There is limited information on the influence of tumor growth on the expansion of tumor-specific TGF-β-producing CD4+ T cells in humans. α-Fetoprotein (AFP) is an oncofetal Ag and has intrinsic immunoregulatory properties. In this study, we report the identification and characterization of subsets of CD4+ T cells that recognize an epitope within the AFP sequence (AFP46–55) and develop into TGF-β-producing CD4+ T cells. In a peptide-specific and dose-dependent manner, AFP46–55 CD4+ T cells produce TGF-β, GM-CSF, and IL-2 but not Th1-, Th2-, Th17-, or Tr1-type cytokines. These cells express CTLA-4 and glucocorticoid-induced TNR receptor and inhibit T cell proliferation in a contact-dependent manner. In this study, we show that the frequency of AFP46–55 CD4+ T cells is significantly higher (p = 001) in patients with hepatocellular carcinoma than in healthy donors, suggesting that these cells are expanded in response to tumor Ag. In contrast, tumor necrosis-inducing treatments that are shown to improve survival rate can shift the Th1/TGF-β-producing CD4+ T cell balance in favor of Th1 responses. Our data demonstrate that tumor Ags may contain epitopes which activate the expansion of inducible regulatory T cells, leading to evasion of tumor control.

There are two types of regulatory T cells (Treg)3: natural and inducible. Natural CD4+CD25+ Treg are derived in thymus and express CD25 and Foxp3. Inducible Treg are generated in the periphery in response to pathogen or self-Ags and produce IL-10 (Tr1 cells) or TGF-β (Th3 cells). Natural and inducible Treg can be beneficial to the host by regulating anti-self-response in autoimmune patients or during infection by preventing pathogen-induced immunopathology. However, the induction or activation of Treg by tumors or pathogens may suppress protective immunity (1).

It has been suggested that some tumors may activate the expansion of inducible Treg (2). The majority of tumor-associated Ags are self-Ags with the ability to stimulate inducible Treg that can inhibit the development of an effective antitumor immunity (3). To avoid unwanted expansion of inducible Treg by vaccines targeting tumor Ags, it is crucial to identify CD4+ Treg epitopes within tumor-associated Ag sequences. In contrast, the expansion of Treg in autoimmune diseases could suppress anti-self-immune responses (4). Therefore, MHC class II-restricted T cell epitopes with the ability to induce the expansion of Treg in vivo could be used in the treatment of autoimmune diseases. Moreover, it is clear that desirable peptides for therapeutic vaccines should be promiscuous T cell epitopes, which could be recognized by CD4+ T cells with different alleles, allowing broad population coverage.

α-Fetoprotein (AFP) is an oncofetal Ag with intrinsic immunoregulatory properties (5, 6) and is also a tumor rejection Ag in hepatocellular carcinoma (HCC) (7). Several immunodominant AFP-derived Th1 and Tc1 epitopes have been recently identified (8, 9, 10). However, there is little information on the ability of AFP to stimulate the expansion of inducible Treg and as yet no AFP-derived Treg epitope has been identified. In this study, we report the identification of the first self-Ag-derived TGF-β- producing CD4+ T cell epitope in humans and demonstrate that overexpression of AFP stimulate the expansion of AFP-specific TGF-β-producing CD4+ T cells in patients with HCC.

In total, 94 peptides spanning the AFP sequence were synthesized by mimotopes. Sixty-two were soluble in DMSO and were tested in this study (Table I).

Table I.

AFP-derived peptides

Amino Acid StartSequenceAmino Acid StartSequence
MKWVESIFL 350 FLASFVHEY 
FLIFLLNFT 365 QLAVSVILRV 
11 FLLNFTESRT 372 RVAKGYQEL 
20 TLHRNEYGI 379 ELLEKCFQT 
30 SILDSYQCTA 385 FQTENPLEC 
35 YQCTAEISL 410 ALAKRSCGL 
37 CTAEISLADL 419 FQKLGEYYL 
40 EISLADLATI 427 LQNAFLVAYT 
46 LATIFFAQFV 431 FLVAYTKKA 
54 FVQEATYKEV 441 QLTSSELMAI 
65 KMVKDALTAI 449 AITRKMAAT 
70 ALTAIEKPT 453 KMAATAATCC 
86 CLENQLPAFL 462 CQLSEDKLL 
89 NQLPAFLEEL 468 KLLACGEGA 
125 FLAHKKPTPA 475 GAADIIIGHL 
137 PLFQVPEPV 485 CIRHEMTPV 
140 QVPEPVTSC 489 EMTPVNPGV 
158 FMNKFIYEI 492 PVNPGVGQC 
164 YEIARRHPFL 498 GQCCTSSYA 
172 FLYAPTILL 507 NRRPCFSSLV 
179 LLWAARYDKI 514 SLVVDETYV 
187 KIIPSCCKA 531 FIFHKDLCQA 
217 SLLNQHACAV 536 DLCQAQGVAL 
235 FQAITVTKL 542 GVALQTMKQ 
249 KVNFTEIQKL 545 LQTMKQEFLI 
277 CLQDGEKIM 548 MKQEFLINL 
298 KITECCKLTT 555 NLVKQKPQI 
306 TTLERGQCII 562 QITEEQLEAV 
325 GLSPNLNRFL 570 AVIADFSGL 
343 SSGEKNIFL 576 SGLLEKCCQ 
347 KNIFLASFV 598 KLISKTRAAL 
Amino Acid StartSequenceAmino Acid StartSequence
MKWVESIFL 350 FLASFVHEY 
FLIFLLNFT 365 QLAVSVILRV 
11 FLLNFTESRT 372 RVAKGYQEL 
20 TLHRNEYGI 379 ELLEKCFQT 
30 SILDSYQCTA 385 FQTENPLEC 
35 YQCTAEISL 410 ALAKRSCGL 
37 CTAEISLADL 419 FQKLGEYYL 
40 EISLADLATI 427 LQNAFLVAYT 
46 LATIFFAQFV 431 FLVAYTKKA 
54 FVQEATYKEV 441 QLTSSELMAI 
65 KMVKDALTAI 449 AITRKMAAT 
70 ALTAIEKPT 453 KMAATAATCC 
86 CLENQLPAFL 462 CQLSEDKLL 
89 NQLPAFLEEL 468 KLLACGEGA 
125 FLAHKKPTPA 475 GAADIIIGHL 
137 PLFQVPEPV 485 CIRHEMTPV 
140 QVPEPVTSC 489 EMTPVNPGV 
158 FMNKFIYEI 492 PVNPGVGQC 
164 YEIARRHPFL 498 GQCCTSSYA 
172 FLYAPTILL 507 NRRPCFSSLV 
179 LLWAARYDKI 514 SLVVDETYV 
187 KIIPSCCKA 531 FIFHKDLCQA 
217 SLLNQHACAV 536 DLCQAQGVAL 
235 FQAITVTKL 542 GVALQTMKQ 
249 KVNFTEIQKL 545 LQTMKQEFLI 
277 CLQDGEKIM 548 MKQEFLINL 
298 KITECCKLTT 555 NLVKQKPQI 
306 TTLERGQCII 562 QITEEQLEAV 
325 GLSPNLNRFL 570 AVIADFSGL 
343 SSGEKNIFL 576 SGLLEKCCQ 
347 KNIFLASFV 598 KLISKTRAAL 

This study was approved by ethical committees and all patients gave written informed consent. PBMCs were isolated from the blood of patients with HCC or healthy donors.

Short-term T cell lines were generated as described previously (8). In brief, PBMCs were resuspended in AIM-V medium (Invitrogen Life Technologies) and cultured with individual peptides (1 μM). rIL-2 (25 IU/ml) was added on days 2 and 3 of culture and the cells were analyzed after a total of 10–12 days of culture. The experiments presented in this study (excluding inhibition assay and ex vivo data) were performed on short-term T cell lines.

To generate long-term T cell lines (for inhibition assay), PBMCs were resuspended in AIM-V medium (Invitrogen Life Technologies) and cultured with individual peptides (1 μM). rIL-2 (25 IU/ml) was added on days 2 and 3 of culture. After 10–12 days of culture, CD4+ T cells were isolated from short-term T cell lines using Dynabeads and the cells (1 cell/well) were cultured with γ-irradiated 5 × 104 allogeneic PBMCs as feeder cells, and rIL-2 (30 IU/ml) culture medium was changed once a week with fresh medium, rIL-2, and feeder cells. On day 21, T cells were tested for GM-CSF production using ELISA. CD4+ T cell lines that produced peptide-specific GM-CSF were selected and expanded.

AFP-specific T cells were incubated for 5 h at 37°C with AFP-derived peptides (1 μM) or peptide-pulsed or protein pulsed APCs and brefeldin A. Cells were surface stained with Abs to CD3, CD4, CD8, CD25, TCR-αβ, HLA-DR, CD62L, CD45, and GITR (BD Pharmingen). The cells were then permeabilized, fixed, and stained for intracellular molecules (GM-CSF, TGF-β, IL-2, IFN-γ, IL-10, TNF-α, IL-5, IL-13, IL-17, and CTLA-4) or isotype controls (R&D Systems), washed twice, and the frequency of peptide-specific T cell responses was quantified by flow cytometry. Anti-TGF-β Abs for intracellular staining were obtained from R&D Systems and IQ Products. Cells were stained with Abs to Foxp3 (eBioscience) as described by the manufacturer’s instructions. An immunological responder was defined as a 2-fold increase in frequency of cytokine-producing cells above control peptides or proteins.

AFP46–55-spcific T cell lines or control T cell lines (AFP364–373) were washed and cultured in serum-free medium in the presence of relevant or irrelevant peptides for 48 h, and the amount of total TGF-β and GM-CSF were measured in culture supernatants by ELISA (R&D Systems).

TGF-β-releasing cells were detected upon specific peptide stimulation using an ELISPOT assay ex vivo. Nitrocellulose-backed plates (96-well, MAHA S45; Millipore) were coated with mouse anti-human latent TGF-β capture Ab overnight at 4°C. The wells were washed five times with PBS and blocked using blocking buffer (1% BSA and 5% sucrose PBS) for 2 h. PBMCs and the peptides were then added into the wells and incubated for 18 h at 37°C in 5% CO2. The wells were washed with wash buffer (0.05% Tween 20 in PBS), then 1 μg/ml secondary biotin-conjugated anti-human latent TGF-β Ab (R&D Systems) was added and incubated at 4°C overnight. The color development was done using ELISPOT blue color module (R&D systems). After 30 min, the wells were washed with tap water, dried, and the spots counted.

CD4+CD25 T cells (2 × 105) isolated from PBMCs by Ab-coated beads were cultured for 5 days in 96-well plates containing 5 × 104 CD3-depleted APCs, 0.5 μg/ml anti-CD3 mAb, and different numbers of regulatory (AFP46–55) or effector (AFP364–373 peptide) CD4+ T cells in medium containing 10% human serum. The proliferation of responder T cells was assessed by the incorporation of [3H]thymidine for the last 18 h of culture. Cells were harvested and radioactivity was counted in a scintillation counter. All experiments were performed in triplicates. For some experiments, Ab against TGF-β (R&D Systems) was added in the assay at a final concentration of 5 μg/ml.

Transwell experiments were performed in 24-well plates with a 0.4-μm pore size (Corning Glass). Purified naive CD4+ T cells (1 × 105) were cultured in the outer wells in medium containing 0.5 μg/ml anti-CD3 Ab and 2 × 105 APCs. Equal numbers of AFP46–55 CD4+ T cells or AFP364–373 CD4+ T cells were added into the inner wells in the same medium containing anti-CD3 and 2 × 105 APCs. The cells in the inner and outer wells were harvested separately and transferred into 96-well plates after 3 days of culture. [3H]Thymidine was added, and the cells were cultured for another 18 h before being harvested for counting the radioactivity with a liquid scintillation counter.

The Mann-Whitney U test (two tailed) was used to compare the frequencies of AFP46–55 -specific GM-CSF- producing CD4+ T cells in healthy and cancer patients. The statistical significance was defined at p < 0.05.

GM-CSF is a cytokine that is produced by different T cell populations, including inducible Treg (11) and TGF-β is an immunoregulatory cytokine produced by Th3 cells. Short-term T cell lines were generated in medium containing rIL-2 with or without purified AFP (5 μg/ml). Cells were washed, counted, and cultured in serum-free medium in the presence or absence of purified AFP (5 μg/ml). The amounts of GM-CSF and total TGF-β were measured in cell culture supernatant using ELISA. T cell lines stimulated with AFP produced TGF-β and GM-CSF (Fig. 1). The depletion of CD4+ cells but not CD8 T cells before restimulation reduced AFP-specific GM-CSF and TGF-β production by T cell lines (Fig. 1), suggesting that CD4+ T cells are the source of GM-CSF and TGF-β.

FIGURE 1.

AFP stimulate GM-CS F and TGF-β production by CD4+ T cells. The amount of GM-CSF (a) and total TGF-β (b) were measured in cell culture supernatants of T cell lines stimulated with AFP (5 μg/ml). CD4+ or CD8+ T cells were depleted from T cell lines before restimulation to determine the source of cytokine production.

FIGURE 1.

AFP stimulate GM-CS F and TGF-β production by CD4+ T cells. The amount of GM-CSF (a) and total TGF-β (b) were measured in cell culture supernatants of T cell lines stimulated with AFP (5 μg/ml). CD4+ or CD8+ T cells were depleted from T cell lines before restimulation to determine the source of cytokine production.

Close modal

Short-term T cell lines were generated as described in Materials and Methods. In short, PBMCs were cultured in the presence of rIL-2 and 62 different peptides spanning the AFP sequence (Table I) for 10 days. The reactivity was analyzed using intracellular cytokine staining for GM-CSF. Among 62 peptides, the AFP46–55 peptide (LATIFFAQFV) stimulated GM-CSF production by CD3+CD4+ T cells (Fig 2, a and b) in a dose-dependent manner (Fig. 2,c). Similar results were observed in T cell lines generated from three other individuals. GM-CSF production by AFP46–55-specific CD4 T cells is peptide specific as determined using ELISA (Fig. 2 d).

FIGURE 2.

Identification of AFP-derived peptide epitope that stimulates GM-CSF production. a and b, Short-term T cell lines were generated from PBMCs isolated from HCC patients in the presence of 62 different peptides spanning the AFP sequence. After 10 days, cells were restimulated with relevant or irrelevant peptides (c) at different concentrations. Numbers indicate percentages of GM-CSF-producing cells within CD3+CD4+ T cells. d, The production of peptide-specific GM-CSF was detected in the supernatant of AFP46–55 CD4+ T cells using ELISA (in triplicate wells ± SD). e, Peptide-pulsed EBV B cells or MHC class II- deficient cells (HepG2 cells) were cultured with AFP46–55 CD4+ T cells, and peptide recognition was analyzed an using intracellular cytokine assay for GM-CSF. f, To determine optimal peptide length required for T cell recognition, AFP46–55 CD4+ T cells were restimulated with AFP46–55, AFP47–55, AFP44–57, AFP42–55, or AFP364–373 peptide and the peptide recognition by CD4+ T cells was determined using an intracellular cytokine assay for GM-CSF. Two independent experiments were performed and the results are confirmed in three other individuals.

FIGURE 2.

Identification of AFP-derived peptide epitope that stimulates GM-CSF production. a and b, Short-term T cell lines were generated from PBMCs isolated from HCC patients in the presence of 62 different peptides spanning the AFP sequence. After 10 days, cells were restimulated with relevant or irrelevant peptides (c) at different concentrations. Numbers indicate percentages of GM-CSF-producing cells within CD3+CD4+ T cells. d, The production of peptide-specific GM-CSF was detected in the supernatant of AFP46–55 CD4+ T cells using ELISA (in triplicate wells ± SD). e, Peptide-pulsed EBV B cells or MHC class II- deficient cells (HepG2 cells) were cultured with AFP46–55 CD4+ T cells, and peptide recognition was analyzed an using intracellular cytokine assay for GM-CSF. f, To determine optimal peptide length required for T cell recognition, AFP46–55 CD4+ T cells were restimulated with AFP46–55, AFP47–55, AFP44–57, AFP42–55, or AFP364–373 peptide and the peptide recognition by CD4+ T cells was determined using an intracellular cytokine assay for GM-CSF. Two independent experiments were performed and the results are confirmed in three other individuals.

Close modal

HepG2 cells express MHC class I and can present peptide epitopes to CD8+ T cells (12), but these cells are MHC class II deficient (13). AFP46–55 CD4+ T cells were stimulated with peptide-pulsed HepG2 cells or EBV B cells for 5 h and peptide recognition was analyzed using an intracellular cytokine assay for GM-CSF. AFP46–55 peptide-pulsed HepG2 cells did not stimulate GM-CSF production by CD4+ T cells, suggesting that HepG2 cells are unable to present AFP46–55 peptide epitope to CD4+ T cells (Fig. 2 e).

To identify the optimal length of peptide sequence, AFP46–55 CD4+ T cells were stimulated with AFP47–55 (9 aa long), AFP46–55, AFP44–57 (14 aa long), and AFP42–55 (14 aa long), and the frequency of peptide-specific GM-CSF-producing CD4+ T cells was analyzed. AFP46–55 CD4+ T cells recognized AFP47–55, AFP46–55, AFP44–57, and AFP42–55 but not an irrelevant peptide (AFP364–373) and produced peptide-specific GM-CSF. The frequency of GM-CSF-producing cells among CD4+ T cells was highest in cells stimulated with AFP46–55 (Fig. 2 f).

To study the role of IL-2, IL-7, and IL-15 on the generation and expansion of AFP46–55-specific CD4+ T cells, PBMCs were cultured in the presence or absence of different combinations of these cytokines. AFP46–55 CD4+ T cells were not expanded in the absence of exogenous IL-2, suggesting that IL-2 is essential for the expansion of these cells. The highest percentage of AFP46–55 CD4+ T cells was detected in cultures expanded in the presence of IL-2 (25 IU/ml), IL-7 (20 ng/ml), and IL-15 (20 ng/ml) (data not shown).

To test the ability of T cell lines to produce Ag-specific TGF-β, AFP46–55 or AFP364–373 T cell lines (Th1 cells) (8) were washed and stimulated (2 × 105 cells/well) with increasing concentrations of AFP46–55 or AFP364–373 peptides in serum-free medium for 48 h. The amounts of total TGF-β were measured in the culture supernatant using an ELISA for TGF-β. AFP46–55 T cell lines stimulated with AFP46–55 produced TGF-β in a dose-dependent manner (Fig. 3,a). The depletion of CD4+ cells or CD2+ cells before peptide restimulation but not the depletion of CD8+ T cells reduced peptide-specific TGF-β production by AFP46–55 T cell lines (Fig. 3,b), suggesting that CD4+ T cells are the source of TGF-β. To test the recognition of purified AFP by AFP46–55 CD4+ T cells, AFP46–55 CD4+T cell lines or a control AFP364–373 T cell line were cultured with APCs pulsed with purified AFP (5 μg/ml) or a control protein (human serum albumin) for 48 h. The amount of total TGF-β was measured in the culture supernatant using an ELISA for TGF-β. AFP46–55 CD4+ T cell lines stimulated with purified AFP but not with control protein produced large quantities of TGF-β (Fig. 3,c). The production of TGF-β by AFP46–55 CD4+ T cells was confirmed using an intracellular cytokine assay (Fig. 3 d). Both anti-TGF-β Abs from R&D Systems and IQ Products stained similar percentages of peptide-specific TGF-β-producing CD4+ T cells (data not shown).

FIGURE 3.

AFP46–55 CD4+ T cells produce TGF-β in a dose-dependent manner. The amounts of total TGF-β were measured in the culture supernatant of AFP46–55 CD4+ T cells restimulated with different concentrations of the relevant or irrelevant peptides (a). b, CD4+, CD2+, or CD8+ cells were depleted from PBMCs and cultured in medium containing AFP46–55 peptide and rIL-2 for 10 days. The levels of total TGF-β were measured in the supernatant of the T cell lines. c, The amounts of TGF-β produced by AFP46–55 CD4+ T cells upon restimulation with purified AFP, human serum albumin, AFP46–55 peptide, or AFP364–373 peptide are shown. d, The percentage of intracellular TGF-β produced by CD4+ T cells is shown. Two independent experiments were performed and the results were confirmed in five other individuals.

FIGURE 3.

AFP46–55 CD4+ T cells produce TGF-β in a dose-dependent manner. The amounts of total TGF-β were measured in the culture supernatant of AFP46–55 CD4+ T cells restimulated with different concentrations of the relevant or irrelevant peptides (a). b, CD4+, CD2+, or CD8+ cells were depleted from PBMCs and cultured in medium containing AFP46–55 peptide and rIL-2 for 10 days. The levels of total TGF-β were measured in the supernatant of the T cell lines. c, The amounts of TGF-β produced by AFP46–55 CD4+ T cells upon restimulation with purified AFP, human serum albumin, AFP46–55 peptide, or AFP364–373 peptide are shown. d, The percentage of intracellular TGF-β produced by CD4+ T cells is shown. Two independent experiments were performed and the results were confirmed in five other individuals.

Close modal

Ag-specific CD4+ T cells can be classified as Th1, Th2, Th17, or Tr1 based on their ability to produce different cytokine profiles. To classify AFP46–55 CD4+ T cells, we analyzed their ability to produce different cytokines upon peptide stimulation. AFP46–55-specific CD4+ T cells were generated from HCC patients and their ability to produce cytokines was evaluated using intracellular cytokine assays. AFP46–55 CD4+ T cells did not produce Th1 (IFN-γ, TNF-α)-, Th2 (IL-5, IL-13)-, Tr1 (IL-10)-, or Th17 (IL-17)-type cytokines. AFP46–55-specific CD4+ T cells recognized the relevant peptide and produced TGF-β, GM-CSF, and IL-2 (Fig. 4,a). As determined using six-color flow cytometry, IL-2, GM-CSF, and TGF-β are produced by the same AFP46–55-specific CD4+ T cells. AFP46–55 T cell lines generated from five other individuals produced similar patterns of cytokine production. A summary of AFP46–55 peptide-specific cytokine- producing CD4+ T cells (GM-CSF, TGF-β, and IL-2) analyzed using an intracellular cytokine assay in six individuals is shown (Fig. 4 b). AFP46–55 did not stimulate IFN-γ, TNF-α, IL-5, IL-13, IL-10, and IL-17 production by CD4+ T cells from these individuals (data not shown).

FIGURE 4.

AFP46–55 CD4+ T cells do not produce Th1-, Th2-, Th17-, or Tr1-type cytokines. Cell lines were restimulated with AFP46–55 peptide or an irrelevant peptide (AFP364–373), and the production of IFN-γ, TNF-α, IL-5, IL-13, IL-17, IL-2, GM-CSF, TGF-β, and IL-10 by CD4+ T cells was analyzed using an intracellular cytokine assay. a, The percentages of peptide-specific cytokine-producing CD4+ T cells are shown. Two independent experiments were performed. b, A summary of cytokines produced by CD4 T cells from six different individuals is shown. Each symbol represents percent cytokine-producing CD4+ T cells from an individual.

FIGURE 4.

AFP46–55 CD4+ T cells do not produce Th1-, Th2-, Th17-, or Tr1-type cytokines. Cell lines were restimulated with AFP46–55 peptide or an irrelevant peptide (AFP364–373), and the production of IFN-γ, TNF-α, IL-5, IL-13, IL-17, IL-2, GM-CSF, TGF-β, and IL-10 by CD4+ T cells was analyzed using an intracellular cytokine assay. a, The percentages of peptide-specific cytokine-producing CD4+ T cells are shown. Two independent experiments were performed. b, A summary of cytokines produced by CD4 T cells from six different individuals is shown. Each symbol represents percent cytokine-producing CD4+ T cells from an individual.

Close modal

We then analyzed the cell surface phenotype of AFP46–55 CD4+ T cells in short-term T cell lines. The expression of surface and intracellular molecules (CD4, CD45RO, CD62L, CTLA-4, and GITR) were analyzed. The majority of AFP46–55 CD4+ GM-CSF-producing T cells expressed surface CD45RO and intracellular CTLA-4, but not CD62L. Nonresponder CD4+ T cells (cells not producing GM-CSF) did not express GITR or intracellular CTLA-4 (Fig. 5 a). AFP46–55 CD4+ T cells expressed CD3, TCR-αβ, and CD25, but not CD8, CD14, or CD16 (data not shown).

FIGURE 5.

Phenotypic characterization of peptide-specific CD4+ T cells. AFP46–55 CD4+ T cell line was restimulated with AFP46–55 peptide and the cells were stained with mAbs to detect surface CD4, CD69L, CD45RO, and GITR molecules, intracellular CTLA-4 and GM-CSF, and intranucleus Foxp3. Isotype control Abs served as negative control staining. a, CD4+ GM-CSF-producing T cells or CD4+ GM-CSF-negative cells were gated and the expression level of surface and intracellular molecules is shown. b, The expression levels of Foxp3 and CD25 in IL-2- and TGF-β-producing AFP46–55 CD4+ T cells are shown. Two independent experiments are performed in T cell lines generated from five other individuals.

FIGURE 5.

Phenotypic characterization of peptide-specific CD4+ T cells. AFP46–55 CD4+ T cell line was restimulated with AFP46–55 peptide and the cells were stained with mAbs to detect surface CD4, CD69L, CD45RO, and GITR molecules, intracellular CTLA-4 and GM-CSF, and intranucleus Foxp3. Isotype control Abs served as negative control staining. a, CD4+ GM-CSF-producing T cells or CD4+ GM-CSF-negative cells were gated and the expression level of surface and intracellular molecules is shown. b, The expression levels of Foxp3 and CD25 in IL-2- and TGF-β-producing AFP46–55 CD4+ T cells are shown. Two independent experiments are performed in T cell lines generated from five other individuals.

Close modal

Foxp3 is expressed by CD4+CD25+ T cells but its expression in inducible Treg is controversial. We analyzed the expression of Foxp3 in AFP46–55 CD4+ T cells using flow cytometry. The majority of AFP46–55 peptide-specific IL-2 and TGF-β-producing CD3+CD4+ did not express Foxp3 and only expressed low levels of CD25 (Fig. 5 b). In some T cell lines, small percentages of AFP46–55 CD4+ T cells were Foxp3 positive (data not shown).

It has been shown that rTGF-β induces certain CD4+ T cells in human peripheral blood to develop contact-dependent suppressive activity that is not antagonized by anti-TGF-β. This suppressive activity was only partially abrogated when rIL-2 was added to the culture (14). In this study, we examined the suppressive effects of AFP46–55 CD4+ T cells that are shown to produce both TGF-β and IL-2. Four long-term CD4+ T cell lines (lines 1–4) were generated from PBMCs of a healthy donor using limited dilution methods as described in Materials and Methods. Ag specificity and phenotypic characterization of AFP46–55 CD4+ T cell lines were analyzed after restimulation with peptide-pulsed APCs (adherent cells) and detection of intracellular GM-CSF. The majority of CD4+ T cells from these lines recognized the peptide and produced GM-CSF, suggesting that these cells are Ag specific. Next, we tested the ability of these T cell lines to inhibit anti CD3-induced T cell proliferation. The proliferation rate of responding cells without T cell lines was considered as 100% proliferation. AFP46–55 CD4+ T cell line 1 moderately inhibited T cell proliferation, suggesting that these cells may have some inhibitory function. AFP46–55 CD4+ T cell line 3 did not inhibit T cell proliferation (Fig. 6,a). The percentages of TGF-β-producing CD4+ T cells for each long-term T cell line were determined using an intracellular cytokine assay. Eighty-four percent of T cells from line 1 produced TGF-β in a peptide-specific manner but only 32% of the T cells from line 3 produced TGF-β (Fig. 6,b). The addition of anti-TGF-β Ab did not reverse or alter the inhibitory effects of the inhibitory T cells (Fig. 6 c).

FIGURE 6.

AFP46–55 CD4+ T cells suppress T cell proliferation to anti-CD3 Ab stimulation in a cell-cell contact manner. a, In an inhibition assay, responding T cells were cultured with or without AFP46–55 CD4+ T cells (lines 1–4) generated from a healthy donor at a 1:1 ratio and anti-CD3 Ab-induced T cell proliferation was measured. The percentages of inhibition recorded in wells containing AFP46–55 CD4+ T cells are shown. b, The percentages of TGF-β-producing AFP46–55 CD4+ cells are shown for each long-term T cell line. c, Anti-TGF-β mAb did not abrogate the inhibitory effects of AFP46–55 CD4+ T cells. d, Cytokine profile in AFP46–55 CD3+CD4+ T cells (lines 1 and 3). The percentages of cytokine-producing CD4+ T cells are shown. e, The suppressive effects of AFP46–55 and AFP364–373 CD4+ T cells (Th1) from a HCC patient were evaluated in an anti-CD3-induced T cell proliferation assay. f, In a Transwell system, AFP46–55 or AFP364–373 CD4+ T cells were cocultured or cultured separately (in inner wells) with responding CD4+ T cells and stimulated with anti-CD3 Ab. Mean [3H]thymidine incorporation indicated as cpm (±SD) in triplicate wells. Two independent experiments were performed.

FIGURE 6.

AFP46–55 CD4+ T cells suppress T cell proliferation to anti-CD3 Ab stimulation in a cell-cell contact manner. a, In an inhibition assay, responding T cells were cultured with or without AFP46–55 CD4+ T cells (lines 1–4) generated from a healthy donor at a 1:1 ratio and anti-CD3 Ab-induced T cell proliferation was measured. The percentages of inhibition recorded in wells containing AFP46–55 CD4+ T cells are shown. b, The percentages of TGF-β-producing AFP46–55 CD4+ cells are shown for each long-term T cell line. c, Anti-TGF-β mAb did not abrogate the inhibitory effects of AFP46–55 CD4+ T cells. d, Cytokine profile in AFP46–55 CD3+CD4+ T cells (lines 1 and 3). The percentages of cytokine-producing CD4+ T cells are shown. e, The suppressive effects of AFP46–55 and AFP364–373 CD4+ T cells (Th1) from a HCC patient were evaluated in an anti-CD3-induced T cell proliferation assay. f, In a Transwell system, AFP46–55 or AFP364–373 CD4+ T cells were cocultured or cultured separately (in inner wells) with responding CD4+ T cells and stimulated with anti-CD3 Ab. Mean [3H]thymidine incorporation indicated as cpm (±SD) in triplicate wells. Two independent experiments were performed.

Close modal

CD3+CD4+ T cell line 1 (with a moderate inhibitory effect) and line 3 (with no inhibitory effect) were selected for further phenotypic studies using flow cytometry. There was no significant difference in the percentages of GM-CSF-producing cells between these two lines. Small percentages of TGF-β or IL-2-producing CD4+ T cells expressed Foxp3 (Fig. 6 d).

An AFP46–55 CD4+ T cell line was generated from PBMCs of a HCC patient. CD3+CD4+Foxp3neg T cells produced TGF-β, IL-2, and GM-CSF in a peptide-specific manner (data not shown). In an inhibitory assay, T cells from the HCC patient suppressed anti-CD3-induced T cell proliferation. AFP364–373-specific Th1 cells (8) generated from the same patient did not inhibit T cell proliferation in vitro (Fig. 6 e).

Transwell experiments were performed to test whether cell-cell contact is required for AFP46–55 CD4+ T cells to exert their suppressive activity. AFP46–55 CD4+ T cells when cultured in the inner well containing medium with anti-CD3 and the purified APCs did not inhibit the proliferative activity of CD4+ T cells cultured in the outer well containing medium, anti-CD3, and APCs (Fig. 6 f). Taken together, these results indicate that AFP46–55 CD4+ T cells exert its T cell inhibitory properties in a contact-dependent manner.

Short-term T cell lines were generated from PBMCs isolated from 10 healthy donors (6 males and 4 females) and 15 HCC patients (12 males and 3 females), and the frequency of GM-CSF-producing AFP46–55 T cells was analyzed using an intracellular cytokine assay. AFP46–55-specific CD4+ T cells were detected in all healthy donors and HCC patients. A significantly higher frequency of AFP46–55 CD4+ T cells was detected in HCC patients than in healthy controls (p = 0.01; Fig. 7 a), suggesting that these cells are expanded in vivo in response to the tumor Ag. Anti-human TGF-β mAb for an intracellular assay was not available to us when we were performing these experiments and there are no data on TGF-β production by AFP46–55 T cells in this group of patients.

FIGURE 7.

AFP46–55 CD4+ T cells are expanded in peripheral blood of HCC patients. a, The frequency of AFP46–55 CD4+ T cells was determined in HCC patients (n = 15) and healthy donors (n = 10) using an intracellular cytokine assay for GM-CSF. The Mann-Whitney U test (two tailed) reveals a significant difference (p = 0.01) between patients and healthy donors. Frequencies of IFN-γ-producing AFP364–373 CD4+ T cells (b) and GM-CSF-producing AFP46–55 CD4+ T cells (c) were determined before and 3 mo after transarterial chemoembolization/TAE treatment in five HCC patients. The percentages of cytokine-producing cells are shown.

FIGURE 7.

AFP46–55 CD4+ T cells are expanded in peripheral blood of HCC patients. a, The frequency of AFP46–55 CD4+ T cells was determined in HCC patients (n = 15) and healthy donors (n = 10) using an intracellular cytokine assay for GM-CSF. The Mann-Whitney U test (two tailed) reveals a significant difference (p = 0.01) between patients and healthy donors. Frequencies of IFN-γ-producing AFP364–373 CD4+ T cells (b) and GM-CSF-producing AFP46–55 CD4+ T cells (c) were determined before and 3 mo after transarterial chemoembolization/TAE treatment in five HCC patients. The percentages of cytokine-producing cells are shown.

Close modal

We have previously shown that induction of tumor necrosis improved survival of HCC patients and expanded AFP-specific Th1 cells in vivo (9). We analyzed the effects of embolization on the frequency of circulating Th1 (AFP137–145, AFP249–258, AFP364–373) and AFP46–55 CD4+ T cell responses before and 3 mo after the treatment in five consecutive patients. Th1 responses to at least one of the three different AFP-derived epitopes were expanded after treatment in all five patients after embolization. AFP364–373 CD4+ T cells were expanded in four of five patients (Fig. 7,b). AFP46–55 CD4+ T cells were detected before and after treatment in all five patients. The responses ranged from 0.2 to 11% of CD4+ T cells producing peptide-specific GM-CSF (Fig. 7,c). In HCC03, the frequency of AFP46–55 CD4+ T cells was reduced from 8% of CD4+ T cells before transarterial embolization (TAE) (serum AFP = 2625 ng/ml) to 1.5% after TAE (serum AFP = 1640 ng/ml). In HCC04, the response was reduced from 11% of CD4+ T cells before transarterial chemoembolization (serum AFP = 7 ng/ml) to 7% after the treatment (serum AFP = 10 ng/ml; Fig. 7 b). Due to a limited number of cells available from this group of patients, an intracellular cytokine assay for TGF-β or IL-2 was not performed.

The frequency of peptide-specific TGF-β-releasing cells was analyzed using ELISPOT assays for TGF-β. PBMCs isolated from four healthy donors (HD-1, HD-2, HD-3, and HD-4) were stimulated with AFP46–55 or AFP364–373 for 18 h, and the frequency of peptide-specific TGF-β-producing cells was analyzed ex vivo. PBMCs from three of four patients (HD-11, HD-2, and HD-3) responded to AFP46–55 and released TGF-β (Fig. 8). AFP364–373 peptide did not stimulate TGF-β production above the background (cells cultured in medium only) (data not shown).

FIGURE 8.

Detection of AFP46–55-specific TGF-β-releasing cells ex vivo. The frequency of AFP46–55-specific-TGF-β-releasing T cells was analyzed in PBMCs of four healthy donors (HD-1, HD-2, HD-3, and HD-4) ex vivo using an ELISPOT assay for TGF-β. The results are presented as spot-forming units (s.f.u) per 3 × 105 cells and average spots with SD are shown. The experiments were performed in triplicate and the results are representative of two experiments performed on different days.

FIGURE 8.

Detection of AFP46–55-specific TGF-β-releasing cells ex vivo. The frequency of AFP46–55-specific-TGF-β-releasing T cells was analyzed in PBMCs of four healthy donors (HD-1, HD-2, HD-3, and HD-4) ex vivo using an ELISPOT assay for TGF-β. The results are presented as spot-forming units (s.f.u) per 3 × 105 cells and average spots with SD are shown. The experiments were performed in triplicate and the results are representative of two experiments performed on different days.

Close modal

In this study, we show that AFP stimulates the expansion of a subset of CD4+ T cells that produce TGF-β, GM-CSF, and IL-2. These cells express CD3, CD4, CTLA-4, and GITR and exert inhibitory effects on T cell proliferation in a contact-dependent manner. An epitope within the AFP sequence (AFP46–55) that stimulates these CD4+ T cells both ex vivo and in vitro are identified and T cells recognizing AFP46–55 produce a unique cytokine profile, suggesting that these cells differ from Th1, Th2, Th17, or typical Ag-induced Treg that produce IL-10 (15). In independent studies, clones derived from mice that have been orally tolerized with a low Ag dose primarily produce TGF-β, and these cells have been termed Th3 cells. Treg that exclusively produce TGF-β have not been observed in other models (15). To our knowledge, this is the first report describing peptide epitope-specific TGF-β- producing CD4+ T cells in humans. Further studies are required to establish their regulatory effects in vivo and these are planned.

TGF-β is an immunoregulatory cytokine that can act on different populations of leukocytes, including T cells. The inhibitory effects of TGF-β on T cell proliferation is thought to be through the induction of Treg in vitro (14). In our system, the addition of neutralizing anti-TGF-β to an in vitro assay did not abrogate the inhibitory effects of AFP46–55 CD4+ T cells on T cell proliferation. However, this may not reflect on immunoregulatory effects of TGF-β produced by AFP-specific T cells in vivo. Depending on the experimental model used, Treg-mediated suppression appears to occur through CTLA-4, cytokine deprivation, TGF-β and IL-10, either alone or in combination. This may reflect either heterogeneity within the population of Treg or an ability of this population to differentially use suppressor mechanisms depending on the context (16). There is a discrepancy between the role of IL-10 and/or TGF-β in the inhibitory function of Treg in vivo vs in vitro. IL-10 and TGF-β play an important role in suppressive function of Treg in vivo in several models, and neutralization of these cytokines produced by Treg reduce their inhibitory function in vivo. In contrast, IL-10-producing Treg inhibit T cell proliferation in vitro in a contact-dependent manner and the addition of neutralizing anti-IL-10 mAb to an in vitro proliferation assay had no effect on the ability of Treg to mediate suppression (17). The inhibitory role of TGF-β produced by Treg on T cell proliferation in vitro is controversial (18, 19).

The regulatory effects of TGF-β could be altered by the presence of other known or unknown stimulatory or regulatory cytokines (20). It is known that the function of some immune cells from HCC patients is impaired and this may influence the inhibitory effects of TGF-β-producing CD4+ T cells in this group of patients. We are currently studying these factors and will analyze their effects on the function of AFP46–55-specific TGF-β-producing CD4+ T cells.

We have shown that AFP contains distinct epitopes that can generate Th1- and TGF-β-producing CD4+ T cell responses in HCC patients. We believe that the responses to these epitopes are generated (Th1) or expanded (TGF-β-producing CD4+ T cells) in different stages of the disease (HCC). Different treatment modalities may also influence the generation or expansion of these cells. For example, tumor necrosis induces dendritic cell activation and maturation in HCC patients (21) and stimulates tumor-specific Th1 responses (9). In contrast, HCC cells and soluble factors released by tumor cells impair APCs (6) that could favor the development of inducible Treg.

We investigated the possibility that AFP 46–55 is recognized by Th1/Tc1 cells. No AFP46–55 peptide-specific IFN-γ-producing T cells were detected in short-term T cell lines generated from 30 HCC patients and 10 healthy donors (data not shown), suggesting that AFP46–55 does not stimulate Th1/Tc1 cells.

In addition to TGFβ, AFP 46–55 CD4+ T cells also produce IL-2 and GM-CSF. To our knowledge, this is the first report suggesting that IL-2-producing CD4+ T cells may have some regulatory function. There is a discrepancy between the inhibitory or stimulatory roles of IL-2 on Treg in vivo vs in vitro. CD4+CD25+Foxp3+ do not produce IL-2 and the addition of rIL-2 to in vitro T cell cultures neutralizes the inhibitory effects of CD4+CD25+ Treg. However, IL-2 is a key growth/survival factor for Treg in vivo and IL-2-deficient mice bear few Treg and spontaneously develop severe autoimmunity (17). AFP-specific TGF-β-producing CD4+ T cells reported in this study produce IL-2 upon peptide recognition. It is possible but not proven that IL-2 produced by AFP46–55 CD4+ T cells partially abrogates the inhibitory effect of these cells on T cell proliferation in vitro.

It has been shown that GM-CSF enhances protection against tumors and infections, but GM-CSF-deficient mice develop inflammatory disease (22, 23). Many tumors constitutively secrete low levels of GM-CSF, which may be linked with disease progression (24). Moreover, it has been shown that the administration of GM-CSF expands regulatory CD4+CD25+ T cells and suppresses autoimmune diseases in animal models. This suppression is believed to be through activation and generation of regulatory APCs (25, 26, 27, 28, 29, 30).

AFP is an oncofetal Ag and has intrinsic immunoregulatory properties (5, 6, 31, 32, 33) and recombinant AFP is being considered for treatment of autoimmune diseases. The administration of the intact Ag would avoid the selection of specific epitopes to suit MHC-disparate individuals. This is not the case for the AFP-derived epitope identified, since the response to this epitope can be detected in all individuals tested. In this study, donors and patients (30 in total) were not selected based on their HLA haplotypes and determination of HLA class II haplotypes from some patients showed that AFP46–55 T cell responses are detectable in patients with completely different HLA class II haplotypes (data not shown), suggesting that AFP46–55 is a promiscuous epitope and its recognition is not restricted to one HLA class II haplotype. We believe that the processing and presentation of this epitope by APCs can take place via the exogenous pathway, rather than by direct recognition of tumor, because HCCs do not express MHC class II molecules on the cell surface (34) and MHC class II-deficient APCs are unable to present the peptide to AFP46–55 CD4+ T cells.

Although AFP46–55 CD4+ T cells were detected after short-term T cell culture in both HCC patients and healthy donors, the frequency of these cells in HCC patients was significantly higher than that in healthy donors. This suggests that these cells are expanded in vivo in response to the HCC Ag. It has recently been shown that the expansion of circulating Treg is directly associated with poor survival in HCC patients (35). Further studies are required to establish any association between AFP46–55-specific TGF-β-producing CD4+ T cell prevalence with HCC progression and patient survival and to determine the presence of these cells in the tumor.

In support of this notion, we have shown that the reduction of tumor mass reduced the frequency of AFP46–55 CD4+ T cells in patients with expanded AFP46–55 T cells. In contrast, this treatment results in the activation and expansion of IFN-γ-producing Th1-specific CD4+ T cells in the same group of patients. Reduction in tumor burden/regulatory factors by embolization may explain in part the observed concomitant expansion of AFP-specific Th1 and reduction of AFP46–55 CD4+ T cells.

In conclusion, we have identified and characterized self-Ag-specific CD4+ T cells from HCC patients and healthy donors. Their cytokine profile, phenotype, and functional characteristics suggest that these cells are Ag-specific TGF-β- producing CD4+ T cells and that they recognize an AFP peptide as a natural ligand. In this study, we show that tumors may stimulate the expansion of AFP-specific T cells and that the removal of the Ag source reduces the frequency of cells in vivo. These results will be instrumental in the development of peptide-based immunotherapy for treatment of cancer as well as autoimmune disease.

We thank A. Ives for technical assistance.

A patent application based on the findings in this report has been filed by the University College London BioMedica PLC and S. Behboudi is listed as inventor. All other authors declare no competing financial interests.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This study was supported by a project grant from Association for International Cancer Research and the de Laszlo Foundation (to S.Be.). This work was undertaken at the University College London, which received funding from the Health’s Biomedical Research Centers funding scheme.

3

Abbreviations used in this paper: Treg, regulatory T cell; AFP, α-fetoprotein; HCC, hepatocellular carcinoma; GITR, glucocorticoid-induced TNF receptor; HD, healthy donor; TAE, transarterial embolization.

1
Bluestone, J. A., A. K. Abbas.
2003
. Natural versus adaptive regulatory T cells.
Nat. Rev. Immunol.
3
:
253
-257.
2
Beyer, M., J. L. Schultze.
2006
. Regulatory T cells in cancer.
Blood
108
:
804
-811.
3
Nishikawa, H., T. Kato, K. Tanida, A. Hiasa, I. Tawara, H. Ikeda, Y. Ikarashi, H. Wakasugi, M. Kronenberg, T. Nakayama, et al
2003
. CD4+CD25+ T cells responding to serologically defined autoantigens suppress antitumor immune responses.
Proc. Natl. Acad. Sci. USA
100
:
10902
-10906.
4
Franco, A., S. Albani.
2006
. Translating the concept of suppressor/regulatory T cells to clinical applications.
Int. Rev. Immunol.
25
:
27
-47.
5
Murgita, R. A., E. A. Goidl, S. Kontianen, H. Wigzell.
1977
. α-Fetoprotein induces suppressor T cells in vitro.
Nature
267
:
257
-259.
6
Um, S. H., C. Mulhall, A. Alisa, A. R. Ives, J. Karani, R. Williams, A. Bertoletti, S. Behboudi.
2004
. α-Fetoprotein impairs APC function and induces their apoptosis.
J. Immunol.
173
:
1772
-1778.
7
Butterfield, L. H..
2004
. Immunotherapeutic strategies for hepatocellular carcinoma.
Gastroenterology
127
:
S232
-S241.
8
Alisa, A., A. Ives, A. A. Pathan, C. V. Navarrete, R. Williams, A. Bertoletti, S. Behboudi.
2005
. Analysis of CD4+ T-cell responses to a novel α-fetoprotein-derived epitope in hepatocellular carcinoma patients.
Clin. Cancer Res.
11
:
6686
-6694.
9
Ayaru, L., S. P. Pereira, A. Alisa, A. A. Pathan, R. Williams, B. Davidson, A. K. Burroughs, T. Meyer, S. Behboudi.
2007
. Unmasking of α-fetoprotein-specific CD4+ T cell responses in hepatocellular carcinoma patients undergoing embolization.
J. Immunol.
178
:
1914
-1922.
10
Liu, Y., S. Daley, V. N. Evdokimova, D. D. Zdobinski, D. M. Potter, L. H. Butterfield.
2006
. Hierarchy of α-fetoprotein (AFP)-specific T cell responses in subjects with AFP-positive hepatocellular cancer.
J. Immunol.
177
:
712
-721.
11
Wang, H. Y., D. A. Lee, G. Peng, Z. Guo, Y. Li, Y. Kiniwa, E. M. Shevach, R. F. Wang.
2004
. Tumor-specific human CD4+ regulatory T cells and their ligands: implications for immunotherapy.
Immunity
20
:
107
-118.
12
Gehring, A. J., D. Sun, P. T. Kennedy, E. Nolte-’t Hoen, S. G. Lim, S. Wasser, C. Selden, M. K. Maini, D. M. Davis, M. Nassal, A. Bertoletti.
2007
. The level of viral antigen presented by hepatocytes influences CD8 T-cell function.
J. Virol.
81
:
2940
-2949.
13
Sartoris, S., M. T. Valle, A. L. Barbaro, G. Tosi, T. Cestari, A. D’Agostino, A. M. Megiovanni, F. Manca, R. S. Accolla.
1998
. HLA class II expression in uninducible hepatocarcinoma cells after transfection of AIR-1 gene product CIITA: acquisition of antigen processing and presentation capacity.
J. Immunol.
161
:
814
-820.
14
Yamagiwa, S., J. D. Gray, S. Hashimoto, D. A. Horwitz.
2001
. A role for TGF-β in the generation and expansion of CD4+CD25+ regulatory T cells from human peripheral blood.
J. Immunol.
166
:
7282
-7289.
15
Shevach, E. M..
2006
. From vanilla to 28 flavors: multiple varieties of T regulatory cells.
Immunity
25
:
195
-201.
16
Aluvihare, V. R., A. G. Betz.
2006
. The role of regulatory T cells in alloantigen tolerance.
Immunol. Rev.
212
:
330
-343.
17
Sakaguchi, S..
2004
. Naturally arising CD4+ regulatory t cells for immunologic self-tolerance and negative control of immune responses.
Annu. Rev. Immunol.
22
:
531
-562.
18
Piccirillo, C. A., J. J. Letterio, A. M. Thornton, R. S. McHugh, M. Mamura, H. Mizuhara, E. M. Shevach.
2002
. CD4+CD25+ regulatory T cells can mediate suppressor function in the absence of transforming growth factor β1 production and responsiveness.
J. Exp. Med.
196
:
237
-246.
19
Nakamura, K., A. Kitani, W. Strober.
2001
. Cell contact-dependent immunosuppression by CD4+CD25+ regulatory T cells is mediated by cell surface-bound transforming growth factor β.
J. Exp. Med.
194
:
629
-644.
20
Wahl, S. M., J. Wen, N. Moutsopoulos.
2006
. TGF-β: a mobile purveyor of immune privilege.
Immunol. Rev.
213
:
213
-227.
21
Ali, M. Y., C. F. Grimm, M. Ritter, L. Mohr, H. P. Allgaier, R. Weth, W. O. Bocher, K. Endrulat, H. E. Blum, M. Geissler.
2005
. Activation of dendritic cells by local ablation of hepatocellular carcinoma.
J. Hepatol.
43
:
817
-822.
22
Huffman, J. A., W. M. Hull, G. Dranoff, R. C. Mulligan, J. A. Whitsett.
1996
. Pulmonary epithelial cell expression of GM-CSF corrects the alveolar proteinosis in GM-CSF-deficient mice.
J. Clin. Invest.
97
:
649
-655.
23
Gillessen, S., N. Mach, C. Small, M. Mihm, G. Dranoff.
2001
. Overlapping roles for granulocyte-macrophage colony-stimulating factor and interleukin-3 in eosinophil homeostasis and contact hypersensitivity.
Blood
97
:
922
-928.
24
Sotomayor, E. M., Y. X. Fu, M. Lopez-Cepero, L. Herbert, J. J. Jimenez, C. Albarracin, D. M. Lopez.
1991
. Role of tumor-derived cytokines on the immune system of mice bearing a mammary adenocarcinoma: II. Down-regulation of macrophage-mediated cytotoxicity by tumor-derived granulocyte-macrophage colony-stimulating factor.
J. Immunol.
147
:
2816
-2823.
25
Vasu, C., R. N. Dogan, M. J. Holterman, B. S. Prabhakar.
2003
. Selective induction of dendritic cells using granulocyte macrophage-colony stimulating factor, but not fms-like tyrosine kinase receptor 3-ligand, activates thyroglobulin-specific CD4+CD25+ T cells and suppresses experimental autoimmune thyroiditis.
J. Immunol.
170
:
5511
-5522.
26
Gangi, E., C. Vasu, D. Cheatem, B. S. Prabhakar.
2005
. IL-10-producing CD4+CD25+ regulatory T cells play a critical role in granulocyte-macrophage colony-stimulating factor-induced suppression of experimental autoimmune thyroiditis.
J. Immunol.
174
:
7006
-7013.
27
Sheng, J. R., L. Li, B. B. Ganesh, C. Vasu, B. S. Prabhakar, M. N. Meriggioli.
2006
. Suppression of experimental autoimmune myasthenia gravis by granulocyte-macrophage colony-stimulating factor is associated with an expansion of FoxP3+ regulatory T cells.
J. Immunol.
177
:
5296
-5306.
28
Rabinovich, G. A., D. Gabrilovich, E. M. Sotomayor.
2007
. Immunosuppressive strategies that are mediated by tumor cells.
Annu. Rev. Immunol.
25
:
267
-296.
29
Gallina, G., L. Dolcetti, P. Serafini, C. De Santo, I. Marigo, M. P. Colombo, G. Basso, F. Brombacher, I. Borrello, P. Zanovello, et al
2006
. Tumors induce a subset of inflammatory monocytes with immunosuppressive activity on CD8+ T cells.
J. Clin. Invest.
116
:
2777
-2790.
30
Jinushi, M., Y. Nakazaki, M. Dougan, D. R. Carrasco, M. Mihm, G. Dranoff.
2007
. MFG-E8-mediated uptake of apoptotic cells by APCs links the pro- and antiinflammatory activities of GM-CSF.
J. Clin. Invest.
117
:
1902
-1913.
31
Murgita, R. A., T. B. Tomasi, Jr.
1975
. Suppression of the immune response by α-fetoprotein on the primary and secondary antibody response.
J. Exp. Med.
141
:
269
-286.
32
Murgita, R. A..
1976
. The immunosuppressive role of α-fetoprotein during pregnancy.
Scand. J. Immunol.
5
:
1003
-1014.
33
Murgita, R. A., D. C. Hooper, M. Stegagno, T. L. Delovitch, H. Wigzell.
1981
. Characterization of murine newborn inhibitory T lymphocytes: functional and phenotypic comparison with an adult T cell subset activated in vitro by α-fetoprotein.
Eur. J. Immunol.
11
:
957
-964.
34
Matoba, K., N. Iizuka, T. Gondo, T. Ishihara, H. Yamada-Okabe, T. Tamesa, N. Takemoto, K. Hashimoto, K. Sakamoto, T. Miyamoto, et al
2005
. Tumor HLA-DR expression linked to early intrahepatic recurrence of hepatocellular carcinoma.
Int. J. Cancer
115
:
231
-240.
35
Fu, J., D. Xu, Z. Liu, M. Shi, P. Zhao, B. Fu, Z. Zhang, H. Yang, H. Zhang, C. Zhou, et al
2007
. Increased regulatory T cells correlate with CD8 T-cell impairment and poor survival in hepatocellular carcinoma patients.
Gastroenterology
132
:
2328
-2339.