TPL2 kinase activity regulates microglial inflammatory responses and promotes neurodegeneration in tauopathy mice

TPL2 (MAP3K8) is a central signaling node in the inflammatory response of peripheral immune cells. We find that TPL2 kinase activity modulates microglial cytokine release and is required for microglia-mediated neuron death in vitro. In acute in vivo neuroinflammation settings, TPL2 kinase activity regulates microglia activation states and brain cytokine levels. In a tauopathy model of chronic neurodegeneration, loss of TPL2 kinase activity reduces neuroinflammation and rescues synapse loss, brain volume loss, and behavioral deficits. Single-cell RNAseq analysis indicates protection in the tauopathy model was associated with reductions in activated microglia subpopulations as well as infiltrating peripheral immune cells. Overall, using various models, we find that TPL2 kinase activity can promote multiple harmful consequences of microglial activation in the brain including cytokine release, iNOS induction, astrocyte activation, and immune cell infiltration. Consequently, inhibiting TPL2 kinase activity could represent a potential therapeutic strategy in neurodegenerative conditions.


99
(IFNg), a stimulus reported to induce strong microglia activation (Straccia et al., 2011), resulted in 134 almost complete neuronal loss as measured by MAP2 immunostaining (Fig. 2A,B). Strikingly, 135 when the TPL2 inhibitor G-767 was added to the co-culture, the neuronal loss induced by the LPS 136 + IFNg stimulation was nearly fully rescued. In addition, when WT neurons were co-cultured with 137 microglia isolated from TPL2-KD mice, stimulation-induced neuronal loss was also rescued (Fig. 138 2A,B). Of note, when neurons were cultured alone and stimulated with LPS + IFNg or co-cultured 139 with microglia at 1:1 ratio without stimulation, no obvious neuronal loss was observed 140 (supplementary Fig. S4).

141
We next wanted to understand the mechanism of neuronal death and the neuroprotective 142 effect of TPL2 kinase inhibition in this co-culture system. Several groups have shown that LPS +

156
We next examined the role of TPL2 in vivo using an acute neuroinflammation model. We injected In addition to directly affecting neurons, microglia have also been shown to indirectly 167 cause neurotoxicity by activating astrocytes. Cytokines released by microglia in response to LPS 168 stimulation can induce the transition of astrocytes from a resting state to an activated/reactive state 169 that can be neurotoxic (Liddelow et al., 2017). Consistent with a reduction of microglia activation 170 of astrocytes, LPS-induced activated astrocyte gene expression (e.g., Cd44, Srgn) was also 171 attenuated in TPL2-KD mice (Fig. 3B,D,E). In parallel to the RNAseq measurements, we also 172 measured brain cytokine levels and found that several cytokines such as IL-1a, IL-6, and CXCL1 173 that were elevated in LPS treated WT mouse brains, had significantly lower levels in TPL2-KD 174 mouse brains (Fig. 3F). Collectively, these results indicate TPL2 plays a key role in microglia and 175 astrocyte activation in vivo, and show that acute neuroinflammation can be reduced by ablation of 176 TPL2 kinase activation.

177
We next investigated if TPL2 kinase deficiency could protect against neuronal damage in 178 acute injury models. We first tested the mouse optic nerve crush (ONC) model. However, although 179 we observed moderate reduction of microgliosis in TPL2-KD mice after ONC, we did not see

197
% area of the brain covered by Iba1 or GFAP in non-transgenic mice. In TauP301S mice, the Iba1 198 and GFAP signals were greatly increased reflecting gliosis, and TPL2 kinase deficiency slightly 199 ameliorated this gliosis (Fig. 4A,B). The effect was observed with either whole brain section 200 analysis or hippocampal region analysis. Interestingly, phospho-Tau pathology (measured by AT8 201 staining) was also somewhat reduced in TauP301S;TPL2KD mice ( Fig. 4A,B), potentially 202 reflecting the interplay between neuroinflammation and Tau pathology (Didonna, 2020;Maphis et 203 al., 2015). Consistent with the gliosis changes, the levels of several cytokines and chemokines that 204 were significantly increased in TauP301S mouse brains were blunted in TauP301S;TPL2KD   205 brains, including CXCL9, CXCL10, IL6 and IFNg (Fig. 4C).

255
To further analyze microglia phenotypes, the ~55,000 microglia in the study were sub-256 clustered into 22 clusters (Fig. 6D). As expected, clusters of microglia annotated as 257 neurodegeneration-related and interferon-related were elevated, and clusters annotated as 258 resting/homeostatic were decreased in TauP301S mice (supplementary Fig. S11, S12). However, 259 the abundance of these clusters in P301S mice were not altered by TPL2KD (supplementary Fig. 260 S11, S12). On the other hand, a subcluster of microglia (C11) expressing high levels of IEGs (e.g.

261
Fos, Egr1-3, Atf3) (supplementary Fig. S11E), was elevated in P301S mice, and this elevation was 262 ameliorated by TPL2KD (Fig 6E). Calculation of IEG gene set score for every defined cell type 263 (pseudo-bulk) showed IEG expression was up in astrocytes, microglia, oligodendrocytes, OPCs  volumes stayed stable and neocortex and whole brain volumes increased between 6 and 9 months, 291 and TPL2 kinase deficiency did not alter the trajectory of the brain volume changes. As expected,

292
P301S mice developed brain atrophy, as indicated by increasing ventricle volumes and decreasing 293 brain volumes between 6 and 9 months. Strikingly, TPL2KD significantly ameliorated the ventricle 294 enlargement and brain volume decline in TauP301S mice (Fig. 7B,C), indicating a robust 295 neuroprotective effect of inhibition of TPL2 kinase activity in this mouse model. We also measured 296 plasma NfL (neurofilament light chain) to see if this potential biomarker of neurodegeneration was 297 reduced in parallel with the protection against brain volume loss. However, while plasma NfL was 298 strongly increased in P301S mice, TPL2KD did not significantly affect NfL levels (Fig. 7D). This

334
Based on the potentially beneficial effects of reduced TPL2 kinase function in response to 335 direct stimuli in vitro and in vivo, we tested the effects of TPL2KD in in vivo neurodegeneration models. That TPL2KD lessened gliosis but did not provide neuroprotection in optic nerve crush 337 and stroke models is perhaps not surprising given the direct severe neuronal damage in these models

368
Overall, our data indicate that reducing the inflammatory state of the brain with approaches 369 such as targeting TPL2 kinase activity could be a potential therapeutic approach for diseases with      in an incubator at 37°C with 5% CO2 and the medium was renewed using 50% exchange every 3-438 4 days to maintain cell health.

439
Primary microglia were cultured from P0-P2 pups as previously described (Yeh et al., 440 2016). Briefly, mouse brains were dissected out and the brain tissue was disrupted by trituration 441 using a 10 mL serological pipette in cold DMEM media. The homogenate was spun at 300 g for 5 442 min. The pellet was resuspended in DMEM media and filtered through a 70 µm cell strainer.

482
The fastq sequence files for all RNA-seq samples were filtered for read quality (keeping 483 reads where at least 70% of the cycles had Phred scores ≥ 23), and ribosomal RNA contamination.

484
The remaining reads were aligned to the mouse reference genome (GRCm38) using the GSNAP          (log2(normCount+1)). All gene sets used in this study are available in Table-S1.

536
T cells were extracted from the dataset, re-normalized and the 3,000 most variable genes 537 were identified. Dimensionality reduction on the variable genes was performed using principal

568
This was defined as genes with at least 10 total UMIs in at least 3 of the analyzed pseudobulks.

569
Gene set scores for each pseudo-bulk profile were calculated as described above for ScRNA-seq 570 data, except NormCount values were used in the place of nUMI values.

571
Cellularity (cellular composition) plots are plotted using geom_boxplot 572 (https://ggplot2.tidyverse.org/reference/geom_boxplot.html). The lower and upper hinges 573 correspond to the first and third quartiles (the 25th and 75th percentiles). The upper whisker extends 574 from the hinge to the largest value no further than 1.5 * IQR from the hinge (where IQR is the inter-575 quartile range, or distance between the first and third quartiles). The lower whisker extends from 576 the hinge to the smallest value at most 1.5 * IQR of the hinge. Data beyond the end of the whiskers 577 are called "outlying" points and are plotted individually (Cleveland and McGill, 1985). P values 578 are based on t-test between indicated groups using ggpubr. 128, 1 average, with a scan time of 11 min/mouse. During imaging, anesthesia for mouse was 600 maintained at 1.5% isoflurane and rectal temperature was maintained at 37 ± 1 o C using a feedback 601 system with warm air (SA Instruments, Stony Brook, NY). Equal number of males and females 602 were included to detect any gender difference. The regional and voxel differences in the brain 603 structure were evaluated by registration-based region of interest (ROI) analysis. In brief, multiple 604 echo images were averaged and corrected for field inhomogeneity to maximize the contrast to noise ratio and the images were analyzed based on a 20-region pre-defined in-vivo mouse atlas 606 (http://brainatlas.mbi.ufl.edu/) that was co-registered to a study template and warped to individual 607 mouse datasets. All the co-registration steps were performed in SPM8 (Wellcome Trust Centre   Associates). Each chamber was equipped with a house light, an IR light, a speaker that was used to 637 deliver white noise during the training and cued phases of the task, and a near-IR camera to record 638 the movement and/or freezing of the mice. The floors of the compartments were equipped with stainless steel metal grids that were connected to a shock generator. Hardware and data acquisition 640 were controlled by Video-Freeze software (Med Associates).

641
During TFC training, the house lights were on and the metal grids served as the floor. TFC 642 training began with 180 s of baseline recording in which the mice were allowed to explore the 643 chamber, followed by 6 CS-US trials, each consisting of a 20 s tone (90 dB white noise) as the 644 conditioned stimulus, followed by an 18 s trace interval, then a 2 s foot shock (0.5 mA) as the 645 unconditioned stimulus, and finally a 30 s delay. After the 6 CS-US presentations were completed, 646 the mice were returned to their home cage.

647
The TFC test was administered 24 hours following training. During TFC testing, the         induce changes in cell morphology, and upregulation of ERK1/2, iNOS and sPLA(2)-907