Aging Rewires Neuronal Metabolism, Exacerbating Cell Death After Ischemic Stroke: A Hidden Reason for the Failure of Neuroprotection

Multiplex immunohistochemical analysis revealed that as early as one day after stroke onset, a substantial proportion of neurons within the penumbra contained activated Caspase 3, the terminal effector of apoptosis. In the young and middle groups, the absolute number of Caspase 3⁺ neurons exceeded that observed in elderly patients. However, when the percentage of dying (caspase-positive) cells among the total number of neurons (NeuN⁺) in the penumbra was calculated, this value increased with age—42% in young patients versus 89% in elderly individuals. Thus, the apparent decrease in the absolute number of Caspase 3⁺ neurons in the elderly likely reflects reduced neuronal density rather than attenuation of apoptosis, which is consistent with the findings of other researchers [25]. Based on these observations, we developed a method for calculating the "Caspase⁺/NeuN⁺ neuron ratio," which can serve as an applied diagnostic criterion for assessing apoptotic neuronal death in brain tissue samples following ischemic stroke. Although Caspase-3 is commonly used as an indicator of apoptotic signaling, its specificity in ischemic tissue is not absolute, and the absence of confocal imaging may theoretically allow minimal signal overlap. These factors should be considered as potential, though not definitive, limitations of our methodology.

Moreover, the high level of neuronal staining with anti-caspase 3 antibodies in the infarct border zone of elderly patients indicates that by the first day after stroke onset, the apoptotic cascade had already been activated in nearly all surviving cells. This represents a fundamental difference from younger patients, in whom a considerable proportion of penumbral neurons remain viable and can potentially be preserved with timely therapeutic intervention [26]. The obtained results partly explain the more severe clinical course of stroke in elderly patients [27]. In other words, their adaptive and compensatory potential is critically low, and the therapeutic window for effective intervention is extremely limited. In addition, in some cases we observed that in young patients, the infarct zone became demarcated early by glial scar formation, whereas in most elderly individuals, the infarct volume continued to expand progressively. This may account for the higher rate of late mortality observed in the elderly cohort [28]. Thus, the near-total activation of Caspase 3 in penumbral neurons of older patients underlies the rapid expansion of the infarct zone due to apoptotic death of neurons that were previously considered conditionally preserved.

Co-localization of the markers NeuN and NSE selected in the present study provides insight not only into the number of neurons but also into their functional and metabolic activity [5,29]. The Neuronal nuclear antigen (NeuN, RBFOX3) is involved in the regulation of mRNA splicing and is normally present in the nuclei and perikaryon of nearly all differentiated neurons [30]. The intensity of NeuN immunolabeling correlates with neuronal integrity, whereas suppression of its expression indicates cellular distress [31]. Neuron-specific enolase (NSE), a key cytosolic enzyme of glycolysis, serves as an indicator of the neuronal energy metabolism level [32]. Multiplex analysis revealed that in young patients, large pyramidal neurons with high NeuN and NSE expression were still present in the penumbra during the first day after stroke onset. This finding indicates enhanced metabolic activity, reflecting compensatory hyperenergization in response to ischemia [33]. Indeed, cytologically preserved neurons within the infarct border zone must adapt to oxygen and glucose deficiency by shifting toward anaerobic glycolysis, which is confirmed by elevated NSE expression. However, paradoxically, some of these neurons were also positively stained for Caspase 3. It is likely that the transient increase in glycolytic activity does not prevent subsequent cell death but rather delays it, providing short-term metabolic support until the genetic program of apoptosis is executed.

In contrast, in the elderly group, most neurons exhibited low metabolic activity, which may be explained by the reduced compensatory and adaptive potential of brain structures [34]. Immunolabeling for NeuN and NSE in this group was markedly decreased, despite the apparent preservation of some cells under routine histological staining. NeuN deficiency may also indicate impaired protein-synthetic function of neurons, particularly in elderly patients, supporting the role of ontogenetic changes [35]. The decline in protein-synthetic capacity may be indirectly related to the fact that NeuN is associated with nuclear chromatin and Nissl substance, that is, with granular endoplasmic reticulum and ribosomes. Its loss likely reflects ribosomal destabilization and cessation of protein synthesis, further contributing to the energetic and trophic deficiency of the cells [36].

In addition, the present study revealed an interesting phenomenon of NeuN redistribution in neurons of the infarct border zone. We observed diffuse accumulation of this protein within the cytoplasm and even dendritic processes instead of the predominantly nuclear localization typical under normal conditions. Such abnormal NeuN distribution corresponds to isolated reports in the literature, where this pattern is associated with calpain-dependent proteolytic cleavage of RBFOX3 combined with disruption of nuclear envelope transport under ischemic conditions [37,38]. Our results suggest that the efflux of NeuN from the nucleus represents an early marker of neuronal functional failure preceding irreversible morphological alterations.

It is noteworthy that in the cerebral cortex of elderly patients, not only the expression of neuronal markers but also the nature of cellular responses within the infarct border zone was altered. In some fragments from this group, pronounced glial hyperplasia (gliosis) was observed along the periphery of the infarct, and several astrocytes and oligodendrocytes were positively stained with antibodies against NSE. Although NSE is traditionally regarded as a neuron-specific protein, it is known that hybrid isoforms of this enzyme (αγ-dimers) can be expressed by glial cells in the central nervous system [5]. The presence of NSE⁺ small cells in the infarct border zone of elderly patients is likely associated with translocation of this enzyme from damaged neuronal processes to macro- or microglia. Another possible explanation for this phenomenon is a metabolic shift in glial cells toward enhanced glycolysis under ischemic conditions [39]. However, this hypothesis requires further clarification in subsequent studies. Nevertheless, the analysis of NeuN and NSE co-localization clearly demonstrates the inability of neurons in older individuals to maintain adequate energy metabolism during stroke compared with younger patients. This finding supports the concept of age-associated decline in neuronal plasticity and the adaptive-compensatory capacity of the cerebral cortex.

The present study revealed significant age-related differences in the activity of the PI3K/Akt/mTOR and PI3K/Akt/FOXO3a signaling pathways, which are responsible for regulating cellular survival and death [40,41]. Quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR) analysis showed that in young patients, ischemic stroke was accompanied by coordinated activation of both the pro-survival PI3K/Akt/mTOR cascade and the stress-induced PI3K/Akt/FOXO3a axis [42]. Previous studies have reported both beneficial and detrimental effects of PI3K/Akt/FOXO3a activation, depending on the affected organ and the nature of the insult [20]. However, our findings suggest that this signaling pathway directly contributes to neuronal death within the infarct border zone.

In cortical homogenates from young patients, expression levels of Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), AKT serine/threonine kinase 2 (AKT2), Mammalian Target of Rapamycin (MTOR), and Forkhead box O3a (FOXO3A) genes were found to increase almost equally during the first 24 h after stroke, followed by a trend toward normalization at later stages. This "balanced" response can be interpreted as the concurrent activation of mechanisms promoting cell survival (primarily through the Akt–mTOR pathway) and mechanisms aimed at damage resolution (via moderate activation of FOXO3a-dependent genes involved in autophagy and apoptosis). Notably, in the middle group, signs of dysregulation were observed: a relative increase in FOXO3A expression compared with MTOR indicated a shift in signaling balance toward a pro-apoptotic response.

Finally, in elderly patients, the overall transcriptional activity of the PI3K/Akt pathway components was markedly reduced, consistent with previous reports describing age-related suppression of the insulin-like growth factor 1 (IGF-1) signaling cascade and associated kinase pathways as potential underlying mechanisms [19,43]. Against this background, FOXO3A expression levels in our study were significantly higher than in young patients, whereas mTOR expression remained decreased. In other words, in the cerebral cortex of elderly individuals, ischemia induces a pathological predominance of the Akt/FOXO3a axis with relative insufficiency of Akt/mTOR signaling [44]. Considering that under normal physiological conditions Akt suppresses FOXO proteins, ischemic stroke may result in insufficient Akt (and consequently mTOR) activity, leading to uncontrolled FOXO3a activation [42]. This, in turn, triggers cell death, dysregulated autophagy, and the expression of proinflammatory interleukins. As a result, immune cell migration into the infarct border zone increases, promoting autoimmune neuronal injury and expansion of the infarct core. These findings are crucial for understanding the molecular mechanisms underlying neuronal death.

Under normal conditions, the presence of trophic factors such as IGF-1 activates the PI3K/Akt pathway, leading to suppression of FOXO3a expression: activated Akt phosphorylates FOXO3a, retaining it in the cytoplasm, while simultaneously initiating mTOR-mediated programs of protein synthesis, cellular growth, and survival [42]. It is likely that in older patients, insufficient Akt activity results in the accumulation of unphosphorylated FOXO3a, which translocates to the nucleus, where it functions as a transcription factor in the stress response. FOXO3a is known to regulate numerous genes associated with apoptosis and inflammation [45]. In particular, it induces the expression of pro-apoptotic proteins (Bim, FasL, Puma, and others), promotes the synthesis of proinflammatory cytokines, and, in some cases, enhances autophagy and ubiquitin–proteasome-mediated proteolysis [46,47].

In young patients, the effects of FOXO3a are partially counterbalanced by mTOR activity and other pro-survival proteins, such as the anti-apoptotic Akt/Bcl-2 axis. In contrast, the data obtained in the present study indicate that in older individuals, ischemia induces uncontrolled nuclear activation of FOXO3a accompanied by insufficient activity of compensatory survival pathways. The consequences of this include the following: enhanced apoptosis through activation of the caspase cascade; chronic inflammation due to hyperactive FOXO3a–mediated induction of proinflammatory cytokine production; and dysregulated autophagy, characterized by autophagic cell death against a background of suppressed mTOR signaling. Collectively, these processes lead to increased neuronal loss in the infarct border zone and progression of necrosis in the cerebral cortex.

Our findings are consistent with the results of previous studies. For instance, pharmacological activation of PI3K/Akt has been shown to reduce infarct volume in experimental stroke models, whereas inhibition of Akt exacerbates neuronal death [15,16]. There is also evidence that FOXO3a hyperactivation is associated with greater lesion volume due to the induction of uncontrolled autophagy, apoptosis, and proinflammatory cytokine synthesis during ischemia [48]. In the present study, this phenomenon was demonstrated for the first time in human brain tissue. The predominance of FOXO3A over MTOR expression in the infarct border zone may thus serve as a molecular marker of insufficient neuroprotective mechanisms in elderly patients. This observation is particularly important for the development of new potential targets for neuroprotective therapy. Modulation of the Akt/FOXO3a axis—such as targeted inhibition of FOXO3a or activation of Akt—may prove effective in older individuals, whereas such an approach may be less beneficial in younger patients. This finding underscores the importance of considering patient age when selecting neurotropic treatment strategies and may partially explain the limited efficacy of standard neuroprotective therapies in age-heterogeneous cohorts of patients with ischemic stroke [49,50].

The results of the present study demonstrate significant age-associated alterations in the mechanisms of neuronal death following ischemic stroke. In young patients, neuronal injury in the cerebral cortex can be considered partially compensated: during the first 24 h, a pannecrotic zone forms, while a portion of penumbral neurons retains normal cytology and partial metabolic activity—a state often referred to as "neuronal lethargy" [21]. These neurons exhibit compensatory enhancement of metabolic activity, as evidenced by increased NeuN and NSE expression, together with moderate activation of the caspase cascade, suggesting a potential for reversibility. At the molecular level, penumbral neurons appear to maintain a partial balance between the PI3K/Akt/mTOR and PI3K/Akt/FOXO3a signaling pathways. Collectively, this enables temporary maintenance of protein synthesis, energy metabolism, and partial restriction of apoptosis in the infarct border zone (Figure 5), which may contribute to the preservation of cells in this region, partial prevention of infarct core expansion, and improved clinical outcomes.

In contrast, in elderly patients, by the first day after stroke onset, nearly complete loss of NeuN- and NSE-positive neurons was observed in the penumbra, accompanied by a sharp increase in the proportion of Caspase 3 expressing cells. This likely reflects reduced adaptive and metabolic capacity of neurons. The loss of NeuN indicates suppression of protein-synthetic activity, while decreased NSE reflects impairment of glycolytic metabolism. Molecular analysis revealed predominance of FOXO3A transcriptional activity combined with reduced expression of its suppressors, Akt and mTOR, suggesting dominance of pro-apoptotic signaling. Moreover, in this patient group, the mechanisms of autophagy, inflammation, and apoptosis lose regulation and become integral components of ischemic stroke pathogenesis, accelerating neuronal death in the infarct border zone and expansion of the infarct zone (Figure 5). This contributes to a more severe clinical course and poorer prognosis.

This set of findings is highly relevant to current efforts aimed at improving functional outcomes after ischemic stroke, particularly in the era of widespread recanalization therapies [51]. Recent Stroke Treatment Academic Industry Roundtable (STAIR) recommendations emphasize that successful reperfusion must be accompanied by effective cerebroprotective strategies capable of preserving the metabolic viability of penumbral neurons [52,53]. Our data suggest that age-dependent dysregulation of neuronal survival pathways (an imbalance between Akt/mTOR and Akt/FOXO3a) may largely explain why many neuroprotective agents fail to translate into clinical benefit in heterogeneous patient populations. These results highlight the importance of developing age-tailored cerebroprotective approaches that enhance pro-survival signaling and stabilize neuronal metabolism following recanalization, especially in older groups.

In summary, with advancing age, the cerebral cortex progressively loses its capacity for metabolic adaptation and cellular survival in response to ischemia. In the cortical tissue of elderly patients, nearly all neurons in the infarct border zone exhibited signs of apoptotic death at early stages, whereas in younger individuals, a population of cells with recovery potential was preserved. This difference reflects an age-associated decline in neuronal plasticity and energetic resilience, leading to rapid expansion of the necrotic focus and worse stroke outcomes. The findings suggest that age-dependent dysregulation of the Akt/mTOR and Akt/FOXO3a signaling pathways, in combination with impaired energy metabolism, underlies the observed differences in the severity and outcomes of ischemic stroke. These results further emphasize the need to develop neuroprotective approaches aimed at restoring the regulatory function of these signaling pathways and enhancing neuronal metabolic stability in elderly patients. The data also indicate that such neuroprotective strategies may be less effective in younger cohorts, highlighting the importance of personalized, age-specific therapeutic interventions targeting metabolic activation, suppression of apoptotic cell death, and/or modulation of the inflammatory response in patients with ischemic stroke.

Limitations. This study used archival autopsy samples of the cerebral cortex, which precluded assessment of infarct volume and location: only cortical tissue was identified based on pathological examination protocols and characteristic structural organization. Also, in the present study, we did not conduct a correlation analysis between the clinical data of patients and the obtained results of morphological analysis, which may be interesting and useful for future research.

Serial sections (3 μm thick) were prepared from paraffin-embedded brain tissue blocks, deparaffinized, dehydrated, and stained with hematoxylin and eosin according to standard protocols. To identify neurons, one section from each case was stained with cresyl violet using the Nissl method following conventional procedures. The resulting histological slides were examined under a Leica DM2000 light microscope (Leica Microsystems, Wetzlar, Germany) equipped with a digital microphotography system. Neuronal counts were performed in 10 randomly selected microscopic fields at ×400 magnification, separately for the infarct core and the penumbra/infarct border zone. All tissue samples included in the study were confirmed to originate from the ischemic stroke region based on macroscopic and microscopic criteria. All morphological quantifications, including neuronal counts in the intact cortex, the infarct border zone, and the infarct core, were conducted by an investigator who was fully blinded to the patients' age groups and clinical characteristics to prevent observer bias.

Multiplex immunofluorescence staining was performed using the Opal™ 7-Color Fluorescence Immunohistochemistry Kit (Akoya Biosciences, Marlborough, MA, USA) according to the manufacturer's protocol. After deparaffinization and rehydration, antigen retrieval was carried out in Tris-EDTA buffer (pH 9.0) using a microwave oven for 15 min, followed by blocking of endogenous peroxidase activity. First staining round. Sections were incubated for 60 min with the primary antibody against NeuN (Clone EPR12763, Abcam, Cambridge, UK; 1:500). After washing, slides were treated with the Horseradish Peroxidase (HRP) Ms+Rb polymer for 10 min, followed by incubation with the Opal 647 fluorophore for 10 min. Bound primary and secondary antibodies were then removed by repeated microwave treatment in Tris-EDTA buffer (pH 9.0) as described above. Subsequent staining rounds. The procedure was repeated with antibodies against NSE (Clone EPR12483, Abcam, Cambridge, UK; 1:500) and Caspase 3 (Clone 9H19L2, Thermo Fisher Scientific, Waltham, MA, USA; 1:500). After incubation with the HRP Ms+Rb polymer for 10 min, the slides were treated with Opal 488 and Opal 540 fluorophores, each for 10 min. Nuclei were counterstained with DAPI for 5 min, and the sections were mounted using ProLong™ Diamond Antifade Mountant (Thermo Fisher Scientific, Waltham, MA, USA). Visualization and quantitative analysis. Multispectral images were acquired using the Vectra^® Polaris Automated Quantitative Pathology Imaging System (Akoya Biosciences, Marlborough, MA, USA) at 20× magnification. Whole-slide images were processed and analyzed using inForm^® Image Analysis Software v2.6 (Akoya Biosciences, Marlborough, MA, USA).

Then, the number of stained neurons was counted separately for the markers NeuN, NSE and caspase 3 in the field of view of a confocal microscope at a magnification of ×400. Based on the obtained average values, the percentage of Caspase 3⁺ cells among NeuN⁺ was calculated using the formula Caspase 3/NeuN ×100 (%), reflecting the proportion of dying neurons.

The expression levels of mRNA for proteins of PI3K/Akt/mTOR and PI3K/Akt/FOXO3a were evaluated using quantitative real-time PCR (qRT-PCR). The material for RT-PCR was collected at a distance of 1 mm from the border of the infarct core in the infarct border zone. For control, tissue was taken from a distant area of the cerebral cortex, which can be considered conditionally intact. Tissue samples were homogenized according to standard protocols. Total RNA was extracted using the RNeasy Plus Mini Kit (QIAGEN, Venlo, The Netherlands). Complementary DNA (cDNA) synthesis was performed using the SuperScript™ VILO™ Master Mix (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA).

The obtained cDNA samples were subjected to qRT-PCR using the ABsolute™ Blue QPCR Mix (Thermo Fisher Scientific, Waltham, MA, USA) with SYBR Green I detection dye. Amplification was carried out on the StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) following the manufacturer's instructions. Gene expression analysis was performed using the comparative threshold cycle (ΔCt) method, and relative expression levels were calculated according to the established protocol.

β-actin was used as a housekeeping reference gene for normalization which was selected based on its well-established stability in human postmortem cortical tissue. ACTB expression demonstrates minimal variability across ischemic and non-ischemic cortical regions and exhibits no significant age-dependent fluctuations, which supports its suitability for use as an internal control in qRT-PCR analysis. Furthermore, preliminary assessment of our dataset confirmed low dispersion of Ct values for β-actin across all examined samples.

Primer sequences were designed based on publicly available DNA and mRNA data from the NCBI database using the Primer-BLAST tool (Table 3).

Quantitative data were processed using SPSS Statistics 12 for Windows (IBM Analytics, Armonk, NY, USA). The normality of data distribution was assessed using the Shapiro–Wilk test. For normally distributed variables, comparisons were performed using Student's t-test or one-way ANOVA with appropriate post hoc tests. In cases of non-normal distribution, the Kruskal–Wallis test with Dunn's post hoc correction was applied. Multiple comparisons in the clinical dataset were carried out using one-way ANOVA with post hoc analysis for normally distributed data and Pearson's χ² test for categorical variables. Pairwise comparisons in the morphological and immunohistochemical analyses were performed using the Mann–Whitney U test with Bonferroni correction to account for multiple testing. For qRT-PCR data, relative gene expression was evaluated using the ΔCt method, and intergroup differences were analyzed using Student's t-test or one-way ANOVA with Bonferroni adjustment when applicable. The results are presented as mean ± standard deviation (SD) and, where appropriate, with a 95% confidence interval (CI). A p-value ≤ 0.05 was considered statistically significant.