- Email: [email protected]
Study of iron metabolism disturbances in an animal model of insulin resistance
Diabetes Research and Clinical Practice 77 (2007) 363–370 www.elsevier.com/locate/diabres
Study of iron metabolism disturbances in an animal model of insulin resistance Guillaume Le Guenno a,*, Emilie Chanse´aume b, Marc Ruivard c, Be´atrice Morio b, Andrzej Mazur a a
INRA, Equipe Stress Me´tabolique et Micronutriments, UNH, UMR 1019, Clermont-Ferrand/Theix, St-Gene`s-Champanelle 63122, France b INRA, Equipe Me´tabolisme Lipidique et Energe´tique, UNH, UMR 1019, Clermont-Ferrand/Theix, St-Gene`s-Champanelle 63122, France c Service of Internal Medicine Hoˆtel-dieu, C.H.U. Clermont Ferrand, France Received 24 October 2006; accepted 10 February 2007 Available online 9 March 2007
Abstract The relationship between iron and insulin-resistance (IR) is documented by the positive correlation between iron stores and IR. Moreover, some patients exhibited a hepatic iron overload associated with IR (HIO-IR) but the mechanism involved in this overload is not known. Thus, we studied the iron metabolism disturbances in an animal model of IR and the influence of provoked hyperglycemia/hyperinsulinemia on plasma iron parameters. Wistar rats were fed a control or a high-fat/high-energy (HF/HE) diet. Plasma glucose, insulin, iron, transferrin and transferrin saturation (TS) were measured during intra-peritoneal glucose test tolerance (IPGTT) compared to saline. Hemogram, tissue iron concentrations and hepatic hepcidin mRNA expression were determined at the end of experiment. HF/HE rats exhibited higher body and liver weights, increased IR-index and hemoglobin concentration. Iron content was lower in the spleen of HF/HE rats and tended to decrease in the liver as compared to controls. Transferrin values were higher and these of TS lower in HF/HE group. The hepcidin mRNA was 3.5-fold lower in HF/HE rats than in controls. IPGTT had no effect on iron status parameters in both groups. As reflected by higher hemoglobin concentration, IR could increase erythropoı¨esis which enhances iron requirement. Iron stores and TS value decreased leading to a down-regulation of hepcidin expression which increased iron absorption. Hepcidin expression should be investigated in metabolic syndrome and hepatic iron overload associated with IR. # 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Iron; Insulin resistance; High-fat/high-energy diet; IPGTT; Hepcidin; Erythropoı¨esis
1. Introduction The relationship between iron metabolism and metabolic disorders has recently gained interest in both research and clinical practice. Indeed, body iron stores are positively correlated with insulin-resistance (IR), * Corresponding author at: 8 impasse Chabrier 63540 Romagnat, France. Tel.: +33 4 73 61 15 45; fax: +33 4 73 62 46 38. E-mail address: [email protected] (G. Le Guenno).
even in the absence of significant iron overload [1–4]. Ferritin, which reflects body iron stores, is closely associated with IR and can be considered a marker for metabolic syndrome [5]. It has been shown that phlebotomy significantly improves insulin sensitivity in type 2 diabetes [6]. In animals, iron deficiency increases insulin sensitivity [7]. Moreover, a common syndrome called insulin-resistance associated hepatic iron overload (IR-HIO) was described in 1997 by Moirand et al. [8]. This is the most frequent cause of
0168-8227/$ – see front matter # 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2007.02.004
364
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
iron overload in France with an estimated prevalence of 1% of the population, 10 times more widespread than haemochromatosis [9]. A study of 269 subjects with metabolic syndrome showed a prevalence of 14.5% of the IR-HIO [10]. IR-HIO combines an isolated hyperferritinemia with a normal transferrin saturation, steatohepatitis and insulin resistance. It represents the most widespread indication for venesection in referral care units for iron overload [9]. However, the mechanism involved in this overload is presently unknown. A few studies of the perturbation of iron metabolism during IR were focused on the genetic models of obesity, the ob/ob mouse [11,12] and Zucker rats [13]. The obese animals exhibited a decreased iron concentration in the liver [11–13], whereas plasmatic iron, transferrin and spleen iron stores were unchanged when compared with controls [11]. In ob/ob mice, the rate of iron absorption was two-fold greater and the haemoglobin concentration was significantly higher than in lean control mice [11]. The authors concluded that increased erythropoı¨esis in the obese animals provoked a higher iron requirement, which explains the improvement in iron absorption and the decreased liver iron stores. An important advance in our knowledge of iron metabolism was made with the discovery of hepcidin. This peptide, initially described as an antimicrobial peptide [14], is a key regulator of iron stores and it inhibits duodenal iron absorption and iron release by macrophages, thereby provoking the internalization of ferroportin [15]. It has been shown that hepcidin expression is upregulated during infection, inflammation and iron overload [16], whereas the expression is down-regulated by hypoxia, anaemia, iron deficiency, erythropoietin and erythropoietic stimulation [17]. However, even if hepcidin plays a key role in iron absorption, it is unknown how hepcidin expression is modulated in IR. Studies of interactions between iron and glucose metabolism have shown that insulin can cause a rapid and pronounced stimulation of iron uptake by adipocytes by redistributing the transferrin receptor from an intracellular compartment to the cell surface [18]. Transferrin receptors co-localize with the glucose transporters and insulin-like growth factor II receptor in the microsomal membranes of cultured adipocytes, this suggests that regulation of iron uptake by insulin occurs in parallel with glucose uptake [19]. Moreover, a study on the effect of extreme hyperinsulinemia, obtained during a hyperinsulinemic euglycemic clamp in five healthy women, showed a progressive improvement of the plasmatic iron during the experiment [20].
In reviewing the current literature, there is no data to support a mechanism for iron accumulation in IR and the aims of the present work were: To study the perturbations of the iron status and hepcidin expression in an IR animal model induced by an experimental diet. To analyze changes in plasma iron parameters during hyperglycemia/hyperinsulinemia, induced by intraperitoneal injection of glucose, in this animal model. 2. Materials and methods 2.1. Animals and experimental diets Sixteen 3 months old male Wistar rats were randomly divided into two groups. The groups consumed a control or a high-fat/high-energy (HF/HE) diet for 6 weeks. Diets were prepared in the experimental diet preparation unit of Jouy-enJosas (UE300, UPAE INRA Domaine de Vilvert 78352 Jouyen-Josas) and distributed in a semi-liquid form in individual ramekins. Food was weighed daily and prepared for each animal. Tap water was available ad libitum. Animals were housed in individual cages with a normal light cycle (Day 8 a.m. to 8 p.m.), in a temperature-controlled room (22 8C). On a caloric basis, the HF/HE diet consisted of 45% fat (6.7% from groundnut oil, 6.7% from canola oil and 31.6% from lard), 37.6% carbohydrate (25.6% from starch and 12% from sucrose), and 17.4% protein (total 4.68 kcal/g), whereas the control diet contained 13.8% fat (6.7% from groundnut oil and 6.7% from canola oil), 68.8% carbohydrate from starch, and 25.8% protein (total 3.9 kcal/g). Iron was provided as iron sulfate and the intake was 7.8 and 7.7 mg/g body weight in the control and the HF/HE group, respectively. The experimental protocol was approved by the institutional animal care and use committee at INRA (Decree 87– 848 modified by decree 2001–464). 2.2. Experimental procedures One week before the animals were sacrificed, an intraperitoneal glucose tolerance test (IPGTT) was performed at 9 a.m. After 12 h of starvation, four rats of each group received an intraperitoneal injection of 1 g of glucose/kg body weight (G50%, volume (ml) = weight (g) 2) and the remaining eight animals received an isotonic saline solution injection. Blood samples were obtained by retro-orbital puncture at 0, 15, 30, 60 and 120 min. The plasma was collected by centrifugation at 1000 g for 10 min and stored at 80 8C until further analyses. After 6 weeks on the specific diets, the rats were anesthetized by intraperitoneal injection of Imalgene (120 mg/kg; Vetranquil, Merial, Lyon, France) and Diazepam (1.75 mg/kg; Valium, Roche, France) and sacrificed by decapitation. Blood was collected after decapitation for hemogram measurements.
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
Liver, spleen and duodenum were collected, weighed and immediately frozen in liquid nitrogen and stored at 80 8C. 2.3. Glycemia and insulinemia measurements Glycemia was measured immediately using fresh blood (Glucometer Gluco Touch, LifeScan, Inc.). Insulinemia was measured using plasma by ELISA (Rat/Mouse Insulin ELISA kit EZRMI-13K, LINCO research, Inc, USA). Insulin-resistance (IR) was calculated by the IR-index [AUC glycemia (mg/dl) AUC insulinemia (ng/ml)], the Homeostasis Model Assessment score [HOMA = glycemia (mmol/l) insulinemia (mUI/ml)/22.5] and the Quantitative Insulin Sensitivity Check Index [QUICKI = 1/log insulinemia (mUI/l) + log glycemia (mg/dl)]. 2.4. Iron status parameters Haemograms were determined with a Scil Vet ABC counter (Animal Blood Counter, Strasbourg, France). Plasma iron and TIBC were measured using the ‘‘ferrimat kit’’ (bioMerieux SA, Marcy-l’Etoile, France) and the ‘‘TIBC additif’’ kit (bioMerieux SA,) in combination with a Progress Plus Chemistry Analyser automat (Kone, Evry, France). The transferrin saturation (TS) was calculated as fasting plasmatic iron/TIBC and transferrin as TIBC/25. Nonheme iron was determined by a modification of the method of Foy et al. [21] as described by Simpson and Peters [22]. For the duodenum, nonheme iron concentrations were expressed relative to the protein concentration (mg/mg of protein). Tissue protein concentration was estimated using the ‘‘protein BCA Uptima kit’’ (Interchim, Montlucon, France). 2.5. Hepcidin mRNA measurements Total RNA were extracted from the liver using the Qiagen RNeasy Mini kit (Coutaboeuf, France) according to the manufacturer’s instructions. We used 3 mg of total RNA for cDNA synthesis, using the Ready To Go Your First Strand Bead kit (Amersham Pharmacia Biotech, Orsay, France). Polymerase chain reaction (PCR) was carried out using the Pure Taq Ready To Go PCR Beads kit (Amersham Pharmacia Biotech, Orsay, France) and a TC-512 Techne Thermal Cycler (MIDSCI, USA). We performed quantitative RT-PCR using the LightCycler Fast Start DNA Master SYBR Green I kit
365
(Roche Diagnostics, Meylan, France) and a LightCycler (Roche Diagnostics). The hepcidin gene expression was normalized to the GAPDH expression in the same sample. The following primers were used for PCR amplification: GAPDH (50 -CAT GAC CAC AGT CCA TGC CAT CAC-30 and 50 -CAT GTA GGC CAT GAG GTC CAC CAC-30 ), hepcidin (50 -ACA GAA GGC AAG ATG GCA CT-30 and 50 -GAA GTT GGT GTC TCG CTT CC-30 ). These primer pairs produced 458- and 201-bp amplification products, respectively. 2.6. Statistical analysis Results were expressed as means S.E.M. Statistical analysis were performed using the Statview software (SAS Institute Inc., SAS campus drive, Cary, NC, USA). If a significant variance difference was observed between the two groups, a log transformation was performed. The statistical significance of differences between means from two studied groups was assessed by the Student’s t-test. A twoway repeated measures ANOVA, followed by PLSD Fisher’s test, was performed to estimate the effect of group and injection on values obtained during IPGTT. Differences were considered as significant at p < 0.05.
3. Results 3.1. Effect of diet on haematological parameters, body and organ weight The effect of diet on the haematological parameters, body and organ weight are shown in Table 1. After 6 weeks, the HF/HE group had significantly higher body ( p < 0.01) and liver weight ( p < 0.01), whereas spleen weight was lower ( p = 0.03), when compared with the control group. Expressing the organ weight relative to body weight, the difference in the liver weight remained at the edge of significance ( p = 0.04), but the spleen weight was more pronounced ( p < 0.01). The haemoglobin concentration was significantly higher in the HF/ HE group when compared with the control group ( p = 0.03). The other haematological parameters (white blood cells and platelet counts) were not significantly different (data not shown).
Table 1 Effect of diet on body and organ weight, and hemoglobin concentration in the control and high-fat/high-energy diet fed groups
Body weight (g) Liver weight (g) Spleen weight (g) Hemoglobin concentration (g/dl) Values are expressed as means S.E.M. for groups of eight rats. * Significantly different from controls ( p < 0.05).
Control (n = 8)
High-fat/high-energy (n = 8)
514.2 9.2 15.9 0.5 1.02 0.2 14.5 0.9
570.3 9.9 * 19.1 0.9 * 0.89 0.4 * 18.5 1.6 *
366
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
Fig. 1. Effect of intraperitoneal injection of glucose or saline serum on glycemia and insulinemia. Values were measured in control and HF/HE diet fed group at T0, T15, T30, T60 and T120 min. (A) Effect of isotonic saline solution injection on glycemia. (B) Effect of glucose solution injection on glycemia. (C) Effect of isotonic saline solution injection on insulinemia. (D) Effect of glucose solution injection on insulinemia. Values are expressed as means S.E.M. for groups of four rats. Using a two-ways ANOVA, there is no significant effect of group’s factor on insulinemia and glycemia values whereas injection’s factor has significant effect on insulinemia ( p < 0.01) and glycemia values ( p < 0.01). Black circles: HF/HE diet fed group. Empty circles: control diet fed group.
3.2. Effect of diet on IPGTT and evaluation of insulin-resistance Fasting glycemia was significantly higher in HF/HE group ( p < 0.01). The effect of intraperitoneal injection of isotonic saline and glucose solution on glycemia and insulinemia values at T0, T15, T30, T60, T120 min in the control and HF/HE groups are shown in Fig. 1. Using a two-way ANOVA, diet factor had no significant effect on glycemia ( p = 0.06) and insulinemia values ( p = 0.09) at the different times of the IPGTT. Insulin-resistance was higher in HF/HE group whether comparing IR index between groups receiving glucose ( p = 0.04) or isotonic saline injection
( p = 0.01), or the QUICKI index ( p < 0.001). The HOMA index did not differ significantly between the two groups ( p = 0.10). These data are shown in Table 2. 3.3. Effect of diet on tissue nonheme iron concentration The tissue iron concentrations are shown in Table 3. In HF/HE diet group, the nonheme iron content was significantly lower in the spleen ( p < 0.01), tended to decrease in the liver ( p = 0.06), but was unchanged in the duodenum ( p = 0.92), when compared with the control group.
Table 2 Effect of diet on fasting glycemia, fasting insulinemia, IR index, QUICKI and HOMA indexes
Fasting glycemia (mg/dl) Fasting insulinemia (ng/ml) IR index (glucose injection)a IR index (saline injection)a QUICKIb HOMAc
Control (n = 8)
High-fat/high-energy (n = 8)
72.12 1.02 0.90 0.29 2.83.106 9.76.105 9.09.105 2.06.105 0.44 0.02 13.28 4.41
77.87 1.88* 1.35 0.25 5.65.106 9.85.105* 1.81.106 2.15.105* 0.30 0.01* 21.44 4.212
Values are expressed as means S.E.M. for groups of eight rats, except the IR index where four rats of each group received an intraperitoneal injection of isotonic saline or glucose solution. * Significantly different from controls ( p < 0.05). a IR index = AUC glycemia (mg/dl) AUC insulinemia (ng/ml). b QUICKI = 1/(log glycemia (mg/dl) + log insulinemia (mUI/l)). c HOMA = insulinemia (mUI/l) glycemia (mmol/l)/22.5.
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
367
Table 3 Tissue non-heminic iron concentration in control and HF/HE diet fed groups
Liver iron (mg/g tissue) Spleen iron (mg/g tissue) Duodenal iron (mg/mg protein)
Control (n = 8)
High-fat/high-energy (n = 8)
66.76 8.14 424.29 45.10 0.21 0.06
51.51 4.20 304.62 18.04* 0.22 0.09
Values are expressed as means S.E.M. for groups of eight rats. * Significantly different from controls ( p < 0.05).
3.3.1. Effect of diet on plasma iron, transferrin and transferrin saturation during IPGTT The effect of group (control or HF/HE) and injection nature (glucose or saline) on these parameters are shown in Fig. 2 and were assessed using two-way ANOVA. In both groups, plasma iron progressively decreased during the test because of circadian cycle effect. Injection had no effect on plasma iron ( p = 0.87), transferrin ( p = 0.89) and transferrin saturation ( p = 0.92). Group had no effect on plasma iron ( p = 0.42), but had a very significant effect on the transferrin values ( p < 0.01). Comparing the mean values measured during the IPGTT, the HF/HE group
had significantly higher transferrin value ( p < 0.01), lower transferrin saturation ( p = 0.04) and no difference in plasma iron values ( p = 0.19), when compared with the control group. These data are shown in Table 4. 3.4. Effect of diet on hepatic hepcidin expression Liver hepcidin mRNA levels in the two groups were determined by RT-PCR as shown in Fig. 3A. The qRTPCR measurements of the hepatic mRNA levels are shown in Fig. 3B and demonstrate that the hepcidin expression was 3.5-fold lower in HF/HE group ( p = 0.03).
Fig. 2. Effect of intraperitoneal injection of isotonic saline or glucose solution on plasma iron parameters. Values were measured in control and HF/HE diet fed group at T0, T15, T30, T60 and T120 min. (A) Effect of injection on plasma iron in control group. (B) Effect of injection on plasma iron in HF/HE group. (C) Effect of injection on transferrin in control group. (D) Effect of injection on transferrin in HF/HE group. (E) Effect of injection on transferrin saturation in control group. (F) Effect of injection on transferrin saturation in HF/HE group. Values are expressed as means S.E.M. for groups of four rats. Using a two-ways ANOVA, there is no significant effect of injection’s factor on plasma iron parameters whereas group’s factor has significant effect on transferrin values ( p < 0.01). Black circles: isotonic saline solution injection. Empty circles: glucose solution injection.
368
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
Table 4 Average values for plasma iron parameters obtained in control and HF/ HE diet fed groups during intraperitoneal injection
Plasma iron (mmol/l) Plasma transferrin (g/l) Transferrin saturation (%)
Control (n = 8)
High-fat/highenergy (n = 8)
28.67 1.33 4.01 0.11 29.61 2.04
26.16 2.47 4.59 0.18* 23.29 2.76*
Values are expressed as means S.E.M. for groups of eight rats. * Significantly different from controls ( p < 0.05).
4. Discussion The relationship between iron metabolism and IR is well-illustrated in clinical practice by the strong correlation between iron stores and IR [1–5]. Moreover, some of the patients with metabolic syndrome or type 2 diabetes exhibit an insulin-resistance which is associated with hepatic iron overload (IR-HIO). However, the mechanisms behind this overload are unknown and the data from animal studies of genetic obesity models are inconclusive. In this study, we have characterized perturbations in iron metabolism using an animal model rendered IR by an experimental HF/HE diet. Previous studies had shown that 3–4 weeks on a high fat or high energy/high fat diet induces IR as judged by the euglycemic hyperinsulinemic clamp [23–25]. In these studies, the animals exhibited the characteristic abnormalities described in patients with the metabolic syndrome or type 2 diabetes, including increased visceral and muscle fat content, overweight, hepatic steatosis, high fasting glycemia with or without high fasting insulinemia, low HDL values and high plasma fatty acid levels. The HF/HE diet produced similar abnormalities in our study: the animals demonstrated overweight, increased liver weight with macroscopic
steatosis and high fasting glycemia. The development of IR was confirmed by a significant increase in the IRindex and a significant decrease in the QUICKI index, when compared with the control group. To our knowledge, iron metabolism has not been assessed in this IR model. In the present study, we assessed current biomarkers for the blood iron status and the iron content within selected organs serving important roles in iron metabolism, i.e. the liver (iron storage and regulation), the spleen (iron recycling and erythropoı¨esis) and the duodenum (iron absorption). Our data shows that IR leads to a lower iron concentration in the liver and spleen. This is consistent with data from genetic models for IR, such as ob/ob mice [11] and Zucker rats [13], except that in ob/ob mice, the spleen iron concentration is unchanged. In addition, our data shows that there is no difference in the duodenal iron concentration between the two groups. Comparing the mean values of the blood iron status parameters between the control and HF/HE groups, a higher transferrin concentration is associated with lower transferrin saturation in HF/HE group. These changes in the blood parameters of the HF/HE group are associated with lower tissue iron stores, which suggest an increased iron need. Erythropoı¨esis is the state where production of red blood cells is sufficient to maintain a normal level of haemoglobin. In the mammals, iron is mainly used for haemoglobin synthesis. In the HF/HE group, the haemoglobin concentration was higher than in the control group. This is consistent with a study using ob/ ob mice [11] and observations from humans, where the haemoglobin concentration is linked to IR [26,27]. This physiological mechanism may involve insulin stimulation of the sympathetic system and the erythropoietic progenitors. However, the lower spleen weight in HF/ HE diet fed rats, also seen in ob/ob mice [11], is
Fig. 3. (A) Hepcidin mRNA abundance analyzed by RT-PCR (23 cycles) using 3 mg RNA from the liver of control and HF/HE diet fed groups. Products of RT-PCR amplification were separated by agarose gel electrophoresis and visualized with ethidium bromide under ultraviolet transillumination. GAPDH mRNA serves as reference. (B) qRT-PCR for liver hepcidin mRNA in control and HF/HE diet fed group. The values of hepcidin mRNA were normalized to that of GAPDH mRNA. Values are expressed as means S.E.M. for groups of eight rats. *Significantly different from controls ( p < 0.05).
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
contradicting this mechanism because spleen is a haematopoiesis organ in rodents. Changes in haematopoiesis during IR in the rat models were not assessed. However, a study on the role of leptin in haematopoiesis using db/db mice [28] showed that lower spleen weight was secondary to a significant decrease in erythrocytic progenitors. The absence of a decrease in the haemoglobin concentration was explained by a compensatory increase in erythrocytic precursors at the medullar level. As obesity is known to induce leptinresistance, the decrease in spleen weight in the HF/HE group may result from a similar mechanism. Erythropoı¨esis and thus, haemoglobin synthesis, requires the contribution of iron from erythrophagocytosis and digestive absorption. The higher haemoglobin concentration in HF/HE group suggests a higher erythropoı¨esis, which would rely on iron for haemoglobin synthesis ‘‘drawn’’ from storage tissues. The increase in plasma transferrin levels allows a more important iron transport to the bone marrow. To better understand the mechanisms behind the iron metabolism changes in IR, we analyzed the changes in the hepcidin expression, the major regulator of iron homeostasis. Hepcidin is a peptide synthesized by the liver [29] and its role is crucial in iron metabolism. It can stimulate the internalization of ferroportin [15] and thereby inhibit digestive iron absorption and plasmatic release from macrophages. Hepcidin liver expression is decreased by hypoxia, iron deficiency, anaemia, erythropoietin, and during stimulation of erythropoı¨esis [17]. In this study, we show that the liver hepcidin mRNA level was 3.5-fold lower in the HF/HE group when compared with the control. This decrease could be explained by the lower hepatic iron concentration and the increase in erythropoı¨esis in this group. Downregulation of hepcidin expression would permit a more important digestive absorption of iron to compensate for increased iron needs. The short-term effects of variations in glycemia and insulinemia on plasma iron parameters were evaluated. In vitro studies, using cell cultures, suggest that insulin can modulate iron uptake by recruiting the transferrin receptors to the plasma membrane along with the glucose transporters [19]. Another study, using rats, showed that 3 days of high dose insulin treatment (4 UI/ kg) induced iron-loading of brown adipose tissue [30]. Moreover, a hyperinsulinemic euglycemic clamp study in five healthy women showed a progressive improvement of the plasma iron status during the experiment [20]. Here, we show that the injection of glucose has no effect on plasma iron, transferrin and TS. This suggests that hyperinsulinemia has no influence on the iron status parameters. The discrepancy between our data and
369
those of other studies [19,30] could be explained by the use of different insulin concentrations and these reached in our in vivo conditions. The aim of this study was to characterize the changes in iron metabolism in IR and the mechanisms involved. We show that a high fat/high energy diet lead to overweight and IR in Wistar rats. The IR may increase the plasma haemoglobin concentrations consistent with stimulation of erythropoı¨esis, which in turn enhances the needs for iron. In agreement with this prediction, we observed a decrease in tissue iron stores and an increase in transferrin. Erythropoı¨esis stimulation and hepatic iron store reduction lead to downregulation of the hepatic hepcidin expression which subsequently increases iron absorption. Futures studies of this IR model, adjusting the iron intake to achieve a similar liver iron concentration as in the control, would clarify the influence of erythropoı¨esis on hepcidin expression during IR. Our results show a decline in tissue iron stores in the IR group, the opposite is observed in patients [1–5]. A long-term study of the HF/ HE animals would be interesting in order to assess if the hepatic iron overload would appear latter. Interestingly, we show for the first time that IR can lead to a downregulation of hepcidin expression. Human studies focused on iron absorption rates and hepcidin expression, in the metabolic syndrome and HIO-IR, should be performed to unravel the mechanisms behind this iron metabolism disturbance. We also show that provoked hyperglycemia/hyperinsulinemia during IPGTT has no effect on plasma iron, transferrin and TS values. This data suggests that under these conditions, there is no influence of glycemia and insulinemia on plasma iron parameters. Acknowledgments We wish to thank Dominique Bayle, Se´verine Thien and Alexandre Teynie´ for technical assistance. This work was supported in part by Prix de Recherche du Centre Evian pour l’Eau. References [1] M. Haap, A. Fritsche, H.J. Mensing, H.U. Haring, M. Stumvoll, Association of high serum ferritin concentration with glucose intolerance and insulin resistance in healthy people, Ann. Intern. Med. 139 (2003) 869–871. [2] M. Jehn, J.M. Clark, E. Guallar, Serum ferritin and risk of the metabolic syndrome in U.S. adults, Diab. Care 27 (2004) 2422– 2428. [3] T.P. Tuomainen, K. Nyyssonen, R. Salonen, A. Tervahauta, H. Korpela, T. Lakka, et al., Body iron stores are associated with serum insulin and blood glucose concentrations. Population study in 1,013 eastern Finnish men, Diab. Care 20 (1997) 426–428.
370
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
[4] J.T. Salonen, T.P. Tuomainen, K. Nyyssonen, H.M. Lakka, K. Punnonen, Relation between iron stores and non-insulin dependent diabetes in men: case–control study, BMJ 317 (1998) 727. [5] J.M. Fernandez-Real, W. Ricart-Engel, E. Arroyo, R. Balanca, R. Casamitjana-Abella, D. Cabrero, et al., Serum ferritin as a component of the insulin resistance syndrome, Diab. Care 21 (1998) 62–68. [6] J.M. Fernandez-Real, G. Penarroja, A. Castro, F. Garcia-Bragado, I. Hernandez-Aguado, W. Ricart, Blood letting in highferritin type 2 diabetes: effects on insulin sensitivity and beta-cell function, Diabetes 51 (2002) 1000–1004. [7] M.J. Borel, J.L. Beard, P.A. Farrell, Hepatic glucose production and insulin sensitivity and responsiveness in iron-deficient anemic rats, Am. J. Physiol. 264 (1993) E380–E390. [8] R. Moirand, A.M. Mortaji, O. Loreal, F. Paillard, P. Brissot, Y. Deugnier, A new syndrome of liver iron overload with normal transferrin saturation, Lancet 349 (1997) 95–97. [9] A. Guillygomarc’h, M.H. Mendler, R. Moirand, F. Laine, V. Quentin, V. David, et al., Venesection therapy of insulin resistance-associated hepatic iron overload, J. Hepatol. 35 (2001) 344–349. [10] C. Bozzini, D. Girelli, O. Olivieri, N. Martinelli, A. Bassi, G. De Matteis, et al., Prevalence of body iron excess in the metabolic syndrome, Diab. Care 28 (2005) 2061–2063. [11] M.L. Failla, M.L. Kennedy, M.L. Chen, Iron metabolism in genetically obese (ob/ob) mice, J. Nutr. 118 (1988) 46–51. [12] M.L. Kennedy, M.L. Failla, J.C. Smith Jr., Influence of genetic obesity on tissue concentrations of zinc, copper, manganese and iron in mice, J. Nutr. 116 (1986) 1432–1441. [13] R.E. Serfass, K.E. Park, M.L. Kaplan, Developmental changes of selected minerals in Zucker rats, Proc. Soc. Exp. Biol. Med. 189 (1988) 229–239. [14] A. Krause, S. Neitz, H.J. Magert, A. Schulz, W.G. Forssmann, P. Schulz-Knappe, et al., LEAP-1, a novel highly disulfide-bonded human peptide, exhibits antimicrobial activity, FEBS Lett. 480 (2000) 147–150. [15] E. Nemeth, M.S. Tuttle, J. Powelson, M.B. Vaughn, A. Donovan, D.M. Ward, et al., Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization, Science 306 (2004) 2090–2093. [16] G. Nicolas, C. Chauvet, L. Viatte, J.L. Danan, X. Bigard, I. Devaux, et al., The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation, J. Clin. Invest. 110 (2002) 1037–1044. [17] M. Vokurka, J. Krijt, K. Sulc, E. Necas, Hepcidin mRNA levels in mouse liver respond to inhibition of erythropoı¨esis, Physiol. Res. (2006).
[18] L.I. Tanner, G.E. Lienhard, Insulin elicits a redistribution of transferrin receptors in 3T3-L1 adipocytes through an increase in the rate constant for receptor externalization, J. Biol. Chem. 262 (1987) 8975–8980. [19] L.I. Tanner, G.E. Lienhard, Localization of transferrin receptors and insulin-like growth factor II receptors in vesicles from 3T3L1 adipocytes that contain intracellular glucose transporters, J. Cell Biol. 108 (1989) 1537–1545. [20] J.E. Nestler, J.N. Clore, M.L. Failla, W.G. Blackard, Effects of extreme hyperinsulinaemia on serum levels of trace metals, trace metal binding proteins, and electrolytes in normal females, Acta Endocrinol. (Copenh.) 114 (1987) 235–242. [21] A.L. Foy, H.L. Williams, S. Cortell, M.E. Conrad, A modified procedure for the determination of nonheme iron in tissue, Anal. Biochem. 18 (1967) 559–563. [22] R.J. Simpson, M. Lombard, K.B. Raja, R. Thatcher, T.J. Peters, Iron absorption by hypotransferrinaemic mice, Br. J. Haematol. 78 (1991) 565–570. [23] B.D. Hegarty, G.J. Cooney, E.W. Kraegen, S.M. Furler, Increased efficiency of fatty acid uptake contributes to lipid accumulation in skeletal muscle of high fat-fed insulin-resistant rats, Diabetes 51 (2002) 1477–1484. [24] E.W. Kraegen, P.W. Clark, A.B. Jenkins, E.A. Daley, D.J. Chisholm, L.H. Storlien, Development of muscle insulin resistance after liver insulin resistance in high-fat-fed rats, Diabetes 40 (1991) 1397–1403. [25] T. Akiyama, I. Tachibana, H. Shirohara, N. Watanabe, M. Otsuki, High-fat hypercaloric diet induces obesity, glucose intolerance and hyperlipidemia in normal adult male Wistar rat, Diab. Res. Clin. Pract. 31 (1996) 27–35. [26] F.S. Facchini, M. Carantoni, J. Jeppesen, G.M. Reaven, Hematocrit, haemoglobin are independently related to insulin resistance and compensatory hyperinsulinemia in healthy, non-obese men and women, Metabolism 47 (1998) 831–835. [27] M. Barbieri, E. Ragno, E. Benvenuti, G.A. Zito, A. Corsi, L. Ferrucci, et al., New aspects of the insulin resistance syndrome: impact on haematological parameters, Diabetologia 44 (2001) 1232–1237. [28] B.D. Bennett, G.P. Solar, J.Q. Yuan, J. Mathias, G.R. Thomas, W. Matthews, A role for leptin and its cognate receptor in hematopoiesis, Curr. Biol. 6 (1996) 1170–1180. [29] C.H. Park, E.V. Valore, A.J. Waring, T. Ganz, Hepcidin, a urinary antimicrobial peptide synthesized in the liver, J. Biol. Chem. 276 (2001) 7806–7810. [30] A. Korac, M. Verec, V. Davidovic, Insulin-induced iron loading in the rat brown adipose tissue: histochemical and electronmicroscopic study, Eur. J. Histochem. 47 (2003) 241–244.
Study of iron metabolism disturbances in an animal model of insulin resistance Guillaume Le Guenno a,*, Emilie Chanse´aume b, Marc Ruivard c, Be´atrice Morio b, Andrzej Mazur a a
INRA, Equipe Stress Me´tabolique et Micronutriments, UNH, UMR 1019, Clermont-Ferrand/Theix, St-Gene`s-Champanelle 63122, France b INRA, Equipe Me´tabolisme Lipidique et Energe´tique, UNH, UMR 1019, Clermont-Ferrand/Theix, St-Gene`s-Champanelle 63122, France c Service of Internal Medicine Hoˆtel-dieu, C.H.U. Clermont Ferrand, France Received 24 October 2006; accepted 10 February 2007 Available online 9 March 2007
Abstract The relationship between iron and insulin-resistance (IR) is documented by the positive correlation between iron stores and IR. Moreover, some patients exhibited a hepatic iron overload associated with IR (HIO-IR) but the mechanism involved in this overload is not known. Thus, we studied the iron metabolism disturbances in an animal model of IR and the influence of provoked hyperglycemia/hyperinsulinemia on plasma iron parameters. Wistar rats were fed a control or a high-fat/high-energy (HF/HE) diet. Plasma glucose, insulin, iron, transferrin and transferrin saturation (TS) were measured during intra-peritoneal glucose test tolerance (IPGTT) compared to saline. Hemogram, tissue iron concentrations and hepatic hepcidin mRNA expression were determined at the end of experiment. HF/HE rats exhibited higher body and liver weights, increased IR-index and hemoglobin concentration. Iron content was lower in the spleen of HF/HE rats and tended to decrease in the liver as compared to controls. Transferrin values were higher and these of TS lower in HF/HE group. The hepcidin mRNA was 3.5-fold lower in HF/HE rats than in controls. IPGTT had no effect on iron status parameters in both groups. As reflected by higher hemoglobin concentration, IR could increase erythropoı¨esis which enhances iron requirement. Iron stores and TS value decreased leading to a down-regulation of hepcidin expression which increased iron absorption. Hepcidin expression should be investigated in metabolic syndrome and hepatic iron overload associated with IR. # 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Iron; Insulin resistance; High-fat/high-energy diet; IPGTT; Hepcidin; Erythropoı¨esis
1. Introduction The relationship between iron metabolism and metabolic disorders has recently gained interest in both research and clinical practice. Indeed, body iron stores are positively correlated with insulin-resistance (IR), * Corresponding author at: 8 impasse Chabrier 63540 Romagnat, France. Tel.: +33 4 73 61 15 45; fax: +33 4 73 62 46 38. E-mail address: [email protected] (G. Le Guenno).
even in the absence of significant iron overload [1–4]. Ferritin, which reflects body iron stores, is closely associated with IR and can be considered a marker for metabolic syndrome [5]. It has been shown that phlebotomy significantly improves insulin sensitivity in type 2 diabetes [6]. In animals, iron deficiency increases insulin sensitivity [7]. Moreover, a common syndrome called insulin-resistance associated hepatic iron overload (IR-HIO) was described in 1997 by Moirand et al. [8]. This is the most frequent cause of
0168-8227/$ – see front matter # 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2007.02.004
364
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
iron overload in France with an estimated prevalence of 1% of the population, 10 times more widespread than haemochromatosis [9]. A study of 269 subjects with metabolic syndrome showed a prevalence of 14.5% of the IR-HIO [10]. IR-HIO combines an isolated hyperferritinemia with a normal transferrin saturation, steatohepatitis and insulin resistance. It represents the most widespread indication for venesection in referral care units for iron overload [9]. However, the mechanism involved in this overload is presently unknown. A few studies of the perturbation of iron metabolism during IR were focused on the genetic models of obesity, the ob/ob mouse [11,12] and Zucker rats [13]. The obese animals exhibited a decreased iron concentration in the liver [11–13], whereas plasmatic iron, transferrin and spleen iron stores were unchanged when compared with controls [11]. In ob/ob mice, the rate of iron absorption was two-fold greater and the haemoglobin concentration was significantly higher than in lean control mice [11]. The authors concluded that increased erythropoı¨esis in the obese animals provoked a higher iron requirement, which explains the improvement in iron absorption and the decreased liver iron stores. An important advance in our knowledge of iron metabolism was made with the discovery of hepcidin. This peptide, initially described as an antimicrobial peptide [14], is a key regulator of iron stores and it inhibits duodenal iron absorption and iron release by macrophages, thereby provoking the internalization of ferroportin [15]. It has been shown that hepcidin expression is upregulated during infection, inflammation and iron overload [16], whereas the expression is down-regulated by hypoxia, anaemia, iron deficiency, erythropoietin and erythropoietic stimulation [17]. However, even if hepcidin plays a key role in iron absorption, it is unknown how hepcidin expression is modulated in IR. Studies of interactions between iron and glucose metabolism have shown that insulin can cause a rapid and pronounced stimulation of iron uptake by adipocytes by redistributing the transferrin receptor from an intracellular compartment to the cell surface [18]. Transferrin receptors co-localize with the glucose transporters and insulin-like growth factor II receptor in the microsomal membranes of cultured adipocytes, this suggests that regulation of iron uptake by insulin occurs in parallel with glucose uptake [19]. Moreover, a study on the effect of extreme hyperinsulinemia, obtained during a hyperinsulinemic euglycemic clamp in five healthy women, showed a progressive improvement of the plasmatic iron during the experiment [20].
In reviewing the current literature, there is no data to support a mechanism for iron accumulation in IR and the aims of the present work were: To study the perturbations of the iron status and hepcidin expression in an IR animal model induced by an experimental diet. To analyze changes in plasma iron parameters during hyperglycemia/hyperinsulinemia, induced by intraperitoneal injection of glucose, in this animal model. 2. Materials and methods 2.1. Animals and experimental diets Sixteen 3 months old male Wistar rats were randomly divided into two groups. The groups consumed a control or a high-fat/high-energy (HF/HE) diet for 6 weeks. Diets were prepared in the experimental diet preparation unit of Jouy-enJosas (UE300, UPAE INRA Domaine de Vilvert 78352 Jouyen-Josas) and distributed in a semi-liquid form in individual ramekins. Food was weighed daily and prepared for each animal. Tap water was available ad libitum. Animals were housed in individual cages with a normal light cycle (Day 8 a.m. to 8 p.m.), in a temperature-controlled room (22 8C). On a caloric basis, the HF/HE diet consisted of 45% fat (6.7% from groundnut oil, 6.7% from canola oil and 31.6% from lard), 37.6% carbohydrate (25.6% from starch and 12% from sucrose), and 17.4% protein (total 4.68 kcal/g), whereas the control diet contained 13.8% fat (6.7% from groundnut oil and 6.7% from canola oil), 68.8% carbohydrate from starch, and 25.8% protein (total 3.9 kcal/g). Iron was provided as iron sulfate and the intake was 7.8 and 7.7 mg/g body weight in the control and the HF/HE group, respectively. The experimental protocol was approved by the institutional animal care and use committee at INRA (Decree 87– 848 modified by decree 2001–464). 2.2. Experimental procedures One week before the animals were sacrificed, an intraperitoneal glucose tolerance test (IPGTT) was performed at 9 a.m. After 12 h of starvation, four rats of each group received an intraperitoneal injection of 1 g of glucose/kg body weight (G50%, volume (ml) = weight (g) 2) and the remaining eight animals received an isotonic saline solution injection. Blood samples were obtained by retro-orbital puncture at 0, 15, 30, 60 and 120 min. The plasma was collected by centrifugation at 1000 g for 10 min and stored at 80 8C until further analyses. After 6 weeks on the specific diets, the rats were anesthetized by intraperitoneal injection of Imalgene (120 mg/kg; Vetranquil, Merial, Lyon, France) and Diazepam (1.75 mg/kg; Valium, Roche, France) and sacrificed by decapitation. Blood was collected after decapitation for hemogram measurements.
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
Liver, spleen and duodenum were collected, weighed and immediately frozen in liquid nitrogen and stored at 80 8C. 2.3. Glycemia and insulinemia measurements Glycemia was measured immediately using fresh blood (Glucometer Gluco Touch, LifeScan, Inc.). Insulinemia was measured using plasma by ELISA (Rat/Mouse Insulin ELISA kit EZRMI-13K, LINCO research, Inc, USA). Insulin-resistance (IR) was calculated by the IR-index [AUC glycemia (mg/dl) AUC insulinemia (ng/ml)], the Homeostasis Model Assessment score [HOMA = glycemia (mmol/l) insulinemia (mUI/ml)/22.5] and the Quantitative Insulin Sensitivity Check Index [QUICKI = 1/log insulinemia (mUI/l) + log glycemia (mg/dl)]. 2.4. Iron status parameters Haemograms were determined with a Scil Vet ABC counter (Animal Blood Counter, Strasbourg, France). Plasma iron and TIBC were measured using the ‘‘ferrimat kit’’ (bioMerieux SA, Marcy-l’Etoile, France) and the ‘‘TIBC additif’’ kit (bioMerieux SA,) in combination with a Progress Plus Chemistry Analyser automat (Kone, Evry, France). The transferrin saturation (TS) was calculated as fasting plasmatic iron/TIBC and transferrin as TIBC/25. Nonheme iron was determined by a modification of the method of Foy et al. [21] as described by Simpson and Peters [22]. For the duodenum, nonheme iron concentrations were expressed relative to the protein concentration (mg/mg of protein). Tissue protein concentration was estimated using the ‘‘protein BCA Uptima kit’’ (Interchim, Montlucon, France). 2.5. Hepcidin mRNA measurements Total RNA were extracted from the liver using the Qiagen RNeasy Mini kit (Coutaboeuf, France) according to the manufacturer’s instructions. We used 3 mg of total RNA for cDNA synthesis, using the Ready To Go Your First Strand Bead kit (Amersham Pharmacia Biotech, Orsay, France). Polymerase chain reaction (PCR) was carried out using the Pure Taq Ready To Go PCR Beads kit (Amersham Pharmacia Biotech, Orsay, France) and a TC-512 Techne Thermal Cycler (MIDSCI, USA). We performed quantitative RT-PCR using the LightCycler Fast Start DNA Master SYBR Green I kit
365
(Roche Diagnostics, Meylan, France) and a LightCycler (Roche Diagnostics). The hepcidin gene expression was normalized to the GAPDH expression in the same sample. The following primers were used for PCR amplification: GAPDH (50 -CAT GAC CAC AGT CCA TGC CAT CAC-30 and 50 -CAT GTA GGC CAT GAG GTC CAC CAC-30 ), hepcidin (50 -ACA GAA GGC AAG ATG GCA CT-30 and 50 -GAA GTT GGT GTC TCG CTT CC-30 ). These primer pairs produced 458- and 201-bp amplification products, respectively. 2.6. Statistical analysis Results were expressed as means S.E.M. Statistical analysis were performed using the Statview software (SAS Institute Inc., SAS campus drive, Cary, NC, USA). If a significant variance difference was observed between the two groups, a log transformation was performed. The statistical significance of differences between means from two studied groups was assessed by the Student’s t-test. A twoway repeated measures ANOVA, followed by PLSD Fisher’s test, was performed to estimate the effect of group and injection on values obtained during IPGTT. Differences were considered as significant at p < 0.05.
3. Results 3.1. Effect of diet on haematological parameters, body and organ weight The effect of diet on the haematological parameters, body and organ weight are shown in Table 1. After 6 weeks, the HF/HE group had significantly higher body ( p < 0.01) and liver weight ( p < 0.01), whereas spleen weight was lower ( p = 0.03), when compared with the control group. Expressing the organ weight relative to body weight, the difference in the liver weight remained at the edge of significance ( p = 0.04), but the spleen weight was more pronounced ( p < 0.01). The haemoglobin concentration was significantly higher in the HF/ HE group when compared with the control group ( p = 0.03). The other haematological parameters (white blood cells and platelet counts) were not significantly different (data not shown).
Table 1 Effect of diet on body and organ weight, and hemoglobin concentration in the control and high-fat/high-energy diet fed groups
Body weight (g) Liver weight (g) Spleen weight (g) Hemoglobin concentration (g/dl) Values are expressed as means S.E.M. for groups of eight rats. * Significantly different from controls ( p < 0.05).
Control (n = 8)
High-fat/high-energy (n = 8)
514.2 9.2 15.9 0.5 1.02 0.2 14.5 0.9
570.3 9.9 * 19.1 0.9 * 0.89 0.4 * 18.5 1.6 *
366
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
Fig. 1. Effect of intraperitoneal injection of glucose or saline serum on glycemia and insulinemia. Values were measured in control and HF/HE diet fed group at T0, T15, T30, T60 and T120 min. (A) Effect of isotonic saline solution injection on glycemia. (B) Effect of glucose solution injection on glycemia. (C) Effect of isotonic saline solution injection on insulinemia. (D) Effect of glucose solution injection on insulinemia. Values are expressed as means S.E.M. for groups of four rats. Using a two-ways ANOVA, there is no significant effect of group’s factor on insulinemia and glycemia values whereas injection’s factor has significant effect on insulinemia ( p < 0.01) and glycemia values ( p < 0.01). Black circles: HF/HE diet fed group. Empty circles: control diet fed group.
3.2. Effect of diet on IPGTT and evaluation of insulin-resistance Fasting glycemia was significantly higher in HF/HE group ( p < 0.01). The effect of intraperitoneal injection of isotonic saline and glucose solution on glycemia and insulinemia values at T0, T15, T30, T60, T120 min in the control and HF/HE groups are shown in Fig. 1. Using a two-way ANOVA, diet factor had no significant effect on glycemia ( p = 0.06) and insulinemia values ( p = 0.09) at the different times of the IPGTT. Insulin-resistance was higher in HF/HE group whether comparing IR index between groups receiving glucose ( p = 0.04) or isotonic saline injection
( p = 0.01), or the QUICKI index ( p < 0.001). The HOMA index did not differ significantly between the two groups ( p = 0.10). These data are shown in Table 2. 3.3. Effect of diet on tissue nonheme iron concentration The tissue iron concentrations are shown in Table 3. In HF/HE diet group, the nonheme iron content was significantly lower in the spleen ( p < 0.01), tended to decrease in the liver ( p = 0.06), but was unchanged in the duodenum ( p = 0.92), when compared with the control group.
Table 2 Effect of diet on fasting glycemia, fasting insulinemia, IR index, QUICKI and HOMA indexes
Fasting glycemia (mg/dl) Fasting insulinemia (ng/ml) IR index (glucose injection)a IR index (saline injection)a QUICKIb HOMAc
Control (n = 8)
High-fat/high-energy (n = 8)
72.12 1.02 0.90 0.29 2.83.106 9.76.105 9.09.105 2.06.105 0.44 0.02 13.28 4.41
77.87 1.88* 1.35 0.25 5.65.106 9.85.105* 1.81.106 2.15.105* 0.30 0.01* 21.44 4.212
Values are expressed as means S.E.M. for groups of eight rats, except the IR index where four rats of each group received an intraperitoneal injection of isotonic saline or glucose solution. * Significantly different from controls ( p < 0.05). a IR index = AUC glycemia (mg/dl) AUC insulinemia (ng/ml). b QUICKI = 1/(log glycemia (mg/dl) + log insulinemia (mUI/l)). c HOMA = insulinemia (mUI/l) glycemia (mmol/l)/22.5.
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
367
Table 3 Tissue non-heminic iron concentration in control and HF/HE diet fed groups
Liver iron (mg/g tissue) Spleen iron (mg/g tissue) Duodenal iron (mg/mg protein)
Control (n = 8)
High-fat/high-energy (n = 8)
66.76 8.14 424.29 45.10 0.21 0.06
51.51 4.20 304.62 18.04* 0.22 0.09
Values are expressed as means S.E.M. for groups of eight rats. * Significantly different from controls ( p < 0.05).
3.3.1. Effect of diet on plasma iron, transferrin and transferrin saturation during IPGTT The effect of group (control or HF/HE) and injection nature (glucose or saline) on these parameters are shown in Fig. 2 and were assessed using two-way ANOVA. In both groups, plasma iron progressively decreased during the test because of circadian cycle effect. Injection had no effect on plasma iron ( p = 0.87), transferrin ( p = 0.89) and transferrin saturation ( p = 0.92). Group had no effect on plasma iron ( p = 0.42), but had a very significant effect on the transferrin values ( p < 0.01). Comparing the mean values measured during the IPGTT, the HF/HE group
had significantly higher transferrin value ( p < 0.01), lower transferrin saturation ( p = 0.04) and no difference in plasma iron values ( p = 0.19), when compared with the control group. These data are shown in Table 4. 3.4. Effect of diet on hepatic hepcidin expression Liver hepcidin mRNA levels in the two groups were determined by RT-PCR as shown in Fig. 3A. The qRTPCR measurements of the hepatic mRNA levels are shown in Fig. 3B and demonstrate that the hepcidin expression was 3.5-fold lower in HF/HE group ( p = 0.03).
Fig. 2. Effect of intraperitoneal injection of isotonic saline or glucose solution on plasma iron parameters. Values were measured in control and HF/HE diet fed group at T0, T15, T30, T60 and T120 min. (A) Effect of injection on plasma iron in control group. (B) Effect of injection on plasma iron in HF/HE group. (C) Effect of injection on transferrin in control group. (D) Effect of injection on transferrin in HF/HE group. (E) Effect of injection on transferrin saturation in control group. (F) Effect of injection on transferrin saturation in HF/HE group. Values are expressed as means S.E.M. for groups of four rats. Using a two-ways ANOVA, there is no significant effect of injection’s factor on plasma iron parameters whereas group’s factor has significant effect on transferrin values ( p < 0.01). Black circles: isotonic saline solution injection. Empty circles: glucose solution injection.
368
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
Table 4 Average values for plasma iron parameters obtained in control and HF/ HE diet fed groups during intraperitoneal injection
Plasma iron (mmol/l) Plasma transferrin (g/l) Transferrin saturation (%)
Control (n = 8)
High-fat/highenergy (n = 8)
28.67 1.33 4.01 0.11 29.61 2.04
26.16 2.47 4.59 0.18* 23.29 2.76*
Values are expressed as means S.E.M. for groups of eight rats. * Significantly different from controls ( p < 0.05).
4. Discussion The relationship between iron metabolism and IR is well-illustrated in clinical practice by the strong correlation between iron stores and IR [1–5]. Moreover, some of the patients with metabolic syndrome or type 2 diabetes exhibit an insulin-resistance which is associated with hepatic iron overload (IR-HIO). However, the mechanisms behind this overload are unknown and the data from animal studies of genetic obesity models are inconclusive. In this study, we have characterized perturbations in iron metabolism using an animal model rendered IR by an experimental HF/HE diet. Previous studies had shown that 3–4 weeks on a high fat or high energy/high fat diet induces IR as judged by the euglycemic hyperinsulinemic clamp [23–25]. In these studies, the animals exhibited the characteristic abnormalities described in patients with the metabolic syndrome or type 2 diabetes, including increased visceral and muscle fat content, overweight, hepatic steatosis, high fasting glycemia with or without high fasting insulinemia, low HDL values and high plasma fatty acid levels. The HF/HE diet produced similar abnormalities in our study: the animals demonstrated overweight, increased liver weight with macroscopic
steatosis and high fasting glycemia. The development of IR was confirmed by a significant increase in the IRindex and a significant decrease in the QUICKI index, when compared with the control group. To our knowledge, iron metabolism has not been assessed in this IR model. In the present study, we assessed current biomarkers for the blood iron status and the iron content within selected organs serving important roles in iron metabolism, i.e. the liver (iron storage and regulation), the spleen (iron recycling and erythropoı¨esis) and the duodenum (iron absorption). Our data shows that IR leads to a lower iron concentration in the liver and spleen. This is consistent with data from genetic models for IR, such as ob/ob mice [11] and Zucker rats [13], except that in ob/ob mice, the spleen iron concentration is unchanged. In addition, our data shows that there is no difference in the duodenal iron concentration between the two groups. Comparing the mean values of the blood iron status parameters between the control and HF/HE groups, a higher transferrin concentration is associated with lower transferrin saturation in HF/HE group. These changes in the blood parameters of the HF/HE group are associated with lower tissue iron stores, which suggest an increased iron need. Erythropoı¨esis is the state where production of red blood cells is sufficient to maintain a normal level of haemoglobin. In the mammals, iron is mainly used for haemoglobin synthesis. In the HF/HE group, the haemoglobin concentration was higher than in the control group. This is consistent with a study using ob/ ob mice [11] and observations from humans, where the haemoglobin concentration is linked to IR [26,27]. This physiological mechanism may involve insulin stimulation of the sympathetic system and the erythropoietic progenitors. However, the lower spleen weight in HF/ HE diet fed rats, also seen in ob/ob mice [11], is
Fig. 3. (A) Hepcidin mRNA abundance analyzed by RT-PCR (23 cycles) using 3 mg RNA from the liver of control and HF/HE diet fed groups. Products of RT-PCR amplification were separated by agarose gel electrophoresis and visualized with ethidium bromide under ultraviolet transillumination. GAPDH mRNA serves as reference. (B) qRT-PCR for liver hepcidin mRNA in control and HF/HE diet fed group. The values of hepcidin mRNA were normalized to that of GAPDH mRNA. Values are expressed as means S.E.M. for groups of eight rats. *Significantly different from controls ( p < 0.05).
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
contradicting this mechanism because spleen is a haematopoiesis organ in rodents. Changes in haematopoiesis during IR in the rat models were not assessed. However, a study on the role of leptin in haematopoiesis using db/db mice [28] showed that lower spleen weight was secondary to a significant decrease in erythrocytic progenitors. The absence of a decrease in the haemoglobin concentration was explained by a compensatory increase in erythrocytic precursors at the medullar level. As obesity is known to induce leptinresistance, the decrease in spleen weight in the HF/HE group may result from a similar mechanism. Erythropoı¨esis and thus, haemoglobin synthesis, requires the contribution of iron from erythrophagocytosis and digestive absorption. The higher haemoglobin concentration in HF/HE group suggests a higher erythropoı¨esis, which would rely on iron for haemoglobin synthesis ‘‘drawn’’ from storage tissues. The increase in plasma transferrin levels allows a more important iron transport to the bone marrow. To better understand the mechanisms behind the iron metabolism changes in IR, we analyzed the changes in the hepcidin expression, the major regulator of iron homeostasis. Hepcidin is a peptide synthesized by the liver [29] and its role is crucial in iron metabolism. It can stimulate the internalization of ferroportin [15] and thereby inhibit digestive iron absorption and plasmatic release from macrophages. Hepcidin liver expression is decreased by hypoxia, iron deficiency, anaemia, erythropoietin, and during stimulation of erythropoı¨esis [17]. In this study, we show that the liver hepcidin mRNA level was 3.5-fold lower in the HF/HE group when compared with the control. This decrease could be explained by the lower hepatic iron concentration and the increase in erythropoı¨esis in this group. Downregulation of hepcidin expression would permit a more important digestive absorption of iron to compensate for increased iron needs. The short-term effects of variations in glycemia and insulinemia on plasma iron parameters were evaluated. In vitro studies, using cell cultures, suggest that insulin can modulate iron uptake by recruiting the transferrin receptors to the plasma membrane along with the glucose transporters [19]. Another study, using rats, showed that 3 days of high dose insulin treatment (4 UI/ kg) induced iron-loading of brown adipose tissue [30]. Moreover, a hyperinsulinemic euglycemic clamp study in five healthy women showed a progressive improvement of the plasma iron status during the experiment [20]. Here, we show that the injection of glucose has no effect on plasma iron, transferrin and TS. This suggests that hyperinsulinemia has no influence on the iron status parameters. The discrepancy between our data and
369
those of other studies [19,30] could be explained by the use of different insulin concentrations and these reached in our in vivo conditions. The aim of this study was to characterize the changes in iron metabolism in IR and the mechanisms involved. We show that a high fat/high energy diet lead to overweight and IR in Wistar rats. The IR may increase the plasma haemoglobin concentrations consistent with stimulation of erythropoı¨esis, which in turn enhances the needs for iron. In agreement with this prediction, we observed a decrease in tissue iron stores and an increase in transferrin. Erythropoı¨esis stimulation and hepatic iron store reduction lead to downregulation of the hepatic hepcidin expression which subsequently increases iron absorption. Futures studies of this IR model, adjusting the iron intake to achieve a similar liver iron concentration as in the control, would clarify the influence of erythropoı¨esis on hepcidin expression during IR. Our results show a decline in tissue iron stores in the IR group, the opposite is observed in patients [1–5]. A long-term study of the HF/ HE animals would be interesting in order to assess if the hepatic iron overload would appear latter. Interestingly, we show for the first time that IR can lead to a downregulation of hepcidin expression. Human studies focused on iron absorption rates and hepcidin expression, in the metabolic syndrome and HIO-IR, should be performed to unravel the mechanisms behind this iron metabolism disturbance. We also show that provoked hyperglycemia/hyperinsulinemia during IPGTT has no effect on plasma iron, transferrin and TS values. This data suggests that under these conditions, there is no influence of glycemia and insulinemia on plasma iron parameters. Acknowledgments We wish to thank Dominique Bayle, Se´verine Thien and Alexandre Teynie´ for technical assistance. This work was supported in part by Prix de Recherche du Centre Evian pour l’Eau. References [1] M. Haap, A. Fritsche, H.J. Mensing, H.U. Haring, M. Stumvoll, Association of high serum ferritin concentration with glucose intolerance and insulin resistance in healthy people, Ann. Intern. Med. 139 (2003) 869–871. [2] M. Jehn, J.M. Clark, E. Guallar, Serum ferritin and risk of the metabolic syndrome in U.S. adults, Diab. Care 27 (2004) 2422– 2428. [3] T.P. Tuomainen, K. Nyyssonen, R. Salonen, A. Tervahauta, H. Korpela, T. Lakka, et al., Body iron stores are associated with serum insulin and blood glucose concentrations. Population study in 1,013 eastern Finnish men, Diab. Care 20 (1997) 426–428.
370
G. Le Guenno et al. / Diabetes Research and Clinical Practice 77 (2007) 363–370
[4] J.T. Salonen, T.P. Tuomainen, K. Nyyssonen, H.M. Lakka, K. Punnonen, Relation between iron stores and non-insulin dependent diabetes in men: case–control study, BMJ 317 (1998) 727. [5] J.M. Fernandez-Real, W. Ricart-Engel, E. Arroyo, R. Balanca, R. Casamitjana-Abella, D. Cabrero, et al., Serum ferritin as a component of the insulin resistance syndrome, Diab. Care 21 (1998) 62–68. [6] J.M. Fernandez-Real, G. Penarroja, A. Castro, F. Garcia-Bragado, I. Hernandez-Aguado, W. Ricart, Blood letting in highferritin type 2 diabetes: effects on insulin sensitivity and beta-cell function, Diabetes 51 (2002) 1000–1004. [7] M.J. Borel, J.L. Beard, P.A. Farrell, Hepatic glucose production and insulin sensitivity and responsiveness in iron-deficient anemic rats, Am. J. Physiol. 264 (1993) E380–E390. [8] R. Moirand, A.M. Mortaji, O. Loreal, F. Paillard, P. Brissot, Y. Deugnier, A new syndrome of liver iron overload with normal transferrin saturation, Lancet 349 (1997) 95–97. [9] A. Guillygomarc’h, M.H. Mendler, R. Moirand, F. Laine, V. Quentin, V. David, et al., Venesection therapy of insulin resistance-associated hepatic iron overload, J. Hepatol. 35 (2001) 344–349. [10] C. Bozzini, D. Girelli, O. Olivieri, N. Martinelli, A. Bassi, G. De Matteis, et al., Prevalence of body iron excess in the metabolic syndrome, Diab. Care 28 (2005) 2061–2063. [11] M.L. Failla, M.L. Kennedy, M.L. Chen, Iron metabolism in genetically obese (ob/ob) mice, J. Nutr. 118 (1988) 46–51. [12] M.L. Kennedy, M.L. Failla, J.C. Smith Jr., Influence of genetic obesity on tissue concentrations of zinc, copper, manganese and iron in mice, J. Nutr. 116 (1986) 1432–1441. [13] R.E. Serfass, K.E. Park, M.L. Kaplan, Developmental changes of selected minerals in Zucker rats, Proc. Soc. Exp. Biol. Med. 189 (1988) 229–239. [14] A. Krause, S. Neitz, H.J. Magert, A. Schulz, W.G. Forssmann, P. Schulz-Knappe, et al., LEAP-1, a novel highly disulfide-bonded human peptide, exhibits antimicrobial activity, FEBS Lett. 480 (2000) 147–150. [15] E. Nemeth, M.S. Tuttle, J. Powelson, M.B. Vaughn, A. Donovan, D.M. Ward, et al., Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization, Science 306 (2004) 2090–2093. [16] G. Nicolas, C. Chauvet, L. Viatte, J.L. Danan, X. Bigard, I. Devaux, et al., The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation, J. Clin. Invest. 110 (2002) 1037–1044. [17] M. Vokurka, J. Krijt, K. Sulc, E. Necas, Hepcidin mRNA levels in mouse liver respond to inhibition of erythropoı¨esis, Physiol. Res. (2006).
[18] L.I. Tanner, G.E. Lienhard, Insulin elicits a redistribution of transferrin receptors in 3T3-L1 adipocytes through an increase in the rate constant for receptor externalization, J. Biol. Chem. 262 (1987) 8975–8980. [19] L.I. Tanner, G.E. Lienhard, Localization of transferrin receptors and insulin-like growth factor II receptors in vesicles from 3T3L1 adipocytes that contain intracellular glucose transporters, J. Cell Biol. 108 (1989) 1537–1545. [20] J.E. Nestler, J.N. Clore, M.L. Failla, W.G. Blackard, Effects of extreme hyperinsulinaemia on serum levels of trace metals, trace metal binding proteins, and electrolytes in normal females, Acta Endocrinol. (Copenh.) 114 (1987) 235–242. [21] A.L. Foy, H.L. Williams, S. Cortell, M.E. Conrad, A modified procedure for the determination of nonheme iron in tissue, Anal. Biochem. 18 (1967) 559–563. [22] R.J. Simpson, M. Lombard, K.B. Raja, R. Thatcher, T.J. Peters, Iron absorption by hypotransferrinaemic mice, Br. J. Haematol. 78 (1991) 565–570. [23] B.D. Hegarty, G.J. Cooney, E.W. Kraegen, S.M. Furler, Increased efficiency of fatty acid uptake contributes to lipid accumulation in skeletal muscle of high fat-fed insulin-resistant rats, Diabetes 51 (2002) 1477–1484. [24] E.W. Kraegen, P.W. Clark, A.B. Jenkins, E.A. Daley, D.J. Chisholm, L.H. Storlien, Development of muscle insulin resistance after liver insulin resistance in high-fat-fed rats, Diabetes 40 (1991) 1397–1403. [25] T. Akiyama, I. Tachibana, H. Shirohara, N. Watanabe, M. Otsuki, High-fat hypercaloric diet induces obesity, glucose intolerance and hyperlipidemia in normal adult male Wistar rat, Diab. Res. Clin. Pract. 31 (1996) 27–35. [26] F.S. Facchini, M. Carantoni, J. Jeppesen, G.M. Reaven, Hematocrit, haemoglobin are independently related to insulin resistance and compensatory hyperinsulinemia in healthy, non-obese men and women, Metabolism 47 (1998) 831–835. [27] M. Barbieri, E. Ragno, E. Benvenuti, G.A. Zito, A. Corsi, L. Ferrucci, et al., New aspects of the insulin resistance syndrome: impact on haematological parameters, Diabetologia 44 (2001) 1232–1237. [28] B.D. Bennett, G.P. Solar, J.Q. Yuan, J. Mathias, G.R. Thomas, W. Matthews, A role for leptin and its cognate receptor in hematopoiesis, Curr. Biol. 6 (1996) 1170–1180. [29] C.H. Park, E.V. Valore, A.J. Waring, T. Ganz, Hepcidin, a urinary antimicrobial peptide synthesized in the liver, J. Biol. Chem. 276 (2001) 7806–7810. [30] A. Korac, M. Verec, V. Davidovic, Insulin-induced iron loading in the rat brown adipose tissue: histochemical and electronmicroscopic study, Eur. J. Histochem. 47 (2003) 241–244.