The current study demonstrated that chronic treatment with GR antagonist RU486 for 10 days dose-dependently decreased postprandial blood glucose, increased serum corticosterone and ACTH, and did not affect serum MCP-1 and IL-6, in db/db mice. Injection of LPS increased circulating TNFα, corticosterone, and ACTH in both normal AKR mice and C57BL/6 DIO mice. RU486 suppressed the lower dose LPS-elevated serum TNFα in both normal AKR mice and C57BL/6 DIO mice, and further increased circulating corticosterone. The RU486 pretreatment did not affect the high dose LPS (20 mg/kg)-mediated changes in serum TNFα and corticosterone. It is worth noting that the data sets from each strain/model examined were not exactly the same in current studies. But all the studies showed that RU486 either decreased cytokine along with increased corticosterone, or did not affect cytokine along with or without change in corticosterone. Thus RU486, at doses which are efficacious on glucose lowering, did not show any exacerbating effect on cytokine production in both chronic and acute settings, and RU486 could actually suppress the lower dose LPS-mediated TNFα release, possibly due to the increased release of GCs. The interpretation of the data sets were discussed below.
Although GR antagonism has been shown to inhibit gluconeogenesis and lower glucose in humans [1
] and in rodent model of T2D [23
], the concern about GR antagonists on pro-inflammation hampers the long-term treatment with GR antagonists in T2D. The current study used db/db mice to determine the effect of chronic RU486 on circulating cytokines, which is a simulative study to mimic drug use in patients with T2D. The authors found that RU486 dose-dependently decreased postprandial blood glucose, increased serum corticosterone and ACTH, but did not affect serum MCP-1 and IL-6, although the basal IL-6 was higher in db/db mice than in the lean littermate controls. The significantly higher level of IL-6 in the db/db mice is consistent with a state of chronic low-grade inflammation. The changes in glucose, corticosterone, and ACTH in RU486-treated db/db mice were consistent with relevant literature reports, and confirmed the efficacious drug exposure (measured serum drug level 2–3 hours after last dose: 14 ± 3, 33 ± 3, 273 ± 17, and 719 ± 57 nM, respectively at doses of 2.5, 7, 20, and 50 mg/kg). These findings suggest that RU486 did not further enhance inflammation in this mouse model of T2D with low grade inflammation. The dose-effect curves of RU486 on glucose, corticosterone, and ACTH were used to decide the RU486 doses 4–100 mg/kg used in the subsequent acute studies.
In the treatment of septic shock, low doses of glucocorticoids have anti-inflammatory properties. GR antagonists might block the anti-inflammatory effect of endogenous cortisol, and deteriorate inflammation. Excessive LPS release from the bacterial wall into the circulation is the primary pathogenic process of septic shock [24
]. To evaluate the effect of RU486 under the condition of septic shock, an LPS-challenge model was used in the current studies. We found that in normal AKR mice, LPS dose-dependently increased serum TNFα and caused increases in corticosterone and ACTH. Since LPS-increased circulating TNFα and other circulating cytokines can trigger the neural circuits that control the HPA axis [25
], the increases in corticosterone and ACTH in response to LPS challenge were an expected stress response of HPA axis [26
]. Interestingly, pretreatment with RU486 dose-dependently decreased serum TNFα, which is contrary to the hypothesis that RU486 would exacerbate inflammation. There was also a dose-dependent increase in circulating corticosterone in response to RU486 treatment, which is a well known compensatory reaction in response to the GR antagonism. Thus the suppression of LPS-mediated TNFα production in RU486-treated mice was most likely a consequence of the increased circulating corticosterone. Because of the compensatory increase in endogenous glucocorticoids, acute GR antagonism with RU486 does not necessarily lead to exacerbation of inflammation.
It has been reported that diet-induced obesity or/and diet-induced insulin resistance in mice is associated with low grade inflammation, as evidenced by macrophage accumulation in adipose tissue and increased serum cytokines [27
]. Chronic low-grade inflammation in these mice could alter the effects of RU486 on LPS-induced TNFα production compared to the effects in normal mice. Therefore, we used C57BL/6 DIO mice to determine the effects of LPS and RU486. The phenotype of the C57BL/6 DIO mice was consistent with the literature [29
]. While the LPS-induced increases in serum TNFα and corticostrone were similar to those in normal AKR mice, the effects of RU486 pretreatment on the LPS-mediated changes in C57BL/6 DIO mice were different than that in the normal AKR mice. RU486 suppressed 0.1 mg/kg LPS-induced TNFα production but did not affect 0.5 mg/kg LPS-induced TNFα production, despite the fact that RU486 further increased serum corticosterone at both LPS doses. Thus, under these conditions, the compensatory increase in glucocorticoids was not able to suppress LPS-induced TNFα production.
Considering that the stress response during sepsis could exhaust adrenal cortex, a very high dose of exogenous LPS could mimic this state and would exhaust the endogenous glucocorticoids. Under these conditions, RU486 may no longer be able to cause compensatory glucocorticoid release. Whether RU486 pretreatment exacerbates LPS-induced TNFα production under this condition is not known. Therefore, we used LPS 20 mg/kg in C57BL/6 DIO mice to observe the effect of RU486 in this stressed condition. We found that this very high dose of LPS led to an exaggerated increase in serum TNFα, but increases in serum corticosterone and ACTH were similar to those observed after LPS 0.1 and 0.5 mg/kg. RU486 pretreatment did not affect the 20 mg/kg LPS-mediated changes in TNFα, corticosterone, and ACTH. Despite this lack of increase in glucocorticoids, RU486 did not enhance TNFα production.
There are a few papers reporting GR agonist activity of RU486 in some in vitro and in vivo systems. Nordeen et al [30
] reported that protein kinase A activators unmasked the agonist effect of RU486 as mediating an induction of hormone-responsive report genes. Zhang et al [31
] reported that the GR agonist activities of RU486 were dependent on the GR levels but not on EC50 values in COS-7 cells transfected with GR expression vector. Schulz et al [32
] reported that the RU486-induced GR agonism was controlled by the receptor N terminus and by corepressor binding. All these reports indicated that RU486 may act as a GR agonist under certain conditions on certain pathways/procedures. This agonist activity of RU486 may contribute to our observation of decreasing LPS-induced TNFα by RU486 in normal AKR mice and C57Bl/6 DIO mice. Although the mechanism and extent of GR agonist activities of RU486 in whole animals is not known, the GR agonist activities of RU486 reduce the extent of, and concerns about, potential inflammatory exacerbation by the drug.
As demonstrated in the current studies, LPS challenge is a strong stress to the subjects. LPS not only increases cytokine production (such as TNFα and IL-6), but also stimulates HPA axis and leads to increases in both serum corticosterone and ACTH. The levels of cytokine, ACTH, and corticosterone are considered to be the homeostatic status under the LPS-challenge. In the HPA axis, GCs have negative feedback effects on hypothalamus and anterior pituitary gland. GR antagonism with RU486 blocks the GR throughout body tissues and even blocks GCs' negative feedback receptors in the hypothalamus and anterior pituitary gland. This blockade leads to activation of HPA, and results in increased rate of synthesis and release of ACTH and GCs. The current studies in AKR mice and C57BL/6 DIO mice demonstrated that the acute RU486 pretreatment resulted in an increase only in corticosterone but not in ACTH. This finding indicates that GCs may have a negative feedback on the adrenal cortex, thus regulating its own synthesis and release.
Although the 5 deaths of septic shock were reported to be associated with use of RU486 [14
], and antagonism of the GR was hypothesized to be the primary cause [16
], the current studies in three different strains of animals with normal and abnormal metabolism have not shown inflammatory exacerbation by RU486 treatment. This suggests, notwithstanding the species and strain differences, that mechanisms other than, or along with, antagonism of the GR may contribute to the inflammatory exacerbation exhibited by RU486 in obstetric/gynecologic clinic. RU486 also blocks progesterone receptors [33
], and progesterone also demonstrates anti-inflammatory properties [34
]. Therefore, the septic shock associated with use of RU486 as abortifacient may not be solely due to antagonism of the GR by RU486. The exact cause of these deaths needs to be further investigated.