Brain p3‐Alcβ peptide restores neuronal viability impaired by Alzheimer's amyloid β‐peptide

Abstract We propose a new therapeutic strategy for Alzheimer's disease (AD). Brain peptide p3‐Alcβ37 is generated from the neuronal protein alcadein β through cleavage of γ‐secretase, similar to the generation of amyloid β (Aβ) derived from Aβ‐protein precursor/APP. Neurotoxicity by Aβ oligomers (Aβo) is the prime cause prior to the loss of brain function in AD. We found that p3‐Alcβ37 and its shorter peptide p3‐Alcβ9‐19 enhanced the mitochondrial activity of neurons and protected neurons against Aβo‐induced toxicity. This is due to the suppression of the Aβo‐mediated excessive Ca2+ influx into neurons by p3‐Alcβ. Successful transfer of p3‐Alcβ9‐19 into the brain following peripheral administration improved the mitochondrial viability in the brain of AD mice model, in which the mitochondrial activity is attenuated by increasing the neurotoxic human Aβ42 burden, as revealed through brain PET imaging to monitor mitochondrial function. Because mitochondrial dysfunction is common in the brain of AD patients alongside increased Aβ and reduced p3‐Alcβ37 levels, the administration of p3‐Alcβ9‐19 may be a promising treatment for restoring, protecting, and promoting brain functions in patients with AD.


Introduction
Alzheimer's disease (AD) is an incurable neurodegenerative disorder that causes progressive dementia in aged subjects. More than 50 million people live with dementia worldwide and AD accounts for 60-80% of these cases (World Alzheimer's Report, 2018). The generation and accumulation of neurotoxic amyloid b (Ab) protein, a pathogenic hallmark in the AD brain, are believed to be the primary cause of neurodegeneration, resulting in cognitive deficits and memory loss (McLean et al, 1999;Hardy & Selkoe, 2002). In neurons, Ab is generated from Ab-protein precursor (APP), a single membranespanning protein, via intramembrane proteolysis of c-secretase after the shedding of the APP extracellular/luminal region by bsecretase/BACE1 (De Strooper et al, 1998;Vassar et al, 1999;Thinakaran & Koo, 2008).
Our results show a novel neuroprotective function of p3-Alcb, which allowed us to propose that the administration of p3-Alcb peptides to individuals in the AD continuum protects neurons by increasing p3-Alcb levels in the brain and consequently is effective in slowing the progression of the disease.

Results
Human CSF p3-Alcb37 level decreases significantly in the early stage of AD patients We previously reported that the p3-Alcb levels in CSF significantly reduced in patients who were clinically diagnosed with AD (Hata et al, 2019). To further investigate the relationship between AD progression and p3-Alcb levels, we examined p3-Alcb37 levels in CSF of patients who had been classified into AD biomarker categories (Jack et al, 2018). Subjects along the AD continuum (n = 131) were categorized according to biomarker-based criteria (Kasuga et al, 2022), and CSF p3-Alcb37 values were quantified (Fig 1) (Appendix Table S1). Interestingly, the level of p3-Alcb37 in AD patients at an early stage (A+TÀNÀ) was found to be significantly lower (P < 0.0001) than that in subjects (AÀ). The finding raises a hypothesis that a decrease in the level of brain p3-Alcb is pathophysiologically important in the development of AD. Hence, we proceeded to analyze the molecular characterization of p3-Alcb and its physiological functions on neurons and the brain milieu.

p3-Alcb increases neuronal viability and protects neurons against Ab42 oligomer-induced toxicity
The p3-Alca and p3-Alcb peptides are generated through cleavage processing of their precursor molecules, Alca and Alcb, by aand csecretases, whereas Ab is derived from the sequential cleavage of APP by band c-secretases (Fig 2A) (Hata et al, 2009). We first examined whether the p3-Alc peptides would possess aggregation properties as does Ab. Synthetic peptides, Ab42, p3-Alca35, and p3-Alcb37 (the amino acid sequences of the peptides are shown in Fig EV1A), were dissolved in PBS and incubated at 37°C for 24 h (+) or not (À), and the samples were analyzed by immunoblotting with specific antibodies (Fig EV1B). Unlike the aggregation/oligomer formation of Ab42, p3-Alca35, and p3-Alcb37 did not aggregate. The oligomerization of p3-Alcb37 was examined by size-exclusion chromatography ( Fig EV1C). The p3-Alcb37 peptide (50 lM in PBS), which was incubated at 37°C for 24 h, was eluted with a monomer retention time of 5.5 min and yielded an identical recovery of peptides at the indicated time points of 0, 2, 4, and 24 h. These findings indicate that p3-Alcb37 does not form oligomers that are labile to SDS electrophoresis. Ab42 aggregation is easily monitored with the Thioflavin T fluorescence assay (Walsh et al, 1999). The fluorescence intensity in the presence of Ab42 (10 lM) increased linearly with the incubation time, whereas the fluorescence intensities in the presence of p3-Alca35 (10 lM) and p3-Alcb37 (10 lM) were below the threshold and did not increase until the incubation time point of 24 h ( Fig EV1D). These results indicate that p3-Alc peptides are not prone to aggregation. Furthermore, in an in vitro study with Thioflavin T, p3-Alcb did not suppress Ab42 aggregation effectively and specifically ( Fig EV1E). Hence, it is likely that p3-Alcb is a nonaggregation peptide that does not act directly on Ab aggregation.
Next, we assessed the effects of p3-Alc on neurons and whether p3-Alc peptides are neurotoxic in the same way as Ab. Mouse primary neurons were cultured for 24 h in the presence of p3-Alca35, or p3-Alcb37, and their cell viability was then examined with MTT assays (Fig EV1F). Interestingly, neurons showed significantly better viability (P < 0.01) in the presence of p3-Alcb37 than in the absence of p3-Alcb37. This effect was not observed in presence of p3-Alca35, and the amino acid sequence of p3-Alcb37 is distinct from that of p3-Alca35 ( Fig EV1A) (Hata et al, 2009), indicating that the effect is p3-Alcb-specific.

p3-Alcb inhibits anomalous Ab42 oligomer-induced Ca 2+ influx in neurons
Abo promotes neuronal dysfunction by triggering excessive Ca 2+ entry into neurons (Hardingham & Bading, 2010;Benilova et al, 2012). To determine how p3-Alcb restores neuronal function and protects neurons against Abo-induced toxicity, we examined Ca 2+ influx, with a labeled calcium indicator Fluo 4-AM, into mouse primary neurons in culture. Abo increased Ca 2+ influx dramatically, and this increase was significantly inhibited in the presence of p3-Alcb9-19 (P < 0.01) or p3-Alcb37 (P < 0.05) (Figs 3A and B,and EV2A). To uncover the role of p3-Alcb on Ca 2+ influx into neurons, mouse primary neurons were first incubated in a Ca 2+ -depleted medium to reduce their excitability. CaCl 2 (2 mM), at a concentration equivalent to that found in CSF (Forsberg et al, 2019), was then added to the culture medium, and intracellular Ca 2+ levels were recorded in the presence or absence of p3-Alcb9-19 or p3-Alcb37 (Fig 3C and D). Intracellular Ca 2+ levels rapidly increased after Ca 2+ administration, but this response was dose-dependently suppressed by the presence of p3-Alcb9-19. This finding was also observed in the presence of p3-Alcb37, indicating that p3-Alcb attenuates Ca 2+ influx.
Data information: Experimental numbers indicate biological replicates (subject numbers). Detailed information including the statistical summary is described in Dataset EV1. Source data are available online for this figure.
Ó 2023 The Authors EMBO Molecular Medicine 15: e17052 | 2023 incubated with the designated antibodies (red) and streptavidin-Alexa488 (green) to detect p3-Alcb. Pearson's R-values revealed that p3-Alcb colocalized with neuronal proteins (Fig EV3B and C). The postsynaptic glutamate receptor subunit GluN2B colocalized better with p3-Alcb than with presynaptic synaptophysin (SYP) or cytoplasmic bIII-tubulin. These results suggest that p3-Alcb acts on a membrane protein(s) in the postsynaptic region in neurons.
p3-Alcb suppresses the function of NMDARs, which are anomalously activated for Ca 2+ influx by Ab42 oligomers Abo preferentially targets neuronal glutamate receptors. The stimulation of postsynaptic N-methyl-D-aspartate receptors (NMDARs) by Abo induces severe neuronal degeneration via intracellular Ca 2+ overload (Zhang et al, 2007;Hardingham & Bading, 2010), which is more detrimental to neurons than Ca 2+ influx mediated by other channels (Tymianski et al, 1993). In mouse primary cultured neurons, Abo-induced Ca 2+ entry increased and was suppressed in the presence of the NMDA antagonist D-AP5 ( Fig 4A). This finding reiterates previous data that Abo-mediated neurotoxicity is largely mediated by NMDARs (Zhang et al, 2007;Hardingham & Bading, 2010  A Schematic diagram of Ab and p3-Alcb generation from their precursors APP and Alcb. The cleavage site by secretases: a, a-secretase; b, b-secretase; c, c-secretase. B The amino acid sequence of p3-Alcb37 in humans, chimpanzees, mice, rats, and dogs. The amino acid sequence of p3-Alcb9-19 is shown in red letters for humans, and gray boxes indicate an amino acid that is different in nonprimates. C Effect of p3-Alcb37 and its partial peptides on the viability of mouse primary cultured neurons. Wild-type (WT) neurons (div 15-20) were incubated for 24 h in the presence (10 lM) or absence (À) of p3-Alcb1-11, p3-Alcb9-19, p3-Alcb20-37, and p3-Alcb1-37. Neuronal viability was evaluated with MTT assays and expressed relative to that of cells cultured in the absence of peptides (assigned a value of 1.0). Statistical analysis was performed using a one-way ANOVA, followed by the Dunnett's multiple comparisons test (mean AE SEM; n = 10-12), and significant P-values (P < 0.01, P < 0.0001) are indicated on the graph. D-F Effect of p3-Alcb on the neuronal toxicity of Ab42 oligomers. WT neurons (div 15-20) were incubated for 24 h in the presence (10 lM) or absence (À) of p3-Alcb9-19 and p3-Alcb1-37 with (+) or without (À) Ab42 oligomers (Abo, 2.5 lM). Neuronal viability was evaluated by MTT (D), ATP generation (E), and LDH release (F) assays and expressed relative to neurons cultured in the absence of p3-Alcb and Abo (assigned a value of 1.0). Statistical significance was determined by Dunnett's multiple comparisons tests (mean AE SEM; MTT: n = 36, ATP: n = 40, LDH: n = 60). The significant P-values (P < 0.05, P < 0.01, P < 0.001, P < 0.0001) versus cells incubated in the absence (À) of p3-Alcb and Abo are indicated in the graphs. The significant P-values ( # P < 0.05, # P < 0.01) versus cells incubated in the presence (+) of Abo and in the absence (À) of p3-Alcb are indicated in the graphs. G Effect of Ab42 oligomers (Abo) and p3-Alcb peptides on the generation of reactive oxygen species (ROS). WT neurons (div 15-17) were incubated for 24 h with (+) or without (À) Abo (2.5 lM) and in the presence (+) or absence (À) of p3-Alcb (10 lM). ROS generation was assessed and expressed relative to that of neurons incubated in the absence (À) of Αbo, which was assigned a value of 1.0. Statistical significance between samples with or without Abo was evaluated with a Student's t-test (mean AE SEM; n = 10), and the significant P-value (P < 0.05) is indicated on the graph.
Data information: Experimental numbers indicate biological replicates. Detailed information including the statistical summary is described in Dataset EV1. Source data are available online for this figure.
In fact, NMDA administration resulted in a very smaller Ca 2+ influx (See the difference between the magnitude of the vertical axis in panel B and that in panel A) into neurons in the absence of p3-Alcb9-19 than into neurons in the presence of p3-Alcb9-19 ( Fig 4B). This indicated that most NMDARs are largely inactive in mouse primary neurons and that p3-Alcb regulates Ca 2+ influx weakly even in the presence of NMDA. This suggests that p3-Alcb does not directly act on NMDARs, which remain mostly silent in the absence of Abo but regulates Ca 2+ influx by targeting other molecules in a milder manner than when NMDARs are activated by Abo.
To determine whether p3-Alcb targets protein(s) other than NMDARs to control Ca 2+ influx, mouse primary cultured neurons were first incubated in a Ca 2+ -depleted medium to reduce their excitability and were then administered CaCl 2 (2 mM) in the presence or absence of D-AP5 and p3-Alcb ( Fig 4C). Ca 2+ influx was slightly, but not significantly, reduced by D-AP5 (5, 50, and 200 lM), indicating that NMDARs were not main contributors to Ca 2+ influx through extracellular Ca 2+ administration in mouse primary neurons (compare experiment numbers/No. 2 to No. 4 with No. 1 in Fig 4C). Interestingly, this Ca 2+ influx was significantly inhibited by p3-Alcb9-19 and p3-Alcb37 (P < 0.001) (compare No. 5 and No. 6 with No. 1). Furthermore, when D-AP5 (200 lM) was combined with p3-Alcb9-19 or p3-Alcb37, the intracellular Ca 2+ levels were the same as that of p3-Alcb alone (compare No. 7 and No. 8 with No. 5 and No. 6). These results clearly show that p3-Alcb does not target NMDARs directly but regulates Ca 2+ influx through a different molecular mechanism.
It is intriguing if p3-Alcb is capable of regulating a significant amount of Ca 2+ influx under the Abo-induced NMDARs activation condition. To this aim, we further investigated whether the unique A, B Suppression of Ca 2+ influx induced by Abo in neurons treated with p3-Alcb9-19 (A) and p3-Alcb37 (B). Mouse neurons (div 11-13) pretreated with Fluo 4-AM were stimulated at 5 min (arrow) with (+) or without (À) Abo (5.2 lM) in the presence (+) or absence (À) of p3-Alcb (50 lM). The fluorescence intensity was recorded at the indicated time (left) and the fluorescence area intensity for 18 min (5-23 min at time points) is shown (right) as the AUC expressed relative to that of cells cultured in the absence (À) of p3-Alcb and Abo (assigned a value of 1.0). C, D Suppression of Ca 2+ influx into neurons followed by Ca 2+ administration in the presence of p3-Alcb9-19 (C) and p3-Alcb37 (D). Mouse neurons (div 11-13) pretreated with Fluo 4-AM in the calcium-depleted medium were administered Ca 2+ (2 mM) at 5 min (arrow) in the presence or absence of p3-Alcb9-19 (C) and p3-Alcb37 (D). The fluorescence intensity was recorded at the indicated time (left) and the fluorescence area intensity for 7.5 or 9 min (5 to 12.5 min in panel (C) and 5 to 14 min in panel (D) at the indicated time points) is shown (right). Statistical significance was determined by one-way ANOVA with the Tukey's multiple comparison test (mean AE SE; n = 14 (A), n = 15 (B), n = 10 (C)). The significant P-values (P < 0.05, P < 0.01, P < 0.001, P < 0.0001) are indicated on the graphs. Statistical significance was determined with a Student's t-test (mean AE SEM; n = 11 (D)), and the significant P-value (P < 0.001) is indicated on the graph.
Data information: Experimental numbers indicate biological replicates. Detailed information including the statistical summary is described in Dataset EV1. Source data are available online for this figure.
Peripheral administration of p3-Alcb restores neuronal viability impaired by Ab accumulation in the brain of the AD mouse model Next, we examined the ability of p3-Alcb9-19 to promote and preserve neuronal health against Ab toxicity in vivo. Rodents were peripherally administered p3-Alcb9-19 and neuronal viabilities were then analyzed with positron emission tomography (PET) imaging. In a separate study in mice, successful transfer of p3-Alcb9-19 into the brain following subcutaneous injection was confirmed by sELISA by quantifying p3-Alcb9-19 levels in the CSF (Fig EV4A-C). An increase in neuronal viability caused by p3-Alcb was confirmed in vivo by monitoring brain mitochondrial function using PET imaging with a [ 18 F]BCPP-EF probe. This probe detects mitochondrial complex I activity, which reflects neuronal viability in the living brain (Tsukada et al, 2014). Rats (8-week-old) that received p3-Alcb9-19 (0, 1, 3, 5 mg/kg body weight) subcutaneously were scanned for 90 min after intravenous injection of [ 18 F]BCPP-EF through a tail vein. Parametric PET images of [ 18 F]BCPP-EF standard uptake value ratios (SUVRs) were superimposed on the CT images of all rats (Fig 5A-D). As described elsewhere (Hosoya et al, 2017), elliptical regions of interest (ROIs) ranging from 12 to 24 mm 2 were placed over the frontal cortex (Fcx), caudate putamen (Cpu), and hippocampus (Hip) by referring to the X-ray CT images (Yamagishi et al, 2019). SUVR levels were compared among brain regions (Fig 5E-G). The SUVR in p3-Alcb9-19-treated rats was significantly higher in all brain regions than in vehicle-treated rats, with the exception of the caudate putamen in animals that received 3 mg/kg p3-Alcb9-19. This finding clearly shows that peripherally administered p3-Alcb9-19, at a dose of 1 mg/ kg body weight, increases neuronal survival while also activating mitochondrial function in the brain. Mitochondrial dysfunction is common in the brain of AD patients (Ridge & Kauwe, 2018), and decreased viability of vulnerable brain regions can be detected by PET imaging with [ 18 F]BCPP-EF (Tsukada et al, 2014). Therefore, we next investigated the reduction in [ 18 F]BCPP-EF SUVR in the brain of AD mice (App NL-F/NL-F ). These animals generate predominantly human Ab42 and exhibit amyloid accumulation, a hallmark of AD pathology . We tested whether p3-Alcb9-19 could restore neuronal activity (Fig 5H-J). AD mice (APP-KI) showed a significant decrease in [ 18 F]BCPP-EF SUVR in the cortex and hippocampus compared with age-matched (12-month-old) wild-type (WT) mice. SUVR values in AD mice were restored to levels similar to those in WT mice after a single subcutaneous injection of p3-Alcb9-19 (1 mg/kg) (APP-KI/p3-Alcb9-19 in Fig 5H-J). This indicates that peripheral administration of p3-Alcb restores the viability of mouse neurons impaired by the accumulation of human Ab42.
The level of neuroinflammation was assessed in AD mice (APP-KI), in parallel with PET imaging using [ 11 C]DPA713, a translocator protein (TSPO) PET radiotracer (Hosoya et al, 2017). The parametric PET images of [ 11 C]DPA713 SUVR were superimposed on CT images (Fig EV4D-F). In these AD mice, SUVRs of [ 11 C]DPA713 showed a significant inverse correlation with those of [ 18 F]BCPP-EF (vehicle) (P < 0.05), demonstrating that the decreased survival and mitochondrial impairment in neurons are associated with an increase in neuroinflammation, which may be evoked by Ab burden (vehicle in Fig 5K). This inverse correlation tended to change to a positive correlation in p3-Alcb9-19 (p3-Alcb, Fig 5K)-injected AD mice, suggesting that the restoration of neuronal viability through p3-Alcb9-19 treatment may be associated with its neuroprotective property.
We further investigated the effect of p3-Alcb9-19 on mitochondrial activity in monkeys (Fig EV4G-I). Three consecutive PET scans in Rhesus monkeys (n = 2) were performed after the transdermal administration of vehicle and p3-Alcb9-19 (0.5 and 1 mg/kg) with PassPort System (Ono et al, 2022). The PET images of [ 18 F]BCPP-EF SUVR, which reflect the degree of mitochondrial activity, show a marked increase in [ 18 F]BCPP-EF binding in monkeys that were treated with 1 mg/kg of p3-Alcb9-19 ( Fig EV4H, dark blue bar). The dose (1 mg/kg) was also used in the current in vivo study in rodents. As shown in the bar graph (Fig EV4I), the percentile increase in binding is observed in the cerebral cortex and hippocampus. The results suggest that experiments in monkeys with transdermal administration of the p3-Alcb9-19 pharmaceutical formulation are sufficient for delivering the peptide into the brain. Although the number of monkeys used in this study was small, the administration of p3-Alcb9-19 (especially in 1.0 mg/kg body weight) indeed raised mitochondrial activity in the brain of monkeys.

p3-Alcb level in the central nervous system decreases with age in monkeys
We previously reported that endogenous p3-Alcb levels in CSF of monkeys decrease with age (Hata et al, 2019). We further analyzed age-related changes in brain p3-Alcb and Ab levels using the monkey brain (Fig EV5). Similar to humans, the monkey brain parenchyma showed increased Ab accumulation with age. For Ab42, the brain Ab level in monkeys was 10 1 to 10 2 fmol/mg protein under 20 years old and 10 3 to 10 4 fmol/mg protein over 30 years old as described previously (Nishimura et al, 2012). By contrast, p3-Alcb levels remain very low (below 10 2 fmol/mg protein) throughout the aging process, eventually falling below a centesimal to that of Ab in aged individuals. Endogenous p3-Alcb levels may be too low to protect neurons against neurotoxic Abo that accumulates in the brain of aged individuals.

Discussion
In this study, we found the neuroprotective function of p3-Alcb, which is generated from Alcb by a metabolism similar to APP (Araki et al, 2004;Hata et al, 2009), and propose a neuroprotective mechanism mediated by p3-Alcb (Fig 6A). Endogenously-derived neuronal p3-Acb suppresses Ca 2+ influx enhanced by Abo-induced NMDAR activation, causing to restore neurons to the healthy state presumably through suppressing the cell death pathway. The detailed molecular mechanisms as to how p3-Alcb regulates Abo-triggered toxic Ca 2+ influx in neurons and upregulates mitochondrial function remain elusive. However, our results suggested that peripheral administration of p3-Alcb9-19 clearly activated brain neurons as shown in PET imaging. Interestingly, a single subcutaneous injection of p3-Alcb9-19 at the dose of 1 mg/kg body weight dramatically restored the viability of AD mouse model neurons. Furthermore, administration of p3-Alcb9-19 (1 mg/kg body weight) improved neuroinflammation triggered by increasing Abo burden in the brain of the AD mouse model. In vivo experiments have is shown (right) as the AUC expressed relative to that of cells cultured in the absence of D-AP5 and Abo (assigned a value of 1.0). Statistical significance was determined by one-way ANOVA with the Tukey's multiple comparisons test (mean AE SE; n = 4), and significant P-values (P < 0.05, P < 0.01) are indicated on the graph. B Suppression of Ca 2+ influx induced by excessive NMDA in neurons treated with p3-Alcb9-19. Mouse neurons (div 11-13) pretreated with Fluo 4-AM were stimulated with NMDA (52 lM) at 10 min (arrow) in the presence or absence of p3-Alcb9-19 (50 lM). The fluorescence intensity was recorded at the indicated time (left) and the fluorescence area intensity for 28 min (10-38 min at the indicated time points) is shown (right) as the AUC expressed relative to that of cells cultured in the absence of p3-Alcb9-19 (assigned a value of 1.0). Statistical significance was determined by the Student's t-test (mean AE SEM; n = 8), and the significant P-value (P < 0.05) is indicated on the graph. C Suppression of Ca 2+ influx into neurons followed by Ca 2+ administration in the presence or absence of D-AP5 and p3-Alcb. Neurons (div 11-14) pretreated with Fluo 4-AM in Ca 2+ -depleted medium were administered Ca 2+ (final 2 mM) at 5 min (arrow) in the presence (5, 50, 200 lM) or absence (0 lM) of D-AP5, p3-Alcb9-19 (+, 50 lM) and p3-Alcb37 (+, 50 lM). The fluorescence intensity was recorded at the indicated time (left) and the fluorescence area intensity for 9 min (5-14 min at time points) is shown (right) as the AUC expressed relative to that of cells cultured in experiment No. 1 (assigned a value of 1.0). Statistical significance was determined with a one-way ANOVA with the Tukey's multiple comparisons test (mean AE SEM; n = 6), and significant P-values (P < 0.05, P < 0.01, P < 0.001) are indicated on the graph. D Nonsynergistic suppression of Ca 2+ influx induced by Ab42 oligomers (Abo) in neurons by p3-Alcb and D-AP5. Neurons (div 14) pretreated with Fluo 4-AM were stimulated with or without Abo (5.2 lM) at 5 min (arrow) in the presence (+) or absence (À) of p3-Alcb9-19 (50 lM) and D-AP5 (50 lM). The fluorescence intensity was recorded at the indicated time (left) and the fluorescence area intensity for 23 min (5-28 min at the indicated time points) is shown (right) as the AUC expressed relative to that of cells cultured in experiment No. 1 (assigned a value of 1.0). Statistical significance was determined with a one-way ANOVA with the Tukey's multiple comparison test (mean AE SEM; n = 12), and significant P-values (P < 0.05, P < 0.01, P < 0.001, P < 0.0001) are indicated on the graph.
Data information: Experimental numbers indicate biological replicates. Detailed information including the statistical summary is described in Dataset EV1. Source data are available online for this figure.
Ó 2023 The Authors EMBO Molecular Medicine 15: e17052 | 2023 confirmed that not only a subcutaneous injection of p3-Alcb9-19 in rodents (Fig EV4A-C) but also a transdermal administration of p3-Alcb9-19 pharmaceutical formulation applied in monkeys (Fig EV4G-I) are sufficient to deliver the peptide into the brain to operate. In our preliminary study, subcutaneous injection of p3-Alcb9-19 (1 mg/kg body weight) into a mouse model of AD, daily for 30 days, did not cause a significant decrease in brain Ab load ( Fig EV6). Therefore, we would like to re-emphasize that the function of p3-Alcb is to increase neuronal viability and to protect neurons against Ab-induced toxicity, and that the target of p3-Alcb may not directly be Ab peptide and Ab aggregates/plaques. This property of p3-Alcb as a druggable candidate for AD therapy is likely to be distinct from immunotherapies that use anti-Ab antibodies. A-D The SUVRs of [ 18 F]BCPP-EF in the frontal cortex (Fcx), caudate putamen (Cpu), and hippocampus (Hip) of rats with (1 mg (B), 3 mg (C), and 5 mg (D) per kg body weight) or without (vehicle (A)) subcutaneous administration of p3-Alcb9-19. The PET data are superimposed on X-ray CT images, and the color bar denotes the SUVR. E-G The SUVRs of groups administered p3-Alcb9-19 versus the vehicle-treated group. Statistical analysis was performed using one-way ANOVA, followed by Bonferroni correction for multiple comparisons (mean AE SEM; n = 6), and significant P-values (P < 0.05, P < 0.01) are indicated on the graphs. H The SUVRs of [ 18 F]BCPP-EF in the cortex and hippocampus of wild-type (WT) and AD mice (App NL-F/NL-F ) with (APP-KI/p3-Alcb9-19) or without (APP-KI) a subcutaneous administration of p3-Alcb9-19 (1 mg/kg body weight). The PET data are superimposed on X-ray CT images, and the color bar denotes the SUVR. I, J The SUVRs of [ 18 F]BCPP-EF in the cortex (I) and hippocampus (J) are compared across the three groups. Statistical analysis was performed using one-way ANOVA, followed by Bonferroni correction for multiple comparisons (mean AE SEM; n = 6-10), and significant P-values (P < 0.05, P < 0.01, P < 0.001) are indicated on the graphs. Furthermore, the function of p3-Alcb to suppress aberrant, Aboinduced, Ca 2+ influx into neurons and to increase neuron mitochondrial activity are novel mechanisms to protect neurons against Aboinduced toxicity, which clearly differs from the therapeutic actions of current drugs such as memantine which does not increase mitochondrial activity (Singh et al, 2017). Since endogenous p3-Alcb levels may be too low to protect neurons against the increased neurotoxic Abo burden in the brain of aged individuals, increasing p3-Alcb9-19 levels via peripheral administration has the potential to increase neuronal viability in aged individuals. Furthermore, our recent analysis shows that endogenous p3-Alcb levels in the CSF of AD patients are significantly lower than in age-matched nondemented subjects (Hata et al, 2019), which is more obvious in early AD patients (Fig 1). Elderly subjects with low levels of p3-Alcb are likely to experience the greater acceleration of AD pathology. Therefore, present results strongly suggest that peripheral administration of p3-Alcb9-19 to AD patients at an early stage, during which endogenous p3-Alcb likely starts to decrease, could constitute a promising therapeutic strategy for the restoration of brain function (Fig 6B).

Animals and mouse primary cultured neurons
All animal studies were conducted in compliance with the ARRIVE guidelines. The mouse study was approved by the Animal Studies Committee of Hokkaido University (#18-0168). WT C57BL/6J (CLEA, Japan, Inc), Alcb-KO (RBRC11514) (Gotoh et al, 2020), App NL-F/NL-F (RBRC06343), and App NL-G-F/NL-G-F (RBRC06344)  mice were housed in specific pathogen-free (SPF) conditions with a microenvironment vent system (Allentown Inc., Allentown, NJ, USA) under a 12 h/12 h light/dark cycle with free access to food and water. Three to five male or female siblings were housed in each cage; cages were equipped with microbarrier tops. All experimental procedures with Cynomolgus monkeys were approved by the Animal Care and Use Committee of Shiga University of Medical Science and were carried out in accordance with approved guidelines (Nishimura et al, 2012). Stored frozen pieces of the cerebral cortex of monkeys were used. Sprague-Dawley rats (8week-old) were purchased from Japan SLC Inc. (Hamamatsu, Japan) and housed with their littermates. Each cage had a maximum of three animals, and food and water were available ad libitum. All animal protocols and related experiments were approved by the Ethics Committees of the Central Research Laboratory at Hamamatsu Photonics and Hamamatsu University School of Medicine. Mixed mouse cortical and hippocampal neurons were cultured for the indicated day (days in vitro in culture/div) using a modification of a previous method (Chiba et al, 2014). Briefly, the cortex and hippocampus of mice at embryonic day 15.5 were removed and neurons dissociated in a buffer containing papain (Cat #LS003119, Worthington, Lakewood, NJ, USA) The cells were then cultured at 5 × 10 4 cells/cm 2 in Neurobasal Medium (Cat #21103049, Gibco/ Thermo Fisher Scientific, Waltham, MA) containing 2% (v/v) B-27 Supplement (Cat #17504044 Invitrogen, South San Francisco, CA, USA), Glutamax I (4 mM, Cat #35050061, Gibco/Thermo Fisher Scientific), heat-inactivated horse serum (5% v/v, Cat #26050088, Figure 6. Possible mechanism of p3-Alcb for suppressing Ab oligomer-induced neurotoxicity and therapeutic strategy. A Schematic mechanism of p3-Alcb to suppress the neurotoxicity by Abo. One of the major targets of Ab oligomers (Abo) is NMDA receptors (NMDARs). Abo triggers unregulated intracellular Ca 2+ influx into neurons, activating the cell death pathway (left). p3-Alcb regulates an unidentified membrane protein X to attenuate intracellular Ca 2+ influx, which inhibits NMDARs that are unusually activated by Abo. This may activate the survival pathway and increases mitochondrial activity in neurons (right). Since p3-Alcb is an endogenous brain peptide, it may help to maintain a healthy brain by reversibly regulating NMDAR-and Abo-mediated Ca 2+ influx. B Therapeutic strategy with p3-Alcb9-19 pharmaceutical formulation. Brain p3-Alcb level decreases and cognitive impairments increase with age. Administration of p3-Alcb9-19 in the early stage of dementia is expected to restore brain function. SMC, subjective memory complaints; MCI, mild cognitive impairment; AD, Alzheimer's disease.

Synthetic peptides and antibodies
Human p3-Alcb37 peptide, which includes the sequence from Val813 to Thr849 of Alcb (Hata et al, 2009), and its partial peptides, p3-Alcb1-11, p3-Alcb9-19, and p3-Alcb20-37, were synthesized and purified to more than 95% purity. Their predicted molecular weights were confirmed by mass spectroscopy, performed at the Peptide Institute (Osaka, Japan). Human p3-Alca35 peptide, which includes the sequence from Ala817 to Thr851 of Alca, was synthesized and purified as described above (Hata et al, 2009). Human Ab42 peptide was synthesized with over 95% purity and purchased from Peptide Institute (Cat #4349-v). The anti-p3-Alca UT135 antibody has been described previously (Hata et al, 2009). A polyclonal rabbit anti-p3-Alcb #854 antibody was raised against an antigen composed of Cys plus the sequence between positions 841 and 849 (C+NSMIPSAAT). This antibody specifically recognizes p3-Alcb and does not cross-react with p3-Alca. Commercially available antibodies used in this study are listed in Appendix Table S2.

Preparation of Ab42 oligomers and detection by immunoblotting
Ab42 oligomers were prepared as described previously (Dahlgren et al, 2002). Briefly, Ab42 peptide was dissolved in hexafluoroisopropanol (FUJIFILM Wako Pure Chemicals Corp., Osaka, Japan) to a concentration of 1 mM. After removing hexafluoroisopropanol under vacuum with a SpeedVac system, the peptide was resuspended in dimethyl sulfoxide Hybri-Max (DMSO) (Sigma-Aldrich, St. Louis, MO, USA) to a concentration of 5 mM. Ham's F-12 (phenol red-free, FUJIFILM Wako Pure Chemicals Corp.) was then added to adjust the final concentration of the peptide to 100 lM. The peptide was then incubated at 4°C for 24 h. Stock solutions (5 mM) of p3-Alca and p3-Alcb peptides were prepared by dissolving each in DMSO. For the detection of Ab42, p3-Alcb37, and p3-Alca35 by immunoblotting, the respective DMSO-containing peptide solutions were diluted in PBS to 10 lM and incubated at 37°C for 24 h. The peptide solutions were then centrifuged at 20,400 g for 10 min at 4°C and the resultant supernatants analyzed with Tris-Tricine polyacrylamide gel SDS electrophoresis followed by immunoblotting with the indicated antibodies.

Monitoring of peptide aggregation with Thioflavin T fluorescence
Aβ42 prepared from a stock solution (1 mg/ml Aβ42 in hexafluoroisopropanol) was distributed in aliquots into tubes, followed by removal of hexafluoroisopropanol as described above. Then, Aβ42 was dissolved in DMSO and diluted with PBS (the final concentration of DMSO was 2%). Similarly, the p3-Alc peptides dissolved in DMSO were diluted with PBS (the final concentration of DMSO was 2%). These peptide solutions (10 μl of 10 μM) were incubated for the indicated time (h) at 37°C. After a 90 μl volume of Thioflavin T (3 μM solution in 100 mM glycine-NaOH buffer, pH 8.5) was added to the peptide solutions, the fluorescence was measured using EnSpire (PerkinElmer, Waltham, MA, USA) with excitation (430 nm) and emission (485 nm) wavelengths (Ex. 430 nm/Em. 485 nm). In a separate study, we also measured the fluorescence (Ex. 430 nm/Em. 485 nm) of the solutions with a 180 μl volume of Thioflavin T incubated for 12 h at 37°C, which were a mixture of a 10 μl aliquot of the Aβ42 solution (20 μM) with each of a 10 μl aliquot of p3-Alcα35 or p3-Alcβ37 solutions with different concentrations (0, 20, 200 μM).

Quantification of intraneuronal Ca
The photo-affinity probe biotin-X-p3-Alcb9-19-K(pBzBz)-NH 2 , consisting of amino-terminal biotin plus MiniPEG3 (X)-conjugated [His-Arg-Gly-His-Gln-Pro-Pro-Pro-Glu-Met-Ala]-Lys modified by benzophenone along with C-terminal amidation (M.W. 2007.3), was synthesized and purified (> 99%) by high-performance liquid chromatography (HPLC). The biotin-X-p3-Alcb1-37-K(pBzBz)-KH2 consisting of 37 amino acids of p3-Alcb37 was similarly synthesized. Mouse primary neurons (div 23) were incubated with 1 lM of biotin-X-p3-Alcb9-19-K(pBzBz)-NH 2 or biotin-X-p3-Alcb1-37-K (pBzBz)-NH 2 for 1 h, followed by UV irradiation for 5 min. The neurons were then fixed with 4% paraformaldehyde in PBS, treated with 0.2% Triton X-100 in PBS, blocked with 4% BSA in PBS, and incubated in primary antibodies for 12 h. This was followed by incubation with an Alexa546-conjugated anti-rabbit or anti-mouse IgG antibody, and streptavidin-Alexa488 to detect biotin-X-p3-Alcb-K (pBzBz)-NH 2 . Fluorescent images were obtained using an all-in-one fluorescence microscope (BZ-X710, KEYENCE, Osaka, Japan) equipped with a Plan Apochromat 100× oil-immersion objective (1.4 numerical aperture (NA), Nikon, Tokyo, Japan). The colocalization rates of proteins and p3-Alcb were calculated from each frame of images (15,750 lm 2 ) of neurons and are indicated as Pearson's R value. Independent cell stainings were performed one to three times per cell preparation, and three to five frames were acquired from each well. All values were combined and subjected to statistical analysis with the indicated number of independent biological repeats.

PET imaging with [ 18 F]BCPP-EF and [ 11 C]DPA713
In the rat study, 24 rats (8-week-old) were divided into four groups with six animals per group: one group (control) was injected subcutaneously with vehicle (saline), and the three other groups were injected subcutaneously with 1, 3, and 5 mg/kg of p3-Alcb9-19. In the mouse study, 12-14-month-old C57BL/6 WT and App NL-F/NL-F mice were injected subcutaneously with p3-Alcb9-19 (1 mg/kg) or vehicle (saline). The vehicle and peptide were administered just before PET measurements. The [ 18 F]BCPP-EF radiotracer was synthesized using a modified CUPID system (Sumitomo Heavy Industry, Tokyo, Japan), and analyzed by HPLC on a GL-7400 low-pressure-gradient HPLC system (GL Sciences, Inc., Tokyo, Japan) as reported previously (Harada et al, 2013). Radioactivity yields, radiochemical purities, and specific radio-activities of [ 18 F]-BCPP-EF were 5.1 AE 0.9 (mean AE SD), 99.1 AE 0.7%, and 139.6 AE 37.0 GBq/lmol. The [ 11 C] DPA713 radiotracer was synthesized by N-methylation of the norcompound N-desmethyl-DPA with 11 C-methyl triflates, as reported elsewhere (Boutin et al, 2007). Radioactivity yields, the radiochemical purity, and specific radioactivity of [ 11 C]DPA713 were 3.5 AE 0.8, more than 99.1 AE 0.9%, and 99.3 AE 32.2 GBq/lmol, respectively. PET measurements were acquired on a high-resolution animal PET scanner Hamamatsu Photonics,Hamamatsu,Japan) with an axial field of view (FOV) of 330 mm, a transaxial FOV of 108 mm, and a transaxial spatial resolution of 2.3 mm in the center as reported elsewhere (Yamagishi et al, 2019). All animals were anesthetized with 1.5-2.0% isoflurane in O 2 for the duration of the entire imaging experiment. A heat pad was used to control body temperature during PET measurements. The animals were placed in the prone position on a fixation plate and then placed within the gantry hole of the PET scanner. After a 15 min transmission scan utilizing an external 68 Ge/ 68 Ga rod source (67 MBq) for attenuation correction, an 80 min serial emission scan was performed immediately after each injection of [ 18 F]BCPP-EF at a dose of 5 MBq. The tracers were injected intravenously through a cannula inserted into the tail vein. The molar activity of each tracer was above 50 GBq/lmol. No arterial sampling was conducted. The PET data were reconstructed using 3D DRAMA (iteration 2, gamma 0.1) with a Gaussian filter of 1.0 mm full width at half maximum (FWHM), yielding a voxel size of 0.65 × 0.65 × 1.0167 mm for the reconstructed images. To obtain anatomical information, X-ray CT scans were performed before the PET measurement, using a ClairvivoCT (Shimadzu Corporation, Kyoto, Japan) (Yamagishi et al, 2019). Using PMOD image analysis software (version 3.7; PMOD Technologies Ltd, Zurich, Switzerland), the SUVR for [ 18 F]BCPP-EF binding was estimated by dividing the target SUV by the cerebellar SUV (Tsukada et al, 2014). The SUV was calculated as the measured radioactivity divided by the ratio of the total injected dose to the mouse body weight. As described elsewhere (Hosoya et al, 2017), elliptical ROIs ranging from 12 to 24 mm 2 were placed over the frontal cortex, caudate putamen, and hippocampus by referring to the X-ray CT images (Yamagishi et al, 2019). A oneway analysis of variance (ANOVA) was applied to compare SUVR levels in brain regions among the groups. The significance level was set at P < 0.05 with Bonferroni correction for multiple comparisons.

Transdermal microporation of p3-Alcb in Rhesus monkey
The transdermal delivery device, PassPort System (PS), was provided by PassPort Technologies, Inc (San Diego, CA, USA). The patches to stick on the skin contained different doses of p3-Alcb9-19 peptide and vehicle (saline) within the matrix were prepared in PassPort Technologies, Inc, and each patch in the aluminum laminated pouch was packed with a desiccant until use. The skin for the patch was exposed by shaving the hair of the monkey using an electric shaver 1 day before the p3-Alcb9-19 administration. Transdermal microporation was applied with a condition at 400 density and 4 mJ/filament, followed by the application of the patch on the skin with a transdermal area of 0.5 cm 2 . Immediately after the patch application, the PET scans with [ 18 F]BCPP-EF were conducted during 90 min following intravenous injection of [ 18 F]BCPP-EF as described above.
ELISA for p3-Alcb9-19 quantification A polyclonal rabbit antibody was raised against p3-Alcb9-19 containing an amino-terminal Cys residue (C+HRGHQPPPEMA) and was conjugated to bovine thyroglobulin. IgG was purified with antigen-coupled resin and conjugated to biotin. Horseradish peroxidase-conjugated streptavidin was purchased from Amersham/ GE Healthcare (Cat #RPN1051, Little Chalfont, UK), and the tetramethyl benzidine (TMB) microwell peroxidase substrate system was obtained from SeraCare Life Sciences Inc. (Cat #5120-0075, Milford, MA, USA). Mice were then anesthetized with 1% isoflurane, CSF was collected from the cisterna magna as described previously (Liu & Duff, 2008), and mice were then sacrificed. To quantify p3-Alcb9-19 levels in mouse CSF and plasma, the samples were diluted in buffer A (PBS containing 1% BSA and 0.05% Tween-20). Using this polyclonal antibody, we developed a sELISA system to quantify p3-Alcb9-19 levels in the range of 25-200 pg/ml. The sensitivity of this ELISA is equivalent to that of other sELISA systems used to quantify p3-Alcb37 and p3-Alcb40 (Hata et al, 2019). Antiserum diluted 1:10,000 was used as a capture antibody in the sELISA. The antibody was affinity-purified with antigen-coupled resin, and biotin-labeled IgG was then used as the detection antibody. No reaction occurred with p3-Alcb37 (Fig EV4A), and the addition of 1,000 pg/ml p3-Alcb37 did not compete with antibody binding to 0-200 pg/ml p3-Alcb9-19, indicating that this sELISA could quantitatively measure p3-Alcb9-19 in body fluids, even in the presence of endogenous p3-Alcb.

Extraction and quantification of Ab and p3-Alcb from monkey brain
Quantifications of Ab40 and Αb42 in temporal cortex tissue from Cynomolgus monkeys were performed with sELISA kits specific for Ab40 (Cat #292-62301, FUJIFILM Wako Pure Chemicals Corp.) and Ab42 (Cat #296-62401, FUJIFILM Waco Pure Chemicals Corp.) as described previously (Nishimura et al, 2012). In a separate study, temporal cortex tissue was homogenized in eight volumes of Trisbuffered saline (20 mM Tris-HCl [pH 7.6], 137 mM NaCl) containing a protease inhibitor cocktail (5 lg/ml chymostatin, 5 lg/ml leupeptin, and 5 lg/ml pepstatin A) with 30 strokes of a Dounce homogenizer and centrifuged at 180,000 × g for 20 min at 4°C. Since the supernatant contained a low level of p3-Alcb peptides that were below the limit of detection, the precipitate was further homogenized in one volume of 6 M guanidine chloride, sonicated twice for 10 s each, and allowed to stand for 1 h at room temperature. The samples were then centrifuged at 180,000 × g for 20 min at 4°C. The supernatant was diluted 12-times in PBS containing 1% (w/v) BSA and 0.05% (v/v) Tween-20 and then assayed with a sELISA specific for p3-Alcb37 and p3-Alcb40, as described previously (Hata et al, 2019).

Detection and measurement of amyloid plaques in mouse brain
App NL-G-F/NL-G-F mice (9-month-old females) were subcutaneously administered p3-Alcb9-19 (1 mg/kg body weight) or PBS daily for 30 days. The mice brains were fixed and cut into 20-lm-thick coronal slices. The brain sections were immunostained with a mouse monoclonal anti-human Ab antibody (82E1) (IBL), and the localization of Ab was visualized with an Alexa Fluor 488-conjugated donkey anti-mouse IgG secondary antibody (green) (Invitrogen). Nuclei were counter-stained with DAPI (blue). The sections were viewed with a BZ-X710 microscope (Keyence).

Extraction and quantification of Ab from mouse brain
App NL-G-F/NL-G-F mice (9-month-old males and females) were subcutaneously administered p3-Alcb9-19 (1 mg/kg body weight) or PBS daily for 30 days. The cerebral cortex and hippocampus were dissected and then homogenized on ice for 30 strokes with a Dounce homogenizer in a 4-fold volume of TBS (20 mM Tris-HCl, pH 7.4 containing 137 mM NaCl) and protease inhibitor cocktail (PIC) (5 lg/ml chymostatin, 5 lg/ml leupeptin, and 5 lg/ml pepstatin). The lysate was centrifuged at 200,000 × g for 20 min at 4°C with a TLA 100.4 rotor (Beckman Coulter Life Science). The resultant The paper explained Problem Neuronal p3-Alcb peptides are generated through proteolytic cleavage of the precursor protein Alcadein b/Alcb (also known as calsyntenin-3/ Clstn3) by aand c-secretases, while neurotoxic amyloid b (Ab) is derived from the sequential cleavage of Ab-protein precursor/APP by band c-secretases. Ab oligomers (Abo) are considered highly neurotoxic and the main culprit playing a role in the pathophysiology of Alzheimer's disease (AD). AD is the most common, incurable neurodegenerative disease in elderly subjects with dementia. A series of our experiments so far have shown that the p3-Alcb level in CSF decreases with age and with Ab accumulation increasing, which causes to reduce the p3-Alcb expression in the brain. Moreover, the CSF p3-Alcb level significantly decreases in AD patients compared with that in age-matched nondemented subjects. Intriguingly, subjects carrying presenilin gene mutations of hereditary and familial AD show a lower p3-Alcb level than noncarriers in the family. These lines of evidence suggest that p3-Alcb may associate strongly with AD pathophysiology, but the exact functions of p3-Alcb remain unclear.

Results
As a proof-of-concept, we asked about the effects of p3-Alcb on neurons. In contrast to the neurotoxic effect of Abo, the p3-Alcb increased neuronal viability and protected neurons against the neurotoxicity of Abo. We identified a functionally active shorter peptide p3-Alcb9-19, composed of 11-amino acid sequence of endogenous p3-Alcb37. Interestingly, p3-Alcb37 and p3-Alcb9-19 inhibited anomalous Abo-induced Ca 2+ influx in neurons and restored neuronal viability by keeping intracellular Ca 2+ homeostasis. Moreover, peripheral administration of p3-Alcb9-19, which shows an excellent BBB crossing property, restored neuronal viability impaired by increasing Ab burden in the brain of the AD mouse model.

Impact
No curative drugs for AD exist yet. Anti-AD immunotherapeutic drugs being developed are confronting difficulties in their effectiveness, safety, cost performance, and so on. This reality gives us to reconsider a different therapeutic strategy other than drugs targeting Ab. Here, we showed that a short peptide p3-Alcb9-19 is transferred into the brain satisfactorily by peripheral administration, which allowed us to develop the transdermal administration procedure with p3-Alcb9-19 pharmaceutical formulation. The p3-Alcb is a brain endogenous peptide and is detected in blood in humans, guaranteeing the safety and regular evaluation of its level in patients. In fact, no toxicity was observed in a series of rodent and monkey experiments. The short peptide is much less expensive and stable at 4°C for years. Novel therapy with p3-Alcb9-19 pharmaceutical formulation is expected as a promising drug for AD patients.
precipitate was further homogenized in a 9-fold volume of TBS with a Dounce homogenizer for 30 strokes and was then centrifuged at 100,000 × g for 20 min at 4°C with a TLA 55 rotor (Beckman Coulter Life Science). The pellet was dissolved in an equal volume of 6 M guanidine-HCl solution in 50 mM Tris-HCl (pH 7.6) with sonication (1 s with a 1 s interval of 17 cycles) and left to stand for 1 h at room temperature. The sample was then centrifuged at 130,000 × g for 20 min, and the supernatant was used for Ab42 assays. Human Ab42 levels were quantified by sandwich ELISA (sELISA) as previously described (Honda et al, 2023).

Cohort information and quantification of CSF biomarkers
We stratified 131 patients whose CSF was collected for diagnostic purposes at Niigata University and related facilities by AT (N) classification according to CSF biomarkers independent of clinical diagnosis. The CSF concentration of Ab42, p-tau181, and total tau (t-tau) was examined at Niigata University, and the cut-off value for Ab42, p-tau181, and t-tau was described previously (Kasuga et al, 2022). The Ethics Committee of Niigata University approved this study . Participants gave informed consent to participate in the study before taking part. The CSF concentration of p3-Alcb37 was quantified with ELISA system as described previously (Hata et al, 2019). The Ethics Committee of Hokkaido University approved this study (2021-003).

Statistical analysis
Statistical differences were assessed using the Student's t-test or one-way ANOVAs combined with the Tukey-Kramer post hoc test and Dunnett's test or Bonferroni's test for multiple comparisons (GraphPad Prism software, version 9.4.0). P-values < 0.05 were considered significant.

Data availability
Data in this article will be shared on reasonable request from any qualified investigator.
Expanded View for this article is available online.