Iron is an essential element for human life [1
] and also a fundamental catalytical co-factor for several enzymes, including aconitase in the Krebs cycle and electron transfer system (ETS) complexes in mitochondria [4
Hence, an impairment of iron homeostasis, in the form of iron deficiency or iron overload, not only leads to increased production of reactive oxygen species (ROS) [1
] but also to inappropriate energy availability. For instance, in iron overload disorders, excess iron accumulates in vital organs, like the liver [8
], where it affects biochemical processes and mitochondrial respiratory capacity, as we have recently shown [9
]. Altogether, these perturbations result in inadequate energy supply and may explain the phenotype of fatigue often reported by hemochromatosis patients [10
Therefore, maintaining the iron balance is fundamental to preserve multiple enzymatic pathways ensuring energy requirements [11
], and represents a public health concern [12
Nevertheless, the impact of iron homeostasis perturbations on the overall body metabolic profiles (the metabolome) have not been explored thus far.
Metabolomics aims to comprehensively study metabolism, through the identification and quantification of small molecules, defined metabolites that reflect the phenotype investigated. As such, metabolomics is a dynamic approach of multiple biochemical pathway analysis, that also takes into account genetics, aging, environmental factors and lifestyle, thus allowing to identify specific metabolic signatures typical of a phenotype [13
]. Metabolomic analysis can be performed in different tissues [16
] and biofluids [17
], and in combination with the information coming from the periphery tissue metabolites, it offers the possibility to better integrate and characterize systemic and tissue specific regulation of enzymatic pathways.
However, investigating peripheral metabolites in a mouse model is still challenging, especially when blood should be collected overtime, given the limited amount of blood that can be taken. Back in the 1960s Guthrie and Susi developed a system for the screening of phenylketonuria in newborn infants where only a few drops of whole blood, collected on filter paper, called dried blood spots (DBS), were sufficient for the diagnosis [19
]. Nowadays, the use of DBS is very common in clinics [21
] nevertheless its application is limited by several factors, including the influence of hematocrit (HCT), the volume of the blood spotted onto the filter paper, and spot homogeneity [17
We have recently validated the use of a commercially available tool, allowing for volumetric absorptive microsampling (VAMS), in combination with a mass spectrometry (MS)-based untargeted metabolomics workflow [17
]. This technology, similar to DBS sampling, is a novel system that allows accurate and representative single-drop blood (10 µL) collection [24
] and overcomes the issues with hematocrit.
Moreover, considering the minimal amount of blood required, using such approach in animal studies supports the principle of the three Rs (3Rs) by reducing animal sample size and minimizing animal pain [26
To date, there are no studies combining VAMS and untargeted metabolomics in animal models. Therefore, in the current study we first aimed to test the suitability of this technology in a mouse model and then to investigate the metabolic signature driven by dietary iron overload. Giving the clinical importance of iron overload disorders, metabolic signature of altered iron homeostasis was investigated to identify circulating biomarkers that correlate with tissue dysfunction.
2. Materials and Methods
All materials were obtained from Sigma-Aldrich (Seelze, Germany), unless stated otherwise. Acetonitrile was purchased from VWR International (Radnor, PA, USA). Water was obtained from a Milli-Q water purification system equipped with a LC-Pak polisher (Merck, Darmstadt, Germany). All chemicals and solvents were of analytical grade or higher purity. The metabolomic amino acid mix standard, containing isotopically labeled amino acids, was purchased from Cambridge Isotope Laboratories (Tewksbury, MA, USA).
2.2. Animal Models
Wild-type C57BL/6 animals were maintained and bred in the central animal facility at the Medical University of Innsbruck (Austria) under specific pathogen-free conditions. All animal experiments were performed in accordance with national and European guidelines and were reviewed and authorized by the committee on animal experiments (Federal Ministry of Science, Research and Economy-BMBWF-66.011/0035-WF/V/3b/2018). At the age of 10-weeks, male mice were randomly assigned either to the control group or the treatment group. The control group received a standard chow diet (n = 3; 180 mg of iron/kg) (Sniff, Soest, Germany), the treatment group received a diet supplemented with 25 g/kg of carbonyl iron (n = 5) (Altromin, Lage, Germany), for 14 days, respectively. At days 0, 7 and 14, blood was collected by means of VAMS (further description is included in the appropriate section). At day 14, animals were sacrificed, plasma heparin and liver tissue were collected for metabolomic analyses (description in the appropriate section).
2.3. Blood VAMS
MITRA VAMS devices were purchased from Neoteryx (Torrance, CA, USA). One drop of mouse blood was absorbed onto VAMS devices at three different time points and processed as described in the appropriate section.
2.4. Blood, Plasma and Liver Metabolite Extraction
Whole blood was collected by means of VAMS and metabolites were extracted as previously described [17
]. Briefly, VAMS tips were dried for 2 h and stored at −80 °C until metabolomics analysis. Next, VAMS were inserted into Eppendorf tubes and extracted with 200 µL of extraction solution containing the following internal standards: Alanine 13
N (0.9 μg/mL), arginine 13
(1.8 μg/mL), aspartic acid 13
N (1.3 μg/mL), cystine 13
(1.2 μg/mL), glutamic acid 13
N (1.5 μg/mL), glycine 13
N (0.8 μg/mL), histidine 13
(1.6 μg/mL), isoleucine 13
N (1.3 μg/mL), leucine 13
N (1.3 μg/mL), lysine 13
(1.5 μg/mL), methionine 13
N (1.5 μg/mL), phenylalanine 13
N (1.7 μg/mL), proline 13
N (1.2 μg/mL), serine 13
N (1.1 μg/mL), threonine 13
N (1.2 μg/ mL), tyrosine 13
N (1.8 μg/mL), and valine 13
N (1.2 μg/mL).
Samples were sonicated for 15 min and then vortexed at 1200 rpm for 60 min, both at 20 °C. VAMS tips were removed, and centrifuged at 1800× g for 10 min at 20 °C. To remove proteins, a protein removal plate (Sirocco Waters Corporation, Milford, MA, USA) was used, where the supernatant was filtered by using a positive pressure-96 processor applying 12 psi pressure (Waters Corporation, Milford, MA, USA). The extract was evaporated to dryness under vacuum at 35 °C for 120 min in a vacuum evaporator (EZ-2 vacuum evaporator, Genevac, Ipswich, UK).
To obtain plasma, whole blood was collected in heparin tubes, and subsequently centrifuged at 8000 rpm for 8 min at 4 °C. Plasma samples were stored at −80 °C until the metabolomic analysis. Metabolites were obtained adding 200 μL of extraction solution to 50 μL of plasma. The extraction solution consisted of acetonitrile (100%) and internal standards in the concentrations described above. Samples were then filtered through a protein removal plate (Sirocco, Waters Corporation, Milford, MA, USA) and the extract was evaporated to dryness under vacuum at 35 °C for 120 min in a vacuum evaporator (EZ-2 vacuum evaporator, Genevac, Ipswich, UK).
Around 30–50 mg of frozen liver samples were homogenized in ice-cold methanol (100%). 50 µL of this homogenate was added to 200 μL of extraction solution, proteins were removed as described above and the extracted metabolites were evaporated to dryness.
Finally, all samples were reconstituted with 150 μL of acetonitrile/water (50:50, v/v) solution.
Quality control (QC) samples were obtained were obtained by pooling together small aliquots (15 µL) from each sample.
2.5. Ultra-High Performance Liquid Chromatography (UHPLC) Combined with MASS Spectrometry (MS)
All samples were analyzed using the metabolomics workflow previously described [17
]. Briefly, ultra high-performance liquid chromatography (UHPLC) (Agilent 1290; Agilent Technologies, Santa Clara, CA, USA) was coupled to a Q-TOF mass spectrometer (TripleTOF 5600+; AB Sciex, Foster City, CA, USA). The chromatographic separation was achieved by hydrophilic interaction liquid chromatography (HILIC) using an Acquity BEH amide, 100 × 2.1 mm column (Waters Corporation, Milford, MA, USA).
Acetonitrile + 0.1% formic acid was used as mobile phase A and water + 0.1% formic acid as mobile phase B. The injection volume was set at 5 μL, and the flow rate at 0.6 mL/min. The following linear gradients were used: 0 min 95% A and 1 min 95% A, 4 min 30% A and 5 min 30% A, and 5.1 min 95% A and 8 min 95% A.
The mass spectrometer operated in full scan mode in the mass range from 50 to 1000 m/z and an accumulation time of 250 ms. In ESI+ mode, the source temperature was set at 700 °C, the declustering potential at 30 V, the collision energy at 6 V, the ion spray voltage at 5120 V, the curtain gas at 25 psi, and the ion source gases 1 and 2 at 60 psi. In ESI-mode, the source temperature was set at 650 °C, the declustering potential at −45 V, the collision energy at −6 V, the ion spray voltage at −3800 V, the curtain gas at 25 psi, and the ion source gases 1 and 2 at 30 psi. The instrument was mass calibrated by automatic calibration infusing the Sciex Positive Calibration Solution (part no. 4460131, AB Sciex, Foster City, CA, USA) for positive mode and Sciex Negative Calibration Solution (part no. 4460134, AB Sciex, Forster City, CA, USA) for negative mode, after every two sample injections. Samples were then analyzed in a randomized order, and pooled QC samples were injected every eight samples.
2.6. Data Analysis
Data were first converted to mzML, using ProteoWizard MS Convert [27
] and then processed in R (v 3.3.3) using the XCMS package [28
]. Pooled QCs and total ion current (TIC) were used to check the intra- and inter-batch variability, as applicable. We applied an acceptance of 30%. Therefore, features/metabolites with RSD higher than 30% in pooled QCs were removed and not considered further in data analysis. Failed injections were checked by monitoring the intensity of the spiked labeled amino acid standards.
Metabolites were identified by verifying retention time, accurate mass, and tandem mass spectrometry against our in-house and/or online databases, including the Human Metabolome Database (HMDB) [29
] and the metabolomics database (METLIN) [30
]. Among the annotated metabolites, we identified 2-hyroxybutyric acid, however we cannot exclude the presence of the isomer 3-hyroxybutyric acid, which might not be separated by our chromatographic system. From now on, we will refer to this metabolite as 2-/3-hyroxybutyric acid. Tentatively identified metabolites were integrated using MultiQuant 3.0 (Sciex, Foster City, CA, USA). Extracted ion chromatograms (EICs) were extracted using a 0.02 mDa window centered on the expected m
for each targeted compound.
Data were normalized by the sum of the features and then log transformed and scaled by using Pareto scaling before principal component analysis (PCA) using MetaboAnalyst (version 4.0) [31
This is the first study that combines the use of VAMS technology to a metabolomics workflow for the investigation of the effects of dietary iron overload on metabolic profiles. Moreover, this study provides new information on the significance of metabolite levels found in plasma and whole blood related to metabolic changes observed in organs, specifically in the liver.
The latter is the central organ for several physiological and metabolic processes, including iron homeostasis. To that, despite the knowledge gained so far, the overall mechanisms leading to patient’s subclinical problems related to insufficient energy metabolism are still insufficiently understood [1
]. For this reason, exploring the effects of iron homeostasis imbalances and the underlying metabolic rearrangements remain a public health concern. In particular, a better understanding of such processes would be of great importance for patient handling, refinement of genetic and transfusional iron loading, and to monitor intravenous iron therapy in order to prevent tissue damage or systemic problems caused by spatial-temporal accumulation of iron [1
]. In this context, the possibility to use VAMS devices would offer the advantage to easily screen subjects at risk, and subsequently follow the affected ones.
In accordance with previous reports from human studies [43
], in this study we found that iron overload caused a metabolic remodeling that primarily involves a compensatory response to increased oxidative stress. Here we also found, that in response to iron overload, concentrations of blood antioxidant metabolites, such as methionine sulfoxide, glutathione, cysteine, were increased. Under such metabolic stress conditions, 2-/3-hydroxybutyric acid, which is a byproduct of cystathionine cleavage to cysteine in the methionine-to-glutathione pathway, can be released [45
]. Indeed, 2-/3-hydroxybutyric acid was found to be associate with excess glutathione demand and disrupted mitochondrial energy mechanisms, occurring during persistent oxidative stress [35
In the current study, dietary iron overload increased liver 2-/3-hydroxybutyric acid synthesis and its excretion into blood and plasma, indicating an augmented hepatic glutathione synthesis, necessary to cope with the ongoing oxidative stress.
Moreover, besides bearing detoxification functions, this metabolite has been also reported as an early biomarker for insulin resistance and glucose intolerance in a non-diabetic population [35
Interestingly, animals receiving the high iron supplement showed higher glucose levels. This is in agreement with the literature that reports an association between hyperferritenemia and glucose [44
]. Nevertheless, the exact mechanisms that link iron to diabetes development still remain to be elucidated.
With regard to that, our experiments suggest two possible explanations. Firstly, iron loading might act as an activator of the Krebs cycle which is fueled by other substrates rather than glucose; in this way the number of molecules of glucose needed for energy production are reduced and also the ones, eventually, used for lactate production [4
Secondly, iron loading affects carnitine levels. Carnitines are small molecules that carry out a plethora of functions. Besides being involved in the transport of long-chain fatty acids from the cytosol into the mitochondria, carnitines play an important role in energy production, through the coordination of the activity of several enzymes of the Krebs cycle, beta oxidation, urea cycle and gluconeogenesis [40
]. If their balance is maintained correctly, carnitines improve glucose tolerance, on the contrary low levels of carnitines are associated with diabetic complications [40
Accordingly, high iron overload led to carnitines consumption as shown by the low levels of carnitines detected both in the periphery and in the liver. In this setting, carnitines may be employed both as antioxidant molecules and seen as a metabolic response to systemic glucose imbalance [40
Altogether, these results support the association between iron loading and altered glucose homeostasis already described in humans [36
]. Moreover, in accordance to what we have previously shown [9
], dietary iron overload caused not only mitochondrial impairment, but also changes to metabolic remodeling, that primarily implies compensatory mechanisms to oxidative stress.
Some of the presented metabolites changing in blood during iron overload, such as glucose, 2-/3-hydroxybutyric, malic acid, aspartic acid, lactic acid and cysteine might represent potential biomarkers for iron imbalances and tissue iron loading. Nevertheless, to confirm these findings further in vivo studies with larger sample sizes are necessary.
Finally, although plasma is usually preferred for diagnostic purposes, this work demonstrates the advantage of using VAMS technology, considering the minimal amount of blood required and the avoidance of additional pre-analytical procedures to obtain plasma. Moreover, considering that for animal research, plasma is often the limiting step for exploratory experiments, we strongly recommend the use of such a device. Our data not only uncovers regulatory metabolic circuits affected by dietary iron loading but also suggests that determination of metabolites in the periphery provide a suitable reflection of comparable metabolic changes in the liver.
In conclusion, metabolomics offers an advantageous platform to monitor metabolic changes occurring during iron overload. In particular, the use of VAMS technology represents a reliable and useful tool for blood sampling in animal research, in accordance with the principles of reduction and refinement (two of the 3Rs principles). Specifically, VAMS technology allows multiple sampling, reducing the number of animals needed for the study on one side, as well as easy blood collection, diminishing the pain and distress of the animals.
Moreover, the combination of VAMS with MS-based untargeted metabolomics facilitates the collection of time-course metabolomics data, allowing to study changes in the metabolic signature over the time during a perturbation.
In the current study, we found a specific blood signature after iron supplementation that involves changes in the systemic redox state. In addition, iron overload affects glucose homeostasis, suggesting ongoing metabolic remodeling, which may be linked to the mitochondrial dysfunction we previously reported [9
], and supporting the association between iron loading and dysregulated glucose homeostasis reported in humans [36
For this reason, further investigations of the metabolic signature of subjects suffering from primary or secondary iron overload will provide new insights into the association between iron imbalance and glucose homeostasis.