Introduction
Autism spectrum disorder (ASD) is a neuropsychiatric and neurodevelopmental disorder and comprises a series of neurological diseases that share serious alterations of the development of the central nervous system.1 ASD is characterized by core symptoms of impairments in social interactions and communication, and restricted/repetitive and stereotyped behavior.2 Although the full pathophysiology of ASD has not yet been clarified, numerous studies indicate disruptions in several neurotransmitter systems, including acetylcholine, histamine, and dopamine.3–5 Neuroinflammatory processes have also been shown to contribute significantly to ASD development.6
Notably, emerging research has strengthened the idea that cholesterol imbalance may underlie a subset of ASD cases. A recent clinical investigation, for example, identified a group of individuals with ASD who exhibited altered lipid profiles, including elevated low-density lipoprotein (LDL), total cholesterol, and triglycerides classifying them as having dyslipidemia-associated ASD.7
Statins act by inhibiting the enzyme 3-hydroxy-3-methylglutaryl–coenzyme A (HMG-CoA) reductase, thereby blocking the synthesis of cholesterol and isoprenoid compounds.8 A number of investigations have explored the neuroprotective actions of statins across various pathological conditions, including neurodegenerative disorders, traumatic brain injury, and stroke.9–12 Notably, many of these neuroprotective effects involve modulation of inflammatory and signalling pathways. In experimental traumatic brain injury models, statins have been shown to lessen neurological impairments and promote processes such as synaptogenesis, angiogenesis, and neuronal survival.13 Beyond these effects, simvastatin has been shown to enhance the performance of healthy rats in tasks assessing object location memory and passive avoidance.14 Other cholesterol-lowering statins, including fluvastatin, simvastatin, and pravastatin, display distinct protective actions depending on the timing of treatment in models of Aβ-induced cognitive impairment, likely through preservation of cholinergic neurons or modulation of glutamatergic signaling.15,16 Clinical observations further suggest that statin therapy may be associated with reduced rates of depression and anxiety, although the underlying mechanisms remain unclear.17
In addition to their cholesterol-lowering effects, statins exert neuroprotective actions, including anti-inflammatory, antioxidant, and immunomodulatory effects, mediated in part through inhibition of isoprenoid-dependent signaling pathways. Statins act earlier in the mevalonate pathway, influencing the synthesis of multiple downstream products. These include Coenzyme Q, essential for mitochondrial respiration; farnesyl and geranylgeranyl pyrophosphate, involved in protein post-translational modification and isopentenyl tRNAs, important for RNA transcription.
Because of these wide-ranging effects on various cellular processes, statins are considered to have “pleiotropic effects”.8,18,19 These multi-target neurobiological actions provide a strong rationale for repurposing statins in neurodevelopmental and neuropsychiatric disorders.
Adjunctive simvastatin treatment has been reported to improve selected behavioral symptoms in individuals with ASD, while clinical studies of lovastatin in Fragile X syndrome have reported potential benefits on behavioral outcomes.20–22 The HMG-CoA reductase inhibitor lovastatin (LOV) was found to reverse the neurofibromatosis type 1-induced learning and attention deficits in mice.23 Likewise, LOV has also been shown to effectively correct electrophysiological and behavioral phenotypes in the mouse model of fragile X syndrome that shares numerous feature similarities with ASD.24,25 Although emerging studies have begun to investigate the effects of statins in experimental models of ASD, such as the valproic acid-induced model, the available evidence remains limited.26 In particular, the effects of LOV in the BTBR T+ tf/J (BTBR) mouse model, an idiopathic model that closely recapitulates core ASD-like behavioral phenotypes have not been previously characterized. The BTBR mouse strain is a widely used idiopathic model of ASD that exhibits core behavioral domains of the disorder. BTBR mice exhibit marked deficits in social approach, reduced reciprocal social interactions, and impaired juvenile play behavior compared with C57BL/6J controls.27–30 BTBR mice also exhibit a neuroinflammatory phenotype, marked by elevated baseline levels of pro-inflammatory cytokines and increased neuroinflammation.
This pattern mirrors immune abnormalities commonly seen in human autoimmune and neurodevelopmental conditions.31 Considering these facts, the current study aimed to evaluate the effects of systemic administration of LOV (10, 20, and 40 mg/kg, i.p.) on ASD-like behaviors in male BTBR mice subjected to different behavioral tasks, marble burying, three arm maze and three chamber test, which assess repetitive and social behaviors in mice. In addition, the effects of chronic systemic treatment with LOV on neuroinflammatory markers were assessed in brain. Moreover, serum and brain cholesterol levels, together with brain HMG-CoA reductase levels, were assessed in LOV-treated BTBR mice.
Materials and Methods
Animals
All studies were conducted using male C57BL/6J (C57), that provides a baseline as controls, and age-matched BTBR T+Itpr3tf/J (BTBR) mice. A total of 34 male mice were used in the study. Of these, 30 mice were allocated to the experimental groups, comprising 6 C57BL/6J mice and 24 BTBR T+Itpr3tf/J mice.
An additional four untreated male mice (2 C57BL/6J and 2 BTBR, aged 8–10 weeks) were used exclusively as stimulus mice for the three-chamber test. These stimulus mice were separate from the experimental treatment groups and were not included in the behavioral or biochemical analyses. At the start of the experimental treatment (LOV or vehicle administration), BTBR mice weighed 27–32 g (8–10 weeks old), while C57BL/6J mice weighed 23–27 g (8–10 weeks old).
All mice were initially obtained from Jackson Laboratories (Bar Harbor, ME, USA) and subsequently maintained as in-house breeding colonies at the College of Medicine and Health Sciences, United Arab Emirates University (UAEU). Mice were housed in a pathogen-free, standard plastic cages within the facility maintained at 22 ± 2 °C under a 12-h light/dark cycle (lights on at 06:00 am). Standard chow and water were provided ad libitum, and trained animal-care staff performed daily health assessments.
All experimental procedures were approved by the Institutional Animal Ethics Committee (Approval No. ERA-2017-5603, approval date: 03 October 2017) of the United Arab Emirates University and were conducted in accordance with Directive 2010/63/EU on the protection of animals used for scientific purposes. Efforts were made to minimize the number of animals used and any potential suffering.
Drug Administration and Treatment Protocol
LOV was purchased commercially from Pharmaffiliates analytics and synthetics (CAS No: 75330-75-5). A concentrated stock of LOV was prepared in DMSO (final concentration <1%) and subsequently diluted with normal saline to achieve the required doses (10, 20, and 40 mg/kg) prior to administration. The mice were divided into 5 groups and received chronic systemic administration as follows: BTBR and C57BL/6J control groups received vehicle (saline) intraperitoneally (i.p.)., while the BTBR treatment groups were administered LOV (10, 20 and 40 mg/kg, i.p.) dissolved in vehicle (continued for 21 days).
Behavioural assessments were initiated after 7 days of treatment. Marble burying, three-arm maze, and three chamber test were conducted on days 8, 10, and 12, respectively. On each test day, the treatments or vehicle were administered 30 minutes before each behavioural assessment.
Drug administration continued throughout the behavioural testing period until day 21. The volume of drugs administered to each animal was 10 mL/kg adjusted to body weight, and all doses were selected based on the results of previous studies and expressed in terms of the free bases.32,33 A formal a priori sample-size calculation was not performed. The group size was selected based on previous studies employing the BTBR mouse model and comparable behavioral paradigms.34–38 Six animals were allocated to each experimental group and were included in the behavioral tests (n=6 per group).
All six animals per group completed the behavioral assessments and were included in the behavioral analyses. Immediately following completion of the treatment regimen, animals were sacrificed and blood and brain samples were collected for biochemical analyses. For the subsequent biochemical analyses, five independent samples per group were analysed (n=5 per group), as sufficient sample volume for the required biochemical assays was available from five animals per group.
No observations were excluded from the final behavioral or biochemical datasets. The biochemical assays were not performed as technical replicates, therefore each value included in the statistical analyses represents one independent sample from one animal. C57BL/6J mice received vehicle treatment and served as a reference control group; LOV was not administered to reference control.
Figure 1 depicts the timeline for experimental injections and behavioral studies conducted in BTBR mice. Figure 1Experimental timeline of the study design and tasks. Mice received treatment of lovastatin (10, 20, or 40 mg/kg) or vehicle for 21 consecutive days.
Behavioural testing was initiated after 7 days of treatment. Marble burying, three-arm maze, and three chamber tests were conducted on days 8, 10, and 12, respectively. On each test day, behavioural assessments were performed 30 minutes after lovastatin administration.
Drug treatment continued until day 21, after which animals were sacrificed, and samples were collected for biochemical analyses.A study timeline diagram of lovastatin dosing, behavioral tests and sacrifice over 21 days.
Behavioral Assessments
Marble Burying Test
Marble burying behaviour was assessed over a 30-minute period as an index of repetitive and compulsive-like behaviour and is also considered a measure of anxiety-like behaviour in rodents.39 The test was performed in standard polycarbonate cages (approximately 30×20 × 13 cm) containing a 5-cm-deep layer of clean bedding. Twenty glass marbles (diameter 1.5–2 cm) were placed gently on the surface of the bedding in a 4×5 grid, with equal spacing between marbles. Mice were individually transferred to the test cage and allowed to explore freely for thirty min under low ambient light.
At the end of the session, animals were removed and the number of marbles buried was scored by an experimenter blind to group allocation. A marble was defined as “buried” when at least two-thirds of its surface area was covered by bedding. The cage bedding was replaced, and marbles were cleaned with 70% ethanol and dried between trials to remove any olfactory cues.
All tests were carried out during the light phase (between 09:00 and 15:00 h) in a quiet, dedicated behavioural room. Animals were habituated to handling for at least 5 days before testing.
Three-Arm-Maze Test
The three arm maze test primarily used to assess spatial working memory was used to assess cognitive inflexibility, where decreased alternation behaviour is indicative of repetitive behavioral phenotypes associated with ASD.40 The maze was constructed from opaque plastic material and consisted of three identical arms arranged at 120° angle. Mice were placed at the centre and allowed to freely explore all three arms for 8 min. Arm entries were recorded using manual scoring, and an alternation was defined as consecutive entries into all three arms without repetition.
Spontaneous alternation performance (SAP) was calculated as: \documentclass[12pt]{minimal} \usepackage{wasysym} \usepackage[substack]{amsmath} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage[mathscr]{eucal} \usepackage{mathrsfs} \DeclareFontFamily{T1}{linotext}{} \DeclareFontShape{T1}{linotext}{m}{n} {linotext }{} \DeclareSymbolFont{linotext}{T1}{linotext}{m}{n} \DeclareSymbolFontAlphabet{\mathLINOTEXT}{linotext} \begin{document}$$\rm SAP \left(\% \right) = \left[{\left({number\ of\ alternations} \right) / \left({total\ arm\ entries - 2} \right)} \right] \times 100.$$\end{document}
The maze was cleaned thoroughly with 70% ethanol between trials to remove all olfactory cues.
Three Chamber Test (TCT)
TCT for sociability and social novelty preference was conducted over three 10-minute sessions as per our previously published studies.39,41 The test included three 10-minute phases: habituation, sociability, and social novelty. Four untreated male mice (2 C57BL/6J and 2 BTBR), aged 8–10 weeks, were used as stimulus mice and were separate from the experimental treatment groups. The C57BL/6J stimulus mice were used for testing the C57BL/6J experimental group, whereas the BTBR stimulus mice were reused across the BTBR vehicle and lovastatin-treated experimental groups.
The mice that had not previously been subjected to behavioral testing or pharmacological treatment were used as stimulus mice. Stimulus mice were housed separately from the experimental animals throughout the study, and each experimental mouse had no prior contact with the stimulus mice before testing. The stimulus animals were not part of the experimental treatment groups and were not included in the statistical analyses.
In the sociability phase, mice chose between a novel mouse (NM) and an empty cage (NO), and time spent near each was recorded. In the social novelty phase, the empty cage was replaced with a new unfamiliar mouse (NM), while the previously introduced mouse became the familiar mouse (FM), to assess preference for social novelty. The sociability index (SI) and social novelty index (SNI) were used to compare social behaviours between the treated groups.
The sociability index (SI) was calculated from [Time exploring NM – Time exploring NO] / [Time exploring NM + Time exploring NO]; while social novelty index (SNI) was calculated from [Time exploring NM – Time exploring FM] / [Time exploring NM + Time exploring FM]. Behavioral activity was recorded and quantified using the EthoVision® Software (Noldus, Netherlands). Specifically, predefined interaction zones were established around each wire cage, and investigation time was quantified as the cumulative duration for which the nose point of the test mouse was located within the corresponding interaction zone.
During the sociability phase, investigation time for the NM and the empty wire cage (NO) were quantified to calculate the SI. During the social novelty phase, interaction time for the novel mouse (NM) and familiar mouse (FM) were quantified to calculate SNI. Since C57BL/6J and BTBR experimental mice were assessed using strain-matched stimulus mice, comparisons between the C57BL/6J and BTBR groups reflect both test-mouse strain and stimulus-mouse strain and should not be interpreted as a pure test-mouse strain comparison.
However, this does not apply to comparisons between BTBR vehicle and LOV-treated groups, as BTBR stimulus mice were used consistently.
Blood and Tissue Collection for Determining Cholesterol and HMG-CoA Reductase Levels
Following completion of treatment regimen, animals were euthanized in accordance with previously established protocols.39 Anaesthesia was induced using pentobarbital (40 mg/kg, i.p). Under anaesthesia, blood samples were collected from mice via cardiac puncture using sterile syringes. Whole blood was transferred to plain 1.5 mL microcentrifuge tubes and allowed to clot at room temperature for 30 min, followed by centrifugation at 2000–3000 × g for 10 min at 4 °C.
The resulting serum supernatant was carefully aspirated, aliquoted into labelled tubes to avoid repeated freeze–thaw cycles, and stored at −80 °C until biochemical analyses were performed. Similarly, after blood was collected, the mouse was perfused transcardially to remove remaining blood. 1× PBS (0.01 M phosphate buffer, 0.0027 M KCl, 0.137 M NaCl) was used for perfusion.
Brains were extracted rapidly onto an ice-cooled petridish and immediately snap-frozen using liquid nitrogen for subsequent analysis of cholesterol levels. Brains were stored at −80 °C until analysis. Whole-brain homogenates were prepared for the biochemical analyses.
Whole-brain tissue was used to obtain an integrated assessment of brain cholesterol-related and neuroinflammatory alterations following LOV treatment. Free Cholesterol assay kit (Elabscience E-BC-K004-M) was used to measure serum and tissue cholesterol following the manual’s instructions. A commercially available 3-hydroxy-3-methylglutaryl-CoA reductase ELISA kit (ImmunoTag -cat: ITLK00431) was used to quantify the HMG-CoA reductase levels, following the manufacturer’s protocol.
Neuroinflammatory Markers
The levels of the proinflammatory cytokines tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) were quantified in the brain tissues using enzyme-linked immunosorbent assay (ELISA) as per the instructions of the manufacturer and our previously published experimental protocols.36 Commercially available ELISA kits (Duo Set) for IL-6 (Catalog No.: DY406) and TNF-α (Catalog No.: DY410) were purchased from R&D Systems.
Statistical Analysis
For the behavioural, cholesterol, HMG-CoA reductase, and neuroinflammatory outcomes, data are presented as mean ± standard deviation (SD). Each outcome was analysed separately using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference. Normality of model residuals was assessed using Q–Q plots and the Shapiro–Wilk test, and homogeneity of variance was assessed using the Brown–Forsythe test.
Individual animal observations and residual plots were examined for potentially influential observations. Statistical analyses were performed using GraphPad Prism, with p < 0.05 considered statistically significant. Exact multiplicity-adjusted p-values and 95% confidence intervals for treatment comparisons are reported.
Individual animal-level data, descriptive statistics, and detailed one-way ANOVA and Dunnett-adjusted multiple-comparison results are provided in the Supplementary file (Tables S1–S10).
Results
To evaluate the effects of LOV on ASD-like behaviours, we conducted behavioural assessments comprising the marble burying test, three-arm maze test and three chamber test. The behavioural experiments included five animal groups: C57, BTBR, and LOV treated groups at doses of 10, 20, and 40 mg/kg. LOV was administered once daily for 21 consecutive days.
C57 mice receiving vehicle injections served as the baseline control group for assessing the extent of behavioural deficits in BTBR mice and were not subjected to LOV treatment. The behavioral tests constituted the principal efficacy outcomes and evaluated core autism-related behavior, including repetitive and social behaviors. HMG-CoA reductase, serum and brain cholesterol measurements and neuroinflammatory markers were secondary biochemical outcomes.
Effects of Lovastatin on Repetitive Behaviours of BTBR Mice in Marble Burying Test
Marble burying behaviour was assessed on treatment day 8 following repeated LOV administration. As shown in Figure 2, vehicle-treated BTBR mice exhibited greater marble-burying behaviour than vehicle-treated C57BL/6J mice. One-way ANOVA revealed a significant overall effect of group on the percentage of marbles buried ((F (4,25) = 16.59), (p < 0.0001)).
Dunnett-adjusted multiple comparisons, using the BTBR vehicle group as the reference, showed that C57BL/6J mice buried significantly fewer marbles than BTBR (mean difference [C57BL/6J − BTBR]: −58.83 percentage points, 95% CI: −80.25 to −37.41, adjusted p < 0.0001). Treatment effects are expressed as LOV minus BTBR vehicle. Compared with vehicle-treated BTBR mice, LOV at 10 mg/kg did not produce a statistically significant reduction in marble burying (mean difference: −6.67 percentage points, 95% CI: −28.09 to 14.75, adjusted p = 0.8326).
However, LOV 20 mg/kg (mean difference: −30.00 percentage points, 95% CI: −51.42 to −8.58, adjusted p = 0.0043) and 40 mg/kg (mean difference: −35.83 percentage points, 95% CI: −57.25 to −14.41, adjusted p = 0.0007) significantly reduced marble-burying relative to BTBR vehicle. Thus, statistically significant reductions in repetitive marble-burying behavior were observed at 20 and 40 mg/kg with the largest observed mean reduction occurring at 40 mg/kg, whereas no statistically detectable effect was observed at 10 mg/kg. Figure 2Effects of lovastatin on marble-burying behavior in BTBR mice.
Marble-burying behavior was assessed on treatment day 8 in vehicle-treated C57BL/6J mice, vehicle-treated BTBR mice and BTBR mice treated with lovastatin (LOV 10, 20, or 40 mg/kg, i.p.). Data are presented as mean ± SD (n=6 per group). Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference.
The overall group effect was significant ((F (4,25) = 16.59), (p < 0.0001)). Dunnett-adjusted comparisons versus the BTBR vehicle group were: C57BL/6J, adjusted p < 0.0001; LOV 10 mg/kg, adjusted p = 0.8326; LOV 20 mg/kg, adjusted p = 0.0043; and LOV 40 mg/kg, adjusted p = 0.0007 (****p < 0.0001, ***p < 0.001, ** p <0.01 Dunnett-adjusted comparison versus vehicle treated BTBR mice).Y axis label, Marbles buried left parenthesis percent right parenthesis, range 0 to 120. X axis categories, C 57, B T B R, 10, 20, 40, with an underline under 10, 20, 40 and a grouped label, L O V left parenthesis milligram per kilogram comma i dot p dot right parenthesis.
Five vertical bars with error bars and overlaid open circle points. Approximate bar heights: C 57 about 22 percent, B T B R about 82 percent, 10 about 75 percent, 20 about 52 percent, 40 about 46 percent. Ranking by height: B T B R highest, then 10, then 20, then 40, then C 57.
Significance brackets above bars: C 57 to B T B R labeled four asterisks; BTBR to 20 labeled two asterisks; B T B R to 40 labeled three asterisks.A bar graph showing marbles buried percent for C 57, B T B R and L O V doses 10, 20 and 40.
Effects of Lovastatin on Alternation Behaviours of BTBR Mice in Three-Arm Maze
Spontaneous alternation was assessed on treatment day 10 following repeated LOV administration. As shown in Figure 3, vehicle-treated BTBR mice exhibited lower alternation than vehicle-treated C57BL/6J mice. One-way ANOVA revealed a significant overall effect of group on spontaneous alternation ((F (4,25) = 25.40), (p < 0.0001)).
Using the BTBR vehicle group as the reference, Dunnett-adjusted multiple comparisons showed that spontaneous alternation was significantly higher in C57BL/6J mice than in BTBR mice (mean difference [C57BL/6J − BTBR]: 25.62 percentage points, 95% CI: 15.66 to 35.58; adjusted p < 0.0001). Treatment effects are expressed as LOV minus BTBR vehicle. Compared with vehicle-treated BTBR mice, LOV at 10 mg/kg (mean difference: −5.16 percentage points, 95% CI: −15.12 to 4.80, adjusted p = 0.4788) and 20 mg/kg (mean difference: −4.46 percentage points, 95% CI: −14.42 to 5.49, adjusted p = 0.5989) did not produce significant changes in spontaneous alternation.
In contrast, LOV at 40 mg/kg significantly increased spontaneous alternation with respect to BTBR vehicle mice (mean difference: 15.30 percentage points, 95% CI: 5.34 to 25.26, adjusted p = 0.0018). Thus, a statistically detectable improvement in spontaneous alternation was observed at 40 mg/kg, whereas no significant effect was observed at 10 or 20 mg/kg. Figure 3Effect of lovastatin on percentage of alternation in BTBR mice assessed by three-arm maze.
Spontaneous alternation was assessed in the three-arm maze on treatment day 10. Data are presented as mean ± SD (n=6 per group). Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference.
The overall group effect was significant ((F (4,25) = 25.40), (p < 0.0001)). Dunnett-adjusted comparisons were: C57 versus BTBR, adjusted p < 0.0001; LOV 10 mg/kg versus BTBR, adjusted p = 0.4788; LOV 20 mg/kg versus BTBR, adjusted p = 0.5989; and LOV 40 mg/kg versus BTBR, adjusted p = 0.0018 (****p < 0.0001, ** p <0.01 Dunnett-adjusted comparison versus vehicle treated BTBR mice).A bar graph showing alternation percent for C57, BTBR and lovastatin dose groups.
Effects of Lovastatin on Sociability and Social Novelty Index of BTBR Mice in TCT
Figure 4 illustrates the effects of administration of LOV at three doses (10, 20 and 40 mg/kg, i.p.) On ASD-like sociability deficits in the TCT conducted on day 12 of experimental study. As seen in Figure 4A, vehicle-treated BTBR mice demonstrated a markedly lower sociability index (SI) than vehicle-treated C57BL/6J mice. One-way ANOVA revealed a significant overall effect of group on SI ((F (4,25) = 46.58), (p < 0.0001)).
Dunnett-adjusted multiple comparisons, using the BTBR vehicle group as the reference, showed that SI was significantly higher in C57BL/6J mice than in BTBR mice (mean difference [C57BL/6J − BTBR]: 0.8867, 95% CI: 0.582 to 1.191, adjusted p < 0.0001). Treatment effects are expressed as LOV minus BTBR vehicle. LOV treatment significantly increased SI relative to BTBR vehicle treated mice at 10 mg/kg (mean difference: 1.170, 95% CI: 0.865 to 1.475, adjusted p < 0.0001), 20 mg/kg (mean difference: 1.212, 95% CI: 0.907 to 1.516, adjusted p < 0.0001), and 40 mg/kg (mean difference: 1.448, 95% CI: 1.144 to 1.753, adjusted p < 0.0001).
Thus, significant improvements in sociability index relative to BTBR vehicle were observed at all three LOV doses (Figure 4A). Figure 4Lovastatin ameliorates social deficits in BTBR mice. Social behavior was assessed using the three-chamber test on treatment day 12 in vehicle-treated C57BL/6J mice, vehicle-treated BTBR mice, and BTBR mice treated with lovastatin (LOV; 10, 20, or 40 mg/kg, i.p.).
(A) Sociability index (SI). (B) Social novelty index (SNI). Data are presented as mean ± SD (n=6 per group).
Statistical analyses were performed separately using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference. For SI and SNI, the overall group effect was significant ((F (4,25) = 46.58), (p < 0.0001)) and ((F (4,25) = 21.93), (p < 0.0001)), respectively. For SI, Dunnett-adjusted comparisons versus BTBR vehicle were: C57BL/6J, adjusted p < 0.0001; LOV 10 mg/kg, adjusted p < 0.0001; LOV 20 mg/kg, adjusted p < 0.0001; and LOV 40 mg/kg, adjusted p < 0.0001.
For SNI, Dunnett-adjusted comparisons versus BTBR vehicle were: C57BL/6J, adjusted p < 0.0001; LOV 10 mg/kg, adjusted p < 0.0001; LOV 20 mg/kg, adjusted p < 0.0001; and LOV 40 mg/kg, adjusted p < 0.0001 (****p < 0.0001 Dunnett-adjusted comparison versus vehicle treated BTBR mice).A) Vertical bar graph of Sociability Index left parenthesis S I right parenthesis. X axis categories: C57, B T B R, 10, 20, 40, with label, Lovastatin left parenthesis milligram over kilogram semicolon i dot p dot right parenthesis under 10, 20, 40. Y axis label, Sociability Index left parenthesis S I right parenthesis, range minus 1.5 to 1.0.
Bars with overlaid points: C57 about 0.3; B T B R about minus 0.6; 10 about 0.55; 20 about 0.55; 40 about 0.85. Significance brackets marked with four asterisks appear over C57 versus B T B R and over B T B R versus 10, 20 and 40. B) Vertical bar graph of Social Novelty Index left parenthesis S N I right parenthesis.
X axis categories: C57, B T B R, 10, 20, 40, with the same Lovastatin label. Y axis label, Social Novelty Index left parenthesis S N I right parenthesis, range minus 1.0 to 1.0. Bars with overlaid points: C57 about 0.35; B T B R about minus 0.3; 10 about 0.35; 20 about 0.35; 40 about 0.5.Two bar graphs showing sociability index and social novelty index across mouse groups and lovastatin doses.
As seen in Figure 4B, vehicle-treated BTBR mice exhibited a lower social novelty index (SNI) than vehicle-treated C57BL/6J mice, indicating impaired social novelty preference. One-way ANOVA revealed a significant overall effect of group on SNI ((F (4,25) = 21.93), (p < 0.0001)). Dunnett-adjusted multiple comparisons, using the BTBR vehicle group as the reference, showed that SNI was significantly higher in C57BL/6J mice than in BTBR mice (mean difference [C57BL/6J − BTBR]: 0.680, 95% CI: 0.430 to 0.930, adjusted p <0.0001).
Significantly higher SNI values were observed when compared with vehicle-treated BTBR mice, following LOV treatment at 10 mg/kg (mean difference: 0.673, 95% CI: 0.423 to 0.923, adjusted p < 0.0001), 20 mg/kg (mean difference: 0.685, 95% CI: 0.435 to 0.935, adjusted p < 0.0001), and 40 mg/kg (mean difference: 0.7767, 95% CI: 0.527 to 1.027, adjusted p < 0.0001). Thus, LOV treatment was associated with significantly higher social novelty indices at all three doses relative to BTBR vehicle.
Effects of Lovastatin on Cholesterol Biosynthesis
Effects of Lovastatin on HMG-CoA Reductase Levels in LOV-Treated Mice
As shown in Figure 5, brain HMG-CoA reductase levels showed significant differences among the experimental groups following the 21-day chronic LOV treatment ((F (4,20) = 9.174), (p = 0.0002)). Dunnett-adjusted multiple comparisons, using the BTBR vehicle group as the reference, showed significantly higher HMG-CoA reductase levels in BTBR mice than in C57BL/6J controls (mean difference [C57BL/6J − BTBR]: −3.372, 95% CI: −5.101 to −1.644, adjusted p = 0.0002). Treatment effects are expressed as LOV minus BTBR vehicle.
Compared with vehicle-treated BTBR mice, HMG-CoA reductase levels were significantly lower following LOV treatment at 10 mg/kg (mean difference: −3.436, 95% CI: −5.164 to −1.707, adjusted p = 0.0001), 20 mg/kg (mean difference: −2.484, 95% CI: −4.212 to −0.756, adjusted p = 0.0039), and 40 mg/kg (mean difference: −2.533, 95% CI: −4.261 to −0.804, adjusted p = 0.0033). Thus, following 21 days of treatment, significantly lower brain HMG-CoA reductase levels were observed at all three LOV doses relative to BTBR mice administered vehicle. Figure 5Effects of lovastatin on brain HMG-CoA reductase levels in BTBR mice.
Brain HMG-CoA reductase levels (ng/mL) were measured following completion of the 21-day treatment in vehicle-treated C57BL/6J mice, vehicle-treated BTBR mice, and BTBR mice treated with lovastatin (LOV; 10, 20, or 40 mg/kg, i.p.). Data are presented as mean ± SD (n=5 per group). Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference.
The overall group effect was significant ((F (4,20) = 9.174, (p = 0.0002)). Dunnett-adjusted comparisons versus the BTBR vehicle group were: C57BL/6J, adjusted p = 0.0002; LOV 10 mg/kg, adjusted p = 0.0001; LOV 20 mg/kg, adjusted p = 0.0039; and LOV 40 mg/kg, adjusted p = 0.0033 (***p < 0.001 and **p < 0.01, Dunnett-adjusted comparisons versus BTBR mice administered vehicle).A bar graph showing brain H M G - C O A reductase levels across mouse groups and lovastatin doses.
Comparative Analysis of the Effects of Lovastatin on Serum and Brain Cholesterol Levels
Levels of serum cholesterol and free cholesterol in the brain differed markedly between C57BL/6 and BTBR mice as can be seen in Figure 6. Serum cholesterol levels (Figure 6A) showed significant differences among the experimental groups following the 21-day LOV treatment ((F (4,20) = 8.358), (p = 0.0004)). Dunnett-adjusted multiple comparisons, using the BTBR vehicle group as the reference, showed significantly higher serum cholesterol levels in vehicle-treated BTBR mice than in C57BL/6J controls (mean difference [C57BL/6J − BTBR]: −1.654, 95% CI: −2.614 to −0.693, adjusted p = 0.0007).
Treatment effects were expressed as LOV minus BTBR vehicle. Compared with vehicle-treated BTBR mice, serum cholesterol levels were significantly lower following LOV treatment at 10 mg/kg (mean difference: −1.210, 95% CI: −2.170 to −0.249, adjusted p = 0.0113), 20 mg/kg (mean difference: −1.436, 95% CI: −2.396 to −0.475, adjusted p = 0.0028), and 40 mg/kg (mean difference: −1.908, 95% CI: −2.868 to −0.947, adjusted p = 0.0002). The largest observed mean reduction in serum cholesterol occurred at 40 mg/kg of LOV relative to BTBR vehicle.
Figure 6Comparison of the effects of lovastatin on serum and brain cholesterol levels. Cholesterol levels were measured following completion of the 21-day treatment regimen in vehicle-treated C57BL/6J mice, vehicle-treated BTBR mice, and BTBR mice treated with lovastatin (LOV; 10, 20, or 40 mg/kg, i.p.). (A) Serum cholesterol levels.
(B) Brain cholesterol levels. Data are presented as mean ± SD (n=5 per group). Statistical analyses were performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference.
For serum cholesterol, the overall group effect was significant ((F (4,20) = 8.358, (p = 0.0004)). Dunnett-adjusted comparisons versus the BTBR vehicle group were: C57BL/6J, adjusted p = 0.0007; LOV 10 mg/kg, adjusted p = 0.0113; LOV 20 mg/kg, adjusted p = 0.0028; and LOV 40 mg/kg, adjusted p = 0.0002. For brain cholesterol, the overall group effect was significant ((F (4,20) = 23.02, (p < 0.0001)).
Dunnett-adjusted comparisons versus the BTBR vehicle group were: C57BL/6J, adjusted p < 0.0001; LOV 10 mg/kg, adjusted p = 0.205; LOV 20 mg/kg, adjusted p = 0.091; and LOV 40 mg/kg, adjusted p = 0.0506 (****p < 0.0001, ***p < 0.001, **p < 0.01, and *p < 0.05, Dunnett-adjusted comparisons versus BTBR vehicle).A) Bar graph of Serum Cholesterol. Bars are vertical. X axis categories: C57, BTBR, 10, 20, 40, with grouped label, LOV left parenthesis mg over kg comma i dot p dot right parenthesis under 10, 20, 40.
Y axis label: Serum Cholesterol left parenthesis m m o l slash L right parenthesis, range 0 to 6 with ticks at 0, 2, 4, 6. Bar heights: C57 about 3.0; BTBR about 4.6; 10 about 3.4; 20 about 3.2; 40 about 2.7. Open circle points per group: 5.
Significance brackets: C57 versus BTBR marked three asterisks; BTBR versus 10 marked one asterisk; BTBR versus 20 marked two asterisks; BTBR versus 40 marked three asterisks. B) Bar graph of Brain Cholesterol. Bars are vertical.
Same x axis categories and grouped label as A). Y axis label: Brain Cholesterol left parenthesis m m o l slash kg right parenthesis, range 0 to 80 with ticks at 0, 20, 40, 60, 80. Bar heights: C57 about 66; BTBR about 49; 10 about 43; 20 about 41; 40 about 41.
Open circle points per group: 5.Two bar graphs showing serum and brain cholesterol across C57, BTBR and lovastatin doses.
For brain cholesterol (Figure 6B), one-way ANOVA revealed a significant overall group effect ((F (4,20) = 23.02), (p < 0.0001)). Vehicle-treated BTBR mice had significantly lower brain cholesterol than C57BL/6J controls (mean difference: 18.31 [C57BL/6J − BTBR], 95% CI: 9.4 to 27.14, adjusted p < 0.0001). Treatment effects were expressed as LOV minus BTBR vehicle.
Compared with BTBR vehicle, no statistically detectable differences were observed at tested doses of LOV (10 mg/kg (mean difference: −6.35, 95% CI: −15.18 to 2.472, adjusted p = 0.205) or 20 mg/kg (mean difference: −7.82, 95% CI: −16.65 to 0.996, adjusted p = 0.091)) or 40 mg/kg LOV (mean difference: −8.80, 95% CI: −17.63 to 0.0209, adjusted p = 0.0506).
Mitigating Effects of Lovastatin on Different Neuroinflammatory Markers, IL-6 and TNF-α
Brain IL-6 and TNF-α levels differed significantly among the experimental groups following the 21-day treatment regimen. For IL-6 (Figure 7A), one-way ANOVA revealed a significant overall group effect ((F (2,12) = 21.90), (p < 0.0001)). Dunnett-adjusted comparisons, using the BTBR vehicle group as the reference, showed significantly higher IL-6 levels in BTBR mice than in C57BL/6J controls (mean difference [C57BL/6J − BTBR]: −59.02, 95% CI: −82.68 to −35.35, adjusted p < 0.0001).
The treatment effect was expressed as LOV minus BTBR vehicle. IL-6 levels were significantly lower in BTBR mice treated with LOV at 40 mg/kg than in BTBR vehicle-treated mice (mean difference: −47.55, 95% CI: −71.22 to −23.88, adjusted p = 0.0006). Figure 7Mitigating effects of lovastatin on IL-6 and TNF-α levels in BTBR mice.
Brain cytokine levels (pg/mg of protein) were measured following completion of the 21-day treatment regimen in vehicle-treated C57BL/6J mice, vehicle-treated BTBR mice, and BTBR mice treated with lovastatin (LOV; 40 mg/kg, i.p.). (A) IL-6 levels. (B) TNF-α levels.
Data are presented as mean ± SD (n= 5 per group). Statistical analyses were performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test, with the BTBR vehicle group as the reference. For IL-6, the overall group effect was significant (F (2,12) = 21.90, p < 0.0001).
Dunnett-adjusted comparisons versus the BTBR vehicle group were: C57BL/6J, adjusted p < 0.0001, and LOV 40 mg/kg, adjusted p = 0.0006. For TNF-α, the overall group effect was significant (F (2,12) = 36.89, p < 0.0001). Dunnett-adjusted comparisons versus the BTBR vehicle group were: C57BL/6J, adjusted p < 0.0001, and LOV 40 mg/kg, adjusted p = 0.0001 (****p < 0.0001 and ***p < 0.001 Dunnett-adjusted comparisons versus BTBR mice treated with vehicle).Two bar graphs showing brain interleukin 6 and tumor necrosis factor alpha levels across three mouse groups.
For TNF-α (Figure 7B), one-way ANOVA similarly revealed a significant overall group effect ((F (2,12) = 36.89), (p < 0.0001)). Vehicle-treated BTBR mice exhibited significantly higher TNF-α levels than C57BL/6J controls (mean difference [C57BL/6J − BTBR]: −168.1, 95% CI: −218.9 to −117.3, adjusted p < 0.0001). The treatment effect was expressed as LOV minus BTBR vehicle.
TNF-α levels were significantly lower following treatment with LOV at 40 mg/kg compared with BTBR vehicle (mean difference: −124.6, 95% CI: −175.4 to −73.76, adjusted p = 0.0001). Thus, the exploratory cytokine analyses showed that 21-day LOV treatment at 40 mg/kg was associated with lower brain IL-6 and TNF-α levels relative to vehicle-treated BTBR mice.
Discussion
Neurobiological studies of ASD indicate disruptions in lipid metabolism, excitatory–inhibitory neurotransmission, and inflammatory and immune pathways.42–44 These abnormalities provide a rationale for exploring statins as adjunctive therapies; however, the precise mechanisms by which statins might influence core symptoms or associated behaviors in ASD remain unclear. Support for statin-based interventions in neurodevelopmental disorders comes from studies in Fragile X syndrome (FXS), a monogenic condition frequently associated with ASD, where LOV has been shown to correct excessive protein synthesis, normalize ERK1/2 signalling, and improve behavioural phenotypes in the Fmr1−/y mouse model.24,25
Statins inhibit the synthesis of cholesterol and downstream isoprenoids in the mevalonate pathway, and it has been shown that cholesterol metabolism is altered in autism.45,46 Dysregulation of cholesterol metabolism has also been reported in other neurodevelopmental disorders associated with ASD, such as Rett syndrome. In Mecp2 mutant mice, cholesterol homeostasis is significantly altered, with reduced brain cholesterol synthesis and systemic lipid abnormalities. Importantly, pharmacological modulation of cholesterol biosynthesis using statins has been shown to improve metabolic dysfunction, ameliorate behavioral deficits, and increase survival in these models.47 The relationship between cholesterol dysregulation and ASD remains complex.
Available evidence highlights substantial heterogeneity in cholesterol alterations across disorders and patient subsets. Smith–Lemli–Opitz syndrome, associated with a high prevalence of ASD features is characterized by reduced cholesterol levels, whereas Rett syndrome presents with elevated plasma cholesterol and alterations in cholesterol metabolism. In addition, studies in ASD cohorts report a subset of patients with broader dyslipidemia.7,47–49
Accumulating evidence indicates the presence of immune and inflammatory alterations in autism.50,51 Statins also exhibit anti-inflammatory and immune-modulatory effects through the reduction of lipid peroxidation, production of reactive oxygen species, while increased oxidative damage and reactive oxygen species are detected in the brains of individuals with autism.52–54 In addition, statins decrease pro-inflammatory cytokines, increase anti-inflammatory cytokines, and inhibit microglial activation.55
In the present study, LOV administration produced robust improvements in behavioural and biochemical abnormalities in BTBR mice, a well-validated idiopathic model of ASD.30 Consistent with earlier reports, BTBR mice displayed hallmark behavioural impairments, including heightened repetitive behaviours, deficits in spontaneous alternation, and disrupted exploratory patterns. These phenotypes were accompanied by clear neuroinflammatory phenotype, reinforcing the relevance of the BTBR strain for investigating ASD-related pathophysiology.37,39,56
In our experiments in BTBR mice, LOV exerted significant behavioural benefits. BTBR mice exhibited markedly increased marble-burying behavior compared to C57BL/6J controls, reflecting enhanced repetitive and stereotyped activity. In the present study, LOV treatment reduced marble-burying behaviour in BTBR mice, with statistically detectable reductions observed at 20 and 40 mg/kg, whereas no statistically detectable effect was observed at 10 mg/kg.
The largest observed mean reduction occurred at 40 mg/kg. These findings suggest that repeated lovastatin treatment attenuates repetitive and compulsive behaviour in BTBR mice. In the three-arm maze, a statistically detectable improvement in spontaneous alternation was observed only at 40 mg/kg, whereas the 10 and 20 mg/kg doses did not differ significantly from the BTBR vehicle group.
Reduced spontaneous alternation in BTBR mice is consistent with impaired behavioral flexibility and working-memory-related performance. Reduced spontaneous alternation is a measure of impaired working memory and increased behavioral rigidity. The observed decrease in marble burying and increased alternation reflects improved behavioral flexibility and reduced compulsive-like behavior, which are core features disrupted in ASD.
Furthermore, in the three-chamber test, LOV treatment improved both sociability and social novelty in BTBR mice. These findings indicate that LOV not only ameliorates repetitive and cognitive deficits but also improves social deficits, which are highly relevant to core ASD phenotypes. Consistent with these findings, reduced repetitive and stereotyped behaviours have also been reported following simvastatin treatment in other rodent models, where simvastatin significantly decreased turning, jumping, and grooming behaviours compared with untreated animals.57 Together, these observations suggest that statins may attenuate repetitive behaviours across models.
The brain contains approximately one quarter of the body’s total cholesterol, underscoring its critical role in neural development and synaptic function.46 Cholesterol is an essential structural component of neuronal membranes and lipid rafts, which regulate neurotransmitter release and synaptic plasticity.19 Disruptions in cholesterol homeostasis may profoundly affect brain development. Additionally, excessive peripheral cholesterol has been associated with neuroinflammation, as demonstrated in hypercholesterolemic mouse models, where increased cholesterol levels triggered inflammatory responses and subsequent synaptic and cognitive dysfunction.58 The BTBR model used in this study exhibits dysregulated cholesterol homeostasis characterised by elevated serum cholesterol and reduced brain cholesterol and supporting its suitability for investigating dyslipidemia-associated ASD mechanisms. Consistent with our findings, previous studies in BTBR mice have demonstrated similar alterations in BTBR mice, reflecting enhanced peripheral cholesterol levels alongside impaired central cholesterol availability.59 Moreover, a clinical study investigating polymorphisms in the HMG-CoA reductase gene reported that the rs12654264-T allele was associated with increased body mass index, triglyceride levels, and fasting glucose in children and adolescents with ASD, implicating altered cholesterol synthesis pathways in ASD-associated metabolic dysfunction.60 Together, these findings suggest that dysregulation of the cholesterol biosynthetic pathway may represent a convergent mechanism underlying both metabolic and neurobiological abnormalities in ASD.
In the present study, LOV significantly reduced the elevated serum cholesterol levels observed in BTBR mice but did not restore brain free cholesterol levels. All doses of LOV treatment produced alterations in brain HMG-CoA reductase levels, which was significantly elevated in BTBR mice, extending previous reports that primarily identified such changes in peripheral tissues.
Evidence suggests that oxidant/antioxidant imbalance and neuroinflammation are implicated in ASD’s pathobiology.61,62 An abnormal cytokine profile, as an increase in levels of neuroinflammatory cytokines in peripheral blood samples of children with ASD, has been reported.63 It has been shown that the level of TNF-α is elevated in the cerebrospinal fluid of children with ASD.64 Accumulating evidence suggests that interventions capable of reducing neuroinflammatory responses may exert therapeutic benefits and potentially alleviate ASD-related symptoms.38,62 While multiple LOV doses were evaluated in the behavioral studies, the 40 mg/kg dose showed the largest observed behavioral effects across the evaluated paradigms and was therefore selected for the subsequent neuroinflammatory analyses. As this dose selection was informed by the behavioral findings, the subsequent cytokine analyses were considered exploratory.
Our findings demonstrate significant increases in the pro-inflammatory cytokines IL-6 and TNF-α in BTBR mice relative to C57BL/6J controls. Exploratory analysis showed that treatment with LOV (40 mg/kg) was associated with significantly lower brain IL-6 and TNF-α levels relative to BTBR vehicle-treated mice. The ability of LOV to attenuate elevated cytokine levels is consistent with previous reports highlighting the anti-inflammatory and neuroprotective properties of statins.65,66 However, the parallel changes in cytokine levels and behavioral outcomes do not establish a causal relationship between these effects.
Cholesterol plays a critical role in the central nervous system, where it is essential for synapse formation, neurotransmission, dendritic development, and synaptic plasticity, largely through its involvement in cholesterol-rich membrane microdomains. Disruption of cholesterol homeostasis has been shown to impair neuronal function and cognition, as evidenced in neurodevelopmental disorders such as Rett syndrome.47
Statins are known to exert pleiotropic actions through modulation of mevalonate pathway, leading to reduced synthesis of isoprenoid intermediates such as farnesyl pyrophosphate and geranylgeranyl pyrophosphate, which are essential for the prenylation of small GTP-binding proteins (Ras, Rho, and Rap).67,68 Small GTPases play a critical role in synaptic plasticity and cognitive function and modulation of these proteins by statins may contribute to their effects on emotional and cognitive processes.45,69 In rats prenatally exposed to valproic acid, Cartocci et al (2018) identified brain-region- and developmental-stage-dependent alterations in cholesterol/isoprenoid homeostasis, including changes in HMG-CoA reductase regulation and prenylation-related signalling involving small GTPases.70 The study also reported alterations in hippocampal oligodendrocytes and myelination, further highlighting the potential importance of lipid homeostasis for neurodevelopmental function. These findings suggest that disruption of cholesterol/isoprenoid homeostasis may extend across distinct preclinical ASD models. Further support for an isoprenoid-dependent mechanism comes from studies in the Fmr1 knockout model of Fragile X syndrome.
By inhibiting the mevalonate pathway, LOV can reduce the availability of intermediates required for protein prenylation and influence Ras-dependent signalling. Previous studies have demonstrated that pharmacological inhibition of Ras or ERK1/2 can normalize excessive protein synthesis and ameliorate electrophysiological abnormalities in Fmr1 knockout mice.24 Moreover, LOV has been reported to reduce Ras–ERK1/2 pathway activity and correct several disease-associated phenotypes in Fmr1 knockout mice, providing experimental evidence that modulation of Ras–ERK1/2 signalling may contribute to its neurobiological effects.24 Thus, modulation of isoprenoid-dependent signaling, Ras prenylation, ERK1/2 activation, and related downstream signaling represents possible mechanism for the central actions of statins; however, its contribution to the behavioral and neuroinflammatory findings observed in the present BTBR study remains to be established.
In addition, statins have been shown to modulate neurotrophic pathways, including CREB activation and upregulation of BDNF, which are essential for neurogenesis, synaptic plasticity, and cognitive function.57,71 In clinical settings, adjunctive treatment with simvastatin has been shown to reduce irritability and hyperactivity in individuals with ASD.21 Furthermore combined cholesterol supplementation and statin therapy in Smith-Lemli-Opitz syndrome, have been associated with improvements in behavioral symptoms72 Additional studies in Fragile X syndrome also suggest beneficial effects of LOV on behavioral and cognitive outcomes.20 The clinical studies in Fragile X patients have reported behavioral improvements following LOV treatment, in combination with minocycline, demonstrating good safety and improved behavioral outcomes, supporting its translational potential.22
Together, these findings provide preclinical evidence that LOV exerts multifaceted therapeutic effects in a validated ASD mouse model. The observed changes across behavioral, cholesterol-related, and inflammatory outcomes support further investigation of LOV in preclinical models of ASD-related phenotypes.
Limitations
An important limitation of the present study is that only male mice were evaluated. Consequently, potential sex-related differences in the observed behavioral and biochemical responses to LOV could not be assessed, and the present findings cannot be generalized to female mice. Future investigations should therefore include both male and female mice to strengthen the translational relevance of these findings.73,74 Second, the findings were obtained using only BTBR mice as ASD animal model, which may limit the generalizability of findings across the heterogeneous biological mechanisms underlying ASD requiring validation in additional ASD models.
Third, molecular characterization was limited, and further studies are required to define the signalling pathways and molecular mechanisms underlying the effects of LOV. In addition, the 21-day treatment regimen does not establish the optimal treatment duration, or long-term safety. The relatively small sample size and absence of an a priori sample-size calculation may also limit statistical power, warranting confirmation in larger cohorts.
Finally, biochemical analyses using whole-brain homogenates did not permit assessment of region-specific effects, warranting further investigation in ASD-relevant regions.
Conclusion
The present study demonstrates that 21-day LOV administration produces significant behavioural and biochemical improvements in BTBR mice, a well-established idiopathic model of ASD. LOV attenuated repetitive behaviours, improved cognitive performance and social interaction, while also reducing elevated peripheral cholesterol levels and decreasing pro-inflammatory cytokine levels. The behavioural improvements occurred without restoration of brain free cholesterol levels, suggesting that the observed effects of LOV may not be solely attributable to cholesterol modulation and may involve additional mechanisms.
However, given the limited molecular characterization in the present study, the specific signalling pathways and mechanisms underlying these effects remain to be established. Collectively, these findings provide preliminary preclinical evidence for the potential beneficial effects of LOV in ASD-like phenotypes and warrant further investigation in additional experimental models, with more comprehensive mechanistic characterization, before its translational relevance can be established.