Alzheimer’s disease (AD) is the most frequent cause of dementia, affecting about the 70% of those 50 million people suffering from dementia world-wide. Moreover, its prevalence is expected to spread dramatically in the coming years, due to the global increase of population mean age. More specifically, AD prevalence is believed to double by 2050 and disease-related costs are estimated to rise from a trillion of dollars in 2018 to 2.5 trillions by 2030, delineating an urgent public health problem (1). Therefore, it is important to develop effective and safe therapies to relieve memory loss and the global cognitive decline in AD patients, thus reducing the burden of this disease (2). Currently, no treatments are available to halt or reverse AD. In fact, the only available drug therapies for AD approved by the Food and Drug Administration (FDA) are targeting symptoms, such as cholinesterase inhibitors (donepezil, rivastigmine and galantamine) and memantine, which is a N-methyl-D-aspartate (NMDA) receptor antagonist, and novel treatments have mostly failed in clinical studies, except for the expensive monoclonal antibody aducanumab (3) whose marketing authorization was however recently rejected by EMA (EMA/112932/2022 Rev.1); additionally, treatment of late stage AD with these drugs shows moderate results (4, 5).

Despite an intense research activity, clinically available drugs for AD that target disease-specific pathways are currently lacking: many promising treatments in phase II clinical trials have not passed to phase III due to lack of efficacy and/or adverse events (6). A possible explanation of the poor clinical benefit of the studied drugs may be the inability to reach the right target at the right stage of AD. Indeed, since β-amyloid (Aβ) may accumulate in the brain for more than 10 years before the first clinical manifestation of AD, it could be difficult to treat these patients, especially those with a longer disease history (6, 7). In this context, the development of disease-modifying agents with pleiotropic and or multitarget activities is urgently needed (7, 8). There is a large literature showing that Aβ complexes and plaques, as well as neurofibrillary deposition, impair neuronal survival leading to altered intracellular axonal transport, cellular shrinking, synaptic loss and apoptosis (7, 9). Moreover, the Aβ plaque accumulation leads to the recruitment of inflammatory cells favoring neuronal loss (10, 11). Further evidence demonstrates an important role of microglia, the resident central nervous system (CNS) macrophages, in Aβ pathology (12). In fact, in healthy brain microglia exert several crucial roles, by sensing and detecting signs of injury, undergoing morphological and biochemical changes leading to an activated phenotype; these processes could be triggered through different targets including melanocortin receptors (MCR). Synthetic melanocortin agonists have been previously reported as a promising therapeutic tool towards a wide range of central neurological syndromes (13–15). In particular, NDP-α-MSH interaction with melanocortin receptor 4 (MC4R) (16), the most abundant MCR in the CNS and the only one present in microglia and astrocytes, has been already shown as an effective treatment for ischemia and early and moderate AD in experimental mouse models (10, 17–22). However, no studies have explored yet the potential of MCR agonism in severe models of AD.

The aim of this study is to assess the effects of MCR agonist NDP-α-MSH in two transgenic mouse models of AD, 5XFAD and 3xTg mice, that markedly differ as regards the relationship between cognitive alterations and neuropathology. In fact, in 5XFAD mice the accumulation of intracellular Aβ (iAβ) is transient and very precocious (2-3 months of age, mo) and is followed by the deposition of extracellular Aβ (eAβ) and neuroinflammation that are associated with the appearance of cognitive decline (5-6 mo) (11). On the other hand, in 3xTg mice, cognitive decline (5-6 mo) precedes accumulation of iAβ and eAβ and neuroinflammation that become apparent at 9 (iAβ) or 14 (eAβ) mo, being indeed markedly less intense than in 5XFAD mice (23). Therefore, while both AD models lead to prominent cognitive impairments at comparable age, the development of neuropathology is much slower and less intense in 3xTg mice. For these reasons, in this study we analyzed the possible effects of chronic NDP-α-MSH treatment on cognitive and neuropathological alterations in 5 and 7 mo 5xFAD mice and in 9 and 14 mo 3xTg mice.

In this study we demonstrated that the NDP-mediated chronic stimulation of MCRs improves cognitive abilities of AD transgenic mouse models at late stage of AD progression. We also showed that these protective effects are associated with decreased hyperphosphorylated tau but not with Aβ burden, that was unaffected in the Hip and in the CTX of AD mice. In addition, an age-dependent NDP effect on microglial reactivity was observed only in 9 mo 3xTg mice whereas a global downregulation of p38MAPK was selectively observed in 7 mo 5XFAD and 14 mo 3xTg mice.

Specifically, hippocampus-dependent disrupted cognitive function and worsened performance of spatial memory task in 5 mo and 7 mo 5XFAD mice ( Figures 1A, B ) and in 9 mo and 14 mo 3xTg mice ( Figures 1C, D ) were significantly improved when NDP was daily administered for 45 days before behavioral testing. These results are in accordance with previous evidence obtained in other pre-clinical models at earlier stage of AD progression; in fact, we and others have demonstrated that stimulation of the hippocampal MC4R circuit relieves synaptic plasticity impairment in a mouse model of early-stage AD (10, 29, 34). We speculate that these changes could be mediated by the stimulation of MC3R and MC4R, two MCRs widespread in almost all cerebral areas including several key areas involved in memory such as the Hip (35). It is well known that the melanocortins are involved in a several physiological functions, including pigmentation (MC1), steroidogenesis (MC2), energy homeostasis (MC3), regulation both food intake and sexual function (MC4), neurogenesis, analgesia and inflammation (MC4), exocrine secretion (MC5) (36). Our present and past results demonstrate that NDP administered i.p. improves peripheral pathologies (hemorrhagic shock, sepsis, myocardial ischemia) and central ones (cerebral ischemia, head trauma and AD) and for the latter it is plausible to hypothesize the involvement of the MC3 and MC4 receptors as they are mainly expressed centrally, but it cannot be excluded an indirect involvement also of the peripheral melanocortin receptors. Other studies demonstrated that the treatment with other MCR agonists, such as α-MSH, restores fear memory impairment induced by IL-1β injection in dorsal Hip (34, 37). Globally, these observations indicate that the stimulation of melanocortin pathways reestablishes memory function even in advanced stages of AD.

To assess if improved behavior is associated with reduced neuropathology, we investigated the expression of AD-linked alterations at both histological and protein levels by means of immunohistochemical and western blot analyses, respectively. First, we measured the 6E10+ ir in cortico-hippocampal regions, the most affected by Aβ deposition in both 5XFAD and 3xTg mice; in addition, since 6E10 Ab recognises APP, iAβ and eAβ (11), the number of 6E10+ plaques as well as the area coved by 6E10 ir were considered. We did not find any significant effect in CTX and hippocampal subregions of either 5XFAD or 3xTg mice at both experimental ages, apart from an isolated increase in 6E10+ plaques in CA3 of 5XFAD mice at 5 months of age; of note, these observations were also supported by ex vivo analyses of relative pAPP T668 expression levels in the Hip of Wt and AD mice ( Figures 7A, B ). Differently from Aβ levels, our western blot analysis demonstrated that NDP treatment significantly reduced the level of hyperphosphorylated tau, specifically pTau S202 and pTau T181, in the Hip of AD mice. Taken together, these data suggest that NDP treatment and MCR stimulation in late phase of AD progression is not able to decrease the Aβ burden in severe mouse models of AD, such as 5XFAD and 3xTg mice, but is able to reduce tau phosphorylation that is known to better correlate with cognitive abilities (38).

Emerging evidence implicates MCRs in neuroinflammation, that results in a reduced activation of astrocytes and microglia (39). Here we reported an increased microgliosis, measured as the number of IBA1+ cells and the area covered by IBA1 ir, in the cortex and in the Hip of both 5 and 7 mo 5XFAD mice without any significant influence of NDP treatment ( Figure 5 ). Analogously, in 3xTg mice a significant microgliosis was observed exclusively at 14 months of age without any effect of NDP treatment in all analyzed areas; at 9 mo, even if we did not detect any significant increase either in the number or in the area coved by microglial cells, the NDP treatment reduces microgliosis in the CA3 field of Hip of 3xTg mice ( Figure 6 ). The results in 3xTg mice suggest that NDP treatment, possibly by rescuing synaptic plasticity (34), reduces microglial activation and cell proliferation that represent a hallmark of inflammation in the brain (40) to the levels of Wt. The differences observed in the two AD models used in this study, apparently contradictory, could reflect their intrinsic differences; specifically, in 5XFAD mice the process of microgliosis represents a precocious event consequent to first amyloid plaque deposition and precedes behavioral deficits (11) whereas in 3xTg mice, amyloid plaque formation occurs later and gradually increases with age and follows declined cognitive functions (23). Since the process of amyloidosis and, as a consequence, microgliosis is slower in 3xTg mice, we hypothesize that NDP treatment is more effective on microgliosis in earlier stages of AD pathology progression. Our results tally well with previous studies showing that induced MC4R activation on microglia and astrocytes elicits anti-inflammatory effects by stimulating the cholinergic pathway and the adenosine monophosphate-activated protein kinase (AMPK)/c-Jun N-terminal kinase (JNK)/p38 MAPK (41, 42); these studies demonstrated a prominent role of MCRs in reducing release of cytokines/mediators, such as TNFα, IL1ß, IL6 and reactive oxygen species, and pro apoptotic marker Bax, in experimental cerebral ischemia/reperfusion, traumatic brain injury and LPS-induced neuroinflammation (22, 42, 43). Of note, in our experimental conditions, we did not detect any global change in MC4R expression due to AD or NDP treatment suggesting a pleiotropic effect mediated by the activation of different MCRs ( Supplementary Figure 2 and Supplementary Table 2 ).

Furthermore, we only tested the expression of IBA1, a widely-used marker for microglial cells, that does not give any information about the functional/transcription state of these cells; even if we did not detect any changes in cell number and the area occupied by these cells, we cannot exclude that NDP treatment could improve microglial functions. In fact, our WB results demonstrated that the increased activation of p38 MAPK occurring in AD mice is reverted by NDP treatment in both 7 mo 5XFAD and 14 mo 3xTg mice ( Figures 7A, B ) confirming its possible anti-inflammatory role.

Finally, NDP stimulation did not cause a significant extensive reduction in GFAP ir, considered to represent an index of astrogliosis, with the exception of SUB in 9 mo 3xTg mice ( Figure 4 ). Of note, also this result could be explained by specific features of amyloidosis development. In fact, the SUB of 9 mo 3xTg mice represents an instance of initial eAß deposition in the presence of iAß accumulation. Other regions of 3xTg mice show only iAß accumulation whereas SUB in 14 mo 3xTg mice (similarly to 5 or 7 mo 5XFAD mice) have extensive deposition of eAß. Therefore, we speculate that NDP treatment could exert some protective effect on astrocytes only at specific time window/stage of AD progression, precisely, at early stage of eAß deposition. In addition, since GFAP ir does not label all astrocytic populations (44, 45), our data do not exclude that NDP treatment may have a more pervasive protective effect on other astrocytic populations. Indeed, it is known that GFAP+ astrocytes contain subpopulations with different functional roles, such as A1 neurotoxic astrocytes. In fact, a recent study demonstrated that MCR stimulation in 7 mo APP/PS1 mice reduces neurotoxic A1 subtype of reactive astrocyte, which represents the major astrocytic subtype that mediates astrocytic toxicity in AD (40). Further studies with A1-specific markers may show alterations in astrocytic subpopulations that have been overlooked in present analysis.

This study has some intrinsic limitations. First, we focused on the effect of NDP on AD progression using two animal models at two ages, and to limit the number of mice in the behavioral experiments, we did not include Wt groups with NDP treatment, since we previously demonstrated the absence of any NDP effects on learning and memory in Wt mice (10, 21). However, we cannot exclude that NDP treatment induces some histological or biochemical effects in Wt mice.

Second, our ex vivo analysis conducted in whole hippocampi does not allow to unveil NDP effects on target proteins in specific cell populations. For example, even if globally we were able to detect few changes in some protein expression (see Supplementary Table 2 ), we cannot exclude that changes occur in restricted cell populations especially in close proximity to AD-like neuropathology. Further studies are needed to study those effects on neuronal and glial population.

Finally, even if this study was conducted in multiple animal models of AD and behavioral results were reproducible among them, these models only mimic some aspects of the typical late-onset AD dementia in humans and, for this reason, the translational predictability of present results into clinic is not immediate.

To conclude, here we investigated and compared, for the first time, the effect of NDP treatment in two animal models of AD, 5XFAD and 3xTg mice, at different phases of AD-like pathology progression. Our results clearly demonstrated that, even if systemic NDP treatment exerted only minor effects on Aβ burden and neuroinflammation, cognitive improvements are evident in both mouse models and at different stages of AD progression.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

The animal study was reviewed and approved by Committee on Animal Health and Care of the University of Modena and Reggio Emilia and Italian Minister of health.

ED, EV, WL, MB, AV: mouse breeding, in vivo investigation, data collection; AO: biochemical analysis; AV and DG: data collection and curation, formal analysis, study supervision and validation; AV: original draft preparation; MZ, ED, EV, DG, CB: writing - review & editing; DG: Conceptualization and funding acquisition. All authors have read and agreed to the published version of the manuscript.