Structural and functional rejuvenation of the aged brain by an approved anti-asthmatic drug

Abstract

As human life expectancy has improved rapidly in industrialized societies, age-related cognitive impairment presents an increasing challenge. Targeting histopathological processes that correlate with age-related cognitive declines, such as neuroinflammation, low levels of neurogenesis, disrupted blood-brain barrier and altered neuronal activity, might lead to structural and functional rejuvenation of the aged brain. Here we show that a 6-week treatment of young (4 months) and old (20 months) rats with montelukast, a marketed anti-asthmatic drug antagonizing leukotriene receptors, reduces neuroinflammation, elevates hippocampal neurogenesis and improves learning and memory in old animals. By using gene knockdown and knockout approaches, we demonstrate that the effect is mediated through inhibition of the GPR17 receptor. This work illustrates that inhibition of leukotriene receptor signalling might represent a safe and druggable target to restore cognitive functions in old individuals and paves the way for future clinical translation of leukotriene receptor inhibition for the treatment of dementias.

Figure 1. Montelukast treatment improves learning and memory in old rats.

(a) Latency times to find the hidden platform in the Morris water maze test after daily oral administration of 10 mg kg⁻¹ montelukast for 6 weeks. (b) Representative tracks of the distances moved in the Morris water maze on day 4 in the different age and treatment groups (scale bar, 50 cm). (c) Swimming speeds of animals in the Morris water maze test. (d–f) Results of probe trial on day 6: (d) time spent in the former platform quadrant; (e) time spent in zone P, a defined circular area of 20 cm diameter that encloses the former platform location; (f) number of entries into zone P. (g) Results of the object location memory test (performed on an additional cohort of rats). (h) Body weight of the animals on days 1 and 42 of treatment. (i) Locomotion as assessed by the distance moved in the open field. (j) Anxiety assessed in the elevated plus maze with analysis of time spent in the closed, centre and open arms. (k) Depression-like behaviour (immobility time in the forced swim test). Data are shown as mean±s.e.m. (a, c) or mean±s.d. (d–k). * indicates P<0.05; # indicates significant differences (P<0.05) compared to young vehicle. Two-way ANOVA (a,c,h,j), One-way ANOVA (d–f,i,k) with Bonferroni post hoc tests, and the unpaired Student's t-test (g) were performed. N per group (a–f; h–k): 10 (young vehicle), 10 (young montelukast), 7 (old vehicle), 7 (old montelukast); (g): 5 (old vehicle), 5 (old montelukast). yg veh, young vehicle; yg mtk, young montelukast; old veh, old vehicle; old mtk, old montelukast.

Figure 2. Montelukast modulates microglia in old rats and neuroinflammatory gene expression.

(a) Iba1⁺ immunostaining in the DG of the different age- and treatment groups (white arrowheads point toward representative Iba1⁺ cells). Control stainings, in which the primary goat Iba1 antibody was omitted, excluded unspecific labelling of the secondary antibody. (b) Total numbers of Iba1⁺ cells, and (c) percentages of proliferating microglia (Iba1⁺/PCNA⁺) in the DG of young (4 months) and old (20 months) rats after a 6-week treatment with montelukast (10 mg kg⁻¹ daily p.o.). (d) Soma sizes of Iba1⁺ cells in the DG of young and old montelukast- and vehicle-treated rats. (e) Representative images of CD68⁺ immunostaining in Iba1⁺ microglia (arrowheads) of the different experimental groups. (f) Percentage of Iba⁺ cells that contain CD68⁺ particles. (g) Average size of CD68⁺ particles in Iba1⁺ cells of in the DG. (h) Immunoreactivity for CysLTR1 and for GPR17 in BV-2 cells. (i) mRNA expression of NOS2, CCL2, TNF, TGF-beta1 and of ARG1 in BV-2 cells 24 h after treatment with 100 nM LTD4 and/or with 15 μM montelukast, and with vehicle control. N per group in (b–d,f,g): 10 (young vehicle), 10 (young montelukast), 7 (old vehicle), 7 (old montelukast). Every tenth section (400 μm interval) of one brain hemisphere (b,c) or four sections (d,f,g) per animal were analysed. In (i), 3–5 biological replicates were performed in technical duplicates. Data are shown as mean±s.d. *P<0.05, **P<0.01, ***P<0.001. Significant differences between the old age groups and the young vehicle group were indicated by # (#=P<0.05; ###=P<0.001). One-way ANOVA (b–d,f,g) and Two-way ANOVA (i) with Bonferroni post hoc tests were performed. Scale bars, 50 μm (a); 100 μm (e). yg veh, young vehicle; yg mtk, young montelukast; old veh, old vehicle; old mtk, old montelukast; DG, dentate gyrus.

Figure 3. Montelukast treatment increases dentate gyrus neurogenesis in old rats.

(a) Number of PCNA⁺ cell numbers in the DG. (b) Number of Sox2⁺ neural stem cells in the subgranular zone of the DG. (c) DCX⁺ cell numbers in the DG. (d) Number of BrdU⁺ cells 4 weeks after BrdU labelling of the cells. (e) The percentages of BrdU⁺ cells that have differentiated into neurons (BrdU⁺/NeuN⁺) or astrocytes (BrdU⁺/GFAP⁺). Arrowheads point to cells stained positively for the respective marker. Pictures illustrate representative images of antibody labelling in the DG of old vehicle and old montelukast-treated rats. Scale bars, 200 μm (a, d), 100 μm (b, c, e). Data are shown as mean±s.d. * indicates P<0.05; # indicates significant differences (P<0.05) specifically compared with young vehicle. One-way ANOVA with Bonferroni post hoc tests was performed. N per group: 10 (young vehicle), 10 (young montelukast), 7 (old vehicle), 7 (old montelukast). Every tenth section (400 μm interval) of one brain hemisphere per animal was analysed. yg veh, young vehicle; yg mtk, young montelukast; old veh, old vehicle; old mtk, old montelukast; DG dentate gyrus.

Figure 4. Montelukast-mediated cognitive improvements in old rats correlate with increased neurogenesis.

(a) Correlation analysis showed that the individual learning scores of old vehicle-treated and old montelukast-treated rats were independent of the soma size of Iba1⁺ cells in the DG (vehicle: R²=0.001; y=81.79+0.034*x; montelukast: R²=0.027; y=137.35 +0.368*x). (b) The correlation coefficient between learning success and number of PCNA⁺ cells revealed a tendency towards correlation in the old vehicle-treated rats (R²=0.361; y=63.75+0.063*x). Montelukast provoked a stronger correlation and led to a steeper slope of the regression line (R²=0.843; y=80.8+0.129*x). (c) Similarly, learning and the number of BrdU⁺ cells showed a trend for correlation in vehicle-treated old rats (R²=0.225; y=64.11+0.111*x), and again, this correlation was strengthened in montelukast-treated rats (R²=0.601; y=69.01+0.261*x). N per group: 6 (old vehicle), 6 (old montelukast). Correlation analysis was performed using the Pearson product moment correlation test. DG, dentate gyrus.

Figure 5. 5-LOX expression is upregulated in the hippocampus of old rats and elderly humans.

(a) 5-LOX mRNA expression was significantly elevated within the neurogenic regions (hippocampus: 2.4-fold; SVZ: 1.8-fold) of old (20 months) compared to young (4 months) rats (n=3 per group). (b) 5-LOX immunoreactivity was strongly increased in the DG of old ('') compared with young (') rats. Although in young rats 5-LOX staining was predominately allocated to granular neurons ('''), in old rats also Iba1⁺ microglia ('''', arrowheads) expressed 5-LOX. Pictures illustrate representative 5-LOX stainings; five rats per group were analysed. (c) In elderly humans (>60 years) (''), intensity of 5-LOX immunostaining in the DG was clearly upregulated compared with young persons (<35 years) (') (lower panels show higher magnifications of the dentate gyri depicted in c',''). Images in c illustrate the DG of a 27 year old human person ('), representing the young age group (hippocampi from five persons<35 years analysed), and the DG of a 67-year-old person (''), representing 5-LOX expression in the elderly group (hippocampi from five persons >60 years analysed). Data are shown as mean±s.d.; ** indicates P<0.01; Student's unpaired t-test (a). Scale bars, 100 μm (b,c). DG, dentate gyrus; HC hippocampus; SVZ, subventricular zone.

Figure 6. GPR17 and CysLTR1, target receptors of montelukast, are expressed within the dentate gyrus of old rats.

(a–f) GPR17 expression pattern in the DG of old (20 months) rats. GPR17 immunoreactivity was localised to cells of the oligodendroglial lineage (Olig2⁺), a), to some DCX⁺ neuronal progenitors (c), to a subset of NeuN⁺ granular neurons (d) and to a fraction of Iba1⁺ microglia (f). GPR17 expression was neither observed in Sox2⁺ neural stem cells (b) nor in GFAP⁺ astrocytes (e). (g–l) CysLTR1 immunoreactivity in the DG of 20-month old rats was localised to a few Sox2⁺ cells (h), to cells of the oligodendroglial lineage (Olig2⁺, g), to GFAP⁺ astrocytes (k), and to some Iba1⁺ microglia (l). CysLTR1 expression was not observed in DCX⁺ neuronal progenitors (i) or in NeuN⁺ granular neurons (j). Scale bars, 50 μm (a–l). Arrowheads indicate receptor co-localisation. DG, dentate gyrus.

Figure 7. GPR17 knockdown and knockout in neurospheres induce hyperproliferation and abolish the effects of montelukast.

(a) GPR17 mRNA expression in mouse neurospheres derived from adult FoxO1,3,4^fl/fl mice and recombined with GPF/Cre retrovirus (FoxO1/3/4^−/− NPCs), compared with neurospheres infected with GFP-only retrovirus (control; CTR NPCs). CysLTR1 mRNA was not expressed in mouse neurospheres and not affected by the FoxO1,3,4 knockout. BV-2 cells served as positive control for CysLTR1 primer specificity. (b) GPR17 immunoreactivity in GFP⁺ cells of FoxO1/3/4^−/− neurospheres compared with control neurospheres. (c) Overexpression of FoxO1 by lentiviral transfection upregulated GPR17 mRNA expression in FoxO1,3,4^−/− neurospheres (FoxO1/3/4^−/− NPCs +FoxO1) to 60%. (d) Proliferative activity of FoxO1,3,4^−/− neurospheres assessed by a neurosphere bulk assay. Cell numbers were 6 × higher in FoxO1/3/4^−/− neurospheres compared with controls. A 7-day montelukast treatment (10 μM) significantly increased cell numbers in control neurospheres, but did not affect FoxO1/3/4^−/− neurospheres. FoxO1 overexpression significantly reduced cell numbers in FoxO1,3,4^−/− neurospheres; Montelukast treatment provoked a significant elevation of cell numbers in the FoxO1 overexpressed FoxO1,3,4^−/− neurospheres. (e) In a single-cell neurosphere assay, neural stem cells with FoxO1/3/4 deletion generated significantly more neurospheres than control cells. A 7-day montelukast treatment (10 μM) significantly elevated the number of neurospheres in control, but not in FoxO1/3/4^−/− cells. FoxO1 overexpression decreased neurosphere numbers in vehicle-treated FoxO1/3/4^−/− NPCs. Montelukast treatment increased the numbers of neurospheres in FoxO1 transfected FoxO1/3/4^−/− NPCs compared with vehicle. (f) In neurospheres isolated from adult GPR17^{−/− GFP} mice, high numbers of GFP⁺ cells are present, indicating active transcription of the GPR17 gene locus in these cells. (g) Optical density (OD) in an MTS assay was 97% increased in GPR17^−/− neurospheres compared with wild-type (WT) neurospheres. A 7-day treatment with 10 μM montelukast significantly elevated OD absorbance in wild-type NPCs, but did not affect GPR17^−/− cells. (h) In the single-cell neurosphere assay, neural stem cells from GPR17^−/− neurospheres generated more neurospheres than control cells (P=0.0561). Montelukast (10 μM, 7 days) significantly elevated neurosphere numbers in control, but not in GPR17^−/− NPCs. Three independent experiments were done in triplicates. Data are shown as mean±s.d. (a,c–e,g,h). * indicates P<0.05, ** indicates P<0.01. Student's unpaired t-test (a) and Two-way ANOVA (c–e,g,h) with Bonferroni post hoc tests were performed. Scale bars (b,f), 50 μm. MTK, montelukast; NPC, neuronal progenitor cells; VEH, Vehicle.