Matrix metalloproteinase inhibition reduces oxidative stress associated with cerebral amyloid angiopathy in vivo in transgenic mice.

Animals

APPswe/PS1dE9 mice ( Jankowsky et al. 2001 ) aged 7-8 months old and aged matched nontransgenic littermate mice were obtained from Jackson Laboratory (Bar Harbor, Maine), and Tg2576 mice aged 12-14 months were used. All studies were conducted with approval of the Massachusetts General Hospital Animal Care and Use Committee and in compliance with NIH guidelines for the use of experimental animals. Reagents Amplex Red (10-actyl-3,7-dihydroxyphenoxazine), Texas Red dextran 70,000 D and DQ™ gelatin, 1-10-phenanthroline monohydrate and collagenase IV from Clostridium histolyticum were obtained from Molecular Probes/Invitrogen (Eugene, OR) as part of the EnzChek gelatinase/collagenase assay kit. Tissue-Tek OCT compound was obtained from Sakura Finechemical Co. Ltd. (Tokyo, Japan). Methoxy-XO4 was a gift from Dr. Klunk (U. Pittsburgh). GM6001 (Llomastat, Galardin) was obtained from USBiological (Swampscott, MA). PBN, minocycline hydrochloride, simvastatin, thioflavin S, peroxidase and common chemical reagents where obtained from Sigma (St. Louis, MO).

Ex vivo antioxidant activity

Paraformaldehyde fixed brain sections of a Tg2576 mouse were used for the ex vivo assays as previously described with minor modifications ( McLellan et al. 2003 ; Garcia-Alloza et al. 2006a ). Mounted tissue was dehydrated and treated for 45 minutes with PBN (100 μM), minocycline, simvastatin or GM6001 (10 and 100 μM). Control tissue was incubated in PBS without drug treatment. Sections were carefully washed and incubated for 45 minutes with 200 μM AR (in the presence of 0.5 mg/ml peroxidase) along with the previously used antioxidants at the same concentrations, whereas control tissue was incubated in AR and peroxidase only. The tissue was covered to minimize light and air exposure. Sections were washed in PBS to rinse excess reagent, aqueously coverslipped, and imaged. Afterwards, brain tissue was washed in PBS and incubated for 20 minutes in thioflavin S (0.01%). After washing, the sections were covered and imaged again. Measuring the ROS-dependent fluorescence of AR from individual senile plaques in the tissue followed by the thioflavin S intensity for each plaque allows plaque-by-plaque normalization of the ROS signal to plaque size and morphology. In situ MMP zymography in fresh frozen tissue Brains from Tg2576 mice were dissected on ice, embedded without fixation in Tissue-Tek OCT compound and snap-frozen in liquid nitrogen. Serial 10 μm thick sections were obtained using a cryostat microtome. The EnzCheck Gelatinase Assay Kit was used for in situ zymography according to the provided protocol. Sections were incubated at 37C and divided into three experimental groups. Sections from Tg2576 and aged matched wild type mice were incubated with DQ™ gelatin, fluorescein conjugate for 90 min. A second set of sections were preincubated for 30 minutes with the general metalloproteinase inhibitor 1-10-phenanthroline monohydrate, and a third group of sections was incubated with DQ gelatin previously cleaved by mixing the DQ™ gelatin with type IV collagenase purified from Clostridium histolyticum. Sections were then viewed with widefield fluorescence microscopy and photographed. In vivo antioxidant and MMP inhibition treatments The effect of antioxidant treatment and MMP inhibiting agents on oxidative stress and MMP activity associated with CAA was measured in APPswe/PS1dE9 and Tg2576 mice. The doses and treatment schedules were selected based on previous reports in the literature for the selected compounds. APPswe/PS1dE9 mice were treated with the spin trap antioxidant PBN (100mg/Kg ip) for 2 consecutive days ( McLellan et al. 2003 ) before implanting a 6 mm cranial window. APPswe/PS1dE9 mice were also treated with agents that inhibit MMP: one group received minocycline (100mg/Kg ip) ( Garcia-Alloza et al. 2007b ) for 7 days before surgery and a second group received the same dose of minocycline for 14 days before surgery was done. Simvastatin was administered ip for 14 days (20 mg/Kg ip in 10% cremophor) ( Obama et al. 2004 ) and a final group received GM6001 100mg/Kg ip for 3 days (in 10%DMSO) ( Lee et al. 2006 ). Tg2576 mice were also used to assess the effect of the treatments on CAA progression, since this animal model shows a much more robust CAA progression than APPswe/PS1dE9 mice ( Garcia-Alloza et al. 2006b ; Robbins et al. 2006 ). Mini-osmotic ALZET pumps 2000 (0.5 μl/hour, 14 days) (Durect Corporatgion, Cupertino, CA) were subcutaneously placed in the back of the mice for long term delivery of PBN (10 mg/Kg in 10% cremaphor EL). Oxidative stress and MMP activity were checked 2 weeks later and pumps were replaced in order to assess the effect of PBN on CAA progression during the two next consecutive weeks. Minocycline (100 or 50 mg/Kg) was administered daily ip for 2 weeks before oxidative stress and MMP activity was assessed. Animals were treated with 50 mg/Kg minocycline for 2 more weeks to assess the effect of MMP inhibition on CAA progression.

Surgical preparation

Surgery was performed as previously described ( McLellan et al. 2003 ; Garcia-Alloza et al. 2006a ) with minor modifications. Mice were anesthetized with isoflurane. Oxidative stress was measured using Amplex Red, and MMP activity was measured using the green fluorescent substrate (DQ™ gelatin) both in APPswe/PS1dE9 and Tg2576 mice. Both compounds are non-fluorescent until oxidized or proteolytically cleaved, respectively. We also included thioflavin S for histochemical confirmation. Since the thioflavin S signal is very strong we imaged the 3 compounds in 3 immediately consecutive sessions so the signal detected was exclusively derived from the marker under study at each time. After removing the dura, 200 μl of Amplex red (1mM) in 0.5 mg/ml peroxidase was locally applied for 20 minutes, adding 50 μl every 5 minutes and covering the head to protect from light and air exposure. The site was then washed, a coverslip was attached over the site and the animal was imaged. Prior to the second imaging session, the coverslip was removed and animals were treated with 50 μl of the green fluorescent substrate DQ™ gelatin (1mg/ml) for 20 minutes. The site was washed and a coverslip was attached to re-image the same sites and vessel segments. Cranial windows were removed a third time and animals were incubated with thioflavin S (0.1 mg/ml) for 10 minutes and the coverslip was replaced and fixed with dental cement. Wild type mice were treated following the same protocol and vessels were imaged after Amplex Red, DQ™ gelatin and thioflavin S as described above in order to assess the specificity of the signal detected in transgenic mice. In separate experiments, to evaluate the time-course of fluorescence generation resulting from proteolytic cleavage of DQ™ gelatin in vivo , we imaged the brain of mice immediately after topical application of the substrate for up to1 h at 5 minute intervals. We measured progression of CAA in Tg2576 mice for 3 consecutive weeks as previously described ( Garcia-Alloza et al. 2006b ; Robbins et al. 2006 ). At day 0, Aβ was imaged with topical thioflavin S application as described above. After assessing oxidative stress and MMP activity, animals were re-imaged on a weekly basis. For these subsequent imaging sessions, Aβ was visualized with methoxy-XO4, a Congo Red derivative that binds fibrillar Aβ in a pattern identical to thioflavin S. Methoxy-X04 was injected i.p. the day before each imaging session ( Klunk et al. 2002 ).

Multiphoton imaging and processing

As previously described ( McLellan et al. 2003 ; Garcia-Alloza et al. 2006a ), two-photon fluorescence was generated with 800 nm excitation from a mode-locked Ti:Sapphire laser (MaiTai, Spectra-Physics, Mountain View, CA mounted on a multiphoton imaging system (Bio-Rad 1024ES, Bio-Rad, Hercules, CA). A custom-built external detector containing three photomultiplier tubes (Hamamatsu Photonics, Bridgewater, NJ) collected emitted light in the range 380-480, 500-540 and 560-650 nm. Ex vivo imaging of plaques in tissue sections was performed using the normal scan speed and multiple z-series were collected after adding AR and again after adding thioflavin S using a 20X water immersion objective (615×615 μm, z/step 2μ, depth of approximately 50 μm).

Using Image-J software

(NIH, freeware) the intensity of the immediate surroundings of the dense-core plaques was subtracted from the dense core fluorescence for AR and thioflavin S to correct for background levels. AR is a very sensitive probe for H 2 O 2 ( Zhou et al. 1997 ), although H 2 O 2 can lead to the production of many other ROS. A ratio between AR intensity and thioflavin S intensity was calculated for each plaque to normalize across images and mice. This ratio provides a quantitative, dimensionless index of plaque-associated oxidative stress. Results are expressed as a percentage of control values ( McLellan et al. 2003 ; Garcia-Alloza et al. 2006a ). In vivo imaging of CAA affected vessels was conducted under the same conditions (615×615 μm; z-step, 5 μm, depth, 200 μm approximately). Images were analyzed with Photoshop and Image J as previously described ( Robbins et al. 2006 ; Prada et al. 2007 ). We measured the effect of PBN and the MMP inhibiting agents minocycline, simvastatin and GM6001 on oxidative stress and MMP activity associated with CAA. Vessels were outlined and the oxidative stress signal (Amplex Red), MMP activity (DQ™ gelatin) signal and CAA deposits (thioflavin S signal), appearing as bright areas circumferential to the vessels, were manually thresholded so that signal-positive regions were set white and negative regions were set black. The number of pixels in white clusters composed of greater than 6 contiguous pixels was counted. For each vessel segment, the Amplex Red signal and the green fluorescent substrate DQ™ gelatin signal were expressed as a percentage of the histochemical confirmation signal from thioflavin S. Image volumes focused on leptomeningeal vessels and not parenchymal plaques in order to obtain sufficient high quality images, however, a small number of plaques that were Amplex red, MMP, and thioflavin S positive were detected, but not included in this study. Quantitative analysis of CAA progression in Tg2576 mice was performed as previously described ( Robbins et al. 2006 ; Prada et al. 2007 ). Vessels were outlined and the CAA deposits were manually thresholded and segmented. Greater than 6 contiguous pixels were required for a segmented region to be counted. The CAA burden was calculated as the percentage of the vessel area affected by CAA. The vessel area from the initial imaging session was used for the calculation of CAA burden because of the tendency for vessels to dilate in subsequent weeks. New growth of CAA deposits was measured from week to week. Vessel diameter was measured in ImageJ at three points along each vessel segment and the average diameter was taken from these three measurements.

Statistical analysis

To compare the levels of oxidative stress and MMP activity in the ex vivo and in vivo experiments, one-way ANOVA followed by post-hoc Tamhane or Tuckey b test as appropriate was used. Correlation between fluorescent signal from thioflavin S, oxidative stress and MMP activity was assessed by Pearson's correlation. To assess the nature of disease progression we used a linear model as previously described ( Robbins et al. 2006 ). In this modeling process, variability was assumed to originate on the basis of mouse-to-mouse variation (internal variation) as well as be influenced by location in the vascular tree (intra-animal variation), as suggested in previous studies ( Domnitz et al. 2005 ). In practice, this amounted to allowing for mouse-specific and vessel segment-specific intercepts in the linear model.