Intra-hippocampal transplantation of neural precursor cells with transgenic over-expression of IL-1 receptor antagonist rescues memory and neurogenesis impairments in an Alzheimer's disease model

Abstract

Ample evidence implicates neuroinflammatory processes in the etiology and progression of Alzheimer's disease (AD). To assess the specific role of the pro-inflammatory cytokine interleukin-1 (IL-1) in AD we examined the effects of intra-hippocampal transplantation of neural precursor cells (NPCs) with transgenic over-expression of IL-1 receptor antagonist (IL-1raTG) on memory functioning and neurogenesis in a murine model of AD (Tg2576 mice). WT NPCs- or sham-transplanted Tg2576 mice, as well as naive Tg2576 and WT mice served as controls. To assess the net effect of IL-1 blockade (not in the context of NPCs transplantation), we also examined the effects of chronic (4 weeks) intra-cerebroventricular (i.c.v.) administration of IL-1ra. We report that 12-month-old Tg2576 mice exhibited increased mRNA expression of hippocampal IL-1β, along with severe disturbances in hippocampal-dependent contextual and spatial memory as well as in neurogenesis. Transplantation of IL-1raTG NPCs 1 month before the neurobehavioral testing completely rescued these disturbances and significantly increased the number of endogenous hippocampal cells expressing the plasticity-related molecule BDNF. Similar, but less-robust effects were also produced by transplantation of WT NPCs and by i.c.v. IL-1ra administration. NPCs transplantation produced alterations in hippocampal plaque formation and microglial status, which were not clearly correlated with the cognitive effects of this procedure. The results indicate that elevated levels of hippocampal IL-1 are causally related to some AD-associated memory disturbances, and provide the first example for the potential use of genetically manipulated NPCs with anti-inflammatory properties in the treatment of AD.

Figure 1

Age-dependent behavioral and neuro-immune alterations in Tg2576 mice. (a) Fear-conditioning scores in different age groups of Tg2576 mice, presented as percent of WT mice at the corresponding ages. Memory functioning in the Tg2576 mice was normal (ie, ∼100% of control) at 3, 6, and 9 months of age, but it was significantly impaired in 12-month-old mice (n=9/group). (b) In the Morris water maze paradigm, 12-month-old Tg2576 mice displayed a significant spatial learning impairment compared with their littermate controls (n=9/group). (c) Hippocampal neurogenesis, assessed by counting the number of doublecortin-labeled cells in the dentate gyrus, was similar in young (3-month-old) Tg2576 and WT mice, whereas at an older age (12 months) WT mice maintained similar rate of neurogenesis but Tg2576 mice displayed markedly reduced neurogenesis (n=6/group). (d) Hippocampal IL-1β mRNA expression levels were similar in young (3-month-old) Tg2576 and WT mice, whereas at an older age (12 months) WT mice maintained the low expression but Tg2576 mice displayed markedly elevated IL-1β mRNA expression levels (n=8/group). *P<0.05 compared with all other age groups. **P<0.001 compared with all other groups.

Figure 2

Effects of NPCs transplantation on memory functioning. (a) Fear conditioning (reflected by % freezing during the memory retention test) was significantly impaired in naive and sham-operated Tg2576 mice, compared with naive WT mice. Transplantation with either WT or IL-1raTG NPCs completely reversed this memory impairment (n=9/group). (b) In the water maze paradigm, naive, sham-operated and WT NPCs-transplanted Tg2576 mice displayed severely impaired spatial learning and memory compared with the WT naive group. This impairment was completely rescued by transplantation with IL-1raTG NPCs. (c) On the last day of water maze training, naive and sham-operated Tg2576 mice exhibited significantly longer latencies to reach the platform than either naive WT or IL-1raTG NPCs-transplanted Tg2576 mice, whereas the WT NPCs-transplanted Tg2576 group did not differ from any of the other Tg2576 groups. (d) In the probe trial, the naive, sham and WT-transplanted NPCs Tg2576 groups displayed significantly fewer crossings over the previous location of the platform than either the WT or the IL-1raTG NPCs-transplanted groups. *P<0.05 compared with the Tg2576 naive and sham groups. **P<0.05 compared with the Tg2576 naive, sham and WT-transplanted groups.

Figure 3

Localization and characterization of transplanted NPCs. (a) A representative picture of the hippocampal DG 1 month following transplantation of BrdU-labeled NPCs. BrdU (red)-positive cells were mainly found in the hilus of the DG. Staining to both BrdU and the astrocytic marker GFAP (green) revealed that 78% of the BrdU transplanted cells expressed GFAP. (b) A representative picture in higher magnification of two cells, expressing both BrdU (red) and GFAP (green). (c) A representative picture taken from an adjacent slice, showing two transplanted cells that express both BrdU (red) and BDNF (green). (d) A representative picture showing two transplanted cells that express both BrdU (red) and IGF-1 (green). (e) A representative picture of two cells expressing both BrdU (red) and hIL-1ra (green). Such cells were found only in brains transplanted with NPCs derived from IL-1raTG mice.

Figure 4

Effects of NPCs transplantation on hippocampal neurogenesis and BDNF expression. (a) Hippocampal neurogenesis, determined by counting the number of DCX-labeled cells, was significantly suppressed in naive and sham-operated Tg2576 mice, as compared with WT mice, but transplantation of either WT- or IL-1raTG-derived NPCs completely rescued this impairment (n=8/group). (b) A representative picture of the DG from a sham-operated mouse demonstrates the labeling of only two DCX-labeled adult-born cells (green; marked by white arrows). (c) A representative picture of the DG from a Tg2576 mouse transplanted with IL-1raTG-derived NPCs demonstrates the labeling of numerous DCX-labeled cells (green; marked by white arrows). (d) The number of BDNF-labeled endogenous cells in Tg2576 mice was elevated by transplantation of WT NPCs, as compared with WT naive mice (n=8/group). This number was further increased by transplantation of NPCs derived from IL-1raTG mice, rising to a level significantly greater than in all other groups (including the WT NPCs-transplanted group). (e) A representative picture demonstrating a relatively low number of BDNF-expressing cells (green) in the DG of a naive WT mouse, as compared with (f) A representative picture of the DG from a Tg2576 mouse transplanted with IL-1raTG NPCs, demonstrating a relatively high number of BDNF-expressing cells (green). *P<0.05 compared with the Tg2576-naive and Tg2576-sham groups. **P<0.05 compared with the WT-naive group. ***P<0.05 compared with all other groups.

Figure 5

Effects of NPCs transplantation in Tg2576 mice on plaque area and microglia density. (a) At 5-6 weeks post NPCs transplantation, the area of the DG covered with plaques was significantly increased in sham-operated Tg2576 mice, compared with naive Tg2576 mice. Transplantation with either WT or IL-1raTG NPCs significantly reduced this operation-induced plaque area increase, although plaque area in the transplanted groups was still significantly greater than in the naive group (n=6/group). (b) A representative picture of thioflavin-labeled plaques (green staining) in the hippocampal DG (delineated by blue Dapi staining) of a sham-operated Tg2576 mouse as well as (c) a mouse transplanted with IL-1raTG NPCs. (d) The area of the piriform cortex covered with plaques was significantly increased in the sham-operated mice, as well as in mice transplanted with WT or IL-1raTG NPCs, compared with naive mice. (e) Microglia density within the DG (cells per mm²), as well as (f) the number of microglia associated with plaques (cells per μm² plaque area) were significantly increased in mice transplanted with WT or IL-1raTG NPCs (n=6/group). (g) A representative confocal microscope image of Iba-1-labeled microglia (red) associated with a thioflavin-labeled plaque (green) in a sham-operated Tg2576 mouse, and (h) in an IL-1raTG-transplanted Tg2576 mouse. (i) A representative confocal image of plaque-associated microglia with double-staining for Iba-1 (green) and MHCII (red). (j) A representative confocal image of IL-10 labeled plaque-associated cells in IL-1raTG-transplanted Tg2576 mouse. (k) No group differences were found in microglial density (cells per mm²) and (l) the number of microglia associated with plaques (cells per μm² plaque area) in the piriform cortex (n=6/group). *P<0.05 compared with the Tg2576-naive, Tg2576-sham and WT-naive (when applicable) groups. **P<0.05 compared with all other groups.

Figure 6

Effects of pharmacological blockade of IL-1 signaling by chronic i.c.v. administration of IL-1ra. (a) hIL-1ra levels in the hippocampus were significantly elevated following chronic i.c.v. administration of hIL-1ra (n=5-6/group). (b) Fear-conditioning memory, expressed as % freezing during the retention test, was significantly impaired in both vehicle- and IL-1ra-treated Tg2576 groups, ie, the IL-1ra treatment had no effect on memory impairment in this task (n=9/group). (c) In the Morris water maze, Tg2576 mice treated with vehicle showed severe spatial memory impairment compared with the WT groups, with no improvement over the entire training period. This deficit was completely rescued in Tg2576 mice treated with IL-1ra, which displayed performance similar to the groups of WT mice. (d) In the probe test of the water maze, Tg2576 mice treated with vehicle showed almost no crossing over the previous location of the platform, and this deficit was also completely rescued in Tg2576 mice treated with IL-1ra, which performed similarly to the groups of WT mice. (e) Hippocampal neurogenesis, expressed as the number of doublecortin (DCX)-labeled cells in the DG, was markedly reduced in Tg2576 mice treated with vehicle, and this reduction was significantly attenuated in Tg2576 mice treated with IL-1ra. (f) The number of BDNF-labeled cells in the DG was significantly increased in both WT and Tg2576 treated with IL-1ra. (g) Plaque area and (h) microglial density in the DG were not altered following chronic i.c.v. administration of IL-1ra. (i) The number of plaque-associated microglia in the DG was significantly reduced following chronic i.c.v. administration of IL-1ra. *<0.05 compared with both groups of WT controls. **<0.05 compared with the vehicle-treated mice of the same strain.