Biological and Pharmaceutical Bulletin
Online ISSN : 1347-5215
Print ISSN : 0918-6158
ISSN-L : 0918-6158
Regular Articles
Protection of Baculovirus Vectors Expressing Complement Regulatory Proteins against Serum Complement Attack
Yusuke KawaiChiaki KawabataMiako SakaguchiTakahiko Tamura
著者情報
ジャーナル フリー HTML

2018 年 41 巻 10 号 p. 1600-1605

詳細
Abstract

Baculovirus vectors (BVs) enable safe and efficient gene delivery to mammalian cells and are useful in a wide range of applications, including gene therapy and in vivo analysis of gene functions. We previously developed BVs expressing malaria sporozoite surface proteins for targeting liver cells or hepatocytes. However, BVs are known to be very vulnerable to complement attack and efforts to overcome their inactivation based on complement are important. In this study, BVs expressing complement regulatory proteins (CRPs) on the surfaces of virions were developed to inhibit complement reactions. Decay accelerating factor (DAF; CD55)-type BVs exhibited significantly higher complement resistance than control BVs without any CRPs in HepG2 cells transduction, although the transduction efficacy of DAF-type BV was low. In contrast, CD46-DAF-CD59 fusion type BVs showed significantly higher transduction efficacy and complement resistance than both control and DAF-type BVs. DAF-type and CD46-DAF-CD59 type BVs repressed formation of the membrane attack complex, a terminal product of complement reaction cascades, induced by BVs. These results suggest that the CD46-DAF-CD59 fusion construct confers complement protection ability superior to that of the DAF construct in gene delivery under complement active serum.

Baculovirus vectors (BVs) are viral vectors useful for gene transduction in mammalian cells.1) As baculovirus is an enveloped insect virus with high species specificity, it is non-pathogenic to humans. However, a major limitation to the in vivo use of BVs is the inactivation of BVs by serum complement. Complement is an innate immune response against micropathogens.2) The complement pathway is comprised of multiple reaction cascades, finally resulting in membrane attack complex (MAC) formation to make pores in the membranes of target cells and destroying them. Complement-treated BVs show severely reduced transduction efficacy.3) The mechanism of inactivation of BVs by complement remains unclear, but a classical and alternative pathway is thought to be involved in the inactivation.3,4) Complement activity is tightly controlled by numerous complement regulatory proteins (CRPs) expressed by host cells. These proteins inhibit complement reaction cascades, preventing unregulated host damage. Decay accelerating factor (DAF; CD55), CD46 (MCP; membrane cofactor protein), and CD59 are expressed on the host cell surfaces and prevent complement reactions in host cells.2) In some viral pathogens, these host CRPs are recruited into viral particles to prevent complement attack on the virus by the host.5) Previous studies demonstrated that BVs, either expressing CRPs such as DAF fused with viral membrane anchor proteins or administered soluble CR1 showed significant resistance to complement attack.4,68)

Our previous study demonstrated that malaria sporozoite surface proteins circumsporozoite protein (CSP), or thrombospondin-related adhesion protein (TRAP) in BV envelopes significantly increased transduction efficacy into hepatic cells.9) We examined whether these CSP/TRAP BVs are vulnerable or resistant to complement attack. Additionally, we explored new BVs that are more resistant against complement attack. We constructed a CD46-DAF-CD59 fusion type BV, evaluated its complement resistance, and compared its resistance to that of the DAF type, which was previously reported in transduction experiments. These results demonstrated that CD46-DAF-CD59 type BV was superior to both control BV without CRPs and DAF-type BV.

MATERIALS AND METHODS

Cell Culture

Sf9 cells were maintained in Sf-900 II SFM cell culture medium (Invitrogen Corporation, Carlsbad, CA, U.S.A.). HepG2 cells, from human hepatocarcinoma cell line, were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), non-essential amino acids, and antibiotics.

Baculovirus Production

BVs were generated using the Bac-to-Bac® Baculovirus Expression System (Invitrogen Corporation) as previously described.9) Construction of the pFast-cytomegalo virus-enhanced green fluorescent protein (CMV-EGFP)-GL3-pol-DsRed control plasmid was performed as follows. The sequence for polyhedrin promoter and DsRed protein was inserted into pFast-CMV-EGFP-GL3 plasmid,9) using NotI and XbaI sites, to construct the pFast-CMV-EGFP-GL3-pol-DsRed plasmid, which serves as a shuttle vector for control-type BV expressing no CPRs. The expression cassette CMV-EGFP-GL3 drives EGFP-luciferase fusion protein expression in mammalian cells. The expression cassette pol-DsRed induces DsRed expression in Sf9 cells, used for virus titration. The CRP protein expression in this study was under p10 promoter. Sequences for the p10 promoter, gp64 signal peptide, myc tag, and ectodomain of human DAF (amino acids (aa) 35–354), fused to the transmembrane (TM) domain of the vesicular stomatitis virus glycoprotein (VSV-G) (aa 421–510) as a membrane anchor, were inserted into pFast-CMV-EGFP-GL3-pol-DsRed using BssHII and NotI sites. This generated the shuttle vector for DAF-type BV, pFast-CMV-EGFP-GL3-pol-DsRed-DAF. The CD46-DAF plasmid was constructed by inserting the sequences for extracellular domain of human CD46 splicing variant C (aa 35–308)10) upstream of DAF, using EcoRI site, resulting in pFast-CMV-EGFP-GL3-pol-DsRed-CD46-DAF. The CD46-DAF-CD59 plasmid was constructed by inserting CD59 fragment (aa 26–102) downstream of DAF of CD46-DAF, using XmaI site, resulting in the shuttle vector for CD46-DAF-CD59 type BV, pFast-CMV-EGFP-GL3-pol-DsRed-CD46-DAF-CD59. The sequences of plasmids were confirmed using standard DNA sequencing techniques. These constructed pFast shuttle vectors were transformed into DH10Bac™ cells (Invitrogen Corporation) to obtain the corresponding bacmids. The bacmids were transfected to Sf9 cells, and the culture medium containing recombinant baculovirus was collected. Large-scale preparations of virus (from approximately 200 mL of culture medium) were purified by ultracentrifugation (Beckman Coulter, Inc., Brea, CA, U.S.A.). Virus titration was performed as previously described.9)

Western Blot Analysis

After purification, the virion samples (5×106 plaque forming units (pfu)) were separated on 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel, and transferred to an Immobilon®-FL membrane (Merck Millipore, Darmstadt, Germany). To detect myc-tagged CRPs, samples were probed with mouse monoclonal anti-myc antibody (My3; MBL, Nagoya, Japan) and in-house manufactured rabbit anti-VP39 antibody, followed by IRDye® 800CW Goat anti-Mouse and IRDye® 680LT Goat anti-Rabbit secondary antibodies (LI-COR Biotechnology, Lincoln, NE, U.S.A.). The membrane was analyzed using the Odyssey® infrared imaging system (LI-COR Biotechnology). To detect human DAF or CD46 proteins, mouse anti-DAF antibody (BRIC-216; Millipore, Bedford, MA, U.S.A.) or rabbit anti-CD46 antibody (H-294, Santa Cruz Biotechnology. Inc., Santa Cruz, CA, U.S.A.) was used. In case of BRIC-216, non-reduced samples were used.

Flow Cytometric Analysis

To detect cell surface expression of CRPs, Sf9 cells (5×105 cells per well) were seeded in 12-well plates and infected (MOI; multiplicity of infection 10) with recombinant BVs. After three days, cells were collected and incubated in anti-CD46 (MEM258), DAF (JS11), and CD59 (p282) antibody conjugated with FITC (Biolegend). Cells were analyzed using a BD FACSVerse™ flow cytometer and data analysis was performed using the BD FACSuite™ software (BD Biosciences, San Jose, CA, U.S.A.).

Transduction

Human complement serum (S1764) was purchased from Sigma (St. Louis, MO, U.S.A.). The complement activity of classical and alternative pathway was measured using Wieslab complement system screen (Euro Diagnostica, Malmo, Sweden), which confirmed that its activity was within average values of normal population. Virus suspension (1×107 pfu/10 µL) was mixed with 5 µL of human complement serum, either intact (not heat-inactivated; complement active) or heat-inactivated (HI) at 56°C for 30 min to inactivate both classical and alternative complement pathways, and incubated for one hour at 37°C. Transduction procedures were as previously described.9) Briefly, HepG2 cells (1×104 cells) were seeded in collagen-coated 96-well plates, and serum-treated virus suspension (1×106 pfu) was added to the cells. After one-hour incubation, virus suspension was replaced with fresh culture medium. Following 24 h of culture time, GFP expression was examined microscopically (Keyence Corporation, Osaka, Japan). For the luciferase assay, the culture medium was removed and Cell Culture Lysis Buffer (Promega Corporation, Madison, WI, U.S.A.) was added. The extracts were subjected to luciferase activity measurement using a GloMax® system (Promega Corporation). The complement resistance rate was calculated from luciferase values in the /intact treatments.

MAC Enzyme-Linked Immunosorbent Assay (ELISA)

To detect MAC formation by BVs, they were incubated with HI or intact human serum (final 33%) for 20 min at 37°C. The solutions containing approximately 2×107 pfu BVs were coated on 96-well plates for 1 h at room temperature. After blocking with 1% BSA, plates were incubated with mouse anti-human C5b-C9 antibody (aE11; Santa Cruz Biotechnology) and with anti-mouse immunoglobulin G (IgG) conjugated-HRP (Bio-Rad, Hercules, CA, U.S.A.) as secondary antibody. Enzymatic activity was detected using ABTS solution and OD414 nm was measured using a plate reader (Thermo Scientific, Carlsbad, CA, U.S.A.). To detect BVs, ELISA experiments were performed. BVs (approximately 6×106 pfu), without incubation with serum, were coated on 96-well plates. After blocking, plates were incubated with mouse anti-gp64 (AcV1; eBioscience, San Diego, CA, U.S.A.), followed by the same procedure as above.

Statistical Analysis

Results are expressed as mean±standard error (S.E.) Statistical analysis was performed using the Mann–Whitney U test for the comparison of two experimental groups, and data computations were performed using GraphPad Prism version 5 (GraphPad Software, San Diego, CA, U.S.A.). A p-value less than 0.05 was considered significant.

RESULTS

Serum Complement Rapidly Impaired Transduction Activity of BVs

We examined whether BVs expressing CSP or TRAP on the virion surfaces were vulnerable or resistant to complement attack. BVs were reacted with human intact serum (complement active) or HI serum (complement inactive) and tested for HepG2 transduction. In Fig. 1A, both BVs were found to be significantly inactivated by complement exposure in HepG2 transduction. Complement pathways comprise of mainly classical and alternative pathways. To investigate the importance of these pathways, serum was heated at 50°C for 20 min and then used for transduction experiments. As this treatment inactivates factor B, a critical component in the alternative pathway, the cascade of alternative pathway was impaired, while the classical pathway was not.11) As shown in Fig. 1B, intact serum treatment significantly decreased BV transduction in HepG2 cells. Serum heated at 50°C for 20 min did not inactivate BV transduction, suggesting that an alternative pathway is significantly involved in the inactivation of BV transduction in HepG2 cells.

Fig. 1. BV Inactivation by Intact Human Serum, with Complement Activity

(A) CSP/TRAP BVs were treated with human serum (intact or HI) and their transduction efficacy in HepG2 cells was examined by measuring luciferase activities and calculating complement resistance rates (intact/HI). (B) TRAP BV was treated with serum, either intact or HI (at 50°C for 20 min) and their transduction efficacy in HepG2 cells was examined, as above. The values were represented compared with HI. Asterisks indicate a significant difference between HI and intact serum conditions (p<0.05). Experiments were performed twice and representative data are shown.

Expression of CRPs on BV Virions

To avoid complement attack, attempts to express DAF in BV virions have been reported,6,7) but the protection was not sufficient. Previous studies demonstrated that fusion of several CRPs such as CD46-DAF or DAF-CD59 resulted in more efficient protection against complement.12,13) Therefore, to develop more complement-resistant BVs, CD46-DAF-CD59 fusion type BV expressing a fusion of three CRPs (CD46, DAF, and CD59) was constructed, along with DAF-expressing BV (Fig. 2). CRPs were expressed under the control of the p10 promoter. Sf9 cells were infected with these BVs and CRP expression on the infected cell surface was confirmed by flow cytometric analysis (Fig. 3A). The expression of myc-tagged CRPs on purified budded virions was confirmed using an anti-myc tag antibody in Western blot analysis (Fig. 3B). The CRPs expressed in BV virions were also detected using anti-DAF antibody and anti-CD46 antibody by Western blot analysis (data not shown).

Fig. 2. Construction of BVs Expressing Complement Regulatory Proteins

Schematic representation of the recombinant BV constructs employed in this study. EGFP-luciferase fusion proteins, expressed under CMV promoter, indicate the transduction efficacy in mammalian cells. The CRPs expressed under p10 promoter are fused to myc tag and VSV-G TM domain and anchored on BV virions.

Fig. 3. Expression of CRPs in the Constructed BVs

(A) Flow cytometric analysis of BV-infected Sf9 cells. The infected Sf9 cells were stained with the indicated antibody and subjected to flow cytometric analysis to detect the indicated CRPs on the cell surfaces. (B) Western blot analysis of virions. The samples were run on SDS-PAGE gel, transferred to the membrane, and probed with anti-myc antibody; anti-VP39 antibody as the loading control. Lane 1: control BV, lane 2: DAF type BV, lane 3: CD46-DAF-CD59 type BV. Arrows indicate the position of surface proteins in the blots. (Color figure can be accessed in the online version.)

BVs Expressing CRPs Showed Significant Resistance to Complement Attack

To investigate the resistance to complement attack and retention of transduction activity in mammalian cells, the constructed BVs were exposed to intact human serum and HI serum before transduction into HepG2 cells. As shown in Fig. 4A, the transduction efficacy of control BV was highly dependent on the serum complement. DAF-expressing BV showed a significant increase in the complement resistance rate (intact/HI) after complement exposure compared to control BV, although transduction efficacy was low (Fig. 4A). CD46-DAF-CD59 type BV, a newly constructed vector in this study, also showed significant resistance to complement attack compared to control BV (Fig. 4A). Moreover, the complement resistance rate and transduction efficacy of the CD46-DAF-CD59 type was significantly higher than those in the DAF type (Fig. 4A). The transduction efficacy of non-serum treatment was not significantly different from that of HI treatment in all BV types (Fig. 4A). The CD46-DAF-CD59 type BV showed complement resistance at higher concentrations of human serum, although its rate was slightly decreased (Fig. 4B). These results suggest that the CD46-DAF-CD59 type is superior to the DAF type as a transgene vector under complement-active circumstances.

Fig. 4. CRPs Increased Transduction Efficacy of BVs in HepG2 Cells by Protecting Human Complement Attack

(A) The constructed BVs, after human serum treatment, were transduced into HepG2 cells. The reporter gene expressions were analyzed. The luciferase values expressed in transduced HepG2 cells are shown (upper). The complement resistance rate (% of intact/HI) is shown in the lower panel. The bars represent the mean±S.E. Asterisks indicate a significant difference, as compared to the control virus or DAF type virus (p<0.05). Experiments were performed twice and representative data are shown. (B) CD46-DAF-CD59 type BV was exposed to higher concentration of human serum and transduced to HepG2 cells. The complement resistance rate (% of intact/HI) is shown. (C) Detection of MAC on BV virions using ELISA. BV solutions were coated on plates and MAC proteins were detected. As control for viral load, gp64 proteins in the virions were examined and no significant difference was observed between control, DAF type and CD46-DAF-CD59 type BV samples. Asterisks indicate a significant difference, as compared to the control virus type virus (p<0.05). Experiments were performed twice and representative data are shown.

To examine the inhibitory activities of the constructed BVs against complement reactions, BVs were reacted with human intact serum and MAC, the terminal product of complement reactions, and detected by ELISA. Without BVs, spontaneous activation of complement was observed, while addition of control BV potentiated the complement reactions and significantly increased MAC formation compared to that without BVs (Fig. 4C). The DAF type or CD46-DAF-CD59 type BV induced significantly less MAC formation compared to the control BV, suggesting that these CRPs inhibited the complement reactions with BVs (Fig. 4C).

DISCUSSION

BVs are safe and promising gene delivery vectors that may be applicable in gene therapy and in vivo analysis of gene functions. However, BVs are easily inactivated by the complement system, one of the most common innate immune systems in mammals, which may limit the application of BVs. In the present study, CRPs were expressed on the envelopes of BV virions. Their expression was confirmed by Western blot analysis (Fig. 3B). Moreover, electron microscopic analysis supported the expression of DAF on virions.14) Therefore, these CRPs may inhibit complement reactions on virion surfaces, resulting in retention of the transduction ability in BVs. The results of this study may guide the development of next-generation BVs, including those for in vivo use and in vitro transduction.

Previous reports demonstrated that complement-treated BVs displayed a disrupted virion morphology, such as fall-off or pores on the virion envelopes.6,14,15) MAC formation and critical damage have also been observed in the membrane pores of target cells. However, BVs are viruses, not cells, and MAC formation may not be an absolute necessity for inactivation of BVs by complement reactions. Other reports demonstrated that depletion of the C8 and C9 components of MAC from the serum may not be absolutely sufficient to prevent the complement attack,3) suggesting that BV inactivation can be mediated by MAC formation and other unidentified mechanisms resulting from complement reactions before MAC formation. Figure 4 shows that CD46-DAF-CD59 type BV and DAF-type BV, inhibited MAC formation to the same extent, although the complement resistance rate of CD46-DAF-CD59 type BV was significantly higher than that of the DAF-type BV in HepG2 transduction. These results suggest that inhibition of MAC formation may be important, but likely not the only process contributing to the retention of transduction efficiency in CD46-DAF-CD59 type BV, which differs from the DAF-type BV. To more comprehensively determine the inactivation mechanisms of BV transduction by serum complement and effectively develop complement-resistant BVs, analysis of C3b deposits on serum-treated BVs and detailed observation of the morphology of BVs after serum exposure by electron microscopy in parallel with transduction efficacy may be important. It would be interesting to investigate which types of morphological changes and complement reactions are critical. Figure 1B suggests that an alternative pathway played an important role in BV inactivation by serum complement, but studies examining the contribution of classical pathways such as using C1q-depleted serum are needed, as the classical pathway may produce C3b and initially trigger an alternative pathway.2)

DAF type BVs showed lower transduction efficacy in cells other than HepG2 cells. We conducted transduction experiments for DAF-type BVs using PXB cells (primary human hepatocytes) and AsPC cells (human pancreatic cancer cells), in which DAF-type BVs showed lower transduction than control and CD46-DAF-CD59 type BVs, alongside HepG2 cells (data not shown). A previous study observed low transduction efficacy of DAF-type BVs,6) suggesting that the low transduction of DAF-type BVs may be intrinsic. The mechanism by which DAF-type BVs show low transduction efficacy in HepG2 cells and other human cultured cells remains unknown. The binding capacity to heparin, the cellular receptor for BV entry into target cells,16) may be attenuated in DAF-type BV, but not in CD46-DAF-CD59 type BV. The DAF protein in BV virion may affect the functions of gp64 protein, which binds to heparin17) as an essential molecule for transduction, more than CD46-DAF-CD59 protein. Because exogeneous protein expression in BV virion surfaces may affect transduction efficacy, optimization of BV constructs may be important.

The CD46-DAF-CD59 type BV examined in this study can be combined with other types of BVs for complement-active conditions, such as in vivo. The CSP/TRAP proteins can be co-expressed with CD46-DAF-CD59 proteins to equip the BVs with both liver targeting and complement resistance characteristics. Attempts to incorporate CRPs into other viral vectors have been made to protect the inactivation by complement.18) The constructs described in this study can be applied for other viral vectors as well to protect against complement attack.

Acknowledgments

We appreciate Dr. Shigeto Yoshida (Kanazawa University, Japan) for his support in this study. We also appreciate Dr. Ryohei Ogawa (University of Toyama, Japan) for providing AsPC cells. This study was supported by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science to T.T. (15K07926).

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES
 
© 2018 The Pharmaceutical Society of Japan
feedback
Top