Riemerella anatipestifer AS87_RS09170 gene is responsible for biotin synthesis, bacterial morphology and virulence

Riemerella anatipestifer is a bacterial pathogen responsible for major economic losses within the duck industry. Recent studies have revealed that biotin biosynthesis is critical for the bacterium’s survival and virulence. We previously found that R. anatipestifer AS87_RS09170, a putative bioF gene, is important for bacterial virulence. In the present study, we characterized the AS87_RS09170 gene in R. anatipestifer strain Yb2. Sequence analysis indicated that the AS87_RS09170 gene is highly conserved among R. anatipestifer strains; the deduced protein harbored the conserved pyridoxal 5′-phosphate binding pocket of 8-amino-7-oxononanoate synthase. Western blot analysis demonstrated that the biotin-dependent enzyme was present in smaller quantities in the mutant strain Yb2ΔbioF compared to that of the wide-type strain Yb2, suggesting that the biotin biosynthesis was defective. The mutant strain Yb2ΔbioF displayed a decreased growth rate at the exponential phase in tryptic soy broth culture and in BeaverBeads Streptavidin treated tryptic soy broth culture, but recovered when biotin was supplemented. In addition, the mutant strain Yb2ΔbioF showed an enhanced biofilm formation, as well as increased adhesion and invasion capacities to duck embryo fibroblasts. Moreover, the mutant strain Yb2ΔbioF exhibited irregular shapes with budding vegetations and relatively thickened cell walls under scanning and transmission electron microscope observation, as well as a reduced capacity to establish systemic infection in a duck infection model. These results provide the first evidence that the R. anatipestifer AS87_RS09170 gene is responsible for biotin synthesis, bacterial morphology and virulence.

cofactor for the biotin-dependent enzymes that are involved in important metabolic pathways such as membrane lipid synthesis, replenishment of the tricarboxylic acid cycle and amino acid metabolism [12][13][14] . Biotin can be synthesized de novo in microorganisms, plants, and fungi. Biotin biosynthesis can be divided into two stages: (1) synthesis of the pimelate precursor and, (2) assemblage of the bicyclic rings of biotin. The final four steps are highly conserved amongst microorganisms and plants. Pimeloyl-CoA is converted to biotin by the activities of AONS, 7,8-diaminopelargonic acid aminotransferase, dethiobiotin synthetase and biotin synthase, which are encoded by bioF, bioA, bioD, and bioB, respectively (supplementary Figure S1). Biotin is intimately associated with lipid synthesis where the products form key components of the mycobacterial cell membrane that are critical for bacterial survival and pathogenesis 15 .
In the present study, we described a mutant strain of R. anatipestifer Yb2ΔbioF in which the homologous bioF gene AS87_RS09170 was disrupted. The bacterial growth, protein biotinylation, biofilm formation, adherence and invasion capabilities, morphology, gene expression as well as colonization and development during infection of the mutant strain Yb2ΔbioF were characterized.
Characterization of the mutant strain Yb2ΔbioF. The mutant strain Yb2ΔbioF was viable when grown in tryptic soy broth (TSB) medium, and reached a similar stationary phase in TSB compared with the wild-type (WT) strain Yb2. However, the growth rate of the mutant was significantly reduced in the logarithmic phase between 8 and 12 h. Transformation of the mutant strain Yb2ΔbioF with the AS87_RS09170 gene complemented the growth defect in TSB medium ( Fig. 2A). To further access the effect of biotin biosynthesis on R. anatipestifer growth, we assessed the growth capacity of the mutant strain Yb2ΔbioF in either TSB, the BeaverBeads Streptavidin treated TSB (designated as TSB-biotin) or the BeaverBeads Streptavidin treated TSB replete with biotin (designated as TSB-biotin+biotin). As shown in Fig. 2B, the growth deficiency of the mutant strain Yb2ΔbioF in the biotin depleted TSB was restored when biotin was supplemented at a final concentration of 1.0 μg/ml.
We also assessed the impact of biotin deficiency on protein biotinylation by anti-biotin immunoblotting. As shown in Fig. 2C, a deficiency of immunoreactivity in two bands corresponding with two proteins was detected in mutant strain Yb2ΔbioF, indicating that the quantity of biotinylated protein decreased. The electrophoretic mobility and anti-biotin immunoreactivity of these proteins suggested that they correspond to two isoforms of acetyl-CoA carboxylase, which are predicted to be biotinylated in R. anatipestifer. Liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) analysis of the BeaverBeads Streptavidin enriched protein further confirmed the anti-biotin immunoreactive protein was acetyl-CoA carboxylase. The levels of biotinylated acetyl-CoA carboxylase were recovered in the complemented strain cYb2ΔbioF (Fig. 2C).
Percoll density gradient centrifugation analyses showed that the density of Yb2ΔbioF was slightly changed compared to its WT strain Yb2 (Fig. 2D). Yb2 cells settled at the 40-50% interface, whereas the Yb2ΔbioF mutant cells settled at the 60% interface. In addition, the appearance was more diffuse for the WT strain Yb2, compared to the mutant strain Yb2ΔbioF.
The mutant strain Yb2ΔbioF presented an increased biofilm formation. The ability of the mutant strain Yb2ΔbioF to form a biofilm on borosilicate glass surface was examined by fluorescence microscopy. As shown in Fig. 3 with Live/dead BacLight Bacterial Viability staining, the mutant strain Yb2ΔbioF presented a significantly increased abundance of live cells, as compared with the WT strain Yb2 and complemented strain cYb2ΔbioF at 24 h incubation. At 48 h incubation, the mutant strain Yb2ΔbioF produced a mature biofilm, which was structured with numerous microcolonies encased in a thick opaque extracellular matrix, whereas the WT and complemented strain cYb2ΔbioF failed to form a layer of singly attached biofilm on the surface of borosilicate glass.
Inactivation of the AS87_RS09170 gene increased the bacterial adherence and invasion abilities. To  on duck embryo fibroblast cells. When infected at 100 MOI, the host cell-associated Yb2ΔbioF bacteria were counted as 65,600 ± 8,015 colony forming units (CFU)/well, which was significantly increased in comparison with that of the WT strain Yb2 (35,933 ± 10,277 CFU/well). After an additional 1 h of incubation with gentamicin, the invaded Yb2ΔbioF bacteria were counted as 11,293 ± 2,367 CFU/well, approximately 3-fold higher than that of the WT strain Yb2 (4,360 ± 457 CFU/well). The results showed that adherence and invasion capacities of Yb2ΔbioF were significantly increased, compared with those of the WT strain. The complemented strain cYb2ΔbioF exhibited WT levels of adherence and invasion (Fig. 4).
Morphology and ultrastructure observation. The morphology of the R. anatipestifer cells was observed under scanning microscopy. As shown in Fig. 5, WT cells were regular rod-shaped with round ends, consistent with a previous report 16 . The mutant strain Yb2ΔbioF cells remained bacilliform, but with irregular shapes and multiple budding vegetations. There were no obvious perforations in the cellular surface. The complemented strain cYb2ΔbioF recovered the morphology and surface features.
The cellular ultrastructure was observed under a transmission electron microscope (Fig. 6). WT Yb2 cells were integral with clear structures, and few were undergoing division. The mutant strain Yb2ΔbioF cells presented as irregular shapes with budding vegetations, and relatively thickened cell walls. The numbers of cells undergoing division were increased compared with the WT strain Yb2.
Extraction and identification of membrane proteins. Membrane proteins were extracted from the WT strain Yb2, mutant strain Yb2ΔbioF and complemented strain cYb2ΔbioF, followed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis. As shown in Fig. 7, a 70-kDa band was significantly displayed in the purified membrane proteins from mutant strain Yb2ΔbioF, but was absent in both the WT strain Yb2 and complemented strain cYb2ΔbioF. The band corresponding to 70-kDa was excised from gel and analyzed by LC-ESI-MS/MS, and identified proteins are summarized in supplementary Table S2. Among the 133 proteins analysed, 93 (70%), 12 (9%), 11 (8%), four (3%) and 13 (10%) were predicted to be cytoplasmic, cytoplasmic membrane, outer membrane, periplasmic, and of unknown locations, respectively. Determination of the differentially expressed genes. Strand-specific Illumina RNA-Seq analysis was used to investigate the differentially expressed genes between the WT strain Yb2 and mutant strain Yb2ΔbioF. In total, 13 and 26 genes were up-regulated and down-regulated in the mutant, respectively, in comparison to the WT strain. Quantitative polymerase chain reaction (qPCR) further confirmed that transcription of ten genes in Yb2ΔbioF were down-regulated by over 2-fold (Table 1). Of these, the proteins encoded  by AS87_RS01370, AS87_RS01365, AS87_RS01360, AS87_RS01355 and AS87_RS01350 were annotated as vault protein inter-alpha-trypsin (VIT) family protein, sulfite exporter TauE/SafE family protein, efflux efflux resistance-nodulation-division (RND) transporter periplasmic adaptor subunit, TolC family protein and CusA/ CzcA family heavy metal efflux RND transporter, respectively, indicating that the AS87_RS09170 gene disruption affected expression of genes primarily responsible for transmembrane transport.

Determination of the bacterial virulence.
Our previous study revealed that disruption of the AS87_ RS09170 gene resulted in a 768,000-fold attenuation of virulence, compared with the WT strain Yb2 17 . To further investigate the role of the AS87_RS09170 gene in systemic invasion and dissemination, the bacterial loading in blood of infected ducks was quantified. Results showed that bacterial recovery of the WT strain Yb2 maintained an increase of up to 1.82 × 10 6 CFU/ml at 36 hpi. In contrast, the recovery of the mutant strain Yb2ΔbioF was gradually decreased post infection, and was measured as 270 CFU/ml, 213 CFU/ml, and 124 CFU/ml at 12, 24, and 36 hpi, respectively. The mutant strain was significantly attenuated (Fig. 8).

Discussion
Biotin, an indispensable nutrient found in all living cells, is synthesized de novo in many microorganisms, plants, and a few fungi. The early steps of the pathway are responsible for the synthesis of pimelic acid, whereas the last four steps are responsible for assembly of the rings. R. anatipestifer possesses all of the putative proteins (BioF, BioA, BioD, BioB) of the late steps of biotin biosynthesis. In contrast, R. anatipestifer lacks homologous genes  of bioCHIW required for pimelate synthesis, suggested that pimeloyl-CoA could be formed by the enzymes of fatty acid synthesis, which is consistent with the findings of Steven Lin 18 . Here, we characterized a mutant strain Yb2ΔbioF in which the putative bioF gene responsible for biotin biosynthesis was inactivated. We screened a library of random transposon mutants in an animal infection model and obtained an attenuated mutant, Yb2ΔbioF, with the AS87_RS09170 gene disrupted by the Tn4351 insertion in the open reading frame 17 . This mutant showed a decreased growth rate in TSB at the exponential phase ( Fig. 2A) and reduced quantities of biotinylated proteins. The growth of the mutant strain Yb2ΔbioF in biotin-depleted TSB was almost halted, but recovered with addition of biotin (Fig. 2B). Subsequently, we showed that the biotin deficiency resulted in an increase in biofilm formation (Fig. 3), as well as increased capacity for adhesion and invasion to duck embryo fibroblast (Fig. 4). Our data showed that biotin deficiency also led to changes of cell morphology (Fig. 5), resulting in irregular shapes with budding vegetations and relatively thickened cell walls (Fig. 6). Genetic complementation with the AS87_RS09170 gene restored the growth rate, protein biotinylation, biofilm formation and adhesion/ invasion capacities, as well as cell morphology, confirming that deficient biotin synthesis is responsible for the observed phenotypes. Biotin acts as a carbon dioxide carrier during carboxylation, which comprises decarboxylation and transcarboxylation reactions responsible for biosynthesis of fatty acids, gluconeogenesis and amino acid metabolism 19 . Based on the genomic annotation analysis, R. anatipestifer genome has two putative biotinylated proteins of acetyl-CoA carboxylase biotin carboxyl carrier protein (AS87_RS06990) and acetyl-CoA carboxylase biotin carboxylase subunit (AS87_RS06995). Our results confirmed that two biotinylated bands at 18 kDa and 24 kDa, which correspond to the sizes of above two proteins, were clearly reduced in the mutant strain Yb2ΔbioF in comparison with the WT strain Yb2 (Fig. 2C). LC-ESI-MS/MS further confirmed that the 24 kDa protein was acetyl-CoA carboxylase, but the 18 kDa protein was not determined due to its lower abundance, suggesting that the AS87_RS09170 gene product participates in biotin biosynthesis and protein biotinylation. Alterations in biotinylated protein levels caused by biotin biosynthesis deficiency were also observed in Mycobacterium tuberculosis and Arabidopsis thaliana 13,20 .
Cell surface adhesion and cell aggregation initiate bacterial biofilm formation 21 . Our results demonstrated that AS87_RS09170 disruption significantly increased the formation of solid biofilms, suggesting that the cell-cell and cell surface interactions of the mutant strain Yb2ΔbioF were increased. The fact that absence of the AS87_ RS09170 protein induced rather than repressed aggregation indicates that AS87_RS09170 protein itself is not directly involved in cell-cell adhesion. In addition, the hydrophobic experiment revealed that disruption of the AS87_RS09170 gene did not affect bacterial hydrophobicity (data not shown), indicating that it is unlikely that the biofilm formation was mediated by bacterial cell surface hydrophobic properties. Moreover, the capacities of mutant strain Yb2ΔbioF to adhere to and invade duck embryo fibroblast cells were sharply increased compared those of the WT strain Yb2, which is consistent with the observation that the mutant strain increased the liquid-solid interface biofilm formation. Further investigation revealed that the dramatic change of the R. anatipestifer cell surface morphology from regular to irregular with budding vegetations was caused by AS87_RS09170 inactivation. Therefore, it is most likely that the AS87_RS09170 protein modulated bacterial biofilm formation and adherence/invasion capacities by affecting the cell surface morphology.
In addition to acting as a coenzyme for carboxylases, biotin also plays a unique role in regulation of genes expression 22 . qPCR analysis revealed that deleting the AS87_RS09170 gene down-regulated ten genes, five of which encoded efflux RND transporter-associated proteins. The differential expression of these genes may critically change bacterial cell surface properties. Further membrane protein analysis revealed the presence of many cytoplasmic proteins in the mutant strain Yb2ΔbioF, suggesting they may have leaked from cytoplasm, and the differentially expressed genes associated with cellular transport played a role in this leakage. Therefore, the cytoplasmic protein leakage in mutant strain Yb2ΔbioF might result in budding vegetations on the bacterial surface. Considered together, these results suggest that the AS87_RS09170 gene is responsible for the biotin synthesis and gene expression, which caused the defects in bacterial cell surface structures, led to leakage of cytoplasmic proteins and consequent increased biofilm formation, bacterial adherence and invasion.
The manifestation of R. anatipestifer infection is characterized as septicemia, and virulence and pathogenesis of this species are tightly associated with the ability of the bacteria to colonize and develop in the host. We found that the mutant strain Yb2ΔbioF failed to colonize and multiply in ducks, evident by decreased bacterial loadings in the blood of Yb2ΔbioF-infected birds. In addition, the mutant strain Yb2ΔbioF exhibited 768,000-fold attenuated virulence. Our results indicated that the biotin synthesis plays a key role in the survival of R. anatipestifer during the infection. A previous study revealed that biotin biosynthesis of Mycobacterium marinum plays an important role in colonization and development of the host 23 .
In summary, the present study demonstrated that the R. anatipestifer AS87_RS09170 gene is responsible for biotin synthesis, which plays important roles in bacterial growth, protein biotinylation and establishment of Bacterial strains, plasmids and culture conditions. The bacterial strains, plasmids and primers used in the present study are listed in Table 2. R. anatipestifer Yb2 is the serotype 2 WT virulent strain, and the mutant strain RA1893 (Yb2ΔbioF in the present study) was derived from this strain by Tn4351 transposon insertion 17 . R. anatipestifer strains were grown on tryptic soy agar (TSA, Difco, NJ, USA) at 37 °C for 24 h in 5% CO 2 or in tryptic soy broth (TSB, Difco) at 37 °C with shaking at 200 rpm for 8 to 12 h. To remove the biotin from TSB, 50 ml TSB was mixed with 300 μl of BeaverBeads Streptavidin (BeaverBio, Suzhou, China) and incubated with shaking at 4 °C for 12 h. After the beads were removed, the TSB medium was passed through 0.22 μm filter, subsequently designated as TSB-biotin. Escherichia coli strains were grown at 37 °C on Luria-Bertani (LB) plates or in LB broth. Antibiotics were added at the appropriate concentrations when required: ampicillin (100 μg/ml), erythromycin (0.5 μg/ml), kanamycin (50 μg/ml) and cefoxitin (5 μg/ml).
The shuttle plasmid pCP29 was used for complementation of mutant strain Yb2ΔbioF 24 . The AS87_RS09170 open reading frame was amplified from the WT strain Yb2 using primers bioF comp-F/bioF comp-R. The PCR product was inserted into pCP29 at XhoI and SphI restriction sites, producing recombinant plasmid pCP29-bioF. Expression of the AS87_RS09170 gene was under the control of the ompA promoter, as described previously 25 . Plasmids were first introduced into E. coli S17-1 by transformation. Next, they were transferred into mutant strain Yb2ΔbioF by conjugation. Transformants were selected on TSA containing 5 μg/ml cefoxitin and 50 μg/ml kanamycin, and identified by PCR amplification using primers bioF comp-F/bioF comp-R and RA 16S rRNA-F/RA 16S rRNA-R. The complemented strain is subsequently designated as cYb2ΔbioF.
Analyses of AS87_RS09170 gene distribution in R. anatipestifer strains and deduced amino acid sequence. Genomic DNA of 25 R. anatipestifer strains with different serotypes was isolated using the TIANamp Bacteria DNA kit (Tiangen, Beijing, China) according to the manufacturer's instruction. The AS87_RS09170 gene in R. anatipestifer strains was amplified using primers bioF comp-F and bioF comp-R, followed by agarose gel electrophoresis. The crystallized AONS sequences [Paraburkholderia xenovorans (pdb5JAY), Francisella tularensis (pdb4IW7), E. coli (pdb1DJE), Burkholderia multivorans (pdb5VNX), and Mycobacterium smegmatis (pdb3WY7)] were retrieved from Universal Protein Resource (UniProt) (http://www.uniprot.org/). Similarity of the predicted R. anatipestifer AONS sequence and other crystallized AONS sequences was analyzed using the Clustal W algorithm in the MegAlign program from the DNASTAR Lasergene suite, and rendered with Espript 26 .
Bacterial growth curves. The WT strain Yb2, mutant strain Yb2ΔbioF and complemented strain cYb2ΔbioF were cultured in TSB at 37 °C with shaking, and the bacterial growth was measured at an optical  density at 600 nm (OD 600 ) as described previously 27 . Equal amounts of each bacterial culture were then transferred into fresh TSB medium at a ratio of 1:100 (v/v) and incubated at 37 °C, with shaking at 200 rpm. The OD 600 value was measured at 2 h intervals for 16 h using a spectrophotometer (BIO-RAD, USA). To further evaluate the effect of biotin on bacterial growth, the growth of the mutant strain Yb2ΔbioF in TSB, TSB-biotin and TSB-biotin with addition of biotin at final concentrations of 1.0 μg/ml were measured as described above.

Characterization of biotinylated protein.
To detect the biotinylated proteins in the bacteria, the whole-cell proteins of the WT strain Yb2, mutant strain Yb2ΔbioF and complemented strain cYb2ΔbioF were extracted using a bacterial total proteins extraction kit (BestBio, Shanghai, China), subjected to SDS-PAGE and transferred into a nitrocellulose membrane (Whatman, Sigma-Aldrich). The membrane was blocked for 1 h at room temperature in phosphate-buffered saline (PBS, pH7.2) containing 1% bovine serum albumin (BSA), rinsed with PBST (PBS containing 0.05% Tween 20), and then incubated with Ultrasensitive Streptavidin-Peroxidase polymer (Sigma-Aldrich) diluted 1:1000 in PBS with 0.05% Tween 20 at room temperature for 1 h. After washing six times with PBST, the membrane was incubated with a basic ECL subsolution kit (share-bio, Shanghai, China) and blots were visualized using the Chemiluminescent Imaging System (Tanon, Shanghai, China). The mouse anti-GroEL monoclonal antibody (prepared in our laboratory previously 28  Briefly, a stock solution of Percoll (Sigma Aldrich, St. Louis, MO, USA) was prepared by dilution with 1.5 mol NaCl at a ratio of 9:1(v/v). Solutions containing 80%, 70%, 60%, 50%, 40%, 30%, 20% and 10% Percoll in 0.15 mol NaCl were further prepared from the stock solution. A 1.2-ml volume of each of these solutions was carefully layered into a 13.2 ml polycarbonate tube to produce a step gradient with 80% Percoll at the bottom and the 10% Percoll at the top. One milliliter of each bacterial culture (OD 600 = 2) suspended in 0.15 mol NaCl was applied to the top of the 10% layer and the gradient was then centrifuged for 1 h at 10,000 × g at 4 °C in a SW41Ti rotor using a Beckman centrifuge (Optima L-100XP, Beckman Coulter, Inc. CA, USA). The gradient was visualized using fuchsine stained layers as the marker (M).

Biofilm formation assay. R. anatipestifer biofilm formation on borosilicate glass was measured by Live/dead
BacLight Bacterial Viability staining as described previously 27 . Briefly, each bacterial culture in mid-exponential phase was adjusted to OD 600 = 0.1 with TSB, and 1 ml of the bacterial suspension was transferred into 24-well polystyrene microtiter plates (Corning, NY, USA) containing sterile glass coverslips. The plates were incubated at 37 °C under an atmosphere of 5% CO 2 and the coverslips were collected at 24 h and 48 h, respectively. After rinsing gently three times with sterile 0.01 mol PBS (pH7.2), the coverslips were stained with 100 μl of Live/ dead BacLight Bacterial Viability staining reagent (Thermo Fisher Scientific, Waltham, MA, USA) for 15 min as according to the manufacture's protocol, and examined using a fluorescence microscope (Nikon Eclipse 80i, Japan) in the dark at room temperature. The image profiles of bacterial shapes were visualized and analyzed using NIS-Elements Viewer software. Dehydration was accomplished by a graded series of acetone. For scanning electron microscopy, the samples were critical-point dried and sputter coated with a thin layer of gold. Photographs were taken using a Tecnai G2 F30 scanning electron microscope (FEI, Hillsboro, OR, USA). For transmission electron microscopy, the samples were fixed, post-fixed and dehydrated under the same conditions described above. Epoxy resin was then infiltrated into the fixed and dehydrated samples and polymerized into a plastic block. Subsequently, the block was sliced into ultra-thin sections and stained with both uranyl acetate and lead citrate. The specimens were observed on a Tecnai 12 transmission electron microscope (FEI, Eindhoven, Netherlands) at 80 kV and images were recorded using Ditabis imaging plates.

Adhesion and invasion
Extraction and identification of membrane proteins. Membrane proteins were isolated from R. anatipestifer WT strain Yb2, mutant strain Yb2ΔbioF and complemented strain cYb2ΔbioF, respectively, using a bacterial membrane protein extraction kit (BestBio, Shanghai, China) according to the manufacturer's protocol. Briefly, R. anatipestifer strains were cultured respectively to logarithmic phase (OD 600 = 1.5), the bacterial pellets were collected and washed twice by centrifugation. The extract buffer was then prepared and added to the bacterial pellets, the mixture was stirred for 2 h at 4 °C to lyse bacteria. The supernatant was collected by centrifugation for 15 min at 4 °C and incubated at 37 °C in water-bath for 30 min for stratification. After the upper layer liquid was removed, menbrane protein dissolution buffer was added to make the membrane proteins. The proteins were separated by SDS-PAGE and stained with Coomassie blue. The differentially expressed protein band in the mutant strain Yb2ΔbioF was excised manually from the gel and subjected to LC-ESI-MS/MS analysis in Shanghai Applied Protein Technology Co. Ltd (Shanghai, China) as described above, and searched for using the online software PSORTb version 3.0.2 to predict the subcellular location of each protein 30 .
Illumina sequencing for RNA-Seq and differential expression analysis. The WT strain Yb2, mutant strain Yb2ΔbioF and complemented strain cYb2ΔbioF were cultured in TSB medium at 37 °C with shaking for 12 h. The total RNA was extracted with TRIzol regent (Invitrogen) according to the manufacturer's instructions. Total RNA quantity and quality were assessed by Agilent 2100 Bioanalyzer (Agilent RNA 600 Nano kit) and ribosome RNAs were depleted using Ribo-Zero Magnetic Gold Kit (epicenter, USA). The illumina RNA-Seq libraries were generated and then validated by the Agilent 2100 bioanalyzer instrument (Agilent DNA 1000 Reagents) and real-time quantitative PCR (qPCR) (TaqMan Probe). The qualified libraries were amplified on cBot to generate the cluster on the flowcell (TruSeq PE Cluster Kit V3-cBot-HS, Illumina), and the amplified flowcell was sequenced pair end on the HiSeq 2000 System (TruSeq SBS KIT-HS V3, Illumina) 31 . Low-quality reads and adaptors were removed from raw reads. Cleaned reads were aligned to the R. anatipestifer Yb2 genome using RNA Sequel software HISAT (Version 2.0.1-beta) 32,33 . Transcript levels were calculated as RPKM (Reads per kilobase cDNA per million fragments mapped) using RSEM software (version 1.2.12) 34 . Differentially expressed genes were analyzed using possionDis with fold change (cutoff = 2.0) 35,36 , and considered statistically significant if the fold change was >2.0 and the FDR (False Discovery Rate) was <0.05.
Real-time quantitative PCR analysis. qPCR was performed to confirm transcriptional levels of differentially expressed genes obtained in the RNA-Seq analysis. Gene-specific primers were designed using primer3 online software Version.0.4.0 37 and are described in supplementary Table S1. The expression of the L-lactate dehydrogenase encoding gene (ldh) was measured using primers RA ldh-F/RA ldh-R, and used as an internal control. Total RNA was isolated from the WT strain Yb2 and mutant strain Yb2ΔbioF using Trizol reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's instructions. All RNA samples were treated with the TURBO DNA-free kit (Ambion, Grand Island, NY, USA) to remove DNA contamination. cDNA was synthesized using PrimeScript RT Master Mix (Takara Bacterial virulence determination. The bacterial loadings in the blood of infected ducks were counted to evaluate bacterial survival in vivo 38 . Eighteen-day-old ducks (six ducks per group) were inoculated intramuscularly with the WT strain Yb2 and mutant strain Yb2ΔbioF at 2.5 × 10 8 CFU in 0.5 ml PBS. The blood samples were collected at 12 h, 24 h and 36 h post infection, diluted appropriately and plated on TSA for bacterial counting 25 .