BNN27 attenuates AD-related pathologies in the hippocampus of the 5xFAD mouse model
We firstly investigated whether the 60-day administration of BNN27 was capable of reducing the Aβ load in the hippocampus of the 5xFAD mice which is the region mainly responsible for spatial working memory. Immunostaining analysis using the Aβ-specific antibody 6E10 showed a significant decrease in the number of Aβ-immunoreactive plaques load in the hippocampus of BNN27 treated mice 5xFAD mice compared to placebo (Fig. 1A, B). Specifically, there were 23.78 ± 2.702 Aβ-immunoreactive plaques in the hippocampal area of the placebo treated 5xFAD mice that were significantly reduced down to 13 ± 1.764 plaques (PStudent’s t test = 0.0042, t = 3.340, df = 16) in the hippocampi of the BNN27 treated animals (Fig. 1C). Noticeably, BNN27 did not restored the genotype-observed reduction in neurofilament levels on the entorhinal cortex (Supplementary Fig. S1, A, B), indicating that the duration of BNN27 administration and/or the age of the mice were not sufficient for maintaining neuronal survival on the neuronal plexus which innervate the hippocampus. The first projection of the hippocampus occurs between the entorhinal cortex (EC) and the dentate gyrus (DG). At the same time the entorhinal area is the first region, where the plaques and tangles are deposited in AD patients [78]. The EC was chosen as a region of interest due to its well described hippocampal innervation. We also examined the myelin condition in the same brain region, where there was no effect of the genotype or the treatment (Supplementary Fig. S1C, D).
Evaluation of the Aβ plaque load within the hippocampus of the 5xFAD animals [A, B] showed that the BNN27 treatment significantly decreased (PStudent;s t test = 0.0042) the formation of Αβ plaques in the area of the hippocampus [C] (n = 9). Immunostaining for astrogliosis (GFAP, top-panel) and microgliosis (IBA1, bottom panel) in the hippocampus of placebo and BNN27 treated animals [D]. There was an overall effect of the treatment [PTwo-way ANOVA <0.0001] and genotype [PTwo-way ANOVA <0.0001] in the astrogliosis and microgliosis. In the 5xFAD animal there is increased activation of the astrocytes [PBonferroni posttests<0.0001] and the microglia [PBonferroni posttests<0.0001] compare to the WT animals [E, F]. Moreover, in the 5xFAD animals BNN27 treatment significantly reduced astrogliosis [PBonferroni posttests<0.0001] and microgliosis in the hippocampus [PBonferroni posttests<0.0001] (n = 7–9). Immunostaining for Synapsin I in the CA2 mossy fibers of the hippocampus of placebo and BNN27 treated animals [G]. There was an overall effect of the treatment (PTwo-way ANOVA = 0.0138) in SYNAPSIN I. Moreover, in the 5xFAD animals BNN27 treatment significantly (PBonferroni posttests <0.05) increased SYNAPSIN I staining in the mossy fibers of the hippocampus (n = 6–9) [H]. Immunostaining for Synaptophysin in the CA2 mossy fibers of the hippocampus of placebo and BNN27 treated animals [I] There was an overall effect of the treatment (PTwo-way ANOVA = 0.0139) and genotype (PTwo-way ANOVA = 0.049) in SYNAPTOPHYSIN staining. Moreover, in the 5xFAD animals BNN27 treatment significantly (PBonferroni posttests <0.05) increased SYNAPTOPHYSIN staining in the mossy fibers of the hippocampus [J]. Graphs showing mean ± SΕΜ; n = 6–9 *P < 0.05, **P < 0.01, ***P < 0.0001, ****P < 0.0001.
Neuroinflammation is a hallmark of AD brain and is characterized by the presence of activated astrocytes and microglia [for review see: [13, 79] and similarly to other APP transgenic models [80, 81] the 5xFAD mice have elevated inflammation markers and gliosis associated with Aβ deposits commencing around two months of age [62]. We thus explored the effect of BNN27 on astrogliosis and microgliosis. Immunostaining analysis for glial activation markers GFAP (astrocytes) and IBA1 (microglia) (Fig. 1D) confirmed that by 3.5 months of age the 5xFAD mice have significantly increased astrocytic and microglial activation in the hippocampus (effect of genotype GFAP: PTwo-way ANOVA < 0.0001, F(1, 27) = 129.6; effect of genotype IBA1: P Two-way ANOVA < 0.0001, F(1, 27) = 65.96). Specifically, in the 5xFAD animals there is increased activation of the astrocytes measured by the percentage area of the hippocampus covered by GFAP positive signal (4.905 ± 0.2101 vs 14.43 ± 0.7730, PBonferroni posttests < 0.0001) and increased microgliosis measured by the percentage area of the hippocampus depicted by IBA1 positive signal (4.494 ± 0.3398 vs 11.75 ± 0.4982, PBonferroni posttests < 0.0001) in comparison with the WT animals (Fig. 1E, F). Interestingly, BNN27 treatment had a significant reducing effect on the extent of astrocytic and microglial activation in the hippocampus of the animals (overall effect of the treatment GFAP, PTwo-way ANOVA <0.0001, F(1, 27) = 58.29; overall effect of the treatment IBA1, PTwo-way ANOVA <0.0001, F(1, 27) = 34.50). Next, we compared the BNN27 treated and untreated (placebo) 5xFAD mice, showing that BNN27 significantly reduced both astrogliosis and microgliosis back to control (wt) levels. Specifically, in the BNN27 treated 5xFAD mice the percentage area of the hippocampus covered by GFAP positive signal (6.8585 ± 0.4484 vs 14.43 ± 0.7730, PBonferroni posttests <0.0001) and the percentage area of the hippocampus depicted by IBA1 positive signal (5.512 ± 0.465 vs 11.75 ± 0.4982, PBonferroni posttests < 0.0001) are reduced when compared to placebo 5xFAD mice (Fig. 1E, F). It should be mentioned here that our initial protocol for the in vivo testing of BNN27 in the 5xFAD mice was by administrating the compound for two months also but starting at 4 months old mice and isolating their brains at the age of 6 months. However, neither Aβ reduction or decrease of neuroinflammatory markers was observed, and thus we proceeded with the earlier administration, resembling to a prevention model.
Reactive astrocytes have been directly associated, both in vivo and in vitro, to neuroinflammation in such a way that reactive astrogliosis is considered a clear sign of neuroinflammation in the neurodegenerating brain. Just like microglia, astrocytes have two distinct activated states: the pro-inflammatory A1 reactive state and the anti-inflammatory A2 reactive state. In AD patients, the ratio between those two reactive states, especially surrounding senile plaques, is highly in favor of the neurotoxic A1 state. This observation is in accordance with the well characterized chronic inflammation in the brains of AD patients. To further evaluate the effect of BNN27 in microglia-astrocyte activity, we performed in vitro studies using isolated mouse microglia and astrocytes. As depicted in Supplementary Fig. S2, microglia cultures were treated with Aβ42 (10 μM) for 24h and their conditioned medium (MCM) was then transferred to naïve astrocytic cultures for another 24h before mRNA was collected. To determine the possible effects of BNN27, the compound (10-7M) was introduced to either microglia cultures at the same time with Aβ42 or to astrocytic cultures at the same time with MCM or to both cultures at the beginning of each treatment. Although, Aβ42 cannot directly activate astrocytes, we hypothesized that induced astrogliosis could be microglia mediated. For this reason, we performed an experiment where we use Aβ42 to activate microglia and after 24h we collect the conditioned medium and transfer it to naive astrocytic cultures. After 24h of treatment with this microglia-conditioned medium (MCM), we collect RNA and repeat the RT-PCR as previously. This time, as a positive control, we performed a 24h LPS (100ng/ml) treatment on microglia cultures and then transferred the MCM to naive astrocytes. Our results (Supplementary Fig. S2) reveal successful astrogliosis as both expression levels of Steap4 and Serpina3n markers were found significantly increased, while BNN27 had no effect on this astrogliosis phenotype, suggesting that factors involved in the anti-inflammatory effects of BNN27 are missing in the in vitro cultures. After having established our protocol for Aβ42-mediated astrocytic activation, we estimated the relative amount of A1 and A2 astrocytes in these cultures. For this reason, we selected specific target genes whose overexpression has been associated with A1 population respectively. For the A1 population, we have chosen Amigo2, Srgn and Serping1 genes. Polypeptide Aβ was able to induce the expression of these pro-inflammatory astrocytic markers while BNN27 reduced their levels, especially Srgn and Serping1. (Supplementary Fig. S2, B, C).
Oakley et al. (2006) [62] reported synaptic degeneration, assessed with whole-brain levels of the presynaptic marker synaptophysin as early as four months of age of 5xFAD mice. Given our observations that BNN27-treated animals show improved Aβ plaque pathology and reduced gliosis in the hippocampus, we hypothesized that these changes might be associated with reduced synaptic degeneration in the BNN27-treated brain. Therefore, we evaluated the synaptic button load using two different markers (Synaptophysin and Synapsin I) in the CA2 mossy fibers of the hippocampus of 5xFAD mice. Immunostaining analysis for Synaptophysin and Synapsin I (Fig. 1G, I) confirmed that four month of age 5xFAD mice show significant synaptic loss in the hippocampus (effect of genotype Synaptophysin: PTwo-way ANOVA = 0.0490, F(1, 21) = 4.365). Interestingly, BNN27 treatment significantly rescued synaptic degeneration in the hippocampus (overall effect of the treatment Synapsin I, PTwo-way ANOVA = 0.0054, F(1, 21) = 9.605; overall effect of the treatment Synaptophysisn, PTwo-way ANOVA = 0.0139, F(1, 27) = 7.197). BNN27 significantly reduced the synaptic loss, measured as percentage area of the CA2 mossy fibers of the hippocampus covered by Synaptophysin (16.68 ± 4.194 vs 46.34 ± 7.641, PBonferroni posttests = 0.0352) or Synapsin I positive signal (21.79 ± 2.109 vs 44.01 ± 6.704, PBonferroni posttests = 0.0433) when compared to the placebo 5xFAD mice (Fig. 1H, J).
BNN27 rescues the cholinergic atrophy at the basal forebrain of the 5xFAD animals
Cholinergic neurons, especially those in the basal forebrain, are known to be particularly susceptible in AD [5,6,7]. In the 5xFAD mice, in particular, the lesion of cholinergic fibers occurs earlier than the cholinergic neuron loss with the basal forebrain being the first brain region where cholinergic neuron loss is observed at the age of 9 months [82].
First, we aimed to evaluate the integrity of the cholinergic neurons at the basal forebrain of the 5xFAD mice using the choline acetyltransferase (ChAT) expression as a marker for identifying cholinergic neurons (Fig. 2A). Indeed, ChAT+ neurons in the BF of the 5xFAD mice have significantly smaller soma size when compared with the WT animals (67.9832 ± 5.722 vs 52.00 ± 2.8008, PBonferroni posttests = 0.0485) (Fig. 2B). The administration of BNN27 effectively restored the size of cholinergic neurons, indicating a protective role and potentially recovery of their functionality.
Representative image of TrkA immunostained cholinergic neurons (ChAT+) in the basal forebrain (BF) [A]. In the 5xFAD placebo mice there was a significant reduction in ChAT positive neurons soma size (P Bonferroni posttests = 0.0485) when compared with the WT placebo animals [B]. No change was observed in the percentage of double positive TrkA+/Chat+ neurons between genotypes or treatments [C]. Representative images of p75NTR immunostained cholinergic neurons (ChAT+) in the basal forebrain [D]. There was a significant increase in the percentage p75NTR +/ChAT+ neurons at the BF of the 5xFAD animals compared to the WT (PBonferroni posttests = 0.0317). Moreover, in the 5xFAD animals, BNN27 treatment significantly reduced the percentage p75NTR +/ChAT+ neurons (PBonferroni posttests = 0.0198) in the basal forebrain compared to 5xFAD placebo animals [E]. Graphs showing mean ± SΕΜ; n = 4 *P < 0.05, **P < 0.01, ***P < 0.0001, ****P < 0.0001.
We have previously shown that the neuroprotective and neurogenic effects of BNN27 are mediated by the TrkA and p75NTR receptors [56, 57]. Thus, we hypothesized that the expression of these receptors may be altered in the ChAT+ neurons of the BF. Interestingly, no change was observed in the expression of TrkA receptor in the ChAT+ neurons since the number of the double positibe TrkA+/ChAT+ neurons remained unaltered (Fig. 2C). However, the number of p75NTR+/ChAT+ neurons at the BF of the 5xFAD animals was significantly increased compared to the WT (PBonferroni posttests = 0.0317). Moreover, in the 5xFAD animals, BNN27 treatment significantly reduced the number p75NTR +/ChAT+ neurons (PBonferroni posttests = 0.0198) in the basal forebrain compared to 5xFAD placebo animals (Fig. 2D, E), indicating the p75NTR significance in disease progress.
Neuroprotective effect of BNN27 in hippocampal neuronal cultures against oligomeric Aβ-induced toxicity
Based on the neuroprotective effects of BNN27 in vivo in the hippocampal region of 5xFAD mice, we sought to investigate its effects on the neurotoxicity of Aβ, in isolated E17.5 mouse hippocampal neurons, expressing TrkA and p75NTR receptors (Supplementary Fig. S3). Due to well documented pro-survival effect of neurotrophins through Trk receptors, we investigated whether BNN27-induced activation of TrkA and/or p75NTR receptors is also neuroprotective against Aβ. The oligomeric neurotoxic Aβ42 allomorph was introduced in the culture, since predominate in the deposits of 5 μM Aβ42 in the human brain [83,84,85]. We used the In Situ Cell Death Detection Kit (TUNEL, Sigma-Aldrich) to measure cell death. The TUNEL reaction indicates DNA strand breaks that occur during apoptosis, distinguishing apoptosis from necrosis. To localize hippocampal neurons in the initial stage of neuronal differentiation we utilized the widely-used neuronal marker of the central and peripheral nervous system, Tuj1 (class III beta-tubulin).
The addition of Aβ for 48 h in the culture managed to increase the levels of TUNEL (+) 8DIV neuronal cells compared to the healthy complete nutrient medium (CTR) condition by a percentage of ≃40% (*P <0.01, Fig. 3A–C). More importantly, BNN27 showed a strong neuroprotective effect significantly reducing the Aβ-induced increase in the percentage of TUNEL (+) cells (*P <0.01).
A Illustrative graph that shows the ‘acute’ 48h administration of BNN27 and proneurotrophin in the E17.5 mature hippocampal neuronal culture and the initiation of the immunocytochemistry (ICC) experiment. B Representative images of immunofluorescence with TUNEL (green), Tuj1 (red), Hoechst (blue) conditions in the presence of Aβ vs the complete medium (CTR) condition. C Quantification of TUNEL (+) / Tuj1 (+) cells under 48 h of Aβ40 (5 μM) administration to E17.5 hippocampal neurons in the presence and absence of BNN27 (1-way ANOVA, Bonferroni multiple comparison test). p75NTR – not Trk receptors – seem to be implicated in the established effect. D Illustrative graph that shows the ‘acute’ 24h administration of BNN27 and proneurotrophin in the E17.5 mature hippocampal neuronal culture and the initiation of the immunocytochemistry (ICC) experiment. E Immunofluorescence with Cleaved Caspase-3 (green), Hoechst (blue) and [F] Tuj1 (green) conditions in the presence or absence of Aβ [Aβ40, 5 μM)] and/or p75NTR inhibitor [anti-p75 Receptor antibody (MC-192) Abcam], proBDNF and BNN27. G Quantification of TUNEL (+) / Tuj1 (+) cells under a 24 h administration of Aβ, p75NTR inhibitor treated as indicated at a 12DIV hippocampal neuronal cell culture (1-way ANOVA, Bonferroni multiple comparison test). H Quantification of TUNEL (+) / Tuj1 (+) cells under a 24 h administration of Aβ, panTrk inhibitor (AZD-1332, Alomone) treated as indicated at a 12DIV hippocampal neuronal cell culture (1-way ANOVA, Sidak’s multiple comparison test). I,J Immunofluorescence with Cleaved Caspase-3 (green), Hoechst (blue) and Tuj1 (green) conditions in the presence or absence of Aβ [Aβ40, 5 μM)] and/or panTrk inhibitor, NGF and BNN27. BNN27 affects the reciprocal extensive exposure to the Aβ40 toxicity at the E17.5 mature hippocampal neurons K Illustrative graph that shows the exact time periods for drug administrations and the immunocytochemistry (ICC) induction [L] Representative images of Cleaved Caspase-3 (red), Tuj1 (green), Hoechst (blue) staining of hippocampal 12DIV neurons after Aβ induction in the presence or absence of NGF, BNN27. The complete medium (CTR) condition serves as positive control. The images were taken at ×32 magnification (Scale bar, 20 μΜ). M Quantification of Cleaved Caspase-3 (+) / Tuj1 (+) after chronic administration (6 days) of NGF, BNN27 in order to test their ‘long term’ effects in the primary neuronal cell culture, with simultaneous injection of Aβ40 (5 μM) (1-way ANOVA followed by Sidak’s multiple comparison test) Data are expressed as mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001, n = 3; Scale bar = 20 μm.
To determine the potential effect of Aβ on more ‘mature’ neuronal cultures, illustrating the differences occurring throughout the development in in vitro neuronal cultures, hippocampal neurons were utilized in a later developmental stage of 12DIV. The development of hippocampal neurons in culture has been the subject of a remarkable number of studies. Thus, we know that 12DIV are regarded an adequate time interval to initiate the treatment, since nerve cell bodies have expanded in size, a complex and intertwined axonal network and sufficient dendrite network [86]. Cleaved Caspase-3 is considered a key factor of apoptotic proteolysis, serving as a marker of programmed cell death. Recent evidence implicates that this protease in involved in the pathogenesis of neurodegenerative diseases [87]. The introduction of Aβ42 in the culture of 12DIV hippocampal neurons increased the percentage of cleaved Caspase-3(+) / Tuj1(+) cells almost twice compared to the complete nutrient medium (CTR), an increase that was significantly reduced in the presence of BNN27 (*P<0.05) (Fig. 3D, G).
The p75NTR inhibitor was able to significantly decrease the levels of cleaved Caspase-3(+) / Tuj1(+) neurons (**P<0.01, Fig. 3F, G). It is of particular note, that the reducing effect of BNN27 on caspase-3 activation was obliterated by the presence of p75NTR inhibitor in the E17.5 hippocampal neuronal culture, suggesting that the p75NTR receptor was mediating its action [Fig. 3D–F]. The involvement of p75NTR receptor in the above effect of BNN27, was further supported by the stimulatory effects of Aβ, NGF or BNN27 on the p75NTR post-receptor TRAF6 signaling pathway (Supplementary Fig. S4).
To simulate the long-term effects of Aβ in the brain of patients with AD, we cultured 12DIV hippocampal cells with 5 μM Aβ42, for 6 days, adding every 2 days in the culture media the NGF, the proNGF, or BNN27. In ‘mature’ 12DIV cultured neurons, only less than 30% displayed apoptosis in basal conditions (Fig. 3K–M). In contrast, Aβ42 increased the percent of apoptotic cells nearly three-fold (***P<0.001). Interestingly, in agreement with our previous observations, BNN27 was able to alleviate the ‘chronic’ Aβ pathology in 6 days 12DIV hippocampal cell cultures (**P<0.01). These findings suggest that BNN27 can effectively control long-term Aβ neurotoxicity.
BNN27 promotes adult neurogenesis in the hippocampus of the 5xFAD mice
The dentate gyrus (DG) is a critical component of the hippocampal circuit involved in episodic and spatial memory [88]. The subgranular zone (SGZ) of the DG is one of the two areas that undergo neurogenesis in the adult brain [89, 90]. Given the well described neurotrophin actions of the BNN27 and its neuroprotective effects in the 5xFAD animals, we investigated whether BNN27 exerts also neurogenic effect in the SGZ of the DG. Initially, we evaluated the incorporation of BrdU in neurons (NeuN positive cells) located at the SGZ by measuring the number of BrdU+/NeuN double positive cells in the DG of the animals (Fig. 4A, B). Interestingly, there was an overall effect of the treatment (PTwo-way ANOVA = 0.0007, F(1, 13) = 19.68) suggesting that BNN27 significantly increases the number of BrdU+/ NeuN double positive cells. In addition, there was an overall effect of the genotype (PTwo-way ANOVA = 0.0011, F(1, 13) = 17.44) in the number of BrdU+/NeuN+ neurons suggesting that the 5xFAD mice have reduced neurogenic potential at the age of 4 months old. BNN27 treatment in the 5xFAD mice significantly increases the number of BrdU positive neurons in the DG of the hippocampus compared to 5xFAD placebo animals (456 ± 56.349 vs 858 ± 4.472. PBonferroni posttests = 0.0032) (Fig. 4C).
Immunostaining for BrdU and NeuN in the dentate gyrus of the hippocampus of placebo and BNN27 treated 5xFAD animals [A, B]. There was an overall effect of the treatment (PTwo-way ANOVA = 0.0007) and genotype (PTwo-way ANOVA = 0.0011) in the number of BrdU+/NeuN+ neurons. Moreover, in the 5xFAD animals BNN27 treatment significantly (PBonferroni posttests = 0.0032) increases the number of BrdU positive neurons in the hippocampus compared to 5xFAD placebo animals [C]. Immunostaining for DCX in the dentate gyrus of the hippocampus of placebo and BNN27 treated animals [D]. There was an overall effect of the treatment (PTwo-way ANOVA = 0.0097) and genotype (PTwo-way ANOVA =0.0007) in the number of DCX+ neurons [2-way ANOVA for treatment and genotype]. Moreover, in the 5xFAD animals BNN27 treatment significantly (PBonferroni posttests = 0.011) increases the number of newly born neurons hippocampus [E]. Graphs showing mean ± SΕΜ; n = 4 *P < 0.05, **P < 0.01, ***P < 0.0001, ****P < 0.0001.
The neurogenic effect of BNN27 was also evaluated by measuring the number of Doublecortin (DCX) positive neurons in the DG of the 5xFAD mouse hippocampus (Fig. 4D). DCX is a microtubule-associated protein expressed by neuronal precursor cells and immature neurons and a marker of newly born neurons. There was an overall effect of the treatment (PTwo-way ANOVA = 0.0097, F(1, 13) = 9.179) and genotype (PTwo-way ANOVA = 0.0007, F(1, 13) = 19.32) in the number of DCX+ neurons suggesting that BNN27 significantly increases the number DCX+ neurons in the 5xFAD animals which have impaired incorporation of DCX+ neurons in their DG under basal conditions. However, in the 5xFAD animals, BNN27 treatment significantly increases the number of newly born neurons in the DG of the hippocampus compared to the 5xFAD placebo animals (82.2 ± 5.38 vs 106.8 ± 5.737, PBonferroni posttests = 0.011) (Fig. 4E).
BNN27 protects the hippocampal Neural Stem Cells against oligomeric Aβ-induced toxicity
To further explore the neurogenic properties of BNN27, we used isolated adult (P7) Neural Stem Cells (NSCs) treated with the Aβ peptide (a 5 μM concentration of oligomeric Aβ42 peptide was utilized), in order to define compound’s effect on neural stem cell fate. The allomorph with the longest length Aβ42 tends to form higher molecular weight oligomers than the other isoforms, followed by conversion to protofibrils and mature fibrils, indicating that the accumulation of Aβ42 could predominate in the early stages of the Alzheimer’s Disease [83,84,85]. The toxicity of the Aβ was evaluated in a bilateral manner.
The effects of BNN27 on neural stem cell proliferation was firstly evaluated. Survival of the progeny of the dividing progenitor cells was assessed by staining for 5-Bromo-2’-deoxyuridine (BrdU) incorporation. NSCs progeny from the hippocampus were plated on poly-L-lysine and Laminin. Nestin is a well-established marker of neuronal progenitor cells in the adult brain [91]. A significant decrease by almost 30% of BrdU(+)/Nestin (+) cells was found in NSC cultures in the presence Aβ42 compared to cells in the Complete medium. However, neither the injected NGF nor BNN27 were able to restore the decreased BrdU (+)/Nestin (+) cell ratio (Fig. 5A, B).
There was no detectable difference on the percentage of the BrdU(+) cells given the presence of BNN27 in the P7 Aβ-introduced NSC culture. A BrdU incorporation (green) evaluation of the Nestin(+) P7 NSCs (red). B Quantification of BrdU(+)/Nestin(+) cells under 48h induction of Aβ at the the P7 NSC hippocampal culture, in the presence or absence of NGF and BNN27 treatment, respectively. The complete condition depicts the positive control [1-way ANOVA for treatment and genotype, Sidak’s multiple comparisons Post Hoc tests]. C Staining with CellTox Green Cytotoxicity Assay. D Evaluation of the in vitro cytotoxicity of Aβ42 oligomer as well as the potential 48 h neuroprotective activity of BDNF neurotrophin and BNN27 in the P7 NSCs of hippocampus. The complete medium (CTR) condition serves as positive control. [1-way ANOVA for treatment and genotype, Sidak’s multiple comparisons Post Hoc tests]. Both Trk receptors and p75NTR seem to be implicated in the observed result. E Immunofluorescence with CellTox (green) and Hoechst (blue) conditions that shows the 24h administration in the P7 hippocampal neural stem cell culture in the presence or absence of Aβ42 (5 μM) and/or panTrk inhibitor (AZD-1332, Alomone), NGF or BNN27. F Quantification of CellTox (+) / Hoechst (+) cells under a 24 h administration of Aβ and/or panTrk inhibitor, treated as indicated at a P7 hippocampal neural stem cell culture (1-way ANOVA, Tukey multiple comparison test). G Immunofluorescence with CellTox (green) and Hoechst (blue) conditions that shows the 24h administration in the P7 hippocampal neural stem cell culture in the presence or absence of Aβ42 (5 μM) and/or p75NTR inhibitor [anti-p75 Receptor antibody (MC-192) Abcam], proBDNF or BNN27. H Quantification of CellTox (+) / Hoechst (+) cells under a 24 h presence or absence of Aβ and/or p75NTR inhibitor, treated as indicated at a P7 hippocampal neural stem cell culture (1-way ANOVA, Tukey multiple comparison test). Data are expressed as mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001, n = 3; Scale bar = 20 μm).
In order to evaluate the cell protective effects of BNN27 in NSCs, hippocampal P7 neurospheres were treated with the BNN27, the Brain-Derived Neurotrophic Factor (BDNF) or the pro-BDNF in the presence of Aβ42. A 48-h exposure of NSCs in oligomeric Aβ42 allele is sufficient to increase the levels of cytotoxicity in the culture (*** P<0.0001, Fig. 5C, D), as detected by the CellTox Green Cytotoxicity Assay (Real-time Cell Death Assay, Promega). BNN27 was effective to protect against Aβ toxicity (**P<0.05), as well as BDNF (** P<0.05) (Fig. 5C, D).
To evaluate further the implication of the neurotrophin receptors in the aforementioned result, the experimental procedure was recapitulated only this time in the presence of panTrk (AZD-1332, Alomone) and p75NTR [anti-p75 Receptor antibody (MC-192), Abcam] inhibitors. Interestingly, the established in vitro neuroprotective effect of BNN27 compound was abolished in the presence of whether panTrk or p75NTR inhibitor (* P<0.01). In particular, BNN27-induced activation of Trk receptors and p75NTR receptors – both – were shown to significantly reduce in vitro the Aβ oligomer-injected cytotoxicity (Fig. 5E–H).
In order to explore a potential direct neurotrophic effect of BNN27 in adult (P7) NSCs, we exposed cultures of NSCs to BNN27 and a low concentration of neurotrophic EGF in the absence of FGF, for 24 and 48 h. BNN27 significantly increased the proliferation of P7 hippocampal NSCs within 24h, even in reduced levels of EGF (5 ng/ml, absence of FGF) (Supplementary Fig. S4, C, D), suggesting that BNN27 might serve as a sufficient trophic factor under conditions where the endogenous trophic support is limited. However, at 48 h, BNN27 was able to induce the proliferation of P7 hippocampal NSC only with reduced concentration of EGF and not under a complete deprivation of growth factors (Supplementary Fig. S5, A–F).
Furthermore, we tested the neurogenic effects of BNN27 on the other neuronal population of proliferating cells, the Oligodendrocyte Precursor Cells (OPCs). Upon isolation of OPCs from P3 pups, we cultured them in proliferation medium (containing PDGF, FGF) for two days, then cultures were exposed to 1 μΜ BNN27 for 48hrs. Subsequently, a 24hr BrdU pulse was performed. Co-staining for PDGFR and BrdU showed that BNN27 significantly promoted OPC proliferation, increasing the number of PDGFR+/BrdU+ cells by ~20% (Supplementary Fig. S6a), establishing BNN27 as a strong neurogenic agent. Furthermore, we also examined the BNN27 effect on oligodendrogenesis by evaluating OPCs maturation. OPCs were cultured for another 48hrs in differentiation medium (containing T3) in the presence of amyloid beta 1-42 (5 μΜ). Immunofluorescence analysis showed that BNN27 was capable of increasing O4+ cells as well as NGF did under neurodegenerative conditions (Supplementary Fig. S6b).
BNN27 reduces spatial memory deficits in the 5xFAD mice
It is well documented that the 5xFAD mice by 4–5 months of age have significantly impaired spontaneous alternation performance which translates into spatial working memory deficits [62]. Based on the aforementioned cellular effects of BNN27 in the 5xFAD brain, we investigated whether BNN27 has a protective effect in hippocampal-dependent spatial working memory too. We assessed spontaneous alternation performance in the T-maze in mice, receiving for 2 months either BNN27 or placebo (introduced as subdermal pellets steadily releasing 10mg/kg/day, Fig. 6A). The T-maze used was an elevated apparatus in the form of a T placed horizontally (Fig. 6B) (see “Methods” part for protocol’s details). This learning task does not involve any training, reward, or punishment and allowed us to assess hippocampus-dependent spatial working memory.
A Graphic representation of the T-maze apparatus. B Working memory was evaluate by measuring the spontaneous alternation of the mice it the apparatus. It was shown that there is an overall significant effect of the treatment [P Two-way ANOVA = 0.0007] and genotype [P Two-way ANOVA = 0.0408] in the spontaneous alternation performance of the mice. Moreover, in the 5xFAD animals BNN27 treatment significantly improved working memory performance (P Bonferroni posttests < 0.01) [Graph showing mean ± SΕΜ; n = 7–9; **P < 0.01).
BNN27 and placebo-treated for 60 days 5xFAD mice as well as age-matched wild type control mice had their spontaneous alternation performance evaluated. Two-way ANOVA analysis, with the genotype and the treatment as the two independent variables, showed that there is an overall significant effect of the treatment (Ptwo-way ANOVA = 0.0007, F(1, 27) = 14.71) and genotype (Ptwo-way ANOVA = 0.0408, F(1, 27) = 4.616) in the spontaneous alternation performance of the mice, suggesting that 5xFAD mice perform significantly poorer than the wild type mice and that BNN27 significantly improves the performance of the 5xFAD mice. Moreover, post hoc analysis using Bonferroni’s multiple comparisons test showed that in the BNN27-treated 5xFAD animals there is a significant improvement in spontaneous alteration performance compared to the placebo-treated 5xFAD animals (0.643 ± 0.025 vs 0.4625 ± 0.026, PBonferroni posttests = 0.0027). This behavioral analysis fully confirmed that the neurorestorative and cell protecting effects of BNN27 are sufficient to overcome AD-derived pathology and may represent a new therapeutic approach against multifactorial neurodegenerative diseases, like AD.
BNN27 has global rescue effects on Aβ induced negative impact
We finally assessed whether there were any significant changes in the proteomes of BNN27-treated 5xFAD mice, comparing the protein expression profile of these animals to untrated 5xFAD and WT mice, using mass spectrometry analysis. We performed direct DIA analysis, considering a peptide as ‘detected’ if the confidence score (for identification) was ≤0.01. Proteomics analysis included 3200 proteins and 16.000 peptides per sample were identified and quantified (Supplementary Fig. S7a). The Coefficient of Variation was between 12 and 16 percent for the different groups and the peptide intensity distributions revealed high quality across all samples and low missing values, meaning that each identified peptide in a sample overlaps with other samples in the sample dataset (Supplementary Fig. S7b). The fold change of all peptides in a sample was compared to their mean value over all samples in the group. Any loading differences were corrected after normalization which visualized how strongly each sample deviates from other samples in the same group that indicated the identification and removal of one outlier sample (WT placebo), according to the mass spec run (Supplementary Fig. S7c). In order to find the significant regulated proteins, we set the q-value threshold at 0.05 and the log2 fold change threshold ≤ − 0.2 / ≥ + 0.2 and these filters were used for statistical analysis.
Aβ induces major changes that reflect the AD pathology
In the comparison between the 5xFAD and WT both treated with placebo mice group, 209 significantly dys-regulated proteins found, with 127 proteins up-regulated and 82 proteins down-regulated in the 5xFAD group. The volcano plot indicated that Aβ-induced proteins like App, Apoe, Hexb, Bace1, Ncstn were significantly up-regulated in the 5xFAD mice (Fig. 7a, Supplementary Table S1) [92,93,94].
a Significantly regulated proteins in 5xFAD/WT comparison (FDR≤ 0.05, log2 fold change ≤ −0.2 or ≥0.2, pink: up-regulated, blue: down-regulated proteins), [b] protein group analysis on significantly down-regulated and [c] on significantly up-regulated proteins in the 5xFAD/WT comparison (enrichment analysis using the whole mouse genome as background).
In order to get a better understanding towards the changes, a g:profiler enrichment analysis was performed using the whole mouse genome as background, for both the down-regulated and the up-regulated proteins respectively (http://biit.cs.ut.ee/gprofiler/).
The enrichment analysis of down-regulated proteins showed among others, proteins implicated in the synaptic signaling as well as the glutamatergic synapse. Moreover, proteins related to neurogenesis (Lingo1, L1cam, Stxbp1) and neuron projection (Abl1, Nectin1, Gap43) were significantly down-regulated [95, 96]. Another pathway in which proteins were significantly changed was retrograde endocannabinoid signaling pathway (Gabra5, Ndufa3). Finally, molecular functions affected, were those of protein binding and calcium-dependent phospholipid binding (Syt5, Syt7, Syt17) (Fig. 7b) [97].
On the other hand, the enrichment analysis of up-regulated proteins showed changes in the complement pathway activation (e.g. C1qa, C1qb and C1qc), cholesterol biosynthesis (Nsdhl, Hmgcs1) [98], as well as, the innate immune system (Cstd, Aldoc) and neurotrophil degranulation (Cst3, Arl8a) [99]. Moreover up-regulation was observed on lysosomal (Hexb, Hexa, Lamp2), metabolic pathway (Pld3, Idi1), mitochondrion (Ank2, Ncstn), and synapse (negatively) related proteins (Bace1, Syt11). Close to that, significant up-regulation was observed in proteins indicating neuronal death (Nefl, Stat3) and proteins involved in Aβ binding and regulation of amyloid fibril formation (Apoe, App, Clu, Itm2c, Itm2b, Bace1) subsisting Alzheimer’s disease. Proteins associated with pseudobulbar signs (Htra1, Sod1) and cerebral amyloid angiopathy (e.g. IItm2b, Cst3) in the human brain, were also differentially regulated in our mouse disease-like genotype (Fig. 7c). Our proteomics data therefore validates major changes that reflecting AD pathology in the 5xFAD mice.
BNN27 has global rescue effects on Aβ induced negative impact in 5xFAD mice
To assess BNN27 administration effects, we compared the BNN27 treated 5xFAD mice and placebo treated WT mice. 277 proteins were significantly dys-regulated in the 5xFAD+BNN27/WT comparison from which 149 significantly regulated only in 5xFAD+BNN27/WT and 128 overlapped in both comparisons (5xFAD+BNN27/WT and 5xFAD/WT). In addition, 81 proteins were significantly dys-regulated only in the 5xFAD/WT comparison and not in the 5xFAD+BNN27/WT (Fig. 8a, Supplementary Table S2). Many of the shared proteins include the typical AD-associated proteins such as App, Clu, Apoe, Gfap, Hexb etc, which implicates that the administration of BNN27 is not capable of restoring the 5xFAD mice to the wild type level. Nevertheless, BNN27 appeared to cause a lesser up-regulation of these AD-associated proteins.
a Venn diagram showing the common proteins that are present in both the 5xFAD+BNN27/WT and 5xFAD/WT comparisons, [b] double volcano blot showing the shift in log2 fold change induced by BNN27 (circle: 5xFAD+BNN27/WT, triangle:5xFAD/WT), [c] all rescued down- (≥10%) and [d] all up- (≤ −10%) rescued regulated proteins in 5xFAD+BNN27/WT compared with 5xFAD/WT shows the rescue effect in fold changes of major AD-induced proteins, in the presence of BNN27 (increase/decrease percentage in log2 fold change of proteins significantly regulated in both sets FDR ≤ 0.05, log2 fold change ≤ −0.2 or ≥0.2).
To reveal the probable partial rescue effect of BNN27, a double volcano plot was created (Fig. 8b) to reveal the “shift” in the log2 fold change of the significantly-changed proteins between the two comparisons. This “shift” between the log2 fold changes reveals a partial rescue if the log2 fold change is closer to zero after the treatment with BNN27 ((log2fc(5xFAD+BNN27/WT)) as it reveals less difference in the protein expression between the two groups (5xFAD+BNN27 and WT)” and is quantified by calculating the percentage of the change between the log2 fold changes as follows:
$$[(log 2{fc} (5{xFAD}+{BNN}27/{WT})-log 2{fc}(5{xFAD}/{WT})){{{rm{ / }}}} left(right. log 2{fc}(5{xFAD}/{WT})]* %$$
BNN27 treatment leads to the rescue of 16 (out of 47) down-regulated proteins by more than 10%. These proteins are implicated in the ATP-dependent protein binding (Dnajc5, Prnp) [100], the trans-synaptic signaling (Bcr, Cd38, Dagla, Gabra5, Prnp, Slc30a1), response to stress, regulation of biological quality and the synapse (Adam11, Bcr, Cd38, Dagla, Dele1, Gabra5, Gap43, Prkca, Prnp, Slc30a1, Gpc1). In addition, rescued were proteins of the somatodentritic compartment (Adama11, Bcr, Dagla, Gabra4, Gap43, Gpc1, Prnp), the cytoplasm (Nudt3, Gnb4, Gap43, Dele1), the organelle membrane (Atp85, Cd38, Dagla, Dele1, Dnajc5, Ndufs1, Prkca, Prnp, Scl30a1) [101] and the cell body (Adam11, Gabra5, Gap43, Gnb4, Gpc1) (Fig. 8c, Supplementary Table S3).
Furthermore, BNN27 treatment rescued 38 (out of 81) up-regulated proteins by leading to a decrease in the log2 fold change in the comparison 5xFAD+BNN27/WT by more than 10% compared to 5xFAD/WT mice log2 fold changes. More precisely, the rescued proteins are related to Alzheimer’s disease (Apoe, Bace1, App), Aβ and amide binding (Apoe, Clu, Bace1, Itm2c, Itm2b, Clstn1), oxidative damage response and complement activation (C1qc, C1qb, C1qa), N-acetyl-beta-D-galactosaminidase and beta-N-acetylhexosaminidase activity (Hexb, Hexa), amyloid precursor protein metabolic process (Apoe, Itm2b, Clu, Bace1, Itm2c, APP), microglial and astrocytic cell activation, as well as neuro-inflammatory (C1qa, Syt11, Clu, App, Gfap), innate immune response (Actr10, Lamp2, Hexb, C1qc, C1qb, C1qa, CTsd, Ctsz), chaperone-mediated autophagy (Gfap, Lamp2, Clu) [102] and autophagy (Arl8b, Gfap, Ctsd, Atg9a, Lamp2, Syt11, Clu), synapse pruning (C1qc, C1qb, C1qa), as well as postsynaptic proteins (Sptbn1, C1qc, C1qb, C1qa, Clstn1, Syt11, App), proteins of the extracellular matrix (Spock2, Htra1, Clu, Ctsz, Ctsd) (Fig. 8d). The most rescued protein was Sptbn1 that is implicated in synapse, axon, cell projection (as well as Gfap), neuron projection and post-synapse cellular component, together with more rescued proteins [103]. In addition, there were some AD-associated proteins that show minor rescue levels, for example, Ncstn (-6%). Finally, in both the down- and the up-regulated proteins, some didn’t change or dys-regulated further more by BNN27 (Supplementary Table S3).
- SEO Powered Content & PR Distribution. Get Amplified Today.
- PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
- PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
- PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
- PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
- Source: https://www.nature.com/articles/s41380-024-02833-w