BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

Spleen Transcriptome Profile of Muscovy Ducklings in Response to Infection WithMuscovy Duck ReovirusAuthor(s): Quanxi Wang, Yijian Wu, Yilong Cai, Yubin Zhuang, Lihui Xu, Baocheng Wu, and YandingZhangSource: Avian Diseases, 59(2):282-290.Published By: American Association of Avian PathologistsDOI: http://dx.doi.org/10.1637/10992-112514-RegURL: http://www.bioone.org/doi/full/10.1637/10992-112514-Reg

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercialinquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

Spleen Transcriptome Profile of Muscovy Ducklings in Response to Infection WithMuscovy Duck Reovirus

Quanxi Wang,AB Yijian Wu,B Yilong Cai,B Yubin Zhuang,B Lihui Xu,B Baocheng Wu,BC and Yanding ZhangAC

ACollege of Life Science, Fujian Normal University, Fuzhou, Fujian 350119, ChinaBCollege of Animal Science, Fujian Agriculture and Forestry University, Fuzhou, Fujian 350002, China

Received 2 December 2014; Accepted 15 January 2015; Published ahead of print 16 January 2015

SUMMARY. Muscovy duck reovirus (MDRV) causes high morbidity and mortality in ducklings. However, the molecular basisfor pathogenesis of this virus remains poorly understood, and the complete genome sequence of Muscovy duck is lacking. Here wereport the transcriptome profile of Muscovy ducks in response to MDRV infection. RNA sequencing technology was employed toobtain a representative complement of transcripts from the spleen of ducklings, and then differential gene expression was analyzedbetween MDRV-YB strain infected ducks and noninfected ducks. This analysis generated 65,536 unigenes. Of these, 6458 geneswere found to be significantly differentially expressed between the infected and noninfected groups. The symptom and pathology ofducks infected with MDRV-YB was correlated with the greater proportion of decreased expression genes (4906) than increasedexpression (1552) level. Gene ontology analysis assigned differentially expressed genes to the categories: ‘‘biological processes,’’‘‘cellular functions,’’ and ‘‘molecular functions.’’ Differentially expressed genes involved in the innate immune system were analyzedfurther, and 128 of these genes showed decreased expression and 86 showed increased expression between the infected andnoninfected groups. These genes represented the Janus kinase–signal transducer and activator of transcription signaling pathway,and the retinoic acid-inducible gene I (RIG-I)–like and Toll-like receptor (TLR) signaling pathways and included interferon (IFN)a, IFNc, interleukin 6, RIG-I, and TLR4. The data were verified by SYBR fluorescence quantitative polymerase chain reaction(SYBR-qPCR). Our findings offer new insight into the host immune response to MDRV infection.

RESUMEN. Perfil del transcriptoma esplenico de patitos reales en respuesta a la infeccion con un reovirus del pato real.El reovirus del pato real (MDRV) causa alta morbilidad y mortalidad en patitos. Sin embargo, la base molecular de la patogenesis de

este virus sigue siendo poco conocida y no se encuentra disponible una secuencia completa del genoma. En este trabajo se muestra el perfildel transcriptoma de patos reales en respuesta a la infeccion con el reovirus del pato real. Se utilizo la tecnologıa de secuenciacion de ARNpara obtener un complemento representante de las transcripciones del bazo de los patitos y posteriormente se analizo la expresiondiferencial de genes entre patos infectados con la cepa MDRV-YB y patos no infectados. Este analisis genero 65,536 unigenes. De estos,se encontro que 6458 genes se expresaron significativamente diferente entre los grupos infectados y no infectados. Los signos y patologıade los patos infectados con la cepa MDRV-YB se correlacionaron con la mayor proporcion de genes con expresion a la baja (4906) encomparacion con los genes con expresion a la alta (1552). El analisis de la ontologıa de genes asigno los genes expresados de maneradiferente en las siguientes categorıas: ‘‘Procesos biologicos’’, ‘‘funciones celulares’’ y ‘‘funciones moleculares’’. Los genes expresadosdiferencialmente implicados con el sistema inmune innato fueron analizados, y 128 de estos genes mostraron una disminucion de laexpresion y 86 mostraron aumento de la expresion entre los grupos infectados y no infectados. Estos genes representan la senal detransduccion quinasa Janus y el activador del mecanismo de senalizacion de la transcripcion, y los mecanismos de senales del geneinducible por acido retinoico I (RIG-I) y del receptor de tipo Toll (TLR), incluyendo interferon (IFN) a, IFNc, interleucina 6, RIG-I, yTLR4. Los datos fueron verificados por reaccion en cadena de la polimerasa cuantitativa con fluorescencia con SYBR (SYBR-qPCR).Nuestros resultados ofrecen una nueva vision de la respuesta inmune del huesped a la infeccion con el reovirus del pato real.

Key words: Muscovy duck reovirus, transcriptome, innate immune

Abbreviations: ARV 5 avian reovirus; cDNA 5 complementary DNA; COG 5 clusters of orthologous groups of proteinsdatabase; DEG 5 differentially expressed gene; DPI 5 days postinfection; FDR 5 false discovery rate; GO 5 gene ontology;HE 5 hematoxylin and eosin; IFN 5 interferon; IL 5 interleukin; ISGs 5 IFN stimulating genes; JAK-STAT 5 Janus kinase–signal transducer and activator of transcription; MDRV 5 Muscovy duck reovirus; MOI 5 multiplicity of infection; mRNA 5messenger RNA; NF-kB 5 nuclear factor kappa-light-chain-enhancer of activated B cells; PRRs 5 pattern recognition receptors;RIG-I 5 retinoic acid–induced protein gene I; RNase 5 ribonuelease; RNA-seq 5 RNA sequencing technology; SYBR-qPCR 5SYBR fluorescence quantitative polymerase chain reaction; TCID50 5 50% tissue culture infective dose; TLR 5 toll-like receptor.

Reoviridae is a family of viruses that consists of nine genusmembers (15). Viruses in the Reoviridae family have double-stranded RNA genomes and have been reported to affect thegastrointestinal system and respiratory tract. One member of thisfamily, the genus orthoreovirus, has a wide host range includingdiverse species such as birds and mice (6). Avian reovirus (ARV) hasbeen reported to cause arthritis and malabsorption in chickens,whereas Muscovy duck (Cairina moschata) reovirus (MDRV) isa highly pathogenic virus in poultry causing acute infectious disease

in ducklings aged between 2 and 4 wk (26). MDRV was first isolatedin France in 1872 (9), and MDRV infection was first described inSouth Africa in 1950 (29). In 1997, MDRV infection was epidemicin Southeast China; this was the first report of this disease in China(29). Ducks infected with MDRV-YB strain (Genbank accessionnumber: KC571174.1) were characterized by extensive ‘‘whitenecrosis’’ in the liver and spleen and high mortality (28). Comparedwith ARV, the genome sequence of MDRV-YB is highlyhomologous to 89026 strain, and protein P17 is absent in strainYB (27). So MDRV exhibits unique pathogenicity and tissuetropism (10). However, little is known about the molecular basis ofpathogenesis in this organism.

CCorresponding authors. E-mails: ydzhang@fjnu.edu.cn andbc20020212@163.com

AVIAN DISEASES 59:282–290, 2015

282

Generation sequencing based on RNA sequencing (RNA-seq) isa method for accurate qualitative and quantitative characterization ofthe expressed complement of a genome (19). It provides millions ofsequences, giving clear data on the transcripts present and theirrelative abundance (23). RNA-seq is expected to reveal an accuraterepresentation of the transcriptome in an infected host (7), andanalysis of the differential expression of genes between infected andnoninfected hosts would give insight into the genetic mechanismsinvolved in disease pathogenesis.

Previous research has indicated the potential for genetic resistanceto Ascaridia galli infection in poultry (24). A greater understandingof the host response to infection and the resulting pathology willallow researchers to identify the genes that best convey protection(7). In this study, we performed RNA-seq–based transcriptomeanalysis on the spleen of a duck. Then, we analyzed differential geneexpression between infected and noninfected ducks. Genes for whichsignificant differential expression was detected were verified byquantitative SYBR fluorescence quantitative polymerase chainreaction (SYBR-qPCR).

MATERIALS AND METHODS

Ethics statement. The animal protocol used in this study wasapproved by the Research Ethics Committee of College of AnimalScience, Fujian Agriculture and Forestry University. All duck experi-mental procedures were performed in accordance with the Regulationsfor the Administration of Affairs Concerning Experimental Animalsapproved by the State Council of China.

Viruses and infection procedure. MDRV-YB strain was isolated andidentified by Wu and colleagues in 2001 (28). It was propagated inMuscovy duck embryo fibroblasts cultured in Roswell Park MemorialInstitute 1640 medium (Hyclone, Logan, UT) supplemented with 2%fetal bovine serum (Hyclone), as previously described (26). The 50%tissue culture infective dose (TCID50) was 1025.40 (8 multiplicity ofinfection (MOI)/ml).

Eighteen 1-day-old French kelimo Muscovy ducklings (blood ofeach 1-day-old duckling was detected with PCR to ensure that the

ducklings have not been infected with MDRV) were equally dividedinto treatment and control groups with three replicates per group.Ducks in the treatment group were infected with MDRV-YB (0.2 ml,0.01 TCID50 5 1025.40, 8 MOI/ml) by intramuscular injectionindividually. The control group was treated with sterile saline. Duckswere euthanatized and the spleen was collected on 5 days postinfection(DPI) (28).

Pathology observation on the spleen. At 5 DPI, ducklings weredecapitated and the spleens were collected. Then they were fixed in 10%formaldehyde for 24 hr. Following dehydration and transparence, thesamples were embedded into paraffin. Samples were cut into 5-mmsections. The sections were stained by hematoxylin and eosin (HE). Thepathology changes were observed and photographed by Leica IM50(Leica, Solms, Germany).

RNA extraction and cDNA synthesis. Total RNA was extractedfrom the spleen samples and treated with ribonuelease (RNase) freeDNase I (Promega, Madison, WI) to remove the contaminating DNA.Messenger RNA (mRNA) was enriched with Oligo (dT) magnetic beads(Illumina, Shenzhen, China). The concentration of total RNA wasquantified using a NanoDropH 1000 spectrophotometer (Thermo,Shanghai, China) and assessed by electrophoresis in a 1% agarose gel.

Illumina sequencing. RNA samples from each group were subjectedto Illumina sequencing using the paired-end RNA-seq method (1,18).Poly (A) mRNA was isolated from total RNA using sera-mag oligo(dT) beads (Illumina) and fragmented to a length of between 200 bpand 700 bp. First strand complementary DNA (cDNA) wassynthesized using random hexamers (Invitrogen, Carlsbad, CA), thefragmented RNA in a final volume of 40 ml, and the SuperScript IIReverse Transcriptase Kit (Invitrogen) following the manufacturer’sinstructions. Second strand cDNA was synthesized using 2 U of RNaseH (Invitrogen), DNA polymerase I (NEB, Ipswich, MA), and 30 nmolof deoxyribonucleotides (dNTPs) incubated at 25 C for 2.5 hr. Thefragments were repaired using the end repair module (NEB) and a 39 Atail was added using the Klenow fragment (NEB).

PCR was performed on the cDNA using the adaptor primers andindex primers and a program of 18 cycles. The PCR products wereseparated by gel electrophoresis, then isolated from the agarose usinga QIAquick gel purification kit (Qiagen, Shanghai, China). Clusteringand sequencing was then performed on a Hi-Seq 2000 (Illumina) by

Table 2 Output statistics of sequencing.

Samples Total readsA Total nucleotidesB (nt) Q20 (%) N (%) GC (%)

Muscovy duck spleen 113,373,834 17,006,075,100 96.89 0.07 52.34ATotal reads 5 Total reads 1 + Total reads 2.BTotal nucleotides 5 Total reads 1 3 Read 1 size + Total reads 2 3 Read 2 size.

Table 1. Sequences of primers in this study.

Genes Primers Length (bp)

MDRV-P10 F: ATGGCTGACGCTTTTGAAGT 288

R: TAGTTAGATCTCGAGAGCCCGIFNa F: AACCCAAGAGCCACCATGCCT 194

R: CTCTTAGCGGATGGTGCGGGTIFNc F: CCTCGTGGAACTGTCAAACC 140

R: TGCATCTCTTTGGAGACTGGIL6 F: ACCCAGAAATCCCTCCTCAC 125

R: AAATAGCGAACAGCCCTCACRIG-I F: ACCCTTACCTCGCTCTCCAT 161

R: CCAACAGAAGCAGTCAAACCTTLR4 F: TCACGGTCTGAACTCTCTGC 100

R: GATGTGGAGGTGGCTCAAGTb-Actin F: CCAAAGCCAACAGAGAGAAGAT 138

R: CATCACCAGAGTCCATCACAAT

Transcriptome profile of ducklings infected with MDRV 283

Guangzhou Gene Denovo Biological Technology Co., Ltd (Guangzhou,China) following the manufacturer’s protocols.

Read assembly and sequence annotation. The original image dataproduced by sequencing were transformed into sequence data through basecalling. The impurities in the raw sequence data (the ‘‘raw reads’’) shouldhave been removed by this process, giving ‘‘clean reads.’’ Then,

transcriptome de novo assembly was carried out using the program Trinityand the Denovo Institute of Genomic Research Gene Indices clustering tools(TGICL2003; http://compbio.dfci.harvard.edu/tgi/software/) (11). Theobtained unigenes were annotated by Blastx with the sequences availablein the nonredundant protein database (http://www.ncbi.nlm.nih.gov/),clusters of orthologous groups of proteins database (COG; http://www.ncbi.

Fig. 1. Functional annotation and classification of unigenes. (A) COG function classification of Cairina moschata unigene. (B) GO functionclassification of Cairina moschata unigene.

284 Q. Wang et al.

nlm.nih.gov/COG/), Kyoto encyclopedia of genes and genomes (http://www.kegg.jp/), and Swiss-Prot (http://www.expasy.ch/sprot; E , 0.00001).

Analyzing the biological changes in the host from thetranscriptomic data. The expression of unigenes was calculated as the

reads per kilobase of the exon model per million mapped reads, and thenmultiple hypothesis testing was performed including false discovery rate(FDR) controls. Differentially expressed genes (DEGs) were screenedusing the following two criteria. First, the FDR should be no larger than

Fig. 2. Pathology observation on the spleen infected with MDRV-YB. (A) Spleen of Muscovy duck infected with MDRV-YB in the 5 days afterinfection showed significant ‘‘white necrosis.’’ Histopathological observations found that the infected cells in splenic white pulp showeddegeneration, necrosis. (B) HE, 403; (C) HE, 1003; (D) HE, 4003. (E) The histology of control duck spleen was intact (F) HE, 403; (G) HE,1003; (H) HE, 4003.

Fig. 3. Screening of different expression genes in spleen infected with MDRV. About 6558 DEGs were obtained in spleen infected ducks,compare with noninfection. Some 4906 DEGs were down-regulated in infected duck spleen, and the other 1552 DEGs were up-regulated, which isshown with (A) histogram and (B) scatter diagram.

Transcriptome profile of ducklings infected with MDRV 285

Fig. 4. GO functional classification of DEGs in MDRV-YB infection spleen compared with control group. Some 16,246 DEGs had GOfunctional classification annotation, of which there were 3939 down-regulated unigenes and 12,307 up-regulated unigenes. The DEGs were involvedin 8525 biological processes unigenes (2049 up-regulated, 6476 down-regulated), 5102 molecular function unigenes (1235 up-igenes, 3867 down-unigenes), and 2619 molecular function unigenes (655 up-regulated, 1964 down-regulated).

Fig. 5. JAK-STAT pathway of spleen infected with MDRV-YB compared with control. Rectangles showing EPO, Cbl, IL2/3, the bottom halfof CytokineR, STAM, CPB, the bottom half of CycD, Spred, Sprouty, and AKT: significantly down-regulated genes (log fold change . 1, FDR, 0.001); Rectangles showing IFN/IL10, IL6, the top half of CytokineR, STAT, SOCS, Pim-1, and CycD: significantly up-regulated genes (logfold change . 1, FDR , 0.001); Other rectangles: expression unchanged genes. Bcl 5 B-cell lymphoma; CIS 5 cytokine-induced Src homology;C-Myc 5 proto-oncogene; CVCD 5 cerebral venous circulation disturbances; EPO 5 erythropoietin; IFN 5 interferon; IL 5 interleukin; JAK 5Janus kinase; MAPK 5 mitogen-activated protein kinase; ORB 5 growth factor receptor-bound protein; P48 5 retinoblastoma binding protein 48;PI3K 5 phosphatidylinositol 3-kinase; Pim 5 proviral integration site; PKB(AKT) 5 protein kinase B; Ras 5 rat sarcoma; SHP 5 src-homologydomain containing protein tyrosine phosphatase; SOCS 5 suppressor of cytokine signaling; SOS 5 son of sevenless; SPRED 5 Sprouty-related;STAT 5 signal transducer and activator of transcription; UPD 5 unpaired.

286 Q. Wang et al.

0.001. Second, the expression difference between two groups should begreater than two times or verified by real-time PCR to be greater thantwo times.

Functional classification of the DEGs was performed by geneontology (GO, http://www.geneontology.org/), and the categories‘‘biological process,’’ ‘‘molecular function,’’ and ‘‘cellular component’’were analyzed.

Hierarchy clustering analysis with the DEGs was performed usingcluster software (http://bonsai.hgc.jp/,mdehoon/software/cluster/software.htm) (4), and the data obtained were displayed using the Java Treeviewsoftware (http://jtreeview.sourceforge.net/).

Based on the annotation data collected, genes involved in pathways ofthe innate immune response, including the Janus kinase–signal trans-ducer and activator of transcription (JAK-STAT), retinoic acid-inducible gene I (RIG-I) like and the Toll-like receptor (TLR) pathway,were manually screened and compared between the infected group andthe noninfected group.

Quantitative real-time PCR analysis of selected genes fromthe transcriptome. Genes of interest were manually screened. SYBR-qPCR was performed with the primers described in Table 1 to analyzethe expression levels of these genes of interest in the infected andnoninfected groups, as well as to verify the transcriptomic data.

RESULTS

Characteristics of the Muscovy duck transcriptome. In thisstudy, analysis of the Muscovy duck spleen transcriptome revealeda total of 113,373,834 reads, which included 17,006,075,100nucleotides. The average read length was 150 bp. The Q20

percentage (the percentage of bases whose quality was larger than 20in clean reads) was 96.89%, the N percentage (the percentage ofnucleotides in clean reads) was 0.07%, and the GC percentage (thepercentage of G/C bases in clean reads) was 52.34% (Table 2). Thedata were reliable because the library concentration and the sizedistribution were both uniform across the samples.

From our analysis of the Muscovy duck spleen, 119,416 contigswere assembled. The N50 was 116,294, and the length distributionof the contigs is shown in Fig. 1A. Furthermore, 65,535 unigeneswere assembled, and the average length of a unigene was 987 bp.

Using the COG database, 41,651 unigenes were annotated, theseincluded 25 categories of biological functions (Fig. 1A). The mostwell-represented category was ‘‘general function prediction only’’(5649 unigenes), followed by ‘‘translation, ribosomal structure andbiogenesis’’ (4812 unigenes), and the least represented category was‘‘nuclear structure’’ (nine unigenes). There were also 3737 functionalunknown unigenes, which may be genes that are unique to Muscovyducks.

GO functional classification of the 107,492 unigenes revealedthree categories: ‘‘molecular function’’ (18,209 unigenes), ‘‘bi-ological processes’’ (51,725 unigenes), and ‘‘cellular compounds’’(37,558 unigenes). As shown in Fig. 1B, 32,403 coding regionsequences of the unigenes were found. The other 1694 coding regionsequences of unigenes were screened using the ESTscan software.

Of the unigenes identified, 16,258 unigenes were classified into240 biological pathways. The most well represented were metabolicpathways (1941 unigenes, 11.94%), with only one unigene each

Fig. 6. RIG-I pathway of spleen infected with MDRV-YB compared with control. Rectangle showing JNK: significantly down-regulated genes(log fold change , 1, FDR . 0.001); Rectangles showing TRIM25, RIG-I, LGP2, MDA5, IPD-1, IKKe, IRF7, IFNa, IFNb, FADD, CASP8,IKKr, IL-8, IL-12: significantly up-regulated genes (log fold change . FDR , 0.001); expression unchanged genes, other rectangles; Atg 5antithymocyte globulin; CYLD 5 cylindromatosis; DAK 5 dihydroacetone kinase; DUBA 5 deubiquitinating enzyme A; FADD 5 fas-associateddeath domain; IKB 5 I-kappa-B-beta; IKK 5 IKB kinase; IP 5 10 kDa interferon-gamma-induced protein; ISG15 5 interferon stimulate gene;JNK 5 jun N-terminal kinases; LGP2 5 laboratory of genetics and physiology 2; MDA5 5 melanoma-differentiation-associated gene 5; MEKK 5mitogen extracellular signal-regulated kinase kinase; MTA 5 metastasis associated protein 2; NAP 5 neutrophil chemotactic protein; NF-kB 5nuclear factor kappa-light-chain-enhancer of activated B cells; NLR 5 nucleotides in natural immune system combined with oligomeric structuredomain receptor; P38 5 10 kDa mitogen-activated protein kinase; RIG-I 5 retinoic acid-induced protein genes; RIP 5 DNreceptor- interactingprotein; RNF 5 ring finger protein; SINTBAD 5 similar to NAP1 TBK1 adapto; TAK 5 transforming growth factor bactivated kinase; TANK 5tumor necrosis factor receptor-associated factor family member associated NF-kB activator; TRADD 5 the tumor necrosis factor-receptor1associated death domain protein; TRAF 5 tumor necrosis factor receptor-associated factor; TRIM 5 tripartite motif.

Transcriptome profile of ducklings infected with MDRV 287

being classified into the pathways for polyketide sugar unitbiosynthesis, and D-arginine and D-ornithine metabolism.

Pathological observations of the spleen infected with MDRV-YB. Large areas of ‘‘white necrosis’’ (Fig. 2A) appeared in thespleens of MDRV-YB–infected ducklings in the 5 days afterinfection. Histological pathological observations found that theinfected cells in the splenic white pulp showed degeneration andnecrosis (Fig. 2B–D). Obvious inflammation was also observed inthe necrotic foci (Fig. 2D). The organizational structure of thecontrol duck spleen was intact (Fig. 2E–H). These findings indicatethat the host activates defense measures against MDRV infection.

Screening of different expression genes in the spleen infectedwith MDRV. Some 6558 DEGs were detected in the spleens ofinfected ducks (Fig. 3A,B). Of these, 4906 DEGs were down-regulated in infected duck spleens and 1552 DEGs were up-regulated.

Of the 16,246 DEGs that were subjected to GO functionalclassification annotation, 3939 were down-regulated unigenes and12,307 were up-regulated unigenes. They were classified as follows:‘‘biological processes’’ (8525 unigenes, of which 2049 up-regulatedand 6476 down-regulated), ‘‘cellular compounds’’ (5102 unigenes,of which 1235 up-regulated and 3867 down-regulated), and‘‘molecular function’’ (2619 unigenes, of which 655 up-regulatedand 1964 down-regulated; Fig. 4).

Of particular interest were the genetic differences in genesinvolved in the immune system. The data revealed 214 DEGsinvolved in the immune system, of these 86 were up-regulatedunigenes and 128 were down-regulated unigenes (Fig. 4). These

findings suggested that the host can regulate the expression of genesinvolved in immune function during MDRV infection.

Pathway analysis and SYBR-qPCR detection. To furtherexplore the innate immune response of the host during MDRVinfection, we analyzed the JAK-STAT, RIG-I, and TLR pathways.Expression levels of type I interferon (IFN-I), interleukin-6 (IL6),STAT (Fig. 5), RIG-I (Fig. 6), interferon promoter stimulator,interferon regulatory factor 3 (IRF3), IRF7, and TLR4 (Fig. 7) inthe spleens of infected ducks were significantly increased comparedwith noninfected ducks. This indicates that as a defense againstMDRV infection, the host promotes the secretion of interferon andIL6 that activate important signal molecules in the three interferonregulation signaling pathways.

In order to obtain more accurate results, we conducted thefluorescence quantitative PCR detection for the expression of IFN,IFNc, IL6, RIG-I, and TLR4. Results of these gene differenceexpressions were consistent with the chip results (Table 3), whichshowed that the chip results was reliable.

DISCUSSION

Muscovy ducks (Cairina moschata) are an important domesticwaterfowl in China (30), with about 150 million Muscovy ducks bredannually in Fujian Province. Huang and colleagues reported thecomplete genome sequence of Peking duck in 2013 (12). However, thisinformation is not available for the Muscovy duck or domesticatedducks (Anas platyrhynchos domesticus) (20). Here we report the

Fig. 7. Toll-like pathway of spleen infected with MDRV-YB compared with control. Rectangles showing CTSK, AKT, MKK3/6, JNK, AP-1:significantly down-regulated genes (log fold change . 1, FDR . 0.001); Rectangles showing TLR1, TLR2, MD-2, TLR4, IFN-b, FADD, CASP8,IKKb, Ikke, OPN, IRF7, STAT1, IL-1b, IL-6, IL-12, IL-8, RANTES, MIP-1a, MIP-1b, CD-80, IFN-a, IFN-b: significantly up-regulated genes(log fold change . 1, FDR , 0.001); Other rectangles: expression unchanged genes. CASP 5 Caspase; CD 5 cluster of differentiation; LBP 5 lowback pain; MIP 5 macrophage inflammatory protein 1; Myd 5 Myeloid cell differentiation; PLS 5 lipopolysaccharide; Rac 5 raeGTP enzymeactivation protein; RANTES 5 regulated upon activation, normally T-cell expressed and presumably secreted; TAB 5 TAK1-binding protein;TIRAP 5 toll/interleukin 1 receptor protein structure domain; TLR 5 toll-like receptor; TollIP 5 toll interacting protein.

288 Q. Wang et al.

transcriptome profile of the Muscovy duck spleen as deciphered byhigh-throughput sequencing (RNA-seq), which should provideessential information for understanding the biology of Muscovy ducks.For vertebrate organisms where a reference genome is not available, denovo transcriptome assembly enables a cost effective insight into theidentification of specific tissue or differentially expressed genes andvariations in the coding regions of the genome.

Transcriptome analysis provides a new research tool for the studyof the interactions between host and virus, because transcriptomedata can serve as a rich genetic resource (3). Huang and colleagues(12) performed deep transcriptome analysis to characterize geneexpression profiles and to identify genes that are responsive to avianinfluenza viruses. They found that avian defenses might be involvedin the host immune response to two H5N1 viruses whose infectionwas examined in ducks. Moreton and colleagues (20) presenteda number of DEGs in duck hepatitis B virus–infected primary duckhepatocyte cultures; their findings contribute valuable insight intolittle known aspects of hepadnavirus biology. Le Goffic andcolleagues (13) used transcriptomic analysis to investigate hostimmune and cell death responses associated with influenza A virusPB1-F2 protein and demonstrated that this protein stronglyinfluences the early host response during influenza A virus infection,providing new insight into the mechanisms by which PB1-F2mediates virulence. In this study, we identified 6458 DEGs bytranscriptome analysis of MDRV-infected duck spleens; of these,4906 DEGs were down-regulated and 1552 DEGs were up-regulated in the infected duck spleen compared with the noninfectedduck spleen. The high number of DEGs in the spleen indicated thatthe host was launching a significant response to MDRV infection.Detailed analysis of the biological function of these DEGs inMDRV infection is now required to gain insight into the hostdefenses against infection by this pathogen.

The host can respond biologically to virus invasion via a numberof different defense mechanisms (9,21). In this respect, ARVinfection has been more closely studied than MDRV infection.ARV-infected chicken cells have been shown to undergo apoptosis,suggesting a correlation between virus replication and apoptosis inchicken cells. P53 was modulated by mitogen-activated proteinkinase pathways and protein kinase C during avian reovirusS1133-induced apoptosis (16). In contrast, little is known aboutthe host response after MDRV infection. Geng and colleagues (10)reported that MDRV could induce host cell apoptosis but themolecular basis for this remains unclear. Transcriptome analysismay provide insight into this and other questions regardingMDRV infection (5).

The ability of the innate immune system to respond to a variety ofstimuli is central to its function in pathogen clearance and tissuerepair (2). Virus invasion is recognized by the host through patternrecognition receptors (PRRs), such as TLRs and RIG-I-likereceptors. PRRs transduce intracellular signals to activate nuclearfactor-kappa B (NF-kB) and/or IFN-I, IRF3, and/or IRF7, leadingto proinflammatory cytokine and IFN-I expression by the infectedcells (1,14). Newly synthesized IFN-I is secreted and binds to IFN-I

receptor inducing the expression of hundreds of IFN stimulatinggenes (ISGs) with direct antiviral effect. To explore whether MDRVinfection activates host innate immune responses, we investigatedthree signaling pathways, JAK-STAT, RIG-I, and TLR, by RNA-seqand SYBR-PCR. Our results indicated that in the face of the MDRVinvasion, host cells identified MDRV RNA through the TLR-4receptor and RIG-I like receptor, which promotes the expression ofIFN and IL6 and the secretion of antiviral protein to resist MDRVinfection.

Interleukin-6 is a multifunctional proinflammatory cytokine thatregulates inflammation and the immune response. IL6 is over-expressed in several inflammatory diseases including rheumatoidarthritis (8). The induction of inflammatory responses mediated byvarious cytokines and chemokines leads to the activation of theinnate immune response and viral clearance (4,22,25). Over-expression of interferon promotes the secretion of interferonregulatory factor, which participates in the inflammatory reaction(17). Our results indicated that IL6 and IFNa are overexpressed,suggesting that IL6 may play an important regulatory function in thehost against MDRV invasion and IFNa may promote theinflammatory reaction and the innate immune response. Futurestudies will confirm and investigate these findings further.

REFERENCES

1. Brennan, K., and A. G. Bowie. Activation of host pattern recognitionreceptors by viruses. Curr. Opin. Microbiol. 13:503–507. 2010.

2. Ceren, C., R. J. John, S. S. Fayyaz, and L. C. Suzanne. Control ofinnate and adaptive immunity by the inflammasome. Microbes Infect.14:1263–1270. 2012.

3. Chu, J. H., R. C. Lin, C. F. Ye, Y. C. Hu, and S. H. Li.Characterization of the transcriptome of an ecologically important avianspecies, the Vinous-throated Parrotbill Paradoxornis webbianus bulomachus(Paradoxornithidae; Aves). BMC Genomics 13:149. 2012.

4. De Hoon, M. J. L., S. Imoto, J. Nolan, and S. Miyano. Open sourceclustering software. Bioinformatics 20(9):1453–1454. 2004.

5. Delgado, J. A., M. D. Jimenez, and A. Gomez. Samara size versusdispersal and seedling establishment in Ailanthus altissima (Miller) Swingle.J. Environ. Biol. 30(2):183–186. 2009.

6. Eili, H., Y. U. Nathalie, M. Tytti, M. Riggert, B. K. Markus, V.Antti, and V. Olli. Novel orthoreovirus from diseased crow, Finland. Emerg.Infect. Dis. 13(12):1967–1969. 2007.

7. Erin, E. S., O. Megan, B. Emma, B. Nate, X. Y. Li, H. J. Zhou, J. J.Timothy, K. Subhashinie, P. Liu, K. N. Lisa, and J. L. Susan. Spleentranscriptome response to infection with avian pathogenic Escherichia coli inbroiler chickens. BMC Genomics 12:469–482. 2011.

8. Frey, N., S. Grange, and T. Woodworth. Population pharmacoki-netic analysis of Tocilizumab in patients with rheumatoid arthritis. J. Clin.Pharmacol. 50(7):754–766. 2010.

9. Gaudry, D., J. Charles, and J. Tektoff. A new disease expressing itselfby a viral pericarditis in Barbary ducks. Comptes rendus hebdomadairesdes seances del’Academie des sciences. Serie D: Sciences naturelles274:2916–2919. 1972.

10. Geng, H., Y. Zhang, P. Y. Liu, G. D. Vanhanseng, Y. Wang, M. Liu,and G. Tong. Apoptosis p10.8 protein in primary duck embryonatedfibroblast and Vero E6 cells. Avian Dis. 53(3):434–440. 2009.

Table 3. Partial differentially expressed unigenes during MDRV-YB infection.

Gene log2 ratio Up (+)/down (2) QPCR fold change Up (+)/down (2)

IFNa 5.85 + 4.73 +IFNc 1.01 + 7.15 +IL6 20.83 + 7.06 +RIG-I 9.07 + 8.37 +TLR4 1.21 + 2.49 +

Transcriptome profile of ducklings infected with MDRV 289

11. Grabherr, M. G., B. J. Haas, M. Yassour, J. Z. Levin, D. A.Thompson, I. Amit, X. Adiconis, L. Fan, R. Raychowdhry, Q. Zeng, Z.Chen, E. Mauceli, N. Hacohen, A. Gnirke, N. Rhind, F. di Palma, B. W.Birren, C. Nusbaum, K. Lindblad-Toh, N. Friedman, and A. Regev. Full-length transcriptome assembly from RNA-Seq data without a referencegenome. Nat. Biotechnol. 29(7):644–652. 2011.

12. Huang, Y., Y. Li, D. W. Burt, H. Chen, Y. Zhang, W. Qian, H. Kim,S. Gan, Y. Zhao, J. Li, K. Yi, H. Feng, P. Zhu, B. Li, Q. Liu, S. Fairley, K.E. Magor, Z. Du, X. Hu, L. Goodman, H. Tafer, A. Vignal, T. Lee, K. W.Kim, Z. Sheng, Y. An, S. Searle, J. Herrero, M. A. Groenen, R. P.Crooijmans, T. Faraut, Q. Cai, R. G. Webster, J. R. Aldridge, W. C.Warren, S. Bartschat, S. Kehr, M. Marz, P. F. Stadler, J. Smith, R. H. Kraus,Y. Zhao, L. Ren, J. Fei, M. Morisson, P. Kaiser, D. K. Griffin, M. Rao, F.Pitel, J. Wang, and N. Li. The duck genome and transcriptome provideinsight into an avian influenza virus reservoir species. Nat. Genet.45(7):776–783. 2013.

13. Le Goffic, R., O. Leymarie, C. Chevalier, E. Rebours, B. Da Costa, J.Vidic, D. Descamps, J.-M. Sallenave, M. Rauch, M. Samson, and B.Delmas. Transcriptomic analysis of host immune and cell death responsesassociated with the influenza A virus PB1-F2 protein. PLOS Pathog.7(8):e1002202. 2011.

14. Le Negrate, G. Viral interference with innate immunity by preventingNF-kB activity. Cell. Microbiol. 14(2):168–181. 2012.

15. Lian, H., Y. Liu, S. F. Zhang, F. Zhang, and R. Hu. Novelorthoreovirus from Mink, China. Emerg. Infect. Dis. 19(12):1985–1988.2011.

16. Lin, H. Y., S. T. Chuang, Y. T. Chen, W. L. Shih, C. D. Chang, andH. J. Liu. Avian reovirus-induced apoptosis related to tissue injury. AvianPathol. 36(2):155–159. 2007.

17. Lin, P. Y., J. W. Lee, M. H. Liao, H. Y. Hsu, S. J. Chiu, H. J. Liu,and W. L. Shih. Modulation of p53 by mitogen-activated protein kinasepathways and protein kinase C d during avian reovirus S1133-inducedapoptosis. Virology 385:323–334. 2009.

18. Luo, J., A. C. Jose, R. M. Kimberly, L. T. Nathaniel, and J. H. Song.Transcriptome analysis reveals an activation of major histocompatibilitycomplex 1 and 2 pathways in chicken trachea immunized with infectiouslaryngotracheitis virus vaccine. Poult. Sci. 93:848–855. 2014.

19. Montgomery, S. B., M. Sammeth, M. Gutierrez-Arcelus, R. P. Lach,C. Ingle, J. Nisbett, R. Guigo, and E. T. Dermitzakis. Transcriptomegenetics using second generation sequencing in a Caucasian population.Nature 464:773–777. 2010.

20. Moreton, J., S. P. Dunham, and R. D. Emes. A consensus approachto vertebrate de novo transcriptome assembly from RNA-seq data: assembly

of the duck (Anasplatyrhynchos) transcriptome. Front. Genet. 5(190):1–6.2014.

21. Neetu, J., V. Satya Pavan Kumar, T. Deepshi, J. Prashant, D.Hemlata, and P. Subrat Kumar. RNA-Seq based transcriptome analysis ofhepatitis e virus (HEV) and hepatitis B virus (HBV) replicon transfectedhuh-7 cells. PLOS One 2(9):e87835. 2014.

22. O’Neill, L. A., and A. G. Bowie. Sensing and signaling in antiviralinnate immunity. Curr. Biol. 20(7):328–333. 2010.

23. Raj, P. K., K. R. Harsha, J. B. Matthew, N. Jacob, W. Jun, Q. Jiang,S. K. Timothy, and S. Anand. Transcriptome analysis using next generationsequencing reveals molecular signatures of diabetic retinopathy and efficacyof candidate drugs. Mol. Vision 18:1123–1146. 2012.

24. Schou, T., A. Permin, A. Roepstorff, P. Sørensen, and J. Kjaer.Comparative genetic resistance to Ascaridia galli infections of 4 differentcommercial layer lines. Br. Poult. Sci. 44(2):182–185. 2003.

25. Unterholzner, L., and A. G. Bowie. The interplay between viruses andinnate immune signaling: recent insights and therapeutic opportunities.Biochem. Pharmacol. 75(3):589–602. 2008.

26. Wang, Q. X., B. C. Wu, C. K. Wang, and F. Su. Mast cell dynamicsin the digestive system during infection by Muscovy duck reovirus. J. Anim.Vet. Adv. 11(13):2204–2210. 2012.

27. Wang, S., G. H. Deng, Y. J. Wu, B. H. Wu, and H. L. Chen.Complete cDNA sequences analysis of the S1 genes of the avian (muscovyduck) reovirus isolates in China. Chinese J. Prev. Vet. Med. 27(2):105–111.2005.

28. Wu, B. C., J. X. Chen, and J. S. Yao. Isolation and identification ofMuscovy duck reovirus. J. Fujian Agric. Forestry Univ. (Nat. Sci. Ed.)3(2):227–230. 2001.

29. Yun, T., B. Yu, Z. Ni, W. Ye, L. Chen, J. Hua, and C. Zhang.Isolation and genomic characterization of a classical Muscovy duck reovirusisolated in Zhejiang, China. Infect. Genet. Evol. 20:444–453. 2013.

30. Zeng, T., X. Y. Jiang, J. J. Li, D. Q. Wang, G. Q. Li, L. Z. Lu, andG. L. Wang. Comparative proteomic analysis of the hepatic response to heatstress in Muscovy and Peking ducks: insight into thermal tolerance related toenergy metabolism. PLOS One 10(8):e76917. 2013.

ACKNOWLEDGMENTS

This work was supported by a program grant from the ChinaNational Natural Science Foundation and the Joint Fund for thePromotion of Scientific Cooperation across the Taiwan Straits(No. U1305212) and a program grant from the China Fujian ProvinceNatural Science Foundation (No. 2013J0169).

290 Q. Wang et al.