A novel Fcγ receptor ligand augments humoral responses by targeting antigen to Fcγ receptors

Authors


Abstract

Generating efficient antibody (Ab) responses against weak antigens remains challenging. Ab responses require antigen (Ag) uptake by antigen-presenting cells (APC), followed by presentation of processed Ag to T cells. Limited uptake of antigenic peptides by APC constrains Ab responses. Here we improve vaccine efficacy by targeting Ag to Fcγ receptors (FcγR) using R4, a recombinant FcγR ligand. R4 has four repeats per chain of the hinge region and CH2 domain (HCH2) of human IgG1. HCH2 encompasses the FcγR binding site. The repeats are linked to the human IgG1 framework. To test R4 in augmenting Ag uptake, we expressed human serum albumin domain 1 (HSA1) at the N terminus of R4 to produce HSA1R4. HSA1R4 (50 μg) administered to mice in Ribi adjuvant induces up to 1100-fold higher HSA1-specific IgG titers than HSA1 (p<0.001). HSA1R4 (250 ng) induces up to 130 times more anti-HSA1 Ab than HSA1Fc, a protein with HSA1 linked to the IgG1 framework (p<0.001). HSA-reactive T cells proliferate more briskly to HSA1R4 than to HSA1Fc (p<0.008). Immunization with HSA1R4 yields greater T cell reactivity to HSA1 ex vivo than immunization with HSA1Fc (p<0.004). Linking antigenic peptides to linear HCH2 polymers may facilitate vaccine development.

Abbreviations:
HCH2:

hinge region and CH2 domain

HSA:

human serum albumin

IC:

immune complex

MPL:

monophosphoryl lipid A

TDM:

trehalose dicorynomycolate, PE: peritoneal exudate

Introduction

Induction of immunity to pathogens, toxins, and peptides expressed by tumor cells, requires the coordinated participation of the innate and adaptive immune systems. An early step is Ag internalization by APC of the innate immune system, notably by dendritic cells (DC), the most potent APC type, and the one best able to present Ag to naive T cells 1. Internalized Ag is processed through the endosomal/lysosomal path. Processed peptides, bound to MHC molecules, are then delivered to the cell surface. Those T cells with appropriate receptors respond to such peptides provided co-stimulatory molecules are expressed by the DC. A second signal is often required to drive DC maturation and efficient co-stimulatory molecule expression. Ag activates B cells bearing appropriate surface Ig directly to produce IgM. CD4+ T cells, having responded to processed Ag, induce Ig class-switching from IgM to IgG.

Limited uptake of soluble antigenic peptide by DC constrains subsequent Ag processing and presentation. Immune responses increase when Ag uptake is facilitated. IgG immune complexes (IC) bind to FcγR expressed on DC and this is followed by internalization of IC with their captured Ag. Thus, stronger Ab responses occur when soluble Ag is complexed to IgG, than when Ag alone is administered 2. IC in Ab excess are more effective at Ag presentation than IC at equivalence or in Ag excess 3. IC driven, FcγR-mediated, Ag internalization favors DC maturation and hence expression by them of co-stimulatory molecules 4. Other means to target Ag to FcγR on APC have been employed in order to elicit strong immune responses against otherwise weak immunogens. Early studies that documented the potential of this approach employed Ag-containing anti-FcγR mAb as a means to facilitate delivery of Ag to APC and hence increase Ag-specific T cell responses and Ag-specific humoral responses 58. Modification of Ig by introduction of epitopes within the CDR region (i.e. antigenized Ig) also enhances immune responses compared to Ag alone 9, 10.

Mice express four classes of FcγR: FcγRI (CD64), FcγRIIb (CD32), FcγRIII (CD16), and the recently characterized FcγRIV 1114. FcγRI is the high-affinity receptor for monomeric IgG though it also binds IgG IC, and IgG aggregates. FcγRIIb, FcγRIII, and FcγRIV have low affinity for monomeric IgG but share with FcγRI the capacity to avidly bind IC and IgG aggregates 11, 12, 14, 15. FcγRI, FcγRIII, and FcγRIV are activatory due to their obligate association with the heterologous FcR γ chain, which contains a cytoplasmic immunoreceptor tyrosine-based activation motif 16. FcγRI and FcγRIII facilitate IC uptake and Ag processing by DC 17, 18. FcγRIIb contains an immunoreceptor tyrosine-based inhibitory motif in its cytoplasmic tail 16. The role of FcγRIIb in DC activation is complex. Cross-linking of FcγRIIb molecules to one another favors receptor-mediated endocytosis but when FcγRIIb cross-links to FcγR of other classes, internalization of IC and DC maturation are inhibited, so that binding to FcγRIIb can facilitate or inhibit Ag presentation depending on the circumstances 1922.

To exploit the potential of FcγR to augment Ag uptake, we have targeted Ag to FcγR using novel FcγR ligands. Previously we described polymeric FcγR ligands that contain linear repeats of the hinge region and CH2 domain (HCH2) of human IgG1 23. The FcγR binding site is situated within HCH2 2426. The polymeric FcγR ligand described herein expresses key structural features of small IC in Ab excess within a defined molecule 23. Our aim was to evaluate the ability of a polymeric FcγR ligand with four HCH2 repeats, designated R4, to augment immune responses to weakly antigenic domain 1 of human serum albumin (HSA1) in SJL/J and C57BL6 mice. We show that immunization with HSA1 linked to R4 (HSA1R4) is followed by efficient Ag presentation and strong humoral responses.

Results

Expression of HSA1R4, HSA1Fc and HSA1

HSA1 was expressed in a construct with four HCH2 repeats per chain (HSA1R4), as an Fc fusion protein (HSA1Fc), or with a 6× His tag (HSA1) (Fig. 1). Each of the HCH2 repeats contains the locus (residues 226–350, EU numbering system) that binds to FcγR. The HCH2 repeats have been integrated into a typical IgG fusion protein consisting of the IgG1 framework (residues 226–457), thus retaining the dimeric structure of IgG. In contrast, HSA1Fc is a typical IgG fusion protein. HSA1, which spans residues 1–197 of the mature HSA polypeptide 27, is 67% identical and 82% similar to its murine homolog. HSA1, although weakly antigenic for mice, contains both T and B cell epitopes and, accordingly, provides a model for study of techniques to facilitate T cell-dependent Ab responses against weak Ag 2830. HSA can be converted into a stronger Ag when presented to APC in an IC, and is widely employed as a carrier for haptens 29, 30.

Figure 1.

Schematic illustration of the structures of IgG, HSA1R4, HSA1Fc, and HSA1. Drawing on the left represents IgG1 polypeptide. The element labeled Fc represents the IgG1 framework composed of the hinge, CH2 and CH3 domains of human IgG1 with one light chain missing to reveal heavy chain structure. The small chain extending from the CH2 region represents N-linked carbohydrate at Asn297. The next drawing shows HSA1R4. The darkened ovals represent four repeated HCH2 units that link the Fc framework to HSA1. To prevent inter-chain disulfide bond formation between repeat units, hinge region cysteines were mutated to serines. The mutations leave intact those hinge residues known to interact with FcγR. The next drawing depicts HSA1Fc wherein HSA1 has been fused to the Fc region. The final drawing shows HSA1.

Binding of HSA1R4 to FcγR expressed on myeloid cells

Binding of HSA1R4 to FcγR was determined by flow cytometry using the human monocytic cell line U937, murine peritoneal exudate (PE) macrophages, and human DC. U937 cells constitutively express FcγRI and FcγRII 31. Increased fluorescence is uniformly observed when U937 cells are incubated with HSA1R4 followed by FITC-conjugated goat anti-HSA polyclonal IgG to detect surface-bound HSA1R4 (Fig. 2A). Murine PE macrophages and immature human DC express FcγRI, FcγRII, and FcγRIII 16. Increased fluorescence is observed in PE macrophages and in DC following incubation with HSA1R4 (Fig. 2B and 2C).

Figure 2.

HSA1R4 binds to FcγR. Flow cytometric analysis of HSA1R4 binding detected with FITC-anti-HSA goat Ig is shown in black, background fluorescence of cells stained with HSA1R4 and FITC-goat Ig in white. (A, B) Pre-incubation with monomeric human IgG plus heat-aggregated human IgG (gray) abrogates HSA1R4 binding to U937 cells (A) and PE macrophages (B). (C) Ab to FcγRI, FcγRII, and FcγRIII partially block binding of HSA1R4 to DC (gray). (D) Ab to FcγRI partially blocks binding of HSA1R4 to U937 cells (gray). (E) Ab to FcγRII partially blocks binding of HSA1R4 to U937 cells (gray). (F) Ab to both FcγRI and FcγRII completely blocks binding of HSA1R4 to U937 cells (gray).

U937 cells and murine PE macrophages were pre-incubated with monomeric human IgG to block FcγRI, and with aggregated human IgG to block FcγRII and FcγRIII, followed by exposure to HSA1R4. Pre-incubation totally blocked binding of HSA1R4 to U937 cells and to murine PE macrophages, indicating that HSA1R4 binds specifically to FcγR. To detect binding to specific FcγR, U937 cells were pre-incubated with blocking mAb to FcγRI, to FcγRII, or to both. Decreased fluorescence is observed following pre-incubation with either mAb while pre-incubation with both reduces fluorescence to background levels (Fig. 2D–F). Thus, HSA1R4 binds exclusively to FcγRI and FcγRII on U937 cells.

Pre-incubation of immature human DC with a combination of blocking Ab to FcγRI, FcγRII, and FcγRIII lessens binding of HSA1R4 to DC, indicating that HSA1R4 binds to FcγR on the surface of DC (Fig. 2C). Note, however, that Ab pre-incubation does not block all HSA1R4 binding. Since HSA1R4 is produced in SF9 insect cells, it lacks terminal sugars. Incompletely glycosylated IgG may bind to mannose receptors while fully glycolsylated IgG does so poorly. To determine the role of mannose receptors in residual binding of HSA1R4 to DC, DC were incubated with mannan to block mannose-binding receptors along with blocking Ab to FcγRI, FcγRII, and FcγRIII. Binding of HSA1R4 was then measured. Fluorescence intensity was not altered by the addition of mannan, thereby indicating that HSA1R4 does not bind to the mannose receptors on DC (data not shown).

HSA1R4 injected intravenously induces greater Ab responses than HSA1Fc

When Ag/Ab complexes are injected intravenously in mice, greater Ab responses are observed than with Ag alone 2, 32, 33. With this in mind, we injected mice intravenously with 50 μg of HSA1, HSA1Fc, or HSA1R4, obtained serum 14 days later, and assayed Ab responses by ELISA. Anti-HSA Ab titers were not detected in mice given HSA1 alone at the minimum 1:200 serum dilution used as a cutoff (Fig. 3). Mice given HSA1R4 or HSA1Fc developed substantial Ab responses to HSA (Fig. 3). HSA-specific titers were fourfold higher in mice injected with HSA1R4 than in mice injected with HSA1Fc (p<0.05). Equal mass weights of the proteins were injected so that the amount of HSA1 in HSA1Fc was 2.5 times that in HSA1R4 (Fig. 3). Anti-HSA IgG1 and IgG2c were increased in response to both immunizations, indicating activation of both Th1- and Th2-type T cells (Fig. 3).

Figure 3.

HSA1R4 increases HSA1-specific Ab responses following intravenous injection of 50 µg of either HSA1R4, HSA1Fc, or HSA1. Titers of HSA-reactive IgG at 2 wk post-immunization (four mice per group) are shown as mean ± SEM. Also shown are IgG1 and IgG2c titers of the same sera. HSA-specific Ab titers are significantly higher in mice receiving HSA1R4 than in mice receiving HSA1 (p<0.001) or HSA1Fc (p<0.05); ND = not done. HSA-specific IgG titers for mice injected with HSA1 were < 0.2 × 1000.

HSA1R4 administered subcutaneously in Ribi adjuvant induces greater Ab responses than HSA1Fc

Vaccines are typically injected subcutaneously. Accordingly, we determined the efficacy of HSA1R4 as an Ag delivery agent. HSA1R4 was emulsified in Ribi adjuvant. Ribi adjuvant contains monophosphoryl lipid A (MPL), which signals through Toll-like receptor (TLR)4 to activate APC maturation and to increase co-stimulatory molecule expression 34, 35. HSA1R4 in the absence of adjuvant does not up-regulate co-stimulatory molecule expression on immature human DC (data not shown). Mice were immunized with HSA1 alone, HSA1Fc, or HSA1R4 (50 μg/mouse), and anti-HSA Ab titers determined in sera obtained 14 days later. In mice immunized with HSA1 alone, anti-HSA Ab titers were only detectable in sera from two of five mice at the 1:200 cutoff threshold employed (Fig. 4A). Ab titers of mice given HSA1R4 averaged 1100 times those of mice given HSA1 (p<0.001), and were sevenfold those of mice given HSA1Fc (p=0.01). Isotype analysis revealed that both IgG1 and IgG2c Ab titers rose following immunization with HSA1R4.

Figure 4.

HSA1R4 in Ribi adjuvant enhances Ag-specific Ab responses. (A) Mice were immunized with 50 µg of either HSA1R4 (n=6), HSA1Fc (n=4), or HSA1 (n=5) subcutaneously. Sera were obtained 2 wk later. Titers of total IgG reactive with HSA are shown as mean ± SEM, as are IgG1 and IgG2c HSA-specific titers. HSA-specific total IgG titers are higher in mice receiving HSA1R4 than in mice receiving HSA1Fc (p=0.01) or HSA1 (p<0.001). Mean HSA-specific Ab titers in HSA1 immunized mice were 0.72 × 1000. (B) Mice were immunized with either 250 ng of HSA1R4 (n=8) or HSA1Fc (n=7) subcutaneously in Ribi adjuvant. Anti-HSA Ab titers are higher in mice given HSA1R4 than in mice receiving HSA1Fc (p<0.001). HSA-specific Ab titers in HSA1Fc immunized mice were <0.2 × 1000.

We next tested a 200-fold lower dosage of immunogens (250 ng), again in Ribi adjuvant, with Ab measured in sera obtained 14 days post-immunization. Mice given HSA1R4 developed 130 times as much HSA1-specific Ab as mice immunized with HSA1Fc (p<0.001; Fig. 4B).

Since SJL mice are known to produce abnormally high levels of IgG 36, C57BL6 mice were immunized with HSA1R4, HSA1Fc, or HSA1 all in Ribi adjuvant. Sera were collected 14 days later and assayed for HSA-specific Ab titers. As expected, C57BL6 mice develop Ab titers that are substantially lower than those observed in SJL mice. Nonetheless, anti-HSA Ab titers in C57BL6 mice receiving HSA1R4 are tenfold those receiving HSA1Fc (p<0.05) and 50-fold higher than those receiving HSA1 (p<0.005) (Fig. 5). To rule out a non-specific adjuvant effect of the HCH2 polymer, mice were immunized with OVA, or OVA together with HSA1R4, both in Ribi adjuvant. Sera obtained 14 days later were assayed for OVA-specific and HSA-specific Ab. Comparable levels of OVA-specific Ab were detected in sera from OVA-immunized, and OVA plus HSA1R4-immunized mice. Mice given OVA plus HSA1R4 had 30-fold higher anti-HSA1 Ab titers than the anti-OVA Ab titers generated in these same mice (data not shown), indicating that HCH2 polymers augment Ab responses only for Ag bound to the polymer.

Figure 5.

HSA1R4 increases HSA1-specific Ab responses in C57BL6 mice. Mice were immunized with 50 µg of either HSA1R4, HSA1Fc, or HSA1 subcutaneously in Ribi adjuvant. Sera were obtained 2 wk later (seven mice per group). Titers of HSA-reactive IgG are shown as mean ± SEM. HSA-specific Ab titers are tenfold higher in mice given HSA1R4 than in mice given HSA1Fc (p<0.05) and 50-fold higher than in mice given HSA1 (p<0.005).

C1q binding to HSA1R4, HSA1Fc, and monomeric IgG

IC uptake can be facilitated by complement binding followed by uptake through complement receptors expressed by APC. Binding of C1q to IC is an early step in complement activation. C1q binding to HSA1R4, HSA1Fc, and human monomeric IgG was measured by ELISA. HSA1R4 binds C1q to the same modest extent as HSA1Fc and monomeric IgG (Fig. 6).

Figure 6.

Binding of C1q to HSA1R4 measured by ELISA. HSA1R4, HSA1Fc, or monomeric human IgG (Sigma) were immobilized onto ELISA plates at 2–10 µg/mL. C1q was added to ELISA plates at 4 µg/mL. Bound C1q was detected using HRP-conjugated goat anti-C1q IgG followed by OPD addition. Data are expressed as OD. Approximately equal amounts of C1q bind HSA1R4, HSA1Fc, and IgG at all concentrations of ligand tested. Data shown are representative of three separate experiments.

HSA1R4 is a potent Ag delivery vehicle for induction of T cell responses

Mice were immunized with 50 µg of HSA1R4 or HSA1Fc in Ribi adjuvant. HSA1-specific splenic T cell proliferative responses were measured 14 days later. Taking a stimulation index of 3 as indicative of response, T cells from mice given HSA1R4 responded to a 20-fold lower concentration of HSA1 than T cells from mice given HSA1Fc (Fig. 7A). T cells from mice given HSA1R4 responded better to all six concentrations of HSA1 tested than T cells from mice given HSA1Fc (p<0.004; Fig. 7A). The potency of HSA1R4 may be understated in our assay as mice given HSA1Fc received 2.5 times as much HSA1 as those given HSA1R4.

Figure 7.

(A) HSA1-induced T cell proliferation is higher in splenocytes from mice immunized with HSA1R4 than in splenocytes from mice immunized with HSA1Fc (p<0.004). Shown are proliferative responses of cells from mice immunized 2 wk previously with HSA1R4 or HSA1Fc in Ribi adjuvant, and challenged in vitro with HSA1. Data shown are the mean ± SEM of four experiments. (B) HSA1R4 augments presentation of HSA1 to HSA-reactive T cells. Shown are proliferative responses of cells isolated from LN of mice immunized 14 days previously with HSA in CFA following in vitro challenge with either HSA1R4, HSA1Fc, or HSA1 (1.6×10–9 M for each). HSA1R4 leads to greater T cell reactivity (p<0.008 vs. HSA1Fc; p<0.001 vs. HSA1). Data shown are the mean ± SEM of four experiments.

HSA1R4 presents Ag to T cells more efficiently than HSA1Fc

Targeting of Ag to FcγR increases Ag uptake by APC, Ag processing by them, and Ag presentation to T cells 9, 37. Accordingly, we tested whether HSA1R4 could increase Ag presentation to T cells. Cells isolated from draining LN of mice immunized with HSA in CFA 14 days earlier were used as a source of HSA-reactive T cells. Mitomycin C-treated splenocytes from naive mice served as a source of APC. HSA-reactive T cells respond more briskly to HSA1R4 than to HSA1Fc (p<0.008) or to HSA1 (p<0.001) at molar equivalents (1.6×10–9 M) (Fig. 7B).

Discussion

We show that robust Ab responses to HSA1, a weakly antigenic peptide, can be obtained by coupling it to the HCH2 polymer R4. HSA1R4 has ten potential FcγR binding regions and two copies of HSA1. IC in Ab excess, known facilitators of Ab responses to weak Ag, bind to FcγR expressed by APC 29, 30. Binding of IC to FcγR triggers IC internalization so that more Ag enters the APC than when Ag alone is given. Augmented Ag processing, and increased presentation of processed Ag to T cells, ensue. In HSA1R4, these properties of IC have been integrated into a single defined molecule. Additionally, placement of Ag at the N terminus of the HSA1R4 molecule renders the Ag fully accessible to processing enzymes. In theory, any antigenic peptide, of any size, or more than one when the goal is to develop a polyvalent vaccine, can be linked to R4 for Ag delivery.

HSA1R4 was emulsified in Ribi adjuvant prior to its subcutaneous administration. Ribi adjuvant contains MPL, an LPS derivative that retains the adjuvant properties of LPS minus its toxicity, and trehalose dicorynomycolate (TDM), a derivatized form of mycobacterial trehalose dimycolate that likewise has adjuvant properties with minimal toxicity 38. MPL binds TLR4, engagement of which activates DC 34, 35. TLR4 signaling drives DC maturation and induces marked up-regulation of co-stimulatory molecule expression on the DC surface, a requirement for T cell response to Ag 34. Despite the immunostimulatory properties of Ribi adjuvant, Ab titers in mice given HSA1 alone remained modest, in keeping with results from other studies that have shown a requirement for booster injections to achieve high Ab titers to weak Ag, including HSA, even when Ribi adjuvant is employed 28, 39. HSA-specific Ab responses in mice immunized once with HSA1R4 were 1300 times those of mice receiving HSA1. A limitation of Ribi adjuvant, as with adjuvants in general, is a lack of receptor-specific targeting essential for optimal Ag presentation. Our results demonstrate that delivering Ag in the context of an HCH2 polymer overcomes this limitation.

HSA1R4 was compared directly to HSA1Fc, an Ig fusion protein which contains the two FcγR binding sites found in native monomeric IgG. Ab titers in response to 50 µg of HSA1R4 given to SJL/J mice intravenously in saline, or subcutaneously in Ribi adjuvant to SJL or C57BL6 mice, were significantly higher than those to 50 µg of HSA1Fc. Strikingly, SJL mice given a low dose of HSA1R4 (250 ng) in Ribi adjuvant generated anti-HSA1 Ab responses 130 times those of mice given a like amount of HSA1Fc. These findings point to superiority of the HSA1R4 construct with ten potential FcγR binding sites over the HSA1Fc construct with two.

HSA1R4 binds specifically to both high- and low-affinity FcγR expressed by the human monocytic cell line U937 and to murine PE macrophages. HSA1Fc also binds specifically to FcγR expressed on U937 cells (data not shown). FcγRI, alone among FcγR, binds monomeric IgG, so that HSA1Fc binding to U937 cells and to PE macrophages is likely to be restricted to FcγRI. This high-affinity FcγR has been implicated as a major route of IC-mediated Ag presentation based on studies in mice deficient in FcγRI, FcγRII, or FcγRIII 17, 18. The basis for the superiority of HSA1R4 over HSA1Fc remains to be established. We postulate that it relates to the ability of HSA1R4 to bind to FcγRIIb and FcγRIII, in addition to FcγRI, and the inability of HSA1Fc to do so (in preparation). A greater ability of HSA1R4 to promote homo- and hetero-aggregation of FcγR may also contribute to the larger responses observed using HSA1R4 rather than HSA1Fc.

HSA1R4 binds to DC. A cocktail of Ab to FcγRI, FcγRII, and FcγRIII decreases HSA1R4 binding substantially but not to background levels. HSA1R4 may bind to DC through interaction with uncharacterized FcγR. Binding of IC to complement receptors on APC is known to be an absorptive pathway that can lead to increased Ag presentation, but significant binding of HSA1R4 to complement receptors is unlikely since HSA1R4, HSA1Fc, and monomeric IgG all bind C1q to a comparably limited extent. An additional receptor-mediated mechanism that we believe we have excluded as an explanation for HSA1R4 binding to DC is via the mannose-binding receptors. HSA1R4 is produced in insect cells which place core sugars onto IgG, but not N-terminal carbohydrates 40. IgG molecules from which the N-terminal carbohydrate residues have been trimmed show increased binding to mannose receptors relative to native fully glycoslyated IgG 41. Binding of carbohydrate-trimmed IgG to mannose-binding receptors can be blocked by addition of mannan 41. HSA1R4 binding to DC is unaltered by prior exposure to mannan, pointing to a minimal role, at best, for mannose-binding receptors in HSA1R4 binding to DC.

The HSA1R4 and HSA1Fc constructs are based on human IgG1 sequences but were tested in mice to establish proof of principle. Humans express a larger repertoire of FcγR than that found in the mouse. This chiefly results from expression in man of receptors from two additional activatory FcγRII genes: FcγRIIA (which shares features with murine FcγRIV) and FcγRIIC 15. Both are expressed on DC and both are activatory so that the magnitude of response to HSA1R4 in humans might even exceed that observed in mice 11.

While the present study has focused on Ab/CD4+ T cell responses, IC also facilitate “cross-presentation” of Ag by MHC class I with activation of cytotoxic CD8+ T cells, a population crucial for an effective response against solid tumors 4, 42. Future studies will be needed to determine whether HCH2 polymers can duplicate this effect of IC. What can be said is that HCH2 polymer-mediated delivery of Ag to APC results in large and rapid humoral responses against an otherwise weak Ag. Our findings suggest that HCH2 polymers may represent an effective method to augment vaccine efficacy.

Materials and methods

Mice

SJL/J mice and C57BL6 mice, 5–6 wk old, from Jackson Laboratories (Bar Harbor, ME) or from Taconic (Germantown, NY), were maintained in a barrier facility and acclimated for 1–2 wk before study. Animal care and experiments were performed according to NIH guidelines, as approved by the animal use committee of the University of Chicago.

HSA1, HSA1Fc, and HSA1R4 cloning

Total RNA was isolated from cell line Hep G2 (ATCC HB-8065) using the method of Chomczynski and Sacchi 43. First-strand cDNA synthesis was primed with 100 pmol random hexamers using 200 U SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA) and 5 μg of total RNA in a 20-μL reaction mixture that was 500 μM in dNTP (Pharmacia, Piscataway, NJ), 1 U RNasin/μL (Promega Corp., Madison, WI), 10 μM in DTT, and 1X in first-strand buffer. Reaction proceeded at 42°C for 50 min. Domain 1 of mature HSA (HSA1) was amplified from Hep G2 cDNA using PCR, the forward primer Dom1-F (5′-GGCCGCATCTCGAGATGAAGTGGGTAACCTTTATTTCC-3′), and the reverse primer Dom1-R (5′-CCGCATGAATTCTCTCTGTTTGGCAGACGAAGCCTT-3′). The leader sequence and the first 197 amino acid residues of mature HSA (i.e. HSA1) 27 were amplified, and flanking 5′ Xho I and 3′ Eco RI sites were introduced. The PCR product was digested with Xho I and Eco RI and ligated into like-digested pBSKS+ cloning vector (Stratagene, La Jolla, CA) to produce clone pHSA-BS.

The HSA1 fragment was subcloned into previously described 23 Fc and HCH2 polymer expression vectors to yield pHSA1Fc and pHSA1R4, respectively. To express HSA1 with a 6× His tag, a short linker that introduces a His tag and a 3′ stop codon was ligated into the Eco RI and Sal I sites of pHSA-BS. Fusion protein cDNA were transferred into the baculovirus expression vector, pFastBac1 (Invitrogen), by digestion with Bam HI and Sal I and subsequent ligation of the isolated cDNA fragments into the same sites on pFastBac1 to produce pHSA1-FB, pHSA1Fc-FB and pHSA1R4FB. The pFastBac1 expression constructs were transformed into DH10Bac competent cells (Invitrogen) following manufacturer's instructions and correctly recombined virus was identified using PCR.

Baculovirus-mediated protein expression and purification

Cell line Sf9 (ATCC CRL-1171) was maintained in ExCell 420 serum-free medium (JRH Biosciences, Lenexa, KS) supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin. For bacmid transfection, 1 × 106 cells were plated into each well of a 6-well cluster and allowed to grow overnight. Transfection medium was replaced with 2 mL fresh ExCell 420 without antibiotics. Two hours later, bacmid DNA (6 μg) was transfected into Sf9 cells using Cellfectin reagent (Invitrogen). After 9 h, the medium was replaced with fresh medium containing antibiotics. Forty-eight hours later, medium containing virus was harvested and used in a second round of viral amplification.

For protein expression, 100 mL of medium supplemented with 1% Pluronic F-68 (Invitrogen) in shaker flasks was seeded with 4×105 Sf9 cells/mL and shaken at 110 rpm at 27°C for 24 h at which time virus was introduced. Conditioned medium was harvested 72 h later, and PMSF (Research Organics, Cleveland, OH) plus pepstatin A (Peptides International, Louisville, KY) were added to a final concentration of 1 mM and 1 μM, respectively. HSA1Fc and HSA1R4 were purified using protein G-Sepharose (Pharmacia) as described previously 23.

The 6× His-tagged HSA1 protein was purified using a Ni2+ immobilized resin (Ni-NTA; Qiagen, Valencia, CA). Prior to application to the column, it was necessary to remove interfering ions and peptides by dialyzing the conditioned medium (12 000–14 000 MWCO Spectrapor tubing) against 20 mM Tris pH 7.9 and 0.5 M NaCl (buffer TN) with 5 mM imadazole for 36 h (one buffer change). Dialyzed conditioned medium was loaded onto a 2.5-mL bed column at a rate of 1 mL/min. The column was washed with buffer TN with 30 mM imadazole, and HSA1 was eluted from the column with 0.5 M imidazole in buffer TN. Eluted proteins were dialyzed extensively against endotoxin-free PBS pH 7.0, tested for endotoxin content using the Kinetic-QCL limulus amebocyte assay (BioWhittaker, Walkersville, MD), aliquoted, and stored at –70°C for future use.

Immunization

HSA1, HSA1Fc, or HSA1R4, dissolved in 0.15 mL saline, was injected into a tail vein. For s.c. injections, HSA1, HSA1Fc, HSA1R4, or OVA (Sigma, St. Louis, MO) were suspended in Ribi adjuvant (Sigma) according to manufacturer's instructions. Ribi adjuvant contains MPL and synthetic TDM incorporated into a mix of squalene and Tween-80, and serves as an immunostimulant with little toxicity. Proteins were dissolved in 2 mL of saline from 0.0025 to 0.5 mg/mL, transferred into vials containing 0.5 mg of MPL and 0.5 mg of TDM, and vortexed for 4 min to create an oil-in-water emulsion. Mice were immunized subcutaneously at two sites, one on each flank. A total volume of 0.1 mL containing 0.125–25 µg of protein was injected at each site. Mice were bled retro-orbitally. To generate HSA1-reactive T cells for in vitro use, mice were injected with 0.1 mL of an emulsion consisting of 0.05 mL saline containing 100 μg of HSA (Sigma) and 0.05 mL of CFA distributed intradermally with 0.025 mL given in each flank and over each scapula.

Flow cytometry

Cells were analyzed using a FACScan II (BD Biosciences, San Jose, CA).

Human cells

The monocytic cell line U937 (ATCC, Rockville, MD) was maintained in RPMI 1640 supplemented with 10% FBS and 2 mM L-Glutamax. Cells were suspended in wash buffer (1% OVA in Dulbecco's PBS) at 1×107 cells/mL. To detect binding, 5 μg of HSA1R4 was added to a 0.05-mL suspension of cells. The cells were incubated at 4°C for 20 min, washed, and resuspended in 0.05 mL of wash buffer containing affinity-purified anti-HSA FITC-conjugated goat IgG (1:100 dilution; Bethyl Laboratories, Montgomery, TX). To show specificity of binding, FcγR were blocked by incubation at 4°C for 20 min with 50 μg of human monomeric IgG (Sigma) to block FcγRI, plus 50 μg of heat-aggregated human IgG to block FcγRII and FcγRIII, and then incubated with HSA1R4 as described above. IgG in saline (50 mg/mL) was aggregated at 63°C for 30 min. IgG aggregates were pelleted by centrifugation at 10 000×g for 10 min. The pellet was resuspended in saline prior to use. U937 cells were pre-incubated with mAb to FcγRI (CD64, clone 10.1; BD Biosciences) and to FcγRII (CD32, clone FLI8.26; BD Biosciences) to block the ability of HSA1R4 to bind to FcγR.

Dendritic cells

Heparinized blood (10 mL) was diluted threefold in HBSS and overlain onto 10 mL of Ficoll-Hypaque. Tubes were centrifuged for 25 min at 300×g. Cells at the HBSS Ficoll-Hypaque interface were collected, washed three times in HBSS by successive suspensions in HBSS and centrifugations, and then suspended in RPMI with 10% FBS. Cell suspension (5 mL) was plated into T25 flasks for 2 h. Nonadherent cells were removed and 5 mL of medium containing 20 ng/mL of GM-CSF (Peprotech, Rocky Hill, NJ) and 20 ng/mL of IL-4 (Peprotech) was added to each flask. After 6 days of incubation, cells were harvested and processed as described for U937 cells. mAb to FcγRI, FcγRII, and FcγRIII (CD16, clone 3G8; BD Biosciences) were used to block binding of HSA1R4 to FcγR. Mannan (100 µg; Sigma) was added to block potential binding of HSAR4 to mannose-binding receptors along with blocking mAb to FcγR 41.

Mouse cells

Mice were injected i.p. with 0.8 mL of 3% thioglycollate medium (Sigma), euthanized 3 days later, and their peritoneal cavities lavaged with 2 mL of ice-cold saline. Recovered cells were washed with saline, suspended in wash buffer, and processed as described above. Forward scatter vs. side scatter properties of PE cells were used to gate on macrophages. FITC-conjugated goat Ig (Bethyl Laboratories) was used as an irrelevant control.

ELISA

ELISA plates (Corning Inc., Corning, NY) were overlain with 0.1 mL/well of carbonate buffer (0.1 M, pH 8.4) containing 5 μg of HSA or OVA and incubated at room temperature for 5 h. Wells were treated by the addition of 0.1 mL/well of 0.25% Sanalac in Dulbecco's PBS to prevent non-specific binding (Conagra, Irvine, CA). After 2 h at room temperature, wells were washed with 0.5% Tween-20 in Dulbecco's PBS (wash buffer), and 1:200, 1:250, 1:500, 1:2500, 1:12 500, 1:62 500, 1:125 000, 1:250 000, 1:500 000, 1:1 000 000, and 1:2 000 000 dilutions of serum samples in 0.25% Sanalac were added to duplicate wells. Sera from naive mice were diluted 1:200 and added to duplicate wells to provide background OD values.

Plates were left overnight at 4°C, then washed, and overlain with a cocktail of biotinylated rat mAb (each at 0.5 μg/mL) specific for murine Ig (mAb clones: anti-IgG1 A85-1, anti-IgG2b R12-3, anti-IgG3 R40-82; all from Invitrogen; and anti-IgG2c 5.7 from BD Biosciences). IgG2c was measured since SJL and C57BL6 mice express IgG2c rather than IgG2a 44. To quantitate levels of HSA1-reactive IgG1 or IgG2c, wells were overlain with biotinylated Ab specific for those isotypes. Following incubation with biotinylated Ab, wells were washed and overlain with 0.1 mL of affinity-purified peroxidase-conjugated goat anti-biotin Ab (1:500 dilution; Zymed, South San Francisco, CA) for 45 min. Wells were washed, and 0.2 mL of o-phenylenediamine (1 mg/mL) and H202 (1 µL/mL) in citrate buffer (0.1 M, pH 4.5) was added to each well. Absorbance was measured 15 min later using a ThermoMax Microplate Reader (Molecular Devices Corp., Sunnyvale, CA). Serum dilutions were considered positive when their OD values exceeded twice the mean OD values obtained from wells containing non-immune sera. As a control, absorbance values were measured from wells not coated with HSA or OVA but overlain with immune sera. Absorbance values of control wells always approximated those found in blanks.

C1q Binding assay

Binding of human C1q to monomeric human IgG, HSA1Fc, and HSA1R4 was determined using modifications of a previously described ELISA protocol 45. Ligands (2–10 µg/mL) were diluted in PBS and coated onto Costar high-binding ELISA assay plates overnight at 4°C. Plates were washed with 0.05% Tween-20 in PBS (PBS-T) and overlain with 4 µg/mL of C1q (Calbiochem) prepared in PBS-T with 0.1% gelatin (PTG) for 4 h at room temperature. Plates were washed with PBS-T and incubated for 1 h with goat anti-human C1q (Calbiochem, La Jolla, CA) diluted 1:1000 in PTG. Plates were washed with PBS-T and incubated for 1 h with rabbit anti-goat IgG conjugated to horseradish peroxidase diluted 1:10 000 in PTG. The rabbit anti-goat IgG-detecting Ab was pre-incubated with 2.5 ng/mL of human IgG to eliminate residual cross-reactivity to human Ig. Finally, plates were washed with PBS-T and developed with 0.5 mg/mL o-phenylenediamine (Sigma) peroxidase substrate. Absorbance was measured at 450 nm using a ThermoMax plate reader (Molecular Devices).

Proliferative responses

LN and spleens were harvested from mice immunized 14 days earlier with (1) HSA in CFA, or with (2) HSA1R4 or HSA1Fc in Ribi adjuvant. LN and spleen fragments were placed in saline, and disrupted mechanically using a tissue homogenizer to obtain a single-cell suspension. RBC were removed by centrifugation on Ficoll-Hypaque gradients. Buffy layers were harvested from the gradients; cells were washed with HBSS, and resuspended in HL-1 Ventrex medium (Fisher Scientific, Pittsburgh, PA) supplemented with 2 mM L-Glutamax, 50 µm 2-mercaptoethanol, 1× MEM amino acids, and 10 µg/mL gentamicin (Invitrogen). Cells were plated in 96-well flat bottom plates at 6×105 splenocytes/well or 4×105 LN-derived cells/well, and HSA1Fc or HSA1R4 added as indicated in the Results. As a source of APC, splenocytes from naive mice, processed as described above, were incubated at 37°C in RPMI containing 75 µg/mL of mitomycin C (Sigma) for 20 min, washed five times in saline, and added (3×105 cells/well) to wells containing LN-derived cells. Cells were incubated at 37°C for 72 h and pulsed for an additional 8 h with 1 μCi/well of [3H]thymidine. Cells were harvested using a Cambridge PHD cell harvester and radioactivity determined by liquid scintillography.

Statistics

Ab titers, and proliferative responses of HSA-reactive LN cells, were compared using Student's unpaired t-test. Proliferative responses of splenocytes were compared using Chi-square analysis of values above or below a stimulation index of 3.

Acknowledgements

This work was supported by grant RG3563A22/1 from the National Multiple Sclerosis Society (B.G.W.A), by grants 1R41AI060257 and 1R21AI058003 both from NIH/NIAID (D.M.W.), and by a generous gift from Jack Schaps.

Footnotes

  1. 1

    WILEY-VCH

  2. 2

    WILEY-VCH

  3. 3

    WILEY-VCH

  4. 4

    WILEY-VCH

  5. 5

    WILEY-VCH

  6. 6

    WILEY-VCH

  7. 7

    WILEY-VCH

Ancillary