Matrix metalloproteinase-9-null mice are resistant to TGF-β-induced anterior subcapsular cataract formation

Abstract

Epithelial-mesenchymal transition (EMT) is associated with fibrotic diseases in the lens, such as anterior subcapsular cataract (ASC) formation. Often mediated by transforming growth factor (TGF)-β, EMT in the lens involves the transformation of lens epithelial cells into a multilayering of myofibroblasts, which manifest as plaques beneath the lens capsule. TGF-β-induced EMT and ASC have been associated with the up-regulation of two matrix metalloproteinases (MMPs): MMP-2 and MMP-9. The current study used MMP-2 and MMP-9 knockout (KO) mice to further determine their unique roles in TGF-β-induced ASC formation. Adenoviral injection of active TGF-β1 into the anterior chamber of all wild-type and MMP-2 KO mice led to the formation of distinct ASC plaques that were positive for α-smooth muscle actin, a marker of EMT. In contrast, only a small proportion of the MMP-9 KO eyes injected with adenovirus-expressing TGF-β1 exhibited ASC plaques. Isolated lens epithelial explants from wild-type and MMP-2 KO mice that were treated with TGF-β exhibited features indicative of EMT, whereas those from MMP-9 KO mice did not acquire a mesenchymal phenotype. MMP-9 KO mice were further bred onto a TGF-β1 transgenic mouse line that exhibits severe ASC formation, but shows a resistance to ASC formation in the absence of MMP-9. These findings suggest that MMP-9 expression is more critical than MMP-2 in mediating TGF-β-induced ASC formation.

Figure 1

Effects of adenoviral gene transfer of active TGF-β1 after 2 to 3 weeks. Histologic sections from mice injected with AdTGF-β1 or AdDL were stained with H&E (A–D) or subjected to immunostaining for α-SMA (E–H, green). Focal multilayering of LECs occurs in both wild-type (A; n = 7) and MMP-2 KO (B; n = 6) mice treated with AdTGF-β and in both cases is associated with induction of α-SMA (E and F). MMP-9 KO mice treated with AdTGF-β maintained a monolayer of LECs (C; n = 10) that did not express α-SMA (G), resembling mouse eyes injected with AdDL control vector (D and H). Note the presence of an intact lens capsule in all lenses after adenoviral injection. All sections were mounted in medium with DAPI to co-localize the nuclei (blue). The small circles in A and D are artifacts. c, cornea; i, iris; le, lens epithelium. Scale bar = 100 μm (A–H).

Figure 2

Effects of TGF-β on E-cadherin localization and α-SMA levels in wild-type mouse LEC explants. Confluent LEC explants when left untreated (A–C; n = 9) maintain a tightly packed, cuboidal, cobblestone-like appearance in a monolayer with E-cadherin expressed at cellular junctions (A) and negligible α-SMA expression (B, green). Mouse LECs treated with 500 pg/mL of TGF-β for 48 hours (D–F; n = 10) exhibit a distinct loss of E-cadherin (D) and marked expression of α-SMA (E). G: TGF-β leads to a significant induction in α-SMA compared with untreated wild-type lens explants. All explants were mounted in medium with DAPI to co-localize the nuclei (C and F, blue). Scale bar = 100 μm (A–F). Data are expressed as means ± SEM. ^∗P < 0.05.

Figure 3

Effects of TGF-β treatment on E-cadherin localization and α-SMA levels in MMP-2 KO mouse lens epithelial explants. MMP-2 KO explanted LECs, when left untreated (A–C; n = 10), maintained a cuboidal arrangement with E-cadherin staining present at intact cellular junctions (A) and negligible α-SMA expression (B, green) similar to wild-type LECs. In the absence of MMP-2, 500 pg/mL of TGF-β treatment for 48 hours (D–F; n = 7) induced a phenotypic transformation with a distinct loss of E-cadherin (D) and marked expression of α-SMA (E), similar to treated wild-type LECs. G: TGF-β leads to a significant induction in α-SMA compared with untreated MMP-2 KO lens explants. All explants were mounted in medium with DAPI to co-localize the nuclei (C and F, blue). Scale bar = 100 μm (A–F). Data are expressed as means ± SEM. ^∗P < 0.05.

Figure 4

Effects of TGF-β treatment on E-cadherin localization and α-SMA levels in MMP-9 KO mouse lens epithelial explants. MMP-9 KO LEC explants when left untreated (A–C; n = 9) exhibited irregular E-cadherin staining (A) with negligible α-SMA expression (B, green). LECs treated with 500 pg/mL of TGF-β for 48 hours in the absence of MMP-9 (D–F; n = 5) maintained a similar E-cadherin profile as the untreated controls (D) and, in contrast to TGF-β–treated wild-type and MMP-2 KO explants, did not express α-SMA (E). G: TGF-β did not significantly induce α-SMA expression when compared with untreated MMP-9 KO lens explants. All explants were mounted in medium with DAPI to co-localize the nuclei (C and F, blue). Scale bar = 100 μm (A–F). Data are expressed as means ± SEM.

Figure 5

E-cadherin protein levels in LEC explant lysates and conditioned media. A: Representative Western blot reveals E-cadherin protein (120 kDa) in MMP-9 KO LEC explant lysates compared with their wild-type littermates. B: Densitometric analysis of Western blot in LEC explant lysates indicate no significant difference in total E-cadherin protein between MMP-9 KO (n = 8) and wild-type (n = 5) LEC explants. Data were normalized with β-tubulin control and are expressed as means ± SEM. C: ELISA measuring E-cadherin in conditioned media of untreated and 500 pg/mL of TGF-β–treated (for 48 hours) wild-type and MMP-9 KO LEC explants. In the media, TGF-β treatment leads to an increase in E-cadherin protein in wild-type explants when compared with untreated wild-type controls. Levels of E-cadherin in the media of MMP-9 KO explants after TGF-β treatment are significantly different from wild-type TGF-β–treated levels (P = 0.019) and comparable to all untreated explant levels. Data are expressed as means ± SEM (n = 3 to 6). ^∗P < 0.05.

Figure 6

Effects of TGF-β treatment on β-catenin localization in MMP-9 KO mouse lens epithelial explants. In untreated wild-type (A and B) and MMP-9 KO (C and D) explants, β-catenin is predominantly localized at the cellular margins in a cobblestone arrangement. With 500 pg/mL of TGF-β treatment for 48 hours, wild-type LECs exhibit primarily cytosolic expression of β-catenin (E and F), in contrast to TGF-β–treated MMP-9 KO explants, which resembled controls with marginal staining of β-catenin at cell borders (G and H). All explants were mounted in medium with DAPI to co-localize the nuclei (blue). Scale bar = 100 μm.

Figure 7

Lens epithelium of TGF-β1 transgenic and TGF-β1/MMP-9 KO mice at 1 to 3 months of age. Histologic sections were stained with H&E or subjected to immunostaining for α-SMA (I–L, green). Unlike control lenses (A and E; n = 12), TGF-β1 transgenics at this time point exhibit reorganization of cells and plaque formation exuding into the anterior chamber (B and F; n = 10), with immunoreactivity of α-SMA confirming the presence of subcapsular plaques (J compared with control, I). In 75% of TGF-β1/MMP-9 KO mice, a protection from plaque formation was observed (C and G) with no immunoreactivity for α-SMA (K; n = 14), whereas 25% of TGF-β1/MMP-9 KO mice did develop subcapsular plaques (D and H) with α-SMA reactivity (L; n = 5). All sections were mounted in medium with DAPI to co-localize the nuclei (blue). c, cornea; le, lens epithelium. Scale bar = 100 μm. Original magnification: ×20 (A–D); ×40 (E–H).