The mixed-lineage kinase 3 inhibitor URMC-099 facilitates microglial amyloid-β degradation

Background Amyloid-β (Aβ)-stimulated microglial inflammatory responses engage mitogen-activated protein kinase (MAPK) pathways in Alzheimer’s disease (AD). Mixed-lineage kinases (MLKs) regulate upstream MAPK signaling that include p38 MAPK and c-Jun amino-terminal kinase (JNK). However, whether MLK-MAPK pathways affect Aβ-mediated neuroinflammation is unknown. To this end, we investigated if URMC-099, a brain-penetrant small-molecule MLK type 3 inhibitor, can modulate Aβ trafficking and processing required for generating AD-associated microglial inflammatory responses. Methods Aβ1-42 (Aβ42) and/or URMC-099-treated murine microglia were investigated for phosphorylated mitogen-activated protein kinase kinase (MKK)3, MKK4 (p-MKK3, p-MKK4), p38 (p-p38), and JNK (p-JNK). These pathways were studied in tandem with the expression of the pro-inflammatory cytokines interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α. Gene expression of the anti-inflammatory cytokines, IL-4 and IL-13, was evaluated by real-time quantitative polymerase chain reaction. Aβ uptake and expression of scavenger receptors were measured. Protein trafficking was assessed by measures of endolysosomal markers using confocal microscopy. Results Aβ42-mediated microglial activation pathways were shown by phosphorylation of MKK3, MKK4, p38, and JNK and by expression of IL-1β, IL-6, and TNF-α. URMC-099 modulated microglial inflammatory responses with induction of IL-4 and IL-13. Phagocytosis of Aβ42 was facilitated by URMC-099 with up-regulation of scavenger receptors. Co-localization of Aβ and endolysosomal markers associated with enhanced Aβ42 degradation was observed. Conclusions URMC-099 reduced microglial inflammatory responses and facilitated phagolysosomal trafficking with associated Aβ degradation. These data demonstrate a new immunomodulatory role for URMC-099 to inhibit MLK and to induce microglial anti-inflammatory responses. Thus, URMC-099 may be developed further as a novel disease-modifying AD therapy. Electronic supplementary material The online version of this article (doi:10.1186/s12974-016-0646-z) contains supplementary material, which is available to authorized users.


Background
Neuroinflammation is a pathogenic driver for Alzheimer's disease (AD) [1][2][3]. Amyloid-β (Aβ)-activated microglia secrete pro-inflammatory neurotoxins that are strongly linked to AD-associated neural injury. Such inflammatory responses are notably tied beyond AD to common age-related neurodegenerative, neuroinfectious, and neuroinflammatory disorders [4][5][6][7][8][9][10][11][12][13]. Activated microglia, in turn, stimulate neurons to produce more Aβ and the microtubule-associated protein tau, in a vicious paracrine loop [14]. The events change the brain's microenvironment and further affect microglial activation, leading to progressive neuronal injuries. The end result is a paracrine feedback of neurotoxin amplification that drives disease with Aβ serving as the principal inducer of innate immune activation [15][16][17]. This has led many researchers to develop the means to harness immunity for therapeutic gain. Notably, Aβ vaccination can effectively clear Aβ in both mouse models of human disease and AD patients [18,19]. For treatment of AD, Aβ immunization reduces the Aβ42 load with immunomodulation, including microglial phagocytosis and changed lysosomal and scavenger markers [20,21]. However, such immunizations can also result in the development of meningoencephalitis (~6 %) [22], spongiosis, and neuronal loss [23]. Since removal of plaques at later stages of disease marked by neurofibrillary tangle formation was not proven to be beneficial, alternative treatment strategies are likely required. These might improve clinical outcomes if administered years before clinical signs and symptoms emerge. Indeed, such a therapeutic approach could speed clearance of Aβ and modify inflammatory activities before disease develops.
Prior therapeutic approaches focused on development of immunomodulatory agents that affect AD microglial responses. For example, nonsteroidal antiinflammatory drugs (NSAIDs) or antioxidants reduce harmful microglial inflammatory activities and protect neurons in animal models of human disease [24,25]. While the drugs are effective in converting microglial polarization from an M1 (classical activation) to an M2 (alternative activation) phenotype and attenuating neurotoxin production, successful human therapeutic translation remains out of reach [1,25,26].
While the "selectively non-selective" brain-penetrant mixed-lineage kinase type 3 (MLK3) inhibitor, URMC-099, can attenuate neuroinflammatory responses and facilitate the actions of long-acting nanoformulated antiretroviral drugs during human immunodeficiency virus-1 (HIV-1) infection [27][28][29], it is not known whether it could also harness autophagy. If operative, this could occur as a secondary response to phagocytosis of Aβ. Here, we show yet another novel therapeutic role of URMC-099 as an immune modulator in microglial inflammation and phagocytosis through accelerating phagolysosomal pathway Aβ degradation. In this manner, we posit that the drug could be developed as a novel AD therapeutic candidate.

Microglia isolation and cultivation
Housing, care, and breeding of non-transgenic mice (B6.129 hybrid background) were approved by the Institutional Animal Care and Use Committee of the University of Nebraska Medical Center. Primary cultured mouse microglia were prepared from the postnatal day 1 newborn mouse brains [30][31][32]. Meninges-free newborn mouse cortices were minced and trypsinized, followed by mechanical dissociation and filtration to remove tissue chunks. Cells were plated onto plastic tissue culture bottles as mixed glial cultures in Dulbecco's modified eagle medium supplemented with heat-inactivated 10 % fetal bovine serum, 50 μg/ml penicillin/streptomycin (Life Technologies, Carlsbad, CA, USA) and macrophage colony-stimulating factor (MCSF). Microglia were released from astrocytes in the tissue culture media by shaking. Non-adherent cells were collected 7-14 days after plating.

RNA extraction and transcript analyses
Total RNA was extracted from microglia using TRIzol (Life Technologies, Carlsbad, CA, USA). For PCR-based gene expression analyses, cDNA was synthesized with 1 μg of total RNA as a template using a Verso cDNA synthesis kit (Thermo Fisher Scientific, Waltham, MA, USA), and quantitative real-time RT-PCR (RT2-qPCR) was performed on a thermocycler (Mastercycler Gradient, Eppendorf Scientific Inc., Westbury, NY, USA) using 2x SYBR Green qPCR Master Mix (Biotool.com, Houston, TX, USA) and gene specific primer sets (Table 1). All primer sequences were obtained from PrimerBank (https://pga.mgh.harvard.edu/primerbank/) [35]. Thermal cycler conditions were as follows: 10 min at 95°C for activation of polymerase, followed by 40 cycles of a two-step PCR (95°C for 15 s and 60°C for 1 min). Relative expression for target genes was determined by the ΔΔCt method and normalized with glyceraldehyde-3phosphate dehydrogenase (Gapdh) gene expression as an internal control. Each ΔCt value was determined by subtracting Gapdh Ct value from the target gene Ct value. The ΔΔCt was calculated by subtracting the ΔCt value of the control from the ΔCt value of other groups. 2 −ΔΔCt represented the average relative amount of mRNA to control for each target gene.

Statistical analyses
All data were normally distributed and presented as mean values ± standard errors of the mean (SEM). In the case of single mean comparison, data were analyzed by Student's t test. In case of multiple mean comparisons, the data were analyzed by one-way ANOVA. When there were significant differences between ≥3 sample means, post hoc comparisons with the Newman-Keuls method was performed using statistics software (Prism 4.0, Graphpad Software, San Diego, CA, USA). A value of p < 0.05 was regarded as a significant difference.

Anti-inflammatory effects of URMC-099 in Aβ42-stimulated microglia
To determine the neuroinflammatory phenotype in URMC-099-treated microglia, total RNA was isolated and RT2-qPCR was performed for genes specific to IL-4 and IL-13. Data were shown as fold change compared to control (Fig. 3). While 30-min Aβ42 stimulation did not alter gene expression of IL-4 and IL-13, URMC-099 significantly induced this expression pattern (increases of 111.1 and 345.7 % in IL-4 and IL-13, respectively, Fig. 3), further supporting the anti-inflammatory effects of URMC-099 on Aβstimulated microglia.

URMC-099 alters scavenger receptor expression in microglia
A previous study demonstrated that microglial phagocytic capacity is linked to scavenger receptor (SR) expression levels in microglia in AD mouse models [45].
To pursue the mechanism of how URMC-099 facilitates Aβ phagocytosis, we examined the expression of SRs. CD36 and CD47 expression was assessed by immunoblotting (Fig. 5a) Fig. 5b), demonstrating URMC-099-specific alteration in SR expression.

URMC-099 increases Aβ co-localization with Rab7 and Lamp1
In response to Aβ binding to SRs, microglia start to engulf Aβ by phagocytosis, and then Aβ enters into the endolysosomal pathway. Thus, we investigated how URMC-099 affects the endolysosomal trafficking underlying Aβ phagocytosis. Microglia treated as described above were immunostained with antibodies to Rab7 (for late endosomes, Fig. 6a) and Lamp1 (for lysosomes, Fig. 6c) at 1-h post-incubation with Aβ. Confocal microscopy demonstrated that co-localization of Rab7 and Lamp1 with Aβ42 was increased in URMC-099-treated microglia, compared to untreated microglia (21.2 and 26.3 % increases in Rab7 and Lamp1, respectively, Fig. 6c, d). To investigate Aβ metabolism, microglia were exposed to Aβ42 for 30 min, washed and cultured in fresh media for additional 1 h, then harvested for immunoblotting (Fig. 7a). Co-treatment with URMC-099 significantly reduced immunoreactivity of monomeric, dimeric, and high molecular weight (HMW) Aβ42, as compared to Aβ42 treatment only (39.3, 30.6, and 42.3 % decreases in monomeric, dimeric, and HMW Aβ42, respectively, Fig. 7b). To validate these results, microglia were exposed to Aβ42 for 30 min, followed by wash and culture in fresh media for additional 1 h, and then Aβ42 release and retention in microglia were quantified using ELISA. While URMC-099 had no effect on Aβ42 release (Fig. 7c), co-treatment with URMC-099 significantly reduced Aβ42 with a 50.2 % reduction (Fig. 7d). These data suggest that URMC-099 promotes microglial endolysosome-mediated degradation.

Discussion
Microglia are the resident immune cells that serve as the first line of immune defense against invading pathogens in the CNS. The immune response is accelerated by the cell's abilities to recognize pathogen-associated microbial  (Table 1) and synthesized cDNA with total RNA isolated from murine microglia (n = 3 per group). Data are presented as mean ± SEM, a,b p < 0.05, a vs control, b vs Aβ, one-way ANOVA, Newman-Keuls post hoc test Fig. 4 URMC-099 facilitates microglial Aβ-uptake. a Primary mouse microglia were incubated with soluble Aβ42 for 30 min, followed by immunofluorescence with anti-Aβ Ab (green) and DAPI (blue) for nuclear staining (blue). Merged captured images were shown. Scale bar, 100 μm. b Dense intensity of Aβ42 fluorescence was measured using ImageJ. Bars represent mean ± SEM. **p < 0.01 by Student's t test  Bars represent mean ± SEM. *p < 0.05 by Student's t test patterns, induce key co-stimulatory molecules, and secrete cytokines that facilitate induction of the adaptive immune response in disease [46][47][48]. A principal microglial function is to act as scavenger cells to first phagocytose then clear substances such as Aβ [48][49][50]. Notably, Aβ is considered to be the initiating factor in AD inducing neuroinflammation, subsequent synaptic and axonal injuries, tau hyperphosphorylation, and ultimately neuronal death [15][16][17]51]. With these in mind, a number of therapeutic strategies have been developed for facilitating Aβ removal with the goal of improving disease outcomes. Indeed, anti-inflammatory drugs, amyloid degradation enzymes, and active/passive immunizations have been evaluated as treatment approaches [52][53][54][55]. In particular, NSAIDs and antioxidants are effective not only in attenuating the neurotoxins secreted by Aβ-stimulated microglia but also in facilitating microglial phagocytic activity for Aβ clearance in animal models of human disease [1,25,26]. To date, there has been no clear validation of any of these approaches for disease-modifying outcomes in patients with AD, thus further studies for alternative disease-modifying strategies are warranted. Hence, we investigated whether URMC-099 could ameliorate Aβ-mediated microglial activation. URMC-099 was discovered as a result of a large screening for MLK3 inhibitors structurally containing a pyrrolopyridine scaffold with an aryl piperazine side chain [28,29]; it can modify the production of pro-inflammatory mediators by inhibiting MAPK signaling cascades [36][37][38]. MLKs are MAPK kinase kinases (MAPKKKs) that function as bonafide serine/threonine kinases, although their catalytic domains have features of both serine/threonine and tyrosine kinases. MLKs regulate the p38/JNK signaling cascades that coordinate and orchestrate numerous immune processes [40]. URMC-099 is an inhibitor specific to MLK3 [28,29] that is one of the most widely expressed members of the MLK family [56]. MLK3 is expressed in immune effector cells including microglia in the CNS and is activated by cellular and metabolic stress [28,[57][58][59][60]. Based on these findings, we posit that URMC-099 is a potential candidate for AD therapeutics. Although it was previously unclear whether Aβ induces classical microglial activation through MLK-MAPK pathways, we showed that URMC-099 reduces phosphorylation of MKK3/MKK4 and p38/JNK and partially protected Aβ-mediated microglial cytotoxicity, suggesting its therapeutic effect targeting MLK-MKK-MAPK pathways in Aβ-mediated neuroinflammation in microglia.
In AD, microglia contribute to tissue injury through eliciting neuroinflammation and changing the CNS microenvironment. Like peripheral macrophages, microglia can be polarized into M1 and M2 phenotypes based on their functional properties [25,26,[61][62][63]. Aβ activation of microglia is associated with production of pro-inflammatory cytokines (IL-1β/IL-6/TNF-α), traits associated closely with the M1 phenotype [7,8,11,25,47,64]. Alternatively, polarization to an M2 phenotype can initiate an anti-inflammatory and repair phase [1,25,26]. The normal homeostatic balance between M1 and M2 phenotypes seems to be disturbed during disease progression in AD with the presence of more M1 microglia appearing coincident with aging and disease [26]. Herein, URMC-099 treatment was shown to inhibit IL-1β/IL-6/ TNF-α expression and coordinate with the up-regulation of IL-4 and IL-13 genes. These data suggest that URMC-099 treatment may initiate phenotypic changes by inducing anti-inflammatory cytokines, thus has neuroprotective activities [42][43][44]. a Microglia were exposed to Aβ42 for 30 min, followed by wash and culture in fresh media for additional 1 h, and harvested for immunoblot using a 10 % SDS-polyacrylamide Tris-Tricine gel and 6E10 antibody. 1-mer, monomeric Aβ42. 2-mer, dimeric Aβ42. Asterisk indicates high molecular weight (HMW) Aβ42. b Band luminescent intensities for monomeric, dimeric, and HMW Aβ42 were quantified by ImageJ software. The amounts of Aβ42 in culture media (c) and microglial cell lysates (d) were measured by human Aβ42-specific ELISA. Bars represent mean ± SEM. (n = 3 per group). *p < 0.05 vs Aβ, as determined by Student's t test Additionally, microglia in the M2 phase speed Aβ phagocytosis and degradation without neurotoxin production [41][42][43][44]. An increase in phagocytosis of Aβ was observed with URMC-099 treatment. Microglial phagocytic capacity is linked to SR expression levels [45]. CD36 is one of the SRs for Aβ and regulates Aβ clearance as well as brain inflammation [65]. While Aβ exposure increased CD36 expression similar to a previous study with BV-2 microglia [66], co-administration of URMC-099 with Aβ failed to reverse this up-regulation. In contrast, URMC-099 increases CD47 expression in microglia both alone and with Aβ. Since CD47 is an integrin-associated transmembrane protein and participates in Aβ uptake and microglial pro-inflammatory responses by forming the receptor complex CD36/ CD47/α6β1 integrin, which in turn stimulates microglial phagocytic activity [67][68][69]. By increasing CD47 activity, our results suggest that URMC-099 facilitates CD36/ CD47/α6β1-integrin-mediated microglial phagocytic activity of Aβ. Moreover, URMC-099 facilitates colocalization of Rab7 and Lamp1 with Aβ42, which may promote endolysosomal-mediated Aβ degradation and metabolism. Our data in aggregate suggests that these pathologic events can be altered with the challenge of how and when to optimally use disease-modifying agents such as URMC-099 to restore homeostasis to the brain's microenvironment.

Conclusions
URMC-099 inhibits Aβ-mediated phosphorylation of MKK3/4-p38/JNK and pro-inflammatory responses, and up-regulates phagolysosomal trafficking and degradation of Aβ. Such inhibition of phosphorylation by URMC-099 correlates with inhibition of microglial activation. URMC-099 serves as an immune modulatory and neuroprotective agent that may be developed to combat AD.