Soluble TREM2 ameliorates tau phosphorylation and cognitive deficits through activating transgelin-2 in Alzheimer's disease

Abstract

Triggering receptor expressed on myeloid cells 2 (TREM2) is a transmembrane protein that is predominantly expressed by microglia in the brain. The proteolytic shedding of TREM2 results in the release of soluble TREM2 (sTREM2), which is increased in the cerebrospinal fluid of patients with Alzheimer's disease (AD). It remains unknown whether sTREM2 regulates the pathogenesis of AD. Here we identified transgelin-2 (TG2) expressed on neurons as the receptor for sTREM2. The microglia-derived sTREM2 binds to TG2, induces RhoA phosphorylation at S188, and deactivates the RhoA-ROCK-GSK3β pathway, ameliorating tau phosphorylation. The sTREM2 (77-89) fragment, which is the minimal active sequence of sTREM2 to activate TG2, mimics the inhibitory effect of sTREM2 on tau phosphorylation. Overexpression of sTREM2 or administration of the active peptide rescues tau pathology and behavioral defects in the tau P301S transgenic mice. Together, these findings demonstrate that the sTREM2-TG2 interaction mediates the cross-talk between microglia and neurons. sTREM2 and its active peptide may be a potential therapeutic intervention for tauopathies including AD.

Fig. 1. sTREM2 attenuates tau hyperphosphorylation in vitro.

Western blots (a) and quantification (b) tau phosphorylation in HEK293-Tau cells treated with Fc, heat-inactivated Fc-sTREM2 (40 nM), or Fc-sTREM2 (40 nM) for 24 h (mean ± s.e.m.; one-way ANOVA, n = 5 independent experiments). Immunostaining (c) and quantification (d) of p-Tau S202 in HEK293-Tau cells treated with Fc (40 nM), heat-inactivated Fc-sTREM2 (40 nM), or Fc-sTREM2 (40 nM) for 24 h. Scale bar, 25 μm (mean ± s.e.m.; one-way ANOVA, n = 10 images from 5 independent experiments). p-Tau S202 immunostaining (e) and quantification (f) of HEK293-Tau cells treated with different concentrations of sTREM2. The control group was treated with Fc. Scale bars, 25 μm (mean ± s.e.m.; one-way ANOVA, n = 10 images from 5 independent experiments; compared with the control group). Western blots (g) and quantification (h) showing the phosphorylation of tau and GSK3β in HEK293-Tau cells treated with different concentrations of sTREM2. The control group (sTREM2 = 0 nM) was treated with Fc (mean ± s.e.m.; one-way ANOVA, n = 4 independent experiments; compared with the control group). p-Tau S202 and p-tau S396 immunostaining (i) and quantification (j) of primary neurons derived from tau P301S mice treated with different concentrations of sTREM2. Scale bar, 20 μm. The fluorescence intensity in the field was normalized to the control group (mean ± s.e.m.; one-way ANOVA, n = 10 images from 5 independent experiments; compared with the control group), *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 2. sTREM2 interacts with TG2.

a Cartoon illustrating the AP-MS workflow. LC-MS/MS, liquid chromatography-tandem mass spectrometry. b MS/MS spectrum showing the identification of TG2. c Volcano plot of neuronal cell membrane proteins of WT mice (left) or tau P301S mice (right) that interact with Fc-sTREM2. The cutoff threshold was set at log₂FC > 2. d The co-localization of neuronal marker (MAP2), TG2, and sTREM2 in the hippocampus of AD patients. Arrows indicate TG2 expression in non-neuronal cells. Scale bars, 10 μm. n = 5 independent experiments. e The co-localization of MAP2, TG2, and sTREM2 in the hippocampus of tau P301S mice. The arrows indicate the colocalization in branches. Scale bars, 10 μm. n = 5 independent experiments. (f) The intensity trace (offset white line) of (e) is potted. g Confocal images of immunostaining for Fc-sTREM2 binding to GFP-TG2. Scale bar, 5 μm. h Quantification of immunostaining in (g). The binding was calculated as the area of Fc immunostaining relative to the cell area (mean ± s.e.m.; Two-tailed t-test, n = 17 images per group; compared with the control group, ***P < 0.001). arb. units, arbitrary units. TG2 is found in the membranous fraction of SH-SY5Y cells (i) and primary neurons of WT mice (j). Pull-down assay using protein A beads indicated that membrane TG2 binds Fc-sTREM2. n = 5 independent experiments. k GST pull-down assay verified the interaction between GST-sTREM2 and His-TG2. n = 5 independent experiments.

Fig. 3. sTREM2 inhibits tau phosphorylation by activating TG2.

a Western blots showing the phosphorylation of tau and GSK3β in HEK293-Tau cells treated with vehicle or TSG12 (25 μM). b Quantification of the immunoreactivity in (a) (mean ± s.e.m.; Two-tailed t-test, n = 4 independent experiments; compared with the control group). c p-Tau S202 immunostaining in HEK293-Tau cells treated with vehicle or TSG12 (mean ± s.e.m.; two-way ANOVA, n = 10 images from 5 independent experiments). Scale bars, 15 μm. d p-Tau S202 immunostaining in vehicle- or TSG12-treated (25 μM) primary neurons derived from tau P301S mice (mean ± s.e.m.; Two-tailed t-test, n = 10 images from 5 independent experiments). Scale bars, 20 μm. e HEK293-Tau cells were pretreated with sh-NC or sh-TG2, and then exposed to Fc-sTREM2 (20 nM) or control (Fc, 20 nM) for 24 h. Shown is the phosphorylation of tau and GSK3β. f Quantification of the immunoreactivity in (e) (mean ± s.e.m.; two-way ANOVA, n = 4 independent experiments). g p-Tau S202 immunostaining and quantification of HEK293-Tau cells pretreated with control shRNA (sh-NC) or sh-TG2 and then exposed to Fc-sTREM2 (20 nM) or control (Fc, 20 nM) for 24 h. Scale bars, 20 μm (mean ± s.e.m.; two-way ANOVA, n = 10 images from 5 independent experiments). h p-Tau S202 immunostaining and quantification of primary neurons derived from tau P301S mice pretreated with control shRNA (sh-NC) or sh-TG2 and then exposed to Fc-sTREM2 (20 nM) or control (Fc, 20 nM) for 24 h. Scale bars, 20 μm. (mean ± s.e.m.; two-way ANOVA, n = 10 images from 5 independent experiments). *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4. sTREM2 attenuates tau phosphorylation by blocking the RhoA/ROCK pathway.

a Schematic illustration of the potential mechanism by which sTREM2 regulates tau phosphorylation: sTREM2 secreted from microglia binds neuronal TG2 and inactivates the RhoA/ROCK/GSK3β pathway, attenuating tau hyperphosphorylation. b, c Western blots showing the phosphorylation of tau and GSK3β in HEK293-Tau cells treated with control (Fc, 20 nM), Fc-sTREM2 (20 nM), or Tat-C3 (RhoA inhibitor, 0.1 μg/mL) for 24 h. d, e Western blots showing the phosphorylation of tau and GSK3β in HEK293-Tau cells treated with control (Fc, 20 nM), Fc-sTREM2 (20 nM), or Y-27632 (ROCK inhibitor) for 24 h. f, g Western blots showing the phosphorylation of tau and GSK3β in HEK293-Tau cells treated with control (Fc, 20 nM), Fc-sTREM2 (20 nM) for 24 h and incubated with RhoA activator (1 U/ml) or solvent for another 30 min. EK293-Tau cells were pretreated with control siRNA (si-NC), si-RhoA (h, i), or si-ROCK (j, k), and then treated with control (Fc, 20 nM) or Fc-sTREM2 (20 nM). Shown are the phosphorylation of tau and GSK3β (mean ± s.e.m.; one-way ANOVA for (c, e, g), two-way ANOVA for (i, k), n = 4 independent experiments; compared with the control group, *P < 0.05, **P < 0.01, ***P < 0.001).

Fig. 5. sTREM2 attenuates tau pathology and memory loss in tau P301S mice.

p-Tau S202/T205 immunostaining (a) and quantification (b) of 7-month-old wild-type mice and tau P301S mice overexpressing AAV-vector or AAV-sTREM2. Scale bar, 250 μm. (mean ± s.e.m.; one-way ANOVA, n = 6 mice per group). c Western blots showing the phosphorylation of tau and GSK3β (mean ± s.e.m.; one-way ANOVA, n = 8 mice per group). d Electron microscopy of synapses in the hippocampus (left). Stars indicate synapses. Scale bar, 1 μm (mean ± s.e.m.; one-way ANOVA, n = 6 mice per group). e Golgi staining revealed the dendritic spines from the apical dendritic layer of the CA1 region (left). Scale bar, 15 μm. Quantification of spine density (right) (mean ± s.e.m.; one-way ANOVA, n = 6 mice per group). Morris water maze analysis of the escape latency (f) and escape latency on day 4 (g) (mean ± s.e.m.; n = 9 mice in WT, P301S-sTREM2 groups, n = 8 mice in P301S-Vector group; one-way ANOVA). h Probe trial of the Morris water maze test (mean ± s.e.m.; n = 8–9 mice per group; *P < 0.05, one-way ANOVA). i Y-maze analysis as time spent in new arms (mean ± s.e.m.; n = 9 mice in WT, P301S-sTREM2 groups, n = 8 mice in P301S-Vector group; one-way ANOVA). j The slope of fEPSPs after HFS recorded in hippocampal slices. The arrow indicates the onset of HFS. k Quantitative analyses of normalized fEPSPs at 60–80 min. Box–whisker plot (displaying the Max/Min at the whiskers, the 75/25 percentiles at the boxes, and the median in the center line). (mean ± s.e.m.; n = 4 mice per group; two-way ANOVA). *P < 0.05, **P < 0.01, ***P < 0.001. AAVs adeno-associated virus, HFS high-frequency stimulation, LTP long-term potentiation.

Fig. 6. Knockdown of TG2 abolishes the inhibitory effect of sTREM2 on tau phosphorylation.

a p-Tau S202/T205 immunostaining and quantification of 7-month-old mice injected with sh-TG2 and sTREM2. Scale bar, 250 μm (mean ± s.e.m.; two-way ANOVA, n = 6 mice per group). b Western blots showing the phosphorylation of tau, GSK3β, and RhoA (mean ± s.e.m.; two-way ANOVA, n = 8 mice per group). c Electron microscopy of synapses in the hippocampus (left). Stars indicate synapses. Scale bar, 1 μm. Quantification of synaptic density (right) (mean ± s.e.m.; two-way ANOVA, n = 6 mice per group). *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 7. A short peptide of sTREM2 suppresses tau hyperphosphorylation in vitro.

a Schematic illustration of sTREM2 fragments. b GST pull-down assay showing the interaction between GST-sTREM2 fragments and His-TG2. n = 4 independent experiments. c, d Western blots showing the phosphorylation of tau and GSK3β in HEK293-Tau cells treated with different sTREM2 peptides (10 μM) for 24 h (mean ± s.e.m.; one-way ANOVA, n = 4 independent experiments; compared with the control group). e p-Tau S202 immunostaining of HEK293-Tau cells treated with different sTREM2 peptides (10 μM) for 24 h. Scale bars, 15 μm. f Quantification of p-tau S202 (mean ± s.e.m.; one-way ANOVA, n = 7 independent experiments; compared with the control group). *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 8. The active sTREM2 peptide suppresses tau pathology in tau P301S mice.

a FITC fluorescence in the hippocampus of tau P301S mice one week after i.p. injection of FITC-labeled sTREM2 (77-89) (Peptide 1). Scale bar, 50 μm. n = 6 mice per group. b p-Tau S202/T205 immunostaining and quantification of tau P301S mice administrated with sTREM2 (89–77) or (77–89). Scale bar, 250 μm (mean ± s.e.m.; n = 6 mice per group; two-tailed t-test). c, d Western blots and quantification showing the phosphorylation of tau and GSK3β (mean ± s.e.m.; n = 8 mice per group; Two-tailed t-test). e Electron microscopy of synapses in the hippocampus. Stars indicate synapses. Scale bar, 1 μm. f Quantification of synaptic density (mean ± s.e.m.; Two-tailed t-test; n = 6 mice per group). Morris water maze analysis of the escape latency (g) and escape latency on day 4 (h) (mean ± s.e.m.; n = 7 mice per group; Two-tailed t-test). i Probe trial of Morris water maze test (mean ± s.e.m.; n = 7 mice per group; Two-tailed t-test). j Y-maze analysis as time spent in new arms (mean ± s.e.m.; n = 7 mice per group; Two-tailed t-test). k The slope of fEPSPs after HFS recorded in hippocampal slices. The arrow indicates the onset of HFS. l Quantitative analyses of normalized fEPSPs at 60–80 min. Box–whisker plot (displaying the Max/Min at the whiskers, the 75/25 percentiles at the boxes, and the median in the center line). (mean ± s.e.m.; n = 4 mice per group; Two-tailed t-test). *P < 0.05, **P < 0.01, ***P < 0.001. AAVs adeno-associated virus, HFS high-frequency stimulation, LTP long-term potentiation.