Experimental Animals and Procedures

Wild type (WT) mice were purchased from The Jackson Laboratory. We cross-bred Timp3^−/− (19) and Mmp2^−/− (20) mice to generate Timp3^−/−/Mmp2^−/− mice. Timp3^−/−/Mmp2^−/− mice are healthy and fertile. All mice are in a C57BL/6 background. Alzet micro-osmotic pumps (Model 1002, Durect Corp.) were implanted subcutaneously in male mice of the indicated genotypes to deliver 1.5 mg/kg/day Ang II (21–23) or saline (control). Broad spectrum MMP inhibitor (MMPi; PD166793) was administered orally by daily gavage (30 mg/kg/day) as before (24, 25). All experiments were conducted in accordance with the guidelines of the University of Alberta Animal Care and Use Committee and the Canadian Council of Animal Care.

Bone Marrow Reconstitution

Bone marrow (BM) reconstitution was performed to generate chimeric mice as described previously (26). Donor WT and Timp3^−/−/Mmp2^−/− mice received 100 μg of purified anti-CD8 antibody (clone 2.43) intraperitoneally on days −2 and −1 before BM harvest to deplete CD8 T-cells. BM was harvested from the femur, tibia, and humerus in EasySep medium (PBS, 2% FCS, 2 mm EDTA) and passed through 70-μm nylon cell strainers (Fisherbrand). Recipient WT and Timp3^−/−/Mmp2^−/− mice were sublethally irradiated (1,000 grays) and received 10 × 10⁶ BM cells of the opposite genotype via tail vein injection. WT BM was reconstituted in Timp3^−/−/Mmp2^−/− mice (Timp3^−/−/Mmp2^−/−-chimera mice), and Timp3^−/−/Mmp2^−/− BM was reconstituted in WT mice (WT-chimera mice). Mice were provided with antibiotic water (40 mg of neomycin and 15 mg of polymyxin/1 liter) for 4 weeks postinjection, and 7 weeks were allowed for reconstitution before Alzet pumps were implanted for Ang II infusion.

Ultrasonic Imaging of Abdominal Aorta

Ultrasonic images of the abdominal aorta were obtained in mice anesthetized with isoflurane using a Vevo 770 high resolution imaging system equipped with a real time microvisualization scan head (RMV 704, Visual Sonics, Toronto, Canada) (27). The aortic diameters were measured by M-mode at the suprarenal region where aneurysm was detected (in Timp3^−/− and Timp3^−/−/Mmp2^−/− mice) (28). The maximum aortic lumen diameter (aortic systolic diameter corresponding to cardiac systole) and the minimum aortic lumen diameter (aortic diastolic diameter corresponding to cardiac diastole) monitored by simultaneous ECG recordings were measured and used to calculate the aortic expansion index ((Systolic aortic diameter − Diastolic aortic diameter)/Systolic diameter × 100) (29). Three independent measurements were made for each mouse.

Histological Analyses and Scanning Electron Microscopy

Abdominal aortas were fixed in 10% formalin, paraffin-embedded, and used for Gomori trichrome (GT) and Verhoeff-Van Gieson (VVG) staining. Tissue processing and scanning electron microscopy were done as before (25). Neutrophil and macrophage staining was performed as described previously (30) using anti-mouse neutrophil antibody with Cy3-labeled anti-rat secondary antibody and F4/80 macrophage antibody with Cy3-labeled anti-rat secondary antibody that was pseudocolored to appear green for visual differentiation from neutrophil staining. We used nuclear staining by DAPI to confirm the presence of infiltrating cells in positively stained regions. Superimposed staining for neutrophils or macrophages with DAPI is presented.

Protein Extraction, Western Blot Analysis, Gelatin Zymography, and mRNA Analysis

Flash frozen abdominal aorta was used for protein or RNA extraction. Western blotting was performed to detect collagen type I (COL1A1) (Santa Cruz Biotechnology) and α-elastin (Abcam Inc.), and in vitro gelatin zymography was carried out as before (30). A Coomassie Blue-stained SDS-PAGE gel (after proteins were transferred to PVDF membrane) was used as a loading control. In situ gelatin zymography was performed on 14-μm cryosections as described (31). Total collagenase and gelatinase activities were measured using Enzchek collagenase (catalog number D-12060) and elastase (catalog number E-12056) activity assay kits (from Millipore) as before (30). Briefly, 50 μg of protein extract was incubated with fluorescently labeled collagen or elastin substrate. The generated fluorescence signal was recorded for 4 h (excitation, 485 nm; emission, 515 nm), and the slopes of the curves were used to calculate the collagenase or elastase activities. mRNA expression was measured using TaqMan real time PCR (Table 1) (30, 32).

Statistical Analyses

The statistical analyses were performed using SPSS software (Version 10.1; Chicago, IL). We performed normality (Shapiro-Wilk) and homogeneity of variance tests (Levene) to confirm normal distribution of all data. To compare data sets with two factors (genotype and treatment), we performed two-way analysis of variance (Figs. 1–6). In Figs. 5–8, to compare three levels of the genotype, two-way analysis of variance was followed by Bonferroni post hoc testing if the results attained statistical significance (p < 0.05). In Fig. 7B, the effects of MMPi in each genotype were tested using an unpaired t test. Survival data (Fig. 4B) were compared using Kaplan-Meier survival curves followed by the log rank test adjusted with the Bonferroni method. Averaged values represent mean ± S.E. Statistical significance was recognized at p < 0.05.

Timp3-deficient Mice Exhibit Adverse Aortic Remodeling after 2 Weeks of Angiotensin II Infusion

In WT mice, 2 weeks of Ang II infusion significantly elevated Timp3 mRNA and TIMP3 protein levels in the abdominal aorta (Fig. 1A, panels i and ii). To examine the role of TIMP3 in Ang II-induced aortic remodeling, we subjected mice lacking Timp3 to Ang II infusion, and we assessed the predominant arterial structural proteins, elastin and fibrillar collagen type I (12). Verhoeff-Van Gieson staining showed disorganized and disrupted elastin fibers with a greater frequency of elastin fiber disruption in the abdominal aorta of Timp3^−/−-Ang II compared with WT-Ang II mice (Fig. 1, B and C). The Ang II-induced reduction in elastin protein content was greater in Timp3^−/− abdominal aorta (Fig. 1D, panels i and ii) despite a larger increase in mRNA synthesis of elastin in these mice (Fig. 1D, panel iii). Gomori trichrome staining similarly revealed disorganized collagen structures (Fig. 1E, greenish blue) and a drastic decrease in collagen type I protein levels (Fig. 1F, panels i and ii) despite a marked rise in its mRNA synthesis (Fig. 1F, panel iii) in Timp3^−/−-Ang II compared with WT-Ang II abdominal aortas. These data suggest that the decrease in elastin and collagen protein levels in Timp3-deficient aorta is due to post-translational degradation of these proteins and not reduced synthesis.

Enhanced MMP2 Activation in Timp3^−/− Mice following 2 Weeks of Angiotensin II Infusion

TIMP3 is a potent inhibitor of a number of active MMPs and can inhibit cell surface activation of MMP2 (33); hence, its absence can lead to uncontrolled proteolytic activities (34). In situ gelatin zymography showed that Ang II infusion resulted in a stronger gelatinase activity in Timp3^−/− compared with WT aorta (Fig. 2A). To determine the contribution of individual gelatinases, namely MMP2 and MMP9, we performed in vitro gelatin zymography, which revealed a markedly greater MMP2 activation in Timp3^−/−-Ang II abdominal aorta as indicated by a stronger 64-kDa band, whereas MMP9 levels rose similarly in both genotypes (Fig. 2B, panels i and ii). The increase in MMP2 activation was also associated with increased mRNA expression (Fig. 2C, panel i), whereas Mmp9 mRNA did not increase (Fig. 2C, panel ii). Expression analysis of other Mmps showed a greater increase in Mmp13 levels in the Timp3^−/−-Ang II group (Fig. 2C, panel iii), whereas membrane type Mmp (MT1-MMP) was elevated similarly in both genotypes (Fig. 2C, panel iv).

Timp3-deficient Mice Developed AAA after 4 Weeks of Ang II Infusion

Next, we examined the long term outcome of these structural remodeling in the Timp3^−/− aorta. After 4 weeks of Ang II infusion, Timp3^−/− mice, but not WT mice, developed an aneurysm in the abdominal aorta and not in the ascending or descending thoracic aorta (Fig. 3A). Aneurysm is defined as a dilation of 50% or more in the abdominal aortic diameter (1, 35), measured by ultrasound imaging. In WT mice, Ang II-induced dilation of the abdominal aorta did not exceed 20% of the original aortic diameter, whereas 60% of Timp3^−/− mice exhibited greater than 50% aortic dilation in the suprarenal region. Representative ultrasonic images (Fig. 3B) and averaged diameter of the abdominal aorta (Fig. 3C, panels i and ii) demonstrate marked aortic dilation at the suprarenal region in Ang II-infused Timp3^−/− compared with the parallel WT group. The aortic systolic expansion index, a measure of aortic elasticity and recoil property during systole and diastole (29), was significantly suppressed in the aneurysmal aorta in Timp3^−/−-Ang II mice (Fig. 3C, panel iii), confirming the compromised structural and functional integrity of the Timp3-deficient aorta.

Deletion of Mmp2 in Timp3^−/− Mice Led to Exacerbated AAA

We observed that after 2 weeks of Ang II infusion MMP9 levels increased similarly in both genotypes, whereas the increase in MMP2 levels was greater in the Timp3^−/− compared with WT abdominal aorta (Fig. 2B). To determine the contribution of the elevated MMP2 activation in development of AAA in Timp3^−/−-Ang II mice, we generated Timp3^−/−/Mmp2^−/− double deficient mice. Saline-infused Timp3^−/− and Timp3^−/−/Mmp2^−/− mice showed a similar aortic structure and expansion index compared with WT mice (data not shown). Intriguingly, Timp3^−/−/Mmp2^−/− mice exhibited more severe AAA after 4 weeks of Ang II infusion compared with Timp3^−/− mice (Fig. 4A) as dilation of the abdominal aorta exceeded 50% of the original aortic diameter in 75% of these mice. In addition, whereas Timp3^−/− mice exhibited compromised survival post-Ang II infusion, mortality was increased further in Timp3^−/−/Mmp2^−/− mice (Fig. 4B, panel i). Routine autopsy examination revealed that aortic rupture at the suprarenal region of the aorta where aneurysmal dilation was detected (Fig. 4B, panel ii) was the cause of the increased mortality in these mice. Morphometric analysis of the full cross-section of abdominal aorta in saline-infused WT compared with Ang II-infused WT, Timp3^−/−, and Timp3^−/−/Mmp2^−/− mice clearly showed the dilation in abdominal aorta in Timp3^−/− mice and a strikingly greater dilation in the Timp3^−/−/Mmp2^−/− mice (Fig. 4C). Histological analyses at higher magnifications further revealed deterioration of the aortic wall at the site of aneurysm as indicated by disrupted elastin fibers (VVG staining) and collagen structure (GT staining) that were more severe in Timp3^−/−/Mmp2^−/− mice (Fig. 4C). Aortic wall structures in saline-infused mice of the three genotypes were comparable (data not shown).

Scanning electron microscopy allowed us to perform a more detailed evaluation of the structural remodeling of the abdominal aortic walls. Ang II infusion led to remodeling of the aortic wall in WT mice as indicated by compactly folded elastin lamellae (Fig. 4D, panel i) that were more evident at a higher magnification (Fig. 4D, panel ii). Timp3-deficient mice on the other hand exhibited disrupted medial elastic lamellae and disrupted fibrillar structures (Fig. 4D, panel i, open arrows), and this structural deterioration was worsened in Timp3^−/−/Mmp2^−/−-Ang II mice (Fig. 4D, panels i and ii). The structure and diameter of the abdominal aorta in saline-infused mice were comparable among the genotypes (data not shown). These findings demonstrate that despite the marked increase in MMP2 activation in Timp3^−/− aortas early in the post-Ang II infusion period, its deletion indeed aggravated the structural deterioration and the adverse remodeling in the abdominal aorta resulting in worsened AAA.

Timp3^−/−/Mmp2^−/−-Ang II Mice Exhibited Heightened Inflammation in Abdominal Aorta

Further examination of the Timp3^−/−/Mmp2^−/− mice revealed that Ang II-induced total collagenase activity was the largest in the abdominal aorta of Timp3^−/− mice and was reduced significantly in the Timp3^−/−/Mmp2^−/− group (Fig. 5A, panel i). This is consistent with MMP2 also being a potent collagenase (36). However, the Ang II-induced rise in gelatinase activity was markedly greater in Timp3^−/−/Mmp2^−/− compared with Timp3^−/− abdominal aorta despite the absence of MMP2 (Fig. 5A, panel ii). MMP2 and MMP9 are the classically known gelatinases, and as such, the rise in total gelatinase activity in the absence of MMP2 implied an increase in MMP9 activity. In vitro gelatin zymography confirmed a striking up-regulation in MMP9 levels in Ang II-infused Timp3^−/−/Mmp2^−/− abdominal aortas compared with all other groups (Fig. 5B, panels i and ii). This was also accompanied by elevated mRNA expression of Mmp9 (Fig. 5B, panel iii). The increase in MMP9 levels in WT and Timp3^−/− aortas after 2 weeks of Ang II infusion (Fig. 2B) was dissipated by 4 weeks. This could perhaps be due to an early inflammatory response that occurred irrespective of the genotype. MMP9 is a well known elastase that can degrade vascular elastin lamella (37), leading to invasion of inflammatory cells and inflammation (38) and exacerbation of AAA (22). Immunofluorescent staining for inflammatory cells, neutrophils and macrophages (superimposed with DAPI nucleus staining), showed enhanced infiltration of neutrophils and macrophages in the abdominal aorta of Ang II-infused Timp3^−/−/Mmp2^−/− mice (Fig. 5C). The Timp3^−/−-Ang II group showed a less severe macrophage infiltration, whereas no infiltrating cells were detected in WT-Ang II abdominal aorta (Fig. 5C). Consistently, mRNA levels of inflammatory markers interleukin-1β (IL-1β), IL-6, and monocyte chemotactic protein-1 were significantly elevated in the aneurysmal aorta of Timp3^−/−/Mmp2^−/− mice compared with parallel WT and Timp3^−/− groups (Fig. 5D, panels i–iii).

Inflammation Is a Key Factor in Exacerbated AAA in Timp3^−/−/Mmp2^−/−-Ang II Mice

To determine the mechanism underlying the worsening of AAA in Timp3^−/−/Mmp2^−/−-Ang II mice and the contribution of the observed heightened inflammation in these aortas, we reconstituted WT bone marrow in Timp3^−/−/Mmp2^−/− mice and vice versa. We generated Timp3^−/−/Mmp2^−/−-chimera mice by reconstituting WT bone marrow in these mice and WT-chimera mice by reconstituting Timp3^−/−/Mmp2^−/− bone marrow in WT mice. After 7 weeks of reconstitution followed by 4 weeks of Ang II infusion, we found that Timp3^−/−/Mmp2^−/−-chimera mice did not develop aortic aneurysm (Fig. 6A) as the diameter of abdominal aorta (Fig. 6B, panels i and ii) and aortic expansion index (Fig. 6B, panel iii) in these mice were comparable with saline-infused animals. Consistently, minimal infiltration of neutrophils and macrophages was detected (Fig. 6C, panel ii), whereas MMP9 levels were significantly reduced (Fig. 6D) in the Timp3^−/−/Mmp2^−/−-chimera-Ang II mice. WT-chimera mice exhibited neutrophil and macrophage infiltration that was not detected in Ang II-induced intact WT mice (Fig. 6C, panel i). In vitro gelatin zymography showed increased MMP9 and a striking increase in MMP2 levels in the WT-chimera mice (Fig. 6D). These data indicate that inflammation was the underlying cause of exacerbated AAA in Timp3^−/−/Mmp2^−/−-Ang II mice.

Multiple MMP Inhibition Prevented AAA in Timp3^−/− and Timp3^−/−/Mmp2^−/− Mice

To further investigate the contribution of the elevated protease activity as the underlying mechanism for AAA in Timp3^−/− and Timp3^−/−/Mmp2^−/− mice, we treated these mice with a broad spectrum MMPi, PD166793 (24, 39), during the course of Ang II infusion. Interestingly, simultaneous inhibition of a number of MMPs with this inhibitor blocked AAA development in Timp3^−/− and significantly suppressed AAA formation in Timp3^−/−/Mmp2^−/− mice (Fig. 7A). Ultrasonic measurement of abdominal aortic diameter showed that MMPi treatment was strongly effective in preventing the Ang II-induced dilation of the abdominal aorta in Timp3^−/− and Timp3^−/−/Mmp2^−/− mice (Fig. 7B, panels i and ii), and the aortic recoil property as measured by systolic expansion index was significantly restored in MMPi-treated Timp3^−/−-Ang II mice and to a lesser extent in Timp3^−/−/Mmp2^−/−-Ang II mice (Fig. 7B, panel iii). These improvements were well reflected in the abdominal aortic wall structures as shown by intact elastin (VVG) and collagen (GT) structures (Fig. 8A) and preserved elastin lamella organization as seen in the scanning electron microscopy images (Fig. 8B). Consistently, MMPi treatment prevented the elevation of collagenase and gelatinase activities in the abdominal aorta of Ang II-infused Timp3^−/− and Timp3^−/−/Mmp2^−/− mice (Fig. 8C). These data demonstrate the protective function of MMPi against Ang II-induced aortic wall deterioration that otherwise leads to AAA in Timp3^−/− and Timp3^−/−/Mmp2^−/− mice.