|Home | About | Journals | Submit | Contact Us | Français|
Treatment of IgE-mediated food allergy with standard protein-based allergen immunotherapy has proven both unsuccessful and hazardous. Allergen gene vaccination represents a promising alternate, but difficulties in gene targeting and expression in antigen presenting cells represent a major limitation for efficient gene vaccination.
Construction of a genetically engineered human Epsilon-PolyLysine fusion protein that binds allergen gene expression systems and targets the gene vaccine complex to antigen presenting cells through the interaction of Epsilon-PolyLysine and the high affinity receptor for IgE for efficient allergen gene vaccination.
Genetic engineering was used to design and produce the Epsilon-PolyLysine fusion gene consisting of the human CH□2-4 linked to 55 lysine residues, and the conventional approaches were employed to characterize the biological features of Epsilon-PolyLysine.
Epsilon-PolyLysine was assembled as functional dimers and capable of binding DNA plasmids in both an Epsilon-PolyLysine protein and plasmid DNA concentration-dependent manner. Epsilon-PolyLysine targeted plasmid DNA to the high affinity receptor for IgE on cell surfaces, and increased the model gene uptake/expression. The Epsilon-PolyLysine -DNA complexes were shown not to trigger mast cell degranulation.
Epsilon-PolyLysine is able to function as a gene carrier system to target allergen gene to the high affinity receptor for IgE expressing cells via ligand-receptor mediated interactions.
Epsilon-PolyLysine platform can function as a universal human Fc□RI positive antigen presenting cells targeting system for gene(s) delivery and efficient allergen gene vaccination to food allergies including peanut allergy.
A genetically engineered Epsilon-PolyLysine has been produced and this protein is capable of binding and delivering genes to the appropriate human antigen presenting cells so as to provide the basis for effective allergen gene vaccination in humans.
DNA - gene vaccination has attracted a great deal of interest in a broad range of biomedical areas including allergy, autoimmunity, cancer, and infection. It has become well established as a therapeutic approach in a variety of animal systems (1). DNA vaccination employing allergen genes represents a particularly promising therapeutic strategy for food allergy, a problem of increasing magnitude (2-4) where traditional protein based immunotherapy has proven too dangerous. However, one major hurdle for gene vaccination is its low immunogenicity resulting in relatively weak immune responses. This less than optimal immunogenicity appears to result primarily from the fact that the injected DNA is taken up mainly by cells other than professional APC. Thus it has became clear that effective DNA vaccination will also require delivering sufficient DNA to APC that play a crucial role in the induction of effective cellular and humoral immune responses as well as maximizing gene expression. A variety of approaches to selectively direct the DNA expression systems to APCs are being undertaken including targeting APC surface molecules, e.g. DEC 205 (5) or the use of viral vectors (6).
We have approached this issue by selectively targeting allergen gene vaccine to human dendritic cells (DCs) through their expression of the high affinity receptor for IgE, FcεRI. DCs as professional APCs can initiate T-cell mediated immunity, control a substantial part of the adaptive immune response, and have proven to be of crucial importance for the initiation of transgene-specific immune responses for all methods of DNA vaccination (7). DCs in humans, as opposed to mouse (8), normally express the FcεRI and the level of expression is increased in allergic humans (9-18). Our overall strategy was to genetically engineer and produce a carrier protein that would bind plasmid DNA and direct their expression to DCs through ligand mediated targeting. Thus we designed and produced an epsilon Ig heavy chain polylysine fusion protein (EPL) composed of the last three constant region domains of human epsilon heavy chain plus a 55 amino acid stretch of positively charged lysine. The epsilon portion of EPL provides the DC ligand targeting while the polylysine tails electrostatically bind the negatively charged DNA expression systems to be expressed in the DCs. In this paper we report the construction, production and characterization of this universal human DC targeting EPL.
The EPL fusion protein expression construct was modified from the previously described expression construct for an IgE-IgG Fc fusion protein (E2G) (19). The Hinge-C□2-C□3 region in the original E2G construct was removed by Bgl II-Not I restriction enzyme digestion. A fragment of 180 bp DNA coding for repeated lysine residues was synthesized by primer extension from two oligo nucleotides, A,5′-AGGAAGATCTAAAAAAAAAAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAAGAA GAAGAAGAAGAAGAAGAAGAAAAAAAAGAAGAAGAAGAAGAAGA-3′, B,5′-ATAGTTTAGCGGCCGCCTACTTTTTCTTCTTTTTCTTCTTTTTCTTCTTCTTCTTTTTCTTCTTTTTCTTTTCTTCTTCTTCTTCTTCTTCTTCTTCTT-3′ annealed to each other through a 20 nucleotide complementarity at their 3′ ends. The primer extension reaction was carried out by bacterial DNA polymerase I at 37°C for 30 min, and the products was digested with Bgl II plus Not I and ligated into the Bgl II-Not I digested vector. The final construct containing a 5′ Ig kappa leader sequence connected to the IgE C□2-C□4 domain and DNA coding for 55 lysine residue
For the construction of the EPL+ expression plasmid, we inserted a synthesized oligo nucleotide (5′-GATCCGGAAGAAAGAAACGACGCGGTCGTAGACGACCTAAGAAGAAGCGTAAGGTAGGATCCTACCTTACGCTTCTTCTTAGGTCGTCTACGACCGCGTCGTTTCTTTCTTCG3′) encoding a fragment of the HIV tat peptide sequence (GRKKRRQRRR) plus the nuclear localization signal peptide (PKKKRKV) 5′ of the polylysine tract at the Bgl II site in the EPL construct backbone.
CHO cells were grown in Dulbecco’s Modified Eagles Medium (DMEM) as described (19). Linearized plasmid DNA (5 μg) was electroporated into 5 × 106 CHO cells for stable transfection under G418 selection (1 mg/ml). EPL expressing clones were isolated and subcloned. The collected supernatants were passed an anti-human IgE affinity chromatography for purification, with the method described (19). The purified proteins were analyzed by SDS-PAGE as previously described (19).
Varying amounts of EPL and the plasmid containing Ara h2 cDNA driven by CMV promoter (provided by Drs. W. Burks and G. Bannon, formerly from the Univ. of Arkansas) were mixed and incubated at room temperature for 30 min. Sample aliquots were electrophoresed at 150V for 20 min. in 0.8% agarose gels containing 0.5μg/ml ethidium bromide. To document the DNA binding capability of EPL, different amounts of EPL were added to a constant amount of plasmid DNA, and varying amounts of DNA plasmid were incubated with a constant amount of EPL to provide a wide range of EPL to DNA ratios. Bands corresponding to plasmid DNA were detected under UV light.
Binding of EPL to FcεRI was assessed by flow cytometry using human FcεRIα transgene expressing 3D10 cells and FcεRI+ human basophil-like cell line, Ku812. Cells (106) were incubated with or without EPL or IgE proteins at several concentrations at 4°C for 1 hour, followed by staining with FITC labeled goat anti-human epsilon chain (Sigma) for 30 min at 4°C. Samples were run on a FACScan® flow cytometer (Becton Dickinson), and the data were analyzed with FCS expression V3 (De nove software).
FcεRI expressing 3D10 cells or CHO cells (5×105) were incubated at 37°C for 30 min. in 100μL RPMI-1640 medium with the pCMV-EGFP plasmid (0.5 μg)(Clonetech) alone or in the presence of 0.02μg, 0.05μg, or 0.5μg of EPL and then were seeded at equal numbers in 24-well plates overnight. PBS was used as negative control. GFP expression was monitored by FACS analysis and fluorescence microscopy (Olympus, equipped with U-RFLT50-100 Fluorescence light). Images were further analyzed by the Motic® Images Plus 2.0 ML software.
The purified EPL and EPL+ proteins (~20 ng) were boiled for 5 min. with sample buffer, in the absence or presence of 2-mercaptoethanol (2-ME), and separated using 10 % SDS-PAGE. The separated proteins were electrophoretically transferred to PVDF membranes (Millipore, Bedford, MA). The membranes were blocked overnight at 4°C in pH 7.4 PBS with 5% non-fat dry milk, followed by incubation with rabbit anti-human IgE (Sigma) for 2 hours at room temperature. The washed blots were further incubated with HRP-labeled goat anti-rabbit IgG for 2 hours at room temperature, developed using enhanced chemiluminescence (ECL) reagents (Amersham), and exposed to BioMax film (Eastman Kodak Co).
Mice expressing the human FcεRI□ chain (with the endogenous murine FcεRI□ knocked out) were used for test if EPL would drive mast cell degranulation on skin testing as described previously (19). Various amounts of EPL, along or mixed with plasmid DNA, positive control (5 ng of anti-NP IgE and 10 μg NP30-BSA mixture), and negative control (saline), in 50 μl volume, were injected intradermally into the back of the mice, followed by intravenous challenge with 300 μl 1% Evan’s blue dye. After 30 minutes, the mice were sacrificed and skin reaction was visualized as leakage of blue dye into the ventral surface of the skin through dilated blood vessels at the injection sites.
As human APC’s express a form of the Fc□RI, we designed a bifunctional fusion protein comprised of the human Fcε(2-4) plus a 55 repeated amino acid string of lysines. This protein was engineered to bind DNA expression vectors via the electrostatic interaction between the positively charged polylysine tails and the negatively charged DNA vectors while the Fcε portion would target the protein-DNA complex (termed a “polyplex”) to human APC’s via the high affinity FcεRI binding (Figure 1).
To produce the fundamental first generation EPL, a truncated IgE heavy chain gene (CH□2-4) was directly joined to a 180 bp DNA sequence encoding 55 repeat lysine residues (Figure 2A). EPL proteins expressed in CHO cells and purified by IgE affinity chromatography were assembled as the expected dimers as shown by Coomassie blue staining gels where it migrated with a molecular mass of about 100 kD under native (non-reduced) conditions, whereas under reducing conditions it was primarily at ~50 kD (Figure 2B, the purified human IgE protein was included as a comparison).
To produce a second generation of EPL, or EPL+, the fundamental construct was modified by the insertion of an HIV tat peptide sequence (GRKKRRQRRR) to potentially enhance cellular uptake and a nuclear localization signal peptide sequence (PKKKRKV) to enhance nuclear transport (20, 21). These additional sequences were inserted between the CH□2-4 and polylysine tract sequences to produce the EPL+ (Figure 2A, lower panel). EPL+ protein expression level was much lower than that of the EPL in the stably transfected clone so that the amount of the affinity purified EPL+ was not sufficient to be clearly visualized in the Coomassie staining (data not shown). However, Western blot analysis demonstrated that the EPL+ was also expressed as dimers as was EPL (Figure 2C, the positions for EPL+ dimer and monomer are noted, and the non-specific band is identified with an asterisk). Due to its low expression level, we have not extensively tested the functional properties of EPL+.
To verify that EPL binds DNA and therefore can function as a ligand mediated gene carrier, we mixed differing amounts of the EPL protein with varying amounts of plasmid DNA as indicated in the Figure 3, and ran the samples in gel retardation assays. Binding of protein (EPL or otherwise) to DNA will retard DNA migration in the gel. Poly-L-Lysine (PLL) served as a positive control, whereas the purified IgE alone served as a negative control (Figure 3A). EPL showed significant DNA binding (retardation in lanes 4 and 5) in a manner similar to PLL (lane 3) and in contrast to IgE protein itself where retardation was not observed (lane 2, compared with lane 1). These experiments demonstrated the DNA binding ability of our EPL construct and showed that this binding is both EPL protein (Figure 3B, the EPL:DNA molar ratios was 100:1 (lane 2), 50:1 (lane 3), 25:1 (lane 4), etc) and plasmid DNA concentration-dependent (Figure 3C). While it is not possible to exactly quantify the DNA binding ability of the EPL proteins, these data do provide a general relationship between EPL amounts and DNA binding.
EPL:DNA polyplexes with a GFP expression plasmid were prepared and incubated with either CHO cells that express a transfected human FcεRIα chain (3D10 cells) or Ku812 cells, a human basophil-like cell line that expresses the entire FcεRI receptor complex (19). Binding was detected with FITC labeled rabbit anti-human IgE by FACS analysis. As shown in Figure 4, EPL:DNA polyplexes bound to both 3D10 cells (Figure 4, upper panel) and Ku812 cells (Figure 4, lower panel), at a level equivalent to native IgE demonstrating the key property of FcεRI binding. Binding was not observed with CHO cells that did not express the FcεRIα.
The ability of EPL to enhance the plasmid uptake via IgE-FcεRI binding was tested in the control CHO cells versus CHO cells expressing the human FcεRIα chain (3D10). Cells were incubated with either an GFP expression plasmid alone or an EPL:GFP plasmid polyplex. GFP protein expression as evidence of the plasmid uptake and expression was monitored by fluorescent microscopy and FACS. FACS analysis showed that EPL significantly enhanced the frequency of GFP positive cells (Figure 5), as the significantly enhanced GFP expression frequencies were observed in GFP complexed with EPL (0.75 %, panel III), but neither the GFP plasmid alone in the Fc□RI□ expressing 3D10 cells (0.03 %, panel II), nor in the CHO cell controls (0.01%, panel V & VI). These results were further confirmed by fluorescence microscopy (Data not shown). These in vitro results show that the EPL:GFP polyplex led to increased plasmid DNA uptake and expression via FcεRI targeting.
A potential concern with the EPL protein polyplexes is whether the negatively charged DNA plasmids would cross-link EPL proteins and thereby have the potential to trigger mediator release of the mast cells and basophils. To examine this possibility directly in vivo, we tested the ability of EPL:DNA polyplexes constructed a wide variety of ratios to trigger mast cell release when injected into the skin of human FcεRIα transgenic mice. As shown in Figure 6, a positive control mixture of human anti-NP IgE and its corresponding antigen NP30-BSA induced marked blue staining as evidence of IgE mediated local mediator release (Figure 6A), whereas the negative controls, saline (Figure 6B) or DNA plasmid alone (Figure 6C), did not trigger mediator release. Neither EPL alone (Figure 6D), nor EPL:DNA polyplexes formed at molar ratios of 1:1, 4:1 10:1 or 20:1 (Figures 6E-H, respectively) triggered local mediator release indicating that neither EPL alone, nor EPL-DNA complexes were able to trigger mast cell mediator release.
DNA vaccination is an attractive approach for food allergy as it has been shown to result in a more balanced Th2/Th1 response, including suppression of Th2 responses and IgE production, and enhanced IFN-γ, IgG2a and Th1 responses (22-27). However poor immunogenicity of the naked plasmid DNA vaccines remains a major obstacle to realizing the considerable potential of this approach. We employed an IgE-focused DNA vaccination strategy to enhance the DNA vaccination efficiency. Such a method takes advantage of the extraordinarily high - picomolar affinity interaction between the IgE and the FcεRI which in humans, as opposed to mice, is expressed on APC. In fact, FcεRI is not only expressed on human APCs in general but also expressed in even higher levels on the APCs of atopic individuals. This differential Fc□RI expression pattern on human APCs between atopic and non-atopic individuals provides the rationale for the design of an IgE-based allergen gene vaccination strategy. The successful construction and production of EPL, and the subsequent characterization of EPL’s features in various experiments assays have provided the solid basis for moving this platform forward.
Chemical cross-linking has been used to couple various protein ligands to PLL to prepare ligand-mediated gene transfer reagents (28). However, this method has serious shortcomings; inter-protein cross-linking is likely to occur. Such a potential is especially a concern in the case of IgE or Fc fragment of IgE, as such cross-linking of IgE would be expected to trigger mast cell and basophil mediator release. Furthermore the degree of the cross-linked epsilon-PLL complexes is difficult to control due to the nature of the chemical reaction resulting in size and composition variations that would result in batch-to-batch variation, variation would make the development of a true human therapeutic difficult scientifically and highly problematic at a regulatory level. Our genetic engineering expressed EPL overcomes this problem, as the Fc□ and its fused polylysine tract are expressed as a single and uniform molecule, thereby avoiding the issue of molecular heterogeneity. Indeed, using the in vivo skin testing, we demonstrated that over a wide range of doses and ratios, EPL:DNA polyplexes were not able to trigger mast cell mediator release and should be safe from this vantage point.
Allergy effector cells such as mast cells and basophils will also be targeted by our EPL proteins as they express abundant surface Fc□RI. To avoid allergen gene vaccine expression in these cells, we have designed an expression system that addresses this issue in vivo. We have incorporated the DC cell type specific promoter Fascin into the vaccine gene expression plasmid. Only maturing and mature DCs, follicular DCs and Langerhans’ cells express Fascin (29,30). It is not present in mast cells and/or basophils. Thus use of a Fascin controlled allergen gene expression vector will therefore ensuring the selective DC-specific expression of the targeted allergen genes.
Up to this point, we have necessarily evaluated the characteristics of this EPL platform primarily in vitro. While we can employ the hFc□RI□ tg+ mice to measure acute effects of EPL, e.g. skin test reactivity following exposure to EPL, both wild type and hFc□RI□+ mice will generate an anti-human epsilon IgH response that will confound the interpretation of any findings of a vaccination protocol with EPL. Fortunately, investigation of the in vivo effects of EPL and EPL+ have now become feasible as Dr. Kinet and colleagues have very recently constructed Balb/c mice that are hFc□RI□ tg+ and have the human IgH epsilon gene knocked-in (personal communication). In those “allergy humanized” animals, we will be able to repeatedly inject our EPL polyplexes without the complicating anti-human epsilon response. Those future experiments in “allergy humanized” mice will extend the current in vitro work to a suitable in vivo experimental model for testing the efficiency of this novel Ch□-targeted allergen gene vaccine platform as a potential immunotherapeutic intervention in IgE mediated food allergy.
Supported by Food allergy and anaphylaxis network (FAAN) and NIAID, NIH grant AI 15251.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.