1 Introduction

Hepcidin is an antimicrobial peptide (AMP) and iron regulatory molecule, and is of importance in host defense against microbial infection. To date, large numbers of hepcidin have been isolated from human blood ultrafiltrate and urine, other mammals and fish. Mature peptide of hepcidin was isolated firstly in fish from the hybrid striped bass in 2002^[1]. In addition, peptide sequences of hepcidins or expressed sequence tags have been predicted in Perciformes fishes, including Lateolabrax japonicus^[2], gilthead sea bream^[3], medaka, rainbow trout^[4], winter flounder^[5], long-jawed mudsucker^[6] and Altlantic salmon^[7] and so on. Fish hepcidin cDNA and genomic DNA organization have been determined in many fish, including hybrid striped bass^[1], winter flounder^[5], Atlantic salmon^[7], zebrafish^[8], rainbow trout^[9], olive flounder^[10], Japanese flounder^[11], red sea bream^[12], Japan sea bass^[2], sea bass^[13], black porgy^[14], gilthead sea bream^[3]. The hepcidin peptides contain unique structure among antimicrobial peptides, sharing rich cysteines at conserved positions with a distinctive disulfide bridge structure^{[1, 15]}. Cysteine-rich antimicrobial peptides of the defensin family have been detected in the fat body of insects and the haemolymph of mollusks^[16-19]. Besides, hepcidin takes part in iron metabolism and relates to disorders in iron homeostasis resulting in iron deficiency or overload^[20-24].

Arguments against hepcidin being predominantly expressed in the liver have recently occurred in a few publications^{[7, 11, 14]}, either in mammalian or in fish hepcidin studies. In mammals, Kulaksiz et al. suggested that hepcidin was not liver-specific but might be expressed also in the kidney^[25-26]. While in the red sea bream and catfish, hepcidin mRNA was widely expressed in multiple tissues^{[27, 21]}, from which the conclusion was derived that in fish hepcidin might have a non liver-specific expression.

Koi (Cyprinus carpio) is a very important economic species of cultured fish in China. Little is known about the hepcidin and its immune regulating mechanism in koi. This study aimed to present and analyze of cDNA sequences of hepcidin (including the deduced amino acid sequence) in koi and evaluate its expression profiles in several tissues before and after pathogenic bacteria Aeromonas veronii challenge.

2 Materials and Methods 2.1 Fish maintenance

Healthy koi (60±2 g) were obtained from a koi farm in Beijing. Fish were acclimatized in laboratory for 7 days at 25.0±0.5 ℃, dissolved oxygen of 6.5% and fed with fish feed once per day.

2.2 Medium

Luria-Bertani (LB) broth and LB Agar were prepared as the document^[28-29].

2.3 Reagents and instuments

PureLink^TM RNA Mini Kit, Life technologies USA; SMARTer^TM RACE cDNA Amplication Kit, Clontech USA; Wizard ^{SV Gel and PCR Clean-Up System, Promega USA; pEASY-T5 cloning vector, TransGen China; PrimeScript RT reagent kit, SYBR Premix ExTaqTM Ⅱ, TaKaRa. 7500 Real Time PCR System, AB Applied Biosystem; Agarose gel electrophoresis, Bio-Rad.}

2.4 Cloning and sequencing of koi hepcidin cDNA

Livers were collected from koi, frozen immediately in liquid nitrogen individually, and stored at -80 ℃ for use. Total RNA from the liver of koi was extracted using PureLink^TM RNA Mini Kit, according to the manufacturers’ instructions. The quality of total RNA was assessed by electrophoresis on 1% agarose gel. In order to amplify the hepcidin cDNA from koi, RT-PCR and rapid amplication of cDNA ends (RACE) were performed following the SMARTer^TM RACE cDNA Amplication Kit manual. The RACE reactions were performed with primers (Table 1). The pair of primers was designed based on the EST fragments from the common carp cDNA library established in the centre for applied aquatic genomics, and was in the open reading frame (ORF) of carp hepcidin (c-hepc). RACE PCR reactions were performed as following: 5 cycles of denaturation at 94 ℃ for 30 s, annealing at 72 ℃ for 2 min; 5 cycles of denaturation at 94 ℃ for 30 s, annealing at 70 ℃ for 30 s, and extension at 72 ℃ for 2 min; 25 cycles of denaturation at 94 ℃ for 30 s, annealing at 68 ℃ for 30 s and extension at 72 ℃ for 2 min. Amplified products were analyzed on agarose gel electrophoresis and purified from gel with a Wizard ^{SV Gel and PCR Clean-Up System. The purified PCR product was cloned in the pEASY-T5 cloning vector. After sequenced, 5′- end and -3′ end sequences were assembled to a k-hepc cDNA contig.}

2.5 Bioinformatic analysis

Translation of the cDNA was performed using DNAStar software. The amino acid sequence of koi hepcidin (k-hepc) was analyzed using the BLAST at the NCBI (http://www.ncbi.nlm.nih.gov/blast). The cleavage site for the signal peptide was predicted by the SignalP (http://www.cbs.dtu.dk/services/SignalP/)^[30]. Charges of the deduced mature peptides of k-hepc were calculated by the ProtParam tool (http://cn.expasy.org/tools/protparam.html)^[31]. Multiple-sequence alignment of the k-hepc with hepcidin from other animals was performed with the ClustalX^[32]. Homology searches were performed using BLASTn and BLASTp by the National Center for Biotechnology Information (http://www.ncbi.nlm.nih. gov/). Deduced aminoacid sequences from 23 hepcidin gene sequences were used to construct a prepro hepcidin multiple alignment by ClustalX software. Besides, a Neighbor- Joining phylogenetic tree of 23 prepro hepcidin sequences with Poisson Correction model, pairwise deletion option and 10 000 replicates of bootstrap was built with MEGA 4.0.2 software.

2.6 Aeromonas veronii challenge of koi and RNA sampling

The challenge experiment for assessing induction of gene expression was designed to intramuscular infection with Aeromonas veronii, a pathogenic bacterium (CGMCC 7231) isolate for this fish species. Twenty fish were injected with approximately 2×10⁵ CFU/mL, and the equal quantities of fish were injected with 0.7% sterile physiological saline solution (PSS) as the control. The two treatment groups were kept in seperate water cages after injection. The different tissues of three fish in each group were sampled at 0, 4, 8, 12, 24 and 48 h post-injection from the A. veronii-challenged and PSS mock-challenged fish. The tissues including the liver, spleen, kidney, intestine, brain, heart, muscle, and gill were separately collected from each individual fish, and frozen immediately in liquid nitrogen individually, and stored at -80 ℃. Tissue samples were homogenized in PureLink^TM RNA Mini Kit, and total RNA was extracted.

2.7 Analysis of hepcidin expression by RT-PCR and real time PCR

The RT-PCR reactions for hepcidin expression were carried out (37 ℃ for 15 min, 85 ℃ for 5 s, followed by 4 ℃) using PrimeScript RT reagent kit. The relative quantity of k-hepc mRNA was evaluated using the comparative C_t (Cycle threshold) method with a 7500 Real Time PCR System and SYBR Premix ExTaq^TM Ⅱ. The primers specific for k-hepc gene and the endogenous gene control are listed in Table 1. The cycling profile was as follows: stage 1: 1 cycle of denaturation at 95 ℃ for 30 s, stage 2: 40 cycles of PCR reaction at 95 ℃ for 5 s and 60 ℃ for 34 s, stage 3: dissociation stage at 95 ℃ for 15 s, 60 ℃ for 1 min and 95 ℃ for 15 s. The validation experiment was performed using three diluted cDNA series from the control liver to verify that efficiencies of the k-hepc mRNA and β-actin gene were approximately equal in real time PCR. Expression of k-hepc gene was normalized to an endogenous reference β-actin and presented as ∆C_t values. Tissue-specific expression of k-hepc gene was investigated in three normal fish, including the liver, spleen, kidney, intestine, brain, heart, muscle and gill. The expression patterns of k-hepc gene at the time course of A. veronii-challenge were investigated in the liver, spleen, kidney, heart, intestine, brain and muscle, using 0 h normal fi sh liver RNA pool as calibrator samples. All data were obtained in comparison with the same calibrator.

2.8 Statistical analysis

7500 software V2.0.6 automatically calculates the relative quantification (RQ) value. RQ=2^-∆∆C_t. All data were analyzed by one-way analysis of variance (ANOVA). In addition, two-way ANOVA was used to determine the interactions of injection treatment and time. Duncan’s multiple range tests and critical range was used to test differences among individual means. Graphs were made by the box-whisker plot. Difference were regarded as significant when P<0.05.

3 Results 3.1 Determination of k-hepc complete coding sequence (CDS)

Koi hepcidin CDS was obtained from the two EST sequences assembly and the primers were in the ORF of the k-hepc. Out of expectation, there was only R-cg3 without F-cg3 in 5′-RACE product of mRNA, which showed that cDNA sequence of k-hepc was not consistent with that of carp hepcidin even in the ORF sequence (Figure 1). The sequence of k-hepc cDNA (GenBank accession no. KC795559) is 755 bp in length, including 5′-untranslated region (5′-UTR) of 112 bp, -3′ untranslated region (-3′UTR) of 367 bp, and ORF of the 276 bp. Compared with other reported cyprinid fish hepcidin, cDNA sequences of k-hepc is similar with that of c-hepc as high as 94%. The deduced amino acid sequence of k-hepc (GenBank no. AGO64769.1) is 91 amino acids in length and consists of three domains: signal peptide (24 residues), prodomain (42 residues) and mature peptide (25 residues) (Figure 2). According to the analysis by the SignalP, the signal peptide cleavage site of the deduced k-hepc was predicted between Ala 24 and Val 25 (Figure 2). The mature peptide region of k-hepc was predicted, with “RX(F/R)R” as characteristic sequence for propeptide convertases, consisting of 25 amino acid residues at the C terminus of ORF. The processed mature peptide of k-hepc is predicted by ProtParam to be positively charged at neutral pH, having a theoretical pI of 8.34. It is thus a cationic protein. The peptide k-hepc molecular weight is 2 892.5 Da.

3.2 Amino acid sequence alignment

The deduced amino acid sequence of k-hepc has 29%-93% similarity to hepcidins of other fish species and mammals, sharing eight cysteines at the identical conserved position (Figure 3). The alignment showed that all listed hepcidins (23 hepcidin or hepcidin-like sequences searched on Genbank), including fish (18 hepcidin sequences), mammalian (4 hepcidin sequences) and reptile (1 hepcidin sequence) hepcidins were most characterized by eight cysteine residues conserved at identical positions in the mature peptide region. The predicted signal peptide was highly conserved between k-hepc and three other fish hepcidins, such as common carp, Puntius sarana and Danio rerio hepcidins (Figure 3), while the other 14 species fish hepcidins were conserved with each other.There was lower similarity in the signal peptide sequence of hepcidin between fish and mammals (Figure 3). The signal peptide cleavage site of all deduced fish hepcidins was between Ala 24 and Val 25. However, the 25th amino acid in some hepcidins was substituted by other amino acids (Ala 24-Ile 25, or Ala 24-Ser 25, or Ala 24-Phe 25) as shown in Figure 3. The deduced amino acid sequence of k-hepc showed 93% homology with c-hepc when analyzed by GenBank BLASTp, the highest in comparison with other fish hepcidins (Figure 3). Both hepcidin sequences are fully identical in the signal peptide. The RX (K/R) R motif typical of propeptide convertases is identified among all the tested animals (Figure 3).

3.3 Phylogenetic analysis of k-hepc

Phylogenetic analysis of the hepcidin-like family indicated that two clusters were present: mammalian and fish hepcidins (bootstrap value>95%). The deduced amino acid sequence from koi was in a branch position with that from common carp, zebra fish and Puntius sarana (Figure 4). The other branch of fish hepcidins were almost percomorpha and parapercomorpha peptides, which seems closer to each other than to cyprinid in hepcidin evolution.

3.4 Tissue expression profiles of k-hepc gene in normal koi

Expression of hepcidin was detected by means of RT-PCR and real-time PCR in all the assayed tissues, including the liver, spleen, kidney, intestine, brain, heart, muscle and gill (Figure 5). The highest amount of k-hepc mRNA transcripts was demonstrated in the liver and the lowest relative expression level was tested in the gill in normal fish (Figure 5).

3.5 Koi hepcidin gene expression in bacterial infected koi

The expression pattern of k-hepc was investigated in the liver, spleen, kidney, intestine, brain, heart and muscle at the time course of 0, 4, 8, 12, 24 and 48 h post injection with A. veronii and PSS, using real-time PCR. All the results are presented with β-actin as an endogenous control. A significant up-regulation of k-hepc expression was observed in liver at 4 h and in heart at 12 h after A. veronii- challenged (P<0.05) (Figure 6). The liver displayed the higher expression level over 20 fold compared with the control, and the heart displayed over 13 fold compared with the control. No significant up-regulation of k-hepc expression was observed in brain, intestine, kidney, spleen and muscle after A. veronii-challenged (Figure 6).

Two-way ANOVA (Table 2) showed that the expression of k-hepc mRNA in liver and heart were significantly affected by bacterial injection (P<0.05), and the expression of k-hepc in brain, intestine, kidney, spleen, and muscle were not obviously affected (P>0.05). Moreover, challenge time significantly affected the heart, brain, intestine, kidney (P<0.01) and muscle (P<0.05), but no obviously affected the liver and spleen (P>0.05). Interaction effects between injection time and bacteria in heart were significant (P<0.01), and not significant (P>0.05) in other tissues.

4 Discussion

In this study, the deduced amino acid sequence of k-hepc from koi was comprised of 91 amino acids. Although several hepcidins from fish have been predicted or determined, there was little information about hepcidins of cypriniformes fish except common carp^[33], Puntius sarana and zebrafish. So the nucleotide sequence of k-hepc might be different from those of other known cypriniformes hepcidins. The hepcidin-like peptide identified in this paper has the same subregions such as signal peptide, prodomain peptide and mature peptide with other hepcidins identified in fish and mammals, including human. The signal peptides were almost invariably 24 aa in length. Although Barnes et al^[34] reported that the signal peptide regions were highly conserved among most of other fish, the cyprinid fish, such as koi, common carp, Puntius sarana and zebrafish, shared another type of signal peptide regions of hepcidin. They have not polar Ser but nonpolar Ala at amino acid position 23 (Figure 3). K-hepc contains a typical endoplasmic reticulum targeting signal sequence, a consensus cleavage site forprohormone convertases^[35], a propeptide convertase site at amino acid position 66, and 8 cysteine residues as a characteristic of many hepcidins in fish and human^{[8, 36]}. These structures are highly conserved among all animals in Figure 3.

The expression of hepcidin genes was significantly induced in the liver and heart of koi after Aeromonas veronii-challenged, but not obviously changed in other tested tissues. Our observation on the hepcidin expression pattern in the liver completely matched reported for bacterially challenged white bass where up-regulation was most dramatic in liver^[1]. Hilton et al^[37] also summarized recently that hepcidin transcript levels in fish challenged with bacteria were increased primarily in the liver. Furthermore, the expression of hepcidin genes was significantly induced in the heart of koi after bacterial-challenged in this study. The Atlantic salmon hepcidin Sal1 and Sal2 transcripts were both up-regulated in multiple tissues with bacterial challenge^[7]. These results indicate that the regulation of hepcidin-like transcripts from fish might be highly diverged in different species. K-hepc mRNA expression in the brain, intestine, kidney, spleen and muscle was not significantly changed during the period of 48 h after bacterial challenge with unchallenged normal fish, suggesting k-hepc was constitutively expressed in these tissues tested. In addition, according to the analysis of two-way ANOVA, we thought that the impact of time upon k-hepc mRNA expression in the brain, intestine, kidney and muscle was greater than the influence of bacteria. Apparently, different tissues of koi produce hepcidin-like peptides in a constitutive or inducible manner.

In conclusion, we have presented the sequences of hepcidin from koi. The k-hepc is close to the other three cyprinid fish hepcidins, with cysteine conformation and propeptide convertase cleavage sites. Hepcidin transcripts are widely distributed in various tissues of koi. Furthermore, the study describes the expression pattern of hepcidin-like gene of koi and their patterns of expression level at different conditions and in different tissues, showing k-hepc might be not induced by bacteria-challenge but innate regulation with time.