Inhibition of lysyl oxidase by pharmacological intervention and genetic manipulation alleviates epilepsy-associated cognitive disorder

, has so far unchar-acterized etiological origins. Our prior work revealed that lysyl oxidase (Lox) acted as a novel contributor of ferroptosis, a recently discovered cell death mode in the regulation of brain function. However, the role of Lox-mediated ferroptosis in ECD remains unknown. ECD mouse model was established 2 months later following a single injection of kainic acid (KA) for. After chronic treatment with KA, mice were treated with different doses (30 mg/kg, 100 mg/kg and 300 mg/kg) of Lox inhibitor BAPN. Additionally, hippocampal-specific Lox knockout mice was also constructed and employed to validate the role of Lox in ECD. Cognitive functions were assessed using novel object recognition test (NOR) and Morris water maze test (MWM). Protein expression of phosphorylated cAMP-response element binding (CREB), a well-known molecular marker for evaluation of cognitive performance, was also detected by Western blot. The protein distribution of Lox was analyzed by immunofluorescence. In KA-induced ECD mouse model, ferroptosis process was activated according to upregulation of 4-HNE protein and a previously discovered ferroptosis in our group, namely, Lox was remarkably increased. Pharmacological inhibition of Lox by BAPN at the dose of 100 mg/kg significantly increased the discrimination index following NOR test and decreased escape latency as well as augmented passing times within 60 s following MWM test in ECD mouse model. Additionally, deficiency of Lox in hippocampus also led to pronounced

improvement of deficits in ECD model.These findings indicate that the ferroptosis regulatory factor, Lox, is activated in ECD.Ablation of Lox by either pharmacological intervention or genetic manipulation ameliorates the impairment in ECD mouse model, which suggest that Lox serves as a promising therapeutic target for treating ECD in clinic.

Introduction
Epilepsy-associated cognitive disorder (ECD) is a well-established comorbidity of epilepsy (Hermann et al., 2021).Generally speaking, it consists of six subtypes of deficits as follows: disturbances in the aspect of language, sensorimotor functions, learning and memory, executive functioning, complex attention, and social cognition (Harvey, 2019).Approximately 50% of patients with epilepsy receiving pharmacological interventions suffer from cognitive and memory deterioration (Helmstaedter et al., 2003).Such learning impairments, coupled with diminished concentration and poor memory, have a serious impact on life quality of patients (Moorhouse et al., 2020;Kanner et al., 2020).Antiepileptic drugs (AEDs) are currently the major medication for seizure control in epilepsy.However, there is evidence supporting that several AEDs like topiramate and sodium valproate can exacerbate ECD in clinic (Dusanter et al., 2023).Other satisfactory anti-seizure regiment such as intake of ketogenic diet also exhibits insufficient improvement for ECD (Gumus et al., 2015).Therefore, there is an urgent need for exploration of novel therapeutic target for counteracting ECD.
The etiology of ECD is largely unknown.Early studies have illustrated that glutamate-mediated excitotoxicity, oxidative stress, mitochondrial dysfunctions, and neuroinflammation act as pivotal contributors to ECD (Singh and Kumar, 2015;Abdel-Zaher et al., 2017).The nexus between oxidative stress and cognitive impairments in epilepsy has especially attracted considerable attention.Recently, Stockwell and colleagues have introduced "ferroptosis" as a novel cell death paradigm intrinsically linked to oxidative stress and accumulation of iron-dependent lethal lipid peroxides (Dixon et al., 2012).Ferroptosis is implicated in various neurodegenerative diseases, including stroke, traumatic brain injury, epilepsy, Parkinson's disease, Huntington's disease, and Alzheimer's disease (AD) (Zhou et al., 2023;Jia et al., 2023;Liu and Chen, 2022;Lin et al., 2022;Song et al., 2023;Huang, 2023).Our prior work found that in a posttraumatic epilepsy mouse model induced by ferric chloride, treatment with ferroptosis inhibitor ferrostatin-1 (Fer-1) could mitigate seizure activity and associated cognitive deficits (Chen et al., 2022).Moreover, Fer-1 was also shown to reduce neuronal impairment and improve cognitive function in kainic acid (KA)-treated rat epilepsy model (Ye et al., 2019).These findings prompt us to postulate that targeting ferroptosis process may alleviate ECD for patients with epilepsy.
Lysyl oxidase (Lox), a monoamine oxidase that traditionally acts on the extracellular matrix's collagen and elastin proteins to enhance intermolecular crosslinks (Mizikova et al., 2017), is a novel ferroptosis factor which was found to trigger neuronal injury and exacerbate seizure activity in our prior work (Mao et al., 2022).However, whether Lox is involved in ECD remains unknown.Therefore, our current work aims to explore role of Lox in ECD mouse model induced by intrahippocampal injection of KA via two approaches: pharmacological inhibition of Lox by a specific inhibitor, namely, β-aminopropionitrile (BAPN), or genetic ablation of Lox gene in mice hippocampus.

Animals and ethic statement
Male C57BL/6 J mice (6-8 weeks, body weight at 18-22 g) were provided by the Experimental Animal Center of Central South University.Other several mouse lines were used in our present study.Camk2a-Cre mice (Cre-recombinase under the control of Camk2a, a previously well-established hippocampus-specific promoter (Tsien et al., 1996), stock no: 005359, The Jackson Laboratory) were crossed with mice carrying a loxP-flanked Lox gene (provided by Biocytogen medical technology Co., LTD) to generate Lox flox/flox ; Camk2a Cre/-mice, a conditional deletion of Lox gene in hippocampus (Lox CKO for short in the following expression).All the animals were maintained under a pathogen-free environment (12 h light/12 h dark cycle, 24 ± 2 • C, 60% humidity) with free access to water and food.Genotypes of mice were conducted by polymerase chain reaction (PCR) using the following primer sequences: (forward) 5'-CCATCCCAGAGCACACAGTTTGG-3'; (reverse) 5'-ATACTTGACATCTACCA AACTGAGGGC-3'.Prior to the initiation of animal experiments, mice acclimated to environment for 3-5 d.The number of animals used in each experiment was described in figure captions.All experimental procedures were in strict accordance with the guidelines set by the Animal Care and Use Committee of Xiangya Hospital of Central South University (protocol number: 2022-0012).

Preparation of ECD mouse model
After anesthetization with sodium phenobarbital (50 mg/kg, i.p.), mice were placed on a stereotactic apparatus.A single dose of 0.5 μL of 500 ng/μL KA (250 ng, dissolved in saline) (Jia et al., 2020) was conducted via intrahippocampal injection using a microsyringe pump (R452, RWD Life Science, Shenzhen, China) at a flow rate of 4.6 nL/6 s according to the following coordinates of mouse: AP-2.0 mm, ML-1.3 mm, V-1.2 mm.The needle stayed in place for another 10 min after completion of KA injection to avoid reflux.Mice in the control group underwent injection of the equal volume of saline instead of KA.Seizure score was evaluated by Racine scale (Racine, 1972).Mice that displayed stage 4 or 5 seizures were selected for the subsequent experiment.ECD was successfully established 2 months later following a single injection of KA.
Experiment 1: Exploration of the effect of pharmacological inhibition of Lox by BAPN (dissolved in saline with dose of 30 mg/kg, 100 mg/kg or 300 mg/kg) on ECD in mice.This experiment was categorized into five groups, namely, sham control group, ECD group, ECD+30 mg/kg BAPN, ECD+100 mg/kg BAPN and ECD+300 mg/kg BAPN.In this experiment, different doses of BAPN were injected intraperitoneally for 14 consecutive days after ECD.The dosage of BAPN was selected according to the previous studies (Liu et al., 2016;Ikenaga et al., 2013) which showed that 100 mg/kg of BAPN was beneficial for disease improvement and the dose gradients were set according to other investigation (Munley et al., 2001).After BAPN treatment, its effect on deficit in ECD was carried out via novel object recognition test (NOR) and the Morris water maze test (MWM) tests.
Experiment 2: Investigation of the effect of hippocampus-specific deletion of Lox on ECD in mice.This experiment was divided into four groups as follows: wild type (WT) group, WT+KA group, Lox CKO and Lox CKO+KA.Two months later following a single injection of KA, NOR and MWM tests were carried out to assess the effect of Lox deficiency in hippocampus on ECD.

Immunofluorescence
As previously described (Yang et al., 2022), the whole brains from each group were carefully dissected and then fixed in paraformaldehyde after transcardial perfusion with PBS.Then, brain slices (8 µm thickness) underwent deparaffinization, followed by antigen retrieval in sodium citrate solution at 96 • C for 20 min.Afterward, they were cooled at room temperature and washed with PBS three times, with each wash lasting for 3 min.Membrane permeabilization was performed via incubation with 0.5% Triton X-100 slowly in a shaker at the speed of 50 rpm at room temperature.After several washes with PBS, the mice brain slices were blocked with 10% donkey serum for nearly 30 min, followed by the incubation of the primary antibody including Lox (1:50 dilution, ab174316, Abcam) overnight at 4 • C and then probed with the secondary antibody (1:200 dilution, A21206, Thermo Fisher Scientific) for 1 h at room temperature.Fluorescent images from each group were captured by an AXIOPHOT Zeiss microscope and the results were analyzed using the Leica Application Suite 4.9.0 software.

NOR test
Two months later following a single injection of KA, NOR test was conducted to evaluate the effects of Lox inhibition on the ability to recognize a novel object as delineated in our previous investigation (Chen et al., 2022).Firstly, in the habituation phase, mice were familiarized within a polyvinyl chloride chamber (45 cm × 45 cm × 45 cm).Subsequently, mice were allowed to explore freely across two identical objects for 10 min.After the completion of the training session, one object was replaced with a novel object, distinct in both color and shape, and animals from each group were explored for 10 min (Colasante et al., 2020).Before each trial, both the compartment and objects were cleaned clearly using 75% ethanol solution.The moving traces for each mouse were monitored by a video tracking system.The time spent on exploration of the novel object (N) and the familiar object (F) was recorded respectively.The discrimination index was calculated using the following formula: (N-F)/(N+F)×100% (Ye et al., 2019).

MWM test
The effects of Lox inhibition on the spatial memory performance were carried out as previously described (Chen et al., 2022).In brief, the water maze pool (120 cm diameter) was filled with water (temperature at 24 ± 2 • C) accompanied with a platform (10 cm in diameter).In the training trials, mice were allowed to search for the hidden platform (1 cm below the water surface), which was placed in the center of one quadrant for 5 consecutive days.Following these trials, the escape latency (s), the mean path length (cm) and time in the target quadrant (s) were recorded.Swimming speed was calculated by dividing path length by escape latency.In the probe trial, the platform was removed 24 h after the training trial.Passing times, namely, the number of times of crossing the platform, were monitored within 60 s (Berlin et al., 2020;Segev et al., 2020).

Statistical analysis
All data were presented as mean ± standard deviation.Results were analyzed using GraphPad Prism 7.0 software.Data normality was assessed via the Shapiro-Wilk test.For data adhering to a normal distribution between two groups, the Student's t-test was employed, whereas non-normally distributed data were analyzed using the Mann Whitney test.One-way ANOVA assessed statistical differences among multiple groups.Subsequent post hoc tests included Tukey's test for normally distributed data and Dunn's analysis for non-normally distributed data.The Kruskal-Wallis test was additionally utilized.Escape latencies from the water maze were evaluated through two-way ANOVA followed by Tukey's post hoc test.A significance threshold was set at P < 0.05.Participants who were responsible for data analysis were blind to the experimental design.The results of statistical analyses were summarized in Table 1.

Ferroptosis process was activated and the protein expression of Lox was upregulated in KA-induced ECD mouse model
Two months after a single injection of KA, NOR and MWM tests were chosen since they served as well-recognized methods for assessing cognitive function.The results of NOR test demonstrated that mice subject to KA treatment exhibited no significant difference in the aspect of familiar and novel objects while longer time was spent on exploration of novel object than of familiar object in control group (Fig. 1A), indicating the existence of cognitive deficit in mice following the chronic treatment with KA.In the meantime, cognitive function was further confirmed by MWM test.There was no difference in the swimming speed of mice in all groups (Fig. 1B).However, KA triggered the increase of escape latency (Fig. 1C).Representative swimming traces regarding probe trial on the sixth day in ECD and matched control groups were displayed in Fig. 1D.Quantitative analysis illustrated that there was a significant decrease of passing times within 60 s in ECD group (Fig. 1E).From the molecular aspect, the phosphorylation of CREB (p-CREB), a well-known marker for evaluation of cognition, was also found to decrease in ECD mice (Fig. 1F and G).Altogether, these results indicate that KA-induced ECD mouse model is successfully established in our present study.
In order to investigate whether ferroptosis was activated and Lox, a novel neuronal ferroptosis regulator, was altered in ECD, the key ferroptotic indices including 4-HNE and COX-2 and the protein expression of Lox were detected via Western blot assay.Significantly, it was demonstrated that 4-HNE and COX-2 were increased in ECD mice model (Fig. 2A-C).A pronounced increase in Lox protein expression was also observed in ECD model (Fig. 2D and E).These data suggest that ferroptosis is activated and the novel ferroptosis regulator Lox is elevated in KA-induced ECD mice model.

Pharmacological inhibition of Lox by BAPN mitigated cognitive dysfunctions in ECD model
Since Lox is changed in ECD, effect of Lox inhibition on this condition was further investigated via pharmacological intervention with BAPN, an irreversible Lox inhibitor.In the NOR test, when treatment with BAPN by the dose of 30 mg/kg or 100 mg/kg in KA-induced ECD mice, the time spent in exploration of novel object was longer than that in familiar object, revealing improvement of BAPN on cognitive function in ECD mice (Fig. 3A).Fig. 3B depicts a notable enhancement in the discrimination index of the KA-induced ECD mice following treatment with 100 mg/kg BAPN.Additionally, MWM test was conducted to further assess cognitive performance.There was no significance in the aspect of swimming speed of all mice (Fig. 3C).Nevertheless, escape latency was remarkably reduced in ECD mice treated with different doses of BAPN (100 mg/kg and 300 mg/kg) compared with KA-induced groups (Fig. 3D).The representative swimming traces provide a clear visualization of these behavioral differences across various groups (Fig. 3E).Specifically, in the context of the KA-induced ECD model, an increased frequency of crossing the platform were noted following BAPN treatment by the dose of 100 mg/kg (Fig. 3F), suggesting the satisfactory recovery of spatial learning and memory in ECD after intervention with BAPN.Additionally, phosphorylation of CREB was also found to increase in ECD after BAPN treatment (Fig. 4A and B).Taken together, our findings support that pharmacological inhibition of Lox by BAPN can provide the relief of cognitive deficits in ECD mouse model.

Generation of Lox CKO mice
Upon ascertaining the ameliorative effects of Lox inhibition in the  significant reduction of Lox fluorescence intensity in the hippocampal CA1, CA3 and DG subregions (Fig. 5C).In addition, Western blot was also employed to detect the protein expression of Lox in Lox CKO mice and littermate control group.It was found that there was an obvious decrease of Lox protein expression in Lox CKO mice (Fig. 5D and E).In summary, these results indicate that Lox CKO mice is successfully generated, which can be utilized for the subsequent experiment.

Hippocampal-specific ablation of Lox gene ameliorated ECD
We further explored whether genetic deletion of Lox in hippocampus could attenuate cognitive disturbances following ECD.In the NOR test, wild-type mice treated with KA resulted in no significant difference of exploration time between novel and familiar objects (Fig. 6A).However, this index was remarkably reversed in Lox CKO mice post KA injection.Further analysis of discrimination ratio also illustrated that Lox CKO mice exhibited better discrimination capacities compared to wild-type mice treated with KA (Fig. 6B).In the MWM test, no alteration in swimming speed between the Lox CKO and wild-type mice were observed (Fig. 6C).Nevertheless, the Lox CKO mice post KA injection had the manifestation of lower escape latency (Fig. 6D and E) and longer time in the target quadrant (Fig. 6F), compared with KA-treated wildtype mice.Following the probe trial, there was evident increase of passing times within 60 s in KA-treated Lox CKO mice (Fig. 6G and H).Collectively, these data suggest that hippocampal-specific genetic ablation of Lox alleviates deficits especially increases of discrimination capacity, time in target quadrant and passing times within 60 s and decrease of escape latency in ECD.

Discussion
The major findings of our present study were shown as follows: (1) Ferroptosis was activated and Lox was upregulated in the ECD mouse model; (2) Pharmacological inhibition of Lox mitigated ECD, evidenced by an increased discrimination index in NOR test, decreased escape latency and increased frequency of crossing the original platform in MWM tests; (3) Deficiency of Lox via CRISPR-Cas9 method in the hippocampus ameliorated ECD.Our findings disclose the contributory role of the ferroptosis regulatory factor Lox in ECD and these data indicate that Lox may serve as a promising therapeutic target for ECD.
Initially, the current study revealed the presence of ferroptosis in ECD.Previous research has reported that ferroptosis is involved in the cognitive impairment of KA-induced temporal lobe epilepsy in rats (Ye et al., 2019), which is partly consistent with our present work.Excessive iron (a key factor inducing ferroptosis) is also observed around insoluble amyloid plaques and neurofibrillary tangles, hallmark features of AD (Roberts et al., 2012), which suggests that ferroptosis-associated signaling acts as a critical event for this disease.Further evidence from early studies have shown that iron-associated proteins such as ferritin-light chain (FTL) and ferritin-heavy chain (FTH) as well as 4-HNE, a byproduct of lipid peroxidation, which are well-recognized ferroptosis factors, are elevated in AD patients (Ashraf et al., 2020).These findings support the notion that ferroptosis process is central to the regulation of cognitive function.Consistently, several other investigations have also indicated that mitigating ferroptosis exerts beneficial effect on the cognitive dysfunction in AD (Bao et al., 2021;Zhai et al., 2023).Collectively, these findings support the notion that ferroptosis may be crucial for the etiology of ECD and targeting ferroptosis process may possess the curative effect on ECD.Furthermore, results from our current work illustrated that Lox-induced ferroptosis was involved in ECD.In fact, previous studies have revealed activation of Lox is positively correlated with AD (Maheshwari, 2023;Gilad et al., 2005) and spinal cord injury (Chen et al., 2022;Gilad and Gilad, 2001).The increase of nuclear Lox was also found to lead to dysfunction of Purkinje cell dendrites (Li et al., 2010).What's more important, Lox was also demonstrated to contribute to accumulation of endothelial oxidative stress (a critical ferroptosis-inducing factor) in a rat model of spontaneously hypertension (Martinez-Revelles et al., 2017) and exacerbate neuronal ferroptosis in KA seizure mouse model (Mao et al., 2022).These investigations again ascertain the key role of Lox in the regulation of brain dysfunction and ferroptosis process.
Our further work showed that blockade of Lox by its specific inhibitor BAPN improved the deficit in a mouse model of ECD.BAPN was previously reported to inhibit Lox and subsequently alleviate neuronal damage in the KA-treated mouse epilepsy model (Mao et al., 2022).In fact, BAPN was the first identified Lox inhibitor (Pinnell and Martin, 1968), which functions as a potent and irreversible Lox inhibitor by covalently binding to the enzymatic domain (Tang et al., 1983;Barry-Hamilton et al., 2010).The inhibitory effect of BAPN on Lox has been extensively evaluated across diverse disease models (Arem et al., 1979;Zhao et al., 2019;Rachman-Tzemah et al., 2017).Our previous research has shown that pharmacological inhibition of Lox by BAPN can significantly block the iron accumulation in epilepsy model, thereby reducing neuronal impairment (Mao et al., 2022).Treatment with 100 mg/kg BAPN is found to facilitate the functional recovery of sensorimotor and locomotor functions in a mouse model of spinal cord injury (Gilad and Gilad, 2001).Additionally, there is also evidence demonstrating that treatment with BAPN remarkably decreases the level of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, a key enzyme for reactive oxygen species (ROS) generation (Alam et al., 2023), in the arteries of spontaneously hypertensive rats (Martinez-Revelles et al., 2017).Since oxidative stress characterized with excessive production of ROS is a key promoter of ferroptosis, these data indirectly suggest the involvement of ferroptosis in BAPN's effect.Altogether, results from our current work give the hint that ferroptosis is a possible mechanism for the beneficial effect of BAPN on ECD.It is worth mentioning that detection of p-CREB protein level was also carried out in our present work to assess the effect of BAPN on cognitive function in epilepsy mouse model apart from behavioral tests.CREB functions as a transcriptional factor which is localized within the nucleus and binds to the DNA cAMP response element (Yu et al., 2017).The phosphorylated form of CREB can often result in its transcriptional activation and subsequently stimulate the transcription of multiple signaling molecules involved in synaptic plasticity required for memory formation (Scott Bitner, 2012).Therefore, p-CREB acts as a memory switch and in this regard the elevation of its level can reflect the improvement of cognitive function.Our current work revealed that BAPN treatment obviously increased the level of p-CREB in ECD, suggesting its benefit on cognition during epilepsy.
Due to the potential limitations via pharmacological intervention which include unpredictably affect a certain target in a variety of tissue organs leading to undesirable side effects, the role of Lox in ECD was further explored with genetic ablation of Lox in specific ECD-vulnerable brain region hippocampus.Under the control of hippocampus-specific  promoter Cre-recombinase (Tsien et al., 1996), Lox was successfully knockdown in our present study.The reason why Lox protein was not completely deleted in hippocampus tissue sample is possibly attributable to primarily influence CA1 subregion (Tsien et al., 1996) with no evident effect on other hippocampal regions such as CA3 and DG.Anyway, hippocampus-specific downregulation of Lox was achieved and this phenomenon can effectively alleviated ECD.
In summary, the current results reveal that the ferroptosis regulatory factor, Lox, is activated in ECD.Furthermore, by inhibiting Lox through pharmacological intervention and genetic manipulation, impairments of ECD, including manifestation in memory and cognitive functions, can be alleviated.It is believed that Lox may emerge as a promising therapeutic target for ECD treatment.Further investigation is warranted to probe the molecular mechanism by which Lox regulates the pathogenesis of ECD.Anyway, our findings showing the critical role of Lox in ECD undoubtedly hold promise for providing the novel therapeutic target for ECD.

Ethical publication statement
We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Fig. 1 .
Fig. 1.Successful assessment of ECD mouse model induced by intrahippocampal injection of KA. (A) Quantitative analysis of exploration time of NOR test in ECD and control groups (n=6 for each group).N represents the novel object, and F represents the familiar object; (B-E) Analysis of indices for evaluation of spatial learning and memory including swimming speed, escape latency and passing times within 60 s in MWM test (n=8 for each group); (F) Representative protein bands of phosphorylated CREB (p-CREB) and total CREB (CREB) in hippocampus of ECD and control groups; (G) Statistical analysis of protein expression of p-CREB/CREB (n=3 for each group).**P < 0.01, ***P < 0.001.

Fig. 2 .
Fig. 2. Ferroptosis is activated and the protein expression of Lox is upregulated in ECD mouse model.(A-C) Analysis of ferroptosis-associated indices including 4-HNE and COX-2 in hippocampus of ECD and control groups (n=3 for each group); (D) Representative protein band of Lox in hippocampus of ECD and control groups; (E) Statistical analysis of Lox protein level in ECD and control groups (n=3 for each group).*P<0.05.

Fig. 3 .
Fig. 3. Pharmacological inhibition of Lox by BAPN alleviates deficits in ECD mouse model.(A-B) Quantitative analysis of effects of BAPN on exploration time and discrimination ratio in ECD mice (n=5 for each group); (C-F) Illustrations of indices related to spatial learning and memory including swimming speed, escape latency and passing times within 60 s in MWM test (n=8 for each group).*P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4 .
Fig. 4. Pharmacological inhibition of Lox by BAPN remarkably decreases the protein expression of phosphorylated CREB in ECD mouse model.(A) Representative protein bands of p-CREB in different groups; (B) Quantitative analysis of effect of BAPN on p-CREB in hippocampus of ECD mice.*P < 0.05, **P < 0.01.

Fig. 5 .
Fig. 5. Successful generation of mouse model with hippocampus-specific knockout of Lox gene.(A) Schematic diagram showing the details for the construction of hippocampus-specific knockout of Lox gene, namely, Lox flox/flox,Camk2a-Cre (Lox CKO for short); (B) The Lox CKO mice exhibited a pair alleles with 357 bp fragment for Lox flox/flox and a 400 bp fragment for Cre under the control of hippocampus-specific promoter Camk2a.The abbreviations of WT and NTC are Wild type and non-template control, respectively; (C) Deletion of protein distribution of Lox in hippocampal CA1, CA3 and DG regions was validated via immunofluorescence.Scale bar: 100 μm; (D) Representative protein bands showing the decrease of Lox protein in hippocampus; (E) Quantitative analysis of protein expression in Lox CKO and matched control groups.**P < 0.01.

Fig. 6 .
Fig. 6.Deficiency of Lox in hippocampus ameliorates deficits in ECD.(A-B) Effects of Lox CKO on exploration time and discrimination index during NOR test in ECD mouse model (n=13 for each group); (C-H) Effects of Lox CKO on indices linked with spatial learning and memory during MWM test including swimming speed, escape latency, time in the target quadrant and passing times within 60 s in ECD mouse model (n=8 for each group).*P < 0.05, **P < 0.01, ***P < 0.001.

Table 1
Summary of statistical tests.