Baculovirus display of single chain antibody (scFv) using a novel signal peptide

Insect cells and baculovirus infection

Spodoptera frugiperda Sf9 cells were maintained as monolayers at 28°C in Grace's insect medium supplemented with 10% fetal bovine serum (FBS, Sigma), penicillin (200 U/ml), and streptomycin (200 μg/ml; Gibco-Invitrogen). They were infected with one, or coinfected with two or more recombinant BVs simultaneously at a multiplicity of infection (MOI) ranging from 2.5 to 20 pfu/cell, as previously described [24, 28].

Construction of scFv against HIV-1 MAp17 and CD147

The hybridoma cell line MH-SVM33C9/ATCC HB-8975 was obtained from the American Type Culture Collection (ATCC, Manassas, VA). The monoclonal antibody MH-SVM33 reacts with the conserved epitope ¹²¹DTGHSSQVSQNY¹³² at the C-terminus of the matrix protein (MAp17) of HIV-1 [22, 23]. Total RNA was extracted from hybridoma cells using the RNeasy Mini kit (Qiagen Inc., Hilden, Germany), and the first strand of cDNA was synthesized using the oligodT-18 primer of the Transcriptor High Fidelity cDNA synthesis kit (Roche, Mannheim, Germany). The variable regions (V) of heavy (V_H) and light chains (V_L) were then amplified from cDNA using specific forward (Fw) and reverse (Rev) primers. Fw-V_HP17 (5'-ATATGCTAGCGGCCCAGGCGGCCCAGATCCAGTTGGTGCAGT-3') and Rev-V_HP17 (5'-CGACCCTCCACCGCGGACCCGCCACCTCCAGACCCTCCGCCACCTGCA GAGACAGTGACCAGAGTCCC-3') were used for the V_H fragment, and Fw-V_LP17 (5'-GGGTCCGGCGGTGGAGGGTCGGATGTTGTGATGACCCAGACTCCA-3') and Rev-V_LP17 (5'-ATATAAGCTTTCATTAAGCGTAGTCCGGAACGTCGTACGGGTACTGGCCGCCCTGGCCTT TGATTTCCAGC-3') for the V_L fragment. The fragment encoding the single-chain antibody to MAp17 (scFv/p17) was constructed by overlapping PCR using Fw-V_HP17 and Rev-V_LP17 primers, both containing a Sfi I site (shown underlined). The construction and characterization of HA-tagged scFv M6-1B9 directed against CD147 have been described in a previous study [38, 39].

Recombinant BV

Foreign genes were inserted into the genome of AcMNPV, under the control of a chimeric AcMNPV-Galleria mellonella MNPV polyhedrin promoter in the case of recombinant HIV-1 Gag polyproteins [24, 28, 29, 31, 32], or under the control of the AcMNPV polyhedrin promoter in the case of pBlueBac4.5-derived vectors. AcMNPV-Pr55Gag, which expressed the N-myristoylated full-length Gag polyprotein (Pr55Gag) of HIV-1, has been described in detail in previous studies [30–32, 36, 40]. The recombinant AcMNPV expressing CAR (BV^CAR), the high affinity receptor for Coxsackie B and Adenovirus, has been described and characterized in a previous study [41]. BV^CAR virions have been shown to display the CAR glycoprotein at their surface [41]. A DNA fragment coding for H₆MA-CA, a non-N-myristoylated, carboxyterminal-truncated version of HIV-1 Gag polyprotein containing the matrix (MA) and capsid (CA) domains with an oligo-histidine (H₆) tag at its N-terminus, was generating using the following pairs of primers: the first pair consisted of Fw primers 5'-CTAGCATGGGTGCGAGAG-3' and 5'-CATGGGTGCGAGAGCG-3' and the second pair consisted of Rev primers 5'-CTTACTACAAAACTCTTGCTTTATG-3' and 5'-GTACCTTACTAC AAAACTCTTGC-3'. Two PCR reactions were performed using the HIV-1 plasmid pNL4-3 as the template. The PCR products from both reactions were then mixed, denatured, and hybridized to obtain DNA fragments with Nhe I and Kpn I cohesive ends, competent for ligation to pBlueBac4.5-His intermediate vector linearized with Nhe I and Kpn I. Plasmid pBlueBac4.5-His was derived from pBlueBac4.5 (Invitrogen, San Diego, CA) by insertion of a sequence coding for the (H₆) tag and a GSGSAS linker upstream to the Nhe I site. Sf9 cells were cotransfected with pBlueBac4.5-H₆MA-CA and linearised BV DNA (Bac-N-Blue™ Transfection kit; Invitrogen). Positive Sf9 cells harboring recombinant BVs were isolated using the blue plaque selection method after beta-galactosidase staining. Recombinant BV, abbreviated BV-H₆MA-CA, was isolated using the blue plaque selection method as above. Recombinant H₆MA-CA protein was produced in Sf9 cells infected with BV-H₆MA-CA. Sf9 cells were harvested at 48 h post infection (pi), and H₆MA-CA protein recovered from clarified Sf9 cell lysate by affinity chromatography on Ni²⁺-NTA-agarose column (Qiagen, Hilden, Germany), as previously described [42]. Protein concentration in H₆MA-CA samples was determined by Bradford protein assay (Pierce; Thermo Fisher Scientific Inc., Rockford, IL, USA). Recombinant H₆MA-CA protein was used as the antigenic substrate for scFvG2/p17 and scFvE2/p17 in ELISA and co-immunoprecipitation experiments.

The DNA fragment encoding scFv/p17, obtained as described above, was cloned into the Nhe I and Hind III sites of the pBlueBac4.5 plasmid. The 5' and 3' ends of scFv/p17 fragment in the pBlueBac-scFv/p17 vector were then modified by oligonucleotide insertion at both Nhe I and Hind III sites, to obtain two versions of the scFv/p17 cDNA. One encoded the dipeptide Met-Gly at the N-terminus of scFv/p17, generating the scFvG2/p17 clone, the other the dipeptide Met-Glu, generating the scFvE2/p17 clone. At the 3' end of both clones, we inserted an oligonucleotide encoding the Influenza A virus hemagglutinin epitope YPYDVPDYA (HA tag). Sf9 cells were then cotransfected with the resulting plasmid harboring the scFv/p17 coding sequence and linearised BV DNA. Recombinant BVs were isolated using the blue plaque selection method, as described above

The DNA fragment for the HA-tagged scFv M6-1B9 was amplified using plasmid pComb3X-scFv-M6-1B9 as the template [39], with the M6-1B9 Fw primer (5'-GAGGAGGAGCTGGCCCAGGCGGCCCAGATCCAGTTGGTGCAGTCTGGAGAGCTAGTGATGACCCAGACTCCAGC-3') encoding the N-terminal 18 amino acids of scFvE2/p17 (¹MEASLAAQAAQIQLVQSG¹⁸), and the M6-1B9 Rev primer (5'-CTCCTCCTCGGCCGCCCTGGCCACTAGTGACAGATGGGGCTG-3'). The scFv-M6-1B9 was then cloned into the Sfi I site of Sfi I-restricted pBlueBac-scFvE2. The recombinant BV was isolated as described above, and abbreviated BV-N18E2/M6.

Isolation of BV particles

Concentrated stocks of recombinant BV expressing scFv/p17 molecules, BV-scFvG2/p17 and BV-scFvE2/p17, respectively, were prepared as follows [41]. Infected Sf9 cell culture supernatants were harvested at 50 to 60 h post infection (pi), and clarified by centrifugation at 2,400 rpm and 4°C for 10 min. Aliquots (11-ml) of clarified culture supernatant were subjected to ultracentrifugation at 28,000 rpm for 1 h at 4°C through a 1-ml sucrose cushion (20% sucrose, w:v in PBS) in Beckman SW41 rotor. Each baculoviral pellet was resuspended by gentle shaking in sterile phosphate-buffered saline (PBS) overnight at 4°C (100 μl per centrifuge tube). The titers of BV suspensions ranged usually between 5 × 10⁹ and 1 × 10¹⁰ pfu/ml, as determined by plaque titration in Sf9 cells (pfu/ml), which corresponded to 1 × 10¹² to 5 × 10¹² BV physical particles per ml [12, 41]. When needed, e.g. for electron microscopy, BV particles were further purified by isopycnic ultracentrifugation in linear sucrose-D₂O gradient [35, 40]. Gradients (10-ml total volume, 30-50%, w:v) were generated from a 50% sucrose solution made in D₂O buffered to pH 7.2 with NaOH, and a 30% sucrose solution made in 10 mM Tris-HCl, pH 7.2, 150 mM NaCl, 5.7 mM Na₂EDTA. The gradients were centrifuged for 18 h at 28,000 rpm in a Beckman SW41 rotor. 0.5 ml-fractions were collected from the top, and proteins analyzed by SDS-PAGE and Western blotting with the required antibodies. BV particles were recovered in fractions with an apparent density ranging from 1.08 to 1.15.

Cell fractionation and scFv purification

BV-infected Sf9 cells were harvested between 48 and 72 h pi. After cell lysis, cellular fractionation was performed using the FractionPREP™ Cell Fractionation System (Medical & Biological Laboratories Co. Ltd., Nagoya, Japan). Four subcellular protein fractions were thus obtained, cytosol, nucleus, membrane/particulate and cytoskeletal fractions. The scFvE2/p17 protein was purified by affinity selection on anti-HA tag antibody Affimatrix gel (Roche Applied Science, IN, USA), following the manufacturer's instructions. Each fraction eluted from the affinity gel, using elution buffer containing HA peptide (Roche Applied Science) as binding competitor to displace the bound proteins, was analyzed by SDS-PAGE, and scFvG2/p17 or scFvE2/p17 detected by Western blotting, using anti-HA tag antibody, as described below.

Gel electrophoresis, membrane transfer and antibodies

Polyacrylamide gel electrophoresis of SDS-denatured protein samples (SDS-PAGE), and immunoblotting analysis have been described in detail in previous studies [34, 35, 40]. Briefly, proteins were electrophoresed in SDS-denaturing, 10%-polyacrylamide gel and electrically transferred to nitrocellulose membrane (Hybond™-C-extra; Amersham Biosciences). Blots were blocked in 5% skimmed milk in Tris-buffered saline (TBS) containing 0.05% Tween-20 (TBS-T), rinsed in TBS-T, then successively incubated with mouse monoclonal antibody to Influenza A virus hemagglutinin epitope YPYDVPDYA (HA tag antibody; Sigma, St Louis, MO, USA), or primary rabbit anti-Gag antibody, followed by relevant anti-IgG secondary antibodies, at working dilutions ranging from 1:1,000 to 1:10,000. Anti-HIV-1 Gag rabbit polyclonal antibody (laboratory-made; [35]) was raised in rabbit by injection of bacterially-expressed, GST-fused and affinity-purified carboxyterminally truncated Gag protein consisting of the full-length MA domain and the first seventy-eight residues of the CA domain (Pst I site; gag_Lai sequence). Monoclonal antibody against baculoviral envelope GP64 glycoprotein, clone AcV1 was purchased from Santa Cruz Biotechnology (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). Apparent molecular weights were estimated by comparison with prestained protein markers (Precison Plus Protein™ Standards, Dual Color; BioRad Laboratories, Inc., Bio-Rad France). For protein quantification, blots were scanned and protein bands were quantitated by densitometry, using the VersaDoc image analyzer and the Quantity One program (BioRad).

Indirect ELISA

The functionality of recombinant scFv/p17 was evaluated by their binding activity to the synthetic MAp17 peptide ¹²¹DTGHSSQVSQNY¹³² (GenScript; Piscataway, USA) and H₆MA-CA protein prepared as above, in standard indirect ELISA procedure. In brief, aliquots (100 μl) of MA p17 peptide solution at 50 μg/ml, or of H₆MA-CA protein solution at 5 μg/ml in coating buffer (0.1 M NaHCO₃, pH 9.6) were incubated overnight at 4°C in 96-well microtiter plates (NUNC, Roskilde, Denmark). The coated wells were blocked with 200 μl of blocking buffer (2% bovine serum albumin in TBS; TBS-BSA) for 1 h at room temperature (RT), then washed five times with washing buffer (0.05% Tween-20 in TBS; TBS-T). Aliquots (100 μl) from clarified cellular lysates of BV-G2/p17- or BV-E2/p17-infected Sf9 cells in blocking buffer was added to each well and incubation proceeded for 1 hr at RT. After five cycles of washing with TBS-T, monoclonal anti-HA tag antibody (Sigma) was added at dilution 1:2,500 in TBS-BSA, and incubated for 1 hr at RT. After extensive washing with TBS-T, the reaction was developed by addition of 100 μl p-nitrophenylphosphate substrate (Sigma), and OD measured at 405 nm using an ELISA plate reader at the optimum time. The antigen-binding activity of scFv-M6-1B9 and chimeric scFvE2/M6-1B9 was assessed in ELISA using immobilized recombinant CD147-biotin carboxyl carrier protein (BCCP) fusion protein as the antigen, produced as previously described [39].

Competition ELISA

Microtiter plates (NUNC) were coated with 50 μl of 10 μg/ml of avidin in coating buffer (0.1 M NaHCO3, pH 6.8) and incubated overnight at 4°C in moist chamber. The coated wells were then blocked with 200 μl of blocking buffer (2% BSA in TBS) for 1 h at RT, then washed four times with washing buffer (0.05% Tween-20 in TBS). 50 μl-aliquots of 50 μg/ml of biotinylated-peptide 17.1 (Bp17.1) in blocking buffer were added to each well and incubated for 1 h at RT. In parallel, scFvE2/p17 purified by anti-HA tag affinity as described above was mixed with p17 peptide at the final peptide concentration of 1 μg/ml, and incubated for 1 h at RT. After washing the wells, the mixture was added to the wells and incubated for 1 h at RT. Bound scFvE2/p17 was monitored by adding 50 μl of HRP-conjugated monoclonal anti-HA tag antibody (Sigma) at dilution 1:1,000 in blocking buffer. The wells were washed four times prior to the addition of 50 μl of 3,3',5,5'-tetramethyl-benzidine (TMB) substrate. Reaction was stopped by addition of 50 μl of 1 N HCl, and optical densities (OD) at 450 nm were measured using an ELISA plate reader. The percentage of inhibition (PI) was given by the following formula: PI = 100-[(B:Bo) × 100], where B and Bo were the OD values for scFvE2/p17 with and without inhibitor, respectively.

Co-immunoprecipitation

Sf9 cells co-infected with BV-E2/p17 (or BV-G2/p17) and BV-H₆MA-CA were harvested by centrifugation at 5,000 rpm for 10 min at 4°C. The cell pellets were washed 3 times with cold TBS, then resuspended in lysis buffer (1% Brij-35 in TBS) and incubated on ice for 45 min. The cell lysates were clarified by centrifugation at 13,000 rpm for 1 h at 4°C, and supernatants added to pre-washed anti-HA tag affinity gel. After incubation overnight at 4°C with gentle shaking, the resin was washed 3 times with washing buffer (20 mM Tris, pH7.5, 0.1 M NaCl, 0.1 mM EDTA, 0.05% Tween20). Immunocaptured proteins were eluted from the affinity gel by heating in SDS-containing buffer (20 mM Tris pH 7,5, 2 mM EDTA, 5% SDS, 20% glycerol, 200 mM DTT, 0.02% bromophenol blue) at 100°C for 5 min. The resin was pelleted by centrifugation and the supernatant analyzed by SDS-PAGE and Western blotting, using anti-His₆ tag and anti-HA tag primary antibodies, both purchased from Sigma.

Immunofluorescence (IF) microscopy

BV-infected Sf9 cell monolayers were harvested at 48 h pi, fixed with 3% paraformaldehyde in phosphate buffered saline (PBS) and permeabilized in 0.2% (v/v) Triton X-100 in PBS. Cells were blocked with 3% BSA in PBS (PBS-BSA), and HA-tagged scFv/p17 detected by reaction with mAb anti-HA (1:10,000 in PBS-BSA) and Alexa Fluor^® 488-labeled goat anti-mouse IgG antibody (Molecular Probes, Invitrogen). For double labeling of scFv/p17 and Gag proteins, cell samples were reacted with rabbit anti-Gag antibody (1:1,000 in PBS-BSA) and Alexa Fluor^® 546-labeled goat anti-rabbit IgG (Molecular Probes, Invitrogen). Samples were treated with DAPI and mounted on slides. For conventional IF microscopy, images were acquired using an Axiovert 135 inverted microscope (Zeiss) equiped with an AxioCam video camera. For confocal microscopy, samples were analyzed using a Leica TCS SP2 confocal microscope.

Flow cytometry

ScFv-expressing Sf9 cells were resuspended in 200 μl PBS and incubated with monoclonal anti-HA tag antibody for 1 h at room temperature (RT), and at the dilution recommended by the manufacturer (Sigma), then pelleted. The cell pellet was resuspended in 200 μl PBS and reacted with Alexa Fluor^® 488-labeled goat anti-mouse IgG antibody (Molecular Probes, Invitrogen). The cell suspension was then diluted with 10 volumes of PBS, and analyzed by flow cytometry using a BD FACSCanto™ II cytometer (Becton Dickinson Biosciences). At least 10,000 events were acquired for each experiment using the DIVA 6 software (Becton Dickinson).

Electron microscopy (EM)

Specimens were processed for EM and observed as previously described [41]. In brief, pelleted virions of BV-E2/p17 and BV-G2/p17 were resuspended in 20 μl-aliquots of 0.14 M NaCl, 0.05 M Tris-HCl buffer, pH 8.2, and adsorbed onto carbon-coated formvar membrane on nickel grids. The grids were incubated with primary antibody (anti-HA tag monoclonal antibody) at a dilution of 1: 50 in TBS for 1 h at room temperature (RT). After rinsing with TBS, the grids were post-incubated with 20-nm colloidal gold-tagged goat anti-mouse IgG antibody (British Biocell International Ltd, Cardiff, UK; diluted to 1: 50 in TBS) for 30 min at RT. After rinsing with TBS, the specimens were negatively stained with 1% uranyl acetate in H₂0 for 1 min at RT, rinsed again with TBS, and examined under a JEM 1400 Jeol electron microscope equiped with an Orius-Gatan digitalized camera (Gatan France, 78113 Grandchamp).

Statistics

Results were expressed as mean ± SEM. of n observations. Sets of data were compared with an analysis of variance (ANOVA) or a Student's t test. Differences were considered statistically significant when P < 0.05. Symbols used in figures were (*) for P < 0.05, (**) for P < 0.01, (***) for P < 0.001, and ns for no significant difference, respectively. All statistical tests were performed using GraphPad Prism version 4.0 for Windows (Graphpad Software).

Expression and characterization of scFvG2/p17 and scFvE2/p17 molecules in recombinant BV-infected Sf9 cells

The monoclonal antibody secreted by the MH-SVM33C9 hybridoma cell line reacts with the highly conserved and accessible epitope ¹²¹DTGHSSQVSQNY¹³² corresponding to the C-terminus of HIV-1 MAp17 [22–24]. A single chain antibody derived from MH-SVM33 (scFv/p17) and expressed intracellularly showed an inhibitory effect on HIV-1 replication and virus release [6]. We generated two scFv/p17 subclones from the MH-SVM33C9 hybridoma cell line, scFvE2/p17 and scFvG2/p17, of which the full sequence could be communicated upon request. The N-terminal octadecapeptide sequence in scFvE2/p17 read ¹MEASLAAQAAQIQLVQSG¹⁸ , and ¹MGLAAQAAQIQLVQSGPE¹⁸ in scFvG2/p17. Both subclones were also modified at the C-terminus by the addition of a HA-tag, the Influenza A virus hemagglutinin epitope YPYDVPDYA, and when inserted into the baculoviral genome, generated two recombinant BVs, BV-E2/p17 and BV-G2/p17, respectively. The rationale for the N-terminal modification was that the dipeptide Met-Gly at the N-terminus of scFvG2/p17 represented a N-myristoylation signal which would promote the addressing scFvG2/p17 to the same compartment as the N-myristoylated MAp17 protein and Pr55Gag polyprotein. We have shown in previous studies that the Met-Gly signal was functional in insect cells in terms of recognition by N-myristoyl transferase, and that N-myristoylated glycine residue at the N-terminus of Pr55Gag is a prerequisite for plasma membrane addressing of unprocessed Pr55Gag, and budding and egress of membrane-enveloped VLP from Sf9 cells [24, 27–36, 40].

Sf9 cells infected with BV-G2/p17 or BV-E2/p17 were harvested at 24, 48 and 72 h post infection (pi), lysed in hypotonic buffer, and the kinetics of synthesis of scFvG2/p17 and scFvE2/p17 was analyzed by SDS-PAGE and Western blotting using anti-HA tag antibody. The extracellular culture medium was analyzed in parallel. We found that scFvE2/p17 protein (migrating with an apparent molecular mass of 30 kDa) was detectable as early as 24 h, and was maximal at 72 h pi. In contrast, scFvG2/p17 was detected at later times (48 h pi), and in 4- to 5-fold lower amounts, compared to scFvE2/p17 at 48-72 h pi (Figure 1a). However, both scFvE2/p17 and scFvG2/p17 proteins were stable over a period of 72 h, with no major breakdown products detected in the Western blot patterns.

Cell lysates were then clarified by centrifugation, and scFvG2/p17 and scFvE2/p17 were probed in soluble fraction (S) and insoluble pellet (P), respectively. We found that 15-20% of the whole recombinant scFvE2/p17 protein synthesized was recovered as soluble material, as determined by scanning and densitometric analysis of the protein bands on blots, with a maximum at 48 h pi (Figure 1b; leftmost half of the panel). The proportion of soluble scFvG2/p17 was lower (Figure 1b; rightmost half), with only 8-10% of scFvG2/p17 protein recovered in the soluble fraction. The difference in the intracellular levels of the two recombinant proteins could not be explained by a higher level of scFvG2/p17 secretion, compared to scFvE2/p17, as no scFvG2/p17 protein was detected in the culture medium at any time pi. Only scFvE2/p17 protein was detectable in the extracellular medium, as described below.

Cellular distribution of scFvG2/p17 and scFvE2/p17 in recombinant BV-infected Sf9 cells

We next performed cell fractionation to determine in which subcellular compartments the majority of scFvE2/p17 and scFvG2/p17 proteins were localized. Of note, the distinction between cytosol (Cy), membranes and organelles (Mb), nucleus (Nu) and cytoskeleton (Sk) was only operational, and did not preclude probable cross contaminations between different fractions. With this restriction in mind, scFvE2/p17 was found to be associated with the cytosolic fraction and nuclear pellet in similar amounts (ca. 20-25% each), but larger quantities (50-60%) were recovered in the insoluble fraction of cytoskeletal proteins (Figure 1c; leftmost half of the panel). Only small amounts of scFvE2/p17 were detected in the membrane fraction (Figure 1c, Mb lane). The pattern of subcellular localization was different for the scFvG2/p17 protein, which was undetectable in the cytosolic and nuclear fractions, but distributed unequally between membrane and cytoskeletal fractions, two-thirds and one-third of the total, respectively (Figure 1c, rightmost half of the panel). Thus, the N-terminal G² mutation conferred two new properties to scFvG2/p17, compared to its scFvG2/p17 counterpart, (i) a lower level of expression, and (ii) a relocation to and strong association with the membranal fraction. The membrane association of scFvG2/p17 determined by cell fractionation was a priori consistent with the membrane targeting expected for a N-myristoylated protein.

Cellular localization of scFvG2/p17 and scFvE2/p17 was then studied in situ. BV-infected Sf9 cells were harvested at 48 h pi and examined in immunofluorescence (IF) microscopy or analyzed by flow cytometry using anti-HA tag antibody, with or without membrane permeabilization with detergent. Fluorescent signal of scFvE2/p17 was detected in both nonpermeabilized (Figure 2A, i) and Triton X100-permeabilized cells (Figure 2A, ii), whereas scFvG2/p17 fluorescence was only detectable in permeabilized cells (not shown). Flow cytometry of HA tag-positive cells confirmed the accessibility of scFvE2/p17 at the surface of nonpermeabilized cells, and the absence of significant amounts of scFvG2/p17 molecules at the cell surface (Figure 2B). Considering the poor recovery of scFvE2/p17 in the membranal fraction upon cell fractionation (refer to Figure 1c), the results of IF microcopy and flow cytometry suggested that the majority of scFvE2/p17 molecules were addressed to the plasma membrane and highly accessible at the cell surface, but proned to dissociate from the membrane upon cell disruption and fractionation, and to relocate into the soluble fraction. In contrast, we observed a major intracellular retention of most of the scFvG2/p17 molecules. Together with the data of cell fractionation (Figure 1c), this suggested that scFvG2/p17 protein was in majority sequestered in the intracellular membrane network.

When Sf9 cells were coinfected with two recombinant BVs to coexpress scFvE2/p17 and N-myristoylated Pr55Gag, we detected no significant change in the IF pattern of scFvE2/p17, compared to single infected cells expressing scFvE2/p17 alone (Figure 2A, compare panels ii and iii). This suggested that there was no major cellular redistribution of the scFvE2/p17 molecules in the presence of the Gag precursor. Similarly, no cellular redistribution of scFvG2/p17 was observed when coexpressed with Pr55Gag (not shown).

Immunological characterization of extracellular scFvE2/p17 molecules: functionality and specificity

(i) Antigen recognition by scFvE2/p17 in vitro

Since a significant proportion of scFvE2/p17 molecules occurred as soluble intrabody in the cytosolic fraction, we tested lysates of BV-E2/p17-infected Sf9 cells for the antigen binding activity of scFvE2/p17 in ELISA. Two types of immobilized antigens were used: a synthetic peptide p17.1, which reproduced the epitope sequence from HIV-1_LAI MAp17 (¹²¹DTGHSSQVSQNY¹³²), and corresponded to the immunogen used to generate MH-SVM33; a recombinant H₆MA-CA protein of 39 kDa, which carried the same epitope as p17.9 (Table 1), but was embedded in the Gag polyprotein. The binding data from ELISA clearly showed that recombinant scFvE2/p17 reacted with its specific epitope in both configurations (Figure 3a, b), and in a dose-dependent manner (Figure 3c). Competition ELISA indicated that scFvE2/p17 bound to its epitope with a higher affinity when it was used as free p17.1 peptide, compared to the H₆MA-CA polyprotein used at equivalent epitope molarities (Figure 4a), suggesting that scFvE2/p17 preferentially bound to cleaved matrix protein MAp17, compared to non processed Gag precursor.

Competitor	Epitope	Binding competition vs p17.1 (%)	Origin or HIV-1 isolate
Unrelated antigen (negative control)	CD147 (M6)	5.2 ± 2.9	-
17.1 (positive control)	DTGHSSQVSQNY	89.5 ± 1.6	LAI (b)
17.3	DTGHSSQISQNY	74.7 ± 10.2	1M-1005 (c)
17.7	DTGHSSQASQNY	44.0 ± 18.0	g22s2 (d)
17.8	DTGHSKQVSQNY	81.5 ± 9.5	4 (e)
17.9	DTGNNSQVSQNY	68.8 ± 2.2	pNL4.3 (f)
Inverted p17.1	YNQSVQSSHGTD	3.8 ± 1.5	-

^(a) Competition for scFvE2/p17 binding between the dodecapeptide p17.1 (corresponding to residues 121-132 in the MAp17 from HIV-1_LAI, used as immunogen) and other epitope variants was determined by ELISA, as examplified in Figure 4c. Data presented are m ± SEM (n = 3). Residues differing from the LAI prototype sequence are in bold.

^(b) [63].

^(c) [64].

^(d) [43].

^(e) [65].

^(f) [66].

Functionality of scFvE2/p17 and scFvG2/p17 as intrabodies

Biophysical status of extracellular scFvE2/p17 protein: soluble versus particulate

As mentioned above, a significant amount of scFvE2/p17 protein was recovered in the extracellular medium of BV-E2/p17-infected Sf9 cells. These extracellular scFvE2/p17 molecules might occur as soluble protein, or as particle-associated material, e.g. released within membrane microvesicles or exosomes, or associated with cell debris. In order to determine the status of extracellular scFvE2/p17 protein, samples from culture mediun were subjected to isopycnic ultracentrifugation analysis in sucrose-D₂O density gradients [35, 40]. Gradient fractions were analyzed by SDS-PAGE and Western blotting, using anti-baculoviral envelope glycoprotein GP64 and anti-HA tag antibodies. We found that the fractions positive for the C-terminal HA tag of scFvE2/p17 coincided with the anti-GP64-reacting fractions, and corresponded to particulate material sedimenting with the apparent density of BV particles of the viral progeny, d = 1.15-1.08 (Figure 5). This suggested that scFvE2/p17 was associated with the BV particles, either coencapsidated with the baculoviral proteins or inserted into the viral envelope. In the latter case, scFvE2/p17 molecules could be exposed at the surface of the BV particles with their active site accessible for epitope binding.

Immuno-EM analysis of baculoviral progeny of BV-E2/p17

In order to confirm the reality of this surface exposure, samples of extracellular medium of BV-G2/p17- and BV-E2/p17-infected Sf9 cells were analyzed by isopycnic ultracentrifugation in sucrose-D₂O gradients, and fractions sedimenting at 1.15-1.08 were deposited on grids, immunogold labeled with anti-HA tag antibody, and observed under the electron microscope. BV-G2/p17 virions were never seen associated with gold grains, and most immunogold labeling was found at distance from BV particles, and corresponded to nonspecific, background labeling (not shown). The absence of anti-HA labeling of BV-G2/p17 virions in immuno-EM was consistent with the absence of detectable scFvG2/p17 protein in the extracellular medium, as mentioned above. Under the same experimental conditions however, anti-HA tag immunogold labeling was found to be associated with virions of BV-E2/p17 (Figure 6). The immuno-EM analysis therefore confirmed that scFvE2/p17 molecules were truely incorporated into the baculoviral envelope. Such incorporation of foreign proteins into the baculoviral envelope has already been described with human membrane glycoprotein CAR [41].

Immunological functionality and topology of scFvE2/p17 displayed on the baculoviral envelope

The observation that BV-displayed scFvE2/p17 was accessible to anti-HA tag antibodies in immuno-EM analysis strongly suggested that the C-terminal HA tag was oriented outwards. This orientation was already suggested by IF microscopy of intact cells expressing scFvE2/p17 (refer to Figure 2A, i). The next experiments were designed to assess the topological orientation of the scFvE2/p17 molecule in the baculoviral envelope, and, more importantly, to determine the degree of accessibility of its antigen-binding regions. BV-E2/p17 virions were immobilized on ELISA plate and probed with anti-HA tag antibody. The positive reaction indicated that the carboxyterminal region of scFvE2/p17 was exposed at the surface of the virions (Figure 7a), and that the insertion of scFvE2/p17 in the baculoviral envelope mimicked the orientation of class II membrane glycoproteins, with the aminoterminal region anchored in the baculoviral envelope, as depicted in Figure 7b.

The antigen binding activity of BV-displayed scFvE2/p17 was then determined by ELISA, using immobilized H₆MA-CA protein or synthetic peptide p17.1. In this assay, BV-E2/p17 virions displaying scFvE2/p17 were used as the equivalent of primary antibodies, and bound virions were detected using monoclonal antibody directed towards the baculoviral glycoprotein GP64. Both H₆MA-CA protein and p17.1 peptide were recognized by BV-E2/p17 virions (Figure 7c and 7d), which indicated that the BV-displayed scFvE2/p17 molecules retained their antigen-binding capacity.

BV-display of HA-tagged anti-CD147 scFv-M6-1B9

ScFv expressed in recombinant BV-infected insect cells are not naturally or spontaneously addressed to the baculoviral envelope, and different methods have been developed to obtain baculoviral envelope insertion of various scFv molecules, including fusion to GP64 or to VSV-G stem [16–21]. Although scFvE2/p17 did not contain any consensus signal peptide for membrane addressing, we postulated that in Sf9 cells, the N-terminal octadecapeptide sequence ¹MEASLAAQAAQIQLVQSG¹⁸ (abbreviated N18E2), was responsible for the intracellular trafficking and targeting of scFvE2/p17 to the site of BV budding, where it became inserted into the baculoviral envelope. To test this hypothesis, N18E2 was fused to the N-terminus of a non-related scFv, the HA-tagged scFv-M6-1B9, which recognizes the membrane glycoprotein CD147 [38, 39]. When expressed in HeLa or 293 cells, scFv-M6-1B9 occurred as an active anti-CD147 intrabody, provoking the intracellular retention of CD147 and the blockage of its surface expression [38, 39]. Our resulting chimeric scFv construct, abbreviated scFv-N18E2/M6, was expressed in Sf9 cells using a recombinant BV vector (BV-N18E2/M6). ScFv-N18E2/M6 molecules were detected in significant levels at the surface of nonpermeabilized cells harvested at 48 h pi, as for scFvE2/p17 (refer to Figure 2B). The baculoviral progeny from BV-N18E2/M6-infected cells was then isolated by ultracentrifugation and analyzed by ELISA (not shown) and immuno-EM (Figure 8) using anti-HA tag antibody. Both methods showed the accessibility of the HA tag at the surface of BV particles. In the electron micrograph shown in Figure 8d, two baculovirus particles were seen as a V-shape dimer pointing to one gold grain. This might represent the cross-linking of two particles via the two Fab domains of a single antibody molecule, although it could not be excluded that one antibody would bind to one particle, while the other would be in close proximity.

We also tested the antigen-binding capability of BV-displayed chimeric scFv-N18E2/M6, using the recombinant CD147-BCCP fusion protein as immobilized antigen in ELISA, as previously described [39]. The data showed that the chimeric scFv-N18E2/M6 molecules present at the surface of BV particles conserved their antigen-binding function and their specificity towards CD147 (Figure 9). These results suggested that the fusion of the N18E2 peptide to scFv-M6-1B9 was able to confer to the chimeric scFv-N18E2/M6 protein the capacity to be addressed to the BV budding sites at the plasma membrane and to be incorporated into the baculoviral envelope. They also suggested that N18E2 could function as a universal N-terminal signal peptide for BV-display of functional scFv of biological interest. A scheme of the general strategy of stepwise construction of BV particles displaying scFv molecules of interest equiped with our signal octadecapeptide N18E2 is presented in Figure 10.