gelatinase species were quantified to aid in assessing the likelihood of reduced availability of pro-MMPs for activation in ageing Bruch's membrane.

Methods.: Bruch's membrane from macular locations (10 eyes; donor age range, 21–84 years) was mounted in Ussing chambers and eluted with phosphate-buffered saline to release the free pool of MMPs. Free and bound pools of MMPs were subjected to gelatin zymography, and individual gelatinase species were quantified by densitometric scans.

Results.: The zymograms displayed six gelatinase species: four corresponding to the pro- and active forms of MMP-2 and -9 and two high-molecular-weight polymeric forms designated HMW1 and -2, corresponding to approximate molecular masses of 195 and 391 kDa, respectively. The ageing of Bruch's membrane was associated with an exponential increase in the percentage of pro-MMPs bound to the membrane (pro-MMP-2: %age bound = 0.54 exp(0.04 × age), r = 0.87, P < 0.01; and pro-MMP-9: %age bound = 5.0 exp(0.03 × age), r = 0.8, P < 0.01). A similar exponential increase was seen in the percentage of bound HMW1 species (%bound = 11.7 exp(0.018 × age; P < 0.05). The HMW2 species was virtually all bound to the membrane, but some release was observed in the very elderly.

Conclusions.: The ageing of Bruch's membrane was associated with progressive sequestration of MMPs reducing the free concentration and potential for activation. These changes may underlie the reduction in degradation that leads to the age-related increase in the thickness of the membrane.

Bruch's membrane is an extracellular matrix (ECM) that houses passive transport pathways that are essential for the bidirectional movement of nutrients and waste products between the retinal photoreceptors and the choroidal blood supply. ¹ As in most ECMs, the structural and functional properties of Bruch's membrane are maintained by tightly coupled processes of synthesis and degradation, culminating in the continuous remodeling of the membrane. ^2–4  

The breakdown of the ECM is mediated by matrix metalloproteinases (MMPs), a family of Zn²⁺ containing Ca²⁺-dependent proteolytic enzymes. ^4–7 The secretion of these enzymes by the retinal pigment epithelium (RPE) and choroidal endothelium has been demonstrated, and their presence in Bruch's membrane has also been confirmed. ^8–11 These enzymes are released as inactive prozymogen forms and, on activation (by enzymatic cleavage of the propeptide), are capable of digesting most components of the ECM. The proteolytic activity of MMPs is restrained by the presence of tissue inhibitors of metalloproteinases (TIMPs). To date, four different TIMP species have been identified, and TIMP-2 and -3 have been localized to Bruch's membrane. TIMP-3 is unique because it is a normal bound component of the membrane. ^11,12 Thus, the degree of ECM breakdown is controlled by the temporal release of MMPs and their inhibition by TIMPs. 

Abnormalities in the biosynthetic and degradative coupling have the potential to alter the turnover status and therefore the structural and functional characteristics of Bruch's membrane. In Sorsby's fundus dystrophy, a mutation in the TIMP-3 gene is associated with an early-onset form of macular degeneration. In this mutation, the abnormal ECM turnover results in a considerably thickened Bruch's membrane (∼30 μm compared with 2–4 μm in normal) containing a large deposition of lipid-rich material. ^13,14 Although the mechanism leading to photoreceptor degeneration remains unknown, the thickened membrane is likely to provide a formidable resistance to nutritional support of the retina. Similarly, mutations in molecules that normally interact with the MMP system can lead to abnormal ECM turnover. ¹⁵ Thus, a mutation in the TIMP-3 binding epidermal fibulin-like extracellular matrix protein 1 (EFEMP1) is responsible for the hereditary macular degenerative disease, Malattia Leventinese. ¹⁶  

Even in normal ageing, the increased thickness, the deposition of normal and abnormal ECM material, increased cross-link formation, and accumulation of lipid-rich debris imply a disturbance in ECM turnover in Bruch's membrane. ^17–21 The gelatinase component (MMP-2 and -9) of the MMP family has been shown to be present in the membrane, with levels of the inactive forms showing an age-related increase. Active forms of MMP-2 and -9, although present in the peripheral regions, have been observed to be conspicuously absent or rarely apparent in macular regions. ⁸ This evidence, together with the increased levels of TIMP-3 in ageing Bruch's membrane, suggests that it is the ratio of MMPs to TIMPs that controls the degree of turnover of the ECM. In advanced ageing, associated with age-related macular degeneration (AMD), TIMP-3 levels are considerably elevated compared with age-matched controls, with the consensus being that elevated TIMP-3/MMP ratios may underlie the increased thickening of Bruch's membrane in this condition. ^2,12  

Little is known about the control and regulatory mechanisms that lead to the presence of active forms of MMP-2 and -9 within Bruch's membrane. Recent work in rat heart microvascular endothelial cells suggests that increased MMP-9 activation and release occur after mitochondrial oxidative stress mediated by calpain 1, with the primary change being an increase in intracellular calcium. ²² Similarly, angiotensin II, by binding to the AT1 receptor on RPE cells, activates G-protein-coupled phospholipase C and inositol-1,4,5-triphosphate, leading to increased intracellular calcium and a subsequent increase in the levels of MMP-2 mRNA and protein. ²³ Other mediators stimulating the release of MMP-2 and -9 include vascular endothelial growth factor (VEGF), fibronectin, and tumor necrosis factor (TNF)-α, but their role in normal physiological homeostasis remains unclear. ²⁴  

The extracellular activation of pro-MMP-2 is better understood and is mediated by another metalloproteinase, a transmembrane enzyme MMP-14, in combination with TIMP-2. ^25–27 The mechanism requires two molecules of MMP-14. The first MMP-14 molecule binds TIMP-2, and this enables the formation of the ternary complex with pro-MMP-2. A second MMP-14 molecule then cleaves the pro form to release active MMP-2. ^26,28 Thus, efficient activation requires the presence of MMPs and TIMPs in optimum concentrations near the basolateral surface of the RPE. For example, insufficient TIMP-2 results in diminished binding of pro-MMP2 to the cell surface for activation whereas excessive TIMP-2 leads to insufficient available MMP-14 for activation. ²⁸ Alcazar et al. ²⁵ have exploited this principle for reversing the hydroquinone-induced reduction in active MMP-2 by overexpression of MMP-14 and TIMP-2. 

Deficits in the concentration of free pro-MMPs can occur if these species are bound within the ECM. The observation of high-molecular-weight gelatinase species (apart from MMP-2 and -9) on zymograms of Bruch's extracts suggests the likelihood of MMP polymerization. ⁸ In addition, the age-related structural alterations and presence of debris-laden regions within Bruch's membrane may trap MMPs, effectively removing them from the activation process. Whether bound, trapped, or polymerized, the net result would be sequestration of MMPs. A decrease in the available free MMP pool is expected to diminish the potential for MMP activation with effects on the turnover of the ECM. The present investigation was therefore designed to assess the dynamics of the gelatinase system within Bruch's membrane and to quantify the free and bound pools of the various species as a function of age. 

Methods

The free pool of gelatinase activity was determined by mounting the Bruch's–choroid preparation in a Ussing-type open chamber and eluting with phosphate-buffered saline (PBS; Sigma-Aldrich, Poole, UK). The issuing fluid was collected at timed intervals over a period of 6 to 8 hours, and individual gelatinase components quantified by gelatin zymography. At the end of the elution period, the exposed tissue was removed, and the bound components extracted in SDS sample buffer followed by zymography for quantification. In some experiments, elution was performed in both directions across the preparation to assess the possibility that MMPs may simply be trapped because of the resistance barrier within the membrane rather than being bound to the matrix. ²⁹  

Tissue Preparation

Human donor eyes (age range, 21–84 years, and postmortem time, 24–48 hours) were obtained from the Bristol Eye Bank. The corneas had been removed at the Eye Bank for use in transplantation surgery, and the remaining globes were transported to the laboratory on saline-moistened pads in an icebox. The donor eyes were managed according to the guidelines in the Declaration of Helsinki for research involving human tissue. 

After a preliminary fundus examination with a dissecting microscope to ensure that the eyes were free of disease and gross handling artifacts, a circumferential incision was made 5 mm posterior to the scleral sulcus, and the remaining anterior segment, lens, and vitreous were discarded. The macular region was located and an 8-mm full-thickness trephined section, centered on the fovea, was removed and transferred to PBS. The neural retina was easily detached and discarded, exposing the monolayer of RPE cells. These were gently brushed away with the aid of a camelhair brush. Finally, under low-power magnification, the Bruch's–choroid complex was carefully removed from the underlying sclera by blunt dissection. This technique has been shown to preserve the structural integrity of Bruch's membrane. ³⁰ For mounting directly into the Ussing chamber, the preparation was floated onto an 8-μm nylon filter (Whatman; Sigma-Aldrich) and withdrawn from the PBS solution. Cassette mounting to allow bidirectional elution was performed without the aid of a filter support, and the procedure is described below. 

Elution of Gelatinase Species from Bruch's Membrane

The open-type Ussing chamber used for the elution studies is shown in schematic form in Figure 1A. It consists of two half-chamber plates (a1 and a2), the bottom (a1) having a central 4-mm diameter aperture providing feed to the preparation, an entry port for eluant, and an exit port leading to a pressure transducer. The top plate has a central aperture 6 mm in diameter, a 1-mm lip to facilitate withdrawal of fluid without touching the preparation, and two holes to accommodate the guiding pins protruding from the bottom plate. 

Figure 1.

 

Schematic of the open-type Ussing chambers used in the elution studies. (A) The isolated Bruch's–choroid preparation with Bruch's membrane facing up was clamped directly between the two Perspex half-chambers (a1 and a2) using the guiding pins (p) for alignment. The lower compartment was filled with PBS, taking care to remove all adherent air bubbles. Elution was performed at a hydrostatic pressure of 200 mm H₂O, and the eluted volume was collected at timed intervals. A 1-mm lip on the top half-chamber ensured that eluted buffers could be removed without touching, and hence damaging, the preparation. (B) The tissue preparation was mounted into a Perspex cassette (c) and then clamped between the two stainless-steel half chambers (b1 and b2). This configuration meant that, by turning over the cassette, elution studies could be conducted in both directions on the same preparation without tissue damage.

Tissue preparations on a nylon filter with Bruch's membrane facing upward were carefully centered over the aperture of the bottom plate. With the locating pins, the top plate was lowered and secured in place with three screws (not shown in Fig. 1) and tightened to a torque of 70 cN. The lower half-chamber was then carefully filled with eluant using a syringe, with occasional tilting of the chamber to remove any trapped air bubbles from the system. Eluant reservoir and transducer lines were connected, and the hydrostatic pressure was adjusted to 200 mm H₂O. At timed intervals of 1 hour, the fluid entering the upper compartment was removed using micropipettes with the tip end being placed on the 1-mm lip of the chamber. The amount of fluid removed was determined by weighing the retrieved sample. Elution experiments were performed for a period of 6 to 8 hours. At the end of the experiment, a 6-mm surgical trephine was inserted through the top half-chamber to cut out the exposed tissue. 

In experiments designed to assess the effect of reversing flow on further release of MMPs after the standard elution procedure was complete, we used a similar Ussing chamber but with the tissue held in a Perspex cassette (Fig. 1B). The cassette had a central 5-mm diameter aperture, and the tissue was loaded as previously described. ³¹ The bottom half-chamber was immersed in PBS, and all fluid lines were cleared of air bubbles. The tissue cassette was then gently maneuvered into the cassette slot in the chamber, followed by the top plate. The whole assembly, still immersed in PBS, was then clamped together with the aid of three screws. Taps to the entry and exit ports were then closed and the chamber removed from the PBS buffer reservoir. PBS remaining in the top half-chamber was removed, entry and exit ports were connected to the PBS buffer reservoir and digital manometer, respectively, and the hydrostatic pressure applied to the membrane was adjusted to 200 mm H₂O, allowing the elution to proceed. Eluant was removed on an hourly basis and weighed as described earlier. After 4 hours, the chamber assembly was disconnected from the entry and exit lines, immersed in PBS, and dismantled. The cassette was then reversed, and the full mounting and elution procedure was reinstigated, allowing flow in the opposite direction. At the end of the experiment, the tissue preparation was removed from the cassette, floated onto filter paper, and the 5-mm diameter central portion trephined and removed for analysis by zymography. 

The absence of divalent ions in the PBS eluant may disturb the equilibrium between bound and free MMP fractions. This possibility was assessed by comparing the MMP elution profiles between PBS and a basal salt mixture containing nutritional supplements (Dulbecco's modified Eagle's medium [DMEM]; Sigma-Aldrich). For these experiments, two adjacent 8-mm diameter sections were obtained from the peripheral fundus (from donors aged 21 and 82 years): one of the pair subjected to PBS elution and the other to DMEM elution. 

Gelatin Zymography for Quantification of MMPs

Sodium dodecyl sulfate (SDS) gel electrophoresis was performed (X Cell SureLock Mini-Cell System; Invitrogen, Paisley, UK). The 10% zymogram gels (1-mm thick) contained a 4% stacking layer and 0.1% gelatin in the separating layer (Novex Gels; Invitrogen). 

Aliquots of the eluant were diluted (1:2 vol/vol) in Laemmli nonreducing sample buffer (2.5% SDS, 4% sucrose, 0.25 M Tris-HCl, and 0.1% bromophenol blue). Tissue samples were placed in 20 μL PBS and 40 μL nonreducing buffer and vortexed for 5 minutes. Twenty microliters of the mixture was loaded onto the gel. Prestained protein molecular weight standards (Invitrogen) and 10% fetal calf serum (FCS; Sigma-Aldrich) was also run on each zymogram, if spare lanes were available. 

After electrophoresis (150 V, 1 hour), the gels were removed, rinsed in distilled water, and incubated for two half-hour periods in 2.5% Triton X-100, to remove SDS and renature the proteins. They were then transferred to reaction buffer (50 mM Tris-HCl, 10 mM CaCl₂, 75 mM NaCl, and 0.02% NaN₃ [pH 7.4]) and incubated at 37°C for 20 hours to allow proteolytic digestion of gelatin. The gels were rinsed again in distilled water and stained (SimplyBlue SafeStain; Invitrogen) containing Coomassie G-250 blue for a period of 3 hours. Destaining was performed with distilled water for 1.5 hours. 

MMP activity was observed as clear bands on a blue background. These gels were scanned at a resolution of 2400 dpi (3490 scanner; Epson, Nagano, Japan) and stored in JPEG format. The color images were uploaded into the software (Quantiscan, ver. 3.0; Biosoft, Cambridge, UK) in grayscale format, the colors were inverted so that the MMPs were visualized as dark bands against a whitish background, and the area under individual gelatinase bands was quantified. 

Eluted fractions and tissue samples from a given preparation were run on the same gel, and gel-to-gel variation was therefore avoided in the calculation of free-to-bound percentages for individual MMP species. Bound activity of a given species was expressed as a percentage of the total activity of the species (i.e., free and bound). 

Statistical Analyses

Standard linear and nonlinear regression analyses were performed with a commercial statistical package (Fig-Sys; Biosoft) that used the Marquardt-Levenberg algorithm. Significance levels of the Pearson's correlation coefficient were obtained by using the t-statistic with n − 2 degrees of freedom. 

Results

Since active forms of MMP-2 and -9 were more likely to be present in peripheral rather than macular regions, for illustrative purposes, a zymogram of combined peripheral Bruch's extract from donors aged 79 and 84 years is depicted in Figure 2. Gelanolytic activity was observed as clear bands against a dark background with a total of six species being discernible. Four of these bands correspond to the monomeric forms of pro- and active MMP-2 (72 and 66 kDa) and MMP-9 (97 and 85 kDa, respectively) species. Two additional high-molecular-weight forms (designated HMW1 and -2) were also present corresponding to approximate molecular masses of 195 and 391 kDa, respectively—values determined from log plots of migratory distance versus logMW. 

Figure 2.

 

Standard zymogram of gelatinase species present in extracts of human Bruch's–choroid preparations shows combined Bruch's–choroid extract from the peripheral regions of donors aged 79 and 84 years. Six gelanolytic bands were clearly identifiable: bands HMW1 and -2 (195 and 391 kDa), pro and active forms of MMP-9 (97 and 85 kDa), and pro and active forms of MMP-2 (72 and 66 kDa).

A typical elution profile for MMPs of Bruch's membrane from an 83-year-old donor is depicted in Figure 3. Most of the free pool of gelatinase activity was released within the first hour of elution, with progressive reduction in amount released thereafter. By the fifth hour of collection, release was reduced to barely detectable levels. Subsequent removal of tissue and analysis showed the considerable presence of bound forms of gelatinase activity remaining in the preparation. After adjustment for the amount of eluant and activity remaining in the tissue, the level of bound pro-MMP-2 and pro-MMP-9 was determined to be 20% and 48% of total activity for individual species, respectively. 

Figure 3.

 

Zymogram of eluted and membrane-bound MMP species in Bruch's membrane from the macular region of an 83-year-old donor showed rapid release of MMPs in the first hour with progressive reduction until only a trace of activity was observed in the fifth hour of collection. However, the tissue fraction extracted after the elution period showed substantial activity bound to the preparation.

Comparative elution profiles between PBS and DMEM as eluant showed the percentage of bound species to be virtually identical in both the young (21-year-old) and the aged (82-year-old) donor (Fig. 4). For the 21-year-old donor, the percentage of bound MMP species between PBS and DMEM, respectively, was HMW2 (73% and 80%); HMW1 (46% and 47%); MMP-9 (34% and 31%); and MMP-2 (4% and 3%). Bound levels were higher in the 82-year-old donor but similar between PBS and DMEM, respectively: HMW2 (99% and 93%); HMW1 (82% and 82%); MMP-9 (64% and 62%); and MMP-2 (30% and 25%). Thus, PBS as eluant was a suitable agent for determining the bound fractions of the various MMP species in this study. 

Figure 4.

 

Comparative elution of MMPs from peripheral Bruch's membrane with PBS and DMEM in donors aged 21 (A) and 82 (B) years. Elution was performed at a hydrostatic pressure of 200 mm H₂O with flow in the 21-year-old specimen averaging 59 ± 6 μL/hour and that in the 82-year-old 14 ± 5 μL/hour. Elution periods: 6 hours (A); 11 hours (B). MMP elution profiles were similar between PBS and DMEM, with most of the free pool being released in the first collection followed by a progressive reduction. The fraction of MMPs remaining bound was virtually identical for the two eluants used in a given donor preparation. *A leaky fraction, such that the MMP components were considerably diluted.

Bidirectional elution experiments were conducted to assess the possibility that the hydraulic resistance barrier in the inner collagenous zone of Bruch's may trap traversing MMP species during the elution process. As shown in Figure 5 in a 71-year-old donor, flow rates were low, and so eluant collections were made every 2 hours for a period of 8 hours. Overall release of MMP species was low in this preparation, and the normally low levels of MMP-2 were undetectable after the first collection. Low levels were partly due to a smaller exposed area of the tissue cassette (5-mm diameter). Reversing the direction of flow after the standard elution from the choroid to Bruch's membrane did not lead to loss of the bound fraction within the membrane. Of interest, a trace amount of the HMW2 species was released with levels higher than in the previous eluant collection (lane 4), suggesting that the high-molecular-weight component could not traverse the resistance barrier. The remaining mobile gelatinase species could cross the entirety of Bruch's membrane, and thus the amount observed in the tissue fraction must be considered bound MMPs. 

Figure 5.

 

Bidirectional elution of MMPs from the Bruch's membrane of a 71-year-old donor. Lanes 1 to 4: timed eluant collections over a 9-hour period with flow occurring from the choroidal side through Bruch's membrane. The cassette holding the preparation was then reversed, allowing flow from the membrane to the choroidal side. Lanes 5 and 6: collections taken at intervals of 1 hour. Lane 7: the gelatinase activity in the tissue sample. Reversing the direction of flow did not result in removal of the bound form of gelatinase activity from the Bruch's–choroid preparation. There was some release of the HMW2 fraction (arrow), suggesting that this species cannot traverse the resistance barrier.

Standard elution profiles across macular Bruch's membrane were constructed as follows for 10 donors in the age range 21 to 84 years. For a given gelatinase species, the areas under the densitometric scans were corrected for the eluant volume and summed with the tissue extract to provide a value of total enzyme activity. Next, the activity in each eluant fraction was expressed as a percentage of total activity for that gelatinase and plotted as the percentage released versus eluant volume (Fig. 6). The elution data were fitted by nonlinear regression to a hyperbolic function:  

 

where B _max is the maximum percentage released, V _{( μ L )} is the elution volume, and K is the volume necessary to obtain 50% of the releasable pool of MMPs. Pro-MMP-2 was released readily, but the maximum release of pro-MMP-9 showed considerable variation between donors. The eluant volume required for 50% release of the free pool for pro-MMP-2 and -9 (from a 6-mm diameter membrane preparation) was 71 ± 40 μL and 100 ± 60 μL, respectively (mean ± SD). The results show that of the total pro-MMP pool, the free releasable pool of pro-MMP-2 was 87% ± 7% and that of pro-MMP-9 was 66% ± 15% (mean ± SD). 

Figure 6.

 

Release of the free pool of MMP-2 and -9 from human Bruch's membrane and choroid. Full MMP elution profiles were obtained in 10 donors, age range, 21 to 84 years. The activity of individual MMP species in each elution fraction is expressed as a percentage of the total activity of that species in the tissue sample. The data for individual runs have been fitted to a hyperbolic function allowing an estimation of K, the volume required to obtain 50% of the releasable pool of MMPs. (A) Elution profile for MMP-2. In all donors examined, MMP-2 was released readily with a K of 71 ± 40 μL. (B) Elution profile for MMP-9. Considerable variation was observed in the amount of MMP-9 released across the different donors. Nonetheless, half the free pool was released within 100 ± 60 μL of eluant. Donors: 21 (○) and 28 (●) years old; 47 (□) and 51 (■) years old; 72 (▵) and 73 (▲) years old; 75 (▿) and 80 (▼) years old; and, 83 (◊) and 84 (◆) years old.

The percentage of pro-MMP-2 and -9 that remained bound after elution has been plotted as a function of age of donor and is presented in Figure 7. Data were fitted by nonlinear regression to an exponential function; for pro-MMP-2, %age bound = 0.54 exp(0.04 × age), r = 0.87, P < 0.01; and for pro-MMP-9, %age bound = 5.0 exp(0.03 × age), r = 0.8, and P < 0.01. Thus, ageing of Bruch's membrane was associated with an exponential increase in the bound fraction of these pro-MMPs. 

Figure 7.

 

Sequestration of MMP-2 and -9 within Bruch's membrane and choroid as a function of the ageing of the donor. The percentage of MMPs remaining bound was determined from the elution experiments of Figure 6 and plotted as a function of age. Ageing was associated with an exponential increase in the bound pool of MMPs (P < 0.01).

In most of the donor eyes, active levels of MMP-2 and -9 were barely discernible on the zymograms and therefore difficult to quantify in comparison to their pro-MMP counterparts. Of the 10 donor preparations examined, active MMP-2 was discernible in six preparations (the 83-year-old is shown in Fig. 3), and analysis showed that the amount remaining bound after elution was 21% ± 14(6) (mean ± SD(n)). This value was not significantly different from the bound fraction of pro-MMP-2 (11.7 ± 6.7(10)). In sharp contrast, active MMP-9 was present in only one donor preparation. 

High-molecular-weight gelatinase species were present in all the preparations examined. HMW1 (MW 195 kDa) binding to the ECM of Bruch's increased with the age of the donor (Fig. 8A). The relationship was fitted with the exponential function %bound = 11.7 exp(0.018 × age) (P < 0.05) but could also have been described by a linear function. The zymogram in Figure 5 suggested that the HMW2 species (MW 391 kDa) was unlikely to traverse the entirety of Bruch's membrane. Thus, to accurately estimate the bound fraction, elution studies should be performed in both directions for a given preparation. However, in the present analysis, elution was performed in the choroid-to-Bruch's direction only; thus, the observed release would be primarily from the inner collagenous zone of the membrane. With this in mind, most of the HMW2 species remained bound up to the sixth decade of life, but appeared to begin to release thereafter in the elderly (Fig. 8B). 

Figure 8.

 

Age-related variation in the bound fraction of high-molecular-weight gelatinase species in Bruch's membrane and choroid. (A) The amount of HMW1 (195 kDa) remaining bound to Bruch's membrane increased as a function of age reaching levels of approximately 60% in the eighth decade of life. (B) On the other hand, HMW2 (391 kDa) remained virtually all bound for most of the lifespan but appeared to undergo some release in later life. Data set: 10 donors, age range, 21–84 years.

Discussion

Age-related enlargement of Bruch's membrane is associated with increased deposition of both normal and abnormal matrix constituents. The latter includes oxidized, cross-linked, and denatured collagen that in the elderly can account for nearly 50% of total membrane collagen. ¹⁹ In advanced ageing associated with age-related macular degeneration (AMD), the extracellular matrix expansion is much greater. ³²  

The continued presence of abnormal constituents is thought to be due to a reduction in the degradation pathway for ECM turnover. There is also some evidence of increased synthetic activity, since connective tissue growth factor (CTGF), localized to Bruch's membrane, leads to induction of fibronectin and laminin in RPE cells. ³³  

Given the extended mass of the ECM, it is perhaps not surprising that the constituents of the degradation pathway—namely, pro-MMPs and TIMPs—show an age-related increase that is further exaggerated in donors with AMD. ^2,8,12 Increased levels could be due to the expansion of the matrix with age. For example, between the ages of 20 and 80 years, pro-MMP-2 and -9 levels are doubled, ⁸ but this doubling is also associated with a roughly twofold increase in matrix thickness. ^17,18 Thus, the effective concentration of MMPs may be relatively unaltered. However, increased prozymogen levels do not in themselves alter matrix degradation; cleavage of the propeptide, to yield active forms, is required for effective degradation. ^4,34,35 The observation that active levels of MMP-2 and -9 are rarely present in ageing macular compared with peripheral locations may underlie the reduced rates of matrix degradation. The situation is further complicated by the presence of AGE-modified matrix proteins that are known to inhibit MMP activity. ^33,36,37 Other studies have also shown reduced collagen susceptibility to proteolytic action due to the age-related increase in intermolecular cross-links in the fibrils. ^38,39  

Although the hydraulic conductivity and diffusional status of Bruch's membrane diminishes rapidly with age, ^30,31,40 the elution profiles of the present study show that the soluble pro-MMPs are fairly mobile and should therefore be available for interaction with their substrates and effectors of activation. However, the present investigation has demonstrated increased binding of MMPs to the ageing Bruch's membrane, and this sequestration is therefore expected to decrease the free concentration of MMPs. Since the activation of MMP-2, which occurs on the basal membrane of the RPE, requires membrane bound MMP-14, TIMP-2, and pro-MMP-2 in the ratio of 2:1:1, ^25–28 a reduction in the level of free pro-MMP-2 would compromise the activation process, thus leading to a reduction in activated MMP-2 for proteolysis. Reduced or absent levels of active MMP-2 and -9 have been documented in ageing Bruch's membrane from macular regions. ⁸  

The high-molecular-weight gelatinase species (HMW1 and -2) are thought to be mono- and heteropolymers of pro-MMP-2 and -9, and the increased age-related binding of the 195-kDa component demonstrates further sequestration of MMPs. The polymeric 391-kDa species (HMW2) was virtually all bound to the membrane in donors up to the sixth decade of life followed by a reduction in the very elderly. This reduction may be indicative of saturation of binding sites, such that further production remained unbound and could be easily eluted. In earlier work, we have shown the presence of a “resistance barrier” in the inner collagenous layer of Bruch's membrane in close apposition to the elastin layer of the membrane. ²⁹ This barrier does not hinder the movement of monomeric MMP-2 and -9 ³⁵ but appears to restrict the mobility of the 391-kDa species (Fig. 5). Since the elution profiles used to construct Figure 8B were obtained with flow from in the choroid-to-Bruch's direction only, the percentage of bound 391-kDa polymer would have to be a slight overestimate of the real situation. The question arises as to whether these high-molecular-weight species are actually bound or are simply trapped in the fibrillar meshwork of the membrane. Entrapment would not result in easy removal from the membrane with SDS for electrophoresis, and therefore we consider them to be bound. Further investigations are needed to determine the relative composition of these large gelatinase activities. 

The age-related sequestration of MMP-2 and -9, together with their polymeric forms, provides a mechanism whereby a progressive reduction in free pro-MMPs diminishes the potential for enzyme activation and the resultant impairment in degradative capacity leading to the observed age-related changes in Bruch's membrane. 

Footnotes

 Supported by the Guide Dogs for the Blind UK Association.

Footnotes

 Disclosure: A. Kumar, None; A. El-Osta, None; A.A. Hussain, None; J. Marshall, None