Astrocyte priming enhances microglial Aβ clearance and is compromised by APOE4

Abstract

The innate immune system can develop a form of memory called priming, where prior exposure to a stimulus enhances subsequent responses. While well-characterized in peripheral immunity, its function in brain-resident cells such as astrocytes under non-disease conditions remains unclear. Here we show that human astrocytes derived from the induced pluripotent stem cells of healthy female donors, but not microglia, acquire a primed state following transient immune stimulations. Upon subsequent exposure to amyloid-β (Aβ), these astrocytes secrete elevated levels of cytokines and promote microglial Aβ uptake. In contrast, astrocytes carrying the Alzheimer's disease (AD) risk allele APOE4 exhibit reduced priming and fail to support microglial phagocytosis. These findings are validated in astrocyte-microglial co-cultures, cerebral organoids, and male mice, where astrocyte priming enhances Aβ clearance in an APOE4-sensitive manner. Our findings identify astrocytic immune memory as a modulator of microglial function and Aβ pathology, providing insights into how early protective responses in AD may be disrupted by genetic risk factors.

Fig. 1. Formation of immune priming in hiPSC-derived astrocytes.

A Schematic overview of the experimental process for inducing and verifying immune priming in hiPSC-derived astrocytes. B Number and ontology of genes that are upregulated by Poly I:C treatment and subsequently downregulated during recovery in human astrocytes. C Patterns of cytokine secretion change by Poly I:C treatment and the subsequent recovery period. D Changes in IL-6, IL-1β, and TNFα transcript levels by Poly I:C treatment, recovery, and second Poly I:C treatment. N = 24, from four independent batches. E Genes with increased reactivity upon second Poly I:C treatment induced by immune priming and their gene ontology. Differentially expressed genes (DEGs) were identified using Cuffdiff. Statistical significance was determined by adjusted q-values (<0.05). F Schematic overview of the experimental process for inducing and verifying immune priming in iMGL. G Changes in IL-6, IL-1β, and TNFα transcript levels by Poly I:C treatment, recovery, and second Poly I:C treatment in iMGL. N = 4, from two independent batches. All schematics were created in BioRender. Seo, J. (2025) https://BioRender.com/09a0byv. *p < 0.05, ****p < 0.0001; ns, not significant. (One-way ANOVA test followed by Tukey’s post hoc analysis). Exact P values are provided in the Source Data. Error bar ± S.E.M.

Fig. 2. Primed human astrocytes amplify inflammatory signals and boost microglial Aβ₄₂ phagocytosis.

A Schematic overview of the experimental process exploring the effects of priming on human astrocyte reactivity to Aβ₄₂. B Decreased Aβ₄₂ uptake capability in primed astrocytes. N = 11, from three independent batches. C Changes in cytokine secretion patterns in primed astrocytes compared to non-primed astrocytes upon Aβ₄₂ treatment. Cytokines that were significantly increased or decreased are indicated in red and blue, respectively. In each experiment, three samples are pooled, and the entire experiment is repeated three times using cells derived from independent batches. D Increased levels of IL-6 and IL-1β transcripts in primed astrocytes in response to Aβ₄₂. N = 12, from three independent batches. E Ontology analysis of genes uniquely upregulated by Aβ₄₂ in primed astrocytes compared to non-primed astrocytes. F Ontology analysis of genes uniquely downregulated by Aβ₄₂ in primed astrocytes compared to non-primed astrocytes. Differentially expressed genes (DEGs) were identified using Cuffdiff. Statistical significance was determined by adjusted q-values (<0.05). G Schematic overview of the experimental process analyzing the impact of cytokines secreted by primed astrocytes in response to Aβ₄₂ on microglial Aβ₄₂ uptake. H Increased microglial Aβ₄₂ uptake in response to astrocyte-conditioned media (ACM) from primed astrocytes exposed to Aβ₄₂. N = 20, from two independent batches. Scale bar = 5 μm. All schematics were created in BioRender. Seo, J. (2025) https://BioRender.com/09a0byv. *p < 0.05, **p < 0.01, ****p < 0.0001; ns not significant. (One-way ANOVA test followed by Tukey’s post hoc analysis or two-tailed Student’s t-test). Exact P values are provided in the Source Data. Error bar ± S.E.M.

Fig. 3. APOE4 disrupts astrocyte priming and its ability to promote microglial Aβ₄₂ phagocytosis.

A Number and ontology of genes upregulated by Poly I:C treatment and subsequently downregulated during recovery in APOE4 hiPSC-derived astrocytes. B Changes in IL-6, IL-1β, and TNFα transcript levels by Poly I:C treatment, recovery, and second Poly I:C treatment. N = 24, from four independent batches. C Genes with increased transcripts upon second Poly I:C treatment induced by immune priming and their gene ontology. Differentially expressed genes (DEGs) were identified using Cuffdiff. Statistical significance was determined by adjusted q-values (<0.05). D Reduced increase in IL-6, IL-1β, and TNFα transcript levels in primed APOE4 astrocytes compared to APOE3 astrocytes by the second Poly I:C challenge. N = 24 (IL-6 and TNFα) and N = 22 (IL-1β), derived from four independent batches. E Changes in cytokine secretion patterns in primed APOE4 astrocytes compared to non-primed astrocytes. Cytokines that were significantly increased or decreased are indicated in red and blue, respectively. In each experiment, three samples are pooled, and the entire experiment is repeated three times using cells derived from independent batches. F Increased IL-1β transcript levels in primed APOE4 astrocytes in response to Aβ₄₂. N = 6, from three independent batches. G Ontology analysis of genes uniquely upregulated by Aβ₄₂ in primed APOE4 astrocytes compared to non-primed astrocytes. H Ontology analysis of genes uniquely downregulated by Aβ₄₂ in primed APOE4 astrocytes compared to non-primed astrocytes. DEGs were identified using Cuffdiff. Statistical significance was determined by adjusted q-values (<0.05). I Increased Aβ₄₂ uptake capability in primed APOE4 astrocytes compared to non-primed astrocytes. N = 11, from three independent batches. J Decreased microglial Aβ₄₂ uptake in response to astrocyte-conditioned media (ACM) from primed APOE4 astrocytes exposed to Aβ₄₂. N = 20, from two independent batches. Scale bar=5 μm. **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant. (One-way ANOVA test followed by Tukey’s post hoc analysis or two-tailed Student’s t-test). Exact P values are provided in the Source Data. Error bar ± S.E.M.

Fig. 4. Verification of the contribution of astrocyte priming to enhanced microglial Aβ₄₂ clearance using astrocyte-microglia co-culture and glia-embedded AD cerebral organoids.

A Schematic overview of the experimental process for the astrocyte-microglia co-culture system. B, C After applying the priming protocol, increased Aβ₄₂ clearance and enhanced Iba1 signal are observed in APOE3 microglia when co-cultured with APOE3 astrocytes, but not with APOE4 astrocytes. Scale bar=5 μm. D Schematic overview of the experimental process for APPswe organoids embedded with APOE3 astrocytes. E Organoids embedded with APOE3 astrocytes contain only GFAP-positive cells, with no Iba1 signal. Similar glial cell embedding was confirmed across all organoids used for the analysis. Scale bar=50 μm. F Priming does not reduce accumulated Aβ₄₂ in organoids embedded solely with astrocytes. N = 12, from four different batches. Scale bar=50 μm. G Schematic overview of the experimental process for APPswe organoids embedded with APOE3 microglia. H Organoids embedded with APOE3 microglia contain only Iba1-positive cells, with no GFAP signal. Similar glial cell embedding was confirmed across all organoids used for the analysis. Scale bar=50 μm. I Priming does not reduce accumulated Aβ₄₂ in organoids embedded solely with microglia. N = 12, from four different batches. Scale bar = 50 μm. J Schematic overview of the experimental process for APPswe organoids embedded with both APOE3 astrocytes and microglia. K Organoids embedded with both cell types contain GFAP-positive and Iba1-positive cells. Similar glial cell embedding was confirmed across all organoids used for the analysis. Scale bar=50 μm. L Priming reduces accumulated Aβ₄₂ in organoids embedded with both astrocytes and microglia. N = 12, from four different batches. Scale bar=50 μm. All schematics were created in BioRender. Seo, J. (2025) https://BioRender.com/09a0byv. *p < 0.05, ***p < 0.001; ns not significant. (Two-tailed Student’s t-test). Exact P values are provided in the Source Data. Error bar ± S.E.M. test.

Fig. 5. Enhanced Aβ₄₂ clearance by astrocyte priming in an in vivo mouse model and its inhibition by APOE4 isoform.

A Schematic overview of the experimental process for inducing and verifying astrocyte priming in an in vivo mouse model. B Confirmation of increased IL-6, IL-1β, and TNFα transcript levels upon second Poly I:C treatment induced by in vivo astrocyte priming. N = 8 (IL-6 and TNFα) and N = 7 (IL-1β) animals. C Schematic overview of the experimental process using a humanized APOE knock-in (KI) mouse model to verify the changes in priming effects due to APOE4. D Increased Aβ₄₂ clearance by in vivo priming in APOE3 KI mice. N = 26 (non-primed group) and N = 32 (primed group), from four animals. Scale bar=200 μm. E Reduced Aβ₄₂ clearance by in vivo priming in APOE4 KI mice. N = 40 (non-primed group) and N = 32 (primed group), from four animals. Scale bar=200 μm. F–H Microglial activation is assessed using Iba1 and CD11b staining. G Microglia in primed APOE3 KI mice exhibit a more activated morphology, characterized by a reduced number of intersections when Aβ₄₂ is injected, whereas microglia in primed APOE4 KI mice show no such changes with Aβ₄₂ injection. N = 11, from four animals. Scale bar=20 μm. H Increased CD11b intensity in primed APOE3 KI mice when Aβ₄₂ is injected, but microglia in APOE4 KI mice show decreased CD11b intensity with Aβ₄₂ injection. N = 20, from four animals. Scale bar=20 μm. I Increased GFAP intensity in primed APOE3 KI mice when Aβ₄₂ is injected, but microglia in APOE4 KI mice show decreased GFAP intensity with Aβ₄₂ injection. N = 23, from four animals. Scale bar=20 μm. All schematics were created in BioRender. Seo, J. (2025) https://BioRender.com/09a0byv. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns, not significant. (Two-tailed Student’s t-test). Exact P values are provided in the Source Data. Error bar ± S.E.M.