Isolation of Goose-origin scFv Antibodies Against Goose Parvovirus from Bacterial Display Antibody Libraries

Zhou Jin-xin, Zhang Xiao-yu, Wang Yu-yang, Huang Tao, Guo Xiao-chen, Li De-shan, , Ma Bo, and Ren Gui-ping, *

1 Biopharmaceutical Laboratory, College of Life Sciences, Northeast Agricultural University, Harbin 150030, China

2 Key Laboratory of Agricultural Biological Functional Gene, Northeast Agricultural University, Harbin 150030, China

3 College of Veterinary Medicine, Northeast Agricultural University, Harbin 150030, China

Abstract: Goose parvovirus (GPV) can cause a highly contagious and fatal gosling plague (GP) disease in goslings and muscoy ducklings. Here, three goose-origin neutralizing single chain variable fragment (scFv) antibodies against GPV SYG-61 were isolated. The genes of scFv antibodies were derived from goslings immunized with GPV SYG-61, and scFvs were subcloned into a pBSD vector for the construction of pBSD-scFv libraries. The pBSD-scFv libraries were screened following three rounds using VP2 (protective antigen of GPV) as the bait by flow cytometry (FCM). After screening, the 15 clones with high mean fluorescence intensity (MFI) were isolated and sequenced. These 15 scFvs were expressed by pET-28a (+) in E. coli. The specificity and affinity of the 15 purified scFvs were successfully confirmed by ELISA. In the preliminary neutralization experiment on primary goose embryo fibroblast (GEF) in vitro, three of the 15 purified scFvs (named scFv-10, scFv-11 and scFv-50) showed significant neutralizing capacities. The study generated the first goose-origin neutralizing scFv against GPV and laid the foundation for the appearance of fulllength goose-origin neutralizing monoclonal antibody against GPV.

Key words: goose parvovirus (GPV), scFv, bacterial surface display technology, antibody library

Introduction

Goose is an economically significant waterfowl that provides meat, eggs and fat for human in consumption. With the increasing of breeding quantity and density, Gosling plague (GP) has been a prominent disease which restricts the development of goose breeding industry globally (Takeharaet al., 1995; Janssonet al., 2007).

Goose parvovirus (GPV) is the etiological agent of GP which infects goose and Muscovy duck. GPV is self-replicating parvovirus of the genusDependovirusof the parvovirus family. The genome of GPV comprises of two open reading frames (ORFs). The left ORF encodes the nonstructural proteins, and the right ORF encodes three capsid proteins of VP1, VP2, and VP3. The capsid proteins are the key factors that determine the tropism, host domain and pathogenicity of GPV (Zádoriet al., 1995). Compared with other parvovirus, GPV can agglutinate with bovine spermatozoa, but not red blood cells of poultry and nursing animals. GPV can be exclusively propagated in goose embryo, Muscovy duck embryo, goose embryo fibroblast and Muscovy duck embryo fibroblast cells (Brownet al., 1995). Similar to other parvovirus, the capsid protein VP2 is the functional and immunological region of GPV that performs as viral protective antigen. VP2 can be self-formed as viruslike particles in the absence of VP1 and stimulates the organism to induce neutralizing antibodies (Yuet al., 2011; Chenet al., 2012; Tuet al., 2015). Therefore, VP2 can be used as an advantageous bait for research strategies aiming at screening of neutralizing scFvs against GPV.

At present, the treatment of GP mainly relies on hyper immune serum and egg yolk antibody. However, there are lots of problems that can't be ignored from production to usage (Calenk, 1991). The productions of these therapeutic agents are limited because there are rare SPF geese and laying and hatching of goose eggs had to follow the reproductive characteristics of seasonal breeding. What's more, these agents have a high risk of carrying other infectious diseases and harming to more geese. The commercial yolk antibody is mostly chicken source, which also has the problem of immune rejection (Shiet al., 2007; Staak, 1996 ). In recent years, with the development of immunological technology, monoclonal antibody against GPV has been the hot spot in researches (Zhaoet al., 2011). However, the activity of monoclonal antibody is limited by the differentiate various strains, and its origin from heterologous host which makes the actual treatment effect decreased sharply. Above all, there is an urgent demand to obtain a natural host antibody to control the spread of GPV.

Currently, the bacterial surface display (BSD) technology has been widely used in isolating high affinity antibodies from the natural-host antibody libraries (Xuet al., 2014; Baiet al., 2016). In this study, scFv BSD antibody libraries from the spleens of GPV immune geese were constructed. After combination with flow cytometry (FCM), the high GPVspecific scFvs were screened and characterized using sequencing and ELISA. Further, the GPV-specific scFvs neutralization activity was confirmed on goose embryo fibroblasts (CEFs). The aim was to provide theoretical and experimental bases for developing a neutralizing and goose natural antibody against GPV.

Materials and Methods

Materials

The primers are listed in Table 1. The plasmid contained theGPV-VP2 gene (submission to GenBank accession No. KC996729.1) was isolated from GPV strains by biopharmaceutical lab of Northeast Agricultural University. The pSUMO vector containing the 6x His-tag in front of SUMO leader and SUMO protease-I was constructed by biopharmaceutical lab of Northeast Agricultural University. These were used to prepare the VP2 protein of GPV. Bovine serum albumin (BSA) was used as a negative control. FITC antibody labeling kit (ThermoFisher Scientific, Cat. No.53027) was used to label VP2.

The goslings and goose embryos were gifts from Harbin Qinglong Goose Industry Development Co., Ltd. The GPV allantoic fluid (GPV SYG-61 strain, TCID50was 10-4.63/0.1 mL) was stocked in biopharmaceutical lab of Northeast Agricultural University. Goose GPV Ab ELISA KIT (Promega Corporation, Cat. No. YS05336B) was used to analyze the antibody levels after immunization. Gosling plague vaccine (SYG41-50 strain) was purchased from Yangzhou Vacbio Bio-Engineering, China.

E. coliDH5αand two vectors for scFv antibodies library were used.E. coliDH5αwere purchased from Shanghai Weldi Biotechnology Co., Ltd. One vector was pYDx containingscFv-peptide-linker (GGGGSG GGGSGGGGS) gene, the other was pBSD vector containing the sequence of NlpA leader+6aa (CDQSSS). Those were generated in biopharmaceutical lab of Northeast Agricultural University (Xuet al., 2014). The vector pET-28a (+)(Novagen, USA) andE. coliBL21(DE3) were used to express the scFv proteins, meanwhile,E. coliBL21(DE3) were also used to express VP2.E. coliBL21(DE3) were purchased from Shanghai Weldi Biotechnology Co., Ltd. The plasmid and protein of anti-HBV scFv, the rabbit anti-GPV VP2 polyclonal antibody and the HRP-goat anti-rabbit antibody (eBiosscienc, USA) were stored in biopharmaceutical lab of Northeast Agricultural University. These were used for characteristics of scFvs. Dulbecco's modified Eagle's medium (DMEM) and fetal bovine serum (FBS) were purchased from Gibco. DMEM and FBS were used to culture the cells.

Table 1 Primer sequences

Methods

Preparation and confirmation of VP2

Using specific primers,VP2 gene fragment of GPV was amplified by PCR from GPV-VP2 plasmid. Then the PCR products ofVP2 were inserted into the downstream of SUMO leader in the pSUMO vector usingBsaI/BamH I. The recombinant plasmid was named pSUMO-VP2. After confirmation by DNA sequencing, the correct plasmid of pSUMO-VP2 was transformed intoE. coliBL21 (DE3). When culturing OD600nmwas between 0.4 and 0.6, the expression of the His-SUMO tagged VP2 protein was induced by iso-propylβ-D-thiogalactoside (IPTG) at 37℃ for 4 h, and then the SUMO-VP2 protein was purified by Ni-Sepharose affinity chromatography. His-SUMO tag was removed by SUMO protease-I digestion and Ni-Sepharose affinity chromatography. SDS-PAGE was used to analyze the purity ofVP2.

VP2 and unrelated BSA subjected to SDS-PAGE and were transferred onto a nitrocellulose membrane. The membrane was blocked with 5% (w/v) skimmed milk in PBS buffer at 37℃. After 3 h, the membrane was incubated with 100 μL rabbit anti-GPV VP2 polyclonal antibody (40 μg • mL-1) in PBS at 4℃ overnight. Then, the membrane was washed three times with PBS-T (PBS, 0.5% Tween20) and incubated with HRP-goat anti-rabbit antibody (1 : 7 500)for 1 h at 37℃. The membrane was developed using an ECL detection system.

The correct and purifiedVP2 was fluorescently labeled following the manufacturers protocol of Pierce FITC antibody labeling kit.

Immunization of experimental animals

All the animal experiments were performed in accordance with the recommendations in theGuide for the Care and Use of Laboratory Animalsof the National Institutes of Health. The protocol was approved by the Institutional Review Board of the Northeast Agricultural University Institute of Bio-medicine. Thirty healthy one-day-old female goslings were identified maternal antibody levels using Goose GPV Ab ELISA kit and following the manufactures protocol. Then,six goslings (named G-2, G-3, G-9, G-11, G-27 and G-28) with lower maternal antibody levels were vaccinated with gosling plague vaccine. Then, six goslings were vaccinated with the 200 μL GPV allantoic fluid by routine procedures every 7 days, twice. And their anti-GPV antibody levels were measured before each injection. On the 3rd day after the 4th injection, spleens of goose were isolated after determining the finally antibody levels.

Construction of anti-GPV scFv antibody bacterial display library

The total RNA of goose was extracted using Trizol (Invitrogen, Karlsruhe). The cDNA was synthesized from the total RNA sample using Superscript II (Invitrogen) and random hexamer oligonucleotide primers. Using cDNA and library degenerate primers, the variable regions of heavy chain (VH) and light chain (VL) genes were amplified by PCR. After enzyme digestion and ligation,VH(withHindIII andNheI) andVL(withBamH I andXhoI) genes were inserted into up and down stream of peptide-linker in the pYDx vector, scFv antibody library plasmid was constructed; then, scFv gene fragments were digested (withSfiI) and inserted into the display vector pBSD to construct pBSD-scFv plasmid. According to standard procedures, pBSDscFv plasmid was electroporated intoE. coliDH5α. The colonies were counted to calculate the transformation efficiency of electroporation and identified specificity by PCR, according to standard procedures (Ahmadet al., 2014).

Preparation of spheroplasts and screening of library

After electroporation, all the pBSD-scFv library-colonies were cultured in Luria-Bertani medium (LB). When the culturing OD600nmwas approximately between 0.4 and 0.6, library was induced with 0.25 mmol • L-1IPTG at 37℃ for 4 h. For spheroplasts preparation, firstly, 1 mL bacterial cells was collected at 12 000 g for 1 min and washed with PBS (pH 7.4). Secondly, the cells were re-suspended in 350 mL of ice-cold solution of 0.75 mol • L-1sucrose and 0.1 mol • L-1Tris-HCl(pH 8.0), then the cells was added 35 μL lysozyme(60 mg • mL-1) and stored at 4℃ for 10 min. Thirdly, cells were treated with 700 μL of ice-cold EDTA (1 mmol) and 50 μL of MgCl2(0.5 mol) at 4℃ for 15 min. Fourthly, the cell pellets were collected at 12 000 g for 1 min and re-suspended in 100 μL PBS gently, and incubated with 2 μL FITC-VP2 (1 mg • mL-1) and 1 μL 1% BSA at 4℃ for 1 h. Finally, the cells were washed twice with PBS and re-suspended in 300 μL PBS for sorting by FACS (BD FACS AriaTMCell Sorter 334078, BD Biosciences). Isolated cells were rescued by plasmid extraction from the collected library and the plasmid electroporated intoE. coliDH5αto subject another round of sorting by FACS. Until the VP2-binding population reached 80%, colonies were picked up at random, grown and prepared as spheroplasts as described above, then analyzed by FCM. The scFv was sequenced for clones with a high mean fluorescence intensity (MFI). The plasmid of anti-HBV scFv was transformed intoE. coliDH5αas a negative control.

Construction, expression and purification of isolated scFv antibodies

The genes of isolated 15 scFv antibodies were amplified by PCR and subcloned into the expression vector pET-28a (+) to construct the pET-scFv plasmids. Then, the plasmids were transformed intoE. coliBL21 (DE3) and cultured in LB. When the culturing OD600nmwas between 0.4 and 0.6, cells were induced with 0.5 mmol • L-1IPTG at 37℃ for 4 h. The 15 antibody cells were expressed as inclusion bodies. Centrifuged and collected the inclusion bodies at 10 000 g for 30 min and then re-suspended in solution buffer (2 mol • L-1urea). Then, washing twice, the inclusion bodies were dissolved in denatured solution (8 mol • L-1urea) at 4℃ overnight. Ten volumes of the renaturation solution (2 mol • L-1urea) were added into the denatured solution slowly and stored at 4℃ for 24 h. The supernatant was harvested by centrifugation and desalted in PBS overnight. SDS-PAGE was used to analyze the purity of the recombinant proteins.

Sandwich ELISA

For ELISA, 96-well microtiter plates were coated with 100 μL purified scFv antibody proteins at different concentrations (500, 100, 50, 10, 1 and 0.1 μg • mL-1in NaHCO3/NaCO3buffer pH 8.7) overnight at 4℃. PBS-T was washed three times, and the remaining nonspecific binding sites were blocked by 5% skimmed milk in PBS at 37℃ for 3 h. After washing, each well was incubated with 100 μL GPV allantoic fluid at 37℃ for 1 h, followed by rabbit anti-GPV VP2 polyclonal antibody (1 : 3 000) and HRP-goat anti-rabbit antibody(1 : 7 500). Three controls included control 1 (without GPV allantoic fluid), control 2 (without rabbit anti-GPV VP2 polyclonal antibody) and control 3 (coated with anti-HBV scFv antibody protein). The assay was used TMB solution and the color product development, which was terminated by the addition of 50 μL 2 mol • L-1H2SO4. The absorbance of each well was measured by an ELISA reader at 450 nm.

Preparation of goose embryo fibroblasts (GEFs) and neutralizing activity of antibodies

GEFs were obtained from the 12-day-old goose embryonated eggs, according to standard procedures (Zi and Jo, 1981) and cultured with DMEM and 10% FBS at 37℃ in the presence of 5% CO2. When the cells filled the plate, it was digested and counted. Then, the cell suspension was added to the 96-well containing 105cells each well, overnight. The 50 μL purified scFv antibody proteins in Log2 dilutions (from 500 to 1.726 μg • mL-1) were mixed with 50 μL GPV allantoic fluid (100 TCID50) at 37℃ for 1 h. Goose embryo fibroblast cells were incubated with the mixtures for 1 h at 37℃ in the presence of 5% CO2. After removing the mixture, the cell was cultured with maintenance solution (DMEM+1% FBS) at 37℃, 5% CO2for 144 h. The cytopathogenic effect (CPE) was observed by using an inverted microscope. Three controls included control 1 (cells only were treated with 100 μL GPV allantoic fluid), control 2 (cells were treated with 100 μL DMEM) and control 3 (anti-GPV scFv antibody was replaced by anti-HBV scFv protein). The experiment was repeated three times independently.

Results

Expression, purifciation and activity analysis of VP2

VP2 gene fragment was correctly constructed to pSUMO and DNA sequencing showed the gene was VP2 sequence (GenBank accession No. KC996729.1). SDS-PAGE assay showed that purified VP2 was soluble and approximately 85 ku. After removing the His-SUMO tag, the recombinant VP2 was approximately 66 ku and the purity was more than 90% (Fig. 1). Western blotting analysis demonstrated that VP2 was specifically recognized by rabbit anti-GPV VP2 polyclonal antibody, whereas the negative controls were negligible (Fig. 2).

Fig. 1 SDS-PAGE analysis of purified recombinant VP2 protein and scFv antibodies(a) M, Prestained protein Marker; Lane 1, SUMO-VP2; Lane 2, VP2; Lane 3, scFv-2; Lane 4, scFv-4; Lane 5, scFv-10; Lane 6, scFv-11; Lane 7, scFv-12; Lane 8, scFv-18; (b) M, Prestained protein Marker; Lane 1, scFv-19; Lane 2, scFv-21; Lane 3, scFv-24; Lane 4, scFv-41; Lane 5, scFv-44; Lane 6, scFv-47; Lane 7, scFv-49; Lane 8, scFv-59; Lane 9, scFv-91.

Fig. 2 Identification of recombinant purified VP2 by western blottingVP2 and BSA (negative control) are subjected to SDS-PAGE and transferred onto membrane, then the membrane is incubated with rabbit anti-GPV VP2 polyclonal antibody, followed by incubation with HRPgoat anti-rabbit antibody.

Determination of anti-GPV antibody concentration and construction of scFv antibody bacterial display library

The anti-GPV antibody levels of goose were increased in the immune process (Table 2). Finally, anti-GPV antibody levels of the six geese were 77.6, 53.2, 96.3, 53.0, 64.0 and 53.1 μg • L-1, respectively. From the mixed cDNA of the six goslings, theVHandVLgenes were successfully inserted into the pBSD and confirmed by PCR (Fig. 3). The pBSD-scFv plasmids were electro-transformed intoE. coliDH5αand 8.9×107transformants were obtained. The random-collected colonies sequencing results showedscFvgenes were sequences of goose (Anser).

Table 2 Antibody levels of six geese analyzed by goose GPV Ab ELISA in immunized periods

Fig. 3 Identification of scFv antibody library by PCRM, DL2000 DNA Marker; Lane 1, VL gene; Lane 2, VH gene; Lane 3,scFv gene. Molecular sizes of PCR fragments are shown in bottom of lane.

Screen of anti-GPV scFv antibodies by FACS

The bacterial display library was subjected to three rounds of screening and VP2-binding population increased to 28.0%, 55.1% and 81.8%, respectively (Fig. 4). The library showed a steady increase in Agspecific binding level by each round of screening, indicating successful enrichment of a highly fluorescent population. Spheroplast expressing anti-HBV scFv antibody was used as a negative control. After three rounds of screening, 15 fluorescence intensity clones were randomly selected by FACS and sequenced. The 15 scFvs were named as scFv-2, scFv-4, scFv-10, scFv-11, scFv-12, scFv-18, scFv-19, scFv-21, scFv-24, scFv-41, scFv-44, scFv-47, scFv-49, scFv-59 and scFv-91, respectively (Fig. 4).

Characterization of 15 purified anti-GPV scFv antibodies

ThescFvgene fragments were correctly constructed to pET-28a (+) and expressed as inclusion bodies. After denaturing and refolding, SDS-PAGE analysis showed that the purified scFv proteins were approximately 35 ku and the proteins' purity was more than 90%(Fig. 1). ELISA analysis demonstrated that all the clones also could specifically bind to the GPV and produce positive ELISA readings at 450 nm (Fig. 5). And both of the two tests were significantly higher than the negative control.

Neutralization of GPV infectivity

GEFs were successfully prepared. The newly prepared GEFs had a typical spindle shape with an oval nucleus in the center. The cells showed a distinct radiate distribution and had clear boundaries (Zi and Jo, 1981). After virus neutralization assay, three purified scFv-10, scFv-11 and scFv-59 showed obvious neutralization to block CPE on CEFs. The minimal concentrations of neutralization for scFv-10, scFv-11 and scFv-59 were 1.954, 15.625 and 31.25 μg • mL-1, respectively. Whereas, the anit-HBV scFv antibody showed no neutralizing properties (Table 3).

Fig. 4 Screening and identification of VP2-specific scFvs by FACS and 15 anti-GPV clones detected for specificity and affinity by FCM A total of 10 000 spheroplasts events are shown. Control is spheroplasts displaying anti-HBV scFv antibody. Three rounds of enrichment of populations with high fluorescence intensity. Fifteen anti-GPV clones are detected for specificity and affinity by FCM. Negative control is anti-HBV scFv antibody in spheroplasts. M means mean fluorescence intensity. M values of all the clones are above those of negative control.

Fig. 5 ELISA analysis of purified scFv antibodies against GPV Data represent mean±S.D. of quadruple samples. **p＜0.01 vs negative control. N1 without GPV allantoic fluid, N2 without rabbit anti-GPV VP2 polyclonal antibody, N3 coated with HBV scFv antibody. Data are mean of three independent experiments.

Table 3 Virus neutralization titers of scFv antibodies against GPV (SYG-61 strain)

Discussion

In China, goose-breeding is the traditional and characteristic industry. Chinese goose marketable fattened stock was at the forefront of the world every year (Changet al., 2007). Geese had the good qualities of growing fast and higher disease-resistance comparing with other poultries. Nevertheless, GPV infectious disease had affected the stable development of geese industry (Guémené, 2012). As an acute, fulminating and high fatal disease, goslings between 4 and 20 days of age were easier to suffer GPV infection (Derzsy, 1967). For the prevention and control of GPV infection, attenuated live and inactivated vaccines had achieved some results (Takeharaet al., 1995; Denget al., 2014). They aimed to give high antibody levels for breeders so that goslings obtained good maternal immunization. However, multiple injections for breeders made the vaccination schedule expensive and time-consuming. More importantly, goslings would be vulnerable once the maternal antibody level drops in the gosling susceptible age (Irvine and Holmes, 2010). The attenuated vaccine was unstable and prone to return to the ancestor independently, and virulent vaccines could cause the risk of virus spreading (Chenget al., 2005). For the therapy of GP-infected goslings, hyper immune serum and egg yolk antibody played a certain degree of protection effect as passive immunization (Linet al., 2010). Homology hyper immune serum was low output and high cost, and worse to exist risk of vertical transmission diseases. Despite heterologous hyper immune serum was high yield and low cost, it had the defect of immune rejection. Egg yolk antibody generally could not provide enough titer for treatment and its absorption by goslings was not ideal. More important, different batches of hyper immune serum and egg yolk antibody qualities could not be unified. Nonstandard production, improper storage and transportation would have a great impact on the effect of them (Schadeet al.,2005; Irvine and Holmes, 2010). This study used genetic engineering technology to create recombinant antibodies against GPV. Unlike the traditional polyclonal antibody, scFv antibody as a nature and genetic recombination antibody established superiority on no immune rejection. Moreover, there was no worry about the contamination of infectious pathogens during the production process and seasonal restrictions on production. At the same time, it could be easily to establish the quality control procedures (Ahmad, 2012). Lots of scFv antibodies had been shown to possess neutralizing capacity against viruses, such as vesicular stomatitis virus (Cortayet al., 2006), canine parvovirus (Yuanet al., 2000) and infectious bursal disease virus (Sapats, 2003). Therefore, the generation and use of goose-origin therapeutic antibody to against GPV will hold many advantages in the future.

For scFv screening, the scFv proteins were expressed outside of the inner membrane because of a new lipoprotein (NlpA) leader peptide and anchored by six amino acids (CDQSSS) (Harveyet al., 2004). After removal of the outer membrane, the fluorescent antigen FITC-VP2 could enter the periplasmic space. The identification of scFv antibody with high antigen-binding affinity from library could be greatly enhanced by FCM. After incubating with FITC-VP2, the VP2 specific binding scFv could be recognized by FCM. Because the protein was displayed on the inner membrane, the amount of protein could not lost during perforation of the outer membrane. The BSD technology could keep genotype and phenotype consistent (Fig. 6) (Chenet al., 2016; Junget al., 2007). The study showed that anti-GPV bacteria antibody libraries were enriched up to 80% only three rounds and the positive scFv clones with different sensitivities were isolated by bacteria display technology within two months.

Furthermore, the 15 scFv antibodies showed sensitivity and affinity to GPV SYG-61 virus. At the same time, using these scFv antibodies, the virus neutralization experiments on GEF cells were performed, and the three scFv (scFv-10, scFv-11 and scFv-59) antibodies effectively exhibited virus infection. In this study, three neutralizing scFv antibodies had significantly higher affinity than other antibodies without neutralizing activity. These data indicated that VP2 protein could be used as the bait to obtain neutralizing scFvs. And BSD technology in combination with FCM could be used for isolating both affinity and neutralizing antibodies. The future objective will be to structure the full-length antibodies which are based on the three selected scFvs. After optimizing, the problem of half-life reduction due to lacking of an Fc domain in the scFv structure will be avoided. Moreover, the affinaty of a full-length antibody was 10- to 20-fold than the corresponding scFv, due to increased avidity (Yuanet al., 2000; Zhanget al., 2017). After making further efforts on enhancing the effects of these antibodies, they could be selected as new potential alternatives for the treatment and precaution of GPV infection.

Conclusions

In summary, natural antibody libraries with high affinity to GPV from immunized goose were generated. By screening of the bacterial display libraries and virus neutralization assay, scFv antibodies with good affinity and neutralization activity were obtained, which might be the efficient productions for detecting viruses and treating gosling plague disease.

Acknowledgments

The authors wish to thank Harbin Qinglong Goose Industry Development Co., Ltd., for animals.

Fig. 6 Bacteria surface display technology(a), VH-GS-VL fragments subcloned into pBSD vector. (b), Libraries of scFv antibodies expressed in periplasm of E. coli are tethered to inner membrane by CDQSSS. After outer membrane permeabilization, scFv antibodies on inner membrane specifically bind with fluorescently labeled VP2. (c), Enrichment of spheroplasts expressing VP2 specific scFvs by gating region defined by distinct scatter of spheroplasts (FSC and SSC) and high FITC-A signal. FSC forward scatter, SSC side scatter, CDQSSS, NlpA amino acids 1-6.

Conflict of interest statement

Zhou Jin-xin and Zhang Xiao-yu contributed equally to this work. The authors declare that they have no competing interests with other people or organizations.