Recombinant Proteins

Example 3

Recombinant Protein Purification

FIG. 5 shows the steps of one of the purifications carried out on the chimera. In the case of GRNLY, this process was shown in an earlier paper [Ibáñez, R., University of Zaragoza. 2015]. It can be seen in FIG. 5A that the P. Pastoris supernatant obtained after induction (lane 1) contains rather diluted proteins. After concentrating same with Pellicom, protein bands are not seen in the permeate (lane 3), but proteins that are much more concentrated than in the supernatant are seen in the concentrate (lane 2). After dialysis (lane 4), the band profile remains similar to the concentrate. Furthermore, protein bands are not seen in the buffer in which the dialysis bag (lane 5) was introduced. Upon addition of the nickel resin, the chimera binds to said resin as it has a histidine tag. After adding the resin (lane 6), the intensity of a band corresponding to a protein of about 40 kDa decreases with respect to the concentrate and dialysate. This band may correspond to the chimera. The fact that this band does not altogether disappear may indicate that the nickel resin was saturated. In the washes performed on the resin, particularly in the first wash (lane 7), it can be seen how the residues of other proteins are removed. Finally, after the elution of the nickel column, a major protein with a molecular weight of about 40 kDa corresponding to the molecular weight of the chimera (lane 11) is clearly observed. As shown in FIG. 5B, it was confirmed by means of immunoblot that this band of about 40 kDa corresponds to the chimera (lane 11). It is also confirmed that the resin was saturated because a band appears in the post-resin dialysis phase (lane 6).

FIG. 6 shows different elution fractions and the pooling of all of them with the exception of elution fraction 1. FIG. 6A shows several bands in the different elution fractions and in the total eluate. The band with the highest intensity has a molecular weight corresponding to the chimera. Furthermore, other bands having intermediate molecular weights are observed, which means that the chimera undergoes partial proteolysis. The band with the second highest intensity has a molecular weight of about 10 kDa, which corresponds to 9-kDa GRNLY, as its molecular weight increases since it is bound to a histidine tag. In FIG. 6B, it was confirmed by means of immunoblot that these bands of about 40 and 10 kDa correspond to the chimeric recombinant protein and to recombinant GRNLY, respectively.

Once the chimera is generated, its functionality must be assured, that is, on one hand the scFv still recognizes the CEA antigen, and on the other hand GRNLY is still cytotoxic.

Example 1

a. Materials and Methods

i. Vector Construction

1. Virus-Like Particle

As most broadly neutralizing HPV antibodies are derived from the highly conserved N-terminal region of L2, amino acids 14-122 of HPV16 L2 were used to create HBc VLPs. L2 with flanking linker regions was inserted into the tip of the a-helical spike of an HBc gene copy which was fused to another copy of HBc lacking the L2 insert. This arrangement allows the formation of HBc dimers that contain only a single copy of L2, increasing VLP stability (Peyret et al. 2015). This heterodimer is referred to as HBche-L2. A dicot plant-optimized HPV16 L2 coding sequence was designed based upon the sequence of GenBank Accession No. CAC51368.1 and synthesized in vitro using synthetic oligonucleotides by the method described (Stemmer et al., 1995). The plant-optimized L2 nucleotide sequence encoding residues 1-473 is posted at GenBank Accession No. KC330735. PCR end-tailoring was used to insert Xbal and SpeI sites flanking the L2 aa 14-122 using primers L2-14-Xba-F (SEQ ID NO. 1: CGTCTAGAGTCCGCAACCCAACTTTACAAG) and L2-122-Spe-R (SEQ ID NO. 2: G GGACTAGTTGGGGCACCAGCATC). The SpeI site was fused to a sequence encoding a 6His tag, and the resulting fusion was cloned into a geminiviral replicon vector (Diamos, 2016) to produce pBYe3R2K2Mc-L2(14-122)6H.

The HBche heterodimer VLP system was adapted from Peyret et al (2015). Using the plant optimized HBc gene (Huang et al., 2009), inventors constructed a DNA sequence encoding a dimer comprising HBc aa 1-149, a linker (G2S)5G (SEQ ID NO. 39), HBc aa 1-77, a linker GT(G₄S)₂ (SEQ ID NO. 40), HPV-16 L2 aa 14-122, a linker (GGS)₂GSSGGSGG (SEQ ID NO. 41), and HBc aa 78-176. The dimer sequence was generated using multiple PCR steps including overlap extensions and insertion of BamHI and SpeI restriction sites flanking the L2 aa 14-122, using primers L2-14-Bam-F (SEQ ID NO. 3: CAGGATCCGCAACC CAACTTTACAAGAC) and L2-122-Spe-R (SEQ ID NO. 2). The HBche-L2 coding sequence was inserted into a geminiviral replicon binary vector pBYR2eK2M (FIG. 3), which includes the following elements: CaMV 35S promoter with duplicated enhancer (Huang et al., 2009), 5′ UTR of N. benthamiana psaK2 gene (Diamos et al., 2016), intron-containing 3′ UTR and terminator of tobacco extensin (Rosenthal et al, 2018), CaMV 35S 3′ terminator (Rosenthal et al, 2018), and Rb7 matrix attachment region (Diamos et al., 2016).

2. Recombinant Immune Complex

The recombinant immune complex (RIC) vector was adapted from Kim et al., (2015). The HPV-16 L2 (aa 14-122) segment was inserted into the BamHI and SpeI sites of the gene encoding humanized mAb 6D8 heavy chain, resulting in 6D8 epitope-tagged L2. The heavy chain fusion was inserted into an expression cassette linked to a 6D8 kappa chain expression cassette, all inserted into a geminiviral replicon binary vector (FIG. 3, RIC vector). Both cassettes contain CaMV 35S promoter with duplicated enhancer (Huang et al., 2009), 5′ UTR of N. benthamiana psaK2 gene (Diamos et al., 2016), intron-containing 3′ UTR and terminator of tobacco extensin (Rosenthal et al, 2018), and Rb7 matrix attachment region (Diamos et al., 2016).

ii. Agroinfiltration of Nicotiana benthamiana Leaves

Binary vectors were separately introduced into Agrobacterium tumefaciens EHA105 by electroporation. The resulting strains were verified by restriction digestion or PCR, grown overnight at 30° C., and used to infiltrate leaves of 5- to 6-week-old N. benthamiana maintained at 23-25° C. Briefly, the bacteria were pelleted by centrifugation for 5 minutes at 5,000 g and then resuspended in infiltration buffer (10 mM 2-(N-morpholino)ethanesulfonic acid (MES), pH 5.5 and 10 mM MgSO₄) to OD₆₀₀=0.2, unless otherwise described. The resulting bacterial suspensions were injected by using a syringe without needle into leaves through a small puncture (Huang et al. 2004). Plant tissue was harvested after 5 DPI, or as stated for each experiment. Leaves producing GFP were photographed under UV illumination generated by a B-100AP lamp (UVP, Upland, CA).

iii. Protein Extraction

Total protein extract was obtained by homogenizing agroinfiltrated leaf samples with 1:5 (w:v) ice cold extraction buffer (25 mM sodium phosphate, pH 7.4, 100 mM NaCl, 1 mM EDTA, 0.1% Triton X-100, 10 mg/mL sodium ascorbate, 0.3 mg/mL PMSF) using a Bullet Blender machine (Next Advance, Averill Park, NY) following the manufacturer's instruction. To enhance solubility, homogenized tissue was rotated at room temperature or 4° C. for 30 minutes. The crude plant extract was clarified by centrifugation at 13,000 g for 10 minutes at 4° C. Necrotic leaf tissue has reduced water weight, which can lead to inaccurate measurements based on leaf mass. Therefore, extracts were normalized based on total protein content by Bradford protein assay kit (Bio-Rad) with bovine serum albumin as standard.

iv. SDS-PAGE and Western Blot

Clarified plant protein extract was mixed with sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.02% bromophenol blue) and separated on 4-15% polyacrylamide gels (Bio-Rad). For reducing conditions, 0.5M DTT was added, and the samples were boiled for 10 minutes prior to loading. Polyacrylamide gels were either transferred to a PVDF membrane or stained with Coomassie stain (Bio-Rad) following the manufacturer's instructions. For L2 detection, the protein transferred membranes were blocked with 5% dry milk in PBST (PBS with 0.05% tween-20) overnight at 4° C. and probed with polyclonal rabbit anti-L2 diluted 1:5000 in 1% PBSTM, followed by goat anti-rabbit horseradish peroxidase conjugate (Sigma). Bound antibody was detected with ECL reagent (Amersham).

v. Immunization of Mice and Sample Collection

All animals were handled in accordance to the Animal Welfare Act and Arizona State University IACUC. Female BALB/C mice, 6-8 weeks old, were immunized subcutaneously with purified plant-expressed L2 (14-122), HBche-L2 VLP, L2 RIC, or PBS mixed 1:1 with Imject® Alum (Thermo Scientific, Rockford, IL). In all treatment groups, the total weight of antigen was set to deliver an equivalent 5 μg of L2. Doses were given on days 0, 21, and 42. Serum collection was done as described (Santi et al. 2008) by submandibular bleed on days 0, 21, 42, and 63.

vi. Antibody Measurements

Mouse antibody titers were measured by ELISA. Bacterially-expressed L2 (amino acids 11-128) was bound to 96-well high-binding polystyrene plates (Corning), and the plates were blocked with 5% nonfat dry milk in PBST. After washing the wells with PBST (PBS with 0.05% Tween 20), the diluted mouse sera were added and incubated. Mouse antibodies were detected by incubation with polyclonal goat anti-mouse IgG-horseradish peroxidase conjugate (Sigma). The plate was developed with TMB substrate (Pierce) and the absorbance was read at 450 nm. Endpoint titers were taken as the reciprocal of the lowest dilution which produced an OD450 reading twice the background. IgG1 and IgG2a antibodies were measured with goat-anti mouse IgG1 or IgG2a horseradish peroxidase conjugate.

vii. Electron Microscopy

Purified samples of HBche or HBche-L2 were initially incubated on 75/300 mesh grids coated with formvar. Following incubation, samples were briefly washed twice with deionized water then negatively stained with 2% aqueous uranyl acetate. Transmission electron microscopy was performed with a Phillips CM-12 microscope, and images were acquired with a Gatan model 791 CCD camera.

viii. Statistical Analysis

The significance of vaccine treatments and virus neutralization was measured by non-parametric Mann-Whitney test using GraphPad prism software. Two stars (**) indicates p values <0.05. Three stars (***) indicates p values <0.001.

b. Design and Expression of HBc VLPs and RIC Displaying HPV16 L2

BeYDV plant expression vectors (FIG. 3) expressing either the target VLP HBche-L2, or L2 and HBche alone as controls, were agroinfiltrated into the leaves of N. benthamiana and analyzed for VLP production. After 4-5 days post infiltration (DPI), leaves displayed only minor signs of tissue necrosis, indicating that the VLP was well-tolerated by the plants (FIG. 4A). Leaf extracts analyzed by reducing SDS-PAGE showed an abundant band near the predicted size of 51 kDa for HBche-L2, just above the large subunit of rubisco (RbcL). HBche was detected around the predicted size of 38 kDa (FIG. 4B). Western blot probed with anti-L2 polyclonal serum detected a band for HBche-L2 at ˜51 kDa (FIG. 4B). These results indicate that this plant system is capable of producing high levels of L2-containing HBc VLP.

To express L2-containing MC, amino acids 14-122 of HPV16 L2 were fused with linker to the C-terminus of the 6D8 antibody heavy chain and tagged with the 6D8 epitope (Kim et al. 2015). A BeYDV vector (FIG. 3) expressing both the L2-fused 6D8 heavy chain and the light chain was agroinfiltrated into leaves of N. benthamiana and analyzed for RIC production. To create more homogenous human-type glycosylation, which has been shown to improve antibody Fc receptor binding in vivo, transgenic plants silenced for xylosyltransferase and fucosyltransferase were employed (Castilho and Steinkellner 2012). By western blot, high molecular weight bands >150 kDa suggestive of RIC formation were observed (FIG. 4C). Expression of soluble L2 RIC was lower than HBche-L2 due to relatively poor solubility of the RIC (FIG. 4C).

After rigorous genetic optimization, the N. benthamiana system is capable of producing very high levels of recombinant protein, up to 30-50% of the total soluble plant protein, in 4-5 days (Diamos et al. 2016). Using this system, we produced and purified milligram quantities of fully assembled and potently immunogenic HBc VLPs displaying HPV L2 through a simple one-step purification process (FIGS. 4A-4C and 6).

c. Purification and Characterization of HBche-L2 and L2 RIC

To assess the assembly of HBc-L2 VLP, clarified plant extracts containing either HBche-L2 or HBche were analyzed by sucrose gradient sedimentation. HBche-L2 sedimented largely with HBche, which is known to form VLP, though a small increase in density was observed with HBche-L2, perhaps due to the incorporation of L2 into the virus particle (FIG. 5A). To demonstrate particle formation, sucrose fractions were examined by electron microscopy. Both HBche and HBche-L2 formed ˜30 nm particles, although the appearance of HBche-L2 VLP suggested slightly larger, fuller particles (FIGS. 5C and 5D). As most plant proteins do not sediment with VLP, pooling peak sucrose fractions resulted in >95% pure HBche-L2 (FIG. 5B), yielding sufficient antigen (>3 mg) for vaccination from a single plant leaf.

L2 RIC was purified from plant tissue by protein G affinity chromatography. By SDS-PAGE, an appropriately sized band was visible >150 kDa that was highly pure (FIG. 5B). Western blot confirmed the presence of L2 in this band, indicating proper RIC formation (FIG. 5B). L2 RIC bound to human complement C1q receptor with substantially higher affinity compared to free human IgG standard, suggesting proper immune complex formation (FIG. 5E).

d. Mouse Immunization with HBche-L2 and L2 RIC

Groups of Balb/c mice (n=8) were immunized, using alum as adjuvant, with three doses each of 5 μg L2 delivered as either L2 alone, HBche-L2 VLP, L2 RIC, or a combination of half VLP and half RIC. VLP and RIC, alone or combined, greatly enhanced antibody titers compared to L2 alone by more than an order of magnitude at all time points tested (FIG. 6). After one or two doses, the combined VLP/RIC treatment group outperformed both the VLP or RIC groups, reaching mean endpoint titers of >200,000, which represent a 700-fold increase over immunization with L2 alone (FIG. 6). After the third dose, both the VLP and combined VLP/RIC groups reached endpoint titers >1,300,000, a 2-fold increase over the RIC alone group. To determine the antibody subtypes produced by each treatment group, sera were assayed for L2-binding IgG1 and IgG2a. All four groups produced predominately IgG1 (FIG. 7, note dilutions). However, RIC and especially VLP-containing groups had an elevated ratio of IgG2a:IgG1 (>3-fold) compared to L2 alone (FIG. 7).

In vitro neutralization of HPV16 pseudovirions showed that the VLP and RIC groups greatly enhanced neutralization compared to L2 alone (FIG. 5, p<0.001). Additionally, VLP and RIC combined further enhanced neutralization activity ($5-fold, p<0.05) compared to either antigen alone, supporting the strong synergistic effect of delivering L2 by both platforms simultaneously.

In this study, by displaying amino acids 11-128 on the surface of plant-produced HBc VLPs, L2 antibody titers as high as those seen with L1 vaccines were generated (FIG. 6). Mice immunized with L2 alone had highly variable antibody titers, with titers spanning two orders of magnitude. By contrast, the other groups had much more homogenous antibody responses, especially the VLP-containing groups, which had no animals below an endpoint titer of 1:1,000,000 (FIG. 6). These results underscore the potential of HBc VLP and RIC to provide consistently potent immune responses against L2. Moreover, significant synergy of VLP and RIC systems was observed when the systems were delivered together, after one or two doses (FIG. 6). Since equivalent amounts of L2 were delivered with each dose, the enhanced antibody titer did not result from higher L2 doses. Rather, these data suggest that higher L2-specific antibody production may be due to augmented stimulation of L2-specific B cells by T-helper cells that were primed by RIC-induced antigen presenting cells. Although treatment with VLP and RIC alone reached similar endpoint titers as the combined VLP/RIC group after 3 doses, virus neutralization was substantially higher (>5-fold) in the combined group (FIG. 8). Together, these data indicate unique synergy exists when VLP and RIC are delivered together. Inventors have observed similarly significant synergistic enhancement of immunogenicity for a variety of other antigens.

Mice immunized with L2 alone had highly variable antibody titers, with titers spanning two orders of magnitude. By contrast, the VLP and VLP/RIC groups had much more homogenous antibody responses, with no animals below an endpoint titer of 1:1,000,000 (FIG. 6). These results underscore the potential of HBc VLP and RIC to provide consistently potent immune responses against L2.

Fc gamma receptors are present on immune cells and strongly impact antibody effector functions such as antibody-dependent cell-mediated cytotoxicity and complement-dependent cytotoxicity (Jefferis 2009). In mice, these interactions are controlled in part by IgG subtypes. IgG1 is associated with a Th2 response and has limited effector functions. By contrast, IgG2a is associated with a Th1 response and more strongly binds complement components (Neuberger and Raj ewsky 1981) and Fc receptors (Radaev 2002), enhancing effector functions and opsonophagocytosis by macrophages (Takai et al. 1994). Immunization with L2 alone was found to produce low levels of IgG2a, however immunization with RIC and VLP produced significant increases in IgG2a titers. VLP-containing groups in particular showed a 3-fold increase in the ratio of IgG2a to IgG1 antibodies (FIG. 7). Importantly, production of IgG2a is associated with successful clearance of a plethora of viral pathogens (Coutelier et al. 1988; Gerhard et al. 1997; Wilson et al. 2000; Markine-Goriaynoff and Coutelier 2002).

The glycosylation state of the Fc receptor also plays an important role in antibody function. Advances in glycoengineering have led to the development of transgenic plants with silenced fucosyl- and xylosyl-transferase genes capable of producing recombinant proteins with authentic human N-glycosylation (Strasser et al. 2008). Antibodies produced in this manner have more homogenous glycoforms, resulting in improved interaction with Fc gamma and complement receptors compared to the otherwise identical antibodies produced in mammalian cell culture systems (Zeitlin et al. 2011; Hiatt et al. 2014; Strasser et al. 2014; Marusic et al. 2017). As the known mechanisms by which RIC vaccines increase immunogenicity of an antigen depend in part on Fc and complement receptor binding, HPV L2 RIC were produced in transgenic plants with silenced fucosyl- and xylosyl-transferase. Consistent with these data, we found that L2 RIC strongly enhanced the immunogenicity of L2 (FIG. 6). However, yield suffered from insolubility of the RIC (FIG. 4C). We found that the 11-128 segment of L2 expresses very poorly on its own in plants and may be a contributing factor to poor L2 RIC yield. Importantly, we have produced very high yields of RIC with different antigen fusions. Thus, in some aspects, antibody fusion with a shorter segment of L2 could substantially improve the yield of L2 RIC.

e. Neutralization of HPV Pseudovirions

Neutralization of papilloma pseudoviruses (HPV 16, 18, and 58) with sera from mice immunized IP with HBc-L2 VLP and L2(11-128) showed neutralization of HPV 16 at titers of 400-1600 and 200-800, respectively (Table 1). More mice IP-immunized with HBc-L2 VLP had antisera that cross-neutralized HPV 18 and HPV 58 pseudoviruses, compared with mice immunized with L2(11-128). Anti-HBc-L2 VLP sera neutralized HPV 18 at titers of 400 and HPV 58 at titers ranging from 400-800 (Table 1), while anti-L2(11-128) sera neutralized HPV 18 at a titer of 200 and HPV 58 at a titer of 400 (Table 1). None of the sera from intranasal-immunized mice demonstrated neutralizing activity, consistent with lower anti-L2 titers for intranasal than for intraperitoneal immunized mice.

Example 4

An overview of the immunization strategies for lectin-binding proteins, such as galectin-3, is shown in Table 18.

BALB/c mice were immunized with 2 mg/kg mRNA, complexed with LNPs, or 20 μg recombinant protein as indicated in Table 18. Plasma anti-galectin-3 IgG titers were assayed 7 days after the final boost, which was delivered at day 55.

FIG. 3 shows that the use of galectin-3 mRNA as a final boosting agent resulted in a significantly higher target-specific IgG titer than when purified recombinant protein (a traditional immunogen) was used. This effect was observed regardless of whether the antigens were delivered subcutaneously or intravenously.

Hybridomas producing galectin-3-specific antibodies were generated, and high affinity monoclonal anti-galectin-3 antibodies were obtained from further screens.

Table 19 provides a target protein-specific summary of the total number of hybridoma wells (generally about one third (⅓) of these wells contain hybridomas) screened and the number of confirmed target-specific antibodies obtained from those hybridomas wells following the use of lipid-encapsulated mRNA as an immunogen.

Table 20 provides a comparison of mRNA-LNP immunization methods with other conventional methods of immunization by number of hybridomas producing target-specific antibodies. In general, these data suggest that mRNA-LNP immunization is an effective method for inducing an immune response to a target protein antigen and for obtaining a higher number/rate of target protein-specific antibodies. In particular, these results confirm that mRNA-LNP immunization is surprisingly more effective than conventional immunization methods for obtaining antibodies specific for transmembrane proteins, e.g., multi-pass transmembrane proteins, such as GPCRs, which are difficult to raise antibodies against, and for poorly immunogenic proteins (e.g., proteins which produce low or no detectable target-specific IgGs in plasma of animals immunized with traditional antigen).

In general, successful generation of hybridomas producing antigen-specific antibodies have been achieved for at least 15 different targets utilizing mRNA-LNP immunization methods as exemplified herein. These results show that the mRNA immunization methods described herein are capable of eliciting an immune response against a wide range of antigens (e.g., transmembrane proteins, for example multi-pass transmembrane proteins, such as GPCRs) in host animals, and are effective methods for producing high affinity monoclonal antibodies, which can serve as parentals for generation of chimeric variants, humanized variants, and affinity matured variants.