Glycoproteins encoded by MHC loci induce the cellular and humoral responses of allograft rejection (1-3). Alloantigen-specific T cells recognize foreign class I and class II MHC products that are presented on donor cells (direct recognition), or that are processed by antigen-presenting cells (APC) into peptides (indirect recognition; 4, 5). MHC class I molecules are expressed on cells as heterodimers of heavy chains bearing α1,α2, and α3 extracellular domains; theα1 and α2 helices form a peptide-binding cleft with a β-pleated sheet on the floor. The α3 domain of the heavy chain provides a flexible elbow that is noncovalently associated with aβ2-microglobulin (β2-m) light chain, which imposes conformational stability on the heterodimer (6, 7). The structure is further stabilized as self-peptides or allopeptides are loaded onto the peptide binding cleft (8). Although some reports find that the absence of a β2-m or peptide abrogates the antigenic characteristics of class I MHC alloantigens arrayed on normal cells(9, 10), other workers note that soluble class I MHC/β2-m alloantigens released from the graft immunize recipients(11). Increased amounts of circulating donor class I MHC/β2-m alloantigens have been associated with rejection episodes in cardiac and simultaneous kidney/pancreas transplant recipients(12-14). Therefore, the capacity of soluble forms of alloantigens to immunize recipients, particularly alloantigens containing an isolated class I MHC heavy chain protein, remains controversial.
An isolated class I MHC RT1.Aa disparity in rats can induce a vigorous alloimmune reaction, namely, rejection of PVG.R8(RT1.AaDuBuCu) heart allografts by PVG.1U(RT1u) congenic hosts (15). The rejection is associated with an abruptly increased frequency of anti-RT1.Aa-specific T helper cells and cytotoxic T lymphocytes (CTL), as well as with the production of anti-RT1.Aa-specific cytotoxic antibody(15-17). However, the molecular antigenic determinants of the soluble RT1.Aa molecules remain unknown. RT1.Aa proteins purified by passage of liver extracts through an immunoaffinity column failed to sensitize allogeneic hosts in vivo(18). In contrast, administration of complete Freund's adjuvant (CFA) with two short peptides, representing sequences in either theα1 or α2 helical regions of the RT1.Aa molecule, only slightly accelerated the rejection of skin allografts that expressed RT1Aa/β2-m/peptide molecules(19, 20). Intravenous administration of peptides(75-84 amino acids) of the human class I MHC sequence in combination with a short course of cyclosporine (CsA) induced tolerance to heart allografts in rats (21). Similarly, when injected into the thymus (day-2), a mixture of peptides that match the donor class II MHC sequence induced indefinite survival of heart allografts (22).
The present studies examined the immunogenicity of secreted soluble RT1.Aa heavy chain proteins that were produced in a baculovirus expression system and whose constructs contained the melittin secretion (ms) signal attached to the RT1.Aa cDNA. A single subcutaneous injection of either the RT1.Aa heavy chain protein itself or of syngeneic peritoneal macrophages that had been preincubated with the RT1.Aa heavy chain protein induced accelerated rejection of heart allografts bearing native RT1.Aa/β2-m/peptide molecules. Intrathymic or intraportal injection of RT1.Aa heavy chain proteins in combination with anti-T cell receptor (TCR) mAb 2 weeks before grafting induced tolerance to liver allografts that expressed RT1.Aa/β2-m/peptide heterodimers.
MATERIALS AND METHODS
Construction of the RT1.Aaand ms/RT1.Aabaculovirus expression vectors. The first RT1.Aa heavy chain protein construct was produced by inserting the full-length RT1.Aa DNA(1.6 kb) from pBS3.3/1 plasmids (23) into theEco RI site of the homologous multiple cloning site of the pVL1393 transfer vector (Pharmingen, San Diego, CA; 24). This transfer vector was used because it bears a strong polyhedron promoter, and therefore facilitates a high level of protein expression(25, 26). The second RT1.Aa heavy chain protein construct was produced by attaching an ms signal(27) to the 5′ end of the RT1.Aa gene in the proper reading frame using polymerase chain reaction (PCR) amplification(28) with two primers that had been synthesized by Genosys (Woodlands, TX). The first primer (5′-AAG GAT CCA TGA AA TTC TTA GTC AA CGT TGC CCT TGT TTT TAT GGT CGT GTA CAT TTCTT ACA TCTA TGC GAT GGA GGC GAT GGC ACC GCG CACG-3′) included the 6-nucleotide sequence of theBam HI site, the 63 nucleotides of the ms signal, and the 24 nucleotides of the 5′ end of RT1.Aa. The second primer(5′-CCG GAA TTC TTT TTT TGG AAG CTT CTC CAT CTG-3′) included the 6-nucleotide EcoRI site and the 24 nucleotides of the 3′ end of RT1.Aa. PCR amplification in the 9600 GeneAmp system (Perkin Elmer, Norwalk, CT) included 20 cycles of 15 sec of denaturation at 94 °C, 30 sec of annealing at 59 °C, and 45 sec of extension at 72 °C, followed by 1.5 min of final extension at 72 °C. The amplified product was ligated into the pCRII vector in order to transform Escherichia coli-competent cells (Invitogen, San Diego, CA). Positive clones prepared using the Magic miniprep kit (Promega, Madison, WI) were identified by the presence of the proper-sized fragments after EcoRI andBam HI restriction enzyme digestion, and confirmed by DNA sequencing. The isolated, verified ms/RT1.Aa fragment was inserted into the pVL1393 transfer vector between the BamHI and EcoRI sites. The combination of transfer vector DNA (2 μl) and linearized wild-type (WT) baculovirus DNA (8 μl; Pharmingen) in Milli Q water (40 μl) was cotransfected with lipofectin (50 μl; Gibco BRL, Gaithersburg, MD) intoSpodoptera frugiperda (Sf9) cells in 750 μl of serum-free Hink's medium (JRH Biosciences, Lenexa, KS) during a 4-hr incubation at 28 °C on a rocking platform. After a 4-day culture with 10% FCS in Hink's medium, the supernatants were subjected to 3 rounds of plaque purification(19): 1 ml of 10-fold serial dilutions of viral stocks was inoculated into Sf9 cells (2×106) seeded in 6-well plates(Corning, NH) for 1 hr at 28 °C. Inocula were replaced with 3 ml of Seaplaque agarose (1.5%; FMC, Rockland, ME) mixed with 2× Hink's medium supplemented with 20% FCS. After a 4-day incubation at 28 °C, the layers were overlaid with a second layer of Seakem agarose (3.0%; FMC) containing 0.02% neutral red (Sigma Chemical Co., St. Louis, MO). Single plaques bearing either the RT1.Aa or the ms/RT1.Aa construct were selected for preparation of high titer viral stocks (>1×108 plaque-forming units/ml). To confirm the veracity of the DNA product, dot blot hybridization was performed with double-stranded RT1.Aa, or synthetic ms DNA fragments (250ng) that had been labeled by the random primer method using 5μl of 32P-dCTP (3000 Ci/mmol, Amersham, Arlington Heights, IL). Labeled DNA fragments were separately applied to duplicate nitrocellulose membranes (Schliecher and Schuell, Keene, NH) that had been infiltrated previously with 100 ng each of RT1.Aa or ms DNA (positive controls), or 200 μl of viral stock of WT- (negative control), RT1.Aa-, or ms/RT1.Aa-baculovirus.
Immunoprecipitation methods(29). Sf9 cells(6×106) infected with 10 MOI (multiplicity of infection) of WT-, RT1.Aa-, or ms/RT1.Aa-baculovirus were cultured in T25 flasks(Corning) in Hink's medium containing 10% FCS (JRH Biosciences) for 30 hr at 27 °C. On the following day, media were replaced with methionine-free Grace's medium for 1 hr and then with 2 ml of methionine-free Grace's medium supplemented with 35S-methionine (60 μl; ICN, Irvine CA) and 0.5% FCS for a 10-hr culture. Thereafter, supernatants of centrifugation(500×g for 10 min) were analyzed by immunoprecipitation. Adherent Sf9 cells were washed once with phosphate-buffered saline (PBS) (136 mM NaCl, 8 mM Na2HPO4, 1.5 mM KH2PO4, and 2.6 mM KCl; pH 7.2), lysed on ice for 30 min with 2 ml of RIPA buffer (150 mM NaCl, 1% sodium deoxycholate 1%Triton X-100, 0.1% sodium dodecyl sulfate [SDS], 1% Trasylo, and 10 mM Tris-HCl; pH 7.2), detached with a rubber policeman, and centrifuged (10,000×g for 30 min) to remove cell debris. Five replicate supernatants (100 μl) from cultures or from cell lysate were preabsorbed with 10 μl of Staph A cells (immunoprecipitation; Gibco BRL), centrifuged (13,000 rpm for 5 min), and transferred to tubes for overnight incubation (4 °C) with the individual mAb(22, 24), namely, F16-4-4, which recognizes an MHC rat class I monomorphic determinant, or YR1/100, R2/10P, R3/13, or R2/15S, each of which recognizes individual polymorphic RT1.Aa determinants(Bioproducts for Science, Indianapolis, IN). The following day, 50 μl ofStaph A cells were added to each tube for a 15-min incubation on ice. After centrifugation (13,000 rpm for 5 min), the supernatants were washed 3 times with buffer A (1 M NaCl, 0.1% NP40, and 10 mM Tris-HCl; pH 7.2), buffer B (100 mM NaCl, 1 mM EDTA, 0.1% NP40, 0.3% SDS, and Tris-HCl; pH 7.2), and buffer C (0.1% NP40 and 10 mM Tris-HCl; pH 7.2). Thereafter, the pellets were resuspended with 1× gel loading buffer (50 mM Tris-HCl, 10% glycerol, 0.03% phenol red, 500 mM urea, 5% 2-mercaptoethanol, and 1% SDS; pH 6.8), boiled at 100 °C for 5 min, centrifuged (13,000 rpm for 5 min), and loaded (25 μl) onto 10% SDS-polyacrylamide gel for electrophoresis at 20 mA for 4 hr. The gels were fixed, enhanced with Autofluor (National Diagnostics, Manville, NJ), dried, and exposed to XAR-2 film (Eastman Kodak, Rochester, NY) at -70 °C for 48 hr. Protein products were identified by their molecular sizes.
Affinity chromatography(30). Supernatants from ms/RT1.Aa-baculovirus-infected Sf9 cells were purified on a CN Br-activated Sepharose-4B (Pharmacia LKB, Piscataway, NJ) column (Biorad Corp, Hercules, CA) precoated with 5 mg/ml of 3 ml total protein of 1 (F16-4-4) or 2(MN4-91-6 and R3/13) anti-RT1Aa mAb (Bioproducts for Science). The column system included 3 precolumns of Sepharose covalently bound to bovine albumin (first column), anti-rat IgG2c Ab (second column), and anti-mouse IgG1 antibody (third column) to remove moieties with nonspecific binding. Each column was prewashed with 40 ml of Tris-buffered saline (pH 8.0) and then 20 ml of Tris-buffer (pH 11.3) before application of the sample in supernatant dialyzed against 0.01 M PBS (pH 7.2) for 6 hr. After stabilization with 20 ml of Tris-buffered saline (pH 8.0), the affinity column was washed with 10 ml of Tris-buffer (pH 8.0). Upon evaluation with 10 ml of triethanolamine buffer (pH 11.3), the RT1.Aa protein was eluted using 0.2-ml fractions. Positive fractions were concentrated using Centricon-10 microconcentrators (Amicon Corp., Lexington, MA). The immunoaffinity column purification yielded approximately 500 μg of RT1.Aa protein per liter of crude supernatant.
Protein blotting(31). Concentrated samples(20 μl) purified on affinity columns were mixed with equal volumes of loading buffer for electrophoresis on 10% SDS-polyacrylamide gels (20 mA for 4 hr) and transferred to nitrocellulose membranes. After electrophoresis at 600 mA for 2 hr, the membranes were incubated overnight in 10 ml of 0.1% bovine serum albumin (BSA)-PBS containing 50 μl of a mixture of rat anti-RT1.Aa polyclonal antibodies. After 3 washes with 0.1% BSA-PBS, the membranes were incubated with 10 ml of horseradish peroxidase-conjugated rabbit anti-rat IgG (H+L chains, 1:2000, Zymed Laboratories, San Francisco, CA) in 0.1% BSA-PBS (2 hr at room temperature). After being washed 3 times with 0.1% BSA-PBS, the membranes were developed with substrate reagent (30 mg of 4-choloro-1-naphthol [Sigma], in 10 ml of methanol, 50 ml of 0.01 M PBS [pH 7.4], and 25 μl of H2O2).
Animals. Adult male Wistar-Furth (WF; RT1u), ACI(RT1a), Lewis (LEW; RT1l), Brown Norway (BN; RT1n), PVG.1U (RT1u), and PVG.R8 (RT1.AaBuDuCu) rats weighing 180-250 g were obtained from Harlan Sprague-Dawley(Indianapolis, IN) and cared for under the treatment guidelines of the institutional Animal Walfare Committee. Rats were housed in light- and temperature-controlled quarters, and received food and water ad libitum. All operations were performed under aseptic conditions; each animal's postoperative condition was monitored daily.
Cardiac transplantation. Heterotopic heart transplantation utilized a standard microsurgical technique with end-to-side anastomoses to recipient aorta and vena cava (32). The cold ischemia times were less than 30 min. Based upon daily examinations performed independently by two investigators, graft survival was defined as the last day of transabdominally palpable cardiac contractions. The results are presented as mean survival times (MST) ± SD; P-values calculated by the Breslow-Gehan-Wilcoxon method were considered significant if <0.05.
Liver transplantation. Liver transplantation was performed using the method of Lee et al. (33). The portal vein, which had been carefully dissected to preserve a long segment, and the inferior vena cava were anastomosed with cuffs to the recipient's vessels; the superior vena cava was anastomosed with 6-0 monofilament suture (Ethicon). The recipients were given intravenous infusions of 5 ml of warm saline at the time of operation. The animals that died within 4 days after grafting (10%) were excluded from the analysis.
Immunization. Culture supernatant or cell lysate was prepared from cultures of Sf9 cells 4 days after infection with WT-, RT1.Aa-, or ms/RT1.Aa-baculovirus. Four days before transplantation of ACI or third-party BN heart allografts to WF rats or PVG.R8 heart allografts to PVG.1U rats, recipients were given subcutaneous injections at between 2 and 4 sites with either unconcentrated (10 ml/rat) or concentrated (1 ml/rat) culture supernatants (20 mg/kg total protein), or with cell lysate (1 ml; equivalent of 15×106 cells/rat). In addition, WF rats either were not immunized (negative controls) or received injections of 1, 5, or 10×106 ACI lymph node (LN) cells or ACI hepatocytes (positive controls) 4 days before transplantation of ACI heart grafts. In other experiments, PVG.1U rats were injected with PVG.1U peritoneal macrophages(50×106 cells/rat) that had been cultured overnight with soluble RT1.Aa protein. In a tolerance model, LEW rats were injected via either the thymus or the portal vein with 8.4 mg (100 μl) of RT1.Aa protein on day -14 either alone or in combination with 2 doses of anti-TCR R73 mAb(obtained from Dr. Kurrle Behringwerke, Marburg, Germany) injected intravenously on days -13 and -14 prior to ACI liver transplantation.
Preparation of macrophages. Thirty minutes after intraperitoneal injection with 10 ml of RPMI 1640 medium, PVG.1U rats were gently massaged to collect peritoneal exudate cells that were then cultured in 10 ml of RPMI 1640 supplemented with 2 mM L-glutamine, 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 10 mM HEPES, 5×10-5 M 2-mercaptoethanol, 100 μ/ml penicillin, 100 μg/ml streptomycin (Gibco BRL, Grand Island, NY) and 10% heat-inactivated FCS (Hyclone, Logan, UT) overnight in the presence of soluble RT1.Aa protein (10 μg/ml). Thereafter, the cells were detached with a rubber policeman and washed 3 times with fresh RPMI before they were injected subcutaneously (50×106 cells/rat) (total volume of 1 ml) into rats. In addition, a control group received macrophages that had been precultured for 15 min with chloroquine (100 μg/ml) in order to block the uptake and processing of RT1.Aa protein. Four days later, PVG.1U rats received PVG.R8 heart allograft transplants.
Preparation of hepatocytes. Hepatocytes were prepared using an in situ collagenase digestion method (34). PVG.R8 livers were first perfused with 0.9% sodium chloride (Baxter, Deerfield, IL) and then with 0.05% collagenase (type IV; Sigma). Hepatocytes were washed 3 times with HBSS (Gibco BRL, Grand Island) supplemented with 1% FCS (Gibco), loaded onto 10-50% Percoll gradients (Pharmacia), and centrifuged at 3000 rpm for 20 min. Purified hepatocytes obtained at the interface between 20-30% Percoll were washed 3 times with HBSS (1% FCS) and then treated with Pronase E type XXV(Sigma) at 37 °C for 10 min.
Limiting dilution analysis. The methods for limiting dilution analysis and its statistical evaluation have been described previously in detail (15, 17). In brief, 24 replicates of 6 dilutions of responder cells (12800-100) were dispensed into 96-well round-bottom microtiter plates. Each well contained 5×104 irradiated (20 Gray; 60Co source, JL Shepard) splenic stimulator cells. The complete medium included RPMI 1640, 2 mM L-glutamine, 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 10 mM HEPES solution, 5×10-5 M 2-mercaptoethanol, 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco BRL, Grand Island), supplemented with 10% heat-inactivated FCS (Hyclone, Logan, UT). Ten units/milliliter of purified rat IL-2(Collaborative Research, Inc., Bedford, MA) served as exogenous growth factor. After 8 days of incubation at 37 °C in 5% CO2, the cytotoxic activity of each well was assessed during an additional 6-hr incubation at 37°C with 1×10451Cr-labeled (Na2[51Cr]04, sp. act. 200-500 mCi/mg Cr; Amersham) RT1a ACI splenocytes. 51Cr release from target cells into supernatant fluids was assessed after harvesting by 51Cr gamma-counting using a cartridge collection system (Skatron, Sterling, VA). Positive wells were identified by values that exceeded the mean spontaneous release by more than 3 SDs.
Production of RT1.Aa- and ms/RT1.Aa-baculovirus. Two RT1.Aa heavy chain protein constructs were produced in the baculovirus expression system. In the first construct, the full-length RT1.Aa gene, includingα1, α2, and α3 extracellular, transmembrane, and cytoplasmic regions (23), was attached at the N terminus by homologous recombination to the polyhedron promoter (26) in order to enhance RT1.Aa gene expression by baculovirus (Fig. 1A). In the second construct, the honeybee ms signal (27) was attached between the leader sequence for the RT1.Aa gene and the polyhedron promoter, using PCR and specific primers (Fig. 1B). The ms signal encodes a 21-amino acid peptide(Met-Lys-Phe-Leu-Val-Asn-Val-Ala-Leu-Val-Phe-Met-Val-Val-Tyr-Ile-Ser-Tyr-Ile-Tyr-Ala) that enhances the production and secretion of proteins in insect cells. The RT1.Aa or ms/RT1.Aa construct within pVL1393 plasmid DNA was inserted into baculovirus by homologous recombination with linearized WT viral DNA during simultaneous cotransfection into Sf9 insect cells. Dot blot hybridization with 32P-RT1.Aa or 32P-ms DNA probes confirmed the presence of RT1.Aa or of ms DNA fragments in high titer viral stocks (Fig. 1C).
Production of RT1.Aaprotein by baculovirus / Sf9 cell system. RT1.Aa protein was produced by infecting log-phase cultures of Sf9 insect cells, which had been conditioned to grow in FCS-free medium, with RT1.Aa- or ms/RT1.Aa-baculovirus. Protein fractions purified from culture supernatant (Fig. 2A) or from cell lysate (Fig. 2B) were immunoprecipitated with mAbs that recognize either monomorphic class I (F16-4-4) or distinctive polymorphic RT1.Aa (YR1/100, R2/10P, R3/13, and R2/15S) determinants. All 5 mAbs detected RT1.Aa proteins only in the culture supernatant from ms/RT1.Aa-Sf9 cells, thereby documenting that the ms signal induced secretion of RT1.Aa proteins (Fig. 2A, top panel). Although the monomorphic F16-4-4 mAb immunoprecipitated only RT1.Aa protein, the polymorphic YR1/100, R2/10P, R3/13, and R2/15S mAbs immunoprecipitated RT1.Aa protein and cross-reacted with a 32-kDa protein present in a supernatant from Sf9 cells infected with WT-, RT1.Aa-, or ms/RT1.Aa-baculovirus. The ≈50-kDa band present in the ms/RT1.Aa cell lysate line may represent a glycosylated form of ms/RT1.Aa heavy chain protein (Fig. 2B). Image analysis of immunoprecipitation bands revealed that supernatants from ms/RT1.Aa-Sf9 cells displayed 20- to 43-fold higher concentrations of RT1.Aa proteins than those from RT1.Aa-Sf9 cells infected with baculovirus (Fig. 2A, bottom panel). Cell lysate after infection with ms/RT1.Aa- or RT1.Aa-baculovirus displayed 46-kDa bands bearing RT1.Aa epitopes (Fig. 2B, top panel). Interestingly, densitometric analysis showed that the ms signal increases the production of class I MHC proteins by insect cells by 7 to 17-fold (Fig. 2B, bottom panel). Detected products included only RT1.Aa heavy chains because β2-m is not produced by Sf9 cells. To estimate the kinetics of RT1.Aa production and secretion, supernatant RT1.Aa concentrations were measured each day after infection; the maximal quantities of purified RT1.Aa protein were detected at day 4 after infection of Sf9 cells with ms/RT1.Aa-baculovirus (not shown).
In vivo activity of the RT1.Aaheavy chain protein. The immunogenicity of RT1.Aa heavy chain proteins, from either the supernatant or cell lysate of RT1.Aa-Sf9 or ms/RT1.Aa-Sf9 cells, was tested by subcutaneous immunization of WF(RT1u) rats with each preparation 4 days before transplantation with ACI (RT1a) heart allografts that expressed RT1.Aa/β2-m/peptide molecules. Unsensitized WF recipients reject ACI heart allografts at an MST of 5.9±0.5 days. Lysate of cells infected with baculovirus constructs induced accelerated rejection of ACI, but not BN, heart allografts that bore RT1.An/β2-m/peptide molecules (Table 1). In contrast, only the supernatant from cell cultures transfected with the ms/RT1.Aa construct produced RT1.Aa-specific sensitization (Table 1). Similar results were obtained when WF recipients were immunized 7 days or both 7 and 4 days before ACI heart transplantation (not shown). In preliminary experiments, the secreted RT1.Aa protein was purified by immunoaffinity separation, using an anti-RT1.Aa-specific mAb (see Materials and Methods). The purified RT1.Aa protein formed a single 46-kDa band and, when injected subcutaneously at doses of 50 ng or 500 ng, it accelerated rejection of ACI heart allografts from 5.9±0.5 days in controls to 4.3±0.5 days and 3.8±0.5 days, respectively (bothP<0.01). Figure 3 shows the dose-dependent results of immunization with ACI LN cells, ACI hepatocytes, as well as cell lysate or supernatant from RT1.Aa-Sf9 or ms/RT1.Aa-Sf9 cells. Although the cell lysate from RT1.Aa-Sf9 or ms/RT1.Aa-Sf9 cells was as immunogenic as ACI hepatocytes bearing RT1.Aa/β2-m/peptide molecules, all of the cell lysate preparations were immunogenically inferior compared with the cell lysate of ACI LN cells, which express both class I and II MHC alloantigens. In contrast, supernatant from ms/RT1.Aa-Sf9 cells, but not from RT1.Aa-Sf9 cells, produced dose-dependent immunization akin to that produced by the cell lysate (Fig. 3). Thus, soluble RT1.Aa heavy chain proteins produced by the baculovirus expression system display the immunogenic specificity of RT1.Aa/β2-m/peptide alloantigens expressed on ACI hepatocytes.
To extend graft survival in control nonsensitized hosts, WF recipients were treated with CsA (10 mg/kg) daily for 5 days. In these hosts, ACI heart allografts were rejected at 12.2±0.8 days. Immunization of WF recipients with cell lysate from ACI hepatocytes accelerated the rejection of ACI hearts to 8.2±2.2 days (P<0.01). Similarly, lysate of RT1.Aa-Sf9 or ms/RT1.Aa-Sf9 cells shortened the survival of ACI hearts to 6.7±1.0 days or 6.7±0.5 days, respectively(P<0.01; Table 1).
These in vivo results were supported by in vitro analyses of the frequency of anti-RT1.Aa-specific CTL (fCTL). LN cells derived from WF hosts that had received subcutaneous injections of ACI LN cells (4 days earlier) displayed a 10-fold increase in the fCTL (1:2554) in comparison to WF rats that had been injected with noninfected Sf9 cells (1:26681) or with WT baculovirus (1:20935; Fig. 4A). Immunization with cell lysate of ACI hepatocytes increased the anti-RT1.Aa fCTL to 1:4710. Similarly, immunization with lysate from RT1.Aa-Sf9 or ms/RT1.Aa-Sf9 cells increased the fCTL 6-fold to 1:4395 or 1:4310, respectively. However, the supernatant from ms/RT1.Aa-Sf9 cells increased the fCTL to 1:5299; the supernatant from RT1.Aa-Sf9 cells was ineffective (1:24226; Fig. 4B). Therefore, the RT1.Aa heavy chain protein generated in the baculovirus-Sf9 cell system was as effective as native ACI hepatocytes that bear RT1.Aa/β2-m/peptide alloantigens.
Indirect presentation of RT1.Aaproteins. The immunogenicity of RT1.Aa proteins was tested in congenic rats that are disparate only at the RT1.Aa locus. Immunization of PVG.1U hosts with RT1.Aa-Sf9 or ms/RT1.Aa-Sf9 cell lysates, or with supernatant from ms/RT-1Aa-Sf9 cells, induced accelerated rejection of PVG.R8 heart allografts, namely from 6.3±0.5 days in untreated controls to 3.8±0.5 days, 3.5±0.6 days, or 4.0±0.0 days, respectively(all P<0.01; Table 2). Furthermore, PVG.1U peritoneal macrophages cultured overnight with secreted RT1.Aa protein sensitized PVG.1U recipients toward PVG.R8 hearts (4.0±0.0 days;P<0.01), thereby documenting the biologic activity of indirectly presented RT1.Aa heavy chain proteins. Similar results were produced by indirect presentation of RT1.Aa alloantigens as cell lysate prepared from PVG.R8 hepatocytes or LN cells (Table 2). The process of indirect presentation requires active endocytosis, not merely surface binding of RT1.Aa alloantigens; chloroquine treatment of macrophages abrogated the RT1.Aa processing and presentation in PVG.1U recipients (6.0±0.5 days; NS). In addition, naive PVG.1U recipients treated for 3 days with CsA (10 mg/kg) rejected PVG.R8 heart allografts at 10.0±1.0 days. Indirect presentation of secreted RT1.Aa protein displayed accelerated rejection of PVG.R8 heart grafts, namely an MST of 5.8±1.0 days (P<0.01). Similar results were obtained after indirect presentation of cell lysate from PVG.R8 hepatocytes expressing the RT1.Aa/β2-m/peptide alloantigens (6.0±1.2 days;P<0.01). These data demonstrate that indirect presentation of RT1.Aa heavy chain proteins induces specific immunization toward RT1.Aa/β2-m/peptide molecules expressed on allografts.
Induction of tolerance by RT1.Aaprotein. Our preliminary experiments showed that RT1.Aa heavy chain proteins injected intrathymically into WF rats (day -14) in combination with R73 mAb(days -14 and -13) failed to prolong the survival of ACI heart allografts(5.8±0.4 days [n=6] in controls vs. 7.7±0.6 days [n=4]; NS). However, recently published data revealed that DA (RT1a) bone marrow cells (with identical RT1a MHC as ACI rats) injected into WF rats (day-14) in combination with antilymphocyte serum (ALS; day -14) did not affect the survival of DA heart allografts (35). Furthermore, normal WF recipients spontaneously accepted ACI liver allografts (>100 days; n=4), which precluded this strain combination from being used in tolerance induction studies. In contrast, untreated LEW (RT1l) recipients reject ACI liver allografts at 10.3±1.8 days(Table 3). Thus, tolerance induction was tested in ACI to LEW liver transplant model following intrathymic or intraportal injection of RT1.Aa heavy chain proteins (day -14) in combination with 2 doses of anti-TCR R73 mAb (days -13 and -14). Injection of RT1.Aa protein alone either intrathymically or intraportally was not effective (9.3±1.5 days and 10.7±1.5 days, respectively; NS). Hosts treated only with R73 mAb displayed modestly prolonged liver allograft survival, namely an MST of 15.3±7.6 days, with 2/9 hosts surviving more than 300 days. In contrast, the combination of secreted RT1.Aa protein injected intrathymically or intraportally with R73 mAb induced indefinite survival of ACI liver allografts in all LEW recipients (>250 days; P<0.01;Table 3). Thus, RT1.Aa heavy chains (produced in the baculovirus-Sf9 cell system) bear epitopes that induce immunity or tolerance to organ allografts depending upon the site and timing of pretreatment.
The studies presented herein showed that the honeybee ms signal significantly improved the production (7- to 17-fold) and most importantly the secretion (20- to 43-fold) of the RT1.Aa heavy chain protein in a baculovirus/Sf9 cell system. Several in vivo observations suggested that the injection of the RT1.Aa heavy chain protein alone, without theβ2-m light chain protein and self-peptides, was specifically immunogenic: first, the extracted and secreted RT1.Aa proteins displayed reactivity with a panel of specific mAb; second, RT1.Aa heavy chain proteins either secreted by or extracted from Sf9 cells infected with ms/RT1.Aa-baculovirus induced accelerated rejection of ACI or PVG.R8 heart allografts; third, immunization with soluble RT1.Aa heavy chain proteins increased the frequency of RT1.Aa-specific CTL; and fourth, soluble RT1.Aa heavy chain proteins injected intrathymically or intraportally induced tolerance to ACI liver allografts. Thus, soluble class I MHC RT1.Aa heavy chain proteins produced similar results in vivo to those produced by intact RT1.Aa/β2-m/peptide alloantigens expressed on cells. However, we could not exclude the possibility that immediately after injection, the RT1.Aa heavy chain protein bind circulating β2-m, thereby forming RT1.Aa/β2-m dimers.
Indirect presentation of RT1.Aa heavy chain proteins produced specific immunization. The RT1.Aa heavy chain proteins were processed endosomally and presented by APC in conjunction with host MHC molecules. Syngeneic macrophages cultured overnight with RT1.Aa heavy chain proteins and injected into PVG.1U rats induced accelerated rejection of PVG.R8 heart allografts. Chloroquine treatment of macrophages inhibited RT1.Aa heavy chain protein processing and presentation by APC. Almost identical results were obtained by indirect presentation of extracts from PVG.R8 hepatocytes. Thus, immunization with the RT1.Aa heavy chain protein alone induced identical results as indirect presentation of RT1.Aa/β2-m/peptide alloantigens.
The present findings demonstrate that indirect presentation of RT1.Aa heavy chain proteins produced potent immunization compared with previously published results of subcutaneous immunization with synthetic peptides that correspond to the polymorphic sequences of RT1.Aa molecules (19, 20, 36). In rats that received 2 subcutaneous injections of a mixture of CFA with 3 short peptides(bearing polymorphic sequences of RT1.Aa), purified CD4+ T cells displayed strong proliferative responses to a single peptide derived from theα1 helical region (amino acids 57-80). The same CD4+ T cells failed to respond to a peptide either from the β-sheet of theα1 domain (amino acids 99-117) or from the α2 helical region (amino acids 143-164). However, 2 immunizations with the same peptides and CFA induced only marginal acceleration of rejection of skin or kidney allografts bearing RT1.Aa/β2-m/peptide molecules(19, 20). Similar results were recently reported for heart allografts (36). Immunization with a mixture of CFA and 2 peptides from the α1 helical region of the RT1.Aa molecule (amino acids 56-72 and amino acids 73-83) at days 24 and 7 prior to PVG.R8 heart allograft transplantation into PVG.1U recipients slightly accelerated the rejection of heart allografts (6.0±1.0 days in controls vs. 5.0±0.5 days; P<0.2). Thus, although synthetic peptides representing the α1 helical regions of RT1.Aa molecules specifically immunized recipients, as shown in in vitro experiments, the in vivo effect was weak. In contrast, the RT1.Aa heavy chain protein or extracts from PVG.R8 hepatocytes (bearing RT1.Aa/β2-m/peptide) produced similarly potent alloantigen-specific immunization after indirect presentation by APC.
Like other peptides that bind to self MHC, the RT1.Aa allopeptides must bear a peptide motif (usually 1 or 2 amino acids) to bind effectively to class I or II MHC before transport to the cell membrane(37). High resolution x-ray crystallography showed that class I MHC/β2-m dimers bind small (8-10 amino acids) peptides to the peptide binding cleft formed by the α1/α2 domains, whereas class II molecules bind larger (12-24 amino acids) peptides to the α/β cleft (37-39). Therefore, it is possible that randomly selected RT1.Aa synthetic peptides are different than those presented after processing of RT1.Aa protein by APC.
Furthermore, the present experiments showed that RT1.Aa heavy chain proteins injected intrathymically or intraportally in combination with anti-TCR R73 mAb induced indefinite survival of ACI liver allografts in LEW recipients. Prolonged survival of organ allografts in rodents may be induced by administration of donor blood (40, 41), purified cells (42, 43), or extracted histocompatibility antigens (44-46) or peptides matching class I or class II MHC sequences(21, 22) alone or in combination with immunosuppressive modalities. Experiments using intact cells have documented the importance of the route of donor alloantigen delivery, namely via intravenous (47), intraportal(48, 49), or intrathymic(50, 51) injection, for the induction of unresponsiveness. Prolonged survival of heart allografts has been achieved by pretransplant portal vein administration of donor spleen cells alone at 7 or 14 days, but not 1 day, before transplantation (52, 53). However, the most long-lasting tolerance was induced across major histocompatibility barriers by direct intrathymic injection of donor bone marrow (54), pancreatic islets(50), isolated glomeruli (55), UV-B-irradiated spleen (56), or muscle cells transfected with donor class I MHC cDNA (57) 7-21 days before grafting, in combination with ALS, sublethal total body irradiation, or CsA-dexamethasone immunosuppression. Tolerant animals display specific clonal deletion of alloreactive T cells (51), and only modest negative regulatory cell activity (54). The present experiments documented that RT1.A heavy chain proteins can produce both potent sensitization and durable tolerance to organ allografts.
Recent studies documented that the baculovirus/Sf9 cell system may be used for the production of recombinant MHC molecules, including mouseβ2-m molecules (58), mouse class I H2Kd heavy chain molecules (59), H2Kd extracellular domains (α1, α2, and α3) fused to β2-m (60), and human class II HLA-DR1 heterodimers (after coinfection with baculovirus carrying truncated DR1α or DR1β genes; 61). Two of these constructs (β2-m and H2Kd heavy chain molecules) produced only small quantities of soluble proteins. Interestingly, H2Kd/β2-m and HLA-DR1 heterodimers, produced by the baculovirus system, displayed an “empty” peptide binding cleft(60, 61). In contrast, fused H2Kd/β2-m and truncated HLA-DR1 heterodimers were efficiently secreted into the supernatant. The present study shows that the ms signal increased the production and secretion of a soluble heavy chain class I MHC protein. Following binding of antigenic peptides, both class I(HKd/β2-m) and II (HLA-DR1) MHC molecules were stabilized, thereby preventing their aggregation and denaturation. Thus, the baculovirus/Sf9 insect cell system permits production of large quantities of MHC proteins to facilitate analysis of structure-function relationships. Indirect presentation of RT1.Aa heavy chain proteins following subcutaneous administration and intrathymic or portal vein administration with R73 mAb induces sensitization or tolerance to native RT1.Aa/β2-m/peptide heterodimers, respectively.
Acknowledgments. We thank Jennifer Rosen-Senft for editorial assistance in the preparation of the manuscript.
1. Amos DB, Gorer PA, Mikulska ZB. An analysis of an antigenic system in the mouse (the H-2 system). Proc R Soc Lond 1995; 144: 369.
2. Ascher NL, Hoffman RA, Hanto DW, Simmons RL. Cellular basis of allograft rejection. Immunol Rev 1984; 77: 217.
3. Jooste SV, Winn HJ. Acute destruction of rat skin grafts by alloantisera. J Immunol 1975; 11: 385.
4. Sherman LA, Chattopadhyay S. The molecular basis of allorecognition. Annu Rev Immunol 1993; 11: 403.
5. Germain RN. The biochemistry and cell biology of antigen processing and presentation. Annu Rev Immunol 1993; 11: 403.
6. Bjorkman PJ, Parham P. Structure, function, and diversity of class I major histocompatibility complex molecules. Annu Rev Biochem 1990; 59: 253.
7. Fremont DH, Matsmura M, Stura EA, Peterson PA, Wilson IA. Crystal structures of two viral peptides in complex with murine MHC H-Kb
[see comments]. Science 1992; 257: 919.
8. Rammenesee HG, Folk K, Rotzschke O. Peptides naturally presented by MHC class I molecules. Annu Rev Immunol 1993; 11: 213.
9. Koller BH, Marrack P, Kappler JW, Smithies O. Normal development of mice deficient in β2
m MHC class I proteins, and CD8+ T cells. Science 1990; 148: 1227.
10. Williams DB, Barber BH, Flavell RA, Allen H. Role of beta 2-microglobulin in the intracellular transport and surface expression of murine class I histocompatibility molecules. J Immunol 1989; 142: 2796.
11. Doxiadis I, Westhoff U, Grosse-Wilde H. Quantification of soluble HLA class I gene products by an enzyme immunoadsorbent assay. Blut 1989; 59: 449.
12. Zavazava N, Bottcher H, Muller-Ruchholtz W. Soluble MHC class I antigens (sHLA) and anti-HLA antibodies in heart and kidney allograft recipients. Tissue Antigens 1993; 42: 20.
13. Tsuji K, Kusama M, Nakatsuji T, Kato S. Biological significance of Ss (serum soluble) HLA class I antigens in bone marrow transplantation. Tokai J Exp Clin Med 1985; 10: 169.
14. De Vito-Haynes L, Jankowska-Gan E, Sollinger HW, Knechtle SJ, Burlingham WJ. Monitoring of kidney and simultaneous pancreas-kidney transplantation rejection by release of donor-specific, soluble HLA class I. Hum Immunol 1994; 40: 191.
15. Stepkowski SM, Ito T. Frequency of alloantigen-specific T cytotoxic cells in high- and low-responder recipients of class I MHC disparate heart allografts. Transplantation 1990; 50: 112.
16. Gracie JA, Bolton EM, Poteous C, Bradley JA. T cell requirements for the rejection of renal allografts bearing an isolated class I MHC disparity. J Exp Med 1990; 172: 1547.
17. Ito T, Stepkowski SM, Shirakura R, et al. Role of interleukin-2 producing CD4+
T cells in acute rejection of class I major histocompatibility complex disparate rat heart allografts. Transplant Proc 1993; 25: 885.
18. Spencer SC, Fabre JW. Bulk purification of a naturally occurring soluble form of RT1-A class I major histocompatibility complex antigens from DA rat liver, and studies of specific immunosuppression. Transplantation 1987; 44: 141.
19. Fangmann J, Dalchau R, Fabre JW. Rejection of skin allografts by indirect allorecognition of donor class I major histocompatibility complex peptides. J Exp Med 1992; 175: 1521.
20. Fangmann J, Dalchau R, Sawyer GJ, Priestley CA, Fabre JW. T cell recognition of donor major histocompatibility complex class I peptide during allograft rejection. Eur J Immunol 1992; 22: 1525.
21. Nisco S, Vriens P, Hoyt G, et al. Induction of allograft tolerance in rats by an HLA class I -derived peptide and cyclosporine A. J Immunol 1994; 152: 3786.
22. Seyegh MH, Perico N, Gallon L, et al. Mechanism of acquired thymic unresponsiveness to renal allografts. Transplantation 1994; 58: 125.
23. Rada C, Lorenzi R, Powis SJ, van den Bogaerde J, Parham P, Howard JC. Concerted evolution of class I genes in the major histocompatibility complex of murine rodents. Proc Natl Acad Sci USA 1990; 87: 2167.
24. Sambrook J, Fritsch EF, Maniatis T. Essential features of plasmids. In: Sambrook J, Fritsch EF, Maniatis T, eds. Molecular Cloning, 2nd ed. Cold Harbor Spring, NY: Cold Spring Harbor Laboratory Press, 1989: 1.
25. Luckow VA, Summers MD. Trends in the development of baculovirus expression vectors. Bio/Technology 1988; 6: 47.
26. Luckow VA, MD. Summers. High level expression of nonfused foreign genes with sutographa californica nuclear polyhedrosis virus expression vectors. Virology 1989; 170: 31.
27. Tessier DC, Thomas DY, Khouri HE, Laliberte F, Vernet T. Enhanced secretion from insect cells of a foreign protein fused to the honeybee melittin signal peptide. Gene 1991; 98: 177.
28. Frohman MA, Dush MK, Martin GR. Rapid production of full-length cDNAs from rare transcripts: amplification using a single gene specific oligonucleotide primer. Proc Natl Acad Sci USA 1988; 85: 8998.
29. Sambrook J, Fritsch EF, Maniatis T. Immunological screening of expression libraries: essential features of plasmids. In: Sambrook J, Fritsch EF, Maniatis T, eds. Molecular Cloning, 2nd ed. Cold Harbor Spring, NY: Cold Spring Harbor Laboratory Press, 1989: 2.
30. Spencer SC, Fabre JW. Water soluble form of RT-1A class I MHC molecules in the kidney and liver of the rat. Immunogenetics 1987; 25: 91.
31. Dreyfuss G, Adam SA, Choi YD. Physical change in cytoplasmic nucleoproteins in cells treated with inhibitors of mRNA transcription. Mol Cell Biol 1984; 4: 415.
32. Ono K, Lindsey E. Improved technique for heart transplantation in rats. J Thorac Cardiovasc Surg 1969; 57: 225.
33. Lee S, Charter AC, Chandler JG, Orloff MJ. A technique for orthotopic liver transplantation in the rat. Transplantation 1965; 16: 664.
34. Berry MN, Friend DS. High-yield preparation of isolated rat liver parenchymal cells. J Cell Biol 1969; 45: 506.
35. Alfrey EJ, Wang X, Lee L, et al. Tolerance induced by direct inoculation of donor antigen into the thymus in low and high responder rodents. Transplantation 1995; 59: 1171.
36. Shirwan H, Leamer M, Wang HK, Makowka L, Cramer DV. Peptides derived from α-helices of allogeneic class I major histocompatibility complex antigens are potent inducers of CD4+ and CD8+ T cell and B cell responses after cardiac allograft rejection. Transplantation 1995; 59: 401.
37. Rotzschke O, Falk K, Deres K, et al. Isolation and analysis of naturally processed viral peptides as recognized by cytotoxic T cells [see comments]. Nature 1990; 348: 252.
38. Bjorkman PJ, Saper MA, Samraoui B, Bennett WS, Strominger JL, Wiley DC. Structure of the human class I histocompatibility antigen, HLA-A2. Nature 1987; 329: 506.
39. Zhang W, Young AC, Imari M, Nathenson SG, Sacchettini JC. Crystal structure of the major histocompatibility complex class I H-Kb
molecule containing a single viral peptide: implications for peptide binding and T-cell receptor recognition. Proc Natl Acad Sci USA 1992; 89: 8403.
40. Perloff LJ, Baker CF. Variable response to donor-specific blood transfusion in the rat. Transplantation 1994; 38: 178.
41. Wasowska B, Baldwin WM, Howell DN, Sanfilippo F. The effects of donor-specific blood transfusion enhancement of rat renal allograft on cytotoxic activity and phenotypes of peripheral blood lymphocytes, splenocytes, and graft-infiltrating cells. Transplantation 1991; 51: 451.
42. Oluwole SF, Ng AK, Reemtsra K, Hardy M. The mechanism of the induction of immunologic unresponsiveness to rat cardiac allograft by recipient pretreatment with donor lymphocyte subsets. Transplantation 1991; 48: 281.
43. Wood ML, Monaco AP, Gottschalk R. Characterization of spleen cells capable of inducing unresponsiveness in ALS-treated mice. Transplantation 1991; 51: 208.
44. Medawar PB. The use of antigenic tissue to weaken the immunological reaction against skin homografting mice. Transplantation 1963; 1: 21.
45. Myburgh JA, Smith JA. Prolongation of liver allograft survival by donor-specific soluble transplantation antigens and antigen-antibody complexes. Transplantation 1975; 19: 64.
46. Yasumura T, Kahan BD. Prolongation of rat kidney allograft by pretransplant administration of donor antigen extract or whole blood transfusion combined with a short course of cyclosporine. Transplantation 1983; 36: 603.
47. Wood KJ, Evins J, Morris PJ. Suppression of renal allograft rejection in the rat by class I antigens on purified erythrocytes. Transplantation 1985; 39: 56.
48. Qian JH, Hashimoto T, Fujiwara H, Hamaoka T. Studies on induction of tolerance to alloantigens. I. The abrogation of potential for delayed type hypersensitivity response to alloantigens by portal venous inoculation with allogeneic cells. J Immunol 1985; 34: 3656.
49. Rao VK, Burris DE, Gruel SM, Sollinger HW, Burlingham WJ. Evidence that donor spleen cells administered through the portal vein prolong the survival of cardiac allograft in rats. Transplantation 1988; 45: 1145.
50. Posselt AM, Barker CF, Tomaszewski JE, Markmann JF, Choti MA, Naji A. Induction of donor-specific unresponsiveness by intrathymic islet transplantation. Science 1990; 249: 1293.
51. Markmann JF, Odorico JS, Bassiri H, Desai N, Kim JI, Barker CF. Deletion of donor-reactive T lymphocytes in adult mice after intrathymic inoculation with lymphoid cells. Transplantation 1993; 55: 871.
52. Kenick S, Lisbona R, Marghesco D, Lowry RP. Prolonged cardiac allograft survival following portal venous inoculation of allogeneic cells: immunologically specific entrapment of allogeneic cells within the liver. Transplant Proc 1987; 19: 3057.
53. Nakano Y, Monden M, Valdivia LA, Gotoh M, Tono T, Mori T. Permanent acceptance of liver allograft by intraportal injection of donor spleen cells in rats. Surgery 1992; 111: 668.
54. Odorico JS, Barker CF, Posselt AM, Naji A. Induction of donor-specific tolerance to rat cardiac allografts by intrathymic inoculation of bone marrow. Surgery 1992; 112: 370.
55. Remuzzi G, Rossini M, Imberti O, Perico N. Kidney graft survival in rats without immunosuppressants after intrathymic glomeruli transplantation. Lancet 1991; 337: 750.
56. Oluwole SF, Chowdhury NC, Fawwaz RA. Induction of donor-specific unresponsiveness to rat cardiac allografts by intrathymic of UV-B irradiate donor spleen cells. Transplantation 1993; 55: 1389.
57. Knechtle SJ, Wang J, Jiao S, Geissler EK, Sumimoto R, Wolff J. Induction of specific tolerance by intrathymic injection of recipient muscle cells transfected with donor class I major histocompatibility complex. Transplantation 1994; 57: 990.
58. Godeau F, Casanova JL, Fairchild KD, et al. Expression and characterization of recombinant mouse β2
-microglobulin type A in insect cells infected with recombinant baculovirus. Res Immunol 1992; 142: 409.
59. Godeau F, Casanova JL, Lueescher IF, et al. Binding of low concentration of peptide to H-Kd
produced in insect cells requires mouse β2
-microglobulin co-expression. Int Immunol 1992; 4: 265.
60. Godeau F, Luescher IF, Ojcius DM, et al. Purification and ligand binding of soluble class I major histocompatibility complex molecule consisting of the first three domains of H-Kd
-microglobulin expressed in the baculovirus-insect cell system. J Biol Chem 1992; 34: 24223.
61. Stern LJ, Wiley DC. The human class II MHC protein HLA-DR1 assembles as empty α/β heterodimers in the absence of antigenic peptide. Cell 1992; 68: 465.