Date Published: May 31, 2012
Publisher: Impact Journals LLC
Author(s): Mikhail V. Blagosklonny.
Calorie restriction (CR), which deactivates the nutrient-sensing mTOR pathway, slows down aging and prevents age-related diseases such as type II diabetes. Compared with CR, rapamycin more efficiently inhibits mTOR. Noteworthy, severe CR and starvation cause a reversible condition known as “starvation diabetes.” As was already discussed, chronic administration of rapamycin can cause a similar condition in some animal models. A recent paper published in Science reported that chronic treatment with rapamycin causes a diabetes-like condition in mice by indirectly inhibiting mTOR complex 2. Here I introduce the notion of benevolent diabetes and discuss whether starvation-like effects of chronic high dose treatment with rapamycin are an obstacle for its use as an anti-aging drug.
If you read the Abstract, you might wonder whether rapamycin extends lifespan despite or because of “starvation-like diabetes”. As described by Lamming et al [1, 2] extending several previous observations [3-6], chronic administration of high doses of rapamycin causes insulin resistance in mice. Yet, at similar doses, rapamycin prolongs life span in mice [7, 8]. Moreover, in several studies, rapamycin prevented complications of diabetes such as nephropathy [9-14]. Also, theoretical considerations indicate rapamycin for retinopathy , which was recently confirmed in an animal model . Rapamycin prevents atherosclerosis in rodents [17-20] and coronary re-stenosis in humans [21, 22]. In contrast, diabetes promotes nephropathy, retinopathy, atherosclerosis and coronary disease. How could this be reconciled? mTOR is a part of a nutrient-sensing pathway [23-27]. Nutrients and insulin activate mTOR. Rapamycin, which inhibits mTOR, is a “starvation-mimetic”, making the organism “think” that food is in a short supply. The most starvation-sensitive organ is the brain. The brain consumes only glucose and ketones. Therefore, to feed the brain during starvation, the liver produces glucose from amino acids (gluconeogenesis) and ketones from fatty acids (ketogenesis). Since insulin blocks both processes, the liver needs to become resistant to insulin. Also secretion of insulin by beta-cells is decreased. And adipocytes release fatty acids (lipolysis) to fuel ketogenesis by the liver. Thus, there are five noticeable metabolic alterations of starvation: gluconeogenesis, ketogenesis, insulin resistance, low insulin levels and increased lipolysis. This metabolic switch is known as starvation diabetes, a reversible condition, described 160 years ago (see for references ). Starvation diabetes could be explained by deactivation of mTOR, which otherwise is activated by nutrients. In theory, rapamycin can cause similar symptoms in the presence of nutrients.
Starvation-diabetes is not a true type II diabetes. Type II diabetes is a consequence of insulin-resistance in part due to excessive nutrients and obesity. Even brief overfeeding may induce insulin resistance . Nutrients and insulin activate mTOR. In turn, over-activated mTOR causes insulin resistance [30-42]. This feedback loop is shown in figure 1A. mTOR activates S6 kinase (S6K), which causes degradation of insulin-receptor substrates (IRS), thus impairing insulin signaling. Also, mTOR causes insulin resistance by an additional feedback mechanism [43, 44].
Type II diabetes and starvation diabetes seem to be the two opposite conditions: the first is associated with activation of nutrient-sensing pathways, whereas the second is associated with deactivation of nutrient sensing pathways such as mTOR. Type II diabetes is dangerous by its complications such as retinopathy, neuropathy and accelerated atherosclerosis and cancer. Long-term effects of prolonged “starvation diabetes” is not known of course: it could not last for a long time, otherwise an animal (or human) would die from starvation. Or would not? An outstanding study by Fontana et al provides some answers . Among individuals who had been practicing sever CR for an average of 7 years, 40% of CR individuals exhibited “diabetic-like” glucose intolerance, despite low levels of fasting glucose, insulin and inflammatory cytokines as well as excellent other metabolic profiles. In comparison with the rest CR individuals, they had lower BMI, leptin, circulating IGF-I, testosterone, and high levels of adiponectin, which are key adoptations to CR in rodents, suggesting severe CR . The authors speculated that the “insulin resistance” in this severe CR group might have the effect of slowing aging, also based on the finding that a number of insulin-resistant strains of mice are long-lived . The same conclusion could be reached from the mTOR perspective (Appendix 1).
There are two types of diabetes, which at advanced stages may become similar. Insulin resistance may develop in type I diabetes (due to high glucose), whereas insulin insufficiency in type II diabetes (due to loss of beta-cells). Both types of diabetes lead to complications. In comparison, starvation diabetes  is only superficially resembles either type of diabetes. Also, diabetes-like symptoms may occur in rapamycin-treated mice and animals with genetically inhibited insulin/IGFI signaling (Fig. 3). To encompass all these cases, I suggest the term type 0 (zero) or benevolent diabetes. It is possible that some patients with diabetes have inactivating mutations in the insulin/IGFI pathway and thus “suffer” from benevolent diabetes. Furthermore, the condition can be imitated by chronic administration of rapamycin at least in some strains of mice. Both calorie restriction and rapamycin extend life span in mice. Rapamycin prevents retinopathy and nephropathy. Also CR prevents type II diabetes and other diseases , , , . One can suggest that type 0 diabetes should prevent type 2 diabetes. Should type 0 diabetes be treated? Perhaps CR-associated type 0 diabetes should not. What about rapamycin-associated diabetes? Definitely, it should not be treated with insulin. It was discussed that in theory the most rational combinations with rapamycin are mild calorie and fat restriction, physical exercise and metformin . Metformin may in theory counteract rapamycin-induced gluconeogenesis in the liver. And this rational drug combination may be also considered as treatment of type 0 diabetes.
As demonstrated by Lamming et al, chronic administration of rapamycin caused insulin-resistance due to deactivation of mTORC2 and Akt . This is consistent with previous data that IRS signaling and AKT activation was impaired in patients treated with rapamycin . However, there are some inconsis-tencies. In another clinical study, rapamycin therapy in contrast caused activation of Akt . Second, whereas Lamming et al found that rapamycin increased insulin levels after feeding , other studies reported that rapamycin in contrast inhibited insulin secretion [3, 4, 65]. Furthermore, inhibition of beta-cell adaptation and insulin production by rapamycin was considered as the main mechanism of rapamycin-induced diabetes in mice [6, 66-69]. On the other hand, selective inactivation of mTORC2 in the liver can cause hyperinsulinemia .
In renal transplant patients, who are prone to diabetes (due to several reasons), chronic administration of rapamycin modestly increases incidence of diabetes [71, 72]. Although the increase is statistically significant, it took many years to detect it. For many years it was thought that, unlike other agents used in these patients, rapamycin either do not increase the incidence of diabetes or increases it in combinations with tacrolimus [73-79]. In the study involving 20124 recipients of kidney transplant sirolimus (rapamycin) was independently associated with new onset diabetes . And although it statistically significantly increases the incidence of diabetes in renal transplant patient, we do not know whether this is true diabetes, which is dangerous by its complications, or starvation-like diabetes, that prevents the complications of true diabetes . Will chronic high doses of rapamycin cause or prevent diabetes in humans without organ transplantation? More investigations are needed.
Is glucose intolerance a part of therapeutic effects of starvation-like drugs such as rapamycin? And may such condition be not only benign but also prevent true diabetes and its complications? Although these questions are very intriguing, the answers are not immediately crucial. Simply, the most rational anti-aging schedule is an intermittent (rather than chronic) administration of rapamycin [53, 80]. First, this will eliminate potential side effects. Second, intermittent administration of rapamycin may in theory rejuvenate stem and wound-healing cells and (in contrast to chronic treatment) improve wound healing . And intermittent administration of rapamycin extended life span in mice [81-86]. Also, brief treatment with rapamycin does not affect mTORC2 .