Production of biologically active forms of recombinant hepcidin, the iron-regulatory hormone Bruno Gagliardo 1 , Audrey Faye 2,3 , Maryse Jaouen 1 , Jean-Christophe Deschemin 2,3 , Franc¸ois Canonne-Hergaux 4 , Sophie Vaulont 2,3 and Marie-Agne ` s Sari 1 1 CNRS (UMR 8601), Universite ´ Paris Descartes, France 2 Institut Cochin CNRS (UMR 8104), Universite ´ Paris Descartes, France 3 Inserm (U567), Universite ´ Paris Descartes, France 4 CNRS (ICSN), Gif-Sur-Yvette, France Hepcidin is a 25 amino acid, cysteine-rich peptide, first identified in human blood [1] and urine [2] as an anti- microbial peptide of the defensin family. Human hep- cidin is the product of the HAMP gene, which encodes an 84 amino acid precursor protein [3,4]. In addition, hepcidin genes have also been identified in several species, including the mouse [5], rat [6], pig [7], dog [8] and fish [9–11]. In mammals, the hepcidin gene is expressed predominantly in hepatocytes and constitutes the master regulator of iron homeostasis. Hepcidin gene expression is positively regulated by iron and inflammation and negatively regulated by anemia and hypoxia [5,12]. There is compelling evidence that dysregulation of hepcidin underlies many iron disorders. Most of the iron overload syndromes (primary hemo- chromatosis and secondary iron overloads) imply a reduction of hepcidin secretion whereas, in contrast, overexpression of hepcidin appears to play a deter- mining role in anemia of inflammation or inflammation of chronic disease [13] and iron-refractory iron deficiency anemia [14]. HAMP and Hepc1 encode precursors of 84 and 83 amino acids, respectively, which contain typical 24 and 23 amino acid endoplasmic reticulum targeting signal motifs at the N-terminus of the human and mouse precursors, respectively. This signal is cleaved to produce prohepcidin, harboring a 35 amino acid proregion and the C-terminus 25 amino acid mature Keywords antimicrobial peptide; Escherichia coli expression; ferroportin; hepcidin; iron homeostasis Correspondence M A. Sari, UMR 8601, Universite ´ Paris Descartes, 45 rue des Saints-Pe ` res, 75006 Paris, France Fax: +33 1428 68387 Tel: +33 1428 62142 E-mail: marie-agnes.sari@univ-paris5.fr (Received 1 April 2008, revised 22 May 2008, accepted 28 May 2008) doi:10.1111/j.1742-4658.2008.06525.x Hepcidin is a liver produced cysteine-rich peptide hormone that acts as the central regulator of body iron metabolism. Hepcidin is synthesized under the form of a precursor, prohepcidin, which is processed to produce the biologically active mature 25 amino acid peptide. This peptide is secreted and acts by controlling the concentration of the membrane iron exporter ferroportin on intestinal enterocytes and macrophages. Hepcidin binds to ferroportin, inducing its internalization and degradation, thus regulating the export of iron from cells to plasma. The aim of the present study was to develop a novel method to produce human and mouse recombinant hep- cidins, and to compare their biological activity towards their natural recep- tor ferroportin. Hepcidins were expressed in Escherichia coli as thioredoxin fusion proteins. The corresponding peptides, purified after cleavage from thioredoxin, were properly folded and contained the expected four-disulfide bridges without the need of any renaturation or oxidation steps. Human and mouse hepcidins were found to be biologically active, promoting ferro- portin degradation in macrophages. Importantly, biologically inactive aggregated forms of hepcidin were observed depending on purification and storage conditions, but such forms were unrelated to disulfide bridge formation. Abbreviations huhepc, human hepcidin encoded by HAMP gene with an extra N-terminal methionine; m1hepc, mouse hepcidin encoded by Hepc1 gene with an extra N-terminal methionine; MES, 2(N-morpholino)ethanesulfonic acid; SELDI-TOF, surface enhanced laser desorption ionization time of flight; TFA, trifluroacetic acid; TRX, thioredoxin. FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 3793 peptide. This prohepcidin is further processed through a furin-type propeptide cleavage site to generate the secreted 25 amino acid hepcidin [2,15]. The 25 amino acid peptide forms a hairpin loop stabilized by four disulfide bonds, one of which is an unusual vicinal bond between adjacent cysteines at the hairpin turn [16]. A recent structure ⁄ function study revealed that the N-terminus of hepcidin is essential for its bioactivity, and that the structure is otherwise permissive for changes [17]. Once in the circulation, the peptide acts to limit gastrointestinal iron absorption and serum iron levels by inhibiting dietary intestinal iron absorption and iron recycling by the macrophages [18]. This is achieved by hepcidin binding to ferroportin, the trans- membrane iron transporter necessary for iron transfer out of intestinal epithelial cells and macrophages [19], resulting in its internalization and subsequent degrada- tion [20,21]. Although the functional role of hepcidin in iron metabolism has been well documented, little is known concerning its cell biology. Chemical synthesis of human hepcidin has been successful but difficult to achieve [22–24] due to folding restrictions imposed by the four-disulfide bridges [10,16]. Other studies have reported the purification of human recombinant hepci- din in the form of fusion proteins [25–28] and, until very recently [29], none of the recombinant hepcidins were shown to be bioactive in iron metabolism. In the present study, we describe an efficient proce- dure for purification of biologically active recombinant hepcidins. This constitutes an important step that will allow for a better understanding of hepcidin biology, which is a prerequisite for the use of hepcidin in medi- cal applications. Results Strategies for hepcidin expression and nomenclature We were able to produce properly folded recombinant hepcidins with the correct four-disulfide bridges with- out the need of a denaturation ⁄ renaturation step. A Novagen pET32-LIC vector expression system was used into which the hepcidin encoding sequence was cloned downstream of a thioredoxin (TRX) sequence to produce a fusion protein under the control of a T7 promoter. TRX was chosen as the fusion protein because it is involved in disulfide bridge formation [30,31]. A double tag system links TRX and hepcidin: a his-tag upstream of a thrombin cleavage site and a S-tag upstream of an enterokinase cleavage site (Fig. 1). Noteworthy, the recombinant hepcidin produced after cleavage exhibited an extra methionine at the N-termi- nus compared to the native peptide, as a result of the nature of the enterokinase cleavage site (Table 1). The major expected characteristics of the cleaved pro- teins ⁄ peptides are presented in Fig. 1, together with the nomenclature used in the present study. Origami B cells were chosen as a host strain because they carry mutations both in TRX and glutathione reductase genes, hence increasing cytosolic formation of disulfide bridges compared to those obtained in the highly reductive environment of a regular Escherichia coli strain. This strain choice was indeed important because mouse hepcidin encoded by Hepc1 gene with an extra N-terminal methionine (m1hepc) purified from a BL21(DE3) strain was found mostly as an insoluble multimer peptide (data not shown). Fig. 1. Scheme of hepcidin production and their expected characteristics. The name of each construct used through the study is given next to its corresponding cartoon, together with the amino acid numbers (AA), isoelectric point and the theoretical average and monoisotopic (in parenthesis) mass obtained using Protein Parameter software (ExPasy; http://www.expasy.org/cgi-bin/protparam). Yields are expressed in % (nmol of purified protein X100 per nmol of TRX-hepc) and are representative of three independent preparations. Active recombinant hepcidin B. Gagliardo et al. 3794 FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS Plasmid construction Cloning of m1hepc and human hepcidin encoded by HAMP gene with an extra N-terminal methionine (huhepc) was performed using synthetic oligonucleo- tides. Complementary hepcidin oligonucleotides were annealed and directly cloned into the pET32LIC vector (Table 1). Upon transformation and amplification in E. coli TG1 cells, pET32LIC-m1hepc and pET32LIC- huhepc vectors were purified, checked by sequencing and used to transform Origami B cells. Fusion protein expression TRX-hepcidin 20 kDa fusion proteins (TRX-m1hepc and TRX-huhepc) were produced under the same conditions. Therefore, only TRX-huhepc production is described here. Transformed pET32LIC-huhepc Origami cells were cultured in high density ZYM-5052 medium at 25 °C, as described previously [32], except that the autoinduction lactose step was replaced by a ‘classical’ isopropyl thio-b-d-galactoside induction step after 30 h of growth (see Experimental procedures). The decrease of temperature from 37 °Cto25°C was found necessary to avoid the formation of inclusion bodies because TRX-huhepc found in these structures was highly aggregated and very difficult to recover. Human hepcidin purification Hepcidin purification was carried out in three steps: (a) TRX-huhepc purification; (b) thrombin cleavage of TRX-huhepc to generate S-huhepc; and (c) S-huhepc enterokinase digestion to generate huhepc. For unknown reasons, TRX-huhepc (as well as TRX- m1hepc) appeared to be resistant to enterokinase digestion, avoiding a direct purification of huhepc from TRX-huhepc. The purification was followed by SDS ⁄ PAGE, for proteins or peptides, and also Experion TM capillary electrophoresis for proteins. Step1: purification of TRX-huhepc After 30 h of growth, 3 h of induction and lysis, 8.5 g of soluble proteins were obtained; 8% of which com- prised TRX-huhepc (approximately 700 mg) (Fig. 2, lane 2). Some TRX-huhepc remained in the insoluble fraction (Fig. 2, lane 1) and could not be recovered, even in the presence of urea or guanidine (data not shown). Lowering the temperature to 15 °C did not enhance the yield of soluble fusion protein. To enrich the fusion protein in the soluble fraction, a 65 °C heat- ing step was performed (Fig. 2, lane 3). This step allowed most of the unwanted proteins to be irrevers- ibly precipitated because 500 mg of soluble proteins were present in the heated supernatant; 70% of which comprised TRX-huhepc (350 mg). Even though 50% of the fusion protein was lost in this process, this step, which takes advantage of the heat resistance of TRX [33], was found to be necessary to obtain pure TRX- huhepc. The TRX-huhepc enriched solution was purified to homogeneity using affinity TALONÔ Co 2+ chroma- tography, with the remaining unwanted proteins flowing through the column (Fig. 2, lane 4) and TRX-huhepc being eluted with 150 mm imidazole followed by 2(N-morpholino)ethanesulfonic acid (MES) buffer (pH 5) (Fig. 2, lane 5), and found to be Table 1. Primers used for hepcidin cloning and PCR, and primary structure of the purified peptides and proteins. Forward and reverse oligonucleotides encoding hepcidin sequence contained 5¢ and 3¢ additional nucleotides (in bold) corresponding to the LIC sequence of pET-32Ek ⁄ LIC vector. Sequence (5¢-to3¢) Primers Mouse hepcidin forward GACGACGACAAGATGGACACCAACTTCCCCATCTGCATCTTCTGCTG TAAATGCTGTAACAATTCCCAGTGTGGTATCTGTTGCAAAACATAA Mouse hepcidin reverse GAGGAGAAGCCCGGTTATGTTTTGCAACAGATACCACACTGGGAATT GTTACAGCATTTACAGCAGAAGATGCAGATGGGGAAGTTGGTGTCCA Human hepcidin forward GACGACGACAAGATGGACACCCACTTCCCGATCTGCATTTTCTGCTG CGGCTGCTGTCATCGATCAAAGTGTGGGATGTGCTGCAAGACGTAA Human hepcidin reverse GAGGAGAAGCCCGGTTACGTCTTGCAGCACATCCCACACTTTGATCG ATGACAGCAGCCGCAGCAGAAAATGCAGATCGGGAAGTGGGTGTCCA Proteins m1hepc MDTNFPICIFCCKCCNNSQCGICCKT huhepc MDTHFPICIFCCGCCHRSKCGMCCKT S-m1hepc GSGMETAAKPERQHMDSPDLGTDDDDKMDTNFPICIFCCKCCNNSQCGICCKT S-huhepc GSGMETAAKPERQHMDSPDLGTDDDDKMDTHFPICIFCCGCCHRSKCGMCCKT B. Gagliardo et al. Active recombinant hepcidin FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 3795 at least 90% pure as measured on the ExperionÔ device. Routinely, 3.5 mg of fusion protein was obtained from 1 g of cell pellet. Step 2: thrombin cleavage After elution, TRX-huhepc was digested for 16 h with 1 U of thrombin per 10 mg of protein, yielding an almost complete cleavage (Fig. 2, lane 6). S-huhepc generated by thrombin digestion was separated away from TRX using an HPLC purification step on a C18 semi-preparative column. The purified S-huhepc (Fig. 2, lane 7) was lyophilized to dryness, resuspended at 1 mgÆmL )1 in enterokinase cleavage buffer and treated for 6 h with 10 U of enterokinase per mg of S-huhepc. Low temperature was avoided to minimize peptide precipitation during cleavage. Finally, purifica- tion of huhepc was achieved by anion exchange chro- matography. As previously noticed for S-huhepc, huhepc had a propensity to aggregate but remained soluble in the presence of salts. Thus, the enterokinase cleaved mixture was simply diluted in one volume of water to adjust the NaCl concentration to below 40 mm. Under these conditions, low salt buffer at pH 7, huhepc flowed through the UnoQ anion exchange column (Fig. 2, lane 8), whereas the other peptides (S and eventually S-huhepc) bound to the column. Again, to avoid irreversible aggregation of huhepc, the huhepc solution was directly lyophilized in the presence of buffer to maintain further solubility of the peptide. The peptide was stored as a dry lyophilized powder until used. When required, remaining salts were withdrawn either upon thorough dialysis (using 1000 Da molecular weight cut-off dialysis tubing) or HPLC chromatography (Fig. 3) or size exclusion chromatography using bio-gelÔ P4 resin (Bio-Rad, Marnes-la-Coquette, France). Routinely, 2 mg of pure huhepc was obtained from 60 g of Origami cells. Mouse hepcidin purification Mouse hepcidin (m1hepc) was obtained as described for huhepc, except for minor modifications related to differences in the pI of the peptide (Fig. 1). Entero- kinase digestion was carried out at pH 6.2 (instead of pH 7 for m1hepc) and the UnoQ chromatography step was run at pH 6.2 rather than pH 7. The production yield of m1hepc was slightly lower than that of huhepc (8% instead of 13%, respectively, starting form Fig. 3. HPLC profile of pure recombinant (A) mouse or (B) human hepcidins. (A) Fifty microlitres of huhepc (0.1 mgÆmL )1 ) was injected on a MOS C8 and (B) 50 lL of m1hepc (0.01 mgÆmL )1 ) was injected on a Vydac C18 column using a stepwise acetonitrile gradient. Fig. 2. Purification of huhepc followed by SDS ⁄ PAGE using 12% Bis-Tris CriterionÔ XT gels in the absence of reducing agent. Large (left) and small (right) molecular weight markers (Bio-Rad) are indi- cated in kDa. Lane 1, 200 lg of crude extract of TRX-huhepc expressing cells; lane 2, 150 lg of TRX-huhepc expressing cells cytosolic fraction (prior heating step); lane 3, 50 lg cleared cyto- solic fraction after removal of the heat denaturated proteins; lane 4, 30 lgofCo 2+ TalonÔ unbound protein fraction; lane 5, 50 lgof Co 2+ TalonÔ 150 mM imidazole eluted fraction, containing TRX-hu- hepc; lane 6, 40 lg of TRX-huhepc thrombin cleavage mixture; lane 7, 6 lg of HPLC purified S-huhepc; lane 8, 3 lg of UnoQ purified huhepc (lyophilized in the presence of buffer). *TRX-huhepc dimer formed upon disulfide bridge between two thioredoxin subunits. Active recombinant hepcidin B. Gagliardo et al. 3796 FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS TRX-hepc), most likely due to the higher propensity of m1hepc to precipitate during the enterokinase cleavage step. Characterization of the recombinant proteins and peptides Purity of the recombinant proteins was assessed by SDS ⁄ PAGE or ExperionÔ capillary electrophoresis, HPLC chromatography and MS. As shown in Fig. 2, the peptides and proteins appeared fairly clean by SDS ⁄ PAGE analysis (in the absence of reducing agent) and, importantly, ran at the expected molecular weights: TRX-huhepc at 20 kDa (Fig. 2, lane 5); S-hu- hepc at 6 kDa (Fig. 2, lane 7); and huhepc appears to run as a 3.5 kDa peptide (Fig. 2, lane 8). The extra sharp band marked with an asterisk, at approximately 37 kDa, corresponded to TRX-huhepc dimers due to disulfide bridges between two TRX domains. This band was indeed absent when the SDS ⁄ PAGE was run in the presence of reducing agents (data not shown). Purified hepcidin remained soluble and largely as a monomer after the anion exchange chromatography or when resuspended in the presence of buffer after lyophiliza- tion (Fig. 2, lane 8). C18 HPLC analysis also displayed a single peak for each peptide (Fig. 3, m1hepc and huhepc), when assessing the purity of the samples. Interestingly, desalted hepcidins, either obtained after HPLC purification or desalted after anion exchange chromatography following dialysis or exclu- sion chromatography, always aggregated over time. This phenomenon was observed with low salt solu- tions, regardless of whether the hepcidins were stored at 4 °Corat)80 °C and in the presence or absence of glycerol or detergents. These polymeric hepcidin forms could not be reverted to the monomer forms, even in the presence of reducing agents such as dithiothreitol or tris(2-carboxyethyl)phosphine (data not shown). Therefore, pure hepcidins were routinely lyophilized immediately after the anion exchange chromatography and kept as a salt containing powder. To quantify the hepcidin content in the lyophilized powder, a sample was injected on a Vydac (Grace, Templemars, France) C18 HPLC column, lyophilized again and weighed. Concentration was also confirmed using UV-visible measurements [2]. To verify that all the cysteines were engaged in disul- fide bridges, Ellman’s reagent was used to measure the content of free cysteines [34]. As indicated in Table 2, S-m1hepc and S-huhepc, m1hepc and huhepc contained less than one free cysteine per monomer, indicating that at least 90% of the hepcidins and hepcidin derivatives contained the expected four-disulfide bridges. Finally, the presence of the four-disulfide bridges was also confirmed using ESI MS (Table 2) for S-m1hepc and S-huhepc. Upon deconvulation, the molecular weight was 6020.6 ±1 Da for S-m1hepc and 6057.4 ± 1 Da for S-huhepc, with both values being in good agreement with the theoretical mass of 6020.82 Da and 6055.94 Da expected for the corre- sponding S-hepcidins containing four-disulfide bridges. Unfortunately, ESI MS on pure hepcidins was impos- sible due to the difficulty of ionizing the peptide. Therefore, surface enhanced laser desorption ionization time of flight (SELDI-TOF) MS, previously described as a powerful technique to analyze hepcidin, was used [35]. Clean mass spectra were observed when peptides were run on regular (NP20) protein chips (Table 2). The calculated mass was 2885.2 ± 1 Da for m1hepc and 2920.3 ± 1 Da for huhepc, with both values being in good agreement with the theoretical values of 2885.49 Da and 2920.60 Da, respectively, expected for fully oxidized disulfide bridged forms, demonstrating that the recombinant hepcidins were pure and con- tained four-disulfide bridges. Interestingly, SELDI- TOF MS performed on aggregated hepcidins (Fig. 4B) gave values identical to those of the non-aggregated form, indicating that the polymeric bands observed by SDS ⁄ PAGE analysis were not related to interchain disulfide bridges but simple aggregation. Table 2. Characterization of purified hepcidin derivatives. The calculated average molecular weight was obtained either using ESI MS or SELDI-TOF MS, as described in the Experimental procedures. ND, not determined; ) , no activity detected or no m ⁄ z signal obtained; +, activity measured: at least 10 mm diameter growth inhibition of E. coli in plate assay using 100 pmol of each peptide, complete disappear- ance of ferroportin signal at the membrane of iron treated macrophage J774 cells in the presence of 700 l M of each of the hepcidin deriva- tives. Free SH content is related to the lack of detection of free thiols using Ellman’s reagent. All data were obtained for two independent preparations of hepcidin (MS and SH contents) and three independent preparations for activity measurements. Name MW (ESI) MW (SELDI-TOF) Free SH (%) Antimicrobial activity Ferroportin activity S-m1hepc 6020.6 ± 1 6018.9 ± 0.5 < 5 + ) S-huhepc 6057.4 ± 1 ND < 5 + ) m1hepc – 2885.2 ± 0.5 < 3 + + huhepc – 2920.3 ± 0.5 < 3 + ++ B. Gagliardo et al. Active recombinant hepcidin FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 3797 Compared activities of hepcidins and S-hepcidins The antibacterial activity of m1hepc, huhepc and their S-tag forms (S-m1hepc and S-huhepc) was tested against E. coli using a plate assay [27]. In our condi- tions, 10 lLofa10lm solution for either of the four peptides routinely resulted in a 10 mm diameter growth inhibition of E. coli whereas 10 lL of buffer or medium resulted in no more than a 3 mm diameter growth inhibition (Table 2). These results demonstrate that recombinant hepcidins displayed obvious antibac- terial activity against E. coli and that similar activity was observed for the S-hepcidins compared to that of the corresponding hepcidins. The iron-related bioactivity of the recombinant pro- teins was tested in Fe-NTA treated J774 macrophages for their potential to degrade the hepcidin receptor, ferroportin. As shown by western blotting (Fig. 5), only iron treated macrophage cells express detectable amount of membrane bound ferroportin (Fig. 5, lane 1 versus lane 2). As a control, the human synthetic 25 amino acid mature peptide (Peptides International, Louisville, KY, USA) was used at 700 nm for 5 h. As expected, a complete disappearance of ferroportin was observed (Fig. 5, lane 3). When tested under the same conditions, the huhepc also provoked a complete dis- appearance of ferroportin (Fig. 5, lane 8). By contrast, the S-hepcidins were completely inactive towards ferro- portin degradation (Fig. 5, lanes 4 and 5). These results were obtained routinely when ‘monomer’ forms of hepcidin were used in the assay. Interestingly, when salt-free hepcidins were used (Fig. 5, lanes 6 and 7), no effect was observed on ferroportin degradation, sug- gesting that the salt-free hepcidin forms have lost their biological activity. The inactivation of hepcidin activity in salt free solu- tion was further investigated. A fraction of active UnoQ purified human hepcidin was desalted using a P4 exclusion chromatography, concentrated, and com- pared with its parent ‘salted’ hepcidin. The ‘salt-free’ hepcidin form was found completely inactive towards Fig. 4. (A) Analysis by SDS ⁄ PAGE of UnoQ purified huhepc before or after ‘desalting’ using a P4 exclusion column settled in water and concentrated by lyophilization. (B) SELDI-TOF spectrum of desalted hepcidin. This result is representative of three differ- ent experiments performed with external or internal calibration using low mass peptide standards (Sigma, St Louis, MO, USA). Fig. 5. Effects of hepcidin treatment on ferroportin expression in J774 cells. Expression of ferroportin was studied in macrophage J774 cells treated with Fe-NTA for 16 h and with the indicated hep- cidin or S-tagged hepcidin for 5 h. Membrane proteins (30 lg per lane) were separated by SDS ⁄ PAGE, electro-transferred onto nitro- cellulose and analyzed with anti-ferroportin serum. The ponceau red staining of the membrane is shown. Lane 1 is the control without Fe-NTA and lanes 2 to 8 correspond to samples treated with Fe-NTA. Lane 3, 700 n M huhepc from Peptides International; lanes 4 and 5, 100 and 700 n M of enterokinase digested mixture of S-m1hepc; lanes 6 and 7, 100 and 700 n M of pure ‘desalted’ huhepc; lane 8, UnoQ purified salt-containing huhepc (700 n M). Active recombinant hepcidin B. Gagliardo et al. 3798 FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS ferroportin. Interestingly, the migration profile by SDS ⁄ PAGE of this inactive form of hepcidin showed an important proportion of multimers (Fig. 4A). Fur- thermore, the polymerization pattern was identical in the absence or presence of a reducing agent (not shown). Altogether, these results clearly demonstrate that human hepcidin aggregates in the absence of salt, that this aggregation is not related to intermolecular disulfide bridge formation (Fig. 4B) and, more impor- tantly, that human hepcidin activity is significantly reduced upon aggregation. Recombinant m1hepc was also shown to degrade ferroportin, although less efficiently than huhepc. At 100 nm, although both huhepc and synthetic hepcidin (Peptides International) induced a complete disappear- ance of ferroportin from the macrophage membranes, m1hepc yielded only a 50% decrease of ferroportin under the same conditions. Finally, aggregation- induced loss of hepcidin activity was also observed with the m1hepc, to a larger extent than the human form, most likely due to its slightly more hydrophobic nature (human hepcidin contains two histidine and one arginine residues instead of three asparagine residues for the mouse form). Discussion The TRX fusion protein approach in association with the use of the Origami strain was found to be ideal for the expression of hepcidins. The recombinant hep- cidin peptides, containing eight cysteines engaged in four intramolecular disulfide bridges, were properly folded, fully oxidized and biologically active. Further- more, the need of denaturation–renaturation or reduction–oxidation steps (often present in other pro- tocols of hepcidin preparation, either recombinant [26–28] or synthetic [10,17,23]) were removed. By con- trast to another report [26], our recombinant products were apparently metal-free. The human and mouse1 hepcidins were purified by two cleavage steps: removal of TRX with thrombin and S-tagging with enterokinase. The recombinant products differed from the native hepcidins by the presence of an extra amino acid (methionine) at the N-terminus. However, this extra amino acid did not appear to affect the biological activities of the recombinant products because their activities were found comparable to that of a commercially available synthetic 25 amino acid human hepcidin (Peptides International). This is in agreement with observations of Nemeth et al. [17], who showed that a human hepcidin of 26 amino acids, bearing an extra alanine in the N-terminal posi- tion, retained its activity. Interestingly, S-hepcidins, which bear 29 extra amino acids upstream of the hepcidin sequence, con- serve an antibacterial activity comparable to that of the corresponding hepcidin but are completely ineffi- cient in degrading the iron exporter ferroportin. This result is in accordance with observations of Nemeth et al. [17], who demonstrated that the N-terminal part of hepcidin was necessary for interaction with ferro- portin, and strongly suggests that the presence of the S-tag prevents the hepcidin-ferroportin interaction. It is thus predictable that any N-terminal tagged recom- binant hepcidin products, such as GST [26], His [28] or other fusions, could interfere with hepcidin activ- ity. Very recently, Koliaraki et al. [29] described the production of a C-terminal his-tagged human hepci- din using Pichia pastoris as a host for expression. Although it was not possible to express untagged hepcidin, the C-terminal his-tagged recombinant pep- tide that they produced was able to bind ferroportin and promote its degradation in Raw 264.7 macro- phages [29]. The results of the present study also emphasize that the determination of hepcidin antimicrobial properties is not relevant to assessing hepcidin bioactivity in iron metabolism and that only a measurement of hepcidin activity towards ferroportin degradation and ⁄ of cellular iron retention should be considered. In the present study, we have demonstrated that aggregated forms of hepcidins (dimers or tetramers), harboring the same SELDI-TOF mass spectrum as their corresponding monomers, are inactive against ferroportin. As previously demonstrated [36], hepci- din derivatives, including the recently described C-terminal his-tagged peptide [29], are very likely to precipitate, yet the aggregation does not involve interchain disulfide bridge formation. This aggregation occurs at neutral pH in the absence of salt or upon storage at 4 °C in solution. This phenomenon is likely to explain the poor reproducibility in hepcidin preparation (including commercial preparations). In our hands, the storage of hepcidins as salted lyophi- lized aliquots at )20 °C (or less) was found to com- prise the best method for preventing loss of activity over time. Reproducible preparation of biologically active hepcidins, either synthetic or recombinant, is a necessary step for investigating the cellular biology of this hormone and the present study has contrib- uted to this aspect. Finally, the strategy described in the present study was also found to be appropriate for the production of prohepcidin (B. Gagliardo, personal communication) and may be useful in the preparation of other cysteine rich peptides or proteins. B. Gagliardo et al. Active recombinant hepcidin FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 3799 Experimental procedures Bacterial strains, media and chemicals The TG1 E. coli strain [D(lac pro) supE thi hsdD5 F’ traD35 proAB LacIq LacZDM15] was used for cloning and plasmid DNA purification. Origami B(DE3) cells (Novagen, Merck Chemicals Ltd, Nottingham, UK) [F ) ompT hsdS B (r B ) m B ) ) gal dcm lacY1 aphC (DE3) gor522::Tn10 trxB (Kan R , Tet R )] were used for protein expression. LB [Bacto-tryptone 1% (w ⁄ v), yeast extract 0.5% (w ⁄ v), NaCl 5% (w ⁄ v)] and ZYM-5052 [Bacto-tryptone 1% (w ⁄ v), yeast extract 0.5% (w ⁄ v), 25 mm Na 2 HPO 4 ,25mm KH 2 PO 4 ,50mm NH 4 Cl, 5 mm Na 2 SO 4 supplemented with 2mm MgSO 4 , 0.5% glycerol, 0.05% glucose and metal salts] (1000 · metal salts solution: 50 mm FeCl 3 ,20mm CaCl 2 ,10mm MnCl 2 ,10mm ZnSO 4 ,2mm CoCl 2 ,2mm CuCl 2 ,2mm NiCl 2 ,2mm Na 2 MoO 4 ,2mm Na 2 SeO 3 , 2mm H 3 BO 3 ,60mm HCl) were used as regular and high density culture media, respectively, in accordance with Studier et al. [32]. Constructions Human and mouse synthetic oligonucleotides (Table 1) cor- responding to the hepcidin sequence fused to a LIC site were synthesized by Eurogentec (Seraing, Belgium). Mice, unlike humans, have two duplicated genes, Hepc1 and Hepc2 [5,37]. Because the mouse Hepc2 gene was found unrelated to iron metabolism [37,38], Hepc1 was chosen for the present study. Each primer (10 lm) was annealed to its complementary partner and heated at 90 °C. Temperature was allowed to decrease to 25 °Cat1°CÆmin )1 (using a MJ research ther- mocycler; Bio-Rad). One microlitre of the double stranded hepcidin DNA was mixed with 1 lL of pET-32Ek⁄ LIC (50 ng) vector (Novagen) at 25 °C for 5 min and one-tenth of the mixture was used to electroporate E. coli TG1cells. Upon transformation, colonies were used to prepare plas- mid DNA and the presence of pET-32Ek ⁄ LIC-Hepcidin vector (either mouse or human) was confirmed by sequencing performed at Genome Express (Meylan, France). Expression The pET-32Ek ⁄ LIC-mouse1hepcidin vector (pET-32LIC- m1hepc) and the pET-32Ek ⁄ LIC-human hepcidin vector (pET-32LIC-huhepc) were used to transform ORIGAMI B(DE3) competent cells. The transformed cells were cultured in 50 mL of LB medium supplemented with kana- mycin (10 mgÆL )1 ) and ampicillin (50 mg ÆL )1 ). The precul- ture was used to inoculate 3.6 L of high-density culture medium ZYM-5052 [32] supplemented with kanamycin and ampicillin. The culture was carried out in twelve 2 L Erlen Flasks at 37 °C for 3 h, then 20–23 h at 25 °C. Isopropyl thio-b-d-galactoside (1 mm) was finally added and culture allowed to continue for an extra 3 h at 25 °C. The cells were harvested at 6000 g for 20 min at 4 °C. Routinely, the biomass obtained represented 17 gÆL )1 of wet cells culture. Fusion protein purification TRX-hepcidin containing cells were resuspended in 5 mL of phosphate buffer (50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 )atpH7 supplemented with NaCl at 300 mmÆg )1 of cell pellet. Bacte- ria were lysed in the presence of BugBusterÔ (0.5·) lyso- zyme (0.5 gÆL )1 ) and DNAse (0.05 mgÆg )1 of cell pellet), then sonicated on ice in the presence of phen- ylmethanesulfonyl fluoride (0.5 mm) and the protease inhib- itors leupeptin, pepstatin and aprotinin (10 lgÆ mL )1 each) using a LabsonicÔ (B. Braun, Melsungen AG, Melsungen, Germany) device operated at 200 W and with a 0.7 s pulse. Cells debris were centrifuged at 10 000 g for 20 min. The resulting supernatant (15–20 mgÆmL )1 of protein) was heated to 65 °C and immediately transferred on ice. The heat inactivated precipitated proteins were removed by centrifugation at 12 000 g for 15 min and the cleared super- natant, routinely containing 5 mgÆmL )1 of protein, was loaded onto a TALONÔ Co 2+ (Clontech, Ozyme, St Quentin-en-Yvelines, France) affinity column (15 mL bed volume). The column was washed with five volumes of phosphate buffer and the TRX-Hepcidin fusion protein was eluted with the same buffer completed with imidazole 150 mm (one volume) and 20 mm MES buffer at pH 5 (one volume). Cleavage and purification of hepcidin Affinity purified TRX-hepcidin containing fractions were pooled and thoroughly dialyzed at 4 °C against 20 mm Tris buffer (pH 7.4 at 25 °C) supplemented with 500 mm NaCl, then 1 unit of biotynilated thrombin (Novagen) was added for 10 mg of fusion protein and cleavage was allowed to proceed for 16 h at room temperature at a protein concen- tration of 1.5–3 mgÆmL )1 . The biotynilated thrombin was extracted using sreptavidin-agarose and the resulting solu- tion containing TRX, S-hepcidin and, eventually, some uncleaved TRX-hepcidin was concentrated to 8 mgÆmL )1 . Fractions (1.5 mL) of the concentrated protein solution were injected onto a BioBasic C18 semi preparative HPLC column (250 mm · 10 mm, 300 A ˚ ; Thermo, Courtabeuf, France) under the control of a Spectra physics HPLC sys- tem. The mobile phase run at 3 mLÆmin )1 consisted of the water ⁄ acetonitrile gradient: 100% H 2 O [trifluroacetic acid (TFA) 0.1%] for 10 min, 0–40% CH 3 CN (TFA 0.1%) for 20 min, 40–60% CH 3 CN (TFA 0.1%) for 20 min and 60– 100% CH 3 CN (TFA 0.1%) for 20 min. The S-tag-hepcidin containing fractions were eluted around 60% H 2 O (TFA Active recombinant hepcidin B. Gagliardo et al. 3800 FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 0.1%) and 40% CH 3 CN (TFA 0.1%) and lyophilized using an Alpha 1-LU Lyophilizer (Avantec, Illkirch, France). S-hepcidin cleavage S-hepcidin was solubilized at 1 mgÆmL )1 of protein in 20 mm Tris buffer (pH 6.7) containing 75 mm NaCl, 2 mm CaCl 2 and 0.05% Chaps and digested for 6 h at room tem- perature with 10 units of Enterokinase (Novagen) per mg of S-hepcidin. Enterokinase was removed using the manu- facturer capture beads and the enterokinase free digestion mixture was diluted in one volume of H 2 O to decrease the ionic strength prior to injection onto a UnoQ column (Bio- Rad) anion exchange column under the control of a Bio- logicÔ (Bio-Rad) device equipped with a 258 nm detector. A salt gradient was performed between 20 mm Tris buffer at pH 7 and 20 mm Tris buffer at pH 7 containing 1 m NaCl for human hepcidin and at pH 6.2 for the mouse form. The cleaved recombinant hepcidin, either mouse or human, was eluted in the void volume. The pure hepcidins were further characterized by HPLC on a Vydac C18 (250 mm · 4.6 mm, 300 A ˚ pores; Grace, Templemars, France) reverse phase column. Fifty microlitres of hepcidin (0.1 mgÆmL )1 ) were eluted with the following gradient at 1mLÆ min )1 from 90% of (H 2 O, 0.1% TFA) and 10% of (H 2 O 30%, CH 3 CN 70%, TFA 0.1%) to 10% of (H 2 O, 0.1% TFA) and 90% of (H 2 O 30%, CH 3 CN 70%, TFA 0.1%) for 30 min. M1hepc was eluted at 18 min 20 s and huhepc at 18 min 45 s. Protein characterization Protein concentration was determined using Bradford (Bio- Rad) reagent [39]; for peptide concentration determination, Bacitracin (1423 Da) was used as a standard rather than serum albumin. Hepcidin concentration was determined either by weighing a salt-free lyophilized powder or using absorbance spectroscopy (A 214 to A 225 ), as described previ- ously [2]. Purification was followed using SDS ⁄ PAGE anal- ysis of peptides and small proteins (< 10 kDa) were run on Criterion 12% BISTRIS XT gels (Bio-Rad) at 200 V and 190 mA for 45 min in MES buffer at pH 7 in the presence or absence of the reducing agent tris(2-carboxyethyl)phos- phine. Immediately after migration, gels were treated with glutaraldehyde (5%, v ⁄ v final) for 20 min and stained using colloidal Blue [40]. Proteins (> 10 kDa) were characterized using Capillary Electrophoresis ExperionÔ PRO260 chips run on an Experion Ô automated electrophoresis station (Bio-Rad). Free thiol determination was performed follow- ing the Ellman methodology [34]. The titration curve was performed using glutathione (4–200 lm) as a standard and measuring absorbance (A 380 to A 480 ) on a Uvikon 420 (Kontron, Zurich, Switzerland) spectrophotometer in the presence of 5 mm 5,5¢-dithiobis(2-nitrobenzoic acid). Free thiol concentration was measured on hepcidin samples using at least 20 lm solutions. Mass spectrometry was run on a LCQ advantage ion trap (Thermo Finnigan, Courta- beuf, France) mass spectrometer under an ESI positive mode (capillary at 275 °C, capillary voltage 21 V, spray 5 kV). SELDI-TOF MS was performed by Photeomics (Noisy le Grand, France) on a PBS IIc ProteinChip Reader (Ciphergen Bioystems Inc., Le Raincy, France) using eight- spot NP20 ProteinChips (Bio-Rad). Cell culture and western blot analysis The mouse monocyte-macrophage cell line J774 was cul- tured as described previously [20]. To increase ferroportin distribution to the cell membrane prior to hepcidin treat- ment, cells were incubated for 16 h with Fe-nitrilotriacetate solution (Fe-NTA; FeCl 3 100 lm-NTA 400 lm). Synthetic human hepcidin (Peptides International), m1hepc and huhepc were used at 700 nm for 5 h as described. Protein extraction and ferroportin detection were performed as previously described [20]. Acknowledgements We would to thank Dr Dan Qing Lou for the murine cDNA prohepcidin construct and Nicole Kubat for critically reading the manuscript. This study was sup- ported by funding from ANR (RO06024KK project) and EEC Framework 6 (LSHM-CT-037296 Euro- Iron1). References 1 Krause A, Neitz S, Magert HJ, Schulz A, Forssmann WG, Schulz-Knappe P & Adermann K (2000) LEAP-1, a novel highly disulfide-bonded human peptide, exhibits antimicrobial activity. FEBS Lett 480, 147–150. 2 Park CH, Valore EV, Waring AJ & Ganz T (2001) Hepcidin, a urinary antimicrobial peptide synthesized in the liver. J Biol Chem 276, 7806–7810. 3 Nicolas G, Bennoun M, Porteu A, Mativet S, Beau- mont C, Grandchamp B, Sirito M, Sawadogo M, Kahn A & Vaulont S (2002) Severe iron deficiency anemia in transgenic mice expressing liver hepcidin. Proc Natl Acad Sci USA 99, 4596–4601. 4 Ganz T (2003) Hepcidin, a key regulator of iron metab- olism and mediator of anemia of inflammation. Blood 102, 783–788. 5 Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, Brissot P & Loreal O (2001) A new mouse liver-specific gene, encoding a protein homologous to human antimi- crobial peptide hepcidin, is overexpressed during iron overload. J Biol Chem 276, 7811–7819. 6 Zhang AS, Xiong S, Tsukamoto H & Enns CA (2004) Localization of iron metabolism-related mRNAs in rat B. Gagliardo et al. Active recombinant hepcidin FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 3801 liver indicate that HFE is expressed predominantly in hepatocytes. Blood 103, 1509–1514. 7 Sang Y, Ramanathan B, Minton JE, Ross CR & Blecha F (2006) Porcine liver-expressed antimicrobial peptides, hepcidin and LEAP-2: cloning and induction by bacterial infection. Dev Comp Immunol 30, 357– 366. 8 Fry MM, Liggett JL & Baek SJ (2004) Molecular clon- ing and expression of canine hepcidin. Vet Clin Pathol 33, 223–227. 9 Huang PH, Chen JY & Kuo CM (2007) Three different hepcidins from tilapia, Oreochromis mossambicus: ana- lysis of their expressions and biological functions. Mol Immunol 44, 1922–1934. 10 Lauth X, Babon JJ, Stannard JA, Singh S, Nizet V, Carlberg JM, Ostland VE, Pennington MW, Norton RS & Westerman ME (2005) Bass hepcidin synthesis, solution structure, antimicrobial activities and syner- gism, and in vivo hepatic response to bacterial infec- tions. J Biol Chem 280, 9272–9282. 11 Park KC, Osborne JA, Tsoi SC, Brown LL & Johnson SC (2005) Expressed sequence tags analysis of Atlantic halibut (Hippoglossus hippoglossus) liver, kidney and spleen tissues following vaccination against Vibrio anguillarum and Aeromonas salmonicida. Fish Shellfish Immunol 18, 393–415. 12 Nicolas G, Chauvet C, Viatte L, Danan JL, Bigard X, Devaux I, Beaumont C, Kahn A & Vaulont S (2002) The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation. J Clin Invest 110, 1037–1044. 13 Ganz T (2006) Hepcidin and its role in regulating sys- temic iron metabolism. Hematology Am Soc Hematol Educ Program 1, 29–35. 14 Finberg KE, Heeney MM, Campagna DR, Aydinok Y, Pearson HA, Hartman KR, Mayo MM, Samuel SM, Strouse JJ, Markianos K et al. (2008) Mutations in TMPRSS6 cause iron-refractory iron deficiency anemia (IRIDA). Nat Genet 40, 569–571. 15 Valore EV & Ganz T (2008) Posttranslational process- ing of hepcidin in human hepatocytes is mediated by the prohormone convertase furin. Blood Cells Mol Dis 40, 132–138. 16 Hunter HN, Fulton DB, Ganz T & Vogel HJ (2002) The solution structure of human hepcidin, a peptide hormone with antimicrobial activity that is involved in iron uptake and hereditary hemochromatosis. J Biol Chem 277, 37597–37603. 17 Nemeth E, Preza GC, Jung CL, Kaplan J, Waring AJ & Ganz T (2006) The N-terminus of hepcidin is essen- tial for its interaction with ferroportin: structure– function study. Blood 107, 328–333. 18 Nicolas G, Bennoun M, Devaux I, Beaumont C, Grandchamp B, Kahn A & Vaulont S (2001) Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proc Natl Acad Sci USA 98, 8780–8785. 19 Donovan A, Lima CA, Pinkus JL, Pinkus GS, Zon LI, Robine S & Andrews NC (2005) The iron exporter ferroportin ⁄ Slc40a1 is essential for iron homeostasis. Cell Metab 1 , 191–200. 20 Delaby C, Pilard N, Goncalves AS, Beaumont C & Canonne-Hergaux F (2005) Presence of the iron exporter ferroportin at the plasma membrane of macrophages is enhanced by iron loading and down- regulated by hepcidin. Blood 106, 3979–3984. 21 Nemeth E, Tuttle MS, Powelson J, Vaughn MB, Dono- van A, Ward DM, Ganz T & Kaplan J (2004) Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science 306, 2090–2093. 22 Ganz T (2006) Hepcidin – a peptide hormone at the interface of innate immunity and iron metabolism. Curr Top Microbiol Immunol 306, 183–198. 23 Kluver E, Schulz A, Forssmann WG & Adermann K (2002) Chemical synthesis of beta-defensins and LEAP- 1 ⁄ hepcidin. J Pept Res 59, 241–248. 24 Murphy AT, Witcher DR, Luan P & Wroblewski VJ (2007) Quantitation of hepcidin from human and mouse serum using liquid chromatography tandem mass spec- trometry. Blood 110, 1048–1054. 25 Achmuller C, Kaar W, Ahrer K, Wechner P, Hahn R, Werther F, Schmidinger H, Cserjan-Puschmann M, Clementschitsch F, Striedner G et al. (2007) N(pro) fusion technology to produce proteins with authentic N termini in E. coli. Nat Methods 4, 1037–1043. 26 Gerardi G, Biasiotto G, Santambrogio P, Zanella I, Ingrassia R, Corrado M, Cavadini P, Derosas M, Levi S & Arosio P (2005) Recombinant human hepcidin expressed in Escherichia coli isolates as an iron contain- ing protein. Blood Cells Mol Dis 35, 177–181. 27 Wallace DF, Jones MD, Pedersen P, Rivas L, Sly LI & Subramaniam VN (2006) Purification and partial char- acterisation of recombinant human hepcidin. Biochimie 88, 31–37. 28 Zhang H, Yuan Q, Zhu Y & Ma R (2005) Expression and preparation of recombinant hepcidin in Escherichia coli. Protein Expr Purif 41, 409–416. 29 Koliaraki V, Marinou M, Samiotaki M, Panayotou G, Pantopoulos K & Mamalaki A (2008) Iron regulatory and bactericidal properties of human recombinant hepcidin expressed in Pichia pastoris. Biochimie 90, 726–735. 30 Prinz WA, Aslund F, Holmgren A & Beckwith J (1997) The role of the thioredoxin and glutaredoxin pathways in reducing protein disulfide bonds in the Escherichia coli cytoplasm. J Biol Chem 272, 15661–15667. 31 Stewart EJ, Aslund F & Beckwith J (1998) Disulfide bond formation in the Escherichia coli cytoplasm: an in vivo role reversal for the thioredoxins. EMBO J 17, 5543–5550. Active recombinant hepcidin B. Gagliardo et al. 3802 FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS [...]... Vaulont S (2006) Targeted disruption of the hepcidin 1 gene results in severe hemochromatosis Blood 108, 1402– 1405 39 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding Anal Biochem 72, 248–254 40 Klafki HW, Wiltfang J & Staufenbiel M (1996) Electrophoretic separation of betaA4 peptides (1-40) and (1-42)... hepcidin assay by mass spectrometry Blood 106, 3268–3270 36 Melino S, Garlando L, Patamia M, Paci M & Petruzzelli R (2005) A metal-binding site is present in the amino terminal region of the bioactive iron regulator hepcidin-25 J Pept Res 66, 65–71 Active recombinant hepcidin 37 Lou DQ, Nicolas G, Lesbordes JC, Viatte L, Grimber G, Szajnert MF, Kahn A & Vaulont S (2004) Functional differences between hepcidin...B Gagliardo et al 32 Studier FW (2005) Protein production by auto-induction in high density shaking cultures Protein Expr Purif 41, 207–234 33 LaVallie ER, DiBlasio EA, Kovacic S, Grant KL, Schendel PF & McCoy JM (1993) A thioredoxin gene fusion expression system that circumvents inclusion body formation in the E coli cytoplasm Biotechnology (NY) 11, 187–193 34 Ellman GL (1959)... Anal Biochem 72, 248–254 40 Klafki HW, Wiltfang J & Staufenbiel M (1996) Electrophoretic separation of betaA4 peptides (1-40) and (1-42) Anal Biochem 237, 24–29 FEBS Journal 275 (2008) 3793–3803 ª 2008 The Authors Journal compilation ª 2008 FEBS 3803 . Production of biologically active forms of recombinant hepcidin, the iron-regulatory hormone Bruno Gagliardo 1 , Audrey. preparation of biologically active hepcidins, either synthetic or recombinant, is a necessary step for investigating the cellular biology of this hormone and the