Pinpointing the tumor-specific T-cells via TCR clusters

Adoptive T cell transfer (ACT) is a promising approach to cancer immunotherapy, but its efficiency fundamentally depends on the extent of tumor-specific T-cell enrichment within the graft. This can be estimated via activation with identifiable neoantigens, tumor-associated antigens (TAAs), or living or lyzed tumor cells, but these approaches remain laborious, time-consuming, and functionally limited, hampering clinical development of ACT. Here, we demonstrate that homology cluster analysis of T cell receptor (TCR) repertoires efficiently identifies tumor-reactive TCRs allowing to: 1) detect their presence within the pool of tumor-infiltrating lymphocytes (TILs); 2) optimize TIL culturing conditions, with IL-2low/IL-21/anti-PD-1 combination showing increased efficiency; 3) investigate surface marker-based enrichment for tumor-targeting T cells in freshly-isolated TILs (enrichment confirmed for CD4+ and CD8+ PD-1+/CD39+ subsets), or re-stimulated TILs (informs on enrichment in 4-1BB-sorted cells). We believe that this approach to the rapid assessment of tumor-specific TCR enrichment should accelerate T cell therapy development.


Introduction
Tumors develop diverse mechanisms of immune evasion, including the generation of hypoxic conditions 1 , inflammation 2 , establishment of an immunosuppressive microenvironment [3][4][5][6] , downregulation of antigen presentation [7][8][9] , promotion of regulatory T cell (T reg ) infiltration 10 and outgrowth 11 , and induction of T cell dysfunction 12 . The infusion of large numbers of expanded autologous tumor-reactive T cells-typically after the implementation of lymphodepleting regimens-represents a powerful therapeutic option that may override these immunosuppressive mechanisms. Clinical protocols for such adoptive cell transfer (ACT) therapeutic strategies 13 , inspired by the pioneering work of Steven Rosenberg's group [14][15][16] , are now being actively developed and used to treat patients [17][18][19][20][21] .
There is accordingly great demand for methods for the enrichment of autologous tumor antigenspecific T cells for use in ACT protocols. Current techniques rely on the identification of patientspecific peptide neoantigens, which are then used for the functional characterization and selection of cultured tumor-infiltrating lymphocytes (TILs) 22,23 . Alternatively, since the identification of unique neoantigens is costly, time-consuming, and functionally limited in terms of the spectra of identifiable antigens, cultured autologous tumor tissue can be used as a source of antigen-specific stimulus 24 . Certain cell-surface markers such as PD-1, CD39, CD69, CD103, or CD137 may also help to delineate T cell subpopulations (typically CD8 + ) that are enriched for clonally-expanded tumorreactive T cells [25][26][27] , therefore culturing selected TIL subsets, such as PD-1 + T cells 28 , is a feasible option. In all of these scenarios, however, there is the need for a robust method that enables estimation of enrichment of the transplanted cells with tumor-specific T cells based on the T cell receptor (TCR) repertoire without prior knowledge of TCR specificities.
The ongoing adaptive immune response is often driven by groups of T cell clones with highly homologous TCR sequences that recognize the same epitopes [29][30][31][32] . If one can properly account for the most public TCR clusters, which arise regardless of their binding properties by virtue of their high probability of being generated in course of V(D)J recombination 33,34 , identification of such convergent TCR clusters becomes a highly efficient approach to capture clonotypes involved in the current immune response 35 .
Here, we have employed cluster analysis to identify groups of TCR clonotypes involved in the antitumor immune response. We demonstrate that this approach successfully pinpoints known TAAspecific TCRs among TIL repertoires in HLA-A*02 melanoma patients. Furthermore, we find that the number of cluster-related clonotypes and the proportion of the bulk TIL repertoire that they occupy grows significantly after anti-PD-1 immunotherapy. We next investigate the TCR content in sorted CD4 + and CD8 + CD39 + PD1 + TILs, and show that these subsets are prominently enriched for TCR clusters, a substantial fraction of which consists of tumor-specific TCRs. These results provide a rationale for focusing on CD39 + PD1 + TILs in adoptive cancer therapy. Finally, we show that repertoire analysis facilitates optimization of TIL culturing conditions, and allows estimation of the extent of tumor-specific T cell enrichment in cultured donor cells and sorted TAA-activated T-cells. Altogether, our findings strongly support the use of cluster TCR analysis as a powerful tool with practical applications in clinical ACT.

TIL clusters include TAA-specific TCRs and grow after immunotherapy
We first analyzed published TIL TCR beta chain (TCRꞵ) repertoires obtained before and after anti-PD-1 immunotherapy for two cohorts comprising 21 and 8 patients with cutaneous melanoma 36,37 . Using the ALICE alrorithm 29 , we were able to identify clusters of convergent TCRꞵ clonotypes in all patients. The number of cluster-related clonotypes significantly increased after therapy in both cohorts (p = 0.019 and 0.038, respectively; Fig. 1a,b), which might reflect treatment-dependent expansion of convergent antigen-specific TILs. A VDJdb database 38,39 search identified clusterrelated clonotypes that are highly similar or identical to known TCRꞵ variants specific to the HLA-A*02-restricted melanoma-associated antigens Melan-A (Melan-A aa26-35 -ELAGIGILTV) and NY-ESO-1 (NY-ESO-1 aa157-165 -SLLMWITQC) in 20% and 50% of patients from the Ref. 36 and Ref. 37 cohorts, respectively. Cluster-related clonotypes were enriched for TAA-specific TCRꞵ variants compared to the bulk TCR repertoire, indicating that cluster analysis can capture clonotypes involved in an ongoing anti-tumor immune response (Fig. 1c,d). Fig. 1d shows identified TCR clusters for one of the patients after immunotherapy. Summary for the count and size of TCR clusters before and after immunotherapy for each patient is shown on  Notably, the HLA genotypes of the patients for these two cohorts were unknown. Since VDJdb currently includes limited diversity of HLA contexts, we believe that a much higher proportion of cluster-related clonotypes will be assigned to TAA-specificities with the accumulation of TCR specificity data from more diverse HLA contexts.
For TILs obtained from HLA-A*02-positive patient mp26, with BRAF wt melanoma, a VDJdb search identified three TCR clusters that included A*02-Melan-A aa26-35 -specific clonotypes (Fig. 2h). Both the Melan-A-specific clonotypes and the TCR clusters comprising such clonotypes were prominently enriched within the CD8 + DP subset, compared to the bulk TCR repertoire obtained from fresh-frozen tumor tissue (FFT) and to the CD8 + non-DP subset (Fig. 2i,j). Similar results were obtained for another BRAF wt HLA-A*02-positive patient, pt41 (Fig. S1b,c). We concluded that the CD39 + PD1 + fraction is enriched for large and convergent T cell clones that are involved in an ongoing tumor-specific immune response, a substantial portion of which are detectable via cluster TCR analysis.
To functionally confirm our findings, we used the CD137 (4-1BB) upregulation assay. Sorted DP and non-DP TIL subsets from patient mp26 were cultured for two weeks and stimulated with autologous monocyte-derived dendritic cells loaded with a TAA peptide mix. CD8 + CD137 high subsets were subsequently quantified with flow cytometry and sorted for TCRꞵ library preparation. As shown in Figure 2k, the proportion of CD137 high cells was higher in cultured DP cells, but we found no difference between T cells stimulated by TAA-loaded or control dendritic cells. At the same time, TCRꞵ repertoire analysis revealed that the CD137 high fraction of TAA-stimulated CD8 + DP-but not non-DP or control DP-cells was enriched with known Melan-A-specific clonotypes (Fig. 2l). These clonotypes included a TAA-reactive TCRꞵ variant, CSARVGNQPQHF-TRBV20-TRBJ1-5, which was previously detected in cluster analysis of non-cultured CD8 + DP cells, and variant CASSGGMGQPQHF-TRBV19-TRBJ1-5, which is homologous to another cluster. TAAspecific clonotypes cumulatively occupied ~13% of the CD137 high fraction of the cultured and TAAactivated DP TILs. These results underscore the importance of TCR repertoire analysis of responding cells, even in the apparent absence of a quantifiable difference between antigen and control conditions, and demonstrate that CD137 marker analysis on its own is insufficiently informative.

TCR cluster analysis facilitates optimization of TIL culturing conditions
We next investigated the effect of four distinct TIL culture conditions on the expansion of tumorreactive T cells: IL-2 high , IL-2 low /IL-21, IL-2 low /IL-21/anti-PD-1, and IL-2 low /IL-21/anti-PD-1/IFNɣ (Fig.  S3a), where the anti-PD-1 agent employed was nivolumab and the concentration of IL-2 in the low and high conditions was 100 IU/mL and 3,000 IU/mL, respectively. For each condition, we analyzed TCRꞵ repertoires of TILs independently cultured from 12 tumor fragments collected from patient mp26. TCR clusters identified from all samples were joined and visualized along with the noncultured DP and non-DP CD8 + subsets described above (Fig. 3a). As a readout, we used the following: i) normalized count of cluster-related clonotypes (Fig. 3b), ii) cumulative proportion of the repertoire occupied by Melan-A-specific TCRꞵ clusters (clusters predominantly comprising VDJdb-defined Melan-A-specific clonotypes) (Fig.  3c), and iii) the number of differentially-expanded clonotypes compared to pan-activating IL-2 high culture conditions, and the proportion of such expanded clonotypes that were also initially detected among CD8 + DP TILs (Fig. 3d).
The IL-2 low /IL-21/anti-PD-1 combination yielded the greatest number of cluster-related clonotypes and the highest cumulative proportion of Melan-A-specific clusters out of all culture conditions (Fig.  3b,c). Addition of IFN did not further enhance the expansion of potentially tumor-specific clones. We utilized edgeR 46 software, which was initially designed for differential gene expression analysis, to identify clonotypes that were significantly expanded in tumor fragment cultures in the presence of IL-21 and IL-2 low (either with or without nivolumab and IFNɣ) compared to classical pan-activating IL-2 high culture conditions. The IL-2 low /IL-21/anti-PD-1 combination yielded the highest number of reproducibly expanded clonotypes, 60% of which were detected among freshly-sorted CD8 + DP TILs (Fig. 3d, Fig. S3b). The overall count of CD8 + DP-matched clonotypes was also highest for this combination (Fig. 3e). These results show the positive influence of PD-1 inhibition on the proliferative potential of CD8 + CD39 + PD-1 + T cells. We also noted a slight increase in the number of CD8 + non-DP matched clonotypes which were expanded in these same conditions-10 clonotypes, compared to three clonotypes in IL-2 low /IL-21 without nivolumab (Fig. 3d)-which can be explained by nivolumab-dependent expansion of PD-1 + CD39 -T cells.

IL-2 low /IL-21/anti-PD-1 culture conditions stimulate TIL proliferation without T reg expansion
T reg s are known to hamper anti-cancer immune responses elicited by ex vivo expanded TILs 47 . This immunosuppression may be overcome by CD25 + T cell depletion 48 . Alternatively, T reg expansion may be suppressed by IL-21 49 . To more comprehensively characterize the distinct TIL culture conditions described above, we compared their capacity to support or suppress T reg expansion and overall T cell expansion. Predictably, IL-2 alone yielded the highest proportion of T reg (CD4 + CD25 high CD127 low FoxP3 high cells) cells. In line with prior findings 50 , IL-2 low created preferential conditions for selective T reg expansion, while the addition of IL-21 drastically reduced the proportion of T reg s amongst CD4 + TILs (Fig. 3f, Fig.S3с,d).
IL-21 was previously demonstrated to promote the expansion of T cells with a memory phenotype 51 .
We also noted a significant increase in the number of memory phenotype 52 CD127 high cells among CD4 + lymphocytes in IL-21 + conditions regardless of the presence or absence of nivolumab (Fig.  S3e). Remarkably, CD8 + TILs displayed significant CD127 high enrichment only upon simultaneous introduction of IL-21 and nivolumab (Fig. S3f). These results reveal the synergistic action of IL-21 and disruption of PD-1-dependent signaling by nivolumab on expansion of CD8 + memory T cells.
Regarding overall proliferative potential, the highest T cell count was evident for IL-2 high conditions, although we also observed roughly comparable numbers in the IL-2 low cultures (Fig. 3g). The presence of IL-21 stifled IL-2-dependent TIL proliferation, as previously reported for human CD4 + T cells 53 . On the other hand, IL-21 was favorable for expansion of CD8 + TILs, whereas IL-2 alone favored CD4 + cell growth (Fig. 3h).

Discussion
Here we demonstrate that rational TCR clustering can be used to identify tumor-specific T cell clones and estimate their relative enrichment among TILs. Starting with published TCR repertoire data from melanoma tumors, we identified clusters of convergent TCR clonotypes, which increased in numbers and total frequency after anti-PD-1 immunotherapy. We next found significant enrichment of convergent TCR clusters in PD1 + CD39 + subpopulations of both CD4 + and CD8 + TILs, which were previously shown to be enriched in terms of tumor reactivity 41,43,54 . These findings are further supported by the data we obtained for cases where we know the HLA context, as well as some TCR clonotypes of interest and their cognate antigens. A VDJdb database search successfully identified TCR variants specific to TAA antigens in HLA-A*02-positive patients, where approximately half of the clusters could be matched to known Melan-A-specific sequences. The cumulative frequency of such Melan-A specific clusters within the CD8 + DP population was only slightly lower than the total frequency of all identified Melan-A specific TCRs, indicating that our approach was able to identify most of the high-frequency tumor-specific clonotypes. It should be noted that TCR clusters are an essential feature of a convergent immune response that involves several homologous clonotypes. For those cases where a single T cell clone dominates in response to a particular antigen and/or homologous "neighbours" are absent due to the very low probability of convergent TCR generation 30 , cluster analysis may miss some of the tumor-reactive TCRs. Nevertheless, our results demonstrate the overall power of this approach, which is applicable in those situations where specific antigens are unknown.
We describe one potential implementation of our approach by using it to optimize conditions for ex vivo TIL expansion. In particular, TILs cultured in IL-21 + conditions demonstrated the highest number of cluster-related TCRs, which is indicative of a more prominent influence of antigen-driven TCR selection. We speculate that IL-21 exerts its influence in our culture system both at the antigen-presentation stage (as MHCI-and, to a lesser extent, MHC II-restricted antigen presentation by tumor cells occurs while tumor fragments are cultured ex vivo) 55 and at the expansion stage, where IL-21 selectively promotes the expansion of non-regulatory T cell populations. The consequences of in vitro PD-1 blockade are less well understood. The in vitro application of an anti-PD1 antibody was previously shown to induce IFNɣ and TNFα secretion by PD1-expressing CD8 + TILs 56 . In our work, we found a greater increase in abundance of clusterrelated and tumor-specific TCRs for cells cultured in IL-21 + /anti-PD-1 conditions compared to IL-21 + alone.
We expect that this straightforward TCR repertoire-based approach to estimation of TIL enrichment with tumor-reactive clones will accelerate clinical development of adoptive T cell therapy. At the R&D stage, this approach could greatly facilitate the selection of optimal TIL subsets and optimization of TIL culturing conditions and downstream enrichment procedures. And within clinical pipelines, such analyses should make it possible to estimate tumor-reactive TIL abundance at the level of individual tumor samples, both before and after culturing and/or enrichment. As the field of T cell therapy advances towards the ability to determine the clonal specificities and phenotypes that can effectively fight each individual patient's tumor, methods for grouping convergent TCRs that respond to the same tumor antigens will become an essential part of the toolbox for rationallydesigned T cell therapy.

Patients
All clinical samples were acquired from the N.N. Blokhin National Medical Research Center of Oncology in accordance with protocol MoleMed-0921, approved by the ethical committee on 30 Jan 2020. All patients involved in the study were diagnosed with metastatic melanoma and signed an informed consent prior to collection of their biomaterial. Most experiments were performed on freshly resected metastatic lymph nodes obtained during surgery. Genotyping for BRAF V600E was performed at the N.N. Blokhin National Medical Research Center of Oncology, Pathomorphology Department. From each patient enrolled in this study, we also obtained 20-30 mL of peripheral blood before the surgery. Patient information is provided in Table S1.

Brief TIL culture
Freshly-resected tumor specimens were dissected into fragments measuring 1-3 mm in each dimension. Several fragments were frozen in liquid nitrogen for further cDNA library preparation and HLA-typing. Individual fragments were seeded into the wells of a 24-well tissue culture plate with 2 mL of complete T cell cultivation medium (CM) supplemented with 5% heat-inactivated human AB serum (PanBiotech, Germany) and 1,000 IU/mL IL-2 (Ronkoleukine, BiotechSpb, Russia) . CM consisted of RPMI-1640 (PanEco, Russia), 25 µM/mL HEPES, pH 7.2 (PanEco, Russia), 100 IU/mL penicillin, 100 µg/mL streptomycin, 10 µg/mL gentamicin, 1x non-essential amino acids mix, 1x GlutaMAX , β-mercaptoethanol (0.55 µM) (all from Gibco, Thermo Fisher Scientific, US) and 110 μg/m sodium pyruvate. On the fourth day of cultivation, TILs were harvested, filtered through a 70-μm mesh, stained with fluorophore-labeled antibodies, and FACS sorted. For the generation of TCR repertoire libraries, T cells were sorted directly into the RLT cell lysis buffer (QIAGEN, Netherlands) and stored at -80ºC until used for RNA isolation. For functional assays, TILs were sorted into 1.5 mL Eppendorf tubes with 0.5 mL RPMI-1640. Live-sorted T cells were seeded into wells of 96-well cell culture plate at 10 6 cells/mL in CM and cultured for at least five days. Here, CM was supplemented with 1,000 IU/mL IL-2, 50 ng/mL IL-21 (SCI-STORE, Russia), 20 μg/mL nivolumab (Bristol-Myers Squibb, USA), and 10% autologous patient-derived serum. Half of the media was replaced three times a week. One day before the functional assays, all the media was replaced with fresh CM without interleukins or nivolumab.

Expansion of sorted T-cells
FACS-sorted T-cells were expanded using non-specific TCR-dependent stimulation with anti-CD3/CD28 Dynabeads (Thermofisher Scientific, USA). Beads were added into the cultivation medium on the next day after seeding the cells, with 2 µl of bead solution per 10 5 cells. Sorted TILs were expanded for 2 weeks. Beads were magnetically removed upon achieving desired cell numbers. Before co-cultivation experiments, cells were allowed to "rest" for one day in interleukinfree CM supplemented with 10% autologous patient-derived serum (i.e., "resting" medium).

PBMC isolation
Peripheral blood mononuclear cells (PBMCs) were derived from patients' blood samples using gradient centrifugation with Ficoll-Paque Plus (GE Healthcare). Briefly, 18 mL of whole blood were diluted to 50 mL volume with sterile 1x PBS. Diluted blood was layered over the Ficoll-Paque solution in 50 mL SepMate tubes (StemCell Technologies, USA), with 25 mL of diluted blood per tube. SepMate tubes were centrifuged for 20 min at 1200 x g with brake off. Afterward, buffy coats were collected and washed two times with 50 mL of sterile PBS.

Monocyte-derived dendritic cells cultivation
Autologous dendritic cells were generated as described in Ref. 57. Briefly, CD14 + cells (monocytes) were isolated from patients' PBMCs with a magnetic enrichment procedure using anti-CD14 MicroBeads (Miltenyi Biotec, Germany). Then, monocytes were seeded into the wells of 24-well tissue culture plates at 5 x 10 5 cells/well. X-Vivo-15 medium (Lonza, Switzerland) with 400 U/mL IL-4 and 800 U/mL GM-CSF was used for the differentiation of monocytes. On the fourth day of cultivation, the medium was renewed, and dendritic cells were loaded with the mix of melanoma TAA peptides (PepTivator Melan-A/MART-1, gp100/Pmel, and MAGE-A3 human; Miltenyi Biotec) at a concentration of 600 nM each. The next day, loaded DCs were matured using 1 µg/mL PGE, 10 ng/mL IL-1β, and 25 ng/mL TNF-α. Following 24 hours of maturation, DCs were harvested and used for co-cultivation with T cells.

CD137 antigen-specific activation assay
FACS-sorted PD1 + CD39 + (DP) and non-DP populations after expansion and two days "rest" without IL-2 were co-cultured with antigen-loaded autologous DCs at a ratio of 10:1 T cells:DCs. The cocultivation medium consisted of 1:1 CM plus AIM-V serum-free medium (Gibco, USA) supplemented with 50 ng/mL IL-21. Both CD137 high and CD137 low T-cells were lysed with RLT buffer for further RNA isolation and TCR library construction. The frequency of CD137 high cells was measured for both CD4 + and CD8 + TILs. Antigen-specific activation was measured as a ratio of CD137 high T cell frequencies in TILs co-cultured with antigen-loaded vs unloaded DCs.

Nuclear staining for FoxP3
First Finally, the fraction of T reg s (CD4 + CD25 high CD127 low FoxP3 high ) out of total CD4 + T cells was determined.

HLA-typing
For patients mp39, mp41, mp42, and mp44, we have only checked for the presence of HLA-A*02. Aliquots of PBMCs or TILs from these patients were stained with anti-HLA-A*02-PE (BD7.2) (BD Biosciences, USA) antibody and analyzed with a Navios flow cytometer. Other patients (Table S1) were HLA-typed using NGS at the Center for Precision Genome Editing and Genetic Technologies for Biomedicine (Moscow).

RNA isolation and TCR library preparation
RNA from fresh-frozen tumor fragments was isolated using TRIzol reagent(Invitrogen). RNA from RLT-lysed cells was extracted using the RNeasy mini kit (Qiagen) according to the manufacturer's protocol. RNA concentration was measured with the Qubit RNA HS Assay Kit (Thermo Fisher Scientific, US). No more than 500 ng of total RNA was used for cDNA synthesis.
cDNA libraries were generated using the Human RNA TCR Multiplex kit (MiLaboratories), according to the manufacturer's protocol. We aimed to achieve coverage of 20 paired-end reads per cell for sorted and cultivated TIL populations, and approximately 2 x 10 6 reads per tumor sample fragment. Sequencing was performed using the Illumina NextSeq platform (2 х 150 bp read length).

Analysis of TCR repertoires sequencing data
TCRseq data was analyzed with MiXCR software (MiLaboratories) in order to extract TCRβ CDR3 clonotypes. VDJtools software was used for processing of MiXCR output TCR repertoire data, calculation of TCR repertoire diversity, and pre-processing of TCR repertoires (i.e., pooling of joined TCR repertoires and down-sampling of TCR repertoires) 59 .

Cluster analysis of TIL TCRβ repertoires
This analysis was performed using the ALICE algorithm 29 . We selected clonotypes with read count >1 to exclude undercorrected erroneous TCR variants that could potentially create false neighbors of abundant clonotypes, distorting the ALICE hit identification process. P gen of amino acid sequences was estimated using Monte Carlo simulation. For each VJ pair, 5 million TCR sequences were simulated, and 20 iterations of the algorithm were performed. Only clonotypes with Benjamini-Hochberg (BH)-adjusted p < 0.001 were selected as significant ALICE hits. As the number of ALICE hits strongly depends on the initial variability of the TCR repertoire, we normalized the number of hits between samples based on the number of the top-frequency input clonotypes.
To visualize the resultant clusters of convergently selected TCRs we used the igraph function 60 . This function utilizes the de Bruijn graph method to calculate the distance between amino acid sequences of CDR3 regions. It creates a graph file in GML format where each node represents an individual TCR clonotype and the distance between nodes is proportional to the difference between CDR3 amino acid sequences. Edges connected nodes representing TCR clonotypes with Hamming distance ≤2. Graphs were visualized using Gephi 0.9.2 network analysis platform 61 . The size of the node represents the frequency of the corresponding TCR clonotype. Upon construction of composite graphs including clonotypes from multiple TCR repertoires, identical clonotypes from different TCR repertoires were displayed by separate nodes.

Matching cluster-related TCR clonotypes to VDJdb
We annotated TCR repertoire data using the VDJdb database of T cell receptors with known specificity 38 . We assumed that TCRs of interest had the same specificity as TCRs from the database if: i) CDR3 regions of compared TCRs differed no more than by one central amino acid substitution, ii) substituted amino acids belonged to the same group based on their R properties (polar, aliphatic, aromatic, positively/negatively charged), and iii) HLA-restriction of TCR clonotypes from the database matched one of the patient's HLA alleles, if known.
TCR clusters consisting predominantly of TAA-specific clonotypes, but not clonotypes of other specificities, were considered TAA-specific as a whole. VDJdb-unmatched members of the TAAspecific clusters were deemed to possess the same specificity as the whole cluster based on structural similarity to VDJdb-matched clonotypes and were included in the subsequent analysis.

Analysis of differentially expanded clonotypes with edgeR software
We used a statistical approach implemented in the edgeR package to identify TCRβ clonotypes that were significantly expanded in bulk TILs of patients mp26 and mp34 as described in Ref. 60. Previously, this approach was used to track vaccine-responding clonotype expansion in time. We implemented edgeR for comparison of TCR repertoires of TILs cultivated in experimental conditions 2-4 (described above) and TILs expanded in IL-2 high conditions. Six and four biological replicate samples of each cultivation setting were used for the analysis of TCR repertoires from mp26 and mp34, respectively. TCR clonotypes were deemed expanded if the false discovery rate adjusted p value was <0.01 and the log 2 fold-change was >1.

Statistical analysis
Statistical analysis was performed using Graph Pad Prism 8.0 (GraphPad Software Inc., USA). All data was reported as mean ± SD. The Shapiro-Wilk test was used for normality estimation in all cases. Names of statistical tests and numbers of biological replicas in each comparison group are provided in the figure legends.