Extracellular CIRP Activates the IL-6Rα/STAT3/Cdk5 Pathway in Neurons

Extracellular cold-inducible RNA-binding protein (eCIRP) stimulates microglial inflammation causing neuronal damage during ischemic stroke and is a critical mediator of alcohol-induced cognitive impairment. However, the precise role of eCIRP in mediating neuroinflammation remains unknown. In this study, we report that eCIRP activates neurotoxic cyclin-dependent kinase-5 (Cdk5)/p25 through the induction of IL-6Rα/STAT3 pathway in neurons. Amyloid β (Aβ)-mediated neuronal stress, which is associated with Alzheimer’s disease, increased the levels of eCIRP released from BV2 microglial cells. The released eCIRP levels from BV2 cells increased 3.2-fold upon stimulation with conditioned medium from Neuro-2a (N2a) cells containing Aβ compared to control N2a supernatant in a time-dependent manner. Stimulation of N2a cells and primary neurons with eCIRP upregulated the neuronal Cdk5 activator p25 expression in a dose- and time-dependent manner. eCIRP directly induced neuronal STAT3 phosphorylation and p25 increase via its novel receptor IL-6Rα. Next, we showed using surface plasmon resonance that eCIRP-derived peptide C23 inhibited the binding of eCIRP to IL-6Rα at 25 μM, with a 40-fold increase in equilibrium dissociation constant (Kd) value (from 8.08 × 10−8 M to 3.43 × 10−6 M), and completely abrogated the binding at 50 μM. Finally, C23 reversed the eCIRP-induced increase in neuronal STAT3 phosphorylation and p25 levels. In conclusion, the current study demonstrates that the upregulation of neuronal IL-6Rα/STAT3/Cdk5 pathway is a key mechanism of eCIRP’s role in neuroinflammation and that C23 as a potent inhibitor of this pathway has translational potential in neurodegenerative pathologies controlled by eCIRP.


Introduction
Cold-inducible RNA-binding protein (CIRP) is an 18-kDa RNA-chaperone which is constitutively expressed and acts as a nuclear regulator of protein translation [1][2][3]. In addition to cold-shock and hypothermia, cellular stressors such as hypoxia and ultraviolet irradiation induce CIRP expression, nuclear to cytoplasmic translocation in stress granules, and extracellular release [4,5]. The released extracellular CIRP (eCIRP) has been identified as a danger-associated molecular pattern molecule (DAMP), promoting inflammation, tissue injury, and mortality in systemic and brain-specific inflammation in several studies from our lab [5,6]. In particular, we have shown that cerebral ischemia and alcohol exposure induce microglial expression of CIRP and release of eCIRP, which then acts as a neuroinflammatory mediator causing neuronal damage and death [7,8]. Recently, we showed that eCIRP also mediates alcohol-induced regional metabolic hypoactivity and memory impairment [9,10] and has a potential to be a mediator of Alzheimer's disease associated with alcohol consumption [11]. But, the precise mechanism of how eCIRP causes neuronal injury is still unknown.
Amyloid β (Aβ) peptides are produced from the neurons and are derived from the proteolytic cleavage of a much larger amyloid precursor protein (APP) [12,13]. APP isoform containing 695 amino acids is mainly expressed in neurons and is the most abundant form in the human brain [14]. In Alzheimer's disease, pathogenic forms of Aβ peptides accumulate as plaques in the brain which are closely associated with microglia [15]. Pathogenic Aβ forms are known to interact with and activate resting microglia in the brain causing the release of chemokines, proinflammatory cytokines, reactive oxygen species, and cytotoxins which are detrimental for surrounding neurons inducing neurotoxicity [15,16]. Whether exposure to Aβ can also cause eCIRP release from microglia has not been shown.
Neurotoxic insults cause calpain-mediated cleavage of p35, a regulator of the serine/threonine kinase cyclin-dependent kinase-5 (Cdk5), to generate a truncated carboxy-terminal fragment p25 [17]. Being more stable, the induction of p25 causes prolonged activation and mislocalization of Cdk5 resulting in tau hyperphosphorylation, neurite retraction, microtubule collapse, and apoptosis and is involved in the pathology of neurodegenerative diseases [18][19][20]. Moreover, several studies showed elevated p25 levels in the human brain correlate with Alzheimer's disease [21,22]. Signaling via the IL-6Rα/STAT3 pathway in neurons has been previously shown to deregulate Cdk5/p35 pathway [23]. Interestingly, we have recently shown that eCIRP directly binds to IL-6Rα and activates the IL-6Rα/STAT3 pathway in macrophages [24]. Therefore, we reasoned that eCIRP may also activate the IL-6Rα/STAT3/Cdk5 pathway in neurons.
In the present study, we used physiological forms of Aβ  and Aβ 1-40 overproduced and released from N2a neuroblastoma cells stably transfected with human APP695 to stimulate BV2 microglial cells and examined the effects of Aβ stress on the microglial release of eCIRP. We also evaluated the harmful effects of eCIRP on neurons by assessing the expression of neurotoxic mediator p25 in N2a cells and primary neurons stimulated with different doses of eCIRP over time. We then examined if eCIRP stimulation results into direct activation of IL-6Rα pathway in neurons. Next, we tested compound 23 (C23) peptide, a small 15-mer competitive antagonist of eCIRP activity, for its ability to physically block the interaction of eCIRP with IL-6Rα and its effect on IL-6Rα/STAT3 signaling and p25 expression. This approach not only further elaborated the role of eCIRP in regulating Cdk5-p25-mediated neuronal damage via IL-6Rα/STAT3 signaling but also evaluated the potential of a peptide inhibitor C23 for future therapeutic strategy targeting eCIRP in neurodegeneration in Alzheimer's disease.

Recombinant Proteins, Reagents, and Antibodies
Recombinant murine (rm) CIRP with His-tag expressed in E. coli was prepared in-house as described previously by us [6]. The quality control assays included LPS detection by limulus amebocyte lysate (LAL) assay (Cambrex, East Rutherford, NJ) and purity evaluation by Coomassie blue staining and Western blotting. The biological activity of the purified endotoxin-free protein was assessed by measuring the TNF-α levels released from rmCIRP-challenged macrophages. All the quality control assays were performed for each lot of purified rmCIRP. Culture mediums and cell culture reagents were purchased from MilliporeSigma (Burlington, MA) and Thermo Fisher Scientific (Waltham, MA

eCIRP ELISA
To measure the effect of neuronal Aβ stress on eCIRP release, released CIRP levels were measured in the conditioned medium samples from BV2 cells, treated for 6 h and 24 h with supernatants from untransfected N2a cells (Ctrl-N2a) or from N2a cells stably transfected with human APP constructs containing extracellular amyloid β (Aβ-N2a), by using ELISA kit (LifeSpan BioSciences) following the manufacturer's instructions.

Immunoblotting Analysis
The conditioned medium from untransfected or stably transfected with human APP695 N2a cells was incubated with 0.02% deoxycholic acid and 10% trichloroacetic acid (TCA) at 4°C overnight for protein precipitation and then subjected to Western blotting. Aβ levels in the TCA-precipitated supernatant were determined by Western blot analysis using anti-Aβ antibody (clone 6E10). N2a cells and primary neurons treated with various concentrations of rmCIRP with or without pretreatment with IL-6Rα neutralizing antibody or C23 peptide were harvested in lysis buffer (10 mM Tris-HCl at pH 7.5, 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100) containing protease inhibitor and phosphatase inhibitor cocktail tablet (Thermo Fisher Scientific), and total protein was extracted. The levels of Cdk5 activator p25 and phosphorylation status of STAT3 were determined by Western blot analysis using p-STAT3, STAT3, p25, and β-actin antibodies. Cell lysates were fractionated on 4-12% Bis-Tris gels and transferred to nitrocellulose membranes. After blocking with 0.1% casein in Tris-buffered saline, the membranes were incubated in respective primary antibodies overnight at 4°C. The target bands were detected by using infrared dyelabeled secondary antibodies and Odyssey CLx image system (Li-Cor BioSsciences). The intensities of the bands were analyzed using Image Studio 5.2 software (Li-Cor BioSciences).
The densitometric analysis of blots was done using ImageJ software.

Biacore Assay
Analysis of IL-6Rα-CIRP and C23 peptide-IL-6Rα-CIRP interactions was conducted using the surface plasmon resonance (SPR) technique on the Biacore 3000 instrument (GE Healthcare). Binding reactions were performed in 1 × HBS-EP (10 mM HEPES, 150 mM NaCl, 3 mM EDTA, 0.05% P20, pH 7. To evaluate nonspecific binding, the flow cell-1 without coating with the rhIL-6Rα protein was used as a control. The flow rate of 30 μl/min at 25°C was used to perform the binding analyses. To evaluate the binding, the analyte rhCIRP protein (OriGene) (500 nM for yes or no binding analysis or ranging from 62.5 to 500 nM with or without 25-50 μM C23 for the kinetics analysis) was injected into flow cell-1 and flow cell-2, and the association of analyte and ligand in the presence or absence of C23 was recorded, respectively, by SPR. The blank channel (flow cell-1) signal was subtracted from the ligand-coated channel (flow cell-2). Data were analyzed by the Biacore 3000 Evaluation Software. Data were globally fitted to the Langmuir model for a 1:1 binding.

Statistical Analysis
Data were analyzed using SigmaPlot12.5 graphing and statistical analysis software (Systat Software Inc., San Jose, CA) and presented as mean ± standard error of the mean (SEM).
One-way analysis of variance (ANOVA) and Student-Newman-Keuls' (SNK) test were performed for multigroup analysis. All data were tested for normality. For comparison of 2 groups, we performed unpaired 2-tailed Student's t tests.
Differences in values were considered significant if p < 0.05.

Microglial Cells Are a Major Source of eCIRP After Exposure to Aβ
We have previously shown that hypoxic stress and alcohol exposure increase microglial cell expression of CIRP and release of eCIRP [7,8]. To determine whether microglial cells act as source of brain eCIRP under Aβ-associated neuronal stress, we evaluated the effects of Aβ exposure on BV2 microglial cells. We first confirmed the presence of extracellular total Aβ in the TCA-precipitated conditioned medium (CM) from N2a cells stably transfected with human APP695 constructs (Aβ-N2a) and its absence in the CM from the control untransfected N2a (Ctrl-N2a) cells (Fig. 1a). We then cultured BV2 mouse microglial cells with CM from either transfected or untransfected N2a cells for 6 and 24 h. The Ctrl-N2a CM and Aβ-N2a CM had only baseline eCIRP levels comparable to CM of control BV2 cells (Fig. 1b), suggesting negligible effects of Aβ on the neuronal release of eCIRP. Aβ-containing CM induced BV2 cells to release eCIRP in a time-dependent manner, while Aβ-free CM had no effect (Fig. 1c). The eCIRP levels in the media of BV2 cells exposed to Aβ-containing CM for 6 h were 1.8-fold higher than in the media of cells exposed to Aβ-free CM, further increasing to 3.2-fold (p < 0.001) after 24-h exposure to Aβcontaining CM compared to Aβ-free CM (Fig. 1c). These findings suggest that microglial cells are a key source of eCIRP in the brain during neuronal stress caused by Aβ.
eCIRP Upregulates the Neuronal Cdk5 Activator p25 Cdk5 activity is regulated by unstable binding with the neuron-specific activator proteins p35. The more stable p25, generated from calpain-dependent cleavage of p35, is responsible for aberrant hyperactivation and deregulation of Cdk5 [17]. The generation of p25 is induced under neurotoxic stress conditions, and Cdk5-p25 complex promotes neurodegeneration and mediates further neurotoxicity [18][19][20]. To determine whether eCIRP stimulation could generate p25 in neurons, we treated N2a cells and primary neurons with increasing concentrations of eCIRP (0, 0.1, 1, and 2.5 μg/ml) for 48 h and analyzed p25 protein expression by Western blotting. We found that eCIRP significantly upregulated p25 in a dosedependent manner in both N2a cells (Fig. 2a) and primary neurons (Fig. 2b). After 48 h of eCIRP treatment in N2a cells, there was 1.9-fold induction of p25 at 1 μg/ml eCIRP, which further increased to 4.1-fold on stimulation with 2.5 μg/ml eCIRP (Fig. 2a). Primary neurons responded to 48 h CIRP stimulation with 49.9-fold induction of p25 at 1 μg/ml CIRP, going up to 136.2-fold with 2.5 μg/ml CIRP (Fig.  2b). Moreover, eCIRP time dependently induced upregulation of neuronal p25 in N2a cells (Fig. 2c) and primary neurons (Fig. 2d). After 16 h of CIRP treatment, N2a cells expressed 12.7-fold higher p25 than control cells which increased to 22.2-fold at 48-h stimulation (Fig. 2c). Similarly, primary neurons demonstrated 11.9-fold induction in p25 levels at 16 h of CIRP stimulation compared to no CIRP stimulation, further increasing to 26.3-fold at 48-h CIRP stimulation (Fig. 2d). Of note, in the absence of CIRP stimulation, there was no noticeable p25 expression in control N2a cells and primary neurons (Fig. 2a-d). These findings indicate that eCIRP upregulates the expression of neurotoxic truncated p25 form which is known to induce aberrant hyperactivation of Cdk5. Thus, the induction of p25 may be the mechanism by which eCIRP induces neuronal damage.

eCIRP Induces STAT3 Phosphorylation and p25 Via IL-6Rα
Next, we interrogated the role of IL-6Rα on the eCIRPinduced neuronal p25 expression. We stimulated N2a cells and primary neurons with 2.5 μg/ml eCIRP for 48 h plus either IgG or IL-6Rα neutralizing Abs. The eCIRP stimulation increased the p25 expression 19-fold which was suppressed by 38.9% on IL-6Rα neutralizing antibody pretreatment in N2a cells (Fig. 4a). Compared with the resting state, primary neurons stimulated with eCIRP induced 7.9-fold p25 which was decreased by 62.5% with IL-6Rα neutralization (Fig. 4b). These results  further support the concept that eCIRP induces the neuronal p25-Cdk5 pathway via the activation of the IL-6Rα/ STAT3 pathway.

C23 Inhibits the Direct Binding of eCIRP to IL-6Rα
Recently, we showed IL-6Rα to be a novel high-affinity receptor for eCIRP in macrophages [24]. We have also generated C23, a 15-amino acid eCIRP inhibitor peptide derived from human CIRP [6]. We postulated that C23 may inhibit eCIRP binding to IL-6Rα. We used SPR to determine whether different concentrations of C23 could disrupt the binding of eCIRP to IL-6Rα. As predicted, rhCIRP bound to rhIL-6Rα with high-affinity with an equilibrium dissociation constant (K d ) value of 8.08 × 10 −8 M (Fig. 5a), which is similar to K d 9.81 × 10 −8 M   (Fig. 5b). At the concentration of 50 μM, C23 completely abrogated the binding between eCIRP and IL-6Rα (Fig. 5c). These results provide direct evidence that C23 inhibits the binding of eCIRP to IL-6Rα and, thus, have the potential to also inhibit eCIRP's activation of the neuronal IL-6Rα/STAT3/Cdk5 pathway.

C23 Inhibits eCIRP's Induction of STAT3 Phosphorylation and Upregulation of p25
To evaluate C23's ability to inhibit eCIRP's biological activity in neurons, we pretreated N2a and primary neurons with C23 and then stimulated the cells with eCIRP. Stimulation with eCIRP induced STAT3 phosphorylation by 2.6-fold in N2a cells which was inhibited by 27.1% with C23 pretreatment (Fig. 6a). Interestingly, 1.96-fold induction of pSTAT3 and # for p < 0.05 vs. eCIRP plus IgG levels in primary neurons by eCIRP treatment was brought back to unstimulated levels by C23 (Fig. 6b). Along the lines of pathway, C23 also significantly prevented the eCIRPinduced upregulation of p25 in both N2a cells (from 19-fold to 11.5-fold, down by 39.5%; Fig. 6c) and primary neurons (from 8.9-fold to 5.4-fold, down by 40.1%; Fig. 6d). These data show that C23 not only disrupts eCIRP's binding to IL-6Rα but also blocks eCIRP's activation of the IL-6Rα/ STAT3/Cdk5 pathway. Thus, C23 is predicted to attenuate eCIRP's effect on the neuronal injury.

Discussion
In spite of advances in the management and care of patients suffering from neurodegenerative disorders including Alzheimer's disease, there is still no approved cure which poses an urgent need to identify new targets and design new therapeutic strategies. The neurodegeneration and dysfunction are essentially caused by the deregulation of multiple complex and diverse signal transduction pathways, many of which are required for normal regulated functions, which get abnormally triggered or repressed [31,32]. Studies from our and other labs have undoubtedly shown that eCIRP is a major contributing factor causing neuroinflammation, neuronal injury, and memory dysfunction in various settings [7-10, 33, 34]. However, the mechanism and signaling pathway involved in eCIRPmediated neuronal injury and dysfunction have not yet been elucidated. Certainly, understanding the signaling pathways activated by eCIRP in neurons causing dysregulation resulting in neuronal damage will impact the search for new targets and therapeutic strategies in treating neurodegeneration.
In the current study, we evaluated eCIRP as a key inducer of the IL-6Rα/STAT3/Cdk5 signaling pathway in neurons for exploring its potential to be a critical mediator of neurodegeneration. First, we confirmed the effect of established neurotoxic insult involved in neurodegeneration on eCIRP release and demonstrated that exposure of BV2 microglial cells to neuronal Aβ stress increased the microglial eCIRP release. Next, we identified eCIRP to be a critical activator of neuronal IL-6Rα/STAT3/Cdk5 pathway. For this, we first showed that eCIRP stimulation of neurons produced Cdk5 hyperactivator p25. We then illustrated that eCIRP directly interacted with IL-6Rα, leading to the activation of the downstream signal transducer STAT3 and upregulation of p25. Furthermore, we discovered that C23, a CIRP-derived 15-mer peptide, effectively inhibited eCIRP's interaction with IL-6Rα. Finally, pretreatment of neurons with C23 abrogated the eCIRPinduced downstream activation of the IL-6Rα/STAT3/Cdk5 pathway.
Alzheimer's disease pathogenesis has been associated with Aβ peptides in aggregated conformation form found in the plaques as Aβ fibrils, as well as non-aggregated, soluble Aβ forms which can also stimulate neuronal dysfunction [35]. The stimulation of microglia with synthetic Aβ monomers, oligomers, and fibrils has revealed very unique and graphs from the densitometric analysis of blots expressed as means ± SEM. ANOVA analysis showing * for p < 0.05 vs. no eCIRP and # for p < 0.05 vs. eCIRP + IgG. The no eCIRP controls in b are from the same blot but not in the continuous order differential activation profiles for these various forms each involving specific signaling pathway [36]. For this reason, we used conditioned medium containing Aβ released from N2a cells stably transfected with human APP695 instead of synthetic Aβ forms to stimulate microglial cells. This approach would include both Aβ 1-42 and Aβ 1-40 peptides in their various physiological forms to better mimic the Aβ species involved in disease condition. Aβ-mediated stress and activation induced the release of eCIRP from BV2 microglia, which is consistent with eCIRP being a stress response protein. We have shown previously that eCIRP was released from the BV2 cells exposed to either hypoxia [7] or high doses of alcohol [8] which supports the current finding. However, since we stimulated the BV2 microglia with CM from N2a cells stably transfected with human APP695 constructs, it is possible that the increase in eCIRP release from BV2 microglia may not have entirely resulted from Aβ in the CM.
Another soluble APP695 derivative secreted in that CM [37,38], other than or along with Aβ, could have contributed to the alterations in eCIRP levels. Therefore, future studies involving Aβ immunodepletion of the N2aAPP695 CM or treatments with synthetic Aβ will be helpful to confirm the Aβ induction of eCIRP release from microglia. We did not use primary microglia for this experiment because they yield relatively low cell numbers limiting the study. Moreover, primary cultured microglia are known to adopt phenotypes that are distinct from their in situ characteristics [39,40], which could influence or interfere with the eCIRP released from them. Further studies using in vivo Aβ models would warrant the eCIRP release from microglia.
Another DAMP, HMGB1, has also been shown to be extracellularly associated with Aβ plaques and released from neurons after Aβ stimulation [41,42]. However, the conditioned medium from the Aβ-producing N2a cell lines used in this study only showed basal eCIRP levels similar to the eCIRP levels from control BV2 cells. Further studies are needed to show the mechanism by which Aβ causes eCIRP release from microglia and if any other cell types in the brain could also release eCIRP on Aβ stimulation. It would be also interesting to explore whether additional neurotoxic insults involved in neurodegeneration other than Aβ such as pathological Tau forms could also induce eCIRP release. The detection of increased eCIRP levels in the cerebrospinal fluid from patients with neurodegenerative diseases would further establish the association of the upregulation of eCIRP release to neurodegeneration.
A variety of neurotoxic insults can cause p25 generation in neurons resulting in their damage and death [17]. This supports our finding that eCIRP stimulation of neurons lead to p25 production suggesting neurotoxic ability of eCIRP. Our initial results suggested that primary neurons might be more sensitive to eCIRP stimulation. However, the differences in stimulation scales were mostly related to differences in background of immunoblots and amount of protein loaded on gels Fig. 6 C23 inhibits STAT3 phosphorylation and p25 upregulation by eCIRP. N2a cells or primary neurons were pretreated with 25 μg/ml C23 and then treated with 2.5 μg/ml eCIRP for 1 h. C23 reduced eCIRP-induced STAT3 phosphorylation in a N2a neuronal cells (n = 6/group) and b primary neurons (n = 4/group). N2a cells or primary neurons were pretreated with 25 μg/ml C23 and then treated with 2.5 μg/ml eCIRP for 48 h. C23 reduced eCIRP-induced p25 upregulation in c N2a neuronal cells (n = 6/group) and d primary neurons (n = 4/group). Data shown as representative Western blot images and bar graphs from the densitometric analysis of blots expressed as means ± SEM. ANOVA analysis showing * for p < 0.05 vs. no eCIRP (0) and # for p < 0.05 vs. eCIRP alone in these set of experiments. Cdk5 activation is well-known to cause neuronal dysfunction including its role in increasing pathological tau phosphorylation and aggregation [43,44]. However, Cdk5 also plays a critical role in regulating signal transmission across neurons, synaptic plasticity, and cognition as part of the normal physiological brain function [44]. In addition, the inhibition of the Cdk5/p25 complex using small-molecule inhibitors has been reported to reduce neurodegeneration and improve cognition [18,20].
Of note, neurons exposed to Aβ fibrils also showed deregulated Cdk5 activity associated with Cdk5-p35 complex stability regulated by Cdk5 phosphorylation [45]. The Cdk5 hyperactivator p25 is generated through the proteolytic cleavage of p35 by calcium-activated calpain [46]. Additional studies are needed to evaluate the detailed mechanisms by which eCIRP upregulates p25, including triggering calpain proteolytic activity via increasing cytoplasmic calcium, which could further be confirmed using pharmacological or genetic inhibition of calpain [47]. Cdk5 is a major contributor to pathological tau phosphorylation [18,20], and the IL-6Rα/STAT3 pathway is known to induce tau hyperphosphorylation via Cdk5 [23,28]. eCIRP-induced IL-6Rα/STAT3/Cdk5 signaling could potentially lead to pathological tau phosphorylation and aggregation to cause functional neuronal injury. Future studies using in vivo and in vitro gain-and loss-of-function approaches and relevant tau mouse models to determine if eCIRP-induced activation of IL-6Rα/STAT3/Cdk5 signaling affects other aspects of neuronal function will be of great interest. Considering that other mechanisms may also exist for the regulation of Cdk5 activity [48], further studies showing eCIRP-mediated Cdk5 activation, pathological tau phosphorylation, and direct assessment of neuronal damage and death will be needed to confirm the indicated signaling outcome. Several proinflammatory cytokines such as IL-1β, IL-6, and TNF-α can also participate in neurodegeneration [49]. We considered IL-6 in particular since it uses the same signaling pathway. However, IL-6 levels were undetectable in the supernatants of N2a cells stimulated with eCIRP, indicating that eCIRP did not activate the IL-6Rα/STAT3/ Cdk5 pathway via induction of IL-6 release. Of note, eCIRP also induces neuroinflammation causing increased proinflammatory cytokine produced from microglia which would further add to the neuronal damage as an independent factor [7,8].
C23 peptide has been demonstrated to be protective in hemorrhage, renal ischemia/reperfusion, and sepsis models reducing inflammation and injury [6,[50][51][52][53]. C23 has been previously reported to work by competitively inhibiting eCIRP binding via binding to TLR4/MD2 complex with high affinity [6]. Remarkably at higher concentrations, C23 also inhibited physical interaction of eCIRP with IL-6Rα. Furthermore, C23 peptide also significantly decreased the eCIRP-induced IL-6Rα/ STAT3/p25 signaling. The ability of C23 peptide to abrogate neuronal damage due to hyperactive Cdk5 signaling and pathological tau phosphorylation needs to be further evaluated to demonstrate its potential to prevent neurodegeneration and may be a potential cognitive improvement. We also need to consider that along with IL-6Rα, eCIRP binds to a variety of other receptors such as TLR-4/MD2, RAGE, and TREM-1, many of which are present on neurons as well [11]. Further studies will be needed to carefully dissect whether and which additional receptors on neurons are activated by eCIRP, if a crosstalk exists among these receptors and what decides which receptor dominates the complex interplay.
In conclusion, our data provided in this study identifies that Aβ stress causes microglial cells to release eCIRP, which then directly activates the IL-6Rα/STAT3/Cdk5 signaling pathway in neurons. eCIRP does so by effectively inducing the generation of Cdk hyperactivator p25. In particular, the eCIRP inhibitor C23, previously known to antagonize eCIRP's binding to the TLR4-MD2 complex, also inhibits physical interaction of eCIRP with IL-6Rα and thus abrogates the eCIRP-induced increase in neuronal p25. Thus, eCIRP is a novel mediator of neuronal dysfunction via IL-6Rα/STAT3/Cdk5 activation which can be mitigated by C23 providing a novel targeting strategy in neurodegeneration in Alzheimer's disease.
Author Contribution AS, MB, and PW conceived and designed the experiments; AS performed all the experiments and analyzed the data; AJ helped with some immunoblots; AS wrote the manuscript; MB and AJ edited and reviewed the manuscript; PM provided reagents and critical input in some experimental designs; PW and PM critically reviewed the manuscript; and PW supervised the research. All authors read and approved the final manuscript.
Funding This work was supported by the National Institutes of Health grants R35 GM118337 (PW) and R01 AA028947 (PW, PM).
Data Availability All data generated or analyzed during this study are included in this article.

Declarations
Ethics Approval The study design of the mouse primary cortical neuron isolation experiments was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at the Feinstein Institute for Medical Research.

Consent for Publication Not applicable
Conflicts of Interest PW is an inventor of patent applications WO/2010/ 120726 and 61/881.798 covering targeting cold-inducible RNA-binding protein with peptides, licensed by TheraSource LLC. PW is a cofounder of TheraSource LLC.