PrimeSTAR^®Max DNA polymerase, PrimeScript™ RT reagent kit, and SYBR^@Premix Ex Taq™ II were purchased from TaKaRa (Tokyo, Japan). Restriction enzymes, T4 DNA ligase, and protein markers were obtained from Thermo Fisher Scientific (Waltham, MA, USA). Chloramphenicol, diaminopimelic acid (DAP), 2, 2’-dipyridine, chrome azurol S (CAS), agar for swimming motility assays, dimethyl sulfoxide (DMSO), and 2’,7’-dichlorodihydrofluorescein (H₂DCFDA) were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other molecular kits and DNA markers were purchased from TIANGEN (Beijing, China). If not specified, all other chemical reagents were obtained from Sangon Biotech (Shanghai, China).

V. anguillarum 775 (ATCC 68554) strains and E. coli strains used in this study are shown in Supplementary Table S1 . V. anguillarum 775 cells were grown at 26°C in an M9 high salt medium (90 mM Na₂HPO₄, 22 mM KH₂PO₄, 18.8 mM NH₄Cl, 345 mM NaCl, and 1 mM MgSO₄) supplemented with 0.2% casamino acid. If not specified, rich iron conditions were achieved by adding 50 μM FeCl₃, and limited iron conditions were obtained by adding 50 μM 2, 2’-dipyridine as described in our previous study (Li et al., 2024). For growth assays under different iron conditions, the inocula were grown under their respective iron conditions to the exponential phase. E. coli strains were incubated in lysogeny broth (LB) at 37°C. When the donor E. coli strain X7213 was used, 0.5% DAP was added.

The VaRyhB sRNA was identified from the genome of V. anguillarum 775 (GenBank number of chromosome 1: CP002284.1; GenBank number of chromosome 2: CP002285.1; and GenBank number of the endogenous plasmid pJM1: AY312585.1) by using Vibrio cholerae RyhB as a query (Di Lorenzo et al., 2003; Naka et al., 2011). Sequence alignment was performed by ClustalW (Larkin et al., 2007).

DNA purification, digestion, ligation, and transformation were carried out based on standard molecular biological techniques (Sambrook and Russel, 2001). PCR products were sequenced by Beijing Tsingke Biotech Co., Ltd. (Qingdao, China) and analyzed by the software Vector NTI Advance 11.5.1 (Invitrogen, Darmstadt, Germany). Oligonucleotide sequences used in this study are listed in Supplementary Table S2 .

To construct the unmarked deletion mutants ΔVaryhB and ΔVafurΔVaryhB, a homologous recombination technique was carried out as previously described (Li et al., 2024). In brief, fusion PCR was performed to obtain the 2,000-bp flanking fragment of VaryhB, and the KpnI/SmaI-digested fragment was ligated into pRE112 to generate the plasmid pLYJ220. Then, the pLYJ220-containing donor strain E. coli X7213 was used to transform the plasmid into the V. anguillarum 775 wild-type (WT) strain and the ΔVafur mutant by conjugation as follows: 500 μL of exponential phase E. coli X7213 cells were gently mixed with 800 μL of exponential phase V. anguillarum 775 strains. The mixture was collected by centrifugation at 3,500 g for 10 min and dropped in an LB plate in the presence of 0.5% DAP. After incubation overnight at 26°C, the mixed cells were suspended in 10 mL of LB medium, and 100 μL was plated in an LB plate in the presence of 5 μg/mL chloramphenicol at 26°C for 48 h. PCR was performed to confirm the colony bearing the plasmid integrated into the chromosome of the V. anguillarum 775 strains, and then the colony was incubated in 5 mL of LB medium overnight at 26°C. After two transfers, 50 μL of culture was plated in the LB plate in the presence of 10% sucrose and incubated at 26°C for 48 h. Finally, the correct mutant was confirmed by PCR, designating the ΔVaryhB mutant and the ΔVafurΔVaryhB mutant, respectively. To complement the ΔVaryhB mutant, the VaryhB gene with its promoter region was ligated into SmaI/ApaI-digested pBBR1MCS-2-Cm to obtain the plasmid pLYJ232. The plasmid pLYJ232 was transferred into the ΔVaryhB mutant by conjugation.

To compare gene expression levels among WT, ΔVaryhB mutant, and ΔVafur mutant, different V. anguillarum 775 cells were cultured under limited or rich iron conditions twice and grown to the exponential phase (OD_{600 nm} value of 0.8) under their respective iron conditions for the qRT-PCR assay. In brief, 1 mL of culture was collected at 11,000 g at 4°C. The RNAprep Pure Cell/Bacteria Kit (TIANGEN, China) was used to extract total RNA according to the manufacturer’s instructions. After DNA elimination, the PrimeScript™ RT reagent kit with the gDNA Eraser (TaKaRa, Japan) was used to obtain cDNA. The housekeeping gene mreB, encoding an actin protein, was used as an internal control. The qRT-PCR reactions were carried out in a 20-μL volume with 10 μL of 2xSYBR^@Premix Ex Taq™ II (TaKaRa, Japan), 1 μL of cDNA template, 8.4 μL of ddH₂O, and 0.3 μL of each of the forward and reverse primers (10 μM) using the real-time PCR system Applied Biosystems 7500 (Thermo Fisher Scientific, USA). The relative expression levels were calculated using the comparative C^T method (2^-ΔΔCT) (Livak and Schmittgen, 2001). Three independent experiments were carried out, and values were obtained from representative experiments in triplicate.

For RNA sequencing, ΔVaryhB cells were grown under rich and limited iron conditions to the exponential phase. After RNA extraction, the RNA quality and quantity were assessed by the Agilent RNA 6000 Nano Kit (Agilent, USA), and rRNA was further eliminated using the Ribo-Zero™ Magnetic Kit (Epicentre). Then fragmented mRNA was primed with random primers. When the first-strand cDNA was synthesized, the second-strand cDNA was obtained by adding DNA polymerase I, RNase H, dNTP, and buffer. After purification, end-repairing, and poly(A)-tailing, fragments were ligated to Illumina sequencing adapters, and ones with a length of 300–500 bp were selected. Illumina HiSeq™ 4000 (Illumina, USA) was used for sequencing, and the collected data were analyzed by Shanghai Personal Biotechnology Co. Ltd. (Shanghai, China). Three independent samples under limited or rich iron conditions were used for RNA sequencing. The raw data from transcriptomic analysis have been submitted to the National Center for Biotechnology Information (NCBI) Sequence Read Archive database (accession No. PRJNA1186938). To identify genes showing different regulation in the WT and ΔVaryhB mutant, the raw data of the WT RNA-sequencing was obtained from the NCBI with an accession number of PRJNA1140836 (Li et al., 2024).

The CAS liquid assay was performed to calculate the production of siderophore as previously described (Li et al., 2024). In brief, different V. anguillarum 775 cells were grown in MM9 medium (0.3 g KH₂PO₄, 1 g NaCl, 1 g NH₄Cl per L) in the presence of 0.2% casamino acid and 100 mM PIPES overnight at 26°C. After centrifugation, 500 μL of supernatant mixed with 500 μL of CAS assay solution (150 μM CAS, 15 μM FeCl₃, 0.6 mM HDTMA, 500 mM piperazine buffer), 10 μL of 0.2 M 5-sulfosalicylic acid was supplemented, and the mixture was incubated for 5 min at room temperature. When siderophore is present, the siderophore could remove iron from the complex and lead to a decrease in the blue color of the mixture. The absorbance at 630 nm (A₆₃₀) was examined. Siderophore units were calculated as follows:

A_r indicates the absorbance of the MM9 medium plus the CAS assay solution plus 5-sulfosalicylic acid; A_s indicates the absorbance of the tested sample. Three independent biological experiments were carried out, and values were obtained from representative experiments in triplicate.

M9 swimming plates (M9 high salt broth with 0.3% agar supplemented with 50 μM 2, 2’-dipyridine or 50 μM FeCl₃) were prepared and air-dried overnight at room temperature. Exponential phase V. anguillarum 775 cells were inoculated with a sterile toothpick on the plates at 26°C for 24 h, and the diameter of the “colony” was measured. Three independent experiments were performed, and values were calculated from representative experiments in triplicate.

The intracellular ROS levels were measured as described by Pasqua et al (Pasqua et al., 2017). with slight modifications. In brief, different V. anguillarum 775 strains were grown under different conditions to reach the exponential phase. Then, bacteria from 1 mL of the culture were harvested by centrifugation at 3,500 g at 4°C for 15 min. After being washed twice with phosphate-buffered saline (PBS), the cells were resuspended in PBS to reach an OD_{600 nm} value of 2.0, and 10 μM of DMSO-diluted H₂DCFDA was supplemented. The mixture was incubated in the dark for 60 min at 30°C and washed twice with 1 mL of PBS. Finally, cells were resuspended in 200 μL of PBS for the fluorescence measurement (excitation: 485 nm; emission: 535 nm) using an Infinite 200 PRO microplate reader (TECAN, Switzerland). The assays were performed in three independent experiments, and values were obtained from representative experiments in triplicate.

The infection assays were carried out on turbot as previously described (Li et al., 2024). In brief, turbot larvae (~20 g per fish) were intraperitoneally injected with 100 μL of a bacterial suspension (~1,000 CFU), and 100 μL of PBS was used as a control. Before injection, different V. anguillarum 775 strains were grown under limited iron conditions at 26°C to exponential phase, washed twice with PBS, and diluted in PBS. After bacterial injection, the turbots were incubated in fresh, filtered seawater at 20°C and observed daily for dead fish. Then gut bacteria of the dead fish were isolated, and mortalities were considered to result from V. anguillarum 775 strains only when the V. anguillarum 775 strain was found in pure culture (Crosa et al., 1977). Virulence was calculated by recording the number of survivors for 10 days post-injection. The assay was performed in three independent biological experiments. To examine the bacterial survival in turbot larvae, after 20 h of infection, spleens and livers were aseptically collected in PBS. After dilution, the supernatant was plated on M9 high salt plates, and the number of bacteria in the liver and the spleen was counted, which was shown as CFU/g. These turbot experiments were performed in accordance with the ethical guidelines of Shandong University.

If not specified, the Student t test (two-tailed) was used for statistical analysis in Microsoft Excel (Office 2021; Microsoft, Redmond, WA, USA). The statistical analysis of the survival rate was performed using the paired t test in GraphPad Prism 8.0.2 (GraphPad, San Diego, California).

A homolog of the ryhB gene, named VaryhB, was identified in the genome of V. anguillarum 775. As shown in Figure 1A , the sRNA VaryhB gene contains 228 bp and is located in the 734-bp intergenic region between VAA_RS01400, encoding a delta-aminolevulinic acid dehydratase (ALAD), and VAA_RS01405, encoding a YihA family ribosome biogenesis GTP-binding protein. Sequence alignment showed that VaryhB is relatively conserved with that in V. cholerae. Compared to E. coli RyhB (EcRyhB), RyhB sRNAs from V. anguillarum and V. cholerae are much longer in the 5’ region, whereas these sRNAs harbor a conserved central region with the E. coli EcRyhB ( Figure 1B ). The additional 5’ region in the V. cholerae RyhB (VcRyhB) is proposed to be responsible for the stability of the sRNA structure (Mey et al., 2005). The longer RyhB sRNAs appear to be the major form among the Vibrionaceae (Davis et al., 2005; Mey et al., 2005). The longest RyhB is from Vibrio parahaemolyticus, with a length of 233 bp. However, it is unclear why longer RyhB sRNAs occur in the Vibrionaceae.

To understand the function of VaRyhB on iron homeostasis, a ΔVaryhB mutant was constructed, and RNA-seq-based transcription analysis was performed under different iron conditions. Compared to the WT cells (639 regulated genes) (Li et al., 2024), much fewer genes (203 genes) were regulated by iron in the ΔVaryhB mutant ( Figure 2A ): 139 genes had increased expression and 64 genes had decreased expression under limited iron conditions (p < 0.05 and |log₂ fold change| ≥ 1, Figure 2C ). However, compared to the ΔVafur mutant (119 regulated genes) (Li et al., 2024), more regulated genes were observed in the ΔVaryhB mutant ( Figure 2B ). Most significantly up-regulated genes under limited iron conditions are associated with iron uptake ( Figure 2C ), which is similar to that in the WT. Clusters of orthologous groups of proteins (COG) analysis suggested that fewer categories were regulated by iron in the ΔVaryhB mutant compared to those in the WT. For example, compared to the WT ( Supplementary Figure S1 ), flagellar assembly, sulfur relay system, and the TCA cycle were not differently expressed in the ΔVaryhB mutant under different iron conditions ( Figure 2D ). However, different regulation modes occurred among these pathways. Compared to the WT, genes involved in the TCA cycle in the ΔVaryhB mutant exhibited higher expression levels mainly under rich iron conditions; genes involved in the Fe-S assembly showed higher expression levels under both rich and limited iron conditions; while genes for chemotaxis and motility displayed lower expression levels under both rich and limited iron conditions ( Table 1 ). Similar RyhB-mediated expression modes were observed in some microorganisms, such as E. coli (Massé and Gottesman, 2002; Desnoyers et al., 2009) and V. cholerae (Mey et al., 2005). Genes involved in denitrification were up-regulated by VaRyhB, and when VaryhB was absent, their expression levels were significantly decreased under rich and limited iron conditions. This is different from the expression mode of the nap operon in E. coli (Wang et al., 2015), which was down-regulated by RyhB. Most genes for iron uptake displayed similar regulation modes in the ΔVaryhB mutant and the WT strain and showed higher expression levels when iron was depleted. Several genes, including angC, angE, and fatA, appear to be activated by VaRyhB, and reduced expression levels were observed when VaryhB was deleted ( Table 1 ). This activation may be important for bacterial growth in the host and thereby for pathogenesis. Consistent with this, qRT-PCR data suggested that the expression levels of angC, angE, and fatA were significantly reduced compared to those in the WT and the ΔVafur mutant ( Figure 2E ). Differently, angT for iron release from the ferric-anguibactin complex and exbB1 and tonB1 for energy transmission for anguibactin and heme uptake were negatively regulated by VaRyhB. The expression mode of the VaryhB gene was also investigated by qRT-PCR. As shown in Figure 2F , the expression of the VaryhB gene was significantly repressed by iron in the WT. In the ΔVafur mutant, the level of inhibition in response to rich iron conditions obviously became lower, indicating that the expression of the VaryhB gene is regulated by iron and VaFur. Two different regulation modes occurred for genes involved in oxidative stress. In addition, genes responsible for T6SS were negatively regulated by VaRyhB, whereas genes for molybdenum cofactor (Moco) biosynthesis were positively regulated by VaRyhB. These two pathways of T6SS and Moco biosynthesis were first observed to be regulated by RyhB. Taken together, VaRyhB is involved in the regulation of genes involved in the TCA cycle, denitrification, Fe-S assembly, anguibactin-mediated iron uptake, oxidative resistance, and some virulence factors (motility and chemotaxis, and T6SS).

Since the abnormal iron regulation in the ΔVaryhB mutant may cause impaired growth, the growth of the ΔVaryhB mutant strain was first examined under different iron conditions. When grown under iron-rich conditions, the ΔVaryhB mutant exhibited slightly decreased growth ( Figure 3 ), similar to that of the ΔVafur mutant. However, different from that of the ΔVafur mutant (an increased growth yield), when grown under iron-poor conditions, the ΔVaryhB mutant strain showed attenuated growth. The impaired growth under limited iron conditions may be caused by reduced expression of genes for siderophore synthesis and transport ( Table 1 ). In line with this, the siderophore content was slightly reduced in the ΔVaryhB mutant compared to that in the WT ( Figure 4A ). This abnormal growth in both poor and rich iron conditions was also observed in the E. coli ΔryhB mutant (Jacques et al., 2006). The complementation of the ΔVaryhB mutant restored the WT-like growth in the presence of 50 μM 2, 2’-dipyridine ( Supplementary Figure S2 ). Our data indicated that worse multiplication likely occurs during the ΔVaryhB infection, thereby leading to reduced virulence in the ΔVaryhB mutant strain.

Since genes for motility in the ΔVaryhB mutant exhibited reduced expression levels, the swimming motility was further tested under different iron conditions. As shown in Figure 4B , no iron-dependent regulation occurred when VaryhB was absent, and the capability of swimming motility in the ΔVaryhB mutant under rich and limited iron conditions was similar to that of the WT under limited iron conditions. Therefore, VaRyhB is essential to iron-regulated swimming behavior. Notably, the swimming motility of the VaryhB mutant was much greater than that of the Vafur mutant under limited and rich iron conditions (Li et al., 2024), although a similar phenotype of no iron-dependent regulation was observed in the two mutants.

In many organisms, such as E. coli, the expression of the superoxide dismutase gene sodB was indirectly regulated by Fur through the RyhB (Massé and Gottesman, 2002; Troxell and Hassan, 2013). On the other hand, in V. anguillarum 775, the ΔVafur mutant is much more sensitive to hydrogen peroxide than the WT strain (Li et al., 2024). Therefore, the impaired oxidative resistance in the ΔVafur mutant may be caused by VaRyhB. To verify this, the sensitivity of the ΔVaryhB mutant to H₂O₂ was examined. As shown in Figure 5A , the presence of H₂O₂ did not affect the growth of ΔVaryhB mutant under rich iron conditions. In addition, the intercellular ROS levels were also tested using the ROS reactive probe H₂DCFDA (Pasqua et al., 2017). In line with the growth in the presence of H₂O₂, the intracellular ROS levels in the ΔVaryhB mutant were similar to those of the WT in media with both 50 μM and 100 μM FeCl₃ ( Figures 5B, D ). Moreover, a ΔVafurΔVaryhB double deletion mutant showed similar growth and ROS levels with WT and ΔVaryhB mutant ( Figures 5B–D ), further demonstrating that the defective oxidative resistance in VaFur-depleted cells results from the activated VaRyhB, the expression of which is repressed by VaFur under rich iron conditions in the WT. Therefore, VaFur in V. anguillarum 775 adopts a “RyhB-dependent” mechanism to defend against oxidative stress.

RyhB in many pathogens is controlled by Fur and mainly functions under poor iron conditions (Chareyre and Mandin, 2018). On the other hand, a significantly reduced pathogenicity of the ΔVafur mutant was observed in V. anguillarum 775 (Li et al., 2024). It was proposed that the decreased pathogenicity of the ΔVafur mutant may be indirectly caused by RyhB. Therefore, to uncover the role of VaRyhB in the pathogenesis, turbot larvae were intraperitoneally challenged using the WT strain and the ΔVaryhB mutant, respectively. Survival was observed up to 10 days post-infection, and no turbot larvae died in the control group. In the WT infection group, 90% of turbot larvae died after 4 days ( Figure 6A ), whereas deletion of VaryhB resulted in a 20% reduction in the lethality rate compared to that of the WT infection group. When infected by the ΔVaryhB complementation strain, although the pathogenicity was greater than that of the ΔVaryhB mutant, the final lethality rate was lower than that of the WT strain. This may be caused by the loss of the complementation plasmid pLYJ232 in the host. In agreement with this, the ΔVaryhB population in the liver and spleen of turbot larvae was also decreased compared to that of the WT strain ( Figure 6B ). The complementation plasmid, pLYJ232, could restore bacterial numbers in the liver and spleen of turbot larvae back to the WT-like level. These data indicated that VaRyhB in V. anguillarum 775 is required for pathogenicity. However, compared to the ΔVafur mutant, which showed a higher than 50% survival rate (Li et al., 2024), VaRyhB has less effect on the V. anguillarum pathogenicity. These indicated that in the ΔVafur mutant, the reduced pathogenicity may not be mostly caused by irregulated VaRyhB.