Reprogramming and redifferentiation of mucosal-associated invariant T cells reveals tumor inhibitory activity

Mucosal-associated invariant T (MAIT) cells belong to a family of innate-like T cells that bridge innate and adaptive immunities. Although MAIT cells have been implicated in tumor immunity, it currently remains unclear whether they function as tumor promoting or inhibitory cells. Therefore, we herein used induced pluripotent cell (iPSC) technology to investigate this issue. Murine MAIT cells were reprogrammed into iPSCs and redifferentiated towards MAIT-like cells (m-reMAIT cells). m-reMAIT cells were activated by an agonist and MR1-tetramer, a reagent to detect MAIT cells, in the presence and absence of antigen-presenting cells. This activation accompanied protein tyrosine phosphorylation and the production of T helper (Th)1-, Th2-, and Th17-cytokines and inflammatory chemokines. Upon adoptive transfer, m-reMAIT cells migrated to different organs with maturation in mice. Furthermore, m-reMAIT cells prolonged mouse survival upon tumor inoculation through the NK cell-mediated reinforcement of cytolytic activity. Collectively, the present results demonstrated the utility and role of m-reMAIT cells in tumor immunity, and will contribute to insights into the function of MAIT cells in immunity.


16
induced a similar set of tyrosine-phosphorylated proteins, they differed in their ability to 1 produce cytokines and chemokines ( Figures 1F-H and 1J). Therefore, 5-OP-RU and 2 mMR1-tet may have activated different signals, leading to differences in the pattern of 3 serine/threonine phosphorylation or dephosphorylation via TCR, and ultima tely a 4 different profile of cytokine and chemokine production (Gaud et al, 2018, Nausch & 5 Cerwenka, 2008. In this respect, the signaling pathway(s) responsible for the 6 production of IL-2, IL-4, MIP1α (CCL3), and MIP1β (CCL4) may be common between 7 5-OP-RU and mMR1-tet. 8 While 5-OP-RU induced protein tyrosine phosphorylation, the identity of these 9 proteins remains unclear, except for LAT ( Figures 1G-H). Since LAT forms a hub that 10 interacts with many signaling molecules, such as transmembrane receptors, 11 phosphatases, kinases, guanine nucleotide exchange factors, GTP-activating proteins, 12 ion channels, transporters, and ubiquitin ligases, further studies will provide insights 13 into the proteins responsible for MAIT-TCR signaling and reveal differences in 14 signaling between conventional T cells and MAIT cells (Malissen et al, 2014). 15 It is important to note that the challenge with mMR1-tet induced the activation 16 of m-reMAIT cells and production of cytokines and chemokines. This result suggests 17 that MAIT cells prepared from animals, including humans and mice, are unintentionally 18 stimulated during experiments. Therefore, the interpretation of data requires caution 1 ( Figures 1E-F and 1J). 2 Although a previous study suggested that MAIT cells comprise MAIT1 and 3 MAIT17 subsets characterized by the Th-1 and Th-17 transcriptome, respectively, the 4 present results revealed that m-reMAIT cells produced Th-1, Th-2, and Th-17 cytokines 5 (Salou et al, 2019). This may reflect a developmental stage at which m-reMAIT cells 6 are en route to functional maturation and m-reMAIT cells generated in vitro represent 7 the most immature stage. Therefore, it is tempting to postulate that the interaction 8 between nascent MAIT cells and double-positive thymocytes favors the differentiation 9 of naïve MAIT cells into MAIT1 and MAIT17 in vivo. 10 The results on the adoptive transfer of m-reMAIT cells indicated that nascent 11 MAIT cells egressed from the thymus migrate into any organs concomitant with the up-12 regulation of CD44, a cell adhesion receptor, memory, and/or activation marker ( Figure  13 2B) (Benlagha et al, 2005). It is important to note that the expression of IL-7Rα in m-14 reMAIT cells became quasi-equivalent to that in endogenous cells in the intestines 15 ( Figure 2C). Since IL-7 from intestinal epithelial cells (IEC) play a pivotal role in the 16 homeostasis of IEC, the up-regulation of IL-7Rα in m-reMAIT cells may mirror an 17 intrinsic and primary role in this homeostasis and/or in the integrity of the intestinal 18 epithelial barrier (Rouxel et al, 2017, Shalapour et al, 2012. Alternatively, the up-1 regulation of IL-7Rα may reflect the role of MAIT cells in IL-17A and Th1 cytokine 2 production, which is dependent on IL-7 (Tang et al, 2013). Together with the time-3 dependent up-regulation of CD44, increases in IL-18Rα and CXCR6 in different organs 4 indicated that m-reMAIT cells mature in the host, which has important implications for 5 cell therapy. Fasl, and Tnfsf10 in m-reMAIT cells suggested that cytolytic activity comprised 17 exocytosis-mediated killing with GrzB, a serine protease, and FasL-and Tnfsf10-18 mediated killing through the activation of caspases (Rossin et al, 2019, Trapani & 1 Smyth, 2002. The present result showing that NK cell depletion compromised the 2 survival of mice that had received m-reMAIT cells upon a tumor inoculation is 3 consistent with the above findings, further highlighting the role of NK cells in boosting 4 the tumor inhibitory activity of m-reMAIT cells in vivo. While NK cells recognize 5 NKG2D ligands and/or the lack of MHC I on tumor cells and eradicate them, the 6 molecular mechanisms by which m-reMAIT cells recognize and eliminate tumor cells 7 with the aid of NK cells warrants further study (Nausch & Cerwenka, 2008). 8 In contrast to the present results, MAIT cells have been shown to induce the 9 metastasis of melanoma B16F10 (Yan et al, 2020). Although the reason for this 10 discrepancy currently remains unclear, differences in the dynamics of MR1 shuttling 11 between ER and the plasma membrane upon a 5-OP-RU challenge or putative ligand(s) 12 present in the tumor milieu (TME) may be a key feature. Since LLC did not strongly 13 up-regulate MR1 on the cell surface, in contrast to B16F10, when stimulated ( Figure  14 4A), 5-OP-RU may have bolstered B16F10 tumorigenicity through MR1. The MR1-15 elicited signal may suppress the function of NK cells for B16F10, whereas signal(s) 16 from putative ligand(s) present in TME via MR1 did not, thereby enhancing or 17 preserving NK cell function. Further studies are needed to elucidate the underlying 18 20 mechanisms, which will provide a novel avenue for tumor immunotherapy with iPSC-1

derived MAIT cells. 2
Although we demonstrated that m-reMAIT cells exhibited cytolytic activity 3 against LLC together with NK cells in vitro , difficulties are 4 associated with demonstrating that m-reMAIT cells isolated from mice that had 5 received the cells exhibit similar lytic activity ex vivo. This is due to the low recovery of 6 m-reMAIT cells and to the compulsory use of mMR1-tet in cell preparations, which 7 may interfere with cytolytic activity. Furthermore, it currently remains unclear whether 8 these results are applicable to other cancer cells, such as B16F10, if the use of 5-OP-RU 9 enhances or inhibits the cytolytic activity of m-reMAIT cells, and also whether MAIT-10 like cells prepared from human MAIT-iPSCs exhibit similar activity. 11 In summary, the adoptive transfer model with m-reMAIT cells used in the 12 present study opens a new avenue for exploiting the function of MAIT cells and 13 providing insights into their interaction(s) with immune cells in immunity. 14 15

Materials and Methods 16
Mice 17 21 All mouse experiments were performed with approval from the Institutional Animal 1 Care and Use Committee of Dokkyo Medical University. C57BL/6N mice were 2 purchased from CLEA Japan (Tokyo, Japan). C57BL/6 (Ly5.1) mice were obtained 3 from the RIKEN Bioresource Center and bred in-house. All mice were housed in the 4 Animal Research Center, Dokkyo Medical University, under specific pathogen-free 5 conditions with controlled lighting and temperature with food and water provided ad 6 libitum. Male C57BL/6N mice aged 6 weeks were used to isolate MAIT cells for the 7 generation of iPSCs. Male and female mice aged between 8 to 12 weeks were used in 8 adoptive transfer experiments and in tumor experiments. 9 Cell lines 10 OP9/DLL1 cells were maintained in αMEM supplemented with 20% FBS. The mouse 11 cancer cell lines B16F10, CH27, CH27/mMR1, EL4, and LLC were cultured in DMEM 12 supplemented with 10% FBS, while RL-♂1, WT3, WT3/mMR1 and Yac-1 were 13 cultured in RPMI1640 supplemented with 10% FBS at 37ºC in 5% CO 2 . 14

Antibodies 15
The antibodies used in the present study are listed in the supplemetal table 2. 16 Oligonucleotides 17

22
The oligonucleotides used in the present study are summarized in the oligonucleotide 1 Table. 2

Preparation of mouse immune cells 3
Spleen, thymus, and lymph nodes: Tissues were prepared by mashing through a 40-4 µm mesh cell strainer with a syringe plunger. Single cells were suspended in RPMI1640 5 supplemented with 10% FBS, 10 mM HEPES pH 7.0, 0.1 mM 2-mercaptoethanol, and 6 100 IU/ml of penicillin/streptomycin (referred cR10) and spun down at 400×g for 4 7 min. To lyse red blood cells, the cell pellet was suspended in autoclaved ice-cold MilliQ 8 water for 15 sec and immediately neutralized with 4% FBS in 2× PBS. After 9 centrifugation, cells were resuspended in cR10. 10 Lungs and liver: Single-cell suspensions from the lungs and liver were prepared using 11 enzymatic digestion. Briefly, tissues were placed into a GentleMACS C-tube (Miltenyi 12 Biotec) and cut into approximately 5-mm 3 pieces. Four milliliters of tissue digestion 13 solution (90 U/ml collagenase Yakult, 275 U/ml collagenase type II, 145 PU/ml Dispase 14 II, and 4% BSA in HBSS) was added per tissue and tissues were homogenized using the 15 GentleMACS dissociator (Miltenyi Biotech) with the following program: 16 m_lung_01_02 for the lungs and m_liver_03_01 for the liver. Suspensions were then 17 incubated at 37ºC for 30 min under gentle rotation, followed by dissociation with 18 m_lung_02_01 for the lungs and m_liver_04_01 for the liver, and subjected to 1 discontinuous density centrifugation over layers of 40% and 60% Percoll at 400×g for 2 20 min. Cells were recovered from the 40%-60% Percoll interface, washed with PBS, 3 and then suspended in cR10. 4 Intestines: The intestines were longitudinally incised and their contents were 5 thoroughly washed out three times by vigorous shaking. Tissue dissected into 1-cm 6 pieces were placed into a 50-ml conical tube and washed vigorously by shaking three 7 times with PBS. After discarding the supernatant, tissues were treated with 40 ml of 8 intraepithelial lymphocyte (IEL)-washing solution (HBSS containing 1 mM DTT, 5 9 mM EDTA, and 1% BSA) by shaking vigorously at 37ºC for 30 min under gentle 10 rotation. After washing three times with MACS buffer (PBS containing 2 mM EDTA 11 and 0.5 % BSA), tissue was washed again with 40 ml HBSS. After removal of the 12 supernatant, tissues were placed into a GentleMACS C-tube (Miltenyi Biotec), cut into 13 small pieces with scissors, and 4 ml of the Tissue digestion solution (see above) was 14 added. Tissues were processed using the following program: m_brain_01_02 and 15 digested at 37ºC for 30 min under gentle rotation followed by dissociation with the 16 program: m_intestine_01_01. Cell suspensions were subjected to Percoll discontinuous 17 density centrifugation and isolated cells were resuspended in cR10. 18

PCR detection of the rearranged configuration of TCR loci in MAIT-iPSCs 13
To confirm that MAIT-iPSCs stemmed from MAIT cells, PCR detecting the rearranged 14 configuration of TRAV specific for MAIT cells was performed with the primer sets 15 ADV19 and AJ33. Genomic DNA was prepared from MAIT-iPSCs with NaOH. To 16 detect TRBV in MAIT-iPSCs, total RNA was prepared from m-reMAIT cells (days 20-17 28 of differentiation, varying according to the clones), using the RNeasy Mini kit 18 (Qiagen). cDNA was synthesized with the First strand cDNA synthesis kit (Thermo 1 Fisher Scientific) and subjected to PCR with the primer sets TRBV13 and TRBC-Rev, 2 and TRBV9 and TRBC-Rev followed by DNA sequencing (Fasmac). The usage of 3 TRBV-D-J was analyzed by IGBLAST (NCBI, NIH 4 https://www.ncbi.nlm.nih.gov/igblast/). 5

Flow cytometry 9
Cells were stained with the antibodies listed in the antibody table. 7-AAD or Zombie 10 Violet (Biolegend) was used to discriminate between live/dead cells. To stain the 11 transcription factors PLZF and RORγt, cells stained with the åsurface markers were 12 fixed and permeabilized with the Transcription factor buffer set (BD Biosciences), and 13 then stained with transcription factor antibodies. In CD107a staining, a fluorochrome-14 labeled CD107a antibody was added to the culture prior to the assay. Cells were 15 analyzed with the MACSQuant cell analyzer (3 lasers, 10 parameters, Miltenyi Biotech) 16 or the AttuneNxT acoustic focusing cytometer (4 lasers, 14 parameters, Thermo-Fisher 17 Scientific). Data were processed using FlowJo software (ver.9.9 or 10.7, BD 18 Biosciences). Cell sorting was performed using a FACSJazz cell sorter (2 lasers, 8 1 parameters, BD Biosciences). were measured with LegendPlex at the indicated ratio. In CD69 and CD107a expression 1 analyses, the above column-isolated NK cells (1 × 10 5 cells) and m-reMAIT cells (1 × 2 10 5 cells, purity> 95%) were cultured individually or cocultured in the absence or 3 presence of the same number of Yac-1 at 37ºC in 5% CO 2 for 18 h. The percentages of 4 CD69 + cells and CD69 + CD107a + cells among NK cells and m-reMAIT cells were then 5 measured using the MACSQuant flow cytometer. 6

Semi-quantitative PCR 7
NK cells cocultured with m-reMAIT cells (NK cells/m-reMAIT cells ratio =1) were 8 sort-purified as NK1.1 + CD49b + NK cells and TCRβ + mMR1-tet + m-reMAIT cells, 9 respectively. RNA from each subpopulation was extracted with the RNeasy Mini kit 10 (Qiagen), and cDNA was synthesized with the first-strand cDNA synthesis kit (Thermo 11 Fisher Scientific) and then subjected to semi-quantitative PCR. PCR was performed 12 with CYBR Green reagent (Nippon Genetics) using the following program: at 95ºC for 13 5 min (90ºC for 15 sec, 60ºC for 60 sec) × 50 cycles (Light/Cycler Nano, Roche). The 14 primer sets used in this study were described in the supplemental Table 1. Zombie + cells among target cells in the absence of effector cells)/100. 10

Quantification and statistical analysis 11
Statistical analyses were conducted using Prism 9 for macOS (GraphPad). The Log-12 rank test was used for survival analyses between the two indicated groups. A two-way 13 ANOVA was employed to assess the significance of differences among the various 14

effector cells (NK cell, m-reMAIT cells, and NK cells plus m-reMAIT cells) in lysis 15
assays. The percentage of CD69 + cells among m-reMAIT cells cultured with CH27/mMR1, 1 challenged as in B in the presence of the anti-MR1 antibody (¢) or the isotype control 2 antibody (¡) 3

D. 5-OP-RU dose-dependent activation 4
The percentage of m-reMAIT cells expressing CD69 upon a challenge with various 5 concentrations of 5-OP-RU. Representative data from two independent experiments are 6 shown. 7

E. mMR1-tet dose-dependent activation 8
The percentage of m-reMAIT cells expressing CD69 upon a challenge with the 9 indicated amounts of mMR1-tet. Representative data from two independent experiments 10 are shown. 11

S1B. MAIT-iPSC detection by PCR 12
The results of PCR using the primer set shown in A. Bands corresponding to rearranged as BamHI sites are shown together with the primer sequences for probe synthesis. 3

S1F. Expansion of MAIT-iPSCs during differentiation 13
Cell numbers at the indicated differentiation stage of L7-1 from iPSCs are plotted.