If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Research Institute for Diseases of the Chest, Graduate School of Medical Sciences, Kyushu University, Fukuoka, JapanDepartment of Infectious Disease, Respiratory, and Digestive Medicine, Graduate School of Medicine, University of Ryukyus, Okinawa, Japan
Research Institute for Diseases of the Chest, Graduate School of Medical Sciences, Kyushu University, Fukuoka, JapanJapan Community Health Care Organization, Kyushu Hospital, Fukuoka, Japan
Corresponding author. Address for correspondence: Isamu Okamoto, Research Institute for Diseases of the Chest, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan.
The interaction of programmed cell death ligand 2 (PD-L2) with programmed cell death 1 is implicated in tumor immune escape. The regulation of PD-L2 expression in tumor cells has remained unclear, however. We here examined intrinsic and extrinsic regulation of PD-L2 expression in NSCLC.
Methods
PD-L2 expression was evaluated by reverse transcription and real-time polymerase chain reaction analysis and by flow cytometry.
Results
BEAS-2B cells stably expressing an activated mutant form of EGFR or the echinoderm microtubule associated protein like 4 (EML4)–ALK receptor tyrosine kinase fusion oncoprotein manifested increased expression of PD-L2 at both the mRNA and protein levels. Furthermore, treatment of NSCLC cell lines that harbor such driver oncogenes with corresponding EGFR or ALK tyrosine kinase inhibitors or depletion of EGFR or ALK by small interfering RNA transfection suppressed expression of PD-L2, demonstrating that activating EGFR mutations or echinoderm microtubule associated protein like 4 gene (EML4)–ALK receptor tyrosine kinase gene (ALK) fusion intrinsically induce PD-L2 expression. We also found that interferon gamma (IFN-γ) extrinsically induced expression of PD-L2 through signal transducer and activator of transcription 1 signaling in NSCLC cells. Oncogene-driven expression of PD-L2 in NSCLC cells was inhibited by knockdown of the transcription factors signal transducer and activator of transcription 3 (STAT3) or c-FOS. IFN-γ also activated STAT3 and c-FOS, suggesting that these proteins may also contribute to the extrinsic induction of PD-L2 expression.
Conclusions
Expression of PD-L2 is induced intrinsically by activating EGFR mutations or EML4-ALK fusion and extrinsically by IFN-γ, with STAT3 and c-FOS possibly contributing to both intrinsic and extrinsic pathways. Our results thus provide insight into the complexity of tumor immune escape in NSCLC.
Immune checkpoint blockade has shown promising clinical activity in the treatment of several types of cancer. Treatment with antibodies that target programmed cell death 1 (PD-1) (also known as CD279) or its ligand programmed cell death ligand 1 (PD-L1) (also known as B7-H1 or CD274) has thus demonstrated durable efficacy for various malignant tumors.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
Programmed cell death ligand 2 (PD-L2) (also known as B7-DC or CD273) is another ligand of PD-1, and the interaction of PD-L2 with PD-1 also inhibits T-cell activation.
PD-L1 is expressed in various immune and nonimmune cell types, including tumor cells, whereas PD-L2 expression was initially thought to be restricted to antigen-presenting cells such as dendritic cells and macrophages.
However, PD-L2 has recently been shown to be expressed by several malignant tumor cell types and therefore to have a potential role in tumor immune escape.
Indeed, some patients whose tumors do not express PD-L1 respond to treatment with antibodies to PD-1, whereas some PD-L1–positive patients do not respond to treatment with antibodies to PD-1 or to PD-L1,
suggesting that the interaction of PD-L2 with PD-1 might contribute to tumor immune escape in some cases. However, little was known of the regulation of PD-L2 expression in tumor cells. We and others have previously shown that PD-L1 expression is intrinsically induced by activating mutations of the EGFR gene or by the echinoderm microtubule associated protein like 4 (EML4)–ALK receptor tyrosine kinase (ALK) fusion gene in NSCLC cells.
Upregulation of PD-L1 by EGFR activation mediates the immune escape in EGFR-driven NSCLC: implication for optional immune targeted therapy for NSCLC patients with EGFR mutation.
We have now investigated the role of activating EGFR mutations and the EML4-ALK fusion gene in regulation of PD-L2 expression in NSCLC cells. In addition, we examined extrinsic regulation of PD-L2 expression by interferon gamma (IFN-γ) in these cells.
Materials and Methods
Cell Culture and Reagents
PC-9, 11_18, and H3122 cells were obtained as previously described.
HCC827, H1975, H1650, H2228, H1299, A549, H23, H2122, H1437, and BEAS-2B cells were obtained from the American Type Culture Collection (Manassas, VA). All cells were maintained under a humidified atmosphere of 5% CO2 at 37°C in Roswell Park Memorial Institute 1640 medium or Dulbecco’s modified Eagle’s medium, each supplemented with 10% fetal bovine serum. Erlotinib (Selleckchem, Houston, TX), alectinib (Selleckchem), and osimertinib (AstraZeneca, Wilmington, DE) were each dissolved in DMSO (Wako, Osaka, Japan). Recombinant human IFN-γ (Peprotech, Rocky Hill, NJ) was dissolved in phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin. All reagents were stored at –20°C or –80°C.
Flow Cytometric Analysis
Cells were stained with either biotinylated mouse monoclonal antibodies to human PD-L1 (eBioscience, San Diego, CA) or an immunoglobulin G1 κ isotype control (eBioscience), followed by phycoerythrin (PE)–labeled streptavidin (BD Biosciences, San Jose, CA), or with PE-labeled mouse monoclonal antibodies to human PD-L2 (BioLegend, San Diego, CA) or a PE-labeled immunoglobulin G2a κ isotype control (BioLegend), for flow cytometric analysis with a fluorescence-activated cell sorting Verse instrument (BD Biosciences). The relative median fluorescence intensity (MFI) ratio was calculated as PD-L1 or PD-L2 MFI divided by isotype control MFI.
RNA Extraction, RT, and Real-Time PCR Analysis
Total RNA was extracted from cells with the use of an RNeasy Mini Kit (Qiagen) and subjected to reverse transcription (RT) with the use of PrimeScript RT Master Mix (Takara, Kusatsu, Japan). Real-time polymerase chain reaction (PCR) analysis was performed in triplicate with the use of SYBR Premix Ex Taq (Takara) and a Thermal Cycler Dice Real Time System (Takara). The PCR primers (forward and reverse, respectively) were as follows: PD-L1 (5'-CAATGTGACCAGCACACTGAGAA-3' and 5'-GGCATAATAAGATGGCTCCCAGAA-3'), PD-L2 (5'-AAAGACCTGTCACCACAACAAAG-3' and 5'-AAAGTGCTGGGTCATCCAAAG-3'), signal transducer and activator of transcription 3 (STAT3) (5'-GGTCTGGCTGGACAATATCATTG-3' and 5'-ATGATGTACCCTTCGTTCCAAAG-3'), c-FOS (5'-AGAATCCGAAGGGAAAGGAA-3' and 5'-CTTCTCCTTCAGCAGGTTGG-3'), signal transducer and activator of transcription 1 (STAT1) (5'-ATCACATTCACATGGGTGG-3' and 5'-CTTCAGGGGATTCTCAGGAATA-3'), and 18S rRNA (5'-ACTCAACACGGGAAACCTCA-3' and 5'-AACCAGACAAATCGCTCCAC-3'). The abundance of each mRNA was normalized by that of 18S rRNA.
Immunoblot Analysis
Cells were rinsed with ice-cold PBS before lysis with sodium dodecyl sulfate (SDS) sample buffer (2% SDS, 10% glycerol, 50 mM Tris-HCl [pH 6.8], and protease and phosphatase inhibitor cocktails), after which the lysates were incubated at 95°C for 5 minutes. Or, nuclear and cytoplasmic extracts of cells were prepared with the use of NE-PER nuclear and cytoplasmic extraction reagents (Pierce Biotechnology, Waltham, MA). Protein was quantitated with the use of a DC Protein Assay Kit (Bio-Rad, Hercules, Ca), portions (30–50 μg) of the lysates or extracts were fractionated by SDS-polyacrylamide gel electrophoresis on a 10% gel, the separated proteins were transferred to a polyvinylidene difluoride membrane, and the membrane was then incubated overnight at 4°C with rabbit primary antibodies. Primary antibodies included those to phosphorylated EGFR (Y1068), EGFR, phosphorylated ALK receptor tyrosine kinase (ALK) (Y1064), ALK, phosphorylated STAT1 (Y701), STAT1, phosphorylated STAT3 (Y705), phosphorylated STAT3 (S727), STAT3, c-FOS, Lamin B1, and β-actin (all from Cell Signaling Technology, Danvers, MA). The membrane was subsequently incubated for 1 hour at room temperature with horseradish peroxidase–conjugated goat antibodies to rabbit immunoglobulin G (Abcam, Cambridge, MA), after which immune complexes were detected with the use of Pierce Western Blotting Substrate Plus (Thermo Fisher Scientific, Waltham, MA) and a ChemiDoc XRS+ system (Bio-Rad).
RNA Interference
Cells were plated at 60% to 70% confluence in six-well plates and then incubated for 24 hours before transient transfection for 48 hours with small interfering RNAs (siRNAs) mixed with the RNAiMAX reagent (Invitrogen, Carlsbad, CA). The siRNAs specific for EGFR mRNA (EGFR-1, 5′-GCAAAGUGUGUAACGGAAUAGGUAU-3′; EGFR-2, 5′-GCAGUCUUAUCUAACUAUGAUGCAA-3′), ALK mRNA (ALK-1, 5′-ACACCCAAAUUAAUACCAA-3′; ALK-2, 5′-UCAG CAAAUUCAACCACCA-3′), STAT3 mRNA (STAT3-1, 5′-UCAUUGACCUUGUGAAAAA-3′; STAT3-2, 5′-GCAAAAAGUUUCCUACAAA-3′), c-FOS mRNA (c-FOS-1, 5′-CUGUCAACGCGCAGGACUU-3′; c-FOS-2, sc-29221), or STAT1 mRNA (STAT1-1, 5′-CCUACGAACAUGACCCUAU-3′; STAT1-2, 5′-GCGUAAUCUUCAGGAUAAU-3′), as well as a control nontarget siRNA (5′-UUCUCCGAACGUGUCACG-3′), were obtained from JBioS and Nippon EGT (Tokyo, Japan), with the exception of c-FOS-2 (sc-29221 from Santa Cruz Biotechnology). Data are presented for one of the two siRNAs corresponding to each target (siRNA-1 in each case) in Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, with those for each second siRNA (siRNA-2) being shown in the Supplementary Figures 1 to 6 as indicated.
Figure 1Upregulation of programmed cell death ligand 2 (PD-L2) and programmed cell death ligand 1 (PD-L1) expression by activating EGFR mutation or echinoderm microtubule associated protein like 4 gene (EML4)–ALK receptor tyrosine kinase gene (ALK) in transfected BEAS-2B cells. BEAS-2B cells stably expressing an exon 19 deletion (Ex19del)-mutant form of EGFR (A) or the echinoderm microtubule associated protein like 4 gene (EML4)–ALK receptor tyrosine kinase (ALK) fusion protein (B), or those stably infected with the corresponding empty virus, were incubated for 48 hours in the presence of erlotinib (100 nM) (A), alectinib (100 nM) (B), or DMSO vehicle. The cells were then lysed and subjected to immunoblot analysis with antibodies to phosphorylated (p) or total forms of EGFR or ALK or with those to β-actin (loading control), as indicated (left panels). The bands detected by the antibodies to phosphorylated or total ALK correspond to the EML4-ALK fusion protein. The cells were also subjected to reverse transcription and real-time polymerase chain reaction analysis of relative PD-L1 or PD-L2 mRNA abundance (middle panels); data are means plus or minus SD of triplicates from one experiment and are representative of three independent experiments. In addition, the cells were subjected to flow cytometric analysis of PD-L1 and PD-L2 expression at the cell surface (right panels). Immunoblot and flow cytometric data are representative of three independent experiments. *p < 0.05, ***p < 0.001 (Student’s t test).
Figure 2Upregulation of programmed cell death ligand 2 (PD-L2) and programmed cell death ligand 1 (PD-L1) expression by activating EGFR mutations or echinoderm microtubule associated protein like 4 gene (EML4)–ALK receptor tyrosine kinase gene (ALK) in NSCLC cell lines. (A, B, and D) PC-9, H1975, and H2228 cells, respectively, were incubated in the presence of DMSO vehicle or either 100 nM erlotinib for 24 hours (A), 100 nM erlotinib, or 100 nM osimertinib for 48 hours (B), or 100 nM alectinib for 24 hours (D). The cells were then lysed and subjected to immunoblot analysis with antibodies to phosphorylated (p) or total forms of EGFR or ALK receptor tyrosine kinase (ALK) or with those to β-actin (loading control), as indicated (left panels). The bands detected by the antibodies to phosphorylated or total ALK correspond to the echinoderm microtubule associated protein like 4 gene (EML4)–ALK receptor tyrosine kinase (ALK) fusion protein. The cells were also subjected to reverse transcription and real-time polymerase chain reaction analysis of relative PD-L1 or PD-L2 mRNA abundance (middle panels); data are means plus or minus SD of triplicates from one experiment and are representative of three independent experiments. In addition, the cells were subjected to flow cytometric analysis of PD-L1 and PD-L2 expression at the cell surface (right panels). (C and E) PC-9 and H2228 cells, respectively, were transfected with nontargeting (NT) or EGFR (C) or ALK (E) small interfering RNAs (siRNAs) for 48 hours and were then subjected to immunoblot, RT and real-time PCR and flow cytometric analyses as already mentioned. All immunoblot and flow cytometric data are representative of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 (Student’s t test); NS, not significant.
Figure 3Upregulation of programmed cell death ligand 2 (PD-L2) and programmed cell death ligand 1 (PD-L1) expression by interferon gamma (IFN-γ) in NSCLC cell lines. A549, PC-9, H1975, or H2228 cells were incubated in the absence or presence of IFN-γ (100 ng/mL) for 24 hours, after which surface expression of PD-L1 and PD-L2 was determined by flow cytometry. Representative profiles and the relative median fluorescence intensity (MFI) ratios (PD-L1–to-isotype or PD-L2–to-isotype ratios) as means plus or minus SD from three independent experiments are shown. *p < 0.05, **p < 0.01, ***p < 0.001 (Student’s t test). Abbreviation: ALK receptor tyrosine kinase gene.
Figure 4Transcriptional control of programmed cell death ligand 2 (PD-L2) and programmed cell death ligand 1 (PD-L1) expression by signal transducer and activator of transcription 3 (STAT3) and c-FOS in NSCLC cells. (A–D) PC-9 cells (A and B) and H1975 cells (C and D) were transfected with nontargeting (NT), STAT3 (A and C), or c-FOS (B and D) small interfering RNAs (siRNAs) for 48 hours, after which cell surface expression of PD-L1 (left panels) and PD-L2 (right panels) was measured by flow cytometry. Representative profiles are shown. (E) The relative PD-L1–to-isotype and PD-L2–to-isotype median fluorescence intensity (MFI) ratios were determined as means plus or minus SD from three independent experiments. *p < 0.05 (Student’s t test).
Figure 5Regulation of programmed cell death ligand 2 (PD-L2) and programmed cell death ligand 1 (PD-L1) expression by IFN-γ signaling molecules. (A) PC-9, H1975, and H2228 cells were transfected with nontargeting (NT) or signal transducer and activator of transcription 1 (STAT1) small interfering RNAs (siRNAs) for 48 hours, for the final 24 hours of which the cells were also exposed to interferon gamma (IFN-γ) (100 ng/mL) or phosphate-buffered saline (PBS) vehicle. The cells were then assayed for surface PD-L1 (left panels) and PD-L2 (right panels) expression by flow cytometry. (B) PC-9, H1975, and H2228 cells were incubated with IFN-γ (100 ng/mL) or PBS vehicle for 24 hours, after which cytoplasmic and nuclear fractions were prepared from cell lysates and subjected to immunoblot analysis with antibodies to phosphorylated (p) or total forms of EGFR, ALK receptor tyrosine kinase (ALK), STAT1, signal transducer and activator of transcription 3 (STAT3), or c-FOS as well as with those to β-actin (cytoplasmic loading control) or to Lamin B1 (nuclear loading control). All results are representative of three independent experiments.
Figure 6Model for intrinsic and extrinsic regulation of programmed cell death ligand 2 (PD-L2) and programmed cell death ligand 1 (PD-L1) expression in oncogene-driven NSCLC cells. Activating EGFR mutations or echinoderm microtubule associated protein like 4 gene (EML4)–ALK receptor tyrosine kinase gene (ALK) fusion intrinsically induce PD-L2 as well as PD-L1 expression through the transcription factors signal transducer and activator of transcription 3 (STAT3) and c-FOS. Interferon gamma (IFN-γ) signaling extrinsically induces the expression of both PD-1 ligands through STAT3 and c-FOS in addition to signal transducer and activator of transcription 1 (STAT1). The regulation of PD-L2 and PD-L1 expression by intrinsic and extrinsic pathways thus shares common transcription factors in NSCLC cells.
Plasmid Constructs for EGFR or EML4-ALK Expression
The plasmid pBabe-EGFR-Del1, which encodes human EGFR with an exon 19 deletion (Ex19del) (E746–A750), was kindly provided by M. Meyerson (Dana-Farber Cancer Institute, Boston, MA; Addgene plasmid 32062, Addgene, Cambridge, MA). An expression vector for echinoderm microtubule associated protein like 4 (EML4)-ALK (variant 3) was established as previously described.
The coding sequences for both the EGFR Ex19del and EML4-ALK proteins were amplified by PCR with PrimeSTAR GXL DNA Polymerase (Takara), and the PCR products were verified by sequencing and then ligated into the pQCXIP retroviral vector (Clontech, Kusatsu, Japan) between the NotI and EcoRI sites with the use of an In-Fusion HD cloning kit (Clontech).
Stable Cell Lines
The pQCXIP vectors encoding the EGFR Ex19del mutant or EML4-ALK were introduced into Amphopack-293 cells (Clontech) by transfection for 48 hours with the Xfect reagent (Clontech). The culture supernatants were then passed through a 0.45-μm filter, incubated overnight at 4°C with a Retro-X concentrator (Clontech), and centrifuged at 1500 g for 45 minutes at 4°C for isolation of retrovirus pellets. BEAS-2B cells were infected with the retroviruses in the presence of polybrene (8 μg/mL) (Nacalai Tesque, Kyoto, Japan) for 24 hours and were then cultured in complete growth medium for an additional 24 hours. The infected cells were then selected by culture in the presence of puromycin (1 μg/mL) (Invivogen, San Diego, CA).
Promoter Constructs
Human genomic DNA was isolated from BEAS-2B cells with the use of a DNeasy Tissue Kit (Qiagen, Hilden, Germany). The candidate promoter regions of the CD274 (–1019 to +110 bp relative to the transcription start site) and PDCD1LG2 (–982 to +99 bp) genes were amplified from the genomic DNA by PCR, and the PCR products were ligated into the pGL4.1 luciferase vector (Promega, Madison, WI) between the KpnI and XhoI sites with the use of an In-Fusion HD cloning kit (Clontech). Mutations were introduced into putative STAT3 or c-FOS binding sites within the PD-L1 and PD-L2 gene promoter regions by site-directed mutagenesis with the use of a KOD Plus Mutagenesis Kit (Toyobo, Osaka, Japan) and the following primers (forward and reverse, respectively PD-L1 STAT3 (5′-GGGGAAGAAAACTGGACTGACATGTTTCA-3′ and 5′-ATGAGATTTTCACCGGGAAGAGTTTC-3′), PD-L1 c-FOS (5′-ATAACAAGGGAAGGAAAGGCAAACAACGAAGAGTCC-3′ and 5′-TCAACTGCAGTTCAAAATACTGCAT-3′), PD-L2 STAT3 (5′-GGGGTGGCACAGCACTAAGACATGCTGGT-3′ and 5′-ATTGACTCATTTCCTAGGGCTTCTGT-3′), and PD-L2 c-FOS (5′-TAACGAGGATTTCCTGGCACAGCACTAAGACATG-3′ and 5′-TTCCTAGGGCTTCTGTAACACATGA-3′). Each promoter region was verified by direct sequencing.
Luciferase Reporter Assay
Cells cultured in 24-well plates were transfected for 24 hours with 2.5 ng of the pGL4.73 Renilla luciferase vector (Promega) and 200 ng of PD-L1 or PD-L2 gene promoter vectors with the use of the Lipofectamine 3000 reagent (Invitrogen). Cell extracts were then assayed for firefly and Renilla luciferase activities with the use of a Dual-Glo Luciferase Assay System (Promega). Firefly luciferase activity was normalized by that of Renilla luciferase.
Statistical Analysis
Data are presented as means plus or minus SD and were analyzed with the unpaired Student’s t test as performed with GraphPad Prism 7 software (GraphPad Software, La Jolla, CA). A p value less than 0.05 was considered statistically significant.
Results
PD-L2 Expression Is Induced by Activating EGFR Mutation or EML4-ALK in BEAS-2B Cells
To investigate the effects of driver oncogenes on the expression of PD-L2 and PD-L1, we established BEAS-2B human bronchial epithelial cells that stably express either human EGFR with an activating (Ex19del) mutation or the EML4-ALK fusion protein. Immunoblot analysis confirmed the expression of total and phosphorylated EGFR (see Fig. 1A) or ALK (see Fig. 1B) in the respective stably transfected cells. Consistent with previous results,
Upregulation of PD-L1 by EGFR activation mediates the immune escape in EGFR-driven NSCLC: implication for optional immune targeted therapy for NSCLC patients with EGFR mutation.
the expression of PD-L1 at both the mRNA and protein levels was increased by expression of the EGFR Ex19del mutant or EML4-ALK in a manner sensitive to inhibition by treatment with EGFR (erlotinib) or ALK (alectinib) tyrosine kinase inhibitors (TKIs). Moreover, we found that the abundance of both PD-L2 mRNA and protein in BEAS-2B cells was also increased by expression of the EGFR Ex19del mutant or EML4-ALK, and that these effects were inhibited by treatment of the cells with erlotinib (see Fig. 1A) or alectinib (see Fig. 1B), respectively. These results thus indicated that the expression of PD-L2 is up-regulated by activated mutant forms of EGFR and by the EML4-ALK fusion protein.
PD-L2 Expression Is Intrinsically Regulated by Activating EGFR Mutations or EML4-ALK in NSCLC Cell Lines
To investigate the regulation of PD-L2 expression in NSCLC cell lines, we measured PD-L2 mRNA and protein levels in five cell lines (HCC827, H1975, PC-9, 11_18, and H1650) that harbor activating EGFR mutations, two cell lines (H3122 and H2228) that harbor the EML4-ALK fusion gene, and five cell lines (A549, H2122, H1299, H23, and H1437) that are wild type (WT) for both EGFR and ALK. HCC827, PC-9, and H1650 harbor an activating in-frame deletion, del(E746–A750), in exon 19 of EGFR; 11-18 harbors an activating point mutation (L858R) in exon 21; and H1975 harbors the L858R point mutation in exon 21 as well as a secondary mutation (T790M) in exon 20. RT and real-time PCR analysis detected PD-L2 mRNA at only a low level in cell lines WT for EGFR and ALK (see Supplementary Fig. 1A). In contrast, most cell lines positive for the driver oncogenes, including HCC827, H1975, PC-9, H1650, and H2228, manifested a high level of PD-L2 mRNA (see Supplementary Fig. 1A). Flow cytometric analysis revealed that the five cell lines WT for EGFR and ALK did not express PD-L2 at the cell surface, whereas PC-9, H1975, and H2228 cells manifested substantial surface expression of PD-L2 (see Supplementary Fig. 1B). We also examined the Cancer Cell Line Encyclopedia database (http://www.broadinstitute.org/ccle) to investigate PD-L2 gene expression in additional NSCLC cell lines. Expression of the PD-L2 gene in cell lines positive for activating EGFR mutations or EML4-ALK, including two additional EGFR-mutated lines, was consistent with our data (see Supplementary Fig. 1C). Analysis of the Cancer Cell Line Encyclopedia database also revealed that some cell lines positive for KRAS mutations or WT for known driver oncogenes manifest a high level of PD-L2 gene expression.
We next examined whether PD-L2 expression is dependent on activated EGFR signaling in PC-9 cells, which harbor an activating EGFR mutation and also express PD-L1 and PD-L2 at high levels. Treatment of these cells with the EGFR TKI erlotinib inhibited EGFR phosphorylation and resulted in downregulation of PD-L2 expression and PD-L1 expression at both the mRNA and protein levels (see Fig. 2A). We also examined the participation of activated EGFR signaling in regulation of PD-L2 expression in H1975 cells, which harbor both an activating EGFR mutation and a secondary mutation (T790M) in exon 20 of EGFR that contributes to EGFR TKI resistance. Treatment with osimertinib, a third-generation EGFR TKI that inhibits the activation of EGFR harboring T790M, attenuated the phosphorylation of EGFR and reduced the amounts of PD-L2 mRNA and protein as well as those of PD-L1 mRNA and protein in these cells, whereas erlotinib had no such effects (see Fig. 2B). To exclude the possibility of off-target effects of erlotinib in PC-9 cells, we instead silenced EGFR expression by transfection with an siRNA specific for EGFR mRNA. Both the abundance of PD-L2 and PD-L1 mRNAs and surface expression of both proteins were down-regulated by transfection with the EGFR siRNA compared with those in cells transfected with a control siRNA (see Fig. 2C). We obtained similar results with a second siRNA targeting a different sequence within EGFR mRNA (see Supplementary Fig. 2A). These findings thus indicated that PD-L2 expression is up-regulated at both the mRNA and protein levels in cells with activating EGFR mutations.
We next tested whether PD-L2 expression is dependent on EML4-ALK signaling in H2228 cells, which harbor the EML4-ALK fusion gene and also highly express PD-L1 and PD-L2. Treatment of these cells with the ALK TKI alectinib inhibited EML4-ALK phosphorylation and induced downregulation of PD-L2 expression as well as that of PD-L1 expression at both the mRNA and protein levels (see Fig. 2D). To exclude potential nonspecific effects of alectinib, we also silenced EML4-ALK expression in H2228 cells by transfection with an siRNA specific for ALK mRNA. Depletion of EML4-ALK was associated with downregulation of both PD-L2 and PD-L1 mRNA and surface protein levels in H2228 cells (see Fig. 2E). Similar results were obtained with a second siRNA targeting a different sequence within ALK mRNA (see Supplementary Fig. 2B). These results thus indicated that PD-L2 expression is also up-regulated as a result of constitutive activation of ALK signaling.
PD-L2 Expression Is Extrinsically Regulated by IFN-γ in NSCLC Cell Lines
Our data revealed intrinsic induction of both PD-L2 and PD-L1 expression by activating EGFR mutations or EML4-ALK fusion in NSCLC cell lines. Although extrinsic induction by IFN-γ, a cytokine that plays a key role in inflammation,
is also an important mechanism for regulation of PD-L1 expression in tumor cells, little was known of such extrinsic regulation of PD-L2 expression. We therefore examined the effect of IFN-γ on PD-L2 expression in NSCLC cell lines. Consistent with previous results,
the surface expression of PD-L1 was up-regulated by IFN-γ stimulation in PC-9, H1975, and H2228 cells, all of which harbor EGFR or ALK driver oncogenes, as well as in A549 cells, which are WT for these genes (see Fig. 3). The surface expression of PD-L2 was also up-regulated by IFN-γ treatment in these four cell lines. These results thus showed that expression of PD-L2, like that of PD-L1, is extrinsically induced by IFN-γ stimulation in NSCLC cells.
Transcriptional Regulation of PD-L2 and PD-L1 Expression by Intrinsic and Extrinsic Pathways
To identify transcription factors that regulate transcription of the PD-L2 and PD-L1 genes in oncogene-driven NSCLC cell lines, we first searched for potential binding sites in the promoter regions of these genes, including 1000 bp upstream from the transcription start site, with the use of the JASPAR database (http://jaspar.genereg.net). The promoter regions of both human PD-L1 and PD-L2 genes were found to harbor putative binding sites for STAT3 (signal transducer and activator of transcription 3) and c-FOS, both of which are downstream transcription factors of EGFR and ALK signaling pathways (see Supplementary Fig. 3A). To investigate whether STAT3 or c-FOS regulates PD-L2 or PD-L1 gene transcription through these binding sites in oncogene-driven NSCLC cell lines, we performed luciferase reporter assays with WT and mutant promoter constructs. Mutation of the putative binding sites for STAT3 or c-FOS resulted in significant attenuation of PD-L2 and PD-L1 gene promoter activity in H1975 cells (see Supplementary Fig. 3B). To investigate further whether driver oncogenes induce PD-L2 expression through STAT3 or c-FOS, we examined the effects of EGFR TKI or ALK TKI treatment on the activation of STAT3 or c-FOS in PC-9 and H2228 cells. Treatment of these cells with the corresponding TKI inhibited phosphorylation of STAT3 and the nuclear abundance of c-FOS in both cell lines (see Supplementary Fig. 4A). In BEAS-2B cells stably expressing the EML4-ALK fusion protein, forced expression of EML4-ALK increased the level of STAT3 phosphorylation and the nuclear abundance of c-FOS (see Supplementary Fig. 4B). We further silenced STAT3 or c-FOS expression in these cells by transfection with specific siRNAs, finding that the upregulation of PD-L2 expression by activated EML4-ALK signaling was attenuated by knockdown of STAT3 or c-FOS (see Supplementary Fig. 4C). To confirm the contribution of STAT3 and c-FOS to regulation of PD-L2 and PD-L1 expression in oncogene-driven NSCLC cells, we investigated the effects of STAT3 or c-FOS knockdown on PD-L2 and PD-L1 expression in PC-9 and H1975 cells. In PC-9 cells, knockdown of STAT3 resulted in downregulation of the surface expression of PD-L2 and PD-L1 (see Fig. 4A and E), whereas depletion of c-FOS reduced PD-L2 surface expression without affecting that of PD-L1 (see Fig. 4B and E). On the other hand, depletion of STAT3 or c-FOS attenuated the expression of both PD-L2 and PD-L1 at the surface of H1975 cells (see Fig. 4C–E). We obtained similar results with additional siRNAs targeting different sequences within STAT3 or c-FOS mRNAs (see Supplementary Fig. 5). Together, these findings thus showed that expression of both PD-L2 and PD-L1 is regulated through STAT3 and c-FOS in oncogene-driven NSCLC cells, although the pattern of regulation appears to differ depending on the cell line.
IFN-γ has been shown to induce PD-L1 expression through STAT1, a key transcription factor in IFN-γ signaling.
Plasma cells from multiple myeloma patients express B7-H1 (PD-L1) and increase expression after stimulation with IFN-{gamma} and TLR ligands via a MyD88-, TRAF6-, and MEK-dependent pathway.
To investigate the regulation of PD-L2 expression by downstream signaling of IFN-γ, we silenced STAT1 expression by transfection with an siRNA specific for STAT1 mRNA in PC-9, H1975, and H2228 cells. The induction of PD-L1 expression by IFN-γ was inhibited by depletion of STAT1 in all three cell lines, whereas that of PD-L2 expression was also attenuated by knockdown of STAT1 but to a lesser extent in PC-9 and H1975 cells than was that of PD-L1 (see Fig. 5A). Similar results were obtained with a second siRNA targeting a different sequence within STAT1 mRNA (see Supplementary Fig. 6). These observations thus indicated that IFN-γ–STAT1 signaling induces PD-L2 as well as PD-L1 expression in NSCLC cells.
Finally, to investigate whether IFN-γ activates additional transcription factors in oncogene-driven NSCLC cells, we examined the effects of IFN-γ on the phosphorylation of STAT3 and the nuclear translocation of c-FOS in PC-9, H1975, and H2228 cells (see Fig. 5B). Consistent with previous results,
Plasma cells from multiple myeloma patients express B7-H1 (PD-L1) and increase expression after stimulation with IFN-{gamma} and TLR ligands via a MyD88-, TRAF6-, and MEK-dependent pathway.
IFN-γ markedly increased the amount of phosphorylated STAT1 in all three cell lines. It also increased the phosphorylation of STAT3 and the nuclear abundance of c-FOS, but not the phosphorylation of EGFR or ALK, implicating both STAT3 and c-FOS, but not EGFR or ALK, as participants in IFN-γ signaling in oncogene-driven NSCLC cells.
Discussion
We and others have previously shown that PD-L1 expression is induced by activating EGFR mutations or EML4-ALK fusion in NSCLC cell lines.
Upregulation of PD-L1 by EGFR activation mediates the immune escape in EGFR-driven NSCLC: implication for optional immune targeted therapy for NSCLC patients with EGFR mutation.
However, little was known of the regulation of PD-L2 expression by driver oncogenes. We have now shown that BEAS-2B cells stably expressing an activated mutant form of EGFR or the EML4-ALK fusion protein manifest increased amounts of PD-L2 mRNA and protein. Furthermore, either inhibition of activated EGFR or ALK signaling by corresponding TKIs or transfection with EGFR or ALK siRNAs suppressed the expression of PD-L2 in oncogene-driven NSCLC cells. These data thus indicate that activating EGFR mutations or EML4-ALK fusion induce intrinsic upregulation of PD-L2 expression and that of PD-L1 expression in NSCLC cells. We further showed that knockdown of STAT3 or c-FOS inhibited PD-L2 expression in oncogene-driven NSCLC cells, thus implicating these transcription factors in the intrinsic regulation of PD-L2 expression. STAT3 is a downstream transcription factor of EGFR and EML4-ALK signaling
We similarly found that STAT3 knockdown resulted in downregulation of PD-L1 expression in oncogene-driven NSCLC cells. The human PD-L1 and PD-L2 genes share 37% sequence identity as well as a similar overall structural organization of their promoter regions.
which is consistent with our findings implicating STAT3 in the regulation of both PD-L2 and PD-L1 expression. c-FOS is activated in response to EGFR signaling and regulates the expression of various genes by forming heterodimers with other transcription factors.
Although the regulation of PD-1 ligand expression by c-FOS has not previously been described as far as we are aware, we identified potential binding sites for c-FOS in the promoter regions of both PD-L2 and PD-L1 genes. We further found that depletion of c-FOS suppressed both PD-L2 and PD-L1 expression in NSCLC cells, thus implicating c-FOS in regulation of the expression of both PD-1 ligands.
We showed that driver oncogenes intrinsically induce PD-L2 and PD-L1 expression through STAT3, c-FOS, or both transcription factors in NSCLC cells. IFN-γ produced by tumor-infiltrating lymphocytes (TILs) has previously been shown to mediate the extrinsic upregulation of PD-L1 expression in tumor cells.
We have now shown that IFN-γ–STAT1 signaling also up-regulates the expression of PD-L2 in NSCLC cell lines. The expression of PD-L1 on the surface of tumor cells in the tumor microenvironment is thought to be induced largely in response to stimulation by IFN-γ released from TILs.
PD-L1 Negative status is associated with lower mutation burden, differential expression of immune-related genes, and worse survival in stage III melanoma.
Our findings that IFN-γ up-regulated the expression of both PD-L2 and PD-L1 in NSCLC cells suggest that expression of both PD-1 ligands is induced in NSCLC by the presence of a high number of TILs in tumor tissue. Our results are consistent with those of a recent study showing that PD-L2 expression generally correlates with PD-L1 expression in human tumor samples.
We have now shown that IFN-γ increased the phosphorylation of STAT3 as well as that of STAT1 in NSCLC cells. Furthermore, we found that c-FOS was also activated in response to IFN-γ stimulation in these cells. These data thus suggest that STAT3 and c-FOS might contribute to the extrinsic induction of PD-L1 and PD-L2 expression by IFN-γ as well as to the intrinsic induction of these PD-1 ligands in oncogene-driven NSCLC cells. Indeed, STAT3 was recently shown to mediate induction of PD-L2 expression by IFN beta or IFN-γ through binding to the PD-L2 gene promoter in melanoma cells.
We also found that knockdown of STAT3 or c-FOS with specific siRNAs inhibited the IFN-γ–induced expression of both PD-L2 and PD-L1 in PC-9 and H1975 cells (data not shown), suggesting that these transcription factors indeed mediate the extrinsic induction of PD-1 ligands. Together, our data thus indicate that STAT3 and c-FOS contribute to both intrinsic and extrinsic regulation of PD-L2 and PD-L1 expression in NSCLC cells (see Fig. 6).
In conclusion, we have shown that PD-L2 expression is induced intrinsically by activating EGFR mutations or EML4-ALK fusion, as well as extrinsically by IFN-γ–STAT1 signaling in NSCLC cells. Furthermore, STAT3 and c-FOS appear to be common transcription factors in both the intrinsic and extrinsic induction of both PD-L1 and PD-L2 expression (see Fig. 6). Our findings thus provide a better understanding of the regulation of PD-L2 expression in tumor cells and of the complex nature of tumor immunity in oncogene-driven NSCLC.
Expression of PD-L2 at mRNA and surface protein levels in NSCLC cell lines. (A) RT and real-time PCR analysis of PD-L2 mRNA abundance (normalized by that of 18S rRNA) in NSCLC cell lines positive or negative for activating EGFR mutations or the EML4-ALK fusion gene. Data are means of triplicates from one experiment and are representative of three independent experiments. (B) Flow cytometric analysis of PD-L2 expression at the cell surface for representative NSCLC cell lines. Data are representative of three independent experiments. (C) PD-L2 gene expression in NSCLC cell lines presented as reads per kilobase of exon per million mapped reads (RPKM). Data are from the Cancer Cell Line Encyclopedia database.
Efficiency of EGFR and ALK knockdown in NSCLC cells. PC-9 (A) and H2228 (B) cells were transfected with nontargeting (NT) or EGFR-1 or EGFR-2 (A) or ALK-1 or ALK-2 (B) siRNAs for 48 hours, lysed, and subjected to immunoblot analysis with antibodies to phosphorylated (p) or total forms of EGFR or ALK or with those to β-actin (loading control), as indicated (left panels). In addition, the cells were subjected to flow cytometric analysis of PD-L2 expression at the cell surface (right panels). Data are representative of three independent experiments.
Regulation of PD-L1 and PD-L2 gene promoter activity by STAT3 and c-FOS. (A) Putative STAT3 and c-FOS binding sites (red letters) identified in the promoter regions of the PD-L1 (–1019 to +110 bp relative to the transcription start site [TSS]) and PD-L2 (–982 to +99 bp) genes by analysis of the JASPAR database. The nucleotides of these sites that were mutated for promoter activity assays are underlined. (B) Luciferase reporter assays performed in H1975 cells for the activity of WT forms of the PD-L1 and PD-L2 gene promoters as well as of mutant (mut) forms of the promoters in which the putative STAT3 or c-FOS binding sites were altered. Data are means ± SD from three independent experiments. **P < 0.01, ***P < 0.001 (Student’s t test).
Regulation of PD-L2 expression by driver oncogenes through STAT3 and c-FOS in NSCLC cell lines or BEAS-2B cells stably expressing the EML4-ALK fusion protein. (A) PC-9 and H2228 cells, respectively, were incubated in the presence of DMSO vehicle or either 100 nM erlotinib for 24 hours, or 100 nM alectinib for 24 hours. Cytoplasmic and nuclear fractions prepared from PC-9 and H2228 cells (A) or BEAS-2B cells stably expressing the EML4-ALK fusion protein (B) were subjected to immunoblot analysis with antibodies to phosphorylated (p) or total forms of EGFR, ALK, STAT3, with those to c-FOS, or with those to β-actin (cytoplasmic loading control) or to Lamin B1 (nuclear loading control). (C) The cells were also transfected with nontargeting (NT), STAT3 (STAT3-1), or c-FOS (c-FOS-1) siRNAs for 48 hours, after which cell surface expression of PD-L2 was measured by flow cytometry. Data are representative of three independent experiments.
Efficiency of STAT3 and c-FOS knockdown in NSCLC cells. (A) PC-9 cells were transfected with nontargeting (NT), STAT3-1, STAT3-2, c-FOS-1, or c-FOS-2 siRNAs for 48 hours, after which the corresponding relative abundance of STAT3 or c-FOS mRNAs was determined by RT and real-time PCR analysis (data are means ± SD of triplicates from one experiment). The cells were also lysed and subjected to immunoblot analysis with antibodies to phosphorylated (p) STAT3 (Y705), total STAT3 or with those to β-actin (loading control). Alternatively, a nuclear fraction prepared from the cells was subjected to immunoblot analysis with antibodies to c-FOS or to Lamin B1 (loading control). (B) PC-9 cells transfected with NT, STAT3-2, or c-FOS-2 siRNAs for 48 hours were subjected to flow cytometric analysis of PD-L2 expression at the cell surface. Data are representative of three independent experiments.
Efficiency of STAT1 knockdown in NSCLC cells. (A) PC-9 cells were transfected with nontargeting (NT), STAT1-1, or STAT1-2 siRNAs for 48 hours, for the final 24 hours of which the cells were also exposed to IFN-γ (100 ng/mL). The relative abundance of STAT1 mRNA was then determined by RT and real-time PCR analysis (data are means ± SD of triplicates from one experiment). The cells were also lysed and subjected to immunoblot analysis with antibodies to phosphorylated (p) or total forms of STAT1 or with those to β-actin (loading control). (B) PC-9 cells treated as in (A) were subjected to flow cytometric analysis of PD-L2 expression at the cell surface. Data are representative of three independent experiments.
References
Herbst R.S.
Soria J.C.
Kowanetz M.
et al.
Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial.
Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): a phase 3, open-label, multicentre randomised controlled trial.
Upregulation of PD-L1 by EGFR activation mediates the immune escape in EGFR-driven NSCLC: implication for optional immune targeted therapy for NSCLC patients with EGFR mutation.
Plasma cells from multiple myeloma patients express B7-H1 (PD-L1) and increase expression after stimulation with IFN-{gamma} and TLR ligands via a MyD88-, TRAF6-, and MEK-dependent pathway.
PD-L1 Negative status is associated with lower mutation burden, differential expression of immune-related genes, and worse survival in stage III melanoma.
Disclosure: Dr. Okamoto reports grants and personal fees from AstraZeneca, Taiho Pharmaceutical, Boehringer Ingelheim, Ono Pharmaceutical, MSD Oncology, Lilly, Bristol-Myers Squibb, and Chugai Pharma; grants from Astellas Pharma and Novartis; and personal fees from Pfizer outside the submitted work. Dr. Azuma reports personal fees from AstraZeneca and grants and personal fees from Boehringer Ingelheim, Ono Pharmaceutical, MSD Oncology, Bristol-Myers Squibb, and Chugai Pharma outside the submitted work. The remaining authors declare no conflict of interest.
30215-6&id=gr1.jpg){kind=link}
30215-6&id=gr2.jpg){kind=link}
30215-6&id=gr3.jpg){kind=link}
30215-6&id=gr4.jpg){kind=link}
30215-6&id=gr5.jpg){kind=link}
30215-6&id=gr6.jpg){kind=link}