Targeting of Itch by clomipramine or gene therapy improves cognitive defects related to Alzheimer’s disease

TL;DR

This study shows that targeting the Itch protein with gene therapy or the drug clomipramine can improve cognitive issues in a mouse model of Alzheimer’s disease.

Contribution

The study introduces Itch as a novel therapeutic target for Alzheimer’s disease using gene therapy and drug repurposing.

Findings

Loss-of-function Itch mutants reversed cognitive defects in AD mice.

Clomipramine inhibited Itch and prevented neuronal apoptosis in AD mouse brains.

Itch inhibition improved learning and memory in AD models.

Abstract

We propose a therapeutic strategy against Alzheimer’s disease (AD), which involves targeting E3 ubiquitin ligase AIP4 or Itch. Previous studies have shown that Itch is aberrantly activated in cortical neurons of a mouse model of AD and contributes to neuronal death. We used a two-pronged approach to target Itch in a mouse model for AD: (1) adeno-associated virus (AAV) expressing loss-of-function mutants of Itch and (2) clomipramine, a tricyclic antidepressant, which is an Itch inhibitor. Both treatments significantly improved learning and memory associated with AD mice. A reversal in neuronal apoptosis was observed in AD mouse brain, which explained the improvement in cognition. Clomipramine was able to inhibit Itch in neurons and prevent their apoptosis in response to Aβ42. Given that clomipramine is being used against psychiatric disorders, this drug may be repurposed for use against AD.

•Ubiquitin ligase Itch is aberrantly activated in response to Aβ42 in the AD mouse model•Loss-of-function mutants of Itch reversed cognitive defects in the AD mouse model•Clomipramine improved cognition and prevented apoptosis in AD mouse brain•Targeting Itch is a promising therapeutic pathway for Alzheimer’s disease

Ubiquitin ligase Itch is aberrantly activated in response to Aβ42 in the AD mouse model

Loss-of-function mutants of Itch reversed cognitive defects in the AD mouse model

Clomipramine improved cognition and prevented apoptosis in AD mouse brain

Targeting Itch is a promising therapeutic pathway for Alzheimer’s disease

Neuroscience; Molecular neuroscience

Figures

AAV-mediated administration of Itch mutants prevents cognitive deficits in TgAD mice(A) Schematic showing Itch domain architecture. It has a C2 domain, proline rich domain, four WW repeats, and a HECT domain. Sites which were mutated in the present studies are indicated: T222 and S232 are aberrantly phosphorylated by JNK, and K393 undergoes autoubiquitination in response to Aβ_42_.[13](#bib13)(B) Schematic illustrating the regimen followed for AAV stereotaxic injections and behavioral tests. 6-month-old female WT and TgAD mice were injected stereotaxically in the frontal cortex with indicated

Itch mutants prevent neuronal apoptosis in TgAD mice(A) Three mice from each group of mice mentioned in [Figure 1](#fig1) were perfused, coronal brain sections were prepared and subjected to TUNEL staining and immunofluorescence assay (IFA) using anti-PCNA, anti-NeuN antibodies, and nuclei were stained with DAPI. Representative images obtained for the frontal cortex are provided.(B) The percentage of neurons (NeuN+) stained for TUNEL and PCNA (TUNEL^+^/PCNA^+^) was counted (A) from a random sampling of the cortex region in five images from three sections/mouse (mean ± SEM, ∗∗∗∗*p* < 0.0001, ns

Clomipramine inhibits Itch and prevents neuronal apoptosis and cell cycle re-entry in neurons from TgAD mice(A) Chemical structure of clomipramine. The chloride moiety is proposed to interact with catalytic cysteine in the HECT domain of Itch.[19](#bib19)(B) Rat cortical neurons were treated with Aβ_42_ and/or clomipramine (75 nM) for 48 h followed by 12 h treatment with the proteasome inhibitor MG132. TAp73 was immunoprecipitated, and western blotting was performed on TAp73-IP with anti-ubiquitin and anti-Itch antibodies. Total protein lysates were also used for western blotting with indicate

Clomipramine improves learning and memory of TgAD mice(A) Schematic showing clomipramine administration regimen and behavioral studies performed on WT or TgAD animals. 6-month-old female WT or TgAD mice received intraperitoneal injections of clomipramine or saline (vehicle). After 45 days of treatment, cognitive function of mice were assessed by MWM, Y-maze, and 8-arm radial maze tests.(B) MWM test was performed by using hidden platform to assess mean latency to reach the platform of 4 groups of mice for WT, TgAD, TgAD-clomipramine (Tg-clo), and WT-clomipramine (WT-clo). Clomipramine-injected

Clomipramine prevents neuronal apoptosis in the cortex and hippocampus of TgAD mice(A) Mice were sacrificed after completion of behavioral tests described in [Figure 4](#fig4). The cortex region was dissected to prepare protein lysates, which were subjected to western blotting with indicated antibodies. Representative blot from two biological replicates is provided in the figure. Bottom: fold change in the levels of PCNA and cl_caspase3 was determined by densitometry of corresponding bands with respect to β-actin (mean ± SEM, ns, not significant, ∗*p* < 0.05, ∗∗∗∗*p* < 0.0001, one-way ANOVA, T

Full text

Introduction

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that results in loss of memory and decline in cognition. The two major pathological hallmarks of AD are β-amyloid (Aβ) plaques and neurofibrillary tangles (NFTs) accompanied by a massive loss of neurons. Only 1%–2% of AD cases are associated with genetic mutations, which are mainly found in amyloid precursor protein (APP) and presenilin (PSEN1/2).1 Currently, the main therapeutic interventions against AD are monoclonal antibody-based approaches that target the clearance of Aβ plaques in patients with AD.2^,^3 While the results of clinical trials indicate that these treatments provide some relief to patients and reverse plaque formation, some reports suggest that they may cause side effects.4^,^5 There are several other targets that are being pursued to develop anti-AD drugs, but these efforts have not yielded much success in the clinic.6

Despite extensive research, current therapeutic strategies have achieved only modest clinical benefits, emphasizing the need for novel mechanistic insights and therapeutic targets. Clearly, it is important to develop novel approaches to tackle AD that target processes other than the ones dependent on Aβ aggregation. For instance, aberrant activation of the cell cycle in neurons, which is accompanied with neuronal apoptosis, has been reported in AD mouse models as well as in patients with AD.7^,^8^,^9 Cell cycle-related neuronal apoptosis (CRNA) is indicated by increased expression of cell cycle proteins and DNA synthesis in neurons undergoing apoptosis,8^,^10^,^11^,^12^,^13^,^14 which suggests crosstalk between the cell cycle machinery and neuronal signaling pathways. Therefore, prevention of CRNA can prove useful in preventing neurodegeneration.

The major functions of the ubiquitin ligase atrophin-1-interacting protein 4 (AIP4, also known as Itch) have mainly been demonstrated in chronic inflammation and immunological functions such as T cell responses.15 Mice with inversion in the Itch locus exhibited excessive scratching and hyperinflammation.16 The lack of Itch results in autoimmune diseases and chronic production of tumorigenic cytokines and exhibits aberrant tumor formation.17^,^18 While Itch has not been implicated directly in neuronal functions, it was found to be a regulator of neuronal apoptosis in Aβ_42_-treated neurons or neurons from an AD transgenic mouse model (TgAD), which it induces by promoting aberrant cell cycle re-entry of TgAD mouse neurons.13 Itch is aberrantly activated as a result of hyperactivation of the JNK pathway, which promotes hyperphosphorylation at T222/S232, leading to Itch autoubiquitination at K393 position. These post-translational modifications of Itch facilitate its interaction with TAp73, resulting in its degradation. These events lead to Itch-mediated CRNA, and its phosphorylation (T222A/S232A) and autoubiquitination (K393R) site mutants were neuroprotective.13 These results strongly indicated Itch to be a promising target for preventing neurodegeneration in AD. To assess the therapeutic potential of targeting Itch, we examined the functional relevance of Itch in a transgenic mouse model APP(Swe)/PS1(dE9) (TgAD), which is commonly used for AD-related studies as it exhibits several features related to AD. Two independent approaches were used for this purpose: gene therapy via adeno-associated virus (AAV)-mediated delivery of loss-of-function Itch mutants into TgAD mice and pharmacological inhibition using clomipramine, an inhibitor of Itch.19

In a screen for identification of Itch inhibitors, clomipramine was identified as a promising candidate.19 The fact that clomipramine, which belongs to the tricyclic antidepressant family, is already being used to treat anxiety, phobia, depression, and obsessive-compulsive disorder (OCD)20^,^21^,^22 presented a possibility of repurposing it against AD. Strikingly, both AAV-Itch mutants and clomipramine administration in TgAD mice caused a significant improvement in learning and memory in TgAD mice. Furthermore, there was a significant decrease in neuronal death in clomipramine-treated TgAD mice.

These findings collectively suggested that aberrant Itch activation plays a role in AD-associated cognitive deficits and neuronal death, and targeting of Itch represents a promising strategy to tackle AD.

Results

Itch loss-of-function mutants rescue cognitive deficits exhibited by TgAD mice

As mentioned earlier, aberrant activation of Itch in response to Aβ_42_ results in neuronal apoptosis.13 Aβ_42_-mediated hyperactivation of the JNK signaling pathway results in aberrant phosphorylation of Itch at T222/S232, which in turn promotes auto-ubiquitination of Itch at K393. As a result, Itch is aberrantly activated, which is reflected by enhanced degradation of its target TAp73. Importantly, the overexpression of phosphodeficient (T222A/S232A) and ubiquitination-defective (K393R) mutants of Itch reversed cell cycle re-entry and neuronal apoptosis in cortical neurons from TgAD mice.13 Given these findings, it was pertinent to evaluate if Itch can be targeted for preventing at least some of the AD-related symptoms. Therefore, we first investigated the effect of overexpressing T222A/S232A and K393R mutants of Itch on learning and memory in the APP(Swe)/PS1(dE9) mouse model of AD (TgAD), which recapitulates several features of AD and is one of the prominent models for screening anti-AD drugs and therapeutics.23^,^24 AAV-mediated delivery of genes to the brain is effective and has a desirable safety profile, strong neurotropism, and efficient and sustained expression of the desired gene in brain tissue.25^,^26^,^27 Therefore, AAV9-based viruses were generated to overexpress wild-type (WT) Itch or its T222A/S232A (TS/A) and K393R (K/R) (Figure 1A) mutants in WT or TgAD mouse brain and were stereotaxically injected into the prefrontal cortex of WT and TgAD mice at 6 months of age (Figure 1B). AAV-tdTomato (control AAV) was injected as a control at the same dose followed by behavior tests after 45 days (Figure 1B). AAVs were administered in 6-month-old TgAD mice as previous studies suggested that these animals exhibit defects in learning and memory from this age.14^,^28 Stereotaxic injections were performed in the frontal cortex due to the following reasons: (1) previous studies established a role of Itch in CRNA in cortical neurons from TgAD mice13 as cortical neurons undergo CRNA in response to Aβ_42_; (2) several AD mouse models have reported aberrant cell cycle re-entry of neurons in the cortex10^,^29; (3) the loss of neurons in AD has been previously reported in both hippocampus and the cortex30^,^31^,^32 and has been implicated in AD-related cognitive defects; and (4) one of our recent studies has demonstrated that the overexpression of miR-449a, which prevents CRNA, in the cortex reversed cognitive defects in TgAD mice.14 After 45 days of AAV stereotaxic delivery, Morris water maze (MWM) and Y-maze tests were performed to evaluate the effect of Itch mutants (Figure 1B) on spatial learning and, working and reference memory.33^,^34^,^35^,^36 Uninjected or ctrl-AAV-injected TgAD mice showed a significant cognitive decline observed with MWM analysis in comparison to WT counterparts (Figures 1C–1F). These tests revealed that both uninjected TgAD and ctrl-AAV-injected TgAD mice exhibited a significant increase in latency to reach the hidden platform at the end of training (Figures 1C and 1D, WT vs. Tg and WT+ctrl-AAV vs. Tg+ctrl-AAV), which indicated that ctrl-AAV does not have an impact on these animals. While WT-Itch AAV administration did not cause an improvement in latency in comparison to ctrl-AAV (Tg+ctrl-AAV vs. Tg+Itch AAV), both Itch mutants (Tg+AAV-TS/A or Tg+AAV-K/R v/s Tg+ctrl AAV) significantly decreased the latency, which was almost comparable to the WT mice (Figures 1D and 1E).Figure 1AAV-mediated administration of Itch mutants prevents cognitive deficits in TgAD mice(A) Schematic showing Itch domain architecture. It has a C2 domain, proline rich domain, four WW repeats, and a HECT domain. Sites which were mutated in the present studies are indicated: T222 and S232 are aberrantly phosphorylated by JNK, and K393 undergoes autoubiquitination in response to Aβ_42_.13(B) Schematic illustrating the regimen followed for AAV stereotaxic injections and behavioral tests. 6-month-old female WT and TgAD mice were injected stereotaxically in the frontal cortex with indicated AAV expressing WT Itch or Itch mutants T222A/S232A (TS/A) or K393R (K/R); a control AAV (Ctrl AAV) lacking Itch or its mutants was also used, and after 45 days, cognitive function were assessed by MWM or Y-maze test.(C and D) Uninjected or mice injected with indicated AAV were subjected to MWM test for 4 days with a hidden probe to assess the mean escape latency to reach the platform at indicated time (C). Escape latency of all the groups of mice in MWM test on day 4 was compared (mean ± SEM. ∗p < 0.05, ∗∗∗p < 0.001, ns, not significant, one-way ANOVA, Sidak’s test, N = 5–10 mice/group) (D).(E) Track plots of a representative from each group of mice in MWM test are described in (C).(F) A probe trial was performed 24 h after the last hidden platform test in the MWM (C). The number of platform area crossings in the target quadrant is shown (mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns, not significant, one-way ANOVA, Sidak’s test, N = 5–10 mice/group).(G) Y-maze test was performed by acclimatizing mice in the maze with one arm closed for 3 min, and then the alternations were recorded for 5 min. The percentage of spontaneous alternation for each mouse was determined (mean ± SEM. ∗p < 0.05, ∗∗∗p < 0.001, ns, not significant, one-way ANOVA, Sidak’s test, N = 5–10 mice/group).

The probe test, which reflects stored memory and is regulated by the frontal cortex,37 indicated a significant decrease in platform area crossings by TgAD mice,23^,^38 which did not change upon injection with ctrl-AAV (Figure 1F, WT + ctrlAAV vs. Tg+ctrl-AAV). The T222A/S232A and K393R Itch mutants (Tg+ctrl-AAV vs. Tg+TS/A-AAV and Tg+K/R-AAV)—but not WT-Itch (Tg+ctrlAAV vs. Tg+Itch-AAV)—significantly increased the crossings when injected in TgAD mice (Figure 1F). There was no significant difference in the swim velocity and food or water intake among all the experimental and control groups (Figure S1). These results suggested marked improvement in spatial learning and memory upon overexpression of Itch mutants. In contrast, the overexpression of WT Itch caused memory and learning defects in the WT mice.

The effect on working memory was assessed by Y-maze test39 (Figure 1B). TgAD mice displayed a significant decrease in spontaneous alternation (Figure 1G, WT/WT+ctrl-AAV vs. Tg/Tg+ctrl-AAV), which is an indicator of working memory. While no significant improvement was observed in TgAD mice upon WT Itch-AAV injection (Tg+ctrlAAV vs. Tg+WT ItchAAV), both mutants of Itch (Tg+ctrl-AAV vs. Tg+TS/A-AAV and Tg+K/R-AAV) mice exhibited a significant increase in alternation in comparison to control AAV-injected mice. There was no significant change in total entry number in any of the conditions (Figure S1B) in the spontaneous alternation test. These results suggested that Itch mutants can improve spatial and working memory, which is compromised in TgAD mice. Since these Itch mutants are defective in Itch activity and prevent neuronal apoptosis in vitro,13 these data establish that aberrant Itch activation contributes to AD-related cognitive defects in the TgAD mice, which can be reversed by these mutants.

Itch mutants reverse neuronal apoptosis in TgAD mice

Given that Itch mutants were able to reverse cognitive defects in TgAD mice, the underlying mechanisms were explored. As mentioned earlier, TAp73 is a key Itch target, and our previous studies suggested that it undergoes degradation in response to Aβ_42_ in TgAD neurons, leading to CRNA.13 Therefore, we evaluated neuronal apoptosis in vivo by performing immunohistochemistry (IHC) in the cortex of TgAD mice. For this purpose, coronal sections of the frontal cortex from TgAD and WT mice were subjected to TUNEL assay to detect apoptotic neurons, which were stained for the neuronal marker NeuN, and staining for PCNA was also performed to assess the status of the cell cycle. IHC confirmed the overexpression of myc-tagged Itch or its mutants in neurons (Figure S2A). It was evident that uninjected or ctrl-AAV-administered TgAD mice had a significantly higher number of neurons that were both TUNEL and PCNA stained (Figure 2), which was consistent with previous studies performed in AD animal models.40^,^41 Strikingly, there was a decrease in the number of total TUNEL^+^ (Figure S2B) as well as TUNEL^+^/PCNA^+^ (Figures 2A and 2B) neurons upon T222A/S232A and K393R-AAV overexpression. These data strongly supported the fact that these mutants of Itch prevented neuronal apoptosis of cortical neurons in TgAD mice. Since the decrease in PCNA^+^ neurons was also observed, it is reasonable to suggest that reduced re-entry of neurons into the cell cycle contributed to this improvement (Figures 2A and 2B). Collectively, these data strongly suggested that the loss-of-function Itch mutants—T222A/S232A and K393R—were able to override its aberrant activation in TgAD mice and prevented neuronal death, which explained the improvement in cognitive functions in these mice (Figures 2A and 2B).Figure 2. Itch mutants prevent neuronal apoptosis in TgAD mice(A) Three mice from each group of mice mentioned in Figure 1 were perfused, coronal brain sections were prepared and subjected to TUNEL staining and immunofluorescence assay (IFA) using anti-PCNA, anti-NeuN antibodies, and nuclei were stained with DAPI. Representative images obtained for the frontal cortex are provided.(B) The percentage of neurons (NeuN+) stained for TUNEL and PCNA (TUNEL^+^/PCNA^+^) was counted (A) from a random sampling of the cortex region in five images from three sections/mouse (mean ± SEM, ∗∗∗∗p < 0.0001, ns, not significant, two-way ANOVA, N = 3). Scale bars, 100 μm.

Clomipramine prevents neuronal apoptosis

The disruption of Itch function by using the two loss-of-function mutants provided significant cognitive improvement for TgAD mice and raised the possibility of targeting Itch in AD. While AAV-mediated gene therapy has increasingly become prominent for the treatment of several disorders including neurodegenerative disorders,42 it has its limitations and needs to be fine-tuned to be successful in the clinic. Therefore, small-molecule inhibitors of Itch are a more viable and cost-effective option for targeting Itch in AD. It was interesting to note that clomipramine, which is a tricyclic antidepressant (TCA) (Figure 3A), emerged as an effective inhibitor of Itch in an inhibitor screen.19 Given that clomipramine is already being used in the clinic against psychiatric disorders like OCD,43 the possibility of repurposing it for targeting Itch in AD was reasonable.Figure 3. Clomipramine inhibits Itch and prevents neuronal apoptosis and cell cycle re-entry in neurons from TgAD mice(A) Chemical structure of clomipramine. The chloride moiety is proposed to interact with catalytic cysteine in the HECT domain of Itch.19(B) Rat cortical neurons were treated with Aβ_42_ and/or clomipramine (75 nM) for 48 h followed by 12 h treatment with the proteasome inhibitor MG132. TAp73 was immunoprecipitated, and western blotting was performed on TAp73-IP with anti-ubiquitin and anti-Itch antibodies. Total protein lysates were also used for western blotting with indicated antibodies. Ubiquitinated TAp73 levels were quantified by densitometry of ubiquitin immunoblot, normalized with respect to the input TAp73, which was also normalized with respect to actin (mean ± SEM, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, one-way ANOVA, Tukey’s test, N = 3).(C) Rat cortical neurons were treated with Aβ_42_ and/or 75 nM clomipramine for 48 h followed by western blotting for indicated proteins. The levels of TAp73, PCNA, and cl_caspase3 were quantified by densitometry, and fold change with respect to untreated (ctrl) neurons was determined (mean ± SEM, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, one-way ANOVA, Tukey’s test, N = 3).(D) Cortical neurons from WT or TgAD mice were treated with 75 nM clomipramine for 48 h followed by western blotting for indicated proteins. TAp73, PCNA, and cl_caspase3 levels were quantified by densitometry, and fold change with respect to untreated WT neurons is provided (mean ± SEM, ∗p < 0.05, ∗∗p < 0.01, one-way ANOVA, Tukey’s test, N = 2).

First, the effect of clomipramine on Aβ_42_-treated cortical neurons was evaluated, and a preliminary experiment was performed to determine the effective concentration of clomipramine that is not toxic to neurons. As an indicator of Itch activity, the levels of TAp73 were assessed as Aβ_42_ promotes ubiquitination of TAp73 by Itch, resulting in its degradation.13 Consistent with previous studies, Aβ_42_ treatment resulted in a loss of TAp73, which was significantly reversed by clomipramine at 75 nM or higher concentration. The levels of cl_caspase3 were inhibited by 10–75 nM clomipramine but increased at higher concentration (Figures S3A and S3B). Therefore, 75 nM clomipramine was used for subsequent in vitro experiments to avoid toxicity. Itch-mediated TAp73 ubiquitination, which is observed upon Aβ_42_ treatment, was significantly reversed by clomipramine (Figure 3B), which confirmed that Itch was inhibited by clomipramine in neurons.

Next, the effect of clomipramine on Aβ_42_-induced neuronal apoptosis was assessed. Clomipramine treatment caused a significant decrease in levels of cleaved caspase-3, which is indicative of apoptosis (Figure 3C). Moreover, PCNA levels, which are elevated upon Aβ_42_ treatment, were also reduced by clomipramine (Figure 3C). Similar results were obtained when experiments were performed with TgAD neurons, as clomipramine reduced both PCNA and cl_caspase3, which are elevated in these cells (Figure 3D). Further, TAp73 levels were significantly restored upon clomipramine treatment, which aligned with its ability to prevent TAp73 ubiquitination (Figure 3B). Collectively, these findings indicated that clomipramine is a potential drug for preventing neuronal cell death of AD neurons, which it may achieve by inhibiting Itch activity. Even though it may block the Itch-mediated TAp73 degradation, it is possible that other Itch targets important for neuronal survival are also protected from degradation by clomipramine.

Clomipramine improves cognition in TgAD mice

Since in vitro studies suggested a reversal of CRNA with clomipramine treatment, it was worth testing if this Itch inhibitor can reverse cognitive defects observed in TgAD mice. For this purpose, ∼6-month-old WT/TgAD mice were intraperitoneally injected every other day with 25 mg/kg body weight of clomipramine for 45 days (Figure 4A). First, MWM test was performed, which revealed that saline-treated TgAD mice took significantly more time than the WT animals to reach the platform (Figures 4B and 4C). Upon administration of clomipramine, TgAD mice showed marked improvement (Figures 4B and 4C), as latency to reach the platform was almost similar to the WT mice on the last day of training (Figures 4B–4D). After 5 days of training, the probe test revealed that TgAD mice exhibited reduced platform area crossings compared to the WT (Figure 4E). In contrast, clomipramine-treated TgAD mice crossed the target quadrant area more frequently, almost similar to WT mice (Figure 4E). These observations suggested improved spatial learning in TgAD mice upon drug treatment.Figure 4. Clomipramine improves learning and memory of TgAD mice(A) Schematic showing clomipramine administration regimen and behavioral studies performed on WT or TgAD animals. 6-month-old female WT or TgAD mice received intraperitoneal injections of clomipramine or saline (vehicle). After 45 days of treatment, cognitive function of mice were assessed by MWM, Y-maze, and 8-arm radial maze tests.(B) MWM test was performed by using hidden platform to assess mean latency to reach the platform of 4 groups of mice for WT, TgAD, TgAD-clomipramine (Tg-clo), and WT-clomipramine (WT-clo). Clomipramine-injected TgAD animals exhibited gradual improvement with time to reach the platform compared to TgAD animals (mean ± SEM [N = 18–19 mice/group]).(C) Latency to reach platform on the final day of training (B) was analyzed (mean ± SEM [N = 18–19 mice/group]; ∗∗p < 0.01, ns, not significant, one-way ANOVA, Tukey’s test).(D) Track plots of a representative from each group of mice in MWM test are described in (B).(E) A probe trial was performed 24 h after the last hidden platform test in the MWM (B). The number of platform area crossings in the target quadrant is shown (mean ± SEM [N = 18–19 mice/group]; ∗p < 0.05; ∗∗p < 0.01, ns, not significant, one-way ANOVA, Tukey’s test).(F) Y-maze test was performed by first acclimatizing mice in the maze with one arm closed for 3 min, and then the alternations were recorded for 5 min. The percentage of spontaneous alternation for each mouse was calculated (mean ± SEM [N = 18–19 mice/group]; ∗p < 0.05; ∗∗p < 0.01, ns, not significant, one-way ANOVA, Tukey’s test). TgAD mice showed decrease in % alternation which was restored in the TgAD-clomipramine group.(G) Schematic depicting the setup of radial arm maze (RAM) test. During RAM test, mice were acclimatized in the maze with food in all arms. The next day, indicated arms were blocked, and food pellets were provided in open arms. On the third day, all arms were left open with food pellets provided in the same arms on second day. Subsequently, the alternations were recorded for each group, and the percentage of error was calculated.(H and I) Short-term (H) or long-term (I) memory errors from RAM test are depicted as percentage of reference memory error (H) or percentage of working memory error (I) (mean ± SEM [N = 18–19 mice/group]; ∗∗p < 0.01, ∗∗∗p < 0.001, ns, not significant, one-way ANOVA, Tukey’s test).

In the Y-maze test, as mentioned earlier (Figure 1G), TgAD mice exhibited reduced alternation in comparison to the WT counterparts. A significant improvement in spontaneous alternation in the clomipramine-treated TgAD mice group was observed (Figure 4F). However, no significant changes were observed in the total number of arm entries made by each group (Figure S4D). These observations indicated a significant improvement in working memory of the TgAD mice upon clomipramine administration.

Further, radial arm maze (RAM) test was performed to assess the long-term as well as short-term memory of mice44^,^45 (Figure 4G). TgAD mice committed more reference (Figure 4H) and working memory (Figure 4I) errors than the WT mice. Strikingly, there was a significant improvement in both these errors in clomipramine-treated TgAD mice. Collectively, these studies indicated that clomipramine significantly improves spatial, working, and reference memory as well as defects in short- and long-term memory in TgAD mice. It was interesting to note that clomipramine does not alter amyloid plaque formation in TgAD mouse brain (Figure S4G), which suggested that it works downstream of Aβ_42_ generation and is independent of plaque formation.

The body weight, average swimming speed, as well as learning and memory abilities between the 2 groups of WT mice treated with clomipramine and saline remained almost unchanged (Figures S4A and S4C). In addition, the average distance traveled and average speed in MWM and RAM test were almost similar within the cohorts (Figures S4B, S4E, and S4F). Based on these observations, it appears that even prolonged clomipramine administration did not cause any major side effects or defects in locomotor activities in mice.

Clomipramine treatment prevented neuronal apoptosis in mice

To determine the effect of clomipramine on apoptosis in the brain of TgAD mice, the animals used in aforementioned behavioral tests were euthanized, and the cortex was dissected and used for western blotting. The levels of cl_caspase3, which were significantly elevated in the cortex of TgAD mice (Figure 5A, lane 3 and 4), were reversed in clomipramine-treated mice (Figure 5A, lane 5 and 6). In addition, increased PCNA expression in the TgAD mouse cortex was also attenuated by clomipramine. It was striking that TAp73, which was almost undetectable in TgAD, was significantly recovered, indicative of Itch inhibition by clomipramine in these mice. However, Itch levels were almost unchanged, suggesting that clomipramine may inhibit Itch activity without altering its expression (Figure 5A). Next, the status of neuronal apoptosis in aforementioned clomipramine-administered mouse brains was assessed by performing TUNEL assays in the prefrontal cortex region. As reported earlier (Figure 2), the TgAD mouse cortex has neurons that exhibit significantly higher TUNEL staining (Figures 5B and S5A), which was largely consistent with previous studies.46^,^47^,^48 In addition, a significant population of TUNEL^+^ cells also exhibited increased PCNA expression. Clomipramine significantly reduced TUNEL^+^ cells (Figure S5A) as well as TUNEL^+^/PCNA^+^ cells (Figures 5B and 5C), which suggested that it was able to reverse neuronal apoptosis and cell cycle re-entry of neurons in TgAD mice.Figure 5. Clomipramine prevents neuronal apoptosis in the cortex and hippocampus of TgAD mice(A) Mice were sacrificed after completion of behavioral tests described in Figure 4. The cortex region was dissected to prepare protein lysates, which were subjected to western blotting with indicated antibodies. Representative blot from two biological replicates is provided in the figure. Bottom: fold change in the levels of PCNA and cl_caspase3 was determined by densitometry of corresponding bands with respect to β-actin (mean ± SEM, ns, not significant, ∗p < 0.05, ∗∗∗∗p < 0.0001, one-way ANOVA, Tukey’s test, N = 5).(B–E) Three mice from each group (mentioned in Figure 4) were perfused; coronal sections to stain cortex (B and C) or the hippocampus (D and E) were prepared, which were subjected to TUNEL staining and IFA using anti-PCNA, anti-NeuN antibodies; and nuclei were stained with DAPI. Representative images obtained for the frontal cortex (B) and hippocampus (D) are provided. (C and E) The percentage of neurons (NeuN+) that contained staining for TUNEL as well as PCNA (TUNEL^+^/PCNA^+^) was counted from a random sampling of five images from three sections per mouse (mean ± SEM, ∗p < 0.05, ∗∗p < 0.01, ns, not significant, two-way ANOVA, N = 3), Scale bars, 100 μm.

We also analyzed the hippocampus of clomipramine-treated TgAD mice by performing IHC. Similar to the cortex, neuronal apoptosis in the hippocampus sections was also assessed by TUNEL staining (Figure 5D). There was a significant reduction in TUNEL^+^ neurons, which was higher in TgAD mice upon clomipramine treatment (Figure S5B). Even though TUNEL^+^/PCNA^+^ neurons in the TgAD hippocampus (∼10%) (Figure 5E) was smaller than the cortex (Figure 5C), clomipramine reduced this cell population, which reflects neurons undergoing cell cycle re-entry. Collectively, these data indicated that clomipramine is able to prevent apoptosis of neurons in both the cortex and the hippocampus. Given that these regions are pivotal for learning and memory, the reversal of neuronal apoptosis explained the improvement in cognition by clomipramine in TgAD mice.

Discussion

Recent advances in AD research suggest that 65%–70% of clinical trials for AD have evolved from the therapeutics directed against Aβ as a pharmacological target.49^,^50 Lecanemab and donanemab are Aβ_42_-mAb-based drugs that have been approved for use in patients with mild disease and were more effective than other mAb drugs targeting Aβ_42_.2 These drugs were shown to reduce plaques and provide some clinical benefits to patients with mild disease. However, side effects have been reported after treatment with these drugs, which include cerebral amyloid angiopathy (CAA), and several patients have suffered amyloid related imaging abnormalities (ARIA)-associated fatalities.51 Clearly, these and other studies highlight the need for other therapeutic strategies against AD.

E3 ligases play an important role in neurodegeneration and are also emerging as druggable targets in various diseases including neurodegenerative disorders.52^,^53 We demonstrated that ubiquitin ligase Itch was aberrantly regulated in cortical neurons of TgAD mice or rat neurons upon Aβ_42_ treatment. While there was no significant change in the Itch expression in TgAD neurons, detailed investigations revealed that JNK-mediated phosphorylation of Itch and its autoubiquitination at T222/S232 and K393, respectively, resulted in sustained hyperactivation of Itch activity.13 Importantly, overexpression of mutants of these sites prevented CRNA in TgAD neurons in culture. In the present study, we have made use of these loss-of-function mutants—K393R and T222A/S232A—and have demonstrated that their AAV-mediated expression in the cortex reversed cognitive defects in TgAD mice in MWM and Y-maze tests (Figure 1). Strikingly, the levels of cortical neuronal apoptosis, which were higher in TgAD mice, were significantly reduced upon overexpression of Itch mutants (Figure 2). Importantly, their administration did not cause a significant change in WT mice, which suggested almost no toxicity. These results provide a strong explanation for the reversal in cognitive defects by these mutants and support the notion that aberrant activation of Itch may be detrimental in AD.

While the MWM and the Y-maze test used in present studies are generally considered to assess hippocampus-dependent tasks, which include spatial learning and working memory,54 more recent studies also implicate a role of cortical neurons in the processing of working memory.55 Furthermore, there are studies which indicate that spatial and working memory tasks such as the MWM may involve cortex-hippocampus neural networks,37^,^56^,^57^,^58^,^59 in which cortical regions, including the medial prefrontal cortex and dorsomedial striatum, play an important role in hippocampus-related spatial memory tasks.60 Thus, rescuing frontal cortical neuronal health may improve cortical-hippocampal regulation, increase coherence of information flow, and enhance the fidelity of hippocampal outputs assessed by Y-maze test. Since neuronal apoptosis was observed in both cortical and hippocampal regions of the TgAD mouse model,46 even a modest cortex-restricted intervention may have network-level effects, which may “release” the hippocampal circuits from dysregulation, improving their function in spatial tests.

While the use of these mutants, in-principle, can possibly be pursued for gene therapy-based applications against AD, there will be technical and economic challenges. Therefore, the possibility of using small-molecule inhibitors of Itch against AD is a more potent and viable option even though inhibitors can have their own issues like toxicity. As mentioned earlier, a previous screen for Itch inhibitors resulted in identification of several compounds that inhibited Itch, and the most potent of them was clomipramine.19 The other added advantage of clomipramine was that it is being used in the clinic as an anti-OCD and antidepressant drug,20^,^21^,^61 with few side effects associated only with long-term use, such as osteoporosis,62 sexual behavior,63^,^64 mood disorders, and oxidative stress.65^,^66 Clomipramine was able to inhibit Itch activity as indicated by reduced TAp73 ubiquitination in Aβ_42_-treated neurons (Figure 3B). Importantly, it was able to reverse neuronal apoptosis at this concentration in TgAD- or Aβ_42_-treated neurons (Figures 3C and 3D). The fact that cell cycle re-entry was also significantly reversed by clomipramine suggested that the major contribution of Itch to neuronal death may be by promoting cell cycle in neurons. Given that clomipramine (Figures 3C and 3D) did not cause a significant change in Itch levels, its effects can be attributed to its impact on Itch activity. Therefore, preventing aberrant activation of Itch may provide benefits in AD.

Behavioral studies using TgAD mice revealed that clomipramine treatment effectively reverses defects observed in MWM, Y-maze, as well as radial arm test (Figure 4). Given that these tests collectively assess spatial, reference, and working memory, which are impaired in patients with AD,67^,^68^,^69 these data strongly suggest that clomipramine may provide these benefits in AD. Importantly, we did not observe any locomotion-related side effects potentially caused by sedation in clomipramine-treated mice, as the average distance traveled and average speed were almost unaltered (Figures S4A, S4B, S4E, and S4F). The fact that clomipramine treatment almost completely abolished cleaved caspase-3 and significantly reversed TUNEL staining in clomipramine-treated TgAD mouse cortex provided evidence for its ability to reverse neuronal apoptosis in the cortex as well as the hippocampus (Figure 5). Importantly, neuronal death was not observed in WT mice upon clomipramine treatment, which suggested that clomipramine may lack neurotoxicity. In addition, as observed in cultured neurons, clomipramine also significantly reversed cell cycle re-entry in TgAD mouse brain (Figures 5B–5E, S5A, and S5B). These results indicated that clomipramine-mediated reduction in apoptosis of neurons in both hippocampus and the cortex leads to cognitive improvement in TgAD mice.

Clomipramine has been effectively prescribed for more than three decades for patients with depression and OCD with minimum side effects.20 Given that this drug is already in use for psychiatric disorders, it will be worth investigating if it can be repurposed for AD treatments. Clomipramine is suggested to inhibit serotonin uptake by blocking serotonin transporters, which is the suggested mechanism of its action in psychiatric disorders.20 Some studies suggest that clomipramine negatively regulates the autophagic flux glucocorticoid functions.70^,^71 However, its targets have not been shown unequivocally in these processes. Clomipramine was shown to directly inhibit Itch catalytic activity in vitro by preventing its transthiolation.19 Present studies demonstrate that clomipramine indeed inhibits Itch activity in neurons as it blocked TAp73 ubiquitination and degradation (Figure 3B). The fact that TAp73 levels were restored in clomipramine-treated TgAD mouse brains suggested that it blocks Itch in vivo as well. The fact that clomipramine did not alter the number of amyloid plaques in TgAD mice confirmed that it blocks processes downstream of Aβ_42_ generation and independent of plaque formation (Figure S4G), which fits well with previous observations demonstrating that Aβ_42_ promotes aberrant activation of Itch, leading to neuronal death.13 Therefore, clomipramine may provide relief to patients despite the presence of plaques and may also be effective at advanced stages of disease, although further validation will be needed.

Itch is known to regulate immune signaling and inflammatory responses.72^,^73^,^74 Since neuroinflammation is also a critical contributor to AD pathology,75^,^76 the observed cognitive benefits of clomipramine may not be limited only to neuroprotection. It will be worth investigating if Itch mutants and clomipramine influence microglia activation, which contributes to AD.77

In conclusion, our findings establish Itch as a key effector of Aβ_42_ toxicity and a promising therapeutic target in AD. Targeting Itch by using loss of function mutants or using Itch inhibitors such as clomipramine may prevent cognitive defects associated with AD. Present studies provide strong evidence supporting the use of clomipramine against AD. Since clomipramine has already been used against psychiatric disorders, its repurposing for AD patients is a viable option.

Limitations of the study

Our study provides an extensive analysis of therapeutic potential of targeting Itch in AD to prevent neuronal apoptosis not only in in vitro culture system but also in vivo using the APP/PS1 mouse model for AD. Although, this model recapitulates pathological features of the disease, validation of these results in other animal models of AD will provide added benefits from a therapeutic perspective. In addition, since AD pathology is multifactorial with significant involvement of neuroinflammatory processes and Itch is known for its role in immune regulation, studying the effect of clomipramine and Itch mutants on neuroinflammation and microglial activation will provide better regulatory insights in AD.

Resource availability

Lead contact

Requests for additional information, resources, and reagents should be directed to and will be fulfilled by the lead contact, Pushkar Sharma (pushkar@nii.ac.in).

Materials availability

Materials generated in this study are available upon request, subject to a completed materials transfer agreement.

Data and code availability

All data reported in this study will be shared by the lead contact upon request. This article did not report the original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

Birendra N. Roy provided help in cryosectioning of mouse brain tissues. The authors thank Dr. Binukumar BK for critically reading the manuscript and his suggestions. This work was supported by Department of Biotechnology, India (grants #BT/PR23577/MED/122/79/2017 and BT/PR44509/MED/122/292/2021) and core funding from BRIC-10.13039/501100011753NII, India. P.S. is a recipient of J.C. Bose Fellowship; M.C. received Senior Research Fellowship from DBT.

Author contributions

Investigation, data curation, resources, visualization and validation, formal analysis, and writing, reviewing, and editing, M.C. and K.S.; supervision, writing manuscript, project administration, and funding acquisition, P.S.

Declaration of interests

P.S. and M.C. are inventors on US patent #11957692 (Clomipramine for treatment of Alzheimer’s disease).

STAR★Methods

Key resources table

REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesItchCell Signaling TechnologyCat#12117; RRID: AB_2797822Myc-tag (9B11)Cell Signaling TechnologyCat#2276; RRID:AB_331783TAp73NovusCat#NBP2-24737; RRID:AB_2927606PCNASanta CruzCat#sc-56; RRID:AB_628110Cleaved Caspase 3Cell Signaling TechnologyCat#9661; RRID:AB_2341188ActinSanta CruzCat#sc-47778; RRID:AB_626632UbiquitinSanta CruzCat#sc-8017; RRID:AB_628423NeuNNovusCat#NBP2-67314Bacterial and virus strainsE. coli DH10βAddgeneStrain# 229395; RRID: CVCL_V592pAAV-CAG-tdTomato (empty)AddgeneStrain# 59462; RRID: Addgene_59462pAAV2/9nAddgeneStrain# 112865; RRID: Addgene_112865pAdDeltaF6AddgeneStrain# 112867; RRID: Addgene_112867Chemicals, peptides, and recombinant proteinsAβ_42_R-peptideCat# A-1163-2MG132MerckCat# 474790Clomipramine hydrochlorideMerckCat#C7291Thioflavin TMerckCat#T3516Critical commercial assaysDeadEnd TUNEL Fluorometric kitPromegaCat#G3250Experimental models: Cell linesHEK293ATCCCat# CRL-1573HEK293TATCCCat# CRL-3216Experimental models: Organisms/strainsMouse:APP(Swe)/PS1(dE9)Jackson LaboratoriesStrain # 004462Sprague-Dawley RatCharles RiverRRID:RGD_70508OligonucleotidespAAV-Itch- Forward-CGGGGTACCATGTCTGATAGTGGATCACAMerckN/ApAAV-Itch- Reverse-AAATATGCGGCCGCTTAGAACAAAAACTCATCTCAGAAGAGGATGGTCTTCTTGGAGCAGGTGGAGGTGGMerckN/ApAAV-TS/AItch- (T222A)-Forward- CCACCTCCACCTGCTCCAAGAAGACCAMerckN/ApAAV-TS/AItch- (T222A)- Reverse- TGGTCTTCTTGGAGCAGGTGGAGGTGGMerckN/ApAAV-TS/AItch- (S232A)-Forward- GCTTCTGTCAATGGCGCACCATCCACGAATTCTMerckN/ApAAV-TS/AItch- (S232A)- Reverse- AGAATTCGTGGATGGTGCGCCATTGACAGAAGCMerckN/ApAAV-K/RItch- (K393R)-Forward-GCTACCTCACAGAACAGAGAATTTGATCCGCTTMerckN/ApAAV-K/RItch- (K393R)- Reverse-AAGCGCATCAAATTCTCTGTTCTGTGAGGTAAGCMerckN/ARecombinant DNApAAV-ItchThis paperN/ApAAV- T222A/S232AThis paperN/ApAAV-K393/RThis paperN/ASoftware and algorithmsGraphPad Prism 10Graphpad Softwarehttps://www.graphpad.com/ImageJNIH–EndnoteClarivatehttps://www.endnote.com/ANY-mazeANY-maze softwarehttps://www.any-maze.com/BiorenderBiorender Softwarehttps://www.biorender.com/Zen BlueZeisshttps://www.zeiss.com/AxiovisionZeisshttps://www.zeiss.com/

Experimental model and study participant details

Animals and cell lines

For the behavioural experiments, our study has taken into account only the female from APP(Swe)/PS1(dE9) mouse model since female mice show more robust AD phenotype compared to male mice.78^,^79 However, for primary cortical neurons culture, embryonic day 16-18 mice or rats (irrespective of gender) were used. All experiments were designed and performed in accordance with the guidelines of Institutional Ethics Committee. Animal work has been approved by Institutional Animal Ethics Committee (IAEC) with IAEC serial #461/18 and #513/19. This study was performed according to ethics committee approval and is compliant with “Declaration of Helsinki”. All experiments related to AAV usage were approved by Institutional Bio Safety committee (IBSC) #346/19.

APP(Swe)/PS1(dE9) (TgAD) Alzheimer’s disease mouse model

APP(Swe)/PS1(dE9)transgenic mice (TgAD) (strain B6C3-Tg APPswe, PSEN1dE9 85Dbo/J; stock number 004462-JAX) maintained at the Jackson laboratory was a gift from National Brain Research Centre, Manesar to National Institute of Immunology (NII), New Delhi. These are mice with a chimeric mouse/human amyloid-β precursor protein containing K595N and M596L Swedish mutations and a mutant human presenilin1 gene carrying the exon 9 deletion which causes production of significantly higher amounts of Aβ_42_, especially in the females.78^,^79 Mice were genotyped using genomic DNA isolated from mouse tail and PCR primers as recommended by Jackson’s laboratory. After genotyping, mice were randomly divided into experimental groups (n=8 per group).

Sprague-Dawley (SD) rats

SD rat colonies were maintained at NII animal facility vivarium. Adult rats were used for timed mating. Male and female rats (6-8 weeks old) were put together in a 2:1 harem. The males were removed after 2 days. After 10 days, the females were checked for pregnancy. Only two mated females were housed together and always handled gently to avoid any handling trauma. Between days 16-18, the pregnant rat was euthanized with CO_2_ overdose and the embryos were collected for primary neuronal cell culture.

Cell lines

HEK293T (Human Embryonic Kidney) cells were purchased from American Type Culture Collection (ATCC) and maintained in DMEM with 10% fetal bovine serum and 1x antibiotic/antimycotic at 37°C in 5% CO2. HEK293T is not a commonly misidentified cell line as suggested by the International Cell Line Authentication Committee (http://iclac.org/databases/cross-contaminations/). Therefore, it has not been authenticated and was tested for the absence of mycoplasma.

Method details

Primary cortical neurons culture

Embryonic day 18 (E18) Sprague-Dawley rats or embryonic day 16 (E16) TgAD mice were euthanized and cortices were dissected out from the brains and neurons were cultured as described previously.11^,^12^,^13 Briefly, the cortical region of the brain was isolated and trypsinized followed by addition of Serum Containing Media (SCM) (10% fetal bovine serum and 10% horse serum). After 12 h, cells were washed with tyrode’s PBS supplemented with glucose and NaHCO_3_ and further maintained in Serum Free Media (SFM) (Neurobasal supplemented with glucose, L-glutamine, B27 and N2 supplements) for five days.

Soluble oligomers of Aβ_1-42_ (R-peptide) was prepared and 0.5 μM was used as described previously and the treatment was typically done for 48 h.11^,^12^,^13^,^80 1 M solution of clomipramine (Merck) was prepared by dissolving clomipramine HCl in ultrapure water. This 1M solution was used at indicated concentrations for treating cultured neurons for 48 h.

AAV production

For preparation of recombinant AAV, mouse WT or T222A/S232A or K393R mutant Itch was subcloned from the constructs described previously13 in pAAV-CAG-tdTomato vector [(pAAV-CAG-tdTomato (codon diversified)], which was a gift from Edward Boyden (Addgene plasmid #59462; http://addgene.org/59462; RRID: Addgene_59462). WT and mutant Itch AAV constructs were generated using primers indicated in Key Resources.

HEK293T cells were co-transfected with pAAV-td Tomato or pAAV-Itch or pAAV-T222A/S232A Itch or pAAV-K393R Itch along with trans-plasmid pAAV2/9n (gift from James M. Wilson; Addgene #112865; http://addgene.org/112865; RRID: Addgene_112865) and a helper plasmid pAdDeltaF6 (gift from James M. Wilson; Addgene #112867; http://addgene.org/112867; RRID: Addgene_112867. We have used AAV2/9n which has genetic material of AAV2 combined with capsid of AAV9 (http://addgene.org/112865). For transgene expression, CAG promoter was used, which has a CMV enhancer with chicken beta actin promoter and its first intron, which makes CAG promoter robust and efficient.81 Cells were harvested after 72 h and were lysed by cycles of freeze-thaw and virus particles were concentrated and resuspended in AAV storage buffer (1x PBS, 35 mM sodium chloride and 5% glycerol).82^,^83

Stereotaxic injections

6-month old TgAD or WT mice were stereotaxically injected with recombinant AAV coding for WT Itch or its T222A/S232A or K393R mutant in the frontal cortex region of the brain. Stereotaxic injections were performed on the six months old TgAD or WT mice as described previously.14 Briefly, mice were deeply anesthetized with ketamine and xylazine and a small incision was made in the head and Hamilton syringe needle was set at bregma. With the help of stereodrive, needle was located at anterior/posterior: +1.8 mm; mediolateral: +/-1.9 mm; dorsal/ventral: -1.2 mm to the bregma position and 3 μl of AAV was injected in both sides of the brain cortex.84 AAV was injected at a dosage of ∼1x10^10^ vg/kg body weight of the mouse through bilateral injections.

Clomipramine treatment

Genotyped TgAD and age-matched WT mice were randomly distributed in four groups. Clomipramine HCl solution was prepared in ultrapure water as a 10 mg/ml solution and diluted with saline to 5 mg/ml. 100 μl (25 mg/kg) of this solution was injected intraperitonially in 6-month old WT or TgAD mice. The saline or clomipramine injections were given every other day for 45 days followed by behaviour tests and euthanasia for tissue analysis. The dose of clomipramine for animal experiments was decided based on a previous study70 and recommended dose in humans is 2 mg/kg per day, which was converted for use in mice as per USFDA guidelines.

Behavioral tests

Morris water maze test

Mice were subjected to three trials per session and one session a day for five days. Each trial was performed for 60 s in which mice had to reach the hidden platform in the water, and the average time per session was recorded as escape latency. After the escape training, the platform was removed for the probe trial test, which tested the retention of spatial memory at 24 h after training.33^,^34 The number of times a mouse crossed the probe, the time spent in the target quadrant, and swim speed were recorded using ANY-maze software (Stoelting Co., USA).

Y-maze test

Mice were habituated in the Y-shaped maze for 3 min with one of the arms blocked then their spontaneous alternations were recorded with all arms open for 5 min.39^,^85^,^86 The number of alternations was recorded manually and % spontaneous alternations were calculated as mentioned below.

$$[eqn]$$

Radial arm (8-arm) maze test

Mice were given restricted amount of food until their body weight reduced to 85% of the initial weight for two days prior to habituation in the maze. In habituation phase, each mouse was allowed to explore and consume food pellets scattered on the entire maze for a 10 min period (one session per mouse). In the training phase, three arms were blocked, and the remaining five alternate arms were baited. On the test day, all arms were opened and same five arms were baited as done in trial sessions and series of entries were recorded manually.87 Subsequently, % working and reference memory errors were calculated. The mouse entering an arm containing food in which it had previously entered was counted as working memory error (short term memory) and if the mouse entered in an arm without bait it was counted as reference memory error (long term memory for position of the baited arms).

Immunohistochemistry (IHC)

After performing behavioral tests, animals were deeply anesthetized with ketamine/xylazine solution (75 mg/kg ketamine +10 mg/kg xylazine, intramuscular injection) and then perfused with 1x PBS followed by 4% paraformaldehyde (PFA). Brains were isolated and kept in 4% PFA for 12 h and then dehydrated in 30% sucrose. Blocks were prepared using Leica Tissue freezing medium and stored at -80°C. 10 μm sections on poly-L-lysine coated glass slides were prepared using a Leica cryotome.

For TUNEL labelling and IHC, slides with coronal brain sections were twice rinsed in 1x PBS for 10 min at room temperature. Next, heat-mediated citrate buffer antigen retrieval (AR) step was performed by boiling for 10 min and gradual cooling at room temperature for 20 min and then rinsed with water. After a 1x PBS wash, 5 min proteinase K treatment was given followed by 4% PFA fixation for 3 min followed by TUNEL labeling performed by using Dead End fluorometric TUNEL system (Promega) as per manufacturer’s guidelines. Subsequently, blocking was done with 3% BSA, 2% normal goat serum and 0.15% Triton X 100 for 1 h followed by incubation with primary antibodies in blocking buffer at 4°C for 12 h. Subsequently, the Alexa Fluor (488 or 594) conjugated or r-PE conjugated secondary antibodies were used. Sections were mounted with DAPI containing medium (VectaShield) for microscopy.

Single-plane four-channel [(DAPI (blue), TUNEL (green), PCNA (red) and NeuN (yellow)] 8-bit images were obtained using a Zeiss AxioImager 1 microscope. The acquisition settings were kept uniform for all images. Five fields per section and three sections per mouse were obtained for quantification of stained cells. Immunofluorescence images were analysed in Zeiss AxioVision or Zen Blue software.

Thioflavin T staining

The cryopreserved brain sections were thawed at room temperature followed by dehydration in 70% and 80% ethanol for one minute each. The slides were then dipped into the filtered solution of Thioflavin T (Merck) (0.002% in 80% ethanol) for 8 min followed by washing in decreasing concentrations of ethanol from 80%-50% and finally in water for 2 min. Sections were mounted using DAPI containing mounting media. ThT-stained brain sections were imaged and processed with Zeiss AxioImager fluorescence microscope and Axiovision software was used for quantification. The intensity of ThT staining was quantified using ImageJ (NIH).

Western blotting and immunoprecipitation

For protein lysate preparations from tissues, mice were deeply anesthetized and perfused with ice-cold 1x PBS. Cortex and hippocampus were dissected out from perfused brain and stored at 80°C for immunoblotting analysis. For immunoblotting, equal amount of tissue from each mouse was minced in RIPA lysis buffer (100 mM Tris-HCl pH 7.4, 50 mM NaCl, 50 mM EDTA and 1% Triton X-100) and supplemented with a protease inhibitor cocktail (Roche), 2% SDS and 1 mM phenyl methane sulfonyl fluoride and homogenized.

Protein lysates were made in 2% SDS from primary neurons and for immunoblotting. Immunoblotting was performed as described previously13 using primary antibodies and secondary antibody conjugated with horseradish peroxidase (HRP). Chemiluminescence reagent West Pico or West Dura (Pierce) was used for detection as per manufacturer's instructions.

Immunoprecipitation was performed as described previously.13 Briefly, cells were washed with 1x PBS and lysed using ice cold complete lysis buffer (100 mM Tris-HCl pH 7.4, 5 mM EDTA, 100 mM NaCl, 1% Triton x100 and 10% Glycerol, 1 mM phenyl methane sulfonyl fluoride (PMSF), 1 mM sodium orthovanadate, 20 mM β-glycero-phosphate and 1x protease inhibitor cocktail. 50-100 μg of protein lysate was incubated with 1 μg of TAp73 antibody followed by incubation with 50 μl of protein A+G Sepharose (cat.no. sc-1001, Santa Cruz Biotechnology). The resin was washed at 4°C to remove unbound proteins and resuspended in lysis buffer and immunoblotting was performed as described above. Image J (NIH) software was used to perform densitometry of Western blots. The band intensity of the loading control (β-Actin) was used for normalization.

Antibodies used

Following are the details of the antibodies used in this work

PCNA (Santa cruz, cat. no. sc-56; 1:500 for WB, 1:50 for IHC), cleaved caspase 3 (CST, cat. no. 966l; 1:1000 for WB), Itch (CST, cat. no. 12117; 1:1000 for WB), TAp73 (Novus, cat. no. NBP2-24737, 1:1000 for WB, 1:50 for IP), myc-tag (9B11) (CST, cat. no. 2276; 1:2000 for WB, 1:100 for IHC), NeuN (Novus, cat. no. NBP2-67314; 1:100 for IHC), Ubiquitin (Santa cruz, cat. no. sc-8017; 1:1000 for WB) and β-Actin (Santa cruz, cat. no. sc-47778, 1:2000 for WB).

Quantification and statistical analysis

All the data are presented as mean of at least 3 independent experiments ± SEM for most experiments. Unless indicated otherwise, student t-test (two-tailed), two-way ANOVA or one-way Analysis of Variance (ANOVA) followed by Tukey’s or Sidak's multiple comparison method was used for statistical analysis by using GraphPad Prism software (GraphPad software Inc USA). Statistical significance was defined only when p<0.05. The significant statistical tests are presented as asterisks ∗ for p<0.05, ∗∗ for p<0.01, ∗∗∗ for p<0.001 and ∗∗∗∗ for p<0.0001.