MMP9 modulation improves specific neurobehavioral deficits in a mouse model of Alzheimer's disease

Abstract

Background: Matrix metallopeptidase 9 (MMP9) has been implicated in a variety of neurological disorders, including Alzheimer's disease (AD), where MMP9 levels are elevated in the brain and cerebrovasculature. Previously our group demonstrated apolipoprotein E4 (apoE4) was less efficient in regulating MMP9 activity in the brain than other apoE isoforms, and that MMP9 inhibition facilitated beta-amyloid (Aβ) elimination across the blood-brain barrier (BBB) METHODS: In the current studies, we evaluated the impact of MMP9 modulation on Aβ disposition and neurobehavior in AD using two approaches, (1) pharmacological inhibition of MMP9 with SB-3CT in apoE4 x AD (E4FAD) mice, and (2) gene deletion of MMP9 in AD mice (MMP9KO/5xFAD) RESULTS: Treatment with the MMP9 inhibitor SB-3CT in E4FAD mice led to reduced anxiety compared to placebo using the elevated plus maze. Deletion of the MMP9 gene in 5xFAD mice also reduced anxiety using the open field test, in addition to improving sociability and social recognition memory, particularly in male mice, as assessed through the three-chamber task, indicating certain behavioral alterations in AD may be mediated by MMP9. However, neither pharmacological inhibition of MMP9 or gene deletion of MMP9 affected spatial learning or memory in the AD animals, as determined through the radial arm water maze. Moreover, the effect of MMP9 modulation on AD neurobehavior was not due to changes in Aβ disposition, as both brain and plasma Aβ levels were unchanged in the SB-3CT-treated E4FAD animals and MMP9KO/AD mice compared to their respective controls.

Conclusions: In total, while MMP9 inhibition did improve specific neurobehavioral deficits associated with AD, such as anxiety and social recognition memory, modulation of MMP9 did not alter spatial learning and memory or Aβ tissue levels in AD animals. While targeting MMP9 may represent a therapeutic strategy to mitigate aspects of neurobehavioral decline in AD, further work is necessary to understand the nature of the relationship between MMP9 activity and neurological dysfunction.

Keywords: Alzheimer’s disease; Apolipoprotein E; Beta-amyloid; Matrix metallopeptidase 9; Social recognition memory.

Fig. 1

Study design for the 4-week pharmacological MMP9 inhibition and MMP9 gene deletion in vivo analysis. A Study 1. MMP9 was pharmacologically inhibited in E4FAD mice through a 4-week treatment with vehicle or SB-3CT (25 mg/kg), injected intraperitoneally (IP). B Study 2. MMP9KO mice were crossed with 5xFAD mice to create 5xFAD/MMP9KO mice, which were tested alongside WT, 5xFAD and MMP9KO mice. Behavioral analysis consisted of the elevated plus maze (EPM), open field test (OFT), three chamber test and the radial arm water maze (RAWM). MMP Matrix metallopeptidase, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, MMP9KO MMP9 knockout, WT Wild-type

Fig. 2

Anxiety-related behavior and locomotor activity in the OFT and the EPM. SB-3CT and vehicle-treated E4FAD mice were tested using A, B, C the OFT and D, E, F, the EPM. A For the OFT, the duration spent in the outer, middle and center areas of the circular arena was measured along with B the total distance travelled and C the average velocity. D For the EPM, the duration spent in the closed and open arms was measured together with E the total distance travelled and F the average velocity. Animals were approximately 5 months of age and values represent mean ± SEM (N = 14, 5 males, 9 females for each group). Statistical significance was determined by A, D ANOVA followed by the BKY procedure and B, C, E, F an unpaired t-test. *p < 0.05. EPM Elevated plus maze, OFT Open field test, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 3

Testing social interaction and social memory using the three-chamber test. SB-3CT and vehicle-treated E4FAD mice were tested for A, B social interaction (one mouse vs empty cage) and C, D social memory (novel mouse vs familiar mouse) in the three-chamber test. Time spent in A, C the whole chamber containing the mouse/empty cage and B, D the proximal zone surrounding the cages was measured (areas shown in green in each accompanying schematic). Animals were approximately 5 months of age and values represent mean ± SEM (N = 14, 5 males, 9 females for each group). No statistical significance was identified in any measures by ANOVA. E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, SEM Standard error of the mean, ANOVA Analysis of variance

Fig. 4

Spatial memory testing using the RAWM. SB-3CT and vehicle-treated E4FAD mice were tested for their ability to find the hidden platform in the RAWM. A A schematic of the maze layout is displayed. Mice were tested in nine trials per day for 5 days. B The number of incorrect entries made and C the time taken to find the maze were recorded and analyzed. D The total distance travelled per trial and E the average velocity while swimming was also evaluated. Animals were approximately 5 months of age and values represent mean ± SEM (N = 14, 5 males, 9 females for each group). No statistical significance was identified in any of the parameters by ANOVA. RAWM Radial arm water maze, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, SEM Standard error of the mean, ANOVA Analysis of variance

Fig. 5

Spatial memory testing of male mice using the RAWM. SB-3CT and vehicle-treated male E4FAD mice were tested for their ability to find the hidden platform in the RAWM. A A schematic of the maze layout is displayed. Mice were tested in nine trials per day for 5 days. B The number of incorrect entries made and C the time taken to find the maze were recorded and analyzed. D The total distance travelled per trial and E the average velocity while swimming was also evaluated. Values represent mean ± SEM (N = 5 males for each group). No statistical significance was identified in any of the parameters by ANOVA. RAWM Radial arm water maze, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, SEM Standard error of the mean, ANOVA Analysis of variance

Fig. 6

Analysis of Aβ-40 and Aβ-42 levels in the whole parenchyma and cerebrovascular brain fractions. Aβ-40 levels were examined in the A whole parenchyma, while Aβ-42 levels were examined in the B whole parenchyma and C cerebrovasculature of E4FAD mice. Males and females were analyzed separately due to large differences in amyloid levels. Animals were approximately 5 months of age and values represent mean ± SEM (N = 14, 5 males, 9 females for each group). Statistical significance was determined by ANOVA followed by the BKY procedure. *p < 0.05, **p < 0.01 ***p < 0.001. Aβ Beta-amyloid, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 7

Analysis of LDLR and LRP1 levels in the cerebrovasculature and the soluble brain fraction. Levels of the A, B LDLR receptor and C, D LRP1 receptor were analyzed in the A, C soluble brain fraction and B, D the cerebrovasculature of SB-3CT and vehicle-treated E4FAD mice. Animals were approximately 5 months of age and values represent mean ± SEM (N = 14, 5 males, 9 females for each group). No statistically significant differences in LDLR or LRP1 levels were identified between vehicle or SB-3CT-treated mice in either brain fraction by ANOVA. LDLR Low-density lipoprotein receptor, LRP1 Low density lipoprotein receptor-related protein 1, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, SEM Standard error of the mean, ANOVA Analysis of variance

Fig. 8

Analysis of MMP9 levels in SB-3CT and vehicle-treated E4FAD mice. A, B Levels of proMMP9 were examined in spleen samples from SB-3CT and vehicle-treated mice by zymography. A Zymography gel showing bands of proMMP9 in SB-3CT-treated and control animals. B Quantification of zymographic analysis. C, D Levels of total MMP9 as measured by ELISA analysis of the C cerebrovasculature and D the whole brain parenchyma of SB-3CT and vehicle-treated E4FAD mice. Animals were approximately 5 months of age and values represent mean ± SEM (N = 14, 5 males, 9 females for each group). No statistical significance was identified in any fraction by ANOVA. MMP Matrix metallopeptidase, E4FAD Apolipoprotein E4 x Familial Alzheimer’s disease, ELISA Enzyme linked immunosorbent assay, SEM Standard error of the mean, ANOVA Analysis of variance

Fig. 9

Anxiety-related behavior and locomotor activity in the OFT and the EPM. Wild type, 5xFAD, 5xFAD/MMP9KO and MMP9KO mice were tested using A, B, C the OFT and D, E, F, the EPM. A For the OFT, the duration spent in the outer, middle, and center areas of the circular arena was measured along with B the total distance travelled and C the average velocity. D For the EPM, the duration spent in the closed and open arms was measured together with E the total distance travelled and F the average velocity. Animals were approximately 6 months of age and values represent mean ± SEM (N = 12, 6 males, 6 females for each group). Statistical significance was determined by (A, D) ANOVA and (B, C, E, F) one-way ANOVA followed by the BKY procedure. *p < 0.05, **p < 0.01. OFT Open field test, EPM Elevated plus maze, WT Wild-type, MMP Matrix metallopeptidase, FAD Familial Alzheimer’s disease, MMP9KO MMP9 knockout, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 10

Testing social interaction and social memory using the three-chamber test. Wild type, 5xFAD, 5xFAD/MMP9KO and MMP9KO mice were tested for A, B social interaction (one mouse vs empty cage) and C, D social memory (novel mouse vs familiar mouse) in the three-chamber test. Time spent in A, C the whole chamber containing the mouse/empty cage and B, D the proximal zone surrounding the cages was measured (areas shown in green in each accompanying schematic). Animals were approximately 6 months of age and values represent mean ± SEM (N = 12, 6 males, 6 females for each group). Statistical significance was determined by ANOVA followed by the BKY procedure. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. WT Wild-type, MMP Matrix metallopeptidase, FAD Familial Alzheimer’s disease, MMP9KO MMP9 knockout, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 11

Testing social interaction and social memory of male mice using the three-chamber test. Male wild type, 5xFAD, 5xFAD/MMP9KO and MMP9KO mice were tested for A, B social interaction (one mouse vs empty cage) and C, D social memory (novel mouse vs familiar mouse) in the three-chamber test. Time spent in A, C the whole chamber containing the mouse/empty cage and B, D the proximal zone surrounding the cages was measured (areas shown in green in each accompanying schematic). Animals were approximately 6 months of age and values represent mean ± SEM (N = 6 males for each group). Statistical significance was determined by ANOVA followed by the BKY procedure. *p < 0.05, **p < 0.01, ***p < 0.001. WT Wild-type, MMP Matrix metallopeptidase, FAD Familial Alzheimer’s disease, MMP9KO MMP9 knockout, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 12

Testing spatial memory using the RAWM. Wild type, 5xFAD, 5xFAD/MMP9KO and MMP9KO mice were tested for their ability to find the hidden platform in the RAWM. A A schematic of the maze layout is displayed. Mice were tested in nine trials per day for 5 days. B The number of incorrect entries made and C the time taken to find the maze were recorded and analyzed. D The total distance travelled per trial and E the average velocity while swimming were also evaluated. Animals were approximately 6 months of age and values represent mean ± SEM (N = 12, 6 males, 6 females for each group). No statistical significance was identified in any of the parameters by ANOVA. RAWM Radial arm water maze, WT Wild-type, MMP Matrix metallopeptidase, FAD Familial Alzheimer’s disease, MMP9KO MMP9 knockout, SEM Standard error of the mean, ANOVA Analysis of variance

Fig. 13

Analysis of Aβ-40 levels in the cerebrovasculature, plasma, and whole parenchyma brain fractions. Aβ-40 levels were examined in the A cerebrovasculature, B whole parenchyma, and C plasma of 5xFAD, 5xFAD/MMP9KO and 5xFAD/MMP9KO-het mice. Males and females were analyzed separately due to large differences in amyloid levels. Animals were approximately 6 months of age and values represent mean ± SEM (N = 12, 6 males, 6 females for each group). Statistical significance was determined by ANOVA followed by the BKY procedure. *p < 0.05, **p < 0.01, ***p < 0.001. Aβ Beta-amyloid, FAD Familial Alzheimer’s disease, MMP Matrix metallopeptidase, MMP9KO MMP9 knockout, MMP9KO-het MMP9KO heterozygous, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 14

Analysis of Aβ-42 levels in the cerebrovasculature and whole parenchyma brain fractions. Aβ-42 levels were examined in the A cerebrovasculature, and B whole parenchyma of 5xFAD, 5xFAD/MMP9KO and 5xFAD/MMP9KO-het mice. Males and females were analyzed separately due to large differences in amyloid levels. Animals were approximately 6 months of age and values represent mean ± SEM (N = 12, 6 males, 6 females for each group). Statistical significance was determined by ANOVA followed by the BKY procedure. **p < 0.01, ****p < 0.0001. Aβ Beta-amyloid, FAD Familial Alzheimer’s disease, MMP Matrix metallopeptidase, MMP9KO MMP9 knockout, MMP9KO-het MMP9KO heterozygous, SEM Standard error of the mean, ANOVA Analysis of variance, BKY Benjamini, Krieger, and Yekutieli

Fig. 15

Analysis of LDLR and LRP1 levels in the cerebrovasculature and the soluble brain fraction. Levels of the A, B LDLR receptor and C, D LRP1 receptor were analyzed in the A, C soluble brain fraction and B, D the cerebrovasculature of wild type, 5xFAD, 5xFAD/MMP9KO, 5xFAD/MMP9KO-het and MMP9KO mice. Animals were approximately 6 months of age and values represent mean ± SEM (N = 12, 6 males, 6 females for each group). No statistically significant differences in LDLR or LRP1 levels were identified between any genotype in either brain fraction by ANOVA. LDLR Low-density lipoprotein receptor, LRP1 Low density lipoprotein receptor-related protein 1, FAD Familial Alzheimer’s disease, MMP Matrix metallopeptidase, MMP9KO MMP9 knockout, MMP9KO-het MMP9KO heterozygous, SEM Standard error of the mean, ANOVA Analysis of variance