Relationship between regulatory and type 1 T cells in dogs with oral malignant melanoma

ABSTRACT

Recent data suggest a decreased prevalence of IFN-γ-producing T lymphocytes (Type 1 T cells) in tumor-bearing hosts. Moreover, it has been reported that Treg have a strong impact on the activation and proliferation of CD4 (+) and CD8 (+) lymphocytes; however, no previous reports have described the relationship between Treg and the progression of tumor, or Type 1 T cell populations in dogs with malignant tumor. In this study, the percentage of Treg, Th1, and Tc1 in the peripheral blood of dogs with oral malignant melanoma and healthy dogs was measured and compared. Although the percentages of Th1 and Tc1 in dogs with oral malignant melanoma were less than those in healthy dogs (Th1: P < 0.01, Tc1: P < 0.05), the percentage of Treg was increased (P < 0.01). A significant inverse correlation between the percentage of Tc1 and the clinical tumor stage (P < 0.01), and a significant correlation between that of Treg and the clinical tumor stage (P < 0.001) was found. Moreover, there was a significant inverse correlation between the percentages of Treg and Th1 (P < 0.05) or Tc1 (P < 0.001). In conclusion, the percentage of Treg increases with the tumor stage in the peripheral blood of dogs with oral malignant melanoma. In dogs, Treg appears to suppress Type 1 immunity, which may be responsible for anti-tumor responses.

List of Abbreviations: 

FITC

fluorescein isothiocyanate

FSC

forward scatter

IFN-γ

interferon-gamma

mAb

monoclonal antibody

PBL

peripheral blood lymphocyte(s)

PBMC

peripheral blood mononuclear cell(s)

PE

phycoerythrin

PGE2

prostaglandin E2

PMA

phorbol 12-myristate 13-acetate

SSC

side scatter

Tc1

Type 1 cyototoxic T lymphocyte(s)

TGF-β

transforming growth factor-β

Th1

Type 1 helper T lymphocyte(s)

Treg

regulatory T cell(s)

Introduction

The emergence of a tumor results from disruption of cell growth regulation as well as from a failure of the host to mount a sufficient immunological anti-tumor response. Multiple mechanisms have been postulated as responsible for preventing an efficient immunological anti-tumor response: from the tumor cell side, different escape mechanisms have been described (e.g. downregulation of the major histocompatibility complex molecule, loss of expression of tumor-associated antigens, lack of costimulatory molecules, secretion of inhibitory cytokines) (1–3); while from the immune cell side, decreased proliferative T-cell responses and loss of cytokine production have been described (4, 5). IFN-γ plays a central role in providing an immunocompetent host with a tumor surveillance mechanism (6). IFN-γ is produced mainly by Th1 and Tc1. Recent studies in humans have demonstrated that a low percentage of Th1 and Tc1 may contribute to the escape of tumor cells from immunosurveillance (7–10). In veterinary medical science, we have reported that the percentage of Th1 in PBMC from tumor-bearing dogs is significantly decreased compared with that in healthy dogs, and that of Th1 in dogs with metastatic tumor is significantly decreased compared with that in dogs without metastatic tumor (11).

Treg, which are of CD4+ T cell lineage, are thought to be a functionally unique subset of T lymphocytes which play an important role in maintaining immune homeostasis (12, 13). Recently, intracellular detection of transcription factor Foxp3 has been shown to uniquely identify a highly enriched Treg population in rodents (14–16), and is considered to be the most specific Treg marker. In dogs, Treg has been identified using an antibody specific for canine CD4 and murine Foxp3 (17). Treg can inhibit the immune response mediated by CD4+CD25- helper T cells and CD8+ cytotoxic T cells and it has been reported that Treg plays an important role in preventing allograft rejection, graft-versus-host disease, and autoimmune disease (18, 19). In addition, it has been shown that, in patients and experimental models with cancer, Treg down-regulates the activity of the effecter function against tumors, resulting in T cell dysfunction in cancer-bearing hosts (20–22). These studies combined with other indirect evidence led us to formulate the hypothesis that tumor-bearing hosts with advanced cancers have an increased population of Treg, which might inhibit the T cell-mediated anti-tumor immuno-response. In fact, an increased population of Treg has been reported in patients with various cancers (23–28). Moreover, Chikamatsu et al. have demonstrated that the percentage of Treg inversely correlates with that of Tc1 in the PBMC of patients with squamous cell carcinoma of the head and neck (29). In veterinary medical science, Biller et al. reported that an increased population of Treg was observed in dogs with various cancers (17); However the relationship between Treg and the progression of tumor or Type 1 T cell populations in dogs with malignant tumor has not yet been described.

In the present study, we compared the percentage of Treg, Th1, and Tc1 in the peripheral blood of dogs with oral malignant melanoma and healthy dogs, and evaluated the correlation between Treg and Th1 or Tc1 and the clinical tumor stage in dogs with oral malignant melanoma.

MATERIALS AND METHODS

Research dogs

Peripheral blood was obtained from 10 normal healthy dogs and 15 dogs with pathologically confirmed oral malignant melanoma. The study was carried out in accordance with the guidelines for animal experiments issued by the College of Bioresource Sciences, Nihon University. The dogs had received no anti-cancer drugs, radiotherapy, surgery, or any immunosuppressive treatment before blood was drawn. The characteristics of the dogs are summarized in Table 1. All dogs with oral malignant melanoma were clinically staged according to the WHO staging system: namely stage I (tumors < 2 cm diameter, negative nodes); stage II (tumors 2–4 cm diameter, negative nodes); stage III (tumors > 4 cm diameter, and/or positive nodes); and stage IV (distant metastatic disease). Three dogs had stage I disease, five stage II, four stage III and three stage IV. The mean age of the dogs with oral malignant melanoma was 10 years (range, 8–15), while the healthy dogs also had a mean age of 10 years (range, 7–13).

Table 1.  Characteristics of dogs

	Oral malignant melanoma (n= 15)	Healthy dogs (n= 10)
Mean age (range) yrs	10 (8–15)	10 (7–13)
WHO stage
I	3
II	5
III	4
IV	3

Cell preparation

Blood samples were collected using a heparinized syringe. PBMC were isolated as previously described (30).

Intracellular cytokine analysis

Th1 and Tc1 were measured by flow cytometric analysis using intracellular cytokine staining. Intracellular cytokine staining was performed as we have previously described (30). In brief, PBMC were stimulated with a combination of PMA (Sigma-Aldrich, Milwaukee, WI, USA), and ionomycin (Sigma-Aldrich) for 4 hr at 37°C in the presence of Brefeldin A (Sigma-Aldrich) for the final 2 hr. After incubation, non-adherent cells were collected as PBL, which were then stained with FITC -conjugated monoclonal antibody as cell surface markers: CD4 as a Th marker and CD8 as a Tc marker.

For staining intracellular cytokines, PBL were fixed and permeabilized using Intraprep (Immunotech, Marseille, France). PBL were incubated with the respective PE-labeled anti-bovine IFN-γ mAb (Serotec, Oxford, UK).

Th1 and Tc1 subpopulations were analyzed by flow cytometry (FACS Cant; Becton Dickenson, San Jose, CA, USA). Lymphocytes were gated by the region determined by FSC and SSC, and then CD4⁺ or CD8⁺ as Th or Tc were gated by FITC intensity. Th1 and Tc1 cell subpopulations were determined by PE intensity. The Th1 or Tc1 percentage was the percentage of IFN-γ-positive cells in the total CD4- or CD8-positive cell population, respectively (Fig. 1a, b).

Figure 1. Dot plot pattern of intracytoplasmic IFN-γ or FoxP3 staining. X-axis: FITC intensity of CD4 or CD8. Y-axis: PE intensity of IFN-γ or FoxP3 Flow cytometric analysis of (a) Th1, (b) Tc1 and (c) Treg in PBMC isolated from a dog with oral malignant melanoma. In Th1 and Tc1, PBMC were stimulated with PMA and ionomycin for 4 hr in the presence of Brefeldin A for the final 2 hr, and stained with FITC-conjugated mAb to CD4 or CD8. The cells were then fixed, permeabilized and stained with PE- conjugated mAb to IFN-γ. In Treg, PBMC were stained with FITC-conjugated mAb to CD4. The cells were then fixed, permeabilized and stained with PE-conjugated mAb to FoxP3. Flow cytometry was performed as described in Materials and Methods. The percentage of Th1, Tc1 or Treg is indicated in the right hand quadrants (CD4 or CD8 positive and IFN-γ or FoxP3 positive).

Intracellular FoxP3 analysis

Treg were measured by flow cytometric analysis using intracellular FoxP3 staining. Intracellular FoxP3 staining was performed as we have previously described (31). In brief, PBMC were stained with FITC-conjugated anti-canine CD4 monoclonal antibody. For staining intracellular FoxP3, PBMC were fixed and permeabilized using the FoxP3 Staining Set (eBioscience, San Diego, CA, USA) following the set protocol. PBMC were incubated with the respective PE-labeled mAb to FoxP3. Anti-FoxP3 mAbs used in this study were anti-murine FoxP3 (clone FJK-16s; eBioscience).

The Treg subpopulation was analyzed by flow cytometry (FACS Canto; Becton Dickenson, San Jose, CA, USA). Lymphocytes were gated by the region determined by FSC and SSC, and then CD4-positive cells were gated by FITC intensity. The Treg cell subpopulation was determined by PE intensity. The Treg percentage was the percentage of FoxP3-positive cells in the total CD4-positive cell population (Fig. 1c).

Statistical analysis

Results are presented as mean values with associated standard deviations. Statistical analyses were performed using MS-Excel (Microsoft). The Mann-Whitney U test was performed for statistical analysis of differences between groups. Spearman's correlation coefficient by rank test was used to analyze correlations. P values less than 0.05 were accepted as significant.

RESULTS

The percentage of Th1 and Tc1 cells in the peripheral blood of dogs with malignant melanoma

The percentages of Th1 and Tc1 cells in dogs with malignant melanoma (Th1: 18.7%± 7.74, Tc1: 37.9%± 7.78) were significantly decreased compared with those in healthy dogs (Th1: 31.7%± 8.10, P < 0.01, Tc1: 44.9%± 5.71, P < 0.05, Fig. 2a, b). The percentage of Tc1 cells in the peripheral blood of dogs with malignant melanoma showed a significant inverse correlation with the clinical tumor stage (rs =−0.73, P < 0.01, Fig. 3b); however, the percentage of Th1 cells in dogs with malignant melanoma showed no significant correlation with the clinical tumor stage (Fig. 3a).

Figure 2. Percentage of Th1, Tc1 and Treg in CD4⁺ cells. There was a significantly smaller percentage of (a) Th1 and (b) Tc1 in dogs with oral malignant melanoma than in healthy dogs (mean ± S.D., Th1 (a); P < 0.01, Tc1 (b); P < 0.05). (c) The percentage of Treg was significantly higher in dogs with oral malignant melanoma than in healthy dogs (P < 0.01).

Figure 3. Correlation between percentage of Th1, Tc1, or Treg subpopulations in the peripheral blood of dogs with oral malignant melanoma and the clinical tumor stage. (a) The percentage of the Th1 subpopulation showed no significant correlation with the clinical tumor stage. (b) The percentage of the Tc1 subpopulation showed a significant inverse correlation with the clinical tumor stage (rs =−0.73, P < 0.01). (c) The percentage of the Treg subpopulation correlated with the clinical tumor stage (rs = 0.81, P < 0.001).

Percentage of Treg cells in the peripheral blood of dogs with malignant melanoma

The percentage of Treg cells in dogs with malignant melanoma (8.5%± 2.05) was significantly increased compared with that in healthy dogs (4.1%± 0.98; P < 0.01, Fig. 2c). The percentages were significantly correlated with the clinical tumor stage (rs = 0.81, P < 0.001, Fig. 3c).

Correlations between Th1, Tc1 and Treg subpopulations in dogs with malignant melanoma

In dogs with malignant melanoma, the percentage of Treg was inversely correlated with that of Th1 (rs =−0.56, P < 0.05; Fig. 4a) or Tc1 (rs =−0.77, P < 0.001; Fig. 4b); however, these correlations were not observed in healthy dogs.

Figure 4. Correlation between percentage of Th1, Tc1 and Treg. The percentage of Treg correlated inversely with that of (a) Th1 (rs =−0.56, P < 0.05) and (b) Tc1 (rs =−0.77, P < 0.001) in dogs with oral malignant melanoma.

DISCUSSION

In this study, we showed a decreased population of Th1 and Tc1 and increased population of Treg in the peripheral blood of dogs with oral malignant melanoma in comparison with healthy dogs. Moreover, in dogs with malignant melanoma, we found a significant inverse correlation between the percentage of Tc1 and the clinical tumor stage, and a significant correlation between the percentage of Treg and the clinical tumor stage. Furthermore, we found a significant inverse correlation between the percentage of Th1 or Tc1 and the percentage of Treg in dogs with malignant melanoma.

We found that the percentage of Th1 and Tc1 in dogs with malignant melanoma was significantly decreased compared with that in healthy dogs. A similar result has been reported in human melanoma (8). It may be a trend that Type1 immune responses of human and canine patients with melanoma are strongly suppressed. IFN-γ, a Type 1 cytokine, acts to enhance the recognition of transformed cells by the immune system (6); thus, IFN-γ plays a central role in providing an immunocompetent host with a mechanism for tumor surveillance. Therefore, the suppression of Type 1 immune responses observed in patients with melanoma may enable tumor cells to escape immuno-surveillance and multiply, resulting in tumor progression.

Furthermore, we identified a significant inverse correlation between the percentage of Treg and that of Tc1 (rs =−0.77, P < 0.001). In human patients with squamous cell carcinoma of the head and neck, the percentage of Treg is inversely correlated with that of Tc1 (29). We considered that the trend for Treg in dogs with a tumor could be similar to that in humans. Liyanage et al. assessed the suppressor function of human Treg by coculturing CD8+ cytotoxic T cells or CD4+CD25- helper T cells with CD4+CD25+ cells. In their report, as the numbers of Treg increase, IFN-γ secretion by CD4+CD25- helper T cells and CD8+ cells is suppressed. Moreover, the proliferation of CD4+CD25- cells and CD8+ cytotoxic T cells, as measured by the ³H thymidine incorporation test, is also suppressed (21). The significant inverse correlation between the percentage of Treg and that of Tc1 or Th1 found in our study was similar to that described by Liyanage et al.

In this study, we found the percentage of Treg in dogs with malignant melanoma to be significantly increased compared with that in healthy dogs. In human medical science, similar findings have been reported. Jandus et al. reported that the percentage of Treg in melanoma patients was significantly increased compared with that in healthy donors (32). Beyer and Schultze described in their review that Treg are increased in most human solid tumors and that there seems to be a stage-dependent increase in Treg (33). In this study, a significant correlation was found between Treg and the clinical tumor stage in canine oral malignant melanoma. This made us consider that the proliferation of tumor cells is related to the expansion of Treg. One possible scenario could be that some cytokines (such as TGF-β or PGE2) that are secreted by tumor cells possibly facilitate the local proliferation of FoxP3⁺CD4⁺ T cells, as well as the in situ conversion of CD4+ T cells into Treg (34, 35). In this way, tumor cells could favor their growth through selective expansion of Treg. According to recent studies, canine malignant melanoma expresses TGF-β and Cyclooxygenase-2-PGE2 (36–38). These factors could provoke the expansion of Treg, as observed in this report. Further studies are necessary to dissect these hypotheses in detail.

There is increasing evidence that Treg play a key role in suppressing T cell-mediated immunity in a cancer-bearing host. In several animal models, it has been suggested that the efficacy of therapeutic cancer vaccination could be enhanced by depleting Treg (39), and that the adoptive transfer of Treg impairs tumor-specific immunity, resulting in tumor progression (40). Considering the present study and previous reports, an increased population of Treg observed in an advanced stage tumor may provide immune tolerance to tumor cells by suppressing Type 1 immunity, and may be connected with tumor progression.

Taken collectively, our results suggest that Th1 and Tc1 subsets in dogs with malignant melanoma appear to be regulated by a subset of Treg. Further understanding of the regulation of T cell-mediated anti-tumor immune responses and the development of new strategies which are able to shift the balance toward Type 1 cytokine polarization may be essential for more effective immunotherapy for cancers.

ACKNOWLEDGMENTS

We thank Mr. Ryudo Gouhara, Mr. Sho Kimijima, Ms. Yoko Okumura, Ms. Madoka Nakamura, Ms. Aya Nishiyama and Ms. Yoko Kurihara for their veterinary technical support.