Open AccessArticle Comparative Effects of Donepezil and Tacrine on Recall-Related Exploratory Behavior in a Subacute Lipopolysaccharide-Induced Neuroinflammatory Model of Cognitive Impairment 1 Faculty of Medicine, “Carol Davila” University of Medicine and Pharmacy, 050474 Bucharest, Romania 2 Service Anesthesie-Reanimation, Groupe Hospitalier de la Pitié-Salpêtrière, CEDEX 13, 75651 Paris, France 3 Faculty of Medical Engineering, National University of Science and Technology Politehnica Bucharest, 060042 Bucharest, Romania * Authors to whom correspondence should be addressed. † These authors contributed equally to this work. Biomedicines 2026, 14(6), 1306; https://doi.org/10.3390/biomedicines14061306 (registering DOI) Submission received: 26 April 2026 / Revised: 5 June 2026 / Accepted: 8 June 2026 / Published: 9 June 2026 Background/Objectives: Neuroinflammation is increasingly recognized as an important contributor to Alzheimer-like cognitive impairment. Lipopolysaccharide (LPS) is commonly used in experimental models to trigger systemic immune activation and behavioral alterations associated with neuroinflammation. This study aimed to validate a subacute LPS-induced model of recall-phase impairment and to compare the effects of donepezil and tacrine on recall-related exploratory behavior in rats. Methods: Male Wistar rats were tested in a two-trial Y-maze paradigm consisting of an acquisition trial followed by a recall trial 24 h later. In the validation experiment, rats received saline or LPS 1 mg/kg intraperitoneally for four consecutive days. In the intervention experiment, rats received saline, LPS, or LPS combined with donepezil 1 or 3 mg/kg or tacrine 3 or 5 mg/kg. The primary recall-phase outcome was the unknown/known arm time ratio (U/K time ratio). Additional outcomes included arm times, arm entries, U/K entry ratios, discrimination indices, and mean time per entry. Results: Repeated LPS administration significantly reduced the U/K time ratio, decreased time- and entry-based discrimination indices, reduced time spent in the unknown arm, and decreased unknown-arm entries, without significantly altering acquisition-phase behavior, total entries, or mean time per entry. In the intervention experiment, donepezil 1 mg/kg and tacrine 5 mg/kg significantly increased the U/K time ratio compared with LPS. Discrimination indices and entry-based measures further supported a treatment-related shift toward novelty-directed exploration, while total arm entries and mean time per entry were not significantly changed. Conclusions: Subacute LPS administration produced a measurable recall-phase exploratory impairment in the Y-maze. Donepezil and tacrine attenuated several components of this impairment, with partially distinct dose-related behavioral profiles. 1. Introduction Among the available LPS preparations, LPS derived from Escherichia coli O55:B5 represents a well-established experimental source of LPS with strong immunostimulatory activity [ 5]. Structurally, LPS is a glycolipid composed of a hydrophobic lipid A region and a carbohydrate chain, which includes the core oligosaccharide and the O-specific polysaccharide region [ 5, 6]. Other E. coli LPS serotypes, such as O111:B4, O26:B6, O127:B8, and O128:B12, are also used experimentally; although they activate broadly similar TLR4-dependent pathways, structural differences may influence inflammatory potency and response kinetics [ 5, 6, 7]. The present study therefore used repeated systemic LPS administration as an impairment-inducing model relevant to the neuroinflammatory component of Alzheimer-like cognitive dysfunction. We aimed to determine whether this subacute inflammatory challenge produces a recall-phase alteration detectable in a two-trial Y-maze paradigm with a 24 h interval and whether this behavioral pattern is modulated by cholinesterase inhibition. After behaviorally characterizing the LPS-induced phenotype, we compared donepezil and tacrine as pharmacologically distinct cholinesterase inhibitors. The analysis focused on recall-related exploratory organization, including time allocation, arm-entry patterns, U/K ratios, discrimination indices, and time-per-entry measures, in order to distinguish changes in arm-selection dynamics from changes in average visit duration. 2. Materials and Methods Ethical approval for the study was obtained from the Ethics Committee of the “Carol Davila” University of Medicine and Pharmacy, Bucharest, Romania (authorization no. 13619 from 24 May 2024). Animal procedures complied with the applicable national legal requirements for scientific animal research and were authorized and supervised by the National Sanitary Veterinary and Food Safety Authority (authorization no. 9 from 5 July 2024). 2.1. Animals Male Wistar rats, aged 6–8 weeks and weighing 200–300 g at the time of behavioral testing, were obtained from the animal breeding facility of the “Carol Davila” University of Medicine and Pharmacy, Bucharest, Romania. Before inclusion in the experiments, the animals underwent a minimum 7-day acclimatization period under local laboratory conditions. During the study, rats were kept in individual cages positioned next to one another in the same animal room. Food and water were provided ad libitum, and the housing environment included natural ventilation, 40–60% relative humidity, and the standard institutional light cycle. This arrangement avoided direct physical interaction between animals, while still allowing indirect sensory contact with conspecifics. The use of individual cages was intended to standardize pre-test housing conditions and reduce potential cage-mate-related effects on exploratory behavior during the experimental period. The animal experiments were conducted and reported in line with the ARRIVE recommendations. All procedures were carried out in the Department of Pharmacology during the 10:00–17:00 interval. Each rat was assigned to a single experiment. Euthanasia was performed under general anesthesia 24 h after the completion of behavioral testing. 2.2. Experimental Design The study was organized into two separate protocols. (A) LPS validation protocol. Two experimental groups were used, with 9 rats assigned to each group: C = control group: saline; LPS 1 = lipopolysaccharide 1 mg/kg. In the validation experiment, animals received one intraperitoneal injection per day for 4 consecutive days. Rats in the control group were administered saline, whereas rats in the LPS 1 group received lipopolysaccharide at a dose of 1 mg/kg. On days 1–3, treatment was administered according to the same daily schedule without behavioral testing. On day 4, the injection was given 2 h before the acquisition trial. The recall trial was carried out 24 h later, without additional drug administration. (B) Pharmacological intervention protocol under LPS challenge. Six experimental groups were used, with 9 rats assigned to each group: C = control group: saline + saline; LPS 1 = lipopolysaccharide 1 mg/kg + saline; D1 = lipopolysaccharide 1 mg/kg + donepezil 1 mg/kg; D3 = lipopolysaccharide 1 mg/kg + donepezil 3 mg/kg; T3 = lipopolysaccharide 1 mg/kg + tacrine 3 mg/kg; T5 = lipopolysaccharide 1 mg/kg + tacrine 5 mg/kg. In the intervention experiment, all animals received two intraperitoneal injections per day for 4 consecutive days. The first daily injection consisted of lipopolysaccharide or saline, according to group allocation. Seventy minutes later, a second injection consisting of donepezil, tacrine, or saline was administered. On days 1–3, treatment was given according to the same schedule without behavioral testing. On day 4, the first injection was administered 120 min before the acquisition trial, whereas the second injection was administered 50 min before the acquisition trial. The recall trial was performed 24 h later, without additional treatment. The group allocation, treatment schedule, behavioral testing sequence, and euthanasia time point for both experimental protocols are summarized in Figure 1. To minimize possible bias related to testing order, treatment administration and behavioral testing were scheduled according to a rotating group order. In the validation protocol, rats were alternated between the C and LPS 1 groups. In the intervention protocol, one rat from each group was treated and tested in the order C, LPS 1, D1, D3, T3, and T5 before the sequence was repeated for the next set of animals. 2.3. Drugs and Reagents The compounds used in the present study were the following: Lipopolysaccharide (LPS) from Escherichia coli O55:B5 (25 mg vials; Sigma-Aldrich/Merck, MO, USA; catalog no. L2880) Donepezil hydrochloride (1 g vials; Sigma-Aldrich/Merck, MO, USA; catalog no. D6821; purity ≥ 98% by HPLC) Tacrine (1 g vials; Sigma-Aldrich/Merck, MO, USA; catalog no. A79922; purity ≥ 99%) 0.9% sodium chloride solution All treatments were delivered by intraperitoneal injection, with the administered volume adjusted to 0.1 mL per 100 g of body weight. 2.4. Y-Maze Apparatus A Y-maze task was used to evaluate spatial recognition memory. The apparatus consisted of three identical wooden arms arranged in a Y configuration; each arm had a length of 47 cm, a width of 15 cm, and a height of 38 cm. Adhesive foil was applied to the inner surfaces to provide a uniform texture throughout the maze. The two-trial Y-maze paradigm was selected because it provides a rapid and relatively low-stress assessment of spatial recognition memory and novelty-directed exploratory behavior. The Y-maze is a simple and widely used behavioral paradigm in learning and memory studies, and two-trial versions allow assessment of exploration of a previously inaccessible novel arm [ 40]. Compared with more demanding behavioral paradigms such as the Morris water maze, the Y-maze requires minimal training and avoids exposure to swimming-related stress, which could potentially influence inflammatory and behavioral responses under repeated LPS administration [ 41]. Compared with novel object recognition, the present Y-maze protocol focuses on spatial-context novelty and exploratory allocation between familiar and unfamiliar arms, rather than object-directed exploration [ 42]. In addition, the two-trial Y-maze protocol allows evaluation of recall-phase exploratory allocation and discrimination-related behavioral indices within a relatively simple experimental design. Lighting conditions were kept uniform throughout the maze, and environmental noise was reduced as much as possible during testing. The maze was placed in a constant position relative to external visual cues, which were kept unchanged across sessions in order to avoid spatial bias. Its orientation was checked regularly throughout the study. A Logitech video camera (Lausanne, Switzerland) was placed beside the start arm to monitor Y-maze activity. Recordings were transmitted to a Lenovo computer (Morrisville, NC, USA) located outside the testing room, allowing the experimenter to observe the trial remotely. This setup provided continuous visualization of the maze and enabled reliable identification of arm entries. Video files were saved and subsequently evaluated offline to support blinded behavioral scoring and reduce observer-related bias. 2.5. Experimental Procedure To minimize procedural variability, animal handling and placement in the maze were performed by the same experimenter throughout the study. After each rat was introduced into the maze, the experimenter exited the testing room. Rats were randomly distributed across groups, and video-based behavioral scoring was performed under blinded conditions. The Y-maze procedure consisted of two 5 min trials: an acquisition trial and a recall trial performed 24 h later. During acquisition, only the start arm (S) and the familiar arm (K) were accessible, while the novel arm (U) remained closed with a wooden panel matching the maze walls and floor in texture and appearance. The familiar/novel arm assignment was alternated between left and right positions across rats. This distribution was kept balanced within each experimental group. During the recall trial, rats had access to all arms, with the first-session familiar/novel configuration kept unchanged. Between trials, the maze was cleaned and allowed to dry to reduce the influence of residual odor cues. 2.6. Behavioral Analysis Behavioral variables were analyzed from the recorded Y-maze sessions. For scoring, the maze was divided into the previously defined S, K, and U arms, together with the central intersection zone (I). For the acquisition trial, the evaluated measures were S-arm time, K-arm time, S-arm entries, K-arm entries, and total entries. For the 24 h recall trial, the analysis focused on K- and U-arm exploration time and entry counts. Derived recall-phase parameters were calculated separately for each animal. The U/K time ratio was defined as U-arm time divided by K-arm time, and the U/K entry ratio was computed as U-arm entries divided by K-arm entries. Mean time per entry (MTPE) was calculated for each arm, and time- and entry-based discrimination indices were calculated as (U − K)/(U + K). Arm transitions were scored using the hind paws as reference points: entry required both hind paws to be inside an arm, whereas exit was recorded after both hind paws had left the arm. Higher U/K ratios were interpreted as indicating a stronger recall-phase preference for the U arm. MTPE was included as an additional descriptor of visit duration. 2.7. Statistical Analysis Statistical analysis was carried out using IBM SPSS Statistics for Windows, version 29.0 (IBM Corp., Armonk, NY, USA). Data are reported as mean ± standard deviation (SD). Variables from the acquisition and recall sessions were evaluated independently. The primary endpoint for the recall phase was the U/K time ratio. The other recall-related measures, including the U/K entry ratio, time- and entry-based discrimination indices, absolute arm times, arm-entry counts, and MTPE, were used as secondary supportive outcomes. Group comparisons were made using one-way ANOVA in both experimental protocols because the study design involved comparisons among independent experimental groups. In the validation experiment, one-way ANOVA was used to compare the C and LPS 1 groups. In the intervention experiment, whenever ANOVA indicated a significant group effect, Dunnett’s post hoc test was used for comparisons against the LPS 1 group, because the main objective was to determine whether donepezil or tacrine modified the behavioral effects induced by LPS. Normality was checked within each group using the Shapiro–Wilk test, and equality of variances was examined with Levene’s test in order to verify the assumptions required for parametric group comparisons. Kruskal–Wallis testing was additionally used as a non-parametric sensitivity analysis and confirmed the same general pattern of results. Potential outliers were assessed for all behavioral variables included in the statistical analysis using boxplot inspection, the 1.5 × interquartile range (IQR) rule, and standardized values calculated within each experimental group. Values flagged by IQR-based inspection were further evaluated using standardized values and biological plausibility. No observation exceeded the conventional standardized-value threshold for outliers (|z| > 3), and no data points were excluded from any analysis. A power analysis was performed for the primary outcome measure (U/K time ratio). Based on the observed ANOVA effect size in the intervention experiment (η 2 = 0.334; Cohen’s f = 0.708), with α = 0.05, statistical power set at 0.80, and six experimental groups, the estimated minimum sample size was approximately six animals per group. Therefore, the sample size used in the present study (n = 9 per group) was considered sufficient to detect the observed primary effect. A p value < 0.05 was considered statistically significant. 3. Results 3.2. Donepezil and Tacrine Effects Under LPS Challenge 3.2.1. Acquisition-Phase Behavior (Session 1) Acquisition-phase behavioral parameters are presented in Table 3. Acquisition-phase variables were comparable across groups, including S-arm time (one-way ANOVA: F(5,48) = 0.16, p = 0.9752), K-arm time (F(5,48) = 0.34, p = 0.8880), S-arm entries (F(5,48) = 0.51, p = 0.7706), K-arm entries (F(5,48) = 0.53, p = 0.7556), total exploration time (F(5,48) = 0.36, p = 0.8718), and total arm entries (F(5,48) = 0.32, p = 0.8964). These findings indicate comparable acquisition-phase exploratory behavior across all experimental groups before recall testing. 3.2.2. U/K Time Ratio In the intervention experiment, the main recall-related measure was the U/K time ratio, used to compare exploration of the U arm relative to the K arm. Absolute arm times, arm-entry counts, U/K entry ratio, discrimination indices, and mean time per entry were evaluated as complementary outcomes to describe the treatment-related behavioral pattern. Figure 7 shows the U/K time ratio values recorded during recall. The U/K time ratio showed a significant overall difference among groups (one-way ANOVA: F(5,48) = 3.75, p = 0.0061). Dunnett’s post hoc analysis, performed against the LPS 1 group, showed higher ratio values in the D1 group ( p = 0.0178) and T5 group ( p = 0.0070). The comparisons between LPS 1 and C ( p = 0.5716), D3 ( p = 0.0780), or T3 ( p = 0.5618) did not reach statistical significance. These results indicate that, under LPS challenge, donepezil 1 mg/kg and tacrine 5 mg/kg increased novelty-directed exploration during recall, as reflected by higher U/K time ratio values. 3.2.3. Discrimination Indices Analysis of the time discrimination index showed a significant group effect (one-way ANOVA: F(5,48) = 6.52, p = 0.000106). The corresponding mean values were 0.076 ± 0.137 for C, −0.222 ± 0.202 for LPS 1, 0.169 ± 0.287 for D1, 0.199 ± 0.140 for D3, 0.078 ± 0.132 for T3, and 0.239 ± 0.224 for T5. In Dunnett’s comparisons against LPS 1, significantly higher values were found for C, D1, D3, T3, and T5. The corresponding p values, in the same order, were 0.0095, 0.0005, 0.0002, 0.0090, and <0.0001, respectively. A significant overall effect was also found for the entry discrimination index (F(5,48) = 8.42, p = ୮.୯୫ ୍ଠ ୧୦ −6). Mean values were −0.016 ± 0.157 for C, −0.276 ± 0.197 for LPS 1, 0.194 ± 0.187 for D1, 0.172 ± 0.258 for D3, 0.152 ± 0.227 for T3, and 0.287 ± 0.217 for T5. Relative to LPS 1, significantly higher values were observed for C, D1, D3, T3, and T5, with p values of 0.0460, 0.000098, 0.0002, 0.0004, and <0.0001, respectively. Overall, the treatment groups showed a more novelty-oriented recall profile than the LPS 1 group, consistent with attenuation of the LPS-related exploratory deficit. The time- and entry-based discrimination indices from the intervention experiment are presented graphically in Figure 8A,B. 3.2.4. Time Allocation Between the K and U Arms During Recall Figure 9 shows recall-session time allocation between the K and U arms. For K-arm time, one-way ANOVA showed a significant overall group effect (F(5,48) = 4.51, p = 0.0019). Mean values were 71.22 ± 13.48 s in the control group, 94.67 ± 18.70 s in the LPS 1 group, 61.67 ± 19.38 s in the D1 group, 63.44 ± 17.71 s in the D3 group, 66.78 ± 22.99 s in the T3 group, and 60.67 ± 14.65 s in the T5 group. Dunnett’s test was then applied with LPS 1 as the comparator. K-arm time was significantly lower in C, D1, D3, T3, and T5, with corresponding p values of 0.0348, 0.0015, 0.0028, 0.0088, and 0.0011, respectively. For U-arm time, one-way ANOVA also showed a significant group effect (F(5,48) = 3.28, p = 0.0126). Mean values were 82.89 ± 13.74 s in the control group, 60.33 ± 17.28 s in the LPS 1 group, 90.44 ± 31.95 s in the D1 group, 95.56 ± 24.69 s in the D3 group, 77.56 ± 27.17 s in the T3 group, and 102.22 ± 28.04 s in the T5 group. Relative to LPS 1, Dunnett’s test showed significantly higher values in D3 ( p = 0.0168) and T5 ( p = 0.0034), whereas the comparisons with C, D1, and T3 were not statistically significant. Overall, these findings indicate that the LPS-related redistribution of recall-phase exploration was attenuated by treatment, with the clearest shift toward the novel arm observed in the D3 and T5 groups. 3.2.5. Recall-Phase Entry Allocation Between the K and U Arms Figure 10 shows recall-session entry counts for the K and U arms. For K-arm entries, the ANOVA result was significant (F(5,48) = 2.66, p = 0.0335). Mean values were 5.78 ± 2.22 for C, 6.56 ± 2.07 for LPS 1, 4.11 ± 1.27 for D1, 5.00 ± 2.00 for D3, 4.33 ± 1.12 for T3, and 4.33 ± 1.73 for T5. Post hoc testing against LPS 1 showed lower K-arm entry counts in D1, T3, and T5, with corresponding p values of 0.0233, 0.0451, and 0.0451, respectively. The comparisons with C and D3 did not reach statistical significance. U-arm entries also varied significantly across groups (one-way ANOVA: F(5,48) = 5.52, p = 0.0004). Mean values were 5.33 ± 1.50 for C, 3.78 ± 1.64 for LPS 1, 6.11 ± 1.69 for D1, 7.11 ± 2.20 for D3, 6.00 ± 1.73 for T3, and 7.56 ± 1.42 for T5. Compared with LPS 1, U-arm entry counts were higher in D1, D3, T3, and T5, with p values of 0.0249, 0.0007, 0.0352, and 0.0002, respectively. The C versus LPS 1 comparison was not significant ( p = 0.2129). Total arm entries were comparable across groups (F(5,48) = 0.98, p = 0.4414). Overall, these results indicate that the reduction in novelty-directed arm selection induced by LPS was attenuated by treatment, without a significant overall change in total entry counts. 3.2.6. U/K Entry Ratio Recall-phase U/K entry ratio values are summarized in Figure 11. One-way ANOVA revealed a significant overall treatment effect on this parameter (F(5,48) = 4.43, p = 0.0021). Mean values were 1.021 ± 0.392 in the control group, 0.602 ± 0.252 in the LPS 1 group, 1.607 ± 0.623 in the D1 group, 1.681 ± 1.013 in the D3 group, 1.544 ± 0.814 in the T3 group, and 2.078 ± 1.037 in the T5 group. In Dunnett’s post hoc comparisons using LPS 1 as the reference condition, significantly higher U/K entry ratio values were observed in D1 ( p = 0.0271), D3 ( p = 0.0157), T3 ( p = 0.0423), and T5 ( p = 0.0006), whereas the difference between C and LPS 1 was not statistically significant ( p = 0.6440). These findings indicate that the treatment groups showed a more novelty-directed entry pattern during recall than the LPS 1 group, as reflected by higher U/K entry ratio values. 3.2.7. Mean Time per Entry Table 4 shows MTPE for the K and U arms during recall. One-way ANOVA did not reveal a significant overall treatment effect for MTPE into the K arm (F(5,48) = 0.29, p = 0.9139) or into the U arm (F(5,48) = 0.56, p = 0.7270). These findings suggest that the between-group differences observed in recall-phase arm times and entries were not accompanied by significant changes in average dwell time per visit. 4. Discussion The present study evaluated the effects of donepezil and tacrine on recall-related exploratory behavior in a subacute lipopolysaccharide-induced model of neuroinflammation-associated cognitive impairment. The findings show that repeated LPS administration produced a consistent recall-phase behavioral alteration, reflected by reduced novelty-directed exploration and lower U/K-based indices. In the intervention experiment, cholinesterase inhibition attenuated several components of this LPS-induced behavioral alteration, with donepezil and tacrine showing partially distinct response patterns across time-based, entry-based, and ratio-derived measures. For model validation, the primary recall-phase endpoint was the U/K time ratio, because it reflects the relative allocation of exploration between the unknown and known arms at the individual-animal level. Repeated LPS administration significantly reduced this ratio, indicating a shift of recall-phase exploration away from the unknown arm and toward the known arm. This finding suggests that the subacute LPS protocol produced a measurable reduction in novelty-directed recall behavior. This interpretation was supported by the secondary outcomes. LPS-treated rats showed reduced time and entry discrimination indices, spent less time in the unknown arm, and made fewer entries into this arm during recall. In contrast, acquisition-phase parameters, total arm entries, and mean time per entry were not significantly altered, suggesting that the observed recall deficit was not primarily driven by differences in baseline exploration, global locomotor activity, or average dwell time per visit. In the intervention protocol, the U/K time ratio provided the main reference for evaluating treatment effects in relation to the LPS-induced deficit identified during model validation. Compared with the LPS 1 group, donepezil at 1 mg/kg and tacrine at 5 mg/kg significantly increased the U/K time ratio, indicating a shift toward novelty-directed exploratory allocation during recall. Although donepezil 3 mg/kg and tacrine 3 mg/kg did not reach significance for the primary ratio, their effects were further considered in relation to the supportive outcomes. These supportive measures refined the interpretation of the primary result. Both time- and entry-based discrimination indices were significantly higher in all treated groups compared with LPS 1, indicating a more novelty-oriented recall profile after cholinesterase inhibition. Absolute arm-time analysis showed that all treated groups spent less time in the familiar arm than LPS 1, whereas the clearest increases in unknown-arm time were observed in the D3 and T5 groups. Entry-based measures showed a similar redistribution, with increased entries into the unknown arm in treated animals while total arm entries remained unchanged. In addition, MTPE values were comparable among groups, suggesting that treatment-related changes were not primarily explained by altered average dwell time per visit. Overall, the intervention experiment suggests that cholinesterase inhibition modified the LPS-related recall phenotype mainly by changing arm-selection and revisiting patterns, rather than by producing a generalized increase in exploratory activity. The novelty of the present approach should be interpreted in relation to previous LPS-based behavioral studies and cholinesterase inhibitor investigations in inflammatory models. Previous LPS-based studies have evaluated cognitive impairment using paradigms such as the Morris water maze and spontaneous alternation Y-maze [ 8, 32]. Earlier studies also evaluated anti-dementia drugs, including tacrine, rivastigmine, and donepezil, in LPS-induced neuroinflammatory models, but mainly through biochemical endpoints such as IL-2 levels and acetylcholinesterase activity [ 43]. Other studies investigated donepezil in LPS- or Aβ-stimulated neuroinflammatory settings, focusing primarily on microglial activation, cytokine expression, and intracellular signaling pathways [ 25]. In contrast, the present study focused on recall-phase exploratory organization in a two-trial Y-maze paradigm and directly compared donepezil and tacrine under the same subacute systemic LPS challenge. Thus, the novelty of the present work resides not in the use of LPS or cholinesterase inhibitors separately, but in integrating a subacute inflammatory challenge with a recall-focused exploratory analysis and a direct behavioral comparison between two pharmacologically distinct cholinesterase inhibitors. At the same time, these findings should be interpreted primarily within a behavioral pharmacology framework. Because donepezil-only and tacrine-only groups were not included, these effects should be interpreted as treatment-related differences under LPS challenge rather than as intrinsic behavioral profiles of donepezil or tacrine in non-LPS-treated animals. The study indicates that subacute systemic LPS administration modifies recall-related exploratory allocation in the Y-maze and that donepezil and tacrine can differentially influence this profile. Because the present analysis focused on behavioral endpoints, the results provide functional evidence of treatment-related modulation of the LPS-induced recall phenotype, while the underlying inflammatory, cholinergic, oxidative, and cellular mechanisms remain to be clarified in future studies. From a pharmacological perspective, the observed effects are compatible with the established role of cholinergic signaling in memory-related processes and exploratory behavior [ 21, 22]. Although donepezil and tacrine both enhance cholinergic neurotransmission, their distinct pharmacological profiles may contribute to different patterns of behavioral modulation under inflammatory challenge conditions [ 24, 29, 30, 31]. In clinical terms, donepezil remains a currently used symptomatic cholinesterase inhibitor [ 24], whereas tacrine has mainly historical clinical relevance as the first approved cholinesterase inhibitor [ 29], while remaining pharmacologically important as a reference scaffold for tacrine-derived multitarget anti-Alzheimer compounds [ 30, 31]. Therefore, the present comparison should be understood as an experimental pharmacology comparison rather than as a direct reproduction of clinical efficacy. From a practical and translational perspective, the present protocol may serve as a preclinical screening framework for detecting treatment-related modulation of inflammation-associated recall deficits. By combining subacute systemic LPS exposure with a two-trial Y-maze recall task, this approach allows novelty-directed exploratory allocation to be assessed while also monitoring total activity and mean time per entry. This may help guide future studies of cholinergic and anti-inflammatory strategies in neuroinflammation-associated cognitive impairment. The non-linear pattern observed with donepezil is also noteworthy. Although donepezil 1 mg/kg significantly improved the primary U/K time ratio, donepezil 3 mg/kg did not reach significance for this parameter, despite showing favorable effects on discrimination indices and unknown-arm time. This suggests that the relationship between cholinergic enhancement and recall-related exploratory behavior may not be strictly dose-dependent. Greater cholinergic stimulation may not necessarily translate into proportional behavioral improvement, particularly in a neuroinflammatory context where performance may depend on the balance between cholinergic modulation, arousal, exploratory drive, and inflammatory signaling. Compared with our previous trihexyphenidyl-based Y-maze study, the present LPS-induced model showed a different response pattern to donepezil and tacrine, suggesting that the behavioral effects of cholinesterase inhibition may depend on the impairment-inducing agent [ 37]. This comparison supports the complementary value of inflammatory challenge and antimuscarinic models, as they address different experimentally induced aspects of Alzheimer-like pathology. Future studies should combine the present behavioral approach with biochemical, molecular, and histological endpoints to confirm the inflammatory status induced by LPS and clarify the mechanisms through which donepezil and tacrine modify recall-related exploratory behavior. In addition, future work could explore whether LPS-induced recall alterations interact with broader neuroimmune and neuromodulatory pathways, including microbiota–gut–brain axis mechanisms, which have been increasingly discussed as modulators of neuroinflammation, neurotransmitter signaling, and cognition-related behavior [ 48]. 5. Limitations Several limitations should be acknowledged. First, the two-trial Y-maze provides useful information on recall-related exploratory allocation, but performance in this task may also reflect factors other than memory, such as anxiety-like behavior, novelty preference, arousal or residual sickness-related effects following inflammatory challenge. The lack of significant differences in acquisition-phase behavior, total arm entries, and mean time per entry argues against a major nonspecific reduction in exploration, but complementary tests such as the open-field, elevated plus maze, or additional memory paradigms would help clarify the contribution of affective and locomotor factors. Although animals were housed individually, cages were placed adjacent to one another in the same room, allowing indirect sensory contact between animals while preventing direct physical interaction. Nevertheless, the absence of direct cage-mate contact may represent a background factor potentially influencing stress-related behavioral or neuroimmune responses. Since this condition was applied uniformly to all groups, relative between-group comparisons remain interpretable, but future studies could further examine whether housing conditions modulate the LPS-induced recall phenotype. Second, the present study relied on behavioral endpoints and did not include biochemical, molecular, or histological validation of neuroinflammation or cholinesterase-related mechanisms. Accordingly, the findings should be interpreted as behavioral evidence obtained under an LPS-induced inflammatory challenge, while the specific inflammatory, oxidative, cholinergic, and cellular mechanisms remain to be confirmed in future studies combining behavioral testing with biochemical, molecular, and histological endpoints. Third, the intervention protocol did not include groups receiving donepezil or tacrine in the absence of LPS. As a result, treatment-related effects should be interpreted as effects observed under LPS challenge, and cannot be fully separated from the intrinsic behavioral effects of these compounds in non-LPS-treated animals. Future studies should include drug-only groups to better distinguish baseline drug effects from effects expressed under inflammatory challenge conditions. Finally, the study used a moderate sample size and a subacute systemic LPS protocol. Although this approach is suitable for modeling selected inflammation-associated cognitive and exploratory alterations, it cannot capture the long-term, progressive, and multifactorial nature of Alzheimer’s disease. Because ratio-based measures may also be influenced by denominator variability, interpretation should rely on the convergence of the primary U/K time ratio with supportive time-, entry-, discrimination-, and time-per-entry outcomes. 6. Conclusions In conclusion, systemic LPS administration at 1 mg/kg altered exploratory behavior during the 24 h recall session of the two-trial Y-maze, mainly by reducing novelty-directed exploration. Under the same experimental conditions, donepezil and tacrine attenuated several components of the LPS-related behavioral profile, with partially distinct patterns across primary and supportive outcomes. These findings support the usefulness of this subacute LPS model for evaluating behavioral pharmacological modulation of recall-related exploratory behavior. However, further biochemical, molecular, and histological validation is required to clarify the inflammatory, oxidative, cholinergic, and cellular mechanisms underlying the observed effects. Author Contributions Conceptualization, A.-F.D. and A.C.B.; methodology, A.-F.D., A.C.B. and A.Z.; software, C.M.S.; validation, O.A.C. and S.S.; formal analysis, S.S.; investigation, A.-F.D.; data curation, A.-F.D., G.B. and M.C.D.; writing—original draft preparation, A.-F.D. and A.C.B.; writing—review and editing, A.Z., C.I.V.G. and S.S.; visualization, O.A.C., C.I.V.G. and S.F.; supervision, A.Z., S.S., O.A.C. and I.F. All authors have read and agreed to the published version of the manuscript. Funding This research received no external funding. Institutional Review Board Statement Ethical approval was obtained from the Ethics Committee of the “Carol Davila” University of Medicine and Pharmacy, Bucharest, Romania (authorization no. 13619 from 24 May 2024). Animal procedures complied with national legal requirements for scientific animal research and were authorized and supervised by the National Sanitary Veterinary and Food Safety Authority (authorization no. 9 from 5 July 2024). Informed Consent Statement Not applicable. Data Availability Statement The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding authors. Acknowledgments Publication of this paper was supported by the University of Medicine and Pharmacy Carol Davila, through the institutional program Publish not Perish. Conflicts of Interest The authors report no conflicts of interest. Abbreviations AD Alzheimer’s disease LPS lipopolysaccharide AChE acetylcholinesterase BuChE butyrylcholinesterase i.p. intraperitoneal ARRIVE Animal Research: Reporting of In Vivo Experiments SD standard deviation ANOVA analysis of variance MTPE mean time per entry U arm novel/unknown arm K arm familiar/known arm S arm start arm References Abdulkhaliq, A.A.; Kim, B.; Almoghrabi, Y.M.; Khan, J.; Ajoolabady, A.; Ren, J.; Bahijri, S.; Tuomilehto, J.; Borai, A.; Pratico, D. Amyloid-β and Tau in Alzheimer’s disease: Pathogenesis, mechanisms, and interplay. Cell Death Dis. 2026, 17, 21. [] [ CrossRef] de Oliveira, J.; Kucharska, E.; Garcez, M.L.; Rodrigues, M.S.; Quevedo, J.; Moreno-Gonzalez, I.; Budni, J. Inflammatory Cascade in Alzheimer’s Disease Pathogenesis: A Review of Experimental Findings. Cells 2021, 10, 2581. [] [ CrossRef] Handra, C.M.; Ghita, I.; Ulmeanu, A.; Enache, A.-M.; Epureanu, F.; Coman, O.A.; Coman, L.; Fulga, I. Depressive Clinical Manifestations Associated with Professional Aluminum Exposure. Rev. Chim. 2019, 70, 2162–2167. [] [ CrossRef] Skrzypczak-Wiercioch, A.; Sałat, K. Lipopolysaccharide-Induced Model of Neuroinflammation: Mechanisms of Action, Research Application and Future Directions for Its Use. Molecules 2022, 27, 5481. [] [ CrossRef] Park, B.S.; Lee, J.-O. Recognition of lipopolysaccharide pattern by TLR4 complexes. Exp. Mol. Med. 2013, 45, e66. [] [ CrossRef] Pieterse, E.; Rother, N.; Yanginlar, C.; Hilbrands, L.B.; van der Vlag, J. Neutrophils Discriminate between Lipopolysaccharides of Different Bacterial Sources and Selectively Release Neutrophil Extracellular Traps. Front. Immunol. 2016, 7, 484. [] [ CrossRef] [ PubMed] Watanabe, K.; Jaffe, E.A. Comparison of the potency of various serotypes of E. coli lipopolysaccharides in stimulating PGI2 production and suppressing ACE activity in cultured human umbilical vein endothelial cells. Prostaglandins Leukot. Essent. Fat. Acids 1993, 49, 955–958. [] [ CrossRef] Zakaria, R.; Wan Yaacob, W.M.; Othman, Z.; Long, I.; Ahmad, A.H.; Al-Rahbi, B. Lipopolysaccharide-induced memory impairment in rats: A model of Alzheimer’s disease. Physiol. Res. 2017, 66, 553–565. [] [ CrossRef] [ PubMed] Zhang, F.; Zhang, J.-G.; Yang, W.; Xu, P.; Xiao, Y.-L.; Zhang, H.-T. 6-Gingerol attenuates LPS-induced neuroinflammation and cognitive impairment partially via suppressing astrocyte overactivation. Biomed. Pharmacother. 2018, 107, 1523–1529. [] [ CrossRef] Ensandoust, T.; Khakpour-Taleghani, B.; Jafari, A.; Rostampour, M.; Rohampour, K.; Ch, M.H. Effect of simultaneous application of adenosine A1 receptor agonist and A2A receptor antagonist on memory, inflammatory factors, and PSD-95 in lipopolysaccharide-induced memory impairment. Behav. Brain Res. 2025, 476, 115210. [] [ CrossRef] [ PubMed] Denver, P.; Cunningham, C. Microglial activation and neuroinflammation in acute and chronic cognitive deficits in sepsis. Neuropharmacology 2025, 267, 110285. [] [ CrossRef] Caruntu, C.; Boda, D.; Musat, S.; Caruntu, A.; Poenaru, E.; Calenic, B.; Savulescu-Fiedler, I.; Draghia, A.; Rotaru, M.; Badarau, A. Stress effects on cutaneous nociceptive nerve fibers and their neurons of origin in rats. Romanian Biotechnol. Lett. J. 2014, 19, 9517–9530. [] Hopp, S.C.; D’Angelo, H.M.; Royer, S.E.; Kaercher, R.M.; Crockett, A.M.; Adzovic, L.; Wenk, G.L. Calcium dysregulation via L-type voltage-dependent calcium channels and ryanodine receptors underlies memory deficits and synaptic dysfunction during chronic neuroinflammation. J. Neuroinflamm. 2015, 12, 56. [] [ CrossRef] [ PubMed] Nürnberger, F.; Ott, D.; Claßen, R.; Rummel, C.; Roth, J.; Leisengang, S. Systemic Lipopolysaccharide Challenge Induces Inflammatory Changes in Rat Dorsal Root Ganglia: An Ex Vivo Study. Int. J. Mol. Sci. 2022, 23, 13124. [] [ CrossRef] Qin, L.; Wu, X.; Block, M.L.; Liu, Y.; Breese, G.R.; Hong, J.-S.; Knapp, D.J.; Crews, F.T. Systemic LPS causes chronic neuroinflammation and progressive neurodegeneration. Glia 2007, 55, 453–462. [] [ CrossRef] Kamila, P.; Kar, K.; Chowdhury, S.; Chakraborty, P.; Dutta, R.; Sowmiya, S.; Singh, S.A.; Prajapati, B.G. Effect of neuroinflammation on the progression of Alzheimer’s disease and its significant ramifications for novel anti-inflammatory treatments. IBRO Neurosci. Rep. 2025, 18, 771–782. [] [ CrossRef] Ellakwa, D.E.-S.; Rashed, L.A.; Ali, O.S.; El-Sabbagh, N.A. A study to determine the effect of nano-selenium and thymoquinone on the Nrf2 gene expression in Alzheimer’s disease. Future Sci. OA 2025, 11, 2458434. [] [ CrossRef] Wu, L.; Xian, X.; Xu, G.; Tan, Z.; Dong, F.; Zhang, M.; Zhang, F. Toll-Like Receptor 4: A Promising Therapeutic Target for Alzheimer’s Disease. Mediat. Inflamm. 2022, 2022, 7924199. [] [ CrossRef] Ishijima, T.; Nakajima, K. Inflammatory cytokines TNFα, IL-1β, and IL-6 are induced in endotoxin- stimulated microglia through different signaling cascades. Sci. Prog. 2021, 104, 368504211054985. [] [ CrossRef] [ PubMed] Kotilinek, L.A.; Westerman, M.A.; Wang, Q.; Panizzon, K.; Lim, G.P.; Simonyi, A.; Lesne, S.; Falinska, A.; Younkin, L.H.; Younkin, S.G.; et al. Cyclooxygenase-2 inhibition improves amyloid-β-mediated suppression of memory and synaptic plasticity. Brain 2008, 131, 651–664. [] [ CrossRef] Walczak-Nowicka, Ł.J.; Herbet, M. Acetylcholinesterase Inhibitors in the Treatment of Neurodegenerative Diseases and the Role of Acetylcholinesterase in their Pathogenesis. Int. J. Mol. Sci. 2021, 22, 9290. [] [ CrossRef] [ PubMed] Grabowska, W.; Bijak, M.; Szelenberger, R.; Gorniak, L.; Podogrocki, M.; Harmata, P.; Cichon, N. Acetylcholinesterase as a Multifunctional Target in Amyloid-Driven Neurodegeneration: From Dual-Site Inhibitors to Anti-Agregation Strategies. Int. J. Mol. Sci. 2025, 26, 8726. [] [ CrossRef] Cacabelos, R. Donepezil in Alzheimer’s disease: From conventional trials to pharmacogenetics. Neuropsychiatr. Dis. Treat. 2007, 3, 303–333. [] [ PubMed] [ PubMed Central] Guo, H.B.; Cheng, Y.F.; Wu, J.G.; Wang, C.M.; Wang, H.; Zhang, C.; Qiu, Z.K.; Xu, J.-P. Donepezil improves learning and memory deficits in APP/PS1 mice by inhibition of microglial activation. Neuroscience 2015, 290, 530–542. [] [ CrossRef] Marinela, C.; Isabel, G.; Ion, F. Current knowledge on bupropion and varenicline clinical efficacy in nicotine dependence. Farmacia 2015, 63, 1–7. [] Crismon, M.L. Tacrine: First Drug Approved for Alzheimer’s Disease. Ann. Pharmacother. 1994, 28, 744–751. [] [ CrossRef] Svobodova, B.; Moravcova, Z.; Misiachna, A.; Novakova, G.; Marek, A.; Finger, V.; Odvarkova, J.; Pejchal, J.; Karasova, J.Z.; Netolicky, J.; et al. Novel tacrine-based multi-target directed Ligands: Enhancing cholinesterase inhibition, NMDA receptor antagonism, and CNS bioavailability for Alzheimer’s disease treatment. Eur. J. Med. Chem. 2025, 292, 117678. [] [ CrossRef] Bahaidrah, K.A.; Alzahrani, N.A.; Aldhahri, R.S.; Mansouri, R.A.; Alghamdi, B.S. Effects of Different Lipopolysaccharide Doses on Short- and Long-Term Spatial Memory and Hippocampus Morphology in an Experimental Alzheimer’s Disease Model. Clin. Transl. Neurosci. 2022, 6, 20. [] [ CrossRef] Baldwin, H.A.; de Souza, R.J.; Sarna, G.S.; Murray, T.K.; Green, A.R.; Cross, A.J. Measurements of tacrine and monoamines in brain by in vivo microdialysis argue against release of monoamines by tacrine at therapeutic doses. Br. J. Pharmacol. 1991, 103, 1946–1950. [] [ CrossRef] Gawel, K.; Labuz, K.; Gibula-Bruzda, E.; Jenda, M.; Marszalek-Grabska, M.; Filarowska, J.; Silberring, J.; Kotlinska, J.H. Cholinesterase inhibitors, donepezil and rivastigmine, attenuate spatial memory and cognitive flexibility impairment induced by acute ethanol in the Barnes maze task in rats. Naunyn-Schmiedeberg’s Arch. Pharmacol. 2016, 389, 1059–1071. [] [ CrossRef] Nirogi, R.; Bhyrapuneni, G.; Kandikere, V.; Benade, V.; Muddana, N.; Saralaya, R.; Irappanavar, S.; Ponnamaneni, R.; Mukkanti, K. Concurrent administration of atypical antipsychotics and donepezil: Drug interaction study in rats. Eur. J. Drug Metab. Pharmacokinet. 2012, 37, 155–161. [] [ CrossRef] Telting-Diaz, M.; Lunte, C.E. Distribution of Tacrine Across the Blood–Brain Barrier in Awake, Freely Moving Rats Using in Vivo Microdialysis Sampling. Pharm. Res. 1993, 10, 44–48. [] [ CrossRef] Dragomir, A.-F.; Stoleru, S.; Zugravu, A.; Poenaru, E.; Dumitrescu, M.C.; Barbu, A.C.; Fratea, S.; Stoleru, C.M.; Coman, O.A.; Fulga, I. Differential Effects of Donepezil and Tacrine on Recall-Phase Exploration in a Trihexyphenidyl-Induced Cholinergic Impairment Y-Maze Model. Biomedicines 2026, 14, 938. [] [ CrossRef] Dastan, M.; Rajaei, Z.; Sharifi, M.; Salehi, H. Crocin Improves Cognitive Impairment in LPS-treated Rats through Anti-Apoptotic, Anti-Inflammatory, and Antioxidant Activities. Mol. Neurobiol. 2025, 62, 5804–5815. [] [ CrossRef] [ PubMed] Dastan, M.; Rajaei, Z.; Sharifi, M.; Salehi, H. Gallic acid ameliorates LPS-induced memory decline by modulating NF-κB, TNF-α, and Caspase 3 gene expression and attenuating oxidative stress and neuronal loss in the rat hippocampus. Metab. Brain Dis. 2024, 40, 12. [] [ CrossRef] [ PubMed] Cleal, M.; Fontana, B.D.; Ranson, D.C.; McBride, S.D.; Swinny, J.D.; Redhead, E.S.; Parker, M.O. The Free-movement pattern Y-maze: A cross-species measure of working memory and executive function. Behav. Res. Methods 2021, 53, 536–557. [] [ CrossRef] [ PubMed] Harrison, F.E.; Hosseini, A.H.; McDonald, M.P. Endogenous anxiety and stress responses in water maze and Barnes maze spatial memory tasks. Behav. Brain Res. 2009, 198, 247–251. [] [ CrossRef] Antunes, M.; Biala, G. The novel object recognition memory: Neurobiology, test procedure, and its modifications. Cogn. Process. 2012, 13, 93–110. [] [ CrossRef] [ PubMed] Tyagi, E.; Agrawal, R.; Nath, C.; Shukla, R. Effect of anti-dementia drugs on LPS induced neuroinflammation in mice. Life Sci. 2007, 80, 1977–1983. [] [ CrossRef] Nielsen, J.A.; Mena, E.E.; Williams, I.H.; Nocerini, M.R.; Liston, D. Correlation of brain levels of 9-amino-1,2,3,4-tetrahydroacridine (THA) with neurochemical and behavioral changes. Eur. J. Pharmacol. 1989, 173, 53–64. [] [ CrossRef] Clarke, P.B.S.; Reuben, M.; El-Bizri, H. Blockade of nicotinic responses by physostigmine, tacrine and other cholinesterase inhibitors in rat striatum. Br. J. Pharmacol. 1994, 111, 695–702. [] [ CrossRef] Adem, A.; Jossan, S.S.; Oreland, L. Tetrahydroaminoacridine inhibits human and rat brain monoamine oxidase. Neurosci. Lett. 1989, 107, 313–317. [] [ CrossRef] [ PubMed] Drukarch, B.; Kits, K.S.; Van der Meer, E.G.; Lodder, J.C.; Stoof, J.C. 9-Amino-1,2,3,4-tetrahydroacridine (THA), an alleged drug for the treatment of Alzheimer’s disease, inhibits acetylcholinesterase activity and slow outward K+ current. Eur. J. Pharmacol. 1987, 141, 153–157. [] [ PubMed] Barbu, A.C.; Stoleru, S.; Zugravu, A.; Poenaru, E.; Dragomir, A.; Costescu, M.; Aurelian, S.M.; Shhab, Y.; Stoleru, C.M.; Coman, O.A.; et al. Dopamine and the Gut Microbiota: Interactions Within the Microbiota–Gut–Brain Axis and Therapeutic Perspectives. Int. J. Mol. Sci. 2025, 27, 271. [] [ CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. © 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license. Share and Cite MDPI and ACS Style Dragomir, A.-F.; Barbu, A.C.; Stoleru, S.; Zugravu, A.; Dumitrescu, M.C.; Bazar, G.; Ghita, C.I.V.; Fratea, S.; Stoleru, C.M.; Coman, O.A.; et al. Comparative Effects of Donepezil and Tacrine on Recall-Related Exploratory Behavior in a Subacute Lipopolysaccharide-Induced Neuroinflammatory Model of Cognitive Impairment. Biomedicines 2026, 14, 1306. https://doi.org/10.3390/biomedicines14061306 AMA Style Dragomir A-F, Barbu AC, Stoleru S, Zugravu A, Dumitrescu MC, Bazar G, Ghita CIV, Fratea S, Stoleru CM, Coman OA, et al. Comparative Effects of Donepezil and Tacrine on Recall-Related Exploratory Behavior in a Subacute Lipopolysaccharide-Induced Neuroinflammatory Model of Cognitive Impairment. Biomedicines. 2026; 14(6):1306. https://doi.org/10.3390/biomedicines14061306 Chicago/Turabian Style Dragomir, Adrian-Florentin, Aurelia Cristiana Barbu, Smaranda Stoleru, Aurelian Zugravu, Maria Carina Dumitrescu, George Bazar, Cristina Isabel Viorica Ghita, Silvia Fratea, Clara Maria Stoleru, Oana Andreia Coman, and et al. 2026. "Comparative Effects of Donepezil and Tacrine on Recall-Related Exploratory Behavior in a Subacute Lipopolysaccharide-Induced Neuroinflammatory Model of Cognitive Impairment" Biomedicines 14, no. 6: 1306. https://doi.org/10.3390/biomedicines14061306 APA Style Dragomir, A.-F., Barbu, A. C., Stoleru, S., Zugravu, A., Dumitrescu, M. C., Bazar, G., Ghita, C. I. V., Fratea, S., Stoleru, C. M., Coman, O. A., & Fulga, I. (2026). Comparative Effect
