Genomic deletion of TLR2 induces aggravated white matter damage and deteriorated neurobehavioral functions in mouse models of Alzheimer's disease

Abstract

Toll-like receptor-2 (TLR2), a member of the TLR family, plays an important role in the initiation and regulation of immune/inflammation response, which is a critical mechanism underlying Alzheimer's disease (AD). To clarify the role of TLR2 in the pathological process of AD, in the present study, TLR2 knockout plus APPswe/PSEN1dE9 transgenic mice (AD-TLR2KO) were generated. Neurobehavioral tests and brain MRI scan were conducted on mice at the age of 12 months. Additionally, neuron loss was evaluated using NeuN staining. Amyloid β protein (Aβ), glial fibrillary acidic protein (GFAP), endogenous ligands for TLR2, and the activation of downstream signaling of TLR2 in mouse brains were detected by immunohistochemistry and Western blots. The results demonstrated that TLR2 deficit induced learning disabilities, decreased spontaneous activity, increased anxiety and depression, and led to white matter damage (WMD), brain atrophy, loss of neurons, and glial activation. Moreover, TLR2 deficit aggravated impaired neurobehavioral functions and WMD in AD mice, but did not affect the Aβ deposition in mouse brains. Our data indicate that the genomic deletion of TLR2 impairs neurobehavioral functions, induces WMD and brain atrophy, and increases the activation of astrocytes, which in turn aggravate the symptoms of AD through a non-Aβ mechanism.

Keywords: Alzheimer’s disease; MRI; TLR2; neurobehavioral function; white matter damage.

Figure 1

Genomic deletion of TLR2 accelerated cognitive impairments in mouse models of AD. Morris water maze tests (MWM) were performed on mice for seven consecutive days of place navigation processes, and a probe test was conducted on the 8^th day for mice aged 12 months. (A) Representative track plots of MWM for place navigation and probe tests. (B) The latency (time to find the hidden platform) increased in TLR2 knockout mice (&: TLR2KO vs. WT, p<0.05), AD mice (#: AD vs. WT, p<0.05), and AD-TLR2KO mice (+: AD-TLR2KO vs. WT, p<0.05). Moreover, the latency was prolonged in AD-TLR2KO mice compared with AD mice (*: AD-TLR2KO vs. AD, p<0.05). (C) Since mice have the habit of swimming along the pool, most of the mice searched the platform using an edge-type strategy on the first day of the test. As the tests progressed, WT mice converted to the tendency-type and straight-type searching strategies. However, compared with WT mice, AD mice exhibited a significantly higher ratio of the edge-type searching strategy (#: AD vs. WT, p<0.05). In addition, the ratio of the edge-type strategy in TLR2KO mice was significantly higher than that in WT mice (&: TLR2KO vs. WT, p<0.05). Importantly, the ratio of the edge-type strategy increased in AD-TLR2KO mice compared with AD mice (*: AD-TLR2KO vs. AD, p<0.05). (D and E) In the probe test, the frequencies of crossing the platform area in AD, AD-TLR2KO, and TLR2KO mice were significantly lower than those in WT mice (p<0.05, D). The time spent in the target quadrant among AD, AD-TLR2KO, and TLR2KO mice was significantly less than that among WT mice (p<0.05, E). However, there was no difference in the frequencies of crossing the platform area or in the time spent in the target quadrant between AD and AD-TLR2KO mice (Note: &: TLR2KO vs. WT, p<0.05; #: AD vs. WT, p<0.05; +: AD-TLR2KO vs. WT, p<0.05; *: AD-TLR2KO vs. AD, p<0.05. n=11~15/group).

Figure 2

TLR2 knockout increased the anxiety and depression states in mouse models of AD. (A) Representative track plots from open field maze test. (B) The exploration times in the central area were significantly shorter in AD, TLR2, and AD-TLR2KO mice compared with WT mice (p<0.05). Moreover, the exploration time in the central area among AD-TLR2KO mice was shorter than that among AD mice (p<0.05). (C) The total traveled distance of AD-TLR2KO mice was shorter compared with WT mice (p<0.05). (D) In the tail suspension test, the rest time for AD-TLR2KO mice significantly increased compared with WT mice (p<0.05). (E) However, the time in enclosed arms did not show a significant difference among the groups in the elevated plus maze test (n=11~15/group).

Figure 3

Aβ deposition in mouse brains. (A) Representative immunofluorescence images of Aβ deposition in mouse brains (4x, 20x: cortex; 10x: hippocampus). (B) Representative Aβ levels in brain tissues detected by Western blots. (C) Results from quantitative analyses of Western blots showed that Aβ levels were significantly higher in AD and AD-TLR2KO mice compared with WT and TLR2KO mice, respectively (#: AD vs. WT, p<0.05; +: AD-TLR2KO vs. WT, p<0.05). However, there was no significant difference in Aβ levels between AD and AD-TLR2KO mice (n=6 / group).

Figure 4

White matter integrity evaluated by DTI images detected by 7 Tesla MRI system. (A) Representative diffusion-encoded-color (DEC) map and fractional anisotropy (FA) map showing white matter injury (yellow arrow). (B) Quantitative analysis showed that the FA value significantly decreased in TLR2KO mice (&), AD mice (#), and AD-TLR2KO mice (+) compared with WT mice (p<0.05, B). (C) Axial diffusivity (Da) value decreased in AD-TLR2KO mice compared with WT mice (+: p<0.05). (D) Radial diffusivity (Dr) value increased in AD (#) and AD-TLR2KO (+) mice compared with WT mice (p<0.05). (E) There was no significant deference in the mean diffusivity (MD) value among the groups. (n= 4~6 / group).

Figure 5

Thickness of cortex evaluated by T2-weighted images detected by 7 Tesla MRI system. The cortical thickness was measured by Image J software on T2-weighted images obtained from Paravision software. (A) Representative T2-weighted images used for measurement of cortical thickness (red line segment indicates the measured regions). (B) Results showed that cortical thickness was reduced in TLR2KO mice and AD-TLR2KO mice compared with WT mice (&: TLR2KO vs. WT, p<0.05; +: AD-TLR2KO vs. WT, p<0.05). (n= 4~6 / group).

Figure 6

Levels of GFAP in mouse brains. (A) Representative immunofluorescence staining of GFAP, a marker of astrocytes. (B) Representative bands of GFAP in brain tissues detected by Western blots. (C) GFAP significantly increased in AD mice and AD-TLR2KO mice compared with WT and TLR2 mice, respectively (#: AD vs. WT, p<0.05; +: AD-TLR2KO vs. WT, p<0.05). Moreover, the level of GFAP in AD-TLR2KO mice was significantly higher than that in AD mice (*: AD-TLR2KO vs. AD, p<0.05). (n= 6 / group).

Figure 7

Levels of synaptophysin (Syn) and PSD95 in mouse brains. (A) Representative image of Western blots for Syn. (B) Representative image of Western blots for PSD95. (C, D) There was no significant difference in the levels of Syn (C) and PSD95 (D) between AD and AD-TLR2KO mice. (n= 6 / group).

Figure 8

TLR2 deficiency resulted in neuronal loss in mouse brains. (A, B) Neuronal density detection and quantification analysis indicated a lower capacity of NeuN⁺ cells in AD, TLR2KO, and AD-TLR2KO mice when compared with WT mice (n=6 for each group, p<0.05).

Figure 9

Expression of endogenous ligands for TLR2. (A) Expression of biglycan in AD-TLR2KO mice increased significantly compared with that in WT, AD, and TLR2KO mice (p<0.05). (B) HMGB1 in the four groups did not show a significant difference (p>0.05).

Figure 10

Activation of MyD88-NFκB signaling. Expression of MyD88 significantly increased in AD mice compared with WT mice (p<0.05). The increased MyD88 was inhibited in AD-TLR2KO mice compared with AD mice (p<0.05). Expression of pNF-kB significantly increased in AD mice compared with WT mice (p<0.05). The phosphorylation of NF-kB was inhibited in AD-TLR2KO mice compared with AD mice (p<0.05).