Binding of hepcidin to plasma proteins.

The hepcidin hormone of 25 amino acid residues (hepcidin25) is a key regulator of iron homeostasis (1 ). Hepcidin has been shown to bind in vitro to 2macroglobulin ( 2M) and albumin in human plasma (2 ). It is not known, however, to what extent hepcidin is bound to proteins in vivo. For purposes of clinical interpretation, it is critical to know whether protein-bound or free hepcidin is being quantified. To characterize the binding of hepcidin to plasma proteins, we used to gel filtration to fractionate 0.5 mL serum samples from both healthy individuals and patients undergoing hematopoietic stem cell transplantation. We similarly fractionated mixtures of hepcidin and 2M or albumin [1 g/L in 10 mmol/L potassium phosphate buffer, pH 7.4, containing 150 mmol/L NaCl (PBS)]. We used a SuperdexTM 200 10/300 GL column (GE Healthcare Biosciences) equilibrated with PBS at a flow rate 0.5 mL/min. The column was calibrated with synthetic hepcidin-25 (Peptide Institute), human albumin, and 2M. We collected 0.5-mL fractions and measured hepcidin in the fractions by HPLC–tandem mass spectrometry (LC-MS/MS) as previously described (3 ). We observed a single peak, corresponding to free hepcidin, eluting at 21–25 mL. In patient sera containing 36.6 nmol/L (102 g/L) and 50.4 nmol/L (141 g/L) endogenous hepcidin, LC-MS/MS analysis revealed very small hepcidin peaks in fractions containing 2M and albumin, respectively (Table 1), but the hepcidin concentration was below the limit of quantification (3 ). The hepcidin peak area corresponded to 0.7% and 0.9% of the main hepcidin peak, respectively, in the 2M fractions and to 0.9% and 1.3% of the main hepcidin peak, respectively, in the albumin fractions. Thus, approximately 98% of hepcidin was free in samples with increased hepcidin. This result contrasts with the findings of Peslova et al., who calculated that only 11% of hepcidin in the circulation was free (2 ). The molar ratio of 2M to hepcidin varied from 0.2:1 to 5:1 in their in vitro study, whereas the mean molar ratio of 2M to hepcidin in the serum of healthy individuals is 620:1. The much higher proportion of hepcidin relative to the 2M concentration that these investigators used can explain the difference between our results and those of Peslova et al. Hepcidin was not detectable in the 2M and albumin fractions of healthy individuals. Hepcidin and the human internal standard ([C18, N3]hepcidin25; Peptide Institute) elute at a retention time of 4.8 min in the LC-MS/MS assay. Most serum samples, however, show an additional minor peak with the same MS/MS transition as hepcidin and the internal standard; it also elutes at 5.1 min. In patient samples with 36.6 nmol/L and 50.4 nmol/L of hepcidin, the portion of hepcidin eluting at 5.1 min was 0.2% and 0.4%, respectively, of the main peak. The proportion of internal standard eluting at 5.1 min was 2.7% and 3.0%, respectively, of the internal standard eluting at 4.8 min. In the albumin fraction of these samples, 66% and 70%, respectively, of the hepcidin eluted at 5.1 min (see Fig. 1 in the Data Supplement that accompanies the online version of this letter at http://www.clinchem.org/content/ vol58/issue7). Of the internal standard added to the albumin fraction, 7.8% and 3.4%, respectively, eluted at 5.1 min (Table 1). No hepcidin peak was observed at 5.1 min in the 2M or hepcidin fractions. A possible explanation for the late-eluting hepcidin peak is that hepcidin binds to albumin coeluting with hepcidin during sample extraction (3 ). To confirm that extracted serum samples contained albumin, we fractionated 10 L of 1 Nonstandard abbreviations: hepcidin-25, hepcidin of 25 amino acid residues; 2M, 2-macroglobulin; LC-MS/MS, HPLC–tandem mass spectrometry. Table 1. Peak areas of the main peaks (tret 4.8 min) and late-eluting peaks (tret 5.1 min) for hepcidin and the internal standard (ISTD) in the LC-MS/MS chromatograms of patient samples containing 36.6 nmol/L and 50.4 nmol/L endogenous hepcidin, and of gel-filtration chromatography fractions of the same samples.