PD-1–PD-1 ligand interaction contributes to immunosuppressive microenvironment of Hodgkin lymphoma

Programmed death-1 (PD-1), a member of the CD28 costimulatory receptor superfamily, inhibits T-cell activity by providing a second signal to T cells in conjunction with signaling through the T-cell receptor.¹  To date, B7-H1 and B7-DC have been identified as ligands for PD-1 (PD-Ls). During chronic viral infection, PD-1 is selectively up-regulated by the exhausted T cells, and blockade of this pathway restores CD8⁺ T-cell function and reduces viral load.²  This signaling system has been recently highlighted in the research of human immunodeficiency virus (HIV) infection.^3-6  In addition, PD-1 is indicated to be involved in the evasion of tumor immunity.^7-10 

Hodgkin lymphoma (HL) is characterized by massive reactive infiltrates surrounding Hodgkin/Reed-Sternberg (H/RS) cells. HL patients are well recognized as having defective cellular immunity; they are susceptible to bacterial, fungal, and viral infections, and in vitro studies show depressed T-cell proliferation and reduced synthesis of Th1 cytokines.¹¹  We report here that PD-1–PD-L signaling system is operative in patients with HL, and tumor-infiltrating T cells around H/RS cells seem to be kept in balance by this inhibitory signaling. Our findings illuminate the mechanism for deficient cellular immunity observed in HL patients, and propose a potentially effective immunologic strategy for the treatment of HL.

The following cell lines were described previously^12,13 : HL cell lines KM-H2, L428, and HDLM-2; anaplastic large cell lymphoma (ALCL) cell line DEL; follicular lymphoma cell line FL-218; diffuse large B-cell lymphoma (DLBCL) cell line KIS-1; Burkitt lymphoma cell lines Daudi, Raji, Middle 91, Tree 92, and Ramos; adult T-cell leukemia/lymphoma cell line HUT 102; and acute T-cell leukemia cell line Jurkat. LCL-OHN is an Epstein-Barr virus (EBV)–transformed lymphoblastoid B-cell line (LCL). Peripheral blood samples were collected from 19 HL patients, 12 B-NHL patients, and 11 healthy volunteers after informed consent was obtained in accordance with the Declaration of Helsinki. This study is approved by the institutional review board of Kyoto University. After removing red blood cells using ACK lysis buffer, leukocytes were subjected to flow cytometry. For immunohistochemistry, tissue specimens were snap-frozen in OCT compound (TissueTek, Tokyo, Japan) and stored at −80°C.

Total RNA was isolated from cells with Trizol (Invitrogen, Carlsbad, CA), and cDNA was synthesized using MultiScribe Reverse Transcriptase (Applied Biosystems, Foster City, CA). PCR assays were performed by the conventional method using Taq polymerase (TaKaRa Biotechnology, Shiga, Japan). For flow cytometry, cells were analyzed on a FACScan (Becton Dickinson, Mansfield, MS). The following antibodies were used: PE-conjugated B7-H1 and B7-DC (eBioscience, San Diego, CA), FITC-PD-1 (BD Pharmingen, San Diego, CA), PC5-CD4 and CD8 (Immunotech, Marseille, France), FITC-CD25 (Becton Dickinson), and PE-Foxp3 (eBioscience). Immunohistochemical staining was performed as described previously.¹⁴ 

Fragments of the human B7-H1 and B7-DC promoter regions and Flag-tagged LMP1 and LMP2A were constructed by PCR amplification or by hybridization of a pair of complementary strands. The fragments of B7-H1 and B7-DC were subcloned into the pGL3-enhancer vector (Promega, Madison, WI) and pcDNA3-FLAG and CMV-FLAG-5a vectors (Sigma-Aldrich, St Louis, MO), respectively.

LMP1, LMP2A, or mock vector was transfected into 293T cells with each reporter vector using FuGene6 Transfection Reagent (Roche, Mannheim, Germany). Forty-eight hours later, cell lysates were measured for luciferase activity using the Dual Luciferase Assay System (Promega).

Single-cell suspensions derived from HL and B-NHL tissue samples were cultured in RPMI 1640 medium supplemented with 10% autoserum in a 200-μL volume at a cell concentration of 2.5 × 10⁶/mL with B7-H1 and B7-DC blocking antibodies (eBioscience) or isotype control. After 48-hour incubation, the supernatants were analyzed for IFN-γ by an enzyme-linked immunosorbent assay (ELISA; Pierce Biotechnology, Rockford, IL). For the detection of SHP-2 phosphorylation, bulk HL cell suspensions were cultured with blocking antibodies, lysed, immunoprecipitated with anti–SHP-2 antibody (BD Pharmingen), and assessed for tyrosine phosphorylation by Western blotting as previously described.¹⁵  To determine which cell type is the main producer of IFN-γ, CD4⁺ or CD8⁺ T cells were depleted or enriched from HL cell suspensions by magnetic-activated cell sorting (MACS) using CD4 or CD8 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany).

We first examined B7-H1 and B7-DC expression in several lymphoid cell lines (Figure 1A,B). Some HL and T-cell lines were positive for B7-H1, and B7-DC showed broader expression. In contrast, most B-cell lines except LCL-OHN lacked their expression.

As PD-L genes were also expressed in another 3 LCLs, we examined the effect of EBV-latent membrane proteins on their promoter activities (Figure 1C). The result showed that both LMP1 and LMP2A enhanced the transcriptional activity of B7-H1 and B7-DC. This finding implies that in cases of EBV⁺ HL, the expression of PD-L genes is possibly led by latent membrane proteins.

In immunohistochemical analysis, H/RS cells were shown to be positive for PD-Ls in both EBV⁺ and EBV⁻ samples (Figure 1D). Thus the expression of PD-Ls was assumed to be a common feature of H/RS cells.

We next analyzed PD-1 expression of tumor-infiltrating T cells of HL tissue samples. PD-1⁺ cells were obviously increased in all 3 samples tested (Figure 2A), although PD-1 expression was previously described as negative in HL tissues.¹⁶  PD-1 expression level was also significantly elevated in the peripheral blood T cells of HL patients, in comparison with those of healthy controls and B-NHL patients (Figure 2B left). Expression levels tended to be higher in patients with active disease and declined along with treatment. In contrast, T cells of CD4⁺CD25⁺ regulatory phenotype were not apparently increased (Figure 2B right), whereas they are reported to play an important role in the pathogenesis of HL.^17-21  We evaluated PD-1 and Foxp3 expression concurrently in 9 other HL patients (Figure 2C), and they were shown to be mostly independent of each other. This result is similar to the result of tumor-infiltrating T cells studied in B-NHL.²² 

We finally examined the effect of blockade of this pathway. After culturing bulk HL tumor cells in the presence of anti–PD-L blocking antibodies, IFN-γ production was measured by ELISA. Blockade of PD-Ls augmented IFN-γ production in 2 HL samples (Figure 2D left). In contrast, among 5 B-NHL tissues tested, we did not detect IFN-γ in 4 samples (1 DLBCL, 2 follicular lymphoma, and 1 marginal zone B-cell lymphoma), irrespective of PD-L blockade. IFN-γ was barely measurable in 1 DLBCL sample, but blockade of PD-Ls did not influence the IFN-γ production (Figure 2D right). Blockade of PD-Ls was accompanied with the inhibition of SHP-2 phosphorylation (Figure 2E), which has been reported as a mediator of the PD-1 signaling pathway.²³  Depletion or enrichment of T-cell subsets from HL cell suspension indicated that PD-L blockade revived mainly the function of CD4⁺ T cells of HL (Figure 2F). Natural killer cell population was less than 0.5%, and they seemed unlikely to be the major source of IFN-γ (data not shown). We concluded that antitumor activity of HL-infiltrating T cells is inhibited via the PD-1–PD-L signaling pathway, and this inhibition can be successfully relieved by PD-L blockade. Our observations indicate that “T-cell exhaustion” is essential to the pathogenesis of HL, which is in line with the recent study of RNA fingerprints.²⁴  Adoptive immunotherapy with tumor antigen–specific T cells has lately been attempted for the treatment of HL.²⁵  Our findings provide a potentially effective and clinically applicable strategy for the immunotherapy of HL.

We are grateful to the following physicians for providing blood samples of the patients: Dr H. Ohno, Takeda General Hospital; Dr C. Ueda, Kyoto Katsura Hospital; Drs K. Yamashita, T. Kitano, and H. Matsubara, Kyoto University; Drs T. Takeoka, M. Tsuji, and W. Kishimoto, Otsu Red-Cross Hospital; Dr Y. Maesako, Tenri Hospital; Drs M. Watanabe, H. Kaneko, and M. Nishizawa, Osaka Red-Cross Hospital. We also thank Dr Y. Kobashi, Tenri Hospital, and Dr H. Haga and Mr Y. Toda, Kyoto University, for histologic examination.

This work was supported in part by grants-in-aid from the Ministry of Education, Culture, Sports, Science and Technology (18790650) of Japan.

Contribution: R.Y., M.N., and T. Kitawaki performed research; M.N. designed research and wrote the paper; T.S. and M.T. contributed vital new reagents; M.H, T. Kondo, K.O., M.K., and T.H. provided clinical samples and analyzed data; T.U. supervised research.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Momoko Nishikori, 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto, 606-8507, Japan; e-mail: nishikor@kuhp.kyoto-u.ac.jp.