Bone marrow iron distribution, hepcidin, and ferroportin expression in renal anemia Liliana Bârsan1, Ana Stanciu 1, Simona Stancu 1,2, Cristina Ca˘ pus¸a˘ 1,2 , Lavinia Bra˘ tescu 3, Eugen Mandache1, Eugen Radu2, Gabriel Mircescu1,2 ‘Dr Carol Davila’ Teaching Hospital of Nephrology, Bucharest, Romania, 2‘Carol Davila’ University of Medicine and Pharmacy, Bucharest, Romania, 3‘Sf Pantelimon’ Hemodialysis Centre - International Healthcare Systems, Bucharest, Romania 1
Objectives: The hepcidin–ferroportin system is involved in both conditions associated with iron-restricted erythropoiesis in renal anemia: iron deficiency and anemia of chronic disorders. As serum hepcidin could aid diagnosis, we investigated its relationships with bone marrow iron distribution, hepcidin–ferroportin expression in bone marrow cells, and peripheral iron indices in non-dialysis chronic kidney disease (CKD) patients. Methods: Fifty-four epoetin and iron naive CKD patients entered this prospective, observational study. According to bone marrow iron distribution (iliac crest biopsy, Perls’ stain), 26 had iron deficiency anemia, 21 anemia of chronic disorders and 7 had normal iron stores. Medullar hepcidin and ferroportin expression (immunofluorescence (IF), semiquantitative scales) and serum hepcidin (Hep25 – ELISA) were the main studied parameters. Results: Low hepcidin and high ferroportin expression by erythroblast and macrophage were seen in iron deficiency anemia, while the opposites were true in anemia of chronic disorders. In regression analysis, higher Hep25 and ferritin predicted hepcidin expression (R 2=0.48; P < 0.0001), while lower ferritin and Hep25 - predicted ferroportin expression (R 2 = 0.29; P = 0.003) by erythroblast; inflammation had no contribution. In ROC analysis, serum hepcidin and ferritin had similar moderate utility in differentiating iron deficiency anemia from anemia of chronic disorders (AUC 0.63 95% CI 0.47–0.79 and 0.76 95% CI 0.61–0.90, respectively). Conclusions: Thus, in anemic epoetin naive non-dialysis CKD patients, hepcidin and ferroportin expression by erythroblast and macrophage are closely related to bone marrow iron distribution. Although the hepcidin–ferroportin system seems regulated by ferritin-driven Hep25, serum hepcidin and peripheral iron indices are of little help in describing bone marrow iron status. Keywords: Bone marrow, Chronic kidney disease, Hepcidin–ferroportin system, Iron distribution, Renal anemia
Introduction Renal anemia shares some features with anemia of chronic disorders (ACD). Iron is sequestered in macrophages, intestinal iron absorption is reduced, and erythropoiesis is impaired by the low iron delivery to maturing erythroblasts.1–3 Altered iron distribution in renal anemia, e.g. iron accumulation in macrophages and eventually bone marrow iron depletion, is suggested by some studies4,5 and could accentuate the ‘functional’ iron deficiency induced by erythropoietin therapy.6,7 On the other hand, ‘absolute’ iron deficiency (IDA) is also common in renal anemia.8–10 Thus, different conditions could induce iron-restricted erythropoiesis and aggravate renal Correspondence to: Cristina Ca˘ pus¸a˘ , ‘Carol Davila’ University of Medicine and Pharmacy - ‘Dr Carol Davila’ Teaching Hospital of Nephrology, 4 Calea Grivit¸ei, sector 1, Bucharest 010731, Romania. Email: [email protected]; [email protected]
© W. S. Maney & Son Ltd 2015 DOI 10.1179/1607845415Y.0000000004
anemia. The distinction between these conditions is clinically relevant, but peripheral iron indices are of little help in diagnosis, which makes guiding iron therapy problematical. Hepcidin limits iron export by binding to the main iron exporter, ferroportin, and could be responsible for the altered iron distribution of renal anemia.3,11 Moreover, high hepcidin levels were reported in chronic kidney disease (CKD).6,12–14 Iron stores repletion, erythropoietic activity, inflammation and the decline in renal function were the main drivers of hepcidin secretion in various clinical studies. Although hepcidin increased after iron therapy and decreased after epoetin administration or after remission of iron-restricted erythropoiesis6,12,15,16 its diagnostic utility in renal anemia is still uncertain. Autocrine/paracrine secretion of hepcidin was described in various territories – splenic17,18 or
Hematology
2015
VOL.
20
NO.
9
543
Bârsan et al.
Bone marrow iron distribution, hepcidin and ferroportin expression
lung19 macrophages – which could participate in altering iron distribution and further complicates the interpretation of serum levels. Recently, Zhang described the iron and inflammation-regulated ferroportin expression by erythroblasts, supposed to allow iron export in case of severe iron deficiency to non-erythropoietic tissues.20,21 Accordingly, the hepcidin–ferroportin system seems to be involved in iron-restricted erythropoiesis, both in IDA and ACD. As there are no studies addressing these issues in renal anemia, we thought to investigate the hepcidin–ferroportin system in relation with bone marrow iron distribution, serum hepcidin levels, peripheral iron indices and inflammation in epoetin and iron naive CKD patients.
Patients and methods Study design Prospective observational single-center study.
Subjects Patients were selected from those admitted to a tertiary nephrology center. Inclusion criteria were age over 18 years, CKD stage 3–5 non-dialysis (eGFR less than 60 ml/min at 2 measurements 3 months apart), anemia (Hb under 11 g/dl 2 times, 2 weeks apart). Patients with anemia of a specific cause (e.g. vitamin B12 or acid folic deficiency, hemolytic anemias), acute inflammatory conditions, recent bleeding, neoplasm, active liver disease (transaminases 2 times normal), severe secondary hyperparathyroidism (PTH higher than 800 pg/ml), or previous iron and epoetin therapy were excluded. Fifty-four subjects (median age 64.5 (range from 30 to 90) years, 56% males, 19% diabetes mellitus (but none with diabetic glomerular nephropathy as cause of CKD), 50% stage 5, 20% stage 4, and 30% stage 3 CKD) with moderate anemia (9.9 (95% CI 9.4–10.2)g/dl) were enrolled. Vascular nephropathies accounted for almost two-thirds of the CKD causes, followed by chronic tubulo-interstitial nephropathies (18%, including three cases of autosomal dominant polycystic kidney disease) and glomerular nephropathies (Table 1). Using bone marrow iron distribution, the patients were classified as having normal iron distribution, iron deficiency or anemia of chronic disorders. Owing to the low number of patients with normal iron distribution, comparisons were made between IDA and ACD groups. The study was conducted in accordance with the Declaration of Helsinki and was approved by the local ethics committee. All patients signed an informed consent to participate in the study.
544
Hematology
2015
VOL.
20
NO.
9
Methods Bone marrow was collected by aspiration from the iliac crest. Hepcidin and ferroportin expression Antibody (Ab) dilutions were performed with Antibody Diluent (Abcam#ab64211, Cambridge, MA). For primary antibodies the following dilutions were used: rabbit anti-Hep 25 (1:1000, Abcam#ab75883, Cambridge, MA, USA), mouse anti-SLC40A1 (1:1000, Abnova#H00030061-B01P, Novus Biologicals, Inc, Littleton, CO, USA), chicken anti-CD34 (1:1000, Abcam#ab14007, Cambridge, MA, USA), rat anti-macrophage antibodies (RM0029-11H3-1:500, Abcam#ab56297, Cambridge, MA, USA). For secondary antibodies, dilutions were prepared just before use in dark, as follows: goat polyclonal to rabbit IgG-H&L FITC (1:1000, Abcam#ab6717, Cambridge, MA, USA), goat polyclonal to mouse IgG-H&L F(ab)2 Fragment (1:400, Abcam#ab7002, Cambridge, MA, USA), rabbit polyclonal to chicken IgY-H&L Texas Red (1:1000, Abcam#ab6751, Cambridge, MA, USA), goat anti-rat IgG and IgM-H&L Alexa Fluor-345 nm (1:1000, Invitrogen Ltd, Paisley, UK). The specificity of each Ab has already been described.22,23 Two indirect IF protocols were designed for both the evaluation of hepcidin and ferroportin expression, and the identification of cell type (Figs. 1 and 2): 1. First with one type of primary Ab and one secondary fluorochrome conjugated Ab, in order to reveal the expression and specificity of hepcidin (rabbit polyclonal to Hep 25 as primary Ab goat polyclonal to rabbit IgG as secondary Ab), ferroportin (mouse anti-SLC40A1 and goat polyclonal to mouse IgGH&L PE, respectively), erythroblast (chicken antiCD34 and rabbit polyclonal to chicken IgY-H&L Texas Red), and macrophage (rat anti-macrophage Ab and goat anti-rat IgG and IgM-H&L Alexa Fluor 345 nm). 2. The second with a mixture of 4 primary antibodies (anti-hep 25, anti-SLC40A1, anti-CD34, anti-macrophage) and 4 secondary antibodies labeled with different fluorochromes as previously described, for additional study of hepcidin and ferroportin distribution at cellular level. The negative control was prepared with phosphate buffered saline (PBS) at pH 7.4 and the same mixture of secondary antibodies. The positive control for hepcidin binding to ferroportin, was performed by a supplementary incubation 1 hour at room temperature (RT) with hepcidin 25 peptides (ab 31875) in dilution of 1 μg/ml before the incubation with the mixture of primary antibodies.
In both protocols the same steps were followed: firstly, specific bone marrow smears were prepared with a mixture of methanol-acetone 1/1 for
Bârsan et al.
Bone marrow iron distribution, hepcidin and ferroportin expression
Table 1 The investigated parameters in study groups Parameters* Age Median (years) >60 years (%) Gender (% male) eGFR Median (ml/min)
Objectives: The hepcidin–ferroportin system is involved in both conditions associated with iron-restricted erythropoiesis in renal anemia: iron deficiency and anemia of chronic disorders. As serum hepcidin could aid diagnosis, we investigated its relationships with bone marrow iron distribution, hepcidin–ferroportin expression in bone marrow cells, and peripheral iron indices in non-dialysis chronic kidney disease (CKD) patients. Methods: Fifty-four epoetin and iron naive CKD patients entered this prospective, observational study. According to bone marrow iron distribution (iliac crest biopsy, Perls’ stain), 26 had iron deficiency anemia, 21 anemia of chronic disorders and 7 had normal iron stores. Medullar hepcidin and ferroportin expression (immunofluorescence (IF), semiquantitative scales) and serum hepcidin (Hep25 – ELISA) were the main studied parameters. Results: Low hepcidin and high ferroportin expression by erythroblast and macrophage were seen in iron deficiency anemia, while the opposites were true in anemia of chronic disorders. In regression analysis, higher Hep25 and ferritin predicted hepcidin expression (R 2=0.48; P < 0.0001), while lower ferritin and Hep25 - predicted ferroportin expression (R 2 = 0.29; P = 0.003) by erythroblast; inflammation had no contribution. In ROC analysis, serum hepcidin and ferritin had similar moderate utility in differentiating iron deficiency anemia from anemia of chronic disorders (AUC 0.63 95% CI 0.47–0.79 and 0.76 95% CI 0.61–0.90, respectively). Conclusions: Thus, in anemic epoetin naive non-dialysis CKD patients, hepcidin and ferroportin expression by erythroblast and macrophage are closely related to bone marrow iron distribution. Although the hepcidin–ferroportin system seems regulated by ferritin-driven Hep25, serum hepcidin and peripheral iron indices are of little help in describing bone marrow iron status. Keywords: Bone marrow, Chronic kidney disease, Hepcidin–ferroportin system, Iron distribution, Renal anemia
Introduction Renal anemia shares some features with anemia of chronic disorders (ACD). Iron is sequestered in macrophages, intestinal iron absorption is reduced, and erythropoiesis is impaired by the low iron delivery to maturing erythroblasts.1–3 Altered iron distribution in renal anemia, e.g. iron accumulation in macrophages and eventually bone marrow iron depletion, is suggested by some studies4,5 and could accentuate the ‘functional’ iron deficiency induced by erythropoietin therapy.6,7 On the other hand, ‘absolute’ iron deficiency (IDA) is also common in renal anemia.8–10 Thus, different conditions could induce iron-restricted erythropoiesis and aggravate renal Correspondence to: Cristina Ca˘ pus¸a˘ , ‘Carol Davila’ University of Medicine and Pharmacy - ‘Dr Carol Davila’ Teaching Hospital of Nephrology, 4 Calea Grivit¸ei, sector 1, Bucharest 010731, Romania. Email: [email protected]; [email protected]
© W. S. Maney & Son Ltd 2015 DOI 10.1179/1607845415Y.0000000004
anemia. The distinction between these conditions is clinically relevant, but peripheral iron indices are of little help in diagnosis, which makes guiding iron therapy problematical. Hepcidin limits iron export by binding to the main iron exporter, ferroportin, and could be responsible for the altered iron distribution of renal anemia.3,11 Moreover, high hepcidin levels were reported in chronic kidney disease (CKD).6,12–14 Iron stores repletion, erythropoietic activity, inflammation and the decline in renal function were the main drivers of hepcidin secretion in various clinical studies. Although hepcidin increased after iron therapy and decreased after epoetin administration or after remission of iron-restricted erythropoiesis6,12,15,16 its diagnostic utility in renal anemia is still uncertain. Autocrine/paracrine secretion of hepcidin was described in various territories – splenic17,18 or
Hematology
2015
VOL.
20
NO.
9
543
Bârsan et al.
Bone marrow iron distribution, hepcidin and ferroportin expression
lung19 macrophages – which could participate in altering iron distribution and further complicates the interpretation of serum levels. Recently, Zhang described the iron and inflammation-regulated ferroportin expression by erythroblasts, supposed to allow iron export in case of severe iron deficiency to non-erythropoietic tissues.20,21 Accordingly, the hepcidin–ferroportin system seems to be involved in iron-restricted erythropoiesis, both in IDA and ACD. As there are no studies addressing these issues in renal anemia, we thought to investigate the hepcidin–ferroportin system in relation with bone marrow iron distribution, serum hepcidin levels, peripheral iron indices and inflammation in epoetin and iron naive CKD patients.
Patients and methods Study design Prospective observational single-center study.
Subjects Patients were selected from those admitted to a tertiary nephrology center. Inclusion criteria were age over 18 years, CKD stage 3–5 non-dialysis (eGFR less than 60 ml/min at 2 measurements 3 months apart), anemia (Hb under 11 g/dl 2 times, 2 weeks apart). Patients with anemia of a specific cause (e.g. vitamin B12 or acid folic deficiency, hemolytic anemias), acute inflammatory conditions, recent bleeding, neoplasm, active liver disease (transaminases 2 times normal), severe secondary hyperparathyroidism (PTH higher than 800 pg/ml), or previous iron and epoetin therapy were excluded. Fifty-four subjects (median age 64.5 (range from 30 to 90) years, 56% males, 19% diabetes mellitus (but none with diabetic glomerular nephropathy as cause of CKD), 50% stage 5, 20% stage 4, and 30% stage 3 CKD) with moderate anemia (9.9 (95% CI 9.4–10.2)g/dl) were enrolled. Vascular nephropathies accounted for almost two-thirds of the CKD causes, followed by chronic tubulo-interstitial nephropathies (18%, including three cases of autosomal dominant polycystic kidney disease) and glomerular nephropathies (Table 1). Using bone marrow iron distribution, the patients were classified as having normal iron distribution, iron deficiency or anemia of chronic disorders. Owing to the low number of patients with normal iron distribution, comparisons were made between IDA and ACD groups. The study was conducted in accordance with the Declaration of Helsinki and was approved by the local ethics committee. All patients signed an informed consent to participate in the study.
544
Hematology
2015
VOL.
20
NO.
9
Methods Bone marrow was collected by aspiration from the iliac crest. Hepcidin and ferroportin expression Antibody (Ab) dilutions were performed with Antibody Diluent (Abcam#ab64211, Cambridge, MA). For primary antibodies the following dilutions were used: rabbit anti-Hep 25 (1:1000, Abcam#ab75883, Cambridge, MA, USA), mouse anti-SLC40A1 (1:1000, Abnova#H00030061-B01P, Novus Biologicals, Inc, Littleton, CO, USA), chicken anti-CD34 (1:1000, Abcam#ab14007, Cambridge, MA, USA), rat anti-macrophage antibodies (RM0029-11H3-1:500, Abcam#ab56297, Cambridge, MA, USA). For secondary antibodies, dilutions were prepared just before use in dark, as follows: goat polyclonal to rabbit IgG-H&L FITC (1:1000, Abcam#ab6717, Cambridge, MA, USA), goat polyclonal to mouse IgG-H&L F(ab)2 Fragment (1:400, Abcam#ab7002, Cambridge, MA, USA), rabbit polyclonal to chicken IgY-H&L Texas Red (1:1000, Abcam#ab6751, Cambridge, MA, USA), goat anti-rat IgG and IgM-H&L Alexa Fluor-345 nm (1:1000, Invitrogen Ltd, Paisley, UK). The specificity of each Ab has already been described.22,23 Two indirect IF protocols were designed for both the evaluation of hepcidin and ferroportin expression, and the identification of cell type (Figs. 1 and 2): 1. First with one type of primary Ab and one secondary fluorochrome conjugated Ab, in order to reveal the expression and specificity of hepcidin (rabbit polyclonal to Hep 25 as primary Ab goat polyclonal to rabbit IgG as secondary Ab), ferroportin (mouse anti-SLC40A1 and goat polyclonal to mouse IgGH&L PE, respectively), erythroblast (chicken antiCD34 and rabbit polyclonal to chicken IgY-H&L Texas Red), and macrophage (rat anti-macrophage Ab and goat anti-rat IgG and IgM-H&L Alexa Fluor 345 nm). 2. The second with a mixture of 4 primary antibodies (anti-hep 25, anti-SLC40A1, anti-CD34, anti-macrophage) and 4 secondary antibodies labeled with different fluorochromes as previously described, for additional study of hepcidin and ferroportin distribution at cellular level. The negative control was prepared with phosphate buffered saline (PBS) at pH 7.4 and the same mixture of secondary antibodies. The positive control for hepcidin binding to ferroportin, was performed by a supplementary incubation 1 hour at room temperature (RT) with hepcidin 25 peptides (ab 31875) in dilution of 1 μg/ml before the incubation with the mixture of primary antibodies.
In both protocols the same steps were followed: firstly, specific bone marrow smears were prepared with a mixture of methanol-acetone 1/1 for
Bârsan et al.
Bone marrow iron distribution, hepcidin and ferroportin expression
Table 1 The investigated parameters in study groups Parameters* Age Median (years) >60 years (%) Gender (% male) eGFR Median (ml/min)
