SAMP8 was backcrossed during at least 12 generations with C57/BL6-p75^{exonIII−/−} (Lee et al., 1992). After 12 generations, heterozygotes SAMP8-p75^{exonIII +/−} mice were bred, and SAMP8-p75^exonIII+/+ and SAMP8-p75 ^{exonIII−/−} littermates were selected for the aging studies. The same procedure was carried out to generate a SAMR1-p75 ^{exonIII−/−} mouse strain with the SAMR1 background, but we were unsuccessful for unknown reasons. SAMP8-p75 ^{exonIII−/−} mice showed a similar life survival (p-value, 0.51) respect to SAMP8-p75^exonIII+/+ (Figure 1A).

SAMP8-p75 ^{exonIII−/−} mice showed a significant reduction in the total weight gain in a normal diet in comparison to the SAMP8-p75^exonIII+/+ mice (Figure 1B). The SAMP8 strain has been reported to show an increase in astrogliosis in the hippocampus and in the cortex (Díaz-Moreno et al., 2013). The astrogliosis of SAMP8- p75 ^exonIII+/+ and SAMP8-p75 ^{exonIII−/−} was quantified as the intensity of GFAP immunofluorescence in the CA1 region and in the dentate gyrus of the hippocampus (Figures 1C–E).An increase in the astrogliosis with age was found in the CA1 but no significant differences were observed between SAMP8-p75^exonIII+/+ and SAMP8-p75 ^{exonIII−/−}.

A previous work showed a reduction in adult hippocampal neurogenesis in the C57/BL6-p75^{exonIII−/−} strain (Catts et al., 2008) that was related to a reduction in the width of the hippocampal dentate gyrus granule cell layer, indicating a role of the p75^NTR in neurogenesis. We decided to characterize adult hippocampal neurogenesis in our newly generated model. In the case of the SAMP8 strain, there were no significant differences in the width of the granule cell layer (at 2 months SAMP8-p75^exonIII+/+ 49.2 ± 2.5 μm, N = 5 vs. SAMP8-p75 ^{exonIII−/−} 46.2 ± 1.9 μm, N = 4; at 10 months SAMP8-p75^exonIII+/+ 50.2 ± 9.2 μm N = 4 vs. SAMP8-p75 ^{exonIII−/−} 49.5 ± 0.3 μm N = 4), however, quantification of the number of BrdU⁺ cells showed that in the SAMP8-p75 ^{exonIII−/−} mice at the age of 2 months there was a slight decrease in the proliferative activity of the neural stem cell niche that did not reach statistical significance (Figures 1F–I). In addition, there was a reduction in the number of BrdU⁺/DCX⁺ newly born neurons in the subgranular zone (SGZ) of the dentate gyrus of the hippocampus during aging (from 2 to 10 months) and in the total number of immature DCX⁺ neurons irrespective of the mouse genotype (Figures 1H,I), supporting previous findings indicating that in aged SAMP8 mice, neurogenesis is impaired (Díaz-Moreno et al., 2013).

As p75^NTR is highly expressed in the BFCNs, we focused on the study of this region (Figure 2). p75^NTR immunohistochemistry using an antibody against the intracellular domain was undertaken on BF sections. We quantified approximately 40% more ChAT⁺ neurons in SAMP8-p75^{exonIII−/−} than SAMP8-p75^exonIII+/+ animals in the basal forebrain [medial septum (MS) and vertical diagonal band (VDB)] (Figures 2A–C). The number of the BFCNs in the SAMP8-p75 ^{exonIII−/−} is highest at 2 months of age, nevertheless, at 10 months of age, the two mouse genotypes showed the same number of BFCNs (Figure 2C), suggesting that the increased number of BFCNs in the SAMP8-p75^{exonIII−/−} degenerate or become ChAT-negative during this time interval.

To study if the accelerated decrease in the number of BFCNs in the SAMP8-p75^{exonIII−/−} is due to the mouse background strain, we compared the number of BFCNs in the C57/BL6-p75^{exonIII−/−} mouse strain (Figure 3). As the SAMP8 mice have accelerated aging, we quantified the number of BFCNs in C57/BL6-p75^exonIII+/+ and C57/BL6-p75^{exonIII−/−} until geriatric ages (ca. 30 months old). We found that the number of BFCNs at birth is higher in the C57/BL6-p75^{exonIII−/−} similar to the SAMP8 background and to previous reports (Figures 3A,B) (Naumann et al., 2002), and the number of BFCNs slowly decreased and approached the same number than C57/BL6-p75^exonIII+/+ at around 12 months old and reaching lower levels at 24 months of age (Figure 3B).

Altogether this indicates that the deletion of p75^NTR induces an increase in the number of BFCNs at birth, but in the long term, there is a decrease in the survival of BFCNs independent of the mouse strain (SAMP8 or C57/BL6). The loss of BFCNs in the SAMP8 background in comparison to the C57/BL6 background could be mediated by the specific characteristics of this accelerated aging mouse model strain.

To determine whether p75^NTR deficiency in the SAMP8 mice affected anxiety and cognitive ability, a cohort of mice were subjected to the following behavioral tests: open-field (OF), Y-maze (YM) and novel object recognition (NOR). In order to assess the role of BFCNs, mice were tested at two different ages, 2 and 6 months. In the OF test, mice are tested for anxiety-related parameters measured as time spent in the center of the box. As animals display a natural aversion to brightly open areas (central zone) but the also have a dive to explore new environments (Kovacsics and Gould, 2010). A significant difference in the % of time spent in the center in the OF test was observed at both 2 and 6 months of age between SAMP8-p75^{exonIII−/−} and SAMP8-p75^exonIII+/+ animals (Figures 4A,B), showing an increase of time the SAMP8-p75^{exonIII−/−} animals. The behavior of SAMP8-p75^{exonIII−/−} in the OF is similar to the control mice SAMR1, indicating that the deletion of p75 ^NTR in the SAMP8 rescues the behavior of the SAMP8-p75^exonIII+/+ mice in this test. The SAMR1 has more BFCNs than the SAMP8 mice at this age and is more similar in number to the SAMP8-p75^{exonIII−/−} (Figure 2C), suggesting that the increased cholinergic innervation from the BFCNs may play a role in this phenotype. During the YM test, mice are positioned within the central point of a maze comprising three opaque arms, arranged in the configuration of the capital letter Y. Mice are allowed to freely explore their surroundings and due to their innate inclination to discover novel environments, the rodents exhibit a preference for exploring new arms of the maze. The primary objective of this test is to test their episodic memory, which is quantified by observing how often a mouse selects a previously unexplored arm of the maze (considered as the correct behavior). This parameter is referred to as SAB (spontaneous alternation behavior), a measure of the mouse’s tendency to alternate between arms without any external cues (Kraeuter et al., 2019). At 2 months of age, SAMP8-p75^exonIII+/+ and SAMP8-p75^{exonIII−/−} mice were not behaving significantly differently (Figure 4C). However, at 6 months of age, the SAMP8-p75^{exonIII−/−} performed significantly worse than at 2 months of age, quantified with a lower % of correct alternations suggesting a worsening of these cognitive abilities. The decrease in the number of BFCNs from 2 to 6 months (around 50%) in SAMP8-p75^{exonIII−/−} may be responsible for this worsening in cognitive ability. This difference is not observed in the SAMP8-p75^exonIII+/+ mice during aging from 2 to 6 months (Figure 4C), which did not display a reduction in the number of BFCNs during this time frame. In the NOR test, the preference for exploring new objects over familiar ones evaluates short and long memory. The test requires a training phase, where familiar objects are presented. This study specifically focused on evaluating long-term memory as the test phase is performed 24 h after the training phase. When tested in NOR (Figure 4D) there were no differences between SAMP8-p75^exonIII+/+ and SAMP8-p75^{exonIII−/−} in the time exploring the novel object, but these two genotypes clearly performed worse than the control mice SAMR1 (Figure 4D). The decrease in the number of cholinergic neurons may impact on the release of the acetylcholine neurotransmitter. We measured the levels of acetylcholine in the different mice genotypes at the two ages of the study (Figure 4E). As shown, acetylcholine levels decrease in the SAMP8-p75^{exonIII−/−} from 2 to 6 months but not in the other genotypes (SAMR1 or SAMP8-p75^exonIII+/+), suggesting that the decrease in the number of cholinergic neurons impact on the synthesis of acetylcholine and in the behavior.

Altogether the deletion of p75^NTR in the pathological strain SAMP8 induces differences in some of the studied tests, probably depending on the neuronal circuit involved.

The phenotype described in the SAMP8-p75^{exonIII−/−} suggests a deleterious effect of the deletion of p75^NTR on the survival of the BFCNs. Although these data suggest a pro-survival role of p75^NTR, it has been described that depending on the mouse strain, the p75^{exonIII−/−} mice express a short isoform of p75^NTR that may play a negative impact on the neurons where it is expressed (von Schack et al., 2001). We analyzed this possibility in the SAMP8 background and found that in the SAMP8-p75^{exonIII−/−}, there was a significant signal in the BFCNs when stained using a specific antibody against the intracellular domain (ICD) of p75^NTR (Figure 5A). To confirm this observation, we performed immunoprecipitation using a p75^NTR -ICD antibody and western blots analysis of basal forebrain tissue from 2- and 6-months old mice. The Figure 5B shows the presence of a short isoform of p75^NTR in the SAMP8-p75^{exonIII−/−}. This short isoform migrates at a similar size as the p75^NTR-C-terminal fragment, p75^NTR-CTF, construct analyzed in the same blot (Figure 5B) and is reminiscent of a short isoform previously described in von Schack et al. (2001). Interestingly the expression of this short isoform is also observed in the C57/BL6-p75^{exonIII−/−} mice by immunostaining and co-immunoprecipitation (Figures 5A,B), indicating that the expression of the short isoform is due to the p75^exonIII genetic cassette construct and not dependent on the mouse strain. As p75^NTR undergoes receptor intramembrane proteoysis (RIP) in physiological conditions, the p75^NTR-CTF is also observed in the wt mice (both SAMP8 and C57/BL6 strains) together with the p75^NTR full length (p75-FL) protein (Figure 5B). Quantification of the ratio p75-CTF/p75-FL (Figure 5C) indicates that there is a significant increase of this ratio in the SAMP8-p75^{exonIII−/−} and C57/BL6-p75^{exonIII−/−} mice that might play an aberrant signaling in these cells.

It has been described that p75^NTR regulates the metabolism of cholesterol in the nervous system (Yan et al., 2005; Korade et al., 2007; Follis et al., 2021) among other tissues (Pham et al., 2019). We quantified the levels of Sterol Regulatory Element-binding Protein-2, SREBP2, a transcription factor involved in the upregulation of key genes for cholesterol biosynthesis and cholesterol uptake (Figure 6A). As it is shown in the Figure 6B analysis by western blot of basal forebrain tissue showed a significant increase in the total expression of SREBP2 in the p75^{exonIII−/−} with respect to p75^exonIII+/+ in both SAMP8 and C57/BL6 mouse strain. SREBP2 plays an important role in the homeostasis of cholesterol by regulating the 3-hydroxy-3-methylglutaryl-coenzyme A, HMGCR, a key enzyme in the synthesis of cholesterol and the low-density lipoprotein receptor, LDLR, that mediates the uptake of extracellular cholesterol. As shown in Figure 6B, the protein levels of HMGCR and LDLR increase significantly at 2 and 6 months in the in the SAMP8-p75^{exonIII−/−} versus SAMP8-p75^exonIII+/+. In the case of the C57/BL6 mice, there is a significant increase at 6 months but not at 2 months. This difference could be related to the fact that the SAMP8 mice have an accelerated aging in comparison to the normal aging of the C57/Bl6. In any case at 6 months old, a similar increase of SREBP2, HMGCR and LDLR is observed indicating that is a general mechanism of the p75^{exonIII−/−} mice and not of the mouse background.

The cellular type responsible of HMGCR expression is important, as it is described that in the adult brain the synthesis of cholesterol takes place mainly in the astrocytes. Immunofluorescence instead showed an increase of the HMGRC staining in the neurons of the basal forebrain almost exclusively in the ChAT+ neurons of the BF of both SAMP8- p75^{exonIII−/−} and in the C57/BL6-p75^{exonIII−/−} mice (Figure 6F), supporting the western blot data. We then analyzed the total levels of cholesterol in basal forebrain extracts from SAMR1, SAMP8-p75^{exonIII−/−}, and SAMP8-p75^exonIII+/+ at 2 and 6 months (Figure 6H) and found an increase in the total levels of cholesterol, although not reaching statistical significant values, in the SAMP8-p75^{exonIII−/−} at 2 and a 6 months of age respect to SAMP8-p75^exonIII+/+.

So far, we have described a correlation between the presence of a short isoform of p75^NTR in the p75^exonIII mice (independent of the mice strain) and an increase in the cholesterol biosynthetic proteins in the BFCNs. To demonstrate that the neurotrophin signaling is able to activate these genes, we used a heterologous system like the PC12 cells that express endogenous levels of p75^NTR and TrkA, stimulated with NGF for 24–48 h (Figure 7). Stimulation of NGF induces an increase in the expression of SREBP2 and HMGRC in 24 h but not of LDLR (Figures 7A,B) and an increase in the cholesterol content, as shown by filipin staining (Figures 7C,D). NGF/TrkA stimulation also induces p75^NTR upregulation (Figure 7A). Interestingly NGF stimulation in the presence of a TrkA kinase activity inhibitor, K252a, prevents the activation of these genes (Figures 7A,B) and the increase of cholesterol content measured by filipin staining (Figure 7C). As TrkA activation by NGF induces the regulated intramembrane proteolysis (RIP) of p75^NTR, we incubated the PC12 cells in the presence of an α-secretase inhibitor (TAPI-1) or a γ-secretase inhibitor (Compound E) (Figures 7E,F). Inhibition of p75^NTR shedding with TAPI-1 reduces the expression of HMGCR induced by NGF/TrkA, suggesting that shedding of p75^NTR is required for the upregulation of HMGCR. The presence of a γ-secretase inhibitor partially impairs the upregulation of HMGCR, indicating that the RIP of p75^NTR plays a role in HMGCR upregulation. Activation of p75^NTR independently of TrkA with BDNF (Figures 7E,F), is not able to upregulate the expression of HMGCR, supporting the data that TrkA activity is required. These experiments suggested that the activation of TrkA by NGF is required for the shedding of p75^NTR, which is the main driver of the upregulation of the cholesterol biosynthesis genes. In order to demonstrate the direct role of p75^NTR-CTF in this process we overexpressed p75^NTR, p75^NTR-CTF and p75-ICD in PC12TrkA/p75DKO cells (Testa et al., 2022) and found and increase in SREBP2 and HMGCR expression independently of NGF and TrkA (Figures 7G,H) in the cells overexpressing p75-CTF suggesting that accumulation of p75-CTF is a causative agent and pointing to the findings observed in vivo in the SAMP8-p75^{exonIII−/−} and C57/BL6-p75^{exonIII−/−} mice (Figure 6).

SAMP8 mice were backcrossed with C57BL6 p75^{NTR exonIII +/−} mice (Lee et al., 1992) (Jackson Laboratories) for 12 generations to create a new SAMP8-p75 ^{exonIII −/−} strain. The animals were housed on a 12 h light/12 h dark circle with food and water provided ad libitum in specific pathogen-free (SPF) at constant 24 degrees temperature. All animal experimentation was controlled following the recommendations of the Federation of European Laboratory Animal Science Associations on health monitoring, European Community Law (2010/63/UE), and Spanish law (R.D. 53/2013) with approval of the Ethics Committee of the Spanish National Research Council (1,246/2022) and the local Government (2022-VSC-PEA-0139 type 2).

Groups of mice of 2 and 6 months conducted 3 different tests in the following order: open field, Y-maze test, and Novel Object Recognition test (NOR). Before performing the behavioral tests, the mice were moved to the behavioral room for habituation. In addition, each mouse was 5 min per day for 4 consecutive days with the experimenter. Every test was separated 3 to 5 days to let the mice rest.Y.

PC12 cells were cultured in DMEM (Gibco) supplemented with 10% FBS (Sigma), 1% Penicillin–Streptomycin, 1% L-Glutamine (Gibco) and 5% Horse Serum (Fisher, X) at 37°C in a humidified atmosphere with 5% CO₂. Cells were treated with vehicle (DMSO), TrkA inhibitor, K-252a, (100 mM, Sigma, 05288), gamma-secretase inhibitor, Compound E, (10 μM, Millipore, 565,790), alpha-secretase inhibitor, TAPI (25 μM, Sigma, SML0739) and 3 h later NGF (100 ng/mL, alomone, N-245) or BDNF (100 ng/mL, alomone, B-250) was added. Cell lysis or fixation was performed 24 h later. For immunoblotting, cells were plated in p6-wells at 80% confluence and analyzed by western blot (as described above). For filipin staining, cells were plated in coverslips previously treated with Poly-d-lysine (10 mg/mL). In both assays, cells were allowed to attach and 24 h later and then starved with serum-free medium for 2 h hours.

A total of 2 × 10⁴ cells/well PC12 were seeded onto a sterile slide placed in a 24-well plate and incubated with the different inhibitor (see above). After 24 h the cells were fixed by 4% PFA for 1 h, the cells were stained for 2 h with filipin (50 μg/mL, Sigma) in 10% of PBS and propidum iodide was used for nuclear staining. UV filter was required to view the filipin staining (340-380 nm excitation, 40 nm dichroic, 430-nm long pass filter). The cells were protected from light during the procedure because the filipin fluorescence photobleaches very rapidly. A confocal microscope SP8 (Leica) was used to scan and record the fluorescence. For filipin intensity quantification, regions of interest (ROIs) were defined thresholding total projections of each condition and signal intensity was measured using ImageJ software.

Endogenous p75^NTR-CTF was detected by immunoprecipitation. For that purpose, BF extracts were incubated with 500 μL TNE lysis buffer (50 mm Tris–HCl, pH 7.5, 150 mm NaCl, 1 mm EDTA, 0.1% SDS, 0.1% Triton X-100, 1 mm PMSF, 10 mm NaF, 1 mm Na₂VO₃, and protease inhibitor mixture) and disrupted by a dounce homogenizer. Samples were centrifuged and supernatant was subjected to immunoprecipitation. Extracts were incubated overnight with 1.5 μL of p75 antibody (Millipore, 07–476) at 4°C in an orbital shaker. The following day, 10 μL of previously TNE washed Protein G Agarose Beads (ABT, 4RRPG-5) were added and incubated for 2 h. Samples were then centrifuged to remove non-bounded proteins at 100 x g for 2 min to precipitate the agarose beads bound to the antibody, and washed three times in TNE lysis buffer with 0.2% Triton-X. Finally, for SDS-PAGE analysis, 30 μL reducing 2X sample buffer was added and the samples were boiled for 5 min at 96°C. Samples are centrifuged at 100 x g for 2 min and agarose beads are removed. Immunoprecipitated samples are then subjected to western blot analysis.

Mice were sacrificed with dislocation, and the BF was extracted. BF extracts were lysed for 30 min at 4°C in 200 μL of TNE. Protein concentrations were determined using the Bradford assay (Bio-Rad) and normalized across all samples. The lysate was subjected to centrifugation at 20,000 × g for 15 min at 4°C. The supernatant was removed, and 80–60 μg of protein were added to 4× SDS-sample buffer with 1.25% β-mercaptoethanol. Samples were denaturalized at 37°C for 15 min and resolved by SDS-PAGE. Proteins were transferred to nitrocellulose membranes, blocked with 5% BSA in TBS + 0.1% Tween (T-TBS) for 1 h and incubated overnight with the indicated antibodies. Primary antibodies used were the following: rabbit α-p75 (07–476, Millipore) 1:1000, mouse α- HMGCR (Abcam, ab242315) 1:1000, mouse α-LDLR (SantaCruz, sc-18823) 1:1000, mouse α-SREBP2 (SantaCruz, sc-13552) 1:1000, mouse α-actine (Sigma, A5441)1:5000. The membrane was incubated with the corresponding primary antibody diluted in T-TBS overnight at 4°C. For signal detection, secondary antibodies Alexa Fluor 800 donkey α-mouse (Invitrogen, A32789) 1:10000 and Alexa fluor 680 donkey α-rabbit (Invitrogen, A32802) 1:10000 were incubated for 1 h. After incubation with the appropriate secondary antibody, the membranes were imaged by LI-COR Odyssey scanner and quantified using ImageStudioLite (LI-COR).

All the statistical analysis were performed with GraphPad Prism software. The results are represented as mean ± standard error of the mean (SEM). The normal distribution of all data sets were confirmed with the D’Agostino & Pearson test. To determine if the differences between 2 groups were significant the unpaired Student’s t-test was performed. For data presented as a fold increase, the One-sample t test was employed. For multiple comparisons one or two-way analysis of variance (ANOVA) test was used. Initially, it was evaluated if there were significant differences between the groups, then the Tukey’s post-hoc test was used to determine the specific differences between groups. In the plots the “N” indicates the number of the independent mice used of each strain and age for each experiment. In all the analysis a p value <0.05 has been considered statistically significant, and represented as: *p < 0.05; **p < 0.01; *** p < 0.001 and **** p < 0.0001.