The present study shows that delayed treatment with minocycline reduces microglial activation and neuronal death in the hippocampus after hypoglycemia. Minocycline treatment also prevented cognitive impairment at later time points. These results suggest that minocycline is an effective therapeutic candidate to prevent neuronal death and subsequent cognitive impairment after hypoglycemia.
The mechanisms of hypoglycemia-induced neuronal injury represent a more complex process than a simple lack of glucose supply to the brain [32, 33]. Rather, several contributing factors are involved in downstream events of hypoglycemia-induced neuronal death such as sustained activation of glutamate receptors , poly(ADP-ribose) polymerase activation , zinc translocation  and NADPH oxidase activation . Although several studies have been reported to intervene in this neuronal death cascade, there is currently no clinically applicable intervention strategy available. Therefore, this study has sought an indirect intervention strategy for preventing hypoglycemia-induced neuronal death and cognitive impairment that may be more clinically relevant.
The brain injury produced by hypoglycemia maturates over a period of several hours or days as seen in ischemia . Especially in hypoglycemia, delayed hippocampal damage is observed 3 to 5 days after the insult in CA1 pyramidal neurons , suggesting that mechanisms that develop slowly after hypoglycemia have a role in hypoglycemic neuronal death. Recent studies have shown that inflammatory cells infiltrate the ischemic area  or the hypoglycemic brain area . Inflammation is now recognized as a significant contributing mechanism in cerebral ischemia because anti-inflammatory compounds or inhibitors of iNOS and cyclooxygenase-2 reduce ischemic damage and improve the outcome of animals after ischemic insult [37, 38].
Microglial activation contributes to ischemia and traumatic brain injury-induced neuronal death. Previously our lab showed microglia are activated after hypoglycemia . Microglia are the major innate immune cells resident in the brain. Once activated by neurological damage or systemic inflammatory events, microglia release neurotoxic substances such as nitric oxide, ROS, cytokines, and chemokines, and undergo morphological change from a ramified to an amoeboid shape [13, 14]. Whether microglial activation is a net harmful or beneficial process is still controversial , however, there is evidence that indicates that early phage inflammation by acute brain injury can contribute to neuronal death. Our previous study shows that hypoglycemia induces microglial activation in the brain, which is affected by body temperature and vesicular zinc release . However, it is unknown whether the prevention of microglial activation by minocycline after hypoglycemia is neuroprotective.
Infiltrating peripheral inflammatory cells play important roles in the development of pathophysiological response following neurological damage. Brain damage, such as ischemic or traumatic injury, triggers physiological changes and neuronal death that induces adhesion of circulating leukocytes, leading to their migration into brain, causing release of pro-inflammatory substances [17, 40]. MPO produced from neutrophils has been used as a marker of infiltrating neutrophils and is involved in brain damage following such events as traumatic brain injury and cerebral ischemia. Numerous studies have been reported in which the accumulation of MPO-positive neutrophils into the ischemic brain is correlated with ischemic brain damage, although MPO gene deletion has been shown to exacerbate brain injury, which is mediated by the peroxynitrite reaction but not MPO [31, 41]. Activation of microglia precedes neutrophil infiltration and seems to play a role in the recruitment of neutrophils following brain damage . In our study, the MPO-positive neutrophils were observed in the hippocampal formation following hypoglycemic brain injury and this recruitment was prevented by minocycline. It suggests that neutrophil infiltration may be involved in brain inflammation and neuronal death after hypoglycemic injury and may be prevented by minocycline treatment.
Although minocycline was developed as an anti-microbial drug for the treatment of various infectious diseases, many other functions such as anti-apoptotic and anti-inflammatory effects have been identified . In particular the anti-inflammatory properties of minocycline have been observed in acute and chronic neurological disease animal models, as well as in human clinical trial studies [22, 23, 44]. In an animal model of multiple sclerosis, minocycline decreased the transmigration of T lymphocytes and inhibited the activation of metalloproteinases that degrade the extracellular matrix proteins of the basal lamina surrounding blood vessels, causing neutrophil infiltration [45, 46]. Based on these studies, our present results suggest that delayed treatment with minocycline can have a neuroprotective effect on hypoglycemic neuronal death by inhibiting microglial activation and neutrophil infiltration.
Since most hypoglycemic patients visit the emergency room several hours after the hypoglycemic episode, we delivered the initial dose of minocycline 6 hours after hypoglycemic insult. Microglial activation was detected in the hippocampus 24 hours after hypoglycemia. Thus we believe that treatment of 6-hours post-hypogecemic insult is a reasonable therapeutic window. Although we injected minocycline from 6 hours after hypoglycemia, microglial activation was significantly reduced, as was neuronal death. These results suggest that delayed treatment with minocycline suppressed microglial activation, which may decrease release of toxic substances such as nitric oxide, and IL6, etcetera. This further suggests that acute microglial activation after hypoglycemia is detrimental to neuronal survival.
Since minocycline is one of the most lipid-soluble tetracycline-class antibiotics, it can easily penetrate into the brain. Minocycline also has a long half-life compared to other tetracycline antibiotics . Thus, one or two doses of 50 to 100 mg per day of minocycline are effective in many patients to treat bacterial infection. A recent clinical study found that patients who received 200 mg of minocycline for five days within 24 hours after ischemic stroke showed significantly better outcome compared with patients receiving placebo . In the present study, we used 50 mg/kg per day for one week in rats. We understand that this concentration of minocycline is fairly high for a single dose. Therefore, for clinical applications, a single dose of minocycline for prevention of hypoglycemia-induced neuronal death should be re-evaluated.
Learning and memory deficits are common neurological sequelae following hypoglycemia in patients with type 1 diabetes and in the relatively younger population with type 2 diabetes [49–51]. Our previous study showed that hypoglycemia-induced hippocampal damage induced impairment of learning and memory . In the present study, the Barnes maze test was performed to evaluate spatial learning and memory. As seen in our previous study using the water maze test, subjects experiencing severe hypoglycemia displayed a longer distance traveled to reach the escape tunnel, implying that spatial learning has been compromised. However, minocycline treatment reduced the travel distance in rats who experienced hypoglycemia. It has been reported that minocycline improves cognitive impairment in focal cerebral ischemia , Alzheimer’s disease models , and other animal models of neurological disease [54, 55].
Because tetracycline derivatives, like minocycline, have anti-inflammatory properties and are clinically well tolerated, we studied whether minocycline could serve as a neuroprotective compound against hypoglycemia-induced brain injury. In the present study, we report that in a rat model of hypoglycemia, 1) minocycline is neuroprotective, even when the treatment is initiated 6 hours after hypoglycemia; 2) minocycline prevents microglial activation after hypoglycemia; and 3) minocycline prevents cognitive impairment even at several weeks after hypoglycemia. Thus, the present study suggests that prevention of microglial activation by minocycline has a strong therapeutic potential for prevention of hypoglycemia-induced brain injury.