Induction of latency-associated peptide (transforming growth factor-β₁) expression on CD4⁺ T cells reduces Toll-like receptor 4 ligand-induced tumour necrosis factor-α production in a transforming growth factor-β-dependent manner

Abstract

CD4⁺ T cells expressing the latent form of transforming growth factor-β [latency-associated peptide (LAP) (TGF-β₁)] play an important role in the modulation of immune responses. Here, we identified a novel peptide ligand (GPC_81–95) with an intrinsic ability to induce membrane-bound LAP (TGF-β₁) expression on a subpopulation of human CD4⁺ T cells (using flow cytometry; ranging from 0·8% to 2·6%) and stimulate peripheral blood mononuclear cells to release LAP (TGF-β₁) (using ELISPOT assay; ranging from 0·03% to 0·16%). In spite of this low percentage of responding cells, GPC_81–95 significantly reduced Toll-like receptor 4 ligand-induced tumour necrosis factor-α production in a TGF-β₁- and CD4⁺ T-cell-dependent manner. The results demonstrate that GPC_81–95 is a useful tool to study the functional properties of a subpopulation of LAP (TGF-β₁)⁺ CD4⁺ T cells and suggest a pathway that can be exploited to suppress inflammatory response.

Introduction

Transforming growth factor-β₁ (TGF-β₁) is involved in the regulation of numerous cellular functions and is produced by most cell types in a latent form. The latent form of TGF-β₁ [LAP (TGF-β₁)] is comprised of latency-associated peptide (LAP) non-covalently bound to mature TGF-β₁. It is known that many immune cells can produce LAP (TGF-β₁) or can express this molecule on their cell surface^1,2 and that LAP (TGF-β₁)-expressing CD4⁺ T cells play an important role in modulation of immune responses.^3–5 It has been shown that oral or nasal administration of anti-CD3 antibodies induces LAP (TGF-β₁)⁺ CD4⁺ T cells and suppresses autoimmune disease in animal models in a TGF-β₁-dependent manner,^3,6 but there is little information on other LAP (TGF-β₁)-inducing ligands or the mechanism involved in the induction of this regulatory molecule on CD4⁺ T cells.

Tumour necrosis factor-α (TNF-α) is a pro-inflammatory cytokine that is produced mainly by monocytes and macrophages after stimulation with endotoxin.⁷ It has many immunostimulatory functions and plays a crucial role in inflammation and immunity. Tumour necrosis factor-α plays a particularly important role in promoting the cascade of pathogenic events in rheumatoid arthritis, Crohn's disease, psoriasis, septic shock as well as other diseases. The majority of anti-inflammatory agents which can inhibit TNF-α, such as cyclosporine⁸ and glucocorticoids,⁹ are also broadly immunosuppressive and are associated with adverse effects. Therefore antibodies to TNF-α have been developed to specifically target the effects of this pro-inflammatory cytokine.¹⁰ Novel anti-inflammatory agents with no or very few adverse effects that specifically inhibit TNF-α production would therefore be desirable to block TNF-α production and could be used in combination with antibodies that block TNF-α function.

We had shown that a peptide derived from alpha-fetoprotein (AFP) stimulates LAP (TGF-β₁) production by CD4⁺ T cells and demonstrated that these TGF-β₁-producing T cells have immunoregulatory properties.¹¹ Glypican-3 and AFP are oncofetal antigens and are over-expressed in hepatocellular carcinoma. Glypican-3, a cell surface-linked heparan sulphate proteoglycan, is highly expressed during embryogenesis and is involved in organogenesis. It is over-expressed by many tumour and non-tumour cells such as melanoma and hepatocellular carcinoma as well as by hepatic progenitor/oval cells.^12–18 It is also a useful diagnostic marker that distinguishes hepatocellular carcinoma from benign hepatocellular mass lesions and is potentially a target for immunotherapy.^12,19 Therefore, it is important to study the functional properties of Glypican-3 and peptides derived from this antigen on immune system cells including CD4⁺ T cells. A monoclonal antibody recognizing membrane-bound LAP (TGF-β₁) is now commercially available and has allowed us to study the effects of peptides derived from Glypican-3 on the expression of LAP (TGF-β₁) on immune system cells. We screened overlapping peptides covering the sequence of Glypican-3 (GPC) to identify peptide ligands with the ability to induce LAP (TGF-β₁) expression on T cells. A 15-amino-acid-long peptide was identified with the capacity to stimulate the expression of LAP (TGF-β₁) on T cells. The findings also demonstrate that GPC_81–95 has anti-inflammatory properties and suppresses Toll-like receptor 4 (TLR4) ligand-induced TNF-α production in a TGF-β₁-dependent manner. This inhibition was abolished by the removal of CD4⁺ T cells, suggesting that GPC_81–95 stimulates the activation of CD4⁺ T cells with anti-inflammatory properties.

Materials and methods

Blood samples

This study was approved by an UCLH ethical committee and all individuals gave written informed consent. Peripheral blood mononuclear cells (PBMCs) were isolated from the heparinized peripheral blood of healthy donors by density grade centrifugation.

Synthetic peptides and cell lines

A total of 58 fifteen-mer overlapping peptides spanning the GPC sequence, along with alanine substituted and truncated forms of the GPC_81–95 peptide were synthesized (Mimotopes, Clayton, Australia). The human leukaemia CD4⁺ T-cell line (Jurkat E6.1; Sigma-Aldrich, Dorset, UK) were purchased and cultured in RPMI-1640 containing 10% fetal calf serum.

Stimulation and inhibition of T-cell responses

The PBMCs were stimulated with GPC-derived peptides or an irrelevant peptide (AFP364–373) at 1–60 μg/ml and incubated for 5 hr at 37° in AIM V containing 10% fetal calf serum. For intracellular cytokine staining, brefeldin A (10 μg/ml; Alomone Labs, Jerusalem, Israel) was added for the last 3 hr. Dead cells were excluded using 7-amino-actinomycin D (7-AAD; Sigma-Aldrich) staining. Human TLR1 to TLR9 ligands (Autogen Bioclear, Calne, UK) were added to cell culture to mimic or modify peptide-induced cytokine production.

ELISPOT assay

The LAP (TGF-β₁)-producing cells were detected upon peptide stimulation after 18 hr using an ex vivo ELISPOT assay (R&D Systems, Abingdon, UK) as described previously.¹¹

Flow cytometry

Cells were surface stained with different fluorochrome-linked antibodies to CD3, CD4, (both BD Pharmingen, Oxford, UK), LAP (TGF-β₁) (clone 27232; R&D Systems) and Foxp3 (eBioscience, Hatfield, UK) or isotype controls (R&D Systems) and assessed by flow cytometry. An immunological responder was defined as a twofold increase in the frequency of cytokine-producing cells above control peptides or proteins. Apoptosis and cell death were assessed using annexin V (BD Pharmingen) and 7-AAD staining.

Functional assay

The PBMCs were cultured with or without peptides, including vasoactive intestinal peptide (VIP; Bachem, St. Helens, UK; 1 μm), for 5 hr in the presence or absence of mouse anti-human TGF-β₁ IgG1 (50 μg/ml), mouse anti-human isotype control IgG1 (50 μg/ml), different concentrations of rTGF-β₁ (R&D Systems) or PBS diluents (negative control). The cells were then stimulated with lipopolysaccharide (LPS; 10 ng/ml) for a further 24 hr. Interleukin-1β (IL-1β), IL-6, regulated on activation, normal T-cell-expressed and secreted (RANTES) and TNF-α concentrations were determined using human FlowCytomix Simplex assays as described by the manufacturer (Bender Medsystem GmbH, Vienna, Austria). CD4 and CD8 T cells were depleted from PBMCs as described by the manufacturer (Dynal, Oslo, Norway).

Results

Identification of a peptide ligand with the ability to stimulate LAP (TGF-β₁) expression on CD4⁺ T cells

We screened overlapping peptides covering GPC to identify a peptide ligand with the ability to stimulate LAP (TGF-β₁) expression. In brief, PBMCs were stimulated with overlapping GPC-derived peptides (58 fifteen-mer peptides in total) and the expression of membrane-bound LAP (TGF-β₁) on CD4⁺ T cells was analysed using flow cytometry. In these experiments, dead cells were excluded from the assays using 7-AAD staining (data not shown). CD4⁺ T cells stimulated with GPC_81–95 (YQLTARLNMEQLLQS), but not the other 57 GPC peptides, expressed membrane-bound LAP (TGF-β1) (Fig. 1a). The results demonstrate that GPC_81–95 peptide, but not an irrelevant peptide (AFP_365–373), stimulates LAP (TGF-β₁) expression on CD4⁺ T cells in a dose-dependent manner (Fig. 1b). LAP (TGF-β₁) could also be released from the cells by GPC_81–95 treatment in a dose-dependent manner as detected by an ex vivo ELISPOT assay (Fig. 1c). As cells other than CD4⁺ T cells could also produce LAP (TGF-β₁), we aimed to determine the relative contribution of CD4⁺ T cells to the production of LAP (TGF-β₁) by PBMCs. CD4⁺ T cells were depleted from PBMCs and the frequency of LAP (TGF-β₁)-producing cells per 1·5 × 10⁵ cells was determined using an ELISPOT assay. The results demonstrate that over 50% of GPC_81–95-induced LAP (TGF-β₁)-producing cells were CD4⁺ T cells (Fig. 1d; 210 responders per 1·5 × 10⁵ total PBMCs versus 99 responders per 1·5 × 10⁵ CD3⁺-depleted PBMCs). Given the important role that CD4⁺ T cells play in modulating an immune response, we focused this study primarily on the effects of GPC_81–95 on CD4⁺ T cells.

Figure 1.

Identification of a peptide ligand with the ability to stimulate expression of the latent form of transforming growth factor-β [latency-associated peptide (LAP) (TGF-β₁)] by primary CD4⁺ T cells. (a) Peripheral blood mononuclear cells (PBMCs) were stimulated for 5 hr with 15-mer GPC-derived peptides and LAP (TGF-β₁) membrane expression was detected using flow cytometry. The percentage of 7AAD⁻ CD4⁺ LAP (TGF-β₁)⁺ T cells is shown. This result is representative of three independent experiments. (b) GPC_81–95 but not an irrelevant peptide (irr. peptide; AFP_365–373) stimulates LAP (TGF-β₁) membrane expression and release in a dose-dependent manner as detected by flow cytometry (c) and ELISPOT. (d) Depletion of CD4⁺ T cells from PBMCs results in a reduction in the frequency of LAP (TGF-β₁)-producing cells. Error bars show standard errors of triplicates. Results are representative of at least three independent experiments.

The percentages of LAP (TGF-β₁)⁺ CD4⁺ T cells in PBMCs of donors 1–4 after stimulation with GPC_81–95 are shown using flow cytometry (Fig. 2a). The release of LAP (TGF-β₁) was also analysed in the PBMCs of donors 5–8 (Fig. 2b). The results demonstrate that all the individuals tested in this experiment responded to GPC_81–95 peptide but not an irrelevant peptide (AFP_365–373) and expressed LAP (TGF-β₁). To clarify whether or not the responsive CD4⁺ LAP (TGF-β₁)⁺ fraction corresponds to the FoxP3⁺ regulatory T-cell population, GPC_81–95-stimulated CD4 T cells were co-stained for intracellular Foxp3 and membrane-bound LAP (TGF-β₁). The results demonstrate that the reacting CD4⁺ T cells do not express Foxp3 (Fig. 2c).

Figure 2.

Peripheral blood mononuclear cells (PBMCs) from different donors respond to GPC_81–95 and express the latent form of transforming growth factor-β [latency-associated peptide (LAP) (TGF-β₁)]. (a) PBMCs from donors 1, 2, 3 and 4 were stimulated for 5 hr with GPC_81–95 or an irrelevant peptide (AFP_365–373) and LAP (TGF-β₁) membrane expression was detected using flow cytometry. The percentage of 7AAD⁻ CD4⁺ LAP (TGF- β₁)⁺ T cells is shown. (b) The frequencies of LAP (TGF-β₁)-producing cells were analysed in donors 5, 6, 7 and 8 using an ELISPOT assay. Error bars show standard errors of triplicates. (c) PBMCs were stimulated with GPC_81–95 and the expression of Foxp3 was analysed on LAP (TGF-β₁)⁺ T cells. Dot plots show gated CD3⁺ CD4⁺ T cells. Results are representative of two independent experiments performed on two individuals.

To examine whether GPC_81–95 can directly stimulate CD4⁺ T cells, we performed two sets of experiments. The ability of GPC_81–95 to stimulate LAP (TGF-β₁) was demonstrated in purified primary CD4⁺ T cells (95% purity as determined by FACS) and Jurkat CD4⁺ T cells (data not shown).

GPC_81–95 has intrinsic ability to induce LAP (TGF-β₁) expression

We used several approaches to confirm that GPC_81–95 has intrinsic ability to induce LAP (TGF-β₁) on CD4⁺ T cells. First, we demonstrated that alanine substitution at positions 81, 82, 83, 84, 85 (alanine to serine), 86, 87, 88, 89, 92, 93 and 94 reduce the ability of GPC_81–95 to stimulate LAP (TGF-β₁) (Fig. 3a). This result suggests that the biological activity of the GPC_81–95 depends on its amino acid composition. Second, we observed that GPC_81–95 peptide with higher purity (> 90%) induced higher percentages of LAP (TGF-β₁) expression than the lower purity peptide (70%) (data not shown), suggesting that non-GPC_81–95 peptide derivatives produced during peptide synthesis (shorter peptides, peptides with amino acid deletions or substitutions) are not the bioactive components. We also found that none of the truncated 10-mer peptides or the reversed form of GPC_81–95 (SQLLQEMNLRATLQY) induced LAP (TGF-β₁) (Fig. 3b,c), indicating that the biological activity of the GPC_81–95 also depends on its length. To confirm that the GPC_81–95-induced LAP (TGF-β₁) expression on CD4⁺ T cells is not the result of contamination with TLR ligands, we tested commercially available TLR1–9 ligands in a broad range of concentrations. None of these treatments had the ability to induce LAP (TGF-β₁) expression (Fig. 3d). To examine whether TLR ligands can improve the ability of GPC_81–95 to induce LAP (TGF-β₁) expression, we performed a series of experiments. Cells were stimulated with different concentrations of GPC_81–95 peptide (1, 5 and 10 μg/ml) and cultured in the presence or absence of TLR1–9 ligands and the expression levels of membrane-bound LAP (TGF-β₁) were analysed using flow cytometry. None of TLR ligands, including LPS, increased the expression of LAP (TGF-β₁) on GPC_81–95 peptide-stimulated T cells (data not shown).

Figure 3.

GPC_81–95 intrinsically induces expression of the membrane-bound latent form of transforming growth factor-β [latency-associated peptide (LAP) (TGF-β₁)]. Peripheral blood mononuclear cells (PBMCs) were stimulated with (a) altered peptide ligands, (b) truncated peptides or (c) reversed form of GPC_81–95 for 5 hr and LAP (TGF-β₁) membrane expression was detected using flow cytometry. The percentage of CD4⁺ LAP (TGF-β₁)⁺ T cells is shown. (d) The percentages of primary CD4⁺ LAP (TGF-β₁)⁺ T cells induced by GPC_81–95, Toll-like receptor (TLR) 1–9 ligands (e) or vasoactive intestinal peptide (VIP; 1 μm), vascular endothelial growht factor (VEGF; 20 ng/ml), PMA/ionomycin, purified protein derivative (PPD; 10 μg/ml) or anti-CD3 antibody are shown. Results are representative of three independent experiments.

Anti-CD3 antibody induces LAP (TGF-β₁) on T cells and suppresses inflammatory condition in a TGF-β₁-dependent manner.³ Moreover, it has been shown that vascular endothelial growth factor (VEGF) induces TGF-β₁.²⁰ To compare the ability of these ligands with GPC_81–95 to induce LAP (TGF-β₁), PBMCs were stimulated with different concentrations of anti-CD3 antibody, VEGF, VIP, PMA/ionomycin, staphylococcal enterotoxin B, purified protein derivative or GPC_81–95 and the percentage of LAP (TGF-β₁)⁺ CD4⁺ T cells was analysed using flow cytometry. GPC_81–95 and anti-CD3 antibody (1 and 5 μg/ml) induced LAP (TGF-β₁) on CD4 T cells, whereas VEGF, VIP, PMA/ionomycin, staphylococcal enterotoxin B and purified protein derivative did not induce LAP (TGF-β₁) expression (Fig. 3e).

GPC_81–95 does not induce apoptosis or cell death

Apoptotic cells are known to produce TGF-β₁.²¹ To determine whether peptide-induced LAP (TGF-β) expression on CD4⁺ T cells is the result of T-cell apoptosis, CD4⁺ Jurkat T cells were treated with different concentrations of GPC_81–95 peptide (5–30 μg/ml). The percentages of cell death and early apoptosis were analysed by 7-AAD and annexin V staining, respectively, 5 and 24 hr after exposure. GPC_81–95 did not induce cell death or apoptosis in Jurkat T cells (Fig. 4a). All the assays were performed in triplicate and the results were confirmed in two independent experiments. Moreover, peptide-induced LAP (TGF-β₁)⁺ CD4⁺ Jurkat T cells were 7-AAD⁻ annexin V⁻, demonstrating that these cells are not dead or dying (Fig. 4b,c).

Figure 4.

GPC_81–95 does not induce cell death and membrane bound latent form of transforming growth factor-β [latency-associated peptide (LAP) (TGF-β₁)] is not expressed by dead cells. (a) CD4⁺ Jurkat T cells were stimulated with or without GPC_81–95 for 5 or 24 hr and annexin V and 7-AAD were detected using flow cytometry. (b) and (c) CD4⁺ Jurkat T cells were stimulated with or without GPC_81–95 for 5 hr and membrane-bound LAP (TGF-β₁), annexin V and 7-AAD were detected using flow cytometry. The percentage of cells within each quadrant is shown. Error bars show standard errors of triplicates. Results are representative of two independent experiments.

The anti-inflammatory properties of GPC_81–95 are mediated by CD4 T cells and is TGF-β₁-dependent

The PBMCs isolated from healthy donors were cultured with GPC_81–95 or an irrelevant peptide (AFP_365–373) and then stimulated with 10 ng/ml LPS. The concentrations of different pro-inflammatory cytokines (TNF-α, IL-1β, IL-6 and RANTES) were measured in the supernatant after 24 hr (Fig. 5a). Treatment of the cells with an irrelevant peptide (AFP_365–373) did not alter the concentration of pro-inflammatory cytokines in comparison with non-treated cells or cells treated with PBS diluents (data not shown), suggesting that the irrelevant peptide has no inhibitory effect. In contrast, GPC_81–95-treatment inhibited the production of TNF-α but not the production of IL-1β, IL-6 or RANTES (Fig. 5a). The average percentages of inhibition from four independent experiments are shown in Fig. 5(b), demonstrating that TNF-α is the only pro-inflammatory cytokine measured that was consistently inhibited by GPC_81–95 treatment (AFP_365–373 was used as an irrelevant peptide). Inhibition of TNF-α by GPC_81–95 treatment was seen over a range of LPS doses (Fig. 5c). To determine the kinetics of GPC_81–95 treatment on the inhibition of TNF-α, PBMCs were treated with GPC_81–95 and supernatants were taken 0, 5, 18, 24, 48 and 72 hr after LPS stimulation. The results demonstrate that the highest percentage of inhibition by GPC_81–95 treatment is observed 24 hr after LPS stimulation (Fig. 5d). The inhibitory effect of GPC_81–95 treatment on the secretion of TNF-α was analysed in 10 independent experiments (performed on different days) and demonstrates that GPC_81–95 treatment significantly suppresses TNF-α production (P = 0·0002). The average inhibition observed in each experiment is shown (Fig. 5f). To compare the inhibitory effects of recombinant TGF-β₁ and GPC_81–95, PBMCs were treated with different concentrations of rTGF-β₁, GPC_81–95, or PBS diluents (as negative control) for 5 hr and the cells were stimulated with LPS. The percentage of TNF-α inhibition by GPC_81–95 treatment was equivalent to the percentage of inhibition seen with a high dose of recombinant TGF-β₁ (Fig. 5e). The inhibitory effects of GPC_81–95 and VIP, which has been shown to possess anti-inflammatory properties in vitro and in vivo,^22–25 were confirmed in our system (Fig. 5g). To study the role of TGF-β₁ in GPC_81–95-mediated inhibition, anti-TGF-β₁ monoclonal antibody (mouse IgG1) was added to the culture and the results demonstrate that this blocking antibody abrogated the inhibition seen with GPC_81–95 treatment. The inhibitory effects of GPC_81–95 treatment were not diminished when a mouse IgG1 isotype control (Fig. 5h), or when anti-LAP (TGF-β₁) monoclonal antibody (mouse IgG1) was added to the culture (data not shown). The results demonstrate that GPC_81–95 suppress TLR4-ligand-induced TNF-α production in a TGF-β₁-dependent manner. The depletion of CD4⁺ T cells from the PBMCs also abolished the inhibitory effects of GPC_81–95 (Fig. 5i), suggesting that the anti-inflammatory effect of GPC_81–95 is mainly mediated by CD4⁺ T cells.

Figure 5.

The anti-inflammatory property of GPC_81–95 is mediated by CD4 T cells and is transforming growth factor-β₁ (TGF-β₁)-dependent. (a) Peripheral blood mononuclear cells (PBMCs) were incubated with GPC_81–95 or irrelevant peptide (irr. peptide; AFP_365–373) for 5 hr. Lipopolysaccharide (LPS; 10 ng/ml) was then added and the PBMCs were incubated for a further 24 hr. The concentration of interleukin-1-β (IL-1β), IL-6, regulated on activation, normal T-cell-expressed and secreted (RANTES) and tumour necrosis factor-α (TNF-α) was measured in the supernatants. (b) The average percentage inhibition of IL-1β, IL-6, RANTES and TNF-α production by GPC_81–95 treatment is shown from four independent experiments. (c) GPC_81–95 inhibits TNF-α production induced by increasing doses of LPS. (d) The time at which supernatants were taken following the addition of LPS was varied as indicated and the percentage inhibition of LPS-induced TNF-α production by GPC_81–95 was calculated. (e) GPC_81–95 is as potent as high concentrations of rTGF-β₁ at inhibiting LPS-induced TNF-α production from PBMCs. (f) The average percentage of inhibition of TNF-α production by GPC_81–95 treatment is shown from 10 independent experiments. (g) Vasoactive intestinal peptide (VIP) and GPC_81–95 inhibited LPS-induced TNF-α production. (h) Anti-TGF-β¹ antibody abrogates the ability of GPC_81–95 to inhibit LPS-induced TNF-α production from PBMCs. (i) Depletion of CD4⁺ from PBMCs abrogates the ability of GPC_81–95 to inhibit LPS induced TNF-α production. Error bars show standard errors of triplicates. Results are representative of at least three independent experiments.

Discussion

In this study, we demonstrate that a 15-mer GPC-derived peptide (GPC_81–95) has the intrinsic ability to stimulate the expression of LAP (TGF-β₁) on CD4⁺ T cells. The bioactivity of GPC_81–95 could not be attributed to potential contaminants such as non-GPC_81–95 peptide derivatives produced during peptide synthesis or TLR ligands. Finally, we show that GPC_81–95 suppresses TLR4 ligand-induced TNF-α secretion, which is dependent on the presence of both TGF-β₁ and CD4⁺ T cells.

Our data show that GPC_81–95 does not induce cell death, which has previously been shown to stimulate TGF-β₁ release and thereby suppresses the production of pro-inflammatory cytokines by monocytes.²¹ GPC_81–95 suppresses TNF-α production but does not inhibit the production of other pro-inflammatory cytokines including IL-1β by PBMCs stimulated with LPS. This is in accordance with the results demonstrating that recombinant TGF-β₁ inhibits LPS-induced TNF-α production but does not alter the levels of IL-1α and IL-1β production.²⁶ This finding may also indicate that these pro-inflammatory cytokines are independently regulated and confirms that the reduction in TNF-α production is not the result of cell death. Anti-inflammatory agents, such as glucocorticoids and VIP, can directly suppress the function of monocytes and macrophages and result in the inhibition of TLR4 ligand-induced TNF-α production.⁹ In contrast, GPC_81–95 inhibits TLR4 ligand-induced TNF-α production by generating CD4⁺ T cells with anti-inflammatory properties.

GPC_81–95 stimulates LAP (TGF-β₁) expression on only a small fraction of primary CD4⁺ T cells (1–2·6%) or Jurkat T cells (3–4%). It is likely that specific receptor(s) are involved in the recognition of the identified peptide and the expression of these receptors may be up-regulated in a small population of primary CD4⁺ T cells. However, this hypothesis may not explain why only a small population of Jurkat T cells responded to the peptide stimulation. It is possible, but not proven, that up-regulation of LAP (TGF-β₁) is confined to the physiological condition of cells such as a stage of cell division. The fact that only small population of CD3⁺ CD4⁺ T cells responded to anti-CD3 antibody and expressed LAP (TGF-β₁) supports this notion. Although, the majority of CD4⁺ T cells express CD3 molecules but only a small population of CD3⁺ CD4⁺ T cells responded to anti-CD3 antibody and expressed LAP (TGF-β₁). Anti-CD3 antibody is the only known ligand that induces LAP (TGF-β₁) expression on CD4⁺ T cells and the administration of this antibody suppresses inflammatory conditions in a TGF-β₁-dependent manner.^3,27 Our data have shown that both GPC_81–95 and anti-CD3 antibody induce LAP (TGF-β₁) on primary CD4 T cells. It has been suggested that GARP (glycoprotein-A repetitions predominant) is essential for surface expression of LAP (TGF-β₁) on activated regulatory T cells.¹ In our hands, no positive cells were detected in resting primary CD4 T cells using the only commercially available anti-GARP antibody (LRRC32 monoclonal antibody; Enzo Life Science, Exeter, UK) (isotype control IgG2b; Enzo Life Science). Using this antibody, no positive cells were detected in GPC_81–95 or anti-CD3 antibody-induced LAP expressing primary CD4 T cells (data not shown). Therefore, we are unable to confirm or exclude the possibility that GARP may be expressed on these cells.

Further studies are planned to demonstrate whether GPC_81–95 can induce LAP (TGF-β₁) expression and inhibit inflammation in an in vivo model. Previously self-derived synthetic peptides that exert immunoregulatory effects via induction of TGF-β₁ and activation of regulatory T cells have been described. These peptides are derived from a conserved region of the MHC class II molecule and are shown to bind to the MHC and alter T-cell receptor (TCR) –MHC interaction, thereby exerting their inhibitory effect via the TCR.²⁸

The 28-amino acid neuropeptide VIP has been shown to have anti-inflammatory properties^24,29 and is a candidate for the treatment of autoimmune disease and septic shock. Our results demonstrate that both GPC_81–95 and VIP can inhibit TLR4 ligand-induced TNF-α. However, no sequence homology was found between GPC_81–95 and VIP, or between GPC_81–95 and other anti-inflammatory neuropeptides (such as calcitonin gene-related peptide, α-melanocyte-stimulating hormone, and adrenocorticotrophic hormone). We have also observed that VIP does not induce LAP (TGF-β₁) and a VIP receptor inhibitor does not block GPC_81–95-induced LAP (TGF-β₁) expression by primary CD4⁺ T cells (S. Boswell and S. Behboudi, unpublished data). In fact, it has been shown that VIP and pituitary adenylate cyclase-activating polypeptide can inhibit TGF-β₁ production,³⁰ suggesting that there is a significant difference in the mode of action between GPC_81–95 peptide and VIP analogues. Similar to VIP, the recognition of GPC_81–95 peptide by CD4⁺ T cells does not require the presence of antigen-presenting cells or accessory cells, suggesting that CD4⁺ T cells recognize the peptide in a TCR-independent manner. This notion is supported by the fact that GPC_81–95 peptide stimulated purified primary CD4⁺ T cells and Jurkat T cells to express LAP (TGF-β₁). To demonstrate that TCR is not involved in the peptide recognition, we examined the ability of GPC_81–95 peptide to stimulate J.CaM1.6 cells (a derivative mutant of Jurkat CD4⁺ T cells with a defect in TCR signal transduction) to express LAP (TGF-β₁) as assessed by flow cytometry (data not shown). The expression of GPC_81–95-induced LAP (TGF-β₁) on both Jurkat CD4⁺ T and J.CaM1.6 CD4⁺ T cells demonstrates that this recognition is not via TCR molecules and professional APCs are not required for this activation.

Taken together, our results demonstrate that a 15-amino-acid-long peptide within glypican-3 sequence that stimulates the expression of LAP (TGF-β₁) on T cells. The finding also demonstrates that peptide-induced LAP (TGF-β₁)⁺ CD4⁺ T cells have immunoregulatory properties and suppress TLR4 ligand-induced TNF-α production in a TGF-β₁-dependent manner.

Disclosures

The authors have no financial conflicts of interest.