Loss of HuD Sensitizes Neuroblastoma Cells to Palmitate-Driven Stress-Induced Premature Senescence via PPARα Downregulation and FAO Impairment

Abstract

Metabolic stress caused by lipid overload is a key driver of cellular dysfunction in aging and disease. Excess saturated fatty acids such as palmitate impair fatty acid oxidation (FAO), promote lipid accumulation, and increase reactive oxygen species (ROS), ultimately triggering premature senescence-like states. Senescence further amplifies vulnerability by worsening mitochondrial dysfunction, enhancing lipid imbalance, and sustaining pro-inflammatory signaling. Here, we investigated the role of the neuron-enriched RNA-binding protein HuD (ELAVL4) in protecting cells against lipotoxic stress. Using Neuro2a neuroblastoma cells, we found that HuD knockdown suppressed FAO, leading to increased lipid accumulation and elevated ROS following palmitate exposure. HuD-deficient cells also exhibited cytosolic mitochondrial DNA release, IRF phosphorylation, and upregulation of senescence markers. Mechanistically, RNA immunoprecipitation revealed that HuD binds directly to PPARα mRNA, sustaining its expression by competing with the PPARα-targeting microRNAs miR-9-5p and miR-22-3p. Loss of HuD reduced PPARα levels, thereby weakening the FAO capacity and sensitizing cells to palmitate-induced lipotoxic stress. These findings identify a previously unrecognized HuD–PPARα–FAO axis that restrains metabolic stress and senescence. By linking post-transcriptional regulation to lipid metabolism and inflammatory signaling, this work highlights stress-induced premature senescence as both an outcome and a propagator of metabolic dysfunction, providing insight into mechanisms of aging-related vulnerability.

1. Introduction

Cells subjected to metabolic stress exhibit profound alterations in mitochondrial function and lipid metabolism, resulting in destabilized energy balance and heightened oxidative stress [1,2]. Saturated fatty acids such as palmitate (PA) are commonly used to model lipotoxic stress, where their excessive incorporation disrupts fatty acid oxidation (FAO), induces abnormal lipid accumulation, and promotes reactive oxygen species (ROS) generation [2,3,4,5]. These perturbations not only compromise immediate cellular function but also set the stage for long-term maladaptive responses that are increasingly relevant to aging and neurodegeneration [6,7,8].

One such maladaptive response is cellular senescence, traditionally defined as a permanent growth arrest program characterized by chromatin remodeling and a senescence-associated secretory phenotype (SASP) [9,10]. Under metabolic and oxidative stress, even non-dividing or neuron-derived cells can undergo a premature senescence-like transition marked by mitochondrial dysfunction, cytosolic DNA leakage, and activation of innate immune signaling [11]. Importantly, senescence should not be regarded as a passive endpoint alone: once established, it actively reshapes cellular physiology. Senescent cells accumulate lipids due to impaired FAO, exhibit defective mitochondrial dynamics, and secrete pro-inflammatory SASP factors that amplify local and systemic inflammation [12,13]. In this way, senescence not only arises from metabolic stress but also feeds forward as a causal driver of chronic metabolic fragility and inflammatory microenvironments, reinforcing a vicious cycle of cellular decline.

The nuclear receptor peroxisome proliferator-activated receptor α (PPARα) is a key regulator of FAO and represents a critical node for protection against lipotoxic stress [14]. By driving fatty acid catabolism, PPARα activity counteracts lipid accumulation, constrains ROS generation, and preserves mitochondrial homeostasis [15]. While PPARα has been extensively studied in peripheral tissues, its regulation in neuron-derived systems and its potential role in buffering against senescence-associated changes remain poorly defined. Reduced PPARα activity may thus represent an initiating event that links lipotoxic stress to senescence and its downstream consequences.

HuD (ELAVL4) is a neuronally enriched RNA-binding protein that enhances the stability and translation of target mRNAs [16]. Its role in neuronal differentiation and plasticity is well established [17,18], but its role in metabolic stress adaptation has not been addressed. Here, using Neuro2a neuroblastoma cells, we identify HuD as a post-transcriptional regulator of PPARα. We show that HuD binds Pparα mRNA and protects it from microRNA-mediated repression, thereby sustaining FAO capacity. HuD knockdown decreases PPARα expression; enhances palmitate-induced lipid accumulation and ROS production; and promotes cytosolic mtDNA release, phosphorylation of TBK1 and IRF3, and the induction of senescence markers. These findings establish a novel HuD–PPARα–FAO axis that protects cells from palmitate-driven metabolic stress and restrains premature senescence-like responses, highlighting senescence not only as an outcome but also as a propagator of metabolic vulnerability and inflammation in aging-related contexts.

2. Materials and Methods

2.1. Cell Culture, Transfections, and Reagents

Mouse neuroblastoma Neuro2a (N2a, ATCC CCL-131) cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Cytiva, Marlborough, MA, USA) supplemented with 10% fetal bovine serum (FBS; Cytiva) and 1% penicillin–streptomycin (Cytiva) at 37 °C under 5% CO₂. N2a cells were produced that stably expressed shHuD (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) or control shRNA, as previously described [19]. Cells were transfected with small interfering RNAs (siRNAs) (Genolution Pharmaceuticals, Inc., Seoul, Republic of Korea), microRNAs (miRNAs) (BIONEER, Daejeon, Republic of Korea), or pHuD (pCMV6-AN-His-HA-HuD) [20] using Lipofectamine™ 2000 (Invitrogen™, Waltham, MA, USA) according to manufacturer’s instructions. Cells were harvested 48 h after transfection. Oligomer sequence information is listed in Table S1. Palmitate (PA) was dissolved in 0.1 N NaOH at 70 °C and conjugated with 1% bovine serum albumin (BSA) at 55 °C to prepare 10 mM stock solution, and then treated with culture medium prior to use.

2.2. Immunoblot Analysis

Cell lysates were prepared in RIPA buffer supplemented with protease inhibitor cocktail (Roche, Basel, Switzerland) and phosphatase inhibitor (FIVEphoton Biochemicals™, San Diego, CA, USA). Protein concentrations in whole-cell lysates were determined by Bradford assay (Bio-Rad, Inc., Hercules, CA, USA). Protein samples were separated by SDS polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Burlington, MA, USA). After blocking, membranes were sequentially incubated with primary antibodies and horseradish peroxidase (HRP)-conjugated secondary antibodies. Luminescence signals were visualized using Clarity Western ECL Substrate (Bio-Rad, Inc.) on ChemiDoc Imaging System (Bio-Rad, Inc.). Band intensity was quantified using Image Lab software version 6.1. Following primary antibodies were used: β-actin (GeneTex, Inc., Irvine, CA, USA; GTX109639), PPARα (Proteintech, Inc., Rosemont, IL, USA; 66826-1-Ig), TBK1 (Cell Signaling Technology, Inc., Danvers, MA, USA; #3504), p-TBK1 (#5483), IRF3 (#4302), p-IRF3 (#79945), HuD (Santa Cruz Biotechnology, Inc., sc-28299), Lamin B (sc-6216), p16 (sc-1661), p21 (BD Bioscience, Franklin Lakes, NJ, USA; #556431), HA (BioLegend, San Diego, CA, USA; 901501).

2.3. Measurement of Oxygen Consumption Rate

The oxygen consumption rate (OCR) was assessed using a Seahorse XFe24 Extracellular Flux Analyzer (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s protocol and as previously described [21]. Briefly, the cells were seeded in XFe24 cell culture microplates and pretreated with or without etomoxir (40 μM) for 18 h. Before the assay, the culture medium was replaced with Seahorse assay medium, and the cells were incubated in a CO₂-free incubator for 1 h. The OCR was sequentially measured after the addition of oligomycin A (1 μM), carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP, 2 μM), rotenone (1 μM), and antimycin A (1 μM; Sigma-Aldrich, Burlington, MA, USA). The measurements were recorded for 100 min for each mixing and measuring cycle, and the data were analyzed using Wave software (Agilent Technologies). The FAO-dependent OCR (ΔOCR) was calculated as the difference between the OCR values of untreated and etomoxir-treated cells (OCR_control − OCR_etomoxir).

2.4. RNA Analysis

Total RNA was extracted using TRIzol reagent (Takara Bio Inc., Shiga, Japan). cDNA was synthesized from 1 μg of total RNA using PrimeScriptTM RT Reagent Kit (Takara Bio Inc., Shiga, Japan) and used for real-time (RT) quantitative PCR (qPCR). RNA quantification was done by relative quantification methods [22]. Gapdh mRNA was used as reference gene for relative quantification. Primer sequence information is listed in Table S1.

For RNA immunoprecipitation, ribonucleoprotein (RNP) complexes were precipitated using anti-HuD antibody or normal mouse IgG (Santa Cruz Biotechnology, Inc.) conjugated to protein A beads (Invitrogen™). RNAs were isolated from complex by sequential incubation with DNase I and proteinase K, and synthesized to cDNA [19]. Enriched mRNAs in HuD-containing RNP complex were analyzed by RT-qPCR compared to normal IgG.

2.5. Biotin Pull-Down Assay

The DNA fragments corresponding to the 3′UTR of mouse Pparα mRNA (NM_011144.6) containing the predicted binding sites for miR-9-5p or miR-22-3p were amplified by PCR using the specific primers listed in Table S1. The PCR products were transcribed in vitro using a MaxiScript T7 kit (Invitrogen™) in the presence of biotin-CTP (Enzo Life Sciences, Inc., Farmingdale, NY, USA) to generate biotinylated RNA probes [19]. The cell lysates were incubated with the purified probe for 30 min at room temperature and the RNA–protein complexes were pulled down using streptavidin-conjugated Dynabeads (Invitrogen™). The bound proteins were eluted from the beads and analyzed by immunoblotting with an anti-HuD antibody. The 3′UTR of Gapdh mRNA was used as a negative control.

2.6. Measurement of Reactive Oxygen Species (ROS)

Intracellular ROS and mitochondrial superoxide levels were detected using 2′,7′-dichlorofluorescin diacetate (DCFDA; Invitrogen™) and MitoSox™ Red (Invitrogen™), respectively, according to manufacturer’s instructions. Briefly, cells were incubated with 5 μM dye in Hank’s balanced salt solution (HBSS) at 37 °C; cells were trypsinized and analyzed by flow cytometry using BD FACSCanto™ system (BD Bioscience). Fluorescence intensity was quantified for each sample and was analyzed using FlowJo™ software version 10 (BD Biosciences).

2.7. Cell Staining and Microscopy

The cells were fixed with 4% paraformaldehyde and permeabilized with 0.4% Triton X-100. To stain the lipid droplets, the cells were incubated with 0.1 μM Nile Red (Sigma-Aldrich) in the dark for 30 min. After counterstaining the nuclei with DAPI solution, fluorescence images were acquired using a ZEISS Axio Observer microscope (Carl Zeiss, Oberkochen, Germany). The quantitative analysis of each image was done using ImageJ software version 1.54p. For the senescence-associated β-galactosidase (SA β-gal) staining, the cells were incubated in an X-gal-containing solution (BEAMS Biotechnology, Seongnam, Republic of Korea) at 37 °C for 16 h in the dark and observed using an EVOS™ M5000 Imaging System (Invitrogen™). For the FAOBlue staining, the cells were incubated with 10 μM FAOBlue (Funakoshi Co., Ltd., Tokyo, Japan) at 37 °C for 1 h. The fluorescence intensity was measured using a BioTek Synergy H1 microplate reader (Agilent Technologies) at Ex/Em = 405/460 nm. The cell viability was determined using a Cell Counting Kit-8 (CCK-8; DOJINDO Laboratories, Kumamoto, Japan) for normalization.

2.8. Measuring Mitochondrial DNA in Cytosolic Extracts

The mitochondrial DNA (mtDNA) in the cytosolic extracts was assessed as previously described in [23]. In brief, the cells were lysed in a cytosolic extraction buffer containing 150 mM NaCl, 50 mM HEPES (pH 7.4), and 20 μg/mL digitonin. The DNA from the cytosolic fraction or whole-cell lysates was extracted using an AccuPrep^® Genomic DNA Extraction Kit (BIONEER). A qPCR was performed to measure the mtDNA levels using primers specific for MT-CytB, MT-ND4, and MT-Dloop. 18s rRNA was used as a reference gene.

2.9. Measurement of Mitochondrial ATP, Mitochondrial Membrane Potential, and Triglycerides

The mitochondrial ATP levels and mitochondrial membrane potential (ΔΨm) were determined by using a Mitochondrial ToxGlo™ Assay (Promega, Madison, WI, USA; G8000) and a JC-1 Mitochondrial Membrane Potential Assay Kit (Abcam Plc., Cambridge, UK; ab113850), respectively, according to the manufacturer’s instructions. The luminescence of the ToxGlo™ Assay was detected using a BioTek Synergy H1 microplate reader (Agilent Technologies). The fluorescence of JC-1 staining was detected at Ex/Em = 530/590 nm.

The intracellular triglyceride levels were assessed with a Triglyceride Assay Kit (Abcam Plc., ab65336). The fluorescence was detected at Ex/Em = 535/587 nm according to the manufacturer’s instruction.

2.10. Statistical Analysis

The data are presented as the mean ± SD from three independent experiments. The statistical analyses were performed using GraphPad Prism software (Version 8), and the significance was determined using a Student’s t-test, one-way ANOVA or two-way ANOVA test (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001).

3. Results

3.1. Loss of HuD Impairs FAO and Exacerbates Lipid Droplet Accumulation in Palmitate-Treated Neuroblastoma Cells

To investigate whether HuD regulates FAO in neuronal cells, we downregulated HuD in N2a neuroblastoma using two independent knockdown strategies—transient pooled siRNA and an antibiotic-selected stable cell line constitutively expressing shRNA—and quantified the lipid handling and FAO (Supplementary Figure S1A). Upon palmitate exposure, we observed increased Nile Red fluorescence that was further amplified by HuD knockdown (KD); we reproduced the same pattern in a stable shHuD cell line (Figure 1A), indicating that loss of HuD potentiates lipid accumulation under palmitate challenge. Consistent with the imaging data, we detected a higher triglyceride content in the HuD KD cells by the triacylglycerol assays (Figure 1B), supporting a shift toward esterification and storage. These findings collectively suggest that HuD restrains lipid accumulation in neuroblastoma. To determine whether the lipid accumulation in Figure 1A,B reflects impaired lipid degradation, we assessed FAO in N2a cells. We found that FAOBlue staining was moderately, but significantly, reduced after HuD KD, indicating decreased cellular FAO activity (Figure 1C). Consistent with this finding, we performed Seahorse respirometry following treatment with the FAO inhibitor etomoxir and observed lower FAO-linked oxygen consumption (ΔOCR) at the baseline, with a further decrease after FCCP-driven uncoupling in the HuD-deficient cells (Figure 1D). Together, these results suggest that HuD promotes mitochondrial FAO in neuroblastoma cells, thereby limiting palmitate-induced lipid accumulation.

3.2. HuD Binds Pparα mRNA and Supports Its Expression in Neuroblastoma Cells

Because HuD is an RNA-binding protein [16] and HuD KD increases lipid accumulation while reducing FAO in N2a cells (Figure 1), we next investigated whether HuD directly binds to and regulates the expression of transcripts involved in lipid metabolism. Using RNA immunoprecipitation followed by an RT–qPCR with an anti-HuD antibody (Figure 2A and Figure S1), we examined the interaction between HuD and a panel of canonical FAO-related transcripts, including Pparα, Cpt1, Cpt2, Acsl1, Acadl, Acadm, and Acads. Among the candidates tested, we found significant enrichment of Pparα mRNA in the HuD IP relative to the control IgG, indicating a specific association between HuD and Pparα transcripts (Figure 2B). To determine whether this interaction affects Pparα abundance, we quantified the Pparα expression after HuD downregulation using siRNA. Notably, we observed a reduction in the Pparα protein levels in the N2a cells upon HuD KD (Figure 2C). These data suggest that Pparα mRNA is a HuD-associated transcript, supporting a model in which HuD regulates Pparα expression and maintains the FAO program in neuroblastoma cells.

3.3. HuD Counteracts miR-9-5p/miR-22-3p-Mediated Repression of Pparα via Its 3′UTR

Because HuD binds Pparα mRNA and positively regulates its expression (Figure 2), we investigated whether HuD modulates this axis through interactions on the 3′UTR of Pparα mRNA with microRNAs, negative regulators of gene expression. The Pparα 3′UTR (~5.6 kb) contains multiple predicted miRNA sites (TargetScan; Figure S2A), and we observed increased miR-9-5p and miR-22-3p levels in the N2a cells after palmitate exposure (Figure S2B). Therefore, we assessed the Pparα level after transfection of miR-9-5p and miR-22-3p and found that overexpression of either miRNA reduced Pparα protein abundance (Figure 3A). To map the potential sites of competition, an in silico analysis identified putative HuD-binding motifs [24] near the miR-9-5p/miR-22-3p binding sites within the Pparα 3′UTR (Figure 3B and Figure S2C). We then performed a biotin pull-down assay using biotinylated RNA probes corresponding to each miR-binding region and detected HuD binding in proximity to these sites (Figure 3C), supporting a physical basis for site-level interplay. Consistent with the competition on the endogenous transcript, RNA immunoprecipitation followed by an RT-qPCR showed that the HuD–Pparα mRNA association decreased upon miR-9-5p or miR-22-3p overexpression (Figure 3D). Functionally, co-transfection of HuD siRNA with miR-9-5p or miR-22-3p precursors produced a greater reduction in the Pparα protein than either manipulation alone (Figure 3E), consistent with competitive repression. Together, these data indicate that HuD positively regulates Pparα expression by limiting miR-9-5/miR-22-3p-mediated suppression on the Pparα 3′UTR.

3.4. HuD-Pparα Axis Links FAO Loss to Mitochondrial Stress and cGAS-STING-IRF3 Activation Under Palmitate Challenge

PPARα is a key transcriptional regulator of FAO and related mitochondrial function and its downregulation is known to compromise mitochondrial fat oxidation and bioenergetic homeostasis in multiple tissues [14,15]. Thus, we asked whether reduced PPARα/FAO by HuD KD underlies mitochondrial stress in palmitate-treated N2a cells. We detected a significant increase in mitochondrial superoxide by MitoSox™ and a larger increase in the total cellular ROS by DCFDA in the HuD KD cells (Figure 4A,B), indicating enhanced oxidative stress when HuD is downregulated. To evaluate the mitochondrial polarization, we quantified the JC-1 aggregate/monomer ratio and found that it decreased with HuD KD, consistent with loss of mitochondrial membrane potential under palmitate exposure (Figure 4C). In addition, measuring the cellular ATP using ToxGlo™ revealed a significant reduction in the ATP level in the HuD KD cells (Figure 4D).

We next investigated whether mitochondrial damage was accompanied by mitochondrial DNA (mtDNA) mislocalization. Using the cytosolic fraction followed by a qPCR for the mtDNA loci, including Dloop, CytB, and ND4, we found increased cytosolic mtDNA in the HuD KD cells after palmitate exposure (Figure 4E). Because cytosolic mtDNA activates the cGAS-STING pathway, including TBK-dependent IRF3 phosphorylation [25], we investigated the pathway readouts and observed higher levels of phosphorylated TBK1 and IRF3 in the HuD KD cells following palmitate exposure (Figure 4F). Taken together, these results indicate that HuD KD converts palmitate stress into mitochondrial dysfunction in N2a cells, linking FAO loss to mtDNA release and cGAS-STING-TBK/IRF3 activation.

3.5. HuD Loss Amplifies Stress-Induced Premature Senescence Under Palmitate Challenge

Building on the mitochondrial defects shown in Figure 4, we asked whether HuD KD promotes a stress-induced premature senescence (SIPS) program in N2a cells under palmitate exposure. We observed an increased abundance of canonical senescence markers, including p16 and p21, by immunoblotting in the HuD KD cells (Figure 5A). In parallel, we performed SA β-gal staining and a quantitative analysis, and observed a higher proportion of SA β-gal-positive cells with increased per-cell fluorescence intensity in the HuD KD cells relative to the control (Figure 5B), indicating a robust shift toward a senescent phenotype after palmitate exposure. We also measured the levels of senescence-associated secretory phonotype (SASP) transcripts by qPCR and found an elevated expression of SASP genes in the HuD KD cells after palmitate exposure (Figure 5C). Together with Figure 4, these findings indicate that HuD depletion sensitizes N2a cells to palmitate, eliciting a SIPS.

To test the converse prediction that restoring HuD mitigates SIPS, we transfected N2a cells with pHuD and further challenged the cells with palmitate. Ectopic expression of HuD reduced the p21 and p16 protein levels (Figure 5D) and lowered the portion of SA β-gal-positive cells (Figure 5E). Taken together, the loss- and gain-of-function results are consistent with HuD counteracting palmitate-induced premature senescence-like phenotypes in N2a cells.

4. Discussion

Metabolic stress-induced senescence has emerged as a unifying mechanism that connects nutrient overload, mitochondrial dysfunction, and chronic inflammation during aging and age-related diseases [26,27]. Under persistent metabolic or oxidative stress, cells undergo profound transcriptional and bioenergetic reprogramming that culminates in a senescence-like state [28]. This process, although initially protective, can become maladaptive when sustained, as senescent cells accumulate lipids, generate excessive ROS, and secrete pro-inflammatory mediators that amplify local and systemic dysfunction [28,29]. Understanding how cells resist or succumb to this transition is therefore central to unraveling the molecular basis of age-associated decline. In this study, we identify the RNA-binding protein HuD as a molecular safeguard against lipotoxic stress-induced senescence. HuD depletion disrupts Pparα expression, leading to impaired FAO, excessive lipid accumulation, and ROS overproduction. These metabolic disturbances are not passive outcomes but active triggers that initiate a premature senescence-like program. Once established, senescence further amplifies metabolic dysfunction, forming a self-sustaining loop that reinforces oxidative and inflammatory stress. Mechanistically, HuD binds and stabilizes Pparα mRNA, thereby maintaining FAO and preventing the convergence of metabolic stress and innate immune activation. In HuD-deficient cells, FAO suppression and ROS accumulation promote mitochondrial perturbation, cytosolic mtDNA release, and TBK1/IRF3 phosphorylation, culminating in the induction of senescence markers. Together, these findings position the HuD-dependent support of PPARα-driven FAO functions as a post-transcriptional defense mechanism that restrains both the onset and amplification of senescence under metabolic stress.

Recent studies have shown that cellular senescence acts as an active amplifier of metabolic imbalance [27,30]. Senescence itself reshapes the metabolic landscape of a cell, beyond being a downstream result [29,31]. Once induced, the senescence-like state perpetuates FAO suppression, enhances mitochondrial inefficiency, and maintains elevated ROS levels. Furthermore, the SASP contributes to a pro-lipogenic and pro-inflammatory environment. In this way, senescence acts as a metabolic amplifier—a self-propagating source of oxidative stress and lipid dysregulation. Our findings thus position senescence as both a product and a driver of metabolic fragility. The loss of HuD sets this cycle in motion by weakening the PPARα–FAO axis, and the subsequent senescence intensifies the dysfunction through persistent ROS generation and inflammatory signaling. This bidirectional relationship between metabolism and senescence provides a mechanistic basis for how transient metabolic stress can evolve into chronic cellular decline.

Previous studies have linked HuD primarily to neuronal differentiation and synapse functions [17,32], but growing evidence indicates that HuD also modulates metabolic and inflammatory pathways. In our earlier works [19,33], we demonstrated that HuD deficiency leads to reduced FAO activity and upregulation of the chemokine CCL2, a pro-inflammatory mediator associated with metabolic stress and early senescence signaling. These findings suggested that loss of HuD disrupts mitochondrial lipid catabolism while simultaneously activating inflammatory transcriptional programs. Although the baseline Pparδ mRNA exceeded that of Pparα in the N2a cells [34,35,36,37], the RIP-qPCR analysis did not detect enrichment of Pparδ mRNA in the HuD immunoprecipitates relative to the IgG, whereas Pparα was consistently enriched. On this basis, we focused the mechanistic analysis on PPARα-regulated FAO in N2a. We recognize that PPARδ likely contributes to neuronal FAO per se and may be broadly expressed across neuron-derived systems; however, our data indicate that PPARδ is not a direct HuD target in this setting. Therefore, we frame our conclusions around HuD’s engagement of Pparα while acknowledging that PPARδ-driven pathways could operate in parallel via HuD-independent mechanisms or through other RNA-binding proteins. Accordingly, to investigate inducer dependence, we exposed N2a cells to IR and low-dose doxorubicin. We found that IR, but not low-dose doxorubicin, reduced HuD and PPARα expression and elicited SIPS (Supplementary Figure S3). These experiments suggest that SIPS responses are not uniform across stimuli and possibly cell type-specific programs, a notion that requires further investigation.

To probe PPARα’s contribution functionally, we performed a pharmacologic rescue in the HuD-deficient cells using fenofibrate and observed a partial reversal—increased FAO-linked ΔOCR (Seahorse) and reduced p21 and p16—consistent with PPARα supporting FAO and opposing SIPS under palmitate. By contrast, the DCFDA-measured ROS did not decline with fenofibrate. This is mechanistically plausible: augmenting FAO can increase H₂O₂ generation, and SOD-mediated dismutation converts superoxide to H₂O₂ rather than eliminating ROS, so DCFDA (H₂O₂/peroxide readout) may remain elevated despite improved FAO [38,39]. These data, together with reports that lower PPARα is associated with higher senescence markers [40,41], indicate that PPARα/FAO restoration ameliorates a subset of SIPS features, while the ROS and additional HuD-regulated nodes likely continue to drive oxidative signaling.

Consistent with this view, the HuD gain-of-function experiments complement the knockdown phenotypes. Ectopic HuD expression under palmitate condition reduced the levels of p21 and p16, and lowered the proportion and per-cell intensity of SA β-gal-positive cells (Figure 5D,E). Taken together with effect of HuD loss, these reciprocal manipulations suggest that HuD counteracts palmitate-induced senescence-like responses in N2a cells. One possible interpretation is that HuD’s post-transcriptional support of PPARα-dependent FAO helps maintain metabolic homeostasis and thereby limits senescence-linked signaling in this lipotoxic context.

The current study builds directly upon that framework by showing that such metabolic and inflammatory alterations converge mechanistically on the senescence network. CCL2 has been recognized as part of the SASP and can promote the paracrine propagation of senescence through the NF-κB and IRF3 pathways [42,43,44]. Consistent with this, HuD-deficient N2a cells displayed increased IRF3 phosphorylation and cytosolic mtDNA release—hallmarks of innate immune activation during senescence onset (Figure 4). The observation that HuD loss simultaneously reduces PPARα-driven FAO and enhances ROS production provides a coherent explanation for how HuD deficiency tips the cellular balance from an adaptive stress response toward a self-amplifying metabolic–inflammatory–senescence loop in response to PA treatment. Thus, our present and prior findings collectively position HuD as a molecular hub that coordinates metabolic restraint and inflammatory containment. Its deficiency dismantles both arms of this regulation—promoting lipid dysregulation through loss of PPARα and fostering inflammatory signaling—thereby predisposing cells to a senescence-like phenotype under metabolic stress.

This study demonstrates that HuD deficiency sensitizes cells to palmitate-induced metabolic stress, amplifying the transition toward a senescence-like state. Under lipotoxic conditions, loss of HuD disrupts Pparα expression, leading to reduced FAO, excessive lipid accumulation, and elevated ROS. These metabolic disturbances, initially triggered by palmitate exposure, become self-reinforcing in HuD-deficient cells, where ROS accumulation and FAO suppression drive stronger senescence-associated phenotypes, including mitochondrial perturbation, cytosolic mtDNA release, and SASP activation. Our findings indicate that the HuD–PPARα axis functions as a safeguard that mitigates metabolic stress and constrains the onset of senescence. Disruption of this pathway under lipotoxic challenge establishes a feed-forward metabolic–senescence loop, providing a mechanistic explanation for how transient lipid overload can evolve into persistent cellular dysfunction. By defining senescence as both a downstream response and an amplifier of HuD-dependent metabolic failure, this study refines the link between post-transcriptional regulation and stress-induced cellular senescence within a lipotoxic context.

All the experiments were performed in N2a neuroblastoma cells, a proliferative neuronal lineage model that differs from post-mitotic primary neurons in their metabolic rate, mitochondrial dynamics, and stress responses. These differences may alter FAO flux, ROS handling, and susceptibility to SIPS, so our findings should not be over-generalized to physiological neuronal aging or neurodegenerative settings. We chose N2a for its tractability—allowing for robust HuD perturbation and quantitative assays of lipid metabolism and mitochondrial function under palmitate—but the model selection may have influenced the effect sizes and the balance between storage, oxidation, and senescence outputs. To strengthen the physiological relevance, future work will test the HuD→PPARα/FAO axis in primary cortical/hippocampal neurons and human iPSC-derived neurons using the same readouts, thereby defining the extent to which the mechanisms mapped in N2a generalize to post-mitotic neuronal contexts.

Finally, while our data establish PPARα as a HuD target in N2a cells—supporting a model in which HuD sustains FAO and mitigates lipotoxic stress—we do not infer that PPARα is the sole mediator of the observed phenotypes. HuD/nELAVL proteins bind and regulate numerous neuronal transcripts implicated in metabolism and stress adaptation, and it is therefore plausible that additional HuD targets act in parallel with PPARα to drive the breadth of effects we observed (mROS elevation, ΔΨm depolarization, ATP decline, cytosolic mtDNA release, TBK1/IRF3 activation, and SIPS). To delineate this broader regulatory network, we plan unbiased RIP-seq/CLIP-seq of HuD-bound RNAs in neuronal cells exposed to palmitate, followed by integrative transcriptomic analyses and functional validation, to determine which targets are necessary and/or sufficient for specific branches of the mitochondrial and senescence responses. In this framework, PPARα/FAO represents a validated axis within a larger HuD-dependent program that links lipid–metabolic stress to mitochondrial dysfunction and neuronal senescence. In human disease, HuD has been linked to neurodegenerative disorders (Alzheimer’s disease, Parkinson’s disease, and ALS); to neurologic/psychiatric conditions (epilepsy and schizophrenia); and to cancer, including neuroblastoma and paraneoplastic anti-Hu (HuD) syndromes associated with small-cell lung cancer [16,17]. Although our experiments were conducted in Neuro2a cells, the mechanisms defined here provide a basis for evaluations primary and human iPSC-derived neurons and their interpretation in disorders marked by HuD dysregulation.

5. Conclusions

Taken together, our results reveal the HuD–PPARα–FAO circuit that protects N2a cells against lipotoxic stress. By binding to the Pparα 3′UTR and counteracting the repression driven by miR-9-5p/miR-22-3p, HuD promotes PPARα-dependent FAO. HuD downregulation decreases PPARα/FAO, increases ROS, exacerbates mitochondrial damage, and promotes senescence-like phenotypes in response to palmitate. These findings underscore the importance of post-transcriptional regulation of oxidative lipid metabolism as a potential target for treating disorders associated with metabolic and inflammatory vulnerability.