The amount of iron in the diet directly influences the composition of the microbiota. Inversely, the effects of the microbiota on iron homeostasis have been little studied. So, we investigate whether the microbiota itself may alter host iron sensing. Duodenal cytochrome b and divalent metal transporter 1, involved in apical iron uptake, are 8- and 10-fold, respectively, more abundant in the duodenum of germ-free (GF) mice than in mice colonized with a microbiota. In contrast, the luminal exporter ferroportin is 2-fold less abundant in GF. The overall signature of microbiota on iron-related proteins is similar in the colon. The colonization does not modify systemic parameters as plasma transferrin saturation (20%), plasma ferritin (150 ng/L), and liver (85 µg/g) iron load. Commensal organisms (Bacteroides thetaiotaomicron VPI-5482 and Faecalibacterium prausnitzii A2-165) and a probiotic strain (Streptococcus thermophilus LMD-9) led to up to 12-fold induction of ferritin in colon. Our data suggest that the intestinal cells of GF mice are depleted of iron and that following colonization, the epithelial cells favor iron storage. This study is the first to demonstrate that gut microbes induce a specific iron-related protein signature, highlighting new aspects of the crosstalk between the microbiota and the intestinal epithelium.

BackgroundSenescent red blood cells (RBC) are recognized, phagocytosed and cleared by tissue macrophages. During this erythrophagocytosis (EP), RBC are engulfed and processed in special compartments called erythrophagosomes. We previously described that following EP, heme is rapidly degraded through the catabolic activity of heme oxygenase (HO). Extracted heme iron is then either exported or stored by macrophages. However, the cellular localization of the early steps of heme processing and iron extraction during EP remains to be clearly defined.Methodology/Principal FindingsWe took advantage of our previously described cellular model of EP, using bone marrow-derived macrophages (BMDM). The subcellular localization of both inducible and constitutive isoforms of HO (HO-1 and HO-2), of the divalent metal transporters (Nramp1, Nramp2/DMT1, Fpn), and of the recently identified heme transporter HRG-1, was followed by fluorescence and electron microscopy during the earliest steps of EP. We also looked at some ER [calnexin, glucose-6-phosphatase (G6Pase) activity] and lysosomes (Lamp1) markers during EP. In both quiescent and LPS-activated BMDM, Nramp1 and Lamp1 were shown to be strong markers of the erythrophagolysosomal membrane. HRG-1 was also recruited to the erythrophagosome. Furthermore, we observed calnexin labeling and G6Pase activity at the erythrophagosomal membrane, indicating the contribution of ER in this phagocytosis model. In contrast, Nramp2/DMT1, Fpn, HO-1 and HO-2 were not detected at the membrane of erythrophagosomes.Conclusions/SignificanceOur study highlights the subcellular localization of various heme- and iron-related proteins during early steps of EP, thereby suggesting a model for heme catabolism occurring outside the phagosome, with heme likely being transported into the cytosol through HRG1. The precise function of Nramp1 at the phagosomal membrane in this model remains to be determined.

No abstract

HO1 (haem oxygenase 1) and Fpn (ferroportin) are key proteins for iron recycling from senescent red blood cells and therefore play a major role in controlling the bioavailability of iron for erythropoiesis. Although important aspects of iron metabolism in HO1-deficient (Hmox1-/-) mice have already been revealed, little is known about the regulation of Fpn expression and its role in HO1 deficiency. In the present study, we characterize the cellular and systemic factors influencing Fpn expression in Hmox1-/- bone marrow-derived macrophages and in the liver and kidney of Hmox1-/- mice. In Hmox1-/- macrophages, Fpn protein was relatively highly expressed under high levels of hepcidin in culture medium. Similarly, despite high hepatic hepcidin expression, Fpn is still detected in Kupffer cells and is also markedly enhanced at the basolateral membrane of the renal tubules of Hmox1-/- mice. Through the activity of highly expressed Fpn, epithelial cells of the renal tubules probably take over the function of impaired system of tissue macrophages in recycling iron accumulated in the kidney. Moreover, although we have found increased expression of FLVCR (feline leukaemia virus subgroup C receptor), a haem exporter, in the kidneys of Hmox1-/- mice, haem level was increased in these organs. Furthermore, we show that iron/haem-mediated toxicity are responsible for renal injury documented in the kidneys of Hmox1-/- mice.

No abstract

The continuous recycling of haem iron following phagocytosis and catabolism of senescent and damaged red blood cells by macrophages is a crucial process in the maintenance of systemic iron homoeostasis. However, little is known about macrophage iron handling in haemolytic states resulting from a deficiency in antioxidant defences. Our observations indicate that the recently described chronic, but moderate regenerative, haemolytic anaemia of aged SOD1 (superoxide dismutase 1)-knockout mice is associated with red blood cell modifications and sensitivity to both intra- and extra-vascular haemolysis. In the present study, we have characterized the molecular pathways of iron turnover in the liver of Sod1-deficient mice. Despite iron accumulation in liver macrophages, namely Kupffer cells, we did not measure any significant change in non-haem liver iron. Interestingly, in Kupffer cells, expression of the rate-limiting enzyme in haem degradation, haem oxygenase-1, and expression of the iron exporter ferroportin were both up-regulated, whereas the hepcidin mRNA level in the liver was decreased in Sod1-/- mice. These results suggest that concerted changes in the hepatic expression of iron- and haem-related genes in response to haemolytic anaemia in Sod1-/- mice act to reduce toxic iron accumulation in the liver and respond to the needs of erythropoiesis.

Key Points Matriptase-2 is required for hepcidin suppression in fetuses and neonates to ensure iron mobilization to avoid iron deficiency and anemia. In matriptase-2–deficient mice, the constant developmental hyperhepcidinemia is not explained by increased level of hepatic hemojuvelin.

Glucagon-Like Peptide-1 (GLP-1) undergoes rapid inactivation by dipeptidyl peptidase-4 (DPP4) suggesting that target receptors may be activated by locally produced GLP-1. Here we describe GLP-1 positive cells in the rat and human stomach and found these cells co-expressing ghrelin or somatostatin and able to secrete active GLP-1 in the rats. In lean rats, a gastric load of glucose induces a rapid and parallel rise in GLP-1 levels in both the gastric and the portal veins. This rise in portal GLP-1 levels was abrogated in HFD obese rats but restored after vertical sleeve gastrectomy (VSG) surgery. Finally, obese rats and individuals operated on Roux-en-Y gastric bypass and SG display a new gastric mucosa phenotype with hyperplasia of the mucus neck cells concomitant with increased density of GLP-1 positive cells. This report brings to light the contribution of gastric GLP-1 expressing cells that undergo plasticity changes after bariatric surgeries, to circulating GLP-1 levels.