To detect the iron requirement of S. iniae, HD-1 cells were cultured. HD-1 was isolated from Plectorhynchus cinctus in China, and had been described previously and characterized thoroughly by Zhou et al. [12] . HD-1 cells were grown in complete medium, brain heart infusion (BHI) at 28˚C, iron-restricted medium was prepared by adding 0.1285 g nitrilotriacetic acid trisodium salt (NTA), 3.6 g brain heart infusion in 100 ml dd H₂O, and supplementing it with 0.0043 g MnCl₂, 0.0038 g ZnCl₂, 0.0031 g CaCl₂, and 0.0033 g MgCl₂ [13] .

Genomic DNA was extracted from the S. iniae HD-1 using the Wizard genomic DNA purification kit, and the products were quantified by measuring the absorbance at 260 nm. PCR was carried out with 1 μg of genomic DNA using the primers 5'-TTTTGTGACATATAGTTGGCGGGCA-3' and 5'-ATGACGCCTAACAAAGAAGATT-3' as described by Zou [4] , and the PCR products were sequenced at Invitrogen corporation to confirm their specificity. Briefly, the cycling conditions used for mtsR ORF were as follows: 1 cycle of 94˚C for 2 min, 35 cycles of 94˚C for 45 sec, 35 cycles of 61˚C for 45 sec, 35 cycles of 72˚C for 1 min, 1 cycle of 72˚C for 7 min. Nucleotide and deduced amino acid homology analysis of MtsR were carried out by NCBI BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The Conserved Domains of mtsR were detected by NCBI CD Search (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi).

Duplicate cultures of S. iniae HD-1 were harvested at early logarithmic phase (OD_{600 nm} = 0.35), mid-logarithmic phase (OD_{600 nm} = 0.65), and final-logarithmic phase (OD_{600 nm} = 0.80) in BHI orin iron-restricted medium. The total RNA was extracted and was reverse transcribed to cDNA. Real-time fluorescence PCR (qPCR) analysis was performed with a LightCycler480 system. The primers of mtsABC, mtsR and gyrA used in qPCRas listed in Table 1 [4] , and the cycling protocol were an initial denaturation at 94˚C for 2 min followed by 40 cycles of 94˚C for 15 s, 56˚C for 50 s, and 79˚C for 10 s. Finally, a melting curve was performed to ensure that there was no contamination. qPCR of the genes of interest, mtsA, mtsB, mtsC, and mtsR and a normalizer gene, gyrA, were performed in triplicate for each sample, and included a no-template control to rule out contamination and primer-dimer formation. Gene gyrA was chosen as a normalizing gene because its expression in other streptococci is stable under different test conditions [14] . The expression fold change of gene was calculated base on the comparison with the normalized gyrA. All statistical analyses were performed using the SPSS 16.0 software (SPSS Inc., USA).

To detect the importance of iron for S. iniae, the growth curves of HD-1 in BHI and iron-restricted medium were characterized. The results showed that the growth of S. iniae HD-1 was inhibited by adding NTA to BHI medium supplemented with additional cations, and this inhibition was the result of iron limitation. In iron-restricted medium, the HD-1 cells needed 9.5 h to reach the stationary phase, which was 3 h more than that of in BHI medium, and the growth level was dropped 10% (Figure 1). The gram stain was used to observe the morphological features of S. iniae HD-1 cells at the logarithmic phase, and the results showed that the cells in iron-restricted medium had long chain length (Figure 2(b)) than that of in BHI medium (Figure 2(a)), which indicated that

*Gene correspond to mtsABC gene designations in S. iniae HD-1. F, forward, R, reverse.

the reproduction of S. iniae HD-1 cells had been affected without the support of iron. These results indicated that iron was the essential nutrient that S. iniae needed to survive.

Iron acted as a regulatory factor having influence on proteins production, but the mechanisms processes in S. iniae for iron homeostasis have not been characterized. Screening of S. iniae genome sequences resulted in the identification of gene mtsR (GeneBank No: JN177478) that shared amino acid sequence homologies with DtxR family which were the metal-dependent transcription regulators. The closest homologs for MtsR are the iron-dependent repressors from Streptococcus pyogenes MGAS9429 (Table 2, http://blast.ncbi.nlm.nih.gov/Blast.cgi). mtsR has 624 bp which located at 5' proximal of the iron-transporter mtsABC, and was transcribed in the opposite direction (Figure 3(a)). The localization of mtsR, and its similarity to other iron-dependent transcriptional regulators suggested that MtsR may function as the mtsABC repressor, which can regulate

the transcription of mtsABC in response to iron availability intracellular.

The TBLASTN analysis showed that the predicted amino acid sequence of MtsR was highly conserved (Table 2), and the NCBI CD Search predicted that MtsR has three conserved domains: WHTH_GntR, Fe_dep_repr_C, and FeoA (Figure 3(c)). WHTH_GntR, and Fe_dep_repr_Cdomains are highly conserved in MtsR, which found in other DtxR homologues thatare responsive to iron, manganese, or both (Figure 1(b)). FeoA domain was found at the C-terminus of a variety of metal-dependent transcriptional regulators, which in most cases likely to be either iron or manganese [15] . Based on these observations, we concluded that mtsR is a member of the metal-dependent transcriptional regulators.

Total RNA was isolated from HD-1 cells in BHI and in iron-restricted medium, and qPCR analysis was performed to determine whether MtsR regulated the expression of iron-transporter mtsABC at the transcriptional level. The housekeeping gene gyrA was used as an internal control in qPCR, and similar levels of

amplification confirmed that the RNA quantities used as the templates in all qPCR reactions were equal. When use the RNA isolated from the HD-1 grown in BHI medium as templates, the expression of mtsA, mtsB, and mtsC were significantly up-regulated and reached the peak at the early logarithmic, de-regulated at the mid-logarithmic, and back to housekeeping gene level at the final-logarithmic phase (p < 0.05, Figure 4). This representation owed to the high-efficiency of procaryotic cells. The cells needed iron for growth, they stored iron through up-regulate iron-transporter, and once they captured enough iron, the expression of iron-transporter was up-regulated [16] . The expression tendency of the mtsABC in iron-restricted medium was consistent with that of in BHI medium, but the expression levels of mtsA, mtsB, and mtsC were showed 1.83, 13.3, and 2.11 times higher than that of in BHI medium, respectively (Figure 5). In contrast, both in Figure 4 and Figure 5 the results showed that the high transcription of the mtsABC is observed when the mtsR remain inactivate, suggesting that MtsR up-regulate mtsABC expression was very likely to lead to an increase in iron uptake by S. iniae. Meanwhile, activation of MtsR resulted in de-regulation of the mtsABC transcription, which demonstrated that mtsR

up-regulated the expression of mtsABC in response to iron availability intracellular to control the iron homeostasis in S. iniae HD-1.