Data have been reported on the in vivo adjuvant role of soluble lymphocyte activation gene-3 (LAG-3) recombinant protein in mouse models and on its ability to support the in vitro generation of human, tumor-specific CTLs. In this study, we show that soluble human rLAG-3 protein (hLAG-3Ig) used in vitro as a single maturation agent induces phenotypic maturation of monocyte-derived dendritic cells and promoted the production of chemokines and TNF-α inflammatory cytokine. When given in association with optimal or suboptimal doses of CD40/CD40L, hLAG-3Ig functions as a strong costimulatory factor and induces full functional activation of monocyte-derived dendritic cells that includes the production of high level of IL-12p70. Moreover, evidence is here provided that this costimulatory function licensing dendritic cells to produce IL-12p70 is also a functional property of LAG-3 molecules when expressed in a physiological context by CD4+ activated T cells. Altogether, these data show for the first time a role of LAG-3 in mediating dendritic cell activation when expressed on the T cell surface or released after specific Ag stimulation in the interspaces of immunological synapses.

The onset of strong adaptive immunity crucially depends on the activation of dendritic cells (DC)4. It is known that the process of DC activation is indeed a continuous integration program of stimuli that gradually adjust the DC status to host environment (1). Under physiological conditions, in a firmly regulated time frame, DC are progressively exposed to different maturation stimuli that include microbial agents, inflammatory cytokines, and interaction with activated T cells inside lymph nodes. Although the up-regulation of costimulatory molecules can be induced by any of these activation pathways, the release of IL-12p70 requires the integration of multiple signals including CD40/CD40L triggering (2, 3), which represents a limiting crucial point for licensing DC to promote Th1 response and CTL priming (4, 5). Moreover, it has been shown that strong and sustained CD40 signaling is required to enable DC to release cytokines while weak and transient receptor activation results only into phenotypic up-regulation of cell surface activation markers (6). In several pathologic conditions such as in tumors, CD40/CD40L signaling is limited or impaired (7, 8) and a suboptimal triggering of this DC activation pathway may elicit an IL-10-predominant response hampering the on setting of host-protective, memory T cells (9). In the absence of the exogenous TLR agonist, such as in the case of an antitumor response, there is therefore the need to integrate the weak CD40/CD40L signaling to achieve optimal DC activation. In addition to CD40L, activated T cells also express a second adaptive immunity ligand, namely the lymphocyte activation gene-3 (LAG-3 or CD223). LAG-3 binds MHC class II molecules with high affinity and induces their internalization. LAG-3/MHC class II interaction is associated to cytokine production by APC (10) and, in its soluble form, LAG-3 recombinant protein binds to MHC class II molecules on APC, inducing their maturation and migration to secondary lymphoid organs (11).

Several data have been reported on the ability of human rLAG-3 protein (hLAG-3Ig) to work as an adjuvant in in vivo mouse model (12, 13) and we and others have recently showed that soluble rhLAG-3Ig is able to enhance the in vitro induction of viral- and tumor-specific CTL in humans (14, 15).

In the present study, we show that LAG-3, either as recombinant soluble protein or in its physiological, natural form expressed by activated T cells, works as a costimulatory molecule and, in association to CD40/CD40L interaction, provides a signal crucial for the achievement of fully functional maturation state of DC-releasing Th1-priming cytokines. In addition, to clarify a physiological function of LAG-3, our data further validate the use of this protein as a potent adjuvant in immunotherapeutic approaches.

A clinical-grade, highly pure recombinant soluble hLAG-3Ig was used in this study identified as the product IMP321 provided by Immutep. The recombinant soluble human LAG-3 fusion protein (hLAG-3Ig) was provided by Immutep. Briefly, Chinese hamster ovary DHFR cells were transfected with a plasmid coding for the D1–D4 extracellular domains of hLAG-3 fused to the Fc tail of a hIgG1 (16). A production clone was selected after amplification in methothrexate. The purified protein has a concentration of 1.89 mg/ml and 0.75 endotoxin units/mg endotoxin as determined by the Limulus amebocyte lysate assay (BioWhittaker). In addition, boiling hLAG-3Ig for 5 min abrogated its biological activity, evaluated as the ability to up-regulate CD83 expression on monocytes purified from PBMCs of healthy donors.

NIH3T3 cells expressing human CD40L (NIH3T3/CD40L) were provided by C. Traversari (Molmed, Milan, Italy). TB39 is a CD4+ T cell clone recognizing a mutated, melanoma-specific Ag (17). Conventional CD4+ T cells were generated by in vitro activation of CD4+CD25 T cells for 10 days in the presence of 1 μg/ml OKT3 anti-CD3 Ab (Orthoclone; Ortho-Biotech) and IL-2. CD4+CD25 T cells were purified from PBMC of healthy donors by immunomagnetic beads (Miltenyi Biotec). Cell surface expression of LAG-3 was detected by FITC anti-LAG-3 mAb purchased from Alexis Biochemicals.

The melanoma-specific T cell clone and conventional CD4+ T cells display an effector-memory phenotype. TB39, originally isolated from a tumor invaded lymph node and maintained in culture in the presence of IL-2 and weekly stimulated with the autologous tumor, is CD45RACCR7CD62L. Conventional CD4+ T cells cultured in vitro with OKT3 and IL-2 display a CD45RA, CCR7CD62L+/− phenotype as assessed by immunofluorescence and FACS analysis (data not shown) (BD Immunocytometry Systems).

Upon informed consent, healthy donors’ blood samples were collected and monocytes were purified from PBMC by immunomagnetic beads (Monocyte Isolation kit II; Miltenyi Biotec), resuspended and cultured for 6 days in RPMI 1640, 10% FCS, 2 mM l-glutamine (Cambrex) with 50 ng/ml GM-CSF (Myelogen; Schering-Plough) and 20 ng/ml IL-4 (PeproTech). DC maturation was induced by overnight culture with 16 μg/ml hLAG-3Ig, 2 μg/ml LPS 0111:B4 (Sigma-Aldrich) or with equal molar concentration of human IgG1 purified protein or human IgG Fc fragment purified protein (Chemicon International). DC maturation was also induced by coculture with gamma-irradiated (10,000 rad) NIH3T3/CD40L cells at various ratios. Cell surface expression of CD40, CD80, CD83, and CD86 molecules was analyzed by using PE-conjugated mAb (BD Biosciences).

Cytokine and chemokine release was analyzed by the BD Biosciences CBA following the manufacturer’s instructions; data were acquired using a FACScan flow cytometer (BD Biosciences) and analyzed using BD Biosciences CBA Software.

UV-triggered apoptosis was induced on a human melanoma cell line (Me 15392) by treating cells with a 60 mJ UVB lamp at a dose of 200 mJ/cm2/s. After 4 h at 37°C, cells were stained with PKH67 green fluorescent dye (Sigma Aldrich) and after 24 h of culture were incubated for 4 h with immature, LPS- or hLAG-3Ig-treated monocyte-derived dendritic cells (MoDC) at 1:1 ratio. Cells were stained with PE anti-human CD11c mAb (BD Biosciences) and DC that have phagocytosed apoptotic melanoma cells were defined by the percentage of double-positive events (CD11c+PKH67+).

Cell migration was evaluated by using a 24-well chemotaxis chamber (Corning). One hundred microliters of cell suspensions (7 × 106/ml in conditioned medium) were added to the upper wells, while medium alone or containing CCL5 or CCL19 at a concentration of 100 ng/ml was added to the lower chambers. After 90 min of incubation at 37°C with 5% CO2, the transwell inserts were removed and cells migrated to the lower chambers were counted. Results are expressed as the mean ± SEM for at least four different experiments.

Immature MoDC (iMoDC) were cocultured for 24 h at a 4:1 ratio with the HLA-matched T cells. In the case of TB39, MoDC were also loaded with 2 μg/ml of specific peptide (PYYFAAELPPRNLPEP) (17). For blocking experiments, T cells were preincubated for 2 h with 10 μg/ml blocking anti-CD40L (clone MK13A4; Acris Antibodies), of anti-LAG-3 Ab (clone 17B4; Alexis Biochemicals), or of purified mouse IgG1κ (BD Biosciences).

Equal volumes of supernatants were incubated with anti-LAG-3 17B4 mAb overnight at 4°C and then for additional 4 h with Sepharose-G (GE Healthcare Europe). Immunoprecipitated proteins were then separated by SDS-PAGE. Blots were blocked in PBS containing 10% milk and 0.1% Tween 20, stained with 17B4 mAb and developed using ECL (GE Healthcare Europe).

TB39 melanoma-specific T lymphocytes were incubated at a 1:1 ratio with an autologous lymphoblastoid cell line alone or loaded with 2 μg/ml of specific peptide; coculture was performed overnight. Conventional CD4+ T cells were stimulated in vitro for 10 days in the presence of 50 IU/ml IL-2, then seeded at a concentration of 2 × 106/ml and cultured for an additional 24 h. Supernatants were then collected and LAG-3 molecules were detected by specific ELISA (Apotech).

We recently showed that hLAG-3Ig enhanced the in vitro induction of viral- and tumor-specific CTL. In this setting, APC of PBMC exposed in vitro to peptide and hLAG-3Ig displayed a more mature phenotype than APC of PBMC exposed to peptide only, suggesting that the hLAG-3Ig adjuvant effect could have been dependent on activation of circulating APC (14). To evaluate whether hLAG-3Ig worked by a direct mechanism in the induction of DC maturation, iMoDC (100% CD11c+ and CD1a+, data not shown) were matured with hLAG-3Ig and assessed for their phenotype. hLAG-3Ig-treated MoDC displayed a broad up-regulation of CD40, CD80, CD83, and CD86 cell surface activation markers and these effects were comparable to those obtained by LPS, although a degree of variability could be observed among the MoDC generated from 20 different healthy donors’ PBMC (Fig. 1, A and B). As controls, human IgG1 molecules or Fc fragments were used at the same molar concentration of hLAG-3Ig (Fig. 1, C and D) and no effect on cell surface expression of DC activation markers was detected.

FIGURE 1.

Soluble human LAG-3 induces partial phenotypic maturation of human MoDC. Immunofluorescence and FACS analysis of CD40, CD80, CD83, and CD86 expression by iMoDC or MoDC matured overnight with 16 μg/ml hLAG-3Ig or 2 μg/ml LPS (A and B). Human IgG1 molecules or Fc fragments were used at equal molar ratio as controls (C and D). Mean fluorescence intensity (MFI) and percentage of expression for each marker are indicated. Figure reports the analysis of 4 of the 20 healthy donors’ MoDC. Higher doses of hLAG-3Ig did not amplify this maturation (data not shown).

FIGURE 1.

Soluble human LAG-3 induces partial phenotypic maturation of human MoDC. Immunofluorescence and FACS analysis of CD40, CD80, CD83, and CD86 expression by iMoDC or MoDC matured overnight with 16 μg/ml hLAG-3Ig or 2 μg/ml LPS (A and B). Human IgG1 molecules or Fc fragments were used at equal molar ratio as controls (C and D). Mean fluorescence intensity (MFI) and percentage of expression for each marker are indicated. Figure reports the analysis of 4 of the 20 healthy donors’ MoDC. Higher doses of hLAG-3Ig did not amplify this maturation (data not shown).

Close modal

To characterize the functional status of hLAG-3Ig-treated MoDC, cytokine and chemokine release was evaluated by CBA assays (Table I). As expected, incubation of iMoDC with LPS dramatically enhanced the production of chemokines and proinflammatory cytokines. hLAG-3Ig treatment increased the amount of CXCL10, CXCL9, CCL2, and CCL8 detectable in the medium and promoted the release of TNF-α although the amount of this cytokine showed some degree of variability among the tested individuals (Table I). hLAG-3Ig-matured DC consistently failed to produce IL-12p70 in agreement with the notion that, to achieve the ability to induce a polarized Th1 response, DC maturation require the integration of multiple signals that only occasionally can be replaced by a sustained TLR triggering (18). Treatment of DC with human IgG1 molecules or Fc fragments did not induce any changes in the release of chemokines and proinflammatory cytokines (Table I, see donors C and D).

Table I.

Cytokine and chemokine production by differently matured human MoDC

CellsChemokine/Cytokine Release (pg/ml)a
IL-12p70TNF-αIL-10IL-6CXCL8CXCL10CCL2CXCL9CCL5
Donor A iMoDC 1.321 ± 28 58 ± 6 102 ± 7 264 ± 14 169 ± 11 
 DC + hLAG-3Ig 456 ± 21 11 ± 2 17 ± 3 >5.000 1.755 ± 81 1.201 ± 45 3.355 ± 152 218 ± 17 
 DC + LPS 122 ± 7 >5.000 419 ± 17 >5.000 >5.000 >5.000 1.672 ± 61 >5.000 4.611 ± 143 
Donor B iMoDC 14 ± 1 564 ± 25 637 ± 26 8 ± 1 505 ± 12  
 DC + hLAG-3Ig 133 ± 6 17 ± 3 13 ± 1 >5.000 36 ± 4 1.538 ± 84 31 ± 3 521 ± 35 
 DC + LPS 1.823 ± 27 >5.000 1117 ± 54 >5.000 >5.000 538 ± 31 1.595 ± 47 44 ± 6 3.830 ± 140 
Donor C iMoDC 28 ± 5 47 ± 5 3045 ± 215 392 ± 28 91 ± 3 61 ± 4 22 ± 5 
 DC + hIgG1 22 ± 3 42 ± 6 2771 ± 109 325 ± 15 101 ± 16 40 ± 4 22 ± 4 
 DC + LPS 360 ± 4 >5.000 183 ± 11 >5.000 >5.000 >5.000 1727 >5.000 >5.000 
Donor D iMoDC 23 ± 2 3482 ± 122 789 ± 14 1441 ± 32 380 ± 12 52 ± 5 
 DC + Fc 26 ± 3 47 ± 7 3413 ± 187 840 ± 31 1252 ± 54 251 ± 19 46 ± 5 
 DC + LPS 2439 ± 51 >5.000 1000 ± 42 >5.000 >5.000 >5.000 >5.000 >5.000 >5.000 
CellsChemokine/Cytokine Release (pg/ml)a
IL-12p70TNF-αIL-10IL-6CXCL8CXCL10CCL2CXCL9CCL5
Donor A iMoDC 1.321 ± 28 58 ± 6 102 ± 7 264 ± 14 169 ± 11 
 DC + hLAG-3Ig 456 ± 21 11 ± 2 17 ± 3 >5.000 1.755 ± 81 1.201 ± 45 3.355 ± 152 218 ± 17 
 DC + LPS 122 ± 7 >5.000 419 ± 17 >5.000 >5.000 >5.000 1.672 ± 61 >5.000 4.611 ± 143 
Donor B iMoDC 14 ± 1 564 ± 25 637 ± 26 8 ± 1 505 ± 12  
 DC + hLAG-3Ig 133 ± 6 17 ± 3 13 ± 1 >5.000 36 ± 4 1.538 ± 84 31 ± 3 521 ± 35 
 DC + LPS 1.823 ± 27 >5.000 1117 ± 54 >5.000 >5.000 538 ± 31 1.595 ± 47 44 ± 6 3.830 ± 140 
Donor C iMoDC 28 ± 5 47 ± 5 3045 ± 215 392 ± 28 91 ± 3 61 ± 4 22 ± 5 
 DC + hIgG1 22 ± 3 42 ± 6 2771 ± 109 325 ± 15 101 ± 16 40 ± 4 22 ± 4 
 DC + LPS 360 ± 4 >5.000 183 ± 11 >5.000 >5.000 >5.000 1727 >5.000 >5.000 
Donor D iMoDC 23 ± 2 3482 ± 122 789 ± 14 1441 ± 32 380 ± 12 52 ± 5 
 DC + Fc 26 ± 3 47 ± 7 3413 ± 187 840 ± 31 1252 ± 54 251 ± 19 46 ± 5 
 DC + LPS 2439 ± 51 >5.000 1000 ± 42 >5.000 >5.000 >5.000 >5.000 >5.000 >5.000 
a

Results of inflammation and chemokine CBA assays of supernatants from healthy donors’ MoDC stimulated with hLAG-3Ig, LPS, or human IgG1 and Fc fragment as controls. Assay sensitivity: 10 pg/ml. SD calculated on three replicates was <10% of the indicated value. Values below the sensitivity threshold were reported to 0; all the values of IL-1β were 0.

Consistent with the phenotypic profile and the cytokine/chemokine pattern described above showing that hLAG-3Ig-treated MoDC are indeed activated but not fully matured, MoDC exposed overnight to hLAG-3Ig recombinant protein marginally down-regulated their capacity of Ag uptake and showed an intermediate ability to phagocytose apoptotic melanoma cells as compared with iMoDC and to LPS-treated MoDC (Fig. 2). To better characterize the functional activity of these MoDC, their migration properties were assessed by chemotaxis assays. As expected, LPS-matured DC strongly migrated in response to CCL19 and immature MoDC migrated in response to CCL5 (19); hLAG-3Ig-treated MoDC were able to migrate in response to CCL19 attraction, while they did not migrate in response to CCL5 (Fig. 3).

FIGURE 2.

Phagocytosis of UV-treated melanoma cells by dendritic cells. iMoDC or LPS- or hLAG-3Ig-matured MoDC incubated for 4 h at 1:1 ratio with PKH67-labeled apoptotic human melanoma cells were stained with PE anti-CD11c mAb. DC that phagocytosed apoptotic melanoma cells were identified as double-positive cells. Numbers reported in each upper right quadrant indicate the percentage of cells that phagocytosed PKH67+ apoptotic bodies. Results are representative of three independent experiments.

FIGURE 2.

Phagocytosis of UV-treated melanoma cells by dendritic cells. iMoDC or LPS- or hLAG-3Ig-matured MoDC incubated for 4 h at 1:1 ratio with PKH67-labeled apoptotic human melanoma cells were stained with PE anti-CD11c mAb. DC that phagocytosed apoptotic melanoma cells were identified as double-positive cells. Numbers reported in each upper right quadrant indicate the percentage of cells that phagocytosed PKH67+ apoptotic bodies. Results are representative of three independent experiments.

Close modal
FIGURE 3.

Chemotaxis of hLAG-3Ig-treated MoDC. iMoDC (□), hLAG-3Ig- (▪), or LPS-matured MoDC (▦) were tested for migration in response to CCL5 and CCL19 chemokines. DC were added to the upper chambers of transwell inserts and cells that migrated into the lower chamber were enumerated after 90 min of incubation in conditioned medium in the absence or presence of chemokines in the lower chamber. Chemokines were used at a concentration of 100 ng/ml.

FIGURE 3.

Chemotaxis of hLAG-3Ig-treated MoDC. iMoDC (□), hLAG-3Ig- (▪), or LPS-matured MoDC (▦) were tested for migration in response to CCL5 and CCL19 chemokines. DC were added to the upper chambers of transwell inserts and cells that migrated into the lower chamber were enumerated after 90 min of incubation in conditioned medium in the absence or presence of chemokines in the lower chamber. Chemokines were used at a concentration of 100 ng/ml.

Close modal

To evaluate whether hLAG-3Ig could synergize with other molecules and lead to fully DC activation in PBMC, attention has been focused on CD40/CD40L(CD154) pathway, for which a key role in DC function and activation is well known (20, 21).

As expected, CD40 triggering on iMoDC induced dose-dependent up-regulation of CD83 expression (Fig. 4) that was further enhanced by hLAG-3Ig treatment when given in association to suboptimal doses of hCD40L (Fig. 4, dashed bars DC:NIH3T3/CD40L 8:1 and 16:1). No additional effect of hLAG-3Ig treatment was seen at saturating doses of CD40L. Surface expression of CD80, HLA-DR, and CD86 was also investigated; however, no significant differences were detected, due to the high levels achieved by these activation markers even at low doses of NIH3T3/CD40L alone (data not shown).

FIGURE 4.

Soluble LAG-3 acts as a costimulatory factor in association to hCD40L. Immature HLA-A2+ MoDC were cocultured with various ratios of gamma-irradiated (10,000 rad) NIH3T3 cells stably transfected with hCD40L (NIH3T3/CD40L) in the absence or in the presence of 16 μg/ml hLAG-3Ig. DC, gated as HLA-A2+ cells, were analyzed for the surface expression of CD83. Data presented are representative of three different experiments.

FIGURE 4.

Soluble LAG-3 acts as a costimulatory factor in association to hCD40L. Immature HLA-A2+ MoDC were cocultured with various ratios of gamma-irradiated (10,000 rad) NIH3T3 cells stably transfected with hCD40L (NIH3T3/CD40L) in the absence or in the presence of 16 μg/ml hLAG-3Ig. DC, gated as HLA-A2+ cells, were analyzed for the surface expression of CD83. Data presented are representative of three different experiments.

Close modal

To further explore hLAG-3Ig/CD40L costimulation, supernatants of MoDC induced to mature with different doses of CD40L alone or together with hLAG-3Ig were analyzed for the production of proinflammatory cytokines by CBA assay (Table II). Addition of hLAG-3Ig induced a dramatic enhancement in IL-12p70, TNF-α, IL-10, and IL-6 release at the maximal dose of CD40L treatment (1:1 DC:NIH3T3/CD40L ratio). Increased cytokine production was also detectable at suboptimal doses of CD40L with the synergistic effect mediated by hLAG-3Ig clearly evident for TNF-α and still detectable for IL-12p70, IL-10, and IL-6 at ratios of 4:1 and 8:1 (Table II).

Table II.

hLAG-3Ig acts as a costimulatory factor in association to hCD40L

CytokineCytokine Release (pg/ml)a
iMoDCiMoDC Cocultured with Gamma-irradiated NIH3T3/CD40L at the Ratio:
1:14:18:116:1
hLAG-3Ig − − − − − 
IL-12p70 622 ± 22 1.006 ± 43 14 ± 1 108 ± 5 31 ± 3 
TNF-α 35 ± 1 2.897 ± 46 >5.000 99 ± 1 3.229 ± 32 79 ± 3 2.769 ± 193 36 ± 2 599 ± 52 
IL-10 501 ± 18 2.283 ± 103 17 ± 1 244 ± 0 59 ± 3 11 ± 2  
IL-6 2.137 ± 29 3.465 ± 141 166 ± 2 319 ± 12 85 ± 6 91 ± 2 45 ± 1 23 ± 2 
CytokineCytokine Release (pg/ml)a
iMoDCiMoDC Cocultured with Gamma-irradiated NIH3T3/CD40L at the Ratio:
1:14:18:116:1
hLAG-3Ig − − − − − 
IL-12p70 622 ± 22 1.006 ± 43 14 ± 1 108 ± 5 31 ± 3 
TNF-α 35 ± 1 2.897 ± 46 >5.000 99 ± 1 3.229 ± 32 79 ± 3 2.769 ± 193 36 ± 2 599 ± 52 
IL-10 501 ± 18 2.283 ± 103 17 ± 1 244 ± 0 59 ± 3 11 ± 2  
IL-6 2.137 ± 29 3.465 ± 141 166 ± 2 319 ± 12 85 ± 6 91 ± 2 45 ± 1 23 ± 2 
a

Assay sensitivity: 10 pg/ml. Results are representative of three different experiments and values below the sensitivity threshold were reported to 0.

The observation that hLAG-3Ig exerts costimulatory effect on DC activation led us to study its role in a physiological context. Because LAG-3 is expressed at high levels by activated CD4+ T cells, we explored the possibility that LAG-3 could cooperate with CD40L expressed by human T cells in inducing DC maturation. A tumor-specific CD4+ T cell clone, TB39, recognizing an HLA-class II presented epitope derived from a mutated receptor-like protein tyrosine phosphatase κ (17) together with conventional CD4+ T cell lines activated in vitro with anti-CD3 mAb and IL-2 were used to this purpose. Results reported in Fig. 5 showed that all these activated T cells did express LAG-3 protein on their cell surface (Fig. 5, A and B, upper panels). Because it has been previously reported that LAG-3 could be shed from the membrane by protease-specific mechanisms in mice (22) and could be secreted as a truncated soluble form through an alternative splicing event of the mRNA in humans (23, 16), the release of a LAG-3 soluble form by these T cells was also investigated. Stimulation of T cell clone TB39 with peptide-loaded autologous lymphoblastoid cell line induced about a two-fold increase in the release of soluble LAG-3 if compared with stimulation with LCL alone (Fig. 5,A, lower panel). Conventional CD4+ T cells upon 10 day of in vitro activation with IL-2 and OKT3 also showed a strong release of LAG-3 soluble protein (Fig. 5,B, lower panel). Moreover, immunoprecipitation and immunoblotting of these CD4+ T cell-conditioned supernatants revealed the presence of LAG-3 protein with the expected m.w. (Fig. 5 C). Overall, these data indicate that LAG-3 protein is expressed by activated T cells and that it could be released at detectable levels in cell culture supernatants upon TCR activation.

FIGURE 5.

Ag-specific and conventional CD4+ T cells express LAG-3 protein on their surface and release a soluble form of the protein in response to TCR stimulation. CD4+ melanoma-specific T cell clone TB39 (A, upper panel) and conventional CD4+ T cells (B, upper panel) were analyzed for the surface expression of LAG-3 protein. Filled histograms, LAG-3 expression as detected by 17B4 mAb; open histograms, isotype-matched mAb control. Medium conditioned by the overnight growth of TB39 T cell clone stimulated with autologous LCL alone or loaded with the antigenic peptide (A, lower panel) and medium conditioned by not activated or in vitro activated conventional CD4+ T cells were analyzed by ELISA for the content of soluble LAG-3. C, Medium unconditioned (lane 1) or conditioned by conventional CD4+ T cells (lane 2) was immunoprecipitated with 17B4 mAb and then analyzed by Western blot for the content of soluble LAG-3.

FIGURE 5.

Ag-specific and conventional CD4+ T cells express LAG-3 protein on their surface and release a soluble form of the protein in response to TCR stimulation. CD4+ melanoma-specific T cell clone TB39 (A, upper panel) and conventional CD4+ T cells (B, upper panel) were analyzed for the surface expression of LAG-3 protein. Filled histograms, LAG-3 expression as detected by 17B4 mAb; open histograms, isotype-matched mAb control. Medium conditioned by the overnight growth of TB39 T cell clone stimulated with autologous LCL alone or loaded with the antigenic peptide (A, lower panel) and medium conditioned by not activated or in vitro activated conventional CD4+ T cells were analyzed by ELISA for the content of soluble LAG-3. C, Medium unconditioned (lane 1) or conditioned by conventional CD4+ T cells (lane 2) was immunoprecipitated with 17B4 mAb and then analyzed by Western blot for the content of soluble LAG-3.

Close modal

To evaluate the role of LAG-3 expressed by activated T cells in T-DC interactions, CD4+ T cell clone TB39 (Fig. 6,A) and conventional CD4+ T cells (Fig. 6, B and C) were coincubated with iMoDC generated from HLA-DR-matched healthy donor and phenotypic and functional activation was analyzed in the presence or absence of anti-LAG-3 or anti-CD40L specific mAb. A considerable up-regulation of CD83 expression was induced by activated conventional CD4+ or by TB39 T cells. Notably, in the case of TB39 T cell clone, DC maturation was only achieved when DC were loaded with its nominal peptide and therefore when the TB39 was antigenically activated (Fig. 6 A). Addition of anti-CD40L mAb in the assay led to a strong reduction of CD83 expression for all the different T cells examined. Anti-LAG-3-blocking mAb did induce a 35% reduction of CD83 expression in the case of the anti-melanoma TB39, but it did not affected the CD83 levels for DC matured by conventional CD4+ T cells. This result suggests that in DC maturation induced by conventional CD4+ T cells, CD40L signaling reaches a saturating point and the percentage of CD83+ DC could not be further improved by the action of LAG-3. LAG-3 is instead crucial for TB39 activation because CD40L signaling mediated by these antitumor T cells is likely less efficient as compared with that delivered by conventional CD4+ T cells and required integration to induce an optimal cell surface expression of MoDC costimulatory molecules.

FIGURE 6.

LAG-3 expressed on activated and Ag-specific T cells exerts costimulatory activity on DC maturation. Tumor-specific CD4+ T cell clone TB39 (A) and conventional CD4+ T cell lines derived from two healthy donors (B and C) were incubated at 1:4 ratio with HLA-matched MoDC alone or loaded with the antigenic peptide (A) or with allogenic MoDC (B and C). Incubation was performed in the absence or in the presence of blocking anti-CD40L, anti-LAG-3, or mouse IgG1 mAb as a control. After 24 h of coculture, cells were harvested and the expression of CD83 molecule was evaluated on DC cells after gating on the CD4+ cell subset.

FIGURE 6.

LAG-3 expressed on activated and Ag-specific T cells exerts costimulatory activity on DC maturation. Tumor-specific CD4+ T cell clone TB39 (A) and conventional CD4+ T cell lines derived from two healthy donors (B and C) were incubated at 1:4 ratio with HLA-matched MoDC alone or loaded with the antigenic peptide (A) or with allogenic MoDC (B and C). Incubation was performed in the absence or in the presence of blocking anti-CD40L, anti-LAG-3, or mouse IgG1 mAb as a control. After 24 h of coculture, cells were harvested and the expression of CD83 molecule was evaluated on DC cells after gating on the CD4+ cell subset.

Close modal

Supernatants of these same DC-T cell cocultures were collected and analyzed for the release of proinflammatory cytokines (Fig. 7). MoDC, after overnight coculture with peptide-activated TB39 melanoma-specific T cell line or conventional CD4+ T cells, released significant levels of IL-12p70, TNF-α, and IL-6 and, to a lesser extent, of IL-10. Interestingly, the addition of anti-LAG-3 mAb significantly reduced the amount of cytokines detected in the supernatant either for TB39 (Fig. 7,A) or for conventional CD4+ T cell-treated DC (Fig. 7, B and C) with this effect more evident for cytokine such as IL-12p70, TNF-α, and IL-6. As expected, hindering of CD40L signaling strongly limited cytokine release. Altogether, these data indicate that maturation of MoDC achieved by T-DC interaction also involved the active role of LAG-3-mediated signaling that synergizes with the well-known CD40/CD40L pathway in licensing DC to produce Th1 cytokines.

FIGURE 7.

LAG-3 expressed on activated and Ag-specific T cells exerts costimulatory activity on DC maturation in terms of cytokine release. Tumor-specific CD4+ T cell line TB39 (A) and conventional CD4+ T cells derived from two healthy donors (B and C) were incubated as described (see legend to Fig. 6). After 24 h of coculture, supernatants were collected and analyzed for the content of the indicated cytokines by CBA assay. Results are expressed as picograms per milliliter; assay sensitivity: 10 pg/ml; SD calculated on three replicates were <10% of the indicated value. IL-1β levels were under the limits of sensitivity of the assay in all the conditions tested.

FIGURE 7.

LAG-3 expressed on activated and Ag-specific T cells exerts costimulatory activity on DC maturation in terms of cytokine release. Tumor-specific CD4+ T cell line TB39 (A) and conventional CD4+ T cells derived from two healthy donors (B and C) were incubated as described (see legend to Fig. 6). After 24 h of coculture, supernatants were collected and analyzed for the content of the indicated cytokines by CBA assay. Results are expressed as picograms per milliliter; assay sensitivity: 10 pg/ml; SD calculated on three replicates were <10% of the indicated value. IL-1β levels were under the limits of sensitivity of the assay in all the conditions tested.

Close modal

Phenotypic maturation of DC and production of proinflammatory cytokines such as IL-12p70 are critical checkpoints for the induction and regulation of the immune response. These two features of DC maturation are differently regulated and, whereas up-regulation of MHC and costimulatory molecules is more easily induced by a variety of different stimuli, the cytokine production, mainly IL-12p70, depends on the strength, persistence, and type of ligand-induced triggering and requires the cooperation and synergistic action of multiple maturation stimuli. In a temporal window of signal integration, the CD40-mediated signaling, occurring in lymph nodes through T cell/DC interaction, complements those received by TLR and leads to the generation of effector DC producing high levels of IL-12p70 and supporting the induction of Th1 responses (1, 2, 6).

In the present study, we show that signaling triggered by LAG-3 cooperates with CD40L in licensing DC to produce IL-12p70. Our data indicate that LAG-3, physiologically expressed by activated CD4+ T cells, is indeed a crucial component of T cells/DC interaction and that LAG-3-mediated signaling is an essential activation pathway for boosting the amount of IL-12p70 released by T cell-matured DC.

Taking advantage of recombinant soluble hLAG-3Ig, we were able to demonstrate that up-regulation of costimulatory molecules and acquisition of the ability to migrate in response to CCL19 can be induced by LAG-3 provided to MoDC as single stimulus. hLAG-3Ig-matured MoDC were boosted to release high level of a subset of chemokines and TNF-α but failed to produce IL-12p70. In agreement with this partial maturation state, hLAG-3Ig-matured MoDC still retain the ability to phagocytose apoptotic cells. However, our data show that recombinant hLAG-3Ig works as a strong stimulatory molecule in association with CD40L. When given in association with optimal or suboptimal doses of CD40L, hLAG-3Ig functions as a strong costimulatory factor and induces full functional activation of MoDC that includes the production of high level of IL-12p70. This property of acting as a costimulatory molecule together with CD40L for DC maturation is also a feature belonging to LAG-3 molecules when physiologically expressed by activated CD4+ T cells. Our data indicate for the first time that LAG-3 molecules expressed on the T cell surface or released upon specific antigenic stimulation are active players in the T cell/DC interaction whose output is the licensing of DC to IL-12 production. Taking advantage of conventional CD4+ activated T cells and Ag-specific CD4+ T cell clone recognizing a mutated melanoma Ag, we were able to show that the amount of IL-12p70 released by DC matured by T cell coculture is dependent by the availability of LAG-3 molecules to interact with the HLA-class II molecules expressed by DC. Most importantly, as shown for the anti-melanoma T cell clone, the positive outcome of DC maturation is strictly dependent by the Ag-specific stimulation of the T cell clone. Interestingly, as for CD40 and its soluble form (24, 25), the shedding of LAG-3 from the cell surface membrane is also dependent on antigenic stimulation of TCR and TCR triggering leads to a strong release of LAG-3 in the culture medium and, likely, also into the interspaces of immunological synapses. In our experimental setting, we are unable to discriminate in which form LAG-3 participates to DC activation and if this molecule works when expressed on the cell surface, present in the cellular interspaces as a soluble factor or both. It has been shown that the murine natural soluble form of LAG-3 molecules is not able to bind with sufficient affinity MHC class II molecules and that the murine LAG-3 need to be anchored to a cell membrane to reach a functional structure (22). Whether this is also true for human LAG-3 still needs to be determined; nevertheless, our data clearly show a physiological crucial function of natural LAG-3 molecule in determining a fully activation of DC.

Data have been reported showing that LAG-3 triggering on CD4+ and CD8+ T cells results in T cell down-regulation (26, 27, 28, 29) and a role of LAG-3 in controlling T cell homeostasis has been clearly demonstrated in vivo taking advantage of LAG-3 knockout mice (30). This additional function of cell surface-linked LAG-3 may play a role as a crucial, natural feedback mechanism controlling the expansion of a specific response. Conversely, the active membrane shedding of LAG-3 produces a strong HLA class II binder that induces HLA-class II internalization. Down-modulation of HLA class II may prevent further T cell expansion by limiting the availability of a correct environment of stimulation. These additional negative regulating roles of LAG-3 may shield the whole system from an overstimulation. It is likely that all these positive and negative signals are occurring simultaneously in vivo and that the resulting effect on the induced immune response is a function of how they are integrated in a common platform by cells composing the immune system.

In conclusion, we show in this study that, through its synergism with an other important molecule such as CD40L and by modulating the availability of cytokines, LAG-3 can influence the quality of a T cell-induced response. Furthermore, our data prove that the role of LAG-3 expressed by activated T cells is particularly important in those setting lacking sustained TLR-mediated signaling and when a CD40/CD40L interaction could be of suboptimal intensity, as in the case of an antitumor response (7, 8, 9).

Altogether, our data provide physiological rationale for the reported role of LAG-3, when used as recombinant protein, in promoting the immune response so far demonstrated in a variety of settings and further encourage its potential usage as a vaccine adjuvant, especially in cancer.

We thank Francesca Rini and Valeria Beretta for expert technical help and we gratefully acknowledge Grazia Barp for her skilful help in editing this manuscript.

The authors have no financial conflict of interest.

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 work was supported by grants from the Associazione Italiana per la Ricerca sul Cancro (Milano) and the European Community (Cancer Immunotherapy, Contract 518234). C. Casati is a fellow of Fondazione Italiana per Ricerca sul Cancro (Milano).

4

Abbreviations used in this paper: DC, dendritic cell; LAG-3, lymphocyte activation gene-3; MoDC, monocyte-derived DC; hLAG-3Ig, human recombinant LAG-3 protein; iMoDC, immature monocyte-derived DC; CBA, cytometric bead array.

1
Macagno, A., G. Napolitani, A. Lanzavecchia, F. Sallusto.
2007
. Duration, combination and timing: the signal integration model of dendritic cell activation.
Trends Immunol.
28
:
227
-233.
2
Schulz, O., A. D. Edwards, M. Schito, J. Aliberti, S. Manickasingham, A. Sher, C. Reis e Sousa.
2000
. CD40 triggering of heterodimeric IL-12p70 production by dendritic cells in vivo requires a microbial priming signal.
Immunity
13
:
453
-462.
3
Warger, T., P. Osterloh, G. Rechtsteiner, M. Fassbender, V. Heib, B. Schmid, E. Schmitt, H. Schild, M. P. Radsak.
2006
. Synergistic activation of dendritic cells by combined Toll-like receptor ligation induces superior CTL responses in vivo.
Blood
108
:
544
-550.
4
Toes, R. E., S. P. Schoenberger, E. I. van der Voort, R. Offringa, C. J. Melief.
1998
. CD40-CD40 ligand interactions and their role in cytotoxic T lymphocyte priming and anti-tumor immunity.
Semin. Immunol.
10
:
443
-448.
5
Staveley-O’Carroll, K., T. D. Schell, M. Jimenez, L. M. Mylin, M. J. Tevethia, S. P. Schoenberger, S. S. Tevethia.
2003
. In vivo ligation of CD40 enhances priming against the endogenous tumor antigen and promotes CD8+ T cell effector function in SV40 T antigen transgenic mice.
J. Immunol.
171
:
697
-707.
6
Luft, T., E. Maraskovsky, M. Schnurr, K. Knebel, M. Kirsch, M. Görner, R. Skoda, A. D. Ho, P. Nawroth, A. Bierhaus.
2004
. Tuning the volume of the immune response: strength and persistence of stimulation determine migration and cytokine secretion of dendritic cells.
Blood
104
:
1066
-1074.
7
French, L. E., B. Huard, M. Wysocka, R. Shane, E. Contassot, J. F. Arrighi, V. Piguet, S. Calderara, A. H. Rook.
2005
. Impaired CD40L signaling is a cause of defective IL-12 and TNF-α production in Sezary syndrome: circumvention by hexameric soluble CD40L.
Blood
105
:
219
-225.
8
Walker, S. R., R. E. Redlinger, Jr, E. M. Barksdale, Jr.
2005
. Neuroblastoma-induced inhibition of dendritic cell IL-12 production via abrogation of CD40 expression.
J. Pediatr. Surg.
40
:
244
-249.
9
Murugaiyan, G., R. Agrawal, G. C. Mishra, D. Mitra, B. Saha.
2007
. Differential CD40/CD40L expression results in counteracting antitumor immune responses.
J. Immunol.
178
:
2047
-2055.
10
Avice, M. N., M. Sarfati, F. Triébel, G. Delespesse, C. E. Demeure.
1999
. Lymphocyte activation gene-3, a MHC class II ligand expressed on activated T cells, stimulates TNF-α and IL-12 production by monocytes and dendritic cells.
J. Immunol.
162
:
2748
-2753.
11
Andreae, S., S. Buisson, F. Triébel.
2003
. MHC class II signal transduction in human dendritic cells induced by a natural ligand, the LAG-3 protein (CD223).
Blood
102
:
2130
-2137.
12
El Mir, S., F. Triébel.
2000
. A soluble lymphocyte activation gene-3 molecule used as a vaccine adjuvant elicits greater humoral and cellular immune responses to both particulate and soluble antigens.
J. Immunol.
164
:
5583
-5589.
13
Cappello, P., F. Triébel, M. Iezzi, C. Caorsi, E. Quaglino, P. L. Lollini, A. Amici, E. Di Carlo, P. Musiani, M. Giovarelli, G. Forni.
2003
. LAG-3 enables DNA vaccination to persistently prevent mammary carcinogenesis in HER-2/neu transgenic BALB/c mice.
Cancer Res.
63
:
2518
-2525.
14
Casati, C., C. Camisaschi, F. Rini, F. Arienti, L. Rivoltini, F. Triébel, G. Parmiani, C. Castelli.
2006
. Soluble human LAG-3 molecule amplifies the in vitro generation of type 1 tumor-specific immunity.
Cancer Res.
66
:
4450
-4460.
15
Brignone, C., C. Grygar, M. Marcu, K. Schakel, F. Triébel.
2007
. A soluble form of LAG-3 (CD223) induces activation of a large range of human effector cytotoxic cells.
J. Immunol.
179
:
4202
-4211.
16
Triébel, F..
2003
. LAG-3: a regulator of T-cell and DC responses and its use in therapeutic vaccination.
Trends Immunol.
24
:
619
-622.
17
Novellino, L., N. Renkvist, F. Rini, A. Mazzocchi, L. Rivoltini, A. Greco, P. Deho, P. Squarcina, P. F. Robbins, G. Parmiani, C. Castelli.
2003
. Identification of a mutated receptor-like protein tyrosine phosphatase κ as a novel, class II HLA-restricted melanoma antigen.
J. Immunol.
170
:
6363
-6370.
18
Macagno, A., M. Molteni, A. Rinaldi, F. Bertoni, A. Lanzavecchia, C. Rossetti, F. Sallusto.
2006
. A cyanobacterial LPS antagonist prevents endotoxin shock and blocks sustained TLR4 stimulation required for cytokine expression.
J. Exp. Med.
203
:
1481
-1492.
19
Sozzani, S., P. Allavena, A. Vecchi, A. Mantovani.
2000
. Chemokines and dendritic cell traffic.
J. Clin. Immunol.
20
:
151
-160.
20
Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, C. J. Melief.
1998
. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions.
Nature
393
:
480
-483.
21
O’Sullivan, B., R. Thomas.
2003
. Recent advances on the role of CD40 and dendritic cells in immunity and tolerance.
Curr. Opin. Hematol.
10
:
272
-278.
22
Li, N., Y. Wang, K. Forbes, K. M. Vignali, B. S. Heale, P. Saftig, D. Hartmann, R. A. Black, J. J. Rossi, C. P. Blobel, et al
2007
. Metalloproteases regulate T-cell proliferation and effector function via LAG-3.
EMBO J.
26
:
494
-504.
23
Annunziato, F., R. Manetti, I. Tomasevic, M. G. Guidizi, R. Biagiotti, V. Gianno, P. Germano, C. Mavilia, E. Maggi, S. Romagnani.
1996
. Expression and release of LAG-3-encoded protein by human CD4+ T cells are associated with IFN-γ production.
FASEB J.
10
:
769
-776.
24
Newman, J. L., A. Hodzic, D. G. Wingett.
2006
. Differential regulation of soluble and membrane CD40L proteins in T cells.
Cell. Immunol.
241
:
47
-58.
25
Matthies, K. M., D. Graf, S. Muller, U. Korthauer, C. van Kooten, C. Weise, R. A. Kroczek.
1995
. A soluble form of TRAP (CD40 ligand) is rapidly released after T cell activation.
Eur. J. Immunol.
25
:
1749
-1754.
26
Huard, B., P. Prigent, F. Pages, D. Bruniquel, F. Triébel.
1996
. T cell MHC class II molecules downregulate CD4+ T cell clone response following LAG-3 binding.
Eur. J. Immunol.
26
:
1180
-1186.
27
Hannier, S., M. Tournier, G. Bismuth, F. Triébel.
1998
. CD3/TCR complex-associated LAG-3 molecules inhibit CD3/TCR signaling.
J. Immunol.
161
:
4058
-4065.
28
Macon-Lemaitre, L., F. Triébel.
2005
. The negative regulatory function of the lymphocyte-activation gene-3 co-receptor (CD223) on human T cells.
Immunology
115
:
170
-178.
29
Gandhi, M. K., E. Lambley, J. Duraiswamy, U. Dua, C. Smith, S. Elliott, D. Gill, P. Marlton, J. Seymour, R. Khanna.
2006
. Expression of LAG-3 by tumor-infiltrating lymphocytes is co-incident with the suppression of latent membrane antigen-specific CD8+ T-cell function in Hodgkin lymphoma patients.
Blood
108
:
2280
-2289.
30
Workman, C. J., D. A. Vignali.
2005
. Negative regulation of T cell homeostasis by lymphocyte activation gene-3 (CD223).
J. Immunol.
174
:
688
-695.