High throughput screening of recombinant antibodies against infectious hematopoietic necrosis virus from a combinatorial antibody library

Aquaculture 460 (2016) 32–36

Contents lists available at ScienceDirect

Aquaculture journal homepage: www.elsevier.com/locate/aquaculture

High throughput screening of recombinant antibodies against infectious hematopoietic necrosis virus from a combinatorial antibody library Li-Ming Xu, Jing-Zhuang Zhao, Miao Liu, Yong-Sheng cao, Jia-Sheng Yin, Hong-Bai Liu, Tongyan-Lu ⁎ Heilongjiang River Fishery Research Institute, Chinese Academy of Fishery Sciences, Harbin 150070, PR China

a r t i c l e

i n f o

Article history: Received 15 February 2016 Received in revised form 5 April 2016 Accepted 5 April 2016 Available online 9 April 2016 Keywords: Antibody library Antibody-antigen co-expression Bacterial display Glycoprotein Infectious hematopoietic necrosis virus

a b s t r a c t Infectious hematopoietic necrosis virus (IHNV) is a significant rhabdoviral pathogen of salmonid fish. In this study, a single chain variable fragment (scFv) antibody library derived from rainbow trout (Oncorhynchus mykiss) and a glycoprotein fragment (named G4, 540 nt, 180 aa) of IHNV-Sn1203 isolate were co-expressed by a bacterial display technology. The library was subjected to three rounds of screening by flow cytometry (FCM) to select IHNV specific antibodies. Seven antibody clones with different mean fluorescence intensities (MFI) were obtained by picking colonies at random. The antibody clone with the highest MFI was expressed and purified. The purified IHNV-specific scFv antibody was used successfully in Western blotting, enzyme linked immunosorbent assay and an immunofluorescence antibody test. This method provides a high throughput means to screen an antibody library by flow cytometry and isolate an antibody that can be used as a potential universal reagent for the detection and confirmation IHNV strains that are prevalent throughout China. Statement of relevance: Outbreaks of infectious hematopoietic necrosis caused severe economic losses to salmon and trout aquaculture in China every year. In this study, a panel of recombinant antibodies against Chinese IHNV isolates was obtained. The isolated antibody was proven can be used as a universal diagnosis regent for IHNV prevalent in China. The study provides a novel method for rapid development of antibodies to emerging diseases. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Infectious hematopoietic necrosis virus (IHNV) is a rhabdovirus that causes acute, systemic disease in salmonid fish and also occurs asymptomatically. As one of the most important viral diseases in aquaculture, outbreaks of IHN result in losses approaching 100% depending on the species and size/age of the fish, the virus strain and environmental conditions. Since the first outbreak of IHN in Washington and Oregon fish hatcheries during the 1950s, IHNV is now found in many countries (Breyta et al., 2013; Enzmann et al., 2005; Jia et al., 2014; Kim et al., 2007; Kolodziejek et al., 2008; Nishizawa et al., 2006; Rudakova et al., 2007; Troyer et al., 2000). In 1985, the first outbreak of IHN was recorded in hatcheries for juvenile rainbow trout in Liaoning Province, China (Niu, and Zhao, 1988). This was followed by a series of IHN outbreaks in cultured juvenile rainbow trout in various districts of China (Jia et al., 2014; Xu et al., 2013). In previous studies, a bacterial display technology was established and successfully applied to antibody library screening (Xu, Li, Zhou, Guo, Liu, Zhao, Cao, Li, 2014). However, by using the bacterial display ⁎ Corresponding author. E-mail addresses: [email protected] (L.-M. Xu), [email protected] (J.-Z. Zhao), [email protected] (M. Liu), [email protected] (Y.-S. cao), [email protected] (J.-S. Yin), [email protected] (H.-B. Liu), [email protected] (Tongyan-Lu).

http://dx.doi.org/10.1016/j.aquaculture.2016.04.002 0044-8486/© 2016 Elsevier B.V. All rights reserved.

technology, purified antigen must be obtained before the screening process. This is time consuming and labor intensive. In this study, an antigen-antibody co-expression display technology was established. The present study was designed to isolate IHNV-specific single chain variable fragment (scFv) antibodies from an antibody library assembled from rainbow trout using an antibody-antigen co-expression system combined with flow cytometry (FCM). 2. Materials and methods 2.1. Materials Escherichia coli DH5α was used for displaying the scFv library, while E. coli Rosetta and pET27b(+) were used for expression of scFv's that were isolated. The pCoex vector was used for scFv library display and was generated in our laboratory. The glycoprotein gene sequence of the IHNV isolates used in this study was submitted to the GenBank database and referred to as follows, IHNV strain Sn1203 (KC660147), LN12 (KF871194), SD-12 (KF871193), GS-12 (KF871194), YN13 (KF871192) and XJ-13 (KF871191). A rabbit anti-IHNV glycoprotein polyclonal antibody (described below) and the cell line Epithelioma papulosum cyprini (EPC) were from lab stocks. An FITC antibody labeling kit was purchased from Thermo (California, USA). CY3-mouse anti-His tag antibody was purchased from eBioscience (Shanghai, China).

L.-M. Xu et al. / Aquaculture 460 (2016) 32–36

33

2.2. Construction of antigen-antibody co-expression vector pCoex A previously constructed vector pBFD (Xu, Liu, Zhao, Cao, Yin, Liu, Lu, 2014) was used as a basic frame to construct the antigen-antibody co-expression vector. The ribosome binding site (RBS) and pelB leader peptide were amplified from plasmid pET-27b and inserted into the pBFD by EcoRI and XhoI to construct antigen-antibody co-expression vector pCoex. The fragment of the IHNV glycoprotein G4 (540 nt, 180 aa) (Xu, Liu, Zhao, Cao, Yin, Liu, Lu, 2014) was fused with the pelB leader peptide, and the antibody library (described below in Section 2.3) was fused with the NlpA leader peptide. Both of these were co-expressed and transported to the periplasm. If the anchored antibody could bind to the antigen an antibody-antigen complex was formed. Upon removal of the outer membrane by spheroplasting, fluorescent antibodies against the glycoprotein entered the periplasmic space where they were recognized by the membrane-tethered antibody-antigen complex. As a result, specific antibody-expressing spheroplasts became fluorescently labeled and could be readily enriched and screened by a fluorescence activated cell sorter (FACS) (Fig. 1). 2.3. Construction of scFv library Head kidney and spleen were isolated from rainbow trout (mean weight, 20 g ± 2.5) that survived an outbreak of IHN that caused approximately 80% cumulative mortality. Tissues from 6 rainbow trout were harvested and pooled two month after the outbreak of IHN, and total RNA was extracted using Trizol. cDNA was synthesized from the total RNA sample using Superscript II (Invitrogen) and random hexamer oligonucleotide primers. Primers used to amplify the VH and VL gene fragment for construction of the antibody library were designed according to a previous study (Zhou and Xie, 2015). PCR products of the VH and VL genes were linked by overlapping PCR to construct the scFv antibody library. The linker peptide used in the study were typical (Gly4Ser)3 sequences. The fragment of IHNV glycoprotein G4 (Xu, Liu, Zhao, Cao, Yin, Liu, Lu, 2014) was fused with the pelB leader peptide, and the antibody library was fused with the NlpA leader peptide. The recombinant plasmids were electroporated into E. coli DH5α according to standard procedures, and the library was labeled pCoex-G-scFv. The colonies were counted to calculate the transformation efficiency of electroporation. Library diversity was determined by sequencing 5 random clones obtained from the scFv library. 2.4. Spheroplast preparation and scFv screening All colonies from the pCoex-G-scFv library-transformed DH5α were collected and cultured in lysogeny broth (LB). Expression of antigen and antibody were induced as in previous studies (Xu et al., 2015). The spheroplasts were prepared as previously described (Jeong et al., 2007). Rabbit anti-full length IHNV glycoprotein polyclonal antibody (Xu et al., 2013) was labeled using an FITC antibody labeling kit according to the manufacturer. The spheroplasts were incubated with 200 nM FITC-rabbit anti-glycoprotein antibody for 1 h at 4 °C and then screened by FACS. The library of cells was sorted on a FACSAria™ Cell Sorter (BD Biosciences, USA). scFv DNA, that was isolated by plasmid extraction from the cells collected, was electroporated into E. coli DH5α and subjected to another round of sorting by FACS. When the antibodybinding population reached 60%, colonies were picked at random, screened by FACS and subjected to DNA sequencing. Spheroplasts expressing random scFv of rainbow trout were used as a negative control for the FACS analysis. 2.5. Expression and purification of scFv antibodies Expression vector pET-27b(+) was used to express the scFv-1 clone that had the highest MFI. The expression and purification of the scFv antibody was performed according to previous studies (Xu, Li, Zhou, Guo,

Fig. 1. The antigen-antibody co-expression display system. (a) Libraries of scFv antibodies and antigen protein are co-expressed in the periplasm of E. coli and antibodies that are specific to the antigen will tether the antigen to the inner membrane of the bacteria by CDQSSS. (b) After outer membrane permeabilization, the scFv antibody-antigen complexes anchored on the inner membrane bind with fluorescently labeled antigen specific antibodies. (c) Enrichment of spheroplasts expressing antigen specific antibody by gating the region defined by the distinct scatter of the spheroplasts (FSC and SSC) and the high FITC-A signal. FSC, forward scatter; SSC, side scatter; CDQSSS, NlpA amino acids 1–6.

Liu, Zhao, Cao, Li, 2014). The scFv-1 was expressed with a His 6 tag at the 3′ end. 2.6. Western blotting Purified scFv-1 protein and bovine serum album (BSA) were electrophoresed in a native-SDS-PAGE and transferred onto a nitrocellulose membrane. The membrane was blocked with 5% (m/v) skim milk in PBS-Tween20 (PBS-T) buffer. The blots were incubated with 100 μl of the IHNV glycoprotein (Xu et al., 2013) in PBS-T (40 μg ml−1) at 4 °C overnight, followed by incubation with rabbit anti-full length IHNV glycoprotein polyclonal antibody (1:200) for 1 h. The final incubation was with HRP-goat anti-rabbit antibody (1:7500) at room temperature for 1 h. The blots were developed using the ECL Detection System.

34

L.-M. Xu et al. / Aquaculture 460 (2016) 32–36

Fig. 2. Screening of glycoprotein-specific scFv antibodies by FACS. A total of 10,000 spheroplast events are shown for each round of screening. M: mean fluorescence intensity. Control: spheroplasts displaying random scFv of rainbow trout. Three rounds of enrichment of populations with high fluorescence intensity. There is obvious increase of fluorescence intensity along the x axis after three rounds of enrichment.

2.7. ELISA analysis 96-well microtiter plates were coated with 100 μl of purified scFv-1 antibodies at different concentrations including 20 μg ml−1, 2 μg ml−1, 0.2 μg ml−1, 20 ng ml− 1 and 2 ng ml−1 in NaHCO3/NaCO3 buffer (pH 8.7) overnight at 4 °C. Triplicate samples were washed three times with PBS-T (0.5%, v/v) and 5% skim milk in PBS was used to block the remaining non-specific binding sites at 37 °C for 2 h. After washing, each well was incubated with 100 μl IHNV-Sn1203 (107 pfu ml−1) at 37 °C for 1 h, followed by rabbit anti-glycoprotein polyclonal antibody (200-fold dilution) and HRP-goat anti-rabbit antibody (7500-fold dilution). Three negative controls were included: control 1 was without IHNV-Sn1203, control 2 was without rabbit antiglycoprotein antibody, and control 3 was without IHNV-Sn1203 and rabbit anti-glycoprotein antibody. The assay was developed using a TMB solution and the development of color product was terminated by the addition of 50 μl of 2 M H2SO4. The absorbance of each well was measured by an ELISA reader at a wave length of 450 nm. 2.8. IFAT Monolayers of EPC cells grown on microscope slides in a 6-well cell culture plate at 25 °C were inoculated with each of 6 IHNV isolates at the multiplicity of infection (MOI) of 0.01 and incubated at 15 °C. The isolates used in this test were from different provinces of China and included Sn1203, LN12, SD-12, GS-12, YN-13, XJ-13 that were described above. When the monolayers on the slides showed early cytopathic effect (CPE) the slides were used to perform an IHNV-specific IFAT described previously (Zhao et al., 2014). However, in this IFAT procedure, the purified scFv-1 was used as the first antibody and the CY3-mouse anti-His tag was used as the second antibody.

sequenced and blast searched in the Genbank database. Each of the 5 clones had high homology with the salmon immunoglobulin that supported the good quality of the scFv antibody library. Additionally, each of the 5 clones had a unique sequence that indicated the wide diversity of the scFv antibody library. 3.2. Screening of scFv antibodies by FACS After removal of the outer membrane of E. coli, spheroplasts were incubated with 200 nM of the FITC labeled rabbit anti-glycoprotein antibody. Spheroplasts expressing random scFv of the rainbow trout were used as a negative control. In the first round of sorting, approximately 5.2 × 104 spheroplasts (1.3% of the highest events) were isolated in 3 h from the library. Approximately 7.8 × 104 events (1% of the highest events) were isolated from the second round of sorting and 2.5 × 105 events (0.7% of the highest events) from the third round of sorting. After three rounds of sorting, there was an obvious fluorescence population shift in the fluorescein emission (FITC-A) channel indicating successful enrichment of a highly fluorescent population (Fig. 2). Seven unique clones were identified by FACS analysis and were designated as scFv1–scFv7. The fluorescence intensity was different between the seven clones: scFv-1 had the strongest fluorescence intensity and scFv-2 had the weakest, which was still much stronger than the negative control (Fig. 3). 3.3. Characterization of isolated scFv

3. Results

SDS-PAGE analysis showed that purified scFv proteins were approximately 26 kDa (Fig. 4a). Western blotting analysis demonstrated that the scFv antibody isolated was specifically recognized by the glycoprotein (Fig. 4b). ELISA analysis demonstrated that the antibody clone could specifically bind to the IHNV-Sn1203 in an antibody concentration dependent manner (Fig. 5).

3.1. Construction of scFv library

3.4. IFAT

The scFv genes and G4 were cloned into the antigen-antibody coexpression vector pCoex to generate the NlpA-scFv fusion and pelBG4 fusion, and were electrotransformed and resulted in 8.5 × 10 7 transformants. Five clones that were randomly selected were

To determine the binding ability of the scFv antibody with the natural glycoprotein of IHNV, the scFv was used as the first antibody in an IHNV-specific IFAT. The results showed that EPC cells inoculated with each of the IHNV isolates and with the scFv antibody gave a strong

L.-M. Xu et al. / Aquaculture 460 (2016) 32–36

35

Fig. 3. The specificity and affinity of the clones detected by FCM. M: mean fluorescence intensity. The M value ranks as scFv-1 N scFv-3 N scFv-7 N scFv-6 N scFv-4 N scFv-5 N scFv-2 N Control.

specific fluorescent signal. No fluorescent signal was observed in negative control groups (Fig. 6). The results indicated that the scFv antibody could recognize all the IHNV isolates that had been obtained from different districts of China. The results strongly supported the potential application of this antibody as a diagnostic regent for detecting and confirming IHNV strains prevalent in China.

Genetically engineered antibodies have a great potential in clinical diagnosis and the treatment of diseases (Chen et al., 2010; Staneloudi et al., 2007). In this study, an antigen and antibody coexpression system was established. Compared with the previously established bacterial display system, the in this system purified antigen is not required which is a significant improvement. With this system, antigen and antibody are co-expressed within the periplasm of E. coli. Using FCM, the system provided a high throughput method to obtain a panel of antibodies against target proteins in a relatively short period of time. Previous studies have proven that proteins will be expressed at similar concentrations in the bacteria's periplasm and accumulate at sufficiently high concentrations to allow labeling by fluorescence in every single E. coli cell (Jeong et al., 2007). The MFI shown by FACS analysis

indicates the binding affinity of the individual antibody. To obtain the highest affinity antibody to IHNV, the clone with the highest MFI must be selected, expressed and purified as was done in this study. This particular scFv antibody expressed very well in bacteria and was very stable against the IHNV glycoprotein and the virus itself. This antibody was also shown applicable in ELISA, immunoblotting or FACS analysis even without any affinity maturation. Additionally, this antibody could cross-reacted with each of the IHNV isolates that were obtained from different areas of China. Previous studies have shown that the IHNV isolates in China belonged to the J genogroup (Jia et al., 2014) and the glycoprotein homology of the Chinese IHNV isolates was as high as 99%. In a previous study (Xu, Liu, Zhao, Cao, Yin, Liu, Lu, 2014), the G4 fragment (271–450 aa) of IHNV-Sn1203 was shown to have the strongest immunogenicity. In order to reduce the burden on the host E. coli cell and get soluble protein, the G4 fragment instead of the full length glycoprotein, was used as a target antigen to screen IHNV-specific antibody. The results of the IFAT analysis further confirmed that the G4-targeting scFv antibody could bind to IHNV isolates from different areas of China. This suggested that this antibody could be utilized as a potential universal reagent for diagnosis of IHNV throughout China. Furthermore, the IFAT results indicated that there is no obvious variation in serological characteristics among the Chinese IHNV isolates although the differences do exist in the glycoprotein of IHNV.

Fig. 4. SDS-PAGE and Western blotting. (a) SDS-PAGE analysis of the purified scFv-1 antibody. M: unstained protein marker. (b) Western blot analysis of the scFv-1 antibody. M: unstained protein marker; Control: bovine serum albumin.

Fig. 5. ELISA analysis of the scFv antibody against IHNV-Sn1203. The data represent the means ± S.D. of quadruple samples. **p b 0.01 vs negative control; *p b 0.05 vs negative control. NC1: without IHNV-Sn1203; NC2: without rabbit anti-full length IHNV glycoprotein polyclonal antibody; NC3: without IHNV-Sn1203 and rabbit anti-full length IHNV glycoprotein polyclonal antibody.

4. Discussion

36

L.-M. Xu et al. / Aquaculture 460 (2016) 32–36

Fig. 6. Use of the scFv-1 antibody in the detection of different IHNV isolates by immunofluorescent antibody technique. The EPC cell line was inoculated with different IHNV isolates and was incubated with the purified scFv-1 antibody followed by the CY3-mouse anti-His tag antibody. NC1: EPC cells incubated with the CY3-mouse anti-His tag antibody; NC2: EPC cells inoculated with IHNV isolates and incubated with CY3-mouse anti-His tag antibody.

In summary, the antigen-antibody co-expression display technology in combination with FACS can be used for quantitative and real time selection of desirable antibody clones to target antigens. The positive clones can be enriched up to 80% after three rounds of screening, and a panel of anti-IHNV scFvs antibodies was produced in this study within one month. This technique provides a novel and high throughput strategy for the rapid development of antibodies against numerous targets. The establishment of relevant techniques to increase the antibody repertoire in fish research could significantly assist in understanding this and other diseases and aspects such as pathogen virulence, host immune response(s), and host-pathogen interactions and along with assisting in the development of potential therapeutic reagents and/or prevention programs. Acknowledgements This study was supported by grants from the Heilongjiang Province Research and Development of Applied Technology (GA13B401), the Natural Science Foundation of Heilongjiang Province (C201462), and the Central-Level Non-profit Scientific Research Institutes Special Funds (HSY201410). We show our thanks to Dr. Scott E lapatra for his generous help in the revision of the manuscript. References Breyta, R., Jones, A., Stewart, B., Brunson, R., Thomas, J., Kerwin, J., Bertolini, J., Mumford, S., Patterson, C., Kurath, G., 2013. Emergence of MD type infectious hematopoietic necrosis virus in Washington state coastal steelhead trout. Dis. Aquat. Organ. 104, 179–195. Chen, Y., Zhu, X., Zhang, X., Liu, B., Huang, L., 2010. Nanoparticles modified with tumortargeting scFv deliver siRNA and miRNA for cancer therapy. Mol. Ther. 18, 1650–1656. Enzmann, P.J., Kurath, G., Fichtner, D., Bergmann, S.M., 2005. Infectious hematopoietic necrosis virus: monophyletic origin of European isolates from North American genogroup M. Dis. Aquat. Organ. 66, 187–195. Jeong, K.J., Seo, M.J., Iverson, B.L., Georgiou, G., 2007. APEx 2-hybrid, a quantitative protein–protein interaction assay for antibody discovery and engineering. Proc. Natl. Acad. Sci. U. S. A. 104, 8247–8252.

Jia, P., Zheng, X.C., Shi, X.J., Kan, S.F., Wang, J.J., He, J.Q., Zheng, W., Yu, L., Lan, W.S., Hua, Q.Y., Liu, H., Jin, N.Y., 2014. Determination of the complete genome sequence of infectious hematopoietic necrosis virus (IHNV) Ch20101008 and viral molecular evolution in China. Infect. Genet. Evol. 27, 418–431. Kim, W.S., Oh, M.J., Nishizawa, T., Park, J.W., Kurath, G., Yoshimizu, M., 2007. Genotyping of Korean isolates of infectious hematopoietic necrosis virus (IHNV) based on the glycoprotein gene. Arch. Virol. 152, 2119–2124. Kolodziejek, J., Schachner, O., Durrwald, R., Latif, M., Nowotny, N., 2008. “Mid-G” region sequences of the glycoprotein gene of Austrian infectious hematopoietic necrosis virus isolates form two lineages within European isolates and are distinct from American and Asian lineages. J. Clin. Microbiol. 46, 22–30. Nishizawa, T., Kinoshita, S., Kim, W.S., Higashi, S., Yoshimizu, M., 2006. Nucleotide diversity of Japanese isolates of infectious hematopoietic necrosis virus (IHNV) based on the glycoprotein gene. Dis. Aquat. Organ. 71, 267–272. Niu, L., Zhao, Z., 1988. The epidemiology of IHN and IPN of rainbow trout in Northeast China. J. Fish. China 12, 327–332. Rudakova, S.L., Kurath, G., Bochkova, E.V., 2007. Occurrence and genetic typing of infectious hematopoietic necrosis virus in Kamchatka, Russia. Dis. Aquat. Org. 75, 1–11. Staneloudi, C., Smith, K.A., Hudson, R., Malatesti, N., Savoie, H., Boyle, R.W., Greenman, J., 2007. Development and characterization of novel photosensitizer: scFv conjugates for use in photodynamic therapy of cancer. Immunology 120, 512–517. Troyer, R.M., LaPatra, S.E., Kurath, G., 2000. Genetic analyses reveal unusually high diversity of infectious haematopoietic necrosis virus in rainbow trout aquaculture. J. Gen. Virol. 81, 2823–2832. Xu, L.M., Liu, H.B., Yin, J.S., Lu, T.Y., 2013. Prokaryotic expression and immunogenicity analysis of glycoprotein from infectious hematopoietic necrosis virus. Bing Du Xue Bao. 29, 529–534. Xu, L.M., Li, T.H., Zhou, B., Guo, M., Liu, M., Zhao, J.Z., Cao, H.W., Li, D.S., 2014a. scFv antibodies against infectious bursal disease virus isolated from a combinatorial antibody library by flow cytometry. Biotechnol. Lett. 36, 1029–1035. Xu, L.M., Liu, M., Zhao, J.Z., Cao, Y.S., Yin, J.S., Liu, H.B., Lu, T., 2014b. Epitope mapping of the infectious hematopoietic necrosis virus glycoprotein by flow cytometry. Biotechnol. Lett. 36, 2109–2116. Xu, L., Zhang, Y., Wang, Q., Zhao, J., Liu, M., Guo, M., Jiang, Y., Cao, H., Li, Q., Ren, G., Li, D., 2015. Bi-specific antibodies with high antigen-binding affinity identified by flow cytometry. Int. Immunopharmacol. 24, 463–473. Zhao, J., Xu, L., Liu, M., Cao, Y., Yin, J., Liu, H., Lu, T., 2014. Expression and immunogenicity analysis of truncated glycoprotein of infectious hematopoietic necrosis virus in fish. Xi Bao Yu Fen Zi Mian Yi Xue Za Zhi. 30, 1238–1242. Zhou, Y., Xie, Z.G., 2015. High throughput screening of scFv antibodies against viral hemorrhagic septicaemia virus by flow cytometry. J. Virol. Methods 219, 18–22.