ALZHEIMER-LIKE CHANGES OF PROTEIN KINASE B AND GLYCOGEN SYNTHASE KINASE-3 IN RAT FRONTAL CORTEX AND HIPPOCAMPUS AFTER DAMAGE TO THE INSULIN SIGNALLING PATHWAY

Insulin and IR signalling participate in a variety of region-specific functions in the central nervous system, through mechanisms not necessarily associated with glucose regulation (Schulingkamp et al., 2000). Among them, growing evidence suggest that IR signalling modulates neuronal excitability and synaptic plasticity (Skeberdies et al., 2001; Wan et al., 1997), consequently affecting cognitive functions like learning and memory (Zhao et al., 2004). In line with this, deterioration of IR signalling has been found to be associated with sporadic Alzheimer’s disease (Hoyer, 2002; Hoyer and Frőlich, 2005). However, IR and its signalling cascade have not been investigated either in STZ-icv treated rats as an experimental model probably related to the sporadic Alzheimer’s disease, or in a model of the disturbed brain glucose sensing system.

Read Also

In the hippocampus of STZ-icv treated rats we found alterations at the level of Act/PKB - GSK-3 enzymes that are downstream the PI-3 kinase pathway of the IR signalling cascade. The level of phosphorylated GSK-3α/β enzyme was significantly increased at one month following the treatment, and fell down below the control values at three months following the treatment. However, these changes were not followed by alterations of non-phosphorylated GSK-3α/β, which remained unchanged. Considering brain GSK-3 research in Alzheimer’s disease in humans, inconsistent results have been reported; increased brain (Pei et al., 1997) and unchanged hippocampal and hypothalamic (Steen et al., 2005) levels of total GSK-3α/β protein, unchanged cortical GSK-3α mRNA levels (Preece et al., 2003), unchanged hippocampal and hypothalamic GSK-3β levels (Steen et al., 2005), and also unchanged (Pei et al., 1997), reduced (Swatton et al., 2004) and increased (Steen et al., 2005) GSK-3α/β activity. In respect to animal models, a significant increase in GSK-3 activity has been reported in the hippocampus of Tet/GSK-3beta transgenic mice that have also been proposed as an experimental model of Alzheimer’s disease (Hernandez et al., 2002). However, our results are the first report of GSK-3 investigation in the brain of the STZ-icv rat experimental model, probably related to the sporadic type of Alzheimer’s disease. Changes found in our experiments could suggest some effects at the level of GSK-3 phosphorylation or pGSK-3 dephosphorylation, since total GSK-3α/β levels were unchanged, both in the hippocampus and in the frontal cortex, corresponding to the findings in human Alzheimer’s disease (Steen et al., 2005), whereas the level of pGSK-3 was increased. Decrement of the relative pGSK-3/GSK-3 ratio in hippocampal tissue of STZ icv treated rats from the first month-measured over-control level (+50%) to the third month-measured bellow-control level (-9%) could indirectly suggests increment of non-phosphorylated, therefore, active GSK-3 form (Fig. 3) with the duration of observational period. This tendency of decrement of phospho/total enzyme ratio was demonstrated in both, alpha and beta GSK-3 isoform, and seemed to be particularly pronounced after the third month in the former isoform (Fig. 3). However, the activity of GSK-3, whose alterations could not be excluded, has not been measured in our experiments. Increased ratio after first month in hippocampus could represent an acute change which, in line with decreased ratio after the third month, could not be compensated with the time in STZ icv rat model, as observed in both isoforms, particularly in GSK-3 alpha one. Indirectly suggested increment in non-phosphorylated, active GSK-3 in the brain of STZ icv rat model could lead to tau hyperphosphorylation. Our preliminary experiments have indeed shown some alterations at the level of tau protein, the meaning of which is not clear yet, but more extensive analyses of tau protein in STZ icv rat model, needed to clarify this issue, are in preparation.

The phospho-GSK-3α/β antibody used in our Western blot analysis detects endogenous levels of GSK-3 when phosphorylated at Ser 21 of GSK-3α or Ser9 of GSK-3β. Ser21/9 phosphorylation and thus inactivation of GSK-3 are mediated by Akt/PKB that is downstream the IR-activated PI3 kinase pathway (Cross et al., 1995). Unchanged level of total Akt/PKB in the hippocampal tissue samples after one month, and a small decrease after three months as well as mild (~+/-10%) changes of this protein in the frontal cortex of the STZ treated rats may suggest this protein itself to be quite resistant to damage in the course of the 3-month period of observation in this experimental model. Consequently, this may have a reflection on total GSK-3 protein, which, in line with this, we found to be unchanged. These results are in agreement with literature data on unchanged Akt/PKB levels in the brain of patients with Alzheimer’s disease post mortem (Steen et al., 2005). The same authors report on decreased levels of pAkt/PKB and pGSK-3, which were also found to have moderately decreased during the 3-month course of the probable experimental Alzheimer’s disease in our experiments. This could suggest the possible relation to differences in the disease staging and severity. Our STZ icv (1 mg) rat model refers to the early pathological changes, contrary to data obtained post-mortem from humans with, in general, severe, end-stage of sporadic AD disease. The issue of early pathological changes could involve the different velocity of structural changes development in STZ icv rat experimental model in comparison to humans. Some structural changes at the level of beta-amyloid peptide accumulation, resembling those in human AD, have been observed in STZ icv rat model. The conformational transition of beta-amyloid peptides from alpha-helices to beta-sheets strongly favours the formation of beta-amyloid fibrils, which give rise of pathological protein aggregates called amyloid plaques. Beta-amyloid fibrils stained by Congo red show green autofluorescence on cross-polarization. Congo red, which is known to specifically bind to beta-amyloid fibrils (Klunk et al., 1989; Balbirnie et al., 2001), has widely been used in histological staining procedures for the evaluation of beta-amyloid aggregates in human (Ladewig, 1945) and murine tissues (Li et al., 2005). A possible relationship of these structural changes to decrement of pGSK-3 α/GSK-3 α ratio which suggests increment of non-phosphorylated, active GSK-3 α known to be involved in Aβ regulation (Phiel et al., 2003) can not be excluded. Dose-dependent severity of STZ-induced effects have been demonstrated following its peripheral (persistent or transient diabetes, morphological alterations of the islet insulin-immunoreactive cells /Ar’Rajab et al, 1993; Junod et al., 1969/), and central (neurochemical alterations of brain monoamine level /Lackovic and Salkovic, 1990/) administration, supporting this stage- and severity-dependent hypothesis of alteration manifestation. Furthermore, regarding human Alzheimer’s disease, a statistically significant positive correlation seen in the human tissue between Akt/PKB activities or pAkt/PKB levels and Braak staging for the neurofibrillary changes supports this explanation (Pei et al., 2003, Rickle et al.,2004). Increased Akt/PKB protein levels found in frontal cortex (Pei et al., 2003) and increased Akt/PKB activity found in temporal cortex but not in frontal cortex (Rickle et al., 2004) of patients with Alzheimer’s disease post mortem, suggest the possible regional pattern of changes. Our results of highly increased pGSK-3α/β level found after one month in the hippocampus but not in the frontal cortex as well as of decreased Akt/PKB level in the hippocampus but increased in the frontal cortex after three months in STZ-icv treated rats are consistent with these reports.

We did not measure the activity of Akt/PKB, and it can not be excluded that the increased level of hippocampal pGSK-3α/β found one month after STZ-icv treatment was the consequence of increased Akt/PKB activity, as reported elsewhere (Rickle et al., 2004). However, beside the Akt/PKB, numerous kinases can phosphorylate GSK-3β at Ser9, such as protein kinase C, involved in signalling of G-protein linked receptors (Kaytor and Orr, 2002). Furthermore, the increased pGSK-3 level could be related to inactivity and/or decreased levels of phosphatase that dephosphorylates GSK-3, among which serine/threonine protein phosphatase 1 (PP1) and 2A (PP2A) has been mentioned (Bennecib et al., 2000; Milward et al., 1999; Hoyer and Frőlich, 2005). PP2A is a negative regulator of the insulin PI3/Akt/PKB signalling pathway that dephosphorylates and thereby inactivates Akt/PKB, and to a minor extent dephosphorylates, and thereby activates GSK-3 (Milward et al., 1999). PP2A mRNA expression was found to be significantly reduced in the hippocampus of sporadic Alzheimer’s disease brain (Vogelsberg-Ragaglia et al., 2001). Immunoblotting analyses revealed a significant reduction in the total amount of PP2A in frontal and temporal cortex that matched the decrease in PP2A activity in the same region, and was further supported by the finding of lower PP2A expression in immunohistochemical studies of the brains from patients with Alzheimer’s disease (Gong et al., 1995; Sontag et al., 2004). A recent finding of up-regulation of the endogenous PP2A inhibitors in the neocortex of patients with Alzheimer’s disease further supports this hypothesis (Tanimukai et al., 2005). Therefore, it could not be excluded that in STZ-icv experimental model, some neurochemical changes are related to the possible lower activity/protein level of PP2A, not investigated in STZ-icv treated rats so far. Furthermore, GSK-3β is involved in phosphorylation of tau protein, which in the hyperphosphorylated form builds neurofibrillary tangles, important pathological features of Alzheimer’s disease (Kaytor and Orr, 2002). Interestingly, recent data demonstrate that involvement of GSK-3 is not necessary to obtain hyperphosphorylated tau in vivo, indicating that inhibition of PP2A, an enzyme that can directly dephosphorylate tau, is likely the predominant factor in inducing tau hyperphosphorylation (Planel et al., 2001).

In line with literature reports of STZ-icv induced cognitive deficits (Hoyer, 2004; Lannert and Hoyer, 1998; Prickaerts et al., 1999; Sharma and Gupta, 2001), a decreased memory function was also found in STZ-icv treated rats in our experiments, demonstrated as a decline in time spent in search for the hidden platform within the appropriate quadrant where the platform had been placed in training trials.

The results of our experiments with GLUT2 blocker, i.e. (TG) icv treatment came as a surprise. A single TG-icv treatment induced longlasting neurochemical effects in the hippocampus, which mostly resembled those induced by STZ-icv treatment, e.g., a tendency to increase in pGSK-3α/β level after one month, unchanged GSK-3α/β levels during the 3-month observation period, and a decrease in Akt/PKB level after three months, which was more pronounced than the one induced by STZ at the same time. Thus, it could be speculated that by blocking the intracellular glucose uptake and consequently its intracellular metabolism and possible glucose sensing, TG-icv treatment induced local conditions in the brain that could be similar to the impaired brain glucose uptake and metabolism found in human sporadic Alzheimer’s disease and in STZ-icv treated rats proposed as a probable experimental model of this disease, as reviewed elsewhere (Hoyer, 2004; Hoyer and Frőlich, 2005). The finding of cognitive deficits on Morris Water Maze Swimming Test in TG-icv treated rats, which were similar or even more severe than in STZ-icv treated rats after one month, supported this hypothesis. These results are in agreement with the finding that 3 weeks after STZ-icv injection the ultrastructure of rat frontoparietal cortical neurons was similar to that observed after iv application of non-metabolizable glucose analogue 2-deoxyglucose (Grieb et al., 2004). A small improvement in cognitive deficits in comparison to STZ-icv treatment, but still persistent deficits in comparison with control treatment, seen at three months of TG-icv treatment, could suggest involvement of some compensatory mechanisms and factors yet unexplored in such an experimental model.

In conclusion, STZ-icv probably induces experimental, sporadic Alzheimer’s disease in rats that is associated with acute increase in pGSK-3α/β level and subsequent decreasing tendency in pGSK-3α/β and Akt/PKB levels in the hippocampus. No changes or less intensive changes found in the same periods of observation in frontal cortex suggest regional specificity of changes in this probable experimental model of Alzheimer’s disease. The icv treatment with the blocker of GLUT2, a glucose transporter suggested to be related with brain glucose sensing, induces neurochemical changes and cognitive deficits that are, in general, similar to those induced by STZ-icv treatment. This is the first report of altered IR-PI3 kinase downstream signalling pathway in STZ-icv rats that supports the hypothesis of the STZ-icv rats being a probable experimental model of sporadic Alzheimer’s disease. Also, a possible role of GLUT2 in the pathophysiology of sporadic Alzheimer’s disease, at least in its probable experimental animal models, has been suggested.