Inflammasome signaling is dispensable for ß-amyloid-induced neuropathology in preclinical models of Alzheimer’s disease

Background Alzheimer’s disease (AD) is the most common neurodegenerative disorder affecting memory and cognition. The disease is accompanied by an abnormal deposition of ß-amyloid plaques in the brain that contributes to neurodegeneration and is known to induce glial inflammation. Studies in the APP/PS1 mouse model of ß-amyloid-induced neuropathology have suggested a role for inflammasome activation in ß-amyloid-induced neuroinflammation and neuropathology. Methods Here, we evaluated the in vivo role of microglia-selective and full body inflammasome signalling in several mouse models of ß-amyloid-induced AD neuropathology. Results Microglia-specific deletion of the inflammasome regulator A20 and inflammasome effector protease caspase-1 in the AppNL-G-F and APP/PS1 models failed to identify a prominent role for microglial inflammasome signalling in ß-amyloid-induced neuropathology. Moreover, global inflammasome inactivation through respectively full body deletion of caspases 1 and 11 in AppNL-G-F mice and Nlrp3 deletion in APP/PS1 mice also failed to modulate amyloid pathology and disease progression. In agreement, single-cell RNA sequencing did not reveal an important role for Nlrp3 signalling in driving microglial activation and the transition into disease-associated states, both during homeostasis and upon amyloid pathology. Conclusion Collectively, these results question a generalizable role for inflammasome activation in preclinical amyloid-only models of neuroinflammation.


Introduction
Neurodegeneration is a complex, multifaceted process that degrades neuronal structure and function, ultimately leading to cognitive and/or motor disability and dementia (1).Alzheimer's disease (AD) is the most common form of neurodegenerative dementia, and currently affects more than 30 million people over the age of 65 worldwide (2).Neuronal damage and subsequent neurodegeneration in AD are associated with the accumulation of b-amyloid (Ab) in extracellular plaques and the intracellular aggregation of hyperphosphorylated tau in neurofibrillary tangles (2).For a long time the focus has been on cell-autonomous processes primarily affecting neurons, whereas glial involvement is progressively taking center stage as an important mediator of neurodegenerative pathology in AD (3).
Microglia are CNS resident macrophages that are distributed across the brain parenchyma (4)(5)(6).When activated by damageassociated or pathogen-associated molecular patterns (DAMPs and PAMPs, respectively), microglia rapidly move toward the site of injury to initiate an innate immune response in order to cope with the insult (6,7).In AD, Ab oligomers that are generated by cleavage of amyloid precursor protein (APP) and amyloid fibrils in the amyloid plaques, bind various microglial cell-surface receptors, which prompt microglial activation and release of inflammatory mediators (8)(9)(10).Some of the most robust evidence for microglial involvement in AD came from the identification of AD risk genes in genome-wide association studies (11)(12)(13)(14)(15)(16)(17)(18)(19), suggesting that microglia are not merely bystanders, but possibly causally involved in AD pathogenesis.However, the role of specific inflammatory pathways and mechanisms of microglial activation in Ab-induced neuroinflammation and neurodegenerative pathology require further analysis.
The NF-kB signaling pathway critically regulates inflammatory responses and has been linked to the pathogenesis of several neurodegenerative disorders (20)(21)(22).Though baseline activity is very low in glia, NF-kB activity is significantly increased in glia surrounding Ab plaques (23).Increased levels of the NF-kB p65 subunit are reported in cortical neurons and glia of AD patients (24).This in turn upregulates transcription of pro-inflammatory target genes that promote neuroinflammation (25), as well as that of BACE1, which promotes APP cleavage to generate additional Ab (26).NF-kB activation is also a prerequisite for priming the NLR family pyrin domain containing 3 (NLRP3) inflammasome, a cytosolic multimeric protein complex that recruits and activates procaspase-1, a protease that subsequently cleaves and releases the pro-inflammatory cytokines interleukin (IL)-1b and IL-18 and drives an inflammatory form of cell death known as pyroptosis by cleaving gasdermin D (GSDMD) (27).Increased levels of NLRP3, caspase-1, IL-1b, IL-18, the inflammasome adaptor protein ASC and cleaved GSDMD have been reported in reactive microglia that surround Ab plaques in brain of AD patients (28)(29)(30)(31).Moreover, in vitro studies have demonstrated that fibrillar Ab stimulates microglia activation and IL-1b production via assembly of the NLRP3 inflammasome and caspase-1 activation (32).Finally, an in vivo role for the NLRP3/caspase-1 inflammasome axis was subsequently proposed in the APP/PS1 model of Ab-induced neuropathology (33).However, the relative significance of microglial NLRP3 inflammasome activation in additional in vivo models of Ab-induced neuropathology has not been explored.
Here, we investigated the role of microglial and full body inflammasome involvement in the App NL-G-F and APP/PS1 mouse models of ß-amyloid-induced neuropathology.APP/PS1 mice ectopically overexpress the KM670/671NL 'Swedish' mutated amyloid precursor protein (APP) concomitant with mutant human PS1 (presenilin-1) in CNS neurons, and develop an early and robust AD pathology (34).In contrast, App NL-G-F mice express humanized sequences and clinical mutations in the endogenous mouse App gene locus such that Ab42 is produced without APP overexpression (35).As a result, Ab-induced pathophysiological readouts develop more slowly in App NL-G-F mice compared to aggressive disease progression in APP transgenic mice (36).First, we hypothesized that microglial deletion of A20, a negative regulator of NF-kB signaling that controls microglia inflammasome activation and CNS inflammation (37), may exacerbate ß-amyloid-induced neuroinflammation and AD outcomes (Supplementary Figure 1A).Secondly, we tested the assumption that microglial deletion of the central inflammasome effector protease caspase-1 may suppress microglia activation, neuroinflammation and amyloid pathology (Supplementary Figure 1B).Unexpectedly, our results in the App NL-G-F and APP/ PS1 mouse models argue against a significant role for microglial inflammasome activation in Aß-associated neuroinflammation and neurodegenerative pathology.Further supporting these hypotheses, we found that full body deletion of caspase-1 in App NL-G-F mice (Supplementary Figure 1B) and full body Nlrp3 deletion in APP/ PS1 mice (Supplementary Figure 1C) both failed to inhibit ßamyloid-induced gliosis, inflammation and plaque burden in diseased mice.Consistent herewith, single-cell gene expression profiling revealed only minor transcriptional changes in microglia from Nlrp3-deficient APP/PS1 mice, negating inflammasome signaling as a driver of microglial cell states during amyloid pathology.Taken together, these findings question a generalizable role for inflammasome activation in mouse models of ß-amyloidinduced neuroinflammation and pathology.

Results
Microglial A20 deletion promotes CNS hyperinflammation but does not increase ß-amyloid-induced pathology in App NL-G-F mice A20 acts as a key inhibitor of NF-kB signaling (38), and its deficiency leads to increased expression of inflammatory mediators in microglia (37,39).Additionally, previous research has demonstrated that A20 acts as a negative regulator in priming and activation of the Nlrp3 inflammasome in microglia (37).To investigate the impact of microglial NF-kB and Nlrp3 inflammasome hyperactivation in Aßassociated neuroinflammation, we crossed mice carrying floxed alleles for A20 (40) to Cx3cr1 CreErt2 knock-in mice to allow Cre recombinase-mediated long-term A20 deletion in microglia following tamoxifen (TAM) treatment (41).Control mice (A20 FL , expressing A20) and mice with tamoxifen-inducible A20 deficiency in microglia (A20 FL/FL Cx3cr1 CreErt2 , hereafter A20 Cx3cr1-KO ) were first crossed to App NL-G-F animals expressing mutant APP from the endogenous locus (35).At 4-6 weeks of age, the progeny was subcutaneously injected with TAM to activate Cre recombinase and generate A20 FL App NL-G-F mice and A20 Cx3Cr1-KO App NL-G-F mice which do (A20 FL ) or do not (A20 CxcCr1-KO ) express A20 in microglia in an amyloid model of AD pathology.These mice were compared with age-matched App WT mice to accurately assign differences emerging from the mutant APP genetic background or A20 genotype, and possible interaction effects (Supplementary Figures 2A-C).
Successful downregulation of A20 in microglia was confirmed at the protein level by Western blotting of ex vivo isolated and fluorescence-activated cell sorted (FACS) microglia (Supplementary Figure 2D), and of primary microglia cultured in vitro (Supplementary Figure 2E).Consistent with enhanced NF-kB activation in A20 deficient conditions, qPCR analysis confirmed that A20 deletion results in increased expression of the proinflammatory cytokines and chemokines Il1b, Il6, Tnf, Ccl2 and Figures 3A-B).This suggests that A20 deletion in microglia promotes inflammatory cytokine production regardless of the mutant APP allele.
We next assessed brain pathology in A20-deficient App NL-G-F animals at the histological level.Brain sections of A20 FL App NL-G-F and A20 Cx3Cr1-KO App NL-G-F mice, as well as age-matched App WT controls were evaluated for differences in amyloid plaque burden, microgliosis and astrogliosis, as well as for differences in plaqueassociated axonal damage.As expected, brains of App WT control animals were devoid of amyloid deposits, whereas amyloid plaques were readily visualized by anti-Ab antibody staining in brain sections of App NL-G-F animals (Figures 1A, B and Supplementary Figure 4).However, microglial deletion of A20 did not significantly alter amyloid plaque deposition in brains of young (20 weeks of age, Figures 1A, B and Supplementary Figures 4A-F) or aged (56 weeks of age, Figures 1C, D and Supplementary Figures 4G-L) App NL-G-F animals, suggesting that hyperactive NF-kB and inflammasome signaling by long-term A20 deletion in microglia may not be a critical modulator of Ab plaque formation in App NL-G-F animals.
Chronic Ab deposition in brain parenchyma also induces chronic glial activation, characterized by higher microglia proliferation and changes in microglia morphology (42).Moreover, microglial A20 deletion is associated with a significant increase in microglia proliferation in non-AD mouse models of neuroinflammation (37).In agreement, we observed an increased number of hippocampal Iba1 + microglia in brains of 20 week-old A20 Cx3Cr1-KO mice compared to A20 FL mice, which was even more pronounced in the App NL-G-F genetic background (Figures 1E, F and Supplementary Figures 5A, B).Aged (56 week-old) App NL-G-F mice displayed increased levels of hippocampal microglia proliferation compared to mice on the App WT background, regardless whether they had A20 FL or A20 Cx3Cr1-KO alleles, confirming that the mutant APP genetic background on itself boosts microglial proliferation (Figures 1G, H and Supplementary Figures 5C, D).
In contrast to the microgliosis, hippocampal astrocyte numbers were unaffected by microglial deletion of A20.The number of GFAP + hippocampal astrocytes was significantly increased in young App NL-G-F mice compared to App WT mice, which likely was a consequence of early Ab-induced brain pathology (Figures 1I, J and Supplementary Figures 5E, F).However, the number of GFAP+ hippocampal astrocytes reached similar levels across all experimental groups at older age (Figures 1K, L and Supplementary Figures 5G, H).
To infer plaque-associated axonal damage, we quantified dystrophic neural projections in the vicinity of the plaque by immune co-staining for N/25 and sAPP (Figures 2M-P).N/25 targets the Ab 1-7 epitope with high affinity and provides a global overview of amyloid deposition, while sAPP targets the amyloid precursor protein (APP).Early stages of Ab-induced brain pathology involve axonal abnormalities, possibly associated with an atypical accumulation of APP and its cleavage products (43,44).Microglial deletion of A20 on the App NL-G-F background did not induce different staining patterns or differences in the number of plaque-associated dystrophic neurites, suggesting that chronic microglial reactivity and neuroinflammation do not exacerbate the abnormal accumulation of full-length APP and Ab at nerve terminals in App NL-G-F mice (Figures 1M-P and Supplementary Figures 5I, J).
Together, these results demonstrate that microglial deletion of A20 results in a condition of chronic hyperinflammation in the

CNS without modulating Ab pathology and axonal damage in the
App NL-G-F mouse model of Ab-induced brain pathology.

Microglial caspase-1 deficiency does not suppress b-amyloid-induced pathology in
App NL-G-F mice To more directly assess the role of microglial inflammasome activation in Ab-induced brain pathology, we next crossed mice carrying a floxed caspase-1 allele (45) to Cx3cr1 CreErt2 mice (Supplementary Figure 6A).Control mice (casp1 FL , expressing caspase-1) and mice with TAM-inducible caspase-1 deficiency in microglia (hereafter named casp1 Cx3Cr1-KO ) were subsequently crossed to App NL-G-F to examine the role of microglial caspase-1 activation in Ab-induced pathology (Supplementary Figure 6B).Ab-induced brain pathology was analyzed at the age of 20, 40 and 70 weeks with wild-type mice (App WT ) included as negative controls (Supplementary Figure 6C).First, successful caspase-1 downregulation in microglia was confirmed at the protein level by Western blotting of FACS-sorted microglia ex vivo (Supplementary Figure 6D) and in primary microglia cultured in vitro (Supplementary Figure 6E).
As expected, amyloid plaques were readily observed in 20 weekold App NL-G-F mice (Figures 2A, B and Supplementary Figures 7A-F).Microglial caspase-1 deletion had no significant impact on amyloid plaque load (Figure 2A and Supplementary Figures 7A-F) or microglia numbers (Figures 2C, D and Supplementary Figures 7G, H) in App NL-G-F mice compared to that of caspase-1sufficient App NL-G-F mice.Astrocyte numbers did not vary significantly across APP WT and App NL-G-F mice, irrespective of caspase-1 expression (Figures 2E, F and Supplementary Figures 7I, J).
No significant impact of microglial caspase-1 deletion could be observed in sAPP-N/25 immunoreactivity (Figures 2G, H and Supplementary Figure 7K), excluding an effect of caspase-1 activity on the extent of plaque-associated axonal damage.Finally, we analyzed potential differences in the number of plaqueassociated microglia in App NL-G-F mice by co-staining for Iba1 and the fluorescent dye PFTAA (pentameric formyl thiophene acetic acid) that stains Ab plaques.However, the number of plaque-associated microglia was unaltered upon caspase-1 deficiency (Figure 2I).
We speculated that the lack of effect of microglial caspase-1 deletion on Ab-induced brain pathology might be explained by the amount of plaque load not having achieved a "critical level" in 20 week-old App NL-G-F mice.To test this hypothesis, we next assessed 40 week-old groups of casp1 FL App NL-G-F and casp1 Cx3Cr1-KO App NL-G-F mice for differences in Ab-induced brain pathology.As App NL-G-F mice develop maximal plaque saturation by 7 months of age (35), this time point represents a late stage of Ab-induced pathology.However, our analyses did not reveal significant differences in brain-wide amyloid plaque load when comparing casp1 FL App NL-G-F animals and casp1 Cx3Cr1-KO App NL-G-F animals (Figures 2J, K and Supplementary Figures 8A-F), nor differences in hippocampal microglia numbers (Figures 2L, M and Supplementary Figures 8G,   H), or plaque and microglia phenotypes (Figures 2J-M and Supplementary Figures 8A-H).Also no differences were observed in hippocampal astrocyte numbers (Figures 2N, O and Supplementary Figures 8I, J), nor did the degree of plaqueassociated axonal damage change significantly (Figures 2P, Q and Supplementary Figure 8K).In addition, we performed immunohistochemistry for Iba1, TMEM119 and Methoxy-04 (a fluorescent amyloid marker) to assess potential differences in the number of plaque-associated microglia in App NL-G-F cortical samples.However, no significant differences were observed between caspase-1sufficient and caspase-1-deficient microglia (Supplementary Figures 8L, M).Microglial caspase-1 deletion did also not affect hippocampal or cortical Ab 42 loads in the casp1 FL App NL-G-F and casp1 Cx3Cr1-KO App NL-G-F groups (Supplementary Figure 8N).Finally, we assessed whole-brain plaque loads in very old (70 weeks) App NL-G- F mice, but again failed to identify significant differences between casp1 FL and casp1 Cx3Cr1-KO animals in relation to Ab-plaque load (Figure 2R).
Taken together, these results demonstrate that microglial caspase 1 deletion has no significant impact on Ab-induced brain pathology in the App NL-G-F mouse model of AD.

Full-body inflammasome inhibition does not suppress ß-amyloid pathology in
App NL-G-F mice Based on our unexpected findings with inducible microgliaspecific caspase-1 deficient mice, we next evaluated whether constitutive full body caspase-1 deletion may display a more prominent impact on Ab-induced brain pathology in the App NL- G-F model.To address potential roles of both canonical and noncanonical inflammasome pathways, we bred full-body caspase-1 knockout mice (46,47) that are also deficient in caspase-11 (47) (casp1/11 KO ) to App NL-G-F mice, and assessed Ab load and neuroinflammation in 12-month-old mice.Similar to our results with microglia-specific caspase-1-deficient mice, constitutive full body deletion of the central inflammasome effector proteases (Casp1/11 KO App NL-G-F mice) did not result in significant differences in Ab plaque load, microgliosis, astrogliosis or in the number of plaque-associated microglia, compared to caspase-1/11sufficient App NL-G-F mice (Figures 3A-C).Consistent herewith, no differences were observed in cortical Ab 40 and Ab 42 load between the two groups (Figure 3D).

Microglial caspase-1 deletion and full-body
Nlrp3 deletion fail to suppress ß-amyloid pathology in APP/PS1 mice Our results suggest that inflammasome activation is dispensable for Ab-induced brain pathology in App NL-G-F mice with inducible microglia-selective caspase-1 deletion, and we extended these findings to App NL-G-F mice with a constitutive systemic deletion of the central inflammasome effector proteases caspases-1 and -11.We reasoned that defective inflammasome activation in the App NL-G-F model may have failed to recapitulate the decreased gliosis and plaque burden outcomes reported in APP/PS1 mice with full-body deletion of caspase-1 or Nlrp3 (33) because APP/PS1 mice overexpress APP, which is a nonphysiological yet characteristic feature of that model (36).To empirically test this hypothesis, we next bred casp1 Cx3Cr1-KO mice to APP/PS1 mice and examined the role of microglial caspase-1 deletion in the Ab-induced pathology of APP/PS1 mutant mice.Unexpectedly, microglial caspase-1 deletion failed to significantly modulate Ab plaque load in 20 week-old APP/PS1 mice (Figures 4A, B).A significant, but minor effect was seen on microgliosis upon caspase-1-deficiency (Figures 4C,  D).However, astrogliosis, plaque-associated axonal damage and the number of plaque-associated microglia were not significantly affected (Figures 4E-J).Also in aged (70 weeks old) APP/PS1 mice, no significant difference in Ab pathology was observed between casp1 FL APP/PS1 and casp1 Cx3Cr1-KO APP/PS1 mice (Figure 4K).Thus, similar to our previous observations in App NL-G-F mice (Figures 1-3), these results demonstrate that Ab-induced brain pathology develops largely independently of microglial infl ammasome activation in the APP/PS1 mouse model.To further expand our observations to APP/PS1 mice with a systemic defect in NLRP3 inflammasome signaling, we finally evaluated the impact of constitutive full body Nlrp3 deletion (Nlrp3 KO ) on b-amyloid pathology by breeding Nlrp3 KO mice to APP/PS1 mice.When assessing amyloid plaque deposition and neuroinflammation in 4, 6 and 10 month-old mutant mice, we found no significant differences in hippocampal or cortical Ab plaque loads when comparing Nlrp3-sufficient and Nlrp3deficient APP/PS1 mice (Figures 5A-F and Supplementary Figures 9A-F).Microgliosis was observed from early age onwards in APP/PS1 mice, but Nlrp3 deletion did not modulate microglia numbers in the hippocampus or cortex (Figures 5G-L and Supplementary Figures 9G-L).At young age (4 months), we observed a trend toward a reduced number of dystrophic neurites in Nlrp3 KO APP/PS1 mice, although it did not reach statistical significance.Furthermore, no differences between the two genotypes were observed at older age in this regard (Figures 5M-R).We also quantified neuronal counts by immunohistochemistry using the NeuN antibody, but again observed no clear differences between Nlrp3-deficient and Nlrp3-sufficient APP/PS1 mice (Supplementary Figures 10A-D).Finally, Ab40 and Ab42 levels were also similar in the two genotypes (Supplementary Figures 10E, F).Contrasting a previous report (33), these results show that full-body Nlrp3 deletion has little measurable impact on Ab-induced neuropathology in APP/PS1 mice.
Nlrp3 signaling does not shape microglial activation during homeostasis or upon amyloid pathology in APP/PS1 mice Single-cell RNA sequencing (scRNA-seq) approaches have revealed the nature of microglial responses to AD-associated amyloid pathology (48)(49)(50).Previous studies have identified a disease-associated microglia (DAM) signature upon amyloid pathology, with DAMs exhibiting a downregulation of homeostatic signature genes and an induction of genes associated with phagocytosis and lipid metabolism (48)(49)(50).
To investigate whether microglial activation in the context of amyloid pathology is shaped by Nlrp3 signaling, we performed scRNA-seq on CD45 + brain immune cells isolated from 4 groups of mice: wild-type (WT), Nlrp3 KO , APP/PS1 and Nlrp3 KO APP/PS1 mice.Furthermore, mice were profiled at the age of 3, 6 and 9 months for WT and Nlrp3 KO mice, and at the age of 3 and 9 months for the APP/PS1 and the Nlrp3 KO APP/PS1 groups.In each of the 10 individual scRNA-seq datasets, microglia were identified based on known signature genes (50) and were subsequently pooled in a single dataset (Figure 6A).In accordance with previous studies (48-50), we were able to identify microglial cell states that correspond to homeostatic microglia (HM), DAMs, microglia transitioning toward DAMs (TM), microglia with an IFN-signature (interferon-response microglia or IRM), proliferating microglia (PM) and a cluster exhibiting an immediate-early gene (IEG) Nlrp3 signaling does not shape microglial activation during homeostasis or upon amyloid pathology in APP/PS1 mice.11A, B).Remarkably, the relative proportion of the different microglial cell states, including DAMs, was comparable between Nlrp3 WT and Nlrp3 KO mice, both in the non-diseased (Supplementary Figures 11A, B) and APP/PS1 backgrounds (Figures 6C, D).Furthermore, in healthy non-diseased brains, microglia from Nlrp3 sufficient and Nlrp3 deficient mice showed only a few differentially expressed genes, across all age groups (Supplementary Figures 11C-E).The only significantly upregulated gene in Nlrp3 KO microglia was the predicted pseudogene Gm8797.The strongest downregulated gene was Nlrp3, which was completely absent in the Nlrp3 KO groups (Figure 6E), confirming its complete deletion.Similarly, only a limited number of genes were differentially expressed between Nlrp3 WT and KO microglia within the APP/PS1 background both at 3 and 9 months of age (Figure 6F; Supplementary Figure 12A), and the top differentially expressed genes included multiple pseudogenes, predicted genes and ribosomal genes (e.g.Gm10076, Gm26510, Rps29, Rps28).The same was observed for DAMs from Nlrp3 WT and Nlrp3 KO APP/PS1 mice (Figure 6G).We also did not identify a robust NF-kB (Rela, Nfkbia, Relb, Ccl2, Ccl5, Il6, Tnfaip3, Tnf, Il1b) or inflammasome signature (Nlrp3, Nlrp1b, Pycard, Casp1, Casp4, Gsdmd, Il1b, Il18) in DAMs or other microglial clusters in APP/PS1 mice, and these genes were not altered upon Nlrp3 deficiency (Supplementary Figure 12B).Overall, these single cell transcriptomics data suggest that inflammasome activation is not a central regulator of microglial activation in the healthy brain or upon amyloid pathology in the APP/PS1 model.

Discussion
Stimulation of microglia with fibrillar Ab leads to a hyperinflammatory response that is characterized by the upregulation and activation of the Nlrp3 inflammasome (32).This activation subsequently triggers the production and secretion of pro-inflammatory cytokines IL-1b and IL-18 (32).Activation of the microglial Nlrp3 inflammasome has been proposed as a key contributor to the progression of AD pathology in the APP/PS1 and CRND8-APP models of Ab-induced neuropathy (33, 52).APP/PS1 mice are double transgenic mice that express a mutant amyloid precursor protein (Mo/HuAPP695swe or APP-K670N/M671L) together with a mutant human presenilin-1 (exon 9-deleted PS1 or PS1-L166P) (34,53), while CRND8 mice transgenically express a mutant APP incorporating the double Swedish (K670N/M671L) and Indiana (V717F) mutations (54).These mutations are all associated with early-onset Alzheimer's disease.
To further investigate the role of the Nlrp3 inflammasome in a different model of Ab-induced neurodegeneration, we first conducted experiments in App NL-G-F mice with a microgliaselective deficiency in A20, an anti-inflammatory protein that suppresses Nlrp3 inflammasome activation (Supplementary Figure 1A).We also examined the effect of microglial deficiency in caspase-1, the central effector protease of the Nlrp3 inflammasome (Supplementary Figure 1B).Notably, our approach utilizing the Cx3Cr1 CreERT2 deletion strategy allows for selective targeting of long-lived brain macrophages, while excluding chronic gene deletion in peripheral myeloid cells (41).Our findings demonstrated that microglial inflammasome activation plays a negligible role in Ab-induced neuroinflammation and brain pathology in App NL-G-F mice.This conclusion is further supported by our investigation into the potential consequences of constitutive full-body deletion of caspases-1 and -11 in App NL-G-F mice (Supplementary Figure 1B), which provided evidence that constitutively impaired inflammasome signaling in all cell types of the body has minimal impact on the development of Ab-induced brain pathology in diseased App NL-G-F mice.
To corroborate our findings and further explore a potential role of microglial inflammasome activation in other mouse models of Ab-induced neuropathology, we extended our studies to the widely used APP/PS1 mouse model of AD.Unexpectedly, our experiments involving the selective inactivation of caspase-1 in microglia of APP/PS1 mice (Supplementary Figure 1B) yielded similar results to what was observed for App NL-G-F mice.Moreover, our results revealed that full body constitutive deletion of Nlrp3 in APP/PS1 mice (Supplementary Figure 1C) did not mitigate the development of Ab-induced neuropathology.These findings further strengthened our conclusion that inflammasome activation has a negligible effect on Ab-induced neuropathology, and provide a strong rationale for questioning a generalizable critical role of inflammasome activation in mouse models of ß-amyloid-induced neuroinflammation and AD-associated amyloid pathology.
Finally, by conducting a comparative analysis of microglia using scRNASeq in Nlrp3 knockout and wild-type APP/PS1 mice, we revealed that the relative proportions of microglial cell states, including disease-associated microglia (DAMs), were similar between Nlrp3-sufficient and Nlrp3-deficient conditions.This intriguing finding suggests that the absence of Nlrp3 and subsequent inflammasome activation does not significantly impact the distribution of microglial cell states in response to Abassociated pathology.
The role of inflammasome activation in the pathogenesis of AD is an area of active research and ongoing debate.It is clear that Ab fibrils activate the Nlrp3 inflammasome in microglia (32), and several reports have identified activated inflammasome effectorsin particular caspase-1 and IL-1ß -in the brains of AD patients and in animal models of AD-associated neuropathology (30, 31, 33, 52, 55-61).However, other published findings suggest a more complex involvement of IL-1ß secretion in AD pathology.Sustained transgenic overexpression of IL-1ß was shown to have a beneficial effect on amyloid burden without overt neurodegeneration in the APP/PS1 and APPswe/PS-1dE9 mouse models of Ab-associated neuropathy (62, 63).IL-1 Receptor (IL-1R1) deficiency also did not appear to modulate Ab loads in aging APP-transgenic (Tg2576) mice (64).Moreover, IL-1ß-induced neuroinflammation was shown to regulate Ab and tau pathology in opposing ways in the triple transgenic mouse model of AD (65).
The work presented here suggests that the Nlrp3 inflammasome is not critically involved in shaping microglial activation states or in driving Ab-induced neurodegeneration in preclinical models of Ab-associated neuropathy.It is worth mentioning that our conclusions are based on observations in two different animal models of Abassociated neuropathy (APP/PS1 and App NL-G-F mice), and involved both microglia-selective as well as constitutive full body deletion of core inflammasome components (caspase-1 and Nlrp3).Although we can only guess the reasons underlying the discrepant results obtained in our study and those of a previous report (33), potential explanations may include variations in experimental and environmental factors (Supplementary Table 1).First, genetic differences in AD models may explain the discrepant results between our and previous studies.The APP/PS1 model used in this study is the widely used model developed by the Jucker lab (34), which is a relatively 'aggressive' Ab model with a high Ab42/40 ratio.In contrast, the Heneka study used a different APP/PS1 mouse strain that is characterized by a slower plaque growth (53).Moreover, the App NL-G-F mice used in our study do not rely on transgenic overexpression as in the APP/PS1 models, but instead carry coding sequence mutations in the endogenous APP gene that drive rapid Ab aggregation, and thus may be less sensitive to microglia-dependent modulation of Ab-induced neuropathology.Second, variations in both the mice's gender and age during analysis could contribute to distinct observations.In our study, mice underwent analysis at both initial and advanced stages of disease progression, while the Heneka study reportedly conducted analysis at 16 weeks of age (Supplementary Table 1).Third, the composition of host microbiota linked to chow and other environmental factors in different animal facilities might contribute to differences in inflammasome responses.Indeed, the intestinal microbiota was shown to modulate immune functions and signaling to a wide variety of distant cells, including microglia (66, 67).It is worth mentioning in this regard that our studies were performed in three different animal facilities (VIB-Ghent University in Ghent, Belgium; Janssen Pharmaceutica in Beerse, Belgium; RIKEN Center for Brain Science, Saitama, Japan) with similar outcomes.Therefore, even though an animal facility-specific effect of differential microbiomes remains possible, additional examination in independent facility environments would be required to endorse putative effects of the intestinal microbiota.
Regardless, by demonstrating that the role of the Nlrp3 inflammasome in Ab-induced neurodegeneration is not as prominent or universal as previously anticipated, our findings contribute to the ongoing debate and highlight the need for further investigation into the complex mechanisms underlying AD.Although our findings demonstrate that inflammasome activation does not play a significant role in mouse models of Abinduced neurodegeneration, these observations leave open the possibility that inflammasomes may be implicated in the response to other mechanisms contributing to AD pathology, including the aggregation of hyperphosphorylated tau and the formation of neurofibrillary tangles.Loss of Nlrp3 inflammasome function was recently shown to reduce tau hyperphosphorylation and aggregation, identifying an important role of microglia and Nlrp3 inflammasome activation in the pathogenesis of tauopathies (68).With the emergence of brain-penetrant Nlrp3 inhibitors, future investigations should focus on examining whether pharmacological targeting of the Nlrp3 inflammasome can effectively suppress the formation of neurofibrillary tangles in AD and other tauopathies.Such studies could provide valuable insights into the mechanisms underlying neurodegeneration and facilitate the development of targeted therapeutic strategies for tau-related diseases.

Animals
App NL-G-F (35), APP/PS1 (34), caspase-1/11 (casp1/11) knockout (46) and Nlrp3 knockout (69) mice have been described.Casp1/11 KO mice were crossed with App NL-G-F mice to generate Casp1/11 KO App NL-G-F mice.Nlrp3 KO mice were crossed with APP/PS1 mice to generate Nlrp3 KO APP/PS1 mice.Conditional A20 (40) and Casp1 (45) knockout mice were generated as previously described, and crossed with Cx3Cr1Ert2-Cre transgenic mice (41) to generate tamoxifen-inducible myeloidspecific A20 and caspase-1 knockout mice, and further crossed with App NL-G-F transgenic mice.A20 FL and Casp1 FL littermate mice, expressing the floxed allele, are used as controls in all experiments.Casp1 FL and Casp1 Cx3Cr1-KO mice were also crossed with APP/PS1 transgenic mice.At 4-6 weeks of age, mice were subcutaneously injected with tamoxifen (20 mg/ml, Sigma-Aldrich T5648) dissolved in corn oil (Sigma-Aldrich, C8267) twice, 48h apart, to activate Cre recombinase.All experiments were performed on mice backcrossed into the C57BL/6 genetic background for at least 8 generations.Mice were housed in individually ventilated cages in a specific pathogen-free facility at VIB-Ghent, Belgium, at RIKEN and Nagoya City University, Japan (for the Casp1/11-App NL-G-F line), or at Janssen Pharmaceutica Beerse, Belgium (for the Nlrp3 KO APP/PS1 mice).All experiments were conducted according to institutional, national, and European animal regulations or in accordance with the guidelines of the RIKEN Center for Brain Science and Nagoya City University.Animal protocols were approved by the ethics committee of Ghent University, by the ethics committee of RIKEN Center for Brain Science and Nagoya City University, or by the ethics committee of Janssen Pharmaceutica.

Histology, brightfield microscopy and image analysis
Mice were transcardially perfused with PBS (Gibco).Brain tissue was dissected and left overnight at 4˚C in formaldehydebased fixative before being embedded in paraffin blocks.Alternatively, for biochemical analyses, hippocampi were dissected out and snap-frozen in liquid N 2 until further processing.Sagittal sections of 5 µm were rehydrated and incubated in antigen retrieval buffer (Dako).Analyses were performed over whole brain, hippocampus (bounded by CA1 and dentate gyrus), and cortex regions ("cortex" refers to areas in the mouse brain labeled by the Mouse Brain Atlas as visual, posterior parietal association areas, somatosensory, somatomotor, and orbital cortices, fiber tracts excluded) as clarified in individual graphs.For 6E10 staining, antigen retrieval was performed by incubating sections with 33% formic acid for 20 minutes.Endogenous peroxidase activity was blocked by treating tissue with 3% H 2 O 2 for 8-10 minutes.Non-specific binding was blocked by treating slides with blocking buffer (0.5% fish-skin gelatin + 2% BSA in PBS + 5% goat serum) for minimum 1h.The primary antibodies Iba1 (1:1000; Wako Chemicals), GFAP (1:5000; Dako), 6E10 (1:500; Biolegend), and 82E1 (1:1000, IBL America) were incubated overnight at 4˚C.Tyramide signal amplification (PerkinElmer Life Sciences) was used, as previously described (70).Brightfield images were obtained using Axio Scan.Z1 (Zeiss, Germany).Iba1 + and GFAP + cell bodies were manually counted in 7-12 representative images per experimental group using Zen (blue edition) (Zeiss).Image visualization parameters were standardized, and hippocampal glia numbers were normalized to hippocampal area (manually demarcated).Quantification of GFAP + area and 6E10 + plaque area over the whole brain was quantified using QuPath software (71).Graphs represent % of total brain area positive for GFAP + or 6E10 + staining.

Immunofluorescence microscopy and image analysis
For fluorescent immunohistochemistry of plaque-associated dystrophic neurites, rehydrated sections were incubated in 70% formic acid for 10 minutes.Endogenous peroxidase was blocked with H 2 O 2 treatment, and slides were incubated overnight with the primary antibodies JRD/sAPP/32 (0.2 µg/ml, kind gift J&J, Beerse-Belgium) and JRF/AbN/25 (2 µg/ml, kind gift J&J, Beerse-Belgium) in antibody diluent (DAKO).After washing, slides were incubated for one hour with secondary antibodies anti-mouse IgG2a Alexa647 (1:500, Life Technologies) and anti-mouse IgG1 Alexa488 (1:500, Life Technologies).For fluorescent immunohistochemistry of plaque-associated microglia, sections were incubated with Iba1 antibody (Wako Chemicals, Richmond, VA, USA) overnight, followed by incubation with the fluorescent dye PFTAA (pentameric formyl thiophene acetic acid, J&J, Beerse-Belgium) for one hour.Endogenous peroxidase activity was blocked with 3% hydrogen peroxide (DAKO, Glostrup, Denmark).Alternatively, 3 µm thick formaldehyde-fixed paraffin-embedded tissue sections were deparaffinized and cooked in EnVision FLEX Target Retrieval Solution High pH (DAKO) for 40 minutes, and blocked in 1% BSA, 1% Triton-X-100 in PBS.The slides were then incubated with antibodies against Iba1 (Synaptic Systems, 1:500) and TMEM119 (Synaptic Systems, 1:200) in 1% BSA, 1% Triton-X-100 in PBS at 4°C overnight.Sections were then washed and incubated for 2 hours with secondary antibodies Donkey Anti-Chicken Alexa Fluor 488 (Jackson Immuno, 1:300) and Donkey Anti-Rabbit Alexa Flour 568 (ThermoFisher, 1:500).Slides were then incubated with Methoxy-X04 (Tocris, 1:3000) to visualize plaques.Imaging was performed with a BZ-X800 fluorescent microscope (Keyence) and a 20x objective.At least 100 plaque-associated and 100 non-plaqueassociated cortical Iba1 positive cells per animal were evaluated for TMEM119 expression by manual inspection.Quantification of sAPP-labeled dystrophic neurites near N/25-positive plaque deposits was performed using Columbus image analysis software.Thresholds and plaque area were kept constant for all samples.Using QuPath software (71), N25/sAPP and Iba1/PFTAA images were sub-sampled into one training image and a pixel classifier that has been trained using Random Tree Forest by manually annotating colocalization and non-colocalization areas.For each slide, tissue area was detected by applying a gaussian blur of 2 pixels followed by a threshold on the DAPI channel, and pixels were classified using the pixel classifier.Colocalization detection was measured by tissue area.

Microglia isolation for in vitro experiments
0-3 day-old pups were used to isolate primary microglia.Brains were stripped of olfactory bulbs, cerebellum, midbrain, and meninges and stored in ice-cold F12 Nutrient Mix Ham medium (Gibco Life Technologies), after which, the tissue was trypsinized and resuspended in DMEM medium supplemented with 10% FCS, 1% penicillin/streptomycin, glutamine, sodium pyruvate, and nonessential amino acids (NEAA).Cells were grown in tissue flasks (75cm 3 ) pre-treated with poly-L-lysine for a minimum of 1h.Medium was refreshed approximately every other day until a confluent astrocyte layer appears (in 7-10 days).Microglia were then generated by changing the medium composition to 75% of DMEM medium + 25% L929 conditioned medium.After 4-6 days, microglia were isolated from this mixed glial culture by shaking the flasks at 100 rpm at 37°C for 1h and plated with 75% RPMI medium (supplemented same as DMEM) + 25% L929 conditioned medium.Cre-mediated deletion was induced by treating plated microglia with 1µM 4-hydroxytamoxifen (4-OH-TAM, Sigma) for 3 days.

Statistical analysis
Each n represents an independent biological sample.All data are represented as mean ± S.E.M. Statistical analysis was done using GraphPad Prism software version 8.0.and Genstat.Data were analyzed by applying statistical tests depending on the distribution of the data.Type of statistical analysis is mentioned for each experiment.
Cellular suspensions were loaded on a Chromium Chip B (10x Genomics, No.1000074) on a GemCode Single Cell Instrument (10x Genomics) to generate single-cell gel beads-in-emulsion (GEM).GEMs and scRNA-seq libraries were prepared using the GemCode Single Cell 3' Gel Bead and Library Kit (v3 10xGenomics, No. 1000075) and the Chromium i7 Multiplex Kit (10x Genomics, No. 120262) according to the manufacturer's instructions.Briefly, GEM reverse-transcription incubation was performed in a 96-deep-well reaction module at 53°C for 45 min, 85°C for 5 min and ending at 4°C.Next, GEMs were broken and complementary DNA (cDNA) was cleaned up with DynaBeads MyOne Silane Beads (10x Genomics, No. 2000048) and SPRIselect Reagent Kit (Beckman Coulter, No. B23318).Full-length, barcoded cDNA was PCR amplified with a 96-deep-well reaction module at 98°C for 3 min, eleven cycles at 98°C for 15 s, 63°C for 20 s and 72°C for 1 min, followed by one cycle at 72°C for 1 min and ending at 4°C.Following cleaning up with the SPRIselect Reagent Kit and enzymatic fragmentation, library construction to generate Illumina-ready sequencing libraries was performed by the addition of R1 (read 1 primer), P5, P7, i7 sample index and R2 (read 2 primer sequence) via end-repair, A-tailing, adapter ligation, post-ligation SPRIselect cleanup/size selection and sample index PCR.The cDNA content of pre-fragmentation and post-sample index PCR samples was analyzed using the 2100 BioAnalyzer (Agilent).
Sequencing libraries were loaded on an Illumina HiSeq4000 flow cell with sequencing settings following the recommendations of 10x Genomics (Read 1: 28 cycles, i7 Index: 8 cycles, i5 Index: 0 cycles, Read 2: 91 cycles, 2.73nM loading concentration).The Cell Ranger pipeline (10x Genomics) was used to perform sample demultiplexing and to generate FASTQ files for read 0, read 2 and the i7 sample index.Read 2, containing the cDNA, was mapped to the reference genome (mouse mm10) using STAR.Subsequent barcode processing, unique molecular identifiers filtering and single-cell 3' gene counting was performed using the Cell Ranger suite and Seurat v.4.3.1.The total number of cells across all libraries was 47 633 cells.The average of the mean reads per cell across all libraries was 45 319.6 ± 22 166 SD, with an average sequencing saturation of 67.05% ± 14.21% SD, as calculated by Cell Ranger.Digital gene expression matrices were preprocessed and filtered using the Seurat and Scater (v.1.22.0)R packages.Outlier cells were identified based on three metrics (library size, number of expressed genes and mitochondrial proportion per cell); cells were tagged as outliers when they were more than three median absolute deviations distant from the median value of each metric across all cells.By means of the Seurat Merge function, the raw counts of all samples were concatenated, yielding a total of 40 547 cells.The resulting dataset was normalized by the Seurat global-scaling normalization method 'LogNormalize' that normalizes the gene expression measurements for each cell by the total expression and multiplies it by a scale factor (10 000), and log-transforms the result.Highly variable genes were detected in Seurat according to the method described in Stuart et al. (73) and the data was scaled by linear transformation.Subsequently, the highly variable genes were used for principal component analysis (PCA).In order to remove batch effects and technical noise, the harmony package (v.1.0)was applied, using a theta parameter of 0. The resulting harmonycorrected PCA embeddings were used downstream for unsupervised Leiden clustering of the cells and UMAP dimensionality reduction, as implemented in Seurat.

1 Microglial 2
FIGURE 1Microglial A20 deficiency does not exacerbate AD pathology in App NL-G-F mice.(A) Quantification of the area covered by 6E10 + amyloid plaque deposits in whole brains of 20 week-old A20 FL (black) and A20 Cx3Cr1-KO App NL-G-F mice (red).Each symbol represents one mouse, n=5-6 per group (males, dark color; females, pale color).Data are represented as mean ± SEM. (B) Immunohistochemistry for 6E10 + amyloid plaque load in the hippocampus (HC) and frontal cortex (FC) of 20 week-old A20 FL and A20 Cx3Cr1-KO App NL-G-F mice.Scale bars: 100 µm (inset: 50 µm).Representative images are displayed.(C) Quantification of the number of 6E10 + amyloid plaque deposits in whole brains of 56 week-old A20 FL (black) and A20 Cx3Cr1-KO App NL-G-F mice (red).Each symbol represents one mouse, n=6 per group (males, dark color; females, pale color).Data are represented as mean ± SEM. (D) Immunohistochemistry for 6E10 + amyloid plaque load in the hippocampus (HC) and frontal cortex (FC) of 56 week-old A20 FL and A20 Cx3Cr1-KO App NL-G-F mice.Scale bars: 100 µm (inset: 50 µm).Representative images are displayed.(E) Quantification of the number of Iba1 + microglia in the hippocampus of 20 week-old A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Each symbol represents one mouse, n=4-5 per group (App WT ); n=9 per group (App NL-G-F ).Data are represented as mean ± SEM.Significant differences are determined using two-way ANOVA (**, p<0.001).(F) Immunohistochemistry for Iba1 + microglia in the hippocampus of 20 week-old A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Scale bars: 100µm (inset: 50 µm).Representative images are displayed.(G) Quantification of the number of Iba1 + microglia in the hippocampus of 56 week-old A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Each symbol represents one mouse, n=4 per group (App WT ); n=13 per group (App NL-G-F ).Data are represented as mean ± SEM.Significant differences are determined using two-way ANOVA (**p<0.001;****p<0.0001).(H) Immunohistochemistry for Iba1 + microglia in the hippocampus of 56 week-old A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Scale bars: 100µm (inset: 50 µm).Representative images are displayed.(I) Quantification of the number of GFAP + astrocytes in the hippocampus of 20 weekold A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Each symbol represents one mouse, n=4-5 per group (App WT ); n=9 per group (App NL-G-F ).Data are represented as mean ± SEM.Significant differences are determined using two-way ANOVA (***p<0.001).(J) Immunohistochemistry for GFAP + astrocytes in the hippocampus of 20 week-old A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Scale bars: 100µm (inset: 50 µm).Representative images are displayed.(K) Quantification of the number of GFAP + astrocytes in the hippocampus of 56 week-old A20 FL and A20 Cx3Cr1- KO App WT and App NL-G-F mice.Each symbol represents one mouse, n=4 per group (App WT ); n=6 per group (App NL-G-F ).Data are represented as mean ± SEM. (L) Immunohistochemistry for GFAP + astrocytes in the hippocampus of 56 week-old A20 FL and A20 Cx3Cr1-KO App WT and App NL-G-F mice.Scale bars: 100µm (inset: 50 µm).Representative images are displayed.(M) Quantification of the number of plaque-associated dystrophic neurites (N25 + sAPP + ) in whole brains of 20 week-old A20 FL and A20 Cx3Cr1-KO App NL-G-F mice.Each symbol represents one mouse, n=8-10 per group.Data are represented as mean ± SEM. (N) Immunohistochemistry for N25 + sAPP + plaque-associated dystrophic neurites in whole brains of 20-week-old A20 FL and A20 Cx3Cr1-KO App NL-G-F mice.Scale bars: 50 µm.Representative images are displayed.(O) Quantification of the number of plaque-associated dystrophic neurites (N25 + sAPP + ) in whole brains of 56 week-old A20 FL and A20 Cx3Cr1-KO App NL-G-F mice.Each symbol represents one mouse, n=10-14 per group.Data are represented as mean ± SEM. (P) Immunohistochemistry for N25 + sAPP + plaque-associated dystrophic neurites in whole brains of 56 week-old A20 FL and A20 Cx3Cr1-KO App NL-G-F mice.Scale bars: 50 µm.Representative images are displayed.ns, not significant.

3 Full
FIGURE 3Full-body casp1/11 deficiency does not suppress AD pathology in 12 month-old App NL-G-F mice.(A) Amyloid deposition and neuroinflammation was detected by triple staining of 12-month-old casp1/11 WT and casp1/11 KO App NL-G-F mice using 82E1 (blue), anti-Iba1 antibody (red) and anti-GFAP antibody (green) as markers of Ab plaque, astrocytosis and microgliosis, respectively.Scale bars represent 100 µm.(B) Quantification of 82E1+ amyloid plaque load, Iba1+ microgliosis, and GFAP+ astrocytosis over whole brains of casp1/11 WT and casp1/11 KO App NL-G-F mice (all females).Data are represented as mean ± SEM. (C) Quantification of the number of plaque-associated microglia in casp1/11 WT and casp1/11 KO App NL-G-F mice.Each symbol represents one mouse.Data are represented as mean ± SEM. (D) Biochemical quantification of Ab 40 and Ab 42 in the GuHCl fractions of cortical tissue from 12-month-old mouse brains from casp1/11 WT and casp1/11 KO App NL-G-F mice, quantified by sandwich ELISA.Each symbol is one mouse (all females), n=4 per group.Data represent mean ± SEM. ns, not significant.

5 Full
FIGURE 5 Full-body Nlrp3 deficiency does not suppress AD pathology in APP/PS1 mice.(A) Quantification of 4G8 + amyloid plaque load in the hippocampus of 4months old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse.Data are represented as mean ± SEM. (B) Immunohistochemistry for 4G8 + amyloid plaque load in hippocampus of 4 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 500 µm (inset: 50µm).Representative images are displayed.(C) Quantification of 4G8 + amyloid plaque load in the hippocampus of 6-months old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse.Data are represented as mean ± SEM. (D) Immunohistochemistry for 4G8 + amyloid plaque load in hippocampus of 6 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 500 µm (inset: 50µm).Representative images are displayed.(E) Quantification of 4G8 + amyloid plaque load in the hippocampus of 10-months old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse.Data are represented as mean ± SEM. (F) Immunohistochemistry for 4G8 + amyloid plaque load in hippocampus of 10 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 500 µm (inset: 50µm).Representative images are displayed.(G) Quantification of the number of hippocampal Iba1+ microglia in 4 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse, n=14-23 per group.Data are represented as mean ± SEM.Significant differences are determined with One-way-ANOVA using Sidak's multiple comparisons test (*p < 0.05; **p < 0.01).(H) Immunohistochemistry for Iba-1+ microglia in the hippocampus of 4 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 500 µm (inset: 50µm).Representative images are displayed.(I) Quantification of the number of hippocampal Iba1+ microglia in 6 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse, n=14-23 per group.Data are represented as mean ± SEM.Significant differences are determined with One-way-ANOVA using Sidak's multiple comparisons test (*p < 0.05).(J) Immunohistochemistry for Iba-1 + microglia in the hippocampus of 6 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 500 µm (inset: 50µm).Representative images are displayed.(K) Quantification of the number of hippocampal Iba1+ microglia in 10 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse, n=14-23 per group.Data are represented as mean ± SEM.Significant differences are determined with One-way-ANOVA using Sidak's multiple comparisons test (****p < 0.0001).(L) Immunohistochemistry for Iba-1+ microglia in the hippocampus of 10 month-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 500 µm (inset: 50µm).Representative images are displayed.(M) Quantification of the number of plaque-associated dystrophic neurites (N25+ sAPP+) in the brain of 4-months old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse, n=14-23 per group.Data are represented as mean ± SEM. (N) Immunofluorescence staining for N/25 and sAPP in 4 months-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 50µm.Representative images are displayed.(O) Quantification of the number of plaque-associated dystrophic neurites (N25+ sAPP+) in the brain of 6-months old Nlrp3 WT and Nlrp3 KO APP/ PS1 mice.Each symbol represents one mouse, n=14-23 per group.Data are represented as mean ± SEM. (P) Immunofluorescence staining for N/25 and sAPP in 6 months-old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Scale bars: 50µm.Representative images are displayed.(Q) Quantification of the number of plaque-associated dystrophic neurites (N25+ sAPP+) in the brain of 10-months old Nlrp3 WT and Nlrp3 KO APP/PS1 mice.Each symbol represents one mouse, n=14-23 per group.Data are represented as mean ± SEM. (R) Immunofluorescence staining for N/25 and sAPP in 10 months-old Nlrp3 WT and Nlrp3 KO APP/ PS1 mice.Scale bars: 50µm.Representative images are displayed.ns, not significant.
(A) UMAP plot of 40547 microglia cells from whole brain tissue of Nlrp3 +/+ mice (3, 6 and 9 months old), Nlrp3 -/-mice (3, 6 and 9 months old), Nlrp3 +/+ APP/PS1 (3 and 9 months old) and Nlrp3 -/-APP/PS1 (3 and 9 months old).(B) Dot plot, showing the expression of key marker genes per cluster in the dataset from (a) The size of the dot represents the percentage of cells within a cluster that express the gene, while the color encodes the mean scaled gene expression level per cluster.(C) UMAP plot of microglia from Nlrp3 +/+ APP/PS1 and Nlrp3 -/-APP/PS1 mice, split by genotype and age.(D) Pie charts visualizing the percentage of cells per cluster in each genotype and age group from (c) E. Violin plot, comparing the expression level of the Nlrp3 gene in the microglia from the four distinct genotypes (all age groups).(F) Volcano plot, showing differentially expressed genes between microglia (all clusters combined) from Nlrp3 -/-APP/PS1 vs Nlrp3 +/+ APP/PS1 mice (9 months old).In red are shown the genes with adj P<0.05 and abs(log2FC)>1.(G) Volcano plot, showing differentially expressed genes between DAMs from Nlrp3 -/-APP/PS1 vs Nlrp3 +/+ APP/PS1 mice (9 months old).In red are shown the genes with adj P<0.05 and abs(log2FC)>1.HM, homeostatic microglia; IRM, Interferon response microglia; TM, transitory microglia; DAM, disease-associated microglia; IEG, immediate early genes response microglia; PM, proliferating microglia.