Since the discovery of hepcidin there has been great interest in this master regulator of iron metabolism . For the last seven years the scientific community has awaited a robust hepcidin assay. To date the two main approaches for measuring hepcidin in biological samples have been ELISA and SELDI. Two ELISA approaches exist, one measuring pro-hepcidin (which has proved controversial [11, 23–25]) and a recently described ELISA for hepcidin measurement in serum . This latter assay is based upon competition for antibody binding between endogenous hepcidin and added biotinylated hepcidin. Although further validation is required to ensure that this assay is solely specific for bioactive hepcidin 25, it likely represents a promising high throughput approach. In contrast SELDI has been widely used to measure urinary and serum hepcidin [5, 9–12]. The advantages of SELDI include the ability to measure different forms of hepcidin, e.g. hepcidin 20 and 25, and furthermore the assay can be conducted under denaturing conditions so that protein-protein interactions should not interfere. Unfortunately, due to the lack of good internal standards SELDI based hepcidin measurements are at best only semi-quantitative. This issue has now been partially resolved by Swinkels and co-workers utilising hepcidin 24 as an internal standard . The use of a stable isotope labelled hepcidin as an internal has so far been limited to LC-ESI-MS experiments . We have now combined the use of stable isotope labelled hepcidin and SELDI-TOF-MS to generate a technically simple high-throughput quantitative hepcidin assay.
The use of stable isotope labelled peptides to make mass spectrometry based proteomic experiments quantitative is well established and widely used as exemplified by the SILAC and ICAT methods [26, 27]. Stable isotope labelling introduces additional neutrons altering the mass of a peptide but does not alter the electronic structure and hence the binding and ionisation/detection during SELDI should be unaffected. In this study we apply this approach to the measurement of hepcidin. We synthesise a stable isotope labelled hepcidin that is 10 Da heavier than endogenous hepcidin. After checking that the synthetic hepcidin adopts the correct folded structure using FTICR-MS, NMR and a bio-assay we validate its use as an internal standard in SELDI experiments and demonstrate the merits of this approach.
The isotopic envelope of the labelled hepcidin does not overlap with the isotopic envelope of endogenous hepcidin and even when using the PBSIIc instrument (a sensitive but low resolution mass spectrometer) there is minimal overlap between the labelled and endogenous hepcidin peaks. A small amount of overlap with oxidised endogenous hepcidin does occur and it is therefore important to minimise sample (and standard) oxidation. Oxidation of hepcidin has previously been reported as an ex vivo artefact  and is usually minimal in serum but does occur in urine over time. Hepcidin oxidation can largely be prevented by rapid sample processing and avoiding long periods at room temperature. It should be noted that, regardless of internal standards, total hepcidin levels cannot be determined in oxidised samples from the sum of the non-oxidised and oxidised peak heights as the relative ionisation efficiencies are unknown. Additional precautions when using SELDI with or without spiking are to collect and average a large number of laser shots and to use appropriate laser power to ensure substantial, but not saturating peak intensities. The experiments presented here show that, when these conditions are met, the peak height ratio approach offers a quantitative assay for hepcidin.
We find a good correlation between SELDI peak heights and the peak height ratio method adding weight to previous reports inferring changes in hepcidin concentration from SELDI peak heights [5, 9–12]. In addition we now report a small but significant increase in systemic hepcidin in breast cancer patients. We find the average hepcidin concentration in healthy females to be ~50 ng/ml similar to the 65 ng/ml that Ganz et al determined by ELISA .
The use of the stable isotope labelled hepcidin alone does not increase the sensitivity or precision of SELDI measurements as demonstrated by intra- and inter-assay CVs similar to those using TIC normalised peak height alone. The major advance is that the labelled hepcidin approach should allow hepcidin levels to be measured in absolute concentration units independent of instrument/operator or background proteome variations. In the absence of labelled hepcidin spiking, conversion of peak heights into concentrations requires external calibration using a series of dilutions of hepcidin. This is not ideal as at high dilutions hepcidin either adheres to plasticware, precipitates or aggregates and mass spectrometry detection becomes variable (Ward et al, unpublished observations). This problem can be largely overcome by storing a concentrated labelled hepcidin solution in aliquots at -80°C, doing a single dilution in 8 M urea/1% CHAPS and spiking directly into samples. Addition of stable isotope labelled hepcidin to serum prior to sample work-up for mass spectrometry will enable multi-step hepcidin enrichment (followed by SELDI or MALDI) which will ultimately improve the sensitivity of the assay over the single step retentate chromatography of SELDI.
We conclude that by spiking stable isotope labelled hepcidin into clinical samples it is possible to turn high-throughput SELDI analyses into robust hepcidin assays. SELDI measurements without spiking reflect hepcidin concentrations but addition of stable isotope labelled hepcidin improves confidence in the data and provides absolute concentrations facilitating inter-study and inter-laboratory hepcidin comparisons.