Our research shows that metformin has a protective role in neuronal damage caused by glutamate. It is suggested that this neuroprotective effect is mediated by a zinc-dependent inhibition of GSK-3β.
Our histopathological findings indicate that significant damage occurred in the cortical and hippocampal regions of rats administered D-glutamic acid. It was concluded that the neuronal degeneration caused by D-glutamic acid was even more prominent in rats with zinc deficiency (Table 4) and this is parallel with the studies [14, 33, 34]. Some studies have reported that zinc inhibits NMDA receptors [35, 36] and glutamate release [37, 38], by interacting with the zinc-binding site on the post-synaptic and pre-synaptic receptors respectively. Thus, the disappearance of inhibition on NMDA receptors and the increase in glutamate secretion from the presynaptic neuron in zinc deficiency may be responsible for the increased damage in the cortical and hippocampal regions of the rats fed with a zinc-deficient diet. Also, Ilouz et al. have shown that the increase in zinc in the cell inhibits GSK-3β [15]. The occurrence of similar events in zinc deficiency and GSK-3β activation may be related to increased GSK-3β activity in zinc deficiency.
In our study, it was determined that there were significant increases in active GSK-3β levels in both blood and brain tissue of rats in groups with D-glutamic acid induced neurodegeneration (G and ZG). Similarly, Hanumanthappa et al. suggest that GSK-3β plays a role in brain damage with increased glutamate secretion as a result of ischemia [9].
Our research suggests that metformin has a neuroprotective effect by inhibiting GSK-3β activity in neurodegeneration caused by glutamate increase. Our histopathological findings reveal that glutamate-induced neurodegeneration (Table 4) and active GSK-3β levels were significantly reduced (Table 2) in metformin-treated groups compared to the non-treated groups. Similarly, Leng et al. showed that inhibition of GSK-3β is protective against cerebral ischemia and prevents glutamate-induced neurodegeneration [39, 40]. In another study, Cross et al. showed that inhibition of GSK-3β weakened apoptotic signals and prevented neuronal death [41].
In addition, intracellular S-100β staining rates, which is an indicator of neuronal survival, were found to be higher in metformin-treated groups (MG, ZMG) than in non-treated groups (G, ZG). (Table 3). Our findings also reveal that these protective and survival effects of metformin on neurons are less in the groups fed with a zinc-deficient diet than in the groups fed with a normal diet (Table 3). These findings suggest that metformin inhibits GSK-3β not only through AMPK-Akt pathways but also by using zinc-dependent pathways. Similarly, Miranda et al. suggested that IRS-1 phosphorylation, which causes Akt activation, may occur not only through AMPK but also directly through intracellular zinc [42]. Metformin may cause zinc to contribute additionally to IRS-1 phosphorylation at this stage. Therefore, the decrease in intracellular zinc levels in rats fed a deficient zinc diet may be responsible for the lower protective and survival effects of metformin. On the other hand, S6K, a zinc-dependent protein kinase, has been reported to inhibit GSK-3β in some studies, while to cause activation in some studies [43, 44]. It has been suggested that the effect of S6K on GSK-3β may differ between tissues [45]. We think that metformin can create GSK-3β inhibition by directly increasing S6K expression in the CNS. Further studies are needed to investigate the effect of metformin on the expression or activation of S6K and its relationship with GSK-3β.
In our study, the increase in blood AGE levels in D-glutamic acid administrated rats is remarkable. Huang et al. stated that increased glutamate may impair the function of β-cells and accelerates their apoptosis by over-activation of NMDA receptors [46]. Also, Shen et al. suggested that inhibition of GSK-3β induces human pancreatic β-cell proliferation [47]. In line with this information, it is suggested that over-stimulation of NMDA receptors in pancreatic beta cells may be responsible for increased blood AGE levels by increasing active GSK-3β levels in D-glutamic acid administrated rats.
AGEs activate reactive oxygen species (ROS) and advanced glycation end-products receptors (AGER) that produce an inflammatory response [48]. This interaction between the AGERs and AGEs causes an increase in the synthesis of the superoxide radical anion. This reduces the activity of catalase and superoxide dismutase [48]. We think that the decrease in Cu-Zn SOD and catalase activities in the blood of D-glutamic acid administrated rats may have developed due to the increased use of these antioxidants in order to remove the increased ROS in peripheral tissues.
Our findings suggest that the increase in oxidative stress observed in glutamate administrated rats (G, ZG) may be related to the increased active GSK-3β levels (Tables 2 and 5). Similarly, Li et al. have reported that GSK-3β increases ROS formation by causing mitochondrial dysfunction in neurons [49]. On the other hand, oxidative stress parameters in blood and tissue were found to be significantly decreased in rats treated with metformin (GM, ZGM) compared to rats not treated (G, ZG). It was determined that there were also significant increases in Cu-Zn SOD and catalase activities in the groups treated with metformin. This suggests that the protective effects of metformin are due to decreased ROS production and the regulation of the balance between pro-oxidants and antioxidants. These findings are parallel to other studies suggesting that metformin has a similar effect in other tissues [50, 51]. Similarly, Zhao et al. suggested that metformin treatment provided cognitive protection in mice with pentylenetetrazole (PTZ)-induced epilepsy models and showed antioxidant properties by reducing oxidative stress [52].
Finally, it is remarkable in our findings that metformin administration to rats fed an insufficient zinc diet could not reduce lipid peroxidation caused by glutamate toxicity in brain tissue. This shows us that metformin needs zinc to prevent glutamate-induced lipid peroxidation, especially in brain tissue.
In conclusion, our study is the first to investigate the relationship of metformin with GSK-3β and zinc in the brain. We can say that the increase in glutamate secretion may cause an increase in oxidative stress via GSK-3β. Our findings confirm that metformin can inhibit glutamate-mediated neurodegeneration and increase neuronal survival by reducing the active GSK-3β levels with zinc-dependent pathways. Further studies are needed to investigate which advanced intracellular pathways metformin effects on GSK-3β in brain tissue.
Limitations
We could not perform measurement of catalase activity in tissue samples due to technical problems during storage but considering other results, we think that the results in the tissue samples would be parallel to results obtained from the serum samples.