Oral administration of a bone morphogenetic protein type I receptor inhibitor prevents the development of anemia of inflammation

Anemia of inflammation (AI), the second most frequent form of anemia, complicates inflammatory and chronic diseases.1 Although typically mild to moderate, AI is associated with increased morbidity and mortality.1 One of the mechanisms contributing to the development of AI is a persistent stimulation of hepcidin synthesis, a critical regulator of iron homeostasis, by inflammatory cytokines, including interleukin-6 (IL-6).2 Hepcidin synthesis occurs mainly in the liver and is predominantly regulated at a transcriptional level.2 Bone morphogenetic protein (BMP) signaling induces transcription of the gene encoding hepcidin (HAMP).3 Activation of the BMP pathway induces the phosphorylation of regulatory SMAD proteins (SMADs 1, 5, and 8), which activate the expression of target genes, such as the genes encoding the inhibitor of DNA binding 1 (ID1) and hepcidin.42 Treatment of AI is typically directed to curing the underlying disease, which is not always feasible. Current adjuvant therapies, including intravenously-administered iron, erythropoiesis-stimulating agents, and erythrocyte transfusions, have limited efficacy and important side effects.75 Research has focused on the development of new therapeutic agents targeting hepcidin,8 including inhibitors of hepcidin synthesis. Among them, LDN-193189, a BMP type I receptor inhibitor that is a derivative of dorsomorphin, has been reported to increase hemoglobin levels in animal models of AI.119 However, in experimental anemia models, LDN-193189 was administered by intraperitoneal injection, which may limit its applicability to treatment of patients with AI.

We sought to determine if LDN-193189 is biologically available when administered orally and whether oral administration of LDN-193189 is sufficient to prevent the development of AI. Although other dorsomorphin derivatives, such as DMH1, have been developed, we chose to study LDN-193189 because its pharmacokinetics12 and efficacy after parenteral administration have been studied by several groups in different disease models, including anemia of inflammation in mice.119

To evaluate the in vivo disposition kinetics and oral bioavailability of LDN-193189, mice received a single dose of LDN-193189 (3 mg/kg) by gavage or intravenously (IV) by tail vein injection, and blood was collected after various times to measure serum LDN-193189 levels (Table 1 and Online Supplementary Appendix). Following a single IV administration, LDN-193189 had a moderate clearance (CL) of 39 mL/min/kg and a large steady-state volume of distribution (V_dss) of 6.9 L/kg, suggesting that LDN-193189 penetrates extensively into tissues. As expected, the peak concentration (C_max) was higher when LDN-193189 was administered intravenously than when it was administered orally. The area under the concentration-time curve from time zero to infinity (AUC_(0-inf)) was greater when LDN-193189 was administered by IV injection than when it was orally-administered. As anticipated, the half-life (t_1/2) of serum LDN-193189 levels was not impacted by the route of administration.

Table 1.LDN-193189 pharmacokinetic properties.

To assess the dose-dependency of orally-administered LDN-193189 to increase serum LDN-193189 levels, blood was collected at various times after mice received a dose of LDN-193189 (1, 3, or 10 mg/kg), and serum LDN-193189 concentrations were measured (Table 1). The in vivo exposure (C_max and AUC_(0-inf)) increased in a dose-dependent manner. The half-life of LDN-193189 was independent of the orally-administered dose. Comparison of the AUC_(0-inf)s revealed that, after oral administration, LDN-193189 bioavailability was 54%–83%.

To study the tissue exposure of orally-administered LDN-193189, serum, liver, and muscle were harvested at various times after mice received a dose of LDN-193189 (1 mg/kg) by gavage, and LDN-193189 levels were measured (Table 1 and Online Supplementary Appendix). The in vivo exposure (C_max and AUC_(0-inf)) was higher in liver than in serum and muscle. The tissue-to-serum AUC ratio was approximately 2.3 for the liver and 0.79 for the muscle. These results suggest that orally-administered LDN-193189 accumulates in the liver.

To confirm that orally-administered LDN-193189 was able to inhibit hepatic BMP signaling, we measured phosphorylation of BMP-responsive Smads (Smads 1 and 5) and expression of genes encoding Id1 and hepcidin. Mice were administered LDN-193189 (1 mg/kg) by gavage, 1–24 h prior to euthanasia (Online Supplementary Appendix). Administration of LDN-193189 rapidly reduced Smad1/5 phosphorylation, an effect that persisted up to 4 h (Figure 1A). Similarly, orally-administered LDN-193189 decreased hepatic Id1 gene expression after 1 h with a nadir at 2 h, whereas hepcidin mRNA levels progressively decreased and were lowest at 4 h (Figure 1B).

Figure 1.Oral administration of LDN-193189 inhibits hepatic Id1 and hepcidin gene expression in a time- and dose-dependent manner. C57BL/6 female mice (n=3–4/group) received a dose of LDN-193189 (LDN, 1 mg/kg), by gavage, at various times (1–24 h) prior to sacrifice. After euthanasia, livers were harvested. (A) Proteins were extracted, and levels of total Smad1 and phosphorylated Smad1/5 (p-Smad1/5) were determined by immunoblots. (B) RNA was extracted to measure the levels of mRNAs encoding Id1 (upper panel) and hepcidin (lower panel) by qRT-PCR (one-way ANOVAs P<0.007; *P<0.03 vs. time 0). Two hours after oral administration of a dose of LDN-193189 (doses ranging from 0.1 to 10 mg/kg) or vehicle (citric acid), mice were euthanized, and livers were harvested. (C) Proteins were extracted, and levels of Smad1 and p-Smad1/5 were determined. (D) RNA was extracted, and levels of mRNAs encoding for Id1 were measured by qRT-PCR (one-way Anova P<0.0001; *P<0.04 vs. mice treated with citric acid; #P<0.05 mice treated with LDN-193189 0.3 vs. 1 mg/kg). (E) Four hours after administration of increasing doses of LDN-193189, mice were euthanized, and livers were harvested. RNA was extracted, and levels of mRNAs encoding for hepcidin were measured by qRT-PCR (one-way Anova P<0.0001; *P<0.002 vs.mice treated with citric acid).

To assess the dose-dependency of orally-administered LDN-193189 to inhibit BMP signaling, mice received LDN-193189 by gavage at doses ranging from 0.1 to 10 mg/kg. Two hours later, mice were sacrificed, and their livers were harvested. LDN-193189 inhibited Smad1/5 phosphorylation at doses above 0.3 mg/kg (Figure 1C). Orally-administered LDN-193189 at doses above 0.3 mg/kg decreased hepatic Id1 gene expression (Figure 1D), whereas hepcidin mRNA levels did not decrease (data not shown). Four hours after oral administration of LDN-193189, hepatic hepcidin gene expression was decreased at doses of 0.1 mg/kg or over (Figure 1E). Taken together, these results demonstrate that orally-administered LDN-193189 inhibits hepatic BMP signaling and hepcidin gene expression. Moreover, we observed that hepatic BMP signaling was sensitive to inhibition by oral LDN-193189, with as little as 0.1 mg/kg able to reduce hepcidin mRNA levels. The ability of orally-administered LDN-193189 to inhibit hepatic BMP signaling at such a low dose could be explained by the accumulation of LDN-193189 in the liver, as reflected by the hepatic tissue-to-serum AUC of 2.3.

To determine whether the levels of LDN-193189 achieved after oral administration were sufficient to increase hemoglobin levels in AI, mice were challenged with turpentine, a well-established animal model of AI, which induces hepatic hepcidin gene expression in an IL6-dependent manner.13 Mice were injected with turpentine weekly for three weeks,149 and were treated daily with LDN-193189 (1 mg/kg) or vehicle (citric acid) by gavage (Online Supplementary Appendix). As controls, mice received injections of saline weekly and received vehicle by gavage daily. Because LDN-193189 does not affect erythropoiesis,9 we did not include a group of mice that received weekly injections of saline and that were treated daily with LDN-193189. Seven days after the last injection of turpentine, and 24 h after the last administration of LDN-193189, mice were anesthetized, blood was collected to measure complete blood counts, serum IL-6 levels and serum iron levels. Mice were then euthanized to collect liver samples to measure hepatic hepcidin gene expression. Mice injected with turpentine developed sterile abscesses and systemic inflammation with increased granulocyte and platelet counts and serum IL-6 levels (Figure 2A). Administration of LDN-193189 did not modulate the development of inflammation, as reflected by similar increases in granulocyte and platelet counts and serum IL-6 levels in turpentine-challenged mice treated with LDN-193189 and vehicle. Mice challenged with turpentine and treated with vehicle developed AI with a decrease in hemoglobin levels, mean corpuscular volumes (MCVs), and erythrocyte counts (Figure 2B). Hemoglobin levels, MCVs, and erythrocyte counts were greater in turpentine-challenged mice treated with LDN-193189 than in turpentine-challenged mice treated with vehicle. Treatment of turpentine-challenged mice with LDN-193189 increased, but did not fully normalize hemoglobin levels. This partial effect of LDN-193189 has been observed in turpentine and other models of AI, when LDN-193189 was given by intraperitoneal injection.15119

Figure 2.Orally-administered LDN-193189 increases hemoglobin levels in turpentine-challenged mice. Eight-week-old C57BL/6 female mice were injected weekly with turpentine (5 mL/kg, subcutaneously) during three weeks. Simultaneously, mice were treated daily with either LDN-193189 (1 mg/kg, black bars, n=11) or vehicle (gray bars, n=10). White bars represent control female mice challenged with saline weekly and treated with vehicle (citric acid) daily (n=11). Seven days after the last injection of turpentine or saline, mice were sacrificed. Blood was collected to measure complete blood counts and serum IL-6 levels. Turpentine injections induced systemic inflammation with (A) an increase of granulocyte (upper panel) and platelet counts (middle panel) (one-way Anovas P<0.0001 and *P<0.0001 vs. control mice), as well as an increase in serum IL-6 levels (lower panel). (B) In vehicle-treated mice, turpentine injections induced AI with a decrease of hemoglobin levels (upper panel), mean corpuscular volumes (MCVs, middle panel), and erythrocyte counts (lower panel) that was attenuated by oral administration of LDN-193189 (one-way Anovas P<0.0001 and *P<0.0001 vs. control mice; ^§P<0.006 vs. mice injected with turpentine and treated with vehicle).

In turpentine-challenged mice, serum iron levels tended to be lower than in control mice and were not corrected by LDN-193189 treatment (Online Supplementary Figure S1A). In contrast, Steinbicker et al.9 reported that, in turpentine-challenged mice, administration of LDN-193189 at a dose of 3 mg/kg increased serum iron levels. These results suggest that a dose of LDN-193189 of more than 1 mg/kg might be required to normalize serum iron levels. A week after the last injection of turpentine, there was no change in hepatic hepcidin gene expression in turpentine-challenged mice treated with vehicle (Online Supplementary Figure S1B). These results are consistent with previous observations of Prince et al.14 In addition, hepatic hepcidin mRNA levels were similar in turpentine-challenged mice treated with LDN-193189 or vehicle (Online Supplementary Figure S1B). In this study, mice were sacrificed 24 h after the last administration of LDN-193189 when hepatic hepcidin gene expression was no longer inhibited (Figure 1B).

In summary, we report that LDN-193189 is orally bioavailable. Oral administration of a BMP type I receptor inhibitor is effective in inhibiting hepatic BMP signaling and reducing hepcidin gene expression, as well as development of anemia in an animal model of AI. These results suggest that LDN-193189 or other dorsomorphin-related molecules may represent a novel orally-effective therapy for patients with AI.