SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information

Human iron homeostasis is regulated by intestinal iron transport, hepatic hepcidin release, and signals from pathways that consume or supply iron. The aim of this study was to characterize the adaptation of iron homeostasis under hypoxia in mountaineers at the levels of (1) hepatic hepcidin release, (2) intestinal iron transport, and (3) systemic inflammatory and erythropoietic responses. Twenty-five healthy mountaineers were studied. Blood samples and duodenal biopsies were taken at baseline of 446 m as well as on day 2 (MG2) and 4 (MG4) after rapid ascent to 4559 m. Divalent metal-ion transporter 1 (DMT-1), ferroportin 1 (FP-1) messenger RNA (mRNA), and protein expression were analyzed in biopsy specimens by quantitative reverse-transcription polymerase chain reaction (RT-PCR) and immunohistochemistry. Serum hepcidin levels were analyzed by mass spectrometry. Serum iron, ferritin, transferrin, interleukin (IL)−6, and C-reactive protein (CRP) were quantified by standard techniques. Serum erythropoietin and growth differentiation factor 15 (GDF15) levels were measured by enzyme-linked immunosorbent assay (ELISA). Under hypoxia, erythropoietin peaked at MG2 (P < 0.001) paralleled by increased GDF15 on MG2 (P < 0.001). Serum iron and ferritin levels declined rapidly on MG2 and MG4 (P < 0.001). Duodenal DMT-1 and FP-1 mRNA expression increased up to 10-fold from baseline on MG2 and MG4 (P < 0.001). Plasma CRP increased on MG2 and MG4, while IL-6 only increased on MG2 (P < 0.001). Serum hepcidin levels decreased at high altitude on MG2 and MG4 (P < 0.001). Conclusion: This study in healthy volunteers showed that under hypoxemic conditions hepcidin is repressed and duodenal iron transport is rapidly up-regulated. These changes may increase dietary iron uptake and allow release of stored iron to ensure a sufficient iron supply for hypoxia-induced compensatory erythropoiesis. (Hepatology 2013; 58:2153–2162)

Abbreviations
AMS

acute mountain sickness

AMS-C

cerebral-sensitive AMS score

CRP

reactive protein

DMT-1

divalent metal-ion transporter 1

EPO

erythropoietin

FP-1

ferroportin 1

GDF15

growth differentiation factor 15

HAPE

high-altitude pulmonary edema

HIF

hypoxia-inducible factor

IL-6

interleukin 6

LLS

Lake Louise Score

MG2 second day at 4559 m at Capanna Regina Margherita; MG4

fourth day at 4559 m at Capanna Regina Margherita

RES

reticulo-endothelial system

rt-PCR

real time polymerase chain reaction

SpO2

peripheral oxygen saturation

TNSC-EGD

transnasal small-caliber esophagogastro-duodenoscopy

ZH

baseline at 446 m in Zurich

Iron is an essential trace element required as a component of various molecules that sense, transport, and store oxygen.[1] Availability of sufficient amounts of iron is critically important for normal and stress-induced erythropoiesis. Circulating iron levels are affected by intestinal absorption from the diet, iron transport capacity of the blood, iron losses via bleeding and cellular desquamation, and the release of iron from cells such as macrophages and hepatocytes.[2] Inorganic iron is absorbed at the brush border of duodenal enterocytes by the divalent metal-ion transporter 1 (DMT-1; SLC11A2) following reduction by a membrane-associated ferrireductase. Cytosolic iron can be exported by the basolateral iron transporter ferroportin (FP-1; SLC40A1)[3, 4] and subsequently undergoes oxidation by the multicopperoxidase hephaestin before being incorporated into circulating transferrin.

Systemic iron content is tightly regulated,[1, 4, 5] because accumulation of intracellular iron causes cell and tissue damage, presumably by iron-catalyzed generation of reactive oxygen species.[5, 6] Hepcidin, a liver-derived 25 amino acid peptide hormone, has been identified as the key regulator of iron homeostasis[7, 8] (reviewed[6, 9]). Hepcidin negatively regulates the main points of entry into the plasma compartment by decreasing the absorption of iron in the duodenum and limiting the release of recycled iron from macrophages.[1] Mechanistically, hepcidin limits systemic iron influx by binding to the basolateral iron exporter ferroportin and triggering its endocytosis and lysosomal degradation.[1] Hepcidin production is regulated by several stimuli, including iron (as a negative feedback loop), the inflammatory cytokine interleukin-6 (IL-6), endoplasmatic reticulum stress, active erythropoiesis, anemia, and hypoxia.[1, 6, 9] Hypoxia-induced erythropoiesis increases the iron demand in the erythropoietic compartment and induces adaptive changes in the human body such as increased intestinal iron uptake, augmentation of serum iron-binding capacity, and enhanced mobilization of iron from cellular stores.[2] Both hypoxia and anemia induce the synthesis of erythropoietin (EPO) and represent the two principal signals that increase intestinal iron absorption.[10, 11] Hepatic hepcidin production is homeostatically suppressed by low hepatic or extracellular iron and by the erythropoietic need for iron during anemia or hypoxia. It is thought that this is triggered by increased EPO levels or erythropoietic activity, liver hypoxia, or increases in iron levels.[1, 12, 13] However, the exact nature of the hepcidin suppressive signal under these conditions is still unknown but may include circulating factors produced by erythroid precursors in the bone marrow such as growth differentiation factor 15 (GDF15) and twisted gastrulation (TWSG1).[1, 3, 9, 14]

To extend the mechanistic understanding of high-altitude iron homeostasis beyond the level of hepcidin modulation, we examined the adaptive enterohepatic regulation of intestinal iron absorption in humans under hypoxemic conditions. Serum plasma samples and duodenal biopsies procured via unsedated transnasal small-caliber esophagogastro-duodenoscopy (TNSC-EGD) were taken from healthy mountaineers at 446 m and after rapid ascent (<24 hours) to the Capanna Margherita mountain hut at 4559 m. We hypothesized that acute hypoxia suppresses circulating hepcidin, which in turn leads to (1) a rapid up-regulation of iron transporters in the enterocytes, and (2) increased iron consumption and mobilization of storage iron by enhanced erythropoiesis.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information
Participants

Overall exclusion criteria included more than 3 nights above 2500 m in the month preceding the study, chronic diseases necessitating regular medication (including arterial hypertension, coronary heart disease, and pulmonary hypertension), patients with malignancy, transplant patients, patients with clinically significant heart valve disease or with congenital heart or lung disease, lactose intolerance, and celiac disease or relevant food allergies (IgE and/or non-IgE-mediated). The study was approved by the local Ethics Committee (Kantonale Ethikkommission Zürich, Switzerland, EK-1677). Twenty-five healthy participants 22-60 years old with no laboratory signs of iron deficiency were included in this study. Detailed characteristics are given in Table 1. Sample size calculations are given in the Supporting Material.

Table 1. Baseline Characteristics and Relevant Proteins Involved in Iron Pathways of the 25 Participants Subdivided by Gender
 Zurich, 446 mZurich, 446 m
 15 Male Participants10 Female Participants
ParameterUnitnMedian25th-75th PercentilenMedian25th-75th Percentile
  1. Only a physiological difference in hemoglobin concentration was detected. Data are given as median and 25th/75th percentile. BMI: body mass index; GDF-15: serum growth differentiation factor 15; IL-6: serum interleukin 6; CRP: serum c reactive protein; transf.-Sat.: serum transferrin saturation; DMT-1: mRNA expression of duodenal proton-coupled divalent metal-ion transporter 1; FP-1: mRNA expression of duodenal ferroportin 1. mRNA was investigated by reverse transcriptase-polymerase chain reaction (RT-PCR). Expression was normalized to villin.

Ageyears154636.0-51.0104136.3-47.8
BMIkg/m21523.623.0-25.41023.222.0-23.8
ErythropoietinmU/mL155.14.1-7.486.24.7-9.2
GDF-15pg/mL15958826-14059936776-1442
IL-6ng/L150.70.5-1.0100.90.6-1.1
CRPmg/L150.70.4-1.0100.70.0-1.2
Hemoglobing/L1515.2 (P = 0.002)15.0-15.71014.113.0-14.5
Hematocrit%1546.646.0-48.01043.141.8-44.5
Ironµmol/L1523.121.0-31.81020.615.6-24.1
Ferritinµg/L1511367-196109139-159
Transferrinµmol/L1528.028.0-34.01027.025.8-31.8
Transf.- sat.%153930-50103530-46
Hepcidinnmol/L153.401.50-7.9093.502.20-6.75
DMT-1118857-152910577-119
FP-1128464-101912078-156
Study Design

Using a prospective nonrandomized and nonblinded study design, all examinations were carried out at low (Zurich, 446 m above sea level, ZH) and high (Capanna Regina Margherita, 4559 m above sea level, MG) altitudes. For the examinations at ZH level, groups of 2-4 participants arrived in the late afternoon at the University Hospital Zurich and examinations for this study were carried out on the following day. For the high-altitude examinations, groups of 2-4 participants ascended on day 0 by cable car from Alagna Valsesia (1212 m, Italy) to 2971 m and then hiked to the Rifugio Gniffeti at 3611 m, where they spent 1 night. On day 1 (MG1), the participants climbed to the Capanna Regina Margherita at 4559 m. The serologic and endoscopic studies followed on day 2 (MG2) and day 4 (MG4).

Study Protocol and Safety Issues

All investigations were performed without serious adverse events. In two participants, a self-limiting vasovagal reaction occurred during transnasal small-caliber esophagogastro-duodenoscopy at low altitude. Fasting venous and arterial blood samples were taken at baseline (ZH) as well as at MG2 and MG4 at 8 am before endoscopy. All venous blood samples were centrifuged immediately and the plasma was stored in liquid nitrogen and at −80°C after return from Capanna Regina Margherita. In addition, peripheral oxygen saturation (SpO2) was monitored by pulse oximetry (finger clip measurement using Infinity by Dräger, 3097 Liebefeld, Switzerland or Colin next BP 88, Mediana Technologies Corporation, San Antonio, TX). Arterial blood gas analysis was performed, including measurement of hemoglobin and hematocrit (ABL, Radiometer, Copenhagen, 8800 Thalwil, Switzerland). Partial pressure of oxygen, hemoglobin, and hematocrit could not be analyzed in two arterial blood samples of one participant because of a technical defect of the blood gas analyzer.

Acute mountain sickness (AMS) scores were determined at baseline on each test day in the Capanna Regina Margherita based on the Lake Louise scoring (LLS) system and the use of a cerebral-sensitive (AMS-C) score of the Environmental Symptom Questionnaire.

This study was not designed to assess the effects of dexamethasone on high-altitude pathophysiology in a randomized double-blind placebo-controlled fashion. For safety reasons subjects with (1) significant high-altitude pulmonary edema (HAPE) susceptibility (defined by experience of an HAPE in participant's history); (2) without HAPE susceptibility but an LLS greater than 5 in the morning or evening of MG2; or (3) symptomatic subjects, as indicated by the responsible study physician, were treated with 2 × 8 mg/day dexamethasone (9-fluor-16a-methylprednisolone, Dexamethasone Galepharm, 4 mg, Kuesnacht, Switzerland) starting on the evening of MG2, i.e., after the last experiment of that day was performed. Overall, 14 subjects were treated with dexamethasone. In addition, one participant with a mild HAPE diagnosed on day 4 had to be treated with Tadalafin (20 mg/day) until descent. Minor symptoms such as moderate headache and nausea were treated with 500-1,000 mg paracetamol (Dafalgan, Bristol-Myers Squibb, Baar, Switzerland). Diagnosis and prescription of medication were done by an experienced senior critical-care physician (M.M.).

Transnasal Small-Caliber TNSC-EGD

Unsedated TNSC-EGD was performed using small-caliber endoscopes (FG-16V with light source LH-150PC Pentax, 2-36-9, Maenocho, Itabashiki, Tokyo, Japan). All participants fasted from 10 pm the day before endoscopy. Endoscopy was performed between 8 am and 9 am. Mucosal biopsies were taken from the gastric antrum (one biopsy) and the second part of the duodenum (six biopsies). Biopsy specimens were directly transferred into plastic cups on ice (0°C) and immediately after the end of the endoscopy procedure (i.e., 5-10 minutes after biopsy) into liquid nitrogen. Endoscopy with biopsies was performed in 24 participants at baseline level (ZH) and in 18 and 23 participants at MG2 and MG4, respectively. Two participants were excluded from endoscopy at study days MG2 and MG4 because of nasal discomfort and vasovagal reaction at baseline endoscopy but underwent all other investigations. Six participants could not be investigated on MG2 because four did not reach Capanna Regina Margherita in time due to bad weather conditions and two participants had severe AMS, precluding them from endoscopy.

Laboratory Analyses

Analyses of the inflammatory markers iron, ferritin, and transferrin were carried out in plasma samples in the Department of Clinical Chemistry at the University Hospital Zurich using standard methods.

TaqMan Real-Time Polymerase Chain Reaction (PCR)

Total RNA was isolated from human duodenal biopsy using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. Messenger RNA (mRNA) was reverse-transcribed to complementary DNA (cDNA) using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Mannheim, Germany). Real-time PCR was performed using PCR-Primers in combination with FAST SYBR Green PCR Master Mix (Applied Biosystems). Expression levels were normalized to both villin and HPRT1 as housekeeping genes. Both were unchanged under the given conditions (data are reported for villin only). For primer design, Primer Express software (Applied Biosystems, Foster City, CA) was used (Supporting Material).

Immunohistochemistry

Frozen unfixed biopsy specimen sections were used for immunohistochemical staining performed as described.[15] Antibodies against FP-1 were raised by immunization of rabbits with the peptide (FPN1 240-254). Serum from the final bleed was used for affinity purification. Sections were incubated with 0.1 mL of 300 μg/mL affinity-purified anti-FP-1 (240-254) antiserum as described[15] and a biotin-coupled goat antirabbit immunoglobulin IgG as a secondary antibody (Dako, Vienna, Austria) in a 1:500 dilution. For a control staining, antibodies were preincubated with ferroportin peptide against which the antibody was raised for 1 hour (Fig. 2B). Subsequently, the reaction product was developed as described.[15, 16] For immunofluorescence, cryosections were fixed for 5 minutes in phosphate-buffered saline (PBS) containing 1% formaldehyde and permeabilized for 10 minutes in PBS containing 0.1% Triton X-100. Blocking of nonspecific binding sites was performed for 1 hour with PBS containing 1.5% bovine serum albumin (BSA) and 0.1% Tween 20. Slides were incubated with 0.1 mL of 3 μg/mL anti-HIF2α (NB100-122, Novus, Cambridge, UK) antibodies overnight. Bound antibodies were detected with Alexa Fluor 488 antirabbit secondary antibody (Life Technologies, Darmstadt, Germany). DNA was visualized with DNA-specific fluorochrome DAPI (Sigma, Taufkirchen, Germany).

Hepcidin Analysis

Plasma hepcidin measurements were performed as described recently.[17, 18] The lower limit of detection (LLOD) of this method was 0.5 nM. The median reference level of serum hepcidin-25 is 4.5 nM for men, 2.0 nM for premenopausal women, and 4.9 nM for postmenopausal women. Samples with a hepcidin level below the LLOD (<0.5 nM) were assigned with a concentration of 0.25 nM for statistical analyses.

Enzyme-Linked Immunosorbent Assay (ELISA)

EPO was measured using immunoassays (Human Erythopoietin, Quantikine, R&D Systems, Abingdon, UK). Serum GDF15 expression was quantified by ELISA (BioVendor, RD191135200R, Heidelberg, Germany). All ELISAs were processed according to the manufacturer's instructions.

Statistics

Statistical methods are explained in the Supporting Materials.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information
Study Population and Baseline Characteristics

Baseline serum iron parameters at 446 m were lower in female subjects with a significant difference only in hemoglobin concentration (Table 1). Serum hepcidin as well as DMT-1 and FP-1 mRNA expression in duodenal biopsies showed no gender-specific difference under either baseline and hypoxic conditions (data not shown). Therefore, the following analyses were not separated by gender.

Oxygenation and AMS

Median arterial oxygen saturation measured at 7 am after the first night at Capanna Margherita at 4559 m altitude was 76% and partial pressure of oxygen was 5.2 kPa. Oxygenation increased at day 4 at high altitude (median oxygen saturation was 83%, P = 0.04 versus day 2; and median partial pressure of oxygen was 5.8 kPa, P ≤ 0.0001 versus day 2) but remained lower than baseline levels. After rapid ascent to Capanna Margherita mountain sickness scores were highest on day 2 and declined on day 4 at high altitude (Table 2). Due to the occurrence of AMS, 14 subjects had to be treated with dexamethasone.

Table 2. Measured Parameters for Diagnosis of Acute Mountain Sickness, Oxygenation, Erythropoietic Response, Systemic Inflammation, and Iron Pathways at Different Altitudes
 ZurichCapanna Margherita Day 2Capanna Margherita Day 4
 446 m1st Examination Day, 4559 m2nd Examination Day, 4559 m
ParameterUnitnMedian25th-75th PercentilenMedian25th-75th PercentilenMedian25th-75th Percentile
  1. Data are given as median and 25th/75th percentile. LLS Global/Self: Global and self-assessment Lake Louise Score; AMS-C: Acute mountain sickness C score; sO2: oxygen saturation; PaO2: partial pressure of oxygen; GDF-15: serum growth differentiation factor 15; IL-6: serum interleukin 6; CRP: serum c reactive protein; Transf.-Sat.: serum transferrin saturation; DMT-1: mRNA expression of duodenal proton-coupled divalent metal-ion transporter 1; FP-1: mRNA expression of duodenal ferroportin 1. mRNA was investigated by reverse transcriptase-polymerase chain reaction (RT-PCR). Expression was normalized to villin. Significant differences versus ZH are indicated by Bonferroni corrected P values (***P ≤ 0.0001).

LLS251.001.00-2.00255.00 (P = 0.006)3.00-7.00253.00 (P = 0.03)1.50-4.00
AMS-C250.000.00-0.07250.56***0.31-1.62250.26 (P = 0.002)0.04-0.71
sO2%259897-982576***74-832583***80-85
PaO2kPa2512.311.1-13.5245.2***5.1-5.6245.8***5.5-6.5
ErythropoietinmU/mL215.54.3-7.72169.0***53.6-94.42125.2***19.4-43.2
GDF-15pg/mL24949828-1341211190***977-1602181059847-1451
IL-6ng/L250.80.5-1.0212.1***0.9-3.9210.70.1-1.9
CRPmg/L250.70.2-1.0212.41.2-4.8212.6 (P = 0.02)0.9-3.7
Hemoglobing/L2515.013.9-15.42413.712.6-14.32412.9 (P = 0.003)12.1-13.7
Hematocrit%2546.042.9-47.42441.9***39.0-43.82439.6***37.5-42.2
Ironµmol/L2521.917.3-27.42122.115.6-30.12111.2***7.7-14.7
Ferritinµg/L2510765-1732180.0 (P = 0.001)48-1562163***34-111
Transferrinµmol/L2528.027.0-32.52131.0 (P = 0.004)28.0-38.02130.0 (P = 0.02)28-35.0
Transf.- sat.%253731-47213626-442118***13-23
Hepcidinnmol/L243.451.75-7.55210.90 (P = 0.009)0.25-1.7200.25***0.25-0.25
DMT-11910160-12310431 (P = 0.003)337-169815689***384-1425
FP-1218671-13011166 (P = 0.02)127-26118290***178-328
Serum Iron Parameters, Erythropoietic Response, and Systemic Inflammation

Hemoglobin concentration showed a minor decrease on day 4 and hematocrit decreased on days 2 and 4. Serum iron concentration, ferritin concentration, and transferrin saturation were all lowest on day 4. Ferritin concentration had already decreased by day 2 in all participants and transferrin levels increased, indicating increased mobilization of storage iron (Table 2, Fig. 1A). Even in subjects with either elevated transferrin saturation or ferritin at baseline the response to hypoxia was similar to all other participants of this study (Supporting Table 1). Serum hepcidin levels decreased following acute exposure to hypoxia on day 2 and were below the LLOD on day 4 (Table 2, Fig. 1B). Serum EPO and growth differentiation factor 15 (GDF-15) levels peaked on day 2, reflecting immediate activation of compensatory erythropoiesis. EPO levels remained elevated on day 4, whereas GDF-15 levels returned to baseline levels on day 4 (Table 2, Fig. 1B). Levels of the proinflammatory cytokine IL-6 peaked on day 2 and returned to baseline levels on day 4, and the inflammatory marker C-reactive protein (CRP) increased on day 2 and day 4 (Table 2, Fig. 1B).

image

Figure 1. Serum iron parameters. Effect of hypoxia (MG2 and MG4) on: (A) serum iron parameters and (B) erythropoiesis (serum EPO, GDF-15), hepatic hepcidin levels and systemic inflammation. Median (line) and single values are plotted. Significant differences versus baseline levels (ZH) are indicated by Bonferroni corrected P values (***P ≤ 0.0001).

Download figure to PowerPoint

image

Figure 2. Duodenal FP-1 transporter expression. (A) FP-1 histological staining of duodenal epithelium on frozen unfixed sections in three representative volunteers. The intensity of FP-1 staining is highly increased at high altitude (MG2 and MG4) compared to baseline levels (ZH) with highest intensity at the basolateral membrane. Broken lines mark the putative position of the basolateral membrane. (N) Nucleus of one enteroycte. Arrows give the orientation of the enterocyte from basolateral to luminal. (B) To exclude nonspecific binding of the primary and secondary antibodies, control staining on sequential sections from the same volunteer were performed, incubated with ferroportin antibodies preincubated with and without ferroportin peptide (FPN1 240-254).

Download figure to PowerPoint

Duodenal FP-1 Protein Expression

To further investigate the adaptive changes in intestinal iron transporter expression in response to hypoxia, duodenal FP-1 expression was visualized on frozen biopsy specimen sections. Immunohistochemical staining for FP-1 protein of duodenal epithelium showed a polarized staining at the basolateral membrane of duodenal enterocytes on all study days (Fig. 2A). There was a marked increase in the intensity of immunostaining under hypoxic conditions on day 2 and day 4, indicating increased duodenal FP-1 protein expression.

Duodenal Expression of DMT-1 and FP-1 mRNA

Next, we analyzed whether an additional increase in mRNA expression of FP-1 contributes to the observed increase in its protein abundance. As a hypoxia-inducible intestinal marker gene, DMT-1 mRNA expression was quantified in parallel. Acute exposure to hypoxia induced a marked increase in DMT-1 as well as FP-1 mRNA expression (Table 2, Figs. 2B, 3A) in duodenal biopsy specimens on day 2 and day 4 (median percentage increase DMT-1 compared to ZH: MG2: 635%, MG4: 691%; FP-1: MG2: 199%, MG4: 328%, Table 2). DMT-1 and FP-1 mRNA expression levels were closely linearly associated (R2: 0.72, P ≤ 0.0001), indicating that duodenal DMT-1 and FP-1 expression is regulated in the same direction (Fig. 3C).

image

Figure 3. Duodenal iron transporter expression. Effect of hypoxia (MG2 and MG4) on: (A) DMT-1 and (B) FP-1 mRNA expression in the human duodenal epithelium. RT-PCR was performed with RNA samples isolated from duodenal biopsies. Expression was normalized against villin. Median (line) and single values are plotted. Significant differences versus baseline levels (ZH) are indicated by Bonferroni corrected P values (***P ≤ 0.0001). (C) Association of duodenal log transformed DMT-1 and FP-1 mRNA expression for all conditions. The regression line (solid) and 95% confidence bands (dashed) are plotted.

Download figure to PowerPoint

Duodenal HIF2a Expression

Hypoxia leads to the induction of the hypoxia-inducible factors (HIFs) that typically mediate the adaptive response in the expression of hypoxia-responsive genes. HIF-2α protein expression and localization was visualized in frozen biopsy specimen sections. No HIF-2α protein could be detected at baseline altitude but was visible on days 2 and 4 at high altitude (Fig. 4A). In parallel, HIF-2α mRNA expression in the human duodenal epithelium increased on day 4 (Fig. 4B), suggesting HIF-2α mRNA stabilization.

image

Figure 4. Duodenal HIF-2α expression. (A) Representative confocal microscopic micrographs of duodenal HIF-2α immunohistochemistry. Frozen sections from duodenal biopsies were stained against HIF-2α (green). Nuclear HIF-2α as well as cytoplasmatic staining was seen only under hypoxic conditions (MG2 and MG4) but not at baseline levels (ZH). For a better histological orientation, a merge with the respective phase contrast images is shown in the lower row. All sections were labeled with the DNA-specific fluorochrome DAPI. (B) Effect of hypoxia on HIF-2α expression in the human duodenal epithelium. The expression of HIF-2α expression was investigated by RT-PCR. Expression was normalized against villin. Median (line) and single values are plotted. Significant differences versus baseline levels (ZH) are indicated by Bonferroni corrected P values.

Download figure to PowerPoint

Effects of Dexamethasone and High-Altitude Illness

Dexamethasone treatment induced no changes in oxygenation and serum iron parameters as analyzed in one mixed-effect model testing the effect of altitude and dexamethasone treatment. However, as expected, the median levels of the proinflammatory cytokine IL-6 were lower in the dexamethasone-treated participants on day 4 (IL-6 without treatment: 1.75 [0.98-4.32] ng/L versus IL-6 under dexamethasone treatment: 0.34 [0.00-0.73] ng/L, P = 0.001). In addition, serum EPO levels were lower in the treatment group on day 4 (EPO without treatment: 41.9 [25.7-50.9] ng/L versus EPO under dexamethasone treatment: 20.8 [15.9-26.5] ng/L, P = 0.003). No differences in intestinal iron uptake, systemic iron homeostasis, or hemoglobin were found in participants with high HAPE susceptibility and those without HAPE susceptibility but an LLS greater than 5 (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information

Hypoxic conditions such as high-altitude or hypoxia-related disorders lead to an activation of compensatory erythropoiesis and subsequent iron consumption.[19] The primary site of control of iron uptake is the small intestine. Protection against iron overload in humans is mediated by liver-derived hepcidin, which contributes to the internalization and degradation of ferroportin-1 protein as the principal basolateral iron exporter.[14] In addition, the reticuloendothelial system (RES) is crucial for iron homeostasis, and is likely to be responsible for acute changes in serum iron indices under hypoxic conditions by short-term ferroportin-induced iron release by macrophages.[19]

However, the integrative mechanisms underlying adaptation of iron homeostasis to hypoxic conditions in humans are still not well defined. The present study investigated these regulatory mechanisms in both the human intestine and the circulation after a rapid ascent to high altitude. Participants underwent unsedated transnasal EGD with duodenal biopsy sampling without serious adverse events. This technique has been shown to avoid the negative side effects of sedoanalgesia compared to peroral EGD and has a higher patient tolerability, acceptability, and safety.[20]

At high altitude (4559 m) a significant decrease in the oxygen saturation and the partial oxygen pressure was observed. As an adaptation to a decreased oxygen pressure, erythropoiesis is thought to be rapidly activated, inducing an increased iron demand in the erythropoietic compartment.[3, 4] A compensatory elevation of erythrocyte volume as reflected by hematocrit was not detected in our participants. This observation might be explained by the study design with a short time interval between rapid ascent and the analyses done. Normally, an increase of hemoglobin levels due to stimulated erythropoiesis takes 2 to 4 weeks. Furthermore, baseline investigations were performed under hot midsummer weather conditions in Zurich, which induced signs of hemoconcentration in the majority of the participants. This is supported by the fact that mild vasovagal reactions occurred only at the baseline level. Piperno et al.[21] presented a slight increase in erythrocyte volume after a longer time interval of adaptation to high altitude including 3 days rest at 3400 m and a 5-day ascent to 5400 m. Similar changes in hematocrit could also be detected at the Capanna Regina Margherita level after a longer time interval to allow adaptive erythropoiesis.[22] However, it is important to note that established markers for an activated erythropoiesis were already present on day 2 and day 4 of our study, as EPO and GDF-15 serum levels were significantly increased compared to baseline levels.[21] A marked EPO secretion has been observed on day 2 at high altitude, reflecting the intense need for oxygen due to fast and stressful ascent to 4559 m. These levels were almost double those in previous reports. In line with these findings was the observation of decreased serum iron levels, reflecting iron consumption for increased hemoglobin and erythropoietic cell output.[21] This demand for systemic iron was followed by a compensatory increase in transferrin levels observed at day 4 in our study. It cannot be ruled out that these findings are a result of the combination of exercise, which could alter muscle iron accumulation and myoglobin homeostasis, and hypoxia. However, most of the findings in this study are in agreement with previous literature from both humans and experimental animals using an experimental setting without physical exertion.[4, 19]

In previous studies, mediators of iron homeostasis have been investigated independently under high-altitude conditions. Hypoxia caused an increase of circulating IL-6,[23] whereas serum hepcidin levels were suppressed under these conditions.[1, 21, 24] In line with these findings, the measured IL-6 serum levels in our study were increased, indicating a subtle systemic inflammatory response, which could be slightly attenuated as expected by treatment with dexamethasone. Suppression of hepcidin expression represents the mechanistic link between hypoxia and the observed changes in systemic iron availability. However, hepcidin suppression at high altitude is not driven by a reduction in iron stores.[25] Despite the up-regulation of IL-6 as an activator of HAMP gene expression, the clearance of serum hepcidin levels under hypoxic conditions indicates a dominant-negative regulatory (iron-independent) impact of hypoxia-induced erythropoiesis over inflammatory cytokines. This could be based either on direct hypoxia-mediated effects on hepcidin expression, or be a consequence of hypoxia-induced erythropoiesis and iron consumption for heme synthesis with a subsequent decrease of circulating iron levels.[32] Our data are in concert with the report of Huang et al.[26] which showed that the erythropoietic drive might inhibit both inflammatory and iron-sensing pathways in mice. Nonetheless, cytokines such as IL-6 can promote iron retention in macrophages by hepcidin-independent pathways, which would also result in low serum iron levels.[27] Such changes are always paralleled by increased circulating ferritin levels. However, the opposite, namely, decreased serum ferritin levels, were observed in our study, thus ruling out IL-6-mediated iron retention under hypoxic conditions. This response to hypoxia was even present in subjects with elevated baseline transferrin saturation or ferritin levels. However, we cannot exclude the presence of a genetic predisposition for later clinically relevant hemochromatosis (e.g., C282Y homozygotes) in these subjects.

Hepcidin represents the principal regulator of iron release from the intestine and iron stores[28] into the circulation and its expression is down-regulated by hepatic HIF-2 through EPO-mediated increase in erythropoiesis.[13] FP-1, the basolateral iron exporter expressed in enterocytes, is regulated by hepcidin. The latter has been shown to trigger internalization and degradation of FP-1 protein, consequently limiting the transfer of iron from the intestinal lumen to the circulation.[4, 29] We observed increased FP-1 protein levels in duodenal biopsies taken at high altitude, due to high iron demand during increased erythropoiesis. Maximum FP-1 levels peaked at day 4 when serum hepcidin was no longer detectable. A 2 to 3-fold increase in FP-1 mRNA levels in duodenal tissue suggests that FP-1 protein accumulation cannot be exclusively explained by an absent hepcidin-induced degradation, but is also at least in part due to a transcriptional regulation. Interestingly, an even more pronounced 6-fold up-regulation of apical DMT-1 mRNA was detected. The linear correlation between DMT-1 and FP-1 transcripts suggests that those two transporters could be under the control of the same regulator in humans, as has been reported for different disease states of iron deficiency and overload.[30] Besides its well-known role of repression of intestinal FP-1, hepcidin could also regulate intestinal DMT-1 by an unknown signaling pathway, as it has been shown that acute changes in hepcidin concentration induce proteasomal-mediated degradation of DMT-1.[31] Furthermore, in hypoxic conditions and in simulated disease conditions (Hepc−/−), this regulator might be HIF-2, possibly overriding the effect of the hepcidin-ferroportin axis.[13, 32]

It was recently shown that in conditional knockout mice lacking either HIF1α or HIF2α, DMT-1 and FP-1 are both target genes activated by the hypoxia-induced transcription factor HIF-2α.[30, 32] HIF-2α protein is degraded under conditions of sufficient iron and oxygen availability but accumulates during iron deficiency and hypoxia (reviewed[19]). In the present study, HIF2α protein could be detected under hypoxic conditions in duodenal tissues at high altitude but was not detectable under normoxic baseline conditions. This activation, together with the linear correlation between the expression levels of different candidate target genes, implies that HIF-2α might be involved in the regulation of human DMT-1 and FP-1, similar to the findings in mice.[33] Furthermore, the possible contribution of HIF-2α as a transcriptional activator of FP-1 is in concert with an increased iron transporter mRNA expression in duodenal biopsies under hypoxia. In the limited remaining tissue HIF-1α expression was less consistent in immunohistochemistry without detectable changes under hypoxia (data not shown).

The present study uncovers the intestinal regulatory mechanisms underlying adaptive changes in iron metabolism under hypoxic conditions for the first time in humans. Most important, all known regulatory events in the circulation and the intestine were assessed within the experimental setting of this single study. Our findings provide a comprehensive overview of hypoxia-induced adaptive mechanisms in humans and could have implications for the treatment of hypoxia-related acute and chronic disorders in the future.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information

This study represents the analysis of adaptive enterohepatic regulation of intestinal iron absorption under hypoxic conditions and is part of a cooperative project (principal investigators: M. Maggiorini and Th. Lutz) supported by the Zurich Centre for Integrative Human Physiology (ZIHP).

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgment
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
hep26581-sup-0001-suppinfo1.docx22KSupporting Table 1: Individual data of 6 different subjects with elevated transferrin saturation (A) or elevated ferritin levels (B). In all three subjects with elevated transferrin saturation a gradual decrease of iron parameters under hypoxia was detected. The same is true for those three subjects with elevated ferritin levels. This fact clearly indicates that the response to hypoxia in these subjects is similar to all other participants of this study.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.