Leptin administration restores euglycemia in rodents with severe insulin-deficient diabetes, and recent studies to explain this phenomenon have focused on the ability of leptin to normalize excessive hypothalamic-pituitary-adrenal (HPA) axis activity. Here, we employed a streptozotocin-induced rat model (STZ-DM) of uncontrolled insulin-deficient diabetes mellitus (uDM) to investigate the contribution of HPA axis suppression to leptin-mediated glucose lowering. Specifically, we asked if HPA axis activation is required for diabetic hyperglycemia, whether HPA axis normalization can be achieved using a dose of leptin below that needed to normalize glycemia, and if the ability of leptin to lower plasma glucocorticoid levels is required for its antidiabetic action. In STZ-DM rats, neither adrenalectomy-induced (ADX-induced) glucocorticoid deficiency nor pharmacological glucocorticoid receptor blockade lowered elevated blood glucose levels. Although elevated plasma levels of corticosterone were normalized by i.v. leptin infusion at a dose that raises low plasma levels into the physiological range, diabetic hyperglycemia was not altered. Lastly, the potent glucose-lowering effect of continuous intracerebroventricular leptin infusion was not impacted by systemic administration of corticosterone at a dose that maintained elevated plasma levels characteristic of STZ-DM. We conclude that, although restoring low plasma leptin levels into the physiological range effectively normalizes increased HPA axis activity in rats with uDM, this effect is neither necessary nor sufficient to explain leptin’s antidiabetic action.
Gregory J. Morton, Thomas H. Meek, Miles E. Matsen, Michael W. Schwartz
Submitter: Gerald Shulman | gerald.shulman@yale.edu
Authors: Gerald Shulman and Rachel J. Perry
Yale University School of Medicine
Published January 20, 2016
We read the study by Morton et al. (1) with great interest as it tests a subset of our findings that leptin’s effect to reverse fasting hyperglycemia and ketoacidosis in an insulin-independent manner occurs through suppression of the hypothalamic-pituitary-adrenal (HPA) axis, reducing rates of lipolysis and hepatic gluconeogenesis (2). Although these authors found, as we do, that leptin treatment ameliorates hypercorticosteronemia in a rodent model of poorly controlled type 1 diabetes, we were surprised that they did not see an acute effect of this intervention to lower plasma glucose concentrations. However on further review we believe that the differences between their findings and ours stem primarily from differences in models: our streptozotocin (STZ)-treated diabetic rats were studied in the fasted state with food withdrawn ~16 hours prior to study, whereas Morton et al. examined the effect of leptin in STZ-treated diabetic rats under postprandial conditions with food withdrawn for only 3 hours. Following a 16 hour overnight fast, plasma glucose concentrations are determined mostly by rates of hepatic gluconeogenesis. In contrast under postprandial conditions plasma glucose concentrations are mostly determined by insulin’s ability to promote glucose uptake in liver and skeletal muscle as well as by glucose absorption from the gut. In contrast to our study, Morton et al. did not perform any glucose, fatty acid, glycerol or acetate turnover measurements nor did they assess rates of hepatic gluconeogenesis so it is not possible to determine whether these factors were altered in their leptin-treated diabetic animals. Another important difference in the models studied is in the plasma insulin concentrations; their diabetic rats had plasma insulin concentrations approximately 5-fold higher (~10 microU/mL) than those in our study (~2 microU/mL). The adipocyte is exquisitely sensitive to insulin, and even these low plasma insulin concentrations may be sufficient to suppress lipolysis enough that leptin suppression of the HPA axis may not have any additional effects. This is reflected by the fact that our rats were in diabetic ketoacidosis and would not survive more than 24 hours without insulin or leptin treatment. Although Morton et al. did not measure plasma ketones or anion gap in their diabetic rats, the fact that their diabetic animals were able to survive for several days without insulin or leptin treatment strongly suggests that there was minimal ketoacidosis. Further studies will be required to assess the contributions of these methodological differences to the disparities between Morton et al.’s findings and ours.
References
1. Morton GJ, Meek TH, Matsen ME, and Schwartz MW. Evidence against hypothalamic-pituitary-adrenal axis suppression in the antidiabetic action of leptin. J. Clin Invest. 2015;125(12):4587–4591.
2. Perry RJ, Zhang XM, Zhang D, Kumashiro N, Camporez JP, Cline GW, Rothman DL, and Shulman GI. Leptin reverses diabetes by suppression of the hypothalamic-pituitary-adrenal axis. Nature Medicine. 2014;20(7):759-63.
Submitter: Michael W. Schwartz | mschwart@uw.edu
Authors: Michael W. Schwartz, Gregory J. Morton, Thomas H. Meek, Miles E. Matsen Title:
University of Washington School of Medicine
Published January 20, 2016
Shulman and colleagues seek to explain differences between our recent findings (1) and their earlier report (2). Both papers confirm earlier evidence that hypothalamic-pituitary-adrenal (HPA) axis activation in rats with uncontrolled diabetes induced by streptozotocin (STZ) is a consequence of leptin deficiency, since it is blocked by physiological leptin replacement. Whereas they reported that 1) physiological leptin replacement normalizes diabetic hyperglycemia, and 2) this effect is mediated via HPA axis suppression, our results indicate that HPA axis suppression is neither necessary nor sufficient to account for leptin’s ability to normalize diabetic hyperglycemia, consistent with a recent report (3).
To explain the different outcomes, Dr. Shulman states, “…our STZ-treated diabetic rats were studied in the fasted state for ~16 hours… whereas Morton et al. examined the effect of leptin in STZ-treated diabetic rats under postprandial conditions with food withdrawn for only 3 hours.” On the contrary, the Methods section of our paper (https://www.jci.org/articles/view/82723/sd/1) clearly indicates that the fasting protocol that we employed is identical to theirs. (We did employ a 3-hour fast prior to measurement of plasma corticosterone levels to confirm successful adrenalactomy surgery). Thus, a difference in fasting duration cannot be invoked to explain the strikingly divergent outcomes.
For this reason, Dr. Shulman’s attribution of relatively lower plasma insulin levels in their STZ-diabetic rats to a difference in fasting duration is also incorrect. Instead, this difference likely reflects the higher dose of STZ that they employed (75 mg/kg/iv vs. 65 mg/kg/sc in our studies). Shulman reasons that despite comparable hyperglycemia, adipocyte lipolysis was relatively suppressed in our study by higher insulin levels. Because lipolysis fuels hepatic gluconeogenesis, Shulman posits that this effect in turn obscured the ability of leptin-mediated HPA axis suppression to inhibit hepatic glucose production. In advance of studies to test this hypothesis, it’s important to point out that differences in diabetes severity may also have contributed to the difference in outcomes. As Shulman et al. note in their letter, the STZ dose that they used is so high as to cause death within 24-48h, and our unpublished experience bears this out as well. We speculate that in the hours preceding death, the extreme duress of these animals heightens their reliance on the HPA axis to support glucose production, a scenario unlikely to apply to the more typical STZ-diabetic model in which rats survive untreated for several weeks. The larger point is that 1) a fundamental difference in disease severity exists between the animal models studied, and 2) this difference likely contributed to the difference in outcomes. Consequently, we are concerned that findings reported by the Shulman group are unique to animals in extremis from severe diabetic ketoacidosis, and hence cannot be generalized to other diabetes models. This possibility warrants additional study.
References
1. Morton, G.J., Meek, T.H., Matsen, M.E., and Schwartz, M.W. 2015. Evidence against hypothalamic-pituitary-adrenal axis suppression in the antidiabetic action of leptin. J. Clin Invest. 2015;125(12):4587–4591.
2. Perry, R.J., Zhang, X.M., Zhang, D., Kumashiro, N., Camporez, J.P., Cline, G.W., Rothman, D.L., and Shulman, G.I. Leptin reverses diabetes by suppression of the hypothalamic-pituitary-adrenal axis. Nature Medicine. 2014; 20:759-763.
3. Neumann, U.H., Denroche, H.C., Mojibian, M., Covey, S.D., and Kieffer, T.J. 2015. Insulin knockout mice have extended survival but volatile blood glucose levels on leptin therapy. Endocrinology epub online. DOI: http://dx.doi.org/10.1210/en.2015-1890.