MHC Class II-Transfected Tumor Cells Directly Present Antigen to Tumor-Specific CD4⁺ T Lymphocytes¹

Mice

C57BL/6, A/J, and (C57BL/6 × A/J) ((B6 × A/J)F₁) mice were purchased from The Jackson Laboratory (Bar Harbor, ME) or bred in the University of Maryland Baltimore County animal facility. Bone marrow donors and recipients were female mice, and chimeras were generated as previously described (15). Briefly, recipient (B6 × A/J)F₁ mice were maintained on tetracycline water (0.2%) for 1 wk before and 5 wk after reconstitution and were given gentamicin sulfate s.c. (500 μg) for 7 consecutive days beginning 1 day before irradiation/reconstitution. Approximately 24 h before irradiation/reconstitution, recipients were taken off food. Recipients were irradiated with 875 rad total body irradiation using a ¹³⁷Cs source (Kewaunee Scientific, Statesville, NC) and reconstituted i.v. with one femur-equivalent of donor bone marrow within 2 to 3 h of irradiation. Chimeras were maintained in a pathogen-free environment for 6 wk before use. All chimeras were tested by indirect immunofluorescence to ascertain bone marrow genotype and assure chimeric status as follows. Concurrent with the in vitro APC assays, spleens of chimeras were removed and stained for MHC class I (H-2K^k for A/J (mAb 16-3-1) (16); H-2K^bD^b for C57BL/6 (mAb 28-13-3) (17)) Ags. Positively staining cells were gated relative to conjugate alone controls, and positive cells were compared with wild-type A/J and C57BL/6 splenocytes stained under identical conditions. The percentage of the donor phenotype was calculated by comparing the percentage of positive chimeric spleen cells vs that of wild-type cells. For example, if 98 and 2% of A/J→F₁² chimeric splenocytes were stained with the 16-3.1 and 28-13-3 mAbs, respectively, and 99 and 1% of A/J wild-type splenocytes were stained with the 28-13-3 and 16-3.1 mAbs, respectively, then the A/J→F₁ chimeras were considered 99% donor phenotype.

Cells, transfectants, and hybridomas

The SaI sarcoma and its transfectants were cultured as previously described (7). SaI sarcoma cells transfected with syngeneic MHC class II A_α^k and A_β^k genes (SaI/A^k cells) have been previously described (7). SaI cells expressing I-A^b class II molecules (SaI/A^b cells) were generated by transfecting SaI cells with the A_α^b and A_β^b cDNAs contained in the pKCR3 plasmid (18) plus the pSV2.neo plasmid using the transfection procedure previously described (7). SaI, SaI/A^k, or SaI/A^b cells expressing an endoplasmic reticulum-retained hen egg lysozyme gene (HEL; SaI/HEL, SaI/A^k/HEL, SaI/A^b/HEL cells) were generated as previously described by transfection with the BCMG-HEL plasmid containing the hygromycin^R gene (19). Transfectants expressing MHC class II genes or HEL were maintained in medium supplemented with 400 μg/ml G418 or 400 μg/ml hygromycin, respectively. Double transfectants were maintained on both drugs. 3A9 is an I-A^k-restricted HEL_46–61-specific T cell hybridoma (20) and was maintained as previously described (19); JK1290 is an I-A^b-restricted HEL specific hybridoma (21) and was maintained in Iscove’s modified Dulbecco’s medium supplemented with 10% Fetalclone I (Hyclone, Logan, UT), 1% penicillin, 1% streptomycin, and 1% gentamicin.

Tumor challenges

For tumorigenicity studies, mice were inoculated i.p. with live tumor cells, observed daily for survival, and killed when they became moribund. The inoculation dose was chosen based on previous titration studies (7). For immunization studies, chimeric mice were inoculated i.p. with 5 × 10⁵ to 10⁶ live tumor cells and killed 6 or 7 days later. For tumor challenge studies, mice were inoculated i.p. with the indicated number of tumor cells and examined three times per week for tumor growth. Ascites tumors usually became palpable within 10 to 14 days of inoculation and grew progressively. Based on our previous experience with the SaI tumor, if mice do not develop palpable tumor within 2 mo of challenge, they will remain tumor free during their lifetime (7). Tumor incidence is the number of mice with progressively growing tumors divided by the total number of mice challenged. Tumor-bearing mice were killed according to University of Maryland Baltimore County institutional animal care and use committee guidelines when they became moribund. The mean survival time is the time between inoculation and sacrifice.

In vitro Ag presentation assays

Splenocytes from immunized mice were prepared from mechanically dissociated spleens, and B lymphocytes were removed by panning as previously described (22). Resulting T cells (5 × 10⁶ cells/well) were cocultured in flat-bottom 96-well plates with fresh naive A/J or C57BL/6 splenocytes in a final volume of 250 μl/well containing 1 mg/ml lysozyme. Responder to stimulator ratios were 10:1 and/or 50:1. Supernatants were harvested after 24 h and assayed for IL-2 content using ELISA kits as described by the manufacturer (Endogen, Boston, MA). Samples were run in triplicate, and the mean ± SD determined for each sample. In most cases, SDs were ≤5% of test values. Background values (IL-2 release in the absence of HEL) were subtracted from experimental values (IL-2 release in the presence of HEL) to obtain specific IL-2 release. Values were converted to picograms per milliliter using a standard curve incorporated into the IL-2 assay. In some experiments splenocytes were depleted of CD4⁺ or CD8⁺ T lymphocytes in vitro (22) or in vivo (12) before use in an APC assay. APC assays using splenocytes from chimeric mice plus T cell hybridomas 3A9 and JK1290 were performed as previously described (19).

Indirect immunofluorescence and flow cytometry

Tumor cells, transfectants, and splenocytes were monitored for cell surface Ag expression by indirect immunofluorescence as previously described (7) and analyzed on an Epics XL flow cytometer (Coulter, Hialeah, FL). The following mAbs were used: I-A^k (10-3-6 or 10-2-16) (23), I-A^b (34-5-3S) (17), K^k (16-3-1) (16), D^d (34-5-8) (24), and lysozyme (HyHEL 7 and 10) (25). Cells monitored for intracellular lysozyme were fixed with paraformaldehyde and stained with a mixture of the HyHEL 7 and 10 mAbs (25) as previously described (19). Isotype controls were performed for surface and cytoplasmically stained cells, and staining was essentially identical with that in fluorescent conjugate alone controls.

SaI sarcoma cells transfected with MHC class II and/or HEL genes express these gene products at the cell surface or intracellularly

SaI/A^k and SaI/A^k/HEL cells were available from previous experiments (7, 19). SaI/A^b and SaI/A^b/HEL cells were generated by gene transfection as described in Materials and Methods. The resulting transfectants were stained for cell surface expression of MHC class II molecules (live cells) or for intracellular expression of lysozyme (paraformaldehyde-fixed and saponin-permeabilized cells). As shown in Figure 1⇓, cells transfected with the MHC class II I-A^k and I-A^b genes expressed comparable levels of these molecules, as measured by staining with the 10-2-16 and 34-5-3S mAbs, respectively (Fig. 1⇓, i and j, and e and f for SaI/A^k and SaI/A^b cells, respectively). Similarly, cells transfected with the HEL construct expressed comparable levels of intracellular lysozyme as measured by staining with the mixture of HyHEL 7 and HyHEL 10 mAbs (Fig. 1⇓, n, p, and r for SaI/HEL, SaI/A^k/HEL, and SaI/A^b/HEL cells, respectively.), while untransfected cells were negative for lysozyme (Fig. 1⇓, m, o, and q). HEL transfectants were also stained for cell surface HEL expression and were negative (data not shown). Supernatants of the transfectants were assayed by ELISA for HEL secretion and showed low levels of soluble HEL (1–5 ng/ml/6.7 × 10⁵ cells/24 h). The transfectants were also stained for MHC class I H-2K^k, H-2D^d, and H-2L^d Ag expression, and these levels were approximately equivalent among all transfectants and parental SaI cells (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Indirect immunofluorescence staining of SaI (a, g, and m), SaI/HEL (b, h, and n), SaI/A^k (c, i, and o), SaI/A^k/HEL (d, j, and p), SaI/A^b (e, k, and q), and SaI/A^b/HEL (f, l, and r) cells. Tumor cells were stained for I-A^b with the 28-13-3 mAb (17) (a–f), for I-A^k with the 10-2-16 mAb (23) (g–l), or for lysozyme with the HyHEL 7 plus 10 mAbs (25) (m–r). MHC class II staining was performed on live tumor cells; lysozyme staining was performed on fixed, permeabilized cells. The second-step, fluorescent conjugate FITC-goat anti-mouse IgG was used with all Abs. The solid line represents staining by fluorescent conjugate alone; the dotted line represents staining by specific Ab plus fluorescent conjugate.

Lysozyme peptides are presented by both I-A^k and I-A^b MHC class II molecules

Numerous in vitro studies have demonstrated that both I-A^k and I-A^b MHC class II molecules present HEL-derived peptides to CD4⁺ T cells (21, 26, 27, 28). To ascertain that HEL peptides are presented by both class II alleles when the alleles are expressed by SaI sarcoma cells, semisyngeneic (C57BL/6 × A/J)F₁ mice were challenged i.p. with parental SaI, SaI/A^k, SaI/A^b, SaI/HEL, SaI/A^k/HEL, and SaI/A^b/HEL tumor cells and followed for tumor incidence. As shown in Table I⇓, wild-type SaI and SaI/HEL tumor cells were lethal in ≥94% of F₁ mice, indicating that HEL expression alone does not cause tumor rejection. SaI/A^k and SaI/A^b cells were significantly less lethal than SaI cells; however, they still caused tumors in a significant number of F₁ mice (50 and 40% lethal, respectively). In contrast, SaI/A^k/HEL and SaI/A^b/HEL tumor cells were rejected by 100% of F₁ mice, suggesting that HEL peptides are presented by the I-A^k and I-A^b MHC class II molecules of the tumor cells and function as nominal Ag for T cell recognition.

Immunization of chimeric mice with tumor cell transfectants generates T cells restricted to both the tumor genotype and the host genotype

If the genetically modified tumor cells are the exclusive APC for tumor Ags, then A→F₁ bone marrow chimeras immunized with B tumor cells will have HEL-specific T cells restricted to the tumor (B) MHC genotype. In contrast, if host-derived cells are the APC, then the chimeras will have HEL-specific T cells restricted to the genotype of the host (A) regardless of the genotype of the tumor. To test these alternatives, A/J→ F₁ and B6→F₁ chimeras were generated and challenged i.p. with SaI/A^b/HEL or SaI/A^k/HEL tumor cells, respectively. Seven days after tumor cell immunization, spleens were removed and depleted of B cells and adherent cells by panning, and the remaining T cells were incubated in vitro with lysozyme plus fresh A/J or C57BL/6 splenocytes as APC. Supernatants were harvested 24 h later and tested for IL-2 content. Since the goal of these experiments is to characterize the earliest Ag-specific response, T cells were tested 1 wk postimmunization. Likewise, T cells were assayed for IL-2 secretion without an in vitro or in vivo boost to the Ag (HEL), so that the primary or initial response would be measured.

A/J→F₁ and B6→F₁ bone marrow chimeras were generated as described in Materials and Methods. The efficiency of chimera formation was determined using indirect immunofluorescence to measure the percentage of donor genotype cells in the recipients’ spleens. Table II⇓ lists the chimeras used in subsequent experiments and shows the MHC class I genotype (H-2K^k for A/J vs H-2D^b for C57BL/6) of splenocytes from the chimeras as measured by immunofluorescence. As shown, all the chimeras used in the following experiments were ≥97.9%, and most chimeras were ≥99.5%, of the donor genotype. Chimeras were also tested functionally for hemopoietic reconstitution. Splenocytes of representative chimeras were used as APC for intact HEL to I-A^k- and I-A^b-restricted HEL-specific hybridomas, 3A9 and JK1290, respectively. A/J→F₁ and A/J splenocytes presented HEL to 3A9 hybridoma cells, but not to JK1290 hybridoma cells. In contrast, B6→F₁ and C57BL/6 splenocytes presented Ag to JK1290 cells, but not to 3A9 hybridoma cells (data not shown). APC of the bone marrow chimeras, therefore, are phenotypically and functionally the donor genotype.

Table III⇓ shows the results of three representative, independent experiments assessing IL-2 production by total T cells from tumor-immunized, chimeric mice. In two experiments (Expt. 1 and 2) A/J→F₁ chimeras were immunized with SaI/A^b/HEL tumor cells, while in one experiment (Expt. 4), B6→F₁ chimeras were immunized with SaI/A^k/HEL tumor cells. In all three experiments HEL-specific T cells restricted to both the tumor and the host genotype were present. IL-2 release by immune T cells cocultured in vitro with lysozyme plus irrelevant genotype APC (SWR, H-2^q) was at background levels (data not shown). Therefore, at 1 wk postimmunization with genetically modified tumor cells, both tumor cells and host bone marrow-derived cells were APC for tumor- encoded Ags.

The genetically modified tumor cells constitutively express MHC class I (H-2K^k, H-2D^d, and H-2L^d) as well as the transfected MHC class II (I-A^k or I-A^b) molecules. Since CD4⁺ and CD8⁺ T cells are activated by Ag in the context of MHC class II and class I molecules, respectively, and since both T cell types can synthesize IL-2, it is possible that the read-out of the splenic T cell experiments reflects HEL-specific, class I-restricted CD8⁺ T cells. To identify the responding splenic T cells, chimeric mice were depleted of CD4⁺ or CD8⁺ T cells before and during immunization, and the remaining T cells were tested for IL-2 secretion following HEL presentation in vitro. Two representative experiments are shown in Table III⇑, one experiment using A/J→F₁ chimeric mice immunized with SaI/A^b/HEL tumor cells (Expt. 3) and a second experiment using B6→F₁ chimeric mice immunized with SaI/A^k/HEL tumor cells (Expt. 5). As shown in Table II⇑, mice depleted for CD4⁺ or CD8⁺ T cells had ≤1.2 or 0.6% of these cells, respectively, demonstrating functional depletion of these populations. As shown in Table III⇑, in both the depletion experiments most of the IL-2 activity was produced by CD4⁺ T cells, since depletion of CD4⁺ T cells reduced IL-2 activity to negligible levels. Consistent with the results of Expt. 1, 2, and 4, in Expt. 3, both tumor genotype and host genotype APC stimulated HEL-specific IL-2 release, indicating that both tumor cells and host-derived cells are APC for tumor-encoded Ag (HEL). In contrast, in Expt. 5, only host genotype (H-2^b) cells stimulated IL-2 secretion, suggesting that in these mice only host-derived cells are the APC for tumor-encoded Ags.

The five experiments shown in Table III⇑ are representative of 25 similar experiments performed using chimeric or (C57BL/6 × A/J)F₁ mice. Results similar to those of Expt. 5, in which only host genotype APC stimulated IL-2 secretion, were noted in only two of these experiments, while the remaining experiments all showed either host and tumor genotype presentation, or tumor genotype presentation alone. It was occasionally observed that tumor-encoded Ags were exclusively presented by host-derived cells; however, such presentation was a relatively rare event. Similar experiments using class II⁻ SaI/HEL tumor cells showed no direct Ag presentation (data not shown), indicating that direct Ag presentation by tumor cells to CD4⁺ T cells requires MHC class II expression by the tumor. Therefore, as measured at 1 wk postimmunization, CD4⁺ T cells specific for tumor-encoded Ags were activated via Ag presentation by both tumor and host-derived cells, indicating that both direct and indirect (cross-priming) pathways are used.