A Strategy for Generating a Broad-Spectrum Monoclonal Antibody and Soluble Single-Chain Variable Fragments against Plant Potyviruses Han-Lin Liu,a Wei-Fang Lin,a Wen-Chi Hu,a Yung-An Lee,b Ya-Chun Changa Department of Plant Pathology and Microbiology, National Taiwan University, Taipei, Taiwana; Department of Life Science, Fu Jen Catholic University, New Taipei City, Taiwanb

Potyviruses are major pathogens that often cause mixed infection in calla lilies. To reduce the time and cost of virus indexing, a detection method for the simultaneous targeting of multiple potyviruses was developed by generating a broad-spectrum monoclonal antibody (MAb) for detecting the greatest possible number of potyviruses. The conserved 121-amino-acid core regions of the capsid proteins of Dasheen mosaic potyvirus (DsMV), Konjak mosaic potyvirus (KoMV), and Zantedeschia mild mosaic potyvirus (ZaMMV) were sequentially concatenated and expressed as a recombinant protein for immunization. After hybridoma cell fusion and selection, one stable cell line that secreted a group-specific antibody, named C4 MAb, was selected. In the reaction spectrum test, the C4 MAb detected at least 14 potyviruses by indirect enzyme-linked immunosorbent assay (I-ELISA) and Western blot analysis. Furthermore, the variable regions of the heavy (VH) and light (VL) chains of the C4 MAb were separately cloned and constructed as single-chain variable fragments (scFvs) for expression in Escherichia coli. Moreover, the pectate lyase E (PelE) signal peptide of Erwinia chrysanthemi S3-1 was added to promote the secretion of C4 scFvs into the medium. According to Western blot analysis and I-ELISA, the soluble C4 scFv (VL-VH) fragment showed a binding specificity similar to that of the C4 MAb. Our results demonstrate that a recombinant protein derived from fusion of the conserved regions of viral proteins has the potential to produce a broad-spectrum MAb against a large group of viruses and that the PelE signal peptide can improve the secretion of scFvs in E. coli.

T

he genus Potyvirus, one of the largest groups of plant viruses, includes 158 species, many of which are economically important, such as Potato virus Y (PVY) and Plum pox virus (1, 2). Most potyviruses infect various host plants and are spread worldwide by aphid vectors and mechanical inoculation (3). Potyviruses have a single-stranded, positive-sense RNA genome of approximately 10 kb in length that encodes a polyprotein translated from a large open reading frame (ORF) (3) and a trans-frame movement-related protein (P3N-PIPO) encoded from the 5=-terminal portion of P3 and the PIPO cistrons (4, 5). During proteolytic processing, the large polyprotein can be separated into 10 mature proteins that are involved in virus replication, movement, and virion assembly (3). Potyviral capsid proteins (CPs) not only form the viral particle but also participate in cell-to-cell and systemic movement (6, 7). The variable N- and C-terminal regions of the CPs exposed on the virion surface are required for long-distance transport, are involved in aphid transmission (DAG motif), and may also assist in cell-to-cell spreading (7–10), whereas the conserved core regions of CPs are necessary for virion assembly and the cell-to-cell movement of potyviruses (6, 7). In Taiwan, the viral disease caused by potyviruses is one of the limiting factors for production of calla lilies. At least five potyviruses, including Calla lily latent virus (CLLV), which is a tentative species, Dasheen mosaic virus (DsMV), Zantedeschia mosaic virus (ZaMV), which is currently classified as an isolate of Konjak mosaic virus (KoMV), Turnip mosaic virus (TuMV), and Zantedeschia mild mosaic virus (ZaMMV), have been reported to infect the calla lily (11–15). To detect these potyviruses, different methods have been developed for individual viruses, such as reverse transcription-PCR (RT-PCR), immunocapture RT-PCR, dot blot hybridization, and enzyme-linked immunosorbent assay (ELISA)

October 2015 Volume 81 Number 19

(16–18). One multiplex RT-PCR protocol has been successfully developed to detect four calla lily potyviruses, DsMV, TuMV, KoMV, and ZaMMV, in one reaction (19). Based on a field survey of calla lilies performed using multiplex RT-PCR, 72% of the samples were found to be infected by potyviruses, and 63.9% of the infected calla lilies had a mixed infection (19). Accordingly, indexing of potyviruses and tissue culture is essential for the establishment of a calla lily certification program. For this goal, ELISA is a better choice than RT-PCR for the indexing of a large number of plant samples because it is easy to perform and cost-effective. However, most antibodies recognize only one specific virus species or even one virus strain. For host plants, such as the calla lily, which can be infected by at least five potyviruses, preparation of antibodies specific to each species is desirable for detecting the target viruses. To overcome the constraint of preparing a specific antibody for each virus and performing ELISA for individual vi-

Received 12 April 2015 Accepted 19 July 2015 Accepted manuscript posted online 24 July 2015 Citation Liu H-L, Lin W-F, Hu W-C, Lee Y-A, Chang Y-C. 2015. A strategy for generating a broad-spectrum monoclonal antibody and soluble single-chain variable fragments against plant potyviruses. Appl Environ Microbiol 81:6839 – 6849. doi:10.1128/AEM.01198-15. Editor: K. E. Wommack Address correspondence to Ya-Chun Chang, [email protected]. Supplemental material for this article may be found at http://dx.doi.org/10.1128 /AEM.01198-15. Copyright © 2015, American Society for Microbiology. All Rights Reserved. doi:10.1128/AEM.01198-15

Applied and Environmental Microbiology

aem.asm.org

6839

Liu et al.

ruses, the utilization of commercially available monoclonal antibodies (MAbs) or generation of our own MAb with broad-spectrum reactivity to all target viruses would be a favorable approach. Although MAbs can be produced with consistent quality and in large amounts, their preparation is disadvantageous because it is time-consuming, expensive, and laborious. Emerging antibody engineering techniques have been used to generate many different antibody fragments, overcoming some of the weaknesses of MAb production (20, 21). The single-chain variable fragment (scFv) is the most popular type of recombinant antibody, and it is also the smallest unit that maintains antigen-binding ability (22). Compared with complete antibodies (MAb and polyclonal antibody [PAb]), scFvs contain only two variable domains (derived from the heavy and light chains), linked by a short flexible peptide to facilitate their correct association after protein expression (23). Hence, the general size of scFvs (24 to 26 kDa) is smaller than that of complete antibodies (150 kDa) (23). scFvs with high specificity and affinity not only can be cloned from the total RNA of hybridoma cells (24) but can also be screened using different platforms, such as phage libraries and ribosome display (25–28). Furthermore, antiviral scFvs fused with alkaline phosphatase (AP) or biotin can be directly applied in diagnostic assays (26, 27, 29, 30). The shorter production time and lower cost of scFvs compared with MAbs are also advantageous because they enable expression of scFvs in heterologous protein expression systems, such as Escherichia coli systems. However, scFvs can easily form insoluble aggregates when they are expressed in the E. coli cytoplasm (31, 32). Therefore, the development of production methods for soluble scFv expression in bacteria would greatly enhance their applicability. In this study, to index five calla lily potyviruses, we generated a broad-spectrum MAb against potyvirus CPs using an immunogen comprised of concatenated highly conserved core regions of three potyvirus CPs. After immunizing BALB/c mice, we screened a C4 MAb that reacted to 14 potyviruses. In addition, the variable regions of the heavy and light chains (VH and VL) of the C4 MAb were cloned and constructed as C4 scFvs for expression in E. coli. Furthermore, the secretory signal peptide of pectate lyase E (PelE, derived from Erwinia chrysanthemi S3-1) was N-terminally fused to the C4 scFvs to enhance secretion from bacterial cells. The soluble C4 scFvs showed binding abilities similar to that of C4 MAb. This report provides a strategy for generating a broad-spectrum MAb and also for employing the PelE secretory signal peptide for secretion of scFvs in a bacterial expression system. MATERIALS AND METHODS Virus source and propagation. In this study, we used DsMV-, TuMV-, KoMV-, and ZaMMV-infected plants for MAb screening. DsMV, KoMV, and ZaMMV were maintained on calla lily plants, whereas TuMV was kept on a systemic host, Nicotiana benthamiana. Nine potyviruses were used for the reaction spectrum test of the MAb. Potato virus Y (PVY) was kindly supplied by Tso-Chi Yang (Taiwan Seed Improvement and Propagation Station, Council of Agriculture). Peru tomato mosaic virus (PTV), Potato virus V (PVV), and Tobacco etch virus (TEV) were purchased from DSMZ, Germany, and maintained on Nicotiana tabacum cv. White Burley, Nicotiana glutinosa, and N. tabacum cv. Samsun, respectively. The remaining viruses were obtained from collections at our laboratory. Basella rugose mosaic virus (BaRMV) (33), Bean yellow mosaic virus (BYMV), and Potato virus A (PVA) were kept on N. benthamiana. PVY and Pepper veinal mottle virus (PVMV) were maintained on N. tabacum cv. Samsun.

6840

aem.asm.org

TABLE 1 Primers used for generating protein and C4 scFv expression constructs Primer name

Sequence (5=–3=)a

CLLV-CPSalF CLLV-CPBamR

GCT AGT CGA CGA TGA GCA GAT GGG TGT GGT GCT AGG ATC CGA GAG CTG CAG CCT TCA TTT G GCT AGA GCT CGA AGA TGA GCA GAT GCA CAT AGT GCT AGT CGA CGA GTG CAG CGG CCT TCA TCT G TCT AGA ATT CAA TGA TGC AGA GAT GGA CGT CA GCT AGA GCT CCA GGG CGG CTG CTT TCA TCT G GCT AGG ATC CAC GGA GGA CAA AAT GCA AAT CAT TC GCT ACT CGA GCA GTG CTG CTG CTT TCA TCT G GCT ACC ATG GAT GAT AGC CAG ATG CAG AGA ATC TCT AGA ATT CTA GCG CTG CTG CCT TCA TCT G CAC CCG ACC CAC CAC CGC CCG AGc cac cgc cac ctg cag aga cag tga cca gag tcc c CTC GGG CGG TGG TGG GTC GGG TGG cgg cgg atc gga tgt ttt gat gac cca aac tcc ac CAC CCG ACC CAC CAC CGC CCG AGc cac cgc cac ctt tta ttt cca act ttg tcc ccg a CTC GGG CGG TGG TGG GTC GGG TGg cgg cgg atc gga ggt tca gct tca gca gtc tgg a GCA TCT CGA GTT TTA TTT CCA ACT TTG TCC CCG A CTC GAG TGC AGA GAC AGT GAC CAG AGT CCC ATG AAC AAC TCA CGT ATG TCT TCC G ATC GCT TTT CAC CGC GTA GAT TTT GG GCA TGT CGA CGG ATC CGA TAT CGC TTT TCA CCG CGT AGA TTT TG GCA TGT CGA CGG ATC CGA TAT CCA GAC TGG CGG CGG AAG CG ATG TAT ATC TCC TTC TTA AAG TTA AAC AAA ATT ATT TC ATG AGT AAA GGA GAA GAA CTT TTC ATG CCT CGA GTT TGT ATA GTT CAT CCA TG

DsMV-CPSacF DsMV-CPSalR KoMV-CPEcoF KoMV-CPSacR TuMV-CPBamF TuMV-CPXhoR ZaMMV-CPNcoF ZaMMV-CPEcoR VH linker (VH-VL) VL linker (VH-VL) VL linker (VL-VH) VH linker (VL-VH) MuJK4-Xho⌱ MuJH3-XhoI PelEsp-F PelEsp-R PelE-R-long PelE-R-short pET29a-R GFP-F1 GFP-Xho-R2

a Underlining indicates created restriction sites used for cloning. Lowercase letters indicate linker sequences added to primers.

Zucchini yellow mosaic virus (ZYMV) was propagated on pumpkin (Cucurbita pepo). Construction of expression vectors by fusing the conserved core regions of multiple CP genes. To generate a broad-spectrum MAb, a bacterially expressed recombinant protein containing several conserved core regions was used as an immunogen for MAb production. Five potyvirus CP clones, DsMV (16), KoMV (18), ZaMMV (17), TuMV (34), and CLLV (a gift from Chin-An Chang from the Taiwan Agricultural Research Institute, Council of Agriculture), were used to construct expression vectors. The specific forward and reverse primers for each virus were designed to amplify the conserved core region of the CP gene from viral clones (Table 1). After amplification, the PCR products of each viral fragment were purified and digested with the corresponding enzymes. To construct pET-P3CP, the SacI-treated PCR products of KoMV and DsMV were ligated, and the ligation product was amplified using the KoMVCPEcoF and DsMV-CPSalR primers. Then, this PCR product was digested with EcoRI and ligated with a ZaMMV fragment that was pretreated with EcoRI. Next, this ligation product was amplified with the

Applied and Environmental Microbiology

October 2015 Volume 81 Number 19

Potyvirus Broad-Spectrum MAb and Soluble scFvs

ZaMMV-CPNcoF and DsMV-CPSalR primers. Finally, the PCR product digested with NcoI and SalI was used for ligation with a pretreated pET29a(⫹) expression vector (Novagen). To construct pET-P5CP, a BamHIand XhoI-treated pET-29a(⫹) expression vector was ligated with a BamHI- and XhoI-treated TuMV fragment. The correct clone was first digested with NcoI and BamHI and then ligated with an NcoI- and SalItreated pET-P3CP fragment and a SalI- and BamHI-treated CLLV fragment simultaneously to create five concatenated core CPs. The correct clones containing recombinant core CPs of three and five potyviruses were used for protein expression in E. coli. Expression and purification of recombinant core CPs. In the time course experiment of recombinant core CP expression, a single bacterial colony was cultured in 3 ml of Luria-Bertani (LB) broth with 50 ppm kanamycin at 37°C. When the optical density at 600 nm (OD600) of the bacterial culture reached 0.4 to 0.6, 1 mM isopropyl-␤-D-thiogalactopyranoside (IPTG) was added to the culture medium. Bacterial lysates were prepared at 0, 1, 2, and 3 h postinduction, and they were analyzed by 12.5% SDS-polyacrylamide gel electrophoresis and stained with Coomassie brilliant blue R250 (CBR). According to the results of the time course experiment of recombinant core CP expression, we chose the best induction time for large-scale recombinant protein expression. The recombinant core CPs were purified in accordance with the protocol of the QIAexpressionist (Qiagen). Generation and isotype determination of the MAb. Equal amounts of purified recombinant core CPs expressed from pET-P3CP and pETP5CP were separately used for immunization of mice. Each of four BALB/c female mice was injected with 50 ␮g of purified protein four times at 2-week intervals. Indirect ELISA (I-ELISA) and Western blot assays were performed to analyze mouse serum samples obtained 1 week after each injection. After antiserum was generated in response to recombinant protein, the mice were administered a final booster of 100 ␮g purified protein. The mice were sacrificed 3 days later, and their spleens were harvested for cell fusion. Single-splenocyte suspensions were prepared from the immunized spleens and fused with SP2 myeloma cells. Seven to 10 days after fusion, supernatants from the culture wells were screened for anti-potyvirus CP antibody secretion by I-ELISA using expressed recombinant proteins and crude antigens extracted from various potyvirus-infected plants. Antibody-secreting hybridoma cells were selected, transferred to 24-well culture plates, and retested by I-ELISA. Stable potyvirus-specific MAb-secreting lines were cloned using the limiting dilution method. The selected hybridoma cell lines were intraperitoneally injected into pristane-primed BALB/c female mice. Ascites fluid was collected at 10 to 15 days after injection. MAb isotype determination was performed according to the protocol of an ImmunoPure monoclonal antibody isotyping kit II (Pierce). Mouse IgG purification. Mouse IgG purification was performed using an ImmunoPure (A Plus) IgG purification kit (Pierce). For the serum samples and ascites fluid, it was necessary to dilute the samples at least 1:1 with ImmunoPure (A) IgG binding buffer or another suitable buffer prior to application to high-capacity AffinityPak Protein A Plus columns. Eluted IgG was desalted using D-Salt Excellulose desalting columns (Pierce) or dialysis tubing and was then concentrated with a Vivaspin 20 instrument (GE Healthcare). Construction of expression vectors for C4 scFv from C4 MAb-secreting hybridoma cells. To clone VH and VL for construction of C4 scFv, total RNA was extracted from C12-C4 hybridoma cells producing the C4 MAb using the TRI-Reagent solution (Ambion), and mRNA was isolated using a Dynabeads mRNA purification kit (Invitrogen). Primers previously designed for the antibody constant region (35) were used to amplify the desired fragments of the heavy and light chains. Then, first-strand cDNA was synthesized using a Superscript III kit (Invitrogen) with primers MuIgG1/2 For and MuCK For, respectively. The VH and VL fragments were amplified with different degenerate primer pairs, and the PCR products were cloned into pGEM-T Easy vectors (Promega) for sequencing.

October 2015 Volume 81 Number 19

The correct constructs were chosen for assembly of VH-VL and VL-VH with 15-amino-acid linker sequences (Gly4Ser)3 by overlap extension PCR using specifically designed linker primers (Table 1). The PCR products of the VH-linker and VL-linker were gel purified and were then used for the next round of PCR. The final assemblies of the VH-linker and VL-linker were also performed with the overlap PCR method to generate VH-VL and VL-VH for TA cloning and sequencing (GenBank accession numbers KT210383 and KT210384). Subsequently, a XhoI site was added to the VH-VL and VL-VH fragments by PCR, and then the XhoI-treated fragments were ligated to EcoRV- and XhoI-pretreated pET29a(⫹) vectors to create pET-S tag vectors for C4 scFv (VH-VL and VL-VH) expression in E. coli. Construction of pET-PelE-L and pET-PelE-S vectors with PelE signal peptides of different lengths. The full-length pelE gene of Erwinia chrysanthemi strain S3-1 (GenBank accession number KR091955) was amplified using primers PelEsp-F and PelEsp-R (Table 1) and sequenced. To determine whether protein secretion was influenced by signal peptide length, two lengths of PelE signal peptide sequences were used. The long PelE signal peptide (PelE-L) sequence was amplified with primers PelEsp-F and PelE-R-long (containing the EcoRV, BamHI, and SalI restriction sites) and then digested with SalI and cloned into a pET29(⫹) vector that had been previously treated with EcoRV and SalI to create a pET-PelE-L expression vector with the N-terminal 80 amino acids of PelE. The short PelE signal peptide (PelE-S) sequence was amplified with the primers PelEsp-F and PelE-R-short and was also cloned into pET29(⫹) vector via the same restriction sites to create a pET-PelE-S expression vector with the N-terminal 44 amino acids of PelE. To remove the S-tag/ thrombin recognition sequence in front of the PelE secretory signal peptide from the expression vectors, inverse PCR was performed using the PelE-F and pET29a-R primers and Pfu DNA polymerase (Fermentas). The PCR fragment was gel purified and transformed into E. coli DH5␣ after kinase reaction and ligation. After sequencing, the correct constructs were used for the following experiments. Cloning and expression of GFP. The green fluorescent protein (GFP) gene was amplified from p35S-GFP (Clontech) using primers GFP-F1 and GFP-Xho-R2, digested with XhoI, and then ligated with pET-PelE-L and pET-PelE-S as well as PelB-containing pET20b(⫹) (Novagen) expression vectors that had been pretreated with EcoRV and XhoI. These GFP expression constructs were transformed into the E. coli protein expression strain Rosetta (DE3). GFP expression was induced by treatment with 0.1 mM IPTG for 18 to 20 h at 25°C. The culture medium was collected by centrifugation and analyzed by SDS-PAGE and Western blotting. Cloning, expression, and purification of C4 scFv. All C4 scFv (VL-VH and VH-VL) constructs (PelE-L-scFv, PelE-S-scFv, and PelB-scFv) were created as the construction strategy of pET-S tag-scFv. The XhoI-treated fragment was cloned into pET-PelE-L, pET-PelE-S, and pET20b(⫹) expression vectors that had been pretreated with the EcoRV and XhoI enzymes. All C4 scFv constructs were transformed into the E. coli protein expression strain BL21(DE3) or Rosetta (DE3). To reduce inclusion body formation, scFv expression was performed by incubation with 0.1 mM IPTG for 18 to 20 h at 25°C. The culture medium was centrifuged, and the supernatant was collected and precipitated using trichloroacetic acid (TCA) at different time points for the time course experiment. The TCA precipitate was resuspended in 4⫻ sample buffer (8% SDS, 200 mM TrisHCl, pH 6.8, 0.08% bromophenol blue, 40% glycerol, 4% ␤-mercaptoethanol, and 50 mM EDTA) and analyzed by Western blotting. For largescale purification of scFv, the supernatant was precipitated with 80% ammonium sulfate, and the pellet was resuspended in phosphate-buffered saline (PBS). After dialysis, the C4 scFvs were purified using Ninitrilotriacetic acid (Ni-NTA) agarose (Qiagen) and stored at 4°C for further experiments. I-ELISA of C4 MAb and C4 scFvs (L-VL-VH and S-VL-VH). To analyze the antigen-binding activity of the C4 MAb, sap was extracted from healthy and virus-infected plants in 10 volumes of coating buffer (15 mM Na2CO3, 35 mM NaHCO3, pH 9.6). Each sample was loaded into tripli-

Applied and Environmental Microbiology

aem.asm.org

6841

Liu et al.

FIG 1 Conserved regions of potyvirus capsid protein (CP) genes and the recombinant core CP expressed in E. coli. (A) Amino acid sequence alignment of the CPs of five potyviruses, CLLV, DsMV, KoMV, TuMV, and ZaMMV, was viewed by GeneDoc software, with the conserved shading mode. The selected 121-amino-acid conserved core regions (arrows indicated) were used to prepare a recombinant core CP of potyviruses. The asterisks and numbers above the sequences mark every 10th residue. (B) Schematic diagram of pET-P3CP. The CP core regions of ZaMMV, KoMV, and DsMV were concatenated in pET-P3CP. T7, T7 promoter sequence; His, polyhistidine tag. (C) Recombinant core CP of potyviruses expressed from pET-P3CP in E. coli BL21(DE3). After induction with 1 mM IPTG, bacterial lysates were prepared at 0, 1, 2, and 3 h postinduction, and they were then separated by 12.5% SDS-polyacrylamide gel electrophoresis and stained with Coomassie brilliant blue R250 (CBR). The recombinant protein expressed by pET-P3CP is indicated by the arrow.

cate wells. I-ELISA was performed according to the ACP-ELISA protocol of Agdia, Inc. (see https://orders.agdia.com/Documents/m21.pdf). The C4 MAb and AP-conjugated goat anti-mouse antibody (QED Bioscience) were used as primary and secondary antibodies, respectively. Optical density was measured at 405 nm with a Spectra MAX 340 (Molecular Devices) after a 30- to 60-min incubation. To test the antigen-binding activities of the C4 scFvs, L-VL-VH and S-VL-VH (abbreviated terms for PelE-LVL-VH and PelE-S-VL-VH, respectively; see Results) were sequentially diluted with ECI buffer for use as primary antibodies. The anti-His antibody (Bioman Scientific) and AP-conjugated goat anti-rabbit antibody (QED Bioscience) were used as secondary and tertiary antibodies, respectively. Absorption was measured at 405 nm with an ELISA reader at 60 min. Western blot analysis of the C4 MAb and C4 scFv. To test the antigen-binding specificity of the C4 MAb, total proteins were extracted from healthy and virus-infected plants by grinding plant tissues in 10 volumes of PBS, separating the protein by 12.5% SDS-PAGE, and, finally, staining it with Coomassie brilliant blue R250 (CBR). The proteins were transferred to a polyvinylidene difluoride (PVDF) membrane (Osmonics), incubated with the C4 MAb, detected using an AP-conjugated goat anti-mouse antibody, and finally reacted with a colorimetric substrate, NBT/BCIP (nitroblue tetrazolium/5-bromo-4-chloro-3-indolylphosphate) (Roche Applied Science) at room temperature. To assess the secretion and antigen-binding specificities of the C4 scFvs, an anti-His anti-

6842

aem.asm.org

body and an AP-conjugated goat anti-rabbit antibody were used as secondary and tertiary antibodies, respectively, for Western blotting. Nucleotide sequence accession numbers. Sequence data have been deposited in GenBank under accession numbers KT210383, KT210384, and KR091955.

RESULTS

Generation of a broad-spectrum MAb against potyviruses with an antigen containing the conserved core regions of three CPs. To increase the possibility of generating a broad-spectrum MAb for detecting potyviruses in calla lily, we fused the conserved core regions of different CPs as an immunogen. Therefore, the CP amino acid sequences of five calla lily potyviruses, CLLV, DsMV, KoMV, TuMV, and ZaMMV, were aligned, and a conserved core region of 121 amino acids was selected for preparing the recombinant fusion protein to be expressed in E. coli (Fig. 1A). Subsequently, the 375-bp fragments of the conserved core regions of the DsMV, KoMV, ZaMMV, CLLV, and TuMV CP genes were separately amplified using specific primer pairs (Table 1) and were then sequentially ligated and cloned into a pET-29a(⫹) vector to construct the expression clones pET-P3CP and pET-P5CP

Applied and Environmental Microbiology

October 2015 Volume 81 Number 19

Potyvirus Broad-Spectrum MAb and Soluble scFvs

FIG 2 Antibody titer of the ascites fluid and specificity of the C4 monoclonal antibody (MAb). (A) The antibody titer of the ascites fluid was determined by I-ELISA. The bacterially expressed CLLV CP and crude sap from healthy, DsMV-, TuMV-, KoMV-, and ZaMMV-infected plants were used as test samples. The ascites fluid of the C4 MAb was diluted with ECI buffer from 1,000- to 100,000-fold for ELISA. Optical density (OD) was measured at 405 nm after a 30-min incubation. (B) Specificity of the C4 MAb was evaluated by Western blot analysis. The C4 MAb produced from the C12-C4 line was diluted 10,000-fold before use. Lane 1, healthy calla lily; lane 2, bacterially expressed CLLV CP; lanes 3 to 6, DsMV-, TuMV-, KoMV-, and ZaMMV-infected plants, respectively.

(Fig. 1B; see Fig. S1A in the supplemental material). Expressed proteins from pET-P3CP (47.2 kDa) and pET-P5CP (75 kDa) were detected by SDS-PAGE after 1 to 3 h of IPTG induction in E. coli BL21(DE3) (Fig. 1C; see Fig. S1B in the supplemental material). The expressed recombinant core CP was purified under denaturing conditions using Ni-NTA agarose and was then used as an immunogen. However, only the antiserum produced by the mice immunized with the protein expressed by pET-P3CP reacted specifically with the CLLV-expressed CP and the CPs of DsMV, KoMV, TuMV, and ZaMMV in the crude extracts of infected plants, and it did not react with any proteins from healthy plants, as demonstrated by I-ELISA and Western blotting. However, the protein expressed by pET-P5CP was not a good immunogen for inducing antibodies in mice and was therefore not used in the antibody production experiment (data not shown).To screen for hybridoma cell lines expressing antibodies with high immunogen specificity, two bacterially expressed proteins from pET-P3CP and pET-X (an irrelevant one) were used to select the correct cell lines by I-ELISA. Subsequently, cell lines with a high OD405 against the recombinant core CP were further screened by I-ELISA using crude antigens extracted from healthy and potyvirus-infected plants. One of the seven stable potyvirus-specific MAb-secreting hybridoma cell lines, C12-C4, was selected for production of ascites. The subclass of immunoglobulin produced by the C12-C4 hybridoma cell line was determined to be the IgG 2a isotype with the kappa light chain (data not shown) using a mouse MAb isotyping kit and was designated the C4 MAb. I-ELISA revealed that a 105-fold dilution of ascites of the C4 MAb still reacted with the bacterially expressed CP of CLLV (10 ␮g/ml) and the crude sap of potyvirus-infected plants (Fig. 2A). In addition, the specificity of the C4 MAb was tested by Western blot analysis. Specific CP bands were detected in crude extracts from the potyvirus-infected samples, and no band was observed in the healthy control (Fig. 2B). Two bands that were smaller than 34 kDa were observed for the ZaMMV-infected calla lily, and they may have been degradation products of the ZaMMV CP. In summary, these results demonstrated that the strategy of using an antigen containing fused CP core regions was successful in generating a MAb with high specificity for detecting five calla lily potyviruses. Reaction of C4 MAb against other potyviruses. To determine whether other potyviruses can be recognized by the C4 MAb, pu-

October 2015 Volume 81 Number 19

rified IgG of the C4 MAb (1 mg/ml) was diluted up to 104-fold for a reaction spectrum test with nine potyviruses, including BaRMV, BYMV, PTV, PVA, PVMV, PVV, PVY, TEV, and ZYMV. The C4 MAb showed strong reactions by I-ELISA with extracts from these potyvirus-infected plants and exhibited no cross-reactivity with healthy plants (Fig. 3A). Further analysis of the plant samples by Western blot analysis revealed results similar to those of I-ELISA, with specific detection of the potyvirus CPs in the total protein extracts from the infected plants but not the healthy one (Fig. 3B). Because the C4 MAb can react with nine potyviruses in addition to

FIG 3 Reaction of the C4 MAb with nine potyviruses, BaRMV, BYMV, PTV, PVA, PVMV, PVV, PVY, TEV, and ZYMV. (A) I-ELISA. Crude plant extracts were used as test samples. OD was measured at 405 nm after a 30-min incubation. H, healthy plant; buf, coating buffer. (B) Western blot analysis. Crude plant extracts were assayed, including (from left to right) those from healthy N. benthamiana plants (lane 1) and plants infected with BaRMV (lane 2), BYMV (lane 3), PTV (lane 4), PVA (lane 5), PVMV (lane 6), PVV (lane 7), PVY (lane 8), TEV (lane 9), and ZYMV (lane 10).The purified IgG of the C4 MAb (1 mg/ml), which was diluted 104 times, was used as the first antibody for ELISA and Western blot analysis.

Applied and Environmental Microbiology

aem.asm.org

6843

Liu et al.

FIG 4 Amino acid sequences of the heavy (VH) and light (VL) chain variable regions of the C4 MAb. The VH and VL fragments derived from C4 MAb mRNA were separately cloned and sequenced. VH is 114 amino acids long, and VL is 112 amino acids long. The complementarity determining regions (CDRs) within VH and VL are shown in bold and underlined.

five calla lily potyviruses, it is considered a broad-spectrum MAb against potyvirus CPs. Construction of C4 scFvs from the hybridoma cell line C12C4. To construct the C4 scFvs, mRNA from C12-C4 hybridoma cells was purified, and then the variable regions of the heavy and light chains (VH and VL) of the C4 MAb were amplified by RT-PCR using previously designed primers (35). After several different primer combinations were tested, the forward primer MuVH5/9 or MuVH8 paired with the reverse primer MuJH3 was found to amplify the expected 340-bp heavy-chain fragment (see Fig. S2 in the supplemental material). Because MuVH5/9 is a degenerate primer that covers the MuVH8 primer sequence, the MuVH8 and MuJH3 primers were used to amplify the VH gene, and then the product was cloned into a pGEM-T Easy vector for further sequencing. The full-length VH gene of the C4 MAb is 342 bp and can code for a 114-amino-acid protein (Fig. 4). The same strategy was applied to clone the VL of the C4 MAb. Unexpectedly, five out of the eight primer pairs were found to amplify the expected ca. 325-bp band by RT-PCR (data not shown). Hence, two strong bands amplified by different primer pairs (MuVK1/MuJK2 and MuVK7/MuJK1) were first chosen for cloning and sequencing. The sequencing data indicated that the obtained VL clones were 329 and 319 bp long and that their sequences were not in the correct reading frame for VL expression. Therefore, we used a mixture of eight forward primers and five reverse primers to amplify the dominant VL sequence (see Fig. S3). The expected VL gene fragment was purified and cloned into a pGEM-T Easy vector, and then eight colonies were randomly chosen for sequencing. The data revealed that all of the sequenced clones were amplified with the MuVK2 and MuJK4 primers and that they contained a 336-bp VL gene fragment that codes for a 112-amino-acid protein (Fig. 4). The complementarity determining regions (CDRs) within the VH and VL were predicted by use of the IMGT/VQUEST online system (http://www.imgt.org/IMGT_vquest/share /textes/) (36, 37). The VH gene of C4 was 90.28% identical to that of a mouse germ line (IGHV1-18*01 F, IMGT accession no. AC090843), and the VL gene of C4 was 98.64% identical to that of a mouse germ line (IGKV1-110*01 F, IMGT accession no. D00080). To link VH and VL, a (Gly4Ser)3 linker that is commonly employed for formation of scFvs was used. First, the VH and VL fragments were separately assembled with a linker by PCR, using the designed VH-linker and VL-linker primers and appropriate primers for generating 382-bp VH-linker and 376-bp VL-linker fragments (see Fig. S4). Then, the VH-linker and VL-linker frag-

6844

aem.asm.org

ments were connected by overlap extension PCR to form fulllength scFvs (see Fig. S5). Consequently, two types of C4 scFvs that were 723 bp long with different orientations of the heavy and light chain variable regions (VL-VH and VH-VL) were constructed (Fig. 5; see Fig. S5 in the supplemental material). Construction and expression of soluble C4 scFvs. scFvs frequently form inclusion bodies when they are overexpressed in bacterial cytoplasm (22, 31, 32). When pET-S tag vectors containing the C4 scFvs (Fig. 5) were tested under the protein expression conditions of different IPTG concentrations (0.1 to 1 mM) and temperatures (37°C and 25°C), the expressed S tag-VL-VH and S tag-VH-VL of the C4 scFvs were consistently aggregated in the cytoplasm of E. coli (Fig. 6A). Because the pectate lyase B (PelB) secretory signal peptide has been used for secretion of scFvs into the periplasm or medium to prevent aggregation in the cytoplasm (31, 38, 39), C4 scFv constructs carrying the PelB secretory signal peptide (PelB-VL-VH and PelB-VH-VL) were created using a commercial pET20b(⫹) vector (Novagen) to improve secretion. Unexpectedly, no bacterial colony growth was observed after the plasmid DNAs were transformed into E. coli strain BL21(DE3) or Rosetta (DE3) (data not shown). In contrast, when pET20b(⫹) was used to express GFP, the transformed Rosetta (DE3) colonies of PelB-GFP grew slowly, but the PelB secretory signal peptide was still able to secrete GFP into the medium (see Fig. S6 in the supplemental material). Fortunately, we discovered that PelE derived from the Erwinia

FIG 5 Schematic diagrams of the different C4 single-chain variable fragment (scFv) expression constructs. A pET-29a(⫹) vector was used to express two types of C4 scFvs (VH-VL and VL-VH), which were fused with three types of N-terminal tags and a C-terminal His tag. T7, T7 promoter sequence; S-tag, S tag/thrombin recognition sequence; PelE-L, N-terminal 80 amino acids of pectin lyase E (PelE) of Erwinia chrysanthemi strain S3-1; PelE-S, N-terminal 44 amino acids of PelE; VH-VL and VL-VH, C4 scFvs with different orientations of VH and VL joined by a (Gly4Ser)3 linker; His, polyhistidine tag.

Applied and Environmental Microbiology

October 2015 Volume 81 Number 19

Potyvirus Broad-Spectrum MAb and Soluble scFvs

FIG 6 Secretion of C4 scFvs is influenced by the presence of the PelE secretory signal peptide. (A) The S tag-VH-VL and S tag-VL-VH formed inclusion bodies in E. coli. The medium and soluble and insoluble fractions of bacterial lysates were analyzed by SDS-PAGE. M, medium; S, soluble fraction of bacterial lysate; P, insoluble fraction of bacterial lysate. (B) The C4 scFv with the PelE signal peptide was secreted into the medium. The presence of PelE-L-VL-VH (L-VL-VH) was verified by Western blot analysis using an anti-His antibody. (C) A comparison of the secretion efficiencies of four C4 scFv constructs with PelE signal peptides of different lengths and different scFv domain orientations. The results were analyzed by SDS-PAGE (lower panel) and Western blotting (upper panel). L-VH-VL, PelE-L-VH-VL; S-VH-VL, PelE-S-VH-VL; L-VL-VH, PelE-L-VL-VH; S-VL-VH, PelE-S-VL-VH; CBR, Coomassie brilliant blue R250.

chrysanthemi strain S3-1 was secreted into the culture medium at a 37°C induction temperature in the E. coli expression system (data not shown). The secretory pathway of PelE has been proven to be the SecB-dependent (SEC) pathway of the type II secretion system (40). Thus, we considered it worthwhile to test whether the PelE secretory signal peptide aids in the secretion of C4 scFvs and prevents the formation of inclusion bodies in bacteria. Hence, the SignalP 4.1 server (http://www.cbs.dtu.dk/services/SignalP/) (41) was used to predict the cleavage site of the PelE signal peptide. The result indicated that this signal peptide may be cleaved between Ala at position 41 and Ala at position 42, immediately after the recognition motif of AXA. Therefore, two different lengths of the PelE signal peptide were used to test the secretion efficiencies of VL-VH and VH-VL of the C4 scFvs (Fig. 5). PelE-L was 80 amino acids long and contained a 41-amino-acid PelE signal peptide and 39 amino acids of PelE. PelE-S was 44 amino acids long and comprised the signal peptide and 3 amino acids of PelE. All expression constructs of pET-PelE had a PelE signal peptide at the N terminus in addition to a His tag at the C terminus for protein purification and detection (Fig. 5). Based on SignalP 4.1 prediction, when the PelE signal peptide is completely cleaved by the signal peptidase, the molecular mass of PelE-L-VL-VH shifts from 34.9 to 30.85 kDa. When the protein expression of pET-PelE-L-VL-VH was induced with 0.1 mM IPTG at 25°C, the cleaved form of PelE-LVL-VH could be specifically detected by an anti-His antibody in the soluble fraction and medium (Fig. 6B). These results indicated that the PelE signal peptide was correctly cleaved by the signal peptidase. To compare the secretion efficiencies of the signal peptides of different lengths and the effect of scFv domain orientation, VL-VH and VH-VL, which contained PelE-L or PelE-S, were expressed under the same induction conditions as described above, and the culture medium and cell pellets were separately collected and analyzed. The expression levels of L-VH-VL (i.e., PelE-L-VH-VL) and L-VL-VH (i.e., PelE-L-VL-VH) were found to be similar according to SDS-PAGE, and those of S-VH-VL (i.e., PelE-S-VH-VL) and S-VL-VH (i.e., PelE-S-VL-VH) were also similar (Fig. 6C). However, Western blot analysis showed that L-VL-VH and S-VL-VH reacted more strongly to anti-His antibody than L-VH-VL and S-VH-VL in the medium and bacterial lysates (Fig. 6C). These data

October 2015 Volume 81 Number 19

revealed that the domain orientation of the C4 scFv did not significantly affect the secretion efficiency; however, this orientation may have influenced the interaction between the His tag and antiHis antibody. Therefore, L-VL-VH and S-VL-VH of the C4 scFvs were further used to optimize the induction conditions and to assess their binding specificities. In addition, the secretion efficiencies of the PelE signal peptides of different lengths were found to be comparable according to the experiments with the C4 scFvs (Fig. 6C) and GFP (Fig. S6). To determine the most suitable induction time, we used 0.1 mM IPTG to induce C4 scFv expression at 25°C and collected samples at different time intervals to assess secretion efficiency. After centrifugation, bacterial cells were removed, and the medium was concentrated by TCA. The Western blot data indicated that the PelE signal peptides of both lengths could secrete VL-VH into the medium after a 5-h IPTG induction. With increased time, VL-VH accumulation increased, until up to 20 h after induction, and the yield of L-VL-VH was similar to that of S-VL-VH (Fig. 7A). Using Ni-NTA agarose, soluble L-VL-VH and S-VL-VH were successfully purified from the culture medium and verified by SDSPAGE and Western blot analysis with an anti-His antibody (Fig. 7B), and they were then stored at 4°C for further study. Antigen-binding specificities of soluble C4 scFvs compared with that of C4 MAb. To analyze whether the purified soluble C4 scFvs still maintained the correct binding activity against potyviruses, healthy and TuMV-infected N. benthamiana plants were used as ELISA samples to coat plates, and the purified L-VL-VH and S-VL-VH were diluted to different concentrations and used as primary antibodies, which were detected by anti-His (secondary antibody) and anti-rabbit (tertiary antibody) antibodies. The result indicated that L-VL-VH and S-VL-VH had similar binding affinities for the TuMV CP at concentrations ranging from 0.0625 to 1 ␮g/ml (Fig. 8). However, their nonspecific binding increased when the C4 scFv concentration was raised to 2 ␮g/ml. Based on the ELISA data, the appropriate dilution of L-VL-VH and S-VL-VH was between 0.5 and 1 ␮g/ml. Because the two C4 scFvs displayed similar reactivities to the TuMV CP, 1 ␮g/ml of L-VL-VH was used as a primary antibody to analyze the specificity of soluble VL-VH by Western blot analysis. Crude saps from KoMV-, DsMV-, and TuMV-infected and healthy plants were used to compare the

Applied and Environmental Microbiology

aem.asm.org

6845

Liu et al.

FIG 9 Specificities of L-VL-VH and C4 MAb compared by Western blot analysis. Lane 1, healthy plant; lane 2, KoMV-infected plant; lane 3, DsMV-infected plant; lane 4, TuMV-infected plant. The duplicated blots were detected by L-VL-VH (left panel) and C4 MAb (right panel).

FIG 7 (A) Secretion of C4 scFvs monitored at different time points. One milliliter of culture medium (supernatant) of L-VL-VH and S-VL-VH was collected at 0, 2.5, 5, 7.5, and 20 h after IPTG induction. The total protein was concentrated and then analyzed by SDS-PAGE (lower panel) and Western blotting (upper panel) with an anti-His antibody. The expression level gradually increased until up to 20 h after induction. (B) Purification of soluble C4 scFvs. L-VL-VH and S-VL-VH, purified by Ni-NTA agarose after a 20-h IPTG induction, were analyzed by SDS-PAGE (right panel) and Western blotting (left panel). CBR, Coomassie brilliant blue R250.

binding specificities of L-VL-VH and C4 MAb. The data showed that the specificity of L-VL-VH was similar to that of C4 MAb and that it could specifically detect potyvirus-infected plants and did not cross-react with healthy plants (Fig. 9). Our results demonstrated that the soluble L-VL-VH and S-VL-VH of the C4 scFvs were correctly folded and that they maintained the same antigen-binding specificity as the C4 MAb. Detection limit assay of C4 MAb and L-VL-VH. To evaluate the detection limits of L-VL-VH and C4 MAb, recombinant TuMV CP was used as an antigen for I-ELISA. To mimic infected plant samples, recombinant CP was diluted with the crude sap of N. benthamiana to achieve 20 to 1,280 ng per well (Fig. 10). ELISA revealed that the detection limit of the C4 MAb was between 80 and 160 ng of the TuMV CP and that the detection limit of L-

FIG 8 I-ELISA of L-VL-VH and S-VL-VH against healthy (H) and TuMVinfected plants. Different concentrations of L-VL-VH and S-VL-VH were used as primary antibodies, and anti-His and anti-rabbit antibodies were used as secondary and tertiary antibodies, respectively. OD was measured at 405 nm after a 60-min incubation.

6846

aem.asm.org

VL-VH was between 160 and 320 ng (Fig. 10). Although the recombinant TuMV CP contained a His tag, the ELISA values of the negative controls obtained using the C4 MAb instead of L-VL-VH as a primary antibody, with detection by anti-His (secondary antibody) and anti-rabbit (tertiary antibody) antibodies, demonstrated that the anti-His antibody specifically reacted with L-VL-VH and did not react with the recombinant TuMV CP (Fig. 10). Accordingly, the detection limit of L-VL-VH was not overestimated. DISCUSSION

The aim of this study was to produce a MAb with broad-spectrum reactivity for indexing calla lily-infecting potyviruses. First, we attempted to purify virus particles from infected plants for use as immunogens and we then screened for MAbs that recognized several potyviruses, similar to previous reports of MAbs prepared against Johnsongrass mosaic virus (JGMV) (42), Tulip breaking virus (43), TEV (44, 45), and Peanut stripe virus (PStV) (46). Unfortunately, it was difficult to purify enough virus particles from calla lily to produce high-quality antiserum or antibody for detection, most likely due to the mucilage in the plant tissues (17, 47). Nevertheless, recombinant CPs expressed in an E. coli system proved to successfully overcome the problems of low yield and contamination by plant proteins during virion purification in calla lily (16–18). The antiserum prepared against the recombinant CP specifically recognized the native and recombinant CPs of target virus, such as DsMV, KoMV, and ZaMMV (16–18). Therefore, we used bacterially expressed protein as an antigen to generate MAb for potyvirus indexing. The tertiary structures of the potyvirus particles are similar due to the similar polypeptide folding and subunit packing of the CPs

FIG 10 Detection limits of L-VL-VH and C4 MAb, as determined by I-ELISA. C4 MAb (1 mg/ml) diluted 10,000⫻ and L-VL-VH (140 ␮g/ml) diluted 200⫻ with ECI buffer were used as primary antibodies to detect different amounts of TuMV recombinant capsid protein. For negative controls (NC), TuMV CPcoated wells were first incubated with C4 MAb (primary antibody), followed by anti-His (secondary antibody) and goat anti-rabbit conjugate for detection.

Applied and Environmental Microbiology

October 2015 Volume 81 Number 19

Potyvirus Broad-Spectrum MAb and Soluble scFvs

(10). The treatment of virions with trypsin has revealed that the Nand C-terminal regions are exposed on the surfaces of viral particles and that the central region is buried inside (48, 49). The variable termini, especially the large N terminus, establish the immunodominant regions, whereas the trypsin-resistant core region is highly conserved and contains potyvirus group-specific epitopes (49). Therefore, many antibodies prepared from intact potyvirus particles only specifically recognize the surface virusspecific epitopes at the N or C terminus of CPs (50, 51). Nevertheless, certain MAbs that are also raised against native virions such as TEV (44) or JGMV (42) cross-react with other potyviruses. However, mixtures of native and denatured virions of 12 potyvirus isolates have been used for immunization, leading to identification of a broad-spectrum PTY1 MAb from 30 screened MAbs that detects at least 55 potyvirus isolates (52). In contrast, we chose a recombinant protein for use as an immunogen, which was formed by fusion of three conserved core CP regions instead of from virus particles; therefore, we avoided the screening of MAbs that react to species-specific epitopes at the N or C terminus and also increased the opportunity for induction of broad-spectrum MAbs. Our strategy is different from that used for the PTY1 MAb. This strategy can aid in presentation of potyvirus groupspecific epitopes that can then be easily mapped. Based on the antigen-binding activity test results, we successfully selected a C4 MAb that was able to detect at least 14 potyviruses, as determined by I-ELISA and Western blot analysis (Fig. 2 and 3). Notably, not all fusion proteins are good immunogens for inducing antibody production. Although we fused five conserved CP fragments to generate the immunogen (see Fig. S1 in the supplemental material), the results of I-ELISA and Western blotting of the mouse antiserum showed that this recombinant core CP displayed poor immunogenicity and did not induce sufficient antibodies (data not shown). Hence, it is important to test the titers of target antibodies in immunized mice before cell fusion to generate the desired MAbs. Most epitopes can be classified as either linear (sequential) or conformational. It is generally accepted that a conformational epitope cannot be recognized by an antibody under denaturing conditions. According to our Western blot results, the C4 MAb showed a high binding affinity to potyviruses (Fig. 2B and 3B), suggesting that the epitope reacting with this C4 MAb should be sequential and not conformational. Jordan and Hammond have proposed that the epitope type may be determined by I-ELISA and triple antibody sandwich (TAS)-ELISA with disrupted and intact virions, respectively (52). MAbs that react best with plate-trapped, disrupted virus particles and CP subunits, as demonstrated by I-ELISA, recognize antigenic sites as cryptotopes, which are hidden within the intact virion (52). Because our immunogen was derived from the conserved core CP domain, the C4 MAb should react with a cryptotope, which is usually recognized after dissociation of the virion. Comparison of the I-ELISA and TAS-ELISA results revealed that the C4 MAb had a lower binding affinity to native virions, as demonstrated by TAS-ELISA (data not shown). This result again indicates that the epitope reacting to the C4 MAb is a cryptotope. Due to our design, the highly conserved CP sequence of calla lily-infecting potyviruses was used as an antigen; therefore, we expected that other potyviruses would be detected by the C4 MAb. Comparison of the reaction spectrum of the C4 MAb and PTY1 MAb (the potyvirus group antibody purchased from Agdia, Inc.)

October 2015 Volume 81 Number 19

by I-ELISA revealed that they both could detect many potyvirusinfected plants but that they had different binding affinities to very few infected samples (data not shown). These results demonstrate that the C4 MAb, similar to the PTY1 MAb, has wide-spectrum reactivity; however, the exact epitope sequence recognized by the C4 MAb requires further study. In this study, we also cloned VH and VL genes from a C4 MAbsecreting stable hybridoma cell line (C12-C4) and used a linker to connect VH and VL to construct C4 scFvs. Unexpectedly, we obtained two abnormal VL gene sequences, one of which contained a stop codon near the 3= ends, as determined by in silico translation. A similar result has been reported in construction of a recombinant anti-sperm antibody (RASA) from an MHS-8 hybridoma cell line, with identification of a frameshift in VL cDNA generated by RPAS kit primers (53). The aberrant light-chain mRNA may be produced by the myeloma fusion partner (54, 55). To increase the chance of obtaining the correct C4 VL, we attempted to use a primer mixture to amplify the dominant VL gene sequence. This sequence, determined by sequencing, was amplified by primers MuVK2 and MuJK4 and was found to encode a 112-amino-acid protein. Consequently, our experimental strategy avoided the interference of myeloma cell mRNAs and resulted in acquisition of the desired cDNA clones. Two formats, including VH-VL and VL-VH, have been previously used to construct scFvs. However, many reports have shown that the orientation of scFv domains may affect the binding affinity or protein expression level. For example, anti-c-Met VH-VL has been shown to exhibit ⬎5-fold increased productivity compared with that of VL-VH (56). Furthermore, the VL-VH format of scFv and bispecific diabody (bDAb) displays substantially better antigen binding activity and productivity than those of the VH-VL format (57). Our results showed that the domain orientation of the C4 scFv influenced binding affinity to anti-His antibody because the His tag of VH-VL exhibited lower binding affinity than that of VL-VH (Fig. 6C). One possible reason for this finding is the His tag may have been blocked by the non-fully denatured VH-VL. Although most evidence has shown that the effect of domain orientation may depend on the VH and VL sequences (56–58), the reason why domain orientation affects the binding affinity or structure of scFvs remains unclear. With regard to the C4 scFvs, our results indicated that VL-VH was more suitable for protein expression and application than VH-VL. Secretory production of scFvs is a common strategy for solving the inclusion body problem in E. coli expression systems. Several reports have shown that the PelB signal peptide can successfully secrete scFvs into the periplasm or medium (31, 38, 39). However, in our study, use of a commercial expression vector with the PelB signal peptide after transformation with PelB-VL-VH or PelBVH-VL resulted in no growth of either E. coli strain BL21(DE3) or Rosetta (DE3) on selection plates. This phenomenon might have occurred because the combination of the PelB signal peptide and the C4 scFv was inappropriate for these two E. coli strains. Hence, to solve the inclusion body problem, we instead used the PelE secretory signal peptide to test the secretion efficiencies of the C4 scFv. The use of a PelE SecB-dependent pathway for self-export to the inner membrane has been demonstrated (40); however, it has not previously been used for secretion of any other proteins. When PelE signal peptides of two different lengths were tested, both peptides were found to be completely cleaved by peptidase and to have similar secretion efficiencies (Fig. 6 and 7). Both pu-

Applied and Environmental Microbiology

aem.asm.org

6847

Liu et al.

rified L-VL-VH and S-VL-VH were functional (Fig. 8), and the C4 scFv showed the same binding specificity profile as that of the C4 MAb by Western blot analysis (Fig. 9). Furthermore, we fused the PelE signal peptides to the N terminus of GFP and demonstrated that this protein could be expressed and secreted into the medium, similar to what occurs following its expression in a pET20b(⫹) vector, which has the PelB signal peptide (Fig. S6). In summary, our data suggest that the PelE signal peptide has the potential to be developed as a feasible secretory signal peptide for secretion of expressed proteins by E. coli. In conclusion, this study generated a broad-spectrum C4 MAb using a strategy involving fusion of the CP core regions of three potyviruses to create an immunogen. Because the C4 MAb has a high antibody titer and specificity, it will be very useful for indexing and quarantining potyviruses. In our experience, it is a good alternative method to produce a highly specific MAb by means of a bacterially expressed protein compared with highly purified virions, the preparation of which is laborious and time-consuming. In addition, our results demonstrate the possibility of generating a broad-spectrum MAb by fusing the conserved regions of several target proteins to create an immunogen, instead of its generation by the traditional approach, in which it is derived from native and denatured forms of target proteins. The quick production of the C4 scFv in E. coli expression systems may replace traditional antiserum production, as long as the inclusion body problem can be solved. The PelE signal peptide was used to increase secretion of the C4 scFv, and functional C4 scFv, as well as other proteins, were found to be secreted into the medium. Because the PelE signal peptide could aid in the secretion of expressed proteins, it will be very useful in the E. coli system. Both the C4 MAb and soluble C4 scFv (VL-VH) showed high binding specificities to the potyvirus CPs. Moreover, to construct a functional scFv, the influence of the VH and VL orientation on the binding activity and structure should also be considered. Some reports have shown that transgenic plants expressing an scFv that specifically detects plant virus proteins exhibit viral resistance and tolerance (59, 60). Therefore, future studies should be performed to examine transgenic plants expressing the broad-spectrum C4 scFv that may achieve wideranging antibody-meditated resistance to potyviruses.

6. 7.

8. 9.

10. 11. 12. 13.

14. 15.

16. 17. 18. 19.

ACKNOWLEDGMENTS This study was supported by grants from the National Science Council (NSC94-2317-B-002-013, NSC96-2317-B-002-002, and NSC97-2317-B002-002), Executive Yuan, Taiwan, Republic of China. We thank Xiao-Chan He for her help in the generation of the monoclonal antibody.

20.

21.

REFERENCES 1. García JA, Glasa M, Cambra M, Candresse T. 2014. Plum pox virus and sharka: a model potyvirus and a major disease. Mol Plant Pathol 15:226 – 241. http://dx.doi.org/10.1111/mpp.12083. 2. Scholthof KB, Adkins S, Czosnek H, Palukaitis P, Jacquot E, Hohn T, Hohn B, Saunders K, Candresse T, Ahlquist P, Hemenway C, Foster GD. 2011. Top 10 plant virus in molecular plant pathology. Mol Plant Pathol 12:938 –954. http://dx.doi.org/10.1111/j.1364-3703.2011.00752.x. 3. King AMQ, Adams MJ, Carstens EB, Lefkowitz EJ. 2012. Virus taxonomy: classification and nomenclature of viruses. Ninth report of the International Committee on Taxonomy of Viruses. Elsevier Academic Press, London, England. 4. Chung BY, Miller WA, Atkins JF, Firth AE. 2008. An overlapping essential gene in the Potyviridae. Proc Natl Acad Sci U S A 105:5897–5902. http://dx.doi.org/10.1073/pnas.0800468105. 5. Vijayapalani P, Maeshima M, Nagasaki-Takekuchi N, Miller WA. 2012.

6848

aem.asm.org

22. 23.

24.

25.

Interaction of the trans-frame potyvirus protein P3N-PIPO with host protein PCaP1 facilitates potyvirus movement. PLoS Pathog 8(4):e1002639. http://dx.doi.org/10.1371/journal.ppat.1002639. Dolja VV, Haldeman R, Robertson NL, Dougherty WG, Carrington JC. 1994. Distinct functions of capsid protein in assembly and movement of tobacco etch potyvirus in plants. EMBO J 13:1482–1491. Dolja VV, Haldeman-Cahill R, Montgomery AE, Vandenbosch KA, Carrington JC. 1995. Capsid protein determinants involved in cell-to-cell and long distance movement of tobacco etch potyvirus. Virology 206: 1007–1016. http://dx.doi.org/10.1006/viro.1995.1023. Jossey S, Hobbs HA, Domier LL. 2013. Role of Soybean mosaic virusencoded proteins in seed and aphid transmission in soybean. Phytopathology 103:941–948. http://dx.doi.org/10.1094/PHYTO-09-12-0248-R. Seo JK, Vo Phan MS, Kang SH, Choi HS, Kim KH. 2013. The charged residues in the surface-exposed C-terminus of the Soybean mosaic virus coat protein are critical for cell-to-cell movement. Virology 446:95–101. http://dx.doi.org/10.1016/j.virol.2013.07.033. Shukla DD, Ward CW. 1989. Structure of potyvirus coat proteins and its application in the taxonomy of the potyvirus group. Adv Virus Res 36: 273–314. http://dx.doi.org/10.1016/S0065-3527(08)60588-6. Chen CC, Chang CA, Tsai HT, Hsu HT. 2004. Identification of a potyvirus causing latent infection in calla lilies. Plant Dis 88:1046. http: //dx.doi.org/10.1094/PDIS.2004.88.9.1046A. Chen YL, Chung FC, Chang YC. 1998. Molecular cloning and characterization of Dasheen mosaic potyvirus from aroid plants in Taiwan. Plant Pathol Bull 7:226 –227. Chen CC, Chao CH, Chen CC, Yeh SD, Tsai HT, Chang CA. 2003. Identification of Turnip mosaic virus isolates causing yellow stripe and spot on calla lily. Plant Dis 87:901–905. http://dx.doi.org/10.1094/PDIS.2003 .87.8.901. Chang YC, Chen YL, Chung FC. 2001. Mosaic disease of calla lily caused by a new potyvirus in Taiwan. Plant Dis 85:1289. http://dx.doi.org/10 .1094/PDIS.2001.85.12.1289C. Huang CH, Chang YC. 2005. Identification and molecular characterization of Zantedeschia mild mosaic virus, a new calla lily-infecting potyvirus. Arch Virol 150:1221–1230. http://dx.doi.org/10.1007/s00705 -004-0488-3. Huang WZ, Chung FC, Chen YL, Huang CH, Chang YC. 2005. Development and comparison of three detection methods for calla lily-infecting Dasheen mosaic virus. Plant Pathol Bull 14:239 –250. Huang CH, Hu WC, Yang TC, Chang YC. 2007. Zantedeschia mild mosaic virus, a new widespread virus in calla lily, detected by ELISA, dot-blot hybridization and IC-RT-PCR. Plant Pathol 56:183–189. Hu WC, Lin WF, Huang WZ, Huang CH, Chang YC. 2007. The production and application of the antiserum against Zantedeschia mosaic virus recombinant coat protein. Plant Pathol Bull 16:31– 40. Hu WC, Huang CH, Lee SC, Wu CI, Chang YC. 2010. Detection of four calla potyviruses by multiplex RT-PCR using nad5 mRNA as an internal control. Eur J Plant Pathol 126:43–52. http://dx.doi.org/10.1007/s10658 -009-9519-y. Lai H, He J, Hurtado J, Stahnke J, Fuchs A, Mehlhop E, Gorlatov S, Loos A, Diamond MS, Chen Q. 2014. Structural and functional characterization of an anti-West Nile virus monoclonal antibody and its singlechain variant produced in glycoengineered plants. Plant Biotechnol J 12: 1098 –1107. http://dx.doi.org/10.1111/pbi.12217. Maleki LA, Baradaran B, Majidi J, Mohammadian M, Shahneh FZ. 2013. Future prospects of monoclonal antibodies as magic bullets in immunotherapy. Hum Antibodies 22:9 –13. Ahmad ZA, Yeap SK, Ali AM, Ho WY, Alitheen NB, Hamid M. 2012. scFv antibody: principles and clinical application. Clin Dev Immunol 2012:980250. http://dx.doi.org/10.1155/2012/980250. Bird RE, Hardman KD, Jacobson JW, Johnson S, Kaufman BM, Lee SM, Lee T, Pope SH, Riordan GS, Whitlow M. 1988. Single-chain antigen-binding proteins. Science 242:423– 426. http://dx.doi.org/10 .1126/science.3140379. Rouis S, Lafaye P, Jaoua-Aydi L, Sghaier Z, Ayadi H, Gargouri-Bouzid R. 2006. Cloning and expression of functional single-chain Fv antibodies directed against NIa and coat proteins of potato virus Y. J Virol Methods 137:1– 6. http://dx.doi.org/10.1016/j.jviromet.2006.05.011. Hanes J, Schaffitzel C, Knappik A, Plückthun A. 2000. Picomolar affinity antibodies from a fully synthetic naive library selected and evolved by ribosome display. Nat Biotechnol 18:1287–1292. http://dx.doi.org/10 .1038/82407.

Applied and Environmental Microbiology

October 2015 Volume 81 Number 19

Potyvirus Broad-Spectrum MAb and Soluble scFvs

26. Harper K, Kerschbaumer RJ, Ziegler A, Macintosh SM, Cowan GH, Himmler G, Mayo MA, Torrance L. 1997. A scFv-alkaline phosphatase fusion protein which detects potato leafroll luteovirus in plant extracts by ELISA. J Virol Methods 63:237–242. http://dx.doi.org/10.1016/S0166 -0934(96)02133-7. 27. Harper K, Toth RL, Mayo MA, Torrance L. 1999. Properties of a panel of single chain variable fragments against Potato leafroll virus obtained from two phage display libraries. J Virol Methods 81:159 –168. http://dx .doi.org/10.1016/S0166-0934(99)00071-3. 28. Hust M, Maiss E, Jacobsen HJ, Reinard T. 2002. The production of a genus-specific recombinant antibody (scFv) using a recombinant potyvirus protease. J Virol Methods 106:225–233. http://dx.doi.org/10.1016 /S0166-0934(02)00166-0. 29. Al-Mrabeh A, Ziegler A, Cowan G, Torrance L. 2009. A fully recombinant ELISA using in vivo biotinylated antibody fragments for the detection of potato leafroll virus. J Virol Methods 159:200 –205. http://dx.doi.org/10 .1016/j.jviromet.2009.03.025. 30. Griep RA, van Twisk C, Kerschbaumer RJ, Harper K, Torrance L, Himmler G, van der Wolf JM, Schots A. 1999. pSKAP/S: an expression vector for the production of single-chain Fv alkaline phosphatase fusion proteins. Protein Expr Purif 16:63– 69. http://dx.doi.org/10.1006/prep .1999.1041. 31. Mukherjee KJ, Rowe DC, Watkins NA, Summers DK. 2004. Studies of single-chain antibody expression in quiescent Escherichia coli. Appl Environ Microbiol 70:3005–3012. http://dx.doi.org/10.1128/AEM.70.5.3005 -3012.2004. 32. Zheng F, Qiu Y, Chen Y, Chen P, Zhu Y, Xie W, Zhu H, Zhu J. 2009. Cloning, purification and bioactivity assay of human CD28 single-chain antibody in Escherichia coli. Cytotechnology 60:85–94. http://dx.doi.org /10.1007/s10616-009-9218-8. 33. Huang CH, Chang YC. 2006. Basella rugose mosaic, a new potyvirus infecting Basella rubra. Plant Pathol 55:819. http://dx.doi.org/10.1111/j .1365-3059.2006.01478.x. 34. Wu CI. 2004. Molecular characterization and antibody preparation of a Turnip mosaic virus isolate infecting calla lily. Master’s thesis. National Taiwan University, Taipei, Taiwan. 35. Leger OJP, Saldanha JW. 2000. Preparation of recombinant antibodies from immune rodent spleens and the design of their humanization by CDR grafting, p 25– 65. In Shepherd P, Dean C (ed), Monoclonal antibodies—a practical approach. Oxford University Press, New York, NY. 36. Brochet X, Lefranc MP, Giudicelli V. 2008. IMGT/V-QUEST: the highly customized and integrated system for IG and TR standardized V-J and V-D-J sequence analysis. Nucleic Acids Res 36:W503–508. http://dx.doi .org/10.1093/nar/gkn316. 37. Giudicelli V, Brochet X, Lefranc MP. 2011. IMGT/V-QUEST: IMGT standardized analysis of the immunoglobulin (IG) and T cell receptor (TR) nucleotide sequences. Cold Spring Harb Protoc 2011:695–715. 38. Dai K, Zhu H, Ruan C. 2003. Generation and characterization of recombinant single chain Fv antibody that recognizes platelet glycoprotein Ib␣. Thromb Res 109:137–144. http://dx.doi.org/10.1016/S0049-3848(03) 00152-X. 39. Sletta H, Tøndervik A, Hakvåg S, Vee Aune TE, Nedal A, Aune R, Evensen G, Valla S, Ellingsen TE, Brautaset T. 2007. The presence of N-terminal secretion signal sequences leads to strong stimulation of the total expression levels of three tested medically important proteins during high-cell-density cultivations of Escherichia coli. Appl Environ Microbiol 73:906 –912. http://dx.doi.org/10.1128/AEM.01804-06. 40. He SY, Schoedel C, Chatterjee AK, Collmer A. 1991. Extracellular secretion of pectate lyase by the Erwinia chrysanthemi out pathway is dependent upon Sec-mediated export across the inner membrane. J Bacteriol 14:4310 – 4317. 41. Petersen TN, Brunak S, von Heijne G, Nielsen H. 2011. SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Methods 8:785–786. http://dx.doi.org/10.1038/nmeth.1701. 42. Shukla DD, Tribbick G, Mason TJ, Hewish DR, Geysen HM, Ward CW. 1989. Localization of virus-specific and group-specific epitopes of plant potyviruses by systematic immunochemical analysis of overlapping peptide fragments. Proc Natl Acad Sci U S A 86:8192– 8196. http://dx.doi.org /10.1073/pnas.86.21.8192. 43. Hsu HT, Franssen JM, van der Hulst CTC, Derks AFLM, Lawson RH.

October 2015 Volume 81 Number 19

44. 45.

46. 47.

48.

49.

50. 51.

52.

53.

54. 55. 56.

57.

58.

59.

60.

1988. Factors affecting selection of epitope specificity of monoclonal antibodies to tulip breaking potyvirus. Phytopathology 78:1337–1340. http: //dx.doi.org/10.1094/Phyto-78-1337. Dougherty WG, Willis L, Johnston RE. 1985. Topographic analysis of tobacco etch virus capsid protein epitopes. Virology 144:66 –72. http://dx .doi.org/10.1016/0042-6822(85)90305-8. Joisson C, Dubs MC, Van Regenmortel MH. 1992. Cross-reactive potential of monoclonal antibodies raised against proteolysed tobacco etch virus. Res Virol 143:155–166. http://dx.doi.org/10.1016/S0923-2516(06) 80100-7. Richter J, Rabenstein F, Proll E, Vetten HJ. 1995. Use of cross-reactive antibodies to detect members of the Potyviridae. J Phytopathol 143:459 – 464. http://dx.doi.org/10.1111/j.1439-0434.1995.tb04554.x. Li RH, Zettler FW, Purcifull DE, Hiebert E. 1998. The nucleotide sequence of the 3=-terminal region of dasheen mosaic virus (Caladium isolate) and expression of its coat protein in Escherichia coli for antiserum production. Arch Virol 143:2461–2469. http://dx.doi.org/10.1007 /s007050050476. Anindya R, Savithri HS. 2003. Surface-exposed amino- and carboxyterminal residues are crucial for the initiation of assembly in Pepper vein banding virus: a flexuous rod-shaped virus. Virology 316:325–336. http: //dx.doi.org/10.1016/S0042-6822(03)00593-2. Shukla DD, Strike PM, Tracy SL, Gough KH, Ward CW. 1988. The N and C termini of the coat proteins of potyviruses are surface located and the N terminus contains the major virus-specific epitopes. J Gen Virol 69:1497–1508. http://dx.doi.org/10.1099/0022-1317-69-7-1497. Desbiez C, Gal-On A, Raccah B, Lecoq H. 1997. Characterization of epitopes on zucchini yellow mosaic potyvirus coat protein permits studies on the interactions between strains. J Gen Virol 78:2073–2076. Keller H, Pomp R, Bakker J, Schots A. 2005. Epitope identification and in silico prediction of the specificity of antibodies binding to the coat proteins of Potato virus Y strains. Eur J Plant Pathol 111:391–397. http: //dx.doi.org/10.1007/s10658-004-6892-4. Jordan R, Hammond J. 1991. Comparison and differentiation of potyvirus isolates and identification of strain-, virus-, subgroup-specific and potyvirus group-common epitopes using monoclonal antibodies. J Gen Virol 72:25–36. http://dx.doi.org/10.1099/0022-1317-72-1-25. Norton EJ, Diekman AB, Westbrook VA, Flickinger CJ, Herr JC. 2001. RASA, a recombinant single-chain variable fragment (scFv) antibody directed against the human sperm surface: implications for novel contraceptives. Hum Reprod 16:1854 –1860. http://dx.doi.org/10.1093/humrep/16 .9.1854. Carroll WL, Mendel E, Levy S. 1988. Hybridoma fusion cell lines contain an aberrant kappa transcript. Mol Immunol 25:991–995. http://dx.doi.org /10.1016/0161-5890(88)90005-3. Shulman M, Wilde CD, Köhler G. 1978. A better cell line for making hybridomas secreting specific antibodies. Nature 276:269 –270. http://dx .doi.org/10.1038/276269a0. Kim YJ, Neelamegam R, Heo MA, Edwardraja S, Paik HJ, Lee SG. 2008. Improving the productivity of single-chain Fv antibody against c-Met by rearranging the order of its variable domains. J Microbiol Biotechnol 18: 1186 –1190. Lu D, Jimenez X, Witte L, Zhu Z. 2004. The effect of variable domain orientation and arrangement on the antigen-binding activity of a recombinant human bispecific diabody. Biochem Biophys Res Commun 28: 507–513. Luo D, Mah N, Krantz M, Wilde K, Wishart D, Zhang Y, Jacobs F, Martin L. 1995. Vl-linker-Vh orientation-dependent expression of single chain Fv-containing an engineered disulfide-stabilized bond in the framework regions. J Biochem 118:825– 831. Boonrod K, Galetzka D, Nagy PD, Conrad U, Krczal G. 2004. Singlechain antibodies against a plant viral RNA-dependent RNA polymerase confer virus resistance. Nat Biotechnol 22:856 – 862. http://dx.doi.org/10 .1038/nbt983. Nölke G, Cobanov P, Uhde-Holzem K, Reustle G, Fischer R, Schillberg S. 2009. Grapevine fanleaf virus (GFLV)-specific antibodies confer GFLV and Arabis mosaic virus (ArMV) resistance in Nicotiana benthamiana. Mol Plant Pathol 10:41– 49. http://dx.doi.org/10.1111/j.1364-3703 .2008.00510.x.

Applied and Environmental Microbiology

aem.asm.org

6849

A Strategy for Generating a Broad-Spectrum Monoclonal Antibody and Soluble Single-Chain Variable Fragments against Plant Potyviruses.

Potyviruses are major pathogens that often cause mixed infection in calla lilies. To reduce the time and cost of virus indexing, a detection method fo...
NAN Sizes 0 Downloads 11 Views