Beneficial effects of novel antagonists of GHRH in different models of Alzheimer's disease

Introduction

Several etiological factors have been implicated in the pathogenesis of Alzheimer's disease. These factors, in this most frequent form of dementia, lead to the activation of a cascade process that brings about neuronal death and serious decline in cognitive function. These bed-ridden patients ultimately succumb to death due to inter-current infections related to aspiration, decubitus and stagnation of urine [1].

This process involves, biochemically, a pathological cleavage of amyloid precursor protein (APP). APP, in normal circumstances, is cleft by α- and γ-secretases and takes part in axonal transport, synapse formation and synaptic repair in the CNS [2]. The abnormal, sequential processing by the beta-site amyloid precursor protein–cleaving enzymes (BACE) and γ-secretase results in amyloid-β, which is highly neurotoxic. Molecules of amyloid-β, especially the amyloid-β_1-42 type, are prone to aggregation and accumulation in the cell membrane forming insoluble aggregates called “rafts”. Subsequently, these impair membrane conductivity, Ca²⁺ fluxes, control of the formation of reactive oxygen species (ROS), τ-protein assembly, axonal transport and the polarity of mitochondrial membrane. Ultimately, the pathologic cascade leads to Ca²⁺ toxicity, activation of apoptotic processes, inflammation and neuronal death [1]. This vicious cycle can be initiated by a wide array of triggering mechanisms that can be traced back to genetic or environmental factors [3]. Monogenic forms represent the infrequent, presenile or early-onset familial Alzheimer's disease (FAD), which is usually characterized by autosomal dominant point mutations of the genes of APP or the presenilin sub-domains of γ-secretase. Both types of mutations facilitate the accumulation of toxic amyloid-β_1-42, due to abnormal processing or breakdown [3]. Studying FAD, in the past few years, has provided us with an indispensable means to elucidate the underlying pathological phenomena. The abundant, senile or late-onset type of Alzheimer's disease is epidemiologically sporadic and can be attributed to polygenic influences and environmental factors [3]. Epidemiological studies show strong correlation between Alzheimer's disease and metabolic syndrome demonstrating that the lipoprotein profile (ApoE₄ homozygous genotype), hyperinsulinemia and type II diabetes mellitus are among the most characteristic prognostic factors [3]. The secretion of CNS neurohormones, the master regulators of these endocrine and metabolic conditions is considerably altered by senescence [4]. Changes in corticotrophin releasing hormone (CRH), luteinizing hormone-releasing hormone (LHRH), and GHRH secretion related to obesity, hyperinsulinemia and altered leptin signaling may play a role in the development of Alzheimer's disease [5]. Besides, hypothalamic neurohormones have been shown to have intrinsic activity on both cognitive processes [6-10] and the development of Alzheimer's disease [6, 7, 9-11]. In the same category, exogenous GHRH has recently been demonstrated to impair hippocampal memory consolidation [12] while antagonists of LHRH and GHRH showed a positive impact on learning and memory [6, 7, 9-11]. Data in the literature are not in complete agreement regarding the etiopathogenetic role of the GHRH-GH-IGF-I system in the development of senile dementia [13-17].

Since centrally released GHRH controls the whole GHRH-GH-IGF-I axis, a novel and very potent antagonist of GHRH (designated MIA-690), synthesized in our laboratory, was tested in vivo and in vitro in different models of Alzheimer's disease to clarify the effects of GHRH on the progress of symptoms and to elucidate the mechanism of action. Transgenic mice (5XFAD strain), that develop neurodegenerative symptoms characteristic of Alzheimer's disease were used in vivo. These genetically engineered mice accumulate amyloid-β_1-42 plaques as a result of presenilin-1 (PSEN-1) and APP mutations. Proper cleavage is hindered due to both substrate and enzyme mutations [18]. In behavioral studies, spatial learning and memory of the transgenic mice treated with the GHRH antagonistic analog were recorded and followed up with the help of Morris water maze (MWM), for 6 months. The studies, in vitro, tested the effects of the analog on human neural degeneration evoked by amyloid-β_1-42. Human cortical neuronal stem cells (HCN-2) were cultured, differentiated and subjected to co-treatment with amyloid-β_1-42 and our GHRH antagonist. The viability of the cells was measured by proliferation assays, while changes in gene expression in autopsied brain samples were detected by “Mouse Alzheimer's Disease” real-time RT-PCR Array. The alterations of nucleic acid metabolism were later confirmed by analysis of protein changes using enzyme-linked immunosorbent assays (ELISA). A special emphasis was placed on the proteomic detection of the expression of IGF-I, IGF-II and brain derived neurotrophic factor (BDNF). Since oxidative stress and inflammation are key pathologic processes in Alzheimer's disease [1], direct assessment of reactive oxygen species formation was also performed by aminophenyl-fluorescein (APF) assays, while the expression of superoxide dismutase 1 (SOD1) and glutathione-peroxidase 1 (GPx1) was determined by ELISA. The extent of the accumulation of amyloid-β and total τ was determined by ELISAs from the necropsied brain samples of 5XFAD mice.

Results

Discussion

Our previous studies clearly demonstrated that the GHRH antagonists developed by us [19, 20] had a strong impact on the GHRH - GH – IGF-I axis [21]. The effects of GHRH antagonists on several other physiologic and pathologic conditions have already been demonstrated [19, 22-28]. The role of the GHRH - GH – IGF-I axis in the regulation of learning processes and the development of Alzheimer's disease, however, has not been completely clarified. Both hyperactivity [29, 30] and hypoactivity [31] of the system have been connected to decreased life-expectancy and impaired cognition. Nevertheless, important recent observations support the view that the suppression of the axis can improve cognition in both physiologic and pathologic conditions [7, 10, 11, 15, 17].

Present experiments demonstrate that the GHRH antagonist, MIA-690, has several beneficial effects in each of the different models of Alzheimer's disease. Regarding the GHRH-GH-IGF axis, MIA-690 decreased the secretion of IGF-I (Fig. 6) in the supernatant of HCN-2 cell cultures which is critical since in Alzheimer's disease one of the most important pathologic phenomena, is the competition of insulin and amyloid-β for insulin-degrading enzyme (IDE). In hyperinsulinemic states, insulin, which has higher affinity to IDE, occupies the binding sites of the enzyme, rendering IDE inaccessible to amyloid-β, hence amyloid-β accumulates [17]. In addition to insulin, IGF-I, which is the most abundant member of the IGF family in the CNS, appears to influence cognitive decline in pathologic conditions [5, 13, 16, 17]. Decreased IGF-I signaling appears to ameliorate, directly, the amyloid-β proteo-toxicity by enhancing DNA repair and increasing resistance to oxidative stress [14, 15].

In the in vivo studies, the GHRH antagonist significantly and dose-dependently delayed the Alzheimer's disease-related deterioration of the acquisition phase in MWM (Fig. 1A, Fig. 3C). The peptide also tended to improve the parameters of cognitive performance by the 6^th month of the follow-up period as reflected by the probe values (especially the cumulative distance and platform crossings) of spatial reference memory (Fig. 1B, C, D, Fig. 2, Fig. 3D). The PCR Array studies (Table 1), revealed that the neuro-peptide analog, beside several possible, long-term activities, may have acute beneficial effects on learning. This is in harmony with our previous findings [7], and confirms the inhibitory activity of intranasal GHRH agonists on hippocampal memory formation [12]. MIA-690 increased the expression of ubiquinol-cytochrome c reductase core protein 2, which suggests that the GHRH antagonist may restore impaired cellular respiration [32]. In contrast, MIA-690 decreased the expression of acetylcholinesterase, which is consequential, considering the key role of acetylcholine in hippocampal learning [33]. Further, the inhibition of acetyl-cholinesterase is one of the most important, currently available, palliative treatment options for Alzheimer's disease [34].

Our previous publications have already demonstrated that different classes of hypothalamic neurohormone analogs could influence CNS functions. For example, the LHRH antagonist, cetrorelix, facilitated memory and had anxiolytic and antidepressive actions in mice [6] and rats [9] exposed to the neurotoxic effects of an amyloid-β fragment (amyloid-β_25-35). In a similar fashion, the GHRH antagonist, MZ-4-71, improved memory consolidation in passive avoidance learning [7, 10], decreased anxiety [7, 35], and proved to be antidepressive [7, 36], in CFLP mice treated with amyloid-β_25-35. In a different model of cognitive decline, the treatment of senescence accelerated mice (SAMP8) with another GHRH antagonist (MZ-5-156) also enhanced cognitive functions [11]. The GHRH antagonist treatment produced significantly better performance in active avoidance T-maze, step-down passive avoidance and object recognition. Also, the peptide improved pole balance and increased survival and telomerase activity. These findings are in complete agreement with our studies, as MIA-690 treatment also showed a tendency to prolong survival (Fig. 4).

Our genomic and proteomic studies shed light on further possible mechanisms of action of MIA-690. The GHRH antagonist influenced the transcription of almost 2 dozen putative Alzheimer's disease markers according to the PCR Array experiments (Table 1). The most notable examples are related to the metabolism of amyloid-β, the microtubule system, apoptosis, neural signal transduction, and energy homeostasis. Regarding the metabolism of amyloid-b, the transcriptional studies revealed a remarkable inhibition of the expression of APPs and the amyloid-β precursor protein-binding proteins (APP-BP)s. Further, the GHRH antagonist decreased the transcription of the amyloid-β generating BACE2 and several components (presenilin 1, presenilin 2, anterior pharynx-defective 1 and nicastrin) of the γ-secretase complex [37]. The pathogenic role of cathepsins in the development of Alzheimer's disease needs further clarification [38] since they can play a role in both the generation of and the degradation of amyloid peptides. In our experiments, substantial decreases could be detected in the “C” and “G” members of the family. The observed down-regulation of low density lipoprotein receptor-related protein (LRP)6 provide further evidence that MIA-690 can influence amyloid-β degradation, since LRPs, in cooperation with α2-macroglobulin and ApoE, are the main factors in the modulation of amyloid-β secretion/breakdown [1, 39].

Similar changes are observed in the expression of the microtubule-associated proteins (MAP)s. Both MAP2 and MAPτ were down-regulated. The final markers of these genomic changes (amyloid-β_1-42 and total τ levels) were verified by proteomic determination. The amyloid-β_1-42 level showed especially dramatic decrease due to MIA-690 treatment (Fig. 5), which finding suggests that further studies are needed and should include the selective measurement of hyperphosphorylated as opposed to total τ levels. This also confirms that the relationship between amyloid-β_1-42 raft formation and τ-based neurofibrillary tangle formation is not direct but is a highly unpredictable and cross-related process of the deranged axonal transport [40]. The importance of hyperphosphorylation is accentuated by the marked down-regulation of cyclin-dependent kinase 1 (CDK)1. CDKs belong to those “culprits”, which take part in the post-translational modifications of the MAPs [41], and catalyze, with phosphorylation, the misfolding of τ fibrils.

An important feature of Alzheimer's disease is the connection of the cerebral amyloid process to neural regeneration via the γ-secretase complex. As a double-edged sword, γ-secretase is the rate limiting enzyme of not only amyloid-β production, but also Notch signaling. This, in turn, regulates cell renewal in the CNS [42]. Therefore, overburdening of γ-secretase by amyloid-b accumulation, not only accelerates apoptosis and neuronal degeneration but may also decrease the neurogenesis mediated by Notch signaling [43]. The decrease in APP and BACE2 expression and amyloid-β_1-42 production evoked by MIA-690 may preserve the regenerative potential of the CNS. This concept is supported by the down-regulation of growth associated protein 43, which seems to play a pathologic role in the abnormal hyper-activation of hippocampal cells and the dysfunctional signal transduction seen in Alzheimer's disease [44]. Regarding regenerative processes, however, the most important finding is that treatment with MIA-690 significantly increased the secretion of BDNF (Fig. 6); this protective neurotrophin is a well-established and adversely affected marker of neurodegenerative processes [1, 45].

The attenuation of oxidative stress by GHRH antagonists [11] or by the decrease in IGF-I signaling [15] also seems to play a crucial role in the protection against amyloid-β proteotoxicity. In our experiments, MIA-690 significantly decreased free radical formation (Fig. 6) of HCN-2 cells. Further, this GHRH antagonist dose-dependently augmented GPx1 levels while it did not have a significant effect on SOD1 secretion (Fig. 6). Although, Banks et al. found a slight decrease in GPx expression, this by no means indicates a contradiction. It appears that MIA-690, in sharp contrast to the earlier GHRH antagonist that he used, directly stimulates the protective antioxidant enzymes in the CNS. Therefore, in contrast to our previous experiments, present findings demonstrate that the activity of the anti-oxidative system in the MIA-690 treated animals is not only passively responsive to the free radical burden but actively up-regulates the relevant ROS catabolic enzymes. The oxidative stress elicited by proteo-toxicity inevitably leads to apoptosis [1, 15]. In full agreement with our anti-oxidative experiments, MIA-690 dose-dependently increased the viability of HCN-2 cells treated by amyloid-β_1-42 (Fig. 6). The GHRH antagonist also attenuated the transcription of caspase and clusterin (Table 1.), both of which play important roles in apoptotic processes in the CNS [1, 46]. Further, GHRH antagonist treatment decreased the expression of APP-BPs and the neural precursor cell-expressed and developmentally down-regulated gene (NEDD)8 activating enzyme E1-subunit 1, which both cooperate in neddylation, one of the post-translational tagging processes that can lead to apoptosis [47].

To properly analyze present results and harmonize them with literature data, it is necessary to differentiate between two effects and mechanisms. Inhibition of the different levels of the GHRH-GH-IGF axis apparently exerts a beneficial impact on the progress of Alzheimer's disease [7, 10, 11, 15, 17]. This phenomenon, however, should not be confused with the effect of these mediators on cognitive performance, per se. Concerning Alzheimer's disease, literature data suggest that the inhibition of the axis can be desirable, at least, in the developing disorder. [7, 10, 11, 14-17]. However, regarding the acute CNS effects, the available findings are more ambiguous, and raise the possibility, that, in cognitive tests, an adverse, acute effect on learning can mitigate the consequences of the activities on neuro-degeneration. Accordingly, while GHRH antagonists enhance [7, 10] and exogenous GHRH impairs [12] memory consolidation, IGF-II appears an important acute stimulator of this process [48]. Although, IGF-II is far less dependent on the GHRH-GH axis than IGF-I [49], it is important to emphasize that in our experiments, chronic peptide administration decreased IGF-II expression in the brain samples (Table 1.). Therefore, albeit acute administration of MIA-690 did not influence IGF-II secretion in HCN-2 tissue cultures (data not published), the direct effect of the peptide on memory can interfere with the beneficial actions on neurodegeneration. This hypothesis is strongly supported by the recent findings of Vitiello et al. [50] and Baker et al. [51]. It seems that the activation of the GHRH-GH-IGF-I axis has rejuvenating action on the cognitive performance of the elderly, similar to its somatic effects [52]. Recently it was shown that the mammalian target of Rapamycin (mTOR) pathway is involved in cellular senescence, organ aging and diseases of aging including Alzheimer's disease [53]. The mTOR pathway was also implicated as a molecular link between growth and aging [54]. The GHRH-GH-IGF-I axis may concurrently pathologically activate mTOR [53] and the cascade mechanisms involved in oxidative, inflam-matory, degenerative and neoplastic processes [1, 11, 14, 15, 17, 19]. Interestingly, while a reduction of insulin/IGF-I signaling can result in diabetes, its reduction can also increase longevity and delay the onset of protein-aggregation-mediated toxicity [14]. This phenomenon, called the “insulin paradox”, can be explained by mTOR mechanisms: if low insulin/IGF-I signaling does not activate mTOR, then it is beneficial for longevity and health [55]. However, if insulin/IGF-I signaling is low, because of the blockage caused by the feedback of the over-activated mTOR, then it is harmful [55].

Taken together, our data suggest that further studies are needed to clarify the mechanism of action of GHRH analogs in models of Alzheimer's disease, with special emphasis on the separation between the acute effects on memory formation and the selective impact on proteo-toxicity itself. The direct and indirect, short-term and long-term activities should also be further elucidated, taking into consideration the bewilderingly complex nature of the GHRH-GH-IGF axis. Our compounds freely penetrate the blood-brain barrier [56] and apparently target different levels of the pathologic cascade of Alzheimer's disease inhibiting aggregation and proteo-toxicity while restoring normal neural metabolism and regeneration. In making the choice of which GHRH antagonist to use, focusing on those properties relating to the CNS undoubtedly will have important clinical therapeutic application.

Methods