________________________________________________________
Non-classical HLA reactive CD8+ T-cells harbor an
innate-like phenotype and can be identified by
Helios expression in NSCLC
________________________________________________________
Bachelorproject Thesis
Version: Final version
Date: 10-7-2020
Student’s name: Yara Yamise Witte
Student’s number: 11639598
Study: Biomedical Sciences (Bachelor)
University of Amsterdam
Supervisor: Pleun Hombrink
Second assessor: Marieke van Ham
Institute: Sanquin Research, Department of Haematopoiesis
Period of internship: 3rd February 2020 - 14 July 2020
EC: 30
Non-classical HLA reactive CD8+ T-cells harbor an innate-like
phenotype and can be identified by Helios expression in NSCLC
ABSTRACT
T-cells that use innate receptors to respond to stimuli despite expressing T-cell antigen receptors are called “innate-like T-cells”. CD1 restricted T-cells, γδ T-cells and Natural Killer T (NKT) cells are examples of innate-like T-cells and function by recognizing fixed ligands such as HLA-E and HLA-G. However, it is not clear to what extend HLA-E and HLA-G peptides can affect T-cells. The expression of unconventional human leukocyte antigens HLA-G and HLA-E was found to be a critical marker of immune tolerance in non-small-cell-lung carcinoma (NSCLC) immune evasion. As such, HLA-G and HLA-E are considered as immune checkpoints. However, only a little is known about the
characteristics of HLA-G and HLA-E reactive T-cells engaging with HLA-E and HLA-G in NSCLC and the functional consequence. Here, we characterized HLA-G and HLA-E reactive T-cells found in blood, healthy lungs and tumors of NSCLC-patients. For this purpose, we developed UV-exchangeable HLA-G and HLA-E tetramers and analyzed reactive cells by flow cytometry. We found that non-classical HLA reactive CD8+ T-cells harbor an innate-like phenotype and can be identified by Helios expression in NSCLC.
INTRODUCTION
The lungs are the main organs of the respiratory system and are vital for oxygen exchange. Due to the constant exposure to airborne pathogens are the lungs under constant treat of infection. T-cells play an important role in the protection of the lungs against airborne pathogens (Woodland & Scott, 2005). Keeping a fine balance between protection and preventing immunopathology is crucial for maintaining the integrity of this delicate barrier tissue. T-cells have also been demonstrated to play an important role in preventing cancer (Ribas, 2015). Lung cancer is the leading cause of cancer death worldwide and is responsible for an estimated of 2,09 million cases and 1,76 million deaths worldwide in 2018 (World Health Organization, 2018). Non-small-cell-lung carcinoma (NSCLC) accounts for approximately 85% of all lung cancer cases (Herbst et al., 2018).
Both the innate and adaptive arms of the immune systems are important in the protection of the lungs against airborne pathogens, but they function in different ways. The innate immune system identifies the presence of pathogens via pattern-recognition receptors (PRRs) (Kumar et al., 2011). Structural components or molecules such as nucleic acids of viruses, bacteria and fungi are microbial ligands that can activate PRRs, leading to the induction of the innate immune response (Medzhitov, 2009). The innate immune response initiates the production of proinflammatory interferons (IFN) and cytokines and induces cell death, autophagy and phagocytosis (Sky et al., 2015). In addition, the innate immune response can activate the adaptive immune response via antigen presenting cells such as macrophages and dendritic cells that are able to present an antigen to a naïve or immature T-cell. (Murphy & Weaver, 2017). T-cells can be divided into CD8+ and CD4+ T-cells that recognize their peptide antigens once presented by Histocompatibility Leukocyte Antigen (HLA) class I and II, respectively, through their variable T-cell receptor (TCR). After binding, T-cells will be activated and differentiated into cytotoxic T lymphocytes (CTLs) that have the ability to kill tumor and virus-infected cells, memory T-cells that provide long lasting protection or different subsets of T-helper cells
(Murphy & Weaver, 2017).
Some cells do not fit within the 2-arms classifications of the immune system, because there are some T-cells with a non-variable TCR. These T-cells are called “innate-like T-cells” and are for example: CD1 restricted T-cells, γδ T-cells, Natural Killer T (NKT) cells and MR1 restricted mucosal associated invariant T-cells (MAIT) which are all also considered as “unconventional” or
“non-classical” non-MHC-restricted T-cells (Godfrey et al., 2015 & Joosten et al., 2016). “Unconventional” is actually not the right name for these T-cells, because innate-like T-cells are very common. Innate-like T-cells have characteristics of the innate immune system, because they have a restricted TCR or NK receptor that can serve as PRR (Martino et al., 2011). In contrast, innate-like T-cells can also be considered as a part of the adaptive immune system, because these T-cells have a TCR, so they belong to the T-cell lineage and they function by recognizing fixed ligands such as HLA-E and HLA-G (Martino et al., 2011).
HLA-E and HLA-G are MHC class I molecules with a restricted sequence variability (Moretta et al., 2010). HLA-E has both an innate and adaptive role, because HLA-E can present self-peptides to NK-cells and pathogen-derived peptides to CD8+ T-cells (Joosten et al., 2016). HLA-E can also serve as ligand for the CD94/NKG2 receptors that are expressed on NK cells and CD8+ T-cells (Pietra et al., 2009). Binding of HLA-E to the CD94/NKG2A or CD94/NKG2B receptor will lead to the inhibition of CD8+ T-cells and NK cells. In contrast, the binding of HLA-E to the CD94/NKG2C or CD94/NKG2D receptors will lead to the activation of CD8+ T-cells and NK cells (Kanevskiy et al., 2019). Moreover, HLA-E can bind to the αβ T-cell receptor (TCR) and γδ TCR on CD8+ T-cells which will lead to activation (Kanevskiy et al., 2019). Thus, the function of HLA-E depends on the receptor to which the HLA-E ligand binds. HLA-G is able to modulate the innate and adaptive immune system, because HLA-G can serve as ligand for the inhibitory KIR2DL4/CD158d receptor expressed on NK cells, inhibitory receptor ILT/CD85/LILRB1 expressed on dendritic cells, T-cells, NK cells, B-cells and monocytes and the
Yan, 2018). However, it is not clear to what extend HLA-E and HLA-G peptides can affect T-cells (Rölle et al., 2018).
While the prime targets of immunotherapy are conventional T-cells, targeting intratumorally innate-like T-cells may be an attractive option since the expression of unconventional human
leukocyte antigens HLA-G and HLA-E was found to be an important marker of immune strength in NSCLC immune evasion and was strongly correlated with disease procedure and prognosis (Lin & Yan, 2018 ; Creelan & Antonia, 2019). As such, HLA-G and HLA-E are considered as immune checkpoints. However, only a little is known about the characteristics of HLA-G and HLA-E reactive T-cells engaging with these unconventional antigens in NSCLC and the functional consequence. Furthermore, NSCLC is commonly insensitive to chemotherapy and the efficacy of immunotherapy with conventional T-cells varies (Rizvi et al., 2015). Thus, insights into possible new treatment options for NSCLC is important. When more becomes clear about the characteristics of HLA-G and HLA-E reactive T-cells in NSCLC, new insights can be provided for immunotherapy strategies for NSCLC-patients. In this study, we characterized HLA-E and HLA-G reactive T-cells found in blood, healthy lung and tumors of NSCLC-patients. For this purpose, we developed UV-exchangeable HLA-G and HLA-E tetramers and analyzed reactive cells using Fluorescence-activated cell sorting (FACS). We found that non-classical HLA reactive CD8+ T-cells harbor an innate-like phenotype and can be identified by Helios expression in NSCLC. In addition, this T-cell characterization has given new insights into possible immunotherapies for NSCLC and COVID-19 that can be used to improve immunotherapy regimens for NSCLC and can be helpful in designing a therapy for COVID-19 in the future.
MATERIALS & METHODS
Subjects
Healthy lung and tumor tissue samples were obtained from 10 NSCLC-patients. As first line therapy, the NSCLC-patients obtained a surgical resection of primary tumors without prior chemotherapy or radiotherapy. The NSCLC-patients were recruited from Onze Lieve VrouweGasthuis (OLVG),
Amsterdam, the Netherlands. Peripheral blood was obtained from anonymized healthy volunteers at Sanquin Blood Supply Foundation, Amsterdam, The Netherlands. Written informed consent was given by all the patients and donors before inclusion into the study. All procedures were approved by the Sanquin Ethical Advisory Board according to the declaration of Helsinki and Dutch regulations.
HLA-E and HLA-G peptide pools
The peptide pool of HLA-G consisted of the peptides: 1. RIIPRHLQL, 2. KIPAQFYIL (reference), 3. RHPKYKTEL, 4. HVPEHAVVL, 5. MQPTHPIRL, 6. KIAGYVTHL, 7. GVPKTHLEL, 9. KGPPAALTL, 9. SYPTRIASL, 10. RLPDGRVVL, 11. KSPASDTYIVF, 12. MRPRKAFLL, 13. KLPKDFVDL, 14. VLPKLYVKL, 15. VVPKDRVAL, 16. RSPVYLTVL, 17. VQVQMKFRL, 18. RIPQGFGNLL, 19. KTPSGIKL. The peptide pool of HLA-E consisted of the peptides: 1. RLPAKAPLL, 2. VMAPRTVLL, 3. VMAPRALLL, 4. VMAPRTLFL, 5. VMAPRTLIL (reference), 6. QMRPVSRVL, 7. VMAPLTLIL, 8. GMKFDRGYI, 9. VMAPRTLLL, 10. IMANRAQVL, 11. MMKYLAFGL, 12. ALPPRFEL, 13. ILSPDAPVL, 14. AISPRTLNA. The details of the peptide pools are listed in Supplementary tables S1 and S2 in the Supplementary Material.
In vitro stimulation of T-cells with HLA-E & HLA-G peptide pools
Cryopreserved Peripheral Blood Mononuclear Cells (PBMCs) from 5 PBMC donors were thawed in a 37°C water bath, leaving a small ice cube. Next, the cells were further thawed with warm “thaw agent” containing IMDM medium (Gibco) with 10% fetal calf serum (FCS) and DNase (diluted 1:1000) and phosphate buffered saline (PBS) with 0,5% FCS (MACS). The “thaw medium” was added dropwise while swirling and next the cells had to settle for 15-20 minutes at 37°C. Thereafter, the cells were washed, supplemented with 10 ml MACS and counted in 10 ml isotone with 20 μl sample using the CASY counter (Innovatis) to determine the viability of the cells. Subsequently, 2x106 cells were added
per well in a 96-wells V-shape PS plate. For the positive control, the cells were incubated with platebound αCD3 (1:100, HIT3A; eBioscience) at 37°C for minimum one hour. Before the cells were incubated with platebound αCD3, the plate had to be coated and incubated with IgG Fc (diluted 1:500 in PBS, Jackson 115-006-071) overnight at 4°C and blocked with IMDM medium with 10% FCS for 30 minutes at 37°C. Thereafter, the cells were incubated with the conditions: αCD28 (diluted 1:1000, s.28; CLB) as a positive control, Cytomegalovirus (CMV) protein (final concentration of 100 ng/ml) as positive control for a viral infection, HLA-E all peptides (final concentration of 100 ng/ml), HLA-E 1,4,5 & 9 peptides (final concentration of 100 ng/ml), HLA-G all peptides (final concentration of 100 ng/ml), DMSO 40%, because the peptide pools were frozen in 40% DMSO and we wanted to investigate whether this had an effect on the cells and IMDM (10% FCS) as negative control at 37°C for 2 hours (see Supplementary tables S1 and S2 for more information about the HLA-E and HLA-G peptides). To determine the cytokine production by T-cells, the cells were incubated in the presence of Brefeldin A (diluted 1:1000, eBioscience) overnight at 37°C.
Processing of in vitro stimulated T-cells for Flow Cytometry
The cells were labelled with a surface staining mix made in MACS (PBS 0,5% FCS) and an intracellular staining mix made in PERM buffer. All the incubations steps were performed on ice and in the dark by covering the plate with aluminium foil. The surface of the cells was stained with a Surface Staining mix containing the following antibodies; 1:1000 diluted NEAR IR-BUV563 (Thermo Fisher Scientific, L10119), 1:400 diluted CD45RA-BUV563 (HI100, BD bioscience 565703) and 1:100 diluted CD27-BV510 (O323, Biolegend 302836) for 30 minutes. Subsequently, the cells were washed with MACS
and fixed and permeabilized using 50 μl fixation solution of the Foxp3/Transcription factor staining kit (eBioscience) for 30 minutes. Next, the cells were washed again with MACS and stained with the Intracellular Staining mix containing the following antibodies; 1:1000 diluted CD3-BUV661 (UCHT1, BD bioscience 565065), 1:200 diluted CD4-BUV737 (SK3, BD bioscience 564305), 1:200 diluted CD8a-BUV805 (SK1, BD bioscience 564912), 1:200 diluted CD137-PE-Cy7 (434, Thermofisher 25-1379-42), 1:400 diluted CD40L-PEDazzle594 (24-31, Biolegend 310840), 1:200 diluted TNF-α-FITC (BD
bioscience 554512) and 1:100 diluted IFN-γ-PE (45.B3, Biolegend 502509) for 30 minutes. Thereafter, Perm buffer (1x) was added for 5 minutes at 4°C and the cells were washed twice with MACS. Finally, the cells were resuspended in MACS and the data was acquired using the FACS Symphony A5 (BD, Temse, Belgium) with the UV laser on (FCS 380 & SCC 240). The analysis of the data was done using Flowjo Version 10 software (Temse, Belgium).
In vitro HLA-G and HLA-E tetramer staining
Cryopreserved Peripheral Blood Mononuclear Cells (PBMCs) from 5 donors were thawed in a 37°C water bath, leaving a small ice cube. Next, the cells were further thawed with warm “thaw agent” containing PBS with 10% FCS and DNase (diluted 1:1000). The “thaw medium” was added dropwise while swirling and next the cells had to settle for 15-20 minutes at 37°C. Thereafter, the cells were washed, supplemented with 10 ml MACS and counted in 10 ml isotone with 20 μl sample using the CASY counter (Innovatis) to determine the viability of the cells. Subsequently, 4x106 cells were added
per well in a 96-wells V-shape PS plate for the HLA-G tetramer staining and 2x106 cells were added
per well for the HLA-E tetramer staining. Next, 142 μl of HLA-G G01:01 tetramer in MACS solution and 142 μl of HLA-G G05:01 tetramer in MACS solution was added to different wells and as a negative control MACS was added. For the HLA-E tetramer staining 154 μl of UV-exchangeable HLA-E 01:03 VMAPRTLIL tetramer in MACS solution and 154 μl of non-exchangeable HLA-E 01:03 VMAPRTLIL tetramer in MACS solution was added to different wells and again MACS was used as a negative control. The tetramer solutions were incubated for 15 minutes on ice in the dark.
Flow Cytometry analysis after HLA-G and HLA-E tetramer staining
The surface of the cells was stained in the dark for 30 minutes on ice with a combination of the following antibodies for the HLA-G tetramer staining: 1:1000 diluted NEAR IR/APC-Cy7 (Invitrogen L-10119), 1:400 diluted CD45RA-BUV563 (HI100, BD bioscience 565703), 1:400 diluted CD27-PECF594 (M-T271, BD 562297), 1:1000 diluted CD3-BUV661 (UCHT1, BD bioscience 565065), 1:200 diluted CD4-BUV737 (SK3, BD bioscience 564305), 1:200 diluted CD8a-BUV805 (SK1, BD bioscience 564912), 1:50 diluted CD94-APC (DX22, BioLegend 305508), 1:50 diluted CD16-BU510 (3G8, BD 563830), 1:200 diluted CD14-PEC-Cy7 (61D3, eBioscience), 1:100 diluted KLRG1-AF488/FITC (eBioscience 13F12F2) and 1:100 diluted HLA-E-PE For the HLA-E tetramer staining a combination of the following antibodies was used: 1:1000 diluted NEAR IR/APC-Cy7 (Invitrogen L-10119), 1:400 diluted CD45RA-BUV563 (HI100, BD bioscience 565703), 1:400 diluted CD27-PECF594 (M-T271, BD 562297), 1:1000 diluted CD3-BUV661 (UCHT1, BD bioscience 565065), 1:200 diluted CD4-BUV737 (SK3, BD bioscience 564305), 1:200 diluted CD8a-BUV805 (SK1, BD bioscience 564912), 1:50 diluted CD94-APC (DX22, BioLegend 305508), 1:50 diluted CD16-BU510 (3G8, BD 563830), 1:100 diluted CD8b-PEC-Cy7 (Thermo Fisher Scientific, 25-5273-42), 1:100 diluted KLRG1-AF488/FITC (eBioscience 13F12F2) and 1:100 diluted TCRγδ-BV421 (B1, BD bioscience 745202). Finally, the cells were resuspended in MACS and the data was acquired using the FACS Symphony A5 (BD, Temse, Belgium) with the UV laser on (FCS 380 & SCC 240). The analysis of this data was done using Flowjo Version 10 software (Temse, Belgium).
In vitro staining assay HELIOS (performed by Pleun Hombrink)
Cryopreserved Peripheral Blood Mononuclear Cells (PBMCs) were thawed in a 37°C water bath, leaving a small ice cube. Subsequently, the cells were further thawed with 50 ml warm “thaw agent” containing 30% FCS in MACS buffer per ampoule. The “thaw medium” was added dropwise while swirling and next the cells had to settle for 15-20 minutes at 37°C. Thereafter, the cells were washed, resuspended in 10 ml MACS and counted in 10 ml isotone with 20 μl sample using the CASY counter (Innovatis) to determine the viability of the cells. Subsequently, the cells were resuspended in the required volume and transferred to a 96-wells V-shape PS plate. The cells were washed again with MACS buffer to collect the remaining cells.
Processing of stained T-cells for Flow Cytometry (HELIOS)
All the following incubations steps were performed on ice and in the dark by covering the plate with aluminium foil. The surface of the cells was stained for 30 minutes with 40 μl per well of a
combination of the following antibodies diluted in MACS buffer: 1:1000 diluted NEAR IR/APC-Cy7 (Invitrogen L-10119), 1:400 diluted CD45RA-BUV563 (HI100, BD bioscience 565703), 1:50 diluted CD69-BUV395 (FN50, BD Bioscience, 564364), 1:50 diluted CD2-BV785 (Biolegend, 300233), 1:200 diluted CD103-BV711 (Biolegend, 350222), 1:50 diluted CD5-BV605 (UCHT2, BD bioscience 563945), 1:100 diluted CCR7-BV510 (G043H7, Biolegend 353232), 1:50 diluted Tigit-PeCy7 (A15153G,
Biolegend 372714), 1:100 diluted CD27-PeCF594 (M-T271, BD 562297), 1:50 diluted CD6-Pe (BL-CD6, Biolegend 313906), 1:50 diluted PD-1-BB700 (BD, 566460), 1:100 diluted MR1-FITC (National
Institutes of Health (NIH)) 1:50 diluted TCRγδ-FITC (B1.1, Thermo Fischer 11-9959-42), 1:100 diluted CD158a-APC (Biolegend 339510), 1:100 diluted CD158b-APC (Biolegend 312612). Next, the cells were washed with MACS buffer, fixated with 50 μl of 1x fixation/perm buffer (1:4) per well for 30 minutes and washed again with MACS buffer. Thereafter, the cells were intracellular stained for 45 minutes with 40 μl per well of a combination of the following antibodies diluted in permeabilization buffer: 1:1000 diluted CD3-BUV661 (UCHT1, BD bioscience 565065), 1:200 diluted CD4-BUV737 (SK3, BD bioscience 564305), 1:400 diluted CD8a-BUV805 (SK1, BD bioscience 564912), 1:400 diluted Granzyme B (GZMB)-AF700 (BD 560213) and 1:25 diluted HELIOS-eF450/BV421 (22F6, Thermo Fischer Scientific 48-9883-42). Next, 150 μl of 1x permeabilization buffer was added to each well and the cells were washed again. Finally, the cells were resuspended in 70 μl MACS buffer and the data was acquired using the FACS Symphony A5 (BD, Temse, Belgium) with the UV laser on (FCS 380 & SCC 240). The analysis of this data was done using Flowjo Version 10 software (Temse, Belgium).
Statistical analysis
Statistical analyses were performed on the data with Graphpad Prism 8.0.1 (244) using the unpaired T test with Welch’s correction. A p-value below 0,05 was considered as statistically significant in all analyses. Significance is indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.
RESULTS
UV-exchangeable HLA-E and -G tetramers can be used to identify HLA-E and HLA-G reactive
T-cells
UV-exchangeable HLA-E and HLA-G tetramers were made to test the binding capacity of different HLA-E and HLA-G peptides to T-cells. The aim was to investigate whether there is a difference in binding capacity between the peptides and whether the binding of T-cells to the tetramer was peptide specific or not. In this first experiment, UV-exchangeable HLA-E01:03 VMAPRTLIL, “control” non-exchangeable HLA-E01:03 VMAPRTLIL, exchangeable HLA-G01:01 KIPAQFYIL and
UV-exchangeable HLAG05:01 KIPAQFYIL tetramers were validated an tested whether they were suitable for future experiments to identify HLA-E and -G reactive cells. This was done by first performing a T-cell stimulation assay and second a staining assay for the HLA-G and HLA-E tetramers.
Some peptides of HLA-G and HLA-E were also able to bind to conventional HLA-A,B or C, so we could not rule out reactivity via other HLA types (Supplementary tables S1 and S2). To test this, a T-cell stimulation assay was performed with PBMC donors with the most common HLA types
providing a large coverage to validate the role of HLA. A binding assay of the HLA-E peptides for the HLA-E01:03 tetramer showed that the peptides 1,4,5 and 9 of had a higher optical density (OD) compared to the mean (red dashed line) (Supplementary figure S1). Therefore, these four peptides were tested separately as one pole in the T-cell stimulation test. The results of the T-cell stimulation assays showed that both CD8+ and CD4+ T-cells did not produce TNF-α or IFN-γ after stimulation with the HLA-E or HLA-G peptide pools compared to the negative control (Supplementary Figure S3 & S4, general gating strategy in Supplementary Figure S2). These results suggested that HLA-G and HLA-E peptides were approximately not recognized by T-cells via other HLA types, so the HLA-G and HLA-E peptides were not immunogenic.
The E tetramer staining assay was performed to validate the UV-exchangeable HLA-E01:03 tetramer by comparing this tetramer with the “control” non-exchangeable HLA-HLA-E01:03 tetramer (See Supplementary figure S7 for the general gating strategy). CD94+ cells from total CD8+ T-cells were able to bind to the “control” HLA-E01:03 tetramer with VMAPRTLIL peptide (Figure 1Af). Similar results can be seen for the UV-exchangeable HLA-E01:03 tetramer with the same peptide VMAPRTLIL (Figure 1Af). Moreover, there was no difference observed in reactivity between the two HLA-E01:03 tetramers. CD3+ T-cells, CD8a+ T-cells, CD8b+ T-cells, CD8a+&CD8b+ T-cells, CD94+ cells, CD16+ cells, CD3+&CD16+ cells and KLRG1+ cells showed a response to both HLA-E01:03 tetramers (Figure 1A). Furthermore, the HLA-E+ cells expressed the effector markers CD45RA and KLRG1, so it can be concluded that the HLA-E reactive CD8+ T-cells express an effector phenotype (Figure 1Ac,d,e
and i). Overall, these results suggest that the UV-exchangeable HLA-E01:03 tetramer can be used in
future experiments to identify HLA-E reactive T-cells.
The same conclusion applies for the UV-exchangeable HLA-G01:01 and HLA-G05:01 tetramers, because CD3+ T-cells (HLA-G01:01 p-value = 0,0399 and HLA-G05:01 p-value = 0,0468), CD3+&CD16+ cells (HLA-G01:01 p-value = 0,0362 and HLA-G05:01 p-value = 0,0410), CD14+ cells (HLA-G01:01 p-value = 0,0004 and HLA-G05:01 p-value = 0,0002) and CD3+&CD14+ cells (HLA-G01:01 p-value = 0,0055 and G05:01 p-value = 0,0066) showed a significant response to both HLA-G01:01 and HLA-G05:01 tetramers. Furthermore, there was no difference observed in response between both HLA-G tetramers (See Supplementary Figure S6 for the HLA-G tetramers results and
Supplementary Figure S5 for the general gating strategy). Because of the COVID-19 outbreak it was
no longer possible to test the effect of different peptides on the recognition of T-cells. Hence, we proceeded to phenotype the T-cells that recognize the HLA-E01:03 tetramer with VMAPRTLIL peptide in blood, lung and tumor samples of NSCLC-patients.
Figure 1| UV-exchangeable HLA-E01:03 tetramer reactivity compared to the non-exchangeable HLA-E tetramer reactivity. (A) Representative pseudocolor plots of donor 2849 demonstrate an example of the expression of HLA-E and (a) CD3, (b)
CD45RA for CD4+ T-cells (c) CD45RA for CD8a+ T-cells (d) CD45RA for CD8b+ T-cells (e) CD45RA for CD8a&CD8b double positive T-cells (f) CD94+ (from total CD8+ T-cells) (g) CD16, (h) CD45RA for CD3&CD16 double positive T-cells, (i) KLRG1 (from total CD8+ T-cells) and (j) TCRγδ for the UV-exchangeable tetramer (left), non-exchangeable tetramer (middle) and negative control (right).
HLA-E reactive CD8+ T-cells express an effector phenotype
Next, we assessed the phenotype of the T-cells that are able to recognize the UV-exchangeable HLA-E01:03 tetramer with VMAPRTLIL peptide using FACS. The T-cells originated from blood, healthy lungs and tumors of NSCLC-patients, so the differences between the HLA-E reactive T-cells in these
compartments could also be investigated (see Supplementary figures S9, S10 and S11 for the general gating strategy of blood, lung and tumor samples, respectively).
The percentage of CD4+ T-cells was significantly higher than the percentage of CD8+ T-cells in blood (p-value = 0,0083) (Figure 2A). The same result applies for the tumor samples, but this
difference was not significant (Figure 2A). In contrast, the percentage of CD8+ T-cells was significantly higher than the percentage of CD4+ T-cells in the lung (p-value = 0,0417). Additionally, analysis of the phenotypes of the CD3+, CD4+ and CD8+ T-cells in blood, lung and tumor samples were determined by the expression of CD27 and CD45RA (see Figure 2B for representative flow cytometry plots). In blood, all four phenotypes were present: 7.6% Central Memory, 15.9% Naive, 30.3% Effector Memory and 39.6% Effector cells for the CD8+ T-cells (Figure 2C). The frequency of CD27+&CD45RA+ (naïve) cells was high in blood CD8+ cells (15.9%) (Figure 2C). In contract, both the lung and tumor CD8+ T-cells showed low frequencies of CD27+&CD45RA+ naïve T-cells (lung: 0.6% and tumor: 0.6 %) and high frequencies of CD27-CD45RA- effector memory cells (lung: 51,5% and tumor: 68.2%) (Figure 2D and
2E). Moreover, CD4+ T-cells showed low frequencies of CD27-CD45RA+ effector cells in the three
compartments (blood: 4.7%, lung: 6.9% and tumor: 3,2%), while this effector population dominates in the CD8+ T-cell population in all the compartments (blood: 39.6%, lung: 44.1% and tumor: 22.8%) (Figure 2C, 2D and 2E). Overall, these results suggest that HLA-E reactive CD8+ T-cells in blood, lung and tumor express an effector phenotype.
Representative flow cytometry plots shown the expression of CD103 and CD69 by CD8+ T-cells and demonstrated that the lung (23,9%) and tumor (51,2%) compartments contain
CD103+CD69+ resident memory T-cells (TRM), but these TRM cells were not present in blood (Figure
3D). These results are summarized for all the donors per compartment and showed the same trend as
the flow cytometry plots, because 22,4% of CD8+ T-cells in the lung are resident memory T-cells and 45,1% of CD8+ T-cells in the tumor (Figures 3A, 3B and 3C). Representative flow cytometry plots of donor 55 showed the expression of CD69 and CD103 by CD4+ T-cells and CD8+ T-cells with the HLA-E reactive CD27-CD45RA+CCR7- effector T-cells indicated in red (Figure 3E). CD69+ effector T-cells were absent in blood, but present in the lung and tumor of CD8+ T-cells. CD69 is a tissue resident marker, which means that these CD8+ T-cells will stay in the lung and/or tumor tissue. The HLA-E reactive effector cells did not show an CD103+CD69+ phenotype, which means that the HLA-E reactive T-cells are effector T-cells and not resident memory T-T-cells (TRM). Together, these results suggest that non-resident effector CD8+ T-cells are found in the tumor.
Figure 2| Distribution and phenotype of HLA-E reactive T-cells in paired blood, lung and tumor samples. (A) Frequencies
of CD4+ (black circles) and CD8+ (white circles) cells of total CD3+ T-cells of paired blood, lung and tumor tissue was analyzed by flow cytometry (**p-value= 0,0083 for blood CD4+ vs. CD8+ cells, *p-value = 0,0417 for lung CD4+ vs. CD8+ cells). (B-E) The expression of CD27 and CD45RA was analyzed on paired blood, lung and tumor CD3+, CD4+ and CD8+ T-cells. (B) Flow cytometry plots show representative examples for the expression of CD27 and CD45RA by CD3+ T-cells (left panel), CD4+ T-cells (middle panel) and CD8+ T-cells (right panel) in blood (top panel), lung (middle panel) and tumor (bottom panel) samples. (C-E) Summarized results of the percentage of the 4 different phenotypes in (C) paired blood samples (D) paired lung samples and (E) paired tumor samples. *p < 0.05, **p < 0.01, ***p < 0.001; unpaired T Test with Welch correction test.
Figure 3| Tissue resident phenotype of HLA-E reactive T-cells in paired blood, lung and tumor samples. (A-C) Summarized
results of the percentages of CD103-CD69- (in grey), CD103+CD69- (in pink), CD103+ CD69+ (resident memory T-cells (TRM) In purple) and CD103-CD69+ cells (in black) of (A) paired blood samples (B) paired lung samples and (C) paired tumor samples. (D) Flow cytometry plots of donor 55 show representative examples for the expression of CD103 and CD27 by CD8+ T-cells in blood (top panel), lung (middle panel) and tumor (bottom panel) samples. (E) Representative flow cytometry plots of donor 55 show the expression of CD69 and CD103 by CD4+ T-cells (left panels) and CD8+ T-cells (right panels) in a blood (top), lung (middle) and tumor (bottom) sample. The CD27-CD45RA+CCR7- effector cells are indicated in red.
Helios expression is enriched in HLA-E reactive effector CD8+ T-cells
Next, we wanted to investigate whether non-resident effector CD8+ T-cells were reactive for HLA-E in the tumor. On average, 19,5% of Helios+ CD8+ TCRγδ- effector cells showed HLA-E expression
compared to 5,4% of Helios- effector cells in blood (Figure 4A & 4B, for the general gating strategy see Supplementary Figure S8). Hence, we used the transcription factor Helios as a proxy for the HLA-E reactive effector T-cells. In this result section, the differences between Helios+ cells among CD4+ and CD8+ T-cells cells and the differences of Helios expression between the three different
compartments were investigated for blood, lung and tumor samples of 10 NSCLC patients (see
Supplementary figures S9, S10 and S11 for the general gating strategy of blood, lung and tumor
samples, respectively).
The frequency of Helios+ CD8+ T-cells was significantly higher than the percentage of Helios+ CD4+ T-cells in blood (p-value = 0,0314 ), lung (p-value = 0,0065) and tumor (p-value = 0,0007) samples (Figure 4C). In addition, approximately 35% of the CD8+ effector T-cells expressed the transcription factor Helios in the three compartments (Figure 4D). Furthermore, it seemed that there was a trend toward Helios expression in the tumor compartment, because the percentage of Helios+ cells in CD8+ T-cells was higher in the tumor (39,3%) compared to the lung (33,0%) and blood (28,0%) compartments, but this difference was not significant (Figure 4D).
Besides the CD8+ T-cells, the differences between Helios expression between the three compartments was also investigated among CD4+ T-cells (Figure 4E). These results were completely different from the results of the CD8+ T-cells, because the percentage of Helios+ cells in CD4+ T-cells was low in lung (6,3%) and tumor (7,4%) tissue compared to blood (12,7%) (Figure 4E). This
difference was again not significant, but the difference between the lung and blood was close to significance (p-value = 0,0609). Overall, these results suggest that Helios expression is enriched in effector CD8+ T-cells.
Figure 4| Differences between Helios+ cells in the compartments blood, lung and tumor. (A) A representative contour plot
demonstrates the expression of HLA-E and HELIOS by CD8+ TCRγδ- effector cells. (B) Quantification of the expression of HLA-E in Helios- (grey squares) and Helios+ (black circles) CD8+ TCRγδ- effector cells in blood. (C) Frequencies of CD4+ (black circles) and CD8+ (white circles) cells of total Helios+ cells of paired blood, lung and tumor tissue was analyzed by flow cytometry (* p-value= 0,0314 for blood CD4+ vs. CD8+ T-cells, **p-value = 0,0065 for lung CD4+ vs. CD8+ T-cells and ***p-value = 0,0007 for tumor CD4+ vs. CD8+ cells). (D-E) Summarized results of (D) the percentage of Helios+ effector CD8+ T-cells and (E) the percentage of Helios+ effector CD4+ T-T-cells in paired blood, lung and tumor samples (p-value = 0,0609 for lung vs blood). *p < 0.05, **p < 0.01, ***p < 0.001; unpaired T Test with Welch correction test.
Helios+ effector CD8+ T-cells express an innate-like phenotype
Next, we assessed the phenotypic characterization of Helios+ effector T-cells with known innate markers such as CD94, NKG2A, NKG2C and NKG2D and we reanalyzed differentially expressed markers such as CD2, CD5, CD6, CD158ab, Granzyme B (GZMB) and TIGIT from a deep single-cell RNA
sequencing experiment of Guo et al. (2018) that characterized 12.346 T-cells from 14 treatment-naïve NSCLC-patients (data not shown).
HLA-E reactive T-cells express an effector phenotype, so in this experiment the Helios+ effector (CD27-CD45RA+CCR7-) CD8+ T-cells were analyzed (see Supplementary figures S9, S10 and
S11 for the general gating strategy for the blood, lung and tumor samples, respectively). The innate
marker NKG2C (p-value= <0,0001 for Helios+ vs. Helios-) was significantly higher expressed in Helios+ cells compared to Helios- cells and the innate marker NKG2D (p-value= 0,0266 for Helios+ vs. Helios-) was significantly lower expressed in Helios+ cells compared to Helios- cells (Figure 5A). Furthermore, innate markers CD94 (20,0% for Helios+ and 10,1% for Helios-) and NKG2A (17,3% for Helios+ and 7,0% for Helios-) showed a higher expression in Helios+ cells compared to Helios- cells, but this difference was not significant (Figure 5A). So, from these results it can be concluded that Helios+ effector CD8+ T-cells express an innate-like phenotype.
Moreover, the frequencies of the expression of the NSCLC-markers CD2, CD5, CD6, CD158ab, Granzyme B (GZMB) and TIGIT by Helios+ and Helios - cells in blood, lung and tumor were analyzed using FACS (Figures 5B, 5C and 5D and see Supplementary figure S12 for the general gating strategy of the markers). Lymphocyte and surface markers CD2, CD5 and CD6 were lower expressed in the Helios+ cells compared to Helios- cells in the three compartments (Figures 5B, 5C and 5D). Lymphocyte marker CD2 showed a close to significance difference in blood (p-value = 0,0840) and lung (p-value = 0,0701) and a significant difference in tumor (p-value = 0,0155) samples (Figures 5B,
5F and FG). Lymphocyte marker CD5 showed a significant difference in lung (p-value = 0,0484) and
tumor (p-value = 0,0056) samples (Figures 5B and 5C). Furthermore, lymphocyte marker CD6 showed a significant difference in blood (p-value = <0,0001) lung (p-value = 0,0009) and tumor (p-value = 0,0011) (Figures 5B, 5C and 5D) samples. In contrast, the regulatory T-cell marker and surface marker TIGIT had a significant higher expression in Helios+ cells compared to Helios- cells in blood (p-value = 0,0014), lung (p-value = <0,0001) and tumor (p-value = 0,0436) (Figures 5E, 5F and 5G). In addition, it seemed that the innate killer-cell immunoglobulin-like (KIR) marker CD158ab showed a higher expression in Helios+ cells compared to Helios- cells in blood (Helios+ : 12,2% vs Helios- : 3,1%), lung (Helios+ : 13,7% vs Helios- : 2,4%) and tumor (Helios+ : 28,9% vs Helios- : 12,9%) samples, but this difference was not significant (Figure 5B, 5C and 5D). The marker Granzyme B was not differentially expressed between Helios+ and Helios- cells in the three compartments.
CD4+ T-cells showed low frequencies of CD27-CD45RA+ (effector) cells in the three
compartments (blood: 4.7%, lung: 6.9% and tumor: 3,2%), so CD8+ T-cells give probably more reliable results, because the amount of effector CD8+ T-cells is higher in the three compartments (blood: 39.6%, lung: 44.1% and tumor: 22.8%) (Figure 2C, 2D and 2E & Supplementary figure S13A). Nevertheless, the same analysis of the expression of the NSCLC-markers in Helios+ and Helios- cells was still done for the CD4+ T-cells (Supplementary figure S13B, S13C and S13D). The CD4+ T-cell results were again completely different from the results of the CD8+ T-cells, because the expression of the lymphocyte markers CD2, CD5, CD6 and regulatory T-cell marker TIGIT were higher expressed in Helios+ cells and the KIR marker CD158ab was higher expressed in Helios- cells or not expressed at all in the three compartments (Supplementary figure S13). All together, these results suggested that Helios+ effector CD8+ T-cells express the innate markers CD94, NKG2A, NKG2C and that the markers CD2, CD5 and CD6 showed a lower expression in the Helios+ cells. Thus, Helios+ HLA-E reactive CD8+ T-cells express an innate-like phenotype.
Figur e 5| Expression of innate & NSCLC-markers by effector CD8+ T-cells in paired blood, lung and tumor samples. (A) The expression of NKG2A markers in Helios+ (black circles) and Helios – (gray squares) CD8+ TCRγδ- effector cells: CD94,
NKG2A, NKG2C (****p-value= <0,0001 for Helios+ vs. Helios-) and NKG2D (*p-value= 0,0266 for Helios+ vs. Helios-). (B-D Frequencies of the expression of the NSCLC-markers CD2, CD5, CD6, CD158ab (CD158), Granzyme B (GZMB) and TIGIT by Helios+ (black circles) and Helios – (gray squares) was analyzed in the three compartments (B) blood (CD2 p-value= 0,0840; CD6 ****value <0,0001 and TIGIT *value= 0,0164) (C) lung (CD5 *value= 0,0484; CD6 ***value = 0,0009; TIGIT p-value = 0,0695) and (D) tumor (CD5 **p-p-value= 0,0056; CD6 **p-p-value= 0,0011). (E-G) The expression of the NSCLC-markers CD2, CD5, CD6, CD69, CD103, CD158ab (CD158), Granzyme B (GZMB) and TIGIT (geometric mean) by Helios+ cells (black circles) and Helios- cells (white circles) was analyzed in the different compartments (E) blood (CD6 *p-value=0,0392 and TIGIT **p-value=0,0014) (F) lung (CD2 p-value=0,0701 and TIGIT ****p-value= <0,0001) and (G) tumor (CD2 *p-value= 0,0115; CD6 *p-value= 0,0410 and TIGIT *p-value=0,0436). *p < 0.05, **p < 0.01, ***p < 0.001, ****p<0,0001; unpaired T Test with Welch’s correction.
CD5-CD2-CD6-TIGIT+ surface marker gating strategy can be used to isolate Helios+ cells in
lung and tumor samples
For future functional experiments in the characterization of Helios+ CD8+ T-cells is it important to know how to isolate Helios+ cells with a gating strategy without the need to obligatory kill cells due to fixation. Therefore, we made a surface marker gating strategy for Helios+ CD8+ T-cells. The surface markers CD2, CD5 and CD6 showed a lower expression and TIGIT showed a higher expression in Helios+ cells compared to Helios- cells in the in the 3 compartments (Figure 5B, 5C, 5D, 5E, 5F and
5G). These results correspond to the representative flow cytometry plots (Figure 6A), because these
plots showed that the Helios+ cells (indicated in red) had a double negative expression pattern for the markers CD5 and CD6, CD2 and CD6 and for CD5 and C2.
Following, a subsequent gating strategy to a CD5-CD6-CD2- and TIGIT+ population was made for a representative lung donor (Figure 6B). This gating strategy was done for all the three
compartments (blood, lung and tumor) and the resulting population is indicated in red (Figure 6C). This red CD5-CD6-CD2- and TIGIT+ population matched with the Helios+ population (in blue) (Figure
6C). The overlapping percentage of the Helios+ population and the CD5-CD6-CD2- and TIGIT+
population was in blood 20,9%, in the lung 79,3% and in the tumor 61,9% (Figure 6C). In addition, especially in the tumor are the CD8- cells excluded via the CD5-CD6-CD2- and TIGIT+ gating strategy
(Figure 6C). These results suggest that CD5-CD6-CD2- and TIGIT+ T-cells harbor a high percentage of
Figure 6| Helios+ gating strategy with surface markers. (A) Representative flow cytometry plots of donor 55 show the
expression of the surface marker CD6 and CD5 (left panel), CD6 and CD2 (middle panel), and CD2 and CD5 (right panel) in blood (top panel), lung (middle panel) and tumor (bottom panel) samples. The Helios+ population is shown in red dots and the Helios- population is shown in blue dots. (B) Representative pseudocolor plots of lung donor 55 demonstrate the subsequent gating strategy to a CD5-CD6-CD2- and TIGIT+ population. (C) Flow cytometry plots of donor 55 show the expression of Helios and CD8 in a blood (top), lung (middle) and tumor (bottom) sample. The CD5-CD6-CD2- and TIGIT+ population made with the gating strategy in (B) is shown in red dots and the Helios+ population is shown in blue dots. The overlapping percentage of these populations is in blood 20,9%, in the lung 79,3% and in tumor 61,9%.
DISCUSSION
The aim of this study was to acquire new insights in the characterization of HLA-G and HLA-E reactive T-cells found in blood, healthy lung and tumors of NSCLC-patients. We found that UV-exchangeable HLA-E and HLA-G tetramers can be used to identify HLA-E and HLA-G reactive T-cells. In addition, we found that T-cells that are able to recognize the UV-exchangeable HLA-E tetramer with VMAPRTLIL peptide express an effector phenotype and that the transcription factor Helios is enriched in HLA-E reactive effector CD8+ T-cells. In addition, we found that CD5-CD6-CD2- and TIGIT+ T-cells harbor a high percentage of Helios+ cells in the lungs and tumor. Overall, it can be concluded that non-classical HLA reactive CD8+ T-cells harbor an innate-like phenotype and that they can be identified by Helios expression in NSCLC.
This study was mainly focused on HLA-E (due to the COVID-19 measures no further studies could be done with the exchangeable HLA-G tetramers). As we showed the feasibility of using UV-exchangeable HLA-G tetramers to identify reactive T-cells by FACS, these reagents may be used for further research to investigate the potential role of peptide-specificity. Future research can be done into whether there is a difference in binding capacity of T-cells to the HLA-G tetramer with peptide and whether there is a peptide with the most effective binding. In addition, future research can be done into whether the binding of T-cells to this tetramer is peptide specific or not by exchanging the peptide to other HLA-G peptides (See Supplementary table S1 for the other possible HLA-G
peptides). The next step can be the characterization of HLA-G reactive T-cell populations.
We showed that the transcription factor Helios is enriched in HLA-E reactive effector CD8+ T-cells. Helios is also expressed by regulatory T-cells (Treg cells) (Thornton et al., 2010). In line with CD4+ Tregs, Helios+ HLA-E reactive effector CD8+ T-cells, MAIT cells and NKT-cells function
independently of their αβTC (Suliman et al., 2019).Therefore, it is interesting to investigate whether the expression of Helios can push cells to an innate phenotype by making them independent of their αβTCR. Further research can be done into whether Helios is also expressed in MAIT cells and NKT-cells and whether this expression is acquired for their function. The Helios+ gating strategy discussed in this study can be used to isolate Helios+ cells and check whether MAIT cells and NKT-cells also express Helios.
Since the percentage of Helios+ HLA-E reactive effector CD8+ T-cells seemed to be higher in tumor compared to blood, it may be an attractive option to target intratumorally HLA-E reactive effector CD8+ T-cells in NSCLC. We showed that Helios+ HLA-E reactive effector CD8+ T-cells in the tumor express the tissue resident marker CD69. This result suggests that HLA-E reactive CD8+ T-cells can perhaps play a role in new immunotherapy strategies for NSCLC, because these innate-like T-cells will remain intratumorally. Therefore, future research can be done into the function of Helios+ HLA-E reactive T-cells in NSCLC. The Helios+ gating strategy discussed in this study for tumor samples can be used for this experiment. Furthermore, the difference between the percentage of Helios+ HLA-E reactive effector CD8+ T-cells in tumor was not significantly higher compared to blood. Therefore, this experiment can be repeated to see if a significant difference exists between these compartments.
NK cells and CD8+ T-cells play an important role in the host defense against tumors, because NK cells produce cytokines and CD8+ T-cells have the ability to kill tumor-infected cells through cytotoxic activity (Bi & Tian, 2017). However, the immunosuppressive NSCLC microenvironment leads to the exhaustion of NK cells, because NK cells are for example inhibited by TGFβ that is produced by Tregs present in the immunosuppressive NSCLC microenvironment (Cremer et al., 2012). In addition, CD8+ T-cells are exhausted due to chronic stimulation by inhibitory markers of the tumor (Jiang & Zhu, 2015).Thus, future research must be done into immune cells that are not downregulated by the immunosuppressive microenvironment and are not downregulated due to chronic stimulation by inhibitory markers of the tumor. We showed that Helios+ HLA-E reactive effector CD8+ T-cells have an innate-like phenotype. These innate-like T-cells have a lot in common with NK cells, because Helios+ HLA-E reactive effector CD8+ T-cells are independent of their αβTC and can be activated by the integration of activating signals from activating receptors CD94/NKG2C and CD94/NKG2D (Bi & Tian, 2017). Thus, perhaps are Helios+ HLA-E reactive innate-like CD8+ T-cells not downregulated due to
chronic stimulation of inhibitory markers, because HLA-E reactive cells recognizes the tumor like NK cells do. In addition, Helios+ HLA-E reactive innate-like CD8+ T-cells belong to the T-cell lineage, because they have a TCR. Thus, perhaps are Helios+ HLA-E reactive innate-like CD8+ T-cells not downregulated by the immunosuppressive microenvironment. Future research must be done into the clinical outcome of using Helios+ HLA-E reactive innate-like CD8+ T-cells in NSCLC. Moreover, further research can be done into whether Helios+ HLA-E reactive innate-like CD8+ T-cells have a negative effect in NSCLC, because HLA-E is overexpressed in NSCLC and is associated with disease progress and bad prognosis for NSCLC-patients (Lin & Yan, 2018 ; Creelan & Antonia, 2019). The Helios+ gating strategy discussed in this study for tumor samples can be used for this functional experiment.
Currently, the world has to deal with the new betacoronavirus SARS-CoV-2, which causes the disease COVID-19 and has a mortality rate of 3,4% (World Health Organization, 2020). Research of Huang et al. (2020) showed that the innate immune response plays a role in COVID-19, because the so called ‘cytokine storm’ or ‘cytokine release syndrome’ (CRS) is a characteristic of COVID-19 in the acute phase of the disease (Wu et al, 2020 & Rokni et al., 2020 & Li et al., 2020). Some of these cytokines, such as IL-10 and IL-6 showed to evoke an increased expression of the exhaustion marker NKG2A in COVID-19 patients (Cho et al., 2011). Research of Zheng et al. (2020) and Antonioli et al., (2020) indicated that exhausted NK cells and CD8+ T-cells were primarily involved in the anti-COVID-19 response and that these cells showed an upregulation of the NKG2A receptor. The NKG2A receptor is an inhibitory receptor expressed on NK cells, CD8+ T-cells and, according to this study Helios+ HLA-E reactive T-cells, that binds to the HLA-HLA-E class molecule (Kanevskiy et al., 2019). This binding leads to the inhibition and reduction of lymphocytes and thus the suppression of the immune response, whereby for example the Sars-Cov-2 virus can spread more easily (Moser et al., 2002). Further research can be carried out into whether it is possible to block the cytokines IL-10 and IL-6 with a monoclonal antibody, so the expression of the NKG2A receptor on NK cells and CD8+ T-cells is no longer increased. Another idea for future research is whether it is possible to block the NKG2A receptor on Helios+ HLA-E reactive T-cells, NK cells and CD8+ T-cells. When the NKG2A receptor is blocked, the lymphocytes are no longer inhibited and the antiviral immune response can be restored against Sars-Cov-2. Research of van Hall et al. (2019) has indicated that the monoclonal antibody Monalizumab is able to target NKG2A receptors expressed on NK cells and CD8+ T-cells. Further research can be done into whether Monalizumab or the combination of Monalizumab with a monoclonal antibody against IL-6 and IL-10 can be used in the treatment of COVID-19.
In addition, we showed that Helios+ HLA-E reactive innate-like CD8+ T-cells are present in the lungs. So, this results together with the fact that the innate immune response plays a role in COVID-19 and that the marker NKG2A is upregulated in COVID-COVID-19 patients, could be evidence that Helios+ HLA-E reactive innate-like CD8+ T-cells are involved in COVID-19 disease. Follow-up research can be done into the function of these Helios+ HLA-E reactive innate-like CD8+ T-cells in COVID-19, whether they have a positive or negative effect on the replication of Sars-Cov-2. If HLA-E reactive T-cells are found to inhibit Sars-Cov-2 virus replication, these T-cells could perhaps be used as new therapy for COVID-19. An advantage of the innate immune system is that the innate immune response is faster than the adaptive immune response. However, a disadvantage of the innate immune response is that this response is independent of antigen recognition, which can be dangerous when they are self-reactive (Berg & Forman, 2006). So, it is very important that this potential problem is kept in mind during follow-up studies into whether these Helios+ HLA-E reactive innate-like CD8+ T-cells can be used in the fight against COVID-19.
In conclusion, non-classical HLA reactive CD8+ T-cells harbor an innate-like phenotype and can be identified by Helios expression in NSCLC. In addition, this T-cell characterization has given new insights into possible immunotherapies for NSCLC and COVID-19 that can be used to improve
immunotherapy regimens for NSCLC and can be helpful in designing a therapy for COVID-19 in the future.
References
1. Antonioli, L., Fornai, M., Pellegrini, C., & Blandizzi, C. (2020). NKG2A and COVID-19: another brick in the wall. Cellular and Molecular Immunology. https://doi.org/10.1038/s41423-020-0450-7 2. Berg, R. E., & Forman, J. (2006). The role of CD8 T-cells in innate immunity and in antigen
non-specific protection. Current Opinion in Immunology. https://doi.org/10.1016/j.coi.2006.03.010 3. Bi, J., & Tian, Z. (2017). NK cell exhaustion. Frontiers in Immunology.
https://doi.org/10.3389/fimmu.2017.00760
4. Brubaker, S. W., Bonham, K. S., Zanoni, I., & Kagan, J. C. (2015). Innate Immune Pattern Recognition: A Cell Biological Perspective. Annual Review of Immunology.
https://doi.org/10.1146/annurev-immunol-032414-112240
5. Cho, J. H., Kim, H. O., Webster, K., Palendira, M., Hahm, B., Kim, K. S., … Sprent, J. (2011). Calcineurin-dependent negative regulation of CD94/NKG2A expression on naive CD8+ T-cells. Blood. https://doi.org/10.1182/blood-2010-11-317396
6. Creelan, B. C., & Antonia, S. J. (2019). The NKG2A immune checkpoint — a new direction in cancer immunotherapy. Nature Reviews Clinical Oncology. https://doi.org/10.1038/s41571-019-0182-8 7. Cremer, I., Fridman, W. H., & Sautès-Fridman, C. (2012). Tumor microenvironment in NSCLC
suppresses NK cells function. OncoImmunology. https://doi.org/10.4161/onci.1.2.18309 8. Godfrey, D. I., Uldrich, A. P., McCluskey, J., Rossjohn, J., & Moody, D. B. (2015). The burgeoning
family of unconventional T-cells. Nature Immunology, 16(11), 1114 - 1123.
https://doi.org/10.1038/ni.3298
9. Grant, E. J., Nguyen, A. T., Lobos, C. A., Szeto, C., Chatzileontiadou, D. S. M., & Gras, S. (2020). The unconventional role of HLA-E: The road less traveled. Molecular Immunology.
https://doi.org/10.1016/j.molimm.2020.02.011
10. Guo, X., Zhang, Y., Zheng, L., Zheng, C., Song, J., Zhang, Q., … Zhang, Z. (2018). Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing. Nature Medicine. https://doi.org/10.1038/s41591-018-0045-3
11. Herbst, R. S., Morgensztern, D., & Boshoff, C. (2018). The biology and management of non-small cell lung cancer. Nature. https://doi.org/10.1038/nature25183
12. Huang, C., Wang, Y., Li, X., Ren, L., Zhao, J., Hu, Y., … Cao, B. (2020). Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The Lancet.
https://doi.org/10.1016/S0140-6736(20)30183-5 13. IEDB (https://www.iedb.org/)
14. Janeway, C. A. (1989). Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harbor Symposia on Quantitative Biology. https://doi.org/10.1101/sqb.1989.054.01.003 15. Jiang, Y., Li, Y., & Zhu, B. (2015). T-cell exhaustion in the tumor microenvironment. Cell Death and
Disease. https://doi.org/10.1038/cddis.2015.162
16. Joosten, S. A., Sullivan, L. C., & Ottenhoff, T. H. M. (2016). Characteristics of HLA-E Restricted T-Cell Responses and Their Role in Infectious Diseases. Journal of Immunology Research.
https://doi.org/10.1155/2016/2695396
17. Kanevskiy, L., Erokhina, S., Kobyzeva, P., Streltsova, M., Sapozhnikov, A., & Kovalenko, E. (2019). Dimorphism of HLA-E and its disease association. International Journal of Molecular Sciences. https://doi.org/10.3390/ijms20215496
18. Kumar, H., Kawai, T., & Akira, S. (2011). Pathogen recognition by the innate immune system. International Reviews of Immunology. https://doi.org/10.3109/08830185.2010.529976 19. Lee, Y. J., Jameson, S. C., & Hogquist, K. A. (2011). Alternative memory in the CD8 T cell lineage.
Trends in Immunology. https://doi.org/10.1016/j.it.2010.12.004
20. Lin, A., & Yan, W. H. (2018). Heterogeneity of HLA-G Expression in Cancers: Facing the Challenges. Frontiers in Immunology. https://doi.org/10.3389/fimmu.2018.02164
21. Li, X., Geng, M., Peng, Y., Meng, L., & Lu, S. (2020). Molecular immune pathogenesis and diagnosis of COVID-19. Journal of Pharmaceutical Analysis. https://doi.org/10.1016/j.jpha.2020.03.001 22. Martino, A., Meraviglia, S., El Daker, S., Dieli, F., & Martini, F. (2011). T-cells cross-link innate and
adaptive immunity in mycobacterium tuberculosis infection. Clinical and Developmental Immunology. https://doi.org/10.1155/2011/587315
23. Moretta, L., Pietra, G., Romagnani, C., Manzini, C., & Mingari, M. C. (2010). The emerging role of HLA-E-restricted CD8+ T lymphocytes in the adaptive immune response to pathogens and tumors. Journal of Biomedicine and Biotechnology. https://doi.org/10.1155/2010/907092
24. Moser, J. M., Gibbs, J., Jensen, P. E., & Lukacher, A. E. (2002). CD94-NKG2A receptors regulate antiviral CD8+ T cell responses. Nature Immunology. https://doi.org/10.1038/ni757
25. Murphy, K., & Weaver, C. (2017). Janeway ’S immunobiology 9th Edition. Garland Science : Taylor & Francis Group. https://doi.org/10.1007/978-981-10-2116-9
26. Net MHC (http://www.cbs.dtu.dk/services/NetMHC/) 27. NetMHCpan (http://www.cbs.dtu.dk/services/NetMHCpan/).
28. Pietra, G., Romagnani, C., Moretta, L., & Mingari, M. (2009). HLA-E and HLA-E-Bound Peptides: Recognition by Subsets of NK and T-cells. Current Pharmaceutical Design.
https://doi.org/10.2174/138161209789105207
29. Ribas, A. (2015). Adaptive immune resistance: How cancer protects from immune attack. Cancer Discovery. https://doi.org/10.1158/2159-8290.CD-15-0563v
30. Rizvi, N. A., Hellmann, M. D., Snyder, A., Kvistborg, P., Makarov, V., Havel, J. J., … Chan, T. A. (2015). Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science. https://doi.org/10.1126/science.aaa1348
31. Rokni, M., Ghasemi, V., & Tavakoli, Z. (2020). Immune responses and pathogenesis of SARS CoV 2 ‐ ‐ during an outbreak in Iran: Comparison with SARS and MERS. Medical Virology.
https://doi.org/10.1002/rmv.2107
32. Rölle, A., Meyer, M., Calderazzo, S., Jäger, D., & Momburg, F. (2018). Distinct HLA-E Peptide Complexes Modify Antibody-Driven Effector Functions of Adaptive NK Cells. Cell Reports. https://doi.org/10.1016/j.celrep.2018.07.069
33. Seliger, B., Jasinski-Bergner, S., Quandt, D., Stoehr, C., Bukur, J., Wach, S., … Hartmann, A. (2016). HLA-E expression and its clinical relevance in human renal cell carcinoma. Oncotarget.
https://doi.org/10.18632/oncotarget.11744
34. Spits, H., & Di Santo, J. P. (2011). The expanding family of innate lymphoid cells: Regulators and effectors of immunity and tissue remodeling. Nature Immunology.
https://doi.org/10.1038/ni.1962
35. Suliman, S., Murphy, M., Musvosvi, M., Gela, A., Meermeier, E. W., Geldenhuys, H., … Scriba, T. J. (2019). MR1-Independent Activation of Human Mucosal-Associated Invariant T Cells by
Mycobacteria. The Journal of Immunology. https://doi.org/10.4049/jimmunol.1900674
36. Thornton, A. M., Korty, P. E., Tran, D. Q., Wohlfert, E. A., Murray, P. E., Belkaid, Y., & Shevach, E. M. (2010). Expression of Helios, an Ikaros Transcription Factor Family Member, Differentiates Thymic-Derived from Peripherally Induced Foxp3 + T Regulatory Cells . The Journal of Immunology. https://doi.org/10.4049/jimmunol.0904028
37. Van Hall, T., André, P., Horowitz, A., Ruan, D. F., Borst, L., Zerbib, R., … Vivier, E. (2019).
Monalizumab: Inhibiting the novel immune checkpoint NKG2A. Journal for ImmunoTherapy of Cancer. https://doi.org/10.1186/s40425-019-0761-3
38. Vardhana, S. A., & Wolchok, J. D. (2020). The many faces of the anti-COVID immune response. Journal of Experimental Medicine. https://doi.org/10.1084/JEM.20200678
39. Woodland, D. L., & Scott, I. (2005). T cell memory in the lung airways. Proceedings of the American Thoracic Society. https://doi.org/10.1513/pats.200501-003AW
40. World Health Organization – Cancer (https://www.who.int/news-room/fact-sheets/detail/cancer, visited on 19-6-2020)
41. World Health Organization 2020 – COVID-19
42. Wu, D., & Yang, X. O. (2020). TH17 responses in cytokine storm of COVID-19: An emerging target of JAK2 inhibitor Fedratinib. Journal of Microbiology, Immunology and Infection.
https://doi.org/10.1016/j.jmii.2020.03.005
43. Zheng, M., Gao, Y., Wang, G., Song, G., Liu, S., Sun, D., … Tian, Z. (2020). Functional exhaustion of antiviral lymphocytes in COVID-19 patients. Cellular and Molecular Immunology. https://doi.org/10.1038/s41423-020-0402-2
ACKNOWLEDGEMENTS
First, I would like to thank Pleun Hombrink for giving me the opportunity to conduct my Bachelor’s internship in his T-cell group at Sanquin, department of Hematopoiesis. Pleun, thank you for your all your time, supervision, help, feedback, new insights for my future and a very successful internship where I learned a lot. I have learned a lot of new things about immunology, how to write a scientific paper, data analysis of FACS data with Flowjo and how I can properly transfer my data in graphs. Thank you very much for everything.
Secondly, I would like to thank Anna Oja, Cherien Chandour and Ruth Hagen for helping me in the lab and your daily availability for questions. In addition, I would like to thank Sanquin Reagents in cooperation with Wim van Esch and Giso Brasser for providing me the HLA-E and HLA-G peptide pools and UV-exchangeable HLA-G and HLA-E tetramers that were essential for my experiments. Moreover, I would like to thank the FACS facility for their help while using the Symphony.
Furthermore, I want to express my gratitude to the department of Hematopoiesis for the nice time and their kindness. I also want to thank the blood donors for their willingness to donate their blood. Finally, I want to thank Marieke van Ham for her time to asses my thesis and final presentation.
SUPPLEMENTARY MATERIAL
Supplementary table S1| HLA-G peptides information. Shown are the sequences of the HLA-G peptides (Sequence), the
amount of amino acids of the peptides (Amino acids), to which HLA-G allele(s) this peptide can bind (HLA + allele), to what other type of HLA this peptide is probably able to bind (Ability to bind to other HLA types?), the affinity of this binding in nM (Affinity nM), whether this binding is a weak binding (in pink), a strong binding (orange) or non-binding (green) (Binding) and which antigen is associated with this peptide sequence (Antigen). This information is obtained with IEDB
(https://www.iedb.org/), net MHC (http://www.cbs.dtu.dk/services/NetMHC/) and netMHCpan (http://www.cbs.dtu.dk/services/NetMHCpan/).
Supplementary table S2| HLA-E peptides information. Shown are the sequences of the HLA-E peptides (Sequence), the
amount of amino acids of the peptides (Amino acids), to which HLA-E allele(s) this peptide can bind (HLA + allele), to what other type of HLA this peptide is probably able to bind (Ability to bind to other HLA types?), the affinity of this binding in nM (Affinity nM), whether this binding is a weak binding (in pink), a strong binding (orange) or non-binding (green) (Binding), which antigen is associated with this peptide sequence (Antigen) and the related pathogen (Pathogen). This information is obtained with IEDB (https://www.iedb.org/), net MHC (http://www.cbs.dtu.dk/services/NetMHC/) and netMHCpan (http://www.cbs.dtu.dk/services/NetMHCpan/).
Supplementary figure S1| Optical density value of the different HLA-E peptides The HLA-E01:03 tetramer has been made
unstable by UV-light. Using an ELISA test, it was measured which HLA-E peptide was able to bind in the unstable complex, resulting in a stable HLA-E complex again. The OD-values are shown on the y-axis for each peptide. Peptides: 1,4,5 and 9 show a higher OD-value compared to the mean (red dashed line). These four peptides were tested separately as one pole in the T-cell stimulation test.
Supplementary figure S2| Representative flow cytometry plots of the general gating strategy for the stimulation assay (a) Lymphocyte gate, (b-c) single cells gate, (d) CD3+ cells gate, (e) CD8+ and CD4+ cells gate.
Supplementary figure S3| Representative flow cytometry plots of the production ofIFN-γ and TNF-α by CD8+ T-cells
The flow cytometry plots of donors 8356 (left panels) and 3411 (right panels) show representative examples for the expression of CD137 and TNF-α (left two panels) and CD137 and IFN-γ (right two panels) by CD8+ T-cells under 8 different conditions in order from top to bottom: rest, 40% DMSO, CD3, CMV, E all peptides, E peptides 1,4,5 and 9 and HLA-G all peptides.
Supplementary figure S4| Representative flow cytometry plots of the production ofIFN-γ and TNF-α by CD4+ T-cells
The flow cytometry plots of donors 8356 (left panels) and 3411 (right panels) show representative examples for the expression of CD40L and TNF-α (left two panels) and CD40L and IFN-γ (right two panels) by CD4+ T-cells under 8 different conditions in order from top to bottom: rest, 40% DMSO, CD3, CMV, E all peptides, E peptides 1,4,5 and 9 and HLA-G all peptides.
Supplementary figure S5| Representative flow cytometry plots of the general gating strategy of HLA-G tetramer staining (a) Lymphocyte gate, (b-c) single cells gate, (d) CD3+ cells gate, (e) CD8+, CD4+ and double positive (DP) cells gate (f) CD14+
cells gate (g) CD16+ and CD3+&CD16+ double positive (DP) cells gate (h) CD3+&CD14+ double positive cells gate (i-j) The phenotype gates of (i) CD4+ and (j) CD8+ T-cells based on the expression of CD27 and CD45RA.
Supplementary figure S6| HLA-G01:01 and HLA-G05:01 tetramer reactivity compared to the negative control. (A-G) The
expression of HLA-G (geometric mean) was analyzed on (A) CD3+ (*p-value= 0,0399 for HLA-G01:01 vs. negative control and *p-value= 0,0468 for HLA-G05:01 vs. negative control) (B) CD4+ (*p-value= 0,0385 for HLA-G01:01 vs. negative control) (C) CD8+ (*p-value= 0,0426 for HLA-G05:01 vs. negative control), (D) CD16+, (E) CD3&CD16 double positive (*p-value= 0,0362 for HLA-G01:01 vs. negative control and *p-value= 0,0410 for HLA-G05:01 vs. negative control), (F) CD14+ (***p-value= 0,0004 for HLA-G01:01 vs. negative control and ***p-value= 0,0002 for HLA-G05:01 vs. negative control) and (G) CD3&CD14 double positive cells (**p-value= 0,0055 for HLA-G01:01 vs. negative control and **p-value= 0,0066 for HLA-G05:01 vs. negative control) of HLA-G01:01 (black circles), HLA-G05:01 (white circles), and the negative control (gray triangles). (H,I) The expression of HLA-G (geometric mean) was also analyzed on the different phenotypes of (H) CD4+ (*p-value= 0,0421 for HLA-G01:01 vs. negative control of Central Memory cells) and (I) CD8+ T-cells (*p-value= 0,03320 for HLA-G01:01 vs. negative control and *p-value= 0,0109 for G05:01 vs. negative control of Effector cells & *p-value= 0,0252 for G01:01 vs. negative control and *p-value= 0,0327 for G05:01 vs. negative control of Central Memory cells) of HLA-G01:01 (black circles), HLA-G05:01 (white circles), and negative control (gray triangles). *p < 0.05, **p < 0.01, ***p < 0.001; unpaired T Test.
Result: The HLA-G tetramer staining assay was performed to examine whether the HLA-G01:01 and HLA-G05:01 tetramer
reagents can be used for further experiments and whether there is a difference in tetramer reactivity between these tetramers (see Supplementary figure S5 for the general gating strategy). CD3+ T-cells (HLA-G01:01 p-value = 0,0399 and
HLA-G05:01 p-value = 0,0468), CD3+&CD16+ cells (HLA-G01:01 p-value = 0,0362 and HLA-G05:01 p-value = 0,0410), CD14+ cells (HLA-G01:01 p-value = 0,0004 and HLA-G05:01 p-value = 0,0002) and CD3+&CD14+ cells (HLA-G01:01 p-value = 0,0055 and HLA-G05:01 p-value = 0,0066) showed a significant response to both HLA-G01:01 and HLA-G05:01 tetramers
(Supplementary Figure S6A, S6D, S6E, S6F and S6G). Furthermore, there was no difference observed in response between
both tetramers. CD4+ T-cells (p-value = 0,0385) and specifically with a Central Memory phenotype, showed a significant response to HLA-G01:01 (p-value = 0,0385) (Supplementary Figure S6B, S6H). Furthermore, CD8+ T-cells showed a significant response to HLA-G05:01 (p-value = 0,0426) and CD8+ T-cells with an Effector (HLA-G01:01 p-value = 0,03320 and HLA-G05:01 p-value = 0,0109) and Central Memory (HLA-G01:01 p-value = 0,0252 and HLA-G05:01 p-value = 0,0327) phenotype showed a significant response to both tetramers (Figure S6C, S6I). Overall, these results suggest that the UV-exchangeable HLA-G tetramers showed reactivity and therefore can be used in future binding experiments.
Supplementary figure S7| Representative flow cytometry plots of the general gating strategy of HLA-E tetramer staining (a) Lymphocyte gate, (b-c) single cells gate, (d) CD3+ cells gate, (e) CD8a+ and CD4+ T-cells gate, f) CD8b+ T-cell gate, (g)
CD8a+&CD8b+ double positive T-cell gate, (h) CD16+ and CD3+&CD16+ double positive cells gate, (i) CD94+ cells gate, (j) KLRG1+ cells gate, (k) TCR γδ+ cells gate, (l-n) The phenotype gates of (l) CD3+ (m) CD4+ and (n) CD8+ T-cells based on the expression of CD27 and CD45RA.
Supplementary figure S8| Representative flow cytometry plots of the general gating strategy (made by Pleun Hombrink) (a) Lymphocyte gate, (b-c) single cells gate, (d) CD3+ cells gate, (e) MR1 (MAIT) and TCRγδ negative cells gate, because
these non-conventional T-cells are already HELIOS+ (f) CD8+ and CD4+ cells gate, (g) phenotype gates, the CD27- CD45RA+ effector cells were taken to the next gating, (j) CCR7- cells gate, because effector cells are CD27- CD45RA+ CCR7-, (k) HELIOS+ and HELIOS- gate.
Supplementary figure S9| Representative flow cytometry plots of the general gating strategy for the blood samples (a) Lymphocyte gate, (b-c) single cells gate, (d) living cells gate, (e) CD3+ cells gate, (f) clean gate, because there were
double positive aggregates in the samples probably due to the tar from smoking, (g) MR1 (MAIT) and TCRγδ negative cells gate, because these non-conventional T-cells are already HELIOS+ (h) CD8+ and CD4+ cells gate, (i) phenotype gates, the CD27- CD45RA+ effector cells (pink) were taken to the next gating, (j) CCR7- cells gate, because effector cells are CD27- CD45RA+ CCR7-, (k) HELIOS+ (black) gate and HELIOS- gate (grey).
Supplementary figure S10| Representative flow cytometry plots of the general gating strategy for the lung samples (a) Lymphocyte gate, (b-c) single cells gate, (d) living cells gate, (e) CD3+ cells gate, (f) clean gate, because there were
double positive aggregates in the samples probably due to the tar from smoking, (g) MR1 (MAIT) and TCRγδ negative cells gate, because these non-conventional T-cells are already HELIOS+ (h) CD8+ and CD4+ cells gate, (i) phenotype gates, the CD27- CD45RA+ effector cells (pink) were taken to the next gating, (j) CCR7- cells gate, because effector cells are CD27- CD45RA+ CCR7-, (k) HELIOS+ (black) gate and HELIOS- gate (grey).
Supplementary figure S11| Representative flow cytometry plots of the general gating strategy for the tumor samples (a) Lymphocyte gate, (b-c) single cells gate, (d) living cells gate, (e) CD3+ cells gate, (f) clean gate, because there were
double positive aggregates in the samples probably due to the tar from smoking, (g) MR1 (MAIT) and TCRγδ negative cells gate, because these non-conventional T-cells are already HELIOS+ (h) CD8+ and CD4+ cells gate, (i) phenotype gates, the CD27- CD45RA+ effector cells (pink) were taken to the next gating, (j) CCR7- cells gate, because effector cells are CD27- CD45RA+ CCR7-, (k) HELIOS+ (black) gate and HELIOS- gate (grey).
Supplementary figure S12| Representative flow cytometry plots of the general gating strategy for the different NSCLC-markers. (a-h) The gating strategy for Helios+ Marker- (black), Helios- Marker- (grey), Helios+ Marker+ (pink) and Helios-
Marker+ (blue) cells for the markers (a) CD158ab, (b) CD5, (c) CD103, (d) CD6, (e) TIGIT, (f) CD69, (g) CD2 and (h) Granzyme B.
Supplementary figure S13| Expression of innate markers by effector CD4+ T-cells in paired blood, lung and tumor samples. (A) Summarized results of the percentage of the different phenotypes of CD4+ T-cells in paired blood, lung and
tumor samples. (B-D) The expression of the markers CD2, CD5, CD6, CD69, CD103, CD158ab (CD158), Granzyme B (GZMB) and TIGIT (geometric mean) by Helios+ cells (black circles) and Helios- cells (white circles) was analyzed in the different compartments (B) blood (C) lung and (D) tumor.
