Sci Transl Med: Niacin receptor HCAR2 regulates microglial response and AD progression

The niacin receptor HCAR2 is selectively expressed by microglia in the brain, while increased dietary niacin intake is associated
with a reduced risk of Alzheimer's disease (AD).
。 Recently, the journal Science Translational Medicine published a paper entitled "The niacin receptor HCAR2 modulates microglial response and limits disease progression in a mouse model of Alzheimer's disease".
The authors' team revealed a novel mechanism linking
the niacin receptor HCAR2 to regulate microglial responses and AD progression.
and activated HCAR2 using an FDA-approved niacin (Niaspan) preparation in 5xFAD mice, thereby reducing plaque load and neuronal dystrophy, reducing neuronal loss, and improving memory deficits, thereby mitigating amyloid-induced pathology, suggesting that the use of Niaspan may be a promising treatment for AD, specifically targeting neuroimmune responses
.

Induction of HCAR2 by microglia in AD patients

To examine whether the expression of GPCR HCAR2 in AD is altered, the authors analyzed the expression of Hcar2 mRNA in a mouse model of amyloid 5xFAD and human AD brain tissue (Figure 1).

In the hippocampus and cortex of 4~6 month-old female and male 5xFAD animals, the expression of Hcar2 increases during the active deposition of plaques (Figure 1A).

To determine whether Hcar2 expression is specifically related to microglia, the authors' team depleted microglia cells in 4-month-old 5xFAD mice
using CSF1R antagonist PLX5622.
It was found that depleting approximately 70% of cortical microglia reduced Hcar2 mRNA expression in the 5xFAD brain (Figure 1B, left).

In addition, when the CSF1R antagonist (PLXon-off) is discontinued, it leads to reproliferation of microglia and restoration of Hcar2 expression (Figure 1B, right).

These results suggest that there is a specific correlation between Hcar2 expression in the induced 5xFAD brain and microglia
.
It is possible that the underlying cause of the increased expression of Hcar2 mRNA is the proliferation of microglia, and the authors' team analyzed the transcriptome dataset of sorted microglia from 8.
5-month-old mice with 5.
5 months and found an increase in Hcar2 expression within microglia in the 5xFAD brain (Figure 1C, top).

Furthermore, incubating primary murine microglia with 5 μM Aβ1-42 aggregates for 24 h resulted in a significant increase in Hcar2 expression (P<0.
001) (Figure 1C, bottom), consistent
with previous findings 。 To observe the induction of Hcar2 in the brain of 5xFAD, the authors' team compared 5xFAD mice with Hcar2 (5xFAD; Hcar2mRFP) mice expressing the monomeric red fluorescent protein reporter gene (mRFP) were previously reported to accurately reflect Hcar2 expression and were used to study Hcar2 expression
in microglia 。 Subsequent 4-month-old B6, 5xFAD, and immunohistochemistry (IHC) of Hcar2mRFP animals showed that Hcar2 expression was limited to Iba1-positive microglia (Figure 1D), but its expression was significantly increased in cells surrounding Aβ plaques compared to microglia not affected by Aβ plaques (Figure S1A).

。 Since both mRFP and Iba1 are colocalized with the microglia-specific marker P2RY12 (Figure S1B), it is confirmed that Hcar2 is induced by brain-resident microglia in the 5xFAD brain, consistent with previous studies and also showing no peripheral monocyte infiltration into the brain parenchyma in this model
.
The authors' team also observed induced Hcar2 mRNA in the brains of a 4-month-old amyloid-producing mouse model of APPPS1 (Figure S1C).

。 In addition, published transcriptome analysis of sorted microglia from the APPPS1 model showed that the transcriptome analysis was associated with Aβ plaques (Clec7a-; Figure S1D) compared to unrelated microglia, Hcar2 mRNA was selectively increased in Aβ plaque-associated microglia [type C lectin domain containing 7a (Clec7a)+], which is consistent
with the authors' team's results in the 5xFAD model.
Transcriptome data analysis showed an increase in Hcar2 expression in microglia in PS19 tau proteopathic mice (Figure S1E), suggesting that tau pathology can also induce HCAR2 expression
in AD.

Fig.
1 Inducing effect of microglia on HCAR2 in AD patients

The authors' team analyzed a human transcriptome dataset
of dorsolateral prefrontal cortex (BA9) tissue from 157 non-dementia control subjects and 310 AD patients (GSE33000) in human transcriptome data.
The results show an increase in HCAR2 expression in AD (Figure 1E).

This finding
was validated by qPCR response (Figure 1E) analysis of HCAR2 expression in postmortem brain tissue samples from control subjects and AD patients.
IHC from non-dementia control (CTRL) and postmortem brain slices from AD patients showed that HCAR2 protein expression was selective for Iba1-positive cells and significantly increased (P<0.
05) in AD (Figure 1F), consistent with the authors' team's findings in the 5xFAD model (Figure 1D).

。 However, as observed in mouse models of amyloid (Figure 1D and Figure S1B), the induction of HCAR2 in the human AD brain does not appear to be limited to the surrounding environment of amyloid plaques (Figure 1F), possibly due to the robust and widespread presence of other triggerers such as tau (Figure S1C), but may also be caused
by nonspecific interactions of antibodies 。 Given the nonspecific immune response of many GPCR antibodies, the authors assessed the specificity of anti-HCAR2 antibodies by small interfering RNA-mediated knockout of HCAR2 in the human microglia line HMC3, which resulted in a corresponding decrease in the HCAR2 immune response (Figure S1E).

Overall, these findings support the selective expression of HCAR2 by microglia in the brain, which is consistent
with RNA sequencing (RNA-seq) datasets and previous confirmation of the HCAR2 gene as part of the human core microglia signature.
In addition, these data suggest the presence of HCAR2-inducing effects
in the brains of AD patients.

Fig.
S1 Brain Hcar2 induction, Hcar2 expression in APPPS1 and PS19 mouse models, and verification of Hcar2 antibody ab81825 in 5xFAD mice

The lack of Hcar2 interferes with microglial activity in the amyloidosis 5xFAD brain

Amyloid deposition begins early in AD progression and has triggered a microglial response
, including increased HCAR2 expression, decades before symptoms appear.
Therefore, the authors' team sought to investigate the role
of HCAR2 in amyloidosis pathology through the method of gene inactivation of HCAR2 in the 5xFAD amyloidosis model.
The authors' team crossed 5xFAD with Hcar2-/- animals to obtain 5xFAD mice with missing the Hcar2 gene (5xFAD; Hcar2-/-)
。 The absence of Hcar2 was confirmed in brain tissue and primary cultured microglia of Hcar2/- animals (Figure S2A).

The glial cell atlas of NANOSTRING was used to analyze RNA
in the hippocampal tissues of 6-month-old female mice with 5xFAD, Hcar2+/+ and 5xFAD, Hcar2-/-.
Differentially expressed genes (DEGs) [adjusted (adj.
) P<0.
05] are shown
in the heat map (Figure 2A).
The inactivation of Hcar2 led to significant downregulation of 40 genes (adj.
P<0.
05), no significantly upregulated genes were found (adj.
P>0.
05)
。 The glial characterization panel contains genes
expressed in microglia, astrocytes, oligodendrocytes, and neurons.
However, compared to other cell types, more than half of the genes downregulated due to Hcar2 inactivation are selectively expressed in microglia, as described in the Barres RNA-Seq dataset (Figure 2A), which is consistent with
the specificity of HCAR2 for microglia in the brain (Figure 1).
。 In addition, enrichment analysis of gene ontology terms for biological processes (GO-BP) and WikiPathways showed that the lack of Hcar2 inhibits pathways associated with immune activation and response, particularly phagocytosis and the transmembrane immune signaling adapter TYROBP pathway (Figure 2B).

Furthermore, GO cell component enrichment analysis showed a decrease in the expression of cell surface and membrane components, consistent with a decrease in immune response and phagocytosis (Figure 2B).

The decrease in synaptic components when Hcar2 is lost also indicates an effect on neuronal function (Figure 2B).

The authors' team validated the 5xFAD by qPCR; Hcar2+/+ and 5xFAD; Reduction
of several DEGs (SPP1, Cst7, CD68, TREM2, Tyrobp, and Ax1) involved in microglial reactions and phagocytosis in Hcar2-/-.
The authors' team also analyzed their effects in non-GMO B6; Hcar2+/+ and B6; Expression in Hcar2-/- (Figure 2C).

As expected, there was a significant increase in these genes in 5xFAD mice compared to the B6 control group; However, this induction is inhibited by Hcar2 deletion (Cst7, CD68, TREM2, and Tyrobp) or even canceled (Spp1 and Ax1) (Figure 2C).

The authors' team observed no change in gene expression between B6, Hcar2+/+ and B6, Hcar2-/- through qPCR, indicating that Hcar2 does not affect these pathways
in non-diseased brain tissue.
Overall, these data suggest that the lack of Hcar2 leads to an inadequate response of microglia to amyloid pathology, which is significantly associated
with a decrease in phagocytic function and TYROBP signaling.

Fig.
2 The lack of Hcar2 interferes with the microglial pathway in the amyloidosis 5xFAD brain

Lack of Hcar2 reduces the contact of microglia with plaques, increasing plaque burden

The author's team on 5xFAD; Hcar2+/+ and 5xFAD: Hcar2-/- mice performed further functional and phenotypic analyses to understand how the observed transcriptional differences (Figure 2) effectively translate into altered microglial responses and severity
of amyloid pathology.
Therefore, the authors' team investigated the effect of
Hcar2 on plaque pathology and microglia-plaque interactions in the brain with 5xFAD.
In the 5xFAD model, the hippocampal suberior nursery region exhibits the earliest and most aggressive amyloid accumulation
.
The authors' team observed a significant increase in the number of thioflavin S-positive plaques (P<0.
05), and that these plaques also occupied an increased area in the lower bracket of female and male 5xFAD mice missing Hcar2 at 4 and 6 months of age (Figure 3A and Figure S 3A-B).

There was a slight increase in plaque load in the hippocampus (excluding the lower nurse) in 4-month-old animals (Figure S3C) and the cerebral cortex (Figure S3A) of 6-month-old female 5xFAD mice with Hcar2 deletion, along with astrocytosis (Figure S3D).

5xFAD at 4 months of age; The increase in plaque load in the Hcar2/- mouse lower bracket is accompanied by a decrease in Iba1 percentage area (Figure 3B).

Consistent with this, the inactivation of Hcar2 resulted in a decrease in Aif1 expression in the hippocampus of 4-month-old 5xFAD, but not in non-transgenic control (B6) mice (Figure 3B), suggesting that this effect was absent in undiseased
brains.
The authors' team further examined the response
of microglia to amyloid pathology in late disease in 6-month-old 5xFAD mice by IHC.
Similar to 4-month-old animals, loss of Hcar2 reduces the area of Iba1 percent in the lower tray (Figure 3C), accompanied by a decrease in the microglial envelope of plaques (Figure 3C).

At 5xFAD; Similar reductions were observed in the cortex of Hcar2/- mice (Figure 3D).

The authors' team quantified microglia abundance within a 25 μm radius around each plaque to assess the recruitment of microglia to the plaque and found a significant reduction in the absence of Hcar2 (P<0.
001) (Figure 3E).

Overall, these data support that the loss of Hcar2 leads to inadequate microglia coverage and inadequate contact with plaques, which is associated with
a higher plaque burden.

Fig.
3 The lack of Hcar2 reduces the contact of microglia with plaques and increases the plaque burden

Figure S3 Hcar2 deficiency increases plaque burden and astrocytosis in 5xFAD mice

Hcar2 is required for efficient proliferation of microglia and amyloid uptake

The authors' team observed 5xFAD; The microglial response in Hcar2-/- mice is inhibited, resulting in an increased
plaque burden.
Because the area occupied by microglia is in 5xFAD; Hcar2-/- with 5xFAD; Hcar2+/+ in mouse brains has reduced areas compared to the former, so the authors' team hypothesized that HCAR2 regulates the proliferation of microglia in response to amyloid pathological changes
.
At the same time, it was analyzed whether the loss of Hcar2 affected the expression
of the proliferation marker Ki-67 in the brains of 4-month-old 5xFAD animals.
With IHC, the authors' team observed a significant reduction in Ki-67-positive microglia (Ki-67+/Iba1+) in the brains of 5xFAD mice lacking Hcar2 (P<0.
05) (Figure 4A), indicating reduced
proliferation.
Ki-67 staining is observed only within the nucleus (Figure S4A).

Because the proliferative capacity of microglia is associated with an increase in phagocytic activity, it is found in 5xFAD; This decreased proliferation was observed in Hcar2-/- along with decreased phagocytic gene expression (Figure 2) and increased plaque load (Figure 3), suggesting that a lack of Hcar2 leads to defects
in Aβ phagocytosis.
Therefore, the authors' team examined how HCAR2 affects the total number
of microglia and amyloid uptake in the 5xFAD brain by flow cytometry.
which passes 5xFAD at 6 months of age; Hcar2-/- and 5xFAD; Hcar2+/+ animals performed in vivo microglia phagocytosis assay with intraperitoneal administration of methoxy-X04 (Aβ marker), 3 h
before flow cytometry as previously described.
Microglia are identified by gated CD11b+ cells (Figure S4B), consistent with reduced microglial coverage and proliferation, which the authors' team observed to be associated with 5xFAD; Hcar2+/+ animals, 5xFAD; The number of microglia in Hcar2/- animals was significantly reduced (P<0.
05) (Figure 4B).

Based on CD45 expression, microglia can be divided into CD45low (lower activity) and CD45int (higher activity) (Figure S4C).

Hcar2 inactivation significantly reduced the number of cells in the CD45int population (P<0.
05) and had no significant effect on CD45low cells (P=0.
09) (Figure 4C).

Thus, as expected, the loss of Hcar2 reduces the proportion of CD45int cells in the total microglial population (CD11b+) while increasing the proportion of CD45low cells, suggesting that the lack of Hcar2 limits the activation of microglia triggered by amyloid pathology (Figure 4C).

A decrease in the number of microglia (CD11b+ cells) is also accompanied by a population in 5xFAD; Overall reduction of Hcar2-/- Aβ uptake (methoxy-X04+) in animals (Figure 4D).

In Hcar2 knockout animals, the percentage of methoxy-X04-positive CD11b+ cells was only slightly reduced (Figure S4C), suggesting that HCAR2 has little
effect on the intrinsic ability of microglia to uptake Aβ.
However, this analysis may not accurately reflect the overall size of
the effect of HCAR2 on Aβ absorption.
5xFAD; The percentage of Hcar2/- cells that phagocytose Aβ in the animals may be elevated, purely because these mice rely on fewer microglia to clear more Aβ, which can mask defects
in their inherent amyloid uptake 。 Therefore, to further investigate whether HCAR2 affects the intrinsic ability of microglia to uptake Aβ, the authors' team cultured primary mouse microglia from Hcar2+/+ and Hcar2-/- mice and cultured cells at 1 mM for 24 h with an FDA-approved niacin (Niaspan) formula to activate HCAR2, followed by 30 min
with 1.
25 uM HiLyte Fluor 488-labeled Aβ1-42 aggregates 。 The results of the authors' team show that basal uptake of Aβ1-42 is reduced in Hcar2-/- microglia (Figure 4E).

In addition, Niaspan treatment increased the uptake of Aβ1-42 in Hcar2+/+ cells but did not increase Hcar2-/- microglia (Figure 4E), confirming that niacin acts via HCAR2 to increase Aβ uptake
.
In addition, treatment of Hcar2+/+ microglia with lower concentrations of niacin (100 uM) also resulted in a significant increase in Aβ uptake (P<0.
05), which was not observed in Hcar2-/- cells (Figure S4E).

These results are consistent with previous findings that HCAR2 stimulates microglia to engulf globuli and myelin fragments
.
Together, these data suggest that HCAR2 is necessary
for efficient proliferation of microglia and Aβ uptake in the context of amyloid lesions.

Fig.
4 Hcar2 is required for efficient proliferation of microglia and amyloid uptake

Figure S4 5xFAD; Hcar2+/+ and 5xFAD; Proliferation and flow cytometric analysis of Hcar2-/- mouse microglia and effect of Niaspan on Aβ uptake in vitro

Hcar2 deficiency exacerbates amyloid-related neuropathology

Transcriptome analysis by the authors' team showed that with 5xFAD; Hcar2+/+ mice compared to 5xFAD; Impaired neuronal function in Hcar2-/- (Figure 2B).

The authors' team assessed the working memory
of B6 and 5xFAD mice (Hcar2+/+ and Hcar2-/-) using the Y maze task.
This assay has been widely used in 5xFAD models, which show that severe defects in working memory begin to appear at about 4 to 5 months of age
.
The authors' team analyzed 4-month-old mice, although working memory impairment was present at 5xFAD; Hcar2++/+ is still not evident in animals of this age, but 5xFAD mice lacking Hcar2 have shown significant defects (P<0.
05), independent of the number of arm entries (Figure 5A).

Female and male mice were analyzed together, as the authors' team observed no sex-dependent differences
in each genotype at this age.
The authors' team also evaluated motor skills (speed and distance) and anxiety-like behavior (time inactivity) for all genotypes during the Y maze task and found no significant difference (P>0.
05; Figure S 5A).

Fig.
5 Hcar2 deficiency can aggravate amyloid-related neuropathology

In the 5xFAD model, the lower nurse experiences severe neuronal loss
as the disease progresses.
IHC analysis of NeuN-positive cells showed that with 5xFAD; Hcar2+/+ mice compared to 4-month-old 5xFAD; Increased loss of Hcar2/- neurons in mice (Figure 5B).

Dystrophic neurites (DNs) are swollen abnormal neurites that are abundant near Aβ deposits within the brains of AD patients and 5xFAD mice
.
DNs are rich in lysosome-associated membrane protein 1 (LAMP-1) and ubiquitin and accumulate neuronal N-terminal APP (N-APP).

IHC analysis showed that despite 5xFAD; Hcar2+/+ and 5xFAD; The percentage area of Hcar2-/- LAMP-1 and ubiquitin in animals remained similar, but at 5xFAD; Within the basal cells of Hcar2-/-, the colocalization of LAMP-1 with ubiquitin and N-APP is increased (Figure S5B).

Consistent with this, the analysis of N-APP staining within individual DNs showed that N-APP was found at 5xFAD; Increased accumulation in Hcar2/- mice (Figure 5C).

The authors' team found a significant correlation between higher N-APP content in DNs and a reduced number of neurons (P<0.
05) (Figure S5C).

These results suggest that the inactivation of Hcar2 exacerbates neuronal pathology in 5xFAD, including increased
neuronal loss associated with accelerated onset of working memory deficits.

Fig.
S5 Behavioral characteristics and dystrophic neurite analysis in Hcar2+/+ and Hcar2-/- mice

The pharmacological effect of Niaspan-activated HCAR2 reduces amyloidosis in 5xFAD mice

To determine whether HCAR2 can stimulate a beneficial microglial response after amyloid pathological onset after drug activation, the authors' team treated 5xFAD animals
with Niaspan, an FDA-approved oral formulation of niacin.
There is good evidence that female 5xFAD mice exhibit more aggressive amyloid pathology
compared to males.
Because 5xFAD males and females both show similar increases in Hcar2 expression and likewise exhibit lack of Hcar2 exacerbating amyloid pathology (Figures 1, 3, and 5), the authors' team chose to eliminate the sex-dependent confounding effects by analyzing only the effects
of HCAR2 activation in male mice.
To mimic treatment during the symptomatic phase of the disease, the authors' team began administering Niaspan to 5-month-old 5xFAD animals that had already exhibited working memory deficits (Figure S6A).

Niaspan (100 mg/kg) is given to animals by oral gavage for 30 days
.
This dose results in a significant increase in niacin concentration in the brain at 5xFAD 30 min after administration (P<0.
001) (≈6-fold) (Figure S6B).

Treatment of 5xFAD mice with Niaspan rescued working memory deficits and associated brain regions, most prominently not only in the inferior brace, but also in the hippocampus and cortex (Figure 6B and Figure S6C).

In addition, plaque size in the basement membrane of treated animals was reduced (Figure S6D).

The authors' team also observed through IHC that Iba1 was completely colocalized with the microglia-specific marker P2RY12 in Niaspan-treated animals, which supports that this treatment does not induce monocyte infiltration of peripheral origin into the brain parenchyma of 5xFAD (Figure S6E).

These data suggest that Niaspan was able to reduce plaque load and neuronal pathology
in 5xFAD mice.
Because the lower nurse proved to be the most sensitive area for Niaspan treatment, the authors' team analyzed the microglial response
in this area.
The decrease in plaque load after Niaspan-treated animals was accompanied by a slight decrease in Iba1 percentage area, although plaque coverage of microglia remained similar between vehicle and treatment animals (Figure 6D).

Because plaque and Iba1 area area decreased with Niaspan treatment, to compare the response of microglia to amyloid deposition between untreated and treated animals, the authors' team measured the ratio
of microglia area to plaque area.
In Niaspan-treated mice, the ratio of Iba1 area/thioflavin S area increases (Figure 6E), suggesting that Niaspan induces microglia to mobilize more efficiently in response to Aβ deposition
.
In addition, found with 5xFAD; Plaque involvement and genes associated with downregulation of amyloid uptake in Hcar2-/- mice (Figure 2) were increased in the hippocampus of Niaspan-treated mice (Trem2, Axl, and Cd68) (Figure 6F).

The authors' team added Mrc1 to the qPCR analysis because this microglial gene involved in phagocytosis has been shown to be regulated by HCAR2 and is not included in
the nCounter glial characterization panel.
MRC1 increases with Niaspan treatment (Figure 6F).

In addition, Clec7a is an important marker of microglial binding to plaques in AD, and the authors' team observed that Niaspan treatment significantly induced (P<0.
01) its expression (Figure 6F), while NanoString analysis of Hcar2-/- mice showed no effect
on Clec7a expression 。 Despite the similarity of plaque envelopes between untreated and treated animals, Niaspan-induced genes associated with plaque contact and amyloid uptake suggest an increased efficiency of microglia in reducing amyloid deposition, supported by the authors' team's in vitro phagocytosis assay (Figure 4E).

。 Furthermore, using the Meso Scale Discovery (MSD) cytokine panel, the authors' team observed that Niaspan's stimulation of microglia did not cause an overall change in brain cytokine concentration, except for a mild increase in interferon-γ (IFN-γ), which γ has been reported to have a potentially beneficial effect on AD
.

Fig.
6 Niaspan stimulates microglial responses and reduces amyloid lesions in AD mice

To verify that Niaspan exerts its beneficial role via HCAR2, the authors' team used 5xFAD; Hcar2-/- mice were evaluated with the same dose paradigm
.
In the absence of Hcar2, the authors' team did not observe improved working memory or reduced neuronal loss in 5xFAD mice treated with Niaspan (Figure 7A-B).

In addition, at 5xFAD; In Hcar2-/- mice, Niaspan treatment did not alter either plaque load or microglia coverage (Figure 7C-D).

Niaspan-mediated gene induction associated with plaque envelope and amyloid uptake and 5xFAD in the absence of Hcar2; The IFN-γ increase observed in Hcar2+/+ is eliminated (Figure 7E-F).

These observations suggest that Niaspan acts via HCAR2 to stimulate beneficial microglial responses and attenuate amyloid pathology, which validates a potential therapeutic strategy
to target this receptor as AD.

Fig.
7 Niaspan treatment has no effect on 5xFAD lacking Hcar2

brief summary

The authors' team reports a therapeutic strategy to modulate AD microglial function via the niacin receptor HCAR2, which is required
for an efficient microglial response to AD-associated pathological responses.
And it was demonstrated that activating HCAR2 with the FDA-approved niacin preparation Niaspan can reduce plaque burden and neuronal pathology, and save working memory deficits
in a mouse model of 5xFAD.
Therefore, niacin is a promising drug for the treatment of AD with high potential
for clinical application.

Original:

Miguel Moutinho, Shweta S Puntambekar, Andy P Tsai, Israel Coronel, Peter B Lin, Brad T Casali, Pablo Martinez, Adrian L Oblak, Cristian A Lasagna-Reeves, Bruce T Lamb, Gary E Landreth.
 The niacin receptor HCAR2 modulates microglial response and limits disease progression in a mouse model of Alzheimer's disease.
Sci Transl Med.
2022 Mar 23; 14(637):eabl7634.
doi: 10.
1126/scitranslmed.
abl7634.