Generation and Evaluation of Chicken IgY-scFv for the Purposes of CPV Diagnosis and Therapy

Background: Canine parvovirus (CPV) can cause acute and highly contagious enteritis in dog, antibodies have been used for diagnosis and therapy of CPV disease, generation of functional antibody fragments by using novel antibody engineering platforms is promising in veterinary practice. Results: The IgY single chain fragment variables (scFv) were generated by T7 phage display system and expressed in E. coli system after immunizing hens with virus-like particles (VLP) of CPV-VP2. The titer of the primary scFv library reached to 1.5×10 6 pfu/mL, and 95% of the phages contained the target fragments. The CPV-VLP and CPV-VP2 protein showed similar reaction values to the puried scFv in the ELISA test, and the results of ELISA analysis using IgY-scFv toward CPV clinical samples were consistent with commercial immunochromatographic assay (ICA) and PCR detection, the scFv did not show cross reactivity with canine distemper virus (CDV) and canine coronavirus (CCV). IgY-scFv was successfully expressed in CRFK cells, and in the virus suppression assay, 55% of CPV infections were eliminated within 24 hours. Docking results demonstrated that the number of amino acids of the binding sides between scFv and VP2 were AA37 and AA40, respectively. Conclusions: This study revealed the feasibility of a novel functional antibody fragment development strategy by generating diversied avian IgY-scFv libraries towards the pathogenic target of interest for both detection and therapeutic purposes in veterinary medicine.


Background
Canine parvovirus (CPV) was rst identi ed in 1978 (1); it has a single-stranded DNA genome of negative polarity, about 5200 bp in length. The CPV capsid is a 25 nm diameter icosahedron containing three structural viral proteins (VP1, VP2 and VP3) and two non-structural proteins (NS1 and NS2), among them, VP2 accounts for 90% of the viral capsid and represents the major determinant of host range and virus-host interactions (2). Although the conventional attenuated and inactivated CPV vaccines have been successful in reducing the disease outbreaks, the genus of parvovirus still causes severe epizootics worldwide and leads to severe economic losses in dogs (3).
Antibody based approach is promising in CPV disease diagnosis, therapy and prevention, and the relevant hyper immune serum (IgG) and full-length monoclonal antibody (mAbs) have been routinely applied in the veterinary practices (4). At present, most commercially available mAbs are produced in mammalian system using hybridoma technology. Functional antibody fragments (i.e.: scFv, Fab) generated by phage-display technology have been not yet routinely evaluated and applied in the veterinary medicine despite it has been well con rmed that such engineered antibodies offer a series of advantages over polyclonal and full-length monoclonal antibodies. Antibody phage display technology is to display polyclonal antibodies on the surface of the phage shell protein and to screen the speci c monoclonal antibodies by bio-panning procedure (5). Phages are more stable and can be stored for years at 4℃; it can be reproduced rapidly, successively and inexpensively with a con rmed sequence in the prokaryote or eukaryote systems without hybridoma and immunization procedure. Furthermore, higher a nity mutants of scFv can be generated through site directed mutagenesis which is much easier and simpler to be performed (6).
Chicken (Gallus domesticus) IgY antibody generated by IgY technology is another conventional approach in generating large amount of high speci c antibody and have been used for broad biomedical purposes owning to a series of advantages such as higher productivity, better animal welfare, higher immunogenicity to mammal conserved proteins and lower cross-reactivity, as compared to the generation and application of mammalian serum IgG (7). Our previous work con rmed that the generated polyclonal IgY targeted to CPV-VP2 could applied into the immunotherapy and immunoprophylaxis for the CPV infection (8). In recent years, as an interesting tendency in IgY technology, generation of IgY-scFv in order to better combine the biological superiorities of both IgY and recombinant antibody fragment, is gaining increasing attention and application (9,10). Our previous studies demonstrated that IgY-scFv can be generated and applied in different immunoassays for the detection of small molecules gentamicin (11) and large molecules pIFN-γ (12) in the veterinary practice.
As an attempt to understand the feasibility of chicken IgY-scFv in diagnosis and therapy of veterinary diseases, this study aimed to construct and characterize chicken sourced scFv against CPV-VP2 virus like particles (VLP) using phage display technology, and to evaluate the speci city, sensitivity and virus inhibition ability of the obtained scFv.

Immunization of chicken
Twelve-week old white Leghorn hens were immunized intramuscularly with CPV-VLP (provided by Dr. Shiqi Sun (13)) mixed with Freund's adjuvant (Sigma-Aldrich, St. Louis, MO, USA) at four different sites of breast muscles. CPV-VLP protein (125 µL, 2 mg/mL) in equal volume of phosphate buffered solution

Construction Of Anti-cpv-vlp Scfv
The hen's spleen was collected to extract the total RNA using Total RNA Kit (Tiangen Biotech, Beijing, China), and the rst-strand cDNA was synthesized using (V L ) genes were ampli ed by PCR with primers ( Table 1). The V H and V L were assembled using primers HF-EcoR I & LR-Hind III by Overlap PCR. The products of overlap PCR (scFv) were puri ed through Gel Extraction Kit (Omega, Norcross, GA, USA).

Bio Panning
The ampli ed library was subjected to four rounds of bio panning on microplates for the enrichment of the speci c scFv phages according to the T7 select system manual. The microplates were coated with CPV-VLP overnight (15 µg, 10 µg, 8 µg and 5 µg per well in each round) at 4℃, and washed with Tween-PBS (PBST, tween 0.5%; 150 µL/well) and PBS, respectively. After blocked with skimmed milk (5%) in PBST, the ampli ed phages from the starting library of each round of panning were added to the microplates and incubated at 37℃ for 1 h, and the bound phage was eluted with SDS (1%) in each round. In the rst three rounds, eluted phages were ampli ed by infecting BL5403 bacterial culture (Novagen, Darmstadt, Germany). The enrichment of speci city was determined from the input/output ration of the phage. A total of random clones of 100 on the plates were selected from the fourth round to identify the positive rate using a pair of universal primers ( Table 1). The ampli ed library was added with 1% chloroform and stored at 4℃.

Expression And Puri cation Of Scfv
After sequencing, the phage-scFv with the highest binding force was selected to ligate with pET-30a (+) vector using T4 DNA Ligase (TaKaRa Biotech, Dalian, China). The recombinant plasmids were transformed into BL21 (DE3) competent cells, and the scFv expression was induced with isopropyl-β-Dthiogalactopyranoside (IPTG). The bacterial cells were collected by centrifugation and sonication, and the supernatants and pellets were used for SDS-PAGE analysis, respectively. Protein was puri ed by HisTrap HP histidine-tagged protein cultured columns (GE, Pittsburgh, PA, USA).
Membranes were incubated with 12.5 mL TBS-T containing 1% NDM and 5 µg prepared scFv proteins at RT under agitation for 1 h. Membranes were washed three times for 10 min in TBS-T, and incubated with 12.5 mL TBS-T containing 1% NDM and mouse anti-His monoclonal antibody conjugated with HRP

Detection Of Clinical Samples
To determine the accuracy of scFv, a total of 28 clinical dog stool samples were collected using sterile swabs in an animal hospital (Xinger, Xi'an, China).
Among them, 24 samples were con rmed CPV positive and 4 were negative by the hospital using commercial colloidal gold test strip (ICA). The samples were homogenized (10%, w/v) in PBS (1 mL, pH 7.2) and centrifuged, the supernatants were used for indirect ELISA and PCR. In order to verify whether scFv has speci city in clinical practice detection, Canine distemper virus (CDV) and coronavirus (CCV) were used to detect the cross-reaction by ELISA.

Construction of scFv library
The lengths of VH-linker, VL-linker and scFv were approximately 370 bp, 420 bp and 800 bp, respectively ( Fig. 1A and B). The titer of the primary anti-CPV scFv library (Fig. 1C) and the ampli ed library were 1.5 × 10 6 pfu/mL, and 8.2 × 10 11 pfu/mL, respectively. There were 95% of the phages containing the target fragments in the primary scFv library.

Screening Of Scfv Phage Library
After 4 rounds of "bind-elute-amplify" bio-panning procedure, the data of input and output phages in each round indicated that the phage library had been enriched 100 times ( Fig. 2 and Table 2). A total of 15 scFv genes showed relatively high binding capacity to CPV-VLP in the reactivity detection of 100 phages, and the sequencing con rmed that all the 15 scFv genes have complementary determining region 3 (CDR3) that was the main mutation region in both VH and VL. There were few clones having limited mutations in framework region (FR, Fig. 3). The phage-scFv (No. 96) showed the highest binding force was chosen for further experiments.

Scfv Expression And Puri cation
The solubility analysis showed that scFv mainly existed in the form of inclusion body (38 kDa; Fig. 4A, lane 2). The denaturation and puri cation of scFv inclusion body protein showed a single protein band with high purity (Fig. 4B). The VP2 protein could bind to scFv with a single binding strip (85 kDa; Fig. 4C).

Scfv Sensitivity Analysis
The immunoreactivity of scFv against soluble VP2 was examined by ELISA. scFv bound in a dose-dependent manner to the soluble VP2 (Fig. 5). The minimum antibody concentration for the detected antigen was 2 ng/µL. The coincidence of ELISA (Fig. 6A) and PCR (Fig. 6C) with ICA (data not show) was 100% and 85.7%, respectively. The scFv showed no cross reactivity with CDV and CCV (Fig. 6B).

Notes
Clinical samples of CPV were analyzed by ELISA (A), PCR (C) and ICA (data not shown); the cross reactivities of anti-CPV-IgY-scFv with CDV and CCV were analyzed by ELISA (B). CPV was CPV-HY strain; VP2 was expressed in the prokaryotic system.

Neutralization Effect Of Scfv To Cpv In Crfk Cells
The sequencing results showed that the scFv and pcMV-3 vectors were successfully connected (Fig. 7A), IFA con rmed that scFv was signi cantly expressed in CRFK cells (Fig. 7B). Immunoblotting further con rmed that the scFv protein was correctly folded and modi ed in the cell (Fig. 7C). The CRFK cells expressing scFv showed a small amount of cytopathic effect (CPE) after CPV infection; the cells without scFv expression were broke away signi cantly from the bottom wall, became round, and some cells even broke up (Fig. 7D). Virus TCID 50 at different time points was determined; the growth rate of virus in cells expressing scFv was signi cantly lower than that in cells not expressing scFv (Fig. 7E). The inhibition rates of scFv on virus growth at 24 h, 48 h, and 72 h were 55%, 38% and 30%, respectively (Fig. 7F). The main structural domains of scFv were VLCDR1, VLCDR2, VLDR3, VHCDR1, VHDCR2, and VHCDR3, with 13 antigen-binding sites on scFv (Fig. 8A). The 3 dimensional scFv model was subsequently constructed (Fig. 8B). A total of 5 highly similar VP2 homologous sequences were simulated by software (Fig. 8C), the VP2 molecular stereo model was established by combining the characteristics of each sequence (Fig. 8D). After analysis on ScFv and VP2 binding mode ( Fig. 8E), the interacting amino acids at the binding sides of scFv (AA37) and VP2 (AA40) were con rmed ( Fig. 8F and 8G).
Notes: All docking analysis were performed by Discovery Studio software. ScFv modeling: Antibody Modeling Cascade program; VP2 modeling: Build Homology Models program; Docking: ZDock program.

Discussion
The mAbs have been widely applied in the biomedical areas owning to their high speci city and homogeneity. However, mAbs produced by mammals may have the side effects of immunogenicity, thrombocytopenia and hypersensitivity reactions, etc, which has greatly limited the application of mAbs in target detection and treatment (14,15). With the development of antibody library technology and humanized antibody modi cation technology, increasing humanized recombinant antibodies have been gradually developed and entered the clinic trials (16). Diversi ed antibody generation strategies could be a future tendency in antibody engineering in order to better combine the characteristics and advantages of antibodies from different sources. As a notable example, Brolucizumab (Beovu) is the rst FDA approved rabbit-derived scFv used as vascular endothelial growth factor (VEGF) inhibitor for the treatment of exudative (wet) age-related macular degeneration (AMD), diabetic macular oedema and macular oedema secondary to retinal vein occlusion, which could better overcome the possible side effects (discomfort and increased tears in the affected eyes, itchy or watery eyes, dry eyes, swelling of the eyelids, etc.) of murine-derived IgG-Fab fragment (Lucentis) (17). In avian IgY, similar attempts have also been made. For instance, the snake venom contains neurotoxic proteins, the urgent administration of hyperimmune serum from horse used to be the most e cient treatment, which can recognize many different antigenic determinants. However, generation of equine anti-venom is costly and associated with several potential side effects. Chicken IgY-scFv has been generated against glutaraldehyde-attenuated Daboia russelii formosensis (DRF) venom proteins for passive immunization, which can identify and neutralize the toxic activity of the venom components, with only small quantity of antigens required to induce a signi cant antibody response in hens, and can be also used as a rapid diagnostic tool for wound secretions to determine snake types (18).
As summarized by previous authors, owning to the unique structure, phylogenetic distance and the mechanism of molecular diversi cation, IgY antibodies provide a series of important advantages over mammal IgG, including the stronger immune responses of chicken system to the proteins conserved among mammals, decreased/avoid cross-reactivities (ie.: rheumatoid factor, human anti-mouse IgG antibody, complement system, Fc receptors) in the mammal systems (19,20), and more convenient to design a primer against IgY-scFv (Table 1), as IgY only has one isotype and it is lack of hinge region (9).
Furthermore, it is noteworthy to address, recent studies con rm that from the glycobiology point of view, recombinant IgY antibody could be a potentially promising immune-therapeutic candidate after proper antibody engineering and expression as IgY is more heavily glycosylated (21), and has higher sialic acid content (22) as compared to mammal IgG. Recombinant functional antibody fragments remove or reduce irrelevant structures, while retain the speci city and main biological activities of natural antibodies, which offers a wider application prospect than natural antibodies (23). Recombinant IgY-scFv could combine the advantages of both IgY molecular and functional antibody fragment (9).
Designing on chimeric antibody could be the next step for IgY-scFv study in order to provide better compatibility of the antibody in the host system, and to recoup the possibly decreased speci city and a nity of antibody fragments as compared to full length antibody. Recent study con rmed that mammalian IgG and avian IgY shared compatible V-C region interfaces, which may be conducive for the design and utilization of mammalian-avian chimeric Abs (24).
In our study, the high consistency of ELISA analysis to PCR and ICA on clinical samples con rmed the speci city of the obtained IgY-scFv (Fig. 6), which offers the potential using IgY-scFv for rapid detection of CPV. scFv neutralized the virus (Fig. 7D), inhibited CPV replication with signi cantly reduced growth rates of the CPV observed in the CRFK cells (Fig. 7E, F), which provides the value of further therapeutic investigation of obtained IgY-scFv.
As alternative to viral components, CPV-VP2-VLP was used as immunogen in this study. With increasing applications in vaccine design and immunization, VLP has been recognized as safe and effective particle to stimulate adequate immune responses for both viral and non-viral diseases by inducing lymphocyte proliferation and speci c antibody with high titer (25). The docking of scFv-VP2 shows that there were 13 antigen-binding sites on scFv, with high binding force to VP2. According to Residues Contact Frequency (RCF) algorithm analysis, 37 binding amino acids on scFv and 40 binding amino acids on VP2 were involved in the binding (Fig. 8), these results provide us con dence that a well-designed CPV-VLP can be used as a potent immunogen to induce quali ed speci c antibodies.
In Conclusion, we demonstrated that speci c IgY-scFv can be generated with high speci city and signi cant inhibition to CPV growth. As a preliminary evaluation, our work revealed the potential of IgY-scFv as a novel approach in veterinary diagnosis and therapy.